Al-Farabi and the History of the Syriac Organon 9781463216733

Table of contents :
AL-FARABI AND THE HISTORY OF THE SYRIAC ORGANON
ADDENDA ET CORRIGENDA

Citation preview

Al-Farabi and the History of the Syriac Organon

Analecta Gorgiana

129 Series Editor George Kiraz

Analecta Gorgiana is a collection of long essays and short monographs which are consistently cited by modern scholars but previously difficult to find because of their original appearance in obscure publications. Carefully selected by a team of scholars based on their relevance to modern scholarship, these essays can now be fully utilized by scholars and proudly owned by libraries.

Al-Farabi and the History of the Syriac Organon

John Watt



 2009

Gorgias Press LLC, 180 Centennial Ave., Piscataway, NJ, 08854, USA www.gorgiaspress.com Copyright © 2009 by Gorgias Press LLC All rights reserved under International and Pan-American Copyright Conventions. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, scanning or otherwise without the prior written permission of Gorgias Press LLC. 2009

‫ܙ‬



ISBN 978-1-60724-041-9

ISSN 1935-6854

This extract originally appeared in George A. Kiraz, ed., Malphono w-Rabo d-Malphone: Studies in Honor of Sebastian P. Brock, Gorgias Press, 2008, pages 751–778.

Printed in the United States of America

AL-FąRąBĩ AND THE HISTORY OF THE SYRIAC

ORGANON

JOHN W. WATT Syriac studies at many points are intertwined with Arabic, but few Arabic texts are as pertinent to Syriac as the account in Ibn Abĩ UʛaibiȨa attributed to al-FĆrĆbĩ of the appearance of philosophy in Islam.1 The text became well known through the celebrated study of Max Meyerhof,2 who in essence accepted the basic thrust of the account, namely, that the group of Baghdad Aristotelians associated with Abu Bišr MattĆ and al-FĆrĆbĩ were direct successors of the School of Philosophy in late antique Alexandria by virtue of that School’s migration via Antioch to the ȨAbbĆsid capital. The fictional character of this basic account, however, is now widely accepted,3 and the reliable historical information to be gleaned from it usually thought to be limited for the most part to that section close in time to al-FĆrĆbĩ himself, identifying his own teachers and contemporaries.4 It has also been argued that al-FĆrĆbĩ’s own bias affected the substance of the account,5 and that a 1 Ibn Abĩ UʛaibiȨa, ed. A. Müller, ȨUyşn al-anbĆ‚ fĩ ʞabaqĆt al-aʞʞibĆ‚ (Cairo, Königsberg, 1882–1884), vol. II, pp. 134–135. English translation in F. Rosenthal, The Classical Heritage in Islam (London, 1975), pp. 50–51. English or German translations of the account, or particular sections of it, are also given in the works cited in the following five footnotes. 2 M. Meyerhof, ‘Von Alexandrien nach Bagdad,’ Sitzungsberichte der preussischen Akademie der Wissenschaften, phil.-hist. Klasse 23 (Berlin, 1930), pp. 389–429. 3 G. Strohmaier, ‘Von Alexandrien nach Bagdad—eine fiktive Schultradition,’ in J. Wiesner (ed.), Aristoteles: Werk und Wirkung. Paul Moraux gewidmet, vol. II (Berlin, 1987), pp. 380–389. 4 F. W. Zimmermann, Al-Farabi’s Commentary and Short Treatise on Aristotle’s De Interpretatione (Oxford, 1981), pp. ciii-cviii. 5 S. Stroumsa, ‘Al-FĆrĆbĩ and Maimonides on the Christian Philosophical Tradition: a Re-evaluation,’ Der Islam 68 (1991), pp. 263–287; J. Lameer, ‘From Alexandria to Baghdad: Reflections on the Genesis of a problematical Tradition,’ in

751 [1]

752

WATT

prominent motif within it, that of the fearful or hostile attitude of Christian leaders to parts of Aristotelian logic, is drawn from a more general antiByzantine polemicalbeit directed towards Muslim opponents of philosophy rather than the Byzantines themselvesprobably originating in the reign of al-Ma‚mşn.6 Concretely, in al-FĆrĆbĩ’s narrative this fearfulness on the part of Christian bishops takes the form of an allegation that the sections of the logic curriculum (the Organon) beyond the assertoric figures (Prior Analytics I 7) should not be studied, or not studied with a teacher,7 because they were harmful to Christianity. Although the Christian or episcopal motivation for this limitation, together with the linear thrust of the basic narrative, is now generally discounted, the assertion that in fact only an abridged Organon was studied for some time in the East before the coming of Islam (or before the rise of Islamic philosophy under the ȨAbbĆsids) is still given some credence, and in particular is thought to be firmly supported by Syriac evidence. It is this aspect of the account which I intend to examine in the present article. At the outset it may be noted that evidence in Syriac of particular effort being directed towards the early parts of the Organon8 is not unique to that language sphere. The study of predications and propositions (derived from the Categories and De Interpretatione) leading to the assertoric syllogism of Prior Analytics I 1–7 appears to have been considered even in Roman times as a relatively self-contained unit within logic as a whole. Special attention to these subjects did not necessarily imply ignorance or total disregard of the other parts of logic. If there was a ‘Syriac tradition’ devoted to these treatises alone, it should therefore neither be taken without further ado as the only Syriac tradition, nor considered a special characteristic of the Syrians rather than a feature of the antique tradition in general. The Latin G. Endress and R. Kruk (eds.), The Ancient Tradition in Christian and Islamic Hellenism (Leiden, 1997), pp. 181–191. 6 D. Gutas, Greek Thought, Arabic Culture. The Graeco-Arabic Translation Movement in Baghdad and Early ȨAbbĆsid Society (2nd-4th/8th-10th centuries) (London and New York, 1998), pp. 83–95; idem, ‘The “Alexandria to Baghdad” Complex of Narratives. A Contribution to the Study of Philosophical and Medical Historiography among the Arabs,’ Documenti e studi sulla tradizione filosofica medievale 10 (1999), pp. 155–193, esp. pp. 174–179. 7 The latter is Gutas’s reading of the text; cf. his ‘Alexandria to Baghdad,’ pp. 164, 179–180. 8 I use the terms ‘early’ and ‘late’ in this article not with reference to date of composition, but to position in the sequence of treatises comprising the Organon.

[2]

AL-FąRąBĨ

753

Peri Hermeneias of (or attributed to) Apuleius, and the ‘Old Logic’ of which it formed a part, are evidence of this particular strand of logical tradition in antiquity. Apuleius deals only with assertoric syllogisms, without any discussion of the modal, and takes his material predominantly from Aristotle’s treatise of that name and the first seven chapters of the Prior Analytics. He doubtless knew the rest of Aristotle’s logical works, but his employment of them is sporadic; he may have used a few chapters from the later parts of the Prior Analytics and from the Topics, but he appears to have made no use at all of the Posterior Analytics or Sophistical Refutations (or indeed of the Categories).9 Marius Victorinus translated Porphyry’s Eisagoge, the Categories, and the De Interpretatione, and wrote treatises on definition and on categorical and hypothetical (i.e., Stoic) syllogisms, but nothing on modal syllogisms or anything from the Posterior Analytics, while Jerome (Ep. 50, 1) criticised a false logician for not having read the Categories, De Interpretatione, and Prior Analytics, making no mention of the rest of the Organon. The Aristotelian treatises included in the Latin ‘Old Logic’ (Logica Vetus) were the Categories and the De Interpretatione, but the ‘Old Logic’ also included Apuleius’ treatise and some of Boethius’ writings.10 The latter’s De Syllogismo Categorico, like Apuleius’ work, does not include anything on modal syllogisms in its treatment of the Prior Analytics. Boethius translated the entire six-volume Organon, but only the first two volumes became well known, and he only wrote commentaries or monographs (apart from that on hypothetical syllogisms, i.e., Stoic logic) on the (Eisagoge and) Categories, De Interpretatione, and the categorical syllogism (an exposition of Prior Analytics to I 7). Thus although he knew and appreciated the entire Organon, when writing a teaching manual he effectively confined himself, like Apuleius and Victorinus, to the ‘abridged’ version terminating at Prior 9 Cf. M. W. Sullivan, Apuleian Logic. The Nature, Sources, and Influence of Apuleius’s Peri Hermeneias (Amsterdam, 1967), pp. 143–146. 10 Ibid., pp. 228–232. The Latin ‘Old Logic’ was brought forward as a parallel to the Syriac by Sh. Pines, ‘A Parallel in the East to the “Logica Vetus,”’ in J. P. Beckman et al. (ed.), Philosophie im Mittelalter: Entwicklungslinien und Paradigmen (Hamburg, 1987), pp. 125–129 (reprinted in S. Stroumsa [ed.], The Collected Works of Shlomo Pines, vol. III [Jerusalem, 1996], pp. 262–266) but he still thought of this truncation of the Organon, following al-FĆrĆbĩ, as determined by Christian bishops; cf. Stroumsa, ‘Al-FĆrĆbĩ and Maimonides,’ pp. 275–276. Pines also laboured under the misapprehension that ‘as far as is known, the whole of the Prior Analytics was not translated into Syriac until the ninth century [and this is also true of the Posterior Analytics];’ cf. below.

[3]

754

WATT

Analytics I 7.11 If the same phenomenon occurs independently in both Latin and Syriac, we may be sure that its origin lies in Greek scholastic tradition, and the commentary of David on the Prior Analytics (preserved only in Armenian translation) has been adduced as possible testimony to the recognition of an abridged Organon even among the Greek commentators on Aristotle, since that commentary appears to treat only the early part of the treatise.12 It is frequently assumed that al-FĆrĆbĩ’s statement on the prohibition of the (public) teaching of the books of logic beyond the assertoric figures relates to the practice of the Syriac schools.13 Al-FĆrĆbĩ himself, however, does not attribute the abridgement of the teaching syllabus to Syrians, but to Christians. In his account, Aristotle’s teaching was carried on at Alexandria and Rome,14 and after the coming of Christianity was continued only at the former and was subsequently curtailed by the bishops. The alleged transfer of the school to Antioch followed the (public) restoration of the full curriculum at or after the coming of Islam. Thus according to alFĆrĆbĩ the curtailment of the (public) teaching occurred in Alexandria,15 which he presumably did not consider a Syriac-speaking city. The text does not therefore make an assertion about Syriac logical studies, but about logical studies in Alexandria in an unspecified period of Christian rule. Why al-FĆrĆbĩ substituted the reduction of the logical curriculum (which had nothing to do with Christianity) for the general hostility to philosophy on the part of Christian leaders in the ȨAbbĆsid-era narrative which he took over16 is a question to which we shall have to return. Here, however, it may 11 Cf. P. Hadot, Marius Victorinus; recherches sur sa vie et ses oeuvres (Paris, 1971), pp. 194–196; H. Chadwick, Boethius. The Consolations of Music, Logic, Theology, and Philosophy (Oxford, 1981), pp. 115–118, 133–141, 163–166. 12 H. Hugonnard-Roche, La logique d’Aristote du grec au syriaque (Textes et traditions, 9) (Paris, 2004), p. 19, n. 1. Porphyry himself may also be mentioned in this connection: in addition to the Eisagoge he also composed two commentaries on the Categories, one on the De Interpretatione, and an introduction to categorical syllogisms. Cf. Hadot, Marius Victorinus, p. 193. On the Armenian tradition, cf. below, n. 61. 13 Gutas, ‘Alexandria to Baghdad,’ p. 185. 14 Rşmiya. Stroumsa, ‘Al-FĆrĆbĩ and Maimonides,’ pp. 267–268, suggests the reference is to Constantinople. 15 This is correctly noted by Stroumsa, ‘Al-FĆrĆbĩ and Maimonides,’ p. 272. Gutas’s point, ‘Alexandria to Baghdad,’ p. 182, that al-FĆrĆbĩ does not refer to Alexandria, overlooks the sequence of the narrative. 16 See above, n. 6.

[4]

AL-FąRąBĨ

755

be noted that what he did say explicitly about the Syriac logical tradition offers no hint that he attributed to the Syrians any abridgement of the curriculum. In his treatise on The Attainment of Happiness, he wrote that ‘this science [of demonstrative logic] existed anciently among the Chaldeans, who are the people of Iraq, subsequently reaching the people of Egypt, from there transmitted to the Greeks, where it remained until it was transmitted to the Syrians and then to the Arabs. Everything comprised by this science was expounded in the Greek language, then in the Syriac language, then in the Arabic language.’17 Whatever ideological motivations one may discern in his reference to the Chaldeans and Egyptians, in the sequence ‘Greek-Syriac-Arabic’ he was in the realm of firm historical tradition, no doubt mediated to him primarily through his membership of the school of the Baghdad Aristotelians under the leadership of Abş Bišr MattĆ. A common misconception which seems often to underlie the discussion of this matter is that before the ȨAbbĆsid era nothing of the Organon beyond Prior Analytics I 7 had been translated or studied by Syriac scholars. This is presumably the source of the suggestion that what the complex of narratives as evidenced in al-FĆrĆbĩ tells us about the curriculum of logic in Syriac is relatively accurate, and that a structural change occurred in the Syriac tradition in Baghdad at the end of the eighth century.18 The evidence, however, does not support this conclusion, and if al-FĆrĆbĩ’s account has any evidential worth, it indicates some change, not in the Syriac curriculum of logic at the end of the eighth century, but in the Arabic near the end of the ninth. The evidence of Syriac interest in the full Organon before the end of the eighth century is not new, but it bears repeating here.19 Athanasius of Balad (died 686) translated into Syriac both books of 17 M. Mahdi, Alfarabi’s Philosophy of Plato and Aristotle (Ithaca, N.Y., 1969/2001), p. 43 (§ 53); Arabic text ed. J. al-YĆsĩn, Taʚʜĩl al-saȨĆda (Beirut, 1981), p. 3814–17. 18 Gutas, ‘Alexandria to Baghdad,’ p. 186; idem, Greek Thought, Arabic Culture, p. 22—but contrast p. 62! Cf. also below, n. 35. 19 It is finely set out in S. Brock, ‘The Syriac Commentary Tradition,’ in C. Burnett (ed.), Glosses and Commentaries on Aristotelian Logical Texts: The Syriac, Arabic and Medieval Latin Traditions (Warburg Institute Surveys and Texts, vol. XXIII), (London, 1993), pp. 3–18, esp. pp. 4–5. Attention has also been drawn to it by Hugonnard-Roche, ‘L’intermédiaire syriaque dans la transmission de la philosophie grecque à l’arabe: Le cas de l’Organon d’Aristote,’ Arabic Sciences and Philosophy 1 (1991), pp. 187–209, esp. pp. 188–190, and La logique d’Aristote, p. 19. Stroumsa, ‘Al-FĆrĆbĩ and Maimonides,’ pp. 271–273, seems aware that the situation is not as

[5]

756

WATT

the Prior Analytics, the Posterior Analytics, the Topics, and the Sophistical Refutations. George, Bishop of the Arabs (died 724), translated both books of the Prior Analytics, while Theophilus of Edessa, who did live into the ȨAbbĆsid period (died 785), translated the Prior Analytics and Sophistical Refutations. The evidence for Athanasius’ translation of the Prior Analytics, Topics, and Sophistical Refutations is found in the marginal and interlinear notes of the manuscript of the Arabic Organon produced in the school of the Baghdad Aristotelians (MS Parisinus arabe 2346),20 and that for his simple as is commonly supposed, but does not deal with the most important evidence. (The ‘Jacob’ mentioned at Stroumsa, p. 271, should read ‘George,’ and the commentary is on the Prior Analytics, not the Posterior Analytics.) 20 These marginal and interlinear notes are reproduced in the edition of ȨA. Badawĩ, Manʞiq Arisʞş, vols. I-III (Cairo, 1948–1952). For Prior Analytics, see Badawĩ, pp. 113 nn. 4 and 6, 115 n. 4, 116 n. 2 (all on I 4), and 284 n. 2 (on II 17and thus beyond I 7). For Topics, cf. ibid., pp. 563 n. 4, 572 n. 1, 579 n. 5, 636 n. 2, 646 n. 1, 682 nn. 2 and 4, 685 n. 1, 686 n.1, 703 n. 2, and 719 n. 4. For the Sophistical Refutations, cf. ibid., introduction pp. 30–31, and text p. 737 . See also K. Georr, Les Catégories d’Aristote dans leurs versions Syro-Arabes (Beirut, 1948), pp 26–28, 30–31, 190–193, 195–200. The importance of this material was first appreciated in the West by R. Walzer, ‘New Light on the Arabic Translations of Aristotle,’ in Oriens 6 (1953), pp. 91ff., reprinted with an addition and cited here from R. Walzer, Greek into Arabic (Oxford, 1962), pp. 60–113. On the notes citing Athanasius, see Greek into Arabic, pp. 85–88 (Prior Analytics), 99 (Topics), and 81–83 (Sophistical Refutations). The evidence of the Paris Organon for the Syriac versions of Aristotle was also noted by F. E. Peters, Aristoteles Arabus (Leiden, 1968), pp. 14, 20–21, and 23. It should be noted that the low opinion of Athanasius’ ability as an interpreter held by Ibn SuwĆr (who did not know Greek) does not necessarily mean that Athanasius was in fact a bad translator; his literal translation style may have caused Ibn SuwĆr difficulty. Phocas of Edessa, who did know Greek and like Athanasius translated Pseudo-Dionysius, did not share Ibn SuwĆr’s opinion; cf. G. Wiessner, Zur Handschriftenüberlieferung der syrischen Fassung des Corpus Dionysiacum, Nachrichten der Akademie der Wissenschaften in Göttingen, phil.-hist. Klasse, 1972, Nr. 3, pp. 198–199 (nn. 12 and 14 referring to pp. 167–168) and S. P. Brock, ‘Two Letters of the Patriarch Timothy from the late Eighth Century on Translations from Greek,’ Arabic Sciences and Philosophy 9 (1999), pp. 233–246, at p. 244. It is quite mistaken (as is frequently done) to assume without question that criticisms made by translators of the ȨAbbĆsid era of the competence of their predecessors (including ʗunain’s celebrated criticism of Sergius) are factually valid. Different perceptions of translation methodology also enter into the picture. Cf. S. Brock, ‘Towards a History of Syriac Translation Technique,’ in R. Lavenant (ed.), III Symposium Syriacum 1980, Orientalia Christiana Analecta, 221 (Rome, 1983) [reprinted in Sebastian Brock, Studies in Syriac Christianity, Hampshire, 1992, chapter X], pp. 9–10;

[6]

AL-FąRąBĨ

757

translation of the Posterior Analytics (and Topics) in a letter of the Patriarch Timothy I.21 George’s version of the entire Prior Analytics is extant,22 and Theophilus’ versions are cited in the notes of the Paris Organon,23 which also cites further anonymous Syriac versions (of unidentified date) in addition to those of ʗunain and IsʘĆq.24 The date of Athanasius’ translations is not known, but he made a revised translation of Porphyry’s Eisagoge in 645. A complete Syriac Organon (excepting the Rhetoric and the Poetics)25 thus certainly existed no later than 686, the death of Athanasius, and the subsequent translations by George (his colleague or pupil at Qenneshre) and Theophilus indicate that Athanasius was not completely alone among Syrians living before the time of ʗunain and his son IsʘĆq in having an idem, ‘The Syriac Background to Hunayn’s Translation Techniques,’ Aram 3 (1991) [pp. 139–162, reprinted in Sebastian Brock, From Ephrem to Romanos (Aldershot, 1999), chapter XIV], pp. 151–153; and Hugonnard-Roche, ‘Les traductions du grec au syriaque et du syriaque à l’arabe à propos de l’Organon d’Aristote,’ in J. Hamesse and M. Fattori (eds.), Rencontres de cultures dans la philosophie médiévale. Traductions et traducteurs de l’antiquité tardive au XIVe siècle (Louvain-la-Neuve/Cassino, 1990 [pp. 131–147]), pp. 134–137. 21 Letter 48, to Sergius. Cf. H. Pognon, Une version syriaque des aphorismes d’Hippocrate (Leipzig, 1903), pp. XXI-XXV (text and translation); S. P. Brock, ‘Two Letters of the Patriarch Timothy,’ pp. 238–239 (translation) and 245–246 (commentary). According to Timothy, Athanasius translated aulētrides (‘fluteplayers,’ Post. Anal. I 13. 78b30–31) by zammĆrĆtĆ (‘women singers,’ ‘women fluteplayers’). According to the note in the Paris Organon, the Syriac (i.e., of IsʘĆq) read (in Arabic) muĠanniyĆt (‘women singers’), thus probably Syriac zammĆrĆtĆ, the same as in the translation of Athanasius. Cf. Pognon, Version syriaque, p. XXI22–23; Badawĩ, Manʞiq Arisʞş, p. 351 n. 4 (referring to p. 3524–5); Walzer, Greek into Arabic, pp. 105–106. 22 Edited by G. Furlani, ‘Il primo libro dei Primi Analitici di Aristotele nella versione siriaca di Giorgio delle Nazione,’ Accademia nazionale dei Lincei Mem. 6, 5, 3 (1935), pp. 143–230, and ‘Il secondo libro dei Primi Analitici di Aristotele nella versione siriaca di Giorgio delle Nazione,’ ibid. (1937), pp. 233–287. 23 On Prior Analytics, cf. Badawĩ, Manʞiq Arisʞş, p. 105 n. 2, 106 n. 3, 110 n. 4, 111 n. 2, 112 n. 1, 113 n. 4, 115 nn. 2 and 4, 116 n. 2, and 284 nn. 2 and 5 (Walzer, Greek into Arabic, pp. 84–86, 88). On Sophistical Refutations, cf. Badawĩ, Manʞiq Arisʞş, pp. 767 nn. 1 and 2, 784 n.11, 785 nn. 2 and 3, and 790, n. 3; Peters, Aristoteles Arabus, pp. 23, 25 n. 1. 24 Walzer, Greek into Arabic, pp. 83–88, 99 (Badawĩ p. 645 should read p. 654) from Badawĩ, Manʞiq Arisʞş, vols. I-II, and from Badawĩ, Manʞiq Arisʞş, vol. III, pp. 676 n. 1, 678 n. 1, 687 n. 2, and 717 n. 6; Peters, Aristoteles Arabus, pp. 20–21. 25 Cf. below on the Rhetoric and Poetics.

[7]

758

WATT

interest in, and access to, the later books of the Organon.26 If there was a distinction in the Syriac monastic schools between what was supposed to be studied privately and what was read with a teacher,27 it must remain hidden to us, unless some reliable evidence of the actual school practice should become available. The translations just mentioned, however, indicate that some Syrians who were able to read Greek thought that these books should be studied by those who did not know the language, or who could only understand it with difficulty. These are the very students who are most likely to have needed a teacher to help them understand Aristotle. It is quite plausible, indeed, to suppose that the translations might have been made precisely for teaching purposes. The distinction that must have existed between these two groups of Syriac speakers, those with sufficient Greek to read Aristotle and those without, is vital to a clear analysis of this issue. It puts in question suggestions that there were two different traditions of instruction, Greek and Syriac, in the late antique and early Islamic periods.28 It is possible to conceive of a Syriac tradition different from Greek, but this would be a tradition confined to those who were entirely dependent upon works already translated into Syriac, and this is not the tradition of the Syriac translators (with the exception of the final stage, the scholars of the school of Abş Bišr MattĆ and his successors who translated from Syriac to Arabic). If separate linguistic traditions are posited, it would be better to think of Graeco-Syriac and ‘Syriac-alone’ traditions; it is the former with which we are dealing for the most part when we consider Syriac writers on the

Walzer, Greek into Arabic, p. 98, recognised the importance of the information in the Paris Organon with respect to the pre-ʗunain (or pre-IsʘĆq) Syriac translations, but was unaware of Athanasius’ translation of the Posterior Analytics. He did, however, think al-FĆrĆbĩ’s assertion was limited to ‘the Nestorian Syriac church,’ and that ‘the Jacobite followers of John Philoponus … had continued to study the important work…’ Although the text of Timothy’s letter was published by Pognon in 1903, it remained largely unnoticed until Sebastian Brock brought it to the attention of the scholarly community (already in 1993 in ‘The Syriac Commentary Tradition,’ p. 5). P. Kraus, ‘Zu Ibn al-MuqaffaȨ,’ Rivista degli studi orientali 14 (1933/4), pp. 12–13, noticed Pognon’s publication, but missed its testimony to the translation of the Posterior Analytics and Topics by Athanasius. 27 Gutas, ‘Alexandria to Baghdad,’ pp. 164, 172, 180. 28 Ibid., pp. 184–186. 26

[8]

AL-FąRąBĨ

759

Organon.29 Translators belong to two circles, those of the originating and those of the target language, both on account of their knowledge of the two languages and on account of the materials at their disposal. In the case of Greek and Syriac, the two circles overlapped both culturally and geographically. The translators of Qenneshre not only had a mastery of the Greek language, but also must have had access there to Greek manuscripts. Translating one or more Greek books (or part of a Greek book) into Syriac does not imply that they subsequently had nothing more to do with the Greek text of that book, or that they knew nothing of, or cared nothing about, those books which they did not translate. After Athanasius of Balad translated the Prior Analytics and Jacob of Edessa the Categories, George subsequently produced further Syriac versions of these books, revising his predecessors’ works using the Greek texts at his disposal. To the best of our knowledge the only book of the Organon rendered into Syriac by Jacob was the Categories, but that does not imply that he had no interest in or access to the others. He could perfectly well read them in Greek (or indeed in the translations of Athanasius and the Anonymous of the De Interpretatione). The translations were presumably made for those who could not cope easily, or at all, with the Greek original, and a fresh Syriac translation was only made when the translator felt he could improve on the work of his predecessor. When we speak about the Syriac Aristotelian tradition, we are usually thinking of men like Jacob and Athanasius, and for them the (growing) Syriac tradition merely constituted a parallel to (part of) the Greek tradition, and was presumably created for the benefit of those with little or no Greek, or possibly for their own use as teachers. Qenneshre was founded in 531 by John bar Aphtonia, who came from a Hellenised family in Edessa and, according to a panegyric of him by a monk of the monasterySyriac in language but Greek in its rhetorical formhe was educated at a monastery in Greek territory, namely St. Thomas at Seleucia on the Orontes.30 No doubt Qenneshre was the greatest of the ‘Graeco-Syriac’ monasteriesJohn himself probably composed literary works only in Greekbut it was by no means the only one, and The most significant scholars belonging to the ‘Syriac-alone’ tradition were Paul the Persian (probably) and the Baghdad Aristotelians (who were bilingual in Syriac and Arabic). 30 J. W. Watt, ‘A Portrait of John bar Aphtonia, Founder of the Monastery of Qenneshre,’ in J. W. Drijvers and J. W. Watt (eds.), Portraits of Spiritual Authority (Leiden, 1999), pp. 155–169. 29

[9]

760

WATT

Aristotle was by no means the only subject of ‘Graeco-Syriac’ study. The Greek fathers, especially Gregory of Nazianzus, were also intensively studied in some of the West Syrian monasteries and repeatedly rendered into Syriac in successive revisions based on the Greek texts. Another notable ‘Graeco-Syriac’ monastery was that of Mar Zakkai, where Marutha of Tagrit spent ten years ‘reading the books of the orthodox doctors and especially Gregory the Theologian … for the monks of that monastery were very assiduous and loved to read Gregory.’31 John of Tella entered it at the age of twenty-five, having earlier in his life been sent by his mother to the praitorion of the dux at Callinicum in order to be brought up ‘in the letters and wisdom of the Greeks.’32 When Marutha left Mar Zakkai to return to Sasanian lands, he settled at the monastery of Mar Mattai, which by the eighth century was known to possess a fine library, as we know from the Catholicos Timothy I, who asked Pethion, his former teacher at the School of Mar Abraham, to enquire there ‘whether there is some commentary or scholia by anyone, whether in Syriac or not, to this book, (that is) the Topics, or to the Refutations of the Sophists, or to that of Rhetors, or to that of Poets.’33 Gregory of Nazianzus, Pseudo-Dionysius, and Aristotle were all subject to repeated translation from Greek to Syriac during the course of the sixth to eighth centuries; Athanasius himself worked on both Gregory and Aristotle (and possibly on Pseudo-Dionysius also).34 In this connection it is pertinent to note that within the wider religio-political context of late antiquity, there are no adequate grounds for positing sharply separate Greek and Syriac traditions. The West Syrian scholars at institutions such as Qenneshre were indeed in conflict with the Dyophysite christology of the emperor at Constantinople, but the Miaphysite movement to which they F. Nau (ed./tr.), Denʚa, Histoire de Marouta, Patrologia Orientalis, vol. 3 (Paris 1909), p. 70. 32 E. W. Brooks (ed./tr.), Vitae virorum apud monophysitas celeberrimorum, Corpus Scriptorum Christianorum Orientalium, vol. 7/8 (Paris-Louvain, 1907), pp. 39–40 and 43–45/27–28 and 30–31. 33 Letter 43. Cf. Brock, ‘Two Letters,’ p. 236 (and commentary pp. 241–242); Pognon, Version syriaque, p. XVII (text), XVIII (translation); O. Braun, ‘Briefe des Katholikos Timotheos I,’ Oriens Christianus 2 (1902), pp. 4–11. 34 Cf. Brock, ‘Two Letters,’ pp. 237 and 242–244; A. Baumstark, Geschichte der syrischen Literatur (Bonn, 1922), pp. 256–257. Baumstark (1922) did not of course know of the material from the Paris Organon on Athanasius brought to light in Georr (1948) and BadawĪ (1951), and neither did Furlani in his numerous articles prior to 1948. 31

[10]

AL-FąRąBĨ

761

belongedJohn bar Aphtonia founded Qenneshre after leaving St. Thomas as a result of the strict Chalcedonian policy of Justinian’s reignwas widespread throughout the Eastern Empire. The letter of Timothy cited immediately above well illustrates the Graeco-Syriac Aristotelian tradition within some Syriac monasteries. Not only does it show Timothy’s interest in the later books of the Organon, but also the note ‘whether in Syriac or not’ indicates that he assumed there could be books at Mar Mattai in a language other than Syriac. There can be little doubt that he was thinking of the possibility of commentaries or scholia in Greek. While Timothy himself was writing towards the end of the eighth century, he could hardly have thought that such commentaries or scholia, whether in Greek or Syriac, might be present in Mar Mattai if the later books of the Organon had only just recently become a subject of study among Syrians. Timothy also notes immediately afterwards that Job the Chalcedonian told him he had seen scholia on some chapters of the Topics.35 In another letter (number 19) he asks Sergius (the addressee of letter 48) to search in the monastery of Mar Zina for ‘two treatises on Poets,’ of which he had one, together with ‘Olympiodorus’ interpretation of the books on logic, or those of Stephen, Sergius, and Alexander.’36 Both the context and the wording (cf. that in letter 43 above) favour the suggestion that the ‘two treatises on Poets’ refer to the two Books of Aristotle’s Poetics, and make it likely, therefore, that a Syriac translation of Book One of the Poetics existed in his day, and thus prior to the translation of IsʘĆq.37 On the evidence of Bar Hebraeus’ commentary in the Cream of Wisdom, we can assume that there was also an ‘old’ (i.e., pre-ʗunain) translation of the Rhetoric.38 The Paris manuscript of the Arabic Organon and the letters of Timothy might thus be thought of as ‘palimpsests,’ revealing in their ‘undertexts’ an earlier engagement with the full Organon on the part of Graeco-Syriac scholarship. A real palimpsest indicating the presence of Greek manuscripts On this letter, cf. further the remarks in my article ‘Syriac Translators and Greek Philosophy in Early Abbasid Iraq,’ Journal of the Canadian Society for Syriac Studies 4 (2004), pp. 15–26, esp. pp. 17–19. 36 O. Braun (ed./tr.), Timothei Patriarchae Epistulae I, Corpus Scriptorum Christianorum Orientalium, vol. 74/75 (Rome-Paris, 1914–1915), p. 129/86. 37 V. Berti, ‘Libri e biblioteche cristiane nell’ Iraq dell’ epistolario del patriarcho siro-orientale Timoteo I (727–823),’ in C. D’Ancona (ed.), The Libraries of the Neoplatonists (Leiden, 2007), pp. 307–317. 38 J. W. Watt, Aristotelian Rhetoric in Syriac (Aristoteles Semitico-Latinus, 18) (Leiden, 2005), pp. 20–29. 35

[11]

762

WATT

and interest in non-theological subjects (although not Aristotle) in a Syriac monastery is to be found among the manuscripts of the British Library (Add. 17,210 and 17,211). This Syriac manuscript of Severus of Antioch’s Against John the Grammarian was copied in the early ninth century by a monk named Simeon at the monastery of Qartamin (and probably taken to that of St. Mary Deipara in Egypt by Moses of Nisibis in 932). The undertext is Greek, and contains portions of the Iliad in a hand of the fifth century, of the Gospel of Luke in a hand of the sixth century, and of Euclid in a hand of the eighth or ninth century.39 There is no evidence for a Graeco-Syriac translation of Euclid,40 or of Homer before Theophilus of Edessa (according to Bar Hebraeus),41 but it is clear from this palimpsest that these Greek texts were present and doubtless studied in some learned Syriac circles at an earlier date, and discarded in the ninth century when, presumably in that location, knowledge of Greek had died out. There is no reason to suppose, therefore, that when Athanasius of Balad cited passages of the Iliad he was quoting from a Greek logic commentary or compendium42 and not translating directly from the text of the Iliad itself. Acquaintance with the Iliad in Greek on the part of earlier Graeco-Syriac scholars enables us to understand (whatever one may make of Bar Hebraeus’ reference to the translation of ‘two Books of Homer’ by Theophilus) how so much of the Iliad was known in Syriac to Antony of Tagrit.43 W. Wright, Catalogue of Syriac Manuscripts in the British Museum acquired since the Year 1838 (London, 1870–1872), pp. 548–550; S. P. Brock, ‘A Syriac Intermediary for the Arabic Theology of Aristotle? In Search of a Chimera,’ in D’Ancona, Libraries of the Neoplatonists, pp. 300–301. 40 G. Furlani, ‘Bruchstücke einer syrischen Paraphrase der ‘Elemente’ des Euklides,’ Zeitschrift für Semitistik und verwandte Gebiete 3 (1924), pp. 27–52, 212–235. 41 Baumstark, Geschichte, p. 341. 42 G. Furlani, ‘Una introduzione alla Logica Aristotelica di Atanasio di Balad,’ Rendiconti della reale Accademia dei Lincei 5, 25 (1916) [pp. 717–778], pp. 732–733, 769; idem, ‘Meine Arbeiten über die Philosophie bei den Syrern,’ Archiv für Geschichte der Philosophie und Soziologie 37 (1926), p. 8. 43 H. Raguse, ‘Syrische Homerzitate in der Rhetorik des Anton von Tagrit,’ in Paul de Lagarde und die syrische Kirchengescichte, ed. Göttinger Arbeitskreis für syrische Kirchengeschichte (Göttingen, 1968), pp. 162–175; J. W. Watt, The Fifth Book of the Rhetoric of Antony of Tagrit, Corpus Scriptorum Christianorum Orientalium, vol. 481 (Louvain, 1986), pp. xix-xx and index p. 75; idem, ‘Guarding the Syriac Language in an Arabic Environment: Antony of Tagrit on the Uses of Grammar in Rhetoric,’ in W. J. van Bekkum, J. W. Drijvers, and A. C. Klugkist (eds.), Syriac Polemics. Studies in 39

[12]

AL-FąRąBĨ

763

The evidence for an old Syriac logical cursus comprising the Organon up to but no further than Prior Analytics I 7 consists of the extant pre-ȨAbbĆsid Syriac versions of, and commentaries on, the first three of the six (or eight) treatises. A brief review of that evidence is therefore now necessary.44 We know of three translations of the Categories,45 two of the De Interpretatione, three of the Prior Analytics, and one each of the Posterior Analytics, Topics, and Sophistical Refutations.46 Of these eight translations of the first three treatises, only three might belong to the period before Athanasius of Balad, the other five being from Athanasius himself (the complete Prior Analytics), Jacob of Edessa (the Categories), and George of the Arabs (Categories, De Interpretatione, and complete Prior Analytics). The three probably earlier ones (all of uncertain date) are the Anonymous of the Categories, an Anonymous of the De Interpretatione (attributed to Probus in one manuscript47), and an Anonymous of the Prior Analytics to I 7. Even if all three belong to the sixth century rather than the seventh, this is hardly strong evidence of a universal intention among Syriac-speaking scholars to create a Syriac Organon differing from the Greek by the rejection of everything beyond Prior Analytics I 7. That the anonymous translators of the Categories and De Interpretatione did not, as far as we know, each translate more than one book of course proves nothing. As far as translations go, therefore, the evidence that the ‘Syriac tradition’ did not proceed beyond the assertoric syllogism amounts to one partial version of the Prior Analytics of uncertain date, but no earlier than the sixth century (and possibly later). However, in the seventh and subsequent centuries the entire Organon, both up to and beyond Prior Analytics I 7, was Honour of Gerrit Jan Reinink (Orientalia Lovaniensia Analaecta, 170) (Leuven, 2007), pp. 133–150, esp. 140–142. 44 Brock, ‘The Syriac Commentary Tradition,’ pp. 3–5, 12–14. 45 A translation by a Jonah is mentioned in the notes of the Paris Organon. This Jonah cannot at present be identified with any certainty. Cf. Hugonnard-Roche, La logique d’Aristote, pp. 35–36 (originally published in ‘Sur les versions syriaques des Catégories d’Aristote,’ Journal Asiatique 275 (1987), pp. 205–222). The three are the Anonymous, that by Jacob of Edessa, and that by George of the Arabs. That by Jonah might or might not be pre-ȨAbbĆsid, and Hugonnard-Roche entertains the possibility that Jonah was the translator of the Anonymous. (The Anonymous was earlier thought to be by Sergius, but that was shown to be impossible by Hugonnard-Roche, ibid., pp. 25–33.) 46 We cannot tell if the anonymous Syriac translations to which reference is made in the Paris Organon are pre-ȨAbbĆsid or not. For the enumeration above I have counted the translations by Theophilus (died 785) as ȨAbbĆsid. 47 MS. Birmingham, Mingana Syriac 606.

[13]

764

WATT

repeatedly rendered into Syriac. Given the evidence from elsewhere outside the Syriac realm of interest in an abridged curriculum, this anonymous version may be considered evidence that the abridgement was adopted by some Syrians. Thus both in Greek and Syriac we can distinguish two traditions, an abridged Organon directed towards the assertoric syllogism, and the tradition of the full Organon. Among the Syrians, however, the written study of the latter was necessarily confined to those able to read the Greek, until such time as the whole became available in Syriac. The evidence of the Syriac commentaries does not alter this conclusion. From the fact that there are no commentaries extant in Syriac of the Posterior Analytics to Sophistical Refutations the conclusion cannot be drawn that they were not studied in Syriac at least from the time of Athanasius, since the translations existed. The only Syriac commentator between Athanasius and al-FĆrĆbĩ whose works are extant is George (on Categories, De Interpretatione, and the complete Prior Analytics). From before the time of Athanasius, we have two commentaries on the Categories by Sergius, commentaries on the De Interpretatione by Paul the Persian and Probus, and a commentary by Probus on the Prior Analytics to I 7.48 Again this suggests that there was interest in the abridged Organon, but cannot be held to provide convincing evidence that the Syrians as a whole abandoned the study of the full Organon.49 And again the evidence for a termination at the close of the assertoric syllogism is based on a single witness, the commentary of Probus, a physician in Antioch of uncertain date, but certainly not prior to the sixth century. Paul the Persian may have been ignorant of Greek50 and thus belong to the ‘Syriac-alone’ tradition, but there is no doubt that with Sergius of I exclude in this enumeration scholia, letters, and treatises on particular points, and brief fragments of unidentifiable origin. These are listed in Brock, ‘The Syriac Commentary Tradition,’ pp. 13–14. Epitomes of logic are discussed separately below. 49 Cf. Stroumsa, ‘Al-FĆrĆbĩ and Maimonides,’ p. 272. Someone working merely from the number of extant Greek commentaries would have guessed that the Categories, not the Posterior Analytics, was the treatise considered to be the most important by the Greek commentators. For a convenient overview (of the editions), cf. M. Chase, ‘Les commentaries grecs et byzantins,’ in Dictionnaire des philosophes antiques: Supplément, ed. R. Goulet (Paris, 2003), pp. 113–117. 50 Hugonnard-Roche, La logique d’Aristote, pp. 234–235, n. 5 (originally in ‘Le traité de logique de Paul le Perse,’ Documenti e studi sulla tradizione filosofica medievale 11 (2000), pp. 59–82). 48

[14]

AL-FąRąBĨ

765

ReshȨaina, the fount of the Aristotelian tradition in Syriac, we are firmly in the Graeco-Syriac stream, even though Sergius himself was not responsible for the translation of any book of the Organon. Sergius, in fact, is the only certain known member of this ‘tradition’ belonging to the sixth century, given the uncertainties surrounding Paul the Persian (as to his knowledge of Greek) and Probus (as to his date and also his knowledge of Greek), the other pre-Athanasian commentators on books of the abridged Organon. That the Categories was the only book of the Organon on which he wrote an (extant) commentary does not of course prove it was the only one in which he was interested. In fact Sergius was interested in the entire Aristotelian (and the pseudo-Aristotelian) corpus, as shown for example by his Treatise on the Causes of the Universe according to the Doctrine of Aristotle (now recognised as a translation of a treatise of Alexander of Aphrodisias)51 and his translation of the pseudo-Aristotelian De mundo.52 Most significant, however, is his statement about the scope of his enterprise, and the connections he sees between the various books of the Organon, in his commentary on the Categories. The relevant passage is worth quoting at length.53 It begins with a comparison with a builder commissioned to build a hut providing protection from wind and rain. The builder’s theŇria, ‘that is, the thinking within his intellect,’ begins with the roof, is then led to the necessity of load-bearing walls to support the roof, and finally to foundations to support the walls. His praxis, however, ‘that is, the manual labour,’ begins with the foundations and ends with the roof. Therefore (2va22) the starting-point of the theŇria (‘thinking’) is the end-point of the praxis (‘action’), and the starting-point of the praxis is the end-point of the theŇria. So also Aristotle in the same way employed the logical art. First he used his intellect, (saying): ‘I wish to create a discriminating instrument which defines for me in praxis good from bad and separates

51

p. 34.

C. F. Genequand , Alexander of Aphrodisias On the Cosmos (Leiden, 2001), esp.

52 Hugonnard-Roche, La logique d’Aristote, pp. 128–129 (originally in ‘Note sur Sergius de RešȨainĆ, traducteur du grec en syriaque et commentateur d’Aristote,’ in Endress and Kruk, The Ancient Tradition, pp. 121–143). 53 I use for this the oldest manuscript of the work, London, British Library, Add. 14,658 (seventh century), the passage in question on foll. 2rb-3rb. The text has not been edited. There is a partial Italian translation by G. Furlani, ‘Sul trattato di Sergio di RêshȨaynâ circa le categorie,’ Rivista trimestrale di studi filosofici e religiosi 3 (1922), pp. 135–172, the passage in question at pp. 140–142.

[15]

766

WATT for me in the knowledge of things (2vb) truth from falsehood. But because this instrument is an art which generates all demonstrations which are established through reason (meltĆ), it is evident that I ought before it to create this art of demonstrations (taʚwyĆtĆ). And because this art of demonstrations comes from the syllogism (quʞʞĆr mamllĆ) which is composed with art, I ought beforehand to teach about it. And because, furthermore, the syllogism comes from the preceding proposition,54 I ought before that to write about this. And because, again, simple namings (kunnĆyē pšĩʞē) precede this, I am obliged before that to teach about these.’ So he began in his thinking from the art of demonstrations and proceeded in a link to simple namings; but he constructed the beginning of the teaching concerning all these from simple namings, after these taught about the proposition, after this wrote about the (3ra) correctly and properly composed syllogism, then gave instruction about the art of demonstrations, and after that about everything useful to it in any way. The end-point of his praxis on these coincided with the starting-point of his theŇria about them, just as the end-point of his theŇria of them coincided with the starting-point of his treatise on them. The book written by him about simple namings is called Categories, that which he wrote about the proposition Peri hermeneias, that about the syllogism is named Analytics, and that about the art of demonstrations is designated Apodeictics.55 Together with this there is that called Topics and that about the refutation of the sophists which was named by him Sophistici Elenchi. With these, therefore, this philosopher completed the whole logical art, which is, as we have said, an instrument of philosophy

Literally, ‘the preceding combination of terms’ (rukkĆbĆ qadmĆyĆ d-šmĆhē). K K K ¬ ¬ Text: .¿ÿØÍϚƒ ¿šÍæâ†~ƒ ÌÙàîƒ †…† .¾ØËÏ~† ¾ÙâÊøK .¾ùÙÒÍàå~ ... § .¾æÜÿâ ¾ùÙÓù؃Íñ~ (‘…Analytics, Prior and Posterior, and that about the art of demonstrations is designated Apodeictics’). I seclude ‘Prior and Posterior’ as a thoughtless insertion of a scribe, who wished to make the point that there were two treatises which could be designated Analytics, but thereby gave the impression that the Posterior Analytics and the Apodeictics were two different treatises. The Prior Analytics was frequently designated simply Analytics, the Posterior (or Second) Analytics designated Apodeictics. An alternative interpretation, taking the existing text as authentic Sergius, is to take it as a somewhat unfortunate way of expressing the fact that both books dealt with the syllogism, the Prior with the syllogism in general and the Posterior with the demonstrative syllogism, the Posterior Analytics being named thus because it dealt with the syllogism, but named Apodeictics because it dealt with demonstration. 54 55

[16]

AL-FąRąBĨ

767

and not a part of it. Some people say that (3rb) the Art of Rhetoric which was composed by him is also part of the same (art) of logic, but we will now return to the subject and start to speak as (well as) we can about the aim56 of each one of these treatises, beginning the chain with that On Categories and similarly treating each one of them in the same way. Then we will go on to his other treatises, those on the parts of praxis, all natures and mathematics, and finally those called theological.57

Sergius’ work follows the general outline of those composed by the Alexandrian commentators, and is particularly close to that of Ammonius, whose lectures he probably attended. For Ammonius and his fellow Greek commentators the Posterior Analytics was the heart of the Organon and the raison d’être for the study of the other books in the collection, either as a necessary preliminary to it (Categories, De Interpretatione, Prior Analytics), or as a protection against its misuse (Topics, Sophistical Refutations, possibly also the Rhetoric and Poetics). The section of Sergius cited above is close to corresponding passages in Ammonius,58 but not even in Ammonius is there such a graphic assertion of the interrelatedness of the first four books of the Organon as that given by Sergius. For Sergius, ceasing the study of the Organon before reaching the method of demonstration in the Posterior Analytics would be akin to constructing a house designed to offer protection from rain without putting on a roof. He certainly planned to cover the whole Organon, although we do not know whether his plan was ever achieved, and only a commentary on the Categories (‘the foundations’) has come down to us from his pen.59 No Syriac reader of this commentary, however, who followed his line of thought could possibly have supposed that the study of the first two books and the first part alone of the Prior Syriac nĩšĆ = Greek skopos. i.e., practical philosophy, physics, mathematics, and metaphysics/theology, the standard division of philosophy according to the Alexandrian commentators. Cf. e.g. D. Gutas, ‘Paul the Persian on the Classification of the Parts of Aristotle’s Philosophy: A milestone between Alexandria and BaĠdâd,’ Der Islam 60 (1983), pp. 261–263. 58 In Aristotelis categorias commentarius, ed. A. Busse (Commentaria in Aristotelem Graeca IV, 4, Berlin, 1895), pp. 428–68, 133–11, 1418–152. English translation in S. M. Cohen and G. B. Matthews, Ammonius On Aristotle’s Categories (Ithaca, New York, 1991), pp. 12–14, 19–20, 21. 59 Hugonnard-Roche, ‘Aux origins de l’exégèse orientale de la logique d’Aristote: Sergius de RešȨainĆ, médecin et philosophe,’ Journal Asiatique 277 (1989), pp. 11–12. 56 57

[17]

768

WATT

Analytics constituted the whole or the most important part of logic. In the century following Sergius (died 536) Athanasius of Balad (died 686) clearly did not think so either, and neither did Severus Sebokht (died 666).60 We have no reason to suppose that in the interim the abridged Organon constituted the one and only ‘Syriac tradition’ and that the full Organon suddenly reappeared among Syriac scholars in the time of Severus and Athanasius. It is much more likely that in Graeco-Syriac scholarship, as also in Greek, there existed two curricula of logic, one treating the material of the full Organon, the other (also known in Latin)61 dedicated primarily to the assertoric syllogism and focussing on the earlier part up to Prior Analytics I 7. While both strands were accessible to the Graeco-Syriac scholars, those unable to read Greek (or without a teacher able to read Greek) were evidently dependent upon whatever Syriac translations or commentaries were available to them. Sergius himself, as far as we know, made no translation of any book of the Organon, and while it remains possible that translations existed for parts of it beyond Prior Analytics I 7 before those made by Athanasius, we have no evidence of them. It is quite likely that even in the seventh century many Syrians who were interested in logic were able to read both Greek and Syriac, but during that century or possibly even earlier, as knowledge of Greek became more restricted, a corpus of logical works seems to have developed limited by the Aristotelian translations existing or known in Syriac at the time. The evidence for this corpus is provided by those Syriac manuscripts containing a body of writings on logic. The earliest of them may be Add. 14,658 of the British Library 60 In his Discourse on (categorical) Syllogisms in the Prior Analytics Severus wrote that ‘syllogistic is useful … for the perfect knowledge of the logical and demonstrative theŇria of those things said in the Book of Apodeictics (i.e., Posterior Analytics), the (Book) in which truth is precisely demonstrated and distinguished from falsehood by means of the logical art.’ See Wright, Catalogue, p. 1160b, and cf. D. Miller, ‘George, Bishop of the Arab Tribes, on True Philosophy,’ Aram 5 (1993) [pp. 303–320], pp. 311–312. For Severus as for Sergius, as also for the Baghdad Aristotelians and no doubt for Athanasius, Severus’ pupil at Qenneshre, the syllogistic of the Prior Analytics was not an end in itself, but the route to the demonstrative art of the Posterior Analytics. 61 See above on the Latin tradition (Apuleius, Marius Victorinus, Jerome, Boethius, and the ‘Old Logic’). The abridged tradition appears also to have been known in Armenian. Cf. V. Calzolari, ‘Aux origins de la formation du corpus philosophique en Arménie: quelques remarques sur les versions arméniennes des commentaries grecs de David,’ in C. D’Ancona (ed.), The Libraries of the Neoplatonists (Leiden, 2007), pp. 261–263. Cf. above, n. 12.

[18]

AL-FąRąBĨ

769

(seventh century), but other important and more convincing witnesses are Vatican syr. 158 (ninth or tenth century) and its descendants, Berlin syr. 88 (dated 1260) and 89, and Mingana syr. 606 (modern). The translations of Aristotle by Athanasius, George, and Theophilus were unknown to or ignored by this corpus, which (although in the case of the London manuscript includes Sergius’ commentary on the Categories) evidently belongs, whether by choice or necessity, to the tradition of the abridged Organon. In addition to translations and commentaries on the books up to Prior Analytics I 7, it also includes translations and commentaries on the Eisagoge (including in some manuscripts the translation by Athanasius), works of grammar, lives of Aristotle, and brief treatises and scholia on assertoric syllogisms.62 It is impossible to say with any confidence whether this corpus or the tradition of the full Organon was more widely known in Syriac scholarly circles between the seventh and tenth centuries. It is this corpus and these manuscripts which have led to the conclusion that ‘compared to the plethora of commentaries on the earlier books of the Organon, the evidence for interest in the later books is admittedly meager.’63 Simply counting the extant manuscripts, however, without identifying the tradition to which they belong or the circumstances leading to their production and preservation, does not yield a certain result. On the other side, it should be noted that the translations of Athanasius and Theophilusnot now extant were known in Baghdad to the Syro-Arabic Aristotelians who produced the Paris Organon (although there is no evidence that they knew those of George, which are still extant).64 Neither is there any evidence that they had any knowledge of or interest in the corpus embodying the tradition of the abridged Organon and the anonymous early translations which it utilised.65 What can be said with some certainty is that 62 Hugonnard-Roche, La logique d’Aristote, pp. 91–97 (originally in ‘Les traductions syriaques de l’Isagoge de Porphyre et la constitution du corpus syriaque de logique,’ Revue d’histoire des texts 24 (1994), pp. 293–312). 63 Stroumsa, ‘Al-FĆrĆbĩ and Maimonides,’ p. 271. 64 The information contained in the Paris manuscript of the Organon as to what the Baghdad scholars knew of the pre-ʗunain Syriac translations and commentaries cannot be assumed to be complete; what is recorded there is essentially a matter of chance. Cf. Hugonnard-Roche, ‘L’intermédiaire syriaque,’ pp. 190–192, and ‘Les traductions du grec au syriaque,’ pp. 134, 140–143. The same, of course, applies to the information contained in Timothy’s letters. 65 The translation of the Categories by Jacob of Edessa is the one translation of Aristotle which was assuredly used both by the Baghdad Aristotelians and the

[19]

770

WATT

both traditions existed within the Syriac-speaking realm, although both were not necessarily known to each and every group.66 To complete the picture, however, one further strand requires mention, that of the epitomes or introductions to logic or philosophy. The two significant examples we possess in Syriac, by Paul the Persian and Athanasius of Balad, exhibit the exclusive or predominant interest in the assertoric syllogism of the tradition of the abridged Organon. That by Paul covers the ground of the Eisagoge, Categories (sections 3–12 in the edition of Land),67 De Interpretatione, and Prior Analytics to I 7 (sections 13–26),68 and that by Athanasius similar ground, touching only very briefly on the material of the Sophistical Refutations. Yet Paul was fully aware of the complete Organon, if the entire text on the classification of the parts of Aristotle’s Syriac corpus of logic based on the abridged Organon. Cf. Walzer, Greek into Arabic, pp. 68–69, 71. In addition to Aristotle’s treatises Athanasius, as noted above, also translated the Eisagoge, and this translation is also represented in the Syriac abridged corpus (cf. Hugonnard-Roche, La logique d’Aristote, pp. 81–91). Cf. also below, n. 74. 66Although the anonymous (probably sixth century) translation of the Prior Analytics to I 7 is likely to have been produced with a curriculum of an abridged Organon in mind, on the basis of the manuscripts representing such a curriculum it cannot be proven that this curriculum is earlier than the seventh century or prior to the creation of a complete Syriac Organon by Athanasius (d. 686). The earliest of the manuscripts assigned by Hugonnard-Roche to this group, London Add. 14658 (cf. above, n. 62, La logique d’Aristote, pp.95–96) belongs to the seventh century, but it differs in significant respects from the others. For example, the only book of Aristotle with which it deals is the Categories (the anonymous translation and Sergius’ commentary), whereas the others all include at least two, and often three, of the sequence Categories—Prior Analytics I 7, either a translation (in most cases Jacob of Edessa’s translation of the Categories) or a commentary. Furthermore, Add. 14658 is not primarily a philosophical manuscript, but contains other material unlike that found in the others. It is therefore very doubtful whether this manuscript should be included within this group, and certainly it cannot be held to prove that the curriculum of the abridged Organon (i.e., the sequence Categories— Prior Analytics I 7) already existed in the seventh century. The earliest of the other manuscripts is Vaticanus syr. 158, of the ninth or tenth century, whereas Timothy in the late eighth century clearly knew the tradition of the complete Organon. It is therefore by no means certain that a curriculum of the abridged Organon in Syriac represented by these manuscripts was widely known earlier than a curriculum of the complete Organon available in Syriac from the time of Athanasius. 67 J. P. N. Land, Anecdota Syriaca, vol. IV (Leiden, 1875). 68 Hugonnard-Roche, La logique d’Aristote, pp. 233–254.

[20]

AL-FąRąBĨ

771

philosophy in an Arabic treatise of Miskawayh is indeed from his hand, and was in agreement with the Alexandrian tradition (including Sergius) in declaring the noblest of the books to be the Posterior Analytics, the first three books of Aristotle being its preparation and the last four (i.e., including the Rhetoric and Poetics) its ‘protection.’69 Athanasius’ Introduction to the Logic and Syllogistic of Aristotle, an Epitome, is designed, as he says at its beginning, to offer the reader an account of the syllogistic science through short extracts (karyĆtĆ), because the logic pragmateia of Aristotle is very difficult on account of his stringent probing (mnassyĆnştĆ). It deals with some themes of the Categories, De Interpretatione, and the assertoric figures of Prior Analytics I 1–7, with some brief concluding remarks on the construction of premisses (Pr. Anal. I 27), on reduction ad impossibile, and on sophisms.70 The fact that Athanasius translated texts of Aristotle (Prior Analytics beyond I 7, Posterior Analytics, Topics, and Sophistical Refutations ) to which he made little or no reference in this introductory epitome on logic, is clear evidence that extant Syriac texts devoted only to the abridged Organon are no proof that interest in the study of the complete Organon was totally absent among the Syrians. Athanasius’ Introduction on the one hand and his translations of Aristotle texts on the other illustrate the double strand of logical study in Syriac, as the translations and monographs of Boethius (referred to above) similarly do so in Latin.71 If therefore a measure of historical reliability is to be sought in alFĆrĆbĩ’s assertion that ‘the books on logic were to be taught up to the end Gutas, ‘Paul the Persian,’ pp. 231–234, 238–250. On the terminology used to describe the function of the last four books, cf. in general Gutas, ibid., pp. 242, n. 24, 264–266; D. L. Black, Logic and Aristotle’s Rhetoric and Poetics in Medieval Arabic Philosophy (Leiden, 1990) pp. 34–36, 50; and for Sergius in particular, the passage from his commentary on the Categories cited above (‘and after [demonstration] about everything useful to it in any way’). 70 G. Furlani, ‘Una introduzione alla Logica Aristotelica di Atanasio di Balad,’ pp. 716–766 (introduction and text), 767–778 (summary in Italian). 71 Mutatis mutandis, the same point may be made with reference to the activity of Severus Sebokht. While from his hand we have an extant Discourse on (categorical) Syllogisms in the Prior Analytics, he clearly regarded the Posterior Analytics as the most important book of the Organon. Cf. above, n. 60. The significant difference between the Syriac and Latin traditions, however, is that while the sixth century translations by Boethius of the Posterior Analytics and other later books of the Organon only became known again in Latin during the twelfth, Athanasius’ seventh century Syriac translations were certainly known (to Timothy and the Baghdad Aristotelians) during the eighth to eleventh centuries. 69

[21]

772

WATT

of the assertoric figures, but not what comes after it,’ it will have to be said that he reflects an important half-truth, namely, one strand of a two-strand tradition. But the question has to be asked as to how he knew of this tradition of an abridged Organon, the origin of which he locates (seemingly correctly, in the light especially of the Latin evidence discussed above) not in the Syriac linguistic area, but in the Greek (which for al-FĆrĆbĩ implied Alexandria). His contacts with (and dependence upon) the Syro-Arabic Aristotelians of the school of Abş Bišr MattĆ in Baghdad are well known, and the suggestion has therefore been advanced that he ‘reflects the practice of the “ancient Syriac tradition” as he heard it from his Syriac associates.’72 The suggestion has the merit of making explicit what seems often to be implicitly assumed, but it is entirely speculative and not without its difficulties. Apart from the fact that study of the abridged Organon is not the ancient Syriac tradition, but merely one strand in it, there is the problem as to how his Syriac associates, the Baghdad Aristotelians, could have known of this strand. What they certainly knew was the tradition of the entire Organon in Syriac translations, the translations of the school of ʗunain and the earlier ones of Athanasius of Balad, Jacob of Edessa, and Theophilus of Edessa. There is no evidence that they knew those probably even earlier ones of the Categories, De Interpretatione, and Prior Analytics to I 7,73 the translations on which modern scholarship bases the theory of a Syriac school curriculum of an abridged Organon. Even if they knew some of the manuscripts of the logical corpus containing these translations and limited to this material, they could hardly have deduced merely from them that an ancient Syriac school tradition existed with this limitation,74 and that they knew an oral tradition to this effect, while not impossible, is nevertheless pure speculation. If they knew of any school centuries before their own time, it is most likely to have been that of Qenneshre, for they had translations stemming from its members, namely those by Athanasius and Jacob of Edessa, but from these scholars of Qenneshre they inherited a Gutas, ‘Alexandria to Baghdad,’ p. 185. Walzer, Greek into Arabic, pp. 68–69. 74 As noted above (n. 65), of the Aristotle translations only that of the Categories by Jacob of Edessa was known both to the Baghdad Aristotelians and the Syriac manuscripts which contain a corpus of logical writings based on the abridged Organon. The Baghdad Aristotelians apparently knew of some ‘old (i.e. pre-ʗunain) Syriac versions’ (al-suryĆnĩ fĩ nuqşl qadĩma) of the Eisagoge, but it is not known to which versions this refers; cf. Georr, Catégories, pp. 193–194. It may include the version of Athanasius. 72 73

[22]

AL-FąRąBĨ

773

complete Syriac Organon.75 Neither is there any evidence that Timothy, active in Baghdad a century or more before the heyday of the Baghdad School, knew anything of a Syriac scholastic tradition terminating the study of the Organon at Prior Analytics I 7. Yet since al-FĆrĆbĩ’s assertion has an element of truth in it, we must assume it is based on something in his environment, and if it did not come to him from his Syriac associates,76 it is likely to have come from his Arabic ones. From the Arabic environment he could have known about epitomes of logic and an abridged Organon, not simply because before the work of Abş Bišr MattĆ some of the books (the Posterior Analytics and the Poetics) had not yet been translated into Arabic, but because (as far as we can judge) the earliest and most popular Arabic book on logic was the epitome composed by (or attributed to) Ibn al-MuqaffaȨ. After a paraphrase of the Eisagoge, the author of this treatise indicates he will proceed with four books, beginning with the Categories, but in fact he deals only with the material to the end of the assertoric syllogism, which he considers the end of the Analytics, maintaining (quite speciously) that the most important points of the Apodeictics have already been treated at the beginning of his treatise.77 On the basis of its technical vocabulary, this treatise can certainly be said to be dependent in some way on Syriac tradition.78 Its content places it firmly within the epitome tradition of the abridged Organon,79 although with his 75 It is appropriate to recall here what al-FĆrĆbĩ did say about the Syriac tradition, which did probably reflect what he heard from his Syriac associates at Baghdad: ‘Everything comprised by this science was expounded in the Greek language, then in the Syriac language, then in the Arabic language’ (cf. above, n. 17). 76 We may recall here the point that al-FĆrĆbĩ’s account does not attribute the truncation of the Organon to Syriac tradition, but to Alexandrian. Cf. above, n. 15. 77 Ibn al-MuqaffaȨ, Al-Manʞiq, ed. M. T. DĆneshpazhşh (Tehran, 1978), pp. 93– 5, 9315–18; cf. Gutas, ‘Alexandria to Baghdad,’ p. 183. Cf. also Kraus, ‘Zu Ibn alMuqaffaȨ,’ pp. 7–8. 78 Hugonnard-Roche, ‘L’intermédiaire syriaque,’ pp. 203–204. 79 The point made by Gutas, ‘Alexandria to Baghdad,’ p. 184, that the author writes as if the Analytics (i.e., Prior Analytics) ended at I 7 (‘Here ends the Book of Analytics,’ Arabic ΎϘϴρϮϟϮϧ΍ ΏΎΘϛ ϢΗ) also links it with the Syriac tradition of the abridged Organon. Cf. the title (‘The First Book of the Analytics of Aristotle,’ Syriac K êÙàÓÓéؘ~ƒ Ìà؃ ¾ùÙÒÍßÍå~ƒ ¾ÙâÊø ¾ÁÿÜ) and the subscription (‘Here ends the Book of the Analytics of Aristotle,’ Syriac êÙàÓÓéؘ~ƒ ¾ùÙÒÍßÍå~ƒ ¾ÁÿÜ äàü) in the anonymous translation of Prior Analytics to I 7, ed. A. Nagy, ‘Una versione siriaca inedita degli Analitici di Aristotele,’ Rendiconti della Reale Accademia dei Lincei, Classe di scienze morali, storiche e filologiche 5, 7 (1898), pp. 322 and 347.

[23]

774

WATT

reference to ‘the four books’ (none of which he had probably read, and none of which had probably been translated into Arabic in his time) the author shows he is aware of the existence of the full tradition.80 The same 80 ‘The four (books)’ (Categories to Posterior Analytics) referred to here by Ibn alMuqaffaȨ (d. 757) are hardly an allusion to a Greek abridged Organon of ‘three and a bit’ books with Prior Analytics I 8–22 or I 8–II 27 omitted (according to the evidence of Ibn al-ʙalĆʘ, below), parallel to a Syriac abridged Organon of ‘two and a bit’ books (Gutas, ‘Alexandria to Baghdad,’ pp. 183–185). The ‘four books’ are, rather, a reference to the first and main part of the complete Organon (dealing respectively with predicates, propositions, syllogisms, and demonstrations). The tradition of the complete Organon (whether Greek or Syriac) always regarded these as the essential part, and the last two or four books (if the Rhetoric and Poetics are included) only as a useful addendum; cf. above n. 69. Patriarch Timothy (d. 823) knew the full Organon, but his request for commentaries or scholia on the Topics, Sophistical Refutations, Rhetoric, and Poetics (cf. above n. 33) suggests that he recognised these last four as a unit within the greater whole; Gutas, ibid., p. 173, cites a ‘four plus four’ division by al-YaȨqşbĩ (c. 872). The study of the full logic curriculum was thus essentially completed with the first four books (as graphically represented by Sergius in the passage from his commentary on the Categories cited above, where the Posterior Analytics is compared to the roof of a building); the remaining four were not devoted to the differentiation of truth from falsehood per se, but to debate (Topics), refutation of error (Sophistical Refutations), and communication with the multitude (Rhetoric and Poetics). Whatever truth there may be, however, in the statement of Ibn RiʡwĆn (d. 1068) that in Alexandria the medical curriculum included the first four books (and only the first four) of the Organon (cf. Strohmaier, ‘Von Alexandrien nach Bagdad,’ p. 386; Gutas, ibid., pp. 164, 174), it is clear from all the philosophers (Greek Alexandrian commentators, Sergius, or Paul the Persian) that philosophical study was not taken to be limited to them alone. There is no reason to suppose, especially given the range of the translations by Athanasius (d. 686), that from the later sixth century philosophical study was restricted to them, even though they were clearly regarded as the most important part. Gutas, ibid., pp. 181–183, puts together the ‘four books’ with the statement of Ibn al-ʙalĆʘ (d. 1153) in his Treatise on the Fourth Figure that modal syllogisms were placed in a separate section of study by ‘the ancients,’ and deduces thereby that Ibn al-ʙalĆʘ knew of the existence of a Greek abridged Organon of ‘three and a bit’ books which arose in Alexandria (according to Gutas in the sixth century). He considers Ibn alʙalĆʘ to have been independent of al-FĆrĆbĩ (ibid., pp. 181–182), but that is not certain (cf. Stroumsa, ‘Al-FĆrĆbĩ and Maimonides,’ pp. 273–274). On the lack of an explicit mention of the Alexandrians in al-FĆrĆbĩ, cf. above n. 15; as for the attribution of the truncation in al-FĆrĆbĩ to Christians and in Ibn al-ʙalĆʘ to ‘the ancients’ (Gutas’s other argument), the latter may simply have modified al-FĆrĆbĩ’s story. Even if the two texts are independent, Ibn al-ʙalĆʘ’s statement may mean no more than that modal syllogisms were studied differently from assertoric, with no

[24]

AL-FąRąBĨ

775

can be said of the sections on logic in al-Kindĩ’s RisĆla on the Number of the Books of Aristotle.81 While aware of the eight books of logic, his epitome scarcely touches on anything beyond Prior Analytics I 7.82 If al-FĆrĆbĩ knew the curriculum of the full Organon from his Christian Syro-Arabic teachers, and an alternative curriculum ending with the assertoric figures from his fellow Muslim Arabs, he faced a delicate task in explaining to the latter why he so enthusiastically embraced the curriculum taught by the former.83 It is not hard to see in these circumstances a motive implications about the extent of the curriculum as a whole. However, if Gutas’s interpretation of Ibn al-ʙalĆʘ is correct, the alleged Greek abridged Organon of ‘three and a bit’ books is not the abridged Organon of the Syriac tradition, or alFĆrĆbĩ, or Ibn al-MuqaffaȨ. The Peri Hermeneias of Apuleius, a work dating probably from the second century AD, and the other Latin evidence mentioned at the outset of this article are evidence that a curriculum of logic based on subject matter ‘up to the end of the assertoric figures’ did not originate in the sixth century and was not confined to (or necessarily originate in) Alexandria. It is questionable whether a twelfth century author can be taken as a reliable guide to scholastic developments many centuries earlier, especially as there is no evidence for a curriculum of ‘three and a bit’ books in any earlier text. 81 Edited with Italian translation by M. Guidi and R. Walzer, ‘Uno scritto introduttivo allo studio di Aristotele,’ Studi su al-Kindi I, Atti della Reale Accademia Nazionale dei Lincei, Memorie della Classe di scienze morali, storiche e filologiche, 6, 5, 5 (1940), pp. 375–419. 82 In Chapter III al-Kindĩ devotes 15 lines (in the edition of Guidi and Walzer) to the Categories and a mere 9 (III, 16–24) to the remaining eight books; in Chapter IX he devotes around 10 lines each to the Categories and De Interpretatione and 16 (IX, 20–35) to the Prior Analytics (ΔϠγήϤϟ΍ ϊϣ΍ϮΠϟ΍, categorical syllogisms, the subject matter of I 1–7), but in Chapter X a mere 13 to the remaining five books (with less than 4 to the Posterior Analytics). Equally significant is the fact that the architectonic thrust of the full Organon tradition is lacking in his epitome, namely, the ‘ascent’ to the Posterior Analytics and the subsequent ‘descent’ to the Sophistical Refutations and Poetics. There is no evidence for the epitomes or commentaries on all books of the Organon attributed by the Fihrist to al-Kindĩ (G. Flügel (ed.), KitĆb al-Fihrist [Leipzig, 1871], pp. 248–250; B. Dodge (tr.), The Fihrist of al-Nadĩm [New York, 1970], pp. 598–602) apart from this risĆla. The version of the Prior Analytics probably made for him by Ibn al-Biʜrĩq (G. Endress, ‘The Circle of al-Kindĩ,’ in Endress and Kruk, The Ancient Tradition, p. 58) is known only at I 3 (25b17); cf. Walzer, Greek into Arabic, p. 85, Badawĩ, Manʞiq Arisʞş, p. 112 n. 5. We do not know, therefore, whether or not it extended beyond I 7. 83 That the narrative in Ibn Abĩ UʛaibiȨa indicates that he studied with YşʘannĆ ibn ʗaylĆn to the end of the Posterior Analytics does not mean that his studies, or those of his teachers, ceased there (Gutas, ‘Alexandria to Baghdad,’ p.

[25]

776

WATT

for his story. His own writings were directed to both Muslims and Christians, and he was therefore required to exercise a degree of sensitivity towards both. To his Christian teachers he acknowledged his gratitude for instruction in the Posterior Analytics, the heart of the complete Organon, and in putting the ‘blame’ for the Organon’s truncation on the emperors and bishops of the Byzantine Empire he no doubt felt rightly that that would arouse little or no antagonism from East and West Syrian Christians living under the dominion of Islam. On the other hand, Muslims familiar with the restricted range of logic presented in Ibn al-MuqaffaȨ or al-Kindĩ now needed to be persuaded to move on to a wider canvas dominated by a treatise only recently made available in Arabic, and a curriculum hitherto largely the preserve of Christians.84 Al-FĆrĆbĩ would therefore have been happy to grant to both curricula an Alexandrian birthplace,85 not wishing to denigrate the abridged tradition, but doubtless implying that only the complete tradition was transmitted by that chain (isnĆd) leading from

185). There is no indication, for example, in his Philosophy of Aristotle, that the end of the Posterior Analytics marked such a termination (cf. Mahdi, Alfarabi: Philosophy of Plato and Aristotle, pp. 71–130, with the transition from the Posterior Analytics to the Topics at p. 87). Abş Bišr MattĆ translated from the complete Syriac Organon both the Posterior Analytics and the Poetics. 84 Since Arabic translations of the complete Prior Analytics, Topics, Sophistical Refutations, and Rhetoric existed prior to al-FĆrĆbĩ (cf. e.g. Peters, Aristoteles Arabus, pp. 14–28), it was clearly possible for Arabic Muslims before al-FĆrĆbĩ to have been acquainted with more of the Organon than is implied by ‘up to the end of the assertoric figures, but not what comes after it.’ The lack of an Arabic version of the Posterior Analytics before that made by Abş Bišr MattĆ, and the evidence of the works of Ibn al-MuqaffaȨ and al-Kindĩ, suggest, however, that the curriculum of the full Organon would still have been perceived by al-FĆrĆbĩ (and his contemporaries) as characteristic of the teaching of the Syro-Arabic Aristotelians. The Arabic translations made prior to those of Abş Bišr MattĆ did not produce a complete (six-volume) Arabic Organon, as Athanasius’ translations of Prior Analytics, Posterior Analytics, Topics, and Sophistical Refutations produced (together with Syriac versions of Categories and De Interpretatione) a complete (six-volume) Syriac Organon. It is perhaps not inappropriate to compare the introduction of the complete Organon into Arabic from Syriac in the tenth century with the introduction of the Logica Nova into Europe from Arabic in the twelfth. 85 Inasmuch as a curriculum directed to the assertoric figures emerged in antiquity, al-FĆrĆbĩ was not totally wrong, but only by chance. It did not necessarily originate in Alexandria, and if Apuleius is taken as a guide, it arose before the period of Christian rule.

[26]

AL-FąRąBĨ

777

Alexandria to his own teachers.86 He could, however, endeavour to remove any sting felt by Muslims at the apparent Christian ‘ownership’ of the full Organon by claiming that its truncation had arisen from the hostility of Christian rulers to philosophya theme already familiar in his environment87and had only been reversed under the rule of Islam. Access Cf. Strohmaier, ‘Von Alexandrien nach Bagdad,’ pp. 388–389. In some earlier articles touching on this theme I considered as plausible both the hypothesis of the creation of the linear chain ‘from Alexandria to Baghdad’ by al-FĆrĆbĩ himself, and the alternative hypothesis that he took it over from his Christian teachers (thus Zimmermann, Al-Farabi’s Commentary, p. cxi). Further reflection leads me to believe that only the former is tenable. The Baghdad Aristotelians knew from the Syriac versions at their disposal (especially those from Qenneshre) that the transmission was not through a ‘linear’ migration, and they had no need to justify to other Christians their teaching of the full Organon; that was already accepted among Christians, as is evident from the witness of Timothy. Al-FĆrĆbĩ, by contrast, had no access to evidence of a ‘diffuse’ transmission of Aristotelian logic in the centuries before his time, and every incentive to create a school tradition legitimising his teaching, as Strohmaier has indicated, and indeed legitimising it over against the truncated epitome teaching terminating at the assertoric syllogism, which was all that was previously well known among Muslims. I doubt we shall ever know why al-FĆrĆbĩ named Antioch as the intermediate station on his imaginary migration of the school from Alexandria to Baghdad; but faute de mieux let me offer the suggestion that the Baghdad Aristotelians knew that the first monks of Qenneshre had migrated from St. Thomas at Seleucia near Antioch (and the even more speculative suggestion that the mention of the man from ʗarrĆn, and in alMasȨşdĩ ʗarrĆn itself, has something to do with Qenneshre). 87 Gutas, ‘Alexandria to Baghdad,’ pp. 176–178. It is much more likely that alFĆrĆbĩ took over this widespread topos than that his attribution of the abridgement to the hostility of Christian bishops in the Christian Empire reflected his own hostility to kalĆm and a low opinion on his part of his Christian teachers and contemporaries whom he regarded as Christian mutakallimşn, as suggested by Stroumsa, ‘Al-FĆrĆbĩ and Maimonides,’ pp. 276–285. It may be the case that he and YaʘyĆ ibn ȨAdĩ disagreed over the question of the eternity of the world, but there is much else on which they had a great deal in common; cf. my ‘Christianity in the “Renaissance of Islam”: Abş Bishr MattĆ, al-FĆrĆbĩ, and YaʘyĆ ibn ȨAdĩ ,’ in M. Tamcke (ed.), Christlich-muslimische Gespräche in Mittelalter (Beiruter Texte und Studien, 117) (Beirut-Würzburg, forthcoming). It was, after all, his Christian teachers and contemporaries to whom al-FĆrĆbĩ owed his knowledge of the complete Organon. The refutation of Aristotle by Christian philosophers had to do not with the Organon, but with the teaching on the eternity of the world (and thus with treatises such as the Physics and Metaphysics). The close connections between al-FĆrĆbĩ and YaʘyĆ ibn ȨAdĩ and his brother are scarcely conceivable if Maimonides’ adverse 86

[27]

778

WATT

to the full Organon was therefore not merely to be attributed to Christian translators and scholars, but above all to Islamic rulers more enlightened than their Christian predecessors. The abridged, apparently ‘Muslim,’ Organon owed its genesis to Christian rulers, the genuine, apparently ‘Christian,’ Organon its restoration to the rule of Islam. Viewed in this light, al-FĆrĆbĩ’s assertion about the study of the Organon sheds light primarily not on the situation centuries before his time, but on that of his own perioda judgement that applies in general to the whole ‘fiktive Schultradition.’

opinion of YaʘyĆ is a reflection of that of al-FĆrĆbĩ; cf. Gutas, ibid., pp. 164, n. 34; 167, n. 46; 176–179.

[28]

ADDENDA ET CORRIGENDA p. 755, n. 17. The Arabic reference (p. 3814-17) is to the Hyderabad edition (1345 A.H./1926 A.D.). The correct reference to the edition by J. al-YĆsĩn is p. 8810-15. pp. 755-6, n. 19. See also H. Daiber, ‘Die Aristotelesrezeption in der syrischen Literatur’, in D. Kuhn and H. Stahl (eds.), Die Gegenwart des Altertums: Formen and Funktionen des Altertumsbezugs in den Hochkulturen der Alten Welt (Heidelberg, 2001), pp. 332-335, who also notes that in Syriac literature the limitation of Aristotelian logic to the material up to Prior Analytics I 7 is not uniform. Daiber attributes the initial absence of Syriac translations of the remainder of the Organon to an orientation of Syriac logical studies towards the Church Fathers. I consider it more likely that it was to do with the status of the Greek text as the basis of advanced study of Aristotle, to which literal Syriac versions were added in later times (by Athanasius of Balad and others) as an aid for Syriac speakers in the reading of the Greek, or as a ‘mirror’ of the Greek, which could be paraphrased in more idiomatic Syriac by a teacher, for those who could not read the Greek text. Cf. my article ‘Commentary and Translation in Syriac Aristotelian Scholarship’ in the forthcoming Proceedings of the Symposium Syriacum 2008, to appear in Parole de l’Orient. p. 767. The statement in the text (following n. 58) that ‘not even in Ammonius is there such a graphic assertion of the interrelatedness of the first four books of the Organon as that given by Sergius’ is too abrupt. The comparison between the construction of a house and the building of the Organon is certainly not present in the commentary issued under the name of Ammonius (referred to in n. 58), but may nevertheless have been part of his teaching as it appears in Philoponus, In Aristotelis categorias commentarium, ed. A. Busse (Commentaria in Aristotelem Graeca XIII 1, Berlin, 1898) pp. 10211133.

[29]