OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

India in the Persian World of Letters

OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

OXFORD ORIENTAL MONOGRAPHS

This series of monographs from the Faculty of Oriental Studies, University of
Oxford, makes available the results of recent research by scholars connected
with the Faculty. Its range of subject matter includes language, literature,
thought, history, and art; its geographical scope extends from the
Mediterranean and Caucasus to East Asia. The emphasis is more
on specialist studies than on works of a general nature.

Editorial Board
Professor Julia Bray, Laudian Professorial Fellow in Arabic
Dr Dominic Brookshaw, Associate Professor of Persian Literature
Dr Elizabeth Frood, Associate Professor of Egyptology
Professor Henrietta Harrison, Professor of Modern Chinese Studies
Professor Christopher Minkowski, Boden Professor of Sanskrit
Professor Alison G. Salvesen, University Research Lecturer in Hebrew
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

India in the Persian

World of Letters
Ḳhān-i Ārzū among the Eighteenth-
Century Philologists
ARTHUR DUDNEY

1
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

3
Great Clarendon Street, Oxford, OX2 6DP,
United Kingdom
Oxford University Press is a department of the University of Oxford.
It furthers the University’s objective of excellence in research, scholarship,
and education by publishing worldwide. Oxford is a registered trade mark of
Oxford University Press in the UK and in certain other countries
© Arthur Dudney 2022
The moral rights of the author have been asserted
First Edition published in 2022
Impression: 1
Some rights reserved. No part of this publication may be reproduced, stored in
a retrieval system, or transmitted, in any form or by any means, for commercial purposes,
without the prior permission in writing of Oxford University Press, or as expressly
permitted by law, by licence or under terms agreed with the appropriate
reprographics rights organization.

This is an open access publication, available online and distributed under the terms of a
Creative Commons Attribution – Non Commercial – No Derivatives 4.0
International licence (CC BY-NC-ND 4.0), a copy of which is available at
http://creativecommons.org/licenses/by-nc-nd/4.0/.
Enquiries concerning reproduction outside the scope of this licence
should be sent to the Rights Department, Oxford University Press, at the address above
Published in the United States of America by Oxford University Press
198 Madison Avenue, New York, NY 10016, United States of America
British Library Cataloguing in Publication Data
Data available
Library of Congress Control Number: 2021948779
ISBN 978–0–19–285741–5
DOI: 10.1093/oso/9780192857415.001.0001
Printed and bound in the UK by
TJ Books Limited
Links to third party websites are provided by Oxford in good faith and
for information only. Oxford disclaims any responsibility for the materials
contained in any third party website referenced in this work.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

For those who have realized that “there is no document of civilization

which is not at the same time a document of barbarism”
and are working to do something about it.
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

Acknowledgments

This study is a revision of a PhD thesis submitted in 2013. One of the

delights of revising it over such a long period was stripping out the
claims—some merited and some the inevitable result of a student’s
naivete—that various aspects of my argument were completely new.
Several excellent scholars have converged on questions that I also address
in this study and there have been rapid advances in “Ārzū Studies” and
adjacent disciplines in that time. I have done my best to integrate
relevant new scholarship, but I am acutely aware of the gaps that remain.
In any case, “Above everyone who has knowledge there is the One who is
all knowing” [wa-fauqa kulli żī ʿilmin ʿalīm] (Qurʾān 12:76). The manu-
script was completed during the coronavirus pandemic, which meant a
strange exile from the library exactly when I would have preferred to be
around books rather than my jumble of PDFs.
I am grateful to the Mellon Foundation and Leverhulme Trust for
funding my Mellon Postdoctoral Fellowship in Early Modern Indian
Cultures of Knowledge (Faculty of Oriental Studies, Oxford University,
2013–15) and Leverhulme Early Career Fellowship (Faculty of Asian and
Middle Eastern Studies, Cambridge University, 2015–18, ECF-2015-
057). Chapter 4 in particular is based on material I gathered when
I was a Mellon fellow. I am also grateful to my post-academic employer,
the Arcadia Fund, for the sudden career stability that allowed me to
finish this study, and to its chief executives Anthea Case and Simon
Chaplin. I am especially grateful to Arcadia’s founders, Lisbet Rausing
and Peter Baldwin, whose generosity has enabled this volume to be made
available open access.
On the subject of open access, the cover image for this volume, a
striking illustration of a school scene from a deluxe edition of Amīr
Ḳhusrau’s Ḳhamsah (Quintet), deserves special mention. The manu-
script was likely made for the emperor Akbar in Lahore in the late
sixteenth century and is now held in the Walters Art Museum
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

viii 

(Baltimore, Maryland) as manuscript W.624. The Walters has an ambi-

tion to make images of most of its collection, including the manuscripts,
available through a CC0 license, which waives all copyright and allows
anyone to use the high-resolution images for free. This commitment to
open access is foresighted and brave since image reproduction charges
are such an entrenched revenue stream for museums.
I thanked seventy-six people in the PhD thesis, and I thank them all
again here. People who helped subsequently, investing significant time,
care, or both, include Abhishek Kaicker, S. Akbar Hyder, Alexander
Jabbari, Amira Bennison, Anand Venkatkrishnan, Ananya Chakravarti,
Aria Fani, Barbara Metcalf, Christopher Bahl, David Lunn, Ed Weech,
Frances Pritchett, Francesca Orsini, Jack Hawley, Jill Ross, Katherine
Butler Schofield, Kevin Schwartz, Mana Kia, Mihir Srivastava, Nathan
Tabor, Neeraja Poddar, Nick McBurney, Nur Sobers-Khan, Owen
Cornwall, Oxford University Press’s anonymous reviewers, Pasha
Khan, Raha Rafii, Roy Fischel, C. Ryan Perkins, Shireen Hamza, Stefano
Pellò, Sudev Sheth, and Walt Hakala. I thank my postdoctoral mentors at
Oxford and Cambridge: Charles Melville, Christopher Minkowski,
Edmund Herzig, and Faisal Devji. I will always fondly remember
Allison Busch (1969–2018) and Shamsur Rahman Faruqi (1935–2020),
two guides whose radiance still lights our path after their departure.
Hannah Archambault deserves particular thanks as the ideal proof-
reader and editor for this volume. Her attention to detail and her
encouragement greatly improved the manuscript.
Lastly, I am grateful to Sahil, for his companionship on this and other
journeys.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

Contents

Abbreviations of Ārzū’s Major Works xi

Notes on Conventions Used xiii
Transliteration xv
Introduction 1
The Cosmopolitan and the Classic 6
On Method, the West, and the Non-West 9
Where This Study Will Take You 13
1. A Literate Life: Placing Ārzū and His Works in Their
Social Context 15
Becoming Ḳhān-i Ārzū 18
Early Life and Family 18
Delhi 22
Final Years in Lucknow 22
Ārzū’s People: Friends, Patrons, and Rivals 23
Ānand Rām Mukhli : s: 25
Tek

Chand Bahār 27
The Ticket to Success: Mīrzā ʿAbd al-Qādir Bedil 29
Other Poets in Ārzū’s Network 32
Patrons 35
Rivals and Enemies 37
Ārzū’s Works 45
Conclusion 52
2. Ārzū’s Fruitful Theory of Language 54
The Fruit of the Fruitful Tree 58
A Brief History of Arabic Philology 59
The History of Persian Literary Culture as Described in Mus̄mir 62
Phonetics, Vocabulary, and Regional Variation 74
Connections between Languages 80
Three Kinds of Connections 81
Figurative Language and Where Meaning Comes From 88
Conclusion 93
3. Innovation and Poetic Authority in Eighteenth-Century
Persian 94
The “Indian Style” and India 101
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

x 

The Texts: Munīr’s Critique and Ārzū’s Responses 114

Dād-i Suk: han’s Prefaces 117
The Commentary and Intertextuality 124
Hazīn’s

Critique and Ārzū’s Response 126
Ārzū and the Persian Cosmopolis 129
4. Dictionaries Delimiting Literary Language 142
Mughal Lexicography before Ārzū 144
Reading Sirāj al-Luġhat and Chirāġh-i Hidāyat together 149
Other Major Eighteenth-Century Persian Dictionaries 154
Ārzū’s Vernacular Lexicography 161
Observations on Indian Religion as “Proto-Anthropology” 176
Later Persian Lexicography in the Sub-Continent 184
5. Building a Vernacular Culture on the Ruins of Persian? 187
Beginning in the Middle: How Āb-i Hayāt

Presents the
Eighteenth Century 193
:
A Who’s Who of the People of Rekhtah 198
Shāh Hātim

(1699–1783) and His “Contemporaries” 201
:
Mīr Muhammad Taqī Mīr (1722–1810) 203
:
Rekhtah in Majmaʿ al-Nafāʾis 206
Defining the (Literary) Vernacular 208
Hindī outside of Delhi and the Language of Delhi 222
Urdu and the Everyday 224
Revisiting the Question of the “Unprivileged Power” of
Indo-Persian 232
6. How Language Actually Works: Contrasting Europe
and the Non-West 235
Language and Early Modern Thought 238
Ārzū’s Philology and Its Possible (But Unlikely) Influence
on European Philology 245
Ancients and Moderns in India and Europe 250
Europe and Vernacular Politics: The Vernacular as Modernity? 256
The Pre-Colonial Language Economy 260
Multilingualism and the Individual 263
Relevant and Irrelevant Linguistic Distinctions for Pre-Colonial
South Asia 266
The Social Mechanisms for Defining Language 270
“Imagine There’s No Countries” 275
Conclusion 278
Bibliography 283
Index 319
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

Abbreviations of Ārzū’s Major Works

ʿAK ʿAt̤ īyah-i Kubra̍

CH Chirāġh-i Hidāyat
DS Dād-i Suk: han
ḲhG Ḳhiyābān-i Gulistān
M Mus̄mir
MN Majmaʿ al-Nafāʾis
MʿU Mauhibat-i ʿUz̤ ma̍
NA Nawādir al-Alfāz̤
SL Sirāj al-Luġhat
SM Sirāj-i Munīr
TĠh Tanbīh al-Ġhāfilīn
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

Notes on Conventions Used

Instead of “Hindi,” this study uses the term “hindī” in pre-colonial

contexts to remind us that the Indic language(s) generally called
“hindī” in the Persian sources should not be confused with today’s
Modern Standard Hindi or uncritically contrasted with Urdu. As late
as 1832, the poet Mirzā Ġhālib’s preface to a volume of what we would
call his Urdu poems refers to them as “hindī.” Even later, Munshī
Bhagwant Rāʾī Kākorwī’s Nal Daman Hindī (1859, reprinted 1869)
uses the word in that sense in its title (see Alam and Subrahmanyam
2012, 205n3).
For clarity, bait is always translated as “couplet” even though the form
should technically be called a “distich” when the two lines do not rhyme.
Years are generally cited in both the sources’ Islamic lunar date () as
well as Common Era (). Converting years according to a formula
leaves a range of error  one year unless the exact date is known because
the Islamic lunar calendar cycles over the solar calendar. The dating
systems are generally obvious from the context (namely that tenth- to
twelfth-century dates are Islamic while seventeenth- to nineteenth-
century dates are Western), and ambiguous dates are marked.
Translations from Arabic, Hindi, Persian, Urdu, and European lan-
guages are mine unless noted. The principal exceptions are passages
from Āzād’s Āb-i Hayāt,
which are cited by page number from the
1907 Urdu edition but with English text provided from Frances
Pritchett and Shamsur Rahman Faruqi’s translation. Quranic passages
are drawn from the M. A. S. Abdel Haleem translation.
The crucial term tażkirah has no good English equivalent so it appears
in the original throughout. The name is derived from the Arabic root
ż-k-r which has to do with remembering, including remembering people
for the purpose of praising them. In Persian and Urdu, tażkirah in the
context of literary communities refers to a book containing a number of
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

xiv    

entries on individual poets, including in most cases anecdotes about

them and selections of their poetry. It has often been rendered into
English as “biographical dictionary” but both parts of that translation
are misleading: “Biographical” obscures the implied purpose of the work,
namely preserving the memory of the poets of the tradition and their
works, while “dictionary” suggests that it is about words rather than
people and is organized alphabetically, which is not necessarily the case.
Especially in Arabic, tażkirah can also mean a “memoir” which refers to
a different sort of work. It has this sense in the title of Shaikh :
:
Muhammad ʿAlī Hazīn’s

autobiography Tażkirat al-A :
hwāl [Tażkirah
of (My) Circumstances]. The rendering of “tażkirah” as “memorative
communication” (as in Hermansen and Lawrence 2000, 150), while
undoubtedly more correct than other options, is too unwieldy for our
purposes.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

Transliteration

Marked or Ambiguous Consonants

t: ‫ﭦ‬ ट :
kh ‫ﺥ‬ sh ‫ﺵ‬ श
t:h ‫ﭨﮫ‬ ठ d: ‫ﮈ‬ ड s: ¹ ‫ﺹ‬
s̄ ‫ﺙ‬ :
dh ‫ﮈﮪ‬ ढ z
‫ﺽ‬
j ‫ﺝ‬ ज ż ‫ﺫ‬ t̤ ‫ﻁ‬
jh ‫ﺟﮫ‬ झ r: ‫ﮌ‬ ड़ z̤ ‫ﻅ‬
ch ‫ﭺ‬ च r: h ‫ﮌﮪ‬ ढ़ ʿ ‫ﻉ‬
chh ‫ﭼﮫ‬ छ z̄ ‫ﮊ‬ ġh ‫ﻍ‬
h: ‫ﺡ‬ s ‫ﺱ‬ स q ‫ﻕ‬

Vowels and special marks

ʾ ‫ء‬ (hamza)
y/ī/e/ai ‫ﻯ‬ “ye” as semi-vowel and vowel (maʿrūf and majˈhūl)
w/ū/o/au ‫ﻭ‬ “wāw” as semi-vowel and vowel (maʿrūf and majˈhūl)
r̥ ऋ the vocalic “r” which cannot be specifically represented
in Perso-Arabic script but appears here in the translit-
eration of some Sanskrit words
a̍ ‫ﯼ‬ :
alif maqsūrah (a word-final “ye” pronounced as/ā/that
appears in some Arabic words)
w: ‫ﻭ‬ quiescent (unpronounced) wāw as in kh : wāndan
:
ṁ ‫ﮞ‬ nūn-ġhunnah (nasalization of the preceding vowel)
an
‫ﺍﹰ‬ tanwīn
ˈ a sign marking the rare cases in which “h” follows a
consonant that could be aspirated but in which the two
are actually separate sounds (e.g. majˈhūl)

¹ Nota bene that in Devanagari sources this transliterates the letter “ष” (as in “bhās:ā”), which
is unrelated to the Perso-Arabic letter “‫ﺹ‬.”
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

xvi 

There is something here for everyone to dislike, but this system appears
to be the best compromise between accurately representing orthography
and approximating pronunciation for the primary research languages of
this study, namely Persian and Hindi/Urdu. Furthermore, it can com-
pletely represent all the letters in Arabic, from which the other languages
in question have heavily borrowed. The great contradiction in scholarly
attention to diacritical marks is that people who know the words already
do not generally need them, and people who do not know the words are
bewildered by the infestation of dots marching across the page—the
present work takes a completist approach and attempts to show all
diacritical marks (except in place names and the names of modern
scholars).
The majˈhūl vowels [“o” and “e”] have been lost in standard modern
dialect of Tehran (the collapse of the vowels probably started in the
seventeenth century) but are preserved in most eastern Persian dialects,
including Indo-Persian, and are therefore marked here (on present
conditions see, for example, Henderson 1975; Hodge 1957; in a historical
context Baevskii 2007, 163).
The unpronounced “he” at the end of some words is marked as [ah] in
both Urdu and Persian sources. Thus, for example, the word ‫( ﺯﻧﺪﮦ‬living)
will appear as “zindah.” The izāfat
construction is written [-i] in both
Persian and in Urdu, as in the famous diamond “koh-i nūr.”
The Arabic prefix “al-,” which assimilates to the following consonant
when it is one of the so-called “sun letters,” has been represented here as
it is written rather than as it is pronounced. Thus, for example, the name
“Sirāj al-Dīn” is pronounced [sirāj ud-dīn]. The Arabic tāʾ marbūt̤ ah (‫)ﺓ‬
is not specially marked and is transliterated as “-t” or “-h” as context
requires. Furthermore, the Arabic case inflections reflected in the pro-
nunciation are not noted except in direct quotations from Arabic.
Devanagari sources follow the transliteration scheme of the Oxford
Hindi-English Dictionary, except that “w” has generally been preferred
to “v.”
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

Introduction

In the eighteenth century, the lamp of intellectual achievement in Persian

literature and scholarship burned brightly in India even as the glory of
the Mughal Empire’s political order faded. In recent decades, scholars
have reexamined the narrative of late Mughal political decline, and have
nuanced our understanding of the implications of the loss of Delhi’s
centralized control over the provinces.¹ This study is not concerned with
that political history. Instead it focuses on the intersection of literary
culture—especially the scholarly apparatus behind it—with social and
political thought. How did the cultured elite who staffed the imperial
bureaucracy make sense of the world through poetry and belles lettres?
The research presented here is intellectual history and biography focused
on that specific question, namely reconstructing how people critically
analyzed and used literary language. This study is not a literary history,
but rather a history of thought as expressed through language that is
largely “imaginative” (such as poetry) or in service of imaginative lan-
guage such as critical and analytical writing.²
Why was there an efflorescence of Persian literary and philological
scholarship in Delhi in the first half of the eighteenth century? Why was
Persianized hindī—what would come to be known as Urdu—emerging
as a literary language (for poetry if not yet for prose) without becoming
an administrative language until a century later? How was India [Hind,

¹ See, for example, Marshall 2003; Alam and Subrahmanyam 1999; Bayly 1993, 35ff.; Alam
1986; Wink 1986.
² I use the word “imaginative” here in the sense of language that creates the possibility of
meaning rather than establishing literal fact. I follow Northrup Frye in my understanding of this
distinction: “In literature the standards of outward meaning are secondary, for literary works do
not pretend to describe or assert, and hence are not true, not false, and yet not tautological
either, or at least not in the sense in which such a statement as ‘the good is better than the bad’ is
tautological. Literary meaning may best be described, perhaps, as hypothetical, and a hypothet-
ical or assumed relation to the external world is part of what is usually meant by the word
‘imaginative’ ” (Frye 2000, 74).

:
India in the Persian World of Letters: Khān-i Ārzū among the Eighteenth-Century Philologists. Arthur Dudney,
Oxford University Press. © Arthur Dudney 2022. DOI: 10.1093/oso/9780192857415.003.0001
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

2       

or more narrowly, Hindūstān] constructed as a cultural space, and how

was it linked to the wider Persian-using cosmopolis that stretched across
South and Central Asia from the Ottoman Balkans in the west to the
Bay of Bengal and frontiers of China in the east? These are enormous
questions, each one deserving of a detailed historical study, and my scope
here is significantly more modest: This study approaches these three
fields through a particular inflection point in each of them, the career of
the poet-philologist Sirāj al-Dīn ʿAlī Khān
: (1688–1756), commonly
:
known by his pen-name [takhallu : Ārzū.³ Ārzū is at the center of all
s],
of these transformations because he was a great Indian Persianist, an
important early teacher of Urdu composition in Delhi, and a keen
cultural observer whose ability to situate India in the rest of the
Persianate world was probably unparalleled in his time. By circumscrib-
ing this research so that it focuses on the thought of one remarkable
individual, I hope to give others an anchor for larger questions about the
early modern South Asian world.
Ārzū’s work is an important entry point into many historical ques-
tions. It is lamentable but not surprising that there is no monograph in
English about him, when Sir William Jones (an eighteenth-century
British philologist to whom he has been compared) has been the subject
of dozens of recent books and articles.⁴ Furthermore, why, given that
Ārzū’s works set a standard for meticulously detailed analysis, were all
but a couple of them ignored by colonial-period scholarship? This
colonial connection is worth noting, if only in passing, as part of what
Mohamad Tavakoli-Targhi has called the “genesis amnesia” of Oriental
studies, namely that not enough attention has been paid to why certain
non-Western texts and not others became part of the West’s scholarly
canon.⁵ The narrative of the eighteenth century as understood by

³ “Sirāj al-Dīn” means “Lamp of the Faith” and is probably an homage to Ārzū’s famous
ancestor, the Sufi master Chirāġh-i Dihlī [“Lamp of Delhi”] (d. 791/1389). (Sirāj is the Arabic
cognate of the Persian word chirāġh.) Likewise, Ārzū’s student Lālah Tek

Chand, known by his
pen-name Bahār, refers to his teacher as “Sirāj al-muhaqqiqīn”
: [Lamp of the Scholars] and
“Sirāj al-shuʿarā” [Lamp of the Poets] (Bahār-i ʿAjam 2001, preface). Khān
: is a title ultimately
derived from the Mongol tradition.
⁴ In 2015, Habib University (Karachi, Pakistan) recognized Ārzū’s importance by inaugurat-
ing a center named for him, the Arzu Center for Regional Languages and Humanities.
⁵ Tavakoli-Targhi 2001.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

 3

Europeans, as John Dowson admits in The History of India as Told

by Its Own Historians (1867–77), is based overwhelmingly on the
most “English” of the Persian histories, Ġhulām Husain’s
Siyar al-
Mutaʾakh : :
khirīn [Deeds of the Moderns, 1783], to the neglect of the
other materials available. Dowson writes: “In fact, the native side of the
history of Ghulám Husain’s days, as it appears in the works of English
writers, rests almost entirely upon his authority,” that is, as told in
Siyar al-Mutaʾakh : khirīn.⁶
: Both James Mill and Lord Macaulay, two key
architects of colonial historiography, approved of this work, and its
adoption was no doubt helped by being translated into English in 1789
before other relevant texts were available in European languages. Why
does engaging with the development of the colonial canon matter
for pre-colonial history? The master narratives that still shape our
understanding—since, after all, many of the translations gathered in The
History of India as Told by Its Own Historians have not been superseded
despite the collection’s obvious flaws—are based on a colonial engagement
with the pre-colonial past. In many cases, this colonial discourse has
silently displaced the tradition’s own narratives, so that ideas about Indic
language and literature presented as “culturally authentic” in fact have a
much newer origin than is generally believed.⁷ My goal here is to strip
away the accretions of interpretation and historical happenstance (that
one text was translated into English before another, for example), and
face the eighteenth-century Indo-Persian milieu on terms that would
have been more familiar to its participants. This involves reconstructing
the circumstances under which philological knowledge was created
and put into practice.⁸ In part, such an analysis requires understanding
pre-colonial India—and indeed the whole of the Persian cosmopolis—as
a multilingual zone, including the recognition that even within Iran the

⁶ Elliot and Dowson 1877, 194ff. In the 1770s, a period in which British colonial scholars
were particularly interested in history, these scholars tended to consult recently prepared
Persian historical summaries instead of the original chronicles (Teissier 2009, 141–2).
⁷ Allison Busch has made a similar case about recent Hindi literary criticism: She observes
that “the voice of the postcolonial speaks in English, whereas the Orientalist voice is still alive,
and speaking in Hindi” (Busch 2011, 14).
⁸ As Roger Chartier has written: “The historian’s task is thus to reconstruct the variations
that differentiate the espaces lisibles—that is, the texts in their discursive and material forms—
and those that govern the circumstances of their effectuation—that is, the readings, understood
as concrete practices and as procedures” (Chartier 1994, 2).
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

4       

cosmopolitan poetic language co-existed with unstandardized local

dialects of Persian. The present analysis begins with the assumption
that everything we think we know about language ideology, particularly
how languages build nations, must be rigorously tested before it can be
applied to the early modern period.
Ārzū was a significant and influential thinker even if few non-
specialists know his name today. He settled in Delhi near the beginning
:
of the emperor Muhammad Shāh’s reign (probably in early 1720) and
had a successful career for just over thirty years in the capital before
taking up a post in Lucknow and dying there soon after. He was survived
by two generations of students and by his works, which included com-
mentaries, a body of Persian poetry, three dictionaries, a tażkirah, and a
treatise on language called Mus̄mir [lit. “fruitful”], which is the subject of
Chapter 2. It is notable that besides the sheer volume and range of his
scholarly and creative works, his colleagues and students were among the
most important Persian scholars and Urdu poets of their generations. He
is regarded as one of the first intellectuals to take Urdu literature
seriously (although it should be noted that not a single line of Urdu
survives that is unquestionably his work) but was also arguably the
greatest Indian Persianist of his day. He was referred to as a “marjaʿ”
[a refuge or point of reference] and as “Lamp of the Researchers” [sirāj
:
al-muhaqqiqīn] and was at the center of most of the famous literary
debates of his time. As a scholar, he drew upon an unusually broad range
of research materials and correspondingly employed an encyclopedic
critical approach that compared numerous sources. He drew deeply
from the resources of the Arabic tradition, for example basing his
Mus̄mir on al-Muzhir fī ʿUlūm al-Luġhah wa Anwāʿihā [The Luminous
Work Concerning the Sciences of Language and Its Subfields] written by
the fifteenth-century Egyptian polymath Jalāl al-Dīn al-Suyūt̤ ī, while
also having easy access to indigenous Indic traditions. Ārzū was a keen,
detail-obsessed observer of society and tradition who sought to correct
cultural misconceptions, whether those held by Indians or non-Indians,
and to record regional usages faithfully, even if he found them inelo-
quent or inappropriate for poetry. However, his ethnography, if we can
call it that, was based on his interest in language, which was in turn
focused on the practice of poetry. Since he does not fit easily into any of
our contemporary disciplines, I will refer to him as a “philologist.”
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

 5

The term “philology” has seen its stock rise and fall since it was coined
in late Antiquity. After hitting rock bottom during the postmodern turn
of the 1980s and 1990s,⁹ there has been recent scholarly interest in
reclaiming the term.¹⁰ Philology literally means, of course, “the love of
words,” but throughout history it has referred specifically to ways of
studying language and literature. According to the Oxford English
Dictionary, it has meant variously a devotion to books, the careful
study of literature (especially Greco-Roman Classical texts), and the
systematic study of language in order to discover its structures.¹¹ In
the last sense, it came to be synonymous with “historical linguistics,”
although this usage is now rare. So why choose as the key term in this
study one which has little resonance for people today? Precisely because
it is capacious and ambiguous: It is a word that forces us to think
historically. For example, we might be tempted to call Ārzū a linguist,
but he certainly would not get hired in a modern university’s
Department of Linguistics, whose members have entirely different pre-
suppositions about the nature of language and its relationship to litera-
ture. Avoiding the term “linguistics” prevents the false assumption that
for someone like Ārzū to be worth our attention, he must have had the
remarkable foresight—before there was such a thing as “linguistics”—to
have thought about the development of language in terms that are
familiar to us. (In fact, his works demonstrate a sophisticated sense
of the historicity of language, but one that is quite unlike ours.)
Furthermore, linguistics is narrowly focused on language as its object,
but philology acts upon a body of texts (rather than on a set of linguistic

⁹ Through a process of sloppy synecdoche, philology has been called to account by post-
modern philosophers like Paul de Man for the sins of all the modern human sciences and even
blamed for the rise of fascism (Pollock 2009). The translator of Ulrich von Wilamowitz-
Moellendorff ’s Geschichte der Philologie laments having had to title his translation History of
Classical Scholarship for the sake of clarity because as far as the name philology is concerned “it is
deplorable that we in England have ceased to use this valuable term correctly” (Wilamowitz-
Moellendorff 1982, vii). For an influential (and famously broad) early eighteenth-century
definition of the term as proposed by Giambattista Vico (1688–1744), see Burke 1985, 84 and
Manson 1969, 46.
¹⁰ See, for example, Philology (Turner 2014), World Philology (Pollock et al. 2015), and the
journal Philological Encounters, which has been published since 2016.
¹¹ It has also, from early on, had a pejorative sense in English as the Oxford English
Dictionary attests, e.g., Henry Cockeram’s The English Dictionarie, or an Interpreter of Hard
English Words of 1623 gives “loue of much babbling” as the definition for “phylologie.” Two
other entries for this sense contrasted philology unfavorably with philosophy since philology (by
this definition) is the study of mere words while philosophy gets at truth directly.
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

6       

data), and indeed Ārzū and his peers generally made their living or their
reputation or both as practicing poets. No one, it goes without saying,
expects linguists to be poets today. The idea that a language can be
abstracted from a literary tradition is therefore totally alien to Ārzū’s
scholarship. It is also important to dispense with the presumption that
composing literature or engaging with it was a leisure-time activity, a
domain of play separate from serious business. While mocking bad poets
is a staple of Persianate literature, there is no sense (before the colonial
encounter) that reflecting on language or composing verse are inherently
indolent activities. The Islamicate tradition never gave up on Aristotle’s
famous formulation that poetry “is a more philosophical and a higher
thing than history: for poetry tends to express the universal, history the
particular.”¹² In seeking to position aspects of Ārzū’s life-world as sep-
arate from basic features of our own experience, whether in the present-
day West or elsewhere, the term “philologist” throws us a lifeline since it
encompasses “literary scholar,” “linguist,” and “litterateur”—all of which
accurately but incompletely describe Ārzū.¹³

The Cosmopolitan and the Classic

This study will return often to the concept of the cosmopolitan, taking as
given that for centuries there existed a Persian cosmopolis stretching
from the Balkans in the west across Central and South Asia to the Bay of
Bengal and Chinese frontier territory of Xinjiang in the east. This vast
region was the zone where Persian language and literature, and its
corresponding ethical and political culture, was common currency, just

¹² Poetics 1451b, trans. Butcher. Indeed, some of the greats of the tradition, like the tażkirah
:
compiler Muhammad ʿAufī (died c. 1232–3 ) and the poet Amīr Khusrau,
: described the craft
of poetry as ʿilm [learning, often associated with religious studies], and Khusrau
: even went so far
as to compare it to divine revelation (Keshavmurthy 2011; Gabbay 2010, 27ff.).
¹³ However, it should be made clear from the outset that there is no single pre-modern
Persian term that captures an equivalent meaning “philologist” because someone like Ārzū
could be slotted into perhaps a dozen or more common categories such as ahl-i qalam [people of
the pen] or suk: han-dānān [knowers of speech/poetry]. He was an adīb, that is someone who
dealt with adab, but while this term contains the sense of philology (see Bonebakker 1990),
nothing distinguishes general poetic practice from inquiry into the nature of language. It is
worth also problematizing the applicability of the term “literature,” which has no exact analogue
in pre-modern Persian since adabiyāt is a nineteenth-century Turkish coinage (De Bruijn 2009).
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

 7

as Latin and Latinity was a “European sign” that maintained a similar

cultural zone in the West.¹⁴ Sanskrit, of course, linked South and
Southeast Asia in the same way for centuries. Part of the appeal of the
cosmopolitan as an object of analysis is precisely that it has no long
scholarly genealogy in contrast to its obvious importance throughout
world history.¹⁵ It is a real but nonetheless difficult-to-locate phenome-
non that can serve as a historiographical bulwark against the nationalist
thinking that is often the default mode for modern people, and which
distorts our understanding of the past. A cosmopolis is not merely
a linguistic zone—for example, where most elites are educated in
Persian—but a textually constituted entity. A textual canon existed across
the whole of the Persian cosmopolis, with some texts virtually universal,
such as the works of the thirteenth-century writer Saʿdī, and others less
common in circulation. Indeed, many texts were written in the cosmo-
politan language but never left the place where they were written. The
texts that make a cosmopolis are not constitutive of a national language
but are anchored in a tradition that cannot be mapped neatly onto a
single extant political unit even if there is a sense of where it originated.
The texts can be plausibly claimed by different groups in forming their
identities. All of the disparate regions in the Persian cosmopolis shared
an appreciation, for example, for Firdausī’s epic Shāhnāmah [Book of
Kings, c. 1000 ]—today claimed as the text that created an Iranian
national identity but in fact composed in present-day Afghanistan on the
Indian frontier—just as the Rāmāyana : and Mahābhārata were touch-
stones across the Sanskrit cosmopolis and localized as required. In the
case of the Persian cosmopolis, the early modern Mughals, Ottomans,
and Safavids—the three great empires of the cosmopolis—all used sim-
ilar curricula largely derived from thirteenth- and fourteenth-century
Central Asia and Iran.¹⁶ The lyric poetry of Hāfiz̤

and ʿAt̤ t̤ ār (to pick two
influential examples) was quoted from Istanbul to Khotan, from
Tashkent to Hyderabad, as was the political philosophy of Niz̤ āmī
ʿArūzī.

The works of Saʿdī were probably read by every single child in
the world educated in the Persian medium for more than half a
millennium.

¹⁴ Waquet 2001, 121. ¹⁵ Pollock et al. 2000, 577. ¹⁶ Robinson 1997, 154.
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

8       

By framing this analysis around cosmopolitanism, I am also necessarily

invoking something that can be called “the classical,” which like “philology”
requires quite a bit of explanation in today’s world. We are perhaps more
familiar with the concept of “classical” as the textual “canon,” a substituted
term that is deliberately not as explicitly honorific as “classical,” but still
denotes a body of texts elevated above others that provided an identity
for a literate elite.¹⁷ This identity based on the mastery of these classical
texts, an identity that could certainly have local inflections but was by
definition cosmopolitan, was on par with or more important than what
we consider standard identity markers today (nation, language, class, and
so on). Not only did the classics build pre-modern elite communities, but
for us they flag important historical moments: When an attempt is made
to redefine which texts are classical and how critics should approach
them, it can be an illuminating historical discontinuity, such as when
Renaissance Latinists became confident enough to question ancient
editors of classical texts.¹⁸ While the classical canon is inevitably repre-
sented as static, even in many cases divinely ordained, the reality is that it
is contested and constructed. The South African novelist J. M. Coetzee
observes that,

The classic defines itself by surviving. Therefore, the interrogation of

the classic, no matter how hostile, is part of the history of the classic,
inevitable and even to be welcomed. For as long as the classic needs to
be protected from attack, it can never prove itself classic.¹⁹

In making this provocative claim, Coetzee is drawing upon a contrast

between the classic as an object of veneration versus that of the classic as
something to be questioned, even derided as a force that retards progress.
Such tension appears as early as the first century  in the works of
the Roman grammarian Quintilian, who is ambivalent about Quintus
Ennius, an early Latin poet who was by then already outmoded: “Ennius
we must venerate as we do groves whose age makes them holy, full of
great oaks that nowadays have less beauty than sanctity.”²⁰

¹⁷ Guillory 1993, 6. ¹⁸ D’Amico 1988, 9. ¹⁹ Coetzee 2001, 14.

²⁰ Institutio Oratoria 10.1.85–90, quoted and translated in Steele 1990, 121.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

 9

This study considers Ārzū’s own interventions in the Persian canon:

He attempted to maintain the unity of the Persian cosmopolis through
careful attention to the classics while at the same time broadening the
range of permitted interpretations of Persian texts. A hallmark of early
modernity was this blend of the continued centrality of the classics
alongside the development of new textualized knowledge. The modern
impulse to cast progress and tradition as opposites can obscure the fact
that historically there was no inherent contradiction between venerating
a textual canon and supporting aesthetic development.

On Method, the West, and the Non-West

Before starting to read texts historically, we should posit that historical

study is necessarily a conversation with a text, in which the ways that
we the interpreters pose the questions condition our understanding of
the replies we receive. The present study is guided by Hans-Georg
Gadamer’s dictum that “at the beginning of all historical hermeneutics,
then, the abstract antithesis between tradition and historical research,
between history and the knowledge of it, must be discarded.”²¹ At first
glance, this seems a willfully obfuscatory position—why can we not just
read a historical text and extract information from it? Furthermore, does
this view not presuppose that only emic analyses (that is, ones based on
the tradition’s own assumptions) are legitimate?²² It is obvious that the
purpose of the study of history is gathering knowledge about the past, but
intellectual history, the goal of this project, is not just a matter of
collecting names and dates but of sustained interpretation.²³ For

²¹ Gadamer 2006, 283–4.

²² Walter Mignolo observes that Gadamer’s hermeneutics is ill-equipped to think transcul-
turally because it is “monotopic,” presuming that the interpreter stands within the tradition
being interpreted. This is frequently not the case, especially in colonial situations. Mignolo
proposes a “pluritropic” approach as an alternative “that reflects on the very process of
constructing (e.g., putting in order) that portion of the world to be known” (Mignolo 2003,
15–25).
²³ “History as a science has, as it is known, no epistemological object proper to itself; rather, it
shares this object with all social and human sciences” (Koselleck 2004, 94).
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

10       

historically conscious interpretation, we must situate ourselves vis-à-vis

the text and the “variety of voices in which the echo of the past is heard”
that constitute the tradition we are studying.²⁴ What we call a tradition is
actually a web of relationships between texts, the strands of thought and
rhetoric that interpreters follow from one text to another. A tradition is
not simply a Great Books curriculum and in no way stands outside of
history—traditions are made and remade; traditions die or, in some
cases, are deliberately killed off. Thanks to tradition we cannot approach
any text, even one entirely foreign to us, with a blank slate. Thus, we must
be conscious of our predisposition to reach certain conclusions (what
Gadamer and the hermeneutical tradition call “prejudices”) that shape—
and sometimes deform—our understanding of history. Of course, a
scholar must be able to research a tradition without necessarily using
its categories; a purely emic interpretative project would have no reso-
nance outside the society to which the particular tradition in question
belongs. However, we must recognize that when our study crosses not
only boundaries of time but of culture, our ability to think historically is
constrained not just by the boundaries of tradition but more significantly
by our individual horizons as scholars.²⁵ When one has access to the
tradition in question by means of a university education (as I do in the
case of Urdu and Persian literature) rather than acquiring it through
familial inheritance, this attention to one’s own position in history is
especially important. Treading carefully given all we now know about
cross-cultural power imbalances and epistemic violence, cultural dis-
tance can be a benefit: It allows us to listen closely to voices in a text
that the mainstream of a tradition tends to drown out. It also side-steps
some thorny questions of taste, principally whether a literary work that
was appreciated in its time remains enjoyable to modern-day inheritors

²⁴ Gadamer 2006, 285.

²⁵ For Gadamer, “horizon” [Horizont] is a technical term: “The horizon is the range of vision
that includes everything that can be seen from a particular vantage point . . . A person who has
an [sic] horizon knows the relative significance of everything within this horizon, whether it is
near or far, great or small” (Gadamer 2006, 301). The Perso-Arabic term “hadd”
: [boundary] is
sometimes used in a similar sense. One way in which Gadamer understood historical interpre-
tation was as a “fusion of horizons” [Horizontverschmelzung], in which meaning is found in the
overlap between the historical subject’s horizon and the interpreter’s.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

 11

of the tradition or not.²⁶ Perhaps the assertion that cultural distance is a

possible benefit will strike some as naïve after interventions like Edward
Said’s Orientalism, but Said never argued that outsiders are unable to
assess a tradition fairly—though many uncritical readers have assumed
exactly that—but rather that their predispositions, both institutional and
individual, must be taken into account.²⁷ No one, internal or external to a
tradition, can claim a neutral vantage point because history itself condi-
tions the questions that motivate our inquiry into the past.
We can nuance Gadamer’s historical philosophy by considering
Quentin Skinner’s method: For Skinner, “the understanding of texts . . .
presupposes the grasp of what they were intended to mean and of how
that meaning was intended to be taken.”²⁸ He cautions us against apply-
ing external standards of rationality to historical texts because such an
approach makes us dependent on the degree of “cognitive discomfort”
we feel in guiding us to read something figuratively or literally.²⁹
Attempting to read texts in relation to other contemporary texts and
the constellation of meanings they contain is entirely compatible with
Gadamer’s thought. However, when it comes to the Indo-Persian and
Urdu tradition, there is a deep structural problem in our current pro-
spects for approaching the past. For reasons having to do with colonial-
ism and the nationalist response to it, scholars have been generally

²⁶ On listening to voices outside a tradition’s mainstream, see Gadamer 2006, 297; compare
Foucault 1972, 9 on “discontinuity.” There is a gap in interpretation when a literary work that
delighted people when it was written is no longer enjoyable in the present because of how tastes
have changed. It is a difficult problem because such impressions cannot—and should not even if
they could—be excluded from interpretation, but they are frequently given so much weight that
people in effect blame writers for not writing in a way that resonates with a certain reader
centuries in the future. The Urdu writer and literary reformer Alt̤ āf Husain

Hālī

(1837–1914)
observed something similar in justifying the vast gulf between the new naturalistic poetic style
[nechral shāʿirī] of which he was a proponent, and the classical tradition: He notes that the
Persian poets Z̤ uhūrī, ʿUrfī, T̤ ālib, Asīr and others were great in their time (the sixteenth
century), but tastes have changed and their work no longer appeals to readers, but that does
not diminish their greatness (Yādgār-i Ġhālib 1986, 124).
²⁷ Said approvingly quotes Gramsci’s Prison Notebooks: “The starting point of critical
elaboration is the consciousness of what one really is, and is ‘knowing thyself ’ as a product of
the historical process to date, which has deposited in you an infinity of traces, without leaving an
inventory” (Said 1979, 25). This footnote is not an appropriate vehicle for a wide-ranging
critique of Said’s theories and their potential for misinterpretation and misuse, but Brennan
2006, 115 is a good summary of my misgivings. Westerners are not the only people in history to
produce essentializing ethnographies (e.g., as cited in Subrahmanyam 1997, 761).
²⁸ Skinner 2002, 86. ²⁹ Skinner 2002, 41.
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

12       

unwilling or perhaps unable to attempt anything like a Skinnerian

reconstruction of the intellectual framework of the Delhi literary scene
in the period immediately before colonialism.³⁰ In other words, ques-
tions are not typically posed in order that we might consider what the
authors of the texts might have intended and what that tells us. More
frequently, the texts are asked to demonstrate momentum towards either
enabling colonial rule or an anti-colonial national consciousness. The
historiographical argument in this study is that many of the questions
that have been framed about eighteenth-century Indo-Persian (and by
extension Urdu) literary culture by later scholars would have had little or
no relevance for those living in the period under discussion. The eight-
eenth century has been thoroughly studied and important arguments
made, but especially in the case of intellectual history there has to some
degree been a failure to engage with eighteenth-century writers on their
own terms. Often the eighteenth century was studied with a focus on
economic and political history as a way station to colonial history,³¹ but
recent work on the cultural and intellectual history of the eighteenth
century in South Asia is beginning to rectify this.
Dwelling on historiography, the philosophy of how we conceptualize
and write history is especially important when the historical subject in
question is commonly remembered as a founder, “the First Person to
have done X,” a claim often made about Ārzū (e.g., first important
teacher of Urdu, first critical lexicographer in Urdu, first to see the
historical affinity between Persian and Sanskrit). We are obliged to
remember that historical subjects were not necessarily particularly
invested in having the world turn out the way that we know it did after
their deaths. Thus, we should renounce any goal of showing what Ārzū
got “right” and “wrong” by measuring his insights against modern
linguistics (which in any case would require the implicit assumption
that our approach to analyzing language is right and any other approach

³⁰ An additional obstacle in applying Quentin Skinner’s method in the Indian context is that,
as Jonardon Ganeri observes, trying to understand how a text would have been received by
contemporaries demands specific knowledge about individuals’ circumstances that will almost
always be lacking for pre-colonial India (Ganeri 2011, 64–5).
³¹ For example, Alavi 2002 and Marshall 2003, two edited volumes that focus on eighteenth-
century economic and political history.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

 13

is primitive or wrong). Our thinking about how to understand language

in the abstract is nonetheless informed by modern linguistics, as well as
by the historical analysis and mythmaking on the relationship between
Latin and the vernacular languages in Europe. My only real idée fixe
when I first undertook this project was the recognition of Ārzū as an
inflection point in Indian intellectual history.

Where This Study Will Take You

Chapter 1 explores Ārzū’s life and social milieu. Who were his patrons
and what was his network of colleagues in Delhi and beyond? The source
material is primarily tażkirahs, but includes miscellaneous biographical
statements gleaned from other works (such as prefaces). Chapter 2
addresses Ārzū’s engagement with the long tradition of literary criticism
in Persian and Arabic primarily through a reading of his treatise Mus̄mir.
He combines ideas by diverse thinkers in order to create a new criticism
which is based on research [tahqīq]: and a sophisticated theoretical
understanding of language, including its transformation over time. It
was arguably the most advanced theoretical study of language in the pre-
modern Persian tradition but was superseded by the Western discipline
of linguistics and largely forgotten. Chapter 3 reconsiders the well-
known but simplistic paradigm that Iranians reacted negatively to a
supposedly distinctly Indian, decadent style of Persian literature, which
was later called the “sabk-i hindī” [Indian style]. Instead of this over-
determined and anachronistically nationalist paradigm, eighteenth-
century poets were far more concerned with old versus new styles in
Persian poetry. Ārzū was arguably the first to develop a rigorous criticism
to address poetic innovation associated with the “fresh speech” [tāzah-
goʾī] movement. Chapter 4 discusses lexicography, a genre in which Ārzū
was influential and which can be understood as a textual tool for relating
different parts of the Persian cosmopolis to each other. It considers
the importance of Ārzū’s dictionary of Indic terms, Nawādir al-Ālfāz̤ .
Chapter 5 argues that Ārzū’s theory of language is intentionally capa-
cious so that it can apply to Indian vernacular language as well as
Persian. Ārzū’s crucial intervention in vernacular poetics was to suggest
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

14       

:
that the Indic poetic practice of rekhtah (which later became a synonym
for Urdu) could be standardized along the lines of Persian, and thus that
an Indic language could be a cosmopolitan literary medium in its own
right. Indeed, Ārzū’s defense of “fresh speech” in Persian carries over
into the vernacular sphere through intermediaries and is most likely the
basis for Urdu literature as it came to be. This discussion frames the
larger question of how language is defined today compared to how it was
understood in the pre-colonial period. It advances the argument that
language was classified according to people’s concept of its function (that
is, what linguists would call a “domain”) rather than as a linguistic
identity or through the phylogeny developed in the wake of colonial-
era linguistic surveys. Chapter 6, the final chapter, proposes a compara-
tive method for understanding roughly parallel social transformations
around knowledge and language in early modern times in Europe and
the Persianate World. Why were changing relationships with the canon
generally assumed to be constitutive of modernity in the West and
evidence of aesthetic and intellectual decline in the Persian cosmopolis?
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

1
A Literate Life
Placing Ārzū and His Works in Their Social Context

In Ānand Rām Mukhli : s: ’s dictionary Mirʾāt al-Is::t:ilāh: [The Mirror of

Expressions, 1158/1745], biographical notices appear alongside defini-
tions of certain words and expressions. Thus, under “ārzū” we find:

“Ārzū” has the meaning “hope” and “desire,” and is also the pen-name
:
of Khān s: āh: ib Sirāj al-Dīn ʿAlī Khān
: (God bless him), whom the author
knew intimately for thirty years, free from hypocrisy, in a friendship
greater than words can describe. Now when his illustrious name has
advanced from the pen’s tongue, the manners of affection demand that
I write down a few lines at this juncture and in the writing of an
example of his excellence I complete it as decoration for these pages:
“His circumstances completely like a rose in color and scent: / Sirāj
al-Dīn ʿAlī Khān : Ārzū” The specifics of his praised rank are a deco-
ration to the folios of night and day and the ornament of the page of
time . . . He is a master of the science [ʿilm] of Arabic, lexicography,
and prosody, and the art [fann] of history, music, and hindī.¹

Much of this finely wrought metaphorical language is lost on us, both as

: s: ’s
a matter of interpretation and aesthetic taste, but nonetheless Mukhli
deep respect for Ārzū comes through clearly. He notes that he has been
Ārzū’s close friend for thirty years, and lists a number of Ārzū’s special
qualities, namely knowledge of Arabic, lexicography, prosody, history,

¹ Mirʾāt al-Is::t:ilāh: 2013, 18–20; checked against the British Library’s manuscript, Mirʾāt
al-Is::t:ilāh: 1850, ff. 20a–21a. The flowery rhetoric does not lend itself to easy translation so
only a sample has been provided here. On the text, see Rieu 1879–83, 997. A sometimes faulty
but nonetheless useful English translation of the whole work is Mukhlis 1993. For more
information on the dictionary, see Chapter 4.

:
India in the Persian World of Letters: Khān-i Ārzū among the Eighteenth-Century Philologists. Arthur Dudney,
Oxford University Press. © Arthur Dudney 2022. DOI: 10.1093/oso/9780192857415.003.0002
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

16       

music, and hindī.² The range is impressive, and suggests that the set of
skills that are brought to bear on poetic composition and criticism is
larger than we might expect. History and hindī may seem like outliers,
but, as we see in the following chapters, they were valued by Ārzū and his
: account is filled with
: s’s
circle as philological and aesthetic tools. Mukhli
complex imagery of a lush garden, an extended tribute to Ārzū’s talent as
: s’s
a poet and a claim that his talent has received divine sanction. Mukhli :
remarks are a good place to start in assessing Ārzū’s life, social milieu,
and works.
Although Ārzū’s career has long been widely recognized as important
by scholars and critics, until recently he had more often been invoked as
a symbol than as an object of study.³ He has, for example, been taken as
the father of Urdu poetry but the contours of his intellectual engagement
with vernacular literature have not been traced extensively. Most often,
he has been represented as a patriot defending India against Iranian
chauvinism, and as the person whose innovative theories on linguistic
origins the colonial philologist and jurist Sir William Jones (1746–94)
stole and made European. Without dismissing out of hand these popular
ideas about Ārzū—to the degree that impressions of an eighteenth-
century philologist can be said to be “popular”—this study seeks to
place them not only in their immediate historical context but in a
comparative early modern framework.
The reasons behind Ārzū’s relative obscurity offer us a useful opening
point in explaining the constitution of present-day literary aesthetics and
linguistic knowledge systems that we take for granted. While nineteenth-
century scholars of Persian still engaged with Ārzū and he was highly

² The critical edition (Mirʾāt al-Is::t:ilāh: 2013) prefers the reading from some manuscripts that
he was a master of “Indian music” [mūsīqī-yi hindī] rather than “music and hindī” [mūsīqī wa
hindī]. However, I doubt the critical edition’s reading because it loses the rhetorical symmetry of
describing three things as “ʿilm” and three things as “fann.”
³ No work in English gives a complete account of Ārzū’s life and social context. In Urdu,
however, Professor Rehana Khatoon’s Ah: wāl-o As̄ār-i Khān-i : Ārzū [The Life and Works of
:
Khān-i Ārzū] provides a useful overview of the available source materials (Khatoon 1987). An
account of Ārzū’s teachers, contemporaries, and students is Khatoon 2004b. The editor’s
introduction to Khatoon’s Mus̄mir is the fullest account in English (Khatoon 1991) and a
substantial entry on Ārzū in Encyclopædia Iranica is the most up to date (Keshavmurthy 2012).
An Iranian scholar who has worked extensively on Ārzū, Mahdi Rahimpoor, has observed that
no comprehensive account of Ārzū’s life exists in Persian (Rahimpoor 2008b), which apparently
remains true.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

   17

respected by some—the German-British Orientalist Henry Blochmann,

for example, calls him the “best commentator whom India has
produced”⁴—his achievements have always been bracketed off as partic-
ularly Indian rather than relevant to and part of the Persian tradition as a
whole. He was after all an Indian scholar, not Iranian, and by that time
national identity had replaced more cosmopolitan forms of belonging.
By the twentieth century, scholars in Iran who tried to modernize the
study of Persian literature completely ignored Ārzū.⁵ ʿAlī-Akbar
:
Dihkhudā (1879–1956), the lexicographer and literary critic whose
Luġhatnāmah is the Persian equivalent of a comprehensive lexicograph-
ical project like the Oxford English Dictionary, is one example of such a
modernizer. Luġhatnāmah relies heavily on dictionaries compiled in
India—because lexicography was far more developed there than in
early modern Iran⁶—and yet in a biographical note on Ārzū, Dihkhudā :
makes four mistakes in succession: He misstates Ārzū’s name, falsely
claims Ārzū was an Iranian resident in India, incorrectly cites the title of
one of Ārzū’s works, and misattributes a famous work written by another
Indian to Ārzū.⁷ Ārzū’s historical placement goes some way towards
explaining why such errors have crept into modern historiography: For
an important and well-connected figure, there is surprisingly little source
material available. Although there was an explosion of tażkirah-writing
towards the end of Ārzū’s life continuing in the decades after his death,
besides these tażkirahs (which often repeat the same general information
or by contrast disagree even on basic facts) there are virtually no other
sources available for a biographical study. Because the historically attes-
table biography is so thin, we turn to his circle of friends and enemies,
which is to say the networks he was a part of and opposed, and glean

⁴ Blochmann 1868, 25.

⁵ An important exception is Muhammad Reza Shafiʿi Kadkani (b. 1939), who reads Ārzū
carefully and often sympathetically (Rahimpoor 2008a, 335).
⁶ Discussed at length in Chapter 4.
⁷ Shamisa 2002, 21. The misattributed work is Wārastah’s Mus::t:alah: āt al-Shuʿarā [Idioms of
the Poets], which is described in detail in a discussion of Wārastah further in the chapter.
:
Dihkhudā’s four mistakes appear under the headword “ārzū” (Luġhatnāmah 1994, 66). There is,
however, a separate entry for “k: hān-i ārzū” (Luġhatnāmah 1994, 8266) where Dihkhudā
: notes
simply that Ārzū is an Indian poet whose poems are often cited in Farhang-i Ānandrāj
[Dictionary for Ānand Rāj, 1306/1888–9] the largest Persian dictionary compiled before the
Luġhatnāmah itself. The mistakes are the same in the original fascicles and the revised 1994
edition.
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

18       

what we can from critical works by Ārzū and his contemporaries that
form the bulk of the texts addressed in the present research.

:
Becoming Khān-i Ārzū

Ārzū’s life follows the typical career path for a Mughal man of letters,
from early education in a familial and small-town environment to
advanced education and urban network building to a career in full flower
in the capital. He spent his early years between Gwalior, a backwater that
was nevertheless of some cultural and administrative significance, and
Agra, a former imperial capital that remained one of the empire’s great
cities even after its eclipse by Delhi. An education and poetic appren-
ticeship in Agra prepared him for a career in Delhi, the capital and focus
of cultural life in the early years of the eighteenth century. He spent more
than three fruitful decades in Delhi before moving to Lucknow in middle
age as the increasingly independent rulers of Awadh transformed the
province into a new seat of culture in northern India through lavish
patronage. It was to be a brief stay in Awadh because he only lived two
more years.

Early Life and Family

The basic facts about Ārzū’s early life are unclear because even reputable
sources offer conflicting accounts. He was born either in Gwalior (in
present-day Madhya Pradesh) or in Akbarabad (today’s Agra), 120 kilo-
meters to the north in present-day Uttar Pradesh. In any case, he had
family ties to both places. The date of his birth was almost certainly 1099/
1687–8 but 1101/1689–90 has also been widely accepted. Ārzū’s account
of himself, his own entry in his tażkirah Majmaʿ al-Nafāʾis [Collection of
Subtleties, 1164/1750–1], does not mention either the date or place of his
birth.⁸ In Safīnah-yi Kh: wushgo
: : wushgo’s
[Kh : Notebook], Ārzū’s close
friend and disciple Bindrāban Dās Kh : wushgo
: refers to 1099 , and

⁸ MN 2004, 1:186ff.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

   19

provides Ārzū’s father’s chronogram for the occasion of his son’s birth,
namely “nuzul-i ġhaib” [a gift from the unseen world], whose letters add
up to 1099. Indeed, Ārzū’s entry in Safīnah-yi Kh: wushgo
: includes a long
quotation from Ārzū describing his own life, so it is the most reliable
source available.⁹ Kh : wushgo
: does not specify where Ārzū was born
but mentions that his father came with the emperor Aurangzeb’s
(r. 1658–1707) army to Gwalior, so Ārzū was either born or spent his
early childhood there. His nisbat [toponymic surname] is at times given as
“Gwāliyārī” [from Gwalior] or “Akbarābādī” [from Agra] as well as more
rarely “Dihlawī” and “Shāhjahānābādī” [from Delhi].
Ārzū had an impeccable mystical and poetic lineage through both of
his parents.¹⁰ On his mother’s side, he claimed descent from the twelfth-
century Iranian mystic poet Farīd al-Dīn ʿAt:::t:ār through Muh: ammad
Ġhaus̄ (d. 1653). Muh: ammad Ġhaus̄ was a politically connected, musi-
cally inclined Shat:::t:ārī Sufi in Gwalior. Ārzū’s illustrious ancestor is
known to history as a translator of Yogic texts, the teacher of Tānsen
(the most famous musician of the emperor Akbar’s court), and as an ally
of the first Mughal emperor, Bābur, when Bābur conquered Gwalior.¹¹
Ārzū’s father was descended from the Chishtī Sufi saint Nas: īr al-Dīn
Mah: mūd (d. 757/1356), known as “Chirāġh-i Dihlī” [The Lamp of
Delhi]. One tażkirah writer begins his entry on Ārzū with the words
“nasab-i sharīfash” [his noble lineage] and nearly all of them mention
something to that effect.¹² Given how frequently such lineages are
referenced, they clearly mattered a great deal in literary high society
because Persian poetry, even on what could be called “secular” subjects,
is shot through with Sufi imagery. Ārzū’s father, Shaikh : Husām

al-Dīn,
whose tak: hallus: [pen-name] was Husāmī,

was a soldier-poet in the

⁹ See Safīnah-i Kh: wushgo

: : wushgo’s
1959, 312ff. Kh : text was first completed in 1147/1734–5,
was corrected by Ārzū after 1155/1742–3, and was later revised up to 1162/1748–9 (Safīnah 1959,
editor’s preface). Mahdi Rahimpoor argues conclusively for 1099 as Ārzū’s birth year
(2008a, 241).
¹⁰ Compare, for example, ʿIqd-i S̄uraiyā (1978, 28) and Tażkirah-yi Riyāẓ al-ʿĀrifīn (1977, 8).
Muntak: hab al-Ashʿār [A Selection of Verses], compiled in 1748 and available as Bodleian ms
Elliott 247, has very short entries but still mentions his native place as Gwalior and his descent
from Muh: ammad Ġhaus̄, suggesting that these facts were important to the compiler (f. 28;
Sachau and Ethé 1889, 239ff.).
¹¹ Ernst 1996; Nizami 2002.
:
¹² Khulā s:at al-Afkār [Essence of Thoughts], a tażkirah begun in 1206/1791–2, Bodleian ms
Elliott 181, f. 31a. On the text, see Sachau and Ethé 1889, 302ff.
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

20       

Mughal mold, and much of what we know about him comes from Ārzū’s
own tażkirah.¹³ He was elevated as a mansabdār : [salaried nobleman]
under Aurangzeb, and was well acquainted with high-ranking nobles
[umarāʾ-i ʿumdah]. He did not frequently read his poetry in front of
people, Ārzū says, because this would be “contrary to the path of soldier-
:
ing” [khilāf-i tarīqah-yi sipāhīgarī]. Although he was not formally edu-
cated in political philosophy, he was a confidant of two important
officials, the Mīr Munshī [imperial chief secretary] Fāz̤ il Khān
: and Mīr
Muh: ammad Amīn. The best known of his works is Husn-o
ʿIshq [Beauty
and Love], a Persian adaptation of the Awadhi [eastern hindī] romance
Madhumālatī.¹⁴
Ārzū’s early education was undertaken by his father and followed the
pattern typical for students across the Persian cosmopolis. Kh : wushgo
:
reports that before the age of five or six Ārzū had read Saʿdī’s Bostān,
Gulistān, and Pandnāmah, and that he studied Arabic until age four-
teen.¹⁵ His father helped him memorize “one or two hundred couplets”
by Modern [mutaʾak: hkhirīn]
: poets, and at fourteen Ārzū developed an
interest in writing Persian poetry. Kh: wushgo
: provides the detail, nestled
within Ārzū’s autobiographical quotation, that Ārzū first wrote poetry
while visiting Mathura. The historical context of this fact is not obvious,
so perhaps it is mentioned by way of glorifying Kh : wushgo’s
: own home-
town, which he compliments hyperbolically as “ground that awakens
Judgment Day and a tumult-exciting place.”¹⁶ It could also be a way for
Ārzū to demonstrate his connection to vernacular poetics, since Mathura
was, and indeed still is, a center of Krishna-worship and hindī poetry. In
any case, the mention of Mathura implies that Ārzū’s childhood was
peripatetic, since he would have spent time in three places: Agra,

¹³ MN 2004, 1:359–60.
¹⁴ Behl 2012, 335. Behl corrects the misunderstanding that this text was a translation of
Padmāwat and notes that there is an illustrated manuscript in the private collection of the late
Simon Digby. A manuscript in Berlin (described as “Padmāwat”) was composed in 1071/
1660–1 and dedicated to Aurangzeb (Pertsch 1888, 929–30). Ārzū writes that his father
composed a qis: s: ah on the well-known—and frequently retold—story of Kāmrūp and
Kāmlatā, but “did not find leisure” (presumably to complete it). Strangely, Kh : wushgo
: claims
that Ārzū wrote a mas̄nawī called Husn-o

ʿIshq himself so perhaps these were conflated even
though they are supposedly on different subjects. The title, “Beauty and Love,” is somewhat
generic and has been used by several poets.
¹⁵ Safīnah 1959, 313.
¹⁶ “k: hāk-i qiyāmat-k: hez wa sar-zamīn-i shor-angez” (Safīnah 1959, 313).
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

   21

Gwalior, and Mathura. One of Ārzū’s childhood teachers was Mīr

Ġhulām ʿAlī Ah: sanī, about whom we know nothing except that he was
active in Gwalior and corrected Ārzū’s verses.¹⁷ Another was Mīr ʿAbd
al-Samad

:
Sukhan, who in his early days was in Gwalior before working
for nobles in Agra and Lahore.¹⁸ Ārzū knew him briefly in Gwalior.¹⁹
When Ārzū was a budding poet of fifteen or sixteen, Sukhan : corrected
his work, and even fifty years later Ārzū kept scraps of paper with
:
Sukhan’s corrections on them.²⁰ Sukhan: was eventually posted to
Gujarat, where he died in Ahmedabad, but at some point his calligraphy
had become famous in Delhi.
Upon his father’s death in 1115/1703, Ārzū joined the entourage of
Prince Aʿz̤ am Shāh, who happened to be an important literary patron in
Persian and hindī.²¹ When the imperial army marched to the Deccan
with a contingent including the prince’s forces, Ārzū was part of the
campaign for nine months. The aged Aurangzeb died shortly thereafter
on March 3, 1707, and the inevitable succession struggle commenced.
Aʿz̤ am Shāh declared himself emperor immediately after his father’s
death in Ahmadnagar (present-day Maharashtra). Though his claim
was made in the imperial encampment, it lasted a mere four months.
His brother, Prince Muʿaz̤ z̤ am (who took the regnal name Bahādur Shāh,
r. June 19, 1707–February 27, 1712), killed him on the battlefield that
June.²² Ārzū himself had already returned to Gwalior because—at least in
his telling—his mother had asked him to come home, which we can
probably read as a convenient excuse to avoid a tricky political situation.
After the bloody resolution of the princes’ competing claims to the
throne, Ārzū relocated to Agra and then back to Gwalior. He resettled
in Agra during the brief reign of Jahāndār Shāh (r. February 1712–
February 1713), and spent five years there as a religious student and
participant in the poetic scene.²³

¹⁷ MN 2004, 1:87–8. Not even Ārzū had access to his dīwān, so he must have been a relatively
marginal figure.
¹⁸ MN 2004, 2:716. ¹⁹ Safīnah 1959, 213. ²⁰ MN 2004, 2:716.
²¹ Busch 2010, 297. ²² See Richards 1996, 253; cf. Chandra 2002, 55.
²³ He studied religious science under Maulānā ʿImād al-Dīn, known as Darwīsh Muh: ammad,
and had his verses corrected by Shāh Gulshan (d. 1140/1727), Mirzā Hātim
Beg Hātim

(note
that this is not the poet Shāh Hātim

Dihlawī), ʿAz̤ matallāh Kāmil, Muh: ammad Muqīm Āzād,
the great Naqshbandī Sufi Nās: ir ʿAlī Sirhindī (d. 1108/1696), and others. While both Kh
: wushgo
:
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

22       

Delhi

Ārzū visited Delhi before settling there. He tells us that he first came in
:
the beginning of the reign of Farrukhsiyar (1713–19) in an unsuccessful
search for work. He then went to Agra and entered the service of Mīrzā
:
Khān in Gwalior. He returned to Delhi shortly after Muh: ammad Shāh
became emperor (that is, after September 1719), and remained there, he
tells us, for thirty years (at the time of writing Majmaʿ al-Nafāʾis in
1751–2).²⁴ During this period he wrote nearly all of his works and made
the personal connections that enabled his career. Soon after settling in
Delhi he must have met Ānand Rām Mukhli : s: , who introduced him to his
future patron Ish: āq Khān
: and facilitated his entrée into literary high
society. He also came to know the great poet Mīrzā ʿAbd al-Qādir Bedil
and his circle before Bedil’s death in 1720. This short-lived association
with Bedil brought him great prestige. While he lived in Delhi, he
witnessed the invasion of the Iranian ruler Nādir Shāh and the general
massacre (qatl-i ʿāmm) of the city’s inhabitants by the invading troops in
March 1739. The violence in the capital and subsequent looting of both
city and imperial treasury are generally regarded as heralding the end of
Mughal power, although the last Mughal emperor would not be deposed
until 1857.

Final Years in Lucknow

The best contemporary sources on Ārzū’s life, namely Safīnah-yi

: wushgo
Kh : and Ārzū’s own Majmaʿ al-Nafāʾis, were both completed in

and Ārzū’s accounts mention Darwīsh Muh: ammad, only Kh : wushgo

: lists the poets with whom
Ārzū was associated in Agra. If Nās: ir ʿAlī’s death date is correct then he obviously could not
have corrected Ārzū’s verses in the 1710s.
²⁴ MN 2004, 1:187. Āzād Bilgrāmī concurs on these facts, giving 1132/1719–20 as the date of
Ārzū’s arrival (Sarw-i Āzād 1913, 228). The year 1719 was a tumultuous one for imperial
politics. The Sayyid brothers, who had helped Farrukhsiyar take the throne in 1713, now
decided to betray him. He was blinded in January and strangled towards the end of April.
They put his young nephew Rafīʿ al-Darajāt on the throne, but he promptly died and was
succeeded by his brother Rafīʿ al-Daulah (who reigned for about four months as Shāh Jahān II).
Afterwards Muh: ammad Shāh took the throne and built a coalition that succeeded in breaking
the power of the Sayyid brothers. (The classic account of these events is Irvine 1922.)
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

   23

Delhi before Ārzū settled in Lucknow in Awadh. Thus, there is little we

can say about this final period of his life.²⁵ The author of Khulā
: : al-
sat
Afkār [The Pith of Thoughts], writing in the 1790s, suggests that Ārzū
left Delhi because of the “desolation” [kharābī]
: of the city.²⁶ Ārzū’s own
writings give no indication of his thoughts on leaving the capital so this
motivation is perhaps an interpretation at several decades’ remove.
Whatever the situation in Delhi, Ārzū no doubt moved to Lucknow in
1754 to be closer to a new patron, Nawab Shujāʿ al-Daulah.²⁷ Ārzū’s
nephew Mīr Taqī Mīr cruelly refers to this period as “chasing in the
desert of greed” and implies that Ārzū was unsuccessful in getting the
Nawab’s patronage.²⁸ All the sources that mention it agree on the date of
Ārzū’s death, 23 Rabīʿ II 1169/January 26, 1756, and most state that his
body was taken to Delhi for burial. We are not told by whom or why such
trouble was taken.²⁹

Ārzū’s People: Friends, Patrons, and Rivals

Given the arc of Ārzū’s career, it is unsurprising that he was tied into
multiple literary and political networks. He was personally linked with
the literary communities of Delhi, Agra, and Gwalior, both in Persian
and in vernacular literary circles,³⁰ and his friends and disciples repre-
sented a remarkable cross-section of the elite. Like the political class
itself, his interlocutors included people from all over India, Central Asia,
and Iran. Two of Ārzū’s closest friends, Ānand Rām Mukhli : s: and Tek

²⁵ The main source is Āzād Bilgrāmī’s tażkirah Khizānah-yi

: ʿĀmirah [Royal Treasurehouse,
1164/1762–3]. Bodleian ms Ouseley Add. 6, p. 206.
:
²⁶ Khulā s:at al-Afkār by Abū T̤ ālib (Bodleian ms Elliott 181, f. 31a). The work was begun in
1206/1791–2 and completed sometime before 1210/1795–6 (Sachau and Ethé 1889, 302ff.).
²⁷ Although several sources are precise in locating Ārzū in Lucknow, some scholars have
speculated that instead of going to Lucknow, Ārzū settled in Faizabad (130 km to the east).
Faizabad and not Lucknow was the first capital of Awadh, but there is no evidence to suggest
that the tażkirah-writers are wrong to refer to Ārzū’s stay in Lucknow (Khatoon 1987, 36; cf.
Khatoon 1991, 18–23).
²⁸ Trans. Naim 1999, 76.
:
²⁹ For example, Khizānah-yi ʿĀmirah (Bodleian ms Ouseley Add. 6, p. 206).
³⁰ People whose connection to Ārzū is only relevant for our purposes because of Urdu poetic
society (whatever Persian they wrote) will be dealt with in Chapter 5. The list includes Ābrū,
Maz̤ har, Maz̤ mūn, Mīr, Mīr Dard, Saudā, and Yakrang.
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

24       

Chand Bahār, were Hindus (more specifically khattris).³¹ Many of Ārzū’s

eventual colleagues and students were first brought together as disciples
of the Sufi poet Bedil, whose death shortly after Ārzū’s arrival in Delhi
led to a new arrangement in which Ārzū (perhaps in part because of his
own Sufi credentials) became teacher to many of them and a sort of
master of ceremonies in Bedil’s annual memorial [ʿurs].³² Tracing the
diverse group of people with whom he interacted is important as back-
ground for our discussions of contemporary literary debates in the
following chapters because disputes are often both personal and
philosophical.
The source material available to us for reconstructing the literary life
around Ārzū is almost exclusively in the form of tażkirahs, a category of
evidence that needs to be interpreted with care. Literary tażkirahs are not
historically minded documents and so have little interest in hard facts
like dates and places, and it is generally fruitless to read them in pursuit
of such information.³³ Because tażkirahs are concerned above all with
transmitting good poetry to posterity, their prefaces often cite the num-
ber of couplets recorded rather than the number of people included. In
many cases, the volume of poetic quotations so outweighs the text of the
biographical entries that they are better understood as poetic anthologies
arranged by author. Soft historical facts often fare no better than bio-
graphical specifics: The language is frequently stereotypical, with flowery
variations on “he was a great poet” rather than real critical content. But if
tażkirahs are not historical documents, then what are they? They are
literary representations of social networks and the memory thereof.³⁴
Tażkirahs were deliberate in their presentation of how poets were con-
nected both personally and aesthetically, and often we can reconstitute

³¹ Many Hindus were written out of the history of Indo-Persian. For example, Mīr Husain


Dost Sanbhalī’s Tażkirah-yi Husainī

(1875) does not contain any of the Hindus mentioned in
this chapter except Chandar Bhān Brahman. For a good overview of Hindu Persian poets, see
Pellò 2008. On Persian-using Hindu administrators, see Kinra 2010.
³² Tabor 2019, 86–7.
³³ However, they gradually become more historical. The watershed moment, the compilation
of Āzād’s Āb-i Hayāt

in 1880, is discussed in Chapter 5.
³⁴ An idea developed in Hermansen and Lawrence 2000 (and applied in, for example, Kia
2011). It should be noted, however, that especially in Urdu tażkirahs, compilers are sometimes
concerned more with particular conceits, such as (in one extreme example) including only
verses that mention parts of the body, than with representing an actually existing network of
poets (Pritchett 1994, 66).
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

   25

these ties into an understanding of poets’ intellectual and aesthetic

concerns. When used correctly, namely by tuning our interpretation to
the rhetoric of representing a community of poets, they can tell us a
great deal.
Ārzū’s own tażkirah Majmaʿ al-Nafāʾis is a guide to the people
whom Ārzū respected as his teachers and students or as poets in general,
and those whom he would rather have held at arm’s length. Mukhli: s: and
Bahār, both intellectuals in Ārzū’s mold, were Ārzū’s closest friends
and we consider them first. Then we can address the influence of Bedil,
who was teacher to all of them, before turning to some of the more
peripheral figures whom Ārzū mentions. We consider his patrons sepa-
rately. Finally, we consider his enemies, and through them the complex-
ity and stakes of eighteenth-century Indo-Persian literary debate.

Ānand Rām Mukhli

: s:

The present chapter began with a lexicographical affirmation by Ānand

Rām Mukhli: s: (d. 1164/1750–1) of his friendship with Ārzū. Mukhli
: s: had
been Bedil’s student, but after Ārzū settled in Delhi and Bedil died, he
became Ārzū’s student and arranged for Ārzū to receive the trappings of
nobility—namely an estate [jāgīr], an imperial rank [mans: ab], and a title
: ::āb] of “k: hān”—required to move in the empire’s highest circles.³⁵
[khit
: s: could accomplish this because he was one of the most important
Mukhli
political functionaries in Delhi: He was wakīl, which is to say the
personal representative at court, both of the imperial wazīr, Qamar al-
:
Dīn Khān (known by his noble title Iʿtimād al-Daulah), and of ʿAbd al-S

:
amad Khān (known as Saif al-Daulah), governor of Lahore and Multan.³⁶
He had received the title “Rajah of Rajahs” [rāʾī-yi rāyān] for his service.
: s: ’s skills clearly commanded the respect of
Despite the fact that Mukhli
those around him, later scholars have been wary of accepting their

³⁵ Kh: wushgo
: reports that Ārzū’s courtly title was “Istiʿdād Khān”
: (Safīnah 1959, 312), but
none of the other texts under discussion here mention it.
³⁶ The best English-language account of Mukhli : s: ’s life is the prefatory material in Mirʾāt al-
Is::t:ilāh: 2013. In Urdu see ʿAbdullah 1967, 122ff. On Qamar al-Dīn Khān : and ʿAbdul Samad


:
Khān, see Beveridge and Prashad 1979, 2:488–90 and 1:71–3.
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

26       

verdict. Sayyid ʿAbdullah, a twentieth-century Pakistani scholar, feels the

need to say that Mukhli : s: could not have had pure [t: het: h] Persian
because he was an Indian [hindūstānī] and worse than that a Hindu.³⁷
Contemporary accounts configure Mukhli : s’s
: relationship with Persian
differently: He was entrusted with composing a letter to the Safavid king
to commemorate the ascension of Muh: ammad Shāh—hardly a task for
someone with imperfect Persian.³⁸ Furthermore, ʿAlī Qulī Khān : Wālih
Dāġhistānī, writing in 1169/1748, praises him as the most highly fluent
Hindu Persian writer of his time.³⁹
: s: to Ārzū are the only specimens of anyone’s
Six letters sent by Mukhli
correspondence with Ārzū that have been preserved, as far as I have been
able to ascertain.⁴⁰ They appear in Mans̄ūrāt-i Ānand Rām [Ānand
Rām’s Prose], an unpublished miscellany of Mukhli : s’s
: writings, which
includes a set of letters gathered in 1149/1736. Unfortunately for histor-
ians, the criteria for inclusion in such a collection have little to do with
historical value but rather with rhetorical force: Such letters were meant
to serve as models of elegant writing (and indeed as a delivery vehicle for
poetry), and not necessarily as a historical record. The letters pertain to
the period when Mukhli : s: was in the Deccan on campaign with Iʿtimād
al-Daulah in 1147/1736–7 fighting the forces of the Maratha Peshwa Bājī
Rāo.⁴¹ Mukhli: s: expresses his longing for Ārzū’s company, sends some
poetry, and describes the itinerary. A decade later, when Mukhli : s: was
again traveling, he recorded in his Safarnāmah [Travelogue], which
covers events in 1745, that he met Ārzū on the road.⁴² Mukhli : s: ’s

³⁷ ʿAbdullah 1967, 125. A recently published monograph on Mukhli : s: (James 2011) is

somewhat disappointing because it likewise cannot conceive of its Hindu subject except as a
brilliant imitator.
³⁸ Though as with much of the ceremonial correspondence at this level, whether it was sent
or not is an open question. It is contained in Mans̄ūrāt-i Ānand Rām [Ānand Rām’s Prose], an
unpublished miscellany of Mukhli: s: ’s writings (Khuda Bakhsh ms HL 882 ff. 56b–67a; see Khuda
Bakhsh Library 1970, 9:109).
³⁹ Tażkirah-yi Riyāẓ al-Shuʿarā 2005, 4:2209.
⁴⁰ A few letters ascribed to Ārzū are to be found in a majmūʿah [miscellany] in the Collection
of the Asiatic Society of Bengal, but I have not been able to access the manuscript (no. 420;
Ivanow 1985, 184).
⁴¹ The letters are Mans̄ūrāt-i Ānand Rām ff. 3a–5a, 5a–6a, 6a–9a, 14a–15a, 19b–20b, and
25a–27a. The third letter is dated 25 Shawwāl of Muh: ammad Shāh’s seventeenth regnal year
(= 9 March 1736) and the fifth is headed with a notice that it dates from the campaign against
Bājī Rāo.
⁴² Alam and Subrahmanyam 1996, 145.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

   27

best-known work, the lexicon Mirʾāt al-Ist:::ilāh: , was edited after his death
by Ārzū who added marginal notes and a preface.
: s: was at the center of literary life in the capital. He was a
Mukhli
favorite disciple [shāgird] of Bedil, and hosted Bedil’s former students
at home.⁴³ He owned a remarkable library because his wealth allowed
him to have a copy made of any book which he came across.⁴⁴ He
borrowed books from Ārzū.⁴⁵ His social circle was broad, as we learn
from the circumstances of the composition of Hangāmah-yi ʿIshq [The
Tumult of Love, 1739]. According to the preface, Ārzū, Muh: ammad Qulī
:
Khān, :
Maʿnīyāb Khān Shāʿir (d. 1157/1744), Rāo Kirpā Rām, Rāʾī Fath:
Singh, and “other friends” accompany the author to the fair of a Sufi
saint.⁴⁶ Afterwards, he cannot sleep and so a servant narrates part of the
hindī poem Padmāwat, which he decides to turn into Persian prose for
“people interested in this art [of Persian composition]” [ahl-i żauq-i īn
fann].⁴⁷ In Chapter 5, we pick up the argument that it is no accident that
stories moved back and forth between hindī and Persian in the company
of such elites.

Tek

Chand Bahār
The closeness of Ārzū’s friendship with Bahār is clear from the repeated
references in Ārzū’s Khiyābān-i
: Gulistān, a commentary on Saʿdī’s
Gulistān, to Bahār’s being “among humble Ārzū’s friends.”⁴⁸ No one
else gets this sort of treatment in that text. Bahār himself says in the
preface to his dictionary Bahār-i ʿAjam [The Spring of Persian, 1152/
1739] that Ārzū had been his close friend for twenty years. He was likely

⁴³ Ārzū mentions one such gathering in which Ārzū, Payām and ʿAt::ā (all disciples of Bedil
like Mukhli : s: ’s house (MN 2005, 64). Mukhli
: s: himself) met Mukhli : s: corresponded with Payām,
as demonstrated by the two letters addressed to him preserved in Mans̄ūrāt-i Ānand Rām.
⁴⁴ Jalibi 1984, 2.1:164.
⁴⁵ Such as the dīwān of the seventeenth-century poet Sālik Qazwīnī (Mirʾāt al-Is::t:ilāh: 2013,
384–5).
⁴⁶ Muh: ammad Qulī Khān: must be the same person as the author of the Muraqqaʿ-i Dihlī. He
was in Delhi in the entourage of Ās: af Jāh I from 1738 to 1741.
⁴⁷ Mans̄ūrāt-i Ānand Rām f. 137a–b. A second translation from hindī into Persian in the
same collection is Kārnāmah-yi ʿIshq [Book of Love, 1731]. The latest work included in the
manuscript is dated 1746.
⁴⁸ “az yārān-i faqīr-i ārzū ast” (KhG
: 1996, 22, 33, 89, 111).
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

28       

Ārzū’s exact contemporary, but we know little about his life.⁴⁹ There is
no entry for him in Ārzū’s Majmaʿ al-Nafāʾis and other slightly later
tażkirahs have sketchy entries.⁵⁰ For example, Mīr Hasan
tells us only
that he knew Persian idioms well [az ist ::: :
ilāhāt-i fārsī bisyār :
khabar
dāsht], was Ārzū’s friend, and was a prolific writer.⁵¹ The idea that he
was a prolific writer is difficult to substantiate in the usual context of
writing a great deal of poetry because he left no dīwān, a book of selected
poetry that generally formed a carefully curated part of a poet’s legacy.
He was the author of at least five critical works, of which the most
important is his massive dictionary Bahār-i ʿAjam. Arguably, simply
completing a dictionary on the scale of Bahār-i ʿAjam would qualify
someone as a prolific writer.⁵²
Bahār-i ʿAjam’s sources are worth discussing. They demonstrate an
encyclopedic grasp of the tradition, as in Ārzū’s own lexicographical
projects, but also a surprising willingness to use texts written by people
who did not get along with one another philosophically or personally.
The Bodleian manuscript provides a list of works from which the
dictionary is derived: It mentions well over 200 poetic collections, dic-
tionaries, and commentaries. Ārzū’s own works in the list include a
dīwān of ġhazals, his Sikandarnāmah and Gulistān commentaries, the
dictionary Sirāj al-Luġhat, and “several treatises” [baʿz̤ rasāʾil] besides.⁵³
Indeed, Bahār highlights in the preface that Ārzū receives the special

⁴⁹ ʿAbdullah 1967, 162ff. The evidence for the composition date is a chronogram (“yādgār-i
faqīr-i h: aqīr bahār”) which yields 1152/1739, twenty years into Ārzū’s stay in Delhi, and so
precisely tracks with Bahār’s statement that he had known Ārzū for twenty years (Bahār-i ʿAjam
2001, 1:xxx). The work was subsequently revised and this has confused some scholars about
when it was first compiled.
⁵⁰ I have checked both the critical edition of MN (2004) and the table of contents of the
comprehensive Bodleian ms (Elliott 399). The omission is puzzling but perhaps Bahār was a
literary scholar rather than a practicing poet since he apparently left no volume of selected
poetry [dīwān].
⁵¹ ʿAbdullah 1967.
⁵² He wrote three treatises called Jawāhir al-Hurūf,

Nawādir al-Mas:ādir, and Ibt::āl-i Zurūrat.
:
Blochmann calls Bahār’s dictionary Bahār-i ʿAjam “one of the grandest dictionaries ever written
by one man” and notes that it was the only one of Bahār’s works that was readily available even
though the others had been lithographed (Blochmann 1868, 28–30). It has most recently been
published in an edition prepared by Kazim Dizfuliyan (Bahār-i ʿAjam 2001). Bahār also wrote a
well-known commentary on Saʿdī’s Bostān called Bahār-i Bostān. He was also a rek: htah poet
(that is, he wrote in the Persianized vernacular) according to Mīr (Nikāt 1979, 164) and Shafīq
(Chamanistān-i Shuʿarā 1928, 44–8).
⁵³ Ms Caps. Or. B 15, f. 2b (on this ms see Sachau and Ethé 1889, 1018).
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

   29

honor of being referred to as “sirāj al-muh: aqqiqīn” [Lamp of the

Researchers] throughout the book. Bahār expresses his deep devotion
to Ārzū and uses works by Ārzū’s friend S̄ābit, and yet also acknowledges
:::alah: āt al-Shuʿarā [Idioms of the Poets],
his great debt to Wārastah’s Must
and refers to Ārzū’s archnemesis Shaikh : Muh: ammad ʿAlī Hazīn
in
highly complimentary terms, namely as “h: az̤ rat-i shaikh : al- ʿ ārifīn” [the
: of the wise].⁵⁴ Bahār is clearly a partisan of Ārzū and yet
exalted shaikh
has no qualms about relying upon the works of Ārzū’s enemies like
Wārastah and Hazīn

(whose particular enmity towards Ārzū is described
in this chapter).

The Ticket to Success: Mīrzā ʿAbd al-Qādir Bedil

Bedil (1054–1133/1644–1720) was arguably the most important Persian

poet of his time anywhere in the Persian cosmopolis. His remarkable
career took him from Patna, where he grew up in a family with Central
Asian roots, to other parts of India, and eventually to Delhi in 1096/
1685.⁵⁵ He mentored dozens of students and set a standard for Persian
literature: Friend or foe, every Indo-Persian writer who came after felt
the need to engage with his style. He counted among his students great
religious figures and some of the most powerful nobles of the empire.⁵⁶
For example, Shaikh : Saʿdallāh Gulshan (d. 1141/1728–9) was an impor-
tant Naqshbandī Sufi, whom Bedil taught and respected as being
especially divinely inspired in his poetry.⁵⁷ Another student was Niz̤ ām
al-Mulk Ās: af Jāh, who eventually carved out an autonomous province in
Hyderabad and the eastern Deccan for himself and his descendants.⁵⁸

⁵⁴ Bahār-i ʿAjam 2001, 1:xxix–xxxi.

⁵⁵ Ārzū notes that Nas: rābādī’s tażkirah incorrectly connects him with Lahore and that he
lived in Delhi for “about thirty years” (MN 2004, 240–1). ʿAbdul Ghani cites a chronogram for
1075/1664 as the year Bedil left Patna but it is unclear how long he stayed in other places
(including Mathura) before settling in Delhi for the first time (Ghani 1960, 30–1). He settled
there permanently in 1096/1685 (Ghani 1960, 61). The standard intellectual biography of Bedil
in Urdu is Hadi 1982.
⁵⁶ A comprehensive list appears in Ghani 1960, 82ff.
⁵⁷ MN 2004, 3:1369. On Gulshan’s role in developing Urdu literature, see Chapter 5.
⁵⁸ Safīnah 1959, 114. However, as Munis Faruqui has argued, Niz̤ ām al-Mulk’s motivation
for gaining political independence in the Deccan emerged from his weakness and marginaliza-
tion at the imperial court (Faruqui 2009).
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

30       

Like Ārzū, Bedil had been a member of the establishment of Prince

Aʿz̤ am Shāh, although much earlier than Ārzū.⁵⁹ Furthermore, tażkirah
writers imply that he was a major proponent of bringing non-Muslims
into Sufi poetic circles.⁶⁰
Ārzū became closely linked to Bedil, despite apparently having little
direct contact with him. He was Bedil’s student twice, first during the
time he stayed briefly in Delhi at the beginning of Farrukhsiyar’s : reign
and again when he settled there permanently.⁶¹ Not long after his arrival
the second time, Bedil died; his death cast a very long shadow. Ārzū took
up the training of many of Bedil’s students, as is clear from Majmaʿ al-
Nafāʾis. He also had a special role in Bedil’s ʿurs [death anniversary], a
major annual event in the Persianate Sufi circles of Delhi.⁶² Bedil’s poetic
circle was part courtly aesthetics—indeed Kh : wushgo
: rhetorically frames
it as a royal court—and part Sufi lodge [khānaqāh]. Kh : wushgo
: gives the
most intimate account of Bedil, including his dining habits. The great poet
would apparently eat two and a half or three ser [approx. 3 kilograms] of
food at a sitting.⁶³
Bedil’s literary style is controversial to say the least. He is the bête noire
of Iranian critics of the twentieth century, and is held up as the exemplar
of the excesses of the so-called “Indian Style.”⁶⁴ Recent critical literature
often implies that he started out as a good poet in the classical mold
before something went terribly wrong.⁶⁵ Without a doubt, Bedil’s poetry

⁵⁹ Siddiqi 1989. ⁶⁰ Pellò 2014.

⁶¹ The phrase in MN is ambiguous, lit. “Humble Ārzū [was] in the service of this great man
twice” [faqīr-i ārzū do bār bah k: hidmat-i īn buzurgwar], which could even be interpreted to
mean that he only met Bedil in person on two occasions. That meaning is unlikely but still the
timeframe in question is undefined.
⁶² Muraqqaʿ 1993, 81. It took place each year on the fourth of the month of Safar.

: wushgo
Kh :
is perhaps implying that Ārzū took up Bedil’s mantle by using virtually the same honorific
formula in his entries for each of them, namely “qiblah and kaʿbah of significations” [qiblah-o
kaʿbah-yi maʿānī] and “qiblah of the word and kaʿbah of significations” [qiblah-yi lafz̤ wa
kaʿbah-yi maʿānī], respectively (Safīnah 1959, 312, 104). The qiblah is the direction in which
Muslims pray and the kaʿbah is the black cubic building in Mecca towards which the qiblah is
directed. On the importance of Bedil’s ʿurs as a socio-cultural event, see Tabor 2019.
⁶³ Safīnah 1959, 109. ⁶⁴ E.g., Shafiʿi Kadkani 1981, 152.
⁶⁵ E.g., Siddiqi 1989. Even ʿAbdul Ghani, who admires Bedil’s style, could not resist putting
decline at the center of his analysis: In his telling, Bedil’s poetic genius was not properly
recognized because of the cultural rot that supposedly set in after Aurangzeb’s death (Ghani
1960, vii–viii).
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

   31

is complicated and not to everyone’s taste, but we cannot assume that his
difficulty was the result of a failure to understand proper Persian as
opposed to an aesthetic choice.⁶⁶ Gardezī, a nineteenth-century tażkirah
writer, calls him “master of his own style” [sāh
: : ib-i ::t arz-i kh
: wud],
: which
was generally thought to be the highest form of poetic achievement.
(Ārzū uses a similar phrase in describing him.) Shamsur Rahman Faruqi
has argued that “it can be said that most of the criticism of the Indian
Style poets’ use of Persian emanates from Bēdil’s (dis)reputation as an
undisciplined writer.”⁶⁷
We can compare Bedil to other roughly contemporary poets, who
were nodes in their own networks of teachers and students: Another poet
with such influence, albeit in the generation before Bedil, was Mīrzā
Muh: ammad ʿAlī Sāʾib

(d. 1080/1669), who had come to India in Shāh
Jahān’s reign but returned to his native Isfahan, where he died. His
students Riẓwān and Sābiq and acquaintance Fit::rat are mentioned in
Majmaʿ al-Nafāʾis. Likewise, ʿAbd al-Lat̤ īf Khān
: Tanhā, who was con-
troversial in his time (which was the latter part of Aurangzeb’s reign, that
is, the end of the seventeenth century), had students (e.g., Maʿjiz and
Nis̄ar) and continued the tradition of his own teacher Mīrzā Jalāl Asīr
Isfahānī (d. 1049/1639). Majmaʿ al-Nafāʾis emphatically declares that
Mīrzā Afẓal Sarkh : wush
: (d. 1127/1715) was not a student of Bedil’s.
Rather Sarkh : wush
: had his own establishment of students and was in
that sense a rival to Bedil.⁶⁸

⁶⁶ Indeed, the most biting comment on Bedil is probably a later fabrication: Hazīn
suppos-
edly said, “Bedil’s prose is unintelligible. If I manage to go back to Iran, there is no better present
to make my friends laugh” [nas̄r-i bedil ba-fahm namī āyad, agar murājaʿat-i īrān dast dahad
barāʾī rīshk: hand-i bazm-i ah: bāb rah āwardī bahtar azīn nīst] (quoted in Ghani 1960, 259). The
difficulty is that Ghani cites Muh: ammad Husain
Āzād’s Nigāristān-i Fārs [Persian Picturebook]
as the source for Hazīn’s

now quite famous insult, and in this work Āzād—who expresses his
contempt for Bedil’s style—tends to make up or at least misattribute quotations. I have not been
able to trace the quip to an earlier work, but scholars have theories about where Āzād might
have found it if it is indeed authentic (Mikkelson 2017, 525n65).
⁶⁷ Faruqi 2004b, 59, cf. 24. It was a contemporary objection as well: Faruqi points to Āzād
:
Bilgrāmī’s stylistic objections in his tażkirah Khizānah-yi ʿĀmirah but notes that Āzād concedes
that Bedil did have the right to experiment, citing Ārzū. A full discussion of that passage appears
in Chapter 3.
⁶⁸ Pellò 2014.
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

32       

Other Poets in Ārzū’s Network

Majmaʿ al-Nafāʾis is an index of other people Ārzū respected, including

many who do not appear frequently in other tażkirahs, but with whom
Ārzū claimed a personal relationship. It includes people from diverse
parts of India and a number of Hindus, especially Punjabis in high
positions in Delhi. If the function of tażkirahs is not recording history
but building a notional community, then Ārzū assembled the most
geographically and confessionally diverse group possible and strove to
keep ties with nobles who moved around India. Majmaʿ al-Nafāʾis aims
to be a comprehensive account of the Persian tradition, with some 1,800
poets represented in the most complete manuscripts, but clearly the
stakes are different in the entries on Ārzū’s contemporaries and near
contemporaries. Indeed, Ārzū carefully represented himself as anointed
successor of the two major poetic networks in Delhi, those of Bedil and of
Sarkh: wush.
:
In these entries, Ārzū situates himself relative to other poets, con-
structing for the reader a map of his close friends, friends with whom he
maintained relationships by letter, students, teachers, and—usually
maintaining the requisite politesse—his rivals and enemies. Ārzū had
an epistolary relationship with Mīr Ġhulām ʿAlī Āzād Bilgrāmī (d. 1200/
1785). Āzād had been educated by his uncle, the noted Islamic scholar
Sayyid ʿAbd al-Jalīl Bilgrāmī, and Ārzū was keen to know Āzād in part
because of his family’s spiritual credentials. Āzād settled in Aurangabad,
and Ārzū refers to their relationship as “friendship in absentia” [ikhlā
: s-i:
ġhāyibānah]. Ārzū sent him a message to get to know him (later for his
own tażkirah, Āzād likewise asks Ārzū for an account of himself).⁶⁹ He
replied with poetry, two ġhazals in Persian and a qasīdah
: in Arabic. Their
relationship is important for our purposes because it gives some sense of
Ārzū’s readership: Āzād was familiar with Majmaʿ al-Nafāʾis and Dād-i
:
Sukhan.⁷⁰ Likewise Ārzū had an epistolary relationship with Shāh
Faqīrallāh Āfarīn and similarly with Sharaf al-Dīn Payām.⁷¹ (Mukhli : s:

⁶⁹ MN 2004, 1:164.
⁷⁰ They are mentioned in the entry on Bedil in Khizānah-yi
: ʿĀmirah.
⁷¹ See Kia 2011, 229–30.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

   33

also corresponded with both of them, as we know from letters preserved

in Mans̄ūrāt-i Ānand Rām.) Another of Ārzū’s acquaintances was Mīr
Haidar

Tajrīd, a soldier who came from Surat to Delhi, spent time with
Ārzū, and eventually left for a posting in Bengal.⁷² Ġhulām Nabī Nasīm
has a similar biography: He was originally from Amroha (present-day
Uttar Pradesh), lived in Delhi before settling in Khudabad (then the
capital of Sindh). Although Nasīm missed his friends in Hindustan
(presumably meaning Delhi specifically), Ārzū reports, he was content
in Sindh.⁷³ He knew Ārzū for a little over thirty years. Mīrzā Afẓal
Sarkh : wush
: (d. 1127/1715) made an impression on Ārzū although they
knew each other only briefly during Ārzū’s first interval in Delhi during
:
the reign of Farrukhsiyar.⁷⁴ He mentions being a young poet and reciting
for Sarkh : :
wush, who recognized his talent (according to Ārzū) by declar-
ing “I have never seen the thought of any young man reach this level” [tā
h: āl fikr-i hīch naujawān bah īn pāyah na-dīdah-am]. Ārzū corroborates
the story by twice invoking Sarkh : wush’s
: sense of “justice” (that is, his
good poetic judgment) and noting that his poetry is read even in Iran.
Ārzū follows his anecdote by mentioning Sarkh : wush’s
: teachers
(Muh: ammad ʿAlī Māhir and later Mīr Muʿizz al-Dīn Fit::rat) as well as
the fact that he was “contemporary and stylistically similar” [muʿas: ir wa
hamt::arah: ] to Bedil. Ārzū praises Sarkh : wush’s
: tażkirah Kalimāt al-
Shuʿarā [Words of the Poets], which was itself a project designed to
corroborate Sarkh : wush’s
: own poetic lineage. Clearly Ārzū was trying to
establish himself as being tied into two major poetic networks in Delhi,
that of Bedil and of Sarkh : wush.⁷⁵
: There are many other poets with whom
Ārzū claims a relationship. Those actively involved in the literary debates
in Delhi will be discussed further in the context of Hazīn:
They include
Qizilbāsh Khān: Ummīd, who defended Ārzū, S̄ābit and his son S̄abāt,
and Rāsikh : (who were all defended by Ārzū), and Girāmī (whose inter-
action with Hazīn
Ārzū recounts).

⁷² MN 2004, 1:315.
⁷³ “az āshnāyān wa k: hwīshān-i
: hindūstān dil-ash giriftah wa bī hīch ranjīdah” (MN 2004,
3:1701).
⁷⁴ MN 2004, 2:676.
⁷⁵ He also notes having met Muh: ammad Husain

Nājī in the company of Sarkh: wush
: (MN
2004, 3:1691; cf. Safīnah 1959, 317).
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

34       

Ārzū’s circle also included a large number of Hindus, including some

who had converted to Islam. Besides Bahār and Mukhli : s,: Ārzū’s most
important non-Muslim confidant was Bindrāban Dās Kh : wushgo.
: He
was a student of Ārzū’s for twenty-five years and also connected to
: wush,
Bedil, Sarkh : and Gulshan. There does not appear to be any extant
: wushgo
collection of his poetry, but his tażkirah Safīnah-yi Kh : provides
important information about Delhi’s poetic scene.⁷⁶ Ārzū wrote a preface
for it and edited it, which then as now implies some degree of endorse-
ment. By the time Ārzū was compiling Majmaʿ al-Nafāʾis (which highly
praises the Safīnah), Kh : wushgo
: had retired from imperial service and
had become an ascetic wandering about Allahabad, Patna, and Banaras.
Another Hindu interlocutor of Ārzū’s was Gurbakhsh :
̤ ūrī. He was a
Huz
Kanbo from Multan who had known Ārzū continuously for forty years,
perhaps the longest of anyone mentioned in Majmaʿ al-Nafāʾis.⁷⁷ His
work was prolific and in the style of Ġhanī Kashmīrī (d. 1069/1668). Jai
Kishan ʿIshrat was a young Kashmiri Brahmin who had been Ārzū’s
friend for fifteen years. Like Ārzū, he was in Mūʾtaman al-Daulah’s
service before joining the service of his son Najm al-Daulah after the
father’s demise.⁷⁸ Ārzū reckons his mas̄nawī Rām-o Sītā [Rām and Sītā],
: story, superior to the well-known mid-
a telling of part of the Rāmāyana
sixteenth-century mas̄nawī of the same title by Shaikh: Saʿdallāh Masīh: ā
of Panipat. The apex of his bureaucratic career was his appointment as
dīwān [chief revenue officer] of all of Kashmir, a plum posting in his
native land. Another Kashmiri Hindu connection is with Sālim, who was
a protégé of Ġhanī and was thought to be a Kashmiri Brahmin who
converted to Islam.⁷⁹ Ārzū mentions that as a child he had studied
Sālim’s dīwān.⁸⁰ Ikhlā: s: Khān
: Wāmiq was a Khattri from Kalanaur
(Punjab) who converted to Islam—Ārzū’s account of his conversion is

⁷⁶ The work provides little information about its author, who probably died not long after
Ārzū. He probably finished it in 1147/1734–5, gave it to Ārzū in 1155/1742–3, and made some
changes up until 1162/1748–9 (Rahman 1959). It has three parts: The first discusses 362 Ancient
poets, the second 545 (or 811 in the Khuda Bakhsh ms) middle-period poets from Jāmī
(1414–1492), and the third—published as Safīnah 1959—deals with 245 Modern poets
(Sprenger 1854, 131). It is obviously the third section that is of most interest to us. In a separate
: wushgo
work, Kh : apparently recorded the sayings [malfūz̤ āt] of Bedil, but this text is no longer
extant. See Pellò 2014, 28.
⁷⁷ MN 2004, 1:396. ⁷⁸ MN 2004, 2:1139.
⁷⁹ Safīnah 1959, 38; on Sālim see Tikku 1971, 140–3. ⁸⁰ MN 2004, 2:603–4.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

   35

extremely complimentary.⁸¹ Ārzū had met him as a child in the company

of his father. While he was a good prose writer possessing his own
personal style [t::arz], he was never properly a poet. One young friend
of Ārzū’s was Bāl Mukund Shuhūd, a Kayasth whose ancestors worked as
administrators in Bengal and Bihar (Ārzū parenthetically notes here
that Kayasths were frequently Lords of the Pen [arbāb-i qalam] or
administrator-secretaries).⁸² He wrote Ārzū a cordial letter two years
before moving to Delhi, which is evidence that Ārzū was in demand as a
mentor in the capital. Indeed, Ārzū gave him his takhallu: : which is the
s,
usual prerogative of a teacher.⁸³
A few of Ārzū’s family members appear in his tażkirah in addition
to his father Husāmī.

:
The Persian and rekhtah poet Mīr was Ārzū’s
estranged nephew. Shaikh :

Hafīz̤ allāh Ās̄im (d. c. 1742) was Ārzū’s mater-
nal cousin who had also been in the service of Prince Aʿz̤ am Shāh and later
became a member of the emperor’s household troops [wālāshāhī]
because of his connections. Ārzū tells us nothing specific about him
except that he was a good poet and died in Agra.⁸⁴ Another poet, ʿĀqil
:
Khān Rāzī, was a distant relation of Ārzū’s through his mother’s connec-
tion to Muh: ammad Ġhaus̄, which Rāzī also claimed.⁸⁵

Patrons

After Ārzū settled in Delhi, he received patronage from two of the most
important noble families in the Mughal Empire. The first was that of
Ish: āq Khān
: and his descendants. They were Mughal administrators
of recent Persian origin, and were connected by marriage to Shujāʿ

⁸¹ MN 2004, 3:1784. He began his career in Aurangzeb’s reign, and under Farrukhsiyar :
received the rank of 7000 and wrote a history of that emperor’s reign. He is mentioned in the
Ruqʿāt-i ʿĀlamgīrī [Epistles of Aurangzeb] in a letter instructing the paymaster to make a record
of his having been promoted (Ruqʿāt-i ʿĀlamgīrī 1908, 159). My thanks to Prof. Muzaffar Alam
for this reference.
⁸² Although later historians have often imagined Hindu administrators during the Mughal
period as having had an internal conflict between their personal identity as Hindus and their use
of supposedly “Islamic” Persian as the medium of administration, Rajeev Kinra has argued
compellingly that this expectation is anachronistic and unsupported by the evidence (Kinra
2015).
⁸³ MN 2004, 2:855. ⁸⁴ MN 2004, 1:181. ⁸⁵ MN 2004, 1:455.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

36       

al-Daulah, the Nawab of Awadh, who was Ārzū’s patron towards the end
of his life.
Ish: āq Khān,
: known by his title Mūʾtaman al-Daulah, was Ārzū’s first
noble patron in Delhi. His father, Ġhulām ʿAlī Khān, : had come from
Shustar in southwestern Iran and risen to the position of imperial
bakāwal [superintendent of the kitchen]. Ish: āq Khān : was born in
Delhi and himself rose to a position of trust in the imperial establish-
ment, namely as the emperor Muh: ammad Shāh’s Khān-sāmān: [house
steward or chief of staff]. This was the position he held for twenty-two
years until his death in 1152/1740–1. In an account of his having
interrupted a meeting between Muh: ammad Shāh and Nādir Shāh—the
latter upholding the pretense that Muh: ammad Shāh was his “guest” even
though he had just routed the Mughal army at Karnal—Ish: āq Khān : is
described as Muh: ammad Shāh’s at::ālīq or childhood protector/tutor.
Nādir Shāh was pleased with Ish: āq Khān’s
: answers to his questions,
and pronounced him worthy of being Wazir of India. According to Ārzū,
he had a poetic temperament, writing under the pen-name Ish: āq.⁸⁶ He
receives considerable, albeit expected, adulation in Ārzū’s Majmāʿ al-
Nafāʾis. Ārzū writes that he was in his service for “just over twenty years,”
which means that he must have entered Ish: āq Khān’s
: employ around the
same time as he relocated to Delhi the second time.⁸⁷
After Ish: āq Khān’s
: demise, Ārzū received patronage from the late
nobleman’s sons and maintained his associations with the most impor-
tant people in the empire. The eldest son, Mīrzā Muh: ammad (later called
Najm al-Daulah), was Ārzū’s patron for a decade until he died on the
battlefield in 1162/1750. His brother Mīrzā Muh: ammad ʿAlī Sālār Jang
took Ārzū to Lucknow and introduced him to the Nawab, Shujāʿ
al-Daulah, who was also his brother-in-law.⁸⁸ Shujāʿ al-Daulah was the
son of Safdar

Jang, who had emigrated from Khurasan to India. Soon

⁸⁶ MN 2004, 1:181–2. The brief account of him in Maʾās̄ir al-Umarā unusually cites a couplet
of his (trans. Beveridge and Prashad 1979, 1:690).
⁸⁷ For biographical details, see Maʾās̄ir al-Umarā 1891, 3:774–6 (trans. Beveridge and
Prashad 1979, 1:690–1). On the meeting with Nādir Shāh, see Irvine 1922, 2:354. On his poetry,
see MN 2004, 1:181–5. Sāʾib

Tabrīzī (d. 1086/1676) and Jalāl Asīr Is: fahānī (d. 1049/1639) were
both influential poets of the recent past. They both referred to the “freshness” [shīwah-yi tāzah,
etc.] in their style (see Chapter 3).
⁸⁸ Khulā
: s:at al-Afkār f. 31a.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

   37

:
afterwards Sālār Jang became imperial bakhshī [paymaster] under the
new emperor Shāh ʿĀlam.⁸⁹ In Lucknow, Ārzū began receiving a
monthly stipend of 300 rupees from Shujāʿ al-Daulah. The arrangement
was cut short, however, when Ārzū died just over a year later.

Rivals and Enemies

Ārzū’s conflicts with his contemporaries are of more than just biograph-
ical interest because they demonstrably shaped Ārzū’s development as a
scholar and reveal the vibrancy of the intellectual scene. The literary
battles waged in Delhi were a tremendous inspiration for Ārzū, and he
turned from relatively neutral critical projects that he had been pursuing
(such as his commentary on Saʿdī’s Gulistān—a text that was read by
nearly every early modern student of Persian in the world) to more
polemical works. He was compelled to face the question of poetic
authority, framed as differences between Ancient and Modern poets.
Chapter 3 considers the substance of these debates, but this section gives
an account of the dramatis personae.
To set the scene, we should observe that major poets were constantly
generating corrections of others’ work, and this practice often developed
from informal critiques into well-defined factions or lineages. There were
those like Zāhid ʿAlī Khān
: :
Sakhā, whose entry in Majmaʿ al-Nafāʾis
begins with praise but then takes a darker turn when Ārzū begins
suggesting corrections to his verses.⁹⁰ Recording the critique in this
way is obviously a play for poetic power and not the friendly advice it
might have been had Sakhā: been Ārzū’s own student (or had the critique
been offered in another forum). Rival networks of poets were shown to
have inherited the faults of their teachers: For example, ʿAbd al-Lat̤īf
:
Khān Tanhā was a linchpin of one such poetic lineage. Ārzū implies that
he did not know Tanhā personally, but many of his close friends met
Ārzū [baʿẓī az yārān-i makh
: s:ūs: -i ū bā faqīr mulāqāt nimūdah], and they

⁸⁹ On the importance of the position of bak: hshī see Richards 1996, 63–4.
⁹⁰ He was an Iranian who came to India during the unrest in Afghanistan. Ārzū had met him
twice (MN 2004, 2:703). Hazīn’s

tażkirah mentions him and notes that his father had stayed
with Hazīn

(Tażkirah-yi Hazīn

1955, 95).
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

38       

reported that he was melancholy and a heavy drinker. Some accused him
of being abstruse in his poetry—apparently a characteristic of the work of
his own teacher Mīrzā Jalāl Asīr (d. 1049/1639)—and Ārzū does not
mount a defense, damning him with silence. He merely mentions that
Tanhā has two important students in India. They are so devoted to his
style [t::arz] that they think of it, according to Ārzū, as “the primary
tradition.”⁹¹ One of them was the Afghan Muh: ammad Niz̤ ām Muʿjiz
:
(d. 1162/1748–9) and the other was Nusratallāh :
Khān Nis̄ār, who was of
Iranian ancestry. Ārzū knew Muʿjiz personally and found his style
problematic, but was more positively disposed towards Nis̄ār.
A more serious network of rivals was centered around Muh: ammad
ʿAlī Hazīn,

who had arrived in India in 1734. Hazīn’s

sometime traveling
companion, ʿAlī Qulī Khān : Wālih (d. 1169/1756), known as Wālih
Dāġhistānī—his name implies a connection to Dagestan in the
Caucasus but he was born in Isfahan—was deeply shaped by Hazīn

although he had disclaimed his friendship with him by the time he had
finished writing his tażkirah Riyāẓ al-Shuʿarā [The Garden of Poets,
1161/1748].⁹² While by then no longer associated with Hazīn,
he was
no friend to Ārzū (although his tażkirah entry on Ārzū is complimen-
tary). He was associated with Faqīr and Wafā, who was probably by
extension also Ārzū’s adversary. Another among Hazīn’s
students was
Mullā Bāqir Shahīd.⁹³
The most famous rivalry, which has attained the status of folklore in
some circles, was between Ārzū and Hazīn
personally. Hazīn’s

circum-
stances are important for understanding both his social position as a poet
and his demeanor. He had seen the Safavid establishment, of which he
was a member (and which was the source of his wealth and prestige),
utterly torn down by Nādir Shāh.⁹⁴ He arrived in Delhi a deeply frus-
trated man, having been chased around Iran and eventually having to
leave his homeland altogether. He writes in his memoirs that an English

⁹¹ “sunnat al-awwalīn” (MN 2004, 1:299).

⁹² This is a significant work because it was the first transregional tażkirah in over a century
(Kia 2011, 218). With over 2,000 entries, it is on the scale of Ārzū’s own MN which would follow
in a few years’ time.
⁹³ On Faqīr see Kia 2011, 228; on Shahīd see Tażkirah-yi Mardum-i Dīdah 1961, 167.
⁹⁴ See Kia 2009.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

   39

official had urged him to go to Europe instead of India, but he chose to

continue on to India.⁹⁵ When he arrived in Sindh, he hoped he would not
be recognized. He traveled through Sindh and Punjab before arriving in
Delhi, where he stayed a little over a year before returning to Lahore. In
1738, with Nādir Shāh’s troops massing in Peshawar, Hazīn
fled with a
troop of hired bodyguards via Sirhind to Delhi where, unfortunately for
him, Nādir Shāh’s troops headed next. In Delhi, he witnessed the fall of
the city and the general massacre of its inhabitants. He wrote his
memoirs there in 1742, feeling unwell and melancholy over the destruc-
tion of the Safavid dynasty and the life he had known. He presided over
lavish poetic gatherings in Delhi, according to Dargāh Qulī Khān, : but
eventually retired to Benares where, according to Hākim,

who met him
twice there, he lived simply and wore a Sufi robe.⁹⁶ Why was Hazīn

accorded the respect he received from so many people in India? Ārzū,
who is obviously not a neutral party, gives us little notion in Majmaʿ al-
Nafāʾis. Three possibilities are Hazīn’s

spiritual authority (he is related to
:
Shaikh ʿAlī Wah : dat of Lahijan, for example), his connections with the
erstwhile Safavid court, and his poetic lineage. The idea that he was
popular in Delhi because he was an Iranian, and that Indian Persian
users were desperate for an Iranian to sort out their failing Persian, falls
well short of a satisfactory explanation.
How were the literary debates in Delhi of this period remembered?
Āftāb Rāy, who wrote in the late nineteenth century and thus took for
granted that Indo-Persian was inferior to Iranian Persian, declares that:
“Whenever [Ārzū] made objections against Shaikh : Hazīn,

he was in the
wrong. Envy is a calamity that makes a man blind.”⁹⁷ This is a statement
of opinion more than fact from well over a century after the events, but it
is nonetheless historically instructive. For Āftāb Rāy, no Indian Persian
writer could stand up to a well-regarded Iranian. Some scholars have

⁹⁵ Hazīn’s

own account of the journey (Life of Sheikh Mohammed Ali Hazin 1830, 251ff.;
analysis at Kia 2011, 186ff.). It is fascinating that Āftāb Rāy’s tażkirah Riyāẓ al-ʿĀrifīn [Garden
of the Wise, 1883] makes Hazīn’s

arrival in India sound accidental since it does not mention
that he was basically driven out from his home (Tażkirah-yi Riyāẓ al-ʿĀrifīn 1977, 1:188–9).
⁹⁶ Muraqqaʿ-i Dihlī 1993, 80; Tażkirah-yi Mardum-i Dīdah 1961, 66.
⁹⁷ “har chand dar bāb-i shaik: h-i lah: ījī kih niwā-hā bar āwardah ammā yaksar k: hārij az
āhang ast. h: asad bad balāʾī ast kih ādam rā kūr mī kunad” (Tażkirah-yi Riyāẓ al-ʿĀrifīn
1977, 8).
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

40       

made the sensible case that Hazīn

diluted Ārzū’s prestige and that this
was the main reason behind his hostility towards the rival master poet.⁹⁸
However, Ārzū’s own case against Hazīn

rested on real stylistic disagree-
ments: Preferring the style of the so-called Ancients, Hazīn

denigrated
recent developments in poetry, positioning himself as able to adjudicate
the difference in quality. By contrast, Ārzū supported contemporary
poetic innovation as an extension of rather than a departure from the
tradition, and accused Hazīn

of inconsistency and capriciousness in his
aesthetic judgments. The effects of this disagreement rippled outwards in
Delhi’s literary community. Qizilbāsh Khān : Ummīd defended Ārzū
when, according to Ārzū’s telling, he received a barrage of criticism
after releasing Tanbīh al-Ġhāfilīn.⁹⁹ Hākim,

who had himself been at
the receiving end of Ārzū’s sharp jabs, merely writes that some of Ārzū’s
criticisms in Tanbīh al-Ġhāfilīn are unfair.¹⁰⁰ Nonetheless, he quotes
Majmaʿ al-Nafāʾis’s back-biting entry on Hazīn
nearly in full.
This section considers the personal enmity between Ārzū and Hazīn

so that we can reconstruct their substantive disagreements in Chapter 3.
In Majmaʿ al-Nafāʾis, Ārzū offers no assessment of Hazīn’s

poetry, which
is one of the most vicious things that the genteel standards of tażkirah-
writing allowed one to do.¹⁰¹ Instead he concentrates on summarizing
Hazīn’s

flight from Nādir Shāh. He practically rants that Hazīn

had been
an ungrateful guest in India after being welcomed (and provided for
from the imperial treasury) and that his purpose in writing a particular
treatise was to denigrate everyone there, “from the beggar to the king.”
Ārzū notes that while Hazīn

was free with his criticism of India before,
with the passing of time, he has accepted that there are consequences to
insulting his hosts. Ārzū is wonderfully sarcastic on this point: “Thank
God that now all the faults of India have turned into graces!”¹⁰² Ārzū
continues by questioning whether Hazīn
had actually written three
dīwāns before the most recent, which he claimed as his fourth (the others

⁹⁸ Kia 2011, 208ff. ⁹⁹ MN 2004, 1:169–70.

¹⁰⁰ Tażkirah-yi Mardum-i Dīdah 1961, 66.
¹⁰¹ MN 2004, 1:379–80. Further analysis of the numerous ways in which Ārzū slights Hazīn

in MN is Kia 2011, 196ff.
¹⁰² “ammā al-h: amdu li-llāh kih alh: āl hamah-yi qabāʾih: -i hind bah h: asanāt badal shud” (MN
2004, 1:379).
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

   41

having been lost in the bloodshed). He recounts the bitter rivalry

between Mīr Muh: ammad Afẓal S̄ābit and Hazīn.
Hazīn’s

apparent
plagiarism was called to the attention of Delhi’s poetic community by
S̄ābit’s son Muh: ammad ʿAz::īm S̄abāt after S̄ābit’s death in 1151/
1738–9.¹⁰³ Ārzū defended both of them.¹⁰⁴ A strange incident when the
poet Girāmī came to Hazīn

with twenty raucous friends and left the
: confused is also related by Ārzū in the entry on Girāmī.¹⁰⁵ These
Shaikh
personal spats and how they are reported are our main source of
knowledge about what Hazīn
stood for: He was a prolific author of
prose tracts but apparently not of literary criticism and whatever critical
works he wrote (with the exception of a brief tażkirah) have probably
been lost, making it difficult to reconstruct his aesthetic program.¹⁰⁶
Hazīn’s

prose works may have been difficult to obtain even during his

¹⁰³ Wālih enters the fray—and demonstrates his complex relationship with his former
friend—by quoting about half of the couplets to which S̄abāt objected, as well as a large
proportion of Ārzū’s tract against Hazīn,
Tanbīh al-Ġhāfilīn (Tażkirah-yi Riyāẓ al-Shuʿarā
2005, 1:632ff.). After devoting pages to these attacks on Hazīn,
he nonetheless avers that
Hazīn

is among the greatest poets of his time [dar īn juzw-i zamān sar-āmad-i suk: hanwarān].
In his entry on Ārzū (Tażkirah-yi Riyāẓ al-Shuʿarā 2005, 1:347) he notes that Ārzū objected to
500 of Hazīn’s

couplets as “disjointed” [nā-marbūt::], a fundamental violation of poetic norms.
¹⁰⁴ He also defends a poet called Rāsikh, : although against “Indian Persianists” who objected
to a technical point in one of his couplets (MN 2005, 82).
¹⁰⁵ MN 2004, 3:1362–3. Ārzū describes Girāmī, the son of a Kashmiri nobleman, somewhat
ambivalently as being “very far-seeking [i.e. creative?]; he is greatly [concerned] with making of
new words and fresh meanings, even from a different language like hindī or ‘European’
[firangī]” and so open-minded, the proverbial “Sunni with a Sunni and a Shiʿa with a Shiʿa”
(and, extending the saying, he is even “a Yogi with a Yogi, a Christian with a Christian, a Jew
with a Jew, and a Hindu with a Hindu”), that he lacks his own “religion and sect.” Girāmī,
several sources including Ārzū suggest, had a reputation for uncouth behavior and being
indulged by hangers-on (personal communication with Nathan Tabor, October 2014). When
Hazīn

first arrived in Delhi, Girāmī went to see him with a group of some twenty students and
friends, and started to recite. Or rather—here Ārzū appears to be telling the story from Hazīn’s


perspective—Girāmī “raised a clamor up to the dome of heaven in an uncouth accent.” Hazīn,

we are told, could only sit in stunned silence. Mukhli : s: also describes Girāmī and his father
(Mirʾāt al-Is::t:ilāh: 2013, 496–8).
¹⁰⁶ His tażkirah, Tażkirat al-Muʿās: irīn [Tażkirah of (my) Contemporaries, 1165/1751], has
been published as Tażkirat al-Muʿās:irīn 1996 and Tażkirah-yi Hazīn
1955. It contains just a
hundred entries and provides little in the way of literary criticism. The poets included are all
Iranians and mostly from Isfahan, in stark contrast to the inclusive project of Ārzū’s MN
released the same year, which contains over 1,500 entries on poets from around the Persian
cosmopolis. Mahdi Rahimpoor has identified and published a small tract by Hazīn
in which he
comments on 24 couplets from Khāqānī’s: qas: īdahs in response to Ārzū (Rahimpoor 2012, 241,
249–56). The arguments in that work are framed in ad hominem terms, namely that only
ignorant people with poor Persian (read: Indians) would raise objections to Khāqānī’s : verses.
On Hazīn’s

prose works, which included tracts on horses [farasnāmah], theology, ethics, and so
on, see Storey 1953, 1,2:845–7. Hazīn
completed a treatise on administrative ethics, Risālah-i
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

42       

lifetime. Such circumstances would explain why Ārzū reproaches him

with the objection—seemingly irrelevant to judging his merit as a poet—
that no work of Hazīn’s

about philosophy or theology [ʿilm-i h: ikmat wa
kalām] has come to Ārzū’s notice. The implication appears to be that
Ārzū objected to having to take Hazīn’s

reputation for scholarship on
trust and wanted to judge for himself as to whether Hazīn’s
work was
actually up to the mark. The aesthetic conversation between the two
major figures in Delhi is therefore one-sided, which is perhaps a reason
for later scholars to have interpreted these debates more as personality
clashes rather than disagreements over fundamental aesthetic questions.
There is no treatise in which Hazīn

lays out a case against Ārzū, and so
we must reconstruct his objections. In contrast, Ārzū develops the case
against Hazīn

across several works, considered in Chapter 3 alongside
the replies of Hazīn’s

proxies.
A lack of extensive primary sources outside of Ārzū’s treatises makes it
difficult to reconstruct how partisans of Ārzū and Hazīn
advanced their
arguments during the conflict or indeed how strongly they supported
their contender. One such proxy for Hazīn

is Siyālkot:ī Mal Wārastah. He
lived in Lahore and Delhi, but the circumstances of his life are something
of a mystery because he does not appear in many tażkirahs.¹⁰⁷ Wārastah
wrote at least one work critical of Ārzū that is still extant. It is titled
Jawāb-i Shāfī [The Categorical Answer], and is unusually specific about
the circumstances of its composition: Wārastah wrote it after a visit in
1163/1750 from Hakīm

:
Beg Khān, known poetically as Hākim,

who now
lived in Delhi but was originally a fellow Lahori. Hākim

brought a copy
of his dīwān, whose margins were filled with Ārzū’s criticisms. Wārastah
tried to answer Ārzū’s objections by writing Jawāb-i Shāfī in defense of
his friend.¹⁰⁸ But it is more complicated than that because Wārastah

Dastūr al-ʿUqalā [Essay on the Manners of the Wise], in Delhi in Rabīʿ II 1153/June 1740. In
noting the work’s place of composition, namely India, the colophon adds in Arabic “God save us
from it” [najānā-llāh minhā] (British Library Delhi Persian ms 1207, 55b).
¹⁰⁷ The fullest account of Wārastah’s life is ʿAbdullah 1967, 139–62, which draws on Gul-i
Raʿnā and other sources. Pellò notes that in his Tażkirah-yi Bīnaz̤ īr [The Unique Tażkirah]
ʿAbd al-Wahhāb Iftikhār
: (d. 1190/1776) defends his own ustād [teacher] Āzād Bilgrāmī against
Wārastah’s criticism by arguing that Hindus should not “imitate Muslims and get involved in
the Islamic sciences” (Pellò 2014). This line of criticism is surprisingly rare, as far as I can tell.
¹⁰⁸ Browne 1896, 234. Based on the incipit, this is the same text I consulted at the
Amiruddaulah Public Library in Lucknow under the title Jawābāt-i I ʿtirāẓāt-i Ārzū [Replies
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

   43

perhaps wrote a second work called Rajm al-Shayāt::īn [Stoning of the

Devils], which is frequently cited as a response to Tanbīh al-Ġhāfilīn,
Ārzū’s broadside against Hazīn.¹⁰⁹

The twentieth-century scholar Sayyid
ʿAbdullah, however, has concluded that Rajm al-Shayāt::īn is the same
work as Jawāb-i Shāfī.¹¹⁰ The identification of Rajm al-Shayāt::īn as a text
written against Tanbīh al-Ġhāfilīn is the strongest evidence that
Wārastah was a partisan of Shaikh : Hazīn—ʿAbdullah

makes precisely
that claim with Rajm al-Shayāt::īn adduced as evidence but then four
pages later observes that the Rajm al-Shayāt::īn does not in fact exist as a
separate work! Since Jawāb-i Shāfī is a defense of Hākim

(and not Hazīn)


against Ārzū, the case that Wārastah was a vocal defender of Hazīn

against Ārzū becomes more circumstantial.¹¹¹
It is beyond doubt, however, that Wārastah had a pro-Iranian orien-
tation since Iranian informants are the key to his lexicographical project,
Mus::t:alah: āt al-Shuʿarā [Idioms of the Poets]. In the preface, Wārastah
describes his research technique: Whenever he could not understand a
word or expression in his study of Persian poetry and could not find a
satisfactory dictionary definition, he would “appeal to the idiom-
knowers of the land of Iran.”¹¹² In other words, he would ask a Persian

to the Criticisms of Ārzū]. It is incorrectly attributed to Hākim

himself (Muradabadi 2000, 8).
The date for copying given by the catalogue, 1107 , appears to be a misprint for 1207 
(= 1792 ).
¹⁰⁹ There was a work known by the name Rajm al-Shayāt::īn extant in the early nineteenth
century, but it is nowhere to be found in manuscript catalogues and has never been published. It
is mentioned, for example, in an obscure tażkirah by Qāẓī Muh: ammad Sādiq Akhtar : called
Āftāb-i ʿĀlamtāb [The World-Inflaming Sun, 1269/1853]. The tażkirah is massive, describing
some 4,000 poets, but the only known copy is held in a private collection in Farrukhabad, Uttar
:
Pradesh. Although Akhtar appears to endorse an idea that is crucial in Ārzū’s work, namely that
Indians can show the same talent in literary Persian as Iranians, he sides with Wārastah in the
particular context of Hazīn

versus Ārzū. He calls Rajm al-Shayāt::īn “extremely good” [bisyār
k: hūb] and states explicitly that it is a response to Ārzū’s Tanbīh al-Ġhāfilīn (Qasmi 2008,
355–8).
¹¹⁰ ʿAbdullah 1967, 145. Cyrus Shamisa concurs, but is concerned by the fact that the phrase
“stoning the devils” does not appear in any manuscript of Jawāb-i Shāfī (Shamisa 2001, 30). The
most accessible text of Jawāb-i Shāfī is Cambridge University ms Add 795, ff. 109–30.
Remarkably, it is bound together with Ārzū’s dictionary Chirāġh-i Hidāyat. It is also available
in the Delhi Persian manuscript collection at the British Library (Delhi Persian 428g).
¹¹¹ Three other works by him, though not relevant to the present discussion, are worth
mentioning: They are an anthology of couplets arranged by keyword (Sprenger 1854, 146), a
selection of ornate prose compiled by him (Rieu 1879–83, 1006–7), and a treatise on rhetoric,
which has been published (Mat::laʿ al-Saʿdain 1880).
¹¹² “rujūʿ bah muh: āwarah-dān-i īrān diyār” (Mus::t:alah: āt 2001, 37).
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

44       

speaker from Iran.¹¹³ This is certainly different from Ārzū’s textually

based approach, but it is not a matter of natural language—Wārastah’s
object is literary language, specifically, he says, that of the “tāzah-
goʾiyān.” Blochmann argues that this proves Wārastah lived in Iran for
the fifteen years he says he was engaged in research, but such a conclu-
sion ignores the fact that there were many Iranians in India whom he
could have asked, and indeed there is no evidence he ever traveled.¹¹⁴
Wārastah’s objectives were considerably different from those of Mukhli: s:
since when it came to gathering data, Mukhli : s: was interested in docu-
mentary and social practices, while Wārastah was narrowly concerned
with the poetic practices of certain people. In fact, Wārastah cites the
authority of Iranian speakers [muh: āwarah-dān-i īrān] only when a word
or phrase does not appear in one of the dictionaries he had at hand, and
he still feels the need to provide poetic quotations for virtually every
entry. This becomes problematic if we assume an unbridgeable gap
between Iranian Persian and Indo-Persian. For example, he cites
T̤ uġhrā several hundred times and although T̤ uġhrā was originally
from Mashhad in Iran, his poetry is famously studded with Indic
words. Indeed, Ārzū often cites him as evidence for how Indic words
are used in Persian. Wārastah’s attention to Indian-born Indo-Persian
poets is slight, but he does quote from Ārzū, Ġhanī Kashmīrī and S̄ābit.
His extremely sparing use of Ārzū’s works as a source was almost
certainly intentional. The text refers to Ārzū only perhaps a dozen
times (the published edition is nearly 800 pages), always as the “author
of Sirāj al-Luġhat” [s: āh: ib-i sirāj al-luġhat], and demonstrates that
Wārastah was familiar with several of Ārzū’s works.¹¹⁵ There are a
meager five references to Chirāġh-i Hidāyat, Ārzū’s lexicon of recent
words and expressions, the most relevant of his works for Wārastah’s

¹¹³ The obvious term in our own context for such a person, namely a “native speaker” of
Persian, is misleading in the pre-modern context (see Dudney 2017b).
¹¹⁴ Blochmann 1868, 30.
¹¹⁵ He cites Sirāj-i Munīr once (Mus::t:alah: āt 2001, 139, headword “bā kasī dast-o baġhal
raftan”). There is an elliptical reference apparently to Nawādir al-Alfāz̤ (discussed in Chapter 4),
or rather a text he calls Ārzū’s “essay of research into hindī words” [risālah-yi tah: qīq-i luġhāt-i
hindī] (Mus::t:alah: āt 2001, 171, headword “pażīrah”). It is possible that he means another work,
perhaps one now lost, since the relevant headword does not appear to be present in NA. There
are two references to Ārzū’s commentary on Saʿdī’s Gulistān, KhG : (Mus::t:alah: āt 2001, 203 and
453, headwords “pas kār nishāndan” and “rozgār ast”).
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

   45

project, and all but one of these are in the margins of the Lucknow
manuscript used for the critical edition.¹¹⁶ While he was definitely not
Ārzū’s friend, it is difficult to conclude more about his program from
what we know, which is that he only used Iranian informants.
In Delhi’s literary circles, loyalties and lineages were crucial. However,
this should not obscure the fact that there were aesthetic and philological
issues at stake that were larger than individual disagreements. As we
consider in the next section, Ārzū was particularly productive as a critic
towards the end of his life.

Ārzū’s Works

Before considering the arguments developed over the course of Ārzū’s

career, I must provide a sketch of his major works, focusing particularly
on his critical rather than literary compositions. The list of critical works
given in Ārzū’s entry on himself in Majmaʿ al-Nafāʾis in 1750–1 corre-
sponds to nearly all of the extant titles (his treatise Mus̄mir is missing
because it had not yet been written), but all of his poetry is listed
unhelpfully as a “kulliyāt” [complete works].¹¹⁷ The list given in the
preface to ʿAt::īyah-yi Kubra̍ [The Great Gift] provides a fuller account
of his poetic output, but contains some ambiguities and entirely leaves
out Ārzū’s lexicographical works.¹¹⁸ The most obvious explanation for
this omission, namely that Ārzū had not completed his dictionaries
before writing that text, is confounded by the fact that ʿAt::īyah-yi
Kubra̍ references Wārastah’s Mus::t:alah: āt al-Shuʿarā, which is a compar-
atively late text (possibly completed in 1752) and in fact itself references
Ārzū’s dictionary Sirāj al-Luġhat. Textual interpolation was common,
but it is a less than satisfying explanation in this case.
Ārzū’s surviving literary works are in a fragmentary state because
there is no surviving kulliyāt in which they are all collected. Tracing
his literary output is outside the scope of this study, but we know that he
wrote ġhazals, rubāʾīs [quatrains] and mas̄nawīs in Persian. The preface

¹¹⁶ Mus::t:alah: āt 2001, 42, 47, 69 [a longish quotation in the text], 132, 308.
¹¹⁷ MN 2004, 1:189. ¹¹⁸ ʿAK 2002, 50.
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

46       

to ʿAt::īyah-yi Kubra̍ lists five mas̄nawīs.¹¹⁹ Of these the only one which
has been traced and published is Shor-i ʿIshq [The Clamor of Love].¹²⁰
A mas̄nawī entitled Mihr-o Māh [The Moon and the Sun] is not men-
tioned elsewhere but Ārzū is noted as the author in the colophon.¹²¹ It is
thought that he completed seven Persian dīwāns, some of which were
later combined. All are rather rare.¹²² Whatever rekhtah : poetry Ārzū
may have written is in even greater doubt. Earlier sources, for example
Shafīq (d. 1808) in his Chamanistān-i Shuʿarā [Garden of the Poets],
:
believe there was a rekhtah dīwān.¹²³ By the late nineteenth century, the
idea had taken hold that Ārzū never collected his vernacular work (e.g.,
as is claimed in Āb-i Hayāt).

However, because no vernacular dīwān is
available we only have tażkirah quotations, which do not appear to be
particularly unusual or interesting.
We can divide Ārzū’s extant critical oeuvre into commentaries, dic-
tionaries, and general critical works (including his tażkirah) before
considering works of uncertain attribution in all categories.
Commentaries: These can be broadly divided into commentaries on
the Ancients and on the Moderns. Three in the first category are a

: wushgo
¹¹⁹ Kh : mentions a mas̄nawī on the theme of Mah: mūd and Ayāz responding to Zulālī
called “Husn-o

ʿIshq” [Love and Beauty] (however, this is apparently a mistake for “Shor-i
ʿIshq”), a Sāqīnāmah called “ʿĀlam-i Āb” [The Drinking Bout], an untitled mas̄nawī, another
called “Josh-o Kharosh”
: [Fervent and Tumult], a then unfinished mas̄nawī in response to S̄anāʾī,
25,000 couplets, a collection of rubāʾiyāt [quatrains], muk: hammasāt [cinquains], tarjīʿ-bands
[poetry based on refrains], chronograms, correspondence called Payām-i Shauq [Message of
Desire], and other mas̄nawīs (Safīnah 1959, 314–15). The prose works mentioned are miscel-
laneous prose, two essays on “maʿānī-o bayān” (obviously ʿAK and MʿU), SL, KhG, : the
Sikandarnāmah commentary and commentary on ʿUrfī’s qaṣīdahs. Kh : wushgo
: mentions
that Ārzū has accomplished all of this by age 47, which means that all of these works
can be safely dated before 1735.
¹²⁰ Ansari 1940.
¹²¹ Raza Library Rampur mss. 4327f. and 4328f. Oddly, the cataloguer refers to the text as
Muhr-o Wafā [Seal and Trust] even though the title is clearly written in the colophon (along
with Ārzū as the author). There is also a copy in Manchester University’s John Rylands Library
(Persian ms 620).
¹²² Some manuscripts include Aligarh Habib Ganj 47/77 (Razvi and Qaisar 1981, 207). Two
in the Asiatic Society’s Curzon collection are identical to each other and apparently with the one
noted by Sprenger, which was in imitation of Mirzā Shafīʿā Shīrāzī who also used the pen-name
As̄ar (d. 1702) (no. PCC 295 and PCC 296; Ivanow 1926–8, 212–13; cf. Sprenger 1854, 337–8;
Khuda Bakhsh Library 1970, 9:220). The only serious investigation of the dīwāns appears to
have been carried out by S.M. Asghar of Aligarh Muslim University. I have not been able to
review this work but some of his conclusions appear in Keshavmurthy 2012. Ārzū’s poetry
appeared in various collections [majmūʿāt], such as one in Berlin (Pertsch 1888, 118) and
another in Salar Jung ms 2359, which belonged to Tīpū Sult::ān (Ashraf 1966–88, 6:131–4).
¹²³ Chamanistān 1928, 6.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

   47

:
commentary on Saʿdī’s Gulistān called Khiyābān-i Gulistān [Road to the
Gulistān, 1708–9 revised 1738–9], another Shikūfah-Zār on Niz̤ āmī’s
Sikandarnāmah, and lastly Sirāj-i Wahhāj [The Blazing Lamp], a com-
mentary on Hāfiz
̤ .¹²⁴ Works dealing with the Moderns are a direct
contribution to the literary debates in Delhi after Hazīn’s

arrival, while
those on the Ancients are probably from earlier in Ārzū’s life and are
somewhat more abstract. Commentaries on Modern writers include one
on ʿUrfī’s qasīdahs,¹²⁵
: Sirāj-i Munīr [A Lamp for Munīr], Dād-i Sukhan:
[Justice in Poetry or A Gift of Poetry], and Tanbīh al-Ġhāfilīn [An
Admonition to the Heedless]. These last three works are discussed at
length in Chapter 3.
Lexicography: Ārzū’s Persian lexicographical project was in a sense
itself a commentary. His Sirāj al-Luġhat [Lamp of Language, 1147/
1734–5] was largely concerned with correcting the earlier dictionaries
Burhān-i Qāt::iʿ [The Decisive Proof, 1654] and Farhang-i Rashīdī (1652).
Chirāġh-i Hidāyat [Lamp of Guidance] is often referred to as the second
volume of Sirāj al-Luġhat (although it is only perhaps one-fifth the
length),¹²⁶ but it has the different goal of elucidating the usage not of
the Ancients but of the Moderns, namely words and phrases that do not

¹²⁴ The Sikandarnāmah commentary was published in the margins of the Bombay litho-
graph edition of the poem (1277/1860–1). Ārzū apparently wrote two different Sikandarnāmah
commentaries, Shikūfah-Zār covering the first part of Niz̤ āmī’s text and a second commentary
(without a specific name) covering the second, but it seems likely that Ārzū considered them a
single work. An edition of Shikūfah-Zār, based on the Aligarh Muslim University manuscript
(no. 89165525), has been published (Shikūfah-Zār 2013). I thank Owen Cornwall for sending it
to me. Other copies in manuscript are Raza Library, Rampur mss. 3965f. and 3985f.; National
Archives of India (Fort William College collection no. 145) where it is given the alternative title
Shigūfah-yi Rāz [Flowering of the Secret]; Hardayal Public Library, Delhi, no. 98; and Berlin, see
Pertsch 1888, 752, 764–5. I have not been able to check whether these mss. cover only the first
part of Niz̤ āmī’s text (like the published edition of Shikūfah-Zār) or the full poem. Khiyābān-i
:
:
Gulistān appears in a recent critical edition (KhG 1996) and is quite common in manuscript. Its
start date is fixed by chronogram (1119 ). Sirāj-i Wahhāj appears in Rampur ms 2452B, ff.
199b–207b, which was copied in 1236/1820–1 and contains a number of Ārzū’s works, and at
Aligarh as ms University 3 Farsi 119.
¹²⁵ There are three mss. of this commentary in the British Library (Delhi Persian 1286A,
1286C, and 1286D). The Salar Jung Library (Hyderabad, India) copy (ms 1765; Ashraf 1966–88,
5:70–1), dated 1803, bears the seals of the ruling family of Awadh.
¹²⁶ The date of SL is fixed by chronogram [yād būd sirāj al-dīn ʿalī k: hān] and presumably
since CH is often referred to as its second volume, CH was written sometime later. SL has never
been printed but appears in a very readable ms copy at the British Library as India Office Islamic
ms 1783 (= Éthé 2513). CH has been published since the nineteenth century in editions of
Ġhiyās̄ al-Luġhāt (a “student’s dictionary,” completed in 1242/1826 by Muh: ammad Ġhiyās̄ al-
Dīn Rāmpūrī). A modern edition is CH 1984.
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

48       

appear in older dictionaries. As I argue in the following chapters,

this attention to new usages and how they fit into a longer tradition
is characteristic of Ārzū’s thought. A third work is Nawādir al-Alfāz̤
[Wonders among Words, 1743], a dictionary of Indic words. Although it
does not define itself as a dictionary of hindī but rather is a dictionary of
Indic words used in a Persianate context—as we will consider in depth in
Chapter 4—it might in fact be the oldest critical dictionary of kharī : bolī
hindī, which is to say the vernacular usage of the Ganges-Yamuna plain
that yielded both Modern Standard Hindi and Urdu.¹²⁷ Taken together,
Ārzū’s lexicons represent an innovative and deliberate project of codify-
ing an expansive Persian literary culture: In Sirāj al-Luġhat, he reconsi-
ders the oldest part of the New Persian poetic tradition with novel
research techniques like the application of tawāfuq al-lisānain (the
recognition that some Indic and Persian words are the same) and thus,
in a sense, makes it new. In Chirāġh-i Hidāyat, he formalizes recent
developments in the tradition and gives them a certain stature, including
finding antecedents, thus making the new old. In Nawādir al-Alfāz̤ he
brings Indic languages into the discussion.
General critical works: Ārzū makes expansive claims of innovation in
his philological works. ʿAt::īyah-i Kubra̍ [The Great Gift], his work on
bayān [clear discourse], an aspect of rhetoric, claims to be the first
treatise in Persian to deal specifically with that subject.¹²⁸ Mauhibat-i
ʿUz̤ ma̍ [The Great Gift], his work on ʿilm-i maʿānī, which we can
imperfectly translate as “semiotics,” does not make such a sweeping
claim but seeks merely to “elevate” [afrāk: htan] the subject.¹²⁹ As the
titles suggest, the two are companion pieces (but were likely written a few
years apart).¹³⁰ Ārzū’s critical magnum opus, Mus̄mir [Fruitful], deals
with “the science of the elements of language” [ʿilm-i us: ūl-i luġhat] and is

¹²⁷ NA has been published in a critical edition as NA 1951.

¹²⁸ ʿAK 2002, 50. Āzād Bilgrāmī agrees on its originality (Sarw-i Āzād 1913, 228).
¹²⁹ MʿU 2002, 95.
¹³⁰ Lithographed together as early as 1832 (Storey 1953, vol. 1, pt. 2:836). The modern critical
edition, edited by Cyrus Shamisa, is ʿAK/MʿU 2002. As far as the dates, ʿAK was written before
MʿU (and probably before TĠh and DS since those would have certainly been mentioned in
preface had they been written). The early 1740s is the likeliest date of composition. MʿU refers to
Sirāj-i Wahhāj (pp. 107, 128), ʿAK (p. 172), KhG
: (p. 181), and TĠh (p. 128). Mention of the last
implies that MʿU was written in the late 1740s.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

   49

the subject of Chapter 2. Although based on an Arabic model, it is, as it

claims to be, the first text to apply those techniques to Persian and Indic
languages. Mus̄mir was almost certainly Ārzū’s final work since it refers
to every other of his extant critical texts, and is obviously the product of a
long career of thinking about the nature of language and literature. It was
possibly completed after he had left Delhi for Awadh, but there is no firm
evidence for this. Ārzū’s tażkirah, Majmaʿ al-Nafāʾis is preserved in
many collections in various fragmentary states.¹³¹ The last work men-
tioned in the catalogue of Ārzū’s oeuvre in Majmaʿ al-Nafāʾis is the text
itself, namely a tażkirah of Ancient and Modern poets together, which if
completed “by the grace of God” would contain some 40,000 couplets.
That translates to several thousand entries (though many of them
have little in the way of biographical information since the work was
intended as a safīnah—a selection of representative verses from different
poets¹³²—and only later repurposed as a tażkirah with biographies).
The most complete manuscript, that in the Khuda Bakhsh Library,
includes listings for 1,835 poets. It is apparently the only one with so
many, suggesting a process of revision (by Ārzū or one of his followers)
that we cannot trace. Majmaʿ al-Nafāʾis draws extensively on an influ-
ential Safavid tażkirah, Tażkirah-yi Nas: rābādī (completed 1091/1680),

¹³¹ Mahdi Rahimpoor argues that MN was started in 1129/1717, completed in 1164/1750–1,
and expanded until 1167/1753–4 (Rahimpoor 2008a, 343). Sprenger’s catalogue mentions an
edition with 1,419 entries and that a certain Mr. Hall had a good copy of the “second half” and
another in album shape and abridged (1854, 132–4). The two latter mss. made their way to the
India Office Library (Ethé 1903, 351–2). The ms in the Salar Jung Library also purports to be
complete (Ashraf 1966–88, 2:157–8). The Bodleian ms (Elliott 399) is written illegibly, but its
table of contents suggests that it contains around 1,400 poets (the catalogue gives precisely 1,419
but apparently on Sprenger’s authority rather than an actual count, Sachau and Ethé 1889, 255).
On the mss. see Storey 1953, vol. 1, pt. 2:834–40. A complete or nearly complete three-volume
critical edition (on the basis of three Pakistani mss. and the Khuda Bakhsh ms) has recently been
published in Islamabad (MN 2004). The recent Tehran edition (MN 2005) is just a recompila-
tion of Abid Reza Bidar’s 1970 edition, including his mistakes. Bidar’s edition is a selection of
109 poets out of 1,835 in the Khuda Bakhsh ms. His selection criteria were that the poets were
Ārzū’s contemporaries (that is, belonging to the twelfth-century ) and that some particulars
of their lives were given.
¹³² A safīnah appears to be more curated than a bayāẓ (a notebook used for recording
interesting poetry at mushaʿirahs), but it does not, as we can infer from Ārzū’s preface to MN,
include a complete set of biographical data on the poets as a tażkirah often does. However, the
fact that Safīnah-yi Kh: wushgo,
: a very biographically complete tażkirah, carries the title it does
undercuts this definition.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

50       

for information about earlier poets.¹³³ The influence of Majmaʿ al-Nafāʾis

in the eighteenth century is demonstrated by the fact that ʿAbd al-Hakīm

Hākim

(who had fallen out with Ārzū) quoted it extensively when he
prepared his own Tażkirah-yi Mardum-i Dīdah [Tażkirah of People
(I Have) Seen, 1761].¹³⁴ The list of works in ʿAt::īyah-yi Kubra̍ mentions
“other prefaces” among Ārzū’s works—we know of his prefaces to
Safīnah-yi Kh: wushgo
: : s’s
and Mukhli : Mirʾāt al-Ist:::ilāh: but he no doubt
contributed introductions to a number of his students’ dīwāns.
Non-extant works and works of uncertain attribution: ʿAt::īyah-yi
Kubra̍ mentions five works that cannot be traced beyond the simple
descriptions given. These are Risālah-yi Ādāb-i ʿIshq [Essay on the Art of
Love] described as “on research in the manners [ādāb] of love,” Miʿyār
al-Afkār [The Touchstone of Thoughts] “on the rules of conjugation and
syntax of Persian,” Payām-i Shauq [Message of Desire] “in answer to the
letters of the dear/excellent” (in other words, a collection of correspond-
:
ence), Gulzār-i Khiyāl [Rose-Garden of Thought] describing the Indian
festival of Holi, and Ābrū-yi Suk: han [Chief of Speech, or Honored in
Speech] “in description of tanks, fountains, fruits, and vines.” There are a
few manuscript texts that cataloguers have attributed to Ārzū for which
there is no confirmation of his authorship: For example, Zawāʾid
al-Fawāʾid [Increases in Useful Things], a dictionary of Persian infini-
tives [mas:ādir] and abstract nouns [mushtaqqāt] derived from them,
appears in the same manuscripts as other works by Ārzū but no author is
attributed in its introduction.¹³⁵ There is also a dictionary of Sufi terms at
Princeton University’s Rare Books and Manuscripts Library purporting

¹³³ Ārzū quotes Tażkirah-yi Nas:rābādī where relevant and sometimes provides commentary
on Nas: rābādī’s information. Tażkirah-yi Nas:rābādī itself absorbed material from two other
important seventeenth-century tażkirahs but unlike Ārzū’s work does not acknowledge them as
sources in the text (Fotoohi 2011).
¹³⁴ The title is a play on words since it could also be translated as the Tażkirah of the Pupil of
the Eye. Hākim’s

text has eighteen entries with quotations directly from MN (with a further
forty-eight in which no quotation from MN is present).
¹³⁵ For example, Raza Library (Rampur) ms 2520f., which is written in thick illegible
shikastah, contains Zawāʾid al Fawāʾid and Mus̄mir together. Another copy appears in the
Asiatic Society of Bengal’s Curzon Collection in a majmūʿah [miscellany] containing three
works by Ārzū, Zawāʾid al Fawāʾid possibly by Ārzū, and works by eleven other authors (no 969,
see Ivanow 1926–8, suppl. 2:27–30). Lastly, another majmūʿah in Iran contains Mus̄mir,
Zawāʾid al Fawāʾid, Dād-i Suk: han, a tract by Hazīn
:
on Khāqānī’s qas: īdahs and Wārastah’s
Jawāb-i Shāfī (Rahimpoor 2012, 243). Rahimpoor reports that there are scattered references to
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

   51

to be by Ārzū.¹³⁶ Likewise, a versified Persian grammar at the Asiatic

Society of Bengal is attributed to him.¹³⁷ However, Sirāj al-Dīn was
hardly an uncommon name, and there are several people with whom
Ārzū is easily confused.¹³⁸ Lastly, Ih: qāq al-Haqq

[Administering Justice,
or Establishing the Truth], another text criticizing Hazīn,

is not men-
tioned in either of Ārzū’s lists of his own works, but the mid-nineteenth-
century scholar Imām Bakhsh : Sahbāʾī

wrote a reply to a text of Ārzū’s
that he cited by that name.¹³⁹
Ārzū is now relatively well published and the major texts in his
philological oeuvre (with the exception of Sirāj al-Luġhat) have been

Ārzū’s other works in the first person (e.g., “as I wrote in Dād-i Suk: han”) in this manuscript
(Rahimpoor 2012, 215). Such references are reasonably strong evidence of Ārzū’s authorship
but it is not impossible to imagine them as interpolations. Because a text under the title Zawāʾid
al Fawāʾid is also attributed to ʿAbd al-Wāsiʿ Hānsawī, it is possible that authorship has been
wrongly assigned to Ārzū. However, there might in fact be two works in the same relationship as
another pair of works by ʿAbd al-Wāsiʿ and Ārzū: Since Ārzū’s NA is often catalogued as
Ġharāʾib al-Luġhāt, the title of ʿAbd al-Wāsiʿ’s work that he revised and expanded into NA, it is
possible that Ārzū actually wrote a Zawāʾid al Fawāʾid revising a text by ʿAbd al-Wāsiʿ. This is
Rahimpoor’s preferred conclusion.
¹³⁶ Islamic Manuscripts, Third Series no. 771. It is the second work bound in the codex (of
four total) and the title is given as “Intik: hāb-i Baʿẓī Is::t:ilāh: āt-i Sūfīyah”

[A Selection of Several
Sufi Expressions]. The text gives the attribution “Sirāj al-Dīn ʿAlī Khān” : and someone (in this
case Charles Rieu of the British Library) has assumed this to be Ārzū. There is no incontro-
vertible evidence for or against attributing this work to Ārzū. The inside cover reads “Presented
to Samuel Bochart, the celebrated traveller and linguist 22nd July MDCXLVIII (1648) by Ch
Tarravins,” but this perhaps only applies to the first text contained in the codex, which is
described as “Fragment of a treatise on Hindi dramas or ballets.” It is more interesting than the
dictionary of Sufi expressions since what is being described is not theater as such but apparently
rāsa-līlā [Krishna’s dance with the cowherdesses]. It uses Indic terms; for example, gopī [cow-
herdess] is glossed as “the beloveds of Krishna” [mah: būb-hā-yi kānhah]. Assuming that the
seventeenth-century date applies to this text, it is probably among the earliest examples of a
European scholar’s engagement with Hinduism. (Bochart, who lived from 1599 to 1657, was a
French Protestant philologist and Bible scholar.)
¹³⁷ Ivanow 1926–8, 387–8. The colophon calls the work Ārzū’s but the ms is incomplete so it
is difficult to judge the correctness of this statement.
¹³⁸ For example, Sirāj al-Dīn Husainī

Aurangābādī (fl. 1169/1755) used the tak: hallus: Sirāj.
He lived around the same time as Ārzū, and like Ārzū was a literary scholar—for example,
compiling selections from some 680 poets—and a poet in rek: htah (Sprenger 1854, 148–51).
Another poet with whom Ārzū is accidentally conflated is Sirāj al-Dīn Sirājī Khurāsānī
(d. 1254). This is somewhat surprising since he lived five centuries before Ārzū and indeed
his dīwān appears to be the oldest extant dīwān of any Indo-Persian poet (Ahmad 1972).
Another Sirāj al-Dīn ʿAlī Khān
: was from Mohan near Lucknow and used the tak: hallus: Mūjid.
He was appointed munshī to the Supreme Court in Calcutta and was a poet there (Sprenger
1854, 171; Rizvi 1986, 2:233–4).
¹³⁹ Rahimpoor 2008a, 345. Some scholars do not believe that Ārzū was the author of Ih: qāq
al-Haqq

(Khatoon 2004a, 96). On other works that have been attributed to Ārzū but for which
there is not good evidence, see Keshavmurthy 2012.
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

52       

made available. Still, a number of his titles do not apparently survive even
in manuscript (some of these may eventually turn up in the archives, but
we are unlikely to find anything truly surprising). Although recent
decades have seen a growing number of Ārzū’s works enter publication,
it is striking that more of his works were not published earlier given the
respect he commanded during his lifetime. The most influential of his
works still only in manuscript form is probably Sirāj al-Luġhat, which
appears in virtually every major Persian manuscript collection. In terms
of published works, the ones that were lithographed in the nineteenth
:
century are Chirāġh-i Hidāyat, Khiyābān-i Gulistān, the Sikandarnāmah
commentary, and ʿAt::īyah-i Kubra̍ and Mauhibat-i ʿUz̤ ma̍. It was only in
the late twentieth century that the most important philological works,
including Mus̄mir, were published. We now have the opportunity to
re-evaluate Ārzū with access to a much wider range of his writing.

Conclusion

This chapter located Ārzū within his overlapping networks of friends,

students, and teachers, and in relationship to rival poetic networks.
Although the complexity of interpersonal relations in this elite commu-
nity is unsurprising, it is important that we understand these relation-
ships as a basis for analysis of the substance of the debates to be discussed
subsequently. Ārzū began his life well positioned for a central role in the
literary society of his time because of his Sufi inheritance, his education
in Agra, and his youthful service under Prince Aʿz̤ am Shāh. Once he
reached Delhi, became Kh : wushgo’s
: friend, and entered Bedil’s circle, he
had the resources he needed to become a popular and influential teacher.
Thus, he consolidated the networks of both Bedil and Sarkh : wush
: around
himself. Arguably it was the particularly acrimonious debate over aes-
: Hazīn
thetics after the arrival of Shaikh
that pushed him from the safety
of commentaries on the classics to the reappraisal of the role of poetic
authority—and of the nature of poetic language—that characterized his
later works. While later critics have concentrated on the personality
politics of Persianate aesthetic bickering, with Hazīn

playing the role of
arrogant but ultimately justified Iranian exile and Ārzū that of the jealous
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

   53

Indian epigone, it is important to see the substantive disagreements

between them as well as the way in which they represented alternative
networks of poetic authority. Ārzū’s circle, as reflected in his Majmaʿ
al-Nafāʾis, was broad, and while many Indians flocked to Hazīn,

his own
tażkirah necessarily represents his community as exclusively Iranian
because it was a project to recover the Safavid cultural legacy.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

2
Ārzū’s Fruitful Theory of Language

Ārzū frames the subject of Mus̄mir, the text discussed at length in this
chapter, as an account of “ʿilm-i usūl-i
: luġhat” or “the science of the
elements of language,” which we can translate more simply as “phil-
ology.”¹ Ārzū drew upon centuries of Arabic and Persian scholarship to
build a philology that was anchored in this tradition with its various
grammatical, rhetorical, and historical sub-fields but far exceeded his
sources in considering multilingualism and other contemporary topics.
Mus̄mir represents arguably the most sophisticated early modern theory
of language in Persian. Unlike linguists of today, however, Ārzū was
concerned primarily with literary language and engaged with non-
literary language only insofar as it potentially influenced the former.
The logic of Ārzū’s philology also contains within it an aesthetic
program—the creation of good poetry—that is foreign to modern lin-
guistics, which attempts not to correct usages but merely to record them
as they are. Reconstructing Ārzū’s thought through Mus̄mir provides
insights into a Mughal-era conceptualization of language that is both
familiar to and radically different from our own understanding. I argue
that a hallmark of the intellectual practices of the global early modern
period was reverence for tradition, particularly for traditional categories
in knowledge systems, combined with the repurposing of such categor-
ies, sometimes radically transforming them. Mus̄mir is heavily invested
in a philological tradition that stretches back a millennium, but it vastly
extends the possibilities of that tradition: It is thus a paradigmatic early

¹ M 1. The editor of Mus̄mir has taken great pains to trace many of the references in the text,
including passages quoted directly from other sources, notably al-Muzhir and Farhang-i
Jahāngīrī. However, the text is written out in her own handwriting and poorly photocopied,
making it difficult to read. Its publication in 1991 came before computer typesetting was
available in Pakistan, and the publisher did not hire a professional scribe to prepare a fair
copy for reproduction. Furthermore, manuscript material that came to light towards the end of
the project was simply appended (again in poor-quality photocopy).

:
India in the Persian World of Letters: Khān-i Ārzū among the Eighteenth-Century Philologists. Arthur Dudney,
Oxford University Press. © Arthur Dudney 2022. DOI: 10.1093/oso/9780192857415.003.0003
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

̄̄’     55

modern text. It also represents the zenith of Ārzū’s own development as

a scholar.
Little can be ascertained about the history of the work itself. It is
undated, but since it makes reference to nearly every one of Ārzū’s
other extant critical works, we can surmise that it was compiled at the
very end of his life, perhaps during the brief period when he lived in
Lucknow before his death in early 1756.² In contrast to Ārzū’s practice in
other critical writing, he appears to cite no contemporary scholars in
Mus̄mir (although he quotes couplets by many recent poets, including
: Hazīn,
Shaikh
as evidence). The text seems to be unique in the Persian
tradition in that it is structured on an Arabic philological treatise that
was never—as far as I can tell—used as a model by anyone else writing in
Persian.³ The work in question is Jalāl al-Dīn al-Suyūt::ī’s al-Muzhir fī
ʿUlūm al-Luġhah wa Anwāʿihā [The Luminous Work Concerning the

² The extant manuscripts of Mus̄mir are not helpful for dating the work. The published
edition is based on two manuscripts at Aligarh Muslim University and one at the Raza Library
(Rampur). None of the three is complete. At Aligarh: Munīr ʿĀlam Collection 3/21 is 336ff. long
at fifteen lines per page and was copied in 1271/1854. University Collection Persian luġhāt 22 is
126ff. long with an average of twenty-six lines per page and is undated. It is written in a poor
shikastah in thick, illegible pen. At Rampur: Raza Library ms 2520f. contains Mus̄mir together
with Zawāʾid al-Fawāʾid [Increases in Useful Things], a lexicon of Persian verbs and nouns
derived from them that is attributed to Ārzū (see Chapter 1). The manuscript consists of 94ff. of
which approximately two-thirds are Mus̄mir. It is written in shikastah and undated. A second
Mus̄mir manuscript at Rampur has come to light, namely 2620f. The text is a brief fragment
(corresponding to the first forty or fifty pages of the printed version, with some omissions and
interpolations) and is written in an elegant nastaʿliq. Some 150 blank pages follow the text so we
can assume that a calligrapher was engaged to transform the other copy into something more
pleasant to read but the task was never finished. There is a manuscript in the Curzon Collection
of the Asiatic Society of Bengal (PCC 550), which is in poor condition and lacks a colophon
(Ivanow 1926–8, 387). Among the manuscripts I was unable to consult are the copy in the Zakir
Husain Library of Jamia Millia Islamia (Delhi), which is no. 1339 [C125/1] (Iran Culture House
1999, 329–30), and in the Punjab University Library (Lahore, Pakistan), no. AP1-2-15, a portion
of which has been appended to the published edition in facsimile. There appear only to be two
manuscripts in any collection outside of South Asia and both of these are in Iran: One is in the
Tehran University Library (Persian ms 4755), according to the union catalogue of Persian
manuscripts in Iran (Dirayati 2010, 8:1272). The other appears to be in the Astan Quds Razavi
Library in Mashhad in an anthology containing two works by Ārzū, one almost certainly by him
(Zawāʾid al-Fawāʾid), and one each by two of his rivals (Hazīn

and Wārastah) that was copied in
1246/1830 (Rahimpoor 2012, 248). Rahimpoor consulted the manuscript as a microfilm (no.
18631) and provides no details about the original, such as whether it is actually held in the same
collection.
³ The work, Al-Muzhir, came to at least one other Indian scholar’s attention in the nine-
teenth century. Nawāb Siddīq

Hasan

:
Khān of Bhopal (1832–1890) prepared an Arabic treatise
based upon it called Al-Bulġha fī Us:ūl al-Luġhah [A Sufficiency in the Elements of Philology],
about which see Haywood 1956. The work was published in Istanbul in 1296/1879 at the Jawāʾib
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

56       

Sciences of Language and Its Subfields].⁴ Ārzū’s title Mus̄mir, meaning

“fruitful” in Arabic, is meant to echo “muzhir” [luminous], and Ārzū
explicitly acknowledges his debt to al-Suyūt::ī in the preface and implicitly
through the text’s structure. Indeed, al-Suyūt::ī claims to be the inventor of
ʿilm-i usūl-i
: luġhat, the same term that Ārzū uses and which I am
translating as philology.⁵ Mus̄mir departs extensively from al-Suyūt::ī’s
model but the fundamental questions concerning (what we would call)
semiotics, phonetics, and morphology are mostly engagements with al-
Muzhir. Ārzū’s key contribution in these sections is to bring Indic
language into the discussion and presumably to make the Arabic trad-
ition better known to Persian-using Indians, many of whom lacked a
strong grounding in Arabic. A particularly important achievement for
Ārzū, as he himself notes, was to be the first scholar writing in Persian to
invoke al-Suyūt::ī’s concept of tawāfuq al-lisānain [correspondence of
languages], which is a rigorous way of thinking about how the same
words can appear in two or more languages. It is likely that Ārzū chose
al-Muzhir as his model precisely because its broad scope, as compared to
other philological or rhetorical treatises, allowed him to make so many
interventions.
Al-Muzhir’s author, Jalāl al-Dīn al-Suyūt::ī (849–911 /1445–1505 ),
was a Cairene polymath thought by modern scholars to be “the most
prolific author in the whole of Islamic literature”; his corpus has been
estimated to include nearly 1,000 works.⁶ Like many intellectuals in

:
Press by Ahmad Fāris Shidyāq, an important figure in the Nahda (the late nineteenth-century
Arab “Renaissance” movement). The work is a remarkable artifact of cosmopolitanism since the
first half is largely an abridgement of Al-Muzhir and the second half is a bibliography of Arabic,
Persian, Turkish, and hindī dictionaries and philological works. It does not appear to engage
with Ārzū’s scholarship in any way, but it does praise the Indian languages Persian and Urdu
(Haywood 1956, 172). I thank Professor Barbara Metcalf for pointing me to Al-Bulġha.
⁴ Al-Muzhir 1998. A Polish scholar has translated the first chapter (Czapkiewicz 1989) but
the non-native English presents an obstacle to interpretation. That al-Muzhir remained import-
ant and even came to the notice of Europeans is proved by the preface to E. W. Lane’s Arabic-
English Lexicon (1863), which notes that al-Muzhir is “a compilation of the utmost value to
students in general, and more especially to lexicographers.”
⁵ “I have originated the science of the principles of language [us:ūl-i luġhat] and its study, and
nobody has preceded me in this. It follows the same lines as Prophetic tradition and principles
of jurisprudence” (quoted in Sartain 1975, 70).
⁶ Geoffroy 2002.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

̄̄’     57

the Arab world in this period, he was of Persian descent.⁷ His father’s family
had settled in Baghdad and later the Upper Egyptian town of Asyūt:: (the
origin of his name). Like his father, he became a well-known jurist in the
Shāfiʿī legal school and his works were read across the Islamic world,
including in South Asia.⁸ His juridical training and reputation matter a
great deal because the study of the philosophy of language in an Islamicate
context was intimately connected with the techniques for research in
Islamic law [fiqh].⁹ Al-Muzhir, like Mus̄mir, is most likely the author’s
swan-song—the evidence being that it is not mentioned in his
autobiography—and so is the product of a lifetime of study.¹⁰ Al-Suyūt::ī’s
career was a high-water mark for Arabic scholarship, but his polymath
endeavors are also a reflection of philology’s interdisciplinarity.¹¹ Al-
Muzhir is actually structured as a compilation of other people’s opinions,
which al-Suyūt::ī adjudicates (or in some cases, such as in the divine or
conventional nature of language, he takes no personal position). This
commentarial structure in which an explication of previous scholars’
views is followed by the author’s own interpretation is precisely the pattern
that Ārzū adopts in those sections of Mus̄mir derived from al-Suyūt::ī. While
the subject of al-Muzhir is a specifically Arabic philology, al-Suyūt::ī none-
theless uses Persian examples when appropriate. For example, in order to
establish that multiple words exist for the same concept, he points out that
“mard” [man] and “sar” [head] in Persian have the same meaning as Arabic
“insān” and “rāʾs.”¹² Ārzū builds on al-Suyūt::ī’s awareness of Persian to
generalize his theories into a framework that addresses primarily Persian
but also has Indic language as an integral component. Ārzū’s ambition to

⁷ On his father’s side. His mother was supposedly a Circassian slave who gave birth to him in
a library, which led to his receiving the odd nickname Ibn al-Kutub [son of the books] (Geoffroy
:
2002). Ibn Khaldūn (732–808/1332–1406), a Maghrebian scholar who lived a couple of gener-
ations before al-Suyūt::ī, noted that most great Muslim intellectuals, both religious scholars and
otherwise, were not Arabs (Muqaddimah 2005, 428).
⁸ Sartain 1975, 40, 48.
⁹ The grammarian al-Fārābī (d. 339/950) was the first to apply the word “qānūn” [law] to
grammar, probably under some Hellenistic influence since the Arabic word is derived from
Greek kanōn (Versteegh 1980, 21–2).
¹⁰ Sartain 1975, 107. His other important works on language are al-Iqtirāh: fī ʿIlm Us:ūl al-
Nahw: [Impromptu Account of the Science of the Rules of Grammar] and al-Ashbāh wa’l-
Nazāʾir

: [Resemblances and Similarities in Grammar] (Balhan 2001, 12). On al-Suyūt::ī
fī’l-Nahw
as a legal thinker, see Hernandez 2017.
¹¹ Loucel 1963–4, 58; cf. Cantarino 1975, 1. ¹² Czapkiewicz 1989, 48.
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

58       

create a general theory of language is striking and ripples throughout his

works.
This chapter begins where the text begins, namely by addressing some
fundamental questions about the origins and extent of Persian literary
culture. It continues with Ārzū’s account of what we would call morph-
ology and phonetics, which are concerned with sound changes within
words: Which of these changes are meaningful and which are incidental?
Can people who speak local varieties of Persian be considered experts of
the language? The next section considers the connections between lan-
guages, of which the concept tawāfuq al-lisānain [correspondence of
languages] is crucial and has been of most interest to scholars. The
discussion concludes with a brief account of semiotics and particularly
of figurative language, important topics in a number of Mus̄mīrʼs chap-
ters. These themes too demonstrate Ārzū’s interest in how meanings may
have changed over time.

The Fruit of the Fruitful Tree

Ārzū’s Mus̄mir is self-evidently an important work, but its place in the

Persian intellectual tradition has not been settled. Although scholars have
recently become more interested in the text, much scholarship has
focused exclusively on its explication of tawāfuq al-lisānain (a concept
that explains how words in different languages are related), which is often
invoked as a totem of Indo-Persian scholarly achievement that can call
into question Orientalist assumptions that the Persian literary and critical
tradition was moribund by the time Europeans encountered it. However,
tawāfuq is addressed in only a small part of the text—albeit one that Ārzū
himself flags as important—and there is considerably more to Mus̄mir.
The various philological questions explored in the text from the broad to
the more abstruse are relevant to other literary and social questions, and
are worth considering if we are to have a complete picture of early modern
Indo-Persian views on language. Mus̄mir has never received a lengthy
treatment in any European language except in the published edition’s
English introduction and a one-off article in Italian. Scholars have written
about it in Urdu and Persian, though not, so far as I have been able to
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

̄̄’     59

discern, comprehensively.¹³ Many scholars have leaned on others’ opin-

ions that Mus̄mir is historically significant, rather than themselves
engaging with the text, its ideas, or its intellectual genealogy in the
tradition of Arabic philology.

A Brief History of Arabic Philology

In the same way that the explication of law in Islam was viewed as a
rigorous science, philology followed precise rules often derived from
methods of legal enquiry. The formulae “tahqīq : ān ast” and “dar
:
tahqīq paiwast” appear scores of times in Mus̄mir to introduce a defini-
tive argument. Both mean more or less “the truth as established through
research.” This is not literary interpretation as we typically think of it; its
rules and norms are unfamiliar to modern English readers both in
content and in rigor.¹⁴ Islamicate philology considers discourse primarily
on two levels, that of the utterance [lafz̤ ], which covers what we would
call phonetics and morphology, and of its meanings [maʿānī], either in
the abstract or as part of a syntactic unit (a literary trope or a sentence).¹⁵
Discourse in its broadest sense (utterances intended to produce a mean-
ing) is known as kalām.¹⁶ A similar term, bayān, literally means “clear

¹³ The most comprehensive treatment in Persian appears to be Mahdi Rahimpoor’s Bar

: wān-i
Kh : Ārzū (In Praise of Ārzū), a collection of essays in which two chapters deal exclusively
with Mus̄mir (Rahimpoor 2012, 109–28, 129–40). The editor’s introduction in M 1991 (in
English) helpfully—if sometimes misleadingly—summarizes the contents of each chapter. The
Italian article Pellò 2004 is narrowly focused on the question of tawāfuq and related topics in
M’s chapters 28, 29, and 32. On Mus̄mir generally, see passim Alam 2003, 175 and Kinra 2011,
374. Mus̄mir’s editor’s biography of Ārzū (in Urdu) contains a lengthy analysis of the text
(Khatoon 1987, 129–39, 157ff.). See also Husain 1940.
¹⁴ Recent work that has considered representation in Rajput and Mughal painting as a formal
system—without the previously all-too-common subtext that formalization presupposes a lack
of creativity— has been illuminating (Aitken 2010; Minissale 2006). From the disciplinary
perspective of comparative literature, a now slightly dated but evocative discussion of “literature
as system” is Guillén 1971.
¹⁵ For the Arabic grammarians, “lafz̤ ” or “qaul” is seen as merely a string of sounds in
contrast to “kalām” [meaningful utterance] (Cantarino 1975, 46; Versteegh 1993, 100). The
present chapter skips over the crucial distinction in Islamicate poetics between ʿilm-i lafz̤ ī [the
science of wording] and ʿilm-i maʿnawī [the science of (making) meaning], which are typically
considered subsets of ʿilm al-badīʿ (see Windfuhr 1974, 335–6).
¹⁶ Poetry and prose are considered not separate kinds of expression but as two different
aspects of “kalām,” which is to say that poetry = prose + meter (Grunebaum 1981, 2:336).
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

60       

speech,” although it too is often translated as “discourse.” The philo-

logical sub-discipline known as the science of discourse [ʿilm al-bayān]
has generally been framed as the study of speech acts, that is, whether
the implication of an utterance in fact corresponds to the situation it is
attempting to describe.¹⁷ Both bayān and maʿānī were seen by rhetor-
icians as falling under a broader category of balāġhah (often translated as
“rhetoric”), which is distinguished from the focus on form in morph-
:
ology [sarf] :
or syntax [nahw].¹⁸
The application of this philological framework to secular literature
and by extension to natural language is a departure from its original
function, which was qur’anic exegesis. Traditional Arabic philology had
just two acceptable sources for data—pre-Islamic poetry and the Qurʾān
itself. Hadīs̄

(the sayings and recorded practice of the Prophet of Islam)
was almost unanimously considered to be unusable as linguistic data
by classical Arabic philologists.¹⁹ There is a stark contrast between this
narrowly focused study of language and philology as it was practiced in
Ārzū’s time. However, philology continued to be guided by its origins
:
in hadīs̄ scholarship because the use of poetic quotations as examples of
permitted usage, known as asnād (sg. sanad), is derived from the concept
:
of a chain of transmission in hadīs̄ scholarship.²⁰ Within the philological
:
tradition (no less than for hadīs̄ scholars), it is crucial to be able to show
that a competent person was the originator of a particular usage (or
practice). A second important philological technique derived from reli-
gious scholarship is that of qiyās (usually translated as “analogy”).²¹ In

¹⁷ Versteegh 1997, 124. Of course, in this context “speech acts” refers not to spoken
communication but to literature. In that sense ʿilm-i bayān is the study of metaphorical
language.
¹⁸ See Baalbaki 1983. The term s: arf is commonly translated as “morphology” while nahw :
refers to grammar generally or in the narrow sense of syntax (on nahw: see Troupeau 2012). Sarf


and nahw : have been used in overlapping senses by different authors so drawing a clear,
universal distinction between them is not possible.
¹⁹ Suleiman 1999, 16–17. ²⁰ Al-Suyūt::ī says as much (Loucel 1963–4, 69).
²¹ See Versteegh 1997, 47; Versteegh 1980. Its complexity as a legal term is clear from Kholeif
1966, 150ff. Sībawayh’s (fl. late eighth century ) use of it implies that for him it was a term
without a settled meaning. In philology, qiyās is “based on drawing inferences about similar
accidental grammatical effects exhibited by divergent structures” (Marogy 2010, 27). Sībawayh
is probably responsible for transforming qiyās: In the early days it was a productive tool of
grammar (that is, a way of creating forms of words that are not attested in the sources) and
indeed even a basis for correcting the forms of words of the Qurʾān, but after Sībawayh it
became purely descriptive (Versteegh 1993, 37–9).
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

̄̄’     61

legal reasoning, qiyās is the process of determining whether a rule

defined with regards to a particular situation also applies in another
situation, such as the famous debate as to whether the Qurʾān’s prohib-
:
ition on wine [khamr] applies to other intoxicants, such as date-wine
[nabīż]. In philology, qiyās is the means by which the rules of language,
specifically the patterns of words, are established on the basis of other
word patterns. Al-Suyūt::ī defines grammar as “nothing but analogy
[qiyās] which is to be followed.”²² This analogical concept of language
is also the basis for Ārzū’s understanding.
Al-Muzhir draws from several different strands of Arabic linguistic
and aesthetic thought. Although untangling this interwoven intellectual
tradition is outside the scope of this chapter, there are a few key devel-
opments that are important for understanding Al-Muzhir and hence
Mus̄mir.²³ The tradition of grammar proper that crystallized in the late
ninth-century grammarian Sībawayh’s Kitāb [lit. Book] was the first
attempt at systematizing Arabic grammar. It was so influential that it
was widely quoted into early modern times.²⁴ Although Sībawayh was
from a Persian-speaking background, he settled in Basra (present-day
Iraq) and was later regarded as an important link in the transmission of
the work of the Basran school of Arabic philological thought. An
important turn in philology came when the Persian scholar Al-Jurjānī
(d. 1078) criticized Sībawayh for concentrating solely on word forms and
instead took the study of language in a more rhetorical direction (i.e.,
towards balāġhah). He in turn influenced the Khwarazmian scholar
Sakkākī (d. 1229) who divided up philology (which he called ʿilm al-
adab) into two grammatical disciplines, ʿilm al-s: arf and ʿilm al-nahw,
: and
two rhetorical disciplines, ʿilm al-maʿānī and ʿilm al-bayān.²⁵ His formu-
lation of bayān was particularly innovative. Lastly, lexicography (ʿilm al-
luġhah) was exceptionally important since dictionaries are frequently

²² “innamā al-nahw: qiyās yuttabaʿ” (translated in Suleiman 1999, 25ff.). The Arabic philo-
logists Ibn Fāris (tenth century) and Al-S̄aʿlabī (eleventh century) referred to the study of
language as “law” [fiqh] (Haywood 1960, 100).
²³ Versteegh 1997 is a comprehensive account of the Arabic grammatical tradition.
²⁴ See Versteegh 1997, ch. 3.
²⁵ Versteegh 1997, 123–4. Al-Muzhir appears uninfluenced by ʿilm al-badīʿ, the study of
rhetorical devices that is usually considered the third sub-category of rhetoric.
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

62       

quoted.²⁶ These developments in the tradition set the parameters in

which al-Suyūt::ī in the fifteenth century and Ārzū in the eighteenth
century added their own understanding of language to the existing
science of philology.

The History of Persian Literary Culture as

Described in Mus̄mir

Mus̄mir begins with a historical analysis derived from Farhang-i

Jahāngīrī [Dictionary for Jahāngīr, 1608] and Dabistān-i Mażāhib
[School of Religions, mid-seventeenth century], along with some signifi-
cant interventions by Ārzū. It seeks to establish where the name
“Persian” [fārsī/pārsī] comes from, what the dialects of Persian are
(both in the past and at present), whether it is theologically permissible
to use Persian as opposed to Arabic, and what the genesis of poetry was.
Taken together, Ārzū’s engagement with these questions goes to the
heart of the Mughal understanding of the history of Persian. He ratified
and in some cases refined the views expressed in an important intellec-
tual project of Akbar’s reign, namely Farhang-i Jahāngīrī. Although
many of the conclusions that Ārzū or the sources he cites had reached
are based upon historical and textual presuppositions that are no longer
tenable, it is nonetheless useful to examine them in detail because of how
they shaped the Mughal worldview.
When discussing the history of the Persian language, Ārzū primarily
relies upon Farhang-i Jahāngīrī but frequently compares it with other
dictionaries to establish alternative pronunciations and meanings.
Farhang-i Jahāngīrī was a remarkable effort in which the emperor
Akbar himself took an interest. It was compiled by the courtier Mīr
Jamāl al-Dīn Husain

Injū, whom Akbar had commanded during a halt in
Srinagar to compile “a book containing all the authentic Persian words,
archaic usages, and idiomatic expressions.”²⁷ (The dictionary is named
for Jahāngīr because death claimed Akbar three years before its

²⁶ On lexicography in Arabic, see Haywood 1960 and Baalbaki 2019.

²⁷ The relevant passage is carefully translated and interpreted in Kinra 2011, 369–70.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

̄̄’     63

completion.) Injū came from a noble family in Shiraz—his cousins were

high officials in the court of the Safavid ruler Shāh Tahmāsp I—and
spent time in the Deccan before entering Akbar’s service. He rose in the
imperial hierarchy and became a confidant to Jahāngīr. He even acted
:
as Jahāngīr’s envoy to his rebellious son, Prince Khusrau, and was
:
appointed governor of the sūbah (province) of Bihar towards the end
of his career.²⁸ His experience shows that for the Mughals, dictionary-
making was not an activity to be undertaken by low-paid secretaries but
could bring prestige to a member of the highest echelons of society. It was,
in fact, a hugely resource-intensive project. Injū used forty-four sources,
which he lists in his preface. Thus, to a large degree, he consolidated the
lexicographical tradition in his one book of some 9,000 entries. The
Farhang-i Jahāngīrī is not just a lexicon but also contains an elaborate
:
introduction [muqaddamah] and appendix [khātimah].²⁹ The introduc-
tion is a substantial work in twelve chapters, including a history of the
Persian language. A detailed description of the methodology of the lexi-
con’s compilation is provided in a separate preface. The appendix consists
of five sub-lexicons, one of which is a list of Middle Persian (“Zend and
Pazend”) words, which was clearly useful for Ārzū.³⁰ Injū tells us that he
relied on a Zoroastrian informant for the Middle Persian material. The
other main sources for Ārzū’s history of Persian are four of the major
dictionaries of the language then available: Majmaʿ al-Furs (1600, compiled
in Isfahan), also known as Farhang-i Surūrī after its compiler, Farhang-i
Rashīdī (1654), Burhān-i Qāt::iʿ (1652, compiled in the Deccan), and Ārzū’s
own Sirāj al-Luġhat [Lamp of Language, 1147/1734–5]. Ārzū discusses

²⁸ Beveridge and Prashad 1979, 1:742. Injū’s son apparently married the daughter of ʿAbd al-
Rahīm : Khān-i
: :
Khānān, the greatest literary patron of Akbar’s reign besides the emperor himself.
²⁹ Farhang-i Jahāngīrī’s odd feature for the modern reader is that it is arranged by the second
letter of each headword then by the initial letter. Indeed, the first page of the modern printed
edition contains a notice that an index, in proper alphabetical order, has been provided at the
end of the book for the reader’s convenience.
³⁰ The five sections of the appendix are (1) Metonyms and Expressions [kināyāt wa
is::t:ilāhāt],
: (2) Compound Words from Persian and Arabic [luġhāt-i murakkabah az pārsī wa
ʿarabī], (3) Words in Which Appear One of the “Eight Sounds” [luġhātī kih yakī az hurūf-i :
hashtagānah dar ān yāftah shudah], (4) Zend and Pazend Words [luġhāt-i zind wa pāzind], and
(5) Strange Words [luġhāt-i ġharībah]. In number (3), the “eight sounds” must refer to the
sounds native to Arabic but not found in Persian before the Arab conquest. In number (5),
“ġharīb” is being used in the technical sense, namely as a loan word or calque. This section
includes Indic words such as “āgrah” [the city of Agra] and “pānī” [water].
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

64       

these various dictionaries and declares Farhang-i Jahāngīrī to be the best,

even though Farhang-i Rashīdī and Burhān-i Qāt::iʿ are later and each
represents something of a repudiation of Farhang-i Jahāngīrī.³¹ The only
two non-lexicographical works drawn into the conversation are
Dabistān-i Mażāhib³² and Muhammad : ʿAufī’s Lubāb al-Albāb [The
Piths of Intellects, 1221 ]. ʿAufī’s text is regarded as the first Persian
tażkirah, although its author does not use that term (instead he calls it
“t::abaqat”). The Dabistān’s author was probably a Zoroastrian who went
by the pseudonym Mūbad Shāh but also used a Muslim name when it
suited him.³³ With its methods that strike us as proto-anthropological
(the author observes, and in some cases infiltrates, various religious
groups), this text is not a rigorous history, although it was taken as
such by later writers such as Ārzū. Instead, its historical descriptions
represent how a certain sect of Zoroastrians memorialized their trad-
ition. While Ārzū approaches the material critically, emending it as
necessary, his richly imagined history of Persian literary culture diverges
from our present understanding because of the constraints of his sources.
Nevertheless, it is a remarkable intellectual achievement. Let us explore
that here.
The origin of the name for Persian (pārsī or fārsī) is tied up with
ancient history, as described by Injū and mostly taken word for word by
Ārzū. Pārs son of Pahlaw son of Sām son of Noah [nūh] was king of a

³¹ M 39ff.; see also Kinra 2011, 373.

³² A complete but inaccurate English translation by Anthony Shea and David Troyer under
the title The Dabistān, or School of Manners was published in 1843. A chapter had earlier been
translated by Francis Gladwin (born 1744 or 1745, died 1812) so it was of interest to British
orientalists, and indeed the Persian text was first published in Calcutta in 1809 in moveable type
(Ali 1999). The first European to study the book was Sir William Jones in 1787. He was pleased
to learn of an “authentic” source of knowledge about Zoroastrianism since he had famously
dismissed as an obvious forgery the other important potential source, the Zend-Avesta. The
Zend-Avesta was brought to the attention of European scholars by Anquetil-Duperron
(1731–1805), who published it in 1771. Although Jones destroyed Anquetil-Duperron’s repu-
tation over the matter, it became clear soon after his death that Anquetil-Duperron had been
right and Jones wrong (Browne 1929, 45ff.).
³³ Shea and Troyers’s translators’ preface repeats the mistaken conclusion of Jones (which
had been repeated by Gladwin) that the Dabistān’s author was Muhsin : Fānī (d. 1081/1670), a
Kashmiri Sufi poet and friend of Dārā Shukoh. In fact, a quatrain of Fānī’s is quoted in the
preface but that of course is not evidence that he wrote the whole work. The identity of the
author and his role as “ethnographer in disguise” has been addressed by the late Aditya Behl
(Behl 2011; cf. Ali 1999; Sheth 2019).
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

̄̄’     65

territory that was named after him.³⁴ All of Iran was contained in
this ancient territory, which supposedly stretched from the Amu Darya
:
[rūd-i jaihūn] in the east to the Euphrates [āb-i furāt] in the west and
from the Gates of Alexander [bāb al-abwāb] (in Derbent, in present-day
Russia near the border with Azerbaijan) in the north to the Persian Gulf
[daryā-yi ʿumān] in the south. Within this great swathe of territory, the
eastern region was known as Khurasan because, Ārzū and Injū argue,
“khurāsān”
: means “East” in “ancient Persian” [pārsī-yi qadīm]—he does
not make the distinction between Old Persian and Middle Persian,
referring to all historical Persian as “ancient”—and this was the region
:
to the east of Persepolis [istakhr].³⁵ After the coming of Islam, the capital
was moved to Isfahan (400 km to the north) and the towns to the west
became known as “ʿirāq al-ʿarab” [Arab Iraq] and the nearer ones as
“ʿirāq al-ʿajam” [Persian Iraq]. The language of Pārs’s kingdom was
“pārsī” and the Arabs, unable to pronounce “p,” turned it into “fārsī.”
The Zoroastrian period in Persian history was apparently a golden age of
beauty, rhetoric, “cleanliness” [pākīzagī], good horsemanship, and so on.
Before the Persians were Zoroastrian, they were supposedly Sabian
[s: ābīʾah], the exact meaning of which is hard to reconstruct.³⁶ What
comes out of all of this is that the distribution of ancient Persian was
thought to track closely with that of contemporary Persian. However,
Ārzū adds a very important caveat, which does not appear in Injū: He
notes that although Persian is “the medium of (careful) speech and
writing” [madār-i takallum-o tarassul] across Iran and Turan (i.e.,
Central Asia), and to some extent India, the language of the masses
[ʿawāmm] across that region is often different. This insight allows Ārzū
to move beyond the idea that language distribution is historically static as

³⁴ M 2–4; Farhang-i Jahāngīrī 1975, 14–15. A similar discussion of Pārs appears in Ārzū’s
Gulistān commentary (KhG : 1996, 11). An alternative lineage for Pārs through Japhet [Yāfis̄] son
of Noah is also given.
³⁵ M 218–19 also discusses “k: hurāsān” as “east.”
³⁶ Sabians are mentioned in the Qurʾān (e.g., 22.17) alongside Jews, Christians, and Magians,
but the term seems to be used in various conflicting ways such as serving as a general term for
non-Muslims who nonetheless believe in God (Pedersen 1922). See also Roberts 2017 on
historically attested Sabians. Ārzū’s rival Shaikh
: Hazīn

claims to have encountered them around
Shustar in south-central Iran, where supposedly the last remaining population lived (Life of
Sheikh Mohammed Ali Hazin 1830, 160). Perhaps here it refers to the proto-Islamic religion of
Abraham.
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

66       

is implied by Injū’s conclusion that Persian is spoken where Pārs’s

ancient kingdom was. Theorizing the considerable linguistic variation
across territories even where Persian is natively spoken is key to Ārzū’s
conception of the Persian cosmopolis. Because the poetic language of
even native speakers may be imperfect—affected as it is by non-standard
local usages—non-native speakers have as much (or in some cases more)
ability to write Persian poetry as native speakers.
After concluding their discussion of how the language got its name,
Ārzū and Injū address the dialects of Persian. It has, they argue, seven
dialects [gūnah], of which three are current and four are historical
curiosities. The four extinct dialects are “hirawī,” “sakzī” or “sagzī,”
“zāwalī,” and “suġhdī.” They are extinct in the sense that one cannot
eloquently write complete sentences in them but if “one brings a word
[from one of them] into a couplet or a poem then this is permissible.”³⁷
Besides this brief reference, these four dialects are not mentioned again
in the text. Of the four, only “suġhdī” or Soghdian, finds a place in today’s
standard dialectology of Persian.³⁸ The three current dialects have the
familiar names “fārsī”/“pārsī,” “pahlawī,” and “darī,” but these have
different meanings from the ones we know today. As defined by Injū,
they are:

Parsi is simply defined as the poetic language of the territories of Pārs

whose capital was at Persepolis, the original city founded by Kiyomars̄
(the mythical first king in the Shāhnāmah and the first human in
Zoroastrian belief).³⁹ He includes a quote from Tafsīr-i Dailamī

³⁷ “dar baitī balkih dar ġhazalī agar yak kalimah biyāwarand rawā bāshad” (M 4; Farhang-i
Jahāngīrī 1975, 15). The transliteration of “zāwalī” is conjecture.
³⁸ “Hirawī,” according to Steingass, refers to an archaic dialect of Herat (in modern western
Afghanistan). On Soghdian, see Windfuhr 2009, 279. It is puzzling that Choresmian
[k: hwārazmī],
: the native language of al-Bīrūnī (d. post 442/1050), is not among these dialects
since al-Bīrūnī cites Choresmian words in his famous Chronology of Ancient Nations (Cereti
and Maggi 2005, 149). Perhaps Soghdian was thought to include all of what we now call the
northeastern group of Middle Iranian languages, of which Choresmian is a part. As is the case
for a great deal of Middle Iranian material, manuscripts in Soghdian were unknown for
centuries until they were rediscovered in the early twentieth century (Cereti and Maggi
2005, 101).
³⁹ Farhang-i Jahāngīrī 1975, 15–16; M 4. Ārzū discusses Kiyomars̄ in his dictionary Chirāġh-i
Hidāyat (CH 1984, 1204).
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

̄̄’     67

[Dailamī’s Commentary, fourth/tenth century] supposedly spoken by

:
God to the Prophet Muhammad in this language.
Dari has had several related meanings in history, including being
defined from as early as the tenth century as New Persian literary
language generally (especially in contrast with Pahlavi or Middle
Persian).⁴⁰ Some people, writes Injū, attribute special eloquence to
Dari because it is apparently the dialect of Persian in which full forms
of words are always used.⁴¹ Another possibility is that it is a dialect of
Balkh, Merv, and Bukhara (the cities are located in a region today split
between northeastern Afghanistan, Uzbekistan, and Turkmenistan).
Injū’s preferred explanation is that it was the language of the imperial
court [dargāh-i kayān], presumably meaning the Sassanian Empire’s
capital Ctesiphon (near present-day Baghdad). The quotation from
:
Injū includes a supposed reference to hadīs̄ which says, improbably,
that the “Arabic language of the people of Paradise [jannat]” is
actually Dari Persian.⁴²
Pahlavi, according to Injū, might be named for Pārs’s father Pahlaw
who was connected with the region known as Pahlah (which consists
of Isfahan, Ray, and Dinawar in western Iran). Another possibility,
which Injū prefers, was that since “pahlaw” can mean “city” (he cites a
couplet by Firdausī as evidence), Pahlavi was thus urban speech. We
now understand Pahlavi to be Parthian, a Middle Persian prestige
dialect which was written in a script derived from Aramaic under the
Parthian Empire (248 –224 ) and adopted by the Sassanians
(224–651 ). In imprecise usage, the term “Pahlavi” can also refer to

⁴⁰ On historical meanings of “darī” (which encompass all the meanings proposed by Injū),
see Lazard 1993 and 1995, and Paul 2005.
⁴¹ Persian has a number of words with a longer and a shorter form, generally maintained for
metrical purposes. Thus Dari prefers “ishkam” [belly] to “shikam,” “ba-raw” [go!] to “raw,” “ba-
go” [speak!] to “go.” Instead of the second, Ārzū uses the example of “abar” [upon] instead of
“bar.” In the eleventh-century etiquette manual Qābūsnāmah, secretaries are warned against
using “unrestrained Persian” [pārsī-yi mut::laq], presumably meaning Persian without Arabic
words, and especially “the Dari Persian which is not well-known” (Qābūsnāmah 1956, 187;
Hanaway 2012, 102).
⁴² Ārzū adds an additional possibility derived from Majd al-Dīn ʿAlī Qūsī’s dictionary (M 8).
Ārzū consulted that lexicon when compiling Sirāj al-Luġhat and praised it highly in the preface,
but it is apparently no longer extant.
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

68       

Middle Persian in general, especially its continued liturgical usage by

Zoroastrians after the Sassanian Empire’s collapse.⁴³

This triad of Parsi/Dari/Pahlavi is, in fact, attested from such an early

moment that it predates the formation of New Persian: It goes back at least
to the eighth-century writer Ibn al-Muqaffaʿ (ca. 721–57), who does not
appear to be mentioned by name in seventeenth- and eighteenth-century
Persian works but was widely cited by earlier Arabic authors, including the
earliest known transmitter Ibn al-Nadīm (ca. 987).⁴⁴ Ibn al-Muqaffaʿ was
an Iranian from a Zoroastrian convert background who settled in Basra,
became an important man of letters in Arabic, and most famously wrote
the Arabic translation of Kalīla wa Dimna. He defines Parsi (written in
Arabic as “fārsī”) as literary Middle Persian, which was still spoken by
some Zoroastrian priests [maubads] (to whom Ibn al-Muqaffaʿ also adds
“the learned [ʿulamāʾ] and those like them”), as well as being the spoken
register of the Pars region. Dari he viewed as the language of Ctesiphon
and environs (“a Persephone island in an Aramaic speaking area,” accord-
ing to John R. Perry), and modern scholars, notably Gilbert Lazard, have
interpreted this to mean that it was a northern dialect of spoken Parsi that
had mixed in Parthian influences. Dari’s spread as far as Balkh in the east,
on the opposite side of the Persian-using world from Ctesiphon, can be
explained by its being an administrative and contract language. Pahlavi (or
“fahlawī” in its Arabicized pronunciation) is the language of five regions
called the Fahlah, consisting of Isfahan, Ray, Hamadan, Nahavand, and
Azerbaijan, a territory in present-day western Iran which largely overlaps
with Ārzū’s description of the Pahlah.
After laying out the various definitions provided by Injū, Ārzū reveals
his position: The division of the three modern dialects is irrelevant; there
is only one, unitary literary dialect of Persian. If Dari is the language of
the court and Pahlavi is the language of the city [pahlaw] then, he argues,
obviously the two are different names for the same thing.⁴⁵ The only

⁴³ Browne 1929, 79ff. ⁴⁴ Lazard 1995, especially the quotation on p. 49; Perry 2009.
⁴⁵ Although the cross-over between Middle Persian and New Persian dialects is contested,
scholars have now come to a conclusion parallel with Ārzū’s, namely that Dari is not an “eastern
Persian” but rather a universal literary dialect integrated into New Persian wherever it was used
(Khanlari 1994; Clinton 1998; Smyth 1994, 292n6).
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

̄̄’     69

distinction is perhaps that each of the two names were used at different
times (he does not elaborate here but suggests in the following pages that
Pahlavi came first and then Dari). Since the usage of the court [urdū] is
the most eloquent and since the Farsi of the court is the accepted variety
then in fact all three, Dari, Pahlavi, and Farsi, are the same.⁴⁶ Thus, he
concludes,

:
it is a fact that [bah-tahqīq paiwast kih] the most eloquent of languages
is the language [zabān] of the court [urdū] and the Persian of that place
is respected, but a dialect [zabān–i khās:ah] of other places is not
accepted in poetry or belles lettres [inshā]. A poet, whichever place
he may be from, for example Khāqānī: from Shirvan, Niz̤ āmī from
:
Ganjah, Sanāʾī from Ghazni and Khusrau from Delhi, composed in the
established [muqarrar] language and that is none other than the
language of the court.⁴⁷

We consider the implications of this sweeping statement of poetic unity

in the next section. It is no exaggeration to call this stylistic flattening of
the whole geography of the Persian cosmopolis the key to Ārzū’s
thought. At the historical cusp of the nationalisms that would start to
divide up the Persian cosmopolis into the nation-states we know today,
Ārzū anchored his criticism in the well-known but apparently never
before so carefully formulated cosmopolitan ideal that every poet and
literary scholar across the cosmopolis was judged by the same standard,
which was not only translocal but transnational.
Having defined Persian, Injū and Ārzū faced a crucial theological
question: If the language of religion is Arabic, then is it permissible
even to use Persian?⁴⁸ They cite a relevant quranic passage to demon-
strate that although Arabic is significant as the medium of the final
revelation (i.e., the Qurʾān), it has not been the sole language of divine

⁴⁶ It is important to note here that the colonial-era identification of Urdu as the “camp
jargon” of India on the basis that “urdū” can mean military camp fails the historical test. The
word is cognate with “horde” (as in the Mongol unit of polity), and so refers to the seat of power
rather than to a common bivouac.
⁴⁷ M 13.
⁴⁸ Discussed in M 14–17; Farhang-i Jahāngīrī 1975, 17–22. See also Browne 1929, 12ff.
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

70       

revelation: “We have never sent a messenger who did not use his own
people’s language to make things clear for them” (14:4). Although Ārzū’s
discussion of the commentarial tradition is somewhat difficult to follow,
in essence many commentators have agreed that Arabic is not the sole
language with divine sanction. According to one commentator, it is even
permissible to use a Persian translation of quranic passages during
prayer [namāz]. Furthermore, the sons of Pārs were referenced in the
:
Qurʾān and Persians mentioned in hadīs̄.
The discussion in Mus̄mir now shifts to the origin of poetry, using
Lūbāb al-Albāb and Dabistān-i Mażāhib as the key sources. The first
poem ever composed was a mars̄iyah [dirge] sung by Adam for his slain
son Abel (it was presumably in Arabic, but this is left open).⁴⁹ According
to Lūbāb al-Albāb, the first Persian poem was composed by the king
Bahram Gūr (the historical Sassanian ruler Bahrām V, r. 420–38).
Dabistān-i Mażāhib offers another possibility: The first Persian poetry
was composed in the age of the Ābādiyān, the followers of Ābād, the first
prophet sent to Persia.⁵⁰ One of the kings in that mythical period,
Farhūsh, had seven incomparable [bī-qiyās] poets who each produced
a poem for every day of the week. Ārzū does not adjudicate between
these various possibilities,⁵¹ but instead changes the subject to one that is
more familiar to us: Ārzū argues that the language of Zoroaster’s time
was “obviously” Pahlavi, and the commentary upon it (Zend) and upon
that (Pazend) were also in the Pahlavi language. From our perspective,
this claim is incorrect, but it is incorrect in a telling way: Ārzū cannot
distinguish between what we call Old Persian and Middle Persian. From

⁴⁹ This is an extension of the story as told in Genesis 4:8 and Qurʾān 5:30. For a good
discussion of this section in Lūbāb al-Albāb and its larger context, see Keshavmurthy 2011,
esp. 109ff. A fascinating Sanskrit cultural parallel on the theme “poetry is born in grief” is
recorded regarding the author of the Rāmāyana: : Vālmīki witnessed a hunter kill one of a pair of
mating birds, and he spontaneously uttered the first śloka [couplet] as he lamented the death
and cursed the hunter.
⁵⁰ M 18. Ārzū reports this as fact but of course we are discussing an event which would have
occurred in mythical time. Ābād was believed to be author of the Dasātīr, which scholars later
concluded is a forgery written during Akbar’s reign. It has its own fascinating theory of language
(Sheffield 2014).
⁵¹ Yet another possibility, which Ārzū hints at in just a few words, is that (as the anonymous
author of the Tārīk: h-i Sīstān reports) the secretary of the Saffarid king Yaʿqūb bin Lais̄ (r.
867–79) composed the first Persian poetry because his king could not understand an Arabic
poem that was read out (Hanaway 2012, 105).
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

̄̄’     71

our perspective, Pahlavi is Middle Persian while the language of

Zoroastrian scripture, Avestan, is a variety of Old Persian (there is a
gap between the two of approximately 1,000 years). Ārzū is correct that
the Zend is a commentary on the ancient Zoroastrian scripture, the
Avesta, however Pazend is not a second-order commentary but rather
a script in which the Zend is written.⁵² Ārzū also incorrectly attributes
authorship of the “Zend” and “Pazend” to Zoroaster, but of course—
assuming that Zoroaster was an actual person and not a historical
composite in the same way that the poet Homer almost certainly
was—the commentaries were written many centuries after Zoroaster’s
life.⁵³ This line of thinking does not appear to be derived either from
Farhang-i Jahāngīrī or Dabistān-i Mażāhib, and represents an attempt
by Ārzū to tie up various historical loose ends: He refers to Zoroaster as
“Ibrāhīm Zardusht” or “Abraham-Zoroaster,” and so melds the Islamic
and Zoroastrian historical traditions by choosing a side in a long-
running debate over whether the two prophets were in fact the same
person.⁵⁴ Furthermore, he brings India into the picture by referring to
“the letter Zoroaster wrote to the emperor of India and which was
repeatedly read.” I have not been able to trace any origin for this
anecdote, but the conclusion Ārzū draws from it is crucial: He argues
that “the basis of these languages is Pahlavi, and after that Dari, after-
wards the common usage [i.e., present-day languages], as the research
concludes, and it [i.e., Pahlavi] has indeed fallen out of use.”⁵⁵ The phrase
“these languages” is ambiguous but since Ārzū has been discussing both
India and an Iranian prophet, it is reasonable to assume that he means

⁵² The prologue to Niz̤ āmī’s Haft Paikar [The Seven Beauties, 1197 ] makes a claim that
the poet’s project is exegetical by invoking Zend, which was widely known among Islamicate
writers to be an exegesis [taʾwīl] on Zoroastrian scripture: “This poem’s design I have adorned /
with seven brides, like Magian Zend” (4:33; translated in Meisami 1987, 299).
⁵³ M 20.
⁵⁴ For an example from c. 1400 in which a Muslim author rejects the Zoroastrian claim that
because both Ibrahīm and Zoroaster dealt with a holy fire they are the same person, see
Yohannan and Jackson 1907, 187. The same conflation of Zoroaster and Abraham appears in
M 3, 82.
⁵⁵ “az maktūbī kih zardusht ba-pādishāh-i hind niwishtah wa mukarrar ba-mut̤ āl ʿah dar
āmadah chunān dar yāft mīshawad kih as: l-i īn zabān-hā pahlawī ast baʿd az ān darī sipas
mutaʻārif hażā āk: hir al-tahqīq
: wa ān bi’l-fiʿl matrūk al-istiʿmāl ast” (M 20).
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

72       

that both Indic languages and contemporary Persian are derived from
“Pahlavi” (or, as we would call it, Proto-Indo-Iranian). While for us it is
evident that over time one language can change into another or branch
off into multiple languages, before the advent of modern historical
linguistics this was far from obvious.⁵⁶
Once these preliminary, Persian-specific matters are dealt with, the
text continues by engaging with al-Muzhir. There are various discussions
of how Arabic concepts entered Persian—for example, a section on
words in Persian for which the native Persian word was “forgotten”
and replaced by an Arabic equivalent and another section on “foreign”
words.⁵⁷ The difficulty for a Persian writer attempting to engage with the
Arabic tradition on this level is that throughout the Arabic philological
tradition, Arabic was generally represented as a closed system. (While
philologists were aware of the various dialects in contemporary Arabic,
this fragmented Arabic was not the object of their study since they were
interested in the Qurʾān and its particular language.) Traditionally, the
most eloquent spoken Arabic had been regarded as that of the Bedouin
tribes who had little contact with people outside their communities. At
the annual oratorical contest at Mecca, these Bedouins almost invariably
bested other Arabs, such as the Levantines, whose Arabic had been
“corrupted” by contact with Syriac speakers, and the Yemenis, whose
speech changed through contact with Indic and other languages.⁵⁸ While
al-Suyūt::ī was aware of the influence of other languages on Arabic and
vice versa—after all, he wrote a lexicon of words in the Qurʾān with
analogues in other languages⁵⁹—others were unwilling to accept Arabic

⁵⁶ Cf. Foucault 1994, 291–4.

⁵⁷ The discussion on “forgotten” Persian words is based on Imām S̄aʿlibī, and Ārzū takes him
to task for being wrong about his examples (M 55ff.). On foreign words, the discussion is M 80ff.
⁵⁸ M 9, cf. Suleiman 1999, 22. “Syriac speakers” is a conjecture on the basis of the word
“nis: ārī,” meaning Nazerene (a term for Christian and Judeo-Christian sects). That Yemen is
singled out is noteworthy because it was, of course, on the maritime crossroads of India, the
Middle East, and Africa, and so would have been teeming with languages.
⁵⁹ Bell 1924. In his Lubb al-Lubāb, al-Suyūt::ī lists seven sound combinations that suggest an
Arabic word was originally Persian (Asbaghi 1987, 7–8). The extremely influential Tafsīr of
T̤ abarī (b. 224–5 /839 ) also deals with the question of the relationship of Persian and
Arabic and more generally addresses loan words from other languages in its introduction
(Tabari 1987, 1:12–15).
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

̄̄’     73

as anything less than an originary language.⁶⁰ Indeed, al-Suyūt::ī’s section

on tawāfuq begins with a quotation from the grammarian Ibn Jumhūr
(fl. third/ninth century) claiming that “there is nothing in God’s book
[the Qurʾān]—praise be upon it!—except the Arabic language.”⁶¹ In
Mus̄mir, such views on “pure” Arabic are not well connected to any
discussion of Persian, but because Ārzū posits Persian as a courtly
language, there is an inherent contradiction: The best Arabic is
unmixed and yet Ārzū acknowledges that courts are places where
languages do mix. For example, the word barsāt [rainy season], “an
obviously Indian” term whose invention Ārzū connects to the “people
of the court” [ahl-i urdū], is derived from the Persian word “bārish”
[rain] plus the Arabic feminine plural ending “-āt.”⁶² Thus Arab views
on linguistic purity cannot apply directly to Persian, but Ārzū hedges
against this inconsistency by making proper Persian equivalent to
courtly usage and therefore unmixed, because whatever expressions
come into the Persian of the court and are accepted there are, by
definition, Persian. While his discussion appears to relate to both
formal and informal language, his focus is on written literary language.
Indeed, in his discussion of the Meccan oratorical contest, he asserts
that all Arabs wrote in an identical style even if they spoke different
dialects. This emphasis on the production of literature over casual
speech requires Ārzū to offer a nuanced account of the phonetics of
regional speech as well as to some degree regional vocabulary, in order
to determine which words are universal (and therefore literary) and
which are local (and therefore unliterary).

⁶⁰ The problem of a divine origin of language versus language as an agreed upon convention
was one of the major threads of debate in Arabic philology. Those inclined to the view that God
handed language to man often pointed to Qurʾān 2:31: “He taught Adam the names [of things],
then He showed them to the angels and said, ‘Tell me the names of these if you truly [think you
can].’ ” The first chapter of al-Muzhir, in which al-Suyūt::ī deals with this question, does not
record his personal opinion (Loucel 1963–4, 68). Early Arabic grammarians noted the foreign
origin of various Arabic words, which is somewhat surprising given how vehemently the
possibility was opposed by later exegetes (Versteegh 1993, 88–91).
⁶¹ “laysa fī kitāb allāhi—subhānahi—shayʾ
: baġhair luġhat al-ʿarab” (al-Muzhir 1998, 209).
⁶² M 212.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

74       

Phonetics, Vocabulary, and Regional Variation

Ārzū’s contention that there is a single literary Persian used across the
vast territory of the Persian cosmopolis is not based on the linguistically
suspect claim that there is no variation across that region. He is clear—as
we should also be—that what is under consideration is literary language
and not everyday communication. For one thing, his project is not pure
description (as in a modern linguistic survey) but also an aesthetic
endeavor: The goal is not only to describe usage but to set a standard
at the same time.⁶³ Ārzū recognized that regionalism is inevitable,
even in a formalized dialect like literary Persian, and so understood
the need to deal with variation in a theoretically sophisticated way. The
cosmopolitan tradition always feels the pull of the local as it simultan-
eously exerts its own force on the local language and culture.⁶⁴ He
posits a structure of authority that flows from the imperial court,
arguing that:

The truth as established by research is that the respected [i.e., standard]

Persian is that of the royal court, which has been established after the
mixture of crowds and troops; thus, in the poetry of the eloquent and
the prose of the articulate there is no other language, and if ever,
because of the demise of its authority, [the royal court] ceases to
exist, then the eloquent and articulate of every city and province use
in conversation that [language] which has been established and do not
mix [it] with the language of their country.⁶⁵

⁶³ The underlying philosophy of grammar, namely the system of Hellenistic and traditional
Arabic grammar, likewise followed rational principles (that is, it strove for an ideal) rather than
empirical (that is, observation-based) ones (Versteegh 1980, 21–2).
⁶⁴ These interactions are not inherently good or bad in Ārzū’s eyes, but they are inevitable
and so must be dealt with. Ārzū’s understanding of them is considered further in this chapter
and again in Chapter 5.
⁶⁵ “wa haqq-i
: :
tahqīq ān ast kih zabān-i muʿtabar-i fārsī zabān-i urdū-yi pādishāhī ast kih baʿd
ik: htilāt::-i firaq wa jumāʿat qarār yāftah lihażā dar shʿir-i fus: ahā
: wa nas̄r-i bulaġhā zabān-i dīgarān
nīst wa agar ahyān :
an
:
sabab-i qalat-i hukm ʿadam dārad wa ānchih muqarrar shudah fas: īh-o: balīġh
az har shahr wa ulkahʾī kih bāshad badān takallum namāyad wa ba-zabān-i mulk-i k: hwud : mak: hlūt::
na-sāzad” (M 9).
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

̄̄’     75

Thus, according to Ārzū, primary responsibility for controlling good

taste lies with the people of the imperial court and then with sub-
imperial elites.⁶⁶ In Persian, in contrast to the Arabic tradition’s vener-
ation of Bedouins, the guardians of language are urbanites.⁶⁷
The existence of regional vocabulary in Persian literature is an issue
because regional vocabulary is by definition unaesthetic—however, it can
later become universal and accepted in Persian. It is likely that Persian
lexicography originated to address the need to explain unfamiliar dialect
words that appeared in poetry, and the process of being codified in
dictionaries made many such originally local words universal across
Persian.⁶⁸ Ārzū engages with al-Muzhir in a number of chapters dealing
with vocabulary: on the eloquent [fasī : h]
: or rather correctly formed
words, on the “despised and base” [raddī-o mażmūm], on the “universal
and rare” [mut:::t:arad-o shāż], and finally on the “unfamiliar and strange”
:
[wahshī-o ġharīb]. Eloquent words, according to Arabic rhetoricians, are
ones that have “purity” [k: hulūs:] and common currency.⁶⁹ This is in part
a historical argument because, as Ārzū notes, every period has a different

⁶⁶ Contemporary Europe had a similar system of linguistic authority: Standard French in the
seventeenth century was thought of as that of “the court and the city” [la cour et la ville], reflecting
a courtly/urban bias just as in Ārzū’s formulation (Burke 2004, 99–100). Claude de Vaugolas
(1585–1650), a founding member of the Académie Française, argued that more specifically the
standard was to be set by “the soundest part of the court” [la plus saine partie de la cour], which
included women, the administrative establishment, and townspeople who had dealings with the
court. This was the accepted linguistic standard of that period in France as well as in Poland (where
it was called “dworski mowy” or “courtly speech”), Denmark, Russia, England, and elsewhere.
⁶⁷ In my discussion of NA, Ārzū’s lexicon of Indic words, in Chapter 4, I argue that this view extends
to the vernacular as well. On early Arabic grammarians’ use of Bedouin data (before the tradition
generally came to acknowledge only the Qurʾān as admissible evidence), see Versteegh 1983, 146–9.
⁶⁸ Baevskii 2007, 29. The creation of dictionaries was a catalyst for maintaining the remark-
able homogeneity of literary Persian over nearly a millennium. Regionalism was a concern even
in the early years of the Persian cosmopolis: For example, the eleventh-century  writer Nās: ir
:
Khusrau met a poet called Qat::rān in Tabriz (northwestern Iran) who “wrote good poetry but
did not know Persian well” [shʿirī nekū mī guft ammā zabān-i fārsī nekū namī dānist] (quoted in
Hanaway 2012, 124). It is not clear whether Qat::rān “did not know Persian well” because he was
an Azeri Turkish speaker or whether he was not familiar with the literary language, but context
implies the latter: He read the dīwāns of Manīk and Daqīqī with the help of Nās: ir Khusrau, : all
:
three of whom were from Khurasan in the east. Thus, Nās: ir Khusrau had an easy access to their
particular usage, which Qat::rān did not.
⁶⁹ M 61–75, esp. 71. The Arabic tradition tends to cast correct and incorrect usage in
aesthetic terms, where for example a “hasan”
: (lit. beautiful) form corresponds to our idea of
grammatically “well formed” structure and “qabīh” : (lit. hideous, ugly) to an incorrect form
(Carter 1973, 148). The terminology is not standardized, even in the same work, but as in
modern linguistics it is an examination of the form itself and not the correctness of the
information being transmitted.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

76       

set of acceptable vocabulary since words pass out of common usage.⁷⁰

The “despised and base” words are glossed as “the worst words” [bad-
tarīn luġhāt], a direct translation from Muzhir, but Ārzū’s discussion
has nothing to do with vocabulary as such but rather with dialectical
pronunciations that are not reflected in written language.⁷¹ In a not
altogether unexpected but still amusing twist of history, two standard
features of modern Iranian Persian are given as examples of this cat-
egory,⁷² namely the interchangeable pronunciation of the letters “qāf”
and “ġhain” (i.e., the sounds “q” and “ġh”) and the tendency to turn “ān”
into “ūn” (e.g., pronouncing “dukān” [shop] as dukūn).⁷³ The discussion
of the “universal and the rare” is derived from Ibn Jinnī’s analysis of the
matter.⁷⁴ Ārzū simply endorses his views on dialect variations in pro-
nunciation as applicable to Persian as well as Arabic. The “unfamiliar
and strange” words are those which are rarely used.⁷⁵ In the Arabic

⁷⁰ M 62. The Roman poet Horace (65–8 ) makes a similar observation in his Ars Poetica:
“As leaves in the forest are changed in the fleeting years / the oldest fall first: thus words die of
old age / and the newly coined flourish and thrive, as in the prime of youth” [ut silvae foliis
pronos mutantur in annos, / prima cadunt: ita verborum vetus interit aetas, / et iuvenum ritu
florent modo nata vigentque] (vv. 60–2).
⁷¹ M 75–77. As dismissive as Ārzū is of these dialectal usages here, it is notable that in a
discussion in an earlier text, TĠh, he allows similar usages to stand if poetic masters had made
them (TĠh 142–3).
⁷² Modern Iranian Persian is, of course, a relatively recent standard based on the local speech
of Tehran, which was not a place of much importance until it became the Qajar capital in 1795.
It was not, apparently, a prestige dialect before the spread of mass media in Iran.
⁷³ One of the fourteenth-century satirist ʿUbaid Zākānī’s anecdotes turns on the confusion
between “q” and “ġh” and demonstrates that it was a common mistake in his time: A person asks
“Do they make [that is, spell] ‘fried food’ [qalīyat] with ‘qāf’ or ‘ġhain’?” The reply is that “They
make fried food with neither ‘qāf’ or ‘ġhain’ but with meat” (Risālah-yi Dilgushā 1955, 114).
I thank an anonymous reviewer for the reference. In his discussion of Yahya̍ : Kāshī in MN, Ārzū
notes that the people of Kashan in particular pronounce “ān” as “ūn” (MN 2004, 3:1843). I am
grateful to Nathan Tabor for the reference. Ārzū’s nephew Mīr relates an anecdote that the
:
emperor Muhammad Shāh set a verbal trap for the nobleman Safdar

:
Muhammad :
Khān, who
was originally from ‘irāq (i.e., western Iran), by quoting a line of Saʿdī. When the nobleman
completed the couplet, his accent turned the words “k-ān sokhtah” (“that this inflamed”) into
“kūn sokhtah” (“inflamed anus”) (Naim 1999, 137). Mīr suggests that the sound change in
question (ān > ūn) is a matter of pretension in poetry, even though it is the colloquial accent of
the nobleman’s birthplace.
⁷⁴ M 78–9. For him there are four kinds of words: (1) Universal in Analogy and Use
[mut:::t:arad dar qiyās-o istʿimāl], (2) Universal in Analogy and Rare [shāż] in Use, (3)
Universal in Use and Rare in Analogy, and (4) Rare in Analogy and Use. The precise
distinctions are not entirely clear.
⁷⁵ “dar mustaʿmalāt kamtar āmadah bāshad” (M 80–3). The term “wahshī” : might have been
used in the technical sense of unintelligible language as early as the sixth century 
(Bonebakker 1970, 81).
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

̄̄’     77

tradition, there is some disagreement as to whether using such words is

permitted at all. According to Ārzū, words which are known but whose
meanings are unclear (because they are so rarely used) are avoided by the
eloquent. However, since Persian is a mixed language, this rule can never
be absolute.
More important for our purposes is a two-part discussion on whether
:
borrowings [tasarrufāt] from Indian (and other) languages are permitted
in Persian.⁷⁶ It begins in a short chapter which frames the question
through al-Muzhir⁷⁷ and returns, without particular reference to al-
Muzhir this time, in several later chapters. When the discussion is not
framed by Arabic, Ārzū is philosophically uncompromising: If Arabic,
Turkish, and Armenian [arāmanah] words could be borrowed into
Persian over the years, what logic could prohibit Indian words?⁷⁸ Great
masters have used hindī words in Persian and their usage of them is
“unquestionably correct” [bī-takalluf durust]. Thus, the question is not
whether Indic words are allowed in Persian in the first place (“in the
doctrine of the present author they are not forbidden in the present time”
[ba-mażhab-i muʾallif darīn zamān mamnūʿ nīst]) but rather in what
frequency and to what degree they can be changed.⁷⁹ He notes that

⁷⁶ Blochmann defines tas:arruf as follows: “The change in spelling, form, meaning and
construction, which an Arabic word, apparently without any reason, undergoes in Persian, or
which an Arabic or a Persian word undergoes in Hindustani, is called tas: arruf” (1868, 32). For
him it is necessarily a kind of corruption, but Ārzū appears to have viewed it as a normal
process. A good reconstruction of the aesthetic consequences of tas:arruf for Ārzū is
Keshavmurthy 2013, 37–8.
⁷⁷ M 36–9; 160–75. In the first chapter, Ārzū summarizes a passage in which al-Suyūt::ī
considers whether Arabic had in fact borrowed Persian words, a possibility that many Arabic
grammarians denied.
⁷⁸ M 160. He takes up this argument again in Mauhibat-i ʿUz̤ ma̍ and the second preface to
Dād-i Suk: han, discussed in Chapter 3 of this study. Early Persian dictionaries were not
concerned with where words came from. As Solomon Baevskii notes, “They make no explicit
distinction between assimilated loanwords in Persian and Fremdwörter or outright foreign lexis;
but in general it is clear that fully assimilated Arabic and (in lesser volume) Turkic vocabulary in
Persian is not labeled for etymology” (Baevskii 2007, 142).
⁷⁹ This translation is based on Cyrus Shamisa’s emendation of M 1991’s “darīn zabān” [in
this language] to “darīn zamān” (Shamisa 2002, 31).The latter is the lectio difficilior (the
preferred reading because of its greater complexity) since it addresses an argument about
chronology which Ārzū introduces just above: He points out that the influence of Arabic on
Persian began during the Arab conquests of Iran (a millennium before) but that Turkish and
Arabic words continued to be taken into Persian subsequently. The ongoing process of
borrowing is the analogy for borrowing Indic words into Persian, and therefore borrowing
them “at the present” [darīn zamān] is not prohibited.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

78       

Arabic and Turkish words have even replaced some common indigenous
Persian words over the centuries. However, there is a sense that care
must be taken in deploying Indic terms. Ārzū’s friend Mukhli : s: is explicit
in reserving the right of bringing hindī words into Persian for “people of
[aesthetic] might and exceptional talent” [ahl-i qudrat wa istiʿdād-i
:
mukhtar].⁸⁰
The considerable variation in pronunciation across the Persian cos-
mopolis does not concern Ārzū except to the degree that it is liable to
corrupt the literary form. For example, by the seventeenth century
Iranians had begun to stop distinguishing majˈhūl vowels (the sounds
“e” and “o”) from maʿrūf vowels (“ī ” and “ū”), but to Ārzū, this change is
worth mentioning but it is not in itself significant.⁸¹ Such a shift is not in
the same category as the “despised and base.” Mus̄mir has several
chapters devoted to various phonetic transformations. The discussions
are largely derived from al-Muzhir. The topics, which in Arabic are
generally taken together as part of nahw, : include ikhtilāf
: [variance in
:
gemination], qalb [transposition], hażf [elision], ibdāl [exchanging, i.e.,
sounds], and imālah [changing “ā” into “ī’”].⁸² The most relevant for our
purposes is ibdāl, which refers to when a sound is exchanged for one with
a similar articulation point [qurb-i mak: hraj] in the mouth.⁸³ It explains a
number of historical sound changes, including changes to hindī words in
Persian, such as “rānā” [hindī for king] being pronounced as “raʿnā”
[Arabic for beautiful or tender] and “garbhsūt” [cloth with cotton warp
and silk weft] turned into “garmsūt.”⁸⁴ The book ends with a lengthy
(and to the modern reader tedious) account of the characteristics of all
the letters in the Perso-Arabic alphabet, going letter by letter.⁸⁵
The phonetics of languages used alongside Persian require a descrip-
tive apparatus in order to consider the integration of foreign words into

⁸⁰ Quoted in Shamisa 2002, 32.

⁸¹ M 84. See Khan 2004, and cf. a longer discussion in TĠh 142–3.
⁸² Ik: htilāf, which Ārzū claims is very common in Persian, involves either reducing a
geminated consonant (that is, one pronounced twice in succession, as in, for example, “n” in
“cannot”) to a single consonant or geminating a single consonant (M 90ff.). A chapter heading
refers to qalb as a subset of ik: htilāf, but in fact it seems to be entirely different, namely when two
contiguous sounds in a word are switched (M 117ff.). Hażf
is, of course, the loss of a sound (M
99ff.). Imālah is described at M 158–60; cf. Owens 2006, 197–229. On Ārzū’s discussion of
phonetics, see Rahimpoor 2012, 109–28.
⁸³ M 135ff. ⁸⁴ M 171, 174. ⁸⁵ M 244ff.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

̄̄’     79

Persian. Expressing non-Arabic sounds in written Arabic had a long

tradition and this was expanded by Persianate intellectuals.⁸⁶ The
descriptive system adopted by Ārzū and by some earlier writers, like
the author of Farhang-i Jahāngīrī, is simple yet capable of expressing
virtually all of the sounds in Arabic, Persian, and Indic languages. The
four Persian letters that Arabic does not have (“p,” “ch,” “z̄,” and “g”) are
analogized to their Arabic letter with the equivalent form.⁸⁷ Thus, for
example, “p” [‫ ]ﭖ‬is referred to as “the Persian ‘b’ [‫]ﺏ‬.” In a manuscript
culture where the marks that distinguish letters from similarly shaped
letters are often not written in or simply get lost as ink fades and paper
crumbles, such long-form orthography [taqaiyud] prevents ambiguity.⁸⁸
Representing Indic language presents a greater challenge because it
includes aspirated and retroflex sounds, which are not phonemes
found in either Arabic or Persian.⁸⁹ The aspirated consonants, which
are consonants spoken with more breath than their unaspirated coun-
terparts, are described aptly as a letter “mixed with ‘h.’” Thus “ph” is
called “the Persian ‘b’ mixed with ‘h’” [bā-yi fārsī mak: hlūt:: bā hā].
Retroflexes, formed by curling the tongue back towards the roof of the
mouth, are unique to Indic languages (and Pashto) within the Persian
cosmopolis, and are simply analogized to Arabic as the uniquely Persian
letters were. Thus, the retroflex “t:” is called “the Indian ‘t’” [tā-yi hindī].
For a sound like “t:h,” which is both retroflexed and aspirated, the two
operations are combined to yield “the Indian ‘t’ mixed with ‘h’” [tā-yi

⁸⁶ Ibn Khaldūn
: proposed a kind of phonetic triangulation: He rejects the usual scribal
practice of replacing a foreign letter with the single closest Arabic equivalent, instead deciding
“to represent such non-Arabic (sounds) in such a way as to indicate the two (sounds) closest to
it, so that the reader may be able to pronounce it somewhere in the middle between the sounds
represented by the two letters and thus reproduce it correctly” (Muqaddimah 2005, 31). He is
primarily concerned with the representation of Berber.
⁸⁷ M 135.
⁸⁸ A letter described by this system is muqaiyad [lit. “fettered”] because it is no longer subject
to misinterpretation (Blochmann 1868, 13). Taqaiyud (which Blochmann calls qaid) was
common well into the twentieth century, even when it had become easy to typeset vowel
marks in Persian. Surprisingly, though, transliteration into roman script was generally preferred
to vowel marks in published dictionaries until recently (Baevskii 2007, 163).
⁸⁹ In hindī/Urdu the aspirated letters are “bh,” “ph,” “th,” “dh,” “t:h,” “dh,”
: “r: h,” “jh,” and
“chh.” The digraphs “lh” [ल्ह], “mh” [म्ह], and “nh” [न्ह] are each considered a single phoneme
in some dialects, e.g., Braj, and of course are not consonants. The retroflex consonants are “t:,”
“t:h,” “d,”
: “dh,”
: “r: ,” and “r: h.” The retroflex nasal (“n”)
: and sibilant (“s:”) are ignored. An
alternative but far clumsier system to distinguish the Indic letters used in Tuhfat : al-Hind is
discussed further in Chapter 4.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

80       

: :: bā hā].⁹⁰ This sort of precision is important when words

hindī makhlūt
pass between languages because sounds change in the process, and the
nature of the phonetic transformation can help us determine the words’
relationship. Ārzū repeatedly refers to the difficulty faced by non-Indians
in pronouncing Indic sounds, which leads to mistakes when the word
becomes a Persian word. Although some of his conclusions are tentative,
he uses the logic of phonetics to probe how Arabic, Persian, and Indic
languages are linked.

Connections between Languages

Ishtirāk or “sharing” describes the existence of the same word in two or

more languages. This is of obvious interest for modern scholars because
it could be interpreted as historical linguistics avant la lettre. It is,
however, important to point out that Ārzū’s project was limited: He
was interested not in drawing up a family tree of languages but merely in
explaining how a word could be shared between languages. The
approach he took, as in much of his scholarship, was to innovatively
extend the tradition as it stood. In Mus̄mir, he describes three processes
that can explain instances of ishtirāk.⁹¹ None of these is entirely

⁹⁰ To my knowledge, there has been no study of the development of modern Urdu orthog-
raphy from Persian orthography. While the aspirates had apparently been written in throughout
the Mughal period (although not with do-chashmī “he” [‫ ]ﮪ‬as in modern Urdu), differentiating
“h” retroflexes from their nearest Persian equivalents does not appear to have been common
before the nineteenth century. The modern method for making retroflexes, the letter “t::” [‫]ﻁ‬
written above the letter as in “t:” [‫]ﭦ‬, was one of at least three competing possibilities. Another
common method was using four dots in a block above the letter as in [‫ ]ﭢ‬for “t:.” An undated
autograph manuscript of the poet Rāsikh : of Azimabad (d. 1824) from approximately 1810 now
in the collection of Shamsur Rahman Faruqi has the unusual feature of a vertical line [ˈ] placed
above letters to mark them as retroflex. In general, there was a great deal of inconsistency in
orthography in Urdu manuscripts produced for personal consumption before around 1810 so it
is not surprising that retroflexes, which would have been obvious from context, were not often
marked (conversation with Shamsur Rahman Faruqi, Allahabad, 5 June 2011).
⁹¹ Elsewhere he proposes a five-part scheme: In his lexicon CH’s entry for “ang” he notes that
words can be shared as a result of (1) tawāfuq, (2) ittifāq, (3) tafrīs, (4) iltizām, or (5) muhannad
(CH 1984, 1017–18). The first three are addressed in Mus̄mir. Iltizām, which has been defined
by one modern scholar as “signification on the basis of an association” (De Bruijn 1988), is an
odd category because Ārzū does not describe it except to say that it is as in the work of the poet
T̤uġhrā (whose penchant for using Indic vocabulary is clear from the number of times Ārzū cites
him, cf. Abidi 1982, 51–2). Perhaps therefore it means something like “arbitrarily using a non-
Persian synonym.” Muhannad is considered further in the chapter in the context of tafrīs.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

̄̄’     81

satisfactory from the perspective of modern historical linguistics, but the

logic is fascinating and self-consistent.

Three Kinds of Connections

The simplest of the three is Persianization [tafrīs], which is the process

by which a foreign word is borrowed into Persian and experiences
phonetic changes that mold it into a Persian word. Persianization is
thus defined as when the root [jauhar] of a word is taken from
another language into Persian and generally undergoes a phonetic trans-
formation.⁹² Ārzū argues that this is a long-standing phenomenon, and
indeed that most of the words had been borrowed into Persian “before
the mixing of Persian and hindī.” This is a surprising statement, but its
meaning becomes clear after some examples⁹³ and the remark that

the words mentioned [as examples] are in the book hindī [hindī-yi
kitābī] of the people of India, which is completely different from the
Indian language current in this country and Persian is not mixed with
it, insomuch as the aforementioned language [i.e., “book hindī”] is still
present in the world.⁹⁴

Context makes it clear that the “book language” is Sanskrit since it differs
from presently spoken languages and is primordial. It is, as Ārzū notes,
unaffected by the large-scale borrowing of Persian vocabulary that the
spoken Indic languages had experienced from the Delhi Sultanate period
into Ārzū’s time. Thus, in another work, he introduces the concept of
muhannad [making Indic], which is the opposite of tafrīs since it refers

⁹² “ān tas:arruf ast dar jauhar-i kalimah az jihat-i taġhayyur-i talaffuz̤ yā ġhair-i ān” (M 211,
cf. 61ff.).
⁹³ Namely “usht:ur”/“ushtur” [camel], “angusht:”/“angusht” [finger], and “nābh”/ “nāf”
[navel]. The first of each of the pairs is the hindī pronunciation and the second the Persian.
⁹⁴ “alfāz̤ -i mażkūrah ba-zabān-i kitābī-yi ahl-i hind ast kih muġhāyarat tamām ba-zabān-i
hindī mustaʿmal-i īn mulk dārad wa fārsī rā badān mut::laq ik: htilāt:: nīst chunānkih bar ʿālam
zabān-i mażkūr z̤ āhir ast” (M 211–12). The word “muġhāyarat” from the Arabic root ġh-ī-r in
addition to opposition has the sense of exchange or bartering, which would change the
interpretation of how book hindī relates to contemporary hindī.
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

82       

to a Persian word that has been borrowed into an Indic language.⁹⁵ He

notes that muhannad is his own term, and observes that such borrowings
of Perso-Arabic terms are especially prevalent in Indic (official) docu-
ments [dafātir-i hindī].
Persianization is an ongoing process as Indic words, especially names,
continue to be brought into Persian, but there is an important distinction
to be drawn between recently Persianized words and those that entered
Persian centuries before. The poet T̤uġhrā, for example, refers to “barwach”
(for “bharūch,” a town in Gujarat near Ahmedabad formerly written in
English as “Broach”) and Ārzū disapproves of the transformation of “ū” to
“wa,” calling it “unmotivated” [bi-zarūrat].⁹⁶

When words have been
recently Persianized, phonetic changes have often occurred because of “a
:
lack of research and inattention to the literature of Indians” [ʿadam-i tahqīq
wa iʿtnāʾ ba-kalām-i hindiyān].⁹⁷ (This probably at least partially explains
why using hindī words in Persian for the first time is reserved for expert
poets.) For example, “arhant” [an enlightened person or Jain saint] has
been incorrectly written in the dictionary Burhān-i Qāt::iʿ and elsewhere as
“arhaft” because in Perso-Arabic script “f ” and “n” (which both have a dot
above) can look somewhat similar.⁹⁸ A common pattern of mistakes is
discernible in the name of the Mahābhārata character Bhīma, which has
been infelicitously rendered into Persian as “bahīm” because Iranians
assumed that “bh” (a single sound in hindī) represented the two separate
sounds “b” and “h.”⁹⁹ Since there are no aspirates in Arabic or Persian, Ārzū
says that Iranians are generally unable to avoid such a “gross mistake”
[k: hat::ā-yi fāhish].
: However, there is an important case in which an allow-
ance is made for usages that are wrong by Indic standards, namely when
an incorrect spelling has been sanctified by usage. For example, in

⁹⁵ CH 1984, 1018.
⁹⁶ M 217. The term zarūrat

(literally “necessity”) came to have a technical meaning in
criticism indicating that the poet was compelled to use a particular form. Blochmann glosses
it as “poetical license” in describing Tek
Chand Bahār’s Ibt::āl-i Zarūrat
: [The Refutation of
“Zarūrat”],
: a treatise that argues that all correct poetic expressions rest on previous authority
and any that do not are simply wrong (Blochmann 1868, 29).
⁹⁷ M 216.
⁹⁸ M 214. This section corrects the spelling of a number of Hindu concepts such as
Maheshwar and Brahma, also providing fascinating definitions.
⁹⁹ M 215. Likewise, “gur: ,” glossed as “rock candy” [qand-i siyāh], is incorrectly rendered as
“gor”—both spellings are current in Persian, but the chief [sar-guroh] of Modern poets, Sāʾib

Tabrīzī, uses the former and only it is correct.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

̄̄’     83

Aurangzeb’s chancery, an order was given that Indian place names written
with a vocalic “he” at the end, such as Bengal (“bangālah”) or Malwa
(“malwah”), should henceforth be written according to the Indic manner
with an “alif” at the end, that is “bangālā” and “malwā.”¹⁰⁰ This change was
wrong, argues Ārzū, because Persian has a history of spelling such names
the former way, as in the famous verse of Hāfiz̤

, when the poet boasts about
the extent of his readership:

shakkar shikan shawand hamah :t:ūt::iyān-i hind

z-īn qand-i pārsī kih bah bangālah mī rawad
[All the parrots of India will turn to sugar crunching
From this Persian candy that goes [as far as] Bengal.]

Because a master-poet [ustād] like Hāfiz̤
has written it this way, this
usage is beyond reproach, even though it is wrong from the perspective
of Indic language. This is a mechanism of authority that operates like the
Sanskrit “ārsa: prayoga” or the usage of a r̥si : [sage] that is invoked as
justification for grammatical exceptions in a language whose very name
: tam, lit. refined] represents it as perfect. In Chapter 3, we con-
[samskr̥
sider how the usage of the Ancients [mutaqaddimīn] structures authority
in Persian discourse. Ārzū implies that there is a limit to what precisely
should be taken from one language to another but does not elaborate.
Elsewhere, however, he writes that “an expert knows that the grammat-
ical logic [qiyās] of one language does not apply to another and that the
:
requirements [ahkām] of each language are distinct—this is a basic
principle.”¹⁰¹ In this case, the “-ah” ending is proper Persian but it
would be wrong to use it in the “idioms of India” [muhāwarāt-i : hind],
by which Ārzū presumably means vernacular languages, or write it that
way in “Indic script” [ba-k: hat:::t:-i hindī], just as it is wrong to write such

¹⁰⁰ M 213. In Hindi-Urdu today, words ending in “-ah” and “-ā” are pronounced identically,
but for Ārzū there was a distinction as there is in modern Iranian Persian.
¹⁰¹ “ammā muhāwarah-dān
: mī dānad kih qiyās-i zabānī ba-zabān-i dīgar nabāyad kard chih
:
ahkām-i har zabān ʿala̍hidah
: ast wa īn as:lī ast” (M 221). For example, Arabic words are
combined in Persian to form compounds that do not exist in Arabic and this is unproblematic
(M 38). Ārzū objects to the mistake (committed even by good lexicographers) of uncritically
correcting Persian text based on Arabic (M 43).
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

84       

names with “-ā” in Persian.¹⁰² Using Indic words in Persian properly

takes finesse and so Ārzū concludes that one should not use an Indic
word unnecessarily if there is a good Persian equivalent.¹⁰³ The sorts of
words that are unquestionably allowed are “the idioms of the court and
chanceries of the emperors of India which are accepted by the nobles of
the court,” and he provides a number of examples, such as “sarbāzī” in
the meaning of “sardār” [ranking officer].¹⁰⁴ There is no judgment
passed on poets who use Indic words (as long as they are used properly),
and Ārzū cites his own verse to show that he himself does so.¹⁰⁵ He is
keen to know where words come from so that their proper usage can be
established.
The concept of Persianization is based on the model of Arabicization
[taʿrīb], the equivalent process by which a word enters Arabic and
changes. Ārzū catalogues at length the shared words in Arabic and
Persian in a section that engages with al-Muzhir called “A Description
of Several Problems and Examples Connected with the Arabicization of
Persian Words.”¹⁰⁶ In fact, the chapter begins with a lengthy list of Arabic
words from al-Muzhir. Ārzū’s contribution here is to demonstrate con-
vincingly that many of the “Arabicized” words that scholars like al-Suyūt::ī,
and before him al-S̄aʿlabī, had assumed were natively Persian were in fact
of Indic origin. These include “filfil” [pepper], the Arabic equivalent of
Persian “pilpil” which in fact was originally Indic “pīpal,” and likewise
“fīl” [elephant] which is Indic “pīl.”¹⁰⁷ The latter example is noteworthy
because Ārzū states that “pīl” is the word in “hindī-yi kitābī” (that is, in
Sanskrit).¹⁰⁸ Ārzū notes that this is the meaning that “has been heard from
several people of knowledge of India” [az baʿzī
ahl-i ʿilm-i hind shinīdah

¹⁰² However, in practice many pre-modern hindī texts written in Perso-Arabic script did not
carefully distinguish between “-ah” and “-ā,” just as they often failed to distinguish between
homonymic Arabic letters (for example, confusing “z”
and “z”).
¹⁰³ M 222.
¹⁰⁴ “mus::t:alahāt-i
: urdū wa dafātir-i pādishāhān-i hind wa qarār dādah-yi buzurgān-i darbār”
(M 222).
¹⁰⁵ M 221.
¹⁰⁶ “as: l: dar bayān-i baʿzī

az masāʾil wa ams̄alah kih mutaʿalliq ast ba-taʿrīb-i alfāz̤ -i fārsiyah”
(M 179–209). The section in al-Muzhir specifically addresses Persian words that have become
Arabic (1998, 211–35).
¹⁰⁷ M 186, 182, respectively.
¹⁰⁸ Following the usual practice when Indo-Persian works cite Sanskrit lexemes, Ārzū has left
the ending off—modern scholars would cite the word as “pīlu.”
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

̄̄’     85

shudah]. This and similar statements are solid evidence that, contrary to
the claims of some scholars who assume that Ārzū himself knew Sanskrit,
he in fact did not but depended on informants.¹⁰⁹ There are dozens of
examples in the lengthy section, which no doubt serve to anchor the
discussion in the prestige of Arabic.
Another way of explaining the existence of the same word in two
languages is coincidence [ittifāq]. This phenomenon is the result of the
independent phonetic transformation of two words in different languages
such that their form becomes identical in each language.¹¹⁰ For example,
the word “jārū” [broom] is found in both Persian and hindī. In hindī,
according to Ārzū, it is a compound formed from “jhārnā”: [to sweep] and
“rūb” [sweeping].¹¹¹ In Persian, it is the shortened form [mukhaffaf]
: of
“jārob” [broom]. For Ārzū, coincidence appears to be the least productive
of the three kinds of linguistic connections because he only gives this single
example and does not discuss coincidence again in Mus̄mir. However, in
the Arabic philological tradition it was an important concept since it is the
term used by T̤abarī to explain non-Arabic words in the Qurʾān. T̤abarī
means something slightly different by “ittifāq” because for him it refers to
a word’s existence in multiple languages (similar to how Ārzū uses the
term “ishtirāk”), not necessarily that the word developed independently in
different languages.¹¹²
The most significant of the three processes of word-sharing is tawāfuq,
a term perhaps best translated as correspondence. Ārzū defines it as the
kind of shared word that is “fixed in its root and neither [word] has been
taken from the other.”¹¹³ What he means is that the root of the word is
identical in both languages—in contrast to Persianization, in which it has

¹⁰⁹ Elsewhere, however, he cites “grammar books of the Indians” [kutub-i nahw-o : s: arf-i
hindiyān] (M 173). He does not give any descriptions or titles so it is unclear what exactly he
means, but presumably this is a reference to the kind of literary manuals available in Braj Bhās: ā
or Sanskrit.
¹¹⁰ We can think of this as equivalent to “convergent evolution,” in which organisms develop
similar traits independently, such as bats and birds both developing wings through separate
evolutionary chains. Obviously Ārzū was not thinking in such Darwinian terms, but for us the
parallel is almost unavoidable.
¹¹¹ The word in hindī is actually “jhār: ū” (with an aspirate and a retroflex unlike in Persian),
but Ārzū ignores the distinction as irrelevant.
¹¹² Gilliot 1990, 95.
¹¹³ “wa ān ishtirāk ast dar as: l wazʿ

kih hīch yakī az dīgarī ak: hż na-kardah bāshand bi-ʿainihi”
(M 209–10).
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

86       

been phonetically transformed in the target language—and that one

language has not borrowed the word from the other. In fact, the word
need not be identical in both languages, but the difference cannot be in
its root (however that is to be defined). Thus, for example, “das”/“dah”
[ten] and “mās”/“māh” [moon] are the result of tawāfuq since the sound
change is trivial.¹¹⁴ This is suggestive because if a word did not develop
independently in two languages as in coincidence and if it was not
borrowed from one language to another as in Persianization then only
one possibility is left for us: The languages in which it appears must have
developed from a common ancestor. Unfortunately for his readers in the
distant future, Ārzū never explicitly draws such a conclusion with regards
to tawāfuq. Perhaps it depends on some kind of theory of similitude that
we, trapped in the mental habits of modernity, cannot access.¹¹⁵ In that
case, some kind of fundamental sameness could exist in two languages
without necessitating the conclusion that they are sister languages.
We cannot assume that when Ārzū describes tawāfuq he has an
understanding of history that would be familiar to us, but he certainly
considers language in the past. For example, noting the aftermath of the
Arab conquest of Persia as a time when Arabic words came into Persian
in large numbers, in identifying the transformation of Pahlavi to Dari to
Farsi, and in recognizing that “book hindī” (Sanskrit) is “unmixed” with
Persian in contrast to modern Indic languages. Unlike Persianization,
which continues into the present, correspondence is limited to the past,
as in the example of “ānk” [notation, number], which Ārzū argues
cannot be the result of correspondence as it does not appear in the
works of the Ancients [qudamā] or the “Middles” (mutawassit::ān, that
is poets who fall between the Ancients and the Moderns), or indeed any
modern work except that of the (now obscure) poet Tās̄īr, but rather is
:
spread by present-day merchants [saudāgarān-i hāl].¹¹⁶

¹¹⁴ M 210. He also gives “nīst”/“nāst” [“is not”] as an example of tawāfuq, so certain more
significant sound changes are also permitted (M 213). A fuller discussion of this particular
example appears in Chapter 5.
¹¹⁵ Michel Foucault describes a parallel similitude in early modern Europe in The Order of
Things (Foucault 1994, esp. 17–30).
¹¹⁶ M 217. The entry for “ang” in Ārzū’s lexicon Chirāġh-i Hidāyat has a similar definition as
given in Mus̄mir and exactly the same Tās̄īr quote (but with the spelling “ang” instead of “ānk”)
He notes in the definition that in hindī it is known as “ānk” and “ang” is Persianized. On Tās̄īr,
see MN 2004, 1:309.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

̄̄’     87

The use of the term tawāfuq itself is not new as it appears in al-
Muzhir, but the purpose to which Ārzū puts it is both innovative and
crucial to his intellectual project. Al-Suyūt::ī’s section on tawāfuq in al-
Muzhir is perfunctory, and upon comparison it is clear that Ārzū has
taken a relatively minor concept and transformed it into a central tenet
of his philology.¹¹⁷ He writes

Until now, no one had discovered the correspondence [tawāfuq]

between hindī and Persian; even in the lexicographers’ every effort,
researchers in this art [i.e., lexicography] were not aware of what is
hindī and what is Persian and otherwise. But humble Ārzū—a person
who is derivative [in thought], an old man, and incapable—established
this for the first time, while correcting some Persian words, as in the
books written by him, such as Sirāj al-Luġhat, Chirāġh-i Hidāyat, and
so on. It is strange that Rashīdī [i.e., ʿAbd al-Rashīd al-Tattawī, author
of the dictionary Farhang-i Rashīdī] and others were in India, and yet
never noticed the degree to which there is correspondence in the two
languages.¹¹⁸

Correspondence is mentioned in Mus̄mir in no fewer than six chap-

ters.¹¹⁹ Since the concept appears in Sirāj al-Luġhat, Ārzū must have
been thinking about it by at least the early 1730s. Besides Sirāj al-Luġhat
and Chirāġh-i Hidāyat, it also appears repeatedly in Nawādir al-Alfāz̤
(his dictionary of Indic words), and is mentioned once briefly in
Mauhibat-i ʿUz̤ ma̍ and once at length in Khiyābān-i
: Gulistān.¹²⁰ This
common thread in Ārzū’s work also provides an opportunity for Ārzū to
research Indian culture and to bring it to a Persianate readership. These

¹¹⁷ Al-Muzhir 1998, 209–11. Shams-i Qais (fl. seventh-century /thirteenth-century )
earlier used the term tawāfuq in a different sense, namely as “harmony” of lines and half-lines in
a couplet (Windfuhr 1974, 337). This appears to be the same concept that the Persian/Urdu
tradition later called rabt::. Ārzū himself appears to use the word in this sense at M 68.
¹¹⁸ “tā al-yaum hīchkas ba-daryāft-i tawāfuq-i zabān-i hindī wa fārsī bā-ān hamah kas̄rat-i
:
ahl-i luġhat chih fārsī wa chih hindī wa dīgar muhaqqiqān-i īn fann mutawajjih na-shudah-and
illā faqīr-i ārzū kasīkih muttabiʿ wa pīr wa īn ʿājiz bāshad wa īn rā as: lī muqarrar kardah wa
bināʾī tas:hī
: h-i
: baʿzī
alfāz̤ -i fārsiyah ba-dīn gużashtah chunānkih az kutub-i mus: annafah-yi
k: hwud
: mis̄l-i sirāj al-luġhat wa chirāġh-i hidāyat wa ġhairah niwishtah-am [sic] wa ʿajab ast
az rashīdī wa ġhairah kih dar hindūstān būdah-and wa hīch lihāz : ̤ na-kardah-and kih dar īn do
zabān chih qadr tawāfuq ast” (M 221).
¹¹⁹ M 59, 65, 115, 171, 175–9, 195, 209, 213, 217, 218, 221, 251, 269.
¹²⁰ MʿU 2002, 100; KhG : 1996, 113–14.
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

88       

discussions lead to fascinating—if occasionally perplexing—observations,

such as a wonderful learned discussion of the etymology of Kandahar
(a city in Afghanistan), in which he floats the possibility that the name is
derived from the name of Queen Gāndhārī in the Sanskrit epic
Mahābhārata. Citing the geographer Mīr Ġhiyās̄ al-Dīn Mansūr : as his
authority, he argues that the city is technically located just within Indian
territory on the basis that India has been defined as the place where the
black antelope [ahū-yi siyāh] is found. The black antelope, or blackbuck
(Antilope cervicapra), is identified with India because of its ritual signifi-
cance to Brahmins, since its skin is made into a mendicant’s seat (kusha
grass is used similarly).¹²¹ Despite this fascinating proto-anthropology,
tawāfuq also leads to what are now historically unpalatable conclusions,
such as Ārzū’s reference to the correspondence of Pahlavi and Arabic, a
linguistic impossibility if tawāfuq means that the languages have a com-
mon origin.¹²² We must be careful, in recognizing the value of Ārzū’s
scholarship, not to impute to him motivations or ideas that he could not
possibly have thought, simply to sooth our discomfort over what (from
our perspective) he got wrong.

Figurative Language and Where Meaning Comes From

Now let us turn to the last of the major themes addressed in Mus̄mir,
meaning and figurative [majāzī] language. The importance of Mus̄mir’s
discussion of these topics lies not in the account itself. Rather its value for
us is in its application: The proper or improper use of metaphorical
language is the substance of the analysis in the critical works to be dealt

¹²¹ M 218. While some of Ārzū’s information comes from scholarly texts including “old
Persian histories of India” (M 172), often it is uncited and presumably derives from Ārzū’s
personal knowledge.
¹²² M 59, 195. Arabic and Persian (whether Pahlavi or any other form) are not genetically
related; the former is a Semitic language and the latter Indo-European. Al-Suyūt::ī appears to
suggest (in a quote from Ibn Jinnī) that the words shared between Arabic and other languages
had a common origin. If al-Suyūt::ī provided a fuller description of tawāfuq that might clarify the
issue, I have not found it. Europeans made precisely the same mistake: The Anglo-Irish
Orientalist and diplomat Gore Ouseley (1770–1844) wrote to his brother William in 1792
that “Arabic is doubtless the mother of Persian; but, by the same rule, we should begin with
Sanscrit, which is mother and grandmother of them both” (Biographical Notices 1846, xviii).
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

̄̄’     89

with in Chapter 3, such as Ārzū’s Tanbīh al-Ġhāfilīn and Sirāj-i Munīr.

For the sake of completeness and because the supposed demerits of a
particular kind of figurative language, namely īhām (loosely translated as
“punning” or more exactly as “amphibology”),¹²³ loom so large in the
history of Urdu criticism, we should sketch the terminology of metaphor
and simile operative in Ārzū’s thought. Western and Islamicate concep-
tualizations of figurative language share an Aristotelian heritage, but the
Islamicate tradition developed a more nuanced—and, in Ārzū’s case,
historically minded—approach to metaphor.¹²⁴ Ārzū refers to tashbīh
[simile], istiʿārah [metaphor/trope], kināyah [metonymy], tams̄īl [alle-
gory], majāz-i mursal [“free trope”], and īhām, all of which are different
basic approaches to rhetorical comparison. However, in the Arabic trad-
ition of rhetoric, which the Persian tradition inherited, the terms relating
to metaphorical language had considerable overlap, creating difficulty for
later scholars.¹²⁵ For example, istiʿārah and badīʿ [literary ornament] were
considered synonyms in early works, but eventually the former came to be
considered a particular subset of the latter.¹²⁶ Furthermore, the basic
concepts can be joined up, as in istiʿārah bi’l-kināyah [metonymic simile].
In four chapters near the beginning of Mus̄mir, Ārzū engages with the
debates on meaning presented in al-Muzhir: He agrees with al-Rāzī and
his followers on the question of whether every meaning has a separate
word for it and vice versa—it does not, they conclude.¹²⁷ The discussion

¹²³ Amphibology is “ambiguity deriving from grammar, morphology, or syntax” (Dupriez

1991, 31–2). Another term that could plausibly translate “īhām,” namely semantic syllepsis, is “a
figure by which a word or expression is used simultaneously in its literal and figurative senses”
(Dupriez 1991, 440–1).
¹²⁴ On Aristotle and Arabic poetics, see Bonebakker 1970, 92–5 and Cantarino 1975, 64ff.
Good overviews of metaphor in Persian literature are Garcin de Tassy 1873; Meisami 2003,
319ff.; Seyed-Gohrab 2012. Shamsur Rahman Faruqi translates “kināyah” as “implication,” but
I have preferred “metonymy” here.
¹²⁵ Even were this not the case, the translations of these concepts will necessarily be
imprecise. Noting Earl Miner’s concern that various common critical terms in the West (for
example, “representation,” “fiction,” and “originality”) presuppose that the goal of literature is
mimesis—which is not necessarily the case for non-Western literatures—Julie Meisami has
argued that we can still use such terms as long as we reflect on the systemic differences (Meisami
2003, 4).
¹²⁶ Meisami 2003, 320.
: al-Dīn Rāzī (d. 606 /1210 ) was, depending on the source, either the
¹²⁷ M 21–3. Fakhr
greatest legal scholar of his time or a dangerous enemy of orthodoxy. He was born in Ray (now a
suburb of Tehran) but traveled throughout Central Asia. He was prolific in Arabic and Persian,
writing more than 100 books primarily on jurisprudence and philosophy (Kholeif 1966).
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

90       

is partially what we might call anthropological, in that Ārzū observes that

hindī does not have an indigenous word for hammām : [bath-house]
because Indians bathe in rivers, while there is a Persian word for bath-
:
house, garmābah. Both the Arabic-derived hammām and the indigenous
Persian garmābah co-exist in Persian and mean the same thing (likewise
the religious terms namāz [prayer] and rozah [fasting] which are avail-
:
able in Persian alongside their Arabic equivalents salāt :
and saum). He
picks up a discussion from Qāzī
ʿA zud

al-Dīn on the philosophical
problem of determining whether a word describes the quality of a
particular person or of a general category to which a particular person
might belong.¹²⁸ A discussion with a surprising relevance for modern
semantics follows, namely whether words share the same nature as their
referents [madlūlāt].¹²⁹ In Structuralist terms, this is asking whether the
relationship between signifier and signified is arbitrary or not.¹³⁰ Here
Ārzū disagrees with his Arabic source (ʿIbād bin Sulaimān al-Zamīrī : as
quoted in al-Muzhir) and notes that while there are onomatopoetic
words in which the connection is clear, for most words the relationship
is arbitrary. The Arabs did not properly consider other languages in this
regard, he argues, and such a comparative method would have demon-
strated that onomatopoetic words in one language are not so in another.
The next discussion is again derived from al-Rāzī, specifically from
Kitāb-i Mah: s: ūl, a text also known as al-Mah: s: ūl fī ʿIlm Us: ūl al-Fiqh
[Collection of the Knowledge of Jurisprudence], and deals with how to
establish the meanings of words.¹³¹ Ārzū summarizes the conclusions
relating to Arabic and writes “because religion is Arabic, one is powerless
to question this.”¹³² He provides some possibilities to explain how
meaning can be established in Persian, which is obviously not con-
strained by being the language of Islamic scripture as Arabic is. These
discussions are taken up in greater detail in Mauhibat-i ʿUz̤ ma̍ [The
Supreme Gift], Ārzū’s treatise on ʿilm-i maʿānī, which roughly translates
to semantics.¹³³

¹²⁸ M 23–5. ¹²⁹ M 25–7. ¹³⁰ See Czapkiewicz 1990.

¹³¹ M 27–34, 49–55. ¹³² “chūn dīn ʿarabī ast nāchār ast ihtiyāt
: :: dar ān” (M 28).
¹³³ The discipline ʿilm al-balāġhat, usually translated as “rhetoric,” is divided into three parts:
“(1) ʿilm al-maʿānī, the semantics of Arabic syntax; (2) ʿilm al-bayān, the theory of figurative
speech proper; and (3) ʿilm al-badīʿ, the remaining forms of rhetorical embellishment” (De
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

̄̄’     91

Considerably later in the text, Ārzū discusses the distinction between

“the real” [haqīqat]
: and “the figurative” [majāz].¹³⁴ This question is
considered part of ʿilm al-bayān, the subject of Ārzū’s ʿAt::īyah-i Kubra̍
[The Greatest Gift]. In Mus̄mir, the description is in keeping with the
tradition: Haqīqat

involves using a word in its narrow, specific meaning,
while majāz refers to its use in any extended sense. A key distinction
:
between the two is that qiyās is operative in haqīqat while in majāz it is
not.¹³⁵ This is a technical discussion in part derived from Ibn Jinnī as
quoted in al-Muzhir, and concludes with another endorsement of al-
Rāzī’s views. ʿAt::īyah-i Kubra̍ is another matter: It begins with the claim
that no one has written a Persian book on ʿilm-i bayān specifically and
suggests—in perhaps the most boastful sentence in all of Ārzū’s oeuvre—
that “this essay is the first [such] book which has been revealed from the
sky of lofty thought onto the terrain of Persian verse.”¹³⁶ It defines the
terminology of metaphorical language (tashbīh, istʿārah, etc.), and while
not in fact the only Persian text to describe such categories, it is rather rare
since most of the theorizing of metaphor in the Islamicate context, though
written by Persian speakers, took place in Arabic works.¹³⁷ The most
interesting intervention is a claim, which is not quite explicit, that meta-
phor is culturally bound: He points out that in Persian poetry unlike in
Indic poetry, a lover’s face is compared to the color red [rang-i :t:ilā], and
in Indic poetry unlike in Persian poetry, the eyes are compared to fish.¹³⁸
He closes this discussion somewhat defensively by arguing that the
examples given may be found in the works of the masters [asātiżah].
The significance of this careful analysis of categories of metaphorical
language is that it is the primary mode of criticism in Ārzū’s debates
with particular poets. The texts in question, which we will consider in
Chapter 3, are Tanbīh al-Ġhāfilīn, Sirāj-i Munīr, and Dād-i Sukhan. : In

Bruijn 1988). An important difference between Western and Arabo-Persian terminology is that
the word “k: hit::ābat” [oratory] is “applied strictly to the spoken word in public addresses, and is
not used in any wider sense” (Bonebakker 1970, 76; badīʿ discussed at 85).
¹³⁴ M 222–30, cf. ʿAK 2002, 69ff.
¹³⁵ “dar haqīqat
: qiyās jārī ast wa dar majāz nah” (M 224).
¹³⁶ “pas īn risālah awwal kitābī ast kih az āsmān-i fikr-i buland bar zamīn-i shʿir-i pārsī nāzil
shudah” (ʿAK 2002, 51). For an analysis of this text, see Dudney 2017a.
¹³⁷ On ʿilm al-bayān versus ʿilm al-maʿānī in Arabic and Persian, see Van Gelder 2009.
¹³⁸ ʿAK 2002, 65.
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

92       

many cases, the arguments take on a temporal aspect because the trad-
itionalist poets against whom he is contending claim that a certain use of
metaphor is new, but Ārzū is able to show that in fact it has a long history.
Although it is not discussed in Mus̄mir, one category of metaphorical
device that deserves special mention is īhām. Ārzū defines īhām as
follows: “Sometimes a word contains two meanings, one literal and the
second extended. Thus, the poets on the amplitude of common meaning
construct the extended one and equally allow both.”¹³⁹ Interestingly he
observes that not just poets, but people employing everyday speech [ahl-i
rozmarrah] also use it.¹⁴⁰ Husain
Wāʿiz
Kāshifī’s fifteenth-century
Badāʾiʿ al-Afkār fī Sanāʾiʿ

al-Ashʿār classifies īhām on the basis of whether
additional words are included that point to the literal meaning or not.¹⁴¹
:
In Mīr’s discussion of types of rekhtah (i.e., Urdu or mixed Persian-hindī
:
poetry) in the conclusion [khātimah] of his tażkirah Nikāt al-Shuʿarā, he
defines īhām as involving a near [qarīb] and a far [baʿīd] meaning in
which the poet actually means the far one and the reader must know to
dispense with the near one.¹⁴² Īhām was also seen by some Indian writers
as being a bridge between hindī and Persian: Ānand Rām Mukhli : s: quotes
a Persian verse by the seventeenth-century poet Salīm in which the word
“chāk-hā” (chāk means “rip” in Persian but “wheel” in hindī) appears
and he notes that it is “not without īhām in the language of the people
of India.”¹⁴³ Oddly, the modern critic Shafiʿi Kadkani claims that
īhām is frequently employed in Indo-Persian poetry because Indians
are non-native speakers of Persian—his explanation that they would
therefore focus on an individual word (and not its larger context) is

¹³⁹ “gāhī lafz̤ mushtamil do maʿnī bāshad yakī haqīqī

: wa duwwum majāzī. goyā shuʿarā banā
bar tausiʿah-yi maʿnī-yi ʿumūm majāz wa mustarak har do jāʾiz dāshtah-and” (DS 1974, 26).
On its history, see Chalisova 2004. Prashant Keshavmurthy has argued, in a personal commu-
nication, that devices like īhām recapitulate Ārzū’s distinction between the interpretation of
poetry by laymen and experts since they invite a simple interpretation and a nuanced inter-
pretation. See Chapter 3 for an analysis of this distinction.
¹⁴⁰ The linguist George Lakoff observes that “the locus of metaphor is thought not language”
and that the traditional distinction between metaphorical and literal language cannot hold
because most discourse involves some kind of metaphor (Lakoff 1993).
¹⁴¹ Chalisova 2004. I thank an anonymous reviewer for the suggestion to point to Kāshifī’s
definition.
¹⁴² Nikāt al-Shuʿarā 1979, 161.
¹⁴³ Mirʾāt 495. The symbolism in Salīm’s quatrain is very convoluted.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

̄̄’     93

unconvincing to me.¹⁴⁴ However, it is certainly true that the ambiguity

and misdirection inherent in īhām became an important aesthetic tool
for early modern Persian and later Urdu poets.

Conclusion

The way in which Ārzū classifies figurative language is, as we consider in

Chapter 6, arguably characteristically early modern since he has rever-
ence for existing categories and the tradition that produced them, while
at the same time filling those categories with innovative ideas. Despite
being a significant scholarly achievement, Mus̄mir’s direct influence on
Ārzū’s contemporaries and later generations appears to have been min-
imal. According to the twentieth-century scholar Sayyid ʿAbdullah,
Wārastah—whom we met in Chapter 1 among Ārzū’s “rivals and
enemies”—mentions Mus̄mir in his treatise Mat::laʿ al-Saʿdain [The
Setting of Venus and Jupiter].¹⁴⁵ That quotation appears to be the only
substantive reference to Mus̄mir before ʿAbdullah’s rediscovery of the
text in the Punjab University Library in the mid-twentieth century.
Mus̄mir is a particularly useful text for scholars today because it distills
virtually all of his important ideas into one text written at the end of his
life, but its influence cannot be accurately gauged because Ārzū’s major
ideas, such as tawāfuq as applied to Persian and hindī, are not restricted
to this one work.

¹⁴⁴ Shafiʿi Kadkani 1981, 160. For further context, see Dudney 2017a.
¹⁴⁵ ʿAbdullah 1967, 142. Mat::laʿ al-Saʿdain was published by Naval Kishore in 1880.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

3
Innovation and Poetic Authority
in Eighteenth-Century Persian

Rhetoric was the common ground of poetry, history and ora-

tory; it could mediate both between the past and present and
between the imagination and the realm of public affairs.
Encouraging men to think of all forms of human discourse as
argument it conceived of poetry as a performing art, literature as
a storehouse of models.
(Greenblatt 1980, 162)

In seventeenth- and eighteenth-century India, Persianate literary culture

was experiencing a transformation that we can call a crisis of authority.¹
Let us define authority (in the context of literature) as broadly and cross-
culturally as possible: It is a tool to judge whether some aspect of a
composition is a success or a failure.² The Persian tradition does not
approach the question of assessing literary merit with such a catch-all
term, but it is useful to combine the concept of sanad, or literary
precedent, with the idea of the consensus [ijmāʿ] of contemporary
poets. The crisis of authority in this period sprang from a poetics that
explicitly valued newness, the “tāzah-goʾī” or “Fresh Speech” movement.
Although “freshness” had been invoked by earlier poets, its

¹ A condensed version of part of this chapter appeared as Dudney 2016.

² We should note Gadamer’s discomfort with this term: “The concept of tradition, however,
has become no less ambiguous than that of authority, and for the same reason—namely that
what determines the romantic understanding of tradition is its abstract opposition to the
principle of enlightenment . . . . It seems to me, however, that there is no such unconditional
antithesis between tradition and reason” (Gadamer 2006, 282). Gadamer’s implication that the
very idea of tradition must be historicized is crucial since the post-Romantic view of tradition as
a constraint to be overthrown by enlightenment (or perhaps “development”) has little relevance
for an emic analysis of a non-Modern, non-Western intellectual tradition.

:
India in the Persian World of Letters: Khān-i Ārzū among the Eighteenth-Century Philologists. Arthur Dudney,
Oxford University Press. © Arthur Dudney 2022. DOI: 10.1093/oso/9780192857415.003.0004
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

    95

crystallization into a wide-spread tendency became an insurgent threat

to the smooth operation of literary precedent because in many cases
poets and readers approved of compositions that contained phrases and
literary images for which there was no obvious precedent.³ Thus critics
had to struggle with the relative value of precedent and consensus.
Specifically, we can consider how Ārzū simultaneously reiterated the
importance of tradition as a unifying factor across the Persian world of
letters while systematizing the expertise of living poets. This chapter will
use his influential disagreements with the long-dead Abū al-Barakāt
Munīr Lahorī (1019–54 /1610–44 ) and with his contemporary
: Muhammad
Shaikh : ʿAlī Hazīn

Lāhījī (1103–80/1692–1766) to offer a
necessarily revisionist account of how this crisis of authority played out.
In Ārzū’s framing, Munīr and Hazīn

both stood for a literary purism that
valorized the works of the pre-tāzah-goʾī poets and claimed themselves to
be the present-day guardians of this earlier poetic style. Ārzū by contrast
defended the new poetics on the basis that its aesthetics were not a
departure from the Persian tradition taken as a whole. He documents
this through careful scholarship, in the process arguing that Munīr and
Hazīn’s

judgments are often capricious rather than anchored in research.
The historiographical difficulty is that the rupture caused by tāzah-goʾī
has almost without exception been unhelpfully framed by nineteenth-
and twentieth-century scholars as a centuries-long contest between
Iranian and Indian aesthetics. From the nineteenth century until
recently, the interpretation that there was a degraded, particularly
Indian Persian (often called the “Indian style” [sabk-i hindī] in Persian
poetry) had been almost universally held (even by Indians themselves).
However, recent scholarship has shown this to be an anachronistic
framing.⁴ We will return to the relevant eighteenth-century critical

³ For example, in Haft Aurang the fifteenth-century poet Jāmī described his fourteenth-
century predecessor Hāfiz
:: as writing fresh [tāzah] poetry (Ingenito 2009, 163). I thank the
anonymous reviewer for the reference.
⁴ Alam nicely frames the issue (2003, 182). Just as Allison Busch has taken modern Hindi
critics to task for their arbitrary assessments of the quality of early modern courtly Hindi
literature (the rīti tradition), we need to revisit the standards used to determine what good
Persian style is (Busch 2011, preface, 14). Sheldon Pollock has called attention to the failure of
modern critics to engage with traditional commentaries—we can take that as inspiration for this
chapter to discuss critical works rather than the poems they analyze (2009, 254–5).
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

96       

texts, including Lut::f ʿAlī Beg Āżar’s tażkirah Ātashkadah [The Fire
Temple, 1174/1760], and trace the later scholarly interpretation that has
accreted to them. On the one hand, I will historicize the anti-Indian
sentiments of nationalist Iranian critics and Indians’ later lack of confi-
dence in their own Persian. On the other, I will use the analogue of the
Quarrel of the Ancients and Moderns, a roughly contemporary debate in
Europe over a roughly similar crisis of authority, in order to reconsider the
narrative of decline and cultural failure that is generally used to explain the
fortunes of Indo-Persian. But first we have a party to attend . . .
Munīr, a young literary luminary during the reign of Shāh Jahān,
records a terrible soirée. In the preface to his Kārnāmah [Commentary],
he describes how it began:

Because I brought only a book of poetry besides myself, although it was

my collected poetry, I chose silence; I was a spectator on the margins of
this assembly and I heard the conversation of the people who see fine
points, and the bright light—the burning of the lamp of cleverness—
entered my intellect through the sight of my eyes and through my ears.
And in the beauty of the conversation I beheld the poetry of the bright-
faced ones.⁵

The recitation [mushāʿirah] quickly goes awry. Soon,

in succession poetry emerged in that assembly, and the speech of the lords
of poetry appeared in it. All began to slander the earlier [pīshīn] poets, and
to destroy the praise of the earlier master versifiers; they undertook the
slandering of the deceased caravan of meanings and then the rest began to
praise the still-living people on the journey of poetry-knowing.⁶

As the assembled poets disrespect the exemplary writers from centuries

before, they heap praise upon their own contemporaries and near con-
temporaries. Someone says of ʿUrfī Shīrāzī (d. 999/1591) that

⁵ Kārnāmah 1977, 3. The text has been published in an edition including Ārzū’s response,
Sirāj-i Munīr [A Lamp for Munīr]. Munīr’s prose was well-regarded enough to appear on a
Mughal syllabus in 1688 (Syed 2012, 289). On his life, see Mohiuddin 1971, 221ff. For a useful
summary of the rhetoric of these passages, see Alam 2003, 182–3.
⁶ Kārnāmah 1977, 3.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

    97

This master versifier is the lord of fresh style [tarz:: -i tāzah] and the
manifestation of unguessable grandeur, writing such fine poetry and
bestowing delicateness on his verse, adorning each hair with the ringlet
[t::urrah] of poems and adorning the face of meanings in a most
pleasing way. The color of his words is the blush on the face of meaning
and the ink of his verse is the best example of the fine-points [mawādd,
pl. of māddat] of poetry.⁷

Similar hyperbolic praise is lavished upon T :: ālib Amulī (d. 1036/1627),

Zulālī (d. 1031/1622 or 1034/1625), and Z:: uhūrī Turshīzī (d. 1035/1626).
Then the assembly turns to denigrating two classical-style poets, Razī⁸

and Kamāl Is: fahānī (d. 635/1237), before putting two of the tradition’s
other revered poets in their place:

:
If [Amīr] Khusrau [d. 725/1325] had managed to converse with [these
Modern poets], he would have acquired the delicacy of their sweet
poems, and if Salmān [Sāwajī, d. 778/1376] had lived in their time he
would have learned Persian from their wives [ahl-i bait].⁹

These comments are hyperbolic, and it becomes clear to the reader that
Munīr is not describing a real event but has invented the occasion to
:
serve his rhetorical purposes. In the face of the absurd idea that Khusrau
and Salmān Sāwajī had something to learn about poetry from more
recent poets’ families, Munīr finds himself obliged to spoil the mood of

⁷ Kārnāmah 1977, 4.
⁸ Identified by the editor, in my view incorrectly, as Mīr Razī
Dānish Mashhadī (d. 1072/
1661–2), a poet of Shāh Jahān’s reign who settled in the Deccan. Since context demands that this
be a classical poet, the likelier Razī
is Ustād Razī
al-Dīn Nīshāpūrī, whom Ārzū notes has
affinities with the omni-talented thirteenth-/fourteenth-century poet Amīr Khusrau, : who
:
praised him (MN 2004, 2:454). In fact, Khusrau mentions Razī

Nīshāpūrī in the famous critical
preface to his dīwān Ġhurrat al-Kamāl. Further evidence comes from the introduction to
Munīr’s kulliyāt, which refers to Razī
al-Dīn Nīshāpūrī as one of the poets whom Munīr
counted as a significant influence (quoted in Akram 2009, 351).
⁹ Kārnāmah 1977, 6. Although “ahl al-bait” can refer to the family of the Prophet
:
Muhammad and his descendants, who especially in Shiʿism are seen as forming a spiritual
lineage from the Prophet to the present day, context here suggests that we must read the phrase
literally as “people of the house” or wives. I thank Nathan Tabor for our helpful dialogue on this
point.
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

98       

the imaginary gathering by jumping to the defense of the older poets.

He declares,

I, who am the mirror-bearer of justice [āʾīnah-dār-i ins: āf-am], saw that

these iron-hearted ones were bandying about ideas that were far off the
mark of justice and their poetry was no more than an imitation [lit. a
face reflected in a mirror] and in no way assumed [its own] form. So
I said, “Justice Knowers! The face reflects badly in the shining of the
mirror of justice and the distraction of its own thoughts. As you speak,
each mirror of yours is in the image of the mirror of imitations and you
view the mirrored face backwards [since] you consider these [Modern]
poets more worthy than those. So don’t talk rubbish and prefer and
exalt these fresh receptacles [māʾbān] over those who have come
before, and don’t follow the path of infidelity to justice!”¹⁰

As this passage and its reference to “fresh receptacles” implies, Munīr

saw a stark divide between the Ancients [mutaqaddimīn], who wrote
well in his estimation, and the Moderns [mutaʾak: hk: hirīn], whose “fresh
speaking” had driven them to disrespect the aesthetic achievements of
the Ancients. By the seventeenth century, the tradition had generally

:: in the late fourteenth
begun to consider as Ancients the poets from Hāfiz
century back to the earliest New Persian poet, Rūdakī, in the tenth.¹¹
Given that Kārnāmah has been cited as the first salvo against “Indian
style” excesses in poetry, it is surprising that Munīr’s framing of the
debate in the preface has nothing to do with Iran versus India. Indeed,
the only mention of either place at the beginning of Kārnāmah is an
expression of hope that Indians and Iranians both will be convinced by

¹⁰ Kārnāmah 1977, 6.
¹¹ Still, Ārzū refers to the poetry of Hāfiz

:: as “namak-i tāzah” [fresh-flavored, lit. fresh salt],
which we can probably read as an echo of tāzah-goʾī even though Hāfiz
:: came well before that
movement (M 1991, 11). Some sources consider the classical period to have ended with Jāmī
(d. 1492), referring to him as “the final poet” [shāʿir-i k: hātam] (see Browne 1959, 26; Losensky
1998, 193). Akbar’s poet laureate Faizī
(d. 1595) writes of Jāmī that: “In his society no one
appeared after him / And he is the seal of prose and poetry” (trans. Sharma 2012a, 239).
A detailed recent appraisal of Jāmī’s place in the canon and his role in shaping it is Lewis 2018.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

    99

his claims.¹² Furthermore, one of the Ancients slighted at the imaginary

:
gathering, Amīr Khusrau, was an Indian, while all four of the Moderns
whom Munīr believes to have received unearned praise were Iranian by
birth. Munīr was concerned with stylistic chronology and not geography.
If Munīr’s attacks on his contemporaries seem intemperate, they are not
at all out of place. The early modern Indo-Persian literary scene was
:
contentious, and research [tahqīq] could be a weapon.¹³ Some two gener-
ations separated Munīr from Ārzū, but nonetheless Ārzū felt the need to
criticize him at length in two works, which will be introduced later in the
chapter. This debate between the living and the dead was hardly the unfair
contest it would seem to be because Munīr had many defenders among
Ārzū’s contemporaries.¹⁴ However, the feud that would become much
better known was Ārzū’s disagreement with Shaikh : Hazīn,

his exact
contemporary. Ārzū’s broadside against Hazīn,
Tanbīh al-Ġhāfilīn
[Admonition to the Heedless, ca. 1157/1744], is a critical tour de force
which has been framed as the most coherent Indian response to the charge
that the Persian used in India was inferior to Iranians’ Persian (although
that was not its stated intention).¹⁵ Driven from Isfahan by political unrest,

¹² “I hope that the poetry-knowers of India and the meaning-makers of Iran will not revile
my correct speaking and will be apologetic” [ummīdwāram kih suk: han-shināsān-i hind wa
maʿnī-rasān-i īrān bar rāst guftārī-yi man kajmaj zabān girift nakunand wa pozish dar
pażīrand] (Kārnāmah 1977, 7).
¹³ For example, Munīr refers to “s: ulh-i
: kull,” the concept of tolerance famously put into
practice by the emperor Akbar, but which refers etymologically to an armistice (Kārnāmah
1977, 7). Similarly Ārzū writes of Mụhammad Afzāl
S̄ābit that “sometimes he is at peace and
sometimes he is at war with me at a poetic recitation” [bā faqīr bar sar-i shiʿr-i bait al-hāl
: gāhī
s: ulh: wa gāhī jang dāsht] (MN 2005, 69). Thanks to Rajeev Kinra for bringing my attention to the
possibilities of “s: ulh.”
: Another set of rhetorical key terms having to do with justice, ins: āf and
dād, will be explained further in the chapter.
¹⁴ Notably Wārastah but also lesser-known poets like Hākim

Lahorī (Akram 1977, 39). The
staying power of the debate between Munīr and Ārzū was such that it would be rekindled in the
mid-nineteenth century by Imām Bakhsh : Sahbāʾī

in a work called Qaul-i Fais: al [The Last
Word] (published by Naval Kishore as Risālah-yi Qaul-i Fais: al n.d., see also Naim 2006).
¹⁵ Hākim

is best known for his Tażkirah-i Mardum-i Dīdah [Biographical Dictionary of
People (I Have) Seen, or Biographical Dictionary of the Pupil of the Eye] composed in 1761–2
and which frames many entries around long quotes from Ārzū’s MN (Tażkirah-i Mardum-i
Dīdah 1961; Storey 1953, I.ii: 829). Ārzū claims Hākim

as a friend, citing a difficult-to-translate
verse by him on their friendship: “zi dunyā wa z māfī-hā, zi dunyā wa z māfī-hā / hamīn yār
ārzū dāram, hamīn yār ārzū dāram” (MN 2004, 1:396). Hākim,
despite his apparent friendship
with Ārzū and respect for him, writes in his tażkirah that Tanbīh al-Ġhāfilīn is mostly unfair,
namely that Ārzū has committed “sitam-sharīkī” or “partaking in oppression” (Tażkirah-i
Mardum-i Dīdah 1961, 66). Ārzū’s criticism of Hākim’s
dīwān, discussed in the context of
Wārastah in Chapter 1, might have strained their relationship.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

100       

Hazīn

came to India in 1147/1734 and was treated as a celebrity, as Ārzū
himself notes in the preface to Tanbīh al-Ġhāfilīn.¹⁶ In Indian cultural
memory, including in folk tales, Hazīn

is the archetypal haughty Iranian
émigré, irascible and judgmental.¹⁷ This narrative is not the whole story
because Hazīn

had numerous Indian supporters, and in any case says little
about India, good or bad, in his autobiography.¹⁸ Likewise, Ārzū does not
object to any Iranian chauvinism on Hazīn’s

part but rather to his literary
conservatism, demonstrating that in many of Hazīn’s

verses, Hazīn

vio-
lates his own precepts and uses expressions for which there is no prece-
dent. The far more explosive accusation was laid by Mīr Muhammad :
ʿAz::īm, known by his pen-name S̄abāt (1122–61/1710–48): He accused
Hazīn

of plagiarizing some 500 verses.¹⁹

¹⁶ TĠh 1981, 1.
¹⁷ Perry 2003; Khatak 1944; Kirmani 1986, 30. Mana Kia has argued convincingly that
Hazīn’s

apparent dislike of India needs to be seen through the lens of his personal experience,
and should be understood not as indicting Indian culture so much as lamenting his inability to
return to his devastated native land (Kia 2009). Faruqi on the other hand points to Hazīn’s

“pure
malice” (Faruqi 2004b, 17).
¹⁸ The chapter in which he purports to describe India (Life of Sheikh Mohammed Ali Hazin
1830, 275–83) is devoted to Iranian rulers’ relations with India and says practically nothing
about the place itself. However, as discussed in Chapter 1 of this study, in Majmaʿ al-Nafāʾis
Ārzū accuses him of having been an ungrateful and insulting guest in India.
¹⁹ Crucially the objection is not that words of others were copied but rather that themes were:
“S̄abāt pointed out 500 couplets from the dīwān of the Shaikh : [i.e., Hazīn]

whose themes were in
the form of others’ themes” [s̄abāt pāns: ad bait az dīwān-i shaik: h rā bar-āwardah kih mazāmīn-i


ānhā ba-jinsah az dīgarān ast]. This is cited by Hazīn’s

friend and erstwhile traveling compan-
ion Wālih in his tażkirah Riyāz
al-Shuʿarāʾ [Garden of the Poets, 1161/1748] (Tażkirah-yi Riyāz

al-Shuʿarāʾ 2005, 647ff; also quoted in Akram 1981, 30; partial trans. Faruqi 2004b, 37).
According to Wālih, what raised S̄abāt’s ire was a comment by Hazīn
on a verse of his father’s.
(He was the son of the well-known poet Mụhammad Afzal
S̄ābit.) Hazīn

wrote that the verse’s
theme [mazmūn]

was not very good and was, in any case, “stolen” [duzdīdah ast] from some
other poet. In a fit of pique at this negative comment on a single verse, S̄abāt copied out 500
verses by Hazīn

along with the verse he had supposedly plagiarized in each case. Wālih’s entry
on Hazīn

is the longest in the tażkirah, taking up nearly forty pages in the printed edition and

:: in second place at just over thirty pages (Ārzū by contrast gets about a page).
leaving Hāfiz
Remarkably he includes a lengthy selection of Tanbīh al-Ġhāfilīn (shortened either intentionally
or because it was copied from a manuscript recension we apparently no longer have) as well as
the couplets S̄abāt accused Hazīn

of plagiarizing (Tażkirah-yi Riyāz
al-Shuʿarāʾ 2005, 635–47,
647–57). Before introducing the selection of Hazīn’s

work, he argues that the previous two
dozen pages of criticism notwithstanding, Hazīn
is a world-renowned poet [sarāmad-i
suk: hanwarān-i ʿālam] as his quoted verses prove. We can speculate as to why so much of
Hazīn’s

entry in a tazkirah

by his friend and traveling companion is taken up by criticism of
Hazīn.

Mana Kia has argued that it was in order to shame Ārzū and S̄ābit rather than to lend
support to their position (Kia 2011, 219). Clearly, however, the Ārzū–Hazīn
debate was
important enough to merit inclusion at length.
Another significant instance of plagiarism was the wholesale inclusion by ʿAt::āʾallāh Nudrat of
entries from SL and CH (as well as from Bahār-i ʿAjam) in his own dictionary (MN 2005, 121).
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

    101

The “Indian Style” and India

From Munīr’s preface in Kārnāmah, it is abundantly clear that his

concern in the work was defending the honor of the Ancients against
the free-wheeling Moderns. But this was not the frame adopted when
later critics and historians reflected upon the seventeenth and eighteenth
centuries. Rather, those who are aware of Munīr’s work have preferred to
view his criticism as the first salvo in an attack on the “Indian style”
[sabk-i hindī].²⁰ The label “Indian style” purports to describe the poetic
modes popular across the Persianate world from the late sixteenth to the
early nineteenth century. However, because the name refers to a specific
place, India, many critics have claimed explicitly that the supposed
degeneracy of the literature of the period springs from the Indian
environment.²¹ There has been an undeniable South Asian influence
on Persian literature for centuries, mostly in the form of word borrow-
ings. It remains an open question whether Indian philosophy and
Sanskrit literary culture (whether mediated through vernacular poetics
or directly) had a more subtle influence.²² However, there is no evidence
that such mixing was marked as foreign to the Persian cosmopolis, and

(Note that this is not the same Nudrat who was a Hindu disciple of Bedil and friend of Mukhli : s: .)
However, it is worth noting that plagiarism in the context of scholarly work was regarded
differently from poetic plagiarism. A full discussion of this contrast is outside our present scope.
²⁰ The editor of Kārnāmah and SM himself writes that “Munir is the first critic who objected
and raised his voice against the Sabk-i-Hindi in Lahore, in the first half of the eleventh century
A.H., whereas this style of poetry was criticised in Isfahān at the end of the twelfth century A.H.”
(Akram 1977, English introduction).
²¹ For example, M. J. Borah connects the historian Firishtah’s claim that Hindus began to
take up the study of Persian seriously during Sikandar Lodī’s reign to the sabk-i hīndī debate by
arguing that “with the growing influence of the Hindu scholars who began to study Persian to
qualify themselves for the service of the State, the difference in the style of India and Persia
proper became more marked” (Borah 1934, 36; cf. Mohiuddin 1960, 24–5). This “Hinduization”
of Persian has often been assumed but never satisfactorily demonstrated. One recent history of
Iranian emigration makes precisely the same claim that sabk-i hindī came about because of
contact with “Hindu philosophy and thought” [ʿaqāʾid wa andeshah-yi hindū-ān] (Rafīʿ 2004,
334ff.). For wider context see Kinra 2007.
²² For example, Faruqi suggests that Bedil’s definition of suk: han [speech, poetry] echoes the
thought of the fifth-century Sanskrit grammarian Bhartr̥hari (Faruqi 2004b, 19). He also
speculates that there might be a connection between Sanskrit poetics and the “meaning-
creation” [maʿnī-āfirīnī] typical of early modern Persian poetry but admits “direct evidence is
lacking as yet” (Faruqi 2004b, 31ff.). If the evidence is not yet available, we can at least consider
the possibility that polyglot litterateurs of the Mughal period like ʿAbd al-Rahīm : (Naik 1966;
Sharma 2009; Lefèvre 2014) might have been a locus for such interactions.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

102       

in the early modern period it apparently produced none of the anxiety

that later Persian critics felt towards anything that could not be given a
properly documented Iranian provenance. The idea that India corrupted
Persian literature has been a convenient license for such critics to voice
their disapproval of all Persian literature of the period. What was so
objectionable to these critics and how did it connect with their inter-
pretation of early modern politics?
Mughal and Safavid-period literature has often been dismissed as the
product of overwrought formalism obsessed with wordplay at the
expense of emotion and truth. One twentieth-century Indian scholar
writes, “Poetry produced under the Moghuls in the twelfth century 
[roughly, the eighteenth century ], in India, is degenerate, stereotyped,
and imitative. There is nothing new, creative or original about it.”²³
Ehsan Yarshater writes of such poetry that “Within its span the Indian
style developed organically, followed a normal curve within certain limits
and constraints, and finally exhausted itself into a lifeless and forced
poetry.”²⁴ Discussions of the “Indian Style” have consistently been
framed in terms of taste, obscuring its rickety conceptual framework.²⁵
Both fair-minded scholars and ideologues have time and again conflated
these literary characteristics with cultural decline. But where does the
formulation that complexity accompanies intellectual or aesthetic

²³ Khatak 1944, 57.

²⁴ Yarshater 1986, 965. We can add to this list Shafiʿi Kadkani, who tries to rescue Iran from
the worst excesses of the Indian Style by positing that however bad things were in Iran, they
were far worse in India (Shafiʿi Kadkani 1981, 150ff.). Yarshater 1988 offers some choice quotes
from the nineteenth- and early twentieth-century critics.
²⁵ Cf. Hasan 1998, 6. The great Cambridge Persianist Edward Granville Browne, for example,
justifies his exclusion of Indian poets on the grounds that his book would have become too long
and because much had already been written about them in British India itself. However, after
this milquetoast protest, he declares that “so far as a foreign student may be permitted to express
an opinion on matters of literary taste, this Persian literature produced in India, has not, as a
rule, the real Persian flavour, the blar as the Irish call it, which belongs to the indigenous
product” (Browne 1956, 106–7). He continues in a vein that proves beyond a doubt that we in
the twenty-first century need to reconsider his judgments. He concludes, “therefore the omis-
sion of Amír Khusraw from this chapter is as justifiable as the omission of Walt Whitman from
a modern English literary history, especially as a very long notice of the former is given in Elliot’s
History of India.” The only Indian literary production he includes in this work, which was
crucial in shaping the later Western understanding of the Persian canon, is that of Mughal-
period Iranian émigrés. An important recent corrective to Browne’s project is Alam 2003.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

    103

decline come from?²⁶ And why “Indian”? The style was neither invented
in India nor particularly associated with India, except to the extent that
Persian poets, both Indian and Iranian, could at that time find better
opportunities for patronage in Mughal India than in Safavid Iran.²⁷ Ārzū
himself notes the importance of Bābā Fiġhānī of Shiraz (d. 1591) in
forming the literary style that would later become dominant, and Fiġhānī
had no ties to India.²⁸ Nor did practitioners of the “Indian Style” ever
refer to themselves as such. In fact, the term itself is no older than the
early twentieth century. It was popularized by Muhammad: Taqī Bahār
(1886–1951) in the introduction to his history of Persian literature,
Sabkshināsī.²⁹ He ties temporal divisions in Persian literature to place,
so the oldest style is “Khurasani,” the middle style is “Iraqi,” and the late

²⁶ The rhetoric of scholarship on Persian as it was used in South Asia involves some
fascinating logical contortions. For example, in his history of thirteenth-century literature,
Mumtaz Ali Khan (who is himself Indian) writes of a particular work that “It is free from
those artificialities, affectations, intricacies and wordplays [sic] which detract from the value of
some of the contemporary works, like the Lubāb al-Albāb and the Tāj al-Māthir” (Khan 1970,
58). By his own admission, the texts he has mentioned as flawed were greatly admired in their
time, so on what basis is he calling the style tedious? He is simply giving voice to his own
unexamined prejudices. (Here I use “prejudice” in Gadamer’s technical sense of a presuppos-
ition “that makes us deaf to what speaks to us in tradition,” Gadamer 2006, 271ff.) Another such
example is Khatak 1944, 57–8.
²⁷ Browne 1959, 26ff. See also Ghani 1930, 278ff.; Dale 2003, 199ff.; Yarshater 1988, 251;
Lewisohn 1999. There is the fact, which troubled Iranian nationalists, that most of the diction-
aries of the Persianate world during that period, arguably the golden age for Persian lexicog-
raphy, were produced in India (Perry 1998, 329, 338–9; Tavakoli-Targhi 2001, 106–7). For a
balanced recent account, see the editor’s preface to the 2001 edition of Tek
Chand Bahār’s
dictionary Bahār-i ʿAjam (Bahār the eighteenth-century lexicographer should not be confused
:
with the critic Muhammad Taqī Bahār mentioned in this chapter).
²⁸ M 11; cf. Losensky 1998. The closest he ever got to the sub-continent was a stint in Herat in
western Afghanistan.
²⁹ Bahar 1942. The pattern for Bahār is primeval cultural greatness followed by decline and
lastly the return to greatness exemplified by his contemporaries. It is a kind of self-
Orientalization that can also be observed in the Urdu tradition. The logic is exactly the same
:
as Muhammad Husain

Āzād’s Āb-i Hayāt

(discussed in detail in Chapter 5 of this study). What
is relevant here is that Āzād strains credulity trying to support his historically impossible
argument that the tradition of Urdu poetry began “naturally,” became decadent, collapsed,
and then recaptured its earlier simplicity under the benevolence of India’s British colonial
masters (Āb-i Hayāt

1907). A dispassionate analysis shows Āzād’s framework to be deeply
lacking. It is certainly a product of its time (1880, the zenith of British colonialism in India) just
as Bahār’s work reflects the nationalism of late-Qajar and post-Qajar Iran (Smith 2009, 196ff.).
Bahār “draws a border around literature of strictly Iranian origin, distinguishing it from—and
elevating it above—Persian poetry written outside those lines” (Smith 2009, 199). This project of
ethnic purification in Persian literature has been so successful that it is only recently that Iranian
scholars have begun to take Indo-Persian seriously.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

104       

style is “Indian.”³⁰ It is a deeply misleading frame of reference; I am

certainly not the first to argue that the term “Indian Style” (as in Bahār’s
teleological formulation) is so freighted as to be irredeemable, and that
we should just stop using it.³¹ However, it represents an attempt to
explain an underlying phenomenon which is both real and worth under-
standing, namely the transformation wrought by tāzah-goʾī. Unlike
“sabk-i hindī,” “tāzah goʾī,” also known as “t::arz-i tāzah” (“fresh style”)
or “ʿibārat-i tāzah” (“fresh expression”), was a contemporary label used
by both its adherents and detractors in the seventeenth and eighteenth
centuries.³² It is important to understand that in Bahār’s formulation
sabk-i hindī is a stylistic category referring to a specific historical period,
and not an analysis of linguistic and stylistic variations particular to
Indo-Persian throughout history.³³ It is therefore not the same as observa-
tions throughout history of an “Indian accent” [lahjah-yi hindūstāniyān] or
a particular “mode” [rawish] of Persian in India.³⁴
While some aspects of this multifaceted debate can and should be
interpreted as an Iranian identity defining itself against an Indian iden-
tity and vice versa, considering the crisis of authority in simple ethnic
terms (especially when we modern readers necessarily understand those
terms anachronistically as corresponding to national identities) stops
well short of a satisfactory explanation. To that end, let us dispense

³⁰ This same scheme is taken up by later scholars, e.g., Heinz 1973. In his earlier work, Ehsan
Yarshater used the term “Safavid style” instead of “Indian style” (Subtelny 1986, 58). This
however further confuses the issue because it implies that India played no role in Persian letters
(at a time when it obviously did). The best compromise, if we accept the periodization, seems to
be the ungainly term “Safavid-Mughal” (as in Losensky 1998).
³¹ Kinra 2007, 142n20. The interested reader can consult Kinra 2007 and Faruqi 2004b on
this question. Good context is also provided in Hanaway 1989 and Smith 2009, 196. Faruqi
memorably writes that Bahār has “a blind arrogance that better suits a provincial administrator
than a literary historian and critic” (2004b, 21). Oddly, Bahār wrote a cheerful poem about how
he pines for India (quoted in Barzegar 2001, x–xi). Bracketing off the question of whether this
was a formal exercise or heartfelt, it nonetheless suggests that he did not develop his tripartite
literary model out of antipathy for India, the namesake of the nadir of literary excellence in his
system.
³² Shafiʿi Kadkani introduces another term, “incidentalist style,” as an alternative (Shafiʿi
Kadkani 1981, 147). On tāzah goʾī in the Ottoman context, see (the delightful but occasionally
uncritical) Andrews and Kalpaklı 2004, esp. 352; and Darling 2012, 179.
³³ The Iranian literary scholar Zabih Allah Safa argues that sabk-i hindī is the correct term on
the basis that South Asia was at the center of Persian letters at the time, but admits that scholars
debate whether the style has an Indian origin (Safa 1984, 523–4).
³⁴ See Alam 2003, 149; Dudney 2017b.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

    105

with the “Indian Style” as a frame of reference and consider instead the
cultural politics of literary innovation: Early modern literary debates in
Persian were framed primarily in terms of temporality, that is, old styles
versus new styles. Cultural geography, which is to say India versus Iran,
is a distant secondary concern in the critical literature until the end of the
eighteenth century.³⁵ As early as the mid-eighteenth century, writers in
Isfahan began to rebel against tāzah-goʾī, and their criticisms were
increasingly inflected by the idea that the place in which Persian litera-
ture is composed matters. They advocated a return to the literary style
that existed before tāzah-goʾī, and because literary trends diffuse
unevenly, India remained a stronghold of tāzah-goʾī.³⁶ By the nineteenth
century, Indians had fully assimilated these critiques and themselves
privileged Iranian Persian over their own.³⁷ Hints of nationalistic senti-
ments in literature, an entirely secondary concern in the eighteenth
century, take on a special salience for us because we, and several gener-
ations of scholars before us, have been accustomed to thinking of
literature as constitutive of a nation rather than of some other cultural
unit, either smaller or larger than a nation. This is why Benedict
Anderson warns us to contrast the modern nation-state, which we take
for granted, with “large cultural systems that preceded it.”³⁸ Despite
sharing a name, eighteenth-century India is not the post-1947 Republic
of India, just as Safavid Iran is not today’s Islamic Republic of Iran. The
understanding that we are dealing with unfamiliar political formations

³⁵ It was a distant concern but not a non-existent one. Shamsur Rahman Faruqi notes that
“the earliest opponents of Indian Persian, like ʿAlī Hazīn

and Wālih Dāghestānī, were them-
selves distinguished poets of the Indian Style and they were disdainful of the Indian register of
Persian, and not of the Indian Style of Persian poetry” (Faruqi 2004b, 17). While he is correct in
the narrow sense that Hazīn

and Wālih would today be considered “Indian Style” poets, his
account must be nuanced by noting that both frequently stated their opposition to contempor-
ary poets while only rarely mentioning India (such as in the example he cites, which is the
exception rather than the rule). To ignore this introduces anachronism. Furthermore, Faruqi
does not consider Munīr in this context.
³⁶ Yarshater 1986, 965. And indeed it can be argued that the then-emerging Urdu literary
tradition, whose relationship to Persian I address in Chapter 4, retained some of the stylistic
complexity which was the hallmark of tāzah-goʾī.
³⁷ Of course there are some exceptions among Indian scholars, who themselves stereotype
:
the other way: Wahid Mirza declares, for example, that Amīr Khusrau has a “peculiar finesse”
lacking in all Iranian poets except Jāmī and Naz::īrī (Mirza 1935, 206).
³⁸ Anderson 1983, 19.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

106       

with relatively familiar names is crucial.³⁹ This also requires reflecting on

where such “fore-meanings” (to use Gadamer’s terminology) come from,
namely at least in part from critical texts that provide a compelling if
problematic framework for the tradition in question.
Lut::f ʿAlī Beg Āżar’s Ātashkadah [Fire Temple, 1174/1760], the
tażkirah associated with the genesis of the bāzgasht-i adabī [literary
Renaissance] movement (itself an anachronistic label popularized by
Bahār in the early twentieth century), was precisely such a category-
establishing work.⁴⁰ It divides up poets by region of their birth and
effectively ghettoizes India as a place where poetry was produced because
so many of the important writers were immigrants.⁴¹ Although he does
not explicitly state his disapproval of Indian poets, either collectively or
as individuals, his blind spots are clear from the content: Hundreds of
pages are devoted to Iranian poets, while Indians are represented in a
section containing a meager seventeen entries.⁴² Most of these are just a
few lines, with the accounts of just three poets, Faizī,

Hasan

Dihlawī, and
:
Amīr Khusrau, spanning more than a page. He mentions just seven poets

³⁹ As Hans-Georg Gadamer argues, “Just as we cannot continually misunderstand the use of

a word without its affecting the meaning of the whole, so we cannot stick blindly to our own
fore-meaning about the thing if we want to understand the meaning of another” (Gadamer
2006, 271).
⁴⁰ Bāzgasht-i adabī was not a label used by poets of its time, but instead, like sabk-i hindī, was
popularized by Bahār in the early twentieth century (Smith 2009, 197; Schwartz 2014). The
structure of Ātashkadah is explained in Matini 2011. In a rich analysis of the work’s concept of
geography, Mana Kia warns against conflating awareness of place with nationalism (Kia 2014b,
90). The 1861 Bombay lithograph has been reprinted as Tażkirah-yi Ātashkadah 1998.
⁴¹ On Āżar and the earlier but in a sense like-minded tażkirah writer Muhammad : :: āhir
T
Nas: rābādī, see Alam 2003, 176. Āżar’s description of India notes simply that it is hot and large,
that “its customs and laws are often contrary to those of the people of Iran and indeed Turan”
[rusūm wa qawāʿid-i ānjā aks̄ar k: hilāf-i ahl-i īrān ast balkih tūrān], has bizarre fruits that are
not found in Iran, and lastly that complete information about it is not available in Iran but can
be constructed on the basis of books and manuscripts (Ātashkadah 1999, 417). No mention is
made of India’s centuries-long role supporting Persian literature or indeed welcoming Iranian
intellectuals. On the contrary, Āżar exoticizes it and interprets it at a distance, perhaps akin to
James Mill (who famously wrote in the preface to his history of India, “As soon as everything of
importance is expressed in writing, a man who is duly qualified may obtain more knowledge of
India in one year in his closet in England, than he could obtain during the course of longest life,
by the use of his eyes and ears in India,” quoted in Majeed 1992, 139).
⁴² Ātashkadah 1999, 423–38. The list of poets in the Bodleian Library catalogue’s entry on
Ātashkadah demonstrates just how stark the difference in numbers is (Sachau and Ethé 1889,
261ff., esp. 288). Dividing Indian and Iranian poets was not new, as this structure had been used in
tażkirahs such as Nas: rābādī’s (1083/1672, enlarged 1092/1681), but it takes on a new significance
under Āżar (since the Tażkirah-yi Nās: rābādī at least includes Iranian-born Mughal nobles in the
India section). Ārzū extensively uses Nas: rābādī as a source in his own MN.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

    107

from Delhi, and the only one whom he quotes at length is Amīr
:
Khusrau.⁴³ Furthermore, many poets deeply associated with India
whom Āżar included were given a non-Indian provenance—after all,
they had not been born in India—and poets who had long stints in India
and were respected in their time, such as Shāh Jahān’s poet laureate Abū
:: ālib Kalīm (d. 1061/1651), were condemned. Kalīm was prolific but
T
Āżar’s assessment is that he “has not a single verse worth remembering”
[shʿirī kih qabūl-i tażakkur bāshad nadārad] except in a few ġhazals.⁴⁴
He goes on to damn Kalīm with faint praise by noting that his poetry was
“accepted” [musallam] because of special favor from the Mughal
emperor and Indian nobles. (This implies that while people read his
work, it was not necessarily any good, objectively speaking.) As a further
example, let us consider the four poets to whom Munīr objects: In the
short entry on T :: ālib Āmulī, Āżar observes that T :: ālib “was for a time
in India among the most respected in the service of Shāh Salīm
[i.e., Jahāngīr]” and “in poetry [he] has a particular style which is not
sought by eloquent poets.”⁴⁵ Z :: uhūrī merits a one-line notice.⁴⁶ ʿUrfī he
acknowledges as a well-known poet whose dīwān he has come across
many times but dismisses him as having “a few verses which are not
empty of eloquence” [chand shʿirī kih k: hālī az fas:āhat : na-būd].⁴⁷ In
contrast, he heartily approves of Zulālī as a poet whose poetry is
“lucid” [roshan-zamīr].

Tellingly, the most obvious difference between
Zulālī and these other poets is that he never spent time in India.⁴⁸
Āżar’s generation of Isfahan-based poets have no connection to the
sub-continent.⁴⁹ By the time he was writing, a major change in the career

⁴³ The work’s twentieth-century publication history appears to recapitulate the problem:

Tehran-based Amir Kabir publishers released the first three volumes of the work, edited by
Hasan Saadat Nasiri, in 1957. The fourth and final volume edited by Mir Hashem Muhaddis,
which contains all of the entries for the poets from India and Turan, was not released until 1999.
⁴⁴ Ātashkadah 1999, 48.
⁴⁵ “wa muddatī dar hindūstān dar k: hidmat-i shāh salīm az muʿtabarīn būdah . . . wa dar
shāʿirī :t:arz-i k: hās:s: ī kih mat::lūb-i shuʿarā-yi fas: īh nīst dārad” (Ātashkadah 1957, 2:870–1).
⁴⁶ Ātashkadah 1957, 1:130. ⁴⁷ Ātashkadah 1999, 191.
⁴⁸ Ātashkadah 1957, 2:767; cf. Ghani 1929–30, 193, 298–9 on the observation that Āżar
harshly judges poets who made a living in India.
⁴⁹ They were the students of Sayyid Mīr-ʿAlī Mushtāq (d. 1757), a group including Āżar
himself, Hātif, Sabāhī, Bidgolī, and Sahbā
(Smith 2009, 200). Indeed, Sunil Sharma has argued
that Ātashkadah was concerned above all with preserving a Safavid legacy in the wake of the
collapse of the Safavid state (Sharma 2012b, 52).
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

108       

paths of Iranian litterateurs had set in, namely that India was no longer
an important rung in the advancement ladder for them. It had become a
strange place onto which critical judgments could be projected at a
distance. No one from their time into ours, notes the Urdu critic
Shamsur Rahman Faruqi, has ever written a study of “the theory and
practice of the language” used by Indo-Persian poets specifically.⁵⁰ The
:
Indian Persianist Muhammad ʿAbdul Ġhanī, writing at roughly the same
:
time as Muhammad Taqī Bahār, argues that there would be no need for
such a study because “the Indian style was essentially Persian, and was
founded on the same model as obtained in or what was brought from
Persia.”⁵¹ He notes that the idea that Indian usage was stylistically
inferior—or even markedly different—from that of the rest of the
Persian cosmopolis “now seems, perhaps more than it was a few years
before, to be making impression [sic] on European minds, for the
persons responsible for expressing such views have a far-reaching and
authoritative voice.”⁵² Yet despite the lack of any rigorous analysis, the
assumption remains on the part of many scholars in the West, in Iran,

⁵⁰ Faruqi 2004b, 61; cf. Alam 2003, 139n9. The difficulty is the slippage between the
categories Indo-Persian and so-called “Indian Style” by any name (see Dudney 2017b). For
:
example, Muhammad Taqī Bahār includes a list of characteristics of “sabk-i tāzah” in his
posthumously published article “Sāʾib

wa Shīwah-yi Ū” [Sāʾib

and His Style] (Bahar 1970).
These are the usual generalities about an obsession with newness in expression that made
eloquence all but impossible for the poets of this period (as well as decrying the ġhazal, the
iconic literary form of the period, as promoting facile thinking). He says of “sabk-i tāzah” that
after it was developed in Safavid Iran “it then created a breach [as in a metaphorical wall]
between Iran and India, and reached its zenith in India” [az īrān bah hindūstān nīz rak: hnah
kard wa dar hindūstān takmīl gardīd]. Thus he implies that the faults in Sāʾib’s
poetry are
somehow related to India without engaging at all with how Indo-Persian poetics might be
different than Iranian Persian poetics.
⁵¹ ʿAbdul Ghani 1930, 3:278.
⁵² ʿAbdul Ghani 1930, 3:278. This is particularly in reference to his teacher, E. G. Browne,
with whom he had a somewhat tense relationship. The preface to ʿAbdul Ghani’s later book,
Pre-Mughal Persian in Hindustan [1941], remarkably includes a facsimile of a 1923 hand-
written letter from Browne in which he writes “Professor ʿAbdu’l Ghanī appears to cling to what
I regard as the delusion that Indian Persian is better than Persian Persian—a matter of constant
conflict between us” (Ghani 1994, xxvii–xxix). ʿAbdul Ghani clarifies that he simply disagrees
with Browne’s assessment that Indo-Persian is automatically inferior to Persian produced in
Iran. Browne’s four-volume history of Persian literary culture was extraordinarily influential in
Europe and in the former Persian cosmopolis. It is telling that the Pakistani scholar Sayyid
ʿAbdullah cites as his authority the colonial researcher Henry Blochmann in his discussion of
“istiʿmāl-i hind” [Indian usage of Persian] (ʿAbdullah 1967, 267ff.; Blochmann 1868, 32ff.).
Blochmann’s analysis, although largely unobjectionable, is still problematic because it rests
mainly upon nineteenth-century sources (and for that matter nineteenth-century attitudes
about standard Persian) while claiming to provide a general history of Indo-Persian usage.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

    109

and in India that Indo-Persian went irreparably wrong when measured

against an Iranian standard of eloquence. A transhistorical Iranian
standard was reified after Ārzū’s lifetime by a group of poets apparently
seeking a break with the past by bracketing off Iranian literary culture
from the traditions of the Persian cosmopolis as a whole, although this
motivation may also be a later accretion to the critical tradition. For
Āżar, Sāʾib

Tabrīzī (1592–1676), who stayed in India during the reign of
Shāh Jahān, was one of the people who had utterly spoiled Persian
literature. Āżar writes that

From when he first tossed off verses, the ways of thinking established
by the eloquent among the ancients were closed off, the accepted
principles of bygone masters were lost as well as the niceties of poetry
after Sāʾib,

who was the inventor of an unpleasant new style [t::arīqah].⁵³

But even Āżar here frames the issue primarily as that of a degraded new
style that should be replaced by a universally accepted old style.⁵⁴ By
contrast, Ārzū’s own assessment of Sāʾib

is extremely positive, since Ārzū
sees him as the “leader” [sar-guroh] of the Moderns.⁵⁵ Views like Āżar’s
became dominant across both India and Iran in the nineteenth century
and were recapitulated throughout the twentieth: Iranian scholars (along
with most Indian Persian scholars) could find little commendable in the
tāzah-goʾī poets. However, after the Iranian Revolution poets previously
dismissed as decadent and unpleasantly difficult to interpret have started
to get their due in the Iranian academy.⁵⁶

⁵³ “az āġhāz-i suk: han-gustarī-yi īshān :t:uruq-i k: hayālāt-i matīnah-yi fus:ahā-yi :

mutaqaddimīn masdūd wa qawāʿid-i musallim-i ustādān-i sābq mafqūd wa marātib-i
suk: hanwarī baʿd az janāb-i mīrzā-yi mushārilaihi kih mubdiʿ-i :t:arīqah-yi jadīdah-yi
nāpasandīdah būd” (Ātashkadah 1957, 122–3).
⁵⁴ Even in the nineteenth century, the Iranian nationalist scholar Rizā :

Qulī Khān Hidāyat
(1800–72) condemns this period in the strongest terms (“the poets, following their sick natures
and distorted tastes, began to write confused, vain, and nonsensical poems”) in the preface to his
tażkirah Majmaʿ al-Fus:ahāʾ without specific reference to India (trans. Parsinejad 2003, 22;
Majmaʿ al-Fus:ahāʾ 1957).
⁵⁵ M 215.
⁵⁶ For example a special issue devoted to Bedil (no. 39–40, Winter 2007/Spring 2008) of
Qand-i Pārsī, the journal of Iran Culture House, New Delhi, contains a number of articles by
important Iranian scholars. Bedil’s complexity has been recast as philosophical depth rather
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

110       

In the Persian cosmopolis, the distinction between mutaqaddimīn and

mutaʾakh: khirīn
: was long-held and took on a special salience in the
seventeenth and eighteenth centuries. Five centuries before Ārzū, it
:
appears in Amīr Khusrau’s programmatic introduction to Ġhurrat al-
Kamāl [The Perfect New Moon, 1294], his third dīwān [book of collected
poems].⁵⁷ This categorical division remains into the eighteenth century,
albeit with the textual canon having been profoundly expanded. For
example, Ārzū’s Persian lexicon is structured around this division: The
first volume deals with words and expressions used by the Ancients while
the second, much slimmer volume deals with the Moderns’ usages.⁵⁸ The
Moderns may be further sub-divided into “muʿāsirīn”: [(living) contem-
poraries] and “mutaʾakh: :khirīn” [Moderns, living or dead], but there is
no significant distinction between contemporaries and Moderns as there
is between Ancients and Moderns.⁵⁹ Hazīn
wrote a tażkirah of roughly
100 contemporary poets—completing it in just nine days apparently—
entitled Tażkirat al-Muʿāsirīn [Tażkirah of [Our] Contemporaries,
1752].⁶⁰ Crucially, views on imitation were very different from those in
our society and so the Ancients had an important role to play in

than cultural degeneracy. The preface to the volume grandly declares “the present decade has
been the decade of Maulānā [the poet also called Rūmī] and the coming decade will without a
doubt be that of Bedil” [qarn-i hā
qarn-i maulānā būd wa qarn-i āyandah bī hīch gumān az
: zir
ān bedil ast] (Qazwah 2008, 8). It is worth observing that Bedil’s popularity, while it waned in
Iran and India, never eroded in Afghanistan, where Persian speakers are said to hold him in as
much esteem as Hāfiz
::.
⁵⁷ In this case the distinction is drawn between “mutaqaddamān” and “muʿās: irān”
(Dībāchah-yi Dīwān-i Ġhurrat al-Kamāl 1975, 38; Kinra 2008, 347). Also Niz::āmī ʿArūzī
writing
in the twelfth-century  advises all would-be poets to memorize 20,000 lines of the Ancients
and 10,000 of the Moderns (Zipoli 1993). In Arabic the division of Ancients from Moderns is
practically primordial. For example, Ibn Qutaybah (d. late ninth-century ) writes “I have not
regarded an ancient with veneration on account of his antiquity nor any modern with contempt
on account of his being modern” (trans. Nicholson 1907, 287).
⁵⁸ Naqvi 1962, 109–18. The second volume generally goes by the title Chirāġh-i Hidāyat. It
consists of words and expression used by the Moderns which do not appear in the major
dictionaries (Ārzū mentions Farhang-i Jāhāngīri, Farhang-i Surūrī, and Burhān-i Qāt::iʿ as
examples in the preface).
⁵⁹ A similar division of convenience has the “Middles” [mutawassit::īn], e.g., in Safīnah-yi
: wushgo;
Kh
:
see Husain 1937, 224 on Amīr Khusrau as the first of the “Middles.” The application
of such a tripartite division in Urdu, as in Mīr Hasan’s

Tażkirah-yi Shuʿarā-yi Urdū [Tażkirah of
Urdu Poets], will be addressed in Chapter 4.
⁶⁰ Storey 1953, 1,2:848. The claim of the improbably short writing time is Hazīn’s

own, but
he does not explain why he compiled it so quickly (Tażkirat al-Muʿās: irīn 1996, 228).
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

    111

contemporary poetic practice.⁶¹ While outright plagiarism [sariqat] was

generally condemned, thematic and formal imitation was institutional-
ized in the practice of writing poems in imitation of earlier masters
[istiqbāl, lit. “welcoming”] and quoting them in order to reply
[tazmīn].⁶²

To choose two examples at random, T :: ālib (sixteenth and
seventeenth century) acknowledges himself a follower of Amīr Khusrau:
(thirteenth century), while Abū al-Fazl

(sixteenth century) acknowledges
his debt to Abū al-Faraj Runī (eleventh century).⁶³ Thus, when Ārzū
takes the side of the Moderns in the debates of his time, he is not in any
way rejecting the Ancients but rather proposing a poetics that could cope
with newness. He is so respectful of the Ancients that, as we saw in
Chapter 2, he is willing to allow their mistakes in “Persianized” [tafrīs]
words borrowed from Indic languages to stand because these have in his
view become standard Persian. Moderns, however, must use recently
borrowed Indic words in Persian correctly.
For us, Persian poetry, especially in the early modern period, is
striking because of its rich and sometimes bewildering intertextuality—
poets frequently reference other poets and depend on them for sanad,
which we can formally translate as “warrant.” A sanad is an example of
usage in which a particular poet used a particular word, phrase, or

⁶¹ This is also true of pre-modern and early modern Europe, where it was a virtue to write in
the style of other people, for example Cicero in prose and Virgil in poetry (Bailey 1930, 205).
“The sixteenth-century theorists of poetry,” argue Nagel and Wood, “had another name for
pastiche: imitatio, or the transformation of text into text. The literary text of the Renaissance
was understood to be the altered double of a predecessor text. Acceptable doubling was literary
creation itself; unacceptable doubling—duplicitous doubling—was plagiarism” (Nagel and
Wood 2010, 297).
⁶² The different levels of plagiarism were theorized as part of ʿilm-i balāġhah (rhetoric). For a
detailed analysis of the tradition’s distinction between permitted borrowings and plagiarism, see
Losensky 1994 and Zipoli 1993. On the influential thirteenth-century critic Shams-i Qais’s views
on plagiarism, see Clinton 1989, 117–25. The Orientalist Francis Gladwin glosses tazmīn
as
“when the Poet applies to his purpose some lines from another author; but in case the author so
quoted be not well known, it is incumbent on him to mention the name, in order to obviate the
imputation of plagiarism” (Gladwin 1801, 33). The classic account of plagiarism in Arabic
theory is Grunebaum 1944 (although his assumption that imitation was equivalent to the
classical Greek concept of mimesis was later criticized).
⁶³ Hadi 1962, 113; Ghani 1930, 63. To this list we could add the nineteenth-century poet
Ġhālib, who acknowledges himself a follower of ʿUrfī, Sāʾib,
and Z:: uhūrī. Ġhālib writes, for
example, “ġhālib az aurāq-i mā naqsh-i :z:uhūrī damīd / surmah-yi hairat
: kashīm dīdah ba-dīdan
dahīm” [Ghalib! The color of Z :: uhūrī shines from our pages / Let us apply to collyrium of
wonder in our eyes and engage them in the act of looking] (quoted in Rahman 1970, 48 with
misprint of “taqsh” for “naqsh”).
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

112       

metaphor in a particular meaning, and like a royal warrant, it implies a

transaction of asking for and receiving approval from one’s betters.⁶⁴
This is central to the enterprise of Persian poetry and yet modern critics
have often failed to “allow intertextuality as a legitimate literary
device.”⁶⁵ Intertextuality—when a literary text cites or otherwise
responds to a previous literary text—is one way in which sanad is
embedded in the tradition. It has often been misinterpreted by later
critics as “mere imitation” rather than a crucial component of the system
of aesthetic control in Persian poetry. If we consider the painstakingly
researched critical works that Ārzū and other poet-scholars have written
then it becomes clear that the search for sanad was the ordering principle
for literary scholarship in Persian.⁶⁶ In Ārzū’s case, it raised questions
about the history of language (discussed in Chapter 2) that we would
recognize as his society’s equivalent of our modern discipline of linguis-
tics. For us, linguistics is an entirely separate realm of enquiry from
literary criticism: It is the study of “languages themselves for their own
sake.”⁶⁷ In the pre-modern Persian tradition, there was no such distinc-
tion between the study of language and the appraisal of literature. We

⁶⁴ It is worth noting that the term’s semantic range spans the secular and the religious. The
Arabic root S-N-D literally refers to “making something rest upon something else.” It refers to
the chain of transmission (usually in the synonym isnad) in scholarship used to establish the
:
authenticity of hadīs̄, but in Ottoman and Indo-Persian usage also refers to an officially sealed
(and therefore authenticated) document or proclamation (Bosworth 2011, 703).
⁶⁵ Faruqi 2004b, 22. Likewise, critics have generally failed to see the importance of humor
and mixed registers in Persian literature (Perry 2012, 90). It is axiomatic that whenever a
literature is elevated to the status of a classic, critics attempt to save the tradition from its
unsavory parts either by omitting them or trying to explain them away. In the Urdu tradition, a
good example is the poet Mīr, who is generally described as a serious and somewhat dour man,
but students of his (Persian prose) Żikr-i Mīr know that that text ends with several pages of
ribald jokes, many about the sexual deviance of Pathans. For many modern readers, it would be
inconceivable that the Mīr who wrote the serious poetry could be the same Mīr who collected
dirty jokes, but of course he was.
⁶⁶ It is useful to consider John Searle’s concept of the “constitutive rule,” namely a principle
without which a system ceases to be itself. For example, imagine a game of chess without the
Queen’s move—it would still be a game of strategy involving pieces moved around a board in
turns but it could not meaningfully be called “chess” (Taylor 1985, 34). Failing to recognize the
function of sanad (or rather dismissing it as the mark of degeneracy) similarly deforms modern
criticism of the Persian poetic tradition. A text by Ārzū’s friend Tek

Chand Bahār called Ibt::āl-i
:
Zarūrat [Refutation of Poetic License], which was lithographed in 1268/1851–2, argues that
good poetry should not allow changes made out of poetic necessity [zarūrat-i

shʿir] (Blochmann
1868, 29–30). In other words, he appears to argue (contrary to Ārzū, who is trying to account for
poetic innovation), that absolutely everything in poetry must rest on proper sanad.
⁶⁷ “les langues en elles-mêmes et pour elles-mêmes” (Auroux 1989, 30).
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

    113

must therefore be receptive to critical concerns that seem very different

from our own and avoid imposing anachronistic judgments on the
material.⁶⁸ Thus we should focus on what appears to have truly con-
cerned Ārzū, namely that new poetic styles presented novel issues of
interpretation. The old system of sanad was threatened by people’s
enjoyment and approval of verse that did not have an obvious precursor.
In Europe, imitation [imitatio] was also a guiding principle in litera-
ture until the Modern period.⁶⁹ The Ancients and Moderns debate in
Europe emerged from Renaissance questions about which literary and
bureaucratic models were to be imitated, the more recent scholastic
tradition or the rediscovered literature of Antiquity. In the words of
Ingrid Rowland,

The drive to purge Latin of its medieval vocabulary also marked a

rebellion against its development into a bureaucratic and technical
language. Medieval Latin was the Latin of laws, contracts and of
traditional university education. Medieval theology and philosophy
had acquired pinpoint precision, but precision could also sound excru-
ciatingly dull.⁷⁰

⁶⁸ Foucault’s eloquent summation of the state of knowledge in pre-modern Europe can be a

guide for the Persian tradition as well: “To us, it seems that sixteenth-century learning was made
up of an unstable mixture of rational knowledge, motions derived from magical practices, and a
whole cultural heritage whose power and authority had been vastly increased by the rediscovery
of Greek and Roman authors. Perceived thus, the learning of that period appears structurally
weak: a common ground where fidelity to the Ancients, a taste for the supernatural, and an
already awakened awareness of that sovereign rationality in which we recognize ourselves,
confronted one another in equal freedom” (1994, 32). Foucault also reminds us of the value of
seeing each Persian critical work as “a node within a network . . . caught up in a system of
reference to other books, other texts, other sentences” (Foucault 1972, 23).
⁶⁹ The Aristotelian formulation of mimesis (the imitation of nature) gave way to the
imitation of other authors. An interesting parallel is that the eighteenth-century Orientalist
Sir William Jones disagreed with Aristotle’s formulation and so apparently coincidentally held a
view on imitation of nature that was compatible with Perso-Arabic literary theory (Mukherjee
1968, 43). (On the pitfalls of applying Western notions of mimesis to Perso-Arabic literature, see
Meisami 2003, 4.) A good eighteenth-century expression from the Western perspective of the
tension between imitation as necessity and as a sign of cultural degradation is Winckelmann’s
Geschichte der Kunst des Altertums [History of the Art of Antiquity, 1764] on the Greeks (Stern
2003, 69).
⁷⁰ Rowland 1998, 14; cf. Auerbach 1993, 121. Likewise, Janet Coleman argues that the
transition between the Middle Ages and the Renaissance is best understood as a “genre shift”
(Coleman 1992, 573ff.).
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

114       

Humanists, particularly those associated with the Roman Academy,

sought to recover what they saw as the natural grace of classical Latin
and from the fifteenth century onward took deliberate steps to bend
literary usage back towards the Classics. Many of them were bureaucrats
in the Papal establishment and could change the Latin of the Church
from within. Famously “nuns” became “Vestal Virgins,” “churches”
became “temples,” and so on.⁷¹ This was not a simple substitution of
vocabulary but rather the smallest unit of an enormous cultural reorien-
tation that found its expression in literature, architecture, the arts and
political philosophy. From the beginning, Renaissance humanists had a
sense that they were separated from the Ancients not just in time but by
different modes of living, but this had not necessarily called into question
the desire to order society and culture on the basis of Antiquity.⁷² By the
seventeenth century, this devotion to the Ancients was itself critiqued.
People began to rebel against what Quentin Skinner has usefully called
“the mythology of doctrines,” namely that for any given subject an
ancient writer “will be found to enunciate some doctrine on each of the
topics regarded as constitutive of the subject.”⁷³

The Texts: Munīr’s Critique and Ārzū’s Responses

Returning to the primary sources, let us consider Munīr’s Kārnāmah.

Throughout the text, his mode of attack is expressing his strong opinions
about which metaphors are meaningful and which are nonsense. He
rejects metaphors not used by the Ancients, although Ārzū shows that in
some cases he has made the mistake of thinking there was no precedent
for a certain usage when there actually had been. When Ārzū comes to
the defense of the Moderns two generations after Munīr’s attack on
them, he does not reject the primacy of the Ancients but rather makes
space for the Moderns by developing a poetics that could assess their
works rather than dismissing them out of hand. Munīr’s haphazard
approach stands in contrast to Ārzū’s careful theorization of literary
criticism, and their personalities differed considerably. Munīr was

⁷¹ Rowland 1998, 199. ⁷² Auerbach 2003, 321. ⁷³ Skinner 2002, 59.

 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

    115

recognized as brilliant at an early age, had a brief and controversial

career, and then died young, aged just thirty-four.⁷⁴ Ārzū on the other
hand invokes his old age in the preface to his Dād-i Sukhan : [Justice in
Poetry, or A Poetic Gift, ca. 1746], noting that he had been a student of
poetry from the “whiteness” of childhood through the “blackness” of
youth into the “yellow sun” of old age.⁷⁵ Temperament perhaps goes
some way to explain why Munīr, the literary enfant terrible of his time,
vaguely states a program to defend the Ancients vigorously against the
Moderns, while Ārzū, the critical éminence grise of his own time, speaks
from a lifetime of poetic experience when he lays out his interest in both
the theory and practice of literature. Ārzū nonetheless has only compli-
ments for Munīr in his tażkirah Majmaʿ al-Nafāʾis. He writes that
Munīr’s “like in the art of poetry has not been seen in India since Faizī.

:
In the art of belles lettres he is a follower of the style of Amīr Khusrau ...
By any measure he is accepted by the proven authorities of the perfect
people of India and Iran.”⁷⁶ Such politeness is de rigeur for the tażkirah
genre, but critical works like Kārnāmah and Dād-i Suk: han had more
scope to be vicious.
The two works that Ārzū wrote in reply to Munīr are Sirāj-i Munīr
[The Shining Lamp, or The Lamp for Munīr] and Dād-i Suk: han. The
dating is uncertain but Dād-i Suk: han was probably written in 1746 and
Sirāj-i Munīr was written at some point before that with Tanbīh al-

⁷⁴ By his own estimate, Munīr had written 100,000 couplets. He adds, demonstrating that he
was not a particularly humble man, that “each and every couplet is elevated by the sparkle of
meaning” [har yak bait al-sharaf-i kaukabah-yi maʿnī ast] (Kārnāmah 1977, 27). Munīr took his
pen-name at fourteen, and entered the service of Saif Khān, : the governor of Agra and brother-
in-law of the empress Mumtāz Mahal, : at age 25 in 1045/1635. He was apparently well received
in Agra. On his patron’s death in 1049/1639, he went to Jaunpur before returning to Agra and
dying there in Rajab 1054/September 1644 (Memon 2011).
⁷⁵ DS 1.
⁷⁶ “mis̄l-i ūʾī bah fann-i shiʿr baʿd-i faizī
bah hindūstān bah ham narasīdah. dar fann-i inshā
muttabiʿ-i :t:arz-i amīr k: husrau ast . . . bah har hāl
: az musallam al-s̄ubūtān-i ahl-i kamāl-i hind
wa īrān ast” (MN 2005, 116–7).
Ārzū also notes the existence of the following works of Munīr: His Sāz-o Barg is a mas̄nawī
“in praise of the betel-leaf and specifically Indian items” [dar sitāyish-i barg-i tanbul wa ġhairah
wa ashyā-yi mak: hs: ūs: ah-yi hind] and he has also written a mas̄nawī “describing the flowers of
Bengal” [dar taʿrīf-i gulhā-yi bangālah] which must be Maz:: har-i Gul [A Show of Flowers], also
called Mas̄nawī dar Sifat-i

Bangālah [A mas̄nawī on the character of Bengal]. Sāz-o Barg was
published in Lucknow 1889 and since reprinted (Memon 2011).
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

116       

Ġhāfilīn falling between them, most likely in 1744.⁷⁷ It is worth sum-

marizing the contents of each because this will help demonstrate how
interconnected the Indo-Persian critical tradition is.
Sirāj-i Munīr is a direct response to Munīr’s Kārnāmah with which
I began this chapter.⁷⁸ In Kārnāmah, Munīr critiques four poets, ʿUrfī,
:: ālib, Zulālī, and Z
T :: uhūrī, who were his elder contemporaries and some
of the most respected poets of the day. Ārzū notes that what unites
Munīr’s four targets is that they are “Modern” [mutaʾakh : khir],
: clearly
setting up a conflict between old and new styles.⁷⁹ Munīr’s criticism is
sharp throughout but he claims in his preface that he is not an enemy
: sm]
[kha : of any of the poets.⁸⁰ In other words, he presents his remarks as
a reasoned critique rather than a personal attack. He analyzes some
fifteen verses by each poet. In Sirāj-i Munīr, Ārzū takes each of those
verses, quotes at least part of Munīr’s commentary, and then comments
on the commentary. The non-polemical commentary that Ārzū wrote on
the qas:īdahs of ʿUrfī before he wrote Sirāj-i Munīr is excluded from this
discussion.⁸¹
:
Dād-i Sukhan is based on a similar principle but with an added layer:
:
First, Muhammad Jān Qudsī (d. 1056/1646–7), a poet laureate under
Shāh Jahān, wrote a qas:īdah. Then his contemporary Mullā Shaidā
(d. 1042/1632) commented on some lines of which he disapproved.
This commentary was counter-commented by Munīr, and finally Ārzū
has added his contribution to the debate in Dād-i Suk: han while citing the

⁷⁷ Dād-i Suk: han must be the latest of the critical works, since it mentions both Sirāj-i Munīr
and Tanbīh al-Ġhafilīn (DS 6) while Sirāj-i Munīr mentions neither of the other texts.
A manuscript at Punjab University (Lahore) has a colophon stating that Dād-i Suk: han was
:
written in Muhammad Shāh’s thirty-eighth regnal year—however, we cannot accept this since
:
Muhammad Shāh’s rule only lasted thirty years (Akram 1974, xx). Ārzū mentions in Dād-i
Suk: han’s preface that he is now an old man, which puts the composition date not earlier than
the 1740s. The editor has concluded that Dād-i Suk: han was written in 1746/1159 on the basis
that the regnal year “38” was probably a scribal error for “28” (Akram 1974, xxi). The preface to
Tanbīh al-Ġhafilīn mentions that it was written approximately ten years after Hazīn
arrived in
India (Hazīn

writes in his autobiography that he arrived in Sindh at the beginning of Shawwal
1146 = March 1734). It was perhaps composed at the same time as Dād-i Suk: han because it
does, in fact, contain a single reference to Dād-i Suk: han (TĠh 31). Otherwise this must be an
interpolation.
⁷⁸ The context is explained in Alam 2003, 182ff. ⁷⁹ SM 33.
⁸⁰ Kārnāmah 1977, 7.
⁸¹ The commentary on ʿUrfī is unpublished and the three manuscripts in the British Library’s
Delhi Persian Collection (Delhi Persian mss. 1286A, 1286C, and 1286D) are all nineteenth-
century copies.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

    117

positions of his predecessors. Additionally, Dād-i Sukhan : begins with

three fascinating prefaces on critical theory that Ārzū himself claims are
unique in the tradition.⁸²
Tanbīh al-Ġhāfilīn, or in its full and gloriously Arabicized title
Risālah-yi Tanbīh al-Ġhāfilīn fī I ʿtirāzāt
ʿala̍ Ashʿār al-Hazīn

[lit. The
Essay of Admonition to the Heedless, in Objections to the Verses of
Hazīn],

is altogether simpler since it consists only of Ārzū’s objections to
some 300 verses of Hazīn.

It follows the same pattern as the two works
against Munīr in that it criticizes specific lines of Hazīn.
Ārzū also
apparently wrote a shorter second tract against Hazīn
called :
Ihqāq al-
Haqq

[Administering Justice, or Establishing the Truth], but no manu-
script of it exists and we only know about it because Imām Bakhsh :
Sahbāʾī

wrote a text in the nineteenth century in response to it. It was
Ārzū’s spat with Hazīn

(rather than with Munīr) that captured the
imagination of Indo-Persian scholars in the nineteenth century because
it fit so well into the narrative of Indians versus Iranians.

Dād-i Suk: han’s Prefaces

Ārzū’s prefaces to Dād-i Sukhan

: attempt to establish the limits of literary
interpretation in a way that the tradition never had before: Where does
poetic authority come from? is the implicit question. Ārzū writes that the
prefaces are based on his own ideas [gumān-i kh : wud]

and deploys a
standard formula for when an idea is untested: (literally) “May God make
it true!”⁸³
The first preface deals with mistakes in poetry and suggests a method
for deciding whether a strange usage by an otherwise qualified Persian
poet is in fact a mistake. Ārzū himself admits that this is a “very difficult”
:
[khailī mushkil] task. He accepts that usage shifts over time, but also that
native speakers inevitably make mistakes because making mistakes is an
integral part of language use.⁸⁴ Ārzū invokes vernacular poetic practice

⁸² DS 2–14. ⁸³ “khudā kunad kih wāqiʿī bāshad” (DS 2).

⁸⁴ This has, of course, become a crucial principle in modern linguistics, which recognizes that
speech errors occur even when the underlying linguistic concept is sound in a person’s mind.
Furthermore, it is worth pointing out that in this preface Ārzū invokes rek: htah, a vernacular
literary practice that would come to be called Urdu, and such a reference to something outside
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

118       

:
(“rekhtah”) in order to make his case, which appears to be a new
development in Indo-Persian intellectual history.⁸⁵ He proposes that
the solution for determining whether a mistake has been committed is
an assessment based on both the record of previous poets’ acceptance of
a usage and consensus among respected poets. Specifically, if a usage is
picked up by poets “whose standing is beyond reproach and accepted by
others” [kih pāyah-ash māfauq-i radd wa qabūl-i dīgarān ast] then it
becomes an accepted usage [dākhil-i: :
tasarruf].⁸⁶ Crucially, the Ancients
are by definition beyond reproach—being cited for centuries has proved
their worth—but the Moderns are not. It may seem obvious that an
expression becomes accepted if people accept it, but Ārzū is making a
more subtle claim: Just as the Ancients’ works became sanad, contem-
porary poets are also producing works that will potentially become
sanad. For example, a formula that appears in Tanbīh al-Ġhāfilīn is
that “the Shaikh: [i.e., Hazīn]

is the only sanad for” a particular usage.⁸⁷
Indeed, Ārzū implies with his careful refutations of various usages that he
is fearful that an ugly expression will become sanad.⁸⁸ He points out a
metrical fault in a Modern poet and then asserts that research [tahqīq]:

the Persian tradition implies that this is a universal formulation about poetry. A thorough
discussion of Ārzū’s ambition to make his literary theory universal and therefore applicable to
Urdu/“hindī”/“rek: htah” will be taken up in Chapter 4.
⁸⁵ Cf. TĠh, in which Ārzū mentions some rek: htah poets and notes that “a Mughal [in this
context, an Indian-born Muslim and therefore a native speaker of hindī] and a non-Mughal is
equivalent in the mistake in his own language” [muġhal wa ġhair-muġhal dar ġhalat:: bah zabān-i
k: hwud

musāwī ast] (TĠh 76).
There are some earlier claims of the vernacular on Persian such as the thirteenth-/fourteenth-
:
century poet Amīr Khusrau’s boast in the mas̄nawī Nuh Sipihr [The Nine Heavens] and the
preface to his third dīwān, Ġhurrat al-Kamāl [The Full Moon] that India is great because its
inhabitants can learn other languages, including Arabic and Persian, but outsiders can never
master Indian languages (see the discussion in Gabbay 2010). The nature of rek: htah as a mixed
form invites some comparison between the vernacular and Persian such as the Deccani poet
Nus: ratī Bījāpūrī’s (1600–74?) statement that “Some beauties of Hindi poetry cannot / Be
transported to Persian properly” (translated in Faruqi 2004b, 33). I have not, however, come
across any rigorous comparison between the two languages before Ārzū.
⁸⁶ DS 7. ⁸⁷ For example, TĠh 124.
⁸⁸ The critical literature on Arabic, from which Persian criticism took important inspiration,
is quite different because Arabic is the language of divine revelation in Islam and therefore
requires special considerations. For theological reasons, Arabic grammarians have throughout
history been concerned with the corruption of language [fasād al-luġhah] (Auroux 1989, 247).
For example, the Arabic literary theorist ʿAbd al-Qāhir Jurjānī (d. 471 /1078 ), who is
arguably the most important theorist on balāġhat in the tradition, frequently refers to the idea
that if someone changes the rules of a language then he can no longer be called a speaker of it—
clearly Ārzū has moved far from this view (Baalbaki 1983, 11–12).
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

    119

points to similar mistakes in the work of some of the Ancients but “to
name them would be a slight against propriety.”⁸⁹ Likewise, he cautions
that “to take up a mistake of one of the greats [buzurgān] is a great
: ::ā-yi buzurgī].” Poets therefore need to be vigilant lest
mistake [khat
future writers treat their slips of the pen as sanad, as has happened in a
few, thankfully rare, cases. After all, as they say in a different literary
tradition, even Homer nods.⁹⁰
The second preface deals explicitly with the question of whether (non-
native) Indian speakers of Persian are competent to make changes in
Persian usage to vocabulary and expressions. Indeed, he refers to the
usage of the poets who might be making such changes as “the idiom
:
[tasarruf] of the mighty lords of India who have brought poetry and
belles lettres [inshā] to the seat of perfection.”⁹¹ He appeals to history to
observe that Iranians had accepted Turkish and Arabic words and
constructions [tarākīb], and as these were pulled into Persian, their
meanings changed considerably from Turkish as it was spoken in
Central Asia [lit. Turan and Turkistan].⁹² He most directly addresses
tensions between Iranians and Indian Persianists when he writes that the
objection of Iranians against Indians’ idiomatic usage [rozmarrah] are
unfounded.⁹³ Mastery, however, is crucial: He notes that a certain

⁸⁹ “burdan-i nām-i īshān sūʾ-i adab ast” (DS 3). Similarly in MN, he points out a metrical
fault in a couplet by a contemporary poet called Girāmī (d. 1156/1743), calling it “apparently a
mistake” [ġhāliban sahw] and also sighing “but this particular mistake indeed appears in the
: baʿzī
meter of several of the masters” [lekin sahw-i mażkūr dar bahr-i
az asātiżah nīz wāqiʿ ast]
(MN 2005, 111). Shams-i Qais, apparently dealing with the same issue that not all sanad is good
sanad, counsels would-be poets to take care to only imitate good poets and not bad ones
(Clinton 1989, 116).
⁹⁰ The Roman poet Horace in his Ars Poetica notes his irritation at what we might call the
“continuity errors” in the Homeric epics with the line “I become annoyed whenever good
Homer nods” [indignor quandoque bonus dormitat Homerus] (359). As Indians became
increasingly self-conscious of their supposedly defective Persian, the scope of sanad accepted
from Indian poets was seen to narrow to the impossible standard of poets whose verse was
absolute perfection: The poet Ġhālib (1797–1869) famously argues that besides Amīr Khusrau,:
he himself is the only good Indian poet of Persian, because even “even master Faizī
[a poet
laureate under the emperor Akbar] fumbles once in a while” (translated in Faruqi 2004b, 5).
⁹¹ “tas:arruf-i s:āhib-i
: qudratān-i hind kih fann-i shiʿr wa inshā rā bah ʿarsh al-kamāl
rasānīdah-and” (DS 7).
⁹² He makes an identical case in Mus̄mir (M 38–9).
⁹³ For example, he writes of a particular expression that differs by a single word in India and
Iran that the Indian form is acceptable: “The criticism of the Persians is unfounded against our
idiom ‘what leaves from his purse [kīsah]’ as opposed to ‘[what leaves] from his purse [girih]’ ”
[iʿtirāz-i

fārsiyān bī-jā-st kih rozmarrah-yi mā az kīsah-yi ū chih mī rawad ast nah az girih-i ū]
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

120       

expression has become standard in Indo-Persian “through strength and

not because of weakness and error.”⁹⁴ Ārzū notes that Amīr Khusrau
: is
one such master [ustād] whose usage was innovative rather than mistaken.
He is recognized by all the great poets of Iran and Turan, and indeed, Ārzū
:
tells us to preclude any possibility that Khusrau is only appreciated by
Indians; his works are cited in important dictionaries by non-Indians
(namely Farhang-i Jahāngīrī “by one of the nobles of Shiraz” and
Surūrī’s Majmaʿ al-Furs “by a poet established in Kashan”).⁹⁵ Abū al-
Fazl’s

Akbarnāmah is likewise acclaimed by Iranian men of letters.
In the third preface, Ārzū divides his contemporaries into groups
according to their different methods of poetic interpretation. The fun-
damental distinction is between common readers [ʿawāmm] and con-
noisseurs [khawās: s],
: although a poem should ideally appeal to both even
if common readers will miss the subtleties of its meaning. However, the
usage of expert poets should not be subject to the whims of common
readers who lack the training to understand literary subtleties. Crucially,
simply being a native speaker of a language does not qualify one as an
expert because literary judgment requires particular training. Ārzū men-
tions a famous quotation from “one of the greats of India [addressed] to
one of the contemporary poets of Iran” (which is elsewhere attributed to
Abū al-Fazl
with ʿUrfī Shīrāzī as his interlocutor) that “We have learned
your language from your most eloquent [i.e., written works by classic
authors], and you have learned it from your old men [pīrzāl-hā].”⁹⁶ He

:
(DS 8). Amīr Khusrau is cited as sanad for this expression in M 38, although the context of the
quotation does not allow us to determine what the expression actually means. On the use of
“rozmarrah” as a technical term, see Chapter 4.
⁹⁴ “az rū-yi qudrat ast nah az rāh-i ʿajz wa dalīl” (MʿU 2002, 98). This is parallel to a crucial
distinction in Persian poetics between two kinds of error, ġhalat::-i ʿāmm and ġhalat::-i ʿawāmm,
where the former is a mistake sanctified by usage (“which all eloquent people use” [tamām
: badān takallum kunand], according to Ārzū) while the latter is a mistake born of simple
fus:ahā
ignorance (MʿU 2002, 98; cf. Blochmann 1868, 33).
⁹⁵ It is worth pointing out that both men in fact had Indian connections: Injū, the author of
Farhang-i Jahāngīrī, although originally from Shiraz, was resident in India when he compiled
his dictionary at the behest of the Mughal emperor. Surūrī, although he wrote and revised his
dictionary in Kashan, compiled the second edition after receiving a copy of Farhang-i Jahāngīrī
from India. Furthermore, there is some evidence, which is perhaps spurious, that he came to
India in Shāh Jahān’s reign and died there (Blochmann 1868, 16–17).
⁹⁶ DS 9, cf. M 33. For the nineteenth-century poet Ġhālib, there is an addendum, namely a
withering response from ʿUrfī, to the effect that of course the great poets of Iran, whose work
Indians study, learned from these very same old people (Faruqi 1998, 27). Faruqi notes that he
has not been able to trace this incident back before Ġhālib but in fact this must be the same—a
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

    121

contrasts “ahl-i zabān,” literally “people of the language” but here most
likely meaning what we would call native speakers, with “ahl-i tahqīq,”
: or
“people of research,” namely those who mine previous poets’ works for
precedent. Even more explicitly, he declares that native speakers [ahl-i
:
muhāwārah] and non-native speakers [ġhair-i īshān] are equivalent
[musāwī] because what matters is the ability to assess precedent.⁹⁷
Furthermore, Ārzū argues that the best kind of interpretation is a holistic
one that does not fall into the trap of following a single narrowly focused
interpretative style. He lists six or seven (depending on how one
counts) different sorts of people whose interpretations or compositions
are restrictive.⁹⁸ For example, one category is that of the “arbāb-i maʿānī”
(lit. “Lords of Meanings”) who focus on questions of semiotics to the
exclusion of other concerns. Another, the arbāb-i bayān (lit. Lords of
Discourse) focus only on simile [tashbīh] to the exclusion of all else
while the badīʿyān (“inventors”) concentrate on literary ornamentation,
including īhām. Ārzū has a particular distaste for the interpretative
practices of schoolmasters known as nāz:: ims in India [mullayān-i
maktabī . . . kih dar hind nāz::im k: hwānand]

whom he considers
charlatans.⁹⁹ On the other hand he considers interpretation “according
to the taste of the poets” [muwāfiq-i mażāq-i shuʿarā] to be a useful
catch-all for other considerations such as whether an expression from
everyday language [rozmarrah] is appropriate for poetry.
The crucial role of expertise in determining whether an expression
is eloquent Persian is echoed by Ārzū’s close friend Ānand Rām

version nearly identical with Ġhālib’s is quoted by DS’s editor from a manuscript of Ārzū’s
contemporary Mirzā Muhammad : ʿAlī Tamannā apparently quoting the Shāh Jahān-period
:: abāt::abāʾī (DS lxi).
critic Jalālā-yi T
⁹⁷ M 30.
⁹⁸ A thorough analysis (in Urdu) of the third preface is ʿAbdullah 1977, 142–7. See also
Keshavmurthy 2013, 35.
⁹⁹ This particular teacher says that in a couplet in which both lines end “yaʿnī chih?” [so
what?], a verb must be expressed rather than implied. The pedantry is obvious in English as well:
No right-thinking person would suggest that the phrase “so what?” is wrong and must instead
be “so what is it?” In this context, Ārzū scornfully brings up the question that began the debate
he is trying to settle in Dād-i Suk: han: the objection by Shaidā to Qudsī’s use in a couplet of the
word “tang” [tight, narrow] to refer to grief rather than to a space. (In the former case, he says, it
is a matter of quality [kaifīyat] of the grief while in the latter it is a matter of quantity
[kammīyat] of the space. He implies that only truly stupid people would not see the obvious
difference.)
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

122       

: s,
Mukhli :::ilāh: [The Mirror of Expressions,
: who writes in Mirʾāt al-Ist
1158/1745] that

: Kāshī has used the word “squirrel” [gilahrī] in his mas̄nawī

Yahya̍
satirizing gluttony even though it is a hindī word. As I have written
:
above, for masterful poets [khudāwandān-i istiʿdād] whatever they say
goes and it is a warrant [sanad] for beginners.¹⁰⁰

We can be certain that the ideas developed in this work have Ārzū’s
stamp of approval because the manuscript was given to Ārzū to correct
: death in 1164/1751. Ārzū writes,
: s’s
three months after Mukhli

Ask not of my dejection:

Now there is no enthusiasm in poetry for me.
My heart was Ānand Rām Mukhli : s: ;
After his death, I find no enjoyment.¹⁰¹

A remarkable parallel to Ārzū and Mukhli: s: ’s views on poetic mastery is

the entry for Bedil in Āzād Bilgrāmī’s tażkirah Khizānah-yi
: ʿĀmirah
[The Royal Treasury], which declares that as a master, Bedil has the
right to innovate.¹⁰² The passage cites as evidence Ārzū’s discussion of
Bedil in Majmaʿ al-Nafāʾis, in which he makes the same claim, as well as
:
Dād-i Sukhan.
Later critics have had some difficulty accepting the formulation that
the masters have free rein. For example, the last lines of Dād-i Suk: han are

¹⁰⁰ “yahya̍
: kāshī lafz::-i gilahrī rā dar mas̄nawī kih ba-hajw-i akūlī āwardah ast wa hāl : ānkih
:
hurūf-i hindī ast hamān harf-i : faqīr ast kih dar aurāq-i guzashtah niwishtah-am kih
ba-k: hudāwandān-i istiʿdād harchih ba-gūyand mīrasad wa īn barāʾī mutabaddiyān sanad ast”
(f. 252b in British Library ms Or 1813 (= Mirʾāt al-Is::t:ilāh: 1850); Mirʾāt al-Is::t:ilāh: 2013, 660; see
Rieu 1879–83, 3:997). The passage to which he refers is an observation [naql] that is missing in
the British Library ms but appears in Mirʾāt al-Is::t:ilāh: 2013. It says: “Let it be known that some
venerable people believe that a hindī word should not be used in a Persian verse because it
degrades the verse. Nevertheless, Hakīm
Hāżiq—a

master of language—has used the word
‘chūrī’ [bangle] in the verse quoted above. Likewise, Mullā T :: uġhrā has a large number of hindī
words in his verses. This means that these [restrictions] are for novices and beginners. It is
allowed for the master-poets [ahl-i qudrat] who are free from obligation.”
¹⁰¹ “afsurdagī-yi marā ma-pursīd / aknūn na-buwad ba-shiʿr shauq-am / dil būd ānand rām
muk: hlis: / az murdan-i ū na-mānd żauq-am” (Mirʾāt al-Is::t:ilāh: 1850, f. 10a).
:
¹⁰² Khizānah-yi ʿĀmirah 252–3.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

    123

paraphrased in the English introduction by the work’s editor Sayyid

Muhammad Akram as follows:

The Persian poets belonging to the countries other than Iran, who are
experts in language and rhetoric and have long experience in poetic
exercises, are qualified to amend or modify the meaning of words and
idioms and use indigenous idioms in case of poetic emergency.

A more literal translation would go like this:

Thus it is clear that if someone neither Iranian nor Turanian [that is, a
Central Asian] has followed excellence and has conversed with the
Masters of Idiom then his utterance is sanad, but to the degree that he
takes all the [necessary] pains, only as God has willed [i.e., only under
exceptional circumstances].¹⁰³

The phrase “poetic emergency” is an overstatement because it makes the

poet’s use of a non-Persian word seem like something that simply occurs
once and has no consequences, but in fact Ārzū observes that the usage
becomes sanad and is thus available for other poets to use. Akram’s
translation also shifts focus away from the question of poetic mastery,
which is a recurring theme in Ārzū’s works. Additionally, leaving out
“Turanians” is significant because Ārzū is not looking specifically to Iran
for guidance but rather across the Persian cosmopolis. Indeed, Akram
makes Ārzū sound more conflicted and cautious than he is: Ārzū asserts
that any expert, native speaker or not, can change cosmopolitan Persian
by introducing non-Persian words under exceptional circumstances. The
default position for us, with our concepts of mother tongue and national
language, is to grant native speakers primacy in linguistic and stylistic
judgments. We cannot assume this holds for the eighteenth century, and
Ārzū’s project is a case in point. The term “ahl-i zabān” [lit. people of the
language], commonly translated into English as “native speakers,” has

ʿ [sic, misprint for ‘wuzū

¹⁰³ “pas bah wuzū : paiwast kih ġhair īrānī wa tūrānī agar tatabbuʿ

h’]
:
bah kamāl dāshtah wa suhbat-i s:āhibān-i
: :
muhāwarah namūdah bāshad qaul-i ū sanad ast,
ammā bah īn martabah rasīdan mashaqqat tamām mī k: hwāhad
illā māshāʾallāh” (DS 64).
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

124       

not been fully explicated by scholars. It exists within a constellation of other

related, albeit somewhat less common, terms like “ahl-i muhāwarah”
: [lit.
people of the idiom]. How do we distinguish between these? Or should we
not in some cases? It is risky to assert that, for example, the idea “that
proper poetic usage was authoritatively guided by the poetic idiom of
Iranian masters was a largely dominant view” in India.¹⁰⁴ At least from
Ārzū’s perspective, we should not refer to “Iranian masters” but “masters of
the Persian language.”¹⁰⁵

The Commentary and Intertextuality

Commentaries [shurūh: sg. sharh] : were an important vehicle for intel-

lectual activity in early modern Persian. However, modern habits of
mind that fetishize originality and the Orientalist prejudice that finds a
“stagnant East” everywhere one looks have conditioned us to see
commentaries as a derivative form of scholarship inferior to the free-
standing treatise.¹⁰⁶ To dismiss commentaries is to overlook a great
deal of intellectual endeavor because the commentary form was every-
:
where, whether as independent works, marginalia [hāshiyah], or the
habit of invoking extracts to anchor a discussion. For example, the
sophisticated philological discourse in Mus̄mir is built upon extracts
from al-Muzhir and the many earlier works al-Muzhir commented
upon. Lexicographical works and tażkirahs frequently digested earlier
works, sometimes to add information but also to nuance or even disagree
with earlier interpretations. Some commentaries represented a particular
engagement with the past such as Āzād Bilgrāmī’s Shifāʾ al-ʿAlīl (Cure for
the Ailing, 1196/1782), a commentary in Arabic correcting 180 infelici-
tous rhetorical choices in the work of the Arabic Golden Age poet

¹⁰⁴ Kia 2011, 191.

¹⁰⁵ In Islamicate scholarly culture there is certainly scope for recognizing the achievements of
non-native speakers, namely that the greatest philologists in Arabic were mainly from Persian-
speaking backgrounds. The first major grammar of Arabic was written in the eighth-century 
by Sībawayh from Fars (Marogy 2010, xi). Ārzū was well aware of such transcultural connec-
tions that enabled both Arabic and Persian literary culture to develop in the early centuries after
the Islamic conquest of Iran.
¹⁰⁶ Recent scholarly interventions include Ahmed and Larkin 2013 and Van Lit 2017.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

    125

al-Mutanabbī (d. 354/965).¹⁰⁷ Such commentaries could serve to show-

case a scholar’s erudition (or to advance a position in a debate, as in the
commentaries discussed in this chapter) but others were important as
pedagogical tools. Persian manuscript libraries are stuffed with commen-
taries on works commonly used by students such as the Gulistān of Saʿdī,
a thirteenth-century didactic work that was read by nearly all Persian
learners across the early modern world.¹⁰⁸ It is likely that for early
modern Persian litterateurs, compiling a commentary was seen as a rite
of passage from student to independent poet as Ārzū’s first commentary
:
Khiyābān-i Gulistān seems to be. While commentaries could take the
form of a simple gloss to help students understand difficult words or
metaphorical constructions, in general they create intertextuality. This
intertextuality could work on several levels, between the source text
[matn] and the commentary [sharh] : but also drawing upon both refer-
ence works and literary works (for example, to compare a metaphor used
by one poet with another poet’s usage). This textual layering has led
some scholars to use the analogy of hypertext, in which a discourse is
built up by cross-references to other texts—the form of a commentary is
a reflection of the arguments its author is making.¹⁰⁹
While Ārzū’s critical prefaces in Dād-i Suk: han are self-evidently
interesting for the modern reader, it takes considerable patience to get
through the bulk of the texts, which in the typical manner for pre-
modern Persian commentaries are a series of discussions of individual
couplets. These analyses usually turn on whether a metaphor is properly
used or not. Most of the entries in both Dād-i Suk: han and Sirāj-i Munīr
are prosaic, basically taking the form “Munīr’s criticism: The metaphor
used in this line has no meaning. Ārzū’s response: In fact, it does have
meaning for the following reason . . .” (On a few occasions, Ārzū admits
that he agrees with Munīr’s critique of a line.) Munīr’s comments are
frequently pointed barbs. For example, one of ʿUrfī’s couplets that he
really dislikes has, according to him, “not a whiff of meaning and is in
need of perfume” [būʾī az maʿnī na-dārad wa shāyistah-yi :t:īb ast], while

¹⁰⁷ Toorawa 2008.

¹⁰⁸ On Mughal Gulistān commentaries, see Kia 2014a. Ārzū’s friend Tek

Chand Bahār wrote
Bahār-i Bostān, a commentary on Saʿdī’s other famous didactic work, Bostān.
¹⁰⁹ See the taxonomies provided in Van Lit 2017.
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

126       

elsewhere he sneeringly proposes that two bad couplets should be Zulālī’s

epitaph.¹¹⁰ Although the Ancients are in dialogue with Moderns through
an unbroken tradition of sanad, there can be disagreement over the
different standards to which they are held.¹¹¹ For example, in a case
where Munīr claims that ʿUrfī has misused a metaphor [istiʿārah], Ārzū
retorts that, in fact, that particular kind of metaphor is only used by the
Moderns so judging it according to the style of the Ancients [tarz::-i
qadīm] is impossible. This formulation crops up several times.¹¹² If one
accepts, as Ārzū necessarily does, that consensus of contemporary poets
is part of the production of sanad, then a survey of poetic interpretation
as practiced by them would be helpful, which is precisely what Ārzū has
provided in the third preface of Dād-i Sukhan. : In any case, it is the
commentary form that allows questions about asnād to be adjudicated.
Commentaries were, of course, the battleground for the debates
described in this chapter. Their engagement with the tradition was by
definition productive and contentious rather than mindless copying of
what had gone before.

Hazīn’s

Critique and Ārzū’s Response

The exact occasion that gave rise to Tanbīh al-Ġhāfilīn is not recorded.
The preface only provides the context that everyone, young and old, was
reciting Hazīn’s

poetry and Ārzū noticed that some of it was incorrect.
So, he put pen to paper to try to make sense of Hazīn’s

poetic missteps.
Ārzū does not lay out a program but instead praises Hazīn

(including as
“Seal of the Moderns” [khātam-i
: mutaʾa : :
kh k hirīn]—perhaps ironically
given Hazīn’s

preference for the Ancients), while slyly introducing the
idea that he is a bigot. When Ārzū notes that Hazīn

fled from Iran to

¹¹⁰ Kārnāmah 1977, 10, 19.

¹¹¹ In Mus̄mir, Ārzū implies that Indic words that had been borrowed into Persian but used
incorrectly by the Ancients were allowed to stand, but new borrowings had to follow the Indic
spelling. For example, Bengal is written in Persian as “bangālah” (i.e., ending with the letter he)
while people in Indic languages write and say “bangālā” (i.e., ending with alif ). Other more
recent interventions, where a Persian poet has mispronounced a borrowed Indic word, are
rejected as unsound (M 1991, 213).
¹¹² SM 47, cf. 43, 53, 66. Alam 2003, 183 analyzes the passage from SM 53.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

    127

India, he quotes a relevant quranic verse: “whoever enters it is safe” [man

:
dakhala-hu kāna āminan], a reference to the holy precincts of Mecca
(3:97). By changing the reference from the Islamic Holy of Holies to
India, Ārzū highlights Hazīn’s

apparent ungratefulness in coming
to India and then disrespecting the poetic achievements of its inhabit-
ants. In the body of the text, Ārzū seems to suggest that Hazīn,
being a
Shīʿah, would be horrified by the ritual pollution if a Hindu fuller [gāżur]
bleached his pīrāhan [loose shirt].¹¹³ Elsewhere Hazīn
is cited as claiming
that no Indian ever says his prayers five times a day (the implication
being that Indian Muslims were bad Muslims).¹¹⁴ As Mana Kia has
argued, the idea that Ārzū did not like Hazīn
simply because he was
not Indian is contradicted by Majmaʿ al-Nafāʾis, in which some entries
for Iranian émigrés are highly positive and some negative.¹¹⁵ Wālih,
however, does accuse Ārzū of targeting Hazīn
out of personal enmity
and a desire for revenge.¹¹⁶ Having addressed their strained personal
relationship in Chapter 1, I will take the literary criticisms at face value in
this section.
Ārzū himself selected the verses of Shāikh : Hazīn’s

that he wished to
critique, unlike in Dād-i Suk: han and Sirāj-i Munīr, in which his choices
were constrained by the verses that Munīr had already picked. Ārzū does
not summarize which kinds of poems to which he objects in Tanbīh al-
Ġhāfilīn but his contemporary Mīr Muhammad : :
Muhsin of Agra does.
:
Muhsin’s Mu :
hākamāt al-Shu ʿ arā [Judgments of the Poets] completed in
1180/1766, the year of Hazīn’s

death, lists three categories of objections
raised by Ārzū:

The first [applies] to an expression outside of usage, which is not found

in the verse of the teachers, each of whom is a mighty lord of the art of
speech, namely ʿAt:::t:ār, Rūmī, Sanāʾī, Khwājah-yi Kirmānī, Saʿdī, Amīr
:
Khusrau, Hasan

:
Dihlawī, Khaqānī,
::, ʿUrfī,
Anwarī, Kamāl Ismaʿīl, Hāfiz

¹¹³ TĠh 133.

¹¹⁴ Qasmi 2004, 76. The most frequently cited evidence that Hazīn
was not a bigot is a
chapter in his autobiography, in which he declares that he learned about other traditions (The
Life of Sheikh Mohammed Ali Hazin 1830, 62–4). Additionally, there are some verses (possibly
spurious) in which he praises the inhabitants of Benares, the place he finally settled in India.
¹¹⁵ Kia 2011, 196.
¹¹⁶ Kia 2011, 220. See Chapter 1 for more on their personal disputes.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

128       

Naz::īrī, Z
:: uhūrī, Hakīm

Shafāʾī, Sāʾib,

Qudsī, Kalīm, and Salīm. The
second [applies] to an unconnected expression that renders the couplet
meaningless and often comes into ʿAlī Hazīn’s

verse. The third con-
cerns idioms that appear in several of Hazīn’s

unsuccessful couplets
and that Ārzū, considering the appropriateness of the word (which is
the edifice of the workshop of speech), has rectified with another [i.e., a
new] hemistich and the verse shows a complete improvement through
these subtleties. Thus as God Almighty has commanded in the Holy
Qurʾān: “Above everyone who has knowledge there is the One who is
all knowing.”¹¹⁷

:
Muhsin’s analysis correctly identifies the sorts of objections that appear
throughout Ārzū’s text, but his list of poets (other than being an index of
important figures in the tradition) is apparently arbitrary, mixing
Ancients and Moderns. A fourth objection, which Muhsin : has not
mentioned, although it appears several times in Tanbīh al-Ġhāfilīn, is
that of plagiarism [sariqat].¹¹⁸ One couplet of Hazīn’s,

for example, is so
obviously the result of plagiarism that Ārzū declares it must be a copying
mistake.¹¹⁹ The not-so-subtle implication is that only a fool would make
such an error.
Ārzū’s overall strategy is to demonstrate Hazīn’s

failure to live up to
Hazīn’s

own apparently conservative standards. He writes “obviously
because Hazīn

is so devoted to the discourse of the Ancients, he takes
nothing at all from the Moderns.” Interestingly, despite Hazīn’s
alle-
giance to the Ancients, he too uses the rhetoric of freshness in his

¹¹⁷ “awwal bar ʿibārat-i ġhair mustaʿmal kih dar shiʿr-i ustādān-i s: āhib-i : qudrat-i fann-i
suk: han mānand farīd al-dīn ʿat:::t:ār, maulwī jalāl al-dīn rūmī, hakīm
: sanāʾī, k: hwājah-yi kirmānī,
shaik: h saʿdī, amīr k: husrau, hasan
: dihlawī, maulānā k: haqānī, anwarī, kamāl ismaʿīl, k: hwājah
: ::, maulānā ʿurfī, maulānā naz:: īrī, maulānā :z:uhūrī, hakīm
hāfiz : shafāʾī, mīrzā s: āʾib, qudsī, kalīm
wa salīm yāftah ni-mī shawad. duwwum bar ʿibārāt-i bī-rabt:: kih shiʿr rā bī maʿnī mī kunad wa
aks̄ar dar shiʿr-i ʿalī hazīn
: āmadah. siwwum az rāh-i tas:arrufāt kih baʿzī
mis:āriʿ-i shaik: h-i
nārasā wāqiʿ shudah wa ha : zrat

ārzū bah sabab-i munāsibat-i lafz::ī kih bināʾī kārk: hānah-i
suk: han [ast] ba-rāh-i rāst mis: raʿ-i dīgar rasānīdah wa shiʿr az īn daqāʾiq taraqqī-yi kull mī
nimāyad. liha̍żā haqq
: taʿāla̍ dar qurʾān-i sharīf irshād farmūdah: fauqa kulli żī ʿilmin ʿalīm
(12:76)” (quoted in Akram 1981, xlv–xlvi). Muhsin’s : text remains unpublished and the editor
has taken this from the Punjab University Library ms.
¹¹⁸ Akram 1981, xlvi. ¹¹⁹ TĠh 87; cf. 73, 125.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

    129

poetry.¹²⁰ In one of the couplets cited by Ārzū, Hazīn

refers to “my fresh
speech.” Ārzū refers to freshness many times and it becomes a sort of
slur, the implication being that the expression [ʿibārat] may be fresh but
also incorrect for one reason or another. Thus, the idea—which animates
:
Muhammad Taqī Bahār’s dim view of the “Indian style”—that the search
for “freshness” in poetry necessarily represents a rebellion against trad-
ition and the rules of composition cannot hold.
We know something of the reception history of the text from accounts
Ārzū gives of his friends. Qizilbāsh Khān,: known by his pen-name
Ummīd, was originally from Hamadan in Iran but had lived in the
Deccan for forty years.¹²¹ Ārzū describes him as a good friend who had
died three years before the composition of Majmaʿ al-Nafāʾis, which
would put the year around 1747. Ummīd was confronted by an Indian
who was “a follower of the poets of Iran” [muʿtaqid-i kalām-i shuʿarā-yi
īrān] who said that someone (namely Ārzū) had criticized Hazīn.

Ummīd
replied that “there is no doubt of the Shaikh’s
: literary acumen but at the
same time it is clear that what this person (namely Ārzū) has said is not to
be dismissed.”¹²² “Subhānallāh!”
: [Thank God!] Ārzū declares parenthet-
ically. He goes on to describe how when Tanbīh al-Ġhāfilīn became
widely copied, “no Mughal or non-Muslim Indian [hindū]” made any
sort of counterattack. Wālih included a large citation of Tanbīh al-
Ġhāfilīn in his tażkirah and sent it to Isfahan. Ārzū admits that the
inclusion of so much of his text was not a ringing endorsement of his
: s: ūs: iʿtirāz-i
views, “especially criticism by an Indian of an Iranian” [khu an


hindī bar īrānī], and yet he is bemused that Wālih went to the trouble.¹²³

Ārzū and the Persian Cosmopolis

Although he never lays it out in exactly these terms, Ārzū’s project in the
critical works considered in this chapter is maintaining the unity of

¹²⁰ Jane Mikkelson argues compellingly that “Hazīn’s

model of freshness appears to be
essentially conservative” (2017, 524).
¹²¹ MN 2004, 1:169–81.
¹²² “dar zabān-dānī shaik: h shubhah nīst ammā īn qadr ham yaqīn ast kih ānchih falānī—
yaʿnī faqīr-i Ārzū—guftah bāshad bī-chīzī na-k: hwāhad

būd.”
¹²³ MN 2004, 1:170.
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

130       

Persianate literary culture. In the second preface to Dād-i Sukhan,: he

appears to deal narrowly with the question of Iranian versus Indian
usage, but actually is making a larger point about the abiding intercon-
nectedness of the Persian-using world. This was not an oversight but a
deliberate choice. As a keen researcher he was aware, perhaps more
acutely than any of his predecessors, of regional differences within the
Persian cosmopolis. For example, in Mus̄mir, he quotes a learned joke:
An Indian emperor asked an Iranian on a lark [az rāh-i shokhī] : if it is
true that Iranians mix up the pronunciation of the letters “qāf”/q/ and
“ġhain”/ġh/. The Iranian’s supposed reply is the punchline: “No, that’s a
mistake [qalat::]!” (The humor is that in the indignant answer he mis-
pronounced “ġhalat::,” proving that Iranians confuse the two sounds—
indeed, in standard Iranian Persian today the two sounds are pro-
nounced identically.) Ārzū observes a number of other phonetic vari-
ations peculiar to Iran such as the loss of the majˈhūl vowels (that is, the
distinction between the sounds “ī” and “e,” and “ū” and “o”) and the shift
in pronunciation of the suffix “-ān” into “-ūn.”¹²⁴ In other places he
notes local lexical differences, such as objecting to one of Hazīn’s

usages,
namely “bīrūn raftan” [to go out] versus simply “raftan” [to go], and
noting that it is not used that way in India.¹²⁵ He concedes that perhaps it
is an idiomatic usage.¹²⁶ However, he believes that literary Persian is tied
to a courtly standard rather than to a locale (samples of ineloquent local
speech from Shiraz, Qazwin, Gilan, and Khurasan, for example, appear
in Mus̄mir).¹²⁷ He observes in Mus̄mir that

:
it is a fact that [bah-tahqīq paiwast kih] the most eloquent of languages
is the language [zabān] of the court [urdū] and the Persian of that place
is respected, but a dialect [zabān–i khās:ah] of other places is not
accepted in poetry or belles lettres [inshā]. The poets of every place
:
(for example, Khāqānī was from Shirvan, Niz::āmī from Ganjah, Sanāʾī

¹²⁴ M 76; TĠh 142–3. In the discussion in Mus̄mir (see Chapter 2 of this study), he implies
that this regional usage should never be used for poetry, but in TĠh he suggests that if relatively
recent poetic masters used such regionalisms (even though this usage would have been
“distasteful to the Ancients” [pīsh-i qudmā makrūh]) then they would be acceptable.
¹²⁵ TĠh 123. ¹²⁶ “shāyad muhāwarah-yi
: ahl-i zabān bāshad” (TĠh 123).
¹²⁷ M 5.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

    131

:
from Qazwin, and Khusrau from Delhi), all composed [literally
“spoke”] in the established [muqarrar] language and that was none
other than the language of the court.¹²⁸

Ārzū’s predecessors were not particularly concerned with defining

regionalism in such terms. Instead, Indians occasionally complained
that Iranian parvenus were taking jobs (for example, Munīr hints that
being Iranian makes it easier to be accepted as a Persian poet).¹²⁹ Munīr
was clearly concerned with style and not geography. But this does not
mean that there was no recognition of differences between Iranians and
Indians, in particular of their role in elite society. In the conclusion
:
[khātimah] to Kārnāmah, Munīr carps about the difficulties he faces in
becoming a respected poet. He mentions that the four social advantages
for a poet (none of which he possesses) are being old, rich, loud-voiced,
and having a “connection to Iran” [nisbat-i īrān].¹³⁰ Being Iranian is an
automatic advantage—he claims that an Iranian can make a hundred
mistakes in Persian and no one will say anything, but an Indian’s brilliant
work will be met with stony silence rather than praise. Munīr declares
himself an Indian, but then somewhat disparages India, saying that it
produced only five excellent poets (Masʿūd-i Saʿad-i Salmān, Abū al-
:
Faraj Rūnī, Amīr Khusrau, Hasan

Dihlawī, and Faizī).¹³¹

Munīr’s obser-
vations on the status of Indian poets, like Ārzū’s need to settle the
question of whether Indians can change Persian usage, must have been
based on some actual social pressure, such as favoritism shown towards
Iranians. The idea that talent-less Iranians were being rewarded was
stated even more openly by Shaidā, who was born in Fatehpur (in today’s
Uttar Pradesh between Kanpur and Allahabad). He is quoted by
: wushgo
Kh
as saying:

¹²⁸ M 13.
¹²⁹ Muzaffar Alam argues that “Mughal India thus virtually emerged as a kind of Iranian
colony” as compared to the considerably fewer numbers of Iranians resident in South Asia
during the Delhi Sultanate period (Alam 2004, 147). However, recognizing both the strong
influence of Persian and the vast number of Iranian migrants to India in this period should not
cause us to lose sight of the fact that even then the vast majority of Persian writers in India were
Indian-born because even local officials were using Persian to some extent (Alam and
Subrahmanyam 2004; Alam 1998, 328).
¹³⁰ Kārnāmah 1977, 25–6. ¹³¹ Kārnāmah 1977, 27.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

132       

Because of my being Indian, Iranians do not respect me . . . The fact is

that being Iranian or Indian is not a reason to boast. A man’s standing
is connected to his personal standing and if Iranians declare “the
Persian language is ours” then their language/tongue does not bring
them success, and if the tongue does find success, it is not in knowing
taste because they do not have poetic talent even though they flail for
it.¹³²

No doubt a long list of other such complaints could be compiled, but the
remarkable circulation of scholars, administrators, and poets in the
triangle formed by India, Central Asia, and Iran must have had a diluting
effect on regionalism over the centuries.¹³³ Movement, both of people
and of texts, ensured that writers kept abreast of developments in other
parts of the Persian cosmopolis. Ārzū has no qualms about passing
judgments on Iranians, such as calling a poet called Tanhā “well-spoken”
:
[khūb-go], which is admittedly a term of art (meaning “he composes
poetry well”) but still maintains its literal force, in which case Ārzū, an
Indian, is implicitly claiming the prerogative to determine whether or
not Tanhā, an Iranian, speaks Persian well.¹³⁴ After all, the right to
appraise native speakers is one he arrogates to himself—and all well-
:
versed poets—in the prefaces to Dād-i Sukhan. In his opinion, the loss of
the majˈhūl vowels in Iran is a unique handicap to following metrical
rules because of the difficulty of constructing correct rhymes.¹³⁵ Ārzū
also apologizes for the fact that Indians have not fully learned the new
kind of Iranian Persian but then turns the apology on its head. He writes
that “We Indian people have not taken up this really new Persian; we
bear the load of the insufficiency of our understanding, but according to

¹³² Safīnah quoted in Akram 1974, xxxiv.

¹³³ See, for example, Ghani 1930 for a side-by-side comparison of Indo-Persian and Iranian
correspondence that shows a remarkable similarity. As far as the mobility of people and texts
during the early modern period, the family of the tażkirah writer Muhammad : :: āhir Nas: rābādī
T
(born ca. 1025/1616) is an excellent example. His grandfather and uncle had settled in India, but he
himself had been born near Isfahan (Sprenger 1854, 88). A recent volume on circulation in South
Asia has the following excellent formulation: “ ‘Circulation’ therefore, in this volume, is also meant
as a kind of shorthand for the capacity of Indian society over the centuries to generate change”
(Markovits et al. 2003, 11). It should be noted that among Indians, Ottomans, and Safavids, the
Ottomans are the outlier in that they traveled outside their respective empire less than people from
the other two empires and received fewer outside visitors (Robinson 1997, 164).
¹³⁴ MN 2005, 66. ¹³⁵ TĠh 142–3.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

    133

the idiom-knowers, this kind of discourse [i.e., Hazīn’s

expression] does
not spring from eloquence.”¹³⁶
Despite Ārzū’s confidence in such transnational judgments, the
Persian cosmopolis fractured irrevocably soon after his death.¹³⁷
Stefano Pellò has argued that Persian ceased to be viable as a cosmopol-
itan language when it became the national language of an emerging
Iranian state.¹³⁸ In the literary sphere, the groundwork was laid by the
regionalism inaugurated in Āżar’s generation, and by the nineteenth
century it necessarily led to the preference of a national usage that
excluded cosmopolitan possibilities.¹³⁹ It remains to be studied whether,
as some have claimed retrospectively, an unbridgeable divide had opened
up between Indian and Iranian usages. One twentieth-century scholar
who has made a case for this, Momin Mohiuddin, tries to lay the blame
on “tasarrufāt,”
: which he glosses as “changes in spelling, form, meaning,
and construction” that inevitably come about because (he claims)
Indians were bad Persianists and thus the “purity of the idiom” was
lost.¹⁴⁰ For example, he contends that there were no proper Persian
grammars available, that Indians were bad Arabists and therefore bad
Persianists, and that while introducing “Hindawi words into pure
Persian was considered unpleasant,” they did it anyway. While for
Mohiuddin tas: arrufāt represent a cultural failure, for Ārzū they are

¹³⁶ “mā mardum-i hindūstān fārsī rā k: hūb nau bar na-kardah-īm, bar qus:ūr-i fahm-i k: hwīsh


:
haml :
mī kunīm pīsh-i muhāwarah-dān īn qism-i kalām az bulaġhā s:ādar na-shawad” (TĠh 122).
¹³⁷ Though prior to this moment, even in the seventeenth century, the Persian cosmopolis
was somewhat fragmentary, no doubt because of its sheer size. For example, the geographer and
:
economist avant la lettre Muhammad Mufīd Mustaufī writes in 1680 in his geography of Iran
prepared in India that he does not really know anything about the rulers of Iran, the Safavids
(Fragner 1999, 100–1). He has no excuse since they had been in power for nearly two centuries!
Āżar himself admits that people in Iran know the poetry of the Delhiites he has deigned to
include in Ātashkadah (1999, 523). This can serve as an index of how aware Iranians were of
Indian-born poets at this time.
¹³⁸ The politics of the Persian language in early modern Iran were more complicated than
most people realize: The Safavid state was shot through with Turkish. The origins of the dynasty
were in (Azeri Turkish-speaking) Azerbaijan, and Turkish remained an important language
even after Shāh ʿAbbas I moved the capital to Isfahan. Recent research (Floor and Javadi 2013)
has demonstrated that it was widely spoken at court, and that literary translations were made
not only from Turkish to Persian but from Persian to Turkish.
¹³⁹ On language planning in nineteenth-century Iran and the debate over whether it was
desirable to cleanse Persian of Arabic, Turkish, and Mongolian accretions, see Parsinejad 2003.
¹⁴⁰ Mohiuddin 1960; cf. Fragner 1999, 100–1. A more critical definition of “tas:arruf” appears
in the discussion of Mus̄mir in Chapter 2.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

134       

inevitable because languages change over time, even at the hands of

educated native speakers. The end of the Persian cosmopolis was not a
matter of mutual incomprehensibility—at least this has never been
satisfactorily demonstrated as a cause—but of the assertion of national
identities whose claims of cultural self-sufficiency overpowered the
claims of the cosmopolis.
The key to Ārzū’s claim for a universal Persian literary culture is that it
does not depend on pretending that India is not different (culturally or
politically) from Iran or elsewhere. Rather it assumes that for poetry, the
difference does not matter: The Persian cosmopolis is capacious enough
to contain the differences and everyone who uses Persian has access to
the same literary tradition, regardless of variation in local dialects and so
on. As we have seen in the discussion of Mus̄mir’s account of the
historical development of Persian in Chapter 2, Ārzū believes in a stand-
ard Persian that is immanent across the whole of the cosmopolis. He
notes that Persians used Arabic and Turkish for centuries, and changed
both languages in the process, which is unobjectionable. Ārzū argues
explicitly—for perhaps the first time in a work of literary criticism—that
Indians are allowed to modify Persian usage, provided they are properly
trained. His lexicon Chirāġh-i Hidāyat is designed in part to serve as an
extended justification of Indians’ correct usage. He states that he has
included two kinds of lexemes:

The first part consists of words that have a difficult meaning and which
Indian people often do not know; the second are vocabules whose
meaning is often widely known but there is debate as to the correctness
of having them in the speech [rozmarrah] of the eloquent of the people
of the language.¹⁴¹

He continues with an even more explicit description of how his diction-

ary is an intervention:

¹⁴¹ “qism-i awwal alfāz::ī-ast kih maʿānī-yi ān mushkil būd wa aks̄ar ahl-i hind bar ān it:::t:ilāʿ
na-dāshtand qism-i duwwum luġhātī kih maʿānī-ān agarchah maʿrūf wa maʿlūm būd lekin dar
: h: būdan-i ān az rozmarrah-yi fus:ahā-yi
s: ahī : ahl-i zabān baʿzī

rā taraddud ba-ham-rasīdah”
(Cambridge University ms Add. 795, 1a; checked with British Library ms Or 2013, Or 264, and
I.O. Islamic 71).
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

    135

because some of the Persian-speakers of India use loanwords/calques

[tas:arruf] in Persian on the basis of influence from hindī, it is necessary
to research a number of words. Thus this text is useful for the Persian-
speakers of India who use old words differently from the language-
knowers of Iran and Turan that are usually equivalent for experts in the
language and other people, but in this text there are many words that
have been researched and these people will be unable to answer
because they [the meanings of the words] are connected at length
with external research.¹⁴²

Thus research is able to demonstrate that deliberate Indian interventions

in Persian are in fact deliberate and not simply the product of a poor
understanding of Persian.
Ārzū’s compromise to keep India an integral but unique part of the
Persian cosmopolis is probably best expressed by the philosopher
Kwame Anthony Appiah. He writes, of our own time, that “the position
worth defending might be called (in both senses) a partial cosmopolit-
anism.”¹⁴³ Contrary to the assertions of some modern South Asian
scholars that Ārzū was a great patriot,¹⁴⁴ he gives little indication of
defending India per se but rather concerns himself with the entirety of
the Persian cosmopolis by defining the relationship of the part to the
whole. The domain of poetry itself requires justice to be done, as
evidenced by the very title of Dād-i Suk: han (Justice in Poetry) or by
Munīr’s earlier claim that he was a “mirror-bearer of justice.”¹⁴⁵ Poetry
can fit into a system of universal values, perhaps akin to Giambattista
Vico’s early eighteenth-century understanding of Ancient Roman law as
a kind of poetry.¹⁴⁶

¹⁴² “chūn bark: hī az fārsī goyān-i hind rā tas: arruf gūnah dar zabān-i fārsī ba-sabab-i ik: htilāt::-i
zabān hindī dast dādah āwardan baʿzī
az alfāz:: każā-ī bar s:āhib-i
: :
tahqīq zurūrī-st

pas in nusk: hah
mufīd ast bar fārsī goyān-i hind kih bah zabāndānān-i īrān wa tūrān rā ba-k: hilāf luġhāt-i
qadīmah kih dar aks̄ar an ba-zabān-dān wa ġhair zabān-dān musāwī-ast balkih dar īn nusk: hah
baʿzī

az alfāz::ī-st kah az chand kas zabān-dān tahqīq : kardah shud wa ānhā dar jawāb ʿājiz
shudand āk: hir kār az jāʾī dīgar ba-tahqīq
: paiwast” (Cambridge University ms Add. 795, 1a).
¹⁴³ Appiah 2006, xvii.
¹⁴⁴ E.g., Zaidi 2004, which argues that Ārzū could not bear to have people speak ill of India
and so wrote out of love for his country.
¹⁴⁵ Kārnāmah 1977, 6. Indeed Ārzū also expresses the hope that his own book will be justly
received (DS 2).
¹⁴⁶ Stern 2003, 79.
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

136       

The Persian cosmopolis as a geographical concept does not map easily

onto our concrete sense of geography. It is best understood not as a place
so much as a set of practices. The formula “Iran, Turan, and Hindustan”
is the most common designation (though Turan, corresponding roughly
to our “Central Asia,” is usually not well defined, and Hindustan typically
did not include southern India), but this should not be read like a list of
modern nation-states. Of course, people were conscious of being from a
place, but this must be carefully distinguished from identities available to
people in our time (including, for example, different ways to be “of” a
place). A cosmopolis should not be understood as homogenous, since
everyone has a sense of a belonging to a smaller locality (whether that be
nation, region, city, or something else) with particular local customs.
Even though the set of cosmopolitan texts may be the same, the inter-
pretation of them may be different in different places because of localized
commentarial traditions, which Sheldon Pollock has called “vernacular
mediations.”¹⁴⁷ Furthermore, the intellectual and economic topography
of a cosmopolis allows culture to flow more strongly in certain direc-
tions.¹⁴⁸ It is made up of many horizons, to use Gadamer’s term for what
a historical subject is able to comprehend around him or her, and there is
no expectation that anyone’s horizon take in the whole cosmopolis or
indeed that anyone even know its full extent.¹⁴⁹ There is also certainly
scope for local cosmopolitanism, which appears paradoxical on its face
but is actually quite reasonable, namely that the cosmopolitan idiom can
be employed in a text meant for a local audience.¹⁵⁰ More study is
warranted in the case of Indo-Persian texts, but Ārzū’s critical works
probably had a broad audience in mind since he is typically careful to

¹⁴⁷ Pollock 2009, 954. ¹⁴⁸ Cf. Brennan 2006, 3.

¹⁴⁹ “A person who has an [sic] horizon knows the relative significance of everything within
this horizon, whether it is near or far, great or small” (Gadamer 2006, 301–2).
¹⁵⁰ Yigal Bronner and David Shulman take issue with what they see as Pollock’s willingness
to flatten regional differences in the Sanskrit cosmopolis since some texts were written in the
cosmopolitan language but not meant to circulate. They write that “Every Sanskrit poem is, of
course, local or regional in that it was composed in a particular place by a poet speaking some
vernacular as his or her mother tongue (and writing in some local script). This is not, however,
sufficient to qualify a text as regional in our terms . . . First and foremost, a regional Sanskrit
work aims at a local audience. It is not meant to travel the length and breadth of the cosmopolis,
nor did it do so” (Bronner and Shulman 2006, 6).
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

    137

gloss Indic words and explain Indic cultural practices that Indians would
have been familiar with.
This brings us back to an interpretative problem. Can we distinguish
between tropes and actual analysis of the local? Scholars have often
drawn conclusions about various poets’ opinions of India on the basis
of poetic tropes. For example, India and its inhabitants are often
described as black. This is precisely the imagery Munīr uses (the five
Indian poets he singles out as great are blazing suns against the black
night of India). Obviously this is a literary commonplace rather than a
rigorous observation—elsewhere India is frequently described as para-
dise [e.g., as “hindūstān jannat nishān”], which is similarly a trope.¹⁵¹ In
an oft quoted couplet, Salīm Tihrānī (d. 1674) declares that

The means of acquiring perfection do not exist in Iran:

Henna has no color until it arrives in India.¹⁵²

In a similar vein but apparently with the opposite sentiment, T:: ālib Amulī
writes “The parrots of India are in fact mortified to see such a nightingale
coming forth from the garden of Iran.”¹⁵³ While such quotations have
often been interpreted as definitive historical evidence, this is an error
stemming from our modern incapacity to engage with a literary tradition
that depends on tropes and a different conception of history from our
own.¹⁵⁴ In the absence of other data, historians have often tried to read

¹⁵¹ Such rhymed epithets, reflective of the sajaʿ tradition of rhymed prose composition,
should not be taken literally. Ārzū for example refers to Qazwin as “jannat-āʾīn” [the heaven-
like] (MN 2005, 70).
¹⁵² “nīst dar īrān zamīn samān-i tah: s: īl-i kamāl / tā nayāyad sū-yi hindūstān hinā
: rangīn na-
shud” (quoted in Khatoon 2004a, 86 and elsewhere; discussed in Alam 2003, 160n99). More
examples at Dadvar 1999, 210.
¹⁵³ Translated in Hadi 1962, 103.
¹⁵⁴ To take another example: The poet Mīr ʿAlī Sher Nawāʾī (1441–1501) of Herat, who later
joined the first Mughal emperor Bābur’s entourage, was an important poet in both Chagatay
Turkish and Persian. Bābur acknowledges that “no one composed so much or so well in the
Turkish language as he did” but as far as his Persian, Bābur observes “some of his lines are not
bad, but most are flat and of low quality” (translated in Thackston, Bāburnāmah 1993, 2:354–5).
:
Nawāʾī’s Muhākamat al-Luġhatain [Judgment of the Two Languages, 1499] is a famous work
defending Turkish against Persian. It also mentions Arabic and hindī as two possible contenders
to Persian but dismisses Arabic as too lofty and hindī as too base. He uses the story of Noah’s
curse on his son Ham to explain Indians’ dark skin and ineloquence. (This was, of course, also a
common explanation among early modern Europeans for differences among human
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

138       

poetry in a straightforwardly biographical or sociological way, while

failing to engage with the complexities of the rhetorical system of
symbolism and form that the poets were actually drawing upon.
A more complex case is the drawn-out argument for India’s greatness
:
in the works of Amīr Khusrau (whose patriotism is even more overde-
termined in the historiography than Ārzū’s). Khusrau
: casts India in the
mold of a heaven-blessed, legitimately Islamic land, and makes an early
argument for Indo-Persian cosmopolitanism. He claims in the preface to
his dīwān Ġhurrat al-Kamāl [The Prime of Perfection, 1293/4] that while
inhabitants of other countries can only master their native tongues,
Indians can pick up any language.¹⁵⁵ They are therefore unquestionably
good at Persian (as well as Arabic and hindī). In another work, Iʿjāz-i
:
Khusrawī, he writes that just as desert Arabs used to drinking brackish
water cannot appreciate rose water, the writers of Khurasan and
Transoxania cannot appreciate good writing such as that produced in
India where the “old mode” [qarār-i qadīm] of writing persists as well as
a delightful new style.¹⁵⁶ This appears to be a serious argument about
literary stylistics imposed upon geography, and yet it is rendered suspect
by the political context in which it was written: The King’s realm must be
shown to be the most civilized on earth.
There is a marked difference between these pre-modern constructions
of Indianness in Persian literature and nineteenth-century ones. The first
comprehensive study of the differences between Indian Persian and the
Persian of Iran seems to have appeared at the turn of the nineteenth
:
century. Muhammad Hasan

Qatīl (1759–c. 1820), a Delhi-born Khattrī

populations.) Nawāʾī writes that “There is none among them whose skin is not black as the black
of ink and whose speech does not resemble the scratching of a broken pen.” He goes on to
describe Indian writing—presumably he means one of the Brahmi-derived scripts like
Devanagari—as like “the footprint of a raven” and incomprehensible except to Indians (trans-
lated in Devereux, 1966, 24–5). The connection drawn here between dark skin and ineloquence
is interesting but tells us little more than that Nawāʾī was somewhat aware of Indian culture (as
we would expect of a Central Asian at this time) but considered it barbaric. It might be a stretch
to connect his moralizing about dark skin with the common trope of the “blackness of the
Hindu” in Persian poetry.
¹⁵⁵ Analyzed at Gabbay 2010, 31. A similar claim appears in Nuh Sipihr [The Nine Heavens],
whose classification of Indian languages has been addressed in Chapter 2 (1950, 172ff.).
¹⁵⁶ Iʿjāz-i Khusrawī
: 1876, 66; in English translation see Iʿjāz-i Khusrawī
: 2007, 35.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

    139

who converted to Shīʿah Islam, wrote three works that dealt with
Indian Persian usage (as compared to Iranian and Turanian usage) and
his views were influential.¹⁵⁷ Although Qatīl was a careful scholar, he was
a prescriptivist who argued that Indians—like himself it should be
noted—had no real claim over Persian usage and must submit to the
judgments of Iranians and Turanians. Views along these lines became
increasingly dominant in the nineteenth century. Munīr’s strategy of
self-aggrandizement by suggesting that few Indians are good Persian
poets is of an entirely different order from that pursued by Mirzā
Ġhālib (1797–1869), who wrote in both Persian and Urdu two centuries
later. For Ġhālib, the only good Indian-born Persian poet—besides
:
himself of course—is Amīr Khusrau.¹⁵⁸ Indeed, Ġhālib went so far as
to apparently invent a native Iranian tutor, ʿAbd al-Samad,

for himself
(there is no independent confirmation of the man’s existence—he was
supposedly a Zoroastrian convert to Islam who stayed with Ġhālib’s
family in Agra). Thus, Ġhālib is entirely dismissive of Indians’ achieve-
ments in Persian literature whereas Munīr offers praise for his homeland,
albeit in a rather backhanded way. Ġhālib was not alone in the nine-
:
teenth century in his disdain for all things Indo-Persian.¹⁵⁹ Imām Bakhsh
Sahbāʾī’s

Qaul-i Fai :
s al [The Definitive Word], the last major text in the
Indo-Persian tradition to engage with Ārzū’s Tanbīh al-Ġhāfilīn, states
that Indians simply took good Persian style from Iran and contributed
nothing in return. His argument, fascinatingly, rests in part on the idea
that Iranian immigrants to India cannot learn Urdu properly and thus
Indians are similarly unable to master Persian.¹⁶⁰ (In fact, in Majmaʿ al-
Nafāʾis Ārzū praises a number of Iranian immigrants, notably Ummīd
:
and Zamīr, for their facility in the Indian vernacular—Ummīd appar-
ently understood subtleties in hindī poetry that even a native Indian

¹⁵⁷ The definitive study of Qatīl’s linguistic observations is Pellò 2016. Pellò convincingly
explains Qatīl’s views as the result of a double “conversion” in which Qatīl’s turn to Islam was
accompanied with a turn to Iranian aesthetic/cultural purism. The Orientalist Henry
Blochmann drew heavily upon Qatīl’s data later in the century.
¹⁵⁸ Faruqi 2004b, 5. ¹⁵⁹ The classic analysis is Faruqi 1998.
¹⁶⁰ Risālah-yi Qaul-i Fais:al, 4ff. Sahbāʾī’s

comments on each couplet have been printed as
footnotes in the published edition of TĠh.
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

140       

[muġhal-bachchah] often did not.)¹⁶¹ The later tażkirah tradition like-

wise comes to judge harshly any perceived deviations from Iranian usage.
Āftāb Rāy of Lucknow’s Riyāz
al-’Ārifīn [Garden of the Wise, 1883] is
such an example of linguistic purism. It was simply unthinkable by then
that Indian-born writers—in particular Hindus—could make any claim
on Persian even though this had been commonplace a century before.¹⁶²
Of Bedil’s numerous Hindu disciples, Āftāb Rāy mentions only Mukhli : s,:
and seems uninterested in Persian-writing Hindus despite the fact that
he was himself a Persian-writing Hindu.
Although eighteenth-century litterateurs were conscious of regional
variation in Persian, the best way to understand their critical debates is
not through the overdetermined concept of sabk-i hindī, or the “Indian
Style.” Instead of a narrowly nationalistic frame of Indians versus
Iranians, we should be thinking in terms of something universal, namely
Ancients versus Moderns, as we have seen with reference to the disagree-
ments between Ārzū and Munīr.¹⁶³ It is tempting to map literary dis-
agreements onto the factionalism of the eighteenth-century Mughal
court, but within the contemporary philological discourse this was not
the framework through which these debates played out. The crisis of
authority in poetry led Ārzū to think about the relationship between
the tradition and contemporary poetic practices, and to consider both
historical usages and regional variations across the Persian cosmopolis.
He recognized that languages are porous and fluid—a crucial develop-
ment in the amplification of Urdu literature to be addressed in

¹⁶¹ Ummīd’s hindī pronunciation (or perhaps fluency in speaking) was apparently poor
because Ārzū notes that “his tongue did not take to the Indian accent” [zabān-ash dar lahjah-yi
hindī k: hwub

na-mī gardad] despite having lived in India for nearly forty years and understand-
ing hindī well (MN 2004, 1:169).
¹⁶² “Although there are still some among ignorant Indians who consider him to be among
the most sublime writers, he is absolutely worthless in the opinion of those who really know the
Persian language. His Persian, like that of Nās: ir ‘Alī [Sirhindī (d. 1694)], is worse than Hindī”
(translated in Pellò 2016, 219 with modifications by me for clarity; original Tażkirah-yi Riyāz
al-
’Ārifīn 1977, 1:123).
¹⁶³ Snobbishness on the part of Indians around Indian Persian is attested only in rare
examples before the nineteenth century. For example, Mīrzā Muhammad : Rafīʿ Saudā
(1713–80), a writer in both Persian and hindī, is shocked that when one of his contemporaries,
:
Mīrzā Fākhir Makīn, was asked to correct a large selection of poetic quotations he refused to
work on the verses by Indian Persian poets and only deigned to consider what Iranians had
written (ʿIbrat al-Ġhāfilīn 2011, 43). Saudā wrote ʿIbrat al-Ġhāfilīn to criticize Makīn and their
falling out is dramatized in Āb-i Hayāt

1907, 156–61.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

    141

Chapter 4—while at the same time arguing for a cosmopolitan standard

determined by the consensus of good poets. To arrive at what Ārzū and
his contemporaries were concerned with, we must peel away the inter-
pretations of the last two centuries that have anachronistically reframed
the debates on authority as discussions of nationalism, socio-political
decline, and, of course, poor taste.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

4
Dictionaries Delimiting Literary
Language

Percy Bysshe Shelley’s “A Defence of Poetry” concludes bombastically

that “poets are the unacknowledged legislators of the world.”¹ If we are
willing to grant poets that expansive role, then it is worth also sparing a
thought for the unacknowledged parliamentary staffers who help the
poetic legislators legislate. In the early modern Persian cosmopolis,
lexicographers played such a consequential supporting role by shoring
up a baseline of words and expressions that counted as eloquent Persian
available to poets. Early modern dictionaries, which with rare exceptions
focus on poetic language, reflect a set of conscious choices about usage in
contrast to the modern comprehensive dictionary, which is typically
(if falsely) represented as a passive record of natural language. Pre-
modern Persian lexicography is a system of knowledge to be unpacked
because, as Walter Hakala has written, “lexicographical works reflect
dominant cosmographies.”² Pre-modern dictionaries represent a form
of standardization, albeit one that is unfamiliar to us today. This
included providing a stock of available poetry since most dictionaries,
Burhān-i Qāt̤ iʿ being the famous exception, provided poetic quotations
to illustrate their definitions. Especially in the early days of Persian
lexicography, authors often referred to their dictionaries as embedded
in a real literary community. For example, they might justify a diction-
ary’s compilation on the basis that friends struggled to read poetry of a
certain kind. An author might also reference a personal engagement with
the canon, for example declaring that from an early age he read poetry

¹ Shelley 1840, 14. The essay was written in 1821 but first published posthumously in 1840.
² Hakala 2016, 12; reviewed in Dudney 2017c. Hakala focuses on nineteenth-century devel-
opments in Urdu lexicography but his approach can be applied to reconstruct the Persian
lexicographical tradition before the colonial encounter.

:
India in the Persian World of Letters: Khān-i Ārzū among the Eighteenth-Century Philologists. Arthur Dudney,
Oxford University Press. © Arthur Dudney 2022. DOI: 10.1093/oso/9780192857415.003.0005
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

    143

but strained to understand its meaning, and so assembled the dictionary

in the course of study for himself and anyone else interested.³
That pre-modern Persian dictionaries connected the various geog-
raphies of the literary world is suggested by the striking consistency of
their being a product of the edges of the Persianate world rather than
of its core. During the Safavid-Mughal period, South Asia produced
by far the most dictionaries in Persian, and the origin of New Persian
lexicography was likewise not in Iran but in Khurasan to the northeast.⁴
Dictionaries at that time were hardly compiled in Iran proper, leaving a
gap of nearly three centuries between Surūrī’s Majmaʿ al-Furs (first
edition 1008/1600, Isfahan) and the next major lexicon, Rizā
Qulī
:
Khān Hidāyat’s Farhang-i Anjuman-ārā-yi Nāsirī : (1288/1871,
Tehran).⁵ Pre-modern Persian dictionaries reflect the tension between
the local and the universal in the Persianate world of letters because their
origins lay in the need to clarify unfamiliar expressions appearing in
literary works that had been written in different places from the ones
where they were being read.⁶ This process was especially pronounced as
the focus of literary patronage shifted from Khurasan (on the eastern
edge of the Persianate world) to Iraq (on its western edge) around the
twelfth-century , and again when the Mughals established themselves
as the world’s premier patrons of Persian letters in the sixteenth century.
The contributions of lexicographers were crucial both for the debates
over style relating to the position of the Ancients vis-à-vis the Moderns
described in Chapter 3, and the expansion of north Indian vernacular
literary culture described in Chapter 5. Dictionaries compiled across the
length and breadth of the Indian sub-continent had always either

: al-Dīn Mubārak Shāh Qawwās Ġhaznawī, the author of Farhang-i

³ For example, Fakhr
Qawwās (late thirteenth or early fourteenth century), the earliest extant Persian dictionary from
the Indian sub-continent, claims he wrote it for friends (Baevskii 2007, 73). Tek

Chand Bahār
calls Bahār-i ʿAjam the fruit of a research project that has taken him from youth [jawānī] up
until his present age of fifty-three (Bahār-i ʿAjam 2001, 1:xxix).
⁴ On dictionaries written in South Asia, see Naqvi 1962 (in Persian) and Blochmann 1868.
Remarkably, the last comprehensive survey in English on the subject seems to be Blochmann’s.
Baevskii 2007 is a survey of the whole Persianate world but only covers the eleventh to fifteenth
centuries, and Perry 1998 addresses the nineteenth century through a narrow selection of works.
⁵ Perry 1998, 338.
⁶ Baevskii 2007, 29–30.
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

144       

included some Indic words or used hindī equivalents to gloss the word
being defined.⁷ The tendency for lexicographers to include hindī
increased from a few stray words at first to relatively frequent glosses
in the seventeenth-century to eighteenth-century dictionaries whose
compilers actively sought out information on vernacular usage.
Here I focus on the Persian lexicographical tradition as it developed in
the first half of the eighteenth century. I begin by sketching the history of
lexicography in Persian, with particular attention to the dictionaries to
which Ārzū directly responds in his own works (namely Burhān-i Qāt̤ iʿ,
Farhang-i Jahāngīrī, Farhang-i Rashīdī, and Majmaʿ al-Furs). The next
section considers the place of Ārzū’s Persian dictionaries (Sirāj al-Luġhat
and Chirāġh-i Hidāyat) within the tradition and measured against each
other. Ārzū’s third dictionary, Nawādir al-Alfāz̤ , is a lexicon of Indic
words and has never been properly understood in light of his commit-
ment to Persian or the state of affairs in Persianate vernacular circles in
his time. Comparing Ārzū’s lexicographical scholarship to that of his
Hindu friends Tek
Chand Bahār and Ānand Rām Mukhli : s: provides an
opportunity to explore the tradition’s understanding of cultural differ-
ence. The chapter concludes with a sketch of Persian lexicography after
Ārzū, arguing that Persian lexicography remained important in South
Asia well after the East India Company’s general abolition of Persian in
administration in the 1830s, the point at which the Persian language is
widely but inaccurately believed to have become moribund in the sub-
continent.

Mughal Lexicography before Ārzū

Across the vast distances of the Persian cosmopolis, dictionaries were

crucial in constructing a shared literary language from as early as the
fourteenth century. In that respect, the works of Ārzū and other
eighteenth-century lexicographers fit into a pattern that was set centuries

⁷ “We may notice that nearly every province of India can point to a lexicographist”
(Blochmann 1868, 2–3). On glossing see Shirani 1966 with helpful commentary in Hakala
2016, 55.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

    145

before. In contrast to our sense of what lexicography seeks to accomplish

in recording language generally (or rather a subset of language repre-
senting “proper” usage), pre-modern dictionaries were concerned pri-
marily, and in most cases exclusively, with poetry. The social practice of
poetry, along with the philological apparatus like dictionaries that poets
depended upon, was arguably the chief affective tie between the disparate
parts of the Persianate cosmopolis.⁸ Given this regulatory function, it
is remarkable that the Indian sub-continent and not Iran was the
locus of pre-modern Persian dictionary-writing, as Henry Blochmann,
a nineteenth-century Orientalist associated with the Calcutta Madrasah
and the Asiatic Society of Bengal who was the author of the last com-
prehensive account in English of Mughal-Safavid period lexicography,
observed in 1868.⁹ Patronage appears to be a key factor, as we would
expect for such time-consuming projects, and some major Indian dic-
tionaries including Farhang-i Jahāngīrī and Burhān-i Qāt̤ iʿ were dedi-
cated to royal patrons, the Mughal emperor Jahāngīr (r. 1605–1627),
and ʿAbdullāh Qut̤ bshāh (r. 1626–72), the seventh ruler in the Shiʿite
dynasty of the Deccani state of Golconda. Besides patronage, something
less tangible, perhaps that dictionary-writing had a prestige in the
sub-continent that it did not have in the Safavid Empire, also explains
the stark contrast in the fate of the tradition in Mughal India and
Safavid Iran.
Between 1600 and the late nineteenth century, just one of the perhaps
two dozen major Persian dictionaries of the time was compiled outside of
South Asia. That work, Majmaʿ al-Furs, was first written in Isfahan in
1008/1600. Even so, Blochmann considers it “half-Indian” because its
author not only spent time in India but revised his dictionary after
receiving a copy of Farhang-i Jahāngīrī in 1621, which had been pro-
duced for the Mughal court in the meantime.¹⁰ Farhang-i Jahāngīrī
[Dictionary for Jahāngīr, 1017/1608] was compiled by the courtier Mīr
Jamāl al-Dīn Husain

Injū for Akbar (but named for Jahāngīr because

⁸ On the history and importance of dictionaries see Baevskii 2007; Perry 2012, and passim
Perry 1998.
⁹ Blochmann 1868, cf. Perry 1998 and Tavakoli-Targhi 2001, 106–7. The best work in
Persian on Indo-Persian lexicography is Naqvi 1962.
¹⁰ Storey 1984, 25.
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

146       

Akbar had died three years before its completion). The other two
important seventeenth-century dictionaries are Farhang-i Rashīdī
(1064/1654), compiled in Thatta, Sindh, and Burhān-i Qāt̤ iʿ (1062/
1651), compiled in Golconda in the Deccan. Both were intended to
provide corrections to Majmaʿ al-Furs and Farhang-i Jahāngīrī.¹¹ Since
these dictionaries are connected to each other and provide the basis for
eighteenth-century Persian lexicography as well as European efforts to
compile reliable Persian dictionaries, it is worth considering the circum-
stances of their composition in more detail.
Farhang-i Jahāngīrī was the great Mughal lexicographical project
because it was supported by the emperor Akbar himself.¹² The general
facts about Injū’s life are sketched in the “who’s who” of the Mughal
nobility, Maʾās̄ir al-Umarā: He first came to the Deccan from his
ancestral home, Shiraz. He arrived in Akbar’s court in the 1580s, was
promoted to the rank of six hundred then of a thousand, and capped his
political career by cementing an important political alliance between
Bijapur and the Mughal Empire. He escorted Ibrāhīm ʿĀdil Shāh’s
daughter to Agra to marry Akbar’s son Prince Daniyāl in 1603/1013.¹³
The dictionary was such an achievement that Maʾās̄ir al-Umarā
mentions it.
Injū frames his background and the lexicographical project in the
preface in tropes familiar across the Persian lexicographical tradition:
He had always had a poetic temperament but was constantly encounter-
ing words in poetry that did not make sense to him. He therefore made a
study of obscure Persian words for thirty years beginning at age ten.
What motivated the composition of the dictionary was an audience in
1005/1596 with Akbar in Kashmir in which the emperor declared that he
wanted “a book containing all the authentic Persian words, archaic
usages, and idiomatic expressions” in Persian (the “pārsī,” “darī,” and

¹¹ However, as Blochmann notes, the lack of words in Burhān-i Qāt̤ iʿ specific to Surūrī’s
second edition means that Burhān must only have used the first edition (1600) of Majmaʿ al-
Furs (Blochmann 1868, 16).
¹² Naqvi 1962, 81–7; Farhang-i Jahāngīrī 1975, ii.
¹³ For the year he entered Akbar’s service, Naqvi says 993–4  (1585–6 ) but Blochmann
(1868, 66ff.) says 1581 (although in fact Maʾās̄ir al-Umarā does not specify so it is unclear what
his source was).
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

    147

“pahlawī” words).¹⁴ For Akbar, the recovery of these words would be a

part of his imperial legacy. The method was, as Rajeev Kinra puts it, a
“scientific turn” in lexicography, as it depended upon forty-four diction-
aries enumerated at the beginning of the work. It was a mining of the
lexicographical tradition but also depended upon histories and commen-
taries as well as poets’ dīwāns to seek out words not mentioned in earlier
dictionaries.¹⁵ As Muzaffar Alam has noted, the text emphasizes purifi-
cation and correctness, but I differ with him in his interpretation that
these dictionaries “were all oriented to update the language in the light of
the current usages in Iran.”¹⁶ The difficulty with focusing on Iran is that,
as Ārzū would argue explicitly in the following century, literary Persian
was not the spoken language of any particular place. Mughal lexicog-
raphy was not meant to reorient Indian Persian specifically towards
usage in contemporary Iran but towards correct usage in general,
whether that was Iranian or otherwise. As we have seen, the modern
concept of the “native speaker” as the source of authentic knowledge of a
language does not map onto the pre-modern idea of an authorized user
of the language.
The three other major sixteenth-century dictionaries are connected
with Farhang-i Jahāngīrī, demonstrating the core unity of the lexico-
graphical tradition of the time. Majmaʿ al-Furs, also sometimes called
Farhang-i Surūrī after its author Muhammad: Qāsim Kāshānī whose
pen-name was Surūrī, has two editions, an Iranian first edition and a
possibly Indian second edition. The first edition (1008/1600) was dedi-
cated to the Iranian ruler Shāh ʿAbbās I and was followed by a revised
edition that engaged with Farhang-i Jahāngīrī. The second edition was
written after 1028/1619, around the time its author left Isfahan for
India, sometime between 1032/1623 and 1036/1626.¹⁷ Burhān-i Qāt̤ iʿ

¹⁴ The preface to Farhang-i Jahāngīrī (1975, 1ff. esp. 4) is contextualized with the relevant
passage translated in Kinra 2011. See Chapter 2 of the present book for a discussion of the
meanings of “pārsī,” “darī,” and “pahlawī” as Ārzū has adapted them from Farhang-i Jahāngīrī.
¹⁵ Farhang-i Jahāngīrī 1975, 5–7, 9–10. ¹⁶ Alam 1998, 336, 340–1.
¹⁷ The European traveler Pietro della Valle met him in Isfahan in 1032 , setting the date
after which he must have left, and he appears to have arrived in Lahore by 1036, if an inscription
in one of the manuscripts at the British Library is authentic (Rieu 1879–83, 2:498). Blochmann
dates the second edition to 1038 but this appears to be a misreading for 1028, the year in which
Surūrī read Farhang-i Jahāngīrī (Blochmann 1868, 16; Rieu 1879–83, 2:499).
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

148       

was dedicated to another ruler, ʿAbdullāh Qut̤ b Shāh of Golconda

:
(r. 1626–72).¹⁸ Its author was Muhammad Husain

:
bin Khalaf Tabrīzī,
known by his pen-name Burhān. Other than his composition of the
dictionary, Burhān is a historical cypher despite having been in a pos-
ition that allowed him to dedicate his dictionary to a ruler.¹⁹ Burhān-i
Qāt̤ iʿ is unusual among the major lexicons of its time in that it omits
poetic quotations, which in other dictionaries serve to provide evidence
for usage. As in tażkirahs, the citations in dictionaries serve a second
important purpose, namely providing a repository of vetted poetic
quotations that readers could draw upon. Burhān-i Qāt̤ iʿ became popular
later because it was expansive (with around 20,000 entries), having
absorbed much of Farhang-i Jahāngīrī and Majmaʿ al-Furs, and is
conveniently presented in strict alphabetical order.²⁰ However, in being
comprehensive it uncritically repeats mistakes in pronunciation and
interpretation committed by earlier lexicographers.²¹ Farhang-i Rashīdī
by ʿAbd al-Rashīd bin ʿAbd al-Ġhafūr is an outlier both geographically,
having been compiled in Sindh rather than in the imperial metropole,
and because it has no dedication. Although Farhang-i Rashīdī appears to
have no dedication, the Arabic dictionary ʿAbd al-Rashīd had written
two decades earlier, Muntak: hab al-Luġhāt-i Shāh Jahān (1046/1636),
was dedicated to Shāh Jahān.²² Farhang-i Rashīdī explicitly lists four
improvements on Farhang-i Jahāngīrī and Farhang-i Surūrī and gives

¹⁸ Naqvi 1962, 93–101.

¹⁹ His birth and death dates have not been established, and although it is assumed that he left
for the Deccan after completing his studies at Tabriz, in fact it might have been a father or
grandfather who emigrated (Dabirsiaqi 1989; Naqvi 1962, 96–7).
²⁰ Pre-modern Persian dictionaries tend to be arranged in the same hierarchy of headings as
other pre-modern works, with a chapter [bāb] for each letter and a sub-chapter [fas:l] for each of
the next letters, with headwords sometimes appearing in more or less random order in the sub-
chapters. Burhān-i Qāt̤ iʿ and Farhang-i Rashīdī are alphabetically arranged but other diction-
aries used more unfamiliar ordering systems: Farhang-i Jahāngīrī is perhaps the strangest since
it is organized by second letter then first letter. Majmaʿ al-Furs is organized by last letter then by
first letter.
²¹ Dabirsiaqi 1989.
²² Although Storey follows several catalogues in stating that Farhang-i Rashīdī is dedicated to
Shāh Jahān, I do not find any such dedication in the printed edition (Storey 1984, 35).
Blochmann implies that this is because the dictionary was compiled when Shāh Jahān “was
the prisoner of his perfidious son Aurangzíb, for whom Rashídí has no words of praise” but it
was completed (according to a very clear chronogram that Blochmann reads correctly) in 1064/
1653 and Shāh Jahān was not imprisoned until June 1658 (Blochmann 1868, 20–1; Richards
1996, 160).
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

    149

specific examples of mistakes from the previous works.²³ He notes that

such mistakes are more prevalent in Farhang-i Surūrī than in Farhang-i
Jahāngīrī. From the perspective of usability, Farhang-i Rashīdī is inter-
esting in that it separates out metaphors [istiʿārāt] into their own cat-
egory at the end of each sub-section.
Just as the important seventeenth-century dictionaries were written in
response to one another, the eighteenth-century lexicons also aimed to
correct and amplify earlier works. Ārzū’s stated project in Sirāj al-Luġhat
(discussed in detail in the following section) was correcting errors in
Farhang-i Rashīdī and Burhān-i Qāt̤ iʿ. In Mus̄mir, Ārzū discusses these
various dictionaries and declares Farhang-i Jahāngīrī to be the most
accurate of them even though Farhang-i Rashīdī and Burhān-i Qāt̤ iʿ
were both written later.²⁴ Some of the prefatory material in Farhang-i
Jahāngīrī is the centerpiece in Ārzū’s construction of the Persianate past,
as we have considered in Chapter 2. Farhang-i Rashīdī’s muqaddamah
[introduction], which is a brief summary of Persian grammar, was the
basis of ʿAbd al-Wāsiʿ Hānsawī’s Persian grammar, perhaps the most
widely used primer in the eighteenth century.

Reading Sirāj al-Luġhat and Chirāġh-i Hidāyat together

A notable trend in lexicography over the Mughal period was the increas-
ing awareness of cultural and linguistic differences between parts of the
Persianate world as well as between contemporary and earlier usage.
Ārzū’s Persian dictionaries, Sirāj al-Luġhat and Chirāġh-i Hidāyat, mir-
ror this recognition because while Sirāj al-Luġhat is a workmanlike

²³ ʿAbd al-Rashīd lists the faults of the two previous dictionaries as: (1) superfluous or
repeated verse quotations; (2) incorrect vowelings and definitions for words; (3) including
Arabic and Turkish words without specifically noting them as Arabic or Turkish rather than
Persian; and (4) repeating the same words under different spellings, which he notes is particu-
larly a problem in Farhang-i Surūrī. An additional category is the outright mistake [sahw-o
ġhalat̤ ], which he states is again particularly a problem in Farhang-i Surūrī, and this includes
misspellings involving letters that look nothing alike. He takes the unusual step of providing in
the preface a sample of the mistakes in this class that he has corrected (Farhang-i Rashīdī 1875,
1:1–3; quoted and translated in Blochmann 1868, 21–4). ʿAbd al-Rashīd claims that he can
support all of his corrections with quotations.
²⁴ M 1991, 39ff.; Kinra 2011, 373.
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

150       

correction of two earlier dictionaries, Chirāġh-i Hidāyat (which is Sirāj

al-Luġhat’s second volume but has generally been treated as a standalone
work, including by its author) contains words that had never appeared in
earlier dictionaries. Sirāj al-Luġhat was completed in 1147/1734 and only
survives in a few manuscript copies.²⁵ Later in life, Ārzū wrote in Mus̄mir
that his Sirāj al-Luġhat was based on “examination and research” [tanqīh:
:
wa tahqīq] and bragged that “some people say that until now no such
book of Persian lexicography had ever appeared” and that it makes over
1,000 corrections to Burhān-i Qāt̤ iʿ and other dictionaries.²⁶ In Sirāj al-
Luġhat’s preface, he praises Farhang-i Rashīdī for its careful use of
quotations and Burhān-i Qāt̤ iʿ for being the most voluminous diction-
ary.²⁷ However, he wrote the Sirāj al-Luġhat to fix their shortcomings,
which he notes are especially prevalent in Burhān-i Qāt̤ iʿ, and so does not
repeat quotations that are in Farhang-i Jahāngīrī, Farhang-i Surūrī, or
Farhang-i Rashīdī (Burhān-i Qāt̤ iʿ, of course, has no quotations to cite).
His list of sources, besides Farhang-i Rashīdī, Burhān-i Qāt̤ iʿ, Farhang-i
Jahāngīrī, and Farhang-i Surūrī, includes Farhang-i Qūsī, Muʾabbad al-
Fuzalāʾ,

Farhang-i Mūnisī (?),²⁸ Kashf al-Luġhāt, and some commentar-
ies on Saʿdī’s Gulistān and Rūmī’s Mas̄nawī. He particularly highlights
Farhang-i Qūsī, a dictionary by someone called Majd al-Dīn ʿAlī Qūsī,
“one of the learned people of Iran,” which in Ārzū’s estimation is very
good but not well known (indeed, I have been unable to trace this work
beyond Ārzū’s mention of it). Kashf al-Luġhāt wa al-Is: t̤ ilāhāt : is a
relatively well-known mid-sixteenth-century dictionary by ʿAbd al-
:
Rahīm :
bin Ahmad Sūr and Muʾabbad al-Fuzalāʾ

a relatively unknown
work by ʿAbd al-Rahīm’s: :
teacher Maulānā Muhammad : Lād
bin Shaikh

²⁵ SL 1747, f. 2a. The manuscript is an early copy dated 1160/1747 made at Delhi by a Hindu
scribe, Ratan Singh Munshī (Ethé 1903, 1353). It was owned by Richard Johnson (1753–1807),
who, given the titles inscribed on f. 1a, probably acquired it during the period he was posted in
Lucknow as assistant to the resident (1780–2). (On Johnson, see Marshall 2004.) The manu-
script later entered the East India Company Library in Calcutta, where it was consulted by Sir
William Jones, and eventually the India Office Library (now housed in the British Library).
²⁶ “chunānkih baʿzī : dar fann-i luġhāt-i fārsī kitābī chunīn dīdah na-shudah”

goyand kih tāhāl
(M 1991, 47).
²⁷ SL 1747, f. 2a.
²⁸ Blochmann reads this name as “Farhang i Múnisí” but in the British Library ms to which
I had access it appears to be—equally improbably—something like “Farhang-i Jūsī” (Blochmann
1868, 26; SL 1747, f. 2a.). Neither title is traceable.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

    151

Dihlawī, completed in 925/1519.²⁹ Sirāj al-Luġhat is a hefty tome—the

British Library’s manuscript weighs in at 682 folios—and it is clearly an
achievement but perhaps not the unique phenomenon that Ārzū implies
that it was in Mus̄mir. After all, both the dictionaries to which Sirāj al-
Luġhat is a response were themselves similar projects of responding
critically to earlier dictionaries, albeit with errors that Ārzū sought to
remedy. Sirāj al-Luġhat plainly states correct pronunciations and defin-
itions but in effect hides its work from the reader because it neither
quotes verses nor introduces the reader to the false pronunciations or
definitions being corrected. It is unsurprising therefore that the text
remained rare and that later European scholars drew upon the often
inaccurate Burhān-i Qāt̤ iʿ without realizing that Burhān’s errors had
been corrected.³⁰
The second volume of Sirāj al-Luġhat is Chirāġh-i Hidāyat, a lexicon
based on a very different principle since it deals with expressions used by
the Moderns not found in earlier dictionaries (or at least not found in a
particular meaning in early dictionaries).³¹ It was apparently completed
in the same year as the first volume of Sirāj al-Luġhat (that is, in 1734)
because its brief preface mentions no details specific to the completion of
the second volume. Chirāġh-i Hidāyat states that the words included are
of two kinds: “The first type are words whose meanings were difficult and
often the people of India did not have information about them; the
second are vocabules whose meanings were well-known, but there is
doubt about their being correct in the usage of the eloquent among the
ahl-i zabān.”³² The strategy in Chirāġh-i Hidāyat is to recognize the
Indian context of its intended readership explicitly: It offers Indic-
language equivalents for the names of objects to clarify what they are,

²⁹ Naqvi 1962, 146–8, 66–8; Ethé 1903, 1:1337–8. Muʾabbad al-Fuzalāʾ
divides words based
on whether their origin is Persian, Arabic, or Turkish. Kashf al-Luġhāt wa al-Is:t̤ ilāhāt : is
apparently also known under the title Kashf al-Luġhāt wa al-Mus: t̤ alahāt.:
³⁰ Blochmann writes that “the critical remarks on the Burhán are so numerous [in Sirāj al-
Luġhat], that the Burhán should never have been printed without the notes of the Siráj”
(Blochmann 1868, 25).
³¹ In his study of Indo-Pakistani Persian dictionaries, Naqvi claims that some words in CH
had in fact appeared in earlier dictionaries (Naqvi 1962, 116).
³² “qism-i awwal alfāz̤ ī-ast kih maʿānī-yi ān mushkil būd wa aks̄ar ahl-i hind bar ān it̤ t̤ ilāʿ na-
dāshtand qism-i duwwum luġhātī kih maʿānī-yi ān agarchih maʿrūf wa maʿlūm būd lekin dar
: h: būdan-i ān az rozmarrah-yi fus:ahā-yi
s: ahī : ahl-i zabān baʿzī

rā taraddud ba-ham-rasīdah”
(British Library ms Or 2013).
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

152       

but more fundamentally observes the variations within Persian literary

language by frequently employing the formula “the people of India say . . .”
or “in India they say . . .” to introduce a particularly Indian usage in
Persian. Thus, Chirāġh-i Hidāyat, a dictionary of contemporary poetry,
engages directly with Indians’ usage, unlike Sirāj al-Luġhat and other
previous dictionaries. For example, nowhere in Chirāġh-i Hidāyat does
Ārzū imply a dichotomy between “Indians” [mardum-i hind or ahl-i hind]
and “ahl-i zabān.” Rather, for parts of the Persianate world populated by
what we would think of as native speakers of Persian, he refers to “ahl-i
tūrān” and “ahl-i wilāyat,” i.e., Central Asians and Iranians.
Chirāġh-i Hidāyat’s pattern of citations gives us information about the
texture of Persian literary culture. One scholar has totaled the number of
named poets cited as asnād in the text as 184 across 2,075 words and
phrases.³³ (Unlike Sirāj al-Luġhat, which has virtually no citations,
Chirāġh-i Hidāyat provides a citation for every single entry.) I have
analyzed the frequency of references in the first 445 citations (through
the letter “b”), which represents roughly a quarter of the total, to discover
the ratio by which poets are cited.³⁴ In those 445 citations, three poets are
quoted more than 50 times each: (1) Taʾs̄īr, (2) Wahīd, : and (3) Salīm.
One is quoted more than 40 times: (4) Ashraf. Two are quoted more than
20 times: (5) T̤ uġhra̍ and (6) Sāʾib.

Quoted more than ten times: (7)
Shifāʾī, (8) As̄ar, then the anonymous poets taken together, (9) Mīr
Najāt, and (10) Z̤ uhūrī. To give a sense of proportion, each of the top
four represents between 10 and 15 percent of the total quotations (that is,
:
Taʾs̄īr, Wahīd, Salīm, and Ashraf together provide just under half the
total citations), while Z̤ uhūrī represents 3 percent of them. Many of the
184 poets are quoted just once. Besides these poets in the top ten, some
others worth mentioning are Kalīm (13) and Yahya̍ : Kāshī (14), two
important Safavid poets who settled in Kashmir, and Ārzū himself
(18). A few poets who are among the Ancients get a look in, for example,
Salmān Sāwajī and Hāfiz̤
are quoted a couple of times each. This list
represents a kind of virtual literary network that is unavailable to us
through any other means. The most commonly cited poets are all

³³ Perry 2002 citing Darvesh Bahriddin, whose Tajik article was not available to me.
³⁴ By compiling a spreadsheet that accounts for every quotation.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

    153

basically from the eleventh century , which is to say roughly the
seventeenth century . All were born in or near the Safavid lands,
indeed some (for example, Sāʾib
and Shifāʾī) spent their careers at the
Safavid court. Some Indian-born poets like Ġhanī Kashmīrī and Munīr
Lahorī appear further down in the list. Most of the poets had a connec-
tion to India, especially to Kashmir during Z̤ afar Khān
: :
Ahsan’s (d. 1073/
1662) governorship there. Sāʾib’s
own sojourn in the Indian sub-
continent was famously brief (he became friends with Ahsan : in 1624
when Ahsan: was governor of Kabul, accompanied him to the Deccan
and to the Mughal court, but left for Isfahan when Ahsan : took up his
post in Kashmir in 1632).³⁵ Taʾs̄īr, As̄ar, Shifāʾī, and Mīr Najāt appear
never to have left the Safavid lands. It is somewhat surprising how many
of the poets here—and among the rest of them if we were to consider
further down the list—are obscure. Taʾs̄īr is a case in point: He is the
most cited poet in the dictionary and must have been important in
Ārzū’s time, but he is virtually unknown today. Of course, literary canons
are reshaped in every generation and the eighteenth century’s assessment
of the significant literature of the seventeenth century is very different
from the view from the twenty-first century. Chirāġh-i Hidāyat gives us a
snapshot, admittedly a crude one, of which poets’ works were considered
worthy of inclusion and available to the author at the time when it was
written.
Chirāġh-i Hidāyat is a half-way house between the conservative lexi-
cography of Sirāj al-Luġhat, which merely seeks to correct earlier dic-
tionaries, and dictionaries of natural language. Although every entry is
accompanied by a literary quotation, Ārzū frequently refers to research,
especially from the “ahl-i zabān,” and provides details that are not
strictly speaking connected with literary language, such as equivalents
in Persian as used by Indians (presumably in less formal Persian) or the
equivalent word in hindī. He sometimes invokes a distinction between
common people [ʿawāmm] and more careful users of the language.³⁶
Literary research is combined with some degree of independent fact-
checking. For example, in defining the word “pādshāh” [king] he notes
that although the word was originally spelled with a “p” Indians changed

³⁵ Losensky 2003. ³⁶ For example, in the entry for “palang” (CH 1984, 1046).
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

154       

the word to “bādshāh” because the sound was “improper” [qabīh] : in

hindī (because, although he does not explicitly say so, “pād” refers to
flatulence).³⁷ The word “pā-yī chirāġh dāshtan” [lit. to have the base of
the lamp], which Ārzū defines as “to have hope of profit” and notes that
“it is [in] the language of merchants [ahl-i bazār].”³⁸ For “jahānābād”
(a shortening of “shāhjahānābād,” which is to say Delhi) he has a theory
that Iranians heard the common people use the short version and
adopted it.³⁹ Some specifically Indian concepts, such as “jigar-kh : wār”
:
(a kind of vampirical witch) and “chandāl” (broadly a member of an
“untouchable” caste but with other more precise meanings), are
explained in a way that is both literary scholarship but also reflects
observation.⁴⁰ In some cases, for example research into words that are
originally Arabic, such as “tamāshā” and “tamīz,” one sees Ārzū’s general
philological method at work.⁴¹

Other Major Eighteenth-Century Persian Dictionaries

Ārzū’s interventions in Persian lexicography came during a fruitful

period in Persian dictionary-making, the 1730s and 1740s. Ārzū’s friend

Chand Bahār was working on Bahār-i ʿAjam, which with around
Tek
10,000 entries and copious citations might be the longest pre-colonial
Persian dictionary ever compiled and it uses Sirāj al-Luġhat as a source.
Ārzū’s rival Wārastah compiled Mus: t̤ alahāt
: al-Shuʿarā [Expressions of
the Poets], which is far shorter than Bahār-i ʿAjam (containing about
2,000 idiomatic expressions) but more polemical because in addition to
engaging with the mainstream dictionary tradition it also draws upon
and privileges the views of Iranian informants. Mirʾāt al-Is: t̤ ilāh: by
Ānand Rām Mukhli : s: is a significant achievement, especially from a
modern perspective of what a dictionary should be, because it goes
beyond Chirāġh-i Hidāyat’s cursory observations on society to employ
what we could call a proto-anthropological method. Although it accom-
panies its definitions with literary quotations like Chirāġh-i Hidāyat, the

³⁷ CH 1984, 1044. ³⁸ CH 1984, 1045. ³⁹ CH 1984, 1074.

⁴⁰ CH 1984, 1071, 1081. ⁴¹ CH 1984, 1061.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

    155

quotations are frequently just excuses to describe something Mukhli : s: has

seen or heard about.
Bahār-i ʿAjam, the peak of the Mughal lexicographical tradition in
terms of sheer volume, invokes Ārzū as an inspiration. Remarkably,
Bahār claims to have used only two works by contemporary writers
[muʾallafāt-i mutaʾakh: khirīn]
: during the initial composition of the dic-
tionary, namely Ārzū’s Tanbīh al-Ġhāfilīn and “a short tract” by Mīr
:
Muhammad Afzal

S̄ābit.⁴² He looked at other contemporary critical
works in the course of revisions. Ārzū’s own Sirāj al-Luġhat comes first
in the list of reference works consulted. One fact worth noting is that
both Ārzū and his intellectual nemesis Shaikh : Hazīn

are both repre-
sented in the text. Indeed, Bahār states that he will use the honorific
“Lamp of the Researchers” [sirāj al-muhaqqiqīn]
: for Ārzū and “Shaikh: of
the Wise” [shaikh : al-ʿārifīn] for Hazīn.

Bahār expresses his deep devo-
tion to Ārzū and uses works by Ārzū’s friend S̄ābit, and yet also acknow-
ledges his great debt to Wārastah’s Mus: t̤ alahāt
: al-Shuʿarā, and refers to
Hazīn

in highly complimentary terms.⁴³ Bahār was clearly a partisan of
Ārzū, and yet like Mukhli
: s: , had no qualms about relying upon the works
of Ārzū’s enemies. Bahār could play both sides of the contemporary
debate over literary style as he was concerned with the contemporary as
it affected the written record, which was what actually interested him.
His access to the tradition was encyclopedic: Some manuscripts, includ-
ing the one held in the Bodleian, list 200 or so sources, including poetic
collections, dictionaries, and commentaries.⁴⁴ This range demonstrates
an encyclopedic grasp of the tradition, as in Ārzū’s own lexicographical
projects, but also a surprising willingness to use texts written by people
who did not get along with one another philosophically or personally.
Ārzū’s own works in the list include a dīwān of ġhazals, his
Sikandarnāmah and Gulistān commentaries, the dictionary Sirāj al-
Luġhat, and “several treatises” [baʿz
rasāʾil] as well.⁴⁵
Some later scholars, following Blochmann, have been confused about
the date of Bahār-i ʿAjam. The preface has an ambiguous chronogram

⁴² Bahār-i ʿAjam 2001, 1:30. ⁴³ Bahār-i ʿAjam 2001, 1:xxix–xxxi.

⁴⁴ Ms Caps. Or. B 15, f. 2b; on this ms see Sachau and Ethé 1889, 1018.
⁴⁵ Ms Caps. Or. B 15, f. 2b.
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

156       

best read as 1152 (= 1739), but Blochmann erroneously claims that the
first edition was not completed until 1752 and the final (seventh) revision
was completed by Bahār in 1782.⁴⁶ The preface also clearly states that
Bahār was fifty-three years old at the time of its completion, meaning
(if the dating proposed here is correct) that he was born in the late 1680s
and thus exactly the same age as Ārzū. By extension if they had been
friends for twenty years in 1740 then that means they met around the
time when Ārzū settled in Delhi. Further evidence is the fact that ʿAt̤ ā-
:
allāh Khān Nudrat’s dictionary ʿAin-i ʿAt̤ a, completed in 1749, cites both
Sirāj al-Luġhat and Bahār-i ‘Ajam, suggesting that those dictionaries
were circulating in Delhi and more importantly that Bahār-i ‘Ajam was
by then long finished.⁴⁷ In fact, Bahār’s pupil Indarman completed an
abridgement in 1182/1768 and context makes it clear that Bahār had died
by then.
Mus: t̤ alahāt
: al-Shuʿarā is somewhat like Chirāġh-i Hidāyat in that
Wārastah’s aim, like Ārzū’s, was to collect expressions that appeared in
poetry but that the lexicographical tradition had not yet captured. He
explains that he came across expressions “unfamiliar to Persian tongues”
[ġharībah-yi fārsī-zabānān] and used research [tahqīq] : during a fifteen-
year search [talāsh] to discover their meanings.⁴⁸ The emphasis on the
authority of Iranians to declare the meaning of an expression, including
in some cases resting authority on “muhāwarah-dānān-i
: īrān” [lit. the
“idiom-knowers of Iran”], is different from Ārzū’s project. Although
Wārastah explicitly mentions checking with Iranians [zabāndān-i īrān-

⁴⁶ The Bodleian catalogue (Sachau and Ethé 1889, 1018) incorrectly calculates the chrono-
gram as 1162 by reading “bā dah sāl” [with ten years (more), i.e., 1152 + 10] for “māddah-yi sāl”
[derivation of the year] following the British Museum Catalogue (Rieu 1879–83, 502), which
repeats the mistakes of Sprenger’s Oudh Catalogue. Sprenger, for example, refers to Bahār’s use
of a treatise by Ānand Rām Mukhli : s: but in fact it is by Mukhli : s: Kāshī (that is, Mīrzā
:
Muhammad : s: of Kashan) and not by Ānand Rām. The argument for the date of the
Mukhli
final edition (1782), furthermore, comes from assuming an incorrectly late date of composition
for Wārastah’s Mus:t̤ alahāt
: al-Shuʿarā. If Bahār revised the text in 1782 then he would have been
approximately ninety-five years old. Naqvi reports the correct date of composition (1962, 154).
⁴⁷ The ms is British Library IO Islamic 1813. Nudrat also mentions Farhang-i Majd al-Dīn
Qūsī—the same untraceable dictionary used by Ārzū in SL—as a source (Ethé 1903, 1:1354).
According to Ethé, Nudrat claims to have worked on his dictionary for twenty years, which is
probably an exaggeration given how heavily it uses Bahār-i ʿAjam and SL, neither of which
would have been finished when he started work. Ārzū complains in MN about how much of
Ārzū and Bahār’s text Nudrat included in the dictionary (MN 2005, 121).
⁴⁸ Mus: t̤ alahāt
: al-Shuʿarā 1985, 1.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

    157

dīyār] there is no reason to assume, as Blochmann does, that this meant

that Wārastah traveled in Iran.⁴⁹ The title is a chronogram for the year he
started writing, 1149/1736–7, which points to a likely completion date of
1164/1751 since he implies the writing took fifteen years.⁵⁰ Wārastah
died in 1180/1766–7.
: s’s
Mukhli : Mirʾāt al-Ist̤: ilāh: (1158/1745) is a favorite source for students
of eighteenth-century social history because it frequently observes actual
practices. Charmingly it tells us details such as which flowers are grown
in Delhi, and revels in seemingly off-topic anecdotes and sayings. Ārzū
revised the work after Mukhli : s’s: death in 1164/1751 so we can assume
his endorsement of the contents. Mirʾāt al-Ist̤: ilāh: is somewhat different
from previous dictionaries in that it contains a great deal of “proto-
anthropological” observations as well as long digressions describing, for
example, particular people that Mukhli : s: knew such as Ārzū or objects
like the Peacock Throne. Additionally, it ends each letter’s section with a
series of adages (ams̄āl). Despite these unusual features, it fits squarely
into the tradition of poetic dictionaries because it generally uses poetic
quotations for authority, as other dictionaries do. This is worth high-
lighting because it was written after Nādir Shāh’s conquest of Delhi had
completely changed the political relationship of the Mughal Empire with
Persia.⁵¹ One sometimes hears the anecdote that Mukhli : s: chased after
soldiers in Nādir Shāh’s army with a notebook to ask them about how
they used Persian (the idea being that they had correct “native-speaker”
Persian), but this is an anachronism—he would not have thought them
worth talking to because they certainly did not have the right educations
or lineages to be master poets. In the work there is a deep attention to
administrative terminology and, as the work’s editors note, to calli-
graphic and painting terminology.⁵² Some evidence for definitions

⁴⁹ Blochmann 1868, 30.

⁵⁰ Naqvi’s interpretation of the chronogram does not take into account the assimilation of
the “al-” in the title so he cites 1180 as the start date, which is impossibly late (Naqvi 1962, 159;
Mus:t̤ alahāt
: al-Shuʿarā 1985, ii).
⁵¹ The editors claim that 1156  (i.e., 1743 ) was the year he started writing it on the basis
that 26 Muharram 1156 is the first named date in the text.
⁵² Mirʾāt al-Is: t̤ ilāh: 2013, 42ff., English introduction. In the entry for “dar-pardah” Mukhli
: s:
mentions his own collection of paintings in the winter bedroom, including one of Holi [majlis-i
holī], which he claims is worthy of Mānī and Bahzād (2013, 354). There is a lot of fascinating
material here: Firstly, that the Hindu festival is described with the word “majlis,” a word with
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

158       

comes from Mukhli : s: himself and some from named people (some of
whom may be traceable), but the favorite technique remains quoting a
verse as sanad as nearly all Persian dictionaries of the past had done.
There is a lot of fascinating material in the text, but there is a risk in
: s: in the role of a present-day anthropologist lest we forget
casting Mukhli
that he is embedded in a literary-cultural system.
Although Mukhli: s: seems interested in Iranian Persian as opposed to
Indo-Persian, I believe that scholars have overinterpreted the reason for
this orientation. He mentions equivalent administrative positions in
Nādir Shāh’s court and in India, but it seems a reasonable assumption
that this was not because the Iranian terms were more “correct” but
rather because the Iranians were, at least for a time, in charge.⁵³ The
editors of the critical edition claim that “It can be seen from a close
reading of the text that after finishing the work, he got it authenticated
from speakers of the language just arrived in India.”⁵⁴ They do not cite
text in support of this proposition and implications of the claim are
somewhat misleading. Mukhli : s: asked authorities about certain head-
words but this is no different from the tradition (in which the second-
best evidence for a definition, failing finding authoritative poetic sanad,
was to ask someone) and certainly does not represent “authentication” of
the text by native speakers in toto.
: s: ’s Mirʾāt al-Is: t̤ ilāh: differs from most other pre-colonial
Mukhli
Persian lexicons in that it focuses on general vocabulary—often bureau-
cratic terminology and flora and fauna—rather than poetic usage, and its
entries are frequently more encyclopedic than straightforward defin-
itions. For example, he might praise someone in an entry (such as
Ārzū, Ās: af Jāh, or Qizilbāsh Khān : Ummīd) or include a tangentially

overtones of Persianate sociability. Secondly, that the quality of the work is referenced with the
Ancient painter of Persianate legend, Mānī, and the by-then-storied Timurid-Safavid painter of
late fifteenth and early sixteenth-century Shiraz and Tabriz, Kamāl al-Dīn Bahzād.
⁵³ The editors note that his work Guldastah-yi Asrār is a compilation of specifically Iranian
chancellery terms—he studied Nadir Shah’s court’s letters to come up with the list of terms.
They mention this in the context of Mukhli : s: ’s introductory statement of his goal for composing
Mirʾāt al-Is: t̤ ilāh,
: namely presenting words not recorded in contemporary dictionaries (precisely
Ārzū’s project in Chirāġh-i Hidāyat). By juxtaposing the two works, the editors introduce an
Iranian standard where Mukhli : s: does not imply one. In any case, I have not come across the text
in any catalogue.
⁵⁴ Mirʾāt al-Is:t̤ ilāh: 2013, English introduction, 33.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

    159

related anecdote, joke, or saying. Indeed, Mukhli : s’s: dictionary can be

thought of as being like a tażkirah in terms of building a community
through memory. The text is crowded with everyday details unavailable
in other works: The entry on coffee [qahwah] complains of high prices in
the two Chandni Chowk coffeehouses, and we know what sorts of
flowers grow in Shahjahanabad from the entry on “gul-i jaʿfarī” [a kind
of yellow flower].⁵⁵ There is a strong sense of the author’s desire to know
about objects and social practices that contrasts with other lexicog-
raphers. Unconstrained by the need to provide evidence on the basis of
poetic quotations, which would be the case for a typical Persian farhang
[lexicon], Mukhli : s: is himself often the authority on the meaning of a
word or phrase. Interestingly, he also calls upon various informants who
had traveled outside of India and even takes the opportunity of Nādir
Shāh’s invasion to learn about Iranian chancellery practices first-hand.⁵⁶
He frequently draws comparisons between Iranian and Indian usages in
Persian but does not give Iran primacy and instead merely provides a
sketch of how terms are used in other parts of the Persian cosmopolis.
Mirʾāt al-Is: t̤ ilāh: is undoubtedly a text intended for an Indian readership
since it frequently quotes hindī equivalents of Persian terms and expres-
sions, including hindī adages [ams̄āl]. Arguably the most delightful is
that “laddus [sweets] are not distributed during war” (the equivalent
Persian saying has “halva” as the sweet in question).⁵⁷ Although Mukhli : s:
was Ārzū’s close friend, he sometimes cites Ārzū’s archrival Hazīn,
for

⁵⁵ Mirʾāt al-Is:t̤ ilāh: 2013, 570–2, 621.

⁵⁶ Some examples from the text: “Dūd-i mashʿalī” means lamp-black (but here is described as
a deep blue color like the nāfarmān flower) and the information was provided by Safdar

:
Muhammad :
Khān, who had acted as an envoy to the Safavid court during Muhammad :
Shāh’s reign (Mirʾāt al-Is: t̤ ilāh: 1850, 260). “S̄aʿlab farush” refers to a seller of a kind of hot
sugar-syrup beverage. Hājī Naz̤ īr, who recently returned from Iran, explained the drink and
noted that he had had it himself in Mashhad but only in the winter months (Mirʾāt al-Is:t̤ ilāh:
1850, 171). “Farmān bi’l-mushāqqah” refers to an oral order of the king that does not require the
chancellery’s seal and the evidence for this usage is provided by documents from Nādir Shāh’s
chancery (Mirʾāt al-Is: t̤ ilāh: 1850, 409). Likewise, “mīrzā-yi daftar” (a clerk in the imperial
establishment) was a term that Mukhli : s: heard repeatedly from people around Nādir Shāh
(Mirʾāt al-Is: t̤ ilāh: 1850, 490–1). The invasion also brought new material things to Mukhli : s: ’s
attention, such as “chūb-dast” (a kind of stick) seen in Delhi when Nādir Shāh’s soldiers came
(Mirʾāt al-Is: t̤ ilāh: 1850, 198–9).
⁵⁷ The Persian expression “dar jang halwā : bak: hsh namīkunand” is rendered into hindī (in
sumptuous calligraphy in the British Library’s copy) as “lar: āʾī meṁ koʾī lad: dū : nahīṁ bat:te”
(Mirʾāt al-Is: t̤ ilāh: 1850, 265). It appears among several hundred sayings.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

160       

example, on the fact that “mahtāb” appears in works of ancients in the

meaning “moon.”⁵⁸ Clearly there were limits to personal animosity if
Hazīn

could find a place in a work by one of Ārzū’s most ardent
partisans. Significantly, Ārzū received the text after Mukhli : s’s
: death
and added marginal notes and a preface.
Although it was especially common in Persian dictionaries of this time
to provide the equivalent hindī term (for example, in Chirāġh-i Hidāyat),
Mirʾāt al-Ist̤: ilāh: perhaps includes them more frequently than in any
other dictionary. Furthermore, Mukhli : s: appears to have shared Ārzū’s
relatively permissive views on how a hindī word can be adopted in
Persian poetry. After the entry for “maniyār” (an Indic word for a kind
of jeweled bangle), Mukhli : s: writes in a “fāʾidah” (a heading which we can
perhaps translate as “useful note”):

It ought to be known that some venerable persons believe that an Indic

word should not be used in a Persian verse because the verse is thereby
devalued. In one of the couplets written by the same master [ustād]
Hakīm

Hāżiq

[who provided the previous example] has used the word
“chūrī” [bangle]. Likewise, Mullā T̤ uġhrā has a large number of Indic
words in his verses. It means that these restrictions are [only] for
inexperienced and beginner poets. It is quite lawful for the master-
poets who possess the ability to use them.⁵⁹

A large number of Indian concepts appear in Mirʾāt al-Is: t̤ ilāh. : For

example, “qashqah,” the sandalwood mark placed on the forehead by
upper-caste Hindus called a “tīkah”; “bīrah-yi pān,” the chewable prep-
aration of areca nut and spices wrapped in betel leaves, which is accom-
panied by a fāʾidah describing of various presentations of pān; and
administrative terminology relating to land grants in India.⁶⁰
Some entries relate to the Iranian local or customs with no analogue
in India. For example, “s̄aʿlab farosh” a seller of hot syrup (i.e., salep)

⁵⁸ Mirʾāt al-Is:t̤ ilāh: 1850, 477.

⁵⁹ Mirʾāt al-Is: t̤ ilāh: 2013, 660–1 (the British Library ms, Mirʾāt al-Is: t̤ ilāh: 1850, has a lacuna
: s: goes on to cite Yahya̍
here). Mukhli : Kāshī’s use of the word “gilahrī” [squirrel] in support of the
same principle of mastery (this part is in the British Library ms on f. 252b).
⁶⁰ Mirʾāt al-Is:t̤ ilāh: 2013, 566, 148, 51–3. In hindī, “bīrah” is written “bīr: ā.”
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

    161

typically found in bazārs in Iran. Hājī
Nażīr, who recently returned from
Iran, is cited as Mukhli: : s’s informant. Others include names of particular
places, such as streets, rivers, or gardens in Isfahan and Herat, such as
“āb-i shīrāzī,” a metonym for wine but also the name of a river in Isfahan,
and “bāġh-i bulbul” [Nightingale Garden], which was apparently in
Isfahan.⁶¹
Others present contrast or loosen comparisons between India and
Iran. For example, “bar sar sang nishāndan,” a punishment in which
hot stones are put on someone’s face, which Mukhli : s: notes is originally
Iranian but now is practiced in India as well. “Kamar-chīn” refers to an
Indian robe [jāmah] which is not worn in Iran. The reason is that it is
pleated and men’s robes are not pleated in Iran, where indeed the only
people who wear pleated robes are “courtesans and dancing boys”
[loliyān wa at̤ fāl-i raqqās]. : In the entry for “dast gardān” [money given
as a loan] Mukhli: : s conflates the festivals of Nowroz and Diwali. In the entry
for “turunj-i t̤ ilā” [lit. golden orange], a groom-picking ceremony of the
“sultans of ʿAjam” is compared to the Hindu swayamwara : [here: sīmbar],
the ceremony in which a princess chooses the groom as in the Sanskrit
: s: cites a number of proverbs in both Persian and hindī.
epics.⁶² Mukhli
Although a smattering of Indic words had always made an appearance
in Persian lexicographical works, by the eighteenth century such words
were frequent even in general purpose literary dictionaries like Bahār-i
ʿAjam.⁶³ Mukhli: s: goes further than some of his contemporaries and
positions his lexicography at the interface of Persian literary culture
and a largely hindī-speaking life-world. Ārzū has a similar project of
exploring the edges of Persian in Nawādir al-Alfāz̤ .

Ārzū’s Vernacular Lexicography

The knowledge that Indic loanwords existed in Persian was certainly not
novel in the eighteenth century—Majmūʿat al-Furs, which probably

⁶¹ Mirʾāt al-Is:t̤ ilāh: 2013, 250, 13, 99.

⁶² Mirʾāt al-Is:t̤ ilāh: 2013, 111–12, 594, 359–60, 211–12.
⁶³ On various hindī-derived words in Bahār-i ʿAjam, see Ali 2002. See Karomat 2014 on Indic
words in fourteenth- and fifteenth-century Persian dictionaries.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

162       

dates to the fourteenth century, marked some words as Indian in

origin⁶⁴—but a critical engagement with this borrowing, such as Ārzū’s
description of linguistic sharing [ishtirāk] between Persian and Indic
languages, only developed in the eighteenth century. The Indic words
that had by then become part of the Persian poet’s available stock include
common items, e.g., “pānī” [water], as well as South Asian toponyms,
e.g., “pānīpat” (a place in modern-day Haryana). The two of those
together leads to delightful examples like Mullā T̤ uġhrā’s tongue-in-
cheek declaration in the mid-seventeenth century that:

Whoever drinks the water [pānī] of Panipat

Sprouts a parrot-like gift of the gab.⁶⁵

Poets of this time reveled in such wordplay, which in the case of this
couplet works on several levels: water is contained within the name of the
town and leads to green foliage (like the proverbial greenness of parrots,
who are also known for their ability to talk), and the expression for
eloquence “rat̤ b al-lisānī” literally means “moistness of the tongue.”
Indic words had by this time become another arrow in a poet’s quiver
of verbal effects.
Ārzū wrote Nawādir al-Alfāz̤ sometime before 1156/1743 towards the
end of an illustrious career as a Persian poet and lexicographer but also—
as we consider in depth in Chapter 5—as a promoter of vernacular
literature.⁶⁶ There is seemingly a contradiction in Ārzū’s legacy in that

⁶⁴ On this dictionary, which should not be confused with the seventeenth-century Majmaʿ al-
Furs, see Baevskii 2007, 64. The preface refers to the difficulty of reading the Shāhnāmah
because of its multidialectal style (Farhang-i Majmūʿat al-Furs 1977, 1–3). By my count there
are three Indic [hindawī] words in the dictionary, namely “laund” [defined as “the impetuous-
ness of the effeminate” (k: hez-i muk: hannas̄)], “land” [penis], and “rāʾī” [an Indian king]
(Farhang-i Majmūʿat al-Furs 1977, 62, 62, 263). Without mentioning the term’s Indic origin,
the dictionary Burhān-i Qāt̤ iʿ defines “laund” as either an “immoral woman” or a “boy who
commits bad acts” [pisar-i badkārah], an obvious sexual euphemism (Burhān-i Qāt̤ iʿ 1850,
2:230).
⁶⁵ “zi pānīpat ānkas kih noshīdah pānī / cho t̤ ūt̤ ī shudah sabz rat̤ b al-lisānī” (quoted in Hasan
1998, 9). T̤ uġhrā (d. before 1026/1667–8) was born in Mashhad and came to India in Jahāngīr’s
reign, becoming munshī to Shāh Jahān’s youngest son, Prince Murād Bakhsh. : He seems to have
been one of the most enthusiastic users of Indic words in Persian (see Dudney 2017b).
⁶⁶ The evidence for the date of composition is found in the definition for “baisākh”
[Vaiśākha, the second month of the Hindu calendar that falls in April–May]. See NA 1951,
96; cf. NA 1951, xvi and ʿAbdullah 1965, 46, which both mistakenly cite the year as 1165 rather
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

    163

he is recognized as a towering figure in rekhtah

: (that is, in hindī poetry
written according to Persian conventions) by his contemporaries and yet
his extant work in the vernacular consists of a few couplets that would
not fill a page and whose attribution is uncertain anyway. What is the
source of Ārzū’s reputation in rekhtah/Urdu
: if not a collection [dīwān] of
poetry such as the cherished volumes left to posterity by his contempor-
aries like Shāh Hātim

and Mīr? The answer to the riddle is not found in
how Ārzū wrote Urdu poetry, but rather must be how he thought about
Urdu. His ideas about vernacular composition were clearly passed down
to other poets, who respected him in the first place because he was a great
Persianist. His embedding of arguments about the vernacular in works
nominally about Persian literature will be considered in the next section.
Here I turn to the importance of Ārzū’s dictionary Nawādir al-Alfāz̤ as a
bridge between Persian literary culture and the burgeoning vernacular
composition that Persian-language poets were increasingly interested in
exploring.
Nawādir al-Alfāz̤ is the first critical dictionary of hindī/Urdu in
Persian, and thus represents an attempt to bring the tools available for
Persian literary criticism to bear upon the vernacular.⁶⁷ The definition of
language in the eighteenth century in India and elsewhere was based not
on structure but rather on what we would call socio-linguistic criteria:
A language was defined less by a set of formal characteristics than by its
users and the contexts in which they used it.⁶⁸ For that reason, people
who had good literary judgment in one language could apply it to
another language. Furthermore, Ārzū—like modern socio-linguists—
acknowledges that languages are fundamentally porous. He recognizes
that from ancient times Indic words had been entering into Persian, and
that hindī freely borrowed Persian words and grammatical structures, a

than 1156. Walter Hakala has calculated that the lunar date 24 Muharram : 1156 corresponds
with a date in the Hindu month Vaiśākha, which is not true for the year 1165, conclusively
showing that 1165 is a repeated typo for 1156 (personal communication, 15 January 2014).
⁶⁷ Here “critical dictionary” is somewhat of a hedge because certainly there were earlier
lexicons, such as a Persian dictionary composed in Gujarat that had a chapter on hindī words
used in poetry (Faruqi 2001, 73) and a hindī-Persian dictionary probably composed in
Rajasthan in 1764 (Truschke 2012b, n148).
⁶⁸ I have tried to theorize this user-based perception for Braj Bhās: ā by considering colonial-
era misinterpretations of how language was used in India (Dudney 2010a).
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

164       

process which arguably intensified in his own time.⁶⁹ The key difference
between the cosmopolitan Persian tradition and the localized hindī
tradition was that the former had been constituted by centuries of both
poetic practice and scholarship while the latter was based only on poetic
practice. We see a parallel in early modern Europe, where Latin existed
alongside vernacular literatures that had flourished for centuries but did
not develop a written critical tradition until they were influenced by
Latin.⁷⁰ The techniques for classifying and assessing the literature (or
rather the words, phrases, and literary tropes) of the cosmopolitan
language shaped the vernacular literature and standardized its usage.
The twentieth-century critic Sayyid ʿAbdullah refers to this process as
“washing out the stain of lack of gravitas [be-iʿtibārī]” that kept Urdu
from being fit for serious writing, but such rhetoric, implying shame over
Urdu’s undeveloped early state, clearly represents a modern Urdu
speaker’s feelings projected onto the past.⁷¹
Shamsur Rahman Faruqi has written that “with his vast erudition in
comparative philology, considerable wit and elegance of style, Khan-e
Arzu left a model in Nawādir-al-Alfāz̤ which our later lexicographers
unfortunately did not follow closely” when they applied themselves to
writing dictionaries of Urdu in Urdu.⁷² In the present discussion of the

⁶⁹ Compilers of much earlier Persian dictionaries were aware of linguistic borrowings.

Numerous individual lexemes in these dictionaries make reference to a language of origin,
such as one surprising entry in Majmūʿat al-Fars [A Persian Collection, fourteenth century?]:
“land bah zabān-i hindī nām-i kīr ast” [“In the hindī language, ‘land’ is the word for penis”]
(quoted in Baevskii 2007, 64).
Beyond individual entries, some prefaces note patterns of borrowing. For example, the Delhi
Sultanate-period Dastūr al-Afāzil
[Canons of the Learned, 1342 ] contains, according to the
preface, “Arabic, Turkish, Mongolian, Pahlavi, Persian, Afghan [Pashto], Jewish [Judeo-
Persian?], Christian [Aramaic?]; the tongues of the Magians, Syrians, philosophers and Tajiks;
Hebrew; words from the dialects of Rayy, Hijaz, and Transoxania, poetical idioms from every
city, scholarly coinages, and popular sayings” (translated in Baevskii 2007, 81). See also the
preface of Burhān-i Qāti̤ ʿ, which has a similar list. This kind of multilingual consciousness also
exists in the Arabic tradition, as in Jalāl al-Dīn al-Suyūt̤ ī’s Mutawakkilī, a lexicon of originally
non-Arabic words found in the Qurʾān and arranged by presumed language of origin (Bell 1924).
⁷⁰ See, for example, Burke 2006. New Persian literature itself seems in the historical record to
appear fully formed in the tenth century but there was no well-established critical tradition in
Persian until the thirteenth (Clinton 1989). Arabic had been a strong influence on Persian
literature long before the Arabic-derived critical tradition began in Persian. This is a point that
certainly would not have been lost on Ārzū and his contemporaries. Obviously Persian
influenced vernacular literature long before people began writing critically about Urdu.
⁷¹ ʿAbdullah 1965, 45. ⁷² Faruqi 1990b, 29.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

    165

work, I will advance two related arguments. Firstly, Ārzū consistently

:
places rekhtah-style hindī into the same linguistic and cultural frame as
Persian, most explicitly through the concept of tawāfuq (the idea laid out
in Mus̄mir that there is an underlying genetic relationship between
Persian and Indian languages). Although he never argues for it
specifically—in part, we can assume, because it would have been obvious
to his readers—this means that Persian discursive practices could be
applied to Urdu poetry just as Arabic practices had been applied to
Persian or Persian had influenced Turkish literary culture. The rich
canon of Persian literary theory could be brought to bear on the ver-
nacular because even if vernacular criticism was underdeveloped at this
point, the deep linguistic bonds between the languages mean that the
theory of Persian literature was not foreign to hindī. Secondly, I build on
the argument in Chapter 2 in the context of Mus̄mir that Ārzū was trying
to establish a Delhi-centered standard for hindī poetry along the lines of
how a standard Persian came into existence. While he is content to
record non-Delhi usages, he rejects any that would not meet with
approval in the capital. His approach is therefore comprehensive in its
research but considerably narrower in its conclusions.
Although its methodology is original, Nawādir al-Alfāz̤ is a correction
of an earlier lexicon, Mīr ʿAbd al-Wāsiʿ Hānsawī’s Ġharāʾib al-Luġhāt
[Rarities of Words, late seventeenth/early eighteenth century] with a
considerable number of additional entries. Ārzū often invokes quota-
tions from the previous work with the phrase “dar risālah . . .” [in the
treatise . . . ], and sometimes full-throatedly disagrees with ʿAbd al-
Wāsiʿ’s interpretations, many of which are indeed simplistic. Some
manuscripts of Nawādir al-Alfāz̤ lack the new title and just call it
Ġharāʾib al-Luġhāt or “a correction” [tas: hī : of Ġharāʾib al-Luġhāt.⁷³
: h]
As he states in his preface, Ārzū has kept all of the words that appeared
in Ġharāʾib al-Luġhāt, even though he occasionally questions ʿAbd

⁷³ For example, the British Library’s copy, ms Or. 12,015, is catalogued as Ġharāʾib al-Luġhāt,
and indeed even has enclosed a letter from one Zulfishan Noor who wrote on February 3, 1938
that Ġharāʾib al-Luġhāt is by Ārzū (with no mention of ʿAbdul Wāsiʿ) and “is a recognised book
among research workers.” He urged Sir Gerard Clauson, who later presented this copy to the
British Library, not to bear the expense of publishing the work because it is widely available in
manuscript. The mss of NA at Aligarh Muslim University are catalogued similarly.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

166       

al-Wāsiʿ’s reasons for including some of them (for example, “takiyā”).

Although he praises ʿAbd al-Wāsiʿ in the preface as “one of the successful
learned men and famous scholars of heaven-resembling India” [yakī az
fuzalā-yi

kāmgār wa ʿulamā-yi nāmdār-i hindūstān jannat-nishān], it is
clear that he finds ʿAbd al-Wāsiʿ’s scholarship lacking because he then
uses four synonyms for “mistake” to describe the research in Ġharāʾib
al-Luġhāt.
Ġharāʾib al-Luġhāt is of historical interest as an old dictionary of the
vernacular but it is not a particularly sophisticated work. For the editor of
the published edition of Nawādir al-Alfāz̤ , Sayyid ʿAbdullah, the primary
explanation for the difference in depth between Ġharāʾib al-Luġhāt and
Nawādir al-Alfāz̤ is that ʿAbd al-Wāsiʿ was a schoolmaster while Ārzū
was a philologist writing for other scholars. Unfortunately, we know little
about ʿAbd al-Wāsiʿ because services to education did not merit inclusion
in poetic tażkirahs of the time.⁷⁴ He composed popular educational texts,
including a nis: āb [a rhyming, multilingual dictionary for schoolchil-
dren], some Persian grammar books, and at least two poetic commen-
taries [sharh: pl. shurūh].⁷⁵
: Indeed, he wrote what appears to be the most
popular Persian grammar of its time in India.⁷⁶ As his name implies, he is
connected with the town of Hansi in present-day Haryana (about 130 km
northwest of Delhi), and Ārzū occasionally hints that his usage is pro-
vincial. Ġharāʾib al-Luġhāt itself is undated and the four manuscripts
I have consulted were copied decades too late to give any indication of
the original date of composition.⁷⁷

⁷⁴ NA 1951, iv. On ʿAbdul Wāsiʿ see Hakala 2016, ch. 2; Dudney 2019a.
⁷⁵ On ʿAbdul Wāsiʿ’s Nis:āb-i Sih Zabān [Nis:āb of Three Languages] see ʿAbdullah 1965,
92–3. In the Aligarh Muslim University Library, his commentary on the Bustān of Saʿdī is ms
J Per. 301 and two copies of his commentary on Yūsuf and Zulaik: ha are mss J Per. 240 and J Per.
302. On some important nis: ābs, see Baevskii 2007, 101, 123–4. The most famous Indian nis: āb,
:
Khāliq :
Bārī, is attributed to Amīr Khusrau but this has generated controversy. Internal evidence
:
suggests it was written centuries after his death by someone else named Khusrau, possibly in the
tenth/sixteenth century (NA 1951, ii). However, more recent scholars have argued that there is
:
no compelling evidence against Amīr Khusrau’s being the original author and that anachron-
istic features of the text are interpolations (Hakala 2010, 259ff.).
⁷⁶ The preface notes that it was based on Farhang-i Surūrī, Farhang-i Jahāngīrī, and
Farhang-i Rashīdī. It was published by the Naval Kishore press (Risālah-yi ʿAbd al-Wāsiʿ
1862).
⁷⁷ Rampur ms 2543 gives 1205/1790 and ms 2544 gives 1281/1864. The oldest ms in
existence appears to be from 1159/1746, which is still far too late (NA 1951, xliii). The undated
manuscript in the Cambridge University Library (Eton Pote 291) came into the collection of
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

    167

Another text, the late seventeenth-century Tuhfat : al-Hind [“A Gift

from India”] by Mīrzā Khān,: is worth mentioning in the context of
Nawādir al-Alfāz̤ .⁷⁸ It is notable because it appears to be the only
Persian treatise on hindī grammar from the pre-colonial period.⁷⁹
:
Some manuscripts include an appendix [khātimah], which is a diction-
ary about the same length as the rest of the work.⁸⁰ The appendix is one
of the oldest lexicons of an Indic language in Persian (other than nisābs, :
which were, after all, not dictionaries but rather hindī-medium teaching
aids for Arabic and Persian).⁸¹ It is of little interest for us as a dictionary
because although it has some 3,000 entries, most are just a spelling
followed by a single-word Persian gloss. A comparison of a number of
entries suggests that none of the definitions match up with any in

E. E. Pote, a colonial officer, before 1788 (Margoliouth 1904, 2–6) and is probably not much
older than that since it has the Swiss adventurer Antoine Polier’s seal (under his Mughal title
Arsalān Jang) dated 1196/1782.
⁷⁸ The dating is uncertain, as is the patron and indeed the author, who is otherwise
unattested. Rieu’s British Museum catalogue claims that the patron was Jahāndār Shāh and
that the text was written before 1676 (1879–83, 1:62). William Irvine argues that Kukaltāsh
:
Khān, Jahāndār Shāh’s foster brother, was the patron and that the work must have been
completed between 1695 and 1706 (Irvine 1898). Ziauddin’s editor’s preface and the corrigenda
to the Bodleian catalogue agree that it was written for Aurangzeb’s eldest son, Prince
:
Muhammad Aʿz̤ am Shāh (Ziauddin 1935, 2–3; Beeston 1954, 102). That seems best supported
by the evidence and would put the date before 1707 when Aʿz̤ am Shāh was killed. I have seen a
manuscript in the Aligarh Centre of Advanced Study in History (ms 67 formerly Univ. Coll. 98),
:
which clearly lists both Kukaltāsh Khān and Jahāndār Shāh as dedicatees. It seems most likely
that these were later insertions. Another possibility is that somehow two texts written for
different patrons were merged (see Bhatia 1987, 17–21), but this is presently speculation and
is rendered unlikely by the fact that the whole work appears to use the same solipsistic system
for describing Indic sounds.
⁷⁹ Ziauddin 1935, 8; cf. Bhatia 1987, 21. A very short section of it is published with a
translation and some useful context as Grammar of the Braj Bhakha 1935. I have consulted
the two Bodleian manuscripts as well as an edition published in Tehran (Tuhfat : al-Hind 1975).
The author does not identify the part of the work dealing with phonology and grammar (the
introduction) or the work as a whole with any of the traditional Arabic linguistic disciplines
(Rampur ms 2543 ff. 2a–3b). I am therefore using “grammar” as a shorthand description of the
work’s contents rather than defining its genre as “s: arf wa nahw,” : the usual Perso-Arabic term
for “grammar.”
⁸⁰ For example, Bodleian ms Elliott 383. The complete appendix (a lexicon of Indic words)
has been published as Tuhfat: al-Hind 1983. Although unpublished before the twentieth century,
it attracted the attention of Sir William Jones who drew on it extensively for his essay “On the
Musical Modes of the Hindus” (this fact is curiously not mentioned in Zon 2006).
⁸¹ The Sharh-i : Sundar Singār [Commentary on Sundar], a lexicon of the work of the Braj
poet Sundar, is several decades older. The India Office Library’s copy was compiled in the
nineteenth century by Garcin de Tassy from a glossary originally dated 1686 /1636  (Ethé
1903, 1:1538). It describes the unvoiced retroflex stop [t:] as “tā-yi hindī” [Indian “t”] meaning
that its transliteration system is identical to or close to Ārzū’s and unlike that of Tuhfat
: al-Hind.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

168       

Ġharāʾib al-Luġhāt or Nawādir al-Alfāz̤ . Furthermore, its system of

describing Indic sounds is unique.⁸² It therefore appears not to have
had any direct influence on either Ġharāʾib al-Luġhāt or Nawādir al-
Alfāz̤ .
As in the case of the more mainstream Persian dictionaries, Nawādir
al-Alfāz̤ and Ġharāʾib al-Luġhāt are not dictionaries in the modern sense
of being comprehensive and general-purpose lexicons. Even though they
are crossing linguistic boundaries, their contents are selective and are
meant to serve as a tool for literary composition and interpretation. The
preface of Nawādir al-Alfāz̤ states (in relation to Ġharāʾib al-Luġhāt’s
purpose) that the goal was to define “Indic words that people of the lesser
provinces use rather than the Persian, Arabic or Turkish” synonyms.⁸³
Both Ġharāʾib al-Luġhāt and Nawādir al-Alfāz̤ spell out Arabic and
Persian synonyms, implying that their purpose is as much about helping
the reader build his Persian and Arabic vocabulary as it is about defining
Indic words, and Nawādir al-Alfāz̤ frequently quotes Arabic and Persian
dictionaries, including Ārzū’s own Sīraj al-Luġhat. ʿAbd al-Wāsiʿ only
rarely quotes dictionaries.⁸⁴ Indeed, many of Ārzū’s corrections have to
do with ʿAbd al-Wāsiʿ’s Perso-Arabic synonyms rather than with any
doubt about the Indic word being defined. For example, Ārzū rejects
ʿAbd al-Wāsiʿ’s inclusion of takiyā [pillow], arguing correctly that it is
itself an Arabic word (and therefore out of place in a dictionary of Indic
:
words), and then gives the hindī synonym, geṅdwā.⁸⁵ Although numerous

⁸² Tuhfat
: al-Hind’s transliteration scheme is an awkward admixture of Arabic terminology
for degrees of “heaviness.” It is described at Ziauddin 1935, 11–12. The terminology used in
Ġharāʾib al-Luġhāt and NA, as well as Ārzū’s other works, is both simpler and more accurate.
A fuller account of Indo-Persian schemes for representing Indic sounds is given in Chapter 2.
⁸³ “luġhāt-i hindī kih fārsī yā ʿarabī yā turkī-yi ān zabān-zad-i ahl-i diyār-i kamtar būd” (NA
1951, 3). Walter Hakala’s analysis of NA is illuminating (Hakala 2016, 51–63). Neither Ġharāʾib
al-Luġhāt nor NA appears in the most comprehensive list of Persian dictionaries compiled in
South Asia (Naqvi 1962, 333ff.). Oddly, Ārzū’s description of Ġharāʾib al-Luġhāt’s function is
more clearly articulated than ʿAbd al-Wāsiʿ’s own vague explanation. Ġharāʾib al-Luġhāt’s
preface is excerpted in NA 1951, iv. I have compared it to Rampur mss 2543 and 2544.
Hakala graciously provided photographs that made this comparison possible.
⁸⁴ For example, for “rāʾitā” he cites a variant from Farhang-i Jahāngīrī (NA 1951, 260). As far
as I have been able to establish, he has only used that dictionary, which is also frequently cited by
Ārzū in M.
⁸⁵ NA 1951, 149–50. Platts and the Oxford Hindi-English Dictionary (McGregor 1993) both
give this word with dental “d” rather than retroflex “d” : as Ārzū has. Another example is the
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

    169

Indic concepts are described in Nawādir al-Alfāz̤ (such as the months), it is

not primarily a dictionary of Indian cultural practices. For example, the
word Diwālī (the Hindu festival) is carefully defined in Mus̄mir but does
not appear in Nawādir al-Alfāz̤ itself, as it certainly would have if Ārzū had
thought of the work as a lexicon of Indian traditions.⁸⁶
Besides adding entries and correcting ʿAbd al-Wāsiʿ’s Arabic and
Persian, Ārzū has made other fundamental improvements. Since
Ġharāʾib al-Luġhāt’s entries are grouped into chapters by first letter but
are randomly arranged within each chapter, it is difficult to locate
particular words. Ārzū himself notes this shortcoming in the preface to
Nawādir al-Alfāz̤ , and has organized Nawādir al-Alfāz̤ according to the
more usable system of chapter by first letter and sub-chapter by the
second letter. Also, Ārzū incorporates a great many learned sources and
highlights ʿAbd al-Wāsiʿ’s lack of research in matters of practical know-
ledge, such as zoology or botany—or more precisely, how words for
animals and plants had been used in literature.⁸⁷ Furthermore, with the
possibilities opened up by the concept of tawāfuq, Nawādir al-Alfāz̤ is
able to make much more sophisticated observations about language than
Ġharāʾib al-Luġhāt. By acknowledging that Persian and Sanskrit are
related, Ārzū can discuss the origins of words and trace their meanings
through history.
While Nawādir al-Alfāz̤ is widely available in manuscript, the unre-
vised Ġharāʾib al-Luġhāt is rare.⁸⁸ As Walter Hakala has argued, Ārzū

entry for “tikkā” which Ārzū notes is just a misspelling of Arabic and Persian “tikkah” [small
piece, as in a meat dish] (NA 1951, 150). For “t:hag” [robber], Ārzū corrects ʿAbd al-Wāsiʿ’s gloss
“mushtag” to “mushtang” [robber] (NA 1951, 164).
⁸⁶ M 1991, 174.
⁸⁷ For example, ʿAbdul Wāsiʿ tries to identify the papīhā [a kind of cuckoo], which “is a small,
sweet-voice bird” [murġhī ast kūchak wa k: hwush
: āwāz], with the shuk: hish bird even though the
: naz̤ ar ast] (NA 1951, 105). The problem,
papīhā itself is not found in Persia [dar wilāyat mahl-i
as Ārzū notes, is that there are many species of “small, sweet-voiced” birds that can be
analogized to the papīhā. Ārzū also throws in a learned reference to the tenth-century 
poet Rūdakī’s use of shuk: hish. He concludes that the papīhā is actually a s: aʿwah [finch]. See also
the entry on “totā” [a kind of parrot], which includes a discourse on which birds are represented
in poetry as eating sugar (NA 1951, 156; cf. SL f. 201a on “totī” noting that the hindī word is
“totah” [sic]).
⁸⁸ The only manuscript of Ġharāʾib al-Luġhāt in a Western collection (and that is not NA
under the wrong title) appears to be Cambridge University Library ms Eton Pote 291. On other
mss., see NA 1951, xliii ff. The published edition of Nawādir al-Alfāz̤ is unfortunately not
satisfactory for the purpose of determining exactly what Ārzū added to the original work. The
problem is that it uses a symbol “[=]” to indicate, according to an editor’s note, that a particular
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

170       

effectively writes ʿAbd al-Wāsiʿ out of the tradition.⁸⁹ Ārzū’s attention to

detail and philological questions is clear, to choose one example, from
the entries for “ajwāʾin.” Ġharāʾib al-Luġhāt in one recension has the
following:

It is a seed that is mixed into bread and baked during the cold period
and is useful for ending flatulence. In Persian: Aniseed.⁹⁰

And in another:

It is the name of a seed: Aniseed.⁹¹

And now Ārzū’s definition:

The name of the well-known seed, aniseed—both nānk: hwāh and

zinnyān (spelled “zi-nn-yā-n”)—and this word is actually cognate
[mushtarak] in Persian and hindī although in Persian it is “jiwānī’”
and “jiwāʾin” with the same meaning. On reflection, it is no secret that
the source is the hindī word because people have written it in the
meaning “life-making.”⁹²

An entry like this demonstrates that the focus is entirely different in the
two works. Ārzū gives two synonyms—it is a “well-known seed” so there
is no reason to define it any more carefully than that or indeed to

entry in Nawādir al-Alfāz̤ incorporates ʿAbd al-Wāsiʿ’s full definition (NA 1951, 5). But upon
consulting manuscripts of Ġharāʾib al-Luġhāt, it becomes clear that a number of the entries
marked “[=]” either do not exist in the earlier work or are in fact completely different.
⁸⁹ Hakala 2016, 59.
⁹⁰ “tuk: hmī bāshad kih ānrā bar rūʾī nān rek: htah pazand dar dafaʿ-i burūdat wa nafk: h ba-
ġhāyat mufīd ast b.f. [=bah fārsī] nān-k: hwāh”
: (Rampur Ġharāʾib al-Luġhāt f. 12). Steingass
defines the synonym nān-k: hwāh : as follows, “Aniseed (in some places it seems to mean caraway-
seed), which frequently is baked in bread on account of its flavour and stomachic qualities;
bishop’s weed; one who begs his bread.”
⁹¹ “nām-i dānah-yīst nān-k: hwāh”
: (Cambridge Ġharāʾib al-Luġhāt f. 6b).
⁹² “nām-i dānah-yi mashhūr nānk: hwāh wa zinnyān ba-kasrah zāʾī maʿjmah wa tashdīd nūn
:
wa tahtānī ba-ālif-i kashīdah wa nūn wa īn lafz̤ nīz mushtarak ast dar fārsī wa hindī balkih dar
fārsī ‘jiwānī’ wa ‘jiwāʾin’ badīn maʿnī āmadah bar mutaʾammil poshīdah nīst kih as:l lafz̤ -i hindī
ast chirākah bah maʿnī-yi zindah kunandah niwishtah-and” (NA 1951, 10).
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

    171

mention indigestion—but then he makes a linguistic argument that the

Persian word is originally Indic because it is related to the Indic word for
life, “jīwan.”⁹³ This sort of reclamation of an Arabic or Persian word as
Indic appears across scores of entries, such as on pān [the betel leaf
chewed in India], both of whose “Persian” synonyms, namely tanbūl and
tānbūl, are originally, in fact Indic [“har do lafz̤ dar asl : hindī ast”].⁹⁴
Furthermore, references to tawāfuq frequently appear and they are often
accompanied by a statement to the effect of “as I have noted in Sirāj al-
Luġhat.”⁹⁵ Ārzū’s project in Nawādir al-Alfāz̤ is therefore fundamentally
linked to the project in his Persian lexicographical works. One such
example, a lexicographical tour-de-force, is “tan sukh” (defined as a
wondrous and rare thing, especially cloth).⁹⁶ Ārzū traces phonetic and
semantic variations through hindī, Arabic, and Persian. He makes refer-
ence to the fact that “dictionaries note” [dar kutub-i luġhāt marqūm ast]
that in both hindī and Persian “tan” means “body” and “sukh” means
“happy,” but that it can particularly mean a kind of fine cloth from
Bengal.⁹⁷
The best example of Ārzū’s attention to language at its most funda-
mental is the entry for “ast,” which does not appear in Ġharāʾib al-
Luġhāt and for which there is absolutely no reason to provide a definition
except to make a case about language:

⁹³ In fact, it seems more likely (as Platts notes) that it is related to Sanskrit “yawānī” which
refers to a similar plant. Just as in European studies of etymology at this time, Ārzū mostly
depends on his intuition to match the forms of words and is working with only of a small set of
sound changes compared to the full panoply now available to historical linguists. That his
intuition is frequently right by modern standards is a testament to his intellect and erudition.
⁹⁴ NA 1951, 104.
⁹⁵ See, for example, the entry on “kes” [i.e., kesh, hair] which he connects by tawāfuq with the
Persian gesū [lock of hair] (NA 1951, 358).
⁹⁶ NA 1951, 153.
⁹⁷ This requires us to speculate as to which dictionaries he means in relation to hindī.
Presumably he is referring to the kind of literary manuals available in Braj Bhās: ā or Sanskrit,
such as “the grammar books of the Indians” [kutub-i nahw : wa s: arf-i hindiyān] mentioned in
Mus̄mir (M 173). It does not appear, however, that he mentions any of these by title. Allison
Busch has noted that in Abū al-Fazl’s
chapter on Indian literature [sāhitya], he instructs his
interested readers to consult “works on this subject” implying that there was a corpus of
reference materials in Persian or perhaps in Braj itself (Busch 2010, 284). Nonetheless, Ārzū
might simply be referring to Persian dictionaries since these do contain stray references to Indic
words.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

172       

“Ast” is connected with “ast” [i.e., “is”] as proven by the Persian “nāst”
[i.e., variant of “is not”] with “n” as the first [letter] like “nāstik”⁹⁸
meaning “atheist” and “denier of God” in hindī because of linguistic
concordance [tawāfuq-i lisānain], and “hast” [i.e., emphatic “is”] in
Persian is a variant of “ast” because the “a” in both languages has
changed into “h.”⁹⁹

While entries like “ast” point to a philosophical project of describing the

nature of language, Ārzū is also obsessed with observed details. He has an
astonishingly precise entry on chhatrī, which ʿAbd al-Wāsiʿ has defined
both as a trellis [baram] for growing vegetables and a particular kind of
bird perch, a meaning which, according to Ārzū, it never has in hindī
[“dar hindī īn rā chhatrī na-goyand”].¹⁰⁰ He goes on to define its
construction precisely as a kind of wooden frame made of small pieces
of wood lashed together. It does not matter whether you put pigeons or
vegetables on it, but the key to the definition is the way in which it is
constructed. And, he helpfully adds, if you seat birds of prey on the
perch, then it is called “patwāz.” Similar erudition is on display in his
discussion of chapātī [flat-bread], which he turns from a common
foodstuff into a historical concept.

[As defined] in the treatise [i.e., Ġharāʾib al-Luġhāt]: “thin bread

cooked on a pan with the hand.” Originally “chapātī” was a Persian
word; chapāt in Persian is a slap and since it is cooked by smacking
with the palm, they call thin cakes chapātī. However, in credible books
“chapātī” means unleavened bread that is cooked with “chapāt” or
spread hands. Some say the “alif” [i.e., the long “ā”] fulfills a metrical
requirement in poetry but it is originally chapātī [i.e., with the long “ā”
already in the word instead of added to fill out the meter in a poem];

⁹⁸ This same Sanskrit-derived word is also adduced as evidence in M (1991, 213). Ārzū notes
that Burhān-i Qātiʿ incorrectly defines it as a particular historical person. Ārzū correctly identifies
it as a philosophy.
⁹⁹ “ast rābt̤ -i kalām ‘ast’ dar is̄bāt chunānkih dar fārsī ‘nāst’ ba-nūn muqābil-i ān lihāżā
nāstik ba-maʿnī-yi nāfī wa munkir-i k: hudā ast dar hindī pas az tawāfuq-i lisānain bāshad wa
‘hast’ dar fārsī mubaddal-i ‘ast’ bāshad chirākih alif ba-har do zabān mubaddal bah hā shawad”
(NA 1951, 23).
¹⁰⁰ NA 1951, 198.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

    173

most likely when they refer to thin cakes in this sense [i.e., cooked with
spread hands], these are not chapātī in the familiar meaning.¹⁰¹

Another such example is lāt:h, which Ārzū defines in the general meaning
of a wooden or stone pestle in hindī but notes that in Delhi it refers
specifically to the stone columns on two buildings constructed by the
fourteenth-century ruler Firoz Shāh.¹⁰² As in the example of lāt:h, careful
observation often reveals a stark contrast between a broad, common
[ʿāmm] meaning and a specific [k: hās: s: ] meaning. The gap presents an
obstacle for either understanding the hindī word or providing an accur-
ate Persian gloss for it.¹⁰³ In the entry for “dan
: d”
: [fine, penalty], for
example, Ārzū constructs a historical argument demonstrating that ʿAbd
al-Wāsiʿ has chosen an overly specific kind of fine in Persian to translate
“dan
: d,”
: which is a more general concept of punishment.¹⁰⁴
Ārzū relies on two different kinds of data, written sources and per-
sonal observations. Although the works cited are from the mainstream
Persian tradition, he considers these texts able to shed light on Indic
concepts. He uses Persian dictionaries (primarily his own Sirāj al-
Luġhat, but with reference to others, especially Burhān-i Qāt̤ iʿ,
Farhang-i Jahāngīrī, and Farhang-i Surūrī), “credible books” [kutub-i
muʿtabarah], and some important Arabic lexicons.¹⁰⁵ He also refers to
technical works like Imām Damīrī’s Hayāt

al-Haiwān

[Animal Life, 773/
1371].¹⁰⁶ Sometimes he quotes Persian poets, such as Rūdakī or Saʿdī.¹⁰⁷

¹⁰¹ NA 1951, 194. Thanks to Prashant Keshavmurthy for offering his exegesis on the humble
chapātī.
¹⁰² ʿAbd al-Wāsiʿ had defined it apparently incorrectly as a structural beam. The word in fact
refers to the two Ashokan pillars brought by Firoz Shāh to Delhi from Meerut and Topra
(Haryana). I thank Zirwat Chowdhury for clarifying the reference. The text says: “A huge and
long stone which had been mounted on two of the buildings of Feroz Shāh in Delhi they call
‘lāt:h’ ” [sang-i kalānī darāzī rā kih bar do ʿimārat az ʿimārāt-i sult̤ ān feroz shāhī dar dihlī nas: b
kardah būdand ān rā lāt:h k: hwānand]
: (NA 1951, 388).
¹⁰³ See the editor’s remarks on NA 1951, vii. ¹⁰⁴ NA 1951, 253.
¹⁰⁵ The phrase “credible books” appears, for example, in the fascinating entry on chaudharī
(NA 1951, 217). The Arabic dictionaries include al-Qāmūs [The Ocean, fourteenth–fifteenth
century ], Muntak: hab al-Luġhāt [Selection of Words, 1046/1636], and Kanz al-Luġhāt
[Treasure of Words, ninth century ] (see Rieu 1879–83, 2:503, 510–1). All three dictionaries
are cited together in, of all things, the entry on panīrwālā [cheese-monger] (NA 1951, 121–2).
¹⁰⁶ NA 1951, 156. This particular text, written in Arabic by Muhammad : Kamāl al-Dīn al-
Damīrī (d. 808/1405), was translated into Persian by Shaikh : Mubārak (the father of Faizī
and
Abū al-Fazl).
See Hadi 1995, 362; al-Damīrī’s death date is from Kopf 2012.
¹⁰⁷ NA 1951, 106, 199.
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

174       

Other entries, especially ones in which Ārzū’s objection hinges on ʿAbd

al-Wāsiʿ’s usage in hindī, depend on personal observation. For example,
ʿAbd al-Wāsiʿ gives “chhanīl” as a headword and Ārzū sputters “no one
knows where this word comes from! We who are among the people of
India and who are in the Royal Court [i.e., Delhi] have never heard
[it]!”¹⁰⁸ He frequently cites himself as a source. For example, in the entry
on chhatrī, he writes, “I have not heard” [na-shinīdah-am] the word used
in a particular meaning. There are a number of entries in which Ārzū
directly criticizes ʿAbd al-Wāsiʿ for his Haryana dialect.¹⁰⁹ One withering
example appears for “gupchup.” Ārzū writes,

But what is known as “gupchup” to the eloquent has the meaning of a

delicate sweet, eating which one is struck dumb; in the meaning given
[by ʿAbd al-Wāsiʿ], it is perhaps the usage of the compiler’s own locale
[wat̤ an].¹¹⁰

The usages appearing in standard texts, such as Persian dictionaries, and

the knowledge of people in the court, including Ārzū himself, necessarily
trump the definitions offered by ʿAbd al-Wāsiʿ. This superiority of the
metropole can be fruitfully connected to the discussion of fas:āhat :
[linguistic purity] in Mus̄mir, namely that members of the courtly elite
refine a language by pruning local usages. The refined language, although
originally the language of a place, becomes a translocal literary standard.
Furthermore, since Ārzū conceives of Persian as being originally
anchored to the royal court [urdū-yi muʿalla̍] but then available in a
standard form across the Persianate world, perhaps the vernacular liter-
ary practices of the royal court would have had the same kind of
portability.
Ārzū’s lexicographical method, which should be counted as a break-
through in Indo-Persian intellectual history, fused a sophisticated

¹⁰⁸ “malʿūm nīst kih luġhat-i kujā ast; mā mardum kih az ahl-i hind-īm wa dar urdū-yi
muʿalla̍ mī bāshīm na-shinīdah-īm” (NA 1951, 214).
¹⁰⁹ The editor gives references for several such entries (NA 1951, ix).
¹¹⁰ “lekin ānchih gupchup mashhūr-i fus: ahā-st
: bah maʿnī-yi shīrīnī ast nāzuk kih ba-k: hw-
:
urdan-i ān āwāz-i dahan bar nayāyad, bah maʿnī kih āwardah shāyad mustaʿmal-i wat̤ an-i
mus: annif bāshad” (NA 1951, 363).
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

    175

historical understanding of language and literature with keen cultural

observations. It is important to remember that Ārzū was not a modern
anthropologist content to record usages as they appear in society, but
rather was making an intervention in the language—one that intermixed
things as they were with things as he thought they should be. The process
of language standardization is inevitably about exercising power, because a
standard is defined by those with the power to define a standard. Of course,
by “power” in the pre-modern context we refer both to the instrumental
exercise of it like chancellery directives on proper usage—which is familiar
in modern language planning theory—and more importantly to the pres-
tige that causes one’s usage to be thought worth following—as Ārzū argues,
poets are constantly creating sanad for later generations. It would be
anachronistic for us to criticize Ārzū for being an elitist (in the sense of
narrowing the criteria for writing good Urdu and limiting it to a small
group of practitioners) because, of course, much intellectual history is the
study of the inner lives of the literate elite of a society.¹¹¹
Tracing his influence up to the present day through dictionaries and
other critical works remains to be done. We can say that Nawādir al-
Alfāz̤ does not appear to have figured into Urdu dictionaries compiled
later by Indians but the colonial state used it.¹¹² In particular, the words
and definitions in Nawādir al-Alfāz̤ could be fruitfully compared with
later Hindi and Urdu with attention to both geographical and temporal
variation in usage, but that is a project whose scope far exceeds my
present aims. It is difficult to assess the correctness of Ārzū’s definitions
by contemporary standards because the language had been thoroughly
transformed by the efforts of indigenous and colonial language reformers
before John T. Platts compiled the now standard Dictionary of Urdu,
Classical Hindi, and English in 1884. When one of Ārzū’s definitions
seems a little too pedantic, there is little we can do. For example, dādā is
defined by ʿAbd al-Wāsiʿ as “paternal grandfather,” a meaning shared in
modern Hindi/Urdu, but Ārzū corrects that to “maternal and paternal
grandfather” [jadd-i mādarī wa pidarī] on the basis of the dictionaries

¹¹¹ Cf. LaCapra 1985, 79.

¹¹² However, an 1825 medical vocabulary draws upon NA and Ġharāʾib al-Luġhāt, the two
sources other than Gilchrist and Shakespear listed for “Hindee” (A Vocabulary). Hakala has also
discovered that NA was used by H. M. Elliot (Hakala 2016, 65ff.).
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

176       

al-Qāmūs (Arabic) and Burhān-i Qāt̤ iʿ (Persian).¹¹³ In this case it is not

clear whether Ārzū’s meaning is actually current or whether he had fallen
prey to the malaise common among lexicographers that was noted by the
German classicist Christian Lobeck (1781–1860): “Who among us does
not have his own Utopia? And where can we rest from the crush of
everyday concerns but in the ethereal land where Etymology reigns?”¹¹⁴

Observations on Indian Religion as

“Proto-Anthropology”

How do the eighteenth-century lexicographers describe Indian religious

practice and philosophy? In the texts under discussion, the modern
understanding of Hinduism as one of the “world religions” does not
appear.¹¹⁵ Nonetheless, there is a concept of community [qaum], broadly
or narrowly defined, which has certain practices and beliefs; there is also
a sense of the sacred as a distinct aspect of human existence, especially in
terms of rituals, festivals, and geography. An obvious first step is to ask
how these sources define who is a Hindu, and the answer is that Hindus
are not Muslims. Ārzū contrasts “mardum-i hindūstān” [the people of
India] with “hindiyān” [Indians] but it is unclear whether by the latter he
means Hindus in a religious communitarian sense (of course with the
caveat that this is not necessarily in the same way we would understand
the term today) or if he is contrasting people resident in India with
native-born Indians including Muslims.¹¹⁶ Bahār, citing Ārzū, is however
explicit on this point:

:
Khan-i Ārzū says that “hindū” is a particular community and thus it
cannot refer to Muslims who are resident¹¹⁷ in this country, and the

¹¹³ NA 1951, 232.

¹¹⁴ “quisnam nostrum non suam Utopiam habet? aut ubi tandem a turba quotidianarum
rerum requiescere possumus nisi in illa aetherea regione ubi Etymologia dominatur?” (quoted in
Allen 1948, 60).
¹¹⁵ The standard account of how Orientalists conceived of Hinduism as a religion rather than
a belief system is King 1999.
¹¹⁶ CH 1984, 1156.
¹¹⁷ When dealing with Persianate pre-modernity, we need to rethink our concept of what it
means to be of or settled in a place. While the term “sākin,” especially in its literal English
translation “resident,” implies a temporary state of affairs, it is important to note that it does not
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

    177

correct term for a resident of India is “hindī”; this is supported by the

writing of many because it is most preferable.¹¹⁸

The term “hindī” is also defined as referring to a wide range of Indian

things, notably a kind of sword.¹¹⁹ The poetic trope of the Hindu as
heart-stealer or slave was still very much in force and thus before
introducing the straightforward definition of Hindus as a community,
Bahār mentions that “Hindu” has been used to mean “thief,” “shepherd,”
“slave,” “infidel,” and “resident of India.”¹²⁰ The literary tradition in
whose service these dictionaries have been compiled—and Sufi thought
itself—has reciprocity built into it: “dīn” [faith] is one pole and the other
is “kufr” [unbelief].¹²¹ In Mirʾāt al-Is: t̤ ilāh,
: faith is delightfully compared
to “bādām-i dū-maġhz” [lit. an almond of two marrows], which refers to
the occasional oversized almond containing two piths. The poet Shaukat
:
Bukhāri (d. ca. 1695) has written: “Unbelief and Islam have a common
root; / A double-almond does not have two flowers.”¹²² There is symbolic
movement across these categories—kufr can be transgressively valorized
in literature—but of course, India and its non-Muslim inhabitants have
generally been associated with kufr in literature, especially with black
: literally means a “liver-eater” and this
magic. For example, jigar-k: hwār

necessarily have this connotation in Persian. The opposite problem in interpretation exists with
the term “wat̤ an,” often translated as “homeland,” but since people can “become settled”
[tawat̤ t̤un] this translation is sometimes misleading. Consider the poet Ahsanī,: whose “ances-
tors chose to settle in Gwalior” [ajdād-ash tawat̤ t̤ un gwāliyār ik: htiyār namūdah] (MN 2005, 49).
One’s native place is “wat̤ an-i as: lī” (Tavakoli-Targhi 2001, 14). The usage of “millat” as “nation”
and “wat̤ an” as “fatherland” is a nineteenth-century development (Parsinejad 2003, 129).
Wat̤ an also has a technical sense in pre-modern Indian polity (namely as heritable rights
which generally not even the king could remove), which need not concern us here except as
further evidence that the term has various meanings (Perlin 1985, 452ff.).
¹¹⁸ “k: hān-i ārzū mī farmāyand kih hindū qaumī-yi mak: hs:ūs: wa lihāżā bar musalmānānī kih
sākin-i īn mulk-and it̤ lāq-i ān namī tawān kard wa s: ahī : h: bah maʿnī-yi sākin-i hind hindī ast wa
ānkih az kalām-i baʿzī
mustafād mī shawad binābar taġhlīb ast. hindūān jamʿ wa buland az
s: ifāt-i ū ast” (Bahār-i ʿAjam 2001, 3:2139).
¹¹⁹ E.g., Farhang-i Jahāngīrī 1975, 2:1915; Bahār-i ʿAjam 2001, 3:2140.
¹²⁰ “gāh bar duzd it̤ lāq kunand wa gāh bar pāsbān wa gāh bar ġhulām wa ʿabīd wa gāh bar
kāfir wa gāh bar sākin-i hind” (Bahār-i ʿAjam 2001, 3:2139).
¹²¹ A relevant definition in Farhang-i Jahāngīrī for dīn is “a path and way and a faith and a
rite” [rāh wa rawish wa kesh wa ʿādat] (1975, 2:2283). Notably, Abū al-Fazl
uses the same word
in the context of Hinduism, referring to people outside the four castes as being “outside of this
religion” [bīrūn azīn dīn] (Āʾīn-i Akbarī 1869, 55).
¹²² “dārad yak as: l kufr-o islām / bādām-i dū-maġhz rā dū-gul nīst” (quoted in Mirʾāt al-
Is: t̤ ilāh: 1850, 70).
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

178       

refers to a witch who consumes the livers of children with her eyes in
order to steal their vitality. As Indic synonyms, Ārzū offers dāyan
: [witch]
and rather more obscurely bhatī, : and argues that although various
Iranian writers (including Sāʾib)

have referred to this practice, it is
specifically Indian. Saʾib

writes: “[The beloved’s] gaze is a running sore
in the wound of my heart / Her dark eye is a Hindu jigar-kh : wār.”¹²³
: This
tropology is well known so we should turn to definitions that seem to
stand apart from the tropes.
It is easy to compile a list of carefully described religious practices and
concepts in these texts: Ārzū discusses “arhant,” “mahīshwar,” and
“birhamā” in the context of correcting earlier dictionaries that had
badly mangled the pronunciation or definition of these words.¹²⁴
“Arhant,” for example, has often been written “arhaft” because the letters
“n” and “f” look similar in Perso-Arabic script. Iranian poets lacked
cultural context so “they have made many mistakes regarding the mean-
ings of these words” [dar miyān-i maʿānī-yi īn alfāz̤ ġhalat̤ bisyār
kardah-and]. Ārzū sets the record straight by defining “arhant” as:
“A person who through the practice of austerities is kept out during
the transmigration of souls and has reached perfection, and this is an
expression of a group of Hindus whom they call sarāwakī [i.e., a
Buddhist or Jain].”¹²⁵ With Brahma and Shiva, he follows the familiar
Islamic pattern of referring to each of them as an angel [firishtah] rather
than a deity. Birhamā is defined simply as “an angel who is entrusted
with creation” while Mahīshwar is “an angel endowed with destruction,
and this is an expression of a different group [gurohī-yi dīgar] of Hindus”
(Maheshwar is an epithet of the god Shiva so presumably this group are
Shaivas, his devotees).¹²⁶ Of course an ethnography written from within
a living culture does not require analysis but can simply reference

¹²³ “dar zuk: hm-i dil-am rīshah-yi dawānīdah nigāh-ash / hindū-yi jigar-k: hwārī : buwad
chasm-i siyāh-ash” (CH 1984, 1071). Perhaps “bhat:ī” is connected with the “bhāt:” tribe, who
are traditionally bards.
¹²⁴ M 1991, 214.
¹²⁵ “shak: hs: ī ast kih az kas̄rat-i riyāzāt

az tanāsuk: h bāz māndah ba-as:l rasīdah bāshad wa īn
is: t̤ ilāh-i
: gurohī az hindūʾān ast kih az ānhā rā sarāwakī goyand” (M 1991, 214). On tanāsuk: h
[metempsychosis], see Walker 1991. On Indo-Persian studies of Hinduism in general, see Ernst
2003, which does not unfortunately deal with how Persian philology approaches Hinduism.
¹²⁶ Ārzū’s account of these figures is rather brief in contrast to Abū al-Fazl’s
more detailed
description (Āʾīn-i Akbarī 1869, 5).
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

    179

something, and so the analysis does not go any deeper than this. Ritual
objects are also discussed. For example, Ārzū notes in Nawādir al-Alfāz̤
that “arthī” refers to a bier used by Hindus but not by Iranian
Zoroastrians or Indian Muslims.¹²⁷ Festivals are described with preci-
sion. For example, he cites ʿAbd al-Wāsiʿ on the idea that since Holī is a
festival that Hindus [hindūʾān] have in the month of Phagun involving
the throwing of rose-water and gulāl [powdered dye] during the day and
the lighting of fires at night, it is identical to the Iranian Sadah festival.¹²⁸
In fact, Ārzū argues, they are different. Sadah is solely Iranian and Holī is
solely Indian although the dates sometimes coincidentally overlap and
both are similar to a Zoroastrian celebration. In another work, he
describes the festival of Dīwālī as a day and night celebration in which
Hindus put lamps in their homes.¹²⁹
These definitions raise two obvious questions: Firstly, is there a dif-
ference in the treatment of Zoroastrianism and Hinduism, or are they
both the non-Muslim Other? Secondly, is there any interest in doctrinal
and social differences among Hindus? Besides Brahmins, whose activities
are sometimes analogized to those of Sufis, the distinctions do not seem
to be particularly important, although Ārzū is aware of Shaivas, Jains,
and Buddhists as particular groups of Hindus (in the broad sense of
“non-Muslim Indians”). Caste is treated as equivalent to community: For
example, Bārīs, who are a torch-bearer [mashʿalchī] sub-caste, are called
“a specific community” [qaumī-yi mak: hs: ūs: ].¹³⁰ Also consider the term
“chandāl,”
: which Ārzū historicizes in Chirāġh-i Hidāyat.¹³¹ He writes
that it originally meant “the most base people” [firomāyah-tarīn mar-
dum] who were watchmen and originally pig-keepers, but also gained
some authority by becoming menial servants to the imperial court and
the nobility. Under Akbar, the “kalāl” [sic], a related group, were licensed
as a community to sell liquor. The poet Mullā T̤ uġhrā uses it to mean
watchman, which is the sense, Ārzū tells us, that it has in Kashmir. For
us, it is significant that the chandāls : are disadvantaged by religious
taboos but Ārzū’s description of them makes no reference to religion at

¹²⁷ NA 1951, 20–1. ¹²⁸ NA 1951, 445.

:
¹²⁹ M 174. He cites a verse by Muhsin Tās̄īr. ¹³⁰ NA 1951, 59.
¹³¹ CH 1984, 1081.
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

180       

all.¹³² Mentions of Shīʿah practices in Chirāġh-i Hidāyat do not seem

much different from references to particular Hindu groups—they are
noted as the practices of a community, which happens to be Muslim.¹³³
Within the philological literature, there are frequent comparisons
between Islamic practices and those of other communities. The key to
cultural study in this tradition appears to be the possibility of universal
similitude. In the European context, we can fall back on Foucault’s
observations in The Order of Things that pre-modern Western know-
ledge was constituted by finding sweeping similarities across words and
things.¹³⁴ (For us, the tendency is to divide things rather than analogize
them.) We can find evidence of a similar habit of mind in the Persianate
tradition to what Foucault has described: In Nawādir al-Alfāz̤ , the entry
for upāsanā gives simply “parastish” and “ʿibādat” (all meaning “wor-
ship”) and for upās it has “rozah” and “saum”
: (all meaning “fasting”).¹³⁵
It is noteworthy that both Persian and Arabic equivalents are given for
the Indic word. Since the Arabic and Persian meanings are seen as
functionally equivalent, adding an Indic term implies that it too is seen
as identical. Religious practices are often analogized, with Sufis and
Brahmins being unproblematically considered in the same frame. For
example, in Chirāġh-i Hidāyat, under the heading “tāʾi jauzī ba-kaf
dāshtan” [lit. to have a bit of nut in the hand], Ārzū explains that holding
a piece of nut to represent a begging bowl is a tradition of the qalandars
and faqīrs (both of whom are religious mendicants) of Iran and Turan so
that they would not have to present an empty hand to people they
meet.¹³⁶ He notes that Brahmins have the same practice in India but

¹³² The idea, presented among others by Abū al-Fazl, :

that chandāls are the offspring of a
Brahmin mother and a Shudra father does not appear in Ārzū’s definition (Āʾīn-i Akbarī 1869,
55). Perhaps this is because that particular definition is not relevant for poetic usage.
¹³³ He refers to Shiʿism as “mażhab-i imāmiyah” (CH 1984), “jarīdah” (1070), “dast-o dahan
ba-āb kashīdan” (1110), “saqīfah sāzī” which is apparently a reference to taqīyah or keeping
one’s true religious denomination hidden (1144), and “kār dast bastah kardan” which claims
that Iranian Shīʿahs who come to India begin to pray like [Sunni] Hanafīs,

1191).
¹³⁴ He writes, “Knowledge therefore consisted in relating one form of language to another
form of language; in restoring the great, unbroken plain of words and things; in making
everything speak” (Foucault 1994, 40, cf. xxii, 17, 218, 232).
¹³⁵ NA 1951, 5.
¹³⁶ CH 1984, 1055; cf. Mirʾāt al-Is: t̤ ilāh: 1850, 147, where it is noted that the “Hindu
mendicant” [hindū darwīsh] called Bābā Majnūn (whose name is of course derived from
Persian!) has a takiyah [an abode of (Islamic) mendicants].
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

    181

using betel-nut or coconut. Another example, which is on the level of

folk religion, involves divination [fāl]. One method, Ārzū tells us,
depends on finding significance in the random utterances of others,
and this is practiced particularly by women in India. In Iran, besides
this method, keys are also thrown for divination although this is not
practiced in India. The verse he quotes as evidence, by Muhsin : Taʾs̄īr,
includes both kinds of prophecy in the same line.¹³⁷
Besides observations of certain analogous rituals, we see figures from
each tradition universalized. One in particular is Adam, who is men-
:
tioned in the benediction [hamd] at the beginning of Bahār-i ʿAjam. The
context is that humankind, the “children of Adam,” are grateful to their
creator “whose kindness has been available to all.”¹³⁸ The invocation of
Adam would not surprise anyone who has read other Persian texts
written in India by non-Muslims. However, it is notable that as part of
Bahār’s traditional declaration of self-deprecation he specifically refers to
his Hinduness. In Mirʾāt al-Is: t̤ ilāh,
: which was likewise written by a non-
Muslim friend of Ārzū’s, Adam appears under the heading “qadamgāh-i
ādam” [lit. step-place of Adam], a location on the island now called Sri
Lanka:

The step-place of Adam [peace be upon him] they say is a place in

Serendip because when Lord Adam [peace be upon him] came down to
the Island of Serendip his blessed foot first touched the hill which is the
step-place, and it is well-known that because of that blessing a mine of
gems was found there.¹³⁹

There is no caveat here such as “Muslims believe . . .” but rather this is

presented as a straightforward statement of fact. The unanswerable
question then is whether Adam has been absorbed into Mukhli : s: ’s own

¹³⁷ CH 1984, “fāl-i gosh” 1180. Taʾs̄īr is a somewhat obscure Iranian poet, on whom see MN
2004, 1:309 and Tażkirah-yi Nas: rābādī.
¹³⁸ Bahār-i ʿAjam 2001, xxix.
¹³⁹ “qadamgāh-i ādam (ʿalaihi al-salām) nām-i jā-īst dar sarāndīp goyand chūn ha
ādam
: zrat
(ʿalaihi al-salām) ba-jazīrah-yi sarāndīp nuzūl farmūdand dar kūhī kih qadamgāh ast awwal pā-
yi mubārak-i īshān ba-ān rasīdah wa mashhūr ast kih ba-barakat-i ān kān-i yāqūt dar ānjā
paidā shud” (Mirʾāt al-Is:t̤ ilāh: 2013, 564). Abū al-Fazl

notes that Indic sources, which otherwise
discuss various ancient events, do not mention Adam’s fall (Āʾīn-i Akbarī 1869, 193).
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

182       

framework of belief as a universal progenitor or whether the use of

Persian enforces an Islamic framework on non-Muslim writers.
When non-Muslim figures have entered the Persianate repertoire,
they have generally been bleached of any religious significance and
made purely historical. This is true for Zoroastrian figures centuries
before and Hindu figures more recently. Thus, Ārzū mentions
Kiyomars̄ as “the first king of Persia [ʿajam],” which is how he is
presented in the Shāhnāmah. Ārzū provides a long discussion of the
history of the name (concluding that it was originally “gayomart” derived
from “speaking man” [go + mard]), but does not mention his religious
significance to Zoroastrians, namely as either the first man or the first
person to worship Ahura Mazda.¹⁴⁰ The appearance of Hindu figures
follows a similar path: Bhīma, a character from the Mahābhārata,
appears occasionally in Persian poetry as a synecdoche for a mighty
warrior (sometimes with his name incorrectly rendered as “bahīm”).¹⁴¹
No mention is made of the fact that in the Mahābhārata his father was
the wind god Vāyu and that this supernatural origin accounts for his
incredible strength. Likewise, Gāndhārī, a character from the same epic,
is thought to have possibly given her name to the city of Qandahar
(in modern-day southwestern Afghanistan).¹⁴² We must, however, be
cautious in interpreting the Sanskrit epics as having anything to do with
the other religious practices described here because Ārzū, like other
Indo-Persian writers, describes the Mahābhārata as a history [tārīk: h].
He writes that “the Indians’ history Mahābhāratah is the complete
account of the circumstances of their war.”¹⁴³ Bhīm is mentioned along-
side King Bhoja, who was a historical king in central India in the eleventh
century.¹⁴⁴ Thus, invoking such figures in poetry presents no religious
problem since they have effectively no religious content.
Returning to the question of religious practice, the willingness to
carefully describe non-Muslim practices and define parallels with Islam

¹⁴⁰ CH 1984, 1204. : writes as quoted by Ārzū.

¹⁴¹ E.g., Farrukhī
¹⁴² M 1991, 218.
¹⁴³ “tārīk: h-i mahābhāratah-i hindiyān tamām-i ahwāl-i
: jang-i īshān ast” (M 1991, 218). On
the Sanskrit epics as history and the context for their translation into Persian, see
Truschke 2012a.
¹⁴⁴ For example, Farhang-i Jahāngīrī’s entry for “bahū-o bahīm” [sic] (1975, 2:2158).
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

    183

does not imply syncretism or an abandonment of Islam as the baseline

against which other religio-cultural systems are judged.¹⁴⁵ Occasionally
there are entries such as Chirāġh-i Hidāyat’s “chirāġhān-i roz-i isfand,”
namely the Persian tradition of lighting lamps on the third day of
Farwardīn (the first month in the Zoroastrian calendar).¹⁴⁶ Ārzū declares
that it is “obviously a Zoroastrian practice that remained in Iran and
under Islam some ignorant people have maintained this practice to the
present, just like some customs of the Hindus, which some ignorant
Muslims practice here.”¹⁴⁷ So much for syncretism. Clearly in Ārzū’s
view, each community has customs proper to itself and although cus-
toms are often shared between communities, a practice with a non-
Muslim origin may be suspect. Ārzū engages in what Stephen Gregg
has called “hierarchical inclusivism,” in which someone recognizes that
there are religious systems besides one’s own but accepts them only on
the terms of one’s own religion.¹⁴⁸
The Persian philological tradition did not, at least in my interpret-
ation, have much of a stake in actual religious practice. This is parallel to
the principle that everyday language only enters into pre-modern Persian
dictionaries if it impacts poetic language because the sort of knowledge
the philological tradition deals in is textualized. However, this approach
is not the same as Orientalist preference for religious texts to be inter-
preted “from the source” as opposed to observing supposedly corrupted
folk religion¹⁴⁹ but rather textualization within the Persianate literary
tradition. Once a practice or belief is pulled into Persian literature then
other sources, whether written or based on personal observation, can be
introduced to nuance the literary interpretation. This is obvious in
Ārzū’s gloss of Mathura, a city associated with Krishna devotion. He

¹⁴⁵ A compact description (and critique) of syncretism in this context is Ernst and
Stewart 2003.
¹⁴⁶ He cites Surūrī as his reference but Steingass’s dictionary claims that the lamps are lit on
the third day of the five intercalary days added to Ābān (the eighth Zoroastrian month).
¹⁴⁷ “wa īn z̤ āhiran rasm-i majūs bāshad kih dar īrān māndah wa dar islām nīz baʿzī
juhalā ān
: dāshtah-and chunānkih bazī
rasm rā ba-hāl
rusūm-i hindūān kih baʿzī
az musalmānān-i jāhil
nīz az īnjā ba-jā ārand” (CH 1984, 1077–8).
¹⁴⁸ Gregg uses the term in the context of Swami Vivekananda (1863–1902) who employs
“a system of thought which valorises one philosophical or religious approach above others, but
that accepts the presence of spiritual or religious truths within other traditions, in relation to one’s
own tradition” (Gregg 2019, 120). I thank Anand Venkatkrishnan for introducing me to the term.
¹⁴⁹ As King 1999 describes at length.
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

184       

calls it “the name of a city which in Hindu belief is a [place] of holy sites
of the Hindus.” But what is really at stake is not this description but
rather that he has checked “histories of India written by Persians” to see
how they spelled the name, and they rendered it “matūrah.”¹⁵⁰ Is this
indifference best thought of as a special flavor of secularism?

Later Persian Lexicography in the Sub-Continent

As a coda to this discussion, I will selectively consider some dictionaries

from the nineteenth century, when colonialism had remade both the
intellectual landscape and the opportunities for patronage. The indigen-
ous tradition of literary lexicography co-existed with the colonial state’s
survey-based approach, and both methods were applied to lexicography
in Indian vernacular languages such as Urdu.¹⁵¹ Literary lexicography
reached its zenith in the Farhang-i Ānandrāj, the largest pre-twentieth-
century Persian dictionary. Farhang-i Ānandrāj was one of the import-
ant sources available to the Iranian compilers of Luġhatnāmah, which
depended heavily on Indian sources without acknowledging the signifi-
cance of this relationship to the project of defining Persian.¹⁵² The
fashioning of a modern national language for Iran took Persian out of
circulation as a cosmopolitan language because the diversity that sustains
transnational intellectual networks is by definition an obstacle to a
nationalist project.
Lexicography continued to be funded by Mughal successor states,
including by the British colonial state, as well as generally remaining a
mark of scholarly achievement. Ġhiyās̄ al-Luġhāt (1242/1826–7) com-
:
piled by the schoolteacher Muhammad Ghiyās̄ al-Dīn bin Jalāl al-Dīn bin
Sharaf al-Dīn Rāmpūrī became widely used as a student’s dictionary.¹⁵³
Several lithographed editions of the work have Chirāġh-i Hidāyat in the
margins.¹⁵⁴ Some dictionaries have been lost or nearly lost such as Miftāh:
:
al-Khazāʾin, which was completed in 1228/1813 in Delhi by Jayarām Dās

¹⁵⁰ M 1991,172. ¹⁵¹ On Urdu lexicography, see Hakala 2016.

¹⁵² Tavakoli-Targhi 2001, 106–7. ¹⁵³ Blochmann 1868, 31; Ġhiyās̄ al-Luġhāt 1842.
¹⁵⁴ Storey 1984, 49.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

    185

and now exists in the Delhi Persian collection at the British Library but
apparently nowhere else.¹⁵⁵ The last of the Persian dictionaries compiled
along traditional lines, Farhang-i Ānandrāj (1889), has approximately
50,000 entries and is therefore probably the largest dictionary compiled in
the Persianate world before the twentieth century. Besides its length, the
work is remarkable because of where it was written, namely at the edge
of the Persian-using world, in what is today the southeastern Indian state
of Andhra Pradesh. The dictionary is dedicated to the maharajah of
Vizianagaram, Ānand Gajapatī Rāj. The maharajah’s mīr munshī (chief
:
secretary) Muhammad Pādshāh bin Ġhulām Muhī : al-Dīn, known by his
pen-name Shād, collated every available Persian dictionary, including
some European sources, to produce the work.
Burhān-i Qāt̤ iʿ enjoyed a somewhat surprising renaissance in the
nineteenth century, despite its drawbacks as a lexicographical source.
The ruler of Awadh, Nawab Ġhāzī al-Dīn Haidar
Shāh, wrote—or more
likely put his name on—a dictionary called Haft Qulzum [The Seven
Seas, 1822], which was largely derived from Burhān-i Qāt̤ iʿ.¹⁵⁶ Because
Burhān-i Qāt̤ iʿ was the first major Persian dictionary to be prepared for a
published edition by Europeans (it was published in Calcutta in three
editions between 1818 and 1834), it became well known in British India
and Europe.¹⁵⁷ By far the most surprising episode is Burhān-i Qāt̤ iʿ’s role
in the career of the Persian and Urdu poet Mirzā Ġhālib (1797–1869). In
1278/1862, Ġhālib wrote and published a pamphlet called Qāt̤ iʿ-i
Burhān, which took Burhān to task for some 400 errors. Ġhālib was
not a lexicographer and argued his case ferociously on the basis of his
own opinion and taste rather than research. The work touched off
a pamphlet war late in Ġhālib’s life and, in Blochmann’s words, “it
seriously damaged his reputation as a critical scholar.”¹⁵⁸
Although there was considerable interest on the part of the British
colonial state to have a convenient general purpose English–Persian

¹⁵⁵ It apparently only survives in the Delhi Persian collection at the British Library (Delhi
Persian 554A–F).
¹⁵⁶ It was published at the royal press in Lucknow with an English title page so it was a
showpiece.
¹⁵⁷ Storey 1984, 32–3.
¹⁵⁸ Storey 1984, 34; Ramezannia 2010, 187ff. The first edition was published in Lucknow and
a second edition was printed in Delhi in 1865 under the title Dirafsh-i Kāwayānī.
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

186       

dictionary, this was not to come to fruition until the end of the nineteenth
century. The indigenous dictionaries, though their poetic focus was not well
suited for administrative purposes, were nonetheless more useful than the
highly criticized European dictionaries with their outright misunderstand-
ings. Blochmann’s critique of Johann August Vullers’s Lexicon Persico-
Latinum Etymologicum (1855) runs for a full ten pages and so negates the
dictionary’s aims of being a carefully researched, scientific dictionary.¹⁵⁹
Another major project, Captain Thomas Roebuck’s edited version of
Burhān-i Qāt̤ iʿ with appendices (1818), is full of errors both from the
source material and introduced by the editor. Burhān-i Qāt̤ iʿ was widely
corrected within the indigenous lexicographical tradition and in choosing it
as a streamlined but reasonably comprehensive source Roebuck appears
to have ignored this inconvenient fact and let the errors stand.
A dependable Persian–English dictionary did not become available until
A Comprehensive Persian–English Dictionary was published by Francis
Joseph Steingass in 1892, which remains the standard dictionary of pre-
modern New Persian in English. Like the dictionaries that came before it,
Steingass compared and digested earlier dictionaries.¹⁶⁰
The history of Iranian lexicography of the nineteenth and twentieth
centuries is outside the scope of this discussion, but the obscured con-
nections between India and Iran need to be understood. Iranian lexicog-
:
raphers like Dihkhudā depended upon research material produced in
India while meticulously placing India outside of the tradition they
considered their own. While Steingass did not apparently make use of
Farhang-i Ānandrāj, it is cited in the Luġhatnāmah. That a dictionary
composed on the southeastern coast of India in a place where the
language used by most of the population was Telugu should be the
culmination of traditional Persian lexicography is a remarkable reminder
of the vibrancy of an institution that is wrongly thought of as having died
out in South Asia decades or more earlier.

¹⁵⁹ As in Blochmann 1868, 41–51. The earliest Persian lexicon in a modern European
language (as opposed to Latin) was John Richardson’s in 1777, revised by Charles Wilkins in
1820 (Baevskii 2007, 3).
¹⁶⁰ It is primarily a revision of Francis Johnson’s 1852 A Dictionary, Persian, Arabic, and
English, which was itself an expansion of Charles Wilkins’s dictionary, which was in turn an
abridgment of an earlier dictionary by John Richardson. Steingass also heavily depended on
Vullers’s 1855 Lexicon Persico-Latinum. See Dudney 2019b, 385.
 OUP CORRECTED PROOF – FINAL, 20/2/2022, SPi

5
Building a Vernacular Culture on
the Ruins of Persian?

Hidāyat, from the time that I began to compose in rek: htah [i.e. Urdu],
The custom of Persian has departed from India.¹
Hidāyat Dihlawī (d. 1805)

Scholars and critics have generally assumed that Indians started using
Urdu more widely in the eighteenth century because Mughal politics
were in a shambles and feckless nobles could no longer fund poetry in
Persian, the language of cosmopolitan prestige.² In this interpretation,
taking vernacular literature seriously was an act of desperation, an
attempt to fill a political and cultural lacuna with a kind of “poor
man’s Persian,” namely a Persianized form of the vernacular. For exam-
ple, S. K. Chatterji has written:

The first Urdu poets, deeply moved by the manifest decay of Muslim
political power in the eighteenth century, sought to escape from a
world they did not like by taking refuge in the garden seclusion of
Persian poetry, the atmosphere of which they imported into Urdu.³

¹ “hidāyat kahā rek: htah jab se ham ne / rawāj ut:h gayā hind se fārsī kā” (quoted in Āb-i
Hayāt

1907, 111n35).
² For example, Jalibi 1984, 2:149 and Sarkar 1920, 147; Sarkar 1949, 16; cf. Syed 2012, 298.
A similar assumption exists for the Ottoman context: Steps towards modernity in the eighteenth
and early nineteenth centuries have generally been seen as synonymous with Westernization
rather than as internal developments (because the Ottomans supposedly would only have
turned towards the West if their politics were in decline). Scholars have begun to question
this tautological reasoning (e.g., Hamadeh 2004, 34ff. and İslamoğlu 2012). On the funding of
cultural activities during times of turmoil, it appears little useful scholarship has appeared since
Lopez 1959 on that issue during the Renaissance.
³ Chatterji 1942, 216; Rai 1984, 242.

:
India in the Persian World of Letters: Khān-i Ārzū among the Eighteenth-Century Philologists. Arthur Dudney,
Oxford University Press. © Arthur Dudney 2022. DOI: 10.1093/oso/9780192857415.003.0006
OUP CORRECTED PROOF – FINAL, 20/2/2022, SPi

188       

However, such escapist sentiments are not actually reflected in the

sources written by Ārzū’s circle. Another common explanation for
Urdu’s rise to prominence is, of course, nationalist sentiment, namely
that proud Indians took the initiative to write in an indigenous Indian
language rather than in “foreign” Persian. But this too is a misleading
framework because it generally reflects assumptions about citizens’ rela-
tionships with modern nation-states onto pre-modern political forma-
tions. Contemporary statements that imply that vernacular language was
part of an Indian identity have been given undue weight, while the actual
process required to develop a literary idiom and its accompanying
language ideology have not widely been considered in the context of
the eighteenth century. The Urdu literary historical tradition credits
Ārzū with having developed such a language ideology, but the relation-
ship between the ideas addressed in Ārzū’s surviving works, which are
nominally about Persian alone, and the Urdu tradition have not been
developed in detail. Here I consider how Ārzū’s ideas on language and
the members of his circle influenced the vernacular literary tradition of
Delhi.
This chapter addresses the development of what came to be called
Urdu with neither nationalist nor declinist preconceptions. Instead, my
proposition is that the standardization of vernacular poetic practice in
the mid-eighteenth century was a catalyst in building the consciousness
of a pre-existing, dispersed community of language users across India. It
is self-evident that a language grammatically similar to that of Delhi and
its hinterlands (the so-called khar: ī bolī or “upstanding speech”) had
made it across the length and breadth of South Asia well before the
eighteenth century.⁴ It was known by a number of different names and
exhibited considerable variation in vocabulary and grammar. When
poetry was composed in this dialect following Persian poetic rules, it
was known as rek: htah. By the end of the eighteenth century, rek: htah,

⁴ The exact mechanism by which khar: ī bolī spread so successfully is unclear, but it is
:
frequently connected with Muhammad bin Tuġhluq’s briefly shifting the capital of the Delhi
Sultanate to Daulatābād (near Aurangabad in present-day Maharashtra). In any case, sources
are patchy until it is a fait accompli: For example, Simon Digby’s research into a set of early
seventeenth-century texts show “heavily Persianized vocabulary” and “uncompromising Urdu
grammatical structures” even though the texts were produced far from Delhi in Gujarat or
Rajasthan (Digby 1995).
 OUP CORRECTED PROOF – FINAL, 20/2/2022, SPi

    189

which had originally referred to a poetic practice, had become synony-

mous with Urdu, a new term denoting a language which was derived
from the phrase zabān-i urdū-yi muʿalla̍ [language of the Royal Court].
As noted in Chapter 2, Ārzū used this phrase in reference to standardized
Persian, and the transference of the phrase strongly implies a parallel
between his idea of how the literary standard functions in Persian
(namely, as an urban literary language that displaces local variations as
the eloquent idiom for poetry) and a vision for the development of
vernacular poetics.
This transformation is both overdetermined and undertheorized: The
:
key intervention in this chapter is that conflating the terms rekhtah and
zabān-i urdū in this way, as most scholars have since, elides crucial
differences between literary style in the vernacular and vernacular lan-
guage itself. A prerequisite to such an analysis is contrasting the ways of
defining linguistic usage in early modernity to later definitions. In the
South Asian context, today’s divisions into languages, dialects, and other
formally bounded categories were promulgated (or at least formalized)
mostly in the colonial period.⁵ Our socio-linguistic terms were obviously
unavailable to people in the eighteenth century and we must tread
carefully in ascribing current perceptions of language usage to early
modern conditions. Let us take the most obvious example, the concept
of zabān, a term shared between Persian and hindī-Urdu. It is generally
translated as “language” but does not necessarily designate what we
mean by that word today.⁶ Instead, it has an inherent ambiguity akin
to that of its literal English translation, “tongue.” Rather than attempting
to show whether definitions of language used in the eighteenth century
were precursors to modern definitions, a better way to understand the
rise of Urdu is as a series of rhetorical acts, both overt and subtle. By
recovering the rhetoric of early Urdu through critical texts that address

⁵ These formal boundaries have been further reified by the reorganization of Indian states on
the basis of language in the late 1950s.
⁶ Formal linguistics defines a language according to “mutual intelligibility” (Campbell and
Mixco 2007, 91–2). Socio-linguistics modifies this definition to require a standardized dialect
that has been defined by a speech community. An example of why historical contextualization of
seemingly obvious terminology is Ronit Ricci’s observation that the definition of translation,
seemingly a simple act of (as the Latin literally means) “bringing over” a text from one language
into another, is itself culturally bound (2011, 31ff.).
 OUP CORRECTED PROOF – FINAL, 20/2/2022, SPi

190       

vernacular composition in eighteenth-century northern India, we come

closer to accepting the polysemy of the tradition and that the intentions
of people then do not necessarily map onto what we, with historical
hindsight, know of the tradition’s later development.
In the early eighteenth century and before, composition in the
Persian-influenced kha: ri
: bolī hindī that was later known as Urdu was a
kind of novelty act in Delhi and did not attract much attention from
scholars and critics. Serious poetry in north-central India had been
written either in Persian or in another literary dialect of hindī, such as
: or Awadhī.⁷ This changed sometime between 1720, when the
Braj Bhāsā
:
collected works of the poet Walī Muhammad, known by his pen-name
Walī, arrived in Delhi from the south where kha : ri: bolī hindī (sometimes
under the names dakanī/dakhinī or gujrī) had already been a medium of
serious poetry for well over a century, and 1752, when Mīr Muhammad :
Taqī Mīr wrote Nikāt al-Shuʿarā [Subtleties of the Poets], the first
northern Indian tażkirah to deal primarily with vernacular poets.⁸ To
the present-day observer, it can seem as though an Urdu literary culture
(with its corresponding social network) appeared fully formed in those
three decades since there is virtually no analytical writing on Urdu before
the 1750s. Ārzū is an important exception: His critical writings deploy a
conceptual framework designed to account for the vernacular although
they are nominally about Persian literature.⁹ Of course, complex literary
systems do not appear ex nihilo but rather reconfigure and amplify what

⁷ Āzād says as much in Āb-i Hayāt

(cf. Jalibi 1984, 2:148). A full account of these dialects is
beyond the scope of the current project, but as with the broader question of what is Hindi and
what is Urdu, defining them is beset with issues of arbitrary determinations and anachronism.
⁸ On Walī and his reception in mid-eighteenth century tażkirahs, see Dhavan and Pauwels
2015. The hedge “north Indian” is required because Khwājah : :
Khān Hamīd

of Aurangabad’s
Gulshan-i Guftār [Flower-Garden of Speech], a brief account of thirty poets, is dated 1165/
1751–2, and might be earlier (Gorekar 1970, 107). It also claims to be the first account of
rek: htah poets. Interestingly, many of the poets Hamīd
discusses are from southern India
(Gulshan-i Guftār 1929, 3). Nikāt al-Shuʿarā is undated but a reference to Ānand Rām
Mukhli: s: , who had died in 1164/1751, suggests it was finished soon afterwards. It was definitely
finished before Ārzū’s death in 1756.
⁹ Sadiq 1964, 79. We should not be concerned by this lacuna because the great programmatic
statement in favor of the vernacular in medieval Europe, Dante’s De Vulgari Eloquentia, is also a
rather limited project: “Dante elevates the vernacular poets to the dignity of a standard.
Promoting the vernacular languages to the status of authority is one of his unique achievements
. . . Yet Dante never mentions more than the first line of any poem or discusses its content in
detail, nor does he articulate a theoretical framework for their evaluation” (Shapiro 1990, 42).
 OUP CORRECTED PROOF – FINAL, 20/2/2022, SPi

    191

exists in their environment. In the short preface to Nikāt al-Shuʿarā, Mīr

deals with the question of origins:

Let it not be hidden that in this art [fann] of rek: htah, which is poetry in
the style of Persian poetry in the language of the exalted court of
Shahjahanabad Delhi, there has been until now no book written that
has placed the lives of the poets of this art upon the page of history.¹⁰

:
This is both a claim about the definition of the practice of rekhtah—and
it is obvious that we must call it a practice or art [fann] rather than a
language given how Mīr has explained it—as well as about the history of
:
criticism concerning rekhtah. It is not the natural language of Delhi but
rather the result of what we could see through the socio-linguistic lens of
“language planning,” namely an attempt to make a language hew to a
particular set of principles.¹¹ Mīr continues by minimizing the contribu-
tion of the Deccan:

:
Although there is rekhtah in the Deccan, since a poet of connected
[verse] has not appeared from there, I have not begun with their
names, and the condition of their lines being imperfectly connected
[in sense to the previous line] means that collecting them is often
nothing but wearisome, but I shall write about some of them.¹²

¹⁰ “pūshīdah na-mānad kih dar fann-i rek: htah kih shaʿirī-st ba-t̤aur-i shiʿr-i fārsī ba-zabān-i
urdū-yi muʿalla̍ shāhjahānābād dihlī kitābī tāhāl: tasnīf na-shudah kih ahwāl-i
: shāʿirān-i īn fann
:
ba-safhah-yi rozgār bamānad” (Nikāt al-Shuʿarā 1979, 9).
¹¹ A standard definition, which is deliberately left broad: “Language planning refers to
deliberate efforts to influence the behavior of others with respect to the acquisition, structure,
or functional allocation of their language codes” (Cooper 1989, 45). Obviously this usually refers
to governmental efforts, but there is no reason not to extend the idea to hegemonic influence
rather than direct administration.
¹² “agarchih rek: htah dar dakkan ast, chūn az ānjā yak shāʿir-i marbūt̤ bar-na-k: hwāstah :
lihażā shurūʿ ba-nām-i ānhā na-kardah wa t̤ abaʿ-i nāqis: -i mas: rūf-i īn-ham nīst kih ahwāl
: aks̄ar
malāl andoz gardad, magar baʿzī
az anhā nawishtah k: hwāhad
: shud” (Nikāt al-Shuʿarā 1979, 9).
Rabt̤ (of which “marbūt̤ ” is the adjective) refers to the logical and aesthetic connection between
the two lines of a couplet. As the couplet is the unit of meaning in much Persian poetry,
especially in the ġhazal, properly constructing this internal connection is one of the key
measures of mastery in the Persian poetic context. For example, Wālih in his entry on Ārzū
notes that Ārzū objected to 500 of Hazīn’s

couplets as “disjointed” [nā-marbūt̤ ] (Tażkirah-yi
Riyāz
al-Shuʿarā 2005, 1:347).
OUP CORRECTED PROOF – FINAL, 20/2/2022, SPi

192       

Delhi poets and Deccani poets were obviously aware that they were
engaged in similar projects of making vernacular poetry follow Persian
rules.¹³ Mīr sidelines the contributions of poets from southern India, a
trend that continued in later Urdu criticism (interestingly, though,
Ārzū’s views are more nuanced on this point). If rekhtah/Urdu
: is defined
as a poetic language of Delhi then influence from further afield is
problematic and is to be ignored or dismissed. One common technique,
used by Mīr and subsequent literary historians, is to declare the poetry of
the Deccan stylistically inferior and therefore irrelevant to discussions of
proper Urdu.¹⁴ Another is to show that it is an epigone of Delhi poetry
and therefore irrelevant. In the latter case, the founder-poet Walī is
famously reputed to have come to Delhi (probably from Gujarat),
learned his craft from Shāh Saʿdullāh Gulshan, a contemporary of the
great Persian poet Bedil, and before leaving Delhi was encouraged by
Gulshan to compile the dīwān of rekhtah: verse that would turn out to be
so influential.¹⁵ In this scenario, “Deccani” literature is really just poetry
transplanted from Delhi and eventually returned to its rightful home
with the posthumous arrival there of Walī’s dīwān. There are holes in
this story, but true or not it serves to remind us that the purity of a Delhi
poetic style is an untenable construct—just as the distinction between
Delhi Urdu stylistics and Lucknow Urdu stylistics, despite having been a
mainstay of Urdu literary criticism, crumbles with the slightest bit of
historical probing.¹⁶ The language of Delhi was always influenced by that
of other regions and vice versa.¹⁷ This chapter, however, is not specifi-
cally concerned with tracing the pre-history of Urdu in the Deccan. Here
my goal is to follow the primary sources to disentangle the relationship
between eighteenth-century conceptions of linguistic geography from

¹³ Jalibi 1984, 1:335. ¹⁴ Qāʾim, for example, does the same (Faruqi 2004a, 840).
:
¹⁵ As argued by Muhammad Husain

Āzād, whose mythologizing of the origins of Urdu is
discussed in the following section (Āb-i Hayāt

1907, 84). On the attempt to minimize Walī,
including by assigning him an improbably late death date, see Faruqi 2001, 129–31.
¹⁶ An Urdu literary historian writing in the early 1990s claims to be the first to question this
dogma, which demonstrates how entrenched it is (Zaidi 1993, x). Carla Petievich has admirably
deconstructed it (Petievich 1992).
¹⁷ Compare Francesca Orsini’s recent call for a literary history of India that considers how
the traces of different languages/literary traditions appear obliquely in the texts of any given
tradition: “Multilingual history, as we have seen, requires a perspective open to elements and
agents not immediately present in the texts, an awareness that each text and author exists in a
context that is more complex and varied than the one he gives us to believe” (Orsini 2012, 243).
 OUP CORRECTED PROOF – FINAL, 20/2/2022, SPi

    193

our present understanding of language in northern India. Critics like

Ārzū, even as they were claiming to develop a Delhi-based literary
standard, themselves considered how Delhi’s language fit into a pan-
Indian vernacular literary tradition.
Nikāt al-Shuʿarā, though it positions itself as originary, is not the first
critical engagement with the literary practice called rekhtah : and the
language that would eventually be called Urdu. Scattered references in
Ārzū’s works and the fact that he composed a lexicon of vernacular
words, Nawādir al-Alfāz̤ , demonstrate a sustained engagement with the
vernacular, in particular one that people of taste (namely Persian users)
can shape. Focusing on metropolitan usage as Ārzū does is a different
approach to that of another important Persian text describing vernacular
language, Tuhfat: al-Hind [A Present from India, late seventeenth cen-
tury], which is about the hindī dialect Braj Bhāsā : rather than the khārī :
bolī that is the basis of Delhi speech. The formulation in Inshāʾallāh Khān:
Inshā and Mīrzā Muhammad: Hasan

Qatīl’s Daryā-yi Lat̤ āfat [The Sea of
Refinements, 1807] approximately a century later appears to build on
Ārzū’s definitions but with a tighter focus that excludes the transregional
possibilities that Ārzū saw.¹⁸
This chapter begins with prosopography, that is, tracing the contours
:
of the community of the rekhtah-goyān [rek: htah poets] of Delhi in the
mid-eighteenth century. It continues with definitions of rek: htah/Urdu/
hindī and its many other names in the context of the richness of available
linguistic forms in northern India, and closes with a reexamination of
attitudes towards Persian and Urdu that follow from debates in the mid-
eighteenth century over what we can call the language ideology of Urdu.

Beginning in the Middle: How Āb-i Hayāt

Presents
the Eighteenth Century

In 1880, roughly mid-way between our time and Ārzū’s, Muhammad :

Husain

Āzād published Āb-i Hayāt

[The Water of Life], the first com-
prehensive history of Urdu poetry. It is a Janus-faced work that looks

¹⁸ Concerning Ārzū’s possible influence on that work, see NA 1951, xxxviii.

 OUP CORRECTED PROOF – FINAL, 20/2/2022, SPi

194       

back nostalgically and reverently towards the Urdu poetic tradition while
also striving to be a “modern” (in other words, British-emulating) literary
history that could help Urdu poetry transcend its supposedly decadent and
immoral past. It is stuffed with well-worn anecdotes about each of the poets
like a traditional Persian tażkirah but attempts to systematically demon-
strate civilizational development over time like a contemporary Western
historical text.¹⁹ Āzād was the first person to have been able to write such a
book—no one before would have felt the need—while at the same time being
one of the last to be able to access the living tradition that had been swept
away by the wholesale cultural reorientation in the wake of the 1857 uprising
against colonial rule.²⁰ Āb-i Hayāt’s

influence cannot be overstated; its fluid
style and wealth of stories are so beguiling that people have frequently
overlooked the fancifulness of many of Āzād’s historical claims. For exam-
ple, he prefaces an absurd declaration about the historical origin of Urdu
with the phrase “everyone knows this much.”²¹ His fashioning of the Urdu
poetic canon, of which Ārzū is a part, is a powerful reminder that when we
read a text our understanding of it is shaped by an interpretative tradition.
Āb-i Hayāt

makes much the same case as this study in that it argues
that Ārzū’s career is a significant historical juncture. He offers the
striking formulation that Ārzū is to Urdu poetry as Aristotle is to
logic.²² Earlier tażkirahs had similar hyperbole, for example referring
to Ārzū as the “Abū Hanīfah”

of Urdu poetry (Abū Hanīfah

being the
founder of the dominant tradition of interpretation in Islamic jurispru-
dence in South and Central Asia).²³ However, because Āzād’s subject is
the development of Urdu literary culture, he ignores Ārzū’s Persianate

¹⁹ The work is divided into five historical ages [daur, pl. adwār], each of which is meant to
show a discrete development (or decline) compared to the last.
²⁰ It is a textbook case of what Pierre Nora has called “historicized memory” (Nora 1996).
²¹ “itnī bāt har shak: hs: jāntā hai.” He argues that Urdu was derived from Braj Bhās: ā, a hindī
literary dialect which he claims entirely lacks artifice. Unfortunately, since Braj Bhās: ā literature
is actually full of complex rhetorical devices, he has to make up his own supporting examples.
²² Āb-i Hayāt

1907, 115.
²³ For example, “On the basis that the Islamic Scholars (may their blessings endure) are
called [descendants of] Abū Hanīfah

Kūfī (may God be pleased with him) thus it is fitting that
:
they call poets of the hindī language [i.e. Urdu] descendants of Khān-i Ārzū” [ba-mas̄ābah-yi
kih ʿulamāʾ-i ahl-i haqq
: rā dāmat barakātuhum imām hamām qiblah-yi anām abū hanīfa : kūfī

allāh ʿanhu mī goyand agar shuʿarā-yi hindī zabān rā ʿayāl-i k: hān-i ārzū goyand sazā-st]
razī
from Majmūʿah-yi Naġhz [A Delightful Collection, 1806] by Qudratallāh “Qāsim” (quoted in
NA 1951, editor’s preface 12; Das 1991, 426).
 OUP CORRECTED PROOF – FINAL, 20/2/2022, SPi

    195

intellectual life.²⁴ Literary histories of Urdu like Āzād’s frame Ārzū’s

career in terms of his foundational role in Urdu literature; but
except in passing they do not connect it with his Persian scholarship,
which forms the vast bulk of what he wrote and significantly is what
earned him the respect of his peers. This omission is not surprising
given the precipitous decline in Persian’s prestige in South Asia, but it
does mean that Urdu literary historians’ assessments of Ārzū are incom-
plete. This is not to say that they get their facts wrong, as judged from a
philosophically suspect “neutral” vantage point, but rather that they
cannot fully answer the questions we wish to pose. However, their inter-
pretations shape our own understanding of the tradition and indeed their
interpretations comprise much of the tradition. Ārzū has not received his
due in the annals of South Asian intellectual history, we can conclude,
because he was a Persian scholar first and a hindī poet a distant second.
Furthermore, any attempt to prove Ārzū to be an Indian patriot who
shaped an indigenous language (Urdu) so that it could replace a foreign
language (Persian) is anachronistic in its conception of national and
linguistic identity, and—more importantly—finds no unambiguous con-
temporary textual support.
If we look more closely at Āzād’s analysis of Ārzū then we see that it
suffers from a problem common to much of his scholarship, namely the
need to slot writers into a narrative of the development of Urdu literature
that the available evidence cannot sustain. According to Āb-i Hayāt,

Ārzū was one of the poets who “took poetry that was founded on
wordplay [jugat] and double meanings [żū-maʿnī] and pulled it into
the Persian style and manner of expression.”²⁵ This statement falsely
implies that there was no wordplay in Persian poetry. When Persian
influenced Urdu poetry, it supposedly stripped away the indigenous
tendency to engage in wordplay. This is nonsensical on its face, and
furthermore contradicts Āzād’s contention that the Indic tradition from

²⁴ Āzād’s Suk: handān-i Fārs [The Persian Poet, 1872] deals primarily with the early history of
Persian and the most recent poet discussed in detail is ʿUrfī (1555–91). Ārzū makes a cameo
appearance as an example of an Indian patriot who cared about the philosophy of language
[falsafah-yi lisān] and had not yet ceded the field to Europeans (Suk: handān-i Fārs 2005, 10).
:
²⁵ “aur jis shāʿirī kī bunyād jugat aur żū-maʿnī lafz̤ om par usse khīnchkar fārsī kī tarz̤ aur
adā-yi mat̤ ālib par le āʾe” (Āb-i Hayāt

1907, 115–16).
OUP CORRECTED PROOF – FINAL, 20/2/2022, SPi

196       

which Urdu is derived lacked wordplay.²⁶ Throughout his work, Āzād

focuses on—indeed, it is fair to say that he is obsessed with—the literary
device called “īhām,” or punning. Īhām has a long history in Persian
poetics and a critical analysis of it first appeared seven centuries
before Āzād in the work of Rashīd al-Dīn Wat̤ wāt̤ (d. 578/1182–3).²⁷
:
Furthermore, Amīr Khusrau, the master poet of thirteenth- and
fourteenth-century Delhi, boasted of having invented a new style of
īhām.²⁸ Great classical poets, including Hāfiz̤
, mastered īhām and it
was an important tool in courtly discourse since it often provided an
opportunity for a clever turn of phrase at the right moment.²⁹ The device
clearly has an impeccable pedigree in Persian and yet Āzād flatly claims
that īhām (or in its Indic variant “jugat”) is a fault that Urdu inherited
from “Hindi” and, even more improbably, that the process of
Persianization of Urdu poetry was mostly about scrubbing the īhām
out of it.³⁰ There was an eighteenth-century conversation about īhām,
but it came later and was far less central than Āzād has made it.³¹
The reason Āzād should frame all of Urdu literary history as a
campaign against wordplay is obvious when we consider that he spent
much of his life working for the colonial state: Nineteenth-century Britons
had imbibed Romanticism so deeply that in general the only kind of
contemporary poetry they valued as serious art was that of poets such as
William Wordsworth with their descriptions of apparently spontaneous

²⁶ The claim is especially nonsensical in the context of eighteenth-century Persian compo-

sition, which was much maligned by later Iranian and colonial literary critics precisely because it
was supposedly baroque and elevated wordplay above sense (as discussed in Chapter 3).
²⁷ See the discussion in Chapter 2.
²⁸ I ʿjāz-i Khusrawī
: 2007, 1:41; see also Gabbay 2010, 36ff.
²⁹ For some examples of Persian īhām, see Hasan 1998, 20–1.
³⁰ “This master of invention [the Urdu poet Mirzā Muhammad : Rafīʿ Saudā (1706?–81)] . . .
compiled verbal devices, Persian constructions, and original themes, and made such excellence
that people forgot punning [īhām], alliteration, and the other verbal devices that were the
foundation for Hindi dohrās” (Āb-i Hayāt
1907, 54). The term “jugat” is derived from Sanskrit
“yuktih”: referring to combining or yoking whereas the etymology for “īhām” is “to make one
suppose” (Platts 1884, 384; Chalisova 2004).
³¹ On how different tażkirahs address īhām, see Gorekar 1970, 122. Ġhulām Hamadānī
:
Mus: hafī’s late eighteenth-century tażkirah Riyāz
al-Fus: ahā
: [Gardens of the Eloquent] notes
that the meanings of Ārzū’s verse were predicated upon īhām because it was the “age [daur]
of īhām” (Tażkirah-yi Riyāz
al-Fus:ahā : 1985, 38). Mīr contrasts īhām-style rek: htah with
“andāz”-style in the afterword of his Nikāt al-Shuʿarā. He does not define “andāz” (which
literally means “measure” or “mode”), but presumably for him it means poetry without
īhām—needless to say, Mīr himself uses īhām in his supposedly “andāz”-style poetry.
 OUP CORRECTED PROOF – FINAL, 20/2/2022, SPi

    197

emotional reactions to natural scenery or events. Thus there was no

room for the perceived artificiality of īhām and similar devices.³² After
Āzād’s full-throated denunciation, this humble rhetorical device has
often been considered a pernicious evil that stifles genuine emotion.
With a few exceptions, twentieth-century critics of Urdu literature have
not historicized Āzād’s thought.³³ For example, Muhammad Sadiq, who
wrote in the 1960s, views the apparently disastrous consequences of
īhām on a poet’s creativity in exactly the same way as Āzād.³⁴
However, the modern critical obsession with perceived artificiality as
epitomized by īhām means that Sadiq has a warped frame of reference
:
for what, for example, the oeuvre of the early eighteenth-century rekhtah
poet Shāh Hātim

(b. 1699) represents.³⁵ Sadiq tries to argue that the
preface [dībāchah] to Hātim’s

second, revised collected works, entitled
Dīwānzādah,³⁶ takes a position against īhām. Indeed, the critic tries to
have it both ways, admitting that Hātim

frequently used punning in his
poetry while implying that the very existence of the Dīwānzādah (in
whose preface the poet explains the aesthetic considerations that com-
pelled him to release a thoroughly revised edition of his collected works)
is in fact a direct attack on īhām. But since īhām was commonly
discussed at the time, and there certainly was a contemporaneous turn
against it, why does Hātim

not mention it even once in the preface?
There are far more interesting aesthetic judgments that Hātim
does
mention, but Sadiq does not consider them because they are irrelevant
to the īhām-centric narrative he wishes to present. Similarly, Āzād
provides no analysis of the significant scholarly interventions that Ārzū

³² It is worth pointing out that the equivalent work for the Indo-Persian tradition, Shiblī
Nuʿmānī’s Shʿir al-ʿAjam [Persian Poetry, 1907–25], is likewise driven by the same unspoken
preoccupation with English poetics (Faruqi 2004b, 11ff.). For some telling examples of how
damaging this assumption has been for Urdu criticism, see Faruqi 2005.
³³ For a trenchant complaint about Urdu critics’ tendency to reproduce mythology as fact,
see Faruqi 2008.
³⁴ Sadiq 1964, 75ff. On the same page he provides a couple of useful definitions of īhām,
which the interested reader can consult. Another recent history of Urdu literature similarly calls
īhām part of an awkward, artificial phase in Urdu poetry (Hasan 1995, 102).
³⁵ He frames the poet’s career in terms of īhām: “Hātim’s

life spanned nearly the entire
eighteenth century, and therefore represents the rise and fall of the īhām ideal” (Sadiq 1964, 78).
³⁶ A dīwān is the term for a poet’s selected works in Persian or Urdu, so Dīwānzādah is a
cheeky title meaning “son of the dīwān.”
OUP CORRECTED PROOF – FINAL, 20/2/2022, SPi

198       

made in Urdu literature except in the most general terms, and it is worth
enumerating them.
If, as Āzād himself has written, Ārzū was primarily a Persian writer,
then why discuss vernacular literary practice in this study at all? Hindī
has to be included precisely to keep us from the temptation of treating
Ārzū as two separate people, namely a great Indo-Persian philologist and
an important figure in Urdu literature. The tendency to address his
achievements in the two languages separately means that we have not
appreciated the fact that his theories on language and aesthetics, osten-
sibly about Persian only, are actually broad enough to encompass hindī
as well. Without his engagement with the vernacular, his philosophy of
language in Persian would not have been as rich, and without his
standing in Persian, he would not have been in a position to influence
vernacular poetry. There is also a practical problem: It is difficult for us to
know how to split up Ārzū’s life-world between the vernacular and
Persian—Āzād states with a brazen certainty that “Khān-i : Ārzū was
not an Urdu poet; nor did people of that time consider Urdu poetry to
be an accomplishment.”³⁷ But did Ārzū himself feel that way about
vernacular composition?

A Who’s Who of the People of Rekhtah

:

The social networks of the Delhi rek: htah community can be recon-
structed largely on the basis of tażkirahs. These are among the few
relevant sources that we have, and they compel us to dispense with the
idea that vernacular poetry represents the language of commoners win-
ning out over that of elites, for these texts record elites in dialogue with
other elites.³⁸ Here, as in many cases in Europe and India, the texts do
not offer examples of the masses undermining the elite through their

³⁷ Āb-i Hayāt

1907, 116.
³⁸ The first dīwān of Urdu poetry to have a prose preface in Urdu was that of ʿAbd al-Walī
ʿUzlat of Surat (1692/3–1775), according to Jalibi (1984, 2:327). Indeed, all criticism on rek: htah
was written in Persian until Mirzā ʿAlī Lut̤ f ’s 1801 Gulshan-i Hind [Rose-Garden of India],
which was the first Urdu tażkirah actually in Urdu (Gorekar 1970, 105). This did not become
common until decades later. As late as 1846, Saʿādat ʿAlī Nās: ir claimed that his Khush
: Maʿrikah-
i Zebā was the first tażkirah of Urdu poets to be written in Urdu, which is not of course true, but
 OUP CORRECTED PROOF – FINAL, 20/2/2022, SPi

    199

language choices. Nor, in that case, do elites appear to be speaking on

behalf of the masses as might be expected. Rather, the people developing
the vernacular for their own purposes were the very same elite who are
:
experts in Persian. A perfect illustration is the rekhtah poet with the
pen-name Āftāb, who happened to be the emperor Shāh ʿĀlam II
(r. 1759–1806) and participated in these very same literary networks (his
:
teacher was Mīrzā Fakhir Makīn).³⁹ This has not stopped later scholars
from making flowery pronouncements on the proletarian nature of ver-
nacular poetry. N. S. Gorekar, for example, argues that “in its infancy, Urdu
was quite simple and homely and could cope with the requirements of the
people whose needs were few and whose outlook on life was limited.”⁴⁰
Likewise, Jadunath Sarkar says of the Mughal period that “the
Muhammadans of that age (except the Hindi-speaking portion) had no
vernacular religious poetry for the masses.” He presumes that Urdu was
invented to fill that role.⁴¹ Many eighteenth-century vernacular poets,
such as Mīr Dard and Maz̤ har, were well-known Sufis, but it is unlikely
that their language choice for verse was motivated by a desire to spread
religious messages.⁴² Bedil, arguably the most influential Indo-Persian

as Pritchett argues, he must have believed it was or at least expected his readers to believe it since
it figures so prominently in the introduction (Pritchett 2003, 882–4). Nās: ir provides no
satisfying explanation of why he should write in Urdu: He merely says that “yakrangī”
[uniformity] is better than “dorangī” [diversity]. (There is apparently no connection to the
British and their language planning that is clear in some other early Urdu critical prose works.)
The Urdu translation in 1842 of Shams al-Dīn Faqīr’s Persian rhetorical treatise Hadāʾiq
al-
Balāġhah by Imām Bakhsh : Sahbāʾī

of Delhi College was an important moment in making Urdu
teaching texts available in Urdu. The preface highlights the fact that Sahbāʾī

not only translated
the text but replaced the Arabic and Persian examples with Urdu ones.
³⁹ He was also a Braj poet as attested by a few surviving snippets (Marshall 1967, 436). Shafīq
identifies Āftāb as the emperor Muhammad
: Shāh (r. 1719–48), which is obviously wrong since
earlier tazkirahs

do not mention him (Chamanistān-i Shuʿarā 1928, 245).
⁴⁰ Gorekar 1970, 101.
⁴¹ Sarkar 2008, 322. His division of Hindi-speaking Muslims from Muslims generally (who,
presumably, would be speaking Urdu or some kind of proto-Urdu) is anachronistic because it
presumes a split between Hindi speakers and Urdu speakers with Urdu, in this view, being an
exclusively Muslim language. Sarkar unhelpfully conflates spoken language with literary lan-
guage. Likewise, I am not convinced by Tariq Rahman’s argument that there was a conscious
“Muslimization” (his term) that is the best framework for explaining the development of Urdu
(Rahman 2010, esp. 90).
⁴² Similar questions are explored in Pollock 1995 and 2000. Richard Eaton has argued that in
the Deccan it was in fact the case that Deccani Urdu was adopted for spreading a religious
message, though he offers some caveats (Eaton 1978; cf. Guha 2004). As Pollock reminds us, the
European case is complicated by the fact that so many people had a basic knowledge of Latin
that it could hardly be said to be restricted to the cloister and the court (2006, 460–1; cf. Briggs
OUP CORRECTED PROOF – FINAL, 20/2/2022, SPi

200       

poet of the late seventeenth and early eighteenth century but also cred-
:
ited as a rekhtah master poet by Mīr and others, was also an acclaimed
Sufi. If language choice is the crucial criterion then why should Dard and
Maz̤ har be remembered as religious popularizers and Bedil not? Another
important point is that being a Sufi, even a self-proclaimed mendicant,
does not make someone a representative of demotic religion. The records
we have of this early Urdu literary culture were kept by and represent an
elite community. However much we should like to know about how
other strata of society used language, there is little data for the pre-
colonial period that could help.
Ārzū came from this same charmed circle, being a part of both the
Mughal aristocracy and having a Sufi lineage. He is recognized in most
tażkirahs of Urdu poets as the first important teacher of Urdu compo-
sition in Delhi. Indeed, he had earned his moniker “the Abū Hanīfah

of
Urdu poetry” within a generation.⁴³ Mīr’s Nikāt al-Shuʿarā uses the
slightly less hyperbolic formulation that all teachers of Urdu were taught
by Ārzū: “All teachers connected with the art of rek: htah are classmates
[studying under] this great man.”⁴⁴ Besides the Persian poets mentioned
in Chapter 1 who also composed in the vernacular, his name is linked to
Mīr, Mirzā Jān-i Jānān Maz̤ har, Mirzā Muhammad : Rafīʿ Saudā, and
: :
Kh wājah Mīr Dard, all of whom attended re :
k htah mushāʿirahs held at
his house.⁴⁵ These younger contemporaries of Ārzū ended up being the
most important Urdu poets of the mid- to late eighteenth century. He
was a personal mentor to Shāh Mubārak Ābrū (1685–1733), Sharaf al-
Dīn Mazmūn

(1689?–1735?), and Ġhulām Mus: t̤ afa̍ Yakrang (d. 1737). In
Tanbīh al-Ġhāfilīn, Ārzū says of them that “each of the three is a master

2003). Given what we know of how Sufis used Persian in India, it seems likely that they too
expected to be understood even if they were delivering much of the content of their message to
their followers in the prestige language.
⁴³ See Majmūʿah-yi Naġhz [A Collection of Delights] by Qudratallāh Qāsim (1750–1830)
dated 1806 (ʿAbdullah 1965, 42). The twentieth-century scholar Nabi Hadi reports that the
phrase appears in an earlier tażkirah by Ārzū’s student Hakīm

Husayn

Shuhrat Shīrāzī but
I have been unable to trace the reference (Hasan 2001, 850). According to Ārzū, Shuhrat was
probably from a Bahraini Arab background but raised in Shiraz before coming to India (MN
2004, 2:848).
⁴⁴ “hamah ūstādān mazbūt̤

-i fann-i rek: htah hamshāgirdān-i ān buzurgwār-and” (quoted in
ʿAbdullah 1965, 43). Likewise Mīr Hasan

in 1191/1777 calls Ārzū “teacher of the teachers of
Hindustan” (Tażkirah-yi Shuʿarā-yi Urdū 1985, 13).
⁴⁵ Āb-i Hayāt

1907, 164.
 OUP CORRECTED PROOF – FINAL, 20/2/2022, SPi

    201

:
and established in the art of rekhtah” and notes that they have all sought
: :
correction [islā h] from him.⁴⁶ Despite Ārzū’s obvious importance in the
social fabric of the Delhi poetic community, the only Urdu composition
of his that we can trace is a dozen or so couplets quoted in eighteenth-
century tażkirahs, none of which is particularly inspired or historically
interesting. While in Ārzū’s case—and that of most of his contempor-
aries, Hātim

being an important exception—there was no interest in
collecting one’s Urdu poetry and distributing manuscripts of it, there was
:
a thriving community of rekhtah poets who were becoming ever more
:
serious about it. The systematization of rekhtah poetry is what concerns
us here, and we can trace it through the careers of poets tied to Ārzū
either directly or indirectly. The first of these is Shāh Hātim,

whose
career is unusual in that he actually wrote (Persian) prose explaining
some of his aesthetic choices.

Shāh Hātim

(1699–1783) and His “Contemporaries”

Z̤ uhūr al-Dīn Hātim

was a native of Delhi. He began composing rek: htah
poetry in his late teenage years and the first volume of his works [dīwān]
was compiled in 1144/1732, which makes it one of the earliest Urdu
dīwāns compiled in Delhi. Some of the ġhazals are dated 1131/1718–19
or one year before Walī’s dīwān is said to have arrived in Delhi, and they
are not appreciably different from later ones.⁴⁷ The influence of Walī on
rek: htah stylistics might therefore be overstated—or perhaps we simply
have the dates slightly wrong. Hātim

was well connected politically, and
wrote two mas̄nawīs for the emperor Muhammad: Shāh (one of which is

⁴⁶ “har sih ustād wa mustanad-and dar fann-i rek: htah” (TĠh 1981, 76). Sayyid Fath: ʿAlī
:
Khān Hus: ainī Gardezī, in his Tażkirah-yi Rek: htah-Goyān [Tażkirah of Rek: htah Poets, 1752]
also mentions that Mazmūn
was Ārzū’s student (1995, 155). Gardezī’s tażkirah was one of the
two important tażkirahs probably written in response to Mīr’s Nikāt al-Shuʿarā. The other,
Qayām al-Dīn Muhammad: Qāʾim’s Mak: hzan-i Nikāt [Treasury of Fine-Points] composed two
:
years later, claims in the preface that its author had not seen any other tażkirah of rekhtah poets,
which was possibly disingenuous since Qāʾim was in Delhi when Mīr and Gardezī wrote theirs
(Gorekar 1970, 108–9).
⁴⁷ Dīwānzādah 1975, 10.
OUP CORRECTED PROOF – FINAL, 20/2/2022, SPi

202       

a delightful account of coffee) in 1749.⁴⁸ The preface to his Dīwānzādah

(1169/1755–6) is an important early intervention in Urdu stylistics. He
writes in it that the need had arisen to replace his first dīwān, which was
by then hard to find and, more importantly, stylistically out of date. He
had pared down the number of poems and so gave the new edition the
cheeky name Dīwānzādah (literally “born of the dīwān” or the “dīwān’s
son”).⁴⁹ He asserts that since certain words have an inherent ugliness/
inappropriateness [qabāhat],: he has tried to give them up. The words he
lists are all Sanskrit-derived, like “jag” meaning “world.” Many have seen
this as the first salvo in the Kulturkampf whose armistice terms in the
early twentieth century were that Hindi was to be “the language of
Hindus” (and hence the national language of India) and Urdu “the
language of Muslims” (and of Pakistan).⁵⁰ The fact that he also argued
for the correct spelling of Persian- and Arabic-derived words has been
interpreted as an “Islamicization” of Urdu. However, there is no com-
pelling reason to read Hātim’s

interventions (namely rejecting certain
words and respelling others) as a rejection of a non-Muslim other.
Instead he appears to follow Ārzū in the concern for setting a literary
standard. He certainly knew Ārzū’s work because although he does not
mention Ārzū by name, the contemporary poets he does mention are
connected with Ārzū.
In Dīwānzādah’s preface, Hātim
invokes Ābrū (1685–1733), Mazmūn

(1689?–1735?), Maz̤ har (1689–1791), Shaikh : Ahsanallāh
: (d. 1737–8),
Mīr Shākir Nājī (1690?–1744/47?), and Yakrang (d. 1737) as his
“contemporaries” [muʿās: irān].⁵¹ This can be taken as an index of the

⁴⁸ Additionally, his Persian dīwān, mentioned by Āzād in Āb-i Hayāt
(1907, 113) but
subsequently thought lost, has been published as Dīwān-i Fārsī-yi Hātim
Dihlawī 2010. The
coffee mas̄nawī and its context in eighteenth-century connoisseurship has been explicated in
Hakala 2011.
⁴⁹ Besides Dīwānzādah itself, see Mus: hafī’s
: Tażkirah-yi Hindī (1985, 88–9). For Shamsur
Rahman Faruqi, Hātim

was the person who brought “elitism” into rek: htah by insisting that it
follow rules that require a knowledge of Persian and Arabic (Faruqi 2004a, 850). Thinking about
his intervention as setting a literary standard is perhaps more useful than the implied criticism
of “elitism.”
⁵⁰ The classic study is King 1994 and for Hindi specifically Dalmia 1997. A notable text from
the debates around the proper national language for independent India is R. S. Shukla’s Lingua
Franca for India (Hind), which attempts to prove that Sanskritized khar: ī bolī Hindi is natural
while Urdu is artificial (Shukla 1947).
⁵¹ In his own 1168/1754 tażkirah, Qāʾim Chāndpūrī mentions Hātim
as an associate of Ābrū
and Mazmūn
(quoted in Dīwānzādah 1975, vi).
 OUP CORRECTED PROOF – FINAL, 20/2/2022, SPi

    203

:
important Delhi-based rekhtah writers of the first third of the eighteenth
century. Ārzū claims to have mentored three of the six of them in
vernacular poetry. Ārzū also knew Maz̤ har and Nājī personally, as he
reports in Majmaʿ al-Nafāʾis.⁵² The leading light of the next generation
was Ārzū’s nephew Mīr.

:
Mīr Muhammad Taqī Mīr (1722–1810)

As Ārzū’s nephew, Mīr had an impeccable introduction into Delhi

literary society.⁵³ He arrived from Agra most likely in 1739 and stayed
with Ārzū. C. M. Naim is undoubtedly right to infer that Ārzū’s inter-
vention transformed Mīr from somewhat of a bumpkin into a confident
metropolitan poet.⁵⁴ Mīr praises Ārzū to the skies in his entry in Nikāt al-
Shuʿarā, which is hardly unexpected in a tażkirah, and acknowledges
him as his teacher.⁵⁵ Despite this public expression of admiration while
Ārzū was alive, Mīr’s quasi-autobiographical narrative, Żikr-i Mīr [The
Recollection of Mīr], which was written perhaps two decades later, tells a
different story. Mīr says nothing of what he learned in Delhi from Ārzū
except that he had read “a few insignificant books” at Ārzū’s house.⁵⁶
While Mīr was staying with Ārzū, Saʿādat ʿAlī, “a Sayyid from Amroha,”
advised him that he should write in the vernacular, whereupon his work
became extremely popular.⁵⁷ At some point during Mīr’s stay, as Mīr

⁵² MN 2004, 3:1583, 1691.

⁵³ Ārzū was his k: hālū. In current usage, this means “mother’s sister’s husband” but previ-
ously it also meant “maternal uncle.” It is in this latter sense that Ārzū and Mīr are related since
Ārzū was his stepmother’s brother (Naim 1999, 66; Āb-i Hayāt
1907, 194).
⁵⁴ Naim 1999, 15.
⁵⁵ Nikāt al-Shuʿarā 1979, 10–1. Mīr also respected Ārzū as a critic in Urdu, at least in one
important case (Faruqi 1996, 99). Thanks to Owen Cornwall for the reference.
⁵⁶ Naim 1999, 66–9.
⁵⁷ One source, Nishtar-i ʿIshq [Love’s Lancet, 1233/1817] by Husain
:
Qulī Khān ʿAshiqī of
Azimabad (i.e., Patna), reports that Ārzū had a similar conversation with Saudā (1968, 84–5;
translated in Alam 2004, 180–1). There is a crucial difference in the two conversations as
reported, namely that Mīr’s conversation with the Sayyid had no sense of pitting Indians against
Persians in their mastery of Persian whereas the putative conversation between Saudā and Ārzū
explicitly states that Indians cannot rise to the level of Iranians in Persian (Indians hold a mere
lamp to the sun of the Persian of Iranians) and furthermore that Persian itself is worn out as a
literary medium. The conversation reported in Nishtar-i ʿIshq is problematic as evidence in my
view: Firstly, it appears unsourced in a tażkirah written more than fifty years after Ārzū’s death.
OUP CORRECTED PROOF – FINAL, 20/2/2022, SPi

204       

himself explains it, Ārzū received a letter from Mīr’s stepbrother and his
entire demeanor towards his nephew changed. Mīr admits that this
situation drove him mad and they eventually had a falling out over
supper, which led him to stumble down the road towards Jama Masjid
in a daze. He made it as far as Hauz Qazi, where he was recognized by
someone who helped him secure a patron and move out of Ārzū’s house.
The exact cause of this family split is not recorded.⁵⁸ Żikr-i Mīr was
almost certainly not meant for public consumption so it is no surprise
that Mīr felt free to render a vicious final assessment of Ārzū in its pages:
Towards the end of his life, Ārzū “went chasing in the desert of greed,
that is to say he journeyed to Shuja-ud-Daulah’s camp [Lucknow] . . . But
he received nothing but a fistful of air and, buffeted by Time, died.”⁵⁹
Mīr expressed some ambivalence about the relative value of the
:
vernacular and Persian. While writing a tażkirah of rekhtah poets repre-
sents a significant commitment to rek: htah poetry, he also spent two years
trying to be a Persian poet exclusively.⁶⁰ The lack of any statement in
prose expressing his opinion of rek: htah justifies a brief diversion into
:
discussing the references to rekhtah in his poetry. He writes, for example,

Why was rek: htah in this lofty rank, Mīr?

Whatever “ground” [zamīn] emerged, I bore it away to the sky.

Secondly, while the statement as reported does not directly contradict the arguments about the
universality of literary language that Ārzū presents throughout his works (as described in
Chapter 4), it is less nuanced than his other invocations of Iranian literary competence and
concedes more than he concedes elsewhere. It is possible that this iconic conversation never
happened, but it has nonetheless been adduced as evidence without mentioning these concerns
(e.g., Syed 2012, 298).
⁵⁸ We can speculate with Naim that Ārzū might have disliked Mīr’s Shīʿah tendencies or that
perhaps Mīr had an affair with a member of Ārzū’s household. The former possibility is also
suggested by Āzād (Āb-i Hayāt
1907, 194).
⁵⁹ Translated in Naim 1999, 76. The evidence that it was not meant for a wide readership
comes first in its rarity since it was lost for nearly a century and a half and only published in
1928 (Faruqi 2005, 173). There is also an unflattering reference to the current king Shāh ʿĀlam II
(r. 1759–1806), who “slanders the title [of Emperor] and is a prisoner of the Firangis” (i.e., the
British are the real power behind the throne) (Naim 1999, 80). Lastly, as Naim has shown,
throughout the text Mīr uses expressions from Ārzū’s lexicon Chirāġh-i Hidāyat with such
frequency—sometimes a dozen appear on the same page—that it appears the text was meant to
be pedagogical (Naim 1999, 13–14, 19). It could have been a lesson in Persian composition
intended for a member of the family enlivened by being cast in the form of an “autobiography.”
In fact, it tells us surprisingly little about Mīr and his life.
⁶⁰ As reported by Ġhulām Hamadānī Mus: hafī : (1747/50?–1824) in his tażkirah ʿIqd-i S̄uraiyā
[The Necklace of the Pleiades] (1978, 95).
 OUP CORRECTED PROOF – FINAL, 20/2/2022, SPi

    205

And also,

What respect does rek: htah receive—although I

in this art was equal to Naz̤ īrī?⁶¹

Trying to discern the poet’s personal feelings on the basis of poetic

utterances is a minefield because of how poetry works rhetorically,
namely as formal speech embedded in a system of metaphor, but let us
try.⁶² In both couplets, the speaker (for simplicity, we can just call him
“Mīr”) is expressing his prowess in composing rek: htah, which is a
continuation of the venerable Persian tradition of poetic braggadocio
: k: hur]. In the first couplet, rek: htah is in an exalted state because
[tafakh
Mīr has raised it from its earth-bound position to the sky. (As a technical
poetic term, zamīn refers to the combination of meter and rhyme.) In the
second, Mīr questions whether rek: htah has any value because even
someone like himself, who writes it as well as Naz̤ īrī (d. 1612) wrote
Persian, supposedly gets no recognition.⁶³ We can surmise that the value
:
of rekhtah was still up for negotiation at the time Mīr was writing. In
another poem he says,

I had fallen in love with a Turkish lad—what rek: htahs

I composed
Gradually going from Hindustan, my poetry went to Iran.⁶⁴

: :
⁶¹ “rek: htah kāhe ko thā is rutbah-yi aʿla̍ mem mīr / jo zamīṅ niklī use tā āsmāṅ maim le gayā”
: :
(1056, 7) and “kyā qadr hai rek: hte kī go maim / is fann mem naz̤ īrī kā badal thā” (1057, 4), as
printed in Kulliyāt-i Mīr 2003.
Such rhetorical anxieties over rek: htah continue into the nineteenth century, as encapsulated in
Ġhālib’s famous line: “If anyone would say, ‘How could rek: htah be the envy of Persian?’ / Just
:
once, recite to him the speech of Ghalib—‘Like this!’ ” [jo yih kahe kih rek: htah kyomkih ho rashk-i
:
fārsī/guftah-i ġhālib ek bār par: h ke use sunā kih yūm] (116, 10), as printed in Dīwān-i Ġhālib 1982.
⁶² We should remember when discussing poetry that “formal speech requires a positional
rather than a personal identity” (Kennedy 1998, 67).
⁶³ In the context of European vernaculars defining themselves against Latin, Peter Burke has
referred to this as the “humility topos” (2004, 18). Allison Busch has noted a similar rhetoric in the
work of the Braj poet Keshavdās (1555–1617): “In a family where even the servants did not know
how to speak / the vernacular, Keshavdas became a slow-witted Hindi poet” (Kavipriyā 2.17,
translated in Busch 2011, 23). He framed his linguistic defection—he was a pandit with full
knowledge of Sanskrit—apologetically as a desire to educate vernacular poets (Busch 2011, 44, 54).
: :
⁶⁴ “turk bachche se ʿishq kiyā thā rek: hte kyā kyā maim ne kahe / raftah raftah hindusitām se
shiʿr mirā īrān gayā” (1554, 6).
OUP CORRECTED PROOF – FINAL, 20/2/2022, SPi

206       

This verse unusually has “rekhtah”

: used in the plural, presumably to
mean “poems written in rek: htah.” Such a usage again cautions us against
thinking of rek: htah as an exact synonym for what later became known as
Urdu.⁶⁵ Unexpectedly, we find the trope of the poem’s making its way to
a distant place, namely Iran—it is safe to assume that there was no actual
:
readership for rekhtah poetry in Iran. We can perhaps read it as a joke
(“I was so desperate that my poetry went everywhere, even to Iran where
people couldn’t understand it”) or perhaps it is a fossilized trope whose
logic we are not meant to question. Now let us turn to how Ārzū
references rek: htah.

Rekhtah
: in Majmaʿ al-Nafāʾis

Mid-eighteenth-century critical works that are nominally about Persian

contain various references to vernacular composition. Ārzū’s Majmaʿ al-
Nafāʾis, a tażkirah of writers in Persian compiled at roughly the same
time as Mīr’s Nikāt al-Shuʿarā, mentions several poets who also com-
posed in the vernacular. Some of these are Ummīd (1678–1746), Mīr
Dard (1722–85), Mīrzā Roshan Zamīr : (seventeenth century), Maz̤ har
(1689–1791), and Nisbatī (fl. mid-eighteenth century). Although Mīr
refers to Ummīd as a “Mughal” (which means either Central Asian-
born or of Central Asian descent), Ārzū states that he is from Hamadan
in western Iran and spent nearly forty years in the Deccan.⁶⁶ Ārzū knew
Ummīd in Delhi and reports that he had difficulty speaking hindī,
literally “his tongue did not turn well in the hindī speech” [zabān-ash
:
dar lahjah-yi hindī khūb namī gardad], which is of course ambiguous as
to whether he lacked fluency entirely or merely had trouble pronouncing

⁶⁵ Imre Bangha has traced the term to a seventeenth-century musicological treatise in which
it means a bilingual text set to both a rāga and a tāla, that is to musical scale and rhythm
(2010, 24).
⁶⁶ Nikāt al-Shuʿarā 1979, 13; MN 2004, 1:169. Ummīd appeared in Chapter 3 of this study as
a defender of Ārzū’s TĠh. In Āb-i Hayāt,

Āzād notes that Ummīd and Mīr Muʿizz al-Dīn
:
Muhammad Fit̤ rat were Iranians who composed in Urdu (Āb-i Hayāt

1907, 75). Fit̤ rat appears
in MN and other tażkirahs but his composition in hindī is not mentioned—in any case, Āzād
only mentions them to make the case that before Walī, rek: htah poetry was an aesthetic failure.
He backhandedly (and anachronistically) suggests that there was no good poetry because native
Indians did not care enough about their own language to write in it.
 OUP CORRECTED PROOF – FINAL, 20/2/2022, SPi

    207

Indic sounds. The latter case, a difficulty with pronunciation, is implied

by Ārzū’s observation that Ummīd was so good in hindī verse that he
noticed subtleties that even a “Mughal” (meaning in this context an
Indian-born Muslim) might not.⁶⁷ Mīr Dard’s father was an important
spiritual leader in the Naqshbandī order. Ārzū reports that Dard was a
talented poet “especially in the rekhtah
: which is now in fashion in India”
:
[siyamā rekhtah : dar hindūstān rawāj dārad], adding that “he
kih al-hāl
writes good Persian as well.”⁶⁸ Zamīr: was thought to have been an
Iranian soldier in Surat, but Ārzū writes that in fact not he but his
ancestors were Iranian and that he was a newsletter-writer [waqāʾiʿ-
nigār] in Surat and not a soldier. Of his hindī, Ārzū says that despite
his official duties he was “nonetheless very assiduous in every linguistic
exercise [zabān-bandī] such that he mostly composed hindī poetry under
the pen-name ‘Nehī’ which means ‘lover.’”⁶⁹ Ārzū concludes the entry
with (what strikes me at least as) a rather awful couplet of his own
composition:

Let those two black curls on desire’s [or “Ārzū’s”] glowing face
:
be the hindī poem of Mīrzā Roshan Zamīr.⁷⁰

:
Another poet whose rekhtah Ārzū mentions is Maz̤ har, a Naqshbandī
Sufi like Dard who in contrast to the general trend wrote in the vernac-
ular but abandoned it in later life. Ārzū writes that “previously he
:
sometimes composed rekhtah—which is poetry of hindawī and Persian
mixed—in a particular style; now having understood it to be against his

⁶⁷ See Dudney 2017b on Ummīd and other Iranian emigres. I thank Mana Kia for bringing to
my attention a significant discrepancy between the two recent editions of MN (MN 2005 says
“mī gardad” following MN 1970 where MN 2004 says “namī gardad” and the latter reading fits
the context far better).
⁶⁸ “fārsī ham k: hūb mī gūyad” (MN 2004, 1:440).
⁶⁹ MN 2004, 2:956. Nehī is derived from Sanskrit “snehī” [lover] which in turn is derived
from “sneha” [oiliness, love]. It goes without saying that Ārzū’s interest in a poet with a
Sanskrit-derived pen-name puts us well outside the Hindi/Urdu paradigm of today.
⁷⁰ “ān do gesū-yi siyah bar rū-yi rak: hshān-i ārzū / shʿir-i hindī buwad az mīrzā roshan zamīr”


(MN 2004, 2:956). (In MN 2005, 97, the second hemistich is reported as “shʿir-i hindī būdah az
mīrzāʾi roshan zamīr.”)

Presumably the two curls are being compared to the two hemistichs of a
couplet.
OUP CORRECTED PROOF – FINAL, 20/2/2022, SPi

208       

nature, he has given it up.”⁷¹ Another poet, Nisbatī, was from

Thanesar—Ārzū begins the entry by explaining where Thanesar is,
namely between Delhi and Lahore—and his father was from Iran.
Along with Ābrū, he was a student of Ārzū’s but has now become
“a peerless master in the art of rekhtah: poetry.”⁷² Another friend of
Ārzū’s hailing from what is today Haryana is Kh : wājah
: :
Muhammad
ʿĀqil (d. 1143/1730) of Sonepat, who composed not only in rekhtah :
but also in Braj using the name Budhwant, which means “clever” like
:
his Persian/rekhtah :
takhallu s: ʿĀqil.⁷³ The range of ʿĀqil’s vernacular
literary activities dramatizes what we consider in the next section,
namely how the ideology of the vernacular differs for us and for the
eighteenth century.

Defining the (Literary) Vernacular

“Vernacular literature” is a term necessarily defined in opposition to a

high-prestige, cosmopolitan literature.⁷⁴ While this seems to lock any
analysis of it into a potentially unhelpful and ahistorical binary, there are
two points to keep in mind: Firstly, the elite of eighteenth-century South
Asia (as well as, for that matter, Latinate Europe) themselves conceived

⁷¹ MN 2004, 3:1583. Gardezī says Maz̤ har is from Bukhara (Tażkirah-yi Rek: htah-Goyān
1995, 152).
⁷² “dar fann-i rek: htah-goʾī ustād bī-mis̄last” (MN 2004, 3:1649).
⁷³ MN 2004, 1:1122–3; Safīnah-i Kh : wushgo
: 1959, 178–82 (which provides the detail of his
Braj pen-name). He was a friend of Ārzū and a disciple of Shuhrat. Kh : wushgo
: quotes a couplet
written by ʿĀqil to Ārzū and Ārzū’s reply. His dīwān is available in Aligarh’s Habib Ganj
collection as ms H.G. 47/66.
⁷⁴ We can say of “vernacular” that “the term describes, not a language as such, but a relation
between one language situation and another, with the vernacular at least notionally in the more
embattled, or at least the less clear-cut, position” (Somerset and Watson 2003, x).
Etymologically, it can be traced back to the Latin grammarian Marcus Terentius Varro’s
(116–23 ) invocation of “vernacula vocabula” in De Lingua Latina [On the Latin
Language], an influential text of which just two books survive. “Vernacula” is derived from
“verna” meaning “a slave born in one’s own house” and so vernacular language is literally the
non-standard Latin vocabulary of the lower class of a particular place. The definition of
vernacular used in this study and in most people’s understanding of the term today has only
the most tenuous link with Varro’s definition.
 OUP CORRECTED PROOF – FINAL, 20/2/2022, SPi

    209

of such a distinction between different categories of language use.⁷⁵

Secondly, “vernacular” is a markedly broad and open category. If the
cosmopolitan is defined as being transregional then a number of vernac-
ular traditions—in that they were “not Persian” and “not Sanskrit”—
such as Braj exhibited a similar capacity for movement, albeit on a
smaller scale.⁷⁶ In this section I catalogue the complexity of vernacular-
cosmopolitan literary cross-influence, while in the pages that follow
I offer a theoretical framework for language ideology in pre-colonial
India. Here I challenge the idea that Urdu is exceptional because its
Persian influence somehow set it in a category apart and made it
incommensurate with other vernacular language use in northern
India.⁷⁷ Instead, in the period under discussion, it was a kind of literary
idiom that was seen in the context of other vernacular literary idioms.
Later definitions of Urdu can be distilled down to their colonial and
post-colonial political motivations, which have little relevance for the
pre-colonial period. One influential early twentieth-century guide to
Urdu poetics, Bahr : al-Fas: āhat [Sea of Eloquence], for example, argues
that Urdu is a creole, that is, a mixture of various languages created in the
course of everyday contact.⁷⁸ It also paraphrases the claim made famous

⁷⁵ The tri-partite scheme of the late seventeenth-century Tuhfat : al-Hind has Sanskrit
[sahāskirt] as the divine language used for technical writing, the vernacular [bhākhā] as the
language of “the world in which we live,” and Prakrit [parākrit] as a mixture of Sanskrit with the
vernacular primarily employed for praise poetry. Of course, this definition of Prakrit is
problematic from our perspective, but in the context of an intellectual system in which the
evolution of one language into another was not obvious, this is a logical conclusion. On the
category “deshī,” another term for vernacular, see Pollock 2006, esp. 397ff.
⁷⁶ This is the guiding principle for R. S. McGregor in his summary of the history of Hindi
(McGregor 2003), and has also been painstakingly developed by Allison Busch, for example in
Busch 2003, 2010, 2011.
⁷⁷ George Grierson, the linguist and literature scholar responsible for the Linguistic Survey of
India, implies as much when he claims that the sixteenth and seventeenth centuries saw a
flourishing vernacular literary culture while the eighteenth century was rather barren (Grierson
1889). In order to make such a claim about the exact time when Urdu literature came into its
own, one would have to wrongly assume that Urdu was excluded from the category “vernac-
ular.”
⁷⁸ “Because the Urdu language is made up of several languages together, they call it ‘rek: htah’ ”
: :
[chūn-kih zabān-i urdū kaʾī zabānom se mil kar banī hai is liye is ko rek: htah kahte haim] (Bahr :
al-Fas:āhat, Khan 2006, 1:36; cf. Syed 2012, 295). This falsely implies some sort of a creole
language situation, when in fact the term rek: htah referred specifically to verse that mixed
Persian and the vernacular until it was conflated with “zabān-i urdū.” The historically improb-
able idea that Urdu was a creolized military language (in some tellings deliberately conceived by
the emperor Akbar as part of a syncretic project) is very common among educated South Asians
today.
OUP CORRECTED PROOF – FINAL, 20/2/2022, SPi

210       

:
in Muhammad Husain

Āzād’s declaration that “Everybody knows this
much—that our Urdu language has emerged from Braj Bhasha. And Braj
Bhasha is a purely Indian language.”⁷⁹ Frances Pritchett is right to
ridicule this linguistically dubious assertion, but we should briefly ana-
: origin for
lyze its rhetorical force.⁸⁰ Positing a creole and a Braj Bhāsā
Urdu is a political claim, namely that Urdu is an authentically Indian
language. This is a rhetorical position that became necessary in the mid-
nineteenth century when Hindi, written in the unimpeachably indige-
nous Devanagari script, began commonly to be contrasted with Urdu,
written in the imported Perso-Arabic script. In the eighteenth century,
:
however, all evidence points to the conclusion that rekhtah and other
hindī vernacular literary idioms like Braj were seen as part of the same
genus in a literary eco-system before they were redefined into Urdu and
Hindi. The other important intervention here is to abandon the idea that
the influence of Persian on Urdu made Urdu irreducibly “foreign,” a
thesis which T. Grahame Bailey’s History of Urdu Literature (1932)
clearly illustrates:

What was called “polishing” the language was “Persianising” it; poetry
became more and more artificial and un-Indian . . . . In Urdu every-
thing now became foreign, artificial and exotic. Urdu critics have
themselves often admitted that the old Hindi poets were far truer to
nature.⁸¹

Leaving aside the dated idea that good poetry is necessarily true to
nature, what remains is an overwrought argument that Persian deraci-
nated whatever it touched in India. This is rooted in the colonial
discourse of the delegitimation of the “foreign invader” Mughals rather
than in any indigenous view of language. In the context of the eighteenth

⁷⁹ Āb-i Hayāt

: al-Fas:āhat, Khan 2006, 1:37.
1907, 6; Bahr
⁸⁰ Pritchett 2001. The claim is “linguistically dubious” because khar: ī bolī and Braj have
different grammars, and clearly evolved (even the sparse records of their early history show)
along parallel paths with mutual influence rather than having one branch off from the other.
Nonetheless, Āzād’s position is that Amīr Khusrau
: wrote some dohrās in a mixture of Braj and
Persian, and that this is the ultimate origin of Urdu (Āb-i Hayāt

1907, 67).
⁸¹ Bailey 1932, 40.
 OUP CORRECTED PROOF – FINAL, 20/2/2022, SPi

    211

century, before Hindi and Urdu were defined against each other, the
logic would have been incomprehensible.⁸²
:
Here I consider rekhtah’s literary culture along two axes. First,
I investigate how the vernacular and Persian were seen as compatible.
:
It is self-evident that they were, since rekhtah was developed as verse that
mixed the two, but it is worth tracing the argument in the context of the
eighteenth century and specifically recovering the ways in which Persian
actually influenced the vernacular.⁸³ That the community of rekhtah :
writers was largely coterminous with that of the most influential
Persian litterateurs in Delhi has been established. Second, I describe
the implications of the fact that many of what today we consider separate
dialects and even languages were referred to simply as “zabān-i hindī”
[Indian language] in the eighteenth century. Since rekhtah : (and later
Urdu) as well as the literature of dialects folded into what is today called
“Hindi” are all premised on “zabān-i hindī,” we should take this as an
invitation to view all the vernacular literary dialects in the same frame.
There had been no census or linguistic survey, so judgments about
language existed in a “fuzzy” rather than an “enumerated” life-world,
to use Sudipta Kaviraj’s distinction.⁸⁴ Language use must therefore have
been defined more by function—whether it was suitable for poetry on
particular subjects or for use in communicating with a particular group,

⁸² In the modern period, though, “linguistic nationalism renders the geographical distribu-
tion of the language coterminous with that of the nation itself, so that any attack on the national
language becomes, by substitution, a violation of the national borders, an illicit incursion. It is a
logic that has a great deal in common with the notion of ‘homeopathic magic’ that Frazer
discerned among tribal peoples: if you can suppress its impurities, you can maintain the
essentialist purity of the nation itself” (Herzfeld 1992, 113).
⁸³ Ārzū’s friend Tek

Chand Bahār defines rek: htah as follows in Bahār-i ʿAjam: “It is a thing
which is mixed in its form and that is very felicitously joined together, and discourse mixed with
two or more languages, and this is metaphorical; and ‘rek: htah hemistich’ and ‘rek: htah meaning’
[refer to when] the hemistich and meaning [in question] are found without deliberation and
without effort” [chīzī ast kih dar qālab rīzanad wa ān k: hailī k: hwush
: qimārah mī bāshad wa
kalām mak: hlūt̤ bah do zabān yā zyādah, wa īn majāz ast, wa mis:raʿ-i rek: htah wa maʿnī-yi
rek: htah, mis: raʿ wa maʿnī kih bī-takalluf wa bī-taʾammul yāftah shawad] (2001, 2:1135). Bahār
here references the earlier definition of rek: htah internal to Persian literature, namely an
effortlessly constructed line of verse. Bailey is thus wrong in claiming that the term “rek: htah”
had no “literary significance” before being applied to Urdu (1932, 3).
⁸⁴ See Kaviraj 1992. Cf. Fumaroli 1984, 144 on functional definitions of language in early
modern France.
OUP CORRECTED PROOF – FINAL, 20/2/2022, SPi

212       

for example—than by any other pre-determined criterion.⁸⁵ At some

point the language of Delhi began to be seen as a fit medium for the
ġhazal and other Persianate genres of poetry rather than just a means for
having informal conversations and communicating with people who
were less likely to have access to Persian.
Although every one of Ārzū’s extant works is written in Persian and is
nominally about Persian literature, he mentions Indian vernacular lan-
:
guage in all of his major critical works, and refers to rekhtah by name in
:
at least two, Dād-i Sukhan :
and Tanbīh al-Ġhāfilīn.⁸⁶ Dād-i Sukhan puts
forth a theory of poetic mastery, as we explored in Chapter 4, that takes
into account the fact that mistakes in language use are inevitable. The
end of the first preface, quoted at length here, shows clearly how this
insight is connected to vernacular poetic practice:

Thus in this regard, whatever we have come to call a mistake if it is

[committed] by some person whose standing [in matters of literary
judgment] is above repudiation and acceptance by others, it shall be a
new idiom [dāk: hil-i tas:arruf ],⁸⁷ not a mistake. [Arabic:] Consider this,
for it is a stumbling-block for understanding minds, and furthermore
when considering that this occurs in the practice of the poets of rek: htah
of India (this is poetry in the hindī language of the people of the Court
[urdū] of India, especially in the style of Persian poetry, and it is
presently popular in Hindūstān [i.e., northern India] and formerly it
was current in the Deccan in the language of that country), and I have
seen many leaders [muqtadā] in this art [i.e., of composing rekhtah]:
who have made mistakes in their own idiom, and this made me aware
that as the people who know hindī and Persian are equivalent in their

⁸⁵ “A bi- or multilingual world of interacting language communities is the historical norm

(and the contemporary one in many parts of the globe). The greatest challenge for a historian’s
history of language, in our own largely monoglot environment, is to recreate that world and
explain its workings” (Evans 1998, 18). Such a reconstruction is the project in Orsini 2012.
⁸⁶ I analyzed both works in Chapter 4. Perhaps significantly, both were written in the same
late period in Ārzū’s career, the mid-1740s.
⁸⁷ Literally, “[something which has] entered into [accepted] usage”; an accurate paraphrase
would be “included among authoritative poetic innovations.”
 OUP CORRECTED PROOF – FINAL, 20/2/2022, SPi

    213

circumstances so an error of the Indians can be considered analogous

to that of the Persians.⁸⁸

This argument on the basis of vernacular poetic praxis shows clearly that
the vernacular is not an incidental inclusion in Ārzū’s work but rather
integral to his critical enterprise in his later years. Nawādir al-Alfāz̤ ,
Ārzū’s vernacular lexicon (discussed in Chapter 4), demonstrates the
porousness of language by recording the Indic origins of vocabulary.
Nawādir al-Alfāz̤ is not actually a dictionary of a vernacular language as
such but rather a tabulation of words of vernacular origin used in
Persian. The formal philosophy of language that underpins this possi-
bility of sameness between Persian and vernacular has been considered
in detail in Chapter 2. Thus, Ārzū can consider the concept of simile
[tashbīh] as common to hindī, Persian, and Arabic but note that poets in
each tradition use different similes (for example, only in hindī do poets
refer to eyes as being like fish).⁸⁹
Besides the tendency to consider Persian and rek: htah poetic practices
to a large degree interchangeable, the tradition frequently compares
Urdu poets stylistically to Persian poets. Such comparisons are prevalent
in Urdu verse itself, as for example in the verse in which Mīr compares
himself to Naz̤ īrī, but are also deeply embedded in the critical tradition.
For example, the critic Qatīl writes in his work on Persian composition
Chār Sharbat [Four Cold Drinks, 1217/1802] that Saudā has the same
stature and style in rek: htah that Z̤uhūrī (d. 1616), a great seventeenth-
century poet active in the Deccan, has in Persian.⁹⁰ In Āb-i Hayāt,

Āzād
dutifully invokes a Persian stylistic predecessor for many of the Urdu

⁸⁸ “pas dar īn s:ūrat ān chih mā ān rā ġhalat̤ guftah āmadīm agar az chunān kasī ast kih
pāyah-ash māfūq rad wa qabūl-i dīgarān ast, dāk: hil-i tas: arruf k: hwāhad: būd nah ġhalat̤ . fa-
taʾammal li-ānna-hu min mazālli aqdām al-āfhām wa nīz ba-dān kih naz̤ ar-i īn majrā ahwāl-i :
shuʿarā-yī rek: htah hind ast wa ān shiʿrī ast bah zabān-i hindī-yi ahl-i urdū-yi hind, ġhāliban bah
t̤ arīq-i shiʿr-i fārsī wa ān alhāl
: bisyār rāyaj hindūstān ast wa sābq dar dakkan rawāj dāsht bah
zabān-i hamān mulk, wa mā bisyār kas rā dīdīm kih maqtadā-yi īn fann būdand wa ġhalat̤ dar
:
muhāwarah-yi k: hwud
: : s: il shudah, wa chūn zabāndān-i
kardah-and wa mā rā bar-ān it̤ lāʿ hā
: musāwī ast ġhalat̤ -i hindiyān maqaiyas ʿalaihi fārsiyān tawānad
hindī wa fārsī az ʿālam-i k: hwud
būd” (DS 1974, 7).
⁸⁹ ʿAK 2002, 65.
⁹⁰ As cited by Āzād (Āb-i Hayāt

1907, 155). The tendency to conflate a poet’s Urdu style with
a Persian model is a hallmark of early tażkirahs, although Nikāt al-Shuʿarā does comparatively
less of this (Gorekar 1970, 121). To take a very different example, we can see the same trend in
OUP CORRECTED PROOF – FINAL, 20/2/2022, SPi

214       

poets he discusses, so this was still an accepted mode of criticism into the
1880s. Perhaps the most remarkable comparison of a Persian and a hindī
poet comes from Kh : wushgo
: quoting Ārzū: In the entry in Safīnah-yi
: :
Kh wushgo for Rūdakī, Ārzū’s marginal note says the equivalent of
Rūdakī in hindī is Sūrdās.⁹¹ He invokes the famous blind Braj poet in
order to make the historical claim that just as there are many spurious
compositions attributed to Rūdakī so there are to Sūrdās.⁹²
Why would Ārzū, an important early proponent of literary style that
became Urdu, refer to the language of Gwalior, his native place (at the
northern tip of modern-day Madhya Pradesh), as “the most eloquent”
: h]
[afsa : language of India? After all, zabān-i gwāliyār [lit. the tongue of
Gwalior], which for our purposes we can treat as synonymous with Braj
: was quite different from the language of Delhi that he was
Bhāsā,
promoting as a literary standard.⁹³ Was it the pride of a native son or
is there more to this? In fact, Ārzū’s statement only makes sense if we
assume a cosmopolitan view of literary language, in which the bound-
aries of dialects as we understand them have little role to play.⁹⁴ Recent
research has called into question the idea that a pure “Hindi” literature
was untainted by Persian (and by extension Muslim) culture while Urdu
soaked up Persian. At the edges of the current Hindi canon of pre-
colonial texts cluster works that brim with Persian vocabulary and
imagery. They defy critics’ efforts to categorize literatures through a
prism of Hindi as a Hindu language and Braj as a literary medium for

early modern English letters: Richard Carew’s The Excellencie of the English Tongue (1595)
claimed that Virgil’s style could be found in the Earl of Surrey’s work and that Shakespeare and
Marlowe wrote like Catullus (Vance 1997, 7).
⁹¹ Keshavmurthy 2013, 34–5.
⁹² On the modern scholarly understanding of how the followers of the sixteenth-century
hindī poet Sūrdās wrote poems in his style and signed them as if Sūrdās had written them, see
Hawley 2009.
⁹³ NA 1951, 26. He conflates the language “of Braj and of Gwalior” as “most eloquent” or
“very elegant” at NA 1951, 48. For further discussion of the term “zabān-i gwāliyār” see Busch
2011, 8n13, 121 and McGregor 2001, 21. Amrit Rai mentions but fails to deal with this
conundrum (Rai 1984, 248).
⁹⁴ We can speculate that hindī regionalism functioned somewhat along the lines of the rītis in
Sanskrit. The formal constraints of these styles, such as Vaidarbhī and Gaudī, : are carefully
defined, but they apparently developed from stereotyped regional forms (Chari 1993, 137). They
never lost their nominal attachment to places even though most uses of them by the time literary
scholars defined them had little or nothing to do with the actual usage of the places with which
they were associated. It is entirely possible that the spoken hindī of elites in Gwalior had taken
on characteristics of khar: ī bolī, while still being imagined as being more like Braj Bhās: ā.
 OUP CORRECTED PROOF – FINAL, 20/2/2022, SPi

    215

poems about Krishna.⁹⁵ Take Sāwant Singh (1699–1764), the ruler of

Kishangarh (in present-day Rajasthan) and a Krishna devotee, who
wrote poetry under the name Nāgarīdās. While later Hindi critics have
tried to present him as a Braj poet in the traditional (Hindu) mode, he in
fact wrote a great deal of Persianized hindī verse, clearly in dialogue with
developments in Delhi and the Deccan, and these literary experiments
have been side-lined.⁹⁶ It is the same for the noted patron of the arts ʿAbd
: Khān-i
al-Rahīm : :
Khānān (1556–1627), whose vernacular poetry written
under the chāp [pen-name] Rahīm : breaks later categories.⁹⁷ The Satasaī
[700 Verses] of Bihārīlāl (1595–1663), a court poet at Amber, is full of
Arabic and Persian loanwords.⁹⁸ A work generally taken by modern
critics to be early Urdu, Mirzā Afzal’s

Bikat: Kahānī [A Tremendous
Story, 1636], is in fact difficult to characterize or indeed to trace as an
influence on later Urdu. It contains long phrases that are pure Persian
but also contains earthy hindī expressions and takes the form of an Indic
bārahmāsa [lit. twelve months] genre, which describes the different
characteristics of the lover’s pining for her beloved over the twelve
months of the year. (The lover here is marked as feminine in the
Indian manner rather than being of ambiguous gender as is typical in

⁹⁵ It is telling that the last monograph written in either English or Hindi on the question of
Persian’s influence on Hindi appears to be from 1936 (Vajpeyi 1936). Fascinatingly that text
refers to Rānī Ketakī kī Kahānī [The Story of Queen Ketaki], an odd literary experiment by
:
Inshāʾallāh Khān Inshāʾ, as “Hindi” that is “yet unsurpassed” (Vajpeyi 1936, 101). Inshāʾ
explains his project at the beginning of the work (quoted in Shackle and Snell 1990, 89–90).
The special circumstances, namely the fact that most of Inshāʾ’s literary output now falls into the
category of Urdu and that he invented the style of Rānī Ketakī kī Kahānī as a literary exercise
akin to writing an English work with no Latin-derived words, appear not to matter to critics
grasping for another pre-colonial work to include in the Hindi canon.
⁹⁶ Pauwels 2012. Pauwels has recently found ms evidence of the direct transcreation of a
Persian poem into hindī.
⁹⁷ His poetry has been published in Devanagari script as Rahīm Granthāwalī. On Rahīm, : see
Lefèvre 2014. The Naval Kishore press published a large number of books under the heading
“bhākhā” that we would consider Urdu today (Phukan 2000b, 18n28).
⁹⁸ Dewhurst 1915. A good illustration of how later interventions in the Hindi tradition
attempted to cleanse it of Persianate influence can be found in Thomas Duer Broughton’s
Selections from the Popular Poetry of the Hindoos (1814), the first printed collection of Hindi
verse. Broughton’s informants were mostly Brahmin soldiers who recited poems that strike
modern readers as a mix of khar: ī bolī and Braj forms. Many of the poems that Broughton heard
contained Persian words and Broughton tells us that he suppressed these as inauthentic, even
though his upper-caste informants were unperturbed by this supposed “inauthenticity.” He
apologizes for including two poems in the collection that do in fact contain Persian words that
escaped his notice before publication (Bangha 2000, 49–50).
 OUP CORRECTED PROOF – FINAL, 20/2/2022, SPi

216       

Persian literature.) Shantanu Phukan has argued that the text is “a highly
self-conscious literary undertaking” in contrast to the view of the influ-
ential early twentieth-century critic Mahmud Sherani, who declares the
mixed style distracting and therefore an aesthetic failure (this judgment
is anachronistic in that it takes a modern view of literary genres and
projects them onto the past).⁹⁹ Persian even finds a place in Braj literary
criticism. Bhikhārīdās (fl. 1740) notes that some Braj moves towards
Persian in an “easy Persian” [sahaj pārsī] style but he condemns “foreign-
ers’ language” [jaman bhākhā], which is apparently over-Persianized.¹⁰⁰
As I discuss in the context of rozmarrah, the very same debate took place
at roughly the same time in the context of what is now seen as Urdu.
Nor were poets who are today associated with Persian and Urdu
uninterested in hindī dialects other than Persianized kharī : bolī. Ārzū’s
: s: translated Muhammad
friend Mukhli : Jayasī’s Padmāwat into Persian
prose under the title Hangāmah-yi ʿIshq [The Clamor of Love, 1739].
: s: records that it was his Deccani servant who nar-
Fascinatingly, Mukhli
rated the story to him even though its hindī was “an eastern dialect” (a fact
: s: calls attention), and Mukhli
to which Mukhli : s: himself was a Punjabi.¹⁰¹

⁹⁹ Phukan 2000a, 42. Francesca Orsini’s study on bārahmāsa observes that the people who
wrote them were typically not connected with the Delhi court (Orsini 2010). Orsini’s attempt to
put all bārahmāsas, regardless of whether they were later called “Hindi” or “Urdu,” into the same
frame is commendable, but even so she refers to Afzal’s

work as “Urdu bārahmāsa.” Using that
frame of reference introduces a risk of anachronism because though the grammar is largely khar: ī
bolī, the literary conventions are far from later Urdu practice. For example, it begins with the
:
invocation “Listen, my [female] friends!” [suno sakhiyom] which is common in Braj poetry but
only finds a place in Urdu in the much later rek: htī genre, in which the (male) poet speaks
mockingly in a female voice (Bikat: Kahānī 1979, 24). On the fact that the first recorded
bārahmāsa is actually in Persian (Masʿūd-i Saʿd-i Salmān’s Māh-hā-yi Fārsī [Persian Months,
eleventh century]), see Alam 2003, 145ff.
¹⁰⁰ McGregor 2001, 28–9. McGregor argues that a trend towards Sanskritizing Braj (that is,
preferring tatsama, or unmodified Sanskrit vocabulary, to tadbhava, or Sanskrit-origin words
that have undergone phonetic changes) at this time foreshadows the nineteenth-century split
between Hindi and Urdu.
:
¹⁰¹ See Phukan 2000b, esp. 67–8. Muhammad Shakīr in the seventeenth century copied the
hindī poem, taking care to mark all of the vowels properly, and made an interlinear Persian
translation (Phukan 2000a, 36). McGregor has studied poets who wrote in both Persian and
Awadhī, that is, in “eastern” hindī (1984, 150–4). Tażkirah-i Hu
s:ainī (1875) refers to the poet
Mīr Ġhulām Nabī (who wrote under the pen-name Ġhulām and served in Safdar
Jang’s army in
Awadh) as a prolific writer of dohrās (hindī couplets) and of a “nāyikābhed in rek: htah” (2008,
332–3). Nāyikābhed is a genre of poetry usually associated with Braj and is a description of types
of beloveds. For such a poem to be “in rek: htah” suggests the possibilities that “rek: htah” was used
in this late nineteenth-century text to refer to hindī besides khar: ī bolī or that nāyikābheds were
actually written in khar: ī bolī.
 OUP CORRECTED PROOF – FINAL, 20/2/2022, SPi

    217

Covering south, east, and west, this simple interaction encompasses the
whole of what we now call the “Hindi Belt.” The career of Mīr Ġhulām ʿAlī
Āzād Bilgrāmī (d. 1785) provides the modern researcher with other
surprises.¹⁰² Unusually for a South Asian litterateur he was a serious
poet and scholar in Arabic, but for our purposes what is relevant is his
Sarw-i Āzād [Āzād’s Cypress], a Persian account of 143 Indian-born
Persian poets along with a further eight who wrote in “hindī bhāshā.”¹⁰³
All eight of the poets are Muslim, yet the voluminous quotations show that
the literary medium is closer to Braj (as the label “hindī bhāshā” implies)
than to later Urdu.¹⁰⁴ The inclusion of the Braj tradition in a book devoted
to Persian poets is significant because it implies a parallel between the
literary activities of poets writing in Persian and those writing in hindī,
:
even if that hindī is not the intentionally Persianate rekhtah.
Āzād Bilgrāmī also wrote an Arabic account of nāyikābhed, an Indic
poetic taxonomy of different kinds of beloveds, under the title Subhat: al-
Marjān, and soon translated part of that work into Persian.¹⁰⁵ The
Persian edition is Ġhizlān al-Hind (Gazelles of India, 1178/1764–5),
whose title refers to beloveds as represented in Indic poetry.¹⁰⁶ Āzād
provides the most exact account of Indic figurative language in Persian,
and in so doing provides an argument for the compatibility of various
: al-Marjān attempts to place
literary traditions.¹⁰⁷ The preface of Subhat
India in the Islamic world, both literally and philosophically. The first
part of the main text catalogues “mentions of India in tafsīr [exegetic

¹⁰² He came from a highly learned community in Bilgram, a small town in what is today
central Uttar Pradesh. He studied in the Chishtī silsilah and was well-connected enough to meet
As: af Jāh Niz̤ ām al-Mulk in 1737, receiving financial support from him to undertake the Hajj.


Unusually for an Indian, he wrote extensively in Arabic (both in poetry and in prose) rather
than just in Persian. After leaving Arabia in 1740, he settled in Aurangabad in the Deccan at
As: af Jāh’s invitation. On his biography and Arabic works, see Toorawa 2009.
¹⁰³ Sarw-i Āzād 1913, 352ff.
¹⁰⁴ Āzād Bilgrāmī appears to be the only early modern writer to consider Indian vernacular
poetics in Arabic, as in his Subhat: al-Marjān [The Coral Rosary, 1764]. His Ġhizlān al-Hind
[Gazelles of India, 1178/1764] is fascinating because it is a Persian-language nāyikābhed text,
that is, a catalogue of different archetypes of beloveds that is quite common in hindī. (It was
published in Tehran in 2003 under the probably erroneous title Ġhazālān al-Hind.)
¹⁰⁵ Ernst 1995.
¹⁰⁶ The editor of the published edition, Cyrus Shamisa, regards the title as Ġhazālān al-Hind,
but most other scholars have opted for Ġhizlān al-Hind without the first alif (Sharma 2009,
96n5).
¹⁰⁷ Sunil Sharma has given us a longue durée perspective on where Āzād fits (Sharma 2009).
OUP CORRECTED PROOF – FINAL, 20/2/2022, SPi

218       

commentary] and hadīs̄”: and Āzād notes that a goal of his was to locate
the more obscure references.¹⁰⁸ Some examples from this section appear
in the preface of Ġhizlān al-Hind. Thus, Āzād refers to Adam and his
having taken his first step after being cast out of Paradise on the island of
Serendīb (or what is today Sri Lanka).¹⁰⁹ Adam’s connection to Sri Lanka
is a touchstone of the debate over India as a properly Islamic place and
Āzād states that Adam’s exile to the sub-continent means not that India
was odious [maġhzūb]
(as some scholars had argued) but rather paradi-
siacal and blessed. The second part discusses the ʿulamāʾ [Islamic scho-
lars] of India, and is therefore not relevant for our purposes, while parts
three and four consist of “some Indic rhetorical figures which have been
Arabicized” [baʿzī :

sanāʾiʿ-yi ʿilm-i badīʿ-yi hindī . . . taʿrīb] and “the art of
Nāyikābhed” [fann-i nāyikābhed] itself. These latter sections are what
have been adapted into the Persian Ġhizlān al-Hind, as the first and
second halves of the text, respectively.
The first half is a catalogue of rhetorical figures [s: anāʾiʿ] which are part
of the discipline of ʿilm-i badīʿ. Some examples of figures shared between
Arabic and hindī, he notes in the preface, are īhām, husn : al-taʿlīl, tajāhul
al-ʿārif, marājaʿt, istiʿārah, tashbīh, jinās, and sajʿ.¹¹⁰ The total analysis
comes to sixty-seven specific rhetorical devices: He offers twenty-seven
devices known in hindī and Arabic. Then he describes thirty-five rhe-
torical devices he has invented [muk: htaraʿ] himself, which is to say that
no previous theorist had ever named them as unique devices. He lists two
miscellaneous hindī devices, one of which is attributed to Amīr Khusrau, :
and three “ancient” [qadīm] devices that are specific to Persian (and
hindī), and thus did not appear in the Arabic Subhat : al-Marjān. He uses
the word “tafrīs” to mean describing these terms in Persian. This is also
the term used by Ārzū and other lexicographers to refer to what we
would call lexical borrowing from other languages into Persian. The
second half of Ġhizlān al-Hind is the nāyikābhed proper. As in the
rhetorical figures section, it contains both traditional categories and

¹⁰⁸ Ġhizlān al-Hind 2003, 23ff.

¹⁰⁹ He mentions this again when discussing Indian beloveds (Ġhizlān al-Hind 2003, 58ff.).
¹¹⁰ “baʿzī

mushtarak dar ʿarab wa hind mis̄l-i īhām wa husn
: al-taʿlīl wa tajāhul al-ʿārif wa
marājaʿt wa istiʿārah wa tashbīh wa jinās wa sajʿ” (Ġhizlān al-Hind 2003, 32).
 OUP CORRECTED PROOF – FINAL, 20/2/2022, SPi

    219

some of Āzād’s own invention. The implication is that he wanted it to be

literarily productive rather than a simple catalogue.¹¹¹
Āzād’s rhetoric on the purpose of his addressing Indic poetry is
forcefully cosmopolitan. He says of nāyikābhed (in the context of his
project in Subhāt: al-Marjān) that “this rare offering, which is a speciality
of the Indians, must be given in service of the Arabic of the Arabs.”¹¹²
Whether he literally means that he wants an Arabic-knowing readership
to have some sense of nāyikābhed is unclear but he makes exactly the
same claim of Ġhizlān al-Hind, namely that it will bring the “delightful-
ness of the parrots of India” to “the people of temperament of Iran”
: hib-i
[sā : tabʿ-i furs].¹¹³ In both cases, at least rhetorically, these were ideas
that were meant to circulate. Strikingly, in this account of hindī poetic
practices, he does not actually quote a single line of Indic poetry. Instead,
he offers us quotations of Persian poets that illustrate the argument he is
making about hindī. Indeed, on several occasions he introduces quranic
:
quotations or hadīs̄ in discussing the characteristics of Indic poetry. The
Persian poets he quotes range from Hāfiz̤
:
and Amīr Khusrau on the
classical end of the scale to recent greats like Ġhanī Kashmīrī, Sāʾib,
and
Bedil to contemporary Indian and Iranian poets (including, unsurpris-
ingly, copious quotations from his own poetry). Mīrzā Khān : by contrast
does provide examples in hindī as well as in Persian in Tuhfat : al-Hind.
While Āzād reports the Indic terms for categories of nāyikās (beloveds),
he only gives the names of the literary devices in Persian (or rather
Arabic). Āzād was obviously not concerned with placing these elements
in the context of hindī literary culture in the way that Mīrzā Khān : had
been, but rather his precision in describing the categories is a way of
fixing them as universal.¹¹⁴

¹¹¹ In Āzād’s tażkirah Khizānah-i ʿĀmirah he makes a statement in the same vein that hindī
(or at least Indic ideas) enrich Persian (translated in Alam 2003, 179). Sharma argues that Āzād
depended on an oral recension of nāyikābhed and not a particular author’s text in Sanskrit or
Braj such as Keshavdās’s Rasikapriyā or Sundardās’s Sundarśr̥ṅgār, which in fact has a dedica-
tion to Shāh Jahān (Sharma 2009).
¹¹² “īn armaġhān-i shigarf rā kih mak: hs:ūs: -i hindiyān ast bah k: hidmat-i ʿarab-i ʿarbā bāyad
sipurd” (Ġhizlān al-Hind 2003, 24).
¹¹³ Ġhizlān al-Hind 2003, 24.
¹¹⁴ However, to get at the universal, the non-universal must be addressed. Particularly thorny
is the problem of hindī gender politics, which are the reverse of Persian and Turkish. In Indian
love poetry (here called “taġhazzulāt”), love is portrayed from the perspective of the woman
(Ġhizlān al-Hind 2003, 116ff.). Essentially, argues Āzād, Indian poetry always conceives of a
 OUP CORRECTED PROOF – FINAL, 20/2/2022, SPi

220       

Several broad trends emerge in this complex vernacular environment.

The first is the matter of the circulation of literary people and with them
languages, styles, perceived functions of such styles, and literary tropes. It
:
is suggestive that the rekhtah poet Ābrū, like Ārzū, hailed from Gwalior,
a center for Braj literature.¹¹⁵ Indeed, Prince Aʿz̤ am Shāh, in whose army
Ārzū served, was a patron of poetry in Braj.¹¹⁶ No doubt many other
prominent figures in Delhi’s world of letters also came to the cultural
metropole from peripheral locales where a strong vernacular literary
tradition existed. The Braj country itself was brought firmly into the
Mughal orbit with road connections built to Agra and Delhi during the
Sher Shāh Sūrī interregnum (1538–45).¹¹⁷ With current evidence it is
probably impossible to show the influence of the textual traditions of, for
example, Gwalior on those of Delhi, but certainly poets of the sub-
imperial courts (such as Kishangarh and Amber) were well aware of
what was happening in the capital. As we saw in Chapter 1, Ārzū and
other Delhi-based poets of his time spent time in the Deccan on military
campaigns. Direct evidence is lacking, but the presence of so many
literary elites from Delhi in the Deccan seems an obvious conduit of
influence for Deccani rek: htah on Delhi rek: htah separate from Walī. Also,
poets who had thought of themselves as composing in hindawī (a generic
term for Indian vernacular poetry in use from before Amīr Khusrau’s:
time) began to represent themselves as rek: htah poets specifically.¹¹⁸ Later
scholars, especially linguists, have read this teleologically as a natural
evolution since the khar: ī bolī upon which rek: htah was based would give

woman as having a husband, and her considering him the “treasure of her life” [sarmāyah-yi
zindagī-yi k: hwud
: mī shumurd]. He further explains this by saying that in the Hindu religion
[dar dīn-i hindū] when a woman’s husband dies, she is obliged to commit “satī” (ritual
immolation). Fascinatingly, he argues that women loving men is not far-fetched [istibʿād] on
the basis of the quranic story of Yūsuf and Zulaikhā. : He declares that it is indeed possible for
love to flow in either direction, and to be mutual or not. This discussion is carefully framed with
:
hadīs̄ proving its Islamic acceptability at different points. Furthermore, he quotes Sāʾib

to prove
that Persian poets sometimes take on a feminine persona. However, this discussion never turns
to the vast amount of Braj poetry that describes Krishna as a lover, which is an odd omission—
as far as I can tell one committed by Mīrzā Khān : as well. One plausible explanation has to do
with genre: perhaps such poetry was excluded from the category of “taġhazzulāt.”
¹¹⁵ According to Gardezī (Tażkirah-yi Rek: htah-Goyān 1995, 39).
¹¹⁶ Ziauddin 1935, 3. ¹¹⁷ Pauwels 2009.
¹¹⁸ As in the entry on Nus: ratī in Gulshan-i Guftār (1929, 6–7).
 OUP CORRECTED PROOF – FINAL, 20/2/2022, SPi

    221

rise to modern Hindi and Urdu.¹¹⁹ In fact, this view confuses spoken
language with literary language. Since, as Allison Busch has convincingly
argued, literary Braj spread outside the region in which it resembles the
local spoken language largely because of Mughal patronage—a counter-
intuitive notion to most modern-day Hindi scholars though well sup-
ported by evidence—then when tastes changed and a new literary vehicle
:
became available in the form of rekhtah based on Deccani or Delhi hindī,
this too could similarly have spread beyond where kharī : bolī was com-
monly spoken.¹²⁰ We should also resist the temptation to see the distri-
bution of vernacular literary practice as static. Over the course of as little
as three years, Ārzū’s definition of rekhtah
: shifts. In Tanbīh al-Ġhāfilīn
(c. 1744) he defines it “as poetry in the hindī or Deccani language
agreeable to [the rules of] Persian poetry,” but in the slightly later Dād-
:
i Sukhan (c. 1746) he writes that it was formerly written in the Deccan.¹²¹
While it is of course entirely unlikely that the Deccan’s vernacular
literary tradition died in that short period of time, the need for Ārzū to
express that it was moribund is a notable shift. These movements of taste
and literary influence from metropole to periphery and vice versa need to
be theorized. One framework is that of “lateral” versus “vertical” stand-
ardization.¹²² In the former, metropolitan elites are influenced by pro-
vincial elites and vice versa in matters of taste and patronage, while in the
latter influence moves from elites to the populace and vice versa. In the
case of rek: htah, the lateral model is far better at explaining its rise than
the vertical model.¹²³

¹¹⁹ For example, S. K. Chatterji, who writes: “Braj-bhakha as the direct descendent of the
Śaurasēnī Prakrit, the most elegant Prakrit of the centuries immediately following Christ,
became the dominant literary dialect in the Upper Ganges Valley, and the most cultivated;
and the Muhammadan aristocracy of Northern India also felt its charm and came under its
sway. Delhi Hindusthani had at first very little chance against Braj-bhakha . . . We have seen how
the Hindusthani stands at the end of a chain—how it represents the latest phase in the history of
a Common Language for Aryan India” (1942, 172).
¹²⁰ Busch 2011, 186–8.
¹²¹ “wa rek: htah shʿirī ast bah zabān-i hindī yā dakkanī muwāfiq-i shʿir-i fārsī” (TĠh
1981, 76).
¹²² Lieberman 1997, 482.
¹²³ Thomas Nairn writes of modernity that “The new middle-class intelligentsia had to invite
the masses into history; and the invitation card had to be written in a language they understood”
(quoted in Anderson 1983, 77). Leaving aside the question of whether there was a “middle class”
in late pre-modern India, there does not appear to be any evidence that the political engagement
of the masses was a conscious goal of developing vernacular literary practices.
OUP CORRECTED PROOF – FINAL, 20/2/2022, SPi

222       

Hindī outside of Delhi and the Language of Delhi

Although Ārzū’s position is unclear in this regard, many Delhi intellec-

tuals from subsequent generations explicitly marginalized vernacular
composition from other parts of the sub-continent, particularly the
Deccan, as they argued for a Delhi-based standard language. Delhi
vernacular literary culture was influenced by places where the vernacular
had a stronger social position vis-à-vis Persian. In the Deccan, for
example, the Sultan Qulī Qut̤ b Shāh composed extensively in hindī in
the late sixteenth century so it had a royal imprimatur. To account for
Deccani vernacular literary culture is outside the scope of this study, but
we can take it as given that it was rich and unjustly neglected.¹²⁴ Khar: ī
bolī hindī poetry was also written in Punjab in Delhi’s western hinter-
lands before it had taken hold in the metropole.¹²⁵ Contemporary
sources tried to minimize the significance of this outside influence, as
with Mīr, and there is some evidence that the tradition was keen to
exclude earthier poetry, as hindī poetry in the Deccan tended to be, in
favor of the more abstract, philosophical style that was typical in
Delhi.¹²⁶ However, since even within Ārzū’s lifetime a coherent network
of Urdu writers developed across India, the only convincing explanation
for this rapidity is that there were many people already writing in khar: ī
bolī hindī across India who gained a consciousness of shared linguistic
enterprise when Ārzū and other Delhi-based intellectuals created a
framework for it. Shamsur Rahman Faruqi makes the related points

¹²⁴ On Deccani literature, see Matthews 1976, 1993; Hashmi 2012. “Dakanī” as a term for a
language or literature is first attested in a mas̄nawī in 1645 (Matthews 1993, 82) but it is an open
question as to whether this was a southern hindī that defined itself against northern (i.e., Delhi)
usage.
¹²⁵ Shirani 2005.
¹²⁶ See Petievich 2017 on how some Deccani poetry was thematically more prosaic than the
philosophical, mystical poetry associated with Delhi. One Delhiite who has been largely written
out of the tradition is the viciously satirical Mīr Jaʿfar Zat̤ allī (d. c. 1125–8/1713–16), who
appears in Mīr Nikāt al-Shuʿarā but not in many subsequent compendia likely because he was
too radical and out of step with “serious” poetry. See Jalibi 1984, 2:90–118. The 2003
Zat̤ alnāmah appears to be the first attempt at a critical edition of Zat̤ allī’s poetry, and
Abhishek Kaicker’s readings of several mock-bureaucratic exercises (Kaicker 2014, 336ff.)
represent arguably the best attempt to placing him in a socio-political context.
 OUP CORRECTED PROOF – FINAL, 20/2/2022, SPi

    223

that not only do we lack evidence that Walī, as later critics would
have us believe, learned his craft in Delhi, but also that his work could
not have been sui generis.¹²⁷ In other words, Walī must have been
embedded in a vernacular literary sphere outside of Delhi that produced
poetry that was very similar to the poetry of Delhi. This cosmopolitan
aspect of Urdu literary culture was suppressed in order to cement Delhi’s
claim to it.
Somewhat ironically, the mid-eighteenth-century exodus of poets
from Delhi, often to the greener pastures of Awadh as in Ārzū’s own
case, cemented Delhi’s unique status as the geographical center of
authentic Urdu.¹²⁸ For example, consider the poet Ġhulām Hamadānī
: (d. 1240/1824–5), who was from Amroha (about 130 km west of
Mus: hafī
Delhi) but studied in Delhi and then spent his career in Lucknow.
: apparently made short visits to Delhi and still felt that it alone
Mus: hafī
had a claim on correct language, writing:

Some people think to themselves “We’re the ahl-i zabān”

But if they have not seen Delhi, how can they be language-
knowers [zabān-dānān]?¹²⁹

Urdu authors around the sub-continent promoted their Delhi

connections in the nineteenth century as that attachment, in contrast
to the philological tools available to Persian writers, had generally
become the criterion of authenticity. In the 1807 treatise Daryā-yi
Lat̤ āfat, Inshāʾ redefined sanad so that it no longer referred to a recorded
usage by a previous poet but rather to a more amorphous judgment of
aesthetic soundness by a person from Delhi, specifically someone born
there.¹³⁰ Ironically, having been born in Murshidabad, Inshāʾ failed his
own test of authenticity.

¹²⁷ Faruqi 2001, 138.

¹²⁸ On this migration of poets and the vexed question of describing it as “decline” or not, see
Dudney 2018b.
: : :
¹²⁹ “Baʿzo

: m kā gumān hai yih kih ham ahl-i zabān haim / dillī nahīm dekhī hai zabāndān
kahān haim?” (quoted in Āb-i Hayāt

1907, 313; my translation).
¹³⁰ Inshāʾ’s intervention is discussed with great clarity in Hakala 2016, 89–93.
OUP CORRECTED PROOF – FINAL, 20/2/2022, SPi

224       

Urdu and the Everyday

In Ārzū’s work, there is an assumption that literary practice is universal

even though individual features in literary traditions may be different.¹³¹
:
The practice of rekhtah, with its mixture of Indic language and Persian
poetics, recognizes this by its very existence. Ārzū’s philological method
formally acknowledges the dividing line between the two traditions as
porous, and brings the centuries-old discourse of Persian literary theory
to bear on the nascent Urdu tradition.¹³² Saudā, one of Ārzū’s acquain-
tances (though not officially his student), paraphrases the advice of an
unnamed friend and litterateur, whom we are probably to understand as
Ārzū, in a poem:

No matter what the language, excellence lies in the quality of

:
the theme [khūbī-yi mazmū

m].
Poetry does not depend on Persian alone.
You cannot always use their language correctly;
You should express colorful ideas in your own language.¹³³

The emphasis in later scholarship has been on the supposedly national-

istic aspects of this and other similar conversations recorded for

¹³¹ This section is adapted from a paper first given at the Association of Asian Studies
conference (Dudney 2012) and also appears as Dudney 2018a. See also the discussion of
metaphorical language in Chapter 2. The theory of tawāfuq, which allows for Persian and
Indian literature to be considered in the same frame, is a tool for making literature universal.
While tawāfuq between Persian and hindī applies only to hindī-yi kitābī [i.e., Sanskrit], there is a
conceptual slippage between Sanskrit and the modern Indic languages. Although Ārzū states
that Sanskrit is different from these modern languages, he cites words from both in explaining
tawāfuq, and therefore we can speculate that he understood that Sanskrit was an earlier iteration
of the present languages. Furthermore, the concepts of tafrīs and muhannad (the borrowing of
words into Persian and Indic languages, respectively) formalized the way in which words cross
over the linguistic boundaries.
¹³² Shāh Hātim

recognizes the newness of the practice of rek: htah, pointing out that Walī’s
dīwān was the first “in this art” (i.e., composing in rek: htah) (Dīwānzādah 1975, 39). He ignores
other literary experiments, including the far earlier dīwān of Muhammad : Qulī Qut̤ b Shāh
(b. 1566; r. 1580–1611), the Sultan of Golconda in the Deccan.
:
¹³³ Translated in Naim 1999, 177. The original is “koʾī zabān ho, lāzim hai k: hūbī-yi mazmū

m
: :
/ zabān-i fars pah kuchh munha
: : s: ir sak: ht to
:
nahīm / kahām tak un kī zabān tū durust bolegā /
zabān apnī mem tū bāndh maʿnī-yi rangīm” (quoted in Jalibi 1984, 2:654–5).
 OUP CORRECTED PROOF – FINAL, 20/2/2022, SPi

    225

posterity: In this reading, it is a matter of starting to use “our” language

(hindī/Urdu) rather than “theirs” (Persian) because using Persian is
artificial and forced in India.¹³⁴ Yet the view that Indians as non-native
speakers of Persian cannot compose in it at the same level as native
speakers is contradicted by the quote in its larger context (Saudā tells us a
few lines before that Makīn, the poet correcting his Persian, could find no
mistake in it).¹³⁵ It is also contradicted by Ārzū’s own writings, as we
have seen in Chapter 4,¹³⁶ but a “crisis of confidence” in Indo-Persian has
been taken as self-evident (perhaps under the influence of scholarship on
linguistic nationalism in nineteenth-century Europe). Bracketing off the
question of whether there was such a crisis in the early and mid-
eighteenth century, let us consider the parameters for a dialogue between
Persian and Urdu poetry. The expression “khūbī-yi
: mazmūn”

in Saudā’s
poem is in fact a term of art. Mazmūn

corresponds to “topos” in Western
rhetoric and so Saudā’s unnamed interlocutor is implicitly arguing that
the construction of topoi is fundamental in poetry and that such topoi
transcend individual languages.¹³⁷ Saudā’s use of the expression “k: hūbī-
yi mazmūn”

in an Urdu poem illustrates the point since it is a completely
Persian phrase. Nor is this literary influence uni-directional (moving
only from the some would say overdetermined Persian tradition to the
less developed Urdu tradition). Ārzū argues at length that the use of
Indic words must be allowed in Persian poetry by analogy with Persian’s
own historical borrowing of Arabic and Turkish words and phrases.
Rozmarrah served as a conceptual tool for mediating between Persian
and hindī. It appears in critical writing on poetry and yet is also anchored
to the world of daily experience, therefore forming a junction between
the largely formal realm of Persian and the largely vernacular world of

¹³⁴ A parallel account cited in Āb-i Hayāt

leaves out the crucial idea that well-deployed
mazmūn

is what matters (1907, 142).
¹³⁵ Interpreting the verse is complicated by its circumstances: It comes from a satire referring
to a specific incident, when the somewhat pedantic Makīn unjustly butchered some Persian
verses of Saudā’s that Saudā had submitted to him for correction, and its bleak conclusion that
:
no Indian Persian poets except Khusrau,
Ārzū, and Faqīr were any good is clearly meant
Faizī,
to be read as a satirical exaggeration.
¹³⁶ In particular we can point to Dād-i Suk: han’s second preface (DS 1974, 7–9) and Mus̄mir’s
chapter “dar bayān ānkih ġhalat̤ az ahl-i zabān s: ādir shawad yā na-shawad” [In the matter of
whether a mistake can arise from the (usage of the) ahl-i zabān or not] (M 1991, 34ff.).
¹³⁷ On the history of the term mazmūn,

see Faruqi 2004a, 852ff.
 OUP CORRECTED PROOF – FINAL, 20/2/2022, SPi

226       

hindī in the eighteenth century. In modern Urdu usage, the meaning of

rozmarrah is almost always “colloquial” or “everyday” language as
opposed to formal language.¹³⁸ On the other hand, in modern Iranian
Persian, it usually refers to a daily allowance or occurrence.¹³⁹ In Persian
in the eighteenth century and before, it meant both, as illustrated by
Bahār-i ʿAjam, the dictionary compiled by Ārzū’s friend Tek
Chand
Bahār.¹⁴⁰ Bahār writes:

rozmarrah: This word is used in two situations: the first in the meaning
of idioms and words that are well-known among the people and the
other in the meaning of ration and obligation of victuals, derived from
“day” [roz] and “marrah” which is an Arabic word in the meaning of a
time/turn, that is, what one receives daily and that which one says [bar
zabān bu-gużarad] daily. Thus the word is shown to be not originally
Persian.¹⁴¹

The difference between Urdu and traditional northern Indian literary

dialects like Braj is that Urdu grammar is based on what later came to be
known as khar: ī bolī, the actual spoken hindī dialect of Delhi.¹⁴² Poetry,
which is what concerns us here, is obviously a linguistic domain bound
by precise rules and is by definition not the prose of normal, everyday
communication. And yet it has the notion of conversation built into it:
:
Sukhan, literally “speech,” is used throughout the Persian tradition as a
metonym for poetry.¹⁴³ One does not typically “write” poetry in classical

¹³⁸ The title of Shamsur Rahman Faruqi’s dictionary, Luġhāt-i Rozmarrah, is a case in point.
¹³⁹ For example, in Hayyim 1934–6. ¹⁴⁰ Rieu 1879–83, 2:502.
¹⁴¹ “rozmarrah: īn lafz̤ rā dar do mauzaʿ :

istiʿmāl kunand: yakī bah maʿnī-yi muhāwarāt wa
alfāz̤ -i mashhūrah bain-i al-nās wa dīgar bah maʿnī-yi rātbah wa wajh-i maʿāsh, murakkab az
‘roz’ wa ‘marrah’ kih lafz̤ -i ʿarabī ast bah maʿnī-yi bār, yaʿnī ānchih har roz bah yak bār barasad
wa ānchih har roz bah yak bār bar zabān bugużarad. pas lafz̤ mustahdas̄ : bāshad nah fārsī as:l”
(Bahār-i ʿAjam 2001, 2:1114). It is a concept that is far older than the eighteenth century, cf.
Madār al-Afāzil
(1001/1592): “rozmarrah (f [ārsī]): ānchih bā-ū rozgār gużarānand wa nīz ʿurf-i
: chunānkih goyand zabān-i rozmarrah” [rozmarrah (Persian): that which one uses daily as
hāl
well as present common usage which they call “everyday language” [zabān-i rozmarrah]]
(1959–70, 2:336).
¹⁴² Khar: ī bolī also happens to be the basis for Modern Standard Hindi. The later history of
the differentiation of Hindi and Urdu (as described, for example, in King 1994) is outside the
scope of this study.
¹⁴³ Faruqi has helpfully quoted Bedil’s philosophical definitions of “suk: han” (Faruqi
2004b, 19).
 OUP CORRECTED PROOF – FINAL, 20/2/2022, SPi

    227

Persian but rather one “speaks” it [shiʿr guftan]. Similarly, the locus of
poetic appreciation (at least in South Asia) was the literary gathering, or
mushāʿirah, an oral performance involving a great deal of audience
interaction.¹⁴⁴ Many poets had a bayāz
[notebook] in which they
recorded appealing poems from mushāʿirahs and selections from these
oral records often circulated in parallel to or in lieu of the “publication”
of poems in a poet’s own curated dīwān [selected works]. However, we
do not know the socio-linguistic specifics of the pre-colonial mushāʿirah:
For example, how much hindī was spoken at a Persian mushāʿirah in
Delhi and in what context?
The later critical consensus, as we have seen in Chapter 3, has been
that Indo-Persian writers sought complexity to the exclusion of compre-
hensibility. Although creating complex imagery was an often-stated goal
of the tāzah-goʾī poets, comprehensibility as measured against the roz-
marrah was in fact also a contemporary concern. Let us first return to the
three prefaces of Ārzū’s Dād-i Suk: han. In the third preface, Ārzū distin-
guishes between two kinds of poetic interpretation:¹⁴⁵ The first is the
“path of the common people [who know] the language” [t̤ arīq-i ʿāmmah-
yi ahl-i zabān]. In some contexts “ahl-i zabān” (literally “people of the
tongue”) refers to native speakers but much of the time it simply means
people who use the language competently—we could think of them as
the “community of language users.”¹⁴⁶ These “ahl-i zabān,” according to
Ārzū, understand the meaning of words and the common interpretations
that they have heard from their elders. Ārzū writes that “both common
people and experts share in this interpretation” [“dar īn t̤ arīq ʿawāmm
wa k: hawās: s: sharīk-and”] but he cautions that this shallow reading is not
the last word. The second kind of interpretation is that to which only true
experts have access. He goes on to mock a number of so-called experts
like schoolmasters or people obsessed with metaphor to the exclusion of
other kinds of interpretation. Ārzū’s argument is simple: Knowing a
language is a prerequisite for interpreting and composing poetry, but
the real work cannot begin until someone masters poetic interpretation.

¹⁴⁴ See Naim 1989 and Zaidi 1989. ¹⁴⁵ DS 1974, 9ff.
¹⁴⁶ When writers wish to refer unambiguously to Iranian native speakers, as does Wārastah,
they use a phrase like “muhāwarah-dān-i
: īrān” [lit. “idiom-knowers of Iran”].
OUP CORRECTED PROOF – FINAL, 20/2/2022, SPi

228       

Ārzū’s complaint against the mullās [religious educators] is telling. He

writes that their comprehension of poetry is “other than that of the
people of the rozmarrah.”¹⁴⁷ In other words, their reading of poetry is
casuistry rather than a commonsense understanding of how the language
is actually used. Indeed, Ārzū’s first preface begins with a slightly tauto-
logical invocation of the rozmarrah. He writes, “Of that which is current
[wāqīʿ] or not for the people of the rozmarrah, it is mostly that which is
current.”¹⁴⁸ What does this mysterious pronouncement mean? Simply
that most poetic rules follow natural speech (with the exception of
metrical requirements for certain words, which Ārzū admits trip people
up). In the third preface, he describes interpretation “according to the
taste of the poets” [muwāfiq-i mażāq-i shuʿarā] and argues that it
depends on comparing one’s own rozmarrah with that of the poet in
order to find the particularities in the poet’s language.¹⁴⁹
A similar concern for poetry’s necessary relationship with the rozmar-
rah presents itself in Shāh Hātim’s

preface to Dīwānzādah. Examining
what Shāh Hātim

says, before jumping to conclusions about his inten-
tions, is important: He rejects “the hindawī which they call ‘bhākhā’”¹⁵⁰
(in other words, hindī dialects like Braj) in favor of “the rozmarrah of
Delhi.” More specifically, he states that he “has chosen purely the
rozmarrah which is understood by common people and acceptable to
experts” (we can note the parallel to Ārzū’s invocation of common
people and experts). Thus, he is arguing not against some kind of
“Hindu language” but rather for the Delhi rozmarrah, which Braj is
patently not.¹⁵¹ On the other end of the cultural spectrum, he condemns
poets who use Persian clumsily in their rek: htah. He lists “dar,” “bar,”
“az,” and “ū” as examples of Persian words that should not be used in the
vernacular. The first three are prepositions and the last is the third-
person singular pronoun (in Persian grammar they are all known as
“harf
: ” or what we would call an indeclinable particle). He approvingly

¹⁴⁷ “ān ġhair-i fahmīd-i ahl-i rozmarrah bāshad” (DS 1974, 10). ¹⁴⁸ DS 1974, 2.
¹⁴⁹ DS 1974, 12.
¹⁵⁰ “hindawī kih ān rā bhākhā goyand mauqūf kardah” (Dīwānzādah 1975, 40).
¹⁵¹ The Braj poetic tradition represents its language as the everyday speech of the rustic Braj
country, but this is undermined by its use in a transregional, cosmopolitan literary tradition.
 OUP CORRECTED PROOF – FINAL, 20/2/2022, SPi

    229

cites an Urdu poem by his contemporary Shāh Mubārak Ābrū mocking

people who use the wrong sort of Persian in their Urdu:

The ones whose time is spent in rek: htah poetry,

To them I say: Mark my words that
Whoever brings Persian verbs or particles into rek: htah
:
His deeds will be trifling; his rekhtah verses will be
questionable.
:
:
[waqt jin kā rekhte kī shāʿirī mem s: arf hai
:
un sitī kahtā hūm būjho harf: merā z̄arf hai
:
jo kih lāwe rek: hte mem fārsī ke faʿl-o harf
: ¹⁵²
: :
laġhw haim-ge faʿl us ke rek: hte mem harf: hai.]¹⁵³

Thus, not only are the conventions of Braj poetry to be eschewed, but so
is using Persian in a forced, artificial way. This is crucial in the self-
definition of Urdu literary culture.
The importance of the invocation of the colloquial in the development
of Urdu poetry has been noted before, for example by Shamsur Rahman
Faruqi regarding Mīr.¹⁵⁴ Managing the colloquial was an aesthetic

¹⁵² When Mīr enumerates the kinds of rek: htah in the conclusion [k: hātimah] of Nikāt al-
Shuʿarā, he notes that what he defines as the third kind of rek: htah, in which Persian verbs and
particles appear, is “unaesthetic” [sīwum ānkih harf: wa faʿl-i pārsī ba-kār mī-burdand wa īn
qabīh: ast] (Nikāt al-Shuʿarā 1979, 161). The Arabic technical use of the word “qabīh” and other
terms for proper and improper expressions is explored in Marogy 2010, 53–4. In Mīr’s
classification, the first two kinds of rek: htah are linguistically mixed in a formal way, the first
:
being in the style typical of works attributed to Amīr Khusrau, namely with one line of the
couplet in Persian and one in “hindī,” while the second has each language alternate at the middle
of each line. The fourth we consider later in the chapter. The fifth and sixth are “īhām” and
“andāz.” Mīr declares himself a poet of “andāz.”
¹⁵³ The word-play on “harf”
: is notable: In the second line, it seems to be used along the lines
of the Persian idiom “harf
: zadan” [lit. “to strike a word”] meaning “to speak.” In the third,
context tells us that it is being used in the technical grammatical sense of an indeclinable
particle. The fourth line uses an idiom defined by Platts as follows: “harf: honā (-par), To be a
stigma, stain, spot, or disgrace (upon); to be derogatory (to).” Likewise, “faʿl” is used in its
grammatical sense as “verb” in line three but is made concrete in line four as “action.” Thanks to
Frances Pritchett for her advice on translating this passage.
¹⁵⁴ In Shʿir-i Shorangez (Faruqi 1990a, 57ff., with condensed translation by Frances Pritchett at
www.columbia.edu/itc/mealac/pritchett/00garden/about/txt_srf_mir_ghalib.html). Muhammad :
Husain

Āzād admires Mīr because his poetry is supposedly similar to actual speech, which
makes it close to nature [nechar] (Āb-i Hayāt

1907, 202). This obviously fits into Āzād’s aesthetic
program nicely but he must also explain Mīr’s borrowings from “artificial” Persian (Āb-i Hayāt

1907, 203ff.).
OUP CORRECTED PROOF – FINAL, 20/2/2022, SPi

230       

problem that concerned Mīr himself. In Nikāt al-Shuʿarā, he notes in the

:
conclusion [khātimah] that:

The fourth [style of rek: htah] is that which they adorn with Persian
constructions [tarkībāt]; often a construction which is conformable to
:
the rekhtah dialect appears and that is allowed, and those other than
poets cannot judge it; a construction which is not familiar in rek: htah
[i.e., which does not seem to fit] is faulty and judging [lit. “knowing”] it
is likewise based on the good taste of a poet. The preference of this
wretch [i.e., the author] is the same: If a Persian form is acceptable to
conversation [guftagū] in rek: htah then there is no harm.¹⁵⁵

In other words, if a Persian expression is already naturalized in rekhtah—:

that is, if it is part of the rozmarrah (although in this case Mīr uses the
similar term guftagū)—then it is automatically acceptable. If not, then a
poet’s judgment determines whether it is good rekhtah : or an unwanted
intrusion of Persian. The debate over the precise amount of Persian
allowed in Urdu continued into the mid-nineteenth century. Sir Sayyid
:
Ahmad :
Khān (1817–98), for example, notes that it is for the ahl-i zabān
to decide.¹⁵⁶
Theorizing the colloquial is not a peculiarity of Urdu literary culture
but was adopted from the Persian criticism of the eighteenth century.
Critics of this period, including Ārzū, frequently comment on whether a
Persian verse follows rozmarrah or not. For example, in Dād-i Sukhan :
associating “hukm”
: [command] with “t̤ uġhrā” [a seal] is called into
question as a problem of rozmarrah since “t̤ uġhrā” is connected with
“farmān” [another kind of command] and not with “hukm.”¹⁵⁷ : It is
mentioned frequently in ʿAt̤ īyah-i Kubra̍, his treatise on “ʿilm-i bayān”
[rhetoric], and Mauhibat-i ʿUz̤ ma̍, his treatise on “ʿilm-i maʿānī”

¹⁵⁵ “chahārum ānkih tarkībāt-i fārsī mī ārand, aks̄ar tarkībī kih munāsib-i zabān-i rek: htah
mī uftad ān jāʾiz wa īn rā ġhair shāʿir na-mī dānand wa tarkībī kih nāmānūs-i rek: htah mī
bāshad ān maʿyūb ast wa dānistan-i īn nīz mauqūf-i salīqah-yi shāʿir ast wa muk: htar-i faqīr ham
:
hamīn ast. agar tarkīb-i fārsī muwāfiq-i guftagū-yi rekhtah būd muzāyaqah

na-dārad” (1979,
161; cf. Naim 1999, 179).
¹⁵⁶ Lelyveld 2011. ¹⁵⁷ DS 1974, 52.
 OUP CORRECTED PROOF – FINAL, 20/2/2022, SPi

    231

[semiotics].¹⁵⁸ He invokes “rozmarrah-dānān” [rozmarrah-knowers] as

judges of whether a metaphor has been properly used.¹⁵⁹ Mocking the
rozmarrah of others was fair game for satirists, such as Mīr Yahya̍: Kāshī
(d. 1653), who apparently identified with Shiraz despite his name
(“Kāshī” means “from Kashan,” a city near Isfahan) and wrote some
vicious lines, quoted by Ārzū, about the speech of Kashan.¹⁶⁰
The apposition of “rozmarrah” and “muhāwarah,”
: two terms which
occupy nearly the same semantic range in this period, is instructive.
“Muhāwarah”
: is usually translated as “idiom” and is used more fre-
quently than “rozmarrah.” Shāh Hātim
uses the two in a telling contrast
since he mentions the rozmarrah of Delhi that people have in their
:
muhāwarah, so for him the latter is clearly a broader category than the
former. Since they are so similar in meaning, drawing sharp distinctions
between the two words is difficult; but perhaps the difference between
them is akin to what we would call diachronic and synchronic analysis.
That is to say that rozmarrah is conceptually the usage that is current at
one time, namely the present, and so a diachronic phenomenon, while
:
muhāwarah can refer to a set of usages over time and therefore be a
synchronic phenomenon. Thus, while muhāwarah : is sometimes used
with a historical reference, rozmarrah is not used in that way, at least in
the texts examined here.¹⁶¹ Ārzū, for example, refers to the “rozmarrah of
the Persians” [rozmarrah-yi fārsiyān] in order to address the differences
between current usages in Indo-Persian and Iranian Persian.¹⁶² This is
obviously a diachronic rather than a synchronic comparison, because it
is concerned with variations over space and not over time. Crucially,
rozmarrah is not an observed speech pattern as we would expect in a
modern linguistic survey—it is always useful to register the differences
between the philology of this period and our present-day linguistics, lest

¹⁵⁸ The glosses of the two terms should be understood as approximate. The texts were
published together in 2002, edited by Cyrus Shamisa. Further context can be found in
Rahimpoor 2008a.
¹⁵⁹ ʿAK 2002, 53, 67, 91; MʿU 2002, 99, 125, 135, 136, 181. ¹⁶⁰ M 1991, 5–6.
:
¹⁶¹ For example in KhG, :
where he contrasts the muhāwarah :
of different periods (KhG 1996,
15, 16). He does refer to “current rozmarrah” [rozmarrah-yi hāl]
: in MʿU, which could imply the
possibility of a “non-current rozmarrah,” but since such a historical formulation never appears
it is safe to assume that “current” is redundant in this context.
¹⁶² M 1991, 38. He also notes that the “people of the rozmarrah” [ahl-i rozmarrah] criticize
Sāʾib (d. 1676), who had by this time just passed out of living memory (M 1991, 79).
OUP CORRECTED PROOF – FINAL, 20/2/2022, SPi

232       

we allow ahistorical expectations to creep into our analysis. Today we

separate gathering linguistic data from judgments about proper usage,
but this was not the case in eighteenth-century Persian, and for
eighteenth-century literati like Ārzū, Hātim

and Mīr, rozmarrah was
subject to criticism.¹⁶³
There is an intriguing possibility that rozmarrah became established as
a concept in Persian criticism because of contact with vernacular litera-
ture, in the same way that Latin literary culture was retooled during the
Renaissance partially in response to vernacular literary movements for
which Latin literature had itself provided the basis. Persian and Urdu
literary culture co-existed for decades and it is important for us to
understand as thoroughly as possible the contours of each. Not only
did Indian vernacular languages absorb a great deal of vocabulary from
Persian, but Persian itself had pulled in Indic words for centuries, as
Ārzū takes pains to elucidate in Mus̄mir.

Revisiting the Question of the “Unprivileged Power”

of Indo-Persian

We know how the story ends: By the mid-twentieth century, Persian had
retreated into the madrasa and a few university departments, while Hindi
and Urdu had become national languages in India and Pakistan. But was
it a failure on the part of Indo-Persian that led to the rise of Urdu? There
is plenty of evidence that Ārzū and his circle saw themselves as equals of
Iranians in matters of Persian style, and this appears to have been the
dominant view well into the mid-eighteenth century. By the nineteenth
century, of course, the situation had changed. The Delhi College Persian
professor Imām Bakhsh: Sahbāʾī

(1802–57) declares in Qaul-i Fais: al [The
Definitive Assertion], the last major traditional work to engage with the
conflict between Ārzū and Hazīn,
that since Indians are not native

¹⁶³ Besides Ārzū’s thoughts on the matter, explored elsewhere, Shāh Hātim

mentions “ġhalat̤ ī
rozmarrah” [erroneous rozmarrah]. The preface of Ārzū’s CH justifies its composition by
noting that although a word’s meaning may be well known, there can still be confusion “in
the rozmarrah of the eloquent” [rozmarrah-yi fus: ahā-yi
: ahl-i zabān] as to its correctness. On
CH see Chapter 4.
 OUP CORRECTED PROOF – FINAL, 20/2/2022, SPi

    233

speakers, everything elegant in Persian necessarily comes from Iran.¹⁶⁴

Faruqi has identified a pervasive value structure in which certain urbane
Indians in the nineteenth century privileged Persian written by Iranians
who had never come to India over Persian written by Iranians who had
come to India. They barely tolerated Persian written by Indians, but even
that supposedly degenerate Persian was better than vernacular writing.¹⁶⁵
Such a hierarchy was not in place during Urdu’s formative period in the
latter half of Ārzū’s life, and Indo-Persian was not under any threat
recognized by Ārzū and his circle, either from Iran or from the vernac-
ular. The nineteenth century then is where we must look for forceful
declarations of the utility and moral good of the vernacular as opposed to
Persian. By then the question is tied up with colonialism and nationalism
in a way that it could never have been in Ārzū’s time.¹⁶⁶
The central issue, to which I have returned again and again in this
chapter, is the protean nature of hindī/hindawī/rek: htah/Hindi/Urdu. If
we consider the relationship of the vernacular and Persian as that of an
innate language versus a learned language (as in Dante’s formulation of
“prima locutio” versus “locutio secondaria”) then this framework is
complicated by the fact that rek: htah was not an “innate” language as
such. Rather it was a consciously taught literary style of an innate
language, which was not necessarily the native tongue of all of its
practitioners. Khar: ī bolī had spread well beyond Delhi by the eighteenth
century, but it was the spark of recognition that it could be a literary
language defined in a particular way that allowed it to become a literary
language associated with the prestige of the imperial court. That this
happened in the milieu of the master-poets of the cosmopolitan language
is in line with what we know of Europe (where the process took several
centuries longer), and is not a cause for alarm over the “artificiality” of
the vernacular. Most languages undergo some sort of language planning,
which is to say a deliberate intervention in a language’s development.

¹⁶⁴ Quoted in Shamisa 2002, 41. ¹⁶⁵ Faruqi 1998.

¹⁶⁶ Although I hesitate to raise a topic that cannot be covered fully here, it is important to
note that Persian did not “die” in India in the eighteenth century but endured as a marker of
elite Indian identity into the twentieth century even as its practical use waned. The ideology
around late Indo-Persian cannot therefore be separated from the colonial encounter. There were
factions in favor and against replacing Persian both among British administrators and Indian
elites in the nineteenth century (see, for example, King 1994, 53–79; Mir 2006).
OUP CORRECTED PROOF – FINAL, 20/2/2022, SPi

234       

This chapter has narrowly focused on a mid-eighteenth-century

moment, insisting that we not make conclusions about linguistic identity
for which there is not contemporary evidence. Recognizing Urdu as
a language available for all uses was a late eighteenth- and early
nineteenth-century development in which rekhtah : practice was conflated
with the Persianized kharī : bolī of everyday life in Delhi. Consolidating
poetic norms and raising a literary style to new prestige actuated a
network of literate kharī : bolī users across South Asia who now could
identify as users of the same language rather than merely practitioners of
the same literary style. Ārzū’s insight that the vernacular could work like
Persian and his influence over the literary community in Delhi were
crucial, but he did not set out to banish Persian from India or even
necessarily to access this network. It was later generations who mobilized
around language identities, simultaneously distancing Persian from
Indian experience (by buying into the idea that Persian is the cultural
patrimony of Iranians), and raising Urdu and later Hindi to the level of
national languages.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

6
How Language Actually Works
Contrasting Europe and the Non-West

The biblical story of the Tower of Babel (Genesis 11:1–9) describes how
humankind, speaking only one language and therefore able to organize
efficiently, began to construct a tower that threatened God’s dominion.
To put a stop to the project, “the Lord did there confound the language of
all the earth” and scattered the arrogant tower-builders. In the Judeo-
Christian tradition, the diversity of human languages was thus often
interpreted as a divine punishment. The Qurʾān also briefly relates a
story along a broadly similar theme of humans disrespecting the Deity
through monumental architecture: Pharaoh asks for a tower to be built
up to the heavens so that he might see God for himself. However, in that
telling there is no mention of language as either enabling the tower’s
construction or being part of the punishment.¹ Like their forebears,
eighteenth-century Europeans still took the confusio linguarum (“mixing
up of languages”) described in Genesis 11 as the logical starting point for
understanding language diversity and traced language families and
nations through the prophet Noah’s sons Japhet, Shem, and Ham.²
(The fact that we still refer to the “Semitic” language family is a vestige
of this system.) However, they also began to amplify an idea that every
nation had a language that singly and properly expressed the character of
its people.³ This view, which was not novel in the eighteenth century but

¹ Qurʾān 40:36–7. ² See Campbell and Poser 2008, esp. chs 2 and 3.
³ Saint Augustine (354–430) considered the profusion of languages in the world as an
unfortunate separation of people and a reminder of humankind’s rightful punishment by
God. He viewed the Roman Empire’s consolidation of (as he saw it) the civilized world under
a single language, Latin, as a remedy (O’Daly 1999, 201). After the Roman Empire, Europeans
only began to take monolingualism for granted again after around 1800 (Evans 1998, 27).
Dante’s distinction between innate languages [prima locutio, lit. primary speech] and languages
whose grammatical rules had to be learned academically [locutio secondaria, lit. secondary

:
India in the Persian World of Letters: Khān-i Ārzū among the Eighteenth-Century Philologists. Arthur Dudney,
Oxford University Press. © Arthur Dudney 2022. DOI: 10.1093/oso/9780192857415.003.0007
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

236       

had previously been limited by its poor fit with the confusio linguarum
narrative, crystallized into the keystone of nineteenth-century national-
ism and offered a structure for linguistics as it developed into a modern
discipline. The bundle of processes by which we came to our present-day
understanding of languages and nations began with a multilingual world
and plainly we still live in a world with a multiplicity of languages.
However, our multilingualism is not their multilingualism. Yasemin
Yildiz has coined the phrase “postmonolingual condition” to highlight
the differences in how people have understood language before and after
the divide of modernity/colonialism.⁴ When faced with an unfamiliar
cosmopolitan linguistic situation, postmodern people cannot but look
for the specter of the monolingual possibility that failed.⁵ (For all the
multilingualism in present-day India, the hand-wringing over the need
for Hindi to be a singular “national” language shows that at least the
expectation of monolingualism now affects even highly multilingual
societies.) Because the modern view of language, like the concept of
modernity, represents itself as universal when in fact it was, at least at
some point, a parochial European perspective, the arguments about
eighteenth-century Persianate South Asia presented in this study are
not complete without some discussion of Western parallels and, in
particular, discontinuities.⁶
In examining Ārzū’s career, I have argued for the historical contin-
gency of several concepts that often seem universal from our position on

speech] had stood for centuries, but was only widely operative from this surprisingly late date
(see Farrell 2001, 16). Friedrich Schleiermacher (1768–1834) expressed a view typical of the
period: “To one language alone does the individual belong entirely.” His rough contemporary
Johann Herder (1744–1803) argued that “because every people is a people, it has its national
culture like its language” [Denn jedes Volk ist Volk; es hat seine nationale Bildung wie seine
Sprache] (quoted in Anderson 1983, 66).
⁴ Yildiz 2012, 5–6. Yildiz argues that although “lingua materna” (mother tongue) was a
medieval Latin term, it was not “emotionalized” until the late eighteenth century.
⁵ The linguist Michel Degraff has observed “creole exceptionalism,” which he defines as the
false “set of beliefs, widespread among both linguists and nonlinguists, that Creole languages
form an exceptional class on phylogenetic and/or typological grounds” (Degraff 2005, 533). For
a parallel in medievalists’ difficult relationship with mixed-language medieval texts, see Schendl
2000, 79.
⁶ On modernity’s false universalism, see Chakrabarty 2000, ch. 1 and Goody 2006. On the
importance of not projecting the nation as a conceptual category onto the past and the difficulty
of doing so within disciplinary structures that are premised around it (in this case a “post-
colonial” analysis of the Middle Ages), see Gaunt 2009.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

    237

the arc of world history. The underpinnings of eighteenth-century

Indo-Persian philology—that is, Ārzū’s society’s assumptions about
the nature of language and how best to analyze it—are emphatically
different from those of modern linguistics or even lay people’s experi-
ence of language and its relationship to identity in the twenty-first
century. The conception of nation that can be implicated in the
fracturing of the Persian cosmopolis had not yet been articulated in
the early eighteenth century. Indeed, the now frequently invoked idea
that it was national pride that caused Indians to throw off the yoke of
“foreign” Persian in the eighteenth century sounds suspiciously like the
anti-colonial nationalism of a much later era. The twin distortions of
colonialism and modernity have made it exceptionally difficult for us to
interpret the pre-colonial past except as filtered through their univer-
salizing tendencies. This chapter turns to the possibility of comparat-
ism and argues that early modern textual knowledge systems in Europe
and the Persian cosmopolis can be understood in dialogue with each
other. Twenty-first-century scholars must ourselves construct this dia-
logue, since it is obviously not available in the sources, but it is
nonetheless an important interpretative tool. By treating the concept
of literary humanism as a universal baseline, we can identify parallel
intellectual trends in Europe and South Asia—the key difference is that
while these developments are generally valorized in the European
context as steps towards modernity and the nation-state, in the non-
Western context they have been adduced as evidence of socio-political
decline. When viewed together, these contradictory interpretations
throw into relief a problematic trend in much of the available histori-
ography for both the West and South Asia. It is also crucial to recog-
nize that eighteenth-century Persianate approaches to understanding
language are foreign to us not only because Orientalist scholarship
separated East and West, but even more because of the divide between
modernity and what came before: While Europe’s eighteenth-century
scholarship was swept into the story of modernity, which allows the
infelicities to be explained away as minor diversions from progress,
eighteenth-century South Asian scholarship leads mostly to the historical
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

238       

dead-end wrought by colonial domination.⁷ Paring back the teleology in

the analysis, early modern European views on and experiences of language
are more alike to Persianate views of the same period than they are to
modern ones.

Language and Early Modern Thought

Scholarship that theorizes early modernity has often excluded cultural

questions, instead considering economic and political structures that
might characterize the period. This section considers the parallels
between early modern European and South Asian philological know-
ledge systems. The intellectual environment in both Europe and South
Asia during the early modern period is arguably distinct from other
historical periods because of the prevalence of a particular approach to
knowledge: Intellectuals sought to extend old categories of knowledge
through radical new approaches without replacing the old categories
themselves. Furthermore, rhetoric had not been dethroned in either
place and thus bound together a variety of disciplines that modernity
has split apart. The difficulty at the outset of such an analysis is that
modernity in South Asia is typically connected with the colonial epi-
stemic break.⁸ However, it is obviously important to take stock of what
pre-dated colonialism. We find that a great deal of what Europeans took
credit for bringing to India as part of “modernization” was there already
in forms that Europeans either did not understand or did not want to
understand. “Early modern” is inherently a teleological term—this is of
course problematic from the perspective of historiography because
people in the early modern period, whether in Europe or elsewhere,
could not peer into the future and see modernity for themselves—but

⁷ In this regard, it is useful to observe the distinction between “modernization” and

“modernity,” the former being the process of developing modern institutions and the latter
the emergence “of a degree of reflective, judgmental thinking about these processes”
(Chakrabarty 2011, 669).
⁸ The problem is nicely posed in Kaviraj 2005. For Partha Chatterjee, early modernity in
South Asian history is the precursor to the “colonial modern” that begins in the 1830s
(Chatterjee 2012, 75–6). This somewhat confuses the issue of comparative early modernities
because it is at once teleological and anti-teleological.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

    239

it should not be taken to imply that anything that did not directly
contribute to (European) modernity should be excluded from consider-
ation. South Asian early modernity is not a failure but a road not taken
for various historically contingent reasons, the most important of which
was colonialism itself. The project of finding Europe’s fundamental
uniqueness that allowed “the Great Divergence,” in which a handful of
European states became rich, unified, and technologically advanced
enough to rule much of the world, only makes sense (and breaks free
of its imperialist roots) if we are prepared to investigate how Europe was
not unique.
Whether the term “early modernity” is useful in the first place has
sparked an academic debate that is worth outlining. The idea of mod-
ernity popular in the 1950s and 1960s, namely that it is based on global
“convergence,” has lost its luster. As Shmuel N. Eisenstadt and Wolfgang
Schluchter argued in an influential special issue of Daedalus, a new
strategy is to think about “multiple modernities” and how the underlying
pre-modern society has made these permutations of modernity.⁹ They
argue that every society will develop some kind of public sphere (an idea
that I take up later in the chapter). This is a proper inquiry for the sort of
intellectual history that concerns us here, but often historians with a
stake in the term “early modern” have not meaningfully considered
cultural production.¹⁰ An intellectual history of early modernity is per-
haps the fuzziest definition for it and recapitulates the problem of how to
define a period which is both not modern and not quite not-modern. We
cannot easily dismiss Randolph Starn’s assessment that the term “early
modern” is obfuscatory because it “seems to diminish the liabilities of
periodization while maximizing the benefits.”¹¹ Depending on how we

⁹ Eisenstadt and Schluchter 1998; cf. Eisenstadt 2000.

¹⁰ Richards (1997) uses the term neutrally but his account is all technology, economy, and
large states. Compare Subrahmanyam (2001), for whom “early modern” is perfectly reasonably
applied to India with minor caveats. Jack Goldstone dislikes the term early modern (whether for
Europe or anywhere) and uses a paradigm of “advanced organic societies” instead, but he does
not consider the cultural life of such societies (Goldstone 1998). He makes the striking argument
that nineteenth-century industrialization in Western Europe was made possible by the dumb
luck of easily accessible coal in the right places. For Frederic Jameson, the history of modernity
is the history of capitalism (Jameson 2002), but if we accept that formulation then most of the
questions posed here would be irrelevant.
¹¹ Starn 2002, 302.
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

240       

define modernity, we can find it everywhere in history—or nowhere.¹²

Moreover, because our template is inevitably Europe we should be wary
of seeking particular analogues to European early modernity in South
Asia. For example, Ārzū might have some claim to be an Indo-Persian
equivalent of the Italian humanist Giambattista Vico (1668–1744), but
that is certainly not an argument I would want to make. A more sensible
analysis would track large patterns and try to avoid teleological thinking
rather than encouraging it. The sociologist Jack Goody takes such a
pattern-based approach in arguing that the European Renaissance was
not the key to modernity or capitalism but rather represents a particular
instance of a stage in a cultural cycle (a “renascence” as he terms it) that
is practically universal in literate societies.¹³ The things that constitute
European early modernity must be generalized so that they make sense
outside of Europe, if they were indeed present. We should not, for example,
be looking for a seventeenth- or eighteenth-century Enlightenment in
India because we will not find it.¹⁴ We can, however, consider if some of
the individual features thought to be characteristic of early modernity are
present in India and whether they are new in the Indian context. If they
are indeed new, then we can speculate as to why they appeared when
they did.
To take an obvious example, how does the idea of a public sphere
apply in the South Asian context? It has expanded beyond Jürgen
Habermas’s original, rather specific conception into virtually any sort
of European public activity.¹⁵ Critics have rightly pointed out that
despite Habermas’s argument that the public sphere depended on uni-
versal access, it actually excluded vast numbers of people (women, ethnic

¹² Indeed, the word “modern” itself, with its overtones of a break with the past, was used in
that sense by Cassiodorus in the fifth century  to mark the difference between the Pagan past
and the Christian present (Habermas 1997).
¹³ Goody 2010, 7ff., 241ff. For Goody, the Renaissance has three main characteristics: revival
of classics, secularization, and economic change. But he is sufficiently vague, especially when
talking about the non-West, that it becomes a cautionary tale about generalizations. For
example, he argues that the European Renaissance was unique in that it drew on a completely
different tradition from the current one, i.e., that it brought Pagan works into a Christian
domain, but then draws no conclusions from this insight (2010, 255–7).
¹⁴ Pollock 2004, 79; cf. 2007.
¹⁵ And this despite Habermas’s clear warning that the concept of a bourgeois public sphere is
tied to its specific time and place (Habermas 1991, vii).
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

    241

and religious minorities, and so on) in any given place.¹⁶ Under a broad
definition of public sphere, the circulation of Indo-Persian intellectuals
must count—the public sphere is the literary language itself and the
community of language users.¹⁷ An indigenous term for this intellectual
:
common is suhbat, which can narrowly mean conversation but also
applies to a range of public interactions.¹⁸ However, even recognizing
that venue and medium of a public sphere cannot help but be different in
:
different societies, suhbat seems to move so far from Habermas’s con-
ception that perhaps we are better off creating a localized genealogy of
the “common” rather than trying to generalize the term “public
sphere.”¹⁹ There happen to have been coffeeshops and newsletters in
eighteenth-century Delhi as there were in Europe, but that is not the
point.²⁰ If we posit that a feature of early modernity is that high culture
becomes more diffused during that period,²¹ then we can see the increas-
ing availability of Persian education in South Asia as a marker of this
trend. Mus̄mir itself deals with the question of where the controlling
authority for Persian might come from given that the language is used
across such a vast and varied terrain. It is also important to consider
whether the medium of a public sphere needs to be vernacular language.
If not, as the Indo-Persian case suggests, then this throws up a major
conceptual problem for Europeanists: Europe has mythologized its own

¹⁶ Or, more precisely, as Sudipta Kaviraj has argued, the public sphere “sloped” against
certain categories of people, who were not expressly forbidden from participating but faced
obstacles to access that effectively made their participation impossible. Thus, while women were
not overtly forbidden from participating in Indo-Persian letters, very few did and these were
extremely high-born women such as Gulbadan Begum, the sister of Humāyūn and author of the
Humāyūnnāmah.
¹⁷ Sanjay Subrahmanyam has preliminarily sketched the mindset of such early modern South
Asian intellectuals (Subrahmanyam 1998, 93–6). In the European Renaissance context, Grafton
uses the philologist Giovanni Pico della Mirandola (1463–94) as a model for how reading
worked as a community-builder: “Confronting ancient authorities was for him an intense and
complex act, at once individual and collaborative, private and public” (1997, 132; cf. Matheson
1998, 28ff. on the sixteenth century).
¹⁸ As it also does in the Ottoman context (Andrews and Kalpaklı 2004, 106ff.).
¹⁹ The printing press, which figures so prominently in the historiography of Europe, is an
irrelevance in South Asia because of its late adoption there. There is clear evidence that even in
the nineteenth century previously printed texts were copied by hand when lithographic plates
wore down (Baevskii 2007, 176).
²⁰ For Habermas, the coffeeshop and the newsletter arise together in a kind of symbiotic
relationship (Habermas 1991, 42, 59), but we do not have enough information about the Indian
case to know whether that principle applies.
²¹ Higman 1997.
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

242       

march towards the vernacular. The transnational public sphere in both

Europe (where its written medium was often still Latin) and the Persian
Cosmopolis were both fragmenting by the end of the eighteenth century.
The classical language—seen in the context of European modernization
theory as retrograde—was universal enough to serve as the medium for a
social common and arguably a public sphere. If we grant in the South
Asian case (as we do in the European) that the public sphere took
centuries to develop then what meaningful transitions do we see? We
observe the domain of Persian expanding through education rather than
contracting at the same time as vernacular literary production was
becoming systematized.
The concept of humanism—the rough equivalent in Persian being
adab—similarly needs to be dealt with comparatively.²² Mus̄mir is a
work that sets the stage for a kind of humanism because it establishes a
cultural baseline for Indo-Persian in that it fixes the relationship of the
three relevant literary cultures: Arabic, Persian, and Indic. Humanism,
argues Stephen Greenblatt, has an “enzymatic function” since it absorbs
culture and integrates it into a coherent discourse.²³ Rhetoric was crucial
because it was the structure of that coherence. History and other human-
istic disciplines were generally thought of as sub-disciplines within
rhetoric. At the same time, the canon of classical texts was proof that
knowledge and eloquence were inexorably linked.²⁴ Indeed, early mod-
ern works often strike modern readers as having been more concerned
with style rather than content, that is, with the literariness of describing a
given historical event rather than conveying to the reader what “really
happened” in a positivist sense.²⁵ This was, of course, a complaint leveled
by British colonial scholars against Persian historical sources because—
in the familiar pattern—they held such sources up to a contemporary

²² It should be noted that although “humanities” and “humanist” have a long history, the
term “humanism” itself did not emerge until the nineteenth century (Burke 1997, 12). On adab
as Persian literary humanism, see Kia 2014a; Dabashi 2012. In my reading (and according to
Sharma 2013), Dabashi’s sweeping work unfortunately recapitulates some of the methodo-
logical problems it argues against.
²³ Greenblatt 1980, 230. ²⁴ Gray 1963, 502.
²⁵ A fascinating example is the painting Flight into Egypt by Nicolas Poussin (1594–1665).
Although the story in the Gospel of Matthew specifically mentions camels as having participated
in the event, Poussin omits them in the painting as they were beneath the dignity of the scene
(Burke 1970, 105–6). On rhetoric and the study of history, see LaCapra 1985, 36ff.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

    243

European standard, forgetting that European scholarship of a century or

two before would have also failed the same test. Attempting to extract the
content while ignoring the subtleties of the form is a well-attested, if
obviously problematic, historical method. Arguably the loss of formal
rhetoric is the greatest intellectual barrier between modern readers and
those of several centuries ago.²⁶ This holds for South Asia just as it does
for the West. Even our sense of poetry and its social function is the result
of the early nineteenth-century Romantic determination that poetry
should be excluded from traditional rhetoric.²⁷ Because the objective of
lexicography in the Persianate tradition was with few exceptions literary
interpretation, Persian humanism at the analytical level is incomprehen-
sible if rhetoric and poetry are separated. The systematization of know-
ledge in literary form was also an important idea in the Arabic tradition.
For example, a ninth-century Baghdadi writer argues that “poetry is the
mine of knowledge of the Arabs and the book of their wisdom, the
archive of their histories.”²⁸ The Renaissance polymath Giovanni Pico
della Mirandola’s (1463–94) argument that philosophers need not be
eloquent would have been virtually unthinkable in the Persianate
world.²⁹
The foundation of pre-modern humanism is a set of texts whose
assumptions, specifically their categories, are taken as given.³⁰ Modernity
represents an attempt to dismantle that foundation while largely keeping
whatever had been built on top of it.³¹ Early modernity can therefore be
seen as a transition in the construction of pre-modern knowledge sys-
tems, while only in rare cases seeking to abandon their structure. We can
say that it is distinguished from previous conflicts of Antiqui and
Moderni (that is, people defending old and relatively newer forms of
knowledge, respectively) by degree since the amount of new knowledge
generated and its relative centrality were much greater. Sheldon Pollock

²⁶ Gray 1963, 497, 514. ²⁷ Ong 1971, 6. ²⁸ Quoted in Cantarino 1975, 24.
²⁹ Gray 1963, 508ff.
³⁰ As Grafton argues, “Renaissance intellectuals shared a commitment to continuous, inten-
sive conversation with ancient texts. This provided the foundation of their efforts to understand
other cultures and religions, devise natural and political philosophies, create a personal code of
conduct, and cultivate a literary style” (Grafton 1997, 6).
³¹ Which in some cases led to an awkward admixture such as the emphasis on memorization
in modern Indian education (Kaviraj 2005, 518).
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

244       

has observed this tension in the context of Sanskrit new intellectuals in

the seventeenth century, namely that there were “remarkable new subtle-
ties of argument and exposition but directed toward the analysis of
ancient categories and the establishment of archaic principles.”³² The
old philosophical categories were no longer entirely fit for new purposes,
but by the same token no one could mount an effective critique that
would dismantle them.³³ Mus̄mir operates within similar constraints,
since it radically expands the scope of its inquiry into language beyond
what the tradition had previously countenanced (integrating, notably,
Indic language), but never grapples with any question of the foundation
of its categories. The transition described by Foucault in The Order of
Things, in which the split of signifier and signified led to new ways of
organizing knowledge and aesthetics in Europe, only appears to have
taken place in India after the advent of colonialism. It is probably useless
to speculate as to whether such a development would have taken place
had Western knowledge practices not supplanted indigenous ones, but
the parallel crisis in traditional categories in South Asia and the West
suggests that it was at least a possibility. Whatever the case, now more
than ever there is a need to “provincialize Europe,” which means that we
should be wary of claims that the European experience was universal.
There must be a double operation in which we also are careful to define
European modernity against European pre-modernity. Marx memorably
wrote that “the tradition of all the dead generations weighs like a
nightmare on the brain of the living” but this cannot be taken as
axiomatic for all time.³⁴ The wonderful imagery short-circuits the neces-
sary historical reflection: Before modernity, the “dead generations” were
practically alive in the texts they left for posterity, and far from being the
stuff of nightmare, their guidance was actively sought.

³² Pollock 2001, 19. Jonardon Ganeri disagrees with Pollock’s assessment (arguing that even
what appears to be respect for tradition in the new intellectuals’ work is actually itself radically
new), but I am unqualified to adjudicate this claim (Ganeri 2011, 100).
³³ Pollock 2004, 32.
³⁴ From The Eighteenth Brumaire of Napoleon Bonaparte (quoted in Grafton 1992, 253).
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

    245

Ārzū’s Philology and Its Possible (But Unlikely)

Influence on European Philology

:
Muhammad Husain

Āzād, previously mentioned in this study as the
author of the Urdu literary history Āb-i Hayāt,

wrote an account of
:
Persian philology and literary history called Sukhandān-i Fārs [The
Persian Poet, 1872].³⁵ It is unquestionably a nationalist work—many
sections begin with the invocation “My Dear Compatriots!” [ʿazīzān-i
wat̤ an!]—and seeks to prove that Europeans do not have a monopoly on
the scientific study of language. He calls philology “falsafah-yi lisān” [lit.
philosophy of language] and by the English word transliterated as
“filālājī.” The centerpiece of the argument is a comparison of Sanskrit
and Persian words to show the historical tie between the two languages,
and his reasoning appears to be an extension of Ārzū’s method.
However, Āzād does not cite his sources and we are not in a position
to say how much of his data was derived from the indigenous philo-
logical tradition and how much from European historical linguistics,
which had matured by this time and would have been known to the
Europeans with whom Āzād interacted. Āzād’s study is framed by a
radical assertion: The Anglo-Welsh polymath and colonial judge
Sir William Jones (1746–94), regarded as the founding genius of
European philology, probably did not come up with his famous obser-
vation on the historical ties between languages on his own but rather
derived it from the works of Ārzū and Ārzū’s friend Tek

Chand Bahār.³⁶
Evocative as the idea was in Āzād’s time—and for that matter still is in
ours—the evidence suggests no particular connection between Ārzū’s
theory and Jones’s except that Jones was a Persian scholar. Indeed, given

³⁵ On this work see Sharma 2012b, 55–6.

³⁶ Note that the exact statement is tentative in its wording, employing a formula that Āzād
often employs when he has no solid evidence to support his contention: “God knows whether
[Jones] came to this conclusion through his own efforts or with the help of the works of these
: :
two [i.e., Ārzū and Bahār]” [k: hudā jāne s: āhib ne apnī t̤ abīʿat ke lagāʾom se yā un donom kī
tas: nīfāt se yih nuktah pāyā] (Suk: handān-i Fārs 2005, 10). Āzād mistakenly writes that Jones
went back to Europe and spread his theory, but of course he died in India, and his Indian
scholarship became popular in Europe once Asiatick Researches, the proceedings of the Asiatic
Society of Bengal, began to be published in Britain in the 1790s.
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

246       

the early modern intellectual climate in both India and Europe, there is
an air of inevitability that a similar discovery would be made in both
places at roughly the same time.
Although Jones certainly knew of Ārzū through at least one of his
works and undoubtedly also from Indian informants, there is no evi-
dence for the proposition that he could not have come to his conclusions
if he had never encountered Ārzū’s work: Not only were there European
antecedents for Jones’s theory but there is also a potential problem of
chronology. The only indication we have of Jones’s encounter with
Ārzū’s work comes from an inscription dated two years after Jones’s
famous proclamation of the historical links between Sanskrit, Latin, and
Greek, as well as with “the Gothic and the Celtic . . . and the old Persian,”
in the Third Anniversary Discourse to the Asiatic Society in Calcutta, on
2 February 1786.³⁷ Jones’s copy of the dictionary Farhang-i Jahāngīrī is
inscribed “The gift of Charles Roddam Esq 17 Febr. 1788 to W. Jones.”³⁸
It continues in a different hand—presumably Jones’s own—noting that
“many corrections of this valuable work & many additions to it, may be
found in the Sirāju’lloghah [i.e., Sirāj al-Luġhat] by Sirājud’dīn Ārzū, and
in the Majmaū’llogah.” Although there is no corroboration that any
eighteenth-century European read or even knew of Ārzū’s Mus̄mir,
Sirāj al-Luġhat itself contains references to tawāfuq al-lisānain [corres-
pondence of two languages]. We can assume, for the sake of argument,
that reading Sirāj al-Luġhat allowed Jones to appreciate the implications
of tawāfuq. But there is no indication that Jones took his argument from
Ārzū. In fact, what little by way of evidence we do have perhaps implies
that he encountered the Sirāj al-Luġhat after his 1786 lecture in Calcutta.
Thus, we have no proof that Jones knew of Ārzū before he made his
famous claim, and in all of Jones’s works, there appears to be no
reference to Ārzū besides that single marginal note. Nor do we have
any copy of Sirāj al-Luġhat that Jones may have owned (his Farhang-i
Jahāngīrī is heavily marked up in the margins so for us not to have his

³⁷ “The Third Anniversary Discourse” 1995.

³⁸ The manuscript is British Library ms RSPA 20. It was donated to the Royal Society by
Jones in 1792 and transferred to the India Office Library in 1876, and then to the British Library
when the India Office Library merged into the British Library. See Ross and Browne 1902,
18–19.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

    247

Sirāj al-Luġhat potentially represents a significant loss for intellectual

history). There are, however, two extant manuscripts to which he might
have had access: The National Archives of India hold a damaged but
readable copy.³⁹ A stamp connects it to Fort William College and
therefore to early nineteenth-century Calcutta, if not exactly to Jones’s
time there in the late eighteenth century. It is not dated and has no
interesting marginalia. Similarly the British Library’s copy has no mean-
ingful marginalia but is inscribed “Mr. Richard Johnson” (1753–1807),
an East India Company official and manuscript collector.⁴⁰ He was a
friend both of Jones and the Governor-General Warren Hastings, and it
is possible that Jones borrowed his copy of Sirāj al-Luġhat, which the
colophon states was completed in 1160/1747.
If we assume, again for the sake of argument, that Jones had in fact
been familiar with Ārzū’s work before early 1786 (perhaps through the
mediation of one of the native scholars of Persian with whom he
worked), then it is still unlikely that he took his ideas from Ārzū’s
thought directly because similar ideas were available elsewhere. Jones
was by no means the first European to consider etymologies as evidence
of possible ties between languages. In his lecture, he mentions several
phonetically obscure etymologies (for example that French “jour” comes
from Latin “dies”) as obvious fact.⁴¹ The Renaissance origin of this sort of
thinking is what strikes us today as the most credulous sort of pseudo-
science, but even wild speculation about the development of languages
contained a kernel of the late eighteenth-century insights that would
eventually become historical linguistics. Take, for example, the learned
discussion by Jan van Gorp (1518–72, also called Johannes Goropius
Becanus) of how “Cimbrian,” the supposed ancestor of Dutch, was also
the ancestor of Hebrew.⁴² While van Gorp was clearly trying above all to
irritate French and Spanish speakers in his adopted city of Antwerp (who
thought their languages were superior to Dutch/Flemish), this kind of
magical thinking about the evolution of language did in fact lead to

³⁹ Fort William College collection ms 109.

⁴⁰ British Library ms IO Islamic 178; Ethé 1903, 1352–3. On Johnson, see Marshall 2004.
Jones’s letters to Johnson are cited in Cannon 1990, 366; Cannon 1970, letters 373, 509 (in
which we learn that Jones borrowed Johnson’s manuscript of the Shāhnāmah).
⁴¹ Jones 1995. ⁴² Olender 1994, 13.
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

248       

sophisticated reflection on how it is possible for different languages to

contain words that are the same or nearly the same. Āzād would have
been unaware of this particular European tradition—which was by his
time an embarrassment to serious historical linguists—when he wrote
:
Sukhandān-i Fārs, so from his perspective Jones must have either
invented or plagiarized the foundation for historical linguistics ex novo.
In fact, at least a half dozen people had written works in the seventeenth
century alone opining that various European languages formed some
kind of a common linguistic area.⁴³ In some cases, this reasoning even
extended to India and Persia. The English Jesuit Thomas Stephens,
writing from Goa in 1583, probably noticed the connection between
the Indo-Aryan languages and Latin and Greek two centuries before
Jones, and the French Jesuit Gaston Cœurdoux certainly did in 1767.⁴⁴
George Hadley’s Introductory Grammatical Remarks on the Persian
Language (1776) concludes with “a small list of words shewing the
analogy between the Persian and European Languages.”⁴⁵ Such ideas
were available in Europe, and nothing in Jones’s construction of the
theory suggests any particular debt to Ārzū rather than to these
European precursors.⁴⁶ Ārzū, as discussed earlier in this study in the

⁴³ Olender 1994, 17.

⁴⁴ Fr. Stephens (d. 1619) wrote to his brother on 24 October 1583: “The languages of these
regions are very many. They have a not unpleasant pronunciation and a construction similar to
Latin and Greek” [Linguae harum regionum sunt permultae. Pronuntiationem habent non
invenustam, et compositionem latinae graecaeque simile] (quoted in Olender 1994, 20). Today
he is remembered for compiling the first printed Konkani grammar (as well as writing Konkani
in Roman letters) and for his literary experiments in Marathi, including the Kristapurāna, : an
epic poem on the life of Christ. Fr. Cœurdoux was more explicit about the connection between
European languages and Sanskrit in his 1767 memoir that was known in French intellectual
circles but not actually published until 1808 (Olender 1994, 20; cf. Arlotto 1969). Another
possible European precursor was a word list compiled by the Leiden classicist Justus Lipsius
(1547–1606), in which the author observes that “they [i.e., Persians] have many of our words
and scarcely different inflexions” [et uoces plures nostras habent et flexus coniugationum haud
nimis diuersos] (Emmerick 1974). See also Tavoni 1994, esp. 45, 54–6.
⁴⁵ Hadley 1972, 215–16.
⁴⁶ Abdul Azim also concludes that Jones was most likely not influenced by Ārzū (Azim 1970,
267–9). It is important to remember that Jones did not sketch a linguistic family tree or explicate
an evolutionary framework for languages. (His statement that the languages he mentions had a
common origin does not imply the multiple evolutionary steps as we know them, e.g., that
Proto-Indo-European became Proto-Indo-Iranian which split into the Indo-Aryan and Iranian
languages which sub-divided further.) Although the first linguistic tree was drawn in the early
nineteenth century, it was not a well-known image before the work of August Schleicher
(1828–68), whose understanding of language was explicitly Darwinian (Auroux 1990, 228;
Timpanaro 2005, 121n4). One of Schleicher’s books is in fact called Die Darwinische Theorie
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

    249

context of tawāfuq, does not explicitly posit a common origin for

Sanskrit and Persian but rather leaves it to the reader to infer the cause
of the linguistic similarity. Jones on the other hand was unequivocal
about there being a “family” of languages with a “common source which
perhaps no longer exists.”⁴⁷ While Jones may have found independent
confirmation of his own thinking in Ārzū’s work, having encountered
Indo-Persian was not a necessary precondition for the development of
the theory, which after all was a more nuanced version of arguments that
Europeans had long made. The Third Anniversary discourse itself opens
by invoking the English scholar Jacob Bryant (1715–1804) “with rever-
ence and affection.” In fact, Bryant’s A New System, or, An Analysis of
Ancient Mythology (published in 1774 and subsequently revised) had
recently been the subject of a dispute with Bryant on one side and Jones
and Richard Johnson on the other. Jones notes in the discourse that “the
least satisfactory part of [Bryant’s book] seems to be that, which relates to
the derivation of words from Asiatic languages.” He locates his own
thinking about language as an extension to and correction of Bryant’s
work. Bryant’s project was demonstrating the development of various
cultures (including that of the “Indii” and “Indo-Scythae”) from an
antecedent, and it is therefore the equivalent in cultural studies of what
Jones proposes regarding languages.
The project of theorizing and ameliorating Orientalism’s “genesis
amnesia”—as Mohamad Tavakoli-Targhi has provocatively framed the
lack of recognition for non-European scholars’ contributions to the
development of Western knowledge about the non-West—is important
and much work remains to be done along those lines.⁴⁸ Nonetheless,
however much sympathy we might have with the goal of highlighting the

und die Sprachwissenschaft [Darwin’s Theory and Linguistics, 1863]. The limitations of Jones’s
project are contextualized in Robins 1996. A readable account of European philology to 1600 is
Law 2003.
⁴⁷ “The Third Anniversary Discourse” 1995.
⁴⁸ Tavakoli-Targhi argues that “the breakthroughs in comparative religion and linguistics,
which were the high marks of ‘the Oriental Renaissance’ in Europe, were in reality built upon
the intellectual achievements of Mughal India” (2001, 21). He cites tawāfuq as an example
(Tavakoli-Targhi 2001, 26; cf. Tavakoli-Targhi 2011, 270–1). Jones, it should be said, seems to
have credited his Sanskrit teachers with introducing him to particular texts and ideas more
frequently than other Orientalists created their informants, but was not as charitable when it
came to his Arabic or Persian studies.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

250       

contributions of Indians to the creation of colonial knowledge and the

modern world itself, the marquee claim that Ārzū through Jones sparked
the development of Western historical linguistics is unsupported by any
particular evidence, although the available evidence does not preclude
the possibility of influence either.⁴⁹ I can only propose that we tread
carefully so that we do not overcompensate for past injustices by redu-
cing the complexity of colonial knowledge production (and the intellec-
tual background of the people involved) to a stereotyped extraction that
renders the nuances of the European half of the equation irrelevant.⁵⁰
Ārzū’s historical importance is undiminished if it turns out that no path
connects his insights to Western linguistics; his work would nevertheless
remain as a road that could not be taken because of the colonial
encounter.

Ancients and Moderns in India and Europe

The distinction between Ancients and Moderns, a workable translation

of mutaqaddimīn and mutaʾak: hkhirīn,
: is one that seems to hold across
many traditions, and can also be thought of as classical writers versus
recent and contemporary ones. It is crucial for the question of poetic
authority because recent and contemporary writers are inevitably held to
different standards than centuries-old ones, a distinction which informs

⁴⁹ Kinra refers to the “strong circumstantial case . . . that some form of Jones’s revolutionary
thesis has its provenance with the likes of Ārzū” (Kinra 2011, 360). However, this ignores the
stronger circumstantial case that Jones, as a trained classicist, was aware of European etymo-
logical research and applied it to the impressive set of languages of which he had some
knowledge. None of the three scholars whom Kinra cites (Muzaffar Alam, Mohamad
Tavakoli-Targhi, and Kapil Raj) as proponents of the influence of Ārzū on Jones appear to
have engaged with possible European genealogies for the theory.
⁵⁰ See, for example, Marchand 2010 for a similar project. There is an alarming tendency, for
which Said is largely, if accidentally, responsible, for scholars to simply ignore the larger context
of Western scholarship when drawing inferences about Orientalist prejudice. An irony in the
goal of Said’s reclamation of the non-West as a culturally dynamic space is that his study is
premised on the West’s being an essentially static engine of domination from Aeschylus’s
portrayal of the Other in The Persians (472 ) to the present. Much of what Said concludes
in Orientalism about Western scholarly attitudes over a vast sweep of history is an extrapolation
from the nineteenth century, as I have argued at length elsewhere (Dudney 2008). For example,
when Said invokes the idea that Westerners thought of the East as unchanging because they
described it in terms of classical antiquity, he fails to allow for the fact that until the nineteenth
century the Classics were still frequently invoked as a template to describe society in the West.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

    251

both theoretical and practical concerns in literature. The works of the

Ancients have survived, and by surviving the ravages of time and neglect
have become classics that can provide guidance to contemporary poets.⁵¹
A discussion of the relationship between Ancients and Moderns in
seventeenth- and eighteenth-century Indo-Persian, as in Chapter 3 of
this study, lends itself to a comparison with the roughly contemporary
crisis of authority in Europe. The most celebrated account of this situ-
ation in English is Jonathan Swift’s satire “The Battle of the Books,”
published in 1704. It imagines the library of St. James’s Palace in London
as a battlefield where books by old authors face off against books by
recent authors. (A spider and a bee also enter the fray.) It is a witty
allegory of the contemporary debates over authority in Western Europe.
A simplified account of the intellectual currents of that period would go
like this: People began to question the worth of classical texts, which no
longer seemed timeless, because, for example, if Aristotle has been
proven wrong through experimental science then why should Homer
be held in higher regard than Modern poets?⁵² After Edward Said, it is
impossible not to notice the tendency to offer broad socio-political
explanations (namely “decline”) for early modern literary trends in the
non-West while European literature supposedly comes into being
because of deliberately taken, positive steps towards modernity. Each

⁵¹ See Lianeri and Zajko 2008. They observe that “the idea of the classic is invested in a
particular model of history, one which allows for a perpetual tension between the enduring and
the transient and for the survival of the past in ways that are comprehensible even to a radically
different present” (Lianeri and Zajko 2008, 4). This is undoubtedly true, but it is worth pointing
out that most societies have seen history in this way, and it is only with modernity that the
exceptionalism of the present (as in Tocqueville’s image that the lamp of history no longer
illuminates the future) becomes a common way of thinking. An early modern and modern use
of the word “classic,” which can be more precisely called “relative classic,” is a work that attains
perfection by the standards of its age but is not venerable enough in age to be on par with the
classics of Antiquity (Lianeri and Zajko 2008, 11). In the Persianate context, discussions of
tāzah-goʾī can be read as hinging on a similar tension between the absolute and the relative
classic.
⁵² For example, Sir Richard Blackmore in 1716 wrote in his Essay upon Epick Poetry (along
similar lines to what Abbé Pons had written two years before) that “Unless the Admirers of
Homer will assert and prove their Infallibility why may they not be deceiv’d as well as the
Disciples and Adorers of Aristotle?” (quoted in Aldridge 1973, 76–7). See also DeJean 1997.
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

252       

case is overdetermined and can shed light on the other.⁵³ When I use the
label “Modern” in these pages to describe an intellectual faction in the
seventeenth and eighteenth centuries, this refers to people whose orien-
tation was towards new configurations of knowledge and a move away
from the Ancients. We should resist the temptation to think of them as
“the makers of the modern world” because then we are locked into a
teleological and Eurocentric reasoning that undermines the effort of
comparison. The turn of the contemporary against the established is
an intellectual attitude that reappears from the coining of the term
“modernus” in Latin in the fifth century through the present-day usage
to refer to the particularities of our own time.⁵⁴ Of course, the constitu-
ents of each category change in every instance, and it is in studying the
transformation of the categories that we find instructive discontinuities.
A brief account of the contours of the Ancients and Moderns debate in
early modern Europe is in order. It finds a place here not in order to
facilitate a rigorous comparison with the situation in the Persianate
world in the same period, but rather because it has helped me to
understand my own historical preconceptions. Additionally, it provides
a sketch of a pervasive historiographical problem, namely that Europe’s
march towards intellectual modernity (however exactly that is to be
defined) is seen as a series of deliberately taken steps while roughly
similar transformations in the non-West are often assumed to be hap-
hazard and accidental. Of course, this is the result of the West’s being the
yardstick of modernity—introducing a historical tautology since the
instrument of measurement is the same as the object being measured—
rather than stemming from an identifiable, operative difference between

⁵³ We can of course find evidence of so-called cultural decline if we just root around in the
European tradition. For example, in 1771 Sir William Jones writes (in Latin) to his friend to
dissuade him from publishing a book of Latin translations but rather to publish in French: “One
can hardly believe how few worthy men there are in England who know Latin” he declares [nam
credible vix est quam pauci sint in Anglia viri nobiles qui Latine sciant] (Cannon 1970, 1:86). My
own attempt to revisit questions of cultural decline or “decadence” in eighteenth-century Urdu
literary culture is Dudney 2018b.
⁵⁴ “Modernus,” according to the Oxford English Dictionary, is derived from “modo” [just
now] + -iernus [Classical Latin hodiernus, the adjective derived from hodie “today”]. In
European intellectual history, the moderni/antiqui distinction is important in the twelfth
century, again in the fourteenth and fifteenth centuries, and again in the period under discus-
sion here (see Coleman 1992, 293, 541ff.).
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

    253

intellectual conditions in Europe and, say, the Persianate world.⁵⁵ We

need to take the (inevitably Western) “makers of the modern world” off
their visionary pedestal and find other heroes.
The idea that the conditions of the present might be better than those
of an Ancient Golden Age was apparently first explicitly formulated in
early modern Europe by Alessandro Tassoni in his Dieci Libri di Pensieri
Diversi [Ten Books of Diverse Thoughts, 1620].⁵⁶ Some years later, after
Tassoni’s book had been translated into French, an attack on the
Ancients was delivered at the recently founded Académie Française.
Charles Perrault (1628–1703), the right-hand man of the finance minis-
ter Jean-Baptiste Colbert, argued in his 523-line poem “La Siècle de Louis
le Grand” [The Age of Louis XIV, 1687] that literature and indeed life in
general were better now than they had been in Antiquity.⁵⁷ In particular,
he put Plato in his place, declaring that “Plato, who was divine in the days
of our ancestors / Is beginning to sound sometimes boring.”⁵⁸ This text
and his subsequent Parallèle des Anciens et des Modernes [Analogy of the
Ancients and the Moderns], published in four volumes in his later years,
were arguably a struggle for personal legitimacy (namely getting out
from under Colbert’s shadow since Colbert had installed him in the
Académie in the first place) even though it had the effect of laying the
groundwork for a major shift in the intellectual life of Europe.⁵⁹ Perrault
was not at the head of a movement but rather a man fighting his own
battles with the help of friends against particular enemies. He tangled in
particular with the classically inclined poet Nicolas Boileau-Despréaux,
who replied to volume 2 of Parallèle with an epigram in 1692:

How is it that Cicero, Plato, Virgil, Homer, and all these great authors
whom the university reveres, when translated in your writings appear
to us to be so stupid? P . . .⁶⁰ it is because in lending to these sublime

⁵⁵ Marshall Hodgson has argued that the key to Western misunderstandings of the
Islamicate world is not the gap between East and West but rather the yawning gulf between
modern and pre-modern intellectual practices that is often ignored (Hodgson 1968, 54; pace
Said 1979).
⁵⁶ Bury 1920, 80. ⁵⁷ Barchilon and Flinders 1981, 43ff.
⁵⁸ Translated in Barchilon and Flinders 1981, 43ff. ⁵⁹ Zarucchi 1989, 13–14.
⁶⁰ The rhyme makes it clear that “P . . .” is meant to be read as “Perrault.”
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

254       

minds your manner of speaking, your crudenesses, your rhymes, you

make them all into P . . . s.⁶¹

Perrault replied in kind by jabbing at Boileau’s well-known misogyny in

another poem.⁶² Boileau perhaps had the last laugh after Perrault’s death:
He said that Perrault’s seat at the Académie would remain vacant
because its next occupant would be obliged to give a eulogy on
Perrault, and who would volunteer to “praise the enemy of Cicero and
Virgil”?⁶³ These ad hominem attacks and the fact that Perrault appar-
ently had no programmatic vision are important to note here in order to
contrast the historiography of the West and the non-West: Literary
debates in Indo-Persian which involve personal rivalries have been
dismissed as having no consequence to a larger historical narrative
(except, of course, as manifestations of the omnipresent specter of
Mughal-Safavid decline), but the evidence is clear that some important
Europeans credited with bringing forth Modernity were throwing liter-
ary brickbats rather than sketching detailed blueprints for an enlightened
future society.
Rather than facing the situation in terms of apparently modern atti-
tudes and practices winning out over pre-modern ones, we should pick
up the common thread of the responses to perceived newness.⁶⁴ In
Europe, one key technique was to bracket off the newly reinvigorated
experimental sciences from the humanities.⁶⁵ Francis Bacon (1561–1626)
argued for a remarkable inversion of history: If the Ancients are charac-
terized by having knowledge which was then lost over the ages, then
clearly the term “Antiquity” is misplaced because the people of Bacon’s
own time had their own sort of knowledge (which is to say, scientific

⁶¹ Translated (as prose) in Zarucchi 1989, 15. ⁶² Barchilon and Flinders 1981, 54.
⁶³ Translated in Zarucchi 1989, 21.
⁶⁴ A major issue in the scholarly debate over the appropriateness of the term “early mod-
ernity” to describe a global historical epoch from roughly 1500 to 1800 is the question of implied
teleology: The term suggests that in the period in question people across the world were
thinking “how can we become modern?” (which is, of course, absurdly anachronistic). In this
debate I tend to side with Richards 1997 rather than with Goldstone 1998.
⁶⁵ In the Persianate world, the long tradition of experimental science was not apparently
marked as new during this period as it was in the West. It has been argued that the Scientific
Revolution could not have taken place in the West without building upon Islamic science (e.g.,
Saliba 2007). Could this infusion of outside ideas into Europe itself have contributed to the
perception of experimentation as a new endeavor?
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

    255

knowledge derived from experimentation) and it vastly exceeded what

people in Antiquity could have known.⁶⁶ In other words, the Moderns
are the true Ancients. Likewise, John Dryden (1631–1700), best known
for his extraordinarily eloquent English translation of Virgil’s Aeneid,
implies that the Ancients studied poetry and the Moderns study science,
noting that all Modern poetry is inferior to Ancient poetry.⁶⁷
Furthermore, Antiquity’s own concerns over hegemonic knowledge
derived from literary works, as expressed in Cicero’s well-known declar-
ation that “the inconsistencies of Plato are a long story” [iam de Platonis
inconstantia longum est dicere], could be deployed against tradition.⁶⁸
Despite its different political formations, the Persianate early modern
world addressed similar aesthetic and intellectual issues but often in a
different idiom. In architecture and art history, both outside the scope of
the present discussion, we can see other ways in which people grappled
with newness without reference to science.⁶⁹ It would, however, be a
stretch to propose exact analogues between Persian and the West, as for
example in calling for the recognition of t̤ arz-i tāzah as the “Persian
Baroque.”⁷⁰ The purpose of placing the traditions side by side should be
to undercut the claims to uniqueness that, particularly in the European
context, are the result of not seeking out parallels in other parts of the
world.

⁶⁶ Bury 1920, 54. ⁶⁷ Steele 1990, 229, 232.

⁶⁸ De Natura Deorum, I.30. Early modern scholars’ faith in their knowledge about the ancient
world gathered by reading classical texts was also shaken by the development of new kinds of
archaeology and textual analysis that demonstrated inconsistencies (Bietenholz 1994).
⁶⁹ For example, development in early modern Turkish architecture need not be seen as
synonymous with Westernization but rather within the framework of old and new outlined here
(Hamadeh 2004, 33). In 1691, the French architectural writer Augustin-Charles D’Aviler
defined “capriccio” as a building that did not follow “the customary rules of architecture, that
is of singular and novel taste” (trans. Stern 2003, 82). This catch-all definition is a response to
newness. On the deliberate reconfiguration of classical forms to mark new architectural taste in
the seventeenth century as specifically French, see Ballon 1989. As Ballon argues, the Ancients
could be invoked in specifically nationalist ways. On newness in art in the Ottoman and Safavid
contexts, see Artan 2010 and Farhad 2001, respectively.
⁷⁰ Riccardo Zipoli in a Persian article whose title translates as “Why Is Sabk-i Hindī Called
‘Baroque’ in the Western World?” (Zipoli 1984). A recent book on the Ottomans brings up
“Mannerism” as a possible framework for comparison between early modern Ottoman and
European intellectual history (Andrews and Kalpaklı 2004, 338ff.).
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

256       

Europe and Vernacular Politics: The Vernacular

as Modernity?

The triumph of a national vernacular language over a classical idiom is

widely agreed to be a constituent of modernity.⁷¹ In the West, we
generally speak the same language in daily life that we use (in a higher
register) for administrative and literary writing. South Asia, however, is
deeply problematic if we adopt this paradigm as universal. The interest in
vernacular poetry in eighteenth-century Delhi was not matched by a
prose tradition until considerably later, and only then largely because of
the demands of the colonial state rather than Indians’ preference for the
vernacular over Persian.⁷² In fact, the princely state of Hyderabad, which
was considered pioneering in modern Urdu-medium education in the
early twentieth century, did not replace Persian with Urdu in its admin-
istration until 1884, unlike the territories directly administered by the
British.⁷³ While early modern Europe had Babel as a model for multi-
lingualism, in which the multiplicity of languages was divine punishment
for human sin, South Asia simply took a complex linguistic landscape as
a given. This returns us to the question of whether a singular language is
required for a modern nation-state to come into being.
The path to linguistically constituted nation-states in Europe is less
clear-cut than we have often been led to believe. Languages were in
constant contact as Latin’s influence waned in the seventeenth century,
and various creolizations caution us against exclusively nationalist read-
ings of language in the period.⁷⁴ Indeed, the multiplicity of languages was
not only a reality, but this hybridity was celebrated in pre-modern
European literature.⁷⁵ The process by which a vernacular went from a
private, unofficial kind of language to being the mainstay of a literary

⁷¹ For example, Benedict Anderson contrasts “classical communities linked by sacred lan-
guages” with modern linguistic formations (Anderson 1983, 20). Elsewhere he argues that “the
most important thing about language is its capacity for generating imagined communities,
building in effect particular solidarities” (Anderson 1983, 122).
⁷² Indo-Persian prose-writing continued late into the nineteenth century even as Indo-
Persian poetry waned (Qasmi 2008, 212).
⁷³ Moazzam 2012. ⁷⁴ Burke 2006.
⁷⁵ See Heller-Roazen 2012 and Gaunt 2013a on the medieval celebration of multilingual
textuality; and Gaunt 2013b on why assumptions about national language do not apply to
medieval French. In a specific but telling case, the Sephardic merchants of Livorno in the
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

    257

common was not necessarily connected with the state. Rather, it had
more to do with control over the technology of writing, which at least
from our historical vantage point appears to be crucial.⁷⁶ The acceptance
of the vernacular comes in two phases, the first being the insight that the
language is writeable. According to Armando Petrucci, the second phase

is the conquest by vernacular languages of the right to “canonization in

books,” by which complex texts in vernacular come to be conclusively
and organically written in book form by themselves (that is, without
being mixed or juxtaposed with other texts in “noble” written
languages).⁷⁷

This second phase does not appear to have had much traction in
northern India until the colonial period because Persian remained the
preferred language of technical writing into the nineteenth century. This
tracks with the reluctance of Sanskrit-knowing intellectuals in northern
India to write technical works in languages other than Sanskrit.⁷⁸
Likewise, the “age of translation” that begins in Europe in the seven-
teenth century and allows for more and more classical literature to be
consulted in vernacular translations rather than in Latin or Greek has no
counterpart in pre-colonial northern India.⁷⁹ Until well into the colonial
period, Urdu was not a language of public monuments; apparently
nothing like the debate in Paris over whether the text on the Arc de

seventeenth and eighteenth centuries chose among several different possible languages in which
to write letters based on the addressee’s identity (Trivellato 2009, 179). Naturally, linguistic
hybridity was a typical feature of this polyglot culture.
⁷⁶ This invites the thought experiment of what we might find if we could send a digital voice
recorder into the past. Our pre-twentieth-century archive is textual, so from our perspective a
language only steps onto the stage of history when it is written down. The method of finding odd
spelling and syntax in texts in order to reconstruct speech is hardly an exact science, and even
then gives us only the faintest expression of the life-world of the past.
⁷⁷ Petrucci 1995, 175. Early vernacular texts were written in the guard pages at the front and
back of Latin books, which is true of hindī written in Indo-Persian mss as well.
⁷⁸ Pollock 2011, 24. In both the case of Persian and Sanskrit, “northern India” is defined
against the Deccan (southern India), where there were various high-level offical roles for
vernacular languages in contrast to the situation in the north (Eaton 2014).
⁷⁹ Waquet 2001, 2. The culture of literary/biblical translation in the Renaissance (see
Newman and Tylus 2015) is very different from the Persianate world, but a worthwhile parallel
to draw is that in Europe works were often translated into Latin for wider accessibility and
likewise translations into Persian (from Arabic or Sanskrit) can be seen as making works more
accessible.
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

258       

Triomphe (constructed 1806–36) should be in French or Latin occurred

in northern India.⁸⁰
The concept of a “mother tongue” is seen as universal and exclusively
positive today, as read through the discourse of rights, such as the right to
education in one’s mother tongue (enshrined, for example, in article
350A of the Indian Constitution). At other times in history, however, it
had an emotionally neutral or even pejorative sense.⁸¹ Lisa Mitchell’s
work in the context of South India on language as an “object of emotion”
shows that present-day views on the connection of language and identity
are historically contingent.⁸² A particular language has to be transformed
from a speech pattern associated with a place or with a particular social
function into an identity marker. It is problematic to assume that
vernacular languages were emancipated from the dead hand of a classical
language simply because they later became part of a national identity.⁸³
Indeed, classical languages can also themselves be identity markers, as
Walter Ong argued in describing Latin as a “puberty rite” for young men
(and a few young women) in Renaissance Europe.⁸⁴ Mastery over Persian
functioned in exactly the same way in South Asia as a symbol that
someone was ready to take his or her place in society (although for
women that social role was restricted to extremely elite women wielding
power from within the zenana or women’s quarters).⁸⁵ The fact that

⁸⁰ Fumaroli 1984, 152–3. In the Deccan (in contrast to north India), it appears that there
were sometimes hindī poetic inscriptions on saints’ tombs in the sixteenth and seventeenth
centuries. Walter Hakala is exploring their significance.
⁸¹ As in medieval France, where, since the twelfth century, the “mother tongues” [maternae
linguae] were seen as the “language of children, women, knights, and people in their secular and
non-religious activities” in contrast to the patrius sermo, the Latin of scholarship and religion
(Fumaroli 1984, 139).
⁸² Mitchell 2009, 15, 19ff. On the difference between “mother tongue” and “home language,”
see Das 2014, 45.
⁸³ “By 1750, the European linguistic system was very different from the medieval system,
which had been divided between a living but non-classical Latin and regional dialects which
were spoken rather than written” (Burke 2004, 61). Peter Burke prefers the idea that vernaculars
were “discovered”—which has somewhat Nehruvian overtones (as in the Discovery of India)—
rather than that there was a “crisis” in the classical language because the timescales in question
are centuries (Burke 2004, 16).
⁸⁴ Ong 1959.
⁸⁵ In 1748, the Earl of Chesterfield wrote to his son, “Classical knowledge, that is, Greek and
Latin, is absolutely necessary for everybody. . . . The word illiterate, in its common acceptance,
means a man who is ignorant of these two languages” (Hall 2008, 318). In this formulation, the
ability to read the Classics is equivalent to the act of reading itself as though there were no
worthwhile books in English!
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

    259

Persian and Latin were cosmopolitan, and therefore by definition not

local, did not invalidate this role because these languages were marked as
an integral part of a local cultural formation. Latin is “a language without
a speech community” [eine Sprache ohne Sprachgemeinschaft] and has
therefore been protean in its ability to fill local contexts, even being seen
as advantageous in diplomacy because no nation has a particular claim
upon it.⁸⁶ Persian, obviously, was a language with native speakers, but it
too could function locally in a place where it was not anyone’s mother
tongue.⁸⁷ Surely, one could argue, hindī, a kind of language use whose
very name implies Indianness, is more Indian than Persian. To us it may
seem that way, but the evidence that Indians might have used vernacular
composition to express an Indian identity is lacking for the late pre-
:
colonial period. Of course, centuries earlier Amīr Khusrau had gleefully
declared his ability to speak “hindawī” a part of his Indianness (though
he wrote the poem in Persian, the language of power, rather than in hindī
itself). However, such sentiments appear to have been uncommon in
eighteenth-century rek: htah, for example in Mīr’s oeuvre, where refer-
ences to composing in rek: htah are not explicitly connected with any
identity except that of being a rek: htah poet.⁸⁸ This mode of vernacular
composition was itself cosmopolitan, and poets made reference to their
linguistic medium to showcase their universal competence as poets and
not to highlight their Indianness.
The technical challenges of switching to vernacular literary compos-
ition in part recapitulates the Ancients versus Moderns debate. The

⁸⁶ Burke 2004, 44–6.

⁸⁷ Waris Kirmani overstates the case when he argues that Indians must “submit to the
aesthetic judgment of Iranians” on the basis that “however rich the Persian literature of India
may be as a store-house of thought and learning, Persian nevertheless remained an alien
language” (1972, vii). While this certainly was the prevailing view by the late nineteenth century,
it was open for debate in the eighteenth.
: :
⁸⁸ An exception to this observation might be Mīr’s couplet “kyā jānūm log kahte haim kis ko
: :
surūr-i qalb / āyā nahīm yih lafz̤ to hindī zabām ke bīch” [How should I know what people mean
by “joy of the heart”?/This word has not yet come into the hindī language] (1370, 5). In this
instance the poet’s persona is claiming to not understand a Perso-Arabic phrase that does not
appear in hindī. Shamsur Rahman Faruqi observes that he composed a Persian verse with
exactly the same conceit. In the Persian verse, “the word does not appear in my dictionary” [īn
luġhat jāʾī nah mī-yāband dar farhang-i mā] (www.columbia.edu/itc/mealac/pritchett/00gar
den/13c/1370/1370_05.html).
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

260       

cosmopolitan languages were tied to a long literary past while the

vernacular languages were by comparison upstarts. Just as tatabbuʿ
[establishing a chain of transmission] was difficult for new imagery in
:
Persian poetry, a new literary language such as rekhtah-style hindī did
not have a stock of available asnād [sg. sanad], or what I have translated
as “warrants.” The solution to this aesthetic quandary was to argue that
vernacular composition and the cosmopolitan literary tradition follow
similar rules.

The Pre-Colonial Language Economy

What was the economy of language in northern India before colonial

interventions changed the equation? Firstly, our conceptual vocabulary
for defining patterns of language use (“language” versus “dialect,” “native
speaker” versus “non-native speaker,” “national” versus “regional,” and
so on) is not at all up to the task of capturing how language was used and
categorized in pre-colonial society. Furthermore, a universal problem in
socio-linguistics, the ineluctable gap between how the language user
perceives his or her communication and how others judge it, is exacer-
bated by the colonial source material subsequently used to describe
language in South Asia. The Linguistic Survey of India (1894–1928)⁸⁹
and previous smaller-scale linguistic survey projects gave the illusion of
solving this problem by removing the user’s subjectivity: Thenceforth,
the user’s perception would not matter because the grammar and
vocabulary of his/her speech would be defined by an outside observer’s
matching it to a pre-defined dialect. These speech patterns, identified as
they were by formal characteristics external to the life-world of the
speakers, did not, of course, correspond with how anyone understood
his or her language. People do not mobilize around others’ perception of
their language unless they themselves internalize that perception.
Ironically, it is the other great colonial linguistic data-gathering oper-
ation, the Census, which proves that the Linguistic Survey was not, in

⁸⁹ On its complexity, see Majeed 2011.

 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

    261

fact, a solution. The identification of people’s mother tongues in the

Census shifted wildly as different names were put to the language in
question. People who were recorded as speaking “Hindustani” in one
census were speaking “Hindi” in the next, and so on. The very process of
census-taking forced people across the sub-continent to engage with a
modern, European sense of how language works: a single mother tongue
that had to be identified with a name. Previously, no one would ever have
been asked what he or she called the language used in the home or in the
bazaar. Furthermore, the Census did not bother with building a picture
of people’s other linguistic habits, e.g., knowledge of Persian or Sanskrit,
or the ability to have a conversation with the people in the next district in
those people’s own dialect. Because the colonial sources have these
inherent limitations, and there was no wide-scale study of Indian lan-
guage use before the colonial period (and specifically before the late
nineteenth century, when the Linguistic Survey and Census were first
carried out across the sub-continent), we must do our best with circum-
stantial evidence to construct a model for how language worked in pre-
colonial society.
The paradox of the pre-colonial Indian eco-system of language as
viewed from the modern perspective can be described as follows: Most
language varieties⁹⁰ were not formally defined (with a grammar, a lexi-
con, and so on) before the colonial period, but particular varieties were
more specifically associated with particular social situations. Today in
the West many people are used to using a single standard variety,
generally the majority language framed officially or unofficially as a
national language, when speaking to their family, writing a document
at work, interacting with a government service, or writing poetry.⁹¹ In
modern India, the situation is somewhat different since society remains
considerably more multilingual. However, a key difference between

⁹⁰ “Variety” is a neutral term in socio-linguistics that sidesteps the problem of defining a

particular kind of language use as a language or a dialect or something else. One definition is “a
linguistic system used by a certain group of speakers or in certain social contexts” (Swann et al.
2004).
⁹¹ Of course these activities would likely be carried out in different registers of that particular
language variety, from the informal register of the home to the highly formal register of a legal
proceeding.
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

262       

modern and pre-modern India is that today there are norms for
Standard Hindi and English as well as standardized forms of regional
languages such as Bengali and Punjabi. These languages have been
developed, largely within the previous century, to function in many
different registers (just as standardized languages in the West aim to
fill all social functions). Previously the only standardized varieties were
literary and bureaucratic languages, whose functional range was limited.
In general, we can assume, this availability of unstandardized varieties
created a situation in which the boundaries of language were more fluid
than we understand them to be today. Indeed, the concept of hetero-
glossia, Mikhail Bakhtin’s term for the multiple voices contained within
the unitary language of the novel, applies equally well to the actual
conditions (as far as we can reconstruct them) of a pre-modern linguistic
environment.⁹² While this might appear to suggest that pre-modern
language was a cacophony, people in the environment would have
understood its protocols through life-long exposure.
I will explore three theoretical claims here. Firstly, multilingualism did
not strike pre-modern Indians as unusual, because different varieties
performed different social functions and one’s mother tongue had no
special status in society. Secondly, the characteristics that defined the
functions of a particular variety are not the ones we might expect. Lastly,
the mechanisms for defining language in this society focused on usage
(for example, composing poetry) rather than attempting to produce a
universally applicable language akin to the national languages of today.
Thus, literary language and the language of the everyday—although they
are intertwined in the life-world of literary people—need to be con-
sidered separately. Such a multilingual intersection operates like a one-
way valve, in that the high-prestige literary language influences everyday
language somewhat freely but there is far less influence in the other
direction. I will avoid the well-trodden ground of how Hindi and Urdu
came to be standardized in the nineteenth and twentieth centuries and
will instead offer a general theory of what came before that process.⁹³

⁹² Bakhtin 2002, 259ff., esp. 273 and 278. ⁹³ Described in King 1994 and elsewhere.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

    263

Multilingualism and the Individual

Our received ideas about language and identity are based on monolin-
gualism because functional multilingualism has generally become the
exception rather than the rule in Western societies, except for second
languages taught in school. More complex societal situations (for
example, diglossia) are understood as degrees of divergence from the
monolingual baseline. This is a problematic presupposition because
multilingualism is the norm across much of the world today, and indeed
was taken for granted in the multi-ethnic empires of Europe’s past.⁹⁴ The
Holy Roman emperor Charles V (1500–56) is apocryphally reputed to
have spoken Spanish to God, Italian to his courtiers, French to his ladies,
and German to his horse.⁹⁵ The linguistic scorched-earth politics of pre-
modern England and France—which depended upon extirpating the
minority languages Welsh and Occitan, respectively—should almost
certainly be seen as historically contingent rather than an inevitable
step in building a modern nation-state.⁹⁶ As languages of empire,
English and French have become widely used second languages around
the world, a fact which is notable in light of the kind of internal
oppression required to make them singular and universal in their
home countries, a pattern of violence that would be echoed in global
colonialism later.
The imposition of a national language has not historically been the
only possible way for linguistic cohesion to exist within a large polity.
China, for example, has a single writing system used nearly universally,
but several different spoken varieties that are not, when spoken, mutually
intelligible. Indian vernacular languages, although broadly defined as

⁹⁴ Evans 1998, 18, 29. Obvious exceptions in Western Europe where multilingualism exists
on a national scale are Belgium and Switzerland, but both are special cases. As far as sub-
national multilingualism is concerned, the Catalan movement in Spain and the resources
devoted to Welsh in Britain are both, especially in the latter case, attempts to reverse the
historical tide. The fate of Irish is similar even though it is constitutionally the national language
of the Republic of Ireland.
⁹⁵ Burke 2004, 28. In other versions of the story the distribution of languages is different.
⁹⁶ The 1536 Welsh Act of Union that joined Wales politically with England was uncom-
promising: It declares that “no personne or personnes that use the Welsshe speche or langage
shall have or enjoy any maner office or fees within the Realme of Englonde, Wales or the other
Kinges dominions” (quoted in Evans 1998, 21).
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

264       

distinct from one another even in pre-modern times (for example, in a

:
list Amīr Khusrau provides in his poem Nuh Sipihr [The Nine Heavens],
considered in detail later in the chapter), nonetheless share a core
predicated on Sanskrit and Sanskritic culture.⁹⁷ Linguists refer to South
Asian languages as a Sprachbund (lit. “federation of languages” in
German) in recognition of this unity. Even stranger in the context of
expectations based on European experience, the New Indo-Aryan lan-
guages formed an unbroken cline stretching from Afghanistan to Assam
(apart from Maldivian and Sinhalese separated by water from the other
languages and spoken on the islands of the Maldives and Sri Lanka). This
is to say that language changes gradually across that vast swath of
territory. If one traveled east to west or west to east from one end of
northern India to the other, stopping in every village along the way, the
speech of each new village might be slightly different from that of the
previous village but would almost never vary from it significantly enough
to mark a clear distinction.⁹⁸ Because one would have to travel far to
encounter a difficult-to-understand language, the view from the village
level would have assumed not difference but similarity to be the defining
feature of people’s language. Indo-Persian scholarship’s tendency to
define all Indic language use as “zabān-i hindī” or “hindawī” is perhaps
not the result of imprecision (or rather ignorance of the complexity of
India’s linguistic eco-system) but rather a reflection of a contemporary
perception of language in India.⁹⁹
Of course, anyone who was literate or traveled encountered different
language varieties. Without standardized language, people were not
concerned with speaking “properly” but being understood, an experience

⁹⁷ The Dravidian languages, a group which includes the major languages of South India, are a
distinct family and not genetically related to Sanskrit or the New Indo-Aryan languages of
northern India such as Hindi and Punjabi. Nonetheless, the Dravidian languages also reflect a
significant Sanskrit influence.
⁹⁸ Masica 1991, 25.
⁹⁹ Pace Talbot 1995, 712. Thus, in Āʾīn-i Akbarī, Abū al-Fazl
writes “Throughout the wide
extent of Hindustan, many are the dialects that are spoken, and the diversity of those that do not
exclude a common inter-intelligibility are innumerable” (trans. Āʾīn-i Akbarī 1947, 133). This is
precisely what I have argued, and it squares with the common aphorism that the language in
India changes over a short distance: “Kos kos pah badale pānī, chār kos pah vānī” : [Every kos
( 3 km), the water changes, and every four kos ( 12 km) the speech]. I have been unable to
determine how old the expression in this and its related forms actually is.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

    265

that the language cline made possible since the view at ground level,
except in the case of people who traveled long distances, was that the
people over there speak differently from us but we understand them and
they us. The discomfort at the core of translation studies, namely that
there is an inherent loss of meaning in any translation, was not operative:
It had not yet been articulated except in the special case of scripture. One
surprising commonality between Indo-Persian and European writers in
the eighteenth century is the shared metaphor of a change of clothes
for translation.¹⁰⁰ A costume change was considerably easier than the
soul-searching that often accompanies translation today. Furthermore,
the idea of diglossia, either as it was originally defined by Charles
A. Ferguson or in its extended form as employed by Joshua Fishman
and others, cannot apply here because it presupposes linguistically meas-
urable boundaries between languages.¹⁰¹ In everyday life, people must
have accepted considerable variation in usage, essentially being what we
would call multilingual without having recourse to such a concept. The
vast majority of people would never have dealt directly with a standard-
ized language and so would have had no ideological difficulty adopting
terms from high-prestige languages like Persian.¹⁰² Of course, even
within the life-world of the relatively monolingual village, there was
sometimes a need to access religious or literary or bureaucratic language,
and in these cases someone like a Brahmin or a Persian-literate village
official could mediate. The exclusionary capacity of standardized lan-
guages should not be underestimated, but by the same token neither
should the potential for language-facilitating transactions be ignored.

¹⁰⁰ Both Ārzū and Jones use the image. See, for example, in the preface to Jones’s Persian
Grammar on Persian works in “European dress” and M 1991, 167 on Persian garb [libās] for
Arabic words.
¹⁰¹ In Ferguson’s original conception, the high-prestige variety (H) in a society was a
standardized language and the low-prestige variety (L) an unstandardized dialect of it
(Ferguson 1959). That obviously cannot apply here because there was no standard, all-purpose
hindī to be H. Joshua Fishman’s extension to the theory, namely that H and L need not be
related varieties, still cannot capture the complexity of early modern northern India where khar: ī
bolī-based rek: htah exhibited features of both H and L. The kind of profound multilingualism
present in that society has not been adequately theorized in the framework of diglossia (one
attempt is Mackey 1986).
¹⁰² Anecdotally, rural speech in India since Independence has tended to have more Perso-
Arabic vocabulary than urban speech because education in “proper” (which is to say,
Sanskritized) Hindi was not as widely available.
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

266       

Literacy, which was comparatively rare, had a strong correlation with

power, as we see in the following sections. Certain languages were not
written, and this would not have struck anyone as odd: Letters were often
composed in Persian not because the spoken language of the sending
party would not have been understood by the recipient but because the
spoken language was not, as a matter of course, written down.¹⁰³ Thus,
having someone draw up a document in Persian does not have the same
implication as having a translation done today. Likewise, poetry was
composed in translocal varieties like Braj not necessarily for any com-
municative reason but because that was seen as the variety proper to
poetry of a certain sort.

Relevant and Irrelevant Linguistic Distinctions for

Pre-Colonial South Asia

Which distinctions between varieties are useful from the perspective of

historical inquiry? For simplicity’s sake, we can take a structuralist
approach and posit a set of linguistic binaries, within which there is
inevitably considerable overlap: The key distinction must be between
local versus translocal varieties. The latter are what I have been calling
cosmopolitan languages, which are different from “national” languages
(even if national languages are often developed to facilitate translocal
communication within the nation-state). Another crucial distinction is
between “literarized” and “non-literarized” varieties, that is between
varieties seen as suitable vehicles for literature and those seen as not
suitable. Then there are commonly written versus non-written varieties,
as well as varieties connected with religious practice and those not. We
could continue drawing ever more subtle gradations, such as whether a
variety is deemed suitable for facilitating commercial transactions or not.

¹⁰³ This is in stark contrast to the sense today that it is a human right to have one’s own
language written. Documentary linguists, sometimes for scientific and sometimes for humani-
tarian purposes, are attempting to give every language in the world, no matter how rare and
localized, a written form.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

    267

We rapidly reach a point at which there is simply no evidence to allow us

to judge one way or another.¹⁰⁴
The local and the translocal map onto the categories vernacular and
classical/cosmopolitan, but the fit is problematic since there are translo-
cal vernaculars (for example Braj) and the possibility, albeit exercised
only under exceptional historical circumstances, of identifying a classical
language with its place of origin. Of note is the phenomenon that
Sheldon Pollock has identified as “the dialectic between cosmopolitan
and vernacular that creates them both.”¹⁰⁵ That is, cosmopolitan and
vernacular are not bounded categories but rather sets of characteristics
interpreted against each other. Whether we think of a literature as
cosmopolitan rather than classical or vice versa creates a subtle difference
in framing. A cosmopolitan language will inevitably have a classical
literature, because in order to become cosmopolitan it needs to exist
long enough to spread and needs to have a standard to support that
spread without being transmuted out of recognition. On the other hand,
the medium of classical literature generally must be a cosmopolitan
language. “Classical” puts an emphasis on history and tradition while
“cosmopolitan” emphasizes the contemporary circulation of people and
ideas (and, for that matter, written materials), and together these create
the hegemonic power of a translocal language.
As I argued in Chapter 5, the boundaries of a language both in an
aesthetic or geographical sense were considerably fuzzier before the
modern period: Persian, like Sanskrit, appears never to have articulated
its political boundaries. (Persian and Sanskrit contrast with Latin, which
had the term Latinitas to conceptualize the boundaries of the territory
Rome had Latinized.) Indeed, only from its development as a national
language in Iran in the early nineteenth century did Persianate culture

¹⁰⁴ Socio-linguistic domains, which are what this thought experiment is attempting to
construct, are defined based on “painstaking analysis and summarization of patently congruent
situations” (Fishman 1971, 51). The pre-modern past gives us extremely little data to work from.
The work of J. N. Adams (for example, Adams 2003) on bilingualism and other socio-linguistic
questions in Classical Latin is remarkable for being able to do so much with so little available
material.
¹⁰⁵ Pollock 2000, 616.
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

268       

really become associated with bounded place.¹⁰⁶ Otherwise, there were

various regional usages of Persian, which Ārzū and others noted were not
appropriate as literary language.¹⁰⁷ In other words, the cosmopolitan
language can have non-cosmopolitan variants of itself. The mutual
constitution of cosmopolitan and vernacular are on display in the fact
that one can very nearly write a high-register Bengali sentence in Sanskrit
and a high-register Urdu sentence in Persian.¹⁰⁸ Pollock’s observation
that literary language is deliberately crafted rather than randomly created
is also important in this context.¹⁰⁹ The existence of cosmopolitan
languages did not in any way prohibit vernacular language from becom-
ing literary but in fact was the catalyst for that transformation. I again
return to the observation that non-literary language in this period served
certain circumscribed functions (which could be expanded by developing
a literature) rather than being an important bearer of identity: The
amplification of the vernacular, that is the process of vernacularization,
was a question of changing domains in which the vernacular variety was
considered proper.
Apparently, no ethnic feeling was constructed out of non-literarized
language in India before the nineteenth century, but literary communi-
ties were important frames of reference even if communities of speakers
were not. If a Punjabi went to Awadh then his non-literarized spoken
language would not be understood, but if he were a Sufi who composed
in Persian (or even hindī) then he could find himself a place in a Sufi
community and eventually pick up the local spoken variety. The role
of the “cosmopolitan vernacular” is important because various vernacu-
lars that became transregional built networks of language users.¹¹⁰
I advanced the argument in Chapter 5 that rek: htah-style hindī formed
a large and important network whose parts assembled themselves into a

¹⁰⁶ The Shāhnāmah’s mytho-poetic invocations of Iran (as “ʿAjam” or “Īrān”) are not
territorial, but rather portray civilization struggling against the non-civilized across a topog-
raphy of both real and imaginary places. The inscription of the real locations mentioned in the
poem onto the Iranian nation-state in the nineteenth century has parallels with the Romantic
nationalism of early nineteenth-century Europe.
¹⁰⁷ For the parallels in Classical Latin, see Adams 2007 on regional variation and Adams 2013
on social stratification.
¹⁰⁸ For the Bengali example, see Kaviraj 2003, 512. ¹⁰⁹ Pollock 2000, 591.
¹¹⁰ The term is Pollock’s (1998).
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

    269

coherent, self-aware whole in the eighteenth century. A cosmopolitan

vernacular that can remain a language of power for long enough eventu-
ally becomes capable of the functions of the cosmopolitan classical
language. A cosmopolitan vernacular gains prestige by developing similar
formal characteristics to an established cosmopolitan variety and repre-
:
senting itself as an outgrowth of the classical tradition.¹¹¹ For rekhtah-
style hindī, an important milestone came when people began frequently
writing it down in the same script as Persian and Arabic. The rapid
:
expansion of rekhtah poets’ collected works came towards the end of
Ārzū’s life and afterwards, but Persian remained the language of admin-
istration and serious (non-poetic) thought until much later: In the
eighteenth-century context and specifically in the context of Ārzū, verna-
cularization occurred primarily in poetry and not much in other domains.
Two other literarized languages in South Asia, namely Turkish and
Arabic, complicate the situation and have, unsurprisingly, received com-
paratively less attention than Persian or Sanskrit.¹¹² Central Asian
Turkish was at times widely spoken among the elite of the Delhi
Sultanate and the Mughal Empire, and while a Chinggisid (that is, a
Mongol) identity was often expressed, it never appears to have extended
to language.¹¹³ Famously, the Mughals quickly lost their grasp of
Turkish: Although Bābur wrote his Bāburnāmah in Turkish, by his
grandson Akbar’s time it had to be translated into Persian (by the
remarkable ʿAbd al-Rahīm : :
Khān-i :
Khānān) so that people could still
access it.¹¹⁴ Arabic in South Asia appears to be largely though not
exclusively connected with Islamic religious contexts,¹¹⁵ but perhaps

¹¹¹ Compare a similar observation on Middle English, namely that in “annexing Latin’s
cultural authority, vernacular literatures demonstrate their ability to do anything Latin can do,
while marking their difference from Latin; asserting the prestige of Latin texts and auctores, they
also seek to assimilate that prestige, in an endless shuttling between gestures of deference and
gestures of displacement whose most obvious effect is to tie the theory and practice of
vernacular writing permanently to the question of its status in relation to Latin” (Evans et al.
1999, 321–2). Such “gestures of deference” and “gestures of displacement” are familiar in
rek: htah when poets refer to Persian.
¹¹² On Arabic, see Qutbuddin 2007 and two recent PhD theses (Bahl 2018; Leese 2019).
¹¹³ See, for example, Kumar 2007, 195ff. Indians apparently learned Turkish to work for the
Turkish-speaking Central Asian nobles of the Delhi Sultanate.
¹¹⁴ Lefèvre 2014.
¹¹⁵ Qutbuddin 2007. A notable exception was Āzād Bilgrāmī (discussed in Chapter 5), who
used Arabic both as a literary and critical language.
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

270       

mastery of Arabic was seen as necessary for mastery of Persian. This has
not been systematically studied, but certainly Ārzū’s Hindu friends Bahār
and Mukhli: s: had facility in a style of Persian that was essentially
Arabic.¹¹⁶ They used what might appear as unambiguously Islamic
phrases and concepts, but for them these must simply have been part
of the conventions of Persian discourse and not necessarily religious
statements. The modern distinction between secular and religious is
impossibly anachronistic in such cases.
Although various distinctions such as these make little intrinsic sense
to us, thinking through them is crucial so that we can consider the ways
in which language was defined in the period in question.

The Social Mechanisms for Defining Language

Who decided what constituted good language? In Ārzū’s conception, we

see that the standard was courtly usage (whether real or imagined) and
he, like apparently every other writer concerned with defining linguistic
norms at this time, focused on literary language rather than the language
of everyday life. Another brief mention of linguistic habits, this time in
:
Mīrzā Khān’s : al-Hind, considers Sanskrit, Prakrit, and “bhāshā”
Tuhfat
:
in their literary functions.¹¹⁷ Mīrzā Khān notes that bhāshā “is the
language of the world in which we live” (and indeed refers to it as
especially pertaining to the language of the Braj country), but he is not
interested in drawing any further distinctions and states that his reason
for compiling the grammar was because poetry is written in the language.
Likewise, when non-literary language comes into Ārzū’s work, it is
almost without exception as a prop for a discussion of literary language.
The cosmopolitan and by extension the cosmopolitan vernacular are

¹¹⁶ Ārzū refers to his young friend Bāl Mukund Shuhūd, a Kayasth, who “although he has not
gained that much knowledge of Arabic sciences” [bā ān kih chandān kasb-i ʿulūm-i ʿarabiyah
nah-kardah], has made a study of Euclid [Iqlīdas] but has not had the leisure to study it further
(MN 2004, II:855). What is notable here is that Ārzū implies that a Hindu from a long line of
administrators should know Arabic.
:
¹¹⁷ Tuhfat al-Hind is discussed in more detail in Chapter 5. The section under discussion
here is Grammar of the Braj Bhakha 1935, 34–5.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

    271

much better represented in accounts of language than is the vernacular of

the masses.
For the mass vernacular (that is low-prestige, non-literary, and non-
written language patterns), linguistic standards were maintained orally,
and members of the speech community would naturally imitate speakers
with higher prestige. This is a familiar pattern in non-literate societies
and would have been the same for such varieties. Even though writing
was an available technology, it was not applied to these varieties, which of
course constituted the vast majority of language used in India or indeed in
the pre-modern world. There is no reason to think that anyone was
concerned with setting a standard for any of these informal varieties,
but of course, there is a problem with the evidence: Imprints of everyday
speech form an infinitesimal fraction of the documents from this time.
When people using such language interacted in formal contexts that were
recorded in writing, their language would be translated, usually without
even marking the act of translation, into the language of written records.
For the purposes of intellectual history, we are concerned with vernacular
language only as the elite used it because that is what is available to us.
The Indo-Persian tradition names various Indic regional varieties but
:
deals with them in the most general terms. Amīr Khusrau famously did so
in the thirteenth century in his mas̄nawī Nuh Sipihr [The Nine Heavens].
His account is not in any sense a linguistic survey and makes no claims as
far as the mutual intelligibility of the language varieties mentioned.
Indeed, the phrasing strongly implies that the list of thirteen regional
varieties is not meant to be exhaustive. The argument is straightforward:
:
In a section of the poem that is a panegyric on India, Khusrau presents the
case for India’s greatness in the context of language: Persian, Turkish, and
even Arabic, are simply not as pleasant as the diverse languages of India.
He provides a list of these and concludes that:

All these [language varieties] are hindawī, which from ancient times
:
has been commonly used for every kind of speech [sukhan].¹¹⁸

¹¹⁸ “īn hamah hindawī-st kih az aiyām-i kuhan / ʿāmah ba-kār ast ba-har gūnah suk: han.”
The passages under discussion here are at Nuh Sipihr 1950, 179–80. See Faruqi 2001, 65ff. on the
problems of identifying the languages mentioned; cf. Narang 1987, 29, who makes several
interpretative mistakes that Faruqi corrects. Further context available in Nath and Gwaliari
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

272       

Thus, all the varieties mentioned are to be considered “hindawī,” which

we can translate as “Indic” to reflect the broadness of its meaning.
According to Shamsur Rahman Faruqi, the term does not, in Khusrau’s :
usage, carry any analytical force or refer to a specific kind of language
use. Faruqi (and my translation) interprets “sukhan”: generally, meaning
that these languages are used for every kind of “speech.” Sukhan: can, of
:
course, mean poetry specifically, and perhaps Khusrau does use the term
in this sense, because he is arguing that the best varieties for “delighting
the heart” [dil ba-t̤ arab kardan] (i.e., through poetry or music) are Indic.
He specifies that each of the Indic varieties is proper to a defined place,
fitting our definition of local (or non-cosmopolitan vernacular) varieties.
Although there is no consideration of writing, the section ends with a
mention of Sanskrit:

But there is a different language from these spoken ones

This is the chosen one for all the Brahmins.
Sanskrit is its name from ancient times;
The masses know nothing of its hidden power.
The Brahmin knows it, but not even every Brahmin
For that matter knows the [full] extent of the discourse.¹¹⁹

This is obviously another rhetorical proof of India’s greatness (that is,

that no other place in the world can boast of having mystical Brahmins
with a secret language) rather than a careful analysis of language ideol-
ogy, but it does illustrate the hegemony of Sanskrit and its transcendence
:
of the languages of place. A list similar to Khusrau’s is the emperor
Bābur’s account of the varieties spoken around Kabul.¹²⁰ Again, this is an
inventory offered without comment on how the languages identified

:
1981. Alyssa Gabbay deals with Khusrau’s legitimation of Indic language but does not appar-
ently consider the problem of what hindawī means—she appears to take it straightforwardly as
the name of a defined language (2010, esp. 20ff.).
¹¹⁹ “lek zabānī-st digar k-az suk: hanān / ān-ast guzīn nazd-i hamah barhamanān / sanskrit
nām zi ʿahd-i kuhan-ash / ʿāmah nadārad k: habar az kann-i mikan-ash / barhaman-ash dānad
wa har barhamanī / nīz nadānad hadd : zānsān suk: hanī” (Nuh Sipihr 1950, 180).
¹²⁰ “Eleven or twelve dialects are spoken in Kabul province: Arabic, Persian, Turkish,
Mongolian, Hindi, Afghani, Pashai, Parachi, Gibari, Baraki, and Lamghani. It is not known
whether there are so many different peoples and languages in any other province” (translated in
Thackston 1993, 2:270).
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

    273

relate to one another, but Bābur concludes that Kabul is probably the
most linguistically diverse province then under his control. At the court
of his grandson Akbar, a list of languages was prepared by Abū al-Fazl
to
be included in Āʾīn-i Akbarī. While Abū al-Fazl
notes that there are
“innumerable” varieties in India that are mutually intelligible, he pro-
vides a list of places whose varieties would not be understood by speakers
of the other varieties. We would understand the languages as (kharī : bolī)
Hindi, Bengali, Punjabi, Rajasthani, Gujarati, Telugu, Marathi, Kannada
(and Tamil?), Sindhi, Pashto, Baloch, and Kashmiri, but he does not
provide these varieties’ names but rather states the places where each
variety is typical.¹²¹ For his purposes, language varieties were akin to
geographical features rather than markers of communities or nations.
This geographical versus communitarian understanding of language
encapsulates the key distinction between our noticing that there is a
mutually intelligible language variety spread across most of northern
India, and Indo-Persian writers of the eighteenth century’s noticing the
same thing. Amrit Rai’s seminal analysis of how pre-modern hindī came
to be Modern Standard Hindi, A House Divided, tries to split the
difference but ends up projecting a linguistic consciousness on the past
that is too much like modern Hindi to be fully credible.¹²² To avoid the
possibility of such a conflation, I have made the slightly grating choice to
use the term “hindī” throughout this study in deference to the Persian
primary sources. In the Persian context, differences within hindī were
only recognized to the degree they needed to be in a given context. For
example, Bengali could be identified as “zabān-i hindī-yi bangālah” (lit.
“the Indian tongue of Bengal”). Literary languages tend to be written
languages, and so are governed by rules that are recorded for pedagogical
purposes and even debated since, as I have argued, early modern writers
did not separate descriptive grammar from what we would bracket off as
aesthetic considerations. There were even grammars of Persian available
in Sanskrit, for example the late sixteenth-century Pārasīprakāsha [The
:
Illuminator of Persian], attributed to Kr̥s: nadāsa. We should remember

¹²¹ That is (in the same order): Delhi, Bengal, Multan, Marwar, Gujarat, Telangana,
“Marhatta,” Carnatic, Sindh, “Afghan” of Shāl (the region between Sindh, Kabul, and
Qandahar), Balochistan, and Kashmir.
¹²² Rai 1984.
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

274       

these are not descriptive grammars, but rather present an idealized view
of the language.¹²³
How does influence move in these languages, and does this movement
of vocabulary/structure reflect the social hierarchy of language varieties
themselves? Power was not exercised through a government’s defining
language for the populace or the nation (unlike, say, the Académie
Française and other comparatively recently established institutions). It
was instead a matter of determining the acceptable forms within the
language of power.¹²⁴ Obviously the practices of the imperial chancellery
set standards for Persian documentary conventions, but it would have
had absolutely no interest in regulating language use outside that narrow
:
compass. Poetry itself was a science [ʿilm], as Khusrau shows through
complex exegetical reasoning.¹²⁵ While poets like Ārzū were deeply
concerned with the systematization of Persian poetry, they did not
seriously consider other purposes to which the language might be put.
It is hard to know whether people outside bureaucratic and poetic circles
appreciated these niceties—or if that would even matter. The difficulty is
that Indo-Persian was undoubtedly a link language often used for rela-
tively informal communication, but the documentary evidence generally
records only its more formal usage. We can surmise that this informal
usage was how Persian came to influence so many vernacular languages
in India. In a society where certain language varieties were associated
with the elite, the language varieties of the non-elite tended to be
undifferentiated, while elite language use filtered down into the con-
sciousness of upwardly mobile low-level elites and non-elites. Building
a majoritarian politics upon language is difficult, perhaps impossible,
without the sort of enumeration that a census provides and the
corresponding realization of strength in numbers. The colonial-era

¹²³ An interesting parallel is the relationship in Sanskrit thought between shāstra [theory]
and prayoga [practical activity], which, as Sheldon Pollock has shown, was viewed in traditional
India in precisely the opposite way from the usual Western understanding (Pollock 1985, 511).
To summarize a complex argument: In the West, the formulation is that practice is followed by
a descriptive theory while in Sanskrit thought practice is an adumbration of the shāstra even if
the practitioner is unaware of it. While Perso-Arabic philological thought generally follows the
Aristotelian (that is, the Western) pattern of description following practice, it also has prescrip-
tive tendencies. More study is warranted of the meeting between the philological traditions of
Sanskrit and Perso-Arabic.
¹²⁴ For example, the Sanskrit praśasti (Pollock 2006, 134ff.). ¹²⁵ See Faruqi 2001, 91.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

    275

consciousness of how many speakers of a language there were brought

the conception of northern Indian language full circle: The Persian
zabān-i hindī was replaced by a similar totalizing construction of Indic
language, namely Modern Standard Hindi.¹²⁶ Unlike the Indo-Persian
concept, this standard Hindi would be based on drawing sharp distinc-
tions between proper and improper usage, and in particular employing
(and in many cases artificially introducing) Sanskritic vocabulary as the
glue that would connect the kharī : bolī core of the language with the other
languages of India. The older notion of Indic language could not func-
tion for nationalist politics, but its subsequent redefinition could. Today
the project of Modern Standard Hindi as a national language is compli-
cated by the vast number of people who nominally speak it as a mother
tongue but whose actual usage is at variance with the standard
language.¹²⁷

“Imagine There’s No Countries”

With apologies to John Lennon, imagining a world without nation-states

is in fact hard for us to do. The desire of those living in the present to find
the ancient seeds of their polity and culture conscripts historiography
into the unforgiving service of a false idol.¹²⁸ The study of the European
Middle Ages has arguably been deeply distorted for two centuries
because of the tendency to force nation-states as a unit of analysis onto
a historical record for which they are ill-suited.¹²⁹ In India, more perni-
ciously, an increasingly widely accepted pseudo-history locates various
aspects of the Hindu nationalist program for modern India in what can
be best described as mythical time. This is a nationalism that rhetorically

¹²⁶ See Das 2014.

¹²⁷ This has sparked regional movements in India agitating for what have been thought of as
dialects of Hindi, such as Maithili, to be officially declared independent languages. Maithili was
added to the Eighth Schedule of the constitution in 2003 and thus is now recognized as a
language instead of a dialect.
¹²⁸ Memorably, Marc Bloch decried historians’ devotion to “the idol of origins” when there
are many other more worthy questions for historical research to address (Bloch 1992, 24ff.).
¹²⁹ Chris Wickham identifies the “two grand narratives” that have caused misunderstanding
of the Middle Ages in recent historiography as “the narrative of nationalism and the narrative of
modernity” (Wickham 2009, 3).
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

276       

erases its own true origins in the colonial encounter and seeks to
represent itself as indigenous when it is patently not. Of the various
possible misuses of history or departure points for pseudo-history, the
identification of a language with nation is particularly tempting and has
also been amplified by other disciplines such as literary studies, phil-
ology, anthropology, and linguistics. Karla Mallette observes that
philology, notable as a pre-modern discipline that survived the transition
to modernity, was cosmopolitan until it turned nationalist:

But more recently, Philology has been drawn into a marriage of

convenience with the European nationalisms, and this has dulled her
senses somewhat—or, at least, has predisposed her to see the grand
sweep of history in a certain light.¹³⁰

This (unhappy?) marriage of history and philology has had several effects
that we still need to pick our way through: Firstly, scholars—even from
multilingual backgrounds themselves—have often discounted multilin-
gualism as a historical phenomenon, because from the nationalist van-
tage point multilingualism must be a kind of false consciousness on the
path to the right and proper state in which people use and identify with
their national language. Hearteningly, given the number of recent studies
in various historical contexts focusing on multilingualism (some of
which are cited in this chapter), this reflex is arguably being widely
corrected. Secondly, scholars have often not given sufficient weight to
cosmopolitan languages because a cosmopolitan language does not
appear to lead to the nation or indeed seems to block the full flowering
of the national language—this jaundiced view is common in scholarship
on both Europe and the non-West. Stepping back from expectations of
the nation-state, it is obvious that people in the past could live their lives
writing and speaking in learned languages like Latin or Persian, and they
were not (as is often implied) yearning to express themselves in a
different, vernacular language that was closer to their hearts. Lastly,
scholars interpreting cultural phenomena have been predisposed to see
cultural chaos outside of Europe and measured, deliberate steps towards

¹³⁰ Mallette 2013, 260.

 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

    277

modernity in Europe. More than anything else, these vastly different

readings of broadly similar cultural developments reflect the fact that
Europe was able to write its own story of modernity. In fact, Europe’s
march to modernity was messy and its colonizing nation-states often the
product of internal colonialism of previous centuries. The cosmopolitan
possibilities of pre-colonial South Asia should therefore be understood
on their own terms rather than fit into an idealized European pattern.
Present-day scholars must turn the interpretative lens on themselves in
order to do so.
 OUP CORRECTED PROOF – FINAL, 18/2/2022, SPi

Conclusion

Contrary to later depictions of the eighteenth century as a dark time of

cultural stagnation waiting to be dispelled by the dawn of European
rationality or by Indian/Pakistani nationalism, this study has shown
the vibrancy of Indo-Persian philological scholarship during this period.
In their analysis of literary language, Ārzū and his circle posed funda-
mental questions about the nature of language in society. Indeed, it is
somewhat surprising—as it was to Ārzū as well—that given the cultural
intermixing of Islamicate northern India, it was not until the eighteenth
century that a systematic enquiry into the relationship between Persian
and other Indian languages took place. I have tried to situate Ārzū in a
global early modern context, which is to say that his methodology
maintained a deep respect for tradition while stretching that tradition’s
fundamental assumptions to breaking point and sometimes beyond. This
study has argued consistently that many of the received ideas about the
late pre-colonial period would have been simply unthinkable to the
people whose milieu we are trying to explain.
Ārzū was well positioned for a career in literature and philological
scholarship. His standing in Delhi’s literary circles came from his lineage
(which included descent from important literary Sufis on both sides), his
family’s tradition of imperial service, and his influential friends and
mentors. It was Ānand Rām Mukhli : s: who facilitated his entry into
Delhi high society. His close ties with Mukhli : s: and Tek

Chand Bahār,
both Hindus, show the influence of non-Muslims in Persian and the
multiconfessional nature of the tradition despite its Islamicate underpin-
nings. Ārzū did not lack for patrons, and although they would begin to
fray not long after his death, the intellectual networks that had sustained
Persianate intellectual culture in India were still in place during his
: s: and Bahār wrote import-
lifetime, despite political turmoil. Both Mukhli
ant dictionaries, and both their projects are connected with Ārzū (in the

:
India in the Persian World of Letters: Khān-i Ārzū among the Eighteenth-Century Philologists. Arthur Dudney,
Oxford University Press. © Arthur Dudney 2022. DOI: 10.1093/oso/9780192857415.003.0008
 OUP CORRECTED PROOF – FINAL, 18/2/2022, SPi

 279

case of the former because Ārzū posthumously edited the manuscript

and in the latter because Bahār used Ārzū’s Sirāj al-Luġhat in preparing
his own dictionary). Ārzū was a mentor to a number of poets including
many of the disciples of the late Bedil, who during his lifetime had
arguably been the doyen of Persian poets in Delhi. Ārzū successfully
established himself as a guardian of Bedil’s legacy while also tapping into
: wush.
the other important literary network in Delhi, that of the late Sarkh :
He makes this positioning clear in his tażkirah Majmaʿ al-Nafāʾis. The
: Hazīn
arrival of Shaikh
formalized an opposition to Ārzū since Hazīn,


like Munīr in a previous generation, was a defender of the classical style
against recent poetic developments. It was perhaps this personal conflict
that caused Ārzū to stop writing relatively staid commentaries on clas-
sical texts and begin theorizing a defense of contemporary poetics, in
some instances making his case vehemently.
Ārzū’s theory of language recognizes the deep ties between languages,
the prerogative of urban elites to set literary standards, the need to
balance tradition with the consensus of living poets in order to account
for changing tastes, and the principle that even a native speaker makes
mistakes and so is not automatically an authority. Ārzū’s expansive
treatise on the theory of language (Mus̄mir) is built upon an Arabic
work (al-Suyūt̤ ī’s al-Muzhir) that he revisited in part so that he could
scrape away whatever was irrelevant to his purposes and build within its
hollowed-out categories a theory appropriate to the complex interaction
of languages in his own life-world. This, I have argued, is a typically early
modern approach to knowledge production. By the end of the eighteenth
century, the Persian cosmopolis had begun to fragment, but Ārzū did not
see the nascent changes in the way that many later scholars have
understood them. Instead, he proposed a flattening of the Persian cos-
mopolis, in which eloquent literary language is recognized as the same
across its vast expanse even though obviously local language and dialect
will be different. His thinking was undeniably historical, and he theor-
ized how words have moved from one language to another. Notably, he
adapted al-Suyūt̤ ī’s notion of tawāfuq al-lisānain [correspondence
between languages] to explain the primordial connection between
Persian and “hindī-yi kitābī” [i.e., Sanskrit], which (by his own state-
ment) is the first time this was addressed in the Persianate tradition.
OUP CORRECTED PROOF – FINAL, 18/2/2022, SPi

280       

From our perspective, it is very much like the fundamental principle of

(Western) historical linguistics that languages branch off from one
another over time, but Ārzū does not define it precisely enough for us
to be sure of that reading. Although Mus̄mir was not a widely known
text, the ideas it presents, such as tawāfuq or its discussion of metaphor-
ical language, are present throughout Ārzū’s earlier texts. Yet claims that
Sir William Jones perhaps appropriated Ārzū’s application of tawāfuq in
the development of his own theories on linguistic development do not
rest on a particularly sound foundation. Instead of fitting Jones into the
tradition of Orientalist appropriation of non-Western ideas, we should
instead consider the remarkable parallel between Indo-Persian and
Western philology that allowed them to come to similar conclusions
about how languages develop over time.
Our terms for explaining the transformations in the world of Persian
letters in the late pre-colonial period are premised on an anachronistic
periodization of Persian literary history and on similarly anachronistic
nationalist assumptions. Ārzū’s main concern was not with Iranian
versus Indo-Persian usages, but rather with the controversy over the
relatively new (and for him not satisfactorily theorized) literary style of
tāzah-goʾī, which needed to be fit into the tradition. Rather than looking
for evidence of cultural decline, we are better served by comparing this
state of affairs to the roughly contemporary Quarrel of the Ancients and
the Moderns in Europe, which is remembered fondly in contrast to the
supposed cultural dead-end of the “Indian Style” in Persian. In the
eighteenth-century Indo-Persian context, the Ancients [mutaqaddimīn]
are the poets through Jāmī (fifteenth century) and the Moderns
[mutaʾak: hkhirīn]
: are the poets who came after, who are especially
identified with the “tāzah-goʾī” style. As in Indo-Persian, the difficulty
faced in Europe was assessing the relative value of tradition compared
with new knowledge that was being produced. Ārzū’s insight was that
contemporary poets are always making literary precedent [sanad], just as
earlier poets made the precedent that must be followed by contemporary
poets. He observes that even native speakers make mistakes (he presents
this by analogy with Indian poets composing in Indic languages), and so
proper training in literary production rather than simply knowing a
language is crucial in his view. He argues against the criticism of the
 OUP CORRECTED PROOF – FINAL, 18/2/2022, SPi

 281

likes of Hazīn

and Munīr by out-researching them: He shows in many
instances that perceived faults of the Moderns to which such critics had
objected were actually acceptable to the Ancients, as proved by a more
careful reading of precedent-establishing works. Being attuned to the
linguistic variation across the Persian cosmopolis allows him to propose
a consolidation of the tradition that incorporates contemporary poetics
into the longer history of literary Persian.
The development of Urdu vernacular poetry in the eighteenth century
should not be understood as a deliberate attempt to replace Persian with
an Indic language or to democratize the literary scene. The great
Persianists of the day were the same people promulgating a standard
for Urdu. Ārzū’s formulation of literary Persian as a dialect that histor-
ically emanated from the royal court is exactly parallel to the definition of
Urdu as the literarized vernacular of the royal court in Delhi. Poets
working in the rek: htah style apparently received advice from Ārzū on
aspects of this standardization. The idea that the motivation for devel-
oping Urdu was that it was Indians’ natural language as opposed to
“foreign” Persian rests on shaky foundations. It was only during the
colonial period, in a very different cultural moment, that this framing
was adopted to justify eliminating Persian, a move which, however it
affected Indians, undeniably helped the colonial state’s political project.
:
Rekhtah was developed as another avenue for poetic expression without
necessarily providing any scope for a national identity, as many later
literary histories claim. Indeed, people’s relationship to language was
entirely different in the era before the nation-state became the base unit
of polity.
Reader, I owe you some parting words on cosmopolitanism. In India
today, many aspects of the cosmopolitan past described in this study are
now coded as “Muslim” and because of the majoritarian chauvinist turn
in Indian politics, they are therefore probably irretrievable. It is an
incalculable loss. Intellectual history can show us possibilities for imagin-
ing society that seem remote under current circumstances, though we
must be vigilant about romanticizing the past as we are engaged in
imagining it. There are conceptual alternatives to the nation-state avail-
able to us, one of which is a transnational political and aesthetic culture
of the sort that the Persian cosmopolis represented. About a decade ago
OUP CORRECTED PROOF – FINAL, 18/2/2022, SPi

282       

when this project was in its early stages, I imagined that, both in the West
and especially in the postcolonial world, we could scrape off the dead
wood of nationalism and expose a living, pre-modern cosmopolitanism
that could be nursed back to health. I no longer think so. No depth of
knowledge of a cosmopolitan past will bring about a cosmopolitan future
because what is relevant is people’s inherent desire to be cosmopolitan or
not. Historical knowledge cannot on its own create that orientation.
Literature, on the other hand, is fundamentally imaginative and is
therefore one of our most powerful tools for inspiring people to make
cosmopolitan connections. However, cosmopolitanism built upon a for-
mal literature entrenches the privilege of an elite and there is simply no
way around that fact because literate cosmopolitans share a canon that
they must have had the opportunity to study. Some people are always
excluded from mastering this canon, either by circumstances or by
institutional gatekeepers. Cosmopolitanism, I was humbled to learn in
the past few years, is not a panacea even when it comes close to its ideal
form. Sometimes I am hopeful about the prospects of building a more
cosmopolitan culture in the communities in which I have a stake, and
sometimes, especially when I consider how universities are struggling,
I despair. This self-reflection may be an odd way to end an academic
book, but these odd times seem to call for it. I wish you well on your own
journey.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

Bibliography

Note: For abbreviations of Ārzū’s major works used here, see Abbreviations of Ārzū’s
Major Works on p xi.

Primary Sources

Āb-i Hayāt

:
of Muhammad Husain

Āzād. Lahore: Naval Kishore, 1907.
Āb-i Hayāt

of Mu :
hammad Husain

Āzād. Translated by Frances Pritchett in associ-
ation with Shamsur Rahman Faruqi. New Delhi: Oxford University Press, 2001.
:
Ahkām-i ʿĀlamgīrī of Aurangzeb. Edited by Jadunath Sarkar. 2nd ed. Calcutta:
M. C. Sarkar, 1926.
Āʾīn-i Akbarī of Abū al-Fazl
ibn Mubārak Allāmī. Edited by Henry Blochmann. Vol. 2.
Calcutta: Asiatic Society of Bengal, 1869.
Āʾīn-i Akbarī of Abū al-Fazl
ibn Mubārak Allāmī. Translated by Henry Blochmann.
Vol. 1. Calcutta: Royal Asiatic Society of Bengal, 1873.
Āʾīn-i Akbarī of Abū al-Fazl
ibn Mubārak Allāmī. Translated by H. S. Jarrett and
revised by J. N. Sarkar. Vol. 3. Calcutta: Royal Asiatic Society of Bengal, 1947.
Ātashkadah-yi Āżar of Lut̤ f ʿAlī Baig Āżar. Edited by Hasan Sadat Nasiri. Vols 1–3.
Tehran: Amir Kabir, 1957.
Ātashkadah-yi Āżar of Lut̤ f ʿAlī Baig Āżar. Edited by Mir Hashem Muhaddis. Vol. 4.
Tehran: Amir Kabir, 1999.
ʿAt̤ īyah-i Kubra̍ of Sirāj al-Dīn ʿAlī Khān
: Ārzū. In Shamisa, ed. 2002.
Bahār-i ʿAjam of Lālah Tek
Chand Bahār. Ms Caps Or B 15. Bodleian Library,
Oxford University.
Bahār-i ʿAjam of Lālah Tek
Chand Bahār. Edited by Kazim Dizfuliyan. 3 vols.
Tehran: Talayah, 2001.
:
Bikat: Kahānī of Muhammad Afzal.

Edited by Nur al-Hasan Hashmi and Masʿud
Husain Khan. Lucknow: Uttar Pradesh Urdu Academy, 1979.
Biographical Notices of Persian Poets of Gore Ouseley. London: Oriental Translation
Fund, 1846.
:
Burhān-i Qāt̤ iʿ of Muhammad Husain

:
ibn Khalaf Tabrīzī Burhān. Bombay, 1850.
Chamanistān-i Shuʿarā of Ray Lachhmī Narāʾin Shafīq. Hyderabad: Matba-yi
Anjuman-i Taraqqi-yi Urdu, 1928.
Chirāġh-i Hidāyat of Sirāj al-Dīn ʿAlī Khān : Ārzū. Mss Or 2013, Or 264, and IO
Islamic 71. British Library.
Chirāġh-i Hidāyat of Sirāj al-Dīn ʿAlī Khān : Ārzū. Edited by Mansur Sarwat. Bound
in Ġhiyās̄ al-Luġhāt. Tehran: Amir Kabir, 1984.
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

284 

:
Dād-i Sukhan of Sirāj al-Dīn ʿAlī Khān
: Ārzū. Edited by Sayyid Muhammad Akram.
Rawalpindi: Iran Pakistan Institute of Persian Studies, 1974.
Dībāchah-yi Dīwān-i Ġhurrat al-Kamāl of Amīr Khusrau. : Edited by Sayyid Vazir
Alhasan ʿAbdi. Lahore: National Book Foundation, 1975.
Dīwān-i Fārsī-yi Hātim

Dihlawī of Z̤ uhūr al-Dīn Hātim.
Edited by Mukhtar al-Din
Ahmad. Rampur: Raza Library, 2010.
Dīwān-i Ġhālib of Mirzā Asadullāh Khān : Ġhālib. Edited by Imtiaz ʿAli ʿArshi. New
Delhi: Anjuman-i Taraqqi-i Urdu, 1982. Original ed., 1958.
Dīwānzādah of Z̤ uhūr al-Dīn Hātim.
Edited by Ghulam Husain Zulfiqar. Lahore:
Maktabah Khiyaban-i Adab, 1975.
Farhang-i Jahāngīrī of Mīr Jamāl al-Dīn Husain
Injū. Ms RSPA 20. British Library.
Farhang-i Jahāngīrī of Mīr Jamāl al-Dīn Husain
Injū. Edited by Rahim ʿAfifi.
Mashhad: University of Mashhad, 1975.
Farhang-i Majmūʿat al-Furs of Abū al-ʿAlāʾ ʿAbd al-Muʾmin Jārūtī known as Safī-yi

: Edited by ʿAziz Allah Jawayni. Tehran: Bunyad-i Farhang-i Iran, 1977.
Kah: hāl.
Farhang-i Rashīdī of ʿAbd al-Rashīd bin ʿAbd al-Ġhafūr. Edited by Maulwi Zulfaqar
ʿAli. Calcutta: Asiatic Society of Bengal, 1875.
Ġharāʾib al-Luġhāt of ʿAbd al-Wāsiʿ Hansawī. Cambridge University Library ms
Eton Pote 291.
Ġharāʾib al-Luġhāt of ʿAbd al-Wāsiʿ Hansawī. Rampur Raza Library ms 2543.
Ġhiyās̄ al-Luġhāt of Muhammad: Ġhiyās̄ al-Dīn. British Library ms Or 260. 1258
(1842).
Ġhizlān al-Hind [as Ġhazālān al-Hind] of Mīr Ġhulām ʿAlī Āzād Bilgrāmī. Edited by
Cyrus Shamisa. Tehran: Sada-yi Muʿasir, 2003 (1382 ).
A Grammar of the Braj Bhakha (Tuhfat : al-Hind) of Mīrzā Khān.: Translated by
M. Ziauddin. Calcutta: Visva-Bharati Bookshop, 1935.
Gulshan-i Guftār of Kh : wājah
: :
Khān Hamīd.

Edited by Sayyid Muhammad.
Hyderabad: Maktabah-yi Ibrahimiyah, 1929.
ʿIbrat al-Ġhāfilīn of Mīrzā Muhammad: Rafīʿ Saudā. Edited by Sharif Husain Qasemi.
New Delhi: Centre for Persian Research, 2011.
:
Iʿjāz-i Khusrawī :
of Amīr Khusrau. Lucknow: Naval Kishore, 1876.
:
Iʿjāz-i Khusrawī :
of Amīr Khusrau. Edited by Habibuddin Ahmed with Mahfooz Ali
Siddiqui and Shahid Ali Abbasi. Vols 1 and 2 translated as The Writings of Amir
Khusrau by Nazir Ahmed and Iqtidar H. Siddiqui. Forest Park, IL: Islamic
Thought and Science Institute, 2007.
ʿIqd-i S̄uraiyā of Ġhulām Hamadānī Mus: hafī. : Edited by ʿAbdul Haq. Karachi:
Anjuman-i Taraqqi-yi Urdu, 1978.
Jawāb-i Shāfī of Siyālkot:ī Mal Wārastah. Ms Add. 795, ff. 109b–130a. Cambridge
University Library.
Jawāb-i Shāfī of Siyālkot:ī Mal Wārastah. Ms 9. Amiruddaulah Public Library,
Lucknow. (Incorrectly attributed to Hākim
in Muradabadi 2000.)
Kārnāmah of Abū al-Barakāt Munīr Lahorī (with Sirāj-i Munīr of Ārzū). Edited by
Sayyid Muhammad Akram. Islamabad: Centre for Research in Persian of Iran and
Pakistan, 1977.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

 285

:
Khiyābān-i Gulistān of Sirāj al-Dīn ʿAlī Khān: Ārzū. Edited by Mehr Nur Muhammad
Khan. Islamabad: Markaz-i Tahqiqat-i Farsi-yi Iran wa Pakistan, 1996.
:
Khizānah-yi ʿĀmirah of Mīr Ġhulām ʿAlī Āzād Bilgrāmī. Ms Ouseley Adds 6.
Bodleian Library, Oxford University.
:
Khulā : al-Afkār of Abū T̤ ālib al-Isfahānī.
sat : Ms Elliott 181. Bodleian Library, Oxford
University.
Kulliyāt-i Mīr of Mīr Muhammad : Taqī Mīr. Edited by Zill-i ʿAbbas ʿAbbasi. New
Delhi: National Council for the Promotion of the Urdu Language, 2003.
The Life of Sheikh Mohammed Ali Hazin of Muhammad : ʿAlī Hazīn.

Translated by
F. C. Belfour. London: Oriental Translation Fund, 1830.
Luġhatnāmah of ʿAlī Akbar Dihkhudā. : Edited by Mohammad Moʿin and Jaʿfar
Shahidi. Tehran: Tehran University Press, 1994.
Maʾās̄ir al-Umarā of Nawāb Sam
s: am al-Daulah Shāh Nawāz Khān. : Edited by Mirza
Ashraf ʿAli. 3 vols. Calcutta: Asiatic Society, 1891.
Madār al-Afāzil
of Allāhdād Faizī
Sirhindī. Edited by Muhammad Baqir. Lahore:
Punjab University, 1959–70.
Majmaʿ al-Fus: ahāʾ of Rizā :

Qulī Khān Hidāyat. Edited by Mazahir Musaffa. Vol. 1.
Tehran: Amir Kabir, 1957.
Majmaʿ al-Nafāʾis of Sirāj al-Dīn ʿAlī Khān : Ārzū. Ms Elliott 399. Bodleian Library,
Oxford University.
Majmaʿ al-Nafāʾis of Sirāj al-Dīn ʿAlī Khān : Ārzū. Mss India Office Islamic 3316 and
3317. British Library.
Majmaʿ al-Nafāʾis of Sirāj al-Dīn ʿAlī Khān : Ārzū. Edited by Abid Reza Bidar. Patna:
Khuda Bakhsh Library, 1970.
Majmaʿ al-Nafāʾis of Sirāj al-Dīn ʿAlī Khān : Ārzū. Edited by Zebunnisa ʿAli Khan.
3 vols. Islamabad: Markaz-i Tahqiqat-i Farsi-yi Iran wa Pakistan, 2004.
Majmaʿ al-Nafāʾis of Sirāj al-Dīn ʿAlī Khān : Ārzū. Edited by Mir Hashim Muhaddis
(based on Bidar, ed. 1970). Tehran: Anjuman-i Asar wa Mufakhir-i Farhangi,
2005.
Majmūʿah [miscellany] containing several works by Ārzū (Dād-i Sukhan, : Tanbīh al-
Ġhāfilīn, Sirāj-i Munīr, and Sirāj-i Wahhāj). Ms 2452b. Raza Library, Rampur.
Mans̄ūrāt-i Ānand Rām of Ānand Rām Mukhli : s: . Ms HL 882. Khuda Bakhsh
Oriental Public Library.
Mat̤ laʿ al-Saʿdain of Siyālkotī : Mal Wārastah. Lucknow: Naval Kishore, 1880.
Mauhibat-i ʿUz̤ ma̍ of Sirāj al-Dīn ʿAlī Khān : Ārzū. In Shamisa, ed. 2002.
:
Miftāh: al-Khazāʾin of Jayarām Dās. British Library ms Delhi Persian 554A-F.
Mihr-o Māh attributed to Sirāj al-Dīn ʿAlī Khān : Ārzū. Mss 4327f and 4328f. Raza
Library, Rampur.
Mirʾāt al-Is: t̤ ilāh: of Ānand Rām Mukhli : s: . Ms Or 1813. British Library, 1850.
Mirʾāt al-Is: t̤ ilāh: of Ānand Rām Mukhli : s: . Edited by Chander Shekhar, Hamidreza
Ghelichkani and Houman Yousefdahi. Delhi: National Mission for Manuscripts,
2013.
:
Muhākamat al-Luġhatain of Mīr ʿAlī Sher Nawāʾī. Translated by Robert Devereux.
Leiden: E. J. Brill, 1966.
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

286 

Muntakhab : :
al-Ashʿār of Muhammad ʿAlī Khān
: Mubtala̍. Ms Elliott 247. Bodleian
Library, Oxford University.
The Muqaddimah: An Introduction to History of Ibn Khāldūn. : Translated by Franz
Rosenthal and abridged by N. J. Dawood. Princeton, NJ: Princeton University
Press, 2005.
Muraqqaʿ-i Dihlī of Dargāh Qulī Khān. : Edited by Khaliq Anjum. New Delhi:
Anjuman-i Taraqqi-yi Urdu, 1993.
Mus̄mir of Sirāj al-Dīn ʿAlī Khān: Ārzū. Ms 2520f. Raza Library, Rampur.
Mus̄mir of Sirāj al-Dīn ʿAlī Khān
: Ārzū. Edited by Rehana Khatoon. Karachi: Institute
of Central and West Asian Studies, University of Karachi, 1991.
Mus̄mir of Sirāj al-Dīn ʿAlī Khān
: Ārzū (fragment). Ms 2620f. Raza Library, Rampur.
: al-Shuʿarā of Siyālkot:ī Mal Wārastah (with an abridgement of Bahār-i
Mus: t̤ alahāt
ʿAjam as marginalia). Edited by Maulawi Jamil Ahmad. Lucknow: 1853 (1270 ).
Mus: t̤ alahāt
: al-Shuʿarā of Siyālkot:ī Mal Wārastah. Tehran: Muʾassasah-yi
Pizhuhishha-yi Farhangi, 1985 (1364 ).
: al-Shuʿarā of Siyālkot:ī Mal Wārastah. Edited by Cyrus Shamisa. Tehran:
Mus: t̤ alahāt
Intisharat-i Firdaus, 2001.
al-Muzhir fī ʿUlūm al-Luġhah wa Anwāʿihā of Jalāl al-Dīn al-Suyūt̤ ī. Edited by Fuad
Ali Mansur. 2 vols. Beirut: Dar al-Kutub al-ʿIlmiyah, 1998.
Nawādir al-Alfāz̤ of Sirāj al-Dīn ʿAlī Khān : Ārzū. Edited by Sayyid ʿAbdullah.
Karachi: Anjuman-i Taraqqi-yi Urdu, 1951.
Nikāt al-Shuʿarā of Mīr Muhammad : Taqī Mīr. Edited by ʿAbdulhaq. Karachi:
Anjuman-i Taraqqi-yi Urdu, 1979.
Nishtar-i ʿIshq of Āqā Husain

:
Qulī Khān ʿAshiqī ʿAz̤ īmābādī. Translated by ʿAta
Kakavi into Urdu. Patna: Patna Centre for Research in Arabic and Persian,
1968.
:
Nuh Sipihr of Amīr Khusrau. Edited by Mohammad Wahid Mirza. London: Oxford
University Press, 1950.
Qābūsnāmah of Kaikāwus. Edited by Amin Abdulmagid Badawi. Tehran: Ebne Sina,
1956.
Rahīm Granthāwalī (in Devanagari script) of ʿAbd al-Rahīm Khān-i : :
Khānān Rahīm.
Edited by Vidyaniwas Mishra and Govind Rajnish. New Delhi: Vani Prakashan,
2010.
Risālah-yi ʿAbd al-Wāsiʿ of ʿAbd al-Wāsiʿ Hansawī. Kanpur: Naval Kishore, 1862.
Risālah-yi Dilgushā in Kulliyāt-i ʿUbaid-i Zākānī. Edited by ʿAbbas Iqbal. Tehran:
Eghbal, 1955 [1334].
Risālah-yi Qaul-i Faisal : of Imām Bakhsh : Sahbāʾī.

Kanpur: Naval Kishore, n.d.
(c. 1860–80).
Ruqʿāt-i ʿĀlamgīrī of Aurangzeb. Translated by Jamshed H. Bilimoria as Ruqa’at-i
Alamgiri, or Letters of Aurungzebe. London: Luzac & Co., 1908.
Safīnah-i Kh : wushgo
: of Bindrāban Dās Kh : wushgo.
: Edited by S. Shah Md. Attaur
Rahman. Vol. 3. Patna: Institute of Post Graduate Studies and Research in Arabic
and Persian, 1959.
Sarw-i Āzād of Mīr Ġhulām ʿAlī Āzād Bilgrāmī. Edited by ʿAbdulhaq. Hyderabad:
Asifiyah Library, 1913.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

 287

: Qasāʾid-i
Sharh-i : ʿUrfī of Sirāj al-Dīn ʿAlī Khān : Ārzū. DP 1296A. Delhi Persian
Collection, British Library.
: Sikandarnāmah of Sirāj al-Dīn ʿAlī Khān
Sharh-i : Ārzū. Ms 145. Fort William College
Collection, National Archives of India, New Delhi.
: Sikandarnāmah of Sirāj al-Dīn ʿAlī Khān
Sharh-i : Ārzū. Mss 3985f and 3965f. Raza
Library, Rampur.
: Sundar Singār (anonymous). Ms India Office Islamic 2075. British Library.
Sharh-i
Shikūfah-Zār of Sirāj al-Dīn ʿAlī Khān : Ārzū. Edited by Zahra Ibrahim Aughali
Khiyavi. Tehran: Safir Ardehal, 2013 (1392 ).
Sirāj al-Luġhat of Sirāj al-Dīn ʿAlī Khān : Ārzū. Ms 109. Fort William College
Collection, National Archives of India, New Delhi.
Sirāj al-Luġhat of Sirāj al-Dīn ʿAlī Khān
: Ārzū. Ms India Office Islamic 1783. British
Library.
Sirāj-i Munīr of Sirāj al-Dīn ʿAlī Khān
: Ārzū. See Kārnāmah.
Sirāj-i Munīr of Sirāj al-Dīn ʿAlī Khān
: Ārzū. See Majmūʿah.
Sirāj-i Wahhāj of Sirāj al-Dīn ʿAlī Khān : Ārzū. Ms University 3 Farsi 119. Aligarh
Muslim University.
Sirāj-i Wahhāj of Sirāj al-Dīn ʿAlī Khān: Ārzū. See Majmūʿah.
:
Sukhandān-i :
Fārs of Muhammad Husain

Āzād. New Delhi: Qaumi Urdu Council,
2005.
Tanbīh al-Ġhāfilīn of Sirāj al-Dīn ʿAlī Khān : Ārzū. See Majmūʿah.
Tanbīh al-Ġhāfilīn of Sirāj al-Dīn ʿAlī Khān : Ārzū. Edited by Sayyid Muhammad
Akram. Lahore: Punjab University, 1981.
Tażkirah-yi Ātashkadah: Az Rū-yi Nuskhah-yi : Chāp-i Sangī-i Bambaʾī-i 1277 of Lut̤ f
ʿAlī Baig Āżar. Tehran: Rawzanah, 1998.
Tażkirah-yi Hazīn

:
of Muhammad ʿAlī Hazīn.

Edited by Muhammad Baqir Alfat.
Isfahan: Parvin, 1955.
Tażkirah-yi Hindī of Ġhulām Hamadānī Mus: hafī. : Lucknow: Uttar Pradesh Urdu
Academy, 1985.
Tażkirah-yi Husainī

of Mīr Husain

Dost Sanbhalī. Translated by Kishwar Jahan
Zaidi into Urdu. New Delhi: Anjuman-i Taraqqi-yi Urdu, 2008.
Tażkirah-yi Mardum-i Dīdah of ʿAbd al-Hakīm
Hākim.

Edited by Sayyid Abdullah.
Lahore: Oriental College, 1961.
:
Tażkirah-yi Rekhtah-Goyān of ʿAlī al-Husainī

Gardezī. Edited by Akbar Haidari
Kashmiri. Lucknow: Uttar Pradesh Urdu Academy, 1995.
Tażkirah-yi Riyāz
al-ʿĀrifīn of Āftāb Rāy Lakhnawī. Edited by Sayyid Husam al-Din
Rashidi. 2 vols. Islamabad: Markaz-i Tahqiqat-i Farsi-yi Iran wa Pakistan, 1977.
Tażkirah-yi Riyāz
al-Fus: ahā : of Ġhulām Hamadānī Mus: hafī. : Lucknow: Uttar
Pradesh Urdu Academy, 1985.
Tażkirah-yi Riyāz
al-Shuʿarāʾ of ʿAlīqulī Wālih Dāġhistānī. Edited by Mohsin Naji
Nasrabadi. Tehran: Asatir, 2005.
Tażkirah-yi Shuʿarā-yi Urdū of Mīr Hasan.
Lucknow: Uttar Pradesh Urdu Academy,
1985.
:
Tażkirat al-Muʿāsirīn :
of Muhammad ʿAlī Hazīn.

Edited by Masumah Salik. Tehran:
Nashr-i Sayah, 1996.
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

288 

“The Third Anniversary Discourse, on the Hindus” of William Jones. In Sir William
Jones: Selected Poetical and Prose Works, edited by Michael J. Franklin, 355–70.
Cardiff: University of Wales Press, 1995.
Tuhfat :
: al-Hind of Mīrzā Khān. Ms 67. Centre of Advanced Study in History, Aligarh
Muslim University.
: al-Hind of Mīrzā Khān.
Tuhfat : Ms Elliott 383. Bodleian Library, Oxford University.
:
Tuhfat :
al-Hind of Mīrzā Khān. Edited by Nur al-Hasan Ansari. Tehran: Bunyad-i
Farhang-i Iran, 1975.
:
Tuhfat :
al-Hind (khātimah :
only) of Mīrzā Khān. Edited by Nur al-Hasan Ansari. 2
vols. Delhi: Department of Persian, Delhi University, 1983.
A Vocabulary of the Names of the Various Part of the Human Body and of Medical
and Technical Terms in English, Arabic, Persian, Hindee and Sanscrit by Joseph
Breton. Calcutta: Government Lithographic Press, 1825.
Yādgār-i Ġhālib of Alt̤ āf Husain

Hālī.

Lucknow: Uttar Pradesh Urdu Academy, 1986.
Zat̤ alnāmah of Mīr Jaʿfar Zat̤ allī. Edited by Rashid Hasan Khan. New Delhi:
Anjuman-i Taraqqi-yi Urdu, 2003.
Zawāʾid al-Fawāʾid attributed to Sirāj al-Dīn ʿAlī Khān
: Ārzū. Ms 2520f. Raza Library,
Rampur.
Żikr-i Mīr of Mīr Muhammad
: Taqī Mīr. Edited by Sharif Husain Qasemi. New Delhi:
National Council for the Promotion of Urdu Language, 2011.

Secondary Sources

ʿAbdullah, Sayyid. 1965. Mabāhis̄.

: Lahore: Majlis-i Taraqqi-yi Urdu.
ʿAbdullah, Sayyid. 1967. Adabiyāt-i Fārsī meṁ Hindūʾoṁ kā Hi
s: sah.
: Lahore: Majlis-i
Taraqqi-yi Adab.
ʿAbdullah, Sayyid. 1977. Fārsī Zabān wa Adab. Lahore: Majlis-i Taraqqi-yi Adab.
Abidi, S. A. R. 1982. India’s Rich and Valuable Contribution to Persian Literature: In
the Light of Some Recent Discoveries. Khuda Bakhsh Annual Lecture Series 5.
Patna: Khuda Bakhsh Oriental Public Library.
Adams, J. N. 2003. Bilingualism and the Latin Language. Cambridge: Cambridge
University Press.
Adams, J. N. 2007. The Regional Diversification of Latin 200 – 600. Cambridge:
Cambridge University Press.
Adams, J. N. 2013. Social Variation and the Latin Language. Cambridge: Cambridge
University Press.
Ahmad, Aijaz. 1987. “Jameson’s Rhetoric of Otherness and the ‘National Allegory.’ ”
Social Text 17 (Autumn): 3–25. https://doi.org/10.2307/466475.
:
Ahmad, Nazir, ed. 1972. Dīwān-i Sayyid Sirāj al-Dīn Khurāsānī Maʿrūf bah Sirājī.
Aligarh: Aligarh Muslim University Press.
Ahmed, Asad Q., and Margaret Larkin. 2013. “The Hāshiya

and Islamic Intellectual
History.” Oriens 41: 213–16. http://doi.org/10.1163/18778372-13413401.
Aitken, Molly Emma. 2010. The Intelligence of Tradition in Rajput Court Painting.
New Haven, CT: Yale University Press.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

 289

Akram, Sayyid Farid. 2009 [1388]. Surūdah-hā wa Niwishtah-hā-yi Munīr Lāhorī.

Tehran: Bunyad-i Mauqufat-i Duktur Mahmud Afshar.
Akram, Sayyid Muhammad. 1974. “Editor’s Preface.” In Dād-i Suk: han, xiii–xlii.
Rawalpindi: Iran Pakistan Institute of Persian Studies.
Akram, Sayyid Muhammad. 1977. “Editor’s Preface.” In Kārnāmah wa Sirāj-i Munīr,
xiii–xl. Islamabad: Centre for Research in Persian of Iran and Pakistan.
Akram, Sayyid Muhammad. 1981. “Editor’s Preface.” In Tanbīh al-Ġhāfilīn, xiii–lxv.
Lahore: Punjab University Press.
Alam, Muzaffar. 1986. The Crisis of Empire in Mughal North India: Awadh and the
Punjab, 1707–48. New Delhi: Oxford University Press.
Alam, Muzaffar. 1998. “The Pursuit of Persian: Language in Mughal Politics.”
Modern Asian Studies 32 (2): 317–49. https://doi.org/10.1017/S0026749X98002947.
Alam, Muzaffar. 2003. “The Culture and Politics of Persian in Pre-Colonial
Hindustan.” In Literary Cultures in History: Reconstructions from South Asia,
edited by Sheldon Pollock, 131–98. Berkeley, CA: University of California.
Alam, Muzaffar. 2004. The Languages of Political Islam: India, 1200–1800. Chicago,
IL: University of Chicago.
Alam, Muzaffar, and Sanjay Subrahmanyam. 1996. “Discovering the Familiar: Notes
on the Travel-Account of Anand Ram Mukhlis, 1745.” South Asia Research 16 (2):
131–54. https://doi.org/10.1177%2F026272809601600202.
Alam, Muzaffar, and Sanjay Subrahmanyam. 1999. “Introduction.” In The Mughal
State, 1526–1750, 1–71. New Delhi: Oxford University Press.
Alam, Muzaffar, and Sanjay Subrahmanyam. 2004. “The Making of a Munshi.”
Comparative Studies of South Asia, Africa, and the Middle East 24 (2): 61–72.
https://muse.jhu.edu/article/181648.
Alam, Muzaffar, and Sanjay Subrahmanyam. 2006. “When Hell Is Other People:
A Safavid View of Seventeenth-Century Mughal India.” In History and
Historiography of Post-Mongol Central Asia and the Middle East: Studies in
Honor of John E. Woods, edited by Judith Pfeiffer and Sholeh A. Quinn in
Collaboration with Ernest Tucker. Wiesbaden: Harrassowitz.
Alam, Muzaffar, and Sanjay Subrahmanyam. 2012. “Faizi’s Nal-Daman and Its Long
Afterlife.” In Writing the Mughal World: Studies on Culture and Politics. New
York: Columbia University Press.
Alavi, Seema, ed. 2002. The Eighteenth Century in India. New Delhi: Oxford
University Press.
Aldridge, A. Owen. 1973. “Ancients and Moderns in the Eighteenth Century.” In
Dictionary of the History of Ideas, edited by Philip P. Wiener. New York: Charles
Scribner.
Ali, Ashfaque. 2002. “Hindi Commercial and Craft Terms in Persian Lexicon, the
Bahar-i- ʿAjam.” Proceedings of the Indian History Congress 63: 378–82. www.
jstor.org/stable/44158105.
Ali, M. Athar. 1997. The Mughal Nobility under Aurangzeb. Revised ed. New Delhi:
Oxford University Press.
Ali, M. Athar. 1999. “Pursuing an Elusive Seeker of Universal Truth: The Identity
and Environment of the Author of Dabistan-i Mazahib.” Journal of the Royal
Asiatic Society 9 (3): 365–73. https://doi.org/10.1017/S1356186300011512.
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

290 

Allen, W. S. 1948. “Ancient Ideas on the Origin and Development of Language.”

Transactions of the Philological Society of London 47 (1): 35–60.
Anderson, Benedict. 1983. Imagined Communities: Reflections on the Origin and
Spread of Nationalism. London: Verso.
Andrews, Walter G., and Mehmet Kalpaklı. 2004. Age of Beloveds: Love and the
Beloved in Early Modern Ottoman and European Culture and Society. Durham,
NC: Duke University Press.
Ansari, Iqbal. 1940. “Khān-i
: Ārzū kī Mas̄nawī ‘Shor-i ʿIshq’.” Nigār December: 41–9.
Appiah, Kwame Anthony. 2006. Cosmopolitanism: Ethics in a World. London: Allen
Lane.
Arlotto, Anthony T. 1969. “Jones and Coeurdoux: Correction to a Footnote.” Journal
of the American Oriental Society 89 (2): 416–17.
Artan, Tülay. 2010. “Eighteenth-Century Ottoman Princesses as Collectors: Chinese
and European Porcelains in the Topkapı Palace Museum.” Ars Orientalis 39:
113–47. www.jstor.com/stable/23075925.
Asbaghi, Asya. 1987. Die semantische Entwicklung arabischer Wörter im Persischen.
Stuttgart: Franz Steiner.
Ashraf, Muhammad. 1966–88. A Concise Descriptive Catalogue of the Persian
Manuscripts in the Salar Jung Museum and Library. Hyderabad: Salar Jung
Museum and Library.
Auerbach, Erich. 1949. “Vico and Aesthetic Historism.” Journal of Aesthetics and Art
Criticism 8 (2): 110–18. http://doi.org/10.2307/426593.
Auerbach, Erich. 1993. Literary Language and Its Public in Late Latin Antiquity and
in the Middle Ages. Translated by Ralph Manheim. Princeton, NJ: Princeton
University Press.
Auerbach, Erich. 2003. Mimesis: The Representation of Reality in Western Literature.
Translated by Willard R. Trask. Princeton, NJ: Princeton University Press.
Original ed., 1953.
Auroux, Sylvain. 1989. La naissance des métalangages en Orient et en Occident.
Histoire des idées linguistiques 1. Liège: P. Mardaga.
Auroux, Sylvain. 1990. “Representation and the Place of Linguistic Change before
Comparative Grammar.” In Leibniz, Humboldt, and the Origins of Comparativism,
edited by Tullio de Mauro and Lia Formigari. Amsterdam: John Benjamins.
Azim, Abdul. 1970. “Khan-i Arzu’s Observations on the Relationship of Sanskrit and
Persian.” Zeitschrift der Deutschen Morgenländischen Gesellschaft 119 (2): 261–9.
Baalbaki, Ramzi. 1983. “The Relation between nahw : and balāġa: A Comparative
Study of the Methods of Sībawayhi and Ğurğānī.” Zeitschrift für arabische
Linguistik 11: 7–23.
Baalbaki, Ramzi. 2019. “8. Arabic to c. 1800.” In The Cambridge World History of
Lexicography, edited by John Considine, 159–81. Cambridge: Cambridge University
Press. http://doi.org/10.1017/9781316827437.009.
Baevskii, Solomon I. 2007. Early Persian Lexicography: Farhangs of the Eleventh to
the Fifteenth Centuries. Translated by N. Killian. Revised by John R. Perry.
Folkestone: Global Oriental.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

 291

:
Bahar, Muhammad Taqi. 1942. Sabkshināsī, yā Tārīkh-i Tatawwur-i Nas̄r-i Fārsī.
Tehran: Chapkhanah-yi Khudkar.
Bahar, Muhammad Taqi. 1970. “Sāʾib
wa Shīwah-yi Ū.” Yaghma 23 (5): 244–5.
Bahl, Christopher D. 2018. “Histories of Circulation: Sharing Arabic Manuscripts
across the Western Indian Ocean, 1400–1700.” Ph.D. thesis, Department of
History, SOAS.
Baig, Farhatullah. 1979. The Last Mushaʿirah of Delhi. Translated by Akhtar Qamber.
New Delhi: Orient Longman.
Bailey, T. Grahame. 1930. “Early Urdu Conversation.” Bulletin of the School of
Oriental Studies, University of London 6 (1): 205–8. https://doi.org/10.1017/
S0041977X00091035.
Bailey, T. Grahame. 1932. A History of Urdu Literature. London: Oxford University
Press.
Bakhtin, M. M. 2002. The Dialogic Imagination: Four Essays. Translated by Caryl
Emerson and Michael Holquist. Edited by Michael Holquist. Austin, TX:
University of Texas. Original ed., 1981.
Balhan, Jean-Marc. 2001. La révélation du Coran selon Al-Suyūt:ī. Rome: Pontificio
Istituto di Studi Arabi e d’Islamistica.
Ballon, Hilary. 1989. “Constructions of the Bourbon State: Classical Architecture in
Seventeenth-Century France.” In Culture Differentiation and Cultural Identity in
the Visual Arts, edited by Susan J. Barnes and Walter S. Melion, 135–48.
Washington, DC: National Gallery of Art.
Bangha, Imre, ed. 2000. Thomas Duer Broughton’s Selections from the Popular Poetry
of the Hindoos, 1814. Delhi: Rainbow.
Bangha, Imre. 2010. “Rekhta: Poetry in Mixed Language: The Emergence of Khari
Boli Literature in North India.” In Before the Divide: Hindi and Urdu Literary
Culture, edited by Francesca Orsini, 21–83. New Delhi: Orient Blackswan.
Barchilon, Jacques, and Peter Flinders. 1981. Charles Perrault. Boston, MA: Twayne.
Barzegar, Karim Najafi. 2001. Mughal-Iranian Relations during Sixteenth Century.
Delhi: Indian Bibliographies Bureau.
Bayly, C. A. 1988. Indian Society and the Making of the British Empire. Cambridge:
Cambridge University Press.
Bayly, C. A. 1993. Rulers, Townsmen and Bazaars: North Indian Society in the Age of
British Expansion, 1770–1870. Expanded Indian ed. Delhi: Oxford University
Press. Original ed., Cambridge University Press, 1983.
Bayly, C. A. 1996. Empire and Information: Intelligence Gathering and Social
Communication in India, 1780–1870. Cambridge: Cambridge University Press.
Beeston, A. F. L. 1954. Catalogue of the Persian, Turkish, Hindustani and Pushtu
Manuscripts in the Bodleian Library, Part III: Additional Persian Manuscripts.
Oxford: Clarendon.
Behl, Aditya. 2005. “Poet of the Bazaars: Naz̤ īr Akbarābādī, 1735–1830.” In A
Wilderness of Possibilities: Urdu Studies in Transnational Perspective, edited by
Kathryn Hansen and David Lelyveld, 192–222. New Delhi: Oxford University
Press.
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

292 

Behl, Aditya. 2011. “Pages from the Book of Religions: Comparing Self and Other in
Mughal India.” In Forms of Knowledge in Early Modern Asia, edited by Sheldon
Pollock, 210–39. Durham, NC: Duke University Press.
Behl, Aditya. 2012. Love’s Subtle Magic: An Indian Islamic Literary Tradition,
1379–1545. Edited by Wendy Doniger. New York: Oxford University Press.
Bell, William Y. 1924. The Mutawakkili of As-Suyuti: A Translation of the Arabic
Text with Introduction, Notes, and Indices. Cairo: Nile Mission Press.
Berlin, Isaiah. 1976. Vico and Herder: Two Studies in the History of Ideas. New York:
Viking Press.
Beveridge, H., and Baini Prashad. 1979. The Maʾthir-ul-Umarā. English Translation
with Notes and Preface. Reprint of 2nd ed. Patna: Janaki Prakashan.
Bhatia, Tej K. 1987. A History of the Hindi Grammatical Tradition. Leiden: E. J. Brill.
Bietenholz, Peter G. 1994. Historia and Fabula: Myths and Legends in Historical
Thought from Antiquity to the Modern Age. Leiden: E. J. Brill.
Bloch, Marc. 1992. The Historian’s Craft. Manchester: Manchester University Press.
Blochmann, Henry. 1868. “Contributions to Persian Lexicography.” Journal of the
Royal Asiatic Society of Bengal 37 (1): 1–72.
Bonebakker, S. A. 1970. “Aspects of the History of Literary Rhetoric and Poetics in
Arabic Literature.” Viator 1: 75–95. https://doi.org/10.1484/J.VIATOR.2.301709.
Bonebakker, S. A. 1990. “Adab and the Concept of Belles Lettres.” In Abbasid Belles
Lettres, edited by Julia Ashtiany, T. M. Johnstone, J. D. Latham, and R. B. Serjeant,
16–30. Cambridge: Cambridge University Press.
Borah, M. J. 1934. “The Nature of the Persian Language Written and Spoken in India
during the 13th and 14th Centuries.” Bulletin of the School of Oriental Studies,
University of London 7 (2): 325–7. https://doi.org/10.1017/S0041977X00085244.
Bosworth, C. E. 2011. “Sanad.” In Encyclopaedia of Islam, edited by P. Bearman, Th.
Bianquis, C. E. Bosworth, E. van Donzel, and W. P. Heinrichs. 2nd ed., Vol. 12.
Leiden: Brill.
Brennan, Timothy. 2006. Wars of Position: The Cultural Politics of Right and Left.
New York: Columbia University Press.
Briggs, Charles F. 2003. “Teaching Philosophy at School and Court: Vulgarization
and Translation.” In The Vulgar Tongue: Medieval and Postmedieval
Vernacularity, edited by Fiona Somerset and Nicholas Watson, 99–111.
University Park, PA: Pennsylvania State University Press.
Bronner, Yigal, and David Shulman. 2006. “ ‘A Cloud Turned Goose’: Sanskrit in the
Vernacular Millennium.” Indian Economic Social History Review 46 (1): 1–30.
https://doi.org/10.1177%2F001946460504300101.
Browne, Edward G. 1896. A Catalogue of the Persian Manuscripts in the Library of
Cambridge University. Cambridge: Cambridge University Press.
Browne, Edward G. 1929. A Literary History of Persia: From the Earliest Times to
Firdawsí. Vol. 1. Cambridge: Cambridge University Press. Original ed., 1902.
Browne, Edward G. 1956. A Literary History of Persia: The Tartar Dominion
(1265–1502). Vol. 3. Cambridge: Cambridge University Press.
Browne, Edward G. 1959. A Literary History of Persia: Modern Times (1500–1924).
Vol. 4. Cambridge: Cambridge University Press. Original ed., 1924.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

 293

Burke, Peter. 1970. The Renaissance Sense of the Past. New York: St. Martin’s Press.
Burke, Peter. 1985. Vico. Oxford: Oxford University Press.
Burke, Peter. 1997. The Renaissance. 2nd ed. Basingstoke: Macmillan Press.
Burke, Peter. 2004. Languages and Communities in Early Modern Europe.
Cambridge: Cambridge University Press.
Burke, Peter. 2006. “The Hybridization of Languages in Early Modern Europe.”
European Review 14 (1): 105–10. https://doi.org/10.1017/S1062798706000093.
Bury, J. B. 1920. The Idea of Progress: An Inquiry into Its Origin and Growth. London:
Macmillan.
Busch, Allison. 2003. “The Courtly Vernacular: The Transformation of Brajbhās: ā
Literary Culture (1590–1690).” Ph.D. thesis, Department of South Asian
Languages and Civilizations, University of Chicago.
Busch, Allison. 2010. “Hidden in Plain View: Brajbhasha Poets at the Mughal Court.”
Modern Asian Studies 44 (2): 267–309. https://doi.org/10.1017/S0026749X09990205.
Busch, Allison. 2011. Poetry of Kings: The Classical Hindi Tradition of Mughal India.
New York: Oxford University Press.
Butcher, S. H., trans. 1902. The Poetics of Aristotle. 3rd ed. London: Macmillan.
Campbell, Lyle, and Mauricio J. Mixco. 2007. A Glossary of Historical Linguistics. Salt
Lake City, UT: University of Utah Press.
Campbell, Lyle, and William J. Poser. 2008. Language Classification: History and
Method. Cambridge: Cambridge University Press.
Cannon, Garland, ed. 1970. The Letters of Sir William Jones. Oxford: Oxford
University Press.
Cannon, Garland. 1990. The Life and Mind of Oriental Jones: Sir William Jones, the
Father of Modern Linguistics. Cambridge: Cambridge University Press.
Cantarino, Vicente. 1975. Arabic Poetics in the Golden Age: Selection of Texts
Accompanied by a Preliminary Study. Leiden: E. J. Brill.
Carter, Michael G. 1973. “An Arab Grammarian of the Eighth Century A.D.:
A Contribution to the History of Linguistics.” Journal of the American Oriental
Society 93 (2): 146–57. www.jstor.com/stable/598889.
Cereti, Carlo G., and Mauro Maggi, eds. 2005. Middle Iranian Lexicography: Proceedings
of the Conference held in Rome, 9–11 April 2011. Rome: Istituto Italiano per l’Africa e
l’Oriente.
Chakrabarty, Dipesh. 2000. Provincializing Europe: Postcolonial Thought and
Historical Difference. Princeton, NJ: Princeton University Press.
Chakrabarty, Dipesh. 2011. “The Muddle of Modernity.” American Historical Review
116 (3): 663–75. http://doi.org/10.1086/ahr.116.3.663.
Chalisova, N. 2004. “Ihām.” Encyclopædia Iranica. www.iranicaonline.org/articles/
iham.
Chalisova, N. 2009. “Persian Rhetoric: Elm-e Badi’ and Elm-e Bayân.” In General
Introduction to Persian Literature, ed. J. T. P. De Bruijn, 139–71. Vol. 1 of
A History of Persian Literature. London: I. B. Tauris.
Chandra, Satish. 2002. Parties and Politics at the Mughal Court, 1707–1740. 4th ed.
New Delhi: Oxford University Press.
Chari, V. K. 1993. Sanskrit Criticism. Delhi: Motilal Banarsidass.
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

294 

Chartier, Roger. 1994. The Order of Books: Readers, Authors, and Libraries in Europe
between the Fourteenth and Eighteenth Centuries. Translated by Lydia G. Cochrane.
Stanford, CA: Stanford University Press.
Chatterjee, Partha. 2012. The Black Hole of Empire: History of a Global Practice of
Power. Princeton, NJ: Princeton University Press.
Chatterji, Suniti Kumar. 1942. Indo-Aryan and Hindi: Eight Lectures. Ahmedabad:
Gujarat Vernacular Society.
Clinton, Jerome W. 1989. “Shams-i Qays on the Nature of Poetry.” Edebiyat N.S.
1 (2): 101–27.
Clinton, Jerome W. 1998. “A Sketch of Translation and the Formation of New
Persian Literature.” In Iran and Iranian Studies: Essays in Honor of Iraj Afshar,
edited by Kambiz Eslami. Princeton, NJ: Zagros.
Coetzee, J. M. 2001. The Humanities in Africa. Munich: Carl Friedrich von Siemens
Stiftung.
Coleman, Janet. 1992. Ancient and Medieval Memories: Studies in the Reconstruction
of the Past. Cambridge: Cambridge University Press.
Cooper, Robert L. 1989. Language Planning and Social Change. Cambridge:
Cambridge University Press.
Czapkiewicz, Andrzej. 1989. The Views of the Medieval Arab Philologists on
Language and Its Origin in the Light of As-Suyuti’s Al-Muzhir. Cracow:
Jagiellonian University Press.
Czapkiewicz, Andrzej. 1990. “The Relation between ‘signifiant’ and ‘signifié’
According to the Mediaeval Arab Philologists.” In Orientalistische Philologie und
Arabische Linguistik, edited by Wolfgang Reuschel, 90–4. Berlin: Akademie-
Verlag.
Dabashi, Hamid. 2012. The World of Persian Literary Humanism. Cambridge, MA:
Harvard University Press.
Dabirsiaqi, Mohammad. 1989. “Borhān-e Qāteʿ.” : Encyclopædia Iranica. www.
iranicaonline.org/articles/borhan-e-qate.
Dadvar, Abolghasem. 1999. Iranians in Mughal Politics and Society 1606–1658. New
Delhi: Gyan Publishing.
Dale, Stephen Frederic. 2003. “A Safavid Poet in the Heart of Darkness: The Indian
Poems of Ashraf Mazandarani.” Iranian Studies 36 (2): 197–212. https://doi.org/
10.1080/00210860305242.
Dalmia, Vasudha. 1997. The Nationalization of Hindu Traditions: Bhāratendu
Hariśchandra and Nineteenth-Century Banaras. New Delhi: Oxford University Press.
D’Amico, John F. 1988. Theory and Practice in Renaissance Textual Criticism: Beatus
Rhenanus between Conjecture and History. Berkeley, CA: University of California.
Darling, Linda T. 2012. “Ottoman Turkish: Written Language and Scribal Practice,
13th to 20th Centuries.” In Literacy in the Persianate World: Writing and the
Social Order, edited by Brian Spooner and William L. Hanaway, 171–95.
Philadelphia, PA: University of Pennsylvania.
Das, Rahul Peter. 2014. “Defining Hindi: An Introductory Overview.” In Defining the
Indefinable: Delimiting Hindi, edited by Agnieszka Kuczkiewicz-Fraś, 11–64.
Frankfurt am Main: Peter Lang.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

 295

Das, Sisir Kumar. 1991. A History of Indian Literature: 1800–1910, Western Impact:
Indian Response. New Delhi: Sahiyta Academy.
Datla, Kavita. 2013. The Language of Secular Islam: Urdu Nationalism and Colonial
India. Honolulu, HA: University of Hawaii Press.
De Bruijn, J. T. P. 1988. “Bayān (1).” Encyclopædia Iranica. www.iranicaonline.org/
articles/bayan-a-noun-common-to-arabic-and-persian-meaning-statement-
exposition-explanation.
De Bruijn, J. T. P. 2009. “Classical Persian Literature as a Tradition.” In General
Introduction to Persian Literature, edited by J. T. P. De Bruijn. A History of
Persian Literature series, 1:1–42. London: I. B. Tauris.
Degraff, Michel. 2005. “Linguists’ Most Dangerous Myth: The Fallacy of Creole
Exceptionalism.” Language in Society 34 (4): 533–91. https://doi.org/10.1017/
S0047404505050207.
DeJean, Joan E. 1997. Ancients against Moderns: Culture Wars and the Making of a
Fin de Siècle. Chicago, IL: University of Chicago.
Dewhurst, R. P. 1915. “Persian and Arabic Words in the Satsai of Bihari Lal.” Journal
of the Royal Asiatic Society of Great Britain and Ireland January: 122–7.
Dhavan, Purnima, and Heidi Pauwels. 2015. “Controversies Surrounding the
Reception of Valī ‘Dakhanī’ (1665?–1707?) in Early Tażkirahs of Urdu Poets.”
Journal of the Royal Asiatic Society 25 (4): 625–46. www.doi.org/10.1017/
S1356186315000255.
Digby, Simon. 1995. “Illustrated Muslim Books of Omens from Gujarat or
Rajasthan.” In Indian Art and Connoisseurship: Essays in Honour of Douglas
Barrett, edited by John Guy. Middletown, NJ: Grantha Corporation.
Dirayati, Mustafa. 2010 [1389]. Fihristwārah-i Dastniwisht-hā-yi Īrān (Dinā). 12 vols.
Tehran: Library of the Iranian Parliament.
Dudney, Arthur. 2008. “Colonial Knowledge and the Greco-Roman Classics:
Resituating the Legacy of Sir William Jones in a Humanist Context.” M.A. thesis,
South Asian Studies, Columbia University.
Dudney, Arthur. 2010a. “Braj Reinvented: Some British Approaches to Hindi
Dialects.” In Problematizing Language Studies: Essays in Honour of Rama Kant
Agnihotri, edited by S. Imtiaz Hasnain and Shreesh Chaudhary, 35–47. Delhi: Aakar.
Dudney, Arthur. 2010b. “The Wonders of Words, or the Role of Khān-i : Ārzū’s
Navādir al-Alfāz̤ in the Development of Urdu.” In An Informal Festschrift in Honor
of Shamsur Rahman Faruqi, edited by Satyanarayana Hegde. www.columbia.edu/
itc/mealac/pritchett/00urduhindilinks/srffest.
Dudney, Arthur. 2012. “Fashioning Everyday Speech into Literary Language in
Eighteenth-Century Delhi.” Paper presented at Association of Asian Studies
conference, Toronto, 18 March.
Dudney, Arthur. 2016. “Sabk-e Hendi and the Crisis of Authority in Eighteenth-
Century Indo-Persian Poetics.” Journal of Persianate Studies 9 (1): 60–82. https://
doi.org/10.1163/18747167-12341294.
Dudney, Arthur. 2017a. “Metaphorical Language as a Battleground for Tradition and
Newness in Late Mughal Persian.” International Journal of Persian Literature 2:
138–60. www.jstor.com/stable/10.5325/intejperslite.2.1.0138.
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

296 

Dudney, Arthur. 2017b. “Going Native: Iranian Émigré Poets and Indic Languages.”
Comparative Studies of South Asia, Africa, and the Middle East 37 (3): 531–48.
Dudney, Arthur. 2017c. Review of Negotiating Languages by Walter Hakala.
Journal of the Royal Asiatic Society 27 (1): 325–26. https://doi.org/10.1017/
s1356186316000596.
Dudney, Arthur. 2018a. “Urdu as Persian: Some Eighteenth-Century Evidence on
Vernacular Poetry as Language Planning.” In Text and Tradition in Early Modern
North India, edited by Tyler Williams, Anshu Malhotra and John Stratton Hawley,
40–57. New Delhi: Oxford University Press. http://doi.org/10.1093/oso/
9780199478866.003.0002.
Dudney, Arthur. 2018b. “Literary Decadence and Imagining the Late Mughal City.”
Journal of Early Modern Cultural Studies 18 (3): 187–211. http://doi.org/10.1353/
jem.2018.0028.
Dudney, Arthur. 2019a. “Persian-Language Education in Mughal India from Qas: bah
to Capital.” In Early Modern India: Literatures and Images, Texts and Languages,
edited by Maya Burger and Nadia Cattoni, 79–91. Heidelberg: CrossAsia. http://
doi.org/10.11588/xabooks.387.
Dudney, Arthur. 2019b. “17. Turkish and Persian from c. 1700: Persian.” In The
Cambridge World History of Lexicography, edited by John Considine, 377–87.
Cambridge: Cambridge University Press. http://doi.org/10.1017/9781316827437.
Dupriez, Bernard. 1991. A Dictionary of Literary Devices: Gradus, A–Z. Translated by
Albert W. Halsall. Toronto: University of Toronto.
Eaton, Richard M. 1978. Sufis of Bijapur, 1300–1700: Social Roles of Sufis in Medieval
India. Princeton, NJ: Princeton University Press.
Eaton, Richard M. 2014. “The Rise of Written Vernaculars: The Deccan 1450–1650.”
In After Timur Left: Culture and Circulation in Fifteenth-Century North India,
edited by Francesca Orsini and Samira Sheikh, 111–29. New Delhi: Oxford
University Press.
Eco, Umberto. 1990. On Literature. Translated by Martin McLaughlin. New York:
Harcourt.
Eisenstadt, Shmuel N. 2000. “Multiple Modernities.” Daedalus 129 (1): 1–29.
Eisenstadt, Shmuel N., and Wolfgang Schluchter. 1998. “Paths to Early Modernities.”
Daedalus 127 (3): 1–18.
Elliot, H. M., and John Dowson, eds. 1877. The History of India as Told by Its Own
Historians: The Muhammadan Period. Vol. 8. London: Trübner.
Emmerick, R. E. 1974. “The Beginnings of Iranian Comparative Philology.” In Neue
Methodologie in der Iranistik, edited by Richard N. Frye, 49–56. Wiesbaden: Otto
Harrassowitz.
Ernst, Carl W. 1995. “India as a Sacred Islamic Land.” In Religions of India in Practice,
edited by Donald S. Lopez, Jr., 556–63. Princeton: Princeton University Press.
Ernst, Carl W. 1996. “Sufism and Yoga According to Muhammad Ghawth.” Sufi 29:
9–13.
Ernst, Carl W., and Tony K. Stewart. 2003. “Syncretism.” In South Asian Folklore: An
Encyclopedia, edited by Margaret A. Mills, Peter J. Claus, and Sarah Diamond,
586–88. London: Routledge.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

 297

Ethé, Hermann. 1903. Catalogue of Persian Manuscripts in the Library of the India
Office. Oxford: Printed for the India Office.
Evans, R. J. W. 1998. The Language of History and the History of Language. Oxford:
Clarendon Press.
Evans, Ruth, Andrew Taylor, Nicholas Watson, and Jocelyn Wogan-Browne. 1999.
The Idea of the Vernacular: An Anthology of Middle English Literary Theory,
1280–1520. Exeter: University of Exeter.
Farhad, Massumeh. 2001. “ ‘Searching for the New’: Later Safavid Painting and the
‘Suz u Gawdaz’ (‘Burning and Melting’) by Nau’i Khabushani.” Journal of the
Walters Art Museum 59: 115–30.
Farrell, Joseph. 2001. Latin Language and Latin Culture from Ancient to Modern
Times. Cambridge: Cambridge University Press.
Faruqi, Shamsur Rahman. 1990a. Shiʿr-i Shor Angez. New Delhi: Taraqqi-yi Urdu
Bureau.
Faruqi, Shamsur Rahman. 1990b. “Some Problems of Urdu Lexicography.” Annual
of Urdu Studies 7: 21–30.
Faruqi, Shamsur Rahman. 1996. Urdū Ġhazal ke Aham Mor: : Īhām, Riʿāyat,
Munāsabat. New Delhi: Ghalib Academy.
Faruqi, Shamsur Rahman. 1998. “Unprivileged Power: The Strange Case of Persian
(and Urdu) in Nineteenth Century India.” Annual of Urdu Studies 13: 3–30.
Faruqi, Shamsur Rahman. 2001. Early Urdu Literary Culture and History. New Delhi:
Oxford University Press.
Faruqi, Shamsur Rahman. 2004a. “A Long History of Urdu Literary Culture, Part I:
Naming and Placing a Culture.” In Literary Cultures in History: Reconstructions
from South Asia, edited by Sheldon Pollock, 805–63. Berkeley, CA: University of
California.
Faruqi, Shamsur Rahman. 2004b. “A Stranger in the City: The Poetics of Sabk-i
Hindi.” Annual of Urdu Studies 19: 1–93.
Faruqi, Shamsur Rahman. 2005. “The Poet in the Poem or, Veiling the Utterance.” In
A Wilderness of Possibilities: Urdu Studies in Transnational Perspective, edited by
Kathryn Hansen and David Lelyveld, 173–91. New Delhi: Oxford University
Press.
Faruqi, Shamsur Rahman. 2008. “A Modest Plea: Please, Could We Have a Proper
History of Urdu Literature?” Unpublished essay. www.columbia.edu/itc/mealac/
pritchett/00fwp/srf/srf_lithistneeds_2008.pdf.
Faruqui, Munis D. 2009. “At Empire’s End: The Nizam, Hyderabad and Eighteenth-
Century India.” Modern Asian Studies 43 (1): 5–43.
Ferguson, Charles A. 1959. “Diglossia.” Word 15: 325–40.
Fishman, Joshua A. 1971. Sociolinguistics: A Brief Introduction. Rowley, MA:
Newbury House.
Floor, Willem, and Hasan Javadi. 2013. “The Role of Azerbaijani Turkish in Safavid
Iran.” Iranian Studies 46 (4): 569–581. https://doi.org/10.1080/00210862.2013.
784516.
Fotoohi, Mahmoud. 2011. “Taḏkera-ye Nasrābādi.”
: Encyclopædia Iranica, www.
iranicaonline.org/articles/tadkera-ye-nasrabadi.
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

298 

Foucault, Michel. 1972. The Archaeology of Knowledge. Translated by A. M. Sheridan

Smith. New York: Pantheon Books.
Foucault, Michel. 1994. The Order of Things: An Archaeology of the Human Sciences.
New York: Vintage.
Fragner, Bert G. 1999. Die “Persophonie”: Regionalität, Identität und Sprachkontakt
in der Geschichte Asiens. Berlin: Das Arabische Buch.
Frye, Northrup. 2000. Anatomy of Criticism: Four Essays. Princeton, NJ: Princeton
University Press. Original ed., 1957.
Fumaroli, Marc. 1984. “L’apologétique de la langue française classique.” Rhetorica
2 (2): 139–60.
Gabbay, Alyssa. 2010. Islamic Tolerance: Amīr Khusraw and Pluralism. New York:
Routledge.
Gadamer, Hans-Georg. 2006. Truth and Method. Translated by Joel Weinsheimer
and Donald G. Marshall. 2nd ed. New York: Continuum.
Galbraith, John Kenneth. 1978. The Affluent Society. 3rd ed. New York: Penguin.
Gallego, María Angeles. 2003. “The Languages of Medieval Iberia and Their
Religious Dimension.” Medieval Encounters 9 (1): 107–39.
Ganeri, Jonardon. 2011. The Lost Age of Reason: Philosophy in Early Modern India
1450–1700. Oxford: Oxford University Press.
Garcin de Tassy, Joseph. 1873. Rhétorique et prosodie des langues de l’Orient musul-
man. 2nd ed. Paris: Maisonneuve.
Gaunt, Simon. 2009. “Review Essay: Can the Middle Ages Be Postcolonial?”
Comparative Literature 61 (2): 160–76. http://doi.org/10.1215/00104124-2009-
004.
Gaunt, Simon. 2013a. “Linguistic Difference, the Philology of Romance, and the
Romance of Philology.” In A Sea of Languages: Rethinking the Arabic Role in
Medieval Literary History, edited by Suzanne Conklin Akbari and Karla Mallette,
43–61. Toronto: University of Toronto Press.
Gaunt, Simon. 2013b. Marco Polo’s Le Devisement du Monde: Narrative Voice,
Language and Diversity. Cambridge: D. S. Brewer.
: In Encyclopaedia of Islam, edited by: P. Bearman, Th.
Geoffroy, E. 2002. al-Suyūtī.
Bianquis, C. E. Bosworth, E. van Donzel, and W. P. Heinrichs. 2nd ed. http://dx.
doi.org/10.1163/1573-3912_islam_COM_1130.
Ghani, ʿAbdul. 1960. Life and Works of Abdul Qadir Bedil. Lahore: Publishers United.
Ghani, Muhammad ʿAbdul. 1930. A History of Persian Language and Literature at
the Mughal Court. Vol. 3 (Babur to Akbar). Allahabad: The Indian Press.
Ghani, Muhammad ʿAbdul. 1994. Pre-Mughal Persian in Hindustan. 2 vols.
Gurgaon: Vintage Books. Original ed., 1941.
Gilliot, Claude. 1990. Exégèse, langue, et théologie en Islam: L’exégèse coranique de
Tabari (m. 311/923). Paris: Librairie Philosophique J. Vrin.
Ginzburg, Carlo. 1999. History, Rhetoric and Proof. The Menahem Stern Jerusalem
Lectures. Hanover, NH: University Press of New England.
Gladwin, Francis. 1801. Dissertations on the Rhetoric, Prosody and Rhyme of the
Persians. London: Oriental Press.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

 299

Goldstone, Jack A. 1998. “The Problem of an ‘Early Modern’ World.” Journal of

Economic and Social History 41 (3): 249–84. www.jstor.com/stable/3632414.
Goody, Jack. 2006. The Theft of History. Cambridge: Cambridge University Press.
Goody, Jack. 2010. Renaissances: The One or the Many? Cambridge: Cambridge
University Press.
Gorekar, N. S. 1970. Indo-Iran Relations: Cultural Aspects. Bombay: Sindhu
Publications.
Grafton, Anthony. 1992. New Words, Ancient Texts: The Power of Tradition and the
Shock of Discovery. Cambridge, MA: Harvard University Press.
Grafton, Anthony. 1997. Commerce with the Classics: Ancient Books and Renaissance
Readers. Ann Arbor, MI: University of Michigan Press.
Gray, Hanna H. 1963. “Renaissance Humanism: The Pursuit of Eloquence.” Journal
of the History of Ideas 24 (4): 497–514. www.jstor.com/stable/2707980.
Greenblatt, Stephen. 1980. Renaissance Self-Fashioning: From More to Shakespeare.
Chicago, IL: University of Chicago Press.
Gregg, Stephen E. 2019. Swami Vivekananda and Non-Hindu Traditions:
A Universal Advaita. London: Routledge.
Grierson, George A. 1889. The Modern Vernacular Literature of Hindustan. Calcutta:
Asiatic Society.
Grunebaum, Gustave E. von. 1944. “The Concept of Plagiarism in Arabic Theory.”
Journal of Near Eastern Studies 3 (4): 234–53. https://doi.org/10.1086/370723.
Grunebaum, Gustave E. von. 1981. Themes in Medieval Arabic Literature. Edited by
Dunning S. Wilson. London: Variorum Reprints.
Guha, Sumit. 2004. “Transitions and Translations: Regional Power and Vernacular
Identity in the Dakhan, 1500–1800.” Comparative Studies of South Asia, Africa
and the Middle East 24 (2): 23–31. https://doi.org/10.1215/1089201X-24-2-23.
Guillén, Claudio. 1971. “Literature as System.” In Literature as System: Essays toward
the Theory of Literary History, 375–419. Princeton, NJ: Princeton University Press.
Guillory, John. 1993. Cultural Capital: The Problem of Literary Canon Formation.
Chicago, IL: University of Chicago Press.
Gulchin Maʿani, Ahmad. 1990. Kārwān-i Hind. 2 vols. Mashhad: Astan-i Quds-i
Razawi.
Habermas, Jürgen. 1991. The Structural Transformation of the Public Sphere: An
Inquiry into a Category of Bourgeois Society. Translated by Thomas Burger with
Frederick Lawrence. Cambridge, MA: MIT Press.
Habermas, Jürgen. 1997. “Modernity: An Unfinished Project.” In Habermas and the
Unfinished Project of Modernity: Critical Essays on the Philosophical Discourse of
Modernity, edited by Maurizio Passerin d’Entrèves and Seyla Benhabib. Translated
by Nicholas Walker, 38–55. Cambridge, MA: MIT Press.
Hadi, Nabi. 1962. Talib-i-Amuli (The Poet-Laureate of Jehangir): His Life and Times.
Aligarh: Aligarh Muslim University Press.
Hadi, Nabi. 1982. Mīrzā Bedil. Aligarh: Aligarh Muslim University Press.
Hadi, Nabi. 1995. Dictionary of Indo-Persian Literature. New Delhi: Indira Gandhi
National Centre for the Arts.
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

300 

Hadley, George. 1972. Introductory Grammatical Remarks on the Persian Language.

Menston: The Scolar Press. Original ed., 1776.
Hakala, Walter Nils. 2010. “Diction and Dictionaries: Language, Literature, and
Learning in Persianate South Asia.” Ph.D. thesis, Department of South Asian
Studies, University of Pennsylvania.
Hakala, Walter Nils. 2011. “Dependence without Sustenance: Coffee
Connoisseurship in Eighteenth-Century Delhi.” Paper presented at Association
for Asian Studies conference, Toronto, 18 March.
Hakala, Walter Nils. 2016. Negotiating Languages: Urdu, Hindi, and the Definition of
Modern South Asia. New York: Columbia University Press.
Hall, Edith. 2008. “Navigating the Realms of Gold: Translation as Access Route to the
Classics.” In Translation and the Classic: Identity as Change in the History of
Culture, edited by Alexandra Lianeri and Vanda Zajko, 315–40. New York: Oxford
University Press.
Hamadeh, Shirin. 2004. “Ottoman Expressions of Early Modernity and the
‘Inevitable’ Question of Westernization.” Journal of the American Society of
Architectural Historians 63 (1): 32–51. www.jstor.com/stable/4127991.
Hanaway, William L. 1989. “Bāzgašt-e adabī.” Encyclopædia Iranica. www.iranica.
com/articles/bazgast-e-adabi-literary-return-a-movement-for-a-return-to-
writing-poetry-in-the-korasani-and-eraqi-styles-that-b.
Hanaway, William L. 2012. “Secretaries, Poets, and the Literary Language.” In
Literacy in the Persianate World: Writing and the Social Order, edited by Brian
Spooner and William L. Hanaway, 95–142. Philadelphia, PA: University of
Pennsylvania.
Haneda, Masashi. 1997. “Emigration of Iranian Elites to India during the 16–18th
centuries.” Cahiers d’Asie centrale 3/4: 129–43.
Hasan, Hadi. 1998. Mughal Poetry, Its Cultural and Political Value. Delhi: Aakar.
Original ed., 1952.
Hasan, Hadi. 2001. History of Indo-Persian Literature. New Delhi: Iran Culture
House.
Hasan, Iqtida. 1995. Later Moghuls and Urdu Literature. Lahore: Ferozsons.
Hashmi, Nasir al-Din. 2012. Dakkan meṁ Urdū. New Delhi: National Council for
the Promotion of Urdu. 1st ed., 1985.
Hawley, John Stratton. 2009. “Introduction.” In The Memory of Love: Sūrdās Sings to
Krishna, 3–44. New York: Oxford University Press.
Haywood, John A. 1956. “An Indian Contribution to the Study of Arabic
Lexicography: The ‘Bulgha’ of Muhammad :
d: dīq
Si : Hasan

Khān Bahādur
(1832–1890).” Journal of the Royal Asiatic Society of Great Britain and Ireland 3/4:
165–80. www.jstor.org/stable/25222839.
Haywood, John A. 1960. Arabic Lexicography: Its History, and Its Place in the General
History of Lexicography. Leiden: Brill.
Hayyim, Sulayman. 1934–6. New Persian–English dictionary. Tehran: Béroukhim.
Heinz, Wilhelm. 1973. Der indische Stil in der persischen Literatur. Wiesbaden:
F. Steiner.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

 301

Heller-Roazen, Daniel. 2012. “The Most Delectable of Languages.” Digital Philology:

A Journal of Medieval Cultures 1 (1): 32–41. http://doi.org/10.1353/dph.2012.0000.
Henderson, Michael M. T. 1975. “Diglossia in Kabul Persian Phonology.” Journal of
the American Oriental Society 95 (4): 651–4. www.jstor.com/stable/601021.
Hermansen, Marcia K., and Bruce B. Lawrence. 2000. “Indo-Persian Tazkiras as
Memorative Communications.” In Beyond Turk and Hindu: Rethinking Religious
Identities in Islamicate South Asia, edited by David Gilmartin and Bruce
B. Lawrence, 149–75. Gainesville, FL: University Press of Florida.
Hernandez, Rebecca Skreslet. 2017. The Legal Thought of Jalāl Al-Dīn Al-Suyūt:ī:
Authority and Legacy. Oxford: Oxford University Press.
Herzfeld, Michael. 1992. The Social Production of Indifference: Exploring the Symbolic
Roots of Western Bureaucracy. New York: Berg.
Higman, Francis. 1997. “1350–1750? The Perspective of Intellectual History.”
Journal of Early Modern History 1 (2): 95–106. https://doi.org/10.1163/
157006597X00055.
Hodge, Carleton T. 1957. “Some Aspects of Persian Style.” Language 33 (3): 355–69.
www.jstor.com/stable/411158.
Hodgson, Marshall G. S. 1968. “Two Pre-Modern Muslim Historians: Pitfalls and
Opportunities in Presenting them to Moderns.” In Towards World Community,
edited by John Nef, 53–68. New York: Humanities Press.
Husain, Iqbal. 1937. The Early Persian Poets of India (A.H. 421–670). Patna: Patna
University Press.
Husain, Iqbal. 1940. “Khān-i
: Ārzū ek Naqqād kī Hais̄iyat

se.” Nigār November: 41–9.
Ingenito, Domenico. 2009. “Tradurre Hāfe
z:
Quattro divān attuali.” Oriente
Moderno 89 (1): 149–69. http://doi.org/10.1163/22138617-08901010.
: t:t̤ ī-yi Fārsī-yi Kitābkhānah-yi
Iran Culture House. 1999. Fihrist-i Nusk: hah-hā-yi Kha :
Duktur Żākir Husain,

Jāmiʿah Millīyah Islāmīyah. New Delhi: Iran Culture House.
Irvine, William. 1898. “Two Proposed Corrections in the ‘Catalogue of Persian Mss.
in the British Museum’ of Dr. C. Rieu.” Journal of the Royal Asiatic Society April:
373–5.
Irvine, William. 1922. Later Mughals. Edited by Jadunath Sarkar. 2 vols. Calcutta:
M. C. Sarkar & Sons.
Isfandagheh, Murtaza Amiri. 2007–8. “Bedil rā Hanūz bah Madrasah Na-
yāwarandah-and.” Qand-i Pārsī 39–40: 447–72.
Islam, Riazul. 1970. Indo-Persian Relations: A Study of the Political and Diplomatic
Relations between the Mughal Empire and Iran. Tehran: Iranian Culture
Foundation.
İslamoğlu, Huri. 2012. “Beyond Histories of Stagnation to ‘Living’ Histories of
Possibilities.” International Journal of Middle East Studies 44: 536–8. https://doi.
org/10.1017/S0020743812000487.
Ivanow, Wladimir. 1926–8. Concise Descriptive Catalogue of the Persian Manuscripts
in the Curzon Collection, Asiatic Society of Bengal. 1 vol, 2 supplements.
Bibliotheca Indica 241, 244, and 248. Calcutta: The Asiatic Society.
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

302 

Ivanow, Wladimir. 1985. Concise Descriptive Catalogue of the Persian Manuscripts in

the Collection of the Asiatic Society of Bengal. Bibliotheca Indica 240. Calcutta: The
Asiatic Society. Original publication, 1924.
Jalibi, Jameel. 1984. Tārīk: h-i Adab-i Urdū. Delhi: Educational Publishing House.
James, George McLeod. 2011. Anand Ram Mukhlis: His Life and Works, 1695–1758.
Delhi: Dilli Kitab Ghar.
Jameson, Frederic. 1986. “Third-World Literature in the Era of Multinational
Capitalism.” Social Text 15 (Autumn): 65–88. www.jstor.com/stable/466493.
Jameson, Frederic. 2002. A Singular Modernity: Essay on the Ontology of the Present.
London: Verso.
Kaicker, Abhishek. 2014. “Unquiet City: Making and Unmaking Politics in Mughal
Delhi, 1707–39.” Ph.D. thesis, Department of History, Columbia University.
Karomat, Dilorom. 2014. “Turki and Hindavi in the World of Persian: Fourteenth-
and Fifteenth-Century Dictionaries.” In After Timur Left: Culture and Circulation
in Fifteenth-Century North India, edited by Francesca Orsini and Samira Sheikh,
130–65. New Delhi: Oxford University Press.
Kaviraj, Sudipta. 1992. “Writing, Speaking, Being: Language and the Historical
Formation of Identities in India.” In Nationalstaat und Sprachkonflikte in Süd-
und Südostasien, edited by Dagmar Hellman-Rajanayagam and Dietmar
Rothermund, 25–65. Stuttgart: Franz Steiner.
Kaviraj, Sudipta. 2003. “The Two Histories of Literary Culture in Bengal.” In Literary
Cultures in History: Reconstructions from South Asia, edited by Sheldon Pollock,
503–66. Berkeley, CA: University of California.
Kaviraj, Sudipta. 2005. “Outline of a Revisionist Theory of Modernity.” European
Journal of Sociology 46 (3): 497–526.
Kaviraj, Sudipta. 2010. “Said and the History of Ideas.” In Cosmopolitan Thought
Zones: South Asia and the Global Circulation of Ideas, edited by Sugata Bose and
Kris Manjapra, 58–81. New York: Palgrave Macmillan.
Kennedy, George Alexander. 1998. Comparative Rhetoric: An Historical and Cross-
Cultural Introduction. New York: Oxford University Press.
Keshavmurthy, Prashant. 2011. “Finitude and the Authorship of Fiction:
Muhammad ‘Awfi’s Preface to his Chronicle, Lubab al-Albab (The Piths of
Intellects).” Arab Studies Journal 19 (1): 94–120. www.jstor.org/stable/23265813.
Keshavmurthy, Prashant. 2012. Ḵān-e Ārezu, Serāj-al-din ʿAli. Encyclopædia Iranica.
www.iranicaonline.org/articles/kan-e-arezu.
Keshavmurthy, Prashant. 2013. “The Local Universality of Poetic Pleasure: Sirājuddin
‘Ali Khān Ārzu and the Speaking Subject.” Indian Economic and Social History
Review 50 (1): 27–45. https://doi.org/10.1177%2F0019464612474168.
Keshavmurthy, Prashant. 2016. Persian Authorship and Canonicity in Late Mughal
Delhi: Building an Ark. London: Routledge.
Keynes, J. M. 1947. “Newton, the Man.” In Newton Tercentenary Celebrations, 15–19
July 1946, 27–34. Cambridge: Cambridge University Press.
Khan, Muhammad Najmulghani. 2006. Bahr : al-Fas: āhat.
: Reprint ed. New Delhi:
National Council for the Promotion of Urdu Language.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

 303

Khan, Mumtaz Ali. 1970. Some Important Persian Prose Writing of the Thirteenth
Century A.D. in India. Aligarh: Department of Persian, Aligarh Muslim University
Press.
Khan, Pasha M. 2017. “What Storytellers Were Worth in Mughal India.”
Comparative Studies of South Asia, Africa and the Middle East 37 (3): 570–87.
www.muse.jhu.edu/article/683285.
Khan, Rashid Husain. 2004. “Fārsī meṁ Majˈhūl Āwāzoṁ se Mutaʿalliq Khān-i: Ārzū
kī Rāʾī.” In Sirāj al-Dīn ʿAlī Khān
: Ārzū: Ek Mut̤ ālaʿah, edited by Shahid Mahuli,
22–33. New Delhi: Ghalib Institute.
Khanlari, Parviz Natel. 1994. “A Three-Thousand-Year-Old Language.” In Persian
Studies in North America: Studies in Honor of Mohammad Ali Jazayery, edited by
Mehdi Marashi and translated by Richard Foltz, 247–54. Bethesda, MD:
Iranbooks.
Khatak, Sarfaraz Khan. 1944. Shaikh Muhammad : ‘Alī Hazīn:

His Life, Times and
Works. Lahore: Sh. Mohammad Ashraf.
:
Khatoon, Rehana. 1987. Ahwāl-o Ās̄ār-i Khān-i
: Ārzū. Delhi: Indo-Persian Society.
Khatoon, Rehana. 1991. Introduction. In Mus̄mir 1991, 4–68.
Khatoon, Rehana. 2004a. “ʿAlī Hazīn

:
wa Khān-i Ārzū kā Adabī Maʿrakah.” In Sirāj
al-Dīn ʿAlī Khān
: Ārzū: Ek Mut̤ ālaʿah, edited by Shahid Mahuli, 82–97. New Delhi:
Ghalib Institute.
Khatoon, Rehana. 2004b. “Khān-i
: Ārzū ke Ustād, Muʿāsir : wa Shāgird.” In Sirāj al-
Dīn ʿAlī Khān
: Ārzū: Ek Mut̤ ālaʿah, edited by Shahid Mahuli, 126–65. New Delhi:
Ghalib Institute.
Kholeif, Fathalla. 1966. A Study on Fakhr al-Dīn al-Rāzī and His Controversies in
Transoxiana. Beirut: Dar el-Machreq.
Khuda Bakhsh Library. 1970. Catalogue of the Arabic and Persian Manuscripts in the
Khuda Bakhsh Oriental Public Library. Patna: The Library.
Kia, Mana. 2009. “Accounting for Difference: A Comparative Look at the
Autobiographical Travel Narratives of Hazin Lāhiji and ʿAbd-al-Karim
Kashmiri.” Journal of Persianate Studies 2 (2): 210–36. https://doi.org/10.1163/
187470909X12535030823814.
Kia, Mana. 2011. “Contours of Persianate Community, 1722–1835.” Ph.D. thesis,
Committee on Middle Eastern Studies, Harvard University.
Kia, Mana. 2014a. “Adab as Ethics of Literary Form and Social Conduct: Reading the
Gulistān in Late Mughal India.” In No Tapping around Philology: A Festschrift in
Celebration and Honor of Wheeler McIntosh Thackston Jr.’s 70th Birthday, edited
by Alireza Korangy and Daniel J. Sheffield, 281–308. Wiesbaden: Harrassowitz.
Kia, Mana. 2014b. “Imagining Iran before Nationalism: Geocultural Meanings of
Land in Azar’s Atashkadeh.” In Rethinking Iranian Nationalism and Modernity,
edited by Kamran Scot Aghaie and Afshin Marashi, 89–112. Austin, TX:
University of Texas.
King, Christopher R. 1994. One Language, Two Scripts: The Hindi Movement in
Nineteenth Century North India. Bombay: Oxford University Press.
King, Richard. 1999. Orientalism and Religion: Postcolonial Theory, India and “the
Mystic East.” Florence, KY: Routledge.
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

304 

Kinra, Rajeev. 2007. “Fresh Words for a Fresh World: Tāza Gūʾī and the Poetics of
Newness in Early Modern Indo-Persian Poetry.” Sikh Formations 3 (2): 125–49.
https://doi.org/10.1080/17448720701733872.
Kinra, Rajeev. 2008. “Secretary-Poets in Mughal India and the Ethos of Persian: The
Case of Chandar Bhan Brahman.” Ph.D. thesis, Department of South Asian
Languages and Civilizations, University of Chicago.
Kinra, Rajeev. 2010. “Master and Munshi: A Brahman Secretary’s Guide to Mughal
Governance.” Indian Economic and Social History Review 47 (4): 527–61.
Kinra, Rajeev. 2011. “This Noble Science: Indo-Persian Comparative Philology, ca.
1000–1800 .” In South Asian Texts in History: Critical Engagements with
Sheldon Pollock, edited by Yigal Bronner, Whitney Cox, and Lawrence McCrea,
359–85. Ann Arbor, MI: Association for Asian Studies.
Kinra, Rajeev. 2015. Writing Self, Writing Empire: Chandar Bhan Brahman and the
Cultural World of the Indo-Persian State Secretary. Oakland, CA: University of
California Press.
Kirmani, Waris. 1972. Evaluation of Ghalib’s Persian Poetry. Aligarh: Aligarh
Muslim University Press.
Kirmani, Waris. 1986. Dreams Forgotten: An Anthology of Indo-Persian Poetry.
Aligarh: Academic Books.
Kopf, L. 2012. “Al-Damīrī.” In Encyclopaedia of Islam. Edited by P. Bearman, Th.
Bianquis, C. E. Bosworth, E. van Donzel, and W. P. Heinrichs. 2nd ed. http://dx.
doi.org/10.1163/1573-3912_islam_SIM_1685.
Koselleck, Reinhart. 2004. Futures Past: On the Semantics of Historical Time.
Translated by Keith Tribe. New York: Columbia University Press.
Kumar, Sunil. 2007. The Emergence of the Delhi Sultanate, 1192–1286. New Delhi:
Permanent Black.
LaCapra, Dominick. 1985. History and Criticism. Ithaca, NY: Cornell University
Press.
Lakoff, George. 1993. “The Contemporary Theory of Metaphor.” In Metaphor and
Thought, edited by Andrew Ortony, 202–51. 2nd ed. New York: Cambridge
University Press.
Lane, Edward William. 1863. An Arabic-English Lexicon. London: Williams and
Norgate.
Law, Vivien. 2003. The History of Linguistics in Europe: From Plato to 1600.
Cambridge: Cambridge University Press.
Lazard, Gilbert. 1993. The Origins of Literary Persian. Bethesda, MD: Foundation for
Iranian Studies.
Lazard, Gilbert. 1995. La formation de la langue persane. Paris: Peeters.
Leese, Simon. 2019. “Longing for Salmá and Hind: (Re)producing Arabic Literature
in 18th and 19th-Century North India.” Ph.D. thesis, Centre for Cultural, Literary
and Postcolonial Studies, SOAS.
Lefèvre, Corinne. 2014. “The Court of ʿAbd-ur-Rahīm : Khān-i Khānān as a Bridge
between Iranian and Indian Cultural Traditions.” In Culture and Circulation in
Early Modern India, edited by Thomas de Bruijn and Allison Busch, 75–106.
Leiden: Brill.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

 305

Lelyveld, David. 2011. “Sauda Sulaf: Urdu in the Two Versions of Sayyid Ahmad
Khan’s Asaru’s-Sanadid.” Annual of Urdu Studies 26: 21–38.
Lewis, Franklin. 2018. “To Round and Rondeau the Canon: Jāmī and Fānī’s
Reception of the Persian Lyrical Tradition.” In Jāmī in Regional Contexts: The
Reception of ʿAbd Al-Rahmān: Jāmī’s Works in the Islamicate World, edited by
Thibaut d’Hubert and Alexandre Papas, 463–567. Leiden: Brill.
Lewisohn, Leonard. 1999. “Sufism and the School of Is: fahān.” In Late Classical
Persianate Sufism: The Safavid and Mughal Period, edited by Leonard Lewisohn,
63–134, Vol. 3 of The Heritage of Sufism. Oxford: Oneworld.
Lianeri, Alexandra, and Vanda Zajko. 2008. “Introduction: Still Being Read after so
Many Years: Rethinking the Classic through Translation.” In Translation and the
Classic: Identity as Change in the History of Culture, edited by Alexandra Lianeri
and Vanda Zajko, 1–23. New York: Oxford University Press.
Lieberman, Victor. 1997. “Transcending East-West Dichotomies: State and Culture
Formation in Six Ostensibly Disparate Areas.” Modern Asian Studies 31 (3):
463–546.
Lockhart, L. 1938. Nadir Shah: A Critical Study Based Mainly upon Contemporary
Sources. London: Luzac & Co.
Lopez, Robert S. 1959. “Hard Times and Investment in Culture.” In The Renaissance:
Medieval or Modern?, edited by Karl H. Dannenfeldt, 50–61. Boston, MA:
D. C. Heath.
Losensky, Paul E. 1994. “ ‘The Allusive Field of Drunkenness’: Three Safavid-Moghul
Responses to a Lyric by Bābā Fighānī.” In Reorientations/Arabic and Persian
Poetry, edited by Suzanne Pinckney Stetkevych, 227–62. Bloomington, IN:
Indiana University Press.
Losensky, Paul E. 1998. Welcoming Fighānī: Imitation and Poetic Individuality in the
Safavid-Mughal Ghazal. Costa Mesa, CA: Mazda.
Losensky, Paul E. 2003. “Sāʾeb

Tabrizi.” Encyclopædia Iranica. www.iranicaonline.
org/articles/saeb-tabrizi.
Loucel, Henri. 1963–4. “L’origine du langage d’après les grammairiens arabes.”
Arabica 10–11: 57–72, 151–87, 188–208, 253–81.
Lunn, David. 2018. “Across the Divide: Looking for the Common Ground of
Hindustani.” Modern Asian Studies 52 (6): 2056–79. http://doi.org/10.1017/
S0026749X1600069X.
Mackey, William F. 1986. “The Polyglossic Spectrum.” In The Fergusonian Impact: In
Honor of Charles A. Ferguson on the Occasion of his 65th Birthday, edited by
Joshua A. Fishman et al., 2:238–43. Berlin: Mouton de Gruyter.
Majeed, Javed. 1992. Ungoverned Imaginings: James Mill’s The History of British
India and Orientalism. Oxford: Clarendon.
Majeed, Javed. 2011. “What’s in a (Proper) Name? Particulars, Individuals, and
Authorship in the Linguistic Survey of India and Colonial Scholarship.” In
Knowledge Production, Pedagogy, and Institutions in Colonial India, edited by
Indra Sengupta and Daud Ali, 19–39. New York: Palgrave Macmillan.
Malik, Zahir Uddin. 2006. The Reign of Muhammad Shah, 1719–1748. New Delhi:
Icon Publications.
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

306 

Mallette, Karla. 2013. “Boustrophedon: Towards a Literary Theory of the

Mediterranean.” In A Sea of Languages: Rethinking the Arabic Role in Medieval
Literary History, edited by Suzanne Conklin Akbari and Karla Mallette, 254–66.
Toronto: University of Toronto Press.
Manson, Richard. 1969. The Theory of Knowledge of Giambattista Vico. Hamden,
CT: Archon Books.
Marchand, Suzanne. 2010. “On Orientalism and Iconoclasm: German Scholarship’s
Challenge to the Saidian Model.” In Cosmopolitan Thought Zones: South Asia and
the Global Circulation of Ideas, edited by Sugata Bose and Kris Manjapra, 260–83.
New York: Palgrave Macmillan.
Margoliouth, D. S. 1904. Catalogue of the Oriental Manuscripts in the Library of Eton
College. Oxford: Oxford University Press.
Markovits, Claude, Jacques Pouchepadass, and Sanjay Subrahmanyam. 2003.
“Introduction.” In Society and Circulation: Mobile People and Itinerant Cultures
in South Asia, 1750–1950, 1–22. New Delhi: Permanent Black.
Marogy, Amal E. 2010. Kitāb Sībawayhi: Syntax and Pragmatics. Leiden: Brill.
Marshall, D. N. 1967. Mughals in India: A Bibliographical Survey. Bombay: Asia
Publishing House.
Marshall, P. J., ed. 2003. The Eighteenth Century in Indian History: Evolution or
Revolution? Oxford: Oxford University Press.
Marshall, P. J. 2004. “Johnson, Richard (1753–1807).” In Oxford Dictionary of
National Biography. Oxford: Oxford University Press. www.oxforddnb.com/
view/article/63514.
Masica, Colin P. 1991. The Indo-Aryan Languages. Cambridge: Cambridge
University Press.
Matheson, Peter. 1998. The Rhetoric of the Reformation. Edinburgh: T&T Clark.
Matini, J. 2011. “Āẕar Bīgdelī.” Encyclopædia Iranica. www.iranicaonline.org/
articles/azar-bigdeli-hajj-lotf-ali-big.
Matthews, David John. 1976. “Dakanī Language and Literature 1500–1700 A.D.” Ph.
D. thesis, School of Oriental and African Studies, University of London.
Matthews, David John. 1993. “Eighty Years of Dakani Scholarship.” Annual of Urdu
Studies 8: 82–99.
McGregor, R. S. 1984. Hindi Literature from Its Beginnings to the Nineteenth Century.
Wiesbaden: Harrassowitz.
McGregor, R. S., ed. 1993. Oxford Hindi–English Dictionary. Oxford: Oxford
University Press.
McGregor, R. S. 2001. The Formation of Modern Hindi as Demonstrated in Early
“Hindi” Dictionaries. Amsterdam: Royal Netherlands Academy of Arts and
Sciences.
McGregor, Stuart. 2003. “The Progress of Hindi, Part I: The Development of a
Transregional Idiom.” In Literary Cultures in History: Reconstructions from
South Asia, edited by Sheldon Pollock, 912–57. Berkeley, CA: University of
California.
Meisami, Julie Scott. 1987. Medieval Persian Court Poetry. Princeton, NJ: Princeton
University Press.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

 307

Meisami, Julie Scott. 2003. Structure and Meaning in Medieval Arabic and Persian
Poetry: Orient Pearls. London: RoutledgeCurzon.
Memon, M. U. 2011. “Abu’l-Barakat Lahuri.” Encyclopædia Iranica. www.
iranicaonline.org/articles/abul-barakat-monir-lahuri-indo-persian-poet-
commonly-known-as-molla-monir-lahuri-b.
Mignolo, Walter D. 2003. The Darker Side of the Renaissance: Literacy, Territoriality,
and Colonization. 2nd ed. Ann Arbor, MI: University of Michigan.
Mikkelson, Jane. 2017. “Of Parrots and Crows: Bīdil and Hazīn
in Their Own
Words.” Comparative Studies of South Asia, Africa and the Middle East 37 (3):
510–30. http://doi.org/10.1215/1089201x-4279224.
Minissale, Gregory. 2006. Images of Thought: Visuality in Islamic India, 1550–1750.
Newcastle: Cambridge Scholars Press.
Mir, Farina. 2006. “Imperial Policy, Provincial Practices: Colonial Language Policy in
Nineteenth-Century India.” Indian Economic and Social History Review 43 (4):
395–427. http://doi.org/10.1177/001946460604300401.
Mirza, Mohammad Wahid. 1935. The Life and Works of Amir Khusrau. Lahore:
University of the Punjab.
Mitchell, Lisa. 2009. Language, Emotion, and Politics in South India: The Making of a
Mother Tongue. Bloomington, IN: Indiana University Press.
Moazzam, Anwar. 2012. “Urdu Inshā: The Hyderābād Experiment, 1860–1948.” In
Literacy in the Persianate World: Writing and the Social Order, edited by Brian
Spooner and William L. Hanaway, 311–27. Philadelphia, PA: University of
Pennsylvania.
Mohiuddin, Momin. 1960. “Sabk-i-Hindi (The Indian Style of Persian Prose) with
Special Reference to Insháʾ.” Indo-Iranica 13 (2): 19–30.
Mohiuddin, Momin. 1971. The Chancellery and Persian Epistolography under the
Mughals, from Bábur to Sháh Jahán, 1526–1658. Calcutta: Iran Society.
Moin, A. Azfar. 2012. The Millennial Sovereign: Sacred Kingship and Sainthood in
Islam. New York: Columbia University Press.
Mukherjee, S. N. 1968. Sir William Jones: A Study in Eighteenth-Century British
Attitudes to India. London: Cambridge University Press.
Mukhlis, Anand Ram. 1993. Encyclopaedic Dictionary of Medieval India: Mirat-ul-
Istilah. Translated by Tasneem Ahmad. Delhi: Sundeep Prakashan.
Muradabadi, Muhammad Shafiq. 2000. Bahumūlya pān: dulipiyāṁ : (Arabī, Fārsī,
tathā Urdū). Lucknow: Amiruddaulah Public Library.
Nagel, Alexander, and Christopher S. Wood. 2010. Anachronistic Renaissance. New
York: Zone.
Naik, C. R. 1966. ʿAbdu’r-Rahīm : Khān-i-Khānān and His Literary Circle.
Ahmedabad: Gujarat University Press.
Naim, C. M. 1989. “Poet-Audience Interaction at Urdu Mushaʿiras.” In Urdu and
Muslim South Asia: Studies in Honour of Ralph Russell, edited by Christopher
Shackle, 167–73. London: SOAS.
Naim, C. M., ed. and trans. 1999. Żikr-i Mīr. New Delhi: Oxford University Press.
Naim, C. M. 2006. “Shaikh Imam Bakhsh Sahba’i: Teacher, Scholar, Poet, and
Puzzle-Master.” In The Delhi College: Traditional Elites, the Colonial State, and
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

308 

Education before 1857, edited by Margrit Pernau, 145–85. New Delhi: Oxford
University Press.
Naqvi, Shariyar. 1962. Farhangniwīsī-i Fārsī dar Hind wa Pākistān. Tehran: Idarah-i
Kull-i Nigarish-i Vizarat-i Farhang.
Narang, Gopi Chand. 1987. Amīr Khusrau : :
kā Hindawī Kalām maʿ Nuskhah-yi
Barlin-i Żak: hīrah-yi Ispringar. Chicago, IL: Amir Khusrau Society of America.
Nath, R., and Faiyaz Gwaliari, eds. 1981. India as Seen by Amir Khusrau in 1318 A.D.
[Nuh Sipihr]. Jaipur: Historical Research Documentation Programme.
Newman, Karen, and Jane Tylus. 2015. “Introduction.” In Early Modern Cultures of
Translation, edited by Karen Newman and Jane Tylus, 1–24. Philadelphia, PA:
University of Pennsylvania.
Nicholson, Reynold A. 1907. A Literary History of the Arabs. New York: Charles
Scribner.
Nizami, K. A. 2002. Gwāliyār. In Encyclopaedia of Islam. Edited by: P. Bearman, Th.
Bianquis, C. E. Bosworth, E. van Donzel, and W. P. Heinrichs. 2nd ed. http://dx.
doi.org/10.1163/1573-3912_islam_SIM_2569.
Nora, Pierre. 1996. “General Introduction: Between History and Memory.” In Realms
of Memory: The Construction of the French Past, edited by Lawrence D. Kritzman.
Translated by Arthur Goldhammer, 1: 1–20. New York: Columbia University
Press.
O’Daly, Gerard J. P. 1999. Augustine’s City of God: A Reader’s Guide. Oxford:
Clarendon Press.
Ogborn, Miles. 2007. India Ink: Script and Print in the Making of the English East
India Company. Chicago, IL: University of Chicago Press.
Olender, Maurice. 1994. “Europe, or How to Escape Babel.” History and Theory 33
(4): 5–25.
Ong, Walter J. 1959. “Latin Language Study as a Renaissance Puberty Rite.” Studies in
Philology 56 (2): 103–24. www.jstor.com/stable/4173267.
Ong, Walter J. 1971. Rhetoric, Romance, and Technology: Studies in the Interaction of
Expression and Culture. Ithaca, NY: Cornell University Press.
Orsini, Francesca. 2010. “Barahmasas in Hindi and Urdu.” In Before the Divide:
Hindi and Urdu Literary Culture, edited by Francesca Orsini, 142–77. New Delhi:
Orient Blackswan.
Orsini, Francesca. 2012. “How to Do Multilingual Literary History? Lessons from
Fifteenth- and Sixteenth-Century North India.” Indian Economic Social History
Review 42: 225–46. https://doi.org/10.1177%2F001946461204900203.
Orsini, Francesca. 2014. “Traces of a Multilingual World: Hindavi in Persian Texts.”
In After Timur Left: Culture and Circulation in Fifteenth-Century North India,
edited by Francesca Orsini and Samira Shaikh, 403–36. New Delhi: Oxford
University Press.
Owens, Jonathan. 2006. A Linguistic History of Arabic. Oxford: Oxford University
Press.
Parsinejad, Iraj. 2003. A History of Literary Criticism in Iran (1866–1951). Bethesda,
MD: Ibex.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

 309

Passannante, Gerard. 2011. The Lucretian Renaissance: Philology and the Afterlife of
Tradition. Chicago, IL: University of Chicago Press.
Paul, Ludwig. 2005. “The Language of the Šāhnāme in Historical and Dialectal
Perspective.” In Languages of Iran: Past and Present, edited by Dieter Weber,
141–51. Wiesbaden: Harrassowitz.
Pauwels, Heidi. 2009. “The Saint, the Warlord, and the Emperor: Braj Bhakti and
Bundelā Loyalty.” Journal of the Economic and Social History of the Orient 52:
187–228. www.jstor.com/stable/25651163.
Pauwels, Heidi. 2012. “Literary Moments of Exchange in the 18th Century: The New
Urdu Vogue Meets Krishna Bhakti.” In Indo-Muslim Cultures in Transition,
edited by Alka Patel and Karen Leonard, 61–85. Leiden: Brill.
Pauwels, Heidi. 2015. Culture Exchange in Eighteenth-Century India: Poetry and
Paintings from Kishangarh. Berlin: EB-Verlag.
Pedersen, Johannes. 1922. “The Sābians.”

In A Volume of Oriental Studies Presented
to Edward G. Browne on his 60th Birthday, edited by T. W. Arnold and Reynold
A. Nicholson, 383–91. Cambridge: Cambridge University Press.
Pellò, Stefano. 2004. “Persiano e hindī nel Musmir di Sirāj al-Dīn ʿAlī Xān Ārzū.” In
L’Onagro Maestro: Miscellanea di fuochi accesi per Gianroberto Scarcia in occa-
sione del suo LXX sadè, edited by Rudy Favaro, Simone Cristoforetti, and Matteo
Compareti, 243–72. Venice: Cafoscarina.
Pellò, Stefano. 2008. “Hindu Persian Poets.” Encyclopædia Iranica. www.
iranicaonline.org/articles/hindu-persian-poets.
Pellò, Stefano. 2012. “Safina-ye Ḵošgu.” Encyclopædia Iranica. www.iranicaonline.
org/articles/safina-ye-kosgu.
Pellò, Stefano. 2014. “Persian as a Passe-Partout: The Case of Mīrzā ʿAbd al-Qādir
Bīdil and His Hindu Disciples.” In Culture and Circulation in Early Modern India,
edited by Allison Busch and Thomas de Bruijn, 21–46. Leiden: Brill.
Pellò, Stefano. 2016. “A Linguistic Conversion: Mīrzā Muhammad: Hasan

Qatīl and
the Varieties of Persian (ca. 1790).” In Borders: Itineraries on the Edges of Iran,
edited by Stefano Pellò, 203–40. http://doi.org/10.14277/6969-100-3/EUR-5-9.
Perlin, Frank. 1985. “State Formation Reconsidered: Part Two.” Modern Asian
Studies 19 (3): 415–80.
Pernau, Margrit, and Yunus Jaffery, eds. 2009. Information and the Public Sphere:
Persian Newsletters from Mughal Delhi. New Delhi: Oxford University Press.
Perry, John R. 1998. “The Waning of Indo-Persian Lexicography: Examples from
Some Rare Books and Manuscripts of the Subcontinent.” In Iran and Iranian
Studies: Essays in Honor of Iraj Afshar, edited by Kambiz Eslami, 329–40.
Princeton, NJ: Zagros.
Perry, John R. 2002. “Čerāḡ-e Hedāyat.” Encyclopædia Iranica. www.iranicaonline.
org/articles/cerag-e-hedayat.
Perry, John R. 2003. “Hazin

Lāhiji.” Encyclopædia Iranica. www.iranicaonline.org/
articles/ hazin-lahiji.
Perry, John R. 2009. “The Origin and Development of Literary Persian.” In General
Introduction to Persian Literature, ed. J. T. P. De Bruijn, 43–70. Vol. 1 of A History
of Persian Literature. London: I. B. Tauris.
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

310 

Perry, John R. 2012. “New Persian: Expansion, Standardization, and Inclusivity.” In

Literacy in the Persianate World: Writing and the Social Order, edited by Brian
Spooner and William L. Hanaway, 70–94. Philadelphia, PA: University of
Pennsylvania.
Pertsch, Wilhem. 1888. Verzeichniss der Persischen Handschriften der Königlichen
Bibliothek zu Berlin. Berlin: A. Asher.
Petievich, Carla. 1992. Assembly of Rivals: Delhi, Lucknow, and the Urdu Ghazal.
Delhi: Manohar.
Petievich, Carla. 2017. “Revisiting that Earth(l)y, Glorious Tradition.” In From Urdu
and Indo-Persian Thought, Poetics, and Belles Lettres, edited by Alireza Korangy,
87–100. Leiden: Brill. http://doi.org/10.1163/9789004341579_005.
Petrucci, Armando. 1995. Writers and Readers in Medieval Italy: Studies in the
History of Written Culture. Translated by Charles M. Radding. New Haven, CT:
Yale University Press.
Phukan, Shantanu. 2000a. “The Rustic Beloved: Ecology of Hindi in a Persianate
World.” Annual of Urdu Studies 15: 3–30.
Phukan, Shantanu. 2000b. “Through a Persian Prism: Hindi and Padmāvat in the
Mughal Imagination.” Ph.D. thesis, Department of South Asian Languages and
Civilizations, University of Chicago.
Platts, John T. 1884. A Dictionary of Urdu, Classical Hindi, and English. London:
W. H. Allen & Co. http://dsal.uchicago.edu/dictionaries/platts.
Pollock, Sheldon. 1985. “The Theory of Practice and the Practice of Theory in Indian
Intellectual History.” Journal of the American Oriental Society 105 (3): 499–519.
www.jstor.com/stable/601525.
Pollock, Sheldon. 1993. “Deep Orientalism? Notes on Sanskrit and Power beyond
the Raj.” In Orientalism and the Postcolonial Predicament, edited by Carol
A. Breckenridge and Peter van der Veer, 76–133. Philadelphia, PA: University of
Pennsylvania Press.
Pollock, Sheldon. 1995. “Literary History, Indian History, World History.” Social
Scientist 23 (10/12): 112–42. www.jstor.com/stable/3517886.
Pollock, Sheldon. 1998. “The Cosmopolitan Vernacular.” Journal of Asian Studies
57 (1): 6–37. http://doi.org/10.2307/2659022.
Pollock, Sheldon. 2000. “Cosmopolitan and Vernacular in History.” Public Culture
12 (3): 591–625. https://doi.org/10.1215/08992363-12-3-591.
Pollock, Sheldon. 2001. “New Intellectuals in Seventeenth-Century India.”
Economic and Social History Review 38 (1): 3–31. https://doi.org/10.1177%
2F001946460103800101.
Pollock, Sheldon. 2004. The Ends of Man at the End of Premodernity. Amsterdam:
Royal Netherlands Academy of Arts and Sciences.
Pollock, Sheldon. 2006. The Language of the Gods in the World of Men: Sanskrit,
Culture, and Power in Premodern India. Berkeley, CA: University of California
Press.
Pollock, Sheldon. 2007. “We Need to Find What We Are Not Looking For.”
International Institute for Asian Studies Newsletter 43 (Spring): 1, 4–5.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

 311

Pollock, Sheldon. 2009. “Future Philology? The Fate of a Soft Science in a Hard World.”
Critical Inquiry 35 (Summer): 931–61. www.jstor.com/stable/10.1086/599594.
Pollock, Sheldon. 2011. “The Languages of Science in Early Modern India.” In Forms
of Knowledge in Early Modern Asia, edited by Sheldon Pollock, 19–48. Durham,
NC: Duke University Press.
Pollock, Sheldon, Homi K. Bhabha, Carol A. Breckenridge, and Dipesh Chakrabarty.
2000. “Cosmopolitanisms.” Public Culture 12 (3): 577–89. http://doi.org/10.1215/
08992363-12-3-577.
Pollock, Sheldon, Benjamin A. Elman, and Ku-ming Kevin Chang, eds. 2015. World
Philology. Cambridge, MA: Harvard University Press.
Powell, Avril A. 2011. “Old Books in New Bindings: Ethics and Education in Colonial
India.” In Knowledge Production, Pedagogy, and Institutions in Colonial India,
edited by Indra Sengupta and Daud Ali, 199–225. New York: Palgrave Macmillan.
Pritchett, Frances W. 1994. Nets of Awareness: Urdu Poetry and Its Critics. Berkeley,
CA: University of California Press.
Pritchett, Frances W. 2001. “Everybody Knows This Much. . .” In Āb-i Hayāt
of
:
Muhammad Husain

Āzād, 1–17. New Delhi: Oxford University Press, 2001.
Pritchett, Frances W. 2003. “A Long History of Urdu Literary Culture, Part 2:
Histories, Performances, and Masters.” In Literary Cultures in History:
Reconstructions from South Asia, edited by Sheldon Pollock, 864–911. Berkeley,
CA: University of California.
Qasmi, Sharif Husain. 2004. “Fārsī meṁ Adabī Tanqīd kī Rawāyat aur Tanbīh al-
Ġhāfilīn.” In Sirāj al-Dīn ʿAlī Khān
: Ārzū: Ek Mut̤ ālaʿah, edited by Shahid Mahuli,
71–85. New Delhi: Ghalib Institute.
Qasmi, Sharif Husain. 2008. A History of Indo-Persian Literature (1220–1274 A.H./
1806–1857 A.D.). Delhi: Department of Persian, Delhi University.
Qazwah, Ali Reza. 2008. “Sih Nigāh bah Jā-yi Yak Muqaddamah.” Qand-i Pārsī
39–40 (Winter–Spring): 7–19.
Qutbuddin, Tahera. 2007. “Arabic in India: A Survey and Classification of Its Uses,
Compared with Persian.” Journal of the American Oriental Society 127 (3): 315–38.
www.jstor.com/stable/20297278.
Rafiʿ, ʿAbd al-Rafiʿ Haqiqat. 2004. Muhājirat-hā-yi Tārīk: hī-yi Īrāniyān: Az Qarn-i
Awwal-i Hijrī tā Imrūz. Tehran: Intisharat-i Kumish.
Rahimpoor, Mahdi. 2008a. “ʿAt̤ īyah-i Kubra̍ wa Mauhibat-i ʿUz̤ ma̍.” Ayene-ye Miras
NS 5 (4 Winter): 334–50.
Rahimpoor, Mahdi. 2008b. “Siyarī dar Ahwāl-o
: Ās̄ār-i Sirāj al-Dīn ʿAlī Khān-i
: Ārzū-
yi Akbarābādī.” Ayene-ye Miras NS 6 (1, Spring): 289–318.
Rahimpoor, Mahdi. 2012. Bar Kh : wān-i
: Ārzū. Qom: Majmaʿ-i Zakhair-i Islami.
Rahman, M. L. 1970. Persian Literature in India during the Time of Jahangir and
Shah Jahan. Baroda: Department of Persian and Urdu, M. S. University.
Rahman, S. Shah Md. Attaur, ed. 1959. Safīnah-i Kh : wushgo.
: Patna: Institute of Post
Graduate Studies and Research in Arabic and Persian.
Rahman, Tariq. 2010. “Urdu and the Muslim Identity: Standardization of Urdu in
the Eighteenth and Early Nineteenth Centuries.” Annual of Urdu Studies 25:
83–107.
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

312 

Rai, Amrit. 1984. A House Divided: The Origin and Development of Hindi/Hindavi.
Delhi: Oxford University Press.
Ramaswamy, Sumathi. 2007. “Conceit of the Globe in Mughal Visual Practice.”
Comparative Studies in Society and History 49 (4): 751–82. https://doi.org/10.
1017/S0010417507000758.
Ramezannia, Mehrdad. 2010. “Persian Print Culture in India, 1780–1880.” Ph.D.
thesis, Centre for Historical Studies, Jawaharlal Nehru University.
Razvi, M. H., and M. H. Qaisar. 1981. Catalogue of Manuscripts in the Maulana Azad
Library: Habibganj Collection (Persian). Vol. 1, Pt. 1 (Poetry). Aligarh: Aligarh
Muslim University Press.
Ricci, Ronit. 2011. Islam Translated: Literature, Conversion, and the Arabic
Cosmopolis of South and Southeast Asia. Chicago, IL: University of Chicago Press.
Richards, John F. 1984. “Norms of Comportment among Imperial Mughal Officers.”
In Moral Conduct and Authority: The Place of Adab in South Asian Islam, edited
by Barbara Daly Metcalfe, 255–89. Berkeley, CA: University of California Press.
Richards, John F. 1996. The Mughal Empire. Cambridge: Cambridge University
Press.
Richards, John F. 1997. “Early Modern India and World History.” Journal of World
History 8 (2): 197–209. www.jstor.com/stable/20068593.
Rieu, Charles. 1879–83. Catalogue of the Persian Manuscripts in the British Museum.
3 vols. London: British Museum.
Rizvi, Saiyid Athar Abbas. 1986. A Socio-Intellectual History of the Isnā ’Asharī Shī’īs
in India. Canberra: Ma’rifat.
Roberts, Alexandre M. 2017. “Being a Sabian at Court in Tenth-Century Baghdad.”
Journal of the American Oriental Society 137 (2): 253–77. www.jstor.com/stable/
10.7817/jameroriesoci.137.2.0253.
Robins, R. H. 1996. “Jones as a General Linguist in the Eighteenth-Century Context.”
In Objects of Enquiry: The Life, Contributions, and Influences of Sir William Jones
(1746–1794), edited by Garland Cannon and Kevin R. Brine, 83–91. New York:
New York University Press.
Robinson, Francis. 1997. “Ottomans-Safavids-Mughals: Shared Knowledge and
Connective Systems.” Journal of Islamic Studies 8 (2): 151–84.
Ross, E. Denison, and Edward G. Browne. 1902. Catalogue of Two Collections of
Persian and Arabic Manuscripts Preserved in the India Office Library. London:
Eyre and Spottiswoode.
Rowland, Ingrid D. 1994. “Revenge of the Regensburg Humanists, 1493.” The
Sixteenth Century Journal 25 (2): 307–22. www.jstor.com/stable/2542883.
Rowland, Ingrid D. 1998. The Culture of the High Renaissance: Ancients and Moderns
in Sixteenth-Century Rome. Cambridge: Cambridge University Press.
Sachau, Eduard, and Hermann Ethé. 1889. Catalogue of the Persian, Turkish,
Hindustani and Pushtu Manuscripts in the Bodleian Library. Oxford: Clarendon
Press.
Sadiq, Muhammad. 1964. A History of Urdu Literature. London: Oxford University
Press.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

 313

:
Safa, Zabih Allah. 1984. Tārīkh-i Adabīyāt dar Īrān. Vol. 5, Pt. 1. Tehran: Shirkat-i
Muʾallifan wa Mutarjiman-i Iran.
Said, Edward W. 1979. Orientalism. New York: Vintage.
Saliba, George. 2007. Islamic Science and the Making of the European Renaissance.
Cambridge, MA: MIT Press.
Sarkar, Jadunath. 1920. Mughal Administration. Calcutta: M. C. Sarkar.
Sarkar, Jadunath. 1949. Anecdotes of Aurangzib. 3rd ed. Calcutta: M. C. Sarkar.
Sarkar, Jadunath. 2008. “The Condition of the People in Aurangzīb’s Reign.” In The
Mughal State, 1526–1750, edited by Muzaffar Alam and Sanjay Subrahmanyam,
301–22. New Delhi: Oxford University Press. Original ed., 1924.
: Biography and Background. Cambridge:
Sartain, E. M. 1975. Jalāl al-Dīn al-Suyūtī:
Cambridge University Press.
Schendl, Herbert. 2000. “Linguistic Aspects of Code-Switching in Medieval English
Texts.” In Multilingualism in Later Medieval Britain, edited by D. A. Trotter,
77–92. Cambridge: D. S. Brewer.
Schwartz, Kevin Lewis. 2014. “Bâzgasht-i Adabî (Literary Return) and Persianate
Literary Culture in Eighteenth and Nineteenth Century Iran, India, and
Afghanistan.” Ph.D. thesis, Department of Near Eastern Studies, University of
California, Berkeley.
Schwartz, Kevin L. 2016. “The Local Lives of a Transregional Poet: ʿAbd al-Qādir
Bidel and the Writing of Persianate Literary History.” Journal of Persianate Studies
9 (1): 83–106.
Schwartz, Kevin L. 2020. “A Transregional Persianate Library: The Production and
Circulation of Tadhkiras of Persian Poets in the 18th and 19th Centuries.”
International Journal of Middle East Studies 52 (1): 109–35. http://doi.org/10.
1017/S0020743819000874.
Seyed-Gohrab, Ali Asghar. 2012. “Introduction: Persian Rhetorical Figures.” In
Metaphor and Imagery in Persian Poetry, edited by Ali Asghar Seyed-Gohrab,
1–14. Leiden: Brill.
Shackle, Christopher, and Rupert Snell. 1990. Hindi and Urdu: A Common Reader.
London: SOAS.
Shafiʿi Kadkani, Muhammad Reza. 1981. “Persian Literature (Belles-Lettres) from the
Time of Jāmī to the Present Day.” In History of Persian Literature from the
Beginning of the Islamic Period to the Present Day, edited by George Morrison,
133–206. Leiden: Brill.
Shamisa, Cyrus, ed. 2001. Mus: t:alahāt
: al-Shuʿarā. Tehran: Intisharat-i Firdaus.
Shamisa, Cyrus, ed. 2002. ʿAtīyah-i
: Kubra̍ wa Mauhibat-i ʿUz̤ ma̍. Tehran:
Chapkhanah-i Ramin.
Shapiro, Marianne. 1990. De Vulgari Eloquentia: Dante’s Book of Exile. Lincoln, NE:
University of Nebraska Press.
Sharma, Sunil. 2003. “The Land of Darkness: Images of India in the Works of Some
Safavid Poets.” Studies on Persianate Societies 1: 97–110.
Sharma, Sunil. 2009. “Translating Gender: Āzād Bilgrāmī on the Poetics of the Love
Lyric and Cultural Synthesis.” The Translator 15 (1): 87–103. https://doi.org/10.
1080/13556509.2009.10799272.
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

314 

Sharma, Sunil. 2012a. “The Function of the Catalogue of Poets in Persian Poetry.” In
Metaphor and Imagery in Persian Poetry, edited by Ali Asghar Seyed-Gohrab,
231–48. Leiden: Brill.
Sharma, Sunil. 2012b. “Redrawing the Boundaries of ʿAjam in Early Modern Persian
Literary Histories.” In Iran Facing Others: Identity Boundaries in a Historical
Perspective, edited by Abbas Amanat and Farzin Vejdani, 49–62. New York:
Palgrave Macmillan.
Sharma, Sunil. 2013. “Review of Hamid Dabashi’s The World of Persian Literary
Humanism.” International Journal of Middle East Studies 45 (4): 840–2. http://doi.
org/10.1017/S0020743813001104.
Sheffield, Daniel J. 2014. “The Language of Heaven in Safavid Iran: Speech and
Cosmology in the Thought of Āẕar Kayvān and His Followers.” In No Tapping
around Philology: A Festschrift in Celebration and Honor of Wheeler McIntosh
Thackston Jr.’s 70th Birthday, edited by Alireza Korangy and Daniel J. Sheffield,
161–83. Wiesbaden: Harrassowitz.
Shelley, Percy Bysshe. 1840. Essays, Letters from Abroad, Translations and
Fragments. London: Edward Moxon.
Sheth, Sudev J. 2019. “Manuscript Variations of Dabistān-i Maẕāhib and Writing
Histories of Religion in Mughal India.” Manuscript Studies 4 (1): 19–41. http://doi.
org/10.1353/mns.2019.0010.
Shirani, Hafiz Mahmud. 1966. “Fārsī-Zabān kī ek Qadīm Farhang meṁ Urdū-Zabān
kā ʿUns: ur.” In Maqālāt-i Hāfiz

̤ Mahmūd: Shīrānī, edited by Mazhar Mahmud
Shirani, 1: 102–31. Lahore: Majlis-i Taraqqi-yi Urdu.
Shirani, Mahmud. 2005. Panjāb meṁ Urdū. New Delhi: National Council for the
Promotion of Urdu. Original ed., 1982.
Shukla, Ravi Shankar. 1947. Lingua Franca for India (Hind). Lucknow: Oudh
Publishing House.
Siddiqi, Moazzam. 1989. Bīdel (Bēdil), Mīrzā ʿAbd-al-Qāder. Encyclopædia Iranica.
www.iranicaonline.org/articles/bidel-bedil-mirza-abd-al-qader-b.
Simidchieva, Marta. 2003. “Imitation and Innovation in Timurid Poetics: Kashifi’s
Badāyiʿ al-Afkār and Its Predecessors, al-Muʿjam and Hadāʾiq

: Iranian
al-Sihr.”
Studies 36 (4): 509–30. www.jstor.org/stable/4311580.
Skinner, Quentin. 2002. Regarding Method. Cambridge: Cambridge University Press.
Smith, Matthew C. 2009. “Literary Connections: Bahār’s Sabkshenāsi and the
Bāzgasht-e Adabi.” Journal of Persianate Studies 2 (2): 194–209.
Smyth, William E. 1994. “Shams-i Qays on the Nature of Poetry.” Edebiyat N.S. 5 (2):
291–302.
Somerset, Fiona, and Nicholas Watson. 2003. “Preface: On ‘Vernacular.’ ” In The
Vulgar Tongue: Medieval and Postmedieval Vernacularity, edited by Fiona
Somerset and Nicholas Watson, ix–xiv. University Park, PA: Pennsylvania State
University Press.
Spooner, Brian, and William L. Hanaway. 2012. “Introduction: Persian as Koine,
Written Persian in World-historical Perspective.” In Literacy in the Persianate
World: Writing and the Social Order, edited by Brian Spooner and William
L. Hanaway, 1–69. Philadelphia, PA: University of Pennsylvania.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

 315

Sprenger, Aloys. 1854. A Catalogue of the Arabic, Persian and Hindustani

Manuscripts of the Libraries of the King of Oudh. Calcutta: Baptist Mission Press.
Starn, Randolph. 2002. “The Early Modern Muddle.” Journal of Early Modern
History 6 (3): 296–307. http://doi.org/10.1163/157006502X00167.
Steele, Timothy. 1990. Missing Measures: Modern Poetry and the Revolt against
Meter. Fayetteville, AR: University of Arkansas Press.
Steiner, George. 2001. Grammars of Creation. New Haven, CT: Yale University Press.
Steingass, Francis Joseph. 1892. A Comprehensive Persian-English Dictionary,
Including the Arabic Words and Phrases to Be Met with in Persian Literature.
London: Routledge and Kegan Paul. http://dsal.uchicago.edu/dictionaries/steingass.
Stern, Ralph. 2003. “Winckelmann, Piranesi and the Graeco-Roman Controversies:
A Late Exchange in the Querelle des ancienes et des modernes.” Architectura 33:
63–94.
Storey, C. A. 1953. Persian Literature: A Bio-Bibliographical Survey. Vol. 1, Pt. 2.
London: Luzac.
Storey, C. A. 1984. Persian Literature: A Bio-Bibliographical Survey. Vol. 3, Pt. 1.
London: Luzac.
Stray, Christopher. 1998. Classics Transformed: Schools, Universities, and Society in
England, 1830–1960. Oxford: Clarendon.
Subrahmanyam, Sanjay. 1992. “Iranians Abroad: Intra-Asian Elite Migration and
Early Modern State Formation.” Journal of Asian Studies 51 (2): 340–63. www.
jstor.com/stable/2058032.
Subrahmanyam, Sanjay. 1997. “Connected Histories: Notes towards a
Reconfiguration of Early Modern Eurasia.” Modern Asian Studies 31 (3): 735–62.
www.jstor.com/stable/312798.
Subrahmanyam, Sanjay. 1998. “Hearing Voices: Vignettes of Early Modernity in
South Asia, 1400–1750.” Daedalus 127 (3): 75–104. www.jstor.com/stable/
20027508.
Subrahmanyam, Sanjay. 2001. Penumbral Visions: Making Polities in Early Modern
South India. Ann Arbor, MI: University of Michigan Press.
Subtelny, Maria Eva. 1986. “A Taste for the Intricate: The Persian Poetry of the Late
Timurid Period.” Zeitschrift der Deutschen Morgenländischen Gesellschaft 136:
56–79.
Suleiman, Yasir. 1999. The Arabic Grammatical Tradition: A Study in Taʿlīl. Edinburgh:
Edinburgh University Press.
Swann, Joan, Ana Deumert, Theresa Lillis, and Rajend Mesthrie. 2004. A Dictionary
of Sociolinguistics. Tuscaloosa, AL: University of Alabama.
Syed, Muhammad Aslam. 2012. “How Could Urdu Be the Envy of Persian (Rashk-i-
Farsi): The Role of Persian in South Asian Culture and Literature.” In Literacy in
the Persianate World: Writing and the Social Order, edited by Brian Spooner and
William L. Hanaway, 279–310. Philadelphia, PA: University of Pennsylvania.
Tabari. 1987. The Commentary on the Qurʾān. Translated by John Cooper. 5 vols.
Oxford: Oxford University Press.
Tabor, Nathan Lee Marsh. 2019. “Heartless Acts: Literary Competition and
Multilingual Association at a Graveside Gathering in Eighteenth-Century Delhi.”
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

316 

Comparative Studies of South Asia, Africa and the Middle East 39 (1): 82–95.
www.doi.org/10.1215/1089201X-7493799.
Talbot, Cynthia. 1995. “Inscribing the Other, Inscribing the Self: Hindu-Muslim
Identities in Pre-Colonial India.” Comparative Studies in Society and History
37 (4): 692–722. www.jstor.com/stable/179206.
Tavakoli-Targhi, Mohamad. 2001. Refashioning Iran: Orientalism, Occidentalism
and Historiography. New York: Palgrave.
Tavakoli-Targhi, Mohamad. 2011. “Early Persianate Modernity.” In Forms of
Knowledge in Early Modern Asia, edited by Sheldon Pollock, 257–88. Durham,
NC: Duke University Press.
Tavoni, Mirko. 1994. Renaissance and Early Modern Linguistics. History of
Linguistics 3. Edited by Giulio Lepschy. London: Longman.
Taylor, Charles. 1985. Philosophy and the Human Sciences. Cambridge: Cambridge
University Press.
Teissier, Beatrice. 2009. “Texts from the Persian in Late Eighteenth-Century India
and Britain: Culture or Construct?” Iran 47 (1): 133–47. https://doi.org/10.1080/
05786967.2009.11864764.
Thackston, W. M., ed. and trans. 1993. Bâburnâma. Cambridge, MA: Harvard
University Press.
Tikku, G. L. 1971. Persian Poetry in Kashmir, 1339–1846: An Introduction. Berkeley,
CA: University of California Press.
Tilly, Charles. 1975. “Reflections on the History of European State-Making.” In The
Formation of National States in Western Europe, edited by Charles Tilly, 3–83.
Princeton, NJ: Princeton University Press.
Timpanaro, Sebastiano. 2005. The Genesis of Lachmann’s Method. Translated by
Glenn W. Most. Chicago, IL: University of Chicago Press.
Toorawa, Shawkat M. 2008. “The Shifāʾ al-ʿAlīl of Āzād Bilgrāmī (d. 1200/1786):
Introducing an Eighteenth-Century Indian Work on Al-Mutanabbī’s Poetry.”
Middle Eastern Literatures 11 (2): 246–64.
Toorawa, Shawkat M. 2009. “Āzād Bilgrāmī.” In Essays in Arabic Literary Biography,
1350–1850, edited by Joseph E. Lowry and Devin J. Stewart, 91–7. Wiesbaden:
Harrassowitz Verlag.
Trivellato, Francesca. 2009. The Familiarity of Strangers: The Sephardic Diaspora,
Livorno, and Cross-Cultural Trade in the Early Modern Period. New Haven, CT:
Yale University Press.
Troupeau, G. 2012. “Nahw.”: In Encyclopaedia of Islam. Edited by P. Bearman, Th.
Bianquis, C. E. Bosworth, E. van Donzel, and W. P. Heinrichs. 2nd ed. http://dx.
doi.org/10.1163/1573-3912_islam_COM_0838.
Truschke, Audrey. 2012a. Cosmopolitan Encounters: Sanskrit and Persian at the
Mughal Court. Ph.D. thesis, Department of Middle Eastern, South Asian and
African Studies, Columbia University.
Truschke, Audrey. 2012b. “Defining the Other: An Intellectual History of Sanskrit
Lexicons and Grammars of Persian.” Journal of Indian Philosophy 40: 635–68.
https://doi.org/10.1007/s10781-012-9163-2.
 OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

 317

Turner, James. 2014. Philology: The Forgotten Origins of the Modern Humanities.
Princeton, NJ: Princeton University Press.
Vajpeyi, Ambikaprasad. 1936. Persian Influence on Hindi. Calcutta: Calcutta
University Press.
Van Gelder, G. J. 2009. “Traditional Literary Theory: The Arabic Background.” In
General Introduction to Persian Literature, ed. J. T. P. De Bruijn, 123–38. Vol. 1 of
A History of Persian Literature. London: I. B. Tauris.
Van Lit, L. W. C. (Eric). 2017. “Commentary and Commentary Tradition.” MIDÉO
32: 3–26. http://journals.openedition.org/mideo/1580.
Vance, Norman. 1997. The Victorians and Ancient Rome. Oxford: Blackwell.
Versteegh, C. H. M. 1980. “The Origin of the Term ‘Qiyās’ in Arabic Grammar.”
Journal of Arabic Linguistics 4: 7–30.
Versteegh, C. H. M. 1983. “Arabic Grammar and Corruption of Speech.” Al-Abhath
31: 139–60.
Versteegh, C. H. M. 1993. Arabic Grammar and Qurʾānic Exegesis in Early Islam.
Leiden: Brill.
Versteegh, Kees. 1997. Landmarks in Linguistic Thought III: The Arabic Linguistic
Tradition. London: Routledge.
Walker, Paul E. 1991. “The Doctrine of Metempsychosis in Islam.” In Islamic Studies
Presented to Charles J. Adams, edited by Wael B. Hallaq and Donald P. Little,
219–38. Leiden: E. J. Brill.
Waquet, Françoise. 2001. Latin or the Empire of a Sign: From the Sixteenth to the
Twentieth Centuries. Translated by John Howe. London: Verso.
Wickham, Chris. 2009. The Inheritance of Rome: A History of Europe from 400 to
1000. London: Allen Lane.
Wilamowitz-Moellendorff, Ulrich von. 1982. History of Classical Scholarship.
Translated by Hugh Lloyd-Jones. Baltimore, MD: Johns Hopkins University Press.
Windfuhr, Gernot. 1974. “A Linguist’s Criticism of Persian Literature.” In Neue
Methodologie in der Iranistik, edited by Richard N. Frye, 331–52. Wiesbaden:
Otto Harrassowitz.
Windfuhr, Gernot, ed. 2009. The Iranian Languages. London: Routledge.
Wink, André. 1986. Land and Sovereignty in India: Agrarian Society and Politics
under the Eighteenth Century Maratha Svarājya. New York: Cambridge
University Press.
Yarshater, Ehsan. 1986. “Persian Poetry in the Timurid and Safavid Periods.” In The
Cambridge History of Iran, edited by Peter Jackson and Lawrence Lockhart, 6:
965–94. Cambridge: Cambridge University Press.
Yarshater, Ehsan. 1988. “The Indian or Safavid Style: Progress or Decline?” In
Persian Literature, edited by Ehsan Yarshater, 249–88. New York: Columbia
University Press.
Yildiz, Yasemin. 2012. Beyond the Mother Tongue: The Postmonolingual Condition.
New York: Fordham University Press.
Yohannan, Abraham, and A. V. Williams Jackson. 1907. “Some Persian References
to Zoroaster and His Religion.” Journal of the American Oriental Society 28: 183–8.
www.doi.org/10.2307/592770.
OUP CORRECTED PROOF – FINAL, 19/2/2022, SPi

318 

: Mushāʿirah. New Delhi: Shan-e Hind Publications.

Zaidi, Ali Jawad. 1989. Tārīkh-i
Zaidi, Ali Jawad. 1993. A History of Urdu Literature. New Delhi: Sahitya Academy.
Zaidi, ʿIraq Raza. 2004. “Wat̤ an Parast Khan-i
: Ārzū kī Is: lāhīn.”
: In Sirāj al-Dīn ʿAli
:
Khān Ārzū: Ek Mut̤ ālaʿah, edited by Shahid Mahuli, 98–107. New Delhi: Ghalib
Institute.
Zarucchi, Jeanne Morgan. 1989. Introduction to Charles Perrault: Memoirs of my
Life, by Charles Perrault, 1–25. Columbia, MO: University of Missouri Press.
Ziauddin, M., ed. 1935. A Grammar of the Braj Bhakha (Tuhfat : al-Hind). Calcutta:
Visva-Bharati Bookshop.
Ziauddin, Muhammad. 2005. “Role of Persians at the Mughal Court: A Historical
Study, during 1526 AD to 1707 AD.” Ph.D. thesis, Arab Study Centre, University
of Balochistan.
Zipoli, Riccardo. 1984. Chirā Sabk-i Hindī dar Dunyā-yi Ġharb Sabk-i Bārūk
: wāndah
Kh : Mīshawad? Tehran: Instituto Italiano di Cultura.
Zipoli, Riccardo. 1993. The Technique of the Ğawāb: Replies by Nawāʾī to Hāfi
z
and
Ğāmī. Venice: Cafoscarina.
Zon, Bennett. 2006. “From ‘Very Acute and Plausible’ to ‘Curiously Misinterpreted’:
Sir William Jones’s ‘On the Musical Modes of the Hindus’ (1792) and Its
Reception in Later Musical Treatises.” In Romantic Representations of British
India, edited by Michael J. Franklin, 197–219. New York: Routledge.
 OUP CORRECTED PROOFS – FINAL, 24/2/2022, SPi

Index

Āb-i Hayāt,

see Āzād, Muhammad
: Husain

cosmopolitan versus vernacular language,
Ancients and Moderns 40, 98, 101, 110, see cosmopolis; vernacular
116, 118–19, 126, 128, 143, 243–4
see also Quarrel of the Ancients :
Dād-i Sukhan 32, 115, 116–24, 130, 212,
and Moderns 227, 230
Arabic philology 55–7, 59–62, 72–3, 90 dictionaries, see lexicography
Ārzū, Sirāj al-Dīn ʿAlī Ḳhān: :
Dihkhudā, ʿAlī-Akbar 17, 184, 186
ancestors 19–20, 35
birth 18–19 early modernity 54, 93, 238–40, 242–4
education 21
in Delhi 22 Farhang-i Jahāngīrī 62–4, 79, 119, 145–9,
in Lucknow 22–3 246
patrons 35–7 Fresh Speech, see tāzah-goʾī
Urdu, role in 200–1, 212–13
works 45–52 Gadamer, Hans-Georg 9–11, 106, 136
Ātashkadah, see Āżar, Lut̤ f ʿAlī Beg Ġharāʾib al-Luġhāt 165–6
ʿAt̤ īyah-i Kubra̍ 45–6, 48, 91 Gwalior 18–19, 21, 214, 220
Āzād Bilgrāmī, Mīr Ġhulām ʿAlī 32, 122,
124, 217–19 Hātim, Z̤ uhūr al-Dīn (Shāh Hātim) 197,
Āzād, Muhammad


: Husain

193–8, 210, 245 201–3, 228
Āżar, Lut̤ f ʿAlī Beg 96, 106–9 Hazīn, Shaikh: Muhammad
: ʿAlī 29, 33,


38–42, 51, 95, 99–100, 110, 118,
Bahār-i ʿAjam, see Bahār, Tek

Chand 125–9, 155
:
Bahār, Muhammad Taqī 103–4, hindī xiii, 15–16, 21, 48, 57, 79, 85,
108, 129 139–40, 144, 159, 163, 165, 190, 198,
Bahār, Tek

Chand 23–4, 27–9, 211, 214, 221, 225, 233, 264
154–6, 181 hindī-yi kitābī 81, 83
Bedil, Mīrzā ʿAbd al-Qādir 22, 24, 25, 27, Hindi, Modern Standard xiii, 48, 202, 273,
29–31, 122, 199–200 275
Blochmann, Henry 17, 145, 186
: 214–16, 220, 228
Braj Bhāsā īhām 92–3, 196–7
Indian style, see sabk-i hindī
Chirāġh-i Hidāyat 44, 47–8, 87, 134,
151–4, 179–80, 183, 184 Jones, Sir William 16, 245–9
classical 8–9, 242,
colonial historiography 3, 11–12, 39, Karnāmah, see Munīr Lahorī, Abū
187–8, 199–200, 210–11, 242–3, al-Barakāt
249–50, 275–6 Ḳhiyābān-i Gulistān 27, 47, 87, 125
commentary (as a philological tool) 124–6 Ḳhusrau, Amīr 69, 97, 99, 106–7, 110,
cosmopolis 2, 6–9, 69, 129–33, 136, 258, 115, 120, 134, 138–9, 196, 219, 259,
267, 281–2 271–2
OUP CORRECTED PROOFS – FINAL, 24/2/2022, SPi

320 

Ḳhẉushgo, Bindrāban Dās 18–19, 34, philology 4–5, 54–6, 166, 237, 276
131–2 see also Arabic philology
plagiarism 41, 100, 111, 128
Latin 7, 8, 113–14, 208, 232, 235, 242, 256,
257–8, 267, 276 :
Qatīl, Muhammad Hasan

138–9
lexicography: Quarrel of the Ancients and the
as source for intellectual history 47–8, Moderns 113–4, 135, 250–5
142, 147
hindī 144, 160, 161–76 :
rekhtah 163, 187–9, 191, 204–6, 211,
Persian in South Asia 63–4, 143–4, 220–1, 224, 233, 259, 269
145–6 rozmarrah 119, 134, 225–6, 228, 230
see also Chirāġh-i Hidāyat;
Luġhatnāmah; Mirʾāt al-Ist̤: ilāh;
: :
S̄abāt, Mīr Muhammad ʿAz̤ īm 41, 100
Mus: t̤ alahāt
: al-Shuʿarā; Nawādir sabk-i hindī 30, 95, 101–5, 108, 129, 140
al-Alfāz̤ ; Sirāj al-Luġhat sanad 94, 111–13, 118, 123, 126, 152, 158,
linguistics, see philology 223, 260
Linguistic Survey of India 260–1 Sanskrit, see hindī-yi kitābī
Luġhatnāmah, see Dihkhudā : sariqat, see plagiarism
:
Sarkhẉush, :
Mīrzā Muhammad Afzal

31,
Majmaʿ al-Nafāʾis 18, 22, 25, 28, 30, 32–33, 34
32, 49–50, 139, 203, 206–8 Sirāj al-Luġhat 28, 44, 47, 87, 149–51, 171,
Mauhibat-i ʿUz̤ ma̍ 48, 90 246–7
metaphorical language 59–61, 88–93, 195, Sirāj-i Munīr 47, 115–16
218–19 Skinner, Quentin 11–12, 114
see also īhām
Mirʾāt al-Ist̤: ilāh: 27, 122, 157–61, 177, 181 Tanbīh al-Ġhāfilīn 40, 43, 99–100,
Mīr, Mīr Muhammad : Taqī 23, 35, 203–6 115–17, 118, 128, 139
see also Nikāt al-Shuʿarā :
tasarrufāt 77, 118, 133, 135, 212
: s,
Mukhli : Ānand Rām 15–16, 23, 25–7, 78, tawāfuq al-lisānain 58, 85–8, 165, 169
121–2 tāzah-goʾī 44, 94–5, 104–5
see also Mirʾāt al-Is:t̤ ilāh: tażkirah xiii–xiv, 17, 24–5, 32, 64, 110,
multilingualism 3–4, 211–12, 236, 262, 159, 190, 194, 204
263–5 : al-Hind 167–8, 193, 219, 270
Tuhfat
Munīr Lahorī, Abū al-Barakāt 95, 96–9,
101, 114–15, 153 Ummīd, Qizilbāsh Ḳhān 33, 40, 129,
Mus̄mir 48–9, 56, 93, 150–1, 174, 242 139–40, 158, 206–7
: alahāt
Must̤ : al-Shuʿarā 43, 45, 156–7 Urdu, scholars’ definitions of 69, 174,
187–8, 192, 198, 202, 209–10, 222–3
Nawādir al-Alfāz̤ 48, 162–6, 179–80 :
see also hindī; rekhtah
Nikāt al-Shuʿarā 92, 190–3, 200, 204, 230
vernacular 136, 187–8, 208–9, 232, 233,
Orientalism 17, 58, 124, 183, 251 241–2, 256–7, 268, 271
see also colonial historiography see also cosmopolis

Pārs (mythical king) 64–6 Wārastah, Siyālkot:ī Mal 42–5

Persian dialects (as described in see also Mus:t̤ alahāt
: al-Shuʿarā
pre-modern sources) 66–9, 146–7 :
Walī, Walī Muhammad 190, 192
see also Persian, regional usage in Wālih Dāġhistānī, ʿAlī Qulī Ḳhān 38, 127
Persian, regional usage in 66, 74–6, 130–1,
153–4 zabān-i hindī, see hindī