Ibn Sina and his Influence on the Arabic and Latin World (Variorum Collected Studies) [1 ed.] 086078987X, 9780860789871

Citation preview

Also in the Variorum Collected Studies Series:

RICHARD M. FRANK (Ed. Dimitri Gutas)

Philosophy, Theology and Mysticism in Medieval Islam Texts and Studies on the Development and History of Kalam, Vol. I

RICHARD M. FRANK (Ed. Dimitri Gutas)

Early Islamic Theology: The Mu'tazilites and al-Ash'ari Texts and Studies on the Development and History of Kalam, Vol. II

RICHARD M. FRANK (Ed. Dimitri Gutas)

Classical Islamic Theology: The Asharites Texts and Studies on the Development and History of Kalam, Vol. Ill

STEPHEN GERSH

Stephen Gersh From Ancient Commentary to Medieval Reception

DIMITRI GUTAS

Greek Philosophers in the Arabic Tradition

JOHN MARENBON

Aristotelian Logic, Platonism, and the Context of Early Medieval Philosophy in the West

FERNAND BRUNNER

Metaphysique d'Ibn Gabirol et de la tradition platonicienne

GABRIEL NUCHELMANS

Studies on the History of Logic and Semantics, 12th-17th Centuries

JAMES MCEVOY

Robert Grosseteste, Exegete and Philosopher

VALERIE IJ. FLINT Ideas in the Medieval West Texts and Their Contexts

GEORGES VAJD A

Etudes de theologie et de philosophic arabo-islamiques a 1'epoque classique

WILLIAM J. COURTENAY

Covenant and Causality in Medieval Thought Studies in Philosophy, Theology and Economic Practice

VARIORUM COLLECTED STUDIES SERIES

Ibn Sina and his Influence on the Arabic and Latin World

Jules Janssens

Ibn Sma and his Influence on the Arabic and Latin World

Routledge

Taylor & Francis Group >ON AND NEW YORK

First published 2006 by Ashgate Publishing Published 2016 by Routledge 2 Park Square, Milton Park, Abingdon, Oxon OX 14 4RN 711 Third Avenue, New York, NY 10017, USA Routledge is an imprint of the Taylor & Francis Group, an informa business This edition © 2006 by Jules Janssens Jules Janssens has asserted his moral right under the Copyright, Designs and Patents Act, 1988, to be identified as the author of this work. All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. Notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe. British Library Cataloguing in Publication Data Janssens, Jules L. Ibn Sina and his influence on the Arabic and Latin world. (Variorum collected studies series) 1. Avicenna, 980-1037 2. Avicenna, 980-1037-Influence 3. Philosophy, Medieval 4. Philosophy, Islamic I.Title 181.9'2 Library of Congress Cataloging-in-Publication Data Janssens, Jules L. Ibn Sina and his influence on the Arabic and Latin world p. cm. — (Variorum collected studies series ; CS843) In English and French. ISBN 0-86078-987-X (alk. paper) 1. Avicenna-History. I. Title. II. Collected studies ; CS843. B751.Z7J36 2006 181'.5-dc22

2005057080

ISBN 9780860789871 (hbk) Transferred to Digital Printing in 2012 VARIORUM COLLECTED STUDIES SERIES CS843

CONTENTS Preface Acknowledgements I

Ibn SIna, and his heritage in the Islamic world and in the Latin West First publication

II

Ibn Slna's ideas of ultimate realities: Neoplatonism and the Qur'an as problem-solving paradigms in the Avicennian system Ultimate Reality and Meaning 10. Toronto, 1987

ix-xi xii

1-14

252-271

III

The problem of human freedom in Ibn SIna Actes del Simposi Internacional de Filosofia de I 'Edat Mitjana. Elpensament anthropologic medieval en els ambits islamic, hebreu i cristia, ed. P. Llorente et al. Vic: Patronat d'Estudis Osonencs, 1996

112-118

IV

Creation and emanation in Ibn SIna Documenti e Studi sulla Tradizione Filosofica Medievale 8. Florence, 1997

455^77

V

Ibn SIna (Avicenne): un projet "religieux" de philosophic? Was ist Philosophie im Mittelalter?, ed. J.A. Aertsen and A. Speer. Berlin/New York: Walter de Gruyter, 1998

863-870

VI

Les Ta 'liqat d'Ibn SIna: essai de structuration et de datation Languages et Philosophie. Hommage a Jean Jolivet, ed. A. de Libera, A. Elamrani-Jamal, A. Galonnier. Paris: Librairie Philosophique J. Vrin, 1997

109-122

VII

Le Danesh-Nameh d'Ibn SIna: un texte a revoir? Bulletin de philosophic medievale 28. Louvain-la-Neuve, 1986

163-177

VI

CONTENTS

VIII

Le ma 'drij al-quds fi maddrij ma 'rifat al-nafs: un element-cle pour le dossier GhazzalT-Ibn Slna? Archives d'histoire doctrinale et litteraire du moyen age 60. Paris, 1993

IX

Al-Ghazzall's Milydr al- lilm fifann al-mantiq: sources avicenniennes et farabiennes

Archives d'histoire doctrinale et litteraire du moyen age 69. Paris, 2002

X

Al-Ghazzah's Tahafut: is it really a rejection of Ibn Sina's philosophy? Journal of Islamic Studies 12. Oxford, 2001

XI

Al-GazzalT, and his use of Avicennian texts Problems in Arabic Philosophy, ed. M. Maroth. Piliscsaba: The Avicenna Institute of Middle Eastern Studies, 2003

XII

Bahmanyar ibn Marzuban: a faithful disciple of Ibn Slna? Before and After Avicenna. Proceedings of the First Conference

27-55

39-66

1-17 37-49

177-197

of the Avicenna Study Group, ed. D.C. Reisman, with the assistance ofA.H. Al-Rahim. Leiden/Boston: Brill, 2003

XIII

Mulla Sadra' s use of Ibn Sma's Ta 'liqat in the Asfdr Journal of Islamic Studies 13. Oxford, 2002

XIV

L'Avicenne latin: particularites d'une traduction Avicenna and His Heritage, ed. J. Janssens and D. de Smet. Leuven: Leuven University Press, 2002

XV

L'Avicenne latin: un temoin (indirect) des commentateurs (Alexandre d'Aphrodise - Themistius - Jean Philopon)

Tradition et traduction. Les textes philosophiques et scientifiques au moyen age latin. Hommage a F. Bossier, ed. R. Beyers et al. Leuven: Leuven University Press, 1999

XVI

Some elements of Avicennian influence on Henry of Ghent's psychology

Henry of Ghent. Proceedings of the International Colloquium on the Occasion of the 700th. Anniversary of his Death (1293). Leuven: Leuven University Press, 1996

1-13 113-129

89-105

155-169

CONTENTS XVII

Elements of Avicennian metaphysics in the Summa Henry of Ghent and the Transformation of Scholastic Thought. Studies in Memory of Jos Decorte, ed. G. Guldentops and C. Steel. Leuven: Leuven University Press, 2003

vn

41-59

1-2

Index

This volume contains xii + 304 pages

PUBLISHER'S NOTE The articles in this volume, as in all others in the Variorum Collected Studies Series, have not been given a new, continuous pagination. In order to avoid confusion, and to facilitate their use where these same studies have been referred to elsewhere, the original pagination has been maintained wherever possible. Each article has been given a Roman number in order of appearance, as listed in the Contents. This number is repeated on each page and is quoted in the index entries.

PREFACE In the classical era of Islam, philosophy, falsafa in Arabic, adopted and adapted the philosophical project as it has been developed in Greece from the PreSocratics to the late Hellenic school of Alexandria. Inside the Islamic world of that time falsafa became an independent discipline. Although they owed a lot to their Greek predecessors, the faldsifa did more than take over the latter's thoughts: when faced with new problems, they introduced innovations in many ways and presented a multitude of new ideas. Ibn Slna - the Avicenna of the Latin West - is perhaps the most representative figure of this tendency. After him, not only in philosophy, but also in other fields of thought, no one in the Islamic world was ignorant of his works and/or ideas. In sum, he became the "auctoritas" in the Islamic East. But in the West, too, some of his major innovative doctrines became very influential. Ibn Slna was a Muslim. Notwithstanding his strong belief in human reason in the search for truth, he felt himself challenged by some affirmations of the Qur'anic Revelation. He did not put them simply aside. On the contrary, he eagerly tried to offer rational, logically coherent solutions to what at first sight appeared to be radically opposed views: human responsibility, and hence freedom, versus divine omnipotence expressed in the Divine Decree; an independent and determined causal order versus God's creative Command which is the immediate cause of each individual's existence. Of course, one may question the coherence of some of the proposed solutions. However, philosophy was never considered to be the handmaiden of theology by Ibn Slna. He was above all a philosopher who derived a large part of his basic inspiration from what may be called the "Neoplatonism" of the late school of Alexandria. But, despite all this, he was not insensitive to the problems raised by his religious beliefs. He made tremendous efforts in order to study the relation between God and the world - and both at the level of production and knowledge - in a way that is not only philosophically satisfactory, but also acceptable from a religious point of view. This formed one of the central issues in his thought, as becomes evident in the Ta 'liqat, a work that mainly seems to consist of student's notes. The latter work also contains elements of what probably constituted the original Arabic text, maybe a primary draft by Ibn Slna himself, of his encyclopaedia Ddnesh-Nameh, the very first philosophical writing in modern Persian so far as is known. If the direct influence of this work in fact all remained

x

PREFACE

very limited, indirectly it became very influential, i.e., through al-Ghazzall's slightly interpretative translation, traditionally known as Maqdsid al-faldsifa. It is commonly believed that al-Ghazzall wrote this work as an introduction to the Tahdfut, in order to offer his readers a better understanding of the philosophical theory he then categorically rejects in the latter work. But there are important terminological differences between the two works and the former is never quoted in the latter. Moreover, the very title of the former of the two poses problems, since al-Ghazzall only uses Avicennian writings. For these reasons - among others - it seems more appropriate to consider the Maqdsid as an independent work of the young al-Ghazzall, when he was studying philosophy. But also the common scholarly qualification of the Tahdfut as a total rejection of philosophy,1 especially Avicennian philosophy, is not as evident as is generally suggested. A closer inspection of the work reveals that al-Ghazzall, above all, reacts against taqlid, blind imitation, and hence that his criticism is in the first place directed against some contemporary philosophers who simply imitated - read: repeated Ibn STna's ideas. This does not mean that al-Ghazzall was not critical of Ibn SIna. Although he undoubtedly regretted the latter's insufficient recognition of the very limits of human reason, there exists strong evidence that he admired him for some of his major doctrines. Whether or not the Ma'drij is a genuine Ghazzalian writing, it still clearly shows that a great deal of Ibn STna's psychology fits well wthin a larger Ghazzalian framework.2 On the other hand, al-Ghazzall, in his authentic workMi'ydral-'ilm, largely bases his logical views on Avicennian texts (besides using a few Farabian texts). And also in several of his other writings, a rather systematic use of Avicennian texts comes to the fore, although he never explicitly mentions this. Hence, his attitude towards Ibn SIna is complex, but it is obvious that he, in the elaboration of his own system of thought, :s largely indebted to Ibn SIna. Contrary to al-Ghazzall, Bahmanyar ibn Marzuban, a first generation student of Ibn SIna, has been considered to be a rather slavish follower of his master. His major work, Kitdb al-Tahsil, copies almost verbatim many pages of Ibn SIna's works and seems to imitate the structure of the DdneshNdmeh. But a close analysis of that very structure already reveals that Bahmanyar while rejecting some of Ibn SIna's most important innovations, he somehow returns to a more genuine Aristotelianism. Hence, he was not totally uncritical of Ibn SIna, although in a way different from that of al-Ghazzall. Still another kind of criticism, once more combined with a profound knowledge of Ibn SIna's 1

The Tahdfut only points to the doctrinal divergence, and hence incoherence between the philosophers on doctrinal issues (e.g., temporality or eternity of the world) which they claim to have demonstrated in the proper sense, but such demonstration in fact exceeds the limits of human reason. 2 The work itself offers a unique, comprehensive view of man, paying particular attention to the soul and its faculties (in the Avicennian line), and also the spiritual vocation of man (in the Ghazzalian line).

PREFACE

xi

works, as shown by the many quotations from the otherwise unnoticed Ta 'liqdt, is obvious in the Asfdr, the major work of the great sixteenth-century Iranian thinker Mulla Sadra. For him, Ibn Slna was very close to the basic tenets of Ishraqiphilosophy, but never clearly expressed them. Ibn Slna's influence was not limited to the East. It also profoundly marked the West. Large parts of his major encyclopaedia, al-Shifd\ were translated into Latin. To translate accurately was - and still is - a difficult task. There exist a lot of double translations in the Avicenna Latinus\ they clearly oppose a very literal against a more "Latinized" translation. Moreover, no single case seems to require the consultation of a second Arabic manuscript that proposes an important variant. Therefore, it is probable that the Latin translator first elaborated a literal translation and thereafter revised it while offering a more "readable" text in Latin. Thanks to these Latin translations, scholastics became familiar with Ibn Slna's thought, and, through them, also - although most of the time in an unconscious way, since Ibn Slna rarely mentions their names - with the at that time unknown ideas of such Greek commentators such as Alexander of Aphrodisias, Themistius, and Philoponus. But of far greater importance was the direct impact of the innovative ideas of Ibn Slna. Historians of philosophy have already indicated the range of this influence, especially on such great scholastics as for example, Albert the Great and Thomas Aquinas. Till recent times, a somewhat neglected figure was Henry of Ghent. Living at the end of the thirteenth century, he may be considered to be one of the scholastics who has best understood Ibn Slna's system. Notwithstanding that the number of Avicennian quotations in his writings is rather limited, he deeply treats several of Ibn Slna's major innovations. The heritage of Ibn Slna is so rich and complicated that many of its aspects still remain unexplored while other await further critical examination. This volume, I hope, may constitute a humble contribution to a better understanding of that heritage. I thank all the editors who granted permission to reproduce the articles contained in this collection. I wish, moreover, to thank Ch. Burnett (Warburg Institute, London) for having suggested that I put together a collection of some of my papers, as well as Dr John Smedley, the publisher of Variorum, for having accepted their publication. JULES JANSSENS Katholieke Universiteit, Leuven, Belgium October 2005

ACKNOWLEDGEMENTS I am grateful to the following persons, publishers, societies and journals for their permission to reprint the papers included in this volume: John F. Perry, editor of the journal Ultimate Reality and Meaning (II); the Patronat d'Estudis Osonencs (III); the editors of the journal Documenti e Studi sulla Tradizione Filosofica Medievale, Sismel-Edizioni del Galluzzo, Florence (IV); A. Speer, editor of the volume Was ist Philosophic im Mittelalter? (V); Librairie J. Vrin, Paris (VI, VIII, IX); M. Hoenen, Secretary General of the society S.I.E.P.M. (VII); the editors of the Journal of Islamic Studies (X, XIII); M. Maroth, editor of the volume Problems in Arabic Philosophy (XI); Brill Publishers, Leiden (XII); Koninklijke Brill N.V, Leuven (XII); Leuven University Press, Leuven (XIV-XVII).

I Ibn Sina, and His Heritage in the Islamic World and in the Latin West Ibn Slna, known in the West as Avicenna, is one of the major figures of the history of human thought. Born around 980 near Bukhara, he died in 1037 in Hamadhan. He lived a very stirring life, travelling from one place to another "out of necessities", such being an euphemism for his search for glory, and almost certainly also for money.1 He was a very gifted person, who busied himself from time to time with politics, who practised now and then as a physician, but who was above all an outstanding philosopher who put forward a great many innovative ideas.2 Especially in this later sense, he exercised a tremendous influence both in the East and the West. Hence, it is no surprise that his heritage is very varied and large. It is impossible to give a full account of it in one paper, but let us concentrate on a few main lines in the fields of physics, psychology and metaphysics. For Ibn Slna, metaphysics is basically the study of "being qua being", but he immediately adds that its most noble, although not its first object is God, the Necessary Being by virtue of itself.3 All this implies that the existence of God is not self-evident, but has to be proven. In order to do so, Ibn Slna develops his theory of the essence/existence distinction. In God, His essence is His existence, while in all other Beings one has to sharply distinguish their existence from their essence.4 In the latter case, he qualifies the existence as being not identical with, 1

See Y. Michot, Ibn Slna. Lettre au vizir Abu Sa'd (Sagesses musulmanes, 4), Beyrouth, Les editions al-Bouraq, 1421/2000, Introduction, especially pp. 31*-51*. 2 It is also true that his medical works subsequently received great attention, both in the East and the West, and such not only in medical but also in philosophical circles, but this merits a study of its own. 3 This is a common doctrine in his writings, clearly expressed in his major writing al-Shifa', al-Ilahiyyat, vol. I, eds. G.C. Anawati and S. Zayed, Cairo, OGIG, 1960, book I, chapters 1-2. This text is also available in a mediaeval Latin translation, see S. Van Riet (ed.), Avicenna Latinus. Liber de philosophia prima sive scientia divina, I—IV, Louvain, Peeters; Leiden, Brill, 1977, Tractatusprimus, capitula 1-2. 4 This idea is omnipresent in Ibn SIna's metaphysical works too. Again, I refer only to al-Shifa\ al-Ilahiyyat, but now vol. II, eds. M.Y. Musa, S. Dunya and S. Zayed, Cairo, OGIG, 1960, book VIII, chapter 4; for the Latin translation, see S. Van Riet (ed.),

I 2

Ibn Sina, and His Heritage in the Islamic World and in the Latin West

but "accidental" to essence. Ibn Rushd saw in it a simple affirmation of the pure accidentality of existence, and therefore vehemently criticized Ibn Sina on this point.5 Till recent times, this latter interpretation remained the standard one in the West. F. Rahman seems to have been the first contemporary commentator to have seriously put into question such an understanding of Ibn Sina's affirmation.6 However, it deserves to be stated that already Henry of Ghent, in the late thirteenth century, was aware of the fact that the restricted Aristotelian notion of "accidentality" was surely not involved here, but a larger one.7 Among the vast majority of the Latin scholastics who followed Ibn Rushd's line of interpretation, was Thomas Aquinas. However, he accepted as most valid Ibn Sina's distinction between essence and existence,8 a fact already evident from the very title of his famous early work De ente et essentia. Thomas Aquinas uses Ibn Sina's theory in order to explain the composite nature of all creatures, especially the immaterial ones, i.e., the angels. He obviously rejects any kind of hylemorphic composition in them. Although he does not mention the Avicennian vocabulary of "necessary in virtue of itself, "necessary in virtue of another", and "possible", he clearly derived his view from Ibn Sina, and not from Aristotle, Boethius, or the Liber de Causis as he suggests.9 But, on the other hand, Thomas wants to distance himself from what he considers to be Ibn Sina's "essentialism". For him, there is not only identity in God between Avicenna Latinus. Liber de philosophia prima sive scientia divina, V-X, Louvain, Peeters; Leiden, Brill, 1980, Tractates octavus, capitulum 4. 5 See e.g., Ibn Rushd, Tafsirma bad al-Tabira, ed. M. Bouyges, Beirut, 1938. Repr. Tehran, Int. Hikma, 1377 H.S., vol. I, commentary on book Gamma, C 3, p. 313; there exists a medieval Latin translation, see Averroes, Opera omnia, Venetiis, Junta, 1562, vol. VIII, f. 67 B. 6 F. Rahman, "Essence and Existence in Avicenna", Medieval and Renaissance Studies 4 (1958), pp. 1-14; see also ID., "Essence and Existence in Ibn Slna. The Myth and the Reality", Hamdard Islamicus4i (1981), pp. 3-14. 7 See P. Porro, "Possibilita ed esse essentiae in Enrico di Gand", in W. Vanhamel (ed.), Henry of Ghent. Proceedings of the International Colloquium on the Occasion of the 70Cfh. Anniversary of His Death (1293) (Ancient and Medieval Philosophy, Ser. I, vol. XV), Leuven, Leuven University Press, 1996, pp. 211-53, especially p. 215. 8 There exists ample literature on this subject. For references, see my An Annotated Bibliography on Ibn Slna (1970-1989) (Ancient and Medieval Philosophy, Ser. I, vol. XIII), Leuven, Leuven University Press, 1991, pp. 244-250 and my An Annotated Bibliography on Ibn Slna. First Supplement (1990-1994) (FIDEM, Textes et etudes du moyen age, 12). Louvain-la-Neuve, FIDEM, 1999, pp. 137-161. 9 See Thomas d'Aquin. Dietrich de Freiberg, L'etre et /essence. Le vocabulaire medieval de 1'ontologie. Traduction et commentaires par A. de Libera et C. Michon (Points. Essais, 339), Paris, Editions du Seuil, 1996, pp. 27-36.

I Ibn Sina, and His Heritage in the Islamic World and in the Latin West

3

essence and existence, but God is above all "esse subsistens", which clearly means that the priority is on the side of the existence. Thomas certainly believed that he thus was radically opposing Ibn Sma's view. Even if he misunderstood the accidentality of existence in the latter's thought, he was right in his opinion that Ibn Sina had not fully appreciated existence as a part of the integral ontological dimension of Beings. It is striking that at the beginning of the seventeenth century, the great IshraqI Iranian thinker Mulla Sadra Shirazi also opposed the essentialist view present in his main predecessors, e.g., al-Suhrawardi and Mir Damad. Against them, he defends the priority of "being" or "existence" (wujud) over essence.10 Does this constitute at once an indirect criticism against Ibn Sina? Not at all! According to Mulla Sadra, Ibn Sina never affirmed the accidentality of existence, or, if some of his sayings offer such an affirmation, they do, at most, apparently.11 Mulla Sadra rightly insists that for Ibn Sina an essence in itself is indifferent toward existence or non-existence; in fact, Ibn Sina had clearly rejected any kind of pre-existence of the essences and had, accordingly, vehemently criticized the Platonic theory of the Ideas.12 Mulla Sadra agrees moreover with his thirteenth-century predecessor Nasir al-Dln al-Tusi that for Ibn Sina existence only qualifies quiddity by way of understanding, not in reality; and he, contrary to Thomas Aquinas in the West, adopts the entire Avicennian vocabulary of necessary/possible. It seems therefore that one may conclude that, according to him, Ibn Sina was certainly on the right path, but that his wording sometimes lacked clarity. This does not mean that Mulla Sadra completely agreed with Ibn Sina's system. Such is in no way the case, since the latter after all remained too peripatetic, notwithstanding the presence of elements in his thought that strongly point in the direction of the doctrine of Ishraq. Generally speaking, one has the strong impression that Ibn Sina was a brilliant mind in the eyes of Mulla Sadra, and that it therefore was hard for him to understand that the former had never developed such kind of "illuminationist" ideas.

10

See Mulla Sadra Shirazi, al-Hikmat al-muta 'aliyya f i 1-asfar al- 'aqliyya al-arba fa, ed. R. Lutfi, Qom, Manshurat Mustafawl, 1397 H.S., I, 1, 1, 4, vol. I, pp. 38-44. The present paragraph is largely inspired by the much more detailed analysis of Mulla Sadra's doctrine of Being in the unpublished Ph.D. thesis of C. Bonmariage, Le reel et les realites. La structure de la realite de letre chez Mulla Sadra Shirazi, Universite Catholique de Louvain, Institut superieur de Philosophic, 1998, vol. I, pp. 32-66. 11 Ibid., I, 1, 2, 22, vol. I, pp. 250-51. 12 See al-Shifa\ al-Ilahiyyat, book VII, ch. 2-3.

I 4

Ibn SIna, and His Heritage in the Islamic World and in the Latin West

As indicated above, God does not form the first object of the science of metaphysics. However, to prove His existence does constitute one of its major objectives. In this sense, Ibn SIna works out an entirely new proof that is essentially different from the Aristotelian proof of the Unmoved Mover.13 He rejects the latter, since, by its very nature, it belongs to physics, and the problem of the divine existence falls completely outside the scope of this latter science. Instead, he offers his famous proof from contingency, based on the distinction between necessary and possible being, stressing that the impossibility of an infinite regress of possible beings necessitates the acceptance of the existence of a necessary being in virtue of itself, which by logical necessity can only be one. Hence, God is the first cause of the existence, not of merely the motion of the world. Of importance, too, is Ibn Sina's presentation of this argument as a proof (dalll) and not as a demonstration proper (burhari). Al-Ghazzall quotes the proof with approval in his Fada'ih al-Batiniyya, also known as al-Mustazhirf, he presents it as an intellectual example of how one may arrive at certain scientific knowledge.14 Surprisingly, he, in his Tahafut alfalasifa, gives the impression that he cannot accept the proof as a genuine demonstration of God's existence.15 If such is the case, there does exist a flagrant contradiction between these two works, due either to a lack of attention or to a conscious change of mind. The former hypothesis is hardly compatible with a brilliant mind as al-Ghazzall's, while the second would imply a radical change in thought in a very brief period, since both works are generally accepted as belonging to approximately the same period. But is there really a contradiction? What al-Ghazzall basically states in the Tahafut is that the proof is non-conclusive insofar as one accepts an actual infinity of Beings, and that one is forced to accept such an actual infinity if one admits the eternity of the world. But, unless I am completely mistaken, the proof is valid if simply one 13

For an essential outline of Ibn Sina's proof, and its reception or rejection in later times in Islamic, as well as in Jewish and Christian authors, see H.A. Davidson, Proofs for Eternity, Creation and the Existence of God in Medieval Islamic and Jewish Philosophy, New York, Oxford, Oxford University Press, 1987, pp. 281-377. In what follows, I will not unnecessarily multiply the references to this work. I will only explicitly indicate those cases where I offer a slightly different interpretation from Davidson's. 14 Al-Ghazzall, Fada'ih al-Batiniyya, ed. A. Badawi, Cairo, 1960, pp. 82-3; English translation in RJ. McCarthy, Freedom and Fulfillment, Boston, Twayne Publications, 1980, pp. 224-5, §§ 166-9. 15 Al-Ghazzall, The Incoherence of the Philosophers. A Parallel English-Arabic Text, Translated, Introduced and Annotated by M.E. Marmura, Provoh, Utah, Brigham Young University Press, 1997, Fourth Discussion, pp. 79—84.

I Ibn Sina, and His Heritage in the Islamic World and in the Latin West

5

accepts a temporal origination of the world. As al-Ghazzali speaks in his own name in the Fada'ih al-Batiniyya, one might surmise that a temporal limitation of the universe forms a basic, although not explicitly stated idea, and thus the proof has logical force.16 While Ibn Rushd many times categorically rejects this proof, but clearly based on a total misunderstanding of its articulation, Fakhr al-Dln al-Razi did not hesitate to use it, although alongside Kalam arguments.17 Al-ShahrastanI, in his turn, restricts himself to copying in a slightly modified way the section of the argument in the Najat, although omitting any explicit reference to an infinite regress of causes. However, he emphasizes the cosmological nature of the proof by stating in a most explicit way that it is based on an inference from the possible (read: contingent) to the necessary.18 Finally, Mulla Sadra mentions it as one of the possible proofs for establishing God's existence. Nevertheless, he insists that it has only value on the level of the concept (mafham) of Being, not on that of its true reality (Aaqfqa).19 But Ibn Sina's proof received attention outside Islamic circles as well. Two twelfth-century Jewish thinkers, Ibn Daud and Joseph ibn Aqnln, developed both an argument from the impossibility of an infinite regress, supplementing it at key points with elements derived from or with a paraphrase of Ibn Sina's proof.20 In the Latin tradition, one already discovers an impact on Gundissalinus, one of the important Arabic-Latin translators of twelfth century Toledo: he develops in his De procesione mundi a proof for God's existence based on the impossibility of an infinite regress in combination with an analysis of the notion of "necessarily existent", which means "necessary being in virtue of itself'.21 In the first half of the thirteenth century, William of Auvergne,

16

I therefore doubt that Davidson is correct, when he states (op. cit., p. 375, n. 215): "Ghazali appeared in many guises, and in another work [than the Tahafut, i.e. Fada'ih alBatiniyya], where he presents his own position, he does accept the proof. 17 For more details, see Davidson, op. cit., pp. 311-35. 18 Al-ShahrastanI, Kitab al-milal wa 1-nihal, ed. W. Cureton, London, 1842^6, p. 376. For the corresponding passage in Ibn Sina, see Ibn Sina, al-Najat, ed. Daneshpazhuh, Tehran, Mu'assasa intisharat va-jap daneshgah, 1364 H.S., pp. 567-8. Davidson, op. cit., p. 308 qualifies Shahrastanl's version as a more straightforward formulation of the proof from contingency, but he seems to overlook its direct verbal dependence on the Najat, and, above all, Shahrastanl's insistence on its cosmological character. 19 Mulla Sadra, As fir, III, 1,1,2, vol. VI, pp. 26-9, especially pp. 26-7. 20 For more details, see Davidson, op. cit., pp. 357-60. 21 Ibid., p. 360. Davidson further (p. 361) refers to William of Auxerre, and reminds us of elements in Albert the Great, Thomas Aquinas (third way) and Duns Scotus.

I 6

Ibn Sina, and His Heritage in the Islamic World and in the Latin West

proposes an argument that relies upon Ibn Slna's dialectic of possible and necessary existence, though also containing a mixture of elements from Boethius and others.22 In the second half of the same century, Henry of Ghent, in his Summa, q. 22 presented a new "metaphysical" proof for the existence of God. Henry does this while quoting extensively from the fifth chapter of book eight of the Metaphysics of Ibn Sina. However, he divides the texts in several parts and now and then introduces modifications. Moreover, he clearly rejects Ibn Slna's identification between "possible per se" and "necessarium ab alio" and holds a much more radical view of contingency. Nevertheless, it cannot be denied that Ibn Slna's argument from contingency had a real impact on Henry's formulation, but we have to wait to carry out a very detailed study in order to trace its precise nature. It may also be observed that Henry agrees with Ibn Sina that the proof of the Unmoved Mover does not have a place in the science of metaphysics.23 I have already stressed that for Ibn Sina, metaphysics was essentially the study of being qua being, i.e., an ontology, but that its most noble object of investigation is God, in other words it also includes theodicy. One may add that for Ibn Sina metaphysics is moreover the science that has to demonstrate the principles of the other sciences, which means that it also deals with the "archaeology" of the sciences. Although Aristotle's metaphysics entails elements of all three of these views, it never linked them together within a substantial unity. Based on the important preparatory work of his predecessor al-Farabi, for the first time in the history of philosophy Ibn Sina had worked out a metaphysical "system" as "system". In this respect this latter rightly deserves to be qualified as a kind of onto-theology even if it does not completely match Heidegger's - inspired by Duns Scotus - conception. In view of all this, metaphysics is certainly the highest and most valuable of all sciences, a fact particularly underlined by Ibn Sina when he discusses, in his Danesh-Nameh, metaphysics immediately after the "instrumental" science of logic, but before "physics", and such contrary to the customary habit, a habit he himself respects in his Arabic encyclopaedias, of presenting it as the latest of all sciences. While it is commonly designated as "meta-physics", i.e., the science of what comes 22

See R. Teske, "William of Auvergne's Debt to Avicenna", in J. Janssens and D. De Smet (eds), Avicenna and His Heritage (Ancient and Medieval Philosophy, Ser. I, vol. XXVIII), Leuven, Leuven University Press, 2002, pp. 153-70, especially pp. 159-60. 23 See J. Janssens, "Elements of Avicennian Metaphysics in the Summd\ in: G. Guldentops and C. Steel (eds), Henry of Ghent and the Trans formulation of Scholastic Thought (Ancient and Medieval Philosophy, Ser. I, vol. XXXI), Leuven, Leuven University Press, 2003, pp. 41-59, especially pp. 50-52 (included in this volume as study XVII).

I Ibn Sina, and His Heritage in the Islamic World and in the Latin West

1

after physics, an appellation sometimes used by Ibn Sina as well, he does not hesitate to call it Ilahiyyat, i.e., "(science) of the Divine Things" as well, and indeed prefers to do so. It is striking that Ibn SIna's "new" metaphysical system was clearly not accepted by one of his immediate and most important pupils, i.e., Bahmanyar ibn Marzuban. In his "metaphysics" proper, i.e., the second part of his Kitab alTahsll, he avoids any reference to specific theological ideas and he moreover blurs the borderlines between the sciences of logic, physics and metaphysics. He thus undermines a few of the pillars of Ibn SIna's metaphysical system and seems to prefer a return to a more "historical" Aristotelianism.24 He might therefore have prepared the way for Ibn Rushd's more radical criticism of Ibn Sina, and his plea for adhering to a "pure" Aristotelianism. Contrary to Bahmanyar, al-Lawkari, a third generation student of Ibn Sina, remains faithful to the master's project. Despite his knowledge of Bahmanyar's Kitab al-Tahsil, he insists that the first object of the science of metaphysics is "Being qua Being", and that the establishment of God's existence and of the general principles of the totality of the Beings do constitute its more noble, but not its primary object.25 It is clear that Ibn SIna's close and less close immediate disciples have different evaluations of his new metaphysical project. It would be certainly worthwhile to examine in detail how both tendencies have influenced later thinkers, but such exceeds the limits of the present paper.26 With respect to the Latin world, it is obvious that Ibn Sina contributed, at least in some respects, to the elaboration of different kinds of onto-theology. But, most of the time, his system is supplemented with elements derived from other authorities, as e.g., Ibn Gabirol, Ibn Rushd, or Augustine. But, whatever the case, the vast majority of scholastics seems to have accepted with Ibn Sina that "Being qua being" constitutes the first object of metaphysics, but that it is also a basic task of this science to offer a proof, or proofs for God's existence. However, in some of them, e.g., Peter of Auvergne, or Henry of Ghent, a strong 24

See J. Janssens, "Bahmanyar ibn Marzuban: A Faithful Disciple of Ibn Sina?", in: D. Reisman, with the Assistance of Ahmed H. al-Rahim (eds), Before and After Avicenna. Proceedings of the First Avicenna Study Group (IPTS, LII), Leiden, Boston, Brill, 2003, pp. 177-97 (included in this volume as study XII). 25 Fadl ibn Muhammad al-Lawkari, Bayan al-Haqq bi-diman al-sidq, al- film al-ilahi, ed. Ibrahim Dibajl, Tehran, ISTAC, 1995,1, 1, pp. 8-11. 26 For a general survey of the influence of Ibn SIna's philosophy on the AraboIslamic world till approximately 1350, see D. Gutas, 'The Heritage of Avicenna: The Golden Age of Arabic Philosophy, 1000- ca. 1350", in J. Janssens and D. De Smet (eds), Avicenna and His Heritage (Ancient and Medieval Philosophy, Ser. I, vol. XXVIII), Leuven, Leuven University Press, 2002, pp. 81-97.

I 8

Ibn Sina, and His Heritage in the Islamic World and in the Latin West

Avicennian dimension comes too the fore in the way they determine the proper object of metaphysics, and make theology part of it.27 Regarding Thomas Aquinas, the Avicennian dimension is perhaps less outspoken, but he agrees with Ibn Sina that the science of metaphysics primarily deals with "Being qua Being", and that the presence of a theology certainly belongs to it even if God is not included as such under its proper subject.28 But Ibn Sina was not only a great metaphysician. He made also important contributions in the field of psychology that did not remain unnoticed in later centuries, both in the East and the West. In what follows, I will limit myself to two of them, i.e., his concept of the soul and his introduction of the faculty of estimation (wahm). Ibn Sina defends a dualistic and spiritualistic image of man:29 the soul exists independently of the body and has its own substantiality. Each soul originates together with the origination of a body and is individualized through the link with the body for which it came into existence. However, each soul can survive after the death of its body; it possesses individual immortality. Hence, despite its temporal beginning, the soul does not perish. Such view clearly implies the rejection of the Aristotelian adagium that whatever has a beginning, must also have an end. This was indeed a very innovative attitude. Also special attention should be given to the famous "flying man" argument that is present in several of Ibn Slna's writings, and is very particular to him. It shows how man, deprived 27

M. Pickave, "Heinrich von Gent iiber das Subjekt der Metaphysik als Ersterkanntes", Document! e Studi sulla Tradizione Filosofica Medievale 12 (2001), pp. 493-52, rightly stresses that there are also differences between Henry and Ibn Sina, especially insofar as the former describes Gcc1 as "primum cognitum". However, he ignores, or, at least, does not mention the idea of an encompassing metaphysical ''system" as common to both Henry and Ibn Sina, see my "Elements of Avicennian Metaphysics in the Summd\ p. 58. 28 See Thomas Aquinas, In Metaphysicam, passim, but especially prooemium. See also J. Wippel, "The Latin Avicenna as a Source for Thomas Aquinas' Metaphysics", Freiburger Zeitschrift fur Philosophic und Theologie 37 (1990), pp. 51-90, especially pp. 53-9. It may be worthwhile to note that Thomas' master, Albert the Great, rather disregards Ibn Sina when he develops his own metaphysical system, which A. de Libera, Metaphysique et noetique. Albert le Grand (Problemes et controverses), Paris, Vrin, 2005, pp. 8-9 has called a "theologie philosophique" and has qualified as a combination of a ontology and a philosophical noetics. 29 For a basic presentation of Ibn Slna's theory of the soul, see my "Ibn Slna's Ideas of Ultimate reality and meaning. Neoplatonism and the Qur'an as Problem-Solving Paradigms in the Avicennian System", Ultimate Reality and Meaning 10 (1987), pp. 252-71, especially pp. 255-62 (included in this volume as study II). For a more detailed expose, the reader may consult M. Sebti, Avicenne. L'ame humaine (Philosophies, 129), Paris, PUF, 2000.

I Ibn Sina, and His Heritage in the Islamic World and in the Latin West

9

of any sensation, still is able to affirm the existence of his core essence, i.e., his soul. It thus opens the way for arguing the substantiality of the soul, but it directly only proves, or, more precisely, "hints at" its independence from the body.30 Generally speaking, Bahmanyar, in his Kitab al-Tahsil, remains faithful to Ibn Slna's conception of the soul. Let me stress that he, in full accordance with Ibn Sina,31 characterizes the "flying man" argument as a "reminder" (tanbiti), but presents it in a slightly modified form. In fact, he insists very strongly on the awareness (shu'ur) of one's essence (dMt).32 Moreover, he omits any reference to a sudden creation in the air, but such was already the case in some of Ibn Slna's own formulations.33 Of much greater complexity than Bahmanyar's is alGhazzall's use and appreciation of Ibn Slna's teachings on the soul.34 In view of the Tahafut, especially question 18, he seems to categorically reject the idea that the soul is a spiritual substance, totally independent of the body. But from the title of this question, it is clear that he claims that the philosophers are unable to give a rational demonstration in this respect, no more no less. Nevertheless, he clearly condemns the philosophers, included Ibn Sina, for their denial of bodily

30

D.N. Hasse, A vicenna 's de Anima in the Latin West (Warburg Studies and Texts, 1), London, The Warburg Institute; Turin, Nino Aragno Editore, 2000, pp. 80-7, offers a most valuable analysis of the argument based on all of Ibn Slna's formulations. 31 (Ibn Sina), A vicenna's De Anima (Arabic Text), being the Psychological Part of Kitab al-Shifa', ed. F. Rahman, Reprint London, New York, Toronto, Oxford University Press, 1970,1, 1, p. 15,1. 20. In the not yet edited psychological part of Ibn Slna's work Al-Mashriqiyyun, the same designation is presented, although two manuscripts offer the variant ishara, "pointer", see D.N. Hasse, A vicenna's de Anima in the Latin West, p. 86 and p. 86, n. 39. 32 Bahmanyar ibn Marzuban, Kitab al-Tahsil, ed. M. Mutaharri (Intisharat Daneshgah Tehran, n°. 2310), 2Tehran, Intisharat Daneshgah Tehran, 1375 H.S., pp. 734-5. 33 See the versions of the Isharat( Part II, Namat 3, chapter 1) and of the al-Risala al-Adhawiyya fil-ma -^(chapter IV). 34 There exist two monographs in Arabic on Ibn Slna's influence on al-Ghazzall's doctrine of the soul: M.H. Abu Sa'dah, Al-athar al-sinawiya fi madhhab al-Ghazali fi lnafs al-insaniyya, Cairo, Dar Abu Hurayba, 1991 and J.R. SayyidabI, Nazariyyat al-nafs bayna Ibn Sina wa l-Ghazzali, Cairo, Al-hay'a al-misriyya al-'amma lil-kitab, 2000. A significant study is also that of B. Abrahamov, "Ibn Slna's Influence on al-Ghazall's Non-Philosophical Works", Abr Nahrain 29 (1991), pp. 1-17. Finally, some interesting ideas are present in R. Frank, Al-Ghazali and the Ash 'ante School (Duke Monographs in Medieval and Renaissance Studies, 15), Durham and London, Duke University Press, 1994, passim.

I 10

Ibn Slna, and His Heritage in the Islamic World and in the Latin West

resurrection. In limiting themselves to accept the sole survival of the human soul, they make themselves guilty of unbelief (kufr), but this later qualification might be a later addition that did not belong to the original version and was only introduced by the late al-Ghazzall.35 In what probably is one of his earlier works, i.e., Ma farij al-quds, he largely uses Ibn Sina's proofs for the existence and the substantiality of the soul; he even presents a slightly modified version of the "flying man" argument by combining elements derived both from the Kitab al-nafsof the Shi fa'and the Isharat.36 Unambiguously critical is Fakhr al-Dln alRazi's evaluation of the "flying man" argument.37 Certainly, he admits with Ibn Slna that each human Being has a constant self-awareness, but Ibn Slna has in his view failed to prove this. However, his main criticism concerns the very nature of the self involved in self-awareness. Even if no link whatsoever with the body is required in order to have the latter, this does not imply that the concerned self is purely "immaterial", as Ibn Slna claims. On the contrary, it is material, although in a subtle manner. Therefore, the self is not identical with the physical body, neither "bodily", and only in this respect Ibn Sina's argument is really conclusive. In the Latin West, nobody adhered to Ibn Sina's psychology in its totality. But this does not mean that the Latin translation of the de Anima received no attention. Not at all! It is quoted many times, and some particular ideas were taken over either by only a few or by a vast majority of the scholastics. For example, John Blund adopted the idea that the soul generates together with the body, although he ascribes the soul's infusion into the body to God instead of to 35

Since the conclusion entails a more radical condemnation of some philosophical ideas than the introductions, or even the expose proper did (see my "Al-Ghazzali's Tahafut Is it Really a Rejection of Ibn Sina's Philosophy?", Journal of Islamic Studies 12 (2001), pp. 1-17, included in this volume as study X), one may wonder whether it has not been added later, since it seems to reflect al-Ghazzall's mature attitude as present in his late work Faysal al-tafriqa bayna 1-Islam wa 1-zandaqa, where he states that Revelation may not be questioned unless a real demonstration does exist, otherwise one is guilty of unbelief. See Fr. Griffel, Apostasie und Toleranz im Islam. Die Entwicklung zu al-Ghazall's Urteilgegen die Philosophie und die Reaktion der Philosophen (I.P.T.S., XL), Leiden, Boston, Koln, 2000, passim, but especially pp. 466-7. 36 See my "Le ma farij al-quds f i madarij ma 'rifat al-nafs*. un element-cle pour le dossier Ghazzali-Ibn Slna?", Archives d'histoire doctrinale etlitteraire du moyen age 60 (1993), pp. 27-55, especially p. 31 (included in this volume as study VIII). 37 In what follows, I summarize the substantial contribution of M.E. Marmura, "Fakhr al-Din al-Razi's Critique of an Avicennan TanbifT, in B. Mojsisch and O. Pluta (eds), Historia philosophiae medii aevi: Studien zur Geschichte der Philosophie des Mittelalters, Amsterdam, Philadelphia, B.R. Gruner, 1991, t. II, pp. 627-41.

I IbnSina, and His Heritage in the Islamic World and in the Latin West 11 the Agent Intellect.38 Also based on Ibn Sina's De Anima, Albert the Great, Thomas Aquinas, Siger of Brabant and Matthew of Aquasparta stated that the soul has a beginning in time together with the body.39 It is worthwhile to note that Thomas Aquinas, in his De ente et essentia, V, 4,40 stresses that for Ibn Sina only with respect to the principle, not to the end the individuation of the souls is dependent upon the bodies of which they constitute the act, in other words that the temporal origination of the soul does not exclude its immortality. Attention was moreover paid to the "flying man" argument, which however was understood as a proof or demonstration.41 Some authors, e.g., Gundissalinus or William of Auvergne, used it in order to show the incorporeality of the soul, while others, e.g., Jean de la Rochelle, Matthew of Aquasparta and Vital du Four, proved by it the self-awareness of the soul, and still others, e.g., Jean de la Rochelle and Anonymous (Vat. Lat. 175) saw in it the demonstration of the existence of the soul. However, this argument had no major impact on the later scholastics. Let me add that despite a superficial similarity, it is quite different from the Cartesian Cogito, insofar that this latter implies a much more natural connection between soul and body than does the former.42 A really significant and major influence originated in Ibn Sina's discussion of the inner senses, especially his innovative idea of estimation.43 Almost every author of the thirteenth century mentions this faculty, as well as two related topics: the notion of "intentio" and the famous example of the fear of the sheep for the wolf, which states that the former flees the latter as soon as it sees it. John Blund already offered a very profound analysis, while at once completing and 38

John Blund, Tractatus de anima, ed. D.A. Callus and R.W. Hunt (Auciorss Britannici Medii Aevi, 2), London, 1970, pp. 98-9. See also J. Jolivet, "The Arabic Inheritance", in P. Dronke (ed.), A History of Twelfth-Century Western Philosophy, Cambridge, Cambridge University Press, pp. 113-48, especially pp. 146-7. 39 See D.N. Hasse, A vicenna 's de Anima in the Latin West, p. 297 sub V, 3, f (Index locorum). 40 See Thomas Aquinas, De ente et essentia, in Thomas d'Aquin. Dietrich de Freiberg, L'etre et 1 'essence. Le vocabulaire medieval de 1'ontologie, V, 4, pp. 110 (Latin)-111 (Frenchtranslation). 41 For a more detailed presentation, see D.N. Hasse, Avicenna's de Anima in the Latin West, pp. 87-92. 42 See Th.-A. Druart, "The Soul and Body Problem: Avicenna and Descartes", in Th.-A. Druart (ed.), Arabic Philosophy and the West. Continuity and Interaction, Washington, Georgetown University, 1988, pp. 27^9; compare also R. Arnaldez, "Un precedent avicennien du cogito cartesien?", Annales Islamologiques 17 (1972), pp. 341-9. 43 See once more D.N. Hasse, A vicenna's de Anima in the Latin West, pp. 141-53.

I 12

Ibn Sina, and His Heritage in the Islamic World and in the Latin West

radicalizing Ibn Sina's theory. He insists that the faculty of estimation directly perceives the "intentiones", and that the latter are nowhere else than in the objects. Michael Scot, on his turn, develops the Avicennian idea in the direction of a "ratio sensibibilis", which enables us to foresee things. Finally, Albert the Great agrees with Ibn Sina's distinction between five internal senses, but he encloses them in a rigid Aristotelian framework of thought. Many more scholastics refer to, and also use this Avicennian complex of ideas, but they present no particular new insights compared to the ones just mentioned. Finally, Ibn Sina also proposes innovative ideas in the domain of physics. Perhaps influenced by the definition of motion as "the first perfection of what is in potency insofar as it is in potency", as offered by Themistius, Philoponus, and, maybe also by Alexander of Aphrodisias,44 Ibn Sina sharply distinguishes between two notions of motion.45 The first is of a mental nature and grasps the totality of the distance covered by a motion from its starting point to its end point; the second concerns the intermediary state between the two extreme points and it, having existence in reality, constitutes motion proper. Unless I am mistaken, the acceptance of motion in the category of "situs" is entirely new with Ibn Sina.46 The same seems to be true for his emphasis on the objective existence of time that is defined as a continuum produced by the passing of the "flowing now".47 But such does not exclude that it is on the mental level 44

See my "L'Avicenne latin: un temoin (indirect) des commentateurs (Alexandre d'Aphrodise -Themistius - Jean Philopon), in R. Beyers, J. Brams, D. Sacre and K. Verrycken (eds), Tradition et traduction. Les textes philosophiques et scientifiqucc au moyen age latin (Ancient and Medieval Philosophy, Series I, vol. XXV), Leuven, Leuven University Press, 1999, pp. 89-105, especially pp. 97-9 (included in this volume as study XV); and A. Hasnawi, "Alexandre d'Aphrodise vs. Jean Philopon: Notes sur quelques traites d'Alexandre "perdus" en grec, conserves en arabe", Arabic sciences and Philosophy 4 (1994), pp. 53-109, especially pp. 63-6. 45 See A. Hasnawi, "La definition du mouvement dans la Physique du Shi fa' d'Avicenne", Arabic Sciences and Philosophy 11 (2001), pp. 219-55, especially pp. 228-37. 46 See Ibn Sina, Al-Shifa\ al-Sama al-tabrt, ed. S. Zayed, Cairo, OGIG, 1983, Maq. II, chapter 3, pp. 103-6. 47 For Ibn Sina's conception of time, see ibid., chapters 10-13. There exists an English translation with commentary by A. Shayegan, A vicenna on Time, (unpublished) Ph.D. Thesis, Harvard, 1986; the translation of chapters 11-12 has been published in S.H. Nasr and M. Aminrazavi, An Anthology of Philosophy in Persia, vol. I, New York, Oxford, Oxford University Pres, 1999, pp. 241-50. There exists another English translation, with commentary, of chapter 12, see J. McGinnis, "Ibn Sina on the Now: Text and Commentary", American Catholic Philosophical Quaterly 73 (1999),

I Ibn Slna, and His Heritage in the Islamic World and in the Latin West

13

potentially divisible in infinity: in relation with an imagined finite motion one can imagine a limited time, the "now" presenting itself as a limit. Bahmanyar, Ibn Sina's close disciple as already mentioned, does not seriously question these innovations.48 But he does neither emphasize the new insights, e.g., he mentions the two notions of motion, but the way he presents them may let the reader believe that this constitutes common doctrine. AlLawkarl, in his turn, in his major work Bayan al-Haqq bi-diman al-sidq, copies almost verbatim pages and pages,49 although he omits passages, or even entire chapters insofar as they — effectively - do not present Ibn Sina's own theory. A more critical attitude comes to the fore in Fakhr al-Din al-Razi. Although he belonging to the school of kalam, not of falsafa, he, especially in his alMabahith al-Mashriqiyya, pays great attention to Ibn Sina's physical ideas, but this does not mean that he always agrees with them.50 For instance, despite the fact that he accepts Ibn Sina's double notion of motion, he does not fully agree with the latter's dynamic conception of motion; in fact, he rather inclines to the Eleatic view. With respect to time, he is "perplexed" as he himself admits. He seems to be aware of the advantage of Ibn Sina's idea that the "flowing now" produces real time, but he also wonders how something that appears to be a "nothing" can produce something real?51 Generally speaking, one gets the impression that he agrees with the Avicennian view of time as a continuum, but tries to avoid characterizing it as a necessary actual infinity since such qualification only belongs to God. Although Mulla Sadra still is inspired by Avicennian ideas - most of the time mediated by F.D. al-Razi - he does not

pp. 73-106. For the double notion of the "now" and its linkage with the double notion of motion, see A. Hasnawi, "La definition du mouvement", pp. 237-8. 48 Regarding the remarks on motion, the reader may find a more thorough investigation in my "La reception de la Physique d'Ibn Slna", forthcoming in the Acts of the SIHSPAI-colloquium held at Namur, 2003. Our analysis is based on Bahmanyar's major work, Kitab al-Tahsil'(see supra, note 31). 49 The physical part of al-Lawkari's text has unfortunately not yet been edited; it is only available in one manuscript, i.e. Paris, BN. 5900 (ff. 20v-28v contain the section on motion and mainly copy chapters 1, 3 and 4 of the Shifa', as shown in the publication referred to in the previous note; ff. 33v-40v offer that on time, while largely reproducing chapters 11-13 of the Shifa'). 50 Fakhr al-DIn al-Razi, al-Mabahith al-Mashriqiyya, 2Qom, Intisharat Baydar, 1411 H., vol. I, especially pp. 547-96 (motion) and 642-80 (time). Regarding the issue of motion, see also my "La reception de la Physique d'Ibn Slna". 51 For a basic summary of this interrogation, see McGinnis, "Ibn Slna on the Now", pp. 101-2.

14

Ibn Slna, and His Heritage in the Islamic World and in the Latin West

hesitate to introduce now and then fundamental modifications.52 This is best illustrated by his famous theory of substantial motion. Nothing in Ibn Slna's thought prefigures this concept; on the contrary, there is simply no room for such an idea in Ibn Slna, since he absolutely denies the existence of motion in the category of substance. Related to the notion of substantial motion is Mulla Sadra's idea of the "eternal now", which includes the idea that time is directly dependent upon God. Such ideas are clearly non-Avicennian ideas. As with his metaphysics and psychology, Ibn Slna's physics also influenced the Latin West, although in a less prominent manner. Among the scholastics, Albert the Great undoubtedly paid most attention to Ibn Slna's physical theories. He deals in a very critical way with them, now accepting, then rejecting them. For example, he agrees to a large extent with Ibn Slna's conception of "real" time, on the one hand; but he does not accept the latter's thesis that motion also belongs to the category of "situs".53 Whether influenced by Albert or not is still open to question, but the fact is that thinkers such as Ulrich of Strasbourg and Peter Olivi adhered to Ibn Slna's basic conception of time, on the one hand;54 and that such a major scholar as Thomas Aquinas refused to admit the existence of motion in the category of "situs", on the other.55 But Albert was perhaps not the first Latin author to have read Ibn Slna's Physics, at least the parts available in Latin (books 1-2, and the very beginning of book 3). Already Richard Rufus seems to have used that work before 1240 since his relational conception of place might ultimately stem from Ibn Slna.56 In the preceding lines I have tried to offer a significant, although very incomplete outline of the rich heritage of Ibn Slna, as a philosopher, both in the East and the West. But I th:nk it shows enough how great a genius he was.

52 See once more my "La reception de la Physique d'Ibn Slna". Mulla Sadra's ideas are best expressed in his major work al-Asfar. 53 For Albert's use of Ibn Slna regarding the notion of time, see U.R. Jeck, Aristoteles contra Augustinum. Zur Frage nach dem Verhaltnis von Zeit und Seele bei den antiken Aristoteleskommentatoren im arabischen Aristotelismus und im 13. Jahrhundert (Bochumer Stud. Phjlos., 21), Amsterdam, Philadelphia, B.R. Gruner, 1994, 219 ff. (A. deals with both the De quattuor Coaequaevis and the Commentary on the Physics). For the rejection of motion in the category of substance, see Albertus Magnus, Physica, ed. P. Hossfeld, Munster, 1987, V, 1, 7, p. 416. 54 See Jeck, Aristoteles contra Augustinum, pp. 289-95 and 404-27. 55 See Thomas Aquinas, Commentarium in VIIILibros Physicorum, IV, 5. 56 See R. Wood, "Richard Rufus: Physics at Paris before 1240", Document! e Studi sulla Tradizione Filosofica Medievale 5 (1994), pp. 87-127, especially pp. 111-12, 117 and 126.

II Ibn Sma's Ideas of Ultimate Realities. Neoplatonism and the Qur'an as ProblemSolving Paradigms in the Avicennian System

1. I N T R O D U C T I O N

Ibn Sma, in the West known as Avicenna, was born A.D. 980 (H. 370) in a small village near Bukhara, called Afshanah. According to his autobiography (Gohlman, 1974), he spent his youth in Bukhara. There he read, aided by a teacher, the Qur'an and many other literary works. At the age of ten, he began, under the direction of an ascetic, called Isma'il, the study of jurisprudence. Thereafter, Abu cAbd Alah al-Natili introduced him to the science of logic. Ibn Sma proudly affirms that he soon surpassed his master. The latter left him, and Ibn Sina, through personal efforts mastered the natural sciences and metaphysics, as well as medicine. At sixteen years old he engaged himself in medical and jurisprudential practice, but his main concern remained philosophy. In reading Aristotle for the first time to his great disappointment, he realized that he could not understand all that an author like Aristotle had meant. Fortunately he found al-Farabl's book entitled The Objects of the Metaphysics (Fi aghrdd ma bacd at-tabo'a, but cf. Gohlman, 1974, p. 122, N. 30) which opened up his comprehension of Aristotle. After having cured Sultan Nuh ibn Mansur he was given access to his library and to many more books. At the age of twenty-one, about the time his father died, he wrote his first book, al-Majmuc, the Compilation (better known as al-Hikma al-farudiyya, conserved in the unique ms. Ups. 364, and partly edited, Salem, 1953 and 1969). Soon afterwards for unknown reasons he had to move to Gurganj. It may have had something to do with the Ismacili-sympathies of his family (although Ibn Sma would not necessarily have shared those - Gohlman, 1974, p. 19). This was the start of extensive journey, which brought him to several cities, and finally to Jurjan. There he received the partial protection of a certain Shirazl. This provisory rest permitted him to return to writing in the medical (Qanun) and in the philosophical field e.g. his K. al-mabda' wa 'l-ma'dd, Book of the Origin and the Return, (Nurani, 1984). And once again, for unknown reasons, he moved to Rayy, and a few years later to Hamadhan. There he became actively involved in politics. After having cured the prince Shams al-Dawla of a colic he was appointed vizier, but later he was dismissed because of a rebellion in the army. Several days later, however, the prince reinstated him, because he needed Ibn Sma's medical skill. It was

II probably at this time that he started to work on his great philosophical encyclopedia K.al-Shifa', Book of Healing (now completely edited in about 20 volumes, Cairo, 1952-1983). After the death of Shams al-Dawla he was imprisoned for four months for his alleged secret correspondence with a rival of the Hamadhan-dy nasty, cAla al-Dawla, the governor of Isfahan. According to his pupil, al-Juzjam, Ibn Sina wrote a great deal while in prison. His famous esoteric tale Hayy ibn Yaqzan (Mehren, 1889; Corbin, 1954; Goichon, 1959) and his K. al-Hidaya, Book of Direction (Abduh, 1974) with others, can be dated to this period. After having been released from prison, he was required to stay with Taj al-Mulk, probably a son of Shams al-Dawla. But he escaped, by dressing as a sufi, Isfahan. cAla al-Dawla received him with many honours, and provided him with great facilities. This was an extremely productive period for him. He lead a very active and not always wise life. During an expedition he became seriously ill, but notwithstanding his illness, he continued his many activities. On the next expedition he once again fell ill and died A.D. 1037 (H. 428). Besides the writings referred to above, the following are his major works: K. al-Najdt, Book of Deliverance (al-Kurdi, 1913; Fakhry, 1984) (Vattier, 1658; Rahman, 1952; Carame, 1926), Ddnesh-NdmehAli'i, Book of Scientific Knowledge (Mo'in, Meshkat, 1952) (Achena, 1986; Morewedge, 1973), K. al-Ishdrdt wa 'l-Tanbihdt, Book of Remarks and Admonitions (Forget, 1892; S. Dunya, 1968-1971) (Goichon, 1951;Inati, 1984), Ta'liqat, Glosses (Badawi, 1973) (on this work, see however Michot, 1982 and Janssens, 1986). 2. M A J O R INFLUENCES ON IBN S I N A ' S T H O U G H T

There is no doubt that his major inspiration comes from ancient Greece. He considered himself a fay lasuf, i.e. someone who wanted to develop the philosophical quest begun in Hellas. The two great pillars of prime importance to him were the legendary Plato and the first master, Aristotle. While he directly borrowed many ideas of the latter, he was indirectly yet profoundly influenced by the former. While this might initially sound strange, the Arabic faldsifa were deeply convinced of a perfect harmony existing between these two major thinkers. They were undoubtedly strengthened in this belief by the false assumption that some obviously Neoplatonic writings were written by Aristotle. The best known are the so-called Theologia Aristotelis (Dieterici, 1882; Badawi, 1955 (a), p. 1-164) and the L/for P'. But this is clearly an intellectualistic reformulation, far removed from any voluntaristic perspective. Hence, any kind of arbitrariness has to be excluded from man's freedom of choice. Thus, this freedom seems to have included in the 25. J. R. Michot (La destinee de I'homme scion Avicennc. Lc rctour a Dieu (ma'ad), ct I'imagination (Ac. Roy. Bclg. Classe des lettres F. Draguet, V), Lovanii, /Ed. Peelers, 1986) deals in a very substantial way with Ibn Sma" s doctrine of resurrection. His work was a major source of inspiration regarding this matter. Comp., also, my "Ibn Sma's Ideas of Ultimate Reality and Meaning. Neoplatonism and the Qur'an as Problem-Solving Paradigms in the Avicennian System", URAM 10 (87) 252271, p. 258-262. 26. Mahiyyat al-salat, ed. M. Mehren, Traites Mystiques I I I , Leyden, 1894, p. 34. Transl. by A. Arbcrry, Avicenna on neology, London, 1951. Repr. Westport (Conn.), Hyperion Press, 1979, p. 54-55. 27. Isharat, p. 188. 28. Sec W. C. Klein, Al-As'ari's al-lbanah 'an usul addiyanah (The Elucidation of Islam's Foundation), transl. with imrod. and notes (Am. Orient, ser., 19) New Haven, 1940 (repr. New York, Kraus Repr. Corp., 1967), p. 113. Comp. I I . Daibcr, Das Thcologisch-Philosophische System des Mu'ammar ihn Abhad as-Sulami (gcst. 830 n. Chr) (BTS, 19), Beirut / Wiesbaden, F. Steincr, 1975, esp. p. 216. 29. R. Al-qadar, cd. M. Mehren, Traites Mystiques IV, Leyden, 1899, p. 10.

Ill 118

universal order of the intelligible. Here, undoubtedly some tension appears inside Ibn Sina's own system, which one might, at least partially, resolve by interpreting Ibn Sina's notion of freedom as an appropriation of the measure (qadar!) which God put into the world-order (also qadar!) which is the best posible order. For the actual realization of this appropriation, man has to open himself to the enlightenmant of the Agent Intellect, but this is not an automatically given one. Moreover, our analysis evidences that Ibn Sina accepts the real contingency of earthly life. He not only presents an emanationism system, but he clearly tries to combine emanationism with creationism —not without some difficulties (may tensions in his system have their final sourse here!30, but nonetheless opening most interesting perspectives for the further history of philosophy. Finally, his doctrine of freedom shows the necessity of religion, and might be considered to be a philosophical counterpart of the Islamic idea of freedom31.

30. L. E. Goodman (Avicenna (Arabic Thought and Culture), London, Routledge, 1992) may give the impression that no such tensions are present in Ibn Sina's system. This is surely not the case. But I must conceed that his expose of Ibn Sina's metaphysics (p. 49-122) is outstanding, and was indeed of great significance for our present research, as well as part of its wording. I also took many profit of A. Ivry, "Destiny Revisited. Avicenna's concept of Determinism", in: M. Narmura (ed.), Islamic Theology and Philosophy, Albany (N. Y.), Suny, 1984, p. 160-171; and, to a lesser extent, of Sh. Inati, "An Examination of Ibn Sina's Theodicy. Dissolving the Problem of Evil", The New Scholasticism 58 (1984) 170-186. 31. I am aware that such an affirmation is somewhat risky, especially since there is no «clerical authority» in Islam to fix the «orthodox» point of view!

IV

Creation and Emanation in Ibn Sma1 Ibn Sma (Avicenna) presents in his metaphysics a triadic emanationistic structure2. Each higher Intelligence is the source of a further Intelligence, and of the soul and the body of the sphere proper to it. This threefold emanation is due to a threefold understanding, i.e., one of God, as the ultimate necessary source, and one of itself as possible, divided however into two moments : possible in se, and necessary ab alio. This doctrine is well known, and therefore deserves no further analysis. Let us just state that : 1. Ibn Sma himself limits the number of Higher Intelligences to ten. As far as I can see, he gives no proper explanation for this fact. However, he refers to the loss of power, which naturally accompanies the emanative process. But, how can one through this general observation arrive at the exact number of ten ? It looks as if Ibn Sma has chosen this number on an a posteriori basis, i.e., the number of celestial movements, according to the prevailing cosmologies of his time. Moreover, Ibn Sina's decision might also have been influenced by some kind of numerical symbolism, or even by some Qur'anic considerations 3 ; 2. Ibn Sma especially emphasizes that from God, as the One, only one Being can proceed, based on the adage : « From the One, insofar as it is one, only one Being can come into existence ». This principle, notwithstanding its obvious Neoplatonic origin, seems to have received its actual formulation 1

This paper is a somewhat reworked version of part II, chapter 2, 2 of my Ph. D.thesis : Avicenna : tussen neoplatonisme en islam (Avicenna : between Meoplatonism and Islam), Leuven 1984, 2 vol., I, pp. 75-112. Its point of departure was constituted by A.-M. GOICHON, La distinction de I 'essence et de I 'existence d'apres Ibn Slnd (Avicenne), Paris 1937, pp. 210 ff. In view of the many new materials, which since that time have been published, a renewed analysis of Ibn Sina's vocabulary of emanation and creation imposed almost itself. It revealed that Goichon's study was indeed very meritorious, but now somewhat outdated, and clearly in need of some important additions, and even corrections. 2 An outline of this emanation-scheme is present in all the major works of Ibn Sina. See e.g. IBN SINA, Al-Sifa'. Al-Ilahiyyat (La Metaphysique), Eds. G.-C. ANAWATI, S. ZAYED M. Y. MUSA, S. DUNYA, 2 vol. in 1, Cairo 1960, b. IX, c. 4 - 5. 3 Regarding the possible influence of numerical symbolism, see: S. H. NASR, Three Muslim Sages. Avicenna, Suhrawardl, Ibn 'Arabl, Cambridge, Mass. 1969, p. 140, n. 55.

IV 456

in Ibn Sma himself 4 , and clearly helps him to explain how a plural universe can derive from the wholly unitary First Cause. Thus far, a very brief evocation of Ibn SIna's emanative scheme. But, what does he understand by « emanation » or « emanating » ? Such question may prima facie sound as redundant. In fact, Ibn Sma at first sight endorses its Plotinian sense, notwithstanding that there exist obvious differences in both's respective emanative schemes. But, as soon as one starts to deal in a systematic way with Ibn SIna's metaphysics, one cannot but observe that Ibn SIna's major aim is to combine a metaphysics of necessity with a metaphysics of contingency, or, expressed in other words, between emanation and creation. Viewed from this perspective, Ibn SIna's use of different terms regarding the very notions of emanation and creation is undoubtedly not devoid of any significance. On the contrary ! It deserves full attention. In what follows, I will examine one by one the major terms, by which Ibn Sma evokes the idea of either emanation or creation. I. IBN SINA'S EMANATIVE VOCABULARY When reading Ibn SIna's works, one finds no less than four expressions, which in one way or another refer to the very notion of emanation, i.e., fayd, tagalll, sudur and inbidt. To these four, a fifth may be added: inbigds, which appears twice in his Commentary on the Theologia Aristotelis5. However, since this term does not appear anywhere else in Ibn SIna's works, one may wonder whether Ibn Sma really did subscribe to it, all the more since the two passages, where this term can be found, seem to belong to the textus itself, on which the comment is made. It has to As to the eventual presence of a Qur'anic background, see: IBN SINA, R. ft lib at alnubuwwd, ed. M. MARMURA, Beirut 1968, pp. 53 - 54. 4 H. A. DAVIDSON, Alfarabi, Avicenna, and Averroes, on Intellect. Their Cosmologies, Theories of the Active Intellect, and Theories of Human Intellect, New York - Oxford 1992, p. 75, n. 3. 5 IBN SINA, Sarh "Kitab Utulilgiyd" al-mansub ila Arista, in ( A. BADAWI (ed.), Aristil 'inda 'l-'Arab, Cairo 1947 (repr. Kuwait 1978), pp. 62.22 and 63.14. Regarding the exact contents of this edition, as well as the difference which normally has to be made between the sarh- and the tafsfr-parts of Ibn SIna's commentary (not taken into account by me in this paper, since of no significance for my actual purpose), cf. D. GUTAS, Notes and Texts from Cairo-Manuscripts, I I : Texts from Avicenna's Library in a Copy by c Abdar-Raizaq as-Signahi, « Manuscripts of the Middle East », 2, 1987, pp. 12-13.

IV CREATION AND EMANATION IN IBN SINA

457

be noted moreover that, whereas inbigds in the Neoplatonica Arabica refers to the direct emanation of the First Intelligence (and of all Being) from God6, Ibn Sma expresses this "generation" in terms of ibda (cf. infra). Finally, it is certain that inbigds never did receive such a central place in Ibn Sma thought as in that of e.g., the Ihwan al-Safa' 7 . Therefore, I will only deal with the four above mentioned notions in greater detail. A. Fayd

(fada)

The radix fy d, in its most basic meaning 8 , refers to the overflowing of water, or the abundancy of water — a not necessarily negative fact, since it gives rise to fertility. It moreover signifies the departing, i.e., the "flowing away" of the soul at the time of death, in other words : the very fact that someone dies. It also means the spreading of news, or even the revelation of a secret. In its substantive form, it may also designate an act of benevolence on behalf of God. The common factor of all this pre-philosophical meanings is clearly the idea of "overflowing", with a basically positive connotation. In the Theologia Aristotelis9, it is one of the main terms expressing the very idea of emanation. Considered to be a positive fact, the idea of overflowing is nevertheless unambiguously expressed in terms of pure necessity. This necessary character is still more predominantly present in the Theologian use of inbagasa, the "gushing forth". In its basic sense10, this term refers to the fact that water searches its way to flow on, or to the water welling from a spring. Especially in this latter case, one faces a plainly automatical process. Therefore, when the Theologia adopts this term, a kind of pantheism can hardly be avoided. Since the author of the Theologia was probably aware of this implication, he somehow 6 See e.g., Die sogenannte Theologie des Aristoteles aus arabischen Handschriften zum ersten Mai herausgegeben von F. DIETERICI, Leipzig 1882 (repr. Amsterdam 1965), p. 136. 7 As far as I can see the notion of inbigds is in the Rasd'il of the Ihwan al-Safa' the most important emanation-concept after that otfayd. 8 E. W. LANE, An Arabic-English Lexicon, 8 vol., London 1885 (repr. Beirut 1968), pp. 2471-2472. 9 Having done a survey of the most important writings of the Neoplatonica Arabica, i.e., Dicta Sapientis Graeci, Epistola de Scientia Divina, Proclus Arabus and Liber de Causis, I have not found any new elements with respect to those discovered in the Theologia Aristotelis, which could contribute in a significant way to our analysis of Ibn Sma's theory. Therefore, I will most of the time limit myself to give references to the Theologia, also in as far as the other terms of emanation and creation are concerned. 10 LANE, Lexicon cit., p. 153.

IV 458

tried to exclude, or, at least, attenuate it. He, therefore, always links with it very specific notions, such as e.g., those of life (haydt), power (quwwa), light (nur) and good (hayr)n. Probably out of monotheistic motives 12 , he even seems to introduce a creationist correction by making use of the very notion of creatio ex nihilo, i.e., ibda (cf. infra for a more detailed analysis of this notion). However, there still remains a predominantly pantheistic undertone in the Theologia Aristotelis, especially where it affirms that God brings forth all things from His very Being alone (bi-annahu faqat)]3. Hence, this regards clearly an act, devoid of any knowledge or will14. As far as Ibn Slna is concerned, I already mentioned that he avoids the stronger notion ofinbagasa. His preference goes clearly tofdda, and its derived notions. Note moreover that in sharp contrast with the Theologia Aristotelis, he never posits that the created things are generated « by God's Being alone ». Finally, one finds in his use of « overflowing » some accentuations, which appear to be totally absent in the Theologia Aristotelis. First of all, he identifies the overflowing of the created things as one of the « concomitants » of God's brilliance15. In other words, the realization of the created realm does not belong to the core of the divine essence. Thus being a « concomitant », it is fundamentally something accidental. Therefore, it is certainly no coincidence that Ibn Slna does neither characterize the many relations, which may be ascribed to God as constitutiva nor as parts of the divine essence16. It is worth mentioning that he always precedes this denial by the very affirmation that God is the One from whom everything is overflowing. Furthermore, the gap between Creator and creation seems to be widened, and, at the same time, even to be radicalized, where Ibn Slna states that the concrete overflowing (fayadan) is separated (mubayin, Lat. discretus) from God's essence17. At least in its very literality, this expression excludes any real participation of the created Beings in the 11

Cf. e.g., DIETERICI (ed.), Die sogenannteTheologie cit., pp. 104, 105 and 111 (life) ; 157 (power); 51 and 107 (light) ; 77 and 111 (good). In all the given references here God functions as the source of the overflowing. 12 C. D'ANCONA COSTA, Per un profile filosofico dell'autore della "Teologia di Aristotele", « Medioevo », 17, 1991, pp. 112 ff. 13 DIETERICI (Ed.), Die sogenannte Theologie cit., pp. 109, 118 and 152. 14 D'ANCONA COSTA, Per un profile cit., pp. 96, 115, 121-126. 15 IBN SINA, Al-Sifa. Al-Ilahiyydt cit., p. 403.1. 16

17

Ibid., p. 344.1-5. Ibid., p. 403.13.

IV CREATION AND EMANATION IN IBN SINA

459

divine essence. In this respect, it is most significant that Ibn Sma himself explicitly characterizes the overflowed Beings as « caused » (malul)n. Moreover, one almost looks in vain for any direct connection — so typical for the Theologia Aristotelis (and also present in the Liber de Causis) — between the very notions of life and power, on the one hand, and the concept of overflowing, on the other19. The notion of light mentioned by Ibn Sma, will be dealt with when discussing the notion of tagalld (cf. infra, 1 B). On the contrary, Ibn Sma clearly prefers to combine the notion of good with the idea of overflowing 20 . Did we not observe the presence of this very same combination in the Theologia Aristotelis ? In fact, we did. Nevertheless, Ibn Sina's use seems to be peculiar to him, and hence does not cope completely with that of the Theologia. He immediately adds the notion of good to that of order, and directly links this order with God's will : « The arrangement of the Beings, emanating from the First, is not because what is aimed at in them is this (litt. the) order, but they only possess this (litt. the) order, because it is willed by the First, and this being willed is identical (nafs) with this (litt. the) order » 21 . In strong opposition with the Theologia Aristotelis, Ibn Sma explicitly includes the attribute of will in God's primary action, and, hence, indirectly also the attribute of knowledge. For him, both attributes of will and knowledge are identical with God's essence. However, this fact does not necessarily mean that any distinction between them is impossible. I believe that Ibn Sina left enough space to keep them distinguished, at least at a logical level. Finally, although Ibn Sma agrees with the Theologia Aristotelis that the overflowing of the created Beings constitutes an eternal divine act, he underlines, in sharp contrast with the « by His very Being alone » of the Theologia, the total absence of any inner or outer motive for the divine acting : « Emanation (fayd) is the action of the actor who is eternal in his 18

Ibid., p. 14.16 and p. 366.3-7. The notion of life is lacking completely, whereas that of power was only once found, and then in connection with the planets (ibid., p. 401.11). 20 E.g., ibid., pp. 366.9 ; 415.5-6 and 425.17. It has to be noted that this combination is most of the time followed by an evocation of the idea of "order". 21 IBN SINA, Al-Ta'llqat, ed. C A. BADAWI, Cairo 1973, p. 49.11-12. 19

IV 460 action. His action is not due to a cause requesting it, nor for a goal except this very action itself » 22 . The emphasis is clearly on the absolute autonomy of God's action, the radical contingency of all created Beings being hereby at once guaranteed. So, the way for the affirmation of a creatio ex nihilo remains open, even if it is absolutely certain that Ibn Sma excludes the possibility of its having started in time. At most, he admits of an eternal creation. And even this is not quite sure, since the idea of overflowing inevitably seems to imply a necessary link between the source of emanation and the emanated. Of course, as I have shown, there are some obvious creationistic corrections, and they are clearly in a more prominent way present in Ibn Sma than they were in the Theologia Aristotelis. Hereby, Ibn Sma's notion of overflowing more emphatically includes the basic pre-philosophical idea of "beneficent abundance". In all this, he certainly develops further a tendency which already showed itself in the Neoplatonica Arabica. However, it would be far too premature to conclude on this basis that he in fact adhered a genuine idea of creation. A further analysis of Ibn Sma's emanationistic/creationistic vocabulary imposes itself. Therefore, I will continue with the examination of the concept of manifestation. B. Tagalll (tagalld). Before thoroughly dealing with the very notion of manifestation, tagalll, I first want to pay some attention to the idea of the overflowing of light. This terminology is only found in two minor treatises, i.e., Fulfilment of Prayer, and How to visit Tombs and Commentary on s. alfalaq23. From their title it is obvious that both treatises concern an explicit religious theme, i.e., a devotional practice, respectively an exegesis of a sura of the Holy Qur'an. This might suggest that Ibn Sma reserves the expression of overflowing of light for a specific religious-mystical context. However, this is far from self-evident, since such idea also fits well a Neoplatonic context. In the former of the two treatises, God is described as « the Well of the overflowing of light (fayaddn al-nur} over what is beneath Him » 24 . In view of the description of the emanation of all the kinds of created Beings following immediately upon this 22

Ibid., p. 81.1-2.

23 por recen t editions, or reprints of these works, see my An Annotated Bibliography on Ibn Slnd (1970 - 1989). Including Arabic and Persian Publications, and Turkish and Russian References, Leuven 1991, pp. 64-65. 24 IBN SiNA, Igdbat al-du'd wa-kayflyat al-ziydra, ed. M. MEHREN, Traites Mystiques.

IV CREATION AND EMANATION IN IBN SlNA

461

statement, Goichon's interpretation 25 of this overflowing in terms of an existentialising act almost becomes obvious. However, a few lines later, it is said : « And the Intellect, proper to the Sphere of the Moon, lets overflow its light over the human souls, in order that they be guided by it in (their) research of the intelligibles, such as the overflowing of the light of the Sun over the corporeal Beings is in order that the eye perceives them » 26 .

Here, the overflowing is clearly situated on a gnoseological, and not on an ontological level. It therefore seems possible, and even probable, that even the previous citation uses this idea in the very same perspective. However it might be that the existentializing meaning is certainly the only possible, where Ibn Sma in his Commentary on s. al-falaq mentions of « an overflowing of the light of existence over the possible quiddities » 27 . Moreover, this overflowing is directly linked by Ibn Sma with God's « bursting » (infilaq). It more and more looks as if, according to Ibn Sma, God has to open Himself, in order that all Being gushes from Him. If this is the right interpretation, then Ibn Sma clearly adheres a kind of pan(en)theism, either consciously, or unconsciously. But such a radical conclusion imposes only itself, when the above statement is taken separately. As soon as one takes into account the more general context of the complete treatise, a far more balanced judgment is possible. Ibn Sma's use of the notion of bursting, infilaq, then almost becomes natural in view of the presence of the sura of the intimately related term offalaq in the title. Where the main topic of the treatise is the problem of evil, infilaq might merely refer to the "breaking of of the divine day. And, in this latter perspective, the statement can be understood as expressing the idea that the existence of the good (or should one say : the good of existence ?) is only possible in the (metaphorical !) light of God. Therefore, Ibn Sina's use of the expression of the overflowing of light, in no way leads to the inevitable conclusion that he defends a strict emanationistic-pantheistic doctrine. Leiden 1894, III, p. 45 ; ed. H. c Asi, Al-tafsir al-Qur'dnl wa 'l-luga 'l-sufftya ft falsafat Ibn Sina, Beirut 1983, p. 284. 25 GOICHON, La distinction cit., p. 213. 26 Ibid., p. 46, resp. pp. 285-286. 27 IBNSINA, Tafsir al-mu'awwiddt al-uld, in : Garni' al-Badd'l, Cairo 1917, p. 25 ; and c in Asi, Al-tafsir cit., p. 116.

IV 462

Perhaps, some more decisive arguments can be obtained from a closer examination of his notion of manifestation. Also this time I have found traces of it in only two of his minor works, i.e., his Treatise on Love28 and his Commentary on the Theologia Aristotelis. Notwithstanding they are limited in size, it has to be mentioned here that of the Commentary on the Theologia Aristotelis only fragments seem to be extant 29 , but their philosophical significance can hardly be doubted. But before dealing with Ibn Sina's use, let us have a brief look at the pre-philosophical meanings this notion has. The radix (t) g I y30 evokes in its very primary sense the nonfigurative idea of clarity, i.e., a thing becoming bright because of its being illuminated by light. Very soon, the figurative sense of clarity was also accepted, and hence tagalld (tagalll) became also the indication for what is lucid for the mind. In the Qur'an, it clearly denotes God's manifestation, i.e., the appearance of His glory. In direct connection with this latter meaning, the mystical tradition developed two further meanings, i.e., that of God's glorious Light, issuing from His essence, and that of transformation, caused by the divine Grace in this world. Let us now turn to Ibn Sma;s interpretation of this notion. In the Theologia Aristotelis, there is a sentence that states that the yoi)^ (in its Plotinian sense) through its self-knowledge knows all other things. This statement could easily be misunderstood, affirming as it were that the you? through its self-knowledge also knows the highest Being, i.e., God. That such cannot be the case is fully stressed by Ibn Sma in his comment : « But it ought to be that their being endowed with reason by the First principle insofar as there is a manifestation of the First principle to them. When the First principle manifests itself to them, they comprehend It, they comprehend themselves, and they comprehend everything (situated) in the further degrees (of Being) » 31 . 28

For recent editions, and reprints, see once more my Bibliography cit., p. 68. From the published fragments, one gets the impression that they only are part of a more encompassing commentary, which however seems no longer extant. Nevertheless, it still may be hoped that more fragments will be discovered in not yet described manuscripts. The importance of, and the need for a more exhaustive catalogue of the Avicennian manuscripts is well shown by J. MICHOT, Un important recueil avicennien du VII/XIII s. : La Majmu'a Huseyin felebi 1194 de Brousse, « Bulletin de philosophic medievale », 33, 1991, pp. 121-129. 30 LANE, Lexicon cit., pp. 446-447. 31 IBN SINA, Sarh cit., p. 49. With G. VAJDA, Les notes d'Avicenne sur la Theologie 29

IV CREATION AND EMANATION IN IBN SINA

463

The Higher Intelligences (note that they replace the voi)9 of the original text !) are not only unable to know God without any previous manifestation on His side, but without this latter they are also unable to know themselves, as well as all the Beings beneath them. Hence, God's manifestation guarantees the possibility of knowledge on their behalf, and, hence, evokes the idea of a figurative illumination, i.e., a brightening of the intellect. In the Treatise on Love, at least at first sight, a more existential interpretation of the notion of manifestation comes to the fore. In this respect, two most relevant statements appear in the seventh chapter. At its very beginnings, it is said : « Indeed, God (litt. He) is loving by His excellence because He is conferring His manifestation 32 , and the existence of (all) the things is together with His manifestation (bi-tagalllhi) » 33 .

Near the end of the chapter, a possibly even stronger affirmation is given : « If no manifestation of the absolute Good could have happened, there could be no obtaining (mayl) from Him, and if nothing were obtained from Him, there would be no existence (wugud). Thus, if there is no manifestation, there is no existence, hence His manifestation is the cause of all Being » 34 .

One naturally is tempted to understand Ibn Sina's concept of manifestation in both passages as being synonymous with the pure Neoplatonic idea of emanation, or to put it in Goichon's words: of a « fluxcreateur ».The very fact that the source of this manifestation is always God's knowledge, in no way constitutes a threat to such an interpretation, since Ibn Sma d'Aristote, « Revue Thomiste », 51, 1951, p. 368, n. 3, I read taliya instead of talit. 32 Or, according to Mehren's version: « He loves by His excellence that His manifestation should be received ». 33 IBN SINA, R. ft 'l-^isq, ed. MEHREN, Traites mystiques cit., Ill, p. 22 ; ed. cAsi, Al-tafslr cit., p. 265. It has to be noted that some manuscripts read tagalllhi instead of bitagalllhi at the end, in which case there is a complete identification between the existence of the things and God's manifestation. If that reading is correct, we have undoubtedly to do with an outspoken pantheistic affirmation. But Ibn Sma always avoids such clear pantheistic statements. Therefore, the version with the pre-added "hi" is probably the more correct one. My English translation of tagalll by « manifestation », and not by "effulgence", is due to an explicit wish to render the text as neutral as possible, in order to leave open as much room as possible for further interpretation. 34 Ibid., p. 26, resp. p. 268.

IV 464

considers the divine knowledge to be a creative knowledge, i.e., for God to know is to create. Moreover, the whole treatise calls for a Neoplatonic structure, since special attention is paid to the second movement of the emanative scheme, i.e., the reditus-movement. It might be correct that we have here to do with a further Neoplatonic development of the connection between « the Good » and « the sun » in Plato's Republic*5. Although I surely do not want to deny the presence of ^- sometimes even strong — (Neo)Platonic elements in this tract, I do not believe that it is right to characterize it as a mere Neoplatonic treatise. In fact, as far as its central issue is concerned, i.e., the explicit linkage between natural love and the strife for assimilation with God in view of one's perfection, one is already confronted with a derivation from Peripatetic sources36. But Ibn Sina, as usual, has not slavishly followed this latter sources. He has somewhat modified them, probably — at least, to some extent — out of religious motives. E.g., he does not hesitate to introduce into the Peripatetic notion of natural love a specific Islamic connotation, i.e., the idea of the possibility of a "free" reception by the creature of God's manifestation. Note that only in doing so, he can make a full distinction between the elevated human souls and the lower ones, both in this life and in the life in the hereafter. Their actual characterization directly depends upon their degree of ability to receive God's manifestation. I cannot but see in this the expression of a will to islamize somehow the Greek Peripatetic sources. But if this is the case, is it then not normal to suppose that also the Neoplatonic elements underwent some similar islamization ? Is it not striking that God's manifestation, viewed from the totality of the treatise, appears to be not so much the efficient cause of the existence of the created Beings, but rather their final cause, ensuring rather their conservation in life than their generation ? In affirming this, I do not wish to deny that God's manifestation does nowhere contribute to the coming into Being of the created realm. Insofar as each created Being (at least in the super-lunar world) functions as a mirror of God's manifestation, it, by becoming diaphanous, and hence its ongoing reflecting of the very same divine manifestation, gives rise to the origination of further Beings(-mirrors)37. 35

F. RUNDGREN, Avicenna on Love. Studies in theRisdlafimdhlyatal-^isq, I, « Orientalia Suecana », 27-28, 1978-79, p. 57. 36 M. SORETH, Text- und Quellenkritische Bemerkungen zu Ibn Slnd's Risdlaft 'l-^isq, « Oriens », 17, 1964, pp. 126 ff. 37 J. MI£HOT, La destinee de I'homme selon Avicenne. Le retour a Dieu et Vimagination, Lovanii 1986, pp. 87-93 (Academic Royale de Belgique. Classe des Lettres. Fonds

IV CREATION AND EMANATION IN IBN SINA

465

But, above all, God's manifestation helps each created Being to realize its real perfection. It functions as a true transforming power inside the creatures. In this last idea, I detect a — according to the real mystics perhaps poor, but nevertheless acceptable — philosophical explanation of the mystical idea of transfiguration. And with genuine Islamic mysticism, Ibn Sina categorically rejects the possibility of the realization of a full unification of the creatures with God. If my interpretation is correct, one may detect here a conflation between the doctrine of the Neoplatonic reditus-movement and the Islamic mystical idea of the search for God's Glory. Is it a coincidence that I have in vain looked for a single trace of the very notion of tagalll(d) in the Neoplatonica Arabica ? It may also be mentioned that at the beginning of chapter seven it is affirmed, at least according to two of the three printed versions, that the divine manifestation is posterior to the existence of the Beings who love God, a reading which finds explicit support in the Commentary on the Theologia Aristotelis, where Ibn Sina twice states that the existence of the Highest Intellect precedes the divine manifestation38. But, as this reading is not absolutely certain, I do not want to overemphasize its importance. Generally speaking, it seems that a metaphorical interpretation of Ibn Sma's notion of manifestation has to be preferred over an existential one. But the latter cannot in all honesty been completely eclipsed, nor can the Neoplatonic-emanationistic undertone completely be ignored. Ibn Sina clearly struggles with this latter. Once more, it is obvious that he introduces outspoken monotheistic-creationistic corrections. But are they sufficient for the affirmation of a genuine idea of creation ? It becomes more and more obvious that no simple answer is available in this respect. It can only be hoped that the analysis of the remaining terminology will help to clarify this matter. C. Inbidt (inbaata) and sudur (sadara). I will first concentrate on the notion of "pouring out", which in Arabic is evoked by the radix (in) b c £39. In its very basic meaning it simply signifies to be sent. Among its derivative meanings, one finds the pouring out of water, as though impelled or propelled, as well as that of becoming roused. In the Theologia Aristotelis, it functions as one of the terms, albeit a very exceptional one, to indicate the Plotinian Draguet, 5), but with a slightly different emphasis then ours. 38 IBN SINA, Sarh cit., pp. 49 and 60. 39 LANE, Lexicon cit., p. 223.

IV 466

idea of TTpooSo?40. However, it has to be noted that most of the time the Theologia situates this pouring out at the mediating level of the Intellect, or of the Soul. This use has explicitly been preserved in the Ismailite circles of the tenth and eleventh century, as e.g., al-Sigistani and, above all, al-Kirmani 41 . Investigating Ibn Sina's use, I have found about ten occurrences in the Ildhiyydt of the Sifd'42. In many of these cases, an outspoken connection with the notion of desire (sawq) is present. But what is more important is the conclusion that from all the passages only two seem to involve an emanationistic undertone. In all the other ones, it is quite obvious that Ibn Sina in no way links the pouring out-notion with the idea of the further generation of things. I will now have a closer look at the two mentioned ones. The context of the first consists in the intellection by the highest spheres of God. Ibn Sina states that due to this intellection, the sphere forgets everything else, but, at the same time, « what lower is in degree is poured out, and there is the desire to resemble Him as far as possible » 43 .

Let me immediately stress that what is at issue here in the first place is not the generation of a new Being, but the movement of a celestial sphere. This movement is the unconscious and unintended result of the strife by the concerned sphere for its highest possible perfection. Moreover, through this movement the maintenance of the sphere's existence is guaranteed. As to its generation, this movement plays no role whatsoever, although it definitely « contributes » to the generation of what is lower in existence. 40

Cf. e.g. DIETERICI (ed.), Die sogenannte Theologie cit., pp. 6, 12 and 90. For al-Sigistani, see: P. WALKER, Early philosophical Shi'ism. The Ismaili NeoplatonismofAbu Ya^qubal-Sijistanl, Cambridge 1993, pp. 53, 85-86, 98 (Cambridge Studies in Islamic Civilization). As to al-Kirmani, see I. NETTON, Allah Transcendent. Studies in the Structure and Semiotics of Islamic Philosophy, Theology and Cosmology, London and New York 1989, pp. 222-228. 42 Thanks to the Lexica of the Avicenna Latinus, I was able to discover the presence of this term in Ibn Sma's writings — a presence which has escaped the attention of such great Avicenna scholars as A.-M. Goichon and L. Gardet. For all its precise occurrences, see S. VAN RIET, Avicenna Latinus. Liber de philosophia prima sive scientia divina, I - X. Lexiques, Louvain-la-Neuve - Leiden 1983, p. 10 sub 75. 43 IBN SINA, Al-Sifa. Al-Ildhiyydt cit., pp. 391.18 - 392.1. 41

IV CREATION AND EMANATION IN IBN SINA

467

But how should this contribution be understood ? Is God delegating His creationistic power to that sphere ? In view of Ibn Sma's Islamic background, which certainly has influenced his thought, as was already previously shown, this sounds improbable, a fact which seems to be confirmed by the affirmation that « the existence of all things is poured out from God (litt. Him), without any desire or purpose » 44 .

This time, God is clearly defined as the generator of all things. He, and nothing else, is the proper source of all Being. In all this, we may detect Ibn Sma's wish to adhere the Neoplatonic idea of a vertical causality, perhaps somewhat modified by the (Proclean inspired ?) introduction of a mediating structure. However, this mediation reveals to be of a rather weak form. Ibn Sma emphasizes in a very strong way that everything depends on God — even the mediating role played by the celestial Beings, since it is neither their autonomous will, nor a blind fatum, but their desire to resemble God, which gives rise to the coming into Being of new creatures. As far as I can see, the mentioning of such a desire is completely absent in the Theologia Aristotelis, as well as in the already mentioned Isma'ilite authors. It might well be that once more Islamic motives, as e.g., the divine omnipotence, and the radical distance between God and man (which makes that the Revelation is given to the Prophet by mediation of the archangel), have pushed Ibn Sma to adopt this modification. However, as afaylasufhe has worked out a pure philosophical solution — and in this sense Proclus might have offered him an interesting basis for a philosophical solution with the problem which he was facing. But let us not forget that the notion of pouring out does not occupy a central place in his works. Sudur, sadara, which in their current use45 mean that an act, respectively a saying emanates from an acting subject, respectively a speaker, certainly constitute on the philosophical level the terminus technicus for the very idea of emanation. Contrary to the notion of pouring out, it is of frequent occurrence in Ibn Sma's writings. However, surprisingly enough in the Ilahiyyat of the Sifd', I did discover one single statement, which links this act of emanating directly with the divine essence :

44 45

Ibid., p. 397.18.

LANE, Lexicon cit., p. 1160.

IV 468

« There is no hindrance to, nor aversion from the emanation of all (Being) from Him (i.e., God) » 46 . Even in this case, Ibn Sma obviously tries to avoid any pantheistic implication by stressing that this emanation belongs to the concomitantia of God's Glory. He moreover adds that it is in no way subject to any outer or inner hindrance. By affirming this latter, he gives the impression that the act of emanation is the sheer result of a free and autonomous decision of God, and, as such, to open the way for the affirmation of the genuine idea oicreatio ex nihilo. This impression is even reinforced by his making the emanation dependent upon a preceding action. In his Treatise of the Throne on divine Unity and Oneness (Tawhid), the chapter concerning the emanation of the created Beings , is significantly entitled: « On the Emanation of the Actions issuing from Him (i.e., God) » 47 . Overlooking the totality of Ibn Sma's emanation-vocabulary, I cannot but conclude that it clearly reveals a tendency to eliminate as much as possible all pan(en)theistic connotations. This tendency shows itself already in the Neoplatonica Arabica, but it appears to be much more prominent in Ibn Sma. The distance to the Greek originals of the Neoplatonic tradition becomes greater, not only in time, but also in contents. This might partially be due to the fact that Ibn Sma wants to inscribe himself more explicitly in the Aristotelian-Peripatetic tradition. But it might also partially be based on some religious, i.e., Islamic motives. Whatever it may be, it is quite obvious that Ibn Sma tries to avoid a purely necessitarian monism. But even if such is his expressed goal, one may wonder whether he really succeeded in realizing it ? I think it is still premature to give a definite answer to this delicate question. The detailed analysis of his creation-vocabulary is certainly needed, before one can attempt to formulate a definite conclusion. II. THE TERMINOLOGY OF CREATION

As with the notion of emanation, four terms are used by Ibn Sma to evoke the very idea of creation, i. e., takwln, generation; ihddt, origination; halq, (material) creation, and ibda , creation from nothing. 46

IBN SINA, Al-Sifa. Al-Ilahiyyat cit., p. 402.17. IBN SINA, R. al-^arslya ft 'l-tawhid, in: Magmu rasd'il al-sayh al-ra'is, Hayderabad 1353 H., p. 15. 47

IV CREATION AND EMANATION IN IBN SINA

469

A. Takwln and ihddt. Since L. Gardet, in his famous study La pensee religieuse d'Avicenne, has given a very fine definition of both of these terms, I will limit myself to a very basic evocation. Gardet defines the former of the two notions as « the mediated production of perishable Beings », while he identifies the latter as «the production of contingent Beings, either eternal, or, more usually, temporary ones »48. As such, they do not really contribute to presenting a solution to the question whether one may ascribe to Ibn Sma the very concept of creation or not. However, regarding the latter notion of ihddt,! want to stress that it is stricto sensu only used by Ibn Sma with respect to material Beings49. Moreover, it seems to imply a strong contingency, but I will postpone my discussion of the very idea of contingency till the final part of my paper. B. Halq

In its most basic sense50, the notion of halq signifies the making of a thing according to its measure. In the Qur'an, it refers to the idea of creatio ex nihilo (in tempore), and, in full accordance with this latter use, it became in the kaldm a technical term for the very idea of temporal creation. Ibn Sina links with it a general sense, i.e., the « making acquire (ifdda) existence », although, more often, a more specific one, i.e., to « make acquire existence, resulting from a matter and a form, whatever it may be » 51 . Since this latter definition is the more specific one, I fully agree with Gardet, when he identifies Ibn SIna's halq-notion as « the production, with or without mediation, of corporeal Beings, either perishable, or imperishable ones » 52 . However, it has to be observed that not only the notion of corporeality, but also that of measure clearly characterizes Ibn Sma's understanding of this term. But even if one takes this latter aspect into consideration, it is certain that Ibn Sma's use of this notion has nothing in common with that of the kaldm, and, hence, of the Qur'an. He thus seems to miss a serious chance to introduce into his system the religious idea of creation. So, are we confronted here with a tendency, directly opposed to the one we discovered in his emanation-terminology, 48

L. GARDET, La pensee religieuse d'Avicenne, Paris 1951, pp. 64-65 (Etudes de philosophic medievale, 41). 49 Cf. e.g., IBN SINA, Al-Sifa. Al-Ilahiyyat cit., p. 181.7. 50 LANE, Lexicon cit., pp. 799-800. 51 IBN SINA, K. al-hudud, ed. A.-M. GOICHON, Cairo 1963, p. 43 (Ar.), p. 60 (Fr.). 52 GARDET, La pensee cit., p. 65.

IV 470 and which consisted in a move towards a monotheistic-creationistic perspective ? This would be undoubtedly a too strong conclusion. But it cannot be denied that things seem to become anew more complicated. At most, Ibn Sina can still adhere a kind of "mediated" creation. But does he do so ? The remaining notion of ibda will have to bring the ultimate clarification. C.Ibda This term, and its related forms, very basically is synonymous with creation, indicating that something, which was totally inexistent before, now originated 53 . In the Qur'an it is not completely absent, and it refers in it to what one could call the very first, absolute creation of Heaven and Earth54. This notion also frequently occurs in the Neoplatonica Arabica55, as well as in the writings of the early Isma c ilite authors56. As already previously observed, in both these latter cases there clearly exists a will to modify somewhat the original Neoplatonic emanationsystem into the direction of a monotheistic creationism. By introducing the very notion of ibda', this process attains its apogee. But, especially regarding the Neoplatonica Arabica, one may wonder whether the way, and the very terminology by which they described the process of emanation, did not after all remain too dependent upon their Greek sources, in order to successfully introduce the very idea of creatio ex nihilo ? Already the Isma'ilite authors apparently experienced some problems with this "innovation" of the Neoplatonica Arabica. They therefore dealt in a much more pronounced way with the very "origination" of the very first Being. Between God, whom they considered to be beyond Being, and the First Intellect, being the first created Being, they most of the time introduced a third instance, i.e., God's Word (kallma)57. Their aim in doing so was unambiguously to arrive at a better harmonization between the two — at first sight radically opposite — conceptions of creation and emanation. They wanted to avoid any possible confusion between God, and His creation. Whether they really succeeded in achieving their goal may be doubted, but deserves a study in itself. I therefore have to leave this question open. 53

LANE, Lexicon cit., p. 166. Quran, s. 2, v. 117 and s. 6, v. 101. 55 Cf. G. ENDRESS, Proclus Ambus. Zwanzig Abschnitte aus der Institutio Theologica in arabischer Ubersetzung, Beirut 1973, p. 231 (Beiruter Texte u. Studien, 10). 56 NETTON, Allah Transcendent cit., pp. 220 ff. 57 W. MADELUNG, Aspects of Is ma ill Theology, the Prophetic Claim and the God 54

IV CREATION AND EMANATION IN IBN SINA

471

Let us turn our attention now to Ibn Sma himself. From what precedes it looks possible, or even probable that he also adopted this notion in order to come to a kind of synthesis between emanation and creation. However, since he considers God not to be « beyond Being », but as being part of the realm of Being, his final theory will substantially differ from the two previous mentioned ones. To grasp it as accurately as possible, I will now elaborate on some of Ibn Sma's most significant statements regarding the very notion of ibda . A very basic definition is given by Ibn Sma in the Book of Definitions. It runs as follows : « ibda is an equivocal word having two meanings : 1. the existentialization (ta'yls) of a thing, not from another thing, nor by mediation of another thing ; 2. its second meaning is that a thing possesses absolute Being from a cause, without mediation, and to its essence belongs that it is not existent, but what from its essence belongs to it is in a perfect manner removed » 58 .

This twofold definition opens some interesting perspectives, although due to its brevity it is so dense that it is difficult to determine its full significance. However, there are some certainties. E.g., it cannot be doubted that in the first definition mention is made of a totally unmediated origination. In a at first sight even superfluous way it is even added that the by ibda realized thing is not from another, verbatim even : a thing. As such, the referent can be only one thing, i.e., the first created Being, it is the First Intelligence. Moreover, the absence of mediation seems to constitute a central item to the concept of ibda , since it is mentioned in both definitions. Finally, in the second definition, the possible, read : contingent character of the Being produced by ibda is explicitly stressed. It is a common Avicennian doctrine that the possible Being in itself inclines rather to non-Being than to Being, and that every created Being is a contingent Being, what is best exemplified by his radical essence-existence distinction. But is there a real difference between the two given definitions ? Certainly, there are some different accents, but taken very literally the First Intelligence, and nothing else, seems Beyond Being, in S. H. NASR (ed.), Ismalll Contributions to Islamic Culture, Tehran 1977, pp. 51-65. 58 IBN SINA, K. al-hudud, pp. 42-43 (Ar.), p. 60 (Fr.). It has to be noted that I read with 4 manuscripts ta'yls instead of ta'sls, especially in view of the close similarity with Al-Sifa'. Al-Ilahiyyat, p. 266.12-13.

IV 472

to function as their ultimate referent. How can in such a case any equivocate be involved ? When the requirement of the absence of any mediation is taken in its strictest sense, this is not the case. Therefore, it looks highly probable that Ibn Sina himself understood it in a somewhat broader sense in the second definition, i.e., as the absence of any mediating "matter". In this latter case, not only the First Intelligence, but also all the other Higher Intelligences would function as a proper object of an origination by ibda . If this interpretation is correct, the second definition appears as an "extension", and, at once, a weaker version of the first. The idea that "without" mediation may eventually signify "without the mediation of matter" receives some substantiation in the Division of the Sciences, where the fourth part of the science of metaphysics is explained as the study of « the first, spiritual substances, who are the Beings created by ibda' »59. The first, spiritual substances are undoubtedly synonymous with the Higher Intelligences. But a definite confirmation, which removes any possible doubt, can be obtained from a passage in the Commentary on the Theologia Aristotelis, which affirms : « That which is produced by ibda is, generally speaking, this Being, the existence of which is from another Being, and from itself it belongs to it not to have existence. Furthermore, (all) this is not by mediation of (any) matter, in which the Being of this (by ibda } realized thing is potentially present »60.

Here, it is clearly affirmed that the Being originated by ibda' may have come into existence by mediation of another thing, as long as no matter is involved. However, at the same time, one gets more and more the impression that ibda' is no more than the mere logical completion of the three other terms of creation, i.e., being the expression for the origination of pure spiritual substances. If this is the case, Ibn Sina fundamentally adheres an emanationistic system, offering at most minor creationistic corrections, which do not go much farther than the Neoplatonica Arabica, or his Isma'ilite contemporaries, notwithstanding the introduction of innovative elements in his emanative vocabulary. But ibda possesses for Ibn Sina a much more specific sense. In a very similar way as in his Book on Definitions, Ibn Sina also in his Commentary 59

IBN SINA, Fl aqsdm al-'ulum al-'aqliyat, in Tis' rasailfl 'l-hikma wa 'l-tabl'iyydt Cairo 1326 H., p. 113. 60 IBN SINA, Sarh cit., p. 59. With VAJDA, Les notes cit., p. 383, I read bi-tawassut instead of yatawassut.

IV CREATION AND EMANATION IN IBN SINA

473

on the Theologia Aristotelis offers a double explanation for it, i.e., a general one, which was discussed in the previous paragraph, and a more particular one. He expresses this latter as follows : « What is originated by ibda , in its particular sense, is such a Being as this (Being) 61 . It (litt. which) has at its very inception such a Being, without any mediation whatsoever. (Moreover), it has its existence from the One who supplies existence to it, insofar as non-Being had previously no mastery over it 62 . Rather, the Creator (al-mubdl'), grants in an absolute manner Being instead of non-Being to it, without having opposed to its non-Being — after it would have had power 63 — an impediment. Ibda is the relation between Creator and created Being (nisbat al-mubdf ila 'l-mubda], as far as concerns this Being » 64 .

In spite of some difficulties of interpretation, the present citation shows unambiguously that an absolute nothingness precedes the act of originating by ibda . Before its coming-into-existence, what is originated by ibda has no ontological reality at all. It is a pure nothingness, and assessing no reality it has no power as well. For its realization, no opposition has to be overcome. As long as it is a pure potentiality, there is no real Being involved. As such, it has not yet any identity : « The true ibda (requires) that no possibility has power over the existence of a thing, (in such a way) that (the possibility), while accompanying the non-Being, would precede the existence according to an anteriority, which, at the same time, had not to be accompanied by its posteriority » 65 .

For Ibn Sina it is clear that the potentiality of a created Being is not a state preceding its existence, but the potential character of such a Being is only revealed at the very moment it exists. Therefore, what is truly created by ibda is preceded by nothing, except God :

61

The reference here is to the Being originated by ibda in its general sense. Such a previous mastery of non-Being is in fact inconceivable, since fundamentally impossible. 63 This constitutes in Ibn Sina's eyes an absurd hypothesis! 64 lBN SINA, Sarh cit., pp. 59-60. The sentence : « The Creator grants... » is the result of a translation ad sensum, since the actual Arabic version offers in its very literality no sense. What has been put into brackets are additions by us, which intend to render the text more intelligible. 65 Ibid., p. 64.12-13. 62

IV 474 « As to ibda' , it means that a thing has its existence from the First alone » 66 . This surely does not mean that it, in its potential state, already partakes of the divine essence. As such it would constitute more than a threat to the divine unity. And for this very same reason, it seems that Ibn Sina rejects the idea of an origination by ibda in time. In the latter case, there is no longer any guarantee that the anteriority of the potential will at once be accompanied by the posteriority of the actual existent — anteriority and posteriority having here an outspoken non-temporal significance. If ibda involves that nothing precedes the coming-into-existence of a thing, it cannot but be synonymous with creatio ex nihilo. In the religious mind, this idea of creation implied always, or almost always a temporal beginning, although from the pure philosophical point of view this temporal dimension may be dismissed. Ibn Sina in any case did not hesitate to do so. But did he adhere the genuine idea of creation ? The previous lines gave a very positive indication in that direction. Nevertheless, they may not satisfy the critical reader. He may — rightly ! — expect more explicit confirmation. I believe to have found it, where Ibn Sina states : « This (i.e., the being together of the real cause and its effect) having been established, (we may affirm) that if one among the things in itself is cause of another thing for (some) duration, it always will be a cause for it, as long as its essence continues to be existent. If it is eternal in existence, its effect will be eternal in existence (as well). Something similar among the causes (i.e., God) is prior (or : worthier) in its causality, because it prevents the caused thing from having absolute non-Being. Hence, this cause is the one which bestows upon the things perfect existence. And this is the intention (ma no), which the sages call ibda , i.e., the existentialization of a thing after an absolute non-Being (tay'ls al-say bad lays mutlaq). In fact, to the caused in itself it belongs to be a non-Being, but (wa) out of its cause it belongs to it to be a Being. What belongs to a thing is essentially, not temporally prior to the mind than what belongs to it out of something else. Therefore, any caused thing is a Being after a non-Being by way of an essential posteriority » 67 .

66 67

Ibid., p. 64.5-6.

IBN SINA, Al-Sifa. Al-Ilahiyyat cit., p. 266.12-15.

IV CREATION AND EMANATION IN IBN SINA

475

In this passage, two ideas are predominant, i.e., the one that an effect cannot exist without its cause, and the other that existentialization by ibda implies that before the existence of the originated Being there was absolutely nothing. In both cases, the radical contingency of all caused, and thus all created Beings is eminently stressed. On the contrary, in God there is no contingency at all. Therefore He is not just a cause among the causes, but something "similar" among the causes. He, and He alone, is the only "cause", who replaces an absolute nonBeing by Being. He really creates something out of nothing, while creatures, who originate further creatures, never confront an absolute, but just a relative non-Being. It has also to be stressed that in the above passage Ibn Sina makes use of a rather ancient terminology, where he expresses himself in terms of tay'ls, aysa and laysa in order to evoke the notions of existentialization, Being and non-Being. As far as I can see, this terminology completely failed in al-Farabi, and was almost out of use in Ibn SIna's time. Therefore, I believe that he consciously re-introduced this older terminology, which was frequent in al-Kindi, and his circle68. Now, al-Kindi explicitly defended the idea of a creatio ex nihilo (in tempore). As to al-Farabi, he clearly had no such an idea. Is it therefore a too far-fetched hypothesis to suppose that Ibn Sina, willing to restore, at least to some extent, the genuine idea of creation falls back on al-Kindi ? However, perhaps inspired by al-Farabi, he may have been fully aware of the philosophical difficulties involved in the idea of a temporal creation. Hence, he seriously modified al-Kindfs conception of creation, not only by eliminating the dimension of "in tempore", but also by limiting the very act of genuine creation to the origination of the first creature, at least taken in its strictest sense. However, in this very first creative act the gift of all further Being is already implied, and in that sense God appears to be the Creator of the complete Universe, and of all Being. But ostensibly we have to deal here with a kind of "mediated" creation. And is such an idea not in direct contradiction with the divine omnipotence? Ibn Sina seems to have known this problem. He tries to avoid it by calling God the "Causator" of the causes (musabbib al-asbdb). This term, being derived from mystical sources, signifies that God exercises His causal action by lending the intermediary causes their causal function, i.e., all causal activity has its ultimate source in God69. Whereas in Aristotle God is above all the 68

ENDRESS, Proclus Ambus cit., pp. 104-105. See my Ibn Sind's Ideas of Ultimate Realities. Neoplatonism and the Qur'an as Problem-Solving Paradigms in the Avicennian System, « Ultimate Reality and Meaning », 69

IV 476

final cause of the creation, in Ibn Sma He is presented in the first place as the efficient cause70. Moreover, one may detect here a combination of the idea of vertical causality, as offered by the (Arabic) Proclus, at the one hand, and of the Islamic mystical idea of God as the ultimately unique cause, on the other, which however permeates the totality of creation by enabling the "horizontal" causality of the secondary causes. I think I may conclude that Ibn Sma, by using the very notion of ibdci , wants to express a genuine theory of creatio ex nihilo. God has created out of nothing, and He did so in a conscious and voluntary way. Between Him and his creatures exists not just a logical analogy of Being, but a transcendental analogy of Being. III. GENERAL CONCLUSION Out of the preceding analysis, it becomes evident that Ibn Sma is willing to combine a theory of creation out of nothing with an emanative scheme. In doing so, he undoubtedly has imposed upon himself a very demanding and risky task. Therefore, one may wonder whether he has really succeeded in offering a satisfying solution for this delicate problem. Although I do recognize that his theory has some appealing force in itself, it is in my view not totally free of tensions. Especially two of them appear to be of such a fundamental nature that they cannot be ignored. I now will briefly discuss them. The first tension arises out of Ibn Sma's identification in God between thinking and creating. Hereby a strict determinism seems unavoidable, and there is no real space left for any act of choice — in the sense, e.g., that the actual world would be the best of all possible worlds71. Ibn Sma was perhaps aware of this difficulty. He accentuates the notion of the divine action , but, above all, he seems to conceive of the divine freedom as an absolute freedom, i.e., a freedom without choice. However, the question then arises whether such a freedom is 10, 1987, pp. 265-266. As to its probable mystical origin, cf. R. FRANK, Creation and the Cosmic Sy stem: Al-Ghazall and Avicenna, Heidelberg 1992, p. 39, n. 64 (Abh. Heidelb. Ak. der Wiss., Philos.-hist. Klasse). 70 J. JOLIVET, La repartition des causes chez Aristote et Avicenne : le sens d'un deplacement, in J. JOLIVET, Z. KALUZA, A. DE LIBERA (eds.), Lectionum Varietates. Hommage a Paul Vignaux (1904-1987), Paris 1991, p. 53. 71 Al-Gazali seems to have adopted this latter thesis. See E. M. ORMSBY, Theodicy in Islamic Thought. The Dispute overal-Ghazall's "Best of all Possible Worlds", Princeton

IV CREATION AND EMANATION IN IBN SINA

477

based on a real autonomy ? And in this respect, Ibn Sma remains silent, and, as far as I can see, there is no real solution. The second tension is linked with Ibn Sina's contention that the created Beings are necessary, i.e., necessary with respect to their cause. If every « possible in itself »is« necessary by something else », is any real contingency then still possible ? And does this not become even more questionable, when, in a somewhat (Neo-Platonic) fashion, a tendency to non-Being is ascribed to the possible in itself ? Surely, an opposite tendency manifests itself, where Ibn Sma clearly suggests that the existentializing cause in no way nullifies the possibility, i.e., the contingency of the possible in itself. But I cannot see how these two opposite tendencies can be combined in a perfectly coherent manner. Was Ibn Sina's theory of creation then a complete failure ? This would be a far too harsh criticism. His theory offers many valuable insights, and is especially interesting as far as its basic project itself is concerned, i.e., to really combine both theories of emanation and creation, which are each in themselves open to serious challenges. Or to put it in Goodman's words : Ibn Sma tried to harmonize a metaphysics of being as eternally given with a metaphysics of radically contingent Being 72 . As such, it appears to be one among the parts of the major synthesis between Aristotelianism and Neoplatonism, which certainly figures as one of the major aims of Ibn Sina's metaphysical project73. Underlying all this, one moreover detects a clear tendency in Ibn Sma to translate into a genuine philosophical theory the data of religion, i.e., in his case, Islam 74 . It is therefore no wonder that, notwithstanding the actual presence of some inconsistencies, Ibn Sina's philosophy, in its various aspects, remained, and perhaps still remains, really appealing both in the East and the West. As to his doctrine of creation, I believe it is still worthy of attention, not at least because it invites us to avoid all too simplistic solutions. 1984, passim. 72 L. E. GOODMAN, Avicenna, London 1992, p. 61 (Arabic Thought and Culture). 73 This fact is well observed and exposed by A. HASNAOUI, Aspects de la synthese avicennienne, in Penser avec Aristote, Toulouse 1991, pp. 227-244. 74 In my view this is a very basic fact, which has to be taken seriously into account when one is dealing with Ibn Sina's thought. Unfortunately, till now it has in general not received the attention it deserves, as in an exemplary way may be shown by the recent — in many other respects however outstanding ! — study of D. GUTAS, Avicenna and the Aristotelian Tradition. Introduction to Reading Avicenna s Philosophical Works, Leiden 1988 (I.P.T.S., 4), where almost no attention is paid to this aspect of Ibn Sina's philosophy.

V

Ibn Sma (Avicenne): un pro jet ,,religieux" de philosophic?

Avicenne, dans PIsagoge du Shifa1, definit le but de la philosophic comme suit: ,,atteindre les essences-verites de toutes les choses selon que c'est humainement possible". Cette definition est tres proche de celle formulee par alKindi au debut de sa ,,Premiere Philosophic", a savoir que la philosophic est ,,la science des choses dans leurs essences-verites en tant que ce soit possible pour Fhomme"2. La source immediate de cette definition se trouve dans la tradition alexandrine3. Mais en plus de cet heritage philosophique, un souci d'ordre religieux semble y constituer la base. Al-Kindi, en effet, ne laisse pas planer le moindre doute sur le fait que la philosophic est une entreprise foncierement humaine, qui ne met nullement en cause des dogmes aussi fondamentaux que p. ex. celui de la creation ex nihilo. Chez Avicenne par contre, un tel souci ne parait pas etre present, p. ex. il n'hesite pas a presenter un systeme hautement emanatif4. On peut done se demander si Avicenne continue a agreer avec al-Kindi quand il s'agit de fixer les limites du pro jet philosophique lui-meme? Notons en premier lieu que la definition avicennienne de la philosophic, que nous avons citee plus haut, a indeniablement les memes origines alexandrines que celle d'al-Kindi. Cela ne surprend guere, car dans son ,,Autobiographie" Avicenne expose comment il a progresse dans Petude des sciences en accord avec la classification en vigueur dans la-dite tradition alexandrine. II conclut: ,J'ai continue ainsi jusqu'a ce que toutes les sciences philosophiques furent fermement etablies en moi, et que je les eusse comprises tant que ce soit humainement possible"5. Quand on sait que les sciences philo1 2

3 4

5

Ibn Sina, Shifa, al-Madkhal. Le Caire 1952, 12. Al-Kindi, Fi al-falsafat al-ula, in: Abu Rida, Rasa'il al-Kindi al-falsafiyyat, Le Caire 19501953, I, 97 — 162, p. 97,1. 8 — 9. Le rapprochement de la definition avicennienne avec celle d'al-Kindi a deja ete mentionnee par M. E. Marmura, Avicenna on the Division of the Sciences in the Isagoge of His Shifa, in: Journal for the History of Arabic Science, 4 (80), 239 — 251, 242, ainsi que par 1'equipe du CNRS de Paris dans le volume: Al-Kindi, Cinq epitres. Paris 1976, 58. A. Ivry, Al-Kindi's Metaphysics, Albany 1974, 116-118. Mais ailleurs, j'ai demontre la complexite de la position avicennienne. Cf. J. Janssens, Creation and Emanation in Ibn Sina, in: Documenti e Studi sulla Tradizione Filosofica Medievale 8 (97), 457-477. W. Gohlman, The Life of Ibn Sma, New York 1974, 30 — 31. Concernant 1'importance de la tradition alexandrine pour Avicenne, cf. D. Gutas, Avicenna and the Aristotelian Tradition,

V 864

sophiques ont pour objet de saisir les essences-verites, on retrouve aisement les deux composantes de la formule initiale. Mais regardons comment Avicenne explicite davantage cette definition de base. II observe d'abord 1'existence de deux sortes d'etres: celles soumises et celles non soumises a notre action et a notre volonte. Ce qui le mene tout de suite a Introduction d'une distinction capitale, a savoir entre philosophic pratique et philosophic theorique, dont il specific enfin le but comme la connaissance (ma'rifa) du bien, respectivement du vrai6. La double division de la philosophic en theorique et pratique nous renvoie une fois de plus a la tradition alexandrine. II ne surprendra personne qu'elle soit aussi presente chez al-Kindi. Mais celui-ci definit le but de la philosophic comme ,,1'atteinte de la verite dans sa connaissance theorique film), et le fait d'agir veritablement dans son action"7. On cherche en vain chez al-Kindi ce qui semble constituer le mot-cle de la specification avicennienne, a savoir le mot de connaissance: ma'rifa. Mais ce qui saute le plus aux yeux, c'est qu'il, contrairement a Avicenne, identifie le but de la philosophic pratique en termes d'agir, en parfait accord d'ailleurs avec Aristote. Quant a Avicenne, il ne prend done pas seulement une certaine distance par rapport a son celebre predecesseur arabe, mais aussi par rapport au Premier Maitre8. Mais quelle en est la portee exacte et quel en est le sens? L',,Epitre sur les divisions des sciences"9 peut nous apporter un premier eclaircissement a ce sujet. Avicenne y utilise hikma au lieu de falsafa, mais tout indique que ces deux termes figurent plus ou moins comme synonymes10. De cette hikma, ,,science", il est specific qu'elle est un art speculatif11. Comme dans Plsagoge, Avicenna adopte ici aussi une interpretation rationa-

6

7 8 9

10

11

Leiden 1988, 157 (I. P. T S., IV) et Fr. Hudry (ed), Liber viginti quattuor philosophorum, Turnhout 1997, XV (Corp. Christ., Cont. Med., CXLIII A). Madkhal, 12-14. II me semble que la notion de ma'rifa, telle qu'elle se presente dans 1'actuel contexte, est utilisee par Avicenne comme indiquant une connaissance purement speculative. On retrouve le meme sens chez al-Ghazzali, cf. F. Jabre, La notion de certitude selon Ghazali, Paris 1958, 2 Beyrouth 1986, 177. Al-Kindi, op. cit., 97. D. Gutas, op. cit., passim, illustre de facon exemplaire 1'attitude critique d'Avicenne vis-a-vis d'Aristote et de la tradition philosophique. Ibn Sina, R. ft aqsam al-'ulum al-'aqliyya, in: TV Rasa'il. Le Caire 1908, 104-118. Comme une edition critique fait toujours defaut, nous nous sommes contentes d'utiliser cette edition. Pour Tensemble d'editions et de re-editions, ainsi que des traductions (globales ou partielles) de ce traite, cf. J. Janssens, An Annotated Bibliography on Ibn Sina (1970 — 1989). Leuven 1991, 45 (Ancient and Medieval Philosophy, XIII). II est peut-etre interessant de signaler qu'Alpago le traduit par ,,scientia", suivi en cela par M. Mahdi, qui donne: ,,science". J. Michot et G. C. Anawati proposent: ,,sagesse" (le sens primaire du mot), tandis que M. Mimoune n'hesite pas d'utiliser: ,,philosophie". Par respect au grand traducteur de la Renaissance qu'Alpago fut indeniablement, j'ai prefere la traduction de ,,science", ce qui ne veut nullement dire que les autres traductions soient fautives. Je lis avec Alpago nazari au lieu de nazar.

V Ibn Slna: un projet ,,religieux" de philosophic?

865

liste. Mais s'agit-il d'un rationalisme radical ou plutot modere? Dans la suite de 1'expose, nous lisons: ,,Cette ,science' est divisee en une partie theorique abstraite et en une partie pratique. La partie theorique est celle dont la fin est la realisation de la conviction certaine (i'tiqad)12 en ce qui concerne 1'etat des etres dont 1'existence ne depend pas de Faction de 1'homme. Et le but recherche n'est (autre) que la realisation d'un discernement rationnel (ra'y)13 seulement, comme la science de I'unite et 1'unicite divine, ou comme la science de 1'astronomie. La partie pratique n'est pas celle dont la fin n'est autre que la realisation de la conviction certaine en ce qui concerne les etres. Plutot, son but n'est que14 la realisation d'un discernement rationnel correct en ce qui concerne une chose, qui est realisee par une acquisition (kasb)15 de 1'homme en vue d'acquerir a partir d'elle ce qui est bien. Par consequent, le but de la partie pratique n'est pas la realisation d'un discernement rationnel seulement, mais la realisation d'un discernement rationnel en vue d'agir. Le but de la (partie) speculative est done la verite, et le but de la (partie) pratique est done le bien"16. Malgre quelques differences terminologiques avec 1'Isagoge, Avicenne ne remet nullement en question le caractere foncierement speculativo-theorique de la philosophic. Celle-ci recherche un discernement, qui est regi par une rationalite certaine. Mais il n'en conclut pas qu'elle arrive a un savoir absolu et complet. Si dans 1'Isagoge la notion de ma'rifa pourrait en principe encore 12

13

14 15 16

Alpago lit ^cognitio certct\ et est suivi en cela par Anawati et Mimoune qui retiennent: ,,connaissance certaine". Mais cette traduction force un peu le sens d'i'tiqad qui est essentiellement celui d'une croyance (ferme). A.-M. Goichon, Lexique de la langue philosophique d'Ibn Sina (Avicenne), Paris 1938, 225 propose simplement: ,,croyance certaine", ce qui risque pourtant d'etre entendu dans le sens d'une croyance de foi. M. Mahdi, sans doute afin d'eviter toutes les ambigui'tes qui entourent dans le present contexte la notion d'i'tiqad, rend toute Texpression par ,,certainty". Non sans hesitation, nous avons en fin de compte opte pour la traduction de ,,conviction certaine", qu'Anawati utilise pour rendre la meme expression quelques lignes plus loin dans 1'expose sur la partie pratique. Elle nous apparait un peu plus neutre que celle de croyance, et elle tient en meme temps compte du sens primaire du terme. Alpago lit: »scire (plus bas: sententia), seu cognitio certa". II poursuit ainsi logiquement la ligne d'interpretation, qu'il avait fait sienne dans sa traduction d'i'tiqad (v. note precedente). Comme ra'y est un terme ayant une assez large signification, les traductions d' ,,opinion (Anawati, Mahdi) ou de ,,conception" (Mimoune) sont aussi bien acceptables 1'une que 1'autre. Si j'ai pourtant opte pour celle de ,,discernement rationnel", qui a premiere vue peut bien surprendre, c'est parce que j'estime que le contexte present de par 1'utilisation de sa terminologie ne peut qu'evoquer un arriere-fond kalamique. Et la, comme les ecrits d'alGhazzali le demontrent a 1'evidence (cf. p.ex. son Munqidh. Ed. F. Jabre. Beyrouth 1969, 15 (Ar), 67 (Fr), ou il mentionne la categoric des mutakallimun qui pretendent au discernement ( ra'y) et a la speculation (nazar) ), la notion de ra'y implique toujours un point de vue rationnel (ou doctrinal). Cp. F. Jabre, op. cit., 550. Je lis avec Alpago innama (non.,. nisi} au lieu du rubbama de 1'edition actuelle. La notion de kasb a joue un grand role dans la kalam classique islamique. Avicenne y faitil allusion? Si c'est le cas, cela ne fait que confirmer la presence d'un ,,vocabulaire kalamique". Aqsam, 105.

V 866

suggerer une telle possibilite, elle parait definitivement eliminee ici par Tutilisation du terme d'i'tiqad dont on voit mal comment il pourrait referer a une science profonde et comprehensive. Meme si la raison philosophique mene vers une certitude, il semble bien qu'elle ait ses limites. Ce qui se confirme nettement au niveau pratique, ou la recherche d'un discernement ne constitue qu'un — mais pas le seul! — moyen, ouvrant le chemin a la realisation du bien. En plus, Texecution de Pacte, Pagir meme, echappe clairement a Ternprise philosophique strictement parlant. Bien sur, Faction humaine n'est pas depourvue d'une base rationnelle dont il est la tache du philosophe de la saisir dans sa fondamentalite, mais elle n'est pas sans plus reductible a cellela. Mais ne faisons-nous pas ici des extrapolations insensees? Afin de demontrer qu'il n'en est nullement ainsi, nous nous tournons vers les Mubahathat ou Avicenne, en commentant le passage de 1'Isagoge qui divise la philosophic en theorique et pratique, insiste que la science (hikma) pratique a un double sens chez les philosophies. L'une concerne une des perfections morales; 1'autre a trait a la sous-division de la philosophic en tant que discipline. A 1'interieur de cet expose, nous retrouvons plusieurs sentences qui sont tres signiflcatives par rapport a la problematique que nous sommes en train d'evoquer. Une premiere s'articule ainsi: ,,Quand ils disent: ,,Certainement, une (partie) de la philosophic est theorique, et une (autre) est pratique", ils ne pensent pas a Faction morale, car celle-ci ne fait nullement partie de la philosophic — en effet, la faculte logique (qiyasiyya) est autre que la faculte morale — mais ils signifient par cela la connaissance (ma'rifa) par rhomme des facultes morales par la voie de la logique (litt. syllogisme — qiyas) et de la pensee (fikr), (p.ex.): Combien sont-elles (a savoir, les facultes morales)? Quelle est leur quiddite?...". Et un peu plus loin Avicenne affirme: ,,11 reste done que la partie ,,derniere", qui est science, est la partie pratique. Par consequent, la philosophic est divisee en (une partie) pratique et (une partie) theorique. La philosophic n'est pas du tout d'ordre moral, mais il se peut qu'elle soit un savoir film) par rapport a la (pratique) morale". Enfin, vers la fin, notre savant remarque: ,,Quand nous disons: ,,la science pratique est une partie de la philosophic", sachez que nous signifions (alors) par elle le savoir des excellences pratiques de facon universelle, et que nous ne signifions pas par elle les excellences morales elles-memes"17. Une insistance particuliere est mise par Avicenne dans Tensemble de ces fragments sur le fait que la philosophic ne concerne ce qui est veritablement universel. II est done clair que Faction concrete morale ne fait pas partie de son domaine. Qui plus est, en morale on n'a pas de veritables elements universels, du moins universels dans le sens que la raison humaine peut les saisir par deduction ou par raisonnement discursif. Contrairement a al-Farabi, Avicenne n'accepte pas la possibilite de Texistence de premisses ethiques Ibn Sina, Mubahathat, Qom 1992, 189-192, spec. § 571, 574 et 578.

V Ibn Sina: un projet ,,religieux" de philosophic?

867

universelles et necessaires18. Est-ce dire que le philosophe n'a pas a s'occuper de cette morale au sens propre, et doit par consequent se contenter de Petablissement de ce qui semble constituer la base rationnelle de 1'agir humain? Une fois celle-ci saisie, a-t-il en main tout ce qu'il faut pour ne faire que le bien? A en croire le dernier fragment cite, cela n'est pas le cas. La connaissance des vertus n'equivaut pas a sa pratique. Quand Avicenne, dans ses 'Uyun al-hikma, nous presente sa classique tridivision de la philosophic pratique: la cite, 1'economie domestique et la morale individuelle, il conclut: ,,Le principe de ces trois (sciences) se deduit du cote de la shari'a divine, et les perfections de leurs definitions deviennent claires par la shari'a divine. Et apres cela, la faculte theorique humaine en dispose librement par le biais de la connaissance (ma'rifa) des codes pratiques ayant trait a eux (a savoir, les trois sciences), ainsi que par Papplication de ces codes dans les choses particulieres". La shari'a divine figure par consequent comme principe et comme fin de Tensemble des sciences pratiques. La raison humaine n'entre en jeu qu'en seconde instance. Elle se voit placee devant une vraie limitation, et la philosophic, en tant que son expression supreme, se voit obligee ainsi de ceder la priorite a la Revelation. La loi divine echappe en quelque sorte au philosophe, non parce qu'elle est irrationnelle ni d'ailleurs a-rationnelle, mais qu'elle est en fait supra-rationnelle. II est vrai, Avicenne ne le dit pas si explicitement. Mais ailleurs, il insiste qu'il n'y a que le prophete, le seul a avoir une perfection totale tant du point de vue theorique que pratique, qui peut guider les hommes vers la connaissance correcte, ainsi que vers la pratique juste19. Le prophete est superieur au philosophe. Est-ce a dire qu'Avicenne abandonne simplement toute attitude rationaliste? Evitons les simplifications, et voyons comment il specific la tache du prophete. En ce qui concerne la connaissance theorique le prophete ne peut reveler de facon claire et distincte qu'un nombre tres restreint de verites ,,essentielles". Afin de ne pas induire la masse des simples gens en erreur, il doit recourrir, et recourt de fait a Putilisation de symboles20. A premiere vue, les hommes de science n'ont done pas besoin de la revelation prophetique dans ce sens. Mais en ce qui concerne les directives prophetiques pour la vie pratique, il ne fait pas de doute que ceux-ci valent invariablement pour tous les hommes21. Du moins au niveau pratique, le philosophe ne peut pas se passer de la shari'a. Mais dans quel sens exactement? 18

19 20 21

Cf. M. E. Marmura, Al-Ghazali on Ethical Premises, in: The Philosophical Forum 1 (1969), 393-403; et Th.-A. Druart, Al-Farabi, Ethics, and First Intelligibles, in: Document! e Studi sulla tradizione filosofica medievale 8 (1997), 403-423, spec. 404. Ibn Sina, Tafsir s. al-A'la (s. 87), in: H. 'Asi, Al-tafsir al-Qur'ani wa '1-lugha as-sufiyya fi falsafat Ibn Sina. Beyrouth 1983, 94-103, 98-99. cf. Ibn Sina , Shifa, Ilahiyyat, 442-443. Pour une analyse plus detaillee de cette affirmation, cf. J. Janssens, Ibn Sina's Ideas of Ultimate Reality and Meaning. Neoplatonism and the Qur'an as Problem-Solving Paradigms in the Avicennian System, in: Ultimate Reality and Meaning 10 (1987), 252-271.

V 868

Ayant presente la division de la philosophic pratique apres celle de la philosophic theorique, Avicenne, dans son Isagoge, affirme: ,,La validite de tout cela n'est etablie que par la demonstration theorique et par le temoignage de la shari'a, mais sa precision (tafsil) et sa determination scion la mesure (taqdir) sont etablies par la shari'a divine"22. A notre avis, le ,,tout cela" du debut refere a rensemble de la philosophic theorique et pratique avec toutes leurs sous-divisions23. En effet, le fragment fait explicitement mention d'une demonstration theorique. Mais comme nous venons de le constater, la theorie avicennienne ne laisse guere de place pour une telle methode au niveau pratique. Par centre, la notion de ,,temoignage de la shari'a" convient parfaitement a la seule philosophic pratique. Notons toutefois qu'Avicenne ne specific pas s'il s'agit de la shari'a divine, ou de la shari'a comme synonyme du nomos philosophique. Le double registre — de theorique et de pratique — se retrouve d'ailleurs ou Avicenne distingue entre tafsil et taqdir. Le dernier de ces deux appartient indeniablement au domaine de la philosophic pratique, car il indique que Dieu a accorde a chaque etre sa juste mesure jusque dans le dernier detail24. Quant au tafsil, il est un detaillement theorique qui — de facon universelle — permet de saisir les choses les plus particulieres a partir des verites universelles25. De nouveau, on decouvre une indication tres nette de la limitation fondamentale de la raison humaine. Non seulement celle-ci ne peut pas fonctionner comme le createur independant d'un nomos, mais elle n'est pas capable non plus de saisir les particularites dans une veritable perspective universelle. Qu'il ne s'agit pas d'une interpretation ,,forcee", nous essayerons de le demontrer par une derniere citation, empruntee a l',,Epitre sur les divisions des sciences". Avicenne y affirme clairement que c'est la Revelation qui paracheve 1'intellect, et que c'est done nullement ce dernier qui vient au secours de la prophetic pour remedier a ses insufficances26: 22 23 24

25

26

Madkhal, 14. M. Marmura, art. cit., 247 y voit par centre une allusion a la seule science pratique. Le terme de taqdir est bien sur de la meme racine que celui de qadar, rendu souvent par predestination, mais impliquant fondamentalement la notion de mesure. Pour son interpretation chez Avicenne, cf. J. Janssens, The Problem of Human Freedom in Ibn Sina, in: P. Llorente et alii (eds.), Actes des Simposi Internacional de Filosofia de 1'Edat Mitjana. Vic 1996, 112-118. C'est dans le meme sens, il me semble, qu'Avicenne affirme que Dieu connait les choses particulieres ,,de fac on universelle". L'interpretation de cette affirmation est loin d'etre facile, et meriterait une etude approfondie en soi. Toutefois, le lecteur peut deja consulter les excellentes etudes de M. E. Marmura, Some aspects of Avicenna's Theory of God's Knowledge of Particulars, in: Journal of the American Oriental Society 82 (1962), 299-312, et: Divine Omniscience in Alfarabi and Avicenna, in: P. Rudavsky (ed.), Divine Omniscience and Omnipotence in Medieval Philosophy, Dordrecht 1985, 81—94. Pour une interpretation differente, cf. neanmoins J. Janssens, Avicenna: tussen neoplatonisme en islam (These de doctorat), Leuven 1984, I, 34-64. J. Michot, Les sciences physiques et metaphysiques selon la Risalah fi aqsam al-'ulum d'Avicenne. Essai de traduction critique, in: Bulletin de philosophic medievale 22 (1980), 62 — 73, 71. II est interessant de lire aussi ce que 1'auteur remarque a propos de ce passage ibid., 63.

V Ibn Slna: un projet ,,religieux" de philosophie?

869

,,La misere spirituelle, il y a pour 1'intellect un moyen de la connaitre, par le biais de la speculation, du syllogisme et du raisonnement tandis que la (misere) corporelle est authentifiee par la prophetic, laquelle est declaree authentique par 1'intellect, rendue necessaire par la demonstration: la (revelation) paracheve Pintellect. En effet, tout ce dont Pintellect n'arrive pas a etablir 1'existence ou la necessite par la demonstration mais dont il sait seulement que c'est possible, la prophetic Finforme de facon peremptoire de son existence ou de son inexistence, (prophetic) dont la veridicite est pour lui authentique. Ainsi s'acheve pour lui ce qu'il avait etc incapable de connaitre".

II est a noter que cet ensemble se situe a Pinterieur d'un expose sur les derives de la science divine, c'est-a-dire sur un plan purement ,,theorique". Cela n'empeche Avicenne pourtant pas a y indiquer Pexistence de limites a la connaissance humaine, ainsi que la necessite de la revelation prophetique pour combler precisement ces limites. Tout cela signifie-t-il qu'Avicenne rejette le rationalisme de quelque forme il soit? Surement pas, car Pexpression claire des verites theoriques reste le privilege de la seule raison philosophique. Mais cette raison n'appartient pas aux communs des mortels. Et a cause de cela la revelation est avant tout symbolique. Observons qu'Avicenne, contrairement a al-Farabi, ne defend pas tellement une approche rhetorique, mais plutot une approche poetique du texte revele. Pour lui, le langage poetique revele des verites a celui qui en connait les regies ainsi que les conventions, car le bon poete tient compte de la realite. En plus, Avicenne semble etre le seul auteur dans la tradition philosophique islarmque a avoir fourni une explication au caractere syllogistique de la poesie27. Tout indique que le Goran, en tant que texte de la Revelation, est considere par lui comme le texte ,,poetique" par excellence. D'autre part, au plan pratique, il fait peu de doute qu'il a cherche a etablir une base rationelle, en creant ce que Gutas a appele ,,the metaphysics of the rational soul"28. Car son but final fut de presenter la philosophie comme un systeme entier, coherent, et strictement rationnel, ayant comme couronnement la metaphysique. Et celle-ci n'est pour autant pas en premier lieu aetiologie, ni theologie, mais — en bon aristotelisme — etude de Tetre en tant qu'etre. En un mot: Avicenne n'abandonne pas la tradition philosophique alexandrinoaristotelicienne, mais il la revisionne d'une facon originale, tout en systematisant tous les elements en un seul systeme comprehensif. N'empeche qu'Avicenne, tout comme son predecesseur arabe al-Kindi, fut bien conscient des limites de la raison humaine 29. Meme s'il adhere a un 27

28 29

L'importance de la poetique dans la pensee logique d'Avicenne a etc mise en evidence par S. Kemal, The Poetics of Alfarabi and Avicenna, Leiden 1991 (I. P. T S., IX), ainsi que par D. Black, Logic and Aristotele's Rhetoric and Poetics in Medieval Arabic Philosophy, Leiden 1990. D. Gutas, op. cit, 254sqq. J. Jolivet, Le deploiement de la pensee philosophique dans ses rapports avec 1'Islam jusqu'a Avicenne, in: L'Islam, la philosophie et les sciences. Paris 1981, 35-58 emet (et souscrit) deja 1'hypothese qu'Avicenne suit la ligne kindienne quant a I'lnterpretation fondamentale du projet philosophique.

V 870 rationalisme plus prononce que ce dernier, ce rationalisme n'est ni exclusif ni radical, mais de toute evidence modere. La Revelation n'est en aucun cas super fetatoire: elle supplee aux exigences de la vie humaine dans tous ses aspects, et meme le philosophe ne peut que lui preter oreille. En ce qui concerne la vie pratique, il lui est largement tributaire. Et alors peut-il etonner qu'aucune oeuvre majeure de philosophic pratique d'Avicenne ne nous soit parvenue? Mais comment expliquer alors qu'il a accepte des fonctions politiques publiques? Sans doute fut-il trop realiste pour ne pas sentir que 1'idee philosophique d'un ,,etat parfait" n'est qu'une simple illusion car elle suppose une rationalite parfaite. Celle-ci n'existant qu'en Dieu, c'est Lui qui doit eclaircir 1'esprit humain en ces matieres. Alors peut-on conclure qu'Avicenne a defendu un projet ,,religieux" de philosophic? La reponse ne peut etre que positive dans la mesure ou il reconnait des limites explicites a ce projet. La philosophic ne rend en aucun moment la Revelation coranique superflue, mais elle reconnait son insufflsance par rapport a elle. Neanmoins, dans la pratique philosophique d'Avicenne cette priorite de la Revelation ne se manifeste pas toujours aussi nette. Certains en deduiront peut-etre qu'il a cache ses veritables opinions, en vue de ne pas heurter les autorites religieuses. Mais nous avons en vain chercher des indications concretes dans ce sens. II existe par contre une explication nettement plus simple. Elle consiste a constater que la modestie ne rut pas le cote fort d'Avicenne, et que par consequent il avait pleine confiance dans sa propre raison philosophique. Peut-etre a-t-il simplement oublie dans 1'exercice de sa passion que le genie est en fin de compte un etre ,,inspire", car sa genialite ne fait pas de doute.

VI

Les Ta'liqat d'Ibn Slna Essai de structuration et de datation

Dans les recherches avicenniennes, les Ta'liqdt ont jusqu'a present regu peu d'attention. II a fallu attendre 1973 pour en saluer la premiere, et toujours seule edition (du moins d'une de ses recensions) par 'A. Badawi ! . Mais, meme apres cette parution, et a quelques heureuses exceptions pres, les Ta'liqat n'ont plus guere eu de place dans les etudes sur la pensee d'Ibn Slna2. Particulierement en ce qui concerne leur structure et leur nature exactes, un examen systematique fait toujours defaut. Par la presente contribution, nous esperons pouvoir combler - ne fut-ce que partiellement - cette lacune importante.

STATUS QUESTIONIS : LES TA 'LIQAT ET LE DANESH-NAMEH Heureusement, nous ne devons pas partir d'une table entierement rase. D'une part, D. Gutas a plaide en faveur d'un rapprochement des Ta'liqdt avec les Mubdhatdt, produit d'echanges ou de discussions entre Ibn Slna et ses disciples, surtout Bahmanyar ibn Marzuban et, dans une moindre mesure, Ibn 1. ' Ibn Slna. Ta'liqdt, ed. A. Badawi, al-Qahira, 1973. Ayant eu a notre disposition les microfilms de deux manuscrits, le ms. Aya Sofia 2390 (date 521 H.) et le ms. British Library, Or. Micr. 6724, Add. 16659 (nous remercions la bibliotheque Suleimaniya et la British Library de nous les avoir procures), nous avons pu effectuer une premiere comparaison entre eux et 1'edition Badawi (le temps nous a manque pour une comparaison systematique). De ce premier examen il ressort que 1'edition Badawi est bonne, mais necessite tout de meme quelques corrections importantes (pour un exemple typique, cf. n. 10). Nous ne pouvons done que confirmer le jugement que nous avons emis dans notre An Annotated Bibliography on Ibn Sind (1970-1989). Including Arabic and Persian Publications, and Turkish and Russian References, « Ancient and Medieval Philosophy », Ser. I, XIII, Leuven, University Press, 1991, p. 25. 2. Parmi elles on peut saluer la presence des deux etudes les plus importantes sur la pensee avicennienne de la derniere decennie, j'ai nomme : J. Michot, La Destinee de Vhomme selon Avicenne. Le retour a Dieu (ma'dd) et I'imagination, « Fonds Rene Draguet» V, Leuven, Peeters, 1986 ; et D. Gutas, Avicenna and the Aristotelian Tradition. Introduction to Reading Avicenna's Philosophical Works, « Islamic Philosophy, Science and Theology » IV, Leiden, Brill, 1988.

VI

no Zayla3. La raise en relation des deux oeuvres semble se justifier par une certaine analogic de structure (surtout 1'elaboration de notes a partir de citations courtes du Shifd), par une evocation commune du nom de Bahmanyar, mais surtout par Fexistence de certains fragments identiques ou similaires4. D'autre part, nous avions nous-memes jadis demontre que les Ta liqdt contenaient ce qui semble constituer 1'original arabe d'une partie de la Metaphysique du Ddnesh-Ndmeh5. Ces fragments se trouvaient groupes (bien qu'entoures par des developpements, absents du Ddnesh-Ndmeh) au debut et a la fin du texte. Ici, une question s'impose. Les fragments que nous venons d'evoquer font-ils partie integrante de notre texte ou sont-ils des ajouts posterieurs ? Un manuscrit de Leyde contient un petit traite en arabe qui correspond litteralement au premier chapitre de la Logique de Ddnesh-Ndmeh6. Etant isole et depourvu de tout commentaire, ce traite ne fait que confirmer la probabilite de Fexistence d'un original arabe pour 1'ensemble du Ddnesh-Ndmeh. Avec les fragments contenus dans les Ta'liqdt, ledit traite a pu faire partie d'un ensemble unique. Toutefois, en ce qui concerne la presence ou 1'absence d'un lien reel entre ces fragments et le corpus des Ta 'liqdt, il n'apporte rien de decisif. L'absence d'un fragment correspondant au texte du chapitre 31 du DdneshNdmeh s'avere, en revanche, significative, voire capitale. Nous avions d'abord emis I'hypothese d'une corruption dans le texte transmis des Ta'liqdt (du moins 3. Ibn Sina, Al-Mubdhatdt, ed. M. Bidar Far, Qom, Int. Bidar Far, 1992, offre une edition de 1'ensemble des Mubdhatdt (y compris cinq epitres), se montrant ainsi plus complete que celle d' 'A. Badawi, Arista 'inda-l-'Arab, al-Qahira, 1947 ; 2e ed. Kuwait, 1978, p. 122149. Quant aux disciples d'Ibn Sina, D. Gutas, op. cit., p. 1-2, note a juste titre qu'une etude serieuse les concernant fait toujours defaut. En ce qui concerne Bahmanyar, les donnees essentielles ont ete rassemblees par H. Daiber, art. « Bahmanyar, Kia », dans : Encyclopedia Iranica, 1989, III, p. 501-503. Cf. aussi J. Michot, « Une nouvelle ceuvre du jeune Avicenne », Bulletin de philosophic medievale 34, 1992, p. 138-154, spec. p. 151-154. 4. Precisons que ce rapprochement ne concerne que les Mubdhatdt 6 (pour cette denomination, cf. ed. M. Bidar Far, p. 12-15). Le nom de Bahmanyar y est mentionne a la premiere ligne (ibid., p. 145). Bien que, dans les Ta'liqdt, celui-ci ne figure pas dans le texte de base, il est evoque deux fois dans 1'addendum, qu'une des recensions anciennes y a rallie, et qui consiste dans une table des matieres elaboree par al-Lawkari (m. 1021 ?), cf. ms. Aya Sofia 2390, f° I v et XV v. En accord avec D. Gutas, op. cit., p. 143, n. 4, nous pouvons confirmer que les remarques ay ant rapport a Bahmanyar et al-Lawkari se trouvent en fait dans ce manuscrit qui est le temoin le plus ancien de cette recension. Quant aux fragments identiques ou similaires entre les Ta'liqdt et les Mubdhatdt, nous pouvons presenter la petite liste suivante (un lexique complet des deux ouvrages faisant defaut, nous n'osons pas pretendre a l f exhaustive): T. p. 80, 24 et 27 = M. § 853-854 (Bad. § 471-472); T. p. 83, 12-14 = M. § 858 (Bad. § 476) ; T. p. 96, 7-12 = M. § 739 (Bad. § 462) ; T. p. 104, 24-105, 18 = M. § 745-746 (Bad. § 333). II est a noter que le signe = n'est pas synonyme d'identite absolue. 5. Cf. notre article « Le Ddnesh-Ndmeh d'Ibn Sina : un texte a revoir ? », Bulletin de philosophic medievale 20, 1986, p. 163-177, spec. p. 164-65. Ayant reexamine les correspondances, deux corrections s'imposent: D.N. § 34 = T. p. 20, 2-11 (avec un remaniement dans 1'ordre du texte) (au lieu de : T. p. 19, 11-20, 10); et: D.N. § 20 = T. p. 176, 12-178, 24 et p. 179,5-15(aulieude:r.p. 176, 12-177,5 ; p. 177, 19-178, 24 et p. 179,8-13). 6. Cf. notre An Annotated Bibliography, p. 17, note (au-dessous de 2). II s'agit du ms. Or. 184 de Leiden (nous remercions la bibliotheque de Leiden de nous en avoir procure" un microfilm). Un examen complet du manuscrit n'a apporte aucun nouvel element supplementaire par rapport au texte edite par M. Mohaghegh. Nous ne pouvons done que (re)affirmer 1'existence en arabe de ce premier chapitre de la Logique du Ddnesh-Ndmeh.

VI LES TA 'LI'QAT D'IBN SlNA

111

dans la redaction sur laquelle se fonde 1'edition Badawi). Ayant effectue une nouvelle analyse du texte entier, nous nous sentons obliges de renoncer en toute humilite a cette hypothese. Le corpus des Ta'liqdt n'a pour objet qu'une problematique, fondamentale : la question de la connaissance des particuliers par Dieu. Cette problematique se trouve aussi au cceur du chapitre 31 du Ddnesh-Ndmeh, qui a pour titre : « Examen de la connaissance du contingent par le sujet connaissant7 ». Mais son bref expose est developpe ici en details. Tant au niveau metaphysique que physique ou logique, 1'analyse de la problematique est plus englobante et plus precise. Dans cette perspective, 1'incorporation (un peu isolee) du chapitre 9 de la meme Metaphysique du Ddnesh-Ndmeh, qui porte sur 1'etat de 1'accident, ne peut plus vraiment surprendre : la particularity n'est-elle pas souvent intimement liee a une certaine accidentalite, au point que 1'une ne se laisse pas comprendre sans 1'autre ? Quant a la division en deux ensembles des fragments relatifs au Ddnesh-Ndmeh - Tun au debut du texte : couvrant les chapitres 3236, 1'autre a sa fin, ayant trait aux chapitres 20-30 -, elle aussi se laisse expliquer. Dans le premier ensemble, Tinvariabilite de 1'essence divine est affirmee de fa^on absolue, tant sur le plan de la connaissance que sur celui de la volonte, de la puissance, de la sagesse et de la generosite. Ainsi, le probleme de la possibilite d'un lien entre Dieu et ce qui n'est pas Lui se pose presque naturellement, mais, en meme temps, de fa$on accrue. Quant a 1'ensemble de cloture, il traite (apres 1'accidentel) du contingent; de 1'etre necessaire en soi (Dieu) - celui-ci etant un etre unique et un, inchangeable, n'ay ant d'autre essence que d'exister, et n'etant ainsi conceptualisable sous aucune categorie ; enfin, de la possibilite de 1'existence d'une relation entre 1'etre necessaire en soi et les etres contingents, ainsi que de la caracterisation specifique de cette relation, a savoir son articulation en termes de causalite eternelle, fondee sur une connaissance invariable. En reponse directe au premier ensemble, la possibilite d'un lien entre Dieu et ses creatures est affirmee ici de fa$on claire. Si cela est possible, c'est qu'entre-temps le probleme de la connaissance des particuliers par Dieu a ete - du moins cela est suppose - resolu. Les fragments relatifs au Ddnesh-Ndmeh sont done organiquement lies au corpus des Ta 'liqdt, qui, a leur tour, se presentent comme un ensemble plus ou moins coherent, centre autour d'une seule problematique. Si cette constatation n'est pas due a une erreur d'interpretation, un abime separe les Ta'liqdt des Mubdhatdt, surtout en ce qui concerne la structure de base. Ainsi les Ta liqdt ne se presentent plus comme un amalgame « arbitraire » de notes8. 7. Nous empruntons la traduction franchise de ce titre a M. Achena, Avicenne. Le Livre de science, traduit et annote par M. Achena et H. Masse, 2e ed. revue et corrigee par M. Achena, Paris, Les Belles Lettres, 1986, p. 198. 8. Nous doutons que D. Gutas, op. cit., p. 143, ait raison quand il qualifie la recension du ms. al-Qdhira, Hikma 6 M comme « a consecutive but haphasard transcription of all the notes ». Nous nous demandons meme si elle ne constitue pas la seule recension qui reflete fidelement la redaction primitive du texte ? En ce qui concerne la division du texte, presente dans tous les autres manuscrits, celle en ta'liq-s (a en juger a base du ms. Brit. Libr., Or. Micr. 6724, Add. 16659) est de toute evidence qualitativement inferieure (pla^ant par exemple une coupure au milieu d'un expose couvrant un seul chapitre du Ddnesh-Ndmeh), et done probablement posterieure a celle articulee selon le systeme a-ba-ga-d, et dont la non-

VI 112 ESSAI CRITIQUE DE STRUCTURATION

Essayons maintenant de decrire les grandes lignes de cette structure. Deux remarques preliminaries s'imposent: 1. La division du texte, telle que nous la proposerons dans la suite, est celle que nous avons personnellement choisie. Elle ne pretend nullement reproduire la division originelle (si une telle division a jamais existe !). Elle presente seulement la commodite de mettre en evidence une certaine articulation, pour ne pas dire une articulation certaine du texte. 2. Dans notre systematisation, nous ne tenons pas compte de tous les details du texte. Certains passages, d'habitude plutot brefs, font en effet probleme si on veut s'en tenir a un ensemble strictement coherent9. Neanmoins, nous ne croyons pas qu'il s'agisse d'une intervention inconsideree, dans la mesure ou les Ta'liqdt sont probablement a mettre en rapport direct avec I'enseignement oral d'Ibn Sina (voir infra), et constituent done en quelque sorte des notes de cours, qui sont rarement d'une coherence parfaite. Ay ant exprime ces deux specifications a la fois importantes et limitatives, entamons maintenant la description proprement dite de la structure. Affirmation de base : I 'invariabilite absolue de I 'essence divine Le contenu du debut du texte a ete indique plus haut: 1'invariabilite de 1'essence divine. Ajoutons seulement que les passages, qui sont incorpores ou ajoutes aux fragments ay ant une correspondance plus ou moins directe avec le Ddnesh-Ndmeh, peuvent etre qualifies de « commentaires explicatifs ». Quelques ressemblances - assez vagues, il est vrai - avec la Metaphysique du Shifd s'y font jour (p. 13-22, 18). Premier mouvement: connaissance creatrice et conceptualisation intellectuelle de I'individu designable L'auteur s'interroge ici sur la possibilite de la conceptualisation de I'individu. Plus specifiquement, il poursuit cette interrogation par rapport a la connaissance de Dieu, qui est decrite comme une connaissance a partir des causes, et done comme fondamentalement creatrice (p. 22, 19-30, 4). Apres la conceptualisation de I'individu, la possibilite de la saisie d'un concept intelligible pur se trouve au cceur des recherches. II est affirme avec originalite ne fait pas de doute, car la tradition la fait explicitement remonter a al-Lawkari. Notons en plus que - a Texception de la recension propre a une tradition iranienne manifestement tardive - 1'ordre du texte se revele sensiblement le meme dans les differentes recensions. Enfin, observons que 1'anciennete' (premiere moitie du xne siecle), ainsi que la signification bien particuliere du ms. Hikma 6 M ne font pas de doute, et que 1'edition Badawi est essentiellement bas6e sur ce manuscrit. Cf. D. Gutas, « Notes and Texts from Cairo Manuscripts, I I : Texts of Avicenna's Library in a Copy by 'Abd al-Razzaq asSignahi », Manuscripts of the Middle East 2, 1987, p. 8-17. 9. La division, proposee par al-Lawkari, couvre parfaitement tous ces passages, mais, a ce qui nous semble, elle fait perdre de vue le fil rouge qui traverse 1'ouvrage entier.

VI LES TA 'LIQAT D'IBN SlNA

113

insistance que, si Dieu, en tant qu'acte pur, a dans sa connaissance creatrice acces a un tel concept, la connaissance humaine, par contre, a cause de sa limitation fondamentale, ne peut pas saisir la veritable essence pure des choses et done, a fortiori de Dieu. II est a noter en outre que dans 1'intelligible pur, il n'y a aucune pluralite, mais que, d'autre part, il n'existe pas de veritable concept intelligible de 1'individu en tant que sensiblement designable (mahsus muSar ilayhi) (p. 30, 5- 38,4). Est-ce a dire qu'aucun individu designable ne peut faire 1'objet d'un concept intelligible ? Une telle affirmation serait excessive. Au moins la nature particuliere des individus, uniques dans leur genre, est connaissable de fagon essentielle. Quant aux autres individus qui se caracterisent par la pluralite a 1'interieur d'une meme espece, ils sont congus dans leurs individualites selon leurs etats universels, c'est-a-dire leur matiere et leur position absolues et, par consequent, non selon leurs etres sensiblement designables. Selon ces modalites precises, Dieu connait chaque chose. En vue de rendre cette conclusion plus plausible, et d'en specifier les contours exacts, 1'auteur avait auparavant souligne la finitude de ce qui existe en acte, plus specifiquement de chaque veritable cause essentielle, ainsi que de 1'ensemble des causes vraies emanant de la pensee creatrice de Dieu selon un ordre universel intelligible10 (p. 38, 5-50, 22). La connaissance et la puissance de Dieu, ipseite absolue, ne peuvent pas etre comparees de quelque fagon que ce soit avec celles de 1'homme. Cela est avant tout illustre a propos de la connaissance. Si dans 1'homme celle-ci se revele fondamentalement discursive et passive, car toujours mise en rapport avec quelque chose, en Dieu elle se presente au contraire comme foncierement instantanee et active, car source des causes. Quelques observations, ayant trait surtout au monde corruptible, accompagnent cette these fondamentale d'une difference radicale entre les attributs de Dieu et ceux de 1'homme. II y est, entre autres, question du fait que la matiere ne constitue qu'une « quasi »-cause de 1'individu ; qu'une notion universelle n'a d'existence que par ses individus ; et que 1'ame humaine, etant tout a fait immaterielle, ne se meut reellement que pour soi-meme, c'est-a-dire en vue de sa propre perfection (p. 50, 23-67, 8). Tout cela est precise davantage. Dans le monde corruptible, 1'individuation des formes se fait dans une matiere preparee. Comme c'est le cas pour tout etre cree, 1'existence n'appartient pas a 1'essence. Seul en Dieu, 1'identification entre essence et existence est valable. Dieu, agent par excellence, et, en meme temps, existence pure, n'est d'ailleurs que cause d'existence, non d'origination. D'autre part, 1'ame humaine, quant a elle, regoit ce qui est intellige de fagon accidentelle. Mais quant a son auto-intellection, celle-ci ne necessite aucune mediation instrumentale (un peu a 1'instar de la connaissance divine). Toutefois dans cette vie terrestre 1'ame ne peut atteindre la veritable connaissance intellectuelle que grace a une illumination de 1'Intellect agent. En corollaire a 10. Nous attirons Inattention du lecteur sur le fait qu'il faut lire p. 37, 25 \fi 'l-fasl au lieu de :fi 'l-fi'l (lecture confirmee par les deux manuscrits dont nous disposons).

VI 114

ceci, il est affirme que les ames non parfaites continuent a la mort a avoir besoin d'elements materiaux pour leur perfectionnement (p. 67, 9-85, 14). Un moment charniere : les categories de relation et de quantite Chaque connaissance implique une relation, tandis que chaque individu, en tant qu'entite, se refere toujours a la quantite. Rien d'etonnant done a ce que toute une sous-section de notre ouvrage soit consacree a ces deux categories (p. 88, 15-96, 6). Sans vouloir nous attarder sur cette sous-section ay ant trait a la relation et a la quantite, signalons tout de meme qu'elle semble exercer une fonction charniere entre deux parties majeures : une premiere, ou une attention particuliere est portee a la « difference », tant celle qui existe entre 1'intellectuel et le sensible, que celle qui separe Dieu de rhomme ; une seconde, ou 1'accent est plutot mis sur la « causalite », tant immediate en Dieu que mediate dans les creatures. Tout en affirmant cela, nous sommes bien conscient du fait que Fidee de difference n'est pas absente dans la deuxieme partie, comme d'ailleurs celle de causalite ne faisait pas defaut dans la premiere. Second mouvement: connaissance creatrice et causalite mediatisee A 1'examen des categories de relation et de quantite fait suite une interrogation sur 1'origine de la pluralite dans la creation. Le role intermediate des Intelligences et des spheres celestes dans 1'origination des etres sublunaires est particulierement souligne. En effet, de Dieu ne decoule directement qu'une seule chose, selon le fameux adage : « Ex uno, secundum quod est unum, non est nisi unum 11 . » Mais tandis que Dieu comprend de par son essence cette premiere creature, et, du coup, tout ce qui la suit, et ainsi particularise les choses de par son essence, la premiere Intelligence ne particularise que de par ses concomitants. Cela vaut aussi pour toutes les autres Intelligences separees et leurs spheres. Ainsi, la volonte continue des ames celestes est cause directe du mouvement circulaire des spheres celestes, celui-ci constituant la perfection premiere, mais entraine aussi comme perfection seconde la generation des etres corruptibles. A cette occasion, ce qui distingue les ames celestes des ames humaines est mis a 1'evidence : tandis que chacune des premieres forme une espece en soi, et est done de par elle-meme particuliere, il n'en va nullement de meme pour les dernieres, etant donne qu'elles appartiennent a une seule et meme espece. Mais malgre cela, notre auteur n'hesite pas a affirmer la survivance individuelle de Tame humaine (p. 96, 7-115, 17). Bien que la creation soit fondamentalement mediatisee, elle n'echappe pas, jusque dans le moindre de ses details, a la connaissance divine. De fait, la creation se realise selon un ordre universel, qui est un concomitant de 1'essence 11. H.A. Davidson, Alfarabi, Avicenna, and Averroes, on Intellect. Their cosmologies, Theories of the Active Intellect, and Theories of Human Intellect, Oxford, Oxford Univ. Press, 1992, p. 75, n. 3, remarque qu'il n'a pas trouve de trace de cette formule avant Ibn Sma, bien que son inspiration neoplatonicienne ne fasse pas de doute.

VI LES TA 'LIQAT D'IBN SINA

115

divine, et qui est done inaccessible en tant que tel a I'homme. Par consequent, cet ordre universel creationnel qui implique une fmitude reelle des formes et des etats tout en admettant 1'existence d'une infinitude de relations entre eux, ne peut etre connu que par celui qui perc,oit intensement en acte (al-darrdk alfa"dl), c'est-a-dire Dieu, qui est un intellect pur et simple. N'ayant aucune puissance en Lui, II est a la fois cause efficiente et cause finale de tout ce qui existe. En un mot, II est veritable cause, ce qui veut dire qu'Il est en tant que cause « avec » son cause - cet « avec » ne devant pas se comprendre trop litteralement dans ce cas precis (p. 115, 18-133, 6). La cause precede toujours le cause par essence et par intention. Quant a la cause de 1'individuality de 1'individu, elle n'est pas constitutive de sa quiddite. Ainsi, dans le monde sublunaire, 1'individu ne se laisse pas connaitre de par une definition, car chaque definition est fondee sur la combinaison de deux elements essentiels : le genre et la difference specifique, qui en tant que tels n'existent pas dans la realite. La saisie intellectuelle n'est en fait possible que par une approche relationnelle. Toutefois, il ne faut jamais perdre de vue la distinction radicale entre le relationnel essentiel limite et le relationnel accidentel infini, ou, exprime d'une maniere differente, entre le relationnel intellige et le relationnel existant. Pour rhomme, tout individu a une double « essence-verite »(haqlqd): comme existant en soi, et comme existant dans F intellect. Seulement dans I'autoperception humaine, il y a une identite absolue entre le percevant et le pergu. Observons aussi qu'a 1'interieur de cet ensemble 1'auteur developpe quelques remarques ay ant rapport au temps et ses differentes acceptations, ce qui le mene a rejeter categoriquement toute these reclamant 1'existence d'un monde atemporel qui serait intermediate entre Dieu et la creation (p. 133, 7149, 17). Contrairement a I'homme, Dieu, acte pur, connait toutes les choses par sagesse ('aid wagh al-hikma) selon leur ordre necessaire, toute possibilite (ou impossibilite) d'exister etant dependante de 1'etre necessaire (wugiib al-wugud) de son essence. Dieu est le « causateur des causes » (musabbib al-asbdb)12. Sa providence n'est rien d'autre que la connaissance de son essence en tant que source de toute existence. En Dieu, le connaissant, le connu et 1'acte de connaissance ne font qu'un. Quant a la connaissance humaine de Dieu, elle ne peut se realiser qu'en ay ant recours a la division et a la negation. Quant aux spheres celestes, elles ne sont capables que d'imagination. Toute existence est precedee d'une conception intellectuelle ou imaginaire (p. 149, 18-167, 13). Le savoir humain s'articule selon differentes sciences dont la methode et 1'objet sont bien specifiques pour chacune d'elles. Les sciences sont hierarchisees. La plus elevee de toutes est la science universelle, i.e. la metaphysique (p. 167, 14-174, 16).

12. Pour cette notion, cf. notre article « Ibn Sina's Ideas of Ultimate Realities. Neoplatonism and the Qur'an as Problem-Solving Paradigms », Ultimate Reality and Meaning 10, 1987, p. 252-271, spec. p. 265-266.

VI 116

L'ouvrage s'acheve sur une affirmation claire de 1'existence d'un lien reel entre Dieu et ses creatures, comme nous 1'avons indique13 (p. 174, 17-194). Conclusion Voila, ce qui a notre avis se laisse entrevoir comme trame essentielle des Ta 'liqdt. Malgre quelques simplifications - sans doute inevitables, quand on essaie de synthetiser tout un ouvrage en quelques lignes - nous avons tente dans ce qui precede de refleter aussi fidelement que possible la teneur reelle du texte original. Nous esperons avoir articule correctement la demarche que 1'auteur a entreprise lors de 1'elaboration de son ouvrage. Neanmoins, on peut se demander si nous n' avons pas impose d'une fa£on trop artificielle une structure - meme s'il n'y a pas de coherence parfaite, il s'agit toujours d'une structure - a une oeuvre dont le titre se limite a un seul substantif, mis au pluriel Ta'liqdt (« gloses »). Mais quelle est alors la portee exacte de ce titre ?

ESSAI D'IDENTIFICATION DU GENRE ET DE LA PATERNITE DE L'OUVRAGE

Notons d'abord que le terme Ta'liqdt ne figure pas comme titre proprement dit dans les deux manuscrits les plus anciens, mais qu'il s'y trouve dans la formule de cloture. Comme aucun de ces deux manuscrits n'indique la division du texte en termes de ta'liq-s, et que Tun d'eux ne comporte meme aucune division, la designation de 1'ouvrage par le terme Ta'liqdt nous semble avoir precede ladite division. Mais alors, a quoi peut-il referer ? G. Makdisi a montre que le genre Ta 'liqdt dans les etudes de droit couvre soit des notes de cours d'un etudiant, soit une combinaison de notes de cours et d'extraits d'ouvrages du maitre effectuee par un etudiant avance, soit meme une premiere publication independante d'un etudiant erudit ou extremement avance14. II note en plus que ce meme genre a etc adapte au xe siecle par le kaldm mu'tazilite, ainsi que par des grammairiens et des medecins. Par consequent, il ne fait pas mention d'une application eventuelle en philosophic. Mais est-ce a exclure a priori ? Nous ne le croyons pas, d'autant moins que notre texte revele indeniablement des caracteres essentiels du genre Ta 'liqdt, tel qu'il vient d'etre evoque. En effet, on y trouve des extraits d'un ouvrage (i.e. le Ddnesh-Ndmeh) du maitre (i.e. Ibn Sma), accompagnees de notes explicatives dont le style indique plutot une base orale. Mais on y trouve aussi des paraphrases de parties d'autres ouvrages avicenniens, tels que les Categories, le De anima et la Metaphysique du Shifd, ainsi que des references implicites, mais evidentes a ces memes ouvrages15. Une fois de plus, nous y trouvons ce que 13. Cf. supra, p. 4. 14. G. Makdisi, The Rise of Colleges. Institutions of Learning in Islam and in the West, Edinburgh, 1981, p. 114-125. 15. L'enumeration de"taillee de toutes ces references depasse largement le cadre de cette etude. Nous comptons en publier le bilan complet plus tard. Remarquons toutefois que, si les

VI LES TA 'LlQAT D'IBN SlNA

117

nous oserions appeler des additions explicatives. En outre, le style - la presence de nombreuses repetitions, tant dans les idees que dans 1'utilisation de certains termes techniques, de digressions, ou de simples intercalations - fait penser a un contexte de cours oral. Enfin, 1'idee que notre ouvrage est Ii6 a une communication au vif est renforcee par la constatation qu'on y trouve de temps a autre des elements de ce qui semble constituer une polemique vivante avec le kaldm mu'tazilite, a premiere vue celle de 1'ecole d"Abd al-Gabbar, contemporaine d'Ibn Sina16. En voici une breve enumeration : impossibilite de connaitre Dieu a partir des creatures (p. 37); rejet de toute analogic entre les attributs divins et les attributs humains (p. 52); rejet de 1'idee que Dieu puisse agir en raison de quelque chose d'externe (p. 53) ; negation que chaque intention (ma'no) est precedee d'un etat de non-etre ('adam) (p. 54); affirmation que la volonte de Dieu peut etre considered comme un accident ('arad) (p. 61); refus de considerer Dieu comme cause d' origination (p. 84 ; p. 131); impossibilite de 1'existence d'un infini en acte (p. 125); enfin, rejet categorique de 1'existence d'un monde intermediate entre Dieu et la creation (p. 138). Dans tous ces cas, il s'agit d'une opposition nette - parfois tres explicite, parfois plutot implicite - a des theses mu'tazilites. Mais ce qui nous interesse davantage ici, c'est la concision avec laquelle ces antitheses sont developpees. En meme temps, on ne peut etre frappe par 1'absence d'une tentative de systematisation de cette critique anti-mu*tazilite. Ces constatations se laissent pourtant aisement comprendre si on suppose a la base un enseignement oral. N'est-il pas normal de faire, dans un cadre pareil, de temps a autre des remarques visant les opposants les plus importants du moment ? Nous concedons volontiers de ne pas avoir livre la preuve definitive que les Ta'llqdt sont Tissue directe de notes de cours. Mais, etant donne les elements signales, cela nous parait probable. Mais qui alors en fut 1'auteur ? Spontanement, on songe a un des disciples intimes d'Ibn Sina, et plus specifiquement a Bahmanyar, d'autant plus qu'une tres ancienne tradition 1'indique explicitement comme rapporteur de ces gloses17. Rappelons que Bahmanyar fut un des disciples les plus importants references aux Categories et a la Metaphysique sont nombreuses, celles au De anima s'averent extremement rares. Observons aussi que les citations litterales, introduites par : raw qawlihi, referent toujours a la Metaphysique, du moins a premiere vue. 16. Nous remercions H. Daiber pour les eclaircissements qu'il nous a procures a propos du caractere anti-mu'tazilite de certaines de ces doctrines. Des elements importants concernant la position mu'tazilite par rapport aux idees evoquees ici se laissent trouver dans son Das theologisch-philosophische System des Mu'ammar Ibn 'Abbdd as-Sulami, « Beiruter Texte und Studien », XIX, Beirut, Wiesbaden, 1975. Notons que certaines formules des Ta 'liqdt sont tres proches du Sarh al-usul al-hamsa de Mankdim Sheshdiv, un des disciples d' 'Abd al-Gabbar. Donnons-en un exemple : T. p. 140, 10-11 est presque identique a 'Abd al-Karim al-Utman (ed.), 'Abd al-Gabbar (sic !). Sarh al-usul al-hamsa, al-Qahira, 1965, p. 95, 10-12. Toutefois, il n'est pas sans interet de voir qu'Ibn Sina, malgre son attitude antimu'tazilite (et anti-kaldm en general) a quand meme subi des influences importantes de ce meme mu'tazilisme (et de ce meme kaldm). Voir l'6tude fondamentale de J. Jolivet, « Aux origines de 1'ontologie d'Ibn Sina », dans J. Jolivet et R. Rashed (ed.), Etudes sur Avicenne, Paris, Les Belles Lettres, 1984, p. 11-28. 17. Cf. supra, n. 3.

VI 118 d'Ibn Slna, et qu'il fut probablement plus jeune que lui 18 . Quant a son oeuvre majeure, le K. al-Tahsildt, elle se base comme les Ta 'liqdt largement sur les Categories et la Metaphysique du Shifd (mais on y trouve en plus des references a laNagdt et aux Ishdrdi), et s'inspire directement du Ddnesh-Ndmeh, du moins pour sa structuration19. Toutefois les ressemblances avec les Ta'llqdt s'averent superficielles - trop vagues en tout cas pour qu'on puisse affirmer qu'un seul et meme auteur ait redige les deux ouvrages. A defaut d'autres indications, nous devons done nous contenter de presenter Bahmanyar comme 1'auteur « possible » de ces notes. Observons que si les Ta 'liqdt sont en fait des notes de cours, ce probleme d'auteur ne revele pas une grande importance - le « veritable » auteur etant plutot 1'enseignant, i.e. Ibn Sina.

ESSAI CRITIQUE DE DATATION

II nous reste encore une derniere interrogation fondamentale : a quelle epoque de la vie d'Ibn Slna peut-on situer ces « gloses » ? Ou bien furent-elles redigees apres sa mort ? Cette derniere alternative nous parait peu probable parce que le nom d'Ibn Sina n'est nulle part evoque, ni au debut, ni a la fin du texte. L'absence d'une reference quelconque au venerable maitre apres sa mort constituerait un phenomene plutot exceptionnel. II semble done justifie de penser que leur redaction eut lieu du vivant d'Ibn Slna. Mais est-il alors possible de preciser davantage la date de cette redaction ? II nous semble que oui. Facteurs externes et internes II n'est guere douteux qu'Ibn Sma avait fini la redaction definitive des Categories, du De anima et de la Metaphysique du Shifd, en raison des multiples references qui y renvoient et surtout des additions explicatives les accompagnant (cf. supra). Quant au Ddnesh-Ndmeh, notre texte ne permet pas de conclure que sa version persane etait totalement achevee, mais demontre

18. La date la plus probable de la mort de Bahmanyar est 1066, done une trentaine d'annees apres celle d'Ibn Sina. Qu'il fut plus jeune qu'Ibn Sina semble d'ailleurs etre confirme par un passage des Mubdhatdt 3 - cf. ed. M. Bidar Far, p. 75, 5-7 (J. Michot a eu la gentillesse de mettre a notre disposition la version exploratoire de sa traduction integrate en frangais de ces Mubdhatdt 3, la publication definitive etant prevue dans Le Museon, 1997). 19. Bahmanyar Ibn al-Marzuban, Al-Tahsil, ed. M. Mutahhiri, Tehran, 1979. Dans 1'introduction (ibid., p. 1) 1'auteur lui-meme affirme explicitement qu'il s'est inspire pour la structuration de son livre de 1'ordre du Ddnesh-Ndmeh. Notons toutefois qu'il designe la metaphysique non comme science divine, 'ilm ildhi, mais comme meta-physique, 7/m md ba 'd at-tdbi'a, la theologie en etant exclue. Celle-ci est en fait placee au debut de la troisieme partie et y est suivie de la physique, qui comprend la psychologic, dans laquelle la doctrine de la resurrection de 1'ame humaine est incorporee. Cette derniere partie ne s'intitule d'ailleurs plus physique, al-'ilm at-tabi'a, mais science des etats des plus nobles parmi les etres, al-'ilm bi-ahwdl a'ydn al-mawguddt. La conception de la metaphysique de Bahmanyar ne recoupe done pas sans plus celle d'Ibn Sina.

VI LES TA 'LI'QAT D'lBN SlNA

119

suffisamment que « la version originale arabe », du moins la partie metaphysique, avait etc terminee bien anterieurement20. Precisons qu'al-Guzgani nous informe qu'Ibn Sma dedia le Ddnesh-Ndmeh a 'Ala' al-Dawla, et qu'Ibn Sina ecrivit cet ouvrage a Isfahan21. A la lumiere de toutes ces donnees, une datation des Ta'liqdt pendant le sejour d'Ibn Sina a Ramadan, a savoir 1015-1024, s'avere difficile, sans etre entierement exclue22. Si ces facteurs externes ne permettent de formuler qu'un doute raisonnable a ce propos, il en va tout autrement quand on y mele des considerations internes d'ordre doctrinal. En effet, les Ta'liqdt contiennent plusieurs elements doctrinaux, dont la teneur est a premiere vue non avicennienne, ou, en tout cas, peu avicennienne. Citons les trbis plus importants : la designation explicite de Dieu comme « acte pur » ; 1'identification de 1'Intellect agent avec le premier etre emane ; 1'utilisation de la notion de « science universelle » comme appellation pour la metaphysique 23 . Pour autant que nous le sachions, aucune des trois idees ne se trouve en cette formulation specifique dans une des grandes sommes philosophiques avicenniennes. On peut done se demander si elles ne sont pas propres au « redacteur » de notre texte. Contrairement aux apparences, rien n'est moins sur. Pour les trois elements doctrinaux, on peut en effet argumenter en faveur d'une paternite d'Ibn Sina lui-meme. Premier element doctrinal: la science universelle En ce qui concerne la notion de science universelle (al- 'ilm al-kulli), cela est hors de discussion. Ibn Sina Futilise dans son Mantiq al-MaSriqiyyin, ainsi que dans ses Ahwdl al-nafs24. Sans doute avait-il emprunte cette notion au petit traite d'al-Farabi, dont il fait tant la louange dans son autobiographie, et qui s'intitule : Sur les buts de la Metaphysique25. Comme chez al-Farabi, la science universelle est definie dans ces deux ouvrages, comme d'ailleurs dans les Ta'liqdt, comme 1'etude de ce qui est commun a toutes les sciences particulieres, et done a tous les etres. Mais dans les Ta 'liqdt, il y est explicitement ajoute que la theologie fait partie integrante de cette science : ne reprenant ainsi qu'une affirmation deja bien presente dans le texte farabien.

20. Cf. supra, p. 2. 21. W. Gohlman, The Life of Ibn Sina. A critical edition and annotated translation, New York, SUNY, 1974, p. 68-69. 22. 'A. Badawi, Ibn Sina. Ta'liqdt, p. 6. 23. Cf. Ta'liqdt, p. 32, 29-33, 1 ; p. 150, 8, 14, 26 ; p. 151, 7, 10, 14 ; p. 153, 11-12 (pour : acte pur} ; p. 100, 21 ; p. 165, 13 (pour I'Intellect Agent comme premier emane); p. 41, 3 ; p. 129, 5, 12 ; p. 168, 27 ; p. 169 (6 fois) ; p. 171 (trois fois) (pour science universelle}. 24. Cf. respectivement Ibn Sina, Mantiq al-Masriqiyyin, ed. M. al-Hatib et 'A. al-Qatla, al-Qahira, 1910, p. 17 ; et A.F. Ahwani (ed.), Ahwdl al-nafs. R.fi al-nafs wa baqd'ihd wa ma 'ddihd lil-sayh al-ra 'is Ibn Sina, p. 114. 25. al-Farabi, Fi agrdd al-hdkimfi kull maqdla rain al-kitdb al-mawsum bil-huruf, dans Fr. Dieterici, Alfarabi'sphilosophische Abhandlungen. Aus Londoner, Leidener und Berliner Handschriften herausgegeben, Leiden, 1890, p. 34-38. La notion de science universelle y est mentionnee trois fois a la page 35. En ce qui concerne la signification particuliere que ce traite a eu pour Ibn Sma, cf. D. Gutas, op. cit., p. 238-254.

VI 120

Pour notre propos, il faut surtout constater que les deux ouvrages d'Ibn Sina, que nous venons de mentionner, et qui, pour autant que nous avons pu le verifier, sont les seuls ou cette notion de science universelle est reprise, semblent appartenir a 1'ceuvre relativement tardive d'Ibn Sina. Le Mantiq fait probablement partie du grand ouvrage sur les « Orientaux », qui selon les renseignements d'al-Guzgani a ete perdu tres tot - a 1'exception probablement de cette section logique. Si ce rapprochement est correct, on peut situer la redaction du Mantiq aux environs de 1027-102926. Quant aux Ahwdl, J. Michot a livre une argumentation solide en faveur d'une datation posterieure a la Nagdt, ce qui nous ramene a la meme periode de 1027 a 1029 ou, eventuellement, un peu plus tard27. Les Ta'liqdt pourraient done se situer a la meme epoque. Mais examinons d'abord les deux autres doctrines « particulieres ». Second element doctrinal: Dieu comme acte pur Sauf erreur de notre part, Ibn Sina ne decrit jamais Dieu en des termes precis d'acte pur (al-fil al-mahd). En tant que telle, cette notion ne fut pas courante dans lafalsafa. Par exemple, Ibn Rusd, dans son Grand Commentaire sur la Metaphysique, ne la mentionne apparemment qu'une seule fois28. Cela est peut-etre moins surprenant qu'il ne le parait a premiere vue. Car, meme si T inspiration hautement aristotelicienne de cette notion ne fait pas de doute, le Stagirite lui-meme, a en croire VIndex Aristotelicus de H. Bonitz, n'en fait pas mention. Pour autant que nous avons pu le determiner, Fexpression auto to energein kathoron apparait pour la premiere fois dans un commentaire neoplatonicien sur le Parmenide, attribue par P. Hadot a Porphyre, ou a un membre de son ecole29. Une des occurrences les plus anciennes, sinon la plus ancienne de cette expression, sous son equivalent arabe d'a/-//7 al-mahd se trouve dans la (Ps.-)Theologie d'Aristote, ouvrage parfois mis en rapport direct avec ce meme Porphyre et son ecole30. Quoi qu'il en soit, ce qui importe pour nous ici, c'est qu'Ibn Sina a lu la Theologie, et qu'il a meme ecrit un commentaire sur elle, dont seulement quelques fragments ont ete conserves. Parmi ceux-ci, le chapitre contenant la 26. Nous nous basons pour 1'ensemble de ces idees sur D. Gutas, op. cit., p. 115-130. 27. J. Michot, La Destinee..., p. 6, n. 29. Son argumentation nous parait plus concluante que celle proposee par D. Gutas, op. dr., p. 99-100 en faveur d'une datation plus ancienne. Observons que la notion de science universelle s'y trouve dans une section n'ayant pas de correspondant dans la Nagdt. 28. Averroes, Tafslrmd ba'dat-tabi'at, ed. M. Bouyges, Beirut, 1938, III, p. 1565. 29. Nous empruntons ces donn6es £ Shi. Pines, « Les textes arabes dits plotiniens et le courant porphyrien dans le neoplatonisme grec », dans Studies in Arabic Versions of Greek Texts and in Medieval Sciences, « The Collected Works of Shi. Pines », II, Jerusalem, Leiden, 1986, p. 264-77, spec. p. 270-271. Nous remercions R. Taylor d'avoir attire notre attention sur 1'occurrence de ladite expression dans ce commentaire anonyme sur le Parmenide. 30. 'A. Badawi (ed.), Utiilugiyd Aristdtdlis, dzmAflutin 'inda 'I-'Arab, al-Qahira, 1955, p. 3-164. A la page 51 la notion d'acte pur se trouve trois fois. Quant au nom de Porphyre, il est explicitement mentionne dans Vincipit de 1'ouvrage, mais quelle credibilite peut-on accorder £ une telle information ?

VI LES TA 'LlQAT D'IBN SlNA

121

notion d'acte pur ne figure malheureusement pas31. Toutefois, des fragments qui nous restent, il apparait clairement qu'Ibn Slna reprend souvent dans son commentaire la terminologie « technique » du texte commente, i.e. la Theologie. Par consequent, n'est-il pas probable qu'il eut fait de meme pour notre notion en question ? Mais I'objectivite nous oblige a reconnaitre que la reprise d'un terme ne semble pas toujours impliquer qu'Ibn Slna souscrit automatiquement a la conceptualisation precise qu'il a regue dans la Theologie32. Toutefois, en raison de sa tonalite aristotelicienne, il ne serait nullement etonnant qu'il 1'ait trouvee valable comme appellation pour Dieu, a cote de tant d'autres33. Le commentaire d'Ibn Slna sur la Theologie (qui fait partie du K. al-Insdf) peut done constituer la source des Ta'liqdt en ce qui concerne 1'utilisation de la notion d'acte pur. Troisieme element doctrinal: L'Intellect agent comme premier emane Qu'il s'agisse d'une hypothese fondee, nous essaierons de le demontrer a partir de 1'examen critique du dernier point doctrinal « inhabituel ». Plus que dans les deux autres cas, ce dernier qualificatif est ici de rigueur. En effet, Ibn Slna affirme partout de fagon absolument non ambigue que 1'Intellect agent est le dernier des dix Intelligences celestes, et n'est par consequent nullement le premier emane. II est vrai qu'un des fragments du Commentaire sur la Theologie evoque vaguement une telle idee, mais tout indique qu'Ibn Slna y prend la partie pour le tout, c'est-a-dire 1'Intellect agent pour 1'ensemble des Intelligences superieures34. Cette reference est done, en fait, insignifiante. Par contre, ce qui ne semble pas etre sans valeur reelle, c'est le fait qu'al-Shahrastani, dans la section sur Aristote de son Livre des Religions et des Secies, place cette opinion dans la bouche du Stagirite meme35. Cela ne peut que surprendre. Neanmoins, rappelons qu'al-Shahrastani base son expose de la pensee aristotelicienne en grande partie sur le commentaire d'Ibn Sina du livre A de la Metaphysique (qui fait aussi partie du K. al-Insdf doni, comme dans le cas du Commentaire sur la Theologie, il nous est parvenu qu'une petite partie)36. On peut done se demander si al-Shahrastanl n'a pas trouve cette affirmation dans une partie perdue de ce commentaire d'Ibn Slna ? Mais n'est-il pas trop facile, voire inconsidere, d'invoquer de nouveau un texte perdu ? Nous ne le croyons pas, malgre 1'absence (provisoire) de preuves 31. La seule edition existante a ce jour se trouve dans 'A. Badawi, Arista 'inda-l- 'Arab, p. 37-74. Quant a 1'existence de deux recensions de ce commentaire, voir D. Gutas, op. cit, p. 138 (en ce qui concerne leur rapport & 1'edition Badawi, cf. D. Gutas, art. cite n. 8, p. 1213). 32. Par exemple, pour designer 1'emanation, Ibn Sina utilise dans son commentaire sur la Theologie le terme d'inbigds, terme technique par excellence dans la Theologie, mais absente dans tous les autres ouvrages d'Ibn Sina lui-meme. 33. Cf. notre article cit6 n. 12, p. 262-269. 34. Cf. 'A. Badawi, Arista..., p. 49. Cette reference est notee et analyse'e par J. Jolivet, Shahrastdnl. Livre des religions et des secies. Traduction avec introduction et notes par J. Jolivet et G. Monnot, II, Leuven, Paris, 1993, p. 292, n. 2. 35. J. Jolivet, Shahrastani, p. 292. 36. De nouveau, 'A. Badawi, Arista..., p. 22-33 offre la seule Edition existante.

VI 122

formelles. En fait, nous pensons pouvoir fournir un indice relativement probant. Nous savons qu'Ibn Slna a utilise, pour son interpretation d'Aristote, les commentateurs grecs, en premier lieu Themistius, mais aussi Alexandre d'Aphrodise. Or, selon les textes de ce dernier conserves en traduction arabe, 1'Intellect agent est en meme temps cause premiere de 1'Univers37. En combinant cette doctrine a un schema emanatiste, on arrive presque naturellement a 1'idee de T Intellect agent comme premier etre emane. Miskawayh, en tout cas, n'a pas hesite a tenir explicitement un tel propos38. Serait-il alors etonnant qu'Ibn Slna, du moins a une certaine etape de sa pensee, en ait fait de meme ? Si cette hypothese s'avere quelque part correcte, elle nous permet de faire un nouveau rapprochement entre les Ta 'liqdt et le K. al-Insaf, dont la redaction est a situer aux environs des annees 1029-1030. Notons tout de suite que ces dates concordent bien avec les dates avancees a 1'issu de 1'examen du premier point doctrinal, c'est-a-dire 1027-1029, ou un peu plus tard. Tout cela s'accorde d'ailleurs parfaitement avec la datation la plus probable pour le Ddnesh-Ndmeh, qui est a situer entre 1027 et 103239. En vue de la coherence de toutes ces donnees, nous osons proposer comme periode la plus probable de la redaction des Ta'liqdt, la periode dite « orientale » d'Ibn Slna, c'est-a-dire approximativement entre 1027 et 103040.

CONCLUSION Malgre la presence de nombreuses hypotheses et suppositions dans ce qui precede, nous esperons avoir ouvert quelques voies directrices pour des recherches futures, et surtout avoir attire 1'attention sur 1'oeuvre par trop meconnue, que sont les Ta 'liqdt.

37. H. A. Davidson, op. cit., n. 11, p. 30. 38. Nous avons signale ce fait dans notre article « Le Livre des religions et des secies d'al-Shahrastam : une conception particuliere de 1'historiographie », Bulletin de philosophic medievale 35, 1993, p. 104-112, p. 106. Toutefois, etant donne 1'affirmation des Ta'liqat, nous doutons maintenant que Miskawayh ait et£ la source directe d'al-Shahrastani. Rien ne plaide pourtant en faveur des Ta Tiqdt comme etant cette source, dans la mesure ou nous n'avons pas detecte une seule autre reference a elles dans le Livre des religions et des sectes. Qu'al-Shahrastani ait puise cette idee dans un fragment, actuellement perdu, du K. al-Insdf, se laisse par contre aisement concevoir. En effet, son expose sur Aristote est largement tributaire du Commentaire sur le livre A de la Metaphysique d'Ibn Sina, comme il ressort clairement des notes tres fournies de J. Jolivet. 39. La premiere date est suggeree par D. Gutas, op. cit., p. 67, tandis que la derniere est proposee par J. Michot, art. cit. n. 3, p. 154. M. Cruz Hernandez, dans une communication privee, incline aussi vers une datation tardive. 40. D. Gutas, op. cit.,passim, a bien specific la signification exacte de cette periode dite « orientale ». Cf. son article « Avicenna. Mysticism », dans Encyclopedia Iranica, 1989, III, p. 79-81.

VII LE DANESH-NAMEH d'IBN SINA : UN TEXTE A REVOIR ? Quelques considerations a propos de la decouverte d'un original (partiel) (dans les Ta?l£qdt) et d'une traduction (interpretative?) (les Maqdsid d'al-Ghazzall) arabes. Le Ddnesh-Ndmehl d'Ibn Slna se revele un ecrit philosophique particulier a plus d'un egard.

II se distingue d'abord par la

langue utilisee, le persan moderne- etant alnsi, si pas le premier, au moins un des premiers textes du genre philosophique ecrits en cette langue ; par sa division generale -I1etude de la metaphysique y precede celle de la physique et des mathematiques ; enfin, par son caractere hautement abstrait et synthetique -ce qui le revele comme une encyclopedic foncierement philosophique.

Ces particularites ne

rendent evidemment pas faciles ni 1'interpretation, ni 1*edition ou la traduction de cette oeuvre.

Par consequent, toute indication pou-

vant contribuer a une meilleure comprehension vaut la peine d'etre examinee en detail. Nous croyons en avoir decouvert deux.

II s'aglt d'abord

de ce qui fut probablement la premiere redaction, faite par Ibn Stna lui-meme, de la partie metaphysique du

Danesh-Ndtoeh, au moins de

quelques fragments qui en ont ete conserves.

Ce 'brouillon' fut re-

dige par lui en langue arabe, et se retrouve dans ses Tacl£qdt Glosses (concernant la metaphysique et la psychologic d'Aristote). En second lieu, nous croyons pouvoir affirmer que les Maqdaid ait un texte d*attribution certaine a al-Ghazzal£, ne 1. Ed. MO'IN-MESHKAT. 1975. 2. Ed. A. BADAWI.

Teheran, 1953.

Repr. Teheran, Dekhoda B.,

Le Caire, 1973.

3. Ed. S. DUNYA. Le Caire, 1961. La traduction latine medievale fut editee, en ce qui concerne les parties metaphysique et physique, par J. MUCKLE, Algazel's Metaphysics (sic!). Toronto, 1933, et, en ce qui concerne la partie logique, par Ch. H. LOHR, Log-lea Algazali Introduction and critical text, in Traditio* 21 (1965), p. 223-290. On peut aussi consulter 1'edition imprimee de Venedig, 1506. Repr. Frankfurt/Main, Minerva, 1969.

VII 164

contiennent nullement un resume par celui-ci des grandes theses philosophiques developpees avant lui -comme la majorite des commentateurs a cru jusqu'a maintenant -, mais tout simplement une traduction, bien de par sa main, en arabe de ce meme Ddnesh-Ndmeh d'Ibn Stna.

Celle-ci ne s'avere pas litterale, mais elle respecte nean-

moins 1'essentiel du contenu -ainsi la meilleure qualification nous paratt etre celle de 'traduction interpretative1. Demontrons maintenant les donnees concretes de cette double Mecouverte1 . Signalons d'abord les paralleles existant entre les TaPliqdt et le °Ilm Ildh£ du Ddnesh-N&neh. Ddnesh-N&neh

TaPltqdt

§

1741.

9 20

18-175 1.9

176 1. 12-177 1.5; 177 1. 19-178 1. 24; 179 1. 8-13

21

181 1. 5-12; 182 1. 9-16

22

182 1. 16- 183 1.8

23

185 1. 22-23

24

184 1. 28 -185 1.7

25

186 1. 8- 187 1. 12

4. Cfr. e.a. parmi les etudes recentes, celle de M. SHERIF, Ghazdli's Theory of Virtue. New York, SUNY, 1975, p. 4-5 ou celle de E. ORMSBY, Theodicy in Islamic Thought. The dispute over al-Ghazdl£'s "Best of al.l possible worlds". Princeton, New Jersey, Princeton Univ. Press, 1984, p. 84. M. STEINSCHNEIDER, Die Hebraeischen Uebersetzungen des Mittelalters und die Juden als Dolmetscher. Berlin, 1893. Repr. Graz, Ak. Druck- u. Verlagsanstalt, 1956, p. 311 avait pourtant deja signale une ressemblance forte avec la grande encyclopedic (i.e. le Shifd') d'Ibn Slna, mais davantage encore avec la petite (probablenient referant ainsi a la Najdt, plutot qu'au Danesh-Nameh). LOHR, Logica Algazalis... Introd., designe par contre de facon non ambigue le Ddnesh-Ndrneh comme source principale, mais il fait etat d'une reelaboration intelligente par Ghazzalt. La meme idee se trouve aussi dans I1introduction a la traduction espagnole des Maqdsid, cfr. M. ALONSO, Maqdsid alfaldsifa o Intenciones de los filosofos. Barcelona, 1963, p. XLVLII. Un lien etroit avec I1encyclopedia persane d'Ibn Stna avait done deja ete reconnu avant nous. Si neanmoins nous parlons de 'decouverte*, c'est par rapport au caractere de ce lien : en premier lieu traducteur, et non interpretatif!

VII Le Ddnesh-Ndmeh d'Ibn Stnd

165

26

187

1. 16- 188 1. 15

27

188

1. 16- 189 1. 5

28

189

1. 6-21 1. 24 - 190 1. 26 ; 192 1. 11-16

29

189

30

193 1. 5-15

32

13-14 1. 2 ; 14 1. 27- 15 1. 9

33

16

1. 10 -17 1. 4 ; 18 1. 8-17 ;

18,1. 23-19 1. 14 1. 15- 20 1. 10

34

19

35

20 1. 23- 21 1. 3 ; 21 1. 19- 22 12 22 1. 3-9.

36

Precisons en toute honnetete que dans certains de ces cas, 11 vaut mieux parler de paraphrase que de traductlon litterale, alnsl pour les paragraphes 23, 24, 27 et 29, le debut des paragraphes 20 et 32, le milieu du paragraphe 33 et la seconde moitie du paragraphe 35. Mais nfest-il pas alors peremptoire, voire injustifie, de designer ces quelques fragments des Taal£qdt comme ayant fait partie d'un original arabe? Nous ne le croyons pas.

II est a remarquer qu'il s'agit bien de deux

ensembles -les paragraphes 20 a 30 ainsl que 32 a 36- repris tels quels dans les TaPltqdt, p. 175-193 et 13-22, c'est-a-dire la fin et le debut de lfedition actuelle.

Ne peut-on pas en deduire raisonna-

blement qu'il y ait eu un glissement dans cette derniere, ou plus probablement dans la tradition manuscrite sur laquelle elle se base? A cette occasion, une partie -le paragraphe 31 et le tout debut du paragraphe 32- se serait perdue, ce deplacement ayant ete accompagne de la perte d'un ou de plusieurs folios.

En plus, la presence d'un

correspondant arabe au paragraphe 9 -et comme nous n'avons effectue qu'une premiere analyse asse native, 11 se peut que certains paralleles nous aient echappe- ne fait qu'accrottre la probabilite de 1'existence d'un original arabe de la partie metaphysique entiere a 1'interieur des TaPltqdt.

VII 166

Remarquons d'ailleurs que 1*edition recente de celles-ci par Badawi apparalt a premiere vue assez defectueuse.

En effet, une

comparaison plus detaillee entre la redaction arabe et la persane, specialement concernant les paragraphes 22 et 33, nous a revele des fautes de lecture, par exemple, p. 17, 1. 1 khayr a^n au lieu de wujudailn ; p. 19, 1.2 na°qalu au lieu de naf°alu ; des omissions, par exemple, p. 175 1.6 la categoric de possession

(ar. malaka ,

pers. rrrulk) a clairement ete oubliee ; p. 183 1. 4 toute une phrase est tombee entre al-khdssa et

fa-idhan ; voire des distorsions, par

exemple, p. 17 1. 3 ou la formulation actuelle arabe est incomprehensible -une omission en est probablement la cause principale-, mais or se demande si 1'ordre actuel des mots est bien correct ; p. 19 1. 2 wa-ld jusqu'a 1. 3 nafi° est, tel quel, un passage entierement superflu -a partir du persan on y attendrait quelque chose dans le genre de wa-in yakun bi-shawq. 11 se peut que certaines de ces fautes soient a incriminer a des copistes anciens.

Comme nous n'avons pas eu le temps de consul-

ter des manuscrits, nous ne sommes malheureusement pas capables d'en fixer les limites exactes^.

Toutefois, on ne peut s'empecher de

croire que M. Badawi a fait son edition un peu a la hate -ce qui expliquerait e.a. I1absence d'un appareil critique, indiquant les variantes de lecture.

Qu'elle reste dans 1*ensemble encore assez hon-

nete n'est probablement du qu'4 sa grande erudition, plutot qu'a un examen minutieux de tous les manuscrits existants.

Ainsi, elle est

utile, mais a utiliser avec precaution. Par contre, 1*edition persane semble avoir ete faite avec beaucoup plus de soin.

Neanmoins, on y rencontre quelques fautes,

mais tres souvent d'ordre plutot secondaire, bien qu'une correction plus fondamentale s1impose parfois, comme par exemple p. 75, 1. 9-10, ou le sens requiert la lecture suivante : andar mdhiyya (au lieu de wujub al-wujud) wdjib al-wujtid wa-£n mdhiyya wujub al-wujud (au lieu de wdjib-al-wujud aat), ce qui est confirme par I1original arabe 5. Mr. J. MICHOT, qui a pu consulter plusieurs manuscrits, nous a confirme I1existence de lacunes importantes dans 1*edition Badavi (communication orale). Nous 1'en remercions.

VII Le Danesh-Ndmeh d'Ibn Stnd

167

(Ta°l., p. 183, 1. 2-3 bien qu'il n f y ait pas de correspondance litterale absolue). Toutefois, le texte persan, tel qu'il est presente actuellement (et qu'il a ete conserve dans la majorite (totalite?) des manuscrits), frappe par un manque de logique assez evident dans le develop pement des idees.

Ainsi, par exemple, dans la partie metaphy-

sique, I1expose du retour a Dieu precede celui de la creation. 11 est vrai qu'Ibn Slna a consciemment adopte une structure particuliere pour cette oeuvre, connne nous 1'avons indique au debut. Mais cela ne nous semble pas justlfier des deplacements aussi inhabituels que celui que nous venons de mentionner.

A

notre grand etonnement, Ghazzali dans ses Maqdsid offre cet ordre naturel, tout en gardant la specificite de la division generale.

Qu'il s'y inspire largement du Ddnesh-N&neh ne

etre mis en doute.

peut

Lea deux ecrlts demontrent une tres forte

ressemblance quant au contenu.

L'apport personnel de Ghazzali

semble a premiere vue se limlter a 1'ajout de quelques resumes et a la concretisation de certaines biais d'exemples).

idees tres abstr.aites (souvent par le

11 est a remarquer en plus qu'il evoque, sans

emettre aucune reserve, quelques theses philosophiques, qu'il qualifie partout ailleurs d'impies, comme par exemple 1'eternite du monde, la resurrection de 1'ame seule, ou qu'il les evite scrupuleusement, comme par exemple la posslbilite d'une vie 'imaginale' dans 1'au-dela . Tout cecl nous fait incliner vers 1'hypothese d'une traduction Interpretative. Mais a quel degre elle est ainsi, depend principalement de 1'explication a donner aux differences d'ordre dans les deux exposes.

Sont-elles dues a un remaniement par Ghazzali lui-meme,

ou, par contre, s'expliquent-elles par une corruption dans la tradition manuscrite du Ddnesh-Nfimeh a un certain moment de 1'histoire? Ce qui plaide en faveur de la premiere alternative, c'est la presence d'autres remaniements -bien que d'envergure plus modeste-.

Mais on peut

6. La seule omission importante, qu'une premiere lecture nous a revelee, est 1'absence de la notion d'Intellect saint (°aql alqudat) dans la Physique (§ 51 du Dan. Nam.).

VII 168

denombrer plusieurs arguments a 1'avantage de la seconde : 1. 11 est assez inconcevable qufun esprit aussi brillant et systematique que celui d'Ibn Slna ait redige un texte si maladroitement structure ; 2. 1'ordre, retenu dans les Maqdsid, recoupe celui des grandes oeuvres avicenniennes hors du Ddnesh-Ndmeh ; 3. une corruption dans la tradition manuscrite est bien concevable, quand on sait que Juzjani le fameux disciple d'Ibn Slna, a reconstitue la partie mathematique, parce que, selon ses propres mots, il n'avait pas pu retrouver I1 original? . leur decisive.

Aucun de ces elements n f a pourtant de va-

Ainsi, afin de pouvoir prononcer un jugement defi-

nitif, une analyse comparative et approfondie s'avere necessaire. En attendant celle-ci^, nous penchons -mais avec les reserves voulues- vers la derniere alternative. Donnons ma in tenant un apercu global des paralleles qui existent entre les Maqdsid et le Danesh-Ncuneh.

Pour faciliter la

tache du lecteur, nous signalerons a la fin de chaque partie, brievement, les glissements les plus import ants entre les deux textes, y ajoutant un premier essai d f explication. (Pour les Maqds-id, nous nous referons a lf edition de S. Dunya) . logique

Ddnesh-fl&neh

Maqdsid

§

p. 33-37, 1. 13

1

2

40, 1. 1-17

3

40, 1. 18-41, 1. 7

4

44-48, 1. 3

5

4^,1. 4-50, 1. 20 et p. 52, 1. 1219

6

50, 1. 21- 52, 1. 11

7

41, 1. 8- 42, 1. 6

8

53-54, 1. 7

7. Cfr. M. ACHENA et H. MASSE, Avicenne. Le livre de science. II. Physiquej Math^matiques. Paris, Les Belles Lettres, 1958, Pref. p. 6. 8. Nous esperons pouvoir 1'effectuer dans un avenir prochain.

VII Le Ddnesh-Nameh d'Ibn Stnd

9

169

54, 1. 8-15

10

55, 1. 16-17 ; 57, 1. 22-60, 1. 5

11

55, 1. 17-57, 1. 21 et 60, 1. 6-18.

et 60, 1. 19- 61

12

62-64, 1. 3

13

64, 1. 4-65

14

66, 1. 9-67, 1. 18

15

67, 1. 19- 69

16

70-72, 1. 12

17

77, 1. 7- 80, 1. 6

18

80, 1. 19- 83

19

84- 86, 1. 8

20

86, 1. 9- 87

21

96-99, 1. 3

22

88

23

89 - 90, 1.4

24

manque

25

90, 1. 5-95

26

99, 1. 4 - 101

27

102- 109

28

110-112

29

122- 123, 1. 9

30

123, 1. 10- 125, 1. 9

31

127, 1. 12 -128, 1. 18 125, 1. 10 -fin; 126, 1. 15- 127,

32

1. 11 et 129, 1. 4-fin 33

120-121

34

118-119

35

113 - 117.

Les deplacements les plus importants a signaler sont : le § 7 (qui continue tout

naturellement 1*expose du terme simple,

entame au § 3), le S 21 (sur les syllogismes composes, qui ressortent ici comme la reponse adequate au raisonnement par analogic des dialecticiens) ; les § 33, 34 et 35 (dont 1'ordre a ete renverse, et qui ont ete ainsi places entre les § 28 et 29 - le § 35, avec ses conseils contre le sophlsme faisant logiquement *

Voir 'Note Additionnelle' p. 177.

VII 170

suite au § 28 sur la dialectique et le § 33 precedant aussi logiquement le § 29, car ce dernier n1offre qu'un complement de commentaire sur le syllogisme demonstratif proprement etudie dans le premier).

En outre,

on constate un double melange : une premiere fois entre les § 10 et 11, une seconde entre les § 31 et 32 (nous ne pouvons pas entrer dans le detail ici, mais la aussi, cela s'avere souvent beaucoup plus logique), ainsi que lfabsence de lvequivalent d'un paragraphe, le 24, ou il est question d'un raisonnement

(faible!) par analogic (Ghazzalt

1'a peut-etre, a cause de cela, juge indigne d'un traite de logique). II. Metaphysique

Danesh-N&neh

Maqdsid

§

1

134 - 137

2

138- 140, 1. 16

3

140, 1. 17- 143, 1. 6

4

144 - 147, 1. 12

5

154, 1. 16- 157, 1. 1

6

147, 1. 13 - 154, 1. 15

7

157, 1. 2-23

8

158 - 162

9

163 - 165

10

166 - 170, 1. 15

11

171 - 174, 1. 4

12

174, 1. 5- 183, 1. 8

13

183, 1. 9 - 187, 1. 12

14

187, 1. 13- 189, 1. 2

15

189, 1. 3 - 192

16

193 - 200, 1. 3

17

200, 1. 4 - 203, 1. 16

18

203, 1. 17 - 205, 1. 11

19

212, 1. 9- 213, 1. 7

20

205, 1. 12- 209

21

216, 1. 1-22

22

213, 1. 8 - 215, 1. 17

23

215, 1. 18 - fin

VII Le Ddnesh-N&neh d'Ibn Stnd

171

24

211,

1. 13 -212,

1. 8

25

216,

1. 23 -219,

1. 13

26

223 - 224,

27

219,

1. 14 - 222,

28

222,

1. 3 - 21

1. 22 1. 2

29

225,

1 . 9 - 229,

30

229,

1. 13 - 232,

31

232,

1. 2- 233,

32

233,

1. 9 - 234

33

235 - 239,

34

239,

1. 15 - 240,

1. 22

35

240,

1. 23 - 241,

1. 19

36

241,

1. 20 - 242,

1. 11

37

242,

1. 12 - 249

1. 12 1. 1 1. 8

1. 14

38

288 - 290,

1. 8

39

253 - 254,

1. 15

40

254,

1. 16 - 255,

41

297,

1. 7 - 299,

1. 8

42

255,

1. 5 - 256,

1. 12

43

256,

1. 13 - 258,

44

258,

1. 11- 261,

1. 10

45

261,

1. 11 - 267,

1. 2

46

271,

1. 12 - 272,

47

272,

1. 21- 274,

48

275,

1. 5-20

319,

1. 4

49

267,

1. 3 - 268,

50

268,

1. 4 - 270

274,

1. 15- 275,

51

21- 279,

1. 4

1. 10

1. 20 1. 14

et Fya., 318,

1. 11-

1. 3 1. 4 et 275, 1.

1. 8

52-53

279,

1. 9 - 283,

54

283,

1. 17 - 287

55

290,

1. 10- 291,

56

291,

1. 15 - 293

57

294 - 297,

1. 16 1. 14

1. 6 et 299,

1. 9- 300,

VII 172

Signalons et commentons les glissements suivants : le § 5 est deplace apres le § 6 -ceci pourtant n'ayant aucune implication doctrinale ; le § 24, ou il est question de 1*essence de I1etre necessaire, est introduit immediatement apres la discussion sur 1'etre contingent (§ 20), tandis que le § 19 (I1impossibilite pour un etre necessaire d*avoir une union essentielle avec quoi que ce soit) fait suite a ces deux §§ ; le § 26, qui parle en termes generaux des attributs de Dieu, se retrouve directement avant 1*expose detaille de chacun d'eux ; le § 38, qui pose les fondements pour determiner la hierarchic des etres, est lie intlmement au § 55, contenant la description concrete du schema emanatif ; de meme que le § 41 (le bien et le mal en general) I1 est au § 57 (la cause du mal dans les etres corruptibles) ; les § 46 et 47, etablissant la necessite d'une amemoteur en vue du mouvement circulaire des spheres celestes, arrivent apres 1'etablissement de la necessite d'un mouvement pour tout ce qui se renouvelle (§ 49-50).

Un melange de deux paragraphes (52-53) se

note aussi dans cette partie de 1'ouvrage.

On remarque enfin que le

paragraphe 48, ou il s'agit de la difference entre le premier corps et les corps corruptibles, se retrouve presque entierement dans la partie physique, ce qui semble en fait plus logique, bien que sa presence dans la metaphysique puisse se defendre. III. Physique Dfineah-N&neh

§

Maqdsid

1

303 - 304, 1. 12

2 3 4

304, 1. 13 - 307, 1. 5 307, 1 . 6 - 308, 1. 2 308, 1. 3 - fin

5 6 7 8 9 10

309 - 311 312 - 313, 1. 4 313, 1. 5 - 315, 1. 2 315, 1. 3 - 16 315, 1. 17 - 316, 1. 24 316, 1. 25 - 317, 1. 14

11

317, 1. 15 - fin

VII Le Ddnesh-N&neh d'Ibn Stnd

12 13

14, 15 i 16 17 18 19 20 21 22

322, 1. 12 - 323, 1. 5 318 - 321, 1. 22 332, 1. 8 - 334 323, 323, 325, 325, 330, 328, 321, 328,

1. 6 - 20 1. 21 - 325, 1. 1 1. 2 - 4 1. 5 - 326 1. 15 - 332, 1. 7 1. 6 - 329, 1. 22 1. 23 - 322, 1. 21 et 327 1. 5

23 24 25

336 - 336, 1. 19 336, 1. 20 - 338 329, 1. 23 - 330, 1. 14

26 27

339 - 342, 1. 13 342, 1. 14 - 344, 1. 7

28 29 30 31 32 33 34 35

344, 1. 8 - 345 346 - 347, 1. 7 347, 1. 8 - 349 350 - 352, 1. 3 353, 1. 23- 355, 1. 5 352, 1. 4- 353, 1. 22 355, 1. 6 - fin 356 - 357, 1. 24

36

357, 1. 25 - 358

37 38

359 - 360, 1. 7 360, 1. 8 - 362, 1. 5

39 40 41 42

362, 362, 364, 367,

43

368, 1. 23- 370, 1. 9

44 45-46 47

371 - 375 376 - 378, 1. 19 378, 1. 20 - 379, 1. 18

1. 1. 1. 1.

6 - 22 23- 364, 1. 17 18 - 367, 1. 21 22- 368, 1. 22

17 S

VII 174

48

379, 1. 19 - 380, 1. 17

49

380, 1. 18 -382, 1. 6

50

382, 1. 7 - 383, 1. 14

51

383, 1. 15 - 385. Un premier grand changement dans 1'ordre se produit apres le

§ 11, qui termine la discussion sur le probleme du lieu, specialement celui de sa definition.

II est suivi ici d'une introduction genera-

le sur les corps simples (§ 13), ainsi que d'une introduction

con-

cernant 1'influence de la nature d'une chose sur son lieu naturel (premiere partie du § 22).

Alors on commence a parler de la froi-

deur et de la chaleur (§ 12).

Un second se situe apres 1*ensemble

des § 16-19, dans lequel trois theories de 1'echauffement sont presentees et refutees.

On y discute la difference entre la forme et

les accidents des corps (S 22, deuxieme partie), la transformation des quatre elements 1'un en 1'autre (§ 21), ainsi que lfinfluence exercee par les corps lumineux (§25 -soleil et lune, suivi du § 20 -un discours complementaire sur ce probleme). la, 11 est question du lieu (naturel)

Seulement apres ce-

que le corps simple possede

(melange des S 14 et 15). Le deplacement du § 32 apres le S 33 et le melange des § 45 et 46 dolvent aussi etre signales, mais sont sans aucune importance doctrinale. II nous semble qu'apres cette mise en parallele, les conclusions suivantes s'imposent (mais soulignons encore une fois que parallelisme ne signifie pas ici traduction lltterale) : 1. L1ensemble du Ddnesh-N&neh (ou presque) se retrouve dans les Maqdsid. 2. Ce dernier ne developpe pas clairement des themes en dehors de ceux qui se trouvent dans le premier. 3. Ce dernier est du point de vue logique tellement mieux structure, qu'on est naturellement incline a croire que cet ordre se trouvait deja chez Ibn Stna lui-meme. En outre, une premiere comparaison plus detaillee entre les deux textes en ce qui concerne les S 22 et 33 de la metaphysique,

VII Le Ddnesh-Ndmeh d'Ibn S£nd

175

nous a demontre a suffisance que Ghazzall reste assez proche des idees d'Ibn Sina.

On

constate surtout un effort a rendre plus

concret, et ainsi en meme temps plus accessible, le texte hautement abstrait du D&nesh-Nameh.

Lfomission de quelques phrases (p.e. D.N.

p. 95, 1. 9-10 et 1. 12-15), ou la creation d'une perspective legerement differente (comme p.e. a la fin du § 22) pourraient se comprendre dans la meme ligne (mais indiquent aussi le caractere foncierement

non litteral), mais ce sont surtout les exemples concrets,

que Ghazzali a clairement ajoutes de sa propre main (p.e. p. 237, 1. 26 - 238, 1. 14

pour illustrer p. 95,1. 1

sqq. du Ddnesh-fldmeh),

qui devoilent sans ambigulte cette tendance (comme le font d'ailleurs aussi les multiples resumes generaux). C'est pourquoi nous croyons -et cecl est notre conclusion finale- que le texte actuel du Ddnesh-N&neh est un texte a revoir, et done a reediter, en tenant compte aussi bien des Maqdsid de Ghazzali (pour 1'entierete), que des Ta^l^qdt (au moins pour une partie de la metaphysique)^. Appendice II nous a semble interessant de faire aussi une premiere evaluation des deux traductions contemporaines qui existent de la partie metaphysique, en tenant compte de la redaction des Ta °l£qdt et de la traduction des Maq&sid.

De facon globale, il s'est avere

que la traduction francaise de M. ACHENA et H. MASSE, Avicenne. livre de science.

Le

2 vol., Paris, Les Belles Lettres, 1955-1958, t. I,

p. 89-216 est assez fiable, mais manque souvent d'une precision absolue.

L'anglaise, par contre, faite par P. MOREWEDGE , The Metaphysica

of Awicenna (Tbn S£nd) • London, Routledge a. Kegan, 1973 (accompagnee d'un commentaire assez etendu) ne semble guere satisfaisante bien qu'elle soit pour beaucoup de details plus precise que la francaise. II nous parait que ce dernier a ete la victime de ses propres idees precongues a propos des lignes fondamentales suivies par Ibn Stna dans 1*elaboration de cette oeuvre.

Par consequent, une revision des

9. Lfinverse est, bien sur, aussi vrai!

VII 176

deux est (plus que) souhaitable. Illustrons-en la necessite par un petit extrait du § 33, concernant la volonte divine.

Les lignes 7-9 de

Q

la p. 93 (va-har keh fa l£. .bi-ddnesh dyad) sont rendues par Morewedge, p. 66 : "If someone knows that an action results from him and his agentness (i.e. his awareness of his role as a knower), then that action becomes known his knowledge)".

due to him (i.e. due to the activity of

Le caractere hautement interpretatif de cette ver-

sion anglaise se revele immediatement par les explications ajoutees entre parentheses.

Achena et Masse, par contre, offrent ostensible-

ment une lecture plus neutre : "Toute personne de qui precede un acte a conscience de cet acte et d'etre cause de cet acte ; c'est (alors) I1acte procedant d'elle par science" (p. I, 160). terme 'cause* ne figure nulle part dans le texte persan.

Mais le Deja dans

notre these de doctorat, Avicenna : tussen neoplatonisme en islam^ 2 vol., Leuven, 1984, polyc., nous emettions ces critiques (p. II, 50, N. 106).

Nous y proposions la traduction suivante : "Chaque fois

qu'un acte precede de quelqu'un, et qu'il en est conscient, ainsi que de lui-meme comme son auteur, cet acte precede de lui par science".

Tacl£qdt, p. 16, 1. 12-13 confirme 1'essentiel de cette lectu-

re (!'equivalent exact de ces lignes manque dans les Maqdsid ) : annahu fd°ildhu! (mais semble necessiter le deplacement du keh apres faol£ : tout acte, procedant de quelqu'un, et dont...). Ceci n'est nullement un cas isole.

Plus loin, dans ce mime

chapitre, i.e. p. 94, 1. 10, on trouve tamdm hast-C.

Morewedge le

traduit : "the complete realm of Being", tandis qu*Achena et Masse retiennent : "toute existence".

Dans les deux cas, on dolt inevita-

blement conclure a la presence de fortes tendances pantheistiques dans la pensee d'Ibn Stna.

A partir du contexte, il est pourtant as-

sez evident qu'il s'agit ici de la these (neoplatonicienne) que Dieu est parfait, voire le plus-que-parfait. lire : le parfait d'etre (en arabe

Par consequent, il faut

il ne se trouve que tdrnm).

D'ail-

leurs tout le contexte a ete mal compris par Morewedge, qul poursult

VII Le Ddnesh-N, Met., 9: 1; 9: 4; 8: 1 and 3; 8: 7; 9: 1; 8: 4; Ishdrdt (ed. Forget), 144 based on preceding questions Ishdrdt, 168 no precise reference Shifd\ Met., 8: 6; no precise references, but the title may be inspired by Shifd>, Met., 381, 1. 11; 9: 2; end Aqsdm24 Shifd', De An., 4: 4, and Ishdrdt, 219 4: 1 and 5: 2, and Najdt, De An., c. 11-13

22 For more detail, see Gutas, Avicenna and the Aristotelian Tradition, 112-14, but let me remark that whereas Gutas believes that Ahwdl al-nafs is earlier than the Najdt, Michot, Destinee, 6, has offered strong evidence for its being later. 23 See Janssens '... His Use of Avicennian Texts' (n. 1, above). 24 See J. Michot, 'Les sciences physiques et metaphysiques selon la R. ft aqsdm al-'ulum d'Avicenne', Bull, pbilos. med., 22 (80), 62-73, App.

X A L - G H A Z Z A L l ' S TAHAFUT

q. 19: q. 20:

II

5: 4, and Najdt, c. 13 Shifd\ Met. 9: 7, and Adhawiyya, c. 3.25

I can now set out a number of observations: 1. Ibn Slna's Shifd* (above all Ildhiyydt, but for questions regarding the soul also K. al-nafs) undoubtedly merits the qualification of basic source of the Tahdfut. However, if one looks for literal, or nearliteral, citations, then most of the time one comes across isolated sentences, and longer passages rather seldom. In this respect, the use of the Shifa* in the Tahdfut is not only extremely different from that of the Ddnesh-Ndmeh in the Maqdsid, but also perceptibly different from that of Avicennian works in other writings of al-Ghazzall, as e.g., Mizdn, Mi'ydr, or, possibly, Ma'drij. In all these cases, the text(s) which function(s) as a source can easily be recognized, notwithstanding the clear interventions that articulate al-Ghazzall's own thinking. Moreover, the source text(s) can also be traced in the structure of the exposition and, often, even in the precise wording. The Tahdfut, for the most part, reformulates elements, based more or less on the Shifd', in a much freer way, to the point indeed that one wonders if al-Ghazzall is still reproducing Ibn Slna's thought correctly. For example in q. 5 (89) he affirms, without adding any further qualification, that for Ibn Slna the quiddity is prior to existence; and in q. 17 (172) he states that particular events are caused by the 'Dator formarum'. Marmura26 rightly insists that in the former case the interpretation given is debatable, even doubtful, and that the latter affirmation does not correctly reflect Ibn Slna's causal theory, since Ibn Slna surely does not say that all particular events are caused by the active intellect. But even discounting these rather problematic interpretations, one has the impression that al-Ghazzall is presenting a kind of personal synthesis, the core of which consists in a few literal, or near-literal, citations. 2. Two major passages from two minor works, Aqsdm and Adhawiyya, function as an important source in q. 1627 and q. 20 25 S. Dunya, in his edition (Cairo, 1949) of Ibn Slna's R. adhawiyya, has already pointed to this derivation, see G.C. Anawati, 'Un cas typique de 1'esoterisme avicennien', La revue du Caire, June 1951, 91, n. 1. (Repr. in Etudes de philosophie musulmane, Paris, 1974, 286, n. 1.) (Unfortunately, I have not myself seen S. Dunya's edition.) 26 M. Marmura, The Incoherence, 237, nn. 6-7, and 242, n. 8, respectively. 27 M. Marmura, ibid. 164-9, indicates the use of the Aqsdm not in q. 16, but in what he describes as the 'Introduction' to the second part of the Tahdfut, i.e., the part concerning the 'natural sciences'. However, al-Ghazzall discusses the topic of the rational soul at some length in q. 16 and makes some reference to the doctrine of the soul in even earlier questions. Further, in q. 20, he places the issue of the ma(dd in a

X 12

respectively. In these cases, al-Ghazzall is using his sources in his more usual way. But their presence surprises in that both texts present views different from those held by Ibn Slna in the Shifd* (as well as in the Najdt and the Ishdrdt}. The division of the natural sciences in the Aqsdm is not in accordance with the actual division of the Shifd\ For example, parts of the Meteorology of the Shifd' are now incorporated into the (untitled) fourth book of the natural sciences, more especially, the De Anima is no longer presented as the sixth, but as the eighth book. The division as given by the Aqsdm is undoubtedly more genuinely Aristotelian28 than that of the Shifd', and has much in common with the order given by Ibn Slna's famous Arabic predecessors, al-Kindl29 and al-Farabl.30 In other words, it seems to reflect the point of view of the Arabic Aristotelian tradition. It is probably for that reason that al-Ghazzall quotes it. But why does he remain silent about Ibn Slna's revision(s)? Is it for the sake of easiness, or is he deliberately seeking to present Ibn Slna as an authentic follower of the Stagirite? For the moment, I prefer to leave this question open. The Adhawiyya, in a way similar to the Aqsdm, entails a radical departure from the Shifd\ insofar as it denies the reality of bodily resurrection without any qualification. As for the Shifd\ Ibn Slna explicitly states that the truth of bodily resurrection cannot be established within the domain of syllogistics or rational demonstrative proof, but must be accepted on the basis of prophecy. Of course, one may wonder whether this latter affirmation is not merely a rhetorical device, and whether the truth is not in what is clearly metaphysical perspective, linking it directly with the problem of the eternity of the world. Finally, as we noted above (p. 3) he explicitly states that his criticism in the Tahdfut is directed only against metaphysical doctrines, or matters related to metaphysics. His discussions on the soul do therefore need to be placed in the perspective of a metaphysics of the rational soul (which is indeed an Avicennian perspective). 28 See M. Mahdi, 'Avicenna. The Division of the Rational Sciences' in R. Lerner and M. Mahdi (eds), Medieval Political Philosophy (Glencoe, 1963), 95-7. 29 Al-Kindl, Kammiyya kutub Aristu in Abu Rlda, Rasd'il al-Kindl (Cairo, 1950), 1: 368, 382-3. 30 Al-Farabi, Ihsd' al-(ulum (ed. A. Gonzalez Palencia, Madrid, 1932), 49-50 (Arabic). 31 One may use the plural since Ibn Slna, in his Ddnesh-Ndmeh, seems to introduce an even more radical change, as already stated. It may be noted that the Aqsdm appears at first sight to be a work by the young Ibn Slna, not least in view of its high Aristotelian contents. But in matters of chronology extreme care is needed, and a later dating cannot at present be excluded. However, even if the later date turns out to be the correct one, the division of the sciences as given would still remain atypical for Ibn Slna. This might be the reason why he states at the beginning of the treatise that he will not go beyond what his interlocutor has asked him.

X A L - G H A Z Z A L l ' S TAHAFUT

13

presented in the minor 'esoteric' writing? Let me observe that things are never as simple as they appear to be at first sight. In fact, Michot32 has offered strong evidence for a rather early date, i.e., Rayy 1014-15, for the Adhawiyya treatise. It may then be that Ibn Slna changed his mind, and in a later development of his thought formulated a position that fits better with the requirements of Islamic belief. On the other hand, irrespective of any chronological consideration, a reconciliation of the formulation in the two works is possible on the basis that Ibn Slna, in the Adhawiyya, limits himself to articulating those matters, and only those, that are susceptible to apodeictic demonstration. Once more, al-Ghazzall remains silent about all this. And the question again arises whether he here wants to discredit Ibn Slna, or whether he just considers this position to be more typical of the general Arabic Aristotelian tradition? 3. One is struck by the (almost) complete absence of references to the Maqasid33 Hence, bearing in mind what I have already argued in earlier papers,34 I believe that I can now affirm without any reserve that the Maqasid was not written as a preparatory work to the Tahafut, and that therefore there is no direct link between the two works. But what then of the similarity in title of the two works? The fact that the Maqasid contains no other sources than Avicennian ones, on the one hand, and the fact that the Latin medieval translation is entitled 'Summa theoricae philosophiae'35 (this title being moreover rather close to the Persian 'Danesh-Nameh'l), on the other hand, would seem to indicate that the present title was not the original one.36 So far, a few formal considerations regarding al-Ghazzall's use of Avicennian texts in the Tahdfut. It is worthwhile keeping in mind that those texts do not constitute al-Ghazzall's only sources. We have already mentioned al-Farabl and Galen. We may add the Liber de Causis, which seems to have functioned as the source for al-Ghazzall's allusion to a mediative eternal circular movement that in one 32

J. Michot, Destinee, 2, n. 10, 6 and 23-4. It remains possible that a yet more thorough investigation may turn up some derived elements, but it is already clear that there is absolutely no evidence of any systematic use. 34 See J. Janssens 'Le Danesh-Nameh d'Ibn Slna: un texte a revoir?', Bull, philos. med., 28-86,163-77, esp. 167-75, and '... His Use of Avicennian Texts' (n. 1, above). 35 see Ch. Lohr, Einleitung to: Ghazzali, Logica etphilosophia. (Venice, 1506; repr. Frankfurt, Minerva, 1969.) 36 The title was not very clear in the Arabic tradition. See M. Bouyges, Essai de chronologic des oeuvres de al-Ghazali, revised edn by M. Allard (Beirut, 1959), 23, n. 5. 33

X 14

respect resembles the Eternal, and in another the temporal,37 as well as Proclus' De Aeternitate mundi (probably mediated through Philoponus' refutation), which appears to have constituted the basis of a proof meant to demonstrate the impossibility of the annihilation of the world.38 However, both the extent and the impact of these nonAvicennian sources is extremely limited compared to the Avicennainspired materials. What is particularly surprising is that there does not appear to be even one direct citation of Aristotle, even though, in the first Introduction, al-Ghazzall presents the latter as the most important of the ancient philosophers. Did he simply forget that basic statement when elaborating his work? I do not think so. It rather looks as if al-Ghazzall considered certain works of Ibn Slna to express in the best possible way the Stagirite's thoughts. If this was indeed his attitude, then he surely did not consciously intend to disguise Ibn Slna's own philosophical ideas, but rather to offer as perfect as possible a formulation of Aristotle' views (as perceived in the Arabic world).39 In light of this, it is easy to understand why he makes use of Ibn Slna's most 'Aristotelian' texts. For al-Ghazzall certainly did not consider Ibn Slna a slavish follower of the Stagirite in every respect. As I already mentioned, al-Ghazzall insists in q. 11 that Ibn Slna's philosophical Brethren disagreed with him on the issue of God's knowledge of the whole. On that issue Ibn Slna clearly took a position that more closely conforms with the data of the Qur'anic revelation. But, at the same time, it is certain that, in al-Ghazzall's eyes, Ibn Slna's demarche was in the end far too timid. In conformity with the essentials of the Aristotelian project, Ibn Slna still continued to try to explain God's knowledge in a purely demonstrative way, which allowed him to qualify it as 'universal'. In that respect, Ibn Slna remains somehow guilty of taqlid, albeit in a weaker form. Ibn Slna's philosophy moved in the right direction, but it failed to realize the final step. This seems to have been a cause of a great disappointment to al-Ghazzall. But where there was disappointment, there was at the same time a great fascination. The presence of so very many Avicennian fragments, from a wide range of works, in a great variety 37

See q. 1, p. 29 §74. A similar idea is expressed in the Liber de Causis, §30-31. See q. 2, p. 49, §13 seq. The relevant proof is proof 5 in the Greek edition of Rabe, and proof 6 in the edition of the Arabic translation by A. Badawi, Al-Afldtuniyya al-muhdatha 'inda l-'Arab (Cairo, 1955). 39 It may be noted that al-Shahrastanl in his Milal also presents Aristotle in 'Avicennicized' terms, his exposition of Aristotle being largely based on extracts from Ibn Slna's Commentary on Book Lambda of the Metaphysics. See the French translation by J. Jolivet in ShahrastanI, Livre des religions et des sectes (Paris, 1993), 2:281 seq., where all the relevant references can be found. 38

X A L - G H A Z Z A L I S TAHAFUT

15

of al-Ghazzall's writings, including Ihya' and Maqdsid40 illustrates that fascination most instructively. Indeed one sometimes gets the impression that al-Ghazzall is convinced that Ibn Slna's philosophy, including its metaphysics, when 'adapted' into the right 'framework', is acceptable.41 A good example is offered in al-Mustazhirt42 (§166-69), where al-Ghazzall illustrates 'necessary knowledge' in the domain of metaphysics (following on from an example in the domain of mathematics) by a (on first view, slightly) reworked version of Ibn Slna's proof of contingency, based on the distinction between necessary and possible. My impression, from a preliminary analysis of the text, is that al-Ghazzall puts a somewhat stronger emphasis43 than Ibn Slna had done in his original proof on the real contingent character of the 'possible', and this might express a shift towards the more classical kaldmi proof for God 'ex creaturis\ However we interpret it, it is clear from this passage that al-Ghazzall does not claim that metaphysics as such must be rejected, and that he judges some of Ibn Slna's formulations in this domain to be worthy of acceptance, albeit somewhat modified. Furthermore, when it is affirmed (as it must be) that this proof for God's existence is original with Ibn Slna, it straightaway becomes probable that Ibn Slna is not the philosopher al-Ghazzall has particularly in mind when he accuses the philosophers of taqlid. A few lines earlier, I said that al-Ghazzall does not reject metaphysics in its totality. But is metaphysics not presented in the Tahdfut as the 'bad guy'? How these two apparently contradictory postures can be reconciled is indicated in the Munqidh44 where the wording is: 'in it one finds most of their [the philosophers'] mistakes', and not: 'it is totally mistaken'. Note moreover that the principal accusation is now no longer expressed in terms of taqlid, but in terms of an over-confidence in one's capacity to offer apodeictic proofs in metaphysical matters. We have already come across this accusation in the Preface of the Tahdfut, in connection with the notions of superior quick wit and intelligence (above, p. 2). It is worthwhile noting that in 40

For further references see J. Janssens, '... his Use of Avicennian Texts' (n. 1, above). 41 Al-Ghazzall probably gave some thought to a synthesis between Avicennism and Ash'arism. See T. Mayer, review of R. Frank, al-Ghazzall and the Ash'arite school (Durham and London, 1994), in/, of Qur'anic Stud., 1(1) (1999), 170-82, esp. 177. 42 Having no direct access to Badawi's Arabic edition, I referred to the English translation by R. McCarthy, Freedom and Fulfillment (Boston, 1980), App. 2, 175-286. 43 Of course the idea of contingency is certainly not absent in Ibn Slna! 44 See the edition and French translation by F. Jabre (Beirut, 1969), 23 (Arabic), 78 (French).

X i6

q. 14, which turns around the affirmation that 'heaven is an animal obeying God', the accusation of bid'a (innovation) does not concern the affirmation itself, but exclusively the philosophers' pretension to have established its truth by demonstrative proof. Ibn Slna's theory of bads makes him a strong candidate for being an object, if not the object, of this particular criticism by al-Ghazzall. However, in the Shifd* and in the Aqsdm, Ibn Slna explicitly states that for such a specific religious matter as bodily resurrection no demonstrative proof can be given, and that it must then be accepted solely on the testimony of the prophet. If that is indeed his final position, it is one quite similar to al-Ghazzall's. But as soon as one looks more closely at how Ibn Slna really dealt with these matters, it becomes obvious that he took a rather different stance. In fact, Ibn Slna tried to do justice 'philosophically' to these religious articles of faith, which would seem to be in flagrant contradiction to the very fundamentals of the philosophical project. Hence, in the final analysis, he simply continued to subjugate Revelation to reason. In that sense, his philosophy undoubtedly remained unsatisfactory for al-Ghazzall.45 But, at the same time it is certain that for al-Ghazzall this same philosophy already contained within it the germs of the right solution—the elaboration of a real Islamic philosophy instead of a philosophy of Islam. As already mentioned, Ibn Slna was aware that no apodeictic proof can be given for all the fundamentals of the religious Revelation. He therefore agreed to renounce a few generally accepted doctrines in the /^/s^/^-movement, but did not go so far as to abandon any of the major fundamentals of its project, above all its acceptance of a causal determinism. To that extent, al-Ghazzall's attack is also directed against him. But may one add 'primarily'? I seriously doubt it. Is it not significant that al-Ghazzall, in the Munqidh,46 when alluding to the danger of philosophy, names not Ibn Slna, but the Ikhwan al-Safa'? It is they who, in citing the prophets, and by using mystical terminology, really create 'confusion', especially among the masses.47 This is evidently not the case in Ibn Slna. Of course, one could object that in q. 11 (131, §18), al-Ghazzall seems to be addressing Ibn Slna directly, when he states: 'as long as you agree with them in denying (God's) will ...'. As Ibn Slna was no longer living, one may read this direct address as a rhetorical device, accompanied by what could be considered a cynical undertone. If this interpretation is correct, then the description of the Tahdfut as an 45

See J. Michot, Destinee, 217-18. See the edition of F. Jabre, 27 (Arabic), 84 (French). 47 Surprisingly enough, al-Ghazzali here adopts the philosophers' distinction between the masses and the elite, which he criticizes in the Preface of the Tahafut. 46

X A L - G H A Z Z A L l ' S TAHAFUT

17

'ad hominem' argument against Ibn Slna can hardly be rejected. However, could it not be that al-Ghazzall is not addressing Ibn Slna himself so much as he is addressing the latter's disciples? Could it not be that he has them in mind when 'speaking to' the master himself—which of course is a rhetorical device? I think that it is at least possible to understand the fragment concerned in this way— the disciples are more likely than the master to be guilty of taqlld. In light of the evidence I have given above and elsewhere, it seems that the Tahafut is not necessarily an (explicit) anti-Avicennian work. I am aware that I have not offered conclusive proof against such a description of it, but I believe that I have put forward several arguments that encourage one to doubt it. I am therefore convinced that no clear-cut answer to whether the Tahafut is an anti-Avicennian work is possible at present. By making the question into a genuinely open one, I hope to have stimulated new research on the nature of the Tahafut and its place in al-Ghazzalfs 'oeuvre', and to have demonstrated that some of the established ideas about it are far from being obvious.

XI

Al-Gazzali, and His Use of Avicennian Texts

It is a well-known fact that al-Gazzali largely copied from some of his most famous predecessors, such as e.g., al-Qusayri and al-Muhasibi. Where al-Gazzali is dealing with philosophy, it is generally admitted that Ibn Sina certainly was his major source of inspiration. But precise indications of the presence of Avicennian texts in his writings, or of the way in which they were used by al-Gazzali are still very rare.1 To show this in a comprehensive way would undoubtedly be a major project in itself, which largely surpasses the limits of this present paper. Therefore, I wish to give only a few primary indications regarding the very presence of Avicennian texts in some of al-Gazzali's works. And even in doing so, I had to face some particular difficulties, as already mentioned in 1975 by Lazarus-Yafeh,2 i.e., the absence of critical editions and concordances for both al-Gazzali's and Ibn Sma's works. Moreover, al-Gazzali never gives any indication that he is basing his expose on a particular source-text. Finally, it may be said that al-Gazzali sometimes combines fragments taken from different texts, or even replaces the basic vocabulary of his source by a vocabulary which finds its origin in another current of thought.3 1 Regarding the identification of Avicennian texts in al-GazzalT, I may refer to my 'Le DaneshNameh d'Ibn STna: un texte a revoir?,' in: Bulletin de philosophic medievale, 28 (86), 163—177; and my: 'Le Mdarig al-quds ft maddrig marifat al-nafs: un element-cle pour le dossier GazzalT Ibn STna?,' in: Archives d'Histoire Doctrinales et Litteraires du Moyen Age, 60 (93), 27-55; for a doctrinal use of Ibn STna by al-GazzalT, I may cite: R. M. Frank: Creation and the Cosmic System: al-Gazzali and Avicenna, Abh. Heidelb. Ak. Wiss., Philos.-hist. Kl., Jahrg. 1992, Abh. 1, Heidelberg, Carl Winter Universitatsverlag, 1992.; and to: M. H. Abu Sacda: Al-atdr al-slnawiyafi madhab al-Gazzdlifi 'l-nafs al-insdmya, al-Qahira, Dar Abu Hurayba, 1990. " H. Lazarus-Yafeh: Studies in al-Ghazzdli, Jerusalem, The Magness Press, 1975, esp. 14 ff. 3 At first sight, the Mihakk al-nazar seems to entail such a change of vocabulary, i.e. from a philosophically to a juridical inspired one, when compared to the Mfydr al-ilm, without implying

XI

But let me now proceed by presenting a broad outline of al-Gazzali's use of Avicennian texts in those of his works, which include a clear philosophical component, i.e. Micydr, Tahdfut, Mlzdn. As to the former, which I have already treated in a previous publication,4 it is a slightly interpretative translation into Arabic of Ibn STna's famous Persian encyclopaedic work, i.e., the Ddnesh-Ndmeh, I am now in a position to add that - at least a part of - al-Gazzali's rewordings consist of, or are inspired by text-fragments, taken from other Avicennian, writings, esp. Isdrdt and Sifd'. Moreover, it may be noted that after a complete re-examination of the text I did not discover any direct reference to a Farabian text. But, in view of the vastness of the materials to be examined, it may be that such a reference has escaped my attention. But should this be so, it is of no major significance, since for more than 90% of the text I can posit its Avicennian source. Finally, based on the previous observation it becomes quite understandable that medieval Latin scholastics have identified alGazzali as a sequax Avicennae. When one has no other Gazzalian text besides the Maqdsid at one's disposal, one cannot but arrive at such conclusion! Let us now concentrate on the Mfydr al-cilm. This work divides into four "books", the first of which, entitled Muqaddimdt al-qiyds, basically deals with singular terms, the categories and propositions, while the second, although simply entitled al-Qiyds, not only treats - it has to be admitted into great detail - syllogistics, but also matters of demonstration, dialectics and sophistics. As to book three, its title: al-Hadd completely covers its contents. Finally, book 4 discusses the categories under the heading: Aqsdm al-wugud wa-ahkdmuhu. Hence, excepted for rhetorics and poetics, the Miydr offers a complete presentation of Aristotle's Organon, as it was conceived in the a real change of doctrine, at least for a large part of its major articulations. In a publication which I am now preparing on this latter work, I will try to deal more thoroughly with this matter. It may be worthwhile to mention also that the Mustasfd almost verbatim reproduces large fragments of the Mihakk. 4 See my "Le Ddnesh-Ndmeh.. " (cited n. 1). M. Bernard: 'al-Gazzali, artisan de la fusion des systemes de pensee', Journal Asiatique, 278 (90), 223-251, esp. 230 ff. has objected to my conclusion, i.e., that the Maqdsid is a slightly interpretative translation of the Ddnesh-Ndmeh, based on some doctrinal observations (mostly concerning the very notion of contingency, but, contrary to what she believes, this very notion is not absent in Ibn SIna, at least in the formulation which she characterises as typical for al-Gazzali), but also on the very fact that, according to her, not only the Danesh-Nameh, but also other Avicennian texts, besides Farabian ones have been used by al-GazzalT in the elaboration of his Maqdsid. As far as the former are concerned, her remark is correct, but it has to be said that one is mostly dealing with very small additions, which do not really affect the basic articulation of the text of the Danesh-Nameh. As to possible Farabian texts, I have not found any till now.

38

XI Al-Gazzall, and His Use of Avicennian Texts

Arabic tradition, i.e., including the Isagoge. However, it is striking that the first two books already present a basic outline of all the books of the Organon. Books 3 and 4, on the contrary, deal with very precise issues, such as definition, respectively the ten categories. In view of the importance of these issues, it is not really surprising that alGazzali has added particular developments on them, and, therefore, there is no reason to doubt this fundamental unity of the work. A confirmation for this latter unity can possibly be derived from al-Gazzali's use in all four books of the logical part of the Nagdt,5 or a work similar to the Nagdt.6 In b. 1-2, as well as in the first part of b. 3, one moreover finds important derivations from the Isdrdt, with some additions taken from the Sifd' and the Maqdsid. However, Ibn Sina's Kitdb al-hudud is only used in b. 3, and such more specifically at the end of the first part, and all along the second part. It has to be stressed that regarding this text, we have to do with an almost literal decalque. The small variants one may come across with respect to some of the definitions seem to be due to errors, either made by ancient copyists or contemporary editors. It is undoubtedly remarkable that even in the metaphysical definitions one cannot find any fundamental changes, although al-Gazzali is giving there additional explanations, indicating inter alia that some terms are used in thefalsafa in a sense different from that of their usual understanding in the kaldm. Regarding b. 4, besides the logical, also the metaphysical part of the Nagdt has functioned as a text-source, more specifically for its second chapter, whereas in the first chapter there are also present a few minor derivations from Sifd' (Maquldt) and Maqdsid. Hence, it is clear that the Mfydr is largely indebted to various writings of Ibn Sma. Nevertheless, the latter does not constitute the only philosophical source. At least two texts of his famous predecessor al-Farabi have been used, and largely copied in the present work, i.e., the Qiyds al-sagir (in b. 2), and the Maquldt (in b. 4, ch. I). 7 If still necessary, this formally does prove al-Gazzali's great familiarity with the writings of both great philosophers. 5

Unfortunately, I cannot offer here the details of this use, nor of the other texts I will later indicate, but, as already mentioned in n. 3, I am now preparing a specific publication on these matters. G. Hana: 'Zur Logik al-Gazalis,' in: ZDMG, Suppl. II. XVIII. Deutscher Orientalistentag (Liibeck, 1972), 1974, 178-185 already offers a survey of the book, with indication of its major sources. However, regarding these latter, the author's indications are rather imprecise, and, are indeed in need of serious correction and/or addition. 6 Several Avicennian works on logic are still no available in published form, as e.g., the alMuhtasar al-awsat fl 'l-mantiq. It may be that al-Gazzali had them at his disposal, and that he based his expose on them. 7 It may stated here that several texts of the Farabian logical corpus are no longer extant, or have not yet been identified in the manuscripts.

39

XI

Although I have only undertaken a rather preliminary survey of the possible sources of the Tahdfut, I can already affirm with almost absolute certainty that alGazzali, when exposing the views of the philosophers, has made use in an important number of cases of Avicennian formulations, derived from Sifd', Nagdt (or a text similar to it) and Isdrdt. It is also obvious that the wording of the Maqdsid in no way whatsoever forms the direct basis of any expose in the Tahdfut. It is perhaps not insignificant that al-Gazzali, in his Tahdfut, clearly does not simply identify Ibn Sina with "the philosophers", since he not only mentions explicitly the names of Plato, Aristotle and Galen, but, above all, presents both al-Farabi and Ibn Sina as the most faithful disciples of Aristotle, and his most original commentators. Finally, based on a few samples, I have the impression that al-Gazzali in the present work does not so much copy his sources, but rather by combining elements out of them creates a kind of "own synthetically summary". In the Mizdn al-camal, Ibn Sina no longer functions as the major source. This "honour" belongs here to Ragib al-Isfahani.8 However, the description of the soul, and its faculties, is directly based on the Nagdt/Ahwdl al-nafs? As to the presence of Avicennian fragments, or ideas in other writings of al-Gazzali, I cannot at the very moment make any serious pronunciation. However, there are evident indications for the presence of at least some Avicennian inspired ideas in some books of the Ihyd',l() as well as in some parts of the al-Maqsad al-asnd.11 But (almost) literal citations are, when, and if present, anyhow rare. One rather comes

8

See W. Madelung: 'Ar-Ragib al-lsfahanT und die Ethik al-GazalTs,' in: R. Gramlich (Hrsg): Islamkundige Abhandlungen. Fritz Meier zum sechzigsten Geburtstag, Wiesbaden, Fr. Steiner Verlag, 1974, 152-163; cp. also: H. Daiber: 'Griechische Ethik in islamischem Gewande. Das Beispiel von Ragib al-Isfahani (11. Jh.)', in: B. Mojsisch u. O. Pluta (Hrsg): Historia Philosophiae Medii Aevi. Studien zur Geschichte der Philosophic des Mittelalter, Amsterdam, Philadelphia, B. R. Griiner, 1991, 181-192. Recently, at the International Symposium 'Avicenna and his Heritage' (Leuven, September 1999) B. Karliga held a communication on a newly discovered ethical text, which he believes to belong to Ibn STna. If his hypothesis reveals to be correct, and he offered indeed good evidence for it, I cannot but admit that the Mizdn has also an outspoken Avicennian basis. 9 For a survey of the identical passages between the Nagdt and the Ahwdl al-nafs, see D. Gutas: Avicenna and the Aristotelian Tradition. Introduction to Reading Avicenna's Philosophical Works (I.P.T.S., IV), Leiden, New York, Kobenhavn, Koln, Brill, 1988, 100. 10 See S. Pines: 'Quelques notes sur les rapports de I'/hyd' al-din d'al-Gazzali avec la pensee d'Ibn STna,' in: Ghazdli. La raison et le miracle (Islam d'hier et d'aujourd'hui), Paris, Maisonneuve, 1987, 11-16; or my: 'Filosofische elementen in de mystieke leer van al-Gazzali,' Tijdschrift voor Filosqfie, 50 (88), 334-342, esp. 338 ff. 11 See R. M. Frank, op. cit. (n. 1), 16.

40

XI Al-Gazzall, and His Use of Avicennian Texts

across paraphrastic, or summarising formulations. Moreover, it is obvious that in these latter "religious-mystical" works al-Qusayri becomes the pre-dominant source. From the thus far given facts, I believe the following conclusions may be drawn: 1. al-Gazzali's use of Avicennian texts, or ideas, is not completely limited to his so-called "philosophical" works; 2. Even if one often finds literal, or almost literal citations, one cannot but recognise that in many instances al-Gazzali is adapting his "source", either by combining passages taken from different works; or by introducing additions, e.g, in the form of explanations; or by recurring to a different type of vocabulary (in all these cases only a careful and very detailed examination can show whether al-Gazzali remains fundamentally faithful to his source, or not); 3. Although al-Gazzali also refers to other Avicennian treatises, he, above all, uses Ibn Sina's four major encyclopaedic works, i.e., Sifd', Nagdt (eventually a work, or works, which show(s) close similarities with it, or a part of it), Isdrdt and DdneshNdmeh (most of the time in its Maqdsid-version). If these three (provisory) conclusions are taken for granted, it becomes very difficult to deny that such a writing as the Macdrig al-quds shows outspoken "Gazzalian" characteristic. As I have posited elsewhere,12 it contains passages taken from, or references to the four mentioned major encyclopaedic works of Ibn Sina. As to the Ddnesh-Ndmeh, it is cited in the version of the Maqdsid. Moreover, references to other Avicennian works are also present, i.e., Maqdlafi 'l-nafs, Mubdhatdt and Commentary on the Theologia Aristotelis. Besides literal, or almost literal citations, one finds reworked versions, as well as additions. Therefore, the way in which Ibn Sina's texts are used, seems to offer a supplementary argument, to those I have already formulated in this paper, in order to accept the attribution of the work to al-Gazzali as being genuine, or, at least, as very probable. After these basic considerations, I would like to deal now with a very specific, although undoubtedly very delicate problem. I mean: is it possible to make al-Gazzali's use of Avicennian texts a (among others!) relevant criterion in deciding the chronological order of the former's works? A priori, there does not exist any strong counter-indication for this supposition. But, of course, it is neither self evident. In order to find out whether it has any chance to be a workable hypothesis, I have decided to concentrate on one particular Avicennian doctrine, and to look at al-Gazzali's formulations of it in different of his works. The doctrine of the inner senses almost immediately revealed itself as a good candidate. Since Ibn Sina's own wording 12

See my 'Le Mcfdrig...' (cited n. 1).

41

XI

of it already shows some minor, but clear variations from one work to another, it is not too difficult to check whether al-Gazzali invariably reproduced always the very same wording, or, on the contrary, made use of several of these Avicennian wordings. Moreover, this doctrine is presented by al-Gazzali in such relevant works as the Maqdsid, Tahdfut, Mizdn and Ihyd', and it is also present in the Mcfdrig. In the Maqdsid,^ as may be expected in view of my former affirmations, Ibn Sina's version of the doctrine, as presented in his Ddnesh-Ndmeh, guides the basic articulation; and several fragments reveal to be perfect translations of the Persian. But, instead of showing this into its smallest detail, it is perhaps more important to note that the denominations for the five inner senses, which al-Gazzali is using here, are identical with these of the Ddnesh-Ndmeh, i.e., al-hiss al-mustarik, al-mu(ta?)sawwir, al-wahm, al-mutahayyil, and al-ddkira. Although al-Gazzali inverted the order of discussing the latter two, he remains faithful to their explanation by Ibn Sma in the Ddnesh-Ndmeh. As to other elements different from, or not present in the DdneshNdmeh, it has to be observed that: 1. the explicit idea that the common sense enables one to combine sensations related to different outer senses, is clearly Avicennian inspired, and may have been related to Sifd', DeAnima, p. 164 or Isdrdt, p. 124,14 2. with respect to the faculty of representation, two examples are added by al-Gazzali. The one of the water is present in the Nagdt (R., 3\)/Ahwdl, p. 63,15 although I could not find a direct source for that of the wax (even if this latter reveals to be a personal addition by al-Gazzali, it nevertheless remains completely inside the "philosophical sphere"),16 3. the affirmation that the destruction of the former part of the brain will lead to the oblivion of all forms is either inspired by Sifd', De Anima, p. 167, L. 9-10 or by Isdrdt, p. 125, 1. 11-12;

13

The precise reference is to part III, M. 4, q. 3. I have made use of the al-Qahira, 1961, ed. by S. Dunya, 356-357. 14 For Sifd', De Anima, I am referring to the London, 1959, ed. by F. Rahman; as to the Isdrdt, references are always to the Leyde, 1892, ed. by J. Forget. 15 Regarding the Nagdt, I have preferred to refer to the excellent English translation of F. Rahman: Avicenna's Psychology, London, 1952. Repr. Westport, Conn., Hyperion Press, 1981, 2 1990; as to the Ahwdl al-nafs, references are to the al-Qahira, 1952, ed. by Al-Ehwany. 16 In this respect, one may compare the Fusus, § 33 (Ed. F. Dieterici, Leyde, 1890 in: al-Farabi: Philosophische Abhandlungen, 66-83). For the probably Avicennian origin of this work, see: S. Pines: 'Ibn SIna et 1'auteur de la Risdlat fusus fi 'l-hikma,' in: Revue des Etudes Islamiques, 19 (51), 121-124.

42

XI Al-Gazzali, and His Use of Avicennian Texts

4. the placement of the faculty of estimation, together with that of memory in the back part of the brain, has probably to be understood in a loose way, or may eventually be due to some carelessness in al-Gazzali (ace. to Ibn Sina its place is always the far end of the middle ventricle of the brain); 5. as to imagination, the text of the Maqdsid reveals to include some replacements with respect to that of the Ddnesh-Ndmeh, but, above all, specifies very explicitly its nature as three-fold, i.e., motive, rapid in translation, and imitative. Whereas for the former two, there do exist indications in the Danesh-Ndmeh, none is present regarding the latter. But cUyun, p. 3917 makes it clear that also for this idea an Avicennian origin may not be excluded, and is even probable. Notwithstanding the presence of all these (small) deviations from the original Persian text, one cannot but recognise that the wording of the Maqdsid is highly dependent on the latter. There is no single passage, which reveals a personal intervention by al-Gazzali, that implies a fundamental change in the basic understanding of the original Persian expose. At most, al-Gazzali elucidates on some obscure passages by either slightly rephrasing, or by adding examples. And even then, the tonality remains Avicennian in a most evident way, and, not seldom direct indications in other Avicennian writings than the Ddnesh-Ndmeh, for these "rewordings" are present. Therefore, I see no valid reason why I should change my previous valorisation of the Maqdsid as a slightly interpretative translation of the Ddnesh-Ndmeh. I now even wonder whether the work should not be ranked among the studies belonging to the category "tacllqa", i.e., a writing showing a close connection with the works, or the teaching activities of the "master".18 Of course, in the case of al-Gazzali there cannot be any reference to Ibn Sina's teaching activity. But should it be excluded that he, being a student at the Avicennian-inspiredya/5-o/a-school of his days, elaborated a kind of advanced student's work while translating into Arabic a difficult text, and, at once, offering a basic elucidation, albeit still in terms close to that of the master? But is this not an overhasty, and, indeed, fanciful conclusion? Has al-Gazzali, in the very introduction to the Maqdsid, not emphasised that he intends no more than to give a "neutral" presentation of the main ideas of the adherents ofthefalsafa? Hence, caution is clearly required. Without an examination of the Tahdfut, no serious conclusion seems to be possible.

17

18

The reference is to the al-Qahira, 1954, ed. by A. Badawi.

See G. Makdisi: The Rise of Colleges. Institutions of Learning in Islam and the West, Edinburgh, 1981, 114-125.

43

XI

Let us then now concentrate on the doctrine of the inner senses, as presented in the Tahdfut, q. 18.19 It immediately becomes apparent that the basic terminology regarding the five inner senses is not the same as that of the Maqdsid. The first inner sense is referred to by two different denominations, i.e., al-haydliyya and al-hiss almustarik. Whereas the latter appellation is still explicitly linked with the assembling of forms derived from different outer senses, the former is presented by al-Gazzali as the expression of the very reception of the impressions by the outer senses. After this first inner sense the faculties of estimation and imagination are discussed, and only hereafter the two "conserving" faculties are brought to the fore. This change in order has however no doctrinal consequences. But on the terminological level, one has once more to do with an important innovation, insofar as the faculty of representation is now designated by the notion of hdfiza, which in Ibn Sina is always used besides ddkira for indicating the faculty of memory. In all this, al-Gazzali appears to take some liberty with his source. Although he does not abandon the Avicennian doctrine, nor terminology - al-haydliyya might be a more "Arabic" expression for the strange bdntdsid, a transliteration of the Greek fantasia -, al-Gazzali clearly does not feel obliged to remain faithful to the Avicennian wording. This becomes even more evident, when one finds (almost) no literal citations.20 However, this does not mean that the present text is based on a vague Avicennian inspiration. On the contrary. Its major part is clearly based on the De Anima of the Sifd', esp. b. IV, c. 1 (combined with a few elements derived from b. I, c. 5, as well as from Nagdt/Ahwdl and Maqdsid). Furthermore, it is worthwhile to mention that al-Gazzali, when he completed the expose on the soul in the beginnings of this question 18 of the Tahdfut, stated, or even insisted that this philosophical doctrine of the soul is not open to rejection, since it is based on facts one may observe, and that it contains nothing contrary to Revelation.21 Precisely this latter assessment makes it difficult to understand why al-Gazzali does not repeat, or, at least, refer to what he has already brought to the fore on these matters in the Maqdsid, if the latter has been meant as an introduction to the former? If there does exist only a small interval of time between the two works, why then has al-Gazzali taken another Avicennian text as basis, and, above all, why has he introI have used the al-Qahira, 1972, ed. by S. Dunya, as well as the Lahore, 1963 English translation by S. Kamali (which has to be used with caution). "° In view of the limited nature of our present examination, I want not to overemphasise this fact, but the Maqdsid undoubtedly remains much more faithful to its source than does the Tahdfut. 21 cf. Tahdfut, 256 (Kamali, 202). In fact, al-Oazzali only rejects the claim of the philosophers that they on solely rational grounds can demonstrate the existence of the soul as a spiritual substance.

44

XI Al-Gazzall, and His Use of Avicennian Texts

duced terminological innovations? If the reader is not an expert in philosophy, will he not be confused by this "different" presentation? It is obvious that we are here dealing with a most fundamental interrogation. Hence, a simple answer cannot be given. Therefore, I want to stress that this delicate matter cannot be settled, based on the sole acceptance of al-GazzalT's statements in the introductions to both works. They may have been written after the works were finished, or even in a still later period.22 But it neither suffices to point to the close connection between Maqdsid and Ddnesh-Ndmeh in order to formulate a definite answer in the negative. What can, and must be conceded, that it is an open question, and that it no longer is an evidence that the Maqdsid have been written in the perspective of the Tahdfut. But, how does al-Gazzali formulate the doctrine of the inner senses in his other works? If one looks at the Mizdn,2* it uses, instead of the Sifd', the Nagdt/Ahwdl-version as its basis. However, as far as the designation of the inner senses is concerned, the very same technical terminology is used. So, here one finds also the notion of al-haydliyya, while the faculty of representation is once more referred to by the notion ofal-hdfiza. Worth mentioning is moreover the statement that the faculty of imagination is rather to be linked with the motive than with the perceptive faculties. As far as I can see such an affirmation is missing in Ibn Sina, but it is present in the Tahdfut. Hence, one cannot but conclude that the Mizdn offers a very similar, albeit not identical version of the Avicennian doctrine of the inner senses. It therefore is most probable that they should be dated in approximately the same period. But one may wonder which of the two preceded the other, and whether the change in basic text has, or has not some deeper doctrinal significance? At present, even more, I prefer to leave these questions open. In the Ihyd',24 more precisely near the end of the Section on the Soldiers of the Heart in book 21 fAgd'ib al-qalb), al-Gazzali very briefly evokes the description of "" G. Hana: 'Die Hochscholastik um eine Autoritat armer', in: Festschrift fur H. Heimpel zum 70. Geburtstag am 19. Sept. 1971 (Veroff. M. Planck Inst. Gesch., 36, II), Gottingen, Vandenhoeck und Ruprecht, 1972, 884-899, 890-894 shows that the introduction to the Maqdsid is almost verbatim present in the Munqid. Therefore, Hana believes that a later author has written this introduction, and he even seems inclined to seriously doubt the attribution of the Maqdsid to alGazzali. On this latter point, I cannot but disagree with him, since al-Gazzali clearly uses the Maqdsid in several of his works. As to the former, he may be right, but it is also may be possible that al-Gazzali has written this text before he started to write his Munqid, and that he has copied it in this latter work. 23 1 have made use of the Beirut, 1989, ed. by A. Sams al-DTn, and in addition the al-Qahira, 1328 h., as well as the al-Qahira, 1964, ed. by S. Dunya.

45

XI

the five inner senses. The reader immediately notices that the first sense is now called: al-hayal, and not al-haydliyya. Note that the former notion is only used by Ibn Sina as another expression for the faculty of representation. However, since the Ihyd' is a work intended for a larger audience, al-Gazzali perhaps prefered to avoid an unusual terminology, such as al-haydliyya, and replaced it by the more common al-haydl, which are both notions belonging to the same family. Furthermore, regarding the faculty of imagination, only the designation of al-tafakkur is mentioned. But since alGazzali in the present context is clearly dealing with the human soul, this is most understandable. However, astonishing, and even problematic is the omission of the notion of wahm. Does this imply that the faculty of estimation has simply been excluded? I do not think so. In fact, al-Gazzali states: "Then the totality of the meanings of the sensed things, (existing) in his (I believe this is referring to the one who senses) imagination (haydl, which here seems to be used in a rather loose sense), are assembled by the sense, which combines between the sensed things." Insofar as there is an explicit reference to "meanings", it is almost impossible to understand this passage as referring to the common sense in its usual acceptance. Therefore, if at least my understanding of the text is correct, I believe that al-Gazzali by this "combining sense" is vaguely evoking the faculty of estimation. It may be added that al-Gazzali, after the previous statement, when giving a recapitulating enumeration of the inner senses, does not say: al-hiss al-mustarik, but: hiss mustarik. Moreover, when offering a similar list in the "introduction" to the Kimiyd-i sacdda,25 al-Gazzali uses the notion of quwwat al-ta\vahhum, instead of hiss mustarik. Since the Kimiyd in all probability is from a later date than the Ihyd' - especially the introduction, which in fact reveals to be a reworked version of the Book of the Marvels of the Heart of the Ihyd', appears in view of its outspoken mystical undertone to belong to a much later period than the concerned book of the Ihyd'26 - it can be safely af24 In view of the absence of any critical edition of the Ihyd', which sharply contrasts with its many editions, I have preferred to give no page-reference, but to indicate the passage by the title of its section, so that the reader, whatever edition he has at his disposal, can easily find the designated location. 25 Imam M. Gazzali: Kimiyd-i sacdda. Introd. by M. cAbbasT. Tehran, 1361 h. s., 52. 26 Generally speaking the Kimiyd shows the same basic structure as the Ihyd', but the order of the first two books has been inverted; b. 28 of the Ihyd' has been split up into two books in the Kimiyd; and b. 35-36, and 37-39 of the former become b. 38-39, resp. b. 36-37 of the latter. But,

46

XI Al-GazzalT, and His Use of Avicennian Texts

firmed that al-Gazzali never abandoned the basic Avicennian doctrine of the inner senses. But how can then that unusual designation of the faculty of estimation be explained? It is tempting to simply ascribe a "scribal" error, either of an ancient copyist, or of the contemporary editor. But thus far I have not found any serious indication which would give substantial support to this conjecture. Therefore, I have tried to find another explanation, and the only one which I can see is that al-Gazzali has consciously avoided the use of a terminus technicus in a work, destined to a very large public. It is perhaps not unimportant to note that the very notion ofwahm in its popular sense always includes a connotation of error, although this connotation is totally absent in Ibn Sina's technical use. Hence, the brief evocation of the inner senses in the Ihyd' has no direct chronological significance, but draws our attention to the importance of the intended audience, when one is dealing with al-Gazzali's works. However, the Book of the Marvels of the Heart has some clear affinities with the Mizdn. Several of its baydn-s, i.e., 3-4 and 7-9 are almost verbatim the same in the latter, i.e., I. 3 = M. 10; I. 4 = M. 4, last part; I. 7 = M. 26; I. 8-9 - M. 7 (in both cases the references are given to the respective baydn-s).21 Hence, it is quite natural to wonder whether it has been written in the very same period, and which of the two is the earlier one? In this respect, the Macdrig may offer some elements of clarification, provided it was written by al-Gazzali himself. Since, at most, I can speak of a strong probability, but surely not of an absolute certainty, as was mentioned, what follows thus only has an indicative value, but I do believe that it is not deprived of any value. As I have shown in an earlier publication,28 it contains chapters, similar to, or almost identical with baydns of both Ihyd', Book of the Marvels of the Heart (Ma. 1= I. 1, but modified; Ma. 15 = I. 6; Ma. 16 = I. 3) and Mizdn (Ma. 14 = M. 16 and Ma. 16 = M. 10). In this list, one discovers the presence of baydn 3 of Ihyd', B. 21. Above, we saw that this baydn was also present in the Mizdn. When we now compare the latter's version with that of the Macdrig, and of the Ihyd', it is remarkable that it, as the former, above all, b. 21 of the Ihyd' has disappeared at the very beginnings of the third part of the Kimiyd, but its major ideas are present in the long introduction of the latter work. Undoubtedly a systematic comparison of both texts will afford important elements regarding an eventual doctrinal evolution in al-Gazzali. 27 It has already be mentioned by several authors that large parts of the Mizdn are also present in the Ihyd', but as far as I can see no systematic account has ever been given in this respect. I now prepare such a systematic account, and I may herewith affirm that the major part of the Mizdn is in fact also present in the Ihyd' (at least in some of its books). 28 See my 'Le Mcfdrig...' (cited n. 1 ), 39-41. This study was based on the Beirut, 1981, ed. by the Revival of Arabic Culture Committee, and the Beirut, 1988, ed. by A. Sams al-DIn.

47

XI

misses the first paragraph of the latter. In that paragraph al-Gazzali states that the irascible and concupiscible forces have to submit themselves to the heart. However, it has to be specified that the heart hereby is defined in rather philosophical than mystical terms, i.e., as being science film), wisdom (hikma) and thought (tafakkur). But, notwithstanding this striking resemblance, the account of the inner senses in the Macdrig is not identical to that of the Mlzdn. It offers almost literally the complete text of the Remark on the Perceptive Faculties,2** although changing somewhat its order, and introducing several additions. These latter can always be conceived inside an Avicennian framework, and one of them, i.e., the addition introduced with respect to the faculty of estimation, has a direct Avicennian source, i.e., Sifd', De Animal It may also be mentioned that the common sense is referred to by the designation of alhiss al-mustarik, or bdntdsid. The presence of the latter notion, seems to situate the work before the introduction of al-haydliyya, and, hence, before Tahdfut and Mlzdn. As to Ihyd', B. 21 it looks as if its final redaction is posterior to both Mlzdn and Mcfdrig. Insofar as the latter work is still faithful in several of its passages to the letter of its Avicennian source, but, at the same time, includes some religious-mystical inspired items, I am inclined to believe it to be posterior to the Maqdsid, since there being elements of the latter present in the former.31 I will now try to formulate some final conclusions. The Maqdsid appears to be an outspoken "Avicennian" work, which, except for the introduction, offers no indication of any disagreement with its philosophical theses on the side of al-Gazzali. Moreover, it cannot but be affirmed that it hardly offers an introduction to the Tahdfut, especially on the terminological level. I therefore am inclined to consider it a work written by the young al-Gazzali in his student days. Hence, this young al-Gazzali was probably an adept of the (Avicennian inspired) falsafa-school of his time. I am aware of the fact that I have not given definite proof for this, but, nevertheless, I consider it to be a valuable hypothesis, worthy of further examination. As to the Tahdfut and the Mlzdn, they may be dated around approximately the same period, and this is also true of the Macdrig, at least if it reveals to be an authentic Ghazzalian work, for which there exist strong indications. It may be observed that whereas the Mlzdn offers an outline of a philosophical ethics, although somewhat modified into a religious-mystical sense, the Macdrig, in a very similar way, offers an outline of a philosophical psychology, once more modified into a religious-mystical sense. I may even add that also the Mfydr enters easily into such a configuration, ™ Isdrdt, 123-125. ™Sifa', DeAnima, 166, 1. 17-168, 1.3. 31 See once more my 'Le Macdrig...\ 42. 48

XI Al-Gazzall, and His Use of Avicennian Texts

offering a philosophical logic, but modified according to religious-legal requirements. It is striking that in all these matters, contrary to those discussed in the Tahdfut, alGazzali accepts basic philosophical theories, by judging them not contrary to Revelation. I therefore wonder whether he did not write all three (or only two?) works before his elaboration of the Tahdfut? If such is the case, in view of the use of Avicennian texts, the Mizdn appears to be the best candidate for preceding immediately the latter, while the Macdrig probably has to be identified as the earliest of them. At once, it becomes a crucial question what exactly al-Gazzali's intention was in writing the Tahdfut? And once again, I must stress that it is not enough to repeat the words of his introductions! As to the Ihyd', its final redaction, may well date from al-Gazzali's period of seclusion, but, at least for some of its books materials of a much earlier period have undoubtedly been included. And hence, one may wonder what was the exact impact, or even the precise sense of his so-called "crisis"? I had preferred to have been able to give some more answers, but the problems regarding al-Gazzali's work and thought are so numerous and so delicate, that I am already glad to be able to offer some sensible hypotheses, and questions.

49

XII

BAHMANYAR IBN MARZUBAN: A FAITHFUL DISCIPLE OF IBN SlNA?

Bahmanyar ibn Marzuban (d. 458/1066) is known as one of Ibn Sfna's first generation students.1 He was clearly involved in a direct correspondence with the Sayh ar-Ra'fs, which became part of what is known as the Mubahatdt? One of the participants in this correspondence was also Abu 1-Qasim al-Kirmani, of whom we know very little,3 except that he was much older than Ibn Sfna and much despised by him. From the Mubahatat, it is obvious that Bahmanyar and Abu 1-Qasim were in close contact with each other. Since the former was undoubtedly much younger than the latter, it might be that he chose Abu 1-Qasim as his tutor.4 Ibn Sfna explicitly regrets that Bahmanyar sometimes seems to take side with this minor thinker, and does not hesitate to reprehend Bahmanyar for such attitude.5 Hence, the question may be raised whether Bahmanyar was a faithful

1 For a detailed account of both ancient and contemporary reports regarding Bahmanyar, especially his origin, life and relationship with Ibn Sma, see D.C. Reisman, "The Making of the Avicennan Tradition: The Transmission, Contents, and Structure of Ibn Sfna's al-Mubdhatdt (The Discussions)" Ph.D. dissertation, Yale, 2001, 310-322. 2 For a precise account of the material related to Bahmanyar (and Abu 1-Qasim al-Kirmani), see Reisman, "Avicennan Tradition," 339-393, as well as 404-410. 3 J.R. Michot, "Une nouvelle ceuvre du jeune Avicenne, note complementaire a propos du ms. Hiiseyin felebi 1194 de Brousse," Bulletin de philosophie medievale 34 (1992), 138-154, 148-150, inclines, but with caution, to identify this Abu 1-Qasim with a secretary of Rayy, called Abu 1-Qasim al-Katib by at-Tawhfdf. See also id. (as Yahya Michot), Ibn Sma, lettre au vizir Abu Scfd, editio princeps d'apres le manuscrit de JBursa, traduction de I'arabe, introduction, notes et lexique, Sagesses musulmanes, 4 (Beyrouth: Les Editions al-Bouraq, 2000), 20. 4 See Reisman, "Avicennan Tradition," 321-322. 5 Ibn Sma, Kitdb al-Mubdhatdt, ed. M. Bidarfar (Qum: Intisarat-i Brdar, 1992), 57, § 52; 63, § 75; and 74-76, §§ 113-116; French translation by J. Michot, "La reponse d'Avicenne a Bahmanyar et al-Kirmanf, Presentation, traduction critique et lexique arabe-francais de la Mubdhatha III," Le Museon 110.1-2 (1997), 143-221 (with reference to the paragraph numbers of Bldarfar's edition).

XII 178 disciple of Ibn Sina, or, on the contrary, did he distance himself from his (major?) teacher in philosophy? When looking at his opus magnum, i.e., Kitdb at-Tahsllf written in all probability between Ibn Sfna's and his own death, one initially has the impression that it is a pure compilation of Avicennan texts, Bahmanyar's own role being limited to the actual choice of the combined fragments. Such an impression is only reinforced when one reads the following at the very beginning of the work, immediately after the dedication of the work by Bahmanyar to his maternal uncle Abu Mansur Bahrain ibn Hursld Yazdyar:7 (I have composed) a book of the wisdom (hikmd) which the Sayh arRa'fs AbO cAli al-Husayn ibn cAbd Allah ibn Sfna—may God have mercy upon him—has taught,8 by imitating the order of the Hikma al'Ala'i? by taking intentions (al-macdm) from the totality of his writings and from what happened through discussion (muhawaratari) between me and him, and by adding what my understanding (na^an) has acquired as derivations (min al-furuc) [while] being analogous to the fundamentals (al-usul). Your looking into his books will furnish you the proof of these derivations.10 This opening statement rings distinctly pro-Avicennan. First, Bahmanyar stresses that the wisdom which he will develop in his work is the very same as the one Ibn Sfna has taught, clearly suggesting that this kind of wisdom was particular to the Sayh ar-Ra'fs himself, in 6 All of my references will be to the edition^ of M. Mutahhan, Second printing (Tehran: Intisarat-i Danisgah-yi Tihran, 1375S/1996). In his edition, Mutahhan often indicates source passages from different parts of the Sifd\ He also occasionally refers to the Nagat or to al-Isdrdt wa-t-tanbihdt. Despite a few (minor) mistakes, these references are generally trustworthy and cover a great deal of Bahmanyar's text. However, since I have found two chapters of the Ddnisndmah, although in Arabic (see infra), it seems possible, not to say probable, that Bahmanyar also drew on that work, and in a systematic manner. I plan a more thorough examination of this issue for a later publication. 7 I follow with Reisman, "Avicennan Tradition," 316, n. 80, the reading of the Aleppan manuscript. 8 D. Gutas, Avicenna and the Aristotelian Tradition: Introduction to Reading Avicenna's Philosophical Works, Islamic Philosophy and Theology: Texts and Studies, 4 (Leiden: EJ. Brill, 1988), 11, n. 14 offers another possible translation, based on the variant reading lubdb instead of Kitdb: "the gist of the philosophy which Avicenna revised." 9 The work is commonly known as the Ddnisndmah. The three parts of the work were published in the series Silsilah-yi Intisdrdt-i Anguman-i Atdr-i Milli, Tddgdr-i gasn-i hazdrah-yi Abu 'All Sind, Collection du millenaire d'Avicenne (Tehran: Danisgah-yi Tihran, 133IS/1951): Risdlah-yi Mantiq, ed. M. Mu c m and M. Miskat, vol. 12; Ildhiydt, ed. M. MucTn, vol. 15; Tabl'iydt, ed. M. Miskat, vol. 13. 10 Bahmanyar, at-Tahsil, 1.

XII

BAHMANYAR: A FAITHFUL DISCIPLE?

179

other words, that the latter has really innovated a very new way of philosophizing. Then Bahmanyar evokes what looks like one of the basic expressions of this new way, i.e., a special ordering of the main parts of the philosophical project as adopted by Ibn Sfria in his Ddnisndmah. Although in most of Ibn Sfna's encyclopaedic works the classical order (logic-physics-mathematics-metaphysics) is indeed respected, this is not the case in his Persian work, which may well be his very last philosophical encyclopaedia.11 There he adopts an evidently new order, i.e., logic-metaphysics-physics. It is striking that there is no mention of mathematics,12 but more interestingly, and contrary to the classical curriculum (in its Platonic-Aristotelian line), metaphysics precedes physics. It is therefore not of secondary importance that Bahmanyar explicitly claims to follow this unique order. In doing so, he seems to valorize it as a major contribution of Ibn Sfna's (mature) thought. This is all the more remarkable since he further claims not to offer a slightly reworked version of the Ddnisndmah itself, as later will be the case with al-Gazalfs Maqdsid^ but rather to have dealt with the totality of the writings of the Sayh ar-Ra'fs and even to have taken into account those ideas which he has learned through direct communication with the Master.14 The latter affirmation clearly implies that Ibn Sfna has not written down all of his ideas. But is Bahmanyar not offering himself a pretext in order to introduce personal ideas, having no Avicennan basis at all? In other words, does this claim not constitute at once both a simple strategic

11 See my "Les Tcfhqat d'Ibn Sma, Essai de structuration et de datation," in Langages et philosophic, Hommage a Jean Jolivet, Etudes de philosophic medievale, 74, ed. A. Elamrani-Jamal, A. de Libera, and A. Galonnier (Paris: J. Vrin, 1997), 109-122, 122 (and n. 39). Let me remark that I do not claim that the Ddnisndmah is Ibn Slna's last encyclopedic work, but rather that this simply may be the case. To fix in a definite way the relative chronological order of Ibn Sfna's works, much systematic research, based both on external and internal evidence, has still to be done, and therefore I refrain from any strong affirmation. 12 Gutas, Avicenna, 113, observes that Ibn Sma always omits in his later works the mathematical part, probably since he considers its traditional expose to be basically sufficient, and, hence, not open to any serious conflict. 13 See my "Le Ddnesh-Ndmeh d'Ibn Sma: un texte a revoir?" Bulletin de philosophie medievale 28 (1986), 163-177, especially 168-175. 14 Although Bahmanyar's statement can be understood as pointing to an oral communication, it may equally refer in a loose sense to any form of communication, either oral or written. If the latter is the case, Bahmanyar may be alluding to his correspondence with Ibn Sfna included in the Mubdhatdt materials.

XII 180 and apologetic move? His final affirmation clearly contradicts such an interpretation. He admits that he will present some ideas, not present in the Avicennan corpus, but he immediately adds that such additions are always derivative of principles the basis of which any serious thinker can easily detect in that very same corpus. Hence, at first sight, Bahmanyar's Kitdb at-Tahsil appears as a basically Avicennaninspired summa. However, a somewhat closer examination of the work immediately elicits a serious qualification of such characterization. In fact, the very basic structure of Kitdb at-Tahsil is only apparently identical with that of Ibn Sfna's Danisnamah. Surely, Bahmanyar still maintains a tripartite division, starting, as did Ibn Sfria, with logic, and then continuing with metaphysics. But, contrary to the Sayh ar-Ra'is, the latter is not designated as "Divine science" (cilm ildhl], but as "metaphysics" (ma bacd at-tabicd). Although this Aristotelian appellation is not necessarily un-Avicennan,15 there is no doubt that Ibn Sma himself largely preferred the expression al-Ilahiyat when dealing with the highest science of philosophy, this designation being the one found in all his later major works. Note that this does not mean that the proper "subject" of metaphysics was, according to Ibn Sma, the "divine." Not at all! Since the existence of God has to be proved, He cannot be its primary subject. For Ibn Sma only "Being qua Being" can fulfill such a role. So, metaphysics is first of all an ontology. But it also includes an aetiology (or: archaeology), since it deals with the principles of all the other sciences, and, moreover, it entails a theology, God being the highest and, hence, most noble Being. Ibn Sma integrates these three elements, which were present but in a scattered and fragmentary way in Aristotle's Metaphysics, into one coherent metaphysical system.16 Bahmanyar, however, seems to limit the metaphysical project to the premisses needed for the totality of the sciences.17 This might explain why he favors the Aristotelian appel15

The expression ma bcfd at-tabica is, for example, mentioned in Ibn Sma, alMabda3 wa-l-macad, ed. CA.A. Nurani, Wisdom of Persia Series, 36 (Tehran: Tehran University, 1363S/1984), 1. 16 See G. Verbeke, Avwenna, Grundleger einer neuen Metaphysik, Rheinisch-Westphalische Akademie der Wissenschaften, Vortrage G 263 (Opladen: Westdeutscher Verlag, 1983), 9—10. The basis for such an integrated vision was already present in alFarabr, more specifically in his treatise Fi Agrdd ma bacd at-tabica, as recognized by Ibn Sma himself in his autobiography; see W. Gohlman, The Life of Ibn Sina, A Critical Edition and Annotated Translation (Albany, New York: SUNY, 1974), 34-35. 17 Bahmanyar, at-Tahsil, 2.

XII

BAHMANYAR: A FAITHFUL DISCIPLE?

181

lation of "metaphysics." But since this affirmation is part of a general introduction, one should not over-emphasize it. A more systematic examination of the contents of this second part of Kitdb at-Tahsil is undoubtedly required for a correct evaluation of the relationship between Bahmanyar's "Metaphysics" and Ibn Sfna's al-Ilahlyat. Before working out a more detailed survey of the metaphysical part of Kitdb at-Tahsil, attention has to be paid to the precise title Bahmanyar gives to the third part of his work. One might naturally expect him to designate it simply by "physics." This is not the case. In fact, Bahmanyar proposes a quite different appellation, i.e., alc ilm bi-ahwdl dydn al-mawguddt, "the science of the states of the most noble of Beings," or, perhaps better, "the science of the states of the essences of the existing Beings." That the latter translation is perhaps to be preferred reveals itself as soon as one takes into account Bahmanyar's further division of this section into two parts. The first deals with "the Necessary Existent in Himself and the enumeration of His attributes" (fi mcfrifat wdgib al-wugud bi-ddtihi wa-ihsd* sifatihi\ while the second treats of "the caused Beings" (al-mawguddt al-maclula).18 Hence, the third part of Kitdb at-Tahsil encompasses the study of the totality of Beings, the divine Being as well as all created Beings. Whereas the examination of the latter is a priori not to be excluded from the domain of what Ibn Slna, with the Classical tradition, called "Physics," this is not the case with respect to God. For Ibn Sma, the existence of God can only be proved in the science of Metaphysics19 and thus His essence and attributes can only be the object of research in this science. As for Bahmanyar, the very structure of the third part of his major work gives, at least at first sight, the impression that he, contrary to Ibn Sma, does not want to place the study of the "divinalia" wholly outside the scope of physics proper. When Bahmanyar states in the introduction20 that "the first subpart (maqdld) indicates (fi daldla) the Existent Being (al-mawgud] who

18

Ibid., 569,/esp. 583. Ibn Sma, Sarh Kitdb Harfal-Ldm, ed. CA.R. Badawf in Aristu cinda l-carab, Second printing (Kuwait: Wikalat al-Matbucat, 1978), 23-24, where he states that it is inappropriate, as Aristotle and the Commentators have done, to argue to the First Truth from motion (the passage has been translated into English by F. Zimmermann in E. Booth, Aristotelian Aporetic Ontology in Islamic and Christian Thinkers, Cambridge Studies in Medieval Life and Thought, 3, Series vol. 20 [Cambridge, Cambridge University Press, 1983], 109-110). 20 Bahmanyar, at-Tahsil., 2. 19

XII 182 has [for His existence] neither secondary nor primary cause (attach Id sabab lahu wa-la cilla), and clarifies (wa-flhd l-ibdna) the ultimate goal (al-garad) of the TTieologia21 and of the book (maqala) Alpha elation"2*2 one may still hesitate somewhat in discerning his real intention. To speak of God in terms of "uncaused" rather than "unmoved" may reflect an Avicennan inspiration, although it may also directly derive from the Theologia, where God is primarily designated as the First Cause (al-cilla al-uld}23 As for the reference to Alpha elation, it is equally somewhat ambiguous. Insofar as Ibn Sfna's proof "ex contingentia" is based on the idea of the necessary finitude of a series of causally connected Beings,24 it may have a partial basis in that very part of the Stagirite's Metaphysics, since Aristotle tries to demonstrate in it the necessity of a First Principle based on the impossibility of an infinity of causes. But Aristotle, contrary to Ibn Sfha, does not radically exclude any kind of possible link with the science of Physics. This is obvious at the very beginnings of his expose, when he states: "The material generation of one thing from another cannot go on in an infinite progression [...]; nor can the source of motion" (my italics).25 Although Bahmanyar makes no mention of motion, it looks as though he somehow wants to return to the original Aristotelian approach by placing the explicit treatment of God in what appears to be a more immediate connection with physics. Such an impression is only reinforced when he presents the aims of the second section of sub-part two as "[offering] knowledge (Ji ma'rifa) of the celestial bodies, their souls, their intellects and their other states, and clarifying the ultimate goal of On Heaven, as well as (wa-} some [items] of Alpha elation and of the TTieologia."26 Unfortunately, Bahmanyar does not specify here these items. However, there is little doubt that he has in mind some issues related to the heavens 21 Bahmanyar refers to the famous Theologia Aristotelis, but the present context does not allow one to decide whether or not he accepted its attribution to Aristotie. 22 Here, the reference is clearly to Aristotle's Metaphysics, but it should be recalled that this book was the first book in the Arabic translation, and not the second, as in the Greek text as it has been transmitted to us. 23 UtulugLya AristutdlTs, ed. CA.R. Badawf in Aristu cinda l-carab, op. cit., e.g., 6 and 110 (title of ch. 10).' 24 See M. Marmura, "Avicenna's Proof from Contingency for God's Existence in the Metaphysics of the Shifa3" Medieval Studies 42 (1980), 337-352, passim. 25 Aristotie, Metaphysics, 994a4-6 (Loeb Classical Library, p. 271), tr. H. Treddenik (repr. Cambridge, MA: Harvard University Press, 1975), 86-87. 26 Bahmanyar, at-Tahsil, 2.

XII

BAHMANYAR: A FAITHFUL DISCIPLE?

183

which had been discussed, at least partly and already by Aristotle himself, in metaphysics. However, again, a more systematic analysis is needed in order to offer a well-founded conclusion. Does the change of title of parts two and three involve more than a simple rewording? Does it reveal a fundamental rupture with Ibn Sfna's thought, and, if this the case, to what extent? To put it briefly: is it really significant, or not? As soon as one starts to read the work in a more systematic way, one cannot but be struck by its overtly Avicennan tone. Almost immediately one discovers that Bahmanyar quotes large passages, sometimes by way of paraphrasis, but even more often very literally, of Ibn Sfna's main philosophical encyclopaedia, the Sifa\ It has to be noted that Bahmanyar uses almost all of its different books, and thus reveals himself to be very familiar with this basic work of the Sayh ar-Ra'Ts. Moreover, one finds several quotations of the Nagat and of the IsaratP Most of them have been identified with due care by the editor of Kitdb at-Tahsil, M. Mutahharf.28 However, while his indications on the whole are correct, they lack precision, insofar as no details are given as to the exact nature of the quotations, nor to their full extent. So, it is not immediately obvious whether one is dealing with literal quotations or rather with paraphrases, or when a quotation concerns a line, a paragraph, a page, or even a chapter. In order to illustrate the enormous difference which exists in this respect between several passages, I present here two examples. The first passage, i.e., book II, maqala I, fann 9 (pp. 320—322), deals with the physical notions of continuity, contiguity, etc. The editor presents as-$ifa\ as-Samdc at-Tabici (designated by him as alfann al-awwal), maqala 3,29 chapter 2 as its direct source. A closer analysis shows that this is indeed the case, but that Bahmanyar omits many passages. The whole chapter turns out to be a collection of very literal quotations of different lengths of chapter 2 of book 3 of

27

Mutahhan indicates in many cases the precise source of such quotations, but most of the time in a rather vague way, since he neither indicates the precise beginning or end of the quotations nor specifies their actual nature. Despite a few minor mistakes, he has performed a great service by establishing all these primary indications. 28 See note 6 above. 29 In fact, he refers to maqala 2, but this is an obvious misprint.

XII 184 as-Samac at-Tabici, as the following concordance of page and line numbers makes perfectly clear: Kitab at-Tahsil 320.3-8

320.8-321.5 321.5-12 322.1 322.2-3

As-Sifa\ as-Samac at-Tabici™ 178.7-8

178.15-179.7 179.9-14 181.9-10 181.12

In all these cases Bahmanyar copies Ibn Sfna verbatim. Certainly, large parts of the concerned chapter are not reproduced by him, but one cannot but observe that, when he does preserve a passage of the text, he remains faithful to the very wording of his master. The second fragment, i.e., book III, maqala II, bdb 2,fann 4 (pp. 657—663), discusses the Divine Decree (qada*) and the problem of evil. The editor does not refer to any Avicennan text, and thus suggests that this chapter is an original contribution of Bahmanyar himself. This is clearly not the case, since many of the expressed ideas are already present in Ibn Sfna's discussion of this topic in as-Sifff, al-Ilahiyat, IX, chapter 6.31 A few wordings, such as "which is often but not the most often" (Tahsil, 659.9 = Sifa\ 422.10), "the good is willed by essence, not by accident" (Tahsil, 569.11-12 = Si/a', 421.1-2), are even literal derivations. But the whole chapter has rather to be qualified as a kind of paraphrasis of Ibn Sma's text, including the very basic ideas such as the link between evil and matter, or evil and the individual; the unavoidable occurrence of evil in this (i.e., the actual) universe; the divine providence having realized the best possible world, etc.—all of these have undoubtedly a genuine Avicennan ring. And even the evocation (661.10—663.1) of the notion of dcfwa, supplicatory prayer, does not necessarily constitute an addition by Bahmanyar himself, since Ibn Sma's Tcfliqat^ includes a rather systematic analysis of this notion which bears some vague resemblances 30 Our references are to the edition by S. Za'id, Cairo: al-Hay'a al-Misnya alAmma li-1-Kitab, 1983. 31 One may also take into account the fourth to last chapter of the Nagat., which is very close in wording to the chapter of the Sifd3. 32 Ibn Sma, at-Tacliqat, ed. CA.R. Badawi (Cairo: al-Hay'a al-Misnya al-cAmma li-1-Kitab, 1973), 47-8. I will not discuss here the different recensions of this work, nor the problem of its authenticity, but I see for the moment no better hypothesis than the one I expressed in my "Les Tcfliqat d'Ibn Sfna," 116—118, scil. they constitute students' notes of lessons by Ibn Sma. c

XII

BAHMANYAR: A FAITHFUL DISCIPLE?

185

to Bahmanyar's expose. While these surely do not suffice to qualify the Tcfliqat as a direct source, they may indicate that Bahmanyar based himself on Avicennan materials, possibly taken from oral discussions, or, at least, that he further developed ideas which had received only limited attention in Ibn Sfna's work(s). As such, we are clearly no longer dealing with literal quotation, but rather with paraphrasis, or, in some cases, elaboration. The above examples well illustrate that Bahmanyar deals with his Avicennan source(s) in different ways. Therefore, a close inspection of every single passage is needed in order to determine the exact kind of dependency on Ibn Sfna. Such an enterprise clearly exceeds the limits of the present investigation, but let us simply note that there is ample evidence that (almost?) every chapter has a direct or at least indirect basis in Ibn Slna's writings. Assuming that this basic impression correctly reflects the facts, why should one still doubt that Bahmanyar was a faithful disciple of Ibn Sfna? Basically because, as indicated above, the titles of books 2 and 3 of Kitdb at-Tahsil seem to imply important changes with respect to the very basic structure of the philosophical project as articulated by Ibn Sma in the Ddnisndmah. A more detailed analysis of the structure of each of these two books will show that a few important changes have indeed taken place. In the first three chapters of maqdla I of book 2, Bahmanyar insists, in complete agreement with Ibn Sfna, that Being qua being constitutes the very subject of the science of metaphysics33 and, once more following his master, he evokes the basic divisions of Being: the ten categories; necessity, possibility and impossibility; truth and falsehood. In all this, he largely bases himself on as-Sifd*, al-Ildhiydt, I, chapters 2, 5 and 8. Since Kitdb at-Tahsil was conceived as a compendium,34 the omission of some parts of the source-text is hardly surprising. Such is the case with as-Sifd', al-Ildhiydt, I, chapters 1 and 3, the former dealing with a negative approach, scil. what metaphysics is not, the latter with its utility and place. However, the systematic 33 It is worthwhile to note that Bahmanyar affirms that Being (wugud) is something general (or: common, amrun cammuri) which is predicated of what is beneath it not univocally but equivocally (Id bi-t-tawatu3 bal bi-t-taskik)', see at-Tahsil, 282.1. In my view, this affirmation somehow sustains the hypothesis I formulated years ago that Ibn Sma adheres to the idea of a transcendental analogy of Being, see my "Avicenna: tussen neoplatonisme en islam," Ph.D. dissertation, Leuven, 1983, 1:133-140. 34 See note 1 above.

XII 186 exclusion of any reference to God, which is predominantly present in the very first book of as-Sifa*, al-Ildhiydt—more precisely in chapters 4, 6, and 7—cannot be qualified as unimportant or, at best, of little significance. For Ibn Slna, God remains the highest (though not the first) object of the metaphysical investigation, and therefore he does not hesitate to evoke in the mentioned chapters such issues as the relationship of God with creation (including the topics of prophecy and mcfad] and God's unity and unicity. By ignoring them, Bahmanyar apparently wants to emphasize that the science of metaphysics studies chiefly, if not exclusively, Being as such. He at once seems to reject the inclusion into metaphysics proper of a (Neoplatonic) exitus-reditus scheme as Ibn Sfna had developed and integrated it in an Aristotelian metaphysical framework. Immediately afterwards, in chapters 4—6, Bahmanyar focuses on the general notion of substance. The Ilahiyat of the Sifa* is used only in a very limited way, i.e., the first part of chapter I 35 of book II, in order to explain the distinction between substance and accidens. The major source of inspiration is, however, as-Sifd\ al-Maquldt, I, 4; II, 4; and III, 1—3. Thus, Bahmanyar integrates a huge part of logic into metaphysics. He may be imitating Ibn Sma, who also discusses the categories, and therefore also substance, both in the Categories and the Metaphysics.^ But even then, one looks in vain for the distinction between mahall and mawduc afforded by Ibn Sfna in the later part of the above mentioned chapter 2 of the Ilahiyat. Body becomes the central topic in chapters 7 and 8, where asSifd\ al-Ildhiydt, II, 2 is largely the source. As for the discussion of continuity/discontinuity at the beginning of chapter 8, it reminds one somewhat of as-Sifd\ as-Samdc at-Tabici, III, 2, and its presence here may be due to Ibn Slna's remark in the Ilahiyat that corporeity is the form of continuity.37 Likewise, the refutation of atomism, presented in chapter 10 and based on as-Sifd*, as-Samd* at-Tabici, III, 3—5 (as well as one chapter of the Nagdf) can easily be explained by Ibn Sfha's observation that he has already treated this issue in the

35

Mutahhan cites ch. 9, but this is clearly a mistake. A systematic study and detailed analysis of the Categories of the Sifa3 (including a thorough comparison with the related parts of the Metaphysics) unfortunately is still lacking, although it almost certainly would contribute to a better understanding of Ibn Sfna's basic philosophical conceptions. 37 As-Sifa3, al-Ildhiydt, 64.6-7. 36

XII

BAHMANYAR: A FAITHFUL DISCIPLE?

187

Physics.^ As for chapter 9, which concentrates on such notions as contiguity and which is completely based on as-Sifd\ as-Samdc atTabi'i, III, 2 as shown earlier, it looks as though Bahmanyar considers it to be of the very same nature as the two surrounding chapters, that is, as being part of the discussion of corporeity. In chapters 11—13 several precise articulations regarding the relation between form and matter are brought to the fore. This time as-Sifd\ al-Ildhiydt, II (more specifically chapters 1, 2, and 4) reveals itself as the major source. But at the end of chapter 13 when Bahmanyar states that the number of bodies has to be finite, he once more quotes from as-Sifd*, as-Samdc at-Tabici, i.e., Ill, 8. To summarize, the first maqdla of the second book of Kitdb atTahsil includes what may be characterized as a summary of books 1 and 2 of the Ildhlydt of the Si/a3. It is striking that Bahmanyar does not hesitate to incorporate passages taken from a logical and a physical part of the Sifd\ i.e., al-Maquldt and as-Samdc at-Tabici. Nevertheless, this in no way involves a radical departure from Ibn Sfna. In fact, the Sayh ar-Ra'Is regularly notes in his Metaphysics that knowledge of these sciences is required for a correct understanding of his metaphysical project. By elaborating on these parts, Bahmanyar facilitates to some degree the task of the reader of this metaphysical part. Thus, their inclusion here is perhaps the result of a pedagogical motive. However, we also detected a very significant omission, one which cannot but elicit the following question: does Bahmanyar completely exclude theology from metaphysics? Does he somehow want to return to Aristotle's original project? But then what about book Lambda? For now these questions remain unanswered. The second maqdla of the part entitled "metaphysics" of Kitdb atTahsil examines one by one the nine categories besides substance, viz. quantity (and related topics; chapters 1—6); quality (and related topics; chapters 7-9); relation (chapter 10); when, where, posture and state (chapter 11); action and passion (chapter 12). For this expose, Bahmanyar has made massive use of Ibn Sfna's presentation of these categories in the Maquldt of the Sifd\ although he also quotes the Ildhlydt^ e.g., in his chapters 3—5 (on "unity") and chapter 9 (on "science"). Moreover, in chapter 6 (on "place," makdri) he combines

38 39

Ibid., 65.8. For more details, the reader may consult the references given by MutahharL

XII 188 elements taken from as-Sifa\ as-Samac at-Tabici, II, 6—9, and in chapter 8 (on the "quality of passion") he reproduces a part of as-Sifd\ Kitdb an-Nafs, III, 2. After the logical and quite brief presentation of action and passion in chapter 12 (417—418.10), Bahmanyar extensively analyses the phenomenon of motion (418.11—431), based on as-Sifa*, as-Samdc at-Tabici, II, 14. Further topics related to motion are discussed in his chapters 13—15 according to IV, 3 of the same work. In chapter 16 attention is paid to the issue of the eternity of motion and time, while in chapter 17 the entire focus is on time. These two chapters also have their source in as-Sifd\ as-Samdc atTabi'f, thus, chapter 16 combines passages taken from III, 11 and IV, 9, and chapter 17 combines passages from II, 10—13. When comparing this second maqdla with Ibn Sfna's expose in the Ildhiydt of the Sifd*',40 three remarks seem to impose themselves: 1. Contrary to Bahmanyar's maqdla, only three categories, i.e., quantity, quality and relation, are discussed in some detail in Ibn Slna's expose. This creates the impression that for Ibn Slna, besides substance, only these three categories possess a real metaphysical relevancy, notwithstanding their being accidents, as is clearly and explicitly emphasized. Bahmanyar, by taking into consideration all categories, undoubtedly weakens the special status accorded the three categories by Ibn Sfna's wording;41 2. In both works, the topic of unity as related to quantity, and the topic of science as related to quality are treated, but they no longer hold such a central position in Bahmanyar's expose as they evidently had in Ibn Sfna's, since they are embedded in a much more extensive account of all categories, and related topics; 3. Bahmanyar's systematic analysis in this framework of three basic physical notions, viz. motion, place and time, can hardly be justified by considering it an additional explanation which was naturally implied by Ibn Sfna's text. What Ibn Sfna had articulated in book 2 of the Sifdy, as-Samdc at-Tablci, in the way of three shorter monographs,42 as outspoken physical doctrines, becomes in Bahmanyar's work an integral part of metaphysics. Moreover, it has to be stressed

40 I refer to this work since it offers the most complete metaphysical expose of Ibn Slna, including his basic ideas in the field. 41 But, as earlier stated (see n. 36 above), a thorough examination of Ibn Slna's doctrine(s) of the categories, both on the logical and on the metaphysical level, is really needed in order to elucidate what appears to be some of his most fundamental philosophical ideas. 42 As far as the form of presentation is concerned, these monographs remind one of what is usually designated corollaries with respect to Philoponus and Simplicius.

XII

BAHMANYAR: A FAITHFUL DISCIPLE?

189

that Bahmanyar, when dealing in book 3 of Kitdb at-Tahsil with asSamdc at-Tabici, no longer makes any mention of time, and evokes only a few particular items regarding motion and place.

Therefore, it is almost natural to conclude that in this second maqdla Bahmanyar rather blurs the boundaries between logic and metaphysics on the one hand, and physics and metaphysics on the other hand. As a disciple of Ibn Sfna, such an attitude is surprising, since the Sayh ar-Ra'Is attached great importance to correct logical reasoning and to keeping distinct the respective domains of physics and metaphysics.43 Maqdla 3, which investigates the notions of anteriority and posteriority, and potency and act, is genuinely Avicennan. It has a solid basis in as-Sifd\ al-Ildhiydt, IV, 1-2. The same evaluation applies to maqdla 4, which focuses on the couples universal/particular, multiplication/individualization and genus/species, and, moreover, on definition, it being based on book 5 of the Ildhiydt of the Sifd\ However, the very first chapter is an exception. It is entitled "On Perception (idrdk), and what is related to it," and contains elements taken from as-Sifd*, Kitdb an-Nafs, II, 2 and V, 2, as well as from asSifd\ al-Ildhlydt, VIII, 6 and V, 5.44 It may still be accepted as strongly Avicennan, but Bahmanyar, before discussing the real metaphysical issue, imports elements from a natural science; in the present case this appears to be preparatory and thus may lead the reader to a better understanding of the relevant topics. Maqdla 5 may also be qualified as profoundly Avicennan. It focuses on the topic of causality, closely following the sixth book of the Ildhiydt of the Sifd\45 But again, one chapter, i.e., chapter 3 (on final causality), encompasses non-metaphysical materials. Based on as-Sifd*, as-Samdc at-Tabtci, I, 13-14, the delicate problem of fate and chance is addressed. Its discussion in a causal context is certainly natural but, as with Maqdla 2, one gets the impression that Bahmanyar simply mixes up physics and metaphysics.

43

Regarding Ibn Sma's insistence on correct logical reasoning, see Lettre au vizir, op. cit, 2; and as for his sharp distinction between physics and metaphysics, see note 19 above. 44 The reference to V, 5 is lacking in Mutahharf's edition. 45 Mutahharf, at-Tahstl, 524, refers to the Isdrat, namat 5, but several elements of the texts are directly derived from as-Sifa3, al-Ildhlydt, VI, 2. It has to be noted that Bahmanyar inverts the order of the chapter.

XII 190 Maqala 6 is of a composite nature. The first chapter, in conformity with as-Sifd', al-Ildhiydt, VII, 1, discusses consequential accidents (lawdhiq), related to one and unity. In chapter 2, finity figures as a central notion: first, according to as-Sifd\ as-Samdc at-Tabicl, II, 8, arguments are adduced for the finitude of bodies and numbers; then it is established that there cannot exist an infinity of causes, on the basis of as-Sifd\ al-Ildhlydt, VIII, 1 (albeit extremely summarized). Once more, one observes the introduction of a physical item, but in this case it clearly has a preparatory function. Based respectively on as-Sifd\ al-Ildhiydt, IX, 7, and IV, 3, one finds in chapter 3 an account of pain and joy, and in chapter 4 one of the notions of perfect and more than perfect. In the latter chapter, one also finds a part on incorruptibility, which undoubtedly has been inspired by asSifd*, Kitdb an-Nqfs, V, 4. In this case, the physical element quite naturally fits into the surrounding metaphysical context, dealing with incorruptibility. Overlooking the totality of maqdlas 3-6, we may conclude that Bahmanyar offers a massively Avicennan-inspired doctrine of the modes of Being, the inclusion of physical elements not present in Avicenna being of a rather limited nature. However, when looks at the entirety of the metaphysics, the often massive presence of logical and physical doctrines is rather problematic. One gets the impression that Bahmanyar does not always respect the borderlines between these philosophical disciplines. Moreover, he does not mention Ibn Sfna's rejection of Plato's theory of ideas (as-Sifd*, al-Ildhiydt, VII, 2—3) and his religious socio-political doctrine (as-Sifd*, al-Ildhiydt., X). In the former case, Bahmanyar perhaps considered the issue too historical and/or too technical for its inclusion in a (so-called?) basic presentation of Avicennan philosophy, whereas in the latter he may have decided that it is not really of metaphysical relevance, since, outside the Sifd\ this topic is never discussed in the metaphysical section of Ibn Sfna's encyclopaedic works. Such reasons may indeed be plausible. However, of a very different nature is the omission of almost any theological item. As previously stated,46 theology forms for Ibn Slna a constitutive element of metaphysics. Hence, by ignoring it, Bahmanyar seems to undermine one of the pillars of the Avicennan metaphysical system. 46

See above and n. 16.

XII

BAHMANYAR: A FAITHFUL DISCIPLE?

191

It has already been noted that Bahmanyar did not entitle the third part of his Kitdb at-Tahsil "Physics," but rather "the science of the states of the most noble of Beings," or, perhaps better, "the science of the states of the essences of the existing Beings," and that he discusses in its first maqdla the Necessary Existent, i.e., the divine Being. It thus becomes clear that he does not simply reject any philosophical theology, but rather he places it in what appears to be a somewhat different context from the one adopted by Ibn Slna. Before concentrating on that context, a brief examination of its contents is worthwhile. Having indicated, rather than proved,47 the existence of one single God, Bahmanyar develops a negative theology, insisting on the unity of the divine essence, denying any quiddity in Him, and rejecting any substantiality of Him.48 As for the divine attributes, they are characterized by him as either negative, relative, or composed of both relation and negation. He then offers a summary proof for God's existence, based on the distinction between possible and necessary, and on the impossibility of an infinite series of causes. Further, he emphasizes that God is at once Knower, Known, and Knowledge, that is, His self-knowledge implies the knowledge of what is other than Him, that this gives rise to the emanation of all things without creating any plurality in His essence,49 that God is absolute perfection and free of any desire, and that God is life and truth. Finally, Bahmanyar observes that from the One only one can proceed.50 All this rings very Avicennan, and actually may be characterized as a rather succinct summary of Ibn Sfna's theology, based on asSifd\ al-Ilaktydt, VIII, chapters 4—7, with a few small additions taken

47

The proof follows only later; see infra. It is somewhat surprising that Bahmanyar in this context does not mention Ibn Sfna's famous saying that in God there is no other quiddity than his annlya (see, e.g., as-Sifd3, alTldhiydt, 344.10), but affirms that the divine essence is a haqiqa Id ism lahu ("a truth having no name"), an expression present in the Ta'liqdt, op. cit, 183.15-16 and 185.26. 49 In this context, Bahmanyar evokes very briefly the notion of providence, stating that it means that God knows all things and is their final cause (at-Tahstl, 579.3-4). A more systematic analysis of this basic notion is only offered in the De Caelo part, see infra. 50 H.A. Davidson, Alfarabi, Avicenna, and Averroes on Intellect, Their Cosmologies, Theories of the Active Intellect, and Theories of Human Intellect (Oxford, Oxford University Press, 1992), 75, n. 3, remarks that despite its Plotinian inspiration, this formulation seems to be original with Ibn Sfna. 48

XII 192 from the same work, chapters 4 and 6.51 Nevertheless, it is striking that he concentrates mainly on the divine essence and the related topic of the divine essential attributes, and that he pays only very limited attention to the very proof of God's existence as well as to the relationship between God and creatures. Moreover, as previously indicated, this theology is no longer a part of metaphysics proper. It is now integrated into a study of Beings. Does this mean that there is no longer any room for Physics in Bahmanyar's view? The second maqdla of the third part makes clear that such is not the case. It is divided into four sections, or bdbs, which constitute a basic survey of the first six parts of the Physics as conceived by Ibn Sfha, i.e., Physics (general principles), Heaven and Earth, Generation and Corruption, Actions and Passions (Meteorologica I), Higher Influences (Meteorologica II), and On the Soul.52 But again one finds significant displacements. In the first section, after what appears to be a general and introductory division of the possible Beings in chapter 1, Bahmanyar elaborates the following physical issues, always based on as-Sifd\ as-Samdc at-Tabt'i: motion (I, 6; IV, 9 and 14); finitude-infinity (III, 8-10); corporeal form (III, 10); direction of movement and situs (III, 14; with an addition taken from the Isdrdfy, natural place (IV, 9-11) and natural body as principle of motion (IV, 12 and 8). It is perhaps worth noting that most of the quotations derive from books 3 and 4 of as-Samdc. Of course, a large use of book 2 had already been made in the metaphysical part, when Bahmanyar discussed the topics of motion, place and time.53 However, of much more importance is the fact that in the eighth and last chapter, Bahmanyar introduces a passage derived from as-Sifa\ as-Samd* wa-l-cdlam, despite his explicit

51 Mutahhan, 570, also refers to as-Sifa\ al-Ilahiyat, I, 7 and V, 5, but the passage of Bahmanyar's text, which affirms the absence of any plurality or changeability in the divine essence, can in my view be related to ibid., VIII, 5. It has to be stressed that jvery literal quotations are rare in the whole maqdla, at least if the Ilahiyat of the Sifd3 did constitute Bahmanyar's major source text for the present expose. Let me simply note that thus far I see no better candidate. 52 The omission of the parts on plants and animals is not overly surprising, insofar as Ibn Sma, as with mathematics (see n. 12 above) never reexamined them in his later works (as with mathematics, probably considering that they had been established in a satisfactory manner in the Aristotelian tradition before him). 53 See above.

XII

BAHMANYAR: A FAITHFUL DISCIPLE?

193

mention in the title of the section that it deals with the intentions of as-Samac at-Tabici. In fact, while focusing on simple body and the reciprocal implication of generation and corruption, Bahmanyar clearly brings to the fore materials taken from as-Sama* wa-l-cdlam, more precisely, chapters 1 and 4. It would appear that he wants to avoid the treatment of such typically and exclusively earthly phenomena in the section on Heaven. Although Bahmanyar presents the second section as an expose (kaldm) on as-Sama" wa-l-cdlam, using the title as given by Ibn Sfna,54 he nevertheless deals only with heavenly matters in its elaboration. In the first chapter, based on as-Sifd\ as-Same? wa-l-cdlam, chapters 2~3, Heaven is described as the first body. Chapters 2 and 3, which are largely identical,55 concentrate on the celestial Beings, viz. Intelligences, souls and bodies. It is insisted that the celestial movements are volitional and that the lower world in no way constitutes the motive for these movements. Moreover, a particular emphasis is placed on the role of the celestial Beings, especially the Giver of Forms,56 in the process of emanation. These ideas remind one of asSifd\ al-Ildhiydt, IX, 2—5. Despite the absence of almost any literal quotation,57 it is clear that Bahmanyar has been inspired by these chapters, or perhaps a text similar to them. Whatever the case, he obviously introduces outspoken (at least in the Avicennan perspective) metaphysical ideas into a clearly physical context. This becomes even more evident when, in chapter 4, divine providence is analyzed in detail. Its formulation is similar to as-Sifd3, al-Ildhiydt, IX, 6 (at the beginning), but Bahmanyar omits, at least in the present

54

As is well known, Aristotle simply mentions On Heaven. Almost identical are the following passages: 641.5-11 = 650.4-7 and 650.17-651.1; 642.2 = 651.5; 642.7-8 = 651.16-17; 642.10-15 = 651.18-652.6; 643.3-12 = 652.9-18; and 643.17-644.7 = 653.1-9. Generally speaking, chapter 3 appears to be an enlarged version of chapter 2. A detailed analysis of both chapters may reveal the exact nature of their mutual relationship, but this falls outside the scope of the present study. 56 This is especially true in chapter 3, where Bahmanyar refers to it by its classical denomination, i.e., wdhib as-suwar, in chapter two, it is less emphasized and, moreover, one finds the unusual expression al-mufid li-s-suwar. 57 Mutahharf gives no reference, but the doctrinal similarity with the indicated chapters is rather straightforward; indeed, the very wording is close to that of the Sifd\ e.g., ch. 2 § 1-IX, 2 (384); § 3-IX, 3 (396), § 7-IX, 2 (386-387), last paragraph-IX, 5 (411); ch. 3-IX, 4 (406). 55

XII 194 context,58 the key notion of knowledge. In the very same chapter the difficult problem of how to explain the existence of evil in light of the divine decree (qadct) is given much attention. Here one encounters genuinely Avicennan ideas, viz. evil is the privation of existence or of the perfection of existence; evil happens often, but not most often; evil is accidental and only occurs in the individual; this world could not have been created free of evil otherwise it would not have been this world. A probable, if not certain, source for these doctrines is the remaining part of the very same chapter 6 of book 9 of as-Sifa\ al-Ildhiydt.59 However, the incorporation of the problematic of evil into a De Caelo context, together with the omission of mentioning any act of knowledge in the activity of providence, hardly allows for any understanding of the divine Decree other than in a purely natural way. In accepting such a view, Bahmanyar sharply deviates from Ibn Slna, who made a major effort to affirm God as the cause of the Universe while at the same time safeguarding a real link between God and his creation.60 Section three deals exclusively with physical matters. Based on asSifa\ al-Kawn wa-l-fasdd^1 one finds a discussion of the four elements, alteration, growth, rarefication, condensation and mixture, more precisely in chapters 1-3 and 5. The phenomenon of light and of light rays forms the central issue of chapter 4. A large part of this chapter is directly based on two chapters of the Tabflydt of the Ddnisndmah^ as is clear from the following table.

58 In the first maqala of the third part, Bahmanyar mentioned a very summary definition of providence, which still included an explicit reference to the divine knowledge; see n. 49 above. 59 As for the chapters 2-3, Mutahharf also provides no reference. It has to be noted that at the end of the chapter, 662-63, Bahmanyar adds a few remarks on c du d\ which vaguely resemble a passage on the very same topic in the Tafliqdt, 47.20-48.12. 60 See my "Creation and Emanation in Ibn Sfna," Documenti e studi sulla tradizione filosofica medievale 1 (1997), 45-477. 61 For details, see Mutahharf's edition, but at 682, n. 2 one has to read "8" instead of "9." 62 I made the discovery of these chapters while finishing the present paper. Due to time constraints, I was unable to investigate further the possible presence of other chapters of the Danisnamah in the Tahsil (but see n. 6 above). Let me add that the present discovery seriously strengthens my hypothesis of the existence of an Arabic original of the Danisnamah. See my "Les Ta'liqat d'Ibn Sma," 110. Mutahharf's reference to as-Sifa\ Kitab an-Nafs, III, 1-2 is not really adequate (perhaps he was misled by the tide of the chapter).

XII

BAHMANYAR: A FAITHFUL DISCIPLE? Kitab at-Tahsil

686.7-11 687.1-6 687.7-688.3 688.3-6 688.7-14 689.5-690.8

195

Ddnisndmah, Tabi'lydt

42.1-8 (beginning of ch. 19) 42.10-43.3 43.7-44.6 45.1-2 45.8-46.4 (end ch. 19) 46.7-49.8 (entire ch. 20)

Some of the passages do constitute a literal translation, while others are of a rather paraphrastic nature. However, the similarities between the two texts are striking. It should be mentioned that at the end of the chapter a passage dealing with color has been added and which is clearly based on as-Sifa\ Kitab an-Nafs, III, 1. Chapter 6, in its turn, mentions the classes of elements, and discusses such phenomena as liquefaction, congelation, etc., by drawing on as-Sifc?, al-Afcdl wa-l-inffdldt, I, 1—2 and 5—9. Bahmanyar does not raise the issue of mizag as developed from an outspoken medical context by Ibn Sfna in the second part of the latter work.63 As for chapter 7, it concentrates on sky-linked items, e.g., clouds, winds, etc., and is based on as-Sifa\ al-Atdr al-culwiya, II, 1-5. In this case Bahmanyar omits the first part of Ibn Sfna's volume where one finds an analysis of earth-linked phenomena such as the origination of wells, mountains, etc. The reason for this omission is not immediately evident.64 Mention has also to be made of the fact that in the title of this third section, Bahmanyar refers to the Book of Meteorology, under the sole title of al-Atdr al-culwiya, the usual denomination of Aristotle's book in the Arabic tradition. Hence, it looks as though he rejects Ibn Sfna's division of the Stagirite's work into two independent treatises. In the fourth, and last, section, Bahmanyar summarizes Kitab anNafs of the Sifd\ Respecting more or less its order, he discusses the following issues: demonstration of the soul, its substantiality and its incorporeality (Nafs, I, 1, 3); the vegetative powers of the soul (Nafs, II, 1); the five outer senses (Nafs, II, 2-5; III, 1, 5 and 8); the five inner senses (Nafs, IV, 1~3); theoretical and practical intellect and the acquisition of knowledge (Nqfs, I, 5 and V, 1); estimation and dreams

63 See S. van Riet (ed.)5 Avicenna Latinus, Liber quartus Naturalium, De actionibus et passionibus (Louvain-la-Neuve: E. Peeters), 28*. 64 It might be that Bahmanyar judged them sufficiently fixed, as with mathematics (see n. 12 above) and with the section on plants and animals (see n. 52 above).

XII 196 (Nafs, V, 7; IV, 2-3); motive powers (Nqfs, IV, 4); intellect and the intellected (Nqfs, V, 2 and 6); classification of the powers of the soul (Nqfs, I, 5); origination and individuation of the soul, and rejection of the idea of transmigration (Nqfs, V, 3—4, 7—8; I, 3); resurrection (Nqfs, V, 4). At first glance, Bahmanyar offers a serious survey of the essentials of Ibn Sfna's psychological views. When overlooking the entire second maqdla of the third part of Kitdb at-Tahsil, one gets the impression that Bahmanyar at least partly seems to "re-Aristotelize" the physical project (in its broad sense) of Ibn Sfna: as with the Stagirite, On Heavens deals exclusively with the super-lunar Beings and their activities. Of course, one can barely detect such a notion as that of providence in the Stagirite5s work, but if one wants to include such a notion in an Aristotelian context, no other work seems more natural in which to do so.65 Also the designation of Meteorology by one single title, as indicated above, points in the very same direction. However, at the same time, one cannot but observe that the omission of the basic treatment of motion, place and time here, and its transferal to the context of metaphysics, simply does not fit with a "return to Aristotle." From a more encompassing point of view, Kitdb at-Tahsil clearly deviates from Ibn Sfna's new philosophical project. It dismisses the extraordinary attempt of that project to elaborate a unified metaphysics. It also seems to reject the unique valorization of theology as expressed by the Sayh ar-Ra'fs. It rather makes God, as well as (His?) providence part of a (natural?) study of Being. In any case, it is definitely not Avicennan in several of its structural demarches, notwithstanding its massive use of Avicennan texts. By limiting metaphysics to a pure ontology, and by re-arranging several physical doctrines, one is occasionally inclined to believe that Bahmanyar had in mind a "re-Aristotelizing" of Ibn Sfna's thought. If this is indeed the case, one may seriously doubt whether he has really understood the Stagirite's thought. The way he mixes up logic, physics and metaphysics clearly has no basis in the great Greek thinker. It rather points to the influence of a figure such as Abu 1-Qasim al-Kirmanf, accused

65 One may think of such passages as I, 4, 271a33 in order to include the very idea of providence in the Aristotelian framework. I thank G. Guldentops, J. Brams and W. Vanhamel for their helpful suggestions in this respect.

XII

BAHMANYAR: A FAITHFUL DISCIPLE?

197

by Ibn Sfna of being a weak logician and a slavish follower of a bad Aristotelianism. It therefore seems possible, if not probable, that Bahmanyar sided more with the latter than with Ibn Sfna.66 Whatever the case in that regard, our analysis has sufficiently shown that Bahmanyar cannot be characterized as a faithful disciple of Ibn Sfna.

66

By way of hypothesis, and in a very provisional way, I would like to suggest that Ibn Sfna tried to have the young Bahmanyar among his disciples (probably during his [first] stay in Rayy and/or his stay in Hamadan), but that already during his lifetime and surely in his later years in Isfahan, it became evident that Bahmanyar had taken sides with al-Kirmanf.

XIII MULLA SADRA'S USE OF IBN SINA'S TA'LIQAT IN THE ASFAR*

A rapid survey of the Asfar1 shows that Mulla Sadra was quite familiar with many of the Avicennian works. He not only cites the major philosophical (e.g. Shifd\ Najdt, Ishdrdt, and Ddnesh-Ndmeh) and medical works (e.g. Qdnun and al-Adtviyya al-qalbiyya], but also a whole range of other works, such as (Uyun al-hikma, Mubdhathdt, K. al-Mabdd' wa-l-Ma(dd, R. Ft l-(ishq, R. al-Adhawiyya ft 'l-ma'dd, R. al-Nayruziyya, and, finally, the correspondence on scientific questions with al-Blrunl.2 The many citations taken from them reveal that he possessed a more than superficial knowledge of them. Though he sometimes criticizes elements of Ibn Slna's doctrine, it is nevertheless apparent, even from a cursory overview, that he considered the latter to be a great philosopher, maybe even the greatest philosopher of all time. In fact, there seems to be even more 'Ibn Slna' in the Asfdr than Mulla Sadra makes explicit, but we will not deal with that matter here. We will concentrate only on the explicit citations, and focus on one particular text that we have not yet mentioned—the Ta'liqdt. It must be stated that this work had received little or no attention in the post-Avicennian tradition, both ancient and modern, perhaps owing to the fact that the attribution of the work to Ibn Slna is not at all obvious. However, elsewhere I have shown from the underlying argument from internal evidence that the work is undoubtedly

* This paper is a revised version of a lecture given at the World Congress of Mulla Sadra (Tehran, 1999). M. S. ShlrazI, al-Hikmat al-muta* aiiyya ft l-Asfdr al-(aqliyyat al-arba'a, ed. R. Lutfl and S. M. H. Tabataba'I, 9 vols. (Q°rn5 n.d.). These Avicennian works are well known from the basic bibliographies (Mehdavl and Anawati). For recent editions, see my An Annotated Bibliography on Ibn Slna (1970-1989) (Including Arabic and Persian Publications, and Turkish and Russian References) Ancient and Medieval Philosophy, ser. I, vol. xiii (Leuven, 1991), ch. 1., and An Annotated Bibliography on Ibn Slna: First Supplement (1990-1994), Textes et Etudes du Moyen Age, 12 (Louvain-la-Neuve, 1999), ch. 1. 3 Ibn Slna, Al-Ta(liqdt, ed. A. Badawi (Cairo, 1973).

XIII 2,

Avicennian-inspired;4 it is therefore no real surprise that Mulla Sadra seems to have accepted it as a genuine Avicennian work. What is surprising is his choice of such a difficult work. On reading the text, one often has the impression that there is no real structure in the work, nor a fortiori in its argumentation. I have been able to find elements of what seems to be the basic structure of the work,5 but only after a long and assiduous search. Of course, Mulla Sadra probably had another version of the work at his disposal than that edited by Badawi, such as the one that al-Lawkarl had divided into ta'liqs. (Mehdavl indicates that al-Lawkarl's division is far from perfect.)6 Nevertheless, the fact remains that Mulla Sadra did use this work, and did so in a clearly first-hand manner: in the Asfar over thirty citations can be found. While the number of citations is not in itself significant, their doctrinal content is. In what follows I will try to determine to what extent Mulla Sadra does or does not accept the basic ideas expressed in these citations. It is immediately evident that Mulla Sadra uses the Ta'liqdt in various ways, such that one may distinguish at least three categories of citation. The first category consists in allusion to an apparently Avicennian idea, but in a wording that seems to have no textual basis in the Ta'liqdt. For example in S. II, M. 1, F. 4, c. 7 (v. 182), the natural corporeal form constitutes the proper object of investigation. In discussing this topic, Mulla Sadra insists that each form that originates from another natural form has its own particularities regarding measure, dimension, and substantial connection, and to substantiate this view he vaguely refers to the Ta'liqat. Chapter 3 of S. Ill, M. 9 (vii. 207-9) focuses on the famous adagium: 'From the One nothing proceeds but the one.' The Ta'liqdt are mentioned in its title. Finally, in S. IV, B. 5, c. 6 (viii. 259), Mulla Sadra points to the fact that Ibn Slna, in his Ta'liqdt and elsewhere, holds that celestial motions are essentially voluntary motions, caused by a soul. In none of these three cases have I found that Mulla Sadra copied a specific fragment in the Ta'liqdt, verbatim, or nearly verbatim. In the second case, it is certain that Bahmanyar's K. al-Tahsil7 is being quoted, and hence is Mulla Sadra's direct source. Nevertheless, it is possible in each of the 4

See my 'Les Ta'liqat dTbn Sina: Essai de structuration et de datation', in A. de Libera, A. Elamrani-Jamal and A. Galonnier (eds.), Langages et philosophic: Hommage ajeanjolivet (Etudes Philosophiqes Medievales, 74 (Paris, 1997), 109-22. 5 Ibid. 112-16. 6 Y. Mehdavl, Fihrist-i mussanafdt-i Ibn-i-Smd. (Tehran, 1954), 60-4, esp. 63. 7 Bahmanyar Ibn Al-Marzuban; al-Tahsil, ed. M. Mutahharl, 2nd edn. (Tehran, 1375 SH), 531.

XIII M U L L A SADRA'S USE OF IBN S!NA'S TA(LIQAT

3

three cases to locate a feasible source in the Ta(llqdt. In this regard, one might refer to pp. 52, corresponding to 100-1 and 106-8. And of course the ideas being presented are indeed Avicennian, or at least are consistent with an Avicennian framework of thought. Therefore, it seems reasonable to describe this category of citations as showing a degree of 'superficial' doctrinal influence. Let us now deal with a second category of citations. Although these too are not literal, they are much closer to the text of the Ta'liqdt than the previous ones. A first instance is given in S. I, M. 3, M. 6, c. 3 (ii. 277). Here, Mulla Sadra specifies that the action of a soul is not visible to what is beneath it, stressing that Ibn Sina has affirmed the very same idea on several occasions in his Ta'liqat. One actually finds several passages in the Ta'llqdt (pp. 102-4) where precisely that idea is emphasized. Another such citation is present in S. Ill, M. 3, c. 6 (vi. 216). Once again 'passages' of the Ta'llqdt are referred to in order to sustain a particular doctrinal point, i.e. that the plurality of things emanates from God not in a temporal but in a causal way. Very similar in content is the affirmation of Ta(liqdt, p. 157,11. 21-5, which begins 'The concomitants of the Necessary Being, which are the things caused by Him, are necessarily together with Him. However, they are posterior to Him, not according to a temporal posteriority, but according to a posteriority [which characterizes the relationship] between cause and caused.' Similar expressions are found on pp. 14, 152, 156, and 163. As far as I can see, there is no other citation of this kind in the Asfdr. These two cases suggest that Mulla Sadra somehow tried to summarize a major Avicennian idea, and more specifically the different wordings in which it is presented in the Ta'llqdt. In this respect, the link with the Ta'liqdt is far more evident here than in the first category of citation. The preceding examples indicate that Mulla Sadra at least had knowledge of some of the major ideas of the Ta(liqdt. He certainly does not hesitate to allude to or summarize them. Taking their contexts into account, one cannot but gain the impression that Mulla Sadra took a predominantly positive view of these Avicennian ideas. But does the preference for paraphrase and summary (instead of citation) suggest some reservation on Mulla Sadra's part? Such an inference would be premature, for Mulla Sadra offers many literal citations, which deserve our attention. Let us first concentrate on the third category, the shorter citations, which I will discuss according to the order of their appearance in the Asfdr. In S. I, M. 1, c. 5 (i. 41), there are two minor citations, reproducing respectively Ta'liqdt, pp. 178, 11. 27-9 and 179, 11. 1-4. (Note that in Badawi's edition they follow one another.) They both express the

XIII 4

same basic ideas, i.e. that the Necessary Being is self-sufficient, and whatever derives its existence from something else requires the latter for its own subsistence. Mulla Sadra embeds these ideas in the broader context of the particularization of existence. Moreover, he sees in them a genuine basis for his doctrine that contingent Beings are 'modes' (shu'un) of the Necessary Being. In M. 2, c. 4 (i. 128), Mulla Sadra states that the most astonishing affirmation of Ibn Slna regarding the absolute necessity of the Necessary Being is present in the Ta'liqdt, where he says: 'All things belonging to the First are necessary ones, and no contingency whatsoever is there' (p. 29, 1. 20). Mulla Sadra reads li-l-awwali, but Badawi's edition gives (inda l-awayi\^ a reading which is confirmed by the two manuscripts I have consulted.8 This wording imposes the following translation: 'All things which are linked to the First Things [i.e. the Higher Intelligences] are necessary, and no contingency whatsoever is there.' Whereas Mulla Sadra's reading links the necessary aspect of Beings with their belonging to God, the actual reading of the Ta'llqdt stresses, on the contrary, that there is no existential contingency in the higher realm of the supra-lunary world. In the latter, everything happens in a necessary way: the motions of the celestial spheres are uniform. However, in a genuine Avicennian perspective, this cannot also mean that on the essential level there is no contingency, given the fact that only the Necessary Being is absolutely necessary. From this point of view, the end of the citation in the Ta'liqdt is not without ambiguity. Was Mulla Sadra aware of this, and did he therefore 'correct' the text? Did he believe that a scribal error might have occured in its transmission? Or, did he simply have another version at his disposal than the one now extant? At present, it is impossible to settle this problem, especially since Mulla Sadra's reading itself is also problematic. In fact, it seems to affirm that the totality of all Beings is included in the very Being of the First Being. Yet Mulla Sadra somehow seems to avoid such a radical pantheism, by combining the citation of the Ta'llqdt with one in the Theologia Aristotelis,9 which states that 'according to necessity' the possible things are present to the First Principle—in other words they are necessary by God, insofar as He is their ultimate cause. It seems therefore that Mulla Sadra has no other intention than to present the well-known Avicennian thesis of the 'necessity ab alio of the possible in se\ 8

i.e. Aya Sofia 2390, and British Library Or. Micr. 6724, Add. 16659. I did not find the citation as such in the Theologia, but something similar in content is expressed in it on p. 156 of the edition by A. Badawi, Afldtun (inda I-'Arab, 3rd edn. (Kuwait, 1977), 1-164. 9

XIII M U L L A S A D R A ' S USE OF IBN S I N A ' S TA(LIQAT

5

In S. I, M. 3, M. 6, c. 24 (ii. 391), in the midst of an analysis of the distinction between liberality (jud) and goodness (al-khayr), Mulla Sadra refers to Ta'llqdt^ p. 18,11. 1-4. This passage brings to the fore the concept that God, as the Necessary Being, is the sole real agent and final cause. Mulla Sadra concurs completely with this idea, he even detects in it an indication of the existence of an (innate) love of creatures for their Creator. Still in S. I—more precisely in M. 3, M. 9, c. 11 (iii. 351)—Ta'llqdt p. 78,11. 12-14 is cited in order to show that God's being both knower and known implies no duality in His essence. (It will become clearer below how essential the issue of the unity of the divine essence was for Mulla Sadra.) Once more, it must be noted that the text of the citation in the Asfdr slightly differs from that of Badawi's edition, but this time no doctrinal difference is implied. This may well constitute more evidence of Mulla Sadra's having used another version. S. II, M. 1, F. 1, c. 2 (v. 20), reproducing Ta'liqdt, p. 90, 11. 3-5, affirms that a part of a body cannot but signify a part of the measure of a body. It is included in a context in which Mulla Sadra tries to specify what is, and what is not, essential for a body. Finally, in S. Ill, M. 2, c. 1 (vi. 120), Mulla Sadra, in an exposition on the divine attributes, quotes Ta'liqdt, p. 49,11. 13-15, stressing that the multiplicity of the divine attributes does not affect the unity of the divine essence. Mulla Sadra had a high regard for Ibn Slna's view on this matter, as will become clearer from the longer citations discussed below. Based on the totality of these shorter citations, the thesis that Mulla Sadra positively evaluates his source text is undoubtedly strengthened. However, we must not forget that these citations concern specific affirmations, hence their doctrinal significance is rather limited. A final decision is impossible without systematic analysis of the longer citations, that is, citations covering more than ten lines of the original text (according to the printed edition), or combining different fragments. Thus we now turn to them, once again according to the order of their appearance in the Asfdr. 'What can man know exactly?' This question is the central issue of chapter 13 of S. I, M. 3, M. 1 (i. 391-2). In it, Ta'llqdt, pp. 34, 1. 17-35, 1. 5 are reproduced literally. Essentially, this fragment states that a human being cannot distinctively grasp the 'true essence' (haqiqa] of a singular thing. Human knowledge as such is limited to understanding the properties or accidents that characterize such a thing. From a formal point of view, one must note that Mulla Sadra adds at the end of the citation: 'till the end of his words'. This addition seems to indicate that for him Ibn Slna's exposition is incomplete,

XIII 6

at least so far as the wording goes. But why does Mulla Sadra stop the citation at that point? Although I have no definite proof, I think that for him the rest of the exposition in the Ta'liqdt entailed nothing really new on a doctrinal level. From Badawi's text, as well as from Lawkarl's division into ta'liqs, it seems highly probable that the very end of the fragment must be placed on line 14 of p. 35. Notwithstanding, the relative length of the citation clearly shows that Mulla Sadra, who stresses that no proper proof (burhdn) exists regarding individual existents, did find Ibn Slna's ideas in this respect worthy of attention, if not adherence. A quite different problem is discussed in S. I, M. 3, M. 7, c. 31 (iii. 118-21), that is, time and motion being determining factors of the finitude of existence. In this context, Mulla Sadra combines elements taken from Ta'liqat, pp. 108, 11. 21-5; 109, 11. 6-7; 105, 11. 19-23; 105, 11. 24-6; and 106, 1. 9. All deal with celestial bodies, specifically with their motion and nature. The citations of the Ta'llqdt strongly emphasize that the ultimate goal of celestial motion is not motion in itself, but the conservation of the nature of the related sphere. Moreover, the celestial soul, which originates the motion concerned, constitutes on the conceptual level not just an individual, but also a species. Mulla Sadra, in his explanation of these citations, confirms these basic ideas by stressing for example the role of celestial souls as principles of celestial motions, the latter being invisible to anything beneath them. He adds, moreover that celestial spheres have no other 'situs' than natural ones, and, in order to reinforce this idea, he quotes Ta'liqdt, p. 101,11. 15-17. All this shows clearly that Mulla Sadra adhered completely to Ibn Slna's ideas on these matters. Of still another nature is the problematic of S. Ill, M. 2, c. 3 (vi. 131). Having criticized the Mu'tazill thesis that 'God's attributes are His essence', Mulla Sadra wonders whether one may conceive of action and reception in God, and, if so, how? In view of a possible solution of this delicate problem, he quotes Ta'llqdt, p. 181,11. 10-22, which states that the divine attributes are concomitant to the divine essence, insofar as action and reception in God are but one. Except for a very small omission (covering wa-innamd ... al-jism of 1.11. 17-18), the quotation is literal. Doctrinally, Mulla Sadra once more agrees with his Avicennian basis. For him, reception in God in no way implies any formal influence or passion, nor the presence of any duality (a fortiori plurality) in His essence. Somewhat further in the same S. Ill, one finds a chapter, M. 3, c. 7, titled, 'How are the forms of things impressed in the Essence of God?'. In this chapter, no less than four major fragments based on the Ta'liqdt are present. The first of these (vi. 11. 206-8) stresses that the

XIII M U L L A S A D R A S USE OF IBN S I N A S T A ' L / Q A T

7

divine perfection automatically entails the very existentialization of all things, and this out of the divine essence; in other words, God, in understanding Himself as the principle of all Being, gives all things their existence. In order to sustain this view, Mulla Sadra quotes Ta'liqat, pp. 155, 11. 7-10; 155, 11. 14-17; 156, 11. 1-3; 156, 11. 4-8; and 156, 1. 13-157, 1. 4. Between the penultimate and final quotations, Mulla Sadra discusses a possible objection, but it is of no importance for the purposes of this article. What is of importance is to consider the bare fact that the last quotation is paraphrastic rather than literal. However, since it almost forms a unity with the preceding citations, I opted to include it in that context. A second major fragment (vi. 210-11) focuses on the topic of action and reception in God. We have already come across this problem in the discussion of S. Ill, M. 2, c. 3 (vi. 131), where we found a quotation of Ta'liqdt, p. 181, 11. 10-22. Almost the same quotation, but now covering the lines 12-23 (and without omission), is given here. But this time, Mulla Sadra adds pp. 181, 1. 24; 182, 11. 1-2; 182, 11. 3-5; and 182,11. 7-8. In his personal remarks, Mulla Sadra stresses that these 'proofs' (bayydndt] are not knowable by ordinary men, though they clearly demonstrate that if a thing has no potency for being—either temporally or essentially—preceded by something else, it cannot have any kind of composition. In all this, Mulla Sadra is once more highly appreciative of his source text. As to the third 'ensemble' (vi. 214-16), it concentrates on the creationist character of divine knowledge. There is no doubt that for God to know is to create. And even if the created things are multiple, this in no way affects the absolute unity of the divine essence. Finally, the relationship between God and the things He created is one of pure intelligibility. A basis for all this can be derived from Ta'liqdt, pp. 153,11. 1-2; 153,11. 3-4; 153,11. 7-10; 60,11. 25-6; 154,11. 21-2; 154,11. 5-10; 159,11. 6-8; and 174,11. 8-11. Having discussed all these issues, Mulla Sadra points to an important difficulty regarding the unity (ittihdd) between the knower and the known. According to Mulla Sadra, this doctrine had been attributed to Porphyry by Ibn Slna, who, after initially accepting it, later rejected it.1 But for Mulla Sadra Ibn Slna's own exposition on these matters demonstrates the undeniable validity of Porphyry's opinion. In spite of this sharp critical observation, Mulla Sadra does not abandon Ibn Slna in his further examination of divine 10 See M. S. ShlrazI, Le Livre des penetrations metapbysiques, ed. and trans, with notes by H. Corbin (Tehran and Paris, 1982), 185-7, n. 71. The importance, as well as the profound significance of this doctrine for Mulla Sadra's thought, is clarified by Corbin in a most magisterial way.

XIII 8

knowledge of creatures. On the contrary, according to a last thesis (regarding divine knowledge) there is no precedence of it over existentialization, nor vice versa. Once again he looks for support in the Ta'liqat, combining this time pp. 48, 11. 13-22; 49,11. 6-8; 48,11. 23-7; 48,1. 28-49,1. 2; and 49,11. 3-5. From the preceding lines, it is obvious that the Ta'liqdt are of great importance for this large seventh chapter, which deals with one of the most delicate philosophical problems: God's knowledge of the particulars. In view of the Islamic stress on God's unity (tawhid], that problem becomes even more puzzling for the philosopher. Despite some criticism, it is obvious that Mulla Sadra greatly admires Ibn Slna's doctrine on this matter. For him, at least at the philosophical level, Ibn Slna has elaborated the best argument, one that is very close to the truth. Since this problem is at the core of the Ta'llqdt, which expresses Ibn Slna's mature thought (the Ta'llqdt being undoubtedly posterior to the Shifd\ since it cites fragments from the latter), it is hardly surprising that Mulla Sadra used this work in such a systematic manner. In S. Ill, M. 4, c. 2 (vi. 313) the difference between power (qudra) in God and man is at stake. While stating that God's power is His being necessary (al-wdjibiyya) and active (al-fa(liyya], Mulla Sadra quotes Ta'liqdt, pp. 53, 11. 26-7 and 50, 11. 23-61, 17. Common to both citations is precisely the idea that the divine action, in contrast to man's, has no cause. Divine will is in no way guided by any kind of preponderance, nor is it in view of any particular goal. Or, to put it in a different way, divine will has no other motive than the divine essence itself, i.e. divine goodness. However, the question inevitably arises whether in such a perspective it is possible to conceive of the world otherwise than being eternal? In his response to this question, Mulla Sadra answers that this is surely not the case so far as the world of generation and corruption is concerned. In formulating his answer this way he—at least in principle—does not deny that the supralunary world is, or even must be, eternal—as it is according to Ibn Slna's doctrine. In chapter 8 of the same M. 4 (vi. 358-63), one finds a very impressive and very long quotation reproducing Ta(liqdt^ pp. 16, 11. 5-9; 16, 11. 26-19, 14; 19, 11. 15-25; 20, 11. 20-22; and 20, 11. 23-21, 16. In these passages, a detailed analysis is given of three major divine attributes, i.e. will (irdda), power (qudra) and wisdom (hikma).11 A systematic and profound examination of this extremely 11 Contrary to what is suggested by the actual edition of the As far, the expression: fa-naqulu (p. 359) in no way finishes the quotation of the Ta(liqdt. Moreover, remember that these pages of the Ta'llqdt seem to constitute the Arabic original of several chapters of the Metaphysics of the Danesh-Nameh, see my: 'Le Danesh-Nameh

XIII M U L L A SADRA'S USE OF IBN S I N A ' S TA(LIQAT

9

long citation is beyond the scope of this article. Suffice it to say that Mulla Sadra quotes in extenso almost everything that is said about these attributes in the Ta'liqdt. Again, it looks as though he considers it to be simply the best possible philosophical presentation, since he does not hesitate to place it immediately after the quotation of a series of Prophetic traditions. In chapter 14, still in M. 4 (vi. 403-6), Mulla Sadra concentrates on the prayer of supplication, du(dy. On this he quotes Ta'llqdt, pp. 47, 11. 28-48, 1. 2; 47,11. 20-28; 48,11. 2-5; 48,11. 5-8; 48, 11. 9-12; 151, 1. 25-152, 1. 7; and 152, 11. 8-17.l These fragments stress in particular that the prayer of supplication has no effect on the superior world, nor (a fortiori) on God. In fact, such kind of prayer forms a part of the all-encompassing totality of qaddy and qadar^ hence, it has its ultimate cause in God Himself. Contrary to Ibn Slna, Mulla Sadra seems to believe that the prayer of supplication may somehow affect the supra-lunary world, i.e. the world of the angels. But, he concedes to Ibn Slna that it cannot have any influence on the divine Being Himself. Thus in this case, he does not agree completely with Ibn Slna, but neither does he completely reject the latter's theory. A very last long citation, taken from Ta'liqdt^ pp. 120,1. 20-121,1. 6, occurs in S. IV, B. 9, c. 5 (ix. 117-18). The relationship between soul and body is the focus of that chapter. As to the Ta(llqdt citation, al-'aql al-baslt, 'simple intellect', reveals its central notion. It is explained as being synonymous with the intellect, which at once grasps all intelligibles according to their causal order. Hence it cannot be but another name for the divine intellect itself, which is the only intellect having such knowledge. Immediately after the Ta'llqdt quotation, Mulla Sadra observes that Ibn Slna, in this very work, repeats his treatment of the simple intellect, the only difference being that he now stresses more systematically the causal, and hence not necessarily temporal, order of intelligibles. One gets the impression that Mulla Sadra here is referring to and summarizing Ta(llqdt^ pp. 121-2. (If this identification is correct, one must relate it to the second category of citation I distinguished at the beginning of this article.) For Mulla Sadra, such a conception of simple intellect is totally mistaken; this time he rejects Ibn Slna's point of view, but specifically on this very particular issue. I believe that this rejection is due to the fact that d'Ibn Sina: Un texte a revoir?', in Bulletin de philosophic medievale, 28 (1986), 163-77, esp. 164-7. 12 Note that the expression wa-qawlubu (p. 405, 1. 12) does not introduce a new citation, but rather continues a citation started a few lines before.

XIII io

the 'presential' aspect of intelligible forms13 cannot be genuinely articulated, once the Avicennian interpretation of the simple intellect is accepted. From this list of long citations, and their brief examination, we can draw some conclusions. First of all, one cannot but be struck by the extraordinary length of some of the citations. Secondly, Mulla Sadra undoubtedly agrees with some of the major ideas they express, and therefore does not hesitate to use them in order to sustain his own doctrine. Thirdly, the Ta'llqdt play a more than marginal role in the Asfdr. But what is the exact extent of the doctrinal impact of the quotations on Mulla Sadra's thought? This is both an important and very difficult question. In order to answer it, a detailed analysis would be required of all the citations, especially regarding the doctrinal elements involved in them. A similar analysis would be necessary regarding Mulla Sadra's comments and/or criticisms. Of course, such an investigation requires a monograph, but here we will have to limit ourselves to a very provisional assessment. From the foregoing, it seems reasonable to believe that Mulla Sadra greatly valued a few basic philosophical ideas of Ibn Slna, but did not agree with him in all respects. Even so, it seems he found some of Ibn Slna's wordings important enough to quote extensively, indicating that they still had some merit for him. After all, is it not a common truth that one refutes only what one considers to be worthy of refutation? But let me repeat that I am offering here only a provisional conclusion. Normally, one expects a paper to finish here. However, I have found in the Asfdr five quotations that Mulla Sadra explicitly attributes to the Ta'llqdt^ which I have tried in vain to locate. I will present them briefly one by one, and then try to determine whether or not it is possible that they do belong to the Ta'liqdt. A first quotation occurs in S. I, M. 1, c. 4 (i. 41), and reads as follows: 'If one asks whether the act of being [or 'existence'— al-wujud] is existent, or not existent, the answer is: it is existent insofar as the "true essence" [al-haqiqa] of the act of being [or 'of existence'] is its being existent. In fact, the act of being [or 'existence'] is "the being existent" [al-mawjudiyya]''. The same citation is given by Mulla Sadra in the K. al-Mashd'ir,14 where it is presented as being derived from the Ta'liqdt. But I have seen no fragment in the Ta'liqdt that might even in the loosest way be said to be its source. Moreover, 13 In this context, Mulla Sadra is not quite clear on this difficult issue. He simply refers 'elsewhere' for a more complete treatment, without giving any precise information. 14 See Corbin (ed.), Le Livre des penetrations metaphysiques § 47 (my translation differs slightly from Corbin's French translation).

XIII M U L L A S A D R A S USE OF I B N S I N A S TA(LlQAT

II

the very notion of mawjudiyya, mentioned at the end of the citation, definitely does not belong to the usual vocabulary of Ibn Slna, and is, as far as I can see, totally absent from the latter's writings. But since one finds other unusual notions and affirmations in the Ta'liqdt,15 one cannot exclude the possibility that Mulla Sadra may well have had that fragment in his manuscript of the Ta'liqdt. In that case, he clearly possessed another version, or at least a more comprehensive version than the one we have at present. A few pages further, in c. 5 (i. 47-8), two more so-called Ta'liqdt citations are offered. This time, they are also present in the K. al-Mashd'ir.16 The problem of existence, and more specifically the very particular nature of its 'accidentally', forms the key issue of both fragments. The first of them states: The existence of accidents in themselves is their existence in their substrates, except for the accident which [one calls] 'existence'. The latter differs from all other accidents: the other accidents are in need of existence for being existent; while existence is not in need of existence for being existent. Hence, it is not correct to affirm that the existence of existence in its substrate is its existence in itself, if this means that existence has existence in the very same way that whiteness has existence. But [one may affirm it] when it means that the existence of existence in its substrate is identical with the existence of its substrate. However, regarding the other accidents, their existence in their substrate is the existence of these accidents [literally 'that other'].

The second citation's wording is as follows: 'The existence, which belongs to a body, is the "act of being existent" [mawjudiyya] of the body. It is not as whiteness and body in the being [or 'becoming'] white, since whiteness and body are not enough for the being [or 'becoming'] white.' One is struck once more by the presence of the notion of mawjudiyya in the last citation. And even if the problem of the accidentally of existence is most central to Ibn Slna's thought, and is conceived therein as a very special kind of accidentality, there are nevertheless no formulations there (and more specifically in the Ta'liqdt) similar to the ones presented here. 15

See my Bonum?

In Metaphysics, VIII, 6, p. 412, 62, Avicenna affirms: "Necesse esse per se est bonitas pura." Because he is always existent, God, the Necessary Being, is both an efficient and a final cause. He not only affords beings their existence, but also preserves their existence as long as they last. Without God, nothing would exist, but God, the fullness of Being in itself, has no need whatsoever for anything outside himself. Hence, creation is a result of his goodness. This goodness is based on a will, but this will cannot but choose the best; otherwise God would not be perfect. In that sense, a necessity is involved. However, the primary sense of God's necessity is to be the fullness of Being,43 a pure 'actuality of existence.' For created being, it also possesses a certain actuality of existence, and therefore deserves to be called 'necessary,' although in a qualified way, i.e., ab alio. Moreover, Avicenna observes that the existence of an actual thing, even if in some respects it appears to be an evil, is always better than its non-existence; in other words, the existence of a creature is always a good.44 According to Henry, the divine ipsum esse is an infinitas, meaning the absence of any restriction in God, and a necessitas, signifying the identity of essence and existence in him, as Gomez Cafferena45 has well estab40. Aertsen recognizes the major importance of the Avicenna Latinus, but he seems to believe that the Avicennian notion of necesse should simply be dismissed (see o.c, p. 82). 41. See ibid., p. 82. Thomas' proper understanding of 'thing' as a transcendental is clarified ibid., pp. 193-198. 42. Henry of Ghent, Summa, art. 1, q. 2 (ed. Badius, fol. 4v). 43. See my "Avicenna: tussen neoplatonisme en islam," pp. 132-133. 44. Avicenna, Metaphysics, IX, 6, esp. pp. 500-501. 45. J. Gomez Caffarena, Ser participado y ser subsistente en la metafisica de Enrique de Gante (An. Greg., 93. Ser. Fac. Philos., B 8), Rome, 1958, pp. 157-180, esp. 163. According to Gomez, the basic texts in Henry's Summa in this respect are art. 23, q. 1 and art. 26, q. 1.

XVII 54 lished. Both constitute together two 'aspects' of one single reality. The notion of 'necessity' evokes an 'existential actuality,' whereas the notion of 'infinity' refers to an 'essential plenitude.' However, if Henry, with Avicenna, admits that one may pass from necessity to infinity, he does so because more fundamentally he has admitted a passage from infinity to necessity. The latter is undoubtedly inspired by Augustine and finds its ultimate source in the identification of the Ipsum Esse with the Ipsum Bonum. For Henry, the bonum ut subsistens belongs exclusively to God, the creator. Only in an analogical way may one say that created being is an hoc bonum, unqualified bonum being common to both creator and creatures.46 In both cases there seems to exist a direct link between necessitas and bonum as well as a radical convertibility with Being. Especially when one reads Avicenna's affirmation of book 8 through an Augustinian prism, as Henry seems to do, it is possible to perceive it as (pre-?)figuring the very idea of 'bonum as a transcendental.' I am aware that I have not offered a definite proof, but I believe that I have opened a valuable way for further research. Let me add that such investigation wil need to concentrate on the historical reception and interpretation(s) of Avicenna's very idea of 'necessity' in the Latin West before Henry. b) Res —> Verum?

According to Avicenna, the sentence "res est id de quo potest aliquid vere enuntiari"47 is not a proper definition of res, but can nevertheless be accepted as a designation8 For him, "res" designates both real things and things existing only in the mind. Related to this most encompassing concept (every existent being is a thing, but not inversely!), Avicenna distinguishes between the universal as universal, on the one hand, and the universal as a thing to which universality is attached, on the other, as exemplified by the notion of equinitas.^ J. Aertsen50 has already shown that Henry's concept of res /ratitudo is 46. Henry of Ghent, Summa, art. 24, q. 6 (ed. Badius, fol. 142v). 47. Avicenna, Metaphysics, I, 5, pp. 33, 37 and 34, 46.

48. Ibid., P. 34, 48-50.

49. For a much more detailed and precise analysis of Avicenna's notion of 'thing,' see J. Jolivet, "Aux origines de 1'ontologie d'Ibn Sina," in: J. Jolivet - R. Rashed (dir.), Etudes sur Avicenne, Paris, 1984, pp. 11-28; R. Wisnovsky, "Notes on Avicenna's Concept of Thingness (sha 'iyya)," in: Arabic Sciences and Philosophy, 10 (2000), pp. 181221; Th.-A. Druart, " Shay' or Res as Concomitant of 'Being' in Avicenna," in: Document! e Studi sulla tradizione filosofica medievale, 12 (2001), pp. 125-142.

XVII ELEMENTS OF AVICENNIAN METAPHYSICS IN THE SUMMA

55

Avicennian (as well as partly Augustinian) inspired. A most significant passage in this respect one finds in Avicenna Latinus, Metaphysics, I, 5, p. 35, 63-64: "res habet certitudinem propriam quae est eius quidditas." Henry observes that for Avicenna the "esse quidditativum" is an "esse divinum quoniam sui esse .. est divina (Av. Lat.\ Dei) intentio," but immediately adds: "intelligo quia habet rationem formalem et exemplarem in Deo."51 Once more, Henry interprets an Avicennian doctrine in an explicitly Augustinian sense. Whatever its precise doctrinal impact, it is sure that res understood in this way can be converted with ens. But can or may res be considered as being equivalent with verum? A possible basis for such an identification might come from Avicenna Latinus, Metaphysics, I, 8, p. 55, 62, respectively, pp. 55, 64 - 56, 67: "necesse esse est id per quod seipsum est veritas semper," "veritas autem quae adaequatur rei, ilia est certa, sed est certa, ut puto, respectu suae comparationis ad rem, et est veritas respectu comparationis rei ad ipsam," combined with the affirmations "certitude est res" (I, 5, p. 35, 72) and "veritas enim cuiusque rei est proprietas sui esse quod stabilitum est ei" (VIII, 6, p. 413, 83-84). These affirmations taken together seem to imply that for Avicenna (at least insofar as the Avicenna Latinus is concerned52), 'thing,' insofar as it is a 'fixed (stabilitum) being,' and 'truth' are convertible. It should be noted that a 'fixed being' must be qualified in Avicennian terms as 'necessary,' either by itself or ab alio. It concerns only real, not purely mental being. Surely, Avicenna (also Avicenna Latinus) never calls God explicitly 'thing,' but the above quoted formulation, which states that the "necesse esse est id...," opens a way for such an affirmation. Moreover, the very fact that for Avicenna, as indicated earlier, res is the primary transcendental, implies inevitably that God is 50. J.A. Aertsen, "Transcendental Thought in Henry of Ghent," in: W. Vanhamel (ed.), Henry of Ghent, pp. 1-18, esp. 9-10; id., Medieval Philosophy and the Transcendental, PP. 198-199. 51. Henry of Ghent, Summa, art. 43, q. 2 (ed. L. Hodl, Opera Omnia, XXIX, Leuven, 1998, p. 53, 60-62), quoting Avicenna, Metaphysics, V, 1, p. 237, 25-26. 52. Especially regarding the first two quotations, the Latin offers a somewhat different formulation from theArabic, which affirms: "The Necessary Existent is the Truth by Itself, always" and "As to the truth by conformity it is like the veracious, except that the veracious, as I believe, [has to be conceived] with respect to its relation to the thing, and the truth [has to be considered] with respect to the relation of the thing to it" (see Shifa,' Ilahiyyat, ed. G.C. Anawati, S. Zayed, M.Y. Musa and S. Dunya, Cairo, 1960, p. 48).

XVII 56

somehow a res.53 Henry, in turn, clearly accepts that res / ratitudo can be converted with 'the true.'54 One of Henry's most fundamental statements in this respect runs as follows: "quanto enim aliquid in re plus habet ratitudinis sive firmitatis, tanto plus habet entitatis, quare et veritatis."55 Regarding res in the sense of reor - reris which includes fictional beings, J. Aertsen has convincingly shown that in Henry it is a kind of supertrancendental concept.56 c) Unum Henry rejects the ultra-realistic interpretation of Avicenna's concept of the one, which because of Averroes' influence was common in the Middle Ages.57 According to Henry, Avicenna defends, as Aristotle had doen before him, the convertibility of "unum" and "ens." Based on Avicenna Latinus, Metaphysics, III, 3, p. 117, 83-85, 87 and p. 119, 20-21 ("[unitas] non enim recipitur in certificatione quidditatis alicuius substantiarum," and "unitas vero est intentio quae est accidens"), Henry concludes (art. 25, q. 1): De quo plane patet quod Avicenna posuit cum Philosopho quod unum non significat re aliud ab ente, et ita quod non significat accidens additum enti, sicut est dispositio in albo et nigro ut imposuit ei Averroes, sed quod significat aliam intentionem additam intention! entis, quam secundum modum suum loquendi appellat ac53. It is worth mentioning that in Islamic kalam (theology), some authors, e.g., Ibn Kullab and Hisham b. Al-Hakam, did not hesistate to designate God by the term shay,' 'thing.' See H. Daiber, Das theologisch-philosophische System des Mu'ammar ibn Abbad asSulami, Beirut - Wiesbaden, 1975, pp. 190-191. 54. See J.A. Aertsen, "Transcendental Thought in Henry of Ghent," p. 16. 55. Henry of Ghent, Summa, art. 34, q. 2 (ed. R. Macken, p. 175). 56. See J.A. Aertsen, Transcendental Thought in Henry of Ghent, pp. 17-18. Although Bonaventure's and Thomas' Commentaries on the Sentences are most probably Henry's direct source (see ibid., pp. 11-14), one may wonder whether the Avicenna Latinus, Metaphysics, e.g., I, 5, p. 36, 81-82 ("intellectus de ente semper comitabitur illam [sc. rem cuius quaerimus intentionem], quia ilia habet esse vel in singularibus vel in aestimatione vel intellectu") has not had direct influence on the elaboration of this notion as well. Let me observe that Avicenna, as Henry does, seems to include fictional beings among things in the broader sense. However, Avicenna characterizes them as being 'impossible.' Herewith a tension, not to say a contradiction, arises in his doctrinal system. See D. Black, "Avicenna on the Ontological and Epistemic Status of Fictional Beings," in: Document! e Studi sulla tradizione filosofica medievale, 8 (1997), pp. 425-453. 57. See A. de Libera, "D'Avicenne a Averroes et retour. Pour les sources arabes de 1'un transcendantal," in: Arabic Sciences and Philosophy, 4 (1994), pp. 149-171. Although this is a very fundamental and encompassing study, no mention of Henry of Ghent is made.

XVII ELEMENTS OF AVICENNIAN METAPHYSICS IN THE SUMMA

57

cidens, vocando accidens rei omne quod convenit rei extra certitudinem quidditatis et intentionem eius.58 For Henry, Avicenna remains faithful to Aristotle. He is aware of the fact that Avicenna's use of "accidens," when used regarding the notion of one, requires a particular interpretation. In fact, the (metaphysical) one is such that, although it may be considered to be an addition of a substance's quiddity, that substance nevertheless cannot exist without it.59 As such, one and being are convertible.60 d) Ens When one interprets Avicenna in a strictly monistic sense, the notion of "Being" can hardly function as a real transcendental in his thought. However, Henry seems to doubt whether this is the correct interpretation: "et habendo respectum ad distinctum intellectum Dei aut creaturae intellexit Avicenna, si bene intellixit, quod intellectus entis prior est intellectu Dei aut creaturae."61 However, Henry remains hesistant — "si bene intellexit" — indicating that Avicenna's position on this matter is not entirely clear. And, indeed, a superficial reading of Avicenna may give the impression that he defends an univocal concept of Being.62 But as soon as one more profoundly examines his metaphysics, one cannot but recognize that there exist serious indications for the acceptance of a transcendental analogy of Being.63 Avicenna takes great care not to confuse the being of the creator with that of the creatures, but 58. Henry of Ghent, Summa, art. 25, q. 1 (ed. Radius, fols. 147r-v). 59. See Avicenna, Metaphysics, II, 3, esp. pp. 118 sqq. 60. It must be observed that Henry, in Quodlibet, I, 9, still adhered the ultra-realistic interpretation of Avicenna's concept of one, see J. Paulus, Henri de Gand. Essai sur les tendances de sa Metaphysique, Paris, 1936, p. 229. 61. Henry of Ghent, Summa, art. 21, q. 2 (ed. Radius, fol. 124v). 62. J. Decorte, "Henry of Ghent on Analogy. Critical reflections on Jean Paulus' interpretation," in W. Vanhamel (ed.), Henry of Ghent, pp. 71-105, esp. 88, believes that according to Henry Avicenna was holding a univocal conception of Reing. However, Henry only mentions Avicenna after having evoked the correct opinion of Aristotle. My interpretation is more in line with J. Gomez Caffarena, Ser participado y ser subsistente en la metafisica de Enrique de Gante, p. 183, but I disagree with him when he states: "acomoda asi Enrique al arabe a su propia opinion." 63. See my Avicenna: tussen neoplatonisme en islam, pp. 133-140, and my "Rahmanyar ibn Marzuban: A Faithful Disciple of Ibn Sina?," in D. Reisman (ed), Avicenna: Before and After. Leiden, Rrill, 2002, n. 33 (in the press). See also G. Verbeke, Le statut de la metaphysique, introd. doctrinale a S. Van Riet (ed), Avicenna Latinus. Liber de philosophia prima seu scientia divina I-IV. Louvain, Peeters; Leiden, Rrill, 1977, p. 67*.

XVII 58

both nevertheless are an integral part of the science of metaphysics, the study of 'being qua being.' If this should be the correct interpretation of Avicenna's thought, 'being' figures undoubtedly as a kind of transcendental in his Metaphysics. At the same time, one must recognize that it surely does not possess the central place it later will occupy in the classical Latin doctrine of the four transcendentals. In conclusion, I had no other intention than to show that Avicenna's primary notions, more particularly as expressed in the Avicenna Latinus, can be 'linked' with each of the four classical transcendentals of medieval Latin thinkers. Especially, if one reads Avicenna the way that Henry understands him, one gets the impression that Avicenna's notions were more than just a 'starting-point.' To what degree exactly, from both historical and doctrinal points of view, they contributed to the elaboration of that doctrine, however, must be left open for further investigation. But it is already clear that Henry offers interesting insights and hints regarding how to explore and interpret the Avicennian materials. (3) Metaphysics as a science

In a recent study,64 Martin Pickave has fundamentally elaborated this topic, indicating in a well-considered way Avicenna's influence on Henry in this particular matter. Let me just add two remarks: 1. The very project of metaphysics, being primarily an ontology, but including also an aitiology and a theology, and forming as such one integrated system, seems to have its fundamental roots in Avicenna. Aristotle had clearly not yet elaborated a metaphysical system as 'system.' 2. The simpliciter of the expression "ens simpliciter" may have been derived by Henry from Avicenna, Metaphysics, I, 2, p. 14, 60: "ens in se abolute non habet principium." III. General Conclusion In his Summa, Henry of Ghent consciously makes use of several passages of Avicenna's Metaphysics. Their amount is perhaps not impressive, but they play an important role in the elaboration of such fundamental doctrines as the establishment of a metaphysical proof for God, the four transcendentals (and, possibly, the idea of a super-transcendental) and the 64. M. Pickave, "Heinrich von Gent liber das Subjekt der Metaphysik als Ersterkanntes," in: Document! e Studi sulla tradizione filosofica medievale, 12 (2001), pp. 493-522.

XVII ELEMENTS OF AVICENNIAN METAPHYSICS IN THE SUMMA

59

very subject of the science of metaphysics. This list is surely not exhaustive. A quick survey reveals that Henry deals in a more than a secondary way with Avicenna's theories of substance, relation and causality, to mention only the most striking examples. What is most surprising is Henry's careful reading of Avicenna's texts, a reading which permits him to reject a too Averroistic reading of Avicenna. One wonders whether such reading is particular to him, or whether some predecessor similarly interpreted these Avicennian doctrines. Even then, one cannot but admire the great efforts Henry made in order to interpret correctly such a difficult text as Avicenna's Metaphysics.

INDEX Names have been standardized according to their English spelling. 'Abd al-Jabbar: VI 117 'Ala al-Dawla: II 253; VI 119 Albert the Great: I, 11-12, 14; XV 93 Alexander of Aphrodisias: I 12; VI 122; XV 90-97, 105 Aristotle: I 2, 6; II 253, 256, 261, 264; IV 475; V 864; VIII 36, 39; IX 43, 58, 62; X 2-3, 5, 8, 10, 14; XI 40; XII 180, 182, 196; XV 89, 97-8, 100-104; XVI 155, 157, 165-6; XVII 51, 56-7 Augustine: I 7; XVI 155; XVII 48, 54 Radius: XVII 45-7 Bahmanyar ibn Marzuban: I 7, 9, 13; VI 109-10, 117-18; XII/*xss*m;XIII2, 12 Bahrain ibn Khurshld ibn Yazdyar, Abu Mansur: XII 178 al-BIrum: XIII 1 Blund, John: I 10-11 Boethius: I 2, 6; XVI 155 Bukhara: I 1 Descartes: II 254 De spiritu et anima: XVI 159-61 Duns Scotus: I 6 Empedoclus: XIV 126 al-Farabl: I 6; II 252, 254-5, 263-5, 268; IV 475; V 866, 869;

VI 119; IX 45, 48, 60, 62-3; X 5, 8-9; 12-13; XI 39-40; XV 89

Galen: X 8, 13; XI40 al-Gharib al-Isfaham: XI 40 al-Ghazzall: 14-5, 9-10; VII163, 167, 170, 175; VIII-XI passim; XII179; XVI163, 168 Gundissalinus: I 5, 11 Hamadhan:Il;VI119 Heidegger: I 6 Henry of Ghent: I 2, 6-7; XVI-XVII passim Hippocrates: X 2 IbnAbIUsaybra:XIII12 Ibn Aqnln, Joseph: I 5 Ibn Daud: I 5 Ibn Gabirol: I 7 Ibn Hazm: II 263 Ibnal-NadIm:XV91 Ibn Rushd (Averroes): I 2, 5, 7; VI 120; XV 89, 93; XVI 155, 158, 168; XVII56 Ibn Slna (Avicenna): I-XVII passim Ikhwanal-Safa':X 16 IbnZayla: VI 109-10; XIII12 Jean de la Rochelle: I 11 al-Juzjani: II253; VI119-20; VII168

2

al-Kindl: II 265, 267-8; IV 475; V 863-4,869;X 8,12 al-Kirmam, Abu 1-Qasim: XII 177, 196 al-Kirmam, Hamld al-DIn: IV 466 al-Lawkarl: I 7, 13; XIII 2, 6 Leibniz: II 254 Liber de Causis: I 2; II 265-6; IV 459; X, 13 Matthew of Aquasparta: 111 Michael Scot: I 12 Mir Damad: I 3 Miskawayh: VI 122 Muhammad: VIII 46-7 al-Muhasibl: VIII 53; XI 37 Mulla Sadra ShlrazI: I 3, 5, 13; XIII passim al-Mutanabbl: VIII 39 al-Natill, Abu 'Abd Allah: II252 Neoplatonica Arabica: IV 465, 468, 470, 472 Peter of Auvergne: 17 Peter Olivi: I 14 Philoponus, John: I 12; II 267; X 8; XV 90-92, 94, 97-105 Plato: II 253, 256; IV 464; X 2, 7; XII 190; XVI155; XVII48 Plotinus: II 253, 256-7 Porphyry: VI 120; XIII 7 Proclus: II 262-3, 266; IV 467; X 14 Qur'an: II 254-5; 260, 263-4, 267-9; IV 460, 469-70; V 869; VIII43 al-Qushayrl: XI 37 Rayy:II252;X13

]INDEX

al-RazI,F.D.:I5, 10, 13 Richard Rufus: I 14 al-Shafi'T: VIII 37 Shahrastani: I 5; VI 121; VIII 36, 39 Shams al-Dawla: II 252-3 Siger of Brabant: 111 al-Sijistanl: IV 466 Socrates: X 2 al-Suhrawardl: I 3 Taj al-Mulk: II 253 Themistius: I 12; VI 122; XV 90-92, 97-102 Theologia Aristotelis: IV 457-60, 462, 465; VI 120-21; XII 182; XIII 4; XV, 89 Thomas Aquinas: I 2-3, 8, 11, 14; II 254; XVII 52-3 Toledo: I 5 al-TusT,N.D.:I3 Ulrich of Strasbourg: I 14 Vital du Four: I 11 William of Auvergne: 15,11 Yahya: XV 92, 98 Manuscripts: Dubrovnik, Bibl. Dom. 20 (36-V-5): XIV 114, 116, 118-19, 125 Mazarine 3473: XIV 122 Meshkat 1037/8: VIII 34-5 Meshkat 1149: VIII 34-5 Paris, B.N. Lat. 16604: XIV 120 Venice, Bibl. S. Marco Lat. 2665: XIV118, 120