Classical and Byzantine Astrology in Sassanian Persia

Author(s): David Pingree

Source: Dumbarton Oaks Papers, Vol. 43 (1989), pp. 227-239
Published by: Dumbarton Oaks, Trustees for Harvard University
Stable URL: http://www.jstor.org/stable/1291610 .
Accessed: 26/11/2013 16:22

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp

.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact [email protected].

Dumbarton Oaks, Trustees for Harvard University is collaborating with JSTOR to digitize, preserve and
extend access to Dumbarton Oaks Papers.

http://www.jstor.org

This content downloaded from 128.206.9.138 on Tue, 26 Nov 2013 16:22:13 PM

All use subject to JSTOR Terms and Conditions
 Classical and Byzantine Astrology in Sassanian Persia
DAVID PINGREE

notorious ninth-century astrologer from circles, revolving back and forth between the
TheBalkh, Abfi Macshar, once proclaimed to his peoples whom he mentions. In this paper I intend
curious student, Abfi SacId Shadhan, that: "The to explore some new evidence relevant to the ques-
Chaldaeans were the first to write about the stars, tion of the extent of scientific intercourse among
their measurements, and observations of them; Indians, Greeks, Persians, Byzantines, Syrians, Ar-
they knew their courses in nature [i.e., astronomy] abs, and Western barbarians. In the course of this
and in prorogations from base-nativities [i.e., as- investigation I hope to exemplify how, in the area
trology]. After them were the Indians, then the of astrology as well as in other domains, the medie-
Syrians, and then the Arabs."1 This two-sentence val tradition has surprising things to reveal to us
history of ancient science, along with the rest of about our classical Greek heritage, and inciden-
Shadhan's Mudhakarat, was being read in Byzan- tally to demonstrate that Paul Lemerle, in his emi-
tium in the eleventh century,2 and in 1260 or nently intelligent Le premierhumanismebyzantin,did
thereabouts the Greek version was made available not do full justice to the Orient in trying to deter-
to Western scholars in Latin dress provided, it mine the origins of the revival of Byzantine schol-
seems most likely, by Stephanus of Messina.3 arship in the ninth century.4
Thereby the history of Shadhan's own text illus- Abfi Macshar, renowned as he was as an astrolo-
trates nicely Aba Macshar's main point, that astron- ger, did not foresee that, after the Arabs, the bar-
omy and astrology were transmitted from culture barians of the Latin West would appear as experts
to culture in the ancient and medieval periods. In in the science of the stars. It is, however, in their
fact, what is wrong about Abfi Macshar's history is tongue-or rather, in a most convoluted distortion
that it is too simple, representing the transmission of Latin-that was written the text that is the key
as being linear when in fact the celestial sciences to our inquiry, a translation from the Arabic exe-
were constantly being transmitted in appropriate cuted by Hugo of Santalla for Michael, the bishop
of Tarazona in northern Spain from 1119 to 1151.
It was Harvard's great medievalist, Charles Homer
'This is my translation of the unpublished Arabic original. Haskins, who first drew attention to this work in
The abbreviated Byzantine version-chapter 32 of Book II of
the MvoTripta of 'AnoRdoap-was published in Catalogus Codi- 1911 ;5 it attracted the interest of Charles Burnett,
cum AstrologorumGraecorum(hereafter CCAG), 5,1, p. 148.
my learned Warburgian friend to whom much of
2One of the two complete Greek manuscripts is Vat. gr. 1056,
wherein the Mudhakardt occupy fols. 194-221. Though this
the research for this paper is due, and myself some
manuscript itself was copied in the 14th century, its source was seventy years later by its strange title. For it is
a collection of texts put together in the 12th century; see D. called the LiberAristotelisde ducentisquinquagintaqu-
Pingree in DOP 18 (1964), 138. But the translation into Greek
seems to have been made simultaneously with the Byzantine inque Indorum voluminibus universalium questionum
translation of Abfi Ma'shar's Kitab ft ahkam tahawil sini al- tamgenetialiumquam circulariumsummamcontinens-
mawMlid,and that translation can be dated ca. 1000; see D. Pin- The Book of AristotleContaining the Sum of Universal
gree, AlbumasarisDe revolutionibusnativitatum (Leipzig, 1968),
VIII.
3The Latin translation was evidently made by the Latin trans-
4See P. Lemerle, Le premier humanismebyzantin(Paris, 1971),
lator of Abfi Macshar'sDe revolutionibusnativitatum, who seems especially ch. 2, "L'hypothese du relais syro-arabe," 22-42.
to have been Stephanus of Messina working in 1262; see Albu- 5C. H. Haskins, "The Translations of Hugo Sanctallensis,"
masaris De revolutionibusnativitatum, vi. Stephanus used aphor- The Romanic Review 2 (1911), 1-15. I have used the revised ver-
isms from the Mudhakarat in the Hermetis Centiloquiumthat he sion found as ch. 4 in his Studiesin the Historyof Mediaeval Science
composed in 1262. (repr. New York, 1960), 67-81; see especially 74-76.

This content downloaded from 128.206.9.138 on Tue, 26 Nov 2013 16:22:13 PM

All use subject to JSTOR Terms and Conditions
228 DAVID PINGREE

Questions, both Genethlialogical and Revolutionary, reading for Euctemon), Erato (who may be
Drawnfrom the 255 Volumesof the Indians.6 Aratus), Antiochus, Welis Egiptius (that is, Vettius
This is a bulky treatise whose popularity can be Valens considered, in the Sassanian-Arab fashion,
gauged by the facts that no manuscript of the Ar- as an Egyptian from Alexandria rather than a na-
abic original is known to exist, and that of Hugo's tive of Antioch), Alwelistus (apparently Erasistra-
Latin version, aside from the basic, but now incom- tus), and the Babylonians. Next is named the Be-
plete, manuscript, Digby 159 in the Bodleian Li- fida, a "codex mirabilis" of the Indians; Befida
brary, which was copied in England in the thir- seems to be an attempt to render Bizidaj, the Ara-
teenth century, there is just one other witness to bic form of the Pahlavi Wizddak,meaning "The
the text, Savile 15, also in the Bodleian, which was Chosen"-a title reflecting that of Valens' work,
copied from the Digby manuscript in the fifteenth the 'AvOokoy(at. After this, Hugo names another
century when the latter was still complete. Both work with a title originating in Sassanian Iran; his
manuscripts, it appears, were on the shelves of the Xaziur represents the last part of the Pahlavi ZMk-i
great library assembled at Mortlake by the cele- Shahriyaran, the Royal Astronomical Tables. Hugo
brated John Dee, to whom we owe the survival of ends his catalogue with a reference to two other
this as of so many other ancient and medieval precious but here untitled volumes on nativities
scientific texts, including the Greek Anthologiesof and on interrogations that have been preserved by
Vettius Valens which will play a central role in our the Indians-surely from Alexander's pillaging of
story. the library at Istakhr lamented by the author of
From the pen of Hugo we have some ten works the Denkart; it is the volume on nativities, he im-
on astronomy, astrology, and divination translated plies, that is the basis of the LiberAristotelis.In the
from Arabic into Latin. In the preface to one of catalogue he has mentioned only 125 rather than
these translations-that of Ibn al-Muthanna's com- 255 books, but still the identification seems reason-
mentary on al-Khwarizmi's Z`j al-Sindhind-Hugo able; in texts like this, what difference could 130
states that Bishop Michael found the Arabic man- extra sources make?
uscript of this treatise "in Rotensi armario et inter The identification, in any case, is confirmed by a
secretiora bibliotece penetralia."'7As Haskins long most remarkable document found toward the end
ago suggested,8 this library at Rota was probably in of a fourteenth-century Byzantine manuscript, Va-
the Muslim stronghold of that name, now called ticanus graecus 1056, a magnificent codex whose
Rueda Jal6n. This fortress was ceded to Alfonso core is an astrological compendium of the Com-
VII by Sayf al-Dawla, the last of the Banfi Had, in nenian period, and whose margins are filled with
1140-1141.9 Assuming that the Arabic manuscript rare Greek translations from the Arabic." On folio
of the LiberAristoteliswas found in the same "secre- 242 of this manuscript is a catalogue of astrological
tiora bibliotece penetralia," Hugo must have made authorities that turns out to be but a shortened
his Latin version of it in the decade between 1141 and slightly variant version of the bibliography
and 1151. found in Hugo's Liber Aristotelis. The Byzantine
About halfway through his tortuous prologue to text ascribes this bibliography to Maodka-that is
the Liber Aristotelis, following the fulsome praise to say, to Mashd'allah ibn Atharli, a Persian Jew
dutifully bestowed upon his patron bishop, Hugo from Basra who participated in casting the horo-
rather abruptly launches into a bibliography of as- scope of Baghdad for its founder, Caliph al-
trology,'0 in which are described and enumerated Mansfir, on 30 July 762, and who lived on till about
the varied treatises composed by Saraphies (who is 810 since his astrological history, in which he pre-
Serapion of Alexandria), Aristotle, Hermes, Ptol- dicted the imminent downfall of al-Mansfir's dy-
emy, Doronius (that is, Dorotheus of Sidon), De- nasty, the cAbbasids, was completed shortly before
mocritus, Plato, Victimenus (undoubtedly a mis- al-Ma'min became caliph in 813.12
The Greek version of the bibliography con-
6An edition with commentary by Prof. Burnett and myself cludes with the statement, missing in Hugo's Latin,
will appear shortly.
SSee E. MillaisVendrell, El comentariode Ibn al-Mutannd' a las
TablasAstron6micasde al-Jwdrizmi(Madrid-Barcelona, 1963), 95.
IIThe relevant texts from this manuscript, which also con-
8Haskins, 70-71. tains Shadhan's Mudhakardt,will be published elsewhere.
9R. Dozy, Histoire des musulmans d'Espagne, 3 vols. (Leiden, '2Concerning Masha'allah see the articles by D. Pingree in the
1932), III, 154 note 1. Dictionary of Scientific Biography, 9 (New York, 1974), 159-62,
'oThis bibliography will be dealt with in detail in the forth- and by F. Sezgin in his Geschichtedes arabischenSchrifttums,VII
coming edition of the LiberAristotelis. (Leiden, 1979), 102-8 (hereafter GAS).

This content downloaded from 128.206.9.138 on Tue, 26 Nov 2013 16:22:13 PM

All use subject to JSTOR Terms and Conditions
 CLASSICAL AND BYZANTINE ASTROLOGY 229

that the author, Masha'allah, has compiled tPv prose paraphrases of it were available until the
agoQf)oav 3q3Xkov &no tOv OAOvtmyv early eleventh century in Byzantium as it was used
Vaticanus
I3LPLCcv the
ovvoTLxWg v too0aoL k6yotg. graecus by astrologer who cast the horoscope of Em-
1056 unfortunately does not preserve those four peror Constantine VII Porphyrogenitus shortly
books, but the Liber Aristotelisdoes indeed contain after his birth on 3 September 905; by Demophi-
four sections, separated by summaries in the text lus, who is known to have been active in 989; and
rather than by formal titles. These four sections by an astrologer who cast an anniversary horo-
are, respectively:on difficult points of astronomy; scope for 13 October 1011. But most of what we
a collection of astrologicaldefinitions; on geneth- have in Greek of Dorotheus' poem is preserved in
lialogy, organized according to the twelve places of excerpts and in prose paraphrases in the 'Anote-
the and on two methods of contin- XEo0LaTLX6 of Hephaestio of Thebes, a gentleman
uous&bEXtoronog;
horoscopy, the revolution of the years of na- whom his mother conceived on 20 February 380
tivities and the periods and sub-periods of the na- and gave birth to on 26 November of that same
tive's life assigned to each of the planets. It seems year.
plausible to hypothesize that the lost Arabic origi- However, we know from a passage of the lost Ki-
nal of Hugo's Liber Aristotelis was the work of tab al-Nahmatan of Ibn Nawbakht quoted by Ibn al-
Masha~'allahto which the bibliography served as a Nadim that Dorotheus' hexameters were trans-
preface, perhaps his al-Kitab al-murdi (The Pleasing lated into Pahlavi under Ardashir I, the Sassanian
Book) that is mentioned by the greatest of Arabic emperor from about 222 to 237, or under his son,
bibliographers, Ibn al-Nadim. The association of Shapir I, who ruled from about 237 to 267; and
the Latin text with Mqsha'allqh is heightened by that it was later elaborated during the reign of
the fact that immediately preceding the LiberAris- Khusro Antishirwan, between 531 and 578. The
totelis in Savile 15 is the unique copy of Hugo's principal Arabic translation of the lost Pahlavi ver-
translation of Misha'allah's Kitab al-mawalid al- sion of Dorotheus was apparently due to the ef-
kabir.3 This work also is not extant in Arabic, but forts of CUmaribn al-Farrukhan al-Tabari in about
was composed by the Basran in the late 780s or in the year 800. This translation proves that Ibn Naw-
the 790s since there are included in it horoscopes bakht was correct; for it contains not only a num-
that can be dated 2 October 770, 7 February 771, ber of Arabic transliterations of Pahlavi technical
and 17 July 784. We shall return to this text pres- terms, but two horoscopes that were added to the
ently. text as illustrations by Sassanian scholars. One is
But in order to test the hypothesis that the Liber datable 20 October 281, the other 26 February
Aristotelisis, in fact, the translation of a lost opus of 381.
Masha~'allah, it seems useful to compare the au- Dorotheus' own examples are eight horoscopes
thorities used by the Arab astrologer in his known cast for natives born between 7 B.C. and A.D. 43.
works with those cited or used in Hugo's text. Most Three of these together with other Dorothean ma-
of these authorities are Greek, but some are Per- terial were incorporated by Mashd'allah into his
sians of the Sassanian period. I shall discuss them shorter Kitab al-mawalid, which also survives only
briefly in chronological order; for they brilliantly in a Latin translation,15 while a long section on lots
illuminate the shady history of ancient and early in Masha~'allah'sKitab al-mawalid al-kabirin Hugo's
medieval astrology. translation is attributed to Dorotheus. This pas-
We begin with Dorotheus of Sidon, who com- sage, however, corresponds only in part to the dis-
posed in about A.D. 75 a poem in leaden hexame- cussions of the various lots scattered throughout
ters setting forth in five books both genethlialogi- CUmar's translation of Dorotheus; from this and
cal and catarchic astrology.14 The original poem other circumstances it is clear that Mashc'allah did
survived until at least the seventh century, and not use CUmar'stranslation-which was probably,
in any case, made only after the Basran had com-
131 am currently preparing an edition with commentary of
this work.
posed his works on genethlialogy-but rather he
14The Arabic translation of the Pahlavi version made read the Pahlavi original, in a fuller form than ap-
by
'Umar ibn al-Farrukhan together with the genuine fragments pears in cUmar's text and one that was accompa-
in Greek and Latin was edited by D. Pingree, Dorothei Sidonii
CarmenAstrologicum(Leipzig, 1976). This volume contains dis-
cussions of all of the material referred to in this and in the suc- '5Edited by D. Pingree in E. S. Kennedy and D. Pingree, The
ceeding paragraph. The Apotelesmaticaof Hephaestio was edited Astrological History of Mash'allah (Cambridge, Mass., 1971),
by D. Pingree, 2 vols. (Leipzig, 1973-74). 145-74.

This content downloaded from 128.206.9.138 on Tue, 26 Nov 2013 16:22:13 PM

All use subject to JSTOR Terms and Conditions
230 DAVID PINGREE

nied by a commentary. Confirmation of the exis- nativities was a Pahlavi version of Dorotheus with
tence of this fuller PahlavI text of Dorotheus is to a commentary. It helps to connect Hugo's LiberAr-
be found in an Arabic manuscript now in the li- istotelis with Masha'allah, then, that a substantial
brary of the University of Leiden, Oriental 891, portion of the first two books of the Latin text is
which contains on folios 1-28 a Kitab ft bayan al- taken from a commentary on Book III of Doro-
ifradat ascribed to Dorotheus.16 Parts of this work theus. Not only that; long passages in Books II and
are derived from Books II and V of Dorotheus' III of the LiberAristotelisare versions, often variant
poem, but this version is different from and more versions, of the Dorotheus we know from CUmar,
expansive than that of CUmar. Embedded in the replete with Pahlavi technical terms transliterated,
Leiden Dorotheus are eleven horoscopic examples through the Arabic, into Latin. This material in-
that can be dated between 13 June 765 and 17 cludes the horoscope of 20 October 281 that had
June 768, a period when Mashd'allah was active, been inserted into the Pahlavi version of Doro-
though I can not yet offer any evidence for either theus, proving that it goes back to the same Sassan-
his authorship or his use of this text. ian text that was translated by CUmar.
What is remarkable, however, is that a number But it is also clear from fragments of Dorotheus'
of the horoscopes in the Leiden Dorotheus are poem preserved by Hephaestio, but not found in
found in a vast Byzantine compilation entitled Eto- his Arabic translation, that CUmar'stext omits Do-
xat OE[EtLov ig TIVv owrQokoylav and rotheus' chapters on professions, on army service,
ay0Oyil
falsely attributed to one Ahmad the Persian.17 This and on friendship. The first and last of these sub-
Introductionto Astrology,divided into four books, is jects are discussed in Book III of the LiberAristotelis
in fact a collection of chapters from Classical and in words that in part reflect the fragmentary hex-
Byzantine astrological texts mixed up with mate- ameters and prose paraphrases presented by He-
rial translated into Greek from the Arabic; much phaestio; this allows us to recover from Hugo some
of the latter comes from the early eleventh-century of the lost instructions of Dorotheus. But these ex-
Byzantine translation of Shadhan's Mudhdkardt. cerpts in the Liber Aristotelis,now shown to be Do-
The whole was apparently put together by Eleu- rothean, also make it possible to identify as para-
therius Elias Zebelenus,is who has inserted into the phrases of the Sidonian's poem a substantial part
second book an interpretation of his own horo- of a brief and anonymous Byzantine astrological
scope, which can be dated 10 November 1343. compendium fortuitously preserved on the four
Eleutherius, it should be noted, is responsible for folios, numbered 238 to 241, of Vaticanus graecus
another astrological compendium of a similar 1056 that immediately precede the astrological
character that he falsely ascribed to Palchus. bibliography of Masha)'allah.19This compendium,
Though some eminent historians of Byzantine as- which includes some excerpts also from the work
trology have asserted that Palchus wrote in about of Rhetorius of Egypt to which we will shortly turn,
A.D. 500, in fact his name is a transliteration of an presumably goes back to the Comnenian exemplar
Arabic al-Balkhi, which Eleuther- from which Vaticanus graecus 1056 was copied,
6volta •6vtx6v,
ius found in the Mudhdkardtwherein Abfi Macshar and perhaps to the earlier prose paraphrase of Do-
is represented as including in a list of astrological rotheus that was used by Byzantine astrologers of
authorities a tarjamdn al-Balkhi, a translator from the tenth and early eleventh centuries. Thereby we
Balkh. One of Eleutherius' sources for pseudo- have now recovered a significant part of the non-
Palchus was the compendium of Rhetorius, an- Hephaestionean paraphrase of Dorotheus in its
other Byzantine astrologer important to our inves- original Greek.
tigation. It is now time to leave Dorotheus, whose tradi-
But to return to Dorotheus. We have shown that tion has been so unexpectedly magnified by Hugo,
one of Masha'allah's sources in his two books on and to turn to Vettius Valens.2? This astrologer of
Antioch was conceived on 13 May 119 and born on
16This text, together with some other late 8th-century Arabic
8 February 120. The nine books of his 'Av0okoyat
works on astrology, I hope to publish soon. have been corruptly and incompletely preserved in
17Anedition of this enormous text is also in hand, but will
take some time to complete.
18Concerning Eleutherius and his works, see D. Pingree, hope to publish this shortly.
191I
"The Astrological School of John Abramius," DOP 25 (1971), 20D. Pingree, VettiiValentisAntiocheniAnthologiarumlibri novem
189-215, especially 202-4, and "The Horoscope of Constanti- (Leipzig, 1986). See also D. Pingree, "The Byzantine Tradition
Studien (Wies-
nople," ILQ(oataTa.Naturwissenschaftsgeschichtliche of Vettius Valens' Anthologies," Harvard Ukrainian Studies 7
baden, 1977), 305-15. (1983), 532-41.

This content downloaded from 128.206.9.138 on Tue, 26 Nov 2013 16:22:13 PM

All use subject to JSTOR Terms and Conditions
 CLASSICALAND BYZANTINE ASTROLOGY 231

a recension made in the fifth century of a version weird Latin, to 31 July 62, 26 July 114, and 8 Feb-
produced in the third. Probably contemporaneous ruary 120, which last is Valens' own nativity. Hugo,
with the third-century version was the Pahlavi of course, being ignorant of Valens' astronomical
translation of Valens' work. This Pahlavi Valens and calendaric data, could make no sense of these
was quoted in both the expanded Sassanian text of horoscopes, but his gibberish suddenly becomes
Dorotheus translated by cUmar and in the PahlavI intelligible when one imagines the Arabic and Pah-
original of the Kitab al-mawalid ascribed to Zara- lavI versions that lie between it and the original
dusht, of which we will presently have more to Greek.
say.21 And, as we know from the bibliographical Moreover, a valuable manuscript preserved in
and biographical dictionaries of Ibn al-Nadim22 the Laleli Mosque in Istanbul, no. 2122, contains
and Ibn al-Qifti,23 Valens' PahlavI Anthologieswere a short compilation on interrogations by
commented on by Buzurjmihr-not, indeed, if Masha:'allah.26 In this treatise, untitled in the man-
we follow Christensen,24 the alleged minister of uscript, Masha:'allah quotes from both Dorotheus
Khusro Anfishirwan, but a contemporary sixth- and Valens-the latter on the subjects of buying
century Sassanian scholar named Burjmihr who, land and of government, which are scarcely topics
among other intellectual feats, translated the San- in our Greek Anthologies, but both of which are
skrit Paficatantrainto PahlavI and introduced chess dealt with in the Arabic translation of a Pahlavi
(caturafiga) into Sassanian Iran-both acts with re- work on catarchic astrology ascribed to Valens, the
markable consequences for us. However, since the Kitab al-asrar, preserved on folios 3 lv-60 of man-
Arabic authorities consistently name him Bu- uscript no. 2920 of the Nuruosmaniye Mosque in
zurjmihr, we will follow their erroneous practice. Istanbul.2v Masha'allah, in his Kitab ft qiyam al-
It is Ibn Hibinta, a Christian astrologer writing in khulaf&'wa maCrifatqiydmkull malik, says that he has
Baghdad in about 950, who informs us that Bu- discussed the matter of government in his Bizda-
zurjmihr's book was entitled Bizidaj, and that it em- jit.28 So MashY'allah's collection of astrological
braced the sayings of the wise men25-i.e., that, dicta on catarchic astrology and interrogations in
though perhaps based on Valens, it was yet an- Laleli 2122 may have borne the title that Bu-
other compendium of astrological doctrines zurjmihr gave to his expansion of Vettius Valens'
gleaned from the ancient classics-though for Bu- Anthologies.
zurjmihr these would have included Iranians and Having seen that Valens is indeed a source for
Indians as well as Greeks. Masha'allah, we are not surprised to find in the
We must now consider whether or not LiberAristotelisa number of references to an astrol-
Masha'allah in his known works utilized Valens', oger named Welis. Some of these references head
either in an Arabic translation of the original passages not to be found in our corrupt and in-
Greek or in Buzurjmihr's Pahlavi version. In Hu- complete Byzantine text of the Anthologies; and
go's translation of Masha'allah's Kitab al-mawalidal- other passages that derive from the Anthologiesare
kabirwe find confirmation that he did indeed study not expressly attributed to Valens by Hugo. More
Valens' Anthologies; for into Masha'allah's work is interesting is the fact that Buzurjmihr, under the
incongruously incorporated an incomprehensible corrupt transliteration Zarmiharus, is said at the
version of chapters 21 and 22 of Book I of Valens, end of Book III of the LiberAristotelisto have been
including the horoscopes that can be dated from a source for this maqala on genethlialogy, and to
the original Greek, though not at all from Hugo's have commended the subject matter of Book IV,
that is, continuous horoscopy. We now need to re-
21 The
Zaradusht text will be published in the same volume as mind ourselves that Sa'id al-Andalusi in his Kitdb
the Leiden Dorotheus mentioned in note 16. Some notice of
this text has been given in D. Pingree, "Mdshd'alldh: Some Sas- tabaqat al-umam records that Valens' Bizidaj was
anian and Syriac Sources," Essayson IslamicPhilosophyand Science concerned with nativities and their revolutions,
(Albany, 1975), 5-14. and had an introduction to these topics.29 This de-
22 Ibn
al-Nadim, Kitab al-fihrist, ed. G. Flugel, 2 vols. (Leipzig,
1871-72), I, 269. scription precisely fits not the Greek 'Av60koyCat,
23Ibn al-Qifti, Ta'rikh al-hukamd', ed. J. Lippert (Leipzig,
1903), 261. 26This also will be published in the collection of
24A. Christensen, "La l6gende du sage Buzurjmihr," Acta Or- early Arabic
astrological texts.
ientalia 8 (1929), 81-128; the passage from Ibn al-Nadim is cited 27This as well will appear in the same collection.
on 92-93. 28Translated in The AstrologicalHistory of Mashd'alldh, 129-35;
25C. A. Nallino, "Tracce di opere greche giunti agli arabi per the reference to his Bizfdajat is on p. 131.
trafila pehlevica," A Volumeof OrientalStudiesPresentedto Professor 29S'id al-Andalusi, Kitdb tabaqatal-umam, trans. R. Blachere
E. G. Browne (Cambridge, 1922), 345-63, esp. 352. (Paris, 1935), 87.

This content downloaded from 128.206.9.138 on Tue, 26 Nov 2013 16:22:13 PM

All use subject to JSTOR Terms and Conditions
232 DAVID PINGREE

but the Latin Liber Aristotelis, in which the intro- epitome, which can be dated 24 February 601.
duction comprises Books I and II, Book III deals Since Rhetorius was already being pillaged in the
with nativities, and Book IV with the revolutions eighth century, by Theophilus of Edessa, and says
of the years of nativities. When we further remem- nothing to indicate that he is working under Arab
ber that Hugo calls the Befida (i.e., the Bizdaj) a rule, it seems most likely that he wrote at the be-
"codex mirabilis" of the Indians, we may conjec- ginning of the seventh century.
ture that much of Masha'allah's work was derived One of Rhetorius' fifth-century horoscopes is
from Burzurjmihr's PahlavI compilation. quoted by Masha'allah in his Kitdb al-mawalid,34
But Masha'allah had other sources in addition to and another, cited by him in a lost work on astro-
the Pahlavi versions of Dorotheus and Valens. The logical history, is found in the compendium that
principal one is a Greek compendium associated Eleutherius Elias attributed to Palchus and that
with the name Rhetorius, who was evidently the contains much from Rhetorius.35 This latter horo-
last astrologer of Egypt to write in Greek before scope is one of a group of three that are associated
the Arab conquest of 640.30 Rhetorius excerpted, with Emperor Zeno as are the horoscopes of Pam-
almost certainly at Alexandria, from a splendid li- prepius and of Emperor Leo's son preserved at the
brary of astrological literature that had also been end of Book V of the first epitome of Rhetorius'
available to the Neo-Platonic philosopher, Olym- work. Incidentally, MashY'allah's treatise on astro-
piodorus, when he lectured on Paul of Alexan- logical history contained five horoscopes that can
dria's Etoayoyr in the summer of 564.31 Rhetorius' be dated between 21 October 766 and 10 January
work survives only in epitomes, of which there are 768, three of which are found in Eleutherius' com-
two principal ones, both probably of the tenth cen- pilation ascribed to Ahmad the Persian; one of
tury.32 The more complete epitome occupies most these three is also found in the Leiden Doro-
of a fourteenth-century manuscript, Parisinus theus.36 Masha'allah wrote this work, then, in the
graecus 2425, while a number of chapters from it 770s, and at that time was able to read Rhetorius'
are found also in Laurentianus 28, 34, a magnifi- work-presumably in the original Greek. It is
cent codex copied in about the year 1000. The likely that he also derived from Rhetorius the
second epitome, preserved in another fourteenth- seven other fifth-century horoscopes and the one
century copy, Parisinus graecus 2506, and in Mar- of the sixth century in his Kitdb al-mawdlid.7 We
cianus graecus 335 of the fifteenth century, was shall consider later how a Persian Jewish astrologer
perhaps compiled by that late tenth-century astrol- writing in Arabic in Baghdad was able to read a
oger, Demophilus, to whom we referred in discuss- Greek astrological text. But first it will be relevant
ing the Byzantine prose paraphrase of Dorotheus. to describe the first epitome of Rhetorius' compen-
Rhetorius' date can be determined from the dium and its relationship to the Liber Aristotelis
horoscopes that he cites-about a dozen from the more fully.
fifth century, including several due to the astrolo- The epitome of Rhetorius is divided into six
ger of Emperor Zeno33 as well as one cast by the books. The first four constitute an independent re-
philosopher Eutocius for a native born in 497. cension of Ptolemy's 'AnoTXGtoatLCxd, having
Rhetorius also quotes a horoscope of the early some connections with the version in Laurentianus
sixth century; but his most revealing example is 28,34. Book V, on genethlialogy, is based largely on
one found in chapter 110 of Book V of the first Dorotheus of Sidon, Vettius Valens, Antiochus of
Athens, Porphyrius, and Paul of Alexandria.
Whole chapters of Book V, including, I stress, the
30D. Pingree, "Antiochus and Rhetorius," CPh 72 (1977),
203-23. Much of the evidence mentioned in this and in the horoscope of 601, are found Latinized in Hugo's
succeeding paragraph is discussed in detail in this article. Liber Aristotelis. Hugo, while he never mentions
31Edited under the name of Heliodorus by E. Boer (Leipzig, Rhetorius' name, from time to time refers to the
1962). Concerning Olympiodorus' authorship see J. Warnon, authors used by him-Durius or Doronius for Do-
"Le commentaire attribue a Hdliodore sur les Etoaycwyd de
Paul d'Alexandrie," Travaux de la FacultMde Philosophie et Lettres rotheus, Welis for Valens, Anteius for Antiochus,
Catholiquede Louvain 2 (1967), 197-217, and L. G.
de l'UniversitM and Marius the Roman, perhaps for Porphyrius.
Westerink, "Ein astrologisches Kolleg aus dem Jahre 564," BZ
64 (1971), 6-21. A new edition that will clarify the structure
and original contents of Olympiodorus' lectures on Paul is
under preparation. 34The AstrologicalHistory, 163 and 172-73.
32An edition of these earliest epitomes is nearing completion. 35"PoliticalHoroscopes," 139-42.
33D. Pingree, "Political Horoscopes from the Reign of Zeno," 36"PoliticalHoroscopes," 135.
DOP 30 (1976), 133-50, especially 144-50. 37The AstrologicalHistory, 155, 158-62, 164-67, and 169-74.

This content downloaded from 128.206.9.138 on Tue, 26 Nov 2013 16:22:13 PM

All use subject to JSTOR Terms and Conditions
 CLASSICAL AND BYZANTINE ASTROLOGY 233

Book VI of the epitome of Rhetorius begins with chapter 1 of the ninth maqalais devoted to al-nujiim
a group of chapters containing astronomical and al-bayabaniya or "desert stars" as described by
astrological notes and definitions culled from such Hermes. The original Arabic work from which
sources as Dorotheus, Valens, Ptolemy, Theon of Abfi Macshar draws his information survives in a
Smyrna, Serapio of Alexandria, Paul of Alexan- manuscript, now at the Chester Beatty Library in
dria, Heliodorus, Julian of Laodicea, and Eutocius. Dublin, Arab 5399, folios 206v-208v, under the
Rhetorius' example of collecting such a hodge- title Kitab asrar al-nujfim,38 of which the Latin trans-
podge of introductory information was followed in lation, entitled De iudiciis et significatione stellarum
Books I and II of the Liber Aristotelis,though I so beibenariumin nativitatibus, is attributed to Salio, a
far have been able to identify only one direct quo- canon of Padua who served as astrologer in the
tation from Rhetorius in those two books. This court of Ezzelino da Romano, who died in 1259.39
comes from the chapter that Rhetorius compiled Salio was actively translating from Hebrew and Ar-
of the opinions of Serapio, whose name Hugo has abic into Latin at Toledo in Spain in 1218.4 Since
transformed into Saraphies. At the end of Book the Arabic word "al-bayabaniya" is an adjective
VI, which unfortunately breaks off in the middle formed from the Pahlavi word, "wiyaban," mean-
of a sentence in Parisinus graecus 2425, Rhetorius ing "desert,";'the hypothesis has already been ad-
presents summaries of the contents of various as- vanced that the Kitab asrar al-nujfim was translated
trological treatises, the majority of which are now from a Sassanian source.41 This conjecture is con-
lost, accompanied by biographical notes concern- firmed by the version of this little work incorpo-
ing their authors; those authors are: Erimarabus rated into Book III of the LiberAristotelis,wherein
the Egyptian prophet, Phoredas (Bhfiridasa) the Hugo gives the names of five of the desert stars;
Indian, Odapsus the priest, Ptolemy, Paul of Al- four of these names are of easily demonstrable
exandria, Demetrius, Thrasyllus, Critodemus, Cal- Pahlavi origin. Thus Hugo has for the star that the
licrates, Balbillus, and Antiochus of Athens. Rhe- Greeks call 1tdXVg, our a Virginis, the name Ha-
torius' summaries are a most important source of cac corresponding to the Pahlavi "hoag," "ear of
information concerning the astrological literature corn"; for the Greek AMea, our a Lyrae, the name
of the first few centuries A.D. It also seems to have Kibar, a misreading of the Arabic transliteration of
been the inspiration for Masha'allah's prefatory the PahlavI "kennar," "lyre"; for the mouth of the
bibliography, though naturally the Basran has Southern Fish, our a Piscis Australis, the name
gathered information concerning a group of au- Sanduol derived from the Pahlavi Sadwas, that
thors different from those summarized by the star's Persian name; and for the Northern Crown,
Egyptian. our a Coronae, Hugo offers the choice of Sarben
Hugo's Latin translation of the compendium au- or Zarben, either one of which is a mutation of
thored, if I may now pretend that my hypothesis is "abesar,"the Pahlavi word for "crown." The fifth
correct, by Masha'allah allows us to recover some name in Hugo's text, Bariegini, represents the
of the previously lost material once included in the head of the first of the twins, that is, x Gemino-
astrological poem of Dorotheus, in the Anthologies rum; I can only guess that it represents the PahlavI
of Vettius Valens, and in the compilation of Rhe- "sar-i doganag," "the head of the twin."
torius the Egyptian, and to repair some tattered In all Hugo lists twenty-seven desert stars, of
passages in the Byzantine witnesses to those texts. which one is repeated to make twenty-eight;
It also permits us, in conjunction with various early twenty-five of these are identifiable among the
Arabic translations of the PahlavI translations of thirty stars in the Kitab asrdr al-nujfim and its Latin
Dorotheus and Valens, to reconstruct something of derivative, where they are given the same longi-
the history of the passage of Greek astrologers
through the hands of their Sassanian successors. I 38 Concerning this text see P. Kunitzsch, "Neues zum 'liber
would like now to discuss briefly some other Sas- hermetis de stellis beibeniis'," ZDMG 120 (1970), 126-30. The
sanian texts pertinent to this inquiry. Arabic text with its Latin translation will be included in the vol-
our man from Balkh, wrote a Kitdb ume of early Islamic astrological treatises.
Abfi Macshar, 39A defective edition was published some five times between
ah~km al-mawalid that survives in but one exemplar, 1484 and 1581; see F. J. Carmody, ArabicAstronomicaland Astro-
Huntington 546 in the Bodleian Library, and there logical Sciencesin Latin Translation(Berkeley, 1956), 55.
40L. Thorndike, "A Third Translation by Salio," Speculum 32
incompletely. In this work, which is a summary of
(1957), 116-17.
the views on specific points of genethlialogy enter- 4P. Kunitzsch, "Zum 'liber hermetis de stellis beibeniis',"
tained by Ptolemy, Dorotheus, and Vettius Valens, ZDMG 118 (1968), 62-74, esp. 64.

This content downloaded from 128.206.9.138 on Tue, 26 Nov 2013 16:22:13 PM

All use subject to JSTOR Terms and Conditions
234 DAVID PINGREE

tudes as they have in the LiberAristotelis.These cir- tury.42 Also relevant is the fact that Dorotheus is
cumstances prove that Mashd'alldh used the Pah- called by pseudo-Zoroaster a king of Egypt; this is
lavI version of Hermes. Indeed, Hugo attributes a misconception found also in 'Umar's translation
this section on the desert stars to Sarhacir astrolo- of the Pahlavi version of Dorotheus, who himself
gus, where the Arabic text reads Hurmus ra's al- came from Sidon, but who called Hermes, not
hukam&',"Hermes the head of the wise." It would himself, a king of Egypt. A revision of the Pahlavi
seem that Hugo's Sarhacir corresponds to the Pah- Kitab al-mawalid during Anfishirwan's reign may be
lavi "sar-i zirak," "head of the wise," words which conjectured on the basis of a horoscope in the text
Mdsha'alldh had transliterated into Arabic. that can be dated 6 October 549.43
Rhetorius also summarizes Hermes' work, But the claim that the work was "translated" by
though he does not mention the Trismegistan by Mahankard in about 637 also finds corroboration
name. His list of thirty stars includes all twenty- in the text. For the translator (al-mufassar)reports
seven found in Hugo's list, and must have coin- that the horoscope cast for the son of Adhin indi-
cided completely with the original Pahlavi list. cated the passing of the reign of Ardashir.44 This
Rhetorius' longitudes for these thirty stars are clearly refers to Ardashir III, who ruled from
greater than those given by Ptolemy in Books VII sometime in 628 till he was dethroned on 27 April
and VIII of the Almagest by 3;400; so that Rheto- 630. The horoscope, in which four planets and the
rius' source, who is presumably the astrologer who ascendent are in Libra, can be dated within that
pretended to be Hermes, wrote in about A.D. 500. brief span of time to November 629.45 As Mahan-
Rhetorius' longitudes of three stars, however, are kard claims to have been instrumental in inter-
corrupt; and for these stars as for most of the oth- preting this horoscope, he could have revised the
ers the longitudes given by Hugo and in the Arabic Kitab al-mawalideight years later. Unfortunately, he
Kitdbasrar al-nujitm agree with Rhetorius' mistaken goes on to discuss the horoscope of a child born in
values. Therefore, the common source of Rheto- Fahraj in the same year that he, Mahankard, was
rius and of the Pahlavi translation of Hermes was born. This horoscope can be dated 1 August 487,46
a corrupt Greek manuscript copied after 500.
The Pahlavi translation must have been made 42See, e.g., D. Pingree, The ThousandsofAbis Ma'shar (London,
during the sixth century, probably during the long 1968), 3-13.
43The positions of the planets in this and the next horoscopes
reign of Khusro Anfishirwan, since it certainly ex- are taken from B. Tuckerman, Planetary, Lunar, and Solar Posi-
isted by the beginning of the next century, when it tions A.D. 2 to A.D. 1649 (Philadelphia, 1964).
was used in the Pahlavi version of the Kitab al- Planets Text 6 October549
mawalid ascribed to Zaradusht. The Arabic version Saturn Capricorn Capricorn 250
of this Kitab al-mawalid that is preserved in two Jupiter Sagittarius Sagittarius 260
Mars Capricorn Capricorn 100
manuscripts-one in the Nuruosmaniye Mosque Sun Libra Libra 150
in Istanbul and the other in the Escorial-claims Venus quartile to Leo 290
that the original treatise, written in Old Persian by ascendent
Mercury Libra Libra 60
Zaradusht, was turned into Newer Persian by Md- Moon Libra Libra 12 o
hankard in, apparently, 637, the year in which Cte- Ascendent Taurus ca. 8 P.M.
siphon was captured by the Arabs. Mahankard's 44This story passed from Zaradusht into the Kitab al-bari' of
Pahlavi version, the story continues, was translated
'Ali ibn Abi al-Rijal, which was translated into Latin by Aegidius
into Arabic by Sacid ibn Khurasankhurrah for the de Thebaldis in ca. 1256. Thence it came to the notice of Nal-
lino ("Tracce,"354-55).
Ispahbadh, Sunbadh, in the time of Abfi Muslim; 45For this horoscope:
this would date the translation to the years be-
Planets Text 1 November629
tween 747 and 754.
Saturn Scorpio 2 o
A reading of this fascinating text quickly assures Jupiter Libra Libra 160
one that its original was indeed a Sassanian pro- Mars Libra Libra 50
duction. Like the other translations to which we Sun Scorpio 120
Venus Libra Libra 00
have referred, it is filled with transliterations of Mercury Libra Sagittarius 00
Pahlavi technical terms. It refers to various Indian Moon Pisces 120
Ascendent Libra 80 ca. 5 A.M.
astrological ideas; and we know from cUmar's Do-
rotheus, from the Denkart, from the Bundahiin, 46For this horoscope:
and from Ibn Nawbakht's Kitab that Planets Text 1 August 487
texts had
al-nahmat.n,
Sanskrit astronomical and astrological Saturn Sagittarius 250 Sagittarius 200
been translated into Pahlavi by the early fifth cen- Jupiter Libra 150 Libra 30

This content downloaded from 128.206.9.138 on Tue, 26 Nov 2013 16:22:13 PM

All use subject to JSTOR Terms and Conditions
 CLASSICAL AND BYZANTINE ASTROLOGY 235

with the result that Mahankard would have been of a third-century astrological text composed by
150 years old in 637. I can only hope that he this native of Harran, perhaps bearing the name
claimed contemporaneity with the nameless native of Zaradusht's pretended master, Aelius. The
of Fahraj out of either ignorance or senility; or, translation would have been made contempora-
better yet, that the horoscope was inserted into the neously with those of Dorotheus and Valens; and
text by the redactor in the time of Khusro Anfi- like them the text was revised and added to in the
shirwan who added the horoscope of 549 and the sixth century, with a final revision taking place in
references to the Pahlavi book of Hermes on the 637, shortly before the downfall of the Sassanian
desert stars. The association of the horoscope of Empire.
487 with Mahankard, on this hypothesis, would be The one Persian astrologer of the Sassanian pe-
due to the Arab translator's confusion. riod besides Buzurjmihr to whom Arabic authors
Zaradusht's Kitab al-mawalid, then, represents a frequently refer-they, of course, thought Zara-
genuine Sassanian work, and shares with the few dusht and Jamasp were much earlier-was al-
other pieces of Sassanian astrology that we have Andarzaghar, that is, the advisor (from Pahlavi
the characteristics of: an overwhelming depen- handarzgar), a scholar named Zddanfarrfikh.49 He
dency on Classical Greek astrology, of which only is said by Ibn al-Qifti and Sa'id al-Andalus! to have
the material from Dorotheus, Valens, and Hermes greatly admired the ten books of the Bizidaj of Val-
has been as yet identified; a smattering of Indian ens. The largest section from al-Andarzaghar's
concepts and technical terms; an emphasis on con- work that we have in Arabic is found in the Majmitc
tinuous horoscopy, developed in Iran from the aqdwilal-hukam&'compiled by al-Damaghamniin the
theories of Dorotheus and Valens; and an intense early twelfth century.50This is devoted to the topics
interest in political astrology, a subject banned in of continuous horoscopy: the revolution of the
the Roman Empire and therefore poorly repre- years of nativities, the chronocrators, and the pe-
sented in our Greek and Latin texts, but one that riods and sub-periods of the native's life domi-
flourished exuberantly in Sassanian Iran and in nated by each of the planets. These methods of
the Arabic, Byzantine, and other astrological tra- astrology have their roots in the Classical texts of
ditions that were influenced by Iran. Dorotheus and Valens, but the Sassanian author
The Kitab al-mawalid also shows some connec- presents a far more elaborate system that differs
tions with Harran, the center of Neo-Platonic pa- in detail from those of the extant Greek authors.
ganism in early Islam. The pretended autobiogra- Almost all of the numerous citations from
phy that Zaradusht gives toward the end of the al-Andarzaghar that can be gleaned from al-
work claims that the prophet studied astrology and
Ddmaghani correspond, when rearranged, to
magic under the wise Iliyfis in that city.47And at much of Book IV of Hugo's Liber Aristotelis; the
the very end of the Kitab al-mawalid is to be found name al-Andarzaghar, in fact, seems to lie behind
the horoscope of a man born in Harran; this hor- the name, Alafragar, of the author whose exposi-
oscope can be dated 9 April 232.48 It is tempting tion Hugo promises to follow. We can, then, by fol-
to conjecture, then, that the core of pseudo- lowing the arrangement of our Latin text, restore
Zoroaster was a Pahlavi translation from the Greek a substantial part of the Arabic translation of al-
Andarzaghar's Pahlavi treatise to its original order.
But other fragments of al-Andarzaghar's Kitab
Mars Capricorn Capricorn 1
Sun Leo 180 Leo 100
al-mawalid that are cited by Arabic astrologers cor-
Venus Cancer Cancer 00 respond to passages in Book III of the LiberAris-
Mercury Leo Cancer 290 totelis. One of these fragments is preserved in yet
Moon Gemini Gemini 230
Ascendent Leo 150 ca. 5 A.M. another work entitled Kitab al-mawdlid, this one
composed by yet another Jewish scholar, Sahl ibn
4This story was read by al-Birfini and is reported by him in
his Kitab al-athar al-baqiya; see J. Fick, "Sechs Erganzungen zu Bishr.51 Sahl had worked for Tahir ibn al-Husayn
Sachaus Ausgabe von al-Birfinis 'Chronologie Orientalischer
Vol61ker',"Documenta Islamica Inedita (Berlin, 1952), 69-98,
esp.75, last line. Venus Aries 240 Aries 200
48 For this Aries 260
horoscope: Mercury
Planets Text Moon Taurus 30 Taurus 60
9 April 232
Ascendent Virgo ca. 4 P.M.
Saturn Aries 230 Aries 190
Jupiter Aries 220 Aries 60 49Sezgin, GAS, 80-81.
Mars Aries Taurus 20 50See the
important article to be published by C. Burnett.
Sun Aries Aries 190 5 Sezgin, GAS, 125-28.

This content downloaded from 128.206.9.138 on Tue, 26 Nov 2013 16:22:13 PM

All use subject to JSTOR Terms and Conditions
236 DAVID PINGREE

when he governed Khurdsan from 820 to 822, and Hibinta compiled his Kitdb al-mughni. In MoOyvqg
later for al-Hasan ibn Sahl, who served as wazir to this bit of fakery is followed by a series of short
al-Ma'mfin, the caliph from 813 to 833. astrological dicta ascribed to the Sassanian favor-
One of Sahl's favorite authorities in all his works ites Dorotheus and Valens as well as to Pythagoras,
was Mashd'allah; in the Kitab al-mawalid, for in- but also to the nations who appear as the heirs of
stance, he quotes in extensoseveral chapters of the antediluvian astrology in the historical myth of
Basran's Kitdb al-mawalid al-kabir, allowing us an Abfi MaCshar:the Babylonians, the Egyptians, the
opportunity to compare Hugo's Latin translation Greeks, the Indians, and the Persians. Indeed,
with the Arabic original. He performs the same AbflaMacshar himself is quoted to warn us that this
service for Book III of the Liber Aristotelis,for he is the product of an Arab's imagination, not a Sas-
cites a high percentage of Masha'allah's conflation sanian's.
of Buzurjmihr's Bizdaj with Rhetorius' compen- We must now use our imaginations to answer
dium. Where Hugo's Latin presents a patchwork two final questions: how did Rhetorius travel from
of sentences drawn from Dorotheus, Valens, and Egypt to Baghdad, and how were Greek astron-
Rhetorius, Sahl's Arabic preserves the same patch- omy and astrology restored to Byzantium after the
work. Unfortunately, both of the extant manu- disasters of the seventh and eighth centuries? The
scripts of Sahl's Kitdb al-mawdlid, that in the Escor- key to the answers to both questions is held by a
ial (Arab 1636, fols. 1-99) and that in Tehran Syrian Maronite Christian born in Edessa in about
(Majlis 6484, fols. 61-142) are defective at the be- 695, Theophilus the Philosopher, the son of
ginning, so that we do not have the preface in Thomas, whose name is commonly corrupted into
which Sahl would have named his sources-pre- Nfifil in Arabic texts and their Latin descendants.54
sumably including Mdsha'alldh-nor do we have Theophilus provides what some may regard as the
his compilation for the first astrological place and only truly Classical note in this paper; for he trans-
for most of the second. But the rest allows us to lated the Iliad and the Odysseyinto Syriac. But his
recover the Arabic originals of Masha'allah's trans- major works are four substantial treatises on as-
lations from Pahlavi and from Greek, and resolves trology composed in Greek. In composing them he
many of the puzzles occasioned by the obscurities demonstrates an ability to use Sassanian material,
of Hugo's Latin. whether he read it in the original Pahlavi or in Ar-
One of the other Arabic astrologers who quotes abic, Syriac, or Greek translations. For his work on
from the lost work of al-Andarzaghar is the Chris- military astrology, the H6vot EtOL xaTaQa6v tokE-
tian scholar named Ibn Hibinta, whom we men- tx<~ov,depends heavily on the Sanskrit Brhadyatra
tioned earlier as having composed a Kitab al- composed by Varahamihira at Ujjayini in about
mughnifl ahkamal-nujiymat Baghdad in about 950. 550, which could only have reached him through
The Kitab al-mughni is probably-I haven't yet a Sassanian intermediary. He also summarizes
been able to check the manuscript in the Zahariye chapters on war and on receiving letters ascribed
Library in Damascus-the original of the author- to Zoroaster that he presumably derived from a
ity named Mo6yvrg in a Byzantine translation Sassanian source. These were later incorporated
from the Arabic preserved in Vaticanus graecus by Eleutherius Elias into his pseudo-Palchus, and
1056.52 This text contains an interrogation con- thence made their way into Bidez and Cumont's
cerning Muhammad and his career that reminds Les mages hellinises.55 Thus they are perfect ex-
us of the similar one falsely ascribed to Stephanus amples of one of the themes of this paper. For the
of Alexandria, but which may be by Stephanus the astrology they represent originated in the Hellen-
Philosopher, whose importance will soon become istic period, was transmitted to Iran, and returned
apparent. The interrogation concerning Mulham- via Baghdad and Syria to Byzantium. The main
mad in
Moiyvrqg
is addressed by the king of the problem with the historical reconstruction offered
Persians to Valens. The attribution is obviously by Bidez and Cumont is that they mistook the
false, but it does indicate the fame which Valens fourteenth-century form of the text for one of the
enjoyed as a Sassanian astrologer The horoscope early sixth century reflecting one of the Hellenistic
of this interrogation can be dated 7 November
939,53 so that it was cast only a few years before Ibn 54An edition of all of Theophilus' surviving Greek works is in
preparation. Concerning Theophilus' HI6vot see D. Pingree,
52 Edited
partially in CCAG 5,3, pp. 110-21; it will be reedited The Yavanajatakaof Sphujidhvaja, Harvard Oriental Series 48, 2
with annotations in the volume of early Arabic astrological vols. (Cambridge, Mass., 1978), II, 389.
texts. 55Fragments 0 79 and 0 81 in J. Bidez and F. Cumont, Les
53Pingree, "The Horoscope of Constantinople," 314. mages hellinisis, 2 vols. (Paris, 1938), II, 208-19 and 225-26.

This content downloaded from 128.206.9.138 on Tue, 26 Nov 2013 16:22:13 PM

All use subject to JSTOR Terms and Conditions
 CLASSICAL AND BYZANTINE ASTROLOGY 237

period. Therefore they failed to recognize why peoples: the Egyptians, he says, use the heliacal ris-
and when these texts were ascribed to Zoroaster. ing of Sirius, the Greek [tOTIt[ltxoC, he claims,
Other alleged fragments of Zoroaster preserved in use the New Moon preceding the Sun's entry into
Cassianus Bassus' Geoponicamay have followed this Aries, that is, 1 Nisan; and Ptolemy, he reports
same route from Iran via Syria to Byzantium.56 with reference to the Tetrabiblos,divides the year
But Theophilus' sources included much more into four seasons, each of which begins at its own
than material derived from the Pahlavi texts of TQo0nT.But, he continues, the Persians throughout
Sassanian Iran. The incomplete copies that we the whole of the Orient, being lovers of wisdom,
have of his fIlp xa~taaQfv &c(6Qwcv show that he have translated the Greek books into their own
had read some form of the poem of Dorotheus, language; and they use the entry of the Sun into
but also the 'Ano0TeXeoFtax6Td of Hephaestio of Aries as do Critodemus, Valens, Dorotheus, and
Thebes. And substantial extracts from Rhetorius Timocharis. Valens and Dorotheus, of course,
appear both in Theophilus' work on military as- Theophilus would associate with the Persians be-
trology and in that on general prognostications. cause he knows that their works existed in Pahlavi;
Into the former he has copied out the three chap- Critodemus is cited by Valens; and Timocharis by
ters on war that are found in Rhetorius' summary, Ptolemy in the Almagest, which also existed in a
in his Book VI, of Julianus of Laodicea's lost TIEQp Pahlavi version. Theophilus, therefore, is histori-
xca~tay6v •x•oya'l xT"QoL[ot;and, in the latter, cally correct in implying that these four names
three chapters-on the dodecatemoria, on mad- were known in Iran though wrong in identifying a
men and epileptics, and on the lot of death-sum- common epoch for them. He goes on to say that
marize chapters in Book V of Rhetorius. In this he will use as epoch for his own annual meteoro-
same work Theophilus also copies Rhetorius' ver- logical and political predictions the local time at
sion of Hermes' treatise on the desert stars, but he which the Sun enters Aries "at the capital city," he
corrects the longitudes of those stars so that they explains, "of the Saracens, where I have made my
correspond to their Ptolemaic longitudes in 768. calculations, which is East of Babylon and is called
We can conclude, then, that by the 760s Theophi- E`l'v6unoXtg (i.e., in Arabic, Dar al-salam), but in
lus had a copy in Greek of Rhetorius; the question the dialect of the Syrians Baghdad."
arises of where he and that manuscript were. We now must address the question of how an
We know that he served as military advisor to al- interest in scientific texts, and particularly in as-
Mahdi, the caliph from 775 to 785, and that he tronomy and astrology, came to be implanted in
predeceased his master by twenty days, that is, he Byzantium. Lemerle57 regards the primary mover
died on 15 or 16 July in 785. In the first preface to to have been Leo the Mathematician who allegedly
his T[6votLTet XatTaQX(v to0XCt{xLv he relates that was wooed by al-Ma'mian to come to Baghdad in
he accompanied ot tlvtxaTra x aToiwveg on an the period between 829 and 833 because of his ex-
expedition to the East, to MaQyLavlj-that is, pertise in Euclidean geometry. One version even
Khurasan-during which he suffered from the claims that Leo's appointment to the archbishopric
cold. This may have been the expedition that al- of Thessalonica in 841 was a reward for his resist-
Mahdi, then just the son of the ruling caliph, al- ance to the Arab's blandishments. In fact, of
Mansir, led against the rebellious governor of course, al-Ma'mian had many excellent scientists in
Khurasan, cAbd al-Jabbdr, in the winter of 758- Baghdad, including geometers; and what little we
759. If this identification is correct, Theophilus have from Leo's hand-mostly astrological bits,
was serving the cAbbasids at Hashimiya near al- though also one scholium on multiplying frac-
Kfifah already before al-Mansfir founded Bagh- tions-indicates that he was not in any way a dis-
dad in 762. Since Mdsha'allah was also in the cal- tinguished or even a very competent mathemati-
iph's employ at this time, it is more than likely that cian.58 His library would have been of greater
he knew Theophilus; and thus the link is estab- interest to al-Ma'muin than his largely self-
lished that brings Rhetorius and Masha'allah to- acquired mathematical talent.
gether in Baghdad. Leo's astrological writings are based on the
Theophilus locates himself in Baghdad in his anonymous commentator on Ptolemy's Tetrabiblos,
'Ewttowxaywymi7tEQLxoo•ttLxv xataqxv, which on Paul of Alexandria, on Hephaestio of Thebes,
deals with general and political astrology. He be-
gins by indicating the year-beginnings of various
57Lepremierhumanismebyzantin, 148-76.
58D. Pingree, article on Leo the Mathematician in Dictionary
56 Fragments 0 37-0 52 in Bidez and Cumont, II, 173-97. of ScientificBiography,8 (New York, 1973), 190-92.

This content downloaded from 128.206.9.138 on Tue, 26 Nov 2013 16:22:13 PM

All use subject to JSTOR Terms and Conditions
238 DAVID PINGREE

and on John Lydus, and on a non-Classical tradi- trology, and history to forge the astrologicalhis-
tion of political astrology. All of the material that tory of pseudo-Stephanusof Alexandria.
he shows a knowledge of, in fact, is found in Lau- Indeed, what we have of his, aside from the ci-
rentianus 28,34, which contains some of Leo's own tations from his Arabic works in AbiaMacshar and
treatises and a reference to the exile of Patriarch others, and from his Greek works in Eleutherius
Photius in 867; this manuscript, then, may be de- Elias' pseudo-Ahmad the Persian, is a short de-
scended from one or more of Leo's. The political fense of astrology as a Christianscience, a defense
astrology that it contains is the adaptation of Sas- similar to, but by no means dependent on, that of-
sanian theories produced by Theophilus, and it is fered by Theophilus in the preface to his second
this type of astrology, though not necessarily di- edition of the I6vot. Stephanus'work in its purer
rectly from Theophilus, that Leo was imitating. form is preserved in Vaticanusgraecus 1056, like
Another scholar who wrote on political astrology so much else that I have discussed in this paper.
in Greek between the time of Theophilus and Leo In the HQIFL Tjg t[aO artTxiLg
was Stephanus the Philosopher.59 He was well T~~v•gStephanus
states that he has come from Persia-presuma-
known in Baghdad in the ninth century since bly he means by this Baghdad-to this happy
Shadhan has Abfi Macshar refer to him frequently, city only to discover that the astronom-
and his sayings are quoted by others in this pe- (e6'a(•t•wv)
ical and astrologicalparts of philosophy have been
riod;60 but he is clearly represented as a figure snuffed out in it. That he speaksof Constantinople
from the early days of Arabic astrology, from the is clear from the fact that he places it in the fifth
decades immediately preceding 800, and is asso- clima. He notes that people find it difficult to use
ciated by Ibn Abi al-Rijal with Theophilus. I have the tablesof Ptolemy and Ammonius because their
already mentioned that Stephanus may be the man calculationserr with respect to the Sun by 5o. This
who wrote in Greek, during the reign of al-Mahdi, is not, as Boll thought,63a reference to precession,
a history of the Caliphate based on the horoscope but rather to the fact that people might forget to
of the year in which the Hijra occurred. The true include in a year'smotion of the Sun the five epa-
authorship of this interesting work is concealed by gomenal days of the Egyptian calendar.That that
a fake preface in which Stephanus of Alexandria is Stephanus' concern is shown by his succeeding
is represented as receiving the news about Mu- remark that learners find it difficult to deal with
hammad from Epiphanius, a merchant who ar- calendars they are not used to; and Stephanus re-
rived in Constantinople from the Yemen on 1 Sep- fers then to Ptolemy'suse of the years of Napovxo-
tember 621.61 The forger is well informed both 86voooQ and the Egyptian months; to Theon's,
about Stephanus' role as commentator on the Heraclius', and Ammonius' use of the years of
Handy Tablesin the early 620s, and about the meth- Philip and the Egyptian months; and to the use by
odology of Sassanian political astrology based on oLvE&teotLof the years of the Persian kings or of
a year-beginning, as was Theophilus' 'Emovv- the Saracens.In considerationof these difficulties,
But pseudo-Stephanus of Alexandria Stephanus saysthat he has computed tablesfor the
aywy0j.
chose none of the epochs for the year mentioned longitude and latitude of this happy city,and based
by Theophilus, but rather the year-beginning of them on the Byzantine calendar.
the Byzantine calendar.62 There is only one set of The tables of oLveTeQot to which Stephanusre-
Greek astronomical tables that I know of before fers are of some importance in determining more
Isaac Argyrus' New Tables of 1367 that used the precisely the date of this treatise. Among the ear-
Byzantine calendar, and that is the lost set com- liest Arabic uses of tables based on the Sassanian
puted and described by Stephanus the Philoso- calendar with the first year of a Sassanianemperor
pher. He is also one of the few men of the period as the epoch, as was characteristicof the Pahlavi
of al-Mahdi's reign (775-785) who had the knowl- Zik-i Shahriydrans,was Masha'allah's use of Khusro
edge of both Greek and Oriental astronomy, as- Anfishirwin's;64
it is not coincidental, then, that
this Z~jis cited in the Liber Aristotelis.But the Per-
59The Greek fragments of his works will be published else- sian Royal Canons had probably been used earlier
where. See, for now, F. Cumont in CCAG 2, pp. 181-86; and D. in Islam, perhaps even in the historical horoscopes
Pingree, "Historical Horoscopes,"JAOS 82 (1962), 487-502. of the early period of Islam computed shortly after
60Sezgin, GAS, 48-49.
61This was edited
by H. Usener, De Stephano Alexandrino
(Bonn, 1880), 17-32. 63CCAG2, p. 181.
62For his horoscope, cast for 1 September 621, see O. Neu- 64See F. I. Haddad, E. S. Kennedy, and D. Pingree, The Book
gebauer and H. B. Van Hoesen, GreekHoroscopes(Philadelphia, of the Reasons behind AstronomicalTables (Delmar, N. Y., 1981),
1959), 158-60. passim.

This content downloaded from 128.206.9.138 on Tue, 26 Nov 2013 16:22:13 PM

All use subject to JSTOR Terms and Conditions
 CLASSICAL AND BYZANTINE ASTROLOGY 239

679.65 But it was only in about 790 that al-Fazari we read: "The Moon in Gemini with the aspect of
computed the first set of tables to use the Muslim the benefics indicates the unsuccessful withdrawal
calendar.66 So Stephanus wrote his Christian apol- of the besieging troops" (Constantine was within
ogy for astrology probably in the 790s, though his the fortress, which was surrounded by the Bul-
Byzantine tables were earlier-probably early gars). At about noon on 20 July 792 the Moon was
enough to have been used in forging the pseudo- in the twentieth degree of Gemini aspected in sex-
Stephanus of Alexandria text in about 780. tile by the benefic planet, Venus, which was in the
That Theophilus' work was known in Byzantium twentieth degree of Leo.
in the 790s, presumably through Stephanus' inter- With Stephanus, then, we have astrology and as-
vention, is indicated by an incident recorded in tronomy restored to Byzantium, historical astrol-
Theophanes' Chronographiaunder the year 6284 ogy introduced from the East, and the mathemat-
(A.D. 792).67 In July of that year the emperor, Con- ical art so stoutly defended as a Christian science
stantine VI, marched against the Bulgars. He built that even the archbishop of Thessalonica felt free
a fortress, Marcellae, on the border. On the twen- to follow it. As Stephanus says in the second chap-
tieth of the month Kardam, the Bulgarian ruler, ter of his little work, in a Christian version of the
led his army across the frontier up to the fortifica- late eighth-century history of astrology that I
tions. The emperor, being advised by his "pseudo- quoted at the beginning of this paper in its devel-
prophet and astrologer," Pancratius, that victory oped form as concocted by Abid Macshar:
would be his, sallied forth to ignominious defeat.
Seth, as we have read, was the first to use this art.
Theophanes does not describe the astrological The Chaldaeans received it from him; then it went
technique used by Pancratius, but it can be guessed from them to the Persians, and from them to the
at from chapter 20 of Theophilus' I6vot,68 where Greeks, from whom it was transferred to the Egyp-
tians, from whom the Romans (i.e., Byzantines)were
65Pingree, The Thousands, 114-21. also initiated. Then finally the Arabs got it. And all
66D. Pingree, "The Fragments of the Works of al-Fazari," the nations that have been mentioned had almostcos-
JNES 29 (1970), 103-23. mocraticand victoriousdynastiesas long as they used
67TheophanisChronographia,ed. C. de Boor, I (Leipzig, 1883), it. Therefore I thought it necessaryto renew this use-
467-68. ful science among the Romans (i.e., Byzantines)and
68Edited by C. O. Zuretti in CCAG 11, 1, p. 206. Further Byz- to implant it among the Christiansso that they might
antine horoscopes interpreted in accordance with Theophilus' be deprived of it nevermore.
methods can be found, e.g., among those published by D. Pin-
gree, "The Astrological School of John Abramius," DOP 25
(1971), 189-215. Brown University

This content downloaded from 128.206.9.138 on Tue, 26 Nov 2013 16:22:13 PM

All use subject to JSTOR Terms and Conditions