MUSLIMS, MONGOLS AND CRUSADERS

The period from about 1100 to 1350 in the Middle East was marked by
continued interaction between the local Muslim rulers and two groups of
non-Muslim invaders: the Frankish crusaders from Western Europe and the
Mongols from Northeastern Asia. In deflecting the threat those invaders
presented, a major role was played by the Mamluk state which arose in
Egypt and Syria in 1250. The Bulletin of the School of Oriental and African
Studies has, from 1917 onwards, published a variety of articles pertaining to
the history of this period by leading historians of the region, and this
volume reprints some of the more important and interesting of them for the
convenience of students and scholars. In making the selection the interests
of those who are not specialists in the history of the Middle East and may
not know the languages of the region have been taken into account. This
volume will be of interest to historians of medieval Europe who are
concerned with the Crusades as well to those interested in the Middle East
and the Mongols. The papers here reprinted include discussion on Arabic
and other sources for the period (including the controversial Marco Polo
and his 'Travels'), innovative studies of military, diplomatic, administrative
and other issues, and wider treatments of such things as the image of
Saladin and historiography on the Mongol Empire. An introduction by the
editor puts the papers in the historical and scholarly context.
G.R.Hawting is Head of the History Department and Professor in the
History of the Near and Middle East at the School of Oriental and African
Studies, University of London. His special interest and most of his
publications relate to the early development of Islam in the Middle East.

MUSLIMS, MONGOLS AND CRUSADERS

An Anthology of Articles Published in The Bulletin
of the School of Oriental and African Studies
Compiled and Introduced by
 G.R.Hawting

LONDON AND NEW YORK

First published 2005

by RoutledgeCurzon
2 Park Square, Milton Park, Abingdon, Oxon, OX14 4RN
Simultaneously published in the USA and Canada
by RoutledgeCurzon
270 Madison Ave., New York, NY 10016
RoutledgeCurzon is an imprint of the Taylor & Francis Group
This edition published in the Taylor & Francis e-Library, 2005.

To purchase your own copy of this or any of Taylor & Francis or Routledge’s collection of thousands
of eBooks please go to www.eBookstore.tandf.co.uk.
Editorial matter © 2005 G.R.Hawting
Contributions © The School of Oriental and African Studies
All rights reserved. No part of this book may be reprinted or
reproduced or utilised in any form or by any electronic, mechanical,
or other means, now known or hereafter invented, including
photocopying and recording, or in any information storage or
retrieval system, without permission in writing from the publishers.
The publisher makes no representation, express or implied, with
regard to the accuracy of the information contained in this book and
cannot accept any legal responsibility or liability for any errors or
omissions that may be made.
British Library Cataloguing in Publication Data
A catalogue record for this book is available from the British Library
Library of Congress Cataloging in Publication Data
A catalog record for this book has been requested
ISBN 0-203-64182-5 Master e-book ISBN
 ISBN 0-203-66665-8 (MP PDA Format)
ISBN 0-7007-1393-X (Print Edition)
Copyright © 2004/2005 . All rights reserved.

Reader's Guide
This ebook has been optimized for PDA.
Tables may have been presented to accommodate this Device's Limitations.
Table content may have been removed due to this Device's Limitations.
Image presentation is limited by this Device's Screen resolution.
All possible language characters have been included within the Font
handling ability of this Device.
 CONTENTS

Introduction
G.R.Hawting
1. A on the destruction of Baghdad by the Mongols
J.de Somogyi
2. Notes on the Arabic materials for the history of the early Crusades
H.A.R.Gibb
3. The influence of Chingiz-Khan's Yasa upon the general organization of
the Mamluk state
A.N.Poliak
4. Studies on the structure of the Mamluk army [I]
D.Ayalon
5. Studies on the structure of the Mamluk army [II]
D.Ayalon
6. Studies on the structure of the Mamluk army [III]
D.Ayalon
7. Saladin and the Assassins
B.Lewis
8. The position and power of the Mamluk sultan
P.M.Holt
9. Cassiodorus and Rashid al-Din on barbarian rule in Italy and Persia
D.O.Morgan
10. The treaties of the early Mamluk sultans with the Frankish states
P.M.Holt
11. The Mongol Empire: a review article
D.O.Morgan
12. Saladin and his admirers: a biographical reassessment
P.M.Holt
13. Some observations on the 'Abbasid caliphate of Cairo
P.M.Holt
14. The 'Great Yasa of Chingiz Khan' and Mongol law in the Ilkhanate
D.O.Morgan
15. The Ilkhan embassies to Qalawun: two contemporary accounts
P.M.Holt
16. The Crusades of 1239–41 and their aftermath
P.Jackson
17. The Secret History of the Mongols: some fresh revelations
T.H.Barrett
18. Ghazan, Islam and Mongol tradition: a view from the Mamluk sultanate
R.Amitai-Preiss
19. Marco Polo and his 'Travels'
P.Jackson
INDEX
INTRODUCTION1
 The period and its historiography

One notable feature of the period in the history of the Middle East with
which the articles collected in this volume are concerned is the migration
into the region, or its invasion, by peoples—Turks, Franks and Mongols—
originating outside it. These groups ruled over and settled among an already
existing population which was diverse but by this time probably
predominantly Muslim, and Arabic or Persian speaking. It may be useful to
begin with a summary account before proceeding to reflect on the way in
which academic scholarship on the period has developed.2
During the ninth and tenth centuries, Turks became the most important
element in the armies of the eastern Islamic world. Some achieved positions
of power to the extent that they were able to establish and maintain dynastic
states independent of the authority of the caliphs in Baghdad. Frequently,
the Baghdad caliphs were little more than figureheads, with real power in
the hands of Turkish generals and commanders.
The Turks had first come into the Islamic world as individual slaves,
selected for training as soldiers and conversion to Islam, to be followed by
manumission and service in the armies of the caliph or other powerful
figures. Islam had not yet spread among the Turkish tribes of Central Asia
and they were still legitimate targets for Muslim raiders and slave traders,
their reputation as fighters high. The institution of the slave soldier
(ghulam, mamluk) was to remain a characteristic feature of Muslim armies
in pre-modern times, and so long as there were still non-Muslim Turkish
populations they were to remain a favoured source of military manpower.
During the second half of the tenth century, however, Islam began to spread
among some groups of Turkish tribes in and beyond what were then the
eastern border regions of the Islamic world. The reasons for, and the nature
of this process of Islamization, are to some extent obscure but, given the
earlier willingness of groups of Turks to adopt religions associated with the
settled cultures with which they were in contact, not too difficult to
envisage. One of the groups which became Muslim at this time was that of
the Oghuz tribes, and in the first half of the eleventh century they began to
migrate west under the leadership of the Seljuk family.
By 1055, they had won control over most of Persia and Iraq, including
Baghdad, the residence of the caliphs who were the nominal leaders of
Sunni Islam. Subsequently, groups of Turks continued to move west into
Syria, over most of which they established control in the 1070s, and
northwest into Asia Minor. There, their victory at Manzikert near Lake Van
in 1071 over the Byzantine army led by the emperor Romanus Diogenes
opened up Asia Minor for the first time to Islamization and Turkification.
That region, before Manzikert largely Christian and Greek speaking, is now
the heart of the country known as Turkey.
These free Turkish migrants into the Islamic Middle East came as Muslims,
and their progress from Central Asia to Baghdad and beyond was relatively
slow. This enabled them, or at least their leaders, to adapt and assimilate to
the predominantly Perso-Islamic culture of the eastern regions of the
Muslim world but without losing their own Turkish identity. The Christian
Franks who came as crusaders from western Europe and who arrived in
Syria at the end of the eleventh century, in contrast, had no interest in
assimilation, although inevitably during the nearly 200 years of their
presence in the region they had to adapt to many of the features of the local
way of life.
Coming from the north in 1099, the Franks quickly established four Latin
Christian polities in the region of Syria: the kingdom of Jerusalem, the
principality of Antioch, and the two counties of Edessa and Tripoli. To a
large extent this intial success must be explained by the disunity and
disharmony of the several Muslim powers who, between them, struggled
for hegemony in Syria and more widely in the Middle East.
The Isma'ili caliphate, ruling in Cairo since 969, partly as a result of
the coming of the Turks, had been forced to abandon its aim of establishing
its authority throughout the Muslim world. The Seljuks proclaimed
themselves champions of the Sunni form of Islam, and their arrival in Syria
ended expansion there. Nevertheless, the still had supporters
in the area and an interest in maintaining a presence at least in Palestine and
Ghaza in order to protect Egypt, the heart of their Shiite state.
In central and northern Syria, geographical fragmentation and the divisions
inside the Seljuk empire following the death of the sultan Malik Shah in
1092 allowed the development of a political and religious patchwork and a
shifting pattern of alliances and hostilities, sometimes involving also the
Frankish states. One other notable element here was that of the Nizari
Isma'ilis, known to the Franks and to the west as the Assassins. In the
1090s, they had separated from those Isma'ilis who continued to recognize
the leadership of the caliphs in Cairo, and they had established
politically active groups in various parts of Persia and Syria.
Slowly during the twelfth century these divisions among the Muslims were
diminished, although never completely overcome. The process culminated
in the rule of Saladin (1169–93), who was able to make himself master of
Egypt and of extensive territories in Syria. Saladin's most famous
achievement was the wresting back of Jerusalem from the Franks following
the battle of in 1187; earlier (1171) he had ended the line of
caliphs in Cairo. His seizure of Jerusalem did not, however, bring the Latin
kingdom of Jerusalem to an end, and at the time of Saladin's death it still
hung on in Acre and Tyre. It was to remain a significant player in the region
for another century. Two of the originally three crusading states to the north,
those in Tripoli and Antioch, also still survived, but the county of Edessa
had been eliminated as early as 1144.
Saladin was succeeded in his Egyptian and Syrian territories by various
members of the family to which he belonged, forming a loose, often
mutually antagonistic, dynasty—the Ayyubids whose name is derived from
that of Saladin's father (see the dynastic table in Peter Jackson's article in
this volume on the crusades of 1239–41). These successors of Saladin were
faced with the incursions of several more crusading expeditions from
Europe and with the continuing presence of the crusading states in Syria.
By this time, however, the crusading movement itself had become more
liable to rivalries and competing aims between the various participants, and
the crusaders in Syria did not pose the threat to the Muslims that they had
before Saladin's time.
During the first half of the thirteenth century, a more immediate and greater
danger to the Muslim rulers came from the rise of the power of the Mongols
and the disturbances that it triggered in east and Central Asia. Between
1218 and 1221, some of the great Muslim towns of Central Asia were
devastated by the raids of Genghis Khan's followers, but the Middle East
itself was spared for some years as the Mongols consolidated their
conquests in China and expanded in the southern Russian steppe. Genghis
himself died in 1227. In the 1240s, renewed Mongol raids led to the
submission of a number of Muslim territories, including the Seljuk sultanate
of Rum which had developed in Asia Minor after Manzikert, and to the
influx into Syria and Palestine of significant numbers of refugees from
Khwarazm, a Muslim state to the southwest of the Aral Sea. This new
ingredient in the situation in Syria was to be an important element in the
relationship between Ayyubids and Franks in the region.
In the 1250s, the Mongol invasion of Persia began and by 1258, after the
destruction of the strongholds in Persia of the Nizari Isma'ilis (the
Assassins), Baghdad was taken. Whereas the Muslim Seljuk Turks, two
centuries earlier, had posed as champions of the Baghdad caliphs and saw
the advantages of maintaining the caliphate and exercising control over it,
the non-Muslim Mongols had no use for it. They killed the reigning caliph,
and effectively ended an institution that dated back to the
earliest Islamic times.
By this time the caliphate had lost its importance in the life of Islam. From
the middle of the ninth century its religious authority had been usurped by
the religious scholars (the ulema), and the caliph in Baghdad had become
little more than a symbol of the unity of Sunni Islam, even though some of
the later caliphs were able to take advantage of temporarily favourable
conditions to revive their political authority and, to some extent, the
prestige of their office. At the level of theory, the writings of several Sunni
Muslim scholars reflect the diminished status of the caliphate and it may be
argued that by 1258 the institution was no longer of fundamental
importance for Muslim religious and political life. Although the idea of the
caliphate survived, and although later individuals claimed to be, and to
some extent received recognition as, caliphs, the institution henceforth had
only an attenuated existence. (See the article in this volume by P.M.Holt on
the Abbasid caliphate in Cairo.)
Mongol ambitions in the Middle East were not limited to Persia and Iraq
and by 1260 they had entered northern Syria and were pushing south into
Palestine. But there they experienced their first significant defeat—at the
battle of 'Ayn Jalut in that year. That proved to be a decisive reverse.
Excluded now from lands west of Iraq, the Mongol ruler Hülegü (d. 1265)
established his capital in Azerbaijan and he and his Mongol descendants as
rulers of Iraq and Persia became known as the dynasty of Il-Khans (1256–
1335).3
The Mongols had been defeated at 'Ayn Jalut by the recently installed
Mamluk sultanate of Cairo. Like earlier Muslim rulers, the Ayyubid
successors of Saladin, especially al-Malik sultan of Egypt 1240–49,
had built up their armies by the extensive use of slave soldiers. At this time
the majority were slaves acquired from among the Turks of the Kipchak
steppe of southern Russia.
died during the crusade to Egypt led by Louis IX in 1249 and shortly
afterwards his son and successor was killed by some of the Turkish soldiers.
By 1257, following a period of puppet rulers and violent intrigues in which
one of concubines, Shajar al-Durr, played a leading part, power in
Egypt was seized by the soldier a Khwarazmian. at first
claimed to act on behalf of the son and successor of a previous sultan but
soon felt secure enough to send the nominal ruler into early retirement. This
was a pattern of succession that became frequent during the following more
than two and half centuries when Egypt and Syria were ruled by a series of
sultans who had originally been recruited as slave soldiers.
It was who commanded the army that defeated the Mongols at 'Ayn
Jalut but soon afterwards he was killed by Baybars, a Turkish Mamluk who
had played an important part in the battle and had been involved earlier in
the murder of successor. Baybars (1260–77) was the first important
sultan in the line of Mamluk rulers of Egypt and Syria (1250–1517).
Baybars' immediate priority was to secure his territories against further
attacks from the Il-Khanid Mongols in Persia. To this end he began to
extend his control over Syria by moving against those elements there—
Franks, Armenians, Nizaris and survivors of the Ayyubid dynasty—who
had been or were potentially allies of the Mongols, and to forge links with
another Mongol power in the region, the Golden Horde.
The Horde controlled the Russian steppe and exercised suzerainty over the
Russian princes to the north. Its rulers were enemies of the Mongol Il-
Khans of Iran. Its Khan Berke (1257–67) had accepted Islam although
generally the Islamization of the Horde proceeded at a slower pace than that
of the Il-Khanid Mongols. From the point of view of Baybars, probably the
most important factor was that the Horde controlled the region which was
the source of the supply of slaves for the Mamluk territories.
Following the killing of the caliph in Baghdad a member of the 'Abbasid
family escaped and made his way to Cairo where, in 1261, he was installed
as caliph by Baybars. This successor caliphate in Cairo continued until the
Ottoman occupation of Egypt in 1517. At a later date the Ottoman sultans
in Istanbul claimed that the last caliph of Cairo (who had been taken to
Istanbul in 1517) had transferred the office into their hands. The caliphate
was proclaimed formally abolished by Kemal Ataturk when he established
the state of Turkey in 1922.
After the death of Baybars in 1277, the Mamluk sultanate eventually passed
to a figure of similar stature, Qalawun (1279–90), who continued the
policies of his predecessor. Largely as a by-product of the need for security
against the Mongols in Iran, Qalawun was concerned to bring Syria more
firmly under his control. His great achievement was the taking of the county
of Tripoli in 1289. He died shortly afterwards while preparing for an attack
upon Acre. That prize finally fell to Qalawun's son and successor, al-Malik
al-Ashraf Khalil b.Qalawun (1290–93).
A convenient conclusion to the period we are concerned with in this volume
is signalled by the destruction of the last remaining Frankish possessions in
Syria during the 1290s and by the conversion to Sunni Islam of the Mongol
Il-Khanid ruler of Persia, Ghazan Khan (1295–1304). The conversion of the
Il-Khans to Islam symbolizes the gradual acculturation of the Mongol
conquerors of Iran and was a significant stage in the disappearance of the
Mongols as a distinct ethnic group in the Middle East. Thirty years or so
after the death of Ghazan, the Il-Khanate itself had disappeared, and when
—some twenty years later—a claimant to the Il-Khanid throne appeared, he
bore the traditional Persian name of Anushirwan.
Nevertheless, the acceptance of Islam by Ghazan Khan did not end the
hostility between the Mongols in Iran and the Mamluks in Egypt and Syria.
Ghazan was even able to take possession of Damascus for a short time in
1299–1300. A second attempted invasion in 1303 was less successful and
Syria thenceforth remained under Mamluk control until the coming of the
Turko-Mongol Timur (Tamerlane) a century or so later.
Clearly, a selection of articles concerned with aspects of Mamluk, Mongol
and Crusading history, taken from the volumes of the Bulletin of the School
of Oriental and African Studies published in the latter two-thirds of the
twentieth century, will not be completely representative of the development
of studies in these fields. There are scholars—Bertold Spuler, Claude Cahen
and Ulrich Haarmann are just three who come to mind—who made major
contributions but did not publish (at least on these subjects) in the Bulletin.4
Equally there are areas, for example, artistic, intellectual and religious
history in the period of this volume, which have received significant
attention elsewhere but are largely unrepresented here. The majority of the
contributions in this volume are concerned with political, diplomatic,
administrative and institutional history, and the methods and evidence used
are also predominantly traditional—that is, the critical analysis of texts,
predominantly the chronicles and biographies written by contemporaries.
Most of the research presented here is based on the appearance or easier
availability of new textual sources, or by fresh exploitation of already
known texts.
Given the tight focus of most academic articles, it is not really possible on
the basis merely of those in this volume to deduce how far and in what
direction views about the significance of the period as a whole may have
developed. It may be of value, however, to consider some views about the
period generally and to see how far the articles collected here indicate
developing understandings of it.
Two of the contributors to this volume, in particular, have been concerned
to view the period within the longue durée of the Islamic Middle East.
H.A.R.Gibb, in an influential article assaying an interpretation of medieval
Islamic history, which appeared in the first issue of the Journal of World
History, characterized this period as one in which the "orthodox institution"
of Islam (he was referring to the body of Sunni scholars concerned above
all with the interpretation and implementation of the Law) lost its ability to
integrate and provide leadership in society, a role that increasingly came to
be taken over by Sufi movements. This development he associated with an
urban and economic decline consequent upon the expansion of nomadism,
and the coming of the Turks and Mongols was part of that phenomenon.
The growing influence of Sufism, the major forms of which, according to
Gibb, had grave intellectual consequences for Islam ("it drew intellectual
energies off into subjective and antirational speculation") was
hastened on by the destruction of the still vigorous centers of Islamic culture in north Persia during
the Mongol invasion of 1220, and the Mongol occupation of all western Asia (except Syria) after the
capture of Baghdad in 1258. The orthodox institution was eclipsed under the rule of heathen princes,
and though it gradually revived in the following century its social and political foundations were too
weak to allow it to recover its former influence.5
Gibb, therefore, tends to view the period unfavourably in cornparison with
the earlier one in which, following the Arab conquest of the Middle East,
Islamic society and culture had formed. Gibb seems to value the Arab
contribution over that of other ethnic and linguistic groups, urban society
over that of agriculturalists and pastoralists, and Sunni "orthodoxy" over
other forms of Islam in the religious sphere.
The view of the period, especially between the coming of the crusades and
the Mongol conquests, as one of stagnation or even decline has been shared
by some other scholars. Historians who focus on political developments
have often emphasized the political fragmentation of the Islamic world
from the ninth century onwards—the period in which the importance of the
Turks first became notable. Those who focus on intellectual and religious
matters sometimes point to a perceived lack of innnovation in the tradition
of Islamic philosophy (Averroes, d. 1197, and Maimonides, d. 1204, are
often seen as the last significant scholars in that tradition); and to an alleged
increasing narrowness and intolerance in the field of religion (Ibn
Taymiyya, d. 1328, is sometimes seen as the key figure here; al-Ghazali,
d.1111,although generally admired for the sophistication of his theology, is
sometimes regarded less favourably for the impact of his attack on the
philosophical tradition).
Bernard Lewis shares the view that it was the already apparent internal
weaknesses of Islamic state and society in the Middle East that made it
susceptible to attack from outside. Those weaknesses were political,
economic and cultural. Internally, the chief threat was posed by Isma'ili
Shiism, externally it came from the invaders from the east—the Turks and
subsequently the Mongols.
However, for Lewis, the Turks were not simply one aspect of the spread of
nomadism at the expense of urban society. He emphasizes the identification
of the Turks with Islam and the role they played in defending the Sunni
form of the religion against the Isma'ilis and the Crusaders, which allowed
it to recover and develop its strength.
He sees Sufism, not in the rather negative way of Gibb, but as strengthening
Sunnism by providing it with an emotional content in addition to its legal
and dogmatic ingredients. The development of the Sufi tradition was not at
the expense of the "religious institution" but accompanied the rise of that
institution to a position of authority which was in fact greater than that it
had enjoyed in early Islam. Under the Turks the religious institution came
to be incorporated into the structure of political authority, a process which
reached its peak in the "gunpowder empires" which emerged in the period
following the withering away of the Mongol presence.
The destruction of the caliphate by the Mongols, according to this view,
was no more than the laying to rest of something already moribund, and
Lewis also plays down the long-term destructive effects of the Mongol
devastation. Without minimizing the depredations of the Mongols in the
lands that they conquered, he stresses too the potential for speedy recovery
and the fact that the Mongols never reached Egypt which, by this time, had
replaced Baghdad as the centre of Islamic civilization in the Middle East.
He regards, however, the coming of the Mongols as deleterious for the
long-term political and agricultural development of Iraq.
The chief legacy of the period other than the resurgence of Sunnism,
according to Lewis, was the division of the Islamic Middle East into two
cultural zones—that dominated by Persian and Turkish to the north and east
and that by Arabic to the south and west—although to some extent this
division was countered by the revival of Sunnism.6
Lewis's presentation, therefore, seems less negative than that of Gibb. For
Lewis, the period seems generally to be one of recovery and development
(although not necessarily constant and consistent), preparing the way for the
flourishing of Islam in the period of the "gunpowder empires" (the
Ottomans, Safavids and Moghuls) before the impact of modernity on the
Middle East from around the end of the eighteenth century.
Our other contributors have been less ready to paint with such broad
strokes, perhaps reflecting a feeling shared by many contemporary
historians that their primary task is to understand rather than to evaluate and
that the broad delineation of periods of history may be at the expense of
their complexity and diversity.
It is perhaps possible to deduce, nevertheless, that few of them would share
the negative views of Gibb. Assuming that scholars do not devote
themselves to the study of topics that they do not consider important and
attractive, the recent revival of interest in the Mongols (illustrated here
especially by the contributions of Amitai, Jackson and Morgan) and in the
Mamluks (Amitai, Ayalon and Holt here, Haarmann and others elsewhere)
points to a more positive evaluation of these groups compared with that of
some earlier writers (and of much popular writing).
At the beginning of his article on "The position and power of the Mamluk
Sultan" (chapter 8, below), Holt quotes the view of Prideaux in 1722 that
"they scarce did anything worthy to be recorded in History". His own article
then goes on to illustrate the power, wealth and sophistication of the
Mamluk state and to argue that it was more stable and united than the
previous Ayyubid regime. The achievement of the Mamluks in their
struggles against the Mongols and Franks has long been recognized, but
Holt's assessment of them seems to go beyond that.7
To say that there has been a clear and marked change in the scholarly
treatment of the period is perhaps to put the case too strongly, but any
attempt to discuss it generally today would take account of, for example,
increased awareness of the richness and diversity of the Shiite (Isma'ili and
Twelver) tradition of Islam, the importance of Sufism in spreading Islam in
Asia and Africa, and the importance too of the Mongol conquests for the
spread of Islam in central and eastern Asia. The Arab, Sunni and urban bias
which characterized the interpretation of Gibb has declined in scholarship,
if not necessarily in popular accounts and in the use of the Crusaders and
the Mongols in political propaganda.
On two particular points modern scholarship has taken issue with views
expressed in popular and politically inspired presentations.
The Crusades have naturally elicited comparisons in some circles with what
is portrayed as the threat from "western" imperialism and Zionism. One
result has been to magnify the importance of the Crusades as a factor in the
history of the Middle East and that has been a consequence too of the
modern idealization of Saladin and the wish on the part of several modern
Arab leaders to be seen as inheritors of his mantle.
In contrast, modern scholarship on the Crusades has tended to diminish
their importance in the development of Islamic and Middle Eastern history
generally, in contrast to their importance in European history. It has been
pointed out that following the liquidation of the last Frankish states in the
1290s the historical memory of them faded in the Middle East until the
nineteenth century when it revived as European historical works came to be
translated into Arabic. The Arabic expression for "crusades"
did not exist prior to the nineteenth century; until that point the
common name for the crusaders in Arabic was the ethnic designation
"Franks", and no historical works on the Crusades appeared in Arabic
before the late nineteenth century.8 The significance of the Crusades in the
history of the Middle East may be debated but modern scholarhip is
inclined to limit it.9
The second point concerns the long-term consequences of the destruction
caused by the Mongol raids and conquests in the first half of the thirteenth
century. That the Mongols were destructive of agriculture and towns is not
questioned, but there has been a reaction against the view, sometimes
expressed in popular and politically motivated writings, that the Mongols
are to be blamed for many of the ills which have been claimed to afflict the
Middle East even into the twentieth century—agricultural and economic
"backwardness", the failure to develop representative political institutions,
the dominance of society by the military, etc.10
The period treated in this volume continues to inspire research and interest,
and it is hoped that the greater accessibility of these articles as a result of
their republication here will contribute to that.
 Notes on the authors and articles

Some of the authors of articles collected in this volume contributed to

Middle Eastern and Islamic studies, and other fields, in areas unconnected
with that which is of interest here. In the following notes I have focused
mainly on their work associated with the Crusades, the Mamluks and the
Mongols.
Joseph de Somogyi (1899–1976) is perhaps best known for his association
with Ignaz Goldziher (d. 1921), often regarded as the real founder of the
academic study of Islam. De Somogyi seems to have been the last of
Goldziher's students to survive, was one of the editors of the Goldziher
Memorial Volumes, was responsible for the publication of his teacher's
Gesammelte Schriften, and for the composition of several obituaries and
tributes. He himself was Professor of Islamic Languages in the School of
Oriental Commercial Studies at Budapest until after the Second World War
when he moved to Harvard. His A Short History of Oriental Trade was
published in 1968.11
The text of the mourning the fall of Baghdad to the Mongols, which
de Somogyi edited and translated in his article (1933) reprinted here, is
taken from the Ta'rikh al-Islam of Shams al-Din al-Dhahabi (d. 748/1347),
a work which was only in manuscript at the time de Somogyi wrote. The
full text is now available in the edition of 'Umar 'Abd al-Salam Tadmuri (52
vols, Beirut: Dar al-Kitab al-'Arabi, 1987–99) and the poem is printed in the
volume for the years AH 641–650, at pp. 37–39.12 Tadmuri indicates that
the poem is also given in the Nujum al-zahira of Ibn Taghribirdi. A
marthiya (in Arabic) on the same subject by the famous Persian poet al-
Sa'di (d. 691/1292) may also be noted.13
For discussion of the Mongol conquest of Baghdad, the precise dating of
which seems to be open to question, see J.A.Boyle, "The death of the last
'Abbasid caliph of Baghdad: a contemporary Muslim account", Journal of
Semitic Studies 6 (1961), and, by the same author, The Cambridge History
of Iran, v, Cambridge 1968, 345–350.
Avraham N.Poliak (b.Kiev, 1910) was Research Professor and Head of the
Department of General and Jewish History at the University of Tel Aviv and
is probably best known for his work (in Hebrew) on the Jewish Khazars14
and for that pertaining to the agrarian and social history of the Middle East
in the late mediaeval period. His Feudalism in Egypt, Syria, Palestine and
the Lebanon15 is the best known of his works in English. With reference to
the subject of the present volume, Poliak's "La caractère colonial de l'état
mamelouk dans ses rapports avec la Horde d'Or", Revue des Études
Islamiques 9 (1935):231–245, may be noted.16
The main thesis of Poliak's "The influence of Chingiz-Khan's Yasa upon the
general organization of the Mamluk State" (1940–42) is that the law and
administration of the Mamluk sultanate, especially as they concerned the
Mamluk element of the population, were consciously derived from and
modelled upon the "great yasa" or code of law promulgated by Genghiz
Khan. The article was published under wartime conditions, and appended at
the end are a number of critical comments contributed by the great
Iranologist, Vladimir Minorsky (d. 1966), the first of which questions the
validity of Poliak's thesis.
Subsequent work, notably that of the late David Ayalon (see especially the
reference at note 1 of Morgan's article on the Yasa below), and the article of
David Morgan which is included in this volume, has served to weaken
much of Poliak's argument and even to question the nature and existence of
the alleged yasa of Genghiz Khan. Nevertheless, his willingness to
contemplate the Mamluk sultanate in a wider comparative context is
refreshing, and some of his suggestions and details seem deserving of
continuing reflection.
H.A.R. (Sir Hamilton) Gibb (d. 1971) was perhaps the foremost British
Arabist and historian of the Middle East of his generation.17 Professor of
Arabic at the universities of London, Oxford and Harvard in succession, his
writings deal with many aspects of Arabic literature, Middle Eastern history
and Islam. He contributed four chapters to the first volume, and one to the
second, of the monumental A History of the Crusades, published by the
University of Wisconsin Press under the general editorship of
K.M.Setton.18 The career of Saladin in particular was a theme to which he
devoted several articles and a posthumous book.
In addition to the chapter "The rise of Saladin, 1169–1189" in the first
volume of Setton's History, mention may be made here of his "The armies
of Saladin", Cahiers d'Histoire égyptienne 3 (1951):304–320; "The
achievement of Saladin", Bulletin of the John Ryland's Library 35 (1952):
44–60; (both were reprinted in Stanford J.Shaw and William R.Polk (eds),
Studies on the Civilization of Islam by Hamilton A.R.Gibb, London 1962);
and The Life of Saladin from the works of 'Imad ad-Din and Baha' ad-Din,
Oxford 1973.
Gibb's "Notes on the Arabic materials for the history of the early Crusades"
(1933–35) is a pioneering analysis of the mediaeval Arabic source material
for the early period of the crusades. It was occasioned by the appearance of
the first of the three volumes of René Grousset's Histoires des Croisades et
du Royaume Franc de Jérusalem (Paris 1934–36).
Gibb argues that on a number of points the understanding of Grousset and
others may be criticized because they interpret the early period of the
crusades in the light of conditions and attitudes that only developed later.
This anachronistic view of the early period is the result of excessive
reliance on sources dating from the time of Saladin and later. To some
extent that was inevitable since it was the later sources, in particular the
universal history (al-Kamil fi'l-ta'rikh) of Ibn al-Athir (d. 630/1233), which
were the first to become known to western scholars. In the second part of
his article, Gibb shows that the understanding of the events of the earlier
period which Ibn al-Athir presents must be corrected by comparison with
earlier material, more contemporary with the events treated. Particularly
important is the Dhayl Ta'rikh Dimashq (Continuation of the History of
Damascus) by Ibn al-Qalanisi (d. 555/1160). Gibb himself had made some
of the important passages of Ibn al-Qalanisi's work available in an English
translation a few years earlier (The Damascus Chronicle of the Crusades,
London 1932).
Bernard Lewis was, at the time of his retirement (1986), Cleveland E.
Dodge Professor of Near Eastern Studies at Princeton University, and
currently holds that title as Emeritus Professor. Before his move to
Princeton in 1974 he was Professor of the History of the Near and Middle
East in the University of London, and he is a former editor of BSOAS. His
many writings range widely over the pre-modern and modern history of the
Middle East, but his earliest interest was in the mediaeval period and
especially in the history of the Isma'ili branch of Shiite Islam, and its
subgroup, the Nizari Assassins.
His revised University of London PhD thesis was published as The Origins
of Ismailism: a study of the background of the Fatimid caliphate
(Cambridge 1940, reprinted New York 1975), and his The Assassins. A
radical sect in Islam, London 1967. He also contributed the chapter on "The
Isma'ilites and the Assassins" to vol. I of Setton's A History of the Crusades.
Lewis's "Saladin and the Assassins" (1953) may be understood as an
exercise in source criticism similar to that in Gibb's article, as much as an
exploration of its avowed theme. Lewis suggests that Saladin's own
presentation of himself as the champion of "orthodox" Sunni Islam against
the Shiite "heretics", an image which seems to be corroborated by many of
the sources, may be called into question. By examining the range of
available sources, including the semi-legendary historical work of the
Isma'ili Abu Firas, Lewis is able to present Saladin's relationship with the
Assassins as more pragmatic than one would guess from the rather
idealizing accounts produced by his own officials and biographers.
David Ayalon, who is represented in this collection by three articles on the
structure of the Mamluk army (1953 and 1954), devoted the majority of his
work to the study of military slavery in Islam and put the study of the
Mamluk state on a new footing. His academic career was associated with
the Hebrew University of Jerusalem where, at the time of his death in 1998,
he was Emeritus Professor of Islamic History.
For a useful summary of his work, see Reuven Amitai, "The rise and fall of
the Mamluk institution: a summary of David Ayalon's works", in M.Sharon
(ed.), Studies in Islamic History and Civilization in Honour of Professor
David Ayalon, Jerusalem and Leiden 1986, 19–30; for a brief obituary and
complete bibliography, see idem, "David Ayalon, 1914–1998", in Mamluk
Studies Review 3 (1999), 1 ff. Most of his articles are now available in four
volumes of the Variorum Collected Studies Series: Studies on the Mamluks
of Egypt, 1250–1517, London 1977; The Mamluk Military Society, London
1979; Outsiders in the Lands of Islam: Mamluks, Mongols and Eunuchs,
London 1988; and Islam and the Abode of War, Aldershot 1994. His best
known book is Gunpowder and Firearms in the Mamluk Kingdom: A
Challenge to a Mediaeval Society, London 1956.
Ayalon's three articles here on the structure of the Mamluk army represent a
major item in his scholarly work and a continuing substantial introduction
to the structure and terminology of the ruling institution of the Mamluk
state. As has been emphasized by Amitai, Ayalon's scholarly strength was
his concern for the correct understanding of terminology, but he was also
willing to stand back and present the details in a broader context.
P.M.Holt, who is represented in this volume by five articles was, at the time
of his retirement in 1982, Professor of the History of the Near and Middle
East in the University of London. Much of his earlier work had concerned
Egypt and the Sudan in the Ottoman and modern periods, but he went on to
make a number of important contributions to the history of the Middle East
in the time of the crusades and the Mamluks.
He is the editor of a collection of seminar papers, The Eastern
Mediterranean Lands in the Period of the Crusades, Warminster 1977; he
translated from the Arabic The Memoirs of a Syrian Prince: Abu'l-Fida',
Sultan of Hamah (672–732/1273–1331), Wiesbaden 1983; and is the author
of The Age of the Crusades: the Near East from the Eleventh Century to
1517, London 1986; and Early Mamluk Diplomacy (1260–1290): Treaties
of Baybars and Qalawun with Christian Rulers, Leiden 1995. Among his
articles are: "Qalawun's treaty with Genoa, 1290", Der Islam 57 (1980):
101–108; "Three biographies of Baybars", in D.O.Morgan (ed.),
Medieval Hisorical Writing in the Christian and Islamic Worlds, London
1982; "Literary offerings: a genre of courtly literature", in Thomas Philipp
and Ulrich Haarmann (eds), The Mamluks in Egyptian Politics and Society,
Cambridge 1998, 3–16; and "The last Mamluk Sultan: al-Malik al-Ashraf
Tuman Bay", in Jerusalem Studies in Arabic and Islam 25 (2001):234–246.
His translation from the French of Claude Cahen's The Formation of
Turkey: The Seljukid Sultanate of Rum, 11th to 14th century also appeared
in 2001 (London: Longman).
By the time of Holt's review article, "Saladin and his admirers: a
biographical reassessment" (1983), which was occcasioned by the
appearance of the work of Lyons and Jackson on Saladin,19 the scholarly
reassessment of Saladin had developed considerably. A study by
A.Ehrenkreutz (publication details in Holt's article) had even presented him
as something of a disaster and a villain. Holt underlines the place of Gibb in
a tradition of idealization of Saladin that in the west goes back to Walter
Scott and earlier, while its ultimate roots may be traced to the success of the
propaganda that Saladin and his panegyrists Ibn Shaddad and 'Imad al-Din
created. The author himself shares the middle of the road position
of Jackson and Lyons rather than the more polemical stance of Ehrenkreutz.
The other four of Holt's articles reprinted here are contributions to Mamluk
history (and in one case to Il-Khanid history also).
His "The position and power of the Mamluk Sultan" (1975) discusses such
things as accession ceremonies, titles, symbols of office, the roles of the
Sultan and his relationship with the Mamluk amirs and the representatives
of Islam. Holt's suggestion that rulers of slave origin could nevertheless
adopt some of the trappings of sacral kingship is especially interesting. The
article emphasizes the complexity and evolving character of the Mamluk
sultanate and argues that, in spite of inherent weaknesses, for two and half
centuries it not only survived but exercised political and military power
more effectively than had its Ayyubid predecessor.
Holt's "The treaties of the early Mamluk sultans with the Frankish states"
(1980) discusses aspects of seven treaties (or "truces" as they are regarded
according to Islamic law) between the Mamluk rulers in Cairo and various
Frankish authorities in the kingdom of Jerusalem and the county of
Antioch-Tripoli in the second half of the thirteenth century. These treaties,
and details about their conclusion, have been preserved in Arabic literary
sources for the period. Particularly interesting is the final account of the
abrogation by Qalawun of his treaty with the Latin kingdom in 1290, which
a year later was to lead to the final extinction of the kingdom of Jerusalem.
"Some observations on the 'Abbasid caliphate of Cairo" (1984) discusses
the circumstances and considerations, different in each case, which led
Baybars to install the first two of the caliphs, and examines the largely
powerless position of the caliph in the Mamluk sultanate. Holt quotes the
saying of a Mamluk amir, "For God's sake—nobody takes any notice of the
caliph!" Nevertheless, the sultans found it useful to try to legitimize their
seizure of power by having the caliph play a role in their accession
ceremonies, and it may be added that some of the Delhi sultans in India,
too, thought it politic to emphasize their allegiance to the Cairo caliphs.
Holt demonstrates the evident falsity of the claim that the last Cairo caliph
transferred his position to the Ottoman sultan after the Ottomans had
captured Cairo and taken the caliph to Istanbul.20
Finally, Holt's "The Ilkhan embassies to Qalawun: two
contemporary accounts" (1986) compares two Arabic accounts of
embassies, letters and gifts sent by the first of the Il-Khanid rulers to be
converted to Islam, Tegüder (681–683/1282–1284), to the Mamluk
sultan Qalawun (678–689/1279–1290).21 acceptance of Islam was a
personal conversion which preceded the later "communal conversion" of
the Il-Khanate under Ghazan Khan (694–704/1295–1304), aspects of which
are considered in the article of Reuven Amitai-Preiss reprinted here.
Holt draws attention to some discrepancies between the accounts, which
agree, however, regarding the display put on by the Mamluk ruler in
attempting to impress his Mongol contemporary with his might and
magnificence. Behind this, however, Holt detects the continuing fear and
mistrust of the Mongols among the Mamluks.
Peter Jackson is currently Reader in History at the University of Keele, and
is especially known for his work on Mongol and Persian history in the
period with which this volume is concerned.
As well as being the author of several articles in the Encyclopaedia Iranica,
he is the author of "The accession of Qubilai Qa'an: a re-examination",
Journal of the Anglo-Mongolian Society 2 (1975):1–10; "The dissolution of
the Mongol empire", Central Asiatic Journal 22 (1978):186–244; "Jalal al-
Din, the Mongols, and the Khwarazmian conquest of the Panjab and Sind",
Iran (1990):45–53; "From Ulus to Khanate: the making of the Mongol
states, C.1220–C.1290", in Reuven Amitai-Preiss and David Morgan (eds),
The Mongol Empire and its Legacy, Leiden 1999; "The state of research:
the Mongol empire, 1986–1999", Journal of Medieval History 26
(2000):189–210; and "The fall of the Ghurid dynasty", in Carole
Hillenbrand (ed.), The Sultan's Turret, Leiden 2000. For another
contribution by him on the crusades see "The crisis in the Holy Land in
1260", English Historical Review 95 (1980):481–513. Jackson also edited
volume 6 of the Cambridge History of Iran, and he is the author of a history
of the Delhi Sultanate (Cambridge 1999). He translated and, together with
David Morgan, provided the scholarly commentary for, The Mission of
Friar William of Rubruck: his journey to the court of the Great Khan
Möngke, 1253–55, London 1990.
Jackson's discussion of "The Crusades of 1239–41 and their aftermath"
(1987) is concerned with a relatively neglected period in the history of the
Franks in Syria. He demonstrates both the substantial increase in source
material, western and eastern, since the time of Gibb's paper, and what can
be done by someone able to exploit the full range of sources. He is able to
exploit in particular volume iv/2 of the Arabic History of the Patriarchs of
the Egyptian Church, the edited and translated text of which became
available in 1974, and which has new information on military history in the
early 1240s. That enables Jackson to present a picture which underlines the
complexities in the military and political situation faced by the Franks and
to counter the propaganda directed against Theobald, especially that
emanating from the Emperor Frederick II who was concerned to maintain
the commercial links between Sicily and Egypt.
"Marco Polo and his 'Travels'" (1998) discusses the problems surrounding
the provenance and nature of the work associated with the name of the
famous Venetian traveller. The article is relevant in the context of the
present volume because, while the Travels have sometimes been regarded
as a valuable source for aspects of Asian and Middle Eastern history in the
last decades of the thirteenth century, recent research has tended to be more
sceptical and to cast doubts on the value of the information it contains. In
particular the work of Frances Wood, Did Marco Polo go to China?,
London 1995, has suggested that the Venetian may not have gone further
east than Constantinople or the Black Sea.
Jackson's conclusions are less sceptical: he argues that the book as it exists
(and the manuscript history is very complex and difficult to reconstruct)
should be regarded as an account of the known world rather than a relation
of the journey of Marco Polo, and that, while the stature of Polo is likely to
have been exaggerated, nevertheless some of the material pertaining to
China and the Mongols must be the result of personal observations.
Although his judgements incline towards the positive, he demonstrates the
difficulties of reaching firm conclusions about the value of the Travels as an
historical source.
David Morgan is Professor of History and Religious Studies at the
University of Wisconsin at Madison. He was until 1999 Reader in the
History of the Near and Middle East at SOAS, a member of the Editorial
Board of BSOAS on various occasions, and Editor of the Journal of the
Royal Asiatic Society.
Morgan is the author of The Mongols, Oxford 1986, and Medieval Persia,
1040–1797, London 1988. In addition to his contribution to The Mission of
Friar William of Rubruck (for which see under Jackson, above) and his joint
editing with R.Amitai of The Mongol Empire and its Legacy, Leiden 1999,
he is the editor of Medieval Historical Writing in the Christian and Islamic
Worlds, London 1982 (which includes his article "Persian historians and the
Mongols"). Among his other articles are: "The Mongols in Syria, 1260–
1300", in Peter W.Edbury (ed.), Crusade and Settlement, Cardiff 1985,
231–235; "Mongol or Persian: the government of Il-Khan Iran", Harvard
Middle Eastern and Islamic Review 3 (1996):62–76; "Rashid al-Din and
Ghazan Khan", Bibliothèque Iranienne 45 (1997); and "Reflections on
Mongol communications in the Ilkhanate", in Carole Hillenbrand (ed.), The
Sultan's Turret.
His "Cassiodorus and Rashid al-Din on barbarian rule in Italy and Persia"
(1977) compares and contrasts the personalities, writings, and the historical
times in which they lived, of two "native" administrators who worked on
behalf of the "barbarian" conquerors of societies with long traditions of
culture and bureaucratic government. In spite of some striking similarities,
the article draws attention to the fundamentally different relationship
between the Ostrogoths and the culture of the Roman aristocratic families,
on the one hand, and the Mongols and the Perso-Islamic tradition, on the
other. One difficulty that Morgan recognizes in the article is the question of
the authenticity of the letters attributed to Rashid al-Din which he uses as a
source. Elsewhere he acknowledges the strength of the arguments against
their authenticity mounted by A.H.Morton.22
Five years before the publication of his own book on the subject (The
Mongols, referred to above), Morgan surveyed some other works of
attempted synthesis and popular historical writing on the Mongols in "The
Mongol Empire: a review article" (1981). This remains informative for its
insights into the problems facing scholars working in the field of Mongol
history.
Undoubtedly the most challenging of the three articles of Morgan in this
volume is his "The 'Great Yasa of Chingiz Khan' and Mongol law in the
Ilkhanate" (1986). The notion of the yasa, understood as a written law code
promulgated by Genghiz Khan which served as the unalterable basis of
Mongol law and administration, has been widely accepted in modern
scholarship. Stimulated by a series of articles in Studia Islamica by David
Ayalon,23 Morgan dissects the evidence on which this notion has been
based, suggests possible explanations as to why and how the notion may
have arisen, and calls into question—without completely rejecting—
whether such an institution ever existed.
Given the prominence of the idea in the literature on Mongol history—and,
as is evident from the article of Poliak included here, on related fields—
Morgan's questioning of it marks an important development. For further
discussion of the topic, see I. de Rachewiltz, "Some reflections on Cinggis
Qan's Jasay", East Asian History 6 (1993):91–104, and the remarks of
Reuven Amitai at pp. 3–6 of his article "Ghazan, Islam and Mongol
tradition: a view from the Mamluk sutanate", reprinted in the present
volume. In the light of these and other recent studies, Morgan has looked
again at the question in "The 'Great Yasa of Chingiz Khan' revisited", in
R.Amitai and M.Biran (eds), Nomads and sedentary peoples in the Middle
East and East Asia (forthcoming).
Reuven Amitai24 is a Professor, and currently Head of Department, in the
Department of Islamic and Middle East Studies at the Hebrew University of
Jerusalem. He is the author of several articles on the history of the Mongols
in the Middle East and of Mongols and Mamluks. The Mamluk-Ilkhanid
War 1260–81, Cambridge 1995. He has edited, jointly with David Morgan,
The Mongol Empire and its Legacy, Leiden 1999. For his appreciations of
David Ayalon, see the note under Ayalon above. Among his other articles
are: "Hülegü and the Ayyubid lords of Transjordan", Archivum Eurasiae
Medii Aevi 9 (1995–97); "Sufis and Shamans: some remarks on the
Islamisation of the Mongols in the Il-Khanate", Journal of the Economic
and Social History of the Orient, 1999; "Northern Syria between the
Mongols and the Mamluks: political boundary, military frontier and ethnic
affinities", in Daniel Powers and Naomi Standen (eds), Frontiers in
Question, London 1999, 128–152; "Al-Nuwayri as a historian of the
Mongols", in Hugh Kennedy (ed.), The Historiography of Islamic Egypt (c.
950–1800), Leiden 2001, 23–36; and "The conversion of Tegüdar Ilkhan to
Islam", Jerusalem Studies in Arabic and Islam 25 (2001), 15–43.
His "Ghazan, Islam and Mongol tradition: a view from the Mamluk
sultanate" (1996), uses Arabic sources, some only available in ms., to throw
light on various aspects of the acceptance of Islam by the Il-Khanid Ghazan
in 694/1295, and its consequences. He refers to the Islam held by Ghazan
and those Mongols who followed him into it as "syncretic", and shows that
it existed together with attachment to Mongol tradition and religion, even
though elements of that tradition and religion were in direct conflict with
demands of the Sharia. As Amitai concludes, this is more historically
convincing than to envisage that Islamization meant a decisive break with
the Mongol past.
T.H.Barrett is Professor of East Asian History in the University of London,
and the current Chair of the Editorial Board of BSOAS. He is a specialist on
the history of religion in China and author of several articles and books on
pre-modern China. He has contributed "Qubilai Qa'an and the Historians:
some remarks on the position of the Great Khan in pre-modern Chinese
historiography", to The Mongol Empire and its Legacy edited by Amitai and
Morgan.
The Secret History of the Mongols, with which his short note is concerned,
is in David Morgan's words, "the only substantial surviving Mongol work
about the Mongol Empire, the only direct insight we possess into how the
Mongols viewed things".25 It has survived in a transcription in Chinese
characters (originally it was written in the script of the Turkish Uighurs
which Genghiz Khan had adopted for the writing of Mongolian) and in an
abridged Chinese translation. The circumstances in which the Chinese
transcription and translation were made are obscure, and Barrett's note
draws attention to a piece of evidence which has not previously been taken
into account in discussions of the genesis of the texts.
The new evidence seems to point to the existence of the Chinese translation
at a date slightly earlier than that accepted previously. More generally,
Barrett suggests that Ming historiography has not been sufficiently
investigated regarding the information it may contain of interest to students
of the Mongols.
 Notes

1 I am grateful to Michael Brett, George Lane and David Morgan for help and advice in preparing
this Introduction.
2 Of course, a summary account is not in any sense neutral. The following is based on the accounts
of the period established by scholars (such as Gibb and Lewis), some of whose assumptions and
approaches are discussed later in this Introduction. Some may feel that the emphasis on invasion,
disintegration and destruction now gives a rather limited view of the period.
3 For some remarks on the significance of this dynastic appellation, see R.Amitai-Preiss, "Ghazan,
Islam and Mongol tradition", ch. 18 of the present volume. [7, n. 43 in original pagination]
4 Note, however, Ulrich Haarmann's "Regional sentiment in medieval Islamic Egypt", BSOAS 43
(1980):55–66, which is not included here because it is slightly tangential to the the theme of this
volume.
5 H.A.R.Gibb, "An interpretation of Islamic history", Journal of World History 1 (1953):39–62; cited
here from the reprint in his Studies on the Civilization of Islam, London 1962, 3–32—see especially
27 ff.
6 This summary is based on chapter 5 ("The coming of the steppe peoples") of B.Lewis, The Middle
East. Two thousand years of history from the rise of Christianity to the present day, London 1995.
Lewis has attempted similar broad treatments elsewhere: see, e.g., his The Arabs in History, 1st edn
London 1950, which naturally interprets the period from the point of view of its importance for the
Arabs and consequently emphasizes decline and decay more than mere change.
7 See too Holt's "Conclusion: retrospect and prospect" at the end of his The Age of the Crusades. The
Near East from the Eleventh Century to 1517, London 1986. For the positive evaluation of the
military achievements of the Mamluks in the Islamic historical tradition, see U.Haarmann, "Der
Segen des Sklaventums", in U.Haarmann (ed.), Geschichte der arabischen Welt, Munich 1987, 217 f.
Regarding the Mamluk victory at 'Ayn Jalut, modern scholarship now sometimes emphasizes the
disadvantages of the Mongol position as much as the military prowess of the Mamluks at the battle.
8 This last point may now need to be modified. In 1981, Suhayl Zakkar published a short work,
apparently based on a manuscript found in the Bibliothèque Nationale, devoted to the Crusades and
entitled Al-I'lam wa'l-tabyin fi khuruj al-firanj al-mala'in 'ala diyar al-muslimin ("Information and
explanation regarding the attacks of the accursed Franks on the lands of the Muslims"). In his
introduction Zakkar attributes this work to b.'Ali and dates it to 926/1520. Possible
questions remain, however, about the provenance and dating of the work (Zakkar describes the ms.
but omits to give the catalogue no.), for my knowledge of which I am indebted to P.M.Holt's review
of Carole Hillenbrand's The Crusades (see below) in BSOAS.
9 One issue on which there is general agreement is the importance of the Crusades in the
development of commerce between Europe and the Middle East. For discussion of the importance of
the Crusades in the history of the Middle East, see the article "Crusades" by Cl.Cahen in
Encyclopedia of Islam, 2nd edn, Leiden 1954 ff. and, most recently, Carole Hillenbrand, The
Crusades, Islamic Perspectives, Edinburgh 1999, especially ch. 9.
10 For discussion and references see Bernard Lewis, "The Mongols, the Turks and the Muslim
polity", in Transactions of the Royal Historical Society, 5th series 18 (1968,):49–68 (reprinted in his
Islam in History, new edition, Chicago 1993, 189–207), and D.O.Morgan, The Mongols, Oxford
1986, 73–83.
11 For an obituary and tribute see Studies in Islam 15 (1978):145–147.
12 I am very grateful to Mr Khaled al-Mufti for sending me a copy of this part of the edited text.
13 Kulliyat al-Sa'di, ed.M.'Ali Furughi, reprint Tehran 1363/1984, 766.
14 See now Kevin Allen Brook, The Jews of Khazaria, Northvale NJ 1999.
15 London 1939; the text was reprinted together with appendices consisting of two of Poliak's
articles on the subject dating from 1936 and 1937, Philadelphia 1972.
16 I know of no obituary of Poliak, and have only been able to ascertain his year of birth. I am
grateful to Shani Allouche for help in enquiries about him.
17 Among the obituaries of Gibb those by A.K.S.Lambton (BSOAS 35 (1972): 338–345) and Albert
Hourani (Proceedings of the British Academy 58 (1972): 493–523) are to be noted.
18 London, Madison and Milwaukee: vol. I, The First Hundred Years, Marshall W. Baldwin (ed.),
1955, 2nd edn 1969; vol. II, The Later Crusades, 1189–1311, Robert Lee Wolff and Harry W.Hazard
(eds), 2nd edn 1969.
19 M.C.Lyons and D.E.P.Jackson, Saladin: the politics of the Holy War, Cambridge 1982.
20 For extensive discussion of the caliphate between the fall of Baghdad and its reinstitution in Cairo
by Baybars, see now Stefan Heidemann, Das Aleppiner Kalifat (A.D. 1261). Vom Ende des Kalifates
in Bagdad über Aleppo zu den Restaurationen in Kairo, Leiden, New York and Köln 1994.
21 See too the article of Amitai-Preiss reprinted in this volume, ch. 18 [8, n. 48].
22 D.O.Morgan (ed.), Medieval Historical Writing in the Christian and Islamic Worlds, London
1982, 123, n. 43. See A.H.Morton, "The letters of Rashid al-Dim: Ilkhanid fact or Timurid fantasy?",
in R.Amitai-Preiss and D.O.Morgan (eds), The Mongol Empire and its Legacy, 155–199.
23 D.Ayalon, "The Great Yasa of Chingiz Khan: a reexamination", details in note 1 of Morgan's
article.
24 In many of his publications he uses the form Amitai-Preiss.
25 D.O.Morgan, The Mongols, 9.
 A on the Destruction of Baghdad by the Mongols

By JOSEPH DE SOMOGYI

(PLATE I)
HARDLY ever has Islam survived a more disastrous and more H mournful
event than the destruction of Baghdad by the Mongols of Hulaghu Khan in
the middle of the month of of the year 656/January, 1258. The
Mongol conqueror, after having subdued the Assassins, turned against the
capital of the 'Abbasids and captured it without any resistance. The fall of
the 'Abbasid caliphate was followed by a veritable reign of terror which
lasted for forty days. Baghdad was plundered during this dismal period, its
entire population was massacred mercilessly with the exception of the
Christians, the co-religionists of Hulaghu Khan's wife and father. The
Caliph and his sons fell victims to the fury of the enraged
conqueror, who put them to death. And to complete the disaster, a great
conflagration destroyed many parts of the city.1
But all the more remarkable is the fact that we possess only very scanty
accounts of this veritable martyrdom of Islam in Arabic literary sources.
The most reliable author on the history of the 'Abbasids, Ibn al-Athir, closes
his Al-kamil fit-ta'rikh as early as the year 628/1230–1. Among the later
historians "neither Abul-Faraj nor Abulfida affords much information on
this subject. Indeed, of the Mongol siege in the seventh century A.H. we
know far less than we do, thanks to of the first siege in the time of the
Caliph Amin in the second century A.H." 2

So far as Arabic literature is concerned,3 we possess only three descriptions

of some length of these disastrous days of the history of Islam. One is by
Ibn who in 701/1301–2 wrote his famous Al-kitah al-Fakhri fil-
adab wad-duwal al-islamiyya,
at the end of which1 he describes the Mongol siege. The second is by Ibn
who lived one century later (died in 807/1404–5), and records the
same event in his hitherto unedited Ta'rikh ad-duwal wal-muluk.2 The third
is by adh-Dhahabi (died in 748/1348), who in his hitherto unedited and
voluminous Ta'rikh al-islam3 devotes a separate chapter to the fall of
Baghdad,4 which not only gives a detailed account of the event, but also
includes a lamenting the decline of the glorious city.
The Author.—The author of this is called by adh-Dhahabi Taqiaddin
Isma'il ibn abi'l-Yusr. His name is not to be found in any European
bibliographical work on Arabic literature, because no literary work bearing
this name has come down to us. In Oriental bibliographical works on
Arabic literature we only find two references to this author. The one is
contained in the Fawat al-Wafayat of ibn Shakir al-Kutubi (died
in 764/1362–3), the continuator of Ibn Khallikan's Wafayat al-a'yan. At the
beginning of his work al-Kutubi gives a short biographical account on the
author of our 5 His name is accordingly Taqiaddin ibn abi'l-Yusr

Isma'il ibn Ibrahim ibn abi'l-Yusr, "musnid ash-Sham." His uncle was a
scribe of the chancery of the Ayyubid Nuraddin, and he himself was scribe
to Da'ud,6 who was also a good poet. He is characterized by al-
Kutubi as being "distinguished in letter-writing, excellent in poetry and very
eloquent in speaking ". He was charged with. the prince's chancery, with the
superintendency of the cemetery, and with other administrative affairs.
Al-Kutubi's record is supplemented by a reference in continuation
of the 7
of adh-Dhahabi, where we read that it was from a
certain Ibn abi'l-Yusr that the grammarian Shamsaddin ibn
ibn 'Abbas ibn abi Bakr ibn Ja'wan (died in 674/1275–6) learnt.
As this scholar lived at the time of Da'ud, this reference
undoubtedly relates to our author, not to his father, who bore the same name
of Ibn abi'l-Yusr.

PLATE I. BULL. S.O.S. VOL. VII, PT. 1.

Our author's excellent qualities as recorded by al-Kutubi, and in particular
his talent for poetry, were certainly well known in his own time. Al-Kutubi
quotes some lines from his poetical works, but does not mention any
independent anthology or other work by him. This is probably due to the
circumstance that his poems were read only by a limited number of
courtiers and scholars in Damascus. In view of this, it is fortunate that adh-
Dhahabi, who lived about half a century later, could still recover a by
him and preserve it in his Ta'rikh al-islam, in the narrative of A.H. 656.
The Poem.—It is owing to adh-Dhahabi's conscientious citation of his
sources that this poem remains as the only work known to be extant of
Taqiaddin Isma'il ibn abi'l-Yusr. Considering the care shown by adh-
Dhahabi in quoting and copying his authorities, there can be no doubt that
this poem also was rendered by him as accurately as possible.
Among the MSS. of the Ta'rikh al-islam we possess two volumes
containing our One is in the Bodleian Library, No. 654 in the
catalogue of Ury, In this MS., which was written by a hand later to adh-
Dhahabi, the is contained on foll. 249–9b. The other MS. is in Istanbul
in the Aya-Sophia library, No. 3013, and has not been yet catalogued. As,
according to Professor O.Spiesz, who has seen this MS., it is an autograph
of adh-Dhahabi himself,1 it is from this latter MS. that I have copied the
text of the 2 to which. I have appended an English translation.
1. The fast-flowing tears give tidings of [the fate of] Baghdad; why stayest
thou, when the lovers have departed?
2. Ye pilgrims to az-Zawra'1 go not forth; for in that sanctuary and abode is
no inhabitant.
3. The crown of the Caliphate and the house whereby the rites of the Faith
were exalted is laid waste by desolation.
4. There appear in the morning light traces of the assault of decay in its
habitation, and tears have left their marks upon its ruins.
5. O fire of my heart, for a fire of clamorous war that blazed out upon it,
when a whirlwind smote the habitation!
6. High. stands the Cross over the tops of its minbars, and he whom a
girdle2 used to confine has become master.
7. How many an inviolate household has the Turk taken captive with
violent hands, though before that curtain were many protecting bastions!
8. How many [youths like] full moons [in beauty] upon al-Badriyya3 have
been eclipsed, and never again shall there be a rising of full moons
therefrom (v.l. "of the tribe or quarter ")!
9. How many treasures have become scattered abroad through. plundering,
and passed into the possession of infidels!
10. How many punishments have been inflicted by their swords upon men's
necks, how many burdens [of sin] there laid down!
11. I called out, as the captives were dishonoured and licentious men of the
enemy dragged them to ravishment—
12. And they were driven like cattle to the death that they beheld, "The Fire,
O my Lord, rather than this—not the shame!"
13. God knows that the people [of Baghdad] were made negligent by what
they enjoyed of divine favours, wherein was abundance,
14. So they grew heedless of the wrath of the Almighty, since they became
negligent, and there came upon them a mighty one of the hosts of infidelity.
15. Who shall aid men against calamities which. tell us of that wherein is
[for us] summons to judgment and warning?
16. After the capture of all the house of al-'Abbas, may no brightening
illumine the face of the dawn!
17. Nothing has ever given me pleasure since their departure save Sayings
of the Prophet that I pass on and Traditions of the Fathers.
18. There remains for neither the Faith nor the world, now that they are
gone, any market of glory, for they have passed away and perished.
19. Truly the Day of Judgment has been held in Baghdad, and her term,
when to prosperity succeeds adversity.
20. The family of the Prophet and the household of learning have been
taken captive, and whom, think you, after their loss, will cities contain?
21. I never hoped that I should remain when they had gone, but destiny has
intervened before my choice.
An Analysis.—As regards its contents, our can be divided into three
nearly equal parts. The first part (ll. 1–6), after a short invocation, describes
Baghdad as a venerated centre of religion which was laid waste by the
enemies of Islam, who are accused of promoting Christianity (1. 6). The
second part (ll. 7–14) poetically describes the sack and plundering of the
once rich. city and the slaughter of its inhabitants, and hints that those
terrors are a punishment inflicted by God for the heedlessness of His people
(ll. 13–14). The third part (ll. 15–21) is a mournful final accord which is not
unlike the "lasciate ogni speranza" of Dante: there is no hope left after the
fall of the 'Abbasids under whose rule the city flourished and the sciences
were cultivated; even the poet himself had not hoped to remain alive after
that veritable Day of Judgment (l. 21).
Our poem is consequently a funeral ode and belongs to a special class of
In their development all the earliest varieties of Arabic poetry
assumed the and the dirge (marthiyya) also shared in this process.
The sentiments felt at the death. of the beloved were first expressed by the
simple unpoetical then by saj'-verses, of which there developed short
metric sayings of some length, and finally the perfect marthiyya in the
metric varieties of the 1 Our consequently belongs to the class of
the
But whereas the marthiyya, as a rule, laments the loss of a prominent
person or a tribe, enumerating his or its qualities, our is a typical
example of a funeral ode lamenting the fall of a city.
Our poem, nevertheless, has all the necessary requisites and characteristic
features common to every Short as it is—consisting only of twenty-
one double verses—it is a fine piece of post-classical Arabic poetry written
in elegant language, and in the metre, the solemn rhythm of which is
especially suited to the dirge.
But, in addition to these common characteristics of the our poem also
shows some peculiarities shared by the only.
(1) The absence of the nasib. Whereas in the ordinary the opening
nasib is an essential requisite, it never occurs in the since the
object of the funeral ode is quite different.1 Instead of the nasib there are
some constant formulæ with which. a marthiyya begins. Thus the poet
sometimes refers to the tears shed on a tragic event, which is also to be seen
in our referring to the tears of those who lament the fall of Baghdad
(1.1).
(2) The repetition of the name of the lamented person,2 which is represented
here by some poetical names of Baghdad, as az-Zawra' (1.2) and Taj al-
khilafa (1.3).
(3) The repetition of the same phrase at the beginning of several
consecutive double-verses. This had been regarded from the beginning as a
peculiarity of the and, retained through its later poetical
development, it was also used in the period of decadence as an archaistic
rhetorical trick employed not only in the but also in other
classes of 3 Thus we see in our the four-fold repetition of the
phrase wa kam "and how many" (ll. 7–10).4
With these characteristic features our is a fine from the
period of decadence of Arabic literature. It is worthy of our attention for
two reasons.
Firstly, it is the only hitherto known work of Taqiaddin Isma'il
ibn abi'l-Yusr and a specimen of post-classical Arabic poetry written in the
refined style of the court-poets.
Secondly, it is to our knowledge the only poem lamenting the fall of
Baghdad and is an excellent poetical expression of the contemporary
sentiment felt at the fall of the 'Abbasids and at the tragedy of their capital.
Despite the decadence of the last 'Abbasids, their prestige was still so great
throughout the Muslim world that even the court-poet of the then
flourishing Ayyubid dynasty in Damascus could not help lamenting that
with them the splendour of Islam had passed away and that after the capture
of the Prophet's family he could not hope either to remain alive. His
presentiment was justified, because one generation later, in 699–700/1299–
1301, his own city, Damascus, and the Ayyubid empire were invaded by the
same Mongols who, after destroying the "crown of the caliphate", swept
over all the Muslim Orient.
 Notes on the Arabic Materials for the History of the Early
Crusades

By H.A.R.GIBB

THE publication of the first volume of M.René Grousset's history of the

Crusades, which is reviewed elsewhere in this issue, brings out again, and
all the more vividly because of its wealth of detail and effort to present a
complete and rounded-off picture, the very serious gaps in Orientalist
research on this period. Whereas the study of the Western and Greek
sources has progressed to a point at which it may be said that little more
remains to be done, research. on the Oriental sources is incredibly
backward. The European scholar has at his disposal, apart from the
topographical studies of van Berchem1 and M.René Dussaud,2 only two
works of any size, Derenbourg's study of Usama ibn Munqidh,3 and
Professor W.B.Stevenson's The Crusaders in the East (Cambridge, 1907),
together with such articles as those on the Syrian cities by Honigmann and
others in the Encyclopædia of Islam. Valuable as these are, they do not carry
him very far. Usama presents a lively picture of certain aspects of Syrian
life, but he was a minor figure and the scope of his material is too restricted.
Professor Stevenson attempted for the first time to situate the Crusaders in
their eastern surroundings, but the main object of his work was the careful
sifting of the Oriental sources for data of political history and chronology.
It is not, however, one or two general works which are required; it is a
whole series of monographs on important figures, on specific aspects of the
political and social life of the time, and on the Oriental sources themselves.
Not a single political figure prior to Saladin and the Third Il-
Ghazi, Zanki, Nur ad-Din—has ever been studied in detail; we know next
to nothing of the composition of the population in the various regions of
Syria, their relations with one another and with 'Iraq and Egypt, or of the
significance of the Shi'ite, and more especially the movements in
Syria; the criticism of the Oriental sources, Arabic, Syriac, and Armenian,
has not even begun. Failing these, the Muslim princes and peoples remain,
even in M.Grousset's work, so many lay figures, a kind of vague patchwork
backcloth against which the Western knights make a brave enough show,
until it presently falls down and envelops them, still valiantly struggling, in
its folds.
It is not the object of the present paper to remedy these deficiencies
forthwith, but to touch on certain points relating firstly to the social
situation in Syria, and secondly to the Arabic sources, which have emerged
in the course of several years' study of the period of the early Crusades.
 I

It is one of the principal services rendered by M.Grousset that, for the first
time in any general history of the Crusades, he brings out the importance of
the Byzantine "Crusades" of the tenth century as the forerunners of the
Latin Crusades, and as establishing a certain juridical claim by the Eastern
Empire to the restoration of its former Syrian territories, the last of which. it
had lost only in 1084. But it has generally escaped notice that the same fact
played a very important part in determining also the nature of the first
Muslim reactions to the Latin Crusades. For more than a century the
Muslims of Egypt, 'Iraq, and Persia had been accustomed to the spectacle of
Christian principalities in Antioch and Mesopotamia, and even of
intermittent Christian protectorates over Aleppo and parts of inner Syria.
The Christian states had taken their place in the normal political framework
of Syria, and the religious aspect of the struggle had long since ceased to
hold any prominent place in the minds of its population. Muslims and
Christians were intermingled with one another, especially after the
extensive immigration of Armenians into northern Syria; Christians ruled
over Muslims and Muslims over Christians, without interference from
either side. Though the Christian states had been temporarily recovered by
the Saljuqids, the report that fresh Christian armies were on their way
through Anatolia to recapture them roused no more than ordinary
apprehensions,1 and was regarded with comparative indifference by all the
Muslim princes except the one directly concerned, the ruler of Antioch
itself, Yaghi Siyan. That the newcomers were Franks, instead of Greeks,
conveyed very little to them. The Crusaders' occupation of Antioch. and
Edessa merely restored, from their point of view, the status quo ante. The
wazir, had been quick to seize the opportunity of renewing
with them the traditional defensive alliance against the
1
Saljuqids, temporarily interrupted in 1055. It is true that the negotiations
came to nothing when the Franks themselves seized Jerusalem from the
Egyptians, but even that failed to inspire an immediate uprush of religious
feeling and of resolve to drive them out. It was not merely the disintegration
of the Saljuqid empire, therefore, which was responsible for the absence of
any vigorous counter-attack from without.2 For a century and a half, Syria
and Mesopotamia had been left to fight their own battles, with some
intervention from Egypt, and for the most part Syria and Mesopotamia were
left to fight them now.
If this view be accepted, it is clearly a false conception to speak, as
M.Grousset has done, of every offensive against the Latin states as a
"counter-crusade". No doubt every war against non-Muslims, from the days
of Heraclius to those of 'Abd al-Karim, has been styled a jihad by its
supporters, but that in itself shows the cheapening of the term. What
distinguished the Crusades was that they were a mass movement, in which
men of all ranks and classes were caught and swept forward by a wave of
emotion. There was nothing corresponding to this amongst the Muslims
until the time of Nur ad-Din at the earliest, perhaps not until the time of
Saladin. Some faint hint of it may doubtfully be detected in the
undertakings of Mawdud, but even these were conducted as routine
expeditions, differing in no respect from any others. Only in one minor
episode of this period does one sense on the Muslim side something of the
Crusaders' exaltation of feeling, namely in the defence of Damascus against
Baldwin II's raid in January, 1126.3
It is almost equally misleading to regard the expeditions of the governors of
as so many instances of Saljuqid intervention, as when Karbuqa, for
example, arrives with a "grande armée seljûqide".1 None of the Oriental
sources suggest that Karbuqa had more than his own troops, together with
those of his minor vassals and of and Damascus. It should be recalled
that, although he was formally recognized as governor of Karbuqa
had in fact captured it for himself with a force of adventurers only two
years before,2 and that on the arrival of the First Crusade the Saljuqid
armies were engaged in Khurasan and almost immediately afterwards in the
long civil wars between Barkiyaruq and It is unlikely that there
was a single Saljuqid squadron in Karbuqa's force, and the size of his
private 'askar may be gauged from that of which his successor Jikirmish
disposed in the battle of namely 3,000 horsemen.3 The governors of
were drawn into the conflict by the Frankish. conquest of Edessa and
the resulting political complications in the Jazira; and even when they held
an official mandate to engage the Franks, it in no way affected the
essentially personal character and objects of their operations, unless perhaps
under Mawdud.4 The one genuine instance of Saljuqid intervention in the
whole history of the Crusades was the expedition under Bursuq b. Bursuq,
the governor of Hamadhan, in 1115; and the authorities are singularly
unanimous that this "counter-crusade" was openly directed against the
Muslim princes, and only as an afterthought against the Franks. It had the
striking result of bringing into temporary existence a Syrian bloc, Franks
and Muslims (except for two minor chieftains) making common cause
against the Eastern invader. Several causes may be and have been assigned
in explanation of this development, but in view of the absence of detailed
studies of the principal characters concerned, it is premature to come to any
definite conclusions. But two points, at least. seem to emerge from the fact
itself: one, that "counter-crusade" was the last idea entertained by the
princes of Syria and 'Iraq alike at that time1; the other that the Franks had
with surprising speed adapted themselves to the traditional atmosphere and
alignments of Syrian politics.
In regard to another aspect of the politico-social situation in Syria, the
Sunni-Shi'a schism, it is still difficult to reach absolute conclusions. A
careful study of the scanty contemporary materials, nevertheless, leads to
the impression that all historians of the Crusades have greatly exaggerated
its significance in Syria at the time of the First Crusade and in the following
decades. This is due partly to the fact that Western historians, seeking a
guiding thread in the labyrinth of Oriental politics, have thought to find it in
the religious schisms, and interpreting these as rival political groups have
used them as a kind of universal clue2; partly (and herein is their excuse)
that Ibn al-Athir and the other writers of the Ayyubid and Mamluk periods
were themselves obsessed to a great extent by an bias. In reality
the lines of political division had little to do with dogmatic differences, and
least of all in eleventh- and twelfth-century Syria. Had the been
inclined to religious intolerance, the case might have been different, but
they were (apart from the personal eccentricity of ) one of the most
tolerant dynasties in Islam. If the Islamic world had been otherwise unified,
the emergence of political Shi'ism would have been disastrous3; but though
it prevented union, it was not in itself a prime cause of disunion. The
healing of the schism was a necessary prelude to the union of forces against
the Crusaders, but the schism was not a factor of importance in their first
success.
The real mainspring of Syrian politics, it can hardly be doubted, is to be
found in the principle of "beggar-my-neighbour" which had governed the
relations of the amirs of Syria and Mesopotamia ever since the
disintegration of the Caliphate. Where ambition, jealousy, and fear were the
dominant motives, questions of religious conformity and belief were of
small account. Religion had long since abdicated the claim to control
political action,1 and the only other restraining force, love of country, while
not absent amongst the general population and possibly even such minor
local chiefs as the Banu Munqidh, was obviously ineffective where foreign
Turkish governors were concerned. No student of Islamic history in the
tenth and eleventh. centuries needs to be reminded that when the Saljuqid
of Aleppo declared for the in 1097 in view of an alliance
against Damascus, he was but following the footsteps of numerous amirs
and princes, who had accepted or rejected the nominal suzerainty of one or
other Caliph for the sake of securing a momentary tactical advantage over a
local rival. Similarly, the readiness of Ibn 'Ammar of Tripolis to assist the
Franks and even to accept a quasi-protectorate, could find more than one
parallel in the history of Syria since the days when the Arab Shi'ite
of Aleppo had invoked the Byzantine protectorate and seen the
great Basil II himself hastening to defend them against their fellow-
countrymen and co-sectaries, the
Thus the refusal of Aleppo and Damascus, and that of Damascus and Egypt,
to co-operate against the Crusaders were due to the same general causes,
into which religion scarcely entered. In the former case, they took a
personal form: the rivalry between the sons of Tutush, and in particular the
resentment of at the loss of Damascus.2 In the second case, they
were rather historical: the spectre of the former Egyptian occupation of
Damascus on the one side, and of the former kingdom of Tutush on the
other. The rulers of Damascus were afraid that the should attempt to
reassert their claim to the city; the Egyptian government feared lest a
restored Saljuqid kingdom should attempt the coup which Tutush may have
planned, but never carried out. Both sides were consequently not ill-
pleased, in the long run, that the Latin kingdom of Jerusalem interposed a
buffer between them—provided the buffer did not turn into a boa-
constrictor. Nothing is more instructive in this connection than to observe
the deliberate inertia of Damascus on the Palestine front between 1099 and
1105, while Godfrey and Baldwin I were engaged in establishing the
kingdom and securing it against the Egyptian counter-attacks, and how, as
soon as was convinced that the Egyptians were unable to dislodge
the Crusaders, he willingly co-operated with them, not in combined attacks
with full forces, but in minor operations designed to harass the Franks and
prevent the expansion of the kingdom. Note, too, how the relations between
Egypt and Damascus grew progressively more cordial, to the extent that
even instigated Egyptian raids (if Ibn Muyassar1 is to be believed),
and that finally he and his successors accepted robes of honour and
diplomas.2 The same indifference to sectarian divisions was shown by
Usama b.Munqidh, who served Zanki and the with equal zeal, by
Ibn 'Ammar of Tripolis, and even by the Egyptian wazir whom the
3
history asserts to have been a fervent Isma'ili but the Damascus
chronicler claims as "a firm believer in the doctrines of the Sunna ".4 Before
the close of the twelfth century, however, there can be little doubt that
Shi'ism was thoroughly discredited in Syria, but it remains to be
investigated how far the activities of the were responsible for this
change, or whether it was a by-product of that waxing religious enthusiasm
which led up to the real Counter-Crusade under the leadership of Saladin.
 II

The second field in which Orientalist research has lagged behind, and
which is a prerequisite for any real study of such problems as are touched
on above, is the critical examination of the Oriental sources. Every historian
of the early Crusades has up to the present used the Kamil of Ibn al-Athir as
the principal Arabic source, and has accepted his version of affairs under
the control of Kamal ad-Din's Chronicle of Aleppo and ibn al-Jawzi's
Mir'at az-Zaman. The recovery of Ibn al-Qalanisi's Damascus Chronicle
completely changes the situation. It is not only that Ibn al-Qalanisi is a
contemporary and reflects the contemporary attitude, whereas Ibn al-Athir
is permeated by the very different mentality of the thirteenth century,5 nor is
it that the former veiws events from the angle of Damascus and the latter
from the wider but more distant angle of The important point for our
present purpose is that Ibn al-Qalanisi is one of the original sources of Ibn
al-Athir—the only one for this period so far recovered—and a comparison
of the two accounts enables us to investigate his methods of compilation,
and to check in some degree the accuracy of his information in this portion
of his chronicle. The results of this examination are not reassuring, and go
to show that while Ibn al-Athir, because of his much wider field than that of
either the Damascus or the Aleppo chroniclers, must always remain a
principal source, he is not to be relied on in details of fact, of chronology, or
of interpretation, and must always be used with caution.1 Outstanding
though his work is, in comparison with the historians of his own age whose
productions have come down to us, he is yet not entirely free from those
romantic and empirical tendencies which are visible over a wide range of
mediaeval Islamic literature.
A detailed analysis being impossible within the limits of an article, it is
proposed in the following paragraphs to examine briefly a few typical
passages, illustrating how Ibn al-Athir's methods may result in misleading
or suspect information, and to touch. still more briefly upon Kamal ad-Din's
work in the same connection.
(1) Ibn al-Athir very frequently suppresses elements of the original
narrative, and occasionally uses the rest to support a false interpretation.
Under A.H. 494 (1100–1) Ibn al-Qalanisi relates an attempt by Sukman
b.Ortuq to recapture Saruj, and its recovery by the Franks. His text reads as
follows:

("In this year the amir

Sukman b.Ortuq collected a great host of Turkmens and marched with them
against the Franks of ar-Ruha [Edessa] and Saruj in the month of First
Rabi'. He captured Saruj and was joined by a large body [i.e. of volunteers],
while the Franks also collected their forces.")2 Ibn al-Athir, apparently
because he found no account in his sources of the capture of Saruj, rewrites
this and misrepresents it as the capture of Saruj:

("In this year, the Franks

captured the town of Saruj in Mesopotamia. The cause of this was the
Franks had already captured ar-Ruha…and at this time Sukman collected a
great host of Turkmens in Saruj and marched against them.")1
(2) The frequency with which. Ibn al-Athir alters the dates given by Ibn al-
Qalanisi (and always does so wrongly)2 raises a question of motive, which
cannot be answered at present. But occasionally also he alters the tenor of a
sentence or phrase in the original. An example will be found in his account
of the Crusaders' capture of Tripolis in 1109 (A.H. 502), which is freely
quoted from Ibn al-Qalanisi. The relevant passage in the latter reads:

("Their spirits were

lowered by universal despair at the delay of the Egyptian fleet in bringing
provisions and reinforcements by sea, for the stores of the fleet had been
exhausted and the direction of the wind remained contrary, through the will
of God that that which was decreed should come to pass.")3 Ibn al-Athir, to
begin with, places this (wrongly) under the year 503, and renders the
passage above by:
("Their spirits were lowered and their weakness was increased
by the delay of the Egyptian fleet in bringing them provisions and
reinforcements. Now the cause of his [presumably dilatoriness in
regard to it (the fleet) was that he did not
give his attention to it and to hastening on its preparations,1 and they [? the
Egyptian ministers] disagreed (or shilly-shallied) about it for more than a
year; and it set off, but the wind drove it back, so it became impossible for
them to reach Tripolis, in order that God should bring about a matter which
was to come to pass.")2
The difference between these two versions is obvious. Ibn al-Qalanisi
implies that the stores and provisioning for the fleet and the town of
Tripolis were not available until the harvest in the spring of 1109, that the
necessary measures were then taken without any stinting (cf. Damascus
Chron., p. 91), and that the delay was a fatality due to the contrary wind. If
he does not explicitly absolve the Egyptian government from the charge of
dilatoriness, at least he says nothing to incriminate it. Ibn al-Athir, on the
other hand, makes a definite accusation against the government, and
particularly asserts that the fleet was detained in Egypt "for more than a
year". There is, fortunately, no dubiety in this instance; Ibn al-Athir's
statement is untrue. For Tripolis fell in July, 1109; in August, 1108, the
Egyptian fleet was in Syrian waters, and had very effectively come to the
assistance of Sidon, defeating a considerable squadron of Italian vessels and
relieving (and presumably reprovisioning) the town.3 It had therefore
returned to Egypt only in the late autumn, and the story that it was kept
back "for more than a year" is a fiction due to bias.4 Whence,
then, did it find its way into Ibn al-Athir's chronicle? That he derived it
from another written source seems to be excluded by his otherwise close
following of the text of Ibn al-Qalanisi. There can therefore, it would seem,
be little doubt that the source is a certain oral tradition current in in
accordance with which Ibn al-Athir "corrected" the statements of his written
authority.
(3) Another of Ibn al-Athir's tricks of compilation is to group together a
number of items, sometimes quite unconnected, sometimes even of
different dates, which as a result of this grouping convey, whether by
accident or design, a certain impression. Thus, under A.H. 504, after
relating Tancred's capture of al-Atharib in that year (December, 1110),1 he
proceeds: "And great was the fear of them amongst the Muslims, whose
hearts rose into their throats, for they were convinced that the Franks were
about to capture all the rest of Syria, for lack of any to defend it and repel
them from it. So the lords of the Islamic cities in Syria set about negotiating
an armistice with them, but the Franks would not agree to any terms except
a tribute in ready money, and that only for a short period." He then appends
a list of rulers and places and the amounts which they undertook to pay:
of Aleppo, 32,000 dinars and other objects; the lord of Tyre, 7,000
dinars; Ibn Munqidh of Shaizar, 4,000 dinars; 'Ali al-Kurdi of 2,000
dinars; and concludes: "the armistice to run only up to the time of the
ripening and harvesting of the crops."2 Note that this passage is inserted
after Mawdud's victory at in July, 1110, and conveys the impression
that even this brought no real relief to the Muslim territories, which were
just as exposed to Crusading attacks as they had been before. The idea that
undoubtedly influenced Ibn al-Athir in so arranging his material (whether
deliberately or not) was his firm conviction that the Muslims in Syria were
beaten from pillar to post until the advent of Zanki,who was the true
champion of the Faith and repeller of the Franks.3 But of the four
agreements which he cites, only one, that between and Tancred,
certainly dates from after the capture of al-Atharib.4 The agreement
between Baldwin and Tyre was concluded in 1107 or 1108,5 that between
Tancred and Shaizar in 1109 or early in 1110.6 The armistice with if
the statement is correct (for no other independent source mentions it),
probably dates from the same period as that with Shaizar. By grouping these
three with the Aleppo agreement, Ibn al-Athir unduly magnifies the effect
of Tancred's victory and to that extent misrepresents the actual situation in
Syria.
(4) It is a habit of Ibn al-Athir to supplement the information contained in
his sources with picturesque anecdotes, some of which may possibly have a
basis of fact, but which more often, probably, serve the purpose of summing
up in a striking sentence or illustration either the historian's own view or the
traditional view of a given situation. Two examples may be quoted.
Immediately after the passage referred to in the preceding section, Ibn al-
Athir inserts, in abridged form, accounts (derived from Ibn al-Qalanisi) of
the riots provoked by refugees from Aleppo at Baghdad against the Sultan
and the Caliph in February, 1111, and of the Greek embassy of the previous
month (December-January).1 To these he adds: "The men of Aleppo used to
say to the Sultan 'Have you no fear of God, that the king of the Greeks
should be so much more zealous than you in the cause of Islam, as even to
have sent an embassy to you to engage you in the Holy War against the
Franks?'"2 The addition is evidently a reflection generated in the lively
imagination of the chronicler by the accidental juxtaposition of the two
items, but possibly in this instance no great distortion of historical fact is
involved.
The second example, which relates to the assassination of Mawdud in the
Great Mosque at Damascus on 2nd October, 1113, is not so innocent. The
Damascus Chronicle (p. 140) leaves the motive of the assassination
unresolved. Ibn al-Athir3 attributes it to a adding: "Some said that the
in Syria feared him and killed him, and others said that on the
contrary it was who feared him and set a man on to kill him."
Having thus (quite justifiably) performed his duty as a historian in
recording the view which was taken at and also, apparently, at the
court, Ibn al-Athir proceeds:

("My father related to me that the King of the Franks wrote a letter to
after the slaying of Mawdud,
which contained the following phrases: 'A nation which has slain its
support, on its festival day, in the house of the Being whom it worships,
justifies God in exterminating it'"). The story has every appearance of being
legendary; it is derived from oral tradition at and attributes to Baldwin
I a pretty taste in Arabic rhymed prose; but it serves the purpose of giving
telling expression to Ibn al-Athir's own conviction without actually
committing him to it in his own words.1 For the rest that conviction
strongly colours his account of the relations between and Mawdud
during the previous campaigns, and leads him even to revise the story of the
assassination itself, where he heightens the dramatic effect (and at the same
time implicitly contradicts Ibn al-Qalanisi's careful description of its actual
setting) by asserting that "they were walking hand in hand".
(5) These, and many other instances which could be cited, raise the very
important question of how far Ibn al-Athir is to be trusted when he is the
sole authority for an alleged event. It is impossible, of course, to lay down
any general principles. While he is likely to be more trustworthy in dealing
with events in and the neighbouring provinces than with those which
took place at a distance, a certain caution is surely justified in receiving his
unsupported statement, and the more sensational and picturesque it is the
greater need there is of hesitation to accept it at face value. Two cases may
be briefly examined here by way of illustration.
In connection with the attack made on Damascus by the united Latin forces
in the late autumn of 1129,2 Ibn al-Athir has a long and circumstantial story
to the effect that the wazir at Damascus, Abu 'Ali al-Mazdaqani, and his
protégés entered into a conspiracy with Baldwin II to deliver
Damascus to the Franks in return for the possession of Tyre. This is
represented as being the cause of the Crusaders' attack, which was,
however, forestalled by the massacre of the in Damascus in the
preceding September. There is no hint of this in Ibn al-Qalanisi (and on this
occasion there is no reason why he should have adopted a reticent "official"
attitude, since the existence of such a plot would have given additional
point to his vigorous denunciations of the sect), nor, more strangely still, in
the Latin historians, and the suggested retrocession of Tyre is highly
suspicious.1 The plot is not necessary to account for the Frankish
attack on Damascus, in view of the death of in 1128 and the arrival
of the new Crusading army. Thus the story, though not impossible, seems to
be nothing more than romantic invention, the starting-point of which was
supplied by the massacre of the in Damascus and their subsequent
surrender of Banyas to the Franks.
The second example is offered by Ibn al-Athir's story of Zanki's capture of
al-Atharib in 1130, which it is the more important to correct since even
Professor Stevenson makes one of his rare slips in this connection.2 Under
the year 523 (1129) Ibn al-Athir inserts, in an abridged form, the narrative
which Ibn al-Qalanisi gives under 524 (1130). There can be no question that
524 is the correct date. Sawar, who was apparently at the time govenor of
for the amir of Damascus, took part with the 'askar of in the
operations against the Crusaders round Damascus in December, 1129, i.e. in
the last days of 523.3 His transference of his services to Zanki is therefore
correctly dated by Kamal ad-Din early in 524, and accounts for the
appointment of Sawinj. to the command of Zanki's "jihad" in this
year (524/1130) consequently consisted of two treacherous assaults on the
possessions and persons of his Muslim allies. But Ibn al-Athir, having
placed all this in 523, is left with the task of finding suitable employment
for his hero in 524. Now it happened that during the conflict between Alice
of Antioch and her father Baldwin in that year a body of Muslims unnamed
made a raid on the suburbs of al-Atharib and of Ma'arrat 4 The raid

may have been made by Zanki's troops, during his stay at Aleppo prior to
the seizure of It is this quite minor expedition which has apparently
been seized upon by the tradition and exultantly magnified into the
full-dress opening of the Counter-Crusade, signalized by the siege, capture,
and dismantling of al-Atharib after a tremendous defeat of the entire
Frankish forces. And with an impressive rhetoric which seems to carry its
own conviction, Ibn al-Athir concludes the detailed narrative of these
mythical exploits with the words: "The fortunes of the Muslims were
revolutionized in those districts; the power of the Infidels weakened and
they realized that there had come into the land that which had never entered
into their calculations, and the most that they could do henceforth was to
hold what they possessed, whereas heretofore they had nursed the ambition
of conquering it outright."1
(6) Kamal ad-Din, in his Chronicle of Aleppo, bases himself largely on
independent sources, but sometimes quotes Ibn al-Athir and sometimes also
Ibn al-Qalanisi, usually without abridgment. He is less sensational than Ibn
al-Athir and more straightforward, probably also more reliable in detail. Yet
he too sometimes adds to his sources, whether with or without justification
can rarely be said. Thus the passage in which he relates the surrender of
Artah by its Armenian garrison is transcribed textually from Ibn al-
Qalanisi,2 but he adds at the end: "And this was all due to the evil conduct
of Yaghi Siyan and his tyrannical government of his lands"
This is clearly an unauthorized
supplement, an attempt to explain an unwelcome fact by the familiar
method of throwing the blame upon an individual. In this instance, the
solidarity which the Armenians of Cilicia and the Taurus had shown with
the Crusaders renders the explanation unnecessary; and even were Yaghi
Siyan a particularly bad governor (and there may well have been a tradition
at Aleppo to that effect), he can hardly be held responsible for their action
at this juncture.
A more complicated problem is offered by the narrative of the unsuccessful
siege of 'Azaz in the year 517 (1123–4), which according to the existing text
of Ibn al-Qalanisi was undertaken by and Aqsunqur in June, 1123,
and according to Kamal ad-Din in January, 1124, by the combined forces of
Balak b.Ortuq and the other two.3 The difference of dating is the more
remarkable since, except for his introductory sentence, Kamal ad-Din
quotes Ibn al-Qalanisi almost textually. The change has therefore been
deliberately made, and for the obvious reason that during June and July,
1123, Balak was engaged in occupying Aleppo and as much as possible of
the territory to the south-west of it.1 It is unlikely, on the other hand, that
Ibn al-Qalanisi was mistaken as to the month, and it is surprising to find no
mention of Balak in his narrative. The explanation is that by some error the
whole paragraph relating to this campaign in Ibn al-Qalanisi's s book (or
some copies of it) was inserted under A.H. 517 instead of A.H. 519 (1125–
6). It followed naturally on Aqsunqur's relief and occupation of Aleppo in
January, 1125, and is mentioned in its proper place by Fulcher of Chartres
(iii, 42), whose description tallies with that of Ibn al-Qalanisi, as well as by
Kamal ad-Din himself and by Ibn al-Athir.2 Moreover, Aqsunqur spent the
year 517 in 'Iraq, where he was engaged in hostilities with Dubais, and did
not return to until 518.3 It is clear also that the paragraph was
accidentally misplaced from the fact that Ibn al-Qalanisi follows up the
account of the battle by relating the despatch of an envoy from Damascus to
Egypt, the reply to which arrived in August, 1126.4 The only possible
conclusion is that Kamal ad-Din, finding this expedition related under A.H.
517 in his copy of Ibn al-Qalanisi, and unable to accept the date there given,
transferred it and combined the narrative with that of an isolated attack
made on 'Azaz by Balak at the close of 517, and thus unwittingly
transformed a minor raid into a major operation terminating in a serious
defeat for the Muslims.
These few examples may serve to show how much there is to be done in the
textual and historical criticism of the Arabic sources, and also that the
materials at our disposal, however incomplete, enable it to be done to a
certain extent. Such a critical scrutiny must, obviously, be made on the
Arabic texts themselves; for this reason, it is not on the historian as such
that the work must fall in the first instance, but on the Orientalist who
possesses an adequate equipment for this new field of "higher criticism".
Not until he does his part will a satisfactory and fully balanced history of
the Crusades become possible.
The Influence of Chingiz-Khan's Yasa upon the General
Organization of the Mamluk State

By A.N.POLIAK
 1

IN an article published several years ago we have collected evidence

corroborating al-Maqrizi's statement (mistrusted by Quatremère in Histoire
des Mongols) that siyasa, the legal code of the Mamluks, was founded upon
the Great Yasa of Chingiz-Khan.1 The Great Yasa was not merely a code of
criminal and civil law but a system of rules governing the entire political,
social, military, and economic life of the community which. adopted it. The
expansion of this system outside the Mongol nation was due to the belief
that it was responsible for the extraordinary military success of the Mongols
in the thirteenth century, and that it might be regarded as a talisman
ensuring victories on the battle-field.2 The Yasa rules concerning communal
organization were even more important from this point of view than the
laws treating of the behaviour of individuals. It is natural, therefore, to
suppose that not only the Mamluk criminal, civil, and commercial law but
also the general organization of the Mamluk state was based upon the Yasa.
The present article, inspired by the attempts made in modern times to
collect and systematize the fragmentary evidence concerning the contents of
the Yasa,3 is intended to show that this organization is indeed
comprehensible only in the light of such evidence.
Some preliminary remarks are necessary. The Turkish states founded within
the Moslem caliphate prior to the advent of the Mongols and to the
promulgation of the Great Yasa already possessed their own Yasas, codes
obligatory exclusively upon the Turkish-speaking military castes, not upon
the natives, and based upon Turkish tribal traditions. The foundation of
these states closely resembled the establishment of Teutonic kingdoms
within the Western Roman Empire: in both cases the fiction of imperial
unity continued to exist,
and the subjugated indigenes were administered, according to imperial laws
and usages, by their own magistrates, who employed the imperial language
as the official one. The conquerors formed a military feudal caste, which
was headed by its kings—who, from the standpoint of imperial constitution,
were regional commanders of troops engaged by the empire—and which
had its own language, laws, and courts of justice. The separation of the two
castes was slightly relaxed by the conversion of the conquerors to the
imperial religion, which necessitated some respect for the native clergy and
the use of the imperial language in religious matters; but the relations
between the two castes were regulated by the law of the rulers. Nur al-Din
b. Zangi was the first who compelled the Turkish military caste to comply
with the demands of Islamic law,1 and he is explicitly stated to be one who
abandoned in favour of this law the siyasa, a term evidently designating
here not the Great Yasa, which. did not yet exist, but the local Yasa of the
Zangid state.2 The early corruption of the word Yasa to siyasa is a
noteworthy point, as also the similarity of that Yasa to the Great Yasa in
recommending more cruel punishments for rebels and highwaymen than.
those ordered by Islamic law. Quite possibly, technical terms used in the
Great Yasa, or at least in its version used by the Mamluks, were largely
derived from these earlier codes; one is tempted to give the rein to fancy
and suppose that the Great Yasa was originally conceived as an improved
edition of the Yasas of Moslem Turkish rulers, which were doubtless more
elaborated and adapted to the needs of a large empire than the kindred
traditions of Mongol herdsmen and of those Turkish tribes which remained
outside the Moslem empire.
It seems that, owing to the semi-magic power ascribed to the Great Yasa,
this code was in general concealed by the rulers from subjugated
populations: among the Arabic, Syriac, Persian, and Armenian authors who
tell us about it, not one explicitly pretends to have seen it himself or appears
to be familiar with the order of its chapters and articles. As to the members
of the military caste, in the Mamluk state they were evidently acquainted
not with the Mongol text written in Uigur characters but with a version
written in their Turkish dialect in Arabic letters. Those of them who did not
specially prepare themselves for the career of military judges, but had
to perform
judicial functions in their quality of sultans and governors, were sometimes
illiterate,1 and could learn by heart oral lessons from this version. At any
rate, this version could not differ considerably from its Mongol source, and
since the materials concerning the organization of the Mamluk state are
more numerous than those regarding any part of the Mongol empire, we
may, if our thesis is correct, consider them as the best instrument for
reconstructing the Great Yasa. We must only remember that, according to
the tradition inherited by the Mamluks from earlier Turkish states, a Yasa
regulated the internal life of the ruling caste and its relations with the
natives, whereas the internal affairs of the latter were governed by Islamic
law and, within limits fixed by it, by religious codes of non-Moslem
communities.
The Mamluks did not consider the Great Yasa as a code of an entirely
strange people. As the Mongols admitted Turkish tribes to their military
caste (cf. the evolution of the term "Tatars" in the Golden Horde), so the
Mamluk- "Turks" regarded the Mongols—and the Circassians, the Alans,
the Russians, and all the peoples regularly recruited for Mamluk troops—as
Turks,2 which meant for them primarily " horsemen of the steppes". The
number of peoples represented in various times in the Mamluk army was
very great, but regular recruiting took place only among populations
acquainted, partly or wholly, with steppe horsemanship. Though most of the
Mamluks were brought through ports of Crimea, and numerous merchants
and religious officials of the Mamluk state were Qirimis, viz. Turkish
natives of the agricultural, mountainous, Crimea regions,3 there was no
Qirimi group in the army. The Turks of the South Russian steppes,
Qypchaq, were, on the contrary, an important element in it4 until the
transformation of these steppes towards the end of the fourteenth century
into an almost depopulated expanse, "the Wild Field," as the Russians later
called it, owing to the excessive emigration into the Mamluk and
Lithuanian possessions, and to the Black Death
of 1348.1 The Russians (Rus, Urus), at first very numerous in the army,
ceased to play there an important part at the same time as the Qypchaqs,
which indicates that they did not originate from forested agricultural
Russian principalities but were nomads (brodniki) of the Qypchaq steppes.2
Ibn Khaldun describes these Slavic nomads as herdsmen of oxen and sheep,
living under the same conditions as Turks.3 If the horticultural inclinations
of the Circassian sultan al-Ghawri show his nostalgia for the Caucasian
mountain-gardens,4 the diet preferred by the first Circassian ruler of the
Mamluks, Barquq, was characteristic of a steppe herdsman of horses:
horsemeat and koumiss.5 It appears that the Circassians, like their eastern
neighbours, the Alans or Ossets, whose Christian religion they shared,6
inhabited then not only the mountains but also adjacent steppes, and that the
subsequent shrinking of the Circassian and Ossetic territories was a
consequence of the mass emigration to the Mamluk state. This situation
may explain the long dependence of Circassia on the Golden Horde,
mentioned by al-Qalqashandi as late as 1412. Owing to it the Circassians
could participate in this emigration from early times, especially in view of
their military reputation which was increased by the similarity of their
ethnic name in sound and in spelling to a
Mamluk word meaning "courageous".1 Significantly, the Georgians, who
became under the Ottomans the principal element of the Egyptian Mamluk
corps,2 did not play an important part in the army of the Mamluk sultans;
the share of the Alans was there more considerable than that of the western
neighbours of Circassia, the Abkhasians, whose connection with the steppes
was smaller3; and those Armenians who participated under Sultan Qalaun in
the Burji regiment of the royal Mamluk corps Were recruited in Armenian
colonies scattered to the north of the Black Sea rather than among the
inhabitants of Armenia proper, military foes of the Mamluks. The first
nucleus of the Mamluk army consisted mainly of northern Oghuz, Turk,
plur. Atrak (the special application of this name to them in the Mamluk
sources recalls its similar use, in the form of "Torks", in early Russian
sources, and seems to have been customary in their area), or Ghuzz; both.
of these ethnic names became general appellations of the military caste, and
their language was the basic element of the Mamluk dialect.4 Under Sultan
they yielded for the first time
their privileged position to the Circassians,1 and after Barquq this change
became firm; but even later their share in the Mamluk army was not
negligible. A striking phenomenon is therefore the relatively humble part
played there by southern Oghuz scattered over Western Asia, Turkuman,
Tarakima, or Tarakimun2: those of them who dwelt within the Mamluk state
had the status of natives, prohibited from wearing "Turkish" uniforms3, and
supplying auxiliary troops only; natives of Mardin and Asia Minor (Rum)
Were admitted to "Turkish"regiments, but not to the Turk faction there.4
The reason seems to be that they had lost those traditions of steppe
horsemanship which ensured the Mamluk military superiority over native
(bedouin) cavalry. Similarly, though the Mamluks admitted at first to their
troops some Kurdish warriors of the former Ayyubid principalities,5 later
this Kurdish. element disappeared, and new immigrants from Kurdistan
received only the status of native auxiliaries. The Mongols were always
welcome in the Mamluk army, and only their high standing prevented them
from becoming there very numerous, because they had to be enlisted as free
and qualified warriors, and not as apprentices who remained temporary
slaves until the completion of their military education. This temporary
slavery was not a great humiliation, because since the time of Nur al-Din
every Mamluk who reached the age of majority had to be manumitted; but
those knights who had never been slaves felt themselves superior to those
who had to pass such an apprenticeship.6 Only towards the end of the
Mamluk state was it imposed on the Mongols as on others.7 The Egyptian
al-Maqrizi admired the physical beauty of the Oyrats settled in Cairo, just
as the Iranian Sa'di admired the Mongoloid features of Central Asian Turks.
 2. The Ruler

As we know from Codex Cumanicus, a manual of West Turkish spoken in

southern regions of the Golden Horde,1 and other sources, there were in the
Golden Horde only three ranks above the common knight: emperor qan or
khan, king and prince or baron bay (beg)=amir. We find the same
graduation in the Mamluk sources, which describe the Mamluk sultan as the
suzerain of the local beysemirs and sometimes as a vassal of the emperor
(khan) of the Golden Horde, who was in his turn a vassal of the higher
emperor of the Mongols, "the Great Qan." Owing to the absence of a rank
intermediate between sultan and emir, strong vassals of the Mamluk sultan,
as princes of Mecca, Medina, and Yanbu', were called in the documents
issued by the royal chancellery emirs only, though they were designated by
their own subjects—and frequently even by Mamluk writers—as sultans.1
The comparison of the Mamluk graduation of titles with that of the Golden
Horde sheds a new light on the action of Baybars I, who called his son
Barka-Khan (al-Manhal v, s.v., explicitly vocalizes this name Barka
and not Baraka) after the latter's maternal grandfather, the emperor of the
Golden Horde: its purpose was evidently not only to please the emperor but
also to promote the son and his successors from kingship to imperial rank.
The promotion was particularly important because the title of sultan was not
Mongol but one conceded by the Mongols to native rulers and chieftains
subjugated by them in the west, while the title of khan or qan was,
according to the testimony of Juwayni and Bar Hebraeus, the only one
accorded to the rulers by the Great Yasa. This object was attained, and the
subsequent emperors of the Golden Horde did not demand their mention as
overlords in the Friday sermon in Mamluk mosques; and if the Mamluk
khan continued to be called in Arabic sources it was because of his
post of al-Islam, administrator of the Moslem community in the name
of the 'Abbasid caliph, so far as the Sunnis were concerned, and as
lieutenant of the caliphs, so far as the Isma'ilis were concerned.2
When the Mamluk ruler became again a vassal of a foreign lord, Tamerlane,
the Mamluk chancellery was compelled to underline the latter's superiority
by applying to him the ancient title of Qara-Qytai's emperors, gurkhan.3
The elective character of the Mamluk monarchy recalls that of the Mongols.
In both cases the ruler was elected by chief dignitaries of the state, who
were simultaneously commanders of the army. The Mamluk state council,
majlis, as the Mongol quriltay, was an advisory body summoned by the
khan whenever he wished to hear the opinion of his electors and other
learned and influential persons on important problems. The numerous and
turgid titles applied to the Mamluk rulers and emirs in Arabic official
documents seem to violate the Great Yasa's rule ordering extreme simplicity
in this respect; but the Mongols regarded this rule as obligatory upon
themselves only, while their clerks, recruited from conquered populations
and composing documents in their languages, had to call the Mongol lords
by titles which. were traditionally applied by these peoples to their
masters.1 One Mongol was a comrade for another, but not for a captured
native; and the Mamluk-"Turks", who considered themselves as a branch.
of the imperial people, followed the usage of its main body.
 3. The Military Caste

The Great Yasa obliged every inhabitant to devote a part of his time to
performance of some service for the state. The Mongols themselves were
not divided into two castes, charged with military service and public works
respectively: Juwayni expresses his admiration for the fact that a Tartar
soldier was a peasant, expected in the time of peace to pay taxes, to
maintain post service, to perform works of public utility, and to prepare
arms, munitions, and everything necessary for warfare. When, however, the
Mongols in foreign countries captured by them became a military caste
dominating the disarmed native peasantry, military service had to be
separated from labour service, and the Mamluks did likewise. Under
Baybars I the knights and emirs still participated sometimes in public
works, such as repairs of Egyptian irrigating canals in 1264–5,2 but later
such cases do not occur. The term bigar, which continued in Iran to
designate the compulsory labour-service of peasantry,3 was used by the
Mamluks as peculiar to the military service of the Turkish-speaking troops,
and hence to these troops themselves,4 while the labour-service of the
Arabic-speaking peasants was called 'awna, the Arabic translation of
bigar.1 By the way, the word bigar was applied from early times to active
military service of Turkish troops in the caliphate's service,2 and might be
used in early Yasas for these troops, just as Juwayni uses it (in its Iranian
connotation) when speaking of the Mongol Great Yasa; we cannot,
however, affirm that it was then already connected with. the idea of general
compulsory service for the state. Similarly, only the "Turks" exempt from
active military service were called the appellation given by the
Great Yasa to persons immune from general state-service. This is another
example of technical terms brought by the Turks to Moslem countries long
before Chingiz-Khan3; adopted by the Great Yasa, which modified their
meaning in accordance with its laws and with Mongol connotations of the
respective words4; and then borrowed again by Moslem Turks from the
Mongols together with the latter's Yasa. In order to ensure the accurate
performance of the general state-service, the Great Yasa ordered that the
population be distributed into fixed divisions attached to specified areas and
charged with specified services and taxes, and that every person be
forbidden to pass from one division to another without permission. The
Mamluks complied with this rule when they compelled every province to
maintain its own troops; forbade the transfers of troopers from one province
to another without sanction of the central government; and located their
fiefs, as a rule. within the respective province only, and the habitual
residence of a trooper in the administrative centre of his province. The
Ottoman feudal system inherited this usage, but an Ottoman feudatory was
more frequently enabled to inhabit his fief. A survival of the Mongol
system of general state-service was the duty of the Mamluk knights and
emirs to know not only how to use arms but also how to make them, not
only how to ride horses but also how to breed them.5 It was perhaps
responsible for the extreme dislike of fire-arms by the Mamluk troops.
These arms never became normal weapons of "Turkish " emirs and knights,
and when al-Ghawri established a mercenary corps of gunners and
musketeers, instead of recruiting such specialists from slaves and serfs, this
action was bitterly resented by Ibn lyas, whose admiration for traditional
arms was doubtless shared by other knights.1
The conquest of Moslem countries by Turks was rendered possible by the
superiority of Turkish cavalry. In the steppes occupied by Turkish tribes
grass annually grew longer than in the more southerly areas of the Arab
bedouins, and therefore these steppes were able not only to support greater
numbers of men, horses, and cattle, but also to accustom horsemen to
longer annual migrations from winter encampments to summer pastures and
vice versa, and, consequently, to extreme mobility. The Arabs could not,
like the Turks, shoot arrows without ceasing or reducing their gallop2; and
while in both cases cavalry was distributed during battles into three
autonomous divisions, called by Arabic sources qalb "heart, centre ",
maymana "right wing", and maysara "left wing",3 the Arabs sought only to
surround, if possible, by their wings the enemy's flanks or one of them,
whereas the Turks, as later the Chingizid Mongols, aimed at surrounding
the entire army of the enemy as by a ring, in Arabic sources.4 The
Turkish system of military education, which was further elaborated by the
Mongols and made a part of the Great Yasa, was founded upon this
manœuvre. It included various exercises during which horsemen had to
form mobile rings,5 and the annual great hunt of the Mongol army, headed
by the khan in person, which took place in early winter, was also such an
exercise, the game being surrounded by the ring of warriors in the same
manner as a hostile army. The game occurring in the deserts of Arab
countries was negligible in comparison with that of the Mongolian, Central
Asian, and South. Russian steppes; but as the annual great hunt was
prescribed by the Great Yasa, the Mamluk sultans practised it in the Libyan
desert, and their Syro-Palestinian governors-general in areas specified by
them.1 The Ayyubid army was divided into "the royal special warriors of
the ring" viz. small feudatories who formed the
2
centre commanded by the sultan, and the knights in the emirs' service, who
constituted the two wings; the Mamluks reorganized these two corps in
accordance with the decimal system required by the Great Yasa, which
necessitated the graduation of emirs according to the number of tens of
Mamluks in their service, and the distribution of into hundreds and
thousands, often nominal only. The adjectives gradually
fell into desuetude, as the ruler became connected with his Mamluk corps
more than with free "warriors of the ring ", who ultimately ceased to be a
military body.
 4. The Serf Caste

The general obligatory service for the state established by the Great Yasa
assumed in the case of native peasants the form of labour-service.
According to Pétis de la Croix, the Great Yasa compelled those who did not
go to war to participate during a specified season of the year in public
works, and to devote a day weekly to some work for the khan. The seasonal
labour-service took in Egypt the shape of annual repairs of irrigating dams
and canals, for which the peasants had to supply money, implements, and
working cattle, in addition to their manual work. The repairs of great dams,
al-jusur were supervised by regional "inspectors of the soil"
kushshaf al-turab, who were emirs appointed to these positions by the
sultan, while small dams, al-jusur al-baladiyya, were under the control of
local feudatories.1 The maintenance of the irrigation-system was always one
of the chief duties of the Egyptian governments, but according to al-Maqrizi
the 'Abbasid governors and the caliphs entrusted it to their
feudatories (he does not make things clear when-speaking of the Ayyubids,
but it appears from contemporary chronicles that they followed the example
of their predecessors). Peasants were mobilized also for construction of new
canals, bridges, and fortresses.2 As to the compulsory regular work for the
khan, it was carried out in demesne farms of the Mamluk ruler and his fief-
holders, to whom he transferred this prerogative.3 Another service imposed
on the peasants was, as in the Mongol state, the maintenance of post-
stations along principal routes. In both cases horses, forage, and money
were supplied by neighbours of such a station.4 Another Mongol institution
adopted by the Mamluks, public granaries, was probably established in the
same manner. They were divided, like irrigating dams, into royal,
and local, baladiyya. In 1315 the former contained in Egypt alone 160,000
irdabbs of grain, i.e. approximately 316,800 hectolitres.5 Their main
purpose was to supply to cultivators loans in kind as seeds and as food until
the harvest (in Egypt also to enable them to use green manure, which. was
not utilized in Syria and Palestine). The royal public granaries must not be
confused with the private granaries of the sultan, or with granaries of the
crown domains.
The law which prohibited the peasants from leaving their respective
villages without permission of their lords, and then only for a specified
time,1 was not brought into existence under Frankish influence, since it was
established not by the last or the Ayyubids, when the Franks were a
strong and imposing power, but by the first Mamluk sultans, when. the
remnants of the Frankish states were moribund. Moreover, in Syria,
Palestine, and the Lebanon it was applied less rigorously than in Egypt.2 Its
source was the Great Yasa, which ordered distribution of the population into
fixed divisions, attached to specified areas. While in the case of a Mamluk
warrior the area within which. he could freely move, when not on duty, was
identical with the province to whose troops he belonged and in whose
centre he dwelt, a native peasant could freely move only in the lands
utilized by his village community. These lands formed for the purpose of
cadastral surveys a unit, whose cultivators could inhabit either the
village (balad, qarya) or small hamlets (kufur) in its vicinity; but if the
population of such a hamlet became numerous, it was transformed into a
particular village-community and 3 The division to which a peasant

belonged, "loaded beast" (the name seems to be a translation of a

Mongol one, like the corresponding Russian tiaglo), could contain either
the entire community or its fixed sub-division, clan, entitled to a fixed share
of the common arable land, which was expressed either as a fraction whose
denominator was 2, 3, 4, 6, 24, less frequently 5 and 7 (redivisions of great
clans into smaller resulted in redivisions of common lands into a greater
number of "shares" ashum or "portions" which was a multiple of the
previous ones), or as some number of faddans (in Egypt 1 faddan was 5,929
square metres, while in Syria, Palestine, and the Lebanon its extent differed
in various villages); and such a division with its lands formed the local
"portion" of an estate-holder, lord of this division (his other
possessions, if any, being usually situated in other parts of the respective
province).4 As we have remarked elsewhere, the main principles of the
Mamluk feudal land-tenre were brought from the Golden Horde, and the
principal tax paid by the peasants, kharaj of Arabic sources, was nothing
but the Mongol qalan.1 Whereas the Islamic kharaj was the tribute imposed
on those lands whose proprietorship was vested in the state and whose
holders had only the right of hereditary lease, as opposed to the tithe levied
on allodial grounds,2 the Mamluk tax was paid by tenants of all cultivated
lands, whether "tribute-paying (kharaji)" or "tithe-paying" from the
standpoint of Islamic law, to their immediate lords. Even in documents
emanating from religious authorities, such as the endowment-deed of Sultan
Qaitbay,3 the term "kharaji lands" was applied to allodial estates whose
tenants paid these rents.
Native townsmen were, from the standpoint of Mamluk law, not a particular
caste having a common status, but they were exempt in groups from duties
imposed on the serf-caste. Small towns were from this standpoint villages,
and cities, military settlements; and native merchants, artisans, clerks, and
physicians, were nowhere citizens and normal inhabitants, but only
exceptions. When during the last Mamluk-Ottoman war the Ottomans
utilized against the Mamluks the slogan of 'adl " justice",4 which meant
from Nur al-Din's time the annulment of Yasas and the return to Islamic
law, it easily attracted the bulk of these townsmen, because it implied their
emancipation, or at least made their own position and that of their property
more certain.
The MS. of this valuable article was sent in two years ago, and, in view of
the difficulty of communications, the proofs have not benefited by Dr.
Poliak's final corrections. Here are a few points on which, in normal times,
the author's opinion would have been duly consulted:— P. 863. The
hypothesis of the existence of "the Yasas of Muslim Turkish rulers" needs
further proofs. The term itself (yasa) does not occur either in the Orkhon
inscriptions, or in Uyghur texts, or in Kashghari (470/1077).
P. 865. It would be more adequate to explain brodniki not as "nomads", but
as "free-lances".
P. 866. The Arabic name of the Circassians, Jarkas, apparently stands for
Persian *Chahar-Kas, i.e. "the Four (tribes) of Kas". Kas was the ancient
name of the Circassian people: in Ossete Kas-ag, in Old Russian Kas-og, in
Arab geographers al-Kas-ak-iyya, see al-'Alam, 446. The popular
etymology of Jarkas, as current in Egypt, should be explained as "[equal to]
four persons".
P. 866. A mistake about the language of the military caste is difficult to
admit. Two centuries before the author of the Bulghat al-mushtaq, the
author of the glossary published by Houtsma also describes the "Qipchaq-
Turkish" language, which he distinguishes from Turkmani.
P. 868. The Turkish-Mongolian name Bärkä (sometimes Bärkäy) apparently
means "a hostage"; Kashghari, i, 357.
P. 868. Barthold's assertion concerning the disappearance of the title
gurkhan is right. The title kurkan has nothing to do with *gür-khan. It must
be read "a son-in-law", as, for example, Tamerlane was surnamed.
P. 871. Nerge has hardly the same meaning as It seems to be a variant
of jirga "a row" in Juvayni: "a line of communications," or "a line of
beaters". In Turkish *y can give both j and n, cf. yumurtqa/nimurtqa.
V.M.
Studies on the Structure of the Mamluk Army1—I

By DAVID AYALON
 The Army stationed in Egypt

CONTEMPORARY sources furnish fairly ample information on the

structure of the Mamluk army and the units from which. it was composed;
but, though some of their definitions and descriptions come near to the
truth, the present writer has found none of them. to be completely accurate.2
He is of the
opinion that the treatment of this question by modern Orientalists is also in
need of some emendation,1 and he here submits the composition of the
Mamluk army which, in his opinion, is the correct one, and the grounds for
his suggestion will be presented in the course of the discussion.
The Mamluk forces stationed in Egypt were divided into three main parts:
I. The Royal Mamluks (al-mamalik very rarely: mamalik ). These
were. of two categories:
(a) The mamluks of the ruling sultan (mushtarawat, ajlab, or julban).
(b) Mamluks who passed into the service of the ruling sultan from the
service of other masters (mustakhdamun). These were divided into two
parts:
1. Mamluks who passed into the service of the reigning sultan from that of
former sultans (mamalik al-mutaqaddima, );
2. Mamluks who passed into the service of the reigning sultan from that of
the amirs, because of the death or dismissal of their masters (sayfiya).
II. The Amirs' Mamluks (mamalik al-umara', ajnad al-umara' 2).
III. The troops of the (ajnad ) a corps of free, i.e. non-mamluk,
cavalry. There was within the a special unit composed of the sons of
the amirs and of the mamluks, called awlad an-nas.
 The Royal Mamluks

The Royal Mamluks constituted the backbone of the Mamluk army. They
were given first-rate training, and in the period numbered not less than
10,000 troops; in the Circassian period they apparently never exceeded this
figure, and usually fell far short of it (see section on the numbers of the
mamluks, pp. 222–8). In the major engagements fought by the Mamluk
kingdom, the brunt of the fighting invariably fell on the Royal Mamluks,
who formed the main force in all military expeditions. In the words of al-
Qalqashandi: 'The Royal Mamluks are the most important and the most
honoured part of the army, the closest to the sultan, and the owners of the
largest feudal estates and it is from among them that the amirs of
various ranks are appointed.'1 The history of the Mamluk kingdom is
actually first and foremost the history of the Royal Mamluks. No other unit
could compare with this corps either in military strength or in political
power.
The Royal Mamluks were not spread out as garrisons in scattered parts of
the kingdom, but were all concentrated in the capital, with a large
contingent of them quartered in barracks in the Cairo citadel. It is only at
rare intervals that some of them, representing but small fractions of the
corps, are found serving as garrison troops. At in Upper Egypt, far from
the centre of the realm and linked to it by poor communications, we find in
the second half of the 7th century A.D. a garrison of Royal Mamluks,
charged with repressing the Nubians.2 After the conquest of Cyprus, there
was also a garrison of Royal Mamluks on that island.3 Towards the close of
the Mamluk era, Royal Mamluks were sent in comparatively small groups
to trouble-spots within Egypt itself.4 Mostly chosen for such. tasks were
members of the corps who had incurred the disfavour of the sultan, or who
were considered as deserving punishment. On the eve of the fall of the
Mamluk state, aged mamluks were sent out to garrison various sectors of
Egypt.5 There was in Mecca, for a large part of the Circassian period, a
garrison of a few dozen Royal Mamluks. The bulk of the corps, however,
remained in Cairo during the most difficult times, and were not even
evacuated during a plague which. wrought havoc among the mamluks.6
They left the capital only when forming part of an expeditionary force.
Mamluk sources offer an immense fund of information on the several
components of the Royal Mamluks. The greater part of this material is,
however, of an unvaried character, and is restricted almost exclusively to
accounts of these units' struggle for power, of their coalitions, and of their
rivalries. The description of political strife among the various army units is,
in fact, one of the pivotal points of Mamluk historiography. Data as to the
military value, efficiency, degree of training, etc. of these units are found
scattered throughout the sources, but are far less abundant than this type of
information. Nevertheless, in view of the fact that the degree of cohesion of
each of the component units and its rivalries with. its sister units were
factors which had a direct influence on the strength and efficiency of the
army, a description of them, at least in outline, cannot be omitted.
Before proceeding to a description of these units, brief mention should be
made of the Royal Mamluks' commanders. These were called muqaddamu
al-mamalik The numerical proportion of the Royal Mamluks to
their commanders is known in one instance only, viz. for the year 715. This
ratio emerges from the list of the army compiled in connexion with. the
redistribution of Egyptian lands between the sultan and his feudatories,
which took place under b.Qalaun (ar-rawk ). The
figures given are 2,000 Royal Mamluks and 40 muqaddamu al-mamalik
i.e. one commander for every fifty mamluks. It should be noted that
the ratio of men to commanders in the given in the same list is also
approximately 50 to 1.1 This is, however, the only piece of information of
its kind, and there is no way of knowing to what extent this proportion was
maintained at other times. The term muqaddamu al-mamalik 2 is
more frequent in the than in the Circassian period.

THE MAMLUKS OF THE RULING SULTAN

(mushtarawat, ajlab, julban)

Those mamluks who were bought and set free by the ruling sultan
constituted the chief support of his rule. The Mamluk system of servitude
(the foundations of which. are discussed elsewhere3) instilled in the
mamluk a feeling of profound loyalty toward his master and liberator
(ustadh) on the one hand, and for his fellows in servitude and liberation
(khushdashiya or khushdashin, sing. khushdash)4 on the other. This twofold
loyalty was one of the principal axes around which revolved the entire
military and social system of Mamluk hierarchy. The sultan and his
mamluks formed a tightly-knit association, whose members were united by
strong bonds of solidarity. There existed between the sultan and his
mamluks a sort of double bond: they were in power only so long as he
ruled, and he ruled only so long as his power was based on them. In
addition, the sultan buttressed his power by appointing his own
khushdashiya to high. positions, and similarly, during the Circassian period,
by appointing his blood relations as well. Sultan Barsbay used to say: 'No
one executes my orders except my own mamluks', 'my commands are
obeyed only by my own mamluks'.1 Elsewhere the sultan is told: ' By God,
were it not for thy mamluks, no one would obey thee'.2 When it was said of
any one that he was betrayed even by his own mamluks,3 it meant that he
was in dire straits indeed.4
The mamluks of the ruling sultan were called mushtarawat,5 sometimes
mushtarawat,6 but it should be noted that this name does not occur
immediately at the beginning of the Mamluk period. It is, for instance,
almost unheard of in the days of Baybars; his mamluks were then
called to distinguish them from the or the the
mamluks of Najm ad-Din Ayyub.7 In the Circassian period, a new
name for the mamluks of the ruling sultan appears which. becomes more
frequent than mushtarawat, without displacing it entirely, viz. ajlab,8 or
julban,9 sing. jalabi 10 or jalab11 (which is a generic name as well).12 The
last two forms are almost non-existent in the sources. The earliest historians
in whose works we find the terms ajlab and julban are Ibn Khaldun 13 and
Ibn al-Furat,14 the latter mentioning them in the closing years of his
chronicles. The mushtarawat, ajlab, or julban are often referred to as al-
mamalik 1 One of the synonyms of mushtarawat is mamalik
2but this appellation is also often synonymous with Royal Mamluks

(mamalik ).3
The new sultan, upon attaining power, attempted to pave the way for the
rise of his mamluks. The most frequent concomitant of a new sultan's
accession to the throne was a ruthless and large-scale purge, and especially
the thorough suppression of the mamluks of the preceding sultan, i.e. those
mamluks who had, until that moment, been the mushtarawat and the most
powerful body of the realm. This procedure is known from the dawn of the
Mamluk era, but had a more moderate character during the period,
because of the vigour of the legitimacy principle still obtaining at the time:
the young sultan could not regard the mamluks of his father, the preceding
ruler, as wholly alien to him, just as his father's mamluks did not consider
him a wholly alien sultan.
Simultaneously with the purge, the new sultan fostered a new generation of
young officers among his mamluks, most of them lacking in experience.
The replacement of the old by the new officer class had, of course, to be
carried out carefully and step by step, for it would have been. impossible
entirely to dispense with the services of the veteran officers without greatly
endangering the very foundation of the army's existence and efficiency.
Thus it was customary for the sultan, in the first years of his reign, to
appoint one group after another of his mamluks to the ranks of the lower
officers (Amirs of Ten) so that these ranks were often filled by his own
mamluks. He would then gradually follow the same procedure with. respect
to the higher ranks.4
The methods employed by the new sultan to destroy the immense power
which his predecessor's mamluks had amassed during their master's reign,
and to undermine their resistance capacity as a cohesive unit, were not
restricted to removing them from influential positions. Much more drastic
steps were taken, including imprisonment and exile, mainly of chiefs and
leaders, or transfer to the service of the amirs.5 The former sultan's youthful
mamluks who had not yet received their freedom (the kuttabiya) were
bought by the new sultan for himself 1; after they had completed their
training and received their liberation papers they became an integral part of
his mushtarawat. One of the most drastic and important steps usually taken
by a new sultan, especially in the Circassian period, was to throw all his
predecessor's mamluks, bag and baggage, out of the barracks of the
Cairo Citadel (qal'at al-jabal) and establish his own mamluks in their
stead.2 The citadel, being the heart of the whole Mamluk empire, and
dominating the capital, thus became the main stronghold of the sultan and
his close supporters.
Not all sultans employed the same degree of severity towards the mamluks
of the preceding ruler. The record for cruelty is held by Amirs and
Yalbugha, who, after deposing Sultan Barquq, nearly rent asunder the whole
body of his mushtarawat. There was in this case, however, an important
racial factor added to the ordinary circumstances; it has been discussed
elsewhere by the present writer.3
There is no doubt that the special structure of the Mamluk army compelled
the sultan to act as he did if he was to hold the reins of control. For it must
be remembered that he was but a newly appointed ruler, and as such, even
had he formerly been commander-in-chief of the army (atabak al-'asakir),
the number of his mamluks was far smaller than that of his immediate
predecessor. The buying of thousands of young mamluks and their training
in military schools required several years at least; during these years he
faced, with but a handful of men, a tightly-knit corps of no less than one
thousand, sometimes several thousand, mamluks of his predecessor who, up
to the accession of the new sultan, had enjoyed almost absolute power. On
the new sultan's side, it is true, were his khushdashiya, who helped him to
reach power, as well as units of the mamluks of the sultans anterior to the
immediately preceding ruler, who welcomed the opportunity of taking
revenge on yesterday's ajlab. But he could in no way count on these
formations as he could on his own mamluks, who were tied to him by the
firmest bonds of union and solidarity. To lay solid foundations to his
authority, two ways only were open to the sultan: weaken the ajlab of his
immediate predecessor as much as possible, and multiply the number of his
own mamluks within the shortest possible time. One amir advised Sultan
b.Qalaun: 'Do not leave a big ram in thy king-
dom, but foster thine own mamluks'. (la tatruk fi dawlatika kabshan
kabiran wa-anshi' mamalikaka.)1 This brief sentence succinctly summarizes
the traditional policy pursued by the Mamluk sultans, especially the
Circassians, in their relations with the cohorts of their predecessors.2 A
most characteristic illustration of the immense power enjoyed by the
mushtarawat, in contrast with the feebleness of the mamluks of the
preceding ruler, is furnished by the historian in connexion with the
mamluks of Sultan Khushqadam. He states that these mamluks
formed the numerically largest unit, the strongest and the most respected,
but also the most oppressive and the most tyrannical, for they drew their
power from the existence of their master-purchaser (the ruling sultan):
ha'ula' hum akthar 'adadan wa-aqwahum
shawkatan wa-aghshamuhum li-'izzihim
bi-wujud ustadhihim'. 3

So long as the Mamluk kingdom was wealthy and powerful, and so long as
its army was well-trained, well-disciplined, and ruled with an iron hand, the
feeling of fellowship or khushdashiya constituted a positive factor. An
esprit de corps, tending to foster moderate competition among the various
units, stimulated the troops and prevented stagnation. When, however, the
foundations and principles of the Mamluk military system began to
crumble, together with the whole Mamluk state structure, when discipline
loosened up and the selfish impulses of the various units burst forth.
unbridled, this feeling of solidarity turned against the interests of the army
and became one of the main factors in its ruin. The way was opened for the
total subordination of all units to the mushtarawat, with. its accompanying
unrestrained extortion and oppression, as well as for indulgence in the
political affairs of the realm and neglect of military duty. This state of
affairs went so far that we are told of sultans who sent forth to war the
members of the veteran units, while their own mamluks were spared that
discomfort.1 Their military training was most ineffective, and they were
lacking in warlike spirit. Ibn Taghribirdi was of the opinion that 100
could put to flight over 1,000 julban, and that, were it not for respect for the
sultan, even the lowest black slaves of Cairo would be sufficient to rout
them.2 Elsewhere we find the ajlab unable to overcome a small troop led by
an amir, in spite of their large number and superior armament, because of
their poor training and their ignorance of the arts of war. They were put to
flight while the engagement was in full swing.3 We are also told of an
encounter between the ajlab and a Cairo mob, in which the ajlab fled
shamefully.4 Helplessness in combat and unwillingness to fight were
common failings among them in the later Mamluk period.5 But whereas in
combat they were wanting both in ability and in courage, they distinguished
themselves in the political intrigues by which they were able to control and
subjugate the weaker units. They similarly distinguished themselves in all
manner of wanton and irresponsible acts. The signs of this deterioration of
discipline appeared early at the beginning of the Circassian period, and had
some sporadic forerunners even in the eriod. But the complete
breakdown began in the second half of the 9th century A.H.
The historians point to the reign of Sultan Aynal as the time when all
restraint was removed from the caprices of the julban. In 858, says Ibn
Taghribirdi, 'is the first appearance of the mamluks of al-Ashraf (Aynal) and
that which is to follow is yet more awesome'.6 The sultan no longer
exercised any control over his mamluks.7 All of Aynal's virtues were
rendered void by the escapades of his ajlab, which caused extreme damage
to the kingdom.8 The people prayed for his death during his illness out of
hatred of his mamluks,9
who committed deeds never perpetrated by the mamluks of his
predecessors.1 The same stories appear in connexion with Sultan
Khushqadam (whose mamluks even surpassed those of Aynal in their
misdeeds)2 and especially in connexion with the mamluks of Sultan
Yalbay.3
From the middle of the 9th century onward, Mamluk sources are permeated
with the terror of the ajlab, and a very great number of pages are devoted to
its description. There are whole years in which little would remain in the
chronicles if the description of these nefarious activities were removed.
Hundreds of stories are told of the expulsion of high state officials (mostly
those connected with payments to the army) the burning of their houses, the
pillaging of the markets and shops of the capital, the burning down of the
townspeople's houses, the abduction of women without any voice being
raised in protest, the amirs' fear of the ajlab, from whom they hide their
treasures, etc.4 The sultan completely loses control of the ajlab,5 who stone
him and put him to shame in public.6 Whenever they wish to extort
something from the sultan, they prohibit his going up to the citadel.7 They
intervene in questions of appointments and depositions of sultans,8 and they
have their way in the appointment of the king of Cyprus and that of the
highest amirs of the kingdom.9 This situation had the effect of terrorizing
the population, for it was known in advance that 'whatever they do will be
permitted them, and the sultan will not protect those oppressed by them '.10
In such. an anarchical state of affairs, the law-courts lost all their value, and
whoever desired anything addressed himself not to the tribunals, but to the
ajlab.11 One of the main reasons for the Circassian sultans' infrequent
departures from the capital was the constant state of upheaval into which
the city was thrown by the activities of the julban. When Qaytbay allowed
himself to leave the city more frequently than his predecessors he came in
for severe criticism.12
For all their misdeeds the julban were given surprisingly light penalties.
when the sultan has one of them soundly trounced his act is described as
being in opposition to accepted usage.1 Elsewhere we are told of an
offender being merely beaten, while his armour-bearer is put to death.2
Some ajlab, who had killed a mamluk amir, were only beaten and
imprisoned.3 We do hear of sultans cutting off the hand and the foot of a
thief from among the julban,4 but this is an exceptional occurrence.
Moreover, even in cases where the sultan had intended firm punishment and
condemned the offenders to death or to having their hands cut off, he
commutes the penalty to lashing.5 The sultan's order to the forbidding
them to marry the ajlab without his permission was totally ignored by the
latter.6

The 7 (bodyguard, select retinue, pages)

The (sometimes called khassakiya,8especially in the period)

were the sultan's corps of bodyguards and select retinue. Mamluk sources
furnish two basic descriptions of this body, due respectively to al-
Qalqashandi9 and to 10 The second is the superior one, and describes

the as those who surround the sultan even in his hours of solitude
(fi khalawatihi, i.e. outside official duties); they lead the ceremonial litter of
the are charged with bringing to the governors the robes of
honour confirming their appointment, and are sent out on political missions.
They are the prospective amirs, and are the closest to the sultan. Al-
Qalqashandi adds that they are differentiated from the other members of the
service in that they bear their swords on them; they wear brocaded bands (
zarkash); they are admitted into the sultan's presence in his private
moments without previous permission; they are meticulous in their dress
and their riding. These descriptions are fully supported by the chronicles.
From the amirs' biographies it can be said that most of them reached the
amirate by way of the corps1 (see Part II). The sources also stress
the great prestige of the and the honour derived from being related
to them.2 Frequent reference is made to their being sent on special missions
to foreign states, their being appointed governors of al-bilad ash-shamiya,3
and their being dispatched to arrest and imprison rebellious amirs and
governors.4 According to there were among them ten pen-box
holders (dawadariya), ten special cup bearers (suqat ), four treasurers
(khazindariya), seven masters of the robe (ra's nawbat jamdariya), four
armour bearers and four shoe-bearers (bashmaqdariya).5 The
same author indicates that ordinary Royal Mamluks, not belonging to the
could hold similar offices.6 The accuracy of this statement is
difficult to ascertain from the biographical dictionaries, as they are
concerned only with those mamluks who reached the amirate, a rank
usually attained via the corps.
The were of course Royal Mamluks,1 and generally belonged to
the mushtarawat.2 The of Sultan al-Ghawri, who numbered 1,200,
were all from among his mushtarawat.3 Nevertheless, one finds among the
a small number of men who not only were not mushtarawat or
Royal Mamluks, but who were not mamluks at all. Thus, for instance,
b. Badlik as-Saqi;4 b.Ibrahim al-Makhzumi at-Taluli.5 The
father-in-law of Sultan Barquq, b.'Ali b.Abi Bakr b.Ayyub, who was
the chief engineer of the kingdom, became a and afterwards an Amir
of Ten.6 The eunuch Shahin. was a 7 The awlad an-nas

will be discussed under the heading 'The ', in Part II of this

article.
As to the numbers of the they underwent great changes from one
period to another. According to they numbered about 40 under
Sultan b.Qalaun, and multiplied to the point of reaching 1,200 in
the days of Sultan Barsbay.8 According to Qalqashandi, they numbered at
first 24, equalling the number of the Amirs of a Hundred; in the days of the
historian (the sultanate of an-Nasir Faraj) they had reached 400.9 Another
source has it that Faraj increased their number to over 1,000; al-Mu'ayyad
Shaykh is then said to have reduced them to 80, the number they reached
under Barquq, and to have reduced to six the number of the dawadariya,
who had numbered 80 under Faraj. This latter sultan similarly reduced the
number of the treasurers (khazindariya), the shoe-bearers (bashmaqdariya),
and the chamberlains 10In 891 the of Qaytbay numbered '40
11
and no more'. Al-Ghawri, who inflated the kingdom's officer corps,
greatly increased the number of the as well. In 908 he fixed their
number at 800, and they subsequently reached 1,200.12 In 922, the last year
of Mamluk rule, the still numbered 1,200 men, all from al-
13
Ghawri's mushtarawat. A description of the offices held by the
and a history of these offices within the Mamluk kingdom, would have to
be based on broader material than we have used here.14 Suffice it to say that
toward the end of the Mamluk era the master of the robes (jamdar) was not
of the but that his office constituted a preliminary stage to
membership in that corps. This may be seen from such expressions as: 'wa-
a'taqahu min jumlat al-mamalik al-jamdariya thumma baqiya
',1 'wa-kan min jumlat mamalik al-jamdariya wa-lam
yakun 2
', ' khamsa wa-jamdariya mi'a wa-khamsin'.3 The
chief of all the jamdariya was called ra's nawbat al-jamdariya.4

THE MUSTAKHDAMUN

Whereas ustadh usually refers to the purchasing master, or the master who
buys and frees, the usual meaning of makhdum is the master into whose
service the mamluk enters after he has received his liberation from the
ustadh.5 The root khadama is found in various forms, having the meaning
of 'service with a master following service with the liberating master', thus:
'huwa 'atiq fulan thumma khadam ' fulan'6 and many other instances.7
The recurrent phrases 'thumma bi-fulan',8 'thumma
fulan',9 refer to a mamluk's transfer from the service of the original master
to that of another. The term istakhdam, as applied to mamluks, generally
refers to bringing into one's service a mamluk liberated by another master.10
Mustakhdamun (rarely: mamalik al-khidma) thus refers to mamluks who
passed from the service of the
original master to that of a new one, and this term is set in opposition to
mushtarawat.1
Within the framework of the Royal Mamluks, both the and the
sayfiya were mustakhdamun, for both groups entered the service of the
ruling sultan after having served other masters.

The Mamluks of Former Sultans

The mamluks of former sultans were called mamalik al-

mutaqaddima, 2 or 3 Very frequently, however, they were not
all included under one appellation, but appeared as separate units under the
surname of their respective masters. Thus, in the days of al-Ashraf Barsbay,
the mamluks of the sultans who had preceded him included the
(Barquq), the (Faraj), the mu'ayyadiya (Shaykh), and others (see n. 2,
p. 218). Much more rarely these units were designated not by the sultan's
surname, but by his first name, e.g., al-khushqadamiya, al-aynaliya, al-
jaqmaqiya, etc.4 Each such unit was generally called 'group' or 'faction' (
 pl. ).5 In contrast to the mushtarawat, who formed a single and
united the mamluks of former sultans did not, naturally enough,
constitute a united front. Their factions were separate and often inimical,
and their sole unifying factor was their hatred of the mushtarawat, who
were their juniors and lorded it over them. The mamluks of ex-sultan B
could not forget how the mamluks of ex-sultan A had forcibly displaced
them from the status of mushtarawat, and the same grudge was borne by
the mamluks of ex-sultan C against those of ex-sultan B, and so on. The
chances for conflict among the various units of former sultans'
mamluks were considerable, especially in the Circassian period, when the
number of these units was very large as a result of the sultans' brief reigns
(al-Maqrizi notes seven such units for his own time).1 This in turn gave rise
to all sorts of coalitions and combinations of forces, of which Mamluk
historiography records many instances. Such coalitions were generally of a
most temporary nature, and the stability of each sultan's rule was to a large
extent dependent on his ability to take full advantage of the rivalry among
the various units.
A detailed presentation of the vast material supplied on this topic by
Mamluk sources is of no special interest. Only the overall conclusion is of
importance; the actual political skirmishes and intrigues offer little variety.
Some of the material gathered in this connexion will be found in the
footnotes.2 We shall content ourselves here with a single example drawn
from the reign of sultan Khushqadam, showing the extent to which these
coalitions were fluid, and the dexterity needed by the sultan to consolidate
his position on such shifting sands. In his obituary notice reviewing
Khushqadam's life, Ibn Taghribirdi relates that it was the constant friction
among the various that gave the sultan peace of mind and a feeling of
security from sudden attack, since he was aware of the disunity prevailing
among the several mamluk units.3 When, however, we examine the career
of that sultan, the picture becomes less idyllic. Ibn Taghribirdi himself
offers some interesting details of Khushqadam's precarious situation in 868,
when he was compelled to administer an oath. of loyalty to his Amirs of a
Thousand. This lack of security derived from his strained relations with the
various mamluk These were, in the order of their seniority: al-
mu'ayyadiya Shaykh, al-ashrafiya Barsbay, Jaqmaq and al-
ashrafiya Aynal. The mu'ayyadiya were the colleagues (khushdashiya) of
the sultan. They were few in numbers (less than fifty) but most of them
were amirs. The governors of Aleppo and Damascus, as well as other
important Damascus amirs, were of their number. In Egypt the amir
majlis (lord of the audience), the amir akhur kabir (grand master of the
stable), the dawadar kabir and one other of the Amirs of a Thousand
belonged to this The rest of the mu'ayyadiya were Amirs of Forty and
Amirs of Ten (see the amirs' offices and ranks below) so that much. power
was concentrated in the hands of the members of this group in spite of their
low number: 'wa-hum kathirun bi-hadhihi al-kayfiya'. As for the ashrafiya
Barsbay, they formed the greater part of the army (or of the amirs?) and
many of them were Amirs of a Thousand, Amirs of Ten, and
office holders (arbab ). They were among the sultan's adversaries, for
many of them had been exiled and imprisoned by him, together with their
important amirs. The Jaqmaq constituted the backbone and main
strength of the army of that time: 'wahum al-'askar al-an'. Many
of them were Amirs of a Thousand, Amirs of Forty, Amirs of Ten, and
office holders. The sultan put their leader Janibak ad-Dawadar in prison, as
well as Tanam and exiled and imprisoned many others of them. They,
like the ashrafiya Barsbay, stored up much resentment against the sultan.
The ashrafiya Aynal, i.e. the mamluks of the sultan's immediate
predecessor, were numerous, but lacked leaders, since the sultan persecuted
them with the utmost ruthlessness, especially since the time they had
conspired against him with his mushtarawat. From that day, the sultan had
pursued, dispersed, and harassed them without respite, going so far as to kill
their leaders or drown them in the Nile. In the same year his relations with
his mushtarawat were greatly strained, and he had a great fear of them. It
thus emerges that he had no military unit left on which he could rely except
the sayfiya; these had no feeling of comradeship, and allied themselves
always with the winning side (see p. 220). Under these circumstances, the
sultan was compelled to attempt a rapprochement with the Jaqmaq,
and to make amends for having persecuted them in the past. His apologies
were accepted for want of a better choice, and the became his allies
in outward appearance. Inwardly they kept their grudge against him,
convinced as they were that he was the source of all their misfortunes. The
sultan was well aware that their allegiance was mere lip service, and that
they would cease supporting him as soon as they had reconciled themselves
with their rivals, the aynaliya. When this occurred, the sultan would have to
depend on the handful of mu'ayyadiya, who were, as has been noted, his
khushdashiya, against the combined forces of these two large groups. But
he could not even rely entirely on the for he suspected one of
their leaders of coveting the throne. In such a situation, the sultan might
very well remain without any supporters at all, for the mu'ayyadiya were the
khushdashiya of the contender as well, and thus quite liable to turn their
backs on Khushqadam and support his rival. Such. instability, Ibn
Taghribirdi concludes, shook to the very foundations the Mamluk kingdom
at that time.1
In spite of the disunity and antagonism which reigned among the it
would be unjustified not to consider them a single unit. Their ill-treatment
at the hands of the mushtarawat and the ruling sultan forced the various
of the into each. other's arms, and they were often moved to
take united action against their common oppressors.

The Sayfiya

It has been pointed out above that the sayfiya were those mamluks who
passed from the service of the amirs to that of the sultan, because of their
master's death or dismissal.1 They may also be said to be, as another
definition has it, Amirs' Mamluks serving at the bab as-silsila.2 That they
were of the Royal Mamluks is clear from the expression al-mamalik
as-sayfiya,3 and, in addition, we are told of a large number of
mamluks who served the amirs but were made Royal Mamluks after they
had been transferred to the sultan's service.4
Of course, the sayfiya, each of whom had served under a different master
before he was thrown together with the others into a single corps, were
indifferent and strange to one another, to the mushtarawat, to the and
to the sultan. They lacked the essential unifying factor of loyalty to the
khushdash on the one hand and to the ustadh on the other. Their allegiance
to any particular sultan was lax and vacillating. The words of the historian
offer a succinct characterization: 'They are as nothing, for they generally
follow the majority; none of them is tied to any particular sultan (wa-la
yaktarith bi-'aynihi) but they serve whoever happens to
ascend the throne much. in the manner of the popular dictum: 'Whosoever
marries my mother, to him I cry: "O my father" (kull man tazawwaj bi-
ummi, lahu ya abi).'5
Al-Mu'ayyad Shaykh, who did not strictly adhere to mamluk criteria of
respectability, but paid greater heed to military talent, introduced into his
service large numbers of sayfiya, claiming that they were time-tested and
battle-tried veterans.6
But he was an exception in that respect, as in many others. In general, the
status of the sayfiya was far inferior to that of the other units of Royal
Mamluks (for data on their inferior pay see Appendix B). The sultan sternly
upbraided the members of an expeditionary force for having appointed as
their commander the atabak of Tripoli, a sayfi and a stranger (rajul sayfi
gharib).7 The appointment of Amir Yusuf as governor of Safed aroused the
hostility of the amirs because he was a sayfi.1 On the other hand, it
happened that membership in a weak body such as the sayfiya was of
assistance in being promoted. Thus, for instance, the appointment of
Barsbay as governor of Damascus suited the sultan better than that of
Birdibak, for Barsbay was a member of the sayfiya and a stranger (rajul
sayfi gharib) from whom nothing need be feared, while Birdibak was
among the leaders of the Barquq.2 But though a few individuals
could indirectly benefit from the fact that they belonged to the sayfiya, the
group as a whole was treated, especially towards the end of the Mamluk
period, with the utmost harshness and cruelty, as can be seen from the
following event: In Sha'ban 903 the plague which burst out in Egypt
some months earlier, reached its peak. On the 20th of that month the julban
caused considerable trouble and mischief in the citadel, and smashed some
of the amirs' saddles, saying that while the plague carried off most of them
(akhadha ghalibahum) only very few of the amirs and the mamalik sayfiya,
were affected by it. So the julban declared: 'If the sayfiya are not afflicted
by the plague, we shall kill them by the sword (idha lam as-sayfiya
naqtuluhum bis-sayf).'3 This threat of the Julban bore immediate fruit. The
sultan decided to diminish the number of the sayfiya by transferring a
considerable part of them back to the amirs. The atabak al-'asakir alone had
to receive 160 sayfiya.4 The sayfiya had hardly time to recover from this
blow when another followed: many of them were forced to take part in a
prolonged campaign in Upper Egypt.5
This attitude of the julban seems to be, on the face of it, rather odd and
unreasonable; but in reality it was based on very sound grounds. The
plagues, which visited Egypt much more frequently during the Circassian
period than during the period which preceded it, played a very important
role in the struggle for power of the various Mamluk units. The julban, who
were comparative newcomers to Egypt, and therefore less immune, suffered
during a plague far heavier losses than the experienced units. Sometimes
more than a third or a half of them would be wiped out.6 The whole
numerical proportion between the various Mamluk factions will thus be
transformed almost overnight to the detriment of the julban. Naturally they
and their master, the ruling sultan, would endeavour to mitigate the effects
of the calamity. Among other means they would try to weaken their
opponents. Of the two rival factions ( sayfiya) they would more
readily turn to the latter who were weaker and whose members lacked the
feeling of mutual solidarity; and this is what they actually did during the
above-mentioned plague.
The sayfiya were extremely weak as a political unit, and seldom acted
independently to improve their position.1 They usually made common cause
with. the mamluks of former sultans in their struggle for a better status.2

FIGURES ON THE ROYAL (AND OTHER UNITS)

We are in possession of fairly abundant contemporary information on the

numbers of mamluks serving in the kingdom. Most of it is, however,
restricted to a single unit of the Royal Mamluks, viz. the mushtarawat. Data
as to the total number of Royal Mamluks are scanty, while no accurate
picture can be formed as to the numbers of the Amirs' Mamluks or their
alterations from one period to another, except for the fact that they were
greatly reduced in the Circassian period (see section 'The Mamluks of the
Amirs' in Part II of this article). The sources are sometimes not clear as to
whether figures cited include non-mamluk soldiery or refer to mamluks
only. So far as is known to the writer, only al-Maqrizi gives a list containing
the numbers of the mamluks from the inception of the Mamluk kingdom
until his time.3 This list is very incomplete, and should be supplemented by
many additions and corrections based on the chronicles and other sources,
as we shall attempt to do below.
In addition, Mamluk sources have handed down two other lists, giving the
numbers of mamluks in service during the reign of a particular sultan. The
first of these, also cited by al-Maqrizi, gives figures on the new organization
and composition of the army encamped in Egypt, which resulted from the
rawk carried out by Sultan b.Qalaun in A.H. 715.4
The second list, of unknown date, includes army figures for the whole
Mamluk kingdom and is cited by Khalil b.Shahin the numbers
seem greatly out of proportion. 5

Below will be found the numbers of the mamluks from the rise of the
Mamluk state until its fall.
Of Aybak, the first Mamluk sultan, Ibn Taghribirdi says that he had soldiers,
mamluks, and retinue exceeding by several times those of the sultans of the
historian's own time, despite the fact that the latter ruled over a much
greater territory. But Ibn Taghribirdi cites no figures whatever.6 As to the
army of Baybars al-Bunduqdari, there is great disparity among the various
accounts. He had 12,000 troops according to one version, 16,000 according
to another,7 but a third claims that, whereas the armies of the later Ayyubid
sultans al-Kamil and his son Ayyub numbered 10,000
soldiers, Baybars raised an army four times as large, which was also given
better equipment, better clothes, and much larger pay.1 This would put the
total figure at 40,000, but there is no doubt that this does not refer to
mamluks only, since the same source states in an earlier passage that
Baybars' mamluks numbered only 4,000.2 Qalaun had 12,000 purchased
Turkish. and Mongol mamluks according to one version, 7,000 according to
another; Mamluk sources themselves tend to accept the lower figure.3 The
historian Baybars Qalaun's devoted mamluk, and the most
important authority on the kingdom during that sultan's time, states that the
number of Qalaun's mamluks, comprising all regiments and ranks, was at
the end of his rule only over 6,000, a number which. he considers to be very
high: 'wa-amma man ba'da min al-mamalik
alladhina ishtarahum bi-anfas al-athman 'alayhim
malabis fa-innahum intahu fi akhir dawlatihi ila ma yanif 'an sittat
alaf mamluk arbab jamakiyat wa-umara' wadhawu
rutab fa-minhum al-jamdariya wal-mafarida
4
wal-muqaddamun walburjiya'. A special corps d'élite, the burjiya, created
by the same sultan, numbered 3,700 mamluks. It is claimed that the number
of Qalaun's mamluks, who were highly disciplined and showed unusually
great respect for their lord and master,5 surpassed that of the mamluks of
any preceding sultan.6 The sources make no mention, so far as is known to
the writer, of the number of mamluks owned by al-Ashraf Khalil. Al-
Maqrizi, in the list referred to above, claims that that sultan had 12,000
purchased mamluks,7 but this figure is doubtful, since it is unlikely that
Khalil could have managed, in his three-year reign, to buy such a large
number of mamluks. Moreover, al-Maqrizi himself states elsewhere (in his
) that al-Ashraf aimed at creating an army of 10,000 men,8
indicating that the mamluks he actually owned numbered less than that
figure. Data concerning the number of the mamluks owned by
b.Qalaun are insufficient, despite the statement that he
bought them on a scale previously unknown.1 Here again the only
information available is that of al-Maqrizi's list, which. gives the figure
12,000 once more.2 The repetition of this figure three times as regards the
mamluks of three different sultans is suspicious. It is only for the beginning
of b.Qalaun's third sultanate, during which he gained
independence, that a reliable account exists. According to this account he
owned in 715, during ar-rawk a total of 2,000 mamluks.3 But of
course this figure in no way reflects the situation during that sultan's rule,
for two reasons: on the one hand, he was unable to buy as many mamluks
as he wished before his third sultanate, hemmed in as he was by the
supervision of his rivals, the amirs Baybars and Salar, who used their best
efforts to restrict his power,4 and, on the other hand, his third sultanate
extended over as long a period as twenty-six years. It was during that
period, in which he was fully independent, that he bought the greater part of
his mamluks. In addition to the explicit statement cited elsewhere5
regarding that sultan's large-scale buying of mamluks, it is important to
recall that he increased the fertility of Egypt so that, as the historian clearly
states, he was able to cover the expenditure required by his large army. He
augmented Egypt's fertile soil area by one half: 'zadat miqdar '.6 (See
pp. 225–7 the intimate connexion between the numerical strength of the
Army, the number of available fiefs, and the economic situation of the
realm.) In the interval between the death of in 741 and the
first years of the reign of al-Ashraf Sha'ban, we have no information as to
the mamluks' numbers. In 769, viz. after four years of rule, Sha'ban had no
more than 200 mamluks.7 This low figure is probably due to the fact that
the sultan did not enjoy independence during the first years of his reign.
The accession to power of the Circassian sultans marks a distinct decrease
in the mamluks' numbers. It is said of Barquq, the first to rely on mamluks
from this racial stock, that he bought many mamluks as soon as he came to
power, and that he owned 3,000 of them within a few years.8 On the other
hand, it is also reported that during the whole of his first reign he bought no
more than 2,000, excluding those he favoured and promoted from among
the great amirs and the who were his khushdashiya.1 As for
figures covering the total duration of his rule, there are diverse accounts:
according to Ibn Tahgribirdi's an-Nujum az-Zahira, he had 5,000
mushtarawat,2 while the same author's al-Manhal credits him with
the same number of mushtarawat and mustakhdamun combined3; whereas
according to al-Maqrizi, the combined total of his mushtarawat and
mustakhdamun was 4,000.4 In the days of Barquq complaints were voiced
as to the 'depleted army of Islam' and the amirs, having consulted with.
each. other as to the best cure for this ill, came to the conclusion that waqfs
dating from b.Qalaun's time or later should be dissolved and
handed over to the army. This was done, in spite of the vigorous opposition
of the clerics. These drastic measures were taken under the pressure of an
impending offensive of Timur Lang against the Mamluk kingdom.5 The
same claim is repeated under Barquq's son, Faraj: the Mamluk armies have
been reduced because of the increase of waqfs, which. must be dissolved so
that it may be possible to re-hire the idle soldiery (al-ajnad ).6
For the reign of al-Mu'ayyad Shaykh, the total number of Royal Mamluks
may be computed, but we have no information concerning the number of
the mushtarawat. In 820 he disbursed 8,000 dinars for the clothes (kiswa) of
the Royal Mamluks7; every mamluk then received 500 dirhams for the
kiswa, so that we arrive at the figure of 5,500–5,700 for the aggregate of the
Royal Mamluks.8 That this tallies with the facts may be judged from the
report that in 824 the sultan, at a parade attended by the majority of the
Royal Mamluks, gave out the pay of 4,000 soldiers.9 The extent to which.
the purchasing of mamluks was reduced during the Circassian period may
be inferred also from the accounts concerning the mamluks of al-Ashraf
Barsbay. It is related that he was addicted to buying large numbers of
mamluks, that in this he emulated Barquq, and that, had it not been for the
plague he would have owned more than 2,000 (!).10 In other words, he
owned even less than that small number. What is even more surprising is
Ibn Taghribirdi's statement that until his own time the mamluks of Barsbay
formed the bulk of the Mamluk forces ('wa-ila al-an mamalikuhu hum
 'askar al-islam').11 This statement is even repeated in his chronicle
for the year 868, viz. twenty-seven years after Barsbay's death.12 To this
must be added the account of al-Maqrizi, who lived under Barsbay,
concerning the strength of the Mamluk army in his days. According to that
author, the troops of the and the Royal Mamluks, if combined, would
amount to no more than 5,000, of which 1,000 or less would be suitable for
combat.1
This sombre picture of the numerical strength of the Mamluk army during
Barsbay's reign seems, however, to be somewhat mitigated by the following
information furnished by Ibn 'Arabshah in his biography of sultan Jaqmaq.
According to this historian 120,000 dinars were distributed among the
Royal Mamluks during one pay parade in 841, the year of Barsbay's death,
and every mamluk received 30 dinars (wa-fi hadha al-yawm unfiqa fi al-
mamalik kull mablagh thalathin fa-kana jumlatuha
mi'at wa-'ishrin alf 2
). This implies that the number of the Royal
Mamlnks in 841 was about 4,000, a figure which is considerably smaller,
indeed, than that of 824, but still it shows that the decline of the numerical
strength of the Mamluk army was not, perhaps, as precipitate as might be
inferred from the overwhelming evidence supplied by other Mamluk
sources.
It is interesting to note the causes to which Ibn Taghribirdi attributes the
numerical decline of the army during the Circassian period. He relates that
when Barquq took power by force of arms, the amirs began buying
up the fiefs for themselves or their mamluks. Not stopping at this,
they obtained from the sultan a monthly salary (jamakiya) for their
mamluks. In this manner, every one of the latter became at the same time a
soldier of the a Royal Mamluk, and an amir's mamluk, so that the
earnings of three persons were pocketed by a single individual. Hence the
decline of the Egyptian forces, 'now three times smaller than formerly'.
There was an additional factor which, in the historian's view, greatly
contributed to the reduction of the army, viz. the loss of many military fiefs
as a result of their being transformed into rizaq (estates granted as pensions)
or into amlak (allodial lands). The number of these was 'enormous and
beyond all bounds'. Ibn Taghribirdi concludes as follows: 'Whoever gives
careful consideration to what we have said will understand the original and
the present state of the Egyptian army. Were it not for the factors
enumerated above, and were it not for the destruction which has befallen
some of the regions of the realm, because of continuous oppression,
increased taxation, and the rulers' neglect of the country's welfare, there
would be no adversary capable of resisting the Egyptian forces, and no
army worthy of comparison with them.'3
Among the factors listed above by Ibn Taghribirdi to account for the
numerical decline of the Mamluk army, two are of prime importance. The
first is the reduction in the number of feudal estates (cf. above the
endeavours of Barquq and his son Faraj to increase it by the dissolution of
the waqfs); and the second was the decline of Egyptian economy.4 On the
other hand, the historian greatly overestimates the role played by the Amirs'
Mamluks, for, as we shall see in Part II of this article, these were far from
occupying the important position which he attributes to them. But, as to the
very fact of the numerical decline of the Mamluk army under the
Circassians, Ibn Taghribirdi speaks in unequivocal language, and his
statements are borne out by all the data that can be gathered on this subject
from Mamluk sources.
We have not been able to obtain any infomation on the number of Jaqmaq's
mushtarawat. Aynal had, at the time of his death, about 1,000 mamluks 'or a
little less or a little more ', not including 200 which he had bought from his
predecessor.1 A most important piece of information—the only one of this
type known to the writer—is available regarding all types of Royal
Mamluks in the services of Sultan Khushqadam: The Faraj
constituted an unimportant quantity; the Shaykh numbered 30,
most of them occupying high posts; many of the ashrafiya Barsbay were
Amirs of and Amirs of Ten, and many were the
Jaqmaq numbered more than 600, including five Amirs of a Hundred; the
ashrafiya Aynal counted some 1,600 (!); Khushqadam himself had 3,000
mushtarawat, of whom 400 were kuttabiya and the rest and
holders of offices.2 Ibn Iyas claims that Khushqadam owned 4,000
mamluks,3 but this author shows a general tendency to exaggerate in
comparison with preceding historians (thus he claims, for example, that the
numbers of the mushtamwat of Barquq, excluding were 7,000,4
of al-Mu'ayyad Shaykh, 5,000,5 of Barsbay, 5,0006). It is related of Qaytbay
that he was fond of buying mamluks, and that, had it not been for the
plague, the number of his mamluks would have reached 8,000.7 It should be
borne in mind that that sultan's reign was very long, 873–901 (1468–1495)
so that he was able to buy a larger number of mamluks than the other
Circassian sultans. According to his biographer Qaytbay had at the end of
877, i.e. about five years after his accession to the throne, more than 2,000
mamluks (wa-ishtara min al-mamalik ma 'ala alfay mamluk).8 As for
we learn that when he was besieged in the Citadel by
the Dawadar, all his supporters deserted him, except his own mamluks
(mamalikuhu; mamalikuhu mushtarahu) who numbered less than 2,000.9
This figure seems to be rather exaggerated at the first glance for a sultan
who ruled only a few months. But it should be remembered that
was sultan Qaytbay's brother-in-law and one of his favourite
amirs, and that during the reign of Qaytbay's son, Abu as-
Sa'adat, he was the dominating figure in the Mamluk kingdom. Under such
exceptionally favourable circumstances he could easily purchase most of
his mamluks before he even came to the throne. The number of the
mamluks of al-Ghawri is not known, but those who completed their
training at the military school during his reign numbered about 2,500; this
figure is lower than the number of the sultan's purchased mamluks, as we
have attempted to show elsewhere.1
As for figures including all categories of Royal Mamluks, three, relating to
the Circassian period, have been cited above: for the reign of al-Mu'ayyad
Shaykh, 5,500–5,700; for the reign of Barsbay, about 4,000; and for the
reign of Khushqadam, approximately 5,500. We may add that
in the first half of the 14th century, built barracks which
2
could accommodate 12,000 Royal Mamluks. In his detailed list of the
kingdom's army (see Appendix A) puts the number of Royal
Mamluks at 10,000. G.de Lannoy, who visited the Mamluk kingdom at the
beginning of the 15th century, also estimates the Royal Mamluks at
10,000.3
It may be pointed out that we have nowhere found any instance in which
the sources mention the numbers of the sayfiya.
The above list concerning the numbers of the Royal Mamluks is, of course,
far from being full; but it should be emphasized here that it is doubtful
whether a similar list can be compiled for many Moslem armies in the past,
with the exception of that of the Ottoman Empire in its later stages.
Moreover, the above list covers the greatest part of the Mamluk period, and
is, most probably, quite accurate as far as it goes. This can be judged by the
smallness of the figures and, what is more important, by the fact that these
figures generally tally with the information furnished by the sources
concerning the numerical strength of the Mamluk military expeditions and
the numbers of mamluks present in pay parades and in other general
parades.4
Studies on the Structure of the Mamluk Army—II

By DAVID AYALON
 The 1

THE term as a name of a military unit, seems to be mentioned for the

first time in 1174, when Turanshah set out, under ad-Din's orders, on
2
his expedition to the Yemen. This unit is also mentioned a few times
during the siege of Acre in 587/1191.3 The sources do not indicate the date
of its founding, and no authoritative explanation of the meaning of its name
is available. Two opinions as to the latter may be submitted with all due
reservations: Quatremère thinks that the was so called because it was a
corps which surrounded the sultan and constituted his bodyguard,4 and
indeed the sources' description of its position and part in combat support the
impression that it was composed of the élite of ad-Din's forces5;
A.N.Poliak disagrees, and holds that the name is derived from the special
tactics the Turkish peoples used to employ in attack, i.e. that of surrounding
the enemy in the form of a ring Poliak supports his view by alluding
to the frequent recurrence of these tactics in the combat manœuvres
described in the furusiya literature and in the hunting expeditions which the
Ayyubid and Mamluk sultans frequently organized.6 In the present state of
our knowledge, both explanations must be viewed as hypothetical, since no
data are available to support either the one or the other; the writer inclines
to favour Quatremère's view.
The underwent a number of transformations over the years, so that
descriptions of this unit by Mamluk sources are incomplete and imprecise.
Qalqashandi says that its soldiers are very numerous, and that it often
admits to its ranks non-military people, such as civil functionaries and
others.7 This description fits the author's own time (end of the 14th and
beginning of the 15th centuries), but can apply neither to the beginning nor
to the end of the history. In order to gain an understanding of the
structure and status of the at its various stages, one must turn to the
chronicles, which supply abundant information on this subject.
As has been stated, the had its most honoured standing under ad-
Din. At the siege of Acre, it occupies a privileged position and escorts the
sultan under the name of or 1Its development

during the Ayyubid period after ad-Din has not been systematically
examined by the writer.2 As for the Mamluk era, its early years saw the
preserve its power and lofty position. It occupied an honoured place at the
various official ceremonies, side by side with the then the élite of the
Mamluk army, and was sometimes mentioned before that unit.3 We find, at
that period, frequent recurrence of the expression 'victorious '
4 a title which gradually disappears at a later period.5 Its
commanders, called muqaddamu were holders of honoured
positions at the beginning of the Mamluk era, and their names appear side
by side with those of the amirs in the most important ceremonies, including
those of oath-taking and coronation. At the allocation of fiefs in 712 they
received their grants even before the Royal Mamluks,6 a procedure which.
would have appeared quite incongruous at a later period. Speaking of the
and its prominent position in former times, al-Maqrizi says that even a
simple trooper would go out with a string of horses and that the
muqaddam would appear as an Amir of Ten: 'wa-kan al-jundi kharaj
ila as-sukkan khayl wa-yakhruj muqaddam ka-amir
'ashara'.7 But what seems most significant is that among the soldiers of the
and especially among their muqaddamun, one finds, at this early
period, not only local citizens and sons of mamluks, but also mamluk
soldiers and even amirs. In 712 b.Qalaun reviewed the
Royal Mamluks and transferred some of them to the 8 That mamluks
were sometimes members of the is apparent from the following
examples: Muqbil b.'Abdallah. was of the 9;

b.'Abd-allah al-Jawili received from Amir Tankiz, governor of

Damascus, an in the 10; the village of was portioned out
to ten of the ajnad 11
including Sunqur as-Sa'di. Mamluk amirs, some
of rank, then served in the some of them as muqaddamun.12
Thus for instance the amirs Bayram Quja,13 Qaraqush al-Kawunduki,14
Lajin 15 Ghurlu al-Baktimuri,16 Jarkas as-Sayfi
Mankalibugha,17 Aybak 18 Baghdi al-Ashrafi,19 were of the

The muqaddamu
. also served as envoys to important states, and as escorts for foreign
envoys on their way home from Egypt.1 These posts, as pointed out above,
were usually reserved for the It must be stressed, however, that
the pay of the muqaddamu even at that early period, was much lower
than that of the amirs. In one case, the grand amir receives from the sultan
10,000 dirhams or less, the Amirs of Ten, 1,000 dirhams, and the
500 dirhams.2 In another instance, Amirs of are
paid some 5,000 dirhams, Amirs of Ten, 1,000 dirhams, and the
muqaddamu 500 dirhams.3 It also occurred that besides Royal
Mamluks, amirs' mamluks were also included in the 4 In 678 it is
reported of one individual that he was the first of the amirs' mamluks to
receive an in the 5

The Turkish and Mongol tribesmen who entered the Mamluk kingdom in
quest of asylum, al-wafidiya, and most of whom were incorporated in the
have been discussed elsewhere by the present writer.6
The commanders of the were, at least in theory, distributed in
proportion to the rank and file as follows: the amir mi'a alf
commanded 1,000 troops; the bash and the naqib, whose exact role is
difficult to ascertain from the sources, were in command of 100 troops; the
muqaddam commanded 40 troops, but his authority was restricted to
actual military expeditions, and lapsed as soon as the expedition was over.7
It is difficult to determine how long this arrangement was carried out in
practice, but it is certain that during the greater part of the Circassian
period, this chain of command had but a paper existence. The had by
that time become greatly reduced and impoverished, and took virtually no
part in combat.
troops of various ranks who were under the command of a particular
amir in combat were called but this appellation was not restricted to
them, and applied to mamluks of amirs under the same commander as
well.1
The members of the were generally called ajnad 2 sometimes

rijal 3 and sometimes simply ajnad, while mamluk troops were

called mamalik. Such expressions as 'al-umara' wa-l-mamalik wa-ajnad
', 'al-mamalik ', 'al-umara'-wa-l-mamatik-wa-l-ajad', are
extremely frequent. 4

BEGINNINGS OF THE DECLINE OF THE

Until the reign of b.Qalaun, we find no clear indications of

the decline of the During the reign of his father, Qalaun, we still hear
of 4,000 soldiers participating in the war against the Mongols in 680 as
élite troops fighting in the centre (qalb) of the front; the number of Royal
Mamluks fighting in the centre was 800 only.5
The first conspicuous sign of a turn for the worse in the status of the
appears during the land redistributions (rawk) conducted in the Mamluk
kingdom at the end of the 7th and the beginning of the 8th centuries of the
Hijra, and accompanied by cadastral surveys. The rawks caused a profound
change in the structure of the Mamluk army, serving especially to reinforce
the position of the Royal Mamluks and to deal a heavy blow to that of the

It does not fall within our purpose to describe here the various rawks and
their importance for the feudal and economic structure of the Mamluk state;
this has already been thoroughly done by A.N.Poliak.6 We merely propose
to indicate the effect the rawks had on the Three such land re-
allotments took place in the Mamluk kingdom within a relatively short
period: the first was ar-rawk in 697, the second ar-rawk in
Syria, Lebanon, and Palestine (al-bilad ash-shamiya), in 713, and the third
ar-rawk in Egypt, in 715. Poliak rightly remarks: 'The aim of these
re-allotments was to render the feudal landlords more and more dependent
upon the central government'. At the beginning of the Mamluk era, the
Mamluk feudal system was still under the influence of the Ayyubid system,
as well as under that of the Latin crusader states, in which the fief was
handed over as inheritance from father to son. The rawk was the means
used by the Mamluk sultans to eradicate the hereditary character of feudal
grants.
The enormous changes which the rawks brought about in the structure of
the army are apparent even from a mere comparison of the lists of feudal
grants during and before the rawks. Before the land surveys, only four of
the 24 units into which Egypt was divided belonged to the sultan and
to the Royal Mamluks; ten units went to the and the remaining ten to
the amirs. Thus by far the greater part of Egyptian lands was equally
divided between the amirs and the while the sultan and his élite corps,
the Royal Mamluks, received a mere sixth of the fiefs. In the rawk
four units were allotted to the sultan alone, nine to the Royal Mamluks (in
payment of the feudal revenue and the monthly salary), and only eleven
units to the amirs and the together. In the rawk the sultan
received ten units, and only fourteen were distributed as fiefs.1
These figures alone clearly show that the aggrandizement of the Royal
Mamluks was carried out at the expense of the amirs and the Indeed,
the sources proclaim the same fact in unequivocal language. We are told by
one of them that' it was an incomparably foul act, and it was the cause of
the weakness of the Egyptian army, especially of the as will be shown
2
later'. Another source points to the rawk as a prime factor in the
enfeeblement of the Egyptian army; it had, moreover, no compensating
advantages, according to that source: on the one hand, no one obtained a
satisfactory quantity, and on the other, the many fiefs that were saved by
means of the rawk were all distributed after the murder of Sultan Lajin.3
The same source holds that, as a result of the rawk the position of
the deteriorated as compared with the reign of Qalaun:under that
sultan, the minimum and maximum incomes from fiefs were 10,000
and 30,000 dirhams respectively, whereas after the rawk they were
5,000 and 20,000 dirhams respectively. The members of the received
their letters of enfeoffment with. marked dissatisfaction, as did the amirs.
Some of the more intrepid amirs threw their letters back at
Mankutimur, Lajin's mamluk and right-hand man, and demanded to be
transferred to the amirs' service, or to be entirely released from military
service.1
The rawk similarly dealt a heavy blow to the army, and aroused
much resentment.2 The land distribution provided for by that rawk was still
in force in the days of the historian Ibn Taghribirdi.3
The great importance of the above information lies, first, in that it
constitutes the earliest clear evidence of the decline of the and the
corresponding rise of the Royal Mamluks at its expense; and second, in that
it shows that the repression of the was carried on systematically and on
a state-wide scale. From that time on, the decline of the proceeds
apace, and degrading restrictions become a matter of common occurrence.
A detailed description of the various stages of this down-grade process after
and of the progressive inclusion of foreign elements in
the is given by al-Maqrizi, who cites exact dates. He states that after
the death of it became usual for members of the to
exchange their feudal estates against payment or compensation
with the result that many foreign elements entered its midst (fa-kathura
addakhil fi al-ajnad bi-dhalik). Pedlars and common people (as-suqa wa-l-
'amma) bought up estates, to such. an extent that in the days of al-Maqrizi
the great majority of the was composed of artisans. These upstarts
caused the rapid ruin of their newly acquired estates. The first to introduce
this exchange system was Sultan al-Kamil Sha'ban b. b.Qalaun,
who came to power in Rabi' ath-Thani 747, and was under the influence of
Amir Shuja' ad-Din Ghurlu, the superintendent of chanceries (shadd ad-
dawawin), and who established for that purpose, in Jumada al-Ula of the
same year, a special department called diwan al-badal.4 Through this
department, a member of the could give up his estate in exchange for a
sum of money, and another person could obtain the estate in exchange for
another sum. This system was abolished as a result of the amirs'
intervention with Sultan Sha'ban. It was restored, however, when Amir
Manjak al-Yusufi became wazir in 749. Members of the would sell
their fiefs for sums ranging up to 20,000 dirhams, according to the size of
the and the wazir would collect a special tax on every sale. These sales
were again abolished, and again reinstated in 753, with the appointment of
Amir Sayf ad-Din Qila (?) as vice-sultan (na'ib ). Corruption was
rife, and estates, including those of the muqaddamun, were bought up
by the shopkeepers and the rabble. A special group of 300 was even
organized, whose members, called al-mahisiyin, would visit the members of
the and persuade them to sell their fiefs. Such activities did much to
weaken the and caused its disintegration. When this trading in feudal
estates had reached dangerous proportions, it was stopped by Amir
Shaykhun al-'Umari while he served as chief of a corps of mamluks (ra's
nawba). He forbade officials of the diwan al-jaysh to collect a tax of more
than 3 dirhams for issuing letters of enfeoffment, whereas the former charge
had been 20 dirhams.1
This account of al-Maqrizi's, which is substantiated by Ibn al-
2
'Asqalani in his biographical dictionary, ad-Durar al-Kamina, is couched
in unequivocal terms, and is indeed confirmed by various other sources.
The disintegration of the reached its peak as early as the beginning of
the Circassian period. In 791, the year of the well-known wars of
'succession' to the sultanate, the sources abound in data on the It is
then already considered, for the most part, unsuited for combat. Amir
says to the 'You are weak and unfit to take part in the
expedition'.3 He selects only the best of them.4 None but holders of fiefs
with annual incomes of 4,000 dirhams or over go into battle,5 and even they
go forth each according to his means, some on foot and some on
horseback.6 It would seem that as early as 791, the principal task of the
was the guarding of vital places in Cairo during the absence of the fighting
forces: some of its members go into battle, some remain behind to guard the
citadel (qal'at al-jabal), some to guard the gates of Cairo, and some the old
city and its suburbs.7 The guarding of Cairo in the absence of the main
force was the chief duty of the during the whole Circassian period.8
Al-Mu'ayyad Shaykh, who reigned from 815 to 824 A.H., attempted to
reorganize the selection of the for participation in military
undertakings. In one of his expeditions, he employed the following twofold
system. First, he combined low income fief holders with high. income fief
holders, viz. those holding fiefs with 3,000 dirham yearly income had to
hand them over to holders of fiefs with. 7,000 dirham yearly income, so that
the latter might take part in the expedition. Second, he combined four
holders of small estates, who were to finance the participation of one of
them, according to their choice. A total of some 400 members of the
rich and poor, great and small, were involved in this arrangement,1 so that
those who actually took part in the expedition were much. fewer than that
number, itself extremely low.
We have pointed out elsewhere in our work on the Mamluk army that all the
reforms carried out by al-Mu'ayyad Shaykh. were short-lived and were
virtually abolished by his son. This holds true of the reform as well. As
early as 832, when Sultan Barsbay organized an expedition against Shah
Rukh, the was sent forth. without any manner of order or control. He
ordered a review in the Royal Square at which the aged,
the young, and the blind were present, and told them that he would not do
as al-Mu'ayyad Shaykh had done, but would require every man to go out:
'Whoever has a horse, let him go to battle on horseback, and whoever has
an ass, let him ride forth on an ass'. 2 Ibn Taghribirdi considers this one of
Barsbay's grave errors. He attributes it to the fault of his dawadar, who was
inexperienced, and cites in contrast numbers of dawadars who handled the
in an appropriate manner.3 The connexion between the dawadar and
the seems to be that it was their function to handle the feudal charters
(manashir) issued by the sultan.4 It is reported of the that they received
their manashir from the sultan, as did the amirs. 5

From then on, the is on a steady down-grade. The very term is

gradually replaced by the term 'awlad an-nas' as will be seen below. 6
What were the reasons for the decline of the ? The principal one was,
of course, that its members were not mamluks, and could in no way
compete with. the military ability of the mamluks, who had been steppe or
mountain dwellers accustomed to the rigours of war. Moreover, it is very
doubtful whether in the Mamluk régime, which opened the gates of the
highest military society only to freed slaves, the would have been able
to hold its ground for any length. of time, even had its members been
endowed with the finest military talents. Another major factor was Egypt's
economic situation: the country being unable to bear the huge cost of
maintaining its armies, reductions in effectives and salaries were
unavoidable. In the event of such curtailments, of which. Mamluk history
shows repeated instances, the was the first victim. Further, beginning
with the reign of b.Qalaun, the slackening of the westward
migration of peoples from the Eurasian steppe completely deprived the
of a very important element, viz. the wafidiya, many of whom were
incorporated in the and the amirs' troops. These were tribesmen of
Turkish, Tatar, or related stocks, and were incomparably superior in warlike
qualities to the population of the Mamluk kingdom. The had an
additional disadvantage in that it lacked the feeling of solidarity linking
companions in slavery and freedom (al-khushdashiya) which constituted
the very foundation of the Mamluk army's structure; its struggle for survival
could in no way have been as grimly determined as that of the mamluk
regiments. The was, in fact, gradually relegated to the side-lines
without any resistance on its part. Except for the dissatisfaction expressed
by its members during the rawk and the threat voiced by some of
them to leave the military service, it seems that there was not any serious
attempt to resist the increasing repression to which the corps was submitted,
or any determined effort to hold back the disintegration of its political and
military importance.

SPECIAL GROUPS WITHIN THE

1. The Awlad an-Nas

There was within the a special unit to which belonged the sons of the
amirs and mamluks. These sons were born and bred in Islam; their great
majority bore Arabic names, and the proportion of theologians among them
was fairly high. In as far as they joined the army they were automatically
cast off from the pure mamluk corps and assigned to the a much lower
unit, in which they formed socially the most select element. They were
known as awlad an-nas 'children of the people', i.e.'of the best people, of
the gentry', for the 'people' were the mamluks, the members of the ruling
class.1
The entry of the sons of the amirs into this unit took place in the following
manner. When the son of an amir came of age, his father would provide him
with 'pay, foodstuffs, meat, and fodder' until he was old enough. to receive a
fief in the Among these amirs' sons, some would ascend to the rank of
Amir of Ten, some to that of Amir of Forty, 'all according to chance'.2
According to al-Maqrizi, Qalaun was extremely strict in his insistence that
no amir's son should enter the before his majority; he acted thus with
his own sons as well, even if they had taken part in combat.3 The historian's
claim that the awlad an-nas attained no higher rank than that of Amir of
Ten or Amir of Forty is accurate on the whole. Nevertheless we find some
members of this unit (especially in al-bilad ash-shamiya, and to a lesser
extent in Egypt) who reached the rank of Amir of a Thousand. That rank
was also reached by some members of the who were not awlad an-
1
nas. The awlad an-nas were sometimes favoured for political reasons: to
reduce the power of the mamluk amirs. Thus Sultan preferred amirs
from the awlad an-nas to mamluk amirs; during his reign, most of the
governors of fortresses (nuwwab al-qila') of the bilad ash-shamiya were of
their number, and it is the chronicler's view that this is the reason why no
rebellion took place in these provinces. In Egypt, eight of the Amirs of a
Thousand were, during that sultan's reign, from the awlad an-nas; these,
together with the sons of the sultan, numbered ten out of a total of 24 Amirs
of a Thousand.2 The privileged position of the awlad an-nas under Sultan
was, however, exceptional, and contrasted sharply with their status
under other rulers. Since theirs was an element which, by its very nature,
was excluded from the ranks of the mamluks, their chances for
advancement and for attaining key positions were seriously handicapped. In
the course of time they declined together with the and saw the same
restrictions applied to them as to the rest of that body, viz. reductions in pay,
sale of their fiefs, exemption from military expeditions in exchange for cash
payment (badil), tests in the use of the how and arrow designed to prove
that they were badly trained and thus not entitled to all the privileges of
full-fledged soldiers.3
Toward the end of the Mamluk era, the name slowly falls into disuse,
while that of the awlad an-nas becomes extremely common. One still,
though rarely, encounters the term 'awlad an-nas min ajnad ',4 but the
name ajnad as such. virtually ceases to exist separately. Nor is that
all: at that period the awlad an-nas are even called Royal Mamluks, and the
expression 'awlad an-nas min al-mamalik ' is fairly frequent.5
We have even found once the term 'awlad an-nas '.1 This should
in no way be taken to indicate an improvement in the status of that unit, for
the sources clearly point to the exact opposite. It is possible, therefore, that
the term mamalik during that period occasionally took on a wider
meaning, as a result of the incorporation of additional units to the Royal
Mamluks into the diwan al-mufrad.
There was, both among the awlad an-nas and the other members of the
a strong leaning toward piety and preoccupation with. other-worldly
affairs. Many of them left the military service and became theologians
(faqihs).2

2. The Sons of the Sultans

The sons of the sultans, who were also included in the awlad an-nas unit of
the 3 constituted its most respected element, and were called al-asyad

or awlad al-muluk. Each. one was addressed as 'sayyidi'. The reigning

sultan would treat the sons of the former ruler with strict severity, precisely
because of their importance and of the fear that they might be used by his
adversaries. Until the reign of Barsbay, most of them were restricted to their
quarters in the Cairo citadel; many had never even seen Cairo and had little
notion of what a city looked like. They were ordered by Barsbay (c. 825) to
come down from the citadel and to take up residence in the city. The
tumultuous life of the metropolis soon corrupted them; many became
impoverished, and all their former splendour left them.4 The most eminent
and most dangerous among them, especially those who had discharged the
office of sultan after their father's death, were usually sent to the Alexandria
prison.5 A sultan might sometimes display his magnanimity by bringing one
of them to Cairo, regaling him with sumptuous feasts and even permitting
him to make the pilgrimage to Mecca6; but these were isolated instances
which imposed no obligations on the sultan, and the status of the sultans'
sons remained unimproved until the end of the Mamluk era.7

3. The al-Mi'atayn

Among the units of the Egyptian which were not stationed in Cairo
one is worthy of special mention: it was called ajnad al-mi'atayn, and was
stationed at Alexandria. Its history dates back to the Frankish raid on.
Alexandria in 767, which caused a number of changes in the defences of the
city. Among these was the posting of a garrison of 200 men of the
called ajnad al-mi'atayn.1 In the days of the historian Khalil b.Shahin
their numbers had increased, but the name was preserved; they
were then 360, headed by twelve commanders each. in charge of 30 men.2
 The Mamluks of the Amirs

The mamluks of the amirs were called mamalik al-umara',3 or ajnad al-
umara'.4 The troops of each. amir were at first registered in the diwan al-
jaysh (q.v. below), but in the days of Qalqashandi this arrangement was
replaced by separate lists prepared by each. of the amirs, who would send
copies to the diwan. The number of mamluks each amir could take into his
service was fixed, and new ones could be introduced only when some of the
original number died or were cashiered. No mamluk could be dismissed by
an amir until the latter had convinced the vice-sultan (na'ib ) that there were
just causes for his dismissal. The amirs' troops received their pay and the
deeds to their estates from their master. According to his feudal charter
(manshur), the amir was to receive one-third of the income of his fief, and
two-thirds were to be distributed among his mamluks. The amir or his
clerks could levy no part of the mamluk's portion for the amir without the
mamluk's consent. It seems probable that this apportionment was not
strictly observed; in the year 767 it even became necessary to enact a law
requiring the amir to share his income equally with his troops, and our
source claims that much good accrued to the army from this procedure.1
Thus we are entitled to assume that before the enactment of this law, the
amir was wont to take for himself more than half the income, though he was
legally entitled only to a third.
Service under an amir was known as bi-abwab al-umara'.2 Such.
service was of course considered much inferior to service under the sultan
(khidma bi-abwab ) and a Royal Mamluk fated for some reason to
serve under an amir was thought ill-starred.3 The amirs' mamluks
constituted no serious political factor in the Mamluk army; rebellions of
great proportions among them were extremely rare.4 In addition, they were
necessarily less well trained than the Royal Mamluks, for they did not have
access to the first-rate military schools in which the latter grew up and
studied. It seems, however, that they attended the military schools of
5
their masters. The political weakness of the amirs' mamluks as a group and
their precarious position, as individuals, even in the highest posts, may be
deduced from the following incident. Qajqar al-Qurdumi, who served as
amir was imprisoned by Sultan with the help of the mu'ayyadiya
and the (Barquq) who were the sultan's 'colleagues' (khushdashiya).
feared that this arrest would cause disturbances, but nothing occurred:
'For Qajqar al-Qurdumi lacked followers, because he was one of the amirs'
mamluks, bereft of power and of colleagues' (wa-dhalika li-'adam
Qajqar al-Qurdumi, fa-innahu mamalik al-umara', laysa lahu shawka
wa-la khushdashin).6
When an amir died or was dismissed, his mamluks passed on to the service
of the sultan or of other amirs, or were divided between the sultan and the
amirs.7 It sometimes also happened that they were joined to the 8 When
an amir was transferred from one province to another, he was almost
invariably unable to take his mamluks with him; when the amir Aljay was
appointed governor of he asked permission to take his mamluks with
him, but this was denied him.1 We know of no instances where such
permission was granted. The same rule was thus apparently applied to the
amir's mamluks as to his estate: he lost both when he was transferred to a
new province.
Little is told us of the mamluk's duties and obligations toward his amir.
Something on this question is learned from the interesting biography of
Amir Sudun (died A.H. 842), whose conduct was the reverse of that
accepted among the amirs. From his unusual behaviour, the more
conventional relations between the amir and his mamluks may be indirectly
inferred. He lived in a barrack in the Cairo citadel even after he had
become a high-ranking amir. When he moved, by order of the sultan, from
the citadel, and settled in town like the rest of the amirs, he continued to act
in his own manner. He would order his mamluks to escort him on horseback
after parades (mawakib) up to the door of his house. He would then arrange
the horses on the left and right of the door, and no one would dismount. He
would leave them and enter his house alone; as Sudun dismounted, he
would be approached by a baba, who served him after the manner of the
for Sudun had neither master of the robe (jamdar) nor armour
bearer He did not serve meals for his mamluks at his house, but
would eat alone. In compensation he allotted each one of his mamluks three
of mutton; and as someone remarked upon this, he replied: 'My
mamluks will reap greater benefit from this arrangement: whoever of them
is married will share his portion with his family. Were he to take his meals
at my table, he would have to incur additional expenses for his family'.
Sudun then gathered his mamluks and asked them whether they would
agree to have their allotments diminished and dine with him instead; but
they replied that they were satisfied with the existing arrangement. (Of
ordinary amirs it is said explicitly that they always took their meals together
with their mamluks.)2 He had in his service 150 mamluks, exclusive of the
kuttabiya, i.e. young mamluks who had not completed their training and
had not received their liberation certificates. He would distribute their pay
and their fodder and meat allotments at the beginning of the month from his
private stocks. They would escort him on horseback on parade days only.3
As for the amir's court, it was a copy on a reduced scale of the court of the
sultan. He had a coat of arms (rank, pl. runuk),4 with a special design
serving as his emblem, such as a cup (hanab), an inkwell (dawat), a napkin
(buqja), a (faransisa), and the like. This coat of arms, which bore
a colour of the amir's choice, was painted on the gates of his house and his
other possessions, such as the grain storehouses, the sugar refineries, the
ships, as well as on his sword, his how, and the caparisons (barkustuwanat)
of his horses and camels.
When the amir rode out of his house, the important members of his corps
would escort him as follows: the ra's nawba, the the amir majlis,
and other high-ranking office holders would precede him; the
would follow him, and the amir akhur would come last, leading the reserve
horses (al-jana'ib). The sultan rode in the same manner.1

THE NUMBERS OF THE AMIRS' MAMLUKS

The contemporary sources are interspersed with. fairly numerous data

concerning the numbers of the mamluks owned by various amirs. These
data pertain, however, mostly to amirs who distinguished themselves by
unusually large-scale acquisitions of mamluks, while virtually nothing at all
has reached us regarding ordinary amirs. The only exception is the very
valuable information we have on the numbers of mamluks owned by the
amirs who took part in the battle of Marj Dabiq.2 There is thus no way of
ascertaining whether or not the amirs actually employed the number of
horsemen to which they were entitled by law (cf. pp. 467–71). A list giving
the numbers of mamluks owned by the most important amirs of the realm
and covering the entire Mamluk era will be found below. This list, together
with. the historians' comments which we shall cite, clearly indicates the
great numerical decrease of the amirs' mamluks in the Circassian period as
compared with. the period.

A. The Period

Qarasunqur had 600 horsemen.3 Sunqur al-Ashqar demanded to be made an

amir disposing of 600 horsemen.4 Asandumur, govenor of Tripoli, had 500
mamluks.5 had 700,6 Shaykhun al-'Umari, 700 7; Ayanbak
assembled, for one of the battles near the Cairo citadel, 200 of his
mamluks.8 There were 1,500 mamluks in the service of Jakam.9 When
Baybars al-Jashnakir was deposed from the sultanate, he asked
for 300 mamluks and obtained only 100.10 The grandest of all amirs with
respect to the numbers of his mamluks was Yalbugha he
11
had, according to one version, 1,500 mamluks, and according to another,
3,000, including four Amirs of a Thousand.12 Ibn Taghribirdi states that the
mamluks of Yalbugha (al-yalbughawiya) constituted at the beginning of the
9th century the overwhelming majority of the Egyptian army, and that the
greatest amirs were from their ranks.13 Manjak had only 75 kuttabiya.14
Nawruz had 1,000 mamluks, each. receiving a monthly pay
(jamakiya) ranging from 10 to 100 dinars; the total of the salaries of his
mamluks and retinue after he had rebelled against the sultan was between
20,000 and 30,000 dinars.1 Mankalibugha had approximately 200.2
had 800.3 In 791 the commander-in-chief (atabak al-'asakir) had
in his (see n. 8, p. 464) 300 mamluks.4 Two hundred out of the atabak
al-'asakir's 500 mamluks take part in the expedition to Upper Egypt.5 There
were 150 mamluks in the of Baybugha and 60 in that of each. of the
Amirs of a Thousand. Kamishbugha had over 300 mamluks.7 Aytamish al-
6

Bajasi, who was commander-in-chief, had 1,000 mamluks.8

B. The Circassian Period

The above-named Aytamish was the last of the amirs who possessed large
numbers of mamluks. Concerning him, the following noteworthy passage is
found in Ibn Taghribirdi: 'He was the last of the mighty amirs of Egypt until
our own days. When my father became commander-in-chief under
Faraj, several people suggested to him that he follow the example of
Aytamish. al-Bajasi. To this he replied,"How small are we in comparison
with those grand seigneurs! (hayhat ma min khayl hadha al-
maydan)". And these words were said by my father when he had in his
service 400 mamluks, and when his daily portion was 1,000 '.9 Thus 400
mamluks in the service of the highest-ranking amir in the days of Ibn
Taghribirdi was considered an unusually high. number. The decline of the
courts of the amirs in the Circassian period is also reflected in such
comments as 'a single ordinary mamluk of Janibak's had better and more
suitable table and furnishings ( wa-barak) than many of our present-day
10
Amirs of a Thonsand '. The same Janibak is said to have followed in the
footsteps of the former kings (al-muluk as-salifa)11 in the large number of
his mamluks and all the varied ceremonies and regalia connected with the
rank of amir.12 With the close of the period, there is a gradual
disappearance of the magnificent amirs owning several hundreds or even
thousands of mamluks. In the Circassian period, their lofty position
becomes a distant and rarely reached ideal. We are told of Yashbak as-
Suduni, one of the important Circassian amirs, that he followed former
generations in his magnificence and the large number of his mamluks 'in
accordance with the times' ( al-waqt),1 i.e. he was considered great
and his mamluks numerous within the limits of the conceptions of his own
day. The number of amirs who possessed large numbers of mamluks had, as
a matter of fact, been greatly reduced at that time. Khudabirdi had 300
mamluks2; Janibak, governor of Jidda and comptroller of its customs, had
between 200 and 300 mamluks3; Aynal al-'Ala'i had 200.4 The only amir of
the Circassian period on a scale reminiscent of the period is Yalbay, a
commander-in-chief, who had over 1,000 mamluks.5 It must be stressed
that the acquisition of a large number of mamluks by any particular amir,
especially in Circassian times, was considered a clear indication that he was
fomenting some rebellion against the sultan.6

THE

In connexion with the amirs' mamluks, attention should be paid to a special

military formation whose members were called These should not
be confused with eunuchs, who were designated by the same term. This unit
had almost completely disintegrated by the early Mamluk period, but it was
still at the peak of its power in the Ayyubid period, especially in its
beginnings. A discussion, therefore, of the under the Ayyubids will
serve to clarify their position among the mamluks.
It is a noteworthy phenomenon that we do not know much concerning
ad-Din's army from published Ayyubid sources.7 The best single description
known to the writer is found in al-Maqrizi's and the source on which
the Mamluk historian bases his statements is In 567 ad-
Din reviewed his troops; 140 8 were present, 20 absent; approximately
14,000 troops were present, most of them and the rest
qaraghulamiya. In the description of this review, the are defined as
follows: 'the is the holder of an income ranging from 700 and 1,000
and 1,200; he has a baggage train (barak) composed of ten or less heads of
animals, including pure-bred horses, ordinary horses, mules, and camels; he
has a page-boy (ghulam), who bears his arms'.1 Ten years later, in 577,
ad-Din again held a review of his troops (after having inspected the feudal
estates and their incomes) and reduced their number to 8,640 horsemen,
distributed as follows: amirs, 111; 6,876; qaraghulamiya, 1,553.2
Thus the formed the majority of ad-Din's regular army, both in
its original and its reorganized forms. The are met on several other
occasions in the Ayyubid period, either participating in military expeditions
or on pay parades; they are also mentioned in the Qawanin ad-Dawawin,
among the units composing the Ayyubid army.3
In the Mamluk period the term gradually falls into disuse. One
comes across it here and there during the first few decades, but afterwards it
disappears almost completely. Below are some examples of its appearance.
The historian Ibn Shaddad, who lived at the end of the Ayyubid and the
beginning of the Mamluk eras, relates that the town of Qurs paid in his day
sufficient land tax (kharaj) to cover the expenses of 40 with their
commanders. Every received 4,000 dirhams, and the commander
4
one-third of the kharaj. When, in 700, Amir Salar goes to Shawbak, 100
are there assigned to him with their feudal estates.5 Amir Shihab
ad-Din al-Qaymari is reported to have given his estate, a fief of 100
to his son.6 One source has it that in 683 he 'was granted a fief for
his inner circle of favourites and for ten '( wa-
li'asharat 7
) while another source states in reference to the same
event simply: ' he was granted (a fief of) ten' ( 'ashara).8 In the Ta'rif of
Ibn Allah al-'Umari, holders of rank are mentioned as having
estates of 40 9
. On the basis of the foregoing, it may well be asked
whether or not the were mamluks. As for the Ayyubid period, there
seems to be no room for doubt; the are a very important military
factor as early as the beginning of that period, and, according to
formed the greater part of ad-Din's regular army. Most of
the Ayyubid sultans did have mamluks in their service; but until the end of
their rule, i.e. the reign of Najm ad-Din Ayyub, who founded the
regiment, their mamluks did not constitute a decisive factor numerically,
militarily, or politically. The
therefore, cannot possibly be identified as mamluks for that period.1 With
respect to the early Mamluk era, it is clear, especially from the last two
quotations, that the were amirs' troops, but the question remains
whether they were mamluks, as Poliak thinks, 2 or not. Nothing in the
above material specifically identifies them as such. Later in this work it is
pointed out that at the dawn of the Mamluk era the sources mention amirs
who had in their service so many horsemen and not mamluks, and there are
other specific indications of horsemen who are not mamluks in the service
of the amirs, as we shall see later. What we know of the seems to
tally with this evidence. The following examples make it quite clear. When
Baybars is promoted in 685 to the rank of Amir of Eighty he says:
'wa-an'ama 'alayya bi-thamanina farisan ',3 while the royal decree
issued in this connexion runs: 'wal-'idda wa-thamanina '.4
In 660 when ad-Din Aghulmish receives an amirate of
Three Hundred on the northern border of the Mamluk kingdom the historian
says: 'wa-'ayyana lahu thalathmi'at fi ar-Rum'.5 Immediately
afterwards he repeats the same information in the following words: wa-
kataba li'l-amir ad-Din al-madhkur manshuran bi-thalath mi'at
Amid wa-a'malaha.6 Thus the identity between and
faris is very evident from these two last examples. It is true that b.
in his Ta'rikh Bayrut, calls the troops serving under the amirs of al-
Gharb mamalik but a detailed list of names makes it clear that these
were not mamluks, but Arabs and sons of Arabs.7This example arouses the
suspicion that even when certain sources specifically mention the word
'mamluk', they may not be referring to mamluks in the technical sense of
the word.
The term in the above-mentioned meaning almost entirely ceases to
appear at a very early date, viz. as far back as the current annals of the first
half of the period. It is, nevertheless, still met with in official
8
documents, whose language tends to be more conservative. It should be
added that by far the greater part of our information on the during
the Mamluk period comes not from Egypt, but from Syria (al-Bilad ash-
Shamiya). Such a peripheral area was more apt to preserve obsolete forms
than the centre of the realm, where innovations originated and from where
they spread forth. We do not encounter the term after the reign of
b.Qalaun, with the following exception: al-Qalqashandi
states that in official letters addressed to an amir's soldier, this soldier would
be called 1
 The Amirs and their Ranks

The officer of the Mamluk army was called amir.2 The rank of officer was
called imra,3 or imriya,4 while the rank of private was called jundiya, the
private himself being called jundi.5
Mamluk encyclopedic literature, especially the works of Ibn Allah al-
'Umari, al-Qalqashandi, al-Maqrizi, and gives detailed
descriptions of the amirs and their various ranks.6
In Egypt, the amirs were divided into the following ranks:—
(a) The highest rank was that of amir mi'a muqaddam alf, viz. an amir
entitled to keep in his service 100 horsemen and to command 1,000 soldiers
of the in the field. This rank is only rarely designated in the sources by
its full title7; a variation of the full title, amir mi'at faris muqaddam 'ala
alf,8 is also found, but very rarely, and that only in the beginning of the
Mamluk period. Much more frequent are the following: muqaddam alf,9
amir muqaddam,10
or simply muqaddam,1 amir mi'a (pl. umara' al-mi'in),2 amir alf,3 and
sometimes simply alf (pl. uluf).4 Some of these titles occur most frequently
in the plural. The designation of the rank itself rarely appears in its full form
of imrat mi'a wa-taqdimat alf 5; much more frequent are the abbreviated
forms taqdimat alf6 or taqdima,7 while the term imrat mi'a is less common.8
The term designating the act of appointing to that office is qaddama,9 and
for receiving such an appointment, taqaddama.10 The number of horsemen
in the service of such amirs might reach. 110 to 120, and we have seen
above that the important amirs actually had much greater numbers. The
holders of the most important posts of the state were selected from among
these amirs, whose total number was twenty-four, nine of whom were office
holders: commander-in-chief (atabak al-'asakir), grand master of the
armour (amir ), lord of the audience (amir majlis), grand dawadar
(dawadar kabir), grand master of the stable (amir akhur kabir), chief of the
corps of mamluks (ra's nawbat an-nuwab), grand chamberlain
grand treasurer (khazindar kabir), and leader of the Egyptian pilgrims'
caravan (amir ); cf. below, the section on office holders. The total of
twenty-four was fixed at the time of the redistribution of Egyptian land
conducted by b.Qalaun (ar-rawk ), but when the
diwan al-mufrad was established by Barquq and many of the Royal
Mamluks were transferred to it, that number was diminished. In the days of
al-Qalqashandi,their number varied between eighteen and twenty, among
whom were included the governor of Alexandria and the governors of
Northern and Southern Egypt.11 The annals give fairly extensive
information on the numbers of the Amirs of a Thousand at various periods;
in 721 there were twenty-four of them1; there were twenty-four under
b.Qalaun and fewer later2; in 791 Yalbugha fixed their
number at twenty-four as of old, in order to stress the extent to which.
Barquq had violated the customs and prescriptions of the former sultans.3 In
827 there were eleven Amirs of a Thousand,4 in 861, eleven,5 in 865,
twelve,6 in 868, thirteen,7 in 872, fourteen,8 in 908, twenty-four,9 in 920,
twenty-seven,10 in 921, twenty-seven,11 in 922, twenty-six.12
(b) The second highest rank was that of Amir of Forty, for which we find
two synonymous appellations in the contemporary sources: amir arba'in13
(pl. umara' arba'in, or arba'inat),14 and amir 15 (pl. umara'

or ).16 Of these two appellations, the latter is much more frequent in

the later Mamluk period. The office itself is called imrat or simply
17 or imrat (imriyat) arba'in.18 Each amir of this rank was generally

entitled to keep in his service 40 horsemen, which. figure was at times

increased to 70 or even 80.19 The number of the Amirs of Forty was not
fixed, but underwent considerable variations. It often happened that one
Amirate of Forty was divided into two Amirates of Twenty or four Amirates
of Ten, or conversely that a number of Amirates of Ten were merged into
one Amirate of Forty.
The Amir of was so called because holders of this and higher ranks
were entitled to have a band playing in front of their houses.
According to the sources, the consisted of a group of musical
instruments, including many drums and some trumpets (abwaq), and flutes
(zumur) of various timbres and playing in a specific style. Every evening,
following the evening prayer, the instruments would be played. According
to Khalil b.Shahin the Amirs of a Thousand had before their houses
eighty 'loads'
of two timbals (dahl), two flutes, and four trumpets (anfira),
etc. The orchestra playing at the gate of the atabak al-'asakir was twice as
large.1 Three 'loads' of had formerly played at the gates of the
Amirs of Forty, but in the days of there were only two drums and
two flutes. The accompanied the of the sultan or the amirs in
wars and expeditions with the aim of heartening the troops and striking
terror into the hearts of the enemy.2
(c) The third rank was that of Amir of Ten. An officer of this grade was
called, besides amir 'ashara, al-'ashrawat (or'ashrawat), al-
'asharat, al-umara' al-'asharat, etc.3 The rank itself was called, besides
imrat (imriyat) 'ashara, simply 'ashara; thus: ' 'ashara' 4 ('he granted
him. an amirate of ten'), ta'ammara 'ashara 5 ('he was appointed Amir of
Ten'). All such. amirs were entitled to keep in their service ten horsemen.
According to al-Maqrizi, al-Qalqashandi, and Ibn Allah al-'Umari,
amirs keeping twenty horsemen were included under the Amirs of Ten, but
puts them in a category apart which he calls al-'isrinat. According
to the latter author, their number had formerly been twenty. Mamluk
sources also frequently make mention of umara' 'ishrin or 'ishrinat.6 The
number of the Amirs of Ten, again, was not fixed, and varied for the
reasons already mentioned in connexion with the Amirs of Forty.
(d) The fourth rank was that of Amirs of Five (umara' khamsa). These were
amirs holding fiefs with incomes equal to half that of the Amirs of Ten.
Their number, according to al-Qalqashandi, was exceedingly low (aqall min
al-qalil), especially in Egypt. Most of the Amirs of Five were the sons of
deceased amirs, and received their titles out of deference for their fathers; in
practice they were on an equal footing with the more honoured private
soldiers.7 This statement of al-Qalqashandi's is accurate, for it is only very
rarely that one encounters Amirs of Five in Egypt during the whole of the
Mamluk era8; they are not very numerous in Syria (al-bilad ash-shamiya)
either.1 b. also mentions an Amirate of Four, imrat arba'a,2 a
rank which occurs much less frequently even than the imrat khamsa.
The numbers of the Amirs of a Thousand during the various Mamluk
periods have been given above. We now turn to the numbers of the amirs of
diverse ranks, especially those of Amir of Forty and downwards, in so far as
they are found in the sources.
According to the numbers of the amirs had 'formerly' (qadiman)
been as follows: Amirs of a Hundred, 24; Amirs of Forty, 40; Amirs of
Twenty, 20; Amirs of Ten, 50; Amirs of Five, 30,3 giving a total of 164
amirs or, exclusive of the Amirs of Five, 134. There is, of course, no way of
knowing to which. period he is referring; at any rate, the figures cited do
not at all conform to those of the list of ar-rawk (cf. Appendix A).
There, the number of the Amirs of a Hundred is indeed 24, but the Amirs of
Forty numbered 200, as did the Amirs of Ten, totalling 424 amirs. In 891
we find: 15 Amirs of a Hundred, 10 Amirs of Forty (i.e. less than the
number of Amirs of a Hundred!), 60 Amirs of Ten, 40 4 or a total

of 85 amirs. In 908 there were 24 Amirs of a Hundred (of whom 7 were

office holders), 75 Amirs of Forty (of whom 10 held offices), and 185
Amirs of Ten,5 totalling 284 amirs. In 912 the sultan appoints, from among
his 40 additional Amirs of Ten, over and above the number of
amirs who had been in his service in 908,6 making a new total of 324. In
922 the total of the Amirs of Forty and the Amirs of Ten alone exceeded
300.7

TERMS FOR AMIRS' RANKS AT THE BEGINNING OF THE

MAMLUK PERIOD

The designations of the amirs' ranks given above are presented in their
crystallized and' definitive forms, which were in use during the greater part
of the Mamluk period. These titles were, however, as yet almost non-
existent at the end of the Ayyubid and the beginning of the Mamluk eras; an
examination, therefore, of the earlier and looser forms is vital for an
understanding of their origin and development.
In the 13th century it is only seldom that one encounters the terms amir
'ashara, ami amir mi'a muqaddam alf, or the variants listed above.8
With, reference to the appointment of amirs, we are confronted most
frequently with expressions such as the following: khubz mi'at faris,
mi'at wa-khamsin faris; imrat khamsim faris, imrat mi'at faris, imrat
sittin faris, imrat thamanin faris; mi'at faris; nabulus
wa-jinin wa-a'maliha bi-mi'at wa-'ishrin farisan; li-imrat sittin faris,
etc.1 What stands out in such. passages is that, in every case, there appears a
feudal grant of a highly variable number of horsemen, whereas the rigid
division of the amirs' ranks into their main categories, as listed above and as
found on virtually every page at a later period, has not as yet made its
regular appearance. It should be emphasized that the above expressions
only designate the appointment to a particular rank, while holders of ranks
mentioned after their appointment are usually referred to, in the current
annals, as al-amir, without the indication of any specific grade. This
represents a system completely different from that in use later on, after the
fixing of the definitive terminology.
At a somewhat later time, we find an intermediate nomenclature between
the loose terms of early Mamluk days and the final ones. Thus: taqdimat alf
faris; taqdimat alf wa-imrat mi'at faris; imrat wa-arba'in
farisan;an'ama 'alayhi bi-imrat mi'at faris wa-qaddamahu 'ala alf, etc.2
Here also the form 'amirate of so many horsemen' consistently recurs.
On the basis of this information, it seems to us feasible to inquire into the
type of troops the amirs had under their orders. It is our opinion that the
amirs' troops did not consist of purely mamluk units, especially in the first
half of the period dealt with. A study of the information handed down by
contemporary sources respecting the division of the amirs into their various
ranks reveals the following picture: later sources indicate that Amirs of a
Hundred, of Forty, and of Ten are respectively entitled to keep 100, 40, and
10 mamluks, while the earlier Ibn Allah al-'Umari states that these
amirs are authorized to have 100, 40, or 10 horsemen under their
command,1 viz. not necessarily mamluks, for there were very many
horsemen in the armies of the Mamluk kingdom who were not mamluks.
This observation of al-'Umari's is no mere accident, and is strongly
supported by the fact that early Mamluk sources almost invariably speak of
'amirates of so many horsemen', and not mamluks, as pointed out above (cf.
footnote 1, p. 472). The existence of non-mamluk soldiers in the armies of
the amirs can be proved by additional evidence: a large percentage of the
wafidiya, who were free horsemen, were transferred to the service of the
amirs, although most of them went to the Instances of troops
dissatisfied with the rawk demanding to be transferred to the amirs,
and of amirs' troops passing over to the have been cited above.
Demands such as the would not, it seems to us, have been raised, had
not the transfer of troops to the amirs been a standard procedure (see
also below). It has been pointed out that generally corresponds to
horseman (and not to mamluk) in the service of an amir. It also seems to us
that the fact that the amirs' troops are called, for most of the Mamluk era,
alternatively mamalik al-umara' and ajnad al-umara' shows that they were
composed of mingled mamluk and non-mamluk elements. It has been
indicated that ajnad is the most common designation of the free, non-
mamluk troops. On the other hand, we have never found the Royal
Mamluks, though they are mentioned thousands of times, to be called by
any such name as ajnad or al-ajnad instead of mamalik
or al-mamalik Similarly, the very frequent combination
'al-mamalik wa-l-ajnad' for 'al-mamalik wa-ajnad ' implies that
mamalik and ajnad are distinct in meaning.2 Of special importance is the
fact that even in the Circassian period there were still soldiers of the in
the service of the amirs, as may be learned from this most interesting piece
of information: in 821 Sultan al-Mu'ayyad Shaykh holds a review of all the
and offers two alternatives to those of them serving with the amirs,
namely, either to remain members of the and leave the service of the
amirs, or to be entirely transferred to the amirs and give up membership in
the Some chose the first, and some the second alternative.3 The
preceding examples indicate that there was considerable overlapping
between the amirs' troops and those of the while the last passage
shows clearly that even in the days of al-Mu'ayyad Shaykh, members of the
could be totally absorbed into the service of the amirs. In other words,
as late as the first half of the 9th century A.H. the army of the amirs was not
yet composed solely of mamluks.
The manner in which the amirs' troops were composed developed, in the
writer's view, somewhat as follows: in the Ayyubid period, the bulk of the
troops of the kingdom consisted of free horsemen; the Ayyubid sultans did
buy some mamluks, but these did not form a predominant part of their
army. The first to purchase mamluks on a very large scale was al-Malik
Najm ad-Din Ayyub, the last of the Ayyubids, but even his corps of
mamluks did not exceed 800 to 1,000 men.1 The gradual
ascendancy of the mamluk element in the realm was, presumably, a rather
slow process going from the centre to the periphery, from the select Royal
Mamluks to the less pre-eminent units. This process was accelerated by the
rawks, carried out 50 to 60 years after the rise of the Mamluk state; but
there nevertheless remained, among the amirs' troops, whose importance
ranked second to the sultan's, a considerable number of free horsemen long
after the rawks, as we have seen above. The slow pace at which the
predominance of the mamluk element in the amirs' armies asserted itself is
due partly to an important economic factor: the buying of a mamluk at a
tender age, his rearing and training until he became a full-fledged soldier,
cost considerable sums of money, and not every amir could afford such
expenses. At the beginning of the Mamluk era, when the was strong
and the wafidiya flocked to the Mamluk kingdom, the amir could include in
his service free soldiers who were not greatly inferior to the mamluks, and
who, on the other hand, required no investment of capital for their training.
Later, however, as a result of the decline of the and the stoppage of the
wafidi influx, the amir no longer had at his disposal any such cheap source
from which. he could draw without considerably undermining the efficiency
of his troops; only the local population was left, and these could not replace
the Kurdish, Turcoman, Turkish, and Mongol free horsemen. Thus the amir
was driven to recruiting more and more soldiers from among the mamluks,
expensive as such. recruitment might be. He had, as a matter of fact, no
other alternative, if he wanted to build up an army of his own that would be
of any value.

THE PROMOTION OF AMIRS

One of the fundamental principles of the Mamluk system was the strict
selection of the fittest of the military school graduates for incorporation into
the élite corps of the followed by promotion at a very slow rate.
This principle was largely adhered to in the period, during which.
accelerated promotion was infrequent; this is, no doubt, one of the distinct
marks of the superiority of that period over the Circassian period.
It is owing to Ibn al-'Asqalani that we know who was the first sultan
who allowed accelerated promotion. According to this historian, Baktimur
al-Abu Bakri was appointed by Qalaun to the rank of
Amir of Forty, and he 'was the first to pass from the rank of private to that
of Amir '.1 Thus we learn that during the first 30 or 40 years of
Mamluk rule there was no case of accelerated promotion. Examination
confirms this information, for we have found no other such instances until
Qalaun's time.2 The basis for corruption in promotions and in the obtaining
of ranks or feudal estates was laid in the reign of Sultan al-Kamil Sha'ban,
of the Qalaun dynasty, when Amir Ghurlu founded the diwan al-badal.3 But
what occurred in the period was very moderate in comparison with the
Circassian period. From the reign of Barquq on, accelerated promotion has
become almost the general rule. Ibn Taghribirdi states that since the days
when al-Ashraf Sha'ban was deposed from the throne, everyone who
achieved greatness and participated in disturbances and political intrigues
had been, during the preceding year, either a private or an Amir of Ten, and
virtually unknown. This situation remained unchanged until Ibn
Taghribirdi's own days.4 Thus the number of amirs who, during the
Circassian period, are stated to have passed directly from the rank of private
to that of the highest amirs, is exceedingly great. The common expression
used for such. elevations is '(promoted) at one stroke' (daf'atan ).5
The only Circassian sultan who vigorously opposed such. rapid promotion
among his mamluks was al-Mu'ayyad Shaykh, who was described as an
expert in the deployment of the army in battle and as a fearless hero.1 He
attempted to bring back to life the principles upon which. the Mamluk
system had been founded, and which had been dealt severe blows by the
Circassian sultans.2 He showed no favouritism: all doors were opened to the
good soldier, and the bad soldier was not even granted a fief with a yearly
income of 10,000 dirhams.3 He refrained from taking into his service amirs'
mamluks who had led luxurious lives at their masters' expense, and who
wore sumptuous clothes.4 A soldier upon whom he conferred the rank of
amir was for years not permitted to wear the takhfifa.5 With. his amirs he
was a-strict disciplinarian.6 He distinguished himself in the upbringing and
training of his mamluks, and he let long periods elapse between
promotions; this is the reason, says Ibn Taghribirdi, why every single one of
his mamluks was promoted to a higher rank and achieved fame after his
death.7 The historian's words are borne out by the great number of al-
Mu'ayyad's purchased mamluks who did, in fact, reach very high. positions
only after his death, whereas during his lifetime they had remained
comparatively obscure.8
STUDIES ON THE STRUCTURE OF THE MAMLUK
ARMY—III

By DAVID AYALON
 Holders of Offices Connected with the Army

OFFICE-HOLDERS in the Mamluk kingdom were, as is well known,

divided into three categories: those who belonged to the Mamluk caste and
were called 'men of the sword' (arbab as-suyuf), those who were civilians
and were known as 'holders of administrative offices' or as 'men of the pen'
(arbab or arbab or al-aqlam), and those who belonged to the clerical
class and were called 'holders of religious offices' or ' men of the turban'
(arbab or al-muta'ammimun). We list below some of those
offices which had a direct connexion with the army; some have been
discussed elsewhere in our work on the Mamluk army, and will therefore be
accorded but the briefest attention. We shall in addition outline the
evolution of the offices of the Mamluk kingdom as well as the
transformations which they underwent in the various periods.1

THE MEN OF THE SWORD (ARBAB-AS-SUF)

The Na'ib (vice-sultan or viceroy in Egypt)

Until the death of b.Qalaun, the na'ib was the highest

ranking amir in the Mamluk kingdom, taking precedence even over the
atabak. According to one source, it was he who signed, in the name of the
sultan, the applications for fiefs and he was authorized to give out
small fiefs ( khafifa) without consulting the sultan.2 Following another
source, he acted as sultan on a restricted scale, appointing the amirs and the
office holders, excepting the highest ones, such as that of wazir, of and
of privy secretary (katib as-sirr). He could, however, propose candidates for
these posts, and his suggestions were but rarely overruled. During reviews
he would ride at the head of the army. The al-jaysh was also under his
supervision; the al-jaysh (q.v., p. 66) was in close contact with
him, while the al-jaysh (q.v., p. 66) kept close contact with the sultan.3
He was called an-na'ib al-kafil, an-na'ib kafil al-mamalik al-
islamiya (or ash-sharifa), or ath-thani, whereas the governor of
Damascus was called kafil and the governors of important
provinces na'ib ash-sharifa,
and those of secondary provinces an-na'ib bi-fulana.1 At the close of the
life of the office of na'ib was abrogated,2 to be
renewed later, but without regaining its former pre-eminence. It was only in
the days of al-Askraf Sha'ban that al-Yusufi, who filled that post, was
granted extraordinary powers: he could, on his own initiative, grant fiefs
with a yearly income not exceeding 600 dinars, dismiss whomever he
pleased and appoint Amirs of Ten and Amirs of Forty throughout al-bilad
ash-shamiya.3 The last to serve in this capacity was, according to one
account, Aqbugha at-Timrazi, in 842.4 According to another version, it was
al-'Uthmani.5 The na'ib served as acting sultan when a
military expedition or other matters required the sultan's departure from the
capital. When the office was abrogated, the duties of acting sultan were
carried out by the highest-ranking amir remaining in the capital, who was
called na'ib al-ghayba.

The Atabak al-'Asakir

After the abrogation of the office of na'ib the atabak was the first
of the Mamluk amirs. It was common, especially in the Circassian period,
for him to succeed the sultan on the throne. He was commander-in-chief of
the army,6 but his functions were much broader, as indicated by the
frequently appended title of mudabbir al-mamalik or mudabbir al-mamalik
al-islamiya.7 He was ordinarily known as atabak al-'asakir, but he was
sometimes also called atabak al-jaysh, or al-juyush.8 It is not clear whether
the title amir al-juyush9 refers to him as well. His title was also abbreviated
to al-atabaki.10 According to Qalqashandi, the original spelling of that title
was and it was only later changed to its accepted form.1 According to
the atabak was called baklar baki, i.e. amir al-umara' 'amir of
amirs', a title used also among the Turcoman dynasties of Anatolia,3 as
2

well as among the Mongols, with respect to some of their high ranking
amirs.4 The designation amir al-umara' is sometimes encountered in
Mamluk sources,5 but it is very doubtful whether it refers to the atabak al-
'asakir. No source other than is known to us to have claimed that
the atabak was called baklar baki. One of the most common titles of the
atabak al-'asakir was al-amir al-kabir, and the first to receive this title was
Shaykhun al-'Umari6 (the term amir kabir is discussed fully in Appendix B
below). The office of atabak was called atabakiya, or atabakiyat al-
'asakir.7 For a short period it was held conjointly by two individuals.8 The
residence of the atabak was generally at or the bab as-
silsila.9
In the interregnum between the and Circassian periods, the post of
atabak was seized a number of times by individuals who rose from
obscurity to greatness at one stroke, for prior to their becoming
commanders-in-chief they were only privates or low amirs. After the
murder of Sultan Sha'ban, the post was held by a succession of such men:
Tashtimur al-Laffaf, Ayanbak ad-Dubri, Barquq, and
Baraka; their example encouraged many others to emulate them.10

The Amir Majlis ( of the )

The amir majlis had charge of the physicians, oculists, and the like. This
office was filled by one person only.11 The sources do not elucidate the
connexion between the rank of amir majlis and this particular task, which
seems to be of no special importance. Although the rank of amir majlis was,
in the first Mamluk period, superior to that of amir (see below), neither
of them was of great significance at that time. In the Circassian period, the
amir majlis, though inferior to the amir was third in importance among
the highest amirs of the kingdom. (For details on the introduction of this
office, see p. 69.)

The Amir (Grand Master of the Armour)

It was the duty of the amir to bear the sultan's arms during public
appearances. He also had charge of the Royal Mamluks' and was
supervisor of the arsenal. The office was held by an Amir of a Thousand.1

The ( Chamberlain)

The main function of the was the administration of justice

among the mamluks of the amirs according to the laws of the Yasa.2 His
authority was independent, but during the time that the office of na'ib
 was in existence, he was sometimes obliged to consult with the
holder of that office. It was also his duty to present guests and envoys to the
sultan, and he was in charge of organizing army parades. It was customary
to appoint five two of whom (the and the ) were
3
Amirs of a Thousand ; the rank of the thani eventually declined to that
4
of Amir of Ten. Wh the office was first created, there were only three
and thani. The first to increase their number to
five was Sultan Barquq; even the lowest of these were at first Amirs of Ten,
'and not the criminal and ignorant riffraff who fill this office to-day'.5 The
office of was known as 6 and that of the chief as
7

The Ra's Nawbat an-Nuwab (Chief of the Corps of Mamluks)

According to al-Qalqashandis definition, this amir had charge of the Royal

Mamluks, supervised their conduct, and executed the sultan's or the amirs'
orders applying to them. He was also responsible for the parades held by
the army before it set out on an expedition. The plural of this title is ru'us
an-nuwab; the ignorant call the holder of this office ra's nawbat an-nuwab,
whereas the correct designation is ra's ru'us an-nuwab.8 It should be noted,
however, that the title indicated as the correct one by al-Qalqashandi is not
found anywhere in the sources; only ra's nawbat an-nuwab or an-nuwwab is
used.9 The ra's nawbat an-nuwab came into existence only after the
abrogation of the office of ra's nawbat al-umara' (see p. 70), and the holder
of this office had previously been known as ra's nawba thani, and had the
rank of Amir of a Thousand.1 The office itself is sometimes called ar-ra's
nawbiya al-kubra.2 The number of the ru'us an-nuwab was four, one of
whom was an Amir of a Thousand, and the rest Amirs of 3

The Wazir4

The wazirate was one of the most important offices of the kingdom during
the early Mamluk period. Sanjar ash-Shuja'i (died 693/1294) was the first
Mamluk amir to be appointed to that office. After him there came a series
of Mamluk wazirs who had bands playing at their gates, in accordance with
the practice of the wazirs in Caliphal times.5 Some of these were so
influential that they could distribute small fiefs in Egypt without consulting
the sultan and appoint Amirs of Ten and Amirs of Forty in Syria.6
According to one source the wazirate functioned well only as long as it was
headed by a mamluk: inna al-wizara in lam-yataqalladha mamluk fasada
7;and, indeed, it was but rarely that a mamluk would be appointed to

this office. Under Barquq the wazirate declined rapidly, after he had created
the al-mufrad8 and transferred to it 5,000 Royal Mamluks. He then
divided the wazirate and its functions into four separate offices: the
wazirate, the ustadariya, the 9 and the kitabat as-sirr (the office

of Privy Secretary).10 It was the office of the which gained most

from the deterioration of the wazirate during the 9th–15th centuries. The
principal and almost only duty of the wazir at that time was to supply meat
to the army.11 Meat merchants and butchers ( mu'amilin )
occupy key positions in that office, and some of them even became
wazirs.12 From the middle of the 15th. century onwards the wazirate
becomes more and more the private domain of butchers and meat dealers.

The Ustadar ( Major )

The ustadar headed the diwan al-mufrad or diwan the main

function of which. was the distribution of the monthly pay (jamakiya) and
fodder ('aliq) and, on rare occasions, clothes (kiswa), to the Royal
mamluks.1 The ustadar's deputy was called al-mufrad.2
According to al-Maqrizi the office of the ustadar was considerably
strengthened in his day, and he discharged many responsible duties.3 There
is no doubt that he was far more important than the wazir. Yet his position
was always precarious because as a rule he was unable to pay the mamluks'
salaries in time. Very often he was beaten and put in custody until he paid a
considerable part of these salaries from his own pocket.4

The Khazindar al-Kabir (Grand Treasurer)

The khazindar had charge of the sultan's treasures, including his funds, his
precious clothes, and the like. This office was generally held by an Amir of
 later to be permanently replaced by an Amir of a Thousand.5 We
have discussed elsewhere the eunuchs filling this office, but we know of no
eunuchs serving in this post who held the rank of Amir of a Thousand.

The Dawadar al-Kabir (Grand Dawadar)

The basic function of the was the bearing and keeping of the royal
inkwell. This office was created by the Saljuqs, and was held by civilians
both under their rule and under the Abbasid caliphs. It was Baybars
6
who transferred it to a Mamluk Amir of Ten. During the period the
dawadar did not rank among the important amirs, but under the Circassians
he became one of the first-ranking amirs of the kingdom.7 Some dawadars
even became sultans.8 One of the dawadar's duties during the later Mamluk
period was to decide which of the members of the were worthy of
9
setting out on a military expedition. In addition, he regularly visited Upper
Egypt, and sometimes the regions of Jabal Nabulus, ash-Sharqiya, and al-
Grharbiya, in order to collect taxes and gather in the crops. These trips
would take place amid great pomp and brilliance, and the sources discuss
them at length. They were accom-
panied by cruel oppressions of the local population.1 At the close of the
Mamluk era, enormous power was concentrated in the hands of the
Amir Yashbak was, in addition to his duties as dawadar, also amir
wazir, ustadar, kashif al-kushshaf (inspector-general), mudabbir al-
mamlaka, and ra's al-maysara. No previous Mamluk amir had accumulated
such a great number of offices.2 The dawadar who later became
sultan, accumulated exactly the same offices.3

The Amir Akhur

The amir akhur was the supervisor of the royal stables. He was generally an
Amir of a Thousand, and resided in He had under his orders
three Amirs of and an undetermined number of Amirs of Ten and
privates.4 It appears that for every department which came under the diwan
of the amir akhur, there was a special official:thus there was an amir akhur
at-tibn 5 an amir akhur al-jimal,6 etc.
 The Amir Jandar

The amir (or pl. janadira7 or jandariya,8 was in charge of the

zardkhana (which. served both as an arsenal and as a detention house) and
of the execution of those condemned to death by the sultan. The zardkhana
was considered a relatively respectable place of detention; its prisoners did
not remain there long, and were either released or executed without great
delay. The jandar also announced to the sultan the amirs' arrival for duty,
and he entered the before them. He also presented the inkwell to the
sultan, together with the and the katib as-sirr. Under his orders were
the the rikabiya, and the He would lead the zaffa
(procession?) around the sultan during his expeditions. This office was held
by an Amir of a Thousand and an Amir of 9 In the Ayyubid period
and at the beginning of the Mamluk era, this was one of the highest
positions of the realm. That office and the post of ustadar were given by the
Ayyubid Turan Shah, upon his coming to power, to the two highest-ranking
amirs.1 In the early Mamluk period, the office of was held by Amirs
of a Thousand, including even Aljay al-Yusufi and Janibak but
later it deteriorated, reaching its nadir in the middle of the 9th century A.H.
From that time until the end of the Mamluk period it was held by privates.2

The Naqib al-Jaysh

The ordinary designation of the holder of this office was naqib al-jaysh or
naqib al-juyush, though. at the beginning of the Mamluk period he had been
called naqib al-'asakir.3 He was a sort of chief of the military police.
According to Qalqashandi, it was his duty to bring to the sultan all the
amirs, members of the etc., who were summoned to his court. He was
also responsible for guarding (the sultan's life?) during official ceremonies
and expeditions. Under him were a number of nuqaba', and his rank was
that of a lower 4 This testimony of Qalqashandi's is confirmed by the

chronicles which, however, give some additional functions for the naqib al-
jaysh: conveying amirs and administrative officials from prison to court-
house,5 keeping them in detention,6 transporting dismissed amirs and
administrative officials to their place of exile or detention,7 taking mamluks
condemned to death to the place of execution,8 announcing to the army that
it was to prepare for a parade or expedition.9 In order to carry out this last
duty, he would send the nuqaba' ajnad who were under his orders,
to Cairo and its environs, and the to the diverse regions of Egypt.10
According to Qalqashandi, the naqib al-jaysh was in Syria (al-mamalik ash-
shamiya) called naqib an-nuqaba,1 but this designation is also found in
Egypt under Qalaun.2

The Naqib al-Mamalik

In addition to the naqib al-jaysh, the early Mamluk period had a naqib al-
mamalik. These two offices were different from one another, as may be seen
from the sources. It is said of one amir that he was naqib al-jaysh, and had
previously been naqib al-mamalik.3 Another amir was naqib al-mamalik,
and was later appointed to be naqib al-jaysh as well.4 Other amirs served
simultaneously in both capacities.5 This office is quite frequently
mentioned.6 Since the naqib al-jaysh often served as naqib al-mamalik, we
may assume that these two offices were similar, but the sources do not
make clear in what manner they may have differed. On the basis of the
designations, it may be presumed that the functions of the naqib al-mamalik
were restricted to the mamluks, or the Royal Mamluks, while those of the
naqib al-jaysh included all (or: all other) branches of the army. It would
seem that the existence of a al-mamalik side by side with a
diwan al-jaysh must be explained in the same manner; see p. 66). The first
office apparently ceased to exist as early as the first half of the 8th century.

The al-Mamalik ( of the Royal Mamluks)

The muqaddam al-mamalik who was usually a eunuch, was head

of the military schools of the Royal Mamluks in the Cairo citadel.7

The Katib al-Mamalik (Scribe of the Royal Mamluks)

The katib al-mamalik (or, more usually, katib al-mamalik) was in

charge of the nominal lists of the mamluks. He read out the names of the
mamluks during pay parades and other official ceremonies.8
 The Malik al-Umara' (King of the Amirs)

This was not an office but a title given to the high-ranking governors of
Syria (al-bilad ash-shamiya), Alexandria, and Upper Egypt (al-wajh al-
qibli). The vice-sultan in Egypt, however, was called kafil al-mamalik, to
stress his superiority over these governors.1 Khayrbak, who bore the title of
malik al-umara' in his capacity as governor of Aleppo, kept it even after the
fall of the Mamluk state and his appointment as ruler of Egypt under the
Ottomans.2

THE MEN OF THE PEN (ARBAB AL-AQLAM)

The al-Jayash (Inspector of the Army)

The distribution of, and supervision over, feudal fiefs were in the hands of a
special department called diwan al-jaysh or diwan The department's
headquarters in Cairo was divided into two principal sections: (a) the diwan
al-jaysh responsible for Egyptian fiefs, and (b) the diwan al-jaysh
ash-shami, in charge of Syrian fiefs. Each. section was headed by an
official known as mustawfi, sometimes as mutawalli, or katib al-jaysh.
There were two lower mustawfis, one of whom was in charge of the
distribution of fiefs to the Beduins, and the other to pensioners. The director
of the diwan, the al-jaysh, was responsible to the sultan, while his chief
assistant, the diwan al-jaysh, was responsible to the viceroy. The
al-jaysh had branch offices in all regions of the kingdom.3 In addition to the
diwan al-jaysh, the al-jaysh had under his orders the diwan
al-mamalik, katib al-mamalik, 4
al-mamalik, and other clerks. It seems
likely that the last three offices were confined to dealing with the mamluks
or the Royal Mamluks only, not with the whole of the army.

The Khaza'in (Supervisor of Arms Stores)

The holder of this office was charged with supervising the armament used
by the army, and with seeing to it that arms manufactured during the year
were transferred, on a fixed date, to the arms stores of the Cairo citadel.5
 The (Supervisor of the Royal Stables)

The holder of this office was responsible for the royal horses, mules, and
camels. He supervised the purchase and sale of these animals, and the
payments to the personnel employed in the stables.6

RELIGIOUS OFFICIALS (AL-MUTA'AMMIMUN)1

The al-'Askar (Judge of the Army)

The main function of al-Askar was to accompany the military

expedition, try its members, and decide on such judicial questions as might
arise during the march, e.g. the division of booty, the inheritance of dead
soldiers, etc.2 His peace-time duties are not clear. There were four chief
judges of the army, one for each of the four schools of orthodox
Islam.3

To conclude the list of offices, two expressions of very frequent oceurrence

in connexion with Mamluk posts should be discussed. The first of these is
'kana', used in the sense of 'formerly, ex-', in relation to officials who ceased
exercising their functions. Thus: atabak al-'asakir kana 'former atabak al-
'asakir' ; na'ib al-karak kana 'ex-governor of Kerak', etc.4 The second is
'sa'ada', used in the sense of ' success and stability of (an amir's) career,
successful service over a long period of time'. It was customary to say:'
ayyamuhu fi as-sa'ada' and similar expressions.5

ON THE EVOLUTION OF THE OFFICES

The offices of the kingdom underwent very considerable changes under the
rule of the Mamluks. They had, at first, been taken over from the Ayyubids,
but under Mongol influence, especially during the reign of Baybars al-
Bunduqdari, new offices were introduced, which partially replaced those in
use since the Ayyubids, without abrogating them entirely. The evolution of
the uppermost offices is of great interest; unfortunately, however, data on
this subject were not systematically gathered by the writer, so that the
description given below will necessarily fall far short of the prospects
offered by the abundant source material. It is, however, his intention to deal
with this subject at length. at another opportunity.
Under Turan Shah, one of the last Ayyubid sultans, the two most important
amirs of the kingdom were the ustadar and the amir jandar.6 Under the
Mamluk sultan 'Ali b. Aybak, the chief amirs were: the na'ib the
7
atabak al-'asakir, the wazir, and the amir jandar. The atabak and the wazir
are mentioned when comes to power.8 When Baybars came to power
in 658, the order of offices was as follows: na'ib atabak, ustadar,
amir jandar, dawadar, amir akhur, wazir, two amir 1 During Qalaun's

reign, mention is made of the amir majlis, amir akhur, and amir
jandar.2 In the period 693–707 the anonymous author published by
Zetterstéen very frequently repeats the list of the offices of the Mamluk
kingdom. He generally mentions eight offices, presented in the same order,
with some minor changes: na'ib wazir, ustadar, jandar,
dawadar, and naqib al-jaysh. The number of the varies
between two and four; that of the jandariya between one and four; that of
the mihmandariya generally remains at two; that of the dawadariya varies
between one and three. It is noteworthy that in the several lists given by the
above historian for the stated period, the atabak al-asakir is not mentioned
at all.3
In 783, i.e. at the close of the period, we encounter an altogether
different order of offices, although it is not yet the one stabilized during the
Circassian period: atabak al-'asakir, ra's nawbat al-umara, amir amir
majlis, kabir, amir akhur, ra's nawba thani. 4

In the Circassian period the sources usually mention seven offices in a fixed
order: atabak al-'asakir, amir amir majlis, amir akhur, ra's nawbat an-
nuwab, kabir, The order of the first four offices was
fixed for the whole of the Circassian period, and the office of
generally, though not always, retained its seventh place. There was
competition between the offices of ra's nawba and kabir for the fifth
5
and sixth places, possession of which alternated irregularly between them.
Thus it is seen that the office roster of the early Mamluk period differs
greatly from that of the Circassian period. As pointed out above, the great
changes occurred under Sultan Baybars, one of the Mongols' greatest
admirers; he introduced the laws of the Yasa into the Mamluk kingdom and,
in the wake of the Yasa, many of the institutions and offices of the Mongol
state.6 Of this, Ibn Tagkribirdi says the following: Some of the offices
introduced by Baybars had indeed existed, previously, but their nature was
considerably altered during his reign. Thus, for instance, the
function before that sultan's time had been the bearing of the inkwell, and
that office was filled exclusively by civilians; the amir majlis was charged
with the guarding and arranging of the sultan's audience, and the was a
sort of gatekeeper. The offices introduced by Baybars were those of
khazindar, amir akhur, suqat, ru'us an-nuwab; amir
amir majlis, amir shikar (master of the hunt). Under Baybars, the
function consisted in the supervision of the as well as the
conveying of arms to the sultan in battle and on other occasions, such as the
Feast of the Sacrifice. At that time, that office did not carry the high dignity
which. it reached under the Circassians, i.e. the right to sit as ra's al-
maysara in the sultan's presence. That latter function was, under Baybars,
reserved for the atabak and, under b.Qalaun, for the ra's
nawbat al-umara'. As for the amir majlis, he was, in Baybars' days, in
charge of the physicians, oculists, and medical assistants. He carried, at that
time, a higher rank than. the amir The office of dawadar went, under
Baybars, to an Amir of Ten, and was still chiefly of a civilian character. The
office of ra's nawba was an entirely new post introduced by Baybars, and
there had previously been no office so named, whether under the Mamluks
or under the dynasties that preceded them. It had been a very lofty post
among the Tatars, and its holder was called, in their language, 'yswl.' The
office of amir akhur had also been greatly revered among the Tatars, and its
holder was called, in their language, As for the the office
founded by Baybars, it steadily grew in importance, until under
b. Qalaun, it equalled that of the vice-sultanate (niyabat
). The rest of the Mamluk offices were introduced by Qalaun and by
his son Ibn Taghribirdi deals with them in his account of
the reign of these two sultans.1
In the period there also existed a supreme council of high-ranking
amirs, called umara' al-mashura or umara' al-mashwar. The president of
this council was called ra's al-mashura.2 This council is almost never
encountered in the Circassian period.3
During the interregnum between the and the Circassian periods, viz. at
the end of the 8th and the beginning of the 9th centuries, a period marked
by severe political crises and upheavals, we find that all the important
offices of the kingdom were occupied jointly by two individuals. The
purpose of such an arrangement was, apparently, to prevent the
accumulation of excessive power in the hands of a single amir by
establishing a counterpoise vis-à-vis each office holder. One member of
each pair was superior in rank to the other and was called 'insider'
while his companion was but a 'partner' (sharik), called 'outsider'(barrani).
The most famous such pair were Barquq and Baraka. Ibn Taghribirdi, in his
account of the appointment of al-Ashrafi as amir and
of Arghun as amir kabir barrani, says that he copied this
information from the historian al-'Ayni, and that other historians concur
with him. He then adds: 'Thus there were in those days an amir kabir
and an amir kabir barrani, an amir and an amir banani; this
is something quite unheard of'.1 It is clear from this passage that these
double offices no longer existed in the days of Ibn Taghribirdi, and that he
unearthed their existence only through reading about them in earlier
sources. We have, in fact, found no trace of their existence later than the
first years of the 9th century.
At the beginning of the Circassian period, under Sultan Faraj, yet another
office is abrogated, namely that of ra's nawbat al-umara', whose full name
was ra's nawbat al-umara' atabak, or ra's nawba kabir (with
apparently to distinguish it from atabak). Below it was the office of ra's
nawba thani, which corresponds to the later ra's nawbat an-nuwab. The
office of ra's nawbat al-umara' was formerly second to that of atabak al-
'asakir.2 The writer has not been able to ascertain the exact nature of this
post, especially in what way it differed from the office of ra's nawbat an-
nuwab.
 Appendix A

1. STRENGTH OF THE ARMY DURING AR-RAWK (715)

ACCORDING TO AL-MAQRIZI

established the number of the total army in Egypt at

24,000 horsemen. These were subdivided as follows: Amirs of a Thousand
and their mamluks, 2,424 horse; the Amirs of a Thousand, numbering 24,
were subdivided as follows: na'ib wazir, 8 amirs, 14 kharji
amirs 3; their mamluks numbered 2,400. The Amirs of numbered,
with their mamluks, 8,200 horse; these amirs, numbering 200, included 54
amirs, 146 kharji amirs, and their mamluks numbered 8,000 horse.
The inspectors (kushshaf) and governors (wulat) in the sub-provinces of
Egypt, together with their mamluks, numbered 574, including 14 kushshaf
and wulat (here the table gives a detailed list by sub-provinces), and 560
mamluks. The number of the Amirs of Ten and their mamluks was 2,200
horse, comprising 200 amirs, of whom 30 were and 170 kharjiya,
and 2,000 mamluks. The wulat al-aqalim1 numbered, with their mamluks,
77 horse; the wulat al-aqalim (here the names of the aqalim are detailed)
were 7, and their mamluks 70. The commanders of the Royal Mamluks and
those of the together with the Royal Mamluks and the soldiers of the
numbered 11,176 horse,2 subdivided as follows: commanders of the
Royal Mamluks al-mamalik 40; commanders of the
180; nuqaba' al-uluf (the writer is unacquainted with
this rank), 24 ; Royal Mamluks and soldiers of the 10,932 horse, of
which the former accounted for 2,000 and the latter 8,932. 3
The total of the mamluks of the amirs, in their three categories, is 12,400, a
figure which is brought to nearly 13,000 with the addition of the governors'
mamluks. This represents over half the total strength. of the army stationed
in Egypt, a situation entirely inconceivable at a later period. Similarly, the
list indicates that every Amir of a Thousand had exactly 100 mamluks,
every Amir of exactly 40, and every Amir of Ten, exactly ten. Such
an exact distribution was no longer in accordance with the facts as early as
the days of Ibn Allah al-'Umari, who wrote only two or three decades
after the rawk (see also below). On the other hand, the low
proportion of the Royal Mamluks is noticeable; they comprised only one-
twelfth of the entire army. As for the its strength in 715 A.H. was in all
probability approximately 9,000, as indicated in the present list, and not
24,000, as al-Maqrizi indicates elsewhere,4 and as stated by in the
list presented below. It would seem that they confused the strength of the
whole army in Egypt with that of the The in spite of the severe
blows it was dealt as a result of the rawks, was still numerically important
in 715, forming as it did more than one-third of the army's total strength.
The extent of the decline at a later period may be gauged from its
enormous numerical losses: it numbered only 1,000 men in the first half of
the 9th century5 or, in other words, its numerical strength dwindled to one-
ninth of its former size in 100 years.

2. THE STRENGTH OF THE ARMY IN THE WHOLE OF THE

MAMLUK KlNGDOM, ACCORDING TO

list of the army of the Mamluk kingdom was composed, according

to the historian's statement, in response to a particular set of circumstances.
During the reign of a certain sultan, an envoy arrived, representing the
Tatars, and threatened that his nation would conquer Egypt. He boasted that
the Tatars had an army of 20 tumans, comprising 10,000 horse each. The
sultan then decided to order a census of his army, the results of which are
here listed: 24,000; Royal Mamluks, 10,000; amirs' mamluks, 8,000.
troops in Damascus, 12,000, troops of the governor and amirs of
Damascus, 3,000; troops in Aleppo, 6,000, troops of the governor and
amirs of Aleppo, 2,000; troops in Tripoli, 4,000, troops of the
governors and amirs of Tripoli, 1,000; troops in Safed, 1,000, troops of
the governor and amirs of Safed, 1,000. In addition, an army of 60,000 men
was encamped in the north of the kingdom and in Egypt.
The Beduin tribes ('urban) contributed the following numbers: Al Banu
Nu'ayr, 24,000; the Beduins, 24,000; Al 'Ali, 2,000;'Arab al-'Iraq,
2,000; Lamlam, 2,000; Banu Saqba and Banu Mahdi,1,000; Al Imra (i.q.
Murra?), 1,000 ; Judham, 1,000; Al 'A'id, 1,000; Fazara, 1,000;
1,000; Qatil, 1,000; Qatab, 1,000; various tribes, each of which. included
more than 100 horse, 3,000. The tribe of Hawwara had formerly mustered
24,000; the tribes of the Turcomans, from Gaza to Diyar Bakr, numbered
180,000 horse (here follows a list of the names of the tribes). Al-'Ashir
(semi-nomadic tribes) mustered 35,000 horse, and their commanders
numbered 35. The Kurds had formerly numbered more than 20,000 horse.
Every village of the Mamluk kingdom was required to muster two
horsemen, and there were 33,000 villages.1
There can be no doubt that of the two lists presented above, the first gives
the more authentic impression. It was put together in response to a great
event in the annals of the Mamluk kingdom, which demanded, as a pre-
requisite, that the strength. of the army and its various subdivisions be
ascertained. Moreover, the figures it offers are generally plausible, except
for the small number of the Royal Mamluks, which is difficult to explain
(see BSOAS., xv, pp, 222–8). It seems likely, therefore, that we have here
an important historical document, which reflects in a fairly reliable manner
the numerical strength of the army stationed in Egypt at the height of the
Mamluk era.
The second list is much. more suspect. First, it is of unknown date, and
even the name of the sultan who ordered it to be compiled is not known;
second, it was composed as a reaction to threats by an external enemy, and
with the purpose of proving to that enemy that the Mamluk army was larger
than his. All the figures given in the second part of the list concerning the
number of horsemen the kingdom could muster from among the Beduins,
the Turcomans, the Kurds, the semi-nomadic tribes, and the villages, as
well as the figures relating to the troops stationed in the north of the
kingdom, are entirely devoid of value. This stands out especially in the light
of the exact accounts furnished by contemporary sources on the poor
participation of all the above-named elements in the wars waged by the
Mamluks. The figures presented in the first half of the list may be taken
more seriously, although here, too, it is clear that the number of troops
was much smaller than the figure listed, and that the share of the troops of
al-Bilad ash-Shamiya in the general strength of the army was much less
significant than might be imagined from figures.
 Appendix B Who Were the

In the body of the present chapter, the have been classified as the
mamluks of former sultans. No evidence was adduced for this contention in
order not to disrupt the continuity of our description; the term, however,
requires a detailed examination since, in the writer's view, our outlook on
the foundations of the Mamluk system depends, in no small measure, on its
correct definition.
The term has already been dealt with by a number of scholars. Von
1
Hammer, who relied on European sources without testing their validity,
described the in a wholly fictitious manner and even distorted their
name to (corsairs!). A.N.Poliak describes them in an entirely
different vein, which, in the present writer's opinion, is just as incorrect (see
below). It seems that the only description which is partially correct is given
by Popper in his critique of Poliak.2
In order to enable the reader to follow our line of argument in defining the
term via a critique of A.N.Poliak's views, we here submit our final
conclusions in advance.
The term 3 in Mamluk sources of the Circassian period,4 had two

interconnected meanings: (a)' veterans, men with long-standing service'; (b)

'mamluks of former sultans' (who were veterans in comparison to the
julban, the mamluks of the reigning sultan).5 The second meaning, which
we consider to have sprung from the first, is by far the more frequent in the
above-mentioned sources.
Poliak, who deals with the in several places in his works, defines
them in an entirely different manner. He says at one point: 'The Mamluk
state may be viewed as a colonial empire of the feudal lords and merchants
of the region north of the Black Sea… During the first period, the feudal
lords and merchants of the metropolis were organized in a united State, the
Golden Horde. During the second period, the Circassian stood at the
head of the feudal element, and the merchants of the Crimea stood at the
head of the mercantile element'.1
A second passage of Poliak's reads: 'In view of the fact that the Mamluk
sultans and amirs married into the principal families of Circassia, and were
doubtless aided by their relatives, who had remained in the home country,
one is led to believe that the ruling power was practically held by the same
group in Circassia and the Mamluk kingdom. This group is called in the
Mamluk sources sing. The were represented
in the diverse ranks of the military hierarchy, but were everywhere held in
greater esteem than their colleagues, even their superiors, and were the first
candidates for promotion. Thus one amir recommended (who later
became sultan) to the sultan as while was still an unliberated
mamluk. There were also among the mamalik i.e. the freed
mamluks who were in the service of the sultan. A small and special corps
was called al-ajnad its members were candidates for the amirate
and holders of large fiefs (arzaq), and their social position was similar to
that of Amirs of Five. This corps was composed of men who had served in
the Mamluk kingdom for a long time. Naturally, there were also
among the julban, the recently arrived mamluks. Needless to say, they were
represented also among the amirs, and here also, it was the fact of
belonging to this nobility more than the military rank which determined the
amir's social standing. The governor of Jerusalem, Khushqadam as-Sayfi (d.
853 A.H.), in spite of his bravery," did not belong to the notables and to
those who are the chiefs of their compatriots", while a private soldier
(jundi), Lajin (d. 804 A.H.), was considered by the Circassians and even the
amirs a certain candidate for the sultanate. The did not need to wear
fine clothes or ride handsome horses in order to gain high esteem but, on
the contrary, many of them (i.e. the thought it an honour to be
distinguished by old and tattered garments. I know of no case in which an
ibn nas was called It is interesting to note that even in al-Jabarti's
chronicle, which was very much influenced by the language of the Mamluk
sources, we meet the word in the sense of "high amirs". Just as
they (the constituted the aristocracy of the Circassians, so the
Circassians formed the aristocracy of the "Turks",2 among whom the
proportion of other races was considerable'.3
In a third passage, Poliak states: 'As particular units within the first corps
[ajnad ] we may mention: (a) (b) al-ajnad
i.e. those Caucasian noblemen who were not yet dubbed amirs, but whose
social position was already equal to that of Amirs of Five'.1
A fourth passage reads: 'Under the Circassian sultans, the Caucasian
nobility had the right of priority to the fiefs, which was often contested by
the freedmen of the reigning sultan'.2
The gist of Poliak's views may be summarized thus: amid the Circassians,
who formed the aristocracy of the Mamluk races, there was a yet more
select aristocracy, called whose influence held sway both in the
Mamluk kingdom and in the mamluks' country of origin. Its members were
the first to be considered for promotion, and did not belong to any particular
military unit, but were represented in the various formations of the army
and in the diverse ranks of the hierarchy.
We shall first take up Poliak's view of the as belonging to various
units and ranks.
The appear in our sources hundreds of times; yet they are called
in the one and only instance of description of them,3 a
description which is not devoid of gross error and distortion, as will be seen
below. As against this solitary instance, they are in all other cases regularly
named mamatik.4 When. their full name is given, they are called al-mamalik
or 5and we have never encountered them in any
military formation outside the mamalik Just as the julban and the
sayfiya, who were also Royal Mamluks, appear in the sources under three
appellations, viz. julban, mamalik julban, mamalik julban (similarly
for 'ajlab' and 'mushtarawat'), and sayfiya, mamalik sayfiya, mamalik
sayfiya, so the appear in the sources under only three
designations: mamalik and mamalik (or
).6No form other than the above three has been encountered by the writer;
we thus conclude that the are Royal Mamluks. As such they can
belong only to one of the three subdivisions of that body, viz. the mamluks
of the ruling sultan, the mamluks of the former sultans, or the sayfiya. Now,
while they are frequently mentioned as different from the julban and the
sayfiya, as having relations with them, as quarrelling with them, or as
persecuted and mistreated by the julban,1 they are never mentioned as
different from the mamluks of the former sultans or as having any contact
or relations with them. The conclusion is inescapable: a group cannot be
mentioned as diiferent from itself or as having relations with itself.
Further, there are public appearances of the sultan at which it is improbable
that the mamluks of former sultans should not participate in a most
honoured capacity. Unless we assume that the and the mamluks of
former sultans are one and the same body, the question arises as to the
whereabouts of the latter, and as to why the occupy precisely the
position which. would be theirs In a series of such public appearances, for
example, we are told of the participation of: ' the greater part of the Royal
Mamluks ' (ghalib al-mamalik ),2'the whole body of the Mamluks'
(jami' al-mamalik ),3 'the whole army' (jami' al-'askar),4 the whole
body of the army' (al-'askar ).5 The only units mentioned in all the
above instances are the julban and the Were we to follow Poliak, we
might well wonder at the fact that in reviews including the whole of the
army or the majority of the Royal Mamluks, the mamluks of the former
sultans do not appear at all; the non-appearance of the sayfiya, we note in
passing, is not surprising in the light of the insignificance of that unit, which
we have already discussed.
Special importance should be attached to the composition of the Mamluk
expeditionary force to the battle of Marj Dabiq. This force, numbering
5,000 men, was composed of julban, and awlad an-nas, while the
army remaining in Egypt was estimated at 2,000 men, also composed of
julban, and awlad an-nas.6 Again, one might be led to think that in
this crucial battle, at which the fate of the Mamluk kingdom was decided,
the veteran and battle-trained mamluks of the former sultans did not
participate; what is even more astonishing, they not remain in Egypt
either. 7

Of decisive importance, in our view, is the following argument: if the

and the mamluks of the former sultans are not identical, we are forced to
conclude that the latter received no pay for nearly 90 years, for they are
totally absent from all pay parades. The payment of the jamakiya took place
once every month; there were, in addition, other categories of pay, of which
space does not permit a full discussion here. From approximately the
thirties of the 9th century until the end of the Mamluk era, these payments
are described in very great detail several hundred times. In a large number
of instances, the names of all units receiving payment are mentioned, while
in many others only one or a few of these units are listed, but when the parts
are joined together, the full number of existing units is easily obtained. In
all these payments, only the following four military units are mentioned as
ever receiving pay: julban, sayfiya, an-nas.1 In all these
instances, we would search in vain for the name of the mamluks of the
former sultans, or of any of the components of that unit, while the are
almost always listed, usually immediately following the julban, i.e.
precisely in the place where we would expect the mamluks of the former
sultans. Moreover, whenever a unit was deprived of all or part of its pay, it
usually raised a great outcry, or at least the historians pointed to the
injustice of such a procedure (which. affected not only the military, but also
the widows and orphans, who received the jamakiya at the time when it was
paid out to the army)2; yet the mamluks of the former sultans forfeit their
pay for nearly a century, without protesting even once against this iniquity.
Even the historians fail to mention such. a glaring injustice, quite in contrast
with their custom with respect to all other units treated unjustly by
comparison with the julban. This grave contradiction holds true, of course,
only if we fail to identify the with the mamluks of the former sultans.
Not less decisive is the following evidence. In connexion with the struggle
for succession to the throne after the death of Sultan Barsbay, the Manhal
relates that the mu'ayyadiya (Shaykh), the (Faraj) and the
(Barquq) supported Jaqmaq al-'Ala'i, and were also joined by a part
of the ashrafiya Barsbay, headed by Aynal al-Abu Bakri. These form a
united front against the ajlab, the ashrafiya Barsbay, most of whom remain
loyal to the son of the deceased sultan.3 On the other hand, we are told
elsewhere in the same source, in connexion with the same events, that the
army was split into two camps. The first faction grouped itself around the
atabak Jaqmaq al-'Ala'i, and was composed of the Royal Mamluks
(akabir al-mamalik for the term kabir, akabir, see below), joined
by Amir Aynal al-Abu Bakri with. his the ashrafiya (Barsbay).
This alliance with a part of Barsbay's mamluks increased the power of the
atabak Jaqmaq. The second faction was headed by Amir 'Ali Bay, and was
composed of the mamalik al-julban al-ashrafiya (Barsbay); it supported the
son of the deceased sultan,'Abd al-'Aziz.1 Here it is beyond any doubt that
the mamluks of the former sultans, viz. the (Shaykh) the
(Faraj), and the (Barquq) of the first version, are identical with the
of the second version.2
Ibn 'Arabshah, in describing the same events, puts it even more briefly and
clearly. He says: 'The army as a whole became divided into two (rival)
sections. One section is said to be called and these are the
(Barquq) and the (Faraj) and the (Shaykh)…and the other
section the mamalik ashrafiya (Barsbay) living in the barracks of the
citadel' al-'askar fi al-jumla qismayn qism yuqal 'anhum annahum
wa-hum wal-mu'ayyadiya…wal-qism al-akhar
al-mamalik al-ashrafiya sukkan bil-qal'a).3
The long series of proofs furnished above by the present writer in support of
his argument that the are identical with the mamluks of the former
sultans finds its full confirmation in an unequivocal statement by Ibn
Zunbul, who says: 'All the julban…were the mushtarawat of (sultan )
al-Ghawri…He endeavoured to teach them (the Art of War), for his purpose
was to create an army of his own mamluks and break the i.e. the
mamluks of the sultans who preceded him' (fa-innahu kana fi ta'tim
al-julban wa-kana an yunshi'a lahu 'askaran min mamalikihi
mushtarawatihi wa wa-hum mamalik al-muluk alladhina
qablahu).4
We shall now return to some of Poliak's other arguments.
The claim that 'there were also among the mamalik 5 does not
bring out any distinguishing feature of this unit since, as mamluks of former
sultans, they were Royal Mamluks (mamalik ) in any case.
Neither does the statement that ' needless to say, they (the were
6
represented also among the amirs' lend them any distinctiveness, for other
units (including even the awlad an-nas and ajnad ) were also
represented among the amirs. In other words, private soldiers were
promoted to the rank of amir from other units, as they were promoted from
among the
As to the claim that the also belonged to the we have already
pointed out that only calls them al-ajnad in contradiction
to all we know of them from other sources. The measure of lack of
accuracy may be judged, inter alia, from his contention that in his days the
numbered less than 100.1 It is not difficult to refute this claim on the
basis of source material. First, the very role played by the as the main
rivals of the ajlab and as the bearers of the brunt of the fighting (see below)
eliminates the plausibility of such a small number. Second, the sources
furnish unequivocal data as to the numbers of the Thus in the
struggle of the against the mushtarawat in 802, the former number
about 1,000 men.2 In 919 every mamluk whose horse had died has his
mount replaced by Sultan al-Grhawri; 1,000 horses are distributed to
the on this occasion.3 In 920 an expeditionary corps of 2,400
horsemen, composed of and julban, is sent out; the julban account for
only 500,4 giving 1,900 In 922, a date close to that of the battle of
Marj Dabiq, the old and feeble (shuyukh wa-awajiz) among the are
posted as garrisons in all parts of Egypt and in some sectors of the Red Sea
coast, so that they might serve as a barrier against the 'urban. They are sent
to ash-Sharqiya, al-Gharbiya, al-Jiza, al-Manufiya, al-
Iskandariya, Rashid, and Upper Egypt to'Aqaba, Aznam,
and Mecca; they are also charged with the guarding of Egyptian dams.5
There can be no doubt that many hundreds of soldiers were needed to carry
out this tremendous task, for it should not be forgotten that to repress a
single attack or rebellion in only one province, 200 to 500 Royal Mamluks
of the first line were required; it stands to reason that the number of 'old and
feeble 'men needed was even greater. We find, in fact, that 500 old
are sent out on one expedition to Sharqiya,Gharbiya, and Upper Egypt,6
whereas 50 are dispatched to Mecca.7
Needless to say, all figures quoted above are but partial figures of the
and are considerably lower than their total number. Nevertheless,
they are much larger than the figure cited for the total number of the
by
As to the claim that 'naturally there were also among the julban',8 it
should be noted that Poliak supports it by a single piece of evidence;
whereas the sources supply us with an abundance of instances showing
clearly that the were different from the julban and that they fought
and clashed with them.1 In reality, however, even that single reference of
Poliak's proves precisely the opposite of what he intended it to demonstrate.
Here is the full text of the passage in question (year 860): 'Then came the
5th day after the 14th of the month, and the wazir2 Ibn fled. On
that day, no one received the meat rations allotted to the mamalik
that is to say, the julban (al-mamalik a'ni
al-julban), and their attendants and black slaves went up (to the Cairo
citadel), but they found nothing. Then they discovered that the julban had
received their rations, and the attendants and black slaves took the matter to
heart. They went down immediately, acted wantonly in the streets of Cairo
and pillaged a few shops.'3
The general meaning of this excerpt is clear and requires no comment. It is
one of the usual stories of the mistreatment of the and the special
favouring of the julban (as will be seen below). The latter received meat
while the former did not, which. provoked their servants to engage in some
mischief in the streets of the capital. If, however, we attempt to translate the
passage literally, we obtain a sentence in which there is an obvious logical
contradiction: the that is to say the julban, did not receive meat,
while the julban did receive meat. It is beyond any doubt that the text is
corrupt; and, indeed, the corruption was caused by the fact that between the
word 'a'ni' and the word 'julban' the word 'ghayr' was omitted by the editor.
This has already been pointed out by Popper in his above-mentioned
critique of Poliak.4 The text should read: ' a'ni ghayr al-julban,' i.e.
'the qaranis, that is to say, those who are not julban' did not receive the
meat rations allotted to them, while the julban did receive them. Thus the
contradiction disappears from the text, and a logical meaning is obtained
which. is opposite to that which Poliak attributes to the source; hence his
claim that may, inter alia, appear among the julban as well, is
automatically dispelled.
We have so far attempted to show the incorrectness of Poliak's claim that
the could belong to diverse units and ranks of the Mamluk army. We
shall now endeavour to prove that the contention that they held a position of
superiority in the kingdom cannot withstand criticism. Here also, we shall
anticipate our conclusions in order to make the argument clearer.
The actual position of the may be ascertained only by a consideration
of the interplay of two opposite tendencies within the Mamluk kingdom.
The first is the tendency to bestow great honour on seniority in both service
and age, and the second the proneness of the ruling sultan to favour his own
mamluks over those of the preceding rulers, who were their seniors in both
respects. With the deterioration of the kingdom, the second tendency
gradually becomes stronger than the first, and eventually displaces it, and
the aura of honour surrounding the veteran wanes considerably in the
course of time.
We have already mentioned elsewhere1 that the young mamluk looked upon
his senior khushdash with the greatest esteem and reverence, especially if
the latter served as his instructor (agha). Such. respect was not restricted to
companions in servitude and liberation, but was much more general. Cases
of lapses in the respect due to one's senior were severely criticized, and
labelled 'unheard-of behaviour'.2
A number of terms relating to seniority are of frequent usage in
contemporary sources, and they are worthy of mention in connexion with
the They are 'qadim', ' hijra','kabir'.
It seems to us that the original intent of the term ' hijra', or ' fi
3
al-hijra' was to indicate that the mamluk had been brought from his home
country to the Mamluk kingdom and embraced Islam a long time since.
There evolved, however, a secondary meaning, viz.'having long standing in
a particular branch of service', thus: 'wa-lahu hijra fi
min ayyam al-malik ibn Qalaun',4 or
'kana al-hijraf '.5The respect accorded to the al-hijra is
reflected also in the following examples. Amir Barsbay proposes the
sultanate to Amir Janibak on the ground of his being senior and veteran
among his colleagues (fa-innaka aghatuna wa-kabiruna
hijratan).6 It is said of two amirs that they are among the most honoured
and with longest service among the amirs (min ajall al-umara
hijratan); they sat at the right and left of the sultan respectively (i.e. in the
most honoured places) in official ceremonies.7 Baybars on giving
the list of Qalaun's highranking and veteran mamluk amirs, who were in his
service before his becoming sultan, says: al-mamalik
al-a'yan kanu fi min zaman
al-imra wa-lahum qidam dl-hijra fi al-'usra wal-yusra.8
From one of the above examples, the similarity between the meanings of
'kabir' and ' hijra' may already be deduced (see also below). The
veteran amir is also called amir kabir, min akabir al-umara', min al-umara'
al-kibar; the individual so called was sometimes an amir of long service
holding a high office (or an undefined honoured position), and sometimes
merely an amir with long service. The words of the historian concerning the
title amir kabir,
are of special importance. He states that this title had formerly been granted
'to all those who had seniority in service and in years' (li kull qadim hijra fi
al-khidma wa-fi ash-shaykhukha); consequently, there was a whole group
of amirs of which every individual was called amir kabir. This situation
remained unchanged until the days of Shaykhun al-'Umari, when the title of
amir kabir became reserved for the atabak al-'asakir only.1 The connexion
between long service and advanced years, on the one hand, and the title
kabir (sing. or pl.) on the other, may be seen from the following instances.
It is said of Kahardas b.'Abd-allah that he was 'min al-umara' wa-
2
akabirihim'. Frequent are the expressions: 'akabir' (or: 'kibar') al-umara'
al-mashayikh' 3 and ' as-sinn min akabir al-umara' '.4 Even after the
days of Shaykhun al-'Umari, the expression 'akabir al-umara'' in the sense
of ' amirs with. long service' is often encountered. Thus it is said of four
individuals holding the not very high rank of Amir of that they
were among the 'greatest' (!) of the amirs of Egypt (min akabir umara' al-
'asakir 5 In this category was included the governor of

Alexandria,6 who was generally an Amir of and sometimes even an

Amir of Ten. There is no doubt that in the last two examples, the meaning
of akabir is 'veterans, men with long service, old men'. It is even said of the
elders of the eunuchs (mashayikh that each was worthy of being
called kabir on account of his great age kullun minhum likibar
sinnihi an yud'a bi-l-kabir).7Seniority was, however, at times accompanied
by high position; thus, for instance, the expression 'the greatest of the amirs
in importance and (the oldest) in years' (akabir al-umara' qadran wa-
sinnan) is of frequent occurrence.8 We are told that Barquq, before he
became sultan, was in fear of the senior amirs (qudama al-umara'), and the
source immediately adds that Aqtimur was 'min akabir al-umara'' and that
Barquq used to sit in a position inferior to his during official ceremonies,
because of Aqtimur's seniority (qidam hijratihi).9 In almost all the instances
in which we succeeded in identifying the akabir al-umara', they were
mamluks of former sultans.10
In the light of the foregoing, the case of must be given an entirely
different interpretation from that placed on it by Poliak,1 who adduces this
case as the sole proof for his contention that the enjoyed greater
esteem than their colleagues and even than their superiors, and were the
first to be considered for promotion.2 The text of the reference is as follows:
'And when al-Malik (Barquq) wanted to give the above-mentioned
his liberation certificate, he caused him to stand in a review, together
with others from among the as yet unliberated mamluks of the military
school (al-mamalik al-kuttabiya). Now was of short stature, and the
sultan thought him too young he therefore returned him
to the military school together with the other young mamluks al-
mamalik). Amir Jarbash asli-Shaykhli a ra's nawba, was present at
the time, and he caught by his shoulder (?) 3 and said: "O our Lord the
sultan this is a learned man, a seeker of knowledge, a (hadha faqih,
'ilm, ) and he is therefore worthy of a horse (yasta'hil al-khayl)"4
Then al-Malik ordered that he be given a horse, and his liberation
certificate was written by Suwaydan, the Qur'an reader'.5
The meaning of this passage is quite clear. was worthy of being freed,
both from the point of view of age and from the point of view of his
learning in Islam.6 But because he was of short stature, the sultan thought
him too young and wanted to send him back to the military school; then he
reversed his decision and freed him as soon as his error had been made
clear to him. There is here, therefore, 110 question of any special privileges
granted to but only the granting of a legal right enjoyed by every
mamluk of the Mamluk kingdom from its inception until its fall: namely,
the right to be liberated and receive the status of a full-fledged soldier when
he had completed his training and reached maturity. Similarly, it is clear that
the term is here used in a sense opposite to 7 with the meaning of

'major, of age, veteran'. This meaning is brought home in a compelling

manner by the words of Ibn Tagkribirdi immediately following the quoted
passage. The historian goes on to say that if al-Maqrizi is right in his claim
that was liberated at a date posterior to that which. he (Ibn Tagkribirdi)
1
proposes, then Amirs X, Y, and Z would have been elders (which is
impossible): 'fayakun ha'ula' bi-n-nisba ila wa-akabir wa-
2
qudama 'hijra'. In other words, it is sufficient for mamluk A to be liberated
from the military school before mamluk B in order for him to be considered
in relation to the latter. Furthermore, we have here incontestable
evidence that is synonymous with kabir and qadim hijra, which terms
we have attempted to define above. It may be pointed out that also
according to definition, the is qadim hijra,3 perhaps the only
entirely accurate statement in that definition.
The following instance is also instructive. During a lance play (la'b )
of Sultan al-Ghawri's mamluks in 909, these mamluks are scorned by
the on account of the poor quality of their play in comparison with
the accomplished technique of their predecessors (ma kanat taf'aluhu al-
aqdamun min al-bunud allati kanat taqa' fi la'b 'ala al-ada, al-
4
qadima). The could make such a comparison because they were
older and had longer service than the mamluks of the ruling sultan, and had
taken part in, or been present at, the contests of previous reigns.
We have attempted to show that and senior or veteran are
synonymous. Since men with long service enjoyed some measure of
respect, the extent to which a was accorded honour and respect
derived from his seniority. The did not, in the long run, retain this
prestige, as a consequence of the ever-increasing ascendancy of the
mushtarawat-ajlab, which, since the reign of Barquq and especially since
the middle of the 9th century, overshadowed every other aspect of the
Mamluk state. There did remain a certain number of who were
honoured and respected until the very end of the Mamluk kingdom as
individuals; but the status of the as a group became progressively
5
worse, although they remained second only to the mushtarawat and
constituted the only body capable of offering them any resistance.1
The sources enable us to determine the status of the in relation to that
of the mushtarawat in three essential areas: in battle, in pay, and in the
distribution of feudal estates. In all of these, severe discrimination against
the is evident. We shall deal separately with each area.

A. and Mushtarawat in Battle

One of the distinct marks of the reduced fighting spirit among the
Circassians was the sultan's practice of relieving his mushtarawat of the
burden of war in order to place it on the shoulders of the mamluks of the
former sultans, especially those of the sultan whom he had succeeded. Thus
Aynal sends the Jaqmaq into battle, while Khushqadam sends the
aynaliya, Qaytbay sends the khusqadamiya, and al-Ashraf al-Ghawri
sends the mamluks of al-Ashraf and al-'Adil
2 Participation in the military expedition was considered a substitute for

banishment and exile.3 Instances are known in which expeditionary forces

were sent out without including a single one of the julban, a procedure
which aroused the wrath of the 4 In one such force, numbering 2,400

men, there were only 500 julban, the rest being 5 In the battle of

Amid (836) Sultan Barsbay wished to send the to lead the attack and
to leave his julban at the rear. He claimed that he was motivated by the
julban's ignorance of the ways of war; but the army was convinced that the
real reason was that he wished to spare his own mamluks and to increase
the number of casualties in the other categories of mamluks.6 It is also
mentioned in connexion with other battles that the ajlab fought indolently
in comparison with other units.7 This was especially evident in the battle of
Marj Dabiq, in which al-Ghawri incited his ajlab not to fight, so that the
brunt of the battle might fall on the And, indeed, most of the losses
suffered during the engagement were among the while the ajlab lost
only a handful of men 'for they did not fight in that battle at all, and showed
no chivalry and no bravery, as though they had been "timbers propped up"'
(fa-innahum lam yuqatilu fi hadhihi al-waq' a shay', wa-la lahum
furusiya fa-ka-annahum khushub musannada).1 The fact that the sultan
protected his own mamluks and spared them from participation in battle had
a disastrous effect on their fighting efficiency.2 According to this historian,
a single is worth ten ajlab in battle.3

B. Payments to the and the Mushtarawat

The iniquitous treatment of the stands out even more clearly in

matters of pay. Whereas one might expect a veteran soldier to receive
higher pay than a more recent recruit of the same rank, we are here
confronted with precisely the opposite procedure. The received at
best the same pay as the julban, and generally much less. They sometimes
received no pay whatever, while the julban received their full or partial
payments; the evidence is presented below.
In 837, while the julban received their customary pay, the and the
4
awlad an-nas received nothing. In 891 the julban received 50 dinars per
man, the 25.5 In 891 there is a bonus of 10 dinars per man for the
julban, 5 dinars for the 3 dinars for the sayfiya.6 In 894 the julban
receive 50 dinars, the 25 dinars.7 In 894 the julban receive 50 dinars,
the 25 dinars.8 In 901 the julban receive 100 dinars, the veterans
among the and the sayfiya receive 100 dinars, and those with
relatively less service, only 50 dinars.9 In 917 the mushtarawat receive 50
dinars, the and the sayfiya receive nothing. This act of the sultan's
almost precipitates a rebellion.10 In 918 only the and the julban receive the
full allowance of 3,000 dirhams for the clothes (Kiswa)11; this is one of the
rare instances in which the received equal pay with the ajlab. In 920
the do not receive payment for their meat rations for some six
months.12 In 920 the julban who were members of the expeditionary corps
were paid in Damascus, while the were not paid.13 In 920 the sultan
agrees to pay the julban 50 dinars, but he refuses to pay anything to the
 and the sayfiya.14 In 921 the julban receive 50 dinars, the
receive nothing.15 In 921 the julban receive 50 dinars, as do the mamluks of
Qaytbay, who are young and black-
bearded, while the older the sayfiya, the awlad an-nas, and the
soldiers of al-khamisa receive nothing.2 In 921 the
1 sayfiya,
and awlad an-nas receive no pay, unlike the julban.3 In 921 the weaker
and awlad an-nas receive no pay.4 In 922, when the sultan was
absolutely certain that the Ottoman sultan was about to attack his kingdom,
he did his best to pacify the and paid them for their meat ration and for
their horses the sums which had been delayed at the diwan.5 Nevertheless,
at Aleppo, on the day of the battle of Marj Dabiq, he pays the julban and
gives nothing to the and the awlad an-nas.6
These passages require no comment. We see here how the sources mention
the the sayfiya, and the awlad an-nas in the same breath as units wronged
and mistreated in matters of pay. Another striking example of the
oppression of the is provided by the following: al-Ghawri, who came to
power in 906, distributed horses to them for the first time only in 9187; thus
he neglected the renewal of this unit's mounts for twelve years.

C. The Distribution of Fiefs

The same picture obtains in connexion with the distribution of feudal

estates. The claim that under the Circassian sultans the 'had the right of
priority to fiefs, which was often contested by the freedmen of the reigning
sultan 'puts matters in a totally wrong light, and is contradicted by a number
of passages adduced by Poliak himself.8 What actually occurred was that
the ajlab robbed the of their estates, and did not shrink from callous and
murderous deeds. This they did without any interference, and generally
with the sultan's support or tacit approval. Some examples follow.
During the plague of 864 Sultan Aynal and his ajlab decided to transfer the
fiefs of the victims of the plague to the ajlab. Those who suffered most
from this plot were the Only those of them who made use of the stratagem
of registering the request for a vacant fief not only in their own name, but
also in the name of one of the ajlab of their acquaintance, received some of
these fiefs. During this plague the ajlab massed enormous quantities of
fiefs, which were taken from them after Khushqadam's rise to power and
were distributed among his mamluks.1 In other words, their estates were
taken from them as soon as they themselves became
In Dhu al-Qa'da 912 one of the mamalik became ill and was about to die.
Some of the ajlab hoped to acquire his fief after his death, but the regained
his health. The ajlab then killed him as he was on his way to the citadel, so
that they might obtain his fief. None was punished, although the fief was
handed over to others.2
The following information is most instructive: in 920, one of the mamalik
was found dead. He had been strangled with a wire, stripped naked, and
thrown upon the highway. His assailants were unknown, but it was said that
the ajlab had killed him, for they had similarly treated a great number
(jama'a kathira) of the on account of their fiefs. Those had
been slain without protest from any quarter (wa-lam fi dhaka skatan). In
those days the situation had become most unstable; the ajlab killed whom
they pleased on account of his fief, and if the murderer was brought before
the sultan, the affair would be ignored.3
From the foregoing, it becomes evident that the twin contentions, (a) that
the belonged to diverse ranks and units, and (b) occupied a specially
honoured position in the Mamluk state, are entirely without foundation.
There remains Poliak's claim regarding the status of the in Circassia or
rather both in Circassia and in the Mamluk kingdom, as expressed in the
following statements: 'the ruling power was practically held by the same
group in Circassia and in the Mamluk kingdom. This group is called in the
mamluk sources', or 'During the second period, the Circassian stood at the
head of the feudal element…of the region north of the Black Sea'.1 This
claim need not be refuted in detail for two reasons, first: Poliak does not
furnish the slightest proof to support it, and second: there is, to the best of
our knowledge, no scrap of information in the Mamluk sources from which
it might be inferred directly or indirectly that the constituted a superior
caste in Circassia. Further, there is no reason whatsoever to assume that the
had any closer connexions with Circassia than the or the sayfiya, or any
other group of mamluks which emigrated from its country of origin to the
Mamluk kingdom.
 LIST OF ABBREVIATIONS

Abu al Fida'
Abu al-Fida', Kitab fi Ta'rikh al-Bashar, Cairo, 1325 H.
Abu Shama
Abu Shama, Kitab Cairo, 1287–8 H.

as-Sakhawi, al-Lami', Cairo, 1353–5 H.

Al-Qalqashandi, Cairo, 1906.

Durar
Ibn al-'Asqalani, ad-Durar al-Kamina, Hyderabad, 1348–1350 H.
Duwal-al-Islam
Adh-Dhahabi, Duwal al-Islam, Hyderabad, 1337 H.
Fawat
Al-Kutubi, Fawat al-Wafayat, Cairo, 1299 H.

Ibn Taghribirdi, ad-Duhur, ed. Popper, Leiden, 1930.

Cairo, 1219 H.
Ibn al-Furat
Ibn al-Furat, Ta'rikh ad-Duwal wal-Muluk, Beirut, 1936–1942.
Ibn Iyas
Ibn Iyas, Bada'i' az-Zuhur, vols. i–ii, Cairo, 1311–12 H.; vols. iii–v, ed. Kahle-Mustapha, Istanbul,
1931–6.
Ibn Kathir
Ibn Kathir, cd-Bidaya wan-Nihaya, Cairo, 1351–8 H.
Ibn Khaldun
Ibn Khaldun, Kitab al-Ibar, Cairo, 1284 H.
Ibn Mamati
Ibn Mamati, Qawanin ad-Dawawin, Cairo, 1943.
Ibn Shaddad
Ibn Shaddad, an-Nawadir Cairo, 1376 H.
Ibn ash-Shihna
Ibn ad-Durr al-Muntakhab, Beirut, 1909.
Ibn Zunbul.
Ibn Zunbul, Kitab Ta'rikh Salim, Cairo, (litho.), 1278 H.

Al-Maqrizi, wal-I'tibar fi Dhikr wal-Athar, Cairo, 1270 H.

Manhal
Ibn Taghribirdi, al-Manhal Paris MS. (de Slane No. 2068–2072). The quotations from this
biographical dictionary refer to photographic reproduction of its MSS. belonging to the library of the
Hebrew University in Jerusalem. This reproduction comprises the Paris MS. of this work in its
entirety and the portions of the MS. of the Egyptian National Library which complete that MS. The
division of the work into volumes in the Jerusalem library is arbitrary. It is to this division that the
quotations figuring below refer.
Nahj
b. Abi An-Nahj as-Sadid (in Patrologia Orientalis, vols. xii, xiv, xx).
Nujum (P) Ibn Taghribirdi, an-Nujum az-Zahira, ed. Popper, Leiden, 1909 and onwards.
Nujum (C) Ibn Taghribirdi, an-Nujum az-Zahira, Cairo, 1934–1942.

ibn al-Jawzi, Mir'at az-Zaman, Chicago, 1907.

Al-Qalqashandi, Cairo, 1913–19.

Suluk
Al-Maqrizi, Kitab as-Suluk li-Ma'rifat ad-Duwal wal-Muluk, Cairo, 1934–1942.
Suluk (Quatremère tr.) Quatremère's translation of as-Suluk, Paris, 1837–1844.
Ta'rif
Ibn Allah al-'Umari, at-Ta'rif fi ash-Sharif, Cairo, 1312 H.
Ta'rikh Bayrut
b. Ta'rikh, Bayrut, Beirut, 1927.
Tibr
as-Sakhawi, at-Tibr al-Masbuk, Cairo, 1896.
Zetterstéen
Zetterstéen (ed.), Beiträge zur Geschichte der Mamlukensultane, Leiden, 1919.
Zubda
Khalil b.Shahin Zubdat Kashf al-Mamalik, Paris, 1894.
 Saladin and the Assassins

By BERNARD LEWIS

IN the year 577/1181–2, in a letter to the Caliph in Baghdad explaining his

activities in Syria, Saladin writes that he is engaged in a struggle for Islam
against a three-fold enemy—the infidel Frankish invader, the heretical and
murderous Assassins, and the treacherous Zangid rulers of Mosul, whom he
accuses of intelligence and even alliance with both Franks and Assassins.1
The story of Saladin's struggle against the Zangids and then against the
Crusaders is well documented and well known. On his dealings with the
Assassins, on the other hand, the sources tell us little, and most of
that refers to three episodes, as follows:— (1) The first Assassin attempt to
murder Saladin; Aleppo, Jumada II, 570/Dec. 1174-Jan. 1175.
Sources: Abu Shama, I, 239–240 (=De Sacy 358–9); Ibn al-Athir, xi, 276–8; Kamal ad-Din, MS. fol.
190a (=Blochet iii 563); Ibn MS. 179; 207; cf. Quatremère 354, Defrémery, 15–16.
The attempt was made during Saladin's siege of Aleppo. The Assassins had
managed to smuggle themselves into the camp, but were recognized by
ad-Din Khumartakin, the amir of Abu Qubais, who had had previous
dealings with them. Khumartakin challenged them, and was killed by them.
In the fracas that followed many soldiers were killed, but Saladin suffered
no harm. 'Imad ad-Din and Ibn Abi as quoted by Abu Shama, say that
it was the rulers of Aleppo who, when hard pressed by the besiegers, sought
the help of the Assassins and promised them estates and other rewards. Ibn
al-Athir, followed by Kamal ad-Din and Ibn 2 is more specific, and

names Sa'd ad-Din Gumushtakin, the regent of Aleppo, as having sent a

messenger to Sinan, the Assassin chief, promising rewards and asking
assistance.
(2) The second Assassin attempt to murder Saladin; 'Azaz, 11 Dhu'l-Qa'da
571/22 May 1176.
Sources: Abu Shama, I, 258 (=De Sacy, 360–5); Baha' ad-Din, iii, 62–3; Ibn al-Athir, xi, 285
(=Recueil, I, 623–4); Kamal ad-Din, MS. fol. 1926 (= Blochet, iv, 144–5); 212; Bustan, 1413;
Ibn MS. 190–1; Michael the Syrian, iii, 366; cf. Quatremère 354, Defrémery, v, 16–19.
During Saladin's siege of 'Azaz, Assassins, disguised as soldiers, penetrated
his camp and joined his army. On the date mentioned they attacked him, but
thanks to his armour he sustained only minor injuries. After a sharp struggle
the Assassins were killed, and thereafter Saladin took elaborate precautions
to protect his life. The sources for this second attempt are more numerous
and more detailed than for the first, but contain no major disagreements.
Abu Shama quotes three accounts, taken from 'Imad ad-Din, Ibn Abi
and a letter of the written to al-'Adil. Of these only Ibn the
latest of the three—accuses the rulers of Aleppo of inciting the attempt:
'When the Sultan conquered the fortresses of Buza'a and Manbij, the rulers
of Aleppo realized that they were losing the strongholds and castles which
they had held, and they returned to their practice of weaving plots against
the Sultan. They wrote a second time to Sinan, the chief of the Assassins,
and induced him with. money and promises to attack the Sultan…' Baha'
ad-Din, Ibn al-Athir, and the others describe the incident in more or less
detail and in much the same terms, but make no reference to any instigation
from Aleppo.
(3) Saladin's attack on 572/July 1176.1
Sources: Abu Shama, i, 261 (=De Sacy 365–6); Ibn al-Athir, xi, 289 (= Recueil, i, 626); Kamal ad-
Din, MS. fol. 193a; Ibn MS. 192–3; 212; cf. Defrémery, 19–20.
After these two attempts on. his life Saladin, thirsty for vengeance, invaded
the Assassin territories. He laid siege to on 20 572/30 July
1
1176. Then, on the mediation of his maternal uncle Shihab ad-Din
ibn Takash, governor of Hama and a neighbour of the Assassins, Saladin
made a truce with Sinan and withdrew his forces. There is some conflict
between the sources on the circumstances of the mediation and the truce.
According to 'Imad ad-Din, as quoted by Abu Shama, Saladin wrought
havoc and destruction in the Assassin lands. The Assassins wrote to Shihab
ad-Din to ask his help as a neighbour. He then interceded with Saladin to
spare them, and Saladin, satisfied with his revenge, agreed to withdraw. Ibn
follows 'Imad ad-Din fairly closely. Much the same story is told by Ibn
al-Athir, who adds the detail—possibly derived from a fuller version of
'lmad ad-Din than that cited by Abu Shama—that Sinan threatened to
murder Shihab ad-Din 'and all the people of Saladin 'if he refused to
intercede on their behalf. Ibn al-Athir hints that Saladin's readiness to
withdraw was due to the weariness of his troops and their desire to return
home to enjoy their booty. Ibn Abi remarks that the main reason for
Saladin's withdrawal was a dangerous Frankish advance in the Biqa'. It was
to meet this threat that 'he made terms with Sinan and returned to
Damascus'. According to Kamal ad-Din Saladin 'advanced into the country
of the Isma'ilis and laid siege to them; then he made peace with them
through his uncle ibn Takash and marched with his troops to
Egypt…' One of the terms of the truce was the release of the brothers Ibn
ad-Daya, who joined Saladin.1
Two questions arise from these events: why did Sinan suddenly take the
offensive against Saladin in 570/1174–5, and what were the circumstances
and the terms of the truce signed between them in 572/1176. Most of the
sources, as we have seen, attribute Sinan's first attack to the instigation and
bribery of Gumushtakin. That Sinan acted in concert with Gumushtakin, or
received help from him against an enemy that threatened both of them, is by
no means unlikely. But the inducements of Gumushtakin can hardly have
been the primary motive of Sinan, who was the leader, not of a mere band
of cut-throats, but of a religious order with far-reaching objectives of its
own.2 A more direct reason for Sinan's action may possibly be found in a
story told by ibn al-Jawzi, though not, oddly enough, by the
contemporary chroniclers. In 570/1174–5, according to 10,000
horsemen of the anti-Shi'ite Nubuwiya order of Futuwwa3 from Iraq raided
the Isma'ili centres in Bab and Buza'a, where they slaughtered 13,000
Isma'ilis4 and carried off much booty and many captives. Profiting from the
confusion of the Isma'ilis, Saladin sent his army against them, raiding
Sarmin, Ma'arrat and Jabal as-Summaq, and killing most of their
5
inhabitants. The raid of the Nubuwiya is also mentioned independently by
Ibn Jubair,6 Ibn Shaddad,7 and the Isma'ili writer Abu Firas,8 though none
of these makes any reference to Saladin's attack. unfortunately does not
say in what month these events took place—the position in which he places
his narrative, shortly after the attempt at Aleppo, is of course no guide to
the real sequence of events. There is therefore nothing to show whether
Saladin's raid on the Assassins took place before or after their attempt to
murder him at Aleppo. There is little to choose as regards probability. The
attempt took place in Jumada II, half-way through the Muslim year 570,
leaving about as much time before as after. It is possible that Saladin sent
his raiders while his army was marching northwards towards Aleppo—it is
equally likely that he sent them down from Aleppo, to give encouragement
and booty to his troops.
Whether or not the first act of aggression came from Saladin, his activities
and policies generally made him a potentially dangerous enemy to the
Isma'ilis, and would be sufficient to explain their attack on him, even if
immediate provocation was lacking.
In 567/1171 Saladin had suppressed the last remnant of the Fatimid
Caliphate in Cairo, and restored the suzerainty of the Abbasid Caliph. The
suppression in itself was of no consequence to the Nizari Isma'ilis, to whom
Sinan and his followers belonged. After the murder of Nizar, Musta'li and
his successors were regarded as usurpers by the Nizaris; the last four
Fatimid Caliphs in Cairo were not accepted as Imams by any part of the
Isma'ili sect. But the circumstances of Saladin's abolition of the Fatimid
Caliphate cannot have failed to mark him down as an enemy of the whole
Isma'ili, indeed the whole Shi'ite cause. The suppression of Isma'ilism in
Egypt; the destruction of the great Fatimid libraries of Isma'ili works—
many of them common to all branches of the sect; above all, the restoration,
after two centuries, of the in the name of the hated Abbasids, all
showed that a new power had arisen who was no longer content to play the
political game of his predecessors, but was determined to restore the unity
and orthodoxy of Islam, and re-establish the supremacy of the Sunni Caliph
in Baghdad as head of the Islamic world.1
In 569/1174 pro-Fatimid elements in Egypt, led by the Yemenite poet
'Umara and some others, organized a conspiracy to overthrow Saladin and
restore Fatimid rule, and, for this purpose, sought the help of the Crusaders.
In a letter to Nur ad-Din, drafted by the Saladin reported on this
conspiracy and its suppression, and stated that the conspirators had written
to Sinan, arguing that their doctrines were basically the same and their
differences trivial, and urging him to attack Saladin.2 Sinan owed no
allegiance to the Cairo Fatimids, but an appeal to him on their behalf is by
no means unlikely. Some half a century previously the Fatimid Caliph Amir
had attempted without success to persuade the Syrian Isma'ilis to accept his
leadership, and had entered into arguments with them to that end.3 That
Sinan, for reasons of his own, agreed to collaborate with the Egyptian
conspirators is not impossible, though it is unlikely that he would have
continued to act in their interest after the definitive crushing of the plot in
Egypt. But the significance of Saladin's policies would have been brought
home to him—the march into Syria and the attack on the Zangid cities
showed that the danger was immediate.
And then after three years of conflict, came the truce at The sources
agree that there was a truce, that Shihab ad-Din ibn Takash acted as
go-between, and that Saladin then withdrew. Most of the sources say that
the request for terms came from Sinan, though Ibn Abi and, more
strongly, Kamal ad-Din imply the reverse. The Isma'ili biography of Sinan
by Abu Firas gives another, more fanciful version of Saladin's attack and
withdrawal. In this Saladin, terrified by the supernatural antics of Sinan and
his henchmen, retreats in disorder, leaving all his arms and equipment
behind. Through. the mediation of the prince of Hama, here called Taqi ad-
Din, Sinan grants a safe-conduct to Saladin, who 'became his friend after
having been his enemy'.1 Abu Firas's book is full of miracles and marvels,
and is obviously legendary. It was written at a time when the Assassins had
become respectable members of Syrian society, and were anxious to defend
themselves against charges of disloyalty to Islam.2 It is therefore natural
that Abu Firas should depict his hero as a friend and collaborator of Saladin
in the Jihad against the Crusaders and thus rebut the accusation that the
Assassins had been traitors to the Muslim cause. Yet with all its absurdities
and its fantasies Abu Firas's narrative of the truce at obviously rests
on a foundation of local historical recollection. In this, as in his other
anecdotes, Abu Firas is independent of the Sunni historians, with whose
works he was probably unacquainted. The very confusion of Saladin's uncle
and nephew—Shihab ad-Din and Taqi ad-Din—suggests that he was
relying on local tradition rather than on the written sources. The same local
recollections underlie some of the stories collected by Kamal ad-Din in his
biography of Sinan in the Bughya. These describe how Sinan and his
emissaries demonstrated their irresistible power, and end with such
significant sentences as: 'We returned to Saladin and informed him of what
had happened, and thereupon he made peace with Sinan' and 'And
thereupon Saladin inclined to make peace with him and to enter into
friendly relations with him'.3
Of the terms of the truce we have no certain knowledge. Kamal ad-Din
mentions only the release by Sinan of the brothers Ibn ad-Daya, who joined
Saladin, but apparently remained in friendly relations with Sinan4—the
other sources add nothing to this scrap of information. But this much is
clear—that for the next seventeen years, until the death of both Sinan and
Saladin, neither of them took any kostile action against the other,1 and
Saladin was left unimpeded to overwhelm first his Muslim and then his
Christian enemies. Sinan and his followers still make a few appearances in
the general histories, which record the suppression by Sinan of a group of
his own extremists in 572/1176,2 the murder of Ibn al-'Ajami in Aleppo in
573/1177,3 Assassin. incendiarism in Aleppo in 575/1179–80, as a reprisal
for the seizure of Hajira by al-Malik 4 and, most striking of all, the

murder of the crusading chief Conrad de Montferrat in Tyre in 588/1192.5

Only the last of these is attributed to the instigation of Saladin, and then
only by Ibn al-Athir and Abu Firas, both suspect for different reasons; while
'Imad ad-Din, on the other hand, points out that Conrad's death came at an
inopportune moment for Saladin. But none of these actions was contrary to
his ultimate interests, and the first, carried out immediately after the truce,
may well have been a direct consequence of it. Four months after the
murder of Conrad a truce was signed between Richard Cœur de Lion and
Saladin in which, at Saladin's request, the Assassin territories were
included.6
References are given to the following editions and manuscripts:—
Abu Firas, Manaqib al-Mawla Rashid al-Din, in Stanislas Guyard, Un grand maître des assassins au
temps de Saladin, Paris, 1877 (reprinted from J.A. 7th series, ix, 324–489).
Abu Shama, Kitab fi Akhbar ad-Dawlatain, 2 vols., Cairo, 1287–8. Abridged German
translation by E.P.Goergens, Zur Geschichte Salahaddins, Berlin, 1879. Several of the relevant
passages were published and translated in Silvestre de Sacy, 'Mémoire sur la dynastie des assassins et
sur l'étymologie de leur nom', Mémoires d'histoire et de littérature orientale, Paris, 1818, 322–403.
Baha'ad-Din, Sirat ad-Din, Recueil H.Or.iii.
Bar-Hebraeus, Chronography, translated by E.A.W.Budge, Oxford, 1932.
Bustan in C.Cahen, 'Une chronique syrienne du VIe/XIIe siècle: le Bustan al-Jami' ', Bull. d'Et. Or.
de l'Inst. fr. de Damas, vii–viii, 113–158.
Ibn al-Athir, Al-Kamil fi't-Ta'ri, ed. J.C.Tornberg, Leiden-Upsala, 1851–1876. Extracts with French
translation in Recueil des historiens des croisades. Historiens orientaux, I, Paris, 1872–1906.
Ibn Jubair, The Travels, ed. W.Wright, rev. M.J.de Goeje, Leiden-London, 1907. English translation
by R.J.C.Broadhurst, London, 1952.
Ibn Shaddad, Al-A'laq fi dhikr Umara' ash-Sham wa'l-Jazira, MS. Istanbul, Revan 1564.
Ibn Mufarrij al-Kurub fi Akhbar bani Ayyub, MS. Cambridge 1079.
'Imad ad-Din, al-Qussi al-Qudsi, ed. C.Landberg, Leiden, 1888.
Kamal ad-Din, (a) Zubdat fi Ta'rikh MS. Paris, 1666; extracts translated into French by
Blochet in Revue de l'orient latin, iii and iv, 1895–6; (b) Bughyat fi
Ta'rikh extracts in B.Lewis, 'Three Biographies from Kamal al-Din', Mélanges Fuad Koprülü,
Ankara, 1953, 325–344.
Michael the Syrian, Chronique, translated into French by J.B.Chabot, 4 vols. Paris, 1899–1910.
ibn al-Jawzi, Mir'at az-Zaman, ed. J.R.Jewett, Chicago, 1907, and MS.Istanbul Saray 2907c xiii.
C.Defrémery, 'Nouvelles recherches sur les ismaéliens ou bathiniens de Syrie', J.A. 5th series, v,
1855, 5–76.
B.Lewis, 'The Sources for the History of the Syrian Assassins', Speculum, xxvii, 1952, 475–489.
E.Quatremère, 'Notice historique sur les ismaéliens', Fundgruben des Orients, iv, Vienna, 1814, 339–
376.
THE POSITION AND POWER OF THE MAMLUK SULTAN

By P.M.HOLT

I could not say much of the Mamalucs, of whom I know no author that has written in particular:
neither did they deserve that any should. For they were a base sort of people, a Colluvies of slaves,
the scum of all the East, who, having treacherously destroyed the Jobidae, their Masters, reigned in
their stead; and bating that they finished the expulsion of the Western Christians out of the East
(where they barbarously destroyed Tripoli, and Antioch, and several other Cities) they scarce did
anything worthy to be recorded in History.
Humphrey Prideaux (1722)1
The Mamluk state, as it was constituted after the defeat of the Mongols at
'Ayn Jalut (658/1260) and the annexation of Muslim Syria, resembles both
territorially and structurally the Ayyubid dominions which had preceded it.
There were, however, important differences. Under the Ayyubids the
territories had been partitioned among members of the ruling clan; at any
given time (at least after the death of Saladin) the territorial settlement was
unstable, although a series of Ayyubid rulers whose power was based in
Egypt—first al-'Adil Sayf al-Din, then al-Kamil, lastly Ayyub—
exercised a somewhat precarious paramountcy over their kinsmen in Syria.
By contrast the Mamluk state was one and indivisible; no part of it (with
one or two temporary and insignificant exceptions) was granted away in
appanage; no rival ruler succeeded in establishing himself in the territories
of the Mamluk sultan.
The Mamluk dominions are usually termed in the contemporary sources al-
diyar wa'l-mamalik al-Shamiyya 'the Egyptian territories and the
Syrian provinces', a distinction which may imply a technical difference
between the Egyptian and the Syrian constituents of the state, or may be
merely an affectation of rhetoric. Throughout the state the sultan's power
was delegated to provincial officials, each of whom bore the title na'ib
'deputy of the sultanate', but who were not of equal standing.
Highest among them was the vicegerent in Egypt, followed by the governor
of Dainascus, then by him of Aleppo, then of Tripoli, and al-
Karak. All these governors were normally Mamluks, although. Ayyubid
princes were permitted to linger on at al-Karak and until 661–2/1263,
and even longer at There in 710/1310, after more than a decade of
rule by Mamluk governors, the sultan b.Qalawun granted the
province to an Ayyubid prince, al-Mu'ayyad Isma'il, bettcr known as Abu 'l-
Fida' the chronicler. He succeeded in retaining the confidence of
who in 720/1320 granted him the insignia and title of sultan,
which in due course were inherited by his son. This signified little: it was a
mark of personal favour, not an experiment in indirect rule. In 742/1341,
after death, the son of Abu 'l-Fida' lost his sultanate and
was given a command of a Thousand (the highest military rank) in
Damascus, while passed again and permanently under a Mamluk
governor.
The Mamluk sultan was in an obvious sense the successor to the Ayyubid
rulers of Egypt and Syria. Three of the early Mamluk sultans, al-Mu'izz
Aybak (648–55/1250–7), Baybars (658–76/1260–77), and
Qalawun (678–89/1279–90), whose combined reigns cover 34 of the first
40 years of the regime, had been members of the Mamluk household of
the last effective Ayyubid sultan in Egypt. The Mamluk rulers
necessarily derived from their predecessors concepts of their office, as well
as of the administrative structure of the state over which they ruled.
The nature of the Mamluk sultan's position is indicated by a series of
observances at the time of his accession. The beginning of al-Mu'izz
Aybak's reign is described by al-Maqrizi in these words: 'The amirs and the
assembled for counsel, and they agreed to install the amir 'Izz al-Din [Aybak], the
commander of the guard, in the sultanate. They gave him the title of al-Malik al-Mu'izz…. They
caused him to ride 011 Saturday at the end of Rabi' II. The amirs in turn bore the saddle-cover [al-
ghashiya] before him to the Citadel, and they sat at the banquet with him'.2
Ibn Taghribirdi gives a slightly different series of events:
'They [sc. the amirs] swore allegiancc to him [baya'uhu], madc him sultan, and seated him upon the
bench. of kingship [dast al-mulk]…. The saddlecover was borne before him, and he rode with the
insignia of the sultanate. The first who bore the saddle-cover before him was the amir al-Din
b.Abi 'AIi, then the great amirs took it in turn, one after another. He was mentioned in the and
proclamation of his sultanate was made in Cairo and Old Cairo'.3
Putting these two accounts together, we arrive at some such pattern of
events as the following.
(1) The election of the sultan by a group of Mamluks—this is also
mentioned by Ibn Taghribirdi before the passage translated above.
(2) A series of events at the time of the election, viz.
(a) the assumption by the sultan-elect of a malik-title;
(b) the taking of an oath of allegiance (bay'a) by the electors;
(c) the enthronement of the sultan.
(3) A state procession, the sultan riding, through the capital, with the
saddlecover borne before the sultan by the amirs in turn.
(4) An accession-banquet
(5) The mention of the sultan's name in the
Some, but not necessarily all, of these observances are noted at the
inauguration of subsequent sultans down to the last of them, al-Ashraf
Bay, who was installed when Selim the Grim had already occupied Syria.
Aybak was raised to the sultanate by a group which al-Maqrizi describes as
consisting of 'the amirs and the i.e. his khushdashiyya who had been
members with him of the household of Ayyub. Election continued to
be the most usual title to succession throughout the Mamluk period,
although the procedure was never formalized, nor was the electoral body
ever defined—no Golden Bull was ever promulgated in the Mamluk
sultanate. It is therefore not surprising that the power to elect was on
occasion arrogated to itself by a victorious faction, and was used to condone
usurpation. Thus in 678/1279 Qalawun, with the agreement of the amirs and
the (i.e. the Mamluks of the palace) deposed the infant sultan al-
'Adil Salamish b.Baybars, and usurped the throne. In 696/1296 a faction of
amirs overthrew al-'Adil Kitbugha (who was himself a usurper), and
installed their own leader, Lachin. A similar coup had been carried out in
693/1293 against al-Askraf Khalil b.Qalawun, who was murdered while
hunting. On that occasion, however, the faction failed to take Cairo, and its
leader (and sultan-elect) was himself put to death.
Although the sultanate was usually elective, some of the stronger rulers
ensured that their sons should succeed them on the throne. There were two
procedures for accomplishing this. The first was the association of the son
as joint (and nominal) sultan with his father. So in 662/126–1
Baybars, apparently on the advice of the amirs, raised his son, al-Sa'id
Baraka Khan, to the sultanate. The same procedure was followed by
Qalawun, first with respect to his son 'Ali, who predeceased him, then
to al-Ashraf Khalil. The second device for ensuring an hereditary
succession was the testamentary nomination of a son by a sultan when
dying. Three days before his death in 741/1341, the great sultan
 b.Qalawun, was persuaded by his senior amirs to nominate a
son as his successor. His choice fell on Abu Bakr, who was enthroned as al-
Malik on the day of his father's death. It may be noted that several
years earlier, in 732/1331, had convoked the amirs, the
judges, and the caliph to recognize another son, Anuk, as his heir, but had
changed his mind at the last minutc. Such hesitation to inaugurate as son,
who might become a rival, is not surprising. Indeed, it seems that al-Askraf
Kkalil's status as joint sultan was technically defective, as his father
repeatedly refused to validate the diploma of appointment. The safer
procedure of nomination in articulo mortis was also utilized by
Barquq, the first of the Circassian line of sultans, in 801/1399. On his
deathbed, he convoked the caliph, the judges, the amirs, and the great
officers to swear to the succession in turn of his three sons, Faraj, 'Abd al-
'Aziz, and Ibrahim.
A cursory glance at the line of Turkish Mamluk sultans, who reigned from
648/1250 to 784/1382 (with a brief sequel seven years later), would suggest
that in this period the hereditary principle was more strongly entrenehed
than in fact it was. Al-Mu'izz Aybak was succeeded by his son 'Ali;
al Baybars by two sons, al-Sa'id Baraka Khan and al-'Adil Salamish;
and Qalawun by four generations of his direct descendants. In mere
longevity the Qalawunid dynasty surpassed its Ayyubid predecessor. This
appearance of dynasticism is, however, specious and misleading.
'Ali b.Aybak was installed as a convenient figurehead—a device which was
frequently to be employed in the following two and a half centuries. Al-
Sa'id Baraka Khan was, as we have seen, joint sultan with his father, but his
brother and successor, al-'Adil Salamish, was installed at the behest of
Qalawun until the latter could usurp the throne. Of the long scries of
Qalawunids, only al-Ashraf Khalil and Abu Bakr were designated
by their fathers to succeed them; the other sultans of this house were
nominees of the great amirs or even of a court faction. The long
continuance of the Qalawunids is hardly to be explained by any residual
loyalty of the Mamluks towards the family, but rather by the convenience of
the nominal sultanate as a for the oligarchy of the amirs.
The appearance of dynasticism in the Turkish Mamluk period forms a
contrast with the line of Circassian Mamluk sultans who reigned from
784/1382 to 922/1517. Concerning them, Stanley Lane-Poole remarked in
his Mohammadan dynaslies, 'As there are seldom more than two kings of a
family…a genealogical table is unnecessary'.4 His statement is formally
correct, but it is unilluminating, since it disregards the fact that in this
period the group within which the succession to the sultanate passed was
not the blood-family but the household, composed of both the heirs of the
body and the Mamluks of the founder. A pedigree of the Circassian sultans
constructed on this basis shows, first, that (with the brief and anomalous
exception of the sultan-caliph al-Musta'in in 815/1412) all the later sultans
down to the Ottoman conquest were linked to Barquq by natural or
Mamluk affiilation; secondly, that of the seven sons who succeeded their
fathers in this period, only one, Faraj b.Barquq, was an effective
ruler, while none of them founded Mamluk households from which later
sultans emerged; thirdly, that the housohold founded by al-Ashraf Qa'it Bay
was in its turn a nursery of sultans, producing Qa'it Bay's son and his five
Mamluk successors, who reigned until the coming of the Ottomans. Seen in
this light, as a synthesis of members by blood and members by Mamluk
recruitment, the succession inaugurated by Barquq resembles the great neo-
Mamluk households of Ottoman Egypt, in which, however, the natural
descendants had a far stronger position vis-à-vis the Mamluks.
Thus in the Mamluk sultanate the concept of an hereditary monarchy failed
to establish itself against a rival view of the state as a crowned republic, an
oligarchy of magnates in which the throne would pass by election or
usurpation to one of the amin. Yet it should be observed that the
competition for the sultanate was not open to all: it is simply not true to say,
as does Wiet, 'C'est un monde bien étrange que ce milieu des Mamlouks,
qui, presque tous, croyaient "porter dans leur giberne "le sceptre du
sultanat".5 It was only from among the Royal Mamluks, and indeed from a
small inner circle of these, that the candidates for the sultanate emerged.
To pass now to a consideration of the accession observances. The
assumption of a malik-title is obviously derived from Ayyubid usage, which
itself had Saljuqid precedents. But there was a significant difference.
Among the Ayyubids, Malik-titles were borne by members of the clan who
were not actually rulers, as well as by the sultan and the ruling princes. This
custom survived the end of Ayyubid domination—an Ayyubid who died in
727/1326–7, and whose highest rank had been that of an amir of a Hundred
in Damascus, was styled al-Malik al-Kamil.6 Among the Mamluks, by
contrast, a maliktitle was hardly ever borne by anyone but the sultan
himself. The only exception known to me before the Qalawunids is the third
son of Baybars, who never reigned but was styled al-Malik al-
Mas'ud. This was a courtesy-title, bestowed on him in 678/1280 by
Qalawun in respect of his appanage of al-Karak, perhaps because this had
been an independent lordship under the Ayyubids.7 Within his own family,
Qalawun seems to have reverted to the Ayyubid practice. In his treaty of
684/1285 with Leon III of Lesser Armenia, two sons are associated with
him in the preamble, and they both have malik-titles, viz. al-Malik 'Ali
and al-Malik alAshraf Khalil.8 It seems, further, that the title of al-Malik
was conferred at birth on another son, 9 the future sultan,
who many years later similarly gave the title of al-Malik to his
own new-born son 10 One other anomaly is the style of al-Malik al-
Amjad, given by Ibn Taghribirdi to a son of who never
11
reigned. Here the title may have been conferred retrospectively, since a
son of did become sultan as al-Ashraf Sha'ban. Apart from these
Qalawunids, there is one other exception in the Turkish Mamluk period:
the son of the usurper al-'Adil Kitbugha, is styled al-Malik al-Mujahid,
although he did not come to the throne.12 The close counexion between
accession and the assumption of a malik-title is borne out by the practice of
unsuccessful usurpers, such as Sanjar (658/1260) and Sunqur al-
Ashqar (678/1279), rebels against Baybars and Qalawun
respectively, and Baydara, the killer of al-Ashraf Khalil (693/1293). All
three took malik-titles.13
The oath taken by the amirs and others on the accession of a sultan might
be either of two kinds: the oath of allegiance to the ruler as sovereign
(bay'a), or the sworn convenant to support him personally It is
perhaps significant that, according to al-Maqrizi, the oath taken to the
puppet-sultan 'Ali b. Aybak was of the latter kind.14 The covenant
might indeed be of a mutual character between the new sultan and the
amirs. There are two good examples of this, both arising out of a usurpation
of the sultanate. The contemporary biographer of Baybars, Ibn 'Abd
describes how, when Baybars had killed and sought
recognition as sultan, 'he said to some of the company of amirs, "Take the covenant ".
The atabak said to the sultan, "Most of this company have been retired and have undergone hardship.
Only let the sultan covenant with them [ lahum] to be as munificent to them as he can, as soon
as he can, and after that they will take the covenant to the sultan "…. The sultan took the
covenant to them, and after that they took the covenant, and the Mamluks [al-nas] took the covenant
until late afternoon'.15
A second example is taken from the coup d'état of 696/1296, when Lachin
ousted his sultan and former colleague, al-'Adil Kitbugha, and usurped the
throne. His associates, however, imposed conditions upon him: that he was
to be primus inter pares, that he was not to act on his sole discretion, and
that he was not to give one of his own Mamluks power over them. Then in
the words of the contemporary chronicler, Baybars al-Dawadar
'Lachin repeated the covenant to them that he would not do so, and
thereupon they covenanted with him [ lahu]'.16
Another contemporary, Abu 'l-Fida', gives a very similar, but independent,
description of the incident, ending with the words: 'Lajin responded to them on this
[i.e. the set of terms], and covenanted with them upon it. Thereupon they covenanted with
him and swore allegiance to him as sultan [ lahu wa-baya'Zuhu bi ]'.17
The distinction and the relation between the two oaths could hardly be
shown more clearly.
The ritual act which marked the inauguration of a new sultan was his
enthronement. Its importance is indicated in the detailcd account which Ibn
'Abd gives of the accession of Baybars, which took place in unsual
circumstances, when the Mamluk army was returning to Egypt after the
victory at 'Ayn Jalut. After the killing of Baybars and his companions
entered the royal pavilion with the atabak, and a discussion. took place. In
the end, the company recognized Baybars's claim to the sultanate, and at
that point Baybars took his seat upon the royal cushion. Normally, of
course, the enthronement was a formal act in the Citadel of Cairo.
We come now to the state procession of the new sultan, during which he
rode through Cairo to the Citadel, accompanied by the amirs (who are often
described as going dismounted on this occasion), and preceded by the
insignia of the sultanate, among which the saddle-cover (al-ghashiya) is
particularly mentioned. In it we have an emblem of sovereignty derived
directly from the Ayyubids and ultimately from Saljuqid usage. Al-
Qalqashandi describes the ghashiya as 'a saddle-cover of leather, decorated with gold, so
that the observer would take it to be made entirely of gold. It is borne before him [sc. the sultan]
when riding in state processions for parades, festivals, etc. The rikabdariyya carry it, the one who
holds it up in his hands turning it to right and left. It is one of the particular insignia of this
kingdom'.18
Becker, who discussed the ghashiya in a short article over 60 years ago, saw
in this an object of religious significance, intended primarily to cover the
hands as a sign of submission. It is surely more appropriately regarded as a
symbol of authority natural to rulers who originated as horse-riding
warriors. If one seeks a Western analogy to the ghashiya, it is not the
episcopal glove, as is suggested in Becker's article, but the golden spurs
borne before the sovereign in the English order of coronation.19
Not only the saddle-cover, but also its display in a state procession, had
Ayyubid and even Saljuqid precedents. An instance from the Ayyubid
period was in 589/1193, when al-'Adil Sayf al-Din, Saladin's brother,
installed his nephew, al-'Aziz 'Uthman, as sultan in Egypt, and 'he walked
before him with the saddle-cover', as Ibn Taghribirdi tells us. Again, when
Ayyub was a claimant for the sultanate in 636/1238–9, the saddle-
cover was borne before him in Damascus by his cousin.20 The two Ayyubid
puppet-sultans of Abu 'l-Fida' and his son, were by one of the ironies
of history accorded the prerogative of the ghashiya by their overlord,
21

As long as there was a caliph in Baghdad, a new sultan was formally

confirmed in office by him. At the time of Baybars's accession there was, of
course, no caliph, but interesting developments ensued when, a few months
later, he received in Cairo a refugee 'Abbasid prince whom, after due
formalities, he installed as caliph with the title of bi'llah. To him
he took the oath of allegiance (bay'a) on 13 Rajab 659/13 June 1261.
in turn invested the sultan with the rule over the Islamic lands and
those which he would conquer from the infidels, i.e. the Mongols and the
Crusaders. The territories were specifically named in the diploma of
investiture, which was publicly read on 4 Sha'ban/4 July, as Egypt, Syria,
Diyar Bakr, the the Yemen, and the lands of the Euphrates, together
with any new conquests. A similar procedure was followed at the
installation of the second 'Abbasid caliph in Cairo, bi-amr Illah, in
22
661/November 1262. There-
after the caliph was technically indispensable at every accession to give
formal authority to the new sultan. A good description is given by Ibn
Taghribirdi of the procedure at the aceession of Abu Bakr on 2
742/18 June 1341: 'The judges went up [sc. to the Citadel] and the Caliph
bi-amr Illab Abu 'l-'Abbas took his seat on the third step of the sultan's throne. He wore a
green robe, and on his turban was a black covering with stripes of gold. Then the sultan came out into
the hall by the secret door, according to custom. The caliph, the judges, and the seated amirs stood up
for him, and he took his seat upon the first step, below the caliph. The caliph stood up and began the

At the conclusion of the the caliph addressed the sultan saying, 'I have
delegated to you all the jurisdiction of the Muslims, and invested you with
that wherewith I am invested in matters of the Faith'. Ibn Taghribirdi
continues: 'He sat down. Immediately a black robe was brought, and the caliph with his own hand
robed the sultan. Then he girt him with an Arab sword, and the judge 'Ala' al-Din 'Ali b.
the confidential secretary, read the caliph's diploma to the sultan in its entirety. He then presented it to
the caliph, who signed it, followed by the signatures of the chief judges as witnesses. The banquet
was then spread. They ate, and the court dispersed'.23
Returning to the delegation of powers by the Caliph to al-
Baybars, one must emphasize that the listing of the territories conferred on
the sultan, all of which he did not actually possess, was not simply
rhetorical hyperbole. Their mention served two purposes, of which the first
was to outline and publicize a programme of expansion, particularly into
lands under Mongol domination. It is perhaps curious that tbere is no
mention of the coastlanda still under the rule of the Latin Kingdom
and Antioch-Tripoli—but at that date the Mongols were a far greater danger
than the Franks. The second purpose of the list is to represent Baybars as
being not merely sultan of Egypt, or even (like his Ayyubid predecessors)
of Egypt and Syria, but as the universal sultan of Islam. This claim may not
have been entirely new. The chronicler Ibn al-Dawadari, writing of events
before his own time, and perhaps not writing advisedly, entitles 'Ali
b.Aybak al-Islam in 657/1258–9 (i.e. after the extinction of the
caliphate in Bagbdad), although in the previous year be merely styles him
'the lord of Egypt' The same author gives the title also to
Baybars's immediate predecessor, 24

In sum, the accession observances in the early Mamluk period bad two
purposes: first, to stress the continuity of the Mamluk with the Ayyubid
sultanate, and hence to manifest the sultan as the legitimate successor of the
Ayyubids; secondly, in more general and traditional terms, to present the
new sultan as the lawful ruler. The element of continuity with the Ayyubids
(and even with the Saljuqids) appears in the sultan's assumption of a malik-
title, and in his state procession with the saddle-cover. These usages
continued with the the momentum of established ritual long after the
Mamluk sultanate had become firmly established and its origin in violent
usurpation had lost significance. The administration of the bay'a by the
caliph, his act of recognition, and the mention of the sultan's name in the
were traditionally associated with the inauguration of a new Islamic
ruler.
A comparison of these observances with the rites which accompanied the
accession of a medieval Christian king shows one great difference: there is
no parallel in the Mamluk usages to the administration of unction, which
from Carolingian times onwards had bestowed on some European rulers a
sacral character. Sacral kingship did not, of course, begin with the anointing
of Pippin the Short; his long-haired Merovingian predecessors were
descended from a seagod, just as the kings of the West Saxons (and
subsequently of the English) sprang from Woden. Such sanctity by blood-
descent or by unction could not be acquired by rulers who were ex
hypothesi slaves by origin and Muslims by religion. Are we then to
conclude that tbere is no trace of sacral kingship in the Mamluk sultanate?
On the contrary, there seems to be an aura of sacredness around certain
individual sultans, although this apparently did not extend to their families,
nor was it inherent in their office. Some evidence of this can be shown in
regard to at least three sultans: Baybars, and
b.Qalawun.
To begin with Baybars. Ibn (d. 692/1292–3), whose panegyric of this
sultan, al-zahir fi sirat al-Malik is one of the principal primary
sources for the reign, at times suggests that Baybars was the object of a
special divine providence. He recounts an adventure of the time when
Baybars, not yet sultan, was in the service of al-Mughith 'Umar, the
Ayyubid lord of al-Karak, and, after defeat in battle, was travelling through
the desert with three companions. When Baybars was about to perish of
thirst, says the writer, God sent rain—and even a little straw for his horse.
After Baybars's usurpation of the throne, the divine assistance was
unremitting. In Ibn 'Abd words, 'Since God gave him the kingdom,
no one imagined evil against him but God acquainted him with it'.25
in spite of his short reign, so swiftly terminated by Baybars after 'Ayn
Jalut, seems to have impressed his contemporaries as possessing a degree of
sanctity. His tomb was a place of pilgrimage until Baybars ordered his
exhumation and its demolition.26 A Maghribi diviner, we are told,
prognosticated the reign of and his defeat of the Mongols nearly 10
years before his accession. In another anecdote, a khushdash of
describes how, as a youth, the future sultan was assured by the Prophet in a
dream that be would rule Egypt and defeat the Mongols. Ibn al-Dawadari,
who transmits these two stories, gives another version of the episode with
the diviner, derived from his own Mamluk father. In this version, the
diviner encounters three young Mamluks, Baybars, and Baktut al-
Atabaki. He prophesies that the first will rule Egypt and defeat the
Mongols, that the second will reign long and slay the first, and that the third
will hold a high amirate, all of which duly came to pass.27
When Ibn al-Dawadari speaks of the birth of b. Qalawun,
the reigning sultan when he wrote, his pages are filled with signs and
portents. He tells of a star in the east, a three-tailed comet seen at Mosul in
the year of birth, which signified that the child would live thrice
thirty years. A holy man saw in a dream the Imam 'Ali b. come from
the to restore for his third reign, after which the sultan would
conquer Baghdad. There is an eschatological flavour about this anecdote;
seems a Muslim counterpart of the Emperor of the Last Days in
medieval Christian millenarianism. Another anecdote (which Ibn al-
Dawadari gives twice in different contexts) tells of a mysterious voice
which spoke to Saladin as a youth, foretelling his victories and those of
Baybars and 28 Much of this may have been propaganda or flattery,

but it is significant that it took the form it did, suggesting a popular

readiness to see a sultan as personally God's chosen and not merely God's
delegate.
The primary function of an early Mamluk sultan was to wage war against
external enemies. Just as the accession observances had a twofold
derivation, from the traditions of the Ayyubids (and ultimately of the
Saljuqids), and from the older traditions of Islamic sovereignty respectively,
so the sultan as warrior can be viewed under a dual aspect. As the leader of
a host composed of his comrades (khushdashiyya) and his own Mamluk
household (a host which was perhaps in its structure and loyalties
essentially a synthetic tribe), he was a Heerkönig, to use an appropriate
German term. As a Muslim ruler and the caliph's delegate, he was a
mujahid, a fighter in the Holy War, a defender of the Community of Islam,
as Baybars is eulogized in the inaugural of the Caliph.
'This sultan al-Malik the most majestic lord, the wise, the just, the fighter in the Holy War, the
guardian of the frontier, the pillar of the world and the Faith, undertook to help the caliphate when
there were few to help, and scattered the infidel armies which had penetrated through the land'.29
This role of the sultan as Heerkönig and mujahid is very apparent in the
first 50 years of the Mamluk sultanate. After the murder of the last Egyptian
Ayyubid, Turan Shah, when the fight against St. Louis and his
Crusaders was still continuing (648/1250), Aybak was installed, first as
atabak al-'asakir (meaning here commander on behalf of the queen regnant,
Shajar al-Durr), then as sultan. Almost a decade later, when the Mongols
were advancing into Ayyubid Syria (657/1259), Aybak's son, 'Ali, was
deposed, and Aybak's Mamluk, became sultan. His murder and
replacement by Baybars (658/1260) merely substituted one warrior-chief
for another. The continuing dangers from the Mongols and the Crusader
states prevented the sons of Baybars from retaining the sultanate. Once
again a warrior, the atabak Qalawun, took the throne (678/1279), and, after
a successful fighting career, was succeeded by his son, al-Ashraf Khalil
(689/1290). Unusually, this sultan and son of a sultan was himself a warrior,
and brought about the final overthrow of the Crusader states. The threat
from the Mongols was passing away at the same period.
Although the early Mamluk sultans were pre-cminently leaders in war,
defending the state (or, as they would have seen it, the Muslim Community)
against external enemies, they were also the mainspring of government.
These governmental functions grew in importance with the ending of the
threat from the Crusaders and the Mongols, and with the evolution of the
sultanate itself into an unchallenged, stable, and pacific monarchy—a
development which was completed in the long third reign of b.Qalawun
from 709/1310 to 741/1341. In respect of these functions, the sultan was
always potentially, and sometimes actually, a despot with arbitrary
discretion, who sought to bring into submission the three categories of his
subjects: the Mamluks themselves (al-nas), the sedentary natives (al-
ra'aya), and the nomads (al-'Zurban). But although an arbitrary despot, the
sultan was not, either in legal theory or administrative practice, an absolute
monarch. As a Muslim, he was as much bound by the Holy Law of Islam as
any of his Muslim subjects, although the absence of means to compel his
submission to the shari'a deprived this concept of effective sanctions. The
caliph and the four chief judges, who headed the official administration of
the Holy Law, deferred to reason of state and the reality of power by
validating actions which they could not oppose and decisions which they
could not upset.
The notional omnicompetence and the divine authority of the Holy Law
resulted, however, in certain practical limitations on the sultan's functions.
He was left formally with no scope in legislation. A similar limitation was,
of course, imposed on the Ottoman sultans, who effectively circumvented it
by their qanuns, which as a systematic corpus of legislation have no
equivalent in Mamluk practice. The Mamluk sultan's judicial function was
of greater significance. The hearing of petitions and consequent redress of
grievances, although in form an administrative act, was so institutionalized,
and so closely associated with the prerogative of the ruler, as to result in the
creation of what was in effect a royal court of justice with a known and
regular procedure. But a comparison with medieval European development
is instructive. The personal and arbitrary power of decision retained by the
sultan prevented the formation. of a body of precedents and case-law. No
royal judiciary came into existence: the administered only the Holy Law
and had no competence in these cases. Admittedly, the sultan did not
personally hear and decide every case, but his prerogative powers were
delegated, not to a body of legal specialists, but to the great officers of the
state and household—the vicegerent in Egypt, the atabak, and the high
steward (ustadar).30
A continuing practical limitation on the power of the sultan was exercised
by the magnates, the great amirs, who might be his khushdashiyya,
originating from the same Mamluk household as he himself, and thus being
his peers in status. Between the sultan and this oligarchy of magnates there
was almost constant tension. Quiescent, even submissive, when occupied in
foreign wars or when dominated by a sultan of strong personality backed by
adequate resources, the amirs at other times were quick to display a factious
spirit, both against the sultan and among themselves. They derived the
means to disobey, to control, even to depose the sultan, in the first place
from their own Mamluk households, the members of which felt loyalty to
their ustadh, their immediate master, rather than to the sultan. In addition,
the tenure by great amirs of provincial governorships in Syria gave them
not only very considerable profits of office but territorial power-bases. Even
in the early days of the Mamluk sultanate, while the threat from the
Mongols and the Crusaders yet remained, both Baybars and
Qalawun were opposed by rebellious governors of Damascus, each of
whom was a khushdash of the sultan against whom he rebelled.
The strength of the magnates and the weakness of the sultan resulted
ultimately from the alienation of great sources of revenue in Under the
early Mamluk sultans the privy purse received only one-sixth of the landed
revenue in cash and kind, the rest being divided between the amin and the
old corps d'élite of the Ayyubids. It was not until 715/1315 that
succeeded in carrying out a cadastral survey (al-rawk ), in
resuming a larger share of the landed revenue for the privy purse, and in
carrying out other fiscal reforms which strengthened his position, In spite of
the splendour and luxury that surrounded him, the strict and formal
ceremonial of his public appearances, and the pompous ritual of his
accession, the sultan occupied a precarious position. The caliph's delegation
of authority counted for little in a crisis. Wh in 709/1310 the usurper al-
Baybars was confronted with a general revolt, and tried to reinforce his
authority with a new diploma from the caliph, his act provoked only the
jeering comment, 'Stupid fellow. For God's sake—who pays any heed to the
caliph now?'.31 At all times the Royal Mamluka were the sultan's main
safeguard, but there were himits to their reliability. Those Mamluks whom a
sultan had himself recruited, trained, and manumitted felt loyalty to him as
their ustadh rather than as their monarch, while no such bond existed
between the sultan and the the Royal Mamluks of his predecessors.
The might indeed form a faction hostile to the new sultan. It was out of
such a situation that the Mamluk sultanate itself arose, when in 648/1250
the resentful of of Ayyub conspired to kill his son and successor,
Turan Shah. Furthermore, the recruitment and training of a
Mamluk household was a slow business; hence one reason for the strength
of Qalawun, who during the long years of his amirate was able to build up
the household that sustained him as sultan. His son, probably profited
similarly from the long period between his first accession in 693/1293 and
the third and final inauguration of his sultanate in 790/1310. Hence on the
other hand the weakness of the later Qalawunids, who, brought from the
harem to the throne, bad virtually no Royal Mamluks of their own
recruitment. Moreover, it was not enough for a sultan to have a large
Mamluk household: it was also necessary for him to place his own amirs in
the key positions of government. This was a most delicate operation, which
taxed to the utmost the capacity and resolution of a sultan. It was an
operation carried out with notable success by at the start of his
third reign.
The Mamluk sultanate was a complex political and social organization. It
had inherent sources of weakness—inevitably clearer to later generations
than to contemporaries—but it was a remarkably durable structure with a
greater and more effective concentration of military and political power
than had existed, except briefly and occasionally, under the Ayyubids. It
was not a static but a changing and developing polity. The two and a half
centuries of the sultanate may be analysed into several periods which differ
in their character, and in which varying historical forces are at work as the
institutions and offices of the state and royal household evolve. We havc to
do, not with 'a Colluvies of slaves', but with a wealthy, powerful, and
sophisticated medieval state.
 CASSIODORUS AND RASHID AL-DIN ON BARBARIAN
RULE IN ITALY AND PERSIA1

By D.O.MORGAN

Italy in the late fifth century A.D. and Persia in the mid-thirteenth were
lands of ancient and deeply-rooted culture. One had been at the centre of
the western Roman Empire; the other had formed a significant part of the
society of medieval Islam, and its traditions stretched back further still.
Both lands found themselves invaded, conquered, and ruled by wandering
'barbarians', Ostrogoths and Mongols respectively, from beyond the then
recognized frontiers of 'civilization'.
Cassiodorus Senator and Rashid al-Din Allah were representatives of
two long-established administrative traditions. Their governmental skills
were not despised by the conquerors: both served under them for many
years in high office. Both, too, left behind them copious and varied
writings, whose range and character is in some ways curiously similar. This
article is concerned with some of these writings. It seeks to study certain
aspects of barbarian rule in the light of what Cassiodorus and Rashid al-Din
have to say, and hopes thereby to bring out both the similarities and the
contrasts between the two régimes.
Cassiodorus's Variae2 are a collection of official correspondence, as are the
Mukatabat-i Rashidi3 the letters of Rashid al-Din. Neither collection can,
perhaps, lay claim to the status of an archive, for both are selections—
Cassiodorus's made by himself, Rashid al-Din's by his secretary,
Abarquhi. But both afford a fascinating insight into the workings of the two
central administrations, the nature of the two régimes, and perhaps, too, into
the minds and attitudes of their authors.
Nor is this all. Cassiodorus wrote a history of the Goths, in twelve books.
Unhappily this is lost, but the substance of it appears to be, to some extent,
preserved in the much shorter work of Jordanes.4 And the posthumous fame
of Rashid al-Din rests chiefly on his 'vast historical encyclopaedia', as
Barthold called it,5 the Jami' al-tavarikh, which includes accounts of the
history of large parts of the known world, including Europe, but is
principally of value as the most important single source for the history of
the Mongol empire. Most of the work seems to have survived, and all the
more significant sections of it have been published.
So, if Cassiodorus may be described as a 'scholar-bureaucrat',6 the
description fits Rashid al-Din no less aptly. Because of their positions in
government, and their literary training and inclinations, they are perhaps
uniquely wellqualified to help their reader to understand their respective
worlds. They show him something of the processes by which barbarian
peoples, lacking experience of government in societies based on cities and
settled agriculture, were able to
achieve their ends through the co-operation of those officials whom, in one
sense or another, they had inherited from the régimes they had supplanted.
Cassiodorus's family was not one of ancient lineage—though he liked to
suggest otherwise.7 But his immediate ancestors for several generations had
served the state in various capacities. His great-grandfather had helped
defend the south of Italy against Vandal raids. His grandfather had served
on an embassy sent by the emperor Valentinian III to Attila the Hun. And
his father had achieved high office, eventually rising to the supreme judicial
and administrative position, that of Praetorian Prefect. Cassiodorus Senator
himself attracted the attention of the Ostrogothic ruler Theodoric, and
became Quaestor 'the mouthpiece of the Emperor',8 and thus an illustris,
one of the highest-ranking ministers, while still under 30. Later he
progressed to the Mastership of the Offices (head of the civil service) and
ultimately, though not until after the death of Theodoric, he became
Praetorian Prefect. Cassiodorus retired from public life during the
Byzantine reconquest of Italy from the Ostrogoths. He lived on for perhaps
another 40 years, dying, we are told, at the age of 95. The scholarly work
done and directed by him at his monastery of Vivarium during those years
was of the first importance for the subsequent history of European culture,
but is not, for the most part, relevant to the theme of this article.
Cassiodorus's life, then, spanned the period between the ending of the
'official' western Roman Empire by barbarian rulers and the return of Italy
to Imperial government under Justinian. In 476, Odoacer, a Rugian chief
holding command in the Roman army, deposed and pensioned off the last
western emperor, Romulus Augustulus, on grounds of superfluity.
Historians have differed over the significance, if any, of this event.
Cassiodorus himself was in no doubt of its at least symbolic importance.
'Thus the Western Empire of the Roman race, which Octavianus Augustus, the first of the Augusti,
began to govern in the seven hundred and ninth year from the founding of the city, perished with this
Augustulus in the five hundred and twenty-second year from the beginning of the rule of his
predecessors and those before them, and from this time onward kings of the Goths held Rome and
Italy.'9
Odoacer's kingdom, though apparently not unsuccessful, was not long to
survive. In 488, Theodoric the Ostrogoth obtained permission from the
eastern Emperor Zeno to remove himself and his people from the Balkans,
to invade Italy and take it from Odoacer. Cassiodorus represents Zeno as
being reluctant to deprive himself of Theodoric's company,10 though in
reality the departure of the Ostrogoths must have been a considerable relief
to him. Of the terms on which Zeno authorized the invasion, 'we have only
the vaguest accounts in our authorities'.11 Possibly Zeno was happy to leave
the terms vague as the price of ridding the eastern Empire of the
Ostrogoths.
After a prolonged struggle, Odoacer was captured, or rather surrendered,
and was treacherously put to death in 493. Cassiodorus records this last
event without comment.
'He (Odoacer) sent an embassy and begged for mercy. Theodoric first granted it and afterwards
deprived him of his life.'12
Theodoric ruled until his death in 526—undeniably a successful and
prosperous reign. But the stability of the Ostrogothic kingdom died with
him. In 535 the Byzantine army under Belisarius landed. The Gothic Wars
began, and the days of Ostrogothic Italy were numbered.
The first Mongol invasion of Persia took place in 1219–23—almost the last
phase of Chingiz Khan's astonishing—and still not satisfactorily explained
—career of conquest.13 The impression that invasion made on
contemporaries is unmistakable. The chronicler Ibn al-Athir is typical:
'…a tremendous disaster such as had never happened before, and which struck all the world, though
the Muslims in particular. If anyone were to say that at no time since the creation of man by the great
God had the world experienced anything like it, he would only be telling the truth…. It may well be
that the world from now until its end…will not experience the like of it again, apart perhaps from
Gog and Magog'.14
But no serious attempt seems to have been made to occupy and absorb the
whole of Persia into the Mongol empire. Nevertheless, the northern and
eastern parts of the country were conquered, and the province of Khurasan,
in particular, was utterly devastated. The great cities of Khurasan—Marv,
Balkh, Harat, Nishapur—were destroyed, and their populations put to the
sword, almost to the last man.
Thirty years later, as part of a grandiose plan for the further extension of the
Mongol empire, Persia was invaded again. The command of the expedition
was entrusted to Hülegü, brother of Möngke, Chingiz Khan's third
successor as Great Khan. Between 1256 and 1260, Hülegü occupied most
of Persia, as well as invading 'Iraq and bringing the 'Abbasid caliphate in
Baghdad to a sanguinary end. His progress was finally stopped by
dissensions within the Mongol royal house, as well as by successful
military resistance on the part of the Mamluk rulers of Egypt, whose lands
remained outside the boundaries of the Mongol empire. But in Persia and
'Iraq, the kingdom of Hülegü and his descendants, known to historians as
the Ilkhanate, was established for the next 80 years.
Rashid al-Din was born between the two invasions, about 1247, in
Hamadan, in western Persia.15 Though himself a Muslim, he was of Jewish
descent, and may have been converted to Islam as late as the age of 30.16
He was by training a physician, and it was in this capacity that he first
entered the service of the Mongol ruling house of Persia, during the reign of
the Ilkhan Abaqa (1265–81). Nothing else is known with certainty of his
career until in 1298, in the reign of Ghazan Khan, he became deputy Vazir.
He remained as one of the two chief
ministers of the Ilkhanate until the death of Ghazan's brother and successor,
Öljeitü, in 1316.
During most of this period, his was the principal influence in government,
and he was almost certainly the architect of the administrative reform
programme which is associated with the name of Ghazan Khan. After the
death of Öljeitü, Rashid al-Din's colleague, Taj al-Din 'Ali Shah, managed
to engineer his dismissal and ultimate execution on a charge of poisoning
the late monarch.17
Rashid al-Din himself, then, was not a member of an old-established
administrative family. His ancestry is obscure, compared with. that of
Cassiodorus. Nevertheless, he may be said to have inherited the traditions
of pre-Mongol government in Persia, as they had been carried on, with little
of a discernible break, by such. families as the Juvaynis. And he founded
something of a dynasty: a number of his sons acted as provincial governors
during the period of their father's Vaziate, and one of them, Ghiyath al-Din
himself became Vazir some years after the death of Rashid al-
Din.
Historians have always been glad to make use of Cassiodorus's writings as
historical evidence. Their opinions of the man himself, however, have
varied. Cassiodorus's Latin style is so involved and flowery that he has
frequently strained the patience of his readers. Sometimes a scarcely veiled
note of exasperation may be detected in discussions of him and of his work.
Consequently, there has been a tendency for his personal character to be
tarred with the same brush as his prose style. This is perhaps not entirely
just—one may note that Cassiodorus's Institutiones, for example, are
written in much more simple and straightforward language than the Variae.
The style of the Variae was not an inseparable adjunct to their author's
character. But Thomas Hodgkin, who undertook the daunting task of
rendering the Variae into English, remarked of Cassiodorus that 'he was
never so happy as when he was wrapping up some commonplace thought in
a garment of sonorous but turgid rhetoric'.18 However, Hodgkin, who could
venture the remarkable description of Attila as 'the squalid and
unprogressive Turanian',19 was no doubt as much as anyone else the
creature of his time. Arnaldo Momigliano has more recently suggested that,
to the generations who have passed through the wars of the twentieth
century, the achievement of Cassiodorus has seemed more worthy of
respect. 'We, the members of the race of iron, have learnt to appreciate the
lesser men—the men who tried to save what could be saved and who did
not disdain the task of elementary teaching when elementary teaching was
needed'.20
Cassiodorus himself explained his reasons for making and issuing his
collection of Variae. His friends, he said, 'wanted me to do this that future
generations might recognise the painful labours I have undergone for the
public good and the hardships of my unbribed conscience'.21 It is hard to
believe, however, that there was anything painful for Cassiodorus in the
composition of the numerous lengthy digressions with which the Variae
abound. He cannot resist the temptation to show off his esoteric knowledge,
particularly
on questions of natural history. His account of the elephant is a well-known
and characteristic example, as is his description of what is known about the
origin of amber, where he concludes
'We have thought it better to point this out to you, lest you should imagine that your supposed secrets
have escaped our knowledge'.22
On occasion, natural history and government business become inextricably
intertwined. Writing to the Praetorian Prefect Faustus, Cassiodorus
demands
'Why are your ships not spreading their sails to the breeze? When the south wind is blowing and your
oarsmen are urging on your vessels, has the sucking-fish fastened its bite upon them through the
liquid waves? Or have the shell-fishes of the Indian Sea with similar power stayed your keels with
their lips: those creatures whose quiet touch is said to hold back, more than the tumultuous elements
can possibly urge forward? …But no. The sucking-fish of these men is their hindering corruption.
The shellfishes that bite them are their avaricious hearts. The torpedo that benumbs them is lying
guile. With perverted ingenuity they manufacture delays, that they may seem to have met with a run
of ill-luck'.23
It is a matter of taste whether this kind of rhetoric is found interminable or
endearing. But even those who, like the present writer, find Cassiodorus's
approach to official correspondence rather attractive, would perhaps have
thought differently had they had to tackle the Variae without the assistance
of Hodgkin's abridged translation.
As for the 'History of the Goths', we again have Cassiodorus's own account
of his motives in writing it. This occurs in the Variae, as part of a eulogy
which Cassiodorus composed on himself (in the name of king Athalaric), to
be sent to the Senate on the occasion of Cassiodorus's promotion to the
Praetorian Prefecture.
'Not satisfied with extolling living kings, from whom he might hope for a reward, he drew forth the
Kings of the Goths from the dust of ages, showing that the Amal family had been royal for seventeen
generations, and proved that the origin of the Gothic people belonged to Roman history, adorning the
whole subject with the flowers of his learning gathered from wide fields of literature.'24
The 'History of the Goths', then, 'presented the tribe in general, and the
family of Theodoric in particular, as co-operative participants in the history
of the Mediterranean, from the time of Alexander the Great onwards'25—or
even earlier: Cassiodorus will have us believe, by means of the not
uncommon identification of 'Getae' with 'Goths', that there were Goths at
the siege of Troy.26
It should be added that there are cogent reasons for supposing that the
'History of the Goths' had, in the minds both of Cassiodorus and of his
summarizer Jordanes, a more immediate political purpose. Professor
Momigliano writes that Jordanes's 'work had a clear political message. It
invited the Goths to cease resistance' to the armies of Justinian, 'but also
gave encouragement to those who worked in Constantinople for a modus
vivendi between Goths and Romans'.27 And Momigliano has demonstrated
the probability that
here, as elsewhere, Jordanes was merely summarizing the arguments of
Cassiodorus himself (on the assumption that Cassiodorus continued to add
to and revise his 'History' for many years after he had described his work in
Athalaric's letter to the Senate).
Rashid al-Din's great Jami' al-tavarikh 'Collection of histories', as it now
survives, falls into two parts. The first part, historically the more important,
relates the history of the Mongols and their ancestors in central and eastern
Asia, the career of Chingiz Khan, and the Mongol conquests. It concludes
with a long and extremely valuable account of the history of Mongol Persia
up to the death of Ghazan Khan in 1304. All this part of Rashid al-Din's
work was commissioned from him by Ghazan Khan himself, and was
consequently named the Tarikh-i Ghazani. Öljeitü, on his accession, asked
Rashid al-Din to write a second part, including histories of all the peoples
with whom the Mongols had come into contact. It included histories of pre-
Islamic Persia, and the Caliphs, the Oghuz and the Turks, China,
India, the Jews, and the Franks. This last section, it has been suggested,'
seems to have been the only attempt made by a medieval Muslim historian
at an outline of the history of the Christian West'.28 That this part is
historiographically of great interest is undeniable, but in terms of historical
information, it is the Tarikh-i Ghazani which is irreplaceable.
This part of the book is the principal authority on the Persia of the author's
own time. And because of his position as chief minister and official
historian, he had access to Mongol records, many of them now lost, which
make his book the nearest thing we have to a primary source on large tracts
of Mongol history outside Persia itself.
Rashid al-Din was fully aware of the unprecedented nature of his great work. 'Until now (he
maintained), there has never been at any time a history which contains narratives and general
accounts of the peoples of the various parts of the world, and of the different races of mankind.'29
Nor was he in any doubt of the worthiness of his subject in deserving the
attention of the historian.
'What event or occurrence (he asked) has been more notable than the beginning of the government of
Chingiz Khan, that it should be considered a new era?'30
He explained the method by which he gathered his information. This
consisted in ransacking written sources as far as possible, having Mongol
chronicles interpreted to him, and in oral interrogation. This might include
both distinguished Mongols—such as Ghazan himself, who was, or so
Rashid al-Din tactfully assures us, a great authority on Mongol tradition—
and whatever foreigners could be intercepted as they passed through
Persia.31
The style of the Jami' al-tavarikh could hardly be further removed from that
of the Variae. Though it is not without its difficulties, the narrative sections
at least are written in clear and simple Persian, with no attempt at elaborate
frills. This was not customary at the time. Much more typical, if an extreme
example, is the Tarikh-i of which E.G.Browne justly
remarked that 'we could forgive the author more readily if his work were
less valuable as an original authority on the period of which it treats, but in
fact it is as important as it is unreadable'.32 Subsequent Persian historians
unfortunately elected to model their work on rather than on Rashid al-
Din.
In view of the immense range of Rashid al-Din's activities as Vazir—
activities fully revealed in his letters, the Mukatabat-i Rashidi—doubt has
been cast on the possibility of the whole of the Jami' al-tavarikh being
completed single-handed by its author. He himself tells us that he did his
writing in the time between morning prayer and sunrise. The probability is
that, like many another historian busy with other concerns, he employed a
large staff of 'research assistants', who did much of the quarrying for
material, and perhaps some of the writing too. This may underlie the claim
of the contemporary historian Abu 'l Qasim Qashani to have been the true,
unacknowledged and unrewarded author of Rashid al-Din's history.33
The Jami' al-tavarikh has the usual strengths and weaknesses of 'official'
history. It benefits through its author's access to sources which would
otherwise have been unavailable, but the reader must look elsewhere for
direct criticism of the Mongol régime, at least so far as the reign of Ghazan
is concerned. On the other hand, there are long and graphic descriptions of
the abuses which, we are told, were put right by Ghazan.
In the Mukatabat-i Rashidi, however, Rashid al-Din several times gives
vent to what we may suppose to have been his true feelings about the
Mongols. For unlike the Variae, the Mukatabat are not an exclusively
official collection. They certainly include official communications with
governors, urban authorities, and so forth, but there are also more personal
letters to, for example, the author's sons—and only partly in their capacity
as provincial governors—and to a considerable number of Muslim men of
religion. The collection includes unexpected items, such as a letter from
Rashid al-Din, allegedly written in India, to which, he says, he had travelled
at the order of the Ilkhan. Nothing is known from any other source of such a
journey on the part of Rashid al-Din; and it is such puzzles as this that have
caused some to doubt the authenticity of the collection.34 But if it is a
forgery, it is an extraordinarily skilful and unbelievably well-informed one,
the purpose of which is by no means readily apparent. The case for
accepting the letters' substantial authenticity seems very strong.
The character of the author which emerges from a study of the Mukatabat is
by no means an unattractive one. It is true that, if he put his mind to it, he
could be as long-winded as, and less diverting than, the author of the
Variae. It may not be entirely fanciful to imagine his sons groaning in
despair when receiving yet another 15-page letter from their distinguished
father, packed full of very orthodox, very detailed, and very conventional
pieces of advice. But there is another side to the coin. Rashid al-Din's
concern with the welfare of the state's subjects, particularly the peasantry,
comes out very strongly—an impressive trait which is not negated by the
consideration that he evidently regarded just treatment of the peasants as
more profitable to the Treasury, in the long run, than ruthless exploitation.
His delight in pious benefactions, his extensive building and land
reclamation programmes, his concern with the promotion of education and
with the maintenance of hospitals, are all amply represented. This is all in
addition to the more obvious Vazir's business—the
organization of armies, the determining of rates of taxation, and so forth.
Even taking into account the possibility of some degree of judicious editing
on the part of Abarquhi, the picture remains one of a many-sided,
generous, cultivated, and benevolent figure—though also ambitious, proud,
and perhaps under no illusions regarding his own great abilities. It is his
long and, some might say, boastful accounts of his great building projects
which. single him out most clearly as a man who intended to make his mark
on the memory of posterity. He was, indeed, remembered; but he deserved
to be.
The most cursory examination of the processes of government illustrated in
the works of Cassiodorus and Rashid al-Din shows clearly enough that the
two barbarian regimes faced, in many respects, similar problems. Some of
these will be discussed. But there were, naturally enough, contrasts, and
these should be borne in mind when considering the similarities.
There was one quite basic difference. Theodoric is reported to have
remarked wryly that 'a poor Roman imitates the Goth, a rich Goth the
Roman'.35 The Ostrogoths—this at least is the impression given by Roman
writers—wanted nothing better than to be integrated thoroughly into the
superior Roman civilization. Thus, 'day after day Cassiodorus tried to give
Roman dignitas to the orders of his Barbarian masters'. He made 'a
sustained effort to present a Barbarian as the embodiment of civilised
justice and wisdom'.36 There is no evidence that the Mongols suffered from
an inferiority complex. They considered themselves, their achievements,
their ancestors, and their way of life to be immeasurably superior to those
of the peoples they had conquered. Even if—especially in China, under the
rule of Qubilai and his successors—there are, eventually, signs of an
appreciation of civilized life which goes beyond the mere material benefits
to be gained from exploitation, the Mongols remained Mongols, and proud
of it. If Ghazan Khan and Öljeitü commissioned Rashid al-Din to write their
history for them, it was not in order that he should fabricate some
connexion between the Mongols and the Persian past, with all its glories,
which would make the Mongols respectable. They regarded their own
history as a fit subject on its own merits. Indeed, if anything, the Mongol
rulers of Persia are likely to have hoped that the writing of the Jami' al-
tavarikh might help to sustain the existence of a distinctive Mongol identity.
This would inevitably come under some pressure when, at the time of
Ghazan's accession, the Mongols of the Ilkhanate were converted to Islam,
the religion of the bulk of their Persian subjects. Among the results of this
'Mongol-consciousness' is that it is far easier to find reference to
specifically Mongol institutions in Rashid al-Din than to anything very
clearly Gothic in Cassiodorus.
This follows, in fact, from an essential difference in the nature of
government in the two countries. The rule of the Mongols in Persia was the
result of straightforward, bloody, alien conquest. Whatever institutions
might survive from the Persian past, whatever native Persians might be
prepared to serve the invaders, nothing could disguise the fact that the
Mongols held Persia by force and by right of conquest. By contrast,
barbarian rule in Italy had an element of consensus built into it. Theodoric
deputized, in theory, for a specifically Roman emperor in Constantinople.
The Ilkhans might recognize the supremacy of the Great Khan in Peking—
but he was a fellow-Mongol: indeed, a close relative. Part of the price the
barbarian rulers of Italy paid for being accepted
as legitimate sovereigns was their association of the Roman senatorial
aristocracy with them in government: 'it was from these men that Theodoric
received his greatest support and in return they gained security and respect
shown by too few emperors. Not since the Principate had the senate been
taken into such close partnership in government'.37 Rashid al-Din served the
Mongols: Cassiodorus and his peers were partners—if junior partners—of
the Ostrogoths. The great men of the senate belonged to families with, in
some cases, centuries of aristocratic tradition behind them. There was an
administrative tradition, certainly, in Persia, but there had been no
hereditary aristocracy at any time since the fall of the Sassanian empire, six
centuries before.
The element of consensus was not an innovation of Theodoric's reign: he
followed the example set by Odoacer. There is interesting evidence for the
latter's reign arising from recent studies of the Colosseum in Rome.
Evidently a focus of loyalty for the Roman aristocracy, it was twice restored
by Odoacer. A number of seat-inscriptions survive, bearing the names of
senators. These illustrate, it has been suggested, the continued privileged
position of the senatorial order, in that they 'continuaient d'avoir droit a des
égards apparaissant au grand jour avec l'usage des places d'honneur qui leur
étaient réservées dans les lieux de spectacle'.38 More generally, it has been
concluded, on the basis of the Colosseum evidence, that Odoacer 'a cherché
à se concilier l'aristocratie et à s'appuyer sur elle. La noblesse romaine joue
sous son règne un role plus important qu'auparavant dans le gouvernement
et 1'administration, et renforce son emprise sociale'.39
But, if these contrasts are remembered, the similarities are none the less
striking. Among these, a major problem inevitably confronting both
barbarian governments was that of the reimposition of order. This involved
the basic question of landownership as much. as anything. The two
administrations seem to have arrived at a very similar rule of thumb.
Theodoric's edict ran:
'If any Barbarian usurper have taken possession of a Roman farm since the time when we, through
God's grace, crossed the streams of the Isonzo, when first the Empire of Italy received us, and if he
have no documents of title to show that he is the rightful holder, then let him without delay restore
the property to its former owner. But if he shall be found to have entered upon the property before the
aforesaid time, since the principle of the thirty years' prescription comes in, we order that the petition
of the plaintiff shall be dropped'.40
The ancestry of this law presumably derives from the emperor Theodosius
II's enactment of 424.41
Ghazan Khan, 40 years after the establishment of the Ilkhanate, found that a
morass of conflicting landownership claims had come into being. He, too,
decided that a 30-year limit must be imposed. The authority of Mongol
tradition was invoked to support this.
'Before this time, past sultans, and Chingiz Khan, in all their farmans and
yarlighs (edicts), made mention that thirty-year old claims should not be listened to.'42
Ghazan Khan's aim was to have this, as was believed, Mongol or Turkish
rule43 given an official stamp of approval by an appropriate Islamic
authority. He therefore persuaded the Fakhr al-Din of Harat to issue a
confirmation of the regulation. This, as transcribed by Rashid al-Din, reads
in part:
'I will not transgress or turn away that which is required by the shari'a of and I will, to
the best of my ability and strength, strive to the utmost limit and extremity in the writing up and
investigation of claims and in examining and verifying title deeds and written shar'i bonds, and I will
not hear any claim made after a period of thirty years, in accordance with the conditions which have
been mentioned, and I will not pay any attention to it, or grant it validity'.44
who would not subscribe to this were threatened with dismissal. But
whether this regulation in fact derives from the custom of the steppe is not
wholly clear. Thirty years is among the limitations on claims to
landownership prescribed in Islamic law.45
Abuses had grown up, too, in the workings of the administrative machinery
of both régimes—notably in the official postal and intelligence systems.
There are several letters on this matter in the Variae, of which the most
interesting is book IV, 47.46 Writing to a Gothic officer, Gudisal the Saio,
Cassiodorus has two main complaints. The first is that large numbers of
people with no right to use the official post, the Cursus Publicus, are
nevertheless doing so. Even those sent on official business are making their
missions an excuse for 'pleasure-tours at the public expense'. The second
complaint is that the post-horses are being more heavily laden than is good
for them. This launches Cassiodorus into one of his characteristic
illustrations from natural history.
'Cranes, when they are going to cross the sea, clasp little pebbles with their claws, in order to steady
without over-weighting themselves. Why cannot those who are sent on public errands follow so good
an example?'
A scale of fines is prescribed for those who transgress these and other
regulations. Details are given of who may use the Cursus Publicus, and
what are the permitted maximum weights for the horses. The fines are to go
to those who operate the postal service, and thus, Cassiodorus portentously
concludes,' the evil will, as we so often see in human affairs, furnish its own
remedy'.
The Ostrogothic government had taken over the Roman Cursus, whatever
its defects, as a going concern. The equivalent official postal system of
medieval Islam, the Barid, had, by contrast, long disappeared by the
thirteenth century A.D. The Mongols brought their own system, the Yam, to
Persia. It owed much to Chinese precedent, as the word Yam itself, derived
from Chinese, shows.47 This serves to emphasize the importance of the
Mongols' Chinese background. It is very clear from the basic, if
problematical, source on the early history of their empire, the 'Secret history
of the Mongols', that in the Mongol mind, Mongolia and China were the
countries that really mattered:
the rest of the world, even the rest of the Mongols' conquests, are treated
vaguely and briefly—evidently of only peripheral interest.
The Mongol horse-post system achieved an unparalleled elaboration
throughout the length and breadth of Asia, as Marco Polo's account of it
vividly shows.48 So far even as China was concerned, the Mongols' system
transcended all precedents.49 It was instituted by Ögedei, the second Great
Khan (reigned 1229–41).50 As early as the reign of Möngke (1251–9), there
are complaints of abuses in the operation of the system very like those of
which Cassiodorus complains. Merchants are illicitly making use of the
facilities, and those legitimately travelling on official business are dallying
on the way. Möngke attempted to put this right.
'As for the more important ilchis (messengers), they should not make use of more than fourteen
ulaghs (post-horses); they should proceed from yam to yam (post-station to post-station) and should
enter no village or town in which they have no specific business; and they should take no more
provisions than each man can eat.'51
Presumably Möngke's reforms had some effect, at least for a time. But in
Persia the situation had deteriorated again by the time of Ghazan's reign.
The depredations of travelling ilchis had once more become a public
menace.
'Even if five thousand mounts had been stationed at each yam, they would not have been enough for
them. Inevitably they took as mounts all the herds of the Mongols that were in the summer and
winter pastures. And they made dismount all (persons travelling in) caravans…from the directions of
Khitay and Hindustan and other quarters far and near…and took their horses, and would leave them
on the road, and some in fearful places with their baggage…. Thieves and robbers masqueraded as
ilchis, and blocked the road, saying "we are ilchis", and having taken their horses as post-horses, they
would suddenly seize them, tie them up, and plunder their baggage. Further, it often happened that
one group of ilchis would take the post-horses of another group Whoever had more weapons and
greater power took the post-horses of the other Whatever they found in the villages they tyrannically
seized.'52
All this, needless to say, did not long survive the reforming zeal of Ghazan
Khan. Doubtless abuses crept in again before long, but the system was still
flourishing under Timur in the early fifteenth century, according to the
account of the Castilian ambassador Clavijo.53
No disorders on the Persian scale are even hinted at in the Variae.
Occasionally, however, there is evidence that things did sometimes go
wrong, whatever the benefits of Theodoric's rule. There are detailed lists of
financial and other abuses which need correction in Suavia,54 and Spain,
then under Theodoric's supervision.55 Merchants in Apulia are granted tax-
exemptions for two years, because of damage they have suffered from
hostile incursions56—
perhaps hardly the fault, directly, of the Ostrogothic government. Nor was
damage caused by the eruption of Vesuvius, also the occasion of a grant of
tax-relief,57 to be laid at Theodoric's door. But one such grant is rather a
different matter. Writing to the Praetorian Prefect Faustus, Cassiodorus (in
Theodoric's name) gives these instructions.
'A wise ruler will always lessen the weight of taxation when his subjects are weighed down by
temporary poverty. Therefore let your Magnificence remit to the Provincials of the Cottian Alps the
as publicum [land tax] for this year, in consideration of their losses by the passage of our army. True,
that army went forth with shouts of concord to liberate Gaul. But so a river bursting forth may
irrigate and fertilise a whole country, and yet destroy the increase of that particular channel in which
its waters run. We have earned new subjects by that campaign: we do not wish them to suffer loss by
it. Our own heart whispers to us the request which the subjects dare not utter to their Prince.'58
This incident has many echoes in the writings of Rashid al-Din. In a letter
to his son, Amir governor of Bam, he reprimands him for
oppressing the people placed in his charge, and orders that, consequently:
'Nothing shall be demanded for a period of three years, by way of qalan and qubchur,59extraordinary
levies and impositions (taklifat), either for the Divan of Kirman or of the Great Urdu (the
Court), so that their ruined places and waste lands may again come into a state of populousness and
cultivation'.60
Then there is the astonishing story, in the Tarikh-i Ghazani, of the landlord
in the region of Yazd who, in the year 1292, travelled to a large village,
Firuzabad, which he owned. He hoped to extract from his tenants some part
of their harvest.
'However hard he tried, for three whole days he could not lay his hands on any of the headmen. But
seventeen tax-collectors holders of drafts (barat) and assignments were sitting
in the middle of the village. They had strung up with ropes a field-guard (dashtban) and two peasants
whom they had seized from the fields and brought to the village. They were beating them, to make
them reveal the whereabouts of the others, and to produce some food for them. But it was all a total
waste of time. All these tax-collectors and their followers had to have provisions and fodder, wine
and beautiful girls. From this it may be deduced what other kinds of tyranny there were.'61
The impression which Rashid al-Din is at pains to give his reader is one of
tyranny, extortion, and incompetent government in the Ilkhanate—until the
accession of Ghazan Khan, after which auspicious event all wrongs were
righted. Arghun Khan (reigned 1284–91), he relates, was impressed by a
report from two officials on the possible collection of tax arrears. He
therefore
'…sent them to Baghdad, to collect the arrears and to demand the wealth due to the Treasury. They
went there, and by using the bastinado and torture, they collected abundant wealth'.62
Writing generally of the period before Ghazan's accession, and discussing
tax maladministration, Rashid al-Din assures us that
'In truth, from those provinces not the smallest gold coin ever reached the Treasury, and not one-fifth.
was paid of the drafts allocated to individuals on the basic revenue. No one ever saw the tamghachi
(in charge of collecting the commercial tax, the tamgha). He had usually either fled, or was a prisoner
in the hands of the who would beat him'.63
The result of the depredations of ilchis, the high level and totally arbitrary
nature of taxation—some taxes being levied, according to Rashid al-Din,
not twice yearly, but 20 or 30 times—the decline in public order, and the
general incompetence of the administration, was virtual government
bankruptcy. This led to such disastrous expedients as the attempt, under the
Ikha Geikhatu (reigned 1291–5), to introduce into Persia paper money,
Chao, on the Chinese model.64 It also inevitably produced a sharp decline
in agriculture, the basis of the economy, and 'a general flight from the land
on the part of the peasants'.65 Rashid al-Din maintains that nine-tenths of
the cultivable land was lying waste.66
What was Ghazan's own attitude? Rashid al-Din seems in two minds over
this. Sometimes the first of the line of Muslim Ilkhans is depicted as an
ideal Muslim ruler, who 'took the greatest personal interest in the welfare of
the state',67 and who had the interests of his Persian subjects very much at
heart. More credible, however, is the speech put into Ghazan's mouth, as he
addresses the Mongol amirs.
'I am not protecting the Persian peasantry. If it is expedient, then let me pillage them all—there is no
one with more power to do so than I. Let us rob them together. But if you expect to collect provisions
and food in the future, and demand this, I will be harsh with you. And you must consider what, if you
commit extortion against the peasants, take their oxen and seed, and cause their crops to be
consumed, you will do in the future. You must think, too, when you beat and torture their wives and
children, that just as our wives and children are dear to our hearts, so are theirs to them; they are
human beings, just as we are.'68
On this evidence, then, Ghazan's attitude was one of hard-headed and
farsighted realism, with a touch of human compassion thrown in.
Nearly half of the section of the Jami' al-tavarikh which deals with
Ghazan's reign is concerned with his great programme of administrative
reforms. Most important of all, it includes copies of the reforming edicts,
the yarlighs, issued by Ghazan. Both Ghazan and his minister, Rashid al-
Din, were strong personalities who evidently knew what they wanted to
achieve, and, apparently, how to set about it. But whatever the results of the
reforms, they did not last. Ghazan himself reigned for only nine years—
hardly long enough to repair the ravages of seven decades. Writing in 1340,
after the death of Abu Sa'id, the last significant Ilkhan, the historian and
geographer Allah Mustawfi remarked, of the revenue of Persia:
'It reached the sum of 21,000,000 odd currency dinars (in the reign of
Ghazan). At the present time it certainly does not amount to half that sum, for most of the provinces
have been thrown into disorder by the usurpation of authority and the coming and going of armies.
The people have withheld their hands from cultivation'.69
The Persian peasants, then, in the opinion of many of the Mongols, were
there simply in order to be exploited as vigorously as possible, for the
benefit of their rulers. If Persian administrators such as Rashid al-Din had a
part to play, it was in using their skills to produce more revenue. When their
usefulness was ended, they would be discarded without a second's thought.
Even the more sympathetic approach attributed by Rashid al-Din to Ghazan
Khan seems to be grounded very largely in a more than usually rational
approach to the same problem of the efficient exploitation of the subject. It
is worth looking at the Mukatabat-i Rashidi for what may be Rashid al-
Din's own personal views about the impact of Mongol rule on his people. In
a letter to Mawlana al-Din concerned with a revised scale of
taxation which is to be applied, Rashid al-Din reveals his feelings on the
nature of Mongol rule explicitly enough. He speaks of 'the time of the
tyrannical Turks' (in this context not to be distinguished from Mongols) 'and
the oppressive bitikchis' (Mongol scribal officials).70 In a letter, also in the
collection, from Mu'in al-Din Parvana of Rum, Rashid al-Din's
correspondent describes Turkish amirs as 'mere deceivers and accomplices
of the Devil'.71 Other similar remarks are to be found scattered about the
Mukatabat. For Rashid al-Din, and other Persians who felt as he did, it was
a matter of making the best of an unavoidably bad job, and mitigating the
evils of Mongol rule as far as possible, as well as feathering their own nests
in the process.
There is little to parallel this in the writings of Cassiodorus. In a letter to
Faustus, Theodoric is made to speak thus about Ostrogothic frontier guards:
'Think what a life of hardship the soldier leads in those frontier forts for the general peace, thus, as at
the gate of the Province, shutting out the entry of the barbarous nations. He must be ever on the alert
who seeks to keep out the Barbarians. For fear alone checks these men, whom honour will not keep
back'.72
There is nothing in this to indicate that Theodoric was himself the most
successful of the barbarian invaders. Cassiodorus totally identifies the
Ostrogothic government with Roman civilitas. Similarly, Theodoric's
newlyacquired Gaulish subjects are urged, in 510, to 'put off the barbarian;
clothe yourselves in the morals of the toga…. Do not dislike the reign of
Law because it is new to you, after the aimless seethings of Barbarism'.73
So much for whatever notions of law the barbarian invaders may have
brought with them to the Roman Empire. By contrast, the Mongol laws, the
Yasa, ascribed to Chingiz Khan, came with the Mongols to Persia. Though
there is much reason for doubting, in the light of the available evidence, that
the Yasa was ever reduced to a coherent, written code of laws, Mongol laws
and customs long remained important and influential in the Ilkhanate and
elsewhere.
Cassiodorus's remarks on the 'seethings of Barbarism' imply, too, a notable
distinction in his mind—and perhaps also in that of Theodoric—between
Ostrogothic Italy, the true heir of the Empire, and the other barbarian
kingdoms which had been set up in the former provinces of the western
Empire. The identification was not, however, by any means total: not even
Cassiodorus could pretend to believe that. King Athalaric, on his accession
to the Ostrogothic throne in 526, swore an oath to the citizens of Rome
'…by the Lord's help to observe justice and fair clemency, the nourisher of the nations; that Goths
and Romans shall meet with impartial treatment at our hands; and that there shall be no other division
between the two nations, except that they undergo the labours of war for the common benefit, while
you are increased in numbers by your peaceable inhabitancy of the City of Rome'.74
And Athalaric wrote to Gothic settlers in terms reminiscent of Ghazan's
common-sense speech to the Mongol amirs.
'If anyone is in need of anything, let him seek to obtain it from the generosity of his Sovereign rather
than by the strength of his own right hand, since it is for your advantage that the Romans be at peace,
who, in filling our Treasury, at the same time multiply your donatives'.75
Further, there are many references to a specifically Gothic officer, the saio,
to some extent fulfilling functions similar to those of a Mongol ilchi, and
also to more senior officials such as the Comes Gothorum, clearly a
necessary figure in a mixed Roman-barbarian society. The formula for his
appointment is recorded by Cassiodorus.76 Gothic officials seem to have
duplicated, in some sense, the regular Roman administration:' in the
provinces Gothic military governors (comites Gothorum provinciae)
appeared at the side of regular civil Roman governors. In the cities other
Ostrogothic counts (comites Gothorum civitatum), technically no more than
the equal of local Roman officials, limited the legal jurisdiction of the latter
and extended their own competence over the supervision of public
buildings and roads and even over the collection of taxes'.77 These are
almost exactly paralleled by a Mongol official, the basqaq, who was
stationed in occupied territory, and also in tribute-paying areas which
retained some degree of local autonomy. He supervised the collection of
revenue, and ultimately became in effect a kind of provincial governor.
Local revolts were often inaugurated by the murder of the basqaq.
Much of the evidence so far cited gives an impression of barbarian rule that,
particularly in the case of Persia, is far from favourable. There is something
to be said on the other side. It is worth pointing out the comparative
religious tolerance of both regimes. In Theodoric's case, though an Arian,
he refrained from persecuting his Catholic subjects until the very end of his
reign, and died before he could do them much harm. The attitude of his
government towards the Jews is notable. Giving the Jews of Genoa
permission to re-roof their synagogue, Cassiodorus, in Theodoric's name,
exclaims: 'Why do you desire what you ought to shun? In truth we give the
permission which you craved, but we suitably blame the desire of your
wandering minds'—tolerant, if irritable; but Cassiodorus concludes with a
remarkable sentence: 'We cannot order a religion, because no one is forced
to believe against his will'.78 Stern measures were taken against cities where
rioting against the Jews took place, as may be seen both in the Variae and in
other sources.79
The Mongols were tolerant of all religions—a tolerance perhaps founded
more on indifference than on any particular philosophy of toleration.80 In
particular, Chingiz Khan insisted that the clergy of all religions should be
exempt from taxation. This is echoed, at least as far as Muslim men of
religion were concerned, in one of Ghazan Khan's edicts, recorded by
Rashid al-Din.
'And since the command of the great yarligh of Chingiz Khan is this, that and learned men and
descendants of 'Ali should not pay qalan and qubchur, we have commanded that for this reason they
should be exempted, and their tax (mal) and qubchur should not be taken; and post-horses and
provisions for travellers (susun) should not be taken from them. No one should billet himself in their
houses, and ilchis should not descend on them.'81
The conversion of Ghazan to Islam meant that toleration was no longer the
automatic right of religions other than Islam: the brief heyday of Jews,
Christians, and Buddhists was over. Buddhist temples and—apparently—
Zoroastrian fire-temples were destroyed, and Buddhists given the
alternatives of accepting Islam or leaving the Ilkhanate.82 As for the
Christians, it is clear from the biography of the Nestorian Catholicus Mar
Yaballaha III83 that their position reverted to that, usual in Islamic societies,
of general tolerance as second-class citizens, interspersed with occasional
bouts of persecution.
It should also be mentioned that both barbarian governments engaged in
extensive programmes of city embellishment and construction. Cassiodorus
includes several letters in which he speaks in enthusiastic terms of the
embellishment of Rome.84 And the building of his great quarter, the Rab'-i
Rashidi, in the Mongol capital of Tabriz, was very dear to the heart of
Rashid al-Din, as many of his letters show.85 There are letters concerned
with the building of new canals, named after both Ghazan Khan and Rashid
al-Din, together with their associated villages, and plans are included in the
letters.86
These, then, are some of the aspects of Ostrogothic and Mongol rule
illustrated by Cassiodorus and Rashid al-Din. To what extent should the
picture painted on the strength of such evidence be accepted at its face
value? So far as Cassiodorus is concerned, Gibbon at least had no doubt of
what were the necessary reservations.
'The volume of public epistles composed by Cassiodorius in the royal name, is still extant, and has
obtained more implicit credit than it seems to deserve. They exhibit the forms, rather than the
substance, of his (Theodoric's) government; and we should vainly search for the pure and
spontaneous sentiments of the Barbarian amidst the declamation and learning of a sophist, the wishes
of a Roman senator, the precedents of office, and the vague professions which, in every court and on
every occasion, compose the language of discreet ministers.'87
Thomas Hodgkin expressed a similar opinion when he observed that 'we are
therefore really without a picture of the Ostrogothic kingdom of Italy from
the
true Ostrogothic point of view'.88 More recently, doubts have been
expressed by Professor J.M.Wallace-Hadrill, who suggests that' the
Theodoric we know from the Variae of Cassiodorus and from Ennodius is a
ruler of Romans devout in the service of Romanitas; the Theodoric his
Gothic followers knew was a Germanic war-leader and a very different kind
of man'.89
It is barbarian government, then, filtered through the mind of a Roman who
was determined, before all else, to stress unbroken continuity with the
Roman past. It would be enlightening to have some kind of Ostrogothic
equivalent to the 'Secret history of the Mongols', as a yardstick with which
to measure the points of view of outsiders. But no such work exists, nor,
perhaps, could be expected to exist.
Nor is acceptance of Rashid al-Din's view of affairs entirely a
straightforward matter. Some of the details in the Mukatabat-i Rashidi
strain the reader's credulity. If the foodstuffs mentioned in letter 45 are, as
seems to be the case, being sent to Shaykh al-Din of Ardabil for a single
feast, the amounts concerned are almost unbelievably colossal. And, in
view of what Rashid al-Din tells us about the general condition of the
Mongol administration in Persia, one sometimes wonders what degree of
connexion exists between theory and reality—especially when Rashid al-
Din sends enormously detailed instructions to some remote part of the
kingdom. How much control did Rashid al-Din, or anyone else, have over
what in fact went on in the provinces—except sporadically, when troops
could be sent to enforce the central government's will? 'The moment control
was relaxed, there was a tendency to relapse into the old habits, and thus it
was a constant struggle to restrain the officials from committing extortion
against those under their power'.90 The letters, or many of them, are
themselves the record of Rashid al-Din's efforts to keep the local officials in
check. We know what orders he sent: what we cannot know is whether their
recipients took any notice, and, if not, what was done about it.
There are, similarly, problems in knowing how to assess Rashid al-Din's
account of the state of affairs before Ghazan Khan's accession, and the
effectiveness of Ghazan's reform programme. The blacker the picture of
things before 1295, the greater would appear Ghazan's achievement in
putting everything right. Ghazan's achievement is, in effect, that of his chief
minister, Rashid al-Din. Certainly, other contemporary writers do not by
any means give the prominence to the reforms which is accorded them by
Rashid al-Din. When all this is said, however, there is no denying that,
subjectively at least, Rashid al-Din's picture of Persia under the early
Ilkhans, immensely detailed and circumstantial as it is, carries a good deal
of conviction, and is confirmed in its general outlines by such other sources
as the Tarikh-i
In the light of these considerations, it is worth asking what real degree of
administrative continuity is to be found between the Ostrogoths, the
Mongols, and their respective predecessors. 'He kept the civil service for the
Romans on the same footing as under the emperors', wrote the author
known as the Anonymus Valesianus of Theodoric.91 This was no doubt true
to a very large extent: 'Theoderic, and Odoacer before him, inherited the
central government of the empire, and Rome itself, with its Senate. While
the other barbarian kings improvised central governments of their own
making, Odoacer
and Theoderic, if only by force of inertia, maintained the ancient offices of
the imperial comitatus and the praetorian prefecture'.92 Some believe,
however, that there were fairly considerable changes. The Variae, it has
been suggested, tends to be misleading. For example, Cassiodorus's formula
for the Mastership of the Offices reveals an ideal conception of the post,
rather than reflecting the realities of the day. And even from the evidence
provided in the Variae, it is possible to see that Gothic saiones were taking
over some of the functions of the agentes in rebus, in the operation of the
postal service. So one may conclude, perhaps, that 'while the Roman
aristocracy was to be respected, it was on important occasions to be given
more of the semblance than the reality of governmental authority'.93
But, with all these qualifications, the administration of Ostrogothic Italy, as
far as may be judged, was in its essentials the administration of the late
western Roman Empire. Was such continuity also characteristic of Mongol
Persia?
Such continuity did indeed exist, and for the same reason as in Italy: inertia.
The Mongols certainly tried out their own ideas on exploitation—these
have been illustrated above. But the basic administration of the country
seems to have continued, staffed by Persian officials, virtually without a
break. The Mongols had their own approach to taxation, and in this they
contrast with the Ostrogoths. The characteristic Mongol qubchur, at first a
tax on pasture or flocks, later a general poll-tax, had to be operated by the
Persian officials in addition to the traditional Islamic taxes such as kharaj.94
The military establishment was Mongol and Turkish, as that of Italy was
Gothic. Mongol law coexisted, in ways not easy to elucidate, with Islamic
law. Provincial governors were sometimes Mongols, sometimes Persians.
Mongol basqaqs kept an eye on what Persian administrators were doing.
Nomadism greatly increased, reducing the amount of cultivated land: this
reduced the scope of the central government in extracting revenue from the
peasants. But essentially the Persians ran the governmental machine on
behalf of their barbarian masters, with interruptions and interference from
those masters. The machine continued to work as it had always done: it
simply did not work as well as before the arrival of the Mongols. According
to Allah Mustawfi's calculations (which, unfortunately, there is no
way of checking), under the Saljuq sultan Malik Shah, at the end of the
eleventh century, the annual revenue of Persia amounted to 500,000,000
currency dinars. At the time of the accession of Ghazan, it was 17,000,000.
As a result of his reforms, it rose to 21,000,000. Whatever we may make of
these figures—and it is worth noting how small a rise in revenue is said to
have resulted from Ghazan's efforts—the author was clear in his own mind
about their implications.
'From the above a comparison may be made between the populousness of the world (in the past) and
its ruin (at the present day). There is no doubt that the destruction which happened on the emergence
of the Mongol state and the general massacre which occurred at that time will not be repaired in a
thousand years, even if no other calamity occurs; and the world will not return to the condition in
which it was before that event.'95
However, it is worth adding that the Mongols to some seemed, with all their
faults, preferable to the anarchy which ensued after the death of the Ilkhan
Abu Sa'id. The author of the Tarikh-i Ruyan, writing only 25 years later,
praised the peace and security of the period of Mongol rule.96 He eulogized
Abu Sa'id—not, in the estimation of most historians, among the more
notable of the Ilkhans. Even accepting that there seems to be a widespread
human tendency to create non-existent Golden Ages in the past, it
nevertheless seems extraordinary that, as early as 1362, the Ilkhanate was
beginning to be seen in such a light. Perhaps there is something to be said,
after all, for the Islamic political theorists' tenet that any government,
however bad, is better than no government at all. No doubt the Romans of
Italy would have agreed, as they watched the armies of Justinian march
back and forth for years on end, and as they too, perhaps, looked back with
nostalgia to their period of barbarian government.
 THE TREATIES OF THE EARLY MAMLUK SULTANS
WITH THE FRANKISH STATES

By P.M.HOLT

Arabic sources have preserved the texts of seven treaties concluded in the
second half of the seventh/thirteenth century between the Mamluk sultans
Baybars and Qalawun on the one hand, and various
authorities in the Latin kingdom of Jerusalem and in Antioch-Tripoli on the
other. Four are given by the Egyptian chancery clerk al-Qalqashandi in his
encyclopaedic compilation, al-a'sha. This was completed in 814/1412,
over 120 years after the extinction of the Frankish states, but the treaties
were transcribed (as al-Qalqashandi tells us) from an earlier work by a clerk
in Qalawun's chancery, b.al-Mukarram. Two of Qalawun's other
treaties are found in his biography, written by the contemporary chancery
clerk, al-Din ibn 'Abd (d. 692/1292). Yet another of Qalawun's
treaties was preserved by the contemporary chronicler, Baybars (d.
725/1325). 1
The purpose of this paper is fourfold: first, to indicate the status of these
treaties in Islamic law; secondly, to present some information from
contemporary Arabic sources on the procedure followed in their
negotiation, drafting and ratification; thirdly, to consider some aspects of
their form and contents; and finally, to illustrate the circumstances in which
they might be abrogated.
The distinction made in Islamic international law between two necessarily
and perpetually hostile domains, Dar al-Islam and Dar was no longer
very appropriate to the situation in the later thirteenth century, when a web
of commercial and political relations linked the Muslim and Christian
powers around the Mediterranean, but at least in form it had to be respected.
Hence these treaties fall into the category of truce (hudna), concluded only
for a limited period, of which the permissible length was a subject discussed
by jurists, and al-Qalqashandi introduces his texts with a summary of legal
opinions on this point. Two positions are stated. If the Muslims are strong, a
truce should not last for more than four months, or in any case one year. If
the Muslims are weak and in fear (i.e. under duress), a truce for ten years
may be concluded.2 It is some indication of the gap at this period between
legal doctrines and practical politics that these treaties with the Frankish
states are all specified as of ten years' duration, although it is absurd to think
of those Turkish warrior-kings Baybars and Qalawun negotiating out of
weakness and fear with the authorities at Acre or the Lady Margaret of
Tyre.
The actual situation is made very clear by the procedure followed in the
negotiation of the truces. The initiative was invariably taken by the
Frankish party, whose ambassadors waited on the sultan for the start of
negotiations. For example, early in Baybars's reign (in 659/1261), the sultan
brought his army into Syria. This action provoked an immediate response
from the Latin kingdom. In his laudatory biography of Baybars, Ibn 'Abd
describes the episode as follows:
When he [i.e. the sultan] reached Damascus, an envoy arrived from Acre to ask for a safe conduct for
the ambassadors despatched by all the military orders. He wrote instructing the governor of Banyas
to enable them to proceed. So the leading men of the Franks arrived and requested peace. The sultan
demurred, and made a number of demands on them. When they refused, the sultan upbraided and
slighted them. The army had already set out from the direction of Ba'albakk to raid their territory, and
they asked for its recall. There happened to be a dearth in Syria, and imports on a large scale could
only come through the Frankish territories, so peace was concluded on the basis existing at the end of
the reign of [i.e. Yusuf, the Ayyubid ruler of Aleppo and Damascus until
658/1260], with provision for the release of captives taken between the end of the said reign and the
time of the truce. Ambassadors went with them to obtain their ratification. A truce was likewise
concluded with the lord of Jaffa and the ruler of Beirut on the basis of truce at the end of
his reign. The roads became safe and imports abounded.3
In a note appended to his transcription of the texts, al-Qalqashandi throws
an interesting light on the way in which the provisions were actually
formulated. He writes:
They are in common language and inelegantly arranged; their like would not be drawn up by any
clerk with the least skill in drafting…. Perhaps the reason was that in those days the Franks were
neighbours to the Muslims in Syria, and terms of agreement would be reached by the two parties
clause by clause. Then a clerk from each of the two parties, the Muslims and the Franks, would set it
down in common, inelegant words for reasons of speed until they concluded the terms of agreement
down to the last clauses of the truce. Then the clerk of the Muslim king would write it according to
the tenor of the draft, to match what the Frankish clerk had written. If the sultan's clerk were to make
any emendation in it, in the arrangement, the improvement of the words and the eloquence of the
composition, this would have departed from what the Frankish clerk had previously agreed to, and
thereupon they would disown it, believing that it was not what had been agreed, owing to their lack
of Arabic. So the clerk had to keep to what the two clerks had agreed to in the draft.4
What perhaps al-Qalqashandi did not realize was that by the late thirteenth
century there had developed a good deal of common form in the treaties
made
between Franks, whether of Syria-Palestine or the Italian republics, and
eastern Mediterranean rulers, whether Ayyubids, Mamluks, Seljuks of Rum
or Byzantines.
Although a truce was negotiated and drafted at the sultan's court, whether at
Cairo or in the field, a copy being filed in the sultan's chancery, it had also
to be ratified (as the passage quoted from Ibn 'Abd has indicated) by
the oaths of the two parties. The Frankish. ratification was obtained by an
embassy sent by the sultan. On such occasions there were two ambassadors,
one a senior Mamluk amir, the other a high chancery official. We see the
procedure in 682/1283, when a delegation from. Acre, consisting of two
Templars, two Hospitallers and two representatives of the Latin kingdom
negotiated a truce in Cairo. Having sworn to the truce, Qalawun then sent
his own embassy consisting of his chamberlain and a judge, to obtain an
oath of ratification from the Franks.5 In this instance we are fortunate in
having not only the text of the truce but also the oaths of ratification.
Qalawun's oath begins with. a ninefold invocation of the name of God, and
undertakes to respect the duration and terms of the truce, under the sanction
of making thirty pilgrimages to Mecca. The Frankish oath begins with a
long and elaborate invocation, confirms the duration and terms of the truce,
and provides as its sanction thirty pilgrimages to Jerusalem and the release
of a thousand Muslim captives.6
An insight into the hazards of negotiation is thrown by Ibn 'Abd
nephew, Shafi' b.'Ali, himself a chancery clerk, in his account of the truce
with Tripoli in 669/1271, where he describes an incident told him by his
uncle in these words:
Al-Malik [Baybars] laid siege to it [i.e. Tripoli], and ambassadors went to and fro between
him and its lord in search of peace. When agreement was reached, the Amir Faris al-Din the
atabak together with the high official al-Din ibn 'Abd went to Tripoli for the
conclusion of the truce on the terms laid down. Al-Malik resolved to disguise himself and
go in behind the atabak like a sword-bearer, in order to have a look at the fortress. When they
attended the court of the prince and the matter was settled, al-Din began to write in these
words, 'Truce is established between the Sultan al-Malik and His Highness the chief'. When
he saw it, the prince said, 'Who is the chief?' He said, 'You are'. He said, 'I am a prince'. He said,' No.
The prince is al-Malik for "the prince" means the lord of Jerusalem, Alexandretta and
Antioch, and these belong to our lord the Sultan al-Malik He was annoyed about this, which
embarrassed him, and his anger appeared. Al-Malik reckoned it as clerks' pettifogging, and
gave him [i.e. the atabak] a gentle kick, which was however sufficient for the purpose. The atabak
turned and said, 'You are right, al-Din. This name belongs to the sultan, and he has granted it
to this person as he has granted him his fortress, his lands and his subjects. I will go surety for him in
this matter'. Thereupon he wrote '…and the prince'.7
Shafi' then goes on to describe the negotiations in Damascus with Qalawun
for a renewal of the truce in 680/1281, when a major difficulty arose.
When the prince's ambassadors presented themselves before our lord the sultan, the Master al-
Din, the head of chancery, was present and I with him. The sultan proceeded to ask the ambassadors
in Turkish why they had come. The [Frankish] minister, Ghurab, got up. He was a pettifogging,
cunning, artful fellow. He answered our lord the sultan's question in this way, 'Bohemond, the friend
of his government, the ally of the glorious sultan, asks the glorious sultan to continue the friendship
which was between him and al-Malik according to his truce'. Our lord the sultan replied,
'Where is the money you have brought? What territory will you give me?' This Ghurab exercised his
skill as an ambassador until he had softened the disposition of our lord the sultan by his pleasant
manner and his agreeable phraseology…. He was bidden to be seated, and he produced the truce of
al-Malik with his signmanual, and we for our part presented the copy filed with us. From
this it was evident what al-Malik had authorized and what he had withheld. A comparison
was made letter by letter, and their agreement obliged one to say that comparison showed its
correctness, until, when we came to the town of 'Arqa, which is one of the best and most profitable of
their lands, the Master al-Din, the head of chancery, said to him, 'However, minister, the terms
of the truce do not apply to this place 'Arqa'.
Shafi' explains the reason for the exclusion of 'Arqa:
Al-Malik had made a truce with them. [i.e. the Franks of Tripoli] on condition that they
supplied thousands of gold pieces and thousands of Muslim captives. The Amir Sayf al-Din Balaban
the dawadar went to take the gold and receive the captives. He stayed six months with them, while
they deceived and tricked him, putting him off with hindrances as is their custom—may God
Almighty curse them. So the Amir Sayf al-Din moved about until he came as a fugitive to al-Malik
and al-Malik confiscated this place 'Arqa from the beginning of the truce until his
death.
Confronted with the exclusion of 'Arqa from the truce, Ghurab began to
argue a case, claiming that the delay in fulfilling the terms of the agreement
resulted from the need to collect the promised tribute and to assemble the
captives. On being asked by al-Din why nothing had still been done in
the second, third and fourth years, he shifted his grounds and undertook the
delivery of the tribute and the captives, but went on:
'We wish for your decree that an account should be made of what al-Malik seized from this
place 'Arqa over a period of ten years.' This was an indication that they would be in credit, not in
deficit. The head of chancery was disturbed, and asked the sultan what to do. He drew his sword on
him, and was about to have his head.
Shafi' then describes his own part in the proceedings:
I said (and the most senior of the amirs of the council, the Amir 'Ala' al-Waziri,
heard me), 'Are we in a court for giving judgment to one another, allotting shares or administering the
law to one another? There is an answer to what he says'. He said to me, 'And what is the answer?' I
said, 'If our lord the sultan commands, I will answer him and argue with him'. On being commanded,
I said to him, 'Ghurab, al-Malik confiscated this place 'Arqa only to annoy you, not to do
you a kindness by taking the tribute in instalments. The labourers, the seeds and the peasantry in this
settlement came from him, not from you; through his officers, not through your officers. Its land is
possessed by the sword—nay indeed, everything is, for the truce has been invalidated by your
breaches of the terms. In taking a part, he let you keep the rest. You have compared the party in
breach with the party in good standing, as if it had been the comparison of two parties in breach'. I
burst out against him like the outburst of a flood. The unbeliever was bewildered, and agreed without
dissembling. The truce was established on these terms. Our lord the sultan ordered the Amir Sayf al-
Din Balaban al-Rumi the dawadar to go and receive the prince's oath and take the tribute, and I was
to go with him.8
Although as this incident shows, the scales were weighted against the
Frankish ambassadors in negotiations, the truces themselves were in form
bilateral treaties as between equal parties, and they follow a regular pattern
of preamble, provisions and concluding formulae. Al Qalqashandi groups
the instruments he presents according to the opening words of their
preambles, and the truces with the Frankish states are characterized by the
formula Istaqarrat al-hudna bayn fulan wa-fulan, 'Truce is established
between A and B'. The titles of the two parties are given, the sultan's heir
being associated with. him as joint ruler in some of the treaties; then
follows the duration and date of commencement of the truce. The ten-year
period is sometimes elaborated (as in Qalawun's truce of 682/1283 with
Acre) to 'ten whole years, ten months, ten days and ten hours', while the
date is invariably specified according to both the Hijri and Seleucid eras.
The use of the latter (designated in the treaties as the era of Alexander, son
of Philip the Greek) suggests that the clerks employed on the Frankish side
were Orthodox Christians or Jews. The territories of the parties to the truce
are then usually detailed. In discussing Qalawun's treaty with Acre, Joshua
Prawer sees an ironic contrast between the immense sweep of the sultan's
dominions and the poor remains of the Latin kingdom.9 The irony is there,
but it may be an unintentional consequence of the myopic legalism of the
drafters of the truce.
The provisions which form the body of the truces naturally vary in the
different instruments, some of which (e.g. that concluded between Qalawun
and Bohemond VII of Tripoli in 680/1281) are very local in their scope.
Several of the provisions, however, recur in over half of the truces. These
have the object of securing the sultan's territories and, conversely, of putting
the Frankish party at a military disadvantage. The danger of a crusade still
haunted the imagination of Muslim rulers, and the denial of assistance to
the sultan's enemies is several times required. The most detailed stipulation
occurs in the treaty with Acre:
If one of the Frankish kings of the sea or others should move by sea with the intention of coming to
harm our lord the sultan, or to harm his son, in the lands to which this truce applies, the bailli of the
kingdom and the grand masters are bound to inform our lord the sultan of their movement two
months before they arrive in the lands. If they arrive after the lapse of two months, the bailli of the
kingdom in. Acre and the grand masters shall be exempt from responsibility under oath in regard to
this clause.
If the Mongols or another enemy move by land, whichever of the parties has first notice shall inform
the other.10
Another common provision forbade the construction or reconstruction of
fortifications, as in Qalawun's treaty of 684/1285 with the Lady Margaret of
Tyre:
The lady Dame Margaret, the lady of Tyre, shall build no new castle, nor renew the wall, nor dig a
new trench, nor renew any fortification for defence or offence.11
An interdiction is laid in some treaties on trade in 'forbidden goods'
(almamnu'at), i.e. military equipment. Finally there is sometimes provision
for the safety of the naval vessels of the two parties.
Another group of provisions is concerned with the security and policing of
the borders between Islamic and Frankish territory. By the late thirteenth
century the situation on the frontiers had been complicated by the
retrocession of former Frankish districts to the sultan, and by the creation
(often as a preliminary to retrocession) of condominia i.e.
districts of which the revenue, and presumably the administration, were
shared between the two parties. In six out of the seven treaties here
considered are clauses dealing with. procedure in the case of homicide or
robbery, and it seems likely that they cover not merely individual crimes but
also border raids from one side or the other. From a collation of the
different treaties, it appears that a fairly elaborate system had developed.
The simplest case was when the killer was taken. Here the truce with the
Lady Margaret of Tyre is the most specific:
Wh anyone of either party is killed, and the killer is found; if he is a Muslim, the delegates of our
lord the sultan (God grant him victory) shall judge him in accordance with the pure administrative
law [al-siyasa] of the august sultanate. If he is a Christian of the people of Tyre, the lady Dame
Margaret, the lady of Tyre, shall judge him. Each party in the presence of a delegate of the other
party shall proceed to judgment in accordance with the laws of the two parties. That shall be the
procedure in regard to all who trespass, damage or commit murder. The delegates of our lord the
sultan shall be charged with the punishment of a Muslim, and the delegates of the lady, the lady of
Tyre, shall be charged with the punishment of a Christian.12
Similarly, when a stolen article or booty could be found, it was to be
restored, or compensation paid.
Difficulty arose when a killer absconded, or stolen goods were concealed.
The procedure in such an event is laid down in similar terms in five of the
treaties. Forty days of grace (muhla) were allowed for the investigation of
the matter. After this, if no discovery had been made, the administrator of
the locality13 against which the accusation lay, and three other persons
chosen by
the accuser were put on oath to say what they knew, or to clear themselves.
If they refused to swear, the accuser's sworn assertion of the value of the
missing goods was accepted. Compensation for a homicide was usually
made by the release of a captive of equal standing with the slain man: 'a
knight for a knight, a Turcopole for a Turcopole, a merchant for a merchant,
a foot-soldier for a foot-soldier, and a peasant for a peasant'.14
The treaty with Tyre (684/1285), however, lays down a tariff of
bloodmoney (diya), levied from the villages where the killing occurred:
The blood-money for a knight of either party shall be 1,100 Tyrian dirhams, for a Turcopole 200
dirhams, for a peasant 100 dinars [sic]; the blood-money for a merchant shall be according to his
nation, origin and standing. It shall be taken from the people of the villages in which that person was
killed as being their fine and a collective punishment.15
There are some variations in two of the treaties. Baybars's first treaty with
the Hospitallers (665/1267), in dealing with the possessions of
 the Ayyubid lord of and those of the Assassins, fixes the
period of grace at fifteen days, after which, if booty is not returned (there is
no mention of homicide) the accused only is put on oath. His second treaty
with the Hospitallers (669/1271), dealing with the condominium in the
territory of al-Marqab, allots twenty days of grace for investigation. Then, if
the culprit is not discovered, 'the ru'asa' of the locality of the highway
robbery, taking of booty or homicide…shall detain in place of the killer or
the thief the nearest of the neighbours to the highwayman or the killer'.16 If
the culprit absconds, and is not brought in within twenty days, a fine of
1,000 Tyrian dinars, divided between the co-domini, is laid upon the nearest
neighbours. Qalawun's treaty with Acre provides that, if justice is denied by
the local administrator, the accuser might petition the authorities on both
sides. If after forty days the administrator had not rendered justice, the ruler
who appointed him was to put him to death and confiscate his chattels.
Another problem of the borders arose from the presence of Muslim
fugitives in Frankish territories and vice versa, a situation rendered more
delicate when a fugitive professed conversion. Here the stipulations vary.
Two of the treaties, Baybars's second with the Hospitallers and Qalawun's
with the Lady Margaret of Tyre, are uncompromising: fugitives must be
sent back with all they brought with them. In Baybars's earlier treaty with
the Hospitallers and Qalawun's with Acre, there is greater discrimination.
Chattels brought by a fugitive are in any event to be returned, but the
fugitive himself (in Baybars's treaty) is to have the option of returning or
staying. This is presumably if he is a freeman, since the clause continues:
If a slave ['abd] flees and abandons his religion, his price shall be returned; if he keeps his religion,
he shall be returned.17
The treaty with Acre specifically extends this condition to all fugitives from
both sides. A later clause lays down the procedure to be followed when a
fugitive is suspected of not having restored all the chattels he brought with
him. A unique clause (which cannot be wholly harmonized with the
provision mentioned above) debars peasants from the rights of fugitives:
Proclamation shall be made in the Islamic lands and the Frankish lands
included in this truce that anyone who was a peasant in the lands of the
Muslims, were he Muslim or Christian, shall return to the lands of the
Muslims; likewise anyone who was a peasant of the lands of the Franks,
were he Muslim or Christian, his domicile being recognized by both.
parties. Anyone who fails to return after the proclamation shall be expelled
by either side. The peasants of the lands of the Muslims shall not be enabled
to reside in the lands of the Franks to which this truce applies, nor the
peasants of the lands of the Franks to reside in the lands of the Muslims to
which. this truce applies. The return of a peasant from the one party to the
other party shall be under a safe conduct.18
The security of travellers is the subject of clauses in several truces and here
there are echoes of the terminology of treaties in European languages.19 The
treaty with the Lady Isabel of Beirut, for example, contains the following:
All those passing from and to these territories shall be safe and secure in respect of themselves, their
chattels and their wares from the Lady [Isabel] and her servants.20
One may compare the phraseology of the capitulations granted by the
Ayyubid Sultan al-'Adil Abu Bakr to the Venetians in 1238:
…et hebeant fidantiam in personis et habere et in mercimoniis, que veuiunt cum eis, venientes,
permanentes et revertentes…21
The treaty of 669/1271 details the arrangements for the convoy of
merchants between the sultan's territories and the Hospitallers' lands around
Margat:
Travelling merchants and those going to and fro with goods from the territories of the Muslims and
Christians shall proceed on leaving the ports defined above under escort of the two parties without
[payment of] duty. Nothing shall be accepted by the escort in regard to their persons until it has
brought them out and delivered them to the land-boundaries of al-Marqab safely and securely under
guard of both parties.
Wh merchants arrive from the kingdom of the sultan at the territory and ports of al-Marqab, both
parties are to organize the escort with patrols of the village headmen [ru'asa'] guarding the roads
both on leaving and entering, so that they may come to the territory of al-Marqab and the ports of al-
Marqab defined above safe and sound in respect of themselves and their chattels under escort of both
parties, as we have set forth.22
It is not possible, in fact, to draw a hard and fast line between political
treaties, such as these truces ostensibly are, and commercial treaties, such as
were concluded with Venice, Genoa and Aragon.23 Four of them have
clauses
concerning dues on merchants, three of them forbidding the imposition of
any new duty. The truce with. Acre provides for the safe keeping of the
chattels of Muslim merchants dying in Frankish territory and vice versa.
These are common form in the commercial treaties of the period. A final
subject of commercial rather than political importance is action in regard to
wreckage, flotsam and jetsam, which. is dealt with in four truces. A
succinct example occurs in the treaty with the Lady Margaret of Tyre:
When a ship of either party is wrecked, if it belongs to a Muslim, it shall be delivered to him, if he is
found; and to the delegates of our lord the sultan, if he is lost. If it belongs to a Christian from the
territory of our lord the sultan (great be his victory), the procedure shall be as in the case of a
Muslim. If he is of the people of Tyre, from the subjects of the lady, the lady of Tyre, the chattels
shall be delivered to him, and to her administration [diwan] if he is lost.24
The continuing validity of the truce throughout the period specified is
guaranteed in six of the seven treaties here under consideration. The death
or removal of one of the parties was not to annul the truce, and other
possible pretexts for annulment are sometimes excluded. The lapse of the
truce was to be followed (according to some texts) by forty days of grace to
give people time to return to their homelands. Nothing could seem more
assured and explicit than this, but the evidence of the Arabic writers
suggests that these undertakings were little more than formalities. For
instance, Baybars's treaty of 665/1267 with the Hospitallers contains a
specific provision against abrogation before the end of its prescribed period,
yet Ibn 'Abd twice states explicitly that 'it was stipulated that the
sultan might abrogate it at will'.25
One may appropriately end with one particular and final example of
abrogation: that by Qalawun in 689/1290 of the treaty he had concluded
with the Latin kingdom. Shafi' b.'Ali, who had himself drafted the treaty,
tells the story both in his biography of Baybars and in that of Qalawun. The
two accounts are substantially the same; the version in the biography of
Qalawun is as follows:
Our lord the sultan suddenly learnt that the Franks in Acre had acted arbitrarily towards a number of
Muslim merchants and others there, and had killed them, esteeming them lightly in spite of their
number. Thereupon our lord the sultan commanded and wrote to them, and reminded them [saying],
'Know that this breaks the covenant and casts out friendship'.
A letter arrived from their chief men saying, 'This happened only because the Franks and Muslims
were gathered in a tavern, and drunkenness caused them to quarrel. We have arrested a number of the
Franks who were in the tavern, and hanged them'. An answer was returned to this effect, 'You have
rightly said that you hanged them—but they were Muslims. We are coming to you by the will of God
Most High, so prepare a hospitable reception'.
Then our lord the sultan summoned the magnates of his council, and discussed with them a campaign
against that people because of this disturbance…The magnates objected on the grounds of the truce,
which was secured by oaths and was equivalent to a guarantee of safety. It implied a covenant, and
the keeping of a covenant is a religious obligation.
Then our lord the sultan indicated to the head of his writing-office, al-Din, that he should go
over the truce. Perhaps he would succeed in finding some reason for attacking that people, and there
might be suggested by it and by what had occurred the means of obtaining his wish concerning
them…I had drafted their truce and had a copy of it. So we met together, I and his father, the
al-Din, and Le himself. I read the truce from beginning to end several times. The
al-Din was convinced that it offered no scope or ground for abrogation either explicit
or implicit. He turned to me and said, 'What do you say?' I said, 'We serve the purpose of our lord the
sultan. If he wishes to abrogate it, there is scope for requiring its abrogation. If it is not his purpose to
abrogate it, its text does not require its abrogation'…So I said, 'Let our lord consider this clause of the
truce, viz. "Provided that the merchants, the ambassadors and those going to and fro shall be safe and
protected by the two parties in travelling, residing, going and coming". This is the text of the clause
contained in the treaty'.
So I said to him, 'Those to whom this applies were merchants, and the provision of the truce has been
broken through the neglect of their case, unless there is some information [?] from a Muslim delegate
appointed to adjudge pleas'. He said, Letters have arrived from the delegates that the matter was not
as they asserted, and that those hanged were Muslims'. So I said, 'The truce has been abrogated in this
stipulation'.26
Thus was set in motion the mobilization of the Mamluk forces which, in the
following year, after Qalawun's death, were to accomplish the capture of
Acre and the extinction of the Latin kingdom of Jerusalem.
 THE MONGOL EMPIRE: A REVIEW ARTICLE

By D.O.MORGAN

The publication of a new history of medieval Central Asia,1 over half of

which is concerned with the empire of the Chingizid Mongols, provides an
opportunity for a survey of a number of books on that subject that have
appeared during the past decade. Professor Kwanten's book is, more or less
avowedly, an attempt to replace René Grousset's L'empire des steppes as the
standard introduction to Central Asian history. I begin, then, with the
English translation of Grousset's celebrated book.2
Denis Sinor rightly remarked of L'empire des steppes, which was first
published in 1939, that 'no other work encompasses as accurately as much
of the multifaceted history of Central Eurasia'3—a judgement which. I fear
must still stand, regrettable though that may be after forty further years of
scholarship. It is still, perhaps, best read in Grousset's elegant and lucid
French, though history teachers will hardly waste their time in telling their
students so. Grousset's weakness was that he knew, apparently, no Oriental
languages, and so was entirely dependent on translations and secondary
works—and, as Sinor interestingly brings out, on the close interrogation of
those, notably Paul Pelliot, who did know the sources in the original
tongues. Still, though it is to some extent, therefore, a tertiary work,
L'empire des steppes remains to this day a far from negligible introduction
to its subject.
But forty years is a long time (though Barthold's Turkestan down to the
Mongol invasion has lasted even longer), and the Mongols have never been
short of authors fascinated by the extraordinary phenomenon of their
conquests. In 1969 appeared E.D.Phillips's The Mongols,4 a well-informed
and nicely illustrated book by a classicist interested in the history of
nomadism. But this is a very short book which because of the pressure of
material tends to degenerate into a catalogue of khans and battles. Much
more important was the publication in 1971 of The history of the Mongol
conquests, by J.J.Saunders. This excellent book must surely now be the first
general work to recommend to anyone interested in the empire of Chingiz
Khan and his successors. Like Grousset's book, it is a work of synthesis,
based on Saunders's extremely wide reading in translated sources (of which
much more was available to him than to Grousset) and the secondary
literature. It is a book not without minor slips (most of them pointed out by
Professor Beckingham in BSOAS, XXXV, 2, 1972, 392–3), but as a whole it
is an impressive achievement, strong on the pre-Chingizid historical
background, and never falling into the sort of bare narrative trap that
ensnared Phillips. And it has most useful and detailed notes, as well as an
admirable annotated bibliography.
Despite the great merits of Saunders's book, Peter Brent's The Mongol
empire (London, 1976), should not be passed over without mention. Mr.
Brent is a professional writer rather than a historian, let alone a specialist in
the history of Central Asia. His book is aimed at the interested general
reader (it is devoid of notes and bibliographical references, but is profusely
illustrated), and is a well-written narrative history. The author is clearly well
informed,
and makes remarkably few factual mistakes—fewer than the more learned
Saunders, to say nothing of Kwanten. His judgements are usually sound and
reliable. I very much enjoyed reading the book, which for a moment
brought back the astonishment and interest aroused by the first book that, as
a schoolboy, I ever read on the Mongols. Mr. Brent deserves a wide sale for
this excellent piece of popularization.
It is instructive to turn from Brent to a more recent book on one aspect of
the Mongol conquests, Gabriel Ronay's The Tartar khan's Englishman
(London, 1978). This is an attempt to reconstruct the career of an
anonymous Englishman in Mongol service whom the author believes
himself to have discovered.5 The 'hard' evidence, such. as it is, consists of a
letter preserved in the Chronica majora of Matthew Paris, the thirteenth-
century chronicler of St. Albans. This recounts the capture of the
Englishman during the latter stages of the Mongol invasion of eastern
Europe in 1242, and his subsequent rather meagre confession of his
nefarious activities. It is certainly an intriguing incident, but on to it Ronay
builds an astonishing structure of nonsense: which he is able to do largely,
so far as I can see, because of his apparent total lack of understanding of the
nature of historical evidence.
The reader is given clearly to know what to expect when in its (I fear
implausible) attempt to identify the Englishman, the book commences with
an account of the troubles of King John's reign, culled with extraordinary
gullibility from the chronicle of Roger of Wendover (cf. Professor W.L.
Warren's remarks on Roger's 'garbled inaccuracy and palpable
implausibility', King John, London, 1961, 16). To this are added details of
yet greater improbability from Matthew Paris: Ronay even believes his
story ('there is little reason to question its veracity', p. 28) of John's alleged
plan to take England over to Islam! (On this, see Warren, op. cit., 14–15.
Ronay might perhaps have heeded A.L.Smith's caveat that Matthew Paris's
evidence 'has to be very carefully scrutinized, for it ranges in value from
first-hand, priceless testimony to the most extravagant and worthless
gossip', Church and state in the Middle Ages, Oxford, 1913, 170.)
As for the rest, the book is a work of imaginative fiction dressed up as
history, in which groundless suppositions on one page have mysteriously
become proven facts on the next. Can the author really have intended it to
be taken seriously? A few further details must serve to characterize the
book and the nature of its author's scholarship:
p. 2: 'The freshly turned-up documents…make fascinating reading, even in
the cumbrous medieval script'. The principal document is Paris's chronicle,
and in particular the letter of Yvo of Narbonne printed as an appendix to
Ronay's book. The 'cumbrous medieval script' he has in mind here, though
unacknowledged, turns out to be J.A.Giles's English translation of Paris,
published in 1852–4. (Compare Ronay, pp. 228–34 with Giles, vol. I, pp.
467–73. All translations, says Ronay,' appear here, unless otherwise stated,
in the author's translations'.)
p. 113: a most eccentric derivation is confidently offered for 'Genghis'—a
name the meaning of which continues to puzzle Mongolists. Ronay's
(unacknowledged) source seems to be Gibbon (Decline and fall of the
Roman
empire, ed. Bury, VII, p. 3, n. 4)—perhaps not the most up to date or
reliable guide to questions of Mongolian etymology. p. 136: 'In this first
ever translation into English of this important document (Güyük's letter to
Pope Innocent IV) I used the Persian, French and German versions of the
missing original text'. This is both misleading and ignorant. The French and
German versions (presumably those of Pelliot and Spuler) are translations
of the Persian, which is the only 'original' text we have. English translations
of this have appeared in C.Dawson (ed.), The Mongol mission, 1955; I.de
Rachewiltz, Papal envoys to the great khans, 1971; and B.Spuler (ed.),
History of the Mongols, 1972.
This is merely a small selection from mountains of pretentious foolishness.
The Tartar khan's Englishman is popular historical writing at its worst.
What a relief it is, then, to welcome James Chambers's The Devil's
horsemen (London, 1979), an account of the Mongol invasion of Europe set
against its Asiatic background. This fills very competently a major gap in
the English historiography of the subject. (The only previous substantial
study appeared in German in 1893.) It may be that historians of eastern
Europe will have complaints to make; and on the Islamic side there are a
few slips (e.g. on p. 42 a pre-Wittek view of Ottoman origins—but even
Saunders has this). But this is not a specialist piece of original research, and
should not be criticized as though. it had such pretensions. The author
writes that 'I have simply attempted to tell the story of an extraordinary
campaign, outline its causes and far-reaching consequences and place it in
its historical perspective' (p. xi). In this aim his readable but serious book
has entirely succeeded. But then, unlike Ronay, Chambers has done his
homework; and unlike him, he knows what the writing of history is about.
So far I have been discussing books which fall at one end or the other of the
critical spectrum. They have been either rather good, like Saunders or
Chambers, or worthless, like Ronay. The work to which I must now pay
some more detailed attention, Kwanten's Imperial nomads, disconcertingly
refuses to be so characterized. It is a curate's egg of a book: parts are
distinctly interesting, but other parts are quite surprisingly bad. The book
comes to us, as I have said, in the hope of becoming a standard work. It is
written by an associate professor at a distinguished American university. It
is replete with scholarly apparatus. So one must measure it by a different
yardstick from that appropriate in other cases.
Let me first say something about the merits of Imperial nomads. Its
principal strength is a linguistic one: the author knows Chinese. Most recent
writers on the Mongol Empire have either not known Oriental languages at
all, or, like the late Professor Boyle6 or Professor Spuler,7 have approached
the Mongols through the medium, in the first instance, of the Middle
Eastern sources. Now while I am not persuaded by the author's belief that
Chinese is the most necessary language for the study of the Mongol Empire
—the priority of Persian seems to me undeniable—it is undoubtedly very
useful, in principle, to have a history of the Empire from the pen of a
scholar well-grounded in the Chinese sources. And so the reader (like
myself) who knows his way a little about the Persian sources, but knows no
Chinese, can learn much of interest from Imperial nomads. Whether what
he learns is always accurate is not so
easy to say; as we shall see, the author's treatment of affairs in the western
half of the Empire does not inspire confidence. But still, the use and citation
of Chinese sources is a strength, and those who concern themselves with the
Middle East, or with Mongol-European relations, do sometimes need to be
reminded of the first place accorded to China in the Mongol world view.
This book will at least do that for them.
It is not only China that gets a fair—perhaps slightly more than a fair—
crack of the whip. Much of Kwanten's own research has been on the history
of Tibet: so here, too, interesting information is provided and a better
balance achieved.
But that, I am afraid, is about it. It has to be said that, essentially, Imperial
nomads is both a nanïve and an unreliable book. So far as the naïvety is
concerned, we must look at the author's stated aims. 'The present volume,'
he writes, 'presents a more objective approach to the subject and provides a
new interpretation that rejects pure chance as a causative historical factor.'
'By training, as well as inclination, I belong to the French historiographical
school commonly known as the Annales School' (p. xiv). Now apart from
the rejection of chance—i.e., presumably, a deterministic framework of
some sort is to be imposed on the evidence—it is not easy to see quite what
this Annales-ism consists in, except perhaps (a fair enough point) the
emphasis Kwanten places on a continuous political tradition of a sort on the
steppe. For the rest, however, we are treated to the following sorts of
assertions, none of them exactly shattering as new insights: 'An empire's
history…does not consist solely of politics and conquests…To survive past
its formative period, an empire needs structures of some kind in almost all
aspects of its internal functioning' (p. 187); 'The study of many factors—
political, economic and social among them—thus leads to the interpretation
of an area's development over a fairly long duration' (p. 286); 'Until the
twentieth century…no economic theory was available to the Central Asian
nomads'; 'An examination of the history of the Central Asian steppe
empires, following the methodology of the Annales school, shows that for
several centuries Central Asia exerted a major influence on the historical
development of the sedentary states' (p. 288).
I submit, then, that despite Kwanten's ostentatious wielding of the mighty
name of Annales, this is not in fact a revolutionary book. It is essentially the
mixture as before: most of it is politics and war, with chunks of culture,
trade and administration thrown in. Were it not for the author's professed
allegiance, no one would dream of thinking this book a product of the same
stable as, say, Braudel's The Meditermnean.
Still, this would not in itself invalidate what the book has to tell us. It is the
unreliability of its information that does that. This may be classed under
two heads: gaps in the author's knowledge of certain areas, and his
provision of actual misinformation. A few examples should suffice to
illustrate these.
Kwanten's bibliography and notes are remarkable for the items they do not
contain: Saunders, for example, is conspicuous by his absence. Now in a
wide-ranging book such as this is, references and perhaps even, to some
extent, the author's own reading are bound to be selective. Yet it is odd that
Kwanten's discussion of the reasons for the Mongol withdrawal from
eastern Europe in 1242 (p. 135) contains no reference to Sinor's well-known
and persuasive arguments—the more so since Kwanten acknowledges the
influence and help of Sinor. Likewise his remark (p. 205) that the Mongols
'conquered the Middle East more by chance than by plan', though it might
be right, ought to be qualified by Lattimore's theory of 'the geography of
Chingis Khan'
(Geographical Journal, March 1963). These may be minor points. Other
deficiencies are more serious. Such bald statements as that 'the records of
the Il-khans…barely mention the Yüan' (p. 225); 'the Islamic world did not
have as strong a historiographical tradition as the Sinitic world'; or that 'the
ruling houses and the entire population… [of the Middle East] were Islamic'
(p. 204), can only be founded on, at best, a liking for misleading over-
simplification; but it may be feared that this will not prevent them from
finding their way into a tümen or two of undergraduate essays over the
years.
Inadequate knowledge must presumably be the explanation of the author's
belief in a 'scholarly neglect of the western part of the Mongol empire' (p.
204). He asserts that 'the situation (i.e. the social and administrative
structure) in the Il-khanid domain has not yet been subjected to a serious
critical examination' (p. 191). This gives very short shrift to the labours of
the likes of Aubin, Ayalon, Boyle, Lambton, Petrushevsky, Smith and
Spuler, some of whom appear in the bibliography, but from the study of
whose works Kwanten seems to have benefited little. No one who supposes
Islam to have been 'a single, homogeneous, structured religion' (p. 219), and
that this was the only religion that confronted the Mongols in the Middle
East, can have read very deeply into the subject. Nor can an author who
suggests that 'the historical sources do not deal with the empire as a whole',
and that 'most of the material dealing with the Mongols is written in
Chinese' (p. 225) know much about Juwayni, Rashid al-Din or (the
last-named being notable for his complete absence from the book).
More serious still are the astonishing number of simple errors and
misleading statements of supposed fact: 'Even though agriculture was one
of the prime sources of revenue, the Il-khans, in sharp contrast to the Yüan,
never expressed any interest in it' (p. 213)—so much for Ghazan Khan; the
Mongols' destroyed the town of Balkh so thoroughly that the city of the
same name in modern Afghanistan is not located on or near the ruins of the
old city; in fact, the location of the old city is not precisely known' (p. 120)
—one can only guess that Kwanten has confused the site of Balkh, and the
modern village of that name on the site, with the nearby city of Mazar-i
Sharif;' the second conflict (with the Ong Khan)…occurred shortly after…
the 1206 quriltai' (p. julban,9 by which time the Ong Khan had been dead
for years; and perhaps most serious of all in its possible implications, the
execution of the Ilkhanid wazir Sa'd al-Dawla is dated after, rather than
before, the death of Arghun (p. 163). This is a tiny point in itself, but when
Kwanten's reference is checked, it is found (a) to be to a Persian source,
Rashid al-Din, in the original (Kwanten disclaims knowledge of Persian);
(b) to be to the wrong page; and (c) when the right page is looked up, the
execution indeed turns out to have occurred five days before Arghun's death
—as reference to any standard secondary source would have established.
The question this little incident leaves in the mind, when one is already
ceasing to rely on Kwanten's facts, is how many other errors of this kind
could be tracked down by checking all his references: worse, how many
there may be lurking in the apparently authoritative eastern Asiatic sections
of the book, for the checking of which such readers as myself do not have
the necessary linguistic competence.
One could go on carping. The book is replete with disagreeable jargon; the
bibliography is full of small mistakes (Drake for Darke, Nitti for Hitti, etc.);
transliterations from the Islamic languages, at least, are well described in
Kwanten's own criticism of those in The empire of the steppes as 'totally
arbitrary': Ayn Jalat, Rahbad al-Shams, wasir, Ummayyads, Sa'ud al-Daula,
Jahan-Guzha, Jami'ut al-tawarikh, Sarbardars. But enough is enough.
It all seems a great pity, and a missed opportunity. It would have been a
pleasure to be able to have said that here, at last, is an accurate,
authoritative history of Central Asia which. one can recommend without
reservation. But the most that can in fact be said is that Imperial nomads
will repay attention for its eastern Asiatic material and for some useful
emphases, mixed in with a mass of misinformation and unsubstantiated
judgements. For the broad sweep of Central Asian history, the reader will
still have to go first to Grousset, or to Gavin Hambly's Central Asia; and for
the Mongols, to Saunders.
This survey inevitably provokes one or two general reflections on the
writing of the history of such subjects as the Mongol Empire. First, and
perhaps most banal, what a prodigious amount of work remains to be done.
Such is the orientalist's customary ritual incantation against the writing of
works of synthesis—it is premature: we must wait for the monographs. This
is not the view I would wish to put forward. It does seem to me that there is
sufficient material, either sources in translation or secondary literature of
worth, to make some kind of synthesis perfectly feasible—something that
has not been true until comparatively recently. Saunders, more than anyone
else I have discussed here, proved that this is so: he knew his limitations,
and those of the material available, and made no exaggerated claims.
Similarly both. Brent and Chambers show that good popular history can be
written about the Mongols.
It is Kwanten's book, though, that casts the problems into the sharpest relief.
Here is an attempt, albeit an unsuccessful as well as a pretentious one, to
write a general survey to the highest scholarly standards; not for Kwanten
the modest aims of a Saunders or a Chambers. Why does it fail? In part the
faults were avoidable: these, I hope, have been sufficiently indicated. But
there are two difficulties which face anyone who contemplates a task
similar to Kwanten's. One is the problem of languages. As Saunders put it,
'in bulk, the original sources are not unmanageable, but they are extant in so
many languages that only a linguistic prodigy could claim a mastery of
them all'.8 Pelliot was such a prodigy, and there are said still to be one or
two, though not, so far as I know, on this side of the Atlantic. Most of those
who elect to study the Mongol Empire must choose their end of Asia, west
or east, and learn the languages accordingly. For the other end, we depend
on translations: and for myself, the mistranslations I have stumbled on
when looking at published versions of Persian texts make me wonder into
what traps I may be falling when I rely on translations from Chinese and
Mongolian.
Still, one can usually call on specialist assistance. What is less easy is to
ensure the soundness of one's judgement when dealing with the end of Asia
and of the Mongol Empire with which one is not so familiar. Kwanten
shows only too clearly that a knowledge (which I presume him to have) of
how eastern Asiatic government and society worked is no guarantee of any
sureness of touch when dealing with the Middle East. And as any orientalist
will tell you, there is no short cut! I can only, then, express the hope that
historians will not allow these difficulties to deter them from writing about
the Mongol Empire: in particular, a new study of the life and career of
Chingiz Khan himself is an obvious need. The Tartar khan's Englishman
and Imperial nomads can at least serve, in their different ways, as object
lessons in how not to go about it.
 SALADIN AND HIS ADMIRERS: A BIOGRAPHICAL
REASSESSMENT1

By P.M.HOLT

'The life and achievements of Saladin constitute one of the great moments in the history of the
crusades. In literature he appears most frequently as a conquering hero, who fought his enemies
victoriously and in the end beat them to a standstill. But a closer examination of his actual life reveals
him not only as a conqueror, but as a man who struggled with enemies of his own side who finally
joined him and fought along with him under his sole command. From this angle we see him as a man
who fought for his ideals, and fought, not victoriously, but in a measure that fell short of his hopes
and ambitions.'2
These opening words of Sir Hamilton Gibb's biography of Saladin are the
latest, and perhaps the last, expression of a tradition in the European
historiography of the Crusades which has been influential at least since the
publication in 1898 of Stanley Lane-Poole's Saladin and the fall of the
kingdom of Jerusalem. The origin of this tradition may indeed go even
further back to Sir Walter Scott's Talisman, which Gibb himself, we are
told, gave 'to students as a work of art from which they could learn much
about Islamic history.'3
This view of Saladin is, however, confronted with the difficulty that from
the beginning of his independent reign in 570/1174 until his death. in
589/1193, Saladin spent twelve years mainly in fighting the Zangids, the
family and partisans of his former lord, Nur al-Din, and five years only in
the Holy War against the Latin kingdom and the Third Crusade. Any
judgement of his career and character must depend largely on the view
taken of those first twelve years. Gibb saw Saladin as inspired throughout
by the resolve to wage the Holy War against the Franks, and indeed as
having wider aims, 'to restore and revive the political fabric of Islam as a
single united empire, not under his own rule, but by restoring the rule of the
revealed law under the direction of the Abbasid Caliphate.'4 This
teleological interpretation of Saladin's career rests upon contemporary
Arabic sources, in which already by the time of his death a legend of his
achievements was coming into existence. Since these sources are only
briefly discussed in the work under review (pp. 2–3), a fuller examination is
perhaps justifiable.
Saladin's two contemporary biographers were Baha' al-Din Ibn Shaddad
and 'lmad al-Din Ibn Shaddad's work, al-Nawadir
al-Yusufiyya, was published in Europe with a Latin translation by
the Dutch orientalist, Albert Schultens, as early as 1732. It was thus a prime
influence in forming the view of Saladin held by European historians. It
was used by Gibbon, who calls the author Bohadin, and eighty years later,
when William Stubbs wrote his introduction to Itinerarium Regis Ricardi,
Bohadin was still the principal, and almost the sole, Arabic source available
to him. wrote a seven-volume chronicle of Saladin's life and
times, entitled
al-Barq al-Shami. Only two complete volumes are now extant, but excerpts
from the others survive, and an abridgement by al-Bundari, Sana al-barq
al-Shami, has recently been published in full. He also wrote an account of
the last glorious phase of Saladin's career from 583/1187 to his death,
entitled al-Qussi al-Qudsi.
These may appear to be impeccable sources, since both men belonged to the
inner circle of Saladin's court officials, Ibn Shaddad having been his army-
judge and his secretary. Such in fact was Gibb's view. But some
caution is necessary. These after all were royal biographies, written by
courtiers about their master was certainly completed within a few
months of Saladin's death, and D.S.Richards has shown that much of it was
'already available in some form during Saladin's lifetime' as 'the work was
probably intended for presentation to Saladin.'5 Al-Nawadir, which Richards
has shown to be partly dependent on was perhaps completed as early
as 1198, and certainly not later than 1216, i.e. at a time when Saladin's
Ayyubid kinsmen were the unquestioned rulers of Egypt and Muslim Syria.
Writing on medieval European historiography, Beryl Smalley has remarked
that,
Royal biographies have one feature in common: they are propaganda pieces. The writers' purposes
and techniques varied, but they had all to find a mould which would contain the unruly facts. The
prince had to be presented as his biographer wished to show him to his readers or hearers.6
This was also true of Ibn Shaddad and al-Isfahani. There is a further
consideration. Both of them had gone over to Saladin's service from the
Zangids. who had been Nur al-Din's secretary, had joined Saladin
in the year after his master's death. Ibn Shaddad's home was at Mosul, and
he negotiated with Saladin on behalf of its Zangid ruler in 581/1186. It was
not until two years later that he entered Saladin's service. Both men may
have sought to justify their shift of allegiance by exalting Saladin's
character and achievements. One must also bear in mind that Ibn Shaddad
could write from personal knowledge only of the last years when Saladin
was engaged in the critical struggle with the Franks.
About the same time as these works by Saladin's partisans, there appeared a
work by the historian Ibn al-Athir (555–630/1160–1233), who spent most
of his life in Mosul, and who gives the Zangid view of events. His dynastic
history of the Zangids, al-Bahir fi ta'rikh atabakat completed in
607/1211, appears to be a counterblast to the writings of Saladin's admirers.
Although the title refers specifically to the atabegs of Mosul, Ibn al-Athir's
patrons, much space is devoted to Nur al-Din, who ruled only in Syria but
was undoubtedly the most distinguished member of the family. In his
introductory remarks, Ibn al-Athir emphasizes the role of the Zangids as
champions in the Holy War. His long and detailed eulogy of Nur al-Din
appears in some respects to retort to statements made by Ibn Shaddad about
Saladin.
Thus within about twenty years of Saladin's death, there came into
existence two rival presentations of the twelfth-century rulers of Muslim
Syria, glorifying the exploits of the Zangids and Saladin respectively. These
two historiographies were reconciled by Abu Shama (559–665/1203–68),
who passed almost his whole life in Damascus. The title of his work, Kitab
fi akhbar al-dawlatayn al-Nuriyya ('The book of the
two gardens con-
cerning the two regimes of Nur al-Din and Saladin'), announces its eirenic
purpose. It was completed in 651/1253 in changed circumstances from
those which the earlier writers had known. The Ayyubids had lost Egypt to
the Mamluks. St. Louis, defeated in Egypt, was at Acre, restoring the
defences of the Latin kingdom. To the Mamluks and to their opponent, the
Ayyubid ruler of Aleppo and Damascus, he was a desirable ally, and both
offered him the retrocession of the territories taken by Saladin. Although
the negotiations ended inconclusively, it must have seemed to Abu Shama,
a pious scholar, unpractised in politics, as if Saladin's work was about to be
undone by his successors. It is not surprising that he drew from history a
lesson for the rulers of his own time, which he sought to convey in this
combined history of Nur al-Din and Saladin.
In this way, sixty years after Saladin's death, the creation of his legend was
completed. Saladin, the usurper of the patrimony of the Zangids, is shown
as their successor in a divinely appointed mission. The contests for
supremacy in Syria between Muslim and Muslim, as well as between
Muslim and Frank, are seen in retrospect, as if the reconquest of Jerusalem
at the end had been Saladin's intention from the beginning. It is an
interpretation of twelfth-century Syrian history which has dominated later
writing, but in essence it is Heilsgeschichte, the salvation of the Islamic
community in the latter days as the Hijra of the Prophet saved it at its
beginning, as implies in 7
A reassessment of Saladin and his achievements by Andrew S.Ehrenkreutz
appeared almost simultaneously with Gibb's biography. In his introduction,
Ehrenkreutz reviewed previous presentations of Saladin, and in the body of
the work he drew on his specialist knowledge of the economic history of the
period to draw attention to aspects of the reign which the political historians
had tended to ignore. His final judgements is severe:
Most of Saladin's significant historical accomplishments should be attributed to his military and
governmental experience, to his ruthless persecution and execution of political opponents and
dissenters, to his vindictive belligerence and calculated opportunism, and to his readiness to
compromise religious ideals to political expediency.
…Rather than the alleged attractiveness of his romantic personality, it was the potent spell of his
tendentious biographers which has clouded the perceptions of most modern writers retelling the story
of the great sultan.8
The revision of the historical role of Saladin and the reassessment of his
personal qualities is carried a stage further by the work under review, which
surveys the whole of his life in the context of his period. To quote the
Foreword:
The object of this work is to re-examine and, where possible, to add to evidence for the career of
Saladin in order to strengthen the frame of reference into which the judgements and conclusions of
his modern biographers can be fitted.
It is an important biography in two respects. In the first place, it gives a
very detailed narrative of Saladin's career, which is a most useful
assemblage of political and military data. In the second place, it draws on
two sources which
have not been exploited by previous writers. One of these is al-Bundari's
abridgement of al-Barq al-Shami, mentioned above. The published edition
by al-Nabarawi9 was not available to the authors, who used her
Cambridge Ph.D. thesis and a partial edition by published in 1971.
The second new source, which. has played a large part in forming the
author's view of Saladin and his policies, is the corpus of letters from
Saladin's court, which to some extent supply the lack of a true archive. They
are thus described:
In the main, these are attributed to Saladin's administrator, the and they comprise both
personal letters sent by himself and others drafted for Saladin. Some are quoted by the
narrative historians or are found in other works; twenty-six are included, complete or in part, in a
Cairo edition, but a large number are still unedited. This collection is supplemented by a manuscript
of letters wrongly attributed to 'Imad al-Din and by the writings of another of Saladin's
contemporaries, the North African al-Wahrani. The scope of their material is, of course, limited and
they cannot compensate entirely for the dearth of official documents, but in addition to the details
that they provide, many show the construction that Saladin himself wished to have placed on his
actions, while others supply this with an unofficial commentary.10
The authors have also drawn on Western and Byzantine sources for the
period, e.g. William of Tyre and Nicetas Choniates. Previous writers,
approaching the subject from an orientalist background, have tended to
neglect this valuable range of materials. One may note in passing that the
significant study by Hannes Möhring, Saladin und der Dritte Kreuzzug,
Wiesbaden, 1980, which draws extensively on medieval European,
Byzantine and Arabic sources, although listed in the bibliography, was
published after the completion of this work.
The Saladin who emerges from these pages is no longer the confident and
dedicated champion of Islam, even if he is not the anti-hero, disastrous to
Egypt, depicted by Ehrenkreutz. The precarious nature of his position
appears constantly. The jihad was a means of legitimating his authority (cf.
pp. 88–9, 97), as also was his marriage to al-Din Khatun (p. 110). As
the daughter of Mu'in al-Din Önör and widow of Nur al-Din, she provided
him with a double personal link with the rulers who preceded him.
Ironically, at the time of this marriage in 1176, Isma'il, her son by
Nur al-Din, still reigned in Aleppo, and was the figurehead of Zangid
resistance to Saladin's usurpation. An interesting paragraph (pp. 152–3)
stimulates reflection as to the nature of Saladin's rule—'the question of
whether Saladin was, or had become, primarily a war-band leader or
whether he should be thought of as a territorial ruler.' His situation was not
unlike that of the Norman and early Angevin monarchs, ruling two distinct
territories with a 'peripatetic administrative nucleus.' His relation to the
administration is summarized (pp. 366–7) as 'an inherited bureaucracy
within whose framework operated a system of patronage with Saladin at its
head. It was patronage, rather than formal administration, that appears to
have occupied his own time'. The ambiguity of his policy as the self-
appointed champion of Islam is brought out at various points in his career:
thus in 1182, after the death of Isma'il:
The policy of Ayyubid expansionism that had been blocked by the peace
treaty of 1176 was about to be renewed. Saladin was laying claim not only to Aleppo, but to any
other town whose troops could be shown to be needed for the Holy War. This could not be accepted
either by Izz al-Din in Mosul or by Zangi in Aleppo and Saladin's sincerity in turning his back on the
Franks to fight his fellow-Muslims was bound to be called in question.11
The nemesis of Saladin's policy is thus characterized:
As his letters show, however, he was finding that the logical end to the cycle of expansion, where
power depended on conquests, attracting recruits to be paid for by further conquests, was a power
monopoly coterminous with the frontiers of Islam.12
Another passage, referring to the situation in 1188, but applicable to the
whole of Saladin's reign, examines another reason for his precarious
position:
A factor that had to be taken into account was the loose structure of his army. His allies had no
reason to give him whole-hearted support. For his own emirs and professional soldiers he and his
family were merely successful members of their own class; his dynasty was bolstered by no divine
right of kings and the religious sanction it had claimed had been denied it by Baghdad. During the
period of its expansion it had been profitable to join his side, but profit and numbers were
inextricably linked. If his military accounts began to show a loss, his numbers could be expected to
diminish and his dynasty in its turn could be threatened by other Muslim expansionists.13
The authors have provided a detailed, considered and perceptive account of
Saladin, setting his acts and policies firmly in their proper context—the
complex and unstable political and military condition of the Near East in
the late twelfth century, a condition in which the Frankish states were one,
but not the sole, factor. The amount of detail and the numerous characters
who make their appearance do not conduce to an easy narrative, and in one
or two respects the authors might have helped the reader more. Although
the month and even the day in which events occurred are usually given, it is
often necessary to range over quite a number of pages to ascertain the year.
A running date-heading (as in Lane-Poole's book) may be old-fashioned but
it is of much assistance in following a full chronological narrative, and still
more for purposes of reference. The provision of the Hijri equivalents to
Christian dates would have facilitated reference to Arabic sources. On the
other hand, to give the equivalent in kilometres to every distance in miles is
hardly necessary. The present reviewer finds the system of abbreviations
used in the notes bewildering and tiresome, and it is curious that the
bibliography, while separately listing Arabic primary sources does not
distinguish original Western (and Byzantine) sources from modern works.
The eight maps provided are useful, although the Dongola shown on Map 6
is New Dongola which developed in the nineteenth century, not the
medieval town (Dunqula al-'Ajuz), which lies about ninety miles further up
the Nile and on the right bank. Plans of the site of the battle of and of
medieval Jerusalem and Acre would have been helpful. These, however, are
criticisms of ancillary details, and in no way detract from the value and
importance of this long awaited work.
 SOME OBSERVATIONS ON THE 'ABBASID CALIPHATE
OF CAIRO

By P.M.HOLT

The re-establishment of the 'Abbasid caliphate in Cairo after its overthrow

on the fall of Baghdad to the Mongols in 656/1258 was, as is well known,
accomplished by the Mamluk sultan, Baybars, who installed
successively two refugee 'Abbasids as caliphs in his capital. What is
perhaps not always realized is that the two pretenders did not appear in turn
and out of the blue at the beginning of Baybars's reign, and that the sultan's
reasons for installing the first, were not identical with his reasons
for installing the second, In certain respects the official account by
Baybars's court biographer, al-Din Ibn 'Abd needs to be
supplemented and corrected. 1

the future had had the more interesting career

2
before becoming caliph. His genealogical claim to the caliphate was not
strong: he was a descendant in the fourth generation of the Caliph al-
Mustarshid (512–29/1118–35). When Baghdad fell, he absconded, and
sought refuge for a time with the tribe of Khafaja. Subsequently he made
his way to Syria, where he enjoyed the protection of 'Isa b. Muhanna, the
powerful chief of Al Relations with Yusuf, the Ayyubid lord of
Aleppo and Damascus, came to nothing with Hülegü's invasion of Syria,
but after the Mongol defeat at 'Ayn Jalut (25 658/3 September
1260), ' Isa b. Muhanna brought him to the notice of the
Mamluk sultan, who promised to restore him, and apparently performed the
bay'a to him by deputy. But the sultanate quickly passed to Baybars by
regicide and usurpation, and the pretender's hopes were again disappointed.
Baybars installed as caliph the alternative candidate, the
son, brother and uncle respectively of the last three caliphs of Baghdad.3
He, with other 'Abbasids who had been held in custody by was
set free by the Mongols. He too sought refuge among the Arab tribesmen,
but appears to have done nothing in particular until Baybars heard of him,
welcomed him to Cairo, and made him caliph.
 finding himself thus anticipated, fled to Aleppo, where a
Mamluk war-lord, Aqush al-Barli, was trying to carve out a kingdom for
himself. Al-Barli performed the bay'a to him, and gave him a force of
Turcomans, with which he set out for and the Euphrates. He had
already made a successful raid in this region with the Arabs of Al in the
previous year. However, at 'Ana he met his kinsman and rival, the Caliph
who had been despatched by Baybars with an expeditionary
force against Baghdad. The two continued their advance until in a battle
with the Mongols in 660/November 1261, disappeared
from history. survived to make his way to Cairo and the caliphate.
What were Baybars's reasons for recognizing as caliph? One was
undoubtedly a desire to legitimate his power. He was not only personally a
usurper, having obtained the throne by the killing of his predecessor
but the Mamluk sultanate itself had supplanted the Ayyubids. Ibn 'Abd
biography bears witness to Baybars's sensitivity in this respect,
since it is careful to present him in a threefold guise as the true successor to
the Ayyubids, the Heerkönig of his Turkish Mamluk comrades, and the
paragon of Muslim rulers. The desire for legitimacy may account for the
speed with which the restoration of the caliphate was carried out. Baybars
received in Cairo on 9 Rajab 659/9 June 1261. Four days later,
his credentials were publicly attested; Baybars performed the bay'a to him,
and in return was invested with the universal sultanate. On 17 Rajab/17
June, the new caliph pronounced the in the Citadel mosque, and on 4
Sha'ban/4 July there was an impressive public ceremony when the sultan
was robed in the black livery of the 'Abbasids and a diploma conferring
plenary power on him was read. In the following month, another 'Abbasid
court usage was revived when the caliph invested Baybars with the futuwwa
in the presence of witnesses.
A further purpose may, however, be seen in the installation of as
caliph. His arrival in Cairo occurred at a time when although the Mongol
conquests had been halted, the frontier between Mamluk Syria and the
Ilkhanate was not yet stabilized. Two Mongol invasions of Syria had been
defeated, at 'Ayn Jalut and again near in 659/December 1260.
There was hostility, which was to culminate in warfare, between Hülegü
and Berke Khan of the Golden Horde. In the circumstances, an offensive
against the Ilkkhanate seemed feasible. Baybars had agents ready to his
hand in the three sons of Badr al-Din Lu'lu', the late atabeg of Mosul, who
had died as Hülegü's vassal. Isma'il, who hoped to obtain Mosul
itself, came in to Baybars in Sha'ban 659/July 1261; his two brothers were
the claimants respectively to the Jazira and Sinjar. The caliph with his
appeal to the tribal Arabs of badiyat al-Sham usefully complemented this
arsenal of potential client-rulers in al-'Iraq, and the four men set out on their
doomed enterprise with the consequences for which have been
described. So ended the first 'Abbasid caliphate of Cairo. It had lasted less
than six months. Baybars, perhaps in self-justification, boasted to Ibn 'Abd
that he had spent no less than 1,060,000 dinars on the caliph and the
princes of Mosul. Ibn 'Abd nephew, Shafi' b.'Ali, writing with the
hindsight of a later generation, expresses his surprise that Baybars should
believe that the numerous and victorious Mongol armies could be
effectively opposed by so small a force.4 But there is no need to suspect bad
faith in Baybars: the political situation in 659/1261 made the enterprise
seem by no means a forlorn hope.
The circumstances of elevation to the caliphate differed
considerably from those which had attended Having escaped
from the battle which saw the end of his kinsman, he again took refuge with
'Isa b. Muhanna. He got in touch with Baybars, and made his way to Cairo,
where he arrived in Rabi' II 660/March 1262. Unlike his predecessor, he
was not hurried into public recognition as caliph; his installation did not
take place until 2 661/16 November 1262. The reasons for this
delay may have been partly personal, partly due to changes in the political
situation. Personally, was clearly a more considerable figure than
He had already twice been recognized as caliph—but by two of
Baybars's defeated rivals, and al-Barli, and was thus to some extent a
potential focus of opposition. Baybars's haste to install may
indeed have been due partly to a desire to anticipate the arrival of this
second and less attractive claimant had gained military experience
and some success in his earlier adventures, and he had a powerful friend
who could provide him with a fighting force in 'Isa b.Muhanna. As far as
the political situation was concerned, Baybars had no further need of
caliphal legitimation, and it is noticeable that he received no new diploma
promulgated in name. Only on the day after his installation did the
caliph pronounce a praising the sultan, and urging the people to
obedience and the Holy War. Furthermore, it was now clear that there could
be no hope of establishing a client-caliph in Baghdad. Mosul had fallen to
the Mongols, and Isma'il had been put to death in the summer of
1262. There was thus no military role for to play, and from
Baybars's point of view it was undesirable that he should enjoy much
publicity. Whereas had been furnished with a caliphal household
and a private army, was given a residence in a tower of the Citadel,
and provided with tutors to improve his religious education.
In one important matter, however, was required to fulfil a political
function, although as a puppet only. The rift between the two western
Mongol rulers, Hülegü and Berke, offered considerable advantages to
Baybars, who by establishing an alliance with. Berke could ease the
pressure on his Syrian frontier. Furthermore, since the territory of the
Golden Horde was at this time the principal recruiting-ground for Mamluks,
an understanding with its ruler was greatly in Baybars's interest. He was not
slow to open diplomatic relations with Berke. In 660/1261–2, he had
written to the khan, urging him to the Holy War against the infidel Hülegü,
who for the sake of his Christian wife was favouring the Christians. Then in
661/November–December 1262, he sent ambassadors to Berke,
from whom in turn he received an embassy in Rajab 661/May 1263. The
diplomatic situation was nevertheless one of some delicacy. Berke was after
all a descendant of Chingiz Khan in the eldest line; Baybars, however
extensive his dominions and great his power, was a self-made ruler of
unknown parentage. It was here that the caliph, the Prophet's cousin, could
usefully serve as a mouthpiece in communicating with the convert to Islam.
The formal entente was conveniently supplemented by the brotherhood of
the futuwwa, here playing a part very like that of the orders of chivalry in
the West. Baybars had been initiated by as has been mentioned.
On 3 661/11 July 1263, through the agency of one of his magnates,
he invested who on the following night, and by the same agent (not
a member of his own entourage) invested Berke's ambassadors. On the
previous Friday, they had heard the caliph pronounce in the the name
of their master after that of Baybars. They had an audience of the caliph, in
which he conveyed appropriate sentiments to Berke, and urged him to the
Holy War. It was the start of a collaboration which survived Berke's death
and the succession of a non-Muslim in 664/1266.5
reigned (but did not rule) for forty years until his death in
701/1301, and was the progenitor of a dynasty of caliphs which finally
vanished only after the Ottoman conquest of Egypt.6 The unbroken
succession, as well as the long reigns of several individual caliphs, indicates
that they stood apart from the turbulent factional politics of the Mamluk
sultanate, and passed their days in retired obscurity and impotence. The
caliph's principal functions were ceremonial, and in one respect at least his
role grew in importance some decades after the restoration of the caliphate,
namely in legitimating by his presence, and sometimes by his word and act,
the accession of a new sultan. The necessity for such symbolic and public
legitimation seems to have been a consequence of the installation of sultans
by rival factions of magnates from the late seventh/ thirteenth century
onwards.
Perhaps the first instance of the caliph's presence at the accession of a sultan
occurred in 698/1299, when was brought back from exile
and reinstated as nominal sultan—he was then about 14 years of age.7 On
this occasion, the sultan was enthroned in the presence of the Caliph
and the chief judges, who do not seem, however, to have played an
active part in the proceedings. Ten years later, when was
again in exile, and the throne was taken by Baybars, the usurper
sought to secure his position (which he had accepted with reluctance) by
obtaining a diploma from the Caliph al-Mustakfi I, conferring plenary
powers on him. When after the death of at the end of his
third reign in 741/1341, twelve of his descendants were set up and cast
down by Mamluk factions in the space of forty years, the participation of
the caliph. in the installation of the sultan is almost invariably noted by the
chroniclers. On some occasions at least an elaborate ceremony was staged,
as on the accession of Abu Bakr, son and first
successor. 8

This nominal sanction by the caliph of the newly installed sultan did
nothing in reality to secure the latter's position or that of the faction which
sponsored him. Abu Bakr reigned less than two months before he
was deposed. The usage, however, continued, not only under the later
Qalawunids but also during the Circassian succession in the ninth/fifteenth
century, when the sultanate passed by usurpation to a series of war-lords.
The accession of the last Mamluk sultan, al-Ashraf in 922/1516
came at a time when the traditional observances were particularly hard to
carry out. The reigning caliph, al-Mutawakkil III, had accompanied
al-Ghawri to Syria, and was a prisoner in the hands of the
Ottomans. Fortunately he had left as his deputy in Cairo his father, al-
Mustamsik, who had abdicated in 914/1508. The delegation of powers to
the deputy caliph was attested, and for the last time an 'Abbasid formally
confirmed the authority of a Mamluk sultan. The manner in which this was
done was a significant indication of the change over two and a half
centuries in the respective positions of caliph and sultan. When
was installed in 659/1261, Baybars performed the bay'a to him as the head
of the Muslim community. In 922/1516 by contrast, as on some (perhaps
all) previous occasions since at least the accession of in
742/1342, the roles were reversed, and the caliph performed the bay'a to the
sultan.9 Khalil writing in the reign of Jaqmaq (842–57/1438–
53), propounded a new constitutional theory in the light of these
developments. In his account of the sultanate, he writes as follows:
The Prophet (the blessing of God be upon him and peace) was truly the controller of the world; then
the caliphate was transferred to the Imam Abu Bakr then the Companions and the caliphs
(may God be pleased with. them all) inherited it one after another, until it is now effected by the
mubaya'a of the Commander of the Faithful by the agreement of the holders of power and the 'ulama'
and the pillars of the August State, and the assent of their lordships the amirs and the divinely-aided
armies [lacuna].10
Elsewhere Khalil bears witness to the degradation of the caliph from
his once unique and supreme position in the Umma. In his chapter on the
caliphate, he begins by deseribing the caliph and his functions in traditional
terms as God's representative on earth, the divinely-appointed ruler, whose
commission alone creates a legitimate sultan. Then in his last few lines on
the subject, he descends to the practical realities of the contemporary
caliphate: His appointment is to concern himself with scholarship and to have a library. If the
sultan travels on some business, he is to accompany him for the benefit of the Muslims. He has
numerous sources of revenue for his expenses, and fine dwellings.11
At best the caliph was a nominal head of the Religious Institution, but
without any jurisdiction. Towards the end of the Mamluk sultanate, he is
most frequently mentioned as accompanying the four chief judges when
they went up to the Citadel at each new moon to congratulate the sultan.
The few interventions of the caliph in Mamluk politics were usually both
involuntary and unprofitable. One such incident was in connexion with. the
usurpation of Baybars, referred to earlier. Not only did the sultan
obtain a diploma on his enthronement in 708/1309, as we have seen, but a
few months later, when the revolt which was to restore was
making headway, al-Mustakfi I furnished Baybars with a second
instrument, confirming his authority as against the hereditary claim of his
opponent, the son of Qalawun. The essential passage is: I have shown you my
pleasure with the slave of God Most High, al-Malik Rukn al-Din, as my deputy in the
kingship of the Egyptian territories and the Syrian lands. I have set him in place of myself because of
his religion, his competence, his ability and his favour to the Muslims. I deposed his predecessor
after learning of his abdication from the kingship. I deemed this to be my function, and the four
judges gave their judgments therein. And know (may God have mercy upon you) that kingship is
childless [al-mulk 'aqim], and does not pass by inheritance to anyone from predecessor to successor
or in order of seniority. I have besought the choice of God Most High, and appointed al-Malik
as governor over you.
Whosoever obeys him, obeys me; and whosoever disobeys him, disobeys me; and whosoever
disobeys me, disobeys Abu'l-Qasim my cousin (the blessing of God be upon him and peace).12
When Baybars sent a report of the promulgation of this diploma to his son-
in-law, Burulghi al-Ashrafi, whose troops were deserting en masse to
the amir's comment to the messenger was disconcertingly realistic: 'Stupid
fellow! For God's sake—nobody takes any notice of the caliph!'
Unfortunately for al-Mustakfi, notice of him was taken by When the sultan
was enthroned for the third time on 2 Shawwal 709/5 March 1310, the
caliph attended in accordance with precedent, and approached to greet him.
The sultan addressed him with heavy irony: 'Why are you here to greet a
rebel? Was I a rebel? Is Baybars a scion of the 'Abbasids?' The caliph's
countenance fell, and he spoke not a word.13 The consequences of his
support of Baybars were indeed unpropitious for him: he never regained the
sultan's favour, was twice incarcerated, and died an exile in in 740/1339.
In the following century, the caliphate and the sultanate were briefly
combined in the person of al-Musta'in, the great-grandson of al-Mustakfi I.
It was an episode which demonstrated the political impotence and
insignificance of the caliph, who served as a puppet in the hands of the
Mamluk magnates.14 The occasion was the failure of Barquq's attempt to
establish an hereditary sultanate. His son, Faraj, succeeded him in
801/1399 at the age of ten, was deposed in 808/1405, restored a few weeks
later, and finally clashed with the great Syrian amirs headed by Shaykh and
Nawruz An expedition which he led against them was defeated on 13
815/25 April 1412, and the caliph, who had accompanied him, fell into the
hands of the rebels. They decided to set up al-Musta'in as sultan to enable
them to overthrow Faraj, who was besieged in Damascus. Al-Musta'in, who
was both fearful and reluctant, was tricked into accepting the sultanate.
Faraj surrendered on 11 May. He was tried by a commission of amirs,
jurists and 'ulama', and sentenced to death. There was deep division over
the propriety of this, since Faraj had surrendered on terms, but the caliph-
sultan had used his influence with the judges and jurists against his deposed
predecessor.
On 2 Rabi' 11/12 July, al-Musta'in entered Cairo in the company of Shaykh,
who henceforth dominated him. The caliph-sultan's attempts to acquire real
power and even public recognition were frustrated. On 8 Rabi 11/18 July,
he invested Shaykh as atabak al-'asakir in Egypt, and formally conferred
plenary powers on him. Shaykh was in fact determined to obtain the
sultanate. His chief opponent, Nawruz, was safely absent as governor-
general of Syria. Al-Musta'in's situation is pathetically described by Ibn
Taghribirdi: The caliph became homesick for his kinsmen in the vast palaces of the Citadel; he
was uneasy at the lack of visitors. In vain he regretted the position into which he had entered. To
speak of his regret would not bring amirs or anyone else to his aid, so he kept silence about his
distress.15
He was not to occupy the position for much longer. On 16 Jumada 1/24
August, his diploma conferring plenary powers on Shaykh was published,
and the atabak began to behave like a sultan. Three weeks later, on 8
Jumada II/ 15 September, the last obstacle was removed from Shaykh's path
when the death of his only serious rival in Cairo, Baktamur Jilliq, enabled
him to canvass the support of the other amirs for his own accession. He was
installed by acclamation on 1 Sha'ban/6 November, and al-Musta'in
resigned the sultanate as he had accepted it six months before—reluctantly
and under compulsion. His remaining career was one of humiliation. In
816/1414, he was deposed from the caliphate in favour of his brother, II.
He was held in the Citadel until 819/1417, when he was sent as a prisoner
(with the three sons of Faraj) to Alexandria. There he died of plague in
833/1430. He was less than 40 years old.
The legend that the Caliph al-Mutawakkil III formally transferred the
caliphate to Selim the Grim on the overthrow of the Mamluk sultanate has
long been recognized as a fabrication. One small scrap of contemporary
evidence of its fictional nature has not perhaps been generally noted. The
Syrian chronicler, Ibn (880–953/1473–1546) introduces each year in his
annals Mufakahat al-khillan with the name of the reigning caliph, and he
continues this practice over the period of the Ottoman conquest down to
926/1519–20, where the fragment breaks off. At the opening of A.H. 924,
Ibn after mentioning 'the Caliph, Commander of the Faithful, al-
Mutawakkil 'ala'llah', adds 'and he has been sent under escort from Egypt to
Istanbul by sea'. In A.H. 926, he adds 'and he is dwelling in Istanbul'.16
Thus not only is there no mention of a transfer of the caliphate, but also Ibn
(and doubtless others) saw no break in al-Mutawakkil's reign when power
passed from to Selim.
Nevertheless with the caliph a state prisoner in Istanbul, the 'Abbasids
quickly passed into obscurity. They left, however, one or two traces in later
history. In his necrology for 1220/1805–6, al-Jabarti notices perhaps the last
notable descendant of the caliphs, 'Uthman Efendi b. Sa'd al-'Abbasi who
had had a lucrative career as a financial official in Egypt.17 Less historically
verifiable is the traditional claim of the Ja'aliyyun, one of the great tribal
groups in the northern Sudan, to be 'Abbasids.18 The claim may be no more
than an attempt to acquire an illustrious ancestry, like that which linked the
royal house of Wessex with the biblical patriarchs through Sceaf, son of
Noah, who was born in the Ark. Or it may have had political implications,
since the Funj overlords of the Ja'aliyyun in the seventeenth and eighteenth
centuries claimed to be Umayyads. Yet the genealogy may preserve a
remote historical memory of the coming to Nubia after the Ottoman
conquest of Egypt of refugees from the caliph's entourage.
 THE 'GREAT YASA OF CHINGIZ KHAN' AND MONGOL
LAW IN THE ILKHANATE

By D.O.MORGAN

One of the odder features of the Persian sources on the history of the
Mongol period is the vagueness and comparative rarity of references to the
'Great Yasa of Chingiz Khan'. This struck me with renewed force after
reading Professor David Ayalon's articles on the Yasa in Studia Islamica.1
My suspicions about the whole matter having thus been aroused, it seemed
to me that it might be an interesting exercise to look again at the origin and
nature of the Yasa before trying to estimate how Mongol law worked in the
Ilkhanate. And so, as I hope to show in this paper, it proved.
 I

Let me first of all outline the view that is usually taken of the Yasa. At some
time during his reign, and probably at the quriltai of 1206, Chingiz Khan
promulgated a code of laws which were to be binding on his people and
their descendants for ever. This was a codification of the ancestral
traditions, customs, laws and ideas of the Mongols, to which Chingiz Khan
added further laws of his own devising. Copies of this great code, the Yasa,
were kept in the treasuries of the Mongol princes for consultation as need
arose. No complete copy has survived, but it is possible to assemble
'fragments' of the code from various sources, and by careful study of these
fragments the general pattern of the Yasa can be recovered.
There are, then, three essential elements in this reconstruction: Chingiz
Khan laid down a coherent code, this was done at the quriltai of 1206, and
the code may be reconstituted from surviving fragments. So far as I have
been able to discover, these three elements seem first to have been brought
together by Petis de la Croix, in his Histoire du grand Gengkizcan,
published in 1710, with an English translation following in 1722.2 Pétis de
la Croix's sixth chapter is entitled 'The Description of the General Diet of
the Moguls, called in their Language Couriltay. The Establishment of the
Yassa, that is to say, the Mogul Laws. Temugin changes his Name for that
of Genghizcan'.3 After giving an account of the quirltai, which he dates in
1205, the author continues:
When he had thanked them all (i.e. those present at the quriltai) for the Marks of Love and Respect
they show'd him, being sensible that the chief Duty of a Prince is to establish good Laws, he declared
to them that he thought to add to the antient Laws some new ones which he desired, and commanded
that they would observe.4
There follows a list of 22 provisions of the Yasa, derived in the main,
apparently, from the Timurid historian Mirkhwand, with additions from the
accounts of European travellers and other sources.
Petis de la Croix's view thereafter determined the pattern of all subsequent
discussions of the Yasa, and achieved classical formulation in V.A.
Riasanovsky's Fundamental principles of Mongol law.5 It remained more or
less unquestioned until Ayalon's articles began to appear in 1971.
The sources for the alleged promulgation of the Yasa at the quriltai of 1206
are in effect two: the anonymous Secret History of the Mongols and the
Jami' al-tawarikh of Rashid al-Din. According to the Secret History, the
most nearly contemporary authority, on the occasion of the quriltai Chingiz
Khan entrusted the supervision of certain important matters to his adopted
brother, Shigi-Qutuqu:
'Divide up all the subject people and apportion them to Our mother, to Us, to Our younger brothers
and sons according to the name of the people,
Splitting up those that live in felt-walled tents,
Separating those that live in dwellings with wooden doors. Let no one disobey your word!'
Further, he entrusted Shigi-qutuqu with the power of judgement over all and said to him, 'Of the
entire people
Chastising the robber,
Checking the liar,
execute those who deserve death, punish those who deserve punishment. Furthermore, writing in a
blue (-script) register all decisions about the distribution and about the judicial matters of the entire
population, make it into a book (i.e. permanent record). Until the offspring of my offspring, let no one
change any of the blue writing that Shigi-qutuqu, after deciding in accordance with me, shall make
into a book with white paper. Anyone who changes it shall be guilty'.6
What are we to make of this account, which even Ayalon regards as a
description of the institution of the Yasa? The first thing to notice, I suggest,
is that nowhere in the passage is the term yasa used to describe what
Chingiz Khan has instituted. Indeed, although the term is found a number of
times in the Secret History, it generally seems to mean 'order' or 'command'.
It never refers to a legal code of any kind. A characteristic example occurs
four years before the quriltai:
Chinggis-qahan engaged these Tatar tribes in battle at Dalan-nemürges… Before fighting, Chinggis-
qahan gave this order to all: 'If we conquer the enemy, we shall not stop to plunder…'.7
Secondly, if we for the moment suspend our commitment to an 'institution
of the Great Yasa', what we find here, on further examination, is something
quite different. By 1206 Chingiz Khan had reached a pinnacle of success.
He
was distributing rewards (soyurqal) to his faithful followers. Shigi-Qutuqu's
reward included the privilege of keeping the population registers that
recorded the distribution of subject peoples among the royal family, and the
exercise of judicial functions. Both the details of the population distribution
and Shigi-Qutuqu's judicial decisions were, apparently, to be recorded by
him in a 'blue book'(kökö debter).8 Such decisions were to be regarded as
unalterable. But there is absolutely no indication that Chingiz Khan himself,
or Shigi-Qutuqu as his newly appointed chief judge, was to lay down a
general legal code. At most, Shigi-Qutuqu was authorized to begin the
establishment of a kind of case law, a body of written legal precedents. I
shall return later to the possible implications of this.
It seems to me, then, that we must conclude that the Secret History of the
Mongols is innocent of any information whatever on the establishment of a
'Great Yasa' at the quriltai of 1206 or at any other time.
Turning now to the evidence to be found in the Jami' al-tawarikh, the
standard discussion of it in Riasanovsky's book may be quoted:
Rashid-Ed-Din wrote that after his victory over Wang Khan (1206), Jenghiz Khan 'convoked a great
assembly and in gratitude for his great success ordered wise and strong yassaks and happily ascended
the throne of the Khans' …In another passage of his work, Rashid-Ed-Din spoke more definitely
'When the year of the hare came, which fell on the 614th year of Zul-ka'da (sic) (1218?) Jenghiz
Khan convoked an assembly, organised the Kurultai, laid the foundation of the yassak which had
been composed of innovations and ancient rules, and undertook a military expedition in the country
of the Khorezmshah' …It is evident that, in the second case, reference was made to the law, or more
exactly Code of Laws compiled from ancient provisions and innovations introduced by Jenghiz Khan
and approved by the Kurultai. It is possible that the resolution to issue a code of Laws was first
adopted in 1206, and that its fundamental rules were confirmed by the Kurultai in 1218 A.D.9
The first of Riasanovsky's quotations is misdated, and should refer to the
year 599/1202–3. It has no connexion with the quriltai of 1206. Rashid al-
Din's account of that event occurs later, and contains no reference of any
kind to the Yasa. In any case, the 1202–3 reference speaks of yasas (yasaq-
ha), not of 'the Yasa'. We clearly have here one of the many instances of
yasa as 'decree'.10
Riasanovsky's second quotation, concerning a quriltai in 1218–19, might be
translated as follows:
He held a quriltai among them; he laid afresh the foundation of the practices (ayin) and customs
(yusun) of the Yasa.11
I regret to say, since I would prefer a tidy solution to the problem, that there
is no escaping the fact that some kind of legal code is apparently implied
here. Still, it is odd that the incident is not reflected in the Secret History of
the Mongols, if this was indeed an event of major importance in the Mongol
Empire's evolution. It is even more puzzling that, both on this occasion and
elsewhere in the
Jami' al-tawarikh, Rashid al-Din's references to the Yasa are so exceedingly
brief and uninformative. Now Rashid al-Din is known to have had
unrivalled access to early Mongolian sources on the career of Chingiz
Khan. It would hardly seem that he deduced from them that the Yasa was a
matter of any great importance. He never gives any kind of list of its
contents, whereas he includes a long chapter giving an account of Chingiz
Khan's biligs, or maxims. Ayalon remarks that 'it is…very unfortunate that
Rashid al-Din has so little to tell us about the Yasa's contents'.12 Not so
much unfortunate, I submit, as distinctly suspicious. If the greatest of all
Mongol-period chroniclers thought the Yasa hardly worthy of comment, one
may very well begin to wonder whether modern historians may not, to say
the least, have markedly overstated its significance. But whatever we may
make of Rashid al-Din's evidence, it is perhaps worth emphasizing at least
that he does not connect the Yasa with. the quriltai of 1206.
 II

So much for the sources on the origin of the Yasa. I would like now to turn
to the question of its contents. Here I must refer again to the first of
Ayalon's articles. The main authorities that historians have used for the
assembling of 'fragments' of the Yasa are Maqrizi, al-'Umari, Bar Hebraeus
and Juwayni. Ayalon demonstrated the 'inescapable elimination' of
Maqrizi's supposed informant, Ibn al-Burhan, who was believed to have
seen a copy of the Yasa in Baghdad and to have reported its contents to
Maqrizi. He then went on to show—quite conclusively, it seems to me—
that Maqrizi derived his information without acknowledgement from al-
'Umari, and al-'Umari and Bar Hebraeus theirs, with acknowledgement,
from Juwayni. Ayalon has thus drastically cleared the ground. Only one
significant source, Juwayni, remains for the contents of the Yasa, and the
total abandonment of the time-hallowed procedure of painstakingly
assembling and classifying 'fragments' from other sources is indicated.
We must therefore give some attention to what Juwayni has to say. The
second chapter of his history is entitled: 'Of the regulations (qaw'id) which
Chingiz Khan framed and the yasas which he promulgated after his rise to
power'. It is a chapter by no means lacking in perplexing features. The
section in which Juwayni allegedly describes the 'institution of the Yasa'
deserves to be quoted at some length:
As his judgement demanded, he laid down a canon (qanun) for every matter and a regulation (dastur)
for every affair; while for every crime he fixed a penalty. And since the Tatar peoples had no script of
their own, he gave orders that Mongol children should learn writing from the Uighurs; and that these
yasas and decrees should be written down on rolls. These rolls are called the Great Book of
Yasas (yasa-nama-i buzurg) and are kept in the treasury of the chief princes. Whenever a khan
ascends the throne, or a great army is mobilized, or the princes assemble and begin (to consult
together) concerning affairs of state and the administration thereof, they produce these rolls and base
their actions thereon; and proceed with the disposition of armies and the destruction of provinces and
cities after that pattern.
At the time of the first beginnings of his dominion, when the Mongol tribes were united to him, he
abolished reprehensible customs (rusum)
which had been practised by those peoples and had enjoyed recognition amongst them; and he laid
down such usages as are praiseworthy from the point of view of reason.13
We may notice in passing that, like the other relevant sources, Juwayni does
not associate the Yasa with the quriltai of 1206—indeed, he offers no
precise date.
Juwayni's s account is set in the context of a chapter in which he discusses
Mongol practices with respect to religious toleration, hunting as a training
for warfare, the organization of the army, official communications (the yam
system), and taxation of the conquered territories. I would argue that it is
this context that explains and makes sense of the chapter's yasa references.
The yasa part of the chapter, it seems to me, amounts to a discussion of the
promulgation by Chingiz Khan of precepts of various kinds, and the writing
down and preservation of those precepts for future consultation. If the
chapter is treated as a whole it is a doubtful proposition that what Juwayni
tells us about Chingiz Khan's yasas should be forced into a pre-cast
framework labelled 'Great Yasa of Chingiz Khan'.
What, then, was it that was written down on those rolls in the treasury?
Apparently it was such precepts as might assist in the planning of 'the
disposition of armies and the destruction of cities'. The manner in which
such actions were to be taken was, it would seem, prescribed by the all-
conquering Chingiz Khan. This is not at all the kind of thing that usually
tends to characterize the lists of 'fragments' of the Yasa of which Ayalon is
so justly scornful—prohibitions of sorcery and adultery, of the Muslim way
of slaughtering animals, of washing in running water—the sort of ancestral
custom that Chingiz Khan's son Chaghatai defended and enforced so
vigorously. No, it is far more likely to be the kind of thing to which, as we
have seen, Juwayni devotes the bulk of his chapter: the hunt; the army; the
yam system.
In his discussion of the sources for the Yasa's contents, Ayalon—having
dismissed Maqrizi, al-'Umari and Bar Hebraeus with ignominy—turns his
attention to the only remaining source, Juwayni. Nor does the unfortunate
Juwayni escape unscathed. He is described as 'a very biased and partisan
source'14—not only because of his 'nauseating' 'servile flattery' of the
Mongols as a whole, but more particularly because he was biased in favour
of one branch of the Mongol royal house, the descendants of Tolui. Ayalon
goes on to cast aspersions on Juwayni's competence as a historian. He
complains that Juwayni continually wanders from the point in his chapter
on the Yasa, discoursing on a variety of important but irrelevant matters,
when he ought to have been discussing the legal code:
The three major sections of the chapter on the Yasa, which occupy its greater part, namely, the
organisation of the army, of hunting and of the horse-post could easily be taken out of that chapter
and form a separate entity with no reference to any kind of law… The fact that according to the title,
the chapter deals with 'The Laws Chingiz Khan framed and the Yasas which he promulgated', cannot
serve at all as a guarantee that al-Juwayni would literally adhere to it.15
This criticism appears to rne to be entirely misconceived. Juwayni says that
his chapter will contain Chingiz Khan's regulations (Ayalon follows Boyle's
misleading 'laws' for qawa'id) and decrees, or orders (yasa-ha). This is
precisely what it does contain. Juwayni never promises to give an account
of the 'Great Yasa', and he is hardly to be blamed because we may consider
that that is what he ought to have been doing. Juwayni's use of the term
yasa, and even of the phrase yasa-nama-i buzurg, does not on the whole
suggest that he was thinking in terms of a formal code of comprehensive
legal enactments. We may notice, for example, the explanatory couplings of
words: qawa'id wa yasa-ha surely, therefore, 'regulations' rather than 'laws';
wa yasa-ha, 'commands' or 'decrees'. A few lines before Juwayni's s
supposed account of the institution of the Yasa he uses a similar phrase: 'In
accordance with the yasa and which he imposed, he utterly
destroyed…'16 The phrase reads yasa wa The use of the final i on
precludes any possibility of it being a proper noun: the same must
therefore surely apply to the other half of the doublet, hence 'a yasa', a
decree, rather than 'the Yasa'.
Consequently it is my view that a more generous attitude towards Juwayni
than Ayalon will allow is entirely consistent with the evidence. It is perhaps
possible that Juwayni was indeed attempting to provide an account of the
institution of a legal code—an attempt largely vitiated by a frequent and
inexplicable wandering from the point, as Ayalon suggests. But it is not
very likely. We may more plausibly believe that Juwayni did in fact know
what he was about. He was concerned with a wide range of Chingiz Khan's
regulations and yasas, some at least of which were written down for the
future guidance of the Mongol princes. Especially important were those
yasas that dealt with great affairs of state and with military matters. In
short, Juwayni's chapter, usually regarded as an essential foundation for the
study of the Great Yasa, is nothing of the sort. It is an account of some of
Chingiz Khan's yasas, certainly, but not of the Great Yasa at all. Ayalon has
done away with all the sources on the Great Yasa's contents with the
exception of Juwayni. It is my contention that, so far as information on the
Great Yasa's contents is concerned, the next step is to discard Juwayni's
chapter too.
 III

If such an interpretation of the evidence is allowed, two questions

immediately present themselves: did the Great Yasa exist at all? And if it
did, was it written down, its contents known and enforced?
It is reasonably clear that a Yasa in the sense of a binding legal code was at
least later believed to have existed, and was attributed to Chingiz Khan.
There seems to have been very little unanimity, however, over what it
contained. Juwayni sometimes gives yasa the sense of the 'fundamental law'
of the Mongols, without specific reference to Chingiz Khan as its author—
or even as being in opposition to an individual yasa of his: 'By the yasa and
custom (ayin) of the Mongols the father's place passes to the youngest son
by the chief wife. Such was Ulugh-Noyan (i.e. Tolui), but it was Chingiz
Khan's yasa that Ögedei should be Khan'.17 Juwayni never uses the term
'Great Yasa' (yasa-yi buzurg), though al-'Umari, who derived his
information from Juwayni, does on one occasion use the phrase al-yasa al-
kabira.18
Some evidence in the Ta'rikh-i tends to support the interpretation of
Juwayni that I have proposed. writes that the Great Khan Qubilai
issued an edict (yarligh) at the beginning of his reign, 'for the renewing of
commands and the strengthening of the yasa-nama of Chingiz Khan,
containing the customs (marasim) of conquest and rule'.19 It may be that
the avowed continuator of Juwayni, understood his predecessor
better than later historians have contrived to do.
Earlier than this, in the reign of Hülegü or Abaqa, al-Din writing
of various financial immunities, had added that 'the yasa-yi buzurg of
Chingiz Khan is similar, and they hold this to be laid down, so that the
property (mal) of men should not decrease'.20 Yasa-yi buzurg is a rare
phrase—it occurs only very occasionally in the Jami' al-tawarikh—but
there are enough instances of its use to testify to the fact that there was a
belief in the existence of such a thing. This, however, is not the same as to
say that the Great Yasa did in fact exist, in the sense of being written down,
its contents generally and indisputably known: though people certainly
thought, or said, that they knew what was in it.
There is a possible interpretation that seems to me to fit the facts, such as
they are, fairly convincingly. This is that, if we concede that belief in the
existence of the code presupposes—even creates—the existence of
something, that that something was not written down in a coherent form at
all. It may well have been no more than the recollection of those of Chingiz
Khan's utterances, or alleged utterances, that were more or less legislative in
character: utterances to which he or his descendants attributed binding
force.
The only real alternative is that proposed by Ayalon: that the code, if
written down, may for some reason not have been available for inspection,
possibly because it was regarded as sacred or taboo. A parallel case would
be that of Rashid al-Din's use of the Mongol chronicle Altan Debter, now
lost. As a non-Mongol he was denied direct access to the text, and had to
use Mongol intermediaries. But as Hambis's comparison of Rashid al-Din's
Altan Debter material with the Chinese version of it showed,21 this process
did not prevent Rashid al-Din from obtaining a perfectly accurate idea of
the Altan Debter's contents. Similarly, then, there seems no reason why he
or Juwayni could not have secured reliable and comprehensive information
about the contents of a 'taboo' Great Yasa, had they so wished.
But there does not appear to be much evidence, in fact, to support the
'taboo' hypothesis—perhaps only Juwayni's account of the taking out of the
rolls on great occasions. It would seem in any case a singularly bizarre way
of treating a series of legal enactments, if the Mongols expected anyone to
obey them. However, we must certainly believe one or the other: either the
Yasa did not exist as a written code, or it was unavailable to those who were
supposed to conform to it. How else are we to account for the fog of
vagueness and uncertainty which appears to engulf all the contemporary
writers when they come to speak about the Yasa? If high officials in the
Mongol administration like Juwayni and Rashid al-Din could not, or did not
bother, to gain access to the text—although they were concerned to write
lengthy histories of the rise and development of the Mongol empire—what
chance had anyone else? Are
we not justified in deducing not only that knowledge of the Yasa was
vaguely and uncertainly diffused, but also that it was not in reality an
institution of any great practical importance? There is other evidence to
point us in this direction. Ayalon discusses the discrepancies in the
application of the Yasa's alleged prohibition of the Muslim method of
animal slaughter. This is hard to account for if the Yasa was embodied in a
written document, but it is much less of a problem if it was not. In the
second instalment of his article,22 Ayalon has a long discussion of what he
(perhaps rightly) sees as Juwayni's biased presentation of how the Yasa was
faithfully observed by the ultimately victorious descendants of Tolui.
Ayalon's interpretation of this is easier to believe if a widespread vagueness
about the Yasa and its contents prevailed. The other side of the dynastic
coin is represented by the claim of the Central Asian Mongol prince Qaidu,
who according to argued (in his own interest) that 'in the yasa-nama
of Chingiz Khan' it is ordained that the Great Khan must be a member of
the family of Ögedei23—so excluding the descendants of Tolui. It was
evidently possible to argue a case on either side from the supposed contents
of the Yasa. Whatever allowances are made for the deficiences of the
sources, it is hard to believe that if the Great Yasa was indeed a legal code,
the contents of which were clear, published, generally known and generally
enforced within the Mongol Empire, one would not be able to find much
more evidence of it, and have much less scope for this kind of discussion. I
propose the following hypothesis: there was probably believed to be a
'Great Yasa of Chingiz Khan', derived in part from Chingiz himself and
perhaps in part from earlier Mongol custom. But this was not written down
in any coherent form, and it was therefore possible to attribute to it a wide
variety of provisions, as was thought necessary or desirable. In practice it
may very well have been a gradually evolving body of custom, not only
beginning before the time of Chingiz Khan but continuing after him. This
certainly seems to be the implication to be drawn from the references in
Chinese sources collected by Professor Cleaves.24 In these the 'Great Yasa'
is frequently ascribed to Ögedei rather than to Chingiz Khan.
What role in all this are we to grant to Chaghatai, Chingiz Khan's second
son who is usually regarded as the stern upholder and custodian of the
Yasa? Juwayni25 tells us that Chaghatai was chosen by his father to
administer and enforce yasa wa siyasat: not 'the Great Yasa' or 'Yasa of
Chingiz Khan'. Later Juwayni relates how Chaghatai's retainers were
constrained to behave themselves well because of their fear of his
(Chaghatai's) yasa wa siyasat.26 And in one of Juwayni's anecdotes about
the benevolence of Ögedei, Ögedei is represented as calling the yasa which
Chaghatai wishes to enforce (on not
washing in running water) our yasa and command (yasa wa ma).27
While it is interesting to note that Rashid al-Din, in his version of the same
anecdote, changes this phrase to 'the Great Yasa' (yasa-yi buzurg),28 I would
take the general tenor of the references to Chaghatai to indicate that he was
regarded as the guardian and expositor of the—probably un written—
Mongol customary law rather than of any 'Great Yasa of Chingiz Khan'.
 IV

What, then, may we say about the operation of the Yasa in the Ilkhanate?
The term yasa is not infrequently to be found in the sources for the history
of the Mongols in Persia, but more often than not its use can be plausibly
explained as requiring the sense of 'decree' or of 'punishment'. The phrase
bi-yasa rasanidan is the most common verb for 'to put to death'. The
sources contain numerous mentions of yasas, but there are very few
indisputable cases of 'the Yasa'. For example, records that revolt
broke out against the Ilkhan Geikhatu because 'he altered (digar kard) the
yasa of Chingiz Khan'.29 Is this the Great Yasa, or does it refer to a specific
decree? Juzjani makes my point for me quite explicitly:
To these (he is referring to Chingiz Khan's prohibitions of telling lies, committing adultery,
washing in running water and so forth) they have given the name yasa, that is wa farman
(translated into) the Mongol language.30
So these, from the perspective of the Delhi Sultanate in 1260, were
individual decrees, not part of a comprehensive legal code. And Juwayni
writes that the Great Khan Güyük
made a yasa that just as Qa'an (Ögedei), at the time of his accession, had upheld the yasas of his
father (Chingiz Khan) and no change or alteration occurred in the commands of those
(yasas), so too the yasas and commands of his own father should be immune from the
contingencies of redundance and deficiency, and free from the corruption of alteration.31
This is a significant passage. As in the Chinese texts referred to above, the
yasas of Chingiz Khan and Ögedei are tacitly equated; and they are given
the sense of individual decrees or commands. All may be described as
or farman-ha.
One might suppose that the Mongol Yasa or yasas would come under some
pressure during the reign of Ghazan, when the Mongols in Persia went over
to Islam and began increasingly to identify with their Persian subjects. This
was certainly the view of al-'Umari, who contrasted the observance of the
Yasa, still strong, according to his information, in the Chaghatai Khanate
and in China, with its decline in Persia and in the lands of the Golden
Horde, which had both been converted to Islam.32 Ghazan himself is
alleged by Rashid
al-Din to have been a fervent exponent of yasaq and yusun in his younger
days. It is interesting, therefore, to note that in his yarligh granting to
the Mongol troops Ghazan begins by praising Chingiz Khan's yasa—
whatever he may have understood by that—in the warmest terms, declaring
that because of it the Mongols had been able to conquer the world.33
Certainly in this context Ghazan, or at least Rashid al-Din, seems to be
expressing a belief in the existence and efficacy of a 'yasa of Chingiz Khan'
of some sort, though the phrase yasa-yi buzurg is not used. However, it
looks to me very much as though this lavish praise of the yasa is designed
to act as a smokescreen device, to divert attention from the otherwise
painfully obvious fact that Ghazan's yarligh is about to erode some yasa or
other attributed to Chingiz Khan. That there was opposition to such
erosions is vividly illustrated by a curious anecdote in Qashani's Ta'rikh-i
Uljaytu. He gives a long account of disputes at Öljeitü's court between
and Shafi'is. At the conclusion of the argument, he writes,
Qutlugh-Shah Noyan said to the other noyans: 'What is this that we have done, abandoning the new
yasaq and yusun of Chingiz Khan, and taking up the ancient religion of the Arabs, which is divided
into seventy-odd parts? The choice of either of these two rites (madhhab) would be a disgrace and a
dishonourable act, since in the one, marriage with a daughter is permitted, and in the other, relations
with one's mother or sister. We seek refuge in God from both of them! Let us return to the yasaq and
yusun of Chingiz Khan'.34
Two things, I suggest, may reasonably be deduced from this story. First,
there was still a considerable amount of feeling for the traditional Mongol
way of doing things, at least among some of the Mongol amirs; and
secondly, that if Qutlugh-Shah was at all typical, the Mongols had not as yet
acquired a very profound knowledge or understanding of the tenets of
Islam.
A similar reaction against Islamisation is ascribed to the Ilkhan Abu Sa'id's
ephemeral successor Arpa Ke'ün:
When he ascended the throne…he used the Mongol jasaq and siyasat and did not pay attention to the
yarligh of Sultan Khudabanda (Öljeitü) and Abu Sa'id which was presented to him.35
 V

I conclude, then, that there are difficulties, possibly insuperable difficulties,

in establishing the nature and contents of the Mongol Yasa, its association
with Chingiz Khan himself, or even whether it ever existed as a written,
coherent, enforceable code of laws. All I have been able to offer so far is a
hypothesis that, I hope, conforms more closely than the conventional view
to such evidence as we have.
But the concept, at least, of the Yasa seems long to have remained a real one
in the minds certainly of western historians but also—so far as it is possible
to judge—in the minds of the men of the later Mongol period and after.
Amuli, writing during Öljeitü's reign, tells us that Chingiz Khan's
precepts maxims (bilik-ha) and yasas were collected into a book
(kitab).36 And Ibn maintains that
Tankiz (Chingiz) had compiled a book on his laws, which is called by them (i.e. the Central Asian
Mongols) the Yasaq, and they hold that if any (of the princes) contravenes the laws contained in this
book his deposition is obligatory… If their sultan should have changed any one of those laws their
chiefs will rise up before him and say to him, 'You have changed this and changed that, and you have
acted in such-and-such a manner and it is now obligatory to depose you'.37
Ibn goes on to relate that the Chaghatai sultan was indeed
deposed for infringing a Yasa regulation about the holding of an annual
feast.
So the founder of the dynasty, it came to be felt, had created an institution
which should be respected and taken into account, even if its precise
provisions were so vaguely known or impractical of execution as to present
few obstacles to their evasion. The Yasa remained, apparently, in the
Mongol consciousness as a symbol of the Shamanist, primitive, simple and
perhaps (to some) 'purer' past, which had gradually been eroded by
conquest and world-empire. If the 'Great Yasa of Chingiz Khan' did not
exist, it was evidently necessary to invent it.
 VI

So far in this paper I have attempted what is very largely a destructive—

though to my mind a necessary—exercise. In conclusion I propose to offer
some tentative suggestions on the other side: that is to say, what judicial
machinery I think may in reality have existed.
For this purpose we must abandon the word yasa and look instead at yarghu
—a very common term in the Persian sources, used in a variety of senses
for some kind of court or investigation. There are nearly forty such
references, for example, in the section of the Jami' al-tawarikh which deals
with the Mongols in Persia. Most of them are rather brief, and tell us little
about precisely what was going on.
According to Rashid al-Din, yarghus were held by the Ilkhans—often with
the use of torture—in order to investigate the cases of fallen ministers or
other enemies of the ruler such as alleged conspirators.38 There were also
courts of enquiry, as into the defeat of Qutlugh-Shah and Chuban at Marj
in 1303, of which a record (yarghu-nama) was kept:39 and into who
was to blame for reverses suffered during Öljeitü's campaign in Gilan, as
Abru records.40 Juwayni refers only occasionally to yarghus, always
with reference to the investigation of plots or of complaints against
officials.41 Such yarghus as these appear to have been ad hoc courts
convened to deal with. specific cases—a phrase frequently used by Rashid
al-Din is 'they held a yarghu for (i.e. to deal with) him' (u-ra yarghu
dashtand).42 And these cases usually involve Mongols or are concerned
with Mongol state affairs.
In the Ta'rikh-i shahi, an anonymous history of the Qara-Khitais of Kirman
written before the end of the seventh/thirteenth century, we have more
detailed accounts of what went on in early Ilkhanid times when a local
yarghu was held. On one occasion a court of enquiry was convened jointly
by the deputies (na'ib) of the Qara-Khitai princess Terken Khatun and the
local basqaqs (representatives in this subject kingdom of the Ilkhanid
government). This was to investigate complaints by certain dissident
persons against Terken. The basqaqs and yarghuchis had the complainers
tied up naked for several days and interrogated, 'as was the Mongol
custom'.43
We are told at greater length about a dispute between Terken Khatun and
the people of Shabankara concerning the rightful ownership of Sirjan. After
much argument, in 663/1264–5 the Ilkhan Abaqa sent representatives to
Sirjan who were to hand it over to Terken's deputies and to hold a land
yarghu (yarghu-yi amlak) to examine the documents and title deeds of both
parties. A principle of division was arrived at,44 but the dispute crops up
again later,45 when we are told that a further yarghu-yi amlak was held—in
the convening of which, interestingly, were prominent. The yarghu was
presided over by an amir-i yarghu.
These detailed accounts are of great interest, but it is difficult to be sure
how far they reflect the practice of the central Ilkhanid government, and
how far what was peculiar to the subject kingdom of the Qara-Khitai in
Kirman.
There is evidence of a very specific kind of yarghu, however, in the
invaluable second volume of Hindushah Nakhjawani's Dastur al-
katib, which contains a long section describing the functions of the amir of
the yarghu.46 The Dastur al-katib provides information on aspects of
Persian administration at the very end of the Ilkhanid period, and it is hence
dubiously appropriate to read back from it to the conditions of a century or
more earlier. But as it happens there is an intriguing and perhaps significant
correlation between the operations of the Dastur al-katib's yarghu and what
I have already argued that we can deduce about the activities of Shigi-
Qutuqu in the judicial capacity to which he was appointed by Chingiz Khan
at the quriltai of 1206.
To recapitulate; Shigi-Qutuqu was to judge certain criminal cases on an ad
hoc basis; and he was to supervise the distribution of subject peoples. All
this was to be recorded in a 'blue book'. The word used for 'case', however
one should translate it, is 47

In his History of the tribes, Rashid al-Din gives an account of the career of
Shigi-Qutuqu who, it may be remembered, was a Tatar orphan who became
an adopted brother of Chingiz Khan; he is therefore included in the section
on the Tatars. Rashid al-Din says this of him:
He conducted courts of enquiry justly (yarghu-ha bi-rasti pursidi), and he was solicitous and helpful
to many criminals and caused his words to be repeated, lest (they) should confess out of terror and
fear; and he said, 'Do not be afraid, but speak the truth'. And in the discussions of the yarghuchis it
became well-known that from that time to this, in the province
of Mughulistan and those regions, the foundations of the yarghus are laid on the regulations
(qawa'id) which he established and followed.48
The point worth noticing here is that there is, as usual, no sign of a yasa,
Great or otherwise—if anything, judges (yarghuchis) follow the case-law
precedents (written or orally transmitted?) of Shigi-Qutuqu, not the
enactments of Chingiz Khan.
Turning now to the Dastur al-katib's evidence, what do we find? The
passage under consideration is not devoid of difficulties and contradictions,
but it is, I think, possible to extract from it a number of revealing points.
According to the writer, the Mongols had caused the decrees of the
yarghu to be laid down as a canon of justice (qanun-i rasti)49 or
alternatively as a qanun called yarghu-nama.50 This was apparently to
regulate solely cases in dispute between Mongols,51 and this system was
regarded as their equivalent of the Shari'a.52 The requisite qualifications for
the amir of the yarghu were that he should be knowledgeable concerning
the customs (rusum) and regulations (qawa'id) of the Mongol sultans and
amirs, and their yasaqs and turas.53 He should settle cases in accordance
with the qutatghu bilik of Chingiz Khan,54 with (according to some
manuscripts) the decrees of Qa'an (i.e. Ögedei) thrown in,55 or in
accordance with the qanun-i yasa wa yasaq of Chingiz Khan,56 the qanun
of the yarghu and the regulation (qa'ida) of the yasaq;57 sometimes with
justice and equity ('adl, ma'dalat, insaf, rasti) added;58 and he was to follow
the example of the great yarghuchis.59 The amir of the yarghu was to give
the successful disputant a yarghu-nama as a record to produce in the case of
further argument60—this yarghu-nama apparently being different from the
more general record already referred to. The amir-i yarghu and the scribe
who had written out the yarghu-nama were to be paid a fee for their
services.61
I suggest as a possibility, therefore, that what we have here is an echo of the
functions which, according to the Secret history, were delegated to
ShigiQutuqu by Chingiz Khan at the quriltai of 1206. Yarghus at least of
this type in Mongol Persia were tribunals for dealing with disputes between
Mongols, and their decisions were recorded in writing. Further, there was
apparently some body of previous practices or decisions available to the
judge for consultation. The cases that such yarghuchis heard might well
have included the same kind of disputes as those that Shigi-Qutuqu was
instructed to deal with,
recording his binding decisions as he went. And as Rashid al-Din's evidence
suggests, Shigi-Qutuqu's practices were regarded as a model for later judges
to follow. Could registers of decisions and proceedings (yarghu-nama) on
disputes between Mongols have been kept from the time of Shigi-Qutuqu's
appointment in 1206 until the writing of the Dastur al-katib, well over a
hundred years later? We do not know, though something of the sort might
be the reality behind later notions of a written 'Great Yasa'. But at least it is
one plausible way of explaining in part how a Mongol judicial 'system' may
have worked. And it does have the virtue of avoiding resort to the desperate
expedient of a fixed 'Great Yasa of Chingiz Khan' laid down in 1206.62
THE ILKHAN EMBASSIES TO QALAWUN: TWO
CONTEMPORARY ACCOUNTS

By P.M.HOLT

With the accession of Tegüder, alias on the death of his brother

Abaqa in 681/1282, the Ilkhanate was for the first time ruled by a Muslim.
Consequently the possibility appeared of the establishment of peaceful
relations with the rival Mamluk sultanate under Qalawun (regn.
678–89/ 1279–90). Two successive embassies were in fact sent to the sultan
during short reign, and accounts of these as seen in Mamluk court
circles are extant in the writings of two contemporaries. The first appears in
the largely unpublished biography of Qalawun, al-ma'thur min sirat
al-Malik by Shafi b.'Ali (649–730/1252–1330),1 and the second in
the published but incomplete biography, Tashrif al-ayyam fi sirat
al-Malik by the maternal uncle of Shafi' b.'AIi, al-Din Ibn
'Abd 2
(620–92/1223–92). Both writers served in the chancery of the
sultan in Cairo.
The account of the Mongol embassies in al-ma'thur forms a
3
continuous portion of the text (ff. 64b–80b), composed in rhymed prose,
and written with the clear purpose of glorifying Qalawun and his
achievements, while denigrating the Mongols. The section is headed 'An
account of what befell our lord the sultan and had befallen no other ruler,
i.e. the submission of the Mongols and their ruler's request for peace'. It
opens with a description of the courage and hardiness of the Mongol
hordes, and their extensive conquests since their appearance in A.H. 614
(1217–18)—one of the very few dates provided by this author. The Mamluk
victory at 'Ayn Jalut is described as having been achieved by under the
inspiration of Qalawun, who then goes on to clear Syria of the Mongols.
Baybars is not presented as the victor of 'Ayn Jalut but mentioned only in
passing in connexion with later campaigns. This is a reversal of the role
usually ascribed to him, but it must be remembered that the primary source
for the received version of events is Ibn 'Abd own court biography
of Baybars, not an independent authority.
Shafi' then passes (f. 66b) to the main subject of the section. He states that
on his accession was converted to Islam by a certain Shaykh 'Abd
of Mosul, who advised him to make peace with the sultan for the
sake of the remnant of the Mongol army (f. 67a). thereupon
proclaimed Islam throughout his dominions, and sent an embassy with a
letter to Qalawun. The ambassadors are named as the Imam al-Din (the
judge of Kayseri) and the Amir Shams al-Din b.al-Tayti al-Amidi. Shafi'
speaks disparagingly of the Persianized Arabic of letter, drafted by
one Jamal al-Din b.'Isa, but gives its text (ff. 67a–70b). In the letter, dated
mid-Jumada 1 681/mid-August 1282, speaks of himself as a convert
in youth to Islam. On his accession, he says, he convoked a quriltai of his
kinsmen, commanders and notables, who wished unanimously to carry out
Abaqa's command to resumé the war against the Mamluk sultanate.
Fortified by the advice of Shaykh 'Abd decided to seek
peace with Qalawun. He announces his establishment of Islam and the
Shari'a, his restoration of waqfs, his encouragement of the Pilgrimage, and
his opening of the roads for the free passage of merchants. Even a spy
disguised as a dervish, who had been arrested by military patrol, had been
sent back, although previously he would have been put to death
protests that such espionage is no longer necessary. The ambassadors were
strictly escorted while passing through Mamluk territory, and were received
by the sultan in state (f. 71). The audience was at nighttime by the light of
many candles. Qalawun sat on his throne, surrounded by his Mamluks in
golden robes. When the ambassadors entered, they were dumbfounded. The
sultan sent for the head of his chancery, al-Din Ibn 'Abd (the son
of al-Din), who appeared and stood before him—although this
deference was not required by his office. At the sultan's command, he took
his seat on Qalawun's right in accordance with precedent, and asked the
ambassadors their business in Turkish through an interpreter. al-Din
(here styled al-Shirazi) said he had an oral message and a letter. The sultan
ordered him to give the letter and repeat the message to al-Din, who
was skilled in correspondence. The ambassador demurred as his
instructions were to give the message to the sultan personally, but his
objections were overborne. The next morning, al-Din read the Ilkhan's
letter to his senior colleagues in the chancery, and told them all to draft an
answer (f. 72a). They all held back except his father al-Din, who
composed an answer but did not put it in writing as it did not win general
agreement. Then comes a characteristic touch. by Shafi'. Although he was
not himself present at the meeting, he produced a draft which was well
received, and a final answer was put together from this and a draft of
al-Din's. Here as on other occasions, Shafi' presents himself as the man with
the bright conclusive idea.4 The text of the letter, dated 1 681/3
December 1282, then follows (ff. 72a–77a).5 After receiving gifts from the
sultan, the ambassadors returned to the ilkhan to report the greatness and
magnificence of Qalawun.
The sultan's letter to the Ilkhan was in measured terms. profession
of faith is welcomed but Qalawun points out that his own conversion to
Islam was earlier, and counters allusion to succession by familial
right (Qalawun himself being a usurper) by the assertion that God has given
the inheritance to the one He has chosen. He notes that has overruled
the wishes of the quriltai and preferred a peaceful policy; had he done
otherwise, his attack would have turned against him. In response to
grant of free passage to travellers, Qalawun has issued similar orders to his
governors in Aleppo, al-Bira and 'Ayntab, but he insists that the
disguising of spies as dervishes was a practice initiated by the ilkhans. He
goes on to answer the oral communications of the ambassadors. To
assurance that he has no territorial ambitions, and that an immediate
agreement on the status quo might be concluded, the sultan replies
somewhat evasively says that there is no further need for warfare
against the Muslims; Qalawun draws his attention to the atrocities
committed in Anatolia, a dependency of the Ilkhanate, by his uncle
Qongqurtai. Finally, suggests that if hostilities are to continue, they
should select a place for an ultimate trial by combat—a proposal which
Qalawun dismisses out of hand.6
For the second embassy to the sultan, sent the shaykh who had been
the instrument of his conversion. Shaykh 'Abd travelled with a
great retinue and a parasol (described by Shafi' as 'a dome of hide') was
borne over his head. He was instructed to conclude peace with Qalawun.
The sultan was informed by his agents of every stage in the ambassador's
progress (f. 77b). Orders were sent to a senior amir, Jamal al-Din Aqush al-
Farisi (the name suggests that he was linguistically qualified for the duty) to
meet the ambassador at al-Bira, the Mamluk frontier-post on the Euphrates,
and to forbid the use of the parasol, which was one of the sultan's insignia.
Shaykh 'Abd deferred to the sultan's command, and made his way
to Aleppo, where the governor, Shams al-Din Qarasunqur had
made preparations for his reception. He and his retinue were held
incommunicado, while the governor sent a report of their numbers to the
sultan. Shafi' was surprised to note that the list included four dervishes for
chanting and religious music. On consulting his senior amirs and advisers,
the sultan decided to meet the ambassador—a decision which received
divine approval. A rhetorical account follows of the joyous entry of the
sultan into Damascus (ff. 77b–78b).
Qalawun took up his residence in the citadel, and sent for the ambassador
and his retinue from Aleppo. They were allowed to rest for three days, then
summoned to an audience. Like the previous ambassadors, Shaykh 'Abd
was confronted with the sultan enthroned among his
magnificently robed Mamluks. He himself was dressed as a dervish, and
was accompanied by the Atabeg (not here explained) and Shams al-Din b.
al-Tayti. He was of a light complexion, and his speech was confused (f.
79a). The reverse of this folio, which is defaced, describes the offering of a
sealed box as a gift to the sultan. When opened, it disclosed a steel pen-box
ornamented with gold and silver (?), allegedly the work of the shaykh
himself. The sultan received it with disdain, and handed it over at once to
his head of chancery. He listened to the ambassador's message, and
dismissed him. Then came the report of death, brought by carrier-
pigeon. The sultan sent for the ambassador, and told him the news. Shaykh
'Abd fell down in a swoon, and died a few days later. Shams al-
Din b. al-Tayti was taken to imprisonment in the Citadel of Cairo, and the
rest of the company were allowed to return home. Qalawun then sent spies
and informers to watch developments after death (ff. 79a–80b).
The account by Ibn 'Abd in Tashrif al-ayyam is more prosaic, factual
and detailed. It is not presented as a single episode but the data are
distributed among several annals. The introductory passage (pp. 2–4) is
headed 'An account of the perdition of Abagha [sic] and the passing of the
kingdom to Takudar called '. It gives the date of Abaqa's death.
680/2 March 1282), and mentions portents of the event. Like
Shafi', Ibn 'Abd asserts that the Mongols professed Islam and made
an approach to Qalawun as a stratagem in view of their weakness in face of
the Muslims. The text is given of a letter sent by to Baghdad in
which he announces his conversion to Islam and the restoration of the waqfs
(pp. 4–5).
The account of the first embassy follows (pp. 5–16). The ambassadors are
named as the chief judge al-Din al-Shirazi (who is here
designated the judge of Sivas), the Amir Baha' al-Din (atabeg of the Seljuk
Mas'ud of Rum) and the Amir Shams al-Din b. who is
described as one of the intimate entourage ( ) of the lord of
Mardin. This last is, of course, the Shams al-Din b. al-Tayti al-Amidi of
Shafi', and he is identified by al-Maqrizi (Suluk, I/3, 723) as being the wazir
of Maridin. Their journey from al-Bira to Aleppo (where they arrived on 21
Jumada II 681/26 September 1282), Damascus and Cairo is described much
as in the first account. As regards their audience with Qalawun, Ibn 'Abd
is brief in the extreme: he merely says that they made obeisance, and
delivered their letter and oral messages. The text of letter and
Qalawun's reply are given. These show some few variants from the version
given by Shafi' but nothing of substance. On their return journey, the
ambassadors reached Aleppo on 6 Shawwal 681/7 Jan. 1283. Ibn 'Abd
like Shafi' emphasizes the measures taken to prevent the Mongol
ambassadors having any contact with the sultan's subjects.
Ibn 'Abd account of the second embassy is broken up among four
separate passages. He first (p. 44) briefly announces the arrival at Damascus
on Tuesday, 2 682/21 February 1284, of the ambassadors, whom he
names as Shaykh 'Abd and —a name which does not
appear in Shafi' b.'Ali's account. In the second passage (pp. 48–50), he
represents 'Abd as advising to profess Islam as a trick, so
that he would have no trouble from Qalawun while he settled accounts with
Arghun and his other kinsmen. 'Abd is said to have been a
Mamluk by origin. He is represented as having acquired ascendancy over
and the Mongols by his charismatic powers, which he expected to be
equally effective in Qalawun's territories. His journey, retinue and reception
at the frontier by Aqush al-Farisi are described much as by Shafi' but
specific dates are given for his arrival at Maridin (4 Rabi' II 682/2 July
1283) and at Aleppo (26 Shawwal 682/17 Jan. 1284). In contrast to Shafi',
Ibn 'Abd says that the ambassadors were sent on to Damascus before
Qalawun's arrival there.
Ibn 'Abd third passage (pp. 61–66) summarizes developments in the
Ilkhanate ending in the overthrow of and the accession of Arghun. It
tells how Qalawun received a despatch in cipher from his chief spy
reporting on events, and gives its contents.
The fourth passage (pp. 68–9) completes the story. The news of
death reached Qalawun at Gaza, when he was on his way to Syria. He
entered Damascus on 2 Jumada II 683/16 Aug. 1284, but instead of a
rhetorical description of the event such as Shafi' provides, Ibn 'Abd
merely remarks, 'The people displayed their customary joy at his arrival'.
'Abd and having been held incommunicado, did not
receive letters telling of death. They were summoned to an audience
with the sultan, and forcibly compelled to kiss the ground before him. They
delivered the Ilkha's letter, and (apparently after leaving the sultan's
presence) were told of their master's death. At a further audience, 'Abd
presented his gift, which Ibn 'Abd simply says was
accepted. The story ends very differently from that told by Shafi': we are
informed that the ambassadors continued to be generously entertained and
kindly treated.
The account concludes with the text of Ahmad's second letter to Qalawun
(not given by Shafi'), which is dated early Rabi' I 683/May-June 1283 (pp.
69–71). It is in fact the credentials of Shaykh 'Abd as ambassador.
In the preamble professes his desire to promote the well-being and
peace of mankind, especially the Muslims. He describes his sending of an
embassy to the chiefs of the Golden Horde, the regent Nogai and the khan
Töde Möngke, deploring the decline of the Mongol empire through its
divisions, and urging a return to the precepts of Jochi Khan, i.e. the
restoration of good mutual relations. Having received a favourable reply,
now sends Shaykh 'Abd at Qalawun's request (as he
asserts), giving him full powers to negotiate a treaty. In the circumstances
there was, of course, no answer to this letter.
The two accounts contain a number of mutually incompatible details, which
are not easily reconciled or explained, but the more sober narrative of Ibn
'Abd is probably the more reliable. Apart from these discrepancies,
both agree in illustrating the attitude of the Mamluk sultan and his servants
towards the Ilkhanate in the late seventh/thirteenth. century. The Mongols
are viewed with extreme mistrust: conversion (which appears to be
dated by both after his accession) and his attempt to establish friendly
relations are presented as due to military and political calculation and
lacking in personal conviction and sincerity. The Mongol ambassadors are
escorted through Mamluk territory in the medieval equivalent of a sealed
train, and then held without communication with either the sultan's subjects
or the Ilkhanate. Their audiences with Qalawun are set-pieces of stage-
management, designed to impress the ambassadors with the superior
magnificence of the Mamluk sultan and his court. Yet behind this
apparently confident presentation of events, one can detect traces of the old
fear of the Mongols as the uncouth but all-conquering barbarians who
destroyed the heartlands of dar al-Islam. This is the note struck by the
opening words of Shafi' b.'Ali (f. 65a): Well known are the great power of this enemy
and his ease in coming and going, his great numbers and his high resolution in spite of his miserable
equipment, his audacity and firmness, his fortitude and capacity to withstand even freezing
conditions, his contentment with clothes that do not cover his nakedness, with food that does not
sustain his strength, with drink that does not quench his burning thirst…. The summer does not repel
them with its heat, nor the cold with its chill, nor an enemy with his numbers.
The lasting fear of the Mongols by the Mamluk sultans and their subjects is
openly expressed by Ibn 'Abd when, after describing the defeat of
by Arghun (p. 63), he concludes: God willing, may they [sc. the Mongols] ever
continue at odds with one of them killing another until God has purged the earth of them. May God
cause every enemy of our lord the sultan to perish, and make His name blessed everywhere.
 THE CRUSADES OF 1239–41 AND THEIR AFTERMATH

By PETER JACKSON

The period of the crusades of Theobald of Navarre and Richard of Cornwall

is a critical one in the history of the Latin Kingdom of Jerusalem. As a
result of truces made by the crusaders with neighbouring Muslim princes
the kingdom came to embrace, albeit briefly, an area more extensive than it
had covered at any time since the losses inflicted by Saladin following his
victory at in 1187. And yet this triumph was but the prelude to an
engagement at La Forbie in October 1244, which was as grave a
catastrophe as and from which the kingdom never recovered. Here the
Frankish army was decimated by the Egyptians and their Khwarizmian
allies, a new and brutal element in the politics of southern Syria; and most
of the newly regained territory was lost within the next three years. In this
paper I propose to examine the events of the years 1239–44, with a view to
re-evaluating the military and diplomatic achievements of the crusades and
to placing the disaster at La Forbie more securely in context.1
 I

Thirteenth-century Latin Syria produced no chronicler comparable in

authority and depth with the great twelfth-century ecclesiastic and
statesman William of Tyre; and the modern-day historian who laments this
dearth will be moved to do so not least when studying the history of the
kingdom of Jerusalem during and immediately after the crusades of
Theobald and Richard. Of the two continuations of William of Tyre's work
which cover the expeditions themselves, the Estoire de Eracles surveys
them at some length and the so-called 'Rothelin' continuation provides a
still longer and more detailed account, especially of Theobald's crusade. Yet
neither, as we shall see below, supplies the complete picture of all the
negotiations in which the Western leaders were involved. For the period,
moreover, from Richard's departure in May 1241 down to La Forbie they
are still less satisfactory. The Eracles furnishes only the most severely
condensed account of these years, containing in any case remarkably little
on external relations, and 'Rothelin' none whatsoever.2
The Islamic sources are scarcely more illuminating. Medieval Muslim
chroniclers are notoriously indifferent to events within the Frankish world,3
and the two principal authorities contemporary with the period 1239–44 are
no exception. It comes as a greater surprise to discover that they fail to do
justice even to the complexities of the Islamic scene. The Ibn al-Jawzi
(d. 1257) and Ibn (d. 1298)—the one the author of a universal
chronicle, the other the historian of the Ayyubid dynasty in particular—
exhibit two major defects. The first is that neither writer provides much
indication on precisely which territories were under the rule of a particular
Muslim prince at any given time—a question of crucial importance in view
of the sizeable territorial concessions obtained by the Franks during the two
crusades. It is only from the topographical-historical work of Ibn Shaddad
(d. 1285)4 that we discover which territories were in the possession neither
of Ayyub, sultan of Egypt (1240–49), nor of his bitter rival
Isma'i, prince of Damascus (1239–45), but of a third prince, Da'ud,
who ruled southern Palestine from Kerak down to 1249 and whose relations
with the Franks, as we shall see, were the most problematical. The second
defect is the way in which the two authors tend to neglect events on the
Syrian-Egyptian frontier. Even though the was in Cairo from 639/1241–
42 and on at least one occasion met and conversed with Ayyub,5 he betrays
a marked preference for his native Damascus, northern Syria and the Jazira.
Ibn for his part, arrived in Egypt from at the beginning of
642/May–June 1244,6 and his information on southern Palestine prior to
that year is consequently often rather vague. The gaps in our knowledge are
filled only to a limited extent by the later, Mamluk chroniclers, of whom the
most valuable are al-Nuwayri (d. 1332) and Ibn Duqmaq (d. 1407). The
fifteenth-century compilator al-Maqrizi (d. 1442) furnishes yet more
material for the period under discussion; but its haphazard arrangement
makes his chronicle one of the most difficult to use, even if at the same time
the most frequently quoted.7
In these circumstances two Christian Arabic sources assume considerable
importance. The major portion of the anonymous work known as the Siyar
al-aba' comprises simply biographies of the Coptic patriarchs of
Alexandria, but the final section of the unique Paris MS arabe 302 is a
detailed history of Egypt from 1215 to 1243 written by someone who was
clearly a

contemporary.8 A complete edition and translation of this section has been

available only since 19749 and has apparently attracted scant notice.
Modern scholars tend to use the Siyar through the medium of extracts
incorporated by Blochet in the footnotes to his translation of al-Maqrizi.10
Blochet's selection was made on highly idiosyncratic grounds, however, and
omitted, from our point of view, a good deal of the most significant
material. This applies especially to the military operations of both Ayyub
and Isma'il during the years 1241–43, concerning which our Muslim
sources are almost totally silent. A recently published work on Ayyubid
history, relying principally on the and Ibn has therefore been able
to describe these years as free of hostilities between Cairo and Damascus.11
The Siyar reveals, in fact, that the case was quite different. As a Christian,
moreover, its author displays a greater interest in Frankish affairs than do
his Muslim contemporaries: he is aware, for example, of the dissensions
over foreign policy within the Frankish camp, and he is the only Eastern
writer to refer specifically to the Emperor Frederick II's embassies to Cairo
in 1241 and 1243.12 This interest is shared by his fellow-Christian, Ibn al-
'Amid (d. c. 1273), the last authority requiring mention at this juncture. It is
only these two writers who tell us of the important role played in Muslim-
Frankish relations by the renegade Ayyubid prince al-Jawwad Yunus and
thus show that the choices facing the Franks were far more complex than
has generally been believed.
 II

In the secondary literature, the verdict on Theobald of Navarre's crusade has

been overwhelmingly unsympathetic. To a certain extent it has suffered
through comparison with the crusade of Frederick II eleven years earlier.
By exploiting the rivalry among the Ayyubid princes with consummate
skill, Frederick—so runs the view—had secured remarkable gains for the
kingdom of Jerusalem.13 Theobald and his colleagues allegedly lacked the
ability to execute a coup of similar proportions.14 They have further been
contrasted unfavourably with the emperor's brother-in-law, Richard of
Cornwall, whose expedition followed immediately on Theobald's departure
from Syria and whose treaty with Egypt is seen as completing the task the
king of Navarre had left half-finished:15 with Richard's treaty, it is claimed,
the work of Frederick II at last began to bear fruit.16 It is but a short step
from here to regarding the Syrian Franks who overturned the Egyptian
alliance after the earl of Cornwall's departure as responsible for the disaster
of 1244.17 Underlying these verdicts is the assumption that Frederick's pro-
Egyptian policy, as expressed in his own treaty with Ayyub's father,
al-Kamil, in 1229 and in Richard's agreement with Ayyub in 1241, was in
the Latin Kingdom's best interests.18 This is in fact a highly dubious
proposition. Among the criticisms levelled by contemporaries at the
emperor's Egyptian treaty, of which some admittedly smack of bitter
prejudice, one in particular appears well grounded. It was that Frederick had
chosen to treat with al-Kamil, who at the time controlled only a small part
of southern Palestine, and had failed to secure an agreement also with
Da'ud, then prince of Damascus and the legal ruler of the territory
ceded to the kingdom.19 This objection had been vindicated by events
following the emperor's departure, for al-Kamil had proceeded to recognize
authority in the regions adjoining the newly surrendered territory,
thus effectively abdicating responsibility for the observance of the treaty by
establishing a hostile prince in the vicinity of the Franks.20 There are sound
reasons for believing, in fact, that the emperor's primary concern in 1229
had been not the strategic interests of the kingdom of Jerusalem but the
commercial advantage of his kingdom of Sicily, from which he was to
maintain relations with the Egyptian court right down to his death in
1250.21
Moreover, one sometimes detects the view that there are bizarre features of
Theobald's crusade which can be put down to nothing more than the leader's
personal inadequacy. No amount of casuistry on my part here will obscure
the fact that Theobald had displayed little judgement or consistency back
home in France; that at a critical juncture he proved incapable of exercising
discipline over his subordinates, with the direct result that around a third of
the crusading knights were annihilated or taken prisoner in an Egyptian
ambush at Gaza on 13 November 1239;22 and that by the time of his
precipitate departure from Syria his credit with the Franks as a whole was
minimal. It might at least be pointed out in his defence that Theobald was
accompanied to the Holy Land by a number of French magnates with whom
he had been at war as recently as the early 1230s. Even though an attempt
had been made to reconcile the various parties in the intervening period,
Duke Hugh of Burgundy had been Theobald's
MAP 1: Syria and Jazira
MAP 2: Palestine

rival for the leadership of the crusade;23 and the resulting tensions may well
have played a part in vitiating the morale of the crusading army. The
question is certainly worthy of investigation, though considerations of space
preclude it in this paper. What I do want to demonstrate initially is that on
two counts Theobald has been criticized unjustly. In the first place, the
strategy adopted at the council of war preceding the advance towards Gaza
has been seen as combining two distinct and mutually exclusive courses of
action.24 And secondly, it has been assumed that the Gaza operations
represent the sum of Theobald's military activity in Syria.25
The diplomatic situation within the Islamic world when Theobald
disembarked on 1 September 1239 would have baffled greater men than the
king of Navarre. Soon after annexing Damascus al-Kamil had died in
March 1238, and his empire had rapidly disintegrated. Initially the
sovereignty of the new sultan of Egypt, his younger son al-'Adil II, was
recognized also in Damascus, where a military junta had set up as governor
a minor prince of the blood named al-Jawwad Yunus. Al-Jawwad's
increasingly obvious aspirations towards independence, however, shortly
provoked measures by al-'Adil to assert direct control over the city;
whereupon al-Jawwad ingeniously placed himself beyond the scope of
reprisal by exchanging Damascus for certain of the Jaziran territories of al-
Kamil's elder son Ayyub at the beginning of 1239. Ayyub aimed at
nothing less than supplanting his brother in Cairo, and in the early spring
advanced at the head of a large army as far as Nablus, where he awaited
some move by disaffected Egyptian grandees that would put al-'Adil at the
mercy of an invading force. Nablus actually belonged to their cousin,
Da'ud, who had himself ruled Damascus from 1227 to 1229; he had
left his dominions around mid-May to visit al-'Adil in Cairo and was to
return to his capital, Kerak, only in September.26 To his rear, on the other
hand, Ayyub's position was less secure. At Ba'labakk his uncle,
Isma'il, who had in turn briefly held Damascus in 1237, prevaricated over
the despatch of auxiliaries to Nablus and was in fact secretly conspiring
with al-Mujahid Shirkuh of to seize Damascus in Ayyub's absence.
Ayyub's only reliable ally was of 27

The choice of Muslim princes to attack (of whom one enjoyed a somewhat
tenuous authority over Damascus and three others all harboured designs
upon it) was therefore highly embarrassing; and it is to Theobald's credit
that he did not simply rush headlong into conflict with Ayyub himself at
Nablus or with his advance forces, which had just occupied Gaza,28 but
instead waited upon events.
His two month delay before embarking on campaign occasioned comment
by Painter,29 though it is worth noting that the Fifth Crusade had been
guilty of precisely the same delay following its appearance in Syria in the
autumn of 1217.30 And as I shall show below, there is strong evidence that
Theobald spent these weeks not in luxurious idleness at Acre but in
negotiations at Tripoli with of It is possible that he had
good reason for not attacking Ayyub in particular. Western sources speak of
a Muslim penetration of Jerusalem which has been confused with the later
assault by but which is dated by the annals of Melk in the first days
of June 1239.31 This can only refer to the operations of Ayyub's forces,
which are known to have overrun the region around the Holy City at this
very time.32 The Dunstable annals allege that the attackers retired when
'imperial envoys' offered a renewal of 'the truce',33 suggesting that Ayyub
may have been glad to reach some temporary agreement with the Franks
that represented an extension of his father's treaty with Frederick.
The plan adopted at the beginning of November by the crusading leaders
and their local Frankish allies was first to fortify Ascalon and then to attack
Damascus.34 As a strategy that would alienate two Muslim rulers—the
sultan in Cairo as well as the prince of Damascus—and drive them to sink
their own differences, this has been seen as the height of folly and
explicable only by the need to compromise between two rival camps, a pro-
Damascene party headed by the Templars and a pro-Egyptian lobby under
the Hospitallers.35 As Bulst pointed out, however, there is no evidence at all
that the two orders did differ over foreign policy at this stage,36 and a far
more probable explanation lies to hand. The council of war met when
dramatic changes had just taken place in Muslim Syria. Late in September
Isma'il and al-Mujahid suddenly appeared before Damascus and stormed
the city, leaving Ayyub without a capital and inducing the majority of his
commanders and their forces to desert him and return home to Damascus in
order to ensure the safety of their families. And on 21 October Ayyub
himself was taken into custody by and imprisoned at Kerak.37 The
Franks cannot have failed to recognize the significance of these events. In
place of Ayyub, who had been committed to hostilities with the Egyptian
sultan, the prince of Damascus was now Isma'il, who at once inserted al-
'Adil's name in the Friday prayers and exhibited every intention of
38
being a dutiful subordinate. With Ayyub's downfall and the establishment
of a new regime at Damascus there was no question of Frankish military
activity causing a rapprochement, since good relations between Cairo and
Damascus had already been restored. The Franks were now confronted,
moreover, with the danger of encirclement. On learning of the fall of
Damascus, Ayyub's advance squadrons at Gaza had dispersed,39 and there
was nothing to prevent al-'Adil's own troops under Rukn al-Din al-Hayjawi,
who had been despatched to guard the Egyptian frontier early in October,40
from occupying the region. The Frankish decision to fortify Ascalon can
therefore be seen as a response to Ayyub's sudden eclipse and as an
essential precaution against any advance by Egyptian forces into Palestine.
There is no evidence that Theobald or the local Franks envisaged a war
against Egypt when they moved south in November 1239,41 and the Gaza
débâcle was brought about simply through the foolhardiness of his
subordinates Henry of Bar and Amaury of Montfort, who sought to emulate
a successful raid by Peter Mauclerc of Brittany a few days before. The
original goal of their chevauchée was apparently to drive off herds of
livestock known to be pasturing in the vicinity of Gaza,42 and this may well
explain why the various local nobles who accompanied the two counts
abandoned them when it became obvious they would have to fight the
Egyptian army.43 Right up until the late spring of 1240 Theobald and the
Syrian Franks, for all that they had initially planned to attack Damascus, in
practice followed a strategy that was essentially defensive. Gaza only
intensified the need to act with caution, since from this point onwards the
fate of the Frankish prisoners held in Egypt was a new factor to be
considered in deciding policy. Moreover, the Franks lacked allies: they had
as yet no understanding with any individual Muslim prince, and the powers
ranged against them were for the most part united. This is surely one reason
why they failed to take immediate reprisals against Da'ud when he
captured Jerusalem and destroyed the Tower of David in December 1239-
January 1240.44 He was not acting merely on his own account but was at
the head of a section of the Egyptian army,45 so that any move against him
might have endangered in turn the lives of the Frankish captives. It is
significant that Theobald, whose sole activity during the six months or so
following the battle appears to have been obscure parleys with Muslim
princes that were widely resented among the Western troops,46 acted with
some energy once Muslim unity was broken and he had acquired an ally.
It was no fault of the king of Navarre's that his first endeavours to secure an
understanding with a Muslim ruler had come to nothing. The only Syrian
Frankish source to refer to his contacts with of the Estoire
de Eracles, places them by implication after the Gaza campaign,47 and
hence secondary authorities have tended to locate these negotiations in the
spring of 1240, But rumours of imminent conversion had been
current even before the crusade set foot in Syria and are mentioned in an
optimistic letter sent to the West by Armand of Pierregort, Master of the
Temple, apparently in the summer of 1239.48 The precise circumstances in
which these reports had gained currency emerge from Ibn Hearing
that Isma'il and al-Mujahid of had designs on Damascus,
decided to send troops under Sayf al-Din 'AIi b.Abi 'Ali al-Hadhbani to
safeguard the city on Ayyub's behalf. To conceal his aim, however, he had
recourse to an extraordinary stratagem. Sayf al-Din was to pretend to
quarrel with on the grounds that the prince planned to surrender
to the Franks, and was to leave with his troops as if in umbrage.
went so far as to admit a body of Franks into the citadel of
in order to lend credibility to the idea. Unfortunately, the rumours were
believed by his subjects and caused a panic. A sizeable exodus from
occurred, seriously reducing capacity to influence events for
some time to come. Ironically, Sayf al-Din and his companions never
reached Damascus, being detained by al-Mujahid and incarcerated at 49

All this clearly occurred immediately prior to the fall of Damascus,

probably late in September, and that Theobald's negotiations with
were taking place early in the autumn is confirmed by a number of details
in Western sources. The Eracles tells us that one of the crusading leaders,
Count John of Mâcon, died at Tripoli during, or possibly after, the abortive
negotiations; and yet the 'Rothelin' continuation specifically mentions that
he was already dead at the time of the council of war preceding the Gaza
disaster.50 Count John is in fact commemorated by the abbey of Saint-Yved
at Braine on 3 November.51 And it is significant that both the count's death
and the collapse of the discussions with are placed under the year
1239 and, apparently, before Gaza by Alberic of Trois-Fontaines, who also
names two other crusaders, Anselm of Trainel and Robert of Courtenay,
who died around the same time: the obits of the two nobles are elsewhere
assigned to the first days of October.52 None of this evidence, of course,
excludes the possibility that al-
may have continued to figure as a potential ally for some months
afterwards. The Eracles links his overtures to the Franks with the threat to
from the regent of Aleppo,53 with whom made peace
'ostensibly', according to Ibn around February or March 1240.54 Even
then was engaged in another subterfuge, endeavouring to dupe al-
'Adil, and Ayyub's other enemies, into believing that he had accepted the
status quo and had abandoned his friend.55
The failure to secure an accommodation with was a bitter
disappointment; but in the spring of 1240 the diplomatic situation in
Muslim Syria changed once again, and this time in the Franks' favour.
had already antagonized the sultan by his refusal to hand over
Ayyub, who was far too valuable a pawn; now, on 21 April, he released his
prisoner and undertook to assist him in his conquest of Egypt in return for
concessions which included the promise of a joint expedition to put
himself back in power in Damascus as Ayyub's subordinate.56 This not
unnaturally caused consternation among Ayyub's enemies, and al-'Adil,
Isma'il and Ibrahim, who had succeeded his father al-Mujahid as
prince of in February, all prepared for war, the two Syrian rulers
advancing with their troops as far as al-Fawwar in the Sawad, east of the
Yarmuk river.57 The extent of Isma'il's alarm may be gauged from the
alacrity with which he entered into correspondence with the Franks. He is
usually assumed to have contacted Theobald at Sephoria ( ), after
news had reached him of the triumphal entry of Ayyub and into
Cairo in June.58 But it is evident from a phrase in the Siyar that Isma'il was
already allied with the Franks prior to this, when the fate of Ayyub's bid for
power hung very much in the balance;59 while we learn from the Master of
the Temple that the first Damascene overture reached the Franks when they
were still encamped in the 'Sablon' outside Acre.60
The Frankish-Damascene treaty was finalized during the summer,61 and we
shall consider its terms below. What needs to be stressed at this juncture is
that Theobald wasted no time in putting the alliance to good use. In July
1240 returned from Cairo to Kerak for the second time, to find his
territories under steady attack. The Franks had raided the Jordan valley (al-
Ghawr) and Nablus; they had again advanced to Gaza and begun work at
last on the fortification of Ascalon; and they had temporarily reoccupied
Jerusalem, where Theobald himself now made the pilgrimage at the head of
an impressive force.62 These almost totally neglected operations constitute
the high-water mark, as it were, of the king of Navarre's crusade and
demonstrate his readiness to adapt to changing circumstances in Muslim
Syria.
 III

The territorial clauses of the Damascene agreement have occasioned no

little uncertainty and debate among present-day scholars.63 But they fall
into two clear categories. Firstly, Isma'il was surrendering important
territories in his own possession: the Muslim-held regions of the Sidon
lordship, Toron (Tibnin) and Châteauneuf (Hunin), all of which had been
ceded by al-Kamil to Frederick II in 1229 but to which the Franks had been
unable, apparently, to make good their claim; 64 Tiberias the
ruins of Safed and, most importantly, the fortress of Beaufort (Shaqif
Arnun), which was intact.65 And secondly, he was recognizing the right of
the kingdom of Jerusalem to all its erstwhile possessions west of the Jordan
held at the time of their conquest by Saladin in 1187.66 Much of the land in
this broader category actually belonged to Nablus, for example,
67
Gaza, Jericho and the Holy City itself. But, as we saw above, the
Damascene alliance provided the Franks with the opportunity of recovering
it by force. If the agreement represented an offensive alliance against
however, it also entailed a defensive alliance against Ayyub, which
in Isma'il's eyes was surely its raison d'être. The Franks were to encamp at
Gaza in order to prevent the Egyptian army from entering Syria; the two
parties were to defend each other's territories against Ayyub, and neither
was to make peace with him without the other.68 Later Muslim sources also
inform us that the Franks were permitted to visit Damascus and buy arms.69
These concessions were to cause Isma'i considerable embarrassment. Not
merely did they provoke the wrath of certain religious leaders in Damascus,
notably the chief preacher Ibn 'Abd al-Salam al-Sulami, whom
Isma'i1 first put under house-arrest and later exiled; they also met with
obstruction on the part of his military subordinates. The commandant and
garrison of Beaufort refused to surrender the fortress, necessitating an
expedition by Isma'il in person in August or September to reduce it and
fulfil his undertakings to the Franks.70 The protracted siege would have left
him little time for the joint expedition with the Franks into southern
Palestine which is usually placed during the latter stages of Theobald's
crusade. Our sole authority for it is the Estoire de Eracles, which includes
of among those present.71 None of the Arabic sources,
however, refers to such military collaboration, except possibly against
town of Nablus;72 and during the summer of 1240, at least,
had good reasons for remaining in northern Syria. His principality of
had twice been raided by the Khwarizmians in recent months,73 and by the
early autumn they gravely threatened neighbouring Aleppo, whose army
suffered a crushing defeat on 2 November. Significantly, Arabic writers
depict as participating in campaigns against Egypt only after his
fierce struggle with the Khwarizmians as commander of the Aleppan army
in the Jazira, culminating in his victory near Edessa in May 1241.74 In all
probability, therefore, the Eracles has conflated the events of 1240 with
those of the summer of 1241, when Isma'il and as we shall see, did
attempt an invasion of Egypt in conjunction with the Franks.75 As far as
1240 is concerned, we can be certain only that Isma'il, in addition to
reducing Beaufort, gave hostages to the Franks as an earnest of his good
faith.76
In any case the Frankish-Damascene alliance swiftly foundered. According
to a well-informed Aleppan chronicler, when learned of the
Khwarizmian crisis early in November 1240 he had been about to raid
Frankish territory, at the head not merely of his own troops but also of a
detachment of the Damascene army.77 Which Frankish territory—Antioch,
Tripoli or the kingdom itself—is not stated. Possibly there is some
connexion here with the rumour, noticed by Matthew Paris, that before the
year was out Isma'il had lost confidence in the Franks and had contravened
the truce by reaching a settlement with their mutual enemies.78 This phase
of events, which coincides with Theobald's precipitate departure from Syria
and the first few weeks of Richard of Cornwall's crusade, is perhaps the
most obscure in the entire period under review. Isma'il may well have had
misgivings at the failure of the Hospitallers to adhere to the truce;79 he is
still more likely to have been discouraged by Theobald's sudden withdrawal
in mid-September; but it appears that the root cause of the breakdown of the
alliance was the capitulation of
For the author of the 'Rothelin' continuation, had participated in the
80
confederacy against Ayyub from the outset. This is misleading, since the
remaining evidence demonstrates that the prince of Kerak made peace with
the allies only at the very end of the summer. According to Richard of
Cornwall, whose letter is reproduced by Matthew Paris in the Chronica
Majora, Theobald had made a truce with but had left Syria before it
81
was finalized. Matthew's own statements elsewhere suggest that one of
the objects of these negotiations was the release of some of the Gaza
captives, allegedly in possession; and modern writers have
accordingly questioned the very authenticity of the truce.82 It is
corroborated, however, by the Arabic sources. The author of the Siyar tells
us that during their subsequent negotiations with the Egyptian Ayyub
the Franks demanded that he recognize the territorial concessions they had
received from 83 Ibn Shaddad, moreover, refers to a treaty between

Isma'il, and the Franks in 638 A.H. (began 23 July 1240)

whereby he surrendered Jerusalem, and Ibn al-Dawadari dates the arrival of
this news in Cairo in the early autumn of 1240.84 There is consequently no
doubt whatsoever that Theobald did reach an agreement with the prince of
Kerak, who was quite simply engaged in a desperate attempt to survive
since he had returned from Egypt diplomatically isolated. Once in power
Ayyub had reneged on promises extracted from him under duress, and
hopes of recovering his Damascene patrimony were frustrated yet
again.85 After an ineffectual gesture of assistance to the mutinous garrison
of Beaufort,86 he appears to have bowed to necessity. He had, after all, little
reason to persist in his friendship with Ayyub and every incentive to seek an
accommodation with his enemies in Syria.
One of the problems we face in trying to evaluate Theobald's crusade is the
tendency of the Frankish sources to muddle his diplomatic activity and so
minimize his achievement. His unpopularity at the time of his departure has
traditionally been linked with an Egyptian truce which the Eracles ascribes
to him, and 'Rothelin' even credits him with the release of all the Gaza
captives, which is known to have been the work of Richard of Cornwall.87
Yet it is noteworthy that neither writer mentions Theobald's dealings with
and that Matthew Paris, conversely, fails to associate him with any
Egyptian truce. There is thus a strong probability that the truce with
—which, as we shall see below, was not without its controversial features—
rapidly became confused in the chroniclers' memory with the even less
satisfactory treaty with Cairo in February-March 1241. The Siyar alleges
that prior to that treaty the Franks had repeatedly sent envoys to Ayyub
seeking peace.88 If, as is possible, these approaches were initiated by
Theobald, he was doubtless simply using his alliance with Isma'il to put
pressure on Ayyub to release his Frankish prisoners and was by no means
necessarily aiming at a settlement which would contravene the defensive
agreement with Damascus. This must unfortunately remain a matter of
conjecture; but it is at least preferable to a view which ascribes to Theobald
a total of three mutually contradictory truces before he left Syria. Not even
someone who had earned the king of Navarre's reputation could have been
so stupid.
More insidious than the confusion in the Syrian Frankish accounts is the
deliberate misrepresentation in Richard of Cornwall's letter. The earl makes
no mention of the military operations which had brought to negotiate
in the first place (nor, for that matter, does he once refer to the Damascene
alliance), and gives us to understand that the French crusading leaders had
opened negotiations with Kerak merely in order to be seen to have done
something.89 It is true that the terms of the proposed truce with were
problematical for the Franks. Isma'i restored to him, albeit reluctantly, the
town of Nablus, until 1187 part of the Latin Kingdom;90 and, according to
the version which reached Cairo at least, Jerusalem was not surrendered
outright but was to be shared with the Muslims.91 The agreement was
therefore open to one of the very objections that had been levelled at
Frederick II's treaty with al-Kamil in 1229.92 Yet if Theobald's truce with
was not ideal, it was far from being, as Richard of Cornwall
described it, 'of little consequence'.93 On the contrary, it rounded off the
territorial concessions won from Isma'il and thereby represented a
considerable advance on the emperor's work. For whereas Frederick's treaty
—as we saw earlier—had been made with al-Kamil alone and had little
effect in many areas where the sultan's authority was minimal,94 Theobald
had contrived a settlement with both the princes who currently occupied the
former territories of the kingdom.
It was his misfortune that entry into the alliance served to generate
dissension over its character. According to the 'Rothelin' continuation, the
Damascene envoys began to press for an invasion of Egypt, to which
Theobald, despite the favourable response among the local Franks, was
unwilling to commit himself.95 Isma'il was doubtless encouraged to plan an
invasion not only by the capitulation of but by reports that important
elements within Egypt desired him as sultan in preference to Ayyub.96 For
the Western knights, however, the prospect of an offensive threw into relief
the plight of their comrades who had been held captive in Egypt since the
battle of Gaza and whose lives an invasion would jeopardize in a way the
defensive alliance had not done; and it seems to have been around this time
that rumours spread concerning their harsh treatment.97 Paradoxically,
therefore, Isma'il may well have played into the hands of those who wanted
a truce with the new Egyptian sultan. The aspirations of this party, which
included the Hospitallers, the duke of Burgundy and Walter of Brienne,
received a further boost when declined to negotiate a fresh truce
with Richard of Cornwall.98 His reasons are not known. Most probably the
security of his new found friendship with Damascus and removed the
necessity for further dealings with the Franks, for whom as a self-
conscious though by no means consistent advocate of the jihad,99 could
only afford to evince disfavour.
Richard's arrival at Acre on 8 October 1240 had in fact tipped the scales on
the side of a Frankish settlement with Ayyub. According to the earl's letter,
the initiative in the ensuing negotiations came from the sultan, who sent a
high-ranking envoy to meet him at Jaffa.100 This was apparently Kamal al-
Din Ibn al-Shaykh, the brother of Frederick II's great friend Fakhr al-Din,
whom the Siyar names as Ayyub's representative.101 The sultan was
doubtless only too eager for a truce, since the coalition against him included
the greater part of Muslim Syria (Aleppo, whose regent Isma'il had been
endeavouring to win over as early as July-August 1240, apparently joined
the alliance around the end of the year102) and it was vital at least to detach
the Franks. But the most powerful force working towards a Frankish-
Egyptian settlement was surely the emperor, who was now in a better
position than at any time since 1229 to influence events. Richard had taken
good care on his journey overland through France and the kingdom of Arles
to remain in touch with his imperial brother-in-law,103 and Frederick's own
words a few years later indicate that the earl came out to Syria as his
accredited representative.104 It appears that Richard fostered the impression
for as long as possible that he was reluctant to commit himself to either the
pro-Damascene or the pro-Egyptian party, and he induced the former to co-
operate in the refortification of Ascalon, which was hardly incongruous
with the terms of the Damascene alliance.105 We know, nevertheless, from
his own correspondence that he had agreed on the advisability of a truce
with Cairo as early as November.106
Regrettably, no text of the treaty between Richard and Ayyub has come
down to us. Yet it seems clear that the only concrete gain it afforded the
Franks was the release of the knights captured at Gaza. The earl gives
prominence in his letter to a long list of places now allegedly surrendered
by the Egyptians.107 Many of these often hopelessly corrupted names
apparently denote villages in the environs of Jerusalem and Bethlehem;108
the list of more important places furnished by the Siyar, however, is broadly
in agreement.109 But it must be emphasized that these territorial clauses
have been misunderstood in two vital respects. Professor Jean Richard, for
example, sees the 1241 agreement as entailing 'no merely formal cession'
and assumes that it was by virtue of this agreement that the Franks were
able at last to reoccupy the areas listed by the earl, such as Tiberias.110 But
Ayyub at this time controlled not a single region of Syria or Palestine—
even Gaza was to be retaken by an Egyptian force only in May 1241.111 In
sharp contrast with Isma'il, therefore, the sultan was yielding territory that
was currently not in his own gift. The Sidon region, Beaufort, Châteauneuf,
Toron, Tiberias and Safed had all been included, as we have seen, in the
Frankish-Damascene treaty. Jerusalem, Bethlehem, Bayt Jibrin (Beth
Gibelin) and Kawkab (Belvoir), on the other hand, lay within
principality. Of these, the Holy City had certainly been one of the places
named in truce with Theobald; and after the departure of Richard,
who may have made a pilgrimage to Jerusalem,112 we find the prince of
Kerak back in possession in May.113 Accordingly, the 1241 treaty merely
recognized the Franks' rule over territories which they had acquired from
Ayyub's enemies—though by virtue of truces that were no longer operative.
Moreover, far from constituting, as has been alleged, an extension of the
Damascene concessions,114 the territories listed in Richard's agreement with
Ayyub actually fell short of them. Where Isma'il had theoretically
acknowledged the pre-1187 boundaries of the Latin Kingdom (though
excluding Kerak and Montréal), Ayyub, according to the Siyar, expressly
reserved for himself three areas of considerable strategic value: namely,
Nablus, al-Khalil (Hebron) and the city (though not the hinterland) of
Gaza.115 When reading the earl of Cornwall's triumphant announcement of
his 'acquisitions' to his English friends, it is difficult to avoid the conclusion
that the political realities in the Near East were being discounted in favour
of a policy already determined hundreds of miles away in Western Europe,
And yet there was a still more powerful objection to the 1241 treaty with
Egypt. As Saladin's victorious career in the 1180s had shown only too
clearly, it was crucial for the survival of Latin Syria that the Muslim
territories should not be concentrated in the hands of one prince. For this
reason, if for no other, Theobald would have been justified in resisting the
pressure to aid Isma'il in the conquest of Egypt. Equally, however, the duke
of Burgundy and the Hospitallers were misguided in their championship of
Richard of Cornwall's agreement with Ayyub. In 1243 Armand of
Pierregort was to write to the West complaining that the sultan sought peace
with the Franks merely in order to proceed with the reduction of Isma'il and
116 He could have voiced the same opinion, and doubtless did, two

years earlier. The terms of the respective treaties highlight the difference
between the two rival princes. For Isma'il Frankish military assistance was a
vital necessity, if only to keep Ayyub's forces out of Syria; hence the
relatively high price he was prepared to pay for it. Ayyub, on the other
hand, was satisfied merely with the Franks' benevolent neutrality, which
would secure his flank while he turned against his Muslim enemies.
 IV

He began with In the late spring of 1241 an Egyptian army under

Kamal al-Din Ibn al-Shaykh and Qarasunqur al-Saqi was sent into Palestine
by way of Gaza with instructions to occupy Nablus. The Egyptians were
surprised by forces under al-Jawwad (see below) in the hills west
of Jerusalem and, in spite of their superior numbers, routed with heavy
losses; Kamal al-Din was taken prisoner. A few weeks later, however, the
prince of Kerak made peace with Ayyub, and the Egyptian general was
released.117 Why chose to abandon the anti-Egyptian alliance at this
stage, we are not told; but one reason is possibly to be found in an incident
reported by Ibn Shaddad. He tells us that the Muslim slaves employed in
the reconstruction of the Templar fortress at Safed planned a mutiny and
sent secretly to lieutenant at 'Ajlun, who requested instructions
from his master. at once notified Isma'il, who proceeded to betray
the conspiracy to the Templars: the slaves were rounded up and
massacred.118 Most probably had intended the Muslim princes to
benefit from the mutiny and Isma'il's reaction was profoundly
disappointing; nor could the prince of Kerak be seen to countenance the
wholesale slaughter of fellow-Muslims. Whatever the place of these events
in policy, the reconciliation was a piece of good fortune for Ayyub,
whose plans to expand into Syria had suffered a humiliating reverse. For he
himself now had to face an offensive from Damascus, which would have
succeeded, according to the Siyar, had not the sultan and been at
peace.119
Isma'il appears to have been encouraged once more to attempt an invasion
of Egypt by reports of Ayyub's notoriously uneasy relations with his
amirs.120 In the early summer, accompanied by by Aybeg
lord of and by a contingent from Aleppo, he advanced
through territories, defeating the prince himself at in the
Bilqa and a further detachment of his at Nablus. Then the allies were joined
by a Frankish army and encamped at Gaza, which had been abandoned by
the Egyptian frontier forces.121 We learn from al-Nuwayri that around this
time Isma'il and the Franks also reoccupied Jerusalem, where the
unfortunate jurist al-Sulami, who had taken up residence there after his
banishment from Damascus, once again suffered a temporary
imprisonment.122 What happened next is uncertain. Later sources speak of
an engagement in which large numbers of the Syrian troops deserted to
Ayyub's army and the Franks were consequently defeated, an episode
usually placed by modern writers in 1240.123 The author of the Siyar, on the
other hand, who refers cryptically at a later juncture to the Franks' desertion
by Isma'il and their defeat by Ayyub, says in his account of the campaign
itself merely that the Syrians retreated.124 As they moved back through
dominions, his forces fell upon Aybeg contingent at al-
Fawwar on 22 August 1241 and routed him with the loss of his heavy
baggage.125
This reverse brought the Franks in turn to heel, and the Templars sent
envoys to Cairo to sue for peace on the basis of the existing territorial
dispositions.126 We are not told whether Ayyub at this point granted the
terms they sought, but subsequent events make it unlikely. Peace with
Egypt was rendered more imperative by the departure of the remnants of
Theobald's and Richard's crusades at the end of the summer. The earl
himself had embarked for Italy on 3 May, but Matthew Paris indicates that a
great many crusaders left Syria at the autumn passage.127 The withdrawal of
these troops represented a serious reduction in Frankish manpower and
obliged the local Franks to fall back from Jaffa, where the Christian army
had taken up position following Richard's departure, to Acre: such, at any
rate, is the testimony of the Eracles, which makes no mention of the
military setback in the summer.128 But the Franks were not to remain
inactive for long. Their return to a forward policy seems to have been
closely connected with the arrival at Acre of the Ayyubid prince al-Jawwad,
whom we encountered earlier as prince of Damascus in 1238–39.129
Dislodged from his new principality in the Jazira, al-Jawwad had returned
to southern Syria in the spring of 1241 with the aim of entering Ayyub's
service and, under his aegis, of recovering Damascus.130 On the advance of
Kamal al-Din, however, he had taken fright and joined who had put
him in command of the small force which defeated the Egyptians. But the
two princes shortly quarrelled, and at the end of June 1241 arrested
al-Jawwad and sent him off under escort to Baghdad. En route al-Jawwad
escaped and made for Damascus, where he attached himself to Isma'il.131 It
seems that Isma'i decided to use al-Jawwad in an attempt to retrieve his
Frankish alliance following the disaster in the summer and the Templars'
entry into negotiations with Cairo, for Ibn al-'Amid says specifically that
the prince was sent off to Acre to woo the Master of the Temple.132 'Li
Johet', who was born of a Frankish mother, used to speak of the Franks as
his brothers and was extremely popular with them.133 But he was to prove a
dangerous choice for such a mission. Initially he served Isma'il well,
accompanying the Franks on an expedition to Gaza, where they looted
whatever they could lay hands on and then withdrew before arrived
134
with his forces to reoccupy the area. Soon afterwards the Franks left al-
Jawwad at Caesarea and fell back on Acre:135 the reason is not given in the
Arabic sources, but their withdrawal seems to have been linked with the
outbreak of disturbances within Acre itself. The imperial bailli Filangieri
chose this moment to arrive secretly from Tyre and endeavour to seize the
city on Frederick's behalf; and we know from the Frankish account of this
episode that prior to their hurried return to Acre the Templars and the army
of the kingdom, under Odo of Montbéliard, had been stationed at
Caesarea.136
Ayyub was quick to seize his opportunity. Al-Jawwad was won over—
chiefly, it appears, by the guarantee of an Egyptian expedition to instal him
in Damascus in Isma'il's place—and was authorized to offer the Franks on
the sultan's behalf whatever territorial concessions they desired. The prince
in turn prevailed upon the Templar Master and other Frankish leaders to
swear to an alliance with Ayyub.137 This represented a considerable triumph
for the sultan, who for the first time, it must be stressed, was at peace with
both parties among the Franks and had thus shattered the links between
Acre and his hated rival at Damascus. He now went on to despatch to Syria
in March-April 1242 a force of three thousand horse under Rukn al-Din al-
Hayjawi, the victor of 1239. Al-Hayjawi's task, prior to the Damascus
expedition, was ostensibly to reconcile al-Jawwad with the sultan's ally
138 but in fact Ayyub had no intention of honouring his undertaking
to al-Jawwad. Very soon al-Hayjawi, who had joined the prince at Jaffa,
received instructions from Cairo to arrest him and send him in chains to
Egypt. Al-Hayjawi duly prepared the way by inducing al-Jawwad to
accompany him to Gaza. But here the prince himself received orders from
Ayyub to arrest al-Hayjawi; whereupon each of the two men revealed his
instructions to his companion and decided, not surprisingly, to abandon the
sultan's service. Al-Jawwad rejoined the Franks at Acre; al-Hayjawi fled to
Damascus, while those of Ayyub's troops who did not follow him retired in
confusion from Gaza to Egypt.139 The Templars now appear to have aligned
themselves once more with Damascus. They were not averse to depicting
the retreat of the Egyptian forces as a great Frankish triumph and sent
inflated reports to the West, where Matthew Paris commented acidly on the
nature of their 'victory'.140
With the Egyptian withdrawal, the Franks were free to continue their
attacks on A series of raids from Bethlehem, among other places,
was checked only when in person advanced on Bethlehem,
slaughtered its Frankish and other Christian inhabitants, and carried off
their children into captivity. Shortly afterwards he massacred a convoy of
pilgrims on their way back from Jerusalem to Acre and plundered their
possessions.141 Not until the early autumn do the Franks appear to have
taken their revenge. Then, on 30 October 1242, they fell upon Nablus. The
sack of the city, in which the congregational mosque was razed and even the
native Christian population, according to the Siyar, perished alongside the
Muslims, lasted for three days, after which the marauders retired to their
own territory—presumably to Jaffa, which had been their base on this
occasion.142 Ayyub was roused to assist his ally though ineffectively
as it turned out. A force of two thousand horse under Shams al-Din
Sirasunqur was sent to Gaza to rendezvous with the Egyptian frontier
detachments and with himself, and together they invested Jaffa. But
the Franks were well prepared, and after the siege had dragged on for some
time Ayyub sent orders to Sirasunqur to withdraw. had no choice
but to retire to Kerak, and the remaining Egyptian forces fell back on
Gaza.143 angry letter to Ibn 'Abd al-Salam al-Sulami, by this time
chief of Egypt, in which he described the Frankish outrages and
complained of the small support he had received from Ayyub, has survived
among his correspondence.144
During the past eighteen months the sultan had made a poor showing. He
had repeatedly prepared expeditions to enforce his sovereignty in Syria, and
on each occasion they had either failed to leave Egyptian territory or had
issued in disaster. The fiasco of Kamal al-Din's campaign in May-June
1241, when Ayyub's troops had been defeated by an army they
outnumbered ten to one,145 had been followed around the beginning of
1242 by preparations for another expedition, which was abandoned for
reasons unspecified.146 Then had occurred al-Hayjawis advance to Jaffa and
the humiliation of his defection to Isma'il. The sultan seems to have
relinquished yet another attempt to move against Damascus in the later
summer of 1242:147 on this occasion the reason was in all likelihood the
death at Gaza, on 12 August, of Kamal al-Din Ibn al-Shaykh, who is known
to have been in command of the invasion force.148 If we are to believe a
letter dated most probably in June or July of this year and addressed to the
amir of al-Gharb in the Lebanon, instructing him to make ready to link up
with the Egyptian forces, Ayyub may have intended to accompany this
expedition in person.149 Its abandonment represents the end of the sultan's
efforts to mount a campaign against his enemies in Syria while relying
principally on the Egyptian army. During the winter and spring of 1242–43
he engaged the Franks in fresh negotiations which lasted for six months or
more but which fell through as a result of his unwillingness to offer
sufficient concessions.150 Nor was Ayyub's standing redeemed by successes
elsewhere. In March or April 1242 his troops had finally been expelled
from Mecca by the ruler of the Yemen, a humiliation the sultan was never
able to avenge.151 The sole triumph registered by the Egyptian forces, in
fact, during the three years preceding La Forbie was a minor victory over a
Frankish naval attack on al-Warrada and east of the Nile delta, in May
1243.152
In the late spring of 1243 Isma'il headed his second campaign to attempt the
conquest of Egypt. Accompanied by of and joined en route by
a Frankish army, he advanced to Gaza, which had been abandoned once
more by the Egyptians.153 This time the allies had reached an understanding
with Disheartened, evidently, by the lack of support he had received
from the sultan, he now invested the Egyptian-held fortress of al-Shawbak
(Montréal) and was within an ace of taking it when the news of his allies'
sudden retreat from Gaza compelled him to retire himself to Kerak.154
Isma'il's campaign, like its predecessors, had accomplished nothing of
worth. The Annales de Terre Sainte remark irritably that he deceived his
Frankish confederates and neglected to swear a truce with them;155 and the
only result of the expedition was the elimination of al-Jawwad, to which
this may be a veiled reference. The fullest account is given by Ibn al-'Amid,
who says that Isma'il wrote to al-Jawwad reproaching him for his
comparatively lengthy stay among the Franks and inviting him to his own
encampment. The prince complied but subsequently, when the terms of the
new Frankish-Damascene truce were being formulated, wrote to the Franks
warning them against Isma'il. The letter fell into Isma'il's hands, and al-
Jawwad was arrested and sent under guard to Damascus.156 Here he died
mysteriously in March or April 1244. According to al-Khazraji, he fell ill
and died a natural death; but the remaining sources all repeat the rumour
that Isma'il had him strangled.157 So ended the career of' this turbulent and
bungling prince', as Blochet called him.158 An intriguer to the very last, he
had even hatched a conspiracy with al-Din Yaghmur, the officer whom
Isma'iI had detailed to arrest him and who was consequently thrown into
gaol at the same time.159 The extent of the Franks' complicity in al-
Jawwad's fate is difficult to assess. A later source alleges that Isma'il had
bribed them to surrender their ally,160 but this is not corroborated
elsewhere: the and Ibn al-'Amid, in fact, both depict the Franks as
concerned about al-Jawwad's disappearance and anxious to have him
back.161 Certainly they had better cause than anyone else to miss al-
Jawwad, whose attachment to them appears to have been the only constant
element in a series of highly volatile allegiances; and his removal from the
Frankish camp must have been a source of relief not merely to Isma'il but to
Ayyub also. It may even have been one factor underlying an extraordinary
reversal of alliances which was briefly in prospect in the early autumn.
The principal factor, however, was almost certainly the developments in the
north, where the Mongols had at last made a decisive thrust westwards into
the dominions of the Seljük sultan of Rum. In the first months of 1243,
alarmed by the Mongol capture of Erzerum, the Seljük sultan Kaykhusrau II
had persuaded the warring rulers of Aleppo, Mardin and Mayyafariqin,
together with the Khwarizmian bands operating in the Jazira, to compose
their differences and ally with him against the newcomers.162 But in June he
and an auxiliary force from Aleppo suffered a crushing defeat by the
Mongol general Baiju at Kösedagh, and a number of important fortresses,
including Sivas and Kayseri, were to fall over the next few months.163
Isma'il, who had caused the Friday prayers at Damascus to be recited in
Kaykhusrau's name since May 1241,164 was hardly impervious to these
events, which very probably occasioned his withdrawal from the Egyptian
frontier. Whether the ensuing negotiations were initiated by him or by
Ayyub, we are not told; but on 11 September 1243165 he replaced
Kaykhusrau's name in the by that of Ayyub, releasing also the sultan's
son al-Mughith from the Damascene prison where he had been held
since Isma'il's seizure of the city four years previously and preparing to
send him back to his father. For his part the sultan now at last recognized
his rival's authority in Damascus. Ayyub's suzerainty was similarly
acknowledged by of and by the ruler of Aleppo, so that the
coalition against him was now dissolved. In its place emerged a new
alliance aimed at the elimination of whose lands were to be divided
between Egypt and Damascus. Isma'il at once set about realizing this
arrangement by sending al- at the head of a Damascene army to
besiege fortress of 'Ajlun in the 166 He proved to have acted
with undue haste. For the settlement was suddenly wrecked by the
discovery that Ayyub had written secretly to the Khwarizmians urging them
to invade Syria and assuring them that he intended his alliance with Isma'il
to last only until his son was safely on his way to Cairo. Isma'il thereupon
returned al-Mughith to his cell, repudiated Ayyub's overlordship, and set
about reviving the coalition, to which also acceded once the
Damascene army had raised the siege of 'Ajlun. 167 During the following
months the Syrian princes negotiated a new truce with Acre, whereby the
concessions made in previous agreements were confirmed (with the
exception of Nablus and the Jordan valley) and Jerusalem was for the first
time surrendered to the Franks in its entirety.168 When Ibn passed
through Palestine around May 1244, he witnessed the allies' preparations
for yet another expedition to invade Egypt, with a division of Isma'il's army
in position at Gaza, near the Franks, and encamped to the west of the
Holy City.169
At first sight it appears puzzling that the Syrian Ayyubids should have
planned to invade Egypt when they must have known that a Khwarizmian
invasion was imminent. But given the notorious unreliability of these
freebooters, who had served many masters over the past ten years, there
was no reason to suppose that they would necessarily leave their present
field of operations in the Jazira and re-enter Ayyub's service. He had
abandoned them there in January 1239 upon coming south to take over
Damascus,170 and for a time they had admittedly continued to regard him as
their paymaster: when in the autumn of that year he took Ayyub into
custody, felt obliged to write to the Khwarizmians explaining that he
had acted only for Ayyub's own good and distracting them with the
recommendation that they attack Aleppo and 171 During the next three

years they were available to the highest bidder, latterly the prince of
Mayyafariqin, with whom they suffered a heavy defeat at the hands of
of and the army of Aleppo in August 1242. Then followed the
general reconciliation in the north at the instance of the sultan of Rum.172
Kaykhusrau's defeat in June 1243 by the Mongols, who had driven the
Khwarizmians from their homeland and with whom they doubtless had no
desire ever to cross swords again, decisively altered their situation; and the
sources implicitly link their return to Ayyub's service with this latest
Mongol thrust westwards. About this very time Ayyub's ally of
despatched an embassy to the Caliph and to a number of Jaziran
princes, among them the Khwarizmian leader Berke Khan. Ibn who
was a member of this mission, tells us that the Khwarizmians were once
more professing an allegiance to Ayyub and that the party discussed with
Berke how his forces might render the sultan assistance.173 It was perhaps
not long after this exchange that Ayyub himself wrote to the Khwarizmians
with a specific proposal: they were to rendezvous with his own troops in the
Jordan valley in preparation for an assault on Damascus, bringing with
them also a Kurdish group, the Qaymariyya, who were moving south from
Aleppo to and had likewise entered his service.174 It seems,
nevertheless, that they eventually set out in May or June 1244 primarily
because the Mongols had just launched a campaign against Aleppo and
were hard on their heels.175
Most probably, therefore, the suddenness of the Khwarizmian descent on
Syria took Ayyub's enemies by surprise. Sweeping down through the
territories of Aleppo and they divided their forces, one half, which
included the Qaymariyya, taking the Biqa' route and ravaging the county of
Tripoli, while the other half under Berke Khan himself made for Damascus
and its fertile plain, the Isma'il sent a force to intercept the first group,
but it was surrounded and practically annihilated at 'Ayn al-Jarr, south of
Ba'labakk. Accompanied by Ayyub's former general, the renegade al-
Hayjawi, he moved out of Damascus to meet the second Khwarizmian
army; but on discovering the size of the enemy they retreated hurriedly into
the city.176 In Palestine the news of the Khwarizmian advance caused panic.
Isma'il's troops at Gaza withdrew towards Damascus, while fled
from the neighbourhood of Jerusalem to the safety of Kerak, leaving the
Franks to resist the invaders unaided.177 With the final emergence of a
coalition against Ayyub which included the Franks and the whole of
Muslim Syria except together with the longdelayed advent of the
Khwarizmians, the forces had taken shape that were to join battle at La
Forbie on 17 October; and we must now consider the nature of the choice
confronting the Franks up to this point.
 V

It would be a mistake to assume that the Franks could afford to trust the
Syrian princes who had persistently sought their alliance. In the first place,
Isma'il and for all their hostility to the sultan, were at no time
impervious to developments further north, in Anatolia and the Jazira, which
might well distract them from commitments on the Egyptian frontier: a case
in point is probably the abandonment of the 1243 campaign, as we saw
above.178
Isma'il himself, moreover, was far from being a dependable ally. His
readiness to jettison his confederates emerges clearly from the short-lived
recognition of Ayyub's suzerainty and the alliance against in the late
summer of that year. Certainly, in assisting Isma'il in his campaigns the
Templars were taking a grave risk that was by no means offset by his
promises of a share of Egypt once it was conquered,179 for there was no
guarantee that he would not subsequently turn against the Latin Kingdom—
and with more formidable resources at his disposal than Ayyub had
possessed. In contrast to moreover, who could be described to
Joinville a few years later as 'one of the best knights that ever was among
the infidels',180 Isma'il personally seems to have held little appeal for the
Franks.181 That they had reservations about the Damascene connexion right
down to La Forbie is evident from Muslim accounts. When their allies
broke under the Egyptian-Khwarizmian attack, says Ibn the Franks
rounded in exasperation and cut down the Syrian troops as they fled.182
Following on three abortive attempts to invade Egypt, it must have seemed
the last in a whole series of betrayals.183 But some of the Franks' suspicions
were almost certainly baseless. In their letters to the West after the
Khwarizmian sack of Jerusalem, they were to complain of the tardiness of
their Muslim allies in answering their urgent requests for help.184 Yet this
was to ignore the fact that the Syrian Ayyubids, as we have seen, were the
first to bear the brunt of the Khwarizmian onslaught. Nor were the
Khwarizmians the only menace to be considered at this juncture, for a
Mongol division under the general Yasa'ur had advanced to the outskirts of
Aleppo in the early summer and had sent to Damascus and demanding
submission. Along with the ruler of Aleppo, Isma'il and appear to
have bribed the Mongols to withdraw; and by August 1244 Yasa'ur's troops
were encamped well to the north-east, in the plain of Mush.185 But like
Bohemond V of Antioch, who had similarly received an ultimatum from the
Mongols, the Muslim princes doubtless hesitated to come to the aid of the
Latin Kingdom out of fear that they would return.186 Had al-
and the Damascene army reached Acre a few weeks earlier than the
end of September to reinforce the Franks,187 the allies might possibly have
forestalled a junction between the Khwarizmians and Ayyub's own forces.
But we need not suspect them of deliberately holding back in order that the
Franks might somehow be weakened: the Mongols were a force nobody
could afford to treat lightly.
But whatever the flaws of the Damascene alliance, the Franks had still less
to hope for from Ayyub. Much of the criticism of Frankish foreign policy
by modern writers rests ultimately on the Emperor Frederick's
denunciations in the wake of La Forbie. His principal charge—that by
joining an overwhelming coalition against Ayyub the Syrian Franks had
driven him to summon the Khwarizmians to his assistance188—has already
been refuted in part by Marie-Luise Bulst, who pointed to the long-standing
ties between the sultan and his auxiliaries.189 Still stronger evidence of the
groundlessness of Frederick's allegations is provided by the diplomatic
manoeuvres we have noticed here. The sultan had no intention of tolerating
the status quo in Syria and never lost his determination to recover
Damascus. We have seen how he repeatedly planned invasions of Syria in
1241 and 1242; how he treacherously endeavoured to remove the Franks'
friend al-Jawwad; and how eventually he attempted to throw Isma'il and
off their guard with a spurious peace settlement at the very time he
was negotiating for Khwarizmian assistance. He was later to disclaim
responsibility for the Khwarizmians' outrages in Jerusalem;190 but there is
no doubt whatever that he had invited them in to attack his enemies. His
sole preoccupation at the time had been to deny the newcomers entry into
Egypt.191
To what extent the two rival factions among the Franks were in any case
influenced by the respective characters of the Muslim princes they
favoured, it is difficult to say. In all likelihood, their foreign policies were
further affected by the way in which they perceived their own local
territorial interests. We cannot view the Templars, for example, as anti-
Egyptian at any price. If their negotiations with Ayyub in the late summer
of 1241 were forced on them by the collapse of Isma'il's invasion plans,
they were still prepared to treat with Cairo from a position of relative
strength, through al-Jawwad, in the following spring, when a short-lived
truce did result, and yet again in the winter of 1242–43. It seems they were
averse not to an agreement with the sultan per se but to the inadequate
terms he offered. The evidence strongly suggests that the stumbling-block
was the city of Gaza, which had been a Templar possession until 1187. It
was excluded, as we noticed, from the territories listed in Richard of
Cornwall's truce with Ayyub and, together with Hebron and Nablus, was
among the localities whose surrender Ayyub refused to countenance in
1242–43, thereby putting an end to negotiations with the Temple and
driving the order to align itself definitively with his enemies. Significantly,
Gaza—alone of these three territories—was in fact surrendered according to
the acceptable terms which the Templars went on to obtain from Isma'il and
192As for the Hospital-
lers, their objections to the Damascene alliance are unclear and cannot be
explained simply on the grounds of the order's recent rapprochement with
Frederick II.193 The Hospitallers may have been closely involved in
Theobald's negotiations in 1239 with Ayyub's faithful ally of
who had been tributary to the Hospital for some years, 194 and
perhaps did not see their interests as served by the subsequent alliance with
enemies at Aleppo and their Master was certainly at war
with Aleppo in 1242. 195 On a wider scale, the aims and interests of the
Temple and the Hospital may have diverged sharply as regards northern
Syria, where the Templars had suffered a serious reverse at Trapesac
(Darbsak) in 1237 and were possibly discouraged from future chevauchées,
finding the south more attractive with the disintegration of al-Kamil's
empire after 1238.196 More than this we cannot deduce from the available
sources. No satisfactory explanation, again, has been offered for the
Hospitallers' readiness to fight alongside the Templars and their Syrian
allies at La Forbie. They may simply have rallied to the defence of the
kingdom against the Khwarizmians, who were a menace when viewed from
any standpoint; or they may have been as shocked by Ayyub's duplicity in
1243 as were his fellow-Muslims; or, finally, it is conceivable that they had
been given some inducement to bring them into the Damascene camp, since
Ibn expressly mentions their fortress of Kawkab (Belvoir) among the
places guaranteed to the Franks by Isma'il and in the final truce.197
The unprecedentedly fluid situation in the years following al-Kamil's death
may have appeared to furnish the Franks with undreamed-of opportunities
for expansion by diplomatic bargaining and military collaboration. But,
apart from the maverick element in the person of al-Jawwad, it yielded no
sure ally. Nor did it necessarily produce all—or even any—of the territory
specified in truces. This is most strikingly obvious when we focus on the
role of who currently held much of the land, including Nablus, Gaza
and the Holy City itself, on which the Franks had designs. The
consequences of this highly inconvenient circumstance can be seen to have
nullified more than one truce prior to 1243–44. In order to accommodate
in the alliance of 1240, the pro-Damascene party had to envisage the
surrender of Nablus and the sharing of Jerusalem. Again, as a means of
securing territory in southern Palestine at expense, the alliance of
Richard of Cornwall and the Hospitallers with Egypt in 1241 might
subsequently have borne fruit; but its value was somewhat diminished when
Ayyub was defeated by and made peace with him within a few
months. And soon after this hostility in turn contributed to the ruin
of the attempted invasion of Egypt by the rival party and its Damascene
allies. When Armand of Pierregort later claimed that his order had been
continually at war with 198 he exaggerated. The statement never-
theless implies a recognition that the prince of Kerak was a major obstacle
to any territorial settlement in Palestine. On balance he seems to have
preferred to ally with Ayyub, who on his deathbed in 1250 was to exonerate
for having been led astray by Isma'il and al-Jawwad.199 in
fact committed himself irrevocably to the sultan's enemies—as also to an
agreement granting the Franks full and exclusive possession of Jerusalem—
only when the exposure of Ayyub's treachery and the prospect of the
Khwarizmian invasion left him no choice.200 Until that point he, more than
anyone else, had been responsible for the constantly shifting diplomatic
balance. All the Franks could hope to do was to meet this fluid situation
with some measure of flexibility. What evidence we have suggests that the
Templars displayed more of that flexibility than their opponents in their
readiness to seek the best deal they could. It was their tragedy that there was
no deal to be had.
 THE SECRET HISTORY OF THE MONGOLS: SOME
FRESH REVELATIONS

As is well known, the Secret History is the only surviving source on the rise
of the Mongol empire produced by the Mongols themselves, yet
controversy continues to surround its value, its purpose, even its date. The
fullest and earliest attested version of the text does not even survive in the
Uyghur script employed by the Mongols, but only in a transcription into
Chinese characters, accompanied by Chinese translation; a transposition
carried out at an unkown date under circumstances which are not entirely
clear. Chinese sources, it is true, have been used to throw a certain amount
of light on the transmission of the Secret History, notably in a lengthy and
detailed article published forty years ago by William Hung,1 but as the
summary by F.W.Cleaves of the problems surrounding this evidence in the
introduction to his translation of the Secret History makes abundantly
clear,2 much has remained a matter for conjecture.
It is worth recalling, however, that Hung did not regard his study as
definitive,3 and that this line of research cannot be said to have reached a
dead end. We shall be considering here some hitherto unused evidence
which points forward to possible more precise conclusions concerning the
production and transmission of the Chinese version of the Secret History.
First, however, it is necessary to recapitulate the main conclusions of earlier
work. This has located the earliest reference to the use of the Secret History
by Chinese in a passage in the Veritable Records (Shih-lu ) of the Ming
apparently suggesting that it was drawn upon in the preparation of an
interpreter's manual, the Hua-i i-yü in 1382.4 Unfortunately, this
precise date marks only the point at which the Hung-wu Emperor ordered
the compilation of the handbook: all that can be affirmed for certain is that
this mention of a Yüan pi-shih must date to before the completion of
the emperor's Veritable Records in 1418.5
A further piece of evidence, however, places the summarized (i.e.
recapitulatory, non-interlineated) Chinese translation of the Yüan-ch'ao pi-
shih as the Secret History is known in Chinese, well before that
point, since a manuscript of this material originally in the Great Vault of the
Grand Secretariat (Nei-ko ta-k'u ) bears upon it the copyist's date
6
1404. It has been surmised that this manuscript copy was prepared in
connexion with the compilation of the massive Yung-lo ta-tien
encyclopedia carried out between 1403 and 1408. This incorporated the
transcribed text with interlinear translation, and the summarized translation:
it seems to have been the ultimate source of most versions of the Secret
History later produced in China, with the exception of a printed edition,
now surviving only in part, which appears to have been associated with the
printing (perhaps, it has been suggested, very early in the Ming) of the Hua-
i i-yü: the Yung-lo ta-tien itself, though frequently drawn upon by later
(especially eighteenth-century) scholars, was never printed.7
Later references to the Yüan-ch'ao pi-shih might, therefore, be seen as
reflecting knowledge of the work disseminated from those two points. The
next earliest reference known to date has been that in the Comprehensive
gazetteer of the Ming (Ta-Ming i-t'ung-chih ) of 1461,8 though
this is not a verbatim quotation but an allusion to the mythical origins of the
Mongol people given at the start of the work. It should be noted that the
wolf-ancestor is described not as 'blue-coloured', ts'ang-se-ti as in
the current text of the Secret History, but as 'blue-white', ts'ang-pai 9 I

adopt, here and below, the translations used by Cleaves, though there would
seem to be room for debate over them; 'whiteness', perhaps, could be taken
to hint at spectral pallor. At all events, the increasing amount of Chinese
evidence from 1461 onwards leaves only the very earliest stages of the
Secret History unattested, specifically the original version, probably but not
necessarily in Uyghur script; the version in which this was provided with
interlinear transcription and glosses; the first summarized version whence
the manuscript of 1404 was copied; and the first version to bring together
interlinear transcription and glosses (without the Uyghur script) and the
summarized translation before the Yüan-ch'ao pi-shih was incorporated in
this form in the Yung-lo ta-tien and the Ming printed edition.10
In all this a hitherto unused source compiled (to judge from its author's
preface) in 1481 gives some very important information. This is the History
through the ages rightly outlined, Shih-shih cheng-kang of
Ch'iu Chün (1420–95). Ch'iu was a prominent scholar-official and
writer who joined the Han-lin Academy in 1454 and worked for a while
with Shang Lu (1414–86), compiler of a chronicle style history of the
11
Yüan period. Ch'iu's own work is chiefly famous for its extremely
negative attitude towards Mongol rule;12 both men worked in the highly-
charged atmosphere following the T'u-mu incident, in which the capture of
China's emperor raised once more the spectre of Mongol conquest and
provoked an outburst of ethnocentric writing.13 Ch'iu apparently believed
that the Mongol rulers had so terrorized the Chinese historians among their
subjects that the existing record of the Yüan dynasty was not as black as it
should rightly have been painted; he raised this point in particular over the
monk Hsiang-mai and his account of Qubilai's destruction of the
Taoist Canon.14 Here we read a comment on the General essay on the
Comprehensive Mirror, T'ung-chien po-lun by the Ming prince
Chu Ch'üan (1378–1448):15
When I first got hold of the Po-lun and read it, and saw that it quoted the Yüan-ch'ao pi-shih as
saying that the first ancestor of the Yüan was the product of a union between a blue-white wolf and a
dull-white doe, I doubted it. After I entered the Grand Secretariat I saw the Yüan-ch'ao pi-shih
written using Mongol letters annotated at the side with Chinese words, and sure enough it had this
very story in it. Looking at things in the light of this, Hsiang-mai's confusion and fear must have been
real enough—how much the more so when the preface clearly states that the work (lun ) was
composed on imperial command; it must have had some basis.
The last sentence here is ambiguous: the modern editors of the Shih-shih
cheng-kang, presumably reflecting the 1563 edition on which they based
their work, leave a blank space before the word 'command', indicating that
their original took this to refer to a command from the Hung-wu Emperor to
his son Chu Ch'üan to write the T'ung-chien po-lun.16 My own feeling is
that this simply refers to Qubilai's command to Hsiang-mai noted in the
preface to his work, the Pien-wei lu 17 'it' would then refer to

Hsiang-mai's feelings when at the mercy of these monsters, not to the story
of the Mongol ancestors itself. Any listing of an edition of the Shih-shih
cheng-kang earlier than 1563 is very hard to find,18 and even were the first
of these editions to exhibit on inspection the same blank space, one would
not be sure that this reflected authorial intention. It should be noted,
however, that one citation of this entire passage, in a source describing itself
as a reprint, dated to 1569, from an editor who died in 1560 (and hence, one
presumes, dependent on an edition of the Shih-shih cheng-kang earlier than
that used by its modern editors) certainly does not exhibit this feature.
This citation occurs in the supplementary portion of a work whose short
title has been catalogued as Digest of the Comprehensive Mirror, T'ung-
chien chieh-yao in Cambridge University Library.19 The full title
is given as Hsin-k'an hsien-t'ai k'ao-cheng kang-mu tien-yin t'ung-chien
chieh-yao hui-ch'eng its
editorship is ascribed to the well-known writer T'ang Shun-chih
(1507–60). A work with a similar title is attested elsewhere as incorporating
Ch'iu's comments; a work of a similar type (but describing itself as a ta-
ch'üan a term which in the Cambridge text does replace chieh-yao in
some chapter headings), is ascribed elsewhere to T'ang;20 but this product
of the Ching-hsien t'ang in the publishing centre of Chien-yang is
not listed in any other bibliography I have consulted so far.
Textually, at any rate, this alternative source is completely identical with the
modern edition of the Shih-shih cheng-kang, except for an interrogative
particle at the end of the last clause, turning it into a rhetorical question.
Earlier in the text, too, the character for annotated, chu, is here written
rather than but otherwise the orthography is also identical. Both these
characters would seem only to refer to the interlineated glosses beside the
original script rather than the transcription into Chinese, but one cannot
infer from this an absence of transcription in the manuscript seen by Ch'iu:
in his short note any particular mention of the transcription would have
been an irrelevance. The term 'Mongol characters', for that matter, is also
tantalizing: the likelihood is that the Uyghur script is meant, but this, too,
cannot be absolutely certain.21
What is of great importance is the confirmation this passage provides of the
existence of a manuscript hitherto only put forward as a hypothetical source
for surviving versions of the Yüan-ch'ao pi-shih; the location it gives for
that manuscript precisely in the archives of the Grand Secretariat; and (less
precisely) a date which may be inferred of somewhere circa 1465 for its
presence there. For though in theory Ch'iu may have had access to the
manuscript for some reason early in his career in the Han-lin Academy, or
later when he came to compile his own Shih-shih cheng-kang (though his
mode of expression implies some lapse of time since his discovery), the
presumption must be that he came across it at a time when officially
engaged in historiographic duties in connexion with the Veritable Records
of the emperor Ying-tsung, a task to which he was assigned in 1464.22
That he saw fit to publicize his discovery in such precise terms, however,
raises another problem: why make such a point of his privileged access to
this document if the substance of what he had to say was already more
widely available in a printed version of the Yüan-ch'ao pi-shih? Though the
continued preservation of this manuscript consulted by Ch'iu may reassure
us that the printed edition could, after all, have been taken directly from this
early and reliable source, the date at which this was done may well have
been rather late, and not (as William Hung was inclined to believe) in the
early Ming at all. It may be objected that the reference to the origin of the
Mongols in the Ta-Ming i-t'ung chih does provide evidence of some wider
diffusion of the Secret History in the late fifteenth century. But it should be
noted that the description of the wolf-ancestor in this text as 'blue-white' is a
peculiarity shared with the T'ung-chien po-lun as quoted in the Shih-shih
cheng-kang (though the orthographic variant for the colour of the doe found
in this quotation is absent),23 suggesting that the gazetteer may be drawing
either on the T'ung-chien po-lun or some other intermediate source rather
than betraying any direct knowledge of the Secret History itself.
These minor linguistic variations do not, as far as it is possible to judge
from the quotation given, warrant any suspicion that Chu Ch'üan was
himself drawing upon a translation of the Secret History other than that
known today: much more probable is that the prince introduced his own
stylistic improvements into the unpolished Chinese of the original
translators. Given, however, that the date of the prince's work is as early as
1396,24 it provides very valuable confirmation that the translation had
indeed been effected by that time; William Hung's own supposition (which,
as we have seen, was subject to uncertainties over the compilation of the
Veritable Records) was that 1398 was the best terminus ad quem which
could be deduced from the evidence he had before him.
Why, then, not verify this quotation from the T'ung-chien po-lun against
one of the exemplars of this text which survive in East Asian libraries?25
Matters are perhaps not quite so simple. For what this initial small piece of
fresh evidence suggests is that a much larger task is actually required: Ming
historiography on the Mongol period may in general be of greater interest to
the student of Chinese intellectual history than to the Mongolist, but it
would seem that is does preserve occasional scraps of useful information,
and it would also seem that the task of hunting down this information has
not so far been carried out in any systematic way. But for this to be done
effectively will surely demand the skills of a specialist in the Ming period.
All that a student of an entirely different era of Chinese history can do is
raise one corner of this problem and invite the attention of those better
qualified to solve it.
T.H.BARRETT
School of Oriental and African Studies, London
 GHAZAN, ISLAM AND MONGOL TRADITION: A VIEW
FROM THE MAMLUK SULTANATE1

By REUVEN AMITAI-PREISS

Hebrew University of Jerusalem

The conversion of Ghazan Khan to Islam in A.H. 694/A.D. 1295 was an
event of great importance for both the Mongol ruling class and the Muslim
subjects of his kingdom. The story of this conversion, based primarily on
semi-official Persian works emanating from the Ilkhanid state itself, has
been retold and analysed in varying detail by several modern scholars.2
Recently, Dr. Charles Melville,3 using contemporary Arabic sources from
the Mamluk Sultanate, has enriched our knowledge of this event; in
addition, he has suggested that the Islamization of the Mongols may have
been well advanced even before Ghazan's conversion. Melville deals
mainly with Ghazan's conversion per se, as well as the events that led up to
it. As for the nature of Ghazan's Islam, he writes: 'It is beyond the scope of
this paper to speculate on the sincerity of Ghazan's conversion, which we
can never know, or on what he actually understood of Islam…'.4 He does
show, however, that Ghazan's conversion to Islam was more than just a
personal decision based on religious conviction: one motive behind this
move was a desire to attract those Mongols who had already become
Muslims, and thus to win their support in his struggle against Baidu.
An additional Mamluk source sheds some light on the workings of
Ghazan's mind, and the nature of his belief in Islam. We also learn that
Ghazan did not see any inherent contradiction between his new religion and
the traditions and laws of the Mongols. This source is al-Din Khalil b.
Aybeg (d. 764/1363), who provides entries on Ghazan in his two
major biographical dictionaries: the more comprehensive al-Wafi bi'l-
wafayat,5 and 6 a shorter work on his
contemporaries. Although most of the evidence on Ghazan is common to
both entries, the latter is somewhat briefer, presents the material in a
different order, and at times differs in detail from the parallel passages in
the former work.7
In this paper, based on the evidence from as well as information
from other sources, both Mamluk and Persian, I will first attempt to
examine Ghazan's commitment to Islam. I will then analyse in some detail
his continued adhesion to two areas of the Mongol imperial legacy: first, the
Yasa, or body of law attributed to Chinggis Khan; and second, allegiance to
the Qa'an or Great Khan. Subsequently, I will briefly touch upon two
additional areas of Mongol tradition: the Chinggisid imperial ideology and
the shamanistic religion. Finally, I shall attempt to draw together this
information in order to paint a portrait of the syncretic nature of Ghazan's
Islam, as well as to discuss briefly the possible implications for the history
of the Islamization of the Mongols in Iran.
initially deals with Ghazan's conversion to Islam in a perfunctory
way: his na'ib (viceroy) Nawruz was responsible for presenting this faith to
him in an attractive manner, and he converted in 694/1294–5; 'and with this,
Islam spread among the Mongols' (wa-fasha bi-dhalika al-islam bi'l-tatar).8
A more detailed—and unique—report about a subsequent crisis in Ghazan's
new faith is provided by further along in the biography. Citing as
his source al-Irbili 9 an immigrant to the Mamluk

Sultanate from the Ilkhanid realm, tells the story of Ghazan's

marriage to Bulughan Khatun,10 the widow of his father Arghun. Following
Rashid al-Din's laconic account, the outline of this story is already known:
after Ghazan's accession to the throne, he married Bulughan Khatun
according to the Muslim even though she had been the wife of his
11
father. Rashid al-Din only hints at the legalistic difficulties of such a
match, but they would have appeared to have been insurmountable, since
Islamic law expressly forbids the practice of marrying one's deceased
father's wives.12 gives a fuller account of this episode: having
become ruler, Ghazan married his father's wives 'in accordance to the Yasa
(law) of the Mongols in this [matter]' (
yasa al-mughul fi dhalika).13 He was particularly enamoured with Bulughan
Khatun, who had been Arghun's senior wife. When Ghazan became a
Muslim, he was told that Islam forbad marriage to one's deceased father's
wives. Finding himself in a legally impossible situation, Ghazan was intent
on abandoning Islam if this religion did not permit his marriage to
Bulughan Khatun. However, one of the 'ulama' offered a legal opinion
which provided a solution to this impasse: since Arghun had been a pagan,
his marriage to Bulughan Khatun was not legal, and therefore Ghazan could
now wed her with impunity. The Khan was happy with this suggestion,
married Bulughan Khatun (in a Muslim manner), and 'he adhered to Islam.
Without this [solution], he would have apostatized.'14 If this evidence is to
be believed, it would indicate that Ghazan's initial commitment to his new
religion was perhaps not very deep. I will return to the implication of this
evidence at the end of the paper.
Mention has been made of the Mongol Yasa or law,15 a subject to which
further reference is made in the biography. writes, evidently citing
his contemporary, the secretary and encyclopedist Ibn Allah 16

[Ghazan] spoke Mongolian and Turkish (al-mughuliyya wa-turkiyya), and he knew Persian,17 but he
did not speak it except with Khwaja Rashid [al-Din] and his like from among the close associates of
his court He understood most of what was said before him in Arabic, but he did
not let it be known that he understood it, out of pride in the deeply-rooted Chinggis Khani and pure
Mongol Yasa ( 'ala yasaq18 al-jinkiz khaniyya wa-mughuliyya ).
19
When he became king, he took up leadership [in] the way of Chinggis Khan and established the
Mongol Yasa [al-yasa al-mughuliyya]. He appointed judges (al-arghujiyya) to carry out the tribunals
(al-arghu).20 He obligated all to keep their rank and not to exceed it. The agha (older brother, i.e.
senior prince) was to be an agha, and the ini (younger, i.e. junior prince) was [to stay] an ini.21
Besides the interesting information on the languages which Ghazan
spoke,22 as well as an indication of his relationship to Rashid al-Din, the
passage contains significant evidence on Ghazan's commitment to the Yasa.
Corroborating evidence of this devotion is found in his imperial command
(yarligh [<Tur. yarli =Mon. ]), recorded by Rashid al-Din, ordering
the distribution of (assignments of land) to the Mongol soldiery.
Ghazan begins the yarligh by praising Chinggis Khan's Yasa, to which he
attributes his forefather's success in conquering the world.23 Ghazan's
commitment to the Yasa evidently preceded his ascension to the throne.
Elsewhere, Rashid al-Din states that Ghazan was already a firm supporter
of the Yasa in his younger days: as a little boy, he would gather his
companions, and teach them the Yasa (text: yasaq) as well as yusun (<Mon.
yosun 'custom').24 Furthermore, 'among them he would appoint the agha
and ini, [as well as] the anda (<Mon. anda "sworn brother")25 and quday
(<Mon. quda "in-law; family member by marriage").26 If anyone had
committed a transgression, then he would be eager to punish him, and
would chastise him according to the way of the Yasa (in different MSS:
yasa/yasaq)…'.27 Whatever the historicity of this passage, it is significant
that Rashid al-Din decided to include it in his work and present it in a
positive light. It must have been clear to him that in the mind of his patron,
devotion to the Yasa, even after his conversion to Islam, was considered a
laudatory thing.
It should be noted that there were fundamental contradictions between the
Yasa and the Muslim certainly in such areas as ritual purity and the
28
slaughter of animals. The disparity between the two codes is further
highlighted by statement on the precept authorizing (or
necessitating) marriage with the widows of one's father. Parenthetically, it
may be mentioned that does not provide detailed information on the
Yasa as a whole (as understood by Ghazan or himself), beyond this
statement, along with the additional remark that Ghazan, out of pride in the
Yasa, refrained from letting it be known that he understood Arabic and
Persian,29 preferring to communicate in public only in the languages of the
steppe nomads, Mongolian and Turkish (N.B. this is not a precept per se).
In spite of the vagueness as to the contents of the Yasa, it is clear, however,
that the intention here is to refer to some kind of corpus of law, i.e. the so-
called Great Yasa of Chinggis Khan, to which great respect is accorded.30
also gives us some evidence regarding the long-term fate of the
Yasa in the Ilkhanid state. He quotes who in turn cites the Mamluk
amir Sayf al-Din Aytamish noted as 'the most knowledgeable
person of his time regarding the affairs of the Mongols.'31 Aytamish states
that after Ghazan's death the 'Yasa of the Mongols (yasa al-mughul) passed
away.'32 Aytamish's comment on the fate of the Yasa brings to mind the
remark of the Mongol general Qutlugh-shah, upon witnessing a debate of
Muslim scholars, c. winter 707/1307–8: What is this that we have done, abandoning the
new Yasa (text: yasaq) and yosun of Chinggis Khan, and taking up the ancient religion of the Arabs,
which is divided into seventy-odd parts? The choice of either of these two rites would be a disgrace
and a dishonourable act, since in the one, marriage with a daughter is permitted and in the other,
relations with one 's mother or sister. We seek refuge in God from both of them! Let us return to the
Yasa and yosun of Chinggis Khan.33
Further evidence on the long-term fate of the Yasa in the Ilkhanid state is
given by who states that the Yasa was maintained in the
Chaghatayid Khanate as well as the realm of the Qa'an, as opposed to the
Ilkhanid kingdom and the Golden Horde.34
It would seem that in spite of Ghazan's attempts to enforce the Yasa as he
understood it, a lapse was perceived not long after his death by at least one
senior Mongol officer. It appears that within several years the had
gained the upper hand over the Yasa. This does not mean that there was a
deliberate wholesale jettisoning of the Yasa by the Khan or the Mongol
leadership, but rather a process, perhaps subconscious in part, during which
the precepts of Islam became more firmly rooted—at least in theory—
among the Mongol leading class. On the other hand, citing Dr. Morgan, if
Qutlughshah's monologue 'was at all typical, the Mongols had not as yet
acquired a very profound knowledge or understanding of the tenets of
Islam.'35
Another area of Mongol imperial culture which might have been influenced
by Ghazan's conversion to Islam was his relationship to the Qa'an/Great
Khan,36 to whom Hülegü and his successors had preserved a formal
allegiance. It should be remembered that just prior to Ghazan's conversion
and his accession to the throne, Qubilai Qa'an had died. Ghazan perhaps,
then, had a convenient opportunity to change the nature of his relations with
the theoretical leader of the Mongol world. In fact, and
unequivocally state in parallel passages that such a transformation indeed
took place. The longer passage in Wafi will be cited:
After this [time], Ghazan started called himself Khan (tasamma bi'lqaniyya), and37 had himself
mentioned in the (Friday sermon) and on coins38 without the name of the Great Khan (al-
qan al-akbar).39 He drove [the Great Khan's] representative (na'ib) from his country. None of his
forefathers and princes of his family had done this before Ghazan. Rather Hülegü and all those who
came after him had lowered themselves to the position of viceroy (na' ib) of the Great Khan. None of
them were called Khan, but rather they were called Sultan so-and-so (fulan). The striking of coins
(sikka) and the was in the name of the Great Khan, not them. If one of them (i.e. Hülegü, etc.)
was mentioned by name, it was in a subservient fashion. Yet they were the kings of the country; they
had the right to collect taxes, and to appoint and dismiss people.
The Great Khan had among them a representative; [in theory] they would enact orders after
consulting with him. In fact, they did not consult him. When Ghazan drove out [the representative of
the Great Khan] and became absolute ruler, he was censured for his. He said: 'I did not take the
kingship [with the help] of Chinggis Khan, or anyone else. I took it only with my sword.' No one
dared answer him. He gained absolute control of the Khanate (al-qaniyya). Those who came after
him followed him [in this manner] until the end of time (i.e., the end of the Ilkhanid state). The Great
Khan could do nothing to dispute this except by words.40
I said: For this reason, there is no mention of 'Hulakuhi (sic),' 'Abaghawi,' 'Arghuni' gold [coins], but
only 'Ghazani,' 'Khurbandi'41 and 'Bu ' gold [coins], since Ghazan is the first in this country
who had the striking of coins [in his] name; those who [came] after him followed him [in this
manner].42
misleads us when he states that Ghazan's adoption of the title khan
was without precedent, since Hülegü had used this title on his coins until
the appearance of ilkhan (usually translated as 'subject khan') c. 658/1260.43
What perhaps meant to say is that in general Ghazan had dropped
from his coins (and elsewhere perhaps) the title ilkhan, which may well
have signified his subservient role vis-à-vis the Qa'an, and had returned to
the perhaps more ambiguous khan. (Interestingly enough, in these entries,
only applies ilkhan to Ghazan once, when listing his titles and
appellations at the beginning of the accounts.) In any case, is
generally correct in his description of Ghazan's changing protocol. Unlike
Hülegü, Abagha and Arghun, Ghazan had dropped the mention of the
unnamed Qa'an (qa'an from the Arabic legends of the coins minted
in his realm.44 This would explain the remark at the end of the third
paragraph, that before Ghazan, none of the Ilkhans had gold coins (i.e.
dinars) named after them, since their coins were minted under the auspices
of the (unnamed) Qa'an.
How much this represented a radical change in Ghazan's relationship
towards the Qa'an is unclear. Certainly, claim that Ghazan drove out
the representative of the Qa'an is untrue. This representative was Bolad
Aqa, who had arrived in the Ilkhanate in 1285, during Arghun's reign
(1284–91) and remained in his position until his death in 1313.45 There is
evidence that some relations were maintained between the Qa'an and
Ghazan. The Qa'an Temür Öljeitü (1294–1307) named Ghazan as 'the
prince who establishes peace in the Western Lands'.46 There is also
evidence of Ghazan sending an embassy to Temür Öljeitü in 1298, in part to
collect dues owed to Ghazan from the manufacture of certain goods in
China.47 It would seem then that Ghazan, in spite of the change in his
protocol and his conversion to Islam, maintained at least a formal allegiance
to the Qa'an. The information related by and regarding the
totality of Ghazan 'declaration of independence' is thus somewhat
misleading.
Two additional points related to this passage need clarification. First,
writes that previous to Ghazan, the Hülegüid rulers had used the
title Actually, this title is only occasionally found in Ilkhanid protocol
before Ghazan's reign. It is not even found in the beginning of
Tegüder's letter to Qalawun, although the Mongol ruler had become a
Muslim.48
The second point refers to the second paragraph. Ghazan is reported to have
said: 'I did not take the kingship [with the help] of Chinggis Khan, or
anyone else. I took it only with my sword.' Although it may appear
otherwise at first glance, this claim is not denigrating Chinggis Khan.
Rather, Ghazan appears to be saying that Chinggis Khan is irrelevant to his
own claim to the kingship vis-à-vis other Mongol princes: in a sense, their
descent from Chinggis Khan gave them all an equal claim to kingship.
What bestowed upon Ghazan the right to rule was that he had defeated his
rival 'with [his] sword'.49 In passing it might be mentioned that this
particular passage is reminiscent of a passage from Ghazan, when
countering Jochid claims to Azerbaijan presented to him by a embassy from
the Golden Horde, is reported to have said: 'I only took the kingship by the
sword, and not by inheritance. Among what I took and gained by the sword
were Tabriz and Maragha. Between you and me only the sword will decide
regarding them.50 This specific version is similar to the one found in Ibn
Although it is possible that Ghazan used the same expression on two
different occasions, it is perhaps more likely that one of the authors took a
catchy phrase and applied it at an appropriate time.
It would seem, then, that although Ghazan's relationship towards the Qa'an
had undergone some changes, formally he still recognized the latter's
suzerainty. Whether the change was occasioned by his conversion to Islam,
reflected Ghazan's view of the reality of the Mongol world-system, or was
merely exploiting the death of Qubilai Khan to improve his own status,
remains an open question. What can be said is that in spite of his
conversion, here Ghazan remained basically loyal to another tenet of the
Mongol imperial legacy.
I will briefly touch upon two other areas which are relevant to the present
discussion. The first is Ghazan's continued commitment to the Mongol
imperial ideology. Briefly, this is the belief that Chinggis Khan had received
a mandate from heaven to conquer the world and place it under his control;
this mandate was to be continued and completed by his successors.51
Elsewhere I have attempted to demonstrate that Ghazan, in spite of his
conversion to Islam, did not in fact eschew a belief in this ideology.52 Here
I shall mention just two points: first, there is the Ghazan's lauding of
Chinggis Khan's conquest of the world, which was cited earlier in this
paper. Secondly, there are Ghazan's extremely belligerent letters to the
Mamluk Sultan, in which he is called upon to surrender or face war and
destruction. In spite of the Islamic terms in which these letters are couched,
and even the Islamic rationalizations contained therein, what we have here
is the traditional Mongol demand to surrender unconditionally or face the
consequences.53
The final area of Mongol tradition to which Ghazan's commitment will be
examined is that of the Mongol folk religion, what is usually called
shamanism. It should be noted that this is the only realm of Mongol
tradition discussed in this paper whose existence clearly precedes the
establishment of the Mongol Empire. Frankly, it should be admitted that the
evidence for Ghazan's continued practice of pagan and shamanistic rituals is
very sparse. Essentially there is only one piece of evidence, provided by
Rashid al-Din: in 1302, Ghazan participated in a traditional Mongol ritual
of hanging cloth streamers to a tree and dancing around it, interestingly
enough after having prayed in a Muslim manner and delivered a speech in a
decidedly Islamic tenor to the assembled officers and ladies. This tree had
been picked out because Ghazan had spent a night by it during a earlier, and
difficult, stage of his career. Any doubts regarding the traditional Mongol
nature of this ritual should be allayed by information provided in this same
passage: such a ritual had been performed in Mongolia by Qutula Khan, an
ancestor of Chinggis Khan, in order to fulfil an oath to his 'ancient god'
(khuday-i qadim) i.e. Tengri of the traditional Turco-Mongolian religion.54
It may be noted that Rashid al-Din thought it politic to relate this story
without comment. There is also evidence that even in the post-Ghazan
period, when obedience to and concern with the Yasa was apparently in
decline, shamanistic rituals were still maintained among the Mongol élite of
the Ilkhanid state.55
One conclusion which can be drawn from the above discussion is that
Ghazan's Islam was a syncretist faith: having converted, he maintained a
belief in various aspects of Mongol custom and tradition, much of which
explicitly contradicted the precepts of his new religion. I would suggest that
this syncretism also characterized the Islam of the Mongols as a whole,
certainly of their élite. This is indicated by the above-mentioned
participation of the Mongol officers and ladies in a pagan ritual. Only
afterwards—gradually and perhaps never fully—were the elements of
Mongolian tradition weeded out. We learn, then, that the conversion of the
Mongols, with Ghazan at their head, fits into the pattern of 'communal
conversion', where—in the words of Professor Nehemia Levtzion—'Islam
was adopted by ethnic groups in their own milieu, while maintaining their
own cultural identity. There was hardly a break with past traditions, and
pre-Islamic customs and beliefs survived.'56
The syncretic nature of the Islam of the newly converted Mongols was also
perceived by a contemporary observer in Mamluk Syria, the theologian Ibn
Taymiyya (d. 728/1328), who condemned them accordingly. In a fatwa
(formal legal opinion), issued probably just before Ghazan's third offensive
against Syria in 702/1303, Ibn Taymiyya cast aspersions on the sincerity
and quality of the Islam of the Mongols. Among the reasons adduced were
that many of the Mongols do not follow the commands of the Muslim
(such as pilgrimage, prayer and fasting), but rather the Law of
Chinggis Khan, referred to not as Yasa, but with the Islamic terms sunna
and . In addition, they put Chinggis Khan on a par with
and pay obeisance to him. They maintain their belief in conquering the
world: submission to them is more important than Islam.57 While it is true
that Ibn Taymiyya was writing anti-Mongol propaganda, he cannot be
accused of inventing slanders against Ghazan to satisfy his Mamluk
masters; Ibn Taymiyya consistently took an independent line and often went
to prison rather than recant.58
It is noteworthy that Ibn Taymiyya's opinion was shared by none other than
Ghazan's brother and successor, Öljeitü, who said of his brother in a letter
to the Mamluk Sultan: 'He was a Muslim on the outside, but an infidel on
the inside.'59 It is admittedly difficult, however, to decide how much of this
statement was a result of Öljeitü's desire to put some distance between him
and his brother, in the interest of diplomacy, and how much was brought
about by a deeply held conviction regarding the quality of the latter's Islam.
Not only was Ghazan's Islam syncretic but following it appears that
this Islam was (initially at least) not very deeply held. This is the conclusion
to be drawn from the story of Bulughan Khatun. In the light of this
evidence, Rashid al-Din's account of the affair would seem to be a cover-up.
There is no reason to reject story. Certainly the latter was writing
long after the end of the Mamluk-Ilkhanid war and the demise of the
Ilkhanate. The Islamization of the Mongols (or their Turkified descendants)
was a fact. Thus there was no compelling political reason for him to cast
aspersions upon the Islam of the Mongol rulers. The result of this story—
especially when taken together with additional evidence presented above—
is that Ghazan's reputation as a devout, orthodox Muslim is somewhat
tarnished. We are left, then, with an inconsistent, even confused, but
certainly more historically convincing convert to Islam.
 Marco Polo and his 'Travels'1

PETER JACKSON

Keele University
The year 1998 marks the seven-hundredth anniversary of the initial
composition of the book associated with Marco Polo, Le devisament dou
monde. As the first European to claim that he had been to China and back
(not to mention that he had travelled extensively elsewhere in Asia), Polo
has become a household name. He has been credited with the introduction
of noodles into Italy and of spaghetti into China. With perhaps greater
warrant, he has been cited as an authority on—inter alia—the capital of the
Mongol Great Khan Qubilai, on the Mongol postal relay system, on the
trade in horses across the Arabian Sea, and on political conditions on the
north-west frontier of India in the mid thirteenth century. The Marco Polo
bibliography published in 1986 contained over 2,300 items in European
languages alone.2
But Marco Polo's reliability has been a matter of dispute from the
beginning. It has recently been proposed that the incredulity he met with on
his return to Venice sprang from an unwillingness to accept his depiction of
a highly organized and hospitable Mongol empire that ran counter to the
traditional Western Christian view of the 'barbarian' and especially the view
of the barbarian Mongols that had obtained since the 1240s.3 Polo has also
met with scepticism from modern commentators. A few years ago, the
approach of the rather fine book by Dr John Critchley was that the Polo
account is a more valuable source for the minds of late thirteenth- and
fourteenth-century Western Europeans than for contemporary conditions in
Asia. For Critchley, therefore, the question of the authenticity of the Polo
material is very much a secondary consideration.4 More recently, Dr
Frances Wood has queried whether Polo was ever in China. She concludes
that the famous Venetian probably never got much further than
Constantinople or the Black Sea.5 The argument tends to be based (1) on
omissions which would supposedly not have been made by anyone who had
genuinely visited the country: Polo's failure to mention foot-binding, tea-
drinking, or the Great Wall, for instance; (2) on the fact that Polo's name has
so far not come to light in any Chinese source; and (3) on what can only be
regarded as deliberate falsehood, such as the alleged participation of the
Polos in the siege and capture of a Chinese city which is known to have
been over one year prior to their arrival. Of these objections,
© School of Oriental and African Studies, University of London 1998
the failure to mention the Great Wall carries little weight, given that we can
be fairly certain it had not yet been built: walls there certainly were, but not
the continuous and impressive structure we see today, which apparently
dates from the sixteenth century, the era of the Ming dynasty.6
In fact, the authenticity of Polo's stay in 'Cathay' was first challenged years
ago, partly for such reasons as these but also on the grounds that the
Chinese section contains remarkably little in the way of personal
reminiscence and that the accounts of Chinese cities are frequently vague
(not to say bland) and hardly compare with the vivid descriptions of life in
the Mongolian steppe.7 Indeed, one could find further grounds for
challenging Polo's firsthand familiarity with the Middle Kingdom: that the
book neglects, for instance, to mention finger-printing, a technique with a
long history in China.8 It seems to me, however, that to consider the visit to
China in isolation is to set about it the wrong way: we need, rather, to take
the work as a whole. In this paper I want to address the following questions.
What is the book we associate with Polo's name? With what purpose was it
written? What claims does it make for itself? To what extent does it purport
to represent Polo's own experiences? Just where did Polo go? This last
question is particularly central to my paper.
 Asia in the era of Marco Polo

First, it is necessary to put the travels in context.9 The voyages of the three
Venetians, Marco Polo, his father Niccolò and his uncle Maffeo, date from
an era when much of Asia lay under the rule of the Mongols; although even
as the elder Polos set off on their first journey in the early 1260s the unitary
Mongol empire was dissolving into a number of rival khanates, of which
those of the Golden Horde (in the steppes of southern Russia) and of Persia
were closest to the territories of the Catholic West. Only the Mongol rulers
of Persia, the so-called Il-khans, acknowledged the Great Khan (qaghan)
Qubilai, whose dominions lay in the east and who was able to compensate
himself for the hostility of many of his relatives by completing the conquest
of southern China in 1279. For all the book's protestations, the mighty ruler
of Cathay immortalized by Polo (and later by Coleridge) was in fact the
first qaghan not recognized throughout the Mongol empire.10
The subjection of much of Asia under a single government had greatly
facilitated the opportunities for both merchants and missionaries to travel
from western Europe across the continent, opportunities which were not
appreciably reduced by the empire's disintegration into a number of
constituent states.11 In the eastern Mediterranean, Italian and other Latin
merchants were active in ports like Ayas (Ajaccio) in the kingdom of Lesser
Armenia, lying at the terminus of one of the overland trade routes through
the Mongol empire.
From the mid thirteenth century merchants from the great Italian
commercial cities,Venice, Genoa and Pisa, were beginning to travel at least
in Persia and the lands of the Golden Horde.12 The appearance of rival
Mongol khanates further gave rise to promising diplomatic contacts. After
the Muslim khan of the Golden Horde reached an understanding with the
Mamluk government at Cairo in 1262, negotiations (ultimately fruitless)
began between the Pope and various Western monarchs on the one hand and
the Il-khans on the other regarding the possibility of military collaboration
against Egypt, as the principal bastion of Muslim power.13 But the shadows
were already closing in on the Latin states in Syria and Palestine. When
Marco Polo accompanied his father and uncle on their second journey in
1271 the great port of Acre was still in Christian hands; but by the time the
Polos came back, the fragile Western settlements had been overwhelmed by
the Egyptians (1291).
 Authors and copyists

Who wrote the book? There has been widespread agreement that the
original language was a form of Old French strongly influenced by Italian.
The style is consonant with the story given in the Prologue to what is
possibly the earliest surviving MS (the Paris MS fr. 1116, known as F), that
Polo dictated his experiences in a Genoese prison in 1298 to a fellow-
captive, the Pisan romance-writer Rusticello.14 But other versions, in other
Western languages, were already being made in the early years of the
fourteenth century. It has been proposed that Rusticello had a hand only in
the production of one version and that subsequently Polo had other co-
authors.15 One hundred and twenty MSS survive in total. Many contain
material not found in others. It seems that F itself is the result of
abridgement, and hence that some of these other versions represent MS
traditions which are in fact older than F; in other words, that F is not the
closest in content to the original.16 The most important traditions, apart
from F, are: T, MSS of a Tuscan version, known as l'Ottimo, 'the Best',
made by Niccolò degli Ormanni, who died as early as 1309;17 P, a Latin
translation made by the Dominican Friar Francesco Pipino of Bologna from
a text in the Venetian dialect, at some time between 1310 and 1314 (and
now represented by the largest single group of MSS); Z, another Latin
version (but quite independent of P), represented primarily by a Toledo MS
of the fifteenth century; and R, the MS used by Ramusio in the mid
sixteenth century as the basis for his printed edition and now lost (the
edition contains a great many, though not all, of the passages otherwise
found only in Z, as well as passages not found in any other extant version).
Many phrases in different MSS may reflect embellishments and accretions
due to particular copyists, working in some cases very soon afterwards but
in others, perhaps, up to a century and a half later. But the discovery of the
fifteenth-century MS Z in Toledo in 1932 revolutionized scholarly thinking
on the subject: the fact that so many passages hitherto found only in
Ramusio's edition were encountered also in Z obviously tended to make the
Ramusio text appear far more dependable.18 And since much of the material
found in Z, but not present in F, would have been too interesting simply to
have been omitted, it is conceivable that these earlier accretions represent
supplementary oral information from Marco Polo himself.19 This had
happened with two previous visitors to the Mongols, both Franciscan friars,
the papal ambassador John of Plano Carpini and the missionary William of
Rubruck. Carpini, returning to the West in 1247, had been in great demand
as a dinner guest, and we know at least that the Italian Salimbene de Adam
obtained further information from him which is not found in his report.20
Rubruck, an unofficial visitor to the Mongol empire, was nevertheless
contacted in Paris a few years after his return by the English Franciscan
Roger Bacon, who exploited the opportunity to check particular details in
the Flemish friar's Itinerarium before incorporating them in his own work.21
If I have spent so long on the issue of Polo MS traditions, it is in order to
make two important and related points at the outset. First, the book—in any
of the forms that have come down to us—is not by Marco Polo. We simply
cannot be certain what was in the work originally drafted by Rusticello on
the basis of Polo's reminiscences in a Genoese prison. Even if we possessed
that original, Polo's own perspective on late thirteenth-century Asia would
be refracted for us through the prism of Rusticello's prose. And secondly,
this means that we cannot afford to lay too much stress on matters that the
book does not mention. Given the kind of material found only in Z, for
instance, but omitted in other texts because some copyist did not find it
sufficiently interesting, we are hardly in a position to claim that Polo was
never in China because he failed to refer to foot-binding or tea-drinking.
They might have been mentioned in some MS (or group of MSS) now lost.
(In fact, it has been overlooked that Z does mention the fact that Chinese
women take very small steps, but gives a somewhat arcane explanation for
it (I, 305), on which I do not propose to expatiate.)
 Corroborative material

What other information is available to supplement the details furnished by

the book about the Polos? How do we know that they actually travelled
anywhere at all? Apart from the information supplied in the MSS
themselves, sources for Polo's experiences are few. There are some that are
near-contemporary, such as the Imago Mundi of Jacopo d'Acqui, which
dates from the fourteenth century. It is d'Acqui who tells us that after his
return Polo was captured in a sea battle with the Genoese in 1296 off Ayas
in Lesser Armenia, and that in his final illness he was urged to excise
passages that were exaggerated and
incredible; Polo allegedly replied that he had not told half of what he had
seen.22 The physician and philosopher Pietro di Abano (d. 1316) claims to
have met Polo, 'the most extensive traveller and the most diligent inquirer
that I have ever known'.23 Pipino claims in his preface that he spoke to
those who had known Marco and also that the latter's father and uncle had
vouched for his veracity.24 It is also worth noticing that according to one
fourteenth-century French MS a version from which it was copied had been
presented by Polo himself to Thibaut de Cepoy, visiting Venice as the agent
of the Capetian prince Charles of Valois in 1307.25 Lastly, documents from
the Polo family have survived. Marco's own will (dated 1324) is less
informative than that of his uncle Maffeo (1310), which refers to 'the three
tablets of gold which were from the magnificent Chan of the Tartars' (a term
that could apply either to the qaghan, to the Il-khan or to the khan of the
Golden Horde); it is presumably one of these which is listed in an inventory
of Marco's property drawn up in 1366, during one of the numerous disputes
among his extremely litigious kinsfolk.26 What became of the tablets
thereafter is, regrettably, unknown.
The fullest source outside the book itself is Ramusio's introduction to his
sixteenth-century edition; but the details found here have to be treated with
caution, since we do not know their provenance (possibly in some cases
genuine traditions preserved at Venice over the previous 250 years) and in a
number of instances Ramusio is demonstrably wrong. On the other hand,
much of the additional material in his text has an authentic ring and is
difficult to account for if it did not in fact emanate from someone who had
visited the Far East (the edition, it should be noted, dates from 1553, some
years before Europeans again began to establish themselves in China). But
what we are to make of the claim, found only in Ramusio's introduction,
that Polo sent home to Venice from his Genoese prison and asked his father
to forward his notes,27 and that he profited from the assistance of a noble
Genoese in writing the book,28 is anybody's guess.
 The aim of the book

With what purpose was the book written? The result, we must presume, of
its having been written by a professional romance-writer is that the style of
the work is heavily formulaic. Of several Chinese cities we learn little more
than that the people are idolators, subject to the Great Khan, use paper
money and live by trade and industry. Particularly towards the end there are
set battle-scenes, in which identical phrases occur with remorseless
regularity: men and horses are slain in profusion, severed arms and legs lie
strewn about, and the din is so great that 'you could not hear God
thundering' (a phrase encountered half a dozen times in F). All the stock-in-
trade of medieval French writers is
naturally imported: 'And what shall I say?…' 'Why make a long story of it?'
(this sometimes a few pages too late.) The book is also rambling and
discursive, at times irritatingly so: 'But I will go on to tell you also a marvel
which I had forgotten to tell' (I, 188; cf. also 216); 'again I will tell you a
thing which I had forgotten…' (I, 244; cf. also 277, 407). Even more
evocative of a thoroughly disordered mind are the abrupt and maddening
changes of direction:
Now since we have told you of these Tartars of the Levant then we will leave them for you and will
turn again to tell about the Great Turquie [i.e. Turkestan] so as you will be able to hear clearly. But it
is truth that we have told you above all the facts of the Great Turquie…and so we have nothing more
to tell of it. So we will leave it and will tell you… (I, 469).
Or still worse:
Now we will leave this and will tell you of the Greater Sea [i.e. the Black Sea]. Yet it is true that
there are many merchants and many people who know it; but there are also plenty more of such as do
not know it and for such as these one does well to put it in writing. And we will do so… [There
follow three lines of text about the mouth of the Black Sea; following which] And after we had begun
about the Greater Sea then we repented of it, of putting it in writing, because many people know it
clearly. And therefore we will leave it then, and will begin about other things… (I, 477).
There can be no unanimity regarding the purpose for which the book was
produced. It may be that Marco Polo conceived of writing a merchant's
handbook in the strict sense—a by no means improbable aim for a member
of an Italian merchant family. The various texts do contain references to
products and their prices, sometimes in Venetian values,29 and the spices
that are not imported into Europe are plainly of no interest.30 But Marco
does not emerge from the book in the guise of a merchant.31 If a merchant's
handbook was ever the aim, it was submerged beneath the priorities of
Rusticello and other copyists. The Z text is content to say that Polo whiled
away his enforced leisure hours in prison by compiling the work 'for the
enjoyment of readers' (ad consolationem legentium). There is a fuller
statement in F:
He says to himself that it would be too great evil if he did not cause all the great wonders which he
saw and which he heard for truth to be put in writing so that the other people who did not see them
nor know may know them by this book (I, 73).
The same theme recurs later, at the beginning of the section on India:
which are indeed things to make known to those who do not know them, for there are many
wonderful things which are not in all the rest of the world, and for this reason it does well and is very
good and profitable to put in writing in our book (I, 353).
So far, then, a concern for the transmission of mirabilia. But Jacques Heers
has drawn attention to the emphasis on the excellence of the Christian faith
and on its triumph.32
It is clear that even in Polo's lifetime the value of the book varied with the
translator or copyist and the era. The fact that a French ambassador, then in
Venice to organize a crusade against Byzantium,33 asked for and received a
copy from Polo himself suggests that it might, even at this early date, have
acquired an interest for would-be crusaders contemplating the Mongol
alliance. And it is worth noticing that some MS copies are found bound up
with crusade treatises or related matter.34
Similarly, for some copyists the information it included that was especially
relevant pertained to the religious beliefs of the various peoples it surveyed.
For that reason the Z scribe frequently noted in the margin adorant ydola,
and the Dominican Pipino, who had composed his Latin translation for the
'reverend fathers' of his Order and replaced the F prologue with one of his
own, lays great stress on the salvation of souls: whether it is the case that
those reading of the marvels of creation in Polo's book will be led to
wonder all the more at the power and wisdom of God, or that the hearts of
'some devoted to religion' will be stimulated to carry the Gospel to 'the
blinded nations of the infidels, where the harvest truly is great but the
labourers are few'.35 The difference of approach is sometimes starkly in
evidence, as when Polo is speaking of marriage customs in Tibet. The basic
text is found in F: 'And there is such a custom of marrying women as I shall
tell you: it is true that no man would take a maiden for wife for anything in
the world…' [unless she has first lain with many men]. In a mid fifteenth-
century Venetian MS, this sentence begins: 'And there is such a pleasing
custom of marrying women'; for Pipino, some generations earlier, it had to
be 'such a custom of marrying women as I shall tell you, an absurd and
most detestable abuse coming from the blindness of idolatry…' (I, 269–70).
Such preoccupations rendered it by no means incongruous for all three
Polos to appear in friars' garb in the illustrations to certain MSS.36 In one
case, indeed, material from the itinerary of the fourteenth-century
Franciscan traveller Odoric of Pordenone is inserted at intervals in an
abridged Polo text.37
The Prologue purports to furnish a framework for the second and main part
of the book, 'a description of the diverse parts of the world'.38 The phrase is
revealing, and has been too often overlooked. Although material is often
introduced by the first person ('I Marco'), the tone is more frequently
impersonal: 'When one leaves this city, one travels…' 'One finds…' In fact,
it is generally unclear whether the Polos' own travels are the sole source for
the information given; the origin of the information is usually left
unspecified. The book is therefore emphatically not a narrative of the Polos'
travels, of the sort that we find, for instance, in the reports of Carpini and
Rubruck.39 If it were an itinerary, the order of places followed in southern
Persia would be bizarre in the extreme. And it is important to note that the
treatment is, if anything, more impersonal in the sections on Persia and
Central Asia than in those on China. In western Asia, Polo virtually
parachutes into a few localities—Sawa and a neighbouring village (I, 113–
16), where he picked up stories about the
Magi; Hurmuz (I, 123–6); a plain in Kirman where he narrowly escaped
capture by the Qara'una Mongols (I, 122); and Badakhshan, where he fell ill
(I, 138, R only):40 there is little sense of an itinerary. By contrast, the points
at which the reader is most strongly under the impression of following in
the footsteps of an individual traveller occur in various journeys within
'Cathay': there is no comparison here with the highly improbable
description given of China in the 1340s by the Moroccan pilgrim Ibn
(demonstrably an authentic traveller as far as India).41
As a whole, however, the Polo book represents an attempt to set out an
encyclopedic survey of the different parts of the world 'in order'. The phrase
'in order' recurs extremely frequently, but the order is manifestly not that of
any particular journey made in the past: the writer and reader, in Critchley's
winsome phrase, 'travel through the book together'42—and it might be
added that they frequently turn aside to places that lie off this imaginary
route. The only chronological framework is to be found in the prologue,
which tells of the departure of Maffeo and Niccolò Polo from
Constantinople in 1260 (1250 in all the manuscripts), recounts their return
to Venice and their second departure, this time with Niccolò's 15-year-old
son Marco, and ends with the three travellers' homeward journey by way of
Persia, as_ambassadors from Qubilai escorting the imperial princess
Kökechin to the Il-khan Arghun, in the early 1290s. And even here there is
no intimation of route other than a brief allusion to Java.
 Personal observation or hearsay?

What claims does the book make for itself and for the Polos? The reader is
at intervals assured that the contents are authentic. 'I shall bind myself for
certain not to say more of it than is according to the truth' (I, 177, from VB).
This statement is found in one of the Venetian MSS, which perhaps
understandably take some interest in trophies that they claim Marco brought
back to Venice, like a specimen of the hair of the wild ox from Ergiuul
(Erji'ül, i.e. Liangchou)43 and the dried head and hooves of a musk-deer (I,
179, from VB). When on the island of 'Lesser Java' (Sumatra), he obtained
some brazil seeds which the travellers took back and planted at Venice,
though 'it did not grow there at all' (I, 376), and when in Maabar (Ma'bar;
the Coromandel coast of southern India) he took some earth from the place
where St Thomas had been martyred and was able to heal many with it back
home in Venice (I, 398). In other contexts it is merely claimed that Marco
witnessed something, as for example that he saw the head of a gigantic fish
in an idol-temple at Quinsai and heard several times the count of the annual
revenues of that city (I, 341, Z only; I, 342); and Ramusio's text adds that he
saw an estimate of the customs dues there (I, 340). According to one
Venetian MS, he had the opportunity to measure a wing-feather of the ruc
which was brought to Qubilai's court (I, 431, from VB).
Marvels, perhaps; but there is remarkably little of the fabulous. Admittedly,
we are treated to an account of men with tails inhabiting the kingdom of
Lambri (I, 376). But even Carpini, a sober diplomat with a brief to inform
the pope of all he had seen in the Mongol world, had mixed in with it well-
worn topoi with a pedigree that went back to the Alexander Romance: tales
about dog-headed people, a race with no heads at all, those with only one
leg who propelled themselves along by cartwheeling, and so on. By
comparison the Polo account is remarkably restrained, at times even
rigorous.44 Great care is taken not to claim that Polo saw the ruc (I, 430–1).
The book seeks to put the record straight regarding the salamander
(asbestos) which is not consumed by fire: not an animal, as hitherto
believed, but a cloth manufactured out of a vein in the earth (I, 156).45 It is
also concerned to explain tales about the unicorn (in this case, evidently a
rhinoceros) and the trade in what purported to be pygmy corpses (actually
the dried and decorated carcases of monkeys) on Sumatra (I, 372). There is
the attempt—by this time obligatory for all European travellers in Asia—to
locate Prester John, and the Polo account is an interesting and not
unintelligent variation on previous themes (I, 181–3).46 There are
improbable stories, certainly: a long tale about the Christian inhabitants of a
village near Baghdad, oppressed by an evil Caliph, who were able by dint
of prayer to induce a mountain to move as promised in the Gospel (I, 105–
12); or the shoemaker of Baghdad who, after lusting after a beautiful
woman who entered his shop, put out his own eye because it was better to
enter Heaven with one eye than to go to Hell with two (I, 108–9); or the
relation of the death of the last caliph of Baghdad, whom the Mongols
allegedly left to starve amidst the treasure he had accumulated instead of
spending it on his army (I, 102–3). Such tales are met with in other sources,
and Polo (and indeed anyone else) could have picked them up in eastern
Christian circles when passing through Persia or Iraq on the way home.47
And it has been suggested that he could have heard a tale about a miracle
involving a church at Samarqand that was threatened by the venom of local
Muslims (I, 144–6) from a Nestorian prelate domiciled at Chên-chiang-fu,
Mar Sargis, whom he mentions elsewhere (I, 323) and who is known to
have originated in Samarqand.48 Generally speaking, the accent, in other
words, is on the edifying and the moralistic, rather than on the zoologically
preposterous.
The book is concerned to detach hearsay from personal experience. At the
outset we find an assurance to this effect:
But there are some things which he did not see, but he heard them from men fit to be cited and of
truth. And therefore we shall put the things seen for seen, and the heard for heard, so that our book
may be right and truthful, with no falsehood… (I, 73).
Although it cannot be said that this laudable aim is adhered to consistently,
it nevertheless does resurface at intervals:
But do not believe that we have treated of the whole province of Catai [north China] in order, nor
indeed of a twentieth part; but only as I Marco
used to cross through the province, so the cities which are on the way across are described (I, 309, Z
only).
Regarding Mangi (Man-tze, i.e. south China) we read in F: 'We have not
told you of the nine provinces ["kingdoms"] of Mangi but of three: these are
Yangiu, Quinsai and Fugiu…'. The Z text adds: 'Of these three, however,
we have told you this in order because Master Marco made his passage
through them, for his way was directed thither. But of the other six also he
heard and learned many things…; but because he was not in any of them as
he was in Quinsai…and because he did not travel over them he would not
have been able to tell so fully as about the others, so we will leave them
aside' (I, 353). Similarly, when we reach the island of 'Lesser Java', to F's
statement that it contains eight kingdoms a fourteenth-century Latin MS
adds: 'of which I Marco was in six, namely in the kingdoms of Ferlec,
Basman, Sumatra, Dagroian, Lambri and Fansur, but I was not in the other
two' (I, 371); and subsequently we find: 'And we will tell you nothing of the
other kingdoms on the other side [i.e. on the southern coasts of Sumatra]
because we were not there at all' (I, 377). The lengthy section on the 'Old
Man of the Mountain' (the head of the sect of the Assassins) is prefaced
with the words: 'I Master Marco Polo heard it told by several men…' (I,
129). The account of the salamander/asbestos in 'Ghinghin Talas' is credited
to a 'companion, named Çulficar, a Turk…, who stayed three years in that
province for the Great Khan to have that salamander brought out'; but, the
author adds, 'I saw them myself' (I, 156–7). At one point, in the account of
the city of Quenlinfu (Chien-ning fu),49 Ramusio's text inserts the words,'
and I was told, though I did not see…' (I, 346).
In the vast majority of cases where there is a personal note, incidentally, it
is Marco's experiences that are transmitted, and not those of his father and
uncle. Exceptions are the reference to all three Polos staying for a year in
the city of Campçio (Kan-chou) 'for their business which is not worth
mention' (I, 160),50 and the discovery, during a visit by Marco and his uncle
to the city of Fugiu, of a strange sect whom they identified as Christians (I,
349, Z only), although modern commentators have seen here an isolated
pocket of Manichaeans.51 But generally speaking the elder Polos are
eclipsed: what they saw on their first journey across Asia to Qubilai's court,
the Prologue tells us, Marco also saw (later), 'and so he will tell you clearly
in the book below' (I, 77).52
So where precisely are we told that Marco Polo went? If we take the book
at face value, Marco himself was employed on numerous missions by the
qaghan, on which he was under orders to write interesting reports, and
these were the means whereby he was so well informed about the world (I,
87). The suggestion has been made that this helps to explain the tone of
much of the information, given the formulaic character of Chinese reports
of this kind—and, incidentally, of those of Venetian ambassadors.53 Some
of this imperial
business seems to have involved land journeys through China. Thus Marco
was sent as messenger to Qarajang, i.e. Yün-nan (I, 86); and he also
travelled westwards for four months' journey from Qubilai's capital,
Khanbaligh, and Pul-i Sangin ('the bridge over the Sang-kan')54 lay en route
(I, 255). Elsewhere we are told that 'from Sindufu [Sheng-tu] one sets out
and rides quite seventy days' journey through provinces and through lands
[in] which we have been and have written them in our book above' (I, 300).
But there are other references to sea voyages on Qubilai's behalf, entailing
visits to the great Chinese ports to which the book devotes so much
attention. Considerable stress is laid upon Marco's experience of India (i.e.
south and south-east Asia in general):
Moreover, I tell you quite truly that Master Marco Polo stays there in India so long and knows so
much of them, of their affairs and of their customs and of their trade, that there was scarcely a man
who would know better how to tell the truth about them… (I, 354).55
In fact a careful reading reveals that an account of India is practically as
vital to the book's purposes as is the account of China: 'our book was not
yet filled with that which we wish to write there,' the reader is assured, 'for
there were wanting all the doings of the Indians' (I, 353). Marco had
allegedly just returned from India when the qaghan's envoys to the Il-khan
Arghun ascertained that the land route across Asia was unsafe, i.e. in about
1290 (I, 89; embellished in some other MSS); which was why he and his
father and uncle were able to latch onto the embassy when it was decided to
go by sea to Persia instead.
The book claims that Polo was in six out of the eight kingdoms on 'Lesser
Java', as we have seen, and was marooned in one of these, 'Sumatra', by
adverse weather for five months (I, 373). It was presumably by sea also that
he visited Çiampa (i.e. part of present-day Vietnam: I, 368), whose king, he
learned, had fathered 326 children. We cannot always be sure whether a
particular visit—that to the island of Lesser Java, for instance—occurred in
the course of the long return journey or formed part of some earlier official
mission; and the various MS traditions render it difficult for us to know, in
some cases, because they contradict one another. Thus a Venetian MS adds
that Polo was part of an embassy which Qubilai sent to Ceylon to ask its
king for an enormous ruby and that he saw the ruby with his own eyes,
though the mission was a failure (I, 380); whereas Ramusio's text specifies
that Polo stopped off in Ceylon on the way home (I, 407). The two
statements are not mutually exclusive, of course. Similarly, the visit to
Maabar, on which Polo says he witnessed its king being pursued for debt
and which is mentioned without embellishment by Z, is alleged in
Ramusio's text to have fallen during the return voyage (I, 389). The halt at
the great port of Tana in Gujarat (I, 421), on the other hand, is most likely to
have occurred just prior to the Polos' arrival in Persia.
 Europeans in Mongol Asia

How plausible, then, is the book as a whole? The notion that Italian
merchants might travel from Europe across the breadth of Asia, and come
back to tell
the tale, is not as far-fetched to us, of course, as it would perhaps have
appeared to the citizens of Venice around 1300. It seems that the arrival of
Europeans of some sort (Fu-lang, i.e. 'Franks') at Qubilai's court is recorded
in the Chinese annals as early as 1261, though the obscure details associated
with them, which include a reference to the Land of the Midnight Sun,
make it unlikely that they were Italians: more probably they hailed from
Scandinavia or from some northern Russian city like Novgorod.56 The
Italian presence in the Far East (as opposed to Persia) was more a feature of
the fourteenth century, and even then the Venetians appear to have been
outnumbered by their rivals the Genoese, who already by c. 1320 had a
reputation for unrivalled daring and curiosity: Polo learned that they had
recently ventured onto the Caspian Sea (I, 99).57 But we do have evidence
that Venetians had got as far as China at least by the time the Polos
embarked on their return voyage. The Franciscan missionary John of
Montecorvino—later, in 1307, to become the first Latin archbishop of
Khanbaligh—speaks in his third letter of a Venetian merchant, Pietro da
Lucalongo, who had travelled out with him to China from Tabriz in 1291
and purchased for him the land on which he had built his church.58 The
remarkable thing about the Polos' two journeys to the qaghan's court is that
they occurred some years before an Italian presence in the Far East is
documented.
What we know of Yüan China from other sources—notably the Yüan Shih,
the dynastic history compiled after the fall of the Mongol regime in 1368,
but from contemporary records—serves to make much of the detail of the
account of Polo's activities there rather convincing. Even the mention of
asbestos deposits appears apt, since Qubilai's finance minister Ahmad had
in 1267 submitted a memorial to the throne in which he advocated their
proper exploitation.59 Nor were far-flung missions on the qaghan's behalf
uncommon either by land or by sea. Qubilai is known to have despatched a
party to explore the sources of the Yellow River in 1281,60 and the Mongol
government's concern to revive trade with the Indian subcontinent and the
islands of the eastern archipelago is apparent from 1278 onwards. Yüan
embassies visited in 1280, Ceylon, and Kawlam (Quilon) in
1281, Ceylon again in 1282, Kawlam in 1283, and in 1285, 1287
(with Ceylon) and 1290. The purpose was not always simply an exchange
of goods. The 1282 mission appears to be that mentioned in the Polo texts
(though with the year 1281 or, in one MS tradition, 1284), since one of its
goals was to inspect or secure the alms bowl and mortal remains of the
Buddha (identified by Polo as 'Adam':
I, 411); and the envoys in 1290 were under orders to bring back men of
learning and interpreters.61
The Polo texts do not claim that the Venetians participated in all of the
embassies mentioned. Thus in the account of the mission sent to Ceylon in
1281 to ask its king for relics of Adam the book makes no mention of
Marco or his father and uncle. Admittedly Qubilai is not known, from
Chinese sources or indeed any others apart from the Polo book, to have
employed expatriate Europeans (and Chinese sources are remarkably
unhelpful in this regard, given their tendency to lump together everyone
from the West, whether Muslims or Central Asian Buddhists, as Hsi-yü,
'Westerners', or Se-mu, 'people of diverse nations').62 On the other hand,
Qubilai's relatives the Il-khans certainly employed Europeans. Such men
are found as early as the 1260s acting as interpreters and envoys on behalf
of Hülegü and Abaqa.63 Ghazan, who eventually married the princess
Kökechin and himself became Il-khan in 1295, had in his service around
1300 a Pisan called Isolo; and he and his father Arghun were represented on
diplomatic missions to Western Europe in 1289–91 and in 1302 by a
Genoese named Buscarello di Ghisolfi.64
It may be that we can begin to explain the frustrating nature of much of the
detail on China. In the first place, Marco spent a significant proportion of
his 17 years in the qaghan's service travelling abroad, possibly in the main
to the ports of southern India. And secondly, both then and during his stay
in China itself, he would have associated largely with non-Chinese: this
would account for the absence of references to tea-drinking. As a foreigner
in the imperial service, he would have been employed as part of a deliberate
policy of reducing dependence on the native Chinese.65 Nor would he have
been required to learn the Chinese language. That he did not seems clear
from his error in interpreting the Chinese title of the Mongol general Bayan
('Cingsang'= 'minister') to mean 'a hundred eyes' (I, 310, 311);66
although this does, incidentally, look like just the kind of mistake that could
only have been made by someone who had visited China. Competence in
other languages was at a premium in the Yüan dominions, as it had been
throughout the Mongol world since the beginning.67 The book does not
specify which were the 'languages and four letters and writings' that Marco
Polo
learned (I, 86). Two of them were almost certainly Persian and Turkish (the
languages of most of the non-Mongol foreigners employed in the
administration), and it has been suggested that the other two were both
Mongolian, but written in two distinct alphabets, the older Uighur script and
the new phagspa script introduced by Qubilai's regime in 1269 and
borrowed from Tibet.68 There is nothing surprising about the fact that the
information in the book has in large measure a Persian slant: similarly,
merchants who told the Master of the Temple in Cyprus some time before
1308 about the great Chinese port of (Hang-chou) employed an
Arabic-Persian form ('Hansa') not too remote from the 'Quinsai' of the Polo
account.69 Persian was by now a lingua franca throughout much of the
Mongol empire,70 and was doubtless the language Marco knew best. He
was an alien who was surely thrown together with other aliens, and it is
through the eyes of aliens that we see late thirteenth-century China in the
book.
 The status of the Polos

It is not so much the main portion of the book that fails to withstand
scrutiny as certain of the details supplied in the Prologue, specifically
regarding the status enjoyed by the Polos. Let us examine the story, which
begins with a commercial expedition by Maffeo and Niccolò from
Constantinople to the lands of the Golden Horde. Here they allegedly found
themselves unable to retrace their steps owing to the war that had broken
out between the khan, Berke, and his southern neighbour, the Il-khan
Hülegü (a war known to have begun in 1261–62), and so travelled east into
Central Asia, where they met an envoy from Hülegü to Qubilai's court and
were persuaded to accompany him to the qaghan (I, 74–7). As Pelliot
pointed out some decades ago, the war between Berke and Hülegü in the
Caucasus region hardly prevented the Venetians from returning via the
Pontic steppes, and the real reason must have been the problems that
developed between Berke and the Byzantine Emperor Michael VIII
Palaeologus and which led to a Mongol invasion of Thrace at some point in
the early 1260s.71
Qubilai sent back the Polo brothers with a gift of some asbestos cloth (I,
157–8) and with the Mongol noble 'Cogatai', as his ambassadors to the
Pope, who was asked to send 100 Christian missionaries; and he instructed
the Venetians in addition to obtain for him oil from the Holy Sepulchre at
Jerusalem (I, 78–9; cf. also I, 201–2). What became of the asbestos cloth,
we are not told; but both the other details of their commission are strikingly
reminiscent of episodes documented elsewhere. When the Nestorian prelate
Mar Yaballaha and his companion Rabban Sauma travelled from China to
Persia in c. 1275, we are told that they went on Qubilai's behalf and that the
qaghan had given them garments with which they were to touch the
Sepulchre after dipping them in the Jordan.72 And again, a few years later,
in 1278, we
learn from a letter of Pope Nicholas III to Qubilai that the qaghan had
asked his short-lived predecessor, John XXI, for the despatch of
missionaries.73 So there is nothing particularly odd about Qubilai's
commission to the Polo brothers; although equally, it raises the possibility
that Rusticello inserted the details about Jerusalem in the light of Rabban
Sauma's own relatively recent visit to western Europe as the ambassador of
the Il-khan Arghun: it is worth noting that the Nestorian had spent the
winter of 1287–88 in Genoa, where his arrival had created a great stir.74
'Cogatai' having fallen ill en route and been left behind, the Polo brothers
continued on to Acre, but were unable to prosecute their business because
their arrival there (probably in April 1269) fell during a three-year vacancy
in the Holy See: this fits well, since no pope was elected for three years
following the death of Clement IV in 1268. On the advice of the papal
legate at Acre, they waited until the election of a successor and occupied
themselves with a visit to their home in Venice. After two years, with the
conclave still undecided, they determined to return to Qubilai before it was
'too late'. Leaving Venice once more, this time with Niccolò's young son
Marco, they stopped off at Acre, where they consulted the legate, Tedaldo
Visconti, archdeacon of Liège. He furnished them with a letter to Qubilai
certifying that they had been prevented from fulfilling their mission by the
vacancy in the Holy See. But they had got no further than Ayas when they
learned that the legate had himself been elected Pope as Gregory X; and
soon afterwards the new Pope himself ordered them to return. The Polos
thus became part of a diplomatic mission from Gregory to Qubilai. Initially
they were accompanied by two Dominican Friars, named as Niccolò da
Vicenza and William of Tripoli. The friars took fright, however, at the news
of a devastating invasion of Armenia by the Egyptian Sultan Baybars,
handed over their letters to the Polos, and returned with the Master of the
Templars. The Polos pressed on, and after a three and a half year journey
reached the qaghan's court (I, 80–4).
M.H.Laurent, who examined the account of the early journeys in the
Prologue in some depth, concluded that the details can all be harmonized
with known conditions, except that the book makes two different clerics
into one: the legate on the first occasion was William of Agen, who died in
April 1270, and not Tedaldo Visconti, who was in Palestine only for
something like 12 months, from the autumn of 1270 until he left Acre for
Italy in November 1271.75 Contrary to the Polo prologue, it does not seem
that Tedaldo was ever papal legate. He is described merely as a pilgrim in
the Estoire de Eracles, which mentions his arrival in Palestine, and in the
dedication of William of Tripoli's De Statu Sarracenorum;76 and it is
difficult in any case to see how he could have been appointed on the death
of William of Agen in 1270, given that there had been no pope for two
years.
Laurent's otherwise admirably meticulous study left two details unexplored.
One relates to the Egyptian invasion. Only a single Armenian source
mentions this campaign, for the good reason that it never materialized.
News reached the capital, Sis, that Baybars was on his way north, and this
caused some alarm; King Leo III was able to deflect him with a placatory
embassy.77 As we learn from Arabic chroniclers, which completely neglect
to mention Armenia in this context, the whole affair had been a false alarm.
At Damascus Baybars received word of advance of a Mongol army from
Ilkhanid Persia in October 1271. From he sent ahead two
expeditionary forces which penetrated as far as and Edessa ( )
respectively, but he himself advanced no further than Aleppo before
withdrawing southwards in response to an attack on Caco (Qaqun) by the
Lord Edward, then on crusade in Palestine. The sultan was back in Cairo by
December.78 We could surmise that the two friars deserted the mission on
hearing the news of the Mongol attack, which at this early stage might have
seemed at least as threatening. But the Mongol advance was in response to
an appeal from the Lord Edward, and one early recension of the Polo book
has Pope Gregory sending the Venetians and the two friars specifically to
secure aid from the Il-khan Abaqa for a future crusade (I, 83).79 On these
grounds Soranzo as long ago as 1930 suggested that Gregory's embassy was
in fact directed to Abaqa. This would rather undermine the need for any
pusillanimous reaction on the part of the Dominicans.80
The second detail relates to the letters which the pope entrusted to the friars.
To write to the qaghan when merely archdeacon of Liège on pilgrimage in
the Holy Land is one thing; it is quite another to do so immediately after
being elected pontiff. We might have hoped that Gregory, mindful of the
importance his correspondence had now acquired, kept a copy to be
inserted in the registers after his arrival in Italy. No such copy has
survived.81 Arguments from silence are always hazardous. But it is odd that
Gregory seemingly failed to recall his letter to Qubilai at the time of the
Second Council of Lyons in 1274, and that, when Pope Nicholas III came to
write to the qaghan on 4 April 1278 (the letter we noticed above), he made
no reference to a correspondence with Gregory: he mentioned only the fact
that Qubilai had sent word to Pope John XXI, via the Il-khan Abaqa, asking
for baptism and for the despatch of missionaries (i.e. in 1276–77).82
It will be observed that on two occasions the Polos are said to have been
part of an embassy but deprived of the company of the official envoys en
route. The Mongol noble 'Cogatai' abandoned the elder Polos on their way
back from Qubilai in the 1260s (we are not told whether or not they still had
a Mongol escort), and the Dominicans selected to head the mission from
Gregory X in 1271 deserted the three Venetians at Ayas on learning of the
northward advance of an Egyptian army. 'Cogatai' is so far unidentified.83
But William of Tripoli appears to be identical with the homonymous
Dominican author of the De Statu Sarracenorum, written at Acre and
dedicated to none other than the future Gregory X: there is no corroboration
of his appointment as an ambassador to the Mongols. To be asked to believe
that two members of a Mendicant Order—men who usually emerge from
the sources as intrepid and conscientious observers of their Rule—panicked
and aborted their mission on the rumours of the Egyptian advance places
the reader's credulity under something of a strain. And taken together with
the disappearance of 'Cogatai' a few years earlier, it leaves the nagging
impression that the Polos had a habit of losing their fellow travellers en
route. It begins to look as if Rusticello (with or without Polo) cobbled
together an apparently plausible background for the outward journey,
Perhaps it was not the first occasion that the Pisan writer had invented
prestigious connections of this kind: the claim in one of his other works to
have been lent an Arthurian romance by the Lord Edward is decidedly
suspect.84 Traders described as mercatores et homines Abagacham were
among those who at Ayas at this very time, in October 1271, received
satisfaction for losses suffered at the hands of the Genoese;85 and it is
possible that the Polos met this party in Armenia and travelled back with
them to Persia. But in any case, it was not unknown for merchants to travel
with friars, though in no sense forming part of their mission: the Venetian
Pietro da Lucalongo, as we have seen, accompanied Montecorvino from
Tabriz to China in 1291.86
Regarding the return voyage in the early 1290s, we seem to be on firmer
ground. The Polos were selected to accompany a mission taking an imperial
princess, Kökechin, from China to Persia as a bride for the Il-khan Arghun:
she was to replace the queen Bulughan, who had recently died. The
authenticity of the mission from the qaghan which brought the princess
Kökechin is confirmed by both Chinese and Persian sources. We know that
the three ambassadors named in the Polo account—the Mongol nobles
Uladai, Abushqa and Qocha—received orders from Qubilai in April–May
1290 to prepare for their departure. The Ilkhanid historian Rashid al-Din
describes the arrival of the embassy, headed apparently by Qocha, in Persia
in c. 1292–93: Arghun having himself died in the interim, they were
received by his son Ghazan, who took delivery of the princess and sent a
share of the qaghan's gifts to his uncle, the Il-khan Gaykhatu.87 The
mention of Qocha in particular appears to corroborate the statement in the
Polo account that he alone of the ambassadors survived the long sea
voyage. Although the Europeans are not mentioned in either of these
eastern accounts, the Polo prologue here seems to do a more
convincing job of linking their return from China with an episode that
actually occurred. Even the detail supplied at one point elsewhere in
Ramusio's text that Marco was in the plain of Hurmuz when the king of
Kirman headed an unsuccessful campaign to capture the city slots
conveniently into the chronology of the return journey, since the expedition
in question can safely be dated to the winter of the Hijri year 691/early
1292.88
But were the Polos part of this embassy to Persia, or did they merely sail in
the same fleet? And what became of the mission with which Qubilai is now
said to have charged them—to the pope and the kings of France and Spain,
among others (I, 90)—but of which we hear nothing further? One problem
here is the insistence of Chinese imperial etiquette that foreign merchants
be treated as tribute missions from subject peoples;89 while merchants were
employed as official agents representing the commercial interests of the
Mongol qaghans.90 The distinction between commercial and diplomatic
missions had long been difficult to make in the world of the steppe,91 and
visiting traders themselves could accordingly be forgiven for concluding, in
error, that their functions had been extended to embrace diplomacy. As we
have seen, there is no evidence for the employment of Europeans on official
diplomatic missions by the qaghan, although such evidence does exist for
Mongol Persia.
The inclusion of the Polos in official embassies, however, looks
suspiciously like part of a wider tendency to magnify their role in the east.
Great stress is laid upon the affection in which the Polos were held. Qubilai
is said to have loved them so much that he repeatedly withheld permission
for their departure; and when he finally gave them leave, it was with great
reluctance (I, 88, 89). So, too, Princess Kökechin regarded each of the Polos
as a father, and 'there was nothing she would not do for them'; she is
supposed to have wept when they left Persia for Europe (I, 92–3). Stuff of
this sort is of a piece with the emphasis on the high esteem which Marco
especially, according to the Prologue, enjoyed at the qaghan's court and
with his alleged capacity to draft more beguiling reports than anyone else (I,
85–7).92 But it should be noted that the more specific manifestations of the
Polos' indispensability are less problematic: their participation, for instance,
in the siege of Saianfu (Hsiang-yang: I, 317–20), which is impossible, given
the date (1273), when they could not have yet reached the Far East, and
where Chinese sources ascribe their role instead to Muslim mangonel
experts;93 and Marco's appointment for three years as governor of Yangiu
(Yang-chou; I, 316), which Pelliot sought to explain away with the proposal
that the Venetian simply had charge of the government salt monopoly in the
city.94 That these—the two most implausible
claims the book makes for the Polos—are not found in certain important
groups of MSS, namely Z, those based on a fourteenth-century Latin text,
and some of those in Venetian dialect, is often forgotten; they may well
represent interpolations of a later date.95 It is a curious coincidence that a
Venetian merchant family is known to have been domiciled in Yang-chou
by 1342, when one of them was commemorated with an impressive Latin
tombstone.96
It is not implausible that the Polos, like other Westerners (this time,
Muslims and Central Asian Turks) we read of in the Chinese and Persian
sources for Qubilai's reign,97 were genuinely employed by the qaghan on
commercial business and accompanied diplomatic missions. Rather, it is, at
the very least, their role as ambassadors which is a fabrication. Despite
Olschki's determination to accept that they served Qubilai in this capacity,98
they may well have belonged to a species with which the Mongol empire
had for some decades been familiar. A generation or so earlier, the
Franciscan William of Rubruck had assured Louis IX of France that bogus
envoys 'scurry about all over the world'. Rubruck told the story of one such
impostor called Theodolus who obtained permission to travel to the Papal
Curia (at some point prior to 1254). He was initially accompanied by a
Mongol envoy, who fell ill, however, and died at Nicaea when the party was
detained by the emperor John Vatatzes; Theodolus was thrown into
prison.99 That there were still attractions in Polo's era in passing oneself off
as an official envoy emerges from a letter written to Edward I of England
by two ambassadors from Mongol Persia in 1276. The ambassadors warn
the king against a couple of Catalans (probably merchants) who, in
company with a Nestorian Christian, had been sent by the Il-khan Abaqa to
purchase gerfalcons in Norway; instead they were travelling further south
and pretending to be Abaqa's envoys to the various courts of Catholic
Europe (the purpose, of course, would have been to obtain free gerfalcons
as gifts and to pocket the Il-khan's money).100 It was only natural,
moreover, for expatriate Westerners to inflate their own consequence when
they could safely do so hundreds of miles away. The Pisan Isolo carries off
the prize in this context, because he seemingly managed to exaggerate his
standing at two completely different courts. He convinced the Ilkhanid
statesman and chronicler, Rashid al-Din, that he was one of the rulers of
Pisa, so that he was described in just those terms when Rashid al-Din came
to write the section of his great historical encyclopedia entitled 'History of
the Franks';101 and when at the Papal Curia in 1301 Isolo claimed to be
Ghazan's 'vicar' for Syria and the Holy Land, which the Il-khan had recently
(and briefly) conquered, although it is more likely that he was merely
deputed to superintend the resettlement of Western colonists in the
region.102 We cannot discount the
probability that the Polos had embarked on a similar—but, in literary terms,
a more successful—imposture.
 Conclusion

The book associated with Marco Polo's name is a description of the known
world rather than the memoirs or itinerary of the traveller himself; and this,
together with an extremely complex and obscure MS tradition, means that
we need not attach too much significance to matters that are omitted. As
regards the areas the Venetians visited, the book itself makes claims that are
not particularly extraordinary, and demonstrates, moreover, a readiness to
distance itself from the outrageous. On several matters, such as the
diplomatic and commercial contacts between Yüan China and southern
Asia, the Polo texts are so well informed that it is difficult to see how
Rusticello might have come by the information without an Italian who had
spent time in the Far East. Marco Polo may not have travelled extensively
in China, and seems to have been employed at least as much on
commissions that entailed lengthy journeys by sea between China and
peninsular India. In any case, apart from the two itineraries within China
which bear the stamp of personal experience, the book is in large measure
only loosely arranged around places which it assures us Marco or all three
Polos visited; and that the Venetians were not in all the places mentioned is
made quite explicit. We have, lastly, to distinguish where the book says the
Polos went from the claims it makes (or some MS traditions make) for their
credentials, and to confine our scepticism to the latter. The fact that Marco
Polo or his co-author or later copyists exaggerated his importance while in
China or on the voyage from China to Persia has long been suspected and
can hardly be in doubt. But it does not in itself demonstrate that he was
never in China or, worse still, never east of the Crimea.
 INDEX

Note on Index Page Hyperlinks

This Index retains the "Print Book Page Numbers" as links to embedded targets within the content.
Navigating from a "Page Number" link will take you to within three Reader "Page Forward" clicks of
the original Index reference point.
This strategy retains the full value of the academic Index, and presents the relative positions and
distribution of Index references within this book. The Index Page Numbers are hyperlink pointers
and have no relationship to the Reader soft-generated page numbers.

Abano, Pietro di 267

Abaqa 153–4, 212, 214, 281
Abarquhi, Muhammad 151, 158
'Abbasids 134;
Baghdad destruction 1, 8, 10; caliphate 191–7
'Abd al-Rahman, Shaykh 212, 214, 215–16
Abu Bakr, al-'Adil 177
Abu Bakr, al-Mansur 140, 141, 145, 194
Abu Firas 135, 136
Abul-Faraj 1
Abu'l-Fida 1, 138, 143, 144
Abu Sa'id 169, 207
Abu Shama 131, 132, 187–8
Abushqa 279
Acqui, Jacopo d' 266–7
Acre 68–9, 171, 172, 174–5, 176, 178, 179, 233, 237–8, 277
Adam, Salimbene de 266
adh-Dhahabi 2, 5
Ahmad 212–16, 274
ajlab 46–7, 48, 50, 51–2, 125–6, 127–8
ajnad 93, 115
ajnad al-mi' atayn 79
akabir al-umara 122
al-'Abbasi al-Ansari, 'Uthman Efendi ibn 197
al-Adil II 26, 224, 226
al-Adil Kitbugha 142, 143
al-Adil Sayf al-Din 144
al-Afdal 13, 17
al-Ahmadi, Arghun 110
al-Amjad, al-Malik 142
Alans 30, 31
al-Ashqar, Sunqur 82, 142
al-Ashrafi Burulghi 196
al-Ashrafi Sirghitmish 110
al-Ashraf Qa'it Bay 141
al-Ashraf Sha'ban 63
al-Ashraf Tuman Bay 139, 194
al-'Asqalani, Ibn Hajar 74
al-Atharib 21, 24
al-Aziz'Uthman 144
al-Barli, Aqush 191
Alberic of Trois-Fontaines 227
al-Bundari 187, 189
al-Dawla, Sa'd 184
al-Din ibn Zangi, Nur 28
Aleppo 16, 22, 25–6, 131
al-Fadil, Qadi 85, 134, 189
al-Gharb 86
al-Ghawri, Qansuh 30, 35–6, 54, 125; lance play 124;
mamluks 67;
qaranis 127;
qirnas 119
al-Hakim 15, 144, 147, 191, 192–3, 194
al-Halabi Sanjar 142
al-Hayjawi 243
al-Hayjawi, Rukn al-Din 226
'Ali Bay 118
Alice of Antioch 24
al-Irbili al-Mutatabbib, al-'Izz Hasan 254
al-Isfahani Imad al-Din 186–7, 188
al-Jabarti 114
al-Jawwad 220, 224, 235, 237, 240–1, 245
al-Jawzi, Sibt Ibn 133, 217–18, 241
Aljay 80–1
al-Kamil 62, 221, 224, 229, 232, 246
al-Kamil, al-Malik 142
al-khassa 37
al-Khassaki Yalbugha an-Nasiri 82
al-Khazraji 240–1
al-Kurdi, Ali 21
al-Kutubi, Muhammad ibn Shakir 2, 5
al-mahisiyin 74
al-Mansur 'Ali 140
al-Mansur, al-Malik see AbuBakr, al-Mansur
al-Mansur of Hims 230, 231, 236, 240, 241–2, 243–5
al-Mansuri, Baktimur al-Abu Bakri 95
al-Mansuri, Shams al-Din Qurasunqur 214
al-Maqrizi 27, 32, 62–3, 64–5, 218; al-Mu'izz Aybak 139, 140;
army strength 110–11;
halqa 69, 73, 74, 76; irrigation system 38;
mamluk numbers 61;
Salah ad-Din 84;
Tatar 124;
Yasa 201, 202
al-Marqab 176, 177
al-Mazdaqani, Abu 'Ali 23
al-Mu'ayyad Shaykh 54, 59, 64, 67, 74–5, 93, 96
al-Mu'azzami, Aybeg 236
al-Mughith Mahmud 241, 242
al-Mughith 'Umar 146
al-Mujahid 225, 227
al-Mujahid, al-Malik 142
al-Mukarram, Muhammed ibn 170
al-Musta'in 196–7
al-Mustakfi I 195, 196
al-Mustansir 144, 145, 191–3, 194
al-Musta'sim 1
al-Mutawakkil III 194, 197
al-Muzaffar, al-Malik 142
al-Muzaffar Mahmud 224, 225, 227–8,243, 246
al-Muzaffar Qutuz 143, 145, 146–7, 191, 212
al-Nasir Ahmad 194
al-Nasir, al-Malik 142
al-Nasir Da'ud 218, 221, 224, 225–6, 228,229, 231–3, 234–40, 241–2, 243–4, 246–7
al-Nasir Faraj 83
al-Nasiri, Sayf al-Din Aytamish 257
al-Nuwayri 218, 236
al-Qalqashandi 30, 54, 79; amir ranks 90;
amir's soldiers 87;
atabaki 98;
ghashiya 144;
halqa 68;
khassakiya 53;
Royal Mamluks 44;
treaties 170, 171–2, 174
al-Qaymari, Shihab ad-Din 85
al-Qurdumi, Qajqar 80
al-Sa'adat, an-Nasir Muhammad Abu 67
al-Safadi, Salah al-Din Khalil ibn Aybeg 253–5, 256–7, 258–60, 262
al-Sa'id Baraka Khan 140, 141
al-Salib, al-Malik 136
al-Salih 'Ali 140, 142
al-Salih Ayyub 62, 94, 139, 144, 218, 220, 221, 224–8, 231, 232–47
al-Salihi, Nasir ad-Din Aghulmishas-Silahdar 86
al-Salih Ismail 189, 192, 218, 220, 224,225–6, 227–32, 234–42, 243–4
al-Saqi, Qurasunqur 235
al-Suduni, Yashbak 84
al-Sulami 236
al-Tayti, Shams al-Din 214, 215
al-'Umari 201, 202
al-'Umari, Ibn Fadl Allah 85, 93, 111, 255–6, 257, 259
al-'Umari, Shaykhun 74, 122
al-'Uthmani, Tunbugha 98
al-wafidiya 70
al-Yusufi, 98
al-Yusufi, Manjak 73–4
al-Zahir, al-Malik see Barquq
al-Zahiri, Jarbash ash-Shaykhi 123
al-Zahiri, Khalil ibn Shahin 61, 89–90, 91,99, 111–13, 115, 118–19, 194–5
Amaury of Montfort 226
Amid battle 125
amir akhur 103, 109
amir arba'in 89
amir jandar 103–4
amir kabir 121–2
amir majlis 99
amir silah 99, 100
amir tablkhana 89, 90, 95, 110–11, 122
Amuli, 207–8
Anas 142
Anselm of Trainel 227
Aqsunqur 25, 26
Aqtimur 122
Arghun 162, 184, 216, 254–5, 275; Kökechin bride 279;
Polo's travels 273
Armand of Pierregort 135, 227, 246
Armenians 25, 31
'Arqa 173–4
Artah 25
Assassins 131–7, 176
atabak 117
atabak al-'Asakir 98–9
Athalaric 165
awlad an-nas 43, 76–8, 127
awna 35
Ayalon, Professor David 198, 199, 201, 202, 204, 205
Ayarghun 210
Aybak 61, 147–8
Aybak, al-Mansur 'Ali ibn 107, 141, 143, 145, 148
Aybak, al-Mu'izz 139, 140
Aynal 50–1, 58, 66, 125, 128
Aynal al-Abu Bakri 117
aynaliya 58, 125
'Ayn Jalut 212
Aytamish al-Bajasi 83
Ayyubids 32, 94, 138, 242; army 37, 61–2;
dynasty 218, 219;
empire invaded 10;
malik-titles 142, 146; saddle cover 144;
tawashiya 84–5
Ayyub see al-Salih Ayyub
'Azaz 25, 26, 132

Bacon, Roger 266

Baghdad, destruction by Mongols 1–26
Baha ad-Din 132
Bahri period 47, 82–3, 95, 108, 109
Baiju 241
Baktut al-Atabaki 147
Baldwin I 16–17, 21, 23, 24
Baldwin II 13, 23
Barbarians 151–69
Bar Hebraeus 33
Barka-Khan 33
Barquq 30, 32, 48, 54, 65, 122, 123, 140, 141; army 64;
diwan al-mufrad 88;
mamluks 63–4;
wazirate 101
Barquq, Faraj ibn 64, 110, 141, 196
Barsbay 46, 54, 56; Amid battle 125;
army 58;
atabak 117;
governor appointment 60;
halqa 75;
mamluks 64–5, 66, 67; sayyidi 78
Basil II 16
basqaq 165, 168
Batinis 17, 22, 23–4
Baybars, al-Adil Salamish ibn 140, 141
Baybars al-Bunduqdari 61–2, 107–9
Baybars al-Dawadar al-Mansuri 143
Baybars al-Mansuri 62, 86, 121, 170
Baybars, al-Muzaffar 82, 149, 195–6
Baybars, al-Zahir 33, 34, 46, 139, 140, 142, 143, 145, 148, 191, 192–3; accession 144;
revolt 149;
sacredness 146–7;
treaties 170–8
Baydara l42
Becker,C.H. 144
Beduin tribes 112
Berke Khan 193, 243, 276
bigar 34–5
Birdibak 60
Blochet, Edgar 220, 241
blood-money 176
Bohadin 186
Bolad Aqa 259
Brent, Peter 180–1
Browne, E.G. 156–7
Buddhism 166
Bulst, Marie-Luise 225, 245
burjiya 62
Bursuq ibn Bursuq 14

Cassiodorus 151–69
castes 28, 29, 34
Cepoy, Thibaut de 267
Chaghatai 205–6
Chambers, James 182
Chingiz Khan 27–41, 153, 159–60, 262; books 180–5;
clergy tax exemption 166;
Ghazan 260;
Yasa 27–41, 198–211
Ch'iu Chün 250, 251–2
Christians 166
Chuban 208
Chu Ch'üan 250, 252
Circassians 30, 43, 46, 63, 109–10; amirs 95–6;
mamluks 83–4, 141;
mushtarawat 125;
qaranis 113–14, 127, 128, 129
Clavijo 161
Cleaves, F.W. 248, 249
Cogatai 276, 277, 278–9
Comes Gothorum 165
Critchley, Dr John 263
Culficar 272

Damascus 16–17;
defence 13;
Franks 225–6, 228, 230; Latin forces attack 23–4;
Mawdud assasination 22–3
Dar al-harb 170
Daral-lslam 170
dawadar 109
dawadar al-Kabir 102–3
de la Croix, Pétis 38, 198, 199
de Lannoy, G. 67
Derenbourg, 11
diwan al-badal 73, 95
diwan al-jaysh 73, 79, 97
diwan al-mufrad 88

Edward I 281
Edward, Lord 278, 279
Ehrenkreutz, Andrew S. 188–9
eunuchs 84, 122

Fakhr al-Din 160

Fath al-Din 173, 179, 213
Fatimids 15, 16, 17, 20; alliances 13;
Caliphs 134
fiefs, distribution 127–9
foodstuffs 167
La Forbie 217, 243, 244, 246
Franks 14–15, 16, 17, 21, 217–47; Damascus 23–4, 225–6, 228, 230; Mamluk treaties 170–9;
Mawdud assasination 22–3;
Saladin 131
Frederick II 220–1, 229, 232, 245–6
Fulcher of Chartres 26

Gaza 221, 224, 226–7, 228, 229, 231–2, 234, 245

Genghis Khan see Chingiz Khan
Georgians 31
ghashiya 144
Ghazan Khan 153–4, 156, 158, 159–60, 161, 162–5, 206–7; Islam conversion 166, 253–62;
Marco Polo's travels 275, 279; reforms 167, 168
Ghurab 173–4
Ghurlu 73, 95
Ghuzz 31
Gibb, Hamilton 186
Gibbon, E. 166, 181–2, 186
Giles, J.A. 181
Godfrey 16–17
Golden Horde 30, 32–3, 40, 267
Goths 155, 158, 165
granaries, public 38–9
Gregory X, Pope 277, 278–9
Grousset, René 11, 12, 13, 180, 185
Gudisal the Saio 160
Gumushtakin, Sa'dad-din 131, 133
Güyük 206
hajib al-Hujjab 100

Hajji 31–2, 142

halqa 43, 45, 64–5, 68–79, 93, 111, 112, 118
Hamah 21, 24, 138–9, 144, 227–8
Hambis, L. 204
Hamdanids 16
Harran 14, 21
Hasan 76
Hawwara tribe 112
Hebraeus, Bar 201, 202
Heers, Jacque 268
Henry of Bar 226
Hodgkin, Thomas 154, 166–7
Hospitallers 176, 177, 178, 225, 233, 235, 245–6
Hsiang-mai 250
hujjab 28–9
hujubiya 109
Hulaghu Khan 1
Hülegü 153, 193, 258, 259, 275, 276
Hung, William 248–9, 252

Ibn 'Abd-allah, Kahardas 122

Ibn 'Abd al-Salam al-Sulami 230
Ibn 'Abdal-Zahir 143, 146, 171, 172, 178, 192, 212, 214–16
Ibn abi'l-Yusr, Taqiaddin Ismail 2, 5, 9–10
Ibn Abi Tayy 131
Ibn ad-Daya 133, 135
Ibn al-'Ajami 136
Ibn al-'Amid 220, 240, 241
Ibn al-Athir 1, 15, 17–25, 26; anti-Fatimid bias 15;
Chingiz Khan 153;
Saladin 131, 132, 136, 187
Ibn al-Burhan 201
Ibn al-Dawadari 145, 146–7, 231
Ibn al-Furat 2, 46–7
Ibn al-Qalanisi 18, 19–20, 22, 23, 24, 25, 26
Ibn 'Ammar 16, 17
Ibn an-Nahhal 120
Ibn 'Arabshah 65, 118
Ibn at-Tiqtaqa 1–2
Ibn Battuta 208
Ibn Duqmaq 218
Ibn Hajar al-'Asqalani 95
Ibn Iyas 36, 66
Ibn Ja'wan 2
Ibn Khaldun 30, 46
Ibn Muhanna, 'Isa 191, 192, 193
Ibn Munqidh 21
ibn Ortuq, Balak 25–6
ibn Ortuq, Sukman 18
Ibn Shaddad 85, 186–7, 218, 231, 238
Ibn Taghribirdi 50, 57, 58, 61, 64, 65–6, 109–10; al-Mu'izz Aybak 139;
al-Musta'in 196;
amirs 95–6;
Aytamish 83;
Barsbay 75;
Baybars 109;
rawks 73;
saddle covers 144;
sultan accession 144;
Tatar 124;
Yalbugha 82
Ibn Takash, Shihab ad-Din Mahmud 132, 133, 135
Ibn Taymiyya 261–2
Ibn Tulun 197
Ibn Wasil 131, 132, 217–18, 227, 228, 242, 243, 244, 246
ilchis 161, 163
Imad al-Din 131, 132, 136, 189
irrigation system 38
Isabel of Beirut, Lady 177
Ismail 189, 192, 218, 220, 224, 225–6, 227–32, 234–42, 243–4
Isma'ilis 134–5
Isolo 275, 281

jamakiya 117
Janibak 58, 83, 121
Jaqmaq 65, 66, 125
Jews, Genoa 165
Jikirmish 14
Jilliq, Baktamur 197
John, King 181
John of Mâcon 227
John of Montecorvino 274
John of Plano Carpini 266, 271
julban 46–7, 50, 51–2, 60, 114, 115–17, 119–20, 125, 126–7
Juwayni 33, 34, 35, 201–4, 205, 206, 208
Juzjani 206

kafil al-Mamalik 105–6

Kamal ad-Din 17, 24, 25, 26, 131, 132–3, 135, 233, 235, 239
Karbuqa 14
katib al-Mamalik as-Sultaniya 105
Kaykhusrau II 241, 242
Ke'ün, Arpa 207
khadama 55
kharaj 40
khassakiya 52–5, 70, 91, 95, 111, 140
Khatun, Bulughan 254–5
Khatun, Ismat al-Din 189
khazindar al-Kabir 102
Khumartakin, Nasih ad-Din 131
khushdash 121
khushdashiya 46, 50, 58, 149
Khushqadam 57, 66, 67, 125
Khushqadam, as-Sayfi 114
Khushqadam, az-Zahiri 49, 53, 54
khushqadamiya 125
Khutba 134
khutba 145, 146, 147, 192, 193
Khwarizmians 230, 241, 242–5
Kitbugha, al-Adil 140
Kökechin 279, 280
Kublai Khan see Qubilai Khan
Kurds 32
Kwanten, Professor Luc 180, 182–5

Lachin 72–3, 140, 143

Lane-Poole, Stanley 141, 186
Lang, Timur 64
Laurent, M.H. 277–8
Leo III 278
Leon III 142
Levtzion, Professor Nehemia 261
Li Johet 237
Louis IX 281
Lucalongo, Pietro da 274, 279

Ma'arrat Masrin 24
Mahmud, Amir 162
majlis 34
makhdum 55
malik al-Umara 105–6
mamalik al-umara 79–82
mamalik 118
Mamluks 27–41;
army structure 42–67, 68–96, 97–130; Frankish treaties 170–9;
halqa 68–79;
peasant movements 39;
public granaries 38;
rawks 72;
sultan power 138–50
Mangi 272
Mankutimur 72–3
Margaret of Tyre, Lady 175, 176, 178
Margat 177
Marj Dabiq battle 82, 116, 125, 127
marthiyya 8–9
Masyaf 132, 135
Mauclerc, Peter 226
Mawdud, 14, 21, 22–3
Melville Dr Charles 253
Michael VIII Palaeologus 276
Mintash 48, 74
Mirkhwand 198
Möhring, Hannes 189
Momigliano, Arnaldo 154, 155–6
Möngke 153, 161
Möngke, Töde 215
Montferrat, Conrad de 136
Morgan, Dr 257–8
mu'ayyadiya 58, 66, 118
mudafun 71
Muhyi al-Din 170, 179, 191, 212
Munqidh, Banu 16
muqaddam al-Mamalik as-Sultaniya 105
muqaddamu 69–70
muqaddamu al-mamalik as-sultaniya 45
mushtarawat 46, 47, 49, 50, 54, 56, 58, 59, 64; in battle 125–6;
numbers serving 61;
payment 86–7, 126–7
Mustakhdamun 55–6, 64
Mustawfi, Hamd Allah 163–4, 168

Nablus 224, 229, 230, 232, 235, 239

na'ib as-saltana 97–8
Naj, ad-Din Ayyub 85
Nakhjawani, Muhammad Hindushah 209
naqib al-Jaysh 104–5
naqib al-Mamalik 105
nasib 9
nasiriya 118
Nawruz 254
nazir al-Istablat as-Sultaniya 106
nazir al-Jaysh 106
nazir Khaza'in as-Silah 106
Nicholas III, Pope 277, 278
Nizaris 134
Nubians 44
Nubuwiya 133
Nur al-Din 32, 134, 186, 187–8
Odoacer 152–3, 159, 167–8
Odoric of Pordenone 269
Ögedei 161, 203, 206
Oghuz 31, 32
Öljeitü 158, 207, 208, 262
Olschki 281
Ormanni, Niccolò degli 265
Ossets 30
Ostrogoths 151, 152–3, 162, 164–8
Ottoman feudal system 35

Painter 225
Paris, Matthew 181, 230, 231, 232, 237
peasants 34–5;
kharaj 40;
movements 39
Pelliot, Paul 180, 185, 276, 280
Phillips, E.D. 180
Pipino, Francesco 265, 267, 269
Poliak, A.N. 68, 71–2, 86, 113–15, 123, 127; qaranis 118, 119–20, 128
Polo, Marco 161, 263–82
Prideaux, Humphrey 138

Qa'an 257, 258–9, 260

qadial-'Askar 107
qadim hijra 121
qadis 148, 160
Qaidu 205
Qalaun, al-Ashraf Khalil ibn 62, 140, 141, 148
Qalaun, al-Kamil Sha'ban ibn 73, 75–6, 95
Qalaun, al-Mansur 31, 71, 72, 95, 139, 140–1, 142, 148, 150; embassies to 212–16;
revolt 149;
treaties 170–9
Qalaun, al-Nasir Muhammad ibn 45, 67, 109, 110, 138, 140, 144, 148, 150, 194, 196; amir numbers
89;
halqa 69;
land redistribution 88;
mamluks 61, 62–3, 64; sacredness 146, 147
Qalaun, Muhammad ibn 48–9, 54
Qalqashandi see al-Qalqashandi
Qansuh, az-Zahir 66
Qara-Khitais 209
qaranis 50, 58–9, 113–29, 149
Qarasunqur 82
qasida, Baghdad destruction 1–10
Qaytbay 66, 125, 126
qirnas 119, 123, 124
Qocha 279
Quatremère 68
qubchur 168
Qubilai Khan 158, 204, 250, 258, 273–5; Polo's travels 264, 270, 276–7, 278, 279, 280–1
quriltai 198–9
Qutb al-Din 213
Qutlugh-Shah 207, 208, 257, 258
Qutula Khan 261
Qutuz 143, 145, 146–7, 191, 212
Qypchaqs 29, 30

Ramusio 265, 266, 267, 270, 272, 273, 280

Rasas, Tanam 58
Rashid al-Din 151–69, 184, 199, 200–1, 204, 206–7, 208, 209–10, 211; Ghazan 254, 256, 261;
Polo's travels 279, 281;
shamanism 261
ra's nawba 109
ra's ru'us an-nuwab 100–1
rawks 71–3
Riasanovsky, V.A. 199, 200
Richard Coeur de Lion 136
Richard of Cornwall 217, 220–1, 231, 232, 233–5, 245
Richard, Professor Jean 234
Richards, D.S. 187
Robert of Courtenay 227
Roger of Wendover 181
Romulus Augustulus 152
Ronay, Gabriel 181–2
Rothelin continuation 217, 227, 231–2
Royal Mamluks 43–5, 59, 61–7, 77–8, 111, 112, 149; amir service 80;
qaranis 115, 116;
rawks 72
Rudwan 16, 21
Rukh, Shah 75
Russians 29–30
Rusticello 265, 266, 268, 279, 282

Safi al-Din, Shaykh 167

saio 165
Saladin, 131–7 186–90
Salah ad-Din 68–9, 84–5
Salar 85
Salih ibn Yahya 86, 91
Saljuqids 12–13, 14, 16
Saruj 18–19
Sauma, Rabban 276–7
Saunders, J.J.180, 185
Sawar 24
Sawinj 24
Sayf ad-Din Qila 74
Sayf al-Din 227
sayfiya 58, 59–61, 115–16, 126–7
sayyidi 78
Schultens, Albert 186
Scott, Sir Walter 186
Selim the Grim 139, 197
Shafi ibn 'AIi 172–4, 178, 192, 212, 213, 214, 215–16
Shaizar 21
shamanism 261
Shigi-Qutuqu 199, 200, 209, 211
Shihab ad-Din 135
silahdar 109
Sinan 131, 132, 133, 134, 135–6
Sinor, Denis 180, 183
Sirasunqur, Shams al-Din 239
Sirjan 209
Siyan, Yaghi 13, 25
siyasa 27, 28
Smalley, Beryl 187
S.m.daghu 215
Smith, A.L. 181
Soranzo, Giovanni 278
Spiesz, Professor O. 5
Stevenson, Professor W.B. 11, 24
Stubbs, William 186
Sudun al-Balati 81
Sunni-Shi'a schism 15

Tabari 1
Taj al-Din 'Ali Shah 154
Tamerlane 33
Tancred 21–2
T'ang Shun-chih 251
Taqi ad-Din 135
tarkhans 35
Tarmashirin 208
Tatars 109, 199
Tatar, Sultan 80, 114, 122–4
tawashiya 84–7
Tedaldo Visconti 277
Templars 225, 235, 236, 238, 244, 245–6
Temür Öljeitü 259
Terken Khatun 209
Theobald of Navarre 217, 220–1, 224–5, 226–7, 228, 229, 230–2, 235, 246
Theodolus 281
Theodoric 152, 155, 158–9, 164, 167–8
Tripoli 19–20, 172–4, 225, 227, 243
Tughtagin 17, 22, 23, 24, 25
Tunanbay 66
Turan Shah 68, 107
Turan Shah, al-Mu'azzam 147, 150
Turks 29, 114;
cavalry 36–7;
tarkhans 35
Tusi, Nasir al-Din 204
Tutush 16
Tyre 21, 23–4, 175, 176, 178

Uladai 279
'Umara 134
Usama ibn Munqidh 11, 17
ustadar 101–2
ustadh 55

Vicenza, Niccolò da 277

Von Hammer, 113

wafidiya 76, 93
Wallace-Hadrill, Professor J.M. 167
Wassaf 204, 205, 206
wazir 101
Wiet, 141
William of Agen 277
William of Rubruck 266, 281
William of Tripoli 277, 279
William of Tyre 217
Wood, Dr Frances 263

Yaballaha, Mar 166, 276

Yaghmur, Nasur al-Din 241
Yalbay 51
Yalbugha 48, 89
Yam 160, 161
yarghu 208–10
Yasa 27–41, 108, 164, 198–211, 255–7
Yusuf 59–60
Yvo of Narbonne 181

zahiriya 58, 118, 125

Zangids 131, 186–8
Zanki 17, 21, 24
Zeno, Emperor 152
 Footnotes

A on the Destruction of Baghdad by the Mongols

1 For the details see G.Le Strange, Baghdad during the 'Abbasid Caliphate, Oxford-London, 1900, p.
343.
2 See Le Strange, op. cit., p. 340.
3 As for Persian literature, the following historical works contain narratives of this event: (1) The
written shortly after 656/1258, is a contemporary authority on the times of
Hulaghu; (2) the Jami' at-tawarikh, Rashid-addin's well-known work, finished in 710/1310–11,
provides a fairly clear account of the siege operations; (3) the history of the historiographer of
Ghazan, the Ilkhan of Persia, written in 700/1300–1, contains only the data related also by
Rashidaddin. See Le Strange, op. cit., pp.340–1.
1 See the edition of W.Ahlwardt, Gotha-Göttingen, 1860, pp. 383–8.
2 See Le Strange, op. cit., p. 343, note.
3 See my paper, "The Ta'rikh al-islam of adh-Dhahabi," JRAS., 1932, pp. 815–855.
4 See the MS. of the Bodleian Library (Ury), No. 654, fols. 248–250, under the title Ka'ina Baghdad.
5 See the edition of Bulaq, A.H. 1299, vol. i, pp. 12–14.
6 See the ta'rikh al-bashar of Abulfida, printed at Istanbul 1286, vol. iii, pp. 204–5,
according to which Da'ud, the son of al-Malik died on 27 Jumada'1-Ula, 656/2
June, 1258.
7 See the edition of Wüstenfeld, xxi, 3.
1 See his "Beiträge zur arabischen Literaturgeschichte", Abhandlungen für die Kunde des
Morgenlandes, Leipzig, 1932, p. 70.
2 I have to thank the obliging courtesy of the direction of the Archaeologisches Institut des
Deutschen Reiches, Abteilung Istanbul, which has been so kind as to have the poem photographed
from the MS. of the Aya-Sophia library and to obtain for this purpose a special permit from the
Ministry of Public Instruction at Ankara. The photograph is reproduced in the accompanying plate.
3 In the MS. of the Bodleian Library
1 Baghdad, said to be so called because one of its inner gates was set askew (izwarrat—so Qamus,
s.v., but for other explanations see Le Strange, Baghdad, p. 11).
2 The zunnar, or cord waistband, was one of the distinguishing marks of Jews and Christians.
3 A quarter of Baghdad near the Bab Badr; Le Strange, op. cit., pp. 270–2.
1 See Goldziher, Bemerkungen zur arabischen Trauerpoesie, Vienna Oriental Journal, vol. xvi, 1902,
pp. 307–311.
1 Ibid., pp. 327–330, where we read that according to Ibn Rashiq in his 'Umda fi ash-shi'r, he
could not find any nasibs in the marathi with the exception of a by Durayd ibn But
even this exception is explained by the circumstance that this poem was written one year after the
death of the lamented person, when the blood-ransom for his sake had been fulfilled already, so that
the poet could employ a nasib to express his other feelings with the deceased person.
2 Ibid., pp. 313–14.
3 Ibid., pp. 314–320.
4 The same wa kam is repeated by Abu Nuwas thirteen times in a (Diwan, ed. by Iskandar
Asaf, Cairo, 1898, p. 140). See the note in Goldziher, op. cit., p. 315.

Notes on the Arabic Materials for the History of the Early Crusades
1"Notes sur les Croisades" in Journal Asiatique, 1902, mai-juin.
2Topographie historique de la Syrie (Paris, 1927).
3 Vie d'Ousama (Ousâma ibn Ire Partie, Paris, 1889).
1 Is William of Tyre a good enough authority for the accusation that the Muslims "brutally
eliminated" the indigenous Christian elements in Jerusalem on the arrival of the Crusaders (Grousset,
284–5)? The statement seems to be contradicted by numerous passages in which Fulcher and others
speak of the native Christian population (e.g. the jubilant passage on the reception of Baldwin I,
quoted G. 213).
1 See Ibn Muyassar, ed. Massé, p. 7, and E.Laurent, Byzance et les Seljoucides, p. 22. The fact that
the calculations of the Fatimid government were based upon the history of the earlier Byzantine
invasions is noted by all historians; but there is a tendency to over-emphasize in this connection the
importance of Jerusalem to the At the time of the First Crusade the possession of Jerusalem
was of little political importance, except as implying control of southern Palestine. It was the
establishment of the seat of the Latin kingdom at Jerusalem that caused it to acquire subsequently the
symbolic significance which it had by the time of Saladin.
2 Although, of course, the disintegration of the local Syrian kingdom of Tutush was responsible for
the absence of a united resistance within Syria.
3 Ibn al-Qalanisi, ed. Amedroz, 213 (Damascus Chronicle, 175).
1 Grousset, pp. 97, 107.
2 Cf. Encyc. of Islam, s.v.Kurbuka.
3 Ibn al-Athir, ed. Tornberg, x, 256, 5–4 from foot; on the same expedition Sukman had 7,000
Turkmen horsemen with him. Cf. the army of Saif ad-Din, prince of early in 1176, when, with
the aid of the Ortuqids of Kaifa and Mardin, "numerous forces assembled to join him, reaching
6,000 horsemen" (ibid., xi, 283, 5–7).
4 In the Damascus Chronicle, p. 99, n. 4, there is a serious error; Mawdud was the son of a certain
Altuntagin, and was not the nephew of Karbuqa.
1 The episode of the émeute at Baghdad in 1111 (cf. Grousset, 460–1) shows the Caliph himself, so
far from being moved by the Syrian appeal, furious at the affront to his personal dignity and only
restrained from taking violent measures against the ringleaders by the tact of the Sultan; see the
original and more detailed account in the Damascus Chronicle, pp. 110–12.
2 M.Grousset, for example, seeks to explain the refusal of of Aleppo to co-operate with the
other Syrian princes and with Mawdud by his patronage of the thereby inverting cause and
effect. The true reason is more probably to be sought in his embitterment at the repeated
disappointment of his ambitions.
3 As it was later on, in the sixteenth century; cf. A.J.Toynbee, A Study of History, vol. i (Oxford,
1934), pp. 347–400.
1 This is, notwithstanding its external conformity, the note sounded by al-mulk in the Siyaset-
Namah, and is frankly acknowledged by no less an authority than al-Ghazali ( 'Ulum ad-Din, ii,
124).
2 The possible influence exerted in this and similar situations by a certain historic antagonism
between the populations of Aleppo and Damascus may be suspected, as a supplementary factor, but
the whole subject awaits investigation.
1 Ed. Massé, p. 63.
2Damascus Chronicle, pp. 179, 280. According to Ibn Muyassar (p. 70) similar advances were made
by the also to Aqsunqur al-Bursuqi after his occupation of Aleppo.
3 Idris 'Imad ad-Din:'Uyun al-Akhbar, vol. vii (MS. of Dr. A.H.al-Hamdani); cf. al-Ishara
ila man nala'l-Wizara, ed. 57–60.
4 p. 164 (Ibn al-Qalanisi, 204, 16: ).
5 The point has already been observed by M.Grousset in a note to p. 510.
VOL. VII. PART 4. 49
1 Somewhat similar conclusions were reached by the writer some years ago after comparing Ibn al-
Athir's narratives of the early history of the Arabs in Central Asia with his sources in and
Baladhuri.
2 Ed. Amedroz, p. 138; Damas. Chr., p. 50.
1 Ed. Tornberg, x, 222 (cf. Grousset, p. 63).
2 e.g. Egyptian capture of Jerusalem: I.A. 489 (wrong), I.Q. 491; attack on Shaizar: I.A. 502,
I.Q. 507 (i.e. after the expulsion of the from Aleppo, which is surely correct); Crusaders' raid
on Damascus: I.A. 520 (wrong), I.Q. 519; and cf. section (5) below. There are many other instances.
3 Ed. Amedroz, 163; Damas. Chr., p. 89.
1 The text is difficult, and I give this translation subject to correction. The reading of the passage in

the Receuil (Hist. Or., i, 273) is: which

in parts makes no sense at all and is rendered in the translation: "Depuis plus d'un an cette flotte était
prête et pourvue de tout, et on ne s'accordait pas sur les instructions qu'on devait lui donner."
2 Ed. Tornberg, x, 334.
3Damas. Chr., 87; cf. Stevenson, 50; Grousset, 253–4.
4 Needless to say, the further reflections on this subject by the author of the Nujum ( ed.
Popper, ii, 335, 3–9; quoted by M.Grousset, p. 357, as confirmatory of I.A.'s statement) are equally to
be rejected; the whole of passage, in fact, deserves to become a classic example of
reckless misstatement. It is noteworthy that Abu'1-Fida, (R.H.C.Or. i, 10) omits the passage entirely.
1 Cf. Damas. Chr., p. 105.
2 x, 338 (R.H.C. Or. i, 278–9); summarized by Röhricht, p. 88, and Grousset, p. 459.
3 That this is an exaggerated view of Zanki's achievement has already been rightly demonstrated by
Stevenson (p. 124).
4 Even here Ibn al-Athir exaggerates the amount of the tribute, which both Ibn al-Qalanisi (Damas.
Chr., 106) and Kamal ad-Din put at 20,000 dinars.
5Damas. Chr., 82.
6 Ibid., 99.
1Kamil, x, 339; Damascus Chronicle, 111–13. In regard to the latter it is a little curious that Ibn al-
Qalanisi does not explicitly mention either Baghdad or the Sultan.
2

3 x, 347–8.
1It is, in any case, impossible to attach to it the weight which it is given by M.Grousset: "A tort ou à
raison, Tughtekin se trouva dès lors suspect aux yeux de tout l'Islam [which is in contradiction with
Ibn al-Athîr's former statement quoted above], et une indication d'Ibn al-Athir prouve que cette
déconsidération l'atteignit aussi aux yeux des Francs" (p. 276).
2I.A. x, 461–2; Damas. Chr., 191–9; cf. Röhricht, 186–7; Grousset, 658–665.
1 It is difficult to see what good Tyre would have been to the and, on the other side, what
would the Venetians have said?
2The Crusaders in the East, pp. 125 and 129. But he decisively rejects the alleged capture of al-
Atharib in 1130 (p. 129, note 3).
3Damascus Chronicle, 197.
4 Kamal ad-Din, R.H.C. Or., iii, 661.
1 x, 466–7. Zanki did not, in fact, reappear in Syria until 1135.
2 R.H.C. Or., iii, 578; Ibn al-Q., 134 (Damascus Chron., 42–3).
3 R.H.C. Or., iii, 640; Ibn al-Q., 210 (Damas. Chron., 169–170); cf. Stevenson, 110; Grousset, 595–
6.
1 Kamal ad-Din, 636–7; Damas. Chron., 167–9. (Note that in the second last line of p. 168 in the
Damascus Chronicle "First Rabi'"is a copyist's error for "First Jumada" [began 26th June].)
2 R.H.C. Or., iii, 651; Ibn al-Athir, x, 443.
3 According to Ibn al-Athir, x, 439.
4Damas. Chron., p. 179.

The Influence of Chingiz-Khan's Yasa upon the General Organization

of the Mamluk State
1Revue des Études Islamiques, 1935, pp. 235–6. Cf. also our Feudalism in Egypt Syria, Palestine,
and the Lebanon, p. 15, n. 1.
2 Cf., e.g., al-Maqrizi, ed. A.H. 1270, ii, p. 221.
3 The long series begins with Pétis de la Croix, The History of Genghizcan the Great (French, 1710,
English, 1722). To cite only his latest followers: Vladimirtsov, Social Organization of the Mongols,
Russian, 1934; Alinge, Mongolische Gesetze, 1934; Riasanovsky, Fundamental Principles of Mongol
Law, 1937; Vernadsky, with Minorsky's collaboration, On the Contents of the Great Yasa, Russian,
Brussels, 1939.
1 Abu Shama, Kitab ed. A.H. 1287, i, p. 6, 1.37, to p.7, 1.2; p.7, 1.12; p. 11, 1.1.
2 Ibid., p. 13, ll. 18–25, etc.
1 Sultans Aynal and Yalbay may be cited as examples: Ibn Iyas, ed. A.H. 1311, ii, p. 64, 1. 29; Ibn
Taghri Birdi, p. 335, 1.17; p. 608, ll. 8–11.
2 This view is endorsed by Ibn Khaldun, Kitab al-'Ibar, ii, p. 10, 11. 6–12; iii, p. 534, 1. 26; iv, p.
501, 1. 8, etc.
3 Al-Sakhawi, al-Lami', ii, p. 276, 1. 17; Ibn Taghri Birdi, Nujum, vi, p. 784, 1. 8; p. 812, 1. 6,
etc.
4 ii, p. 221, ll. 14–15, etc. Such evidence must be carefully distinguished from that which
applies the term Qypchaq to the Golden Horde. In the Zangid and Ayyubid times the same name
designated a province of Kurdistan (after one of its rulers), whose lords were Turcomans: Abu
Shama, ii, p. 138, 1. 33.
1Nujum, v, p. 63. Ibn al-Wardi in Ta'rikh of Abu l-Fida', ed. A.H. 1286, iv, p. 150.
2 Ibn Khaldun, v, p. 372. Al-Qalqashandi, iv, pp. 182, 216."Wladmr" in Ibn al-Durar al-
Kamina, iv, p. 408, 1. 2=Vladimir? But in general the Russians, as others, adopted Turkish names: cf.
Baybugha Rus or B.Urus, d. 1353. During the ethnic conflict among the Mamluks in A.H. 870, there
was no longer a Russian faction: p. 525, ll. 14–16. The "Slavic" regiment of "Turkish"
troops in the Yaman, which existed in the fifteenth century (Sakhawi, x, p. 215, No. 937), possibly
retained its name from earlier times.
3 i, p. 102, ll. 18–21.
4 Ibn Iyas, iv, p. 56, ll. 8–9; p. 102, ll. 8–15; p. 149, ll. 8–17; p. 195, ll. 2–8, The appellation al-
Ghawri is derived not from the place of birth but from the barracks where he received his military
education, al-ghawr: al-Ghazzi, al-Kawakib al-Sa'ira, MS.Damascus i,
f. 123 (the author, who lived in the seventeenth century, already ignores the meaning of the word
! Cf. on tabaqat al-ghawr, situated in the citadel of Cairo, Ibn Taghri Birdi, al-Manhal
MS. Paris, v, f. 113a,]. 16).
5 Ibn Iyas, i, p. 269, ll. 10–15, 17; p. 309, 1. 23.
6 Cf., e.g., Abu l-Fida', Ta'rikh, p. 97, 1. 28; Ibn Khaldun, ii p. 282, ll. 8–9; iv, p. 462, 1. 5.
Under the Circassian sultans the citadel of Cairo contained for some time a Christian church,
probably for Circassian women, who were not compelled to give up their faith as were men: al-
Sakhawi, al-Tibr al-Mastak, ed. 1896, p. 73, ll. 2–5. The native biographers of Mamluk emirs and
knights were not allowed to write about their Christian past, not even to mention the date of their
conversion to Islam.
1Charkas=jarkas=jarkas, lit."[owner of] four souls": al-Manhal i, f. 173a. Used as personal
name. Under the Ayyubids we still find its original form, chaharkas=jaharkas: Ibn Khallikan,
Wafayat al-A'yan, i, p. 397, 1. 3.
2 C.F.Volney, Voyage en Syrie et en Égypte, Paris, 1787, i, p. 166.
3 The Alans ('Alan, al-Lan, As, al-Az) were numerous there throughout the Mamluk
epoch: cf., e.g., Ibn Khaldun, v, p. 372; iv, p. 182; p. 525. The Abkhasians (
Abaza) only during its last century: loc. cit., etc.
4 Cf. our remarks in Revue des Études Islamiques, 1935, pp. 236–8. Unfortunately, owing to the
opposition of the military caste to the study of its tongue by the natives (Feudalism, p. 1, n. 2), the
surviving Arabic-Turkish glossaries of the Mamluk epoch have been written not by members of this
caste but by outsiders, who had little connection with it. Perhaps they do not reflect the official
dialect but are due to the fact that the natives did not distinguish one Turkish dialect from another, the
official being, of course, not the only one used in the Mamluk state. The only published glossary
whose author claims to be a Turk by birth, Bulghat al-Mushtaq of 'Abdullah al-Turki (ed.
Warsaw, 1938), is the work of one whose Arabic name shows clearly that he did not
belong to the military caste (cf. ), p. 616, ll. 5–8; al-Sakhawi, iii, p. 71), and who
honestly admits that the Turkish words compiled by him became known to him through the medium
of earlier books and of persons better acquainted with Turkish than himself. He too evidently
believed in the identity of "Turkish " proper with the language of Qypchaq (cf. the text, p. 1, ll. 1,
10), the latter being in this case probably not one of South Russian Qypchaqs but the tongue of North
Mesopotamian Qypchaq, which included that region of Irbil in which he was particularly interested
(cf. p. xiii). This tongue belonged, indeed, to the same Ghuzz family as the" Mamluk Turkish, though
it hardly was quite identical with the latter. The claim of North Syrian Turcomans to be Qypchaqs (
 xiii, p. 37, 1. 7) might be also a reminiscence of their being descendants of that Qypchaq, since,
according to many evidences, this Turcoman tribe was brought to North Syria by the Atabeg Zangi
from the east:
cf., e.g., Abu Shama, i,pp.43–4(on the tribal name cf. vii,p.190; Zubdat Kashf al-
Mamalik, ed. Ravaisse, p. 105). At any rate, though the material cited in such works is possibly of
more heterogeneous stock than the official Mamluk vocabulary, it also proves the latter's Ghuzz
origin.
1Nujum, v, p.56.
2 Cf.,e.g., p. 549, 1. 18.
3 iv, p. 182, 1. 12.
4 p. 525, ll. 14–16.
5 i, p. 95, 1. 4. Suluk,Quatremère's transl., i, p. 24, 1. i, ii, p. 45, n 53.
6Ibn Iyas, i, p. 168,1. 6.
7 loc. cit.
8Cf. the chapter Nobilitas hominum et mulierum in the editions of Kuun, 1880, and Grønbech, 1936.
1 Zubdat Kashf al-Mamalik, ed. Ravaisse, p. 16, ll. 11–13, etc.
2 Poliak in RÉI., 1935, p. 236, n. 3; Feudalism, p. 1, n. 4; and JRAS., 1939, pp. 429–430.
3 xiv, p. 102,1. 18; p. 103, 1. 7 (rendered as kurkan). While in Central Asia this title was so
forgotten that V.Barthold supposed it to have never been used after the Qara Qytai epoch (12
Vorlesungen über die Geschichte der Türken Mittelasiens, p. 123), the Mamluks preserved the
ancient tradition of the steppe.
1 We find this tendency already in a charter of Hulagu of A.H. 658; Ibn Ta'rikh Bayrut,
2nd ed., p. 57, ll. 2–8.
2Suluk, Quatremère's translation, i, ii, pp. 19, 25.
3 Cf. Minorsky in BSOS., ix, p. 950, after Juwayni, i, p. 22, 1. 20.
4 xiii, p. 94, 1. 8; p. 97, ll. 6, 13. Arabic sources render it, of course, as bikar.
VOL. x. PART 4. 58
1 This term survived in Egypt until the nineteenth century: Jabarti, iv, p. 207, ll. 17–20; the Earl of
Cromer, Modern Egypt, the chapter on the corvée.
2 Cf., e.g., Abu'l-Fida', ii, p. 172, ll. 18, 20, 21; Abu Shama, ii, p. 185, 1. 31; p. 203, 1. 12.
3 Cf., e.g., the personal name Ulugh (Ibn Khaldun, ii, p.297, 1. 21), third century A.H.
H.W.Bailey, Turks in Khotanese Texts (JRAS., 1939, p. 91), sums up some non-Moslem materials.
4 Mong. darkhan "freedman"; cf. Vladimirtsov, pp. 69, 93.
5 Cf. the sources cited in Feudalism, p. 15, n. 6.
1 He was a knight of and a fief-holder: iv, p. 136, ll. 10–12, etc.
2 Cf. Poliak in RÉI., 1934, p. 262, n. 4.
3 Cf., e.g., the application of these terms by Abu Shama (ii, p. 82, 1. 6; p. 144, 1. 20; p. 179, ll. 13–
18) to Ayyubid cavalry. The centre was commanded by the ruler, the two wings by senior generals.
On the corresponding terms used afterwards by Iranian Turks—manqalay Mong. "forehead, front ",
sagh " right" and sol " left "—cf. Minorsky in BSOS., x, p. 165.
4 Cf., e.g., Abu Shama, ii, p. 82, ll. 6, 11; p. 144, ll. 3–4; p. 202, 1. 24, and numerous Arabic treatises
on physical culture (furusiyya) and hunting. The corresponding term employed by Juwayni when
speaking on Mongol imperial hunts is nerge. Those treatises on warfare whose authors were faithful
disciples of the Arab military school (e.g. Ibn Khaldun, i, pp. 226–233) ignore halqa.
5 Thus, Kitab fi'Ilm al-Furusiyya, MS. Aleppo f. 18a, mentions a play during which horsemen
surrounded a single rider, who had to flee from their ring. Lachin al-Mujahidin fi'l-'Amal bi-'l-
Mayadin, MS. Aleppo mentions al-faris, two concentric rings constituted by "rival"
parties; al-muqabala, two rings formed by them behind their respective commanders, who met each
other in the middle; maydan al-ahilla, two concentric crescents; and of horsemen round their own
commanders or round the line on which they were formerly placed. These two small treatises (for the
use of excellent photographs of them I am indebted to Dr. S.Reich),
describe plays performed on public festivals by Arabic-speaking professional lance-players These
lance-players were a hereditary corporation, anxious to conceal the secrets of their professional
education from the general public (Furusiyya, f. 27b); and since the "Turkish " warriors concealed in
their turn technical particulars of their military art from natives, the art of was a conservative one,
claiming descent from Sasanian and early Islamic warriors (ibid., ff. 24, 34), not from the Turks and
Mongols. The author of Furusiyya calls various exercises (f. 4b), Khurasani (f. 28b), Shami (f. 32a),
never by names indicating a Turkish or a Mongol origin. Still, some Turkish influence infiltrated
through the medium of those Mamluks discharged from service who had to derive their subsistence
from lessons of horsemanship given to despised natives and adapted to their needs. The author of
Furusiyya evidently had only native teachers, as Badr al-Din (f. 5a) and Najm al-
Din Ayyub (f. 15b ff.), and for him a private Mamluk, Sayf al-Din who
condescended to write something on this subject, was a most venerable person (f. 14b); but the name
of Lachin indicates that he was some time a Mamluk, though he prefers not to speak about
his past; and the fact that he is much more versed in the halqa plays than the native author
corroborates our view that at least these plays were adopted by the from the military caste.
By the way, the word is always vocalized by him as possibly after its popular pronunciation.
1 iv, p. 22; xiv, pp. 166–171. Abu l-Fida', iv, pp. 30, 31, 93, 134.
2 Abu Shama, ii, p. 179, ll. 17–18; p. 180, ll. 4, 36, etc. Quatremère's explanation of the term in this
case as the guard surrounding the ruler and protecting him (Suluk, his translation, i, ii, pp. 200–2,
cited by Dozy, Supplément aux dictionnaires arabes, s.v.), is not founded by him on any sources. In
fact, did not surround the ruler but fought beside or, more frequently, before him.
1 iii, p. 449. i, p. 101. p. 129.
2Suluk, Quatremère, i, ii, p. 26, n. 29. Abu'l-Fida', iv, p. 120, ll. 19–22.
3 Cf. Feudalism, pp. 71–2.
4 Cf. on the Mamluks: Matériaux pour un Corpus Inscriptionum Arabicarum, ii, 1, 62; Gaudefroy-
Demombynes, La Syrie á l'épogue des Mamelouks, p. 249, n. 7; xiv, p. 377. ll. 1–3.
5 i, p. 91, ll. 18–19.
1 Cf. Feudalism, pp. 64 ff., and RÉI., 1936, pp. 261–3.
2 Al-Subki, Mu'id al-Ni'am, ed. Myhrman, p. 48, 1.15.
3 Ibn al-Ji'an's al-Saniyya is a list of Egyptian On the term kufur, sing. kafr, cf. ibid., p. 9, 1.4; p. 15,
ll. 22, 27, etc.; Jabarti, i, p. 346, ll. 26–8. On transformation of kufur into particular see Ibn al-Ji'an p.
9, 1. 4; p.13, 1. 4; p. 15, ll. 22, 27; p. 22, 1. 6; p. 59, 1. 16, etc.
4 Ibn Ta'rikh Bayrut (the cited feudal charters). Ibn al-Ji'an (the cases of villages divided among
several lords). The Buildings of Qaytbay, ed. Mayer, i (text), 1938, pp. 52–60. Ibn al-Durr al-
Muntakhab, ed. 1909, p. 75, ll. 10–11;
p. 119, ll. 1–6. Suluk, Quatremère, ii, i, p. 89. Matériaux pour un Corpus Inscriptionum Arabicarum,
i, p. 354, etc. Poliak in JRAS., 1937, pp. 105–6, and Feudalism, pp. 19, 69–70.
1RÉI., 1935, pp. 238–241; 1936, p. 264; Feudalism, pp. 65–7.
2 We refer to its final form. The article devoted to the early history of this tax in The American
Journal of Semitic Languages and Literatures, 1940, pp. 50–62, written long ago and never seen by
me in proofs, utilizes but a part of materials proving that this form differed from early ones.
3The Buildings of Qaytbay, i, p. 52, right margin.
4 Ibn Iyas, iv, p. 463, 1. 12; p. 471, ll. 18–23; v, p. 159, 1. 11.

Studies on the Structure of the Mamluk Army1—I

1 The present paper is a chapter from a work on the Mamluk Army. It deals only with the regular
forces, and is devoted for the most part to the army stationed in Egypt, which constituted the main
force of the Mamluk kingdom. The has been included in the regular forces because it formed, before
the Mamluk era and during a considerable part of that period, one of the most important components
of the kingdom's army. The troops stationed in Palestine, Syria, and the Lebanon are dealt with in
broad outline only. The army of the whole kingdom was called al-'asakir or al-'askar (Nujum(C),
viii, p.162. Nujum (P), vi, p. 703, 1. 20; vii, p. 91, 1. 3; p. 97, 1. 19. p. 645, ll. 16–18. Ibn al-Furat, ix,
p. 270, 1. 3. Ibn Khaldun, v, p. 496, 1. 6). This appellation was usually reserved for the army during
military expeditions, not in ordinary circumstances. It was sometimes also known as al-'asakir al-
islamiya (Abu al-Fida', iv, p. 45 (twice on that page); p. 70, 1. 13; p. 88, 1. 8. Ibn al-Furat, viii, p.
192, 1. 16). The army stationed in Egypt was known as al-'askar or al-'asakir (Nujum (P), vii, p.67,
1. 22; p. 68, 1. 1; p. 92, 1. 19; p. 95, 1. 8. p. 300, 1. 4; p. 301, 1. 19; p. 550, ll. 4–5; p. 633, 1. 14, ll.
19–20; p. 641, 1. 14. Ibn Iyas, iii, p. 243, 1. 3. Ibn al-Furat, vii, p. 41, 1. 13; viii, p. 223, ll. 17–18; ix,
p. 64, 1. 1. See also references given in next note.) Contingents stationed in Palestine, Syria, and the
Lebanon (al-Bilad ash-Shamiya) were called al-'asakir ash-shamiya (Zetterstéen, p. 60, 1. 23; p. 80,
ll. 1–2. Abu al-Fida', iv, p. 21; p. 22, 1. 11; p.23, 1. 4; p. 58, 1. 4. Ibn Kathir, xiv, p. 25, 1. 13; p. 202.
Nujum (P), vi, p. 645, 1. 15; p. 646, 1. 6; p.688, ll. 5–6; p. 695, 1. 18; p. 712, 1. 9; vii, p. 92, ll. 7–8;
p. 97, ll. 12–13. p. 514, 1. 3; p. 630, ll. 21–2; p. 633, 1. 22; pp. 634, 1. 15–635, 1. 9; p. 646, ll. 2–3;
647, 1. 16, ll. 22–3. Ibn al-Furat, vii,p. 117, 1.8; p. 176, 1. 7. Tibr, p. 63, 1. 10. iii, p. 9, 1. 4).This
name was, however, sometimes used in the narrower sense of 'troops of the governorship of
Damascus', as distinguished from al-'asakir etc. (Zetterstéen, p. 144. Abu al-Fida', iv, p. 49, ll. 14–
15. Nujum (P), vii, p. 479, ll. 12–13. p.490, 1. 21; p. 645, 1. 14, 1. 18; p. 647, ll. 19–20; p.
653, ll. 9–10. Ibn al-Furat, ix, p. 237, ll. 1–2. Ibn Iyas, v, p. 71, ll. 4–5).The term'askar (Abu al-Fida',
iv, p. 57, 1. 6, et al.) is encountered for a brief period only, viz. during the first years following the
expulsion of the Crusaders from the Muslim-held littoral.
2 IV, p. 14, ll. 8–16, 1. 11. pp. 244, 1. 18–245, 1. 18. Nujum (P), vi, pp. 386–387. Zubda, p. 113, ll.
4–18; p. 116, ll. 7–19. i, pp. 91–5; ii, pp. 216–17. ii, pp. 111–13. The following division of the
Mamluk army given by al-'Asqalani is worthy of note: al-'askar kana qabla ad-dawla (i.e.of
Barquq) thalathat aqsam, al-awwal mamalik wa-hum 'ala mustakhdamin wa-mamlukin; wa-li-kullin
minhum jawamik wa-rawatib 'ala wa-min al-mustakhdamin huna wa-hunaka an la yakunu min al-
qism ath-thalith, wa-hum ajnad (Suluk [trsl. Quatremère], i, pt. 2, p. 161). It is hard to translate this
passage literally, as its language is very confused (the author uses mustakhdamin in two different
meanings, and though he speaks of three units he mentions only two, the first of which is divided into
two sub-units); but, in our opinion, its idea is quite clear:, the Mamluk army was divided into three
units: (a) mustakhdamun; (b) mamlukun (i.e. mushtarawat); (c) The two first units, who were
paid and maintained by the sultan, belonged to the same category (mamalik =mamalik ) from which
the third unit, the was excluded. Also of interest is the division, quite erroneous for the most part but
containing some glimpses of truth, offered by de Lannoy, who visited Egypt and Palestine at the
beginning of the 15th century. He states that the Mamluk army was divided into four parts, as
follows: (a) the who had distinguished themselves in the use of arms, and from among whom were
selected the commanders of fortresses, the captains, and the governors of the towns, some
VOL. XV. PART 2. 17
received their pay from the diwan of the sultan in ready cash, while the rest shared among themselves
the profits from the villages and citadels; (b) the sayfiya, who were foot-soldiers and carried no arms
but the sword (sayf); they received their pay from the diwan of their master; (c) who were held on
call, received their foodstuffs outside the rations regularly allotted to the soldiery, and were given
nothing over and above their expenses; when one of the regular mamluks died, one of the would take
his place; (d) the jalab, who were recent arrivals, knew no Turkish or Arabic (Moorish), displayed no
bravery and had no opportunity to show their strength and personal qualities (Archives de l'Orient
Latin, vol. iiA, p. 90).
1 Quatremère, in his various comments to Suluk. Sobernheim, Enc. of Is., art. 'Mamluks' (a
classification full of errors). Demombynes, in his introduction to La Syrie à l'époque des Mamelouks.
A History of Egypt in the Middle Ages, pp. 242 ff. Blochet, Patrologia, xiv, p. 570, n. 3 (an incorrect
division). G.Wiet, Précis de l'Histoire d'Égypte, Cairo, 1932, vol. ii, L'Organisation militaire et
administrative, pp. 237–249. W.Popper in his glossaries to Ibn Taghribirdi's an-Nujum az-Zahira and
ad-Duhur. The best classification hitherto given is that of Poliak, Feudalism, p. 2, although it
contains some serious inaccuracies. Its main fault is that it relies too extensively on the
'encyclopaedic' literature without sufficiently testing it in the light of the information supplied by the
chronicles.
2 The amirs' soldiers were also called but this designation is very rarely mentioned in the sources,
and only for a brief period of time. See Part II of this article.
1 p. 245, ll. 11–12.
2 Ibn al-Furat, viii, p. 52, ll. 19–23.
3 Ibn Iyas, ii, p. 67, ll. 13–14.
4 p. 444, ll. 18–23.
5 The dispatching of Royal Mamluks to serve as garrison troops in various parts of the Mamluk
kingdom is treated in greater detail in other chapters of the author's work on the Mamluk army. See
also Appendix B, to appear in Part III of this article.
6 See D.Ayalon, 'The Plague and Its Effects Upon the Mamluk Army', J.R.A.S., 1946, pp. 67–73.
1 ii, p. 218, ll. 7–8, 1. 31.
2 Note that the muqaddamu al-mamalik were not eunuchs. The head of the sultan's military schools
was indeed a eunuch called muqaddam al-mamalik but there existed other posts bearing identical
titles and not occupied by eunuchs. Thus it is known that the commander of the Mamluk garrison of
Mecca was also called muqaddam al-mamalik (one of a series of many titles pertaining to this post),
though he was not a eunuch; see the chapter dealing with the Mecca garrison. As for the muqqadamu
al-mamalik who were the commanders of the Royal Mamluks, it is clearly seen that they were not
eunuchs from their distinctly Mamluk names (Jarkas, Bakilmish), their titles (Sayf ad-Din), their
offices (amir akhur, ustadar, ra's nawba), their high ranks (Amir of a Thousand). The writer deals
with the eunuchs in a special chapter of his work on the Mamluk army. There he discusses, inter alia,
their names, titles, offices, and ranks. Cf. Ibn al-Furat, ix, p. 58, ll. 13–14, ll. 16–17. Ibn Shuhba, fol.
47b, 1. 7. Manhal, viii, fol. 259b, ll. 10–13; as well as from the fact that they had sons (Manhal, iii,
fol. 17a, ll.16–23. iii, p. 100, ll. 2–6. Nujum (P), v, p. 486, 1. 12; p. 633, ll. 10–12). For additional
data on the muqaddamu al-mamalik indicating that they were the commanders of the Royal Mamluks
both in official reviews and in the battlefield, see: Suluk, i, p.286; p.612, 1. 5; p.935, ll. 16–18. Nujum
(C), viii, p.162, ll. 13–15. Ibn al-Furat, vii, p.7, ll. 15–18; ix, p.163, ll. 11–12. Abu al-Fida', iii, p.167,
ll.14–15. ii, pp. 111, 1. 39–112, 1. 1; p.218.
3 In D.Ayalon, L'Esclavage du Mamelouk, The Israel Oriental Society (Oriental Notes and Studies),
Jerusalem, 1951.
4 Cf. ibid., pp. 29, 30, 31, 33, 36, 37, 59, 63. Glossary to Nujum, vol. vi, p. xxiii.
1Nujum (P), vi, p. 709, ll. 14–15; vii, p. 423, 1. 2.
2Nujum (P), vi, p. 641, 1. 13.
3Nujum (P), vii, p. 262, 1. 12.
4 Cf.also Manhal, viii, fol. 434b, 1. 10.
5Nujum (P), v, p. 157, ll. 6–7; vi, p. 641, 1. 13; vii, p. 262, 1. 12.
6 Ibn al-Furat, ix, p. 235, 1. 10; vii, p. 462, notes. Ibn Iyas, iii, p. 168, 1. 10; iv, p. 241, n. 4; the editor
here remarks that this spelling occurs frequently in the text in question; v, p. 5, 1. 6.
7 The term mamalik itself had not yet become stabilized. The term is encountered as early as 636
(Suluk, i, p. 281, 1. 11; p. 286; p. 343, ll. 7–8), viz. in the Ayyubid period, but this may be an
anachronism on the part of the later Mamluk historian, since Ayyubid sources do not appear to
employ this designation. For additional material on mamalik see: Zetterstéen, p. 164, ll. 12–13.
Suluk, ii, p. 34, 1. 12. Nujum (C), viii, p. 172, ll. 3–4. Nujum (P), v, p. 112, 1. 5; p. 149, 1. 19; p. 213,
1. 16; pp. 230–1; p. 235; p. 320, 1. 22; p. 321, 1. 1; p. 396, ll. 11–13; p. 582, 1. 21; vi, p. 38, 1. 2. Ibn
Iyas, iv, p. 129, 1. 11. i, pp. 65, 1. 26–66, 1. 1. Ibn Khaldun frequently refers to the mamluks by the
name of mawali, thus: 'Baybars min mawali Aydakin al-Bunduqdari mawla Ayyub' (Ibn Khaldun, v,
p. 381, ll. 6–7), and similarly many other passages (Ibn Khaldun, v, p. 143, 1. 12; p. 358,1. 5; p. 361,
ll. 2–4; p. 384, 1. 5; p. 394, 1. 28; p. 395, 1. 1; p. 409, 1. 10, 1. 20; p. 411, 1. 11, 1. 24; p. 451, 1. 15;
p. 452, 1. 12; p. 422, 1. 16).
8Nujum (P), vii, p. 457, 1. 3, 1. 9, 1. 14; p. 509, note. p. 191, 1. 20; p. 301, 1.6. Tibr, p. 314, ll. 2–3.
iii, p. 43, 1. 28. Ibn Iyas, iv, p. 342, 1. 15.
9Nujum (P), vii, p. 191, 1. 9, 1. 14, 1. 17; p. 192, 1. 17; p. 198, 1. 13; p. 205, 1. 14. Ibn al-Furat, ix, p.
6, 1. 12. Ibn Iyas, ii, p. 214, 1. 17; iii, p. 142, 1. 19, 1. 20; p. 150, 1. 1, 1. 3,1. 4, 1. 5; p. 161, ll. 13–
14; p. 314, ll. 16–18; iv, p. 443, ll. 14–16. The first one is preferred by Ibn Taghribirdi, while Ibn Iyas
uses the second almost exclusively. For identity of ajlab and mushtarawat, cf. Nujum (P), vii, p. 123,
1. 2, also notes, as well as p. 479, ll. 7–15. Ibn Zunbul, pp. 13, 1. 24–14, 1. 2. That the ajlab or
julban were mamluks owned by the ruling sultan may also be inferred from Ibn al-Furat, ix, p. 6, 1.
12. Ibn Iyas, ii, p. 65, 1. 17; p. 77, ll. 2–3; p. 137, 1. 8; iv, p. 404, 1. 13, 1. 14, 1. 15; v, p. 22, 1. 19; p.
29, 1. 17.
10 Ibn Iyas, v, p. 15, 1.5.
11 p. 534, ll. 11–17.
12Nujum (P), vi, p. 203, ll. 1–2.
13 Ibn Khaldun, v, p. 457, 1. 14, 1. 18; p. 458, 1. 1, ll. 8–9, 1. 24.
14 Ibn al-Furat, ix, p. 6, 1. 12.
1Nujum (P), v, p. 219, 1. 23; p. 401, ll. 9–10; vi, p. 757; pp. 768–9; vii, p. 189, ll. 6–7; p. 190, ll. 1–2;
p. 457 and note; p. 527, ll. 7–8; p. 530, ll. 1–2; p. 776, 1. 3. p. 21, ll. 14–15; pp. 37, 1. 8–38, 1. 3; p.
205,1. 1; p. 273, 1. 3; p. 301, 1. 6; p. 532, 1. 1. Manhal, ii, fol. 31b, ll. 19–20. Ibn Iyas, v, p. 43, ll. 4–
5. On very rare occasions, ajlab is used to designate amirs' mamluks as well (Manhal, i, fol. 193b, 1.
3; v, fol. 49a, ll. 16–17).
2 p. 460, 1. 14. Ibn Iyas, p. 400, 1. 8; v, pp. 35, 1. 23–36, 1. 1; p. 48, 1. 11. Cf. also Patrologia, xx, p.
41, 1. 2. Suluk, i, p. 433, 1. 12. Nujum (P), vi, p. 723, ll. 3–9; vii, p. 797, ll. 10–11. Ibn al-Furat, ix, p.
250, 1. 21.Tibr, p. 279, 1. 13; p. 293, ll. 24–5; p. 310, 1. 7. Ibn Iyas, ii, p. 170; p. 178, ll. 1–2; p. 184,
1. 21; p. 221; iii, p. 121, 1. 14, 1. 16; p. 267, 1.4, 1.6; p. 281, ll. 6–7. iii, p. 10.
3 Zetterstéen, p. 26, ll. 24–5. Ibn al-Furat, viii, p. 170, ll. 11–12. ii, p. 317, ll. 28–9. Cf. also ii, p.
218, ll. 7–8, with 1. 13.
4 Cf. also references in following footnotes, as well as the description of the sultan's first years of
rule in the Mamluk chronicles.
5Nujum (C),viii, pp. 48, 1. 16–49, 1. 7; p. 277, ll. 4–7. Suluk, ii, p. 81. Nujum (P), v, p. 457; p. 459, ll.
1–2. Ibn al-Furat, ix, p. 107, ll. 18–23; p. 108, ll. 19–21; p. 115, ll. 14–16. Ibn Khaldun, v, pp. 461–2;
p. 462, ll. 27–9; p. 486, ll. 22–3. x, p. 280, ll. 5–6. See also references given in note 2, p. 210 below.
1 Material on this subject has been gathered in L'Esclavage du Mamelouk, pp. 5, 19–20.
2 The following references contain information on the julban as the dwellers in the barracks of the
citadel, and on the driving out of the mamluks of the immediately preceding sultan (these took up
their abode in Cairo): Nujum, vi, p. 514, ll. 11–13; vii, pp. 12–18 (and also the previous pages from
the beginning of the volume); p. 193, 1. 7; p. 452, ll. 1–5; p. 467, ll. 4–10; pp. 491, 1. 10–492, 1. 3; p.
745, ll. 12–15; p. 836, 1. 14. p. 191, ll. 20–3; pp. 203, l. 14–206, 1. 3; pp. 239, 1. 18–240, 1. 1; p.
251; p. 443, 1. 15. Manhal, ii, fol. 193a, ll. 1–2; viii, fol. 451b, ll. 12–14 (and also the description
preceding these lines). Ibn al-Furat, ix, p. 302, ll. 6–18. Ibn Shuhba, fol. 79b, ll. 22–5. Tibr, p. 41, ll.
7–8. Ibn Iyas, iii, p. 341, ll. 1–2; p. 391, ll. 1–2; p. 397, ll. 15–18; iv, p. 41, ll. 15–17; p. 359, ll. 6–8;
p. 428, ll. 14–22; p. 484, ll. 5–6; v, p. 48, ll. 19–20; p. 63, 1. 15.
3 See my 'The Circassians in the Mamluk Kingdom', J.A.O.S., 1949, pp. 135–147.
1Suluk, ii, p. 94. Nujum (C), ix, p. 27, ll. 12–15.
2 Mamluk sources furnish extremely abundant material on the methods followed by the new sultan,
on his accession to the throne, in order to establish his rule. Because these methods produced most
profound effects upon the relative strengths of the mushtarawat and the mamluks of the former
sultans, a selection of references larger than the usual is here given. Zetterstéen, p. 151. Patrologia,
xiv, pp. 596, 1. 5–597, 1. 5. Abu al-Fida', iv, p. 34, ll. 28–9. Suluk, i, p. 384, ll. 7–8; p. 658, ll. 5–6; p.
671, 1. 14; p. 792, 1. 6; p. 808, pp. 826–7; p. 833, ll. 10–13; ii, p. 20; p. 77; p. 118; p. 119; p. 207; p.
230, note; p. 405, 1. 11; p. 455, ll. 4–10. Nujum (C), vii, p. 84, ll. 1–6; pp. 99–101; p. 269, ll. 14–15;
ix, p. 13, 1. 13; p. 16, ll. 1–2; p. 34, ll. 10–11. Nujum (P), v, pp. 42–3; p. 46; pp. 55–6; p. 117, ll. 4–6;
p. 150, ll. 14–16; p. 153, 1.5; p. 155, ll. 4–8; pp. 206–7; p. 295, ll. 8–9; pp. 295–6; p. 319; pp. 373–4;
p. 380, ll. 15–18; p. 403, ll. 8–10; p. 448, ll. 10–12; pp. 454–5; p. 456, ll. 11–13; p. 457, ll. 6–8; p.
459, ll. 1–2; pp. 469–470; p. 470, ll. 3–5; p. 489, ll. 8–9; p. 492, ll. 17–18; p. 567, ll. 13–15; p. 587,
ll. 2–4; p. 588; vi, 2, 239, ll. 15–21; pp. 264–5; pp. 312–13; p. 333; p. 343; p. 354; p. 363, ll. 2–5; p.
373; p. 384, ll. 4–5; p. 507, ll. 4–9; p. 537; p. 621, ll. 22–3; vii, pp. 6–9; p. 31, ll. 12–13; p. 51, ll. 13–
16; pp. 72,1. 20–73,1. 2; p. 77, ll. 13–15; p. 105; p. 147, ll. 5–7; p. 337, ll. 4–5; p. 388, ll. 1–2; pp.
558, 1. 19–559, 1. 7; p. 662, ll. 3–17; p. 664, ll. 2–3; pp. 718–19. p. 175, ll. 10–13; p. 237, ll. 19–21;
p. 238, ll. 3–4, ll. 6–8; pp. 439, 1. 13–440, 1. 4; p. 444, ll. 11–17; pp. 445–6; p. 607, ll. 18–22; p. 618,
ll. 8–22; p. 619, ll. 10–20; p. 627, ll. 11–19. Manhal, v, fol. 46b, ll. 12–16; fol. 54b, ll. 1–25. Ibn
Shuhba, fol. 39b, ll. 7–10; fol. 55a, l. 24. Ibn Khaldun, v, p. 409, ll.8–9; pp. 457–9; p. 486, ll. 22–3.
Ibn al-Furat, viii, p. 98, ll. 9–12; p. 150, ll. 5–7; pp. 166,1. 26–167,1. 2; p. 170, 1. 8; pp. 173–4; p.
195, ll. 22–3; p. 223, ll. 7–8, ll. 9–11; p. 229, ll. 18–20; ix, p. 56, ll. 3–5; pp. 96–101; p. 106; p. 107,
ll. 18–23; p. 108, ll. 9–21; p. 117, ll. 3–7; p. 115, ll. 14–16; pp. 125–6; p. 130, 1. 3; p. 131; p. 143, ll.
13–25; p. 185, ll. 9–10; p. 192, 1. 4; p. 194, ll. 4–7, ll. 12–15; p. 290, ll. 12–13; pp. 299–301; p. 370,
ll. 16–17. Ibn Iyas, ii, p. 5, 1. 15; p. 10; p. 12, ll. 2–6; p. 13, ll. 21–3; p. 16; p. 25, ll. 20–1; p. 35, 1. 4;
pp. 37–8; p. 41, ll. 14–19; p. 52, ll. 18–20; p. 72, ll. 6–7; p. 73, ll. 2–4; p. 74, ll. 2–3; p. 76, ll. 20–1;
pp. 84–5; p. 86, ll. 9–10; p. 99, 1. 25; p. 152; p. 190, ll. 1–4, ll. 16–19; iii, p. 4, ll. 11–21; p. 6, 1.17; p.
19, ll. 1–6; p. 153, ll. 4–5, ll. 12–16; p. 154, ll. 17–20; p. 155, ll. 7–8; p. 199, ll. 16–20; p. 281, ll. 17–
20; p. 290; pp. 309, 1. 20–310, 1. 2; p. 327, ll. 12–15; p. 431, ll. 2–3; p. 466, ll. 9–16. Durar, i, pp.
482, 1. 21–483, 1. 1. ii, p. 309, 11. 3–5; p. 310, ll. 5–7. When the sources wish to indicate that the
sultan fostered or favoured a certain mamluk, they make use of the term 'ansha'a': Zetterstéen, p.
101, ll. 17–20. Suluk, i, p. 797,1.4; p. 858, 1.13. Nujum (C), vi, p. 319, 1. 14; vii, p. 242, 1. 8. Nujum
(P), v, p. 305, 1. 1. Manhal, i, fol. 163a, ll. 11–12; fol. 164b, ll. 1–4; fol. 191b, ll. 10–11, ll. 18–19;
fol. 194a, ll. 20–2; fol. 196b, ll. 13–14; fol. 210a, 1. 3; ii, fol. 8b, ll. 15–16; fol. 11b, ll. 2–3; fol. 25a,
11. 8–9; fol. 40a, ll. 3–4; fol. 50a, 1. 8; fol. 111b, ll. 12–13; fol. 133b, ll. 15–16; fols. 138a, 1. 23–
138b, 1. 1; iii, fol. 189b, ll. 10–11. iii, 276, 1. 17; p. 284, 1. 4; p. 296, 1. 18.
3 p. 551, ll. 20–2.
1 See Appendix B, to appear in Part III of this article.
2Nujum (P), vi, p. 641, ll. 2–5. See also errata on p.lii in same vol.of Nujum.
3Nujum (P), vi, p. 643, ll. 2–4.
4 pp. 531, 1. 14–533, 1. 10.
5Nujum (P), vii, p. 25, ll. 8–14.
6Nujum (P), vii, p. 452, ll. 6–7.
7Nujum (P), vii, p. 550, ll. 4–11.
8Nujum (P), vii, p. 559, ll. 8–11.
9 p. 348, ll. 22–3.
1 Ibn Iyas, ii, p. 57, 1. 23; pp. 64, 1. 28–65, 1. 17; p. 69, 1. 24.
2Nujum (P), vii, pp. 735, 1. 11–736, 1. 2; p. 760, ll. 5–7. p. 554, ll. 13–14. Ibn Iyas, ii, p. 82, ll. 25–6.
ii, p. 329, ll. 7–15.
3Nujum (P), vii, p. 822, ll. 5–8; p. 828, ll. 15–17; p. 839, ll. 13–17.
4Nujum (P), vii, p. 10, ll. 17–21; pp. 175–180; p. 469, ll. 13–15; p. 473, ll. 15–17; p. 507, ll. 1–4; p.
509, ll. 10–11 and note; pp. 525, 1. 6–527, 1. 5; p. 744, ll. 3–4; p. 320, ll. 11–20. Manhal, viii, fols.
495–9; fol. 496a, ll. 3–5. Ibn al-Furat, ix, p. 102, ll. 21–4. Ibn Shuhba, fol. 115a, ll. 6–7. Tibr, p. 311;
p. 346, ll. 1–4. Ibn Iyas, ii, p. 141, ll. 23–7; p. 144; p. 188; p. 214, ll. 16–18; p. 215, ll. 22–4; p. 280,
ll. 6–7; iii (KM), p. 79, ll. 1–4; p. 150, ll. 1–6; p. 190, ll. 8–10; p. 192, ll. 2–4; pp. 196, 1. 19–197, 1.
9; iv, p. 313; p. 315, ll. 15–18; pp. 385–6 p. 463, ll. 14–19; v, p. 6, ll. 14–16.
5 p. 583, ll. 10–11.
6Nujum (P), vii, p. 475, ll. 1–3, ll. 12–15; p. 476, ll. 1–9.
7Nujum (P), vii, p. 123, ll. 1–9. p. 460, ll. 12–21. Ibn Iyas, ii, p. 99, ll. 1–3, and many other
references.
8Nujum (P), vii, pp. 857–869.
9 p. 188, ll. 15–20; p. 339, ll. 1–9. Ibn Iyas (KM), iii, pp. 331, 1. 23–332, 1. 1.
10 Nujum (P), vii, p. 469, ll. 4–6.
11Nujum (P), vii, p. 494, ll. 1–11.
12 Ibn Iyas (KM), iii, p. 152, ll. 17–21.
1 p. 343, ll. 9–10; see also pp. 494, 1. 16–495, 1. 17.
2 p. 497, ll. 1–13.
3Nujum (P), vii, pp. 518, 1. 15–519, 1. 2.
4 Ibn Iyas, ii, p. 229, ll. 23–4. p. 278, ll. 5–17.
5 Ibn Iyas (KM), iii, p. 72, ll. 5–7; pp. 90, 1. 17–91, 1. 6; p. 213, ll. 13–15; iv, p. 98, ll. 1–4. See also
Nujum (P), vii, p. 471, notes; pp. 528, 1. 17–529, 1. 16. Ibn lyas, ii, p. 137, ll. 8–10; (KM), iii, pp. 93,
1. 24–94, 1. 2. For additional material on the absolute hegemony and unceasing wantonness of the
ajlab, see Nujum (P), vi, pp. 757–8; vii, p. 716, ll. 1–4; p. 717, ll. 1–12; pp. 761, 1. 2–762, 1. 4. p.
191, ll. 18–23; p. 219, ll. 12–18; p. 221, ll. 9–10; p. 223, ll. 1–23; p. 231, ll. 7–15; p. 273, ll. 3–4; p.
301, ll. 6–9; pp. 307, 1. 23–308, 1. 9; p. 324, ll. 16–20; p. 338, ll. 13–15; p. 409, ll. 16–24; pp. 495–6;
p. 608, ll. 1–5; p. 659, ll. 19–20. Tibr, p. 97, ll. 22–5. Ibn Iyas, ii, p. 47, ll. 24–6; p. 54, ll. 1–3; p. 56,
ll. 4–6; p. 57, ll. 13–23. Ibn Iyas (KM), iii, pp. 92, 1. 17–93, 1. 2; p. 95, 1. 5; p. 200, ll. 1–3; p. 214, ll.
6–7; p. 237, ll. 4–7; p. 315, ll. 1–5; p. 341, ll. 16–20; p. 363; pp. 365, 1. 14–366, 1. 2; p. 378, ll. 1–5;
pp. 378, 1. 23–379, 1. 8; p. 427, ll. 20–1; iv, p. 7, ll. 6–8; p. 26; p. 123, ll. 18–23; p. 127, ll. 13–22; p.
156, ll. 2–9; p. 166, ll. 10–18; p. 241, ll. 16–20; p. 277, ll. 22–3; p. 400, ll. 1–15; p. 432, ll. 4–16; p.
464, ll. 18–19; p. 465, ll. 15–22; p. 482, ll. 19–23; p. 484, ll. 15–21.
6 Ibn Iyas, ii, pp. 228, 1. 15–229, 1.1. On the depredations and escapades of the julban see also
M.Mostapha, 'Beiträge zur Geschichte Ägyptens', Z.D.M.G., 1935, pp. 221–3.
7 On the see: Quatremère, Histoire des Sultans Mamelouks par Makrizi, Paris, 1837–1842, vol. i,
part i, p. 11; part ii, p. 158. M.Van Berchem, Matériaux pour un corpus inscriptionum Arabicarum,
Première Partie, Égypte, Paris, 1903, p. 287, p. 543. Gaudefroy-Demombynes, La Syrie à l'Époque,
des Mamelouks, Paris, 1923, p. xxxiii; p. 1; p.c. L.A.Mayer, Saracenic Heraldry, Oxford, 1933, p. 5,
p. 11, p. 60, n. 1, p. 63, etc. A.N.Poliak, Feudalism in Egypt, Syria, Palestine and the Lebanon,
London, 1939, p. 2, p. 6. G.Wiet, L'Égypte Arabe, Paria, 1937, p. 569. J.Sauvaget,' Décrets
Mamelouks de Syrie', Bulletin d'Études Orientales, 1933, p. 24, p. 25. M.Mostapha, 'Beiträge zur
Geschichte Ägyptens', Z.D.M.G., 1935, pp. 212–14.
8 Zetterstéen, p. 14, 1. 7; p. 22, 1. 4; p. 25; p. 27, 1. 11; p. 30, 1. 7; p. 135, 1. 2; p. 164, 1.23.
Zetterstéen calls attention to this spelling in his annotations, p. 1, and cites further examples.
Patrologia, xiv, p. 463, ll. 5–6. Ibn al-Furat, ix, p. 28,1. 8; p. 54, 1. 10; p. 250, ll. 5–6; p. 270, 1. 4.
Ibn al-Furat and Ibn Shuhba (fol. 95a, 1. 22; fol. 114b, 1. 8) are the latest instances of the use of this
spelling.
9Suluk (trsl. Quatremère), ii, pt. 1, p. 159. See also Nujum (C), vii, p. 179, n. 4. Suluk, i, p. 644, n. 4.
10Zubda, pp. 115–16.
1 See, for instance, Manhal, ii, fol. 139b, ll. 4–7; fol. 46a, margin. ii, p. 324, ll. 27–8; p. 328. ii, p.
113, 1. 1. Nujum (P), vii, p. 35, ll. 16–17; p. 398, ll. 8–11; p. 590; p. 691, ll. 8–10; p. 824, ll. 3–10. p.
154, ll. 17–21; p. 378, ll. 11–13. v, p. 168. See also ii, p. 267; p. 270; p. 273; p. 311, p. 312; p. 315;
p. 318; p. 319; p. 324; p. 328; iii, p. 2; p. 6; p. 7; p. 10; p. 12; p. 23; p. 35; p. 36; p. 39; p. 42; p. 44; p.
53; p. 56; p. 60; p. 62; p. 63; p. 64; p. 66; p. 175; p. 230; p. 273; p. 277; p. 280; p. 284; p. 285; iv, p.
214; p. 219; p. 298; vii, p. 235.
2Nujum (P), vi, p. 7; p. 511; p. 512. iv, p. 7, ll. 7–9. Ibn Iyas, v, p. 21, 1. 23.
3 See, for instance, Ibn Iyas, ii, p. 162, 1. 20; (KM), iii, p. 245, ll. 9–13. vii, p. 324, 1. 6. iii, p. 91, ll.
13–16, and many other passages.
4Nujum (P), vii, pp. 74–5; p. 703.
5 On the saqis see: C.I.A., l'Égypte, p. 36. Saracenic Heraldry, p. 11 and n. 1; p. 5; p. 29; p. 31; p. 33,
etc. On the dawadars: Quatremère, vol. i, part i, p. 118. C.I.A., l'Égypte, p. 363. La Syrie, pp. lvii–
lviii. Heraldry, p. 4; p. 5; p. 12; p. 60, n. 1; p. 65; p. 77; p. 87, n. 1; p. 127, etc. On the khazindars: La
Syrie, p. lxi. Heraldry, p. 60; p. 135; p. 142; p. 162; p. 244; p. 248. On the jamdars: Quatremère, vol.
i, part i, p. 11. La Syrie, P.C. Heraldry, p. 5; p. 11, n. 1; p. 14; C.I.A., l'Égypte, p. 390. La Syrie, p.c.
Z.D.M.G., 1935, p. 202; p. 212, n. 4. On the Quatremère, vol. i, part i, p. 159. C.I.A., l'Égypte, p.
195. La Syrie, p. lvii. Heraldry, p. 4; p. 5; p. 13; p. 14; p. 58; p. 65, etc. On the bashmaqdars:
Quatremère, vol. i, part i, p. 100. La Syrie, p.c. Heraldry, p. 5; p. 264.
6 Zubda, pp. 115–16. Al-Maqrizi gives a very confused account of the According to him, al-Ashraf
Khalil specially selected the Kipchakis and to enter the hall called adh-dhahabiya and az-
zumurrudiya, appointed masters of the robe (jamdariya) and cup-bearers (suqat) from among them,
and called them Similarly, from among the burjiya, who belonged to the races of the and the Jarkas,
he appointed armour-bearers jamakdariya, tasters (jashnikiriya), and pages (ushaqiya) ( ii, p. 214).
One gets the impression from this passage that al-Maqrizi attributes to al-Ashraf Khalil the founding
of the The term is, however, encountered fairly frequently before his reign, for example in the days
of al-Malik as-Sa'id Berke Khan (Suluk, i, p. 644; p. 645; p. 650; p. 651; p. 652. Patrologia, xiv, p.
765). We have not encountered the term during Baybars' reign, though and jamdariya are mentioned
(Suluk, i, p. 458). Further, it is not clear why al-Maqrizi restricts the to the offices of jamdariya and
suqat, since it is known that the other offices, which he attributes to the burjiya, were also held by
On the ushaqis see: Quatremère, vol. i, part i, p. 108. C.I.A., l'Égypte, p. 619. Heraldry, p. 148. On
the jashnakirs: Quatremère, vol. i, part i, p. 2. Heraldry, p. 4 and n. 4; p. 5; p. 11; p. 15, n. 5, etc.
1 Ibn al-Furat, vii, p. 95, 1. 20. Nujum (P), v, p. 213, ll. 2–3.
2Nujum (P), vii, p. 518, ll. 8–9. Ibn Iyas, iv, p. 413, 1. 5, 1. 14.
3 Ibn Iyas, v, p. 5, ll. 5–7. Nevertheless, it is not clear why Ibn Iyas remarks that al-Ghawri, unlike
his predecessors, cared for his (Ibn Iyas, iv, p. 358, ll. 11–16), for the were the sultan's favourites at
all periods.
4Durar, i, p. 114, 1. 9.
5 iii, p. 91, ll. 13–16.
6 i, p. 222, 1. 27.
7 i, p. 33, 1. 18.
8Zubda, p. 116.
9 Suluk (trsl. Quatremère), i, pt. 2, p. 159. Unfortunately data on the number of the during the whole
of the period are extremely deficient.
10Nujum (P), vi, pp. 429–439.
11 Ibn Iyas (KM), iii, p. 218,1. 11.
12Ibn Iyas, iv, p. 34, ll. 14–15.
13 Ibn Iyas, v, p. 5, ll. 5–7.
14 See, for instance, Zetterstéen, p. 75; p. 184, 1. 14; p. 188, ll. 3–4. Patrologia, xx, p. 130. Suluk, i,
p. 368, 1. 3; p. 433; p. 743, n. 2; ii, p. 75, 1. 9; p. 156, 1. 12; p. 183, notes; p. 531, 1. 1. Ibn Kathir,
xiii, p. 225, 1. 17, 1. 20. Nujum (C), vii, p. 5, ll. 1–3; p. 126,1. 4. Nujum (P), v, p. 458, ll. 5–7; p. 555;
vii, p. 317, ll. 15–17; p. 318, 1. 2; p. 354, 1. 14; p. 430, ll. 6–9; p. 691,
ll. 11–13. p. 178, ll. 1–6; p. 399, ll. 10–13. Manhal, iii, fol. 4b, 1. 12. Ibn Iyas (KM), iii, p. 324, n. 3;
iv, p. 19, 1. 19; p. 29, ll. 4–8; p. 45, ll. 15–17; p. 50, 1. 16; p. 309, ll. 9–10. Ibn al-Furat, ix, p. 111, ll.
3–9. v, p. 454, ll. 14–15; pp. 458–460; p. 463. p. 314; pp. 343–5; pp. 345–6; p. 348. Zubda, pp. 124–
5.
1 Ibn Iyas, v, p. 2; p. 200.
2 Ibn Iyas (KM), iii, p. 339.
3 Ibn Iyas, iv, p. 467, 1. 3.
4Manhal, iv, fol. 222, ll. 19–20. iii, p. 56, 1. 29; p. 294, ll. 22–3.
5 iii, p. 7, ll. 6–10; p. 18; p. 30; 11. 2, 9, 11, 13, 23. Cf. also ii, p. 342, ll. 14–15. Nujum (P), vi, p.
310, 1. 10. p. 645, 1. 7. The bulk of the material dealing with this question has been gathered in
L'Esclavage du Mamelouk, pp. 28, 29, 33, 57, 58, where it has also been pointed out that the sources
are sometimes lax in their use of the terms ustadh and makhdum, the meanings of which are in some
cases reversed. An outstanding example of such laxity is provided by Baybars who calls Qalaun
throughout his chronicle al-makhdum, though this sultan was the master who purchased and set him
free.
6 Ibn al-Furat, ix, p. 180,1. 17.
7Nujum (P), vi, p. 505, 1. 18, 1. 21; p. 621, ll. 22–23; vii, p. 337, ll. 4–5. Ibn Shuhba, fol. 49b, 1. 3.
iii, p. 61, 1. 8; p. 177, ll. 20–5; p. 277, ll. 14–15.
8Tibr, p. 279,1. 8. ii, p. 68, ll. 1–2. Durar, iv, p. 489,1. 13. iii, p. 11, ll. 3–4; p. 36; p. 285, 1. 10, 1.
29; vi, p. 194, ll. 3–4; p. 224, 1. 14; x, p. 165, ll. 4–5.
9Fawat, i, p. 232. Manhal, i, fol. 18b, 1. 7, 1. 10; fol. 143b, 1. 7; fol. 192a, 1. 17; ii, fol. 190a, 1. 7;
vii, fol. 260b, 1. 3. Ibn Iyas, ii, p. 162, ll. 8–9. iii, p. 174; p. 289, ll. 23–4; p. 295, ll. 24–8; vi, p. 194.
In connexion with notes 1 to 5, cf. material gathered on the same question in L'Esclavage du
Mamelouk. It is likely that the phrase 'thumma intama li-' has a meaning similar to ' thumma bi-
khidmat…' (cf. iii, p. 16; p. 17; vi, p. 231, 1. 8; pp. 211, 1. 29–212, 1. 1). To indicate passing from
rank to rank, from duty to duty, the sources use such expressions as ' tanaqqal fi al-khidam',
'tanaqqal fi wa-l-imriyat', 'taqallabat bihi '. (Manhal, i, fol. 203a, ll. 21–2. Ibn al-Furat, ix, p. 43, 1.
12; p. 277, ll. 12–13, ll. 15–16, ll. 22–3; p. 279, 1. 23; p. 293, ll. 4–5; p. 318, 1. 14; p. 356, 1. 21; p.
447, 1. 11. Ibn Shuhba, fol. 72a, 1. 6. x, p. 279, 1. 17.)
10 iii, p. 82, ll. 27–9. Nujum (P), v, p. 105, ll. 18–19. Manhal, v, fol. 46b, ll. 12–16.
1Nujum (P), v, p. 452; pp. 454–5; vi, pp. 264–5; p. 384, ll. 4–5. Ibn al-Furat, ix, p. 56, ll. 3–5; p. 89;
p. 106; p. 107, ll. 18–23; p. 143, ll. 13–25; p. 162. Ibn Shuhba, fo. 39b, ll. 7–10. Ibn al-Furat twice
mentions mamalik mustakhbara as a body antagonistic to the mushtarawat (ix, p. 88, 1. 3; p. 93); the
context would seem to indicate that the term is synonymous with mustakhdama, but its etymology is
not clear. It is possible that the mustakhbaza mentioned by Ibn Taghribirdi (Nujum (P), vii, p. 450,
notes) have some connexion with al-mustakhbara. For mustakhdamun, cf.: Quatremère, vol. i, pt. i,
p. 64; p. 160. La Syrie, p. xxxiii. Feudalism, p. 55. W.Popper, Glossary to Nujum, vol. vi, pp. xxiii–
xxiv.
2 p. 116, ll. 13–14. The epithet which and after him Poliak (Feudalism, p. 2) apply to the mamluks
of the former sultans is not accurate, for mamalik was the accepted appellation of the Royal
Mamluks, as shown by the references presented above and below, which form but a very small part
of the material supplied on this subject by the sources. It must, however, be indicated that the writer
has eneountered a few isolated cases in which the term seems to apply to the mamluks of the former
sultans (Ibn Iyas, ii, p. 25, ll. 14–16; iii (KM), p. 362, ll. 4–10), but these are extremely rare.
3 Evidence as to the identity of the with the mamluks of the former sultans will be presented in
Appendix B, to be included in Part III of this article.
4 Ibn Iyas (KM), iii, pp. 2, 1. 21–3, 1. 3; p. 73, 1. 13; p. 92, 1. 21; p. 132, ll. 22–3. So far we met the
designation of a mamluk unit by the sultan's first name and not by his surname only in Ibn Iyas
chronicle.
5Nujum (P), vi, p. 770, 1. 9; vii, p. 29, 1. 18; pp. 656, 1. 15–657, 1. 5; p. 666, ll. 2–18. p. 205, ll. 17–
18; p. 443, ll. 8–9; Ibn Iyas (KM), iii, pp. 2, 1. 21–3, 1. 3; p. 92, 1. 21; p. 132, ll. 22–3. The tightness
of the bonds which tied to each other the members of the same may be judged from the fact that
Mamluk history knows of no. single instance of the merger of the mamluks of one with those of
another to form a single Every Mamluk kept its separate existence and disappeared only with the
death of the last of its members.
1 i, p. 95.
2Nujum (P), vi, p. 374, ll. 3–5; pp. 425–6; p. 532; p. 538; vii, pp. 12–19; p. 13, ll. 18–20; p. 16, ll.
16–17; p. 19; p. 29, ll. 18–22; p. 75; p. 87, 1. 9; p. 392; p. 396, ll. 11–12; p. 398, ll. 11–12; p. 756; p.
460, ll. 5–10; p. 461, ll. 12–16; pp. 657, 1. 18–658, 1. 9; pp. 663, 1. 16–664, 1. 13; p. 666, ll. 2–18;
pp. 666–8; p. 672, ll. 5–18; p. 697; p. 699, 1. 2; p. 701, ll. 2–3; p. 720; p. 754, ll. 4–5; pp. 834–8; p.
854, ll. 9–11. p. 183, ll. 10–18; pp. 233, 1. 21–234, 1. 5; p. 371; p. 372; p. 410; pp. 442–1; pp. 520,
1. 20–521, 1. 6; p. 550, ll. 4–9; pp. 550, 1. 20–551, 1. 14; p. 553, ll.19–21; pp. 610–15; p. 643, ll. 8–
9. Ibn Iyas, ii, p. 71, ll. 23–5; p. 76, ll. 16–17; p. 93, ll. 2–3, ll. 5–8; p. 176, ll. 24–6; (KM) iii, p. 73,
1. 13. The fact of belonging to any military group or political unit is denoted in the sources by the
term 'kan min ' or by similar expressions containing the word Nujum (C), vii, p. 30, ll. 12–15; (P) v,
p. 360; p. 403, 1. 14; vi, p. 213, 1.12; p. 524; vii, p. 44, ll.15–16; p. 789, 1. 6. p. 352, 1. 14; p. 596, l.
10; p. 719, 1. 22. Manhal, i, fol. 200b, 1. 18; fol. 201a, ll. 2–3; fol. 203a, 1. 10; ii, fol. 32a, ll. 6–7;
fol. 128a; iv, fol. 110a, 1. 1. Ibn Shuhba, fol. 85a, 1. 14. ii, p. 318, 1. 9; iii, p. 36; p. 41, ll. 25–6; x,
p. 345, ll. 21–2. To denote grouping around a certain individual for common action, political or
otherwise, the sources use the term 'iltaff 'ala' (Ibn Kathir, xiv, p. 363. Manhal, i, fol. 3a, 1. 14; ii, fol.
17b, 1. 17).
3 p. 550, ll. 4–9.
1 pp. 442, 1. 7–444, 1. 10.
VOL. XV. PART 2. 18
1Zubda, p. 116, ll. 14–15. It seems plausible that the sayfiya were so called because during the
Circassian period almost all the Mamluk amirs bore the title of Sayf ad-Din.
2Nujum (P), vii, p. 543, note. Cf. also p. 443, ll. 16–17.
3 p. 334, ll. 4–5. Cf. also Nujum (P), vi, p. 425, 1. 10.
4 See, for instance, Nujum (P), v, p. 216; p. 513, ll. 22–3. Manhal, ii, fol. 59b, ll. 3–6. Ibn Iyas, ii, p.
24, ll. 21–2. iii, p. 209; p. 287, ll. 13–14; vi, p. 231, ll. 22–3; x, p. 165. Durar, ii, p. 196, 1. 16. Ibn
Iyas, iv, p. 209, 1. 14; v, p. 14, ll. 21–2. In fairly rare instances, one encounters the expression al-
mamalik was-sayfiya (Ibn Iyas, ii, p. 25, 11. 14–16; (KM) iii, p. 362, ll. 4–10), but it seems that this
is mere laxity in terminology.
5 p. 443, ll. 15–19.
6 Nujum (P), vi, p. 430, ll. 16–18.
7 p. 448, ll. 7–11.
1 Ibn Iyas, iv, p. 46, ll. 22–3.
2 p. 454, ll. 7–9.
3 az-zaman wa-wafayat ash-shuyukh wal-aqran. Cambridge MS., Dd. 11, 2, fol. 21b, ll. 5–10.
4 Ibid., fol. 24b, ll. 8–13.
5 Ibid., fol. 25b, 1. 13–27a, 1. 4; fol. 29a, ll. 4–10.
6 See the author's 'The Plague and its Effects upon the Mamluk Army', J.R.A.S., 1946, loc. cit.
1 Ibn Iyas, ii, p. 43, 1. 28. On the cutting down of the wages of both the sayfiya and the awlad an-
nas, see Ibn Iyas, iv, pp. 65, 1. 23–66, 1. 1.
2Nujum (P), vi, pp. 425–6; pp. 770, 1. 15–771, 1. 5; vii, p. 13, ll. 18–20; p. 396, ll. 11–12; p. 298, ll.
11–12; p. 672, ll. 13–18; p. 836, 1. 3; p. 837, 1. 7. So far as we know, the term sayfiya appears only in
the Circassian period. In the period, mention is made of al-mamalik as-sayfiya (Suluk, i, p. 821, ll.
2–3), but it is not clear whether this refers to amirs' mamluks.
3 i, pp. 94–5.
4 To be described under the heading 'The ', in Part II of this article.
5 Both lists will be reproduced in full in Appendix A, in Part III of this article.
6Nujum (C), vii, p.15, ll.3–4.
7Zubda, p.116. Suluk, i, p. 638.
1Nujum (C), vii, p. 192, ll. 5–10. Fawat al-Wafayat, i, p. 115. Baybars' army was completely
equipped; during one review, the entire army marched past him, so that it would not be said that a
single soldier had had to borrow anything (Suluk, i, p. 517, ll. 5–17). During that same review, he told
the political envoys that that was the army of the capital only (Suluk, i, p. 519, ll. 6–8), but this was
doubtless great exaggeration, since the first-class troops of the whole kingdom were, for the most
part, concentrated in the capital. Al-Maqrizi is of the opinion that the Mamluks imitated the Ayyubids
in all matters of military organization.
2Nujum (C), vii, p. 179, ll. 15–17. In the passage cited here, it is specified that these were amirs and
office holders; we are unable to determine whether the source meant that all of Baybars' mamluks
were holders of offices and ranks, or whether the figure quoted here refers only to those among them
who did hold such ranks and offices.
3Nujum (C), vii, p. 327, ll. 3–7. Manhal, i, fol. 133a, ll. 20–3. Suluk, i, p. 755, 1. 20. Ibn al-Furat, viii,
p. 97, ll. 21–6. i, pp. 94–5; ii, p. 214. Ibn Iyas, i, p. 20.
4 Baybars Zubdat al-Fikra fi Ta'rikh al-Hijra, B.M. MS., Add. 23, 325, fol. 99a, 1. 13—99b, 1. 1.
5Nujum (C), vii, pp. 327, 1. 15–328, 1. 2.
6 Ibn al-Furat, viii, p. 97, ll. 21–6.,
7 i, P. 95.
8 ii, p. 214.
1Suluk, ii, p. 524, 1. 13–525, 1. 15.
2 i, p. 95.
3 ii, p. 218, 1. 10.
4 On the basis of this figure of 2,000 Royal Mamluks, Poliak draws an unwarranted conclusion as to
the accuracy of the figures of mamluks cited by the sources (Feudalism, p. 6 and n. 7). On numbers
of mamluks see also: G.Wiet, Précis de l'Histoire d'Égypte, Cairo, 1932, vol. ii, p. 242.
5 Cf. L'Esclavage du Mamelouk, pp. 2–8.
6Nujum (C), ix, p. 198, 1. 4; pp. 192–3. It is related of b.Qalaun that he knew his own and his
father's mamluks by name, as well as the rank and pay of each of them (Nujum (C), ix, p.173, ll.12–
14). The Mamluk amir Azdamur al-Mujiri, who was brought before the Khan Ghazan, told him that
the above-named sultan possessed 10,000 Turkish mamluks like himself, but one of the Mongol
Khan's courtiers contested the accuracy of this figure (Zetterstéen, p. 103, ll. 10–20). On virtues in
comparison with subsequent sultans, see Nujum (C), ix, p. 191, ll. 3–7; p. 195, ll. 1–14.
7Nujum (P), v, p. 208, ll. 17–19.
8Manhal, ii, fol. 61b, ll. 17–18.
1Nujum (P), v, p. 420, ll. 13–15. Ibn al-Furat, ix, p. 89, ll. 7–8.
2Nujum (P), v, p. 592.
3Manhal, ii, fol. 72b, ll. 3–4.
4i, p. 95.
5Nujum (P), v, p. 312, ll. 7–11; p. 384, ll. 14–21; vi, p. 47, ll. 4 ff.
6Nujum (P), vi, p. 47, ll. 4 ff.
7 Nujum (P), vi, p. 422, ll. 11–12.
8 For the ratio between the dirham and the dinar in the Mamluk period, see L'Esclavage, du
Mamelouk, p. 42, and E.Strauss, 'Prix et Salaires à l'Époque Mamelouke', R.E.I., 1949, pp. 49 ff.
9Nujum (P), vi, p. 481.
10Nujum (P), vi, p. 773, ll. 6–12.
11Nujum (P), vi, p. 773, ll. 12–17.
12 p. 442, ll. 15–17.
1 i, p. 95, ll. 9–14.
2 B.M. MS., Or. 3026, fol. 115a, ll. 10–11.
3Nujum (P), vi, p. 387, ll. 10–21.
4 The Mamluk sources furnish ample and very convincing information about the terrible decline
which has befallen the whole Egyptian economy since the beginning of the 9th century A.H. till the
destruction of the Mamluk kingdom (this problem is discussed elsewhere by the present writer).
There can hardly be any doubt that this economic decline was one of the main causes for the drastic
reduction in the numerical strength of the Mamluk army.
1 Nujum(P), vii, p. 671, ll. 15–20.
2 pp. 550, 1. 22–551, 1. 10. As-Sakhawi states that Sultan Khushqadam bought many mamluks ( iii,
p. 176, 1. 1).
3 Ibn Iyas, ii, p. 81, ll. 9–10.
4 Ibn Iyas, i, p. 315.
5Ibn Iyas, ii, pp. 13–18.
6 Ibn Iyas, ii, p. 21, ll. 1–2.
7 Ibn Iyas (KM), iii, p. 318, ll. 7–9.
8Ta'rikh Qaytbay at-tarjama ash-sharifa al-ashrafiya, B.M. MS., Or. 3028, fol. 15a, ll. 4–8.
9 az-zaman wa-wafayat ash-shuyukh wal-aqran. Cambridge MS., Dd. ll. 2, fol. 54a, ll. 1–11.
VOL. XV. PART 2. 19
1 See L'Esclavage du Mamelouk, pp. 18–20.
2Zubda, p. 27, ll. 5–7.
3Archives de l'Orient Latin, vol. iiA, p. 91.
4 The figures of the auxiliary armies in the Mamluk kingdom are far less reliable.
 Studies on the Structure of the Mamluk Army—II
1 On the see: Quatremère, vol. i, part i, p. 7; p. 246; part ii, p. 158. C.I.A., 'L'Égypte', p. 458. La
Syrie, pp. xxxii–iv. Z.D.M.G., 1935, p. 219. Feudalism, p. 2; p. 3; pp. 5–10; p. 13; p. 16; p. 19, p. 21;
p. 24; pp. 27–9; p. 31; p. 33; p. 40. J.Sauvaget, La Chronique de Damas d'Al-Jazari, Paris, 1949, p.
46; p. 68.
2 H.A.R.Gibb, 'The Armies of Saladin', Cahiers d'Histoire Égyptienne, Cairo, 1951, p. 305. This
article became available to the writer after the present paper was ready for publication, and therefore
it is only sporadically referred to.
3 Ibn al-Athir, xi, p. 349; p. 369; xii, p. 33. Abu Shama, Kitab ii, p. 179, ll. 17–18; p.
180, 1. 4.
4Suluk (trsl. Quatremère), i, part 2, pp. 200–2.
5 See n. 1 above, and n. 1, p. 449.
6B.S.O.S., x, p. 872.
7 iv, p. 16. p. 245, ll. 14–16.
1Suluk, i, p. 122, ll. 6–7.
2 See, for instance, the information given for the year 646 (Suluk, i, p. 330).
3 Ibn al-Furat, vii, p. 7, ll. 1–2.
4 Zetterstéen, p. 158, ll. 5–6; p. 172, 1. 10; p. 220, ll. 10–13.
5 Ibn al-Furat, ix, p. 451, 1. 1.
6 Zetterstéen, p. 164, ll. 10–14. cf. Glossary to Nujum, vol. vi, p. L.
7 i, p. 87, ll. 38–9.
8 b. Abi an-Nahj as-Sadid (in Patrologia Orientalis), xx, pp. 221, 1. 5–222, 1.
2.
9Ibn al-Furat, ix, p. 451, ll. 10–11.
10Manhal, ii, fol. 16a, ll. 1–2.
11 i, p. 250, ll. 4–5.
12 Zetterstéen, p. 219, ll. 14–15.
13Zetterstéen, p. 167, 1. 5.
14 Zetterstéen, p. 172, 1. 10.
15Zetterstéen, p. 200, ll. 16–17.
16 Zetterstéen, p. 219, ll. 7–8.
17Zetterstéen, p. 221, ll. 1–4.
18Suluk, ii, p. 403, ll. 13–14.
19Nujum (C), vi, p. 376, ll. 12–13; vii, p. 43, 1. 8. Manhal, i, fol. 3b, 1.15. cf. also Ibn al-Furat, viii,
pp. 136, 1. 23–137, 1. 1; ix, p. 245, ll. 3–8.
1 Zetterstéen, p. 17, ll. 9–10; p. 161, ll. 5–6, ll. 14–16.
2 Zetterstéen, p. 161, ll. 20–3.
3 Zetterstéen, p. 228, ll. 9–12. For details on the position of muqaddamu who were not
mamluks, see Zetterstéen, p. 220, ll. 10–13. Durar, iv, p. 381, ll. 9–10, and many other passages. On
the muqaddamu in the early Mamluk period, and their civil and military functions, see
Zetterstéen, p. 24, ll. 19–20; p. 30, 1. 5; p. 32, ll. 13–14; p. 33, 1. 10; p. 41, 1. 13; p. 42, 1. 10; p. 43,
1. 8; p. 51, 1. 12; p. 54, 1. 12; p. 106, 1. 23; p. 107, 1. 13; p. 131, 1. 1; p. 144, 1. 14; p. 190, ll. 3–4; p.
218, ll. 9–10. Patrologia, xii, p. 166, ll. 6–7; xiv, p. 487, 1. 7; p. 534; p. 596, 1. 5; xx, p. 41, ll. 1–2.
Ibn Kathir, xiii, p. 264, ll. 12–14; p. 290, ll. 20–21; xiv, p. 212, 1. 5; p. 222, 1. 27; p. 240, 1. 12.
Suluk, i, p. 493, ll. 1–2; p. 507; p. 518, ll. 5–6; p. 534, ll. 16–17; ii, p. 499, 1. 3, 1. 9. Nujum (C), vi, p.
125,1. 9; vii, pp. 160, 1. 17–161, 1. 2; viii, p. 102, l. 15; p. 173, 1. 1; p. 180, 1. 26; p. 213, ll. 12–13.
Durar, iii, p. 229, 1. 5. Ta'rikh Bayrut, p. 58, ll. 14–15; p. 95, ll. 6–15. ii, p. 112,1. 33; p.
209,11. 37–9. vii, p. 159, 1. 15. In the early Mamluk period, one often encounters a military
eategory designated by the term mufradiya or mafarida, sing. mufradi. Its nature is not known to us,
and it is not clear whether it belonged to the or to another unit. It seems, however, that its
members ranked as honoured privates or lower amirs. They are mentioned as fairly regular
participants in various official ceremonies together with the amirs, the muqaddamu and others
(Suluk, i, pp. 507, 518, 520, 612). After a long interval, they suddenly reappear in 791 (Ibn al-Furat,
ix, p. 166), then vanish again. cf. also Quatremère, vol. i, part 1, p. 187.
4Suluk, ii, p. 20, 1. 20. cf. also Zetterstéen, p. 132, ll. 3–6.
5Suluk, i, p. 673, ll. 16–17.
6 'The Wafidiya in the Mamluk Kingdom', Islamic Culture, Jubilee Number, 1951, pp. 89–104, cf.
also Quatremère, vol. ii, part i, p. 245.
7 Zubda, p. 115. ii, p. 216. iv, p. 16. ii, p. 110, ll. 22–4.
1Suluk, i, p. 838; p. 922, 1. 21; p. 930, 1. 9; p. 932, 1. 17; p. 949, 1. 6; ii, p. 33, 1. 6; p. 63, l. 10; p.
90, ll. 16–19; p. 109, l. 11, 1. 16; p. 139, l. 15; p. 236, 1. 2. Nujum (P), vi, p. 255, 1.18. Ibn al-Furat,
vii, p. 169. A term of frequent occurrence in the sources is alzam, sing. lazim ('retinue, escort, troops
attached to an amir'). The exact position of the alzam is not entirely clear. Thus, 'mamalikuhum wa-
ajnaduhum wa-alzamuhum' (Ibn al-Furat, viii, p. 180, 1. 20), or ' wa-alzam'(Suluk, ii, p. 69, ll.
11–12). For additional material on this term, see Suluk, i, p. 346, ll. 1–2; ii, p. 34, 1. 6; p. 70, 1. 1; p.
385,1. 2. An-Nahj aa-Sadid (in Patrologia Orientalis), xiv, p. 574, ll. 5–6. Abu al-Fida', iv, p. 81, 1.
15; p. 96, 1. 22. Nujum (C), viii, p. 221, ll. 4–5. Nujum (P), vi, p. 243, 1. 3. Ibn al-Furat, vii, p. 101, 1.
17. Ta'rikh Bayrut, p. 165, 1. 6. Durar, i, p. 540, 1. 4; iv, p. 288, 1. 11.
2Suluk, ii, p. 146, 1. 2. Ibn al-Furat, xi, p. 444, 1. 15; p. 451, ll. 10–11. i, p. 250, ll. 4–5. See
also Feudalism, p. 5.
3 Ibn al-Furat, ix, p. 477, ll. 14–15; p. 450, ll. 2–3. Duwal al-Islam, ii, p. 15, ll. 7–8.
4 See, for instance, Abu al-Fida', iv, p. 57, 1. 25. Nujum (C), viii, p. 157, 1. 8; p. 158, 1. 6. Suluk, i, p.
512, 1. 13; p. 518, 1. 5; p. 527, 1. 9; p. 743, 1. 4; p. 768, n. 1; p. 846, 1. 1; ii, p. 49, l. 18; p. 146, 1. 2;
p. 156, 1. 8; p. 356, ll. 5–6. Ibn al-Furat, vii, p. 5, 1. 14; ix, p. 123, 1. 2; p. 186, 1. 20; p. 348, ll. 7–8.
5 On the battlefield the Mamluk army was divided into three principal bodies: the centre (al-qalb),
the right wing (al-maymana) and the left wing (al-maysara). The centre was the most important part,
and in it the best troops were concentrated.
6 See especially Feudalism, pp. 23–5; p. 27, p. 68. History of the Land-tenure Relations in Egypt,
Syria and Palestine in the late Middle Ages and Modern Times (in Hebrew: Toldoth ha-
qarqa'iyim suriya ve-eretz yisrael be-sof yemey ha-beynayim uvazzeman )
Jerusalem, 1940, pp. 20 ff. R.E.I., 1935, pp. 239–241.
1 See references listed in preceding note.
2 Zetterstéen, p. 45, ll. 16–19. The author of the text, unfortunately, does not return to the subject, in
spite of a promise to do so.
3Nujum (C), viii, p. 95, ll. 12–15.
1Nujum (C), viii, pp. 95–9. cf. also i, pp. 87, 1. 23–88, 1. 3.
2 i, p. 90, ll. 1–11; pp. 90, 1. 35–91, 1. 1. Ta'rikh Bayrut, pp. 95,1. 9–96,1. 2. Suluk, ii, p. 146,
ll. 5–6, ll. 13–17. For data on the rawk until 806 and on the decline of Egypt from that date
onward, see i, p. 91, ll. 12–16. For additional data on the rawks, see i, p. 82, 1. 11; pp.
87–91; ii, pp. 206, 1. 39–207, 1. 3. Zetterstéen, p. 164, ll. 10–14. Ibn Kathir, xiv, p. 69, ll. 18–22; p.
75, ll. 25–6; p. 865. Durar, i, p. 251. iii, p. 302, ll. 15–17; p. 387, ll. 25–35; p. 436, ll. 6–7. An-
Nahj as-Sadid (in Patrologia Orientalis), xiv, p. 601, ll. 1–2; xx, p. 236, ll. 1–3; pp. 255, 1. 1–256, 1.
1. Suluk, i, pp. 841–6; p. 858, ll.16–18. p. 865, ll. 5–7; ii, pp. 146–7; pp. 149–150; pp. 154–7; pp.
174–5; pp. 176–7; p. 264, ll. 4–5. Zetterstéen, pp. 160, 1. 25–161, 1. 1. Manhal, v, fol. 55a, ll. 7–19.
Abu al-Fida', iv, p. 37, ll. 1–3. Nujum (C), ix, p. 36, ll.1–2; pp.42–55 p.163, ll. 2–3; p. 177, ll. 4–19;
p.296, ll. 9–11.On incomes from feudal estates, see Suluk, ii, p. 146, ll. 5–6. i, p. 88, ll. 15–20;
pp. 90, l. 36–91, 1. 1; p. 97, ll. 1–22; ii, pp. 216–17. For the rawk, cf. also: Quatremère, vol. ii, part ii,
p. 65. Feudalism, pp. 5–23; p. 27; p. 68. Sauvaget, Jazari, p. 69.
3Manhal, v, fol. 203b, ll. 3–4.
4 On diwan al-badal see: La Syrie, p. xlv. Feudalism, p. 29.
1 ii, p. 219.
2 Durar, iv, p. 361, ll. 6–10.
3 Ibn al-Furat, ix, p. 165.
4 Ibn al-Furat, ix, p. 165.
5 Ibn al-Furat, ix, p. 163.
6Nujum and Ibn al-Furat, in the chronicles for the year 791. cf. also references listed in n. 6, p. 455.
7 Ibn al-Furat, ix, p. 163.
8Zubda, p. 116. unlike his predecessors, is said to have relieved the from guarding the
Cairo citadel in the absence of the army's main body (Ibn Iyas, v, p. 48, ll. 1–10).
1Nujum (P), vi, p. 389.
2Nujum (P), vi, p. 739.
3Nujum (P), vi, pp. 739–740.
4 ii, p. 222.
5Suluk (trsl. Quatremère), ii, part 1, p. 200. ii, p. 216, 1. 1.
6 On the status of the during the Circassian period, see: Nujum (P), v, p. 120, ll. 20–1; p. 407, ll.
20–1; vi, p. 55, ll. 18–20; pp. 70–1; p. 72, 1. 15; p. 385, ll. 3–5; pp. 388–9; pp. 391–2; p. 394, ll. 8–
10; pp. 481–2; p. 483, ll. 13–14; pp. 738–9; p. 740, ll. 4–6. pp. 697, 1. 18–698, 1. 5. Manhal,
viii, fol. 441b. Ibn al-Furat, ix, p. 11, ll. 8–10; p. 150, 1. 12; p. 160, ll. 20–1; p. 163, ll. 5–10; p. 165,
ll. 6–11; p. 350, 1. 20; pp. 362, 1. 25–363, 1. 3. Ibn Iyas, ii, p. 22, ll. 10–13; p. 47, ll. 3–6; p. 101; p.
104; p. 110, ll. 2–3; (KM) iii, p. 20, ll. 4–14; pp. 24, 1. 21–25, 1. 1; pp. 40, 1. 24–41, 1. 4; p. 323, ll.
6–9; v, p. 26, ll. 6–12; p. 27, ll. 1–5.
1The term 'an-nas', however, is also found in its ordinary meaning, i.e.'the public, the people'. The
two meanings should not be confused; they occur in the sources with approximately equal frequency.
For the second meaning, see, for instance: Suluk, i, p. 789, 1. 6. Nujum (P), v, p. 13, 1. 16; p. 16, 1.
19; p. 32, 1. 15; p. 40, 1. 1; p. 312, 1. 10; p. 322, ll. 12–13; p. 404; p. 407, 1. 16; p. 418, 1. 1; p. 522,
1. 8; vii, p. 684, 1. 5, 1. 7. Ibn Iyas (KM), iii, p. 53, 1. 22; v, p. 75, 1. 14. Cf. also Z.D.M.G., 1935, pp.
217–18. Glossary to p. xxix. Feudalism, p. 10; p. 14; p. 20; p. 33; p. 38; p. 40; p. 54.
2 vi, p. 51, ll. 10–12. p. 258, ll. 19–21. ii, p. 216, ll. 18–24.
3 ii, p. 216, ll. 18–24. cf. also Nujum (C), viii, p. 174, 1. 17. Ibn Iyas, ii, p. 177, 1. 20; iv, p. 47,
ll. 11–17; p. 195, ll. 2–3; p. 363, 1. 9; v, p. 26, 1. 1; p. 176, 1. 3; p. 347, 1. 15; p. 395, 1. 4; p. 402, ll.
9–10.
1 See, for instance, Zetterstéen, p. 157, 1. 22. Suluk, i, p. 770. Nujum (C), ix, p. 262, ll. 6–8. Nujum
(P), v, p. 206, ll. 21–3; p. 301, ll. 13–14; vi, p. 145, ll. 1–2; p. 173, ll. 7–14; p. 393, ll. 9–10; p. 475, 1.
14; vii, p. 225, ll. 2–5; p. 625, ll. 16–18. p. 107, ll. 6–8; p. 318, ll.10–16. Manhal, i, fol.
147b, ll. 8–10; vii, fol. 354a, ll. 10–11. Ibn al-Furat, ix, p. 165, ll. 15–16; p. 173, ll. 16–17; p. 239, ll.
3–4; p. 275, ll. 14–20; p. 276, ll. 1–5; p. 477, ll. 15–16. Tibr, p. 354, ll. 9–11. Ibn Iyas, ii, p. 40, ll.4–
5; p. 104, ll. 9–23; p. 143, 1. 11; iii, p. 23, ll. 8–12; iv, p. 354, ll. 17–19. Durar, i, p. 115, 1. 1; ii, pp.
50–1. iii, pp. 100, 1. 17–101, 1. 6; p. 106, 1. 29. The governors of the provinces of Alexandria,
Kerak, Jerusalem, etc., were sometimes non-mamluks as well ( p. 603, 1. 14. Ibn al-
Furat, ix, p. 297, ll. 9–11. i, p. 66, ll. 19–21; p. 226, ll. 13–18; iii, pp. 100, 1. 27–101, 1. 6; p.
102, ll. 12–13; p. 106, ll. 26–7, 1. 29; p. 131, ll. 12–14).
2Nujum (P), v, pp. 159–160.
3Suluk, ii, p. 228, ll. 14–18. Ibn Iyas, ii, p. 21, ll. 5–10; p. 137, ll. 2–4; v, p. 48, ll. 1–10. In the
relevant chapter of our work on the Mamluk army we discussed in detail the curtailments and cuts in
the payments to the and the awlad an-nas; see also: Ibn Iyas, ii, p. 118, 1. 9; p. 174, ll. 28–9;
(KM) iii, p. 130, ll. 3–4; p. 266, ll. 11–12; p. 432, ll. 6–7. Ibn Shuhba, fol. 18a, ll. 24–5. Ibn al-
Furat, ix, p. 79, ll. 11–13; p. 219, ll. 11–13; p. 350, ll. 19–21; p. 379, ll. 8–9.
4 Ibn Iyas, ii, p. 21, ll. 5–6; p. 212, 1. 11; iv, p. 150, 1. 13; v, p. 48, ll. 10–11.
5 See, for instance, Nujum (P), vii, p. 140, ll. 4–5; p. 850, ll. 7–9. p. 175; ll. 10–13; p. 616,
1. 1; p. 681, 1. 8. Ibn Iyas, v, p. 43, ll.3–4; as well as Feudalism, p. 29, and n. 10.
1 p. 175, in the notes.
2 See, for instance, Nujum (C), vii, p. 236, 1. 13. Nujum (P), vi, p. 11, ll.7–10; p. 854, ll. 5–6; vii, p.
339, ll. 6–8. Ibn Shuhba, i, fol. 71b, ll. 1–2; fol. 89a, ll. 11–12. Durar, i, p. 27,1. 17; ii, p. 91, ll.
6–7. ii, pp. 97–8; iii, p. 120, ll. 1–2; v, p. 155, ll. 1–2; vii, p. 147; viii, pp. 171, 1. 28–172; 1. 6.
The terms indicating that a civilian or an amir's son belonged to, or joined, the were: tazayya bi-
ziy al-jund, labisa ziy al-jund, kan bi-ziy al-jundiya, dakhala fi al-jundiya, etc. (Nujum (C), vii, p. 27,
ll. 15–16. Nujum (P), vi, p. 292, ll. 1–2; p. 328, 1. 1; p. 853, ll. 20–1. Manhal, i, fol. 103a, 1. 22. Ibn
al-Furat, ix, p. 283, 1. 17; p. 471, ll. 3–5. Ibn lyas, v, p. 105, ll. 1–6. Durar, i, p. 231, ll. 16–19; p. 239,
ll. 17–19; ii, p. 86, 1. 10; iv, p. 280, 1. 7; p. 345, ll. 16–19; pp. 370, 1. 21–371, 1. 1; p. 489, ll. 9–10.
Tibr, p. 48, 1. 15. i, p. 146, 1. 4; iii, p. 205, 1. 18; v, p. 36; viii, p. 230, ll. 18–20; x, p. 318, ll. 8–
9).
3Nujum (P), v, pp. 159–160; vii, p. 293, notes. p. 142, ll. 2–3.
4Nujum (P), vi, p. 683. Manhal, i, fol. 55a, ll. 2–13; fols. 179a, 1. 21–179b, 1. 2. Ibn Iyas, ii, p. 15, ll.
11–14.
5 Ibn Iyas, ii, p. 14, ll. 25–6, and many other passages.
6 pp. 706, 1. 10–707, 1. 15. Ibn Iyas(KM), iii, p. 66, ll. 12–16; p. 150, ll. 6–14; p. 152, ll.
21–4.
7 For material on the asyad and their status in the Mamluk kingdom, see Suluk, ii, p. 490, 1.13.
Nujum (P), v, pp. 216–17; p. 228, ll. 18–19; p. 229, 1. 3; p. 282, 1. 3; p. 320, 1. 21; p. 397,
ll. 15–17; p. 505, 1. 12; vi, p. 266, ll. 2–3; p. 432, ll. 5–7; p. 514, ll. 7–8; p. 545, ll. 8–9; p. 772; vii, p.
320, 1. 1; p. 426, 1. 1; pp. 508–9; p. 511; p. 644, ll. 1–8; pp. 664–5; p. 678, ll. 8–9. p. 149,
1. 1; p. 305. Ibn al-Furat, ix, p. 56, ll. 15–18; p. 91, ll. 18–21; p. 176, l. 21. Ib'n Iyas, ii, p. 15, ll. 1–2;
p. 60, 1. 4; p. 79; p. 108, 1. 1; p. 113, ll. 6–10; iii, p. 188, ll. 10–12; p. 195, 1. 12; iv, p. 9, 1. 7; p. 399,
ll. 15–23; p. 406, 1. 9. iii, p. 53, ll. 8–10; p. 87, ll. 6–7; p. 124, ll. 2–3; p. 201; p. 217; vi, p. 73;
vii, p. 274. Zubda, p. 111, ll. 5–12. xiii, p. 167, ll. 16–19. One of the surprising facts connected
with the Circassians is that even the later rulers, who were well aware of the fate met by sultans' sons
who were appointed to the sultanate by their fathers, grew no wiser from experience and followed the
same policy, knowing full well that their own sons would be deposed. This fact arouses the
amazement of Ibn Taghribirdi, who can find no explanation for it. He says in one passage: 'We have
seen the same retribution meted out time and again from the day when Barquq deposed
down to our own day. All are made to drink the same cup by the atabak, and the beverage contained
therein is prepared by the mamluks of their fathers. This matter has already been discussed by us in
many places, but silence is more fitting'. (Nujum (P), vii, p. 419, ll. 2–6). Elsewhere, he states that he
fails to understand why the sultan appoints his son as successor at the last minute, knowing as he
does with certainty that his son will be dealt with as he himself dealt with the son of the previous
sultan (Nujum (P), vii, p. 394, ll. 9–13. See also Nujum (P), v, pp. 228–230; vii, pp. 394–6.
p. 134, ll. 1–2).
1 p. 265, 1. 15. iv, p. 24, ll. 6–16; p. 63, ll. 11–12; p. 64, ll. 3–9. ll. 9–15.
2Zubda, p. 134, ll. 11–13.
3Nujum (C), viii, p. 261, 1. 11. Nujum (P), v, p. 229, 1. 4.
4 See, for instance, Zetterstéen, p. 132, ll. 3–6; p. 163, 1. 21; p. 168, 1. 23; p. 170, 1. 8; p. 177, 1. 1.
Suluk, ii, p. 176, ll. 6–7.
1 ii, p. 216, ll. 1–4, ll. 13–14. iv, p. 62, ll. 12–16. Ibn Kathir, xiv, p. 318, ll. 15–17. In
connexion with the repartition of fief incomes between the amir and his mamluks, in the proportion
of two-thirds to one-third, it is interesting to note that Amir Barquq willed one-third
of his possessions to his freed mamluks and his freed slave-women (Manhal, v, fol. 34a, ll. 18–19).
2 Ibn Iyas, ii, p. 62, 1. 12. iii, p. 36; p. 276, 1. 12; p. 284, 1. 13.
3Nujum (P), vii, p. 581, ll. 17–18; p. 597, ll. 18–19; p. 616, ll. 3–5; p. 772, 1. 1; p. 798, notes.
iii. p. 36; p. 276, 1. 12; p. 284, 1. 3; x, p. 205, 1. 28. Al-abwab are called in the sources also al-
abwab ash-sharifa or al-abwab al-'aliya (Abu al-Fida', iv, p. 62, ll. 13–14; p. 79, 1. 3; p. 96, 1. 27.
Ibn Iyas, iv, p. 46, 1.5; p. 119, 1. 14; p. 130, 1. 21. vi, p. 60, ll. 18–20). The houses of the amirs were
also called al-buyut al-karima ( iv, p. 60, ll. 18–20).
4 cf. Ibn Iyas, ii, p. 218.
5 Ibn Khaldun, v, p. 472, 1. 19. Only additional data will make it possible to reach more definite
conclusions on this point.
6Nujum (P), vi, p. 478, ll. 6–11.
7 See, for instance, Manhal, ii, fol. 61a, ll. 17–21. An-Nahj as-Sadid (in Patrologia Orientalis), xiv,
p. 535, ll. 2–4. cf. also section on the sayfiya in this chapter.
8 Zetterstéen, p. 132, ll. 3–6.
1Nujum (P), v, p. 219, ll. 19–20.
2 i, pp. 87, 1. 37–88, 1. 3.
3Manhal, iii, fol. 132b, ll. 4–17. For additional information on this amir and his peculiar ways see
iii, pp. 277, 1. 13–278, 1. 15.
4 For rank, cf. Quatremère, vol. ii, part i, p. 14; p. 15. Heraldry, p. 5; p. 26, p. 34, n. 4. Glossary to
Nujum, vol. v, p. xxvi.
1 iv, pp. 60, 1. 11–63, 1. 3; p. 61, ll. 12–15; pp. 61, 1. 20–62, 1. 5. pp. 264–5. cf. La Syrie, p. ciii.
2 Ibn Iyas, v, pp. 42–3.
3Suluk, ii, p. 100, ll. 4–5. ii, p. 390, ll. 1–2.
4Suluk, i, p.687.
5Durar, i, p. 388, 1. 1.
6Durar, iii, p. 258, ll. 2–3.
7Manhal, iii, fol. 178b, 1. 17. Nujum (P), v, p. 135, ll. 6–7.
8Nujum (P), v, p. 301, ll. 11–13.
9Zubda, p. 132, 1. 20.
10 Abu al-Fida', iv, p. 57, ll. 28–9.
11Nujum (P), v, p. 208,1. 19. Manhal, i, fol. 193b, ll. 8–9.
12Zubda, p. 148, ll. 12–14. Durar, iv, p. 438, ll. 12–13. Ibn Iyas, i, p. 219. Nujum (P), v, p. 200.
13Nujum (P), v, pp. 460–1.
14Nujum (P), v, p. 63, 1. 16.
1Nujum (P), vi, pp. 442–3. Manhal, viii, fol. 390b, ll. 6–9. For the expenses of other amirs, see ii, p.
169, ll. 27–30. Manhal, ii, fol. 412, ll. 8–9.
2Manhal, viii, fol. 367b, ll. 13–15.
3 Ibn Iyas, i, p. 208.
4 Ibn al-Furat, ix, p. 59.
5Ibn al-Furat, ix, p. 148.
6 Nujum (P), v, pp. 80–1.
7Nujum (P), vi, p. 141.
8Manhal, ii, fol. 33a, ll. 16–17.
9Manhal, ii, fol. 33b, ll. 4–5. In another passage, the same historian reports that the mamluks owned
by Taghribirdi exceeded 300 (Nujum (P), vi, p. 141, ll. 5–6), and that when he served as governor of
Damascus, their number reached 970 (Manhal, ii, fol. 125b).
10 p. 659, ll. 17–18. This statement of Ibn Taghribirdi should not, of course, be taken at its face
value, but it contains undoubtedly a substantial grain of truth.
11 The historians of the Circassian period call the sultans of the first Mamluk period muluk as-salaf
or al-muluk as-salifa. To them they attribute all noble virtues, and set them in contrast to the sultans
of their own day.
12 pp. 566, 1. 20–567, 1. 1. The term muluk referred not only to sultans, but also to important amirs,
cf. for instance, Nujum (P), v, p. 611, ll. 10–13.
VOL. xv. PART 3. 34
1 p. 142, ll. 17–18. Amir Baraka had an ustadar who was an Amir of a Thousand, 'an
unheard-of thing' (Nujum (P), v, p. 311, ll. 5–8).
2 Ibn Iyas, v, p. 118, ll. 1–3.
3 iii, p. 58, 1. 25.
4 ii, p. 328, ll. 17–18.
5Nujum (P), vii, p. 478.
6Nujum (C), vii, p. 272, ll. 5–6, ll. 7–8. Ibn al-Furat, ix, p. 36, 1. 7. Manhal, ii, fol. 12b, ll. 12–13; fol.
62b, ll. 20–1. p. 389, ll. 10–12; p. 391, ll. 9–17; p. 577, 1. 22.
7 See now Gibb, op. cit.
8 The was the military unit which the sultan or amir would lead during the military expedition, or
during processions and parades. Al-Maqrizi ( i, p. 86) alone among Mamluk sources defines this
term, but his definition is unsuited to the actual usage of the term during the Mamluk period. cf.
Quatremfère, vol. i, part i, pp. 34–5; part ii, pp. 271–2. Glossary to Nujum, vol. vi, p. xxxix. Gibb, op.
cit., pp. 308–9.
1 i, p. 86, ll. 26–34. Gibb, op. cit., p. 309, n. 31. The whole question of the in ad-Din's army is
comprehensively dealt with by Gibb.
2 i, p. 86, ll. 34–9; p. 87, 1. 1. The same list appears in Suluk, i, p. 75, ll. 3–7, but is much more
corrupt than that given in Both should be read together in order to obtain a correct picture. On the
kinaniya (as well as kitamiya?) mentioned together with the in the above list, see also Nujum (C), vi,
p. 17. Suluk, i, p. 150, 1. 12. For the etymology of the term see Blochet's view (Patrologia
Orientalis, xii, p. 494, n. 4).
3Nujum (C), vi, p. 12,1. 4. Suluk, i, p. 76, ll. 1–3. ii, p. 120, ll. 13–15. Ibn p. 225, ll. 10–14. Al-
Mashriq, 1935, p. 212, ll. 14–18. Qawanin ad-Dawawin, p. 356, 1. 2.
4 Al-Mashriq, 1935, p. 212, ll. 14–18. Ibn p. 225, ll. 10–14.
5 Ibn Khaldun, v, p. 424, ll. 22–5.
6Suluk, i, p. 509, ll. 2–5.
7 Ibn al-Furat, viii, p. 2, 1. 3.
8Suluk, i, p. 722, 1. 8.
9Ta'rif, p. 89, ll. 3–4.
1 To put it in other words: if the are mamluks this means that ad-Din had by far a greater
number of mamluks than Najm ad-Din Ayyub, the founder of the regiment, from which the
Mamluk kingdom sprang up (see below, p. 474). The writer did not succeed in establishing the
connexion between the and the under ad-Din.
2Feudalism, p. 3, n. 4. The whole question concerning the term 'amirs' horsemen' is discussed in pp.
471–5.
3 Baybars Zubdat al-Fikra, fol. 157a, 1. 17.
4 Ibid., fol. 158a, ll. 6–7.
5 ad-Din ibn'Abd Sirat Baybars, B.M. MS., Add. 23, 331, fol. 41a, ll. 3–5.
6 Ibid., fol. 41b, ll. 15–17. See another interesting example: ibid., fol. 42a, ll. 7–9.
7Ta'rikh, Bayrut, pp. 96–98. cf. also p. 79, ll. 6–10; p. 89, 1. 11, 1. 14; pp. 92–4.
8 cf. references cited in the preceding note, as well as some of those given by Poliak, Feudalism, p. 3;
R.E.I., 1935, p. 247, n. 4.
1 vii, p. 159, ll. 15–16. The writer is not certain as to the meaning of in Ibn Kathir, xiv,
p. 287, 1. 26, and Nujum (P), vii, p. 487, ll. 23–5.
2 Ibn Iyas, v, p. 205, ll. 5–6.
3Nujum (P), v, p. 536, 1. 7; vi, p. 7, 1. 14. iii, p. ll, 1. 4.
4 Zetterstéen, p. 168, 1. 3. Nujum (P), v, p. 91, 1. 6; p. 523, 1. 22; vi, p. 803, 1. 5. Ibn Shuhba,
fol. 70b, 1. 25. Manhal, iv, fol. 110b, 1. 3. Ibn Iyas, ii, p. 25, 1. 4; p. 48, 1. 22; p. 92, 1. 28; (KM) iii,
p. 296, 1. 1; p. 314, n. 2; p. 374, 1. 23; iv, p. 105, 1. 6; p. 450, 1. 19.
5 Thus: 'wa-lam li-Lajin hadha imrat 'ashara, wa-mata wa-huwa jundi' (Manhal, v, fol. 56a, ll.
20–1); cf. also: Zetterstéen, p. 199, 1. 12. iv, p. 63, 1. 11. Fawat al-Wafayat, i, p. 99. Manhal, i,
fol. 195b, 1. 1; fol. 196b, ll. 6–7; iii, fol. 18a, ll. 19–20. That jundiya referred to the rank of private
may be inferred from the following examples: wa-intaqala Barquq min al-jundiya ila imrat
daf'atan (Manhal, ii, fol. 61a, ll. 21–2) and ' ayyamuhu fi al-jundiya ila an ta'ammara
'ashara' (Manhal, v, fol. 16a, ll. 3–4); cf. also Zetterstéen, p. 101. Nujum (C), ix, p. 297. Ibn al-Furat,
ix, p. 100, ll. 17–18; p. 319, ll. 8–9; p. 325, 1. 2; p. 419, ll. 4–5. Nujum (P), v, p. 304, 1. 19; pp. 332–
3; p. 450, ll. 12–16; p. 575, ll. 20–2; vi, p. 536, 1. 20; p. 675, 1. 15; vii, p. 301, 1. 5. Manhal, ii, fol.
176b, ll. 11–12; viii, fol. 420b, ll. 14–15. Ibn Shuhba, fol. 73b, ll. 3–4; fol. 78b, 1. 13. Durar, i, p.
216, ll. 1–2; p. 514, 1. 11. iii, p. 60, 1. 26; vi, p. 215, 1. 22. We may point out here that jundi and
jundiya, meaning 'private soldier 'and 'rank of private' respectively, ought not to be confused with
ajnad, meaning, as stated above, 'soldiers of the '. For the term jundi, cf. C.I.A., 'L'Égypte,' p. 544. La
Syrie, pp. xxxiv–v. Heraldry, p. 5. Glossary to p. xxv.
6Ta'rif, pp. 73–4. iv, pp. 14 ff. pp. 244–5. i, pp. 95 ff.; ii, pp. 215 ff. ii, pp. 110–113. Zubda, pp.
111–116. Nujum (P), vi, pp. 386–7.
7Nujum (P), v, p. 204, 1. 2. Ibn al-Furat, ix, p. 393, 1. 15. For muqaddam alf, taqdimat alf, cf.:
Quatremère, vol. i, part i, p. 26; p. 112. C.I.A., ' L'Égypte,' p. 410; p. 545. La Syrie, p. 38. Z.D.M.G.,
1935, p. 214. Glossary to Nujum, vol. v, p. 1.
8 Ibn p. 259, 1. 1.
9 Zetterstéen, p. 163,1. 5. Suluk, ii, p. 338, ll. 20–1. Nujum (P), vii, p. 362,1. 8. Ibn Iyas, ii, p. 42, 1.
3; iv, p. 230, ll 5–6; p. 398, ll. 7–8.
10 Ibn Iyas, ii, p. 113, 1. 9; p. 137, 1. 21; p. 165, 1. 16; p. 195, 1. 5; v, p. 53, 1. 2; p. 76, 1. 4; p. 83, 1.
2. iv, p. 55, ll. 7–8.
1 Zetterstéen, p. 157,1. 21; p. 162, 1. 13. Suluk, i, p. 681, ll. 12–15. p. 62, ll. 3–5. Tibr, p. 357, 1. 15.
Ibn Iyas, ii, p. 163, 1. 16; v, p. 37, 1. 23. ii, p. 275, 1. 7; p. 276, ll. 5–6; (KM) iii, p. 382, 1. 25; vi, p.
212, 1. 10.
2 iv, p. 10, 1. 7. p. 244, 1. 21; p. 261, ll. 5–6. ii, p. 110, 1. 10, 1. 25. p. 202, 1.2.
3Suluk, ii, p. 326, ll. 8–9; p. 470, 1. 14. Nujum (P), v, p. 47; p. 155, l. 23; p. 278, ll. 1–2; p. 428, 1. 21;
p. 429, 1. 20; p. 430, 1. 10; pp. 444, 1. 2–445, 1. 19; p. 520, ll. 12–15; p. 634, l. 13; iv, p. 7, 1. 4; p.
228, 1. 20; p. 234, ll. 20–1; p. 249, 1. 16; p. 250, 1. 8; p. 278, 1. 8; p. 825, 1. 21; vii, p. 166, 1. 14; p.
391, ll. 12–13; p. 410, 1. 12; p. 722, 1. 16.
4Suluk, ii, p. 405, 1. 12. iv, p. 61, 1. 6, 1. 9.
5Nujum (P), vii, p. 264, 1. 15; p. 426, 1. 5.
6Nujum (P), v, p. 28; vi, p. 825, 1. 4, 1. 13. Ibn Iyas (KM), iii, p. 89, 1. 6; iv, p. 63, 1. 3.
7Suluk, ii, p. 237. Nujum (P), v, p. 5, 1. 22; p. 33, 1. 21. p. 452, ll. 16–22. Tibr, p. 122, 1. 11. Ibn Iyas,
ii, p. 60, ll. 21–2; p. 114, 1. 27. ii, p. 312, 1. 14; iii, p. 59, 1. 23; p. 196, 1. 7; p. 282, 1. 4; x, p. 279, 1.
12, ll. 28–9.
8 vi, p. 218, 1. 14.
9 ii, p. 267, 1. 17; p. 273, 1. 6; p. 311, 1. 11; p. 318, ll. 6–7; iii, p. 36; p. 41, 1. 25; p. 60, 1. 13; p. 66,
ll. 27–8; p. 277, 1. 27; vi, p. 164, 1. 7; p. 194, 1. 24; p. 201; p. 214, 1. 19; p. 215, 1. 5; p. 224, ll. 3–4;
x, p. 271, 1. 6; p. 275, 1. 3, ll. 6–8; p. 280, 1. 15; p. 288, 1. 7; p. 290, l. 28; p. 345,1. 21; xi, p. 150, ll.
7–8.
10Suluk, ii, p. 320, ll. 2–3. Nujum (C), vii, p. 280, 1. 7. ii, p. 312, 1. 13; p. 315, 1. 2; p. 316, 1. 28; iii,
p. 8, 1. 3; p. 10; p. 28; p. 29, p. 67, 1. 6; p. 161, 1. 8; p. 295, 1. 15; iv, p. 10, l. 16; vi, p. 195, 1. 24; p.
233, 1. 4.
11 This description is based on the material cited in note 6, p. 467, above; cf. also p. 244, ll. 23–4.
1Suluk, ii, p. 221, 1. 15; cf. also p. 280, ll. 6–7. Nujum (C), ix, p. 65, ll. 9–10.
2 p. 244, ll. 23–4.
3Nujum(P), v, p. 457, ll. 15–18.
4Nujum (P), vii, p. 237, ll. 1–10.
5 p. 281, ll. 11–12.
6 p. 344, 1. 3.
7 p. 452, ll. 21–2.
8p. 631, ll. 16–17.
9 Ibn Iyas, iv, p. 30, 1. 14; p. 277, 1. 8.
10 Ibn Iyas, iv, p. 358, ll. 6–7.
11 Ibn Iyas, iv, p. 434,1. 7.
12 Ibn Iyas, v, p. 2, ll. 13–14.
13Manhal, iii, fol. 155a, 1. 6. Ibn al-Furat, ix, p. 276, 1. 1. Ibn Iyas, ii, p. 3, 1. 3; p. 87, ll. 4–6; p. 221,
1. 27; (KM) iii, p. 396, 1. 18; p. 440, 1. 17. Durar, i, p. 418, 1. 5; p. 424, 1. 12; iii, p. 255,1. 8. ii, p.
275, 1. 12.
14Durar, i, p. 418, 1. 5. Ibn p. 260, 1. 17. iii, p. 26; p. 281, 1. 9; vi, p. 196, 1. 2.
15 Ibn Iyas, ii, p. 16,1. 5.
16Nujum (P), vi, p. 515, ll. 5–6. x, p. 289, 1. 16. For and amir cf. Quatremère, vol. i, part i, p. 129;
p. 173. C.I.A., ' L'Égypte,' p. 543. La Syrie, p. xxxiv; pp. xxxvii–viii; p. liv. Feudalism, p. 3; pp. 7–9;
p. 13; p. 14; p. 21; p. 31; p. 54. G.Wiet, Syria, 1926, p. 160; p. 171; p. 175. Glossary to Nujum, vol.
vi, p. liv. Glossary to p. xxx.
17 Ibn al-Furat, ix, p. 128, ll. 1–3; p. 400, ll. 12–14. Ibn Shuhba, fol. 85a, 1. 15. Durar, i, p. 201, ll.
16–17. iv, p. 9, 1. 29; vi, p. 211, 1. 26.
18 See references listed in n., p., above.
19 Besides basic data cited in n., p., above, cf. also amir khamsin (Abu al-Fida', iv, p. 103, 1. 1; p.
106, 1. 24. p. 378, ll. 19–20) and amir thamanin (Nujum (P), vi, p. 9, ll. 10–11, ll. 19–20; vii, p. 570,
notes); those belonged, by definition, to the Amirs of
1Zubda, p. 113.
2 iv, pp. 8, 1. 17–9, 1. 4. On see Suluk, ii, p. 521, n. 2. Nujum (P), v, p. 209, 1. 2; p. 221, 1. 23. p.
244. Ibn Iyas (KM), iii, p. 338, ll. 8–10. See also Suluk, ii, p. 326, 1. 11. Nujum (P), vi, p. 9, ll. 10–11.
Zubda, p. 113, 1. 8. On the kusat, see Patrologia Orientalis, xx, p. 153, 1. 4. Ibn Kathir, xiii, p. 347,
ll. 7–9. Suluk, i, p. 136, 1. 2. p. 402, 1. 5.
3 Zetterstéen, p. 177, 1. 13; p. 208, 1. 24. Suluk, ii, p. 376, 1. 18. Nujum (P), vii, p. 597, ll. 16–17. Ibn
Iyas, ii, p. 40, 1. 18; p. 44, 1. 6; p. 56, 1. 1; p. 58, 1. 26; p. 100, 1. 26; p. 108, 1. 6; p. 191, 1. 21. Ibn
al-Furat, ix, p. 44, 1. 16. iii, p. 296, 1. 6; vi, p. 198, 1. 27; x, p. 205, l. 13; p. 289, 1. 17. For
amir of Ten, cf. C.I.A., 'L'Égypte,' p. 543. Z.D.M.G., 1935, p. 214. Glossary to Nujum, vol. vi, xliii.
4 Ibn Shuhba, fol. 71b, 1. 14; fol. 75b, 1. 15.
5Nujum (P), vii, p. 617, ll. 12–13. vi, p. 221, 1. 10.
6Nujum (P), v, p. 572, 1. 7; vi, p. 16, 1. 17; p. 25, 1. 15; vii, p. 426, 1. 14. p. 358, ll. 17–19. Manhal,
i, fol. 165a, 1. 18. Ibn al-Furat, ix, p. 61, 1. 21; p. 67, 1. 21; p. 135, 1. 7; p. 164, 1. 14. Tibr, p. 147, 1.
8. viii, p. 221, 1. 1. vi, p. 224, 1. 17.
7 iv, pp. 16 ff. p. 245, ll. 9–10. cf. C.I.A., 'L'Égypte,' p. 543.
8 See, for instance, Nujum (P), vi, p. 182; vii, p. 453, 1. 8; p. 597, 1. 15. p. 299, ll. 5–10; p. 554, 1.
10. Manhal, i, fol. 166, 1. 2; fol. 209, 1. 21; ii, fol. 113b; iii, fol. 109b, 1. 20.
1Ta'rikh Bayrut, p. 142, 1. 14; p. 184, 1. 3.
2Ta'rikh Bayrut, p. 149, 1. 15.
3Zubda, p. 113, ll. 4–18.
4 Ibn Iyas (KM), iii, p. 218, ll. 1–13.
5 Ibn Iyas, iv, pp. 30–4.
6 Ibn Iyas, iv, p. 105, ll. 5–7.
7 Ibn Iyas, v, p. 5, ll. 8–9. Following are references on the places of residence of the amirs and of the
army generally: Abu al-Fida', iv, p. 67, ll. 24–5. Suluk, i, pp. 341, 1. 18–342, 1. 1; p. 668, n. 1. Nujum
(C), vii, p. 72, n. 2; p. 191, ll. 3–4; ix, p. 121, ll. 11–12. Nujum (P), vi, p. 8, 1.12; pp. 523–4; vii, p.
416, 1. 18. Manhal, i, fol. 193a, 1. 3; iii, fol. 132b, ll. 1–4; iv, fol. 171a, ll. 3–4; viii, fol. 432b, ll. 6–7.
Ibn al-Furat, ix, p. 343, 11. 5–8. Ibn Shuhba, fol. 91b, ll. 15–17. Ibn Iyas (KM), iii, p. 404, ll. 1–
2. i, p. 125, ll. 19–21; p. 342, ll. 24–9; ii, p. 23, ll. 20–8; pp. 68 ff.; p. 73, 1. 33; p. 116, ll. 34–8; p.
131, ll. 3–6; pp. 133, 1. 28–134, l. 24; p. 135, ll. 8–13; p. 205, ll. 5–8. Zubda, pp. 28, 1. 22–29,1. 1.
8 Baybars fol. 99a, ll. 7–10; 99b, 1. 13—l00a, 1. 4; 114a, ll. 6–7; 122a, 1. 13; 183a, ll. 7–8; 245a, ll.
8–9, ll. 15–16; 259b, ll. 7–8.
1 See, for instance, Nujum (C), vi, p. 362, ll. 13–14; vii, p. 88, ll. 7–8; p. 89, 1. 11; p. 99, ll. 12–13; p.
320, ll. 5–6; p. 350, 1. 6; viii, p. 13, ll. 12–14; ix, p. 11, ll. 17–18; p. 228, 1. 15; p. 282, 1. 1; p. 287, 1.
2. Ibn al-Furat, viii, p. 37, 1. 7. Suluk, i, p. 239, 1. 1; p. 415; ll. 7–8; ii, p. 185, 1. 4. Zetterstéen, p.
128, 1. 22; p. 173, ll. 3–4. Ibn Kathir, xiv, p. 312, ll. 3–5. Suluk, i, p. 580, ll. 11–12; p. 587, 1. 1; p.
681, 1. 6; p. 687, 1. 18; p. 702, 1. 8; p. 770, 1. 5; p. 794, ll. 8–9; ii, p. 47, 1. 8; p. 97, 1. 12. We could
not ascertain which of these expressions is the earliest.
2 Ibn Kathir, xiii, p. 309,1. 18. Zetterstéen, p. 224, ll. 19–20. Abu al-Fida', iv, p. 29, ll. 24–5. Suluk, i,
p. 735, ll. 5–6. In Suluk (ed. Ziada) we encounter, as early as the days of ad-Din, a sentence such as
the following: 'umara' mi'a 'ashara', which, if taken at face value, might lead us to conclude that the
office of amir mi'a was already in existence at that early period. Such a conclusion would, however,
be erroneous, for the text is here corrupt, and ought to read: 'umara'—mi'a 'ashara'. This has already
been pointed out in this chapter, in connexion with the As early as the year 680, the term
amir mi'a muqaddam alf (Ibn al-Furat, vii, p. 238,1. 14) is encountered. We find, in the early Mamluk
period, the following expressions which do not seem to recur during the later period: amir
(Zetterstéen, p. 46, 1. 17; p. 101, 1. 14; p. 166, 1. 25); rakiba li-imrat (Zetterstéen, p. 201, 1. 5, 1. 7;
p. 202, 1. 21; p. 206, 1. 1, 1. 10; p. 214, 1. 24). In the year 661 we find the following: ' al-'Aziza ibn
al-Malik al-Mughith imrat mi'at faris wa-khala'a'alayhi ' (Suluk, i, p. 493, ll. 2–3). Must we conclude
from the above quotation that at that period the identity of Amir of Forty with Amir of was not yet
firmly fixed? Ibn Taghribirdi tends, indeed, to assume that the title of Amir of was granted at the
beginning of the Mamluk period to an Amir of a Thousand as well, since the played at his gate as
well as at the gate of Amirs of Forty (Mankal, iii, fols. 181b, 1. 23–182a, 1. 8). See, in this
connexion, the interesting passage in Ibn 'Abd fol. 57a, ll. 3–8. cf. also Manhal, ii, fol. 17a,
ll. 1–6; v, fol. 12a, ll. 1–4. Nujum (C), ix, p. 287, ll. 4–10, and references in n. 8, p. 471.
1 See references listed in n. 6, p. 467, above. When later sources copy from earlier ones they use, of
course, the word faris (horseman).
2 See n. 4, p. 451, above, and Nujum, vi, pp. 386–7. Zubda, pp. 115–16. ii, pp. 111–13.
3Nujum (P), vi, p. 386, ll. 8–16.
1 The rise, decline, and disappearance of this regiment which played such a decisive role in wiping
out the Ayyubid kingdom and in establishing the Mamluk kingdom in its stead is discussed by the
present writer in 'Le Régiment dans l'Armée Mamelouk' (R.E.I., 1952, pp. 133–141). We shall here
only allude to a most interesting passage in Ibn Khaldun's Kitab al-'Ibar (vol. v, pp. 371, 1. 27–372,
1. 8) which stresses that though mamluks were bought by the Ayyubids in considerable quantities
from the days of ad-Din onwards, it was Najm ad-Din Ayyub who by far surpassed all his
predecessors in this respect. This passage also mentions a very important factor which greatly
facilitated the buying of mamluks on a grand scale by Najm ad-Din Ayyub, viz. the attack of the
Tatars on the steppes lying to their north-west (al-janib al-gharbi min ash-shimal) which uprooted
the Kipchakis, Russians, Alans, and others, many of whom were sold as slaves. This statement of Ibn
Khaldun is confirmed by Ibn Duqmaq, who says in his description of the rise of the Mamluk
kingdom (ibtida' ad-dawla ash-sharifa at-turkiya) that God has expelled them (the mamluks) from
their vast and spacious countries of origin, and led them to Egypt, by a wisdom which is beyond the
comprehension of man. God has decreed the appearance of the Tatars and their conquest of the
eastern and northern countries (al-bilad al-mashriqiya wash-shimaliya), and their attacks on the
Kipchakis. The Tatars killed the Kipchakis and captured and sold their offspring. These were carried
by the merchants to far-off places (ila al-afaq), and when Najm ad-Din Ayyub became king he
bought about a thousand mamluks (al-jawhar ath-thamin fi ta'rikh al-khulafa' Oxford MS.,
Pocock, 352, fol. 35b, ll. 14–21). Such ideal conditions for buying mamluks existed only at the end of
the Ayyubid reign, and this was one of Najm ad-Din's greatest advantages over his predecessors. It is
worth while to note here, in passing, that the very inception of the Mamluk kingdom is thus closely
connected with the Mongols and their invasions.
1Durar, i, p. 482, ll. 8–9.
2 In Qalaun's time cases of accelerated promotion were still very infrequent, for this sultan was
usually very slow and careful in promoting his amirs (Baybars fols. 99b, 1. 13—l00a, 1. 4).
3Mankal, i, fol. 197b, ll. 13–16. Nujum (P), v, p. 235, ll. 17–20,
4Nujum (P), v, pp. 305–6.
5 See, for instance, Nujum (P), v, p. 295, ll. 9–10; p. 306, ll. 11–12; p. 333, ll. 17–20; p. 345, ll. 5–6;
p. 355, ll. 7–14; vi, p. 432, ll. 21–2; p. 785, ll. 21–2; vii, p. 825, ll. 14–16. p. 485, ll. 14–16. Manhal,
iv, fol. 172b, 1. 15. Ibn Iyas (KM), iii, p. 174, ll. 5–6. These constitute but a few examples. cf. also
the description of the office of atabak al-'asakir in Part III of this article.
1 See references listed in notes 2 and 6, below.
2 See the author's 'The Circassians in the Mamluk Kingdom', J.A.O.S., 1949, pp. 135–147. On the
personality of al-Mu'ayyad Shaykh, see ibid., p. 142.
3Nujum (P), vi, p. 428, ll. 3–6. Ibn Iyas, ii, p. 8, ll. 25–6.
4 Manhal, v, fol. 18a, ll. 12–17.
5Nujum (P), vi, p. 43, ll. 1–2.
6Nujum (P), vi, p. 428, ll. 10–12.
7Nujum (P), vi, p. 430, ll. 14–18.
8 See, for instance, Nujum (P), vi, p. 430, ll. 14–15; vii, p. 377, 1. 15; p. 592, ll. 10–11; p. 687, 1. 14;
p. 688, 1. 1. p. 371, ll. 3–7; p. 379, 1. 4; p. 658, ll. 4–7; p. 660, ll. 8–9; p. 667, ll. 17–18; p. 716, ll.
10–14; p. 717, ll. 18–21; p. 718, ll. 4–7. Manhal, i, fol. 167a, ll. 16–20; fol. 205a, ll. 17–19; ii, fols.
113b, 1. 22–114a, 1. 2; iii, fol. 134, ll. 12–19; viii, fol. 416b, ll. 3–8; fol. 429a, ll. 19–21. Tibr, p. 129,
1. 7; p. 189, ll. 15–16. Ibn Iyas, ii, p. 70, ll. 5–6; p. 84, ll. 8–9. ii, p. 324, ll. 11–12; iii, p. 33; p. 56, 1.
1; p. 60, ll. 18–19; ll. 25–6; p. 76, ll. 26–8; p. 175, ll. 21–2; p. 210, ll. 5–7; p. 276, ll. 23–4; vi, p. 165,
ll. 25–7; pp. 200, 1. 28–201, 1. 1; x, p. 268, ll. 25–7. The fact that al-Mu'ayyad Shaykh favoured the
jins at-turk, the race despoiled of its ascendancy by the Circassians, also indicates that he was
exceptional among Circassian sultans. He also favoured the sayfiya (see above). Both the turk and the
sayfiya were underprivileged elements, but for entirely different reasons; the first because of racial
considerations, and the second because of their low standing in the Mamluk hierarchy, as already
pointed out. In both, al-Mu'ayyad saw pliable material which could be fitted into any mould that
might suit his purposes. (Nujum (P), vi, p. 430, ll. 14–18. pp. 378, 1. 19–379, 1. 13. Manhal, iii, fol.
168a, ll. 4–5.)

STUDIES ON THE STRUCTURE OF THE MAMLUK ARMY—III

1 Basic material in Mamluk sources dealing with the various offices is found in: iv, pp. 16–22;
v, pp. 461–2; vii, pp. 158–9. pp. 245–9; p. 343. Zubda, pp. 114–15. ii, pp. 111–13.
2 Ibn Iyas, ii, p. 127, ll. 5–9. Zubda, p. 112, ll. 15–21. On his functions, see ii, p. 215. iv, pp.
16–22; xi, p. 134, ll. 17–21.
3 ii, p. 111, ll. 8–12. iv, pp. 16–22.
VOL. XVI. PART 1. 5
1Nujum (P), v, p. 454, 1. 5. ii, p. 111, 1. 12. p. 273. For na'ib cf.: C.I.A., L'Égypte, p. 60; p. 210; p.
211; p. 212; p. 215. La Syrie, pp. Iv–vi. For an-na'ib al-kafil, kafil al-mamalik, cf. C.I.A., L'Égypte, p.
208; p. 215; p. 216; p. 227. G.Wiet, Syria, 1926, p. 155. Glossary to Nujum, vol. vi, p. lvi.
2 Ibn Khaldun, v, p. 443, ll. 2–4. ii, p. 111, ll. 8–9. The name of the last holder of this office is not
mentioned in the sources.
3Manhal, viii, fol. 366a, ll. 8–14.
4 Ibn Iyas, ii, p. 127, ll. 5–9.
5Zubda, p. 116, ll. 16–21. The na'ib al-ghayba was the Mamluk amir who took the sultan's place
while the latter was abroad, especially on military expeditions. For additional material on the na'ib
see Suluk, i, p. 664, ll. 6–8; p. 665; p. 715, n. 3. Ibn Kathir, xiv, p. 76, ll. 20–1; p. 140, ll. 15–16.
Nujum (C), viii, p. 233, ll. 11–12. Nujum (P), v, p. 92, 1. 13; p. 175, ll. 8–9; p. 200, ll. 3–4; p. 217; p.
223, 1. 6; vi, p. 294.
6 See, for instance, Ibn al-Furat, vii, p. 148; n. 1. Suluk, i, p. 656, 1. 8; p. 657, 1. 7. For atabak cf.
Quatremère, vol. i, pt. i, p. 2. C.I.A., L'Égypte, p. 290; p. 396. La Syrie, p. xxvii; p. Ivi. Heraldry, p.
56; p. 85; p. 88, etc. Feudalism, p. 1; p. 14. G.Wiet, 'Notes d'Épigraphie Syro-Musulmane', Syria,
Paris, 1926, p. 155; p. 164.
7Nujum (C), viii, p. 286, ll. 6–7. Ibn al-Furat, vii, p. 148, n. 1. Nujum (P), p. 165, ll. 4–5. Ibn Iyas
(KM), iii, p. 454, 1. 19, and many other passages. For mudabbir al-mamlaka (al-mamalik), cf. C.I.A.,
L'Égypte, p. 420; p. 455. Heraldry, p. 65; p. 151; p. 177; p. 252f.
8Zetterstéen, p. 29, 1. 10. Ibn Shuhba, fol. 52b, ll. 6–7. p. 364, 1. 2.
9 Ta'rif, pp. 103–4. p. 318.
10 vi, p. 5. Ibn Iyas, ii, p. 10, 1. 27; (KM), iii, p. 17, 1. 1; p. 126, 1. 18; p. 363, 1. 8. Manhal, i, fol.
111a, 1. 7.
1 vi, p. 6, 1. 1; cf. also Suluk, i, p. 794, ll. 7–8.
2Zubda, p. 112, 1. 22.
3 iv, p. 149, ll. 7–8.
4 vii, p. 262, ll. 13–19.
5 Ibn Khaldun, v, p. 485, 1. 6.
6Nujum (P), v, p. 165, ll. 10–12. Manhal, iii, fol. 154a, ll. 10–11. Tibr, p. 7,1. 14. For amir kabir, cf.
C.I.A., L'Égypte, p. 276; p. 290; p. 452; p. 593. Glossary to Nujum, vol. v, p. xii; vol. vi, p. liv.
7Nujum (P), vi, p. 804, 1. 7; vii, p. 256, 1. 9.
8 Ibn Khaldun, v, p. 458,1. 24. Nujum (P), v, p. 175, ll. 5–6; p. 209, ll. 8–11; p. 210, ll. 7–9.
9 Nujum (P), v, p. 319, ll. 4–5; p. 447, ll. 17–20; vi, p. 315, ll. 6–7. For additional material on the
atabak, see Abu al-Fida', iv, p. 75, 1. 13. Suluk, i, p. 146, 1. 3; ii, p. 663, 1. 10. Nujum (P), v, p. 1, 1.
14; p. 54, ll. 7–8; p. 124, 1. 15; p. 610, 1. 10; vi, p. 144, ll. 4–9; vii, p. 7, ll. 1–3; p. 420, 1. 10.
Manhal, iv, fol. 208b, ll. 19–23. Ibn Iyas (KM), iii, p. 310, ll. 4–5; p. 333, ll. 9–10; iv, p. 8, ll. 9–10;
p. 485, 1. 7.
10Nujum (P), v, p. 355.
11 iv, p. 18. For amir majlis, cf. Quatremère, vol. ii, pt. i, p. 97. C.I.A., L'Égypte, p. 274; p. 585.
La Syrie, p. lvii. Heraldry, p. 69; p. 101, etc.
1 iv, p. 18. ii, p. 111. The amirs of the were called az-zardkashiya, and their chief was called az-
zardkash al-kabir ( iv, p. 18). cf. p. 110, n. 6a.
2 This has already been dealt with in detail by A.N.Poliak in Feudalism, p. 14, p. 15, p. 65, and in
R.E.I., 1935, pp. 2/35–2/36. On the Yasa, see p. 68 and n. 6 below.
3 iv, p. 19. ii, p. 111.
4 p. 504, ll. 7–9. It seems that judicial authority was at first vested in the exclusively, and
that it was only later conferred upon the thani as well: Nujum (P), v, p. 5, ll. 18–21.
5Nujum (P), v, p. 369, ll. 13–15; vii, p. 442, ll. 17–21. iii, p. 288, 1. 6.
6Nujum (P), v, p. 308, 1. 7.
7Nujum (P), v, p. 189, 1. 21. For cf. C.I.A., L'Égypte, p. 567. La Syrie, pp. Iviii–ix. Heraldry, p. 5; p.
18; p. 58; p. 97, p. 116; p. 135, etc. Feudalism, p. 14; p. 15; p. 65.
8 iv, p. 18; v, p. 455, ll. 10–15. Ibn al-Furat, ix, pp. 162–3.
9Nujum (P), vi, p. 5, 1.16. Ibn Iyas (KM), iii, p. 2, 1. 15. Toward the close of the Mamluk period, we
find a ra's nawbat an office whose nature is not clear (Ibn Iyas, iv, p. 450, ll. 1–2; p. 481, 1. 5).
1Nujum (P), v, p. 295, ll. 7–8.
2 Ibn Iyas (KM), iii, p. 381, ll. 22–3.
3 iv, p. 18. For ra's nawba, cf. Quatremère, vol. ii, pt. i, p. 13. C.I.A., L'Égypte, p. 241; p. 537. La
Syrie, p. Ivi. Z.D.M.G., 1935, p. 203. Feudalism, p. 38. Heraldry, p. 69; p. 84, n. 1; p. 91; p. 172, etc.
Glossary to Nujum, vol. v, p. xxxvi; vol. vi, p. Ixiv.
4 The offices of wazir, ustadar, muqaddam al-mamalik katib al-mamalik, and al-'askar are
described here only in general outline since they were originally discussed in other chapters of the
writer's work on the Mamluk army.
5Suluk, i, p. 671. ii, p. 168, ll. 1–2. Zetterstéen, p. 97, ll. 1–19.
6 ii, p. 169, ll. 9–13.
7Durar, ii, p. 252, ll. 11–12.
8 For the establishment and functions of the diwan al-mufrad see Poliak, Feudalism, p. 4 and index;
cf. also the office of the ustadar below.
9 cf. the office of ustadar below.
10 ii, pp. 222–3.
11Zubda, p. 97, ll. 23–4. Nujum (P), vii, pp. 801–2. pp. 225, 1. 21–226, 1. 7. Ibn Iyas, iv, p. 200, ll.
18–20.
12 cf. references in note 8.
1Zubda, p. 107, ll. 15–16. iii, p. 457, ll. 2–7.
2 Ibn al-Furat, ix, p. 300, 1. 7. Nujum (P), vi, p. 577, ll. 1–2. p. 232, ll. 3–5.
3 ii, p. 222.
4Nujum (P), vi, pp. 533–4; vii, p. 216, ll. 4–18; p. 520, ll. 14–16; p. 413, ll. 15–16. In
connexion with payments to the army the office should be mentioned. His main duty was the
distribution of clothes (kiswa) to the army pp. 489–490, 491–2).
5 iv, p. 21, ll. 4–6. cf. p. 110, n. 6a.
6 iv, pp. 16–22. ii, p. 113, ll. 19–22. One sometimes encounters the spelling dawat dar (Abu al-Fida',
iv, p. 140, 1. 17) and a diminutive form duwaydar. In the latter, no diminutive or contemptuous
connotation is intended; even the chief dawadar is called ad-duwaydar al-kabir, without any
belittling implication (Zetterstéen, p. 187,ll. 2–3. Ibn al-Furat, ix, p. 363, 1. 21. Durar, ii, p. 230, 1. 7;
iii, p. 109, 1. 30. iii, p. 10).
7Nujum (P), iv, p. 571, ll. 17–19.
8 For example: Barsbay (Ibn Iyas, ii, p. 15, ll.8–27) and the last Mamluk sultan.
9Nujum (P), vi, pp. 739–740.
1 Ibn Iyas, ii, p. 112, ll. 26–7; (KM), iii, pp. 138–9; p. 191, ll. 12–13; p. 196, ll. 15–18; p. 274, ll. 19–
20; p. 400, ll. 3–6; iv, p. 26, ll. 4–5; p. 160, ll. 5–7; p. 191, ll. 12–15; p. 210, ll. 16–20; pp. 261, 1. 21–
262, 1. 1; p. 264, ll. 16–17; p. 280, ll. 1–8; p. 298, ll. 7–8; p. 388, ll. 17–19. Feudalism, pp. 45–6.
2 Ibn Iyas (KM), iii, p. 145, ll. 3–6.
3 Ibn Iyas (KM), iii, p. 436, ll. 8–11. In the last years of the Mamluk period, one finds an office
called dawadar sakin, with an apparently fairly large number of holders (Ibn Iyas [KM], iii, p. 429,
ll. 8–9; iv, p. 62, 1. 10; p. 133, 1. 3; p. 274, 1. 11; p. 304, ll. 5–6; p. 301, 1. 19; p. 395, 1. 19. p. 485, 1.
21). The nature of this office is not clear. For interesting material on the amir akhur, see his letter of
appointment in xi, pp. 170–2. Ta'rif, pp. 99–101. On the sultan's stables and the rikabkhana, see
Zubda, p. 124, ll. 9–13; p. 125; p. 126, ll. 1–8. cf. p. 110, n. 6a.
4 iv, pp. 18–19. ii, p. 113, ll. 16–17.
5 Ibn Iyas, ii, p. 108, 1. 9; (KM) iii, p. 28, 1. 22.
6 vi, p. 236,1. 6. For amir akhur, cf. Quatremère, vol. i, pt. i, p. 160. C.I.A., L'Égypte, p. 301. La
Syrie, p. Ivii. Heraldry, p. 5; p. 25; p. 130; p. 172.
7 Ibn al-Furat, ix, p. 147, 1. 9; p. 184, 1. 22.
8Suluk, i, p. 133, 1. 4, 1. 11. Nujum (C), vi, p. 132, 1. 18. Ibn Kathir, vi, p. 255, 1. 9.
9 ii, pp. 111, 1. 25–112, 1. 1. iv, p. 20, ll. 5–12.
1Suluk, i, p. 359; cf. also p. 222, 1. 12; p. 134, 1. 16. Nujum (C), vii, p. 37, ll. 6–7. For developments
and changes in the offices during the Mamluk period, see below.
2 p. 473, ll. 3–10. For additional material on this office, see Suluk, ii, p. 377, 1. 4. Nujum (P), vii, p.
237, 1. 18. Ibn al-Furat, viii, p. 156, 1. 23; p. 215, 1. 16; ix, p. 43, 1. 10; p. 48, 1. 4; p. 170, 1. 4; p.
406, 1. 19. Durar, i, p. 387, 1. 18; p. 50, ll. 19–20. iii, p. 522. iii, p. 273, 1. 25. For jandar and amir
jandar, cf. Quatremère, vol. i, pt. i, p. 14. C.I.A., L'Égypte, p. 77; p. 78; p. 291; p. 390. La Syrie, p.
lix; p. c. Heraldry, p. 58, p. 183.
3 Zetterstéen, p. 19, 1. ll. Abu al-Fida', iv, p. 111, 1. 16. Suluk, i, p. 765, 1. 9. Ibn al-Furat, viii, p. 132,
ll. 20–1. xii, p. 453, 1. 14.
4 iv, p. 22; v, p. 456. Additional details in xii, pp. 431, 1. 19–432, 1. 16; this is, however, written in
flowery style, and the extent of its accuracy is difficult to gauge.
5 p. 29, ll. 17–20; p. 519, ll. 8–23. Ibn al-Furat, ix, p. 7, 1. 22.
6 p. 516, ll. 10–11; p. 519, ll. 8–23. Ibn Iyas, iv, p. 124, 1. 19; p. 205, ll. 18–25.
7 Ibn Iyas (KM), iii, p. 329,ll. 1–5; iv, p. 13, ll. 13–16.
8 Ibn Iyas, v, p. 15, ll. 17–18.
9 Ibn al-Furat, viii, p. 154.
10 Ibn al-Furat, ix, p. 5, ll. 18–21; p. 155, ll. 16–17; pp. 365, 1. 26–366, 1. 3. For additional material
on the naqib al-jaysh, see Zetterstéen, p. 1, 1. 2; p. 24, ll. 8–9; p. 43, ll. 23–4; p. 57, 1. 16; pp. 168, 1.
23–169, 1. 1. Suluk, i, p. 800, 1. 4; p. 846, n. 2; p. 850, 1. 3; ii, p. 194, 1. 14; p. 199, 1. 11; p. 455, 1.
12; p. 480, 1. 15. Nujum (P), vii, p. 195, 1. 3, 1. 7; p. 443, 1. 5; p. 448, l. 3; p. 662, ll. 4–7. p. 29, ll.
17–20; p. 166, ll. 1–6; p. 516, ll. 10–11; p. 519, 1. 11. Ibn al-Furat, viii, p. 54; ix, p. 17, 1. 22; p. 80,
ll. 10–11; p. 155, ll. 9–11; p. 159, ll. 9–10; p. 336, ll. 2–3; pp. 365, 1. 26–366, 1. 3. Tibr, p. 183, 1. 3.
Ibn Iyas, ii, p. 69, ll. 2–3; p. 23, ll. 6–8; p. 150; p. 166, 1. 5; (KM), iii, p. 50, ll. 4–6; iv, p. 13, ll. 13–
16; p. 124, 1. 19; p. 205, ll. 18–21; p. 249, 1. 23; p. 256, 1. 6; p. 289, ll. 14–16; p. 446, ll. 14–15; p.
453, ll. 19–21; v, p. 15, ll. 17–18. Durar, i, p. 81, 1. 4; ii, p. 176, 1. 16; p. 229, 1. 9. i, p. 90, ll. 18–19.
1 p. 456. cf. also Durar, i, p. 425, 1. 16; ii, p. 197, 1. 20.
2Manhal, iii, fol. 169b, ll. 14–15. For naqib al-juyush, cf. La Syrie, p. lxii. Heraldry, p. 50; p. 83; p.
213.
3Durar, i, pp. 350, 1. 20–351, 1. 1.
4Suluk, ii, p. 377, ll. 1–2.
5 Zetterstéen, p. 150, ll. 7–8; p. 178, ll. 7–8; p. 188, ll. 11–12; p. 195, ll. 2–5.
6 Zetterstéen, p. 178, ll. 7–8; p. 214, ll. 19–20. Nujum (C), viii, p. 161, 1. 20; p. 204, ll. 8–9. Suluk, i,
p. 946, 1. 11; ii, p. 165, ll. 1–2; p. 246, 1. 19; p. 353, 1. 17. Durar, i, p. 430, ll. 4–5; p. 498, ll. 2–3; iii,
p. 259, ll. 6–7. i, p. 250, 1. 15.
7 For a detailed description of this office and its bearer, cf. the author's L'Esclavage du mamelouk,
pp. 14–15.
8Nujum (P), vii, p. 218, ll. 10–20. p. 113, ll. 7–23. Ibn Iyas, iv, p. 291, ll. 3–6 ; p. 307, ll. 8–
11; p. 353, ll. 20–3; p. 413, ll. 6–7; p. 416, ll. 16–19; v, p. 19, ll. 16–18; p. 78, ll. 5–8.
1 ii, p. 215, ll. 24–6. Ibn al-Furat, ix, p. 32, ll. 5–6. cf. also Feudalism, p. 26, and n. 1.
2 Ibn Iyas, v, p. 58, 1. 15; p. 63, ll. 1–4; pp. 199–200; p. 201, 1. 22; p. 205, 1. 10; p. 207, l. 20; p. 208,
1. 3, 1. 16; p. 217, 1. 4. cf. C.I.A., 'L'Égypte', p. 450. G.Wiet, Syria, 1926, p. 155.
3 This description of the diwan al-jaysh is based on Poliak's Feudalism, pp. 20–1.
4 iv, pp. 30–1; p. 33. On the mutawalli diwan al-jaysh, see Ibn Iyas, ii, p. 174, 1. 23; (KM), iii, p.
129, 1. 17. On the katib al-jaysh and the kuttab al-jaysh, see Suluk, ii, p. 433, 1. 13; p. 496, 1.1. Ibn
al-Furat, vii, p. 158, ll. 11–18. On the mustawfi al-jaysh and the office of istifa' al-jaysh, see
p. 332, 1. 18. Ibn Iyas (KM), iii, p. 354, 1. 22; iv, p. 149, 1. 8; p. 181, ll. 7–8. On 'amil al-jaysh see
Abu al-Fida', iv, p. 108,1. 17. For and diwan al-jaysh, cf. La Syrie, p. xxxiii; p. Ixxii; p. Ixxvi.
Heraldry, p. 46; p. 121 f. Feudalism, p. 21; p. 30. C.I.A., L'Égypte, p. 345. For mustawfi, cf.
Quatremère, vol. i, pt. i, p. 202.
5iv, p. 32.
6 iv, p. 32. cf. also Suluk, ii, p. 438. Nujum (P), v, p. 109, ll. 9–10; p. 313. p. 189, ll. 18–19.
1 For the muta'mmimun, cf. Quatremfère, vol. i, pt. i, p. 245. C.I.A., L'Égypte, pp. 446–8.
2 xi, p. 96, ll. 5–12; pp. 204–7; xii, pp. 206, 1. 17–207, 1. 12; pp. 359–361. Ta'rif, pp. 123–4.
3 xi, p. 204, ll. 16–17.
4 See, for instance, Suluk, ii, p. 391, 1. 16. Nujum (P), v, p. 533, 1. 3; p. 534, 1. 12; p. 537, 1. 14; p.
544, 1. 6; p. 545, ll. 21–2; p. 549, 1. 6; vi, p. 29, ll. 4–5; p. 30, ll. 8–9 ; p. 168, 1. 7; p. 191, 1. 2; p.
195, ll. 9–10; p. 232, 1. 22; p. 341, 1. 4; 1. 18; p. 378, 1. 19; p. 382, ll. 17–18; p. 490, ll. 9–11; vii, p.
545,1. 2. Ibn al-Furat, ix, p. 132, 1. 6; p. 248, 1. 2; p. 249, 1. 9; p. 398, 1. 15. Ibn Shuhba, fol.
59b, 1. 3. Ibn Iyas (KM), iii, p. 79, 1. 23; p. 338, 1. 2. vi, p. 202, ll. 15–16. cf. also: C.I.A.,
L'Égypte, p. 221, p. 539.
5 Abu al-Fida', iv, p. 25, 1. 25. Nujum (P), v, p. 63, 1. 4; vi, p. 151, ll. 6–7; p. 205, 1. 6; p. 777, 1.14.
Ibn Iyas (KM), iii, p. 143, ll. 23–4; p. 149, 1. 10; iv, p. 209, 1. 13. iii, p. 280, 1. 3; vi, p. 201, 1.
5; p. 227, 1. 28.
6See pp. 63–64 and n. 1 on p. 64.
7 Suluk, i, p. 405, ll. 8–12.
8Suluk, i, 418, ll. 8–11.
1Suluk, i, p. 438, ll. 3–9. ii, p. 301. When the Abbasid Caliph comes to Egypt and asks for
Baybars' help, the Mamluk sultan appoints the following office-holders in his service: khazindar,
dawadar, ustadar, wazir (Suluk, i, pp. 452–9).
2Suluk, 699, ll. 4–11.
3 Zetterstéen, p. 24; p. 37; p. 43; p. 57; p. 81; p. 108; p. 130; p. 134.
4Nujum (P), v, p. 349, ll. 8–16. See similar list, pp. 367–8.
5 See, for instance, Nujum (P), vii, p. 237; pp. 259–260; pp. 440–1. pp. 1–3; pp. 22–4; pp.
343–5; pp. 433–4; p. 544. Ibn Iyas (KM), iii, p. 218; p. 386; iv, p. 110; v, pp. 2–3; pp. 90–1.
6Nujum (C), vii, pp. 182–7. ii, p. 113. The pages from an-Nujum az-Zahira cited
above constitute the most important and most detailed testimony in all the published Mamluk
literature on Baybars' role in introducing the laws of the Yasa into the kingdom. According to Ibn
Taghribirdi, yasaq is equivalent to tartib. The origin of the word is si yasa, a word composed of a
Persian and a Mongolian element: si, in Persian 'three', and yasa, in Mongolian 'tartib', and together,
'at-taratib ath-thalatha'. This name emerged from Jinkiz Khan's partition of his domains among his
three sons, and his designating the yasa as legal foundation for the three kingdoms (Nujum (C), vii,
pp. 182, 1. 16–183, 1. 10). It is extremely doubtful whether this is the correct etymology, since Ibn
Taghribirdi himself elsewhere (Nujum (C), vi, pp. 268–9) gives a different explanation of the term,
and a third one in the biography of Jinkiz Khan in his al-Manhal
1Nujum (C), vii, pp. 183–6. The creation of a new office, enterprise, or institution is referred to in the
sources by the verb istajadda (Zetterstéen, p. 102, 1. 18; p. 160, ll. 22–3. Nujum (C), vi, p. 20, 1. 14;
p. 180, ll. 3–4; vii, p. 133, 1. 11. Nujum (P), v, p. 369, ll. 15–20; p. 379, ll. 11–13. Manhal, iii, fol.
64b, ll. 14–15. Suluk, i, p. 269, 1. 6. Ibn al-Furat, ix, p. 94, 1. 1. Ibn Iyas, iv, p. 206, 1. 4. Ta'rif, p.
190, 1. 5). For the evolution and transformation of diverse Mamluk offices, see also Nujum (P), v, p.
311, ll. 2–3; vi, pp. 26–7; p. 356, ll. 6–12; vii, pp. 441, 1. 8–442, 1. 6; p. 442, ll. 6–10, ll. 12–17; p.
443, ll. 11–12. p. 282, ll. 1–2; p. 340, ll. 7–10.
2 Zetterstéen, p. 128, ll. 23–4; p. 210, 1. 19. An-Nahj as-Sadid (in Patrologia Orientalis), xx, p. 99, 1.
1. Suluk, ii, p. 485, 1. 12; p. 497, n. 1; p. 522, ll. 12–19. Nujum (P), v, p. 47. 1. 7; p. 82, 1. 20; vi, p.
15, 1. 22; vii, p. 104, 1. 18, and notes. Manhal, i, fol. 197a, ll. 6–9. Durar, i, p. 406, 1. 7; p. 483, ll.
18–19; iv, p. 367, 1. 4. iv, p. 54, 1. 16.
3 See the few references pertaining to the Circassian period in note 2 above. cf. also C.I.A., L'Égypte,
p. 585. La Syrie, p. 54. Heraldry, p. 121 and n. 3.
1Nujum (P), v, p. 221, ll. 7–13; see also p. 191; p. 195; p. 209; p. 218; p. 222; p. 343; p. 364.
ii, p. 399, 1. 23.
2Nujum (P), v, p. 221; p. 230; pp. 294–5; p. 299; p. 309, ll. 11–13; p. 324, ll. 6–7; p. 344, 1. 20; p.
367, 1. 19; p. 456; p. 521, 1. 5; p. 546, ll. 18–19; vi, p. 56; p. 201; p. 255; p. 312. Manhal, ii, fol. 32a,
ll. 13–14.; fol. 77a, ll. 14–18. ii, p. 313, ll. 15–16. Ibn Taghribirdi's explanation that
means 'amir-father' while atabak means 'amirmother' is quite incomprehensible (Manhal, ii, fol. 42b,
ll. 6–14). cf. also Heraldry, p. 91; p. 123. Glossary to Nujum, vol. vii, p. xvi.
3 On the kharjiya see Appendix B, p. 83 n. 7.
1 The difference between 'wulat' and 'wulat al-aqalim' is not clear.
2 ii, p. 217. Through what is clearly an accidental omission, the source mentions the soldiers
and commanders of the but not the Royal Mamluks and their commanders; we have re-
established the omitted words, rendered obvious by the context.
3 ii, p. 218, 1. 8.
4 i, p. 95,1. 11.
5 i, p. 95, 1. 12.
1Zubda, pp. 103, 1. 22–106, 1. 8.
1Geschichte des Osmanischen Reiches, 2nd ed., i, p. 757.
2 Glossary, art.
Popper's definition of the is as follows: 'Those Mamluks of the sultan who have been long in
service and are in line for promotion to the rank of amir'. (Nujum (P), vi, Glossary.) The second half
of the definition is inaccurate, but this will be discussed below. The first half gives the original
meaning of the term but not the meaning which it later acquired and which, in our view, is
much more frequently encountered in the sources. (See immediately below in the text.) In his
Glossary (pp. xLix ff.) Popper mentions the later meaning (Li). M.Mostapha who, like von Hammer,
writes alludes to it also (Z.D.M.G., 1935, p. 221). But no systematic description of the based on the
bulk of the material supplied by the mamluk sources, has yet been attempted.
3 The plural is much more frequent than the sing.
4 The term is found, in the published Mamluk sources, only in the Circassian period.
5 The writer's elucidation of the term is based solely on its use in Mamluk sources. Its etymology is
unknown to him, and he can express no opinion as to whether the etymology suggested by Popper is
accurate.
VOL. XVI. PART 1. 6
1R.E.I., 1935, p. 244.
2 The term Turks is synonymous here with Mamluks.
3M.E.I, 1935, pp. 243–4.
1Feudalism, p. 2. The references on which Poliak bases his description of the have not been included
in the passages quoted, in order to avoid confusion in the numbering of footnotes. All references of
any importance are, however, dealt with below in our critique of his statements.
2 Feudalism, pp. 28–9.
3Zubda, p. 115, ll. 18–19.
4 See, for instance, Nujum (P), vii, p. 13, ll. 10–20. Ibn Iyas, ii, p. 169, ll. 17–20; (KM), iii, p. 120, ll.
11–14; p. 241, ll. 7–10; iv, p. 60, 1. 11; p. 107, ll. 11–20; p. 281, 1. 9, ll. 12–13; p. 285, ll. 17–20; p.
324, ll. 14–15; p. 358, 1. 17; p. 359, ll. 8–10; p. 427, ll. 21–2; p. 428, ll. 14–22; pp. 443, 1. 21–444, 1.
2; p. 444, ll. 18–20; p. 479, 1. 22; v, p. 12, ll. 8–12; p. 23, ll. 7–8,1. 23; p. 28, ll. 23–9; p. 43, ll. 3–6.
These are only scattered examples; see references in following notes.
5 See, for instance, Nujum (P), vi, p. 16, ll. 10–11; p. 768, ll. 7–9; vii, p. 39, ll. 15–17. p. 175, ll. 10–
12. cf. also Ibn Iyas, iv, p. 383, ll. 2–5, with p. 432, 1. 5; v, p. 43, ll. 3–4.
6 See all notes in which the are mentioned below, as well as n. 4 above.
1 For as distinct from julban, see Nujum (P), vi, p. 768, ll. 7–9; vii, p. 458, ll. 1–2, pp. 4–5; pp. 260–
1; pp. 334–6. Tibr, p. 41, ll. 7–25. Ibn Iyas (KM), iii, p. 231, ll. 12–16; p. 258, ll. 18–21; pp. 256, 1.
11–257, 1. 2; p. 286, n. 2; iv, p. 285, ll. 17–20; p. 358. ll. 16–23. For as distinct from julban and
sayfiya, see Ibn Iyas, ii, p. 169, ll. 17–19; p. 120, ll. 11–14; p. 241, ll. 7–10; p. 312, n. 2; iv, pp. 242,
1. 18–243, 1. 1. See also references in following notes, especially in the section dealing with the
injustices suffered by the in campaigns, payments, and allotment of fiefs.
2Nujum (P), vi, p. 768, ll. 8–9.
3Nujum (P), vi, p. 769, ll. 13–15.
4 Ibn Iyas (KM), iii, p. 286, n. 2.
5 Ibn Iyas, v, p. 27, ll. 6–7.
6 Ibn Iyas, v, p. 43, ll. 3–6.
7 The omission of the sayfiya's name in the campaign of Marj Dabiq can be easily explained by the
insignificance of this unit. It is hardly ever mentioned in any other campaign (note the interesting
exception in connexion with the plague of 903, mentioned above). As for the participation of amirs'
mamluks in the campaign of Dabiq, the source mentions them separately and gives the number of
mamluks under the command of each important amir. The total of these was 944 (Ibn Iyas, v, pp. 42–
3). See below, p. 77, n. 1, for participation of amirs' mamluks in pay parades.
1 See below, on the curtailments of the pay of the The term is no longer frequent in the later
Mamluk period, as already mentioned. The amirs' mamluks received their pay from the amirs, not
together with the Royal Mamluks and the awlad an-nas; it is, therefore, natural that they are not
mentioned in pay parades.
2Nujum (P), vii, p. 450; p. 852, ll. 13–16. p. 678, ll. 7–15; pp. 681, 1. 22–682, 1. 3; p. 682, ll. 12–14.
Ibn Iyas, iii (KM), p.20, ll. 22–23; p. 21, ll. 1–5, ll. 4–9; p. 31, ll. 13–17, p. 271, ll. 12–14; p. 323, ll.
3–5; iv, p. 22; p. 25, ll. 6–15; pp. 65–66.
3Manhal, ii, fols. 192a, 1. 12––192b, 1. 3.
1Manhal, viii, fol. 451a, ll. 7–11.
2 The same conclusion is to be reached from Nujum (P), vii, pp. 13–25; cf. especially pp. 13, 1. 10–
17, 1. 7. cf. also Popper, Glossary to p. Li. The splitting up of the julban into factions, one of them
allying itself with the as happened after the death of al-Ashraf Barsbay, is a very rare phenomenon in
the history of the Mamluk kingdom.
3 Ibn 'Arabshah, at-ta'lif fi shiyam al-malik al-qa'im Abi Sa'id Jaqmaq. Br. Mus. MS., Or.
3026, fol. 116a, ll. 3–10.
4 Ibn Zunbul, pp. 13, 1. 24–14, 1. 2.
5R.E.I., 1935, pp. 243–4.
6 ibid.
1 True, Poliak is of the opinion (R.E.I., 1935, pp. 243–4) that the unit spoken of by (Zubda, p. 115,ll.
17–20) formed only part of the The words of the original, however, do not warrant the conclusion
that this was the Mamluk historian's meaning. mentions the once throughout his work, viz. when he
describes the general composition of the Mamluk army. It would be quite as justifiable to claim that
the other units mentioned by in the above description constituted part of larger units. There is,
incidentally, an obvious contradiction between Poliak's definition of the in R.E.I, 1935, pp. 243–4,
and that given by the same author in Feudalism, p. 2.
2 Nujum (P), vi, pp. 12–18. cf. especially p. 12, ll. 1–3, with pp. 15, ll. 5–8; 16, ll. 10–11; 17, ll. 1–3;
and with p. 18, ll. 1–4.
3 Ibn Iyas, iv, p. 324, ll. 14–15.
4 Ibn Iyas, iv, p. 383, ll. 2–5.
5 Ibn Iyas, iv, p. 448, ll. 4–8; p. 453, ll. 13–19; p. 479, ll. 15–23; p. 480, ll. 10–21; v, p. 23, ll. 1–8;
pp. 28, 1. 23–29, 1. 2; p. 45, ll. 15–22.
6 Ibn Iyas, v, pp. 28, 1. 23–29, 1. 2.
7 Ibn Iyas, iv, p. 480, ll. 10–21.
8R.E.I., 1935, pp. 243–4.
1 See references already listed above, and also others below.
2 The wazir was responsible, inter alia, for the distribution of meat to the Mamluks.
3pp. 250, 1. 19–251, 1. 6.
4 Glossary, art. p. Li. The present writer recognized this distortion before the Glossary of was made
available to him. In a footnote on the same page of the quoted passage ( p. 250), we find that
according to another MS. of the above-mentioned source is to be inserted between 'a'ni' and 'al-
julban', clearly a distortion of
1 See L'Esclavage du mamelouk, pp. 31–4.
2Nujum (P), v, p. 588, ll. 1–3; vii, p. 392, ll. 5–13. Durar, ii, p. 171, ll. 6–8.
3Nujum (C), viii, p. 42, 1. 3. Ibn Kathir, xiv, p. 40, 1. 27. Suluk, i, p. 867, ll. 12–13; ii, p. 19, ll. 4–5;
p. 20, 1. 5. Nujum (P), v, p. 140, ll. 16–22; p. 300, ll. 9–10; pp. 355, 1. 21–356, 1. 7; p. 613, ll. 4–6;
vi, p. 543, ll. 20–1. Manhal, ii, fol. 94b, ll. 2–5; fol. 191a, 1. 1; iv, fol. 210b, ll. 1–2. Ta'rikh Bayrut, p.
54, 1. 1. Durar, i, p. 515, ll. 8–11. cf. also Dozy, Supplément aux dictionnaires arabes, under qadim
hijra.
4 Ibn al-Furat, ix, p. 176, ll. 8–10.
5 Zetterstéen, p. 222, ll. 15–17.
6Nujum (P), vi, p. 543.
7Nujum (P), v, p. 349, ll. 15–16.
8Zubdat al-Fikra, fol. 98b, ll. 7–9.
1Nujum (C), ix, p. 264, ll. 3–6. Nujum (P), v, p. 140; p. 148, ll. 16–22. Manhal, i, fol. 6a, ll. 9–13; fol.
94b, ll. 2–5.
2Manhal, v, fol. 50a, 1. 3.
3Nujum (P), v, p. 300, ll. 9–10; p. 307, 1. 21; vi, p. 40, ll. 11–12. ii, P. 308, ll. 30–1; p. 309, ll. 10–11;
p. 310, ll. 28–9.
4 ii, p. 200, ll. 3–4; p. 209, ll. 6–7. iv, p. 44, ll. 14–15; p. 54, ll. 18–19. p. 254, ll. 21–2; p. 261, 1.
21.
5 Ibn al-Furat, ix, p. 58, ll. 14–15.
6 Tibr, p. 408, 1. 9.
7 xiv, p. 156, ll. 7–8.
8 See some of the references given in n. 4 above.
9Nujum (P), v, pp. 355, 1. 21–356, 1. 7.
10 For additional material on the term akabir al-umara', see Suluk, i, pp. 761–2; p. 788, 1. 14; ii, p.
45, ll. 20–1; pp. 313–314; p. 523, 1. 3. Abu al-Fida', iv, p. 94, ll. 25–6. Zetterstéen, p. 52, ll. 12–16; p.
53, 1. 18; p. 147, 1. 5; p. 176, 1. 4; p. 177, ll. 20–1; p. 182, ll. 19–20; p. 213, ll. 21–2; p. 220, 1. 4; p.
224, ll. 7–9. Ibn Kathir, xiii, p. 293, ll. 5–6. Nujum (P), v, p. 208, 1. 20; p. 209, 1. 5; p. 323, ll. 10–12;
p. 359, 1. 22; p. 360, 1. 7; p. 413, ll. 1–5; p. 541, 1. 23; vi, p. 33, ll. 6–7; p. 93, ll. 11 ff.; p. 166, 1. 8.
Manhal, ii, fol. 18b, 1. 18. Ibn al-Furat, vii, p. 96, 1. 22; p. 97, 1. 5, 1. 9; viii, p. 58, ll. 17–18; p. 222,
ll. 13–14; ix, p. 277, ll. 4–5. ii, p. 316, ll. 14–15.
1R.E.I., 1935, pp. 243–4.
2 Popper gives, in his critique of Poliak Glossary, p. L), a correct though partial explanation of the
misunderstanding which arose in connexion with during the pass-out and liberation parade.
3 The text reads but the correct reading is perhaps
4 The correct reading is undoubtedly as in the second copy of the manuscript of Nujum, and not as
in the first. This can be inferred from the context of the very same line, as well as from the fact that it
was customary to give the kuttabi his horse immediately upon liberation. (See Esclavage du
mamelouk, p. 17.) The word 'khayr' is devoid of any meaning in the context.
5 Nujum (P), vi, pp. 509–510.
6 The elements of Islam and the arts of war were the two principal subjects taught to the young
mamluk in the military school.
7 It should be noted that Mamluk sources often speak of al-umara' as opposed to al-akabir or to al-
umara' al-mashayikh. While the latter are generally mamluks of former sultans, the first are mamluks
of the ruling sultan, whose careers as amirs are but of recent date. The umara' are also called al-
umara' because they were generally chosen from among the of the preceding sultan (Nujum (C), vii,
pp. 265–7. Nujum (P), v, p. 30, ll. 3–5; vi, pp. 12 ff.; p. 35; p. 144, ll. 7–9; p. 772, 1. 4; vii, p. 44, ll.
4–5; p. 235, 1. 10. Manhal, ii, fol. 9a, 1. 18; vii, fol. 35b, ll. 4–5. Abu al-Fida', IV, p. 11, ll. 4–5. Ibn
Kathir,
xiii, p. 280; p. 290, ll. 3–7. An-Nahj as-Sadid (in Patrologia Orientalis), xiv, pp. 466, 1. 2–467, 1. 1.
Zetterstéen, p. 165, ll. 17–18; p. 201, 1. 3; p. 203, 1. 12. Ibn al-Furat, vii, p. 96, ll. 6–17; p. 117, ll. 9–
10; pp. 140 ff.; viii, p. 169, ll. 9–10; p. 171, 1. 16; ix, p. 404, ll. 11–17. Ibn Iyas, v, p. 126, ll. 1–2. i,
p. 91, 1. 9; ii, p. 113, 1. 1. vii, p. 150,1. 7). The sources also speak of umara' as opposed to
umara'kharjiya or barraniya, and as superior to them in prestige. Much material has been gathered by
the writer in this connexion (see, for example, Suluk, i, p. 686, ll. 7–17 and the note; ii, p. 313, ll. 9–
10. Ibn al-Furat, vii, p. 207, ll 5–6; ix, pp. 162, 1. 23–163, 1. 1; p. 163, 1. 6. Nujum (P), vi, pp. 6–7.
Manhal, viii, fol. 437b, ll. 7–10. ii, p. 200, ll. 16–17; pp. 217–19 ; p. 305, 1. 26. iii, p. 376, 1. 10; iv,
p. 48, 1. 9; p. 56, ll. 7–9. cf. Glossary to Nujum, vol. v, pp. xvii–xviii; vol. vi, p. xiii) but, since no
conclusions have as yet been reached, treatment of the question has been omitted from this paper.
1 Al-Maqrizi estimates that was brought to Egypt in 801, and completed his military schooling and
received his liberation certificate in 808, that is, in the days of Sultan Faraj, the son of Barquq. Ibn
Taghribirdi, on the other hand, claims that was already liberated under Barquq, viz. in 801 at the
latest, Barquq having died in that year.
2Nujum (P), vi, pp. 510–511.
3Zubda, p. 115, ll. 17–20.
4 Ibn Iyas, iv, p. 60, ll. 9–13.
5 See above, sections on the mamluks of former sultans and on mamluks of the ruling sultan.
1 After Barsbay's death, the still appeared as a force of considerable weight in the struggle against
the mushtarawat, and were even termed 'notables' (a'yan) (Nujum (P), vi, pp. 12 ff. Manhal, ii, fols.
112a, 192a, 1. 12–192b, 1. 3; viii, fol. 451a, ll. 7–11), but from then their resistance grew weaker.
2 p. 627, ll. 6–8. Ibn Iyas, ii, p. 50, ll. 6–18; pp. 60, 1. 15–51, 1. 27; pp. 67, l. 29–68, 1. l; p. 92, ll.
24–5; pp. 94, 1. 29–95, 1. 2; p. 195, ll. 19–22; (KM), iii, p. 6, ll. 6–7; p. 9, ll. 14–15; p. ll, 1. 5; p. 26,
ll. 1–6; p. 152, ll. 17–21; p. 153, ll. 12–16; p. 154, ll. 4–5, ll. 17–20; p. 165, ll. 7–8; p. 161, ll. 20–3;
iv, p. 19, ll. 3–7.
3 Ibn Iyas, ii, p. 92, ll. 4–5; (KM), iii, p. 6, D. 6–7. iii, p. 207, ll. 6–9. Durar, i, p. 477.
4 Ibn Iyas, ii, pp. 67, 1. 29–68,1. 1. Nujum (P), vii, pp. 660, 1. 12–661, 1. 6.
5 Ibn Iyas, iv, p. 383, ll. 2–5.
6Nujum (P), vi, p. 700, 9–13.
7Nujum (P), vii, p. 677, ll. 7–8; p. 678, ll. 2–5.
1 Ibn Iyas, v, p. 67, ll. 20–1; p. 68, ll. 2–4; p. 70, ll. 8–10. It is interesting to note that the historian
here applies to the julban the words of the Qur'an concerning the hypocrites, al-munafiqun, at the
battle of cf. also Ibn Iyas, ibid., p. 124, 1. 8, ll. 13–15; p. 127, ll. 15–16, and Ibn Zunbul, pp. 13, 1.
23–14, 1. 5; pp. 15, 1. 21–16, 1. 11.
2 cf., for instance, Nujum (P), vii, p. 14, ll. 1–4, ll. 20–2; p. 411, ll. 5–12. p. 171; p. 553, ll. 10–13.
Ibn Iyas, v, p. 22, ll. 10–23; p. 23, ll. 7–8.
3Nujum (P), vi, p. 641, ll. 2–5. cf. also errata on p. Lii of same volume.
4 pp. 174, 1. 16–175, 1. 2; p. 175, ll. 10–12.
5Ibn Iyas (KM), iii, p. 231, 1. 12.
6 Ibn Iyas (KM), iii, p. 131, 1. 12.
7 Ibn Iyas (KM), iii, pp. 256, 1. 11–257, 1. 2.
8 Ibn Iyas (KM), iii, p. 258, ll. 18–21.
9 Ibn Iyas (KM), iii, p. 312, n. 2; p. 313, n. 1. Most of the oppressive measures were, of course, taken
against the new viz. yesterday's ajlab, who had been displaced by the mamluks of the reigning
sultan.
10 Ibn Iyas, iv, pp. 242, 1. 18–243, 1. 3.
11Ibn Iyas, iv, p. 285, ll. 17–20.
12Ibn Iyas, iv, p. 359, ll. 8–10.
13Ibn Iyas, iv, p. 401, ll. 14–23; p. 404, ll. 11–16.
14Ibn Iyas, iv, pp. 430,1. 22–431, 1. 17.
15Ibn Iyas, iv, p. 434, ll. 9–14.
1 Soldiers of the al-khamisa were the arquebusiers of the Mamluk army and held a very inferior
status.
2 Ibn Iyas, iv, pp. 443, 1. 21–444, 1. 3.
3 Ibn Iyas, iv, p. 446, ll. 2–4.
4 Ibn Iyas, iv, p. 448, ll. 4–8.
5 Ibn Iyas, v, p. 12, ll. 9–12.
6 Ibn Iyas, v, p. 61, ll. 8–15.
7 Ibn Iyas, iv, p. 281, ll. 8–13. It is important to note that it was sometimes required of the to pay the
badil, the payment for exemption from military service, in the same manner as the awlad an-nas (Ibn
Iyas, ii, p. 232, ll. 21–2; (KM), iii, p. 214, ll. 21–2; pp. 256, 1. 11–257,1. 2). Such treatment of the is
infrequent, it is true, but in connexion with the julan it is totally unheard of. On account of their
higher age, there were among the a greater proportion of men unfit for military service, and that is
probably the reason why some of them were sometimes required to pay a sum of money instead of
going forth to battle. This fact in no way contradicts the sources' claim that the military competence
of the was higher than that of julban.
8 From Manhal, iii, fol. 186a, ll. 18–23 (more correctly fol. 185b), on which Poliak bases his claim
that the had the right of priority to fiefs (Feudalism, p. 29 and n. 1), precisely the opposite is to be
inferred. Ibn Taghribirdi does not mention the at all in that passage, but he complains bitterly of the
fact that the affairs of the kingdom were turned upside down and its good old usages were totally
transformed, as may be seen from the preference of the Circassians over the other races, and of the
ajlab over the veteran mamluks. It was the ajlab who received the larger fiefs, and it was this
granting of precedence to the over the kabir which, in the historian's view, was one of the main
causes of the decline of the Mamluk kingdom. (See my 'The Circassians in the Mamluk Kingdom',
J.A.O.S., 1949, p. 140). Poliak calls the 'Caucasian nobility 'or 'Caucasian noblemen' (Feudalism, p.
2; cf. also p. 29, top, with the references given in n. 1 and n. 2 on the same page).
1 pp. 334–6.
2 Ibn Iyas, iv, p. 207, ll. 11–20.
3 Ibn Iyas, iv, p. 358, ll. 16–23. Some of Poliak's secondary conclusions must also be contested. The
arguments he adduces from Amir Khushqadam as-Sayfi and from the private soldier Lajin bear no
relation whatever to the If Khushqadam lacked prestige among the members of his own race in spite
of his bravery (Manhal, iii, fol. 48a, ll. 9–10), it was probably because he was a sayfi. Lajin was
admired by the Circassians, but not for being a (cf. 'The Circassians in the Mamluk Kingdom',
J.A.O.S., 1949, pp. 143–4). As for the unpretentious dress worn by Sudun according to the custom of
the former ('ala qa'idat as-salaf min (Manhal, iii, fol. 135b, ll. 9–17), it simply teaches us that
'predecessor' or 'veteran' is synonymous with and that in Ibn Taghribirdi's opinion the Mamluks of
former times were of modest bearing in comparison to his own contemporaries. The historians of the
time clung tenaciously to the view that the Mamluks of bygone days in all respects excelled those of
later times. The mentioned by al-Jabarti (i, p. 412, 1. 31; ii, p. 150, 1. 30) in no way support Poliak's
thesis.
1R.E.I., 1935, pp. 243–4.
VOL. XVI. PART 1.

Saladin and the Assassins

1 Abu Shama, ii, 23–4 (=Goergens, 27–8). Repeated in 234. Cf. Defrémery, 29–30.
2 Ibn otherwise follows fairly closely on 'Imad ad-Din, with some variants. His text begins as
follows:—
It will be noted that here Khumartakin is amir of
3 The Bustan confuses the two attempts.
1 Precise dates for this and the preceding are given only by 'Imad ad-Din (apud Abu Shama) and Ibn
De Sacy, following a different text of Abu Shama, says Friday 19th According to Ibn Saladin set out
from Aleppo on 10th (cf. Maqrizi, Suluk, Cairo, 1934, i, 62).
1
2 Cf. Lewis, 'Sources', 489.
3 On the Nubuwiya see H.Thorning, Beiträge zur Kenntnis des islamischen Vereinswesens, Türkische
Bibliothek, vol. 16, Berlin, 1913, 212–13, and F.Taeschner, 'Das Futuwwa-Rittertum des islamischen
Mittelalters', in Beiträge zur Arabistik, Semitistik und Islamwissenschqft, Leipzig, 1944, 352, n. 17,
where further references are given.
4 Thus the MS. The Jewett version says, absurdly, 13,000 Isma'ili leaders—
5 p. 208. MS. fol. 181a. The form in the Jewett text is an obvious error for as in the MS.
6 pp. 249–250 (translation 259–260). Ibn Jubair, writing in 580 A.H., speaks of these events as
having taken place 'eight years ago'.
7 fol. 146.
8 Guyard 97 and 149. In this version the Isma'ilis are of course victorious.
1 H.A.R.Gibb,' The Achievement of Saladin', Bulletin of the John Rylands Library, vol. 35, 1952,
44–60. On the humiliation of the Isma'ilis at this time see Abu Shama, I, 197.
2 Abu Shama, i, 221.
3 Cf. S.M.Stern, 'The Epistle of the Fatimid Caliph al-Amir'. JRAS., 1950, 20–31.
1 Guyard, 77 ff. and 137 ff.
2 Cf. H.Laoust, Essai sur les doctrines sociales et politiques de b.Taimiya, Cairo, 1939, 124–5, 266–
7, for such accusations in Abu Firas's time.
3 Lewis, 'Three Biographies', 344. One of these stories, that of the threatening letter, is well known
from Ibn Khallikan, Wafayat al-A'yan, Cairo, 1882, ii, 115–6 (=M. de Slane, Biographical
Dictionary, Paris, 1842–1871, iii, 339–341), cf. Lewis, 'Sources', 487.
4 Lewis, 'Three Biographies', 341. Kamal ad-Din quotes a letter of condolence from Sinan to Sabiq
ad-Din 'Ammar ibn ad-Daya, lord of Shaizar, on the death of his brother Shams ad-Din, lord of Qal'at
Ja'bar.
VOL. xv. PART 2. 20
1 A source quoted by Kamal ad-Din in the Bughya (Lewis, 'Three Biographies', 343) mentions a third
attempt on Saladin, in Damascus. But this does not appear to be mentioned by the other authorities.
2 Kamal ad-Din, MS. fol. 1936 ff. (=Blochet, iv, 147–8); cf. Lewis, 'Three Biographies', 338;
Quatremère, 354–5; Defrémery, 8–9.
3 Abu Shama, i, 274–5; Ibn al-Athir, xi, 294–5; Kamal ad-Din, MS. fol. 193b ff. (=Blochet, iv, 148–
9); Bustan, 142; 219; Ibn 200–1, Ibn Shaddad, fol. 128b; cf. Quatremfère, 355–6; Defrémery, 20–
23.
4 Kamal ad-Din, MS., fol. 196 ; Abu Shama, ii, 16 (=Goergens, 22); cf. Quatremère, 356–7,
Defrémery, 24–5.
5 Baha' ad-Din 165; Abu Shama ii, 196 (=Goergens, 185–6); Ibn al-Athir, xii, 51 (=Recueil ii, 58–9);
Bar-Hebraeus, 339; 'Imad ad-Din, 420–2; 269; Ibn 396–7; Quatremère, 357; Defrémery, 25–30;
Lewis, 'Sources', 487–8.
6 Abu Shama, ii, 203; Defrémery, 29.

THE POSITION AND POWER OF THE MAMLUK SULTAN

1 Anon., The life of the Reverend Humphrey Prideaux, D.D., Dean of Norwich, London, 1748, 268–
9.
2 Al-Maqizi, al-Suluk; I, pt. II, Cairo, 1956, 369.
3 Ibn Taghribirdi, al-Nujum al-zahira, VII, Cairo, n.d., 4–5.
4 Stanley Lane-Poole, The Mohammadan dynaslies, Paris, 1925, 83.
5 Précis de l'histoirc d'Égypte, II, Le Caire, 1932, 238.
6 Ibn al-Dawadari, Kanz al-durar, VII, Cairo, 1073, 5; Ulrich Haarmann, Quellensludien, zur frühen
Mamlukenzeit, Freiburg im Breisgau, 1970, p. 229, n. 3.
7Nujum, VII, 273.
8 Ibn 'Abd Tashrif al-ayyam, Cairo, 1961, 94–5.
9Tashrij, 110.
10Kanz al-durar, ix, Cairo, 1960, 126.
11Nujum, XI, 6.
12 Nujum, IX, 261.
13Nujum, VII, 104, 294; VIII, 19; cf. Kanz ul-durar, VIII, Cairo, 1971, 348.
14suluk, I, pt. II, lahu wa al-'askar.
15 S.F.Sadeque, Baybars the First of Egypt, Dacca, 1950, 17 (Arabic text).
16 Cited in Nujum, VIII, 99.
17 Isma'il Abu 'l-Fida', fi akhbar al-bashar, Cairo, n.d., Iv 34.
18 Al-Qalqashandi, al-a'sha, Cairo, n.d., IV, 7.
19C.H.Becker, 'Le "Ghâshiya "comme emblême de la royauté', Centenario della nascila di Michele
Amari, Palermo, 1910, 148–51. Cf. [E.] Quatremère, Histoire des sultans mamelouks de l'Égypte, I, I,
Paris, 1837, p. 3, n. 7.
20 Nujum, vi, 124, 306.
21 IV, 87; Quatremère, op. cit., n. 7 at p. 6.
22 Sadeque, Baybara, 35–41, 61–3 (Arabic text).
23Nujum, x, 4–5.
24Kanz al-durar, VIII, 34, 38, 45.
25Sadeque, Baybara, 9–10, 31 (Arabic text).
26Nujum, VII, 87.
27Kanz al-durar, VIII, 40–3.
28Kanz al-durar, VIII, 271–6; cf. ibid., VII, 8–9.
29 Sadeque, Baybars, 62 (Arabic text),
VOL. XXXVIII. PART 2. 20
30 cf. S.M.Stern,' Petitions from the Mamluk period', BSOAS, XXIX 2, 1960, 268–75.
31Nujum, VIII, 202.

CASSIODORUS AND RASHID AL-DIN ON BARBARIAN RULE IN

ITALY AND PERSIA1
1I am greatly indebted to Professor A.K.S.Lambton, Professor P.R.L.Brown, Mr. R.M. Burrell, Mr.
M.A.Cook, and Mr. J.C.Gough, who read and commented on an earlier draft of this paper.
2 ed. T.Mommsen, MGH, Auctores Antiquissimi, XII, Berlin, 1894: abridged English translation by
T.Hodgkin, The letters of Cassiodorus, London, 1886.
3 ed. M.Shafi', Lahore, 1947.
4Getica, ed. T.Mommsen, MGH, Auctores Antiquissimi, v, Berlin, 1882; English translation by
C.C.Mierow, The Gothic history of Jordanes, Princeton, 1915.
5Turkestan down to the Mongol invasion, London, 1928, 46.
6 P.R.L.Brown: The world of late antiquity, London, 1971, 128.
7 Cassiodorus's life and works are discussed e.g. by L.W.Jones, in the introduction to his translation
of the Institituiones (An introduction to divine and human readings, New York, 1946, 1–64).
8 ibid., 8.
9Getica, 243: tr. Mierow, 119.
10Getica, 290–2: tr. Mierow, 136–6.
11 H.M.Jones, 'The constitutional position of Odoacer and Theoderic', Journal of Roman Studies,
LII, 1962, 126–30. I quote from the reprint in A.H.M.Jones, The Roman economy, ed. P.A.Brunt,
Oxford, 1974, 367.
12Getica, 294–5: tr. Mierow, 136.
13 For a general account of the Mongol period, see J.J.Saunders: The history of the Mongol
conquests, London, 1971. On the invasions of Persia, Cambridge History of Iran, v, ed. J.A. Boyle,
Cambridge, 1968, ch. iv (J.A.Boyle) and vi (I.P.Petrushevsky).
14Al-Kamil, ed. C.J.Toraberg, XII, Leiden, 1853, 234; repr. Beirut, 1966, 358–9. Translation from
B.Spuler, History of the Mongols, London, 1972, 30.
15 On Rashid al-Din's life and works, see e.g. J.A.Boyle's introduction to his translation of part of the
Jami'. dl-tavarikh, The successors of Genghis Khan, New York and London, 1971, 3–13.
16 ibid., 3.
17 This was not an unusual conclusion for a Vazir's career in Mongol Persia. The eventual death of
Taj al-Din 'Ali Shah, still in office, in his bed, of natural causes, was noted with some surprise.
Contemporaries maintained that he was the only Vazir of the Mongols who had the good fortune to
die naturally. See Allah Mustawfi, Tarikh-i Guzida, ed. 'Abd Nava'i, Tehran, 1339/1960, 616.
18 T.Hodgkin, Theodoric the Goth, London and New York, 1891, 162.
19 ibid., 25.
20Cassiodorus and Italian culture of his time', Proceedings of the British Academy, XLI, 1966, 207–
45. I cite the reprint in A.Momigliano, Studies in historiography, London, 1966, 183.
21 P.Llewellyn, Rome in the dark ages, London, 1970, 33.
22Variae, v, 2: tr. Hodgkin, Letters, 265.
23Variae, I, 35: tr. Hodgkin, Theodoric, 169–70.
24Variae, ix, 25: tr. Hodgkin, Letters, 412.
25 Brown, op. cit., 128.
26Getica, 60: tr. Mierow, 67.
27 Momigliano, art. cit., 192–3.
28 B.Lewis, 'The Muslim discovery of Europe', Islam in history, London, 1973, 98.
29 Rashid al-Din, Histoire des Mongols de la Perse, ed. and tr. E.Quatremère, Paris, 1836, 38.
30 ibid., 60–2.
31 It has been inferred, from the remarkably Papacy-centred account of European history to be found
in the 'History of the Franks', that Rashid al-Din's Frankish informant was probably a cleric—perhaps
a friar. See Histoire des Francs, ed. and tr. K.Jahn, Leiden, 1951.
32A literary history of Persia, III, London, 1920, 68.
33Tarikh-i Uljaytu, ed. M.Hambly, Tehran, 1969, 240.
34 See R.Levy, 'The letters of Rashid al-Di ', JRAS, 1946, 1–2, pp. 74–8.
35 Anonymus Valesianus, in Ammianus Marcellinus, ed. and tr. J.C.Rolfe (Loeb Classical Library),
III, London, 1939, 546.
36 Momigliano, art. cit., 190–1.
VOL. XL. PART 2. 22
37 Llewellyn, op. cit., 31. Cf. C.S.Lewis: 'His (Theodorie's) reign in Italy was not a sheer monstrosity
as, say, the rule of Chaka or Dingaan in nineteenth-century England would have been. It was more as
if a (popish) highland chieftain (who had acquired a little polish and a taste for claret in the French
service) had reigned over the partly Protestant and partly sceptical England of Johnson and Lord
Chesterfield'. (The discarded image, Cambridge, 1964, 75–6.)
38 A.Chastagnol, Le sénat romain sous le règne d'Odoacre, Bonn, 1966, 52. I owe this reference to
Professor Brown.
39 ibid., 56.
40Variae I, 18: tr. Hodgkin, Letters, 154.
41 B.Nicholas, An introduction to Roman law, Oxford, 1962, 122.
42 Rashid al-Din, History of Ghazan Khan, ed. K.Jahn, London, 1940, 221–2: Jami' al-tavarikh, III,
ed. A.Alizade, Baku, 1957, 431.
43 Rashid al-Din also ascribes it to the Saljuq sultan Malik Shah.
44 ed. Jahn, 225: ed. Alizade, 434.
45 J.Schacht, An introduction to Islamic law, Oxford, 1964, 138. If in this instance Islamic law owed
anything to Roman law, the enactments of Theodoric and Ghazan may possibly have a common
ultimate origin. (I am indebted to Mr. Cook for this interesting suggestion.)
46 tr. Hodgkin, Letters, 259–60.
47 G.Clauson, An etymological dictionary of pre-thirteenth-century Turkish, Oxford, 1972, 933.
48 See H.Yule and H.Cordier, The book of ser Marco Polo, I, 3rd ed., London, 1903, 433–8.
49 On the Mongol postal system in China, see P.Olbricht, Das postwesen in China unter der
Mongolenherrschaft, Wiesbaden, 1954.
50Histoire secrète des Mongols, ed. L.Ligeti, Budapest, 1971, 257–8.
51 Juvayni, Tarikh-i Jahan Gusha, III, ed. M.Qazvini, London, 1937, 76–7: tr. J.A.Boyle, The history
of the world-conqueror, II, Manchester, 1958, 599. I have modified the translation slightly.
52History of Ghazan Khan, 271–2: Jami' al-tavarikht, III, 480–1.
53An embassy to Tamerlane, tr. G.Le Strange, London, 1928, 155–7, 177–9.
54Variae, v, 14: tr. Hodgkin, Letters, 273–4.
55Variae, v, 39: tr. Hodgkin, Letters, 287–8.
56Variae, II, 38: tr. Hodgkin, Letters, 191.
57Variae, IV, 50: tr. Hodgkin, Letters, 261–2.
58Variae, IV, 36: tr. Hodgkin, Letters, 253.
59 See below, p. 319, n. 94.
60Mukatabat-i Rashidi, 12.
61History of Ghazan Khan, 251: Jami' al-tavarikh, III, 460. See also A.K.S.Lambton, Landlord and
peasant in Persia, London, 1953, repr. 1969, 83.
62Geschichte der Abaga bis ed. K.Jahn, 's-Gravenhage, 1957, 68: Jami' al-tavarikh, III, 209.
63History of Ghazan Khan, 246: Jami' al-tavarikh, III, 455.
64 See K.Jahn, 'Paper currency in Iran', Journal of Asian History, IV, 2, 1970, 101–35.
65History of Ghazan Khan, 248: Jami' al-tavarikh, III, 457.
66History of Ghazan Khan, 350: Jami' al-tavarikh, III, 558. See also I.P. Petrushevsky in Cambridge
History of Iran, v, 491.
67History of Ghazan Khan, 183: Jami' al-tavarikh, III, 392.
68History of Ghazan Khan, 269: Jami' al-tavarikh, III, 478. See also Petrushevsky, op. cit., 494, and
'Rashid al-Din's conception of the state', Central Asiatic Journal, xiv, 1–3, 1970, 155.
69Nuzhat al-qulub, ed. G.Le Strange, London, 1915, 27.
70Mukatabat-i Rashidi, 33.
71 ibid., 274.
72Variae, II, 5: tr. Hodgkin, Letters, 173.
73Variae, III, 17: tr. Hodgkin, Letters, 206.
74Variae, VIII, 3: tr. Hodgkin, Letters, 349–50.
75Variae, VIII, 26: tr. Hodgkin, Letters, 375.
76Variae, VII, 3: tr. Hodgkin, Letters, 321–2.
77 W.G.Sinnigen, 'Administrative shifts of competence under Theodoric', Traditio, XXI, 1965, 458.
78Variae, II, 27: tr. Hodgkin, Letters, 186.
79 See e.g. Anonymus Valesianus, 558–60.
80 The famous debate in which the Franciscan William of Rubruck argued before the Great Khan
Möngke against Muslim and Buddhist spokesmen might be held to suggest more than this. See
Sinica Franciscana, I, ed. A.van den Wyngaert, Quaracchi-Firenze, 1929, 294–7.
81History of Ghazan Khan, 218: Jami' al-tavarikh, III, 427.
82History of Ghazan Khan, 188: Jami' al-tavarikh, III, 396–7.
83 tr. E.A.W.Budge, The monks of Kublai Khan, London, 1928. Syriac text ed. P.Bedjân, Histoire de
Mâr Jab-aldha, patriarche, et de Rabban Sauma, 2nd ed., Paris, 1895.
84 e.g. Variae, I, 20; IV, 49: tr. Hodgkin, Letters, 156; 263.
85 e.g. Mukatabat-i Rashidi, letters 17, 32, 51.
86 ibid., letters 38, 39.
87Decline and fall of the Roman Empire, ch. xxxix.
88Italy and her invaders, III, 2nd ed., Oxford, 1896, 247.
89Early Germanic kingship in England and on the continent, Oxford, 1971, 9.
90 Lambton, op. cit., 92.
91 op. cit., 544.
92 Jones, op. cit., 373.
93 Sinnigen, art. cit., 466.
94 On Mongol taxation in Persia, see J.Masson Smith, 'Mongol and nomadic taxation', HJAS, xxx,
1970, 46–85, and the references contained therein, and A.K.S.Lambton, article kharadj, EI, 2nd ed.
(forthcoming).
95Nuzhat al-qulub, 27.
96 Awliya Allah Amuli, Tarikh-i Ruyan, ed. M.Sutuda, Tehran, 1348/1969, 178–80.
 THE TREATIES OF THE EARLY MAMLUK SULTANS WITH THE
FRANKISH STATES
1 The treaties referred to are the following:
(a) Baybars and the Hospitallers: 4 665/29 May 1267. Al-Qalqashandi, dl-a'sha, Cairo
n.d., xiv, 31–9.
(b) Baybars and the Lady Isabel of Beirut: 6 667/9 May 1269. XIV, 39–42.
(c) Baybars and the Hospitallers: 1 669/13 April 1271. XIV, 42–51.
(d) Qalawun and Bohemond VII of Tripoli: 17 Rabi' 1680/6 July 1281. Baybars Zubdat al-
fikra, excerpted in al-Maqrizi, Kitab al-suluk (ed. M.M.Ziada), r/3, Cairo, 1970, 975–7. Ibn al-Furat,
Ta'rikh, excerpted in Ibn 'Abd Tashrif al-ayyam (ed. Murad Kamil), Cairo, 1961, 210–11.
(e) Qalawun and the Templars: 5 681/16 April 1282. Ibn 'Abd Tashrif, 20–2. Arabic
text and French translation in E.Quatremère, Histoire des sultans mamlouks, Paris, 1837–45, I/1,
177–8, 221–3. English translation (through Italian) in Francesco Gabrieli, Arab historians of the
Crusades, London, 1969, 323–6.
(f) Qalawun and the authorities in Acre: 5 Rabi' I 682/3 June 1283. XIV, 51–63. Ibn 'Abd
Tashrif, 34–43. Zurayq (ed.), Ta'rikh Ibn al-Furat, VII, Beirut, 1942, 262–70.
Quatremère, Histoire, I/1, 179–85, 224–30. Gabrieli, Arab historians, 326–31.
(g) Qalawun and the Lady Margaret of Tyre: 14 Jumada I 684/18 July 1285. Ibn 'Abd
Tashrif, 103–10. Quatremfère, Histoire, I/1, 172–6, 212–21.
2 XIV, 8.
3 Ibn'Abd al-zahir (ed. 'Abd al-'Aziz 1396/ 1976, 118. The lord of
Jaffa and the ruler of Beirut were respectively John of Ibelin, count of Jaffa (1250–66) and his
namesake the lord of Beirut (1247–84).
4 xiv, 70–1.
5Tashrif, 34, 43.
6 XIII, 311–14.
7 Shafi' b.'Ali, al-ma'thur min sirat al-Malik Bodleian MS Marsh 424, fols.
106b-107b. Shafi' gives another account of the incident in al-manaqib al-sirriyya al-muntaza'a
min al-sira (ed. 'Abd al-'Aziz ), 1396/ 1976, 127–8.
8 al-ma'thur, fols. 107b–109a; al-manaqib, 132–4.
9 J.Prawer, Histoire du royaume latin de Jérusalem (second ed.), Paris, 1975, II, 523.
10 xiv, 60; Tashrif, 43.
11Tashrif, 109.
12Tashrif, 108.
13 The Arabic is wali tilka al-wilaya/-jiha, wali al-makan. Wali is probably here synonymous with
ra'is; cf. below (p. 73), where the ru'asa' of the locality act in the event of robbery or homicide.
Clauses in Baybars's first treaty with the Hospitallers dealing with procedure in cases of missing
property are introduced with the phrase 'the discharge of the ru'asa''. The term ra'is (in the
form rays) was specifically used of a village headman in the Latin kingdom; cf. Jonathan Riley-
Smith, The feudal nobility of the kingdom of Jerusalem 1174–1277, [London, 1973], 47–9.
14 XIV, 57.
15Tashrif, 108.
16 XIV, 48.
17 XIV, 38.
18 XIV, 61. This clause is not given in Tashrif.
19 Such clauses derive their validity from the Islamic legal concept of aman (safe-conduct), which
was sometimes embodied in specific instruments: of. J.Wansbrough, 'The safe-conduct in Muslim
chancery practice', BSOAS, XXXIV/1, 1971, 20–35.
20 XIV, 41.
21 G.L.F.Tafel and G.M.Thomas (ed.), Urkunden zur älteren Handels- und Staatsgeschichte der
Republik Venedig, III, Vienna, 1857, 336.
22 XIV, 47.
23 Specimens of the decrees of Mamluk sultans, which were the instruments embodying the final
results of commercial negotiations, have been published and examined by J.Wansbrough, 'A Mamluk
commercial treaty concluded with the Republic of Florence 894/1489', in S.M. Stern (ed.),
Documents from Islamic chanceries, Oxford [1965], 39–79; idem, 'Venice and Florence in the
Mamluk commercial privileges', BSOAS, XXVIII/3, 1965, 483–523.
24Tashrif, 108.
25 Ibn 'Abd 266, 283.
VOL. XLIII. PART 1. 6
26 Shafi' b. 'Ali, al-ma'thur, ff. 127a-128b; al-manaqib, 138–40.

THE MONGOL EMPIRE: A REVIEW ARTICLE

1 Luc Kwanten, Imperial nomads: a history of central Asia 500–1500, XV, 352 pp., 10 plates.
Leicester: Leicester University Press (1979). £13.
2The empire of the steppes, tr. N.Walford, New Brunswick, 1970.
3 Introduction to Grousset, Conqueror of the world, Edinburgh and London, 1967, XIV.
4 Published in the Ancient peoples and places series, London.
5 For earlier discussion of the Englishman, see e.g. Henry Kingsley, Tales of old travel, London,
1869. Since then he has often been mentioned, e.g. by Brent, by Saunders in his article on 'Matthew
Paris and the Mongols', in T.A.Sandquist and M.R.Powicke (ed.), Essaya in Medieval history
presented to Bertie Wilkinson, Toronto, 1969, 116–32, and by Chambers in the book discussed below.
6 See, e.g., his articles collected as The Mongol world empire 1206–1370, London, 1977.
7 As illustrated, e.g., in the useful collection of texts he edited as History of the Mongols, referred to
above.
VOL. XLIV. PART 1. 9
8The history of the Mongol conquests, 1.

SALADIN AND HIS ADMIRERS: A BIOGRAPHICAL

REASSESSMENT1
1 Malcolm Cameron Lyons and D.E.P.Jackson, Saladin: the politics of the Holy War. (University of
Cambridge Oriental Publications.) viii, 456 pp. Cambridge, etc.: Cambridge University Press, 1982.
£25.
2 Hamilton [A. R.] Gibb, The life of Saladin, Oxford, 1973, 1.
3 Albert Hourani, 'H.A.R.Gibb: the vocation of an orientalist', Europe and the Middle East, [London,
1980], at p. 106.
4 H.A.R.Gibb, 'The achievement of Saladin', Studies on the civilisation of Islam, Boston, [1962], at p.
100.
5 D.S.Richards, 'A consideration of the sources for the life of Saladin', Journal of Semitic Studies,
xxv, 1, 1980, at p. 61.
6 Beryl Smalley, Historians in the Middle Ages, London, [1974], 67.
7'Or j'ai adopté, pour départ de ma chronologie, une seconde hégire qui atteste que le terme de la
première sera marqué par la résurrection et que sa promesse est la vraie, celle qu'on ne réfute pas,
celle qui est évidente, celle qu'on ne fausse pas. Cette hégire, c'est l'émigration de I'lslam vers
Jérusalem. Son acteur est le sultan ad-Dîn Abu Yûsuf b Ayyûb.'
'Imâd ad-Dîn Conquête de la Syrie et de la Palestine par Saladin, tr. Henri Massé, Paris,
1972, 6.
8Andrew S.Ehrenkreutz, Saladin, Albany, N.Y., 1972, 238.
9 al-Nabarawi, Sana al-barq al-Shami, Cairo, 1979.
10Lyons and Jackson, Saladin, 1–2.
11 Lyons and Jackson, Saladin, 172.
12 Lyons and Jackson, Saladin, 201.
13 Lyons and Jackson, Saladin, 286.

SOME OBSERVATIONS ON THE 'ABBASID CALIPHATE OF CAIRO

1 al-Din Ibn 'Abd al-zahir sirat al-Malik (ed. 'Abd al-'Aziz
), [1396/1976], 99–112, 141–8.
2 al-Wafi bi'l-wqfayat, vi (ed. Sven Dedering), Wiesbaden, 1972, 317–18 (no. 2819).
3 Wafi, VII (ed. 'Abbas), Wiesbaden, 1969, 384–6 (no. 3378).
4 Shafi' b.'AIi, al-manaqib al-sirriyya al-muntaza'a min al-sira (ed. 'Abd al-'Aziz
b.'Abdallah ), [ 1396/1976], 46.
5 Ibn 'Abd 173–4.
6 For a valuable study of the dynasty, see J.-C.Garcin, 'Histoire, opposition politique et piétisme
traditionaliste dans le de Suyûti', Annales Islamologiques, VII,, 1967, 33–90.
7 Garcin, 'Histoire', p. 55, n. 1, states that 'L'investiture accordée par le calife au sultan est notée par
Maqrizi à partir de Malik 'Adil Katbugha en 694'.
8 Ibn Taghribirdi, al-Nujum al-zahira (Cairo edn.), x, 4–5; tr. P.M.Holt, 'The position and power of
the Mamluk sultan', BSOAS, XXXVIII, 2, 1975, 244.
9 e.g. on the accession of : baya'ahu al-khalifa (Ibn Taghribirdi, Nujum, x,
60); (783/1381): fa-baya'ahu al-khalifa lahu al-umara' (Nujum, XI, 207);
(791/1389): wa-qad al-khalifa wa-baya'uhu (Nujum, XI,
319): (824/1421): fa-lamma rahu al-khalifa gama lahu wa-ajlasahu bi-janibihi wa-
baya'ahu (Nujum, xiv, 211); b.Qayitbay
(901/1496): fa-baya'ahu al-khalifa (Ibn Iyas, Bada'i' al-zuhur, III, 324);
(904/1498): fa-baya'ahu al-khalifa (Bada'i', III, 405); al-Ashraf (922/1516): fa-
baya'a amir al-mu'minin Ya'qub niyabatan 'an waladihi al-Mutawakkil (Bada'i', v,
105), tr. Gaston Wiet, Journal d'un bourgeois du Caire, II [Paris], 1960, 96, 'Le sultan done le
serment de 1'émir des croyants Ya'qûb, mandaté par son fils Mutawakkil'. It should be noted,
however, that on the accession of the Caliphs al-Mustakfi II (845/1441) and al-Qa'im (855/1450), Ibn
Taghribirdi says kanat mubaya'at al-khalifa [fulan] bi'l-khilafa (Nujum, xv, 349, 432) without
specifying who performed the bay'a to the caliph on these occasions.
10 Khalil b.Shahin Zubdat kashf al-mamalik (ed. Paul Ravaisse), Paris, 1894, 54.
11 Khalil Zubdat, 90.
12 Ibn Taghribirdi, Nujum, VIII, 263.
13 Ibn Taghribirdi, Nujum, IX, 8.
14 Ibn Taghribirdi, Nujum, XIII, 189–208.
15 Ibn Taghribirdi, Nujum, XIII, 204.
16 Ibn Mufakahat al-khillan (ed. Mohamed Mostafa), II, Cairo, 1964, 78, 90. The
corresponding passage for A.H. 925 is lost.
17 'Abd al-Jabarti, 'Aja'ib al-athar, III, 355–6.
18 Various references in the genealogies (none ancient) translated in H.A.MacMichael, A history of
the Arabs in the Sudan, Cambridge, 1922, ii. The text of a genealogy compiled in 1361/1942,
'Abdallah al-Khabir, Hadha jami' nasab al-J'aliyyin al-musamina bil-sur has
recently been published by 'Abdallah 'Ali Ibrahim of the Institute of African and Asian Studies,
Khartoum [n.d.]. See also P.M.Holt, 'The genealogy of a Sudanese holy family', BSOAS, XLIV, 2,
1981, 262–72.

THE 'GREAT YASA OF CHINGIZ KHAN' AND MONGOL LAW IN THE

ILKHANATE
1 D.Ayalon, 'The Great Yasa of Chingiz Khan: a reexamination', A, Studia Islamica, 33, 1971, 97–
140; B, 34, 1971, 151–80; C(l), 36, 1972, 113–58; C(2), 38, 1973, 107–56. (Hereafter Ayalon.)
2 Paris, 1710. English translation as The history of Genghizcan the Great, London, 1722. I quote
from the translation.
3 ibid., 78.
4 ibid., 79.
5 Tientsin, 1937. See also C.Alinge, Mongolische Gesetze, Leipzig, 1934; G.Vernadsky, 'The scope
and contents of Chingis Khan's Yasa', Harvard Journal of Asiatic Studies, III, 1938, 337–60; idem,
'Juwaini's version of Chingis Khan's Yasa', Annales de l'Institut Kondakov, XI, 1939, 33–45; idem,
The Mongols and Russia, New Haven and London, 1953, 99–110; M.Haider, 'The Mongol traditions
and their survival in Central Asia (XIV–XV centuries)', Central Asiatic Journal, XXVIII/1–2, 1984,
57–79. P.Ratchnevsky, 'Die Yasa Cinggiskhans und ihre Problematik', Schriften zur Geschichte
und Kultur des alten Orients 5: Sprache, Geschichte und Kultur der altaischer Völker, Berlin, 1974,
471–87, is an interesting discussion of the problems. See also his Cinggis-khan sein Leben und
Wirken, Wiesbaden, 1983, 164–72.
6 L.Ligeti, ed., Histoire secrète des Mongols, Budapest, 1971, 173–4. Translation from I. de
Rachewiltz, 'The Secret History of the Mongols', Papers in Far Eastern History, 21, March 1980, 27.
For another translation see F.W.Cleaves. The Secret History of the Mongols, Cambridge, Mass., 1982,
143–4.
7 Tr. de Rachewiltz, PFEH, 13, March 1976, 46–7; ed. Ligeti, 109; tr. Cleaves, 81.
8 On the 'blue book' see P.Pelliot, 'Notes sur le "Turkestan" de M.W.Barthold', and 'Les kökö-däbtär
et les hou-k'eou ts'ing-ts'eu', T'oung Pao, 27, 1930, 38–42 and 195–8. Pelliot does not suggest a link
between the incident of 1206 and the Yasa.
9 op. 9 27 Riasanocsky's italics.
10 Rashid al-Din, Sbornik letopisei, ed. and tr. I.N.Berezin, Trudy vostochnago otdêleniya
Imperatorskago Russkago Arkheologicheskago Obshchestva, XIII and xv, 1868–88. The text only is
cited. (Hereafter Berezin). XIII, 238–9.
11 Berezin xv 65.
12 Ayalon, A, 139.
13 Juwayni, Ta'rikh-i Jahan Gusha, ed. M.M.Qazwini, 3 vols, Leiden and London, 1912, 1916, 1937
(hereafter Juwayni), I, 17–18; tr. J.A.Boyle, The history of the world conqueror, 2 vols, Manchester,
1958 (hereafter Boyle), I, 25. I have made a number of changes in Boyle's translation of this passage.
14 Ayalon, A, 133.
15 Ayalon, A, 134–5.
16 Juwayni, I, 17; Boyle, I, 24.
17 Juwayni, III, 3; Boyle, II, 549.
18 K.Lech, Das mongolische Weltreich, Wiesbaden, 1968, text, 8, tr., 95.
19 ed. Bombay, 1852–3, 17.
20Majmu'a-i rasa'il-i khwaja ed. Tehran, 1957, 31. See V.Minorsky, ' al-Din
on finance', in his Iranica, Tehran, 1964, 70.
21 P.Pelliot and L.Hambis, Histoire des campagnes de Gengis Khan, I, Leiden, 1951, introduction.
22 Ayalon, B.
23 66.
24 F.W.Cleaves, 'The "fifteen 'palace poems'" by Literatur Chiu-ssu', HJAS, xx, 1957, p. 428, n. 10;
pp. 429–33, nn. 14–15. I owe this reference to the late Professor Joseph Fletcher. The remarks on the
Yasa in P.Ch'en, Chinese legal tradition under the Mongols: the code of 1291 as reconstructed,
Princeton, 1979, esp. 4–8, while accepting the authority of Riasanovsky and being 'pre-Ayalon' on the
Yasa's contents, do not seem to show that the evidence of the Chinese sources is irreconcilable with
the arguments advanced in this paper. Indeed, it has been said that the Mongol Yüan dynasty was
unique in Chinese history in that it did not have a formal penal code. It is even suggested that the
notion of such codes was meaningless to the Mongols, and that they preferred to rule through
individual regulations and legislation in China. See J.D.Langlois, Jr., in Langlois, ed., China under
Mongol rule, Princeton, 1981, p. 10, n. 20, citing Uematsu Tadashi.
25 Juwayni, I, 29; Boyle, I, 40.
26 Juwayni, I, 227; Boyle, I, 272.
27 Juwayni, I, 162; Boyle, I, 206.
28Jami' al-tawarikh, II/1, ed. A.A.Alizade, Moscow, 1980, 183–6; tr. J.A.Boyle, The successors of
Genghis Khan, New York and London, 1971, 77.
29 284.
30 ed. ' 2 vols, 2nd ed., Kabul, 1342–3/1964–5, II, 152; tr. H.G. Raverty,
2 vols, London, 1881, II, 1108.
31 Juwayni, I, 211; Boyle, I, 256. I have altered the translation.
32 Lech, op. cit., text, 41, tr., 118–19.
VOL. XLIX. PART 1. 12
33 Jawi' al-tawarikh, III, ed. A.A.Alizade, Baku, 1957 (hereafter Alizade), 511.
34 ed. M.Hambly, Tehran, 1969, 98, with additional word (omitted from the edition) from the unique
MS, Aya Sofya 3019, f. 178a.
35 Abu Bakr al-Ahri, Ta'rikh-i Shaikh Uwais, ed. and tr. J.B.van Loon, The Hague, 1954,
text, 158, tr., 59.
36 Amuli, Nafa'is al-funun, ed. A.Sha'rani et al., 3 vols, Tehran, 1377–9/1957–
60, II, 250.
37The travels of Ibn tr. H.A. R.Gibb, in, Cambridge, Hakluyt Society, 1971, 560–1.
38 Alizade, 171–2, 199, 202, 205, 226, 305, 313, 317, 327, 343, 363.
39 ibid., 359.
40Dhayl-i jami' al-tawarikht, ed. K.Bayani, 2nd ed., Tehran, 1350/1972–3, 76.
41 Juwayni, I, 36; II, 233–6; III, 48, 52.
42 e.g. Alizade, 202, 204, 313.
43 ed. M.Bastani-Parizi, Tehran, 2535/1976–7, 156.
44 ibid., 192.
45 ibid., 275–6.
46 ed. A.A.Alizade, Moscow, 1976 (hereafter Dastur), 29–35. Parts of this material, without any
precise identification, were quoted by Riasanovsky, op. cit., 41–2.
47 See above, n. 6. The word also occurs elsewhere in the Secret History, ed. Ligeti, 202. This adds
little except the information that Shigi-Qutuqu might have assistants in hearing
48 Jami' al-tawarikh, I/1, ed. A.Romaskevich, L.Khetagurov and A.A.Alizade, Moscow, 1965, 180.
49Dastur, 30.
50 ibid., 32.
51 ibid., 31, 32, 33.
52 ibid., 30, 32. An interesting later parallel is found in a document from the shrine at
Ardabil, dating from 793/1390–1. In this it is claimed that a horse was made off with in accordance,
not with the shar' of but with the yarghu (not yasa) of Chingiz Khan. (Document no. 282:
unpublished. I am indebted to Mr. A.H.Morton for the loan of his transcript of the document.)
53Dastur, 30. Tura is a synonym for yasa.
54 ibid., 31. As was remarked above, a large selection of Chingiz Khan's biligs (bilik) is preserved by
Rashid al-Din. See Berezin, xv, 178 ff. On how these were taken down and recorded, see Boyle in
The successors of Genghis Khan, 13.
55Dastur, p. 33, n. 11.
56 ibid., 31.
57 ibid., 35.
58 ibid., 31, 33, 34.
59 ibid., 31.
60 ibid., 31.
61 ibid., 32, 34, 35.
62 A remote ancestor of this article formed part of an unpublished doctoral thesis, the research for
which was supervised by Professor A.K.S.Lambton.

THE ILKHAN EMBASSIES TO QALAWUN: TWO

CONTEMPORARY ACCOUNTS
1 Bodleian, MS Marsh 424. A portion was published with translation by Axel Moberg, 'Regie-
rungspromemoria eines ägyptischen Sultans', Festschrift Eduard Sachau, Berlin, 1915, 406–21.
2 ed. Murad Kamil, [Cairo, 1961].
3sic in MS, where two successive folios are numbered 39.
4 Shafi' modestly describes how he clinched negotiations with Tripoli in 669/1271 and discovered a
casus belli with Acre in 689/1290; cf. P.M.Holt, 'The treaties of the early Mamluk sultans with the
Frankish states', BSOAS, XLII, 1, 1980, at pp. 70–1, 76; idem. 'Some observations on Shafi' b.'Ali's
biography of Baybars', Journal of Semitic Studies, xxix, 1, 1984, at pp. 127–9.
5 The letter and Qalawun's reply were printed from the then unpublished MS of Tashrif al-ayyam
with a French translation by E.Quatremère, Histoire des sultans mamlouks de l'Égypte, Paris, 1845,
II, 158–66, 185–201.
6 By a curious coincidence, a similar proposal was made in the same year (1282) by Charles of
Anjou to Peter of Aragon, when disputing over Sicily; cf. Steven Runciman, The Sicilian Vespers,
Cambridge, 1958, 236.

THE CRUSADES OF 1239–41 AND THEIR AFTERMATH

1 Two studies devoted specifically to these crusades are Reinhold Röhricht,' [Die] Kreuzzüge [des
Grafen Theobald von Navarra und Richard von Cornwallis nach dem heiligen Lande]', Forschungen
zur Deutschen Geschichte, XXVI, 1886, 67–102; and Sidney Painter, 'The Crusade of Theobald of
Champagne and Richard of Cornwall, 1239–1241', in The later crusades 1189–1311, ed. R.L.Wolff
and H.W.Hazard (A history of the crusades, II), 2nd edition (Madison, Wisconsin, 1969), 463–85. For
the years 1241–44, see Marie-Luise Bulst(-Thiele), '[Zur Geschichte der] Ritterorden [und des
Königreichs Jerusalem im 13. Jahrhundert his zur Schlacht bei La Forbie am 17. Okt. 1244]',
Deutsches Archiv, XXII, 1966, 197–226.
2'[L'Estoire de] Eracles [Empereur et la Conqueste de la Terre d'Outremer]', R[ecueil des]
H[istoriens des] C[roisades.] Historiens Occidentaux (Paris, 1844–95), II, 413–22, 427–
31.'[Continuation de Guillaume de Tyr, de 1229 à 1261, dite du manuscrit de] Rothelin', ibid., 526–
56, 561–6. The version of the Eracles is reproduced, with some additional details, in '[Les] Gestes
[des Chiprois]', RHC Documents Arméniens (Paris, 1869–1906), II, 725–8.
3 Bernard Lewis, 'Ifrandj', E[ncyclopaedia of] I[slam], new ed. (Leiden and London, 1954–), III,
1045.
4 Ibn Shaddad's al-A'laq fi dhikr umara' al-Sham wa'l-Jazira is available in a number of
partial editions, notably those of Sami Dahhan, L[iban,] J[ordanie,] P[alestine: Topographie
historique d'Ibn Šaddad] (Damascus, 1963), and of 'Abbara (Damascus, 1978). The latter text,
which covers the Jazira, is still incomplete, and so it will be necessary on occasions to refer to the
Bodleian Library MS Marsh 333.
5 Ibn al-Jawzi, Mir'at al-zaman fi ta'rikh al-a'yan, facsimile ed. J.R.Jewett (Chicago, 1907), 481–
3; repr. (with different pagination) by Dairatu'1-Ma'aref-il-Osmania Press (Hyderabad, Deccan,
1951–52), Mir'at al-zaman, VIII/2, 727–9 (references will be given in both editions). The erroneous
year 636 found in the text is corrected by later authors, beginning with the redaction of the work
by his pupil al-Yunini (d. 1326): Topkapi Sarayl Müzesi, Istanbul, MS III Ahmet 2907 C, XIII, fo.
375r.
6 Ibn Mufarrij al-kurub fi akhbar bani Ayyub, ed. G.Shayyal et al. (Cairo, 1953-), v, 333–4.
7 All three of these Egyptian writers include scattered details which are derived neither from the
nor from Ibn but in all probability from some thirteenth-century chronicler similarly based in
Egypt. The style and chronological precision of these extracts suggests that the original source may
hane been Ibn Muyassar (d. 1278), of whose work an earlier portion has survived in a MS copied by
al-Maqrizi himself and who was used by al-Nuwayri at least for the twelfth century: see Claude
Cahen, 'Ibn Muyassar', EI, new ed., 111, 894. Other later sources include al-Khazraji (late thirteenth
century), whose work is largely derived from the but contains some additional material; al-
'Izzi al-Khazandari (fl. c. 1330); Ibn al-Dawadari (fl. c. 1334); al-Dhahabi (d. 1348); and Ibn al-Furat
(d. 1405), whose Ta'rikh al-duwal wa'l-muluk is partially available as A[yyubids,] M[amlukes and]
C[rusaders], ed. and tr. U. and M.C.Lyons with historical introduction and notes by J.S.C.Riley-
Smith (Cambridge, 1971). The oft-quoted Abu 'l-Fida (d. 1332) for this period merely abridges Ibn
We are fortunate, finally, in possessing a selection from the correspondence of Da'ud,
[al-]Fawa'id[al-jaliyya fi'l-fara'id ], made by one of his sons, and a biography of the prince in al-
Yunini's continuation (dhayl) of the Mir'at.
On all these Islamic sources, see Cahen, [La] Syrie du Nord à [époque des croisades et la
principauté franque d'Antioche] (Paris, 1940), ch. ii, passim. I am most grateful to the librarians of
the Bibliotheek der Rijksuniversiteit, Leiden, and of the Gotha Forschungsbibliothek for supplying
me with microfilms of the works of al-Nuwayri and respectively, and to Professor Malcolm
Lyons for lending me his microfilm of the Vatican MS of Ibn al-Furat.
8 B[ibliothèque] N[ationale], Paris, MS arabe 302, fos. 155v-223r. See Catalogue des manuscrits

arabes, Ire partie: manuscrits chrétiens, ed. G.Troupeau (Paris, 1972–74), I, 265–6. The author was
formerly believed to be Severus (Sawirus) b.al-Muqaffa', who is now known to have lived in the
tenth century and to have composed only the earlier biographies: G.Graf, Geschichte der christlichen
arabischen Literatur (Vatican City, 1944–53), II, 301–6.
9H[istory of the] P[atriarchs of the] E[gyptian] C[hurch], iv/2, ed. and tr. A.Khater and O.H.E.Khs-
Burmester (Cairo, 1974). E.Renaudot, Historia Patriarcharum Alexandrinorum (Paris, 1713), had
simply abridged this work, and the only author to publish extracts prior to this was Michele Amari,
B[iblioteca] A[rabo-]S[icula] (Leipzig, 1857), 322–6; op. cit. Versione Italiana (Turin and Rome,
1880–89), I, 518–23.
10 Edgar Blochet,' Histoire d'Égypte d'al-Makrizi', R[evue de l']O[rient] L[atin], x (1903–4), 248–
351, passim.
11 R.Stephen Humphreys, From Saladin to the Mongols: the Ayyubids of Damascus, 1193–1260
(New York, 1977), 271–2: the Siyar does not appear to have been used at all. See also Sir Hamilton
Gibb, 'The Aiyubids', in Wolff and Hazard, II, 707–8, who suggests that Ayyub's military activities
were confined to the Yemen.
12HPEC, IV/2, text 114–15, 145, tr. 235–6, 294, for the embassies; text 107, 113, tr. 222, 233–4, for
Frankish disunity.
13 T.C.Van Cleve, 'The crusade of Frederick II', in Wolff and Hazard, n, 461.
14 Joshua Prawer, Histoire du royaume latin de Jérusalem, tr. G.Nahon (2nd ed., Paris, 1975), II,
270; cf. also pp. 225, 265, where the absence of anyone gifted with the emperor's political acumen is
noticed. Jean Richard, The Latin Kingdom of Jerusalem, tr. J.Shirley (Amsterdam, 1979), 323.
15 Painter, 484–5.
16 Prawer, II, 287.
17 Richard, 334. See also Prawer, II, 306.
18 Painter, 484–5.
19H[istoria] D[iplomatica] F[riderici] S[ecundi], ed. J.L.A.Huillard-Bréholles (Paris, 1852–61), III,
87, 108, 137; see also p. 106 for the despatch of envoys to in 1229, and n. 64 infra.
20 This important point is well made—for the first time, so far as I am aware—by K.R.Giles, 'The
Treaty of Jaffa, 18 February 1229: a reassessment', Keele University B.A. dissertation (1982), 55.
21 On these contacts, see Blochet, 'Les relations diplomatiques des Hohenstaufen avec les sultans
d'Égypte', Revue Historique, LXXX, 1902, 51–64. Frederick was expecting an embassy from Cairo
in October 1239: HDFS, v/1, 433. For commercial relations, see W.Heyd, Histoire du commerce du
Levant au moyen-âge, tr. Furcy Raynaud (Leipzig, 1885–6), I, 406–9; Subhi Y.Labib, Handels-
geschichte Ägyptens im Spätmittelalter (1171–1517) (Wiesbaden, 1965), 31, 33. Of the primary
sources reporting subsequent commercial agreements explicitly based on that of 1229, see especially
Ibn 'Abd (d. 1292), Tashrif al-ayyam fi sirat al-malik ed. Murad Kamil
(Cair0; 1961), 156; tr. Amari, BAS Vers. Ital., III, 548–51.
22 For this episode, see René Grousset, Histoire des croisades et du royaume franc de Jérusalem
(Paris, 1934–6), III, 379–83; Prawer, II, 272–4; also Painter, 474–7, who seems, however, to
underestimate the losses suffered at Gaza. The counts of Bar and Montfort were accompanied by 400
knights: 'Annales monasterii de Theokesberia', A[nnales] M[onastici], ed. H.R.Luard (London,
1864–69. Rolls Series), I, 114; cf. also 'Eracles', 414 ('Gestes', 725); 'Rothelin', 539, gives 670. Al-
Nuwayri, Nihayat al-arab fi funun al-adab, Leiden MS Or. 21, 325, and al-Maqrizi, al-Suluk li-
ma'rifat duwal al-muluk, ed. M.M.Ziada, I/2 (Cairo, 1936), 292, tr. R.J.C.Broadhurst, A history of the
Ayyubid sultans of Egypt (Boston, Mass., 1980), 251, give the captives as 80 knights and 250 foot
and the slain as 1,800; HPEC, IV/2, text 96, tr. 197, 15 knights and 500 foot captured and twice as
many slain. The total number of Western knights on Theobald's expedition seems to have
been about 1,000–1,500: see Reinhold Röhricht, G[eschichte des] K[önigreichs] J[erusalem 1099–
1291] (Innsbruck, 1898), p. 839 and n. 2, for references.
23 Philip Mouskès, Chronique rimée, ed. Baron F.A. F.T.de Reiffenberg (Brussels, 1836–8), II, 661
(verses 30,401–4). For the in-fighting among the French magnates during the minority of St. Louis,
see Elie Berger, Histoire de Blanche de Castille reine de France (Paris, 1895), ch. iv, In 1236
Theobald himself had revolted against the Crown and come to an understanding with certain of his
former enemies, such as Peter Mauclerc of Brittany; but they had then left him in the lurch: ibid.,
245–53.
24 Grousset, m, 377–8; hence Richard, 323. Painter, 473. J.S.C.Riley-Smith, [The] Knights of St.
John [in Jerusalem and Cyprus c. 1050–1310] (London, 1967), 176. Prawer, 11, 270–1, and his
'Military orders and crusader politics in the second half of the XIIIth century', in Die Geistlichen
Ritterorden Europas, ed. J.Fleckenstein and M.Hellmann (Sigmaringen, 1980), 221. M.L.Bulst-
Thiele, [Sacrae Domus Militiae Templi Hierosolymitani] Magistri[. Untersuchungen zur Geschichte
des Templerordens 1118/19–1314] (Göttingen, 1974), 199.
25Pamter, 463. Only Hans Eberhard Mayer, The Crusades, tr. J.Gillingham (Oxford, 1972), 248,
hints otherwise. For a military reverse which has been traditionally placed during Theobald's crusade
but which really belongs to 1241, vide infra, p. 49.
26 For his arrival in Cairo, see HPEC, IV/2, text 91, tr. 188:19 Bashans 955 Era of Martyrs/8
Shawwal 636 A.H./14 May 1239. Al-Nuwayri, 324, gives the date of his departure for Kerak as 5
637/6 September 1239, whereas al-Maqrizi, 1/2, 284 (tr. Broadhurst, 245), has
27 For a detailed survey of these events, see Humphreys, 239–56; more briefly in Prawer, II, 263–4.
For the relationships of the various Ayyubid princes, see the genealogical table at p. 34.
28 Ibn v, 215–16, 230.
29 Painter, 473–4.
30 Oliver of Paderborn, Historia Damiatina, ed. H.Hoogeweg, Die Schriften des Kölner
Domscholasters…. Oliverus (Tübingen, 1894), 163–4.
31'Annales Mellicenses. Continuatio Lambacensis', M[onumenta] G[ermaniae] H[istorica.]
Scriptores, ed. G.H.Pertz et al. (Hanoyer etc., 1826–1934), ix, 559, wrongly stating that the Tower of
David was taken: cf. 'Annales Sancti Rudberti Salisburgensis', ibid., 787 ('preter turrim Davit, quam
milites imperatoris defendunt'), and 'Annales prioratus de Dunstaplia', AM III, 150. 'Rothelin', 529–
30, clearly confuses this episode with attack, while placing it immediately prior to the Gaza
campaign. Painter, 472–3 and n. 13, recognized that there were two distinct assaults on the Holy
City; Prawer, II, p. 278, n. 35, reached a similar conclusion, though by dint of misreading 'pseudo-
Yafi'i' (actually the fifteenth-century chronicler al-'Ayni).
32 Ibn v, 215. Ibn al-'Amid, Kitab al-majmu' al-mubarak, ed. Cl.Cahen, 'La "Chronique des
Ayyoubides" d'al-Makin b. al-'Amid', B[ulletin d']É[tudes] O[rientales de l']l[nstitut] F[rançais de]
D[amas], xv, 1955–7, 147. Al-Maqrizi, I/2, 283 (tr. Brpadhurst, 244). The dates coincide almost
exactly, since the news of Ayyub's arrival at Jinin (Gérin) reached Cairo on 20 Shawwal 636/ 26 May
1239: al-Nuwayri, 323.
33'Annales prioratus de Dunstaplia', 150.
34'Rothelin', 531–2. 'Eracles', 414 (hence 'Gestes', 725), mentions only Ascalon; but cf. the
anonymous letter summarized in Matthew Paris, C[hronica] M[ajora], ed. H.R.Luard (London,
1872–83. Rolls Series), IV, 25 ('Damascus non capitur, ut dictum est prius…').
35 Prawer, II, 271–2.
36 Bulst(-Thiele), 'Ritterorden', 204–11; and pace Mary N.Hardwicke, 'The Crusader States, 1192–
1243', in Wolff and Hazard, II, 552. Prawer, 'Military orders and crusader politics', 221, demurs,
however.
37For these events, see Humphreys, 257–61. The date of Ayyub's capture is given by Ibn Khallikan,
as quoted in Röhricht, GKJ, 846 and n. 4, and in W.B.Stevenson, The Crusaders in the East
(Cambridge, 1907, repr. Beirut, 1969), p. 316, n. 5.
38 Ibn v, 220. Al-Nuwayri, 331. Ibn al-'Adim (d. 1262), Zubdat min ta'rikh ed.
Sami Dahhan (Damascus, 1951–68), 111, 246; tr. E.Blochet, 'L'histoire d'Alep, de Kamal-ad-dîn',
ROL, v, 1897, 107. Humphreys, 255, 257–8. Prawer, II, 264 (though at p. 271 he assumes that al-
'Adil and Isma'il were rivals).
39 v, 238.
40 HPEC, IV/2, text 95, tr. 195. Ibn v, 267. For the date of his departure for Gaza, around the
time of al-'Adil's return to Cairo on 17 Rabi'I 637/17 October 1239, see al-Maqrizi, 1/2, 289 (tr.
Broadhurst, 250).
41Pace Richard, 323, who states that the Syrian barons 'wanted to go and attack Egypt on her own
ground'.
42'Rothelin', 541; and see also 540, where quite clearly a mere raid is in question ('il iroient jusques a
Gadrez et lendemain revendroient en l'ost a Escalonne').
43 Rumours which reached both the West and Egypt accused the local Franks of abandoning their
Western confrères: 'Rothelin', 549; HPEC, IV/2, text 96, tr. 197 (cf. also Prawer, II, p. 275, n. 29).
But we know that the duke of Burgundy was among the deserters:' Rothelin', 543–4; CM, IV, 25. For
a list of local barons who avoided battle, see 'Gestes', 726.
44 For the dates, see Ibn Shaddad, LJP, 225; hence Ibn al-Furat, AMC, 1(text), 76, II, (tr.), 62 (and n.
3 at p. 203). Richard, 323, incorrectly places campaign in September 1239.
45 Al-Nuwayri, 325. Al-Maqrizi, 1/2, 291 (tr. Broadhurst, 251). Ibn v, 259, indicates that
had the Friday prayers read in al-'Adil's name right down until April 1240. Prawer, II, 278, is
therefore surely wrong to suggest that his seizure of the Holy City was an act of defiance towards the
sultan.
46 'Rothelin', 550. The reconstruction of the ruined Templar fortress at Safed was contemplated, but
not actually begun, at this time: De constructione castri Saphet. Construction et fonctions d'un
château fort franc en Terre Sainte, ed. R.C.B.Huygens (Amsterdam, 1981), 34–5.
47'Eracles', 415–16 (hence 'Gestes', 726–7).
48 Armand of Pierregort to Walter of Avesnes, in Alberic of Trois-Fontaines,' Chronica', MGH
Scriptores, XXIII, 945: the tense shows that the crusade had not yet arrived ('in iocundo cruce
signatorum adventu…subiciet'). Röhricht, 'Kreuzzüge', 99, dated this letter in the winter of 1238–39,
which is probably too early; Bulst-Thiele, Magistri, 210, in the spring of 1240, which is certainly too
late. For the form of the Master's cognomen, see ibid., 189.
49 Ibn v, 222–8, 239; hence al-Maqrizi, I/2, 285–7 (tr. Broadhurst, 247–8). This episode is
briefly and inaccurately summarized by Humphreys, 256–7, and more exactly by Emmanuel Sivan,
L'Islam et la croisade (Paris, 1968), 153.
50 'Rothelin', 531. 'Eracles', 416 ('Gestes', 727). The contradiction was noticed by Röhricht, GKJ, p.
845, n. 2.
51 B.N.MS lat. 5479, p. 136. The printed version of this text unaccountably reads 'Moiascon.' for
'Matiscon.': 'Obituaire de l'abbaye de Saint-Yved de Braine', ed. Emile Brouette, Analecta
Premonstratensia, xxxiv, 1958, 319. The correct reading had been given long ago by André Du
Chesne, Histoire généalogique de la maison royale de Dreux (Paris, 1631), preuves, 258.
52 Alberic, 946. For these obits, see Obituaires de la province de Sens, ed. A.Molinier et al. (Paris,
1902–23), I/1, 511 (2 Oct.), and III, 107 (5 Oct.), for Robert; IV, 424 (1 Oct.), for Anselm.
53'Eracles', 416; hence 'Gestes', 727. For the siege of by Aleppan troops in 635–6 A.H./ 1238–
9, see Ibn al-'Adim, Zubda, III, 238, 244 (tr. Blochet, in ROL, v, 100–101, 104–5); Ibn v, 182,
198.
54 Ibn v, 257 soon after the death of al-Mujahid of which occurred on 20 Rajab
637 A.H./15 February 1240 484/732).
55 Ibn v, 252. Humphreys, 262.
56 Humphherys, 262–3. The date of Ayyub's release is supplied, again, by Ibn Khallikan: Röhricht,
GKJ, 847 and n. 2; Stevenson, p. 316, n. 5.
57 482/728; for al-Mujahid's death, vide supra, n. 54. Ibn Wasil, v, 259–60.
58 The suggestion in 'Eracles', 416, 419 ('Gestes', 727), and in 'Rothelin', 552.
59HPEC, IV/2, text 104 (tr. 214), speaking of Ayyub's movements following his release by al- but
prior to al-'Adil's deposition: wa-min wara'ihi'l-Ifranj wa-ma'ahum Dimashq ('and to his rear were
the Franks and with them the ruler of Damascus'; my translation).
60 Armand of Pierregort to Robert de Sandford, preceptor of the Temple in England, in CM, IV, 64
('cum jacuisset [sc. Christianus exercitus] diu in sabulo'). Theobald appears to have been in Acre in
April and May 1240: H.d'Arbois de Jubainville, Histoire des ducs et des comtes de Champagne
(Paris, 1859–69), IV, 315–16, note b, 321, note b.Al-Khazraji names Isma'il's envoy to the Franks as
Jamal al-Din al-Rumi: Ta'rikh dawlat al-Akrad wa'l-Atrak, Süleymaniye Kütüphanesi, Istanbul, MS
Hekimoglu Ali 695, fo. 150v.
61 Early in 638 A.H. (began 23 July 1240): Ibn Shaddad, LJP, 134, 155, 159; 485/732, also implies
the outset of the year. CM, IV, 65, says that the messenger bearing the news of the treaty passed
Richard of Cornwall on his way out to Syria, thus suggesting a date in July–August for the
agreement.
VOL. L. PART 1. 4
62 The only source to list all these operations is HPEC, IV/2, text 105, tr. 217. For Ascalon, see also
'Rothelin', 553, and cf. Ibn Shaddad, LJP, 262. For Jerusalem, see 'Rothelin', 554. For Nablus, see al-
Nuwayri, 332, who additionally supplies the date of return from Cairo, 637 A.H.; al-
Maqrizi, 1/2,299 (tr. Broadhurst, 258); cf. also Ibn Shaddad, as quoted n. 72 infra.
63 Painter, 479. Prawer, II, 279 and n. 38. Humphreys, n. 49 at p. 457. See also Stevenson, p. 318, n.
1.
64 For Toron in the 1229 treaty, see Richard, 234, and Prawer, II, 199; for other fortresses in the
north, see Chronique d'Ernoul et de Bernard Le Trésorier, ed. Comte L. de Mas Latrie (Paris, 1871),
464, and 'Eracles', 375, variant readings. Ibn Shaddad's account of Toron makes no mention of 1229
but says expressly that Isma'i handed it over to 'Sir Filit', i.e. Philip of Montfort, in 638 A.H./
1240(-1): LJP, 153 (omitted by Ibn al-Furat, AMC, I, 123; H, 97).
65 The fullest list is to be gleaned from Ibn Shaddad, LJP, 100 (Sidon), 134 (Tiberias), 147 (Safed),
153 (Toron and Châteauneuf, 155 (Beaufort), 159 (Shaqif Tirun/' Cavea de Tyron'); hence Ibn al-
Furat, AMC, I, 55–6, 81, 112, 123, 138; II, 46, 66, 88, 97, 109. For a slightly different list, see Ibn al-
'Amid, 153: Safed, Beaufort, Tiberias, the Jabal'Amila (northern Galilee) and half of Sidon. He is
followed by al-Nuwayri, 334; by Ibn Duqmaq, Nuzhat al-anam fi ta'rikh al-Islam, B.N.MS arabe
1597, fo. 49r; and by al-Maqrizi, I/2, 303 (Broadhurst, 262, wrongly applies the 'half' to Tiberias as
well). Ibn v, 302 (sub anno 639 A.H.), and Abu Shama (d. 1268), al-Dhayl 'ala'l- ed. M.Z. al-
Kawthari, Tarajim rijal al-qarnayn al-sadis wa'l-sabi' (Cairo, 1947), 170, name only Beaufort and
Safed, as does the 485/732, 493/745 (sub anno 642 A.H.: vide infra, n. 70). 'Rothelin', 552,
mentions Beaufort alone.
66'Eracles', 418; hence 'Gestes', 727. Armand of Pierregort, in CM, IV, 65; see also Paris's H[istoria]
A[nglorum], ed. Sir Frederick Madden (London, 1866–9. Rolls Series), II, 440–1; Annals of
Southwark, Bodleian Library MS Rawlinson B 177, fo. 224r. 'Annales de Terre Sainte', ed.
R.Röhricht and G.Raynaud, Archives de l'Orient Latin, II, 1884, documents, 440, version B ('et toute
la terre de Jerusalem'), although both recensions also specify Safed and Beaufort. See finally
'Chronicon S.Medardi Suessionensis', Spicilegium sive collectio veterum aliquot scriptorum, new ed.,
Et.Baluze et al. (Paris, 1723), II, 491 ('omnis terra quam Christiani tenebant tempore perditionis',
except Kerak and Montreal; and see next note).
67 Ibn Shaddad, LJP, 225,234 (Jerusalem), 246 (Nablus), 265 (Gaza, which had been restored to him
by al-'Adil after the Egyptian victory in Nov. 1239). Nablus, Gaza and Jericho (in Jordan valley
territory) are all specified in 'Chronicon S.Medardi', loc. cit.
68'Eracles', 418; 'Gestes', 727. CM, IV, 65; HA, II, 441; Annals of Southwark, fo. 224r.
69 Al-Nuwayri, 334; cf. al-Subki (d. 1370), al-shafi'iyyat al-kubra, ed. al-Qadiri 1906), v,
100(biography of al-Sulami), here following either al-Nuwayri or a common source. Ibn Duqmaq, fo.
49r. Al-Maqrizi, 1/2, 304 (tr. Broadhurst, 262–3).
70 Humphreys, 266–7. Sivan, 150–1. Prawer, II, 280. For Beaufort, see 'Rothelin', 552–3. 493/745,
refers to this episode in the context of the later alliance of 642 A.H./1244, and is followed by Ibn al-
Furat, Vatican MS ar. 726, fo. 41v. But cf. Ibn Shaddad, LJP, 155–6; al-Khazraji, fos. 150v-1r; Ibn
Duqmaq, fo. 48v, with the date Rabi' I 638 A.H (began 20 September 1240) for the arrival of the
news in Egypt.
71'Eracles', 419 ('Gestes', 727). For the data in the Islamic sources which have been taken to apply to
this supposed campaign, vide infra, nn. 75, 123.
72Ibn Shaddad, LJP, 246, says that Isma'il seized Nablus during absence in Egypt; cf. n. 62 supra.
73HPEC, IV/2, text 88, tr. 182 (1239); text 96, tr. 198 (early 1240).
74 On these campaigns, see Humphreys, 269–71.
75vide infra, p. 49. It was in this campaign that the Franks were let down by their Syrian allies.
Secondary authorities placing it in 1240 include Röhricht, 'Kreuzzüge', 80, and GKJ, 848; Grousset,
III, 389; and Prawer, II, 281.
76HPEC, IV/2, text 105, tr. 217.
77 Ibn al-'Adim, Zubda, III, 253 (tr. Blochet, in ROL, VI, 1898, 6); hence Ibn v, 286.
78CM, IV, 79. The 'Rooch' with whom he is alleged to have made the truce can only be Rukn al-Din
al-Hayjawi, the Egyptian general and victor of Gaza; but he was then in temporary eclipse, having
been arrested by Ayyub early in June 1240 and sent to Cairo: al-Maqrizi, I/2, 299 (tr. Broadhurst,
259). In any case, Isma'il's truce was made 'cum quodam potente sibi consanguineo', which hardly
fits al-Hayjawi. Röhricht, GKJ, p. 840, n. 3, was therefore right to see here.
79'Eracles', 419('Gestes', 727).
80'Rothelin', 552.
81 Richard of Cornwall to Baldwin de Redvers, earl of Devon, et al., in CM, IV, 140. There is a
further echo of the truce in Alberic, 949, though he seems to confuse with Isma'il: 'Treuge…dicuntur
esse ad soldanum de Damasco seu Nascere…secundum conpositionem regis Navarre…'. It is
noteworthy, however, that he has referred to as sultan of Damascus on a previous occasion (p. 948).
82F[lores] H[istoriarum], ed. H.R.Luard (London, 1890. Rolls Series), II, 242–3. Cf. Röhricht,
'Kreuzzüge', pp. 81, n. 7, 85, and GKJ, p. 849, n. 6; Bulst(-Thiele), 'Ritterorden', 203, and Magistri,
p. 202, n. 66. Prawer, H, 279, is unclear on stance, owing to a confusion with the campaign
of the summer of 1241 (vide infra, p. 48).
83HPEC, IV/2, text 107 (tr. 221): 'ala'ibqa'i'l-biladi'llati bin al- iyyaha bi-aydihim.
84 Ibn Shaddad, LJP, 234. Ibn al-Dawadari, Kanz al-durar wa-jami' al-ghurar, ed. S. 'Ashur, Die
Chronik des Ibn ad-Dawadari, VII, Der Bericht über die Ayyubiden (Cairo, 1972), 344–5: the month
Rabi' II 638 A.H. (began 20 Oct. 1240) is surely too late, and al-akhir is probably an error for al-
awwal, as frequently happens. Rabi' I began 20 September. Ibn v, 278, and al-Maqrizi, I/2, 302 (tr.
Broadhurst, 261), report alliance with Isma'il and but omit the Franks.
85 Humphreys, 265.
86IbnShaddad, LJP, 156.
87'Rothelin', 554–5. 'Eracles', 419 ('Gestes', 727). On this complex question, see Stevenson, p. 319,
n.2.
88HPEC, IV/2, text 107 (tr. 221): rusul al-Ifranj taraddadat ila This was prior to the despatch of
Kamal al-Din to the Franks, on which vide infra, p. 46. But it should be noted that the last event
mentioned (tr. 220) is the establishment of the Palace of Justice, which is dated by al-Maqrizi, I/2,
306–7 (tr. Broadhurst, 265), in Rabi' II/October-November 1240.
89CM, IV, 140 ('ut aliquid fecisse viderentur'); cf. also FH, II, 243.
90 Ibn Shaddad, LJP, 246.
91 Ibn al-Dawadari, VII, 345: yakund'l-Quds baynahum Armand of Pierregort dated the
exclusively Christian occupation of Jerusalem from 1243: CM, IV, 290. Ibn Shaddad, LJP, 234, is
therefore apparently in error in dating the expulsion of the Muslim inhabitants in 638 A.H./1240.
92HDFS, III, 88–9, 108, 148. Richard, 234–5. Prawer, II, 199, 201.
93CM, IV, 140 ('licet parum attineret').
94vide supra, pp. 36, 42 and n. 64.
95 'Rothelin', 553, erroneously including the Hospitallers among those who urged an invasion.
96 Ibn v, 263–4.
97CM, IV, 78–9; HA, II, 443; FH, II, 242.
98CM, IV, 140–1.
99 See Sivan, 140–1.
100 CM, IV, 141 ('quidam magnus potens valde ex parte Soldani Babiloniae'); cf. also FH, II, 452.
101HPEC, IV/2, text 107, tr. 221. On Kamal al-Din and his family, see Hans L.Gottschalk, 'Awlad al-
Shaykh', EI (new ed.), I, 765–6; more fully, 'Die Aulad (Banu Wiener Zeitschrift für die Kunde des
Morgenlandes, LIII, 1956, 57–87.
102 The regent apparently joined the coalition in return for further Damascene aid against the
Khwarizmians in Jumada II 638 A.H. (began 18 Dec. 1240): Ibn al-'Adim, Zubda, III, 254 (tr.
Blochet, in ROL, VI, 7); Ibn v, 288; cf. also Gibb, 'The Aiyubids', p. 708, n. 17. Ibn v, 300, indicates
that Aleppo had joined by the beginning of 639 A.H.For Isma'il's earlier attempts, see Ibn al-'Adim,
m, 247–8 (tr. Blochet, in ROL, VI, 1–2); hence Ibn v, 268–9.
103CM, IV, 47. 'Annales prioratus de Dunstaplia', 152, where we read that the earl lingered in French
territory 'donec imperatori consuleret'.
104HDFS, VI/1, 239: 'R[icardo] comite Cornubie…in ultramarinis partibus vices agente nostras…'.
105'Eracles', 421 ('Gestes', 728): 'il ne vost faire ne l'un ne l'autre.' This may possibly be the
significance of the statement by Gervase of Canterbury's continuator that the earl reconciled the
Temple and the Hospital: The historical works of Gervase of Canterbury, ed. W.Stubbs (London,
1879–80. Rolls Series), II, 179.
106CM, IV, 141; and for the date of his despatching envoys to Cairo, see p. 143. 'Rothelin', 556, is
closer to the truth than the Eracles in recognizing Richard's early commitment to peace with Ayyub,
though the question is complicated by the assumption of both chronicles that such a peace had
already been effected by Theobald (supra, pp. 44–5).
107CM, IV, 141–3.
108 e.g., 'Bersamul'=Nebi Samwil, a few miles north of Jerusalem, on which see Guy Le Strange,
Palestine under the Moslems (London, 1890), 433; 'Kocabi'=Deir el-kobebe, S.W. of Beth Gibelin,
on which see Gustav Beyer,' Die Kreuzfahrergebiete von Jerusalem und S.Abraham (Hebron)',
Zeitschrift des Deutschen Palästina-Vereins, LXV, 1942, 184.
109HPEC, IV/2, text 107, tr. 221–2.
110 Richard, 326.
111HPEC, IV/2, text 110, tr. 227. For this campaign, vide infra, p. 48. Stevenson, p. 320, n. 2,
suggested that Ayyub still disposed of the revenues of Gaza at this time; but cf. Ibn Shaddad, LJP,
265.
112'Rothelin', 556; this is not mentioned, however, in Richard's letter. John of Columna alleges that
one of the clauses in the treaty secured safe-conduct to Jerusalem for the crusaders: 'E Mari
Historiarum', Recueil des Historiens des Gaules et de la France (new ed. L.Delisle, Paris, 1869–
1904), XXIII, 110.
113 487/736: it was from here that he despatched al-Jawwad against the Egyptian army (vide infra, p.
50). The history of Jerusalem during the previous months is confused. Prawer, II, 278, believes that
Jerusalem had been reoccupied by the Egyptians following its capture by in the winter of 1239–40;
cf. also p. 282. But Ibn v, 259, shows that it was still in hands in April 1240; see also Humphreys,
263. Stevenson, 320, ignores assuming that all the places listed were in Isma'il's possession.
114 Richard, 325. Sir Steven Runciman, A history of the Crusades (Cambridge, 1951–54), III, 218.
Bulst-Thiele, Magistri, 202. Painter, p. 479, n. 20, assumed that the Damascene and Egyptian treaties
conveyed identical territories.
115HPEC, IV/2, text 107, tr. 222.
116 Armand of Pierregort to Robert de Sandford, in CM, IV, 289. For the date of this letter, vide
infra, n. 168.
117 The fullest account is in HPEC, IV/2, text 110–11, tr. 227–30, where the battle is located at Ra's
al-'aqaba, on the road from Jerusalem to Bayt Nuba. Only al-Maqrizi, I/2, 305 (tr. Broadhurst, 264),
supplies the precise date of the campaign, Dhu'l-Qa'da 638 A.H. (began 14 May 1241). 487/736–7, is
briefer, locating the battle at Bayt Furayk, east of Nablus; cf. also p. 489/739. Both he and Ibn
v, 300–1 (followed by Ibn al-Dawadari, VII, 347, and Ibn al-Furat, Vatican MS, fos. 2v–3v),
incorrectly date it 639 A.H. (began 12 July). At a later juncture al-Maqrizi, I/2, 309 (tr. Broadhurst,
267), gives a second account, derived from Ibn and so sub anno 639. On the basis of these confused
data Stevenson, p. 321, n. 1, was misled into identifying the campaign with that of 1242 (vide infra,
p. 51).
118LJP, 148; hence Ibn al-Furat, AMC, I, 112–13, II, 88–9. Since and Isma'il must have been allies
at this time, the incident may safely be dated between December 1240, when the first stone was laid
at Safed (Richard, 326), and the beginning of the summer, when made peace with Ayyub: he did not
ally with Isma'il again until the late spring of 1243 (infra, p. 53).
119HPEC, IV/2, text 112 (tr. 231 incorret): fa-tawaffaqat illa an ma'a'l-malik
120 Ibn al-'Amid, p. 153, ll 9–11; though the latter section of the passage belongs not to 1241 but to
the 1243 campaign (vide infra, n. 156). Al-Maqrizi, I/2, 303 (tr. Broadhurst, 262). For the fullest
account of Ayyub's dealings with his grandees, see Ibn v, 274–6; cf. also Humphreys, 264, 268.
121HPEC, IV/2, text 111, tr. 230, omitting the force from Aleppo and mentioning the Nablus clash.
Ibn al-'Amid, p. 153, ll. 11–15, followed by Ibn Duqmaq, fo. 46v, and al-Maqrizi, 1/2, 304 (tr.
Broadhurst, 263), includes the Aleppan contingent and describes the engagement at alone; cf.
preceding note.
122 Al-Nuwayri, 341. Al-Subki, v, 101; briefer version in Ibn al-'Asqalani (d. 1449), Raf'
ed. 'Abd al-Majid et al., revised by Ibrahim al-Abyari (Cairo, 1957–61), II,
351.
123 Ibn Duqmaq, fo. 49v. Al-Maqrizi, I/2, 305 (tr. Broadhurst, 264). Stevenson, p. 320, n. 3,
expressed doubts as to the reliability of this account, as does Bulst-Thiele, Magistri, p. 204, n. 71, but
they assume that these events belong to 1240. Both Ibn Duqmaq and al-Maqrizi, however, specify
that the Frankish prisoners taken in the encounter were employed by Ayyub on his new colleges'
between the two palaces' (madaris ), on which work began only in 639 A.H.: Ibn al-Dawadari, VII,
347; Ibn Duqmaq, fo. 53r; al-Maqrizi, I/2, 308 (tr. Broadhurst, 266); cf. also HPEC, IV/2, text 119,
tr. 246 (near the end of 639/late spring 1242).
124HPEC, IV/2, text 112, tr. 231. The later reference is at text 113 (tr. 234 slightly misleading):
lamma ra'u khidhlanahum 'alayhim 'when they beheld their abandonment and the victory of the ruler
of Egypt over them'.
125Al-Maqrizi, I/2, 309 (tr. Broadhurst, 267–8), alone supplies the date, 12 Al-Kutubi, 639 A.H.,
along with the other details. HPEC, IV/2, text 113, tr. 233, says merely that defeated the Syrian
forces; the news seems to have reached Egypt at the very beginning of 957 E.M./September 1241.
126HPEC, IV/2, text 113 (tr. 233–4 again confused): wa-hum rusul al-Diwiyya wa-illa'l-mugharrab
[?] 'Asqalan wa-ghayrahum min akbar al-Faranj kanu, [so MS arabe 302—printed text has kunu in
error] ma'a wa-ha'ula'i' lladhina kanu ma'a Dimashq 'they were the envoys of the
Templars—not of the Westerners who controlled Ascalon and of the others among the Frankish
grandees who were at peace with our master the Sultan—and it was these who were allied with the
ruler of Damascus'.
127CM, IV, 167; ibid., 144, for Richard's departure. The crusading leaders who left at this time
included the duke of Burgundy and the count of Nevers: for a list, see P.Jackson, 'The end of
Hohenstaufen rule in Syria', Bulletin of the Institute of Historical Research, LIX, 1986, 32–33.
128 'Eracles', 422. The passage may nevertheless contain a blurred reference to Isma'il's expedition:
'Et ous que li oz des Crestiens aloit, li sodans di Domas o tout son ost estoit toz jorz herbergez pres
d'eaus.' Cf. also 'Gestes', 728; and for the probable conflation of the events of 124–41 in the Frankish
tradition, supra, p. 43.
129vide supra, p. 37. The chequered career of this prince is here tentatively reconstructed.
Humphreys, 271–2 and n. 58 at p. 458, abandons the attempt to make sense of the conflicting details
in the sources.
130HPEC, IV/2, text 106, 110, tr. 219–20, 227. Ibn al-Amid, p. 152, ll. 10–18. Ibn v, 281–2, 296–7.
487/736. For al-Jawwad's career in the Jazira, see Ibn Shaddad, ed. 'Abbara, 203–6; summary in
Cahen, 'La "Djazîra" au milieu du treizième siècle, d'après 'Izz ad-din Ibn Chaddad', Revue des
Études Islamiques, VII, 1934, 118–9.
131HPEC, IV/2, text 110–11, tr. 228–30. 487/736–7. Ibn al-'Amid, p. 152, ll. 18–21 (hence Ibn
Duqmaq, fo. 45v), omitting all mention of the battle. Al-Maqrizi, 1/2, 303 (tr. Broadhurst, 261),
furnishes the date of al-Jawwad's arrest, 18 638 A.H./30 June 1241; more details, greatly confused,
at p. 305 (tr. Broadhurst, 264).
132 Ibn al-'Amid, p. 152, ll. 21–4 (Ibn Duqmaq, fo. 45v). HPEC, IV/2, text 111, tr. 230, includes him
among the princes who accompanied Isma'il on his Egyptian campaign; see text p. 115, tr. p. 238, for
his joining the Franks around October (last datable events mentioned are a solar eclipse on 6 October
1241 and the arrival of Frederick's envoys in Egypt for the winter)
133 Ibn al-'Amid, p. 152, ll. 24–5: wa-kana yaqulu inna'l-Faranj ikhwatun lahu li-anna ummahu
kanat Faranjiyyatun wa-lihadha malu ilayhi maylan kathiran. Al-Dhahabi confirms that his mother
was a Frank: Ta'rikh al-Islam, Süleymaniye Kütüphanesi MS Ayasofya 3013, fo. 12r.
134HPEC, IV/2, text 115, tr. 238.
135 Ibn Duqmaq, fos. 45v-46r. Ibn al-'Amid, p. 152, 1. 24, says only that the Franks and al-Jawwad
encamped at Caesarea (Qaysariyya), and it seems that Ibn Duqmaq has preserved a fuller version
copied from their common source. Neither mentions the raid on Gaza, for which see HPEC, IV/2,
text 115, tr. 238.
136 'Gestes', 729–30.
137 Ibn al-'Amid, p. 152, ll. 25–8. Ibn Duqmaq, fo. 46r, gives a slightly different account and does
not mention that a treaty was actually sworn. The proposal to restore Damascus to al-Jawwad may be
inferred from HPEC, IV/2, text 117, tr. 242.
138Al-Maqrizi, I/2, 310 (tr. Broadhurst, 268), who alone gives the date, 639 A.H.(began 5 March

1242). Ibn al-'Amid, p. 152, ll. 28–9. HPEC, IV/2, text 117, tr. 241–2.
139 Ibn Duqmaq, fos. 46r-v, provides the fullest account: the parallel passage in Ibn al-'Amid, 152–3,
omits Ayyub's secret instructions to al-Jawwad and consequently presents a non sequitur. HPEC,
IV/2, text 118, tr. 242–3, gives no reason for the fears of the two men. Al-Maqrizi, I/2, 310 (tr.
Broadhurst, 268), furnishes the date of al-Hayjawis flight, 15 Dhu'l-Qa'da 639 A.H./17 May 1242,
but erroneously makes him leave from Cairo rather than Gaza.
140CM, IV, 197: 'Templarii…plus miraculose quam humana fortitudine inopinabili victoria gloriose
triumpharunt.' This comment has been taken to refer to the sack of Nablus and its aftermath (vide
infra, p. 52), e.g. by Röhricht, GKJ, p. 854 and n. 4; Grousset, m, p. 397, n. 1, and Runciman, m, p.
220, n. 1. But Paris introduces it in the context of the late spring or early summer of 1242. Stevenson,
p. 321, n. 1, rightly connected it with events in May, but on the basis of confused data in al-Maqrizi
which really apply to the battle of May 1241 (supra, n. 117). The vague reference in Philip Mouskès,
Chronique rimée, ed. Baron F.A. F.T. de Reiffenberg (Brussels, 1836–38), II, 683, can probably also
be linked with these events.
141HPEC, IV/2, text 118, tr. 243.
142ibid., text 131, tr. 268–9. Al-Maqrizi, 1/2, 310–11 (tr. Broadhurst, 269), is briefer but supplies the
date, 4 Jumada I 640 A.H. There is another account in the commentary on correspondence: Fawa'id,
Süleymaniye Kütüphanesi MS Ayasofya 4823, fos. 46v–47v; summarized in the obituary of
in al-Yunini, al-Dhayl 'ala Mir'at al-zaman, ed. Dairatu'l-Ma'aref-il-Osmania Press (Hyderabad,
Deccan, 1954–61), I, 157. The raid on Qalansuwwa, a dependency of Nablus, mentioned by the
492/743, may have been part of this campaign; if so, al-Jawwad was present.
143HPEC, IV/2, text 131–2, tr. 269. It seems that a garbled account of this episode is preserved in
'Annales de Terre Sainte', A, 440, and B, 440–1, though both versions include on the Franks' side at
a time when he is known to have been bitterly hostile to them: this misled Gibb,' The Aiyubids', 709.
But the correct wording has been retained in the Castilian version, ed. A.Sánchez Candeira, 'Las
cruzadas en la historiografia española de la época. Traducción castellana de una redacción
desconocida de los "Anales de Tierra Santa"', Hispania, xx, 1960, 358 (I am indebted to Dr. Peter
Edbury for bringing this article to my attention): 'fueron los Templeros…e Malech Joet a Escalon, e
Le Naser e la hueste de Babillonna asalioron la casa del Temple…'. Bulst(-Thiele), 'Ritterorden', 213,
and Magistri, p. 204, n. 71, was rightly suspicious of the Old French recensions at this point, but her
conclusion that the 'Annales' muddle the events of different years is groundless.
144Fawa'id, fos. 47v–49r; summary in al-Yunini, I, 157–9. See Sivan, 140. For al-Sulami's arrival in
Egypt and appointment first as (10 Rabi' II 639 A.H./18 October 1241) and then as 1242), see
al-Nuwayri, 341–3; al-Maqrizi, 1/2, 308 (tr. Broadhurst, 266–7).
145 So according to HPEC, IV/2, text 111, tr. 229:300 under al-Jawwad as against 3,000 Egyptian
troops; cf. also text p. 110, tr. pp. 227–8. But al-Khazraji, fo. 151 v, gives 700 and 2,000 respectively.
146HPEC, IV/2, text 115, tr. 238: its abandonment is followed immediately by the news of the raid
on Gaza (supra, p. 50 and n. 134).
147488/738, indicating a date very early in 640 A.H. (began 1 July 1242). HPEC, IV/2, text 120, tr.
247–8, speaks of preparations at this juncture for an expedition into Syria, but for the purpose of
meeting an envoy of the Caliph. Cf., however, al-Dhahabi, as quoted in next note.
148 Al-Dhahabi, Ta'rikh al-Islam, Süleymaniye Kütüphanesi MS Ayasofya 3012, fos. 215v, 254v; cf.
also his Duwal al-Islam, tr. Arlette Nègre (Damascus, 1979), 244. For the date of Kamal al-Din's
death, 13 640 A.H., see Abu Shama, 172; 489/739, gives simply the month. Ibn v, 301,
followed by Ibn al-Furat, Vatican MS, fos. 3r–v, states misleadingly that he had died shortly (bi-qalil)
after his release by (supra, p. 48). He is possibly the source of al-Maqrizi's incorrect date 639
A.H. in wa'l-i'tibar bi-dhikr wa'l-athar (Bulaq, 1853–4), II, 34.
149 T'arikh Bayrut, ed. F.Hours and K.S.Salibi (Beirut, 1969), 49–50. The letter is dated 6
in an unspecified year: it can belong only to the period 637–9 A.H., however, since the recipient,
Najm al-Din was killed on 6 Rabi II 640 A.H./3 October 1242, and of the three years 639 (when the
corresponding Julian date was 8 June 1242) is the most likely. See further Salibi, Maronite historians
of mediaeval Lebanon (Beirut, 1959), 205–6.
150 Armand of Pierregort to Robert de Sandford, in CM, IV, 289 (for the date of this letter, vide infra,
n. 168): Ayyub opened negotiations 'post mala quae a nobis recepit [sc. Soldanus] et Nasserus', i.e.
following the sack of Nablus and the abortive siege of Jaffa.
151HPEC, IV/2, text 117, tr. 242. Al-Maqrizi, I/2, 310, 312–13 (tr. Broadhurst, 268, 270–1), and The
pearl-strings: a history of the Resúliyy Dynasty of Yemen, tr. J.W.Redhouse and ed. E.G. Browne et
al. (Leiden and London, 1906–8), I, 109, both give 639 A.H.
152HPEC, IV/2, text 141–2, tr. 288.
153 ibid., text 142, tr. 289: the date of Isma'il's advance may be inferred from the events that
immediately follow, dated Ba'una 959 E.M./June 1243.
154ibid., text 142, 145, tr. 289, 294–5. Ibn v, 323, lists among Ayyub's enemies at the outset of 641
A.H. (began 21 June 1243), but makes no mention of these campaigns.
155 'Annales de Terre Sainte', B, 441: '…et li sires de Damas vint as Moulins de Turs pour affremer
la triue o nos gens; mais il les engana et n'en fist point'; Sánchez Candeira, 358–9. Bulst,
'Ritterorden', p. 213, n. 51, assumes that this entry has been misplaced from 1240 or 1241. For 'Les
Moulins des Turs', on the lower 'Awja river, see Gustav Beyer, 'Die Kreuzfahrergebiete
Südwestpalästinas', Beiträge zur biblischen Landes- und Altertumskunde, LXVIII, 1946–51, 179–80
(and map, p. 188).
156 Ibn al-'Amid, p. 153, ll. 15–20; Ibn Duqmaq, fo. 46v. Both authors record these events sub anno
638 A.H./124(0-)1, although the former adds that some of them occurred after that date. Cf. supra, n.
120, for the probable conflation of two distinct campaigns. The auxiliaries from Aleppo must surely
belong to the 1241 expedition, since in June 1243 Aleppo was too absorbed with the
Mongols to concern herself with Egypt and had just sent a force to assist the Seljük Sultan against
them: Ibn al-'Adim, Zubda, III, 268 (tr. Blochet, in ROL, vi, 19); Ibn v, 314. HPEC, IV/2, text 145,
tr. 295, reports the news of al-Jawwad's arrest and of the retreat of Isma'il and towards Damascus:
MS arabe 302 breaks off at this juncture.
157 Al-Khazraji, fo. 152r. Ibn v, 297. 492/743–4, with the date, Shawwal 641 A.H. Ibn al-'Amid, p.
153, ll. 20–21; Ibn Duqmaq, fo. 46v. For al-Jawwad's epitaph in the at Damascus, see
Repertoire chronologique d'épigraphie arabe, XI, ed. J.Sauvaget et al. (Cairo, 1941–42), 117 (no.
4176).
158ROL, x, p. 339, n. 3.
159 492/743–4.
160Bodleian Library MS Pococke 324, fo. 138r: wa-baddala malan lil-Faranj wa-tasallama'l-
Jawwad minhum. This MS has been identified—though not with total certainty—with an
abridgement of the Zubdat al-fikra of Baybars (d. 1325).
161 492/743. Ibn al-'Amid, p. 153, 1. 21.
162 Ibn al-'Adim, Zubda, m, 267–8 (tr. Blochet, in ROL, vi, 19): between Jumada II and Shawwal
640 A.H./December 1242 and April 1243. Cf. also Ibn v, 314.
163 See Cl. Cahen, Pre-Ottoman Turkey, tr. J.Jones-Williams (London, 1968), 137–8. Gibb, 'The
Aiyubids', 708.
164 Al-Maqrizi, I/2, 308 (tr. Broadhurst, 266): 11 Dhu'l-Qa'da 638 A.H./24 May 1241.
165 Abu Shama, 173:24 Rabi' I 641 A.H.Al-Maqrizi, 1/2, 310 (tr. Broadhurst, 269), erroneously
places this event under Rabi' I 640.
166 Al-Dhahabi alone names as commander of the Damascene army at the siege: Ta'rikh al-Islam,
MS Ayasofya 3012, fo. 254v, margin; cf. also Duwal al-Islam, tr. Nègre, 246. Ibn v, 328, 331, refers
to the siege only briefly.
167 For all these events, see Humphreys, 272–4; Prawer, II, 307. The main source is Ibn v, 327–32.
168 Armand of Pierregort to Robert de Sandford, in CM, IV, 289–90. Pace Röhricht,' Kreuzzüge',
100, and GKJ, p. 860 and n. 1, and Prawer, II, p. 307, n. 41, this letter appears to have reached
England in the first months of 1244; it states, moreover, that Jerusalem has not been in exclusively
Christian hands for 56 years, and hence clearly belongs to 1243, most probably to the late autumn or
early winter. An alternative possibility is that the truce referred to is the one formulated with Isma'il
during the abortive campaign of June 1243. We should in any case expect the territorial clauses in the
two truces to be identical. This letter specifies that Hebron, Nablus and Beisan were to remain in
Muslim hands, whereas 'Annales de Terre Sainte', B, 441, and 'Gestes', 740, both referring to the
1244 agreement, have Nablus and Jericho. Nevertheless, it seems that in both cases rights in the
Jordan valley were being safeguarded. That Jerusalem had been surrendered outright in the 1243
truce is clear from MGH Epistolae saeculi XIII e regestis pontificum Romanorum selectae, ed.
C.Rodenberg (Berlin, 1883–94), II, 6 (no. 6); CM, IV, 307–8 ('circa principium aestatis proximo
praeteritae').
169 Ibn v, 332–3; cf. also Ibn al-'Amid, p. 155, ll. 5–6, regarding the plan to conquer Egypt. Prawer,
II, 310, is incorrect in stating that Isma'il in person advanced to Gaza.
170HPEC, IV/2, text 88, tr. 181.
171 Ibn v, 249. Humphreys, 262.
172 For a survey of Khwarizmian activity from 1240 to 1243, see Cahen, Syrie du Nord, 646–9; 'La
"Djazîra" au milieu du treizième siècle', 119.
173 Ibn v, 323–4, 325 (hence Ibn al-Furat, Vatican MS, fo. 31r): between and Jumada I 641
A.H./late June and early November 1243.
174 Al-Khazraji, fos. 152r–v. For the Qaymariyya, see also Ta'rikh majmu' al-nawadir,
Forschungsbibliothek Gotha MS Or. 1655, fo. 28r; Ibn v, 336.
175 Al-Khazraji, fo.152v: fa-wafaqa dhalika al-Tatar al-Khwarizmiyya. Ibn Shaddad, ed. 'Abbara,
137. fo. 27r. Ibn v, 336, gives the date of their crossing of the Euphrates as the beginning of
642 A.H. For the Mongol campaign against Aleppo, vide infra.
176 The fullest account is in al-Khazraji, fo. 152v. For 'Ayn al-Jarr, see R.Dussaud, Topographie
historique de la Syrie antique et médiévale (Paris, 1927), 400–2. The attack on the Tripoli region is
mentioned only in 'Eracles', 428.
177 Ibn v, 337. fo. 27v.
178 Gibb, 'The Aiyubids', 709; also supra, p. 54. It should be noted, however, that who had
promised to aid the Seljük sultan against the Mongols in June 1243, failed to do so: Bar Hebraeus (d.
1286), tr. E.A.Wallis Budge, The chronography of Gregory Abu'l Faraj (Oxford and London, 1932),
I, 406–7.
179 HPEC, IV/2, text 111, tr. 230 (for 1241). Ibn v, 332, 338, and Ibn al-'Amid, 155 (for 1244).
180 John of Joinville, Histoire de Saint Louis, ed. N.de Wailly (Paris, 1868), 189.
181 Al-Nuwayri, 341, tells how, when the allies occupied Jerusalem in the summer of 1241, the jurist
al-Sulami was imprisoned not far from Isma'il's quarters. During a visit to the prince, the Frankish
leaders heard al-Sulami reciting the Qur'an in a loud voice, and asked who he was. Isma'il explained
that this was a member of the clergy who was undergoing a second spell of incarceration for his
opposition to the surrender of Muslim-held fortresses to the infidel. 'Were he a priest of ours,' the
Franks replied, 'we should have washed his feet and drunk his broth [with him].' There is a similar
version in al-Subki, v, 101.
182 Ibn v, 338–9. See also fo. 28v.
183 'Memoriale potestatum Regiensium', Rerum Italicarum Scriptores, ed. L.A.Muratori (Milan,
1723–51), VIII, col. 1113, contains what purports to be a summary of a letter from the patriarch of
Jerusalem, condemning the treachery of the Muslim forces at La Forbie. But such sentiments are
lacking in the full texts of his letters given in 'Annales monasterii de Burton', A M, I, 257–63, and
CM, IV, 337–44.
184 Robert, patriarch of Jerusalem, et al. to Innocent IV, in Chronica de Mailros, ed. J.Stevenson
(Edinburgh, 1835), 158 ('quorum…adventus fuit ultra omnium opinionem protelatus').' Annales
monasterii de Burton', 258. CM, IV, 339.
185 Ibn Shaddad, Bodleian Library MS Marsh 333, fo. 110v (brief summary in Cahen,' La "Djazîra"
au milieu du treizième siècle', 119, without the month): latter part of 642 A.H. For the raid, see also
Bar Hebraeus, tr. Budge, I, 409; in the Arabic version of this chronicle, Ta'rikh al-duwal,
ed. (Beirut, 1890), 446, it is dated 641 A.H., i.e. before mid-June 1244; cf. also Ibn (d. 1258),
Nahj al-balagha, ed. M.A.Ibrahim (Cairo, 1959-67), VIII 238(erroneously Placed Prior to Kösedgh).
Thes authours mention only the money paid to the mongols by Aleppo. For that sent by Damscus and
see chronica de mailros, 158 ('non sin multa effusione pecunie') CM, IV, 390, speak merely of
Mongol ultimatums issued 'quibusdam potentibus Sarracenorum soldanis'. On Yasa'ur, see
J.A.Boyle,' Kirakos of Ganjak on the Mongols', Central Asiatic Journal, VIII, 1963, p. 211, n. 95.
186Chronica de Mailros, 158. for the ultimatum to Bohemond, see also CM, IV, 389–90: 'aestate
declinante', however, seems a trifle late.
VOL. L. PART 1. 5
187 They appear to have arrived shortly before 4 October, when the allies moved out of Acre towards
Jaffa: 'Annales monasterii de Burton', 260; CM, IV, 341.
188HDFS, VI/1, 237, 239, 256 (=CM, IV, 302): 'Soldanum…ad evocandum auxilium
Choerminorum…coegerunt.'
189 Bulst-Thiele, 'Ritterorden', 219, and Magistri, 207 and n. 78.
190Die Beziehungen der Päpste zu islamischen und mongolischen Herrschern im 13. Jahrhundert
anhand ihres Briefwechsels, ed. K.Ernst Lupprian (Vatican City, 1981), no. 27 (p. 174).
191 Ibn al-'Amid, p. 155, 1. 3. 'Rothelin', 562. 'Eracles', 430 (though suggesting at p. 427 that Ayyub
had lured them south with the promise of land in Egypt). 'Annales monasterii de Burton', 258; CM,
IV, 338. Chronica de Mailros, 157.
192CM, IV, 289–90.
193 On their relations with the emperor, see Riley-Smith, Knights of St. John. 173–4.
194 ibid., 137–9.
195 'Gestes', 729. This expedition coincided with Filangieri's attempt on Acre, on which vide supra,
p. 50–1; for the date, see further Jackson,'The end of Hohenstaufen rule in Syria', 34.
196 Cahen, Syrie du Nord, 650–1. Riley-Smith, 'The Templars and the Teutonic Knights in Cilician
Armenia', in The Cilician Kingdom of Armenia, ed. T.S.R.Boase (Edinburgh, 1978), 110. Ibn al-
'Adim, Zubda, III, 232 (tr. Blochet, in ROL, v, 96), suggests that the order never recovered from this
reverse. For a brief discussion of the local interests of the military orders, see Prawer, II, 280–1, and
'Military orders and crusader politics', 222.
197 Ibn v, 332; he alleges that the Hospitallers now began to refortify Kawkab. But Ibn Shaddad,
LJP, 161, says nothing of Kawkab's fate after its destruction by father al- in 1220,
merely that it was never restored. Had Kawkab been included in Theobald's truce of 1240 with
The preceptor of the Hospital was killed in the Khwarizmian attack on Jerusalem in August
1244: Chronica de Mailros, 159.
198CM, IV, 289: 'quem non cessavimus pro liberatione Terrae Sanctae pro viribus expugnare.'
199 Al-Nuwayri, 363: see the printed text in 'Le testament d'al-Malik Ayyub', ed. Cl. Cahen
and I.Chabbouh, BEOIFD, xxix, 1977, text 101, tr. 108.
200 Frederick II's claim that (who was not present in person, however, at La Forbie) went
over to the Egyptians during the engagement (CM, IV, 303= HDFS, VI/1, 256–7) is without
foundation. general al-Din Sunqur was among the prisoners taken to Cairo:
494/746.

THE SECRET HISTORY OF THE MONGOLS: SOME FRESH

REVELATIONS
1 William Hung, 'The transmission of the book known as The Secret History of the Mongols',
Harvard Journal of Asiatic Studies, 14, 1951, 433–92. On the author of this piece, see Susan Chan
Egan, A latterday Confucian: reminiscences of William Hung (1893–1980) (Cambridge, Mass.:
Harvard University Council on East Asian Studies, 1987).
2 Francis Woodman Cleaves, The Secret History of the Mongols, Vol. 1 (Cambridge, Mass.: Harvard
University Press, 1982), pp. xvii–lxv. This summary stands out for its clarity, concision and caution; I
have preferred to follow it below, rather than try to relate my own account to Hung's more extended
discussion.
3 Note Hung's remarks at the end of his n. 16 on p. 441 of his study.
4 Cleaves, Secret History, p. xxiii, provides the passage in Chinese.
5 Cleaves, p. liv.
6 Cleaves, pp. lviii–lxix.
7 Cleaves, pp. xxiv, lv–lxiii.
8 Cleaves, p. xxv, again provides the Chinese text.
9 As pointed out by Cleaves, Secret History, p. xxvi.
10 Following the listing in Cleaves, Secret History, p. lx.
11For Ch'iu, see L.Carrington Goodrich and Fang Chao-ying, Dictionary of Ming biography, 1368–
1644 (New York: Columbia University Press, 1976), 249–52, and for Shang Lu, pp. 1161–3.
12 His views are touched upon in Jao Tsung-i's survey Chung-kuo shih-hsüeh shang chih chengt'ung
lun (Hong Kong: Lung-men shu-tien, 1977), 43, 154–6.
13On this incident, which took place in 1449, and on its severe political repercussions, see Ph. de
Heer, The care-taker emperor (Leiden: E.J.Brill, 1986). Two studies of the historiographic
consequences are mentioned in the Introduction to John D.Langlois (ed.), China under Mongol rule
(Princeton: Princeton University Press, 1981), 19.
14 For a brief survey of this episode and its background, see K.Ch'en, Buddhism in China (Princeton:
Princeton University Press, 1964), 421–5, and note the literature listed on pp. 541–2: this has been
expanded over recent yeaers, most notably by P.van der Loon, Taoist books in the libraries of the
Sung period (London: Ithaca Press, 1983), 55–6.
15 On this prince, see the Dictionary of Ming biography, 305–7.
16 Ch'iu Chün, (ed.) Kuo Hsin and Lu Ta-chieh Shih-shih cheng-kang (Ching-
ch'üan: Kuo-shih chia-shu, 1936), 31.19a.
17 Hsiang-mai, Pien-wei lu, p. 751c, in ed. of Taisho Canon, vol. 52.
18 The only such listing I have encountered so far is the Sonkeikaku bunko kanseki bunrui mokuroku
(Tokyo: Sonkeikaku bunko, 1934), 151, which only gives the date as 'Hung-chih period'. This
may show that the library held the first edition, since a postface to the Shih-shih cheng-
kang which has been transmitted with it (reprinted also in Jao, Cheng-t'ung lun, p. 157) dating to
1488, the first year of that period, indicates that the work had not been printed up to that point. It is,
of course, also possible that the work was printed more than once in the Hung-chih period.
19T'ung-chien chieh-yao, hsü-pien, 18.32b. This work entered the library from Japan in the early
1960s (it bears a certain amount of manuscript annotation by a Japanese); it is not the different work
under the same short title preserved in the library of Sir Thomas Wade, which is a Korean print.
20 See, respectively, pp. 101 and 100 of Wang Chung-min, Chung-kuo shan-pen-shu t'i-yao
(Shanghai: Shanghai ku-chi ch'u-pan-she, 1983). As Tang's entry in the Dictionary of Ming
biography, 1252–6, makes clear, his reputation did attract a certain number of false attributions.
Though in this case it would seem hard to judge whether or not T'ang was involved in the work's
compilation in any meaningful sense, the original editing must have taken place during his lifetime
for the attribution to be plausible.
21 Note that the preface of the Hua-i i-yü distinguishes the Uyghur writing used by the Mongols
as'Kao-ch'ang chih shu' and reserves the term used here for the 'Phags-pa script: Hua-i
i-yü (Ssu-pu ts'ung-k'an, series 3 reprint of 1918 Shanghai Han-fen-lou edition), preface, 1b.
22 See de Heer, Care-taker emperor, 139, for this, and note also the mention of Ch'iu on pp. 144,
146. Though earlier scholarship linking the Hua-i i-yü with the Secret History has surmised that the
existing Chinese version was produced as a translation exercise or exemplar, the hypothesis that
Ch'iu found it in the Grand Secretariats's historical archive entails the corollary that the specific
document he saw had been preserved for its historical rather than its linguistic value.
23 In both versions of Ch'iu's remarks that I have consulted, he uses Mathews's no. 6688, rather than
no. 6686, which is in all the texts of the Secret History as well as in the Ta-Ming i-t'ung chih. It is at
least conceivable that this substitution may simply mark a pedantic correction by Ch'iu himself.
24This date is given e.g. by Fu Tseng-hsiang. Ts'ang-yüan ch'ün-shu ching-yen lu (Peking: Chung-
hua shu-chü, 1983), 516.
25 See the preceding note, and also Sonkeikaku bunko kanseki bunrui mokuroku, p. 150, for some
prewar listings, and Pei-ching t'u-shu-kuan ku-chi shan-pen shu-mu (Peking: Shu-mu wen-hsien
ch'u-pan-she, n.d.; preface 1987), 1151, and Kuo-li chung-yang t'u-shu-kuan shan-pen shu-mu
(revised ed., Taipei: National Central Library, 1967), 403, for current holdings.

GHAZAN, ISLAM AND MONGOL TRADITION: A VIEW FROM THE

MAMLUK SULTANATE1
1 This is an expanded version of a paper read at the 204th Annual Meeting of the American Oriental
Society at Madison, WI, on 23 March 1994. I would like to thank Dr. David Morgan and Dr. Igor de
Rachewiltz for reading drafts of this paper and their many useful comments.
2 H.H.Howorth, History of the Mongols, III (1888), 383–4; A.C.M.D'Ohsson, Histoire des Mongols,
IV (repr., Tientsin, 1940, of La Haye and Amsterdam, 1835), 132–3; J.A.Boyle, 'Dynastic and
political history of the Il-Khans', in The Cambridge history of Iran, v, ed. J.A. Boyle (Cambridge,
1968), 378–80; A.Bausani, 'Religion under the Mongols', Cambridge history of Iran, v, 541–3.
3'Padshah-i Islam: The conversion of Sultan Ghazan Khan', Pembroke Papers, 1, 1990, 159–
77.
4ibid., 171.
5 MS Topkapi Sarayl, Ahmet III 2920/25, fols. 60b–Z65b. This entry is found under the entry: '
b.Arghun al-Mughuli al-Jinkiz Khani' with 'Ghazan al-Mughuli' in the margin; was
the Muslim name which Ghazan adopted upon his conversion. In another unpublished volume of al-
Wafi bi'l-wafayat (MS Bodleian Arch. Seld. A.28, fol. 75b), there is a very short entry for 'Ghazan',
but this is a cross-reference, sending the reader to the article b.Arghun…' upon which the
present article is based.
6Two Istanbul manuscripts of were consulted: MS Süleymaniye, Aya Sofya 2968, fol.
3b–7b [henceforth: MS AS]; MS Topkapi Sarayl, Emanet Hazine 1216, fol. 128b-130b [henceforth:
MS EH]. A facsimile edition of based on MS Süleymaniye, Atif Ef. 1809, has recently
been published by F.Sezgin (Frankfurt a.M., 1990). This last mentioned manuscript, however, does
not contain the biography of Ghazan. Sezgin, in the introduction to his edition (p. vii), writes: 'We
had to substitute five missing pages (vol. 2, pp. 326–330), which we took from the eighth part of the
autograph in the Aya Sofya collection (no. 2968), fols. 1b-3b.' Actually these folios are taken from
MS Aya Sofya 2967, which also lacks the entry for Ghazan, as found in MS Aya Sofya 2968.
7 On the relationship between al-Wafi bi'l-wafayat and see D.P.Little, as
biographer of his contemporaries', in D.P.Little (ed.), Essays on Islamic civilization presented to
Niyazi Berkes (Leiden, 1976), 190–210 (repr. in D.P.Little, History and historiography of the
Mamluks [London, 1986], art. I). is clearly the source for the entry on Ghazan in the
fifteenth-century al-Durar al-kamina fi al-mi'a al-thamina by Ibn Hajar (d.
852/1449). In this paper I have used the five-volume Cairo edition (1966) of the work; Ghazan's
biography appears in vol. m, 292–4 (no. 3313). This corresponds with the Hyderabad edition (1348–
50/1929–32, four volumes), III, 212–14.
8Wafi, fol. 60b; MS EH, fol. 128b; Ibn III, 212.
9 Thus in Wafi, fol. 62b; the appellation al-mutatabbib can be translated as 'the practitioner of
medicine'. In MS EH, fol. 128a, he his called instead On this personality, who died in
Damascus in 726/1326, see the introduction to K.Lech, Das mongolische Weltreich:
Darstellung der mongolische Reiche in seinem Werk Masalik
(Wiesbaden, 1968), 29 (there called al-Arbili). cites him several times for information on
the Chaghatayid Khanate (ibid., 75–7).
10On her, see Ch. Melville, 'Bologhan Katun', Encyclopaedia Iranica, IV, 339. F.D.Lessing (ed.),
Mongolian-English dictionary (Berkeley and Los Angeles, 1960), gives the spelling of this word
(meaning 'sable') as bulayan, but cf. G.Doerfer, Mongolische und türkische Elemente im
Neupersischen [henceforward: TMEN] (Wiesbaden, 1962–75), I, 215.
11 Rashid al-Din, al-tawarikh, III, ed. (Baku, 1957), 301;=K.Jahn (ed.),
Geschichte aus dem des Rašid al-Din… (London, 1940), 80 (left
column). As far as I can tell, this story is not mentioned in history (Ta'rikh-i
=waTajziyat wa-tazjiyat [repr. Tehran, 1338S./1959, of Bombay 1269H/1852–3]).
12 Boyle, 'Il-Khans', 380, cites the Qur'an, IV, 26: 'And marry not women whom your fathers have
married: for this is a shame, and hateful, and an evil way; though what is past may be allowed to
happen.' Boyle wonders how any self-respecting Muslim dignitary could have officiated at such a
ceremony, which blatantly contradicted the See also J.Schacht, ' I in the classical
Islamic law', EI2, VIII, 27.
13 The matter of the Yasa will be discussed below; in passing it might be mentioned that the term
'yasa' could also refer to a particular precept and not just the entire legal corpus. This particular rule is
mentioned by Rashid al-Din (ed. 6), who writes that Hülegü thus married his father's
widows 'in accordance with the Yasa (ba-rah-i yasaq)'. On this practice, see William of Rubruck, in
A.van den Wyngaert, Sinica Franciscana, I (Quaracchi-Firenze, 1929), 184–5; translation in
P.Jackson (tr.), and P.Jackson and D.Morgan (introduction, notes and appendices), The mission of
Friar William of Rubruck (London, 1990), 91–2, and n. 1 on p. 92. Boyle, 'Il-Khans', 380, shows that
this custom was well established among the Mongol royal family in Iran, as well as being an ancient
practice among the tribes of the Eurasian steppe.
14Wafi, fols. 62b–63a; MS EH, fol. 128a; MS AS, fol. 3b; cf. the summary in Ibn Hajar, III,
292–3, which differs in details from that presented by It is interesting to note that the
unnamed scholar mentioned in this passage was subject to some criticism for his permissive
interpretation of the law, but he replied that adopting an indulgent position and thus preventing
Ghazan's apostasy and his subsequent antipathy to Islam was the best solution. This cogent
explanation was accepted.
15 On the Yasa, see P.Ratchnevsky, Genghis Khan: his life and legacy, ed. and tr. T.N. Haining
(Oxford and Cambridge, Mass., 1991), 187–96; D.Ayalon, 'The Great Yasa of Chingiz Khan: a
reexamination', Studia Islamica, pt. A, vol. 33 (1971), 97–140; pt. B, vol. 34 (1971), 151–80 (the
four parts of this article—of which parts C1 and C2 deal with the position of the yasa in the Mamluk
Sultanate—have been republished in D.Ayalon, Outsiders in the lands of Islam [London, 1988]);
D.Morgan, The Mongols (London, 1986), 96–9; idem, 'The "Great Yasa of Chingiz Khan" and
Mongol law in the Ilkhanate', BSOAS, XLIX, 1, 1986, 163–76; I. de Rachewiltz, 'Some reflections on
Cinggis Qan's ', East Asian History, 6, 1993, 91–104. The transliteration of Yasa presents some
problems. I have eschewed the original Mongolian form in favour of that based on the Turkic
derivative yasa which is used in the Muslim sources. This is usually rendered yasa or yasa in the
Arabic and Persian texts, although yasaq is also found.
16 On the previous folio of Wafi, cites on Ghazan's name and genealogy, as well as
his physical characteristics and personality. The mention of the languages which Ghazan knew, etc.,
would seem to be a continuatipn of this description. According to Little, 203–4, virtually all
of the information which derived from was transferred orally and not through the
latter's written works.
17 Thus in Wafi. MS AS, has 'he spoke Turkish, Mongolian and Persian'; MS EH, omits
Turkish from the list.
18 This form is closer to the Mongol li-ajal yasa jinkaz khan
19Wafi: ahadha nafsahu fi al-siyasa ma 'khadh jinkiz khan. The parallel passage in is slightly
different, but more clear: ahadha nafsahu jinkiz khan. At the beginning of the entry, however,
the latter work does have the following sentence: lamma malaka, akhadha nafsahu fi'l-mulk
ma'khadh jinkaz khan.
20 On the arghu (generally yarghu<Mongolian ), the Mongol combination of committee of
inquiry and court-martial, and see D.Morgan, 'The "Great Yasa of Chingiz Khan ",' 173–6.
21Wafi, fol. 61a; some minor differences are found in MS EH. 128a-b; MS AS, fols. 3b-4a,
which are noted above; there is a short summary in Ibn III, 212. For agha (<Mo. aga; Turkish
ary) and ini (<Tu. ini), see Doerfer, TMEN, I, 133–40, II, 226. The sentence would seem to mean that
cadet member of the Mongol royal family in the Ilkhanate were to defer to the senior members, with
the added implication that the present hierarchy, with Ghazan at the summit, was to be maintained.
22 According to Rashid al-Din (ed. 'Alizadah, 379;=ed. Jahn, 171), Ghazan knew besides Mongolian
some Arabic, persian, 'Hindi', 'Kashmiri', Tibetan, Frankish and other languages. B.Spuler,
Die Mongolen in Iran, 4th ed. (Leiden, 1985), 380, n. 59, notes that Turkish should have been
mentioned in this passage. Whether Ghazan actually spoke all of these languages remains a moot
point. It is possible that (and ) may have known of Ghazan's linguistic skills and
just did not bother to list the non-Islamic languages. Spuler, it should be added, doubts Ghazan's
knowledge of Arabic.
23 Rashid al-Din, ed. 'Alizadah, 511 [=ed. Jahn, 303], cited in Morgan, "The "Great Yasa of Chingiz
Khan",' 172, but compare his comment.
24J.A.Boyle (in the glossary of The successors of Genghis Khan [New York and London, 1971],
341), writes of the yosun: 'Mongol customary law, as distinct from the yasa of Genghis Khan.'
25 See The Secret History of the Mongols, tr. F.W.Cleaves (Cambridge, MA, 1982), I, 271; Doerfer,
TMEN, I, 149–52.
26 See Secret History, tr. Cleaves, I, 276; Doerfer, TMEN, I, 424.
27Rashid al-Din, ed. 251;=ed. Jahn, 8. This passage is mentioned briefly in Morgan, 'The
"Great Yasa of Chingiz Khan",' 172.
28See the comment of P.Jackson, 'Chaghatayid Dynasty', Encyclopaedia Iranica, v, 344.
29I understand that the expression refers to both Persian and Arabic. Thus he did not
reveal that he understood Arabic and spoke Persian only with a few close associates 'out of pride in
the yasa'.
30 The term yasa is subsequently used in the biography in a meaning different from a corpus of laws,
but rather as an individual command; Wafi, fols. 61b-62b; MS. EH, fol. 129a; MS. AS, fol. 4a.
For the application of the term yasa to the individual commands of a particular Qa'an, and in one case
to the orders of Chaghatai, see Morgan, 'The "Great Yasa of Chingiz Khan",' 166–73.
31 Wafi, fol. 62b; MS EH, fol. 129b; MS AS, fol. 5a. Aytamish (or Etmish), a trusted mamluk
of b.Qalawun (709–741/1310–1341), was an adviser on Mongol affairs and served
as an envoy several times to Abu see Ayalon,' The Great Yasa of Chingiz Khan', pt. C2,
131–40; D.P.Little, 'Notes on Aitamiš, a Mongol Mamluk', in U.Haarmann and P.Bachmann (ed.),
Die Islamische Welt zwischen Mittelalter and Neuzeit: Festschrift für Hans Robert Roemer zum 65.
Geburtstag (Beirut, 1979), 386–401 (repr. in Little, History and historiography, art. VI).
32Wafi, fol. 62b; MS EH, fol. 129b; MS AS, fol. 5a-b. The expression yasa al-mughul is found
only in in Wafi, the word yasa is missing. Only mentions that is the ultimate
source of this information.
33Qashani, 98, cited in Morgan, 'The "Great Yasa of Chingiz Khan",' 172, on which this translation is
based.
34 ed. Lech, p. 41 of Arabic text.
35Morgan, 'The "Great Yasa of Chingiz Khan",' 172. Qutlugh-shah also had a confused idea of some
aspects of Mongol history: in a conversation with the Syrian theologian Ibn Taymiyya, he stated that
Chinggis Khan (who he claimed was his ancestor) was a Muslim; Ibn al-Dawadari, al-Kanz al-durar
al-ghurar, ix, ed. H.R.Roemer (Cairo, 1971), 32, citing the historian al-Birzali, who in turn
recorded this from Ibn Taymiyya's testimony.
36 The Qa'an, sometimes called the Great Khan, was the supreme ruler of the Mongol empire, while
'khan' was applied to lesser Mongol princes who ruled the uluses (royal appanages which eventually
became independent states). For these two titles, see I. de Rachewiltz, 'Qan, Qa'an and the seal of
Güyüg', in K.Sagaster and M.Weiers (ed.), Documenta Barbarorum: Festschrift für Walther Heissig
zum 70. Geburtstag (Wiesbaden, 1983), 281–98.
37 MS Wafi: aw, which is a mistake for wa-; the latter particle is found in
38 The mentioning of a name in the sermon and on the coinage are the two major symbols of
sovereignty in traditional Islamic political life.
39 cf. 78, where the title 'Great Khan' is rendered al-qan al-kabir. In general, both
and use qan instead of the more usual khan found in the Mamluk sources. Qan is closer to
the Middle Mongolian qan, while khan resembles the Turkic form of this title. Both authors write
jinkiz khan for Chinggis Khan, as this must have been a fixed expression. Besides the usage of khan
in this case, uses khan only twice in the part of work edited by Lech (see index,
s.v.khan).
40 Much of this information in these two paragraphs is found, albeit in a more condensed form (but
with some additions) in al-'Umari, who cites Shams al-Din Lech, Das mongolische
Weltreich, 19 of Arabic text; 32–4 of introduction (for information on ). wrongly
ascribes to Arghun the addition of his name to that of the Qa'an on coins. In reality, Hülegü had
already begun doing this.
41Kharbanda (Per. 'Ass-Herd') was the original name of (Mon. 'Lucky One'), Ghazan's brother
and successor who ruled 1304–16. He was also known by the name Khudabanda (Per. 'Servant of
God'), as well as See Boyle, 'Il-Khans', 398.
42Wafi, fol. 62a-b; cf. MS EH, 129a–b (missing a line of text on fol. 129b); MS AS, fols. 4b–
5a, for some minor differences. Cf. also the shorter version in ed. Lech, Arabic text. 19.
43See N. and R.Amitai-Preiss, 'Two notes on the protocol on Hülegü's coinage', Israel Numismatic
Journal, 10, 1988–89 [1991], 117–21; R.Amitai-Preiss, 'Evidence for the early use of the title ilkhan
among the Mongols', JRAS, NS, 1, 1991, 353; in idem, 'An exchange of letters in Arabic between
Aba a Ilkhan and Sultan Baybars (A.H. 667/A.D. 1268–9)', Central Asiatic Journal, 38, 1994, 11–
33, there is a discussion of other possible translations of this term: I am now less certain that 'subject
khan' is the correct translation.
44 See the discussion in T.Allsen, 'Changing forms of legitimation in Mongol Iran', in G.Seaman and
D.Marks, Rulers from the steppe: state formation on the Eurasian periphery (Los Angeles, 1991),
230–1, who also mentions a Mongolian legend on Ghazan's coins minted in Georgia which mention
the Qa'an.
45 Spuler, Die Mongolen in Iran, 221–4; E.Blochet, Introduction à I'histoire des Mongols de Fadl
Allah Rashid-ed-Din (Leiden and London, 1910), 231–2; T.Allsen, 'Two cultural brokers of medieval
Eurasia: Bolad Aqa and Marco Polo', in M.Gervers and W.Schlepp (ed.), Nomad diplomacy,
destruction and religion from the Pacific to the Adriatic (Toronto Studies in Central and Inner Asia,
no. 1, Toronto, 1994), 63–78.
46 Spuler, Die Mongolen in Iran, 222; Blochet, Introduction à l'histoire des Mongols, 230.
47 P.Pelliot, Notes on Marco Polo, I (Paris, 1959), 120–1; for the return of this embassy, see ibid, I,
393. See Allsen, 'Legitimization', 241, n. 67 for evidence of other missions.
48 Appendix to al-Maqrizi, Kitab al-suluk fi al-duwal wa'l-muluk, I, ed. M.M.Ziyada (Cairo,
1934–9), 978.
49 This brings to mind the tanistry thesis suggested by J.F.Fletcher, 'Turco-Mongolian monarchic
tradition in the Ottoman Empire', Harvard Journal of Ukrainian Studies, 3–4, 1979–80, 236–51; see
also Morgan, The Mongols, 38–9.
50 III, 293. Ibn however, says that Ghazan said this when he drove out the Jochid
representatives from Rum (Anatolia; according to the editor, one MS reads Iraq). ed. Lech,
79.
51 On this ideology, see the discussion and bibliography in R.Amitai-Preiss, Mamluks and Mongols:
the Mamluk-Ilkhanid war, 1260–1281 (Cambridge, 1995), 10–11.
52See R.Amitai-Preiss, 'Aims and motivations of Ilkhanid strategy towards Syria and the Mamluks',
in D.Morgan (ed.), The Mongol Empire and its legacy, forthcoming.
53 Besides the discussion in the article mentioned in the previous note, see T.Raff, An anti-Mongol
fatwa of Ibn Taimiya (privately printed, Leiden, 1973), 33–5. On p. 30, Raff discusses the use of both
Islamic and Chinggisid motifs in Ghazan's proclamations to the population of Damascus in 699/1300.
54Rashid al-Din, ed. 'Alizadah, 350–1;=ed. Jahn, 141–2; cited in Boyle, 'Il-Khans', 392–3. For
khuday-i qadim (and khuday-i da'im) as Tengri, see S.Heidemann, Der Aleppiner Kalifat (AD 1261):
vom Ende des Kalifates in Bagdad über Aleppo zu den Restauration in Kairo (Leiden, 1994), 332–3,
336, 338.
55 See Boyle, 'Il-Khans', 402.
56 N.Levtzion,' Towards a comparative study of Islamization', in N.Levtzion, Conversion to Islam
(New York and London, 1979), 19.
57 Ibn Taymiyya, fatawi ibn taymiyya (Beirut, n.d.), IV, 280–98, esp. 286–8. This passage is
analysed in detail by Raff, An anti-Mongol fatwa of Ibn Taimiya, 44–59; see ibid., 5–7, for a
discussion of the dating of this fatwa. A summary of Ibn Taymiyya's approach is found in Ibn Kathir,
Tafsir (Cairo, 1342/1923), n, 67, who uses the word yasaq (sic) for the Mongol ; see the
discussion in E.Sivan, Radical Islam: medieval theology and modern politics (New Haven, 1985),
96–7.
58 See D.P.Little, 'The historical and historiographical significance of the detention of Ibn Taymiyya',
International Journal of Middle Eastern Studies, 4, 1973, 311–27 (repr. in Little, History and
historiography, art. VII). Cf. Sivan, Radical Islam, 96–7, who suggests that Ibn Taymiyya 's hostility
to the Mongols was due to the fact that he had fled as a child from Mongol-controlled territory, and
was thus suffering from a 'refugee syndrome'.
59 Ibn al-Dawadari, IX, 127.

Marco Polo and his 'Travels'1

1 Earlier versions of this study were read to my colleagues in the History Department at Keele
University, and to the Seminar on the History of the Middle East at the School of Oriental and
African Studies, London, in April 1996.1 am grateful for the stimulating questions and discussions
that followed.
2 Hiroshi Watanabe (comp.), Marco Polo bibliography 1477–1983 (Tokyo, 1986).
3 Martin Gosman, 'Marco Polo's voyages: the conflict between confirmation and observation', in
Zweder von Martels (ed.), Travel fact and travel fiction: studies on fiction, literary tradition,
scholarly discovery and observation in travel writing (Leiden, 1994), 72–84 (see especially pp. 76–7,
83–4). For earlier views of the Mongols, see Gian Andri Bezzola, Die Mongolen in abendländischer
Sicht: ein Beitrag zur Frage der Völkerbegegnungen (1220–1270) (Berne and Munich, 1974);
Felicitas Schmieder, Europa und die Fremden: die Mongolen im Urteil des Abendlandes vom 13. bis
in das 15. Jahrhundert (Beiträge zur Geschichte und Quellenkunde des Mittelalters, 16, Sigmaringen,
1994). Similarly, the delay in the West's absorption of the new information from the' sophisticated'
East is compared with the easy assimilation of the material on the relatively 'uncivilized' Canary
islanders: J.K.Hyde, 'Real and imaginary journeys in the later Middle Ages', Bulletin of the John
Rylands Library, LXV, 1982, 138–40.
4 John Critchley, Marco Polo's book (Aldershot, 1992), xiv; also the 'Epilogue' (pp. 178–9). My debt
to Critchley's book will be apparent to anyone who has read it.
5Frances Wood, Did Marco Polo go to China? (London, 1995): see especially her 'Conclusions' (pp.
140–51).
6 Arthur Waldron, 'The problem of the Great Wall of China', Harvard Journal of Asiatic Studies,
XLII, 1983, 643–63; idem, The Great Wall of China: from history to myth (Cambridge, 1990). For a
brief defence of Polo in respect of the other omissions, see J.R.S.Phillips, The medieval expansion of
Europe (Oxford, 1988), 118–19.
7 For example, by John W.Haeger, 'Marco Polo in China? Problems with internal evidence', Bulletin
of Sung-Yuan Studies, xiv, 1978, 22–30.
8 Rashid al-Din, II, ed. E.Blochet (Leiden and London, 1911), 481–3, and transl.
J.A.Boyle, The successors of Genghis Khan (New York, 1971), 280–1; see further E.Chavannes,
review of Berthold Laufer, History of the finger-print system (Washington, 1913), in xiv,
1913, 490–1.
9 Leonardo Olschki, Marco Polo's Asia, (tr.) J.A.Scott (Berkeley and Los Angeles, 1960), does a
good job of placing the Polos' journeys in historical context, though the book is marred by a tendency
to be too uncritical and at times excessively eulogistic.
10David Morgan, The Mongols (Oxford, 1986), 118–19, 156–8.
11 For what follows, see generally Phillips, chs. 5–7.
12 Luciano Petech, 'Les marchands italiens dans l'empire mongol', Journal Asiatique, CCL, 1962
549–74.
13 J.A.Boyle, 'The Il-khans of Persia and the princes of Europe', Central Asiatic Journal, xx, 1976,
25–40. Denis Sinor, 'The Mongols and western Europe', in K.M.Setton (general ed.), A history of the
crusades, III (ed. H.W.Hazard). The fourteenth and fifteenth centuries (Madison, Wisconsin, 1975),
530–9. For Egypt and the Golden Horde, see S.Zakirov, Diplomaticheskie otnosheniya Zolotoi Ordy s
Egiptom (Moscow, 1966).
14 Carl Theodor Gossen, 'Marco Polo und Rustichello da Pisa', in Manfred Bambeck and Hans
Helmut Christmann (ed.), Philologica Romanica Erhard Lommatzsch gewidmet (Munich, 1975),
133–43.
15Critchley, 18–19, 52.
16ibid., 9, 139. For an example of a seemingly abridged passage, on 'Caragian', see the composite
translation by [A. C.] M[oule and Paul] P[elliot, The description of the world,] I, [(London, 1938, 2
vols; II is an edition of the Z version)], 278, n. 3: all future references are to this translation.
17 The most recent edition of this text is by Ruggiero M.Ruggieri (ed.), Il Milione (Florence, 1986).
18 See, for instance, the plea of Valeria Bertolucci Pizzorusso, 'A propos de Marco Polo et de son
livre: quelques suggestions de travail', in Essor et fortune de la Chanson de geste dans l'Europe et
l'Orient latin: Actes du ixe Congrès international de la Société Rencesvals pour l'étude des épopées
romanes. Padoue-Venise, 29 août-4 septembre 1982, II, (Modena, 1984), 797.
19 Hyde, 'Real and imaginary journeys', 130–1.
20 Salimbene de Adam, 'Cronica', ed. O.Holder-Egger, in Monumenta Germaniae Historica.
Scriptores (Hanover etc., 1826–1913), XXXII, 210, 213.
21 Roger Bacon, Opus Majus, ed. J.Bridges (Oxford, 1897–1900, 3 vols.), I, 305. See generally Jarl
Charpentier,' William of Rubruck and Roger Bacon', in Hyllningsskrift tillägnad Sven Hedin på hans
70-årsdag den 19 Febr. 1935 (Stockholm, 1935), 255–67.
22 MP, I, 31–2, 34–5. It is improbable, incidentally, that Polo was captured in the battle off Ayas in
1296; a minor sea engagement, at a slightly later date, has been proposed.
23 Sir Henry Yule, Cathay and the way thither, new edn. by Henri Cordier (Hakluyt Society, 2nd
series, XXXIII, XXXVII, XXXVIII, XLI, London, 1913–16, 4 vols.), III, 195; a fuller quotation in
Paul Pelliot, Notes on Marco Polo (Paris, 1959–73, 3 vols with continuous pagination), I, 601–2.
24Translated in MP, I, 60; also reproduced in Sir E.Denison Ross, 'Marco Polo and his book',
Proceedings of the British Academy, xx (1934), 201 (text), 202–3 (transl.).
25 This MS was used by M.G.Pauthier as the base for his edition, Le Livre de Marco Polo (Paris,
1865): its preface appears ibid., 1–2, and is translated in MP, I, 61–2. Ross, 'Marco Polo', 192, was
too dismissive of the 'De Cepoy legend', but it should be pointed out that the date of the gift, August
1307, is impossible, since De Cepoy had left Venice for Brindisi by May: Joseph Petit, 'Un capitaine
du règne de Philippe le Bel: Thibaut de Chepoy', Le Moyen Âge, x=2e série, I (1897), 231–4.
26 MP, I, 28, 556 (and cf. 555, n.l).
27 Critchley, 21. This detail is not found in Jacopo d'Acqui, as Wood claims (pp. 42, 142).
28 Jacques Heers, Marco Polo (Paris, 1983), 290–2.
29 Examples in Critchley, 34.
30 ibid., 49, citing MP, I, 276.
31 A point well made by Olschki, Marco Polo's Asia, 97–9, 111; see also Heers, Marco Polo, 165–
85, 258. But for a more positive assessment of the mercantile point of view as found in the Polo
book, see Antonio Carile, 'Territorio e ambiente nel "Divisament dou monde" di Marco Polo', Studi
Veneziani, n.s., I, 1977, 13–36; Ugo Tucci, 'Marco Polo, mercante', in Lionello Lanciotti (ed.),
Venezia e l'Oriente (Florence, 1987), 323–37.
32 Heers, Marco Polo, 112–17.
33 Critchley, 38. Angeliki E.Laiou, Constantinople and the Latins: the foreign policy of Andronicus
II 1282–1328 (Harvard Historical Studies, LXXXVIII, Cambridge, Mass., 1972), 206–9.
34 See Critchley, 71, 136; though he also points out (pp. 72–5) that the book's attitude towards the
Mongol alliance is less than enthusiastic.
35 MP, I, 59–60; and see Ross, 'Marco Polo', 200–1 (text), 202 (transl.).
36 Olschki, Marco Polo's Asia, 111, 115 (and see his fig. 3, facing p. 117).
37 John J.Nitti (ed.), Juan Fernàdndez de Heredia's Aragonese version of the Libro de Marco Polo
(Madison, Wisconsin, 1980).
38 For what follows, see Jacques Heers, 'De Marco Polo à Christophe Colomb: comment lire le
Devisement du monde?', Journal of Medieval History, x, 1984, 125–43.
39 For an attempt to outline an itinerary for the Polos, see Olschki, Marco Polo's Asia, 12–38.
40 Wolfgang Lentz, 'War Marco Polo auf dem Pamir?', ZDMG, n.F. XI, 1933, 1–32, concluded that
the visit to Badakhshan was authentic.
41 See the remarks of Yule, Cathay, ed. Cordier, IV, 48–9, 130 n. l, 140 n. l; Cordier's note on the
great mosque at Canton, ibid., 122, n. l; Ross Dunn, The adventures of Ibn Batuta, a Muslim traveler
of the fourteenth century (London, 1986), 252–3, reviews the problems, but gives Ibn the
benefit of the doubt. See my review of vol. IV of the translation by H.A.R.Gibb and
C.F.Beckingham, in Journal of the Royal Asiatic Society, 3rd series, vi, 1996, 262–6.
42 Critchley, 81.
43 Pelliot, Notes, II, 646–7.
44 This did not prevent fabulous creatures finding their way into the illustrations in some of the Polo
MSS: see R.Wittkower, 'Marco Polo and the pictorial tradition of the marvels of the east', in Oriente
Poliano. Studi e conferenze tenute all'IsMEO in occasione del VII centenario della nascità di Marco
Polo (1254–1954) (Rome, 1957), 155–72; John Block Friedman, The monstrous races in medieval
art and thought (Cambridge, Mass., 1981), 154–8.
45 For the old misconception, see Berthold Laufer, 'Asbestos and salamander: an essay in Chinese
and Hellenistic folk-lore', XVI. 1915, 299–373.
46 Louis Hambis, 'Le voyage de Marco Polo en Haute Asie', in Oriente Poliano, 183–4. David
Morgan, 'Prester John and the Mongols', in Charles F.Beckingham and Bernard Hamilton (ed.),
Prester John, the Mongols and the Ten Lost Tribes (Aldershot, 1996), 165–6.
47 Critchley, 83–4.
48 Pelliot, Notes, II, 774–5.
49ibid., II, 814–15.
50For Kan-chou, see ibid., I, 150–3. It is noteworthy that the Tuscan version (Ruggieri, 150) omits
Marco's name here, which might suggest that the visit fell during the first journey.
51Paul Pelliot, 'Les traditions manichéennes au Fou-kien', XXII, 1923, 193–208; and in
his Notes, II, 726–8.
52 This reference to Marco is omitted in the Tuscan version (Ruggieri, 106).
53 Critchley, 78–9, and cf. also 82–3 for pilgrims' guides; for Chinese geographical writing, see ibid.,
xii. Heers, Marco Polo, 241–2, draws analogies between the tone of Polo's book and the spiced-up
account of Hülegü's invasion of Persia, based on the report of Te and presented to Qubilai by
Liu Yu in 1263: for the text, see Emil Bretschneider, Mediaeval researches from eastern Asiatic
sources (London, 1888, 2 vols.), I, 122–56.
54Pelliot, Notes, II, 812.
55 For a resumé of the material on India, see K.A.Nilakanta Sastri, 'Marco Polo on India', in Oriente
Poliano, 111–20.
56 Herbert Franke, review of Olschki, Marco Polo's Asia, in ZDMG, CXII, 1962, 229–31; and his
'Sino-Western contacts under the Mongol empire', Journal of the Hong Kong Branch of the Royal
Asiatic Society, VI, 1966, 54–5.
57 Guillaume Adam, 'De modo Sarracenos extirpandi', in Recueil des Historiens des Croisades
[hereafter RHC]. Documents Arméniens, II (Paris, 1906), 553. See also the comments in B.Z. Kedar,
Merchants in crisis (New Haven and London, 1976), 10–11; Michel Balard, 'Les Génois en Asie
centrale et en extrême-orient au XIVe siècle: un cas exceptionnel?', in Économies et sociétés au
moyen âge: Mélanges offerts à Édouard Perroy (Paris, 1973), 681–9.
58'Epistolae Fr. Iohannis de Monte Corvino', in Anastasius Van den Wyngaert (ed.), Sinica
Franciscana, I. Itinera et relationes Fratrum Minorum saeculi XIII et XIV (Quaracchi-Firenze,
1929), 352–3; tr. in Christopher Dawson (ed.), The Mongol mission (London, 1955), 229. For the
date of Montecorvino's departure from Tabriz, see his second letter, in Van den Wyngaert, 345 (tr.
Dawson, 224).
59 Laufer, 'Asbestos and salamander', 365.
60 Emilio Bottazzi, 'Un'esploratione alle sorgenti del Fiume Giallo durante la dinastia Yüan', Annali:
Istituto Orientale di Napoli, n.s. XIX, 1969, 529–46, with the year 1280 in error. Herbert Franke,
'The exploration of the Yellow River sources under emperor Qubilai in 1281', in G.Gnoli and
L.Lanciotti (ed.), Orientalia Iosephi Tucci memoriae dicata (Rome, 1985), 401–16; repr. in Franke,
China under Mongol rule (Aldershot, 1994).
61 W.W.Rockhill, 'Notes on the relations and trade of China with the eastern archipelago and the
coast of the Indian Ocean during the fourteenth century: part I', T'oung Pao, xv, 1914, 429–42.
62Ch'ên Yüan, Western and Central Asians in China under the Mongols, tr. Ch'ien Hsing-hai and
L.Carrington Goodrich (Monumenta Serica Monographs, xv, Los Angeles, 1966), 1–2. Igor de
Rachewiltz, 'Some remarks on the language problem in Yüan China', Journal of the Oriental Society
of Australia, v, 1967, 65.
63 Denis Sinor, 'Interpreters in medieval Inner Asia', in Marcel Erdal (ed.), Studies in the history and
culture of Central Eurasia (Jerusalem, 1982=Asian and African Studies, XVI), 307–16.
64Jean Richard, 'Isol le Pisan: un aventurier franc gouverneur d'une province mongole?', Central
Asiatic Journal, xiv, 1970, 186–94; repr. in his Orient et Occident au Moyen Âge: contacts et
relations (XIIe-XVes.) (London, 1976). Jacques Paviot, 'Buscarello de' Ghisolfi, marchand génois
intermédiaire entre la Perse mongole et la Chrétienté latine (fin du XIIIme-début du XIVme siècles)',
in Storia dei Genovesi, XI (Genoa, 1991), 107–17. Sinor, 'The Mongols and Western Europe', 534–7.
65 Morris Rossabi, 'The Muslims in the early Yüan dynasty', in John D.Langlois (ed.), China under
Mongol rule (Princeton, 1981), especially 257–60, 270–95; idem, Khubilai Khan: his life and times
(Berkeley and Los Angeles, 1988), 70–5. For a revisionist view of the effectiveness of this policy,
especially in the later decades of Mongol rule, cf. Elizabeth Endicott-West, Mongolian rule in China:
local administration in the Yuan dynasty (Harvard-Yenching Institute Monograph series, 29,
Cambridge, Mass., 1989), 78–88; see ibid., 122, for the social isolation of the Mongols (and
presumably, therefore, of other foreign officials) in Yüan China.
66 Pelliot, Notes, I, 68. Francis Woodman Cleaves, 'The biography of Bayan of the Barin in the Yüan
Shih', Harvard Journal of Asiatic Studies, XIX, 1956, 186–8. Polo (or perhaps Rusticello) transferred
to Bayan's Chinese rank an incorrect explanation of the Mongol name bayan ('rich') as deriving from
Chinese po-yen, 'hundred eyes'.
67 Sinor, 'Interpreters', 307–16.
68On this script, see N.N.Poppe (ed.) and John R.Krueger (tr.), The Mongolian monuments in
ags-pa script (Göttinger Asiatische Forschungen, VIII, Wiesbaden, 1957); De Rachewiltz, 'Some
remarks', 71–3; Rossabi, Khubilai Khan, 155–60. Heers, Marco Polo, 234, suggests that Arabic was
one of Polo's four languages.
69 'Les Gestes des Chiprois', RHC Documents Arméniens, II (Paris, 1906), 842.
70 Huang Shijian, 'The Persian language in China during the Yuan dynasty', Papers on Far Eastern
History, XXXIV (September 1986), 83–95.
71 Pelliot, Notes, I, 94–5. For the outbreak of hostilities between Berke and the Emperor Michael,

see Marius Canard, 'Un traité entre Byzance et l'Égypte au XIIIe siècle', in Mélanges offerts à
Gaudefroy-Demombynes (Cairo, 1939–45), 213–19.
72 M.M.Siouffi, 'Notice sur un patriarche nestorien', Journal Asiatique, 7e série, XVII, 1881, 90. The
evidence is discussed by Pelliot, Recherches sur les chrétiens d'Asie centrale et d'extrême-orient
(Paris, 1973), 257–9, and by Morris Rossabi, Voyager from Xanadu: Rabban Sauma and the first
journey from China to the West (Tokyo and New York, 1992), 43–6.
73 Karl-Ernst Lupprian (ed.), Die Beziehungen der Päpste zu islamischen und mongolischen
Herrschern im 13. Jahrhundert anhand ihres Briefwechsels (Vatican City, 1981), 237–41 (no. 47); cf.
also Nicholas's letter to the Il-khan, 1 April 1278, ibid., 233–6 (no. 46). For Rabban Sauma, see Jean
Richard, 'La mission en Europe de Rabban Çauma et l'union des églises', in XII Convegno Volta
(Rome, 1957), 162–7; repr. in his Orient et Occident.
74 E.A.Wallis Budge (tr.), The monks of Kûblâi Khân Emperor of China (London, 1928), 181.
Rossabi, Voyager from Xanadu, 134–8.
75 M.H.Laurent, 'Grégoire X et Marco Polo (1269–1271)', Mélanges d'Archéologie et d'Histoire de
l'École Française de Rome, LVIII (1941–6), 132–44. 'Annales de Terre Sainte', B, ed. R.Röhricht and
G.Raynaud, Archives de l'Orient Latin, II (1884), documents, 455, has 10 November; 'L'estoire de
Eracles empereur', RHC Historiens Occidentaux, II (Paris, 1859), 471, gives the date of Gregory's
embarkation as the octave of St. Martin, i.e. 18 November. For William of Agen, see Bernard
Hamilton, The Latin Church in the Crusader states: the secular church (London, 1980), 270–5.
76 'L'estoire de Eracles', 471. Hans Prutz, Kulturgeschichte der Kreuzzüge (Berlin, 1883), 575.
77La chronique attribuée au Connétable Smbat, (tr.) Gerard Dédéyan (Paris, 1980), 134; also transl.
in A.G.Galstian, Armianskie istochniki o Mongolakh izvlecheniya iz rukopisei XIII-XIV vv. (Moscow,
1962), 64.
78 Peter Thorau, The Lion of Egypt: Sultan Baybars I and the Near East in the thirteenth century,
(tr.) P.M.Holt (London, 1992), 208–9. Reuven Amitai-Preiss, Mongols and Mamluks: the Mamluk-
Ilkhanid war, 1260–1281 (Cambridge, 1995), 125.
79 Critchley, 66–8. For the Lord Edward's appeal, see 'L'estoire de Eracles', 461; R.Röhricht, 'Études
sur les derniers temps du royaume de Jérusalem, A. La croisade du prince Édouard d'Angleterre
(1270–1274)', Archives de l'Orient Latin, I (1881), 623.
80Giovanni Soranzo, Il papato, l'Europa cristiana e i Tartari (Milan, 1930), 217 and n. 2.
81 Burkhard Roberg, 'Die Tartaren auf dem 2. Konzil von Lyon (1274)', Annuarium Historiae
Conciliorum, v (1973), 288, n. 268, suggests that Gregory wrote to the Il-khan Abaqa, at least, from
the Holy Land prior to his departure for Italy, in order to notify him of his plans to convene the
council.
82 See Jean Richard, La papauté et les missions d'Orient au Moyen-Âge (XIIe-XVe siècles) (Rome,
1977), 85–6. Salimbene, 'Cronica', 210, who names the papal envoys, says in error that they were
sent by John XXI.
83 Louis Hambis, 'Le prétendu "Cogatal" de Marco Polo', in Nel VII. centenario della nascità di
Marco Polo (Venice, 1955), 235–40.
84 Critchley, 4, 6–7.
85 Cte. L.de Mas Latrie, Histoire de l'île de Chypre sous le règne des princes de la maison de
Lusignan (Paris, 1852–61, 3 vols.) II, 78–9; Cornelio Desimoni (ed.), 'Actes passés en 1271, 1274 et
1279 à l'Aïas (Petite Arménie) et à Beyrouth par devant des notaires génois', Archives de l'Orient
Latin, I (1881), documents, 441. Catherine Otten-Froux, 'L'Aïas dans le dernier tiers du XIIIe siècle
d'après les notaires génois', in B.Z.Kedar and A.L.Udovitch (ed.), The medieval Levant: studies in
memory of Eliyahu Ashtor (1914–1984) (Jerusalem, 1988=Asian and African Studies, XXII), 154–5.
86'Epistolae Fr. Iohannis de Monte Corvino', in Van den Wyngaert, 352–3 (tr. Dawson, The Mongol
mission, 229); and see n. 58 above.
87 Francis Woodman Cleaves, 'A Chinese source bearing on Marco Polo's departure from China and
a Persian source on his arrival in Persia', Harvard Journal of Asiatic Studies, XXXVI, 1976, 181–
203. Rashid al-Din, al-Tawarikh, ed. A.A.Alizade and tr. A.K.Arends, m (Baku, 1957), text 280
(and see 281), gives no date for the arrival of the embassy from China, though clearly placing it prior
to the winter of 1293–4.
88 Jean Aubin, 'Les princes d'Ormuz du XIIIe au xve siècle', Journal Asiatique, CCXLI, 1953, 88.
89 See generally John K.Fairbank (ed.), The Chinese world order (Cambridge, Mass., 1968). On the
attitudes of the Mongols' predecessors, the Sung emperors, see Herbert Franke, 'Sung embassies:
some general observations', in Morris Rossabi (ed.), China among equals: the Middle Kingdom and
its neighbours, 10th–14th centuries (Berkeley and Los Angeles, 1983), 117; for the post-Yüan era,
Henry Serruys, C.I.C.M., Sino-Mongol relations during the Ming, II: The tribute system and
diplomatic missions (1400–1600) (Mélanges Chinois et Bouddhiques, xiv, Brussels, 1967), 19–21.
90 For Qubilai's reign, see Elizabeth Endicott-West, 'Merchant associations in Yüan China: the Orto
', Asia Major, 3rd series, II (1989), part 2, 127–54; for the pre-Qubilai era, Thomas T. Allsen,
'Mongolian princes and their merchant partners 1200–1260', ibid., 83–126.
91 Denis Sinor, 'Diplomatic practices in medieval Inner Asia', in C.E.Bosworth et al. (ed.), The
Islamic world from Classical to modern times: Essays in honor of Bernard Lewis (Princeton, 1989),
342–3.
92 Critchley, 38–41, suggests that Marco Polo was inflating his own capacities in a bid to secure a
job with the French.
93 Texts conveniently assembled in A.C.Moule, Quinsai and other notes on Marco Polo (Cambridge,
1957), 75–6.
94 Pelliot, Notes, II, 876.
95 A point made by Ronald Latham in his introduction to the Penguin translation (1958), 14, n.
These passages, however, do appear in abbreviated form in the Tuscan version (Ruggieri, 230–1).
96 Francis A.Rouleau, S.J., 'The Yangchow Latin tombstone as a landmark of medieval Christianity
in China', Harvard Journal of Asiatic Studies, XVII, 1954, 346–65.
97 Igor de Rachewiltz, 'Turks in China under the Mongols: a preliminary investigation of Turco-
Mongol relations in the 13th and 14th centuries', in Rossabi (ed.), China among equals, 281–310.
Rossabi, 'The Muslims in the early Yüan dynasty'.
98 Olschki, Marco Polo's Asia, 121–4.
99Rubruck, 'Itinerarium', xxix, 7–13, in Van den Wyngaert, 253–6; (tr.) Peter Jackson and David
Morgan, The mission of Friar William of Rubruck (Hakluyt Society, 2nd series, 173, London, 1990),
184–7.
100 Ch.Kohler and C.V.Langlois (ed.), 'Lettres inédites concernant les croisades (1275–1307)',
Bibliothèque de l'École des Chartes, LII (1891), 57.
101 Karl Jahn (ed.), Die Frankengeschichte des Rašid ad-Din (2nd ed., Vienna, 1977), Persian text.
Tafel 45, German transl., 53.
102Richard, 'Isol le Pisan', 188–90. For Ghazan's brief reoccupation of Syria and Palestine, see
Sylvia Schein, 'Gesta Dei per Mongolos 1300: the genesis of a non-event', English Historical Review,
XCIV, 1979, 805–19 (especially 815 ff.).