(Scientia Graeco-Arabica, 7) Dag Nikolaus Hasse, Amos Bertolacci, (Eds.) - The Arabic, Hebrew and Latin Reception of Avicenna's Metaphysics (2011, Walter de Gruyter)

The Arabic, Hebrew and Latin Reception of Avicenna’s Metaphysics
Scientia Graeco-Arabica
herausgegeben von
Marwan Rashed
Band 7
De Gruyter
The Arabic, Hebrew and
Latin Reception of
Avicenna’s Metaphysics
edited by
Dag Nikolaus Hasse and Amos Bertolacci
De Gruyter
Gedruckt mit freundlicher Unterstützung der VolkswagenStiftung.
ISBN 978-3-11-021575-5
e-ISBN 978-3-11-021576-2
ISSN 1868-7172
Library of Congress Cataloging-in-Publication Data
The Arabic, Hebrew and Latin reception of Avicenna’s Metaphysics /

edited by Dag Nikolaus Hasse and Amos Bertolacci.
p. cm. ⫺ (Scientia graeco-arabica)
Includes bibliographical references and index.
ISBN 978-3-11-021575-5 (hardcover : alk. paper)
1. Avicenna, 980⫺1037. Ilahiyat. 2. Metaphysics 3. Islamic philo-
sophy. 4. Jewish philosophy. 5. Philosophy, Medieval. I. Hasse,
Dag Nikolaus. II. Bertolacci, Amos.
B751.I483A73 2011
110⫺dc22
2011007822
Bibliographic information published by the Deutsche Nationalbibliothek

The Deutsche Nationalbibliothek lists this publication in the Deutsche
Nationalbibliografie; detailed bibliographic data are available in the Internet
at http://dnb.d-nb.de.
쑔 2012 Walter de Gruyter GmbH & Co. KG, Berlin/Boston

Printing and binding: Hubert & Co. GmbH & Co. KG, Göttingen
⬁ Printed on acid-free paper
Printed in Germany
www.degruyter.com
1
2
3
4 Preface
5
6 The articles of this volume are presented by scholars who convened in 2008 to
7 discuss their research on the influence of Avicenna’s metaphysics in the Villa
8 Vigoni, Centro italo-tedesco, Menaggio, Italy. We are grateful to the participants
9 and chairs of the congress who do not contribute to this volume: Rdiger
10 Arnzen, Gad Freudenthal, Dimitri Gutas, Maarten Hoenen and Andreas Speer.
11 Special thanks go to two research assistants at the University of Wrzburg:
12 Anna-Katharina Strohschneider, who prepared the papers for type-setting, and
13 Jon Bornholdt, who extinguished mistakes in the texts of contributors who are
14 not native speakers of English. We gratefully acknowledge the generous and
15 unbureaucratic funding of the conference by the VolkswagenFoundation, as
16 part of the Lichtenberg professorship grant to Dag Nikolaus Hasse. Finally, we
17 would like to thank Gregor Vogt-Spira and the staff of the Villa Vigoni who
18 created an ideal atmosphere for the discussion of a research issue of considerable
19 complexity.
20
21 Dag Nikolaus Hasse Amos Bertolacci
22 Julius-Maximilians-Universitt Scuola Normale Superiore
23 Wrzburg di Pisa
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
Contents
Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . V
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
Jules Janssens
Al-Lawkarı̄’s Reception of Ibn Sı̄nā’s Ilāhiyyāt . . . . . . . . . . . . . . . . . . . 7
Robert Wisnovsky
Essence and Existence in the Eleventh- and Twelfth-Century Islamic
East (Mašriq): A Sketch . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
Stephen Menn
Fārābı̄ in the Reception of Avicenna’s Metaphysics: Averroes against
Avicenna on Being and Unity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
Peter Adamson
Avicenna and his Commentators on Human and Divine Self-Intellection 97
Heidrun Eichner
Essence and Existence. Thirteenth-Century Perspectives in
Arabic-Islamic Philosophy and Theology . . . . . . . . . . . . . . . . . . . . . . . 123
Mauro Zonta
Avicenna’s Metaphysics in the Medieval Hebrew Philosophical Tradition 153
Resianne Fontaine
‘Happy is he whose children are boys’: Abraham Ibn Daud and
Avicenna on Evil . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159
Mauro Zonta
Possible Hebrew Quotations of the Metaphysical Section of Avicenna’s
Oriental Philosophy and Their Historical Meaning . . . . . . . . . . . . . . . . . 177
Amos Bertolacci
On the Latin Reception of Avicenna’s Metaphysics before Albertus
Magnus: An Attempt at Periodization . . . . . . . . . . . . . . . . . . . . . . . . . 197
Dag Nikolaus Hasse
Avicenna’s ‘Giver of Forms’ in Latin Philosophy, Especially in the Works
of Albertus Magnus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 225
Kara Richardson
Avicenna and Aquinas on Form and Generation . . . . . . . . . . . . . . . . . 251
VIII Contents
Pasquale Porro
Immateriality and Separation in Avicenna and Thomas Aquinas . . . . . 275
Gabriele Galluzzo
Two Senses of ‘Common’. Avicenna’s Doctrine of Essence and Aquinas’s
View on Individuation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 309
Martin Pickav
On the Latin Reception of Avicenna’s Theory of Individuation . . . . . 339
Giorgio Pini
Scotus and Avicenna on What it is to Be a Thing . . . . . . . . . . . . . . . . 365
Index of Avicenna’s Works with Passages Cited . . . . . . . . . . . . . . . . . . 389
Index of Names . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 395
1
2
3
4 Introduction
5
6 Many centuries passed after the composition of Aristotle’s Metaphysics before a
7 metaphysical work of similar size and ambition was written in the Peripatetic
8 tradition: Avicenna’s Ila-hiyya-t (Metaphysics, or: Divine Things), the fourth and
9 last part of the summa Kita-b aš-Šifa- (Book of the Cure), dating to 1020 – 27 CE.
˘
10 The Ila-hiyya-t is only one of more than a dozen metaphysical works by Avicenna,
11 but in terms of comprehensiveness, systematic effort and influence, it is his most
12 important metaphysical text. It is rivaled only by the Kita-b al-Išārāt wa-l-
13 tanbı̄hāt (Book of Pointers and Reminders), a late summa dating to ca. 1030 – 34
14 CE that contains a substantial metaphysical section of considerable influence in
15 the Arabic tradition.
16 The study of Avicenna’s metaphysics has made important progress in the
17 past few years, due in part to the appearance of studies in monograph format.1
18
Much, however, remains to be done. Above all, critical editions of Avicenna’s
19
metaphysical works are still lacking,2 and the study of their manuscript tradition
20
is still at a preliminary stage.3 The present book sheds light on Avicenna’s
21
metaphysics itself, but its proper theme is the reception of his metaphysics in
22
three different cultures: Arabic, Hebrew and Latin.
23
In the past few decades, it has increasingly become recognized that
24
Avicenna’s philosophy, and in particular his metaphysics, was of overwhelming
25
influence in the Arabic-speaking world from the eleventh to, at least, the
26
27
28 1 Among recent studies, particularly important are R. Wisnovsky, Avicenna’s Metaphysics in
29 Context, London: Duckworth, 2003, and A. Bertolacci, The Reception of Aristotle’s
‘Metaphysics’ in Avicenna’s ‘Kitāb al-Shifā ’, Leiden: Brill, 2006. A helpful tool is the
˘
30
collection of articles by M.E. Marmura, Probing in Islamic Philosophy, Binghamton:
31 Global Academic Pub., 2005, as well as recent translations of the Ila-hiyya-t into English
32 by M.E. Marmura (Avicenna, The Metaphysics of ‘The Healing’, Provo, Utah: Brigham
33 Young University Press, 2005) and Italian by O. Lizzini with a preface by P. Porro
34 (Avicenna, Metafisica: la scienza delle cose divine, 2nd edn, Milan: Bompiani, 2006) and
A. Bertolacci (Avicenna, Il libro della guarigione: Le cose divine, Torino: Utet, 2008).
35
2 See the list of Emendanda of Anawati’s Arabic edition of the Ila-hiyya-t (Avicenna, al-Šifā’,
36 al-Ilāhiyyāt, Cairo: al-Hay’a al-‘āmma, 1960) in Bertolacci, The Reception (as in n. 1).
37 3 On the Ila-hiyya-t, see A. Bertolacci, On the Manuscripts of the Ilāhiyyāt of Avicenna’s
Kitāb al-Šifa- , in A. Akasoy, W. Raven, eds, Islamic Thought in the Middle Ages. Studies in
˘
38
39 Text, Transmission and Translation, in Honour of Hans Daiber, Leiden: Brill, 2008,
pp. 59 – 75, and on the Šifa- in general, the papers presented at the International
˘
40
Colloquium The Manuscript Tradition of Avicenna’s Kita-b aš-Šifa- : The Current State of
˘
41 Research and Future Prospects, Scuola Normale Superiore di Pisa, 22 – 24 September 2010
42 (proceedings forthcoming in Oriens, 40, 2012).
2 Introduction
1 sixteenth century. Even those thinkers who fundamentally disagreed with

2 Avicenna often developed their intellectual standpoint in confrontation with the
3 Avicennian tradition. Indications of the scope of Avicenna’s influence in Arabic
4 are the many extant commentaries on his works, the lively reception of his
5 metaphysical theories by Islamic theologians, and the reading of his philosophy
6 in the madrasa.4 Research on the later period of Arabic philosophy after
7 Avicenna is still in its infancy. This is particularly true of the centuries after Fahr
8 al-Dı̄n al-Rāzı̄, who died 1210 CE. Lists have been drawn up of the authors and ˘
9 works that should be studied and researched, but since the details of the
10 intellectual history of the period are still unknown, the time has not yet come
11 for generalizations. It is an open question, for example, which of Avicenna’s texts
12 transmitted which theories. As in the case of al-Lawkarı̄, it may well be that an
13 author read several metaphysical texts by Avicenna: Ila-hiyya-t, al-Mabda’ wa-l-
˘
14 ma‘ād, Išārāt, Ta lı̄qāt, etc. (see the article by Janssens in this volume). Also, it
15 can be shown that interpretations of Avicenna were often influenced by previous
16 readers (see the articles by Wisnovsky and Menn). Turning to thirteenth- and
17 fourteenth-century sources, one realizes that Avicennian theories were often
18 transmitted through intermediate sources, such as the influential works of Fahr
19 al-Dı̄n al-Rāzı̄, or through philosophical handbooks (see the articles by Adamson ˘
20 and Eichner). Despite the fact that so much remains unknown about this
21 period, it is already apparent that the reception of Avicenna’s metaphysical
22 theories in later Arabic thought gave rise to a wealth of metaphysical discussions
23 of impressive intellectual quality.
24
The textual transmission of Avicenna’s metaphysics in the Latin speaking
25
world is better known. There are solid grounds for believing that Dominicus
26
Gundisalvi, also called Gundissalinus by the scholastics, an archdeacon and
27
canon of the cathedral of Toledo, was the translator of Avicenna’s Ila-hiyya-t from
28
Arabic into Latin between 1150 and 1180.5 The Latin title was Liber de
29
30
4 Recent studies: D. Gutas, The Heritage of Avicenna: The Golden Age of Arabic
31 Philosophy, 1000 – ca. 1350, in J. Janssens, D. de Smet, eds, Avicenna and His Heritage,
32 Leuven: Leuven University Press, 2002, pp. 81 – 97; R. Wisnovsky, The Nature and
33 Scope of Arabic Philosophical Commentary in Post-Classical (ca. 1100 – 1900 AD)
34 Islamic Intellectual History: Some Preliminary Observations, in P. Adamson et al., eds,
Philosophy, Science and Exegesis in Greek, Arabic and Latin Commentaries, London:
35
Institute of Classical Studies, 2004, vol. 2, pp. 149 – 91; G. Endreß, Reading Avicenna in
36 the Madrasa: Intellectual Genealogies and Chains of Transmission of Philosophy and
37 the Sciences in the Islamic East, in J.E. Montgomery, ed., Arabic Theology, Arabic
38 Philosophy. From the Many to the One: Essays in Celebration of Richard M. Frank, Leuven:
39 Peeters, 2006, pp. 371 – 422; H. Eichner, Dissolving the Unity of Metaphysics: From
Fakhr al-Din al-Razi to Mulla Sadra al-Shirazi, Medioevo, 32, 2007, pp. 139 – 97.
40
5 Dominicus Gundisalvi is identified as the translator in the colophon of three of the 25
41 manuscripts. See Avicenna, Liber de philosophia prima sive Scientia divina, I – IV, ed.
42 S. Van Riet, Louvain/Leiden: Peeters/Brill, 1977, p. 123*, n. 2.
Introduction 3
1 philosophia prima sive scientia divina. This text gradually found its readers in the
2 Latin West (see the article by Bertolacci) and reached the high point of its
3 influence in the period from Thomas Aquinas to John Duns Scotus (as
4 evidenced in the articles of Richardson, Porro, Galluzzo, Pickav and Pini). The
5 manuscript transmission of the Philosophia prima thins out considerably after
6 1400 CE: 15 manuscripts are extant from the thirteenth century, 7 from the
7 fourteenth, 3 from the fifteenth.6 But Avicenna’s doctrines continued to be
8 discussed in the Renaissance (see the article by Hasse).
9 The Latin reception of Avicenna’s metaphysics was also influenced by the
10 translation of al-Ġaza-lı̄’s Maqa-sid al-fala-sifa (Intentions of the Philosophers) into
11 Latin. The Maqa-sid are to a large ˙ degree an intelligent reworking of Avicenna’s
˙
Persian Da-nešna-me-ye Ala- ı̄ (Philosophy for Ala--al-Dawla) and thus exhibit basic
˘
12
˘
˘
˘
13 teachings of Avicenna, though not always faithfully. Since, until the early
14 fourteenth century, this was the only text by al-Ġaza-lı̄ known in the Latin West,
15 the scholastics read al-Ġaza-lı̄ as a sequax Avicennae. Another influence on the
16 Western reading of Avicenna’s metaphysics was Averroes’ Long commentary on
17 Aristotle’s Metaphysics, which contains several passages that engage in criticism of
18 Avicenna’s metaphysical theories.
19 There exist a good number of in-depth studies on the reception of Avicenna
20 by individual scholastic authors and on the reception of certain Avicennian
21 theories, e. g. on the primary notions or the subject matter of metaphysics,7 but
22 scholarship has not yet arrived at a comprehensive picture of Avicenna’s
23 influence on Latin metaphysics. The present volume is meant as a contribution
24 to such a picture. The importance of the issue is widely recognized, in view of
25
the pivotal significance of Avicenna for the formation of metaphysical thought
26
in high scholasticism.
27
The Ila-hiyya-t of The Cure was not translated into Hebrew, but Avicenna’s
28
metaphysics nevertheless influenced medieval Jewish thought.8 This influence
29
30
6 Avicenna, ibid., p. 127*.
31 7 To mention only a few studies: J. Aertsen, Avicenna’s Doctrine of the Primary Notions
32 and its Impact on Medieval Philosophy, in Akasoy, Raven (as in n. 3), pp. 21 – 42;
33 A. Zimmermann, Ontologie oder Metaphysik? Die Diskussion ber den Gegenstand der
34 Metaphysik im 13. und 14. Jahrhundert; Texte und Untersuchungen, Leuven: Peeters,
1998; J.F. Wippel, The Latin Avicenna as a Source for Thomas Aquinas’s Metaphysics,
35
Freiburger Zeitschrift fr Philosophie und Theologie, 37, 1990, pp. 65 – 72. See the
36 bibliography in P. Porro’s preface to Lizzini’s Italian translation of the Ila-hiyya-t (as in
37 n. 1) and J. Janssens’ An Annotated Bibliography on Ibn Sı̄na-, Leuven: Leuven University
38 Press, 1991, with its First Supplement, Louvain-la-Neuve: FIDEM, 1999. See also the
39 collection of articles on the Latin transmission of Avicenna’s works by M.-Th. d’Alverny,
Avicenne en occident, Paris: Vrin, 1993.
40
8 S. Harvey, Avicenna’s Influence on Jewish Thought: Some Reflections, in Y.T.
41 Langermann, ed., Avicenna and his Legacy. A Golden Age of Science and Philosophy,
42 Turnhout: Brepols, 2009, pp. 327 – 40.
4 Introduction
1 occurred through Jewish philosophers reading Avicenna in Arabic, the partial

2 translation of Avicenna’s Kita-b an-Nağa-t (Book of Salvation) into Hebrew, and –
3 indirectly – the Hebrew translations of Averroes and al-Ġaza-lı̄, in particular the
4 translation of al-Ġaza-lı̄’s Intentions of the Philosophers (see the articles by Zonta).
5 Some topics of this reception have received detailed attention in scholarship,
6 such as Maimonides’ usage of several Avicennian doctrines: the proof of God’s
7 existence, the concept of the necessary being, and the distinction between
8 essence and existence.9 Other issues remain to be settled, such as the
9 identification of the Avicennian texts read and quoted by Jews directly from
10 the Arabic, for example by Abraham ibn Daud, the Andalusian philosopher of
11 the twelfth century (see the article by Fontaine), or the transportation of
12 Avicennian ideas through Hebrew commentaries on al-Ġaza-lı̄’s Intentions of the
13 Philosophers. An important question is whether there is evidence for a current of
14 ‘Jewish Avicennism’ in the fourteenth century rivaling the dominant philo-
15 sophical current influenced by Averroes, as Mauro Zonta has suggested.10 It is
16 clear, however, that Avicenna’s overall influence in Hebrew philosophy was never
17 on the same scale as that of Averroes.
18 Some articles of the present volume are devoted to the textual transmission
19 of Avicenna’s metaphysics in different cultures (Janssens, Zonta, Bertolacci).
20 Others study the reception of several prominent doctrines of Avicenna: the
21 distinction between essence and existence (Wisnovsky, Eichner, Porro); the
22 doctrine of primary notions (Menn, Pini); the theory of individuation and
23 universals (Galluzzo, Pickav); emanation theory and the related topics of the
24 necessary being, the emanation of the intelligences, the nature and function
25 of the giver of forms, and the origin of evil (Adamson, Hasse, Richardson,
26 Fontaine). It is the explicit aim of the present volume to facilitate the comparison
27 between the reception processes in distinct cultures and times, thus contributing
28 to our knowledge both of Avicenna’s metaphysics itself, through the lenses of his
29 medieval readers, and of its culturally complex reception history.
30
31
32
33 9 Among the many studies on this topic, only two shall be mentioned: M. Zonta,
34 Maimonides’ Knowledge of Avicenna. Some Tentative Conclusions About a Debated
Question, in G. Tamer, ed., The Trias of Maimonides/Die Trias des Maimonides. Jewish,
35
Arabic, and Ancient Culture of Knowledge/Jdische, arabische und antike Wissenskultur,
36 Berlin/New York: de Gruyter, 2005, pp. 211 – 22; and the chapter on the influence of
37 Avicenna’s metaphysical proof of God in the classic study by H.A. Davidson, Proofs for
38 Eternity, Creation and the Existence of God in Medieval Islamic and Jewish Philosophy, New
39 York/Oxford: Oxford University Press, 1987, pp. 378 – 406.
10 M. Zonta, The Role of Avicenna and of Islamic ‘Avicennism’ in the 14th-Century Jewish
40
Debate Around Philosophy and Religion, Oriente moderno, 80, n.s. 19, 2000, pp. 647 –
41 60, and M. Zonta, Avicenna in Medieval Jewish Philosophy, in Janssens, De Smet (as in
42 n. 4), pp. 267 – 79.
Introduction 5
1 The Avicennian doctrines studied in this volume are important, but they
2 also represent a somewhat accidental choice, since they reflect what several
3 contemporary scholars in the field are currently working on. Equally important
4 issues of Avicennian metaphysics (for example, the subject-matter of meta-
5 physics, the proof of God’s existence, the metaphysically grounded
6 prophetology, and the theory of substance and accident) are only treated
7 cursorily. In this connection it is important to remember that Avicenna’s
8 metaphysics is very rich in content and coherent in structure and ought not to
9 be reduced to a sample of famous doctrines. There are many chapters in the
10 Ila-hiyya-t and the Išārāt that still await more detailed analysis – not to speak of
11 studies on the reception of these chapters. On the other hand, the reception
12 history shows that two doctrines in particular were extremely successful in all
13 three cultures considered here and in both philosophical and theological
14 milieus: the distinction between essence and existence, and the concept of the
15 necessary existent by itself. The medieval readers of Avicenna thus testify to the
16 remarkable philosophical originality of these doctrines.
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
1
2
3
4 Al-Lawkarı̄’s Reception of Ibn Sı̄nā’s Ilāhiyyāt
5
6 Jules Janssens
7
8 Al-Lawkarı̄, who was born at Lawkar near Merw (Iran) at an unknown date and
9 who died most likely at the beginning of the twelfth century (in 1123 at the
10 latest), is presented by al-Bayhaqı̄ as a disciple of Bahmanyār, a first-generation
11 student of Ibn Sı̄nā1. If this is correct, he is a second-generation student of the
12 Šayh al-ra’ı̄s. However, if he really died in 1123, it would be surprising that he
13 had˘ been a direct disciple of Bahmanyār, who died in 1066 at the latest.
14 Whatever the case, he is known as the author of a fihrist, i. e. a list, of Ibn Sı̄nā’s
˘
15 Ta lı̄qāt, but, above all, as the author of a major philosophical encyclopedia,

16 entitled Bayān al-haqq bi-dimān al-sidq 2. This latter encompasses three major
17 ˙ ˙ ˙
parts: logic, physics and metaphysics. The latter divides into two major parts,
18 i. e., al-‘Ilm al-kullı̄ (Universal Science), and Rubūbiyyāt (Lordly Things), followed
19
by al-Risāla l-mulhaqa bi-‘ilmi al-rubūbiyyati, the appended exposition of the
20 ˙
lordly science, and al-Fusūl al-muntahabat min ‘ilm al-ahlāq, the selected
21 ˙ ˘ division into ‘universal
˘
chapters of the moral science. The basic science’ and
22
‘lordly science’ is, as such, not mentioned by Ibn Sı̄nā in the Ilāhiyyāt of his Šifā’,
23
but might ultimately stem from his al-Mašriqiyyūn 3. As to the appellation
24
‘lordly science’, it probably derives from the prologue to the Kitāb al-Mabda’
25
wa-l-ma‘ād 4. In any case, al-Lawkarı̄’s division differs substantially from that of
26
Bahmanyār, which distinguishes between metaphysics, mā ba‘d al-tabı̄‘a, and
27 ˙
28
science of the states of the essences of the existing beings, al-‘ilm bi-ahwāl a‘yān
˙
29
30
31 1 See al-Lawkarı̄, Bayān. Mantiq. al-Madhal. Muqaddima, pp. 57 – 71. For Bayhaqı̄’s af-
32 ˙ , pp. 145
firmation, see al-Bayhaqı̄, Tarı̄h ˘ – 6.
2 ˘
For a more detailed survey of al-Lawkarı̄’s life and works, see Marcotte, Life and Work of
33
34 al-Lawkarı̄, pp. 134 – 57.
3 See Ibn Sı̄nā, al-Mantiq al-mašriqiyyı̄n, pp. 8, 9 – 10 (but note that Ibn Sı̄nā uses the
35
expression ‘ilm ilāhı̄, ˙ not rubūbiyyāt). Bertolacci, Reception of Aristotle’s ‘Metaphysics’,
36 pp. 188 – 9, stresses that the bipartition in question breaks with the unitary view of the
37 Šifā’, but can be regarded as the expression of the more original and independent way of
38 exposition that Ibn Sı̄nā avowedly follows in the Mašriqiyyūn. Regarding the names of
39 metaphysics as a discipline, as well of its parts, see ibid., pp. 599 – 605.
4 See Ibn Sı̄nā, Kitāb al-Mabda’ wa-l-ma‘ād, pp. 1, 7 – 8, where it is stated that ‘the fruit of
40
the branch of metaphysics is Theology, which treats [the subjects of ] lordship, i. e., the
41 First Principle, and the relationship which beings bear to it according to their rank’
42 (translation, slightly modified, taken from Gutas, Avicenna, p. 31).
8 Jules Janssens
1 al-mawğudāt 5. However, both Ibn Sı̄nā’s Ilāhiyyāt of the Šifā’ and Bahmanyār’s
2 Kitāb al-Tahsı̄l are largely used as sources in these basic parts, as is already shown
3
˙˙ his edition of the Bayān. In the part entitled Universal Science, they
by Dı̄bāğı̄ in
4 even constitute the only sources. He reproduces verbatim entire chapters or, at
5 least, large parts of them of both works. With respect to the Ilāhiyyāt of the Šifā’,
6 the following correspondences come to the fore:
7
8 Bayān Ilāhiyyāt of Šifā’
9 c. 1 – 4 I, 1 – 4 (metaphysics as science)
10 c. 14 III, 1 (accidental categories in general)
11 c. 18 – 21 III, 7 – 10 (quality and relation)
12 c. 24 IV, 3 (complete/incomplete)
13 c. 29 – 37 V, 5 – 6, 8 – 9; VI, 1 – 4 (species, differentia and definition)6
14 c. 40 – 41 VII, 2 – 3 (refutation of Platonic Ideas).
15
16 As to Bahmanyār’s Kitāb al-Tahsı̄l, one may point out these derivations:
17
˙˙
18 Bayān Tahsı̄l
˙˙
19 c. 6 – 13 II, 1, 3 – 4, 7 – 8, 11 – 13 (truth, substance, matter and form)7
20 c. 15 – 17 II, 2, 3 – 5 (quantity)
21 c. 22 – 3 II, 3, 1 – 2 (anteriority/posteriority; potency/act)
22 c. 25 – 8 II, 4, 2 – 5 (universal/particular; genus/matter)
23 c. 38 II, 5, 3 (chance/fortune/final cause)
24 c. 39 II, 6, 1 (unity/multiplicity).
25
26 It has to be noted that many of these chapters drawing on Bahmanyār are clearly
27 influenced also by Ibn Sı̄nā’s Ilāhiyyāt of the Šifā’, as indicated by al-Dı̄bāğı̄.
28 Some sources have escaped the editor: regarding chapter 7 (Tahsı̄l, II, 1, 4) there
29
˙˙
exists, besides an inspiration from Ilāhiyyāt, II, 1, a strong influence from the
30 Maqūlāt of the Šifā’ (I, 4). On the other hand, one looks in vain for any
31
32 5 See Bahmanyār, Tahsı̄l, pp. 277 and 567. Regarding the significant deviation from
Avicenna that is involved˙˙ in this distinction, see Janssens, Faithful Disciple, pp. 188 – 9.
33
34 6 In his Bayān, al-Lawkarı̄ has surprisingly reversed the order of the chapters 6 and 8 of the
Ilāhiyyāt. Moreover, he has divided the text of c. 3 (on the compatibility between the
35
efficient causes and their effects) over two chapters, presenting the second part as a
36 detailing (tafsı̄l) of the basic affirmation of the first.
37 7 Despite the fact˙ that the title of c. 6 of the Bayān is derived from the Ilāhiyyāt of his Šifā’,
38 I, 8, the text of the chapter reproduces Tahsı̄l, II, 1, 2, pp. 291,12 – 293,5. C. 13
39 corresponds to c. II, 1, 13 of the Tahsı̄l, not 3˙˙as indicated in al-Lawkarı̄, Bayān. al-‘Ilm
al-ilāhı̄, p. 70, n. 1. Finally, it has to˙be
˙ observed that c. II, 1, 11 of the Tahsı̄l (corporeal
40
matter is not devoid of form) has been copied in c. 10 of the Bayān, except ˙˙ for its end
41 (see Bahmanyār, Tahsı̄l, p. 336,1–12 – on the natural form), a literal copy of which
42 constitutes c. 11. ˙˙
Al-Lawkarı̄’s Reception of Ibn Sı̄nā’s Ilāhiyyāt 9
1 reference to Ilāhiyyāt, II, 4, with respect to chapter 12 (Tahsı̄l, II, 1, 12), or to V,

2 6, 8 and 9 regarding chapter 28 (Tahsı̄l, II, 4, 5). Finally,˙chapter
˙ 38 (Tahsı̄l, II,
3
˙˙
5, 3) has not only been inspired by chapter 14 of book 1 of al-Samā‘ al-t˙abı̄‘ı̄
˙ of
4 the Šifā’, but also by chapter 13 of the same book . 8 ˙
5 As to chapter 5, it is also based on both works, but it is distinctive insofar as
6 it first (pp. 27,2 – 28,9) quotes the very beginnings of I, 5 (regarding the
7 primary concepts and principles) of the Ilāhiyyāt of the Šifā’ (Šifā’, Ilāhiyyāt,
8 pp. 29,3 – 30,4), and then continues (pp. 28,9 – 32) with the reproduction of
9 chapters from Bahmanyār’s Kitāb al-Tahsı̄l: the entire chapter II, 1, 2 (on šay’,
10
˙˙
thing, as distinguished from being and non-being) and the beginning of chapter
11 II, 1, 3 (on possibility, impossibility and necessity; Tahsı̄l, pp. 290,14 – 291,11)9.
12 In Bahmanyār’s wording one easily detects several ˙˙ direct links with Šifā’,
13 Ilāhiyyāt, I, 5.
14 Surveying the totality of these derivations, one sees that al-Lawkarı̄ covers
15 almost the entirety of books 1 – 7 of the Ilāhiyyāt of the Šifā’. Only four chapters
16 seem to have been completely omitted, i. e., I, 6 – 7 (on the Necessary Being);
17 III, 3 (on unity and multiplicity, and the accidentality of number) and V, 4
18 (what among the things contained in the genus render it a species or not a
19 species?). The omission of the former two is certainly not by chance, since a
20 treatment of the very idea of the Necessary Existent evidently has its natural
21 place in the section on lordly things, not in that of the general ontology of the
22 universal science. As to chapter III, 3, it does not so much offer new ideas as
23 rather study in depth the difficult issues of unity and accidentality of number. In
24 fact, it deals with the aporetic nature of the notions of unity and multiplicity
25 and moreover proves the inseparability of unity and substance (despite the fact
26 that unity does not constitute the quiddity of the substance) by way of diaeresis
27 and reductio ad absurdum10. Similarly, chapter V, 4 can be considered as dealing
28 with a particular examination of a difficult topic, i. e., the relation between a
29 genus and its differentiae, both notions being dealt with in other chapters. In
30 ignoring them, al-Lawkarı̄ wanted perhaps to avoid too technical questions. For
31
32 8 I have limited myself here to mentioning only the most important omissions or mistakes
33 by al-Dı̄bāğı̄. For more details regarding Bahmanyār’s derivations from Avicennian texts
34 regarding the chapters II, 1, 4 – 13 and II, 2, 3 – 5, see Janssens, Revision, pp. 99 – 117
(102 – 4 and 110). In c. 28 (Tahsı̄l, II, 4, 5), one detects an inspiration of Ilāhiyyāt, V, 8;
35
V, 6 and V, 9 on pp. 185 – 7 in˙˙the order as indicated here. As to c. 38 (Tahsı̄l, II, 5, 3),
36 the inspiration of elements of c. 13 of al-Samā‘ al-tabı̄‘ı̄, see Ibn Sı̄nā, Šifā’,˙˙al-Samā‘ al-
37 tabı̄‘ı̄, pp. 63 – 6 (detectable on pp. 239 – 41 of the˙ Bayān). It has to be noted moreover
˙that the opening paragraph is also identical with the final paragraph of Ilāhiyyāt, VI, 4,
38
39 see Ibn Sı̄nā, Šifā’, Ilāhiyyāt, p. 283, 4 – 9.
9 The rest of c. II, 1, 3, which deals with truth and nullity, will be reproduced in c. 6 of the
40
Bayān.
41 10 Lizzini rightly stresses this double method of analysis in the given context, see Ibn Sı̄nā,
42 Metafisica, pp. 207 – 9.
10 Jules Janssens
1 the very same reason, he might have dismissed the last part of VI, 3 (Šifā’,
2 Ilāhiyyāt, pp. 274,5 – 278,8), where a rather technical objection against the
3 theory that the recipient of an act cannot be equal to its agent – one is more
4 burnt when one puts one’s hand in molten metal than in fire, hence the molten
5 metal is hotter than fire, although it became hot by the fire – is discussed.
6 However, in this case the omission can be explained (perhaps primarily) by Ibn
7 Sı̄nā’s remark that this discussion more properly belongs to the art of physics
8 (Šifā’, Ilāhiyyāt, pp. 275,18 – 276,1)11.
9 But why has he preferred on occasion reproducing Bahmanyār’s rewording
10 to giving the very text of the related chapters of the Ilāhiyyāt of the Šifā’? One
11 could be tempted to answer: in order to avoid being shown up as a plagiarist.
12 But why has he then not modified Ibn Sı̄nā’s text more significantly? Moreover,
13 one may not forget that in his lifetime authors’ rights did not exist, and scholars
14 were constantly copying large extracts from their predecessors’ works. Certainly,
15 al-Lawkarı̄ has taken the practice to an excess. So, could it be that he, in acting
16 this way, wanted to partake in what I have qualified elsewhere as a revision of
17 Ibn Sı̄nā’s metaphysics? The inclusion of two chapters, one of which is linked
18 with the Maqūlāt, the other with al-Samā‘ al-tabı̄‘ı̄ of the Šifā’, as indicated
19 above, might at first sight suggest that this is the ˙ case. Indeed, it seems to blur
20 the limits between metaphysics, on the one hand, and logic and physics, on the
21 other, just as Bahmanyār had done.
22 However, as soon as one looks more carefully, it appears immediately that
23 this is anything but evident. Indeed, when Bahmanyār in his Tahsı̄l, II, 1, 4,
24 used rather Maqūlāt, I, 4 than Ilāhiyyāt, I, 2 of the Šifā’, he simply ˙followed
˙ Ibn
25 Sı̄nā’s own indications. The Šayh al-ra’ı̄s, in the latter text says: ‘This (i. e., to
26 claim that something can be both ˘ a substance and an accident with respect to
27 two things) is a grave error. We have discussed it fully in the first parts of Logic.
28 For, even though it was not the [proper] place [for discussing it], it was there
29 that they committed this error’12. The reference is clearly to the Maqūlāt, which
30 is the second book of the logical section of the Šifā’ 13. The exact identification
31 of the chapter referred to is not easy, but Bahmanyār obviously has identified it
32 as I, 4 – and in contemporary scholarship M. Marmura fully agrees with him14.
33
34
11 The omission of a great part of Šifā’, Ilāhiyyāt, VI, 5 (pp. 288,12 – 300,9) is undoubtedly
35
due to the reliance on Bahmanyār, but can as well be explained in a similar line.
36 12 Ibn Sı̄nā, Šifā’, Ilāhiyyāt, II, 1, p. 58,14 – 16; English translation of Marmura (Ibn Sı̄nā,
37 The Metaphysics of ‘The Healing’, p. 46,29 – 31).
38 13 The explicit reference to the first parts, seems to have misled Van Riet (Ibn Sı̄nā, Liber de
39 philosophia prima, V–X, p. 66, n. 37 – 8). She offers what is in my view a mistaken
reference to Logyca. Prima pars (i. e., Isagog), Venice 1508, fol. 4r. This reference has
40
been taken over by Lizzini in Ibn Sı̄nā, Metafisica, p. 1080, n. 11.
41 14 See Marmura (Ibn Sı̄nā, The Metaphysics of ‘The Healing’, II, 1, p. 389, n. 4). Contrary to
42 him, Horten (Ibn Sı̄nā, Metaphysik, p. 91, n. 6), Anawati (Ibn Sı̄nā, La mtaphysique du
1 As to the introduction of two chapters of al-Samā‘ al-tabı̄‘ı̄, it once more has an

˙
2 explicit basis in the Ilāhiyyāt, this time VI, 5, where one reads: ‘Regarding
3 coincidence and its being an end, we have finished [discussing] it in the
4 Physics.’15 This time, there is no doubt about the locus referentiae, i. e. I, 13, to
5 be complemented naturally with I, 14, insofar as this latter chapter refutes
6 mistaken conceptions regarding luck and coincidence that had been exposed in
7 the former16. So, in both cases the introduction of textual materials that are
8 foreign to the Ilāhiyyāt of the Šifā’ are supported by Ibn Sı̄nā’s own words.
9 Moreover, it has to be noted that al-Lawkarı̄ never quotes a chapter of
10 Bahmanyār’s Kitāb al-Tahsı̄l, where this latter deviates perceptibly from Ibn
11 ˙˙
Sı̄nā’s basic perspective. In this sense, it is most relevant that he, contrary to
12 Bahmanyār, does not mention in the present context any of the other categories
13 than the four (i. e., substance, quality, quantity and relation) treated by Ibn Sı̄nā
14
himself in the context of metaphysics. Herein I detect a fidelity to the original
15
metaphysical project of the venerable Master, i. e., Ibn Sı̄nā, and, at the same
16
time, a rejection of the revised version elaborated by one of the latter’s most
17
important immediate disciples, i. e., Bahmanyār.
18
However, thus far I have only dealt with the part on universal science.
19
Hence, one may wonder whether the above conclusion remains also valid for
20
that on the lordly things. It is immediately striking that in this part, one finds,
21
with the exception of two small fragments, no reference to Bahmanyār’s Kitāb
22
23
al-Tahsı̄l. The vast majority of the chapters (which are enumerated anew from 1)
˙˙
24
are once again a literal, or, at least, almost literal quotation of chapters of the
25
Ilāhiyyāt of the Šifā’:
26
27 Bayān Ilāhiyyāt of Šifā’
28 c. 1 – 2 VIII, 1, 3 (finitude of causes)
29 c. 3 – 7 VIII, 4; I, 7; VIII, 5 – 7 (Necessary Existent and attributes)
30 c. 8 IX, 1 (activity of God)
31
32 Šifā’, I–V, p. 334, n. 58,14) (based on the commentary of Mullā Sadrā al-Šı̄rāzı̄) and
33 Bertolacci (Ibn Sı̄nā, Libro della guarigione, p. 231, n. 11), refer to˙ I, 6 (Horten adds
34 moreover II, 1 – 3). Based solely on its title, this latter chapter appears to be the best
candidate, but if one looks at what is really at issue I, 4 may be preferable. To settle this
35
delicate issue clearly exceeds the limits of the present paper.
36 15 Ibn Sı̄nā, Šifā’, Ilāhiyyāt, VI, 5, p. 284,8 – 9; English translation of Marmura (Ibn Sı̄nā,
37 The Metaphysics of ‘The Healing’, p. 220,28 – 9).
38 16 Among contemporary scholars, Lizzini (Ibn Sı̄nā, Metafisica, p. 1187, n. 178), and
39 Bertolacci (Ibn Sı̄nā, Libro della guarigione, p. 549, n. 162) refer to both chapters,
whereas Marmura (Ibn Sı̄nā, The Metaphysics of ‘The Healing’, VI, 5, p. 410, n. 3) and
40
Van Riet (Ibn Sı̄nā, Liber de philosophia prima, p. 327, n. 76) refer to the sole c. 13. As to
41 Horten (Ibn Sı̄nā, Metaphysik, p. 416, n. 3), he refers to chapters 12 – 13. Anawati offers
42 no reference.
12 Jules Janssens
1 c. 9 IX, 2 (pp. 381,15 – 384,12 and 386,7 – 17) (soul as proximate mover of
2 the heavens)
3 c. 14 IX, 2 (pp. 392,7 – 393,10) (one single mover for the whole universe, but
4 each sphere has a specific mover)
5 c. 17 IX, 5 (pp. 313,7 – 414,13) (generation of elements)
6 (2. part)
7 c. 20 IX, 6 (pp. 415,8 – 418,12) (evil)
8 c. 21 IX, 7 (return, ma‘ād)
9 c. 28 X, 3 (acts of worship).
10
11
A few remarks have to be made:
12
1. Inside chapter 1 (pp. 272,6 – 273,2), a fragment is inserted into the text
13 of Šifā’, Ilāhiyyāt, VIII, 117. It starts with the words ‘according to another way of
14 consideration, he says’. This is surprising, insofar as before one looks in vain for
15 a previous occurrence of ‘he says’, let alone an explicit mention of a particular
16 author. But given that what precedes is a literal quotation of Ibn Sı̄nā’s Ilāhiyyāt
17 of the Šifā’, one suspects that the reference is to this latter. This turns out to be
18 correct, since the fragment reproduces – once again almost verbatim – part of a
19 text of the Šayh al-ra’ı̄s, taken not however from the Šifā’, but from al-Išārāt wa-
l-tanbı̄hāt. More ˘ precisely, it concerns the chapters 12 – 15 of namat 4 of part
20
˙
21 II18.
22 2. At the end of chapter 7, two passages of Bahmanyār’s Tahsı̄l, i. e., one
˙˙
23 (Bayān, pp. 316,10 – 317,9) covering III, 1, 1, pp. 576,12 – 577,16 (on
24 absolute perfection), another (Bayān, pp. 317,10 – 320,2) covering II, 6, 3,
25 pp. 559,10 – 561,14 (on pain and pleasure, sensible and intellectual), have been
26 added19. They seem to have been inspired by Šifā’, Ilāhiyyāt, VIII, 7,
27 pp. 368,13 – 369,10.
28 3. Chapter 9 opens with a proof that a celestial motion cannot be by force.
29 This proof is explicitly linked with Aristotle’s On the Heavens, designated by its
30 common Arabic title (fı̄) al-Samā’ wa-l-‘ālam (On Heaven and Earth). The
31 passage (Bayān, p. 333,5 – 14) is not a literal quotation, in spite of its opening
32 word qāla (‘he has said’), but offers a paraphrase of a fragment of the latter
33
34
17 The quotation of Ibn Sı̄nā, Šifā’, Ilāhiyyāt, VIII, 1 ends at Bayān, p. 272,2 with the
35
reproduction of line 2 of p. 329 and reopens at p. 273,2 (starting with fa-qad) with that
36 of line 3 of the very same page (not at p. 273, 4 with line 7, as indicated by al-Dı̄bāğı̄, see
37 al-Lawkarı̄, Bayān, p. 273, n. 4).
38 18 The former of these chapters is characterized as a šarh, the remaining ones are designated
as išāra, see Ibn Sı̄nā, al-Išārat, III, pp. 23 – 7. ˘
39
19 Al-Dı̄bāğı̄ seems to have forgotten to indicate the beginnings of this second fragment on
40
p. 310,10 of his edition of al-Lawkarı̄’s Bayān (at p. 320, n. 5 he remarks: ‘this is the end
41 of what has been transmitted from the Tahsı̄l ’, but this cannot be the end of the former
42 fragment, the only one to which he is referring ˙˙ to on p. 316, n. 14).
1 work, i. e., I, 2, 269a9 – 18. Al-Lawkarı̄ agrees with this, but adds that Ibn Sı̄nā
2 has surpassed the Philosopher in adding a proof that such motion is not natural
3 either. Somewhat later in the chapter (pp. 335,9 – 336,4), al-Lawkarı̄ introduces
4 a saying of (pseudo-)Ptolemy’s Kitāb al-Tamara (Book of the Result), which states
5 that there is no difference between what¯ chooses the best and the natural. The
6 saying is followed by an explanation due to a certain Abū al-‘Abbās Ahmad ibn
7 ‘Alı̄ al-Isfahānı̄20. In doing this, al-Lawkarı̄ wants to stress that Ibn Sı̄nā˙was right
8 – in spite˙ of refuting radically any natural circular motion – to accept that
9 somehow circular motion may be called natural, namely insofar as what moves a
10 body in a circular motion is not alien to that body (Šifā’, Ilāhiyyāt, IX, 2,
11 p. 383,1 – 3).
12 4. The second part of chapter 17 corresponds to pages 362,8 – 364,9 of the
13 Bayān. The first part (p. 362,1 – 8) shows affinities with Šifā’, Ilāhiyyāt, IX, 5,
14 p. 413,1 – 7, but stems ultimately from another source as immediately after-
15 wards will be indicated.
16 All the remaining chapters of the part on lordly things, except for two of
17 them, have their ultimate source in another Avicennian text, i. e., his Kitāb al-
18 Mabda’ wa-l-ma‘ād:
19
20 Bayān al-Mabda’ wa-l-ma‘ād
21 c. 10 – 13 II, 1 – 4 (emanation from the One; ibdā‘; first caused is one and is
22 intellect; multiplicity out of first intellect)
23 c. 16 II, 5 (generation of what is beneath the spheres)
24 c. 17 II, 6 (evocation of a certain theory on the generation of elements)
25 (1. part)
26 c. 18 – 19 II, 7 – 8 (providence, especially regarding beings of generation and
27
corruption)
28
c. 23 – 7 III, 16 – 20 (prophecy, and related issues).
29
30
Very close resemblances with the Ilāhiyyāt of the Šifā’ are present in chapters 10
31
(pp. 339,4 – 340,4; Šifā’, Ilāhiyyāt, IX, 3, pp. 402,6 – 403,13); 12 (pp. 345,3 –
32
346,5; Šifā’, Ilāhiyyāt, IX, 4, pp. 403,13 – 404,8), 13 (pp. 347,6 – 353,13; Šifā’,
33
Ilāhiyyāt, IX, 4, 405,10 – 409,20), 16 (pp. 359,3 – 361,6; Šifā’, Ilāhiyyāt, IX, 5,
34
pp. 410,4 – 12 and 411,16 – 412,15) and 17, as has already been noted21. With
35
respect to chapters 23 – 7, a doctrinal similarity shows up – at least, in a broad
36
sense – with Šifā’, Ilāhiyyāt, X, 1 – 2, and this in spite of the absence of any literal
37
correspondence. It looks as if al-Lawkarı̄ has opted for the version of the Kitāb
38
al-Mabda’ wa l-ma‘ād because of its being more detailed (e. g., the three
39
20 I was unable to identify this scholar.
40
21 Dı̄bāğı̄ (see al-Lawkarı̄, Bayān, p. 343, n. 1; p. 365, n. 2 and p. 391, n. 2) refers also – in
41 my view in an unjustified way – for the chapters 11, 18 and 23 to parallels in the
42 Ilāhiyyāt of the Šifā’.
14 Jules Janssens
1 properties of prophecy are presented in an elaborated way, not just briefly

2 referred to, as was the case in Šifā’, Ilāhiyyāt, X, 1, p. 435,14 – 15). In a similar
3 vein, one can explain the use of the Kitāb al-Mabda’ wa-l-ma‘ād in chapters
4 18 – 9 as a more detailed exposition of the definition of providence given at the
5 beginnings (pp. 414,17 – 415,7) of Šifā’, Ilāhiyyāt, IX, 6 (including a specifi-
6 cation with regard to corruptible beings).
7 With respect to chapter 15, I looked in vain for a precise source. The
8 chapter contains a question: ‘does the soul’s desire cause only the motion of the
9 spheres, nothing else?’, and offers a mistaken answer (‘motion originates in the
10 spheres when its soul desires to assimilate with the intellect in perfection, in the
11 same way as it originates in us, humans, when we desire our beloved, hence as a
12 kind of upsurge’) as well as the correct one (‘the soul’s desire to imitate the
13 intellect makes the motion the perfection itself, insofar as it brings into act what
14 was in potency in a continuous, eternal manner’). One easily detects elements of
15 resemblance with Šifā’, Ilāhiyyāt, IX, 2, pp. 389,4 – 392,6, but there is no word-
16 for-word correspondence. The formulation might therefore be al-Lawkarı̄’s own,
17 although I do not exclude that it copies a fragment of still another – Avicennian
18 or Avicennian inspired – work.
19 The probability that this is the case significantly increases as soon as one
20 realizes that a passage of Ibn Sı̄nā’s Commentary on the Theology constitutes the
21 source of chapter 22. It has to be noted that al-Lawkarı̄, contrary to his usual
22 habits, avows (Bayān, p. 388,3 – 4) that he is quoting Ibn Sı̄nā, and more
23 precisely the latter’s Kitāb al-Insāf. As is well known, this work had been lost
24 during Ibn Sı̄nā’s own lifetime, ˙but fragments of it survived, among which the
25 Commentary on the Theology22. Al-Lawkarı̄ quotes the beginning of Ibn Sı̄nā’s
26 commentary on Mı̄mar 2, where it is stressed that the soul, after its separation
27 from the body, understands all things, both universals and particulars, all at
28 once, since it grasps the particulars from their causes, i. e., as universals23. This
29 way of intellection in a universal way is strongly stressed and sharply contrasted
30 with the bodily mediated perception of (material) particular things in the ten
31 last lines of the chapter (pp. 389,14 – 390,7), which, as far as I can see, are not
32 present in the edited version of Badawı̄24. Whatever be the case, they are fully
33
34
22 With respect to the Kitāb al-Insāf, see for more details Gutas, Avicenna, pp. 130 – 40.
35 ˙
Regarding the Theology, I may refer to Adamson, Arabic Plotinus, passim.
36 23 See Ibn Sı̄nā, Šarh Kitāb Utūlūğiyā, pp. 35 – 74 (47,8 – 49,2). Despite the explicit
37 ˙ al-Insāf,
reference to the Kitāb ¯ al-Dı̄bāğı̄ has not identified any specific passage as a
˙
source. Even more surprisingly is the affirmation by Marcotte, Life and Work of al-
38
39 Lawkarı̄, p. 146 that ‘it is unclear what may have been Lawkarı̄’s source for the passage
from al-Insāf to which he refers and quotes (in the Metaphysics) on the posthumous life
40 ˙
of the soul’.
41 24 Although one cannot a priori exclude the possibility that al-Lawkarı̄ has himself added
42 these lines, it looks nevertheless more probable that he quotes a longer version than the
1 congruent with what precedes. They reflect moreover a genuinely Avicennian

2 spirit. In this respect, one may point to the Šayh al-ra’ı̄s’ Risāla l-adhawiyya fı̄ l-
˘ ˙˙
3 ma‘ād, where the universal character of the perception of the rational soul, as well
4 as the specification of the object of its perception in terms of the stable intentions
5 (al-ma‘ānı̄ l-thābita) are both clearly expressed25. However, there remains one
6 fundamental question: why has al-Lawkarı̄ inserted this passage of the
7 Commentary on the Theology? As far as I can see, he considered it in all
8 likelihood a natural complement to the preceding chapter – and hence to Šifā’,
9 Ilāhiyyāt, IX, 7. Indeed, the passage specifies the true happiness of the fully
10 accomplished soul in the hereafter. But, more importantly, he himself stresses in
11 the introduction to the section on divine science (pp. 3 – 4) that he wants to
12 limit his exposition to the basic principles, except for one topic, i. e., the nature
13 of the knowledge of the soul in the hereafter. Al-Lawkarı̄ emphasizes that he
14
does this because the usual manuals do not pay any attention to this topic (in
15
fact, they are always restricted to the issues of punishment and reward, pain and
16
joy). But this does not mean that he includes among these manuals Ibn Sı̄nā’s
17
works. On the contrary, al-Lawkarı̄ is aware that the latter, in his writings, has
18
dealt with it in a significant manner and, hence, makes clear that he owes this
19
idea to his ultimate master by the explicit mention of the latter’s name, followed
20
by a formula of great blessing: ‘May the spirit of God cover him and sanctify his
21
soul (or: him)’ (‘rūh Allāh ramasahu wa-qaddusa nafasahu’; p. 388,3).
22 ˙
Surveying both basic parts, one must admit that there is no real originality
23
24
in al-Lawkarı̄’s work. It is more than just inspired by the Ilāhiyyāt of the Šifā’; in
25
fact, it quotes verbatim many chapters, or parts of them, of the latter. On
26
occasion, it replaces some of them by other fragments, taken from elsewhere in
27
Ibn Sı̄nā’s œuvre, or in the Kitāb al-Tahsı̄l of the latter’s disciple, Bahmanyār –
˙˙
28 once again by way of literal copying. But even then, there is no appearance of a
29 single rupture with the exposition of the Metaphysics of Ibn Sı̄nā’s major
30 encyclopedia. It is therefore surprising that al-Lawkarı̄ (p. 3) qualifies this part
31 of his work as being both in the way of a talhı̄s, concise exposition, and a šarh,
˙ ˙
32 commentary, since there is no real additional˘ input of himself. As indicated at
33 the beginning, the division of the divine science into ‘universal science’ and
34 ‘lordly science’ might reveal a personal accent, although possibly still inspired by
35 Ibn Sı̄nā. It is interesting to see how al-Lawkarı̄, in his introduction (p. 4),
36 articulates this division. First of all, he specifies that the ‘universal science’ deals
37 with the principles of all the sciences. As to the other part, it exposes the
38 intentions of the Lordly Book, called Utūlūğiyā (Theology), – clearly a reference
¯
39 to the famous pseudepigrapic work called Theologia Aristotelis. A little later, the
40
one edited by Badawı̄. It might be present in the so-called longer version, which till now
41 has not yet been edited.
42 25 See Ibn Sı̄nā, Risāla l-adhawiyya fı̄ l-ma‘ād, p. 197,5 – 9.
˙˙
16 Jules Janssens
1 proper object of this part is more precisely articulated as the establishment of

2 the First Principle and Its attributes, and of the spiritual, angelic separated
3 beings. In between, al-Lawkarı̄ has noted that the name ‘meta-physics’ (mā ba‘d
4 at-tabı̄‘a) applies in common understanding to both parts. In the introduction
5 to˙ ˙the part on the lordly things (p. 265), he repeats that the lordly science is
6 usually considered as a part of metaphysics, but now he adds that this is not out
7 of necessity, but for the sake of brevity (li-l-ihtisār). In this sense, the
8 ˘ ˙
establishment of the First Principle is part of the science of meta-physics, the
9 subject of which is ‘being qua being’ and which therefore can be called ‘general
10 divine science’ (al-‘ilm al-ilāhı̄ al-‘āmm). From all this, one gets the impression
11 that al-Lawkarı̄ does not doubt the authenticity of the Theologia, and, moreover,
12 seems to consider the inclusion of a theology in the Metaphysics, i. e., in lambda,
13 as more a matter of convenience than of necessity. This might appear to sharply
14 contradict Ibn Sı̄nā’s basic dmarche of an integrated project, as present in the
15 Ilāhiyyāt of the Šifā’, but this is not necessarily the case, insofar as Ibn Sı̄nā
16 elsewhere, as indicated above, may have open the way to this kind of loosened
17 link between metaphysics proper and theology.
18 Al-Lawkarı̄, in the first introduction (p. 4), mentions also that he will attach
19 to the section of the Lordly Things selected aphorisms (fusūl muntaza‘a) on
20 ethics, especially on the virtues of the soul and making firm˙ the perfect state,
21 and that he will conclude the book with questions and remarks (masā’il wa-
22 nukat) regarding the secrets of meta-physics, i. e., the divine attributes, the agent
23 intellects and the celestial souls. It strikes one immediately that the addition of
24 an ethical section fully corresponds with the structure of the Ilāhiyyāt of the
25 Šifā’, where X, 4 – 5 treat of the morals of the city, the household and the
26 individual, although very briefly. But has al-Lawkarı̄ effectively quoted these
27 chapters? Al-Dı̄bāğı̄ has not included the above-mentioned additional parts in
28 his edition. They have been preserved in a single manuscript, i. e., Tehran
29 University, Central Library 250 (= 108)26. Unfortunately, this manuscript is
30 frequently damaged in the margins, especially near the end of the text. In spite
31 of this damage, I succeeded in identifying the source of the vast majority of the
32 passages. Before presenting the results of my research, I want to stress that the
33 actual order is not that indicated by al-Lawkarı̄ in his introduction. In fact, one
34 first finds the appendix regarding the secrets of metaphysics; it is divided into
35
36 26 Many thanks to Frank Griffel and Meryem Sebti, who each kindly provided me with a
37 separate CD-Rom of the manuscript, thus giving me two scanned versions. For the
38 analysis of the appendices, I have used both CD-Roms, since in Griffel’s copy folios are
39 lacking, whereas in that of Sebti the order of the different parts of the work is changed.
Moreover, there is no numbering on the folios, but based on the information given by
40
Nağğār in al-Fārābı̄, Fusūl muntaza‘a, p. 18, I fixed the folio numbering. However, even
41 ˙
if it would be found incorrect, I am confident that the indications I offer will permit the
42 reader to identify beyond any doubt the passage in the manuscript.
1 four chapters (abwāb): the attributes of the Creator, intellects, souls, sanctity
2 and prophecy (wilāya wa-nubuwwa) respectively. Only afterwards does the
3 ethical section follow. This, in its turn, contains five chapters, i. e., health and
4 sickness of soul; social relations (mu‘āšira); politics (siyāsa); theoretical and
5 practical intellect; and classes of the virtuous city27.
6 Let us now concentrate on the first appendix. Its first chapter (fols 212r-
7 v
214 ) starts with a rather long discursus on the divine attributes of knowledge,
8 will, providence, power, wisdom and liberality. It is a literal reproduction of the
9 discursus on these subjects in the Ta‘lı̄qāt, which offers what I have characterized
10 as the Arabic original of the corresponding chapters in the Dāneš-Nāmeh28. The
11 three following subdivisions (fusūl), while focusing respectively on divine
12 ˙
goodness, on the essential (not temporal) priority of God towards His action,
13 and on the divine Light, reproduce three fragments derived from Ibn Sı̄nā’s
14
Commentary on the Theology29. Finally, three subdivisions have their source in
15
Ibn Sı̄nā’s Mubāhatāt30 : the first deals with God’s being necessary (wāğibiyya),
16 ˙ ¯
the second insists that from the One only one can follow (yalzimu) and the third
17
denies any multiplication in the divine essence. The divine tawhı̄d, unity and
18 ˙
unicity, plays a central role in this chapter. In spite of a possible (logical)
19
distinction between several attributes in the divine essence, and in spite of God’s
20
acting resulting in a creation outside Him, He has a unique, indivisible fullness
21
of being, more precisely of being necessary. This idea is certainly present in the
22
23
Ilāhiyyāt of the Šifā’, but is now stressed in a more pronounced way, based on
24
texts of the Šayh al-ra’ı̄s himself!
25
The second ˘chapter (fols 214v-216r) contains additional information on the
26
higher intellects. Its first two subdivisions relate to the highest of them. This
27 latter is designated as the simple intellect (al-‘aql al-bası̄t) and both its essence
˙
28 and its action are discussed, once again based on the Mubāhatāt (pp. 301 – 2,
˙ ¯
29 § 844, respectively p. 302, §§ 845,1 – 2 and 846). As to the third subdivision,
30 which contains five proofs (i. e., necessity of essential unity of first emanated
31 being; impossibility of existence of celestial bodies; essential separate nature of
32 human soul; actualisation of human soul; eternal motion of universe), in order
33 to justify the existence of the agent intellects, it copies verbatim an entire chapter
34
35
27 The list of these headings is present in al-Lawkarı̄, Bayān. Mantiq. al-Madhal, pp. 90 –
36 91. ˙ ˘
37 28 See Ibn Sı̄nā, Ta‘lı̄qāt, pp. 13,4 – 22,9. Regarding the link with the Dāneš-Nāmeh, see
38 Janssens, Le Dānesh-Nāmeh, pp. 163 – 5.
39 29 See Ibn Sı̄nā, Šarh Kitāb Utūlūğiyā, pp. 46,4 – 15; 47,1 – 5 and 56,14 – 57,8.
˙
30 See Ibn Sı̄nā, Mubāh atāt,¯ respectively pp. 140 – 41, §§ 386 – 90; p. 226, § 674;
40
p. 271 – 2, § 787; p. 112,¯ § 261,1 – 3 (al-Lawkarı̄ introduces the two last fragments with
˙
41 aydan, ‘also’) and p. 366, § 1141. Unless otherwise indicated, page and paragraph
42 ˙
references are in what follows always to the edition by Bı̄dārfar.
18 Jules Janssens
1 of the Risāla marātib al-mawğūdāt, also known as Risāla fı̄ itbāt al-mufāraqāt31.
2
¯
As to the fourth subdivision, it specifies that the higher intelligences are only
3 able to grasp the First Principle thanks to a divine illumination (tağallı̄) by
4 quoting the Commentary on the Theology (pp. 49,6 – 50,10). The fifth
5 subdivision emphasizes that these intellects do not act in view of what is
6 beneath them and reproduces Ta‘lı̄qāt (p. 49,16 – 18 and 11 – 12). Their
7 possible nature is examined in the two following subdivisions, which have their
8 source in the same work (p. 52,9 – 10, respectively p. 54,7 – 14)32. Eventually,
9 the unavoidable presence of a multiplicity in their intellection is affirmed, once
10 more based on the Ta‘lı̄qāt (p. 62,20 – 24). In this chapter, the mediating – but
11 indispensable – role of the higher intellects occupies a central place. Also this
12 time, the idea itself is not foreign to the Ilāhiyyāt of the Šifā’, and, again, one has
13 to do with a clarification expressed in Ibn Sı̄nā’s own words.
14 As to the third chapter (fol. 216r-v), it concerns the soul. The Ta‘lı̄qāt forms
15 its unique source. Only the first subdivision pays attention to the celestial souls
16 and their role in the different circular motions of the spheres (Ta‘lı̄qāt,
17 p. 54,18 – 25). The seven remaining subdivisions all deal with the human soul:
18 the identity in it between active and final cause; the fact that the goal of its
19 motion is nothing outside itself; that its perception is not in view of what is
20 perceived; that the perfection of the vegetative soul is not the real end of man;
21 the goal-directedness of the soul as basis for the difference between voluntary
22 and natural motions; perception (idrāk), not sensation, as activity proper to the
23 soul; and the impossibility for an embodied soul to perceive itself directly as a
24 separate being (Ta‘lı̄qāt, pp. 63,20 – 8; 63,3 – 7; 63,8 – 9; 63,10 – 19;
25 53,20 – 25; 23,1 – 19 and 23,23 – 8). Except for the discussion of the celestial
26 souls, the metaphysical relevance of this chapter is less evident. Nevertheless,
27 one can make a link between the emphasis on the perfection of the human soul
28 in its quality as separate substance, on the one hand, and the evocation of the
29
ma‘ād in Šifā’, Ilāhiyyāt, X, 3, on the other.
30
Regarding the fourth chapter (216v–218r), it opens with the idea that a
31
human soul has to prepare itself adequately in order to achieve its perfection,
32
reproducing once again Ta‘lı̄qāt (p. 37,22 – 4). The second subdivision is rather
33
long. It spotlights the pure soul (al-nafs al-zakiyya), the significance of prayer
34
35
31 Not only the title, but also the authorship of the treatise is doubtful: although other
36 names are given as well, the best candidates for the latter are undoubtedly Bahmanyār
37 and al-Fārābı̄. This problem certainly deserves a profound analysis that cannot be offered
38 in the present paper. Let me simply note that al-Lawkarı̄ copies c. 3, which corresponds
39 to Bahmanyār (?), Risāla marātib al-mawğūdāt, pp. 63,8 – 64,15.
˘
32 In the former of the two fragments, al-Lawkarı̄ adds to the affirmation of the Ta lı̄qāt
40
that this possibility is not like that of the generable beings. This addition was maybe
41
˘
already present in the copy of the Ta lı̄qāt he had at his disposal. A systematic study of
42 this latter work, and its different redactions, remains a major desideratum.
1 and extraordinary events such as magic, talismans and miracles. It copies Ta‘lı̄qāt
2 (pp. 47,20 – 48,12), but before quoting the final sentence (lines 10 – 12)
3 introduces a large section of Namat 10 of part II of the Išārāt33. The
4 subdivisions 3 – 6 copy once more ˙ verbatim passages from the Ta‘lı̄qāt
5 (pp. 69,25 – 70,3; 77,3 – 5; 79,27 – 80,17; 82,22 – 5; 81,18 – 22; 81,26 – 8;
6 82,1 – 5; 82,17 – 18; and 193,25 – 194,5). Here, the soul is presented as a stable
7 entity, as having a separate essence and as being self-perceptive. It is affirmed,
8 moreover, that the body is a condition only with respect to the existence, not the
9 survival of the soul; that the soul cannot reach anything of the Malakūt, i. e., the
10 intelligible celestial world, unless it is entirely spiritual; and that the simple
11 intellectual representation (al-tasawwur al-bası̄t al-‘aqlı̄), which is a gift of the
12 Giver of Forms (Wāhib al-suwār), ˙ brings our ˙intellects from potency into act.
13
˙
The latest of these affirmations is supplemented with two additional remarks:
14 intelligible things are devoid of change, and hence of individuality in the way
15 sensible ones are; human beings cannot know intelligible things unless through
16 a conjunction with the agent intellect. Regarding both, I looked in vain for a
17 source. However, these additions might have been present in the copy of the
Ta lı̄qāt he had at his disposal34. In the following subdivisions, i. e., 7 – 11, al-
˘
18
19 Lawkarı̄ extensively deals with the (self-)knowledge of the soul, paying special
20 attention to its way of understanding after the separation of the body and to the
21 way the soul links with the agent intellect. This time, he combines different
22 passages taken from the Mubāhatāt (pp. 155 – 6, §§ 427 – 8; p. 316, § 888;
23 pp. 87 – 8, §§ 150 – 5435 ; p. 318,˙ ¯ § 893; and p. 318, § 892). The last
24 subdivision insists that the perfect human soul enjoys after its separation
25 from the body a purely intellectual life; it offers a literal quotation of the final
26 part of the Risāla marātib al-mawğūdāt36. All in all, the chapter concentrates on
27 the human soul and its ultimate perfection. Given its title, i. e., On Sanctity and
28 Prophecy, this is somewhat surprising. Certainly the passage, which mentions a
29 reaching of the Malakūt, deals with the mode of the prophet’s receiving
30 revelation37, but generally speaking almost no attention is paid to the specific
31 issue of prophecy (or sanctity). On the sole basis of the title, one might suspect
32 that the chapter was meant as somehow supplementing Šifā’, Ilāhiyyāt, X, 1 – 2.
33 Based on its actual wording, this is, however, far from being evident.
34 Let us now turn to the second section, on ethics. It was already noticed by
35 Fawzı̄ M. Nağğār that this section in our manuscript corresponds to the Fusūl
36
˙
37 33 Ibn Sı̄nā, al-Išārat, IV, pp. 153,2 – 159,6.
38 34 See above, n. 32.
39 35 The beginning of the quotation corresponds to the version of the fragment as published
by Badawı̄, p. 227, § 457.
40
36 See Bahmanyār (?), Risāla marātib al-mawğūdāt, pp. 65,19 – 66,7: the actual wording of
41 the text corresponds to the one given as variant p. 65, n. 14.
42 37 See Michot, Destine de l’homme, p. 127.
20 Jules Janssens
1 muntaza‘a, i. e. §§ 1 – 61 (except §§ 3, 15, 23 and 40) commonly attributed to

2 al-Fārābı̄38. Nağğār also observed that a large passage (from fol. 219r to the
3 beginning – in fact line 3 – of fol. 219v) has no direct counterpart in the actual
4 text of the Fusūl39. However, he mentions another Iranian manuscript, i. e.,
˙
5 Tehran University, Faculty of Theology, 695d, which contains the part of the
6 Fusūl preserved in al-Lawkarı̄’s Bayān. It omits the same paragraphs and also has
7 ˙
a large – even larger than in our manuscript – section that does not derive from
8 the Fusūl 40. Also more important is the fact that it does not stop, as does our
9 ˙
manuscript, in the midst of § 61 (p. 70,10), but at § 62 (still incomplete, since
10 the last words correspond to p. 73,2). In none of the two manuscripts does there
11 seem to be a closing formula, so that one suspects that al-Lawkarı̄’s original text
12 was longer and probably included still further aphorisms41. Anyway, for our
13 present study the way in which he introduces this ethical section is more worthy
14
of attention. Indeed, al-Lawkarı̄ affirms42 :
15
16 These are the sentences and aphorisms chosen from the science of morals. They
17
comprise: acquiring the virtues of the human soul, avoiding its vices; moving the
human being from his bad habits to fine habits; making firm the virtuous city;
18
making firm the household, i. e., the rulership over its members. They are all
19 brought together in this note [in five chapters].
20
21 It is obvious that this is not a neutral statement. On the contrary, an outspoken
22 interest in human virtue manifests itself. Certainly, attention is also paid to the
23
24 38 See al-Fārābı̄, Fusūl muntaza‘a, p. 18, where it is given as one of the manuscripts used to
establish the text.˙ It has to be noted that the omitted paragraphs are exactly the same as
25
the ones not present in Dunlop’s edition (see al-Fārābı̄, Fusūl al-madanı̄). It should be
26 noted moreover that the wording of the passages corresponding ˙ to §§ 6 and 26 is exactly
27 the same as in Dunlop’s edition, hence omitting the last sentence in both cases.
28 39 In fact, the last two lines of fol. 218v are already not present in the Fusūl, but they most
29 naturally continue the discursus on the nutriment of the body. This˙ suggests that al-
Lawkarı̄ had at his disposal another version than the one actually available to us. It has to
30
be noted moreover that this addition shows certain similarities with § 10 of the Fusūl.
31 Nağğār obviously forgot to mention that al-Lawkarı̄’s text, in turn, does not quote ˙the
32 paragraphs 8 – 11 of the Fusūl.
40 Unfortunately, I had no access ˙ to this manuscript and Nağğār’s information remains very
33
34 limited in this respect (see al-Fārābı̄, Fusūl muntaza‘a, pp. 18 – 19).
41 The actual state of affairs does not permit ˙ one to decide up to which point al-Lawkarı̄
35
quoted the Fusūl. In MS Tehran University, Central Library 250 (= 108), it ends in the
36 middle of § 61, ˙ whereas in MS Tehran University, Faculty of Theology, 695d, it ends in
37 the middle of § 62 (see Nağğār’s introduction to al-Fārābı̄, Fusūl muntaza‘a, pp. 18 – 19).
In view of the title of the section one might guess that al-Lawkarı̄˙ continued till § 67,
38
39 and not further, since afterwards other issues come to the fore.
42 For the Arabic text, see al-Fārābı̄, Fusūl muntaza‘a, pp. 18 – 19, as well as p. 23, n. 2; I
40
quote the English translation (slightly ˙ modified) of Butterworth (al-Fārābı̄, Political
41 Writings, p. 6). Note that the addition ‘in five chapters’ occurs only in MS Tehran
42 University, Faculty of Theology, 695d.
1 social and political order, but this is not presented as the exclusive or primary
2 object of the exposition that will follow. In this respect, it is most relevant that
3 the long addition mainly deals with the issue of the possibility of changing one’s
4 moral conduct, i. e., one’s habits. At the end, it is even stressed that the
5 governance of the soul (read: of oneself ) has precedence over all other kinds of
6 governance and that one has to make a jihād in order to perfect one’s habits.
7 Before, a brief remark has been offered on the phenomenon of magic and other
8 extraordinary arts. In my view, this fits better an Avicennian than a Farabian
9 perspective43. This impression is only reinforced when one looks at the division
10 into chapters.
11 The first (fols 218r–220r), containing fifteen subdivision (or aphorisms?),
12 discusses the health and sickness of the soul in parallelism with the health and
13 sickness of the body. The use of a medical metaphor is rather unusual –
14
although not completely lacking44 – in al-Fārābı̄, while it is quite natural in Ibn
15
Sı̄nā. Suffice to say that the latter calls his major philosophical encyclopedia
16
Book of Healing – his ‘healing’ being that of the soul. In the second chapter (fols
17
220r–221r), which has seven parts, the medical metaphor is maintained in the
18
explanation of the social relations. One detects moreover a strong emphasis on
19
the necessity of being virtuous for each individual person. As to the third
20
(fol. 221r), it briefly – in only three subdivisions – deals with the notion of
21
malik, king. Unsurprisingly, the kingly craft is compared to the medical craft of
22
23
the physician. It may be noted that the largest of the three subdivisions offers a
24
survey of different historical opinions about the goal intended in kingship. This
25
kind of historical doxography is once again more typical of Ibn Sı̄nā than of al-
26
Fārābı̄45. The fourth chapter (fols 221r–222v) deals with both theoretical and
27 practical intellection and is divided into no less than twenty subdivisions46. In
28 this section, there is no room for any medical metaphor, but it is striking that in
29 one passage (fol. 222r, lines 3 – 8; Fusūl, § 41) the idea is evoked that a sick
˙
30 person imagine what is sweet bitter, and vice-versa, an idea one also encounters
47
31 in Ibn Sı̄nā’s Kitāb al-Nafs of the Šifā’ . Regarding chapter five, which might be
32
33 43 Regarding the primacy of the governance of one own’s soul, see e. g. Ibn Sı̄nā, Fı̄ l-siyāsa
34 l-manziliyya, pp. 232 – 60 (240). With respect to magic and other occult phenomena, see
for example Ibn Sı̄nā, al-Išārāt, IV, pp. 158 – 9, and compare moreover above, p. 19. The
35
same basic idea is also present in the Kitāb al-nafs of the Šifā’ (especially IV, 4).
36 44 See al-Fārābı̄, Ihsā’ al-‘ulūm, pp. 67 – 76 (71), where the need of the king in his political
37 ˙˙
practice for experience is compared to that of the physician in his medical practice; see
38 Janssens, Experience, pp. 45 – 62 (49). Charles Burnett kindly informed me that this kind
39 of metaphor is the subject of a Ph.D. thesis by Badr el Fekkak.
45 See Janssens, Ibn Sı̄nā, pp. 83 – 93.
40
46 Due to heavy damage, the title of the chapter is not readable, but in view of the space
41 seems to have been ta‘aqqul, ‘intellection’.
42 47 See Ibn Sı̄nā, ‘De Anima’, p. 62,16 – 17.
22 Jules Janssens
1 incomplete as indicated above, it might be worthwhile to note that it seems to

2 involve in its second subdivision (Fusūl, § 57) a conception of jihād that
˙
3 substantially differs from that of al-Fārābı̄’s al-Madı̄na al-fādila48.
˙
4 Certainly, none of these elements definitely excludes a Farabian origin.
5 However, if the work is authentically Farabian, then Ibn Sı̄nā has largely used it
6 in his early work on ethics, i. e., al-Birr wa-l-itm (Piety and Sin)49. Indeed, a part
7 ¯
of the Risāla l-birr wa-l-itm, which seems to be a part of the otherwise lost work,
8 ¯
shows a very close similarity with Fusūl, §§ 1 – 19, respectively 33 – 56, while
˙
9 omitting as al-Lawkarı̄ §§ 3, 15 and 4050. If this were correct, the young Ibn
10 Sı̄nā would have been profoundly indebted to his great predecessor. However,
11 one may wonder then why he has chosen a rather small, and, moreover, not very
12 typical work of the Second Master? Or did there exist a nucleus of aphorisms
13 assembled by al-Fārābı̄ and has Ibn Sı̄nā enlarged it? But the attribution of the
14 Fusūl to al-Fārābı̄ may be mistaken, since it is, after all, only explicitly present
15 ˙
with the title in one single manuscript, namely Bodleian, Hunt 30751. A
16
thorough examination is needed in order to settle this complex issue. It clearly
17
exceeds the limits of the present paper. But whatever hypothesis one favours, it
18
is clear that for al-Lawkarı̄ the Fusūl was sufficiently Avicennian in contents to
19
be included in a kind of florilegium ˙ of Avicennian texts. Even more strikingly
20
Avicennian is his decision to make this ethical tract an appendix of Divine
21
Science. This corresponds to the fact that Ibn Sı̄nā himself, in his al-Birr wa-l-
22
itm, probably treated ethics in the context of what D. Gutas has labelled
23 ¯
‘metaphysics of the rational soul’52.
24
25
Looking over the whole, one has to admit that al-Lawkarı̄ gives an
26
overwhelming impression of a desire to reproduce to the utmost possible degree
27
an encompassing synthetic survey of Ibn Sı̄nā’s Ilāhiyyāt. Still, one may remain
28
somewhat hesitant, insofar as al-Lawkarı̄, contrary to Ibn Sı̄nā in the Šifā’, seems
29
to dissolve the unity of metaphysics in favour of a sharp bipartition. The
30 intimate link between universal Science and theology appears to have been
31 dramatically loosened. But this might have already happened in Ibn Sı̄nā, as
32 indicated at the beginning53. Moreover, one must not forget that the latter, in
33
34 48 See Dunlop’s introduction to al-Fārābı̄, Fusūl al-madanı̄, p. 13.
49 According to his autobiography, he wrote the ˙ work when he was still in Buhārā, just after
35
having finished at the age of twenty-one the Compilation for ‘Arūdı̄ (see ˘Ibn Sı̄nā, The
36 Life of Ibn Sina, pp. 38 – 41) This information permits one to date ˙ al-Birr wa l-itm
37 around the year 1000. ¯
38 50 See Ibn Sı̄nā, Risāla l-birr wa l-itm, pp. 360 – 68. It has to be observed that the wording
¯
in this Risāla is rather of a summarizing nature, but that a close link between both texts is
39
beyond any doubt. I hope to publish later a more detailed comparison.
40
51 See al-Fārābı̄, Political Writings, p. 5.
41 52 See Gutas, Avicenna, pp. 95 and 254 – 61.
42 53 See above, p. 7.
1 the general prologue to the work, has characterized the Šifā’ as ‘accommodating
2 his Peripatetic colleagues’54.
3 On the other hand, al-Lawkarı̄’s work clearly lacks originality. Most of the
4 time, not to say always, he copies verbatim large passages, and even entire
5 chapters of different works of Ibn Sı̄nā, supplemented with fragments taken
6 from the Kitāb al-Tahsı̄l of the latter’s immediate disciple, Bahmanyār and of
7 the Fusūl, which might ˙˙ be a work of the latter’s master, al-Fārābı̄ (if it is
8 ˙ not an Avicennian text, which, in my view, in the actual state of
definitely
9 affairs, cannot be proven). Anyhow, in all cases, the quotations are so literal that
10 al-Lawkarı̄’s text may be used as an independent testimony, besides available
11 manuscripts, for the establishment of the critical edition of the respective works.
12 In this respect, it is worthwhile to note that the oldest known manuscript of the
13 Bayān, i. e., Tehran University, Central Library 250 (= 108), is explicitly dated
14 601 H. (beginnings twelfth century) and is hence very old55. A first rapid survey
15 has permitted me to detect a large number of interesting variants in comparison
16 with the existing editions of the works in question. Of course, a thorough
17 investigation is needed for a final judgement. But just to show how really
18 interesting some of these variants are, I will quote three cases by way of example.
19 They all are related to the Ilāhiyyāt of the Šifā’.
20 The first occurs at p. 171. Having affirmed that the relationships in
21 irrational roots and in numerical relations are easily accessible to the soul, Ibn
22 Sı̄nā continues to state that ‘it does not follow that the soul in one state would
23 intellectually apprehend (an takūna … tu‘qila) all of these’, adding a little later
24 ‘within its proximate power to intellectually apprehend (an tu‘qila) this’ (Šifā’,
25 Ilāhiyyāt, p. 211,4 – 5). In al-Lawkarı̄ ‘intellectually apprehend’ is twice replaced
26 by ‘do’ (an yuf‘ala). Even if the context favours rather the reading of the edition
27 of the Ilāhiyyāt of the Šifā’, the variant of al-Lawkarı̄, which seems not to occur
28 anywhere else, is not totally devoid of sense. Moreover, it sheds light on why the
29
Latin translation has ‘agat’, or ‘agere’; a rendering that corresponds perfectly to
30
al-Lawkarı̄’s variant, and hence might not constitute a free rendering of the
31
Arabic verb ‘aqala (normally translated as intelligere)56.
32
The second concerns the possibility of an intellectual knowledge of an
33
individual entity, namely when it is unique in its species. According to the
34
35
54 Ibn Sı̄nā, Šifā’, al-Madhal, p. 10,14.
36 ˘
55 It is almost as old as several of the rather old known manuscripts of the Ilāhiyyāt of the
37 Šifā’, see Bertolacci, Reception of Aristotle’s ‘Metaphysics’, p. 486, although approximately a
38 century later than the oldest one, i. e., Tehran, Malik 1085, see ibid., p. 486, n. 21.
39 56 See Van Riet (Ibn Sı̄nā, Liber de philosophia prima, V-X, p. 243,58 – 9). Van Riet limits to
indicate in the second apparatus (Latin-Arabic) that agat/agere is here to be understood
40
in the sense of intelligat/intelligere, suggesting that the Latin translator has used a not very
41 literal, but nevertheless acceptable translation of the Arabic verb ‘aqala, there being no
42 variant present in the apparatus of the Cairo edition.
24 Jules Janssens
1 Cairo-edition of the Ilāhiyyāt – once again without any variant attested – Ibn
2 Sı̄nā affirms that ‘if the mind intellectually apprehends that species through its
3 individual instance (bi-šahsihi), it will have knowledge of it’ (Šifā’, Ilāhiyyāt,
4 p. 247,1 – 2). In al-Lawkarı̄’s ˘ ˙ version (Bayān, p. 195,3), this becomes: ‘if the
5 mind intellectually apprehends that species and represents it as individual (wa-
6 tašahissu), it will have knowledge of it’. In this case, this variant seems to be
7 ˘ ˙˙ proper to al-Lawkarı̄, although the Latin translation ‘et eius individuum’
entirely
8 (Philosophia prima, p. 277,12) also presupposes the presence in the Arabic (at
9 least, in the manuscript on which it was based) of the conjunction wa-.
10 Finally, al-Lawkarı̄ (Bayān, p. 261,8) confirms the reading fā’ilin instead of
11 kāmilin (Šifā’, Ilāhiyyāt, p. 318,12) in accordance with many other testimonies57.
12 I am aware that I do not do full justice to the relevance of al-Lawkarı̄’s text,
13 since it contains hundreds of variants. Its significance is undoubtedly still much
14 higher with respect to Ibn Sı̄nā’s Commentary on the Theology, since for that text
15 thus far only two manuscripts are known, i. e., Cairo, Hikma 6M, and Bursa,
16 Hseyin Çelebi, 119458. I therefore believe that the Bayān ˙ may help us in
17 editing better, and thus in understanding better a large variety of Avicennian, or
18 related, fragments of texts. On the doctrinal level, its value is much more
19 limited. Nevertheless, it draws our attention to the delicate way in which Ibn
20 Sı̄nā seems to have conceived the relationship between the different parts of the
21 science of metaphysics.
22
23
24 Bibliography
25
26
P. Adamson, The Arabic Plotinus. A Philosophical Study of the ‘Theology of Aristotle’,
London: Duckworth, 2002.
27
al-Bayhaqı̄, Tarı̄h hukamā’ al-Islām, ed. Mamdūh Hasan Muhammad, Cairo: Maktaba
28 ˘ ˙
al-taqafa al-dı̄niyya, 1996. ˙ ˙ ˙
29 ¯
al-Fārābı̄, Fusūl al-madanı̄. Aphorisms of the Statesman, transl. and ed. M. Dunlop,
Cambridge: ˙ Cambridge University Press, 1961.
30
31 –– , Fusūl muntaza‘a (Selected Aphorisms), ed. Fawzı̄ M. Nağğār, Beirut: Dār al-Mašriq,
˙
1971.
32
–– , Ihsā’ al-‘ulūm, in Kitāb al-Milla wa-nusūs uhrā, ed. M. Mahdi, 2nd edn, Beirut: Dār
33 ˙˙
al-Mašriq, 1991, pp. 67 – 76. ˙ ˘
34 –– , The Political Writings. ‘Selected Aphorisms’ and Other Texts, transl. and ed. Ch.
35 Butterworth, Ithaca, New York/London: Cornell University Press, 2001.
36 al-Lawkarı̄, Bayān al-haqq bi-dimāni al-sidq. al-‘ilm al-ilāhı̄, ed. I. Dı̄bāğı̄, Tehran:
37 ISTAC, 1995. ˙ ˙ ˙
38
39 57 For more details, see Bertolacci, Reception of Aristotle’s ‘Metaphysics’, p. 529. It is
worthwhile to note that many of Bertolacci’s proposals of correction to the Cairo edition
40
are confirmed by the Bayān, although not all of them. But it has to be stressed that
41 Bertolacci has formulated his proposals of correction with due care.
42 58 See Michot, Recueil avicennien, p. 125.
1 –– , Bayān al-haqq bi-dimāni al-sidq. Mantiq. al-Madhal, ed. I. Dı̄bāğı̄, Tehran:

˙
Mu’assasat Intišārāt ˙
Amı̄r Kabı̄r,˙ 1986. ˙ ˘
2
3
Bahmanyār ibn Marzubān, al-Tahsı̄l, ed. M. Mutahharrı̄, 2nd edn, Tehran: Mu’assasa-ye
intišārāt ve-chāp dānešhah-e˙Tehran,
˙ 1375 H. ˙
4 ˘
Bahmanyār ibn Marzubān (?), Risāla marātib al-mawğūdāt, ed. I. Javādı̄, Sophia
5 Perennis, 32, 1977, pp. 56 – 73.
6 A. Bertolacci, The Reception of Aristotle’s ‘Metaphysics’ in Avicenna’s ‘Kitāb al-Šifā’’. A
7 Milestone of Western Metaphysics, Leiden/Boston: Brill, 2006.
8 D. Gutas, Avicenna and the Aristotelian Tradition, Leiden: Brill, 1988.
9
Ibn Sı̄nā, al-Išārāt wa-l-tanbı̄hāt ma’a šarh Nası̄r al-Dı̄n al-Tūsı̄, ed. S. Dunya, 2nd edn,
4 vols, Cairo: Dār al-ma’ārif, 1968. ˙ ˙ ˙
10
–– , al-Mantiq al-mašriqiyyı̄n, Cairo: Matba‘a l-mu’ayyid, 1910.
11 –– , al-Mubāh ˙ atāt, in Aristū ‘inda l-‘Arab,˙ ed. A. Badawı̄, Cairo: Maktaba al-nahda al-
12 misriyya, ˙1947,¯ pp. 122 ˙ – 239. ˙
13
˙
–– , al-Mubāhatāt, ed. M. Bı̄dārfar, Qom: Intišārāt Bı̄dār, 1992.
˙ ¯
–– , al-Šifā’, al-Ilāhiyyāt, eds G.C. Anawati, S. Zayed, M. Musa and S. Dunya, Cairo: al-
14
15
Hay’a al-‘āmma li-šu‘ūn al-matābi‘ al-amı̄riyya, 1960.
–– , al-Šifā’, al-Madhal, eds G.C. ˙ Anawati, M. Hodeiri and F. Ahwani, Cairo:
16
al-Mutba’a al-amı̄riyya,˘ 1952. ˘
17 ˙ al-Samā‘ al-tabı̄‘ı̄, ed. S. Zayed, Cairo: al-Hay’a al-misriyya al-‘āmma
–– , al-Šifā’,
18 li-l-kitāb, 1983. ˙ ˙
˘
19 –– , al-Ta lı̄qāt, ed. A. Badawı̄, Cairo: al-Hay’a al-misriyya al-‘āmma li-l-kitāb, 1973.
–– , ‘De Anima’ being the Psychological Part of ‘Kitāb al-Šifā’’, ˙ ed. F. Rahman, London/
20
21
New York/Toronto: Oxford University Press, 1959, reprinted 1970.
–– , Die Metaphysik Avicennas, transl. and ed. M. Horten, Leipzig, 1907; repr. Frankfurt
22
am Main: Minerva, 1960.
23 –– , Fı̄ l-siyāsa l-manziliyya, in ‘A. Šams al-Dı̄n, al-Madhab al-tarbawı̄ ‘inda Ibn Sı̄nā
24 min ğilāli falsafatihi al-‘amaliyyati, Beirut: al-Šarka ¯ al-‘ālamiyya lil-kitāb, 1988,
25 pp. 232 – 60.
26 –– , Kitāb al-Mabda’ wa l-ma‘ād, ed. ‘A. Nūrānı̄, Tehran: The Institute of Islamic
27
Studies, McGill University, Tehran Branch, 1984.
–– , La mtaphysique du Šifā’, I-V, transl. and ed. G.C. Anawati, Paris: Vrin, 1978.
28
–– , Liber de philosophia prima sive Scientia divina, V-XS, ed. S. Van Riet, Louvain/
29 Leiden: Peeters/Brill, 1980.
30 –– , Libro della guarigione. Le cose divine, transl. and ed. A. Bertolacci, Torino: UTET,
31 2008.
32 –– , Metafisica. La ‘Scienza delle cose divine’ (al-Ilāhiyyāt) del ‘Libro della Guarigione’
33
(Kitāb al-Šifā’), transl. O. Lizzini, pref. P. Porro, 2nd rev. edn, Milano: Bompiani,
2006.
34 –– , Risāla l-adhawiyya fı̄ l-ma‘ād, ed. Fr. Lucchetta, Padova: Antenore, 1969.
35 –– , Risāla l-birr˙ ˙ wa l-itm, in ‘A. Šams al-Dı̄n, al-Madhab al-tarbawı̄ ‘inda Ibn Sı̄nā min
36 ¯
ğilāli falsafatihi al-‘amaliyyati, ¯
Beirut: al-Šarka al-‘ālamiyya lil-kitāb, 1988, pp. 353 –
37 68.
38 –– , Šarh Kitāb Utūlūğiyā, in Aristū ‘inda l-‘Arab, ed. A. Badawı̄, Cairo: Maktaba al-
nahd˙a al-misriyya, ¯ 1947, pp. 35 ˙ – 74.
39 ˙ ˙
–– , The Life of Ibn Sina. A Critical edition and Annotated Translation, transl. and ed.
40 W. Gohlman, Albany, New York: State University of New York Press, 1974.
41 –– , The Metaphysics of ‘The Healing’, transl. and ed. M. Marmura, Provo, Utah:
42 Brigham Young University Press, 2005.
26 Jules Janssens
1 J. Janssens, Bahmanyār, and his Revision of Ibn Sı̄nā’s Metaphysical Project, Medioevo,
2 32, 2007, pp. 99 – 117.
3
–– , Bahmanyār ibn Marzubān: A Faithful Disciple of Ibn Sı̄nā?, in Before and After
Avicenna, ed. D. Reisman with the assistance of Ahmed H. al-Rahim, Leiden: Brill,
4 2003, pp. 177 – 97 (now reprinted in J. Janssens, Ibn Sı̄nā and his Influence on the
5 Arabic and Latin World, Aldershot: Ashgate, 2006, pp. 177 – 79.).
6 –– , Experience (tajriba) in Classical Arabic Philosophy (al-Fārābı̄–Avicenna), Quaestio,
7 4, 2004, pp. 45 – 62.
8
–– , Ibn Sı̄nā: An Important Historian of the Sciences, in Uluslararası İbn Sn
Sempozyumu. Bildiriler. International Ibn Sina Symposium. Papers, II, eds M. Mazak
9 and N. zkaya, Istanbul: İstanbul Bykşehir Belediyesi Kltr A.Ş. Yayınları,
10 2008 – 9, pp. 83 – 93.
11 –– , Le Dānesh-Nāmeh d’Ibn Sı̄nā: un texte revoir?, Bulletin de philosophie mdivale,
12 28, 1986, pp. 163 – 77 (now reprinted in J. Janssens, Ibn Sı̄nā and his Influence on
13
the Arabic and Latin World, Collected Studies Series 843, Aldershot, Hampshire:
Ashgate, 2006, pp. 163 – 77).
14 R. Marcotte, Preliminary Notes on the Life and Work of Abū al-‘Abbās al-Lawkarı̄
15 (d. ca. 517/1123), Anaquel de Estudios rabes, 17, 2006, pp. 133 – 57.
16 Y. Michot, La destine de l’homme selon Avicenne, Louvain: Peeters, 1986.
17 –– , Un important recueil avicennien du VIIe/13e sicle: La Majmū‘a Hseyin Çelebi
18
1194 de Brousse, Bulletin de philosophie mdivale, 33, 1991, pp. 121 – 9.
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
1
2
3
4 Essence and Existence in the Eleventh- and
5 Twelfth-Century Islamic East (Mašriq): A Sketch*
6
7 Robert Wisnovsky
8
9
Introduction
10
11
In a well known section of his Hikmat al-išrāq (Philosophy of Illumination),
12 ˙
Šihābaddı̄n as-Suhrawardı̄ (i. e., aš-Šayh al-Maqtūl, d. 587 H./1191) attacked
13 ˘
˘
the doctrine that existence (wuğūd) is something superadded to (ma nan zā idun
˘
14
˘ ˘
alā) the substance or quiddity of things in the concrete, extramental world (fı̄ l-
15
a yān)–a doctrine he associates with those he refers to as the followers of the
16
Peripatetics (atbā al-Maššā ı̄n).1 Suhrawardı̄ maintains, by contrast, that
˘
˘
17 ˘
18
existence is among those aspects (i tibārāt) of a thing that belong purely to
19
the intellect.2 Partly because of Suhrawardı̄’s insistence on the subjective nature
20
21
22 * Earlier versions of this paper were presented at the American Philosophical Association,
23 Eastern Division Meeting, New York (December 2005); at the conference entitled ‘The
24 Arabic, Hebrew and Latin reception of Avicenna’s metaphysics’, Centro Italo-Tedesco
Villa Vigoni, Menaggio, Italy (July 2008); and as part of McGill’s Philosophy De-
25
partment Colloquium series (November 2009). I am grateful for the feedback I received
26 on those three occasions, as well as to the students and colleagues who participated in
27 two graduate seminars: the first, at Harvard in 1998, was devoted to issues of ontology in
28 Islamic philosophy; the second, at McGill in 2009, was devoted specifically to Suhra-
29 wardı̄. More particularly, Heidrun Eichner first helped me work through Rāzı̄’s position
and graciously shared photocopies of relevant manuscripts that she had acquired; and
30
Reza Pourjavady and Stephen Menn each made crucial suggestions that largely shaped
31 the final form of my argument. Needless to say, all mistakes are my own.
32 1 Suhrawardı̄, Hikmat al-išrāq, I.3, § 59, p. 46,8-ult.
2 Suhrawardı̄, H ˙ ikmat al-išrāq I.3, § 56, p. 45,1-ult., and § 60, p. 47, 1 – 13. The term
33
˙
˘
34 i tibār is difficult to translate into English. In his extended discussion of this issue, T.
Izutsu cites the famous passage in the Madhal (Isagoge) section of the Mantiq (Logic) of
35
Avicenna’s Kitāb aš-Šifā (The Healing) where ˘ Avicenna claims that quiddity ˙ (māhiyya)
˘
36
˘
has three i tibārāt: as a universal existing in the mind; as an essence existing in a concrete
37 individual; and taken in and of itself, i. e., as neutral with respect to either mental or
38 concrete existence. In light of this passage, and in light of Suhrawardı̄’s uses of the term
˘
39 in the Hikmat al-išrāq, Izutsu takes i tibār to mean ‘… a subjective manner of looking at
a thing,˙ something produced or posited through the analytic work of the reason. It is an
40
aspect of a thing which primarily appears in the subject and which, then, is projected
41 onto the thing itself as if it were an objective aspect of the thing.’ See T. Izutsu, The
42 Distinction between essentia and existentia, pp. 49 – 70 at 65. The Avicenna passage is
28 Robert Wisnovsky
1 of existence, his view was later used as an essentialist foil by Mullā Sadrā and
2 other members of that school of Islamic metaphysics which saw˙ itself as
3 upholding ‘the fundamentality of existence (asālat al-wuğūd)’.3
˙
4 Suhrawardı̄’s use of the phrase ‘the followers of the Peripatetics’ in this
5 context is usually taken by medieval as well as modern commentators to refer to
6 Fārābı̄ and especially Avicenna. In a broad sense this is perfectly plausible. After
7 all, Suhrawardı̄ attacks the idea that existence is an attribute that itself has real
8 existence in the concrete world, in the context of pointing to ‘the fact that some
9 followers of the Peripatetics construct their entire metaphysical project on [the
˘
10 basis of ] existence (anna ba da atbā i l-maššā ı̄na banaw kulla amrihim fı̄ l-
˘
ilāhiyyāti alā l-wuğūdi)’. This˙ must at least partially refer to Avicenna, who
˘
11
12 explicitly claimed, in the final chapter of Section 4 (‘On existence and its
13 causes’) of his al-Išārāt wa-t-tanbı̄hāt (Pointers and Reminders), to have created a
14 new proof of God’s existence (viz., burhān as-siddı̄qı̄n) that was superior, by
15 virtue of its basis in existence alone, to proofs ˙of˙ God’s existence from motion,
16 such as Aristotle’s proof of the need for an Unmoved Mover based on the
17 impossibility of an infinite regress of movers and moved things. Since existence
18 provides us with such a shaky foundation, Suhrawardı̄ argues, we need to turn
19 elsewhere, and create an alternative metaphysical basis in the form of ‘light’
20 (nūr).
21 Nevertheless, a question arises, because to my knowledge Avicenna never
22 explicitly committed himself to the thesis that existence is something ‘super-
23
added to’ (zā id alā) a thing’s quiddity.4 True, there is one passage in the Ta lı̄qāt
˘
˘
˘
24
(Marginal Notes) where Avicenna states that ‘The existence of each category is
25
extrinsic to its quiddity and superadded to it (fa-inna kulla maqūlatin fa-
26
˘
wuğūduhā hāriğun an māhiyyatihā wa-zā idun alayhā); whereas the quiddity of

˘
27
the Necessary ˘ of Existence is its thatness; hand its thatness is noti superadded to
28
[its] quiddity’.5 But given our current uncertainty about the circumstances in
29
˘
which the Ta lı̄qāt were composed, it would be rash to extrapolate a full-fledged

30
theory from this isolated instance. In Book IV, Chapter 3, of the Ilāhiyyāt of his
31
Kitāb aš-Šifā (The Healing), Avicenna does use the phrase al-wuğūd az-zā id. But
˘
˘
32
˘
Kitāb aš-Šifā /Mantiq (1): al-Madhal I.2, p. 15,1 – 7. J. Walbridge translates i tibārāt
˘
33
˙ ˘ the basis of an analogy with ‘legal fictions’: The
˘
34 aqliyya as ‘intellectual fictions’, on

Science of Mystic Lights, pp. 45 – 6, n. 43.
35
3 This label would probably have surprised Suhrawardı̄, given that in the passages just cited
36 he maintains the subjective nature not only of existence but also of the various ways of
37 conceptualizing essence, including quiddity (māhiyya), thingness (šay iyya) and inner-
˘
38 reality (haqı̄qa). On Suhrawardı̄’s role in Mullā Sadrā’s historiography of Islamic

˙ see S. Rizvi, An Islamic Subversion of the
philosophy, ˙ Existence-Essence Distinction?,
39
pp. 219 – 27.
40
4 Walbridge raises this question but does not offer an answer: Science of Mystic Lights,
41 pp. 47 – 8.
˘
42 5 Ibn Sı̄nā, at-Ta lı̄qāt, IV.32, p. 164,18–ult.

Essence and Existence in the Eleventh- and Twelfth-Century 29
1 what he is clearly referring to in this passage is the extra existence that God, who
2 is above perfection (fawqa tamām), does not need for Himself and which He
3 therefore passes on to other, lower beings.6
4 In a broader sense, Avicenna’s ontology could doubtless be interpreted as
˘
5 implying the thesis that existence is ‘superadded to’ (zā id alā) a thing’s quiddity.
˘
6 As I have discussed extensively in other publications, Avicenna’s general position
7 on essence (or quiddity) and existence is that essence and existence are
8 extensionally identical but intensionally distinct. In other words, every essence
9 must either be an individual existing in the concrete, extramental world (fı̄
˘
10 l-a yān), or a universal existing in the mind (fı̄ d-dihn). Even so, essence and
11 ¯ ¯
existent have different meanings: essence refers to what a thing is, whereas
12 existence refers to the fact that a thing is. More important for my discussion here
13 is the series of hints, given by Avicenna, that despite the fact that essence and
14 existence are co-implied (the term he uses in Kitāb aš-Šifā /Ilāhiyyāt I.5 is
˘
15 mutalāzimāni), essence nevertheless enjoys some kind of logical priority over
16 existence. The sense that essence is logically prior to existence is conveyed by –
17
among other clues – Avicenna’s frequent uses of the terms lāzim (‘is logically
18
˘
entailed [by]’), ārid (‘attaches accidentally [to]’), lāhiq (‘is a concomitant [of ]’)
19 ˙ [to]’) to describe how existence
and mudāf (‘is related ˙ connects to essence.7 An
20 ˙
interpreter could reasonably infer that describing existence as zā id (‘is
˘
21
superadded [to]’) would be perfectly in line with these other descriptions of
22
how existence connects to essence – despite the fact that, apart from its lonely
23
˘
appearance in the Ta lı̄qāt, zā id is never used by Avicenna in this way.

˘
24
Given the prominence of Suhrawardı̄’s critique of the thesis that existence is
25
something superadded to quiddity, and given the uncertainty about its
26
Avicennian genealogy, we should still try to find out more precisely who
27
Suhrawardı̄ was referring to when he targeted ‘the followers of the Peripatetics’
28
29
in this context. Avicenna may well have been in Suhrawardı̄’s sights, as has been
30
commonly assumed. But the fact remains that the most prominent exponent of
31
the thesis that existence is superadded to quiddity was Fahraddı̄n ar-Rāzı̄ (d.
606/1210), a contemporary of Suhrawardı̄’s and a fellow ˘ alumnus of
32
33
Mağdaddı̄n al-Ğı̄lı̄’s (n.d.) circle in Marāġa. My hypothesis is that the balance
34 of evidence compels us to think that Suhrawardı̄ was not so much targeting
35 Avicenna’s own ontology as he was targeting an emerging Avicennian ontology –
36 that is, the systematic reconstruction of Avicenna’s ontology that Rāzı̄ was just
37 beginning to undertake. Because Rāzı̄ appears to have left Azerbaijan in 580/
38
6 Ibn Sı̄nā, Kitāb aš-Šifā /Ilāhiyyāt (1), IV.3, p. 188,11 – 13.
˘
39
7 I discuss Avicenna’s developing ideas about the relationship between essence and
40
existence in my Notes on Avicenna’s concept of thingness (šay iyya), pp. 181 – 221;
˘
41 Avicenna’s Metaphysics in Context, pp. 143 – 80; and Avicenna and the Avicennian
42 tradition, pp. 105 – 13.
30 Robert Wisnovsky
1 1184 for Transoxania, where he wrote most of the works in which he claims
2 that existence is superadded to essence; and because Suhrawardı̄ finished his
3 Hikmat al-išrāq in 582/1186, it is unlikely that Rāzı̄ himself was in Suhrawardı̄’s
˙
4 sights. But regardless of the identity of the particular person whom Suhrawardı̄
saw himself as opposing (Ǧı̄lı̄ as well as Rāzı̄’s father and his Aš arite circle in
˘
5
6 Rayy present themselves as possibilities, but further research will be needed in
7 order to determine this), Suhrawardı̄’s arguments in favor of the conceptual
8 nature of existence clearly recapitulate an earlier attack, by the mathematician-
poet Umar Hayyām (d. 517/1123), against the tendency of certain Aš arite
˘
˘
9
10 mutakallimūn,˘ such as Imām al-Haramayn al-Ğuwaynı̄ (d. 478/1085), to appeal
11 ˙
to the theory of modes (ahwāl) – a theory associated with the Basran Mu tazilite
˘
12 mutakallim Abū Hāšim˙ al-Ğubbā ı̄ (d. 321/933) and his followers, the
˘
13 Bahšamites – as the best way to construe and promote Avicenna’s concept of
14 existence.
15
16
17 Avicenna’s Two Distinctions
18
19 To begin this story, I must turn first to Avicenna himself. Avicenna’s two key
20 metaphysical distinctions were between essence (or more properly, ‘quiddity’,
21 māhiyya) and existence (wuğūd); and between the necessary of existence in itself
22 (wāğib al-wuğūd bi-dātihi) and the necessary of existence through another
¯
23 (wāğib al-wuğūd bi-ġayrihi), which Avicenna appears to have taken as
24 convertible with the possible (or ‘contingent’) of existence in itself (mumkin
25 al-wuğūd bi-dātihi). Avicenna’s distinction between essence and existence, in its
¯
mature formulation in al-Ilāhiyyāt (Metaphysics) I.5 of his Šifā and in Išārāt 4,
˘
26
27 can be seen from one angle as a compromise position, stated in Arabic-
Aristotelian terminology, between the view of the early Mu tazilite mutakallimūn
˘
28
and that of al-Aš arı̄.8 Like Aš arı̄, Avicenna maintains – in Šifā , Ilāhiyyāt I.5 –
˘
˘
29
that thing (šay ) and existent (mawğūd), and by implication quiddity and
˘
30
31 existence, are extensionally identical: every existent will also be a thing, and vice
versa. This is in contrast to the position of the early Mu tazilites, who believed
˘
32
33 that thing was a broader category than existent, in that thing subsumes both the
˘
non-existent (ma dūm) and the existent. To the Mu tazilites, entities that had not
˘
34
35 yet come to be, and concepts in the mind, are examples of non-existent things:
therefore, non-existents as well as existents possess thingness (šay iyya). By
˘
36
37 contrast, Avicenna argues that things such as concepts in the mind do enjoy a
38 kind of existence – they simply possess mental existence (al-wuğūd ad-dihnı̄ or
¯ ¯
39 al-wuğūd fı̄ d-dihn) as opposed to the concrete existence found in individuals
¯ ¯
˘
40 (al-wuğūd al- aynı̄ or al-wuğūd fı̄ l-a yān; also referred to as ‘extra[mental]
41
42 8 Ibid.
existence’, al-wuğūd al-hāriğı̄). But unlike Aš arı̄, who maintained that thing and
˘
1
2
˘
existent were also intensionally identical, in the sense that thing means no more
3 or less than existent, and vice versa, Avicenna claimed that quiddity or thingness
4 (abstracted from thing) on the one hand, and existence (abstracted from
5 existent) on the other hand, were intensionally distinct. As I mentioned above,
6 for Avicenna, thingness and quiddity refer to what X is (i. e., as opposed to what
7 Y is); existence, by contrast, refers to the fact that X is (i. e., as opposed to X’s
8 not existing).
9 Unlike his distinction between essence and existence, which appears to have
10 evolved over the course of his career but only in a subtle way, Avicenna’s
11 distinction between intrinsically and extrinsically necessary existence underwent
12
˘
some dramatic developments from its first appearance in al-Hikma al- Arūdiyya
13 ˙ ˙
˘
(Philosophy for Arūdı̄) to its final appearance in the Išārāt. More directly relevant
14 ˙
to this chapter is the fact that in addition to articulating each of these two
15
distinctions in slightly different ways in books that he wrote at various points in
16
his life, Avicenna appears to have bound the two distinctions more closely
17
together as his ideas developed over time.9 Thus in the very early al-Hikma al-
18 ˙
˘
Arūdiyya, the distinction between essence and existence had hardly crystallized,
19 ˙
and the distinction between intrinsically and extrinsically necessary existence
20
was not thought through; and the two distinctions do not touch upon one
21
another at all.10 In the slightly later al-Mabda wa-al-ma ād (Origin and
˘
˘
22
Destination), the distinction between quiddity and existence is still only latent,
23
24
while the distinction between necessary and possible existence is quite fully
25
articulated; still, neither is linked directly to the other.11 In chapters I.5 and I.6
of the Ilāhiyyāt of his Šifā , from Avicenna’s middle period, the two distinctions
˘
26
27
receive their fullest expression.12 And while in those chapters neither distinction
28 is brought directly to bear on the other, they are later, in Book VIII of the
29 Ilāhiyyāt. 13 There Avicenna buttresses the distinction between God, the
30 Necessary of Existence in itself, and all other beings in the universe, which
31 are necessary of existence through another, by appealing to the notion that in
32 God quiddity and existence are identical, while in all other beings, quiddity and
33 existence are distinct. In his final major work, the Pointers and Reminders (al-
34 Išārāt wa-t-tanbı̄hāt), the two distinctions operate entirely in tandem, and the
35 distinction between quiddity and existence lays the basis for, and leads directly
36 to, the distinction between intrinsically and extrinsically necessary existence.14
37
38 9 Wisnovsky, Avicenna’s Metaphysics in Context, pp. 245 – 63.
Ibn Sı̄na, al-Hikma al- Arūdiyya (MS Uppsala Or. 364), fol. 2v8 – 10 and fols 3v16 – 4r12.
˘
39 10
˙ ˙ ād, pp. 2,4 – 3,15.
˘
11 Ibn Sı̄nā, al-Mabda wa-l-ma

˘
40
12 Ibn Sı̄nā, Kitāb aš-Šifā /Ilāhiyyāt (1), I.5, pp. 31,5 – 33,18 and I.6, pp. 37,7 – 38,5.
˘ ˘
41 13 Ibn Sı̄nā, Kitāb aš-Šifā /Ilāhiyyāt (2), VIII.4, pp. 343,10 – 347,16.
42 14 Ibn Sı̄nā, Kitāb al-Išārāt wa-t-tanbı̄hāt, pp. 138,2 – 139,13 and pp. 140,12 – 141,2.
32 Robert Wisnovsky
1 Why did Avicenna decide to bind the two distinctions together in his
2 middle and later works? I maintain that it is because he realized, during the
process of writing the Ilāhiyyāt of the Šifā , that he could use the intensional
˘
3
4 distinctiveness between quiddity and existence to show how beings other than
5 God were composites, that is, composed of quiddity and existence. God, by
6 contrast, could be held to be simple by virtue of the identity of quiddity and
7 existence in Him. Because every composite needs a composer to bring its
8 components together, and because of the impossibility of an infinite regress, the
9 chain of composites and composers must originate in a non-composite
10 composer, namely, God. Partly because Avicenna’s use of the quiddity-existence
11 distinction to support the intrinsically necessary-extrinsically necessary distinc-
12 tion was most obvious in the Išārāt, and partly because of that work’s
13 abbreviated and allusive style, which invites decompression and commentary,
14 the Išārāt received more attention from subsequent Muslim philosophers
15 (including the mutakallimūn) than any other of Avicenna’s writings – at least
16 until the sixteenth century CE, when the attention of commentators shifted to
17
the Šifā .
˘
18
Avicenna’s pressing of the essence-existence distinction into the service of his
19
intrinsically necessary-extrinsically necessary distinction was a crucial event
20
in the history of metaphysics. This is because it provided a method of
21
distinguishing God from both eternal and non-eternal beings that was based on
22
God’s simplicity and all other beings’ compositeness; and because, when
23
understood as the Necessary of Existence in itself, whose essence is not even
24
conceptually distinct from its existence, Avicenna’s God enjoyed a more
25
watertight simplicity than that of the Neoplatonists, whose God as One could be
26
held to be conceptually distinguishable from their God as Good. 15 As will
27
become apparent, Fahraddı̄n ar-Rāzı̄’s ontology can be seen as a continuation of
28 ˘
29
this trend in Avicenna’s own thought, that is, the trend towards using the
30
essence/existence distinction to explain the compositeness of all extrinsically
31
necessary beings.
32
33 Theological Ramifications
34
35 To be sure, it was not Avicenna’s proof of God’s existence from the
36 distinctiveness (and hence compositeness) of essence and existence in all beings
37 other than the Necessary of Existence in itself, which first made Avicenna’s
38 metaphysics attractive to Sunni mutakallimūn (specifically, those of the Aš arite
˘
39 and Māturı̄dite schools) from the two or three generations immediately

40
41 15 On this see my Final and Efficient Causality in Avicenna’s Cosmology and Theology,
42 pp. 97 – 123, and Avicenna’s Metaphysics in Context, pp. 181 – 95.
1 following his death. Instead, it was the use that they could make of Avicenna’s
2 distinction between intrinsically and extrinsically necessary existence in order to
3 solve some serious problems with their theology, and particularly with their
4 conception of how God’s attributes (sifāt) were related to the divine self (dāt).
˙ ¯
5 Sunni mutakallimūn from the generations immediately following Avicenna’s
6 death in 1037 CE, appropriated his distinction between intrinsically and
7 extrinsically necessary existence with a view to repairing a two-century-old
8 weakness of Sunni theology. More specifically, Sunni mutakallimūn such as the
˘
9 Aš arite Ǧuwaynı̄ – and to a lesser extent the Māturı̄dite al-Bazdawı̄ (d. 493/
10 1099) – sought to use new conceptual tools from Avicenna’s metaphysics in
11 order to modify and thereby strengthen the theory of divine attributes that they
12 had inherited from Ibn Kullāb (d. ca. 241/855) and which, among other
doctrines, served to distinguish their theology from that of their Mu tazilite
˘
13
14 competitors.16
15 The theological problem facing Ğuwaynı̄ and other Sunni mutakallimūn
16 was an unfortunate legacy of their inferring from Ibn Kullāb’s formula that God
˘
˘
17 is knowing ( ālimun) by virtue of a knowledge (bi- ilmin) that is neither identical
18 to nor other than Him (lā huwa wa-lā ġayruhu), and powerful (qādirun) by
19 virtue of a power (bi-qudratin) and speaking (mutakallim) by virtue of a speech
20 (kalām) that are neither identical to nor other than Him – so therefore God is
21 also eternal (qadı̄m) by virtue of an eternality (bi-qidamin) that is neither
22 identical to nor other than Him. The first part of the Sunnis’ theological
23 problem arose because God’s attribute of eternality could not be treated in the
24 same way as His other attributes. This is because the Sunnis held that not only is
25 God Himself eternal, but God’s attributes, such as His knowledge and His
26 power and His speech, are eternal too.17 The Sunni mutakallimūn viewed
27 eternality as both an attribute (or first-order predicate) and as a meta-attribute
28 (or second-order predicate). In other words, eternality is a divine quality that is
29 predicable not only of God’s self but of other divine qualities too.
30
˘
Early Mu tazilites, by contrast, tended to collapse all the divine attributes

31 into God’s self. According to the formula associated with Abū l-Hudayl (d.
32 ca. 226/841), for example, God is knowing through a knowledge ¯that is
identical to Him ( ālimun bi- ilmin huwa huwa).18 The Mu tazilites were
˘
33
˘
34 motivated by the fear of violating their cardinal tenet of tawhı̄d, or divine

35 ˙
36 16 A broader discussion of this development is contained in my One Aspect of the
37 Avicennian Turn in Sunni Theology, pp. 65 – 100.
38 17 This was not a two-way street, however: just because God’s speech was eternal, it did not
39 mean that conversely, God’s eternality was speaking. Nor was every attribute eternal: the
˘
‘attributes of the act’ (sifāt al-fi l) such as God’s ‘providing’ (rizq, i. e., food and water for
40
humans and animals),˙ which appeared to require a creaturely object on which to act,
41 were held (by the Aš arites, though not by their Māturı̄dite colleagues) to be originated.
˘
18 al-Aš arı̄, Maqālāt al-islāmiyyı̄n, p. 165,5.

˘
42
34 Robert Wisnovsky
1 uniqueness. For if God’s attributes were meaningfully distinct from God’s self –
2 in this case, distinct enough that each of the attributes could itself be seen as a
3 subject of which the predicate ‘eternal’ holds – then the attributes would share
4 the quality of eternality with God’s self, and a picture would emerge of a
5 pleroma of eternal, and hence quasi-independent, divine beings, rather than of a
6 single God, isolated in His sole possession of eternality. In a vain attempt to pre-
7 empt this Mu tazilite critique, early Sunnis tended to stick to Ibn Kullāb’s
˘
8 formula ‘neither identical to God nor other than Him’ (lā hiya llāhu wa-lā hiya
9 ġayruhu).19
10 The first problem facing Sunni mutakallimūn, therefore, was that God’s
11 eternality did double-service, both as a first-order predicate of God’s self and as
12 a second-order predicate of God’s other attributes. As soon as the other
13 attributes themselves became subjects of which ‘eternal’ was predicated, God no
14
longer enjoyed the complete isolation He had before, since each of His eternal
15
attributes had become sufficiently real and distinct that Sunni theology had
16
opened itself up to the criticism of polytheism. Just as serious was the danger of
17
positing an infinite regress of eternalities. For if each attribute is eternal, is it
18
eternal in itself, or through a further eternality? If the Sunni mutakallim answers
19
that each attribute is eternal in itself, then he will have made an explicit claim
20
that the attributes really are independent entities in their own right, and this in
21
˘
turn will provide his Mu tazilite competitor with clear evidence of polytheism.
22
But if an eternal attribute is eternal through yet another eternality, then each of
23
24
the second-order meta-eternalities will itself be eternal through a third-order
25
eternality, and an infinite regress will ensue.20
26
As if this first challenge was not enough of a headache, Ğuwaynı̄ and other
27
Sunni mutakallimūn had to address another weakness. The traditional kalām
28 proof of God’s existence relied on the contradictory nature of the opposition
29 between eternal (qadı̄m) and originated (muhdat). Since everything that is
˙ ¯
30 originated needs an originator (muhdit) to bring it into existence, and since an
˙ ¯
31 infinite series of originators and originated things is impossible, the chain of
32 originators and originated things will originate in an originator that is not itself
33 originated; and since everything in the world is either eternal or originated, this
34 non-originated originator will be eternal. That eternal originator is God. This
proof worked fine for the Mu tazilites, who, as mentioned above, collapsed
˘
35
36 God’s attributes into the divine self.21 But the Sunnis’ commitment to the
37
19 al-Aš arı̄, Maqālāt al-islāmiyyı̄n, pp. 169,2 – 170,3.
˘
38
20 This worry is expressed by, e. g., al-Qušayrı̄ (d. 466/1072), Šarh asmā Allāh al-husnā,
˘
39 ˙ ˙
p. 55,8 and pp. 392,5 – 7.
40
21 See, for example, the Basran Mu tazilite Abd al-Ğabbār’s (d. 415/1025) description of
˘
41 his predecessor Abū Alı̄ al-Ğubbā ı̄’s (d. 303/915) position, in al-Muġnı̄ fı̄ abwāb at-
˘
˘
˘
42 tawhı̄d wa-l- adl, V, p. 233,5.

˙
1 eternality of the divine attributes entailed that God, for them, was not the only
2 eternal thing in the universe: the divine attributes were eternal too. This meant
3 that the Sunnis found themselves with a category of beings – the divine
4 attributes – that were in one sense eternal but which were not themselves
5 originators. With the sudden appearance of this shadowy middle ground, the
6 traditional proof no longer works, since the opposition between eternal and
7 originated is now one between contraries, not between contradictories.22 If the
8 proof is to work, there must be an identity between ‘eternal’ and ‘non-originated
9 originator’; but now there seem to be eternal, non-originated things – the divine
10 attributes – that are not originators.
11 Faced with philosophical challenges such as these, it is not surprising that
12 Sunni mutakallimūn such as Bāqillānı̄ and Ğuwaynı̄ desperately sought tools to
13 help them confront or at least side-step the problematic implications of their
14
doctrinal commitments. One of those tools was Avicenna’s distinction between
15
intrinsically and extrinsically necessary existence. Since necessity can be
16
construed as modifying the predication itself, as opposed to serving as a
17
second-order attribute in the way eternality had, Sunni mutakallimūn could
18
come up with a new formula that did not predicate second-order attributes of
19
the first-order attributes, and hence was in less danger of inadvertently positing
20
the attributes as subjects or selves. Thus a Sunni mutakallim could assert the
21
modal propositions ‘God is necessarily an existent’ and ‘God is necessarily a
22
23
knower’, and thereby avoid stating that ‘God is necessary’ and ‘God’s knowledge
24
is necessary’ – a pair of attributions that would create the same infinite-regress
25
problem with necessity that ‘God is eternal’ and God’s knowledge is eternal’ had
26
created earlier with eternality.23 Nevertheless, the new formulas raised further
27 challenges, one of which was explaining how the divine attribute of existence
28 (wuğūd) connected to the divine self or essence (dāt) in such a way that it
¯
sounded like Sunni attribute-theory and not Mu tazilite attribute-theory.
˘
29
30
31
32 The Status of Modes
33
34 Partly because of these dilemmas, Ğuwaynı̄ turned for help not only to
35 Avicenna’s metaphysics. He, and (apparently) Bāqillānı̄ before him, also seized
upon the theory of ahwāl, or modes, proposed by Abū Hāšim al-Ğubbā ı̄, the
˘
36
˙
37
22 As the Aš arite mutakallim al-Bāqillānı̄ (d. 403/1013) appears to admit; see his Kitāb at-
˘
38
39 Tamhı̄d, pp. 29,17 – 30,2.
23 See, for example, al-Ğuwaynı̄, aš-Šāmil fı̄ usūl ad-dı̄n, p. 292,19 – 20, p. 308,9 – 10,
40
p. 358,11 – 13 and p. 365,7 – 11 (where Ğuwaynı̄ ˙ credits Bāqillānı̄ for this move);
41
˘
al- Aqı̄da an-nizāmiyya, p. 23,4 – 5; and Luma fı̄ qawā id ahl as-sunna wa-l-ğamā a,
˘
42 p. 137,9 – 10. ˙
36 Robert Wisnovsky
arch-rival of Aš arı̄, who founded Ğuwaynı̄’s and Bāqillānı̄’s theological school.
˘
1
2 The theory of modes has commonly been viewed as a theological compromise
3 designed to blunt the horns of the dilemma produced by construing the
4 assertion ‘God is powerful’ (Allāhu qādirun) as implying either that God is
5 power (Allāhu qudratun) or that God has power (Allāhu lahu qudratun). As
mentioned above, early Mu tazilites such as Abū al-Hudayl took the former tack
˘
6
7 and collapsed the attributes into God’s self, while early¯ Sunnis – following Ibn
8 Kullāb – took the latter tack and viewed the attributes as entities that were real
9 enough to be meaningfully distinct from God’s self and from each other.
10 Abū Hāšim appears to have been at least partly motivated by the worry that
the early Mu tazilites’ position entailed that the semantic content of all
˘
11
12 attributive assertions, such as ‘God is power’, is reducible to the assertion ‘God
13 is God’, and that all such assertions are therefore meaningless. So he appealed to
14 an element of classical Arabic grammar – the hāl – for a solution. A hāl, which
15 often takes the form of an adverbial accusative ˙ or phrase, refers to ˙a state or
16 condition that modifies the subject or object of a verb at the moment (hāl)
17 when the event described by the verb is taking place. Thinking of a divine ˙
18 attribute as a hāl, as a mode which describes God in His ‘act’ of being, allowed
19 Abū Hāšim to˙ hold (unlike the early Mu tazilites) that God’s being powerful
˘
20 (kawnuhu qādiran) is semantically distinct from, say, His being alive (kawnuhu
hayyan), just as the hāl ‘riding’ in ‘Smith came riding’ (ğā a Zaydun rākiban) is
˘
21
˙semantically distinct˙ from the hāl ‘walking’ in ‘Smith came walking’ (ğā a
˘
22
23 Zaydun māšiyan); but without ˙implying (as the Sunnis had) the separate
24 existence of real entities such as ‘power’ and ‘life’ or ‘riding’ and ‘walking’.
25 In recent decades, scholars’ focus has extended from the theological
26 motivations and implications of mode theory to its ontological motivations and
27 implications.24 Similarly, although Ğuwayni’s prime concern in appropriating
28 aspects of Bahšamite (i. e., Abū Hāšim’s) mode-theory was theological rather
29 than ontological, his appropriation had implications beyond an analysis of the
30 relationship between God’s self and His attributes. Ğuwayni’s most extensive
31 discussion of modes comes in his Kitāb al-Iršād, where he states baldly that ‘The
32 mode is an attribute which belongs to an existent and which is [itself ] qualified
33 by neither existence or non-existence’.25 To be precise, Ğuwaynı̄ held that
˘
34 ‘knowledge’, ‘power’ and ‘speech’ are real objects (ma ānı̄) that are possessed by a
35 being – God – and in which the modes (ahwāl) ‘being-a-knower’, ‘being-
36
˙
37 24 See A. Alami, L’ontologie modale; the classic treatments are by R. Frank, Abu Hashim’s
38 Theory of ‘States’, and more generally, Beings and Their Attributes; and most
39 comprehensively as well as most comprehensibly, H. Wolfson, The Philosophy of the
Kalam, pp. 167 – 234.
40
˘
25 Cf. al-Iršād fı̄ usūl al-i tiqād, transl. and ed. J. Luciani, Paris, 1938, p. 80,6 – 13 and
41 p. 47,11 – 13; cited˙ by M. Allard, Le problme des attributes divins dans la doctrines
˘
42 d’al-Ash arı̄ et de ses premiers grands disciples, p. 389.

˘
1 powerful’ and ‘being-a-speaker’ are grounded (mu allala). While the attributes
˘
2 understood as ma ānı̄ really exist in the being that possesses them, the attributes
3 understood as ahwāl are neither existent nor non-existent.
˙
4 The intermediate ontological status enjoyed by modes, which are neither
5 existent nor non-existent, lies at the bottom of a series of objections that
6 eventually led to Suhrawardı̄’s attack on the conception of existence held by ‘the
7 followers of the Peripatetics’. The ahwāl clearly violate the law of the excluded
8 middle, which holds that a thing S˙ is either P or not-P, with no third option
9 available. Forcing those who uphold a hāl-type theory of attribution to commit
10 themselves to the existence or non-existence ˙ of the attribute in question was a
11 theme running through eleventh-and twelfth-century discussions of ontology;
12 and skepticism about the real existence of attributes such as existence motivated
13 the mathematician and poet Umar Hayyām (d. 517/1123), who wrote a brief
˘
14 but acute Essay on Existence (Risāla ˘fı̄ l-wuğūd). 26 In the first chapter of that
15 work, Hayyām distinguishes between different types of attribute (sifa): that
16 which is˘ essential (dātı̄) to the characterized thing (mawsūf ), and that˙ which is
17 ¯ ˙
˘
accidental ( aradı̄) to it; that which is a necessary concomitant (lāzim) of the

18 ˙
characterized thing, and that which is not; and last but not least, that which is
19
separable from the characterized thing only in the faculty of estimation (mufāriq
20
bi-l-wahm) – that is, conceptually – and that which is separable in the faculty of
21
estimation as well as in (concrete) existence (mufāriq bi-l-wahm wa-bi-l-
22
wuğūd).27
23
This last division of attributes is most relevant to my argument, because it
24
lays the ground for Hayyām’s distinction between attributes that are conceptual
25 ˘ are existential (wuğūdı̄). Existential attributes are real in
˘
(i tibārı̄) and those that

26
the sense that they exist concretely in the characterized thing, whereas
27
conceptual attributes are not real, because they do not exist concretely in the
28
29
characterized thing, but instead attach to it only in our minds. Hayyām’s
˘ with
˘
30
example of an existential-accidental ( aradı̄ wuğūdı̄) attribute is ‘black’
˙
respect to ‘black body’. Here the attribute ‘blackness’ (sawād) is ‘something
31
˘
32
superadded to the black thing itself [and] existing in concrete reality’ (ma nan
˘
zā idun alā dāti l-aswadi mawğūdun fı̄ l-a yāni). To be precise, Hayyām holds
˘
33 ¯ ˘ attribution
34 that ‘black’ (i. e., corresponding to Ǧuwaynı̄’s hāl) is an existential
˙
˘
35 (wasf wuğūdı̄), whereas ‘blackness’ (i. e., corresponding to Ǧuwaynı̄’s ma nā) is

˙
an existential attribute (sifa wuğūdiyya). Hayyām takes the real existence of this
36 ˙
type of attribute to be evident ˘
at the intellectual, estimative and sensory levels,
37
38
26 Umar Hayyām, Risāla fı̄ l-wuğūd, ed. R. Ridāzāda-yi Malik, pp. 398,9 – 403,10. This
˘
39 ˙
edition ˘contains quite a few mistranscriptions, and should be compared with that
40
contained in S.G. Tı̄rtha, transl. and ed., The Nectar of Grace. Hayyām’s discussion in the
41 ˘
˘
Risāla is reiterated in his Darūriyyat at-tadādd fı̄ l- ālam wa-l-ğabr wa-l-baqā .

˘
42 ˙
27 This and what follows summarize ˙ Risāla fı̄ l-wuğūd, pp. 398,13 – 399,16.
Hayyām,
˘
38 Robert Wisnovsky
˘
1 and in this sense he agrees with Ğuwaynı̄ on the reality of ma ānı̄ (such as ilm –
˘
2 knowledge) in which the ahwāl (such as kawnuhu āliman – being-a-knower) are
3 grounded. ˙
4 However, when he treats the type of attribute he calls ‘essential-conceptual’
˘
5 (dātı̄ i tibārı̄), Hayyām departs from Ğuwaynı̄. Hayyām’s example is ‘being-a-
6
¯
color’, ˘
for ‘blackness’. ˘
Hayyām reasons that colorness (lawniyya) is not an
7
˘
attribute existing in concrete reality and superadded to blackness (sawādiyya),
8 because the fact of being superadded to something else in concrete reality entails
9 that the superadded thing be an accident; yet blackness is itself accidental, and
10 how can one accident serve as a subject for another accident that is predicated of
11 it?
12 According to Hayyām’s schema, then, an existential-accidental attribute such
13 as blackness is indeed ˘ something superadded, in concrete reality, to the black
˘
14 thing itself (fa-ammā l-qismu l-wuğūdiyyu l- aradiyyu fa-huwa ka-wasfi l-ğismi bi-
˙ wuğūdiyyatun ay huwa
˙
˘
15 l-aswadi idā kāna aswada fa-inna s-sawāda sifatun ma nan
¯ ˙
˘
˘
zā idun alā dāti l-aswadi mawğūdun fı̄ l-a yāni). But in the case of a conceptual-
˘
16
17
¯
essential attribute such as the fact that blackness is a color, the attribute colorness
18 is not an attribute superadded in concrete reality to the blackness itself (wa-
˘
19 amma l-qismu l-i tibāriyyu d-dātiyyu fa-huwa ka-wasfi s-sawādi bi-annahu

¯ ¯ bi-sifatin zā idatin alā
˙ dāti s-sawādiyyati fı̄
˘
lawnun…wa…l-lawniyyatu laysat
˘
20
l-a yāni). Hayyām proceeds directly to criticizing those who¯ neglect this crucial
˙
˘
21
22 distinction˘ between conceptual and existential attributions:
23
Those investigating this topic who do not take these conceptual attributions [awsāf
24 ˙
˘
i tibāriyya] into account have erred very far indeed, as is the case with some reckless
˘
25 moderns [al-muta assifı̄na l-muta ahhirı̄na], who posit colorness and accidentality
˘
26 and existence [my italics] and similar˘ ˘states as modes (ahwāl) that obtain in what can
27 be characterized by neither existence or non-existence.˙ The doubt that makes them
fall into this grave mistake pertains to the greatest of First Premises: that there is no
28
middle ground between negation and affirmation, the self-evident nature of which
29 needs no discussion by us, nor is there any way for idiots to contradict it or explain
30 it away.28
31
32
Given the Sunni appropriation of hāl-theory discussed above; given Hayyām’s
˙
insistence in another treatise that Avicenna ˘ of the
is to be regarded as ‘the best
33 29
Modern [philosophers] (afdal al-muta ahhirı̄n)’; given that Hayyām is
˘
34 ˙ (admittedly,˘ ˘ one who is known for˘embellish-

35
described by a bio-bibliographer
36
ing his notices with apocrycphal anecdotes) as having gotten himself into
37
trouble by defending Avicenna against Abū l-Barakāt al-Baġdādı̄’s (d. 560/1165)
attacks;30 and given that Avicenna himself attacked the Mu tazilite conception of
˘
38
39
28 Dānišnāma-yi hayyāmı̄, p. 399,18 – 24.
40 ˘ al-kawn wa-t-taklı̄f, p. lxxxvii,8 – 13.
29 Hayyām, Risālat
41 30 ˘
See al-Bayhaqı̄’s notice on Alā ad-Dawla Farāmarz, King of Yazd, in Tatimmat siwān al-
˘
42 hikma, pp. 110,7 – 111,11. ˙

˙
1 ahwāl for violating the law of the excluded middle;31 Hayyām’s use of ahwāl
˙ by implication ma ānı̄) in the passage above makes ˘it likely that his target ˙
˘
2 (and
3 in this long attack is not Avicenna himself, but Bahšamizing Sunni
4 mutakallimūn such as Bāqillānı̄ and Ğuwayni.32
The appropriation of ahwāl in late-tenth and eleventh-century Aš arite
˘
5
6 ˙
kalām is well attested by twelfth-and thirteenth-century authors, who level
7 much the same accusation against mode-theory that Hayyām does, namely, that
8 the ahwāl violate the law of the excluded middle by falling˘ in between existence
9 ˙
and non-existence. For example, Ibn Hazm (d. 456/1064), a Maġribı̄ near-
10 ˙
contemporary of Ğuwaynı̄ (and thus almost certainly referring here to al-
11 Bāqillānı̄), stated in his Fisal that ‘One piece of Aš arite nonsense is their claim
˘
˙
˘
12 that people can believe in modes and objects (al-ma ānı̄) which are neither
13 existent nor non-existent’.33 One generation after Umar Hayyām, aš-Šahrastānı̄,
˘
14 in his Milal, describes Abū Hāšim as having ‘postulated modes˘ as attributes that
15 are neither existent [nor non-existent] and neither knowable nor unknowable’.
16 In a passage that may have served as a bridge between Hayyām’s attack and
17 ˘ Hāšim’s modes are
Suhrawardı̄’s in the Hikmat al-išrāq, Šahrastānı̄ says that Abū
18 ˙
˘
19 … intellectual aspects (wuğūh) and [subjective] considerations (i tibārāt), that is,
they are the conceptual (mafhūma) [products of when] when we judge things as
20
being alike by commonality (al-ištirāk) or as being unlike by disjunction (al-iftirāq).
21 But these aspects are like relationships (al-nasab), correlationships (al-idāfāt),
proximity (al-qurb), remoteness (al-bu d) and the like, which, according ˙to the
˘
22
23 consensus of opinion, are not to counted among the real attributes.34
24
Šahrastānı̄’s criticism is echoed in his Nihāyat al-iqdām, where he describes Abū
25
˘
Hāšim’s ahwāl as intellectual aspects (al-wuğūh al- aqliyya) and mental and
estimative ˙considerations (i tibarāt dihniyya wa-taqdı̄riyya).35 In his Musāra a,
26
˘
˘
27 ¯ ˙
28
Šahrastānı̄, like Hayyām, uses colorness as an example of a mental consideration
˘
29
31 Ibn Sı̄nā, al-Mubāhatāt, § 181–§ 182, p. 94,1 – 7.
30 ˙ ¯
32 It is tempting to speculate that another factor at play here may have been the fact that
31 Ğuwaynı̄’s pupils were among Hayyām’s rivals for Selğuq infrastructure-funding, with
32 some money from the Selğuq ˘court going to the new Nizāmiyya madrasas (where
˙
Ğuwaynı̄ was employed), and other Selğuq money going to Mālik-Šāh’s new observatory
33
34 in Isfahan (where Hayyām was employed).
˘ wa-l-ahwā wa-n-nihal, no ed., (Cairo: 1317 – 1327 H.), vol. IV,
33 Kitāb al-Fisal fı̄ l-milal
˘
35 ˙
pp. 208,5 – 6; cited by Wolfson, ˙ Philosophy˙ of the Kalam, p. 170.
36 34 Šahrastānı̄, Kitāb al-milal wa-n-nihal (Book of Religious and Philosophical Sects, W.
37 Cureton, ed., London, 1846; reprinted ˙ Leipzig, 1923), p. 56,3 – 4 and pp. 56,16 – 57,1;
38 cited by Wolfson, Philosophy of the Kalam, pp. 171 – 172 and pp. 198 – 199.
˘
39 35 Šahrastānı̄, Nihāyat al-iqdām fı̄ ilm al-kalām (The Summa Philosophiae of al-Šahrastānı̄,
A. Guillaume, ed., London, 1934), p. 135,2 – 5; cited by Wolfson, Philosophy of the
40
Kalam, p. 199. In his correspondence with Īlāqı̄, Šahrastānı̄ also alludes to the fact that
41
˘
existence is something conceptual (ma nan i tibāriyyun): Guftgū-yi Šahrastānı̄ va-Īlāqı̄,

42 p. 101,21.
40 Robert Wisnovsky
(i tibār dihnı̄).36 In the Nihāya Šahrastānı̄ also reports that Bāqillānı̄ and
˘
1
¯
2 Ğuwaynı̄ appropriated Abū Hāšim’s mode-theory; and in the Milal he
complains that mode-theory is vulnerable to al-Aš arı̄’s own criticism that
˘
3
4 ahwāl violate the law of the excluded middle.37
˙
5 In the thirteenth century, aš-Šahrazūrı̄ (d. 7th/13th century), in his Divine
Tree (Šağara Ilāhiyya), explicitly names Abū Hāšim al-Ğubbā ı̄ of the Mu tazilites,
˘ ˘
6
7 and Ğuwaynı̄ and ‘Qādı̄ Abd al-Ğabbār’ from among the Aš arites (sic; ‘ Abd al-
˘
8 ˙
Ğabbār’ is almost certainly a mistaken scribal insertion after Qādı̄, which in
9 ˙
contexts such as this normally refers to Bāqillānı̄), as those mutakallimūn who
10 adopted the idea of ahwāl, which appear to fall between existence and non-
˙
11 existence.38 This is echoed by the Aš arite mutakallim Sayfaddı̄n al-Āmidı̄ (d. 631/
˘
12 1233), who claims that Qādı̄ Abū Bakr (i. e., Bāqillānı̄) and Imām Abū l-Ma ālı̄
˘
13 (i. e., Ğuwaynı̄) agreed with˙ Abū Hāšim regarding ahwāl.39
14 ˙
15
16 Rāzı̄ and Suhrawardı̄
17
18
As mentioned above, Rāzı̄ (and perhaps his teacher Ğı̄lı̄ – we cannot be sure
19
because only one of Ğı̄lı̄’s writings, a treatise on the fourth figure of the syllogism,
20
appears to have survived) followed the late trend in Avicenna’s metaphysics
21
towards pressing the essence/existence distinction into the service of the
22
intrinsically necessary/extrinsically necessary distinction. Motivated by his desire
23
to reinforce the compositeness of contingent beings by hardening the distinction
24
between essence and existence in them, by his Sunni theological commitment to
25
the distinctiveness between self (or quiddity) and attribute, and by his
26
27
commitment to the univocity of existence (on which more below), Rāzı̄ recast
28
˘
29 36 wa-anta ta rafu anna l-lawniyyata wa-l-bayādiyyata tibārāni aqliyyāni fı̄ d-dihni lā fı̄
l-hāriği wa-allā fı̄ l-wuğūdi lawniyyata l-bayād ¯ ¯
˙ i ġayra bayādiyyatihi. Šahrastānı̄, Kitāb
30 ˘ ˙ ˙
˘
al-Musāra a, p. 36,3 – 4 (Arabic) and p. 39 – 40 (English).

31 ˙
37 Šahrastānı̄, Nihāya, p. 131,5 – 9; Milal, p. 57,5 – 6 and p. 67,2 – 8; cited by Wolfson,
32 Philosophy of the Kalam, pp. 175 – 6 and 199 – 200.
38 Rasā il aš-šağara al-ilāhiyya, pp. 31,6 – 35,18.
˘
33
˘
34 39 al-Āmidı̄, Ġāyat al-marām fı̄ ilm al-kalām, p. 29,11 – 13; Abkār al-afkār fı̄ usūl ad-dı̄n,
vol. II, pp. 458,19 – 459,1; at Abkār al-afkār, vol. II, p. 604,6 – 9, Āmidı̄ claims ˙ that
35
Bāqillānı̄ flip-flopped on the issue of the ahwāl, sometimes affirming them and other
36 times denying them. In an apparent reference ˙ to Hayyām’s discussion, al-Āmidı̄ groups
37 wuğūd with lawniyya on the basis of the fact that both ˘ are superadded to quiddity: Ġāyat
38 al-marām p. 32,4 – 6, and Abkār al-afkār, vol. II, p. 607,5 – 8. Rāzı̄’s and Suhrawardı̄’s
39 Maġribı̄ contemporary Ibn Rušd (d. 595/1198) says in his Tahāfut at-Tahāfut that ‘those
who reject modes (al-ahwāl) reject the belief in existence in general and color in general,
40
whereas those who affirm ˙ modes say that existence in general and color in general are
41 neither existent nor non-existent’: Tahāfut at-Tahāfut, vol. III, ed. M. Bouyges, Beirut,
42 1930, p. 258,10 – 11; cited by Wolfson, Philosophy of the Kalam, pp. 170 – 171.
1 Avicenna’s ontology by holding that existence is something superadded to a thing’s

2 quiddity. This in turn raised the question of whether, on Rāzı̄’s account, existence
3 is a real attribute or merely a conceptual attribute, to use Hayyām’s terms.
4 Holding that the existence that is superadded to quiddities is itself˘ neither existent
5 or non-existent would invite accusations that Rāzı̄ had violated the law of the
6 excluded middle. But explicitly affirming that existence is a real attribute
7 superadded to quiddities, i. e., that the existence that is superadded to quiddities is
8 itself existent, would open Rāzı̄ up to the criticism that he had inadvertently
9 posited an infinite regress of existences – the criticism that Suhrawardı̄ was to
10 level, and which the Sunni theory of the attributes’ eternality had been vulnerable
11 to.
12 Without wading into the uncertainty and controversy over the chronology of
13 Rāzı̄’s writings, I have selected a significant sample of treatises that are generally
14 assumed to be from his middle period: Šarh al-Išārāt wa-t-tanbı̄hāt (Commentary
15 on [Avicenna’s] ‘Pointers and Reminders’),˙ al-Mabāhit al-mašriqiyyah (Eastern
16
˙ ¯ al-Mulahhas fı̄ l-hikma
Investigations), Lubāb al-išārāt (The Gist of the ‘Pointers’),
˙
wa-l-mantiq (Epitome of Philosophy and Logic), al-Arba ı̄n fı̄ usūl˘ ˘ad-dı̄n ˙(Forty
˘
17
˙ ˙
[Chapters] on the Fundamentals of Religion), and ar-Risālat al-kamāliyya fı̄ l-haqā iq
˘
18
19 al-ilāhiyya (Complete Epistle on Metaphysical Realities).40 These treatises are ˙largely
20 consistent in their ontology.41
21 In these works, Rāzı̄ normally begins his discussions of ontology by stating
22 that existence is univocal, that is, that existence applies with a commonality of
meaning (ištirāk ma nawı̄) to God and to the rest of the universe.42 Rāzı̄ maintains
˘
23
24
25
40 For a discussion of Rāzı̄’s writings and his intellectual development, see A. Shihadeh,
26 From al-Ghazālı̄ to al-Rāzı̄, pp. 141 – 79; Shihadeh offers a chronology in his The
27 Teleological Ethics of Fakhr al-Dı̄n al-Rāzı̄, pp. 4 – 11. As Shihadeh admits, his chronology
28 is tentative because it is based mainly on internal cross-references in Rāzı̄’s writings,
29 references that may have been inserted not during the actual composition of a given
work, but at a later date, when Rāzı̄ was revising that work. Developments in Rāzı̄’s ethics
30
are treated in Shihadeh’s book passim.
31 41 To my knowledge the only extended discussion of the development of Rāzı̄’s ontology is
M. A. Zarrukān, Fahraddı̄n ar-Rāzı̄ wa-ārā uhu al-kalāmiyyah wa-l-falsafiyyah, Cairo,
˘
32
˘
33 1963, pp. 170 – 74; ˘cited by the editors of two of Rāzı̄’s works: Šarh al-Išārāt wa-t-
34 tanbı̄hāt, ed. Nağafzādeh, pp. 47 – 51 and al-Mabāhit al-mašriqiyyah,˙ ed. M.M. al-
Baġdādı̄, p. 114, n. 1. The passages I am summarizing ˙ ¯are Šarh al-Išārāt wa-t-tanbı̄hāt,
35 ˙ and pp. 199,4 – 204,7;
˘
no ed. [ad Namat 4: Fı̄ al-wuğūd wa- ilalihi], pp. 190,7 – 192,13
36 ˙
Lubāb al-išārāt, p. 79,7-ult.; al-Mabāhit al-mašriqiyyah I.1.2 – 5, pp. 106,ult.–130,10;
37 al-Mulahhas fı̄ l-hikma wa-l-mantiq, Berlin˙ ¯ Staatsbibliothek Ms. or. oct. 629, I.2 – 3 (see
˘ ˘ ˙ ˙ ˙
H. Eichner’s chapter, ‘Essence and existence. 13th-century perspectives in Arabic-Islamic
38
˘
39 philosophy’, in this volume, for a translation of these two chapters); al-Arba ı̄n fı̄ usūl
ad-dı̄n, pp. 53,22 – 58,20; and ar-Risālat al-kamāliyya fı̄ l-haqā iq al-ilāhiyya, pp. 33,5 ˙–
˘
40
34,19. ˙
41 42 Although see Asās at-taqdı̄s (The Basis for Glorifying [God]), p. 89,8 – 17, for an attack on
42 the Karrāmiyya’s claim of strong univocity between the visible (aš-šāhid) and the invisible
42 Robert Wisnovsky
the Kullābite line by arguing, against Avicenna and the Mu tazilites, that God’s
˘
1
2 existence, like the other divine attributes, cannot simply be subsumed in the
divine self. But Rāzı̄ also argues, in open opposition this time to Aš arı̄, that the
˘
3
4 existence of each being is distinct from its quiddity. Rāzı̄ outlines three possible
5 positions on the relation between existence and quiddity, positions that he takes to
6 be exhaustive. The first possible position is that (a) existence is identical to
7 quiddity (al-wuğūd nafs al-māhiyya or al-wuğūd dāt al-māhiyya); the second is
¯
˘
8 that (b) existence is extrinsic to quiddity (al-wuğūd hāriğ an al-māhiyya); and the
9 ˘
third is that (c) existence is intrinsic to quiddity (al-wuğūd dāhil fı̄ l-māhiyya).
Rāzi assigns position (a) to Aš arı̄ (as well as – in Rāzı̄’s Arba˘ ı̄n – to the later
˘
10
˘
11 Mu tazilite Abū l-Husayn al-Basrı̄); position (b) to Avicenna and the falāsifa in
˘
12 ˙ ˙
general;43 and position (c) to no one, it being just a logical possibility.44 Rāzı̄ then
13 argues as follows in favor of position (b):
14
15
i. quiddity (māhiyya) is the principle of difference;
16
ii. existence (wuğūd) is the principle of identity, on the basis of the Avicennian
17
doctrine – which Rāzı̄ accepts and promotes – that all beings share
18
existence, or have existence in common (mušārakat al-wuğud);
19
iii. given i and ii, existence will not be identical to quiddity, and hence position
20
(a) is incorrect;
21 iv. given i and ii, existence will not be part of quiddity, and hence position (c)
22 is incorrect;
23 v. given that positions (a), (b) and (c) are the only possible positions, and
24 given that positions (a) and (c) are incorrect, it follows that position (b) is
25 correct, and that existence is therefore extrinsic to quiddity.
26
˘
(al-ġā ib). For a brief but helpful sketch of how the distinction between ištirāk ma nawı̄
˘
27 and ištirāk lafzı̄ maps onto the distinction between univocal and equivocal, see T. Mayer,
28 Fakhr ad-Dı̄n˙ ar-Rāzı̄’s Critique, n. 48.
˘
29 43 It is possible that in articulating position (b) as al-wuğūd hāriğ an al-māhiyya, Rāzı̄ may
have thought he was reverting to a classical Avicennian ˘formula, since in addition to
30
˘
appearing in the Ta lı̄qāt passage mentioned earlier, the same phrase is used by Avicenna’s
31
˘
favorite student, Bahmanyār, in his Tahsı̄l (viz., fa-inna kawna l-wuğūdi ma nan hāriğan
ani l-māhiyyati arafnāhu bi-bayānin ˙wa-burhānin),
˙ at least according to one˘ of the
˘
32
33 manuscript copies: Kitāb at-Tahsı̄l, p. 285, n. 3, ult.
34 44 Rāzı̄ might have been thinking˙˙here of Discussion V.16 of Ġazālı̄’s Tahāfut al-falāsifa,
pp. 151,7 – 152,10, where Ġazālı̄ rehearses the argument that existence cannot be a
35
constituent of (muqawwim bi-) quiddity. The genus and differentia are in a way parts of
36 the quiddity, given that the quiddity is the species of a thing. Rāzı̄ could be taking as
37 absurd the idea that the quiddity or species ‘caused [thing]’ contains the parts wuğūd
38 (‘existence’, i. e., as genus) and mumkin (‘possible’, i. e., as differentia). This is implied in
˘
39 his Šarh Uyūn al-hikma, pp. 80,1 – 82,12; immediately before that passage
˙
(p. 77,21–22), ˙
Rāzı̄ quotes Avicenna as denying that existence is ‘intrinsic to’ (dāhil fı̄)
40 ˘
essence. In his Epitome of the Metaphysics (I, pp. 34 – 43 and III, pp. 34 – 48), Averroes
41 cites Aristotle, Metaph. 3, as denying that being and unity are parts of the being of a
42 thing.
1 Rāzi further asserts that the correct way to articulate position (b) – the position
2 that existence is extrinsic to quiddity – is that existence is superadded to
˘
quiddity (zā id alā l-māhiyya). Rāzı̄ clearly sees himself as extrapolating from,
˘
3
4 and in some sense systematizing, Avicenna’s ontology. In his Commentary on the
5 ‘Pointers’ I.11, Rāzı̄ interprets the term mudāf (‘related [to]’), which Avicenna
6 had used in the passage being commented˙ upon to describe how existence is
7 connected to quiddity, in a stronger and less neutral sense, as ‘superadded’:
8
His [viz., Avicenna’s] statement ‘Like humanity; for in itself it is a quiddity …’, up
9
to the end of the chapter: Know that in defining the essential, he needs to
10 distinguish between the constituents of quiddity and the constituents of existence,
11 so that neither will be confused with the other. This is attained only by explaining
12 that existence is distinct from quiddity and that it is superadded to it [my italics]
˘
(anna l-wuğūda muġāyirun li-l-māhiyyati wa-annahu zā idun alayhā). He argues for
˘
13
this by [citing the fact] that quiddity may be known at a time when its existence is in
14
doubt; given that the known is other than what is in doubt, existence is superadded
15 to quiddity.45
16
17 Given Rāzı̄’s hard line on the univocity of existence, and given his apparent
desire to oppose Mu tazilite theology by placing himself in the Kullābite
˘
18
19 tradition of holding that the divine attributes are somehow meaningfully
20 distinct from the divine self, we should not be surprised by his argument that in
21 the case of God, as with other beings, existence is also distinct from and
22 superadded to quiddity. But in contrast to a contingent being, which needs a
23 cause to bring its quiddity and existence together, God’s quiddity is sufficient to
24 cause its own existence.46
25 It is worth noting that Rāzı̄’s position is in direct opposition to the doctrine
of his school founder, Aš arı̄. Aš arı̄ held that in created beings, quiddity and
˘
26
27 existence were intensionally as well as extensionally identical; and by some
accounts Aš arı̄ also treated the attribute of existence as an exception to Kullābite
˘
28
29 attribute-theory, by maintaining that in the case of God, too, quiddity and
30 existence were intensionally as well as extensionally identical. Rāzı̄’s departure
from Aš arı̄ is perhaps less surprising than it might first appear, since in the 12th
˘
31
32 century, when Rāzı̄ was active, the strong identification of essence and existence
was associated with a newly formed branch of the Mu tazilites as well as with
˘
33
Aš arı̄. These Mu tazilites were the followers of Abū l-Husayn al-Basrı̄ (d. 436/
˘
34
˙ ˙ Ibn al-
35 1044), most prominent of whom was the Hwārazmian mutakallim
36 ˘
37 45 Rāzı̄, Šarh al-Išārāt, ed. Nağafzādeh, I.11, p. 53,7 – 11.
46 Here Rāzı̄˙ may be appropriating the distinction Avicenna made in Kitāb aš-Šifā /Ilāhiyyāt,
˘
38
39 I.6, pp. 38,17–39,4, between a quiddity that is sufficient for a thing to exist and a
quiddity that is not sufficient for a thing to exist (the former will be uncaused, the latter
40
˘
caused). In any case Rāzı̄ is not entirely consistent; he says in his Šarh Uyūn al-hikma
41 (p. 76,14–15) that ‘we have discussed in our other books decisive proofs ˙ that ˙God’s
42 essence (haqı̄qa) is identical to His existence (wuğūd)’.
˙
44 Robert Wisnovsky
Malāhimı̄ (d. 532/1144). These Mu tazilites were openly opposed to Abū
˘
1
˙
2 Hāšim’s hāl-theory, at least as it was construed by its most famous proponent,
˙
Qādı̄ Abdalğabbār (d. 415/1025), and they subjected it to many of the same
˘
3
4
˙
critiques that Hayyām and Šahrastānı̄ and others had directed at the
5 Bahšamizing Sunnis˘ such as Bāqillānı̄ and Ğuwaynı̄.47
6 What might Rāzı̄’s motivations have been for embracing Avicenna’s
7 ontology in the case of contingent beings, while rejecting it in the case of
8 God? Rāzı̄ may have been unconvinced by Avicenna’s radical understanding of
9 divine simplicity, which Avicenna safeguards by completely identifying essence
10 and existence in God. For Avicenna, God’s simplicity, and thus His non-
11 compositeness, ensures that He is the only uncaused being, in contrast to all
12 other beings, whose composition of essence and existence requires a composer
13 and hence a cause. For Avicenna, in other words, God’s simplicity explained His
14
causal self-sufficiency, His immunity from causedness. But God’s causal
15
productivity, His causation of other beings, was another matter. At most, it
16
could be said that on Avicenna’s account, God’s being wāğib al-wuğūd (necessary
17
of existence) conveyed a weak form of transitivity that was perhaps implicit in
18
the active participle wāğib (necessary). But even if we were to construe wāğib as
19
synonymous with the more clearly transitive fourth-form active participle mūğib
20
(necessitating), the kind of transitivity conveyed by w-ğ-b remained one of
21
syllogistic necessitation rather than one of causal production – let alone one of
22
voluntary agency. And God’s necessitating the world just as a true premise
23
24
necessitates the conclusion that follows from it, was not a robust enough notion
25
of divine causation for Rāzı̄. Avicenna’s theology immunized God from
26
causedness very convincingly. But it made God’s causation of the world too
27 automatic, too mechanistic. By contrast, holding – as Rāzı̄ does – that God’s
28 quiddity is sufficient to cause His existence, allowed God to retain His attribute
29 of will (irāda) and thereby be plausibly seen as a voluntary agent (fā il) rather ˘
30 than as a necessitator. In sum, while Rāzı̄ adopts Avicenna’s ontology, he rejects
31 Avicenna’s theology by maintaining that God’s quiddity is not identical to His
32 existence.
33 With this Rāzı̄an background in our minds, we can now turn back to
34 Suhrawardı̄, from whence we started, and examine, with fresh eyes, his critique
35 of the notion of existence put forward by the ‘followers of the Peripatetics’. To
36 recapitulate: given that Avicenna was not explicitly committed to the thesis that
37
47 See Ibn al-Malāhimı̄, Kitāb al-Fā iq fı̄ usūl ad-dı̄n, pp. 46,3 – 48,ult. (Fı̄ anna wuğūda
˘
38
š-šay i hal huwa ˙ dātuhu aw hālatun zā˙idatun alā dātihi) and pp. 68,10 – 76,15 (Fı̄
˘
˘
˘
39 ¯ ¯ ānin wa-lā ahwālin); and Tuhfat

ālimun˙hayyun li-dātihi lā li-ma
˘
annahu ta ālā qādirun

40 ˙ ¯ ˙ by Samarqandı̄
˙ as
˘
al-mutakallimı̄n fı̄ r-radd alā l-falāsifa, pp. 61,19 – 62,19. Aš arı̄ is cited
˘
41 holding the view that in God essence and existence are identical: as-Sahā if al-Ilāhiyya,
˘
42 p. 298,5–6. ˙˙ ˙
˘
existence is something superadded to (zā id alā) a thing’s quiddity (even though
˘
1
2 Avicenna could plausibly be construed as implying it), whose doctrine might
3 Suhrawardı̄ have been pointing to here?
4 In both his Talwı̄hāt (at least as interpreted by his commentator Ibn
5 Kammūna [d. 676/1277]) ˙ and Hikmat al-išrāq (at least as understood by the
6 commentators Šahrazūrı̄ and Qut ˙ baddı̄n al-Šı̄rāzı̄ [d. 710/1310]), Suhrawardı̄
˙
accepts the incorrectness of Aš arı̄’s old position, namely, that existence is
˘
7
8 indistinguishable from essence or quiddity either in concrete reality (i. e.,
9 extensionally) or in the mind (i. e., intensionally). However, Suhrawardı̄ insists
10 on the incorrectness of the position of ‘the followers of the Peripatetics’, namely,
11 that existence is distinguishable from essence or quiddity both in concrete reality
12 and in the mind. By Suhrawardı̄’s reckoning, essence and existence are
13 distinguishable in the mind but not in concrete reality. This sounds a lot like
14 Avicenna’s position that essence and existence are intensionally distinct but
15 extensionally identical, that essence and existence can be separated conceptually
16 while remaining locked together in actual beings. Where Suhrawardı̄ goes
17 beyond Avicenna is in inferring from this that existence as well as quiddity are
˘
18 merely i tibārāt aqliyya, ‘mental constructs’ or ‘expressions of the intellect’.

19 According to Suhrawardı̄, the position of the followers of the Peripatetics
20 suffers from an infinite-regress problem: If existence really existed concretely, it
21 would itself have to be a concrete existent; otherwise the existence would be
22 non-existent, and a contradiction would ensue. But if existence were itself a
23 concrete existent, it would exist through yet another existence; and this would
24 lead to an infinite regress of existences. Suhrawardı̄ and his commentators
25 reason that, given this and other weaknesses, existence cannot be a real attribute
26 of concrete things, but must instead be understood as a mental construct.
27 Existence is therefore not something really superadded to quiddity in order to
28 make a concretely existing being.
29 It seems clear that Suhrawardı̄ and Hayyām were part of the same line of
30 critics of the Avicennian theory of existence ˘ put forward by Sunni mutakallimūn
31
such as Ğuwaynı̄ and Rāzı̄. Not only does Suhrawardı̄, like Hayyām, deny that
32
existence is an attribute superadded to quiddity in concrete ˘ reality; he later
33
reproduces Hayyām’s distinction between attributes that are merely conceptual
34
and attributes˘ that are real as well as conceptual.48 In a brief passage in his
35
Talwı̄hāt, Suhrawardı̄ also attacks the theory of modes, on the basis of its
36 ˙ of the law of the excluded middle; and the example he gives of a
violation
37
˘
mental construct (i tibār dihnı̄) that falls into this trap is ‘colorness’ (lawniyya) –
38
the same example that ¯Hayyām had given before him. Finally, Suhrawardı̄
39 ˘
40
˘
48 fa-idani s-sifātu kulluhā tanqasimu ilā qismayni sifatin ayniyyatin wa lahā sūratun fı̄ l- aqli
41 ¯ ifatin
˙ ˙ wuğūduhā fı̄ l- ayni laysa illā nafsu˙wuğūdihā fı̄ d-dihni wa-laysa
˙ lahā fı̄ ġayri
˘
… wa-s
42 ˙
d-dihni wuğūdun, Hikmat al-išrāq I.3.68, pp. 50,5 – 51,5. ¯ ¯
¯ ¯ ˙
46 Robert Wisnovsky
1 famously places his main attack on the conception of existence put forward by
2 ‘the followers of the Peripatetics’, in the section of his Hikmat al-išrāq that is
3 devoted to exposing sophistries and logical fallacies. This ˙ too may have been
4 anticipated by Hayyām, given the prominent appearance of hadayān (‘bab-
5 bling’), a term of ˘ art in Arabic sophistics, in Hayyām’s discussion. ¯ 49 Avicenna
6
˘
also uses h-d-y (specifically, mā hadaw bihi min aqāwı̄lihim) in a well known
passage from ¯ Ilāhiyyāt I.5 of his¯ Šifā , when referring to the Mu tazilites’
˘
˘
7
inconsistency on the issue of whether a non-existent (ma dūm) is a thing (šay ).50
˘
8
9 These Avicennian uses of h-d-y, and particularly that contained in the
10
¯
Ilāhiyyāt I.5 passage, are possible antecedents to Hayyām’s use of hadayān in his
11 Risālat al-Wuğūd. For Hayyām and Avicenna appear ˘ to have a common ¯ target:
12
˘
those who commit some sophism by, in this case, violating the law of the
excluded middle. In Avicenna’s case, it is the early Mu tazilites whose ‘non-
˘
13
14 existent thing’ neither is nor is not. In Hayyām’s case, it is the Bahšamizing
Aš arites, whose position that existence is˘ superadded to quiddity entails that
˘
15
˘
16 existence be construed as a mode (hāl) (or more precisely, the grounds [ma nā]
17 of a mode) that falls between existence ˙ and non-existence. The targets are
18 different, but the sophism they commit is ultimately the same. This apparent
19 link between Avicenna and Hayyām; the apparent link between Hayyām and
20 Suhrawardı̄ discussed earlier;˘ and the basic similarity between Avicenna’s˘ and
21 Suhrawardı̄’s distinctions between essence and existence; all increase the
22 likelihood that, as I hypothesized at the beginning of this chapter, Suhrawardı̄
23 was targeting the Avicennian ontology of mutakallimūn such as Rāzı̄, rather than
24 Avicenna’s own ontology; and that Suhrawardı̄ therefore intended to make an
˘
25 implicit distinction when he criticized the followers of the Peripatetics (atbā al-
Maššā ı̄n), and not the Peripatetics themselves, for holding that existence is
˘
26
˘ ˘
something superadded to quiddity (ma nan zā idun alā l-māhiyya). Interestingly,
˘
27
28 in a late work of his, the Nihāyat al- uqūl, Rāzı̄ himself appears to have struck
29 back, targeting Hayyām and Suhrawardı̄ by recapitulating Suhrawardı̄’s infinite
30 regress argument˘ against existence’s being something superadded to quiddity,
31 classifying it as a type of sophism used by radical skeptics who deny existence,
32
33
34
35
49 Risāla fı̄ l-wuğūd 401,11. H-d-y is one of a cluster of roots (the others are h-d-r, h-t-r, h-ğ-
36 r, h-m-z and h-m-r) used to render ¯ Aristotle’s term !dokeswe?m, for example¯ in Yahyā ibn
˙
˘
37 Adı̄’s translation of Aristotle, Sophistical Refutations (SE 3:165b15): see, ad loc., Mant iq
Aristū, vol. III, ed. A. Badawı̄. In the corresponding section of the Safsata (Sophistics)˙of
˘
38
his ˙Šifā (I.1, 7,5), Avicenna paraphrases Aristotle but uses the terms al-had ˙ ayān wa-t-
˘
39 ¯
takrı̄r; Kitāb aš-Šifā /Mantiq (6): as-Safsata, ed. A.F. al-Ahwānı̄; see also SE 13:173a31 –
˘
40
173b16, and Ibn Sı̄nā, Safsat ˙ a II.2: pp. 67,12 ˙ – 69,5, where hadayān and hadā appear at
41 ˙ ¯ I.1, p. 9,1. ¯
˘
68,1.2 tris.3.7.9 bis.10; cf. Kitāb aš-Šifā /Tabı̄ iyyāt (6): an-Nafs,
˘
42 ˙
50 Ibn Sı̄nā, Kitāb aš-Šifā /Ilāhiyyāt (1), p. 33,16 – 18.
˘
˘
1 and criticizing those skeptics for recklessness (ta assuf ) – the same flaw that
2 Hayyām had seen in the Bahšamizing Sunni mutakallimūn. 51
3
˘
4
5 Conclusion
6
7 The two opposing positions on this issue largely determined the metaphysical
8 doctrines of the following century. Rāzı̄’s Avicennian understanding of existence
9 as something superadded to quiddity is embraced in toto by as-Samarqandı̄ (fl.
10 ca. 690/1291) and al-Baydāwı̄ (d. ca. 716/1312), who follow Rāzı̄ in ontology
11 as well as in theology; and˙Rāzı̄’s ontology (though not his theology) is accepted
12 by al-Qazwı̄nı̄ al-Kātibı̄ (d. 675/1276) as well as by al-Abharı̄ (663/1264), at
13 least during one phase of his career.52 As was seen above, Suhrawardı̄’s
14 commentators Ibn Kammūna, Šahrazūrı̄ and Qutbaddı̄n Šı̄rāzı̄ all endorse
15 ˙
Suhrawardı̄’s view. But it would be a mistake to label this simply as the ‘Išrāqı̄’
16 view, given what appear to be Hayyām’s and Šahrastānı̄’s roles in propelling it
17 ˘
along, as well as the fact that prominent non-Išrāqı̄s also rejected the position
18
that existence is superadded to quiddity. These include the Mu tazilite followers
˘
19
of Abū l-Husayn al-Basrı̄, such as Ibn al-Malāhimı̄, as mentioned above. They
20 ˙ ˙ ˙
also include Rāzı̄’s and Suhrawardı̄’s Maġribı̄ contemporary, Averroes.53 The
21
22
51 Rāzı̄, Nihāyat al- uqūl wa-dirāyat al-usūl, Istanbul Ayasofya ms 2376, fols 22r9–v24.
˘
23
˙
˘
52 See as-Samarqandı̄, as-Sahā if al-Ilāhiyya, I.1.1.2 (Fı̄ anna l-wuğūd zā idun alā l-
˘
˘
24
māhiyyāt); al-Baydāwı̄,˙ T ˙ awāli
˙
˘
al-anwār min matāli al-anzār I.1.2.2 (Fı̄ kawnihi [viz.,

25
al-wuğūd] muštarak) ˙ and ˙ I.1.2.3 (Fı̄ kawnihi [viz.,
˙ al-wuğūd]
˙ zā idan hilāfan li-š-Šayh
˘
26 ˘ ˘
˘
[viz., al-Aš arı̄] mutlaqan wa-l-Hukamā fı̄ l-wāğib); but note I.1.4.1 (Fı̄ annahā [viz., al-
˘
wuğūb, al-imkān, ˙al-qidam and˙ al-hudūt] umūr aqliyya lā wuğūda lahā fı̄ l-hāriğ). Al-
˘
27
˙ ¯ (Fı̄ anna l-wuğūd muštarak); I.1.1.3 ˘(Fı̄ anna l-
˘
28 Qazwı̄nı̄ al-Kātibı̄, Hikmat al- ayn I.1.1.2

˙
˘
wuğūd zā id alā māhiyyāt al-mumkināt) and I.1.1.2 (Fı̄ anna l-wuğūd nafs haqı̄qat al-
˘
29 ˙
wāğib al-wuğūd). Al-Abharı̄, Kašf al-haqā iq 1 (Fı̄ kawnihi [viz., al-wuğūd] muštarakun
˘
30 ˙
˘
bayna l-mawğūdāt) and 2 (Fı̄ anna wuğūd al-mumkināt zā id alā māhiyyātihā); but note
˘
31 his opposite view in Muntahā l-afkār fı̄ ibānat al-afkār I.1.1.1 (al-mašhūr anna l-wuğūd
˘
32 amr muštarik min ğamı̄ al-mawğūdāt … wa-l-kull da ı̄f) and I.1.1.2 (al-mašhūr anna
wuğūd al-mumkināt fı̄ l-a yān zā id alā māhiyyātihā˙ … wa-huwa da ı̄f ). For a brief
˘
˘
˘
33
discussion of the major 13th-century developments in Avicennian ontology, ˙ along with
34
translations of key passages, see Eichner, Essence and existence. 13th-century perspectives
35
in Arabic-Islamic philosophy.
36
˘
53 fa-inna-mā buniya l-qawlu fı̄hā alā madhabi bni Sı̄nā wa-huwa madhabun hata un wa-
˘
¯ wa-huwa kawnu š-šay ¯i mawğūdan˘ ˙ šay un

˘
37 dālika annahu ya taqidu anna l-anniyyata

˘
¯zā idun alā l-māhiyyati hāriğu n-nafsi … ammā qawlu l-qā ili inna l-wuğūda amrun
˘ ˘
˘ ˘
38
39 zā idun alā l-māhiyyati …˘ fa-qawlun muġallatun ğiddan … wa-huwa madhabu bni Sı̄nā,
Averroes, Tahāfut at-Tahāfut, pp. 302,13 – 304,14;˙ cf. p. 197,15 – 16. Also¯ see Averroes,
40
˘
Tafsı̄r mā ba da t- tabı̄ a, pp. 1279,12 – 1280,11. Averroes’ critique of Avicenna’s ontology
41 is discussed fully ˙ ˙by S. Menn in his Fārābı̄ in the Reception of Avicenna’s Metaphysics:
42 Averroes against Avicenna on Being and Unity, in this volume.
48 Robert Wisnovsky
1 only difference between Suhrawardı̄ and Averroes is that Averroes assigns this
2 position to Avicenna himself, while Suhrawardı̄ associates it to unspecified
‘followers of the Peripatetics’. The Maġribı̄ philosophical mystic Ibn Arabı̄ (d.
˘
3
4 638/1240) similarly condemns construing existence as something superadded;
although to be precise, what Ibn Arabı̄ denies is that existence is superadded to
˘
5
6 the existent, not that existence is superadded to the essence.54
7 It remains an open question whether any of the critics of what I have labeled
8 Avicennian ontology felt that Avicenna’s own theory of existence could be
9 salvaged if it had been interpreted in a different way. However fertile they
10 doubtless were to subsequent Muslim thinkers, Avicenna’s various articulations
11 of the essence-existence distinction, and his attempts to press that distinction
12 into the service of his distinction between the necessary of existence in itself and
13 the necessary of existence through another, were perhaps so tentative and
14
inconsistent that they resisted systematization. Suhrawardı̄’s creation of a new
15
metaphysics of light might have partly resulted from his frustration with the
16
flaws that emerged from recent and ongoing attempts to systematize Avicennian
17
ontology.
18
19
20 Bibliography
21
˘
˘
22 Abd al-Ğabbār, al-Muġnı̄ fı̄ abwāb at-tawhı̄d wa-l- adl, eds M.M. Hilmı̄ et al., Cairo: al-
Mu assasa al-Mis. riyya al- Āmma, 1958 ˙ f. ˙
˘
23
A. Alami, L’ontologie modale: tude de la thorie des modes d’Abū Hāshim al-Jubbā ı̄, Paris:
˘
24
Vrin, 2001.
25
˘
al-Āmidı̄, Ġāyat al-marām fı̄ ilm al-kalām, ed. A.F. al-Mazı̄dı̄, Beirut: Dār al-Kutub al-
26 Ilmiyya, 2004.
˘
27 –– , Abkār al-afkār fı̄ usūl ad-dı̄n, vol. II, ed. A.F. al-Mazı̄dı̄, Beirut, 2004.
˙
al-Aš arı̄, Maqālāt al-islāmiyyı̄n, I-II, ed. H. Ritter, Istanbul: Devlet, 1929 – 30.
˘
28
˘
29 al-Bayhaqı̄, Tatimmat siwān al-hikma, ed. M. Šafı̄ , Lahore: L. Ishwar Das, Registrar,
˙
University of the Panjab, ˙
1935.
30
al-Bāqillānı̄, Kitāb at-Tamhı̄d, ed. R.J. McCarthy, Beirut: al-Maktaba aš-Šarqiyya, 1957.
31 al-Ġazālı̄, Tahāfut al-falāsifa, ed. M. Bouyges, Beirut: Imprimerie Catholique, 1927.
32
˘
al-Ğuwaynı̄, aš-Šāmil fı̄ usūl ad-dı̄n, ed. A.M.M. Umar, Beirut, 1999.
˙ ed. M.Z. al-Kawtarı̄, Cairo: Matba at al-Anwār, 1948.
˘
–– , al- Aqı̄da an-nizāmiyya,

˘
33
–– , Luma fı̄ qawā id ahl as-Sunna wa-l-ğamā¯ a, in al-Ğuwaynı̄,
˙ ˙ Textes apologétiques de
˘
34
35 Ǧuwaynı̄, ed. M. Allard, Beirut: Dār al-Mašriq, 1968.
˘
36
–– , al-Iršād fı̄ usūl al-i tiqād, transl. and ed. J. Luciani, Paris: Leroux, 1938.
˙
˘
M. Allard, Le problme des attributes divins dans la doctrine d’al-Ash arı̄ et de ses premiers
37 grands disciples, Beirut: Imprimerie Catholique, 1965.
38
39
˘
54 fa- lam anna l-wuğūda wa-l- adama laysā bi-šay in zā idin alā l-mawğūdi wa-l-ma dūmi
˘
40
˘
lakinna huwa nafsu l-mawğūdi wa-l-ma dūmi lakinna l-wahma yatahayyalu anna l-
41 ˘ Arabı̄, Kitāb
˘
wuğūda wa-l- adama sifatāni rāği atāni ilā l-mawğūdi wa-l-ma dūmi, Ibn
˘
42 ˙ – 17.
Inšā ad-dawā ir, p. 6,14
˘
˘
al-Qušayrı̄, Šarh asmā Allāh al-husnā, eds T. A. Sa d and S.H.M. Alı̄, Cairo: Dār al-
˘
˘
1
˙
Haram li-t-Turāt , 2001. ˙ ˙ ˙
2 ˙ ¯
˘
3
al-Rāzı̄, al-Arba ı̄n fı̄ usūl ad-dı̄n, no ed., Hyderabad: Dā irat al-Ma ārif al- Utmāniyya,
1934 – 35. ˙
4
–– , al-Mabāhit al-mašriqiyyah, vol. 1, ed. M.M. al-Baġdādı̄, Beirut: Dār al-Kitāb al-
5 ˙¯
˘
Arabı̄, 1990.
6 –– , ar-Risālat al-kamāliyya fı̄ l-haqā iq al-ilāhiyya, ed. A. Muhyuddı̄n, Beirut: Dār al-
˘
˘
Kutub al- Ilmiyya, 2002. ˙ ˙
˘
7
8 –– , Asās at-taqdı̄s, ed. A.H. Saqqā, Cairo: Maktabat al-Kulliyyāt al-Azhariyya, 1986.
˙ Atiyah, Cairo, 1936.
˘
9
–– , Lubāb al-išārāt, ed. A.S.
–– , Šarh al-išārāt wa-t-tanbı̄hāt,˙ ed. A.R. Nağafzādeh, Tehran, 1384 H.
˘
10
–– , Šarh˙ al-Išārāt wa-t-tanbı̄hāt, no ed., in Kitāb šarhay al-išārāt li-l-Hwāğa Nası̄raddı̄n
11 at-T˙ūsı̄ wa-l-Imām Fahraddı̄n ar-Rāzı̄, Qum, 1983/4. ˙ ˘ ˙
˙ ˙ ˘
˘
12 –– , Šarh Uyūn al-hikma, ed. A. as-Saqqā, Tehran, 1994.

Aristotle,˙ Mantiq Arist ˙ ū , ed. A. Badawı̄, vol. III, Cairo: Maktabat an-Nahda al-
˘
13 .
14 Mis. riyya, 1952. ˙
˙
˘
aš-Šahrastānı̄, Kitāb al-Musāra a, ed. W. Madelung, transl. T. Mayer, London: Tauris,
15 ˙
2001.
16
aš-Šahrastānı̄ and Īlāqı̄, Guftgū-yi Šahrastānı̄ va-Īlāqı̄, ed. M.T. Dānišpazhūh, Nāma-yi
17 Āstān-i Quds 3/9, 1348 H., pp. 97 – 104.
aš-Šahrazūrı̄, Rasā il aš-šağara al-ilāhiyya, vol. III, ed. N. Habı̄bı̄, Tehran: Mu assasa-i
˘
18
˘
Pažūhišı̄-i Hikmat wa Falsafa-i Īrān, 1385 H. ˙

19
Bahmanyār, Kitāb ˙ at-Tahsı̄l, ed. M. Mutahharı̄, Tehran: Da-nishkadah- i Ila-hı̄ya-t wa
˘
20 ˙˙ H. ˙
Ma a-rif-i Isla-mı̄, 1349
˘
21
R. Frank, Abu Hashim’s Theory of ‘States’: Its Structure and Function, in Actas do quarto
22
congresso de estudios rabes e islmicos, Coimbra-Lisboa, 1 a 8 de setembro de 1968,
23 Leiden: Brill, 1971, pp. 85 – 100.
˘
24 –– , Beings and Their Attributes: The Teaching of the Basrian School of the Mu tazila in the
25 Classical Period, Albany, New York: State University of New York Press, 1978.
Umar Hayyām, Darūriyyat at-tadādd fı̄ l- ālam wa-l-ğabr wa-l-baqā , in S.G. Tı̄rtha,
˘
˘
26 ˙ The Nectar of˙Grace: Omar Khayyām’s Life and Works, Hyderabad:
˘ and ed.,
˘
27
transl.
Kitabistan, 1941, pp. xcix–civ.
28 –– , Risālat al-kawn wa-t-taklı̄f, in S.G. Tı̄rtha, transl. and ed., The Nectar of Grace:
29
˘
Omar Khayyām’s Life and Works, Hyderabad: Kitabistan, 1941, pp. lxxxiii–lxxxix.
30 –– , Risāla fı̄ l-wuğūd, in Dānišnāma-yi hayyāmı̄, ed. R. Ridāzāda-yi Malik, Tehran,
1377 H., pp. 398,9 – 403,10. ˘ ˙
31
˘
32 –– , Risāla fı̄ l-wuğūd, in S. G. Tı̄rtha, transl. and ed., The Nectar of Grace: Omar
Khayyām’s Life and Works, Hyderabad: Kitabistan, 1941, pp. cx-cxiv.
33
Ibn al-Malāhimı̄, Kitāb al-Fā iq fı̄ usūl ad-dı̄n, ed. W. Madelung and M. McDermott,
˘
34 Tehran:˙Mu assasa-i Pažūhišı̄-i .Hikmat

˙ wa Falsafa-i Īrān, 2007.
˘
35 –– , Tuhfat al-mutakallimı̄n fı̄ r-radd alā l-falāsifa, ed. H. Ansari and W. Madelung,
36 ˙
Tehran, 2008.
˘
Ibn Arabı̄, Kitāb Inšā ad-dawā ir, in Kleinere Schriften des Ibn al- Arabı̄, ed. H. Nyberg,
˘
37
38 Leiden: Brill, 1919.
Ibn Rušd, On Aristotle’s ‘Metaphysics’: An Annotated Translation of the So-called ‘Epitome’,
39
ed. R. Arnzen, Berlin: De Gruyter, 2010.
40
˘
–– , Tafsı̄r mā ba da t- tabı̄ a, ed. M. Bouyges, 3 vols, Beirut: Imprimerie Catholique,

41 1938 – 48. ˙ ˙
42 –– , Tahāfut at-Tahāfut, ed. M. Bouyges, Beirut: Imprimerie Catholique, 1930.
50 Robert Wisnovsky
˘
Ibn Sı̄nā, al-Mabda wa-l-ma ād, ed. A. Nūrānı̄, Tehran: The Institute of Islamic
˘
˘
1
2 Studies, McGill University, Tehran Branch, 1984.
3
–– , al-Mubāhatāt, ed. M. Bı̄dārfar, Qum: Intišārāt Bı̄dār, 1371 H.
˙ ¯ed. H.M. Al- Ubaydı̄, Baghdad, 2002
˘
–– , at-Ta lı̄qāt,
4 ˙
–– , Kitāb al-Išārāt wa-t-tanbı̄hāt, ed. J. Forget, Leiden: Brill, 1892.
5 –– , Kitāb aš-Šifā /Ilāhiyyāt (1), eds G. Anawātı̄ and S. Zāyid, Cairo: al-Hay’a al-‘āmma
˘
6 li-šu‘ūn al-matābi‘ al-amı̄riyya, 1960.
˙
–– , Kitāb aš-Šifā /Ilāhiyyāt (2), eds M.Y. Mūsā, S. Dunyā and S. Zāyid, Cairo: al-Hay’a
˘
7
8
al-‘āmma li-šu‘ūn al-matābi‘ al-amı̄riyya, 1960.
–– , Kitāb aš-Šifā /Mantiq (1): ˙ al-Madhal I.2, eds G. Anawātı̄, M. al-Hudayrı̄ and F. al-
˘
9 ˙
Ahwānı̄, Cairo: al-Matba‘a al-amı̄riyya, ˘ 1952. ˘ ˙
10 ˙
–– , Kitāb aš-Šifā /Mantiq (6): as-Safsata, ed. A.F. al-Ahwānı̄, Cairo, 1958.
˘ ˘
–– , Kitāb aš-Šifā /Tabı̄˙iyyāt (6): an-Nafs.
˙ Avicenna’s De anima; Being the Psychological
˘
11
˙
Part of Kitāb al-Šifā , ed. F. Rahman, London: Oxford University Press, 1959.
˘
12
13
T. Izutsu, The Distinction between essentia and existentia, in The Fundamental Structure
of Sabzawari’s Metaphysics, in (viz., the introduction to) Sabzawārı̄, Šarh-i
14 manzūmah, eds M. Muhaqqiq and T. Izutsu, Tehran, 1969, pp. 49 – 70. ˙
15 T. Mayer,˙ Fakhr ad-Dı̄n ar-Rāzı̄’s ˙ Critique of Ibn Sı̄nā’s Argument for the Unity of God
16 in the Išārāt, and Nası̄r ad-Dı̄n at-Tūsı̄’s Defence, in D. Reisman, ed., Before and
˙
After Avicenna, Leiden/Boston: ˙ ˙ 2002, pp. 199 – 218.
Brill,
17
18
S. Rizvi, An Islamic Subversion of the Existence-Essence Distinction? Suhrawardı̄’s
Visionary Hierarchy of Lights, Asian Philosophy 9/3, 1999, pp. 219 – 27.
19 A. Shihadeh, The Teleological Ethics of Fakhr al-Dı̄n al-Rāzı̄, Leiden/Boston: Brill, 2006.
20 –– , From al-Ghazālı̄ to al-Rāzı̄: 6th/12th Century Developments in Muslim
21 Philosophical Theology, Arabic Sciences and Philosophy 15/1, 2005, pp. 141 – 79.
22 Suhrawardı̄, Hikmat al-išrāq, transls and eds J. Walbridge and H. Ziai, Provo, Utah:
23
Brigham ˙Young University Press, 1999.
J. Walbridge, The Science of Mystic Lights: Qutb al-Dı̄n Šı̄rāzı̄ and the Illuminationist
24 Tradition in Islamic Philosophy, Cambridge, ˙ Massachusetts: Harvard University
25 Press, 1992.
26 R. Wisnovsky, Avicenna and the Avicennian tradition, in P. Adamson and R. Taylor, eds,
27 The Cambridge Companion to Arabic Philosophy, Cambridge: Cambridge University
28
Press, 2005, pp. 92 – 136.
–– , Avicenna’s Metaphysics in Context, Ithaca, New York: Cornell University Press,
29 2003.
30 –– , Final and Efficient Causality in Avicenna’s Cosmology and Theology, Quaestio: The
31 Yearbook of the History of Metaphysics 2, 2002, pp. 97 – 123.
–– , Notes on Avicenna’s Concept of Thingness (šay iyya), Arabic Sciences and Philosophy
˘
32
33
10/2, 2000, pp. 181 – 221.
–– , One Aspect of the Avicennian Turn in Sunni Theology, Arabic Sciences and
34 Philosophy 14/1, 2004, pp. 65 – 100.
35 H. Wolfson, The Philosophy of the Kalam, Cambridge, Massachusetts: Harvard
36 University Press, 1976.
37
38
39
40
41
42
1
2
3
4 Fārābı̄ in the Reception of Avicenna’s Metaphysics:
5 Averroes against Avicenna on Being and Unity
6
7 Stephen Menn
8
9
It may seem peculiar to describe Fārābı̄ as part of the reception-history of
10
Avicenna’s metaphysics, given that Fārābı̄ died before Avicenna was born. But
11
I’ll try to show that Fārābı̄ became part of that reception-history, because
12
Averroes made him part of that reception-history.1 Averroes is harshly critical of
13
14
Avicenna’s attempt to rewrite Aristotle’s Metaphysics in what Avicenna thinks is
15
the more appropriate logical order, systematically developing the central
16
ontological concepts and making the theological conclusions depend on
17
properly ontological rather than physical demonstrations: Averroes thinks that
18
Avicenna’s improvements are disimprovements which weaken the demonstrative
19 force of Aristotle’s arguments, and leave the philosophers open to Ġazālı̄’s
20 criticisms. Now on some issues Averroes groups Fārābı̄ and Avicenna together, as
21 deviating from Aristotle and his ancient interpreters; but on some central
22 ontological issues Averroes chooses to follow Fārābı̄ against Avicenna, and
23 indeed takes Fārābı̄ to have diagnosed Avicenna’s errors before Avicenna made
24 them. Avicenna has of course great respect for Fārābı̄, and credits Fārābı̄ with
25 showing him the object [sjop|r, ġarad] of metaphysics as a science and of
˙
26 Aristotle’s Metaphysics as a treatise; so it will be particularly effective if Averroes
27 can show that Fārābı̄ had already warned against the errors Avicenna was to
28 make, and that Avicenna had disregarded his warning. And in the Epitome of the
29 Metaphysics, Tahāfut al-Tahāfut, and Great Commentary on the Metaphysics,
30 Averroes will try to show that Avicenna was involved in fundamental confusions
31 about the central ontological concepts, being and unity – that he confused
32 different senses of these concepts with each other, and that he misconstrued the
33 logical syntax of these concepts – and that Fārābı̄ had already warned against
34 these confusions in his Kitāb al-Hurūf and Kitāb al-Wāhid wa’l-wahda. These
˙
35 passages in Averroes have remained very obscure, in part ˙because it was ˙ hard to
36
37 1 I am grateful to Peter Adamson, Amos Bertolacci, Dag Nikolaus Hasse and Robert
38 Wisnovsky for their comments on a draft of this paper, as well as to everyone at the Villa
39 Vigoni conference. Rafael Njera first alerted me to the texts on unity in Averroes’
Epitome of the Metaphysics. Conversations with Marwan Rashed and Richard Taylor were
40
also helpful. This paper forms part of the promised sequel to my ‘Fārābı̄’s Kitāb al-Hurūf
41 and his Analysis of the Senses of Being’. I will often have to summarize here briefly ˙
42 arguments which are developed more fully in that article.
52 Stephen Menn
1 understand what Averroes was saying before Muhsin Mahdi edited these
2 treatises of Fārābı̄; even since then, work on the Kitāb al-Hurūf has concentrated
3 on its religio-political rather than its metaphysical aspects, ˙ and almost nothing
4 has been done on the Kitāb al-Wāhid, and so Mahdi’s work has not had the
5 payoff that it might.2 I will try to˙ build here on my reconstruction of the
6 metaphysical program of the Hurūf in ‘Fārābı̄’s Kitāb al-Hurūf and his Analysis
7 of the Senses of Being’. While ˙ some of what I said in ˙ that article may be
8 controversial, I think it will be clear that Averroes is reading the Hurūf ’s central
9 theses about being in pretty much the same way that I was, and ˙ also that he
10 takes Avicenna to have fallen into the same errors about being that Fārābı̄ is
11 criticizing (which I argued were views of the Kindı̄ circle). Furthermore,
12 Averroes’ use of the Kitāb al-Wāhid will show that he takes these two Fārābian
13 treatises to be very closely connected˙ (as Mahdi had suggested); and Averroes’
14 use of the Kitāb al-Wāhid in refuting metaphysical errors may give us a clue to
15
˙ treatise.3
Fārābı̄’s intentions in that
16 While Aristotle speaks of unity and other supremely universal attributes as
17 per se accidents of being (e. g. Metaphysics C1 1003a21 – 2), Avicenna’s
18 considered opinion seems to be that both ‘one’ and ‘being’ or ‘existent’
19 [mawǧūd] are both per se accidents of the quasi-genus ‘thing’.4 Aristotle says that
20
21 2 I will cite Fārābı̄’s Kitāb al-Hurūf, ed. Mahdi, either by Mahdi’s bāb and paragraph
22 numbers or by his page and line ˙ numbers, and his Kitāb al-Wāhid wa’l-wahda by Mahdi’s
23 paragraph numbers. Translations of the two treatises, by ˙Charles Butterworth ˙ and
24 Ahmed Alwishah respectively, have been announced.
3 I will cite the Metaphysics of the Šifā from Marmura, Avicenna, The Metaphysics of ‘The
˘
25
Healing’; the Tahāfut al-Tahāfut from Bouyges, Averroes, Tahafot at-Tahafot, cited as
26 ˘
˘
‘TT’; the Great Commentary from Bouyges, Averroes, Tafsı̄r mā ba d al-tabı̄ at, cited as
27 ‘Tafsı̄r’. I will cite the Epitome of the Metaphysics, where there is no satisfactory˙ edition,
28 from Quirs Rodrguez, Averroes, Compendio de Metafisica, by book and paragraph
29 numbers. See Davidson, Alfarabi, Avicenna and Averroes, pp. 221 – 2, n. 5, for the state of
the text of the Epitome; see also Davidson’s comments against Bruno Nardi on its
30
authenticity, and on Averroes’ later supplements and corrections, incorporated in some of
31 the manuscripts but not others. On the problems of the Epitome see also Puig Montada,
32 Cuanto se encuentra ms all de la naturaleza. I have not yet seen Arnzen, Averroes, On
33 Aristotle’s ‘Metaphysics’.
˘
4 So in the Naǧāt: ‘lakinna tabı̄ ata ’l-wāhidi min al-a rādi ’l-lāzimati li’l-ašyā i, wa-laysa ’l-
˘
34 ˙
wāhidu muqawwiman li-māhiyyati šay˙in min al-ašyā˙i, bal takūnu ’l-māhiyyatu šay an
˘
35 ˙
˘
imma insānan wa-imma farsan aw aqlan aw nafsan, tumma yakūnu dālika mawsūfan bi-
36 annahu wāhidun wa-mawǧūdun’ (Kitāb al-Naǧāt, ed.¯ Fakhry, p. 245). ¯ For other˙ relevant
37 passages and ˙ discussion, see Wisnovsky, Avicenna’s Metaphysics, pp. 158 – 60. As
Wisnovsky points out, Šifā Metaphysics I,5 says only that ‘thing’ [šay ] and ‘being’
˘
38
39 [mawǧūd] are coextensive and mutually entailing but different in meaning, without
specifying one of them as underlying and the other as its attribute, but other texts
40
including the Naǧāt passage make it clear that ‘thing’ underlies and that ‘being’ and ‘one’
41 are its attributes. (This becomes particularly clear in texts, such as those Wisnovsky cites
42 from Šifā Metaphysics V,1 and VII,1, where the same thing can have either wuǧūd in the
˘
Averroes against Avicenna on Being and Unity 53
1 the definition of a thing does not include ‘being’ or ‘one’, and Avicenna takes
2 the underlying reason for this to be that if X is a thing other than God, being
3 and unity are not contained in the essence of X, which is what the definition
4 expresses. (And God does not have a definition.) The essence of (say) horse has
5 one kind of existence in an individual horse, another kind of existence as a
6 universal in someone’s mind, but both individual and universal existence are
7 superadded to the essence: neither individuality nor universality, neither unity
8 nor multiplicity belong to the essence considered purely in itself, and this is why
9 none of these are mentioned in the definition of horse. ‘In itself it is neither one
˘
10 nor many, neither existent in re [fı̄ ’l-a yān] nor existent in the soul, neither
11 potentially nor actually in any of these things in such a way that this would enter
into horseness; rather, of itself, it is only horseness’ (Šifā Metaphysics V,1,4). The
˘
12
13 accident of existence outside the soul will be important because this is what
14 God, or some subordinate cause of being, adds to an essence in creating a thing;
15 and the accident of unity is important because numbers, in the accidental
16 category of quantity, are just collections of such accidents of unity (III,3 and
17 III,5 – 6). Averroes thinks this is all a mistake: he rejects Avicenna’s theses that
18 being and unity are accidents of horseness, and that there is a combination in re
19 of these accidents with the essence, and a cause for this combination; or rather,
20 Averroes thinks that being and unity are accidents only in senses of ‘being’ and
21 ‘one’ in which being and unity are accidents existing only in the soul, and there
22 is no composition in re. (And he thinks that these are not the senses of ‘being’
23 and ‘one’ which are objects of metaphysics, and which Aristotle is discussing in
24 texts like Metaphysics C1 – 2.) And Averroes supports all of these claims out of
25 Fārābı̄.
26
27
28
29
30
31
32
33 mind or wuǧūd in individuals.) Avicenna’s comparison of thing, being and one contrasts
34 with Aristotle, who never mentions ‘thing’ (‘ti’?) as a universal attribute, and always
speaks of being as the substratum of which unity and so on are per se accidents.
35
Wisnovsky sees a tension between Šifā Metaphysics I,5 and the other texts he cites,
˘
36 suggesting that the Metaphysics I,5 thesis that šay and mawǧūd are mutually entailing and
˘
37 coextensive but different in meaning would be undermined if šay were more

˘
38 fundamental and mawǧūd were its attribute, but I see no tension here, just as Aristotle
39 sees no tension in Metaphysics C1 – 2 in saying that being and one are mutually entailing
and coextensive but different in meaning and that unity is an attribute of being. Šay is
˘
40
more fundamental in that it is what the others are predicated of, but šay and mawǧūd
˘
41 and one are equally fundamental in the sense that none of them can be reduced to or
42 defined in terms of the others.
54 Stephen Menn
1 I Averroes against Avicenna on Essence and Existence in the

2 Tahāfut al-Tahāfut and the Great Commentary on the Metaphysics
3
4 Averroes’ most direct and extended confrontation with Avicenna’s metaphysics
5 (not necessarily the Metaphysics of the Šifā , which he may not have read)5 comes
˘
6 in the Tahāfut al-Tahāfut. Ġazālı̄ in the Tahāfut al-falāsifa assembles the
7 philosophers’ arguments – for which he mainly follows Avicenna – in support of
8 the three theses on which they are infidels and the seventeen on which they are
9 heretics, and tries to show that their arguments are not demonstrative. Averroes
10 on some points defends the philosophers, but on a remarkable number of points
11 he agrees with Ġazālı̄ against Avicenna, and tries to restate what he thinks are
12 the authentic Aristotelian arguments, which he thinks Avicenna either
13 misinterpreted or else deliberately replaced with arguments of his own
14 invention; the authentic Aristotelian arguments, so Averroes claims, will be
15 immune to Ġazālı̄’s critique.
16 In three related passages, in the fifth, seventh and eighth discussions in the
17 Tahāfut al-falāsifa, Ġazālı̄ discusses the opinion of the philosophers that, in
18 things other than God, there is a distinction between being or existence [wuǧūd
19 or inniyya] and quiddity or essence [māhiyya or haqı̄qa], that in such things
20 being is not contained in their essence, whereas ˙God is free of all kinds of
21 plurality including the essence-existence distinction. Averroes answers that this is
22 not the opinion of the philosophers, and that Ġazālı̄ has been misled by
23 following Avicenna, who deviates from the consensus of the philosophers in
24 thinking that ‘the existence [inniyya] of a thing, that is, the thing’s being existent
25 [mawǧūd], is a thing superadded to the quiddity outside the soul and as if an
26 accident in it’ (Discussion 5, TT 302,14 – 15). As Averroes complains,
27
as for the account of someone who says that existence [wuǧūd] is something
28 superadded and that the existent [mawǧūd] is not constituted by it in its substance
29 [i.e. wuǧūd is not a substantial constituent of the mawǧūd], it is a very mistaken
30 account, since it necessitates that the name “mawǧūd” signifies an accident common
31 to the ten categories outside the soul: this is the opinion of Avicenna. And it will be
asked of this accident, when it is said of it that it is mawǧūd, whether this signifies
32
the meaning of “true” [sādiq] or an accident which is mawǧūd in this accident; and
33 ˙
there will be [tūǧadu] infinitely many accidents, and this is absurd (304,13 – 305,2).
34
35
36
37 5 For evidence that Averroes’ knowledge of the Šifā is a reconstruction from at best partial
˘
38 versions and from secondary sources including Ġazālı̄’s Tahāfut al-falāsifa, see Davidson,
39 Proofs for Eternity, pp. 318 – 20 and pp. 334 – 5. Still, while Averroes sometimes seriously
misinterprets Avicenna, and often shows that he does not know the details of Avicenna’s
40
account, sometimes (as we will see below) he gives surprisingly accurate statements of
41 Avicenna’s arguments as we find them in the Metaphysics of the Šifā , whether from first-
˘
42 hand knowledge or otherwise.

1 The criticism of Avicenna (apart from the question about ‘true’) is clear enough.
2 Avicenna says if X exists, its existence must either be contained in its essence, or
3 an accident superadded to its essence. If the existence of X is contained in its
4 essence, so that ‘X exists’ will have the status of ‘man is an animal’, then X is
5 through itself necessarily existent, and Avicenna claims to be able to prove that
6 there is only one such thing, God. So anything that exists other than God must
7 exist through some accident of existence [wuǧūd]. But this wuǧūd too exists: if
8 its existence is contained in its essence, then this wuǧūd is God; if its existence is
9 an accident superadded to its essence, there will be an infinite regress of
10 wuǧūdāt. (If we try breaking the regress by saying that accidents, especially odd
11 accidents like wuǧūd, don’t have their own wuǧūdāt but rather exist through the
12 wuǧūd of the substance they inhere in, then the wuǧūd of the wuǧūd of X is the
13 wuǧūd of X, and so the wuǧūd of X is through itself necessarily existent, and so
14 the wuǧūd of X is God.) One possible solution, adopted both by some Arabic
15 and by some Latin writers after Avicenna, is to say that the wuǧūd of things is
16 God, so that when I say ‘Suqratu mawǧūdun’, ‘Socrates exists’, the word
17 ‘mawǧūd’ signifies God in the same ˙ way that when I say ‘Suqratu abyadu’,
18 ‘Socrates [is] white’, the word ‘abyad’ signifies whiteness:6 but both ˙ Avicenna
˙
19 and Averroes reject this solution. ˙
20 What is not so clear in Discussion 5 is Averroes’ own account (the account
21 that he attributes to the real philosophers from whom Avicenna deviated) of
22 what the term ‘mawǧūd’ signifies – an account which he thinks is necessary if we
23 are to understand and avoid the error that Avicenna fell into, and not merely to
24 refute his conclusions. The philosophers’ account turns on distinguishing two
25 senses of ‘mawǧūd’, and correspondingly of ‘wuǧūd’ or ‘inniyya’:
26
27
Existence [inniyya] in reality [haqı̄qa] in the things-that-are [mawǧūdāt] is a mental
intention, namely a thing’s being˙ outside the soul as it is in the soul, and what
28 signifies it is equivalent [murādif ] with the true [sādiq], which is what is signified by
29 the hyparctic [wuǧūdiyya] copula7 in a predicative ˙ judgement. For the expression
30 ‘wuǧūd’ is said in two senses: one is what is signified by the true, like our saying
‘does the thing exist [hal al-šay u mawǧūdun] or does it not exist?’ and ‘is such-and-
˘
31
32 such such-and-such [hal kadā yūǧadu kadā] or is it not such-and-such?’, and the
other is what occupies among ¯ the things-that-are
¯ the place of a genus, as when being
33
[al-mawǧūd] is divided into the ten categories or into substance and accident. And if
34 there is understood by ‘mawǧūd’ what is understood by ‘true’, then there is no
35
36 6 Nobody thinks that God inheres in Socrates: the view is that Socrates is called ‘mawǧūd’
37 by extrinsic denomination, as a diet is called healthy, not on account of a health in it, but
38 because of its relation to an extrinsic health, existing in the person or animal who eats it.
39 There are clear statements of this view in post-Avicennian thinkers notably by Eckhart of
Cologne in his Prologues to the Opus Tripartitum and by Abd al-Ġanı̄ al-Nābulusı̄
˘
40
(1641 – 1731) in his Īdāh al-maqsūd min wahdat al-wuǧūd.
41 7 ˙ ˙
That is, a non-modal copula, ‘S is˙ P’, as opposed
˙ to a modalized copula ‘S is necessarily
42 P’ or ‘S is possibly P’.
56 Stephen Menn
1 multiplicity [of existence and quiddity] outside the soul, and if there is understood
2 by it what is understood by ‘essence [dāt]’ and ‘thing’, then the name ‘mawǧūd’ is
said of the necessary of existence and ¯ of what is other than it by priority and
3
posteriority, like the name ‘heat’ which is said of fire and of hot things: this is the
4 opinion of the philosophers. (TT 302,1 – 12)
5
6 Several things here should seem at first face puzzling (and have as far as I know
7 never been discussed in the modern literature): the distinction between two
8 senses of being, the identification of one of them with truth, the assertion that
9 this is the sense we use when we ask whether X exists [hal X mawǧūdun], and the
10 claim that wuǧūd in this sense is a merely mental entity, whereas wuǧūd in the
11 other sense is a quasi-genus of the ten categories, so that neither of them is a real
12 accident really inhering in the things-that-are. What should be especially
13 puzzling is that all this is attributed to ‘the philosophers’, which (especially when
14 ‘the philosophers’ are contrasted with Avicenna) seems to mean Aristotle.
15 Aristotle distinguishes the senses of being in Metaphysics D7, but he distinguishes
16 four senses, not two: being per accidens, being per se, being as truth, being as
17 actuality and potentiality. Of these, being per se is divided into sub-senses
18 corresponding to the different categories, and so perhaps it is the quasi-genus of
19 the categories, although Aristotle’s sentences illustrating being per se in D7 are
20 not 1-place assertions ‘S is’ or ‘S is a being’ [S mawǧūdun] but 2-place assertions
21 ‘S is P’. Being as truth (the sense Averroes distinguishes from being as the quasi-
22 genus of the categories) is another sense mentioned in D7, but nothing in D7
23 would suggest that when we ask ‘does S exist?’ we are asking about being as
24 truth: all of Aristotle’s examples for being as truth (like his examples for being
25 per se) are 2-place assertions ‘S is P’, and it is hard to see what it could mean to
26 interpret a 1-place assertion ‘S is’, ‘S exists’, as ‘S is true’. Furthermore, D7 seems
27 to be interested just in the things which are said in different senses to be, and
28 which are thus emta or mawǧūdāt, and not in the status of the wuǧūd/eWmai on
29 account of which things are called mawǧūdāt/emta, as things are called white on
30 account of whiteness: D7 says nothing about whether such a wuǧūd/eWmai is an
31 accident superadded to the things which are, much less about whether such a
32 wuǧūd/eWmai might exist only in the soul and not in the things themselves. So
33 where is Averroes getting the account of mawǧūd and wuǧūd which he attributes
34 to ‘the philosophers’ and puts forward as an alternative to Avicenna’s false
35 account?
36 The key to the answer can be found in another difficult passage, in
37 Discussion 7:
38
39
Avicenna erred in this [sc. in holding that mawǧūd is not the genus of the
categories, even a genus predicated of them per prius et posterius rather than
40 univocally] only because he thought that the name ‘mawǧūd’ signified the true in
41 the language of the Arabs [i.e. of the Bedouin ; and thus that this is the lexically
42 original meaning of the word in Arabic] and that what signifies the true signifies
1 an accident – in reality8 it signifies a second, i. e. a logical, intelligible – he

2 thought that wherever the translators used [‘mawǧūd’] it signified only this meaning.
3 But it is not so. Rather, the translators intended to signify by it only what is signified
by the names ‘essence’ [dāt] and ‘thing’. And Abū Nasr [al-Fārābı̄] has shown this in
4
his Kitāb al-Hurūf and¯made it known that one of ˙the causes of the error which
5 arises about ˙this is that the name ‘mawǧūd’ is in the shape of a paronymous
6 [muštaqq] [term], and paronymy signifies an accident. Indeed, it is in the original
7 [Arabic] language paronymous, but when the translators did not find in the
8 language of the Arabs an expression which signifies this meaning which the ancients
9 had divided into substance and accident and into potentiality and actuality, I mean
an expression which is a first imposition, some of them signified it by the name
10
‘mawǧūd’.9 But they did not understand by it the paronymous sense, so that it
11 would signify an accident; rather [they used it to signify] what is signified by the
12 name ‘essence’ [dāt] – it is a technical term and not a term of [ordinary] language.
13
¯
But other translators saw fit, on account of the ambiguity which arises from this
14 [word], to express the sense which was meant in the language of the Greeks by
15 deriving, from the pronominal expression which signifies the connection of the
predicate with the subject [i.e., in Arabic, huwa/hiya as ‘pronoun of separation’], [a
16
word] which signifies this meaning, since they thought that this was closer to
17 signifying this meaning. So they used, in exchange for the name ‘mawǧūd’, the name
18 ‘huwiyya’, even though they thus had to use a [grammatical] form which does not
19 exist in the language of the Arabs; and for this reason the other school inclined to
20 the name ‘mawǧūd’. And what is understood by ‘mawǧūd’ in the sense of ‘true’ is
21 other than what is understood by ‘quiddity’, and for this reason someone can know
the quiddity who does not know the wuǧūd [i.e. someone can know what X is
22
without knowing that X is]. This meaning is necessarily other than the quiddity in
23 composite things, but in simple things it and the quiddity are one. [This is] not the
24 meaning which the translators signified by the name ‘mawǧūd’, since this is identical
25 with the quiddity. And when we say that some mawǧūd is substance and some is
26 accident, by the name ‘mawǧūd’ must be understood the meaning which the
27
˘
˘
28 8 Translating the text as Bouyges prints it, yadullu alā aradin, wa-lā budda bal fı̄
’l-haqı̄qati alā ma qūlin min al-ma qūlāti ’l-tawānı̄, a nı̄ ’l-mant˙iqiyyati. But lā budda bal
˘
29 ¯ ˙ Bouyges notes that the

is strange, and something may have gone wrong in the pointing.
30
Cairene editio princeps of 1885 prints, instead of lā budda bal, lā yadullu, and while this
31 does not yield a reasonable sense, it will make sense if an illā has dropped out, i. e. if
˘
32 Averroes wrote wa-lā yadullu fı̄ ’l-haqı̄qati illā alā ma qūlin min al-ma qūlāti ’l-tawānı̄,
¯
˘
33 a nı̄ ’l-mantiqiyyati, ‘but in reality it signifies only a second, i. e. logical, intelligible’.

9 ˙
Averroes means to contrast ‘expression of first imposition’ here with a paronymous or
34
derived term [muštaqq], i. e. a term derived by a morphological change from a term of
35
first imposition, rather than with a term of second imposition, in the sense of a
36 metalinguistic term such as ‘verb’ or ‘accusative’. Crudely speaking, concrete adjectives
37 and active and passive participles as well as finite verbs are paronymous, while the masdar
(infinitive or nomen actionis) of a verb, the abstract noun associated with an adjective, ˙
38
39 and ordinary concrete nouns not derived from verbs, are non-paronymous. On the
somewhat tortuous history of first and second impositions in Greek and Syriac and
40
Arabic see Zimmermann (ed.), al-Fārābı̄, Commentary, pp. xxx–xxxiv. On the importance
41 of the distinction between paronymous and non-paronymous terms see Menn, Analysis
42 of the Senses of Being, pp. 65 – 6 and more broadly.
58 Stephen Menn
1 translators signified, the signification which is said by priority and posteriority of

2 the essences [dawāt] of different things; but when we say that substance [is]
3
mawǧūd, there¯must be understood by it what is understood by ‘true’. (TT 371,9 –
373,3)
4
5 Here again Averroes speaks of two senses of ‘mawǧūd’ (and two corresponding
6 senses of the masdar, the infinitive or abstract noun, ‘wuǧūd’), of which one is
synonymous with ˙ ‘true’, and the other is divided into the ten categories and is
7
8 synonymous with ‘essence’ and ‘thing’. He thinks that Avicenna was led into
9 error by confusing these two senses, more precisely that he took occurrences of
10 the word ‘mawǧūd’ in (Arabic translations of ) Aristotle and other Greek
11 philosophers which in fact meant ‘essence’ or ‘thing’, as if they meant the true.
12 He adds an important element which was not in Discussion 5, that Avicenna
13 was here captive to systematically misleading translations: Avicenna went wrong
14 because he took the word ‘mawǧūd’ in a sense closer to its normal Arabic
15 meaning, not realizing that the translators, finding no natural Arabic equivalent
16 for the Greek word for whatever it is that is divided into the ten categories, had
17 resorted to various compromises, one of which was to artificially impose a new
18 sense on the word ‘mawǧūd’. ‘mawǧūd’ most literally means ‘found’, and when it
19 is extended beyond this basic sense perhaps its next-most-basic sense is ‘present’,
20 but this doesn’t seem to be what Averroes is worried about. What he seems to
21 find most misleading about the translators’ choice of ‘mawǧūd’ is not so much
22 the content of the word as its grammatical form, the fact that it is grammatically
23 paronymous (on the theory that Averroes is presupposing, every active or passive
24 participle, and also every finite verb-form, is paronymous from the masdar of
25 the verb). When a word is paronymous, its grammatical form suggests˙that it
26 signifies an accident: just as ‘white’ signifies that something has an accident (a
27 quality) of whiteness in it, and ‘runner’ or ‘runs’ signifies that something has an
28 accident (an action) of running in it, so ‘mawǧūd’, according to its grammatical
29 form, would signify that something has an accident of wuǧūd in it.
30 Indeed, Averroes more-or-less agrees that when the word ‘mawǧūd’ signifies
31 the true – as surely it sometimes does even in Arabic translations of Greek
32 philosophical texts – it signifies an accident. Only ‘more-or-less agrees’, because
33 the wuǧūd through which a thing is mawǧūd-in-the-sense-of-the-true is not an
34 accident in one of the nine accidental categories, really existing in its subject (if
35 it were, there would be an infinite regress of wuǧūdāt), but rather ‘a second, i. e.
36 a logical, intelligible’, as Averroes says here, or ‘a mental intention’, thus
37 something existing only in the mind, as he said in Discussion 5. But when
38 someone says that X is mawǧūd in the sense in which ‘mawǧūd’ is divided into
39 the categories, this doesn’t signify anything beyond the essence of X, and the
40 corresponding wuǧūd just is the essence of X: indeed, Averroes says that
mawǧūd in this sense signifies the same as ‘thing’, ‘šay ’, and so wuǧūd should
˘
41
signify the same as šay iyya, and Avicenna agrees that the šay iyya of a thing is
˘
42
1 precisely its essence.10 In sum, the wuǧūd of a thing, depending on the sense in
2 which it is taken, is either something mental or else the essence of the thing: it is
3 never something really existing in the thing and really distinct from its essence.
4
5
6 Fārābı̄ and Averroes on mawǧūd and wuǧūd
7
8 Anyway, this is what Averroes thinks. But why does he believe it, how can he
9 attribute all this to ‘the philosophers’ (i. e. to Aristotle), and where is he getting
10 the linguistic claim about a non-paronymous word in Greek for which the
11 translators substituted a paronymous Arabic word? The key is the reference to
12 Fārābı̄ and the Kitāb al-Hurūf. Averroes’ reference does not make it clear how
˙
13 much of this story he is taking from Fārābı̄, and before the Hurūf was printed
14 there was no way to know. But it is now clear that Averroes is˙ taking the whole
15 thing from Fārābı̄. As I argued in ‘Fārābı̄’s Kitāb al-Hurūf and his Analysis of the
16 Senses of Being’, one of Fārābı̄’s main concerns ˙in the Hurūf is with what
17 ˙
Aristotle calls sophisms of sw/la t/r kńeyr, logical errors arising when the
18 grammatical form of some word or sentence does not correspond to its logical
19 form and so misleads us about its logical form: he is particularly concerned with
20 errors that arise when a term is grammatically but not logically paronymous, i. e.
21 when ‘F’ does not signify ‘that in which an F-ness is present’, where ‘F-ness’ is
22 the masdar corresponding to the paronymous term ‘F’. Fārābı̄ thinks that this
23 happens ˙ especially when a term that is not paronymous in Greek is translated by
24 a paronymous term in Arabic, and he thinks that this is what has happened to
25 the Greek word ‘astı̄n’, which is rendered into Arabic by some translators as
26 mawǧūd and by others by terms derived from huwa in its use as a pronoun of
27 separation; Averroes’ story about the translators is entirely taken over from
28 Fārābı̄.11
29
Fārābı̄ also thinks, like Averroes, that being is said in two main senses,
30
being-as-truth and being as divided into the categories: this agreement is
31
particularly important since, as noted, Aristotle lists four senses of being in
32
Metaphysics D7, and no one before Fārābı̄ had reduced them to two in this way.
33
Most importantly, Fārābı̄ gives an entirely new interpretation – in which
34
35
10 See Wisnovsky, Avicenna’s Concept of Thingness; developed in Avicenna’s Metaphysics,
36 chapters 7 – 9.
37 11 Fārābı̄’s account of the translators is in Kitāb al-Hurūf I,83 – 6; see the discussion in
38 Menn, Analysis of the Senses of Being, pp. 74 – 6˙and more broadly pp. 71 – 6. Ustāt,
whose translation of the Metaphysics Averroes normally takes as his base for ˙his ¯
39
commentary, favors huwiyya, but Averroes generally rephrases in terms of mawǧūd and
40
wuǧūd, which had been the standard terms in Arabic philosophy at least since the time of
41 Fārābı̄. For everything we know about the translators of the Metaphysics see Bertolacci,
42 On the Arabic Translations.
60 Stephen Menn
1 Averroes follows him without reservation – to ‘being as truth’. From Metaphysics

2 D7 we might think that this was only a sense of 2-place being, i. e. a sense of ‘is’
3 in contexts like ‘S is P’ (in the philosophers’ artificial Arabic ‘S mawǧūdun
4 Paccusative’), but Fārābı̄ thinks that being-as-truth is also a sense of 1-place being,
5 i. e. a sense of ‘is’ in contexts like ‘F is’, ‘F exists’ (‘F mawǧūdun’): for Fārābı̄, to
6 ask whether F is, in the sense of the true, is to ask whether the concept of F is
7 instantiated, and he thinks that this is what we are very frequently asking when
8 we ask whether there are F’s.12 Fārābı̄ also, like Averroes, is very much concerned
9 with the senses of ‘wuǧūd’ corresponding to these two senses of ‘mawǧūd’: this is
10 the connection with his story about how the Arabic translators of Greek
11 philosophical texts chose a paronymous word to render ‘astı̄n’, since the
12 paronymy of ‘mawǧūd’ tempts Arabic-speakers into thinking that, when F exists,
13 it does so through a wuǧūd present in it. Fārābı̄, like Averroes, thinks that this is
14 not in fact correct for either sense of mawǧūd. Averroes is following Fārābı̄ in
15 saying that when ‘mawǧūd’ is used in the sense of the true the wuǧūd that it
16
˘
signifies is a ‘second intelligible’ [ma qūl tānı̄]: Fārābı̄ is apparently the inventor
17 ¯
of this phrase, and he thinks that such second intelligibles are the domain of
18 logic, which is why Averroes speaks of a ‘second, i. e. logical, intelligible’.13 What
19 Fārābı̄ means by saying that ‘mawǧūd’ in the sense of the true signifies a second
20
intelligible is that when I say ‘F mawǧūdun’ = ‘F exists’ in this sense, I am not
21
asserting something of a thing F, but rather asserting something of a concept of F
22
in the mind, namely that that concept is instantiated outside the mind or, as
23
Fārābı̄ says, that it ‘is outside the soul as it is inside the soul’ (Hurūf, p. 116,5),
24 ˙
25
12 For all this see Menn, Analysis of the Senses of Being. Fārābı̄ distinguishes the whether-
26 it-is question (eQ 5sti, hal mawǧūdun) which Aristotle discusses in Posterior Analytics II –
27 are there phoenixes, are there eclipses – into two questions, the first asking whether the
28 concept of F is instantiated, and then, if the answer to the first question is yes, a second
29 question asking whether F has a real essence outside the mind. See below for discussion.
˘
13 Fārābı̄ apparently coins the phrase ‘second intelligible’ [ma qūl tānı̄] (for which Avicenna
30
substitutes the equivalent ‘second intention’ [ma nā tānı̄]) on¯ the model of the older
˘
31 notion of a term of second imposition [wa d tānı̄], in ¯the sense of a metalinguistic term;
˘
32 see, as before, Zimmermann (ed.), al-Fārābı̄, ˙ ¯ Commentary, pp. xxx–xxxiv, and Menn,
33 Analysis of the Senses of Being, p. 81. A second intelligible is a thought about other
˘
34 thoughts, or a predicate of other predicates. Wa d awwal/tānı̄ translate Greek pq¾tg/

˙ ānı̄ ¯do not seem to translate any
˘
deut´qg h´sir, but ma qūl awwal/tānı̄ and ma nā awwal/t

35 ¯ ¯
Greek terminology. The phrase ‘de¼teqai 1p¸moiai’ which Zimmermann p. xxxiii cites
36 from Dexippus, In Categorias 12,14 – 15, may be an antecedent, but in context Dexippus
37 is talking about concepts signified by compound nouns and the like. Dexippus also,
38 without the phrase ‘de¼teqai 1p¸moiai’, talks about other terms that do not signify things
39 in the categories, including what Latin writers will call syncategorematics and also
including some metalinguistic terms, but he does not seem to have the idea of
40
metaconceptual concepts analogous to metalinguistic terms, as Fārābı̄ very clearly does,
41 see esp. Kitāb al-Hurūf I,7 – 10. Still, this kind of thing in the Greek commentators, one
42 ˙
way or another, gives Fārābı̄ his start.
1 the phrase which Averroes also uses14 in the passage cited above from Discussion
2 5 to paraphrase being-as-the-true in both 1-place and 2-place contexts. So
3 although the fact that F exists in this sense is non-tautological, it does not
4 involve any composition of subject and attribute in re: the wuǧūd in question
5 can exist only in the mind, being an attribute of a concept or ‘first intelligible’
6 which itself exists in the mind.
7 By contrast, the wuǧūd which is signified by ‘mawǧūd’ in the other sense, the
8 sense which is divided into the categories, is something real outside the mind,
9 but it is not an accident inhering in the thing which is mawǧūd, being rather the
10 quiddity of the thing (or perhaps a part of the quiddity). Fārābı̄ glosses this
11 sense of ‘mawǧūd’ as ‘having an essence outside the soul’ or more literally
12 ‘delimited by [munhāz bi-] some essence outside the soul’ (Hurūf, p. 116,7).15
˙
13 At first hearing this˙ sounds close to the gloss on being-as-the-true as ‘being
14
outside the soul as it is inside the soul’, but it is importantly different. Fārābı̄
15
insists that being-as-the-true, i. e. ‘being outside the soul as it is inside the soul’,
16
is predicated univocally of everything – what it is for a concept to be instantiated
17
is the same whether the concept is of something in a category like ‘man’ or
18
‘white’, a negation like ‘not-white’ or a privation like ‘blind’, or presumably an
19
ens per accidens like ‘white man’. But being as ‘having an essence outside the soul’
20
will not apply univocally to all these things, since while both man and white
21
have essences, accidents have essences only derivatively from substances (what-
22
23
it-is-to-be-white is just what-it-is-for-some-substance-to-be-white), and white
24
man and not-white and blind do not have essences at all; so being as ‘having an
25
essence outside the soul’ will apply non-univocally to things in the different
26
categories, and will not apply at all to things outside the categories. Averroes is
27 following all this in Discussion 5, a little below the passage that I cited above,
28 when he describes the sense of being other than the true as ‘what privation is
29 opposed to [i.e. what applies only to positive things and not to negations and
30 privations] … what is divided into the ten genera and is like their genus’ (TT
31 303,12 – 13). It is like a genus in that mawǧūd in this sense is an essential
32 predicate of things in all ten categories, only like a genus because ‘it is said of the
33 ten categories by priority and posteriority’ (303,15), whereas being in the sense
34 of the true is said equally of all the categories (303,17 – 304,1) and presumably
35 also of privations and other things outside the categories.
36
37 14 Although not quite verbatim: Averroes says that being as the true is kawnu ’l-šay i hāriǧa
˘
˘ qūl
˘
38 ’l-nafsi alā mā huwa alayhi fı̄ ’l-nafsi (TT 302,2 – 3), Fārābi that it is said of every ma
˘
39 which kāna hāriǧa ’l-nafsi wa-huwa bi- aynihi ka-mā huwa fı̄ ’l-nafsi. But both Fārābı̄ and
Averroes allow˘ themselves some variation in expression.
40
15 In ‘Analysis of the Senses of Being’ I said ‘circumscribed’ for munhāz. ‘Delimited’ is less
41 awkward, but runs the risk of falsely suggesting a close connection ˙with ‘limit’, nihāya (=
42 p´qar). To avoid this risk I will translate the latter, somewhat less naturally, as ‘boundary’.
62 Stephen Menn
1 I argued in ‘Fārābı̄’s Kitāb al-Hurūf and his Analysis of the Senses of Being’
2 that Fārābı̄ in the Kitāb al-Hurūf ˙ is restating and reworking, for an Arabic-
3
˙
speaking Muslim audience, what he sees as the central achievements of
4 Aristotle’s Metaphysics; that Part I of the Kitāb al-Hurūf in particular rework
5 Metaphysics D, and that the chapters on mawǧūd and˙ wuǧūd (I,80 – 103) rework
6 D7. Since Averroes accepts Fārābı̄’s analysis of the different senses of mawǧūd,
7 we would expect him to draw on these chapters of the Hurūf in commenting on
8
˙
D7 in his Great Commentary on the Metaphysics. And indeed he does, although
9 without mentioning Fārābı̄’s name as he does in the Tahāfut al-Tahāfut;16 but
10 much of what he says is closely dependent on Fārābı̄, and much of it is closely
11 parallel to discussions we have seen from the Tahāfut al-Tahāfut; and the aim is,
12 as there, to avert what he sees as Avicenna’s confusions. While Averroes,
13 constrained by Aristotle’s text, has a discussion of being per accidens at the
14 beginning (pp. 553 – 4, commenting on the text of Aristotle p. 552), and a page
15 on being actually and potentially at the end (562,5 – 563,7), his overwhelming
16 interest is in being per se and being as the true: specifically in the relations
17 between the terms ‘mawǧūd’ (and cognates) and ‘huwiyya’ (and cognates), which
18 the Arabic translators used to render the same Greek word; in the danger of
19 being misled by the paronymous grammatical form of the word ‘mawǧūd’ into
20 thinking that it signifies an accident; and in the difference between the
21 ‘mawǧūd’ or ‘huwiyya’ which signifies the essence of the thing and the ‘mawǧūd’
22 or ‘huwiyya’ which signifies the true (he says indifferently ‘haqq’ or ‘sādiq’).17 All
23
˙ he tells
of this is taken directly from the Kitāb al-Hurūf, as is the story ˙ about the
24
˙
dilemmas of the translators (should they say ‘mawǧūd’, despite its misleading
25 paronymous form, or coin a new word and say ‘huwiyya’?), very closely parallel
26 to Tahāfut al-Tahāfut 371,9 – 373,3, cited above; here as in the Tahāfut al-
27 Tahāfut passage he explicitly names Avicenna as having fallen into the trap of
28 taking ‘mawǧūd’ to signify an accident.18 (He does not say here, as he does in the
29
30
16 According to Bouyges’ index, Averroes in the Great Commentary cites Fārābı̄ by name (or
31 rather by his kunya, Abū Nasr) only three times, at 886,1 and 1498,6 and 1499,1: none
32 of these mentions are relevant˙ to the Hurūf.
17 Averroes also gets occupied in another ˙ long polemic against Avicenna, on whether a
33
34 concrete accidental term like ‘white’ primarily signifies the accident or the underlying
substance (558,7 – 559,14). Averroes’ argument here turns on a misconstrual of
35
Aristotle’s text for which the translator is at least partly to blame. I hope to return to
36 these issues elsewhere.
37 18 ‘You must know that the name “huwiyya” is not the form of an Arabic name in its origin.
38 Rather, some of the translators were compelled to this, and derived this name from the
39 connective particle [i.e. the copula, i. e. “huwa” or “hiya” as a pronoun of separation], I
mean what signifies among the Arabs the connection of the predicate with the subject in
40
its substance: this is the particle “huwa” in their saying “Zayd, huwa an animal, or a
41 man.” For the utterance of someone who says “man, huwa an animal” signifies what is
42 signified by our saying “man, his substance or essence is that he is an animal.” And when
1 Tahāfut al-Tahāfut parallel, that Avicenna thought that ‘mawǧūd’ in the

2 translations meant the true, but he does say this in the commentary on Iota 2, in
3 a passage which also parallels the Tahāfut al-Tahāfut infinite regress refutation of
4 making existence a superadded accident.)19 In the commentary on D7 (559,15 –
5
6 they found that this particle was of this character, they derived this name from it,
7 according to the custom of the Arabs in deriving one name from another, for [the Arabs]
8 do not derive a name from a particle; and this name signifies what is signified by “the
9 essence of the thing.” Some of the translators were compelled to this, as we have said,
since they thought that it would signify in the translation what was signified by the
10
expression which was used in the language of the Greeks corresponding to “mawǧūd” in
11 the language of the Arabs, indeed that it would be more apt to signify this than the name
12 “mawǧūd”, since when the name “mawǧūd” is used in the sciences to signify the essence of
13 a thing, then since it is a paronymous/derived [muštaqq] name in the speech of the Arabs
14 and paronymous/derived names only signify accidents, it is imagined that it signifies an
accident in [the thing which is said to be mawǧūd]: this is what happened to Avicenna.
15
And so some of the translators turned away from this expression to the expression
16 “huwiyya,” since this [misunderstanding, or grammatical feature giving rise to a
17 misunderstanding] does not happen with this [expression]. And if the word “mawǧūd” in
18 the speech of the Arabs signified what “thing” signifies, then it would be truer [i.e. more
19 adequate] in signifying the ten categories than the name “huwiyya”, since this name [sc.
“mawǧūd”] does occur in the speech of the Arabs. But since this [misunderstanding or
20
grammatically misleading feature] happens with the name “mawǧūd”, some of [the
21 translators] preferred to it the name “huwiyya”. For this reason, when it is used here, there
22 must not be understood by it anything of the meaning of paronymy, even though its
23 form is the form of a paronymous name, I mean the name “mawǧūd”’ (557,5 – 558,6).
24 19 ‘Avicenna thinks that “being [mawǧūd]” and “one” signify of a thing something
superadded to its essence, since he does not think that a thing exists [that it {is} mawǧūd]
25
by its essence, but rather by an attribute superadded to it, like our saying of a thing that it
26 is white. “One” and “being” according to him signify an accident in the thing; we have
27 already in another place mentioned the absurdities which follow necessarily from this
28 thesis. The first thing that follows necessarily is that it is said to him, “this attribute or
29 this accident by which the one becomes one or the existent [mawǧūd] becomes existent,
does it also become one or existent by something superadded, or by its essence?”. If he
30
says “by something superadded”, there necessarily follows an infinite regress; and if he
31 says “by its essence”, he has agreed that something exists [and is] one by its essence. [This
32 argument is also given in the commentary on C2, 314,19 – 315,9.] The man was
33 deceived by two things: first, that he thought that the one which is the principle of
34 quantity was the one which is equivalent [murādif ] to the name “being”, so he thought,
instead of [li-makāni] [thinking] that this one is enumerated among accidents, that the
35
“one” which signifies all the categories was an accident; and second, that he confused the
36 name “being” which signifies the genus [i.e. the quasi-genus said per prius et posterius of
37 the ten categories] and [the name “being”] which signifies the true: for [the name
38 “being”] which signifies the true is an accident, but [the name “being”] which signifies
39 the genus signifies each of the ten categories by a signification corresponding to the way
that “huwiyya” is said’ (1279,12 – 1280,11). I’ll come back below to Averroes’ criticism of
40
Avicenna on unity, as developed in the Epitome of the Metaphysics. Note that Averroes can
41 block the infinite regress argument also in the case of being-as-truth, since the wuǧūd in
42 this sense is just an accident in the mind: even if there is an infinite sequence of such
64 Stephen Menn
1 560,14) Averroes insists, following the Kitāb al-Hurūf, that ‘mawǧūd’ or

2 ‘huwiyya’ can signify the true in both 1-place [mutlaq] ˙ and 2-place [murakkab]
3 contexts, even though all of Aristotle’s examples of ˙ being as truth in D7 are 2-
4 place. Averroes also stresses that being as truth is something in the mind, which
5 D7 does not say but the Kitāb al-Hurūf does.20 He also stresses, following the
6 Kitāb al-Hurūf without any text in˙ Aristotle, that being as truth, unlike being
7
˙
per se, signifies an accident, ‘and for this reason the commentators on the second
8 book of the Topics differ about the simple question, i. e. about our asking [not
whether S is P but] whether the thing exists [hal al-šay u mawǧūdun], whether
˘
9
10 this falls under questions of accident or under questions of genus. Whoever
11 understands by the word “mawǧūd” here what he understands by “true” says that
12 it falls under the question of accident’ (561,8 – 13). Averroes is here copying the
13 claim about the Topics commentators directly from Fārābı̄ Kitāb al-Hurūf
14 III,246, which refers to Alexander’s Topics commentary: Fārābı̄ is thinking ˙ of
15 Alexander In topica 53,2 – 10, which does mention a disagreement about
16 whether ‘does X exist?’ is a question of genus or of accident, but the
17 interpretation of the dispute as turning on two different senses of being is
18 entirely Fārābı̄’s invention.
19
20
21 Fārābı̄ and Averroes on Posterior Analytics II:
22 a Scientific Treatment of Essence and Existence?
23
24 Both for Fārābı̄ and for Averroes, this distinction between two senses of being is
25 bound up with an interpretation not only of Metaphysics D7 but also of Posterior
26 Analytics II.21 Aristotle says that we can give a scientific definition of a thing (say
27 t¸ 1sti, grasp its quiddity) only after we know that it exists. More specifically,
28 after we know that F exists, we can ask why F exists, and grasping this cause will
29
thoughts about thoughts, the higher ones are derivative of the lower ones, not
30
presupposed by them, cf. Fārābı̄’s discussion in Kitāb al-Hurūf I,7 – 10 of why there are
31 ˙ when e. g. ‘accusative’ is put
no ‘third intentions’, just as there are no ‘third impositions’
32 in the accusative. It may not be as easy to force Avicenna to a regress of unities as to a
33 regress of wuǧūdāt: if the wuǧūd of X is mawǧūd by its essence it’s an intrinsically
34 necessary existent and there’s supposed to be only one of those, but why can’t the oneness
of X be one by its essence? Perhaps because there would be nothing to individuate things
35
whose essence consists just in their being one, which is a point Aristotle makes against
36 the Academics; but Avicenna can reply that they’re accidents individuated by their
37 subjects, unlike the Academics’ units, and also unlike anything which is mawǧūd by its
38 essence.
39 20 So Averroes says that Aristotle in his examples of being as the true ‘means only that there
is a difference between the expression “huwiyya” which signifies the copula in the mind
40
[dihn] and [the expression “huwiyya”] which signifies the essence [dāt] outside the mind’,
41 ¯
561,17 – 562,1. ¯
42 21 For all of this in Fārābı̄ see Menn, Analysis of the Senses of Being, pp. 84 – 90.
1 be grasping the essence of F: thus once we know that there are lunar eclipses, we
2 can ask why there are lunar eclipses, and the answer (‘because the earth blocks
3 the sun’s light from reaching the moon’) also gives the definition of lunar eclipse
4 (‘darkening of the moon due to interposition of the earth between moon and
5 sun’). And although lunar eclipse is an accident, Aristotle thinks a similar
6 procedure will work for matter-form composite substances. So, as Averroes says,
7 ‘the knowledge of the quiddity of the thing cannot be sought until it is known
8 that [the thing] exists’ (TT, Discussion 5, 304,3 – 4). But how can we know
9 whether there are lunar eclipses, if we don’t yet know what a lunar eclipse is? As
10 Averroes says, ‘as for the quiddity which precedes the knowledge of existence in
11 our minds, it is not really a quiddity; it is only the explanation of the meaning
12 of some name’ (TT 304,4 – 6): thus if we can explain the name ‘lunar eclipse’ as
13 ‘darkening of the moon at opposition [i.e. when it is roughly opposite to the sun
14 and so would normally be full]’, this is enough for us to be able to determine
15 whether there are lunar eclipses, i. e. whether the moon is sometimes dark at
16 opposition, although we do not yet know the quiddity of lunar eclipse. So we
17 start with a nominal definition spelling out the concept of F, then ask whether
18 this concept is instantiated, then (if the answer is yes) we investigate why F
19 exists, and in so doing we discover the quiddity of F.
20 So far this is all Aristotelian, and does not turn on a distinction in senses of
21 being. But Fārābı̄ and Averroes add that when we come to know the wuǧūd of F
22 (i. e. come to know that F exists) without knowing the quiddity of F, what we
23 know is the wuǧūd of F only in the sense that F is true, i. e. that the concept of F
24 is instantiated, and this is a predicate of the concept of F rather than of the
25 thing F: ‘the mawǧūd which is in the sense of the true is a meaning [or “entity”]
26 in the minds, namely the thing’s being outside the soul as it is in the soul, and
27 this knowledge [sc. that the thing is mawǧūd in this sense] precedes the
28 knowledge of the quiddity of the thing’ (TT 304,1 – 3). Before I have
29 investigated the quiddity of lunar eclipse I can’t know that lunar eclipses are
30 mawǧūd in the sense of having an essence, the sense of mawǧūd said by priority
31 and posteriority of the different categories, since I won’t yet know which
32 category lunar eclipses are in, or whether they are in no category at all (as if an
33 eclipse is a negation or a privation, or a being per accidens like white man); but I
34 can know that lunar eclipses are mawǧūd in the sense of truth, since mawǧūd in
35 this sense is univocal, to things in any category and to things in no category.
36 Once we know that F is mawǧūd in the sense of truth, we can go on to ask
37 whether it is mawǧūd in a second sense.
38
39
The meaning of our saying ‘is the thing mawǧūd?’, in the case of what has a cause
which determines its existence [wuǧūd], has the same force as our saying ‘does the
40 thing have a cause or does it not have a cause?’, as Aristotle says at the beginning of
41 Book II of On Demonstration [i.e. the Posterior Analytics]; and if it does not have a
42 cause, the meaning is ‘does the thing have some necessary attribute [lāzim] which
66 Stephen Menn
1 determines its existence?’. When there is understood by ‘mawǧūd’ what is

2 understood by ‘thing’ and ‘essence’, it follows the pattern of a genus which is said per
3
prius et posterius; whatever [sense of ‘mawǧūd’] in which what has a cause and what
does not have a cause do not differ, it does not signify anything superadded to the
4
mawǧūd, and this is what is meant by ‘true’, and if it signifies something superadded
5
˘
to the essence, it is only something mental [ma nan dihniyun] which has no existence
6 outside the soul except potentially, as is also the ¯ case for the universal. (TT
7 Discussion 8, 392,10 – 14).
8
This second kind of mawǧūd, unlike mawǧūd as truth, has different senses for
9
uncaused things (i. e. for God or immaterial substances) and for caused things,
10
and indeed for different kinds of caused things: the cause will in each case give
11
the quiddity, and to say that it has some cause is just to say that it has some
12
quiddity (unlike negations and privations and beings per accidens), and the
13
investigation into wuǧūd in this sense will be completed by determining more
14
precisely that cause and that quiddity.
15
In all of this Averroes is following Fārābı̄: in particular, both are
16
supplementing Metaphysics D7 with Posterior Analytics II to give a systematic
17
18
account of the meanings of mawǧūd and wuǧūd and of the relations between the
19
existence [wuǧūd] of F and essence or quiddity of F. But Averroes would not
20
follow uncritically any authority except Aristotle’s – he has very high regard for
21
Alexander of Aphrodisias, but he is perfectly capable of criticizing him, and
22
Fārābı̄ as well. And while Fārābı̄’s exegesis of Posterior Analytics II is to a large
23
extent (certainly not entirely) faithful to the text, his account especially of being-
24 as-truth (and his reading it into the Posterior Analytics) is, at the least, a
25 construction sufficiently remote from the surface of the text that we would
26 usually expect Averroes to be suspicious. The reason why Averroes is willing to
27 accept all this from Fārābı̄ is that it gives him a way of dealing with the problem
28 of Avicenna. At the most obvious level, in writing the Tahāfut al-Tahāfut
29 Averroes is confronted with Ġazālı̄’s summary of the arguments of ‘the
30 philosophers’, taken mainly from Avicenna and full of Avicennian metaphysical
31 concepts, essence and existence and contingency and necessity and so on. His
32 strategy, often, is to say that while Ġazālı̄ is right that Avicenna’s metaphysical
33 arguments do not work, these arguments are Avicenna’s innovations, and the
34 authentic Aristotle is immune to Ġazālı̄’s objections. But Averroes still has to say
35 what he thinks about essence and existence and contingency and necessity and
36 so on, so he needs an ‘authentically Aristotelian’ theory of these things to
37 counterpose to Avicenna, and Fārābı̄ shows him how to find unsuspected
38 metaphysical depths in Posterior Analytics II, not merely a methodology for
39 proving existence and defining essences, but a theory of essence and existence
40 that supports that methodology.
41 More than this, Fārābı̄ offers Averroes an explanation of how Avicenna
42 could go so badly wrong. This is a real problem for Averroes, just as it is a
1 problem for him how Ġazālı̄ could go so badly wrong. Averroes’ epistemology
2 does not admit the possibility of radically different world-views, there are just
3 different degrees of knowledge and ignorance, and anyone sufficiently familiar
4 with scientific concepts should spontaneously recognize the truth of the
5 fundamental scientific principles, from which the other propositions of the
6 sciences can be deduced. But Avicenna and Ġazālı̄ are unquestionably intelligent
7 and familiar with the sciences; so Averroes needs some explanation in their
8 cases. (In the case of Ġazālı̄ he sometimes says – perhaps with some justification
9 – that Ġazālı̄ really agrees with the philosophers and is for some reason
10 disguising his true opinion.)22 Fārābı̄, writing before Avicenna’s birth, gives
11 Averroes some explanation for Avicenna’s errors. In the first place, as we have
12 seen, Averroes accepts Fārābı̄’s story about the difficulties of the Arabic
13 translators in rendering the Greek non-paronymous word for being; Avicenna
14 would not be aware of the discrepancy between the translation and the original,
15 and would assume that what is designated by the grammatically paronymous
16 ‘mawǧūd’ is also logically paronymous, so that things are mawǧūd through the
17 presence of a wuǧūd really distinct from them.23 As we have seen, Averroes says
18 that Avicenna thought on the basis of Arabic grammar and idiom that the word
19 ‘mawǧūd’, as used by the translators of Greek philosophical texts, signified the
20 true, which is indeed logically paronymous, in the sense that ‘F is mawǧūd’
21 meaning ‘there is an F’ cannot hold on account of the essence of F unless F is per
22 se necessarily existent, and so Avicenna thought that ‘mawǧūd’ was neither a
23 genus over the ten categories or a non-univocal quasi-genus over them. But
24 Avicenna assumes that ‘mawǧūd’ is a first intelligible, signifying something in re,
25 as indeed it must be if al-mawǧūdu bi-mā huwa mawǧūdun, being qu being, is
26
27 22 At TT 347,10 – 13 he speaks of the ‘insolence’ [qubh] of Ġazālı̄’s proceeding against the
28 philosophers when he agrees with them on most of˙ their opinions; at TT 117,6 – 8 he
29 says that ‘it is clear from the books attributed to him that he relies in the divine sciences
on the opinion of the philosophers’, and that this is clearest in the Miškāt al-anwār
30
[Niche for Lights]. Similar ideas in the extended personal attack at TT 352,14 – 354,5,
31 where he says in particular that most of Ġazāli’s fame and the brilliance of his books
32 derived from his reading of the books of the philosophers. In none of these texts can
33 Averroes bring himself to name Ġazālı̄, who is just ‘this man’; but at TT 159,10 – 160,2,
34 after complaining about sophistical reasoning, he says ‘but Abū Hāmid [Ġazālı̄’s kunya]
was of greater stature than this, but perhaps the people of his time ˙ compelled him to
35
[write] this book so that he could divert from himself the suspicion that he followed the
36 opinion of the sages [i.e. the philosophers].’
37 23 In the case of God, whose wuǧūd is identical with himself, Avicenna must admit a
38 discrepancy between the grammatical and the logical form of ‘mawǧūd’. But if F is
39 contingent, Avicenna cannot say that F’s wuǧūd is F itself, or he would make ‘F is
mawǧūd’ a per se necessary truth. Averroes thinks that he can avoid this problem, where
40
Avicenna cannot, by saying that ‘mawǧūd’ in this context means the true, and that in this
41 sense the wuǧūd is a mere second intention, so neither really identical with F nor really
42 distinct from it.
68 Stephen Menn
1 an object of metaphysics rather than of logic (all parties seem to accept Fārābı̄’s
2 characterization of logic as the science of second intelligibles); so he is left with
3 wuǧūd as a real accident really distinct from contingent beings, and thus he falls
4 into the problem of an infinite regress of wuǧūdāt and other problems that
5 Averroes is delighted to raise for him.
6 The fact that Fārābı̄ has exposed all of these errors of Avicenna before
7 Avicenna made them makes his exposure all the more persuasive: this shows (so
8 Averroes can say) that what Averroes is taking from Fārābı̄ is not an ad hoc
9 response to Avicenna, rather Fārābı̄ is pointing to objectively misleading features
10 of the situation, and notably of the Arabic translations of Greek philosophical
11 texts, which make it understandable that even someone as brilliant as Avicenna
12 might be misled. And Fārābı̄ speaks with such authority about the structure of
13 the Greek language, the way it expresses the concept of being, and the
14 difficulties of the translators, that Averroes probably assumes in all innocence
15 that Fārābı̄ knew Greek, or at least had reliable sources for what he said about
16 the language.
17 It is probably also important for Averroes that, while Avicenna says that we
18 can know the essence of F without knowing whether F exists, on the alternative
19 account which Averroes (following Fārābı̄) extracts from the Posterior Analytics,
20 what we can know prior to knowing that F exists is only the explication of the
21 name ‘F’, and not the real essence of F. So Averroes can conclude that when
22 Avicenna thinks that he has grasped a real essence, independently of its existence
23 in re, he has in fact grasped only a mental concept which he wrongly regards as a
24 real nature. Fārābı̄ in the Kitāb al-Hurūf draws a contrast between dialectic,
25 which proceeds from the what-F-is question˙ (asking for the explication of the
26 name) to the whether-F-is question (asking for being-as-truth, i. e. for whether
27 the concept of F is instantiated), and science, which picks up once the work of
28 dialectic is done, and proceeds from the distinctly scientific whether-F-is
29 question (asking for being-as-having-an-essence, i. e. asking whether F has a real
30 essence, once we know that F exists in the sense of truth) to the what-F-is
31 question (investigating that essence). If Averroes takes this over from Fārābı̄, he
32 would also be taking over and applying against Avicenna the charge that Fārābı̄
33 brings against his opponents in the Kitāb al-Hurūf, that they are taking their
34 metaphysics of essence and existence from the˙ practice of dialectic, when their
35 should be taking it from the practice of science, as analyzed in the Posterior
36 Analytics.
37 In ‘Fārābı̄’s Kitāb al-Hurūf and his Analysis of the Senses of Being’ I argued
38 (or sketched an argument) ˙ that Fārābı̄ drew these various distinctions, and
39 reconstructed the syntax of being in Greek as he did, in order to attack what he
40 saw as the metaphysical error of the Kindı̄ circle and Arabic neo-Platonica like
41 the Liber de causis which take God as the being-itself in which other things
42 participate in order to exist, the wuǧūd through which other things are
1 mawǧūdāt: Fārābı̄ argues that if we mean mawǧūd-as-having-an-essence, the

2 wuǧūd of F is not extrinsic to the essence of F, and that if we mean mawǧūd-as-
3 the-true, the wuǧūd of F is a mere second intention, and so in either case it
4 cannot be God. Avicenna, of course, does not think that the wuǧūd of F is God
5 (although, as we have seen, he may be logically forced to that view to avoid an
6 infinite regress of wuǧūdāt), rather he thinks it is an accident of F, but Averroes
7 thinks that this is still committing the same fundamental error, and he turns
8 against Avicenna the machinery (including the interpretation and harmoniza-
9 tion of Metaphysics D7 and Posterior Analytics II, and at least some of the
10 speculative reconstruction of the Greek language and of the difficulties facing
11 the translators) that Fārābı̄ had apparently first designed against the Kindı̄ circle
12 and the neo-Platonica. And, as I have argued elsewhere, Averroes’ Fārābian
13 criticisms of Avicenna on being, and the need to save Avicenna’s essence-
14 existence distinction in face of them, are the starting point for Thomas Aquinas’
15 distinction between the senses of esse; and doubtless for much other Latin
16 scholastic ontology as well.24
17
18
19 Avicenna’s Response to Fārābı̄?
20
21 Thus far in this section when I have spoken of Avicenna I have been speaking
22
entirely of Averroes’ construction of Avicenna, which does not seem to be based
23
on intimate knowledge of the Metaphysics of the Šifā , and which in any case
˘
24
serves Averroes’ polemical and apologetic ends.25 But it is reasonable to ask how
25
fair Averroes’ picture of Avicenna is, and, in particular, how far he is justified in
26
making Fārābı̄ his ally against Avicenna. Avicenna too has enormous respect for
27
Fārābı̄, which makes Averroes’ use of Fārābı̄ all the more rhetorically effective.
28
Avicenna famously claims to find in Fārābı̄’s On the Aims of the Metaphysics the
29
key to understanding the object and structure of Aristotle’s Metaphysics and thus
30
the object and structure of metaphysics as a science. Averroes, however, claims to
31
find affinities rather between Avicenna and the people Fārābı̄ is criticizing in the
32
Kitāb al-Hurūf, that is (whether Averroes knows it or not) mainly the Kindı̄
33 ˙
34
24 See my Review of Kenny, Aquinas on Being. I develop these points further in work in
35
progress.
36 25 The most notorious discrepancy between Averroes’ Avicenna and the Metaphysics of the
37 Šifā is that Averroes takes Avicenna’s things ‘contingent in themselves, necessary through
˘
38 another’ to be the heavenly bodies or their movers, not recognizing that Avicenna gives
39 this status to everything except God (see Davidson, Proofs for Eternity, pp. 318 – 20). Also
Averroes says he thinks that for Avicenna ‘mawǧūd’ is said univocally of things across
40
categories (TT 370,4 – 7; and, without the hesitation, Epitome of the Metaphysics
41 III,38 – 9), whereas Avicenna in Šifā Metaphysics I,5,21 says that it is said of them by
˘
42 priority and posteriority.

70 Stephen Menn
1 circle. And it is not out of the question that, despite Avicenna’s debts to Fārābı̄,
2 he does have important agreements with the Kindı̄ circle or the Arabic neo-
3 Platonica against Fārābı̄, not on the issue of the object of metaphysics, but on
4 the analysis of the concept of being. (And, of course, he may perfectly well be
5 right to disagree with Fārābı̄ on this or other issues.) This is a very large issue on
6 which I will make only a few basic comments here.
7 While Avicenna had of course read Fārābı̄, I have not found clear evidence
8 that he had read the Kitāb al-Hurūf. If he did, evidently he did not agree with
9 ˙
everything he read. In particular, Avicenna seems never to interpret ‘being as the
10 true’ from Metaphysics D7 as Fārābı̄ does, as the sense of being found in
11 assertions like ‘there is a phoenix’, ‘there are lunar eclipses’, and as signifying a
12 second intelligible rather than something in re. 26 Indeed Avicenna, like Kindı̄,
13 seems to take Metaphysics a as a model for metaphysical argument, and in his
14
brief version of Metaphysics a1 at the beginning of Šifā Metaphysics I,8 he says
˘
15
˘
that ‘true’ [haqq] is said of what exists in re [mawǧūd fı̄ ’l-a yān], and is said in a
16 ˙
stronger degree of what exists eternally and of the causes of other things’
17
existing, or of propositions which are true eternally and which cause other
18
propositions to be true (Šifā Metaphysics I,8, following Aristotle Metaphysics a1
˘
19
993b23-31–contrast Fārābı̄’s insistence that being as truth is univocal). Avicenna
20
says here that every contingent existent is ‘true through another [sc. and thus
21
ultimately through God] and false in itself ’ (I,8,1), which sounds close to
22
23
Kindı̄.27 On the other hand, Avicenna comes strikingly close to Fārābı̄ when he
24
distinguishes two senses of mawǧūd and of wuǧūd, wuǧūd in the sense of
25
affirmation or positing [wuǧūd itbātı̄] and the wuǧūd proper to a thing [wuǧūd
¯
26
hāss : perhaps, roughly, a thing’s being what it is], corresponding respectively to
˙˙
˘Fārābı̄’s
27 wuǧūd-as-truth and wuǧūd-as-having-an-essence (I,5,9); and he agrees
with Fārābı̄ that mawǧūd in the second of these senses is equivalent with šay
˘
28
29
30
26 This despite the fact that Avicenna takes over Fārābı̄’s theory of second intelligibles as the
31
˘
objects of logic. Usually Avicenna says ‘second intention’ [ma nā tānı̄] but sometimes he
says ‘second intelligible’ [ma qūl tānı̄] like Fārābı̄, and there is no¯difference in meaning;
˘
32
¯ the subject-matter of logic is al-ma ānı̄ al-ma qūla al-
˘
at Šifā Metaphysics I,2,4 he says that

˘
33
˘
34 tāniya which depend on al-ma ānı̄ al-ma qūla al-ūlā. For Avicenna on second intentions
¯as the object of logic, see Sabra, Avicenna on Logic; but Sabra does not have a full
35
picture of the Fārābian background, and does not discuss the Kitāb al-Hurūf.
36 27 ‘The cause of the existence [wuǧūd] and affirmation [tabāt] of everything ˙ is the Truth
37 [al-haqq–i.e. God], since everything that has a being ¯ [inniyya] has a truth/reality
˙
[haqı̄qa]; so the Truth exists [mawǧūd] for beings [inniyyāt] which exist. And the most
38
˙
noble part of philosophy and its highest in degree is first philosophy, I mean the
39
knowledge of the first Truth which is the cause of every Truth’ (Kindı̄, On First
40
Philosophy 9,12 – 14 in Kindı̄, Oeuvres Philosophiques et Scientifiques, eds Rashed and
41 Jolivet). Kindı̄ is here reflecting on the same passage of Aristotle, Metaphysics a1
42 993b23 – 31, as Avicenna in Šifā Metaphysics I,8.
˘
1 [thing or something], and that the wuǧūd in this sense is identical with the
2 quiddity.
3 Thus Avicenna’s differences with Fārābı̄ seems to be concentrated on the
4 status of wuǧūd in the sense of affirmation: Fārābı̄ thinks it exists only in the
5 mind and produces no real composition with the essence of the thing, whereas
6 Avicenna thinks that it is really composed with the essence in re. And the reason
7 for this difference seems to be a different view of what subject wuǧūd is being
8 predicated of when I say ‘F is mawǧūd’ in sentences like ‘there is a phoenix’,
9 ‘thunder exists’. Fārābı̄ and Avicenna agree that ‘F is mawǧūd’ must be capable of
10 being false without being meaningless, and therefore that ‘F’ in such sentences
11 must signify something which it can signify whether it exists in re or not:
12 whatever ‘F’ signifies here will be the subject of which wuǧūd is being
13 predicated. Fārābı̄ says that in this context ‘F’ signifies the concept of F, and to
14 say that F is mawǧūd means that this concept is instantiated. Avicenna rejects
15
this: perhaps his reason is that, when I say that thunder exists, I am not saying
16
that some concept exists. At the same time, Avicenna agrees with Fārābı̄ that
17
when I say ‘F is mawǧūd’ in this sense, I am not referring to some individual F
18
˘
(some instance of the concept of F) which might be either mawǧūd or ma dūm
19
[non-existent], as some Mu tazilites thought (basing themselves on the Qur ānic
˘
˘
20
‘when God decrees a thing, he just says to it “be!” and it is’ [3:47, 40:68,
21
slightly differently 36:82] where what God is addressing must be a non-existent
22
˘
thing [šay ma dūm]; for Avicenna’s attack on the Mu tazilite thesis see I,5,12 – 18
˘
˘
23
and cf. I,5,25 – 7). Avicenna’s alternative is to say that what ‘F’ signifies here is an
24
25
quiddity which is neutral to existence in re or in the soul: it can never be
26
absolutely non-existent, it must always exist either in re or in the soul, and it is
˘
27
this neutral quiddity rather than a concept (and rather than a Mu tazilite not-
28
yet-actualized individual) of which we affirm that it is mawǧūd not only in the
29 soul but also in re; and so it is this quiddity to which wuǧūd is superadded to
30 constitute a really existing individual.
31 Assuming Avicenna’s objection to Fārābı̄’s account is as I have suggested,
32 Fārābı̄ might reply that when I say ‘F is mawǧūd’, I am not saying that the
33 concept of F exists, but rather than the concept of F is instantiated: although
34 the grammatical structure of the sentence suggests that I am asserting of
35 something that it exists, the underlying logical structure does not assert either of
36 some object or of some concept that it exists, but of some concept that it is
37 instantiated. This is what Fārābı̄ ought to say, but it is not clear that he actually
38 does: Fārābı̄ says that the mawǧūd as the true is what ‘is outside the soul as it is
39 inside the soul’ (Hurūf, p. 116,5, cited above), but what is the ‘it’ that is both
˙
40 outside and inside the soul? If Fārābı̄’s wording is to be taken at face value, the
41 ‘it’ of which being-as-truth is predicated is not the concept but an Avicennian
42 neutral quiddity which can be realized either in the soul or outside the soul; if
72 Stephen Menn
1 Fārābı̄’s wording is not to be taken at face value, it is incumbent on him to

2 explain what the correct formulation would be.
3
4
5 II Averroes against Avicenna on Being and Unity
6 in the Epitome of the Metaphysics
7
8 The Epitome of the Metaphysics gives the clearest picture of how Averroes sees the
9 main results of the Metaphysics. The Tahāfut al-Tahāfut is not about Aristotle,
10 and discusses metaphysical issues only piecemeal as they are raised by Ġazālı̄; the
11 Great Commentary is of course full of information, but in most places sticks so
12 closely to the text that it is hard to see the larger picture. But in the Epitome of
13 the Metaphysics, as in his other Epitomes, Averroes does not feel bound to follow
14
the order of Aristotle’s text, but rather restates what he takes to be the chief
15
scientific contributions of Aristotle’s text in whatever he finds the most
16
appropriate scientific order, eliminating (for instance) refutations of now
17
deceased errors, but adding discussions of more pressing errors that have arisen
18
since Aristotle’s time – and, in particular, those of Avicenna. I have argued
19
elsewhere, and recalled above, that Fārābı̄ also sees himself in the Kitāb al-Hurūf
20 ˙
as restating central contributions of the Metaphysics in this way for a new
21
linguistic-religious community, and in particular sees himself as reworking
22
Metaphysics D (far more central to the Metaphysics for Fārābı̄ than it has been for
23
most readers) in Part I of the Hurūf. What emerges from Averroes’ Epitome of
24 ˙
25
the Metaphysics is that he too read Part I of the Hurūf in this way, and took it as
˙
26
his immediate model for the part of the Epitome corresponding to D ; and that
27
he took the confusions Fārābı̄ was warning against to be close enough to
28 Avicenna’s confusions that he thought that a close restatement of Fārābı̄ would
29 serve to refute Avicenna as well. Because he is trying to bring out what he sees as
30 Aristotle’s main underlying points in the Metaphysics, without being bound to
31 Aristotle’s text, he makes more systematic and structural use of Hurūf here than
˙
32 in the Tahāfut al-Tahāfut or the Great Commentary on the Metaphysics, although
33 without ever naming Fārābı̄; but the dependence on the Hurūf is often very
˙
34 close. I will not document this in full (it is generally immediately obvious on
35 comparing the texts, and was noted by Mahdi in his introduction to the Hurūf,
˙
36 p. 39). Instead, I will first briefly review what Averroes says in the Epitome about
37 mawǧūd and wuǧūd – which adds no real content to what we have already seen,
38 but which makes his use of the Hurūf and its relation to D more obvious – and
˙
39 then go on to something that does add real content, the Epitome’s very
40 interesting account of the senses of unity and multiplicity, and criticism of
41 Avicenna on unity and multiplicity, drawn not from the Kitāb al-Hurūf but
˙
42 from the Kitāb al-Wāhid wa’l-wahda.
˙ ˙
1 Senses of Being and Senses of Unity in Epitome I

2
3 In the Epitome of the Metaphysics, before settling down to the systematic study of
4 substances and their attributes and principles, Averroes gives in Book I first a
5 relatively brief discussion of the science of metaphysics and its subdivisions and
6 their objects, corresponding structurally to Metaphysics E1 (Epitome I,1 – 17),
7 and then a much longer discussion of the different meanings of each of the
8 terms that signify the things treated in the science (I,18 – 66). This latter part of
9 Book I corresponds to Metaphysics D, but in fact for much of this text Averroes is
10 immediately dependent not on D but on Part I of the Kitāb al-Hurūf, much of
˙
11 which he extracts or closely paraphrases, without ever naming his source. In
12 particular, in his account of the meanings of ‘mawǧūd’ and ‘huwiyya’ (I,19 – 23)
13 is heavily dependent on Fārābı̄’s. Like Fārābı̄, he distinguishes between two main
14 senses of ‘mawǧūd’, the sense that is said of the ten categories per prius et
15 posterius and the sense that is said of the true, glossed as in Fārābı̄ as ‘what is in
16 the mind as it is outside the mind’; as in Fārābı̄, being-as-the-true is illustrated
17 with examples of the eQ 5sti question from the Posterior Analytics, whether
18 nature exists, whether void does not exist (I,19).28 Because of the paronymous
19 form of the word ‘mawǧūd’, and because it is transferred from the sense of
20 ‘found’, some people thought that it signified an accident in the thing rather
21 than its essence [dāt], but Averroes explains that in fact we are using it the word
22 ¯
(in one of its senses) to signify what is signified by ‘thing’ and ‘essence’ and
23
other non-paronymous terms (I,21). Averroes also adds that some people
24
thought that the two main philosophical senses of ‘mawǧūd’ were the same,
25
where presumably this is the error that he attributes to Avicenna in Tahāfut al-
26
Tahāfut Discussion 7, of thinking that the translators always used ‘mawǧūd’ to
27
mean the true. Also, some argued that ‘mawǧūd’ must signify an accident, since
28
if it signified the essence, then when we say ‘substance is mawǧūd’ (i. e. ‘a
29
substance exists’), it would be a tautology (still I,21); this is in fact an argument
30
that Avicenna makes in Šifā Metaphysics I,5,10 – 11 to show that the wuǧūd of a
˘
31
thing in the sense of affirmation [itbāt] is different from its essence [haqı̄qa] or
¯
thingness. Averroes has presumably been thinking mainly of ˙ Avicenna
32
33
throughout I,21, but I,22 is explicitly a refutation of Avicenna’s thesis that
34
‘mawǧūd’ signifies an accident in the thing: if it is a first intelligible, then either
35
it must belong to one of the nine categories of accidents, in which case being
36
won’t be predicable of substance or of the other categories of accidents except
37
38
inasmuch as this accident inheres in them, or else it is an accident somehow
39
28 Averroes subdivides the first main sense of ‘mawǧūd’ into the things that have a quiddity
40
outside the soul and the quiddities of these things, but he notes that the ‘mawǧūd’ applies
41 to the ten categories in both of these senses (I,19). Beyond the two main senses, he also
42 mentions being per accidens (I,20).
74 Stephen Menn
1 predicated in common of the ten categories, which is the thesis that in the
2 Tahāfut al-Tahāfut (Discussion 7, 370,4 – 11) he says Avicenna probably held,
3 and here as there he says that it is absurd. So if ‘mawǧūd’ signifies an accident,
4 the only possibility is that it is a second intelligible, i. e. an intelligible whose
5 existence is only in the mind; Averroes says that this is possible and indeed
6 correct for ‘mawǧūd’ in the sense of the true, but that this is very different from
7 ‘mawǧūd’ meaning what has a quiddity outside the mind; ‘all this is clear on the
8 slightest reflection, but that is the nature of this man [sc. Avicenna] in much of
9 what he brought forth out of himself ’ (all I,22).
10 Thus far, as we have seen, Averroes is closely following the Kitāb al-Hurūf.
11 But when in Epitome of the Metaphysics Book I, in the section corresponding ˙ to
12 Metaphysics D6, Averroes comes to deal with the concepts of unity and
13 multiplicity, he takes as his model, not the Kitāb al-Hurūf, which indeed has no
14 account of unity, but Fārābı̄’s Kitāb al-Wāhid wa’l-wah˙ da.29 Metaphysics D of
15
˙ ˙
course has an account of unity (D6) alongside the accounts of being (D7) and of
16 the other concepts that Averroes discusses in Epitome of the Metaphysics Book I;
17 since Averroes’ discussions follow the Kitāb al-Wāhid wa’l-wahda for unity and
18 multiplicity, and the first Part of the Kitāb al-H˙urūf for pretty ˙ much all the
19 other concepts, it seems that he thought of ˙ these two texts together as
20 constituting Fārābı̄’s updated version of D, and the model for his own updated
21 version. With unity, as with being, Averroes’ overriding aim in the Epitome is to
22 expose what he sees as Avicenna’s errors, and again he thinks that the errors
23 Fārābı̄ is exposing are the same or close enough that Fārābı̄’s distinctions will still
24 be the relevant ones to draw.
25 The main relevant texts here are Epitome I,34 – 43 and III,34 – 48. As usual,
26 Averroes mentions Avicenna by name only a few times, but in such a way as to
27 make clear that Avicenna is on his mind throughout. The main aim is to refute,
28 or restrict the scope of, Avicenna’s claim that being and unity, as predicated of
29 things other than God, signify something superadded and accidental to the
30 essence of the thing. Avicenna goes beyond Aristotle’s theses that being and
31 unity are not parts of the essence of a thing (and therefore should not be
32 mentioned in its definition), and that unity is a per se attribute of being, to the
33 conclusion that both being and unity are per se attributes of the most universal
34 quasi-genus, al-šay , ‘thing’, consequences of it rather than contained in its
˘
35 essence. Nor are they contained in the essence of any particular species of al-šay :
˘
36 ‘in itself it is nothing at all but horseness: in itself it is neither one nor many,
˘
37 neither existent in re [fı̄ ’l-a yān] nor existent in the soul, neither potentially nor
38 actually in any of these things in such a way that this would enter into horseness’
39 (Šifā Metaphysics V,1,4). Such an essence is capable of receiving existence in re
˘
40 and thus individuality, or existence in the mind and thus universality, and it is
41
42 29 As Mahdi notes in his Arabic introduction to the Kitāb al-Wāhid wa’l-wahda, p. 19.
˙ ˙
1 likewise capable of receiving unity or multiplicity. Averroes’ response in the case

2 of unity is parallel to his response in the case of being. He thinks that Avicenna
3 has misinterpreted the reasons why being and unity are not mentioned either as
4 genera or as differentiae in the definitions of things; and Averroes thinks that
5 the idea of something that is in itself not existent or one, and receives existence
6 and unity as accidents, involves the same absurdities that Aristotle attributes to
7 the Platonists who generate the many things from being and what is not in itself
8 being but comes to participate in being, or from the one and what is not in itself
9 one but comes to participate in unity. And Averroes’ way of avoiding these
10 absurdities, and of diagnosing and avoiding Avicenna’s conclusion, is again
11 closely parallel in the case of unity and in the case of being. With unity, as with
12 being, he distinguishes two main senses of these attributes. In the sense of
13 mawǧūd as having a quiddity outside the mind, man is mawǧūd by his essence,
14 and the reason that mawǧūd is not mentioned in the definition alongside
15 wingless and biped and animal is that a wingless biped animal is already
16 mawǧūd, that ‘mawǧūd’ signifies the whole quiddity rather than signifying some
17 part that could be superadded to another part that is not already mawǧūd by its
18 essence; and Averroes will say that there is likewise a sense of ‘one’ in which man
19 is one by his essence, and in which ‘one’ signifies the whole quiddity rather than
20 a part of the quiddity or an accident superadded to the quiddity. In the sense of
21 ‘mawǧūd’ as the true, Averroes concedes to Avicenna that ‘mawǧūd’ signifies an
22 accident not included in the quiddity, but he argues that this is something
23 existing only in the mind, and he likewise concedes that there is a sense of ‘one’
24 in which it signifies something not included in the quiddity, but again he argues
25 that this is something existing only in the mind. Averroes seems to think that
26 these conclusions, in the case of unity as in the case of being, follow from
27 Fārābı̄’s distinctions. Fārābı̄ does indeed distinguish senses of ‘one’, and he does
28 indeed think that one of these senses of ‘one’ is equivalent to the sense of
29 ‘mawǧūd’ as signifying the categories and to ‘thing’, and that a thing’s unity in
30 this sense is just its quiddity; it is less clear that he thinks, as Averroes does, that
31 ‘one’ has a second main sense in which the thing’s unity exists only in the mind.
32
33
34 Senses of Unity and Ways of Being Delimited
35
36 Epitome of the Metaphysics I,34 – 43 are devoted to unity, corresponding
37 structurally to Metaphysics D6; much of the content is also drawn, directly or
38 indirectly, from Metaphysics D6. But as Averroes’ discussion of mawǧūd differs
39 from Aristotle’s by his interest in determining the wuǧūd by which things are
40 mawǧūd, and in particular whether this wuǧūd is a real accident of the things or
41 is identical with the things or exists only in the mind, so his discussion of ‘one’
42 differs from Aristotle’s by his interest in the unity by which each thing is one,
76 Stephen Menn
1 whether this is a real accident or the essence of the things or exists only in the
2 mind. Both here in Epitome I,34 – 43 and later in III,34 – 48 (which corresponds
3 structurally to Metaphysics Iota 1 – 2)30 this interest leads Averroes to focus on
4 two particular issues neither of which is really Aristotelian: first, the thesis,
5 taken from Fārābı̄’s Kitāb al-Wāhid, that various kinds of things are called ‘one’
6 because they are ‘delimited’ [munh ˙ āz] from other things, either by a place or by
7 a boundary or by a quiddity; and ˙second, the question how ‘the one which is the
8 principle of number’ (and which is, like number, an accident in the category of
9 quantity) arises and how it is distinguished from other senses of ‘one’. On both
10 of these issues, Averroes is in direct confrontation with Avicenna.
11 Epitome I,34 – 43 is a rearranged and rationalized version of Metaphysics D6.
12 Averroes skips the long treatment of one per accidens which Aristotle insists on
13 putting at the beginning of D6 (1015b16 – 36), and he also plays down the
14 treatment of unity in species or genus or by analogy (D6 1016a17-b1,
15 1016b31 – 1017a3), inserting short versions of both of these toward the end, in
16 I,41: he concentrates heavily on the different ways in which something is called
17 one in number (corresponding roughly to D6 1015b36 – 1016a17 and
18 1016b11 – 31). In ordinary usage, something is said to be one in number in
19 the strongest sense if it is continuous (either per se, like a one-, two- or three-
20 dimensional continuous quantity, or through an accident present in it, like a
21 homoeomerous natural body which is continuous through possessing a three-
22 dimensional continuous quantity), especially if it is something complete, like a
23 circle (Aristotle’s example, D6 1016b16 – 17); something is called one in
24 number in a weaker sense if it is composed of several juxtaposed homoeomerous
25 bodies bound together in such a way that they all move together when any of
26 them is moved (it is one in a stronger sense if it is composed by nature, like a
27 hand, in a weaker sense if it is composed by art, c. D6 1016a4 – 6); and also ‘an
28 individual one in form’, like Socrates, is called one, presumably not just in the
29 way that his right hand is called one, but also because, while he is made up of
30 different juxtaposed homoeomerous bodies, they are all contained by a single
31 complete natural form. Thus far (Epitome I,34 and the beginning of I,35)
32 Averroes is broadly following his Aristotelian model, but he then adds a
33 reflection on the kinds of unity or oneness on account of which ordinary people
34 [al-ǧumhūr] call things one in number: these people call these things one only
35 ‘inasmuch as they are delimited [munhāz] from other things and isolated
36 [munfarid] in themselves’ (I,35).31 Different ˙ senses of ‘one’ correspond to
37
38
39 30 The parallel with Metaphysics Iota continues through Epitome III,63. Epitome III,34 – 48
correspond to Iota 1 – 2 on unity, and then III,49 – 63 discuss multiplicity, sameness,
40
difference, contrariety, opposites and intermediates, corresponding to Iota 3 – 10.
41 31 The word ‘munfarid’, which I have translated literally enough as ‘isolated’, is sometimes
42 (but not consistently) used in the translation-literature for ‘jah’ 6jastom’, ‘particular’ or
1 different ways in which a thing can be delimited from other things. Most
2 obviously, something can be delimited by a place which contains it. Other
3 things are delimited not by a place but just by their boundaries [nihāyāt], ‘and
4 they are contiguous’: the sense is apparently that the boundary of body A and
5 the boundary of body B are ‘together’, i. e. that these boundaries are in the same
6 place, so that bodies A and B, without being separated, are divided from each
7 other by real boundaries, which become two really distinct surfaces, although in
8 the same place, when the division is effected (see Aristotle, Physics V,3). Still
9 others have ‘their delimitation by imagination alone, and in this way number
10 attaches to what is continuous’: what is continuous has no actual internal
11 dividing boundaries (but only the potentiality for being so divided) and thus a
12 three-foot length does not contain three feet actually delimited from each other
13 by boundaries, but nonetheless we can measure it by delimiting these parts in
14
the imagination. In all of these cases – in all the cases recognized by the
15
ǧumhūr–‘one in number’ signifies ‘things which are external to their essences,
16
and in general, accidents attaching to them which are inasmuch as they are
17
undivided’ (all I,35).
18
Averroes argues that Avicenna was misled by these ordinary-language cases
19
and overgeneralized to the conclusion that every numerical or individual unity,
20
that is, every delimitation [inhiyāz], is an accident of the thing which is one. ‘In
21 ˙
this art’ – metaphysics, as opposed to ordinary-language contexts – ‘the one is
22
used equivalently [murādifan] with being’ (I,37), and Averroes argues that unity
23
24
in this sense is not an accident. ‘By the one in number is sometimes signified an
25
individual which cannot be divided inasmuch as it is an individual, like our
26
saying one man and one horse’ – a human being can be divided, but not into
27
two human beings (the example comes from Aristotle, D6 1016b3 – 6). But to
28 show more clearly that the relevant kind of unity is not an accident, and is
29 distinct from the ordinary kinds of unity that he has listed, Averroes turns to a
30 different kind of example, a blend such as oxymel (a blend of honey and
31 vinegar, used as a medicine). The concern is evidently not with the sense in
32 which the oxymel itself is one, but with the sense in which the honey and the
33 vinegar are each one and delimited from each other. Because oxymel is a real
34 blend and not just a juxtaposition, the honey and the vinegar are not delimited
35 from each other by spatial boundaries, like contiguous things; but they are also
36 not like the parts of a continuous magnitude, which have no determinate
37 demarcation in themselves, since honey and vinegar are distinct by their own
38 natures (all I,37). So, while ‘the delimitation of continuous magnitudes is
39 something external to their substance, this is not so for the delimitation of a
40
‘individual’ (so for instance at Metaphysics Iota 1 1052a32 and a35). But the Arabic,
41 unlike the Greek, strongly connotes isolation or separation from other things, and, as a
42 passive form, raises the question, ‘isolated by what?’.
78 Stephen Menn
1 blended thing from what is blended out of it’ (ibid.): what makes the honey in
2 oxymel one thing, and the vinegar in oxymel one other thing, is that each
3 possesses, and is ‘delimited’ from the other by, its own intrinsic nature, not by a
4 distinct place or boundary but also not merely in the imagination.
5
It is clear that by ‘one’ here, when there is meant by it what is individually one, there
6 is signified only the delimitation of the designated [mušār ilayhi, rendering ‘t|de ti’]
7 individual by its essence and its quiddity, not the delimitation by anything external
8 to its essence, as when we say about this designated water that it is one in number:
9 for the delimitation, in something like this, is only an accident in the water, and
therefore the water remains the same when it is delimited and when it is not
10
delimited, in the way that accidents succeed each other in the substratum without it
11 changing in its substance. (I,38)
12
13 That is: I can say that this water is one, in the sense that it is continuous: this is
14 to assert that it is distinguished from everything else by being delimited by a
15 determinate bounding surface, and, negatively, that it is not divided by any
16 further boundaries internal to that surface. The unity of the water, in this sense,
17 is an accident superadded to the water, since the water can persist in its
18 substance whether it is divided into many drops or reunited into a continuous
19 mass. But any individual can also be called ‘one’ as being ‘delimited’ by its own
20 quiddity, an attribute which all of it possesses and which distinguishes it from
21 everything else; and its unity in this sense is essential to it, and the thing could
22 not continue to exist if this unity were removed.32 Averroes thinks that
23 Avicenna, overgeneralizing from cases like the one-as-continuous, thought that
24 ‘one in number’ always signifies a delimitation which is an accident superadded
25 to the essence of a thing. He would be doing this, not exactly because he did not
26 recognize a distinctively metaphysical sense of ‘one’ in which it is equivalent to
27 ‘being’, distinct from the ordinary-language senses, but because he did not
28 recognize that this metaphysical sense had a distinct logical syntax, and assumed
29 that the syntax of the ordinary-language senses applied to it as well.
30
31
32
33
34
35
36
37 32 But note that this holds only if the object has a quiddity, that is, if it is mawǧūd in the
38 sense divided into the categories. Averroes in the Great Commentary on the Metaphysics
39 (following a misleading translation) takes Iota 2 1053b20 – 21 to say that ‘mawǧūd’ is
more widely extended than ‘one’, and he explains this by saying that mawǧūd in the sense
40
of the true need not be one (Tafsı̄r 1271,12 – 17). Thus e. g. the concept ‘pile of stones’
41 can be instantiated, and yet the pile of stones, not being delimited by a quiddity outside
42 the mind, will be not one but many.
1 Against Avicenna: Why Unity in the Metaphysical Sense is not an Accident

2
3 However, Averroes thinks that Avicenna also had a supplemental argument to
4 support this conclusion, and therefore that to avoid Avicenna’s conclusion (and
5 the ensuing difficulties), we need not only to point out the difference between
6 the popular and the metaphysical senses of ‘one’, but also to respond to this
7 argument:
8
For this reason Avicenna thought that ‘one in number’ signifies only an accident in
9 substance and in anything else which is delimited, and that it cannot signify the
10 substance of a thing, I mean [that it cannot signify] a delimitation which is not
11 superadded to the essence of the substance. This is because he thought that if he
12 granted that ‘one in number’ signifies a delimitation which is an accident in the
13 accident and a substance in the substance [i.e. that what ‘one in number’ signifies,
when it is said of X, is really identical with X and so falls under the same category as
14
X], then number would be composed of accidents and substances, and so would not
15 fall under a single category, much less under the category of quantity, and this would
16 be absurd. He also said that if we stipulate that it signifies substance alone, another
17 absurdity would follow, namely that substances would inhere in accidents; and
18 otherwise how will we say about a designated accident that it is one in number? And
19 he went wrong in this only in that he paid attention to the popular [ǧumhūrı̄]
signification and thought that the delimitations and unities of things were accidents
20
in all delimited things. We will explain this further in our account of the one and
21 the many [i.e. in Epitome III,34 – 63, corresponding to Metaphysics Iota]. (I,39)
22
23 Avicenna’s point would be: a number is an accident in the category of quantity
24 (it is a peculiar kind of accident, because it seems to be predicated not of any
25 one subject but of many at once, but Aristotle indisputably says that the
26 category includes both continuous quantities and discrete quantities). But a
27 number is composed of ones, and these ones must be homogeneous with what is
28 composed out of them, which is to say that they too must be accidents of
29 discrete quantity, i. e. unities by which some thing is one. If things were one by
30 their essence, so that the unity of X would be in whatever category X is in, then
31 these unities could not combine to compose numbers. This is in fact a fairly
accurate summary of some of Avicenna’s arguments in Šifā Metaphysics III
˘
32
33 (although Avicenna seems not to use the concept of ‘delimitation’, which
34 Averroes takes from Fārābı̄). In particular, Avicenna argues that the unity of an
35 accident cannot be a substance inhering in the accident (III,3,10); and while he
36 does not seem to raise the possibility that a single number might be composed
37 out of some substantial unities and some accidental unities, he does argue that if
38 ‘one’ were said non-univocally of substances and accidents, so that it would
39 signify a substance in substances and an accident in accidents, then there would
40 be some numbers composed out of substantial unities and others composed out
41 of accidental unities (III,3,13), which he takes to be absurd, presumably for the
42 reason Averroes suggests, that these numbers could not all fall under the same
80 Stephen Menn
1 category. (Avicenna also gives other arguments that unity in substances cannot
2 be a substance, e. g. by arguing that it is not a constituent in their definition
3 either as a genus or as a differentia, and therefore must be an accident, III,3,10.)
4 Averroes, in response, agrees that numbers must be accidents composed out
5 of accidents of unity, and that if things were simply one by their essence they
6 could not be counted by numbers belonging to the category of quantity. But
7 Averroes says, as usual, that Avicenna has confused two different senses of ‘one’,
8 in this case the distinctively metaphysical sense and ‘the one that is the principle
9 of number’. (The metaphysical sense, ‘individual unity’, can also be called ‘one
10 in number’, but not in the sense that there are numbers composed out of it, at
11 least not numbers in the category of quantity, the objects of arithmetic.) And
12 indeed Avicenna is very insistent that there is a univocal sense of ‘one’ that
13 applies (as an accident) to beings in all categories (III,3,13 – 15), and that the
14 unity from this universal metaphysical predicate is also the unity out of which
15 numbers are composed (III,3,17). Averroes agrees with Avicenna that the kind
16 of one out of which numbers are composed is an accident of unity, but he tries
17 to show that this kind of unity, and also the numbers arising from it, exist only
18 in the soul, contrasting with other senses of unity, including the metaphysical
19 sense equivalent to being, which exist in re independently of souls. (By contrast,
20 Avicenna not only thinks that the unity which is the principle of number exists
21 in re, he also insists that numbers themselves exist in re independently of souls,
22 directly contradicting Aristotle, Physics IV,14 223a22 – 5: at III,5,2 he rejects the
23 formula ‘number has no wuǧūd except in the soul’ and corrects it to say ‘number
˘
24 has no wuǧūd abstracted from the numbered things which are in re [fı̄ ’l-a yān],
25 except in the soul’ – just as horseness has wuǧūd in an individual, and wuǧūd in
26 the soul, but no wuǧūd abstracted from both.) Averroes says a bit about this in
27 Epitome I,36, as part of his survey of the various ordinary-language senses of
28 ‘one in number’ in I,34 – 6 before he gets to the distinctively metaphysical sense
29 in I,37 – 40; he adds more detail in Epitome III.
30
31
32 Against Avicenna: Quantitative Unity is Mind-Dependent
33
34 Averroes says in I,36 that ‘in this way [i.e. from accidents attaching to things
35 inasmuch as they are undivided] there arises in the mind [fı̄ ’l-dihn] the one
36
¯ from these
which is the principle of number, because when the intellect abstracts
37 individuals this meaning, “indivisible into two or more individuals”, this is the
38 one which is the principle of number, and when the mind repeats it, number
39 arises.’ Like the unity of what is delimited by a place or a boundary, this is one
40 of the senses of individual unity that are recognized by ordinary people [al-
41 ǧumhūr]; in all of these cases, something is one through an accident, and so they
42 all contrast with the metaphysical sense of unity, or as Averroes also calls it ‘one
1 absolutely [al-wāhid al-mutlaq]’, in which a thing is one by its essence. Unlike

2 the unity of what˙ is delimited
˙ by a place or a boundary, the unity that is the
3 principle of number is an act of the soul: ‘the one in number and numerical
4 unity is only something which the soul produces in existing things; were it not
5 for the soul, there would not exist here numerical unity or number at all’
6 (III,41). As this text says, the soul produces this unity in existing things: the
7 thing that is one is something existing outside the soul, but the unity through
8 which it becomes one is an act of the soul – just as, more famously, intellectus est
9 qui facit universalitatem in rebus, so that the things which are universal exist
10 outside the intellect but it is an act of the intellect which makes them
11 universal.33 Why do numerical unity and number depend on the soul in this
12 way, when (Averroes says by contrast, III,41) continuous quantities of whatever
13 dimension do not? All of these things, as the mathematician studies them,
14 depend on some sort of act of abstraction – the surface which the mathematician
15 treats may be the boundary of a natural body, but he abstracts from this and
16 does not treat it qu boundary of a natural body (III,37) – but number is ‘more
17 abstracted from matter’ than continuous quantities (III,41). The reason emerges
18 from III,35: ‘the numerical one is the thing designated [al-mušār ilayhi] in
19 thought, not divided in it into quantity’ (a parallel in III,36 adds ‘or into quality
20 or position’). And, although Aristotle never says that unity depends on the soul,
21 this rests on what he says in Metaphysics Iota, that the one in each genus is
22 something either indivisible or taken as indivisible that measures the things in
23
that genus: in many cases, as with the foot as a measure of length or the
24
semitone as a measure of musical intervals, no genuinely indivisible measure is
25
available, and we must treat something divisible as if it were indivisible; and
26
Averroes reasonably concludes that the indivisibility or undividedness which
27
makes it one is conferred by an act of the soul, not merely abstracting from
28
irrelevant features of the object, but treating the object as if it possessed a feature
29
which it does not possess apart from the soul’s act of considering it as an
30
indivisible unit.
31
Averroes stresses the difference between ‘the one absolutely’ and ‘the one as
32
the principle of number’ that the one absolutely is always predicated of a subject
33
in some particular category, and, because it is predicated of its subject essentially,
34
always itself belongs to the same category as its subject, whereas ‘the one which
35
is the principle of number, if it is grasped inasmuch as it is one, does not
36
manifest its dependence on the subject, and only in this way can it fall under the
37
38
category of quantity’ (III,38). ‘The numerical one is the meaning of the
39
33 This is the form in which the phrase is standardly cited in Latin; it’s sometimes cited as
40
coming from Averroes’ Great Commentary on the De anima, Book I, text 8, but there the
41 Latin apparently has ‘intellectus est qui agit in eis universalitatem’ (so in Averroes,
42 Commentarium magnum in Aristotelis de anima libros, ed. Crawford, p. 12, lines 25 – 6).
82 Stephen Menn
1 individual, abstracted from quantity and quality, I mean that by which the
2 individual is an individual, since it is an individual only by the meaning
3 “undivided”, and the mind abstracts it from matters and grasps it as a separate
4 meaning’ (III,41).34 Once we have abstracted such unities from various
5 individuals, we can collect them into numbers by the act of counting, even if the
6 individuals themselves belong to different categories, and so cannot be called
7 ‘one’ (or anything else) univocally: entities that really exist in such diverse
8 subjects could not be aggregated, but entities produced by the soul in
9 considering them can be.35 Because Avicenna (says Averroes, III,41) failed to
10 distinguish between the one absolutely and the one that is the principle of
11 number, he thought that the one in number was an attribute really existing in its
12 subject (Avicenna ‘wanted to make the case of number like the case of line and
13 surface, I mean that it would have a nature even if there were no soul’, III,41): it
14 would have to be an accident predicated univocally of its diverse subjects, so that
15 the different unities could be aggregated into numbers, and he was thus
16 confronted with the problem what the single subject is of which ‘one absolutely’
17 is univocally predicated. Evidently, it is being, but being is not a single univocal
18 genus of the different categories; and so, Averroes says, Avicenna was forced to
19 posit that the being which is the subject of the one is an accident predicated
20
univocally of things in the ten categories (III,38 – 9)–that is, although Averroes
21
does not say so here, that this being is the univocal accident being-as-truth,
22
wrongly construed as really present in the things rather than as dependent on
23
the soul.36
24
25
34 Compare I,36 cited above. The phrase I have translated ‘only by the meaning
26 “undivided”’ might instead be translated ‘only by an undivided meaning/entity’ – we
˘
27 could read either ‘bi-ma nā “ġayri munqasimin”’ or ‘bi-ma nan ġayri munqasimin’. But I
28 think Averroes means the former: this seems to be what makes the individual an
29 individual, and it seems to be an appropriate building-block of discrete quantities. There
is a similar ambiguity in I,36, where the phrase I translated ‘this meaning, “indivisible
30
into two or more individuals”’ might instead be translated ‘this meaning/entity which is
31 indivisible into two or more individuals’.
32 35 Averroes is here probably thinking of Aristotle’s point in Metaphysics M6 – 7 that
33 mathematical ones are all ‘comparable’ or ‘associable’ [sulbkgt\] with each other and so
34 can be combined into numbers, while other kinds of ones need not be, so that it may not
be possible (say) to form a two out of a one of this type and a one of that type.
35
36 So too, in the Great Commentary on Metaphysics Iota, Averroes takes the last colon of Iota
36 1, 1053b7 – 8, to distinguish one as indivisible in substance from one as indivisible in
37 quantity, and to say that the ignoring of this distinction is responsible for the error of
38 those who say that the one in discrete quantity is a substance, and therefore that numbers
39 are substances (1267,8 – 14); and then Averroes adds on his own behalf that the same
reason led to Avicenna’s believing ‘that the one which is the principle of number is a
40
genus to the ten mawǧūdāt [i.e. the categories], in such a way that it signifies an accident
41 common to them’, since he did not distinguish the one which is the principle of number
42 from the one which is equivalent [murādif ] with being (1267,15 – 1268,3).
1 Thus for Avicenna unity as a per se attribute of being (more precisely, unity
2 as consequent on being and, like being, a per se attribute of ‘thing’) would be an
3 accident in the category of quantity; and manyness or multiplicity (katra), as the
4 per se attribute predicated of several beings jointly, would be number, ¯ as an
5 accident in the category of quantity composed out of quantitative unities
6 aggregated together. As Averroes correctly reports (III,43), Avicenna objected (in
7 Šifā Metaphysics III,3,5) to the standard definition of number as ‘multiplicity
˘
8 composed of unities or ones’, on the ground that multiplicity is not the genus of
9 number but simply is number, and that it is essential to all multiplicity to be
10 composed of unities. But, Averroes says, just as unity absolutely is more general
11 than quantitative unity, so multiplicity absolutely is more general than number
12 or quantitative multiplicity: even things in different categories can be called
13 ‘many’ in the first sense (thus Aristotle can argue against Parmenides that if any
14 substance has an accident, beings are more than one), whereas an accident in the
15 category of quantity could not belong to things in several different categories,
16 except that as a mental act of counting it can depend on the unities formed by
17 the mind in abstracting from these categories. Indeed, if number were a real
18 univocal accident as Avicenna describes it, then a number three, being three
19 units, would itself be three in just the same way that three horses or three virtues
20 are three, and there would be a regress to a further accident of threeness; if the
21 number is the soul’s act of counting the three things, it is not itself (in the same
22 sense) three and there is no regress.37
23
24
25
26
27
28
29
37 The infinite regress argument is not explicit, but this looks like what Averroes means in
30
Epitome III,43 when he speaks of a difficulty arising from the fact that, on Avicenna’s
31 view, number would be numbered like other numbered things. (Averroes says that this is
32 in fact possible, but only so far as a number is an act of the soul in the things that are
33 numbered.) Avicenna’s answer would be that the number three is not three units,
although it is composed out of three units (see Šifā Metaphysics III,5,4 – 8, drawing on
˘
34
Aristotelian texts such as Metaphysics Z17 and H3 which argue that the whole is not the
35
same as all its parts but is something over and above them, and specifically that a number
36 is not the same as all its parts but is something over and above them). So the three units
37 will be three through a further accident, namely the number three, and the number three
38 will not itself be three through any further accident. However, perhaps Averroes’ point is
39 not that on Avicenna’s view the number three would have to be three through a further
accident of threeness, but rather that numbers can, like other things, be counted (e. g. we
40
can say that there are four prime numbers less than 10), that this will require a further
41 accident of number, and that such an infinite sequence is not absurd in successive acts of
42 the soul, but is absurd in real accidents.
84 Stephen Menn
1 Averroes on Aristotle’s Aporia about

2 Whether Being and Unity Exist Separately
3
4 Averroes in Epitome III,44 – 8 steps back to put the issues about unity in a larger
5 perspective by giving his version of an aporia in Aristotle’s Metaphysics B (B4
6 1001a4-b25), which Aristotle raises there both about being and unity: are being
7 and unity subsisting things whose natures are just being and unity, and which
8 could be the first principles from which other things are derived, as Plato and
9 the Pythagoreans (and Eleatics) say,38 or is it as the pre-Socratic physicists say,
10 that ‘being’ and ‘one’ are always attributes of some other underlying nature, such
11 as fire or whatever the thinker in question takes the first principles to be?
12 Averroes gives his version of the history of this disagreement among the
13 ‘ancients’, which is considerably more reminiscent of debates in Arabic
14 philosophy than Aristotle’s original account was. He is willing to attribute a
15 surprising degree of consensus to the ancients. Both the pre-Socratic physicists
16 and their Platonist opponents agreed with Aristotle that ‘being’ [mawǧūd] and
17 ‘one’ are equivalent, and, more surprisingly, they agreed that there is a first one
18 which is a principle of all other mawǧudāt and a cause to them of their existing
19 [their wuǧūd] and also of their being counted and measured and therefore being
20 knowable. The physicists noticed that ‘there is in each genus a first one, which is
21 the cause of the wuǧūd of each species of that genus, and the cause of these other
22 species’ being measured and known’ (III,45), e. g. the semitone among musical
23 intervals or the daily motion of the heavens among motions (the idea and the
24
examples, but not the attribution to the physicists, are from Aristotle,
25
Metaphysics Iota 1 – 2 and parallels). Each of these genera is said per prius et
26
posterius, and is said first of some first thing, as ‘the hot is said of fire, and of the
27
other things related to fire, per prius et posterius, and fire is the cause of the
28
wuǧūd of other hot things and of their being measured and counted’ (ibid.), fire
29
being the homogeneous unit by which hot things are measured. And since
30
beings, too, are called beings per prius et posterius, substances primarily and
31
accidents derivatively, they speculated that ‘there is a first mawǧūd which is the
32
cause of all the other mawǧūdāt’s being mawǧūd and numbered and known, as
33
the one is the cause of all the other species of number being mawǧūd and
34
numbered’ (ibid.).
35
These are surprising things to attribute to pre-Socratic physicists: they
36
sound more like Arabic interpretations of Aristotle. The example of fire and the
37
38
analogy between heat and being, fire and the principles of being, comes from
39
38 Presumably the Pythagoreans say this about the one, not about being. Aristotle doesn’t
40
mention the Eleatics in the formulation of the question, but in the course of the aporia
41 he introduces Parmenides’ argument that (as Aristotle interprets it) if there is something
42 which is just being itself, nothing else can exist (1001a29-b1, cf. Physics I,2 – 3).
1 Aristotle Metaphysics a1, a text (not especially the fire comparison) favored
2 notably by Kindı̄, and while Aristotle there says nothing about the one or
3 measuring (indeed Aristotle never apparently never says that God is one at all,
4 preferring ‘simple’, see K7 1072a31 – 4), for Kindı̄ it is the ‘true One’ which is
5 the first cause of all beings as fire is of all hot things; and Averroes’ ‘first mawǧūd
6 which is the cause of all the other mawǧūdāt’s being mawǧūd’ is virtually a
7 quotation of the first sentence of Fārābı̄’s Perfect City, ‘the first mawǧūd is the
8 first cause of the wuǧūd of all the other mawǧūdāt.’ But Averroes represents
9 these as matters of broad agreement: what is distinctive of the physicists is that,
10 believing in the priority of sensible individuals, and being unaware of any but
11 the material cause, they made the first mawǧūd and first cause of wuǧūd
12 whatever they thought the first underlying matter was, e. g. fire (III,45; Averroes
13 is perhaps combining the merely analogical mention of fire in Metaphysics a1
14 with the mention of fire as a candidate for the underlying nature of the one or
15 being at B4 1001a15). By contrast, Plato (not explicitly named, III,46), who
16 was aware of the formal cause but wrongly conceived it as an intelligible or
17
˘
intention or thought [ma qūl = m|gla] existing apart from the mind and apart
18
from the individuals that fall under it, ‘said that the universal one, common to
19
everything which is called one, is the cause of wuǧūd of all the other mawǧūdāt
20
which are said to be one, and the cause of their being measured’ (III,46).
21
This, then, is Averroes’ version of the opposition between the Platonic and
22
materialist sides of the aporia of Metaphysics B4 1001a4–b25, filled out with the
23
help of the idea of a first cause of wuǧūd from Metaphysics a1 as interpreted by
24
Arabic philosophers;39 he then gives in Epitome III,47 – 8 what he sees as the
25
distinctively Aristotelian solution to the aporia. Against Plato, Aristotle
26
˘
˘
concluded that ma qūlāt or universals have their existence qu ma qūlāt (as
27
opposed to their existence in sensible instances) only dependently on the mind:
28
29
so the universal ‘one’ cannot be realer than other things and the cause of wuǧūd
30
to them. Sensible individuals fall into the ten categories, and there is no real one
31
beyond these categories; but within each category of accidents there is some first
32
one, not as a universal applying to everything in the category, but as a unit (with
33
the usual examples), which is the cause of wuǧūd and of being measured to other
34 things in that category; and so Aristotle concluded that within substance too
35 there must be a first one which is the cause of wuǧūd to all other substances, and
36 thus to all other mawǧūdāt as well, since their wuǧūd is only through substance.
37 And since he concluded that it is a one separate from matter which is most
38
39 39 Including Averroes himself, Tafsı̄r 14,13 – 15,5, culminating ‘al-wuǧūdu wa’l-haqqu
˙ ātihi
˘
innamā ’stafādathu ǧamı̄ u ’l-mawǧūdāti min hādihi ’l- illati fa-huwa ’l-mawǧūdu bi-d
40 ¯ ¯
˘
faqat wa’l-haqqu bi-dātihi wa-ǧamı̄ u ’l-mawǧūdāti innamā hiya mawǧūdātun wa-haqqun

41 ˙ ˙ ¯
bi-wuǧūdihi wa-haqqihi’ (15,2 – 5). Aristotle himself in Metaphysics a1 is more concerned˙
42 about causes of ˙truth than about causes of being.
86 Stephen Menn
1 deserving of the names ‘one’ and ‘mawǧūd’, this investigation of the first one
2 comes to the same as the main investigation of the Metaphysics, ‘whether there is
3 a separate substance which is the principle of sensible substance, or whether
4 sensible substance is sufficient in itself for wuǧūd’ (III,48).40 Indeed, Averroes
5 seems to be mimicking Fārābı̄’s discussion, in Kitāb al-Hurūf I,92, of how to
˙
6 investigate whether there is a first mawǧūd,41 but rewriting the discussion for
7 ‘one’ instead. Either way, Averroes’ point is that all that an honest Aristotelian
8 can do by way of finding a first mawǧūd or a first one is to investigate causally,
9 step by step and category by category, rather than leaping from the many
10 instances to a single universal: there is no universal mawǧūd or universal one
11 beyond the categories, either existing separate from the categories, as he thinks
12 Plato believed, or a universal mawǧūd common to the categories with a universal
13 one as its per se attribute, as he thinks Avicenna believed.42 Avicenna is thus
14
represented as a halfway step to Plato, ignoring the fundamental Aristotelian
15
lesson that being and unity are not anything ‘beyond the categories’ [paq± t±
16
c]mg]; presumably Averroes also thinks that Avicenna has answered the part of
17
Aristotle’s aporia about being (rather than about unity) in a Platonic way in
18
saying that God is just a separate wuǧūd that is not the wuǧūd of any essence.
19
Averroes’ ‘Aristotelian’ answer to the aporia is much closer to what he has
20
attributed to the physicists: it differs in recognizing non-material causes and
21
immaterial substances, but it posits outside the mind only individuals, and it
22
23
posits no being or unity that is not the being or unity of some determinate
24
essence in one of the ten categories.
25
(However, to listen to Avicenna rather than Averroes: from his own point of
26
view, he too is giving an anti-Platonic answer to the aporia about unity. Indeed,
27
28 40 Much, although not all, of this has parallels in the Great Commentary on Metaphysics Iota
29 1 – 2. Averroes sees Aristotle in these chapters as searching for a ‘one in substance’ (for the
Aristotelian basis of this phrase see Iota 2 1054a4 – 9), analogous to the ones in other
30
categories or in sub-categorial genera like colors and sounds, and he thinks that this one
31 in substance will be the first mover (Tafsı̄r 1277,15 – 1278,5). Perhaps due partly to a
32 misleading translation, Averroes takes Iota 2 1053b11 – 15 as asking, not whether the one
33 is itself a substance or an attribute of some other underlying nature, but whether ‘one’
34 signifies a non-sensible substance or a sensible substance (1269,12 – 1270,4); Aristotle
will correct both the physicists’ answers that it is some sensible substance, and Plato’s
35
answer that it is a non-sensible substance which is the universal one, by giving the correct
36 non-sensible substance.
37 41 See Menn, Analysis of the Senses of Being, pp. 93 – 6.
38 42 ‘We have stipulated that the subject of the one absolutely is nothing over and above the
39 ten categories … for [otherwise] the subject of the numerical one would have to be either
something common to all ten categories, as Avicenna says, or something separate, as
40
many of the ancients thought about the nature of the one that they [sc. the one and
41 being?] are separate things’ (III,38). Averroes says that Aristotle will refute the Platonic
42 view, and he himself turns to refute the Avicennian view in III,39.
he develops his version of this Aristotelian aporia at some length, Šifā
˘
1
2 Metaphysics III,3,11 – 17, in support of his claim that unity is an accident, i. e.
3 that unity could not subsist if separated from the substance to which it belongs.
4 There cannot be an indivisible which is just indivisible, without any underlying
5 subject which is indivisible: at a minimum it must be an indivisible wuǧūd. If
6 unity is just indivisibility, it can never be separated from its subject, since it
7 would then have to have some other subject, and its former subject would have
8 to have some other unity, and absurdities would follow [III,3,12]. If unity is not
9 just indivisibility but indivisible wuǧūd, then perhaps it might subsist by itself,
10 but such a self-subsisting indivisible wuǧūd could not be predicated both of
11 substances and of accidents – an understatement, since Avicenna has argued that
12 there can be only one subsisting wuǧūd, God – and so if there is such a unity, it
13 is less universal than the unity that is predicated of beings in all categories, and
14
this maximal universal is what we really mean by ‘unity’ [III,3,13 – 15].
15
Avicenna thus gives an anti-Platonic answer to the aporia on unity but a
16
Platonic answer to the aporia on being.)43
17
While Averroes’ critique of Avicenna on unity is logically complex, his
18
fundamental point is that unity, in the sense in which it is an attribute of every
19
mawǧūd, is not an accident of that mawǧūd (except in a sense in which it would
20
be mind-dependent), but rather is constituted by the essence of that mawǧūd,
21
just as the wuǧūd of a mawǧūd (except its mind-dependent wuǧūd-as-truth) is
22
23
not an accident but is constituted by the essence. There is no essence neutral to
24
existing-in-the-mind and existing-outside-the-mind, or neutral to unity and
25
multiplicity, and there is no cause bestowing wuǧūd or unity on such an essence,
26
just the usual Aristotelian causes to bestow form on a preexisting substratum.44
27 ‘Wāhid’, like ‘mawǧūd’, is grammatically paronymous, so there will be a
˙
28 grammatical appearance that things are wāhid through a wahda other than their
˙ ˙
29 essence, as that they are mawǧūd through a wuǧūd other than their essence. But,
30 more importantly, Averroes thinks Avicenna is led astray by the fact that
31 ordinary-language senses of ‘wāhid’, like ordinary-language senses of ‘mawǧūd’,
˙
32
33 43 You might argue, mimicking Avicenna’s argument about unity, that even if there is a
34 subsisting wuǧūd, it cannot be the wuǧūd that is predicated of beings in all categories,
and that this maximal universal is what we really mean by ‘wuǧūd’. But, despite what
35
Averroes says, Avicenna seems not to believe that there is a universal wuǧūd predicated
36 univocally of substances and accidents, whereas he does believe this about unity. Averroes
37 insists that, since unity is a per se attribute of being, X and Y cannot be univocally one if
38 they are not univocally beings, but Avicenna seems not to accept that inference.
39 44 Compare Aristotle, Metaphysics H6, arguing that there are no causes of unity to an
essence except, if the thing has matter and form, the efficient cause which actualizes the
40
potentiality of the matter; if the thing has no matter, ‘the essence is immediately a one,
41 just as it is immediately a being, and hence there is no other cause to any of these things
42 of being one or of being a being’ (1045b3 – 5).
88 Stephen Menn
1 are indeed logically paronymous; Averroes spells out these ordinary-language

2 senses of (individual) unity, except the mathematical sense of one as the
3 principle of number, as the delimitation [inhiyāz] of a thing by some accident (a
4
˙ being divided, and he argues that
place or a boundary), and negatively as its not
5 a thing is one, in the distinctively metaphysical sense of ‘one’ equivalent to
6 being, through being delimited by its own essence, rather than by an accident
7 which might perish while the thing continues to exist.
8
9
10 Fārābı̄ and Kindı̄ on Unity
11
12 Now Averroes’ analysis of one sense of ‘one’ as ‘delimited by its quiddity’
13 [munhāz bi-māhiyyatihi] is taken directly from Fārābı̄’s Kitāb al-Wāhid wa’l-
14 wahda, ˙ § 17, and is as far as I know entirely original to Fārābı̄. Fārābı̄ has
˙ argued
15
˙
in the preceding paragraph that something can be called one either because it is
16 delimited [munhāz] either by its boundary [nihāya] or by its place [makān]. The
17
˙ in these senses seem to be the continuous and the contiguous
things that are one
18 respectively, since Fārābı̄ has said that the continuous is called one because its
19 boundary is one (§ 12), and says later that ‘the continuous is not divided in its
20 boundaries, and the contiguous is not divided in the places of its boundaries, for
21 both [boundaries of two contiguous things] are in one primary place’ (in the
22 summary § 95, drawing on Aristotle’s definitions of continuous and contiguous
23 in Physics V,3). ‘The continuous becomes one only because something else in it
24 is one’, namely its boundary (§ 12), and the contiguous also is one only through
25 whatever binds its parts together, ‘and the meaning of its unity is the bond
26 [ribāt] by which some of them are attached [lazima] to others, whether by
27 nature˙ or by art’ (§ 13). So things are one in these senses, not through their
28 essence, but through something else present in them (and this something else
29 will have to be one in a stronger sense than they are, for the contiguous to avoid
30 a regress of bonds, for the continuous because its boundary will be of one fewer
31 dimension). By contrast,
32
‘one’ is also said of what is delimited by its quiddity – whatever quiddity it is,
33
divided or undivided, conceived or outside the soul – and it is what is delimited by
34 … its share [qist] of wuǧūd. For the one in this meaning naturally accompanies
35 ˙
[yusāwiqa] mawǧūd, like ‘thing’, for there is no difference between saying ‘every
36 thing’ and ‘every one [kullu wāhidin = “each”]’. And it is said like this of all the
˙ thing, and of other things – if there are any –
categories, and of this designated
37
outside the categories. (§ 17)
38
39 And if a thing is one inasmuch as it is mawǧūd, and its oneness is its ‘share of
40 wuǧūd’, and as we know from the Kitāb al-Hurūf the wuǧūd of a thing,
41 inasmuch as it is something having a quiddity˙ outside the soul, is just that
42 quiddity, then also the oneness of a thing in this sense will be its quiddity and
1 not an accident superadded to it. Indeed, the Hurūf explains the non-mind-
2 dependent sense of ‘mawǧūd’ as ‘delimited by some ˙ essence outside the soul’
3 (Hurūf, p. 116,7) – the same explanation that the Kitāb al-Wāhid gives for the
4
˙
relevant ˙
sense of ‘one’. This is not the most obvious ordinary-language sense
5 either of ‘mawǧūd’ or of ‘one’, but Fārābı̄ thinks that if we imitate Aristotle’s
6 method in Metaphysics D6 (on unity) and D7 (on being), starting from the
7 ordinary-language meanings and pushing deeper, we will see that the deepest
8 meanings of ‘one’ and ‘mawǧūd’, as applied to mind-independent things,
9 converge. In this sense ‘one’ will not be an accident of ‘mawǧūd’, nor will they
10 both be accidents of ‘thing’, rather the oneness, the wuǧūd, and the thingness or
11 essence will all be really identical outside the mind. Fārābı̄ draws precisely this
12 conclusion in the Principles of the Opinions of the People of the Perfect City: ‘[the
13 First’s] wuǧūd by which it is delimited [munhāz] from the mawǧūdāt other than
14 it cannot be other than that by which it is ˙in itself mawǧūd; and therefore its
15 delimitation [inhiyāz] from what is other than it is through a unity which is its
16 essence. For one˙ of the meanings of unity is the specific [hāss] wuǧūd by which
17 each mawǧūd is delimited [munhāz] from what is other than ˘ ˙˙ it, and this is that
18
˙
by which each mawǧūd is called one inasmuch as it is mawǧūd with respect to
19 the wuǧūd which specifies it; and this is the meaning of “one” which
20 accompanies [yusāwiqu] mawǧūd. So the First is also one is in this way, and
21 more deserving of the name “one” and its meaning than any one other than it’
22 (68,7 – 13).45 The Perfect City is here taking up the technical terminology, and
23 the definitions of different senses of ‘being’ and ‘one’, from the Kitāb al-Hurūf
24 and Kitāb al-Wāhid, and applying them to the case of the divine first principle,˙
25
˙
in very much the way that, on Fārābı̄’s reconstruction, Metaphysics K applies the
26 conceptual distinctions and definitions of Metaphysics D ; and it gives Fārābı̄’s
27 constructive alternative to uncritical theories of the senses in which God is the
28 first mawǧūd and the first one and the cause to all other things of wuǧūd and
29 unity.46
30
31 45 I cite the Perfect City (Mabādi ārā ahl al-madı̄na al-fādila) from Fārābı̄, On the Perfect
˘
˘
32 State, ed. Walzer, by page and line numbers. Fārābı̄ says˙ that the First is ‘also one in this
33 way’ because he has just said that the First is also one by being indivisible, and
34 specifically ‘indivisible in its substance’ and therefore ‘one in substance’ (68,6), ‘since one
of the meanings in which “one” is said is what is indivisible’ (68,2 – 3); and because he
35
had said previously that it is ‘isolated [munfarid] in its wuǧūd’ (i. e. that there is nothing
36 else of its species) and also ‘isolated in its rank’ (meaning esp. that it has no contrary),
37 and that it is therefore one in these ways (66,5 – 7, cf. 62,6 – 7). Both ‘indivisible’ (in
38 various ways) and ‘isolated by some wuǧūd’, as well as ‘delimited by some quiddity’, are
39 senses of ‘one’ discussed in the Kitāb al-Wāhid; for ‘isolated by some wuǧūd’, and how it
differs from ‘delimited by some quiddity’, ˙see below.
40
46 The technical sense of what Fārābı̄ is saying here in the Perfect City seems never to have
41 been noticed before, and it could not have been noticed until both the Kitāb al-Hurūf
42 and Kitāb al-Wāhid, with their respective accounts of ‘mawǧūd’ and ‘one’ as having ˙ the
˙
90 Stephen Menn
1 It seems, then, that Averroes in the Epitome of the Metaphysics is taking the
2 Kitāb al-Wāhid as his model for his accounts of unity and multiplicity,
3 corresponding ˙ to Aristotle’s Metaphysics D6 and connected passages in Iota, as he
4 takes the Kitāb al-Hurūf as his model for the rest of Epitome I,18 – 66,
5 corresponding to the˙ rest of Metaphysics D, and especially for his account of
6 mawǧūd and wuǧūd, corresponding to D7; and it seems that he takes the Kitāb
7 al-Hurūf and Kitāb al-Wāhid to be parts of a single Fārābian project of
8
˙
distinguishing ˙
senses of fundamental metaphysical terms and averting the errors
9 that arise from confusing their different senses or from confusion about their
10 logical syntax. Furthermore, in both cases it seems that Averroes sees Fārābı̄’s
11 main contribution as in averting a metaphysical error of Avicenna, before
12 Avicenna even made it. And this suggests that Averroes sees the Kitāb al-Wāhid,
13 like the Kitāb al-Hurūf, as devoted to exposing and averting the metaphysical ˙
14 ˙
errors of some earlier philosopher or philosophers whose errors were close
15 enough to Avicenna’s that the same arguments would work against both. And,
16 while the Kitāb al-Wāhid keeps a mysterious silence about its motivations and
17 the grander conclusions ˙ it might support, it seems very likely that Averroes is
18 right in this, that the Kitāb al-Wāhid, like the Kitāb al-Hurūf, is directed against
19 ˙
˙ and the Arabic adaptations
Kindı̄ or philosophers of his circle of neo-Platonic
20 texts associated with them. And this helps to remove a difficulty about the
21 interpretation of the Kitāb al-Hurūf as directed against Kindı̄: why is Fārābı̄ so
22 concerned to avert the idea that ˙ other things are mawǧūd through participation
23 in God as wuǧūd-itself, when this seems to be a relatively minor commitment of
24 Kindı̄’s, while Fārābı̄’s version of Metaphysics D in Part I of the Hurūf entirely
25 leaves out D6 on unity and multiplicity, although Kindı̄ is˙ much more
26 concerned with God as the ‘true One’ than as wuǧūd-itself ? The answer is that
27 Fārābı̄ is not leaving that out but saving it for the Kitāb al-Wāhid, and that the
28
programmatic parts of the Kitāb al-Hurūf are needed to explain ˙ the motives of
29 ˙
the Kitāb al-Wāhid. The two texts must be read as part of the same project:
30
perhaps, as Mahdi ˙ had suggested (in his Arabic introductions to the Kitāb al-
31
Hurūf, pp. 42 – 3, and to the Kitāb al-Wāhid, p. 25) the Kitāb al-Wāhid was
32 ˙ ˙ transmitted in a fragmented ˙ and
originally part of the Kitāb al-Hurūf (now
33 ˙
disordered state); or perhaps Fārābı̄ originally intended it as part of the Hurūf,
34
but it grew out of proportion and he spun it off as a separate treatise. ˙
35
Fārābı̄, and Averroes to whatever extent he may have been aware of the
36
Kindian position, might reasonably regard Kindı̄ as giving the Platonist answer
37
38 sense of ‘delimited [munhāz] by some quiddity’, had become available. The picture that
˙ as applying Fārābı̄’s technical ontology to theology contrasts
emerges of the Perfect City
39
sharply with Mahdi’s picture (in Foundation of Islamic Political Philosophy) of the Perfect
40
City as an exoteric or ‘political’ work with a non-serious metaphysics, logically
41 disconnected from the esoteric ontology of the Kitāb al-Hurūf. I hope to develop this
42 picture of the Perfect City in other work. ˙
1 to Aristotle’s aporia at Metaphysics B4 1001a4-b25. Kindı̄ describes his aim, in

2 the major extant part of the On First Philosophy, as ‘to make clear the distinction
3 among ones, so that the true One may be manifest […] and [so that it may be
4 manifest] what are the ones by license [bi’l-maǧāz, i. e. improperly or
5 metaphorically one], I mean [one] by participation [ifāda]47 in the true One’
6 (97,21 – 99,1). So Kindı̄ wants to show that anything other than God is only
7 improperly one, and then that the cause of whatever unity it has is the true One.
8 If a thing can be multiplied or divided in any way, Kindı̄ will conclude that its
9 unity can be separated from it (since it can continue to exist while ceasing to be
10 one), that it has unity only as an accident and not by its essence, that it is one
11 only by convention (both 47,17 – 23), that it is one only by license (95,25 – 6).
12 He will then argue that any such thing has its unity by participating in
13 something that is one by its essence (53,6 – 15 and 95,20 – 25): this must be
14 entirely simple, outside all the categories and all the predicables, and ‘therefore it
15 is just pure unity, I mean nothing other than unity, and every one [kullu wāhidin
16
˙
= each thing] other than it is multiplied’ (95,13 – 14). Furthermore, since ‘there
17 is no being [huwiyya] except inasmuch as there is unity in it’ – because any
18 multiplicity must consist of unities – ‘and its being unified [tawahhud] is its
19 receiving being [tahawwı̄]’ (97,16), therefore ‘the emanation of unity ˙ ˙ from the
20 first true One is the receiving-being [tahawwı̄] of every sensible and of what
21 attaches to sensibles, and so every one of them exists [yūǧadu] when it receives
22 being through [the true One’s] giving it being’ (97,8 – 10). By contrast, Fārābı̄,
23 and Averroes following him, will determine senses in which each thing, even if it
24 is divisible, can be called ‘one’. These senses are not by license, and they are not
25 all through an accident, since ‘delimited by its essence’ applies to every mawǧūd.
26 Something may still be one ‘through another’, i. e. through some cause other
27 than itself, just as it may exist through some cause other than itself, and we may
28 be able to trace it back causally, step by step, to a first One, just as we may trace
29 it back causally to a first mawǧūd. But the thing will not be one by participating
30 in this first One, as if it were a formal cause of unity, rather the thing will be one
31 formally through its own essence, and dependent on something else as its
32 subject or its efficient cause; and the first One will not be a universal one
33 beyond the categories and participated in by things in all the categories, rather it
34 will be reached by inferring causally from accidents to substances and then from
35 substances to their first efficient cause.48
36
37 47 The last phrase might be more literally translated ‘by the benefaction [ifāda] of the true
38 One’, i. e. by the true One’s bestowing something (namely, unity) on it. But it seems to
39 be fairly fixed terminology (for instance, in the Liber de causis, rendering Proclus’
language of participation) that the mustafı̄d is the participant, the mustafād is what the
40
participant receives, and the mufı̄d is what it participates in.
41 48 However, Fārābı̄ thinks we can infer causally back from substance to a first cause which is
42 not in any of the categories and thus in particular is not a substance (see Kitāb al-Hurūf
˙
92 Stephen Menn
1 Avicenna of course does not think that things other than God are ‘one’ only
2 metaphorically, nor does he think that they are one by participating in God as a
3 formal cause of unity, any more than he things that they are mawǧūd by
4 participating in God as a formal cause of wuǧūd; when he abstracts away from a
5 thing’s unity he discovers, not a plurality or an original non-being, but an
6 essence neutral to individual existence in matter and universal existence in a
7 mind, which receives numerical or specific unity by coming to be in matter or
8 being thought by a mind. But he too thinks that unity is an accident superadded
9 to the essence of things, and Averroes might reasonably think that the
10 distinctions in unity that Fārābı̄ develops in order to avert Kindı̄’s errors will
11 serve to avert Avicenna’s errors as well.
12
13
14 Whence the Thesis that Quantitative Unity is Mind-Dependent?
15
16 The only main point of Averroes’ account of unity that does not seem to be in
17 Fārābı̄’s Kitāb al-Wāhid is the thesis that the one as a principle of number exists
18 only through the mind’s ˙ act of abstraction, so that unity in this sense, like being
19 as truth, is a second intention. Presumably Averroes was led to this by reflecting
20 on Avicenna’s thesis that number or multiplicity is an accident in the category of
21 quantity composed out of many unities, each of which is an accident in the
22 category of quantity belonging to an underlying subject in some category.
23 Avicenna, responding to an opponent who says ‘multiplicity is sometimes
24 composed out of things other than unities, such as men and beasts’, replies that
25 ‘as these things are not unities, but things which are subjects of unities, likewise
26 also they are not [collectively] a multiplicity, but things which are subjects of
27 multiplicity; and as these things are ones [but] not unities, likewise they are
28
[collectively] many [but] not a multiplicity’ (Šifā Metaphysics III,3,6 – 7).
˘
29
Averroes will distinguish multiplicity as a per se attribute of being, which is not
30
distinct from the essences of its subjects, from the accident numerical
31
multiplicity, but he will agree that if Socrates and Plato are numerically two,
32
their twoness is composed not out of Socrates and Plato but out of the unity of
33
Socrates and the unity of Plato, which must be in the same category, quantity, as
34
the twoness they compose; and this will lead Averroes to reflect on the status of
35
this quantitative unity, which cannot be the same as the unity which is a per se
36
attribute of being. But it may be that here too he found a clue in Fārābı̄.
37
38 I,92; translated and discussed in Menn, Analysis of the Senses of Being, pp. 93 – 6),
39 while Averroes thinks that God is a substance, and has a determinate essence, namely
intellect. (Averroes is quite unusual within the Muslim context in saying that God is a
40
substance: credal statements typically say that God is neither a body [ǧism] nor a
41 substance [ǧawhar], by which they mean in the first instance that he is neither a
42 compound of atoms nor himself an atom.)
1 Recall that Averroes says ‘the numerical one is the meaning of the
2 individual, abstracted from quantity and quality, I mean that by which the
3 individual is an individual, since it is an individual only by the meaning
4 “undivided”’ (Epitome III,41, cited above), and that ‘when the intellect abstracts
5 from these individuals this meaning, “indivisible into two or more individuals”,
6 this is the one which is the principle of number’ (I,36, cited above). This notion
7 of individuality or undividedness seems to be what Fārābı̄ means by the one as
8 ‘that whose essence is not shared [in such a way that] some two things would
9 resemble each other in respect of it’ (Kitāb al-Wāhid § 18) or ‘what has no
10 partner [qası̄m] in the meaning [= predicate] which ˙ is attributed to it –
11 whatever meaning it is – in [the sense] that this is a quiddity which belongs to it,
12 as if it were isolated [munfarid] by wuǧūd, and nothing other than it shares with
13 it the quiddity which belongs to it’ (§ 25); this sense of unity ‘accompanies
14 “delimited by some quiddity”, for it is delimited only by that in which it has no
15 partner’ (§ 26). This sense applies to immaterial substances, which are
16 necessarily unique in their species, but also to
17
every designated thing [mušār ilayhi = t|de ti] which is not in a subject [i.e.
18 individual substances], and every designated thing which is in a subject [i.e.
19 individual accidents as described in Categories 1a23 – 9], since none of these is
20 predicated of more than one thing. And ‘one’ is also said of what is not divided into
21 a subject lower than it [sc. as the individual is ‘lower’ than the species and the species
than the genus] but rather it is the lowest subject which underlies a predicate, and
22
the division of each more general predicate terminates with it and does not go
23 beyond it; and many people are wont to call it ‘one in number’, and it is called49
˘
24 individuals [a yān] and individuals [ašhās]. (§ 18)

25
˘ ˙
26
It seems that this covers both what is predicated of only one thing (or is
27
predicated of nothing, except tautologously of itself ), and what has some
28
predicate that is predicated only of it; and it seems that the more basic notion is
29
‘predicated of only one thing’, while ‘has some predicate that is predicated of
30
only one thing’ is derived from this. But ‘predicated of only one thing’ is a
31
second intention, since ‘predicated of ’ is a second intention, applicable only to
things in the mind (Kitāb al-Hurūf, p. 64,11 – 14). To say that something is one
32 ˙ is only one of it, and it is properly only for a
in this sense is to say that there
33
34
predicate-term F that I can say that there is only one F: if I say of something c
35
which is not a predicate that it is unique, that there is only one of it, this is just a
36
loose way of saying that for some appropriate predicate F (perhaps c’s species, if
37
c is an immaterial substance), c is F and there is only one F.50
38
39
49 Perhaps we should add a hā at the end of the verb, reading yusammı̄hi instead of
˘
40
yusammā; then not ‘it is called’ but ‘[Aristotle] calls it’.
41 50 So something can be ‘one’ in the sense of ‘unique’ in respect of one predicate and not in
42 respect of another: ‘the moon is isolated by the meaning of moonness [since] it has no
94 Stephen Menn
1 It seems likely that Averroes is reflecting on these passages of Fārābı̄, and

2 finding in them a response to what Avicenna says about the unities of which
3 accidents of number are composed; and thus the concept of individual unity
4 which Averroes takes to be the principle of number, i. e. the concept we use in
5 counting, would be this logical indivisibility rather than quantitative indivisi-
6 bility into spatially distinguished parts. I can legitimately count to three and say
7 ‘there are three people is the room’ (logically similar to Fārābı̄’s example, ‘there is
8 one dirham in my hand’, Kitāb al-Wāhid § 25), despite the fact that each of the
9
˙ and lower parts, because the predicate
people in the room is divisible into upper
10 ‘is a person in this room’ is truly predicated of three subjects which are not truly
11 predicated of any further subjects.51 Just as ‘existing outside the mind’, which
12 Avicenna takes to be an accident attaching in re to a neutral essence, is only a
13 second intention, saying of some predicate F ‘there is at least one F’, so
14 numerical unity, which Avicenna likewise takes to be an accident attaching in re
15 to a neutral essence, is only a second intention, saying of some predicate F ‘there
16 is exactly one F’; and presumably threeness is also a second intention, saying of
17 some predicate F ‘there are exactly three F’s’. And just as Averroes’ criticisms of
18 Avicenna on mawǧūd and wuǧūd are the starting point for Latin scholastic
19 discussions of essence and existence, so his criticisms of Avicenna on unity and
20 number will be the starting point for the scholastic discussions of numerical and
21 transcendental unity, and of categorematic and syncategorematic quantity. But
22 even without pursuing the Latin discussions, the Arabic history makes it clear
23 that the logical syntax of unity, like the logical syntax of being, has a long and
24 complicated history; in particular, the thesis that both unity and being have
25 important senses in which they are second-order predicates has a history going
26 back far beyond the nineteenth-century logicians who are usually said to have
27 invented it.52
28
29 partner of its species, although it has a partner in the meaning that it is a star, which is its
genus’ (Fārābı̄, Kitāb al-Wāhid wa’l-wahda § 25).
30
51 Or, perhaps, ‘is a person’, when˙ restricted
˙ to things in this room and to the present time,
31 is truly predicated of three such subjects; or ‘is in the room’, when restricted to persons
32 and to the present time. Fārābı̄ gives ‘there is one dirham in my hand’ as an example of
33 being isolated by some predicate ‘relatively and for a time’, following Aristotle’s
34 discussion in Topics V,1 of predicates which are propria of some subject – i. e. predicates
in respect of which that subject is unique – ‘relatively and for a time’.
35
52 Since the Epitome of the Metaphysics was not available in Latin until the Renaissance, it
36 will be important for the Latin reception that much of the Epitome’s argument about
37 unity can be found, if not as clearly developed, in the Great Commentary on Metaphysics
38 Iota 1 – 2 (some parallels have been noted above). But the Great Commentary, sticking
39 closer to Aristotle’s text, seems not to describe unity as ‘delimitation’ [inhiyāz], or to say
that mathematical unity arises from an act of the soul. In addition to the˙ parallels noted
40
above, note that at Tafsı̄r 1282,7 – 12, on Iota 4 1052a13 – 19, Averroes says that
41
˘
Avicenna wrongly thought that if ‘one’ and ‘mawǧūd’ and ‘thing’ signified the same ma nā
42 then e. g. ‘the man is one’ would be a tautology or negation like ‘the man is [a] man’.
1 Bibliography
2
3 Averroes, Compendio de Metafisica, ed. C. Quirs Rodrguez, Madrid: E. Maestre, 1919.
4 –– , Commentarium magnum in Aristotelis de anima libros, ed. F.S. Crawford,
5
Cambridge, Massachusetts: The Mediaeval Academy of America, 1953.
–– , On Aristotle’s ‘Metaphysics’: An Annotated Translation of the So-called ‘Epitome’, ed.
6
R. Arnzen, Berlin: De Gruyter, 2010.
7
˘
–– , Tafsı̄r mā ba d al-tabı̄ at, ed. M. Bouyges, 4 vols, Beirut: Imprimerie Catholique,
8 ˙
1938 – 48 [‘Tafsı̄r’].
9 –– , Tahafot at-Tahafot, ed. M. Bouyges, Beirut: Imprimerie Catholique, 1930 [‘TT’].
˘
10 Avicenna, Kitāb al-Naǧāt fı̄ ’l-hikmat al-mantiqiyya wa’l-tabı̄ iyya wa’l-ilāhiyya, ed.
M. Fakhry, Beirut: Dar al-Afaq ˙ al-Jadidah, ˙1985. ˙
11
–– , The Metaphysics of ‘The Healing’, transl. and ed. M. Marmura, Provo, Utah:
12
Brigham Young University Press, 2005.
13 A. Bertolacci, On the Arabic Translations of Aristotle’s Metaphysics, Arabic Sciences and
14 Philosophy, 15, 2005, pp. 241 – 75.
15 H.A. Davidson, Alfarabi, Avicenna and Averroes on Intellect, Oxford: Oxford University
16 Press, 1992.
17
18 Compare also the Great Commentary on Metaphysics C2: ‘and Avicenna erred greatly in
19 this: he thought that “one” and “mawǧūd” signified attributes superadded to the essence
of the thing. The astonishing thing about this man is how he committed this error, given
20
that he listened to the Ash arite mutakallimūn who mix his metaphysics with their kalām,
˘
21
˘
and they say that some attributes [hold] through an [inhering] entity [are ma nawı̄
22 attributes] and others [hold through] the thing itself [are nafsı̄ attributes], and they say
23 that “one” and “mawǧūd” refer to the essence of the thing to which they are attributed,
24 and they are not attributes signifying something superadded to the essence, as is the case
with “white” and “black” and “knowing” and “living”. But this man was compelled to his
25
˘
opinion by his saying that if “one” and “mawǧūd” signified the same ma nā, then our
26 saying “the mawǧūd is one” would be nugation, like our saying “the mawǧūd is mawǧūd”
27 or “the one is one”. But this would follow only if it were said that our saying about a
28 single thing that it is mawǧūd and that it is one signify a single thing in a single respect
29 and in a single manner; but we say only that they signify a single essence in different
manners, not different attributes superadded to it. Thus according to this man there is no
30
difference between significations which signify about a single essence in different
31
˘
manners without signifying ma ānı̄ superadded to it, and significations which signify
32 about a single essence attributes superadded to it, i. e. different from it in actuality. And
33 this man was led astray by several things: one was that he found that the noun “one”
34 [wāhid] was a paronymous noun and these nouns signify an accident and a substance
[i.e.˙they signify some accident present in some substance, as “white” signifies whiteness
35
in a substantial substratum]; another was that he thought that the noun “one” signifies
36
˘
an indivisible ma nā in the thing, and that this ma nā is not the ma nā which is [the
37 thing’s] nature [since, if it’s a material thing, this will be divisible]; another was that he
38 thought that this one which is predicated of all categories was the one which is the
39 principle of number, and number is an accident, and so he concluded that the noun
“one” signifies about the mawǧūdāt an accident. The one which is the principle of
40
numbers is only [one] among the beings of which the noun “one” is said, even if it is the
41 most deserving of this [name], as you will learn in the ninth book of this treatise [= Iota,
42 not Theta]’ (313,6 – 314,11).
96 Stephen Menn
1 –– , Proofs for Eternity, Creation and the Existence of God in Medieval Islamic and Jewish
2 Philosophy, Oxford: Oxford University Press, 1987.
3
Master Eckhart, Parisian Questions and Prologues, transl. A. Maurer, Toronto: Pontifical
Institute of Mediaeval Studies, 1974.
4 Abū Nasr al-Fārābı̄, Commentary and Short Treatise on Aristotle’s ‘De Interpretatione’, ed.
5 F.W.˙ Zimmermann, Oxford: Oxford University Press, 1981.
6 –– , Kitāb al-Hurūf = Alfarabi’s Book of Letters, ed. M. Mahdi, Beirut: Dar El-Mashreq,
7 1969 – 70.˙
8
–– , Kitāb al-Wāhid wa’l-wahda = Alfarabi’s On One and Unity, ed. M. Mahdi,
˙
Casablanca: Editions ˙
Toubkal, 1989.
9 –– , On the Perfect State (Mabādi ārā ahl al-madı̄na al-fādila), transl. and ed. R. Walzer,
˘
˘
10 Oxford: Clarendon Press, 1985. ˙
11 al-Kindı̄, Oeuvres Philosophiques et Scientifiques, vol. 2, Mtaphysique et Cosmologie, eds.
12 R. Rashed and J. Jolivet, Leiden: Brill, 1998.
13
M. Mahdi, Alfarabi and the Foundation of Islamic Political Philosophy, Chicago:
University of Chicago Press, 2001.
14 S. Menn, Fārābı̄’s Kitāb al-Hurūf and his Analysis of the Senses of Being, Arabic Sciences
15 and Philosophy, 18, 2008, ˙ pp. 59 – 97.
16 –– , Review of Anthony Kenny, Aquinas on Being, Philosophical Review, 115, 2006,
17 pp. 391 – 5.
Abd al-Ġanı̄ al-Nābulusı̄, Īdāh al-maqsūd min wahdat al-wuǧūd, Damascus: Matba a
˘
˘
18 ˙ ˙ ˙ ˙ ˙
˘
al- Alam, 1969.

19 J. Puig Montada, Cuanto se encuentra ms all de la naturaleza, introduction to the
20 reprint edition of Averroes, Compendio de metafsica, ed. C. Quirs, Sevilla:
21 Fundacin El Monte, 1998, pp. IX-XXXIV.
22 A.I. Sabra, Avicenna on the Subject Matter of Logic, Journal of Philosophy, 77, 1980,
23
pp. 746 – 64.
R. Wisnovsky, Avicenna’s Metaphysics in Context, London: Duckworth, 2003.
24 –– , Notes on Avicenna’s Concept of Thingness (šay iyya), Arabic Sciences and Philosophy,
˘
25 10, 2000, pp. 181 – 221.

26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
1
2
3
4 Avicenna and his Commentators
5 on Human and Divine Self-Intellection*
6
7 Peter Adamson
8
9
It is a notorious feature of Avicenna’s metaphysics that he begins from the
10
notion of existence, rather than progressing smoothly from natural philosophy
11
to divine science. Avicenna was duly criticised by Averroes for this departure
12
from the Aristotelian tradition.1 And certainly it is true that Avicenna proves the
13
existence of God without premises drawn from natural philosophy, in his
14
famous demonstration of the Necessary Existent. This strategy in turn bears on
15
Avicenna’s view of the subject of metaphysics, since it is well-established
16
Aristotelian doctrine that no science proves the existence of its own subject-
17
matter. These familiar points are in danger of obscuring a noteworthy aspect of
18
Avicenna’s metaphysics which is rather in tension with the idea that metaphysics
19
is autonomous from physics. When Avicenna is concerned to say not merely
20
that God exists, but what He is like, he does in fact make use of premises
21
established in natural philosophy. In particular, his discussions of God in both
22
the Healing (al-Šifā ) and Pointers and Reminders (al-Išārāt wa-l-tanbı̄hāt)
˘
23
depend on points established in the psychological sections of those works
24
(psychology, of course, being a part of natural philosophy both in the previous
25
tradition and in the rubrics of the various sections of the Healing itself ).
26
The dependence of Avicenna’s metaphysical theology on his psychology did
27
not escape the notice of his two most famous commentators, Fahr al-Dı̄n al-
28
Rāzı̄ (d. 606/1210) and Nası̄r al-Dı̄n al-Tūsı̄ (d. 672/1274). Both˘of them say
29
more or less the same thing˙about a crucial˙ tanbı̄h in the Pointers,2 which reads
30
as follows:
31
32
33
34
35
* This paper has benefited greatly (though, no doubt, not enough) from the input of
36 numerous colleagues. My thanks to audiences at the Villa Vigoni and in Freiburg in July
37 2008, and to those who kindly read and commented on earlier versions: Amos Berto-
38 lacci, Dag Nikolaus Hasse, Heidrun Eichner, MM McCabe, and David C. Reisman.
39 1 See Bertolacci, Avicenna and Averroes.
2 Avicenna, al-Išārāt, cited as Pointers followed by section number, with volume and page
40
number from Dunyā’s edition in brackets. All citations are to part two of the Pointers, so
41 I will not bother to indicate this. There is no English translation of the relevant parts of
42 the Pointers, but for a French translation see Avicenna, Livre des directives et remarques.
98 Peter Adamson
˘
1 Pointers § 4.28 [III.53]: The First is self-intelligible (ma qūl al-dāt) of its
subsistence.3 For it subsists free from attachments, lacks ( uhad), ¯ materials
˘
2
3
(mawādd), or anything else that would bestow anything additional that would
inhere (hāll zā ida) in the essence. And you know that whatever is like this intellects
˘
4 ˙ is intellected by itself.
itself and
5
6 Now, neither of these two commentators is notable for his concision in dealing
7 with the text of the Pointers; al-Rāzı̄’s commentary is especially voluminous.4 So
8 it is all the more striking how little either of the two commentators has to say
9 about the passage just cited. They simply refer us back to the third section or
10 namat of part two of the Pointers: this is the namat which deals with
˙
psychology. 5 ˙
At this critical juncture, where we are establishing the intellective
11
12 nature of God, the commentators follow Avicenna in allowing the argument to
13 rest on claims proven in the study of psychology.
14 The psychological premise in question is made quite explicit here: anything
15 immaterial must be an intellect, and moreover an intellect that intellects itself.
16 Given the immateriality of God, it follows immediately that God is a self-
17 intellecting intellect. Although the idea of immaterial, self-thinking intellect is a
18 familiar theme from the Neoplatonic tradition, the origins of the crucial
19 premise are already to be found in Aristotle. Of course Metaphysics book
20 Lambda is the source for Avicenna’s idea that God is a self-thinking intellect.
21 But the inference from immateriality to intellection and self-intellection is
22 drawn from De anima III.4. There, Aristotle not only establishes a link between
23 intellection and immateriality, but also discusses the fact that an intellect can be
24 the object of its own thought. So in what follows I will first look briefly at this
25 Aristotelian text. But I will spend most of the paper discussing how Aristotle’s
26 ideas were used and expanded upon by Avicenna and his commentators. In the
27 first section I will discuss the connection between intellection and immateriality,
28 in the second section, the idea that every self-subsisting intelligible is an intellect
29 that can engage in intellection of itself. In the third section, I deal with the claim
30 that God is just such a self-thinking intellect.
31
32
33
34
35
˘
3 The Dunyā edition has al-Awwal ma qūl al-dāt qā imihā, which Goichon translates ‘Le
˘
36 ¯
Premier a une essence intelligible, il la fait subsister.’ One might instead read qā im bi-hā;
˘
37 al-Tūsı̄ has something similar, since his commentary says instead qā im bi-nafsihi. In that
˘
38 case˙ we could translate more simply ‘the First is self-intellecting and self-subsistent’. My
39 thanks to David Reisman for discussion of this passage.
4 For al-Rāzı̄’s commentary I have used the edition in Šarhay al-Išārāt. For al-Tūsı̄’s
40 ˙ cited respectively ˙as al-
commentary see Dunyā’s edition in Avicenna, al-Išārāt. Hereafter
41 Rāzı̄, Commentary and al-Tūsı̄, Commentary.
42 5 al-Rāzı̄, Commentary I.214˙ lines1 – 2; al-Tūsı̄, Commentary III.54 line 10.
˙
1 Immateriality and Intellection

2
3 Turning then to De anima III.4,6 we find Aristotle making the observation,
4 often repeated in the Arabic tradition, that what is extremely intelligible does
5 not incapacitate intellect, the way a bright light or strong scent incapacitates the
6 faculty of sight or smell.7 The reason given for this is that ‘sensation is not
7 without body, whereas intellection is separate from body’ (429b5, p. 73 lines
8 8 – 9 in the Arabic version). In the next sentence, he explains that a man who has
9 learned an intelligible has that intelligible potentially, but in a higher
10 potentiality than the one prior to learning it, ‘and at that point it can
11 intellectually grasp itself ’ (429b9, p. 73 line 12 Arabic).8 The relationship
12 between self-thinking and immateriality is not yet clear, but it is at least
13 suggestive that the two points follow in such close succession. And later in the
14 same chapter Aristotle does make an explicit link between thinking, self-
15 thinking, and immateriality. This is occasioned by the question whether
16 ‘thought is also intelligible’ (429b26, p. 74 line 7 Arabic). His answer is as
17 follows, translating from the Arabic version:
18
De anima III.4, 430a2 – 9 (p. 75 lines 13 – 18 Arabic): It [sc. intellect] is also
19 intelligible,9 just like the other intelligible things. And regarding those things that
˘
20 are immaterial, the intellecting and the intelligible (al- āqil wa-l-ma qūl) are one and
21 the same thing … But I claim that the intelligible in what is material is only
22 potentially intelligible. Therefore material things have no intellect, because intellect
˘
in respect of potentiality is not in matter (al- aql min ğihat al-quwwa laysa fı̄ hayūlā).
23
But the intelligible belongs to the intellect, being ascribed (mansūb) to it.
24
25 This passage sets up a clear correlation, if not mutual implication, between
26 immateriality, thinking, and self-thinking. Because it is immaterial, the intellect
27
28 6 I have been greatly helped in understanding this chapter by McCabe, Some
29 Conversations.
7 De anima III.4, 429a31 – 429b4, and for the Arabic version, Aristūtālis fı̄ l-nafs, ed.
30
Badawı̄, p. 73. This translation is ascribed to Ishāq b. Hunayn but ˙ ˙the attribution is
31 controversial. See Elamrani-Jamal, De anima. ˙ ˙
8 The Arabic has ‘wa-yumkinuhu fı̄ dālika l-waqt an ya qila nafsahu’ and thus clearly
˘
32
33 accepts the reading ja· aqt¹r d³ art¹m¯ t|te d}matai moe?m in the Greek manuscripts. Ross
34 follows Bywater in emending d³ art¹m to d¸ artoO, which would yield ‘and at that point
it can think through itself.’
35
˘
9 A word here about the term ma qūl, here translated ‘intelligible’: like other Arabic words
36 of this form it is ambiguous between factive and potential readings. In other words a
˘
37 ma qūl might be either that which is in fact being intellected or that which is not being
38 intellected, but could be intellected. I translate the term as ‘intelligible’ throughout, but
39 more often than not Avicenna will mean by it something that is indeed occurently an
object of intellection, and the reader should bear this in mind. (For instance, in the
40
passage already translated above, Pointers § 4.28, Avicenna of course means that the First
41 actually intellects itself, not merely that it can do so.) My thanks to Gad Freudenthal for
42 discussion of this point.
100 Peter Adamson
1 can both think and be thought. Aristotle makes it especially clear that it is
2 materiality that prevents thought (and self-thinking), because he says that some
3 intelligibles do not think precisely because they are in matter. What he has in
4 mind here is that a form in a stone, for instance, may be intelligible – in
5 Avicenna’s terms, the stone could provide me with a basis for abstracting the
6 intelligible form granite – but the stone does not itself think, or think about
7 itself, because it is a material object.
8 Avicenna raises a hypothetical objection in the Pointers that builds on this
9 puzzle:
10
Pointers § 3.20 [II.422]: Perhaps you will say that when the form that subsists
11 materially (al-sūra al-māddiyya fı̄ l-qawām) is abstracted in the intellect, then the
12 ˙ prevents [it from itself engaging in intellection] is removed. So
characteristic that
13 what stops us from ascribing intellection to it?
14
As we just saw, in De anima III.4 Aristotle was worried about two kinds of
15
intelligibles: the intelligible object outside the intellect, and the intellect itself.
16
The former does not think, because it has matter,10 the latter does think, because
17
it has no matter – and for the same reason it is an object of self-thinking. The
18
objection then is that there is a third kind of intelligible: the abstracted
19
intelligible that is in the intellect. If it is no longer enmattered, why will it not
20
think? For example, if I abstract the intelligible form granite from a stone, then
21
should the form granite in my intellect not be able to engage in intellection in its
22
own right? Avicenna responds as follows:
23
24 Pointers § 3.20 [II.422 – 4] The answer is: because it is not independently subsisting
˘
(mustaqilla bi-qawāmihā), nor receptive of intelligible concepts (ma ānı̄) that inhere
25
in it. The representations (amtāl) of [the forms that subsist materially] are simply
26 ¯
conjoined (yuqārinuhā) with intelligible concepts, in which they are not inscribed –
27 rather [they are inscribed] in that which receives both of them. Neither is more
28 fitting to be inscribed in the other, instead of the other [being inscribed] in it. The
29 two are not put together the way that form and what has form are conjoined.
30 Here Avicenna makes a concession to the opponent: suppose that I grasp the
31 form granite and also the form iron. 11 In that case the form granite will be
32 ‘conjoined’ with the form iron, because they will both be objects of the same
33 intellect. That is, they will in some sense be compresent with one another, since
34 they are together in the same subject (so in this context the word ‘conjoin’ just
35
36 10 If the external object is itself immaterial, it too will be an intellect. We will see Avicenna’s
37 proof for this in the next section of this paper.
38 11 For the sake of simplicity, I speak rather loosely of ‘forms’ even though Avicenna is using
39 more varied terminology: he speaks of intelligible ‘concepts’ and also of ‘representations’.
The former seems, in this passage, to refer specifically to the form when it is in the
40
˘
intellect, hence my translation of the notoriously difficult ma ānı̄. Meanwhile the use of
41 amtāl (‘representations’) seems to refer to the form in the mind insofar as it is related to
42 the¯ external form in matter.
1 indicates that the two things are somehow together). But Avicenna insists that
2 the form granite will not itself grasp the form iron, because it is not a self-
3 subsisting thing. Their mere compresence is not enough to give either of them
4 an intellectual grasp of the other.
5 This is a question-begging response, insofar as Avicenna is merely
6 stipulating that self-subsistence is a necessary condition for engaging in
7 intellection. But he does add a further argument, which is that granite and iron
8 would have an equal claim to be the subject of intellection. There is no
9 principled reason why granite should think about iron instead of the other way
10 around. But they cannot grasp each other, because intellectual grasping is
11 assumed to be asymmetrical. The sort of asymmetry involved here, as Avicenna
12 says, is the asymmetry between ‘form and what has form’. Al-Rāzı̄’s criticism of
13 this tanbı̄h 12 focuses on this aspect of Avicenna’s argument. He does not
14 question the assumption that the intellection relation must be asymmetrical. But
15 he then points out that there could be such an asymmetry, not only between
16 intellect and intelligible, but also between two abstracted forms in my intellect.
17 If the two forms are distinct, then could one form be a fit subject (mahall) for
18 the other, but not vice-versa? For example, it is possible for motion to ˙be slow,
19 but not for slowness to move. So it seems the form slowness in my intellect could
20 be asymmetrically ‘conjoined’ with the form motion in my intellect too. In
21 which case the form motion could intellectually grasp the form slowness.
22 The objection illustrates a feature of Avicenna’s account that al-Rāzı̄
23 frequently exploits in his criticisms, which is that the ‘conjunction (muqārana)’
24 between intellect and intelligible is an instance of the relation between mahall
25
˙
and hāll. This pair of terms is difficult to translate with two forms of the same
26
˙
English word, but the basic idea is that the hāll is something that ‘inheres’ or is
27 present in the mahall. The primary example ˙of the inherence of hāll in mahall is
28 the ontological case˙ of form inhering in matter. But the terminology˙ can be˙ used
29 to cover other cases; for instance, as we just saw, al-Rāzı̄ speaks of motion as the
30 mahall of slowness. As we have just seen, the inherence relation is always
31
˙
asymmetrical. This explains why both Avicenna and al-Rāzı̄ are happy to assume
32 that two things cannot inhere in one another, as then they would relate
33 symmetrically.13 On the other hand, it is not yet clear what the intellect-
34 intelligible relation amounts to, beyond its being asymmetrical. This is the point
35 exploited by al-Rāzı̄: insofar as an intelligible that I grasp can be a subject of
36
37 12 Al-Rāzı̄’s discussion runs from I.174 line 4 to the end of the page.
38 13 Thus, as al-Rāzı̄ makes explicit in his commentary on this tanbı̄h, the only way an
39 intelligible quiddity grasped by my intellect can be the subject of intellection is if that
quiddity also exists outside of me as a self-subsisting intellect in its own right. If I grasp
40
the quiddity of a celestial intellect, the quiddity of that intellect qua object of my intellect
41 would not be able to engage in intellection, even though the same quiddity engages in
42 intellection outside of me.
102 Peter Adamson
1 inherence for some hāll or other (e. g. motion is the subject of slowness when I
2 predicate the latter of˙ the former), it is hard to see why it could not be a subject
3 of intellection too. What we need is a principled distinction between
4 intellection and other sorts of inherence.
5 We have already seen Avicenna’s answer to this demand, but not his entire
6 rationale for that answer. The answer is that intellection occurs when an
7 intelligible object inheres in a subject that is self-subsisting and immaterial. We
8 can gain some insight into why the subject of intellection must be self-
9 subsistent, if we consider the following. Suppose that two items, A and B, are
10 conjoined such that A is thinking about B. In this case B subsists in A, by being
11 its object of intellection. As we have seen, this relation is asymmetrical, so we
12 can rule out that A conversely subsists in B. It is the mahall, not the hāll. But
13 could A not subsist in some further subject, C? This would˙ happen if the ˙ subject
14 of intellection (item A) depended on some material substrate (item C). But that
15 is ruled out by the second part of Avicenna’s answer, namely his insistence that
16 the subject of intellection is immaterial. We can easily point to Avicenna’s source
17 for the claim. He is following Aristotle’s assertion in De anima III.4 that it is
18 matter which prevents a form from engaging in intellection, of both other
19 things and itself. (Notice the implication that self-subsistence is necessary, but
20 not sufficient, for being the subject of intellection. After all, physical objects are
21 self-subsistent, and so some things can inhere in them: for instance colour might
22 inhere in a piece of paper. But these things are material, and so cannot be
23 subjects for the special sort of inherence that is intellection.)
24 Of course we will want an argument to show that materiality precludes
25 intellectivity. For such an argument we can look earlier in the same chapter of
26 the De anima (III.4, 429a18 – 27), where Aristotle reasons that because intellect
27 potentially thinks every form, it cannot actually have any given form before it
28 thinks. Hence, it cannot be ‘mixed’ with body, since every body actually has
29 determinate form. Whatever we make of this argument, Avicenna seems to have
30 thought it could be improved upon. He does not make use of it in either the
31 Healing or the Pointers. Instead, he argues in both texts that the intelligible form
32 is ‘undivided’, and thus cannot be received in body, which is ‘divided’ in the
33 sense of actually having discrete parts. (If this is inspired by Aristotle, then the
34 source text is not De anima III.4, but rather III.6, where Aristotle discusses the
35 question of how indivisible objects are grasped by mind.) Here is the version of
36 the argument in the Pointers:
37
Pointers § 3.16 [II.404 – 8]: If you want now to have it shown to you that the
38
intelligible item (ma nā) is not inscribed in something divided (munqasim)14 or in
˘
39 something that has spatial position, then listen. You know that something undivided
40 may be conjoined with a multiplicity of things without thereby having to become
41
˘
42 14 Or ‘divisible’; cf. my remarks on ma qūl, n. 9 above.

˘
1 spatially divided (munqasima fı̄ l-wad ), so long as their multiplicity is one that is
˙ a multicoloured object.15 But something that
not spatially divided, as are the parts of
2
3 is divided according to spatial differentiation cannot be conjoined with something
undivided. Inevitably there are undivided concepts in the intelligibles – unless the
4
intelligibles are put together from actually infinite principles. But even then, in
5 every multiplicity, whether finite or infinite, there must be some actual unity. So in
6 the intelligibles there is something actually one, which is grasped intellectually
7 insofar as it is one. Thus it will be grasped intellectually precisely insofar as it is
8 undivided. Therefore it will not be inscribed in what is spatially divided. But every
9 body, and every power (quwwa) in a body, is divided.
10 In keeping with the design of the Pointers, this passage is rather compressed, but
11 the thrust of the argument is clear. An intelligible form cannot be grasped by
12 dividing it up and sharing it out part-by-part in the parts of a subject. Rather,
13 the subject of intellection as a whole needs to grasp the form as a whole.
14 Al-Rāzı̄’s commentary (I.163 – 4) begins by supplementing Avicenna’s
15 argument with a further proof (I.163 lines 8 – 27), which is reminiscent of
16
the Sail passage from the first part of Plato’s Parmenides. 16 The argument goes
17
exhaustively through the different ways an undivided thing can be in a divided
18
thing. It might be wholly present in each part. Or different parts of it might be
19
in different parts of the subject. But in the first case it will appear in the subject
20
an infinite number of times, since the subject is infinitely divisible.
21
Furthermore, what we want is for it to be received by the whole subject, not
22
just each part of the subject. In the second case it will be received as divided, not
23
as undivided – but the goal was to explore how an undivided subject is received.
24
Now al-Rāzı̄ shifts to a critical stance. He says that Avicenna’s claim seems to be
25
26
disproven by the undivided point, which is present in a divided body, and goes
27
on to discuss this possibility at length (I.163 line 27 to end of page). It is
28
interesting to note that the example of the point (for which see also De anima
29
III.6, 430b20), and also the idea of an intelligible being present in parts of a
30 divided subject, can be found in the Psychology section from Avicenna’s Healing.
31 In a chapter devoted to proving that ‘the rational soul does not subsist by being
inscribed (muntabi ) upon bodily material’ (V.2),17 Avicenna has a long
˘
32 ˙
33
34 15 For bulqa as a ‘multicoloured object’ see Lane, Lexicon, vol. I, p. 253. The term is used
for instance of a black-and-white horse.
35
16 Parmenides 131c-e. If there is a historical connection to the Parmenides it would most
36 likely be via the commentary tradition on Aristotle. As early as al-Kindı̄, the Arabic
37 tradition takes up this question of how a form or species is present – by being divided
38 into parts, or as a whole – in its instances. The passage in al-Kindı̄ is in a text related not
39 to any Platonic work, but the Categories. See Adamson and Pormann, Aristotle’s
Categories and the Soul.
40
17 De anima, ed. Rahman, pp. 209 ff. French translation in Psychologie, ed. Bakoš,
41 pp. 148 ff. Cf. the translation at McGinnis and Reisman, Classical Arabic Philosophy,
42 pp. 188 – 92.
104 Peter Adamson
1 discussion of the point as the only indivisible thing that inheres in a body, where
2 he argues that the relation between point and magnitude is not the same as the
3 relation between intelligible and intellect.18 He then goes on to consider whether
4 the intelligible could be present in parts of a body, and introduces a dichotomy
5 like that of al-Rāzı̄: the intelligible might be present by having either similar or
6 dissimilar parts present in the different parts of the body.19 So it seems likely
7 that al-Rāzı̄ consulted the Healing when writing this part of his commentary on
8 the Pointers. However, al-Rāzı̄ is less sanguine than Avicenna about the prospects
9 of showing a disanalogy between the point and the intelligible. He also
10 mentions other undivided things that inhere in bodies. For instance unity,
11 relation and existence all inhere in bodies, but it makes no sense to divide
12 something like existence in half (Commentary, I.163 – 4). If all these things
13 inhere in bodies, what rules out that the intellect be a body, yet still have an
14 intelligible object inhering in it?
15 Since al-Rāzı̄’s arguments have again turned on analogies between
16 intellection and other sorts of inherence, it is unsurprising that al-Tūsı̄’s more
17 ˙
sympathetic commentary strives to distinguish intellection from these other
18 sorts. Again, any such relation is an asymmetrical one between a hāll and a
19 ˙
mahall. Both the subject of inherence and that which inheres can be ‘divided’ in
20 ˙
a number of ways, and many of these ways are consistent with incorporeality.
21 For example black might be divided into genus and specific difference; but this
22 division on the side of the hāll implies no division on the part of the mahall. 20
23
The reverse is also true: the ˙ fact that body has a genus and a species˙ (e. g.
24
extension and three-dimensional) shows nothing about whether there is some
25
division of what inheres in body. Al-Tūsı̄ claims that it is only spatial division of
26 ˙
a mahall which necessitates the division of the hāll, and vice-versa.
27 ˙ ˙
In fact, it is not even the case that everything that inheres in a spatially
28
divided subject will itself be divided. In such a case ‘what inheres in [a subject]
29
does not do so qua this subject (min hayt huwa dālika l-mahall), but insofar as a
¯ ¯
different nature belongs to it (min h˙ayt luhūq tabı̄ a uhrā)’˙ (II.406 lines 3 – 4).
30
˘
31 ˙ ¯ ˙ ˙
This accounts for the cases al-Rāzı̄ mentioned, such as˘ the point, relation, and
32
unity: ‘for instance the line: the point is not divided along with [the line],
33
because [the point] does not inhere [in the line] insofar as [the line] is a line, but
34
insofar as [the line] is limited’ (II.406 lines 5 – 6). Or, to use another of his
35
examples, a figure does not inhere in a plane insofar as the plane is a plane, but
36
insofar as the plain is bounded. Analogously, an intelligible form does not inhere
37
38
in a body insofar as the body is divisible, but insofar as the body instantiates that
39
18 The argument runs from p. 210 line 6 to p. 211 line 14.
40
19 The argument runs from p. 211 line 15 to p. 214 line 3.
41 20 He later points out (I.408, line 1) that the soul itself has genus and specific difference,
42 but of course we do not want to say that the soul is divided.
1 form. Therefore, the same form can inhere in the intellect without the intellect’s
2 being divided:
3
Commentary II.407, lines 9 – 10: Saying that it is ‘grasped intellectually’ means that
4 it is inscribed in a substance that apprehends it. This inscribing (irtisām) in that
˘
5 substance is not insofar as some other nature (tabı̄ a) belongs to [the thing it grasps],
6
˙ (bi-dātihi).
because it apprehends [that thing] in its essence
¯
7
Here the ‘other nature’ is the divisibility which has accidentally accrued to
8
granite when granite is bodily. Since intellection is a grasp of granite in its
9
essence, rather than with this accidental nature, the grasping can occur in an
10
immaterial subject – in fact, as we have seen, it must do so, if the form is to be
11
grasped as a whole by the whole subject.21 The upshot is that in Avicenna’s
12
conclusion, ‘[the intelligible] will not be inscribed in what is spatially divided’,
13
al-Tūsı̄ thinks the emphasis should be placed on the word ‘spatially’ (II.408 lines
14 ˙
18 – 19).
15
16
17 Self-Intellection
18
19 So far, we have seen Avicenna argue that a subject of intellection must be self-
20 subsisting and immaterial. His success in doing so has been called into question
21 by al-Rāzı̄, who has argued that Avicenna fails to distinguish intellection from
22 other cases of inherence. But even if, with al-Tūsı̄, we agree that Avicenna has
˙
23 argued successfully, this would not yet be sufficient. As we saw, later in the
24 Pointers Avicenna will infer that God is a self-thinking intellect, on the basis that
25 God is immaterial. For this inference to go through, the premise that every
26 intellect is immaterial will not do. We need the converse claim, that every self-
27 subsisting immaterial thing is an intellect and self-intellective to boot. Avicenna
28 is well aware of this, and asserts explicitly that he has provided the necessary
29 arguments by the time we get to Pointers § 3.22 (II.429 – 30). There he says, ‘if
30 you take in what I have established for you, you will know that (1) whatever is
31 such as (min ša nihi) to become an intelligible form and is self-subsistent is also
˘
32 such as to intellect; and from this it follows that (2) it is such as to intellect
33 itself.’22 This raises the question of where in the third namat Avicenna has
34 established the two claims he mentions here: ˙
35
36 21 An interesting intermediate case is the estimative faculty. In estimation (wahm) the
˘
37 intelligible item (ma nā) is received cognitively, but in a physical object, namely the
38 brain. As al-Tūsı̄ points out (II.408, line 9 – 10), this sort of reception of an intelligible
˙ with in the immediately following tanbı̄h, namely Pointers § 3.17.
concept is dealt
39
˘
22 I hereby apologize for the use of the word ‘intellect’ as a verb to translate forms of aqala,
40
rather than ‘intellectually grasp’ as used in the previous section. ‘Intellectually grasp’ is
41 better English but would lead to considerable awkwardness in translating some of the
42 passages to be considered in this section.
106 Peter Adamson
1 (1) Every self-subsistent intelligible object is also a subject of intellection

2 (that is, it is able to grasp intelligible objects).
3 (2) Every intellective subject can intellect itself.
4 For claim (2), one might immediately think of the Flying Man thought
5 experiment, which is given a very prominent position in both the Healing and
6 the Pointers. Avicenna uses it to begin the psychology section of the Healing and
7 the third namat on soul in the Pointers. The thought experiment is meant to
8 ˙
show that someone innocent of sensory experience will nonetheless have
9 certainty regarding his own being (tubūt anniyyatihā).23 Obviously this is not
10 ¯
irrelevant to the topic we are considering, but I will pass over the Flying Man
11 and its treatment by al-Rāzı̄ and al-Tūsı̄. For one thing, this section of the
12 ˙
Pointers and the two commentaries on it have been discussed in the secondary
13 literature, albeit only in one article by Michael Marmura.24 More importantly, it
14
is not clear that the Flying Man example is really about self-intellection as
15
opposed to something like self-awareness.25 In neither the Healing nor the
16
˘
Pointers version does Avicenna say that the flying man would ‘intellect (ya qilu)’
17
anything.26
18
So I would like instead to turn to another tanbı̄h from the Pointers, which
19
explicitly sets out to establish both of the claims just mentioned.27 It reads as
20
follows.
21
22
23 23 Pointers, § 3.1 [II.344 – 5].
24 24 See Marmura, Fakhr al-Din ar-Razi’s Critique. My thanks to Dag Nikolaus Hasse for the
reference.
25
25 For self-awareness in Avicenna see Kaukua, Avicenna on Subjectivity. An interesting
26 question here is what faculty is responsible for self-awareness, if self-awareness is not self-
27 intellection. For Avicenna self-awareness seems to be so basic to human cognition that it
28 underlies even sensation and imagination. Thus one might wonder whether it could
29 really be an act of the intellect. (My thanks to Amos Bertolacci for raising this point.) See
further Black, Avicenna on Self-Awareness.
30
26 In the version in the Healing, De anima I.1, he uses terminology similar to that found in
31 the Pointers: the flying man will have itbāt li-dātihi mawğūda (p. 16, line 7). The verb
32 used repeatedly in the Healing version is¯ ‘affirm¯ (tabata)’. The closest Avicenna comes to
speaking of intellection is when he stipulates, in the¯ Pointers, that the flying man must be
33
˘
34 ‘of sound mind (sahı̄ha l- aql)’.

27 The Pointers is much˙ ˙ ˙ more deliberate than the Healing in setting out an account of
35
human self-intellection as a basis for the theory of divine self-intellection. Having said
36 that, there is one intriguing passage in § V.6 of the Psychology in the Healing (239.3 – 10):
37 ‘We say that the soul intellects that it takes into itself the form of the intelligibles,
38 separated from matter. The form is separated either by the intellect’s abstracting it or
39 because that form is in itself separate from matter, in which case the soul has no need to
abstract it. The soul conceives (tatasawwara) itself, and its conceiving itself makes it
40
intellect, intellecting and intellected. ˙But its conceiving these forms does not make it like
41 this, because in its substance in the body, it is always potential intellect, even if it
42 becomes actual with respect to some things.’
1 Pointers § 3.19 [II.415 – 20]: You know that whatever intellects something intellects
2 – through a potency close to actuality – that it is intellecting28. In so doing (min-hu)
3
it intellects itself. So whatever intellects something may intellect itself (fa-kull mā
˘
ya qilu šay an fa-lahu an ya qilu dātahu). And whatever is intellected is such that its
˘
4
essence (māhiyya) is conjoined ¯ to another intelligible. Thus it is also intellected
5 along with something else, and inevitably the intellective power intellects it precisely
6 in conjunction [with this other thing]. If [this other thing] is something that subsists
7 through itself, then nothing in its true nature (haqı̄qa) prevents it from conjoining
8 with the intelligible concept. Unless in fact (fı̄˙ wuğūd) it is hindered from this,
9
because it is afflicted by being conjoined with items (umūr) that are material (min
mādda), or by something else, as the case may be. But if its true nature is
10
unimpaired (musallama), it will not be hindered from conjunction with intellectual
11 forms. And so, [even in the case where it is hindered] it will have this as a possibility,
12 and within this (dimna dālika) is the possibility of its intellecting itself.
˙ ¯
13
This passage is obviously related to Aristotle’s treatment of self-intellection in De
14
anima III.4, and brings out something important about the original Aristotelian
15
16
text. As I said above, the end of De anima III.4 discusses two aporiai: why
17
intellects are objects of thought for themselves, and why external intelligibles are
18
not all able to engage in intellection. But there is a third aporia mentioned by
19
Aristotle, which I omitted in my previous translation of the passage: ‘the reason
20
why intellect is not always perceiving (mudrikan) should be considered’
21 (430a5 – 6, p. 74 line 15 – 16 Arabic). It is not a stretch to think that the
22 infamous following chapter, De anima III.5, is in part an attempt to resolve this
23 aporia, since it includes the claim that the agent intellect always thinks. Nor is it
24 a stretch to think that the explanation for an intellect’s failing always to think is
25 the same explanation that resolved the other two aporiai: matter. In other
26 words, it will be embodiment that results in an intellect’s being only
27 intermittently intellecting.29 It’s clear that Avicenna has the Aristotelian passage
28 in mind in Pointers § 3.19, just as he does in the texts we examined above. As we
29 will see shortly, al-Tūsı̄ also thinks that III.4 is relevant to Avicenna’s discussion.
˙
30 Pointers § 3.19 begins with an explicit endorsement of claim (2): intellection
31 of any intelligible object always brings with it the possibility of self-intellection.
32 If we suppose, as I have just suggested, that self-intellection is not just self-
33 awareness, then what is it? Perhaps Avicenna is making an epistemic point,
34 namely that if I know something, then I must know that I know it. I suspect
35 that Avicenna would accept this internalist constraint on knowledge, which was
36
37
38
˘
39 28 Or ‘that it is intellecting it [sc. that thing]’, if we read ya qiluhu with al-Tūsı̄’s lemma; as
we will see this is in any case how he interprets the passage. ˙
40
29 Some support for this interpretation might be had from Nicomachean Ethics X.8,
41 1178b33 – 5: ‘[human] nature is not sufficient for contemplation (pq¹r t¹ heyqe?m), but
42 the body must have health, food, and other care.’
108 Peter Adamson
1 explicitly set out by al-Fārābı̄ earlier in the Arabic tradition.30 Indeed, Avicenna
2 might even think that the possibility of self-intellection implies the possibility of
3 knowing that one knows. But even if Avicenna has this in mind in § 3.19, he is
4 also talking about something more ambitious, namely having oneself as an
5 intelligible object. This seems to be broader than simply knowing that one’s
6 present judgement is an act of knowledge. Perhaps much broader: it might
7 imply that one must at least potentially have a full grasp of one’s own essence in
8 order to engage in intellection at all. Certainly it is this maximal interpretation
9 that Avicenna seems to employ later on, when he discusses God as a self-
10 intellecting intellect. Avicenna wants, then, to show that intellection implies at
11 least the possibility that one is an intelligible object for oneself. This possibility
12 may or may not be realised in human intellects at any given time, since human
13 intellects are connected to matter. For an immaterial intellect like God, though,
14 it will be a permanent feature of intellection.
15 But why think any of this is true? Let us look at the beginning of the passage
16 in more detail, alongside the discussions from our two commentaries. The
17 tanbı̄h begins with the remark that self-intellection occurs through a potency or
18
˘
power (quwwa) which is ‘close to actuality (qarı̄ba min al-fi l)’ (II.416). As al-
19 Tūsı̄ explains, this is to be understood in terms of Aristotle’s distinction between
20 ˙
levels of actuality:
21
22 al-Tūsı̄, Commentary II.416, lines 4 – 11: [Avicenna] posits three degrees of potency:
˙
distant, which is material intellect; intermediate, which is habitual intellect; and
23
‘close (qarı̄ba)’, which is actual intellect. This is the one required in order that the
24 subject of intellection may behold its object whenever it wishes. His point is that
25 whatever actually intellects something may, whenever it wishes, intellect that it itself
26 is intellecting this thing. For its intellection of this thing is this thing’s occurring
27 (husūl) to it. And its intellection of the fact that it itself is intellecting this thing is
˙ ˙occurring of this occurrence for it (husūl dālika l-husūl la-hu).
the
28 ˙ ˙ ¯ ˙ ˙
29 He seems to be taking ‘close (qarı̄b)’ here to mean something like ‘available’ or
30 ‘ready to hand’, seeing an allusion to De anima III.4. I believe al-Tūsı̄ is drawing
31 ˙
an analogy. When one has previously grasped some intelligible and thus has a
32 habitual grasp of it, one has that intelligible ‘close’ or ‘available’. This means that
33 one can grasp it again at will: Aristotle makes the point at De anima III.4,
34 429b4 – 9. In just the same way, when one is actually grasping that intelligible,
35 one is ‘close’ to grasping that one is grasping it. Or in other words, one may at
36 will intellect that one is intellecting. The reason al-Tūsı̄ gives is that ‘knowledge
37 ˙
˘
through assertion (al- ilm bi-l-tasdı̄q)’ involves ‘conceptualizing the subject

38 ˙
˘
[of the assertion] (tasawwur al-mawdū )’ (al-Tūsı̄, Commentary, II.217.4 – 5).

39 ˙ ˙ ˙
40
41 30 As discussed by Black, Knowledge and Certitude. See also Adamson, Knowledge of
42 Universals.
1 Despite the technical terms,31 what he means is fairly clear. Suppose I grasp that
2 horse is animal. This requires, indeed involves, my grasping horse. Similarly, if I
3 grasp that my intellect is grasping something, then I also grasp my intellect.
4 So, according to al-Tūsı̄, the argument in full goes like this (II.417, lines 8 – 11):
5
˙
Whatever intellects something may, at will, intellect the fact that it is
6 intellecting that thing.
7
Whatever intellects that it is intellecting something also intellects itself.
8
Therefore whatever intellects something intellects itself.
9
10 And the conclusion of this syllogism is equivalent to claim (2). But
11 unfortunately, al-Rāzı̄ has a clever point which would undermine or at least
12 weaken the second premise of the syllogism:
13
al-Rāzı̄, Commentary I.172, lines 26 – 9: [Avicenna’s] saying ‘it thereby intellects
14 itself ’ suggests that a thing’s intellecting that it is intellective of something else is the
15 same as its intellecting itself. But this is not the case. For my knowledge that I am
16 knowing something is knowledge concerning a specific relation between myself and
17 the knowledge of something. And the knowledge of that relation is different from
18
the knowledge of either of the two relata. So how can one say that the knowledge of
that relation is knowledge of myself ? Rather, one must say that my knowledge that I
19
am knowing something includes (yatadammana) or comes along with (yaltazimu)
20 my knowledge of myself. ˙
21
22 Al-Rāzı̄ may already be thinking ahead to Avicenna’s claim that God’s
23 knowledge of other things is identical to His knowledge of Himself. (This is
˘
24 suggested by the shift of terminology from ‘intellection’ to ‘knowledge’, ilm.)
25 Here he sows a doubt that self-intellection could be like this, because
26 intellection or knowledge is a relation – as we saw above, an asymmetrical
27 relation. Of course it is true that one must grasp X in order to grasp that X is
28 related to Y. But the grasp of X, the grasp of Y, and the grasp of the relation are
29 three different things. One might add, though al-Rāzı̄ does not,32 that the
30 knowledge of the relata should really be prior to the knowledge of the relation.
31 If that were so, then I would already have to have self-intellection before
32 intellecting that I intellect some other object.
33 Our two commentators both devote considerably more time to the rest of
34 § 3.19, in which Avicenna seeks to prove claim (1): every self-subsisting
35 intelligible object is also a subject of intellection. This is precisely the converse
36
37 31 The contrast between tasdı̄q and tasawwur goes back at least to al-Fārābı̄. See the
38 discussion of these terms ˙in Black, al-Fārābı̄
˙ on Meno’s Paradox, p. 25, n. 32.
39 32 Though he makes a similar point at the end of his commentary on this tanbı̄h: ‘its
intellection of itself is not a part of its intellection of something else; and whatever is not
40
part of something is not included within that thing. Rather, one must say that its
41 intellection of things has as an implication the possibility that it may intellect itself, where
42 this implication is from cause to effect’ (al-Rāzı̄, Commentary I.173, lines 11 – 18).
110 Peter Adamson
1 of a situation we considered above. We saw that intelligibles that subsist in

2 matter, or insofar as they are grasped by an intellect, cannot engage in
3 intellection. Here, we have the converse claim that every self-subsisting
4 intelligible can engage in intellection.33 Avicenna’s proof, as I understand it, is as
5 follows. Suppose that we have an intelligible object; call it O. If O is intelligible,
6 then we know that it can conjoin with other intelligibles in two ways. First,
7 when I grasp O, then O will be conjoined with the other intelligibles that I
8 grasp. As we saw above, Avicenna will say explicitly in the next tanbı̄h, Pointers
9 § 3.20, that this is the case. (The example I used above was that granite is
10 ‘conjoined’ or co-present with iron even though neither granite nor iron grasp
11 each other intellectually.) Second, if I intellectually grasp O then O will also be
12 conjoined with my intellect, and my intellect is intelligible. For we have just
13 seen that my intellect is intelligible to itself. The conclusion to be drawn from
14 this is that there is in principle no obstacle to O’s being conjoined to other
15 intelligible objects. Now suppose that O is self-subsistent. Then it can itself
16 engage in intellection, because intellection is nothing more than a self-subsistent
17 thing’s conjoining with intelligibles. So long, that is, as O is ‘unimpaired’. If it is
18 hindered, for example by coming into a relation with matter,34 then this may
19 prevent it from engaging in intellection. Furthermore, O will be able to engage
20 in self-intellection. For Avicenna just showed with claim (2) that whatever
21 engages in intellection can also intellect itself. But again, matter (or other
22 hindrances, which here remain unspecified) can prevent this from happening.
23 Thus matter accounts for the fact that intelligibles – including our own intellects
24 – do not always intellect or self-intellect. Yet this is always theoretically ‘possible’
25 for any self-subsisting intelligible, as Avicenna says at the end of the tanbı̄h. 35
26
27
28 33 It stands to reason that these are the only three ways for an intelligible to subsist. It must
29 subsist either by itself, or in something else; and if in something else, then in matter or
not in matter, i. e. in an intellect.
30
34 Note that this relation will not be one of subsisting in matter, because then of course it
31 would not be self-subsistent. What Avicenna has in mind is not an intelligible like granite,
32 which never subsists except in matter or in an intellect, but an intelligible like a human
33 intellect, which is self-subsisting but is sometimes ‘afflicted’ by being related to a body.
34 35 The two commentators’ remarks on the nature of this possibility show the sophistication
with which they handled modal notions. Al-Rāzı̄ points out the innate possibility of
35
intellects for self-intellection is also realised by God, who however is necessarily self-
36 intellective. Thus the ‘possibility’ in question is ‘possibility in the general sense (imkān
˘
37 āmm)’ (al-Rāzı̄, Commentary I.172, line 23). I take this to mean ‘one-sided possibility’,
38 i. e. the possibility which includes both the contingent and the necessary. Al-Tūsı̄ rejects
this, because if the possibility in question included the contingent it would ˙ include
39
things that never engage in self-intellection, though they might have. As he puts it,
40
‘“possibility in the general sense” applies to the remote possibility [that is present even]
41 when something is always non-existent, but not necessarily so’ (al-Tūsı̄, Commentary
42 II.416, lines 18 – 19). Instead, we should take imkān here to refer to the˙ capacity for self-
1 In objecting to this argument al-Rāzı̄, in a move that is by now familiar,

2 exploits the idea that intellection is a conjunction of mahall and hāll. He
˙ ˙
3 distinguishes between three types of conjunction: (a) the hāll’s conjunction to its
˙
4 mahall; (b) the mahall’s conjunction to its hāll; (c) the conjunction of two hālls
˙ ˙ ˙ ˙
5 in a single mahall (al-Rāzı̄, Commentary I.171, lines 25 – 6). To put the point in
˙
6 English, he is envisioning three relations: (a) the relation borne towards a
7 subject by what inheres in it; (b) the converse relation which the subject bears
8 towards what inheres it in; (c) the relation between two things that inhere in a
9 single subject. Al-Rāzı̄ then argues that something’s being capable of entering
10 into a conjunction of type (a) or (c) does not mean that it can enter into a
11 conjunction of type (b). In other words, the fact that the intelligible object can
12 conjoin to my intellect and to other intelligibles grasped by my intellect does
13 not mean that it can itself be the subject of intellection. He adds a further reason
14
to think that the inference is unproven. Quiddities have features when they are
15
‘mental (dihnı̄)’ that they lack when they are ‘externally real (hāriğı̄)’. For
16 ¯ ˘ does not.
instance when man is a hāll, it requires a mahall, but the real man
17 ˙ ˙
Conversely the real man is capable of sensation, motion, and so on, but man in
18
my mind is not (I.172, lines 1 – 3).36 Thus, ‘from the fact that the mental
19
quiddity is conjoined to the intelligibles, it does not follow that the real quiddity
20
is the same’ (I.172, line 4).37
21
22
23 intellection, i. e. in light of the Aristotelian concept of first and second actuality (lines
24 19 – 20).
36 In response to the point, al-Tūsı̄ quite reasonably distinguishes between man considered
25
as an essence and man considered˙ as a mental item. When I grasp man intellectually I am
26 grasping the essence mentally, not grasping a mental essence. Al-Tūsı̄ does admit that it is
27 possible to consider man precisely insofar as it is a mental item, ˙ but this would be an
28 additional act of the intellect, in which it would reflect on its own concepts (II.421, lines
29 13 – 16). For a very similar discussion in Yahyā b. ‘Adı̄, see Adamson, Knowledge of
Universals, pp. 153 – 4. ˙
30
37 For an independent discussion discussing proofs that the intelligible must be intellective,
31 see Fahr al-Dı̄n al-Rāzı̄, Mabāhit, I, pp. 369 – 72. In this passage al-Rāzı̄ sets out three
32 ˘
arguments ˙¯
to show that every separate substance is self-intellective. The first is explicitly
˘
drawn from Avicenna, but from al-Mabda wa-l-ma ād (Origin and Return) rather than
˘
33
34 the Pointers. The passage concludes with a fourth argument which al-Rāzı̄ apparently
endorses; this looks as if it is inspired by Pointers § 3.19 – 20. To summarize the
35
argument (Mabāhit I, p. 371, line 13 to p. 372, line 5): any given intelligible object O1
36 can conjoin with˙ ¯another intelligible O2 in an intellect. But this conjunction cannot
37 presuppose the intellect’s intellection of O1 and O2, because the intellection just is the
38 conjunction of the intellect with O1 and O2. Therefore the objects must be conjoined
39 independently of this intellect which grasps them. In order for this to be the case, O1
must exist outside the intellect in question: ‘the intelligible quiddity always exists among
40
the individuals, subsisting through itself ’ (372, line 2). And if O1 exists as self-subsistent,
41 then its conjunction with other intelligibles will be due to ‘the inscribing of their forms
42 in it’ (p. 372, line 3).
112 Peter Adamson
1 Al-Tūsı̄ is unimpressed by this whole line of objection. Regarding the

2
˙ between types of conjunction, he says that the proof needs only to
distinction
3 show that the intelligible is capable of conjunction at all (mutlaqan; II.421, line
4 6). This seems to be a disagreement over where the burden˙ of proof lies. Al-
5 Tūsı̄’s point is that if the intelligible object can enter into conjunctions of type
6
˙ and (c), then some good reason would have to be provided for why it cannot
(a)
7 enter into a conjunction of type (b). But I think al-Rāzı̄ has the better of the
8 argument here, at least if we assume that the intellect-intelligible relation is
9 asymmetrical (an assumption Avicenna himself seems to make, as we have
10 seen).38 The ability of an intelligible to play the passive role in this relation does
11 not obviously imply any ability to play the active role in a relation of the same
12 type. So there is a good reason for distinguishing between things that enter into
13 conjunctions of type (a) and (c) – that is, things that inhere in something else –
14 and things that enter into conjunction type (b), that is, things that are subjects
15 of inherence.
16 Let us take stock. In the brief compass of Pointers § 3.19, Avicenna has laid
17 two important parts of the foundation for his theory of God. He has shown (2)
18 that every intellect is capable of self-intellection: whenever it grasps an
19 intelligible object it thereby grasps itself or has the ability to do so. Furthermore,
20 Avicenna has shown (1) that any intelligible object is capable of intellection –
21 and hence self-intellection – so long as it is self-subsisting, and so long as it is
22 not hindered by matter or any other factor. Again, al-Rāzı̄ and al-Tūsı̄ disagree
23 about the success of Avicenna’s arguments. Regarding claim (1), the˙ issue seems
24 to boil down to a question of the burden of proof. Whereas al-Rāzı̄ demands to
25 be shown that an object of intellection will be the sort of thing which can itself
26 engage in intellection, al-Tūsı̄ seems to think the absence of an argument to the
27 contrary is sufficient. Here˙ al-Rāzı̄’s position looks more intuitive. On the other
28 hand, regarding claim (2), al-Rāzı̄ suggests that grasping that one knows
29 something is not the same thing as grasping oneself. But al-Tūsı̄’s reconstruction
30 of Avicenna’s argument renders it fairly plausible: to know ˙ something does
31 imply at least the possibility of knowing oneself.
32
33
34
35
36
37
38
39 38 There is however one reason to question that the intellection relation is always
asymmetrical. In self-intellection, the subject and object are the same thing, so it is hard
40
to see how there could be any asymmetry. In the next section of the paper, we will see
41 that al-Tūsı̄ uses a similar argument to refute the idea that intellection is a kind of
42 ˙
inherence.
1 God as a Self-Intellective Intellect

2
3 At any rate, Avicenna takes himself to have established claims (1) and (2): any
4 intelligible object can itself engage in intellection, and whatever can do this
5 can also engage in self-intellection. He now need only add that God is a self-
6 subsisting, immaterial and unhindered intelligible to reach the desired conclusion
7 that God is an intellect who intellects Himself. Al-Rāzı̄ already tries to block this
8 inference in his commentary on Pointers § 3.19, by questioning the basic
9 assumption that God is an intelligible. For, he insists, man cannot have an
10 intellectual grasp of God (I.170, line 31), as Avicenna himself would admit
11 (I.171, line 5 – 6). Of course if God is not intelligible to us, then we have no
12 evidence that God is intelligible at all. In which case we have no reason for
13 applying the argument of § 3.19 to the divine case. This objection leads to an
14 interesting debate between al-Rāzı̄ and al-Tūsı̄ (cf. the latter’s Commentary,
15 II.419) as to whether Avicenna’s demonstration ˙ of the Necessary Existent
16 constitutes a case of God becoming intelligible to man.39 But I will pass over this
17 issue so as to move on finally to the seventh namat of part two of the Pointers,
18
˙
which is partially devoted to the topic of God’s self-intellection. Before I do so I
19 should emphasize that I am not going to deal with the problem of God’s
20 knowledge of particulars. I have discussed this elsewhere, albeit not with
21 reference to the commentaries on the Pointers. 40 Instead I want to look at the
22 more fundamental problems that arise even in saying that God grasps
23 intelligibles (never mind particulars) by grasping Himself.
24 The first threat to guard against is that, as we have now seen at length, God
25 is not the only immaterial entity that grasps both intelligibles and itself.
26 Avicenna is thus in danger of making God’s intellection too much like the
27 intellection enjoyed by creatures. Consider for instance the following passage,
28 from the Metaphysics (Ilāhiyyāt) of Avicenna’s Healing:
29
The Necessary Existent is pure intellect because He is an essence dissociated
30 (mufāriqa) from matter in every respect. You have known that the cause that
31 prevents a thing from being apprehended intellectually is matter and its attach-
32 ments, not [the thing’s] existence … Likewise He is a pure intelligible, because that
33 which impedes a thing from being an intelligible is its being in matter and its
attachments. This is the impediment preventing [the thing] from being an intellect
34
… Hence, that which is free of matter and [its] attachments [and is] realised
35
˘
through existence separate [from matter] is an intelligible for itself (ma qūl li-dātihi).
36 Because it is itself an intellect, being also intellectually apprehended by itself,¯ it
˘
37 [itself ] is the intelligible [belonging] to itself ( aql bi-dātihi wa-huwa aydan ma qūl
¯ ˙
38
39
40
39 Al-Rāzı̄ makes a similar point, and describes a possible response which anticipates al-
41 Tūsı̄’s, at Mabāhit I, 371 line 8 – 12.
42 ˙
40 Adamson, ˙¯
On Knowledge of Particulars.
114 Peter Adamson
˘
1 bi-dātihi fa-huwa ma qūl dātihi). Its essence is, hence, [at once] intellect, intellectual
¯
apprehender, ¯ – not that there are multiple things here.41
and intelligible
2
3 The striking thing about this passage is that Avicenna is relying solely on God’s
4 immateriality to establish the nature of divine intellection. Everything said here
5 should, then, apply equally well to any immaterial entity. This may be why
6
Avicenna is so emphatic that God is separated from matter in every respect (min
7
kull ğiha). But it is doubtful whether even this can distinguish God’s mode of
8
intellection from that of humans after their death, or from that of the celestial
9
intellects, which have no direct connection to their celestial bodies.
10
Avicenna responds to this threat in a sequence found in the seventh namat,
11
where he builds a case for the sui generis character of divine intellection. The ˙
12
first step is to say that God’s intellection precedes the existence of the intelligible
13
outside Him, whereas humans normally acquire their intelligibles from what
14
already exists on the outside:
15
16 Pointers § 7.13 (III.275 – 6): The intellectual forms may in one way be acquired
17 from external forms, for instance whenever one acquires42 the form of the heavens
from the heavens. But it may be that the form first arrives at the intellective power,
18
and thereafter external existence comes to it, for instance when one intellects a
19 figure, and thereafter makes it exist. The Necessary Existent’s intellection of
20 anything must be in the second way.
21
22
Both al-Rāzı̄ and al-Tūsı̄ adopt a terminological contrast between ‘active’ and
˙
˘
23
‘passive’ knowledge ( ilm fi lı̄ vs. ilm infi ālı̄) for Avicenna’s distinction. As the
24 Avicennan text itself illustrates with its example of the human geometer, ‘active
25 knowledge’ is not the privilege of God alone. But in the case of God all His
26 knowledge is productive, never acquired from ‘outside’ Him. This point is
27 illustrated nicely by a remark in the Metaphysics of the Healing, where Avicenna
28 says that God’s creative activity is like ours would be ‘if the very existence of the
29 intelligible forms occuring in us […] were sufficient for the generation from
them of the artificial forms (al-sūwar al-sinā iyya).’43
˘
30
But surely Avicenna’s system ˙ recognizes
˙ other things that produce on the
31
32 basis of grasping intelligibles? What about the Active Intellect, whose grasp of
33 its intelligibles allows it to bestow the corresponding forms on sublunary
34 matter? Admittedly, there will be a difference here, in that the Active Intellect
35 receives forms and existence from superior entities, and gives form only when
36 matter is suitably disposed. Neither is true of God. Still, Avicenna seems to be
37
38 41 Avicenna, Metaphysics, VIII.6 § 6 – 7. I quote here from Marmura’s translation.
39 42 As Dag Nikolaus Hasse has pointed out to me, the term ‘acquisition’ here probably
indicates that Avicenna is explicitly ruling out the case of first principles, which are not
40
‘acquired’ from experience but are known innately. On this see Hasse, Avicenna’s ‘De
41 anima’, p. 179, and Adamson, Non-Discursive Thought, pp. 96 – 7.
42 43 Ibid. VIII.7, § 11, Marmura’s translation.
1 aware that he is in danger of making God too similar to the celestial intellects.
2 He addresses the issue in the next several sections, Pointers § 7.14 – 16. First, he
3 introduces a regress argument not unlike his famous argument for the existence
4 of God. It cannot, he says, be the case that every intellect receives intellection
5 from outside. Rather, just as something must exist through itself, so must
6 something intellect through itself. And this will be God (§ 7.14). Furthermore,
7 since God’s intellection consists in ‘active’ knowledge, His intellection will be
8 the cause for what comes after it (§ 7.15). The upshot is that there are in fact
9 three types of intellection, as follows:
10
Pointers § 7.16 (III.279 – 80): The First’s perception (idrāk) of things is from
11 Himself and in Himself, and occurs in the best way that something can perceive and
12 be perceived. Succeeding this is the perception the consequent intellective
13 substances have of the First, through the illumination (išrāq) of the First, and of
14 what is after Him and from His essence. And after these two [kinds of perception]
15
are the perceptions of souls (idrākāt al-nafsāniyya), which are a figure and sketch
(naqš wa-rasm) of natures that are intellective and are dispersed as principles and
16 relation.44
17
18 Here, then, we have the difference between divine and non-divine self-
19 intellection. All types of intellect are capable of self-intellection, as established
20 by Pointers §3.19. But only God’s intellection consists entirely in self-
21 intellection. The other separate intellects, the lowest of which is the Active
22 Intellect, do grasp intelligibles and grasp themselves in doing so. Since they are
23 immaterial, we can infer that they in fact engage in self-intellection permanently.
24 But their grasp of the intelligibles is nonetheless derived from God, much as
25 their existence is derived from Him, by a kind of ‘illumination’. Furthermore,
26 Avicenna reminds us that these intellects will also have a grasp of God Himself,
27 and this will be neither ‘active knowledge’ (the celestial intellect does not, of
28 course, produce God) nor the consequence of self-intellection.
29 Al-Rāzı̄ makes several interesting points in his commentary on these sections
30 of the Pointers. Regarding § 7.15, al-Rāzı̄ challenges the idea that God will know
31 His effects just by knowing Himself. For God can conceive of Himself simply as
32 Himself, or as the cause of His effects. In the former case, it is begging the
33 question to say that God’s grasp of Himself will include the knowledge of His
34 effects. But in the latter case, what God is grasping is a relation between Himself
35 and His effects. And, reminiscently of his criticism of Pointers § 3.19, grasping
36 this relation will require a prior grasp of the relata, yielding a vicious circle (al-
37
38 44 This last phrase is difficult to understand. Goichon sees a distinction between two types
39 of knowledge : one knowing the effect by the cause, the other vice-versa (Avicenna,
Livre des directives et remarques, ed. Goichon, p. 452). Here she is following al-Tūsı̄’s
40
commentary (III.281) quite closely. Though the details are obscure, it’s clear ˙ that
41 Avicenna has in mind the sort of intellection that is accompanied by discursive
42 reasoning, i. e. human intellection. On this see Adamson, Non-Discursive Thought.
116 Peter Adamson
1 Rāzı̄, Commentary II, p. 69, line 35-p. 70, line 2). But the commentary is not
2 always critical. Regarding the taxonomy of three types of intellection in § 7.16,
3 al-Rāzı̄ gives a rather elegant explanation of why Avicenna is right to say that a
4 separate intellect’s grasp of itself will not entail a grasp of God. By considering
5 its own essence, the separate intellect will realize that it is possible-in-itself, and
6 therefore know that its own existence required a cause. But any cause will do. It
7 will not, by considering its own essence, come to an understanding of the
8 specific cause that has brought it into existence, namely God (al-Rāzı̄,
9 Commentary II.70 lines 8 – 30). This argument on Avicenna’s behalf meets
10 with approval from al-Tūsı̄, who repeats the point in his own commentary
11 ˙
(III.280).
12 There remains one other problem to deal with before Avicenna can turn his
13 attention to the more famous difficulty about God’s knowledge of particulars.
14 The problem is already signalled in the citation from the Metaphysics of the
15
Healing given above. There, after Avicenna describes God as intellect,
16
intellecting, and intelligible, he immediately adds, ‘not that there are multiple
17
things here’. In fact, as any student of Plotinus will know, there are two
18
potential sources of multiplicity that arise when we describe God as an intellect.
19
First, there will be the duality implied by His being both a subject and an object
20
of intellection. This is apparently the kind of multiplicity Avicenna has in mind
21
in this passage from the Healing. Avicenna’s position on this differs from that of
22
Plotinus, in that he takes self-intellection to imply no multiplicity. (As we will
23
see al-Tūsı̄ tries to explain why, in the case of self-intellection, knowledge is not
24 ˙
25
something distinct from the knower.) But there is also the multiplicity of
26
intelligibles that God must grasp, if He is to grasp His effects by grasping
27
Himself. Let me stress again that this is not a worry about knowing particulars.
28
Before we can concern ourselves with whether a simple God can grasp changing
29 particulars like Zayd and ‘Amr, we need to understand how a simple God could
30 grasp both man and horse.
31 There is an irony here. When Avicenna begins to describe God in the
32 Healing, almost the first thing he says is that God is not multiple (VIII.4, § 2).
33 Among the consequences of His simplicity is His immateriality. And as we have
34 seen, God’s immateriality, together with His intelligibility, implies that He is an
35 intellect. But as an intellect, God will grasp a multiplicity of objects. And this
36 suggests that God is Himself multiple after all. In other words, Avicenna has
37 begun from the premise that God is not multiple and on that basis apparently
38 argued to the conclusion that God is multiple. For simplicity entails
39 immateriality; immateriality entails intellectivity; intellectivity entails having a
40 multiplicity of intelligible objects; and having a multiplicity of intellective
41 objects implies a lack of simplicity. Clearly, something has gone wrong here.
42 Avicenna’s diagnosis of the problem is that the last inference is fallacious:
1 Pointers § 7.17 [III.281 – 5]: Perhaps you will say, ‘the intelligibles are unified
2 neither by the subject of intellection nor by each other, according to what you have
3 said. Furthermore, you have admitted that the Necessary Existent intellects
everything. So He will not be truly one, but, in that case, multiple.’ We say that,
4
because He knows Himself through Himself, the fact that He makes things subsist
5 (qayyūmiyyatuhu), as an intellect of Himself through Himself, furthermore implies
6 that He intellects multiplicity. Multiplicity arises as a posterior consequence, not as
7 something internal to the essence which makes it subsist. And it arises too in a
8 certain order. A multiplicity of consequences from the essence – whether these are
9 distinct or not – does not sully the unity [of the essence]. A multiplicity of
consequences belongs to the First, both relational and non-relational, and a
10
multiplicity of negations, and for this reason a multiplicity of names. But from this
11 there is no effect on the unity of His essence.
12
13 Al-Rāzı̄’s commentary on this tanbı̄h begins with a rather polemical restating of
14 Avicenna’s view: for Avicenna ‘the essence of God is the subject (mahall) for all
˙
15 those many forms’ (II.71 line 7). Yet again, al-Rāzı̄ deploys the idea that intellect
16 relates to intelligible as a mahall relates to its hāll. His motive in doing so
˙ ˙
17 becomes clear as he proceeds: he is trying to place Avicenna’s position within the
18 context of the theological dispute over divine attributes. He says that Avicenna
19 here departs from the views of the philosophers (falāsifa) on two points (II.71
20 line 30). Firstly, the philosophers say that nothing simple can be both active and
21 passive in the same respect. But Avicenna is saying that God both produces and
22 receives the intelligibles. Secondly, al-Rāzı̄ finds it quite striking that Avicenna
23 allows for ‘a multiplicity of consequences, both relational and non-relational’.
24 This implies that God has ‘fixed attributes (sifāt tubūtiyya)’, which the
˙ ¯
25 philosophers deny, according to al-Rāzı̄. He adds, ‘and how could he avoid
26 conceding this, given that according to him God knows the quiddities, and
27 [given that] knowledge, according to him, consists in the occurrence of the
28 forms of [those quiddities] in the knower?’ (II.71, lines 33 – 4). This is not so
29 much a criticism as an attempt to foist on Avicenna a view on the notorious
30 kalām problematic regarding divine unity and attributes. Specifically, he claims
31 that Avicenna’s position is closer to that of the Ash‘arites than to that of the so-
32 called falāsifa. While al-Rāzı̄’s remarks here might seem tendentious, it must be
33 said that Avicenna has invited this treatment by referring to God’s names.45
34 Al-Tūsı̄’s commentary on this section consists largely in a scornful response
35 ˙
to al-Rāzı̄. First, he elaborates on Avicenna’s solution, insisting that the
36
multiplicity involved in divine intellection is posterior to that intellection: ‘the
37
multiplication of consequences and effects does not nullify the unity of their
38
cause which gives rise to them as consequences, regardless of whether the
39
40
45 Indeed al-Rāzı̄’s interpretation here is quite plausible. For Avicenna’s response to the
41 kalām tradition of disputes over divine attributes, see also Wisnovsky, Avicenna’s
42 Metaphysics.
118 Peter Adamson
1 consequences are established in the essence of the cause or are distinct from [that
2 essence]’ (III.282, lines 13 – 14). Then he moves on to rebutting al-Rāzı̄’s claims
3 (III.282, lines 21 ff.). He flatly denies that God is in any sense ‘passive’ as the
4 object of His own intellection. He then rejects the idea that the intelligible is
5 present in God as a distinct attribute. His argument here takes us back to the
6 mechanics of human intellection:
7
al-Tūsı̄, Commentary III.284, lines 7 – 11: You do not think that your being a subject
8 (mah ˙ all) for that form is a condition for your intellecting it. After all you intellect
9 ˙
yourself, without being a subject for yourself! Rather, your being a subject for that
10 form is simply a condition for the occurrence (husūl) of that form for you. And this
11 is a condition for your intellecting it. So if that˙ form
˙ occurs to you in another way
besides its inhering in you, then intellection happens without inhering in you (hulūl
12 ˙
fı̄-ka).
13
14 In passages examined above, al-Tūsı̄ has subtly distanced himself from the
˙
assumption so often deployed by al-Rāzı̄, that intellection is just an example of
15
16 inherence. Here, he finally criticizes this assumption explicitly. For al-Tūsı̄, when
I grasp an intelligible form, that form does not strictly speaking ‘inhere’ ˙ in me,
17
46
18 but rather ‘occurs’ to me. The proof lies in self-intellection, for it is absurd to
19 think of something’s inhering in itself, but perfectly reasonable to think of
20 something’s being ‘occurrent’ to itself. Al-Rāzı̄’s view remains seductive, because
21 in the human case a form must indeed inhere in an intellect in order for it to
22 occur to that intellect. Presumably this is because the form comes to the human
23 intellect from outside. But we are capable also of self-intellection, and this is
24 akin to God’s intellection which is entirely self-directed. Humans are thus never
25 so much like God as when they intellect themselves. In this case, there is no
26 inherence involved, but simply the occurrence of one’s intellect to itself. One
27 might say that Avicenna, at least on al-Tūsı̄’s reading, has brought together the
28 Delphic imperative ‘know thyself ’ with˙ the Platonic idea of imitating God.47
29 For, if humans come closest to divine thought by thinking about themselves,
30 then self-thinking and imitation of the divine turn out to be the same thing.
31 On the other hand, al-Tūsı̄ sees significant differences between human
˙
intellection and divine intellection. Not only is our intellection partial and
32
33 intermittent, but it requires that we receive a form which then ‘occurs’ in our
34 intellects. In this respect human thought is normally ‘passive’ and posterior to an
35 external object. Both self-intellection and ‘active intellection’ (which precedes its
36 external object) are for us the exception, not the rule. Still, as al-Tūsı̄ says
˙
37
38 46 In his commentary on Pointers § 3.19, al-Tūsı̄ likewise spoke of ‘occurrence’ rather than
˙
inherence; see e. g. II.416, lines 4 – 11, quoted above.
39
47 In fact, as Heidrun Eichner has suggested to me, Avicenna may have developed his
40
interest in human self-intellection precisely because he was already interested in the
41 question of divine self-intellection. In which case it is no surprise that al-Tūsı̄ is able to
42 make a convincing case for resonances between the two types of thinking.˙
1 (III.284, lines 5 – 6), if humans are capable of a kind of intellection where

2 subject and object are identical, how much more will it be the case for God?
3 Indeed, he argues (III, p. 284, line 16-p. 285 line 12), God is as it were the limit
4 case of this intimate relation, because there is no difference (taġāyur) at all
5 between God and God’s intellection of Himself. Or rather, there is a difference,
˘
6 but only a conceptual (i tibārı̄) one. Since God and His self-intellection are in
7 reality one and the same, the effects of these two things are also one and the
8 same. That is, ‘the existence of the first effect is the same as the First’s
9 intellection of it, without needing any additional form that would inhere in the
10 essence of the First.’ (By ‘the existence of the first effect’ I take al-Tūsı̄ to mean
11 ˙
God’s production of the first emanated intellect, and not the intellect itself.)
12 According to this argument, the production of God’s effects and His intellection
13 of those effects is posterior to God’s essence. But God’s self-intellection is not
14 posterior to His essence – rather, the two are really identical, and only
15 conceptually distinct.48
16 To recapitulate this section, we have seen Avicenna argue that God’s
17 immateriality implies that He engages in self-intellection, but also that there
18 are principled reasons for distinguishing between His intellection and that
19 performed by other immaterial beings. Rather polemically, al-Rāzı̄ suggests that
20
Avicenna’s view here is much like theological views which accept attributes
21
distinct from the divine essence. He also ascribes to Avicenna the view that God
22
is in some sense passive, since He receives intelligibles. Of course al-Tūsı̄ takes
23 ˙
issue with this, and in doing so makes some interesting distinctions between the
24
standard case of grasping an intelligible object, and the special case of self-
25
intellection – which is only one possible act of intellection for humans, but the
26
very nature of thought in the case of God.
27
28
29
Conclusion
30
31
The thread of argument I have just followed, taking us from the third to the
32
seventh namat of the Pointers, is a classic example of Avicenna’s method. He is
undoubtedly ˙inspired by Aristotle, to the point that we can specify the
33
34
Aristotelian passages he has in mind. In the present case, we can think of
35
Avicenna as building a bridge from the discussion of self-intellection in De
36
anima III.4 to the claims about divine self-intellection found in Metaphysics
37
38
book Lambda.49 On the other hand, it is equally typical that he adds complex
39
48 This part of the commentary should be read alongside the Healing, Metaphysics VIII.7,
40
§ 4. The Healing also affirms that the intelligibles are not present to God as ‘forms’
41 (VIII.7, § 1), which lends support to al-Tūsı̄’s claim that they do not inhere in God.
42 49 See further Bertolacci, The Reception. ˙
120 Peter Adamson
1 new arguments for Aristotle’s claims. Thus he gives a much more elaborate
2 rationale for the claim that any intellect can engage in self-intellection, and is
3 more explicit about the inhibiting role of matter that Aristotle mentions in De
4 anima III.4. He also stakes out entirely new positions of his own. For instance,
5 as far as I can see, the view that every self-subsisting intelligible object must be a
6 subject of intellection is not clearly present in Aristotle.50 All of this gives
7 Avicenna a richer psychology to draw on when he comes to describe divine
8 intellection.
9 I hope that the foregoing has also illustrated the value of reading Avicenna
10 alongside his commentators. Al-Rāzı̄ is perhaps the more philosophically
11 interesting of the pair, with his ingenious criticisms of Avicenna. These are
12 usually made from within the Avicennan mindset, and thus provide an
13 ‘immanent critique’ that would be difficult for a modern scholar to duplicate. At
14 the same time, al-Rāzı̄ occasionally provides arguments that would help to
15 support Avicenna’s position (for instance, when he supplies a reason to think
16 that a celestial intellect will not grasp God just by grasping itself ). Because of the
17
highly dialectical nature of al-Rāzı̄’s commentary, it is not always easy to say
18
what al-Rāzı̄’s own position on the various issues would be. To establish this
19
would require reading further in his voluminous corpus, something I have not
20
attempted here. As for al-Tūsı̄, he provides a useful balance to the commentary
21 ˙
of al-Rāzı̄. More often than not, he has powerful counter-arguments to offer on
22
Avicenna’s behalf. In my admittedly limited experience, he also seems to be
23
more helpful in explaining the wording and structure of the Avicennan text.
24
In general, the main value of the commentators is their careful attention to
25
the structure, premises and potential weaknesses of Avicenna’s arguments. This
26
is indispensible in the case of the Pointers, which was written in a deliberately
27
allusive way such that, for instance, additional premises often need to be
28
29
supplied by the reader. The commentators also expose links between different
30
parts of the Pointers and, in al-Rāzı̄’s case, the relationship between Avicennan
31
arguments and parallel disputes in Islamic theology. On the specific issues I have
32
examined here, perhaps the most philosophically significant aspect of the
33
commentaries is their attention to the relationship between an intellect and its
34 object. As we have seen, al-Rāzı̄ frequently exploits the potential parallel
35 between this relation and other types of inherence – for instance, the inherence
36 of a property in a physical object. Al-Tūsı̄ makes efforts to specify the kind of
˙
37 inherence involved in intellection. But, in my judgement at least, he is not able
38
39 50 Admittedly, one might infer it from Aristotle’s rejection of Platonic Forms and embrace
of separate intellects; but Avicenna seems to innovate in arguing for it as a distinct
40
principle. Perhaps De anima III.5 might be thought to endorse the claim in question.
41 But, given the history of contentious debate over the meaning of III.5, it seems fair to say
42 that whatever it is that Aristotle is endorsing in that chapter he is not doing so clearly.
1 to show against al-Rāzı̄’s criticisms that any object of intellection must be

2 capable of being a subject of intellection when unimpaired by matter. More
3 successful, I think, is al-Tūsı̄’s analysis of self-intellection, in which the subject
4 and object of intellection˙ completely coincide without any need for inherence.
5 This allows him to say precisely how God’s thinking is, and is not, like that of
6 humans. If the topic pursued here is a typical case, there is good reason to have
7 both of these commentaries at hand while reading Avicenna, just as it is
8 becoming increasingly common to have Aristotle’s Greek commentators at hand
9 while reading Aristotle.
10
11
12
Bibliography
13
P. Adamson, Non-Discursive Thought in Avicenna’s Commentary on the Theology of
14 Aristotle, in Interpreting Avicenna: Science and Philosophy in Medieval Islam, ed.
15 J. McGinnis, Leiden: Brill, 2004, pp. 87 – 111.
16 –– , On Knowledge of Particulars, Proceedings of the Aristotelian Society, 105, 2005,
17 pp. 273 – 94.
18
–– , Knowledge of Universals and Particulars in the Baghdad School, Documenti e Studi
sulla Tradizione Filosofica Medievale 18, 2007, pp. 141 – 64.
19 P. Adamson and P.E. Pormann, Aristotle’s Categories and the Soul: an Annotated
20 Translation of al-Kindı̄’s That There Are Separate Substances, in The Afterlife of the
21 Platonic Soul. Reflections of Platonic Psychology in the Monotheistic Religions, eds J.M.
22 Dillon and M. Elkaisy-Freimuth, Leiden: Brill, 2009, pp. 95 – 106.
23
Aristotle, Aristūtālis fı̄ l-nafs, ed. A. Badawı̄,Cairo: Maktabat an-Nahd. a al-Mis. riyya,
1954. ˙ ˙
24
˘
Avicenna, al-Išārāt wa-l-tanbı̄hāt ma a Šarh Nası̄r al-Dı̄n al-Tūsı̄, 4 vols, ed. S. Dunyā,
25 ˙
Cairo: Dār al-Ma ārif, 1377/1957 – 1380/1960.˙ ˙
˘
26 –– , De anima, ed. F. Rahman, London/New York/Toronto: Oxford University Press,

27 1959.
28 –– , Livre des directives et remarques, transl. and ed. A.-M. Goichon, Beirut/Paris:
Commission Internationale pour la traduction des chefs d’œuvres/Vrin, 1951.
29
–– , The Metaphysics of ‘The Healing’, transl. and ed. M.E. Marmura, Provo, Utah:
31 –– , Psychologie d’Ibn Sı̄nā (Avicenne) d’aprs son oeuvre aš-Šifā , transl. and ed. J. Bakoš,
˘
32 Beirut: Editions de Patrimoine Arabe et Islamique, 1982.

33 A. Bertolacci, The Reception of Aristotle’s ‘Metaphysics’ in Avicenna’s ‘Kitāb al-Shifā ’, ˘
Leiden: Brill, 2006.
34
–– , Avicenna and Averroes on the Proof of God’s Existence and the Subject-Matter of
35 Metaphysics, Medioevo, 32, 2007, pp. 61 – 97.
˘
36 D.L. Black, Knowledge ( ilm) and Certitude (yaqı̄n) in al-Fārābı̄’s Epistemology, Arabic
37 Science and Philosophy, 16, 2006, pp. 1 – 45.
38 –– , Avicenna on Self-Awareness and Knowing that One Knows, in The Unity of Science
39
in the Arabic Tradition, eds S. Rahman, T. Hassan, T. Street, Dordrecht: Springer,
2008, pp. 63 – 87.
40
–– , al-Fārābı̄ on Meno’s Paradox, in In the Age of al-Fārābı̄: Arabic Philosophy in the
41 Fourth/Tenth Century, ed. P. Adamson, London: The Warburg Institute, 2008,
42 pp. 15 – 34.
122 Peter Adamson
1 A. Elamrani-Jamal, De anima. Tradition arabe, in Dictionnaire des Philosophes antiques,

2 Supplement, ed. R. Goulet, Paris: CNRS, 2003, pp. 346 – 58.
3
D.N. Hasse, Avicenna’s ‘De anima’ in the Latin West. The Formation of a Peripatetic
Philosophy of the Soul, 1160 – 1300, London/Turin: The Warburg Institute/Nino
4 Aragno Editore, 2000.
5 J. Kaukua, Avicenna on Subjectivity, Phd Diss., Jyvskyl, 2007.
6 E.W. Lane, An Arabic-English Lexicon, London: Williams and Norgate, 1863 – 93,
7 reprinted Beirut: Libraire du Liban, 1968.
8
M.E. Marmura, Fakhr al-Din ar-Razi’s Critique of an Avicennan Tanbı̄h, in Historia
Philosophiae Medii Aevi, eds B. Mojsisch and O. Pluta, Amsterdam/Philadelphia:
9 Grner, 1991, pp. 627 – 37.
10 M.M. McCabe, Some Conversations with Plato: Aristotle, Metaphysics Z.13–16, in
11 Aristotle and the Stoics Reading Plato, eds V. Harte, M.M. McCabe, R.W. Sharples
12 and A. Sheppard, London: Institute of Classical Studies, 2010, pp. 73–100.
13
J. McGinnis and D.C. Reisman, Classical Arabic Philosophy: an Anthology of Sources,
Indianapolis: Hackett, 2007.
14 al-Rāzı̄, Fahr al-Dı̄n, al-Mabāhit al-mašriqiyya Eastern Studies, 2 vols, Tehran: Maktabat
15 ˘ 1966.
al-Asadı̄, ˙¯
16 –– , Šarh al-Išārāt wa-t-tanbı̄hāt, no ed., in Kitāb Šarhay al-Išārāt li-l-Hwāğa Nası̄raddı̄n
17 al-T˙ūsı̄ wa-l-Imām Fahraddı̄n al-Rāzı̄, Qum, 1983/4. ˙ ˘ ˙
˙
R. Wisnovsky, ˘
Avicenna’s Metaphysics in Context, London: Duckworth, 2003.
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
1
2
3
4 Essence and Existence. Thirteenth-Century Perspectives in
5 Arabic-Islamic Philosophy and Theology
6
7
Fahr al-Dı̄n al-Rāzı̄’s al-Mulahhas fı̄ al-hikma and the Arabic Reception
8 ˘ ˘˘ ˙
of Avicennian ˙
Philosophy
9
10 Heidrun Eichner
11
12
The al-Mulahhas fı̄ al-hikma, ‘Compendium on Philosophy’, by Fahr al-Dı̄n
13 ˘ ˘ ˙ can be
al-Rāzı̄ (d. 1210) ˙ identified as one of the most influential works
˘ in the
14
Arabic reception of Avicennian philosophy from the late thirteenth century
15
onwards. Al-Rāzı̄’s presentation of Avicenna’s arguments provides the
16
conceptual basis of the majority of later texts stemming both from the
17
philosophical and the theological tradition.1 Fahr al-Dı̄n al-Rāzı̄ is an Aš arite
˘
18
theologian whose importance for the interaction ˘ between philosophical and
19
theological traditions in the Post-Avicennian period has been pointed out
20
repeatedly.2 However, a detailed evaluation of the chronology, the context,
21
and the intellectual outlook of his various works remains a major challenge.
22
The complex problems surrounding these aspects of al-Rāzı̄’s œuvre need to
23
be kept in mind ; however, they will not be addressed in my present analysis.
24
Instead, this analysis will focus on some aspects of the role which al-Rāzı̄ has
25
played for shaping the reception and interpretation of Avicennian ontology.
26
27
In describing the al-Mulahhas fı̄ al-hikma as a presentation of Avicennian
philosophy, I do not mean ˘that ˘ ˙ in the
˙ al-Mulahhas al-Rāzı̄ always holds
28 ˘˘ ˙
29
1 Works in the philosophical tradition heavily influenced by the presentation of the al-
30
˘
Mulahhas fı̄ al-hikma include Sirāǧ al-Dı̄n al-Urmawı̄’s Matāli al-anwār, ‘Places of
31 ˘ ˘ ˙ Lights’,
Ascendant ˙ al-Abharı̄’s Kašf al-haqā iq, ‘Unveiling Truths’,
˙ his Muntahā al-
˘
32 afkār, ‘Outmost Aim of Thoughts’, ˙and his Tanzı̄l al-afkār, ‘Sending down of
˘
33 Thoughts’; al-Kātibı̄’s Hikmat al- ayn, ‘Philosophy [part] of the Core [of logical rules]’.
˙
˘
34 Theological works include al-Baydāwı̄’s Tawāli al-anwār, ‘Ascendant Lights’, Šams al-
˙ ˙
Dı̄n al-Samarqandı̄’s al-Sahā if al-ilāhiyya, ‘Divine Folios’, al-Īǧı̄’s K. al-Mawāqif,
˘
35 ˙ ˙
‘Book of Stations’, al-Taftāzānı̄’s Maqāsid al-maqāsid, ‘Intentions of Intentions’. On
36 these writings and the adaptation of the ˙ literary structure
˙ of the al-Mulahhas fı̄ al-
37 hikma by Muslim theologians in the environment of the Marāġa-Observatory ˘ ˘ ˙see my
˙Philosophical and Theological Summae, pp. 97 – 132 (Chapter IV, Hybrid forms in the
38
39 philosophical tradition) and 351 – 498 (Part III, Chapters XIII to XVI).
2 For this context and a comprehensive analysis of al-Rāzı̄’s ethical theories see Shihadeh,
40
Teleological Ethics, passim. Shihadeh’s study also discusses the role of individual writings
41 of al-Rāzı̄ (for a discussion of chronology see ibid. pp. 5 – 13). For this and a discussion
42 of al-Rāzı̄’s biography see also Griffel, Patronage, passim.
124 Heidrun Eichner
1 Avicennian positions. Rather, at many opportunities, he argues against

2 Avicenna, for example when he denies the concept of ‘mental existence’ or
3 when he holds that in God, as in contingent beings, essence and existence are
4 distinct. Even more serious divergences from Avicenna’s philosophy result
5 from the fact that al-Rāzı̄ applies a kind of ‘agenda’ in his presentation of and
6 commentaries on Avicennian teachings. In the course of a philosophical
7 argument (as in other texts) al-Rāzı̄ identifies a series of clearly distinct
textual units which deal with specific problems and questions (masā il).3
˘
8
9 Analyzing these questions, al-Rāzı̄ uses a terminology which frequently does
10 not render the conceptual framework of Avicenna’s philosophy faithfully.
11 Discussing Avicenna’s position, possible objections and problems implied in
12 the argument, al-Rāzı̄ relies on a unified terminology which partly obscures
13 the conceptual divergences of the positions discussed. On the one hand, this
14 frequently results in considerable simplifications of more complex philo-
15 sophical problems, simplifications which perhaps explain the overwhelming
16 success which al-Rāzı̄’s writings have had in the course of the later reception.
17 On the other hand, these simplifications also open perspectives on new
18 problems and allow solutions that had not been acceptable to the earlier
19 tradition.
20 As its title suggests, the al-Mulahhas fı̄ al-hikma – like al-Rāzı̄’s very
21 similar more comprehensive work al-Mabāh ˘ ˘ ˙ it al-mašriqiyya
˙ ‘Eastern Inves-
22 tigations’ – is an exposition of philosophical ˙ ¯topics and arguments. By and
23 large, these two works cover the issues dealt with in Avicennian works such as
24 the K. al-Šifā and the K. al-Naǧāt in the Ilāhiyyāt (‘Divine Science,
˘
Metaphysics’) and tabı̄ iyyāt (‘Physics’).4 However, their structure no longer

˘
25
26 ˙ these two disciplines but rather relies on a classification
uses the division into
27 of existents into ‘necessary’ beings (i. e., God, dealt with in kitāb 3) and
28 ‘contingent’ beings (dealt with in kitāb 2). The first kitāb is devoted to a
˘
29 discussion of ‘common things’ (al-umūr al- āmma), i. e., things which are
30 common both to necessary and contingent beings.5
31 In this structure, topics covered by Avicenna in the Ilāhiyyāt can be found
32 at various places of the al-Mulahhas. Central notions of Avicenna’s
33 ontological system, most notably those ˘ ˘ ˙relating to ‘existence’ and ‘essence’,
34 are discussed by al-Rāzı̄ in the context of the ‘common things’. The section
35
36 3 On this see my Philosophical and Theological Summae, pp. 61 – 80 (Chapter III,
37 Observations on Fakhr al-Dı̄n al-Rāzı̄’s Method).
38 4 See the concluding remark in the al-Mabāhith al-mašriqiyya: ‘Now, since God has
granted us to collect these problems relating˙ to physics and divine science (al-masā il
˘
39
˘
al-tabı̄ iyya wa-l-ilāhiyya) according to this order and refinement ( alā hādā al-tartı̄b
40 ˙
wa-l-tahd ¯ Fahr al-
ı̄b) in which we are not preceded by anyone, we conclude this book’,
41 ¯
Dı̄n al-Rāzı̄, al-Mabāhith al-mašriqiyya 2:557,10 – 12. ˘
42 5 On this see my Dissolving ˙ the Unity, passim.
Thirteenth-Century Perspectives in Arabic-Islamic Philosophy and Theology 125
1 on ‘common things’ consists of five chapters, devoted to (1) existence, (2)

2 essence, (3) unity and multiplicity (4) contingent and necessary (5) pre-
3 eternal and temporally originated. Here, I shall discuss mainly how two
4 paragraphs (paragraphs 2 and 3) from the beginning of the chapter on
5 ‘existence’ are reflected in some closely related texts from writings by Atı̄r al-
¯
6 Dı̄n al-Abharı̄ (d. ca. 1265) and his student al-Kātibı̄ al-Qazwı̄nı̄ (d.
7 ca.1276).6 These two authors represent the earliest phase of the reception of
8 al-Rāzı̄’s al-Mulahhas. In the course of the thirteenth century, the
9 ˘˘ ˙
‘illuminationist’ philosophy of Šihāb al-Dı̄n al-Suhrawardı̄ emerges as an
10 alternative to Avicenna’s teachings in a way which has not yet been clarified.
11 Although we certainly cannot label al-Abharı̄ as a representative of an
12 Illuminationist tradition in the sense of a well-defined tradition of
13 intellectual affiliation, it will be seen that al-Abharı̄’s discussion is marked
14
by a familiarity with al-Suhrawardı̄’s writings and is partly influenced by his
15
arguments. The interaction between al-Rāzı̄’s presentation with al-Suhra-
16
wardı̄’s views (much more than the original wording of Avicenna’s writings)
17
motivates al-Abharı̄ and al-Kātibı̄ to elaborate a complex ontological and
18
epistemological discussion.
19
20
21
Al-Rāzı̄ and Avicenna’s Ontology :
22
23
The Discussion on mušārakat al-wuğūd
24
25
A most important feature of al-Rāzı̄’s attitude towards Avicenna’s ontology in the
26
al-Mulahhas is that he denies the very concept of ‘mental existence’.7 Further he
˘ ˘ ˙existence is something which is shared by all existents (see text I, f. 2;
asserts that
27
28 § 1), and that this requires that both in contingent beings and in the Necessary of
Existence existence is superadded (zā id) to essence (see text I, f. 3 and text I, f. 4;
˘
29
30 § 2 and § 3, the text of the al-Mulahhas contains further arguments regarding the
Necessary of Existence not translated ˘ ˘ ˙ here). According to al-Rāzı̄, only two
31
32 possible theories on the relation between essence and existence can be qualified as
consistent: (1) Those who assume that ‘existence’ is superadded (zā id) to ‘essence’
˘
33
34 assume that existence is shared (muštarak) – this is the position held by al-Rāzı̄;
35 while (2) those who assume that essence and existence are identical have to assume
36
37 6 Works by al-Abharı̄ referred to in this study include the following: Kašf al-haqā iq MS
˘
Tehran, Maǧlis-i Šūrā-yi Millı̄ 2752, pp. 1 – 212; Marāsid al-maqāsid MS ˙ Istanbul,
38
˙ ˙
Serez 1963; Muntahā al-afkār fı̄ ibānat al-asrār MS Tehran, Maǧlis-i Šūrā-yi Millı̄
39
2752, pp. 213 – 358; R. al-masā il MS Istanbul, Ragıp Paşa 1461. Among al-Kātibı̄’s
˘
40
˘
writings, the Ǧāmi al-daqā iq is to be mentioned: al-Kātibı̄ al-Qazwı̄nı̄, Alı̄ b. Umar

˘
˘
˘
41
˘
Naǧm al-Dı̄n, Ǧāmi al-daqā iq MS Paris, Bibliothque nationale 2370.

˘
42 7 For a discussion of some of al-Rāzı̄’s arguments see my ‘Knowledge by Presence’.

126 Heidrun Eichner
1 that existence is not shared between existents. This is the case because – according
2 to his reasoning – an identity of essence and existence together with the
3 assumption that existence is being shared would make it impossible that existents
4 can be differentiated from each other: Essence, being identical to existence would
5 have to be shared as well.
6 While earlier theories regarding the status of non-existent entities in Islamic
7 rational theology (kalām) provide one additional context in which this discussion
8 is to be placed, Avicennian concepts constitute the framework for al-Rāzı̄’s
9 arguments. Al-Rāzı̄ interprets the relation between Avicenna’s concepts ‘essence’
10 and ‘existence’ by a terminology that he also uses in his analysis of the theory of
11 divine attributes. Al-Rāzı̄’s notion of mušārakat al-wuğūd is very much influenced
12 by this: Using the Avicennian essence-existence distinction for describing
13 ‘existence’ in a framework of attribution, al-Rāzı̄’s understanding of existence’s
14 being shared does not include gradations of existence. Mušārakat al-wuğūd is not
15 to be identified with homonomy of existence.
16 The arguments provided by al-Rāzı̄ dominate the later tradition as well.
17 However, by distinguishing between the existence of things ‘in the mind’ and ‘in
18 external reality’, other authors use more differentiated argumentative strategies.
19 Al-Rāzı̄’s view that not only in contingent beings but also in the Necessary of
20 Existence essence and existence must be distinct has not gained influence in the
21 course of later reception. Moreover, already during the earliest phase of reception
22 of al-Rāzı̄’s analysis, al-Abharı̄ returns to accepting ‘mental existence’. By
23 integrating al-Suhrawardı̄’s views regarding the ontological status of mental
24 constructs he establishes an elaborate conceptual framework. Al-Abharı̄, however,
25 is also the first author who – when proposing a classification of positions
26 regarding the relation of essence and existence – neglects more subtle conceptual
27 and terminological considerations. He establishes a pattern which derives from al-
28 Rāzı̄’s schema and which is followed in the later philosophical and theological
29 tradition up to the Ottoman period. The following tables show how al-Abharı̄
30 and a thirteenth century theological author, Šams al-Dı̄n al-Samarqandı̄ describe
31 pertinent positions. As I am going to discuss, philosophical positions interact also
32 with an interpretation of the ontological position of theological schools, i. e. most
notably the Aš arite and the Mu tazilite positions.
˘
33
34 The nexus between the notion of mušārakat al-wuǧūd and the essence-
35 existence distinction remains dominant in the ontological expositions of later
36 authors. Another context in which the discussion of mušārakat al-wuǧūd is
37 placed is whether ‘existence’ encompasses both essentially contingent and
38 essentially necessary beings – in other words, whether we can label God as
39 ‘existent’ or whether he rather stands above existence.
40
41
42
1 Table 1: Positions as described by al-Abharı̄:a)

2 essence and essence and mental existence is
3 existence are existence are existence shared
4 identical in identical in affirmed
5 contingent beings necessary being
6 al-Rāzı̄ – – – +
7
Ibn Sı̄nā – + + +
8 al-Abharı̄ in the Kašf
9 al-haqā iq
˘
10
al-Abharı̄ in the + + (+) existence as
11 Muntahā al-afkār ‘existence is a mental
12 al-Suhrawardı̄ (as a mental construct is
13 quoted by al- construct’ shared
14 Abharı̄) between
15
Abū al-Hasan al- things in the
Aš arı̄ ˙
˘
mind
16
˘
17 ‘The Mu tazila’ –
18 a)
˘
For al-Aš arı̄ and the Mu tazila see the R. al-masā il quoted in nn. 15 – 20; for al-Rāzı̄
˘
19 and Ibn Sı̄nā see text II, l. 3 [E] (§ 13); for al-Suhrawardı̄ see text II, l. 3 [D] (§ 12); for
al-Abharı̄ in the Kašf al-haqā iq see text II, l. 3 [E] (§ 12), for the Muntahā al-afkār see
˘
20
21
text III, l. 1, b. 2 (§ 17).
22
23 Table 2: Positions as described by Šams al-Dı̄n al-Samarqandı̄:a)
24 essence and essence and mental existence is
25 existence are existence are existence shared
26 identical in identical in affirmed
27 contingent beings necessary being
28 those who verify – – + +
29 things (al-
30 muhaqqiqūn)
˙
31 philosophers – + [‘some –
32 among the
˘
al-Aš arı̄ + + later ones –

33 Abū al-Husayn al- deny mental
34 Basrı̄ (a ˙Mu tazilite
˘
˙ existence’]
35 theologian)
36 a)
See al-Samarqandı̄, al-Sahā if al-ilāhiyya (‘The Divine Folios’), pp. 72 – 82. al-Sa-
˘
37 ˙ ˙ influence al-Īǧı̄’s K. al-Mawāqif which in its turn be-

marqandı̄’s al-Sahā if al-ilāhiyya
˘
38 ˙ ˙
comes one of the most influential theological summae in the Sunnı̄ world. On al-
39 Samarqandı̄ see ‘al-Samarqandı̄’ in EI2 and my Philosophical and Theological Summae,
40 pp. 379 – 424 (Chapter XIV, Shams al-Dı̄n al-Samarqandı̄ and 13th-century kalām).
41
42
128 Heidrun Eichner
1 Breaking up al-Rāzı̄’s Matrix of Distinctions

2
3 Arguing against the matrix set up by al-Rāzı̄ remains an important motivation for
4 later authors to develop a more refined conceptual framework. As pointed out
5 supra, the first major step in this context is the reception of al-Suhrawardı̄’s
6 philosophy by Atı̄r al-Dı̄n al-Abharı̄ and al-Kātibı̄ al-Qazwı̄nı̄. An analysis of the
7 relevant passages ¯from works by these two authors reveals that al-Abharı̄ modifies
8 his earlier views (as laid out in the Kašf al-haqā iq) under the influence of arguments
˘
9 associated with al-Suhrawardı̄, a trend which ˙ is continued by his student al-Kātibı̄
10 al-Qazwı̄nı̄.
11 In the Kašf, presumably the earliest among al-Abharı̄’s writings introduced here,
12 al-Abharı̄ does not accept two features of al-Rāzı̄’s ontology but opts for the
13 ‘Avicennian’ position instead: al-Abharı̄ affirms mental existence, and he assumes an
14 identity of essence and existence in the Necessary of Existence (see text II, l. 3 [E];
15 § 13). His discussion of mušārakat al-wuǧūd and that existence is superadded (zā id)
˘
16 to essence follows al-Rāzı̄’s course of argument. At the end of his discussion of
17 essence and existence in contingent beings, al-Abharı̄ reports al-Suhrawardı̄’s
18 argument that a distinctness of essence and existence would require an infinite series
19 of hypostasized existences (see text II, l. 3 [D]; § 12 and text II, l. 3 [E]; § 13).
20 Al-Suhrawardı̄’s argument is dismissed by al-Abharı̄ in the Kašf while in the
21 al-Muntahā he adopts this position (on this see infra; see text III, l. 1, b. 2;
22 §§ 17 – 18).
23 Fundamental divergences between the ontological system of al-Rāzı̄’s al-
24 Mulahhas and al-Abharı̄’s Kašf are reflected by al-Abharı̄’s discussion of arguments
25 ˘ ˘ ˙from his return to an Avicennian position while retaining the basic layout
resulting
26 of the exposition of the al-Mulahhas. Al-Abharı̄ adduces arguments for the concept
27 of ‘mental existence’, based on ˘ ˘ ˙the possibility of conceptualizing impossible
28 8
entities. In part, these arguments were already known and discussed, though
29 rejected, by al-Rāzı̄.9
30 The relations of essence and existence in contingent beings and in the
31
Necessary of Existence constitute two different contexts in which new
32
arguments are developed and tested. Most notably, the Kašf contains a series
33
of eight objections against the very conception of the Necessary of Existence
34
which are refuted afterwards by al-Abharı̄ (see text II, L3 [A] and [B]; §§ 7 – 9).
35
This series of (albeit hypothetical) objections comes very close to arguing against
36
the existence of the Necessary of Existence, to the best of my knowledge an
37
absolutely singular exception in the Arabic-Islamic philosophical reception of
38
Avicenna’s theological ontology. Most of these arguments center on the problem
39
40
8 See e. g. al-Abharı̄, Kašf al-haqā iq, ‘fı̄ itbāt al-wuǧūd al-dihnı̄’, pp. 110 – 11 (not translated
˘
41 here). ˙ ¯ ¯
v r
42 9 See al-Rāzı̄, al-Mulahhas fı̄ al-hikma, fols 78 ,17 – 79 ,21.
˘˘ ˙ ˙
1 of how ‘existence’ can be instantiated in concrete beings when assuming that

2 some of these concrete beings are essentially necessary while others are
3 essentially contingent. Al-Abharı̄ complements the refutation of these objections
4 by an argument which is an elaboration of Avicenna’s ‘ontological’ proof for
5 God’s existence (see text II, l. 3 [C]; § 10, other versions of this proof are not
6 translated here).10 This argument, too, is further elaborated by al-Abharı̄ and al-
7 Kātibı̄ in later writings. Al-Kātibı̄’s exchange of letters with Nası̄r al-Dı̄n al-Tūsı̄
8 ˙
on this topic has gained considerable prominence in the later tradition. 11
In˙the
9 passage from the Kašf translated here, based on Avicenna’s ‘ontological’ proof,
10 al-Abharı̄ distinguishes the ‘totality of contingent beings’ vs. the ‘essentially
11 necessary’. Al-Abharı̄’s remarks in the Kašf show that his understanding of
12 Avicenna’s ontological proof (‘the proof of the philosophers’) is influenced by al-
13 Rāzı̄’s rejection of this proof. As we further hear, al-Rāzı̄’s rejection is connected
14 to his rejection of the identity of essence and existence in the Necessary of
15 Existence (see text II, l. 3 [C]; § 11).
16 In later writings by al-Abharı̄, the influence of al-Suhrawardı̄’s criticism of
17 the essence-existence distinction and his insistence on the importance of mental
18 activity for constituting the being-there of mental constructs gains momentum
19 (among the texts translated here these are the Muntahā al-afkār, ‘Outmost Aim
20 of Thoughts’ and the K. al-Šukūk, ‘Book of Doubts’). This development
21 manifests itself on the level of terminology, not so much by al-Abharı̄ and al-
22 Kātibı̄ adopting al-Suhrawardı̄’s terminology but rather by their readiness to
23
break up the clear-cut distinctions set up by al-Rāzı̄. But also apart from al-
24
Suhrawardı̄’s influence, returning to the ‘Avicennian’ concept of ‘mental
25
existence’ after it had been rejected by al-Rāzı̄, both al-Abharı̄ and al-Kātibı̄ are
26
forced to present more elaborate theories of the constitution of ‘mental
27
existence’ which integrate the contemporary debate critical of this concept.
28
Al-Abharı̄ and al-Kātibı̄ discuss relevant arguments primarily in the context
29
of the relation of essence and existence in contingent beings, because for them in
30
contingent beings – other than in the Necessary of Existence – essence and
31
existence are not simply identical. Following al-Suhrawardı̄, they point out that
32
an identity in external existence can go along with a distinctness in the mind.
33
The Kašf al-haqā iq and the Marāsid al-maqāsid (‘Lookouts of Intentions’; not
˘
34
translated here,˙ cf. note 16) are ˙familiar with ˙ al-Suhrawardı̄’s criticism of al-
35
Rāzı̄’s position regarding the relation of essence and existence but do not follow
36
it. The K. al-Šukūk likewise reports objections against al-Rāzı̄’s position. But
37
38
since the K. al-Šukūk as a whole is devoted to objections against the al-Mulahhas
˘˘ ˙
39
10 On Ibn Sı̄nā’s ‘ontological’ proof and its reception in the 13th century see Mayer, Burhān
40
al-Siddı̄qı̄n.
41 ˙ exchange of letters has been edited under the title Mutārahāt falsafiyya bayna Nası̄r
11 This
42 al-Dı̄n al-Tūsı̄ wa-Naǧm al-Dı̄n al-Kātibı̄. ˙ ˙ ˙
˙
130 Heidrun Eichner
1 it is not easy to determine with sufficient reliability whether the criticism in the
2 K. al-Šukūk represents al-Abharı̄’s own philosophical position while writing the
3 work, or whether the arguments are rather motivated by the overall critical
4 outlook of the book. The Muntahā al-afkār, however, in the two chapters ‘On
5 Existence’ and ‘On Essence’, accepts the arguments of al-Suhrawardı̄. In the Kašf
6 al-haqā iq al-Abharı̄ had reported these arguments and dismissed them. Now, in
˘
7 the˙ Muntahā al-afkār he adopts them (see text III, l. 1, b. 1 – 2; §§ 15 – 18).
˘
8 This approach is also adopted by al-Kātibı̄ in his Ǧāmi al-daqā iq (‘What
˘
9 Collects Subtleties’) (see text V, f. 2; §§ 20 – 21).
10 A problem is the evaluation of al-Abharı̄’s position in those of his writings
11 which do not follow the Rāzı̄an model. These works contain no comparable
12 discussion of the essence-existence relation, in particular the four works in the
13 MS Kçprl 1618 which al-Kātibı̄ has copied and studied with al-Abharı̄ in the
14 years 628/9 H. (1228/9 AD).12 To some extent, in these works a discussion of
15 ‘universal’ and ‘particular’ seems to replace the discussion of the essence-
16 existence relation. In some passages, these four works reveal an influence of
17 Suhrawardian arguments in other contexts.13 Therefore I tend to assume that in
18 these works the absence of a discussion of the essence-existence distinction
19 following the Rāzı̄an model is the result of al-Abharı̄ emancipating himself from
20 al-Rāzı̄’s position. This would also be in accordance with al-Kātibı̄’s familiarity
21 with al-Suhrawardı̄’s criticism, and with his acceptance and further elaboration
22
˘
of al-Suhrawardı̄’s position in his Ǧāmi al-daqā iq (see text V, f. 2; §§ 20 – 21).

˘
23 Following this line of an argument e silentio one might likewise argue that the
24 absence of a discussion of the essence-existence distinction in al-Abharı̄’s
25
˘
Hidāyat al-Hikma and al-Kātibı̄’s Hikmat al- ayn (a very popular compendium,
26 not translated˙ here) is to be explained
˙ by their increasing distance from the
27
˘
position of the al-Mulahhas. This would be the case although the Hikmat al- ayn
28
traces many features of˘ ˘its˙discussion immediately to the al-Mulah ˙ has. Further
29
clarification and a more precise understanding of the chronological ˘ ˘ ˙ develop-
30
ment of al-Abharı̄’s position regarding the constitution of ‘mental existence’ and
31
his analysis of the essence-existence relation is to be expected from an evaluation
32
of his theory of universals. His theory of universals is discussed not only in the
33
context of the Ilāhiyyāt but in his logical writings as well.
34
35
12 The works contained in the MS bear the titles Bayān al-asrār, ‘An Explanation of Secrets’,
36
˘
Talhı̄s al-haqā iq, ‘A Compendium of Truths’, Kitāb al-Matāli , ‘Book of Places of

˘
37 ˘˙
Ascendants’, Zubdat al-haqā iq, ‘Cream of Truths’. The MS˙ Kçprl 1618 contains
˘
˙ by the hand of al-Abharı̄ for al-Kātibı̄ at the end of each of

reading certificates signed
38
39 these works.
13 This is the case, for example, in the discussion of God’s knowledge of particulars: In the
40
Bayān al-asrār and in the Zubdat al-asrār al-Abharı̄ displays a familiarity with
41 Suhrawardian arguments. On the underlying Suhrawardian arguments see my
42 ‘Knowledge by Presence’.
1 The explicit quotations from al-Suhrawardı̄’s K. al-Talwı̄hāt in the Kašf al-

˙
haqā iq and their acceptance in the Muntahā al-afkār establish an immediate
˘
2
˙
3 textual link between al-Abharı̄ and al-Suhrawardı̄. In particular the concluding
4 remark in the Kašf shows that the positions of Ibn Sı̄nā, Fahr al-Dı̄n al-Rāzı̄ and
5 al-Suhrawardı̄ are the major points of reference for al-Abharı̄.˘ In the Kašf, he
6 explicitly sides with Ibn Sı̄nā (see text II, l. 3 [E]; § 14).
7
8
9 Philosophical Positions and Theological Affiliations
10
11 An evaluation of the elaboration of philosophical arguments, however, has to
12 take external factors into consideration as well. Most notably, there is the
13 problem of how these arguments relate to theological positions. As we posses
14 few precise information regarding the chronological order of most writings, it is
15 often not easy to determine whether differences between positions held in
16 various works reflect a development in the thought of an author, or whether
17 varying circumstances and conventions (i. e. whether a work belongs to a
18
theological or a philosophical context) has influenced the author’s exposition.
19
The case of al-Abharı̄’s Marāsid al-maqāsid and his R. al-masā il can show
˘
20 ˙ ˙
how al-Abharı̄’s adaptation of Suhrawardian arguments is situated in a more
21
complex context of intellectual affiliations. Thus, the Marāsid al-maqāsid
22 ˙ ˙
describe the arguments for the identity of essence and existence in external
23
reality as one of the arguments of the ‘natural philosophers’ against the ‘basic
24
rules of the divine [philosopher]’ – groups which so far cannot be identified as
25
associated with known individual authors. Possibly this refers rather to the
26
27
discipline mentioned.14 In addition to philosophical affiliations, theological
schools, most notably the mutually antagonistic Aš arites and Mu tazilites, are
˘
28
˘
referred to in the debate. Thus, in the R. al-masā il al-Abharı̄ portrays the
˘
29
30
position of an identity of essence and existence as the position of the founder of
the Aš arite school Abū al-Hasan al-Aš arı̄ (d. ca. 935) – an interpretation which
˘
31 ˙
32 might have constituted an important motivation for al-Abharı̄ to integrate al-
33 Suhrawardı̄’s criticism of the contemporary interpretation of Avicennian
34 philosophy in his philosophical interpretation. He portrays his own ontological
position as the position of Abū al-Hasan al-Aš arı̄ and contrasts it with a
˘
35
˙
Mu tazilite position which incidentally is identical to the ontological position
˘
36
laid out by (the Aš arite) Fahr al-Dı̄n al-Rāzı̄ in the al-Mulahhas. Thus al-Abharı̄
˘
37
presents his interpretation ˘of Avicennian positions through ˘ ˘ ˙the prism of al-
38
39 Suhrawardı̄’s arguments. He presents his detour from al-Rāzı̄’s ontology not as
40
41 14 Marāsid al-maqāsid fol. 33v : ‘al-marsad al-hāmis […] yahtawı̄ alā al-i tirādāt allatı̄
˘
˙ bihā min ˙ǧihat al-hukamā al-t˙abı̄ iyyı̄n

˘ alā qawā id ˙al-ilāhı̄’. ˙
˘
42 yuqdah
˘
˙ ˙ ˙
132 Heidrun Eichner
1 an antagonism between a ‘philosophical’ position and a kalām-position but as a

return to orthodox Aš arite teachings set against Mu tazilism.
˘
2
3 This can be seen from the introductory statements to relevant questions in
4 the Marāsid al-maqāsid which refer to positions which turn out to be identical
5
˙
to the position ˙
as depicted in some of al-Abharı̄’s philosophical writings (e. g. in
6 the Muntahā al-afkār). Thus, at the beginning of the second question, al-Abharı̄
7 states:
8
˘
The Imām Abū al-Hasan al-Aš arı̄ teaches that the existence of every contingent
9 being is identical to˙ its external essence. The Mu tazila says that existence is
˘
10 superadded to the external essence.15
11
In a similar vein, al-Abharı̄ describes the assumption of ‘mental existence’ as
12
˘
supporting the Aš arite position that ‘what is not existent is not a thing’ –
13
otherwise essences would have to be persistent in outward reality with existence
14
later being superadded to them.16 al-Abharı̄ further points out that the
15
assumption of an identity of essence and existence in God is the position of al-
16
Aš arı̄.17 Applying Avicennian terminology and analytical categories to the
˘
17
interpretation of the position of authors – most notably theologians – predating
18
Avicenna is an anachronism typical of how the amalgamation of theological and
19
philosophical positions in the thirteenth century takes place.
20
As we have seen, in the context of these ontological problems, al-Abharı̄ uses
21
Avicennian positions for interpreting the Aš arite position and for distancing
˘
22
˘
himself from what he describes as Mu tazilı̄ positions. In part, al-Abharı̄ regards

23
˘
the Aš arite position and arguments as identical to that expounded in his (later)
24
philosophical writings.
25
The complexity of the situation regarding the evaluation of intellectual
26
affiliations in al-Abharı̄’s writings may be highlighted by another very similar
27
case. This is likewise a problem heavily disputed between the philosophers and
28
the theologians, i. e. whether atomism or hylomorphism is to be accepted as
29
basic ontological explanation underlying physical reality. Typically, atomism is
30
described as the position of the mutakallimūn who oppose (Aristotelian)
31
hylomorphism. R. al-masā il, question (11) ‘On the atom’ first gives arguments
˘
32
for atomism, then arguments against it. Then, al-Abharı̄ concludes his
33
enumeration of anti-atomist arguments by the statement:
34
35
36
37
15 al-Abharı̄, R. al-masā il, fol. 366r,2 – 4.
˘
38
˘
16 See the discussion of mas ala (3) fı̄ anna al-ma dūm laysa bi-šay (what is non-existent is
˘
39
not a thing), R. al-masā il, fols 366v,15 – 367v,5. This runs very much counter to standard
˘
40
expositions of Avicennized Aš arism in late 13th-century kalām.
˘
41 17 See the discussion of mas ala (5) fı̄ anna wuǧūd al-bārı̄ nafs haqı̄qatihı̄ (the existence of
˘
42 the creator is identical to his core-essence), R. al-masā il, fol.˙ 368v,11 ff.
˘
1 This question rightly causes one to hesitate (yastahiqqu al-taqawwuf fı̄hā) – or to

2 admit that the atom has to be negated because of ˙these aspects.18
3 Here, al-Abharı̄ is ready to accept the arguments against atomism, and he does
4 not describe the opposition between pro- and anti-atomist positions as an
5 antagonism between mutakallimūn and the philosophers. Nevertheless, in the
6
next question (12) he seems to prefer a rejection of hylomorphism. At least, this
7
is what the title of the question ‘On rejecting hylomorphism’ suggests. In his
8
philosophical writings, however, al-Abharı̄ rather defends hylomorphism.19 He
9
associates hylomorphism with the philosophers (falāsifa) and ascribes the
10
rejection of hylomorphism to the mutakallimūn. Again, al-Abharı̄ expresses his
11
doubts and hesitation at the end of the question:
12
13 This question, too, rightly causes one to hesitate because the arguments bring each
14 other to a standstill (li-ta ārud al-adilla). God is the one who provides guidance.20
˘ ˙
15 The problem of the ontological status of mental constructs and the antagonism
16 between atomism and hylomorphism represent ‘classical’ questions disputed
17 between philosophers and theologians. While by the thirteenth century these
18 problems have a centuries long history, another set of discussions touching upon
19 related basic ontological assumptions begins to emerge only at that time. These
20 are questions centering on how individuation (tašahhus) and concretisation
21 ˘ ˘ ˙ mental and external
˘
(ta ayyun) take place. The Avicennian distinction between

22 existence changes the role which (corporeal) matter plays for the conception of
23 individuation and concretisation. From this result very complex discussions in
24
later authors. For example, discussions of the ontological status of mental
25
abstractions which are described as concepts (and hence taken as concrete beings
26
existing in the mind) and their relation to other concrete beings, both in
27 ˘
external and mental reality, motivate later authors such as Sa d al-Dı̄n al-
28
Taftāzānı̄ (d. 1390) in his Maqāsid al-Maqāsid to highly ˙ complicated
29 ˙
elaborations. Interactions with ontological ˙ typically associated with
traditions
30
Islamic ‘mysticism’ are to be considered in this context as well. In addition to the
31
tradition associated with Šihāb al-Dı̄n al-Suhrawardı̄, this includes the tradition
32
˘
going back to Ibn al- Arabı̄. In the texts translated here, some elements of this
33
tradition are dealt with in the discussions of how God as the Essentially
34
Necessary of Existence relates to the instantiations of contingent beings.
35
36
Only a minor percentage of relevant texts has been published so far, and few
37
38
reliable translations of these highly complex documents exist. Already from this
39
18 al-Abharı̄, R. al-masā il, fol. 373v,14 – 15.
˘ ˘
40
19 For the Kašf al-haqā iq see for example the first book of the section of physics, pp. 169 –
41 73. ˙
42 20 al-Abharı̄, R. al-masā il, fol. 375r,13 – 14.
˘
134 Heidrun Eichner
1 material it can be discerned that the discussion reaches a level of complexity

2 which is mirrored on the level of terminology as well. The texts selected and
3 presented in translation here may serve to document some features of this
4 discussion in Arabic texts of the thirteenth century. The selection of texts centers
5 on the discussion of mušārakat al-wuğūd and its relation to the essence-existence
6 distinction, including some aspects of ‘ontological’ proofs for the Necessary of
7 Existence.
8 The al-Mulahhas fı̄ al-hikma by Fahr al-Dı̄n al-Rāzı̄ (Text I) provides the
9 ˘ ˙ discussions
starting-point for˘ the
˙
of later˘authors, al-Abharı̄’s Kašf al-haqā iq and
˘
10 ˙
his Muntahā al-afkār are important specimens (Texts II and III). Al-Abharı̄’s K.
11 al-Šukūk (Text IV) documents how closely he has studied the al-Mulahhas fı̄ al-
12 ˘˘ ˙
˘
hikma. Al-Kātibı̄’s Ǧāmi al-daqā iq (Text V) shows how al-Abharı̄’s arguments
˘
13 ˙
are continued and refined by his student al-Kātibı̄ al-Qazwı̄nı̄.
14
15
16 Primary Texts – Translations
17
18
(I) Fahr al-Dı̄n al-Rāzı̄, The Compendium on Philosophy (al-Mulahhas fı̄ al-
19
hikma,˘ MS Berlin, Staatsbibliothek MS Or. Oct. 629): ˘˘ ˙
20 ˙
21
(§ 1) fasl (2): On existence being shared (fı̄ anna al-wuǧūd muštarak) because of
22 ˙
two reasons:
23
waǧh (1): If we know that something contingent has a cause then we judge
24
that this cause exists. Our doubts (taraddud) whether it [i.e., this cause] is
25
‘necessary’ or ‘contingent’, or ‘substance’ or ‘accident’ does not affect our first
26
judgement. Then, when we believe ‘that it is necessary’, this belief ceases by our
27
28
belief ‘that it is contingent’. If existence were not common to all divisions (bayna
˘
˘
29
ǧamı̄ al-aqsām), it would cease when the belief of some specifics ceases ( inda
˘
30
zawāl i tiqād al-husūsiyyāt), in the same way as the belief about each single one
˙˙
31
of the specifics ˘ceases by the belief about another specific.
32 waǧh (2): The referent of negation (mafhūm al-salb) is one single referent
33 insofar as it is negation (salb). If what is contrary to it (al-muqābil lahū) were
34 affirmations with various referents (ı̄ǧābāt muhtalifa al-mafhūm) the exhaustive
enumeration (hasr) would be false. If they have ˘ a unified referent (muttahida l-
35
˙ ˙ ˙
36 mafhūm): q. e. d.
37
38 (§ 2) fasl (3): On existence being additional to the essences of contingent beings
˙
˘
(fı̄ anna al-wuǧūd zā id alā māhiyyāt al-mumkināt):

˘
39
40 The existence of blackness (sawād), for example, either is identical to its
41 being blackness (nafs kawnı̄hı̄ sawādan) or intrinsic to it (dāhilan fı̄hi) or
˘ because of
˘
42 extrinsic to it (hāriğan anhu). The first two possibilities are false

˘
1 several reasons. Some of these [reasons] pertain to both [possibilities, 1 and 2],
2 some of them pertain only to one.
3 Those which pertain to both of them are two aspects:
˘
4 waǧh (1): Intellection (ta aqqul) of a heptagon is possible while having
˘
5 doubts about its outward existence ( inda al-šakk fı̄ wuǧūdihı̄ al-hāriğı̄) – as it is
6 obvious – and about its mental existence (al-wuǧūd al-dihnı̄) – if˘ it [i.e., mental
7 existence] were to be affirmed (law tabata), (for even¯ if it follows from being
8 aware of it it does not follow that this¯ occurs as part of being aware of it (wa-in
˘
9 kāna lāziman li-l-šu ūr bihı̄ lakinnahū ġayr lāzim fı̄ al-šu ūr bihı̄). Therefore, for
˘
10 someone who knows a heptagon ( alima al-musabba ) it is possible to deny its
11 mental existence). What is doubted about is not identical to what is not doubted
12 about (al-maškūk fı̄hi laysa nafs ġayr al-maškūk fı̄hi), and it is not part of it (dāhil
13 fı̄hi). ˘
14 As to doubting about existence (tašakkuk fı̄ al-wuǧūd): If by this it is meant
15 doubting about existence being affirmed for existence (tašakkuk fı̄ tubūt al-
¯
˘
16 wuǧūd li-l-wuǧūd): This is impossible because ‘non-existence’ ( adam) and
17 ‘existence’ cannot be predicated of ‘existence’. If doubting about its being
18 present for an essence (fı̄ husūlihı̄ li-l-māhiyya) is meant, what we have said is to
19 be applied. ˙ ˙
20 waǧh (2): If we take blackness together with existence, under this condition
21 it is not susceptible (qābil) of non-existence; and vice versa. If we take it
22 regardless of ‘existence’ and ‘non-existence’, then it is susceptible of both of
23 them. Hence its ipseity (huwiyya) which is susceptible of both of them is
24 something different (muġāyir) from both of these opposite stipulations (al-
˘
25 qaydayn al-muta ānidayn).

26
27 (§ 3) Now the specific [aspects]:
28 waǧh (1): What falsifies existence being identical to essence is: If existence of
29 blackness (sawād) were identical to its being blackness, then whiteness would
30 not be sharing its existence, in the same way as it is not sharing its blackness-ness
31 (sawādiyya).
32 waǧh (2): If we were to say that ‘a substance ( ǧawhar) is existent’, this would
˘
take the same rank of bearing no information (fı̄ adam al-fā ida) as if we said
˘
33
34 that ‘a substance is a substance’.
35 waǧh (1): What falsifies existence being part of it is: If this were the case,
˘
36 then it would be the most general of [all] shared essential things (a amm al-
˘
37 dātiyyāt al-muštaraka). Hence it would be a genus ( ǧins), and the species (anwā )
38
¯which fall under it (dāhil tahtahū) would be distinguished from each other by
39
˘ ˙ the genus would form part of the nature of the
differences that exist. Therefore,
40 differentia, and its being differentiated from the species would require another
41 differentia, and so on to infinity. But the differentia which constitutes the
˘
42 division (al-fasl al-muqassim) has a cause ( illa) for its existence, and thus
˙
136 Heidrun Eichner
1 existence would have a cause for its existence. Hence, existence would have
2 another existence – this according to what Ibn Sı̄nā says. Then, the
3 differentiation of ‘the necessary’ from ‘the contingent’ would be by a differentia
4 which constitutes it, and hence ‘the necessary’ would be composed. Existence
5 would be constituting the things which fall under it (al-umūr al-mundariǧa fı̄hi).
6 If it were in itself independent from a substrate, it would be a substance, and it
7 would be part of an accident. Then, an accident would be a substance. This is
8 absurd.
9
10 (II) Atı̄r al-Dı̄n al-Abharı̄, Unveiling Truths regarding Settling Subtleties (Kašf al-
11 ¯
haqā iq fı̄ tahrı̄r al-daqā iq, MS Tehran, Maǧlis-i Šūrā-yi Millı̄ 2752):
˘
˘
12 ˙ ˙
13 (§ 4) lāmi (1): On its [= existence] being shared between existents (al-
˘
14 mawǧūdāt).
15 This may be clarified by various aspects:
16
˘
waǧh (1): We conceive the notion (natasawwar ma nā) of existence
17 ˙
intuitively (bi-l-badı̄ha), and we judge that it is true (naǧzam bi-sidqihı̄) for
18 ˙
each of the existing things. If this what is conceptualized (al-mutasawwar)
19 ˙
would not be common between them, the judgement that it is true (al- ǧazm bi-
20 21
sidqihı̄) for each single one would be impossible.
21 ˙
waǧh (2): We conceive the referent of existence and non-existence (mafhūm
22
˘
al-wuǧūd wa-l- adam) intuitively, and we pass the judgement that if the second
23
is false for something, then the first is true. If existence were not something
24
25
shared between existents, then from the falseness of the second for a thing there
˘
26
would not follow that the intellect judges that the first is true for it (ǧazm al- aql
˘
27
bi-sidq al-awwal alayhi), because then it would be possible that both of them
˙
28
are false.
29 waǧh (3): If the judgement that something is present among concrete
30 beings, then the judgement about its existence is there (hasala). If existence were
˙ ˙
31 to have two referents, then from the judgement of the mere fact that something
32 is among concrete beings would not follow that existence is true for it. For it
33 would be possible that another referent [of ‘existence’] would be false for it,
34 when there is the mere judgement that it is among concrete beings.
35
(§ 5) lāmi (2): On existence of contingent beings being additional (zā id) to
˘
36
37 their essences, because if it were not additional, then it would either be identical
38 to the essence or intrinsic to it.
39 Both possibilities (qismān) are false. The first [option is false] because of
40 several aspects:
41
42 21 Read: la-stahāla instead of li-stihāla.
˙ ˙
1 waǧh (1): Existence is shared between all existents, and no one of the
2 contingent essences is so.
3 waǧh (2): Under the condition of existence, the existence of a contingent
4 essence is necessary, but insofar as it is what it is (min haytu hiya hiya) its
5 existence is not necessary. Hence ‘essence under the condition ˙ ¯ of existence’ is
6 different (muġāyir) from ‘essence insofar as it is what it is’. Hence ‘existence’ is
7 different (muġāyir) from ‘essence insofar as it is what it is’.
8 waǧh (3): If existence were identical to the contingent essence, if we say that
9 ‘blackness is existent’, for example, and that ‘blackness is blackness’ and that ‘an
10 existent is existent’, this would have one meaning. The consequent22 is false
11 because the judgement of the truth of the last two [examples] realizes only the
˘
12 conception of that about which the predication is made ( unwān al-qadiyya),
13 other than the first. ˙
14
15 (§ 6) The second [option is false] because of several aspects:
16 If existence were part of essence, then – if the essence were simple – the
17 composition of simple things would follow. This is absurd. If it were composed,
18 the existence of an essence would be preceding it because it is necessary that the
19 part precedes the whole.
20 If someone says: ‘If existence were additional to the contingent essence, it
˘
21 would have a ipseity (huwiyya) among concrete beings (a yān) behind the ipseity
22 of the essence in which it inheres. Then, its substrate (mahall) would have an
23 existence which precedes it by existence, and essence would ˙ have existence
24 before its existence. This is absurd. For if it were additional, it would follow that
25 existence would be subsisting (qiyām) by something which is not existent’ – we
26 say: ‘We do not concede that it follows that the substrate of the ipseity of
27 existence has another existence. Why is it not possible that it precedes it by its
28 self (bi-nafs dātihā), not by yet another existence?’ As to his claim that the
29
¯
subsistence (qiyām) of existence by something which is not existent would
30 follow, we say: ‘We do not concede this. What follows is the subsistence of
31 existence by the essence while essence is existent. Yet, its existence is different
32 from it’.
33
(§ 7) lāmi (3): On establishing the Essentially Necessary.
˘
34
35 If none of the existences were essentially necessary, then the nature of
36 existence insofar as it is what it is either would be essentially independent from
37 something else, or it would be in need of it. The second is impossible, because if
38 it [i.e., the nature of existence] were essentially in need of something else it
39 would be essentially contingent. Inevitably it would need a cause which precedes
40 it by existence. Then, existence would precede it [i.e., the nature of existence]. If
41
42 22 Read: al-tālı̄ instead of al-tānı̄.
¯
138 Heidrun Eichner
1 it [i.e., the nature of existence] were identical to existence, something would

2 precede itself. This is absurd. If it [i.e., the nature of existence] were a specific
˘
3 existence (wuǧūd hāss), a concretised existence (wuǧūd mu ayyan) would precede
4 ˘ ˙˙ This is absurd. The first [option], too, is absurd because
the nature of existence.
5 if existence were essentially not in need of something else it would either follow
6 for it an existence which is concretised by itself or by something else, or nothing
7 at all would follow. In the first case, the species of existence would be restricted
˘
8 to one individual (naw al-wuǧūd munhasir fı̄ šahsihı̄). In the second case, and if
9 ˙ ˙ ˘ ˙ nature, then, is not in need of
they are not two things caused by a cause (for the
10 something else) follows that it is possible that the nature can be split from all
11 existences. This is absurd.
12
13 (§ 8) [A] It might be said: ‘Nothing is essentially necessary because of several
14 aspects’:
15
(1) If some among the existences were necessary, the nature of existence
16
either would be essentially not in need of something else, or it would be in need
17
of it. The second is absurd, as you have explained. The first, too, is absurd,
18
because if existence were essentially not in need of something else then from it
19
would follow either an existence which is concretised by itself, or by something
20
else, or nothing at all would follow. The first is absurd, as you have explained.
21
The second, too, is absurd. Otherwise it would follow that it is possible that the
22
nature of existence is split from both these existences. This is absurd.
23
(2) If some one among the existences were essentially necessary, its
24
25
concretisation would be either because of the nature of existence or not. The
26
first is absurd; otherwise the species of existence would be restricted to one
27
individual so that existence would not be shared. This is absurd. The second is
28
absurd; otherwise it would follow that the Necessary of Existence needs for its
29 concretisation something else. This is absurd.
30 (3) If concretised existence were essentially necessary its concretisation
31 would be either ‘something relating to existence’ or ‘something relating to non-
˘
32 existence’ (wuǧūdiyyan aw adamiyyan). The second is absurd because

33 concretisation is a part of what is concretised and what is concretised is
34 existent, so that its part is existent. The first is absurd because if it were
35 ‘something relating to existence’ it would be so because of a cause which
36 precedes it. If this cause were this existence itself it would be concretised before
37 its concretisation. If it were something else, the essentially necessary would be in
38 need of something else in its concretisation. This is absurd.
39 (4) If concretised existence were essentially necessary, then ‘existence insofar
40 as it is existence’ would be either essentially in need of something else, or not.
41 The second is absurd; otherwise no need would accede to it so that the existence
42 of contingent beings would not be in need. This is absurd. The first is absurd
1 because otherwise the essentially necessary would be ‘in need’ because of its
2 essence. This is absurd.
3 (5) Existence insofar as it is what it is either requires that it is connected to
4 an essence or it requires that it is not connected or it does not require one of
5 these two things. The second is absurd; otherwise the existence of contingent
6 beings would not be connected to their essences. This is absurd. The third is
7 absurd; otherwise both ‘connection’ and ‘being-abstracted from anything else’
8 (muqārana wa-taǧarrud) would be because of a cause, and then the essentially
9 necessary would be in need of something else in its abstraction. Hence the first
10 has to be taken into consideration, and so no one among the existences is
11 essentially necessary.
12 (6) If the concretised existence were necessary it would be self-subsistent. If
13 its subsistence by itself were by existence itself or by something which follows
14
from it, every existence would be abstract. This is absurd. If it were by
15
something which follows from it, its existence would be connected to some
16
essence while the opposite has been postulated. If it were by something separate
17
(mubāyin), the existence of the Necessary [of Existence] (al-wuǧūd al-wāǧibı̄)
18
would in its abstraction be in need of a cause which is distinct. This is absurd.
19
(7) If the essentially necessary were existence then every contingent being
20
would be attributed with something identical to the necessary existence (bi-mitl
21 ¯
al-wuǧūd al-wāǧib). This is absurd.
22
(8) If the essentially necessary were abstract existence, then the essentially
23
24
necessary would be composed out of existence and non-existence, i. e.
25
abstraction in the sense of ‘non-existence of its being connected with an
26
essence’. This is absurd.
27
28
(§ 9) [B] The answer:
29 (1) We say: We do not concede that if nothing of the existences follows
30 from existence, it follows necessarily that it is possible that the nature can be
31 split from both existences. This follows only if no concretised existence is
32 essentially necessary. Its necessity, however, is evident because we argue based on
33 this assumption.
34 (2) We say: We do not concede that if its concretisation were not because of
35 the nature of existence it would be because of a separate cause. This follows only
36 if concretisation were ‘something relating to existence’. Why do you say that it is
37 ‘relating to existence’? For according to us, to the nature of existence accede
38 different concretisations, some of them because of the essence which is
39 susceptible of them, and some of them because of the absence of something
40 susceptible. Those which accede to it because of the absence of something
41 susceptible are something ‘relating to non-existence’ which is ‘abstraction from a
42 substrate of inherence’ and ‘not-being-mingled with contingent beings’.
140 Heidrun Eichner
1 (3) We say: Why is it not possible that the concretisation is ‘something

2 relating to non-existence’? As to the statement, that ‘what is concretised is
3 existent, and concretisation is a part of what is concretised, so it must be
4 existent’, we say: If you mean by your statement that ‘what is concretised’ is two
5 existences and that existence accedes to what is concretised insofar as it is
6 concretised, then it is not permissible. For existence is predicated of this
7 existence like the species is predicated of its individuals. If you mean by it that
8 this existence is concretised existence we concede this. But why do you say that
9 the concretisation of what is judged to be a concretised existence is something
10 existent? This, because – according to us – the pure [nature] of existence (sirāfat
11 al-wuǧūd) is something by which this existence is concretised, and˙ it is
12 ‘something relating to non-existence’. It is true for pure existence that it is
13 essentially necessary according to its essence which is existence.
14 (4) We say: Why do you say that the need accedes to existence insofar as it is
15 what it is? It accedes only to specific existences. This, because the contingent
16 essences have the dispositions for specific existences so that on them emanate
17 these existences from the necessary existence. ‘What is in need of ’ is only the
18 concretised existence. ‘Existence insofar as it is what it is’ is essentially not in
19 need of something else.
20 (5) We say: Why do you say that the third alternative is absurd? Regarding
21 the statement that both connection and abstraction are because of a distinct
22 cause, we say: We do not concede this. This would follow necessarily only if
23 abstraction were ‘something relating to existence’ which is in need of a cause.
24 This is not the case. Rather this is a ‘stipulation relating to non-existence’ (qayd
˘
25 adamı̄) and it is not caused by a cause.

26 (6) From this results the answer to the sixth argument.
27 (7) We say: Why is it not possible that each of the contingent beings is
28 attributed by one of the individuals of the nature (mawsūf bi-fard min afrād al-
˙
˘
29 tabı̄ a)? This individual is according to its specificity essentially contingent. The
30
˙claim that it is absurd is not permissible, because there is no demonstration.
˘
31 Rather it is a mere assumption that it is unlikely (istib ād), nothing else.

32 (8) We say: We do not concede that composition must follow. This, because
33 ‘the essentially necessary’ is existence which is acceded by the non-existence
˘
34 ( adam) of an essence. ‘Non-existence’ is not a part of it but an accident, so that

35 it does not lead to composition.
36
(§ 10) [C] You must know that the philosophers (al-hukamā ) have explained
˘
37
38 that some existent is essentially necessary by another ˙method. They say: The
39 totality of contingent beings must have something which exerts an influence
(mu attir) from outside. What is outside of the totality of contingent beings is
˘
40
41
¯¯ necessary. The existence of the essentially necessary must be identical
essentially
42 to its essence because if it were additional it would be in need of the essence.
1 Then it would be essentially contingent, so that it would need something which

2 exerts an influence. If what exerts an influence on it were identical to this
3 essence, it follows that it precedes existence by existence because what exerts an
4 influence on existence must precede it by existence. If it were not identical to it
5 [i.e., the essence], the essentially necessary in its existence would be in need of
6 something else. This has to be deliberated [critically], for when they say that the
7 totality of contingent beings must have something which exerts an influence
8 outside of it, this is not permissible. This would need to be demonstrated.
9
10 (§ 11) The Imām [Fahr al-Dı̄n al-Rāzı̄] does not accept what they say that if
11 what exerts the influence˘ on it is identical to this essence it follows that it
12 precedes existence by existence. What the rejection rests on is that a contingent
13 essence insofar as it is what it is accepts existence, but it does not precede it by
˘
14 existence. It is possible that the agent (fā il) is like that. The philosophers say
15 that this is a sophistry (mukābara). We know by necessity that what exerts an
influence (mu attir) on the existence of something must precede it by existence.
˘
16
17 The essence which¯¯ is susceptible [of existence] does not exert an influence on
18 existence, and therefore it does not have to precede existence by existence.
19 He has already explained that some concretised existence is necessary of
20 existence, and that it is not connected to an essence. Otherwise it would be in
21 need of something else, and what is essentially necessary would be essentially
22 contingent. This is absurd.
23
24 (§ 12) [D] The author of The [Philosophy of] Illumination (sāhib al-išrāq) [al-
25 Suhrawardı̄] has another method regarding existence. He says˙ that ˙ the existence
26 of contingent beings in external [reality] is identical to the essences, for were
27 they distinguished from each other, the essence would have another ipseity, and
28 so both of them would be existent in external [reality]. Then, essence would
29 have another existence and existence would have another existence, and essence
30 would have yet another ipseity. Thus, both of them would be existent in external
31 [reality]. Hence it would follow that the essence has infinite existences.
32 Therefore, essence in concrete beings is one thing, and only the intellect splits
33 external essence into two things, ‘essence’ and ‘existence’, so that two forms are
34 there in intellect which correspond (mutābiq) to external essence. ‘Common
˙ place in concrete beings but it is
˘
35 existence’ (al-wuǧūd al- āmm) does not take

36 only in the mind. The existence which takes place is the existence pertaining to
37 the necessary (al-wuǧūd al-wāǧibı̄) which is abstract from essence. When it is
38 there in the intellect, intellect does not split it into two things, ‘essence’ and
39 ‘existence’, rather only existence results from it. Its existence among concrete
40 beings is not connected to some kind of essence. [This is so] because if it were
41 connected to some kind of essence this essence would be a universal essence so
˘
that it would have particulars in intellect (ǧuz iyyāt aqliyya). Then, the relation
˘
42
142 Heidrun Eichner
1 of essence to [all of ] them would be identical, and it would not require the
2 existence of one of them because this would be a preponderance without an
3 agent (tarǧı̄h bi-lā muraǧǧih). The essentially necessary would not be existent
˙ ˙
4 because of its essence, and the essentially necessary would not be essentially
5 necessary. This is absurd.
6
7 (§ 13) [E] This must be deliberated [critically] for we say: We do not concede
8 that if essence is distinct from existence and both of them exist, essence must
9 have another existence. Why is it not possible that the essence is existent by an
10 existence which is distinct from it while the existence of existence is identical to
˘
11 itself (al-māhiyya mawǧūda bi-l-wuǧūd al-mutamayyiz anhā wa-yakunu wuǧūd
12 al-wuǧūd nafs dātihı̄)? As to the statement that ‘the intellect splits the essence in
13 ¯
external reality into two things, “essence” and “existence”, and so two forms are
14 there in intellect which correspond to one thing, and that this is something
15
absurd because to one thing cannot correspond two different forms’, and further
16
as to the statement that ‘“common existence” does not take place in concrete
17
˘
beings’, this means that existence is a mental consideration (i tibār dihnı̄) which
18 ¯
is not there in concrete beings. This is not true. Otherwise, the nature of
19
existence would be ‘something relating to non-existence’ in concrete beings.
20
This is absurd.
21
The statement that ‘if existence pertaining to the necessary (al-wuǧūd al-
22
wāǧibı̄) were connected to some essence this essence would be a universal
23
24
essence which has intellectual particulars whose relations to it are identical and
25
the essentially necessary would not be existent by its essence’ is not permissible.
26
This would follow only if the relation of this essence to what provides its
27
external individualization (al-mušahhis al-hāriǧı̄) were identical to its relation to
˙ ˘
˘ ˘ you
28 the intellectual particulars. Why do say that this is like this? This would
29 need to be demonstrated.
30
31 (§ 14) The truth is what the Šayh [Ibn Sı̄nā] teaches, i. e. that existence is shared
32 among existents, that existence˘ in contingent beings is additional to their
33 essences, and that the existence of the essentially necessary is not connected to
34 some essence, as we have established. The Imām [Fahr al-Dı̄n al-Rāzı̄] disagrees
35 with him regarding his abstraction from essence and˘claims that it is connected
36 to it and that this essence is a cause for it and that it does not precede it by
37 existence.
38
39 (III) Al-Abharı̄, The Outmost Aim of Thoughts on Clarifying Secrets (Muntahā al-
40 afkār fı̄ ibānat al-asrār, MS Tehran, Maǧlis-i Šūrā-yi Millı̄ 2752):
41
(§ 15) lāmi (1): On existence.
˘
42
1 baht (1): It is widespread (mašhūr) that existence is common between all

2
˙ ¯ They argue for this by several aspects:
existents.
3 waǧh (1): The intellect divides existence into ‘necessary’ and ‘contingent’.
4 That on which the division is carried out (mawrid al-taqsı̄m) must be common
5 between the parts.
6 waǧh (2): ‘Non-existence insofar as it is non-existence’ is one referent.
˘
7 Therefore, its annihilation (raf ) is one single referent. ‘Annihilation of non-
8 existence’ is ‘existence’. Hence, existence is one referent.
˘
9 waǧh (3): If the judgement of the existence of an existent is there ( inda
10 husūl al- ǧazm bi-wuǧūd mawǧūd), and if we are believed that ‘it is necessary’,
11
˙then
˙ this belief ceases by the belief that ‘it is contingent’. The same [is the case]
12 if we are believed that ‘it is substance’. This ceases by our belief that ‘it is
13 accident’, but the belief ‘that it is existent’ does not cease. If existence were not
14 common between all parts, then the belief of the existence would cease when the
15 belief of the specifics ceases. All this is weak.
16
17 (§ 16) As to (1): We do not concede that the intellect divides existence into
18 ‘necessary’ and ‘contingent’. Rather, the intellect passes a judgement that every
19 existing essence is either ‘essentially necessary’ (wāǧiba bi-l-dāt) or ‘essentially
20 contingent’. This does not point to existence being common¯ between the two
21 things (al-amrayn).
22 As to (2): We do not concede that ‘non-existence insofar as it is non-
23 existence’ is one referent. Who is believed that the existence of every thing is
24 identical to its essence is believed that non-existence has various referents. Even
25 if we were to concede this, it is not possible that the annihilation of non-
26 existence as such (nafsuhū) is existence as such. Otherwise, all existents would be
27 realized when there is one existence. This would be necessary because existence
˘
28 as such is realized which is the negation (raf ) of non-existence as such which

29 necessitates (mustalzim) the realization of all existents.
30 As to (3): We do not concede that, if existence is not common, then the
31 belief of all existences would cease if the belief of the specifics ceases. This
32 follows only if predication of existence over the existents is not by ‘sharing in
33 expression’ (ištirāk lafzı̄). If it is like this, then the belief of one of the referents
34 of existence would remain ˙ there (hāsil) when the belief of one of the specifics
35 ceases. Hence it does not follow that ˙ ˙the belief of existence ceases by the fact that
˘
36 the belief of specifics ceases. The truth is that in concrete beings (fı̄ al-a yān)
37 existence is not one common nature between ‘the necessary’ and ‘the
38 contingent’. Otherwise, it would either be ‘essentially necessary’ or ‘essentially
39 contingent’. The first is absurd, because – if its ipseity were because of its
40 essence (dāt) – its species would be restricted to one individual, and it would
41 impossible¯ that it is common. If it were because of something else (ġayr), then it
42 would need it, and what needs something else is essentially contingent. The
144 Heidrun Eichner
1 second is impossible, otherwise each individual (fard) of it would be essentially

2 contingent because it needs something contingent. Therefore, the existence
3 which is essentially necessary would be essentially contingent. This is absurd.
4 Whether or not that what is conceived (mutasawwar) from each existence is
˙ – this is something which
˘
5 one thing not considering (ma a hadf ) the accidents
6 neither of them proves. ˙ ¯
7
8 (§ 17) baht (2): The opinion is widespread that the existence of contingent
9
˙ ¯ concrete beings is additional to their external essence. They argue
beings among
10 that existence is either identical to essence or intrinsic to it or extrinsic to it. The
11 first two [options] are false:
12 The first [option is false], because of several aspects: (1) If we intellect
˘
13 (na qul) blackness while having doubts about its outward existence then ‘what is
14 known’ is different from ‘what is doubted about’. (2) ‘Blackness insofar as it is
15 what it is’ is susceptible of non-existence. ‘Existent blackness’ is not susceptible
16 of non-existence. Hence ‘blackness insofar as it is what it is’ is different from
17 ‘existent blackness’. Therefore, existence is different from essence. (3) Existence
18 is common to all existents. Blackness is not common. Hence existence is
19 different from essence. (4) If existence were identical to blackness, if we say that
20 ‘blackness is existent’ this would have the same status as the claim that ‘blackness
21 is blackness’. This is not the case.
22
23 (§ 18) The second [option is false], because if existence were intrinsic to the
˘
24 essences it would be the most general essential [property] (a amm al-dātiyyāt)
25 which is common to them. So it would be genus for them, and the essentially ¯
26 necessary would be composed out of genus and differentia. This is absurd.
27 Further, because if it were intrinsic, it would be either substance or accident.
28 If it were substance, then a substance would be intrinsic to an accident. If it
˘
29 were accident, then the contrary (bi-l- aks) [would be the case]. If these both
30 possibilities are false, then the third is to be considered true – q.e.d. All this is
31 weak.
32 As to the argument that one can know blackness while doubting about its
33 existence, we say: We do not concede that from this follows that the existence of
34 blackness in concrete beings is different from the external essence. Rather it
35 follows that ‘the referent of blackness in the intellect’ is different from
36 ‘existence’.
37 As to the argument that ‘blackness insofar as it is what it is’ is susceptible of
38 non-existence and ‘the existent blackness’ is not susceptible of non-existence, we
39 say: If you mean by its ‘being susceptible of non-existence’ that it is possible that
˘
40 it can be annihilated (irtifā ) from the outward, then we do not concede that
41 ‘existent blackness’ is not susceptible of non-existence according to this
42 terminology (tafsı̄r). If you mean by it that blackness in the state (hāla) of non-
˙
1 existence is there (hāsil) in the outward and attributed with non-existence, this is
˙ )˙ because it is impossible that blackness has a ipseity in the
˘
2 impossible (mamnū
3 outward during the state of non-existence.
4 As to the argument that existence is common between all existent things:
5 You already know that it is weak.
6 As to the argument that if existence were identical to blackness, if we say
7 that ‘blackness is existent’ this would have the same status as the claim that
8 ‘blackness is blackness’, we say: If you make the substrates of both judgements
˘
9 (mawdū al-qadiyyatayn) external blackness, we do not concede the difference
10 between˙ both ˙judgements. If you make its substrate the referent of blackness,
11 from this does not follow that what is aimed at results, because then from this
12 follows their being different in the intellect, not in the outward [reality].
13 As to the argument that if existence were part of an essence it would be the
14 most general essential [property], we say: We do not concede this. This would
15 only follow if existence were common to all existents, and that this is weak has
16 just been shown. From the assumption (taqdı̄r) of its being common follows its
17 being the most general essential [property] only if it is essential for the essence of
18 the Necessary of Existence. Its being essential only follows if it has an essence
19 beyond existence. Why do you say that this is the case?
20 As to the argument: If it were intrinsic, it would be either substance or
21 accident etc., we say: If you mean by ‘substance’ the essence which, if it is
22 existent among concrete beings, is not in a substrate, and by ‘accident’ ‘existent
23 in a substrate’, then we do not accept this enumeration [to be exhaustive] (hasr),
24
˙ ˙
because a third part can be realized. This is, that – if it exists in concrete beings
25 – then it is in a substrate. If you mean by ‘substance’ the essence which if it exists
26 in concrete beings then it is in a substrate, then existence is neither a substance
27 nor an accident, because it does not have an essence beyond existence.
28 If you mean by ‘substance’ ‘what is independent (ġanı̄) from a substrate’ and
29 by ‘accident’ ‘what needs a substrate’ – why is it not possible that according to
30 this definition (tafsı̄r) substance makes the accident subsist (muqawwim)
˘
31 because it is possible that the whole (maǧmū ) needs a substrate while its parts do
32 not need it? The truth is that in concrete things existence is identical to external
essence because otherwise existence would either be a part (ǧuz ) of it, or an
˘
33
34 attribute (sifa) of it. The first is impossible, otherwise the existence of a thing
35
˙
would precede it because it is necessary that the part precedes the whole (al-
36 kull). This is absurd. And also, because, if existence were intrinsic to the external
37 essences, all simple essences would be composed. This is absurd. The second is
38 also absurd because, if existence were an attribute of essence, it would need it
39 (kāna muftaqiran ilayhā). What is needed by something has to have a ipseity
40 which comes before its ipseity. Then, essence would have a ipseity in outward
41 [reality] before existence. This doctrine is what one of the eminent scholars
˘
42 reports from Aristotle (al-mu allim al-awwal).

146 Heidrun Eichner
1 (IV) Atı̄r al-Dı̄n al-Abharı̄, The Book of Doubts (Kitāb al-Šukūk, MS Istanbul,
2
¯ 2319):
Ayasofya
3
4 (§ 19) He says: ‘If we know that a contingent being has a cause we judge that
5 this cause is existent etc.’ We say: Why do you say that if existence were not
6 shared this belief would cease? This, because we have a belief of some existence
˘
7 which is concretised in itself (wuǧūd mu ayyan fı̄ nafsihı̄) whose self is unknown
8 (maǧhūl nafsahū). According to us, the ‘concretised existence in itself ’ is
9 identical to the essence of this existence (nafs māhiyyat dālika al-wuǧūd). Why
10 do you say that if existence were not shared, then the belief ¯ of an existence which
11 is concretised in itself would cease because ‘considering other specifics’ ceases?
˘
12 Don’t you see that if we know one of the referents of the expression ayn [source,
˘
13 eye, concrete being] is existence, we are convinced of some concretised ayn as
14 such? The belief of the specifics of its concretised referent leads to the
˘
15 disappearance of all others. Yet, given this (ma a hādā), why is it necessary that
¯ referents?
˘
16 the referent of ayn is shared by the remainder of its

17 He says: ‘The referent of negation is one referent insofar as it is negation’.
18 We say: We do not concede this, but according to us, negation (salb) is a shared
19 expression (lafz muštarak) which is predicated of different referents in a way like
20 existence, too,˙ is a shared expression.
21 He says: ‘It is possible to intellect a triangle while doubting its external
22 existence’. We say: This points to the fact that external existence is different
23 from the essence which is intellected from the triangle. But why does it follow
24 from this that external existence is different from the external essence? We
25 concede that external existence is different from the intellected essence and also
26 from essence itself (nafs al-māhiyya). We say that external existence is in (fı̄) the
27 external essence. What you have mentioned does not point to this. We say the
28 same about the remaining aspects which point towards existence being not
29 identical to essence. Our answer is that we do not claim that existence is
30 different from outward essence. Rather we claim that outward existence is
31 different from the essence itself insofar as it is what it is, according to what we
32 have mentioned. We say that existence is identical to the outward essence
33 because, if it were additional to it in the outward, the essence would have a
34 ipseity in the outward other than the ipseity of existence in the outward. Both
35 personalities would be following each other (mutalāzimāni) in the outward, and
36 each of the two things following each other would become substratum for the
˘
37 other. So it would be necessary that one would be the cause ( illa) for the other.
38 Otherwise, there would not be resulting from them one composed thing. If
39 existence were the cause for essence it would be preceding essence, and then
40 existence which is accidental for something would be preceding this thing. This
41 is absurd. If essence were the cause for existence the existence of a contingent
42 being would be caused by the essence. The cause must precede the caused by
1 existence. Therefore, the existence of essence before its existence would follow.
2 This is absurd. What results from the teachings (madhab) of the philosophers is
3
¯
that external existence is different from the essence itself, and that it is identical
˘
4 ( ayn) to the external essences. Both are different in the intellect. Many things
5 are united in the outward and different in the intellect, as the author himself
6 admits in the case of the difference between mental and external composition. If
7 someone were to say: What is known from the teachings of the philosophers is
8 that all existents share in existence. If existence were identical to the external
9 essence, how can it be shared? We answer: What is meant when they say that
10 existence is shared between all existing things is that in intellect they share in
˘
11 existence and the general (al- āmm) which has no occurence (wuqū ) in concrete
12 beings.
13
˘
(V) Naǧm al-Dı̄n al-Kātibı̄ al-Qazwı̄nı̄, Ǧāmi al-daqā iq (MS Paris, Biblioth-
˘
14
15 que nationale 2370):
16
17 (§ 20) fasl (2): On existence of contingent beings being not identical to the
18 essence nor˙ part of it.
19 We conceptualize a triangle while doubting about its external existence. So,
20 in this state, one passes the judgement about the triangle in the intellect that it is
21 a triangle, but one does not pass the judgement that it is existent in outward
22 [reality]. If its external existence were identical to its being a triangle or part of
23 it, then the judgement would be impossible that it is a triangle without the
24 judgement that it is existent in outward [reality]. Therefore, existence is not
25 identical to the triangle and not part of it. Likewise it is possible for the
26 remainder of essences to intellect them while neglecting their external existence.
27 So, existence is not identical to contingent essences, nor is it part of them.
28
29 (§ 21) If someone were to say: ‘If existence were not identical to the contingent
30 essence, the external essence would be attributed with it, and existence would be
31 an attribute of it. An attribute needs “something which is attributed”, and what
32 is needed must have priority. Hence, the essence must have priority by [its]
33 existence before existence and then it would have another existence whose
34 existence must have priority over this existence, too. Thus, between essence and
35 existence would be an infinite number of existences, and something infinite
36 would be included between two ends, and this is absurd.’
37 We say: ‘We do not concede that if existence were not identical to essence,
38 then the external essence would be attributed with existence in outward
39 [reality]. This would follow only if the external essence were different from
40 existence. Why do you say that it follows from existence being different from
41 “essence as such” (nafs al-māhiyya) that it is different from the “external
42 essence”?’
148 Heidrun Eichner
1 This is so because in outward [reality] essence and existence are one thing
(šay wāhid), and when they are in intellect, intellect splits (fassala) them into
˘
2
3
˙ ‘essence’ and ‘existence’.
two things: ˙˙
4 If you say: ‘If it is affirmed that existence is different from essence in
5 intellect it follows that they are different in the outward [reality], too.
6 Otherwise, the judgement of intellect that they are different would not
˘
7 correspond to what actually is the case (hukm al- aql bi-l-muġāyara ġayr mutābiq
8 li-mā fı̄ nafs al-amr).’ ˙ ˙
9 We say: ‘We do not concede this because intellect passes the judgement
10 about them that they are different in intellect and united in outward [reality].
11 That this judgement corresponds to what actually is the case rests only on their
12 being different in intellect, but not in outward [reality]’.
13 If you say: ‘If they are different in intellect and unified in outward [reality]
14 from one thing in outward [reality] would result two representations (mitālān)
15 in intellect, one of them being “essence”, one of them being “existence”’.¯
16 We say: ‘Why is it not possible that from one thing in outward [reality]
17 result two representations in intellect? This, because from an isosceles triangle
18 result two representations in intellect, one of them being the representation of
19 “triangle absolutely”, and one of them being the representation of “isosceles
20 triangle”. In a similar vein, from “blackness” result two representations in
21 intellect while it is one thing in outward [reality] which does not have two
22 aspects so that to one of them would correspond “triangle” and to the other
23 “isosceles triangle”. Likewise, from blackness result two representations in
24 intellect, one of them being the representation of “colour-ness absolutely” (al-
25 lawniyya al-mutlaqa) and the other being the representation of “blackness-ness”
26 (al-sawādiyya) ˙while in outward [reality] they are one thing. There are many
27 similar things which will follow for you – God willing so.’
28
29 (§ 22) fasl (3): The existence of the essentially necessary is not different from its
30
˙
essence, neither in external [reality] nor in the intellect.
31 In external [reality] this is so because if its existence were different from its
32 essence it would either be part of it or extrinsic to it. Both options are false. The
33 first, because if it were part of it its essence would be composed and everything
34 which is composed is in need of its parts which are something other than it.
35 What needs something else is essentially contingent. Then, the essentially
36 necessary would be essentially contingent. This is absurd. The second [is false]
37 because if it were extrinsic to it, it would be an attribute which subsists in an
38 essence. Then it would need it, and what needs something else is essentially
39 contingent. Hence this existence insofar as it is existence would be essentially
40 contingent and it would have a cause. If its cause were not this essence, then the
41 essentially necessary would in its existence be in need of something else. This is
42 absurd. If it were identical to this essence (and the cause must precede by
1 existence what is caused), then this essence precedes existence by existence. Then
2 it would have another existence before existence and it follows that it precedes
3 by existence this existence, too. Thus it follows that between the essence and its
4 existence are infinite existences. This is absurd.
5 In intellect [this is absurd] for if intellect were to split the essentially
6 necessary into essence and existence it would have a universal essence whose very
7 referent would not prevent that it has infinite particulars. Then no one of these
8 particulars would exist by this essence itself because if some particulars were to
9 exist because of the essence itself but others not, this would be a preponderance
10 without an agent (tarǧı̄h bi-lā muraǧǧih). If no one of its particulars were to exist
11 by this essence itself, the ˙ existence of˙ the particular which is realized among
12 concrete beings would be there not because of the essence itself but because of a
13 distinct cause. Then, the Necessary of Existence would be existent because of a
14 separate cause. This is absurd. So, in the intellect the essentially necessary is not
15 split into two things, essence and existence. Rather its existence is identical to its
16 essence both in the intellect and in external [reality].
17 If it were said that if the existence of the Necessary of Existence were not
18 different from Its essence in concrete beings then its abstraction either would be
19 because of the nature of existence and its concomitants or because of a distinct
20 cause. The second is false because if its abstraction were due to a cause which is
21 distinct from the nature of existence and its concomitants, this would be due to
22 a cause which is something else than the essence of the Necessary of Existence
23 and its concomitants. Then, the Essentially Necessary in its abstraction would
24 be in need of something else. This is absurd. The first is also false because if its
25 abstraction were because of the nature of existence, anything which is there
26 among concrete beings would be abstract. The consequent is not true, and so
27 neither is the premise.
28 We say: We do not concede that if its abstraction were because of the nature
29 of existence everything which is there among concrete beings would be
30 permanently abstracted from existence. This would follow only if existence
31 among concrete beings were one thing. Why do you say that it is like this?
32 This, because existence of the contingents among concrete beings (wuǧūd
˘
33 al-mumkināt fı̄ al-a yān) is identical to the essence. It [i.e., this essence] is
34 different from the existence pertaining to the Necessary [of Existence] absolutely
35 (al-wuǧūd al-wāǧibı̄ al-muǧarrad). What results among concrete beings from the
36 singular instantiations (afrād) is existence which is predicated of the existences
37 which result in intellect and are individualized in the Necessary of Existence.
38 The existence of contingents does not have a ipseity among concrete beings, but
39 what results (al-hāsil) among concrete beings are their essences. If essences which
are concretised ˙(al-māhiyyāt
˙
˘
40 al- ayniyya) are there (hasala) in intellect, intellect

41
˙ ˙
splits them into ‘essence’ and ‘existence’. Then the personalities of the existences
42 result only in mind (fa-l-wuǧūdāt innamā tahsulu huwiyyātuhā fı̄ al-adhān).
˙˙ ¯
150 Heidrun Eichner
1 What is meant when the philosophers say that existence is shared among ‘the
2 necessary’ and ‘the contingent’ is that if contingent essences are there in intellect
3 and if intellect splits them into ‘essence’ and ‘existence’, – if one considers their
4 existence – there is a sharing with the existence of the Necessary [of existence]
5 (al-wuǧūd al-wāǧibı̄) in what is called ‘existence’ (fı̄ musammā al-wuǧūd). This is
˘
6 ‘common existence’ (al-wuǧūd al- āmm) which takes place only in mind (lā
˘
7 wuqū a lahū illā fı̄ al-dihn).
8 ¯
By what we mention is refuted what our Imām and the Imām of all
9 researchers, the recreation of the souls of those who seek for knowledge, says
10 against the philosophers (al-hukamā ) in the Kitāb al-Mulahhas where he says:
˘
11 They agree that it is impossible ˙ that some individuals of˘ ˘the˙ nature of the
˘
˘
12 species (al-tabı̄ a al-naw iyya) are abstracted from matter and others are
13 connected. ˙So, existence, too, is one nature. If it is not in need of connection
14 with an essence then it is like that absolutely. If it is possible that it is sometimes
15 abstract and sometimes connected–why is it not possible in one nature of a
16 species that it is sometimes abstract from matter, and sometimes connected?
17
Because we say: Why do you say that existence is not independent from being
18
connected to an essence absolutely? It is only then not independent if some of
19
the existences are there among concrete beings as inherent (hāll) in an essence.
20
This is not the case. The existence of contingent beings takes ˙ place only in
21
mind, not in concrete beings. The existence of the Necessary of Existence is
22
abstract from essence.
23
The result is: Existence in concrete beings is identical to the concretised
24
˘
essence (al-māhiyya al- ayniyya), and it is not identical to the essence in intellect.
25
Existence which is predicated of the Necessary of Existence has among concrete
26
beings only the form of the existence of the Necessary [of Existence]. Intellect
27
splits the contingent which takes place in concrete beings into ‘essence’ and
28
29
‘existence’ after this. The individual instantiation (ifrād) of existence takes place
30
in intellect, and intellect divides the form of absolute existence into existence of
31
the Necessary [of Existence] and into these existences which are assumed
32
(mutaqaddira) in intellect. That on which the division is carried out must be
33
shared among the parts. So, existence is shared between the individual
34 instantiations (afrād) of the existents which are there in intellect.
35
36
37 Bibliography
38
al-Abharı̄, al-Mufaddal b. ‘Umar Atı̄r al-Dı̄n, Kašf al-haqā iq, MS Tehran, Maǧlis-i šūrā-
˘
39 ˙ 1 – 212. ¯ ˙
yi millı̄ 2752, ˙pp.
40 –– , Marāsid al-maqāsid, MS Istanbul, Serez 1963.
41 ˙
–– , Muntahā ˙ fı̄ ibānat al-asrār, MS Tehran, Maǧlis-i Šūrā-yi Millı̄ 2752,
al-afkār
42 pp. 213 – 358.
–– , R. al-masā il, MS Istanbul, Ragıp Paşa 1461.
˘
1
2 H. Eichner, Dissolving the Unity of Metaphysics: From Fakhr al-Din al-Razi to Mulla
3
Sadra al-Shirazi, Medioevo 32, 2007, pp. 139 – 97.
–– , ‘Knowledge by Presence’, Apperception and the Mind-Body Relationship: al-
4 Suhrawardı̄ and Fakhr al-Dı̄n al-Rāzı̄ as Representatives of a 13th Century
5 Discussion, in In the Age of Averroes, ed. P. Adamson, London: The Warburg
6 Institute, 2011, pp. 117–40.
7 –– , The Post-Avicennian Philosophical Tradition and Islamic Orthodoxy: Philosophical and
8
Theological Summae in Context, unpublished professorial dissertation, submitted
Halle 3/2009.
9 F. Griffel, On Fakhr al-Dı̄n al-Rāzı̄’s Life and the Patronage he Received, Journal of
10 Islamic Studies 18, 2007, pp. 313 – 44.
11 T. Mayer, Ibn Sı̄nā’s Burhān al-Siddiqı̄n, Journal of Islamic Studies 12, 2001, pp. 18 – 39.
al-Kātibı̄ al-Qazwı̄nı̄, Alı̄ b. ˙ Umar Nağm al-Dı̄n, Ǧāmi al-daqā iq, MS Paris,
˘
˘
˘
12
13
Bibliothque nationale 2370.
–– , Mutārahāt falsafiyya bayna Nası̄r al-Dı̄n al-Tūsı̄ wa-Naǧm al-Dı̄n al-Kātibı̄, ed.
14 Muh˙ammad ˙ Hasan Āl-Yası̄n, Baghdad:
˙ Maktabat˙ al-ma ārif, 1956.
˘
15 al-Rāzı̄, ˙Muh. ammad˙ b. Umar Abū Abd Allāh Ibn al-Hatı̄b Fahr al-Dı̄n, al-Mabāhit al-
˘
16 ˘ ˘ ˙ -dı̄.˘ Beirut: Dār al-kitāb

mašriqiyya, ed. Muh. ammad al-Mu‘tas. im bi-lla-h al-Baġda ˙ ¯ al-
˘
17 arabı̄, 1990.
18
–– , al-Mulahhas fı̄ al-hikma, MS Berlin Or. Oct. 623.
al-Samarqandı̄, ˙ ammad
˘ ˘ Muh ˙ b. Ašraf al-Hasanı̄ Šams al-Dı̄n, al-Sahā if al-ilāhiyya, ed.
˘
19 ˙
Ahmad Abd al-Rahmān al-Šarı̄f, Kuwait, ˙ 1985. ˙ ˙
˘
20 ˙
A. Shihadeh, ˙
The Teleological Ethics of Fakhr al-Dı̄n al-Rāzı̄, Leiden: Brill, 2006.
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
1
2
3
4 Avicenna’s Metaphysics in the Medieval Hebrew
5 Philosophical Tradition
6
7
A Short Historical Sketch of its Evident Traces
8
9 Mauro Zonta
10
11
Avicenna’s metaphysics and, in particular, Avicennian works in which
12
metaphysical themes are studied in detail had a limited and mostly indirect
13
impact on medieval Jewish philosophy.1 Previous studies on this subject have
14
shown that some major Jewish philosophers probably knew and hinted at
15
various points of Avicenna’s metaphysical thought without mentioning the
16
author’s name. As a matter of fact, they might have not read Avicenna directly,
17
but through the intermediation of al-Ġazālı̄’s The Intentions of the Philosophers
18
(Maqāsid al-falāsifa), which was very well known by late medieval Jewish
19 ˙
philosophers. 2
This might be suggested, e. g., from an examination of Abraham
20
Ibn Daud’s The Exhalted Faith (Kitāb al-‘aqı̄da al-rafı̄‘a), whose original Arabic
21
text, written in Spain in 1161, is not extant, but was twice translated into
22
Hebrew in the second half of the fourteenth century.3 Moses Maimonides, too,
23
certainly knew and employed some aspects of Avicenna’s metaphysics, though
24
possibly through a perusal of al-Ġazālı̄’s work.4 Later on, in 1354, the Spanish
25
Jewish philosopher Moses ben Judah employed al-Ġazālı̄’s work as the main
26
source for the structure and contents of his encyclopedia, Love for Pleasures
27
28
29
30
1 See Zonta, Linee del pensiero islamico nella storia della filosofia ebraica medievale,
31 pp. 450 – 62. This particular fact does not imply that Avicenna’s philosophical thought as
32 a whole had no impact on late medieval Jewish philosophy; on the contrary, as observed
33 there (p. 462), ‘il quadro storico che abbiamo cercato di tracciare induce a pensare che,
34 diversamente da quanto generalmente si crede, nel quadro delle influenze islamiche sulla
filosofia ebraica medievale, ad Avicenna spetti un ruolo pressoch pari a quello di
35
Averro’. See also the historical sketches in Zonta, The Role of Avicenna and of Islamic
36 ‘Avicennism’; Zonta, Avicenna in Medieval Jewish Philosophy.
37 2 This fact was apparently not limited to physics (as pointed out by Harvey, Why Did
38 Fourteenth-Century Jews Turn to Alghazali’s Account of Natural Science?), but probably
39 involved metaphysics too. The whole text of al-Ġazālı̄’s work was translated into Hebrew
three times in the first half of the 14th century.
40
3 See in particular ‘Eran, Me-’emunah tame-ah le-’emunah ramah, pp. 27 and 76.
41 4 See the conclusion suggested in Zonta, Maimonides’ Knowledge of Avicenna,
42 pp. 221 – 2.
154 Mauro Zonta
1 (Ahavah ba-ta‘anugim); by this way, he shows to follow many aspects of

2 Avicenna’s metaphysical doctrines.5
3 There are really very few clear traces of a direct and acknowledged use of
4 Avicenna’s metaphysics, and of that of The Cure (al-Šifā’) in particular, as a
5 source.6 In this short historical sketch, I will give a general summary of the
6 results of research on this subject, according to the chronological order of the
7 Jewish authors.
8 Probably the first such author who evinces a direct knowledge of the
9 contents of Avicenna’s metaphysics was Moses ha-Levi (Mūsā al-Lāwı̄), a
10 neglected thirteenth-century Spanish Jewish philosopher. In his major extant
11 philosophical treatise devoted to metaphysical and theological matters, the
12 Divine Treatise (Maqāla ’ilāhiyya), he displays a familiarity with Avicenna’s
13 thought about those subjects, as Georges Vajda has shown.7
14 However, a number of direct quotations of Avicenna’s metaphysics are
15 found in some of the works of another Jewish author, probably living in the
16 same time and milieu of Moses ha-Levi: Shem Tov Ibn Falaquera (1225 – 95
17 ca.). In those quotations, Ibn Falaquera openly admits his dependence on
18 Avicenna, although the direct sources of these quotations cannot be identified in
19 all cases. I have listed elsewhere Avicenna’s philosophical works quoted by Ibn
20 Falaquera.8 The most interesting cases are the passages of the metaphysical
21 section of The Cure found in one of Ibn Falaquera’s main philosophical works,
22 his commentary on Maimonides’s Guide of the Perplexed: The Guide to the Guide
23 (Moreh ha-Moreh). Most of them were first identified by the Israeli scholar Ya’ir
24 Shiffman, who has recently published a critical edition of Ibn Falaquera’s work.9
25 In the Appendix here below, I will give a list of them.
26 As pointed out by Shiffman, another important source of Avicenna’s
27 metaphysics in Ibn Falaquera’s Guide to the Guide is the metaphysical section of
28 The Salvation (al-Nağāt). 10 However, there was a partial Hebrew translation of
29 this text, made by the ProvenÅal Jewish philosopher Todros ben Meshullam ben
30 David Todrosi of Arles only later, in the period 1330 – 40. This translation,
31
32 5 See Zonta, The Role of Avicenna and of Islamic Avicennism, pp. 657 – 9.
33 6 See Zonta, La tradizione ebraica medievale dei testi arabi, pp. 553 – 4.
34 7 See Vajda, Un champion de l’avicennisme.
8 Zonta, Hebrew Transmission of Arabic Philosophy and Science, pp. 128 – 9.
35
9 Shem Tov ben Joseph Ibn Falaquera, Moreh ha-Moreh; about the influence of Avicenna’s
36 metaphysics on Ibn Falaquera’s Guide to the Guide, see in particular Shiffman’s
37 introduction to his edition of this work, on pp. 56 – 8.
38 10 See the list of quotations of Avicenna’s works in Shem Tov ben Joseph Ibn Falaquera,
39 Moreh ha-Moreh, p. 381. As pointed out by Shiffman, in some cases Ibn Falaquera’s
Avicennian quotations seem to result from a mixture of passages of The Cure with the
40
corresponding passages of The Salvation: see, e. g., Shiffman’s notes to his edition on
41 pp. 238 and 291. About this fact, see Bertolacci, The Reception of Aristotle’s ‘Metaphysics’
42 in Avicenna’s Kitāb al-Šifā’, p. 587.
1 bearing the title The Soul’s Salvation (Hasalat ha-nefeš) and being still
2 unpublished, is found in two handwritten copies; ˙ 11
it includes the complete
3 physical section and some of the first chapters of the metaphysical section of
4 Avicenna’s Salvation,12 so showing the interest of the author, one of the main
5 representatives of a sort of fourteenth-century ‘Jewish Avicennism’, for
6 Avicenna’s natural and metaphysical philosophy in general, and for his
7 theological doctrines in particular.
8 In the same period (second quarter of the fourteenth century), some explicit
9 quotations of Avicenna, allegedly drawn from the lost metaphysical section of an
10 Avicennian philosophical work, The Oriental Philosophy (al-Hikma al-mašri-
11 ˙
qiyya), can be found in four Medieval Hebrew texts by a Castilian Jewish (and
12 13
subsequently Christian) author, Avner of Burgos. Finally, it should be pointed
13 out that at least one literal quotation of the metaphysical section of The Cure
14 has been found by Vajda in a mid-fourteenth-century work by another Spanish
15 Jewish philosopher, Joseph Ibn Waqqar: The Treatise Reconciling Philosophy and
16 Religious Law (al-Maqāla al-ğamı̄‘a bayna al-falsafa wa-l-šarı̄‘a).14 Ibn Waqqar’s
17 apparent ‘Avicennism’ in this theological-philosophical work might justify the
18 explicit employment of Avicenna’s metaphysics as one of its sources.
19 From the above data, it seems that there were two phases of the reception of
20
Avicenna’s metaphysics among Medieval Jewish philosophers. The former is
21
found in the twelfth-century Jewish Aristotelians like Abraham Ibn Daud and
22
Maimonides, and in some of their later followers, where Avicenna’s metaphys-
23
ical doctrines were apparently transmitted in an indirect way, through al-Ġazālı̄’s
24
25
11 These copies are found in fols 208r–301v of the manuscript of London, British Library,
26 Add. 27559 (Margoliouth 890), probably written by a Sephardic hand in the 15th
27 century; and in fols 87r–159v of the manuscript of Paris, Bibliothque National de
28 France, hbreu 1023, probably written by a Sephardic hand in the 15th-16th centuries.
29 The former was the only copy known to Moritz Steinschneider: see Steinschneider, Die
hebrischen bersetzungen, p. 285. A microfilm copy of both manuscripts is found in the
30
Institute of Microfilmed Hebrew Manuscripts of the National Library of Israel in
31 Jerusalem, under the signatures F 6094 and F 15717; in that place, there is also a CD
32 copy of the former, under the signature CD 177. See now Berzin, The Medieval Hebrew
33 Version.
34 12 The first chapters of the metaphysical section of The Salvation are found in fols
277v–301v of the London manuscript, and in fols 141r–159v of the Paris manuscript.
35
13 For an identification, edition and English translation of these quotations, see here below,
36 Zonta, Possible Hebrew Quotations of the Metaphysical Section of Avicenna’s Oriental
37 Philosophy and Their Historical Meaning.
38 14 See Vajda, Recherches sur la philosophie et la kabbale, pp. 115 – 297, and p. 132 in
39 particular, where Ibn Waqqar’s quotation of Avicenna’s Cure, Metaphysics, treatise 9, c. 6
(on providence), is mentioned. Ibn Waqqar probably knew the contents of Avicenna’s
40
metaphysics through Fahr al-Dı̄n al-Rāzı̄’s commentary on Avicenna’s Remarks and
41 ˘
Admonitions (al-Išārāt wa-l-tanbı̄hāt) too, The Cores of the Remarks (Lubāb al-išārāt): see
42 Vajda, Recherches sur la philosophie et la kabbale, p. 125 – 6.
156 Mauro Zonta
1 interpretation of them in particular. The latter is found in some thirteenth- and

2 fourteenth-centuries Jewish authors, most of whom are Spanish: they show to
3 have read Avicenna’s metaphysics and often admit their dependence on it.
4 Among them, a paramount role was played by Shem Tov Ibn Falaquera, as it
5 results from the detailed examination of some of his works.
6
7
8 Appendix
9
10 Quotations of the Metaphysics of Avicenna’s Cure (al-Šifā’) in Shem Tov Ibn
11 Falaquera’s Guide to the Guide (Moreh ha-Moreh)15
12
13 * Shem Tov ben Joseph Ibn Falaquera, Moreh ha-Moreh, p. 136, lines 50 – 53
14 (on GP, I, 34) = Avicenna, Metaphysics, p. 14, line 36–p. 15, line 4 of the
15 English translation; p. 14, line 19–p. 15, line 2 of the Arabic text (about the
16 objects of mathematical sciences and their relationship to the contents of
17 metaphysics).
18 Shem Tov ben Joseph Ibn Falaquera, Moreh ha-Moreh, p. 162, lines 57 – 8
19 (on GP, I, 58) = Avicenna, Metaphysics, p. 296, lines 14 – 17 of the English
20 translation; p. 296, lines 9 – 10 of the Arabic text (about the definition of God
21 as ‘intellect, and subject and object of intellection’).
22 * Shem Tov ben Joseph Ibn Falaquera, Moreh ha-Moreh, p. 220, lines 249 –
23 55 (on GP, II, Introduction) = Avicenna, Metaphysics, p. 30, lines 7 – 8, 19 – 26
24 of the English translation; p. 30, lines 4, 11 – 14 of the Arabic text (on the
25 nature of the Necessary Existent, i. e. God).
27 (on GP, II, Introduction) = Avicenna, Metaphysics, p. 131, line 10–p. 132, line 8
28 of the English translation; p. 131, line 6–p. 132, line 4 of the Arabic text
29 (summarized; about the philosophical meaning of the Arabic term qūwa,
30 ‘potency’).
32 (on GP, II, 4) = Avicenna, Metaphysics, p. 311, line 27–p. 312, line 12 of the
33 English translation; p. 311, line 15–p. 312, line 7 of the Arabic text (partially
34 altered; about the characteristic of the soul, which is the proximate mover of the
35 heavens).
37 (on GP, II, 4) = Avicenna, Metaphysics, p. 314, lines 11 – 17 of the English
38 translation; p. 314, lines 5 – 8 of the Arabic text (about the characteristic of the
39 infinite power which moves the heavens).
40
41 15 The quotations that escaped Shiffman’s notice are signed as *; GP is Maimonides’s Guide
42 of the Perplexed.
4 opinion that two things should exist in one thing).16
5 Shem Tov ben Joseph Ibn Falaquera, Moreh ha-Moreh, p. 289, line 32-p.
6 290, line 44 (on GP, II, 40) = Avicenna, Metaphysics, p. 364, line 27–p. 365,
7 line 25 of the English translation; p. 364, line 16–p. 365, line 13 of the Arabic
8 text (about the necessity of a law, given by a prophet).17
12 fact that the principles of all matters descend from God).
14 (on GP, III, 19) = Avicenna, Metaphysics, p. 288, lines 3 – 6 of the English
15 translation; p. 288, lines 2 – 3 of the Arabic text (about God’s omniscience).
16 Shem Tov ben Joseph Ibn Falaquera, Moreh ha-Moreh, p. 334, line 133–p.
17 335, line 150 (Appendix 1) = Avicenna, Metaphysics, p. 350, line 15–p. 351, line
18 26 of the English translation; p. 350, line 8–p. 351, line 13 of the Arabic text
19 (about the perfection of the man’s natural soul).18
20
21
22 Bibliography
23
24 Avicenna, al-Nağāt fı̄ al-mantiq wa-l-’ilāhiyyāt ta’lı̄f li-šayh al-ra’ı̄s … Ibn Sı̄nā, ed.
25 ‘A. ‘Amı̄ra, Beirut: Wālid ˙al-mağı̄d, 1992/1412. ˘
26
27 A. Bertolacci, The Reception of Aristotle’s ‘Metaphysics’ in Avicenna’s ‘Kitāb al-Šifā’’.
28 A Milestone of Western Metaphysical Thought, Leiden/Boston: Brill, 2006.
29 G. Berzin, The Medieval Hebrew Version of Psychology in Avicenna’s Salvation (Al-Najāt),
30 Ph.D. Thesis, Harvard University, 2010.
31
32
33
34 16 In this case too, the first sentence of Avicenna’s text is not entirely identical with that
translated by Ibn Falaquera. Avicenna writes: fa-zanna qawmun ’anna al-qismata tūğibu
35 ˙
wuğūda šay’ayni fı̄ kulli šay’in, ‘(One) group thought that the division necessitates the
36 existence of two things in each thing’. Ibn Falaquera seems to translate here a partially
37 different text, as follows: fa-zanna qawmun ’anna al-’ilāha yūğibu wuğūda šay’ayni fı̄ kulli
38 šay’in, ‘(One) group thought˙ that God necessitates the existence of two things in each
39 thing’.
17 It should be pointed out that an identical passage is found in al-Nağāt fı̄ al-mantiq wa-l-
40
’ilāhiyyāt … li- … Ibn Sı̄nā, vol. II, p. 165, line 24-p. 166, line 16. ˙
41 18 An identical passage is found in al-Nağāt fı̄ al-mantiq wa-l-’ilāhiyyāt … li- … Ibn Sı̄nā,
42 vol. II, p. 153, line 3–p. 155, line 6. ˙
158 Mauro Zonta
1 E. Eisenmann, Ahabah ba-Ta‘anugim: A Fourteenth-Century Maimonidean Encyclo-

2 pedia, in Traditions of Maimonideanism, ed. C. Fraenkel, Leiden: Brill, 2009,
3
pp. 213 – 22.
A. ‘Eran, Me-’emunah tame-ah le-’emunah ramah. Haguto ha-qdam-Maymonit šel ha-
4 Rabad (From Simple Faith to Sublime Faith. Ibn Daud’s Pre-Maimonidean Thought),
5 Tel Aviv: ha-Qibbus ha-me’uhad, 1998.
6
˙
S. Harvey, Why Did Fourteenth-Century ˙ Jews Turn to Alghazali’s Account of Natural
7 Science?, Jewish Quarterly Review 91, 2001, pp. 359 – 76.
8
Shem Tov ben Joseph Ibn Falaquera, Moreh ha-Moreh, ed. Y. Shiffman, Jerusalem: The
World Union of Jewish Studies, 2001.
9 M. Steinschneider, Die hebrischen bersetzungen des Mittelalters und die Juden als
10 Dolmetscher, 2 vols, Berlin: Kommissionsverlag des Bibliographischen Bureaus,
11 1893.
12 G. Vajda, Recherches sur la philosophie et la kabbale dans la pense juive du Moyen ffge,
13
Paris/Le Haye: Mouton & Co., 1962.
–– , Un champion de l’avicennisme. Le problme de l’identit de Dieu et du premier
14 moteur d’aprs un opuscule judo-arabe indit du XIIIe sicle, Revue tomiste 56,
15 1948, pp. 480 – 508.
16 M. Zonta, Avicenna in Medieval Jewish Philosophy, in Avicenna and His Heritage, eds
17 J. Janssens and D. De Smet, Leuven: Leuven University Press, 2002, pp. 267 – 79.
18
–– , Hebrew Transmission of Arabic Philosophy and Science: A Reconstruction of Shem
Tov Ibn Falaquera’s Arabic Library, in L’interculturalit dell’Ebraismo, ed. M. Perani,
19 Ravenna: Longo Editore, 2004, pp. 121 – 37.
20 –– , La tradizione ebraica medievale dei testi arabi, in Lo spazio letterario del Medioevo. 3.
21 Le culture circostanti, vol. II, eds M. Capaldo, F. Cardini, G. Cavallo and B. Scarcia
22 Amoretti, Rome: Salerno Editrice, 2003, pp. 531 – 67.
23
–– , Linee del pensiero islamico nella storia della filosofia ebraica medievale, Annali
dell’Istituto Universitario Orientale di Napoli, 57, 1997, pp. 101 – 44, 450 – 83.
24 –– , Maimonides’ Knowledge of Avicenna. Some Tentative Conclusions About a
25 Debated Question, in The Trias of Maimonides/Die Trias des Maimonides. Jewish,
26 Arabic, and Ancient Culture of Knowledge/Jdische, arabische und antike Wissenkultur,
27 ed. G. Tamer, Berlin/New York: Walter de Gruyter, 2005, pp. 211 – 22.
28
–– , The Role of Avicenna and of Islamic ‘Avicennism’ in the 14th-Century Jewish
Debate Around Philosophy and Religion, Oriente moderno, 80, n.s. 19, 2000,
29 pp. 647 – 60.
30
31
32
33
34
35
36
37
38
39
40
41
42
1
2
3
4 ‘Happy is he whose children are boys’:
5 Abraham Ibn Daud and Avicenna on Evil
6
7 Resianne Fontaine
8
9
1. Introduction
10
11
The title of this paper is the first part of a quotation from the Babylonian
12
Talmud, which reads: ‘Happy is he whose children are boys, unhappy is he
13
whose children are girls.’1 As we shall see, the twelfth-century philosopher
14
Abraham Ibn Daud, the main protagonist of my paper, adduces this Rabbinic
15
statement in his treatment of the problem of evil. Ibn Daud’s approach to the
16
problem of evil differs markedly from that of other pre-Maimonidean Jewish
17
18
thinkers, and this is mainly due to his use of Avicenna. In order to explain this, I
19
will first provide some background information on Ibn Daud and then present a
20
brief outline of Avicenna’s discussion of evil. Following this I will turn to the
21
main topic of this paper: the reception of Avicenna’s views on evil by Ibn Daud.
22 Born in Andalusia somewhere around 1110, Abraham Ibn Daud received an
23 education both in religious learning and ‘Greek wisdom’, that is, non-Jewish
24 science and philosophy. Later in life, and presumably in the wake of the
25 Almohad invasion in Andalusia, he moved to Toledo, where he died ca. 1180.
26 In Toledo he composed two works, a chronicle entitled the Book of Tradition,
27 and his philosophical The Exalted Faith, which is the subject of this paper.
28 Written around 1160, both were intended as a defense of Judaism. While the
29 former became a classic in Jewish literature, The Exalted Faith, subtitled ‘[The
30 book] that brings harmony between philosophy and religion’, was soon
31 overshadowed by Maimonides’ masterpiece The Guide of the Perplexed, which
32 was written some thirty years later. Ibn Daud wrote the Exalted Faith (Hebrew
33 title: Ha-Emunah ha-Ramah, henceforth: ER) in Arabic (al-‘Aqı̄da al-rafı̄‘a).
34 The Arabic original is now lost, but the work is extant in two Hebrew
35 translations, both dating from the end of the fourteenth century.2 The book has
36 long been neglected in modern scholarship. However, the past two decades have
37
38
39 1 Babylonian Talmud, Qidd. 82b; Pes 65b.
2 For more detailed information on Abraham Ibn Daud and his thought see the surveys in
40
Encyclopedia Judaica, ed. 2006, under Ibn Daud, Abraham; The Stanford Encylopedia of
41 Philosophy under Abraham Ibn Daud; and the monographs Eran, From Simple Faith, and
42 Fontaine, In Defence of Judaism.
160 Resianne Fontaine
1 borne witness to a new interest in Ibn Daud’s thought, and he continues to

2 attract the attention of scholars of Latin philosophy too, since there are
3 indications that our Ibn Daud is the Avendauth who cooperated with
4 Dominicus Gundissalinus in translating philosophical texts from Arabic into
5 Latin.3
6
7 In his introduction to The Exalted Faith Ibn Daud tells us that he was prompted
8 to write his book to address a query about the problem of human freedom.
9 Some Biblical verses seem to teach that man has free will, whereas others, such
10 as: ‘I will harden Pharaoh’s heart’ (Ex 7:3), seem to imply that God determines
11 human action. According to Ibn Daud, the problem can only be solved through
12 a correct interpretation of the verses that contradict reason when interpreted
13 literally. In other words, reason should be the ultimate arbiter between
14 conflicting biblical texts. Regrettably, Ibn Daud continues, many people of his
15
generation fail to investigate the principles of the Jewish religion philosophically
16
and, as a result, they are confused when it comes to solving theological
17
problems, such as that of free will. This is why he seeks ‘to open the eyes of the
18
learned men of our nation regarding the principles of faith by providing proofs
19
from biblical verses and demonstrations from true philosophy.’4
20
By the term ‘true philosophy’ Ibn Daud means the philosophy as taught by
21
Alfarabi and, in particular, Avicenna. Although some earlier Jewish thinkers
22
display familiarity with views of the falāsifa, Ibn Daud is the first Jewish
23
24
philosopher to make extensive and systematic use of them. In fact, his Exalted
25
Faith can be seen as an attempt to transplant the thought of the falāsifa to
26
Jewish soil. Therefore he is commonly considered to be the first Jewish
27
Aristotelian, although the question to what extent his Aristotelianism is mixed
28 with Neoplatonism is a matter of debate. What is undisputed, however, is that
29 with Ibn Daud a ‘wind of change’ enters medieval Jewish thought, primarily
30 because of his use of Avicenna whose views are present throughout Ibn Daud’s
31 book, even though the Muslim philosopher is never mentioned by name. For
32 example, Avicenna’s notions about the soul pervade Ibn Daud’s presentation of
33 the subject, and the same applies to his conception of God as the Necessary
34 Existent.5 Ibn Daud’s approach to the problem of evil is another case in point.
35 To substantiate this claim, I will now highlight the main points of Avicenna’s
36
37 3 For the most recent information on this issue, see Hasse, Avicenna’s ‘De Anima’, pp. 4 – 7,
38 and Bertolacci, Albert the Great, pp. 133 – 7. See also Essudri, R. Avraham bn Daud,
39 R. Efraim al-Naqawa we-sefer ‘al-‘Aqı̄da al-rafi‘a’.
4 References to ER are to ed. Weil 1851 with references to ed. Samuelson/Weiss 1986
40
(henceforth: S) given in parentheses. ER Intro, ed. Weil, p. 3.7 – 8 (S 5b11 – 13).
41 5 On Ibn Daud’s use of Avicenna’s concept of the Necessary Existent see Fontaine and
42 S. Harvey, Jewish Philosophy on the Eve of the Age of Averroism.
Abraham Ibn Daud and Avicenna on Evil 161
1 approach to the subject and then proceed to discuss how Ibn Daud responds
2 to it.
3
4
5
2. Avicenna on Evil
6
Avicenna addresses the problem of evil in various works, for example in Kitāb
7
8
al-Šifā’, Kitāb al-Nağāt, Kitāb al-Išārāt wa’l-tanbı̄hāt and Sirr al-qadā’.
¯
9
Unfortunately, our information about the circulation of his writings in al-
10
Andalus and their use by Jewish philosophers is as yet incomplete.6 Yet, there are
11 several good indications that Ibn Daud derived his knowledge of Avicennian
12 views on evil from the discussion in chapter IX.6 of the Ilāhiyyāt of his Šifā’.7 To
13 begin with, the Šifā’ was available in Toledo, as it was translated into Latin there
14 in Ibn Daud’s day. Furthermore, Ibn Daud displays familiarity with Šifā’ and/or
15 Nağāt in various other sections of ER. Finally, a comparison of the account of
16 evil in the four aforementioned Avicennian texts shows that Ibn Daud’s
17 discussion of the issue has much in common with how the topic is dealt with in
18 both Šifā’ and Nağāt. Kitāb al-Išārāt wa’l-tanbı̄hāt and Sirr al-qadā’ do not
¯
19 contain all the issues that appear in Ibn Daud.8 The account in Nağāt is virtually
20 identical to that in Šifā’, Ilāhiyyāt IX.6.9 Another likely source of Ibn Daud is al-
21 Ġaza-lı̄’s rendering of Avicenna’s views in Maqāsid al-falāsifa.10
˙
22 Chapter IX.6 of the Ilāhiyyāt in Šifā’ is entitled: ‘On providence and how
11
23 evil enters divine predetermination’. The title thus makes it immediately clear
24 that Avicenna treats the problem of evil as related to his conception of
25
26 6 See Zonta, Avicenna in Medieval Jewish Philosophy, and S. Harvey, Avicenna’s Influence
27 on Jewish Thought. For the Andalusian background, see Gutas, What was there for the
28 Latins to Receive?.
29 7 Avicenna, Šifā’, Ilāhiyyāt IX.6, References to the text are to the Cairo edition, 1960, with
the corresponding pages in the text published by Marmura, 2005, given in parentheses.
30
8 For the latter, see Hourani, Ibn Sı̄nā’s Essay on the Secret of Destiny, pp. 27 – 31. For Kitāb
31 al-Išārāt wa’l-tanbı̄hāt, see ed. S. Dunyā, pp. 729 – 46. For the Nağāt I have consulted ed.
32 M. Fakhry 1985, pp. 320.5 – 326.
33 9 This is consistent with Bertolacci’s findings with respect to Šifā’, Ilāhiyyāt VIII.6–IX.3,
34 see Bertolacci, Reception, p. 486.
10 Cf. Eran, From Simple Faith, index under al-Ghazali, and Fontaine, In Defence of
35
Judaism, index, under (al-)Ghazali. See Maqāsid, ed. S. Dunyā, p. 296 – 9.
36 11 Ilāh. IX.6, p. 414.16 (339.3). For a full study ˙ of Avicenna’s treatment of evil and its
37 Greek sources, see Inati, The Problem of Evil. This book is based primarily on Šifā’, but
38 also adduces other Avicennian texts. A very useful overview is Steel, Avicenna and
39 Thomas Aquinas, based on the Latin translation of the Šifā’, see p. 172, n. 6. See also
Janssens, Problem of Human Freedom, who treats the issue on the basis of various
40
Avicennian texts; Belo, Chance and Determinism, especially pp. 38 – 53 and 113 – 20
41 where the problem of evil is discussed within the framework of an examination of
42 Avicenna’s views on chance, and Michot, La destine de l’homme, pp. 59 – 68.
1 providence. He first explains how he understands providence, which comes

2 down to this: the First Cause knows Itself as absolutely good and as the cause of
3 all goodness and perfection of the lower beings. From this apprehension of
4 goodness and order the universe emanates from the First Cause in the best
5 possible way, and this is what is meant by providence.12 Thus, according to
6 Avicenna, God’s knowledge lies at the basis of providence. But if God is the
7 cause of the good and perfect order in the universe, how are we to explain evil?
8 Avicenna starts his theodicy by clarifying the various meanings of the term
9 ‘evil.’ We call evil that which is a deficiency (naqs), like ignorance, weakness, or
˙
10 a deformity. We also use the term for things like pain and grief.13 This last kind
11 of evil is not just privation like the first kind, for pain is perceived as an existing
12 reality, but it follows upon deficiency and privation.14 Later on Avicenna
13 provides a somewhat different ‘classification’ of evil, where he explains that ‘evil’
14
is said of: 1) reprehensible acts, which have their principles in 2) moral
15
dispositions; 3) pain and distress; and 4) a defect (nuqsān) or a loss (fuqdān) of a
16 ˙
natural perfection.15 Reprehensible acts are evil in so far as they cause privation
17
to the person who performs such acts, as in the case of someone who commits
18
an act of injustice. Moral dispositions are not evil in themselves, but only
19
because of the evil acts that proceed from them.16 What emerges from both
20
passages is that evil is privation and lack or loss of perfection. Furthermore,
21
Avicenna distinguishes between essential evil (al-šarr bi’l-dāt) and accidental evil
22 ¯
23
(al-šarr bi’l-‘arad). Essential evil is privation (‘adam), that is, not just any
˙
24
privation, but the privation of a perfection that is required by the nature of the
25
thing. Accidental evil is that which produces privation17 or that which withholds
26
perfection from that which deserves it.18 The difference is that essential evil is
27 nothing but evil, while accidental evil is not evil in itself. For example, clouds
28 that withhold sunshine from plants that need the sun for their growth are not in
29 themselves evil. The effect, however, is privation.
30
31 12 Ilāh. IX.6, pp. 414.17 – 415.7 (339.4 – 12); see Steel, Avicenna and Thomas Aquinas,
32 p. 174.
33 13 Ilāh. IX.6, p. 415.8 – 10 (339.13 – 15).
34 14 Ibid. p. 419.6 – 7 (343.16 – 17).
15 Ibid. p. 419.4 – 6 (343.14 – 16).
35
16 Ibid. p. 419.9 – 10 (344.1 – 2).
36 17 That which produces privation: this rendition presupposes the reading al-mu‘dim as
37 adopted by Bertolacci, Reception, p. 549. The Cairo edition has al-ma‘dūm, p. 416.7
38 (340.12). The Latin translation renders the Arabic passive participle, cf. Steel, Avicenna
39 and Thomas Aquinas, p. 176. For an evaluation of the two different readings, see Inati,
The Problem of Evil, pp. 86 – 7. Unfortunately, Ibn Daud’s ER cannot shed light on the
40
issue as he does not reproduce this sentence, which has to do with his silence on the
41 notion of accidental evil, cf. below.
42 18 Ilāh. IX.6, pp. 415.18 – 416.5 (340.6 – 13).
1 Since evils are privations they can only occur in things that have potentiality
2 and this is because of matter.19 Matter may or may not receive its ultimate form,
3 but if it does not, this is due to the lack of receptivity of matter, or because of
4 some external impediment (such as the clouds that withhold sunshine from that
5 which needs the sun for its growth), and not because of an agent who denies the
6 perfection. However, Avicenna also concedes that a more active external agent
7 may be at work, for example, an opposing element, as when cold strikes plants
8 and thus removes their perfection.20 Yet since evil is connected with matter, it is
9 found only in the sublunary world, that is, the world of generation and
10 corruption. This implies that evil is in fact insignificant as the sublunary world
11 is only a small part of the universe.21 Avicenna has more good news: evil afflicts
12 only individuals, not the species, and only at certain times. In fact, it is rather
13 the exception than the rule. His stock-example is the effect of fire: fire can
14 indeed cause great harm to individuals, but on the whole it is beneficial. Evil is
15 thus a necessary concomitant of the good. For Avicenna a good can only be a
16 good if evil also can ensue from it and with it.22 It is the good thing that is willed
17 by God primarily and this is what happens for the most part.
18 Yet even if good far outweighs evil, why should there be evil at all? Could
19 not the First Governor have brought into existence a world that is entirely free
20 from evil? This would be possible, Avicenna agrees, in ‘absolute existence’, but
21 such a world would not be our world. By ‘absolute existence’ Avicenna means
22 the intellectual, psychological and celestial things that emanate from God, that
23 is the superior, good causes. However, if emanation stopped at the level of these
24 causes, there would be no need for the activity of the higher causes, and the
25 universe would be less complete than it is now. Withholding evil, Avicenna
26 concludes, would be the greatest fault in the universal order of the good.23
27 Finally, he argues that even though evils may be numerous, like illnesses,
28 they are not preponderant. Indeed there is a kind of evil that is preponderant,
29 like ignorance in a certain field (for example, geometry) or a lack of radiant
30 beauty, but these fall under the category of superabundance and excess. Being an
31 outstanding geometer or extremely beautiful does not belong to the category of
32 primary perfections. Again, such evils have to do with a deficiency in the
33 recipient, and are not due to action on the part of an agent, but rather to
34 inaction.24
35
36 19 Ibid. p. 416.9 – 10 (340.16).
37 20 Ibid. pp. 416.1 – 417.4 (340.18 – 341.7).
38 21 Ibid. p. 417.4 – 5 (341.8 – 9).
39 22 Ibid. p. 418.5 – 6 (342.9).
23 Ibid. p. 418.12 – 20 (342.17 – 343.8). As Steel, Avicenna and Thomas Aquinas, p. 185
40
puts it: ‘the universe would be less perfect than it is now. For a possible level of existence
41 would be lacking.’
42 24 Ibid. p. 422.9 – 18 (347.3 – 12).
1 In sum, the key-notions in Avicenna’s theodicy are the following: evil is

2 privation, and is associated with potentiality and matter. As for its presence in
3 this world, Avicenna justifies it by stating that: 1) the sublunary world is only a
4 small part of the universe; 2) it is rare since it only strikes individuals; 3) it is a
5 necessary concomitant of the good; 4) without it the universe would be less
6 perfect. In view of these ‘solutions’ Carlos Steel has characterized Avicenna’s
7 metaphysics as ‘optimistic.’25 If Avicenna is indeed an optimist, then Ibn Daud
8 also qualifies as one, as I will try to show in what follows.
9
10
11 3. Ibn Daud’s Use of Avicenna’s Account of Evil
12
13 It can immediately be noted that Ibn Daud’s discussion of evil bears a great
14 similarity to that of Avicenna as articulated in the Šifā’.26 I shall briefly mention
15 its main features before addressing them in greater detail. Like Avicenna, Ibn
16 Daud treats the problem of evil in connection with the question of God’s
17 knowledge and providence, and as in the Šifā’, the discussion is found towards
18 the end of his book.27 The two expositions display roughly the same order of
19 treatment and share the same basic views: evils are privations; evil is due to
20 matter, and thus found only in the sublunary world. Ibn Daud also accepts
21
Avicenna’s ‘justifications’ for the presence of evil in this world and they
22
constitute in fact the main ingredients of his theodicy.
23
Such a theodicy is not found in the writings of earlier Jewish philosophers.
24
One of Ibn Daud’s predecessors, Abraham bar Hiyya (d. 1136?) of Barcelona,
25 ˙
mentions the theory of privation in his ethical-philosophical treatise The
26
Meditation of the Sad Soul (Hegyon ha-nefeš ha-‘asuvah) as a solution put forward
by ‘the philosophers’ to account for the presence˙of evil. He does not refer to the
27
28
term ‘privation’, however, and he rejects the theory.28 Another ‘Spanish’
29
Abraham, the many-sided scholar Abraham Ibn Ezra, who was an older
30
contemporary of Ibn Daud, views the imperfection of the recipient as the origin
31
of evil, as does Avicenna, and also holds that it is not fitting for superior wisdom
32
33
to preclude the greater good because of small evil.29 However, the combination
34
25 Steel, Avicenna and Thomas Aquinas, p. 181.
35
26 For Ibn Daud’s discussion of evil, see also Eran, From Simple Faith, pp. 236 – 51 and
36 Fontaine, In Defence of Judaism, pp. 196 – 204. These monographs treat Ibn Daud’s
37 account of evil within the framework of an overall analysis of his thought, and in relation
38 to other sections of his book. The present paper focuses more specifically on Ibn Daud’s
39 use of Ibn Sina.
27 Abraham ibn Daud, Ha-Emunah ha-Ramah II.6.2, ed. Weil 1851, pp. 93 – 8 (S 201a12 –
40
208b11).
41 28 Hegyon ha-nefeš, ed. Wigoder, pp. 123 – 4.
42 29 Comm. on Ecclesiastes, Introd.
1 of views on evil as found in The Exalted Faith is new to Jewish philosophy, and
2 therefore, we can safely assume that Ibn Daud drew on Avicenna’s exposition. In
3 fact, this conclusion is wholly consistent with Ibn Daud’s stated declaration,
4 referred to above, that he needed ‘true philosophy’, that is, the philosophy of
5 Avicenna, to solve the problem of free will. After all, this problem is closely
6 connected to that of evil, and Avicenna is the first Muslim philosopher to
7 provide a systematic treatment of the problem. However, as we will see, Ibn
8 Daud’s treatment of the topic is far from a slavish copy of that of Avicenna.
9 Ibn Daud addresses the issue in the penultimate chapter of his book under
10 the heading ‘On the origin of good and evil’,30 which is subtitled: ‘On the order
11 and number of causes, providence, and the secret of [God’s] power.’ To this he
12 adds: ‘This is the chapter because of which we have composed the book.’ Ibn
13 Daud does not provide an explicit formulation of what he understands by ‘the
14 problem of evil’, nor does he offer anything that resembles a ‘classification’ of
15 evils. However, it may be inferred from his account that like Avicenna he stands
16 in the tradition of the Stoics and Neoplatonists to whom the question was: ‘how
17 can evil exist in a world that proceeds from the One who is perfect?’31 His
18 starting point is the thesis that God knows His own essence. He then goes on to
19 sketch the structure of the universe: the farther a thing is removed from matter,
20 that is, potentiality, the more complete is its knowledge, since ‘Every deficiency
21 and every evil originates in what has something in potentiality’, a statement that
22 is a close rendering of Avicenna’s aforementioned view: ‘Evil attaches only to
23 that which has in its nature what is potential.’32 God, who is furthest removed
24 from potentiality, knows His essence in the most perfect way and He also knows
25 that His perfection is not intended towards Himself alone, but that it emanates
26 to others in progressive degradation. Thus, existence has two extremes: at the
27 one end that which is perfect and does not admit of any deficiency, and at the
28 other that which we call ‘matter’, which is removed from perfection to the
29 utmost degree.33 All this is more or less a summary of his views as developed
30 earlier in his book, views that are taken from the falāsifa. Yet at the same time
31 this point of departure with its emphasis on God’s knowledge, perfection,
32 emanation and the hierarchical order of the universe displays all the ingredients
33 of Avicenna’s conception of providence. Unlike Avicenna, however, Ibn Daud, is
34 careful not to present this position as entailing providence, because, as we shall
35 see, he has a different understanding of what providence is.
36
37 30 Or, according to some mss, ‘on free will’.
38 31 For a survey of the treatment of this dilemma by Greek and Latin thinkers, see J.
39 Opsomer and C. Steel, Evil without a cause; Proclus’ doctrine on the origin of evil, and
its antecedents in Hellenistic philosophy, in Zur Rezeption der hellenistischen Philosophie
40
in der Sptantike, eds T. Fuhrer et al., Stuttgart: Franz Steiner Verlag, 1999, pp. 229 – 61.
41 32 ER p. 93.28 (S 201b3 – 4), cf. Ilāh. IX.6, p. 416.9 – 10 (340.16).
42 33 ER p. 93.25 – 41 (S 201a16 – 202a2).
1 Ibn Daud’s next step is to emphasize that according to reason, Scripture, and
2 the Jewish exegetical tradition, it is impossible that evil or deficiency proceed
3 from God.34 Despite Ibn Daud’s categorical assertion, it should be noted that
4 the position that God produces evil alongside the good was in fact defended by
5 one of his compatriots, the aforementioned Abraham bar Hiyya, in an exegetical
˙
6 context. This philosopher based himself on Isaiah 45:7 where God is said to be
7 ‘the maker of peace and the creator of evil’, accepting the literal meaning of the
8 text.35 Given Ibn Daud’s overall aim to reject the literal meaning of biblical
9 verses when they contradict reason he obviously could not adopt this position.36
10 As he argues, God would be composite and not One if two opposites were
11 united in him. God is not like man, from whom can proceed both good (on
12 account of his rational soul) and evil (on account of his lower faculties).37 Since
13 it is likewise impossible that only evil proceeds from God, the question of the
14
origin of evil imposes itself.
15
Thus far we have heard that deficiency and evil are due to matter and
16
potentiality, and to the lower faculties of the soul. Ibn Daud then introduces
17
rather abruptly the notion of ‘privation’ with the statement that ‘among the
18
attributes of deficiencies some are privative (he‘ederiyyim) and some are positive
19
(hiyyuviyim).’38 This statement is puzzling, since Ibn Daud does not make clear
20 ˙
what he means by ‘positive’ attributes of deficiencies, nor does he provide any
21
examples of them. In my view, this distinction can be taken to reflect the
22
23
distinction made by Avicenna between evils like ignorance and deformities that
24
are privations of being, and evils like pain and distress, which are apprehended
25
as very present and real, that is, as existing things.39 If this interpretation is
26
correct, perhaps a better translation of hiyyuviyim would be ‘existing.’40 Ibn
˙
27 Daud shows himself to be more interested in the privative evils, of which
28 darkness is an example. As he explains: ‘light is something in the air that is
29 produced (mithaddeš) when a luminous body shines on it, and when the
˙
30 luminous thing is removed, it remains dark, and nothing is produced in it,
31 rather the produced thing departs from it. Thus, darkness is the privation of a
32
33
34
34 ER p. 93.42 – 3 (S 202a3 – 5).
35
35 Abraham bar Hiyya, Hegyon ha-nefeš, pp. 119 – 22.
36 36 ˙ – 4 (S 2b22 – 5)
ER intro, p. 1.31
37 37 ER pp. 93.45 – 94.3 (S 202a6 – 11).
38 38 ER p. 94.4 – 5 (S 202a11 – 12).
39 39 Ilāh. IX.6, p. 419.6 – 7 (343.16 – 17), cf. also p. 415.12 – 13 (340.4 – 5). Cf. Steel,
Avicenna and Thomas Aquinas, pp. 174 – 6.
40
40 The second Hebrew translation, Ha-Emunah ha-Nisa’ah, reads mehuyyavim (necessary).
41 The Arabic has: wuǧūdiyya, which seems to underlie the reading of ˙ER. The term hiyyuvi
42 also occurs in ER I.3 in the sense of ‘actualization’, ‘realization’. ˙
1 thing, but it is not a thing [itself ].’41 Privations are not made by God, or by any
2 other agent, because they are not beings that can be produced. Thus, one cannot
3 say that God made the privation of elephants in Spain, but rather that He did
4 not make an elephant in Spain. In other words, there is inaction on the part of
5 God, not action. If this absence is a deficiency in Spain, God did not produce
6 it; He just did not make the perfection that removes the deficiency, and left it
7 undetermined whether or not there would be elephants in Spain.42 From
8 darkness and absence of elephants, Ibn Daud moves on to ignorance. The same
9 applies here: when a person’s intellect remains intellect in potentiality, it cannot
10 be claimed that God ‘produces’ ignorance. God merely did not decree that this
11 intellect would become intellect in actuality, so that it remains on the level of
12 matter. Hence, since most vices and defects have their origin in a lack of
13 intellect, God cannot be their maker.43
14 At this point, however, Ibn Daud anticipates an objection. It could be
15 argued, he says, that the two cases are not comparable, ‘because it does not
16 belong to the nature of Spain to have elephants walking around there, and their
17
absence is not viewed as a deficiency, whereas it certainly belongs to the nature
18
of man to possess intellect and perfection, and their absence is seen as a
19
deficiency.’44 Such a privation seems incompatible with God’s goodness, for
20
withholding perfection may be seen as evil.
21
Ibn Daud’s response to the objection is based on Avicenna’s contention that
22
imperfection is to be ascribed to matter. It is found only in that which is
23
composed of the elements, in other words, in the sublunary world. This forces
24
him to account for the occurrence of evil in our world, which is why he argues
25
that the reception of form by matter may be prevented by some external
26
impediment, so that a thing will not attain its natural end. However, instead of
27
28
Avicenna’s example of clouds that withhold sunshine, Ibn Daud evokes another
29
image:
30 When a sower sows seed on the land, a part will fall in fertile moist ground, well-
31 positioned with respect to the sun, so that the mixture of the seed can benefit from
32 it and lo, it will spring up and grow well. But the part that falls unto dry barren
ground, or ground that is not positioned thusly that the mixture can benefit from
33
the sun45 will grow up only weakly. The part that will fall onto stony ground, will
34
35
41 ER p. 94.7 – 8 (S 202a13 – 16). Another example is poverty, which is the privation of the
36 opposite, namely the presence of money. The Hebrew term used for privation (he‘der)
37 renders the Arabic ‘adam, and like the Arabic it can also be translated as non-being or
38 absence.
39 42 ER p. 94.8 – 16 (S 202a16 – 202b8).
43 ER p. 94.16 – 24 (S 202b8 – 203a2).
40
44 ER p. 94.25 – 9 (S 203a2 – 5).
41 45 The reading ‘sun’ (ha-šemeš ) is supported by various manuscripts. Weil’s printed text has
42 sadeh (field).
1 not spring up at all, it will be blown away by the storm scattered or eaten by birds or
2 it will become foul.46
3 The ultimate source of this image is of course the parable of the sower of the
4 evangelists.47 As Amira Eran has shown, however, Ibn Daud’s direct sources were
5 al-Ġaza-lı̄’s Maqāsid and the collection of stories entitled Kitāb al-malik wa’l-
6 nāsik (The Prince ˙ and The Ascetic) in Arabic.48 We can add that in another
7
context Avicenna himself employs the example of seed that does not fall on this
8
particular spot of land always or for the most part, namely in his expositions on
9
chance in the Physics of his Šifā’.49
10
11
Next he brings up Avicenna’s argument according to which evil is small when
12
compared to the good, which is far more widespread. This contention
13
introduces Ibn Daud’s attempt to minimize and even legitimize evil. In so
14
doing, he follows in the footsteps of the Muslim philosopher, for Avicenna’s
15
various strategies to this purpose appear in The Exalted Faith. Ibn Daud,
16
however, gives them a different flavor by choosing his own examples, which are
17
drawn from Rabbinic literature. Here we come across Avicenna’s question
18
whether the First Governor could not have brought into being a world that is
19
entirely free from evil. Ibn Daud’s wording, however, is slightly different: the
20
Creator is under no obligation to create a world in which everything reaches its
21
final perfection, and he goes on to say:
22
23 Whosoever says that this should have been the case and that the Creator should have
24 chosen the best is saying something that is absurd. If the Creator had chosen the
25
best for everything, then necessarily all plants would be like animals, all animals like
man, and all man like Moses, our Rabbi, and Moses our Rabbi, like the most
26 sublime of the angels.50
27
28 In such a perfect world there would be no order, or hierarchy. All existence
29 would be uniform for everything would have reached its final perfection. Like
30 Avicenna, Ibn Daud thus believes that a world in which there is hierarchy and
31 order is better than a world in which there is only good. As Ibn Daud puts it
32 rather poetically: ‘existence would come to an end by the very power of its
33 perfection, like something that is cut in its prime’, a phrase that is inspired by a
34 verse from Job.51 He adds that in such a world God’s perfection would emanate
35 only on few existents while God’s goodness is great and is intended to emanate
36 on many things.
37
38 46 ER p. 94.34 – 9 (S 202b8 – 203a2).
39 47 Cf. Matt. 13:3 – 9; Mk. 4:3 – 9; Lk 8:5.
48 Eran, The Parable of the ‘Sower and the Seeds’, pp. 139 – 51.
40
49 See Belo, Chance and Determinism, p. 43.
41 50 ER pp. 94.42 – 5.
42 51 ER pp. 94.ult.–95.1 (S 203b11 – 12); cf. Job 8:12.
1 At this juncture Avicenna’s thesis that evil is a necessary concomitant of the

2 good makes its appearance in Ibn Daud as well. The Jewish thinker refers to
3 tiqqun ha-kelal (improvement of the whole), which renders Avicenna’s nizzām
4
˙ ˙ by
al-hayr al-kullı̄ (universal order of the good).52 Ibn Daud supports the thesis
5 ˘
means of Rabbinic dicta. It is in this context that the Talmudic quotation of the
6 title of this paper is found. The full quotation is:
7
The world cannot do without perfume-makers and tanners; blessed is he whose
8
occupation is that of perfume-maker and unhappy he whose occupation is that of
9 tanner. The world cannot do without men and women. Blessed is he whose children
10 are boys and unhappy he whose children are girls.53
11
12
To this he adds that every individual wishes to have boys, for this is generally
13
regarded as perfection, but luckily God considers the good of the whole. For the
14
same reason, he continues, the High Priest used to implore God not to heed the
15 prayer of travelers who always ask for fine weather, for without rain all
16 harvesting would come to an end.54 This particular example may also have been
17 inspired by al-Ġaza-lı̄’s rendering of Avicenna’s views on evil in the Maqāsid, for
˙
18 he likewise writes that without rain all sowing would come to an end.55
19 In his elaboration of tiqqun ha-kelal, Ibn Daud goes a long way in agreeing
20 with Avicenna. As was to be expected, he also employs the stock-example of fire,
21 claiming that damage wrought by fire or wind is a necessary concomitant of the
22 good, and, again, that these evils are not predominant.56 Moreover, like
23 Avicenna he accepts that universal order necessitates the existence of individuals
24 with some kind of evil in their bodies or souls. Ibn Daud, however, goes into
25 more detail with respect to what he calls ‘deficient species’. These species include
26 people whose lack of intellect is counterbalanced by a strong physical
27 constitution, which makes such people fit to do hard labor, like animals,
28 which is beneficial to society as a whole.57 Ibn Daud thus refers both to
29 universal order and to order in society, which may explain the use of the term
30
31
32
33 52 Avicenna means by al-nizzām al-kullı̄: ‘that the whole [is such] that the active and passive
˙˙
powers, celestial and terrestrial, natural and psychological, have been organized in it in
34
such a way only so that they would lead to the universal order’ (tr. Marmura, p. 346).
35
53 Cf. n. 1. ER p. 95.5 – 10 (S 204a4 – 8). The existence of women is thus not an evil in
36 itself, as it serves a good end, but individuals may view the birth of daughters as evil,
37 since all people wish to have sons. Cf. Aristotle’s notion that the formation of the female
38 is in fact a deviation, even though nature requires the existence of females, see his
39 Generation of Animals XVIII, 767b8 – 10.
54 ER p. 95.13 – 16 (S 204a810 – 13), cf. b Yoma 53b and Ta‘anit 24b.
40
55 Maqāsid, ed. S. Dunyā, p. 298.
41 56 ER p.˙ 95.26 – 33 (S 204b8 – 13); cf. Ilāh. IX.6, p. 418.3 – 4 (342.6 – 8).
42 57 ER p. 95.16 – 26 (S 204a13 – 204b8).
1 tiqqun (improvement) of the whole here for Avicenna’s nizzām.58 Ibn Daud even
2 goes so far as to say that people whose temperament is ˙such ˙ that they quarrel
3 and kill due to an excess of heat in their bodily constitution have their own place
4 in creation just like bears and dogs:
5
In the matter out of which the wicked sinful person comes to be there is an excess of
6 heat, be it in the constitution of the seminal fluid or in the womb, or both, and
7 following the heat that has assembled in [this person’s] heart in an unbalanced
8 proportion, a certain disposition will occur, which is called anger. As a result, [such a
9 person] will fight with people and be ready to bite them, as dogs do, and it may
10 happen that he will slay and kill them, like wild animals do. But just as it is not
strange that there should exist dogs and wild animals, it is not strange that there
11
should be among the species man a species which resembles man, but whose nature
12 is like that of animals.59
13
14 This position may appear surprising at first sight, for it seems to leave little hope
15 for those who are the unfortunate possessors of such a constitution. A possible
16 explanation would be that Ibn Daud’s intention is to draw a parallel between
17 metaphysical and moral evil, and to finally state his view on human freedom.
18 His problem is again, or still, the absence of intellect. It is this particular issue on
19 which Ibn Daud’s account of evil focuses and to which it gravitates. In the
20 passage discussed above (cf. above, text to n. 44) we learned that lack of
21 intellect, viewed on the physical level, does not come from God; instead it
22 represents a case of a potentiality in matter that has failed to become actualized.
23 In the present passage Ibn Daud argues that moral evil cannot come from God
24 either: evil actions result from a deficiency in one’s physical constitution, in
25 other words, they also go back to matter. However, he hastens to add that people
26 who unfortunately lack knowledge and cannot distinguish right from wrong are
27 not left to their own devices, for God has sent people who are entirely free from
28 evil, that is, the prophets who brought commandments and warnings.60
29 It is at this point that Ibn Daud turns to his defense of free will. Here his
30 understanding of the ‘possible’ plays a crucial role, the possible being a class of
31 existents alongside necessary and impossible things.61 According to Ibn Daud, it
32 is not impossible for man to oppose his bad character traits, for God does not
33 command the impossible. Nor is it necessary, like breathing. Instead it is
34 possible. Just as God left it undetermined whether or not there would be
35
36 58 The more literal rendering siddur (order) also occurs in the Hebrew text. Of course we
37 do not know what the Arabic original had.
38 59 ER p. 95.33 – 6 (S 205a4 – 8). I adopt the reading zar (strange) here, with most of the
39 manuscripts. Weil’s printed text has zeh (this), which does not make sense. (S 205a5 – 6
also reads zeh, but translates ‘strange’).
40
60 ER pp. 95 ult.–96.3 (S 205a15 – 16).
41 61 Earlier on in ER Ibn Daud had already argued that the impossible in fact does not exist,
42 ER p. 47.41 – 4 (S 126b4 – 5).
1 elephants in Spain, He left it undetermined which course of action man will

2 choose for himself. Freedom of action is guaranteed since God created the
3 possible as possible and knows it only as possible.62 At this point of his
4 discussion, Ibn Daud introduces the notion of causality. Advancing a fourfold
5 division of causes into divine, natural, accidental and free causes, he holds that it
6 is thanks to the existence of the last type of cause that man is free to act.63
7 While Ibn Daud’s focus in this section is on free will, the problem of evil
8 also plays a part here, for he contends that man can protect himself against the
9 evil results of natural and accidental causes, for example by refraining from
10 certain kinds of food and lethal drugs. The introduction of accidental causes at
11 this point is not without difficulty in view of Ibn Daud’s earlier account of evil.
12 As has been shown above, there are many points of contact between his and
13 Avicenna’s treatment of the problem. There is, however, a significant difference
14
in that Ibn Daud nowhere mentions Avicenna’s aforementioned distinction
15
between essential and accidental evil, a distinction that is crucial for Avicenna.
16
Similarly, Ibn Daud bypasses Avicenna’s position that evil was willed by God in
17
an accidental way.64 I suspect that his silence has to do with his insistence that
18
evil has no agent and can by no means proceed from God. To be sure, both
19
thinkers hold that privations are not due to an action on the part of God, but
20
rather to inaction, but nonetheless Avicenna seems to designate the agent as a
21
cause of evils that are attached to good things.65 Ibn Daud is more emphatic
22
with regard to inaction by the agent, especially when it comes to accounting for
23
24
the absence of intellect in certain individuals, where he underlines that God is
25
not the cause of ignorance. Avicenna does not single out this particular point for
26
discussion, but for Ibn Daud it is most relevant, since it bears on his defense of
27
free will. Stating that evil proceeds from God’s will, albeit accidentally, would
28 imply that to some extent God is the agent of ignorance either by action or
29 inaction, and hence, that He may ultimately be held responsible for any bad
30 actions that result from ignorance. It is therefore possible that it is Ibn Daud’s
31 eagerness to avoid this notion that led him to ignore Avicenna’s distinction
32 between essential and accidental evil.66 In this regard it is also noteworthy that
33 Avicenna’s classification of evils has no counterpart in Ibn Daud: Ibn Daud
34 seems primarily interested in one particular evil: lack of intellect.
35 However, the notion of accidental evil, or of evil brought about by an
36 external agent is not absent from ER, as is clear from the aforementioned parable
37
38 62 ER p. 96.21 – 4 (S 206a5 – 8).
39 63 ER 96.33 – 97.8. (S 206b1 – 207a7).
64 Ilāh. IX.6, pp. 420.19 – 421.1 (345.11).
40
65 Ibid., lines 11 – 15, Cf. Inati, The Problem of Evil, pp. 91 – 2.
41 66 Belo, Chance and Determinism, p. 117 concludes that, according to Avicenna’s Naǧāt,
42 ‘evil too is willed and created by God’.
1 of the sower. Ibn Daud obviously admits that some effects do not attain their
2 intended goal. Though he ascribes this failure to lack of receptivity of matter, it
3 is difficult not to view the storm or the birds of Ibn Daud’s example as external
4 agents. Moreover, in the present passage on causality the implication that evil
5 ultimately comes from God, albeit indirectly, or, as Ibn Daud puts it not ‘by
6 primary intention’, because God entrusted some causes to nature, seems
7 unavoidable. Our author, however, does attempt to avoid it by claiming that it is
8 given to man to be on his guard against evil effects and that he is helped in this
9 by revealed law.
10
11 At any event, in his defense of free will Ibn Daud cannot follow Avicenna any
12 longer. Of course it can be and has in fact been argued that Avicenna also seeks
13 to absolve God from any blame for moral evil and even that he allows for some
14 human freedom, but this does not come to the fore in the Šifā’, Ibn Daud’s most
15 likely source.67 In the relevant chapter of the Šifā’ Avicenna quotes with approval
16 prophetic traditions like ‘I created these for the fire and I care not; and I have
17 created these for paradise and I care not’,68 and, given Ibn Daud’s overall aim,
18 the Jewish philosopher cannot but reject Avicenna’s position. He has to defend
19 human freedom forcefully because it is on free will that man’s acceptance of the
20 biblical commandments hinges, and for Ibn Daud the end of ‘theoretical
21 philosophy’ is practice, that is, religious practice.
22 At the end of his discussion Ibn Daud offers his own view of providence,
23 and here he definitely departs from Avicenna. While for Avicenna, as may be
24 recalled, providence consists in God knowing Himself as the cause of the good,
25 for Ibn Daud providence lies in the assistance on the part of God and the angels
26 that is granted to man following man’s own initial choice. When man chooses
27 the good, he will be helped to pursue his goal. However, when one opts for evil,
28 one will be directed on that path. In this manner, the biblical verse ‘I will harden
29 Pharaoh’s heart’ must be understood, for Pharaoh, by his own free will, refused
30 to let the Israelites go, and he was helped to persist in his choice accordingly.69
31 We can thus conclude that despite the obvious similarities between
32 Avicenna’s and Ibn Daud’s treatments of evil, there are important differences
33 between the two philosophers. Ibn Daud treats the topic of evil in a wider
34 conceptual framework, since his ultimate goal is the defense of free will, the key-
35
36 67 Avicenna’s expositions on determinism have given rise to divergent interpretations: while
37 scholars like G. Hourani and C. Belo view Avicenna’s position as precluding human
38 freedom, A. Ivry and J. Janssens have argued that Avicenna’s system leaves some room for
39 free human action. See Hourani, Ibn Sı̄nā’s Essay on the Secret of Destiny, pp. 40 – 41;
Belo, Chance and Determinism, pp. 52 – 3 and pp. 116 – 20; Ivry, Destiny Revisited,
40
pp. 167 – 70; Janssens, Problem of Human Freedom, pp. 112 – 18.
41 68 Ilāh. IX.6, p. 422.7 – 8 (347.1 – 2).
42 69 ER p. 98.4 – 7 (S 208b2 – 6).
1 theme of The Exalted Faith. This circumstance explains his focus on the lack of
2 intellect, for this problem constitutes a major obstacle in upholding human
3 freedom. Yet at the same time his discussion is narrower and less detailed than
4 that of his source precisely because of his selective use. In other words, Abraham
5 Ibn Daud follows Avicenna only to the extent that the viewpoints of the Muslim
6 philosopher suit his own purposes, while at the same time he displays a
7 considerable independence from his source.
8 In struggling with the dilemma evil versus providence Avicenna and Ibn
9 Daud put forth views and arguments that are well known from Neoplatonic and
10 Stoic sources.70 It is difficult to determine to which tradition exactly Ibn Daud
11 belongs. His account reveals a mixture of different and sometimes conflicting
12 views: with Plotinus he maintains that evil is to be ascribed to matter and
13 privation; the notion that evil is a side effect, some kind of collateral damage
14 that has no agent echoes Proclus’ position, while Stoic thought is the source for
15 the view that evil when viewed from a universal perspective is only apparent evil,
16 and Ibn Daud adduces biblical passages in support of all these viewpoints.
17 In view of his use of these stock arguments it can be claimed that to some
18 extent Ibn Daud’s explanation of evil is ‘unoriginal’. Yet, from the perspective of
19 the history of medieval Jewish philosophy, it can be called innovative, precisely
20 because of his use of Avicenna and the manner in which he adapts his doctrine.
21 As we have seen no Jewish author before him had attempted to provide a
22 philosophical systematic discussion of the issue. The question whether Ibn
23 Daud’s use of Avicenna has left traces in the works of later medieval Jewish
24 philosophers remains to be explored. A few decades later, Maimonides addressed
25 the problem of evil in his Guide, taking issue with several of the views discussed
26 above, such as the existence of physical and moral evil, divine providence and
27 omniscience. Much has been written already on Maimonides’ positions on these
28 themes.71 However, what is still lacking is a comprehensive study that takes into
29 consideration Maimonides’ use of Avicenna’s theodicy in relation to Ibn Daud’s
30 use. Such a study, it is expected, will throw more light on the reception of
31 Avicenna and his theodicy in medieval Jewish philosophy.
32
33
34
35
36 70 For discussions of Ibn Sı̄nā’s sources, see Hourani, Ibn Sı̄nā’s Essay on the Secret of Destiny,
37 and Inati, The Problem of Evil, pp. 54 – 65. The question to what extent the ‘solutions’
38 that the Muslim and the Jewish philosopher put forth were acceptable to their respective
39 audiences is intriguing, but does not fall within the scope of this paper.
71 See Burrell, Maimonides, Aquinas and Gersonides; Rosenberg, Good and Evil, pp. 24 –
40
31; W.Z. Harvey, Maimonides and Spinoza on the Knowledge of Good and Evil; Pines,
41 Truth and Falsehood; Dobbs-Weinstein, Matter as Creature; Motzkin, Maimonides and
42 Spinoza on Good and Evil; Leaman, Evil and Suffering, pp. 64 – 101.
1 Bibliography
2
3 Primary Sources
4
5 Abraham bar Hiyya, Hegyon ha-nefeš ha-‘asuvah, ed. G. Wigoder, Jerusalem: Mosad
Bialiq, 1971.˙ ˙
6
Abraham Ibn Daud, Ha-Emunah ha-Ramah, ed. S. Weil, Frankfurt a. M., 1852.
7
–– , Ha-Emunah ha-Ramah, The Exalted Faith, eds N.M. Samuelson and G. Weiss,
8 London/Toronto: Associated University Presses, 1986.
9 al-Ġaza-lı̄, Metaphysica, Maqāsid al-falāsifa, ed. S. Dunyā, Cairo: Da-r al-Ma a-rif, 1960.
˘
10 ˙
Avicenna, Kitāb al-Šifā’, Ilāhiyyāt, vol. 2, eds M.Y. Mūsā, S. Dunyā, Cairo: al-Hay’a al-
11 ‘āmma li-šu‘ūn al-matābi‘ al-amı̄riyya, 1960.
12 –– , The Metaphysics of ˙‘The Healing’, transl. and ed. M. Marmura. Provo, Utah:
13
˘
–– , Kitāb al-Nağāt fı̄ al-hikma al-mantiqiyya wa-al-tabı̄ iyya wa-al-ilāhiyya, ed.
14
M. Fakhry, Beirut: Dar al-Afaq ˙ ˙ 1985.
al-Jadidah, ˙
15 –– , Kitāb al-Išārāt wa’l-tanbı̄hāt, ed. S. Dunyā, Cairo: Dār al-Ma ārif, 1958.
˘
16
17
18 Secondary Sources
19
20 C. Belo, Chance and Determinism in Avicenna and Averroes, Leiden/Boston: Brill, 2007.
21 A. Bertolacci, Albert the Great and the Preface of Avicenna’s Kitāb al-Šifā’, in Avicenna
22 and his Heritage, eds J. Janssens and D. De Smet, Leuven: Leuven University Press,
23
2002, pp. 131 – 52.
–– , The Reception of Aristotle’s ‘Metaphysics’ in Avicenna’s Kitāb al-Šifā’: a Milestone of
24
Western Metaphysical Thought, Leiden: Brill, 2006.
25 D.B. Burrell, Maimonides, Aquinas and Gersonides on Providence and Evil, Religious
26 Studies, 20/3, 1984, pp. 335 – 51.
27 I. Dobbs-Weinstein, Matter as Creature and Matter as the Source of Evil: Maimonides
28 and Aquinas, in Neoplatonism and Jewish Thought, ed. L.E. Goodman, Albany, New
29
York: State University of New York Press, 1992, pp. 217 – 35.
A. Eran, From Simple Faith to Sublime Faith. Ibn Daud’s Pre-Maimonidean Thought, Tel
30
Aviv: Ha-kibbutz ha-me’uhad, 1998 (in Hebrew).
31 –– , The Parable of the ‘Sower and the Seeds’ in the Exalted Faith, Tarbiz, 73, 2003,
32 pp. 139 – 51(in Hebrew).
33 Y. Essudri, R. Avraham bn Daud, R. Efraim al-Naqawa we-sefer ‘al-‘Aqı̄da al-rafi‘a’ (in
34 Hebrew, forthcoming).
35
R. Fontaine, Abraham Ibn Daud, Encyclopedia Judaica, vol. 9, 2nd edn, Jerusalem: Gale
Cengage Learning, 2006, pp. 662 – 5.
36
–– , Abraham Ibn Daud, The Stanford Encylopedia of Philosophy, at http://
37 plato.stanford.edu/entries/abraham-daud.
38 –– , In Defence of Judaism: Abraham Ibn Daud. Sources and Structure of ‘ha-Emunah ha-
39 Ramah’, Assen: Van Gorcum, 1990.
40 R. Fontaine and S. Harvey, Jewish Philosophy on the Eve of the Age of Averroism: Ibn
41 Daud’s Necessary Existent and his Use of Avicennian Science, in The Age of
Averroism, ed. P. Adamson, London: The Warburg Institute, 2011, pp. 215–27.
42
1 D. Gutas, What was there for the Latins to Receive? Remarks on the Modalities of the
2 Twelfth-Century Translation Movement in Spain, in Wissen ber Grenzen,
3
Miscellanea Mediaevalia 33, ed. A. Speer and L. Wegener, Berlin: de Gruyter,
2006, pp. 3 – 21.
4 S. Harvey, Avicenna’s Influence on Jewish Thought: Some Reflections, in Avicenna and
5 his Legacy: A Golden Age of Science and Philosophy, ed. Y.T. Langermann, Turnhout:
6 Brepols, 2009, pp. 327 – 40.
7 W.Z. Harvey, Maimonides and Spinoza on the Knowledge of Good and Evil, Binah 2,
8
1989, pp. 131 – 46.
9 Philosophy of the Soul, 1160 – 1300, London/Turin: The Warburg Institute/Nino
11 G.F. Hourani, Ibn Sı̄nā’s Essay on the Secret of Destiny, Bulletin of the School of Oriental
12 and African Studies, 29, 1966, pp. 25 – 48.
13
S. Inati, The Problem of Evil: Ibn Sı̄nā’s Theodicy, Binghamton, New York: Global
Publications, 2000.
14 A. Ivry, Destiny Revisited: Avicenna’s Concept of Determinism, in Islamic Theology and
15 Philosophy: Studies in Honor of George F. Hourani, ed. M. E. Marmura, Albany:
16 SUNY Press, 1984, pp. 160 – 71.
17 J. Janssens, The Problem of Human Freedom in Ibn Sı̄nā, in Ibn Sı̄nā and his Influence
18
on the Arabic and Latin World, ed. J. Janssens, Aldershot: Ashgate, 2006, pp. 112 –
18.
19 O. Leaman, Evil and Suffering in Jewish Philosophy, Cambridge: Cambridge University
20 Press, 1995.
21 J. Michot, La destine de l’homme selon Avicenne: le retour Dieu (ma’ad) et l’imagination,
22 Leuven: Peeters, 1986.
23
A.L. Motzkin, Maimonides and Spinoza on Good and Evil, Da‘at 24, 1990, pp. v-xxiii.
S. Pines, Truth and Falsehood Versus Good and Evil: a Study in Jewish and General
24 Philosophy in Connection with the Guide of the Perplexed I,2, in Studies in
25 Maimonides, ed. I. Twersky, Cambridge, Massachusetts/London: Harvard Univer-
26 sity Press, 1990, pp. 95 – 157.
27 S. Rosenberg, Good and Evil in Jewish Thought (Hebr.), Tel Aviv: Ministry of Defence
28
Publishers, 1985.
C. Steel, Avicenna and Thomas Aquinas on Evil, in Avicenna and his Heritage, eds
29 J. Janssens and D. De Smet, Leuven: Leuven University Press, 2002, pp.171 – 96.
30 M. Zonta, Avicenna in Medieval Jewish Philosophy, in Avicenna and his Heritage, eds
31 J. Janssens and D. De Smet, Leuven: Leuven University Press, 2002, pp. 267 – 80.
32
33
34
35
36
37
38
39
40
41
42
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
1
2
3
4 Possible Hebrew Quotations
5 of the Metaphysical Section of Avicenna’s Oriental Philosophy
6
7 and Their Historical Meaning*
8
9
Mauro Zonta
10
11 Introduction
12
13 Among Avicenna’s philosophical works that have the general structure of an
14 encyclopedia, there is The Oriental Philosophy (al-Hikma al-mašriqiyya), whose
title has been usually interpreted in two different ˙ ways: either as ‘Eastern
15
16 Philosophy’, or as ‘Illuminative/Enlightening Philosophy’.1 Although the former
17 interpretation of the title has often been criticized, a number of considerations
18 militate in its favour. Some of these might become clear from the examination
19 of some Hebrew quotations of apparently lost passages of the work, neglected
20 until now. These quotations are found in some fourteenth-century Hebrew
21 philosophic-theological works, whose texts have been published only recently
22 (mostly in medieval Spanish translations), and whose contents and sources have
23 been examined in detail very recently by Ryan Szpiech in a very interesting
24 article.2 According to Szpiech, these quotations were incorrectly ascribed to
25 Avicenna, but were in reality taken from the Hebrew version of Ibn Tufayl’s
26 Living Son of the Watchful (Hayy ibn Yaqzān), made in Provence ˙at the
˙
beginning of the fourteenth century. However,˙ Szpiech does not point out that
27
28 the quotations even show some surprising similarities with corresponding ideas
29 of one of the many Indian philosophies and of Hinduism as well, as these latter
30 traditions might have been known in Avicenna’s time and milieu. I will re-
31 examine all these fragments, and will try to suggest some tentative hypotheses
32 about them, which differ in part from those suggested by Szpiech. The general
33
34
35
* Many thanks to Amos Bertolacci and Dag Nikolaus Hasse for their useful observations,
36 which have helped me in improving some points of the article, which is a partially
37 revised version of the original paper What Remains of the Metaphysical Section of
38 Avicenna’s al-Hikma al-mašriqiyya? read at the International Conference on ‘The Arabic,
Hebrew and Latin˙ Reception of Avicenna’s Metaphysics’ (Villa Vigoni, Loveno di Me-
39
naggio, 3 – 5 July 2008).
40
1 These two interpretations were first discussed in Nallino, Filosofia ‘orientale’ od
41 ‘illuminativa’.
42 2 See Szpiech, In Search of Ibn Sı̄nā’s Oriental Philosophy.
178 Mauro Zonta
1 structure and contents of The Oriental Philosophy, as well as what is probably the
2 correct meaning of its title, have been recently and excellently examined in
3 detail by Dimitri Gutas.3 I will suggest some possible improvements to a part of
4 Gutas’s study: the contents of the metaphysical section of The Oriental
5 Philosophy, which, as far as I know, have not yet been studied.
6 What are the contents of Avicenna’s Oriental Philosophy? It seems that in
7 this work Avicenna himself wanted not to give a summary of Aristotle’s
8 philosophy but to examine some of the most discussed points of it. Judging
9 from the already published parts of Avicenna’s work as well as from what
10 Avicenna himself states in the introduction to it,4 The Oriental Philosophy
11 should have included a number of sections, as follows:
12 1. A section on logic (mantiq), including the contents of Porphyry’s Eisagoge
13
˙
and of Aristotle’s De interpretatione, Prior Analytics and Posterior Analytics (the
14 last part being no longer extant). According to Gutas, also such parts of logic as
15 dialectics, sophistics, rhetoric and poetry should have been found in the
16 complete original version of this section;
17 2. A section on metaphysics, which according to Avicenna’s intention should
18 have included two parts: one on the so-called ‘universal science’ (‘ilm kullı̄) and
19 the other one on theology (‘ilm ’ilāhı̄). According to Amos Bertolacci, this
20 division seems to reflect the substantial originality of the metaphysics of The
21 Oriental Philosophy with respect to other metaphysical works by Avicenna,
22 which are more adherent to Aristotle’s views;5
23 3. A section on physics (‘ilm tabı̄‘ı̄), including the contents of natural
24
˙
sciences according to medieval Arab-Islamic philosophical tradition, i. e. physics
25 proper, the nature of heavens, generation and corruption of things, actions and
26 passions of natural things, meteorology, psychology, and probably zoology;
27 4. A section on ethics, called ‘practical science’ (‘ilm ‘amalı̄). According to
28 Gutas, this part of the work included neither ethics proper, nor economics, but
29 only what concerns ‘prophetic legislation’, i. e. man’s salvation in the afterlife.
30 The Oriental Philosophy, then, was divided into four sections. But Avicenna’s
31 work is apparently not preserved as a whole. How much of each of these
32 sections is definitely extant, and what has already been published?
33 As for the logical section, most of it has been preserved in some manuscripts
34 and has been published a number of times: the first edition of it, under the
35 fictitious title Mantiq al-mašriqiyyı̄n (The Logic of the Oriental People), appeared
36
˙
37 3 See Gutas, Avicenna’s Eastern (‘Oriental’) Philosophy. In this article, Gutas improves on
38 his treatment of the philosopher in his previous book: Gutas, Avicenna and the
39 Aristotelian Tradition, pp. 115 – 30 (see also pp. 43 – 9 for an English translation of the
introduction of The Oriental Philosophy).
40
4 Gutas, Avicenna’s Eastern (‘Oriental’) Philosophy, pp. 168 – 9.
41 5 See Bertolacci, The Reception of Aristotle’s ‘Metaphysics’ in Avicenna’s ‘Kitāb al-Šifā’’,
42 pp. 305 and 605.
Possible Hebrew Quotations 179
1 in Cairo in 1910. The contents of this section appear to be partially different

2 from those of other logical works by Avicenna.6
3 As for the physical section, it seems that the zoological part is lost; the other
4 parts are found in some Arabic manuscripts. These have been published in 1993
5 in a Ph.D. thesis,7 and a critical edition of them should appear soon.8 By
6 perusing these manuscripts, Gutas has concluded that, in contrast to the logical
7 section of The Oriental Philosophy, most of the physical section of it is ‘copied
8 verbatim from the Šifa’’. 9 This fact does not exclude a number of minor
9 alterations with respect to the contents of the corresponding section of
10 Avicenna’s Cure (al-Šifā’), some of which have been identified and studied by
11 Gutas himself.10
12 As for the ethical section, apparently nothing of it has been found until now,
13 although the possibility that some passages from it might be still extant cannot
14 be totally excluded.
15 As for the metaphysical section, as well, the question of the survival of at
16
least some fragments of it appears to be open, since some apparent quotations
17
are found in late medieval philosophic-theological works. The author of these
18
works is Avner of Burgos, a Spanish Jewish philosopher and theologian, living
19
in Castilia between 1270 and ca. 1345; he converted to Christianity (probably
20
about 1320), assumed the name of ‘Alfonso de Valladolid’, and disputed about
21
this fact with another Spanish Jew, Isaac Polgar.11 Avner wrote in Hebrew a
22
number of works on religious and philosophical questions; some of them have
23
been lost in their original text, but are still extant in medieval Castilian versions.
24
The existence of Avner’s quotations of Avicenna’s Oriental Philosophy was first
25
noted by Walter Mettmann, the critical editor of all the extant Castilian
26
translations of his works, who, in a marginal note, pointed out this fact.12
27
28
However, Mettman did not examine these quotations in detail, nor did he study
29
6 See Gutas, Avicenna’s Eastern (‘Oriental’) Philosophy, p. 177: ‘For the Logic part of al-
30
Hikma al-mašriqiyya, Avicenna apparently wrote the text anew. Although it is quite
31 ˙
parallel to his numerous other compositions on logic, there is no sustained literal
32 correspondence with – i. e., copying from – earlier works’.
33 7 zcan, İbn Sina’nin el-hikmetu’l-meşrikiyye. I owe a copy of this work to the kindness of
34 Jules Janssens.
8 Prof. Fehrullah Terkan (University of Ankara) is preparing a critical edition of the
35
physical section of Avicenna’s Oriental Philosophy, based upon the extant six manuscripts
36 of it (email of the author, July 8th, 2009). I am very grateful to Prof. Terkan for this kind
37 information. It should be pointed out that the almost entire part of this section about
38 optics has been translated into English in Hasse, Avicenna’s ‘De anima’, pp. 107 – 23.
39 9 Gutas, Avicenna’s Eastern (‘Oriental’) Philosophy, p. 177.
10 Gutas, Avicenna’s Eastern (‘Oriental’) Philosophy, pp. 174 – 7.
40
11 On Avner of Burgos, see the recent collection of bibliographical data found in Roth,
41 Dictionary of Iberian Jewish and Converso Authors, pp. 121 – 4.
42 12 See Alfonso de Valladolid, Mostrador de Justicia, vol. I, p. 160, n. 251.
180 Mauro Zonta
1 their close relationship, if any, to Avicenna’s philosophy and thought and to his
2 possible sources. Gutas too knew the existence of them, but affirmed that they
3 might not even be by Avicenna, since no trace of them was apparently found in
4 any of Avicenna’s other works; finally, he left open the question as to whether
5 The Oriental Philosophy was known in Andalusia, or in Maghreb at least.13
6 The four extant works by Avner of Burgos, in which Avicenna’s Oriental
7 Philosophy is more or less explicitly quoted, possibly from the metaphysical
8 section of it, are as follows:
9
- Replies to the Critic (whose Hebrew original title was Tešuvot la-Meharef );
10 ˙
- Offering of Zeal (Oferta de Zelos, in Hebrew Minhat Qena’ot);
11 ˙
- Book of the Law (Libro de la Ley, whose original Hebrew title might have been
12
Sefer ha-Torah or Sefer ha-Mišpat);
13 ˙
- Master of Justice (Monstrador de Justicia, in Hebrew Moreh Sedeq).
14 ˙
15 The first of these works is found both in its original Hebrew text, and in its
16 medieval Castilian translation, which might have even been made by the author
17 himself. The other three works are found in Castilian translations only. While
18 the Castilian versions of all four works were published by Mettmann in his
19 critical edition of them, the Hebrew text of the Replies to the Critic has been
20 edited in a still unpublished Ph.D. thesis only.14 I have consulted the text and
21 copied the relevant passages directly from the original unique manuscript of
22 Parma, Biblioteca Palatina, parmense 2440.15 I will examine the passages (twelve
23 in all) as found in their original Hebrew texts (if extant) or in their Castilian
24 versions, together with my English translation of them. I will briefly discuss the
25
contents of each passage, pointing out some possible similarities to correspond-
26
ing passages of Avicenna’s extant works, and to the section of The Cure devoted
27
to metaphysics in particular. Finally, I will point out some possible
28
correspondences with ideas and passages found in Indian philosophical and
29
religious texts, in particular the Veda, and with some elements of Indian
30
thought. A number of them are found in a famous Arabic source contemporary
31
to Avicenna: the Book of Enquiry about Indian Things (Kitāb tahqı̄q mā li-l-
32 ˙
33
34 13 See Gutas, Avicenna’s Eastern (‘Oriental’) Philosophy, pp. 171 – 2: ‘I have been informed
[by Charles Manekin (University of Maryland)] that the 13th-14th century Rabbi Abner
35
of Burgos, who converted to Christianity, mentions in some of his works that survive in
36 Spanish Avicenna’s al-Hikma al-mašriqiyya (filosofia oriental)’; however, according to
37 ˙
Gutas, ‘the passages mentioned by Rabbi Abner as coming from Avicenna’s Eastern
38 philosophy look suspect and cannot be readily identified in any of the extant portion of
39 the work’.
14 Hecht, The Polemical Exchange Between Isaac Pollegar and Abner of Burgos (non vidi).
40
15 Two reproductions of this manuscript are found in the Institute for Microfilmed Hebrew
41 Manuscripts of the National Library of Israel: the microfilm F 13444, and the CD Rom
42 149.
1 Hind) by the Persian astronomer Abū Rayhān al-Bı̄rūnı̄ (973 – 1048),16 with
2 whom Avicenna was in correspondence.17
3
4
5 Quotations
6
7 1. Alfonso de Valladolid (Abner of Burgos), Tešuvot la-Meharef. Spanische
˙
8 Fassung, ed. W. Mettmann, Opladen: Westdeutscher Verlag, 1998:
9
10 Passage 1.1., page 33, lines 25 – 7:
11 Et como el ssabio Avicena le [i.e. Dios] asemej al sol quando fablava en
12 aquellas luminarias divinales en la ‘Philosophia Oriental’. E l llmale sienpre
13 ‘vierbo de Dios’.
14 MS Parma, Biblioteca Palatina, parmense 2440, fol. 20v, lines 19 – 21:
15 U-kemo še-ha-hakam Ben Sina’ hemšilo ’el ha-šemeš, / ke-še-hayah
˙
16 medabber be-’otan ha-me’orot ha-’elohiyyot ba-Filosofia’ ha-Mizrahit.18 / We-
17 ˙
hu’ qore’ ’oto tamid ‘devar ’Elohim’.
18 English translation:
19
‘Just like the sage Avicenna compared him (i. e. God) to the sun, when he
20
speaks about those divine luminaries in The Oriental Philosophy; and he always
21
calls him “God’s word”’.
22
23
This passage is very similar to passage 3.2. A comparison not of God, but of the
24
separate active intellect (‘aql bi-l-fi‘l) to the sun is found in Avicenna’s Cure, On
25
the Soul, treatise 5, chapter 5, first lines, where Avicenna affirms that ‘its (i. e. the
26
separate active intellect) relation to our souls is like the sun’s relation to our
27
28
29 16 The below quotations are taken from al-Bı̄rūnı̄’s India; see also the English translation of
al-Bı̄rūnı̄’s work by Edward C. Sachau in Alberuni’s India.
30
17 The exchange of eighteen letters between Avicenna and al-Bı̄rūnı̄, apparently going back
31 to ca. 1000, has been published in Nasr and Muhaqqiq, Abū Rayhān Bı̄rūnı̄ wa-Ibn Sı̄nā;
32 ˙
see also their commented English translation in˙ Berjak and Iqbal,˙ Ibn Sina – al-Biruni
33 Correspondence. I owe a copy of the former to the kind courtesy of Amos Bertolacci.
34 The letters do not include any clear reference to the subjects treated here by Avicenna.
However, these subjects might have been discussed in a later correspondence, now
35
apparently lost, which should go back to the period when al-Bı̄rūnı̄ was preparing his
36 book on India (finished in 1030) and Avicenna was writing The Oriental Philosophy
37 (ca. 1027 – 9, according to Gutas, Avicenna and the Aristotelian Tradition, p. 128).
38 18 See the Hebrew version of Ibn Tufayl’s work, in MS Parma, Biblioteca Palatina,
39 parmense 2442, fol. 54v, lines 24 – 5,˙ where he ascribes to Avicenna the statement that:
mi-heqer ’or ha-’emet … ke-’ilu hayu beraqim, ‘from the examination of the light of the
40 ˙ (i. e. God) (there are things) … like lightnings’. See also the Arabic text in ’Abū
Truth
41 Bakr ibn Tufayl, Hayy ibn Yaqzān, p. 17, line 22: min ’ittilā‘i nūri al-Haqqi … ka-
42 ˙
’annahā burūqan. ˙ ˙ ˙˙ ˙
182 Mauro Zonta
1 visions’.19 The phrase ‘vierbo de Dios’, here ascribed to Avicenna himself,20 is a

2 translation of the Hebrew davar ’elohim, which in its turn might be either a
3 literal translation of the Arabic kalām ’Allāh, ‘God’s speech’,21 or even an
4 imperfect translation of the Arabic ’amr ’ilāhı̄, ‘divine thing’ (or ‘divine order’);
5 in the latter case, this translation might result from Avner’s tentative adaptation
6 of the concept to the Christian idea of Jesus as ‘God’s speech’ (John 1,1).
7 However, the latter concept is found, e. g., in a passage of The Cure, Metaphysics,
8 treatise 10, chapter 1 (about the typical Avicennian theme of al-mabda’ wa-l-
9 ma‘ād, ‘the beginning and the return’ of the human soul from and to God),
10 where a reference to the ‘first divine decree and command’ (al-’amr al-’ilāhı̄ al-
11 ’awwal) is found.22 This fact, together with the reference to the ‘divine
12 luminaries’ (which is not found in On the Soul, but seems to concern a
13 metaphysical question here), might suggest that the passage was taken from, or
14 at least inspired by, the metaphysics of The Oriental Philosophy.
15
17 Et escrivi el Avicena en la ‘Philosophia Oriental’ que el omne conplido ha
18 inteligenÅia separada; que si pudiesse ser que se partisse la inteligenÅia agente,
19 diriemos que esta es parte della. E sinon porque fue nueva despues de que non
20 era, diriemos que era aquella misma, e sinon porque se apropri en su cuerpo
21 quando sse ffizo de nuevo, diriemos que non se fizo de nuevo. Fasta aqui fue su
22 palabra.23 Et dio a entender en esto que la novedat que ha la inteligenÅia del
23 omne separada es apropriarsse la inteligenÅia agente en su cuerpo. E como lo
24 que dixo que la ssemejanÅa de la luz del sol proporÅionada al cuerpo que
25 alunbra, que si sse toliesse el cuerpo, tol[i]erse-ýa su luz, de ssin que se toliese la
26 luz.
28 U-katav Ben Sina’ ba-Filosofia’ / ha-Mizrahit še-ha-’adam ha-šalem yeš lo
29
˙
śekel nivdal, še-’illu hayah / ’efšar še-yithalleq ha-śekel ha-po‘el hayyinu ’omerim
30
˙
24
še-hu’ heleq / mimmennu; we-luley še -nithaddeš aharey še-lo’ hayah hayyinu
31
˙ ˙ ˙
32 19 See Avicenna, De Anima, p. 234, line 19: wa-nisbatuhū ’ilā nufūsinā nisbatu al-šamsi ’ilā
33 ’absātinā.
34 20 On˙ the contrary, on passage 3.2. the same phrase is explicitly ascribed not to Avicenna,
but to Plato. See below, p. 189.
35
21 See Avicenna, Metaphysics, p. 359, line 6 of the English translation = line 2 of the Arabic
36 text.
37 22 See Avicenna, Metaphysics, p. 363, lines 5 – 6 of the English translation = lines 3 of the
38 Arabic text.
39 23 The sentence ‘fasta aqui fue su palabra’, which seems to be a literal translation of the
typical Hebrew expression ‘ad ka’n devarayw, ‘up to this point, his words’ has no
40
correspondence in the Hebrew text, as it is found in the Parma manuscript. Very
41 probably, it was erroneously omitted by the copyist of the Hebrew manuscript.
42 24 In the manuscript: ke-še, ‘when’.
1 ’omerim še-hu’ / hu’ be-‘asmo, we-luley še-hityahed be-gufo ke-še-nithaddeš

˙ ˙ ˙
2 hayyinu ’omerim / še-lo’ nithaddeš.25 Ki hodia‘ ba-zeh še-ha-hidduš ’ašer la-śekel
˙ ˙
3 ha-nivdal / šel ha-’adam hu’ hityahed ha-śekel ha-po‘el be-gufo, u-kemo še-
˙
4 ’amar / ba-śekel ’or ha-šemeš ha-ne‘erak ’el ha-guf še-zoreah ‘alayw še-’illu
˙
5 hayah / sar ha-guf hayah sar ’oro mibli še-yasur ha-’or.
7 ‘Avicenna wrote in The Oriental Philosophy that the perfect man has a
8 separate intellect. If it might be possible that the active intellect be divided into
9 parts, we would say that it is a part of it; were it not that it is created ex novo, we
10 would say that it is identical per se; were it not that it has not become proper
11 (hityahed) to his body when it is created, we would say that it is not created.26
12 ˙
His words go up to this point. And he (i. e. Avicenna) means that the creation of
13 man’s separate intellect is the fact that the active intellect becomes proper
14
(hityahed) to his body,27 like he (i. e. Avicenna) says that the intellect is the light
15 ˙
of the sun arranged for the body on which it shines, and if the body goes away,
16
its light (too) goes away, but the light (in itself ) does not go away’.
17
18
Something partially similar to the doctrine found in the first sentence (‘the
19
perfect man has a separate intellect’) can be found e. g. in Avicenna’s Cure, On
20
the Soul, treatise 5, chapter 6, as follows: ‘But if one escapes from the body and
21
from the appendages of the body, then it is possible for him to come into perfect
22
contact with the active intellect’28 – although this passage does not refer to the
23
24
presence of the active intellect inside the human body. However, the general
25
connection between this concept and what is found in the following statements
26
(as well as in what appears to be Avner’s personal comment on them) is not
27
found in this passage, nor in any other passage of The Cure. One might suggest
28
29 25 As pointed out by Szpiech, a similar passage is found in the Hebrew version of Ibn
Tufayl’s work. See MS Parma, Biblioteca Palatina, parmense 2442, fol. 158v, lines 24 – 8:
30 ˙ ar kak ra’ah le-‘asmo ‘asmut nifredet, lo’ ‘avad še-yihyeh qesat ‘asmut ba‘al ha-šiv‘im ’elef
Ah
31 ˙
panim, ˙
’amarnu še-hayyinu ˙ qesatah. … We-luley še-zo’t ha-‘as˙ mut˙ hit haddešah ’ahar še-lo’
32 ˙
hayetah ’amarnu še-hi’ hi’, we-luley hityahedah be-gufo ’es˙el hiddušo˙ ’amarnu ˙še-hi’ lo’
tithaddeš. This version literally depends upon ˙ the Arabic ˙text˙in ’Abū Bakr ibn Tufayl,
33
˙ ˙
Hayy ibn Yaqzān, p. 85, lines 8 – 9, 12 – 13: tumma šāhada li-nafsihı̄ dātan mufāriqatan,
34 ˙ ğāza ’an ˙tataba‘‘ada dātu al-sab‘ı̄na alfi ¯wağhin, la-qulnā ’annahā¯ ba‘duhā. … Wa-
law
35
lawlā ’anna hādihi al-d˙ āta¯ hadatat ba‘da ’an lam takun, la-qulnā ’annahā hiya,
˙ wa-lawlā
36 ¯ ¯ ˙ ¯
’ihtisāsuhā bi-badanihı̄ ‘inda hudūtihı̄, la-qulnā ’annahā lam yahdut.
37 ˙ ˙ above
26 The ˙ three points seem˙to be,¯ in the author’s intentions, ˙three ¯ opinions which are
38 opposite to his own.
39 27 This sentence seems to be not part of the quotation, but Avner’s personal interpretation
of the above passages, ascribed to Avicenna.
40
28 See Avicenna, De anima, p. 248, lines 6 – 7: fa-’inna halasa ‘an al-badani wa-‘awāridi al-
41 ˘ ˙ tamāma al-ittisāli. I owe˙ this
badani fa-hı̄na’idin yağūzu ’an yattasila bi-l-‘aqli al-fa‘‘āli
42 ˙
information to Dag Nikolaus Hasse. ˙ ˙
184 Mauro Zonta
1 that here, in the first lines at least, Avner of Burgos is literally quoting an
2 original passage of the apparently lost metaphysical section of Avicenna’s
3 Oriental Philosophy. As a matter of fact, a very similar passage is found in
4 Avicenna’s Notes on Some Critical Points of Aristotle’s De anima (al-Ta‘liqāt ‘alā
5 hawāšı̄ kitāb al-nafs li-’Aristātālı̄s): ‘The complete perfection in knowledge is
˙ ˙˙
6 only via the actual conjunction with the active intellect’.29
7 A similar doctrine is found in the medieval Indian philosophical school of
8 Sāmkhya, which was well-known to al-Bı̄rūnı̄.30 Al-Bı̄rūnı̄ ascribed to this school
9 ˙
the doctrine of purusa, in origin ‘man’, but according to Sāmkhya ‘pure
˙ ˙
10 consciousness, soul’.31 In Sāmkhya, purusa is the first of a list of twenty-five
11 ˙ ˙
12
13
14 29 See Badawı̄, ’Arisţū ‘inda al-‘Arab, p. 95, lines 14 – 15: al-istikmālu al-tāmmu bi-l-‘ilmi
’innamā yakūnu bi-l-’ittisāli bi-l-fi‘li bi-l-‘aqli al-fa‘‘āli.
15
30 Al-Bı̄rūnı̄ translated an ˙Indian philosophical work from Sanskrit into Arabic: the 5th-
16 century PataÇjali’s Yogasūtra, whose Arabic version by al-Bı̄rūnı̄ has been edited in Ritter,
17 al-Bı̄rūnı̄’s bersetzung des Yoga-Sūtra, and translated into English and commented in
18 Pines and Gelblum, al-Bı̄rūnı̄’s Arabic Version of PataÇjali’s Yogasūtra. The Yogasūtra was
19 a product of the Yoga Indian philosophical school, although its contents seem to reveal
traces of the Sāmkhya school too (see Encyclopedia of Indian Philosophies. Volume IV,
20
pp. 165 – 6). Al-Bı̄rūnı̄ apparently knew other Indian philosophical works, among which
21 there might be the Sāmkhyakārikābhāsya by the Sāmkhya Indian philosopher Gaudapāda
22 (ca. 500 – 600). This fact˙ can be suggested˙ ˙
by a passage ˙ lines
of al-Bı̄rūnı̄’s India, p. 63,
23 16 – 17, where al-Bı̄rūnı̄ lists among the Indian books known to him kitābu ‘amilahū
24 Ġawra al-zāhidu wa-‘urifa bi-’ismihı̄ wa … Sānka ‘amilahū Kapila fı̄ al-’umūri al-
’ilāhiyyati, ‘the book which composed Gauda(pāda) the monk and which is known in its
25
name, and the Sāmkh(y)a which composed ˙ Kapila on divine subjects’ (see also the
26 English translation in Alberuni’s India, p. 132). Here, al-Bı̄rūnı̄ might have separated the
27 unique work by Gaudapāda into two, giving an incorrect name to what Gaudapāda
commented on: the Sām ˙ khyakārikā by the 4th-century Indian philosopher Īśvarakr ˙ sna,
28
29 which he might have identified as a lost work by one of the founders of the Sām˙khya
˙ ˙˙
school, Kapila. On the possible origin of this reference, see Alberuni’s India, p. 267; ˙ see
30
also Encyclopedia of Indian Philosophies. Volume IV, p. 210. A tentative list of al-Bı̄rūnı̄’s
31 Indian philosophical sources is also found and examined in Govind, al-Beruni’s
32 Observations on Indian Philosophical Concepts, pp. 15 – 17; see also Karmakar, Hindu
33 Philosophical Literature Known to al-Beruni, p. 246. According to Govind, al-Bı̄rūnı̄
34 knew the works of a number of Indian philosophical schools (not only the Sāmkhya, the
Yoga, and the Vedantic one, but also the Cārvāka and the Pūrva-Mı̄māmsā); but ˙ the text
35
mentioned by al-Bı̄rūnı̄ would be the Gaudapādakārikā by the Vedantic Indian ˙
36 philosopher Gaudapāda (ca. 600?). However, Govind ˙ too (pp. 17 – 18) thinks that al-
37 Bı̄rūnı̄ made some ˙ mistakes in his references to Indian philosophers and philosophies,
38 since the list of Indian philosophical works whose knowledge is usually ascribed to al-
39 Bı̄rūnı̄ appears to be not always clear to him.
31 See al-Bı̄rūnı̄’s India, p. 19, lines 19 – 20: fa-’innahum yusammūna al-nafsa Puruša, wa-
40
ma‘nāhū al-rağulu, ‘since they (i. e. the Hindus) call the soul puruša, and its meaning is
41 “man”’; see also Alberuni’s India, p. 40. About the various meanings of the term, see also
42 Encyclopedia of Indian Philosophies. Volume IV, pp. 74 – 5.
1 principles, and the third of them is buddhi, ‘intellect’.32 Of course, this does not
2 mean that the affirmation that ‘the perfect man has a separate intellect’ was
3 surely influenced by Indian philosophy; it means that a possible influence of
4 this philosophy, as found in other passages (see in particular 3.1.), cannot be
5 denied.
6
8 Et assi escrivi el Avicena en la ‘Philosophia Oriental’ que la verdaderia de la
9 substancia del omne conplido, aquella es la verdaderia de Dios, sin ningun
10 mudamiento nin contrariedat.
12 U-ken katav Ben Sina’ ba-Filosofia’ / ha-Mizrahit še-’amittat ‘esem ha-’adam
13
˙
ha-šalem hi’ ’amittat ‘esem / ha-’eloah beli śum šinnuy we-hilluf.˙33
14 English translation: ˙ ˙
15 ‘So Avicenna wrote in The Oriental Philosophy that the truth of the
16 substance of the perfect man is, without any alteration and substitution, the
17 truth of the substance of God’.
18
19 The contents of this passage are found in some other quotations of Avicenna’s
20 Oriental Philosophy in Avner’s works (see the passages 2.2., 3.4. and 3.6.). No
21 similar statement is found in The Cure; however, a possible connection between
22 the perfect man and God can be found in The Cure, Metaphysics, treatise 8,
23 chapter 6, where Avicenna states that ‘the necessary existent (wāğib al-wuğūd)
24 per se must be the furnisher of all existence, and every perfection of existence …
25 and all that is a necessary existent is true’.34 Although there is no explicit
26 identification between God and the perfect man here, the reference to ‘all
27 existent’ and to ‘every perfection of existence’ might apparently include a
28 reference to the existence of perfect man too. This might justify the hypothesis
29 that the passage quoted by Avner of Burgos was really taken from Avicenna’s
30 metaphysical section of The Oriental Philosophy.
31
32
33
34
35
32 See Encyclopedia of Indian Philosophies. Volume IV, p. 49.
36 33 A similar passage is found in the Hebrew version of Ibn Tufayl’s work. See MS Parma,
37 Biblioteca Palatina, parmense 2442, fol. 151v, lines 5 – 6:˙ ’amittat ‘asmuto hi’ ‘asmuto
(sic). See also the Arabic text in ’Abū Bakr ibn Tufayl, Hayy ibn Yaqzān, ˙ p. 81, line˙ 22:
38
haqı̄qatu dātihı̄ hiya dātu al-Haqqi. ˙ ˙ ˙
39
34 ˙Avicenna,¯al-Šifā’, al-’Ilāhiyyāt,
¯ ˙ p. 356, lines 7 – 9: wāğibu al-wuğūdi yağibu ’an yakūna li-
40
dātihi mufı̄dan li-kulli wuğūdin, wa-li-kulli kamāli wuğūdin … fa-lā ’ahaqqu ’idan min
41 wāğibi al-wuğūdi; the above English translation is found in Avicenna, ˙ Metaphysics,
42 p. 284, lines 15 – 19 = lines 9 – 11 of the Arabic text.
186 Mauro Zonta

2 Et como escrivi el Avicena otrossi que las obras que vienen de los que han
3 las formas non son suyas dellos segund verdat, mas son de obrador que obra con
4 ellos las obras que son anonbradas a ellos. Et como el dicho del que dizia: ‘Fuy
5 oyr al qui oy conmigo, e ver al qui vio conmigo’.
6 MS Parma, Biblioteca Palatina, parmense 2440, fol. 24r, lines 13 – 17:
7 U-kemo še-katav Ben Sina’: ‘Od / še-ha-pe‘ullot ha-be’ot min ba‘aley ha-
8 surot / ’eynan lahem ba-’emet, ’aval hen le-po‘el še-yif‘al bahem ha-pe‘ullot / ha-
9 ˙meyuhasot ’aleyhem, u-ke-ma’amar ha-’omer: “Hayyiti šema‘ le-’ašer / šema‘ bi,
10 ˙
u-re’ot le-’ašer ra’ah bi”’.
12 ‘As Avicenna wrote: “Since the possessors of the forms do not truly possess
13 the actions coming from them, but those actions are in possession of the actor
14 which performs through them the actions proper to them, according to what is
15 said: I was (the sense of ) hearing to Him who heard through me, and sight to
16 Him to saw through me”’.
17
18 This passage is found in a wider form in the passage 2.1. I have not been able to
19 find any other passage by or ascribed to Avicenna about this point.35 However, a
20 rather similar concept is found in the Sāmkhya philosophy, as it is reported by
21 ˙ Hindus) this is also the form of the
al-Bı̄rūnı̄: ‘And according to them (i. e. the
22 liberated man, who is not different from Him (i. e. God) in those (things), apart
23
from his beginning, since he does not exist in the previous eternity like (Him),
24
because of the fact that, before it, he is in the “place of confusion”, knowing the
25
object known, and his knowledge is like an imagination, acquired via an
26
endeavour’.36 One can also compare a passage of the Yogasūtra of PataÇjali,
27
which al-Bı̄rūnı̄ translated into Arabic: ‘Pure awareness is just seeing, itself;
28
although pure, it usually appears to operate through the perceiving mind’.37
29
30
35 The final quotation is taken from al-Buhārı̄’s collection of ’ahādı̄t : see ’Abū Bakr ibn
31 Tufayl, Hayy ibn Yaqzān, p. 56, n. 1. ˘ ˙ ¯
32 ˙ al-Bı̄rūnı̄’s
36 See ˙ India, ˙p. 39, lines 20 – 21: wa-hādā ’aydan sifatu al-mutahallasi ‘indahum,
fa-lā yanfasila ‘anhū fı̄hā ’illā bi-l-mabda’i fa-’innahū˙ ˙ lam yakun ˘ fı̄ ˙al-’azali al-
33
˙
mutaqaddimi ka-dālika, min ’ağli ’annahū kāna qablahū fı̄ mahalli al-irtibāki ‘aliman
34
bi-l-ma‘lūmi wa-‘ilmuhū ka-l-hayāli muktasabun bi-l-’iğtihādi;˙ see also the English
35
translation in Alberuni’s India, ˘p. 81.
36 37 This passage is found in Yogasūtra, 2, 20: drastā drśi-mātrah śuddho ’pi pratyayānupāsyah.
37 The text and the above English translation are ˙˙ taken
˙ from Hartranft, The Yoga-Sūtra of
38 PataÇjali. As for the partially different Arabic version by al-Bı̄rūnı̄, it is as follows: al-
39 ‘ālimu bi-ġayri ma‘lūmin yakūnu fı̄ dātihi ‘āliman bi-l-quwwati wa-lā yahruğu ’ilā al-fi‘li
˘ lines 10 – 11).
’illā bi-l-ma‘lūmi (Ritter, al-Bı̄rūnı̄’s bersetzung des Yoga-Sūtra, p. 181,
40
See the English translation in Pines and Gelblum, al-Bı̄rūnı̄’s Arabic Version of
41 PataÇjali’s Yogasūtra, p. 525 (and see also p. 541, nn. 97 – 8): ‘The knower, without the
42 known, is in his essence a knower in potentia: he is actualized only through the known’.
1 2. Alfonso de Valladolid (Abner of Burgos), Ofrenda de Zelos (Minhat Kena’ot)

2
˙
und Libro de la Ley, ed. W. Mettmann, Opladen: Westdeutscher Verlag, 1990:
3
4 In the Ofrenda de Zelos:
6 E esto es como lo que escrivi el Abicepna en la ‘Philosofia Oriental’, que las
7 obras que vienen de las cosas fformadas, quiere dezir que an fformas con que
8 obren, non son por ellas ssegund verdat, nin sson dellas, mas sson de obrador
9 que obra con ellas las obras que son anonbradas a ellas, e como el dezir del que
10 dixo: ‘Ffuy oyr del que oy comigo, e veer del que vio comigo’. E dixo: ‘Non
11 ech quando ech, mas Dios es el qui ech’.38
13 ‘And this is how Avicenna wrote in The Oriental Philosophy, that the actions
14 coming from formed things, i. e. those (things) having forms by which they act,
15 are not truly by them, nor they are possessed by them, but are in possession of
16 the actor which makes through them the actions proper to them, according to
17 what is said: “I have heard what He (i. e. God) hears together with me, and have
18 seen what He sees together with me”; and it (also) says: “I have not pushed
19 when I have pushed, but God is who has pushed”’.
20
21 As said before, this passage looks like a wider version of passage 1.4.
22
23 In the Libro de la Ley:
25 Y esto segund lo que escrivi el sabio AviÅienna en la ‘Filosofia Oriental’ que
26 la verdaderia del omne conplido, esso es la verdaderia de Dios sin ningun
27 departimiento.
29
30
31
32
33 38 See the Hebrew version of Ibn Tufayl’s work in MS Parma, Biblioteca Palatina, parmense
˙
2442, fol. 115v, lines 19 – 23: We-hitba’er lo še-ha-pe‘ullot ha-mesuddarot mehem ’eynam
34
lahem, we-’omnam hem le-po‘el yif‘al bam ha-pe‘ullot ha-meyuhasot ’aleyhem, we-zeh ha-
35
‘inyan ’ašer hušqaf lo hu’ ‘inyan ’omero: ‘Hayyiti šema‘ ’ašer šema‘˙ bo, u-re’ot ’ašer yir’eh bo’,
36 we-’amar be-’El hašak: ‘Lo’ hišlakti bi hišlakti, ’aval ha-’El hišlik’. See the Arabic text in
37 ’Abū Bakr ibn T˙ufayl, Hayy ibn Yaqzān, p. 55, line 24–p. 56, line 3: wa-tabayyana lahū
˙ ādira˙ ‘anhā laysat˙fı̄ al-haqı̄qati lahā, wa-’innamā hiya li-fā‘ilin yaf‘alu
’anna al-’af‘āla al-s
38
˙
bihā al-’af‘āla al-mansūbata ˙ ā al-ma‘nā ’alladı̄ lāha lahū, huwa qawlu
’ilayhā. Wa-hād
39
Rasūli ’Allāhi: ‘Kuntu sam‘ahū ’alladı̄ yasma‘u¯ bihı̄, wa-basarahū ’alladı̄
˙ yabsaru bihı̄’. Wa-
40 ¯ ˙
fı̄ muhkami al-tanzı̄li: ‘Fa-lam taqtulūhum, wa-lākinna ’Allāha qatalahum, ˙ wa-mā
41 ˙
ramayta ’id ramayta, wa-lākinna ’Allāha ramā’. The latter quotation is taken from al-
42 Qur’ān, 8, 17.
188 Mauro Zonta
1 ‘This is according to what the sage Avicenna wrote in The Oriental

2 Philosophy, that the truth of the perfect man is the truth of God, without any
3 difference’.
4
5 This passage seems to reproduce almost literally what is found in passage 1.3.
6
7 3. Alfonso de Valladolid (Abner of Burgos), Mostrador de Justicia, 2 vols., ed.
8 W. Mettmann, 1994 – 6:
9
10 Passage 3.1., in volume I, page 160, lines 2 – 6:
11 E esto mostr el AviÅena encubiertamientre en la ‘Filosofia Oriental’, en que
12 dixo que la sustanÅia del entendimiento obrador39 a ssetenta mill caras, e que en
13 cada cara a setenta mill bocas, e en cada boca setenta mill lenguas, con que
14 alaban ssienpre al Criador, e que aquella muchedunbre que paresÅe non es
15 muchedunbre segund verdat.40
17 ‘Avicenna explains this in a hidden way in The Oriental Philosophy, where he
18 affirms that the substance of the active intellect has seventy thousand faces, and
19 each face has seventy thousand mouths, and in each mouth there are seventy
20 thousand tongues, by which they always praise the Creator; and that abundance
21 which appears is not the(ir) real abundance’.
22
23 This passage appears to include a statement (the fact that here Avicenna speaks
24
‘in a hidden way’) and some images which might be interpreted as having an
25
almost mystical character, so being apparently at variance with Gutas’s
26
interpretation of Avicenna’s work.41 However, it might have been found either
27
in the metaphysical section or, even better, in the ethical section of The Oriental
28
Philosophy. In fact, one should remember that the main theme of the ethical
29
30
39 In this case, ‘entendimiento obrador’ appears to be a different Spanish version of the
31 Hebrew term śekel po‘el, ‘active intellect’; in other cases, the same term has been
32 differently rendered as inteligenÅia agente (see e. g. here above, passage 1.2.).
34 parmense 2442, fol. 158v, lines 17˙– 20: U-le-zo’t ha-‘asmut šiv‘im ’elef panim, be-kol
˙
panim šiv‘im ’elef peh, be-kol peh šiv‘im ’elef lešon, we-kol lešon mešabbehet ‘asmut ha-’emet
35
u-meqaddešah u-mehallelah, lo’ tanuah, we-ra’ah le-zo’t ha-‘asmut ’ašer yah ˙ šov ˙ha-ribbuy we-
36 ˙ ˙
’eyno ribbuy. See also the Arabic text in ’Abū Bakr ibn Tufayl, Hayy ibn ˙ Yaqzān, p. 85,
37 ˙
lines 2 – 4: Wa-li-hādihi al-dāti sab‘ūna ’alfi wağhin, fı̄ kulli ˙
wağhin sab‘ūna ˙’alfi famin,
¯
wa-fı̄ kulli famin sab‘ūna ¯ lisānin, tusabbihu bihā dāta al-Wāhidi al-Haqqi, wa-
’alfi
38
tuqaddisuhā wa-yumağğiduhā, lā tafturu. Wa-ra’yā ¯ al-dāti ’allatı̄
˙ li-hādihi ˙ tūhamu ˙ fı̄hā al-
39 ¯ ¯
katratu, wa-laysat katı̄ratan.
40
41 See¯ Gutas, Avicenna’s¯ Eastern (‘Oriental’) Philosophy, pp. 160 – 66. However, about the
41 possible influence of Indian philosophies on medieval Islamic mysticism see Massignon,
42 Essai sur les origines du lexique technique de la mystique musulmane, pp. 81 – 98.
1 section, according to Gutas, should have been a treatment of ‘the life to come’ –
2 a theme which might be found in this passage.
3 Here, an apparent trace of the influence of Indian philosophical and
4 theological tradition is found. In two fundamental Vedic texts, the Atharvaveda
5 and the Rgveda, there are passages that include a rather similar idea: ‘A thousand
6
˙ Purusha, a thousand eyes, a thousand feet …. This Purusha is all that
heads hath
7 yet hath been and all that is to be’.42 An even more similar passage can be read in
8 the Bhāgavata Purāna, a Hindu religious text probably written only one or two
9
˙
centuries before Avicenna: ‘The Lord [Mahā-Visnu], although lying in the
10 Causal Ocean, came out of it, and dividing ˙ ˙Himself as Hiranyagarbha
11 [Brahmā], He entered into each universe and assumed the virāt-rūpa, ˙ 43 with
12
44 ˙
thousand of legs, arms, mouths, heads, etc.’. Of course, the author of this
13 passage ascribed to Avicenna tried to adapt this concept to his religious
14 believing, by identifying the many things that the Indian gods are with the
15 many things that praise God.
16
17 Passage 3.2., in volume I, page 160, lines 28 – 9:
18 E assi Platon le llam ‘Verbo de Dios’, assi como el AviÅena le asemej al sol
19 en la ‘Filosofia oriental’.45
21 ‘So Plato called him (i. e. God) “God’s word”, just like Avicenna compared
22 him to the sun in The Oriental Philosophy’.
23
24 This passage is very similar to the passage 1.1., quoted above. It should be noted
25 that the sun was regarded to be one of the most important gods of Hinduism,
26 especially under the name of Āditya. The sun is regarded by the Hindus as the
27 main divine ‘light’ (jyotis), being son of Aditi, ‘infinity’, the father of the seven
28 principal deities of Vedic Hinduism. This information might have come to
29 Avicenna via al-Bı̄rūnı̄.46 Moreover, it should be pointed out that a similar idea
30
31 42 See Rgveda, X, 90, 1a and 2a (sahasraśı̄rsā purusah sahasrāksah sahasrapāt … Purusa
32 evedam˙ sarvam yad bhūtam yac ca bhavyam), ˙ as translated
˙ into˙ English in The Hymns ˙of
33 the Rgveda, vol. II, p. 517. See also the partially corresponding passage in Atharvaveda,
34 XIX,˙ 6, 1.
43 I.e. ‘excellent form’.
35
44 See Bhāgavata Purāna, II, 5, 35 (sa eva purusas tasmād andam nirbhidya nirgatah
36 ˙ ah sahasrānana-śı̄rsavān),
sahasrorv-ańghri-bāhv-aks ˙ as translated ˙ into English by Swami
37 ˙
Prabhupāda on www.srimadbhagavatam.org. ˙
39 parmense 2442, fol. 158v, lines 21 – ˙2: u-ke-’ilu zo’t ha-‘asmut surat ha-šemeš. See also the
Arabic text in ’Abū Bakr ibn Tufayl, Hayy ibn Yaqzān, p.˙ 85,˙line 5: wa-ka-anna hādihi
40
al-dāta sūrata al-šamsi. ˙ ˙ ˙ ¯
41 ˙
46 About al-Bı̄rūnı̄’s knowledge of this concept, see e. g. al-Bı̄rūnı̄’s India, p. 56, lines 1 – 2:
42 wa-min al-’asnāmi al-mašhūrati sanamu Mūltān bi-’ismi al-šamsi wa-li-dālika summiya
˙ ˙
190 Mauro Zonta
1 is found in the Vedantic Indian philosophy.47 Obviously, if the metaphysics of

2 The Oriental Philosophy, as it was read by Avner, really included such reference,
3 its author should have adapted the Indian polytheistic concept of the sun as a
4 god to the Islamic concept of the unique God.
5
6 Passage 3.3., in volume I, pages 234, last line-235, line 2:
7 E la rrazon daquellos espejos e de aquellas fazes que nonbraron los sabios del
8 Talmud, essa es misma la razon de los espejos e de las fazes que nonbr el
9 Avicena en la ‘Filosofia oriental’.48
11 ‘And the reason of those mirrors and those faces mentioned by the wise men
12 of the Talmud is the same reason of the mirrors and the faces mentioned by
13 Avicenna in The Oriental Philosophy’.
14
15 The contents of this passage appear to be rather similar to what can be found in
16 the passage 3.1. The reference to the Talmud by Avner might in reality hint to a
17 passage found in Genesis Rabbah, chapter 4, where the role played by the ‘big
18 and little mirrors’ (in Hebrew, mar’ot and bevu’ot gedolot and qetannot) in the
19 world is mentioned.49 As for the reference to Avicenna, one ˙ can compare
20 Avicenna’s idea of the two faces of the human soul, as well as his mention of the
21 thesis that the soul is like a mirror (although this thesis seems to be refuted by
22 Avicenna himself ).50 The reference to ‘mirrors’ is found also in the Indian
23 philosophy, as pointed out by al-Bı̄rūnı̄.51
24
’Āditya; see also the English translation by in Alberuni’s India, p. 116: ‘A famous idol of
25
their (i. e. of Hindus) was that of Multn, dedicated to the sun, and therefore called
26 ffditya’.
27 47 See Encyclopedia of Indian Philosophies, Volume III, p. 24, where it is said that, according
28 to the Vedantic philosophy, ‘the “way of the gods” (devayāna) … leads to the sun’.
29 48 See ’Abū Bakr ibn Tufayl, Hayy ibn Yaqzān, p. 86, line 20, where a reference to al-marāyā
˙
wa-l-suwar, ‘the mirrors ˙ the images’,
and ˙ is found.
30
49 ˙
See Jastrow, Dictionary of Talmud Babli, vol. II, p. 835a.
31 50 See Avicenna, De anima, p. 47, lines 17 – 18 (about the existence of two sides of human
32 soul, the inferior and the superior ones) and p. 245, line 16 (where the statement that
33 fatakūnu al-nafsu ka-mir’ātu, ‘the soul is like a mirror’, is found).
34 51 See al-Bı̄rūnı̄’s India, p. 22, lines 18 – 19 (’intibā‘u sūrati al-šamsi wa-hiya wāhidatun fı̄
‘iddati marāyā mansūbatin ‘alā muhādātihā ˙ ’aw˙ miyāhin masbūbatin fı̄˙ ’awānin
35
mawdū‘atin ‘alā muwāzātihā), as translated in Alberuni’s India, p.˙ 46: ‘The image of
˙
36 ˙ though he is only one, is represented in many mirrors which are placed opposite
the sun,
37 to him, as also in the water of vessels placed opposite’. This philosophical interpretation
38 of the mirrors might come from the Veda, where the term ādarśa, ‘mirror’, as ‘ideal
39 perfection’, is found. As a matter of fact, it should be noted that a very similar concept is
discussed in the Upadeśasāhasrı̄ by the well-known Vedantic Indian philosopher Śamkara
40
(ca. 750). See Encyclopedia of Indian Philosophies. Volume III, pp. 242 – 3: ‘The reflection˙
41 of a face is different from the face since it conforms to the mirror, whereas the face is
42 different from the reflection because it does not conform to the mirror …. The reflection
1 Passage 3.4., in volume II, page 16, lines 10 – 13:

2 E segund esta lengua misma dixo AviÅena, como prov por sus palabras, que
3 las intelligenÅias separadas no sson esso mismo que Dios, nin otre cosa; e otrossi
4 que la substancialidat del omne conplido non es essa misma que la intelligenÅia
5 separada, nin otra cosa.
7 ‘And in the same way Avicenna said, as it is proved by his words, that the
8 separated intellects are nothing but God, and nothing else; and also that the
9 substance of the perfect man is nothing but the separated intellect, and nothing
10 else’.
11
12 What can be read here is partially similar to what is found in passage 1.3., and
13 identical to what is found at the beginning of passage 1.2.
14
16 E sta ffue la entenÅion de Aben Rrost ssin dubda en aquella palabra en que
17 alabava aquel ayuntamiento divinal, como lo ovo Moysen e los otros omnes que
18 llaman Åoffim, e como lo alabava el AviÅena en el ‘Libro de la Filosofia oriental’,
19 donde aprendi Aben Rrost aquella unidad de Dios, que nonbr, ssegund
20 pareÅe, en la manera de su fablar.
22 ‘This was surely the intention of Averroes when he praised that divine
23 conjunction, like that Moses and the other men called sofim had, and like that
24 Avicenna praised in the Book of the Oriental Philosophy,˙where Averroes learned
25 that unity of God which he called, as it appears, according to his own words’.
26
27 Unfortunately, this passage seems to be too vague to be useful for our purpose
28 here, and it is probably impossible to identify a similar passage in another work
29 by Avicenna. However, it should be pointed out that an explicit reference to a
30 very similar concept (that the Necessary Existence, i. e. God, is not separated
31 from the celestial bodies), ascribed to Avicenna’s The Oriental Philosophy, is
32 found in Averroes’s The Incoherence of the Incoherence (Tahāfut al-Tahāfut). 52 Has
33 Avner taken this reference from Averroes, or directly from the work ascribed to
34 Avicenna?
35
36
37 of the face in the mirror is a property of neither the face nor the mirror …’
38 (Upadeśasāhasrı̄, II, 18, 32 and 37). See also a passage of the Hastāmalakaślokāh by
39 Hastāmala (ca. 800?), as translated into English by E. Harzer: yathā darpanāthāva
āthāmahānau mukham vidyate kalpanāhı̄namekam, ‘when the mirror vanished then ˙ also
40
the reflection perishes; only the face remains, free from confusion, alone’ (Encyclopedia of
41 Indian Philosophies. Volume III, p. 602).
42 52 See Averroes, Tahafot at-Tahafot, p. 421, lines 4 – 9.
192 Mauro Zonta

2 La verdaderia de la sustancia del omne conplido es la verdaderia de Dios,
3 ssin ningun demudamiento, como es provado en la ‘Filosofia oriental’.
5 ‘The truth of the substance of the perfect man is the truth of God, without
6 any difference, as it is proved in The Oriental Philosophy’.
7
8 This passage is identical to passage 1.3.
9
10
11 Some Tentative Hypotheses
12
13 In the footnotes to this article, I have quoted all the passages in Ibn Tufayl’s
14 ˙
Living Son of the Watchful which are more or less similar to the passages
15 explicitly ascribed by Avner to Avicenna’s Oriental Philosophy, basing myself on
16 Szpiech’s article and a direct examination of the Ibn Tufayl’s work, both in the
17 ˙
Arabic original and in the anonymous Medieval Hebrew version. As can be seen
18 from a close linguistic comparison between the Arabic and the Hebrew, the
19
latter is a very literal translation of the former, so that the evident analogies
20
between those passages and the quotations ascribed to Avicenna by Avner
21
cannot, in my view, rule out the possibility that both of them were directly
22
based upon an Arabic text. Now, the question is: was this common Arabic
23
source Ibn Tufayl’s work? Or was it the Arabic text of Avicenna’s Oriental
24 ˙
Philosophy, which, if we assume that Ibn Tufayl’s and Avner’s statements are
25 53 ˙
true, would have been in its turn a common source of both of them? The
26
second hypothesis might be confirmed by the fact that, after having quoted one
27
28
of the above passages ascribed to Avicenna (n. 3.5.), Avner himself points out
29
that a similar passage is found in Ibn Tufayl.54 In doing so, he apparently reveals
˙
30
that he had indeed read something of The Oriental Philosophy as ascribed to
31
Avicenna and compared it to what he found in Ibn Tufayl. Such comparison, if
˙
32
33 53 According to Gutas, Ibn Tufayl did not really know the contents of The Oriental
34 Philosophy (see Gutas, Ibn˙ Tufayl on Ibn Sı̄nā’s Eastern Philosophy), although he
˙
explicitly mentions it in two passages of his own work. However, it should be pointed
35
out that part of Ibn Tufayl’s work was influenced by a work usually ascribed to Avicenna,
36 which had the same ˙title (Living Son of the Watchful) and was well known in Medieval
37 Jewish Spain, so that it was even rendered into Hebrew in a paraphrastic version by
38 Abraham Ibn Ezra (see Abraham Ibn Ezra, Igeret Hay ben Mekitz, ed. I. Levin); can we
39 really exclude that the same thing happened in the case of The Oriental Philosophy?
54 This reference is found in Alfonso de Valladolid, Mostrador de Justicia, vol. II, p. 54, line
40
30; see Szpiech, In Search of Ibn Sı̄nā’s Oriental Philosophy, pp. 204 – 5. According to
41 Szpiech, this reference would prove that the passage 3.5. was evidently taken from Ibn
42 Tufayl’s introduction to his own work.
˙
1 proved true, might even explain the evident parallelisms between Ibn Tufayl’s
2 and Avner’s passages. Therefore, in my opinion, the question of the source ˙ of
3 Avner’s quotations remains partially open, as it is possible that the text of The
4 Oriental Philosophy, including at least the section on physics (as suggested by the
5 existence of a Judeo-Arabic manuscript of it)55 and that on metaphysics (as
6 suggested by Averroes’s reference mentioned here above, n. 3.5.), might have
7 circulated among some Jewish scholars in Spain or in Maghreb during the
8 Middle Ages.
9 Some of the above quotations allegedly taken from Avicenna’s Oriental
10 Philosophy found in Avner of Burgos’s works are about a similar subject (see e. g.
11 nn. 1.2., 3.4. and 3.6.; 3.1. and 3.3.); some other are almost identical (see e. g.
12 1.1. and 3.2.; 1.3., 2.2. and 3.6.; 1.4. and 2.1.). Some of them are vague (see
13 e. g. nn. 3.3. and 3.5.), while some others appear to be precise: Avner affirms
14 that one of them at least (n. 1.2) is very explicit (as shown by the closing
15 formula found in the Spanish version), so that it seems to be a literal translation
16 of a passage of an original Arabic text ascribed to Avicenna.
17 Although none of these quotations literally corresponds to any extant
18 passage by Avicenna, there are similarities between some of them and some
19 passages or terms of The Cure, in particular to statements found in On the Soul,
20 treatise 5, and in the Metaphysics. This fact might confirm the possible
21 Avicennian origin of these passages. As a matter of fact, although it is certain
22 that the physical section of The Oriental Philosophy is almost identical to the
23 corresponding section of The Cure, this does not mean that the same is true for
24 the metaphysical section. Of course, the fact that Avner was writing works with
25 theological and apologetical purposes might justify his employment of the
26 metaphysical section of The Oriental Philosophy, where the theme of God’s
27 nature was surely discussed, as it results from what Avicenna affirms in the
28 introduction to his work.56
29 Finally, it should be pointed out that there are some apparent, although not
30 literal, correspondences between the passages ascribed to Avicenna by Avner and
31
some concepts of Hinduism and classical Indian philosophies, as found in the
32
Veda and in the works by the Vedantic, Sāmkhya and Yoga school. The general
33 ˙
contents of these texts were known by al-Bı̄rūnı̄, as it results from his major
34
book on India. Al-Bı̄rūnı̄, who exchanged a number of letters with Avicenna,
35
might have transmitted to the latter some information about contemporary
36
Indian thought, so being a possible cultural mediator between Indian
37
philosophy and Avicennian one. These references to Indian philosophies, if
38
they were confirmed, might even explain what could have been the original
39
40
55 This is the MS Oxford, Bodleian Library, Poc. 181 (Neubauer 1334), folios 61v–152v ;
41 see Neubauer, Catalogue of the Hebrew Manuscripts in the Bodleian Library, c. 475.
42 56 See Gutas, Avicenna’s Eastern (‘Oriental’) Philosophy, pp. 167 – 8.
194 Mauro Zonta
1 meaning of the title of this work, ascribed to Avicenna: The Oriental Philosophy,
2 i. e. a philosophy more or less inspired by some ideas which came from an
3 eastern place with respect to the Arabic world.
4
5
6 Bibliography
7
8 al-Bı̄rūnı̄, Alberuni’s India. An account on the religion, philosophy, literature, geography,
9 chronology, astronomy, customs, laws and astrology of India about AD 1030, 2 vols,
transl. E.C. Sachau, London: Trubner and Co., 1888.
10
–– , al-Bı̄rūnı̄’s India. An account on the religion, philosophy, literature, geography,
11 chronology, astronomy, customs, laws and astrology of India about A. D. 1030, ed. E.
12 C. Sachau, London: Trubner and Co., 1887 (repr. Leipzig: Harrassowitz, 1925).
13 Alfonso de Valladolid (Abner of Burgos), Mostrador de Justicia, 2 vols, ed. W.
14 Mettmann, Opladen: Westdeutscher Verlag, 1994 – 6.
–– , Ofrenda de Zelos (Minhat Kena’ot) und Libro de la Ley, ed. W. Mettmann, Opladen:
15
Westdeutscher Verlag, ˙1990.
16 –– , Tešuvot la-Meharef. Spanische Fassung, ed. W. Mettmann, Opladen: Westdeutscher
17 Verlag, 1998. ˙
18 Averroes, Tahafot at-Tahafot (L’incohrence de l’incohrence), ed. M. Bouyges, 3rd edn,
19 Beirut: Dar el-Machreq, 1992.
Avicenna, al-Šifā’, al-Ilāhiyyāt, ed. I. Madkūr, Cairo: al-Matba‘a al-’Amı̄riyya, 1380/
20
1960. ˙
23 –– , De anima (Arabic Text), ed. F. Rahman, London/New York/Toronto: Oxford
‘A. Badawı̄, ’Aristū ‘inda al-‘Arab, al-Kuwayt: Wakālāt al-matbu‘āt, 1978.
25
R. Berjak and M.˙ Iqbal, Ibn Sina–al-Biruni Correspondence,˙Islam and Science, 1, 2003,
26 pp. 91 – 8, 253 – 60; 2, 2004, pp. 57 – 62, 181 – 8; 3, 2005, pp. 57 – 62, 167 – 70;
27 4, 2006, pp. 197 – 212; 5, 2007, pp. 53 – 60.
28 A. Bertolacci, The Reception of Aristotle’s ‘Metaphysics’ in Avicenna’s ‘Kitāb al-Šifā’’. A
29 Milestone of Western Metaphysical Thought, Leiden/Boston: Brill, 2006.
V. Govind, al-Beruni’s Observations on Indian Philosophical Concepts with References
30
to Their Christian, Greek and Islamic Parallels, Bharata Manisha, 4/3 – 4, 1978 – 9,
31 pp. 13 – 25.
32 R.T.H. Griffith, transl. and ed., The Hymns of the Rgveda, 2 vols, Varanasi:
Chaukhamba Amarabharati Prakashan, 1971. ˙
33
34 D. Gutas, Avicenna and the Aristotelian Tradition, Leiden: Brill, 1988.
–– , Avicenna’s Eastern (‘Oriental’) Philosophy. Nature, Contents, Transmission, Arabic
35
Sciences and Philosophy, 10, 2000, pp. 159 – 80.
36 –– , Ibn Tufayl on Ibn Sı̄nā’s Eastern Philosophy, Oriens, 34, 1994, pp. 222 – 41.
37 C. Hartranft,˙ The Yoga-Sūtra of PataÇjali. Sanskrit-English Translation and Glossary, on
38 hrih.net/patanjali/archive/ysp-skrit-eng-chip-hartranft.pdf.
39 D.N. Hasse, Avicenna’s ‘De anima’ in the Latin West. The Formation of a Peripatetic
40
Aragno Editore, 2000.
41
42
1 J.L. Hecht, The Polemical Exchange Between Isaac Pollegar and Abner of Burgos/Alfonso de
2 Valladolid according to Parma MS 2440 ‘Iggeret Teshuvat Apikoros’ and ‘Teshuvot la-
3
Meharef ’, Ph.D. Thesis discussed at the New York University, New York, 1993.
Ibn Ezra, Abraham, Igeret Hay ben Mekitz. A Critical Edition Supplemented with a
4 Hebrew Translation of the Arabic Original Hay Ibn Yaqizian by Abu Abi Alhusain Ibn
5 Abdalla Ibn Sina, ed. I. Levin, Tel Aviv: Katz Institute for the Study of Hebrew
6 Literature, Tel Aviv University, 1983.
7 Ibn Tufayl, ’Abū Bakr, Hayy Ibn Yaqzān, ed. A. Nader, Bayrūt: Dār al-mašriq, 1993.
˙
M. Jastrow, Dictionary of˙ Talmud Babli,
˙ Yerushalmi, Midrashic Literature and Targumim,
8
2 vols, New York: Pardes Publishing House, 1950.
9 R.G. Karmakar, Hindu Philosophical Literature Known to al-Beruni, Annals of the
10 Bhandarkan Oriental Institute (Poona), 38, 1957, pp. 245 – 8.
11 G.J. Larson and R. S. Bhattacharya, eds, Encyclopedia of Indian Philosophies. Volume IV.
12 Sāmkhya. A Dualist Tradition in Indian Philosophy, Delhi: Motilal Banarsidass,
13
1987.
L. Massignon, Essai sur les origines du lexique technique de la mystique musulmane, Paris:
14 Vrin, 1954.
15 S.H. Nasr and M. Muhaqqiq, Abū Rayhān al-Bı̄rūnı̄ wa-Ibn Sı̄nā, al-’as’ala wa-l-’ağwiba,
16
˙ 1352/1973.
Tehran, ˙
17 C.A. Nallino, Filosofia ‘orientale’ od ‘illuminativa’ d’Avicenna?, Rivista degli studi
18
orientali, 10, 1923 – 5, pp. 433 – 67.
A. Neubauer, Catalogue of the Hebrew Manuscripts in the Bodleian Library, Oxford:
19 Clarendon Press, 1886.
20 A. zcan, İbn Sina’nin el-hikmetu’l-meşrikiyye adli eseri ve tabiat felsefesi, Ph.D. Thesis
21 discussed at the Marmara University, Istanbul, 1993.
22 S.L.A. Pandey, Comparative Study of Indian and Persian Philosophy, Islamic Culture,
23
33, 1959, pp. 81 – 7.
S. Pines and T. Gelblum, al-Bı̄rūnı̄’s Arabic Version of PataÇjali’s Yogasūtra, Bulletin of
24 the School of Oriental and African Studies, 29, 1966, pp. 302 – 25; 40, 1977,
25 pp. 522 – 49; 46, 1983, pp. 258 – 304; 52, 1989, pp. 265 – 305.
26 K.H. Potter, ed., Encyclopedia of Indian Philosophies. Volume III. Advaita Vedānta Up to
27 Śamkara and His Pupils, Delhi: Motilal Banarsidass, 1981.
28
H. Ritter, al-Bı̄rūnı̄’s bersetzung des Yoga-Sūtra des PataÇjali, Oriens, 9, 1956,
pp. 165 – 200.
29 N. Roth, Dictionary of Iberian Jewish and Converso Authors, Madrid/Salamanca: Aben
30 Ezra Ediciones, 2007.
31 R. Szpiech, In Search of Ibn Sı̄nā’s Oriental Philosophy in Medieval Castile, Arabic
32 Sciences and Philosophy, 20, 2010, pp. 185 – 206.
33
34
35
36
37
38
39
40
41
42
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
1
2
3
4 On the Latin Reception of Avicenna’s Metaphysics before
5 Albertus Magnus: An Attempt at Periodization
6
7 Amos Bertolacci
8
9
Introduction
10
11
The Latin Middle Ages are a relatively well-known area of the reception of
12
Avicenna’s philosophy.1 For at least a hundred years, the precise mode of this
13
reception has attracted scholarly attention and raised a lively debate in which
14
different labels involving the name of Avicenna have been proposed to
15
characterize philosophical authors and currents variously indebted to Avicenna’s
16
thought. Thus, expressions such as ‘Avicennizing Augustinism’, ‘Latin Avicenn-
17
ism’, ‘Avicennizing Aristotelianism’, etc., are quite common.2 This proliferation
18
of labels – in some cases very different from one another – can be taken as a
19
symptom of a still immature stage of research; more positively, however, it also
20
shows the multiplicity of modes and the different areas of the transmission of
21
Avicenna in Latin. Although Avicenna’s philosophical writings did not enter the
22
official curricula of medieval universities, and were therefore less frequently
23
copied than Aristotle’s works, and never commented upon as such (with the
24
exception of some parts of the section of the Šifā on meteorology),3 they were
˘
25
extensively used by philosophers and theologians from the late twelfth century
26
onward. Thus, the temporal scope of their influence surpassed the limits of the
27
28
29
30
31
32
33 1 I wish to thank warmly Dag Nikolaus Hasse for his insightful remarks on a first draft of
34 the present article.
2 These formulae were coined, respectively, by Gilson, Les sources greco-arabes, De Vaux,
35
Notes et textes, and Van Steenberghen, La philosophie au XIIIe sicle, pp. 451 – 8. The
36 expression ‘Avicennizing Boethianism’ is used to designate Gundissalinus’ epistemology
37 in Fidora, Die Wissenschaftstheorie des Dominicus Gundissalinus, pp. 89 – 95.
38 3 The Latin translation of three excerpts of the fifth section on natural philosophy of the
Šifā (taken from chapters I, 1 and I, 5), under the cumulative title of De mineralibus, was
˘
39
appended to the Latin translation of Aristotle’s Meteorologica. For this reason, this was by
40
far the most often copied philosophical text by Avicenna in Latin translation (Kishlat,
41 Studien, p. 53, counts 134 mss.; Schmitt, Knox, Pseudo-Aristoteles Latinus, p. 44,
42 mention 148 codices).
198 Amos Bertolacci
1 Middle Ages, and reached modern authors such as Descartes, Spinoza and
2 Leibniz.4
3 Thus far, studies have focused mainly on the Latin reception of Avicenna’s
psychology in the Kitāb al-nafs of the Šifā , whose translation into Latin (De
˘
4
5 anima) has been critically edited as first in the series Avicenna Latinus.5 The
reception of some other parts of the Šifā available to Latin medieval readers is
˘
6
7 comparable, in terms of diffusion and impact, to that of the De anima,6 but an
8 overall study of their influence is still a desideratum. Avicenna’s metaphysics, as
expressed in the Ilāhiyyāt of the Šifā , is a case in point: a comprehensive history
˘
9
10 of the influence of its Latin translation (Philosophia prima) in the Middle Ages
11 has yet to be written.7 Previous scholarship on the Latin reception of the
12 Philosophia prima has provided insightful accounts of the influence of this work
13 on single authors of the second half of the thirteenth and of the fourteenth
14 century, such as Albertus Magnus (d. 1280), Thomas Aquinas (d. 1274), Henry
15 of Ghent (d. 1293), John Duns Scotus (d. 1308), and others.8 The picture that
16 emerges from these studies, however, is incomplete, if compared with the
17 diffusion of Avicenna’s metaphysics both before and afterwards. The present
18 contribution tries to fill the lacuna a parte ante by providing a tripartite
19 periodization of the circulation of the Philosophia prima in Latin philosophy
20 before the middle of the thirteenth century (§ 1), a detailed analysis of the first
21 of these three periods (§ 2), and an account of the evidence attesting the first
22 diffusion of Avicenna’s metaphysics in the University of Paris, shortly before its
23 employment by William of Auvergne (§ 3).
24
25
26
27
28 4 On the reception of Avicenna’s philosophy after the Middle Ages, see, among other
29 studies, Davidson, Proofs for Eternity, pp. 388 – 405 (‘Proofs of the existence of God as a
necessarily existent being in modern European philosophy’); Gaskill, Was Leibniz an
30
Avicennian?; Jolivet, L’pistmologie de Descartes; Hasnawi, La conscience de soi;
31 Rashed, Thodice et approximation; Hasse, Arabic Philosophy and Averroism; Yaldir,
32 Ibn Sı̄nā (Avicenna) and Ren Descartes (further bibliographical information on
33 Avicenna and Descartes in Hasse, Avicenna’s ‘De anima’, p. 80, n. 5).
34 5 The use of Avicenna’s De anima by Latin thinkers has been thoroughly investigated by
Hasse, Avicenna’s ‘De anima’.
35
6 Whereas the De anima is preserved in 50 known manuscripts, the De animalibus is
36 attested by 33 codices, the Philosophia prima by 25, the Liber primus naturalium
37 (chapters I–III, 1) by 22, the Logica by 13, and the De diluviis by 11 (see d’Alverny,
38 Notes; Bertolacci, A Community of Translators, and the bibliography quoted therein).
39 On the manuscript dissemination of the De mineralibus, see above, n. 3.
7 The overviews of the Latin impact of the Ilāhiyyāt in Anawati, La Mtaphysique
40
d’Avicenne, and Verbeke, Avicenna’s Metaphysics, are selective and cursory.
41 8 See in this volume the contributions of Galluzzo, Hasse, Pickav, Pini and Richardson,
42 and the further bibliography indicated therein.
On the Latin Reception of Avicenna’s Metaphysics before Albertus Magnus 199
1 § 1 The Influence on Latin Authors until Albertus Magnus:

2 Preliminary Remarks
3
4 § 1.1 The Issue of Continuity
5
6 Mainstream scholarship on the reception of the Philosophia prima has been
7 governed by two main assumptions. The first is that the earliest significant
8 recipient of this work, around the third or fourth decade of the thirteenth
9 century, is William of Auvergne (d. 1249), who is regarded as a forerunner of
10 the authors of the second half of the thirteenth century who fully display the
11 influence of this work; on this view, Avicenna’s De anima – whose translation is
12 coeval with that of the Philosophia prima – would have influenced Latin culture
13 long before the Philosophia prima. This assumption, which can be traced back to
14 1926, the year of Roland-Gosselin’s edition of Thomas Aquinas’ De ente et
15 essentia and Gilson’s first fundamental study on the ‘Avicennizing Augustinism’,9
16 posits a decided discontinuity between the translation of the Philosophia prima
17 around the middle of the twelfth century and its full reception in the second
18 half of the thirteenth. This hiatus has been variously explained. According to
19 some, it was due to the initial resistance of traditional Latin metaphysics to the
20 new Avicennian metaphysics.10 According to others, it was a consequence of the
21 inner logic of Avicenna’s philosophy, in which the theory of knowledge naturally
22 precedes metaphysics.11 According to a further explanation, it reflected a general
23 shift in the cultural climate at the middle of the thirteenth century from more
24 concrete, physiological issues, to more abstract, metaphysical concerns.12 The
25
26 9 In the doctrinal-historical studies on the principle of individuation and the distinction of
27 essence and existence that complement his edition of the De ente et essentia of Thomas
28 Aquinas (Introduction, Notes et tudes historiques), Roland-Gosselin takes William of
29 Auvergne as the starting point of the Latin reception of Avicenna’s metaphysics on these
two issues. In Pourquoi saint Thomas, conversely, Gilson regards psychology as the
30
doctrinal core of the Latin reception of Avicenna’s philosophy from its very beginning.
31 See the frequent references to Roland-Gosselin’s aforementioned study in Goichon, La
32 philosophie d’Avicenne.
33 10 Jolivet, The Arabic Inheritance, pp. 130 – 31, remarks that Avicenna’s central distinction
34 of essence and existence entered the philosophy and theology of Latin thinkers only in
the second half of the 13th century, because until then its adoption was prevented by the
35
key-notion of traditional Latin metaphysics (ultimately deriving from Boethius and
36 instantiated by Gilbert of Poitiers), namely the idea of the complementarity of quo est
37 (subsistence) and quod est (that which subsists).
38 11 Goichon, La philosophie d’Avicenne, p. 103: ‘Chronologiquement la thorie de la
39 connaissance [d’Avicenne] fut la premire influencer la pense occidentale’; ibid.,
p. 117: ‘Dans l’ensemble, l’tude d’Avicenne philosophe de l’Þtre est postrieure l’tude
40
d’Avicenne thoricien de la connaissance’.
41 12 Hasse, Avicenna’s ‘De anima’, contends that, contrary to Avicenna’s De anima, the
42 Philosophia prima had greater impact in the second half of the 13th century than in the
200 Amos Bertolacci
1 second assumption of previous scholarship on the reception of the Philosophia

2 prima is that the impact of Avicenna’s philosophy in general, and of his
3 metaphysics in particular, began to decrease when the Latin translations of
4 Averroes’ commentaries on Aristotle started to gain success.13
5 These two assumptions look incompatible, since the first posits the starting-
6 point of the Latin reception of Avicenna’s metaphysics more or less when the
7 second places the beginning of its decline. In fact, both are in need of some
8 revision. As to the first, whatever the reasons adduced in its support, the thesis
9 of a temporal gap between the translation of the Philosophia prima, on the one
10 hand, and its reception, on the other, needs to be complemented by a more
11 documented investigation: the available data show that the translation of this
12 work and its subsequent employment by philosophers and theologians are two
13 steps of a continuous process.14 The circulation of the manuscripts of the
14
translation confirms this impression.15 The second assumption, likewise, does
15
not match the available evidence: the diffusion of Averroes’ Aristotelian
16
commentaries appears to have prompted not a progressive eclipse of Avicenna’s
17
thought in Latin philosophy, but a better grasp of Avicenna’s philosophy, an
18
outspoken acknowledgment of its value, and a strenuous defense of Avicenna’s
19
positions against Averroes’ frequent and harsh criticisms. This attitude emerged
20
21 first (p. 225), because at that time the interest of scholars shifted from physiology to
22 more theoretical problems (p. 77).
23 13 Goichon, La philosophie d’Avicenne, p. 92: ‘[Le courant avicennien] s’affaiblit lorsque
24 grandit l’influence d’Averros’; De Libera, D’Avicenne Averros, p. 179: ‘pass l’ge
d’or de la fine du XIIe sicle et de l’‘augustinisme avicennisant’ de la premire moiti du
25
XIIIe, Ibn Sı̄nā s’efface devant Averros’.
26 14 The thesis of an earlier diffusion of Avicenna’s De anima with respect to the Philosophia
27 prima seems to project on history the precedence of the study of the Latin impact of
28 Avicenna’s psychology over that of his metaphysics in modern scholarship. The first
29 fundamental study on the Latin reception of Avicenna’s De anima is Gilson, Pourquoi
saint Thomas; the diffusion of the Philosophia prima has started to be studied later. De
30
Libera, Penser au Moyen ffge, p. 112, rightly remarks that ‘Avicenne a t lu et exploit ds
31 la fin du XIIe [sicle]’.
32 15 The most ancient extant manuscript of the translation dates from the first half of the 13th
33 century (ca. 1240; see Van Riet, Traduction latine et principes d’dition, pp. 125*–6*);
34 not being an archetype of the tradition, this codex attests the existence of previous
manuscript circulation. The translation of a chapter of the work (III, 5) has circulated
35
earlier, attached to the Latin translation of al-Ġazālı̄’s Summa, in a manuscript of the
36 beginning of the 13th century copied in Spain, as well as in four other codices. D’Alverny,
37 Les traductions d’Avicenne, pp. 154 – 5 (followed by Van Riet, Traduction latine et
38 principes d’dition, p. 125*, and n. 12), surmises that the five manuscripts that include
39 Ilāhiyyāt III, 5 (placed at the end of al-Ġazālı̄’s Summa) might reproduce a collection of
works constituted ‘ la source mÞme des traductions’, on account of the presence of the
40
same collection in a number of distinct codices (a fact that indicates an ancient common
41 ancestor) and of the heterogeneous character of the collection (a fact that suggests a
42 dependence on the quaterni of the scholares coming to Toledo from all over Europe).
1 in the first half of the thirteenth century but continued in different ways also
2 later. The replacement of Avicenna by Averroes as ‘Commentator’ of the
3 Metaphysics and the other Aristotelian writings was gradual16 and did not imply
4 a total dismissal of Avicenna’s philosophy, but only a change in the view adopted
5 toward the latter.
6 To summarize: in this as in other cases, historia – like natura – non facit
7 saltus: an uninterrupted line of interpreters can be traced, which starts before
8 William of Auvergne, and continues after the diffusion of Averroes’
9 commentaries. This line begins with the probable translator of the Philosophia
10 prima (Gundissalinus) in the second half of the twelfth century, involves
11 significant authors of the very beginning of the next century, such as John Blund
12 and Michael Scot, passes through a series of fundamental figures of the first half
13 of the following century, such as Robert Grosseteste in Oxford and William of
14 Auvergne and Roger Bacon in Paris, and continues with Albertus Magnus and
15 the other main authors of the second half of the thirteenth century.
16
17
18 §1.2 A Three-fold Periodization
19
20 The Latin reception of Avicenna’s metaphysics presents two main features. First,
the Philosophia prima, i. e. the Latin translation of the Ilāhiyyāt of the Šifā , is the
˘
21
22 only work of Avicenna by means of which Avicennian metaphysics was
23 transmitted into Latin. Second, the fate of Avicenna’s metaphysics in Latin is
24 closely related to the more or less parallel reception of Aristotle’s Metaphysics.
These two features are mutually linked: since the Šifā , by Avicenna’s own
˘
25
26 admission, is the summa of his in which the endorsement of Peripatetic
27 philosophy is most evident, and the Ilāhiyyāt is a reworking of Aristotle’s
28 Metaphysics, it is not surprising that the areas of diffusion of Avicenna’s and
29 Aristotle’s work came to overlap. Moreover, these two traits are peculiar: they
30 sharply distinguish, for example, the Latin side of the reception of Avicenna’s
31 metaphysics from its Arabic counterpart, in which the success of Avicenna’s
stance is not exclusively linked with the Ilāhiyyāt of the Šifā , but is primarily
˘
32
33 connected with other works, and Avicenna’s metaphysics soon replaces, rather
34 than interacting with, Aristotle’s work.
35
36
37
38 16 William of Auvergne, for example, still regards Avicenna as an expositor of Aristotle (De
39 universo II, 8, in Opera omnia, vol. I, p. 690BH: ‘… et Avicenna post eum [sc.
Aristotelem] … Similiter et alii expositores eiusdem Aristotelis’), and refers often in
40
effect to Avicenna when quoting by name Aristotle (as noticed, among others, by Hasse,
41 Avicenna’s ‘De anima’, p. 44 and n. 184; Teske, William of Auvergne’s Debt to Avicenna,
42 pp. 154 – 5).
202 Amos Bertolacci
1 The early Latin reception of the Philosophia prima is reconstructed here on

2 the basis of the relationship that this work, in its different recipients and uses,
3 holds with Aristotle’s Metaphysics. On the basis of this criterion, three main
4 modalities of the reception of the Philosophia prima before Albertus, roughly
5 corresponding to three chronological phases, can be distinguished. The first
6 phase goes from the second half of the twelfth century, when the Philosophia
7 prima was translated, until the beginning of the following century, when the
8 first attestations of its use in European universities occur. This first phase is
9 geographically centered, either directly or indirectly, in the Spanish city of
10 Toledo.17 The second phase is documented since the beginning of the thirteenth
11 century, whereas the third started around 1240: both phases were institutionally
12 linked, in different ways, with the Universities of Paris and Oxford, although
13 they followed distinct paths until Albertus Magnus.
14 In the first phase (Gundissalinus; De causis primis et secundis; the
15 anonymous treatise published by M.-T. d’Alverny in 1940 – 1942; Michael
16 Scot’s writing on the classification of the sciences18), the Philosophia prima is
17 both quoted and silently reproduced within independent treatises, of which it
18 represents the main text, or one of the main texts, on metaphysics. Recourse to
19 Aristotle’s Metaphysics, on the contrary, is absent or very scanty, since the Latin
20 translations of this work have still a very limited diffusion. Besides Avicenna, the
21 other metaphysical sources are works by Arabs and Jews (al-Ġazālı̄’s
22 metaphysics, the Liber de causis and Ibn Gabirol’s Fons vitae), although the
23 Latin metaphysical tradition (Augustine, Boethius, Eriugena) is also influential.
24 In the absence of the metaphysical text par excellence (Aristotle’s Metaphysics),
25 the Philosophia prima, on account of its comprehensiveness and articulacy,
26 performs the role of ‘vicarious’ canonical text. Averroes’ Long Commentary on the
27 ‘Metaphysics’ is not yet available.
28 The second phase (John Blund; Robert Grosseteste; William of Auvergne;
29 Roland of Cremona; Roger Bacon) is marked by the joint consideration of
30 Aristotle’s Metaphysics and Avicenna’s Philosophia prima by philosophers and
31
32 17 It is almost certain that the works of Gundissalinus and Michael Scot belonging to this
33 phase were accomplished in Toledo. On the other hand, the anonymous Liber de causis
34 primis et secundis (a work formerly ascribed to Gundissalinus) might have been written
either in Toledo or in England (see below, n. 37), whereas the place of composition of
35
the anonymous treatise published by D’Alverny might be either Toledo or Bologna (see
36 below, n. 44). Cultural exchanges between Toledo and the rest of Europe were frequent
37 at the time, as the cases of Gerard of Cremona and Daniel of Morley, among others,
38 witness.
39 18 The thesis of Vicaire, Le Porrtains et l’avicennisme, according to which Avicenna would
have influenced the school of Gilbert of Poitiers, is dismissed by Jolivet, The Arabic
40
Inheritance, p. 131 and n. 58, pointing out that ‘when one finds in a twelfth-century
41 writer an idea or a formula that recall … Avicenna, one must not immediately assume
42 that he has been influenced by Avicenna’.
1 theologians in universities. Aristotle’s Metaphysics is now regarded as the main

2 text on the subject, but Avicenna’s Philosophia prima represents the privileged
3 way of access to Aristotle’s work and its main tool of interpretation. Traces of
4 this tendency can be found in Robert Grosseteste in Oxford; its full
5 development occurs, however, in philosophical and theological works produced
6 in Paris. Here, the Philosophia prima is frequently mentioned together with
7 Aristotle’s Metaphysics by masters of arts such as John Blund, and professors of
8 theology such as William of Auvergne, Roland of Cremona and Roger Bacon:
9 all these authors read the Metaphysics through the lenses of the Philosophia
10 prima. The university of Paris documents a progressive acceptance of Avicenna’s
11 work: initially used with no restriction in the arts faculty, as John Blund
12 witnesses, and possibly involved in the Parisian condemnations of 1210 and
13 1215, the Philosophia prima was critically scrutinized, but also widely endorsed,
14
by prime exponents of the faculty of theology such as William of Auvergne and
15
Roland of Cremona, and enthusiastically received, with very few provisos, by
16
Roger Bacon. Averroes’ Long Commentary on the ‘Metaphysics’, on the other
17
hand, once it becomes available, is substantially ignored or even criticized.
18
The third phase (Roger Bacon’s commentaries on the Metaphysics; Oxford
19
commentators of the Metaphysics)19 attests the recourse to the Philosophia prima
20
within the exegesis of the Metaphysics. Averroes’ Long Commentary on the
21
‘Metaphysics’ replaces Avicenna’s Philosophia prima in the role of authoritative
22
23
interpretation of the Metaphysics. Yet, both in Oxford and in Paris,
24
commentators of the Metaphysics continue to refer to the Philosophia prima,
25
even though their references to Avicenna are much less frequent and systematic
26
than those to Averroes’ Long Commentary.
27 For the sake of brevity, these three phases can be labeled, respectively,
28 ‘Philosophia prima without Metaphysics’, ‘Philosophia prima and Metaphysics’,
29 ‘Philosophia prima in the exegesis of the Metaphysics’. They correspond grosso
30 modo to three literary genres (independent treatises influenced by the format of
31 the translation literature; philosophical and theological works produced in
32 universities; commentaries) and to three modalities of the reception of the
33 Philosophia prima (doctrinal endorsement; instrumental use for philosophical
34 and theological purposes; occasional recourse for the explanation of Aristotle).
35 From a sociological point of view, they are linked with different institutional
36 contexts (non-universitarian centers of instruction; faculties of arts and faculties
37 of theology within universities; faculties of arts only). Seen diachronically, they
38 reflect an increasing assimilation of this work: the introduction of the
39 Philosophia prima within the doctrinal debate in the first phase is followed by a
40 period of critical evaluation, which allows the use of the main doctrinal points
41
42 19 The earliest extant Latin commentaries on the Metaphysics date from about 1240.
204 Amos Bertolacci
1 of this work either in philosophical and theological writings in the second phase,
2 or in the exegesis of the Metaphysics in the third phase.
3 Obviously, the proposed periodization is not perfectly rigorous. The chosen
4 arrangement, however, seems to provide a sufficiently coherent and systematic
5 way of understanding the wide and complicated historical event under
6 consideration.
7
8
9 § 2 Philosophia prima without Metaphysics
10 (Gundissalinus; De causis primis et secundis; Anonymous d’Alverny;
11 Michael Scot)
12
13
§ 2.1 The Early Diffusion of the Philosophia prima and of the Latin Translations
14
of Aristotle’s Metaphysics
15
16
According to a widespread contention, the Philosophia prima was known in the
17
Latin world before Aristotle’s Metaphysics. 20 This contention is substantially
18
correct, although it is true with respect to the diffusion, rather than the
19
composition, of the translations of the works under consideration. The
20
Philosophia prima was translated into Latin between 1150 and 1175 in
21
Toledo.21 Two Latin versions of the Metaphysics were produced before or at the
22
same time of the Philosophia prima: the earliest Latin version of the Metaphysics,
23
the so-called Translatio Iacobi sive Vetustissima by James of Venice (active
24
between 1125 and 1150), and the translation called Anonyma sive Media,
25
accomplished by an unknown author of the twelfth century.22 Thus, with regard
26
to their composition, the translation of the Philosophia prima is not
27
28
chronologically prior to that of the Metaphysics.
29
20 De Vaux, Notes et textes, p. 10, states that the works of Avicenna translated into Latin
30
were ‘un ensemble comme on n’en possdait point d’autre alors, pas mÞme d’Aristote,
31 dont les œuvres physiques et mtaphysiques n’arrivrant que plus tard et par tapes’;
32 Goichon, La philosophie d’Avicenne, p. 90: ‘La Mtaphysique d’Avicenne a t connue un
33 demi-sicle avant celle d’Aristote … La philosophie d’Avicenne … tait le premier
34 ensemble de doctrine vraiment constitu qui parvint l’Occident’; De Libera, Penser au
Moyen ffge, p. 112: ‘le texte d’Avicenne est la premire grand œuvre philosophique qui
35
soit parvenu en Occident’.
36 21 See Bertolacci, A Community of Translators.
37 22 Vuillemin-Diem, Praefatio, in Aristoteles Latinus XXV 1 – 1a, p. xxvi. The translation
38 called Vetus, accomplished before 1230 (when it starts to be quoted), is just a revision of
39 the Vetustissima in the form in which this latter is extant (see Vuillemin-Diem, ibid.,
pp. xxix–xxxii; Vuillemin-Diem, Praefatio, in Aristoteles Latinus XXV 3.1, pp. 4 – 5).
40
Burnett, A Note on the Origins, advances a new hypothesis on the origin of the Media:
41 according to him, this translation would have been composed in Antioch, in the second
42 quarter of the 12th century.
1 If we consider, instead, the diffusion of the translations, the situation is

2 different. At an unknown stage of its early transmission, the Vetustissima,
3 originally more encompassing or even complete, underwent the loss of its
4 second part (only the portion corresponding to A–C, 4, 1007a31 is fully extant,
5 whereas excerpts of the following books may have survived as glosses in some
6 manuscripts of the Media).23 On the other hand, the more comprehensive
7 Media (books A–I, K–M) apparently had a limited circulation before the middle
8 of the thirteenth century, when it started gaining diffusion.24 Signs of
9 acquaintance with the Metaphysics can be detected in various authors and
10 works of the second half of the twelfth century, but it is difficult to establish
11 whether these quotations (which often do not concern specific passages of the
12 Metaphysics or, if they do, do not refer explicitly to this work, or which concern
13 doctrines that occur also elsewhere in the Aristotelian corpus) are first-hand or
14
second-hand.25 Even before the translations, Latin scholars could draw
15
information on the Metaphysics from the writings of Boethius and the
16
quotations of these latter in subsequent authors (Abelard, Liber sex principio-
17
rum).26 The lack of a complete Greek-Latin translation of the Metaphysics in the
18
first decades of the thirteenth century (due to the incompleteness of the
19
Vetustissima and the late circulation of the Media) is confirmed by the
20
immediate and wide success of the Arabic-Latin translation of the Metaphysics
21
known as Nova, namely the collection of lemmata of Aristotle’s text taken from
22
23
Michael Scot’s Latin translation of Averroes’ Long Commentary.27
24
23 See Vuillemin-Diem, Praefatio, in Aristoteles Latinus XXV 1 – 1a, pp. xxiv–xxv,
25
Vuillemin-Diem, Praefatio, in Aristoteles Latinus XXV 3.1, p. 3. Since the two
26 manuscripts preserving the Vetustissima in its incontaminated extant form (Avranches,
27 Bibl. Munic., 232; Oxford, Bibl. Bodl., Seld. sup. 24) are of the 12th century, the loss of
28 the second part of this translation probably occurred in this same century (‘schon
29 frhzeitig verlorenen vollstndigeren Text’, Vuillemin-Diem, Praefatio, in Aristoteles
Latinus XXV 3.1, p. 3). Thomas Aquinas might have quoted as alia littera (additional
30
translation) in his commentary on the Metaphysics some fragments of the lost parts of the
31 Vetustissima (see Reilly, The Alia Littera).
32 24 See Vuillemin-Diem, Praefatio, in Aristoteles Latinus XXV 2, pp. xxv-xxxiv; Vuillemin-
33 Diem, Praefatio, in Aristoteles Latinus XXV 3.1, pp. 5 – 7.
34 25 See Vuillemin-Diem, Praefatio, in Aristoteles Latinus XXV 1 – 1a, pp. xv-xvi; Vuillemin-
Diem, Praefatio, in Aristoteles Latinus XXV 3.1, pp. 31 – 2. Also the quotation of
35
Aristotle in Gundissalinus’ De divisione philosophiae, despite its resemblance with a
36 passage of the Metaphysics, is taken more probably from the Physics (see below, § 2.2).
37 26 See Vuillemin-Diem, Praefatio, in Aristoteles Latinus XXV 3.1, pp. 30 – 31; Speer, The
38 Hidden Heritage. Burnett, The Blend of Latin and Arabic Sources, pp. 42 – 3, shows
that themes of the Metaphysics are echoed in Abū Ma šar’s Great Introduction to Astrology
˘
39
(mid 9th century), translated into Latin by Hermann of Carinthia in 1141.
40
27 See Vuillemin-Diem, Praefatio, in Aristoteles Latinus XXV 3.1, pp. 7 – 8. This
41 translation even contaminated the archetype of all the extant codices of the Metaphysica
42 media (see Vuillemin-Diem, Praefatio, in Aristoteles Latinus XXV 2, pp. xxx–xxxi, xlii).
206 Amos Bertolacci
1 All this implies that for a few decades – from the loss of the second part of
2 the Vetustissima, sometime in the second half of the twelfth century, until
3 1220 – 1224, the probable date of Michael Scot’s translation of Averroes’ Long
4 Commentary – the Philosophia prima might have been the only comprehensive
5 account of Aristotelian metaphysics available to Latin philosophers. Later on,
6 the diffusion of the Philosophia prima intersected with the spread of Aristotle’s
7 Metaphysics, known first through the Translatio nova and Averroes’ Long
8 Commentary, then through the Translatio media.
9
10
11 § 2.2 Gundissalinus
12
13 The influence of Avicenna on works of Gundissalinus (d. after 1190) such as
14 the De anima, and on areas of his thought such as epistemology, has already
15 been noticed.28 The Philosophia prima exerted a similar influence on his
16 metaphysics. Gundissalinus is, so to say, ‘originally’ linked with the Latin
17 transmission of Avicenna’s metaphysics: if we accept his traditional identifica-
18 tion with Dominicus Gundisalvi, he was responsible, alone or in cooperation
19 with another scholar, for the translation of this work into Latin.29 Thus, it is not
20 surprising to find that at least two of his original works depend visibly on the
21 Philosophia prima. 30 In the first of these, the De divisione philosophiae, the
22 account of metaphysics – both in its themes and its structure – is based on
23 continuous and extensive implicit quotations of Philosophia prima I, 1 – 3, thus
24 reflecting all the main aspects of Avicenna’s preliminary characterization of the
25 science of metaphysics.31 Since, in this context the silent citations of the
26
27
28 28 Hugonnard-Roche, La classification des sciences; Hasse, Avicenna’s ‘De anima’, pp. 13 –
29 18; Fidora, Die Wissenschaftstheorie des Dominicus Gundissalinus.
29 The distinction of Gundissalinus (or Gundisalvus), author of original works, from
30
Dominicus Gundisalvi, the Latin translator of al-Kindı̄, al-Fārābı̄, Avicenna, al-Ġazālı̄
31 and Ibn Gabirol, proposed by Rucquoi, Gundisalvus ou Dominicus Gundisalvi?, is
32 convincingly rejected by Fidora, Die Wissenschaftstheorie des Dominicus Gundissalinus
33 pp. 14 – 18, and Hasse, The Social Conditions, p. 73 and n. 30.
34 30 Among the other works by Gundissalinus, the De scientiis (a treatise on the classification
of the sciences probably antedating the De divisione philosophiae) is a paraphrase/
35
adaptation (not a bare translation, as sometimes it is portrayed) of al-Fārābı̄’s Ihsā al-
˘
36 ˙˙ De
˘
ulūm (see Hugonnard-Roche, La classification des sciences, p. 41 and nn. 6 – 8). The
37 unitate et uno relies mainly, on the one hand, on Boethius and Augustine, and, on the
38 other hand, on Ibn Gabirol’s Fons vitae (see Jolivet, The Arabic Inheritance, p. 135).
39 31 Gundissalinus, De divisione philosophiae, apparatus fontium ad p. 35, 15-p. 42, 17.
Jolivet, The Arabic Inheritance, p. 136, aptly contends that in this work ‘the influence of
40
al-Fārābı̄ persists, but that of Avicenna is much more prominent’. According to
41 A. Fidora, the influence of the Philosophia prima in Gundissalinus’ account of
42 metaphysics in the De divisione philosophiae is limited to the discussion of the subject-
1 Philosophia prima are occasionally accompanied by equally silent citations of al-

2 Ġazālı̄’s metaphysics,32 we may wonder whether Gundissalinus is not indirectly
3 relying on Avicenna’s authority also when, in other parts of the work, he
4 implicitly quotes al-Ġazālı̄ (as main source) together with Avicenna (as
5 complementary evidence).33 Remarkably, in the account of metaphysics in the
6 De divisione philosophiae Aristotle’s Metaphysics is never quoted.34
7
8 matter of this discipline (see Fidora, Die Wissenschaftstheorie des Dominicus Gundissa-
9 linus, p. 27, n. 21; p. 129; pp. 141 – 2, n. 51), whereas Gundissalinus’ Latin translation
˘
of al-Fārābı̄’s Ihsā al- ulūm (De scientiis) would be the main source of the overall account
˘
10 ˙˙ Verhltnis, p. 72 and n. 16). Although al-Fārābı̄’s Ihsā al- ulūm lies
˘
(see Fidora, Zum
˘
11 certainly in the background of Ilāhiyyāt I, 1 – 3 (see Bertolacci, The Reception,˙˙ p. 464 and
12 nn. 114 – 5), the latter rather than the former seems to be the main and direct source of
13 the account of metaphysics in the De divisione philosophiae. For the influence of the
14 Philosophia prima on other parts of the De divisione philosophie, see Fidora, Die
Wissenschaftstheorie des Dominicus Gundissalinus, p. 155, n. 39.
15
32 See Gundissalinus, De divisione philosophiae, apparatus fontium ad pp. 36 – 7.
16 33 Ibid., apparatus fontium ad pp. 9 – 19 (Prologus), 20, 3 – 10 (Sciencia naturalis).
17 34 Aristotle is quoted in the prologue of the work, with no explicit mention of the work’s
18 title, with regard to the tripartition of theoretical philosophy in physics, mathematics and
19 metaphyics: ‘Unde Aristoteles: ideo scienciarum sunt species tres, quoniam una
speculatur quod movetur et corrumpitur ut naturalis, et secunda quod movetur et non
20
corrumpitur ut disciplinalis; tercia considerat quod nec movetur nec corrumpitur ut
21 divina’ (Gundissalinus, De divisione philosophiae, p. 15, 12 – 15). Neither the apparatus
22 fontium, nor the commentary of the edition (pp. 188 – 90), provides information on the
23 exact provenience of the quotation. This renowned point of Aristotle’s epistemology,
24 however, is reported by Gundissalinus differently than in the Metaphysics: the idea of
corruptibility is totally absent in Metaph. E, 1, 1026a13 – 16, where the objects of the
25
three theoretical sciences are distinguished according to their possession or lack of
26 separation, on the one hand, and motion, on the other. Corruptibility and motion
27 determine, at different levels, the tripartite classification of substances in Metaph. K, 1,
28 1069a30–b2, but this classification conveys only an epistemological bipartition (between
29 two branches of physics and metaphysics, to the exclusion of mathematics) rather than a
tripartition. Gundissalinus’ quotation resembles rather, in a reverse order, Aristotle’s
30
tripartition of theoretical sciences in Phys. B, 7, 198a29 – 30, where metaphysics is
31 portrayed as the science of immovable things, mathematics as the science of movable but
32 incorruptible things, and physics as the science of corruptible things (I wish to thank
33 Resianne Fontaine for having brought this point to my attention; cf. Gundissalinus, De
34 divisione philosophie. ber die Einteilung der Philosophie, p. 68, n. 19). This quotation is
markedly different from the report of Metaph. E, 1, 1026a13 – 16 in Boethius’ De
35
trinitate II, p. 8, 5 – 18, or from the tripartition of the theoretical sciences in Avicenna’s
36 Philosophia prima and in al-Ġazālı̄’s Maqāsid al-falāsifa (Lat. transl. in Algazel’s
37 ˙
Metaphysics, p. 2, 31–p. 3, 32). No specific tripartition of the theoretical sciences occurs
˘
in al-Fārābı̄’s Ihsā al- ulūm or in the treatise De ortu scientiarum associated with the name
˘
38
39 of al-Fārābı̄ in ˙the
˙ Latin tradition. The term disciplinalis used in the quotation to indicate
mathematics echoes the terminology of the Latin translations from Arabic, where this
40
˘
adjective renders two Arabic terms (riyādı̄, ta lı̄mı̄) expressing mathematics (see, for
41 example, Avicenna Latinus, Liber de philosophia ˙ prima sive Scientia divina, I–X, Lexiques,
42 p. 204b). The quotation might therefore be indirect, depending on the Latin translation
208 Amos Bertolacci
1 In the second relevant work by Gundissalinus, the De processione mundi (an

2 original cosmogonical treatise written after 1160), implicit quotations of
3 Philosophia prima I, 6 – 7 can be found in the initial discussion of necessarium
4 esse and possible esse.35 In this work, neither the Metaphysics nor any other
5 Aristotelian writing is referred to, and Avicenna, together with Ibn Gabirol’s
6 Fons vitae, is the main philosophical authority.36
7
8
9 § 2.3 De causis primis et secundis
10
11 The anonymous De causis primis et secundis et de fluxu qui consequitur eas, a
12 treatise on the procession of the world from the first causes through successive
13 stages of emanation, dates to the end of the twelfth, beginning of the thirteenth
14 century.37 Its massive dependence on Avicenna is witnessed by its presence
15 (ascribed to Avicenna and under the title De intelligentiis) in the 1508 Venetian
16 edition of Avicenna’s works. Its editor, R. De Vaux, regarded it as the first and
17 clearest expression of ‘Latin Avicennism’.38 M.-T. d’Alverny and J. Jolivet
18 stressed, more recently, its nature of synthesis between Latin authors (Augustine,
19
Boethius and Eriugena, the latter conveying doctrines of Pseudo-Dionysius) and
20
21 of an Arabic text. Since the description of mathematics in the quotation applies properly
22 to astronomy, the source of the quotation might be a Latin translation of an Arabic text
23 on astronomy.
24 35 De processione mundi, pp. 227 – 30 (§§ 7 – 13 of the list of sources, with reference to
pp. 126 – 38 of the edited text; in § 7, the reference to Philosophia prima I, 7, 26 – 37, is
25
in fact to I, 6, p. 44, 24 – 37). See Jolivet, The Arabic Inheritance, pp. 138 – 40; Soto
26 Bruna, Estudio fil sofico, pp. 34 – 42. A detailed survey of the sources of the De
27 processione mundi is available in The Procession of the World; Burnett, The Blend of Latin
28 and Arabic Sources, pp. 52 – 60 and nn. 33 – 5, has amended the edited text by taking
29 into account a source of the De processione mundi previously disregarded (Hermann of
Carinthia’s De essentiis).
30
36 See Jolivet, The Arabic Inheritance, p. 139 (‘But throughout this treatise the influence of
31 the philosophers plays a major role … he exploits to the full two rich mines of
32 speculative writing: Avicenna’s Metaphysics and the Fons vitae of Ibn Gabirol’); Soto
33 Bruna, Estudio fil sofico, pp. 81 – 95.
34 37 On the date, sources and doctrine of this treatise, see De Vaux, Notes et textes, pp. 63 –
80. D’Alverny, Deux traductions latines, pp. 129 – 30, supposes that ‘Master Maurice’
35
(d. 1238), archdeacon of Toledo and bishop of Burgos, might have been the author of
36 this treatise, on account of his interest in Islamic theology and the philosophy of Pseudo-
37 Dionysius; later on (d’Alverny, Une rencontre symbolique, p. 177), she states that ‘les
38 philosophes et thologiens anglais ont quelque responsabilit dans la diffusion du
39 Pseudo-Avicenne, et peut-Þtre dans sa redaction’, since most of the manuscripts
preserving the De causis primis et secundis are of English origin. See also Jolivet, The
40
Arabic Inheritance, pp. 145 – 6, and Davidson, Alfarabi, Avicenna and Averroes on
41 Intellect, pp. 210 – 11.
42 38 De Vaux, Notes et textes, p. 12 (see above, n. 2).
1 Arabic sources, prominent among which is Avicenna.39 The De causis primis et

2 secundis reproduces in different extents and contexts, for the most part
3 implicitly, a large amount of the Philosophia prima: more precisely, chapters I, 1,
4 II, 4, and III, 1 of its ontological part, and chapters VIII, 7 and IX, 2 – 5 of its
5 theological part.40 Significantly, in its few explicit quotations the Philosophia
6 prima is called Methaphisica or Liber de methaphisica. 41 This expression is
7 applied also to Aristotle’s Metaphysics, which, however, is quoted indirectly
8 through al-Fārābı̄.42 The use of the same title for both suggests that the author
9 of the treatise regarded the two metaphysical works of Avicenna and Aristotle as
10 one and the same writing. Remarkably, the theological part of the Philosophia
11 prima is quoted jointly with the Liber de causis. 43
12
13
14 § 2.4 Anonymous d’Alverny
15
16 The anonymous treatise on human nature and man’s afterlife that M.-T.
17 d’Alverny discovered in MS Paris, BNF, lat. 3236 A, fols 85v–87, and published
18 in 1940 – 42, dates to end of the twelfth century. Its place of composition might
19
be either Spain or Northern Italy.44 D’Alverny describes it as ‘un des tmoins les
20
plus curieux de la conjonction du no-platonisme arabe avec la culture
21
22
23
24 39 D’Alverny, Une rencontre symbolique; Jolivet, The Arabic Inheritance, p. 145.
40 The apparatus fontium of the edition is quite accurate, except for a few imprecisions (the
25
reference to Philosophia prima II, 3 at p. 105, n. 1, is too vague to be considered a
26 quotation; the same can be said of the reference to IX, 2 at p. 116, n. 2; the two
27 quotations at p. 114, nn. 3 – 4, refer to c. IX, 5, rather than IX, 2, of the Philosophia
28 prima).
29 41 Ps.-Avicenna Latinus, Liber de causis primis et secundis, p. 102, 19; p. 107, 8.
42 ‘Prima igitur creaturarum est intellectualis, et est intelligentia de qua est sermo apud
30
philosophum in libro de metaphisica’ (Ps.-Avicenna Latinus, Liber de causis primis et
31 secundis, p. 98, 4 – 7). De Vaux, Notes et textes, p. 98, n. 1, rightly points out that this
32 quotation mirrors a passage of al-Fārābı̄’s De intellectu et intellecto (‘et hoc est intelligencia
33 quam ponit Aristoteles … in libro de metaphisica’, Liber Alpharabii de intellectu et
˘
34 intellecto, p. 126, 393 – 4, cf. p. 115, 9), corresponding to Risāla fı̄ l- aql, p. 36, 1 (cf. p. 4,
3). The philosophus in question, therefore, appears to be Aristotle, not Avicenna, as De
35
Vaux surmises (p. 71; p. 98, n. 1).
36 43 Ps.-Avicenna Latinus, Liber de causis primis et secundis, pp. 108 – 23 (chapters 6 – 8),
37 especially p. 110, n. 1; p. 116, n. 1.
38 44 Whereas the wide array of Arabic sources and the inter-confessional approach point to a
39 Toledan (or Catalan) milieu (see D’Alverny, Les prgrinations de l’me, pp. 266 – 7),
the frequent medical references, as well as some codicological features of the manuscript
40
in which the work is preserved, indicate Bologna as a possible place of composition (see
41 d’Alverny, Les traductions d’Avicenne (Moyen Age et Renaissance), p. 79; d’Alverny,
42 Avicennisme en Italie, pp. 121 – 2).
210 Amos Bertolacci
1 chrtienne occidentale’.45 The Arabic Neoplatonism reflected in this treatise is

2 represented mainly by Avicenna’s Philosophia prima, together with al-Ġazālı̄’s
3 metaphysics, the Liber de causis and Ibn Gabirol’s Fons vitae. Among Latin
4 authors, d’Alverny stresses its dependence on Gundissalinus.46 In this writing,
5 the Philosophia prima is always quoted implicitly, sometimes ad litteram,
6 elsewhere ad sensum. 47 To the first category belongs the quotation of the first
7 sentence of Philosophia prima IX, 7 at the beginning of the treatise.48 The
8 second category includes the similarities with Avicenna’s hierarchy of celestial
9 intelligences in Philosophia prima IX, 4, his description of God’s attributes in
10 VIII, 4 – 7, and his view of the misery of afterlife in IX, 7.49 Aristotle’s
11 Metaphysics, on the other hand, is never quoted.50
12 This treatise is not the only example of the tendency to conjoin the
13 Philosophia prima with Latin metaphysics. D’Alverny points at the existence of
14
other similar, still unedited and uninvestigated, witnesses of the synthesis of
15
Islamic (mainly Avicennian) and Christian Neoplatonism.51
16
17
18
§ 2.5 Michael Scot
19
20
The dependence of Michael Scot (d. 1235 ca.) on Avicenna’s works, with
21
particular regard to the De anima, has been already documented.52 The
22
introduction to philosophy that is fragmentarily preserved in Vincent of
23
Beauvais’ (d. 1264 ca.) Speculum doctrinale witnesses a wide recourse to the
24
Philosophia prima. 53 Michael wrote this introduction probably before his
25
26
27 45 D’Alverny, Les prgrinations de l’me, p. 240.
28 46 D’Alverny, Les prgrinations de l’me, pp. 242 – 3.
29 47 Avicenna might be one of the ‘philosophiae peritissimi’ mentioned in the first pages
(Anonymous, Homo, cum in honore esset …, p. 282, 11).
30
48 Anonymous, Homo, cum in honore esset …, p. 282, 8 – 10.
31 49 Anonymous, Homo, cum in honore esset …, pp. 291 – 3, 297, 299; cf. d’Alverny, Les
32 prgrinations de l’me, pp. 245, 247.
33 50 Aristotle is mentioned only once as author of the Liber de causis (Anonymous, Homo,
34 cum in honore esset …, p. 281, 11 – 12).
51 D’Alverny, Une rencontre symbolique, pp. 178 – 9, mentions an unedited treatise (MS
35
Munich, Bayerische Staatsbibliothek, Lat. 380, fols 21v-24r, described in d’Alverny, Alain
36 de Lille, p. 220; Oxford, Bodleian Library, Digby 28, fols 74r-81v) on the First Cause, the
37 Trinity and the emanation of creatures, in which the account of the Trinity ‘utilize
38 largement des terms tirs de la Mtaphysique d’Avicenne’.
39 52 See Hasse, Avicenna’s ‘De anima’, pp. 23 – 30.
53 Michael Scot, Introduction to Philosophy, fragments 1 – 6. Thus, the dependence of this
40
work of Michael on Arabic sources, pointed out by Baur (Gundissalinus, De divisione
41 philosophiae, pp. 366 – 7; cf. Van Steenberghen, La philosophie au XIIIe sicle, p. 113), can
42 be regarded as a reliance, among others, on Avicenna.
1 translations of Averroes’ long commentaries.54 The Philosophia prima is quoted

2 not only indirectly, through the citations that Michael finds in his main source,
3 Gundissalinus’ De divisione philosophie,55 but also directly. The four-fold
4 division of metaphysics in fragment 5, for example, is taken directly from
5 Philosophia prima I, 2.56 In this case, Michael mentions Avicenna explicitly:
6 however, he refers to the metaphysics of both Avicenna and Aristotle (‘Hae
7 quattuor partes continentur in metaphysica Aristotelis et Avicennae’), as if the
8 metaphysical views of these two authors were one and the same.57 He also adds,
9 accessorily, the names of al-Ġazālı̄ and Ibn Gabirol, together with, but in
10 distinction from, Aristotle and Avicenna (‘… et in Algazel et in Avicebronte’).58
11 This quotation is significant in many respects. First, it is the first known
12 attestation of a joint mention of Aristotle and Avicenna about a metaphysical
13 doctrine: since, in fact, Michael deals with a doctrine of Avicenna, he shows that
14 he takes Avicenna’s metaphysics as representative also of Aristotle’s. Second, this
15 is one of the first cases in which other important exponents of Arabic thought,
16 such as al-Ġazālı̄, are mentioned together with Avicenna on a metaphysical
17 topic. Aristotle’s Metaphysics and al-Ġazālı̄’s summary of Avicenna’s philosophy
18 start to be known, but the Philosophia prima remains for Michael, as for
19 previous authors, the main text on metaphysics.
20
21
22
23
24
25
26
27 54 The absence of any reference to Averroes in these fragments, and their dependence on
28 Gundissalinus, suggest that this work was composed in Toledo in the early period of
29 Michael’s career, i. e. before 1220 (cf. Burnett, Michael Scot, p. 105).
55 L. Baur documents the dependence of Michael’s works on the De divisione philosophie in
30
Gundissalinus, De divisione philosophiae, pp. 365 – 6; see also Fidora, Die Wissenschafts-
31 theorie des Dominicus Gundissalinus, p. 13.
32 56 Gundissalinus, De divisione philosophiae, p. 400, fr. 5; cf. Philosophia prima I, 2, p. 14,
33 68–p. 15, 74 [p. 14, 14 – 18] (the tripartition of metaphysics in Avicenna’s original text is
34 rendered as a quadripartition in the Latin translation). Gundissalinus reports Avicenna’s
view on the articulation of metaphysics differently (see Gundissalinus, De divisione
35
philosophiae, p. 37, 10 – 17), in accordance with another passage of the Philosophia prima
36 (I, 2, p. 16, 2–p. 17, 11 [p. 15, 17–p. 16, 5]). Square brackets refer to the Arabic text.
37 57 Whereas the explicit quotation of Aristotle in fragment 6 is taken from Gundissalinus,
38 the ones in fragments 1 – 2 do not derive from the corresponding passages of the
39 Divisione philosophiae: since in these passages Gundissalinus implicitly quotes doctrines
by Avicenna, Michael might have added the name of Aristotle, on account of the identity
40
of views that he ascribes to Aristotle and Avicenna in the aforementioned passage of
41 fragment 5.
42 58 No division of metaphysics occurs in al-Ġazālı̄’s Summa.
212 Amos Bertolacci
1 § 3 The Entrance of the Philosophia prima at the University of Paris

2 (John Blund; Prohibitions of 1210 and 1215)
3
4 The first evidence at our disposal regarding the second phase of the
5 periodization proposed above comes from masters of arts and professors of
6 theology at the University of Paris during the first two decades of the thirteenth
7 century, who attest the introduction of Avicenna’s Philosophia prima in the
8 university environment and the recourse to this latter as a complement of
9 Aristotle’s Metaphysics. Whereas the exponents of the faculty of arts considered
10 here (John Blund) show a positive attitude towards Avicenna’s work, the faculty
11 of theology has probably expressed a veto against the Philosophia prima,
12 including it in the first Aristotelian condemnations of 1210 and 1215.
13
14
15 § 3.1 John Blund
16
17 The Tractatus de anima of John Blund (1175 ca.–1248) was written in Paris (or
18 in Oxford shortly after the author’s stay in Paris) in the first years of the
19 thirteenth century. It is an important witness of the early diffusion of the
20 Philosophia prima, since it provides the first signs of its consideration as a
21 complement of Aristotle’s Metaphysics. Some of the frequent explicit mentions
22 of Avicenna in this work are quotations of the Philosophia prima, of which both
23 the ontological (I, 1, I, 5, II, 1) and the theological part (IX, 2) are cited.59
24 Aristotle’s Metaphysics, on the other hand, is never referred to directly. In two
25 cases Blund mentions not only Avicenna, but also the title of his work. These
26 mentions are revealing, for in both cases the doctrine quoted is the same – i. e.
27 the classification of substances, including the soul, at the end of Philosophia
28 prima II, 1 – but the cited work of Avicenna, namely the Philosophia prima, is
29 differently described: in one case, Blund calls it Metaphysica, as it was called in
30 the De causis primis et secundis;60 in the other case, by contrast, he portrays it as a
31 ‘commentary’ (commentum) on the Metaphysics. 61 This latter is one of the first
32
33
34 59 See, besides the quotations reported in the following footnotes, also those occurring in
John Blund, Tractatus de anima, § 13 (Philosophia prima IX, 2, p. 387, 12 – 13 [p. 455,
35
10 – 14]), § 18 (Philosophia prima I, 2, p. 10, 6 – 8 [p. 9, 59 – 62]), § 85 (Philosophia
36 prima I, 5, p. 29, 5 – 6 [p. 31, 2 – 3]).
37 60 John Blund, Tractatus de anima, § 316, p. 85, 26-p. 86, 1: ‘ut dicit Avicenna in
38 Metaphysica’. See above, n. 41.
39 61 John Blund, Tractatus de anima, § 32, p. 9, 12 – 13: ‘Ab Avicenna habemus in commento
Metaphysice …’. In another passage, Blund ascribes a commentum prime philosophie to
40
both Avicenna and al-Ġazālı̄ (see John Blund, Tractatus de anima, § 101, p. 27, 28 – 9:
41 ‘sicut testatur tam Avicenna quam Algazel in commento prime philosophie’). Although it
42 reveals awareness of the doctrinal similarity of al-Ġazālı̄’s Summa with respect to
1 known attestations in Latin philosophy of the consideration of Avicenna’s

2 Philosophia prima as a commentary on the Metaphysics. In the Tractatus, Blund
3 establishes an identical relationship of commentary-text commented upon
4 between Avicenna’s and Aristotle’s De anima; in this context, Avicenna is also
5 named Commentator. 62 The term commentum in Blund’s Tractatus does not
6 mean a work providing the literal exegesis of another, but a writing recasting in a
7 different form, and therefore ‘explaining’, the doctrine of another.63 Thus,
8 whereas the first quotation considered here is reminiscent of the phase in which
9 Aristotle’s Metaphysics was substantially absent from the philosophical scenario
10 and replaced with Avicenna’s metaphysics, the second quotation indicates that
11 the Metaphysics starts circulating in the University of Paris at the beginning of
12 the thirteenth century and that scholars grasp the doctrinal affinity between it
13 and the Philosophia prima, regarding this latter as a sort of ‘companion’ to
14
Aristotle’s work. Significantly, in a few cases of joint quotation of Aristotle and
15
Avicenna, the doctrine at stake is phrased according to Avicenna’s text, rather
16
than Aristotle’s.64 However, apart from this and other scattered statements, the
17
actual use of the Philosophia prima in Paris as a tool for the interpretation of the
18
Metaphysics at the turn between the twelfth and the thirteenth century remains
19
undocumented.
20
21
22 § 3.2 The Prohibitions of 1210 – 1215
23
24 Furthermore, the two interdictions of the teaching of Aristotle at the University
25 of Paris in 1210 and 1215, if attentively examined, attest that in the second
26 decade of the thirteenth century Aristotle’s Metaphysics is viewed as distinct from
27 Avicenna’s Philosophia prima, and that this latter is considered as a commentary
28 on the former. In this regard, the interdictions mark the transition from the first
29 to the second phase of the Latin reception of Avicenna’s work.
30 The first prohibition – issued by the council of the ecclesiastic province of
31 Sens, held in Paris in 1210 under the lead of Peter of Corbeil – bans lectures on
32
33 Avicenna’s works (cf. the joint quotation of Avicenna and al-Ġazālı̄ also in § 13), the
34 ascription is puzzling, since the reference is to Avicenna’s and al-Ġazālı̄’s psychology
(doctrine of vision), rather than metaphysics.
35
62 See the loci mentioned by Hasse, Avicenna’s ‘De anima’, p. 20, n. 41.
36 63 See the remarks in Hasse, Avicenna’s ‘De anima’, p. 20. Blund’s consideration of
37 Avicenna’s De anima as a commentary on Aristotle’s text may have influenced his view of
38 the relation of Avicenna’s Philosophia prima vis--vis Aristotle’s Metaphysics.
39 64 See, for instance, John Blund, Tractatus de anima, § 18: ‘Praeterea. Ab Aristotele et ab
Avicenna habemus quod subiectum physici est corpus mobile in quantum ipsum est
40
subiectum motus et quietis’; Philosophia prima I, 2, p. 10, 6 – 8 [p. 9, 59 – 62]: ‘Dico
41 autem quod suum subiectum scientiae naturalis est corpus … inquantum est subiectum
42 motui et quieti’.
214 Amos Bertolacci
1 ‘Aristotle’s physical books’ (libri Aristotelis de naturali philosophia) and the

2 ‘commentaries’ (commenta) thereupon.65 Five years later, the renewed prohib-
3 ition – introduced in the university statutes of 1215 by the delegate of the Pope,
4 Robert of CourÅon – clarifies the range of the books of Aristotle under
5 interdiction (mentioning not only the physical but also the metaphysical books),
6 and describes differently the literary genre of their forbidden explanation,
7 speaking of ‘summaries’ (summae de eisdem).66 By comparing the first with the
8 second prohibition, two noteworthy differences emerge. One is that Aristotle’s
9 Metaphysics is mentioned only in the second prohibition, not in the first. The
10 other is that the writings associated with Aristotle’s works are named commenta
11 in the first prohibition, summae in the second.
12 Since the second prohibition is formally the renewal of the first, the
13 precisions that it introduces can be either clarifications of, or additions to, the
14
previous interdiction. Thus, with regard to the first difference we have noticed,
15
the second prohibition appears to clarify the scope of the books of Aristotle
16
being interdicted, imprecisely reckoned in the first ban: it makes clear that the
17
‘physical books’ of Aristotle are to be taken in a wide meaning and include also
18
the metaphysical writings of the Philosopher.67 This being the case, the
19
commenta on the physical books of Aristotle mentioned in the first prohibition
20
include in all likelihood also ‘commentaries’ on the Metaphysics. On the other
21
hand, with regard to the second difference, it is less clear how the summae in the
22
23
second prohibition relate to the commenta in the first. One possibility is that
24
both expressions refer to the same writings, differently described: we know, for
25
example, that the title of summa (as well as those of compendium or tractatus)
26
was used since the twelfth century by masters of logic to designate explanations
27
28
29 65 Chartularium Universitatis Parisiensis, vol. I, p. 70, § 11: ‘… nec libri Aristotelis de
naturali philosophia nec commenta legantur Parisius publice vel secreto …’.
30
66 Chartularium Universitatis Parisiensis, vol. I, pp. 78 – 9, § 20: ‘Non legantur libri
31 Aristotelis de methafisica et de naturali philosophia nec summe de eisdem …’.
32 67 Although it cannot be excluded that the Metaphysics gained popularity in the University
33 of Paris only after 1210, the historical evidence at our disposal suggests that it was
34 already included in the first prohibition. Grabmann, I divieti ecclesiastici, p. 44, aptly
notices that the first prohibition rests on a division of philosophy in rationalis, naturalis
35
and moralis, in which the philosophia naturalis encompasses also metaphysics. At pp. 11 –
36 12, Grabmann reports a text of William Brito, mentioning, with regard to the first
37 prohibition, the interdiction of libelli quidam ab Aristotele, ut dicebatur, compositi, qui
38 docebant metaphysicam. This text is recorded among the witnesses of the first Latin
39 diffusion of the Metaphysics by Vuillemin-Diem, Praefatio, in Aristoteles Latinus XXV
1 – 1a, p. xvi (Sharpe, A Handlist, p. 756, § 1998, places Brito’s activity towards the end
40
of the 13th century, rather than at its beginning, as Vuillemin-Diem does). The many
41 respects in which Brito’s report differs from the actual text of the proscription are
42 evidenced by Grabmann, ibid.
1 of Aristotle’s works.68 Another possibility, by contrast, is that the summae in the

2 second prohibition are not the same as the commenta in the first: since the
3 second prohibition informs us of the existence of summae not only of Aristotle’s
4 books, but also of the writings of masters such as David of Dinant, Amalricus of
5 Bene and ‘Mauricius Hispanus’ (whoever this latter may be),69 one might take
6 the summae in question to be abstracts and summaries of the works of Aristotle
7 and the other condemned authors, made by professors of the faculty of arts in
8 order to escape the limitations imposed by the proscription of 1210.70 If this is
9 the case, however, it seems difficult to identify the summae with the commenta.
10 Coming to Avicenna, scholars generally agree that both the commenta and
11 the summae mentioned in the prohibitions can be identified with the Latin
12 translations of the Šifā .71 This point of view is, with some provisos, acceptable.
˘
The term commenta in the first prohibition very likely designates the Latin
13
14 translations of Avicenna, for three reasons. First of all, the consideration of the
15 Šifā as intimately connected with Aristotle’s corpus, by being its derivation,
˘
16 complement or explanation, is shared by the first Latin translators of Avicenna72

17
18
19 68 See the evidence discussed in Hasse, Der mutmaßliche Einfluss.
69 ‘Non legantur libri Aristotelis de methafisica et de naturali philosophia nec summe de
20
eisdem aut de doctrina magistri David de Dinant, aut Almarici heretici aut Mauricii
21 hyspani’ (Chartularium Universitatis Parisiensis, vol. I, pp. 78 – 9, § 20; emphasis added).
22 70 This hypothesis is advanced by Mandonnet, Siger de Brabant. The counter-argument of
23 Grabmann, I divieti ecclesiastici, p. 30, namely the fact that no coeval abbreviation of
24 Aristotle’s texts is extant, does not seem conclusive. Daniel of Morley (d. 1210 ca.), in his
famous polemical report on the status of teaching in the university of Paris towards the
25
end of the 12th century, mentions the habit of studying jurisprudence by means of
26 summaries (sub compendio) rather than on the original texts (Daniel of Morley,
27 Philosophia, p. 212).
28 71 De Vaux, Notes et textes, pp. 45 – 52; Grabmann, I divieti ecclesiastici, pp. 49 – 51;
29 d’Alverny, Les prgrinations de l’me, p. 242; Van Steenberghen, La philosophie au XIIIe
sicle, p. 85; Van Riet, La traduction latine, p. 54*; Elamrani-Jamal, La rception,
30
pp. 34 – 5; Bianchi, Censure et libert intellectuelle, p. 94 and n. 20. The initiator of this
31 interpretative tendency, R. De Vaux, points at the compatibility between the literary
32 genre of the Šifā and the terms commenta/summae, in order to include Avicenna’s works
˘
33 in the condemnations and corroborate in this way the hypothesis of the existence of a
34 ‘heretical’ Latin Avicennism, condemned by Parisian theologians. Grabmann, I divieti
ecclesiastici, pp. 49 – 50 (cf. pp. 12 – 13), adds further evidence to De Vaux’s arguments,
35
namely the reference to Spain and Toledo (place of composition of the Latin translations
36 of Avicenna) that can be found in the report of the Parisian condemnations in the
37 Speculum ecclesiae of Gerald of Wales (1146 – 1226); on this author, see Sharpe, A
38 Handlist, pp. 134 – 7 (§ 350).
39 72 Several significant examples can be adduced. (1) In the Prologue of the Latin translation
of Avicenna’s Liber de anima, Avendeuth portrays this work as a book that, in the most
40
complete form (plenissime), gathers and replaces what Aristotle says in his De anima and
41 De sensu et sensato (Avendeuth mentions also the pseudo-Aristotelian De intellectu et
42 intellecto): ‘Habetis ergo librum … ex arabico translatum: in quo quidquid Aristoteles
216 Amos Bertolacci
1 and is widespread at the time.73 Secondly, an author active in Paris a few years
2 before the condemnations like John Blund expressly portrays Avicenna’s De
3 anima and Philosophia prima as commenta of, respectively, Aristotle’s De anima
4 and the Metaphysics, as we have seen. Thirdly, the commentaries par excellence on
5 Aristotle’s works, namely Averroes’ long commentaries, cannot be the commenta
6 referred to in the 1210 prohibition, since they were translated into Latin only
7 later (around 1220 – 1235).74 Thus, the Philosophia prima is quite probably
8 alluded to in the first prohibition as commentum on the Metaphysics. The
9 summae in the second prohibition, on the other hand, do (or do not) designate
10 the Šifā , depending on whether they are (or they are not) the same as the
˘
11
12 dixit in libro suo de anima et de sensu et sensato et de intellectu et intellecto, ab auctore
13 libri sciatis esse collectum; unde, postquam, volente Deo, hunc habetis, in hoc illos tres
14 plenissime vos habere non dubitetis’ (Avicenna Latinus, Liber de anima, p. 4, 21 – 5). On
the Latin translation of the De intellectu of Alexander of Aphrodisias, see Burnett, Arabic
15
into Latin: the Reception of Arabic Philosophy into Western Europe, p. 392 and n. 11.
16 (2) Likewise, Alfred of Sareshel prolonged Aristotle’s meteorology with the mineralogy
17 contained in chapters I, 1 and I, 5 of Avicenna’s corresponding section, thus revealing an
acute perception of the Aristotelian inspiration of the Šifā , especially if he regarded these
˘
18
19 chapters as written by Avicenna rather than by Aristotle himself (according to Otte,
Alfred of Sareshel’s Commentary, Alfred did not know Avicenna’s authorship of these two
20
chapters, a suggestion discarded by Mandosio, Di Martino, La Mtorologie d’Avicenne,
21 pp. 413 – 15). Significantly, Alfred commented not only on Aristotle’s Meteorology, but
22 also on Avicenna’s aforementioned chapters, and described Avicenna as imitator of
23 Aristotle and as the second most important philosophical authority after the Stagirite
24 (‘imitator Aristotelis precipuus, immo ipso Aristotele excepto, philosophorum maximus’,
Alfred of Sareshel, Commentary, p. 50, 18 – 19). (3) Michael Scot’s translation of
25
Avicenna’s reworking of Aristotle’s Historia animalium, De partibus animalium and De
26 generatione animalium followed later his translation, from Arabic, of these three
27 Aristotelian works (as a single unit, with the title De animalibus). It is reasonable to
28 suppose that, after having translated Aristotle’s works on zoology, Michael wanted to
provide, with the translation of the part of the Šifā on this topic, its interpretative tool.
˘
29
(4) By inserting the Latin translation of fragments of the rhetoric of the Šifā into his
30
Arabic-Latin translation of Aristotle’s Rhetoric, also Hermann the German aimed at ˘
31 clarifying the obscurities of the Arabic text of Aristotle’s work (see d’Alverny, Notes,
32 pp. 339, 347). On account of all this evidence, it is safe to conclude that the first
translators of the Šifā ‘intended to provide Western scholars with a commentary on
˘
33
34 Aristotle’s works’ (D’Alverny, Translations and Translators, p. 451).
73 See De Vaux, Notes et textes, p. 10: ‘Et cet ensemble [des traductions d’Avicenne] passait
35
pour Þtre un commentaire autoris – le meilleur, mieux: le seul – de toute la philosophie
36 aristotelicienne’; Goichon, La philosophie d’Avicenne, p. 90: ‘Celui-ci [i.e. le Šifā ] passait
˘
37 pour un commentaire’; Van Steenberghen, La philosophie au XIIIe sicle, p. 173: ‘les

38 paraphrases d’Avicenne ont servi d’instrument de travail aux premiers exgtes d’Aristote,
39 jusq’au jour o
elles ont t dtrnes par les commentaries littraux d’Averros’; Soto
Bruna, Estudio fil sofico, p. 38: ‘Este ultimo [i.e. el Šifā ] fue considerado en su
˘
40
momento como un comentario de la propia Metafsica de Arist teles’.
41 74 Denifle contends the contrary in Chartularium Universitatis Parisiensis, vol. I, p. 71,
42 n. 15.
1 commenta in the first prohibition. Also apart from stylistic considerations (the
term summa fits the literary format of the Šifā quite well, and it is even
˘
2
3 contained in the title of the Latin translation of al-Ġazālı̄’s Maqāsid, i. e. Summa
4 theoricae philosophiae), it is not impossible to take the occurrence ˙ of summae in
75
the second prohibition as referring to the Šifā .
˘
5
6 Thus, it seems safe to conclude that the Philosophia prima entered the
7 curriculum of the faculty of arts of the University of Paris some time before
8 1210, playing there the role of an authoritative text to be read together with the
9 Metaphysics in order to convey its interpretation. The quotations of the
10 Philosophia prima in John Blund fit into this scenario.76 Together with the
11 Metaphysics and the other writings of Aristotle, Avicenna’s work must have
12 aroused the suspicion and alarm of the members of the faculty of theology, who
13 promoted the condemnations of 1210 and 1215 in front of the ecclesiastic
14 authorities. This is confirmed a posteriori. When the prohibitions lose their
15 validity, and Aristotle’s writings were ‘rehabilitated’ in Paris in 1231, the
16 Philosophia prima regained its role of interpretive tool of the Metaphysics. The
17 Parisian ‘Guide of the Student’ of 1230 – 1240, for example, reveals a certain
18 silent influence of Avicenna in metaphysics.77 The authors active in Paris in the
19 fourth decade of the thirteenth century will rely massively on the Philosophia
20 prima, providing the first known attestations of the use of this work no longer
21 without, but together with the Metaphysics.
22
23
24
25
26
27
28
29 75 In this prohibition, the term summae might have two different interrelated connotations:
one indicating the explanatory summaries of Aristotle’s books on metaphysics and
30
natural philosophy, in accordance with the attested practice of naming summae the
31 exegetical works produced within the faculty of arts (see above, n. 68); another
32 designating more specifically the abridgements of the teaching of the masters involved in
33 the condemnation. In its first meaning, the term summae would encompass the Šifā . ˘
34 76 Other works written in the faculty of arts of the University of Paris at the beginning of
the 13th century – like the surviving fragments of the Quaternuli of David of Dinant,
35
condemned in 1210 – show acquaintance with the Metaphysics, but no significant
36 recourse to the Philosophia prima (see Anzulewicz, Person und Werk des Davids von
37 Dinant; Anzulewicz, David von Dinant, pp. 81, 90). Vuillemin-Diem, Zum Aristoteles
38 Latinus, p. 30, remarks, however, that the extant fragments of David’s Quaternuli
39 ‘reprsentieren zweifellos nur einen sehr kleinen Teil aus der verlorenen wissenschaft-
lichen und literarischen Produktion Davids’.
40
77 See De Libera, Structure du corps scolaire, p. 75: ‘Cette conception du systme de la
41 mtaphysique comme science [dans la Guide] n’est pas trangre la presentation de la
42 prima philosophia dans la Mtaphysique du Shifā d’Avicenne’.
˘
218 Amos Bertolacci
1 Conclusion
2
3 The stage of the Latin reception of the Philosophia prima that antedates William
4 of Auvergne is quite rich and interesting. Several authors and works are
5 involved, and virtually the entire Philosophia prima is taken into account.
6 Historically, this initial phase connects the period of the translation of the
7 Philosophia prima in the second half of the twelfth century with its employment
8 by theologians in Oxford and Paris from the third decade of the following
9 century onward. Doctrinally, the reception of the Philosophia prima in this early
10 phase is worth considering. On the one hand, it is still, in a way, immature,
11 since it mainly consists in the repetition, often silent, of Avicenna’s views on
12 scattered topics, rather than in their critical evaluation and theoretical
13 refinement. On the other hand, however, authors focus on some crucial points
14 of Avicenna’s metaphysics: this is the case of chapter I, 6 of the Philosophia
15 prima (the distinction of necessary and contingent, and of necessary per se and
16 necessary in virtue of something else) in Gundissalinus’ De processione mundi;
17 and of chapter I, 5 (the idea of ‘existent’ as first intelligible) in John Blund’s
18 Tractatus de anima. Thus, distinctions and doctrines that are central in
19 Avicenna’s metaphysics, and provide evidence of the endorsement of Avicenna’s
20 thought in later authors, are already at stake in this early stage. More than its
21 features, the very existence of this stage is significant. It attests that the
22 transmission of Avicenna’s metaphysics into Latin represent a historical and
23 doctrinal continuum.
24 Future research will have to investigate whether the absence of Aristotle’s
25 Metaphysics at this early stage is just accidental, or rather is causally linked with
26 the diffusion of the Philosophia prima. The Metaphysics is occasionally
27 mentioned by Latin authors at this stage, but knowledge of it remains little
28 more than virtual. The fact that two Latin translations of the Metaphysics were
29 made in the twelfth century, but underwent a partial loss or remained
30 unexploited until later, calls for an explanation. One might think that interest in
31 metaphysical issues at this stage was, in general, not too strong. But one might
32 also surmise that the success of the Philosophia prima somehow prevented the
33 diffusion of the Metaphysics, by providing a metaphysical system that was
34 regarded by Latin scholars as more coherent and complete than Aristotle’s.
35 These are questions that wait to be answered. The fact remains that Avicenna’s
36 Philosophia prima seems to have spread in Latin philosophy before and, initially,
37 without the Metaphysics.
38
39
40
41
42
1 Bibliography
2
3 Primary Sources
4
5 al-Fārābı̄, Liber Alpharabii de intellectu et intellecto, in Gilson, Les sources greco-arabes
6 de l’augustinisme avicennisant, Archives d’Histoire doctrinale et littraire du Moyen
Age, 4, 1929, pp. 5 – 149 (pp. 115 – 26).
7
˘
–– , Risāla fı̄ l- aql, ed. M. Bouyges, Beirut: Imprimerie Catholique, 1938 (reprinted
8 Beirut: Dār al-Mašriq, 1983).
9 Alfred of Sareshel, Commentary on the ‘Meteora’ of Aristotle, ed. J.K. Otte, Leiden/New
10 York/Copenhagen/Cologne: Brill, 1988.
11 al-Ġazālı̄, Algazel’s Metaphysics. A Mediaeval Translation, ed. J.T. Muckle, Toronto:
12 St. Michael’s College, 1933.
13
Anonymous, Homo, cum in honore esset, … (MS Paris, Bibliothque Nationale,
Lat. 3236 A, fols 85v-87), in M.-T. d’Alverny, Les prgrinations de l’me dans
14
l’autre monde d’aprs un anonyme de la fin du XIIe sicle, Archives d’Histoire
15 doctrinale et littraire du Moyen Age, 15 – 17, 1940 – 42, pp. 280 – 99.
16 Avicenna Latinus, Liber de anima seu Sextus de Naturalibus, I-II-III, ed. S. Van Riet,
17 Louvain/Leiden: Peeters/Brill, 1972.
18 –– , Liber de philosophia prima sive Scientia divina, I-IV, ed. S. Van Riet, Louvain/
19
Leiden: Peeters/Brill, 1977.
–– , Liber de philosophia prima sive Scientia divina, V-X, ed. S. Van Riet, Louvain/
20
21 –– , Liber de philosophia prima sive Scientia divina, I-X, ed. S. Van Riet, Louvain/Leiden:
22 Peeters/Brill, 1983.
23 Ps.-Avicenna Latinus, Le Liber de causis primis et secundis et de fluxu qui consequitur eas,
24 in R. De Vaux, ed., Notes et textes sur l’avicennisme latin aux confins des XIIe-XIIIe
25
sicles, Paris: Vrin, 1934, pp. 83 – 140.
Boethius, De trinitate, in The Theological Tractates, transls and eds H. F. Stewart, E. K.
26
Rand, London/Cambridge, Massachusetts: Heinemann/Harvard University Press,
27 1962, pp. 2 – 31.
28 Chartularium Universitatis Parisiensis, ed. H. Denifle, A. Chatelain, 4 vols, Paris 1889 –
29 97; reprinted Bruxelles: Culture et Civilisation, 1964.
30 Daniel of Morley, Philosophia, ed. G. Maurach, in Mittellateinisches Jahrbuch, 14, 1979,
31 pp. 204 – 55.
Gundissalinus (Dominicus Gundisalvi), De divisione philosophiae, ed. L. Baur, Mnster:
32
Aschendorff, 1903.
33 –– , De divisione philosophiae. ber die Einteilung der Philosophie. Lateinisch-Deutsch,
34 transl and ed. A. Fidora und D. Werner, Freiburg/Basel/Vienna: Herder 2007.
35 –– , De processione mundi. Estudio y edicin crtica del tratado de Domingo Gundisalvo, eds
36 M. J. Soto Bruna, C. Alonso Del Real, Pamplona: Ediciones Universidad de
37 Navarra, 1999.
–– , The Procession of the World (De processione mundi), transl. and ed. J. A. Laumakis,
38
Milwaukee: Marquette University Press, 2002.
39 John Blund, Tractatus de anima, ed. D. A. Callus, R. W. Hunt, London: Oxford
41 Michael Scot, Introduction to Philosophy (Fragments), in De divisione philosophiae, ed.
42 L. Baur, Mnster: Aschendorff, 1903, pp. 364 – 7, 398 – 400.
220 Amos Bertolacci
1 William of Auvergne, Opera omnia, 2 vols, ed. F. Hotot, Supplementum, ed. B. LeFeron,
2 Paris/Orlans 1674; reprinted Frankfurt a. M.: Minerva, 1963.
3
4
5 Secondary Sources
6
7
M.-T. d’Alverny, Alain de Lille. Textes indits. Avec une introduction sur sa vie et ses œuvres,
Paris: Vrin, 1965.
8
–– , Avicennisme en Italie, in Accademia Nazionale dei Lincei, Fondazione A. Volta, eds,
9 Oriente e Occidente nel Medioevo: Filosofia e Scienze (Atti del Congresso
10 Internazionale, 9 – 15 aprile 1969; n8 13), Rome: Accademia Nazionale dei Lincei,
11 1971, pp. 117 – 39; reprinted in Avicenne en Occident, Paris: Vrin, 1993, XVI.
12 –– , Deux traductions latines du Coran au Moyen Age, Archives d’Histoire doctrinale et
13 littraire du Moyen Age, 22 – 23, 1947 – 48, pp. 69 – 131; reprinted in La
connaissance de l’Islam dans l’Occident mdival, ed. C. Burnett, Aldershot: Ashgate,
14
1994.
15 –– , Les prgrinations de l’me dans l’autre monde d’aprs un anonyme de la fin du
16 XIIe sicle, Archives d’Histoire doctrinale et littraire du Moyen Age, 15 – 17, 1940 –
17 42, pp. 239 – 99.
18 –– , Les traductions d’Avicenne (Moyen Age et Renaissance), in Problemi attuali di
19 Scienza e di Cultura, Quaderno n8 40 (Avicenna nella storia della cultura
medievale), Rome: Accademia Nazionale dei Lincei, 1957, pp. 73 – 87; reprinted in
20
Avicenne en Occident, Paris: Vrin, 1993, V.
21 –– , Les traductions d’Avicenne: quelques rsultats d’une enquÞte, in R. Marchal, ed.,
22 Actes du Ve Congrs international des Arabisants (Bruxelles, 1970), Bruxelles: Centre
23 pour l’tude des Problmes du Monde Musulman Contemporain, 1971,
24 pp. 151 – 8; reprintred in Avicenne en Occident, Paris: Vrin, 1993, VII.
25 –– , Notes sur les traductions mdivales d’Avicenne, Archives d’Histoire doctrinale et
littraire du Moyen Age, 19, 1952, pp. 337 – 58 reprinted in Avicenne en Occident,
26
Paris: Vrin, 1993, IV.
27 –– , Une rencontre symbolique de Jean Scot Erigne et d’Avicenne. Notes sur le De
28 Causis Primis et Secundis et Fluxu qui consequitur eas, in S. O’Meara, C. Bieler,
29 eds, The Mind of Eriugena, Dublin: Irish University Press, 1973, pp. 170 – 81.
30 –– , Translations and Translators, in R.L. Benson, G. Constable, eds, Renaissance and
31 Renewal in the Twelfth Century, Cambridge, Massachusetts: Harvard University
Press, 1982, pp. 421 – 62.
32
G.C. Anawati, La Mtaphysique d’Avicenne dans l’Occident latin, in Avicenna, La
33 Mtaphysique du Shifā . Livres I V, transl. and ed. G.C. Anawati, Paris: Vrin, 1978,
˘
34 pp. 56 – 79.
35 H. Anzulewicz, David von Dinant und die Anfnge der aristotelischen Naturphilosophie
36 im Lateinischen Westen, in L. Honnefelder, R. Wood, M. Dreyer, M.-A. Aris, eds,
37 Albertus Magnus und die Anfnge der Aristoteles-Rezeption im lateinischen Mittelalter.
38
Von Richard Rufus bis zu Franciscus de Mayronis, Mnster: Aschendorff, 2005,
pp. 71 – 112.
39
–– , Person und Werk des Davids von Dinant im literarische Zeugnis Alberts des
40 Grossen, Mediaevalia Philosophica Polonorum, 34, 2001, pp. 15 – 58.
41 A. Bertolacci, A Community of Translators: The Latin Medieval Versions of Avicenna’s
42 Book of the Cure, in J.N. Crossley, C.J. Mews, eds, Communities of Learning:
1 Networks and the Shaping of Intellectual Identity in Europe 1100 – 1450, Turnhout:
2 Brepols, 2011, pp. 37–55.
–– , The Reception of Aristotle’s ‘Metaphysics’ in Avicenna’s ‘Kitāb al-Šifā ’: A Milestone of
˘
3
Western Metaphysical Thought, Leiden: Brill, 2006.
4 L. Bianchi, Censure et libert intellectuelle l’universit de Paris, Paris: Les Belles Lettres,
5 1999.
6 C. Burnett, A Note on the Origins of the Physica Vaticans and Metaphysica Media, in
7 R. Beyers, J. Brams, D. Sacr, K. Verrycken, eds, Tradition et traduction. Les textes
8
philosophiques et scientifiques grecs au Moyen ffge latin. Hommage Fernand Bossier,
Leuven: Leuven University Press, 1999, pp. 59 – 68.
9
–– , Arabic into Latin: the Reception of Arabic Philosophy into Western Europe, in
10 P. Adamson, R. Taylor, eds, The Cambridge Companion to Arabic Philosophy,
11 Cambridge: Cambridge University Press, 2005, pp. 370 – 404.
12 –– , Michael Scot and the Transmission of Scientific Culture from Toledo to Bologna
13 via the Court of Frederick II Hohenstaufen, Micrologus, 2, 1994, pp. 101 – 26.
–– , The Blend of Latin and Arabic Sources in the Metaphysics of Adelard of Bath,
14
Hermann of Carinthia, and Gundisalvus, in M. Lutz-Bachmann, A. Fidora,
15 A. Niederberger, eds, Metaphysics in the Twelfth Century. On the Relationship among
16 Philosophy, Science and Theology, Turnhout: Brepols, 2004, pp. 41 – 65.
17 H.A. Davidson, Alfarabi, Avicenna and Averroes on Intellect. Their Cosmologies, Theories
18 of the Active Intellect, and Theories of Human Intellect, New York/Oxford: Oxford
University Press, 1992.
19
A. De Libera, Penser au Moyen ffge, Paris: ditions du Seuil, 1991.
20 –– , D’Avicenne Averros, et retour. Sur les sources arabes de la thorie scolastique de
21 l’un transcendental, Arabic Sciences and Philosophy, 4, 1994, pp. 141 – 79.
22 –– , Proofs for Eternity, Creation and the Existence of God in Medieval Islamic and Jewish
23 Philosophy, New York/Oxford: Oxford University Press, 1987.
–– , Structure du corps scolaire de la mtaphysique dans la premire moiti du XIIIe
24
sicle, in C. Lafleur, J. Carrier, eds, L’enseignement de la philosophie au XIIIe sicle.
25 Autour du ‘Guide de l’tudiant’ du ms. Ripoll 109, Turnhout: Brepols, 1997, pp. 61 –
26 88.
27 R. De Vaux, Notes et textes sur l’avicennisme latin aux confins des XIIe-XIIIe sicles, Paris:
28 Vrin, 1934.
29
A. Elamrani-Jamal, La rception de la philosophie arabe l’univesit de Paris au XIIIe
sicle, in C.E. Butterworth, B.A. Kessel, eds, The Introduction of Arabic Philosophy
30 into Europe, Leiden/New York/Kçln: Brill, 1994, pp. 30 – 39.
31 A. Fidora, Die Wissenschaftstheorie des Dominicus Gundissalinus. Voraussetzungen und
32 Konsequenzen des zweitens Anfangs der aristotelischen Philosophie im 12. Jahrhundert,
33 Berlin: Akademie Verlag, 2003.
34
–– , Zum Verhltnis von Metaphysik und Theologie bei Dominicus Gundissalinus, in
M. Lutz-Bachmann, A. Fidora, A. Niederberger, eds, Metaphysics in the Twelfth
35 Century. On the Relationship among Philosophy, Science and Theology, Turnhout:
36 Brepols, 2004, pp. 67 – 84.
37 T.E. Gaskill, Was Leibniz an Avicennian? Leibniz and Avicenna on the Principle of
38 Sufficient Reason, Proceedings of the Patristic, Medieval and Renaissance Conference,
39
16 – 17, 1992 – 3, pp. 105 – 13.
. Gilson, Les sources greco-arabes de l’augustinisme avicennisant, Archives d’Histoire
40 doctrinale et littraire du Moyen Age, 4, 1929, pp. 5 – 149.
41 –– , Pourquoi saint Thomas a critiqu saint Augustin, Archives d’Histoire doctrinale et
42 littraire du Moyen Age, 1, 1926 – 7, pp. 5 – 127.
222 Amos Bertolacci
1 A.-M. Goichon, La philosophie d’Avicenne et son influence en Europe mdivale, 2nd edn,
2 Paris: Maisonneuve 1951.
3
M. Grabmann, I divieti ecclesiastici di Aristotele sotto Innocenzo III e Gregorio IX,
Miscellanea Historiae Pontificiae, Vol. V, Collect. 7: I Papi del Duecento e
4
l’Aristotelismo, fasc. I, Roma: Saler, 1941.
5 A. Hasnawi, La conscience de soi chez Avicenne et Descartes, in J. Biard, R. Rashed, eds,
6 Descartes et le Moyen ffge, Paris: Vrin, 1997, pp. 283 – 91.
7 D.N. Hasse, Avicenna’s ‘De anima’ in the Latin West. The Formation of a Peripatetic
9
–– , Arabic Philosophy and Averroism, in J. Hankins, ed., Cambridge Companion to
10
Renaissance Philosophy, Cambridge: Cambridge University Press, 2007, pp. 113 –
11 36.
12 –– , Der mutmaßliche arabische Einfluss auf die literarische Form der Universittsliter-
13 atur des 13. Jahrhunderts, in L. Honnefelder, ed., Albertus Magnus und der Ursprung
14 der Universittsidee, Berlin: Berlin University Press, 2011, pp. 241–58 and
15
pp. 487–91.
–– , The Social Conditions of the Arabic-(Hebrew-)Latin Translation Movement in
16
Medieval Spain and in the Renaissance, in A. Speer, ed., Wissen
ber Grenzen.
17 Arabisches Wissen und lateinisches Mittelalter, Berlin/New York: De Gruyter, 2006,
18 pp. 68 – 86.
19 H. Hugonnard-Roche, La classification des sciences de Gundissalinus et l’influence
20 d’Avicenne, in J. Jolivet and R. Rashed, eds, tudes sur Avicenne, Paris: Les Belles
21
Lettres, 1984, pp. 41 – 75.
J. Jolivet, L’pistmologie de Descartes dans le Regulae et celle d’Avicenne, in J. Biard
22
and R. Rashed, eds, Descartes et le Moyen ffge, Paris: Vrin, 1997, pp. 187 – 97.
23 –– , The Arabic Inheritance, in P. Dronke, ed., A History of Twelfth-Century Western
24 Philosophy, Cambridge: Cambridge University Press, 1987, pp. 113 – 48.
25 H. Kishlat, Studien zur Verbreitung von bersetzungen arabischer philosophischer Werke in
26 Westeuropa 1150 – 1400. Das Zeugnis der Bibliotheken, Mnster: Aschendorff, 2000.
27
P. Mandonnet, Siger de Brabant et l’averrosme latin au XIIIe sicle, Fribourg (Suisse),
1899; 2nd edn: Louvain: Institut suprieur de philosophie, 1908 – 11 (2 vols).
28 J.-M. Mandosio, C. Di Martino, La Mtorologie d’Avicenne (Kitāb al-Šifā V) et sa
˘
29 diffusion dans le monde latin, in A. Speer, ed., Wissen
ber Grenzen. Arabisches
30 Wissen und lateinisches Mittelalter, Berlin/New York: De Gruyter, 2006, pp. 406 –
31 24.
32 J.K. Otte, Alfred of Sareshel’s Commentary on Avicenna’s De congelatione et
conglutinatione lapidum, in G. Freibergs, ed., Aspectus et affectus. Essays and Editions
33
in Grosseteste and Medieval Intellectual Life in Honor of R. C. Dales, New York: AMS
34 Press, 1993, pp. 105 – 11.
35 M. Rashed, Thodice et approximation: Avicenne, Arabic Sciences and Philosophy, 10,
36 2000, pp. 223 – 57.
37 J.P. Reilly, The Alia Littera in Thomas Aquinas’ Sententia Libri Metaphysicae, Medieval
38 Studies, 50, 1988, pp. 559 – 83.
M.-D. Roland-Gosselin, Introduction, Notes et tudes historiques, in Le ‘De ente et
39
essentia’ de S. Thomas d’Aquin. Texte tabli d’aprs les manuscripts parisiens, Paris:
40 Vrin, 1926; 2nd edn: 1948.
41 A. Rucquoi, Gundisalvus ou Dominicus Gundisalvi?, Bulletin de Philosophie Mdivale,
42 41, 1999, pp. 85 – 106.
1 C.B. Schmitt, D. Knox, Pseudo-Aristoteles Latinus. A Guide to Latin Works Falsely

2 Attributed to Aristotle before 1500, London: The Warburg Institute, 1985.
3
R. Sharpe, A Handlist of the Latin Writers of Great Britain and Ireland Before 1540,
Turnhout: Brepols, 2001.
4 M.J. Soto Bruna, Estudio fil sofico. Introducci n y comentario al De Processione Mundi,
5 in Gundissalinus, De Processione Mundi. Estudio y edicin crtica del tratado de
6 Domingo Gundisalvo, eds M.J. Soto Bruna, C. Alonso Del Real, Pamplona:
7 Ediciones Universidad de Navarra, 1999, pp. 17 – 107.
8
A. Speer, The Hidden Heritage: Boethian Metaphysics and Its Medieval Tradition,
Quaestio, 5, 2005, pp. 163 – 81.
9 R. Teske, William of Auvergne’s Debt to Avicenna, in J. Janssens and D. De Smet, eds,
10 Avicenna and His Heritage. Proceedings of the International Colloquium ‘Avicenna
11 and his Heritage’, Leuven-Louvain-la-Neuve, 8 – 11 September 1999, Leuven:
12 Leuven University Press, 2002, pp. 153 – 70.
13
S. Van Riet, La traduction latine, Annexes, in Avicenna Latinus, Liber primus
naturalium. Tractatus primus de causis et principiis naturalium, ed. S. Van Riet,
14 Peeters/Brill, Louvain/Leiden 1992, pp. 53*-105*.
15 –– , Traduction latine et principes d’dition, in Avicenna Latinus, Liber de philosophia
16 prima sive Scientia divina, I–IV, ed. S. Van Riet, Louvain/Leiden: Peeters/Brill,
17 1977, pp. 123*–66*.
18
F. Van Steenberghen, La philosophie au XIIIe sicle, Louvain/Paris: Publications
Universitaires-Nauwelaerts, 1966; 2nd edn: Louvain/Paris: Peeters, 1991.
19 G. Verbeke, Avicenna’s Metaphysics and the West, in M. Wahba, ed., Islam and
20 Civilization, Cairo: Ain Shams University Press, 1982, pp. 53 – 64.
21 M.H. Vicaire, Le Porrtains et l’avicennisme avant 1215, Revue des sciences philosophiques
22 et thologiques, 26, 1937, pp. 449 – 82.
23
G. Vuillemin-Diem, Praefatio, in Aristoteles Latinus XXV 1 – 1a, Metaphysica lib. I-IV.4.
Translatio Iacobi sive ‘Vetustissima’ cum Scholiis et Translatio Composita sive ‘Vetus’, ed.
24 G. Vuillemin-Diem, Bruxelles/Paris: Descle de Brouwer, 1970, pp. xi–lix.
25 –– , Praefatio, in Aristoteles Latinus XXV 2, Metaphysica lib. I–X, XII–XIV. Translatio
26 Anonyma sive ‘Media’, ed. G. Vuillemin-Diem, Leiden: Brill, 1976, pp. ix–lxix.
27 –– , Praefatio, in Aristoteles Latinus XXV 3.1, Metaphysica lib. I–XIV. Recensio et
28
Translatio Guillelmi de Moerbeka, ed. G. Vuillemin-Diem, Leiden/New York/
Cologne: Brill, 1995.
29 –– , Zum Aristoteles Latinus in den Fragmenten der Quaternuli des David von Dinant,
30 Archives d’Histoire doctrinale et littraire du Moyen Age, 70, 2003, pp. 27 – 136.
31 H. Yaldir, Ibn Sı̄nā (Avicenna) and Ren Descartes on the Faculty of Imagination,
32 British Journal for the History of Philosophy, 17.2, 2009, pp. 247 – 78.
33
34
35
36
37
38
39
40
41
42
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
1
2
3
4 Avicenna’s ‘Giver of Forms’ in Latin Philosophy,
5 Especially in the Works of Albertus Magnus
6
7 Dag Nikolaus Hasse
8
9
The giver of forms (wa-hib as-suwar, dator formarum) is a piece of Avicennian
10 ˙˙
philosophy that went against the grain of most scholastic philosophers.1 It is part
11
of Avicenna’s emanation theory, which the Latins knew from books 8 and 9 of
12
the Ila-hiyya-t (Divine Things) of Kita-b aš-Šifa- (Book of the Cure): the emanation
˘
13
14
of intelligences and accompanying celestial spheres from the first cause, the
15
necessary being (wa-ğib al-wūğūd, necesse esse), which is an eternal efficient cause.
16
In a number of passages, Avicenna calls one of the celestial intelligences the
17
‘giver of forms’.2 He apparantly refers to the lowest intelligence, from which
18 emanate the substantial forms of the sublunar world.3 This intelligence is called
19 ‘the active intellect’ in other passages.4 The forms emanate from the lowest
20 intelligence when the elemental mixture reaches a certain disposition towards a
21 form.
22 In most of his writings, Avicenna uses the concept of a giver of forms not in
23 an epistemological but in an ontological sense: the wa-hib as-suwar is not the
˙˙
24 giver of intelligible forms, but of the forms that combine with prepared matter.5
25
26 1 I am grateful for the advice of Amos Bertolacci, Jon Bornholdt, Katrin Fischer, Jçrn
27 Mller, Adam Takahashi and for suggestions from the audiences in Menaggio, Jena and
28 Berlin (Leibniz-Kreis), where the paper was presented. Research on this paper was
29 funded by the Volkswagen Foundation.
Avicenna mentions the ‘giver of forms’ in five passages outside the Ta liqa-t: (1) Avicenna,
˘
2
30
Metaphysics, c. IX,5, p. 335, line 18 (‘the principles giving forms’); (2) ibid., c. IX,5,
31 p. 337, line 26 (‘When it becomes prepared, it attains the form from the giver of forms’).
32 These two passages appear in the same wording in Avicenna’s Nağa-t (The Salvation). (3)
33 Avicenna, al-Kawn wa-l-fasa-d (On Generation and Corruption), c. 13, p. 187, line 3 (‘the
34 giver of forms’); (4) ibid., c. 14, p. 190, line 14 (‘the giver of forms’); this passage is cited
below, see n. 44. (5) Avicenna, Fı̄ l-af a-l wa-l-infi a-la-t (On Actions and Passions), p. 256,
˘
35
line 10 (‘the giver of forms’). See also n. 7 below for one occurrence in the Da-nešna-me.
36 3 Avicenna, Metaphysics, c. IX,5, p. 335: ‘It follows necessarily, then, that the separate
37 intellects – rather, the last of them, which is close to us, is the one from which there
38 emanates, in participation with the celestial movements, something having the
39 configuration of the forms of the lower world … ’.
4 Avicenna, Metaphysics, c. IX,4, p. 331: ‘This is the state of affairs in each successive
40
intellect and each successive sphere, until it terminates with the active intellect that
41 governs our selves’.
42 5 As I have argued in my Avicenna’s ‘De anima’, pp. 187 – 9.
An exception is Avicenna’s late treatise Ta liqa-t (Notes), where the expression
˘
1
2 appears more than twenty times in various contexts, some of them
epistemological. In the Ta liqa-t, the wa-hib as-suwar supplies substantial forms
˘
3
4
˙˙
in the first place, but also provides first principles of knowledge, the forms of
-
˘
5 the things known (suwar al-ma lūmat), an excellent moral disposition and the
6 actualisation of light.˙ 6 In the inflationary usage of the expression ‘giver of
forms’, the Ta liqa-t resemble a text by a later author: al-Ġaza-lı̄’s Maqa-sid al-
˘
7
8
˙ the
fala-sifa (Intentions of the Philosophers) of the late eleventh century AD. Here
9 expression is used, for instance, in the context of the theory of odours and visual
10 forms.7 It is likely, therefore, that the epistemological interpretation of the
11 expression was developed by Avicenna toward the end of his life and adopted by
12 some of his readers, such as al-Ġaza-lı̄. When the scholastics refer to the dator
13 formarum, they do this in the context of theories of substantial forms and not of
14 intelligible forms (with very few exceptions).8 In modern literature, however,
15 Avicenna’s concept is often misrepresented as epistemological.9
16 Around 1160 in Toledo, Dominicus Gundisalvi translated the metaphysics
part of aš-Šifa- into Latin under the title Liber de philosophia prima sive scientia
˘
17
18 divina. Among its first Latin readers in the twelfth and early thirteenth century,
19 there are some who adopt central doctrines of Avicenna’s emanation system:
20 Gundisalvi himself in his treatise De processione mundi and the anonymous
21 author of The Book of First and Second Causes (Liber de causis primis et
22 secundis). 10 But the great majority of the later scholastic tradition considers
23 Avicenna’s emanation theory to be in conflict with the idea that the world is
24 created. This creation is not a necessary process, it is argued, but depends upon
25
26
27
28 6 As shown by Janssens, The Notions, pp. 551 – 62, esp. pp. 554 – 7.
29 7 al-Ġaza-lı̄, Maqa-sid, p. 350, line 17; p. 352, line 14; p. 359, line 5; p. 369, line 12. One
might suspect that ˙ the Maqa-sid reflect Avicenna’s original usage of the term, since
30 ˙
Avicenna’s Danešna-me-ye Ala-ı̄ (Philosophy
- for Ala--al-Dawla) is the ultimate source of the
˘
˘
˘
31 -
Maqasid (see Janssens, Le Dānesh-Nāmeh, pp. 163 – 77). But, in fact, only one of the
32 eight ˙occurrences of the term ‘giver of forms’ in al-Ġaza-lı̄’s text has a parallel in
33 Avicenna’s Da-nešna-me (see Janssens, The Notions, p. 552; the Persian expression is: sūra
dinanda). ˙
34
8 Hasse, Avicenna’s ‘De anima’, p. 189, n. 620. Possible exceptions are the following:
35
Thomas Aquinas, De veritate, q. 11 a. 1, p. 349 (‘formas omnes sensibiles esse ab agente
36 extrinseco quod est substantia vel forma separata, quam appellant datorem formarum vel
37 intelligentiam agentem’) and Anonymous (Van Steenberghen), Quaestiones de anima,
38 2.19, p. 228, line 47 (‘… et datricem intelligibilium et naturalium quam dixit [sc.
39 Avicenna] motricem decimi orbis’).
9 Examples are: Weisheipl, Aristotle’s Concept, p. 150: ‘to be receptive of new concepts
40
from the dator formarum, the “agent intellect”’; Dales, The Problem of the Rational Soul,
41 p. 8: ‘intelligible objects provided by the Giver of Forms’.
42 10 Anonymous (de Vaux), Liber de causis primis et secundis.
Avicenna’s ‘Giver of Forms’ in Latin Philosophy 227
1 the will of God, and it is not dependent upon intermediaries such as angels and
2 intelligences, and hence not upon a giver of forms.11
3 It is remarkable that, in spite of this, the dator formarum is often mentioned
4 in Latin sources, well into the seventeenth century. The theory, which is usually
5 attributed to Plato or Avicenna, was obviously thought to be important – so
6 important that it could not be passed over in silence. I suspect that the reception
7 was not so entirely negative as it appears. I have therefore been searching for
8 authors and passages with a positive reaction to Avicenna’s theory of the giver of
9 forms – indications that the theory was thought to be a strong theory, even if it
10 was refuted. I start with a brief overview of the Latin fortuna of the concept and
11 then discuss the rare positive reactions to it, four briefly – those of William of
12 Auvergne, John Buridan, Marsilio Ficino and Tiberio Russiliano – and one at
13 length: that of Albertus Magnus.
14
15
16 I The Latin fortuna of the Giver of Forms
17
18 The Avicennian theory of the giver of forms never firmly set foot on Latin soil.
19 This contrasts with Avicenna’s theory that the active intellect is a separate
20 substance, which was adopted by a good number of authors, especially in the
21 thirteenth century. Some of them identified this separate substance with God,
22 thus forming what tienne Gilson has called the position of ‘Augustinisme
23 avicennisant’. These authors combine Avicenna’s teaching of ‘abstractions
24 emanating from the active intellect’ (De anima V,5) with Augustine’s theory of
25 illuminatio. Early exponents of this current are Jean de la Rochelle, the Summa
26 fratris Alexandri and Vincent of Beauvais; later in the thirteenth century, the
27 active intellect was identified with God by Roger Bacon, John Pecham, Roger
28 Marston, Vital du Four, and also Henry of Ghent (though only in parts of his
29 work).12 As far as I can see, the epistemological current of ‘Augustinisme
30 avicennisant’ did not have a parallel in ontology. The two Avicennian concepts,
31 that is, the active intellect as the source of intelligible forms, and the giver of
32 forms as the source of substantial forms, saw a very different Latin reception. It
33 is remarkable that the dator formarum concept was unsuccessful even within the
34 Franciscan tradition that favoured Avicennian epistemology.
35
36 11 See, for example, Thomas Aquinas, Summa contra gentiles, c. II.26, vol. 13, p. 332: ‘Per
37 haec autem exluditur quorundam philosophorum positio dicentium quod ex hoc quod
38 deus seipsum intelligit, fluit ab ipso de necessitate talis rerum dispositio: quasi non suo
39 arbitrio limitet singula et universa disponat, sicut fides catholica profitetur.’ Ibid.,
c. II.42, vol. 13, p. 365: ‘Excluditur autem ex praedictis opinio Avicennae, qui dicit
40
quod deus, intelligens se, produxit unam intelligentiam primam … Et sic inde procedens
41 diversitatem rerum causari instituit per causas secundas’.
42 12 Hasse, Avicenna’s ‘De anima’, pp. 203 – 23.
1 The scholastics preferred other explanations for the origin of forms. An early
2 example is provided by a passage from Bonaventure, in the second book of his
3 commentary on the Sentences, dating from about 1248.13 There are four
4 opinions on the coming-to-be of forms (eductio formae in esse), Bonaventure says
5 – and many similar divisions of opinions can be found in later scholastic
6 literature,14 all the way until Francisco Suarez’ Metaphysical Disputations: 15 First,
7 the theory of latitatio or latitudo formarum of Anaxagoras (as presented by
8 Aristotle in Physics, 187a26–b7): the forms are latent in matter and are only
9 made manifest by an agent. Second, the theory of more modern philosophers
10 (philosophorum magis modernorum – here Avicenna is implied) that all forms
11 derive from a creator. The efficient cause of everything is God; the particular
12 causes only prepare matter for the reception of a form.16 Third, the position of
13 Aristotle and of the doctores in philosophia et theologia that the forms are in the
14
potentiality of matter and are made actual by the particular agent. There are two
15
variants of this position, according to Bonaventure: either you say that the form
16
derives from an agent which multiplies its own form, or – and this is opinion
17
four – you say that the form is already in matter before it is actualized.
18
Bonaventure favours this last position, the pre-existence of forms. One
19
advantage of this position, in the eyes of Bonaventure, is that it accords with
20
Augustine’s well-known theory of ‘seminal reasons’ which exist in matter
21
(rationes seminales).
22
23
The scholastic discussion of substantial generation is mainly about the last
24
two alternatives: do the forms preexist in matter somehow, as Bonaventure,
25
Albertus Magnus and others say, or: is the role of matter purely passive, as
26
Thomas Aquinas insists?17 Whether there is a small or a large difference between
27
28 13 Bonaventure, In quatuor libros Sententiarum, lib. II dist. VII p. II a. 2 q. 1, pp. 197 – 8.
29 14 See the references in the editors’ scholion: Bonaventure, ibid., p. 200.
15 Suarez, Disputationes metaphysicae, disp. XV sect. II, pp. 505 – 12, esp. p. 508 (on Plato
30
and Avicenna).
31 16 Bonaventure, ibid., p. 198: ‘Alia fuit positio philosophorum magis modernorum, quod
32 omnes formae sunt a creatore. Et haec positio potest dupliciter intelligi: uno modo quod
33 deus sit principaliter agens et producens in omnis rei eductione, et sic habet veritatem;
34 vel ita quod deus sit tota causa efficiens, et agens particulare non faciat nisi materiam
adaptare, ut sicut producit animam rationalem, ita et alias formas; et iste intellectus
35
videtur fuisse illorum philosophorum. Et iste intellectus est impossibilis, quia agens
36 particulare aut inducit aliquid aut nihil. Et si nihil, ergo nihil agit. Si aliquid inducit,
37 ergo videtur quod aliquam efficiat dispositionem; sed qua ratione potest in unam et in
38 aliam? Quare ista positio non est rationabilis.’
39 17 See, e. g., Thomas Aquinas, De veritate, q. 11 a. 1, p. 350: ‘Et ideo secundum doctrinam
Aristotilis via media inter has duas tenenda est in omnibus praedictis: formae enim
40
naturales praeexistunt quidem in materia, non in actu, ut alii dicebant, sed in potentia
41 solum de qua in actum reducuntur per agens extrinsecum proximum, non solum per
42 agens primum, ut alia opinio ponebat’.
1 the position of Albertus and Thomas is a matter of dispute in modern

2 scholarship, but that need not concern us here.18 Both sides of the medieval
3 discussion jointly maintained that all substantial forms are educed from matter,
4 with the exception of human souls, which are created by God.
5 This is the general picture of the reception of Avicenna’s theory. Let me
6 point to some distinctive features of it. A noteworthy feature is the association
7 of the giver of forms with the term colcodea (or colcodrea or colchodea).19 This is a
8 Latin term derived from Arabic astrological literature. In many Arabic sources,
9 the length of life is calculated by al-kadhuda-h, a planet with specific attributes
10 ¯
on the birth chart. The source of the˘ association of the two concepts was
11 probably Pietro d’Abano, the early fourteenth-century author of the Conciliator,
12 as Bruno Nardi has shown.20 Pietro explains the astrological term alcocoden with
13 a reference to the giver of forms: quando [scil. Saturn and Mars] fuerint
14 alcocoden, idest datores formarum vitae. 21 In Pietro’s eyes, the alcocoden can be
15 called a ‘giver of forms’ because it is a source of life and a secondary cause (the
16 first and universal cause being the heaven); both are true also of the giver of
17 forms. The term colcodea was presumably derived from alcocoden, but not
18 (according to the present state of our knowledge) by Pietro d’Abano, since a
19 phrase with the term cholcodea in Pietro’s Conciliator is missing in the earliest
20 two Renaissance editions and in at least three manuscripts and thus seems to be
21 a Renaissance addition.22 It is usually assumed that Agostino Nifo coined the
22 term for his commentary on Averroes’ Destructio destructionum of 1495. But in
23 fact Niccolo Tignosi already uses colcodrea several times in 1474 when
24 discussing Avicenna’s giver of forms theory in his De anima commentary, as in
25 the sentence: ‘Avicenna has posited the Colcodrea, that is, an intelligence which
26 is the giver of forms to the things below’ (ipsum posuisse colcodream, id est,
27 intelligentiam datricem formarum istis inferioribus).23 Later, the term also appears
28 in texts by Agostino Nifo,24 Pietro Pomponazzi,25 Marcantonio Zimara,26
29
30
18 On this difference see Suarez, Disputationes metaphysicae, disp. XV sect. II, p. 506;
31 Nardi, La dottrina, pp. 69 – 101, esp. pp. 92 – 8; Weisheipl, The Axiom, pp. 455 – 6.
32 19 On this term see Nallino, La ‘Colcodea’ d’Avicenna, pp. 84 – 91, Wolfson, Colcodea,
33 pp. 573 – 6, and Porro, Colcodea, pp. 2009 – 2010.
34 20 Nardi, Origine, p. 234, n. 1.
21 Pietro d’Abano, Conciliator, diff. 10, propter tertium, fol. 16vb.
35
22 This was again shown by Nardi. See Pietro d’Abano, Conciliator, diff. 71, propter
36 tertium, fol. 108rb (‘quam cholcodeam vocabat’) and Nardi, Origine, p. 235, n. 1.
37 23 Tignosi, In De anima commentarii, p. 384: ‘Ad Avicennam dicatur ipsum posuisse
38 colcodream, id est, intelligentiam datricem formarum istis inferioribus et ipsam irradiare
39 supra intellectum possibilem’.
24 Nifo, In librum Destructio destructionum Averroys, fols 97vb, 98ra (‘dator formarum latine,
40
arabice vero colcodea’), 101ra ; Nifo, In librum de anima, c. II.34.
41 25 Pomponazzi, In XII Metaphysicae, quoted from Nardi, Origine, p. 234n: ‘… et tandem
42 venit ad Colchodeam omnium creatricem formarum istorum inferiorum; quae
1 Tiberio Russiliano27, Julius Caesar Scaliger,28 Daniel Sennert29 and, in the early
2 seventeenth century, Tommaso Campanella.30 It was employed because it was
3 thought to be the original Arabic term. As Agostino Nifo writes: colchodea quam
4 latine dator formarum exponitur: ‘Colcodea, which in Latin is rendered as dator
5 formarum.’31 The term also appears in Hebrew sources of the sixteenth and
6 seventeenth centuries that are influenced by Nifo.32
7 Another remarkable feature is that the Latin reception of Avicenna’s concept
8 was much influenced by Averroes. Averroes, in his Long Commentary on the
9
10
11
12 Colchodea est intellectus agens. … Avicenna autem, quia tenet quod formae
13 substantiales non possunt agere immediate et quia accidentia non possunt agere
14 substantis, ideo oportet ponere datorem formarum. Non enim ipse videbat agens
immediatum formarum, quia non accidens neque substantia; ergo est Colchodea’; and
15
Pomponazzi, Utrum deus concurrat …, quoted from Nardi, Origine, p. 234: ‘… aut in
16 opinionem Avicennae, qui tenuit quod immediate [sc. anima creetur] a Colcodea …’,
17 and p. 237: ‘… non quia creetur anima nostra a deo, ut solvit Scotus, aut hai
18 Colcodheia, ut voluit Avicenna’.
19 26 Zimara, Contradictionum solutiones, fol. 421vb : ‘… quia secundum Avicennam aliae
formae, quae de novo inducuntur in materia, non sunt eductae de potentia materiae, sed
20
sunt ab extrinseco motore, quem datorem formarum appellat seu colcodeam’.
21 27 Russiliano, Apologeticus, disp. 5, p. 177: ‘… cum illa [sc. anima rationalis] secundum
22 Avicennae mentem fuerit infusa ex colcodea omnium formarum generatrice; modo
23 colcodea, dum sufficientem dispositionem in materia habeat, semper formam inducit,
24 vel illa sit ex seminis habita praeparatione vel ex putrefactione’. I discuss Russiliano’s
reception of the giver of forms theory in section II of this article.
25
28 Scaliger, De subtilitate, c. 97, p. 333: ‘Colcodea, nescio quae, ut aiunt, ab Avicenna ficta
26 est, quae formarum conda, et proma, imo vero fabra esset, tuus iste liber, qui etiam
27 dictamo putrido vitam molitur, etiam sesquicolcodea dici mereatur’.
28 29 Sennert, Hypomnemata physcia, lib. IV c. 2, p. 150: ‘Avicennas animas viventium non a
29 parentibus, sed a quadam formarum datrice seu, ut Scaliger, exerc. 97. loquitur,
formarum proma conda intelligentia, quam Colcodeam nominat, provenire statuit …
30
Procul dubio autem istam sententiam ex Platone et Platonicis hausit Avicennas.’ See
31 Hirai, Atomes vivants, pp. 479 – 80.
32 30 Campanella, De homine, c. I.1, p. 14: ‘Avicenna autem Colchodeae hoc munus
33 permandat utenti elementorum materia et qualitatibus: propterea putat omnia animalia
34 et homines posse oriri sponte, sicubi tellus sit apta ad Colchodeae sigillum suscipiendum,
quod philosophi multi, licet animae mundi vel casui hoc opus adscribant, olim et nunc
35
docent’; ibid. c. V.5, p. 70: ‘Verum cum oblivio contingat et scientia deleatur, putavit
36 Avicenna quod, licet ab ideis sit scientia, non tamen, inquit, ab innatis (sic enim nulla
37 fieret oblivio), sed a defluentibus a Colchodea, quae sit ultimus intellectus aut anima
38 mundi secundum alios, in nostram animam, quae a sensibus movetur ut ad illas respiciat
39 ideas’; ibid., c. V.5, p. 78: ‘… per quas excitatur ad species ex Colchodea effluentes
considerandas’.
40
31 Nifo, In librum Destructio destructionum Averroys, fol. 97vb. Cf. n. 24 above.
41 32 See Wolfson, Colcodea, pp. 573 – 6. On the Hebrew reception of the giver of forms
42 theory, see Goldstein, Dator formarum, pp. 107 – 21.
1 Metaphysics, VII.31 and XII.18,33 criticizes Avicenna for holding that all
2 substantial forms derive from the active intellect, ‘which he calls “giver of
3 forms”’.34 Averroes refutes the theory and adds that al-Fārābı̄ and Avicenna are
4 in fundamental agreement with Plato on this issue. As a result, Latin scholastics
5 often attack Plato and not Avicenna for holding the dator formarum theory.35
6 Averroes’ association of the giver of forms with Plato influenced the
7 understanding of Avicenna’s theory in the West.36 Latin knowledge of Plato
8 was, for the most part, confined to the Timaeus. And hence, in the scholastic
9 view, the Platonic standpoint was that the forms are given by the second gods of
10 the Timaeus, as Albertus Magnus puts it: ‘everything is generated by the second
11 gods (a diis secundis), who were given the seed of generation by the god of gods
12 (deus deorum)’37 – or, in the Renaissance interpretation, by the world soul,
13 anima mundi. 38 Hence it came that Plato’s anima mundi and Avicenna’s colcodea
14 were thought to mean the same and that Plato and Avicenna were considered to
15 be the major exponents of a giver of forms theory.
16 That the Platonic association tainted the understanding of Avicenna’s
17 Metaphysics is evident in that the scholastics often use the term creare for the
18 activity of Plato’s and Avicenna’s dator formarum. An example is provided by the
19 passage by Bonaventure which I cited above: the second of the four possible
20 positions on the generation of forms was the theory that all forms derive from a
21
22 33 Averroes, Tafsı̄r, c. VII.31, pp. 881 – 6, c. XII.18, pp. 1496 – 8; Averroes, Commentarium
23 in libros Metaphysicorum, fols 181ra-vb, fols 304ra-va. The latter passage is translated in:
24 Genequand, Ibn Rushd’s Metaphysics, pp. 107 – 9.
34 Averroes, Tafsı̄r, c. VII.31, p. 882, line 19; Averroes, Commentarium in libros
25
Metaphysicorum, fol. 181ra : ‘Et ideo quia Avicenna oboedit istis propositionibus, credidit
26 omnes formas esse ab intelligentia agente, quam vocat datorem formarum’.
27 35 Hasse, Plato Arabico-Latinus, pp. 42 – 5. For the early reception of Averroes’
28 commentary in general see Bertolacci, The Reception, pp. 457 – 80.
29 36 Averroes’ understanding of Plato was influenced by Themistius (d. 388 AD), who had
argued in his paraphrase of Aristotle’s Metaphysics, book XII, that the forms of the living
30
beings are implanted in matter by the gods and that this is Plato’s theory; see Hasse,
31 Spontaneous Generation, pp. 154 and 158 – 9.
32 37 Albertus, Metaphysica, lib. II tr. 1, c. 8, p. 468b: ‘Platonis igitur sententia est omnia fieri
33 a diis secundis, quibus deus deorum dedit sementem generationis. Dii autem secundi
34 sunt stellae et orbes caelestium moventes materiam ad omnium generabilium
productionem. Et illa sementis dicitur forma quaedam formans materiam ad conveniens
35
sibi in nomine. Hanc enim et huiusmodi formam dicit communicari materae per
36 datorem formarum et ipsam materiam aptari formae recipiendae per qualitates activas et
37 passivas’.
38 38 Nifo, Expositiones in libros Metaphysices, lib. 7 disp. 12, p. 201r : ‘Virtus autem generandi
39 est in anima mundi apud Platonem, quae ab Avicenna dicitur cholchodea.’; Campanella,
De homine, c. V.5, p. 70: ‘Colchodea, quae est ultimus intellectus aut anima mundi
40
secundum alios’. Cf. the pseudo-Paracelsian Apocalpysis Hermetis of ca. 1560, as quoted
41 by Jantz, Goethe’s Faust, p. 176: ‘Dieser Geist [sc. the quintessence] wirdt von Avicenna
42 genandt die Seel der Welt’.
1 creator.39 But ‘creation’ does not adequately describe Avicenna’s theory of

emanation. He had redefined the term ‘creation’ (ibda- ) as the permanent
˘
2
3 causation of the existence of a thing, in Ila-hiyya-t, chapters VI.240 and VIII.341.
4 There is existence after non-existence, but the posteriority is essential, not
5 temporal. The giver of forms does not create forms, but continuously reacts
6 with the emanation of forms if the material disposition in the sublunar world
7 requires it. When the elemental qualities change and exceed certain limits,
8 argues Avicenna, matter becomes disposed towards a new form, which flows
9 upon matter from the giver of forms: ‘The augmentation and reduction [of the
10 elemental qualities] has two well-defined limits; when they are exceeded, the
11 entire disposition of the matter towards its form is extinguished, and it becomes
12 completely disposed towards a different form. It is characteristic of matter that
13 when it is completely disposed towards a form, that this form flows upon the
14 matter from the giver of forms to matter, and that it receives this form.’42 The
15 Latin term creare fails to capture the necessity and automatism of the process.
16 The Platonic colouring of Avicenna’s theory is obvious in Thomas Aquinas’
17 presentation of it: Plato and Avicenna, says Thomas in De potentia, posit an
18 agens supernaturale, ‘a supernatural agent’, which is able to produce ex nihilo. 43
19 This, however, is in disaccord with Avicenna, who clearly holds that the active
20
21 39 See n. 13 above. Another example is Albertus, Super Dionysium De divinis nominibus, c.
22 4, p. 194b: ‘Quidam enim, ut Plato et Avicenna et plures alii, formas dicunt advernire ab
23 extrinseco, ponentes eas dari a datore, et sic ponebant esse per creationem, non quod non
24 fiant in aliquo subiecto, sed quia non fiunt ex aliquo suae essentiae praeexistente.’
40 Avicenna, Metaphysics, c. VI.2, p. 203: ‘This, then, is the meaning that, for the
25
philosophers, is termed “creation” (ibda-). It is the giving of existence to a thing after
˘
26 absolute nonexistence. For it belongs to the effect in itself to be nonexistent and [then] to
27 be, by its cause, existing. That which belongs in the thing intrinsically is more prior in
28 essence for the mind ([though] non in time) than that which belongs to it from another.
29 Hence, every effect constitutes an existence after non-existence, in terms of essential
posteriority’. Cf. ibid., p. 204: ‘It is good [however] to call everything not coming into
30
˘
existence from a previous matter not “generated” (mutakawwin), but “created” (mubda ).’
31 For context, see Marmura, Efficient causality, p. 184.
32 41 Avicenna, Metaphysics, c. VIII.3, p. 272: ‘This is the meaning of a thing’s being created
˘
33 (mubda ) – that is, attaining existence from another. … Thus, origination from absolute
34 nonexistence, which is creation, becomes false and meaningless [sc. if this posteriority
were temporal]. Rather, the posteriority here is essential posteriority’.
35
42 Avicenna, al-Kawn wa-l-fasa-d, c. 14, p. 190; Avicenna, Liber tertius naturalium de
36 generatione et corruptione, c. 14, p. 139.
37 43 Thomas Aquinas, De potentia, q. 3 a. 8, p. 61: ‘Et quia operatio naturae non potest esse
38 ex nihilo, et per consequens oportet quod sit ex praesuppositione, non operabatur
39 secundum eos natura nisi ex parte materiae disponendo ipsam ad formam. Formam vero,
quam oportet fieri et non praesupponi, oportet esse ex agente qui non praesupponit
40
aliquid, sed potest ex nihilo facere; et hoc es agens supernaturale, quod Plato posuit
41 datorem formarum. Et hoc Avicenna dixit esse intelligentiam ultimam inter substantias
42 separatas’.
1 intellect is part of nature. Like all existents of the sublunar world, it is a possible
2 existent per se and a necessary existent only through something else. Avicenna in
3 this sense downplays very much the difference between the supralunar and the
4 sublunar world – a position which Averroes criticizes with vehemence in the
5 Taha-fut at-Taha-fut, because he prefers to describe the supralunar world, in
6 contrast to the sublunar, as ‘necessary through its substance’.44 The term
7 supernaturale shows that the gist of Avicenna’s theory was lost to Thomas
8 Aquinas, partly because it was understood through the eyes of Averroes.
9
10
11 II Positive Reactions: William of Auvergne, John Buridan,
12 Marsilio Ficino and Tiberio Russiliano
13
14 If Avicenna’s theory never set firm foot in the West, what was the context in
15 which it nevertheless was found attractive? The first context is theories which
16 attribute a greater power of daily creation to God. As was said above, the
17 principal scholastic line was to reserve the generation of souls to God, whereas
18 all other forms are educed from matter. William of Auvergne and John Buridan
19 diverge from the mainstream position in that they extend God’s role to the
20 forms of all animate beings.
21 William of Auvergne, who is writing in the 1230s, is among the first readers
22 of Avicenna’s Metaphysics, after the translator, Dominicus Gundisalvi.45 William
23 adopts from Avicenna the description of God as the necesse esse per se 46 and as
24 that whose existence is its being.47 But William at the same time criticizes the
25 Arabic followers of Aristotle, as he calls them, for denying the freedom of the
26 creator48 and for describing his creative activity as eternal.49 He also rejects the
27
28
29
30
44 Averroes, The Incoherence of the Incoherence, vol. 1, p. 238.
31 45 See also the article by Amos Bertolacci in this volume. On Avicenna’s influence on
32 William’s metaphysics, see the articles collected in Teske, Studies.
33 46 William of Auvergne, De universo, IIa IIae c. 10, p. 853b: ‘Proprium nomen vero seu
34 propria nominatio est quam impossibile est naturaliter praedicari de multitudine, quare
necesse esse per se est propria nominatio ipsius’; id., De trinitate, c. 3, p. 25: ‘Iam igitur
35
incipit nobis elucere ens essentiale esse necesse, aeternum et incorruptibile, non
36 causatum.’
37 47 Willliam of Auvergne, De trinitate, c. 1, p. 17: ‘… ens, cuius essentia est ei esse et cuius
38 essentiam praedicamus cum dicimus “est”’. See Teske, Individuation, p. 77.
39 48 William of Auvergne, De universo, Ia Iae c. 27, pp. 623b-4a. The critique is levelled
against ‘sequaces Aristotelis et qui famosiores fuerunt de gente Arabum in disciplinis
40
Aristotelis’ (ibid., p. 618b).
41 49 Willliam of Auvergne, De trinitate, c. 10, p. 66: ‘… opinati sunt inter philosophantes
42 praecipui, scilicet peripatetici, eiusdem operationes aeternas esse’.
1 idea that the tenth intelligence is the source of the causation of many things and
2 of the human souls in particular.50
3 Despite this criticism, Avicenna’s influence is still felt. Everything is educed
4 into being and falls back into non-being through God, William says, or through
5 an intermediate cause dependent upon God. There is no being in the world
6 which is not from God and not sustained through him51 – which reminds one of
7 the Avicennian ontological theory of causation. God fills the world in the way
8 the light of the sun illuminates the universe. At first sight, William simply
9 appears to continue a Christian tradition holding with the apostle Paul that
10 ‘everything is from him, through him and in him’ (omnia ex ipso, per ipsum et in
11 ipso, Rom. 11:36).52 But, in fact, William attributes to God what Avicenna had
12 claimed for the active intellect: according to William, God reacts upon the
13 preparedness of matter by giving forms fitting to that part of matter. A telling
14
case are animals that are generated spontaneously, that is, without there being
15
any parents, such as worms in decay: These animals are created a virtute
16
omnipotentissima creatoris:
17
18 [The fire which is said to lead to generation] prepares matter by removing from it
19 the dispositions which offer resistance to the generated form and which deter it from
the matter in which they are. But the most generous and virtuous goodness of the
20
creator is ready to immediately give the form (dat formam) which is adequate to the
21 part of matter. And this appears clearly in the generation of animals. Wherever
22 matter is prepared to receive life or soul, the creator immediately infuses it into the
23 matter. There is no room for any idiocy whatsoever to hallucinate or feign that there
24 is some power in cheese or wood or in very solid rock which could infuse or bring
25 life or soul into the aforementioned animals [sc. animals generated without
parents].53
26
27
28 50 See Teske, Individuation, pp. 84 – 5.
29 51 William of Auvergne, De universo, Ia Iae c. 27, p. 624a: ‘Non intellexerunt …
fortitudinem virtutis eius [sc. creatoris] qua attingit a summo universi usque deorsum …
30
omnia continens, tenens et retinens, prout vult et quamdiu vult, alioquin reciderent in
31 non esse, unde educta sunt ab ipso et per ipsum’; id., De trinitate, c. 5, p. 35: ‘Omne
32 igitur ens debet suum esse et omne ens debet se primo enti, cum non sit ens nisi ab ipso
33 et per ipsum, et per hoc manifestum est, quod universum est fluxus et exuberantia esse
34 eius, quod est fons universalis essendi.’ Cf. ibid., p. 45, line 11; p. 47, line 46.
52 Cited in: William of Auvergne, De trinitate, c. 7, p. 48.
35
53 William of Auvergne, De anima, c. V.1, p. 112a-b: ‘[sc. ignis qui dicitur generans ad
36 generationem] materiam praeparat removendo ab ea dispositiones quae repugnant
37 formae generati et prohibent eam a materia in qua sunt. Praesto autem est largissima
38 bonitas ac virtuosissima creatoris quae in materiae parte statim dat formam
39 convenientem illi. Et hoc apparet evidenter in generationibus animalium; ubicumque
enim materia parata est ad recipiendum vitam vel animam, statim eam illi creator
40
infundit. Non enim est qualiscunque desipientia delirare vel fingere virtutem aliquam in
41 caseo vel ligno esse vel rupe durissima quae vitam vel animam praenominatis animalibus
42 infundere valeat vel praestare.’
1 It is apparent that Avicenna’s theory of the giver of forms had several advantages
2 for William: it is part of a system of causation in which the entire universe is
3 understood as permanently dependent upon God; it explains generation with
4 the preparedness of matter for the reception of life; and it offers a solution for
5 the problem of spontaneous generation.54
6 I suspect that these advantages were also seen in the later scholastic
7 tradition. In the fourteenth century, commentaries on Aristotle’s Metaphysics
8 book 7 sometimes discuss a question with the title: ‘Whether because of the
9 generation of inferior substances it is necessary to posit separate substances’.
10 Unfortunately, only a few Metaphysics commentaries have been published, so
11 that it is difficult to spot Avicennian influences. I am aware of two authors
12 addressing the question directly: John of Jandun and John Buridan. John of
13 Jandun (d. 1328) flatly rejects Avicenna’s theory, in the footsteps of Averroes’
14 critique:
15
In view of this, one has to answer to this question in accordance with Aristotle and
16 the Commentator that it is not necessary to posit abstract substances, such as ideas
17 or a giver of forms, for the sake of the generation of inferior beings, and this is
18 shown by four arguments of the Commentator …55
19
John Buridan (d. 1361), in contrast, takes the opposite position:
20
21 One has to answer to this question that the most important reason, it seems, for
concluding that there are separate substances (or at least one separate substance) can
22
be drawn and inferred from the generation of the sense-perceptible substances.56
23
24 The principal argument in support of this conclusion is that spontaneous
25 generation cannot be explained without assuming the existence of separate
26 substances. It is not sufficient to assume that material principles in combination
27 with heavenly bodies are responsible for the generation of the forms of inferior
28 substances. The material principles do not have the degree of perfection which a
29 substantial form has, and hence there must exist an immaterial generating
30 principle (principale generans), which produces the substantial forms. This
31 principle is God57 – and not ideas, as Plato thought.
32
33
34 54 For the history of the spontaneous generation problem in Greek, Arabic and Latin
philosophy, see Hasse, Spontaneous Generation, pp. 150 – 75 (on William pp. 162 – 3).
35
55 John of Jandun, In duodecim libros Metaphysicae, lib. 7 q. 22, p. 101vb : ‘His visis
36 dicendum ad quaestionem secundum intentionem Aristotelis et Commentatoris quod
37 non oportet ponere substantias abstractas, ut ideas vel datorem formarum, propter
38 generationem inferiorum, et hoc probatur quatuor rationibus Commentatoris …’
39 56 Buridan, In Metaphysicen, lib. VII q. 9, fol. 46va : ‘Ad questionem respondendum est
quod sicut mihi videtur ratio maxima ad concludendum substantias separatas vel saltem
40
substantiam potest sumi et argui ex generatione substantiarum sensibilium.’
41 57 Buridan, In Metaphysicen, lib. VII q. 9, fol. 46vb : ‘Illa substantia separata assistit
42 presentialiter et indistanter toti mundo et cuilibet eius parti, et sic erat sufficienter simul
1 The thrust of this argument is that substantial forms, even if they are the
2 forms of inferior substances, cannot be generated by material principles only.
3 Authors of the scholastic mainstream position would reply that Buridan
4 mistakenly thinks that forms are generated – whereas in fact only the compound
5 of form and matter is generated. This argument comes from Aristotle’s
6 Metaphysics Zeta 8 (1033b17 – 18). Since there is no generation of forms, there
7 is no need to posit a giver of forms. All we need are material principles, because
8 inferior substances like animals, plants and stones, always remain within the
9 limits of natural agency, as Thomas Aquinas puts it (De potentia q. 3 a. 11,
10 Summa theologiae Ia q. 118 a.1). Buridan would probably reply that, even in a
11 compound, the formal information has to have an origin that is not material.
12 When turning to the Renaissance, it is difficult to encounter a favourable
13 attitude towards the Avicennian theory of the giver of forms, even within
14
Renaissance Platonism. The dator formarum theory is mentioned regularly, but
15
is usually refuted. Marsilio Ficino (d. 1499) is an exception to this trend. He
16
praises Avicenna as ‘the prince of the Arabic theologians’58 and Avicenna and al-
17
Ġaza-lı̄ as thorough friends of Plato.59 In the Platonic Theology, he refers
18
approvingly to Avicenna’s theory that the substantial forms are imprinted by
19
‘some divine mind’, a producer of forms (formatrix), into properly disposed
20
matter and that likewise the human mind, when properly disposed, turns
21
toward a divine intelligence, by which it is ‘informed’. Ficino’s true interest,
22
however, is not Avicenna’s theory of substantial generation, but of intellectual
23
24
knowledge. The context is epistemological: Ficino seeks support for his claim
25
that we cannot attain intellectual knowledge if we are not ‘informed’ by divine
26
ideas.60
27 cum materia rane, dico simul per indistantiam, ita quod ipsa posset ex illa materia
28 producere formam substantialem rane. Et credo quod illa substantia separata est ipse
29 deus omnipotens’.
58 Ficino, Theologia platonica, c. XV.2, p. 26: ‘Avicenna theologorum arabum princeps’.
30
Avicenna is cited as holding that the mind is a form of the body while also being
31 incorporeal. His hierarchy of intelligences is approvingly referred to a little later (ibid.,
32 p. 40): ‘Praeterea multum probanda videtur distinctio illa platonica in Metaphysicis
33 Avicennae, videlicet in mundo intelligibili procedendum esse ab intelligibili summo ad
34 intellectus multos, tamquam a formatore ad vires inde formabiles.’
59 See next n.
35
60 Ficino, Theologia platonica, c. XII.1, pp. 10 – 11: ‘Huic autem Platonicorum mysterio
36 similis ex quadam parte videtur esse Avicennae Algantelisque opinio. Opinantur enim
37 materiam tum elementalem tum intellectualem sub luna divinae cuidam menti tamquam
38 formatrici subesse, cuius instrumenta sint ad elementalem materiam disponendam
39 formae corporeae, sed ipsa tandem praeparatae materiae substantiales imprimat formas.
Similiter humanam mentem per imagines corporum per sensus phantasiamque acceptas
40
ita saepe disponi, ut in divinam illam intelligentiam se convertat atque ab illa quatenus
41 convertitur eatenus formari quotidie … Sed mittamus Arabes in praesentia, quamvis
42 Platoni satis amicos. Ad Platonica redeamus.’
1 One author, however, shows open sympathies for the ontological side of
2 Avicenna’s theory: Tiberio Russiliano (d. after 1519). Russiliano, in a series of
3 public disputations in 1519, defended a number of provocative philosophical
4 theories: on the value of magical knowledge about Christ, on the eternity of the
5 world, and on the trinity. He barely escaped the inquisitorial prosecution which
6 followed. His fifth disputation discusses phenomena of spontaneous generation,
7 among which he counts the first human being ever – at least, he says, ‘if we
8 discuss the case in purely natural terms’ (cum phisice tantum disputemus) – and
9 Americans, because the human beings on these newly discovered islands cannot
10 have reached them by boat.61 In this context, he defends Avicenna’s theory of the
11 spontaneous generation of human beings as most probable philosophically. It is
12 sensible to assume that spontaneous generation is the result of material mixtures
13 that trigger the deliverance of forms from the first craftsman and creator, which,
14 Russiliano finds, is equivalent to Avicenna’s colcodea:
15
In accordance with Aristotle, we can argue in two ways, either by holding that the
16 rational soul is perishable and mortal … or by holding that it is immortal, and then
17 we will say that just like the Colcodea necessarily creates appropriate forms due to a
18 certain disposition of matter, likewise the first craftsman produces a rational soul, be
19 it in the semen or in decay, as long as both preparations are sufficient for attaining
it.62
20
21 Again, as in Buridan, spontaneous generation is a problem that led Western
22 thinkers to adopt Avicenna’s giver of forms theory.63 The alternative was to say
23 with Averroes and Thomas Aquinas that spontaneous generation is due to the
24 influence of the stars. But then it remains unclear where the formal information
25 comes from that explains the generation of a specific animal.
26
27
28
29
30
31 61 Russiliano, Apologeticus, disp. 5, p. 174: ‘… hominibus noviter in insulis incognitis
32 repertis …’, and p. 175: ‘Unde secundum omnem philosophie semitam cogimur dicere
33 hominem et cuncta animalia ex terra habuisse originem’. Cf. also p. 177: ‘… cuncta
34 animalia tum perfecta tum imperfecta prima generatione ex putrefactione prodiere’.
62 Russiliano, Apologeticus, disp. 5, p. 177: ‘Secundum vero Aristotelis mentem dupliciter
35
dicere possumus vel tenendo illam [sc. animam rationalem] esse caducam et mortalem
36 … vel illam esse immortalem, et sic dicemus quod sicut colcodea necessario habita
37 materiae dispositione creat convenientes formas, sic etiam erit de primo opifice, quod ita
38 indifferenter producit animam rationalem in semine et putrefactione, dummodo ambae
39 praeparationes sint sufficientes ad illam capescendam.’
63 See Hasse, Spontaneous Generation, and, with special reference to the Renaissance:
40
Hasse, Arabic Philosophy and Averroism, pp. 125 – 9. Cf. also Thorndike, History, vol. 5,
41 p. 236, on the mentioning of Avicenna’s datrix formarum in an astrological prediction for
42 1521.
1 III Albertus Magnus: the Origin of the Intellectual Soul

2
3 Albertus Magnus’ attitude towards the giver of forms theory is interesting
4 because it is not clearcut and because over his long and very productive life
5 Albertus comes back to the topic several times.64 He was instrumental in
6 formulating the scholastic mainstream position on the generation of forms: that
7 only the rational soul comes from outside, ab extrinseco. This position is
8 formulated already in his early De homine of 1243,65 but it also appears in later
9 works, e. g., in the commentary on the Metaphysics, which dates ca. 1262.66
10 The formula ab extrinseco derives from Aristotle’s Generation of Animals,
11 chapter II.3: ‘It remains that intellect (nous) alone enters from outside
12 (thurathen) and it alone is divine’ (736b27 – 28). One problem for Albertus is to
13 specify this external cause. In De homine, he says, like many scholastics after
14 him, that the rational soul is created by God.67 But Albertus changes his position
15 on this point, and one can observe that the theological interpretation that God
16 is the origin of the souls is rivalled by an Avicennian interpretation that the
17 origin is the separate active intellect. In De animalibus (dating 1256 – 60),
18 Albertus interprets Aristotle’s term ab extrinseco variously as meaning: ‘from the
19 light of the active intellect’, or: ‘from the light of the first active intellect’, or:
20 ‘from the principle of generation, which is the intellect whose work is the work
21 of nature’.68 Some interpreters conclude that the reference here is to the divine
22
23
24
64 Albertus’ reception of this Avicennian doctrine is discussed in: de Libera, Albert le Grand
25
et le Platonisme, pp. 101 – 2; de Libera, Albert le Grand ou l’Antiplatonisme, pp. 260 –
26 62; de Libera, Mtaphysique et notique, pp. 156 – 68; Anzulewicz, Pseudo-Dionysius,
27 pp. 258 – 64; Hasse, Avicenna’s ‘De anima’, pp. 188 – 9; Hasse, Plato-Arabico-Latinus,
28 pp. 42 – 5; Takahashi, Nature, p. 476. On the textual traces of Avicenna’s Prima
29 philosophia in Albertus Magnus’ Metaphysica see Bertolacci, ‘Subtilius speculando’,
pp. 261 – 339, and id., Le citazioni implicite, pp. 179 – 274.
30
65 See n. 67 below.
31 66 Albertus, Metaphysica, lib. 2 tr. 1 c. 9, p. 473b: ‘Ex his etiam patet, cum omnes formae
32 educantur de potentia ad actum, sicut diximus, et solus intellectus adeptus sit ingrediens
33 ab extrinseco, quod …’.
34 67 Albertus, De homine, c. 3.3, p. 141, line 7: ‘Dicimus quod anima rationalis, hoc est
anima hominis cum omnibus potentiis suis, hoc est vegetabilibus et sensibilibus et
35
rationabilibus, non est in semine sicut in effectivo neque per substantiam, sed creatur a
36 deo et infunditur corpori.’
37 68 Albertus, De animalibus, lib. XVI tr. 1 c. 11, p. 1094, § 63: ‘Propter quod a toto
38 extrinseco materiae spermatis et virtutum eius a luce intellectus qui secundum
39 Anaxagoram et Aristotelem est primum agens in omnibus praeinductis virtutibus, in
conceptum inducitur anima rationalis et intellectualis’; ibid., § 64: ‘Et ideo principium
40
ipsius nichil aliud est nisi lux primi intellectus agentis. Intellectus enim hic purus est et
41 immixtus et impassibilis omnino, sicut ostendimus in libro tertio De anima’; ibid., c. 12,
42 p. 1096, § 67: ‘Sequitur necessario quod ipse solus sit ab extrinseco materiae datus a
1 intellect, which is called ‘active’ in this context.69 But Albertus himself adds that
2 this is the intellect which he has described as purus, immixtus and impassibilis
3 (following Aristotle, 430a17 – 18) in his own De anima, where the active
4 intellect is clearly distinct from the First Cause.70
5 Albertus, in his own De anima, which was written in the same years as De
6 animalibus, that is, in the late 1250s, explicitly and approvingly uses the term
7 dator formarum:
8
Likewise, the claim (of Alexander of Aphrodisias) that the intellectual soul is educed
9 from semen, is completely false. Rather, (the intellectual soul) enters from outside
10 from a giver (dator) into matter, and it is the likeness of the giver of forms
11 (similitudo datoris formarum), which is the first intelligence and unmixed with the
12 body, whereas the power or form of the body is mixed, and hence this (argument)
too is wrong. It is true, however, that (the intellectual soul) is the aim of generation,
13
but this aim is not brought about in matter through the power of primary qualities
14 which transform matter, but an intelligence is giving it when the matter is (properly)
15 disposed through natural principles.71
16
This is one of the very few passages in scholastic literature where the ontological
17
concept of the giver of substantial forms is used approvingly. Note that it is
18
combined with Avicenna’s doctrine of the preparedness of matter: quando
19
materia est diposita per principia naturalia. Earlier in the same treatise, Albertus
20
had formulated the ab extrinseco theory not in Avicennian terms, but with
21
reference to ‘some philosophers’, in fact to the Neoplatonic Liber de causis
22
translated from Arabic, holding that ‘the soul is created by mediation through
23
an intelligence’.72
24
25
26
27
28
29 principio generationis quod materiae non commiscetur, et hoc est intellectus cuius est
opus naturae sicut primo moventis et causantis.’
30
69 Weisheipl, The Axiom, p. 451.
31 70 See n. 68 above.
32 71 Albertus, De anima, lib. 3 tr. 2 c. 4, p. 183: ‘Similiter autem, quod dicit, quod
33 intellectualis anima educatur de semine, falsum est omnino, sed potius ipsa est ingrediens
34 ab extrinseco a datore in materiam, et est similitudo datoris formarum, qui est
intelligentia prima et non commixta corpori, sicut commiscetur virtus vel forma
35
corporis; et ideo hoc etiam est falsum. Verum est tamen, quod ipsa est finis generationis,
36 sed hic finis non efficitur in materia virtute qualitatum primarum transmutantium
37 materiam, sed potius intelligentia dat eum, quando materia est disposita per principia
38 naturalia.’
39 72 Albertus, De anima, lib. 1 tr. 2 c. 13, p. 54: ‘[sc. intellectualis anima] est similitudo
quaedam agentis primi. Propter quod dixerunt philosophi quidam mediante intelligentia
40
animam creari’. Cf. Pseudo-Aristotle, Liber de causis, § 3, p. 166: ‘causa prima creavit
41 esse animae mediante intelligentia’. On this passage in Albertus see de Libera,
42 Mtaphysique et notique, pp. 278 – 81.
1 Some years later, in the Metaphysics commentary (dating 1262 – 63),

2 Albertus arrives at a terminological clarification. He now distinguishes an active
3 intellect in the soul and a separate intellect outside.73 About the latter he writes:
4
The entirely separate intellect is the intellect of the intelligence, of which the human
5 intellect is an image (imago), that surrounds its movable (carrier), which is the
6 human body, just as the separate intelligence surrounds its movable (carrier), which
7 is the sphere. The human intellect is in (the separate intellect), just like the inferior
8 lights are in the superior light, from which they receive both the forms and the
9 movement (formas et motum) by way of influence. This influence of form and
10
movement continues until the first cause, which moves the first, universally
influential sphere and is thoroughly pure light that does not receive anything from
11
any other (light).74
12
13 In this passage, Albertus does not call the origin of the forms a ‘giver of forms’
14 or an ‘active intellect’ anymore. The Avicennian influence is nevertheless
15 apparent. The separate intellect is not the first cause, but an inferior intelligence
16 which belongs to a heavenly sphere. In Avicennian terms, this would be the
17 active intellect, and indeed, the separate intellect gives ‘forms and movement’ to
18 the human intellect.
19 It is true that ‘forms’ here means ‘intelligible forms’, but the term imago
20 indicates that the passage is a comment also on the origin of the intellectual
21 soul: The human intellect is the imago of the separate intelligence, and it exists
22 in it. The term imago is important, because it is a repeated claim of Albertus that
23 the intellectual soul is the likeness (imago or similitudo) of an intelligence – and
24 not the likeness of God, as one might expect. This claim appears in the
25 commentaries on De divinis nominibus, De animalibus and Metaphysics. One
26 such passage, in the commentary on De divinis nominibus, will be discussed
27 below.
28 The similitudo theory shows that Albertus in these works does not adopt
29 either a naturalistic standpoint or the theological view on the origin of the soul.
30 The intellectual soul’s most intimate link is with an inferior intelligence, even in
31 terms of origin. The strength of Avicenna’s position is apparent in these
32
passages: the intelligence is the source of intelligible forms and of movement in
33
the sublunar world. It is the proximate cause of many effects in the sublunar
34
35
73 For further passages in Albertus’ works on the active intellect, within or outside the soul,
36 see Anzulewicz, Entwicklung und Stellung, pp. 198 – 9 and 207 – 8.
37 74 Albertus, Metaphysica, lib. II tr. 1 c. 9, pp. 472 – 3: ‘Intellectus autem omnino separatus
38 est intellectus intelligentiae, cuius intellectus hominis est imago quaedam ambiens suum
39 mobile, quod est corpus hominis, sicut intelligentia separata ambit suum mobile, quod
est sphaera. Et intellectus hominis est in illo, sicut lumina inferiora sunt in lumine
40
superiori, a quo recipiunt et formas et motum per influentiam. Et continuatur
41 huiusmodi influentia formae et motus usque ad causam primam, quae movet sphaeram
42 primam universaliter influentem, quae est penitus lux pura, quae a nulla aliquid recipit.’
1 world. It is reasonable, therefore, not to attribute the origin of the soul to the
2 remote cause, God, but to the more proximate cause, the intelligence. Avicenna
3 thus enables Albertus to move halfway from a theological towards a naturalistic
4 theory of the origin of the rational soul. The full way would have been
5 Alexander of Aphrodisias’ element theory or Aristotle’s ‘man is begotten by man
6 and by the sun’ (Physics, c. II.2, 194b14).75
7
8
9 IV Albertus: the Origin of the Forms of Inferior Substances
10
11 We have seen that Avicenna’s theory of the giver of forms falls on fertile ground
12 with Albertus when it comes to the origin of the intellectual soul. In one work,
13 De anima, Albertus even openly adopts the giver of forms theory. When we turn
14 our attention to the other substantial forms – that is, to those of inferior beings
15 such as animals, plants, minerals etc. – the case seems to be more
16 straightforward, because here Albertus clearly favours the theory of inchoatio
17 formae: everything comes to be out of indeterminate beginnings of its essence,
18
which preexist in matter.76 The origin of form lies in matter’s never-ending
19
desire for successive forms.77 This theory owes much to Averroes.78
20
Nevertheless, Avicenna’s theory appeals to Albertus in three contexts.
21
(1) The first context is the question of whether God could have created
22
things better. In his commentary on the first book of the Sentences (dating
23
ca. 1245), Albertus says that with respect to substantial being the answer is: ‘no’.
24
God could not have given a greater capacitas to the things (that is, greater
25
powers tied to the forms) because the giving of forms is dependent upon the
26
disposition of matter, ut dicit Avicenna. The dator formarum deus fills everything
27
28
with forms according to the disposition (of matter): forms of elements, forms of
29
elementary compounds, forms of plants, animals and human beings, depending
30
upon the degree in which the material mixture reaches a balance. Albertus
31
32 75 Albertus’ theory of the rational soul is indebted to Avicenna also in other ways; see
33 Hasse, The Early Albertus Magnus, pp. 232 – 52.
34 76 To quote a representative passage: Albertus, Metaphysica, lib. 2 tr. 1 c. 8, p. 470:
‘Quartum est quod nihil fit ex nihilo penitus secundum naturam, sed quaecumque fiunt
35
procedunt ex indeterminatis et confusis incohationibus suarum essentiarum, quae indita
36 sunt materiae’. On Albertus’ doctrine see Nardi, La dottrina, pp. 69 – 101; Snyder,
37 Albert, pp. 63 – 82; Takahashi, Nature, pp. 451 – 81.
38 77 Albertus, De generatione et corruptione, I, tr. 1 c. 22, p. 130: ‘[sc. materia] non desiderat
39 formam unam tantum, sed omnem formam successive, cum simul eas habere non possit.
Hoc autem desiderium formae incohatio est in materia, quae educitur de ipsa’.
40
78 Cf. Averroes, Commentarium medium in De generatione, c. 17, pp. 26 – 7: ‘tunc necesse
41 est ut generatio non abscindatur quoniam per successionem formarum super subiectum
42 quod est materia non denudatur illud ex quo generatio fit simpliciter …’.
1 concludes: Et per hoc patet solutio ad totum. 79 This is pure Avicennian

2 philosophy – employed, however, for a theory not of the active intellect, but of
3 the Christian creator God. The advantage of Avicenna’s theory is that it offers an
4 explanation of the diversity and variety of sublunar nature. The variety is the
5 result of the combination of forms with corresponding material mixtures.
6 Some years later, in the commentary on Pseudo-Dionysius’ De divinis
7 nominibus (dating 1250), Albertus has changed his attitude. He now rejects the
8 Avicennian (and, in his view, Platonic) principle that all forms are given
9 according to the preparedness of matter, because the position cannot account for
10 the origin of matter. What proceeds from the first cause, is only forms. Hence,
11 one would have to posit an eternal matter, but this is against faith.80
12 (2) However, the giver of forms theory is still present in De divinis
13 nominibus, albeit in a different context: Albertus extends the above-mentioned
14
similitudo theory to all forms, not only to the intellectual soul. All forms arise
15
from matter when the light of the intelligence meets matter and turns it into
16
actuality. The form is a likeness of the first cause:
17
18 It has to be said that, according to Avicenna and to those who posit the giver of
19 forms, the form of each thing is nothing else but the ray of the intelligence, or of the
first cause … We, in contrast, maintain in a manner more appropriate for theology
20
and philosophy according to the opinion of Aristotle that all forms are educed from
21 the potentiality of matter. Hence, the form, in virtue of its essence, is not the
22 embodied light of the first cause. However, since all actions of those things which
23 are composed of a mover and a moved, have in them the power of the mover and
24 the moved, as is obvious in the action of the natural heat which digests that which is
25 not turned into ashes, but moves towards the form of flesh according to the power
of the soul, likewise in the action of the heaven, insofar as it educes forms from the
26
potentiality of matter, as it is said that ‘man begets man and the sun’, there is the
27 power of the first mover, towards whose likeness matter rises through being turned
28 into actuality, as far as it can. Hence, the form is not the embodied light of the first
29 cause, but its likeness caused by it (similitudo eius causata ab ipsa). And it is in this
30 way that one should understand what Dionysius says, that the divine ray appears
31
32
33 79 Albertus, In I Sententiarum, dist. 44 B a. 2, p. 392: ‘Et sic est etiam de datore formarum
34 deo quod omnia implet esse substantiali secundum eorum capacitatem, quaedam forma
elementi, quaedam forma mixti, quaedam autem anima vegetabili et quaedam sensibili et
35
quaedam rationali, secundum quod magis recedunt ab actu contrarietatis ad temper-
36 amentum et uniformitatem complexionis, quia in hoc accedunt ad naturam coeli, ut ipse
37 [sc. Avicenna] dicit. Et per hoc patet solutio ad totum.’
38 80 Albertus, De divinis nominibus, c. 2, p. 73: ‘Et hoc est quod Plato dixit quod formae
39 omnes dantur a datore secundum meritum materiae … Secundum hoc autem … non
inveniretur modus quo procedert materia ab ipso; unde oporteret ponere materiam
40
aeternam, quod est contra fidem. Et ideo sequimur opinionem Aristotelis, quae magis
41 videtur catholica’. This passage is discussed in Anzulewicz, Pseudo-Dionysius, pp. 258 –
42 64.
1 above all beings, not as embodied, but so that each thing rises towards its (sc. the
2 divine ray’s) likeness, as far as it can.81
3 Remember that Albertus, in the Metaphysics commentary, has described the
4 human soul as the imago of an intelligence associated with a heavenly sphere. In
5
the present passage, the likeness is not with an inferior intelligence, but with the
6
divine cause. Radius divinus is the term Albertus adopts from Pseudo-Dionysius
7
Areopagita; he likens it to Avicenna’s ray of the intelligence and to the power of
8
the first mover in the heaven. In Albertus’ view, it is not enough to say with
9
Aristotle that the forms of inferior substances arise from the potentiality of
10
matter; they are also caused by an immaterial supralunar cause, whose likeness
11
they are.82 Hence, it is true that Albertus rejects Avicenna’s giver of forms theory
12
and the idea that forms are ‘embodied light’ (lux incorporata), but he follows
13
Avicenna’s basic assumption that the origin of forms has both sublunar and
14
supralunar causes.
15
16
(3) The third context is causation theory. In his Physics commentary (dating
17
1251 – 57), Albertus contrasts Avicenna’s giver of forms theory with the position
18
of ‘most Peripatetics’ that the forms are educed from matter. ‘This is the position
19
of the two parties’, Albertus says, ‘and everybody may choose as he likes. We,
20 however, say, as it appears to us, that both opinions are true in some way.’83
21 Albertus argues as follows: In all things moved by the first cause, there is one
22 essence (essentia), but the being (esse) is manifold. Avicenna maintains that in the
23
24 81 Albertus, De divinis nominibus, c. 1, p. 15: ‘Solutio: Dicendum quod secundum
Avicennam et ponentes datorem formarum forma uniuscuiusque nihil aliud est quam
25
radius intelligentiae sive causae primae … Nos autem aliter dicimus convenientius
26 theologiae et philosophiae secundum opinionem Aristotelis, quod formae omnes
27 educuntur de potentia materiae. Unde forma per suam essentiam non est lux primae
28 causae incorporata. Sed cum omnis actio eius quod compositum est ex motore et moto,
29 habeat se in virtutem motoris et moti, sicut patet in actione caloris naturalis digerentis
quod non incineratur, sed agit ad formam carnis secundum virtutem animae, ita in
30
actione caeli, secundum quam educit formas de potentia materiae, sicut dicitur, quod
31 “homo generat hominem et sol”, est virtus primi motoris, in cuius similitudinem
32 consurgit materia per reductionem in actum, quantum potest. Unde forma non est lux
33 primae causae incorporata, sed similitudo eius causata ab ipsa. Et sic est intelligendum,
34 quod dicit Dionysius, quod superapparet radius divinus omnibus existentibus, non
tamquam incorporetur eis, sed ut in cuius similitudinem consurgit unaquaeque res,
35
quantum potest.’
36 82 Compare the parallel passage in: Albertus, De ecclesiastica hierarchia, c. 5, p. 123:
37 ‘Quicquid autem sit verum de hoc [sc. Avicenna’s theory], in hoc tamen est simile, quod
38 similiter divinus radius, secundum quod se tenet ex parte infundentis, supereminet et
39 manet in sua simplicitate, diversificatur autem secundum quod recipitur in diversis
dissimiliter proportionatis ad ipsum’.
40
83 Albertus, Physica, lib. II tr. 2 c. 3, p. 103, line 29: ‘Ecce, haec est sententia utrarumque
41 opinionum, et eligat unusquisque quod vult. Nos autem dicimus, prout nobis videtur,
42 quod utraque istarum opinionum vera est secundum aliquem modum.’
1 universe there must be one being which, in virtue of its essence, is the efficient
2 cause of everything.84 It is the essence of the first being which exerts all efficient
3 causality in the world, just like warm objects and lucid objects are such only
4 because the essence of heat warms in them and the essence of light shines in
5 them.85
6
Hence we want to say that there is one essence through which the first mover and all
7 subsequent things are moving … Therefore, with respect to the essence through
8 which the first mover moves, Avicenna is right, because in this way the first mover
9 alone educes from potentiality to actuality and perfects matter.86
10
But the other Peripatetics are right too in saying that the essence has manifold
11
existence (diversum habet esse). 87 And, in fact, it is closer to the truth (verius) to
12
say that matter is perfected by natural causes, and not by divine causes.88 It is
13
typical of Albertus that he tries to harmonize Aristotle and the Arabic
14
philosophers. But it remains astonishing that he does so in this question, since
15
the scholastics usually treat the two alternatives – giver of forms and material
16
principles – as antagonistic. Albertus shows his preferences for the proximate
17
cause, which is the material principles, rather than the first cause, and he does
18
not adopt the giver of forms theory. But he is convinced that a theory of
19
generation has to integrate a first essence which is the cause of all efficient
20
causality in the world.
21
22
In these three contexts, Albertus qualifies his often repeated theory of inchoatio
23
formae: that everything is generated from preexisting formal information in
24
matter. His qualification is due to Avicennian influence and may be summed up
25
as follows: The generation of forms also depends upon a higher immaterial
26
cause. The relation between the forms and the higher cause (a relation called
27
similitudo by Albertus) is of varying degrees due to the different degrees of
28
preparedness of matter. The natural causes, which determine the generation
29
process, would not work without an essence that is the principle of all efficient
30
causality. The essence of things receives greater variety with increasing distance
31
from the first cause, as a result of the diversity of matter.
32
33
34
35
36 84 Albertus, ibid., p. 102, lines 43 – 51.
37 85 Albertus, ibid., p. 102, lines 51 – 6.
38 86 Albertus, ibid., p. 103, lines 67 – 76: ‘Sic ergo intendimus dicere quod una est essentia
39 qua movet primum movens et omnia consequenter moventia … Ed ideo quantum ad
essentiam qua movet primum movens, verum dicit Avicenna, quia sic solum primum
40
movens educit de potentia ad actum et perficit materiam’.
41 87 Albertus, ibid., p. 103, line 69.
42 88 Albertus, ibid., p. 103, line 83.
1 In view of this, we realize that Albertus did not fully take over the eductio
2 formarum theory from Averroes, as is sometimes claimed.89 It is true that
3 Albertus distances himself from Plato and Avicenna and that he criticizes the
4 dator formarum theory.90 But his position nevertheless remains close to
5 Avicenna.91 From Albertus’ standpoint, matter has its own dynamism and is
6 in need of a supralunar immaterial causality in order to be actualized and to
7 reach the similitudo which is the result of the generation process. Note, however,
8 that Albertus discusses the functions of the giver of forms under the label causa
9 divina, as distinct from causa naturalis, and that this is in disaccord with
10 Avicenna’s concept of a giver of forms, which is not a supernatural entity. The
11 scholastics were not able to, or did not want to, integrate Avicenna’s basic idea
12 that the sublunar and supralunar worlds are united in one system of the
13 causality of existence. It remains remarkable, though, that Albertus, in very few
14
passages, identifies the cause of the human soul not with God, but with a
15
separate intelligence surrounding a heavenly sphere.
16
We are now in a better position to understand why the scholastics bothered
17
to discuss the theory they attributed to Plato and Avicenna, which almost all of
18
them rejected. First, of course, because Plato and Avicenna are famous
19
philosophers. But a second reason appears to be that it seemed, after all, a good
20
philosophical idea to make the lowest intelligence, and not God, responsible for
21
the origin of souls, and to assume that the generation of inferior substances was
22
23
dependent not only on lower but also on higher causes. If the scholastics did not
24
see the advantages themselves, they found them lurking in the much read works
25
of Albertus Magnus, where the attraction of the Avicennian standpoint is clearly
26
felt.
27
28 Bibliography
29
30 Albertus Magnus, De anima, ed. C. Stroick, in Opera omnia … edenda … curavit
31 Institutum Alberti Magni Coloniense, 7,1, Mnster: Aschendorff, 1968.
32
33 89 de Libera, Mtaphysique et notique, p. 167. But I agree with the following (de Libera,
34 Albert le Grand et le Platonisme, p. 102): ‘Tout en rejetant le Dator formarum, le
Colonais [sc. Albertus Magnus] n’a aucune raison de rejeter l’intgralit d’une doctrine –
35
mÞme imparfaite – de la cration; une doctrine qui, faisant place un Dieu des dieux et
36 une influence des toiles et des sphres, laisse ouverte la possibilit d’une thologie des
37 Intelligences qui, on le sait, reste, ses yeux, la fine pointe du pripattisme. Une fois
38 amend par l’eductio formarum le platonisme est une philosophie peu prs viable’.
39 90 Especially in: Albertus, De divinis nominibus, c. 2, p. 73 (n. 81 above), and id.,
Metaphysica, lib. II tr. 1 c. 8, pp. 468 – 71.
40
91 As was also pointed out by de Libera (as in n. 89 above) and Takahashi, Nature, p. 476:
41 ‘… it [sc. the giver of forms theory] is evidently not so far removed from his own theory
42 of the generative role of the prime intellect of the universe’.
1 –– , De generatione et corruptione, ed. P. Hossfeld, in Opera omnia … edenda … curavit

2 Institutum Alberti Magni Coloniense, 5,2, Mnster: Aschendorff, 1980.
3
–– , De homine, ed. H. Anzulewicz and J.R. Sçder, in Opera omnia … edenda … curavit
Institutum Alberti Magni Coloniense, 27,2, Mnster: Aschendorff, 2008.
4
–– , Commentarii in primum Sententiarum (dist. XXVI–XLVIII), ed. A. Borgnet, in Opera
5 omnia, vol. 26, Paris: Vivs, 1893.
6 –– , Metaphysica, ed. B. Geyer, in Opera omnia … edenda … curavit Institutum Alberti
7 Magni Coloniense, 16, Mnster: Aschendorff, 1964.
8 –– , Physica, ed. P. Hossfeld, in Opera omnia … edenda … curavit Institutum Alberti
9
Magni Coloniense, 4, 2 vols, Mnster: Aschendorff, 1987/1993.
–– , Super Dionysium De divinis nominibus, ed. P. Simon, in Opera omnia … edenda …
10
curavit Institutum Alberti Magni Coloniense, 37,1, Mnster: Aschendorff, 1972.
11 –– , Super Dionysium De ecclesiastica hierarchia, ed. M. Burger, in Opera omnia …
12 edenda … curavit Institutum Alberti Magni Coloniense, 36, 2, Mnster: Aschendorff,
13 1999.
14 Anonymous (Van Steenberghen), Quaestiones de anima, ed. F. Van Steenberghen, in
15
B. Bazn, M. Giele and F. Van Steenberghen, eds, Trois commentaires anonymes sur le
trait de l’me d’Aristote, Louvain: Publications universitaires, 1971, pp. 123 – 348.
16
Anonymous (de Vaux), De causis primis et secundis et de fluxu qui consequitur eas, ed. R.
17 de Vaux, (Notes et textes sur l’avicennisme latin), Paris: Vrin, 1934, pp. 88 – 140.
18 H. Anzulewicz, Pseudo-Dionysius Areopagita und das Strukturprinzip des Denkens von
19 Albert dem Grossen, in T. Boiadjiev et al., eds, Die Dionysius-Rezeption im
20 Mittelalter, Turnhout: Brepols, 2000, pp. 251 – 95.
21
–– , Entwicklung und Stellung der Intellekttheorie im System des Albertus Magnus,
Archives d’histoire doctrinale et littraire du moyen ge, 70, 2003, pp. 165 – 218.
22
Pseudo-Aristotle, Liber de causis, ed. O. Bardenhewer, Freiburg im Breisgau: Herder,
23 1882, reprinted Frankfurt a.M.: Minerva, 1961.
Averroes, Tafsı̄r maba da at-tabı̄ a, ed. M. Bouyges, 3 vols, Beirut: Imprimerie
˘
24
25 Catholique, 1938 – 48. ˙ ˙
26 –– , Commentarium medium in Aristotelis De generatione et corruptione libros, eds F.H.
27
Fobes and S. Kurland, Cambridge, Massachusetts: The Mediaeval Academy of
America, 1956.
28 –– , Commentarium in libros Metaphysicorum, in Aristotle/Averroes, Aristotelis Stagirite
29 omnia quae extant opera … Averrois … commentarii aliique ipsius in logica,
30 philosophia et medicina libri, Venice: Giunta, 1562, reprinted Frankfurt a.M.:
31 Minerva, 1962, vol. 8.
32 –– , Tahafut al-tahafut (The Incoherence of the Incoherence), transl. and ed. S. Van den
33
Bergh, 2 vols, London: Luzac, 1954.
Avicenna, The Metaphysics of ‘The Healing’, transl. and ed. M.E. Marmura, Provo, Utah:
–– , aš-Šifa-, at-Tabı̄ iyya-t, al-Kawn wa-l-fasa-d (On Generation and Corruption), ed.
˘
35
˘
M. Qassem,˙ ˙Cairo: al-Hay a al-Misriyya al- Āmma, 1969.

˘
˘ ˘ ˘
36
–– , aš-Šifa- , at-Tabı̄ iyya-t, Fı̄ l-af a-l ˙ wa-l-infi a-la-t (On Actions and Passions), ed.
˘
˘
˘
37
˙
M. Qassem,˙Cairo: al-Hay a al-Misriyya al- Āmma, 1969.
˘
38 ˙
–– , Liber tertius naturalium de generatione et corruptione, ed. S. van Riet, Louvain-la-
39
Neuve/Leiden: Brill, 1987.
40 A. Bertolacci, ‘Subtilius speculando’. Le citazioni della Philosophia Prima di Avicenna nel
41 Commento alla Metafisica di Alberto Magno, Documenti e Studi sulla Tradizione
42 Filosofica Medievale, 9, 1998, pp. 261 – 339.
1 –– , Le citazioni implicite testuali della Philosophia prima di Avicenna nel commento

2 alla Metafisica di Alberto Magno: analisi tipologica, Documenti e Studi sulla
3
Tradizione Filosofica Medievale, 12, 2001, pp. 179 – 274.
–– , The Reception of Averroes’ Long Commentary on the Metaphysics in Latin Medieval
4 Philosophy until Albertus Magnus, in L. Honnefelder et al., eds, Via Alberti. Texte –
5 Quellen – Interpretationen, Mnster: Aschendorff, 2009, pp. 457 – 80.
6 Bonaventure, Commentaria in quatuor libros Sententiarum, in Doctoris Seraphici S.
7 Bonaventurae … opera omnia, 4 vols, Quaracchi: Typogr. Coll. S. Bonaventurae,
8
1882 – 89.
Campanella, Tommaso, De homine (Theologicorum liber IV), ed. R. Amerio, 2 vols,
9 Rome: Centro Internaz. di Studi Umanistici, 1960 – 61.
10 R.C. Dales, The Problem of the Rational Soul in the Thirteenth Century, Leiden/New
11 York/Cologne: Brill, 1995.
12 Dominicus Gundisalvi, De processione mundi, ed. G. Blow (Beitrge zur Geschichte der
13 Philosophie und Theologie des Mittelalters, Band 24, Heft 3), Mnster: Aschen-
dorff, 1925.
14
Ficino, Marsilio, Theologia platonica (Platonic Theology), transls and eds M. J. B. Allen,
15 J. Hankins, 6 vols, Cambridge, Massachusetts: Harvard University Press, 2001 – 06.
al-Ġaza-lı̄, Maqa-sid al-fala-sifa, ed. S. Dunya-, Cairo: Da-r al-Ma a-rif, 1960.
˘
16
17 Ch. Genequand,˙ Ibn Rushd’s Metaphysics: A Translation with Introduction of Ibn Rushd’s
18 Commentary on Aristotle’s Metaphysics, Book Lām, Leiden: Brill, 1986.
H. Goldstein, Dator formarum: Ibn Ruschd, Levi ben Gerson, and Moses ben Joshua of
19
Narbonne, in I. al-Faruqi, ed., Islamic Thought and Culture, Washington DC:
20 International Institute of Islamic Thought, 1982, pp. 107 – 21.
21 D.N. Hasse, Avicenna’s ‘De anima’ in the Latin West: The Formation of a Peripatetic
–– , Plato arabico-latinus: Philosophy-Wisdom Literature-Occult Sciences, in S. Gersh
24
and M.J.F.M. Hoenen, eds, The Platonic Tradition in the Middle Ages: A
25 Doxographic Approach, Berlin/New York: deGruyter, 2002, pp. 31 – 65.
26 –– , Spontaneous Generation and the Ontology of Forms in Greek, Arabic, and
27 Medieval Latin Sources, in P. Adamson, ed., Classical Arabic Philosophy: Sources and
28 Reception, London/Turin: The Warburg Institute/Nino Aragno Editore, 2007,
29
pp. 150 – 75.
–– , Arabic Philosophy and Averroism, in J. Hankins, ed., Cambridge Companion to
30 Renaissance Philosophy, Cambridge: Cambridge University Press, 2007, pp. 113 –
31 36.
32 –– , The Early Albertus Magnus and his Arabic Sources on the Theory of the Soul,
33 Vivarium, 46, 2008, pp. 232 – 52.
34
H. Hirai, Atomes vivants, origine de l’ me et gnration spontane chez Daniel Sennert,
Bruniana & Campelliana, 13, 2007, pp. 477 – 95.
35 J. Janssens, Le Dānesh-Nāmeh d’Ibn Sı̄nā: un texte revoir?’, Bulletin de philosophie
36 mdivale, 28, 1986, pp. 163 – 77, reprinted in id., Ibn Sı̄nā and his Influence on the
37 Arabic and Latin World, Ashgate: Ashgate Variorum, 2006, art. VII.
–– , The Notions of Wa-hib Al-Suwar (Giver of Forms) and Wa-hib Al- Aql (Bestower of
˘
38
39 intelligence) in Ibn Sı̄na-,˙ in M.C. Pacheco, J.F. Meirinhos, eds, Intellect et
imagination dans la Philosophie Mdivale, 3 vols, Turnhout: Brepols, 2006, vol. 1,
40 pp. 551 – 62.
41 H.S. Jantz, Goethe’s Faust as a Renaissance Man: Parallels and Prototypes, Princeton:
42 Princeton University Press, 1951.
1 John Buridan, In Metaphysicen Aristotelis questiones argutissimae, Paris: Badius Ascensius,

2 1518, reprinted Frankfurt a.M.: Minerva, 1964 (with wrong date 1588 on the
3
titlepage).
John of Jandun, In duodecim libros Metaphysicae, Venice: Scotus, 1553, reprinted
4 Frankfurt a.M.: Minerva, 1966.
5 A. de Libera, Albert le Grand et le platonisme. De la doctrine des ides la thorie des
6 trois tats de l’universel, in E.P. Bos, ed., On Proclus and His Influence in Medieval
7 Philosophy, Leiden et al.: Brill, 1992, pp. 89 – 119.
8
–– , Albert le Grand ou l’antiplatonisme sans Platon, in M. Dixsaut, ed., Contre Platon,
2 vols, Paris: Vrin, 1993, vol. 1, pp. 247 – 71.
9 –– , Mtaphysique et notique: Albert le Grand, Paris: Vrin, 2005
10 M.E. Marmura, The Metaphysics of Efficient Causality in Avicenna (Ibn Sina), in M.E.
11 Marmura, ed., Islamic Theology and Philosophy: Studies in Honor of George F.
12 Hourani, Albany, New York: State University of New York Press, 1984, pp. 172 –
13 87.
C.A. Nallino, La ‘Colcodea’ d’Avicenna e T. Campanella, Giornale critico della filosofia
14
italiana, 6, 1925, pp. 84 – 91.
15 B. Nardi, La dottrina d’Alberto Magno sull’ ‘inchoatio formae’, in id., Studi di filosofia
16 medievale, Rome: Edizioni di Storia e letteratura, 1960, pp. 69 – 101.
17 –– , Origine dell’ anima humana, in id., Studi su Pietro Pomponazzi, Florence: Le
18 Monnier, 1965, pp. 231 – 46.
Nifo, Agostino, In librum Destructio destructionum Averroys commentationes, Venice:
19
Bonatus Locatellus, 1495.
20 –– , In librum De anima Aristotelis et Averoys commentatio, Venice: Petrus de Quarengiis
21 Bergomensis, 1503.
22 –– , Expositiones in libros Metaphysices, Venice: Scotus, 1559, reprinted Frankfurt a.M.:
23 Minerva, 1967.
Pietro d’Abano, Conciliator controversiarum quae inter philosophos et medicos versantur,
24
Venice: Giunta, 1565, reprinted Padua: Antenore, 1985.
25 P. Porro, Colcodea, in V. Melchiorre, ed., Enciclopedia filosofica, Milan: Bompiani,
26 2006, vol. 3, pp. 2009 – 2010.
27 Russiliano, Tiberio, Apologeticus adversus cucullatos. Una reincarnazione di Pico ai tempi
28 di Pomponazzi, ed. P. Zambelli, Milan: Il Polifilo, 1994.
29
Scaliger, Julius Caesar, Exotericarum exercitationum liber XV de subtilitate ad
Hieronymum Cardanum, Frankfurt: Wechel, 1582.
30 Sennert, Daniel, Hypomnemata physica, Frankfurt: Schleich, 1636.
31 S.C. Snyder, Albert the Great, incohatio formae, and the Pure Potentiality of Matter,
32 American Catholic Philosophical Quarterly, 70, 1996, pp. 63 – 82.
33 Suarez, Francisco, Disputationes metaphysicae, in Opera omnia, ed. C. Berton, vols
34
XXV–XXVI, Paris: Vivs, 1866, reprinted Hildesheim: Olms, 2009.
A. Takahashi, Nature, Formative Power and Intellect in the Natural Philosophy of
35 Albert the Great, Early Science and Medicine, 13, 2008, pp. 451 – 81.
36 R.J. Teske, William of Auvergne on the Individuation of Human Souls, Traditio, 49,
37 1994, pp. 77 – 93, reprinted in id., Studies in the Philosophy of William of Auvergne,
38 Bishop of Paris (1228 – 1249), Milwaukee, Wisconsin: Marquette University Press,
39
2006, pp. 121 – 43.
–– , Studies in the Philosophy of William of Auvergne, Bishop of Paris (1228 – 1249),
40 Milwaukee, Wisconsin: Marquette University Press, 2006.
41 Thomas Aquinas, Summa contra gentiles, in Opera omnia iussu Leonis XIII P.M. edita, vol.
42 XIII–XV, Rome: Editori di San Tommaso, 1918 – 30.
1 –– , Quaestiones disputatae de veritate, in Opera omnia iussu Leonis XIII P.M. edita, vol.
2 XXII, Rome: Editori di San Tommaso, 1970 – 76.
3
–– , De potentia, in Quaestiones disputatae, eds R. Spiazzi and P.M. Bazzi, vol. 2, Turin/
Rome: Marietti, 1953.
4 L. Thorndike, A History of Magic and Experimental Science, 8 vols, New York: Columbia
5 University Press, 1923 – 58.
6 Tignosi, Niccol
, In libros Aristotelis de anima commentarii, Florence: Ex Bibliotheca
7 Medicea, 1551.
8
J.A. Weisheipl, The Axiom ‘Opus naturae est opus intelligentiae’ and its Origins, in
G. Meyer OP and A. Zimmermann, eds, Albertus Magnus – Doctor Universalis
9 1280/1980, Mainz: Matthias-Grnewald-Verlag, 1980, pp. 441 – 63.
10 –– , Aristotle’s Concept of Nature: Avicenna and Aquinas, in L.D. Roberts, ed.,
11 Approaches to Nature in the Middle Ages, Binghamton, New York: Center for
12 Medieval and Early Renaissance Studies, 1982, pp. 137 – 60.
13
William of Auvergne, Opera omnia, 2 vols, ed. F. Hotot, with Supplementum, ed. B. Le
Feron, Orlans/Paris: Billaine, 1674, reprinted Frankfurt a.M.: Minerva, 1963
14 (vol. 1: De universo, vol. 2: De anima et al.).
15 –– , De Trinitate, ed. B. Switalski, Toronto: Pontifical Institute of Mediaeval Studies,
16 1976.
17 H.A. Wolfson, Colcodea, 4=79K@9K, in id., Studies in the History of Philosophy and
18
Religion, 2 vols, Cambridge, Massachusetts: Harvard University Press, 1973 – 77,
vol. 2, pp. 573 – 6 (originally: The Jewish Quarterly Review, n.s. 36, 1945, pp. 179 –
19 82).
20 Zimara, Marcantonio, Contradictionum solutiones in hos Metaphysicorum libros, in
21 Aristotle and Averroes, Aristotelis opera cum Averrois commentariis, Venice: Giunta,
22 1562, reprinted Frankfurt a. M.: Minerva, 1962, vol. 8, fols 401 – 24.
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
1
2
3
4 Avicenna and Aquinas on Form and Generation
5
6 Kara Richardson
7
8 Avicenna’s Agent Intellect – the last of a series of productive intellects in his
9 emanationist creation story – is implicated in several controversies in the
10 Medieval Latin West.1 One of these debates has to do with substantial
11 generation, i. e., the coming to be of new corporeal substances, such as plants
12 and animals. According to Aquinas, Avicenna and the Platonists hold that the
13 substantial form needed for the generation of a corporeal substance is produced
14 ex nihilo by a creative agent.2 For Avicenna, this agent is the ‘Giver of Forms’
15 (wāhib al-suwar/dator formarum), i. e., the Agent Intellect.3 (I will refer to this
16 view as the˙ Infusion Model of substantial generation. On this model, when a
17 new compound of form and matter is made, an incorporeal substance produces
18 its form and bestows it upon matter made ready to receive it by corporeal
19 causes.) Aquinas rejects the Infusion Model in favour of a view he attributes to
20 Aristotle: substantial forms are educed from the potentiality of matter by
21 corporeal causes.4 (I will refer to this view as the Eduction Model of substantial
22 generation. On this model, when a new compound is made, its form is
23 produced per accidens in the process of eduction; the compound is produced per
24 se and is the terminus of the process of eduction.) Thus Aquinas claims for
25 himself the territory of Aristotelian naturalism and casts Avicenna’s Agent
26 Intellect in the role of deus ex machina.
27
28
29
1 This paper is developed in part from my doctoral dissertation. I would like to thank my
30
dissertation advisors, Deborah Black and Marleen Rozemond, as well as Martin Pickav,
31 all of the University of Toronto, for very helpful comments on that work. I would also
32 like to thank the participants in the Villa Vigoni conference on the reception of Avi-
33 cenna’s Metaphysics for their engagement with an earlier version of this paper. Finally, I
34 am especially grateful to Dag Nikolaus Hasse and Amos Bertolacci for their comments
and suggestions on the paper in its final form.
35
2 Hasse, Plato arabico-latinus, discusses Aquinas’ association of Plato and Avicenna in De
36 potentia q. 3, a. 8 and elsewhere. See especially pp. 42 – 5.
37 3 Hasse, Avicenna’s ‘De Anima’, p. 188 points out that Avicenna does not explicitly identify
38 the ‘Giver of Forms’ (wāhib al-suwar) and the ‘bestower of intelligibles’ (wāhib al-‘aql),
˙ The Notions, p. 558 also raises this issue, but he
i. e., the Agent Intellect. Janssens,
39
concedes that the roles played by the Giver of Forms indicate that this principle is the
40
Agent Intellect.
41 4 Aquinas attributes this view to Aristotle in De potentia q. 5 a. 1 ad 5: ‘Si autem ponamus
42 formas substantiales educi de potentia materiae, secundum sententiam Aristotelis, …’.
252 Kara Richardson
1 This paper raises several objections to Aquinas’ reckoning of the matter,

2 which has been very influential.5 First, it distorts Avicenna’s view of the unity of
3 a hylemorphic composite substance. Second, it invites an occasionalist reading
4 of Avicenna, which is foreign to his thinking. Third, it conceals Aquinas’
5 agreement with Avicenna that an incorporeal agent causes the existence of the
6 composite through its form, and his subsequent use of Avicennian ideas to
7 reconcile this view with the Eduction Model of generation. Aquinas’ account of
8 Avicenna’s view of substantial generation suggests a very negative reception of
9 Avicenna’s metaphysics. This paper argues that Aquinas’ criticisms of Avicenna
10 are based on a weak reading of Avicenna’s discussions of the causal roles of the
11 Agent Intellect, and shows that Aquinas’ reception of several aspects of these
12 discussions was quite positive.
13
14
15 I Aquinas’ Argument against the Creation of Form in Generation
16
17 Since Aquinas’ most detailed argument against the view of generation he
18 attributes to Avicenna appears in De potentia q. 3 a. 8, I will focus primarily on
19
that text.6 In De potentia q. 3 a. 8, Aquinas treats the question whether God
20
works in nature by creating. ‘Creation’ is a technical term for Aquinas. It refers
21
not to any act of making, such as the builder’s act of making a house out of
22
bricks, but rather to acts of making ex nihilo. Thus the question of De potentia
23
q. 3 a. 8 is not the broad one whether God works in nature at all – which
24
Aquinas answers in the affirmative in De potentia q. 3 a. 7, Summa theologiae 1a
25
q. 105 and Summa contra gentiles lib. 3 c. 67 – but the narrower one whether
26
natural processes, such as the generation of plants and animals, involve the
27
making of something from nothing by God. Aquinas himself denies that God
28
works in nature by creating.
29
30
Of the great many arguments Aquinas recounts against his own view, two
31
are especially instructive for our purposes. The first relies on the following
32
premises: (1) ‘whatever has no material part cannot be made of matter’ and (2)
33
34
35
5 See, for example, Suarez, Disputationes Metaphysicae 18.1 – 2, Leibniz, De ipsa natura,
36 p. 505, Zedler, Saint Thomas, Lee, St. Thomas, Weisheipl, Aristotle’s Concept of
37 Nature, and Brand, Book of Causes. Aquinas’ interpretation of Avicenna has roots in
38 Averroes’ Long Commentary on the Metaphysics, Book Zeta, comm. 31. I discuss Averroes’
39 view of Avicenna below on p. 268. Also relevant are Albert the Great’s arguments against
an Infusion Model of generation he attributes to ‘Platonist’ philosophers. See Snyder,
40
Albert the Great, especially pp. 65 – 71.
41 6 Aquinas also presents this view of Avicenna in Summa contra gentiles lib. 3, c. 69, and in
42 Scriptum super Sententiis lib. 2 d. 1 q. 1 a. 4 ad 4.
Avicenna and Aquinas on Form and Generation 253
1 ‘a form has no material part’.7 (The second premise reflects the Aristotelian view
2 that form is distinguished both from matter and from the composite.) These
3 premises seem to yield the conclusions that substantial forms cannot be made of
4 matter and that if they are made, they must be made ex nihilo. Thus the
5 argument suggests that some natural processes involve creative activity: namely,
6 cases of generation, in which matter acquires a substantial form. It also suggests
7 that the Infusion Model of generation, which Aquinas attributes to Avicenna
8 and the Platonists, is correct.8
9 A second argument threatens the Eduction Model of generation, which
10 Aquinas champions. This argument relies on the principle that ‘nothing is
11 educed from that in which it is not’.9 If this is true, then new forms cannot be
12 educed from matter unless they pre-exist in it. This argument supports a
13 position referred to as Anaxagorean, which holds apparently new forms to be
14 previously latent in matter and construes generation as their uncovering.10
15 Aquinas does not reject the principle behind the Anaxagorean threat to the
16 Eduction Model – sc., that ‘nothing is educed from that in which it is not’ – but
17 he claims that it suffices (in cases of generation, at least) that what is educed
18 exist potentially in that from which it is educed. Thus Aquinas counters
19 Anaxagoras on the ground that he fails to distinguish potentiality and actuality:
20
For he considered it necessary that what is generated pre-exist in act. But it is
21 necessary that it pre-exist in potency and not in act. For if it did not pre-exist in
22 potency, it would come to be out of nothing. And if it did pre-exist in act it would
23 not come to be, since what is does not come to be.11
24
The force of this argument, as I understand it, depends on a prior commitment
25
to the reality of substantial change. If truly new material substances come to be,
26
then either the principle ‘nothing is educed from that in which it is not’ is
27
satisfied by the potential existence of form in matter prior to generation or new
28
forms are produced ex nihilo.
29
30
7 Aquinas, De potentia, q. 3 a. 8 arg. 6: ‘Praeterea, id quod non habet materiam partem
31 sui, non potest ex materia fieri. Sed formae non habent materiam partem sui: quia forma
32 distinguitur et contra materiam et contra compositum, ut patet in principio secundi de
33 Anima. Cum ergo formae fiant quia de novo esse incipiunt, videtur quod non fiant ex
34 materia; et sic fiunt ex nihilo, et per consequens creantur.’ English translations from
Aquinas’ De potentia are my own, but I have benefited from consulting On the Power of
35
God, transl. English Dominican Fathers, 1952.
36 8 Aquinas does not identify the source of the objection at De potentia, q. 3 a. 8 arg. 6.
37 9 Aquinas, De potentia, q. 3 a. 8 arg. 8: ‘nihil educitur de aliquo quod non est in eo’.
38 10 Aquinas, De potentia, q. 3 a. 8 arg. 8 – 9. Aristotle attributes this view to Anaxagoras at
39 Physics, 187a26–b7.
11 Aquinas, De potentia, q. 3 a. 8: ‘putabat enim oportere quod actu praeextiterit [sic] illud
40
quod generatur. Oportet autem quod praeexistat potentia et non actu; si enim non
41 praeexisteret potentia, fieret ex nihilo; si vero praeexisteret actu, non fieret: quia quod
42 est, non fit.’
254 Kara Richardson
1 Aquinas then describes the Infusion Model of generation he attributes to

2 Plato and Avicenna:
3
But since a generated thing is in potency through matter and in act through its
4 form, others held that a thing becomes with respect to form, while its matter pre-
5 exists. And since the work of nature cannot be out of nothing, and thus must be out
6 of something presupposed, nature, according to them, works only on the side of
7 matter by disposing it for form. But form, which must become and is not
8 presupposed, must be from an agent who does not presuppose anything but can
make something out of nothing. And this is a supernatural agent, which Plato
9
posited to be a giver of Forms. And Avicenna called this the lowest intelligence
10 among separated substances. But certain moderns following them consider this to be
11 God.12
12
13
This view employs a distinction between potential and actual being: a
14
hylomorphic composite substance exists potentially prior to generation in the
15
sense that its matter exists prior to generation. On this last point, it diverges
16 from the Eduction Model, which holds that the form of the composite exists
17 potentially in matter prior to generation. Thus on the Infusion Model, the role
18 of nature is not to educe form from matter but rather to dispose matter for
19 form. Form itself is produced ex nihilo by a creative agent, which for Avicenna is
20 the Agent Intellect, but for some of Aquinas’ contemporaries is God.13
21 Aquinas contends that the proponents of the Infusion Model misunderstand
22 form: if a creative principle produces ex nihilo the form of the composite, then
23 this form must exist per se, i. e., through itself or in its own right. (If the form is
24 made in its own right, it must be something in its own right.) Aquinas claims
25 that the form of the composite does not exist per se, but rather it exists because
26 something else – the composite – exists by it. Since the form of the composite
27
28 12 Aquinas, De potentia, q. 3 a. 8: ‘Sed quia res generata est in potentia per materiam, et in
29 actu per suam formam; posuerunt aliqui, quod res fiebat quantum ad formam materia
praeexistente. Et quia operatio naturae non potest esse ex nihilo, et per consequens
30
oportet quod sit ex praesuppositione, non operabatur, secundum eos, natura, nisi ex
31 parte materiae disponendo ipsam ad formam. Formam vero, quam oportet fieri et non
32 praesupponi, oportet esse ex agente qui non praesupponit aliquid, sed potest ex nihilo
33 facere: et hoc est agens supernaturale, quod Plato posuit datorem formarum. Et hoc
34 Avicenna dixit esse intelligentiam ultimam inter substantias separatas. Quidam vero
moderni eos sequentes, dicunt hoc esse Deum.’ While Aquinas says here that Plato
35
posited a dator formarum, he claims in De potentia, q. 5 a. 1 arg. 5 that Plato held that
36 forms in matter are from forms without matter and that Avicenna posited a dator
37 formarum.
38 13 Aquinas attributes this view to ‘moderni’ in De potentia q. 3 a. 8. Hasse, Avicenna’s ‘De
39 Anima’, pp. 203 – 23, argues that the ‘moderni’ Aquinas refers to are Jean de la Rochelle
and Alexander of Hales (or rather: the anonymous author of the Summa fratris
40
Alexandri, book 2), and also identifies William of Auvergne as one of those who take up
41 the view that the agent who produces the form in generation is God. See also Hasse,
42 Plato arabico-latinus, p. 43, and Hasse, Spontaneous Generation, p. 163.
1 does not exist per se, it cannot be produced ex nihilo by a creative agent. In
2 support of his view, Aquinas invokes the following principle: ‘that which is
3 made is said to become according to the way in which it is: because its being is
4 the term of its making; whence that which properly becomes per se is the
5 composite’.14 Since the way things are made must accord with the way they are,
6 the form of the composite ‘is not properly made per se but is that whereby a
7 thing is made, i. e., that through whose acquisition something is said to be
8 made’.15
9 Aquinas’ argument against the Infusion Model of generation relies on the
10 following account of the unity of a hylemorphic substance. The formal and
11 material constituents of the composite are one in the way that a piece of wax
12 and its shape are one.16 Neither the shape of the wax, nor the form of the
13 composite exist per se: rather the shape together with the wax and form together
14 with matter comprise a single per se existent. It follows from this view that we
15 must deny that the shape of the wax can be made per se, and that we must deny
16 that the form of the composite substance can be made per se. 17 Thus, the form
17
18
19 14 Aquinas, De potentia, q. 3 a. 8: ‘Unumquodque autem factum, hoc modo dicitur fieri
quo dicitur esse. Nam esse est terminus factionis: unde illud quod proprie fit per se,
20
compositum est’.
21 15 Aquinas, De potentia, q. 3 a. 8: ‘Forma autem non proprie fit, sed est id quod fit, id est
22 per cuius acquisitionem aliquid dicitur fieri.’ Here I read ‘quo’ instead of the ‘quod’ of
23 the Marietti edition.
24 16 I use Aristotle’s example from De anima. Having determined that the soul is ‘a substance
in the sense of the form of a natural body having life potentially within it’ and that the
25
soul qua substance is the actuality of a body, Aristotle says, ‘That is why we can dismiss as
26 unnecessary the question whether the soul and the body are one: it is as though we were
27 to ask whether the wax and its shape are one, or generally the matter of a thing and that
28 of which it is the matter. Unity has many senses (as many as “is” has), but the proper one
29 is that of actuality’ (Aristotle, De anima 2.1 412a20 – 21; 412b5 – 9).
17 Averroes, in his Long Commentary on the Metaphysics, Book Zeta, comm. 31, makes a
30
similar argument on behalf of Aristotle: ‘The forms are not generated in themselves,
31 because if they were, then the generation would be without the matter of the enmattered
32 thing. Consequently, what is generated is something informed, but if that is so then what
33 generates it is that which moves the matter until it receives the form, that is, that which
34 causes [the form] to emerge from potency to act. Now what moves matter must be either
a body possessing an active quality or a power of a substance that acts through a body
35
possessing an active quality. If what generates the subject of the form were other than
36 what generates its form, then the subject and its form would be actually two things,
37 which is impossible. Thus, the subject does not exist without the form, unless it is said by
38 homonymy. So because the subject of the form has existence only through the form, the
39 agent’s activity is associated with [the subject] only due to [the subject’s] association with
the form. Since the agent’s activity is neither associated with the form alone, nor with the
40
subject without the form, consequently then, the agent’s activity is clearly associated with
41 the subject only on the part of its association with the form. So what generates the form’s
42 subject is what generates the form; in fact, there would be no subject if it were not for
256 Kara Richardson
1 of the composite cannot be produced ex nihilo by a creative agent; it is

2 generated only per accidens through the process of eduction in which the
3 composite is generated per se. Generation does not involve creative acts by God.
4 Aquinas’ claim that proponents of the Infusion Model of generation
5 misunderstand form is somewhat ambiguous. He may mean that they hold the
6 mistaken view that form exists per se. Or he may mean that they hold the
7 mistaken view that the generation of form per se is compatible with an account
8 of hylemorphic unity on which form and matter comprise a single per se
9 existent. In Section II, I will consider Avicenna’s own views in this light.
10
11
12 II Avicenna’s Hylemorphism
13
14 Avicenna discusses form, matter and their relationship in a composite substance
15 in Book 2, chapter one of the Metaphysics of the Šifā’, which focuses on
16 substance. He defines substance as what is not in a subject.18 This definition
17 encompasses the form of the composite and its matter, the composite itself, as
18 well as separate souls and intellects.19 We might assume that since a substance is
19 not in a subject, every substance exists per se. (I.e., we might assume that what is
20
21
22 [the agent’s] generating the form and generating both of them simultaneously. If the
23 subject of the form were to be generated from one agent and the form from another
24 agent, then a single effect insofar as it is one would be generated from two agents, which
is impossible; for it will not be associated with a single act, unless it is an act of a single
25
agent. So one should rely on this [demonstration] in this situation, namely, that on
26 which Aristotle relied’ (Averroes, Commentary on Metaphysics, Zeta 9, p. 334). This
27 argument is discussed in Davidson, Alfarabi, pp. 245 – 50. Earlier in his career, Averroes
28 himself held the view that when a new compound is generated, its form is produced by
29 an incorporeal agent. The merits of Averroes’ later explanation of generation, which does
not appeal to an incorporeal producer of form, are discussed in Freudenthal, The
30
Medieval Astrologization.
31 18 He says that every existent belongs to one of two divisions: ‘One of them is the existent
32 in some other thing, which is a thing having subsistence and species in itself. Its existence
33 isn’t that of a part of this other thing, but it cannot be separated from that other thing.
34 And this is the existent in a subject. And the second is the existent which is not in some
other thing in this way. So it isn’t in a subject at all. And this is substance’ (Avicenna,
35
Metaphysics 2.1, p. 45 (my transl.); Š. Il., p. 57; Pr. ph., p. 65). English translations from
36 the Metaphysics of the Šifā’ are Marmura’s unless otherwise noted.
37 19 He says, ‘Every substance is either a body or not a body. And if it is not a body, then it is
38 either a part of a body or not a part of a body. And if it is not part of a body, then it is
39 separate from bodies entirely. And if it is a part of a body, then it is either form or matter.
And if it is separate, rather than a part of a body, then either it has an administrative
40
relation which is in bodies by means of motion, and is called soul, or it is free from
41 matter in every respect and is called intellect’ (Avicenna, Metaphysics 2.1, p. 48 (my
42 transl.); Š. Il., p. 60; Pr. ph., pp. 68 – 9).
1 not in a subject exists through itself or in its own right.) But Avicenna holds that
2 some substances – namely, form and matter – do not exist per se.
3 Avicenna’s view that form and matter, although substantial, do not exist per
4 se is illustrated in his account of the way form exists in matter. This account
5 relies on the following distinction between a subject (mawdū‘) and a receptacle
6 (mahall): ˙
7
˙
By ‘subject’ is meant that which becomes subsistent in itself and, in terms of being
8 the species, becomes thereafter a cause for something to subsist in it ([but] not as a
9 part of it); and that the receptacle is anything in which something dwells [and
10 which] becomes, by virtue of that [indwelling] thing, [the possessor] of a certain
11 state.20
12 A subject is a subsistent thing of some species, and is the bearer of accidents. A
13 receptacle is merely a repository: something can exist in a receptacle even
14 though that receptacle is not a subsistent thing of some species but rather
15 becomes one through what exists in it.21 This is the way form exists in matter: it
16 exists in a receptacle, which is not a subject; through its reception of form, this
17 receptacle becomes a subsistent thing of some species, i. e., a subject.22 Here the
18 view that matter does not exist per se is quite evident: if matter is not a subsistent
19 thing of some species, it cannot exist per se, i. e., through itself or in its own
20 right. The view that form does not exist per se is less clear, but it is suggested by
21 the claim that form exists in matter as in a receptacle, as well as by two later
22 claims. First, form and matter are conceptually distinct but coexistent; neither
23 can exist apart from the other (Metaphysics 2.4). Second, form is a part (ğuz’) of
24
a body rather than something separate and not part of a body, such as a separate
25
soul or an intellect (Metaphysics 2.1).23 Thus on Avicenna’s view, the form of the
26
composite is a substance (as is its matter), but the composite, not its form or its
27
matter, is a subsistent thing of some species, i. e., a subject (mawdū‘).
28
Indeterminate matter becomes a subsistent thing of some species through˙ its
29
reception of form. With respect to this claim, it is worth noting that Avicenna
30
describes prime matter as that which is ‘in potency, all [material] things’.24
31
This account of form, matter and their relationship suggests that Avicenna
32
denies that form exists per se, and it suggests that he shares with Aquinas the
33
following view of the structure of substantial change: matter first has a form
34
35
20 Avicenna, Metaphysics 2.1, pp. 46 – 7; Š. Il., p. 59; Pr. ph., p. 67.
36 21 Avicenna, Metaphysics 2.1, p. 47; Š. Il., p. 59; Pr. ph., p. 67.
37 22 He says, ‘[a]s for establishing this thing which exists in a receptacle but not in a subject,
38 this is something incumbent on us to show shortly. Once we establish it, [it will be seen]
39 that it is the thing to which, in this place, the name ‘form’ is properly attributed’
(Avicenna, Metaphysics 2.1, p. 47; Š. Il., p. 59; Pr. ph., p. 68). Stone, Simplicius and
40
Avicenna, pp. 77 – 8 discusses Avicenna’s denial that prime matter is a subject.
41 23 Avicenna, Metaphysics 2.1, p. 48; Š. Il., p. 60; Pr. ph., pp. 68 – 9.
258 Kara Richardson
1 potentially and later, through the activity of some agent or agents, it has it
2 actually.25 This account of form, matter and their relationship conflicts with the
3 Infusion Model of generation.26 If Avicenna endorses the Infusion Model, then
4 his hylemorphism conflicts with his view of generation. In Section III, I
5 consider whether Avicenna’s discussions of causal roles played by the incorporeal
6 ‘Giver of Forms’ support the view that he endorses an Infusion Model of
7 generation.27
8
9
10
25 This account of the structure of substantial change is a cornerstone of the Eduction
11 Model of generation: a form is educed from matter in which it pre-exists potentially. A
12 second feature of Aquinas’ Eduction Model is that eduction is the work of a corporeal
13 agent. Aquinas suggests in De potentia q. 5 a. 1 that incorporeal agents cannot move
14 matter and thus that corporeal agents are needed to dispose and transmute matter; he
also equates transmuting matter and educing form from matter. Averroes remarks
15
explicitly that eduction must be the work of a corporeal agent: the generator is ‘that
16 which moves the matter until it receives the form, that is, that which causes [the form] to
17 emerge from potency to act. Now what moves matter must be either a body possessing
18 an active quality or a power of a substance that acts through a body possessing an active
19 quality’ (Averroes, Commentary on Metaphysics, Zeta 9, p. 334).
26 Avicenna’s doctrine of ‘corporeal form’ (sūra ğismiyya) might seem to complicate his
20
views on the unity of the form/matter ˙compound. This doctrine aims to explain a
21 defining feature of body qua body, namely, its receptivity to the three dimensions of
22 length, breadth and depth. Avicenna explains this signal characteristic by positing in
23 body the form of corporeity. Prime matter and this form together comprise body
24 considered in this very general way. But no individual body is a composite of prime
matter and the form of corporeity alone. Individual bodies are members of some species.
25
This might seem to suggest that Avicenna holds that a body has two distinct forms,
26 namely, the form of corporeity, which it shares with all other bodies, and an additional
27 form, which makes it a certain kind of body. I will call this second form the ‘species
28 form’. Hyman, Aristotle’s ‘First Matter’, p. 404 seems to advocate this view, but he notes
29 that it is not clear how Avicenna understood this doctrine of the multiplicity of forms;
see p. 404, n. 85. This interpretation of Avicenna’s doctrine of corporeal form is
30
incompatible with the account of hylemorphic unity I attribute to him based on his
31 discussion of form and matter in Metaphysics 2.1. I reject the view that Avicenna holds
32 that a body has two distinct forms. In my view, Avicenna holds the form of corporeity to
33 be merely conceptually distinct from the species form. So every species form confers on
34 the composite the power to take on the three dimensions of length, breadth and depth,
as well as its specific features. (Zedler, Saint Thomas, pp. 127 – 8 and Stone, Simplicius
35
and Avicenna, p. 100 also take this position. Stone argues: ‘If a substantial form were in
36 another substantial form as its subject, then it would be an accident. But as (according to
37 Avicenna) the same thing can never be both substance and accident, it follows that
38 substantial form is never in any subject whatsoever, but always in prime matter’ Stone,
39 Simplicius and Avicenna, p. 100.)
27 I examine texts from three parts of al-Šifā’: Metaphysics, Physics and Generation and
40
Corruption. I exclude Avicenna’s discussions of the causal roles of the Agent Intellect in
41 his De anima, since these are not directly relevant to the issue of the generation of
42 compounds of form and matter.
1 III Avicenna on the Causal Roles of the Incorporeal ‘Giver of Forms’

2
3 Avicenna refers three times to an incorporeal Giver of Forms in the Metaphysics
4 of the Šifā’. He mentions a separable cause of the existence of form and matter
5 in Metaphysics 2.4, a separable Giver of Forms, which causes the existence of the
6 composite in Metaphysics 6.2 and a separate intellect from which the forms of
7 our lower world emanate in Metaphysics 9.5. In Physics 1.10 of the Šifā’, he says
8 that ‘it seems that that which gives the constitutive form belonging to natural
9 species is extrinsic to natural things’.28 And in his account of elemental
10 transformation in On Generation and Corruption, Avicenna states that ‘it is of
11 the nature of matter when prepared with a complete preparation for a form that
12 it attains this form from the Giver of Forms to matter and accepts it’.29 Taken
13 together, these passages provide an ambiguous view of the division of labour
14 between corporeal agents and the Agent Intellect in generation.
15
In Metaphysics 2.4, where Avicenna aims to establish the ontological priority
16
of form to matter, he also determines the cause of the existence of form and
17
matter. He argues first that matter cannot be the cause of the existence of
18
form.30 He then argues that form alone cannot be the cause of the existence of
19
matter.31 He concludes that ‘the cause of the existence of matter is something
20
conjoined with form, so that the existence of matter emanates only from that
21
thing’.32 The reason the cause must be conjoined with form is that matter and
22
23
24 28 Avicenna, al-Šifā’: al-Samā‘ al-tabı̄‘i [Physics], 1.10, p. 49.
˙
29 Avicenna, al-Šifā’: al-Kawn wa-’l-fasād [Generation and Corruption], c. 14, p. 190.
25
30 Avicenna rejects the claim that matter is the proximate cause of the existence of form for
26 three reasons. He says first that ‘matter is only matter because it has the potency for
27 reception and preparedness. But the prepared insofar as it is prepared is not a cause of
28 existence of what it is prepared for. If it were a cause of existence [of what it is prepared
29 for], then it would have this continuously, without preparedness’ (Avicenna, Metaphysics
2.4, p. 65 (my transl.); Š. Il., p. 83; Pr. ph., p. 96). His second argument relies on the
30
absurdity that ‘the same thing be a cause for something in actuality while it remains in
31 potentiality’ (Avicenna, Metaphysics 2.4, p. 65 (my transl.); Š. Il., p. 83; Pr. ph., p. 96).
32 Matter, as the principle of potentiality in change, cannot actualize itself. Third, ‘if matter
33 were the proximate cause of form, and [if ] matter has no variance in itself, and [if ]
34 whatever was necessitated by that which has no variance has no variance at all in it, it
would then necessarily follow that corporeal forms have no variance in them’ (Avicenna,
35
Metaphysics 2.4, p. 65; Š. Il., p. 84; Pr. ph., p. 97).
36 31 This would be possible were a form never separated from its matter, but is not possible
37 given the corruptibility of the composite: for if ‘this form were alone in itself a cause,
38 matter would cease to exist after [the form] ceases to exist and the commencing form
39 would have another matter that comes to exist through it. That matter would then be
originated, and [this origination] would require another matter’ (Avicenna, Metaphysics
40
2.4, p. 67; Š. Il., p. 85; Pr. ph., p. 98). Avicenna takes this scenario to entail a vicious
41 regress and rejects it.
260 Kara Richardson
1 form cannot exist apart from one another. So form is a necessary condition for
2 the existence of matter. Thus he refers to form as the ‘partner’ of a ‘separate’
3 thing, which causes the existence of matter.33 This separate thing – i. e., the
4 Agent Intellect – is an efficient causal principle of the existence of both form and
5 matter; but it causes the existence of matter through form. The Agent Intellect’s
6 role as cause of existence is also discussed in Book 6 of the Metaphysics of the
7 Šifā’.
8 In Metaphysics 6.2, Avicenna discusses the relative roles of corporeal agents
9 and the Agent Intellect with respect to the existence of corporeal substances.
10 This discussion occurs in the context of an analysis of efficient causality, which
11 aims to correct misconceptions of the efficient cause or agent (al-fā‘il), especially
12 the following one:
13
Someone may think that the agent and the cause are needed only for a thing to have
14
existence after nonexistence and that, once a thing is brought into existence, the
15 thing would [continue to] exist as sufficient unto itself [even] if the cause is no
16 longer present. Thus, someone has thought that a thing is in need of the cause only
17 for its origination, but that, once it is originated and comes to exist, it no longer
18 needs the cause. For such a person, the causes are thus only the causes of origination,
being [temporally] prior [to their effects], not simultaneous with them.34
19
20 This ordinary but mistaken view holds that that role of an efficient cause is to
21 make something, which persists through itself once it is made. Avicenna argues
22 that what is made needs a cause of its existence so long as it exists. This
23 argument establishes a distinction between causes of existence (wuğūd/esse) and
24 causes of origination (hudūt/fieri). The cause of the existence of sublunar things
25
¯
is the separate Giver of˙ Forms, i. e., the Agent Intellect. Avicenna develops this
26 point in Metaphysics 6.2, where he answers two questions raised by the
27 distinction between causes of existence and causes of coming to be: What sort of
28 causal relationship holds between corporeal agents, such as builders and fathers,
29 and the existence of their effects? What is the efficient cause of a thing’s
30 existence post-origination?35
31 Avicenna identifies builders as causes of motion:
32
[a]s for the builder, his movement is the cause of a certain motion. Thereafter, his
33
immobility and refraining from motion, or his ceasing to move and affect
34 transportation after having transported, constitute a cause for the termination of
35 that motion. [Now,] that very act of transporting and the termination of this
36
37 33 Avicenna, Metaphysics 2.4, p. 68; Š. Il., p. 87; Pr. ph., p. 100. Lizzini notes, and I agree,
38 that although Avicenna mentions there only that the cause of the existence of form and
39 matter is a ‘separable principle’, the role of this separable principle indicates that he refers
to the Agent Intellect (Lizzini, The Relation, p. 183).
40
34 Avicenna, Metaphysics 6.1, p. 198; Š. Il., p. 261; Pr. ph., pp. 296 – 7.
41 35 Avicenna’s answers to these questions are discussed in Marmura, The Metaphysics, and
42 Druart, Metaphysics, pp. 327 – 48.
1 motion are a cause of a certain combination, and that combination is a cause for a
2 certain shape taking place; and each of [the things] that constitutes a cause coexists
3
with its effect.36
4 The movements of the builder aim at the form or shape of the building: they
5 terminate in a certain combination of materials, and the combination of
6 materials is the cause of an artifact with a certain shape.37 A father too is
7 identified as a cause of motion: ‘[a]s for the father, he is the cause of the
8 movement of the sperm’, and the (proper) termination of this movement results
9 in a new human being.38
10 According to this analysis, builders and fathers are efficient causes of
11 movements, which temporally precede, and terminate in, the existence of
12 buildings and sons. In terms of the distinction drawn in Metaphysics 6.1 between
13 causes of existence and causes of origination, builders and fathers count as causes
14 of origination: they bring something into being. But such causes are merely
15 accidental or helping causes of existence. This view accords with a principle
16 Avicenna holds true: every cause coexists with its effect.39 He concludes that a
17 correct analysis of the causes of the existence of a thing reflects the simultaneity
18 principle:
19
Thus, the true causes coexist with the effect. As for those that are [temporally] prior,
20 these are causes either accidentally or as helpers. For this reason, it must be believed
21 that the cause of the building’s shape is combination, the cause of [the latter] being
22 the natures of the things being combined and their remaining in the way they are
23 composed, the cause of [these natures] being the separable cause that enacts the
24
natures. The cause of the son is the combination of his form with matter through
the cause that endows forms…We thus find that the causes coexist with [their]
25 effects.40
26
27
28
29
36 Avicenna, Metaphysics 6.2, p. 201; Š. Il., p. 264; Pr. ph., p. 301.
30
37 In Metaphysics 6.4 Avicenna says, ‘[y]ou have known that one and the same thing may, in
31 different respects, be form, purpose, and efficient principle. This is also the case in art.
32 For art is the form of the artifact in the soul. For building is in itself the movement
33 toward the form of the house. This is the principle from which the realization of the
34 form in the matter of the house proceeds’ (Avicenna, Metaphysics 6.4, p. 219; Š. Il.,
p. 283; Pr. ph, p. 325).
35
38 Avicenna, Metaphysics 6.2, p. 201; Š. Il., p. 264; Pr. ph, p. 301.
36 39 The simultaneity principle has Aristotelian origins. Aristotle claims in Physics 2.3 that
37 ‘causes which are actually at work and particular exist and cease to exist simultaneously
38 with their effect, e. g. this healing person with this being-healed person and that
39 housebuilding man with that being-built house; but this is not always true of potential
causes – the house and the housebuilder do not pass away simultaneously’ (Aristotle,
40
Physics 2.3. 195b16 – 21). Avicenna argues for the simultaneity principle in Metaphysics
41 4.1.
262 Kara Richardson
1 The existence of a building is due to its components and to the efficient cause of
2 the natures of its components. The existence of a human being is due to the
3 combination of form and matter and to a separable Giver of Forms, i. e., the
4 Agent Intellect. He then argues that the ‘assisting and preparatory causes’ of the
5 existence of a thing are infinite, one [temporally] preceding the other’, but its
6 ‘essential causes’ are finite and coexist with it.41
7 Since Avicenna’s discussions of the causal roles of the Agent Intellect in
8 Metaphysics 2.4 and 6.2 are concerned with existence, rather than coming to be,
9 they do not provide any clear evidence of his view of substantial generation.
10 That he calls the incorporeal cause of the existence of form, matter and the
11 composite a Giver of Forms is unremarkable given his view of form as that by
12 which indeterminate matter becomes a subsistent thing of some species; this
13 epithet does not indicate that he endorses the Infusion Model of generation.
14 Moreover, Avicenna’s distinctions between causes of origination and causes of
15 existence, and between accidental, assisting or preparatory causes of existence
16 and essential causes of existence, which are a crucial part of his account of the
17 Agent Intellect’s role as the cause of the existence of composite substances, are
18 also a crucial part of Aquinas’ reconciliation of the Eduction Model of
19 generation with the view that God is the cause of the existence of composite
20 substances. This issue is discussed in Section IV.
21 Avicenna’s reference to the Agent Intellect in Physics 1.10 of the Šifā’ is also
22 ambiguous with respect to his model of generation. In that text, Avicenna first
23 defines the efficient cause in natural things as a principle of motion in another,
24 and states that motion is the going out from potentiality to actuality in matter.42
25 He then draws a distinction between the principle of motion considered as a
26 ‘preparer’ (muhayyi’) and the principle of motion considered as a ‘perfecter’
27 (mutammim). The preparer ‘is that which puts the matter in order’, or, in other
28 words disposes the matter.43 Avicenna says that the perfecter ‘is that which gives
29 the form’.44 The referent of ‘perfecter’ here is a matter of controversy, since after
30 stating that the perfecter is that which gives the form, Avicenna says that ‘it
31 seems that that which gives the constitutive form belonging to natural species is
32 extrinsic to natural things’.45 Here Avicenna clearly states that the form shared
33 by members of a natural species has a cause extrinsic to natural things. But he
34 stops short of claiming that whenever a new compound of form and matter is
35 made, an incorporeal substance produces its form ex nihilo. 46 In my view,
36
38 42 Avicenna, al-Šifā’: al-Samā‘ al-tabı̄‘i [Physics], 1.10, p. 48.
39 43 Avicenna, al-Šifā’: al-Samā‘ al-t˙ abı̄‘i [Physics], 1.10, p. 49.
44 Avicenna, al-Šifā’: al-Samā‘ al-t˙ abı̄‘i [Physics], 1.10, p. 49.
40
45 Avicenna, al-Šifā’: al-Samā‘ al-t˙ abı̄‘i [Physics], 1.10, p. 49.
41 46 Lee, St. Thomas, p. 40, n. 1 and ˙ Brand, Book of Causes, p. 46, n. 7 identify Physics 1.10
42 as support for the view that Avicenna endorses what I call the Infusion Model of
1 Avicenna here refers to the Agent Intellect as a cause of species, rather than a
2 cause of individuals. Avicenna distinguishes these two types of cause in
3 Metaphysics 6.3.
4 In Metaphysics 6.3, Avicenna challenges the causal likeness principle – the
5 view that agent and effect are always alike in kind – as well as the belief that
6 agents are always ‘worthier and stronger’ than their patients with respect to
7 whatever it is the agent gives the patient.47 Avicenna argues that this ‘is neither
8 evident nor true in every respect, unless what it bestows is existence itself, and
9 reality. For then the giver is more worthy of what it bestows than the recipient of
10 what is bestowed’.48 He elaborates the view that the agent is superior to the
11 patient when the agent bestows existence and reality by drawing a distinction
12 between efficient causes of a thing insofar as it is a member of a species and
13 efficient causes of a thing insofar as it is an individual. He argues first that the
14 causes of a thing insofar as it is a member of a species must differ from it in
15 kind. His support for this view rests on the idea that if the cause of a thing
16 insofar as it is a member of some species X were itself a member of species X,
17 then the cause of species X would remain unexplained:
18
the two species [i.e., the species of the cause and the species of the effect] are not
19 one [and the same], since what is being sought after is the cause of that species …
20 These [latter] would be essential causes of the thing absolutely caused with respect
21 to the species of the effect.49
22
By contrast, the cause of a thing qua individual may be the same in kind as it:
23
24 this fire is not the cause of that fire in that it is the cause of the specificity of fire,
but in that it is the cause of some fire. If considered in terms of specificity, it would
25
be the cause of specificity accidentally. The case is similar with [the causal relation]
26 of father to son, not inasmuch as this is a father and that a son, but with respect to
27 the existence of humanity.50
28
29
30
generation. Lee’s and Brand’s interpretation is contested by Robert Wisnovsky in his
31 Ph.D. dissertation, Avicenna on Final Causality. Wisnovsky states his own position on the
32 respective roles of corporeal agents and the Agent Intellect in substantial generation
33 briefly: he says that ‘the Agent Intellect gives existence to the form/matter compound,
34 but the matter is set in motion toward the form by its natural principle of motion, i. e.
the perfecter. It is thus through the mediation of the perfecter in nature (i. e. the form
35
itself ) that the Agent Intellect causes the existence of the compound’ (Wisnovsky,
36 Avicenna, pp. 98 – 9). My own interpretation of Physics 1.10 is generally in accord with
37 this one. But I question the view that the perfecter in nature is the form itself, since
38 Avicenna says that the perfecter gives the form. I consider semen and seeds to be the
39 ‘perfecters in nature’ to which Wisnovsky refers.
40
264 Kara Richardson
1 He argues that agents of the second type are not superior to their patients: for
2 example, when this fire causes that fire, agent and patient are equal, since the
3 fiery form does not admit of degree.51 But agents of the first type are superior to
4 their patients: though ‘existence inasmuch as it is existence does not vary in
5 terms of strength and weakness’, still it varies in terms of ‘priority and
6 posteriority, absence of need and need, and necessity and possibility’.52 In these
7 ways, the cause of thing qua member of a species is superior to that thing, since
8 it is the source of its very existence and reality.
9 The distinction between causes of species and causes of individuals suggests
10 that a superior principle – in my view, Avicenna’s Agent Intellect – is the cause
11 of a type replicated by tokens of the type. This view does not entail an Infusion
12 Model of generation. (Aquinas himself holds this view to be compatible with
13 the Eduction Model of generation. This issue is discussed in Section IV.) If
14 Avicenna’s claim in Physics 1.10 that ‘that which gives the constitutive form
15 belonging to natural species is extrinsic to natural things’ refers to an incorporeal
16 cause of the species of natural things, then this claim is compatible with the view
17 that corporeal agents educe form from the potency of matter in generation.53
18 The latter view also accords with Avicenna’s claims about the powers of semen
19 and seeds in generation, which suggest that he considers these corporeal agents
20 to be ‘perfecters’ who play a form-giving role, rather than ‘preparers’, who
21 dispose matter for form.54
22
23
52 Avicenna, Metaphysics 6.3, p. 213; Š. Il., p. 276; Pr. ph., pp. 317 – 18.
25
53 Avicenna, al-Šifā’: al-Samā‘ al-tabı̄‘i [Physics], 1.10, p. 49.
26 54 For example, in Metaphysics 4.2, ˙ he claims that that what is in potency depends on a
27 proximate mover which is for the most part something of the same kind (muğānis)
28 ‘existing in actuality before the action – as in the case of the hot that heats and the cold
29 that cools’ (Avicenna, Metaphysics 4.2, p. 142; Š. Il., p. 184; Pr. ph., p. 212).
Furthermore, ‘it is often the case that that which is in potency, inasmuch as it is the
30
subject bearing potency, comes into existence through that which is in act, where the act
31 is prior to the potency in time, not [contemporaneous] with it. For sperm comes to be
32 from the human, and seed from the tree, so that, from [the former], a human comes to
33 be, and, from [the latter], a tree. Hence, the supposition that the act in these matters
34 precedes potency has no priority to the supposition that potency precedes the act’
(Avicenna, Metaphysics 4.2, p. 142; Š. Il., p. 184; Pr. ph., p. 212). Here Avicenna seems
35
to endorse Aristotle’s belief that biological organisms generally come to be from
36 corporeal things, which are like them in kind. In Metaphysics 5.9, he argues that sperm
37 has the potency for the formation of the human. He says so while illustrating a point
38 about potencies, namely that a potency can be in potency: ‘potency, inasmuch as it is, is
39 the potency of an existence in act. Potency sometimes also exists in potency, this being
the potency which is remote from act but which becomes in act a proximate potency. For
40
the proximate potency for the formation of the human exists potentially in nourishment.
41 Then, if [the nourishment becomes] sperm, this proximate potency exists in act. It is
42 only that its action [while it is still a proximate potency] does not exist’ (Avicenna,
1 The passages that offer the most compelling textual evidence for the view that
2 Avicenna adopts an Infusion Model have to do with elemental bodies. In
3 Metaphysics 9.5, Avicenna discusses the generation of the elements by the first
4 causes in the context of his emanationist account of creation. He claims first
5 that since the elements are generable and corruptible, ‘their proximate causes
6 must be things that receive a species of change and motion’ and so ‘that which is
7 a pure intellect is not alone a cause for their existence’.55 He identifies these
8 proximate causes as the celestial spheres, or, more specifically, their motions.
9 These motions play a role in the generation of the elements because they
10 contribute to the preparedness of some portion of matter for the reception of
11 one elemental form, rather than another, e. g., for the reception of the form of
12 fire, rather than the form of water. Avicenna refers to this preparatory function
13 as the act of ‘specifying matter’:
14
15 [t]he things that specify matter are the things that prepare it. The preparer is that
through which there comes to be, in the thing prepared, something by virtue of
16
which its appropriateness for [the reception] of a specific thing is more appropriate
17 than [the reception of some] other thing. This act of preparing renders
18 preponderant the existence in it of the more appropriate [form] from the principles
19 that bestow forms.56
20
He then illustrates preparedness by the example of the transformation of water
21
into fire:
22
23 This is similar to water when its warming is made excessive, whereby the alien
24 warmth and the watery form combine, [the former] being remote in appropriateness
25 from the watery form [but] greatly appropriate for the fiery form. If that [warming]
is rendered excessive and the appropriateness intense, the preparedness becomes
26
intense. It thus becomes aright for the fiery form to emanate and aright for this
27 [watery form] to cease.57
28
29 In these passages Avicenna illustrates his view that the elements turn into one
30 another when they surpass a certain degree of warmth/coldness or moistness/
31 dryness. This view is clarified in On Generation and Corruption:
32
33 Metaphysics 5.9, p. 193; Š. Il., p. 252; Pr. ph., pp. 289 – 90). Here Avicenna claims that
34 nourishment, which can become sperm, has the power to acquire the power proper to
sperm, i. e., the power for human formation. In this passage, the sperm’s potency seems
35
clearly to be an active power: Avicenna claims that the sperm’s potency is for the
36 formation of the human. Moreover, the view that male semen has active power is the
37 ground for Avicenna’s distinction between the roles of male semen and female semen in
38 the Kitāb al-Hayawān (Book of Animals). On this aspect of Avicenna’s account of sexual
generation see ˙ Musallam, The Human Embryo, pp. 32 – 4 and McGinnis, On the
39
Moment, p. 55.
40
41 56 Avicenna, Metaphysics 9.5, p. 335; Š. Il., p. 411; Pr. ph., pp. 489 – 90.
266 Kara Richardson
1 [e]ach one of the elements has latitude for receiving increase and decrease in its
2 quality. For it may increase and decrease in its natural or accidental quality. And it
3
may do this while maintaining still its form and species. But increase and decrease in
this has two extreme limits. When they are surpassed, the complete disposition for
4 its form is annulled in matter. And it is prepared with a complete preparation for
5 another form. And it is of the nature of matter when prepared with a complete
6 preparation for a form that it attains this form from the giver of forms to matter and
7 accepts it. And it is for this reason that indistinct matter is distinguished such that it
8
is matter for successive forms. And this from the giver of forms.58
9 Avicenna’s account of the roles played by preparers and by the principles that
10 bestow form in these passages are suggestive of the Infusion Model of substantial
11 generation, and seem to support the view that he endorses this model, at least in
12 cases of elemental transformation. But Avicenna’s intention in such passages may
13 be to identify the Agent Intellect as the efficient cause of the four species of
14 elemental bodies, which are the basis of the system in which elemental
15 transformation occurs. Within this system, fire, for example, has the power to
16 heat water to such a degree that water is completely prepared for the airy form.
17 But the emanation of the airy form is not attributed to the fire that heats it
18 because the fire is not the cause of the fact that water heated to a certain degree
19 becomes air. On this view, elemental transformation is more the product of a
20 system (and so of the efficient cause of this system), than of the causal activity of
21 any individual corporeal agent.
22 The view that Avicenna adopts the Infusion Model of generation, at least in
23 cases of elemental transformation, might find additional support in his account
24 of elemental forms.59 Avicenna denies that the form of an elemental body can be
25 conflated with the two primary qualities, which are characteristic of that body,
26 i. e., he denies that the form of fire can be conflated with the qualities of heat
27 and dryness.60 Avicenna’s view relies on the following claims. First, qualities are
28 accidents and accidents are ontologically posterior to their subjects. Second,
29 substantial forms are ontologically prior to their subjects. Given that substantial
30 forms cannot be identified with their sensible qualities, we cannot know them
31 directly: they are occult entities.61 Stone argues that this view of the elements, as
32 well as a related one about the mixtures they form, supports Avicenna’s
33 ‘occasionalist’ solution to the question of the origins of ‘secondary qualities’,
34 such as colours, of ‘faculties’, such as the magnet’s power to attract iron, as well
35
36 58 Avicenna, al-Šifā’: al-Kawn wa-’l-fasād [Generation and Corruption], c. 14, p. 190.
37 59 This was suggested to my by Jon McGinnis in his appraisal of my dissertation and by
38 Gad Freudenthal at the Villa Vigoni conference on the reception of Avicenna’s
39 Metaphysics.
60 On this aspect of Avicenna’s view, see Avicenna, al-Šifā’: al-Kawn wa-’l-fasād [Generation
40
and Corruption], c. 6. This issue is also discussed in Stone, Simplicius and Avicenna, and
41 Stone, Avicenna’s Theory.
42 61 This point is identified as an Avicennian innovation in Stone, Avicenna’s Theory.
1 as of substantial forms.62 Since these occult entities cannot emerge from

2 mixtures of the elements, they must be ‘called down’ from the heavens.63
3 Stone has shown that Avicenna holds that forms, faculties and qualities such
4 as colours cannot be reduced to the primary qualities (hot, cold, wet and dry)
5 and so that their existence is not explained by the existence of these qualities but
6 rather requires independent explanation. But these claims are compatible with
7 the view that through possession of these non-derivative entities, composite
8 substances exercise causal powers, including generative ones, and thus that
9 generation occurs through natural processes.64 (The fact that these entities are
10 occult does not bar them from acting or being principles of activity, even if their
11 occultness renders them useless for explanatory purposes.) In my view,
12 Avicenna’s elemental theory and his related account of mixture do not suffice
13 to show that he offers an ‘occasionalist’ account of generation. Moreover, I find
14
questionable the application of the term ‘occasionalist’ to Avicenna. It seems to
15
me that he has no occasionalist axe to grind. Even where he emphasizes the
16
superiority of divine causality, he recognizes lesser agents as genuine causes.65 He
17
very clearly holds that corporeal substances are endowed with natures, i. e.,
18
19
20
62 Stone, Avicenna’s Theory, pp. 117 – 8.
21 63 Stone, Avicenna’s Theory, p. 118.
22 64 Stone’s claim that Avicenna adopts an occasionalist view of the origin of substantial
23 forms may overlook the possibility that the Eduction Model of generation is compatible
24 with a commitment to creation. For he claims that in De potentia q. 5 a. 1, Aquinas too
adopts an occasionalist view: ‘although Thomas is generally known as an opponent of
25
Avicenna’s occasionalism, he takes over intact the key part of it: namely, that sublunar
26 causes can only dispose the matter to receive substantial forms, while the actual source of
27 those forms is the angels and celestial bodies’ (Stone, Avicenna’s Theory, p. 119). The
28 phrase ‘actual source’ is ambiguous: Aquinas thinks that God is the creative source of
29 form, but he also thinks, as we have seen, that substantial forms are educed from matter
by corporeal agents in generation. This issue is discussed in Section IV.
30
65 For example, in Metaphysics 6.2, Avicenna argues that creative efficient causality, i. e.,
31 causing existence after absolute non-existence, is superior to non-creative efficient
32 causality, i. e., causing existence after privation in matter; but he does not deny that non-
33 creative efficient causes have power: ‘The originated in the sense [of the thing] that does
34 not require time must be such that either its existence is after absolute nonexistence or its
existence is after nonexistence which is not absolute, being after a specific opposing
35
privation in existing matter, as you have known. If its existence were after absolute
36 nonexistence, then its proceeding from the cause in this manner would be ‘creation,’ and
37 it would represent the highest mode of the giving of existence, because nonexistence
38 would have been utterly prevented, existence being fully empowered over it. If
39 nonexistence is firmly established in a manner whereby it precedes existence, then its
formation would be impossible except out of matter. In this case, the empowerment of
40
bringing into existence – I mean, the existence of one thing from another – would be
41 weak, of short duration, and intermittent’ (Metaphysics 6.2, p. 204; Š. Il., p. 267; Pr. ph.,
42 p. 305).
268 Kara Richardson
1 internal principles of motion and change. And he appears to reject the

2 epistemological arguments for occasionalism.66
3 Avicenna’s discussions of the causal roles played by the Agent Intellect do
4 not provide a clear answer to the question whether he endorses the Infusion
5 Model of generation. In my view, these texts suggest that he points to this
6 incorporeal agent to explain the existence of composite substances, and their
7 species, rather than to explain individual cases of substantial generation. While
8 Aquinas’ argument against the Infusion Model of generation in De potentia q. 3
9 a. 8 is persuasive, it is problematic insofar as it takes Avicenna as one of its
10 targets. In my view, the attribution to Avicenna of the Infusion Model of
11 generation is based on a coarse reading of Avicenna’s discussions of the causal
12 roles played by the Agent Intellect, which conflicts with his account of the unity
13 of form/matter compounds. This interpretation has roots in Averroes’ Long
14
Commentary on the Metaphysics.
15
According to Averroes, Avicenna’s alleged commitment to creation is the
16
source of his (and al-Fārābı̄’s) divergence from Aristotle’s view that forms are
17
educed from matter:
18
19 It is only because the earlier group of men [namely, Ibn Sı̄nā and al-Fārābı̄] neither
20 understood Aristotle’s demonstration [that the forms are not generated in
themselves] nor accepted its truth that they were undone. The conceit does not
21
belong to Ibn Sı̄nā alone, but also to al-Fārābı̄; for it is evident in his book on the
22 two philosophers [Plato and Aristotle] that he had problems concerning this
23 account. This earlier group of men was inclined toward the thought of Plato only
24 because it was an opinion very much akin to that upon which the theologians of our
25 religion rely in this account, namely, that the agent of all [generated] things is one,
26 and that some of the [generated] things do not bring about an effect in others. In
other words, they believed that from some of them creating others they would be
27
committed to the infinite series of actual causes, and so they asserted an incorporeal
28 agent.67
29
30 Averroes’ account of Avicenna’s fall is misleading. Avicenna does not posit an
31 incorporeal agent in order to avoid an infinite regress of corporeal causes of
32 generation. Indeed, as mentioned above, Avicenna endorses an infinite series of
33 corporeal causes of generation in Metaphysics 6.2. There he posits an incorporeal
34 agent in order to explain the very existence of generated things. Nevertheless,
35 there is a tension between the view that there is an incorporeal cause of the
36 existence of generated things and the view that corporeal agents educe form
37
38 66 As Marmura notes, Avicenna endorses the claim that sense experience establishes only
39 concomitance, and not the necessary connections needed to establish causal relation-
ships, but he does not conclude that we cannot know that natural things have causal
40
powers. Marmura suspects that Avicenna’s arguments related to this issue are in response
41 to Aš‘arite occasionalism. See Marmura, The Metaphysics, pp. 183 – 7.
42 67 Averroes, Commentary on Metaphysics, Zeta 9, comm. 31, p. 334.
1 from the potency of matter in generation. This tension is not peculiar to

2 Avicenna’s metaphysics; it is a feature of Aquinas’ metaphysics as well. In
3 Section IV, I examine Aquinas’ treatment of this tension in De potentia q. 5 a. 1,
4 and in Summa theologiae 1a q. 104 a. 1. I show that these texts reveal a very
5 positive reception of some of Avicenna’s discussions of the causal roles of the
6 Agent Intellect, and that Aquinas’ resolution of the tension between the
7 Eduction Model of generation and God’s role as the cause of the existence of
8 generated things integrates several Avicennian ideas.
9
10
11 IV Aquinas’ Arguments for Divine Conservation
12
13 Aquinas’ argument in De potentia q. 5 a. 1 for the view that God preserves all
14 things in being relies on the following account of the relationship between a
15 thing made and an efficient cause:
16
The existence of a thing made depends on an efficient cause insofar as it depends on
17 the form itself of the thing made. But there is some efficient cause on which the
18 thing made depends not per se and according to the account (ratio) of [its] form but
19 only per accidens: thus the form of the generated fire depends on the generating fire
20 not per se and according to the account of its species, since they hold the same rank
21 in the order of things, nor is the form of fire otherwise in the thing generated than
in the generator, but rather it is distinguished from it only by a material distinction
22
(divisio) according as it is in different matter. Whence since the form of the
23 generated fire has some cause, the form itself must depend on a higher principle,
24 which is the cause of the form itself per se and according to the account proper to its
25 species.68
26
Aquinas’ first premise – namely, that ‘the existence of a thing made depends on
27
an efficient cause insofar as it depends on the form itself of the thing made’ –
28
reflects his view that the form of the composite is that by which the composite
29
exists and thus that the existence of the composite depends on its form. This
30
suggests that the efficient cause of the existence of the composite and its form
31
are identical. Aquinas then distinguishes two types of efficient cause, one on
32
which a thing made depends per se and another on which it depends per accidens.
33
In the case of a generated fire, for example, the form of that fire depends on the
34
35
68 Aquinas, De potentia, q. 5 a. 1: ‘Secundum hoc ergo esse rei factae dependet a causa
36 efficiente secundum quod dependet ab ipsa forma rei factae. Est autem aliquod efficiens
37 a quo forma rei factae non dependet per se et secundum rationem formae, sed solum per
38 accidens: sicut forma ignis generati ab igne generante, per se quidem, et secundum
39 rationem suae speciei non dependet, cum in ordine rerum eumdem gradum teneat, nec
forma ignis aliter sit in generato quam in generante; sed distinguitur ab ea solum
40
divisione materiali, prout scilicet est in alia materia. Unde cum igni generato sua forma
41 sit ab aliqua causa, oportet ipsam formam dependere ab altiori principio, quod sit causa
42 ipsius formae per se et secundum propriam speciei rationem.’
270 Kara Richardson
1 generating fire only accidentally. It cannot depend on the generating fire

2 essentially, since the generating and the generated fire share the same type of
3 form. Thus it must depend on a ‘higher’ principle, the per se cause of the form
4 itself (God). Here Aquinas adapts Avicenna’s argument in Metaphysics 6.3 for a
5 distinction between causes of species and causes of individuals. (This argument
6 is discussed in Section III.)
7 The ground for Avicenna’s distinction between causes of species and causes
8 of individuals is the idea that the cause of an effect’s specificity cannot be
9 another member of its species. Aquinas offers the following explanation in
10 Summa theologiae 1a q. 104 a. 1:
11
Now it is clear that of two things in the same species one cannot be essentially the
12 cause of the other’s form as such, since it would then be the cause of its own form,
13 since both forms have the same nature; but it can be the cause of this form
14 inasmuch as it is in matter – in other words, it may be the cause that this matter
15 receives this form. And this is to be the cause of becoming, as when man begets man,
and fire causes fire.69
16
17 Aquinas argues that if the generating fire were the essential cause of the
18 existence of any fiery form, then it would be the cause of the existence of its own
19 form, which is impossible. (Nothing is the cause of itself.) He then alludes to a
20 distinction between causes of coming to be and causes of existence, which
21 Avicenna defends in Metaphysics, Book 6, chapters 1 – 2: corporeal agents cause
22 the becoming, but not the existence of generated things.70 The cause of the
23 existence of generated things, Aquinas argues (in Summa theologiae 1.104.1, as
24 well as in De potentia 5.1), is a ‘higher’ principle, which is the essential or per se
25 cause of form: this is God.71
26
27
28 69 Aquinas, Summa theologiae, 1a q. 104 a. 1: ‘Manifestum est autem quod, si aliqua duo
29 sunt eiusdem speciei, unum non potest esse per se causa formae alterius, inquantum est
talis forma, quia sic esset causa formae propriae, cum sit eadem ratio utriusque. Sed
30
potest esse causa huiusmodi formae secundum quod est in materia, idest quod haec
31 materia acquirat hanc formam. Et hoc est esse causa secundum fieri; sicut cum homo
32 generat hominem, et ignis ignem.’
33 70 The provenance of the distinction is noted by Doolan, The Causality, p. 399, n. 29 and
34 Wippel, Thomas Aquinas, p. 202.
71 In De potentia 5.1, the argument for this conclusion – sc., that the essential or per se
35
cause of the existence of the form of the composite is God – relies on two key claims.
36 First, ‘the existence of a form in matter implies no movement or change except
37 accidentally’ and, second, bodies act only by being moved (Aquinas, De potentia, q. 5 a. 1
38 co.): ‘Cum autem esse formae in materia, per se loquendo, nullum motum vel
39 mutationem implicet, nisi forte per accidens; omne autem corpus non agat nisi motum,
ut philosophus probat, necesse est quod principium ex quo per se dependet forma, sit
40
aliquod principium incorporeum; per actionem enim alicuius principii dependet effectus
41 a causa agente.’ I take Aquinas to reason as follows: if bodies act only by being moved,
42 then the effects of their actions will be movements or changes. Since the existence of a
1 The claim that corporeal agents cause the becoming but not the existence of
2 corporeal things, together with the claim that that existence of the composite
3 depends on its form might seem to provide support for the Infusion Model of
4 generation. Aquinas raises this objection to his claim that God preserves all
5 things in being in De potentia q. 5 a. 1:
6
Someone might say that lower agents are causes of becoming and not of being. For
7
this reason the existence of an effect remains once the action of its cause of
8 becoming has ceased. But God causes not only the becoming but also the being of
9 things. For this reason the existence of things cannot remain if divine action ceases.
10 – On the contrary, every generated thing has existence through its form. If therefore
11 lower generating causes are not causes of existence, they will not be causes of form
12
and thus the forms which are in matter are not from forms which are in matter, as
according to the opinion of the Philosopher, who says that the form which is in this
13 flesh and bones is from the form which is in those flesh and bones, but rather it
14 follows that forms in matter are from forms without matter, as according to the
15 opinion of Plato, or from the giver of forms, as according to the opinion of
16 Avicenna.72
17
Here the objector assumes as true the view of the relative roles of lower
18
generating agents and God with respect to the existence of a thing, which
19
Aquinas adopts from Avicenna: lower generating agents are causes of becoming,
20
but not of being. The objector suggests that on this view, God is the cause of the
21
form of the thing generated: for if God is the cause of the existence of a thing,
22
and the existence of that thing is due to its form, then God is the cause of its
23
form. This conclusion seems to conflict with the view that corporeal agents
24
educe form from the potency of matter, and to support the Infusion Model of
25
generation. Aquinas rejects this line of reasoning. He replies that as agents of
26
becoming, lower generating causes educe form from matter. In this way (i. e., by
27
28
educing), they are causes of substantial forms and thus they are ‘principles of
29
existence considered in terms of its beginning not as considered absolutely’.73
30
In De potentia q. 5 a. 1, Aquinas supports his claim that the Eduction Model
31
is compatible with God’s role as the cause of the existence of generated things
32
through a cooperative account of the roles of God and corporeal agents in
33 generation. Insofar as a corporeal agent is a cause of form it acts in the power of
34
form in matter does not involve movement or change, no body can be the cause of the
35
existence of a form in matter. So the cause of the existence of a form in matter must be
36 an incorporeal principle, which Aquinas identifies as God.
37 72 Aquinas, De potentia, q. 5 a. 1 arg. 5
38 73 Aquinas, De potentia, q. 5 a. 1 ad 5: ‘Si autem ponamus formas substantiales educi de
39 potentia materiae, secundum sententiam Aristotelis, agentia naturalia non solum erunt
causae dispositionum materiae, sed etiam formarum substantialium; quantum ad hoc
40
dumtaxat quod de potentia educuntur in actum, ut dictum est, et per consequens sunt
41 essendi principia quantum ad inchoationem ad esse, et non quantum ad ipsum esse
42 absolute.’
272 Kara Richardson
1 an incorporeal principle as if it were its instrument.74 This agent thus makes a

2 genuine causal contribution to generation, even though it is merely a per
3 accidens cause of the existence of generated things. In Section III, I discussed
4 Avicenna’s very similar claims Metaphysics 6.2: due to their causal roles in
5 generation, corporeal agents are identified as accidental or helping causes of the
6 existence of the composite, while the Agent Intellect is identified as the essential
7 cause of the existence of the composite. According to Aquinas, the difference
8 between his view of generation and Avicenna’s is that when Avicenna identifies
9 builders and fathers as accidental or helping causes of existence, he means to
10 deny that they educe form from matter. In Section II, I argued that this
11 interpretation conflicts with Avicenna’s view of hylemorphic unity, and in
12 Section III, I argued that it is based on a questionable reading of Avicenna’s
13 discussions of the causal roles of the incorporeal Giver of Forms.
14 Aquinas’ arguments for divine conservation borrow several distinctions from
15
Avicenna’s metaphysics in order to support the view that there is an incorporeal
16
agent who causes the existence of the composite by causing its form, and to
17
explain how the latter view accords with the Eduction Model of generation.
18
Aquinas’ use of these distinctions shows that key parts of Avicenna’s view of the
19
causal roles of the Agent Intellect are adopted as important features of
20
Thomistic metaphysics. Aquinas’ positive reception of these aspects of
21
Avicenna’s metaphysics may be obscured by his very negative critique of
22
Avicenna’s account of generation in De potentia q. 3 a. 8, which accuses
23
Avicenna of serious metaphysical errors. Through examining Avicenna’s account
24
25
of hylemorphic unity, and his discussions of the causal roles of the Agent
26
Intellect, I hope to have shown the demerits of this critique, and also to have
27
challenged the view that Avicenna endorses the Infusion Model of generation.
28
29
30
Bibliography
31
Thomas Aquinas, Summa contra gentiles, Book Three: Providence, Part 1, transl. and ed.
32
V.J. Bourke, Notre Dame: University of Notre Dame Press, 2001.
33 –– , Summa theologiae, Part One (Basic Writings of Saint Thomas Aquinas, vol. 1), ed.
34 A. Pegis, Indianapolis: Hackett, 1997.
35 –– , On the Power of God, transl. and ed. English Dominican Fathers, Westminster,
36 Maryland: The Newman Press, 1952.
37 –– , Quaestiones disputatae, 9th rev. ed., Rome: Marietti, 1953.
–– , Opera omnia iussu impensaque Leonis XIII P. M. edita, t. 4 – 5: Pars prima Summae
38
theologiae, Rome: Ex Typographia Polyglotta S.C. de Propaganda Fide, 1888 – 9.
39
40
74 Aquinas, De potentia, q. 5 a. 1: ‘Et si aliquod principium corporeum est per aliquem
41 modum causa formae, hoc habet in quantum agit virtute principii incorporei, quasi eius
42 instrumentum.’
1 Aristotle, The Complete Works of Aristotle, ed. J. Barnes, Princeton: Princeton University
2 Press, 1984.
3
Averroes, Commentary on Metaphysics, Zeta 9, in Classical Arabic Philosophy, transl.
and ed. J. McGinnis and D.C. Reisman, Indianapolis: Hackett, 2007, pp. 330 – 35.
4 –– , Tafsı̄r mā ba‘d at-tabı̄‘iyya, 3 vols, ed. M. Bouyges, Beirut: Imprimerie Catholique,
5 1938 – 52. ˙
6 Avicenna, al-Šifā’: al-Ilāhiyyāt, vol. 1, eds G. C. Anawati and S. Zāyed, Cairo:
7 Organisation Gnrale des Imprimeries Gouvernementales, 1960.
8
–– , al-Šifā’: al- Ilāhiyyāt, vol. 2, eds M.Y. Moussa, S. Dunyā and S. Zāyed, Cairo:
Organisation Gnrale des Imprimeries Gouvernementales, 1960.
9
–– , al-Šifā’: al-Kawn wa-’l-fasād, ed. M. Qāsim, Cairo: Organisation Gnrale des
10 Imprimeries Gouvernementales, 1983.
11 –– , al-Šifā’: al-Hayawān, ed. ‘A. Montasir, S. Zāyed and ‘A. Ismā‘ı̄l, Cairo: Organisation
12 Egyptienne˙ Gnrale du Livres, 1970.
13 –– , al-Šifā’: al-Samā‘ al-tabı̄‘i [Physics], ed. S. Zāyed, Cairo: Organisation Gnrale des
˙
Imprimeries Gouvernementales, 1983.
14
–– , Avicenna Latinus: Liber de philosophia prima sive Scientia divina, 3 vol., ed. S. Van
15 Riet, Leiden: Peeters; Louvain-Brill, 1977 – 83.
16 –– , Avicenna Latinus: Liber tertius naturalium de generatione et corruptione, ed. S. Van
17 Riet, Leiden: Peeters; Louvain-Brill, 1987.
19
D.J. Brand, transl. and ed., Book of Causes, 2nd rev., Milwaukee: Marquette University
20 Press, 1984.
21 H.A. Davidson, Alfarabi, Avicenna, Averroes, on Intellect, Oxford: Oxford University
22 Press, 1992.
23 G.T. Doolan, The Causality of the Divine Ideas in Relation to Natural Agents,
International Philosophical Quarterly, vol. 44, 2004, pp. 393 – 409.
24
T. Druart, Metaphysics, in The Cambridge Companion to Arabic Philosophy, ed.
25 P. Adamson, Cambridge: Cambridge University Press, 2005, pp. 327 – 48.
26 G. Freudenthal, The Medieval Astrologization of Aristotle’s Biology: Averroes on the
27 Role of the Celestial Bodies in the Generation of Animate Beings, Arabic Sciences
28 and Philosophy, vol. 12, 2002, pp. 111 – 37.
29
31 –– , Plato arabico-latinus: Philosophy-Wisdom Literature-Occult Sciences, in The
32 Platonic Tradition in the Middle Ages. A Doxographic Approach, eds S. Gersh,
33 M.J.F.M. Hoenen, Berlin/New York: Walter de Gruyter, 2002, pp. 31 – 65.
34
–– , Spontaneous Generation and the Ontology of Forms in Greek, Arabic, and
Medieval Latin Sources, in Classical Arabic Philosophy: Sources and Reception, ed.
35 P. Adamson, London/Turin: The Warburg Institute/Nino Aragno Editore, 2007,
36 pp. 150 – 175.
37 A. Hyman, Aristotle’s ‘First Matter’ and Avicenna’s and Averroes’ ‘Corporeal Form’, in
38 Harry Austryn Wolfson Jubilee Volume, English Section, vol. 1, ed. S. Lieberman,
39 Jerusalem: American Academy for Jewish Research, 1965, pp. 385 – 406.
J. Janssens, The Notions of Wāhib al-Suwar (Giver of Forms) and Wāhib al-‘Aql
40
(Bestower of Intelligence) in Ibn Sı̄nā, in Actes du XIe Congrs International de
41 Philosophie Mdivale de S.I.E.P.M., 2002, eds M. Pacheco et J. Meirinhos,
42 Turnhout: Brepols, 2006, pp. 551 – 62.
274 Kara Richardson
1 P. Lee, St. Thomas and Avicenna on the Agent Intellect, Thomist, 45, no. 1, 1981,
2 pp. 41 – 61.
3
G.W. Leibniz, De ipsa natura, in Die philosophischen Schriften von Gottfried Wilhelm
Leibniz, vol. 4, ed. C.I. Gerhardt, Berlin: Weidmannsche Buchhandlung, 1880,
4 pp. 504 – 16.
5 O. Lizzini, The Relation Between Form and Matter: Some Brief Observations on the
6 ‘Homology Argument’ (Ilāhiyyāt, II.4) and the Deduction of Fluxus, in Interpreting
7 Avicenna: Science and Philosophy in Medieval Islam, ed. J. McGinnis, Leiden: Brill,
8
2004, pp. 175 – 85.
M. Marmura, The Metaphysics of Efficient Causality in Avicenna (Ibn Sı̄nā), in Islamic
9 Theology and Philosophy, ed. M. Marmura, Albany, New York: SUNY Press, 1984,
10 pp. 172 – 87.
11 J. McGinnis, On the Moment of Substantial Change: A Vexed Question in the History
12 of Ideas, in Interpreting Avicenna: Science and Philosophy in Medieval Islam, ed.
13
J. McGinnis, Leiden: Brill, 2004, pp. 42 – 61.
B. Musallam, The Human Embryo in Arabic Scientific and Religious Thought, in The
14 Human Embryo: Aristotle and the Arabic and European Traditions, ed. G.R. Dunstan,
15 Exeter: University of Exeter Press, 1990, pp. 32 – 48.
16 S.C. Snyder, Albert the Great, Incohatio Formae, and the Pure Potentiality of Matter,
17 American Catholic Philosophical Quarterly, LXX, no. 1, 1996, pp. 63 – 82.
18
A. Stone, Avicenna’s Theory of Primary Mixture, Arabic Sciences and Philosophy, vol. 18,
2008, pp. 99 – 119.
19 –– , Simplicius and Avicenna on the Essential Corporeity of Material Substance, in
20 Aspects of Avicenna, ed. Robert Wisnovsky, Princeton: Princeton University Press,
21 2001, pp. 73 – 130.
22 J.A. Weisheipl, Aristotle’s Concept of Nature: Avicenna and Aquinas, in Approaches to
23
Nature in the Middle Ages, ed. L. D. Roberts, Binghampton, New York: Centre for
Medieval & Early Renaissance Studies, 1982, pp. 137 – 60.
24 J.F. Wippel, Thomas Aquinas on Creatures as Causes of Esse, International Philosophical
25 Quarterly, 40, 2000, pp. 197 – 213.
26 R. Wisnovsky, Avicenna on Final Causality, Ph.D. Dissertation, Princeton University,
27 1994.
28
B.H. Zedler, Saint Thomas and Avicenna in the De Potentia Dei, Traditio, 6, 1948,
pp. 105 – 59.
29
30
31
32
33
34
35
36
37
38
39
40
41
42
1
2
3
4 Immateriality and Separation
5 in Avicenna and Thomas Aquinas
6
7 Pasquale Porro
8
9
If we consider the traditional, Scholastic classification of the speculative
10
sciences, it seems rather obvious that first philosophy, or metaphysics, has little
11
to do with matter or with material beings, and has much to do, or has only to
12
do, with beings separate from matter and motion. This is explicitly affirmed by
13
many medieval masters, including Thomas Aquinas. Just to mention a few
14
examples:
15
16 For since each thing has intellective power by virtue of being free from matter, those
17
things must be intelligible in the highest degree which are altogether separate from
matter. … Now those things are separate from matter in the highest degree which
18 abstract not only from designated matter, ‘as the natural forms taken universally, of
19 which the philosophy of nature treats’, but from sensible matter altogether; and
20 these are separate from matter not only in their intelligible constitution, as the
21 objects of mathematics, but also in being, as God and the intelligences.1 (Thomas de
22
Aquino, In Metaph., Prooemium, transl. Rowan).
23 Consequently, separation from matter and motion, or connection with them,
24 essentially belongs to an object of speculation, which is the object of speculative
science. As a result, the speculative sciences are differentiated according to their
25
degree of separation from matter and motion. … There are still other objects of
26 speculative knowledge that do not depend upon matter for their being, because they
27 can exist without matter; either they never exist in matter, as in the case of God and
28 the angels, or they exist in matter in some instances and not in others, as in the case
29 of substance, quality, being, potency, act, one and many, and the like. The science
that treats of all these is theology or divine science, which is so called because its
30
principal object is God.2 (Thomas de Aquino, Super Boetium De Trinitate, q. V, a. 1,
31 transl. Maurer).
32
33
34 1 Thomas de Aquino, In Metaph., Prooemium, eds Cathala and Spiazzi, p. 1: ‘cum
unaquaeque res ex hoc ipso vim intellectivam habeat, quod est a materia immunis,
35
oportet illa esse maxime intelligibilia, quae sunt maxime a materia separata. … Ea vero
36 sunt maxime a materia separata, quae non tantum a signata materia abstrahunt, sicut
37 formae naturales in universali acceptae, de quibus tractat scientia naturalis, sed omnino a
38 materia sensibili. Et non solum secundum rationem, sicut mathematica, sed etiam se-
39 cundum esse, sicut Deus et intelligentiae.’
2 Thomas de Aquino, Super Boetium De Trinitate, q. V, a. 1, ed. leon., p. 138, lines 135 –
40
63: ‘Sic ergo speculabili, quod est obiectum scientie speculatiue, per se competit
41 separatio a materia et motu, uel applicatio ad ea; et ideo secundum ordinem remotionis a
42 materia et motu scientie speculatiue distinguntur. Quedam ergo speculabilium sunt que
276 Pasquale Porro
1 It must be understood, therefore, that there are some things whose existence
2 depends upon matter, and which cannot be defined without matter. Further there
3 are other things which, even though they cannot exist except in sensible matter, have
no sensible matter in their definitions. … And this is true of all the mathematicals,
4
such as numbers, magnitudes and figures. Then, there are still other things which do
5 not depend upon matter either according to their existence or according to their
6 definitions. And this is either because they never exist in matter, such as God and
7 the other separated substances, or because they do not universally exist in matter,
8 such as substance, potency and act, and being itself.
9 Now metaphysics deals with things of this latter sort. Whereas mathematics
deals with those things which depend upon sensible matter for their existence but
10
not for their definitions. And natural science, which is called physics, deals with
11 those things which depend upon matter not only for their existence, but also for
12 their definition.3 (Thomas de Aquino, In Phys., I, lect. 1, transls Blackwell, Spath,
13 Thirlkel).
14
Of course, we could say that this view had already been expressed in Aristotle’s
15
Metaphysics, and especially in E 1, which is nevertheless a text to be read with
16
some caution, as we shall see. Yet, if the medieval masters generally agree on the
17
18
fact that metaphysics concerns what is immaterial, they disagree about the
19
proper subject of metaphysics, i. e., the question of whether it coincides with
20
God and other separate substances or with being qua being.4 There is no doubt
21
that this alternative can also be found in Aristotle, for his new science or ‘the
22 science we are searching for’ (1pifgtoul]mg 1pist^lg, as Aristotle says in B 1,
23 995a24) has at least a twofold structure: on the one hand, in C, as is well-
24 known, the subject of this science is identified with being qua being; on the
25 other, for instance in E 1, the same science is called heokocij^: in other words, it
26 dependent a materia secundum esse, quia non nisi in materia esse possunt. … Quedam
27 uero speculabilia sunt que non dependent a materia secundum esse, quia sine materia
28 esse possunt, siue numquam sint in materia, sicut Deus et angelus, siue in quibusdam
29 sint in materia et in quibusdam non, ut substantia, qualitas, ens, potentia, actus, unum et
multa, et huiusmodi; de quibus omnibus est theologia, id est scientia diuina, quia
30
precipuum in ea cognitorum est Deus. Que alio nomine dicitur metaphisica …’ [The
31 italics are mine].
32 3 Thomas de Aquino, In Phys., I, lect. 1, ed. Maggiolo, p. 3, §§ 2 – 3: ‘Sciendum est igitur
33 quod quaedam sunt quorum esse dependet a materia, nec sine materia definiri possunt:
34 quaedam vero sunt quae licet esse non possint nisi in materia sensibili, in eorum tamen
definitione materia sensibilis non cadit. … Quaedam vero sunt quae non dependent a
35
materia nec secundum esse nec secundum rationem; vel quia numquam sunt in materia,
36 ut Deus et aliae substantiae separatae; vel quia non universaliter sunt in materia, ut
37 substantia, potentia et actus, et ipsum ens. De huiusmodi igitur est Metaphysica: de his
38 vero quae dependent a materia sensibili secundum esse sed non secundum rationem, est
39 Mathematica: de his vero quae dependent a materia non solum secundum esse sed etiam
secundum rationem, est Naturalis, quae Physica dicitur.’
40
4 Suffice it to consider the material discussed in this regard by Zimmermann in his
41 well known volume Ontologie oder Metaphysik. See also Porro, ed.,
42 Metaphysica–sapientia–scientia divina.
Immateriality and Separation in Avicenna and Thomas Aquinas 277
1 is a theology (or divine science) which concerns what is separate from matter
2 and motion (peq· wyqist± ja· !j_mgta). The history of the reception of
3 Aristotle’s Metaphysics is above all the history of all the attempts – from the
4 ancient commentators to Heidegger’s famous theses on the onto-theological
5 constitution of Western metaphysics – to find a way to balance or to combine
6 these two distinct subject-matters: being qua being and the divine. It is generally
7 agreed that this twofold structure originates in the fact that Aristotle attributes
8 to his new science two distinct features: metaphysics should be both universal
9 and first (or primary), and so it should deal, at one and the same time, with that
10 which is not a genus, because it encompasses all possible genera (being qua
11 being), and with that which is a genus, the supreme or highest genus, the divine,
12 i. e. that which is eternal, immaterial and immovable. But how can a qualified
13 subject (being qua immovable, eternal and immaterial, or being qua divine)
14 coincide with an unqualified subject (being qua being)? Or, to reverse the
15 question, how can being qua being exclude from itself material being, if matter
16 is one of the principles which is included in the universality of being? In the
17 thirteenth century, Thomas Aquinas sought out a possible solution to this
18 question, and found it in Avicenna. But while borrowing from Avicenna the
19 essential part of the solution (i. e. the distinction between two different
20 meanings of ‘immaterial’ and ‘immateriality’, as we shall see), some of his
21 conclusions seem radically different from those of Avicenna. This is precisely the
22 issue I would like to deal with here.
23 My exposition consists of five parts. (1) I will show how the question
24 concerning the immateriality of the subject of metaphysics is raised in Aristotle
25 and in his ancient commentators. (2) I will explain the way in which Thomas
26 Aquinas deals with this issue in his commentary on the Metaphysics and, above
27 all, in his commentary on Boethius’ De trinitate, in which he introduces a
28 distinction between two degrees of abstraction proper to physics and
29 mathematics, and a further distinct operation, called separatio, proper to divine
30 science. (3) I will show how Aquinas draws this distinction from Avicenna, and
31 yet distorts the intentions of Avicenna himself in order to reach a different and
32 perhaps divergent conclusion. (4) I will consider the main logical tool used by
33 Avicenna in his re-interpretation of metaphysics and his criticism of Platonism:
34 the distinction between plain negation and metathetic affirmation (or negation
35 by equipollence). (5) I will briefly investigate the possible reasons for the
36 peculiar attitude Aquinas adopts towards this issue, with respect to Avicenna’s
37 legacy.
38
39
40
41
42
278 Pasquale Porro
1 1. Ancient Problems: Aristotle and his Commentators

2
3 As mentioned above, and as is well-known, Aristotle admits two conflicting
4 requirements for the new science: universality and eminence. That is: the new
5 science must deal with the highest and supreme causes and with the most
6 universal things. This conflict can be easily detected in A or in the first 5 aporiai
7 of B, and explodes, of course, in the transition from C to E :
8
But if there is something which is eternal and immovable and separable (eQ d³ t_
9 1stim !ýdiom ja· !j_mgtom ja· wyqist|m), clearly the knowledge of it belongs to a
10 theoretical science, – not, however, to natural science (for natural science deals with
11 certain movable things), nor to mathematics, but to a science prior to both. For
12 natural science deals with things which are inseparable from matter but not
immovable [sed cf. Schwegler: peq· wyqist± l³m !kkûoqj !j¸mgta], and some parts of
13
mathematics deal with things which are immovable, but probably not separable, but
14 embodied in matter (peq· !j¸mgta l³m oq wyqist± d³ Usyr !kkû¡r 1m vk,); while the
15 first science deals with things which are both separable and immovable (peq· wyqi-
16 st± ja· !j¸mgta). … There must, then, be three theoretical philosophies,
17 mathematics, natural science and theology (heokocij¶) … (Arist., Metaph., VI, 1,
1026a10 – 19; transl. Ross).
18
19 I shall leave aside all the problems concerning the well-known correction
20 suggested by Schwegler and accepted by many subsequent editors, according to
21 which, in order to preserve the grammatical balance of the sentence, the objects
22 of physics should be characterized, in this passage, as wyqist², just like the
23 object of divine science.5 As a matter of fact, whether we accept Schwegler’s
24 correction or maintain the reading of the manuscripts and of the Bekker
25 edition, the concept of separation, in this crucial text, and in Aristotle generally,
26 is rather unclear, especially if we consider that ‘separation’ (wyqislºr) is precisely
27 the core of Aristotle’s criticism towards Plato’s theory of forms (Metaph., XIII, 9,
28 1086b6 – 7). Aristotle, for his part, seems to use wyqislºr, wyq_feim and wyqi-
29 stºr, -^, -|m in four different principal meanings:
30
31
32
33 5 See Aristotle, Die Metaphysik, ed. Schwegler, pp. 14 – 16; Metaphysics, ed. Ross, ad loc.;
34 Metaphysica, ed. Jaeger, ad loc.; see also Jaeger, Aristoteles, p. 225, n. 1. Against the
correction see for instance: Apelt, Beitrge zur Geschichte der griechischen Philosophie, p.
35
231; Cousin, A Note on the Text of Metaphysics 1026a14, pp. 495 – 6; Trpanier, La
36 philosophie de la nature porte-t-elle sur des spars ou des non-spars?, pp. 206 – 9;
37 Owens, The Doctrine of Being in Aristotelian Metaphysics, p. 296, n. 44; Wundt,
38 Untersuchungen zur Metaphysik des Aristoteles, p. 49; Dcarie, La physique porte-t-elle sur
39 des ‘non-spars’ ( !w¾qista)?, pp. 466 – 8; Halper, One and Many in Aristotle’s
Metaphysics, pp. 258 – 9; Zekl, Topos: Die aristotelische Lehre von Raum, p. 85, n. 94;
40
Pietsch, Prinzipienfindung bei Aristoteles, esp. pp. 276 – 7, n. 93; Cleary, Emending
41 Aristotle’s Division of the Theoretical Sciences, pp. 33 – 70; Cleary, Aristotle &
42 Mathematics, esp. pp. 426 – 32.
1 – with respect to place (time, quantity);

2 – with respect to definition (formula, logos);
3 – with respect to matter;
4 – as a capacity of independent existence (‘separate’ is what is capable of separate
5 existence).6
6
If we do not accept Schwegler’s correction, we should read the text quoted above
7
according to the third meaning (according to which ‘separate’ means ‘separate
8
from matter’, and thus immaterial). In this case, physics concerns those things
9
which are material and movable, mathematics those things which are material
10
but immovable, and divine science, or theology, concerns those things which are
11
immaterial and immovable. If we accept Schwegler’s correction, we should read
12
the use of wyqist² in this text, at least insofar as physics is concerned, according
13
to the fourth meaning: separate is what is capable of independent existence.
14
Consequently, physics would concern those beings which possess independent
15
existence but are movable; mathematics would concern those beings which do
16
not possess independent existence and are immovable, while divine science, or
17
theology, would concern those beings which possess independent existence and
18
are immovable. According to this reading, the distinction between the three
19
theoretical sciences is not immediately deduced from the degree of separation
20
from matter, but has mainly to do with different kinds of substantial and
21
accidental being. In other words, we would have here a distinction between
22
these three different modes of being: substantial being in motion; accidental
23
being not in motion; and substantial being not in motion. Immateriality would
24
still play a role in this distinction, but only in an indirect way, since the
25
possibility of motion is always rooted in the potentiality of matter.
26
However, in the famous conclusion of E 1, Aristotle himself draws an
27
explicit connection between the study of universal being and the study of
28
supreme being. This science, so Aristotle writes, is universal because it is
29
primary:
30
31 One might indeed raise the question whether first philosophy is universal, or deals
32 with one genus, i. e. some one kind of being. […] We answer that if there is no
substance other than those which are formed by nature, natural science will be the
33
34
6 For the different meanings and textual occurrences in Aristotle see Philippe, *va_qesir,
35
pq|shesir, wyq_feim dans la philosophie d’Aristote, pp. 461 – 79; Berger, ‘Wegnemen’ en
36 ‘scheiden’, pp. 207 – 78; Morrison, Wyqistºr in Aristotle, pp. 89 – 105. See also the
37 debate on the exact meaning of separation in Aristotle between Gail Fine and Donald
38 Morrison, and the subsequent contributions in this regard by Charlotte Witte and Lynne
39 Spellmann: Fine, Separation, pp. 31 – 87; Morrison, Separation in Aristotle’s
Metaphysics, pp. 125 – 57; Fine, Separation: A Reply to Morrison, pp. 159 – 65;
40
Morrison, Separation: A Reply to Fine, pp. 167 – 73; Witt, Substance and Essence in
41 Aristotle; Spellmann, Substance and Separation in Aristotle. For a survey, see also my
42 Astrazione e separazione: Tommaso d’Aquino e la tradizione greco-araba, pp. 527 – 80.
280 Pasquale Porro
1 first science; but if there is an immovable substance, the science of this must be
2 prior and must be first philosophy, and universal in this way, because it is first
3
(jahºkou ovtyr fti pq¾tg). And it will belong to this to consider being qua being –
both what it is and the attributes which belong to it qua being (Arist., Metaph., VI,
4 1, 1026a23 – 32; transl. Ross).
5
6 Now, the main problem in this twofold structure of first philosophy seems to be
7 represented precisely by matter. A science of being qua being cannot exclude
8 matter, since matter is one of the principles which both constitute and explain
9 many substances (as is stated, for instance, in F 3). But in the text of E 1
10 mentioned above, Aristotle clearly states that theological science concerns what
11 is eternal, immaterial and immovable. It was precisely this contradiction which
12 lead Jaeger to postulate that the final passage of E 1 was a later attempt to
13 reconcile Aristotle’s Urmetaphysik (A, B, C, E 1, L 9 – 10, M) with the
14 Sptmetaphysik (F, G, H), or which lead Natorp to the conviction that – due to
15 this unleidlicher Widerspruch, this unbearable contradiction – E 1 and J must be
16 expunged from the authentic text of the Metaphysics. 7 If the sphere of being qua
17 being includes material beings, we cannot say that first philosophy has only to
18 do with immaterial beings, with those beings which are separate from matter.
19 I shall leave aside the debate amongst contemporary scholars, though I shall
20 briefly consider the position of the ancient commentators. As is well-known, we
21 have three ancient commentaries on the Metaphysics: one by Alexander of
22 Aphrodisias, one by Syrianus and one by Asclepius (i. e. a commentary by
23 Ammonius reported by his pupil Asclepius). But since the commentary
24 attributed to Alexander is authentic only for books A–D, we could add to this
25 list a fourth commentary, that of Pseudo-Alexander, i. e. books E–M included in
26 Hayduck’s edition of Alexander’s commentary. There are very good reasons to
27 suspect, together with Karl Praechter and Concetta Luna, that Pseudo-
28 Alexander is in actual fact Michael of Ephesus, who was a twelfth century
29 Byzantine commentator.8 Now, only two of these four commentaries include
30 book E and discuss the above-mentioned passage of E 1, and these are the
31 commentaries by Pseudo-Alexander and Asclepius. Their interpretation on this
32 particular issue is more or less uniform: ‘theology’ deals with what is separate
33 (from matter) and immovable.9
34
35
7 See Natorp, Thema und Disposition der aristotelischen Metaphysik, pp. 37 – 65 and
36 540 – 74; Jaeger, Aristoteles, esp. pp. 170 – 99 (Die Urmetaphysik) and 200 – 36 (Die
37 Entwicklung der Metaphysik).
38 8 See Praechter, Review of Michaelis Ephesii In libros De partibus animalium, De
39 animalium motione, De animalium incessu, pp. 861 – 907; Luna, Trois tudes sur la
tradition des commentaires anciens la ‘Mtaphysique’, esp. pp. 53 – 71, 197 – 212.
40
9 See for instance [Ps.-]Alexander Aphrodisiensis, In Metaph., E 1, p. 446, line 35–p. 447,
41 line 3; Asclepius, In Metaph., E 1, p. 360, line 31–p. 361, line 6; Syrianus, In Metaph., C
42 2, p. 61, lines 17 – 28. We find this very same tri-partition in many Neoplatonic
1 We might be tempted to discard Pseudo-Alexander, since he is probably a

2 very late commentator, yet his commentary deserves some consideration because
3 it often reports many theses coming from Syrianus.10 As already mentioned, we
4 do not possess Syrianus’ own commentary on E, but we can detect in C a
5 threefold distinction between different kinds of being based on their respective
6 degree of separation from matter and motion.
7 To summarize, we can say that, roughly speaking, the standard interpre-
8 tation of Aristotle by his ancient commentators relied upon two main theses:
9 i) first philosophy is mainly theology;
10 ii) the subject of this science is obtained through a real separation from
11 matter and motion, while abstraction, i. e. the possibility of conceiving
12 something apart from its real existence in matter, gives us the subject of
13 mathematics.
14
15
16 2. From Boethius to Thomas Aquinas
17
18 In the Latin West, too, even before Aristotle’s Metaphysics came to be known and
19 read, the choice appears already made. The most significant text concerning the
20 division of the sciences before the introduction of the Greek and Arabic works is
21 undoubtedly one of Boethius’ small theological tracts – the De trinitate. Here
22 Boethius suggests a tripartition of the sciences, which apparently recalls that of
23 Aristotle in E 1, but is actually consistent with the interpretation of the
24 Neoplatonic commentators:
25
Nam cum tres sint speculativae partes, naturalis, in motu inabstracta, !mupena¸qetor
26 (considerat enim corporum formas cum materia, quae a corporibus actu separari
27 non possunt: quae corpora in motu sunt, ut cum terra deorsum ignis sursum fertur,
28 habetque motum j forma materiae coniuncta), mathematica, sine motu inabstracta
29 (haec enim formas corporum speculatur sine materia ac per hoc sine motu: quae
30 formae cum in materia sint, ab his separari non possunt), theologica, sine motu
31 commentaries on the Categories or on the Isagoge: see for instance Ammonius, In Isag.,
32 p. 11, line 22–p.12, line 11; Olympiodorus, In Categ., p. 7, lines 30 – 34; Ioannes
33 Philoponus, In Categ., p. 5, lines 1 – 6; Elias, In Isag., p. 27, line 36–p. 28, line 7 and In
34 Categ., p. 115, lines 18 ff.; David, In Isag., p. 57, line 26–p.58, line 25. See also
Simplicius, In Phys., p. 1, line 14–p. 2, line 2. On the Neoplatonic interpretation of
35
Aristotle’s Metaphysics see Kremer, Der Metaphysikbegriff in den Aristoteles-Kommentaren
36 der Ammonius-Schule; O’Meara, Le problme de la mtaphysique dans l’antiquit
37 tardive, pp. 3 – 22; Steel, Theology as First Philosophy. The Neoplatonic Concept of
38 Metaphysics, pp. 3 – 21.
39 10 In addition to the contributions by Luna and O’Meara quoted above [notes 5 and 6] see
also Luna, Syrianus dans la tradition exgtique de la Mtaphysique d’Aristote. I. Syrianus
40
entre Alexandre d’Aphrodise et Asclpius, pp. 301 – 309 and D’Ancona, Syrianus dans la
41 tradition exgtique de la Mtaphysique d’Aristote. II. Antcdents et postrit, pp. 311 –
42 27.
282 Pasquale Porro
1 abstracta atque separabilis (nam Dei substantia et materia et motu caret): in

2 naturalibus igitur rationabiliter, in mathematicis disciplinaliter, in divinis intellec-
3
tualiter versari oportebit …. (Boethius, De trinitate, II, ed. Moreschini, p. 168 – 9,
lines 68 – 80).
4
5 Interestingly enough, the qualifications are here attributed to the sciences
6 themselves, and not to their objects: thus, natural philosophy is in motion, is
7 not abstract and ‘cannot be removed’; mathematics is not in motion and is not
8 abstract; divine science, or theology, is not in motion, is abstract and separable.11
9 The denomination here perfectly fits the thing: metaphysics is, strictly speaking,
10 only theology or divine science. But this text too is puzzling. We find !mu-
11 pena¸qetor instead of the troublesome oq wyqist², but the new term increases, if
12 anything, the confusion, since if its meaning is ‘what exists in matter’ (literally:
13 ‘what cannot be removed or taken away’), it should also be attributed to
14 mathematics, although Boethius does not use it in this case. Furthermore the
15 same term inabstracta means two opposite things, depending on whether it
16 refers to physics or to mathematics: in the first case, it means that the objects of
17 physics are considered together with their matter (considerata cum materia); in
18 the second case, it means that the objects of mathematics are investigated
19 without matter (speculata sine materia). Finally, we can observe that this text
20 does not even mention a possible science of being qua being: there is no room
21 for an ontological science in such a system.
22 Now, the text by Boethius is also crucial because it has been commented by
23 Aquinas (Aquinas was the only great master of the thirteenth century to write
24
commentaries on Boethius’ Opuscula sacra) and, together with the famous
25
prologue to his commentary on Aristotle’s Metaphysics, it is the most significant
26
text in allowing us to understand Aquinas’ position on the issue of the true, or
27
proper, subject of metaphysics.
28
The commentary on the De trinitate is well-known to all readers and
29
scholars of Aquinas due to a textual peculiarity: we possess, at least for part of
30
the work, an autograph version by Aquinas himself, contained in ms. Vat.
31
Lat. 9850, fols 90r–103v. And this manuscript shows a laborious transition from
32
a first redaction, in which Thomas Aquinas still admitted three different kinds
33
of abstraction (corresponding to the three kinds of being mentioned above),12 to
34
a final redaction, in which we find only two degrees of abstraction, while the
35
third degree is replaced by a different operation of the intellect called separatio.
36
It is precisely this distinction between abstraction and separation that is usually
37
38
taken as one of the most distinctive features of Aquinas’ metaphysics: while
39
11 On Boethius’ division of sciences see d’Onofrio, La scala ricamata, pp. 11 – 63 (esp.
40
pp. 52 – 9); Speer, The Hidden Heritage: Boethian Metaphysics and its Medieval
41 Tradition, pp. 163 – 81. See also Merlan, From Platonism to Neoplatonism, esp. c. 3.
42 12 See Thomas de Aquino, Super Boetium De Trinitate, ed. leon., p. 146, app. ad lin. 86.
1 natural philosophy and mathematics, in order to define their own subject,

2 require two different kinds, or degrees, of abstraction (respectively, the
3 abstraction of the universal from the particular, and the abstraction of form
4 from sensible matter), ‘divine science’, in order to define its own subject,
5 requires a different operation, namely separation; and this consists in the
6 negative judgement by which our intellect states that being is free of any
7 dependency upon matter, not only in thought, but also in reality. Aquinas
8 himself bases this distinction on two different operations of the human intellect:
9 the first is the so-called ‘understanding of indivisibles’ (intelligentia indivisibi-
10 lium), by which the intellect knows what a thing is (i. e., the quiddity or nature
11 of a given thing); the second is that by which the intellect joins or divides its
12 own contents, by forming affirmative or negative statements. Thus, the first
13 operation concerns the nature itself of a thing, while the second has to do with
14 the being of a thing (esse).
15
Sic ergo intellectus distinguit unum ab altero aliter et aliter secundum diuersas
16 operationes: quia secundum operationem qua componit et diuidit distinguit unum
17 ab alio per hoc quod intelligit unum alii non inesse, in operatione uero qua intelligit
18 quid est unumquodque, distinguit unum ab alio dum intelligit quid est hoc, nichil
19 intelligendo de alio, neque quod sit cum eo, neque quod sit ab eo separatum; unde
ista distinctio non proprie habet nomen separationis, set prima tantum. Hec autem
20
distinctio recte dicitur abstractio, set tunc tantum quando ea quorum unum sine
21 altero intelligitur sunt simul secundum rem: non enim dicitur animal a lapide
22 abstrai si animal absque intellectu lapidis intelligatur. Vnde cum abstractio non
23 possit esse proprie loquendo nisi coniunctorum in esse, secundum duos modos
24 coniunctionis predictos, scilicet quo pars et totum uniuntur, uel forma et materia,
duplex est abstractio: una qua forma abstraitur a materia, alia qua totum abstraitur a
25
partibus (Thomas de Aquino, Super Boetium De trinitate, q. 5, a. 3, ed. leon.,
26 p. 148, lines 159 – 79).
27
28 There has been a long and heated debate over the correct interpretation of what
29 separation means and implies in Aquinas. Roughly speaking, the doctrinal focus
30 of the debate could be summarised as follows: does separation imply that we
31 must first prove the effective existence of separate substances (God and the
32 angels) in order to show that being, as the proper subject of first philosophy, can
33 be considered apart from matter? Or rather, is it sufficient, for this same aim, to
34 show that matter, or materiality, is not included in the definition of being qua
35 being? If we support the first thesis (maintained by Geiger, for instance),13 we
36
37
38 13 See Geiger, Abstraction et sparation d’aprs S. Thomas in De trinitate q. 5, a. 3, pp. 3 –
39 40. But see also, more or less along the same interpretative line, Robert, La
mtaphysique, science distincte de toute autre discipline philosophique, pp. 206 – 22;
40
Merlan, Abstraction and Metaphysics in St. Thomas’ Summa, pp. 284 – 91; Moreno,
41 The Nature of Metaphysics, pp. 109 – 35; Owens, Metaphysical Separation in Aquinas,
42 pp. 287 – 306; Nordberg, Abstraction and Separation in the Light of the Historical
284 Pasquale Porro
1 emphasise the theological aspect of metaphysics; while, if we endorse the second

2 thesis (suggested by J.F. Wippel and Jan Aertsen, for instance),14 we stress above
3 all the ‘transcendental’ aspect of first philosophy: that is, the proper subject of
4 metaphysics, being qua being, or common being (though we must note that the
5 two expressions do not strictly coincide in Aquinas), can be obtained without a
6 preliminary theological investigation. In other words, we are not obliged to
7 admit God’s existence in order to prove that being can effectively be separate
8 from matter, precisely as in the case of God.
9 But I shan’t deal any longer with this question. Rather, I would like to raise
10 some doubts about the real importance of the doctrine of separatio within
11 Aquinas’ metaphysical project. The way in which Aquinas arrives at the
12 distinction between abstraction and separation is in itself laborious. First of all,
13 we must attend to an oft-neglected detail: the text of the De trinitate used by
14 Aquinas in his commentary presents a slight, yet significant, difference when
15 compared with the text we usually read today. Where Boethius writes that
16 theology is without motion, abstract and separable (sine motu abstracta atque
17 separabilis), the text possessed by Aquinas gives: without motion, abstract and
18 inseparable (sine motu abstracta atque inseparabilis). Thus, in his literal
19 commentary on this section of Boethius’ text, Aquinas is forced to distinguish
20 initially between abstraction and separation (or separability) by referring the
21 first term to the real relation of a thing to matter, and the second to a simple
22 mental relation:
23
24 Theologia, id est tertia pars speculatiue, que dicitur diuina, uel metaphisica, uel
philosophia prima, est sine motu, in quo conuenit cum mathematica et differt a
25
naturali; abstracta, scilicet a materia, atque inseparabilis, per que duo differt a
26 mathematica: res enim diuine sunt secundum esse abstracte a materia et motu, set
27 mathematice inabstracte, sunt autem consideratione separabiles, set res diuine
28 inseparabiles, quia nichil est separabile nisi quod est coniunctum; unde res hdiuinei
29 non sunt secundum considerationem separabiles a materia, set secundum esse
abstracte, res uero mathematice e contrario (Thomas de Aquino, Super Boetium De
30
trinitate, hExpositio capituli secundii, ed. leon., p. 134, line 102–p. 135, line 14 [the
31 italics are mine]).
32
33 According to this tentative and slightly awkward explanation, mathematical
34 objects are not abstract from matter in existence, while they are separable in
35 thought; divine things, on the contrary, are not separable from matter in
36 Roots of Thomas’ Tripartition of the Theoretical Sciences, pp. 144 – 53; Tavuzzi,
37 Aquinas on Resolution in Metaphysics, pp. 199 – 227.
38 14 See Wippel, Metaphysics and separatio according to Thomas Aquinas, pp. 431 – 70;
39 Aertsen, Was heißt Metaphysik bei Thomas von Aquin?, pp. 217 – 39; Aertsen, Medieval
Philosophy and the Transcendentals. The Case of Thomas Aquinas, esp. pp. 113 – 58. For
40
further bibliographical indications and a survey of the debate see Porro, Metafisica e
41 teologia nella divisione delle scienze speculative del Super Boetium De Trinitate, pp. 467 –
42 526.
1 thought, but they are abstract in existence, and therefore are ‘inseparable’, for
2 nothing is separable unless it is conjoined. The use of inseparabilis is interpreted
3 as a kind of lucus a non lucendo, as the grammarians would say: it is impossible,
4 or at any rate meaningless and superfluous, to separate in thought what is
5 already separate in existence. Now, if we compare this passage with the previous
6 one (taken from q. 5, a. 3, which comes later in Aquinas’ commentary), we may
7 note that the terminology has been completely reversed: abstraction can take
8 place only for those things which are conjoined or one in reality (and thus
9 inseparable), while separation is the operation by which we distinguish two
10 things by understanding that one does not exist in the other. In other words,
11 what in the literal commentary is qualified as ‘inseparable’ (divine things) has
12 now become the object of separation.
13 But this is not the only problem. In the first article of the same q. 5,
14 Aquinas specifies that those things which do not depend upon matter for their
15 being, and not only for their being understood, can be subdivided in two
16 classes: there are some beings (or better, some objects of speculation) which
17 never exist in matter, as in the case of God and the angels, and other objects
18 which do not necessarily exist in matter, as in the case of being, substance,
19 quality, potency and act, one and many, and the like. I refer to the same passage
20 quoted at the very beginning of this article:
21
Quedam uero speculabilia sunt que non dependent a materia secundum esse, quia
22 sine materia esse possunt, siue numquam sint in materia, sicut Deus et angelus, siue
23 in quibusdam sint in materia et in quibusdam non, ut substantia, qualitas, ens,
24 potentia, actus, unum et multa, et huiusmodi, de quibus omnibus est theologia, id
25 est scientia diuina (Thomas de Aquino, Super Boetium De trinitate, q. 5, a. 1, ed.
leon., p. 138, lines 154 – 61).
26
27 The distinction is between a ‘positive’ sense of immateriality (God and other
28 separate substances are positively immaterial), and a ‘negative’ or ‘neutral’ sense
29 of immateriality: that is, matter does not figure in the definition of being qua
30 being, and this means, on the one hand, that being can be conceived
31 independently of or prior to matter and, on the other, that being qua being is in
32 itself neither material nor immaterial, for in the first case we would have only
33 corporeal beings, in the second only separate substances. Both classes of
34 immaterial objects are considered, as is clearly stated in the above passage, by
35 one and the same science – the science we have so far called theology. Thus,
36 according to the distinction mentioned above, we should be able to obtain both
37 classes of immaterial objects through the same operation; in other words,
38 through separation. We must remember, however, that separation distinguishes
39 truthfully only what is not united or conjoined in reality. Therefore it seems to
40 apply, properly speaking, to only one of the classes mentioned above, that is, to
41 the class constituted by God and other separate substances, while the other class
42 consists in those objects which can also be in matter or with matter, though
286 Pasquale Porro
1 matter does not figure in their definition. Yet it is precisely the latter indication
2 (i. e. the fact that matter does not belong to the definition of these objects)
3 which shows us how, for this class, abstraction is more suitable than separation.
4 Aquinas explicitly states that we can only abstract what can be conceived
5 independently, whether in reality two things are united or separated. There is no
6 essential dependence between being and matter, and this is precisely what is
7 required, according to Aquinas, in order to abstract something:
8
Si uero unum ab altero non dependeat secundum id quod constituit rationem
9 nature, tunc unum potest ab altero abstrai per intellectum ut sine eo intelligatur non
10 solum si sint separata secundum rem … set etiam si secundum rem coniuncta sint
11 (Thomas de Aquino, Super Boetium De Trinitate, q. 5, a. 3, ed. leon., p. 147, lines
12 147 – 52 [the italics are mine]).
13 Matter does not figure in the definition of being qua being and so, it would be
14 tempting to say, being can be conceived independently of matter; that is, it can
15 be abstracted from matter. Yet, as we have noted, Aquinas insists on saying that
16 the subject of divine science, or metaphysics, is obtained through separation and
17 not through abstraction.
18 Finally, things become more complicated because what has so far been
19 considered as belonging to one and the same science suddenly gives rise, in a. 4
20 of the same q. 5, to two different sciences:
21
22 Sic ergo theologia siue scientia diuina est duplex: una in qua considerantur res
diuine non tamquam subiectum scientie, set tamquam principia subiecti, et talis est
23
theologia quam philosophi prosequntur, que alio nomine metaphisica dicitur; alia
24 uero que ipsas res diuinas considerat propter se ipsas ut subiectum scientie, et hec est
25 theologia que in sacra Scriptura traditur. Vtraque autem est de his que sunt separata
26 a materia et motu secundum esse, set diuersimode, secundum quod dupliciter potest
27 esse aliquid a materia et motu separatum secundum esse: uno modo sic quod de
ratione ipsius rei que separata dicitur sit quod nullo modo in materia et motu esse
28
possit, sicut Deus et angeli dicuntur a materia et motu separati; alio modo sic quod
29 non sit de ratione eius quod sit in materia et motu, set possit esse sine materia et
30 motu quamuis quandoque inueniatur in materia et motu, et sic ens et substantia et
31 potentia et actus sunt separata a materia et motu, quia secundum esse a materia et
32 motu non dependent sicut mathematica dependebant, que numquam nisi in
materia esse possunt quamuis sine materia sensibili possint intelligi. Theologia ergo
33
philosophica determinat de separatis secundo modo sicut de subiectis, de separatis
34 autem primo modo sicut de principiis subiecti; theologia uero sacre Scripture tractat
35 de separatis primo modo sicut de subiectis, quamuis in ea tractentur aliqua que sunt
36 in materia et motu, secundum quod requirit rerum diuinarum manifestatio
37 (Thomas de Aquino, Super Boetium De Trinitate, q. 5, a. 4, ed. leon., p. 154, lines
175 – 206).
38
39 Of course, Aquinas still assumes (as in his later Prologue to his commentary on
40 the Metaphysics) that ‘philosophical theology’ deals with both classes of separate
41 beings (respectively as subject and as principles of its subject), yet, insofar as the
42 two classes of immaterial being now constitute the subjects of two different
1 sciences, the original tri-partition of Aristotle and Boethius appears absolutely

2 inadequate. I shall not dwell upon the consequences of this reduplication, but I
3 shall focus rather on the procedure followed by Aquinas, a procedure that can be
4 summarised in three main steps, or theses:
5 a) We obtain the subject of theology/divine science through a special operation
6 called separation, which gives us access to real immateriality or trans-
7 materiality.
8 b) There are, however, two different kinds, or types, of immateriality: the
9 ‘positive’ immateriality of God and the angels, and the ‘negative’ or ‘neutral’
10 immateriality of being qua being.
11 c) These two different classes of the immaterial objects of speculation give rise
12 to two different sciences: the theology of Sacred Scripture, on the one hand,
13 and ‘philosophical theology’, or metaphysics, on the other.
14
15 Now, what is quite amazing in this threefold articulation is that the doctrine of
16 separation (thesis a), which is considered by many scholars as the keystone of
17 Aquinas’ metaphysics, is perhaps the least proper way both to designate the
18 essential feature of metaphysics (philosophical theology) and to justify the third
19 conclusion (thesis c). Indeed, it would have been much easier to distinguish
20 from the beginning the two classes of immaterial objects, instead of associating
21 them both under the operation of separation. One could even say that in order
22 to reach the third conclusion, it would have been better for Aquinas to maintain
23 three degrees of abstraction, as in the original redaction of the autograph text,
24 for the three speculative sciences of the philosophical tradition, and to appeal to
25 separation only for the fourth non-philosophical science – Christian theology.
26 The hypothesis I would like to suggest is that Aquinas is confronted here
27 with a conflict between what he draws from a given tradition (thesis a – insofar
28 as it postulates a distinction between what is abstract and what is separate – and
29 thesis b) and his own conclusion (thesis c), which is that there are two different
30 theologies – a philosophical theology and a ‘theological’ theology, if one could
31 concede a similar tautology. In other words, I assume that the latter thesis –
32 thesis c – is what Aquinas has mainly at heart, and in order to reach it he is ready
33 to quarrel, at least partly, with his own sources. I would like to verify my
34 hypothesis by showing that theses a and b – contrary to the emphasis placed on
35 separation in the field of Thomistic studies – are not original doctrines by
36 Aquinas. And I would like to investigate their true origin by raising two parallel
37 questions. First, is Aquinas really the only one, or the first one, to make use of
38 the distinction between separation and abstraction? And second, is Aquinas
39 really the only one, or the first one, to distinguish between two different kinds
40 of immateriality or transmateriality?
41
42
288 Pasquale Porro
1 3. Avicenna and the Immateriality of Being

2
3 Luckily enough, the first of these two questions has already received an answer.
4 Claude Lafleur and Joanne Carrier have shown beyond all doubt that the
5 distinction between abstraction and separation was quite common in the
6 writings of the Masters of the Parisian Faculty of the Arts in the first half of the
7 thirteenth century, and especially in the so-called accessus, or introductions, to
8 philosophy composed between 1230 and 1250:
9
notandum est quod entium triplex est genus: quedam sunt entia coniuncta motui et
10
materie sensibili secundum esse et secundum intellectum, abstracta tamen ab
11 indiuiduis, et de talibus est naturalis philosophia; alia sunt que sunt coniuncta
12 motui et materie sensibili secundum esse, abstracta tamen per intellectum, et de
13 talibus est mathematica; alia sunt separata penitus a motu et materia secundum esse
14 et secundum rationem, et de talibus est methaphisica.15
15
Thus we can at least say that in his commentary on Boethius’ De trinitate
16
Aquinas complies with this common usage. But whence do the Masters of the
17
Arts derive this distinction? Following Alain de Libera, Lafleur and Carrier
18
suggest that the remote source of the distinction between abstraction and
19
separation might be al-Fārābı̄, and they refer more precisely to al-Fārābı̄’s tract
20
De intellectu et intelligibili, in which, however, the distinction is not clearly
21
expressed. Yet, in another text by al-Fārābı̄, his short treatise Maqāla … fı̄ aġrād
22 ˙
al-hakı̄m fı̄ kull maqāla min al-kitāb al-mawsūm bi-l-hurūf (Treatise … on the
23 ˙ ˙ 16
24
Goals of the Sage in Each Treatise of the Book Named by Means of Letters), – a
25
text which was not translated into Latin, but the contents of which were
26
nonetheless introduced in the Latin West through Avicenna – is explicitly stated
27
that the objects of metaphysics (or, at least, some of them) ‘have no existence at
28 all (be it imaginary or real) in natural things. It is not that imagination has
29 abstracted them from natural things; rather, their existence and nature [itself ] is
30 abstracted [i.e. immaterial].’17 The terminological distinction between abstrac-
31 tion and separation might thus be the result of the overlapping, in the Latin
32
33 15 [Anon.], De communibus artium liberalium, § 204, pp. 190 – 91, lines 980 – 87. See also
34 Lafleur and Carrier, Abstraction, sparation et tripartition de la philosophie thortique:
quelques lments de l’arrire-fond farabien et artien de Thomas d’Aquin, pp. 248 – 71.
35
16 The text has been edited by F. Dieterici in Alfarabi’s philosophische Abhandlungen,
36 pp. 34 – 8 (another anonymous edition has been published in Hyderabad: Dā irat al-
˘
Ma ārif, 1349H/1930). A complete English translation is now available in Bertolacci,

˘
37
The Reception of Aristotle’s ‘Metaphysics’ in Avicenna’s ‘Kitāb al-Šifā ’, pp. 66 – 72 (but for
˘
38
39 an analysis of the contents of the treatise and of its influence on Avicenna see the whole
c. 3, pp. 65 – 103). A partial English translation was already available in Gutas, Avicenna
40
and the Aristotelian Tradition, pp. 240 – 42. As is well-known, Avicenna refers explicitly
41 to al-Fārābı̄’s treatise in his Autobiography.
42 17 See Bertolacci, The Reception of Aristotle’s ‘Metaphysics’ in Avicenna’s ‘Kitāb al-Šifā ’, p. 69.
˘
1 Scholastic vocabulary, of al-Fārābı̄’s idea that some objects are not only
2 abstracted through ‘imagination’, but ‘totally abstracted’ in their existence and
3 nature, and Aristotle’s use of separation in Metaphysics, E 1.
4 Moreover, for the second of the above-mentioned questions (is Aquinas
5 really the only one, or the first one, to distinguish between two different kinds
6 of immateriality, or transmateriality?), I would like to suggest the same
7 genealogy: indeed, even the distinction between two forms of immateriality is
8 an elaboration by Avicenna of a theme mentioned by al-Fārābı̄ in the passage
9 referred to above:
10
Of the subject-matters of this science, on the other hand, some have no existence at
11 all (be it imaginary or real) in natural things. It is not that imagination has
12 abstracted them from natural things; rather, their existence and nature [itself ] is
13 abstracted [i.e. immaterial]. Others exist in natural things, even though they are
14 imagined as abstracted from them. However, they do not exist in natural things
essentially, i. e. in such a way that their existence is not independent from these and
15
they are things whose subsistence is due to natural things. Rather, they exist both in
16 natural things and in non-natural things (these latter being separate either really or
17 in imagination).
18 Therefore the science which deserves to be called by this name is [only] the
19 present one. It alone, all other sciences excluded, is ‘metaphysics’.18
20 Yet, since the Fı̄ aġrād remained unknown to the Latins, and Avicenna’s
21 ˙
treatment of these doctrines is much deeper and wider than al-Fārābı̄’s, we can
22 consider directly Avicenna’s Metaphysics (Ilāhiyyāt, The Science of Divine Things,
23 i. e. the metaphysical section of The Cure).
24 Briefly, what is the role played by separation in Avicenna’s Metaphysics? In
25 order to arrive at an answer to this question, we should probably turn to book
26 VII of the Ilāhiyyāt, because the latter is exactly the place dedicated to an explicit
27 and severe criticism of the Platonic theory of ideas and the Platonic notion of
28 separation.
29 Now, book VII has always been considered relatively eccentric with regard
30 to the overall structure of Avicenna’s Metaphysics, if not completely out of place.
31 After having established, in book I, the true subject of the divine science, and set
32 forth all the prolegomena required for its treatment, the work follows the
33 programme which is announced in I, 4, and which envisages first a
34 consideration of the ‘quasi’-species of the existent, and then a consideration
35 of its properties. With regard to the latter, Avicenna first deals, in book IV, with
36 the pairs prior/posterior and potency/act, and with the concepts of ‘complete’,
37 ‘incomplete’, ‘whole’ and ‘total’; then, in book V, with the notions of universal
38 and particular; and finally, in book VI, with causality. Still according to the
39
40
41
42 18 See Bertolacci, The Reception of Aristotle’s ‘Metaphysics’ in Avicenna’s ‘Kitāb al-Šifā ’, p. 69.
˘
290 Pasquale Porro
1 programme expounded in I, 4, Avicenna announces that, at this point, it will be

2 necessary to introduce an analysis of the one and the many.
3 And this is, in actual fact, what we find in book VII. But what puzzles the
4 reader is that this discussion of the one and the many interrupts the transition
5 from the consideration of causes in general to the demonstration of their
6 finiteness, and consequently of the existence of a cause which is absolutely first.
7 It would have been more congruous – so it seems – to place the analysis of the
8 one and the many after that regarding the universal and the particular, and
9 before the section dedicated to causality.
10 Moreover, book VII appears to be rather heterogeneous even in itself. As is
11 well-known, the treatise consists of three chapters, or sections. In the first
12 chapter, Avicenna really deals with the one and the many (or unity and
13 multiplicity); in the other two sections, he discusses and rejects the opinions of
14 the Platonists and Pythagoreans, according to which the principles of all things
˘
15 should be found in ideal forms (mutūl) or in numbers (a dād).
¯
16 The second section, in particular, begins with a quite abrupt transition to a
17 different topic:
18
The time has come for us to devote ourselves exclusively to opposing opinions that
19 have been uttered about forms, mathematics, separated principles, and universals
20 that are contrary to our principles which we have established (Avicenna, The
21 Metaphysics of ‘The Healing’, VII, 2, transl. Marmura, p. 243).
22
Again, such programmes refer more to the contents and aims of book V than to
23
the first section of book VII. And in a way it is Avicenna himself who concedes
24
that this discussion has already been partially raised in the earlier parts of his
25
work:
26
27 [We will engage in this] even though, in the correctness of what we have said and
28
the rules we have provided, there is, for the discerning, a directing of attention
toward the resolution of all their doubts and [toward] showing their falsehood and
29 the contradictions of their doctrines. Nonetheless, we will help by undertaking this
30 ourselves because of what we hope will ensue from this [endeavour] by way of
31 benefits that we will mention in the course of combating them – [benefits] which
32 we may have missed [mentioning] in what we had [previously] presented and
33
explained (Avicenna, The Metaphysics of ‘The Healing’, VII, 2, transl. Marmura, p.
243).
34
35 Interestingly enough, Avicenna introduces the issue with a kind of general
36 historiographical remark, which obviously echoes Aristotle’s attitude, but can
37 also be considered as a possible source (though with some important differences)
38 for the famous account of the historical progress of metaphysics given by
39 Thomas Aquinas in his Summa theologiae, I, q. 44, a. 2:
40
Every art has a genesis wherein it is raw and unripe, except that after a while it
41 matures and after some more time, it develops and is perfected. For this reason,
42 philosophy in the early period of the Greek’s occupation with it was rhetorical. It
1 then became mixed with error and dialectical argument. Of its divisions, it was the
2 natural which first attracted the masses. They then began to give attention to the
3
mathematical [division], then to the metaphysical. They were involved in
transitions from one part [of philosophy] to another that were not sound.
4 When they first made the transition from what is apprehended by the senses to
5 what is apprehended by the mind, they became confused. [One] group thought that
6 the division necessitates the existence of two things in each thing – as, for example,
7 two humans in the idea of humanity: a corruptible, sensible human; and an
8 intellectually apprehended, separate, eternal, and changeless human. For each of the
two they assigned an existence. They termed the separable existence ‘exemplary
9
existence’, and for each of the natural things they made a separable form that is
10 intellectually apprehended, being the [very] one that the mind receives, since the
11 intelligible is something that does not undergo corruption, whereas every sensible of
12 these [natural things] is corruptible. They [further] rendered the sciences and
13 demonstrative proofs move in the direction of [the incorruptible intelligibles], these
being the ones they treat (Avicenna, The Metaphysics of ‘The Healing’, VII, 2, transl.
14
Marmura, pp. 243 – 4).
15
16 According to this passage, the first mistake that ancient philosophers made in
17 the transition from sensible to intellectual knowledge was that of considering the
18 intelligibles as separate from sensible things, so that they had to admit two
19 different kinds of existence for each thing. The responsibility for this mistake is
20 clearly and explicitly indicated:
21 It was known that Plato and his teacher, Socrates, went into excess in upholding this
22 view, saying that there belongs to humanity one existing idea in which individuals
23 participate and which continues to exist with their ceasing to exist. This [they held]
24 is not the sensible, multiple, and corruptible meaning and is therefore the
25
intelligible, separable meaning (Avicenna, The Metaphysics of ‘The Healing’, VII, 2,
transl. Marmura, p. 244).
26
27 Another group, Avicenna continues, attributed a separate existence not to
28 intelligible forms, but to mathematical entities: more precisely, they rendered
29 mathematical entities that are separable in definition (magnitudes, shapes and
30 numbers) as deserving to be separable in existence. Avicenna does not mention
31 any particular thinker in this regard, but it is likely that he is referring to the first
32 Academicians, such as Speusippus and Xenocrates. Others – and Avicenna refers
33 here explicitly to the followers of Pythagoras – took mathematical entities as
34 principles, but did not make them separate. The Pythagoreans composed
35 everything from unity and duality: they made unity within the bounds of good
36 and what is restricted (limited), and duality within the bounds of evil and what
37 is unrestricted (unlimited). Still another group, which remains anonymous,
38 considered ‘excess’, ‘defect’ and ‘equality’ as principles.
39 It is evident that, in its structure, book VII of Avicenna’s Ilāhiyyāt
40 corresponds in general to themes which Aristotle deals with in at least six
41 different books of his Metaphysics: A, C, D, I, L, M. As is well-known, a scrupulous
42 and detailed comparison between Aristotle’s Metaphysics and Avicenna’s Ilāhiyyāt
292 Pasquale Porro
1 has been carried out by Amos Bertolacci,19 so that I can confine myself to
2 interpreting his results (as far as book VII is concerned), and to making a couple
3 of additional remarks.
4 1. First, the references to Aristotle are split into two well-defined blocks: C,
5 D, I for section 1; A, L, M for sections 2 – 3. Thus, even the distribution of
6 Aristotelian quotations shows the composite, not to say heterogeneous, structure
7 of the book, which collects, or refers to, completely different parts and
8 exigencies of Aristotle’s Metaphysics. This lack of homogeneity (in a book which
9 is also relatively short) raises the problem of understanding the real purpose of
10 Avicenna’s treatise – a question which is connected to that of its position, and
11 which we shall postpone, as mentioned above, to our conclusions.
12 2. Second, it is interesting to note that – despite the fact that sections 2 – 3
13 are constructed in close reference to Metaphysics A, L, M – when Avicenna
14 comes to describe explicitly the causes of the errors of the Platonists and
15 Pythagoreans, he does not employ the chief arguments elaborated by Aristotle,
16 but suggests five reasons which appear to be completely original, in the sense
17 that they have no parallel in Aristotle’s text. Avicenna hence derives from
18 Aristotle the doctrines he seeks to criticise, and some of the specific objections
19 against the Pythagoreans and Academicians, but the real background of his
20 criticism is rooted in something quite different and original: Avicenna’s doctrine
21 of the indifference of the essences, and above all, his distinction between plain
22 negation and negation by equipollence, or metathesis.20
23 We can thus move quickly to the consideration of the five roots indicated by
24 Avicenna in order to explain the errors of the Pythagoreans and Academicians
25 (‘If you give thought [to this matter], you will find the bases of the cause of
26
error, in all the things wherein these people have gone astray, to be five [in
27
number]’). We can perhaps confine ourselves to the consideration of the first
28
and fifth ‘roots’, and of the corresponding arguments.
29
The first runs as follows:
30
31 One of them is their belief that, if a thing is abstracted such that the consideration
32
of another thing is not connected with it, then it is separated from it in existence. It
is as though, if attention is paid to the thing alone – [a thing] that has an associate –
33 in a manner that gives no attention to its associate, [this] would render it not
34 adjoining its associate. In short, if it is considered without the condition of [its]
35 conjunction with another, then it is believed that it is considered with the condition
36 that there is no conjunction [with another], so that [according to this view] it was
37 only suitable to be examined because it was not conjoined, but separate. For this
reason it was believed that, since the mind attains the intelligibles existing in the
38
world without attending to what is conjoined to them, the mind attains nothing but
39
40
19 See Bertolacci, The Reception of Aristotle’s ‘Metaphysics’ in Avicenna’s ‘Kitāb al-Šifā ’,
˘
41 pp. 309 – 73, esp. p. 366.

42 20 We shall come back soon on the meaning of this distinction; see also below, n. 22.
1 [what is] separate among them. This, however, is not the case. Rather, each thing has
2 one consideration with respect to itself and another consideration with respect to its
3 relatedness to something conjoining it.
If we apprehend the form of the human intellectually – for example, inasmuch
4
as it is the form of the human alone – we would have apprehended intellectually an
5 existent alone with respect to its essence. But, inasmuch as we have [so]
6 apprehended it, it does not follow that it alone is separate [in existence]. For that
7 which is mixed with another, inasmuch as it is itself, is inseparable from [the other]
8 by way of negation, not by way of equipollence, in terms of which separation in
9 subsistence is understood. It is not difficult for us to direct attention through
perception or some other state to one of the two things whose role is not to separate
10
from its companion in subsistence – even though it separates from it in definition,
11 meaning, and reality, since its reality is not entered in the reality of the other. For
12 conjunction necessitates connectedness, not permeation in meanings (Avicenna, The
13 Metaphysics of ‘The Healing’, VII, 2, transl. Marmura, p. 247).21
14
The first error of the Platonists was that of considering what is separable by the
15
mind as necessarily separate in being. The error thus consists in confusing two
16
different notions of separation: a purely eidetic separation, and a real
17
18
ontological separation. But this confusion implies, in turn, a specific logical
19
(and linguistic) mistake, i. e. the confusion between plain negation and negation
20
by equipollence (i. e. metathetic affirmation), or better, between separation by
21
negation and separation by equipollence, or metathesis, as Avicenna himself
22 clearly and explicitly states.
23
24 21 See the text of the Avicenna Latinus, Liber de philosophia prima sive Scientia divina, VII,
2, pp. 363 – 4, lines 00 – 26: ‘Tu autem, cum diligenter consideraveris hoc, invenies quod
25
radices occasionis omnis erroris in quem inciderunt isti viri sunt quinque. Una est opinio
26 eorum quod, cum res est exspoliata ab aliquo nec est adiunctus ei respectus alius,
27 profecto exspoliata est in esse ab eo, quemadmodum si id cui aliquid adiunctum est
28 consideraveris per se sine consideratione eius quod sibi adiunctum est, iam enim
29 considerasti illud non adiunctum illi; et omnino, cum consideraveris illud sine condicione
coniunctionis, iam putabis te considerasse illud cum j condicione non coniunctionis, ita ut
30
non oporteat considerare illud nisi non coniunctum, quamvis sit coniunctum. Sed, quia
31 intellectus apprehendit intellecta quae sunt in mundo sine consideratione eius cui
32 adiunguntur, ideo putaverunt quod intellectus non apprehendit nisi separata ab eis. Non
33 est autem ita; immo omnis res, secundum quod in seipsa est, habet unum respectum, et,
34 secundum quod coniuncta est alii, habet alium respectum. Nos enim cum intelligimus,
verbi gratia, formam hominis inquantum est forma hominis solummodo, iam
35
intelligimus aliquid quod solummodo est secundum quod est in se, sed ex hoc quod
36 intelligimus, non oportet ut sit solum et separatum. Coniunctum enim, ex hoc quod est
37 ipsum, est non separatum secundum modum negationis, non secundum modum privationis
38 qua intelligitur separatio existentiae. Non est autem nobis difficile intelligere per
39 apprehensionem vel per reliquas dispositiones unum ex duobus quorum unum est scilicet
quod non est de natura eius separari a sibi coniuncto in existentia, quamvis separetur ab eo
40
in definitione et intentione et certitudine, cum fuerit eius certitudo non contenta intra
41 certitudinem alterius, quoniam esse cum illo facit debere esse coniunctionem non
42 contineri in intentionibus.’ [The italics are mine].
294 Pasquale Porro
1 The fifth ‘root of error’ concerns more the separation of mathematical

2 principles than the separation of ideal forms:
3
The fifth [error] is their belief that if material things are caused, then their causes are
4 necessarily any of the things that can separate. For it is not the case that, if material
5 things are caused and mathematical things are separate, then it follows necessarily
6 that the mathematical things are their causes. Rather, [their causes] may be other
7 substances that are not among the nine categories. Nor have they ascertained the
core of the truth that the definitions of geometrical [figures] among mathematical
8
[objects] absolutely do not dispense with matter, even though they can do without
9 some kinds of matter (Avicenna, The Metaphysics of ‘The Healing’, VII, 2, transl.
10 Marmura, pp. 248 – 9).
11
This error presupposes the confusion between abstraction and separation, and in
12
this way it is also connected to the distinction between separation by (plain)
13
negation and separation by metathesis, or equipollence. Moreover, there is no
14
need to admit that the causes of material things should necessarily be separate
15
things, whereas it seems rather awkward that an accidental category (quantity)
16
could be the cause of substances. Finally, Pythagoreans and Academicians do not
17
take into account that mathematical entities are not completely devoid of matter
18
(they lack corporeal matter, but they do possess, in Aristotelian terms,
19
intelligible matter).
20
21
22
23
4. The Main Logical Tool of Avicenna’s Criticism Against Platonism:
24
the Distinction Between Plain Negation and Negation by
25 Equipollence (Metathetic Affirmation)
26
27 As we have seen, all five arguments are connected to the doctrine of the
28 indifference of the essences. More precisely, we might add, they refer to the
29 logical tool underlying this doctrine–that is, an accurate analysis of the role
30 which a negation plays in a statement according to its position. In every negative
31 judgement–and this is really a crucial point for Avicenna–the sense changes
32 according to the position of the negation, or (what is really the same thing)
33 according to the purpose (sjopºr) of the negation itself. The negation can refer
34 only to that which is predicated of a thing, or it can refer to the whole
35 predication. In other words (since in classical logic the verb is not usually
36 included in the predicate), the negation can affect either the verb or only the
37 predicate. This distinction is explicitly posed by Avicenna not so much in his
38 Metaphysics, as for instance in a passage of his Remarks and Admonitions (Kitāb
39 al-Išārāt wa l-tanbı̄hāt), which it might be useful to quote:
40
41
42
1 Remark: Concerning equipollence and positiveness.

2 Sometimes the composition consists of a negative particle with another [term],
3
as in the statement, ‘Zayd is non-sighted.’ By ‘non-sighted’ we intend ‘blind’, or a
concept more general than that.
4
In short, if ‘non’ is made as one thing with ‘sighted’, or with what resembles it,
5 and, is, then, affirmed or negated, ‘non’ – as well as any other negative particle [in its
6 place] – is, then, a part of the predicate. Thus if you affirm the whole, that would be
7 an affirmation. And if you negate it, that would be a negation, as when you say,
8 ‘Zayd is not non-sighted.’ […]
9
[In Arabic], if a negative particle precedes the copula, as in the statement, ‘Zayd
laisa huwa bası̄ran’ (Zayd is not sighted), the negation has been applied to the
10 ˙ eliminating it and negating it. But if the copula precedes the
affirmation, thus
11 negative particle, it makes a part of the predicate: And the proposition is an
12 affirmation as in the statement, ‘Zayd huwa ġayr bası̄r’ (Zayd is non-sighted)
(Avicenna, Remarks and Admonitions. Part One: Logic, transl. ˙ Inati, pp. 83 – 4).
13
14
The passage refers to the doctrine of propositions (relying at least indirectly on
15
Aristotle, De int., 10, 19b27 ff.; Anal. Pr., I, 46, 51b36 ff., and perhaps on al-
16
Fārābı̄)22, and in particular to the composite and divided sense of a negation. As
17
already mentioned, Avicenna makes a distinction between plain or simple
18
negation and negation by equipollence (or perhaps, better, metathesis). Plain
19
20
22 On the logical meaning of metathesis in al-Fārābı̄ and its possible Greek sources
21 (presumably, Theophrastus) see al-Fārābı̄, Commentary and Short Treatise on Aristotle’s ‘De
22 interpretatione’, ed. F.W. Zimmermann, esp. p. 98: ‘In languages, however, in which a
23 noun is not predicated of a noun unless they are expressly connected by a hyparctic verb,
24 negation is distinct from metathesis. For if a speaker attaches the negative particle to the
hyparctic verb the result is a plain hnegationi; and if he attaches it to the predicate noun
25
instead, the result is like Zaydun mawjūdun lā ādilan, which is a metathetic affirmation.’
˘
26 See also Zimmermann’s comments in this regards, esp. p. lxiii: ‘The bulk of al-Fārābı̄’s
27 exposition of the bewilderingly ramified body of doctrine surrounding Aristotle’s cryptic
28 remarks at the beginning of section three (ch. 10, 19b19 ff.) conspicuously resembles
29 what may tentatively be called Theophrastus’ theory of metathesis, because Ammonius’
account … would suggest that to Theophrastus belonged not only the term “metathesis”
30
for the shift from “… is P” to “… is not not-P” (or simply from “P” to “not-P”) but also
31 the broader framework of ideas of which it forms a part’. The use of ‘metathesis’ in this
32 particular context is explained by Zimmermann himself in a footnote on the same page,
33 on the basis of Ammonius: ‘ “In arranging the propositions in a diagram …, once the
34 indefinite negation has been placed underneath the plain affirmation, there is nothing
left but to place the indefinite affirmation underneath the plain negation. And this is
35
why Theophrastus has called them [sc. the pair with an indefinite predicate] “metathetic”
36 [1j letah]seyr], for their order is reversed [letat]heitai] in the diagram – or else
37 because the definite is replaced [letatehe_r] by an indefinite predicate” (Ammonius
38 161.24 – 32; cf. Stephanus 40.22 – 5, where the alternative explanation is that the
39 negative particle is transferred [letat_hetai] from the copula to the predicate).
˘
Theophrastus’ 1j letah]seyr entered the Arabic tradition as ma dūl [deviated,

40
deflected], which in deference to the original I have rendered as “metathetic”. The
41
˘
opposite of ma dūl is bası̄t “plain” [tr. "pkoOr].’ On single and composite negative
42 expressions in Avicenna see˙ also S. Inati, Ibn Sina on Single Expressions, pp. 148 – 59.
296 Pasquale Porro
1 negation denies the verb, and thus renders a proposition negative, as in the case:
˘
2 ‘Zayd is not sighted’; whereas negation by metathesis ( udūl: equipollence,
3 according to Inati’s and Marmura’s translations) is that negation which denies
4 the predicate and, in this sense, is equivalent to a (metathetic) affirmation such
5 as ‘Zayd is non-sighted’.23
6 This logical tool – that is, the distinction between plain negation and
7 negation by equipollence (or metathetic affirmation) – is employed with great
8 consistency and efficacy throughout the whole system of Avicenna’s Metaphysics,
9 and represents, in my opinion, one of its most profound, original and
10 innovative features. I shall confine myself here to pointing out just three cases in
11 which negation by equipollence, which implies a position and hence an
12 ontological separation, is contrasted with plain negation, which implies only a
13 form of eidetic separation.
14 1. The first basic example concerns the determination of the subject of
15 metaphysics: the existent, or being, that represents the subject of metaphysics is
16 not im-material in the sense of ‘on the condition of not being material’, but
17 rather in the sense of ‘not on the condition of being material’, in other words,
18 without all the concomitant conditions that accompany it, but that do not enter
19 into the definition of the existent itself: ‘The primary subject matter of this
20 science is, hence, the existent inasmuch as it is an existent; and the things sought
21 after in [this science] are those that accompany [the existent], inasmuch as it is
22 an existent, unconditionally.’ (I, 2; transl. Marmura, p. 10).24 As is wellknown,
23 Avicenna enumerates many different senses of immateriality, while always
24 maintaining a clear distinction between immateriality by equipollence or
25 metathesis (separation which posits something, hence ‘positive’ immateriality –
26 the immateriality which belongs to God and the separate substances) and
27 immateriality by plain or simple negation, which only implies an eidetic
28 separation (matter is not included in the intention of being qua being).
29
2. A second example is offered by the very same doctrine of the indifference
30
of the essences: the indifferent essence is not a separate entity (a third entity
31
with respect to the concept in the mind or the individuals in the physical
32
world); it is indifferent only because it can be considered independently of the
33
conditions that always, unavoidably, accompany it. Here again the distinction is
34
35
23 Likewise, in his Kitāb al-Naǧat, Avicenna explains that the difference between a ‘plain’
36
˘
proposition (bası̄ta) and a metathetic, or equivalent, proposition (ma dūla) lies in the fact
37 ˙ the negation is part of the predicate, so that, for instance, ‘non-
that, in the second,
38 seeing’ or ‘non-sighted’ is equivalent to ‘blind’; see Ibn Sı̄nā, Kitāb al-Naǧat, ed.
39 M. Fahrı̄, pp. 54 – 5.
˘ Latinus, Liber de philosophia prima sive Scientia divina, I, 2, p. 13, lines 36 – 8:
24 Avicenna
40
‘Ideo primum subiectum huius scientiae est ens, inquantum est ens; et ea quae inquirit
41 sunt consequentia ens, inquantum est ens, sine condicione.’ (The English translation in
42 the text is that of Marmura).
1 between essence taken on the condition of not being associated with its
2 concomitants (the ontological separation of Platonic forms) and essence taken
3 not on the condition of being associated with its concomitants (eidetic
4 separation). In the Avicenna Latinus (i. e. in the Latin version of the Ilāhiyyāt),
5 this doctrine is expressed through an extremely precise and technical phrasing:
6 to take something cum condicione non rei alterius, ‘on the condition of no other
7 thing’, is quite different from taking the same thing non cum condicione rei
8 alterius, ‘not on the condition of any other thing’. The first formula (to take
9 something ‘on the condition that no other thing is added to it’) leads to a
10 separate existence; the second formula (to take something ‘not on the condition
11 of a possible addition’) leads only to the possibility of considering something
12 apart from the different and even innumerable conditions which might
13 accompany it in reality.25
14
This is precisely what we find in the explication of the first reason for the
15
error of the Platonists: ‘In short, if it is considered without the condition of [its]
16
conjunction with another, then it is believed that it is considered with the
17
condition that there is no conjunction [with another], so that [according to this
18
view] it was only suitable to be examined because it was not conjoined, but
19
separate.’ (VII, 2; transl. Marmura, p. 247).
20
3. Finally – and this is the only case in which, according to Avicenna, it is
21
legitimate to appeal to negation by equipollence, or ontological separation – the
22
23
same distinction is employed in book VIII to point out the difference between
24
God (the Necessary Existent) and common being:
25 The First, hence, has no quiddity. Those things possessing quiddity have existence
26 emanate on them from Him. He is pure existence with the condition of negating
27 privation and all other description of Him. … The meaning of my statement, ‘He is
28
29 25 Avicenna Latinus, Liber de philosophia prima sive Scientia divina, V, 1, p. 236, line
92–p. 7, line 10: ‘Hic est autem quiddam quod debet intelligi, scilicet quia verum est
30
dicere quod de animali, ex hoc quod est animal, non debet praedicari proprietas nec
31 communitas, nec est verum dicere quod de animali, ex hoc quod est animal, debet non
32 praedicari proprietas vel communitas; scilicet nam si animalitas faceret debere non
33 praedicari de eo proprietatem vel communitatem, tunc nec esset animal proprium nec
34 esset animal commune; et secundum hoc debes intelligere magnam esse distantiam inter
illa, et ob hoc etiam interest an dicatur quod animal, ex hoc quod est animal per se, sine
35
condicione alterius, et an dicatur quod animal, ex hoc quod est animal per se, cum
36 condicione non rei alterius. Si enim concederetur quod animal, ex hoc quod est animal per
37 se, esset cum condicione quod non haberet esse in sensibilibus istis, non tamen
38 concederetur quod platonitas esset in sensibilibus istis; esse enim animalis cum
39 condicione non rei alterius in intellectu tantum est; animal vero per se, non cum
condicione rei alterius, habet esse in sensibilibus. Ipsum vero in se in veritate sua est sine
40
condicione alterius rei, quamvis sit cum mille condicionibus quae adiunguntur ei
41 extrinsecus. Animal ergo per se ex sua animalitate habet esse in istis sensibilibus; hoc
42 autem non facit debere ipsum esse separatum per se.’ [The italics are mine].
298 Pasquale Porro
1 pure existence with the condition of negating all other additional [attributes] of
2 Him,’ is not that this is the absolute existence in which there is participation [by
3
others]. If there is an existent with this description, it would not be the pure existent
with the condition of negation, but the existent without the condition of positive
4 affirmation. I mean, regarding the First, that He is the existent with the condition
5 that there is no additional composition, whereas this other is the existent without
6 the condition of [this] addition (Avicenna, The Metaphysics of ‘The Healing’, VIII, 4,
7 transl. Marmura, pp. 276 – 7).
8
The passage concerns how we should conceive divine simplicity in comparison
9
with that of common being, yet it also implies, very clearly and precisely, the
10
difference between a ‘positive’ and a ‘negative’ meaning of immateriality (or
11
separation). The Avicenna Latinus has a very interesting formulation here: God
12
is ‘naked being not on the condition of affirming [something of it]’ – esse
13
exspoliatum non condicione affirmandi – in other words, God is pure or
14
indeterminate being ‘on the condition of not adding a composition’ – cum
15
condicione non addendi compositionem. Common being is indeterminate being
16
‘on the condition of denying [something of it]’ – condicione negandi – in other
17
words, ‘not on the condition of being considered together with those
18
dispositions that accompany it’ – non condicione additionis. 26 This text has
19
many parallel passages in Thomas Aquinas.27 Just to mention one example:
20
21 dicendum quod aliquid cui non fit additio potest intelligi dupliciter. Uno modo, ut
22 de ratione eius sit quod non fiat ei additio; sicut de ratione animalis irrationalis est,
ut sit sine ratione. Alio modo intelligitur aliquid cui non fit additio, quia non est de
23
ratione eius quod sibi fiat additio: sicut animal commune est sine ratione, quia non
24 est de ratione animalis communis ut habeat rationem; sed nec de ratione eius est ut
25 careat ratione. Primo igitur modo, esse sine additione, est esse divinum: secundo
26 modo, esse sine additione, est esse commune (Thomas de Aquino, Summa theol., I,
27 q. 3, a. 4, ad 1).
28 To sum up, Avicenna seems to employ plain or simple negation to indicate the
29 immateriality of being (example 1) and the indifference of the essences (example
30
31 26 Avicenna Latinus, Liber de philosophia prima sive Scientia divina, VIII, 4, p. 402, lines
32 48 – 60: ‘Primus igitur non habet quidditatem, sed super habentia quidditates fluit esse
33 ab eo; ipse igitur est esse exspoliatum, condicione negandi privationes et ceteras
34 proprietates ab eo. Deinde cetera alia quae habent quidditates sunt possibilia, quia
habent esse per ipsum. Intentio autem de hoc quod dicimus quod ipse est esse
35
exspoliatum condicione negandi ceteras additiones ab eo, non est quod ipse sit esse
36 exspoliatum in quo communicet aliquid aliud esse, si fuerit esse cuius haec sit proprietas:
37 ipse enim non est illud ens exspoliatum condicione negandi, sed est ens non condicione
38 affirmandi, scilicet de primo, quod est ens cum condicione non addendi compositionem,
39 sed hoc aliud est ens non condicione additionis, et, quia illud fuit universale quod
praedicatur de omni re, istud vero non praedicatur de eo in quo est additio, ideo in omni
40
quod est praeter illud est additio.’
41 27 See for instance Thomas de Aquino, De potentia, q. 7, a. 2, ad 4 and ad 6; C. gent. I,
42 c. 26.
1 2), which represent two cases of eidetic separation, and negation by equipollence
2 or metathesis to indicate the immateriality of God and the separate substances
3 (example 3), which represents a case of ‘positive’, ontological separation. The
4 distinction adopted by the Latin Scholastic masters between abstraction and
5 separation has thus conceptually (though not linguistically) a specific origin: the
6 anti-Platonic polemic expounded by Avicenna in his theory of the indifference
7 of the essences, especially in the seventh book of his Metaphysics.
8 Now, that this is the actual doctrinal context and the actual function of the
9 distinction is indirectly, and surprisingly, confirmed by Aquinas himself in the
10 conclusive passage of the responsio of q. 5, a. 3 of his commentary on Boethius’
11 De trinitate – a passage which interpreters often neglect to consider:
12
Et quia quidam non intellexerunt differentiam duarum ultimarum [scil. distinctio-
13
num] a prima, inciderunt in errorem ut ponerent mathematica et uniuersalia a
14 sensibilibus separata, ut Pittagoras et Platonici (Thomas de Aquino, Super Boetium
15 De Trinitate, q. 5, a. 3, ed. leon., p. 149, lines 287 – 90).
16
17
This remark is indeed crucial, and not only because it proves that Avicenna’s
18
Metaphysics is the real source of the distinction between abstraction and
19
separation (the three distinctions mentioned in the text). What scholars often
20 fail to point out is that the context in which Aquinas proposes this distinction
21 has to do with the epistemological status of mathematics, and not of
22 metaphysics or theology. Therefore, the real and primary aim of the distinction
23 is not that of affirming that some beings (God and the angels) are absolutely
24 separate from matter (which is rather obvious), but, the exact opposite, that of
25 denying that other beings (i. e. mathematical objects and universals) are
26 effectively separate, as happens in all forms of Platonism or ontological realism.
27
28
29 5. Thomas Aquinas and the Rethinking of Avicenna’s Legacy
30
31 Thus, the first two of Aquinas’ theses mentioned above – a and b – are attested
32 in Avicenna. Yet, we immediately detect a clear difference between Aquinas and
33 Avicenna: the distinction between two classes of immaterial beings does not
34 imply, in Avicenna, any kind of split within the field of divine science; on the
35 contrary, in Avicenna’s Metaphysics we find a gradual transition from what is
36 taken ‘not on the condition of being material’ to what can be taken only ‘on the
37 condition of not being material’. Or, as we can rephrase it: we find a gradual
38 transition from what is abstract from matter to what is effectively separate. This
39 is possible, however, only because Avicenna assumes that the proper subject of
40 metaphysics – being qua being – is prior to the division between cause and
41 effect, or between God and creatures. For Avicenna, God is not the cause of the
42 subject of first philosophy; rather, God is part of the subject of metaphysics
300 Pasquale Porro
1 (being qua being), and He is the cause of another part of the same subject
2 (created being). For Aquinas, on the contrary, God enters philosophical
3 theology, or metaphysics, as principle and cause of the subject of this science,
4 which in Aquinas’ view is common being qua created being.
5 We can now return to our initial question: why did Aquinas employ
6 separation in order to define metaphysics? The final matter to which we need to
7 attend is the fact that in Aquinas’ later commentary on the Metaphysics the
8 distinction between abstraction and separation seems to be weakened, and the
9 adjective separata is also used to refer to the objects of mathematics:
10
Sed tamen hoc est manifestum, quod scientia mathematica speculatur quaedam
11 inquantum sunt immobilia et inquantum sunt separata a materia sensibili, licet
12 secundum esse non sint immobilia vel separabilia. … In hoc ergo differt
13 mathematica a physica, quia physica considerat ea quorum definitiones sunt cum
14 materia sensibili. Et ideo considerat non separata, inquantum sunt non separata.
15
Mathematica vero considerat ea, quorum definitiones sunt sine materia sensibili. Et
ideo, etsi sunt non separata ea quae considerat, tamen considerat ea inquantum sunt
16 separata. …
17 Physica enim est circa inseparabilia et mobilia, et mathematica quaedam circa
18 [Al. circa quaedam] immobilia, quae tamen non sunt separata a materia secundum
19 esse, sed solum secundum rationem, secundum vero esse sunt in materia sensibili
20
(Thomas de Aquino, In Metaph., VI, lect. 1, eds Cathala, Spiazzi, p. 297, §§ 1161
and 1163 [The italics are mine]).
21
22 Moreover, in the Prologue to his commentary, Aquinas refers implicitly to
23 different degrees of separation (and not of abstraction): the expression maxime
24 separata implies that there might be beings more or less separated (though in
25 connection with physics and mathematics the verb abstrahhio is still used). How
26 should we read this attenuation? I believe that an answer can be given on two
27 different levels. On the one hand, we could assume that Aquinas no longer uses
28 the term ‘separation’ in its narrow, or strict, sense precisely because he realises
29 progressively that separation is inessential, or superfluous, to define a science
30 which has to do with two different classes of immaterial being and which,
31 moreover, finds its proper subject, strictly speaking, only in what is ‘negatively’
32 immaterial. On the other hand, while commenting on Aristotle, Aquinas has no
33 need to distinguish between two theologies, but can confine himself to the level
34 of mere natural, or philosophical, theology.
35 The case of the commentary on Boethius’ De trinitate is quite different for
36 at least two reasons: 1. the De trinitate is a theological, not a philosophical,
37 treatise; 2. the text is not just in this case a literal commentary, but also a
38 commentary per modum quaestionis, which gives Aquinas the opportunity to
39 discuss the issues he prefers. In this context, Aquinas has a very precise aim: that
40 of making room for a new science in the system of theoretical sciences –
41 theologia nostra or theologia sacrae scripturae. If this project entails some
42 ambiguities, it is because Aquinas avails himself of Avicennian doctrines in order
1 to reach a conclusion that has nothing to do with Avicenna. Actually, Aquinas is

2 well aware of the real solution adopted by Avicenna – that of applying to
3 metaphysics a procedure very similar to that used for mathematics:
4
Advertendum est autem, quod licet ad considerationem primae philosophiae
5 pertineant ea quae sunt separata secundum esse et rationem a materia et motu, non
6 tamen solum ea; sed etiam de sensibilibus, inquantum sunt entia, Philosophus
7 perscrutatur. Nisi forte dicamus, ut Avicenna dicit, quod huiusmodi communia de
8 quibus haec scientia perscrutatur, dicuntur separata secundum esse, non quia semper
9 sint sine materia; sed quia non de necessitate habent esse in materia, sicut
10
mathematica (Thomas de Aquino, In Metaph., VI, lect. 1, eds Cathala and Spiazzi,
p. 298, § 1165).
11
12 Yet the fact is that Aquinas needs separation, in its strict sense, not in order to
13 define the subject of metaphysics as philosophical theology, but, outside the
14 latter, in order to define the subject of a new science. Consequently, Aquinas
15 draws from Avicenna (and from the Masters of the Arts) the notion of
16 separation, but transposes it to a different place (moreover, to a place Aquinas
17 himself recognises as improper): not the context of the confutation of
18 Platonism, but that of the discussion of the subject of metaphysics. Aquinas
19 thus applies to the case of metaphysics as a whole what was originally destined
20 only for its theological part, and what he finally applies to a science other than
21 philosophy.
22 The result is somehow paradoxical: by postulating two different sciences
23 which both concern what is separate, Aquinas divides what in Avicenna was
24 united (metaphysics as a single science that, starting from the negatively
25 immaterial, comes to demonstrate the positively immaterial), and unites what in
26
Avicenna must remain distinct (separation and abstraction: the former is that
27
which is ‘on the condition of no addition’; the latter, that which is ‘not on the
28
condition of an addition’). The stake of this doctrinal game is, in my view, clear:
29
Aquinas intends to broaden and force a traditional tripartition which was by
30
then inadequate to contain Christian theology as a distinct speculative science.
31
To summarise: by seeking to distinguish separation from abstraction,
32
Aquinas upsets the delicate balance obtained by Avicenna (from al-Fārābı̄) in his
33
rethinking of Aristotle’s Metaphysics; this conscious choice by Aquinas implies
34
35
some ambiguities, as I have tried to point out, but it also offers some undeniable
36
advantages: it allows theology to constitute itself as a science independent of
37
philosophy, while at the same time it allows metaphysics to remain ‘mere’
38
natural theology, without any undue or immediate admixture with revealed
39 doctrines. Thus, there would be no need to make room in metaphysics, as a
40 philosophical science of being qua being, for an eschatology, a prophetology, nor
41 for a theory of sexual ethics according to revealed laws, as we find in the last two
42 books of Avicenna’s Metaphysics.
302 Pasquale Porro
1 Yet in order to obtain these results, Aquinas loses along the way one of the
2 most sophisticated tools elaborated by Avicenna in order to make sense of
3 Aristotle’s Metaphysics: that is, the distinction between plain negation and
4 negation by metathesis; in other words, the distinction between ontological
5 separation and what we may call eidetic, or intentional, separation.
6 On the other hand, the issue of separation also casts a different light on the
7 position and purpose of book VII within Avicenna’s Metaphysics – a position
8 which has always puzzled his interpreters and readers, for it interrupts the
9 transition from the general consideration of causes to the consideration of the
10 First Cause (the only real ontological cause). A possible solution has been
11 suggested by Amos Bertolacci in his rethinking of the structure of Avicenna’s
12 Metaphysics. According to Bertolacci, ontology in Avicenna’s Ilāhiyyāt always
13 intersects with henology, i. e. with a discussion of species and the properties of
14 the one and the many. In other words, the concrete structure of Avicenna’s
15 Metaphysics is a result of the coexistence and interaction of two axes, one
16 fundamental, the other complementary. The main axis concerns the existent, or
17 being, from four different perspectives: its role as the subject of metaphysics, its
18 species, its properties and its causes. The second axis regards the one and the
19 many, whose species and properties are also taken into account; accordingly, it
20 encompasses a henology of species and a henology of properties.
21 It might be useful to quote Bertolacci himself here: ‘the core of metaphysics
22
is the result of two vertical axes (dealing respectively with the existent, and the
23
one and the many), intersected by four horizontal lines (subject-matter, species,
24
properties, causes). In the actual structure of Avicenna’s Metaphysics these two
25
axes coexist and partially overlap’.28 This overlapping concerns especially the
26
treatment of the species of the existent and the treatment of the species of one
27
and many in book III. Book VII, on the contrary, should not be considered as a
28
case of overlapping, since it would correspond to the section devoted to the
29
henology of properties. In this sense, chapters 2 – 3 of book VII represent, for
30
Bertolacci, ‘a sort of complement to the proof of the accidentality of quantity
31
provided in chapters III, 3 – 5’, and therefore ‘a sort of pars destruens of the
32
treatment of “one” and “many”’.29
33
I think that this is fundamentally correct, and yet it does not fully explain
34
why book VII starts with a discussion of the one and the many, and their
35
properties, and then turns into a more radical criticism of the notion of
36
separation (also independently of quantity). One might say that, when speaking
37
38
about numbers, it makes perfect sense to show that they do not possess a
39
28 See Bertolacci, The Reception of Aristotle’s ‘Metaphysics’ in Avicenna’s ‘Kitāb al-Šifā ’, p. 209.
˘ ˘
40
29 See Bertolacci, The Reception of Aristotle’s ‘Metaphysics’ in Avicenna’s ‘Kitāb al-Šifā ’, p. 178.
41 See also Marmura, Avicenna’s Critique of Platonists in Book VII, Chapter 2 of the
42 Metaphysics of his Healing, pp. 355 – 70.
1 separate existence. This is true, but why then also discuss here Plato’s theory of
2 ideas? We would have expected such a criticism in book V. Well, my hypothesis
3 is that this discussion is neither casual nor accidental. On the contrary, what
4 Avicenna needs in the transition from the general account of causality to the
5 theological part of his Metaphysics is precisely a clear distinction between
6 separation by (plain) negation and separation by metathesis (or equipollence),
7 i.e. – as we may rephrase it – between eidetic separation and ontological
8 separation. In other words, in order to approach the theological part, Avicenna
9 has to take leave, once and for all, of Platonism: it is only by pointing out the
10 ‘roots’ of the errors of the Platonists (and Pythagoreans) that Avicenna can
11 finally move from what is separate by plain negation to what is separate by
12 equipollence, thus establishing a consistent route from the subject of
13 metaphysics to its end, or purpose. Book VII is therefore certainly a treatise
14
on the one and the many, but at the same time it is also a treatise on separation.
15
The latter is absolutely essential in order to clarify both that metaphysics is a
16
science of that which is separate (as Aristotle states in E 1) and that ‘separate’
17
may be understood in two different ways: separate is both that which is taken on
18
the condition of not being associated with matter, and that which is not taken
19
on the condition of being associated with matter. Without this distinction, the
20
unity of first philosophy as onto-theology remains uncertain and problematic.
21
Instead of being out of place or marginal, book VII would thus serve as the
22
23
essential link in keeping together that which in Aristotle (or, at least, in
24
Aristotle’s Metaphysics as we know it) remained merely juxtaposed: metaphysics
25
as a science of being qua being and metaphysics as a divine science dealing with
26
immovable and separate realities.
27
28
29
Bibliography
30
31
Primary Sources
32
Alexander Aphrodisiensis, In Aristotelis Metaphysica commentaria, CAG I, ed. M. Hay-
33
duck, Berlin: Reimer, 1891.
34 Ammonius, In Porphyrii Isagogen sive V voces [= In Isag.], CAG IV/3, ed. A. Busse,
35 Berlin: Reimer, 1891.
36 –– , In Aristotelis De interpretatione commentarius, CAG IV/5, ed. A. Busse, Berlin:
37 Reimer, 1897.
38 [Anon.], De communibus artium liberalium, § 204, eds C. Lafleur, J. Carrier, in: Un
instrument de rvision destin aux candidats la licence de la Facult des arts de
39
Paris, le “De communibus artium liberalium” (vers 1250?), Documenti e studi sulla
40 tradizione filosofica medievale, 5, 1994, pp. 129 – 203.
41 Aristotle, Opera, ed. I. Bekker (Academia Regia Borussica), Berlin: Reimer, 1831 – 70, 5
42 vols; repr. Berlin: De Gruyter, 1960 – 61.
304 Pasquale Porro
1 –– , Die Metaphysik [= Metaph.], ed. A. Schwegler, Tbingen: Fues, 1847 – 48, repr.
2 Frankfurt a. M.: Minerva, 1960 and 1968.
3
–– , Metaphysics [= Metaph.], ed. W.D. Ross, Oxford: Clarendon Press, 1924.
–– , Metaphysica [= Metaph.], ed. W. Jaeger, Oxford: Clarendon Press, 1957.
4
–– , Metaphysics [= Metaph.], transl. W.D. Ross, in The Complete Works of Aristotle. The
5 Revised Oxford Translation, ed. J. Barnes, Princeton: University Press, 1995 (6th
6 printing, with corrections).
7 Asclepius, In Aristotelis Metaphysicorum libros A–F commentaria [= In Metaph.], CAG
8 VI/2, ed. M. Hayduck, Berlin: Reimer, 1888.
9
Avicenna [Ibn Sı̄nā], The Metaphysics of ‘The Healing’, transl. and ed. M.E. Marmura,
Provo, Utah: Brigham Young University Press, 2005.
10
–– , Kitāb al-Naǧat, ed. M. Fahrı̄, Beirut: Dār al-afāq al-ǧadı̄da, 1985.
11 –– , Remarks and Admonitions.˘ Part One: Logic, transl. and ed. S.C. Inati, Toronto:
12 Pontifical Institute of Mediaeval Studies, 1984.
13 –– , Liber de philosophia prima sive Scientia divina, I–IV, ed. S. Van Riet, Louvain/
15
–– , Liber de philosophia prima sive Scientia divina, V–X, ed. S. Van Riet, Louvain/
16
Boethius, De trinitate, II, ed. C. Moreschini, in id., De consolatione philosophiae –
17 Opuscula theologica, Mnchen/Leipzig: K.G. Saur, 2000.
18 David, Prolegomena et in Porphyrii Isagogen commentarium [= In Isag.], CAG XVIII/2,
19 ed. A. Busse, Berlin: Reimer 1904.
20 Elias, In Porphyrii Isagogen et Aristotelis Categorias commentaria [= In Isag., In Categ.],
21
CAG XVIII/1, ed. A. Busse, Berlin: Reimer 1900.
al-Fārābı̄, Commentary and Short Treatise on Aristotle’s ‘De interpretatione’, transl. and ed.
22
F.W. Zimmermann, London: Oxford University Press, 1981.
23 –– , Maqāla […] fı̄ aġrād al-hakı̄m fı̄ kull maqāla min al-kitāb al-mawsūm bi-l-hurūf, in
24 ˙ ˙ Abhandlungen, ed. F. Dieterici, Leiden: Brill,˙ 1890,
Alfarabi’s philosophische
25 pp. 34 – 8.
˘
–– , Maqāla fı̄ aġrād mā ba da l-tabı̄ a, anonymous edition, Hyderabad: Dā irat al-
˘
26 ˙ ˙
˘
27
Ma ārif, 1349H/1930.
Ioannes Philoponus, In Aristotelis Categorias commentarium, CAG XIII/1, ed. A. Busse,
28 Berlin: Reimer 1898.
29 Olympiodorus, Prolegomena et in Categorias commentarium [= In Categ.], CAG XII/1,
30 ed. A. Busse, Berlin: Reimer 1902.
31 Simplicius, In Aristotelis Physicorum libros quattuor priores commentaria [= In Phys.],
32 CAG IX, ed. H. Diels, Berlin: Reimer 1882.
33
Stephanus, In librum Aristotelis De interpretatione commentarium, ed. M. Hayduck,
Berlin: Reimer, 1885.
34 Syrianus, In Metaphysica commentaria [= In Metaph.], CAG VI/1, ed. W. Kroll, Berlin:
35 Reimer, 1902.
36 Thomas de Aquino, Commentary on Aristotle’s ‘Physics’, transls and eds R. Blackwell,
37 R. Spath, W. Thirlkel, Notre Dame, Indiana: Dumb Ox Books: 1999.
38 –– , Commentary on Aristotle’s ‘Metaphysics’, transl. and ed. J. Rowan, Notre Dame,
Indiana: Dumb Ox Books, 1995.
39
–– , In duodecim libros Metaphysicorum Aristotelis Expositio [= In Metaph.], eds
40 M.–R. Cathala, R. Spiazzi, Turin/Rome: Marietti, 1950.
41 –– , In octo libros Physicorum Aristotelis Expositio [= In Phys.], ed. P. Maggiolo, Turin/
42 Rome: Marietti, 1965.
1 –– , Liber de veritate catholicae Fidei contra errores infidelium seu Summa contra gentiles,
2 eds P. Marc, C. Pera, P. Caramello, t. I: introductio; t. II–III: textus, Turin/Rome:
3
Marietti, 1961 – 67.
–– , Quaestiones disputatae de potentia, ed. P.M. Pession, in Quaestiones disputatae, Turin/
4 Rome: Marietti, 196510, t. 2, pp. 1 – 276.
5 –– , Summa contra gentiles, eds Fratres Praedicatorum [= ed. leon.], Sancti Thomae
6 Aquinatis opera omnia, vols 13 – 15, Rome: Commissio Leonina, 1918/1926/1930.
7 –– , Summa theologiae [= Summa theol.], ed. P. Caramello [cum textu ex recensione
8
leonina], Turin/Rome: Marietti, 1948.
–– , Super Boetium De trinitate, eds Fratres Praedicatorum [= ed. leon.], Sancti Thomae
9 Aquinatis Opera Omnia, vol. 50, Rome/Paris: Commissio Leonina/ditions du
10 Cerf, 1992.
11
12
13
Secondary Sources
14
J.A. Aertsen, Medieval Philosophy and the Transcendentals. The Case of Thomas Aquinas,
15 Studien und Texte zur Geistesgeschichte des Mittelalters 52, Leiden/New York/
16 Cologne: Brill, 1996.
17 –– , Was heißt Metaphysik bei Thomas von Aquin?, in Scientia und ars in Hoch- und
18 Sptmittelalter, eds I. Craemer-Ruegenberg, A. Speer, Miscellanea Mediaevalia 22/1,
19
Berlin/New York: De Gruyter, 1994, pp. 217 – 39.
O. Apelt, Beitrge zur Geschichte der griechischen Philosophie, Leipzig: Teubner, 1891,
20 reprinted Aalen: Scientia, 1975.
21 H.H. Berger, ‘Wegnemen’ en ‘scheiden’: Aristoteles’ verdeling van den theoretische
22 wetenschappen, Tijdschrift voor Filosofie, 25, 1963, pp. 207 – 78.
A. Bertolacci, The Reception of Aristotle’s ‘Metaphysics’ in Avicenna’s ‘Kitāb al-Šifā ’. A
˘
23
24
Milestone of Western Metaphysical Thought, Islamic Philosophy, Theology and
Science 63, Leiden/Boston: Brill, 2006.
25 J.J. Cleary, Aristotle & Mathematics. Aporetic Method in Cosmology & Metaphysics,
26 Philosophia antiqua 67, Leiden/New York/Cologne: Brill, 1995.
27 –– , Emending Aristotle’s Division of the Theoretical Sciences, Review of Metaphysics,
28 48, 1994, pp. 33 – 70.
29
D.R. Cousin, A Note on the Text of Metaphysics 1026a14, Mind, 49, 1940, pp. 495 – 6.
C.D’Ancona, Syrianus dans la tradition exgtique de la Mtaphysique d’Aristote. II.
30 Antcdents et postrit, in Le commentaire entre tradition et innovation, ed. M.-O.
31 Goulet-Caz, Paris: Vrin, 2000, pp. 311 – 27.
32 V. Dcarie, La physique porte-t-elle sur des ‘non-spars’ ( !w¾qista)?, Revue des Sciences
33 philosophiques et thologiques, 38, 1954, pp. 466 – 8.
34 G. d’Onofrio, La scala ricamata. La philosophiae divisio di Severino Boezio, tra essere e
conoscere, in G. d’Onofrio, ed., La divisione della filosofia e le sue ragioni. Atti del
35
VII Convegno della SISPM (Assisi, 14 – 15 novembre 1997), Quaderni di Schola
36 Salernitana, Cava de’ Tirreni: Avagliano 2001, pp. 11 – 63.
37 G. Fine, Separation, Oxford Studies in Ancient Philosophy, 2, 1984, pp. 31 – 87.
38 –– , Separation: A Reply to Morrison, Oxford Studies in Ancient Philosophy, 3, 1985,
39 pp. 159 – 65.
L.-B. Geiger, Abstraction et sparation d’aprs S. Thomas in De Trinitate q. 5, a. 3,
40
Revue des Sciences philosophiques et thologiques, 31, 1947, pp. 3 – 40; reprinted in
41 id., Philosophie et spiritualit, Paris: ditions du Cerf, 1963, I, pp. 87 – 124, and
42 more recently in id., Penser avec Thomas d’Aquin. tudes thomistes prsentes par
306 Pasquale Porro
1 Ruedi Imbach, Vestigia, 26, Paris/Fribourg: ditions du Cerf/ditions Universitaires

2 de Fribourg, 2000, pp. 139 – 83.
3
D. Gutas, Avicenna and the Aristotelian Tradition. Introduction to Reading Avicenna’s
Philosophical Works, Leiden/New York/Copenhagen/Cologne: Brill, 1988.
4 E. Halper, One and Many in Aristotle’s Metaphysics, Columbus: Ohio State University
5 Press, 1989.
6 S.C. Inati, Ibn Sina on Single Expressions, in M. E. Marmura, ed., Islamic Theology and
7 Philosophy. Studies in Honor of George F. Hourani, Albany: State University of New
8
York Press, 1984, pp. 148 – 59.
W. Jaeger, Aristoteles. Grundlegung einer Geschichte seiner Entwicklung, 2nd revised ed.,
9 Berlin: Weidmann, 1955.
10 K. Kremer, Der Metaphysikbegriff in den Aristoteles-Kommentaren der Ammonius-Schule,
11 Beitrge zur Geschichte der Philosophie und Theologie des Mittelalters 39/1,
12 Mnster: Aschendorff, 1961.
13 C. Lafleur and J. Carrier, Abstraction, sparation et tripartition de la philosophie
thortique: quelques lments de l’arrire-fond farabien et artien de Thomas
14
d’Aquin, Super Boetium “De Trinitate”, question 5, article 3, Recherches de Thologie
15 et Philosophie mdivales, 67, 2000, pp. 248 – 71.
16 C. Luna, Syrianus dans la tradition exgtique de la Mtaphysique d’Aristote. I. Syrianus
17 entre Alexandre d’Aphrodise et Asclpius, in M.-O. Goulet-Caz, ed., Le
18 commentaire entre tradition et innovation, Paris: Vrin, 2000, pp. 301 – 309.
–– , Trois tudes sur la tradition des commentaires anciens la ‘Mtaphysique’ d’Aristote,
19
Philosophia antiqua 88, Leiden/Boston/Cologne: E.J. Brill, 2001.
20 M.E. Marmura, Avicenna’s Critique of Platonists in Book VII, Chapter 2 of the
21 Metaphysics of his Healing, in J.E. Montgomery, ed., Arabic Theology, Arabic
22 Philosophy. From the Many to the One: Essays in Celebration of Richard M. Frank,
23 Leuven: Peeters, 2006, pp. 355 – 70.
P. Merlan, Abstraction and Metaphysics in St. Thomas’ Summa, Journal of the History of
24
Ideas, 14, 1953, pp. 284 – 91.
25 –– , From Platonism to Neoplatonism, The Hague: Martinus Njihoff, 1954, 2nd ed. 1960,
26 3rd ed. 1975.
27 A. Moreno, The Nature of Metaphysics, The Thomist, 30, 1966, pp. 109 – 35.
28 D. Morrison, Separation in Aristotle’s Metaphysics, Oxford Studies in Ancient Philosophy,
3, 1985, pp. 125 – 57.
29
–– , Separation: A Reply to Fine, Oxford Studies in Ancient Philosophy, 3, 1985,
30 pp. 167 – 73.
31 –– , Wyqistºr in Aristotle, Harvard Studies in Classical Philology, 89, 1985, pp. 89 – 105.
32 P. Natorp, Thema und Disposition der aristotelischen Metaphysik, Philosophische
33 Monatshefte, 24, 1988, pp. 37 – 65 and 540 – 74.
34
K. Nordberg, Abstraction and Separation in the Light of the Historical Roots of
Thomas’ Tripartition of the Theoretical Sciences, in R. Tyçrinoja, A. Inkeri
35 Lehtinen, D. Føllesdall, eds, Knowledge and the Sciences in Medieval Philosophy.
36 Proceedings of the Eighth International Congress of Medieval Philosophy, vol. III,
37 Helsinki: The Finnish Society for Missiology and Ecumenics, 1990, pp. 144 – 53.
38 D.J. O’Meara, Le problme de la mtaphysique dans l’antiquit tardive, Freiburger
39
Zeitschrift fr Philosophie und Theologie, 33, 1986, pp. 3 – 22.
J. Owens, Metaphysical Separation in Aquinas, Mediaeval Studies, 34, 1972, pp. 287 –
40
306.
41 –– , The Doctrine of Being in Aristotelian Metaphysics, Toronto: Pontifical Institute of
42 Mediaeval Studies, 1951.
1 M.-D. Philippe, )va_qesir, pq|shesir, wyq_feim dans la philosophie d’Aristote, Revue

2 thomiste, 48, 1948, pp. 461 – 79.
3
C. Pietsch, Prinzipienfindung bei Aristoteles. Methoden und erkenntnistheoretische
Grundlagen, Beitrge zur Altertumskunde, 22, Stuttgart: Teubner, 1992.
4 P. Porro, Astrazione e separazione: Tommaso d’Aquino e la tradizione greco-araba, in
5 Tommaso d’Aquino, Commenti a Boezio, ed. P. Porro, Milano: Bompiani, 2007,
6 pp. 527 – 80.
7 –– , Metafisica e teologia nella divisione delle scienze speculative del Super Boetium De
8
Trinitate, in Tommaso d’Aquino, Commenti a Boezio, ed. P. Porro, Milano:
Bompiani, 2007, pp. 467 – 526.
9 –– , ed., Metaphysica–sapientia–scientia divina. Soggetto e statuto della filosofia prima nel
10 Medioevo, Turnhout/Bari: Brepols/Pagina, 2006 [= Quaestio, 5, 2005].
11 K. Praechter, Review of Michaelis Ephesii In libros De partibus animalium, De
12 animalium motione, De animalium incessu, ed. M. Hayduck, CAG XXIII/2,
13
Gçttinger Gelehrte Anzeigen, 168, 1906, pp. 861 – 907.
J.-D. Robert, La mtaphysique, science distincte de toute autre discipline philosophique,
14 selon saint Thomas d’Aquin, Divus Thomas (Piacenza), 3rd ser., 24, 1947, pp. 206 –
15 22.
16 A. Speer, The Hidden Heritage: Boethian Metaphysics and its Medieval Tradition, in
17 Metaphysica–sapientia–scientia divina. Soggetto e statuto della filosofia prima nel
18
Medioevo, ed. P. Porro, Turnhout/Bari: Brepols/Pagina, 2006 [= Quaestio, 5, 2005],
pp. 163 – 81.
19 L. Spellmann, Substance and Separation in Aristotle, Cambridge: Cambridge University
20 Press, 1995.
21 C. Steel, Theology as First Philosophy. The Neoplatonic Concept of Metaphysics, in
22 Metaphysica–sapientia–scientia divina. Soggetto e statuto della filosofia prima nel
23
Medioevo, ed. P. Porro, Turnhout/Bari: Brepols/Pagina, 2006 [= Quaestio, 5, 2005],
pp. 3 – 21.
24 M. Tavuzzi, Aquinas on Resolution in Metaphysics, The Thomist, 55, 1991, pp. 199 –
25 227.
26 E. Trpanier, La philosophie de la nature porte-t-elle sur des spars ou des non-spars?,
27 Laval thologique et philosophique, 2, 1946, pp. 206 – 209.
28
J.F. Wippel, Metaphysics and separatio according to Thomas Aquinas, The Review of
Metaphysics, 31, 1978, pp. 431 – 70, reprinted in id., Metaphysical Themes in
29 Thomas Aquinas, Studies in Philosophy and the History of Philosophy, 10,
30 Washington, D.C.: The Catholic University of America Press, 1984, pp. 69 – 104.
31 C. Witt, Substance and Essence in Aristotle: An Interpretation of ,Metaphysics‘ VII–IX,
32 Ithaca, New York: Cornell University Press, 1989.
33
M. Wundt, Untersuchungen zur Metaphysik des Aristoteles, Stuttgart: Kohlhammer, 1953.
H.G. Zekl, Topos: Die aristotelische Lehre von Raum, Hamburg: Meiner, 1990.
34 A. Zimmermann, Ontologie oder Metaphysik. Die Diskussion ber den Gegenstand der
35 Metaphysik im 13. und 14. Jahrhundert. Texte und Untersuchungen, 2nd expanded
36 ed., Leuven: Peeters, 1998.
37
38
39
40
41
42
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
1
2
3
4 Two Senses of ‘Common’. Avicenna’s Doctrine of Essence
5 and Aquinas’s View on Individuation*
6
7 Gabriele Galluzzo
8
9
Introduction
10
11
The question I shall try to answer in this paper can be phrased as follows: Is
12
Aquinas’s endorsement of Avicenna’s doctrine of essence compatible with his
13
views on individuation?1 In Sections 1 and 2 I shall try to explain why some
14
tension might be perceived between these two aspects of the Dominican
15
Master’s thought. Roughly speaking, the potential conflict I see is the following.
16
According to Avicenna’s doctrine an essence in itself is neither universal nor
17
particular but rather indifferent to both characteristics. Universality and
18
particularity, in other words, are properties an essence acquires as a result of a
19
certain mode of existence, as a result of its existing, respectively, in the intellect
20
or in the extra-mental world, and not properties it possesses in itself. Thus,
21
within Avicenna’s framework an essence is universal or common only when
22
existing in the intellect and in no other sense. Aquinas’s view on individuation,
23
by contrast, seems to require that an essence be in some sense extra-mentally
24
common. For Aquinas standardly analyses a sensible substance into essence,
25
which accounts for the properties a sensible substance shares with all co-specific
26
substances, and principle of individuation, which is responsible for a sensible
27
substance’s being the individual substance it is. This suggests that co-specific
28
individuals are made of a common constituent, the essence, and of a further
29
30
31
32 * I wish to warmly thank Amos Bertolacci for his valuable explanations concerning several
33 aspects of Avicenna’s doctrine of essence and Dag Nikolaus Hasse for his comments and
34 suggestions on a first draft of this paper. I am also grateful to all the other participants in
the Villa Vigoni conference, and in particular to Giorgio Pini, whose comments helped
35
me to sharpen my views about some crucial points of my argument.
36 1 I have already dealt with some aspects of this question in: Galluzzo, Aquinas on
37 Common Nature. Although nothing of what I have argued for in my previous paper is
38 strictly incompatible with my considerations in the present one, the following pages
39 should be seen as a substantial refinement of my argument for the conclusion that
Avicenna’s doctrine of essence and Aquinas’s views on individuation yield two different
40
notions of commonality or universality. In particular, the distinction in Section 3
41 between the level of abstraction and the level of modal considerations is something that
42 was not so clear in my mind when I drafted the 2004 paper.
310 Gabriele Galluzzo
1 principle which accounts for their individuality. From the point of view of the
2 metaphysical analysis of material substances, therefore, the essence seems to
3 enjoy some kind of extra-mental commonality. But how can it do so, given that
4 Avicenna’s doctrine demands that an essence be common only in the intellect?
5 In Section 3 – 4, I shall suggest one possible way out of the conflict. In brief,
6 my solution consists in distinguishing two levels of analysis (Section 3) and,
7 consequently, two senses of ‘common’ or ‘universal’ (Section 4). Avicenna’s
8 doctrine of essence concerns the level of the actual existence of an essence. It
9 explains, in other words, how an essence can actually exist in different ways (i. e.
10 in the extra-mental world and in the mind) and hence take on incompatible
11 properties. Thus, when actual existence is concerned, an essence exists as
12 universal only in the intellect, which knows an essence by abstracting it, i. e. by
13 separating it out from the individuating conditions along with which it exists in
14 the extra-mental world. The metaphysical analysis of material substances, by
15
contrast, does not concern the actual existence of an essence. This is shown by
16
the fact that the commonality of an essence is revealed through a series of modal
17
considerations concerning the way in which an essence would exist if it could
18
exist without the principle of individuation: if we could, in other words, strip
19
the individuating conditions away from all co-specific substances we would be
20
left with only one essence for each species. Modal considerations show that,
21
although an essence is individual in the extra-mental world in that it exists as an
22
individual, its individuality is not primitive but rather derivative, i. e. due to
23
24
some principle external to it. Thus, the extra-mental commonality of an essence
25
is simply the idea that the actual, extra-mental individuality of an essence is not
26
metaphysically primitive. And being non-primitively individual is compatible
27
with being actually individual.
28 Accordingly, we should distinguish two senses of ‘common’ and, con-
29 sequently, two senses of the expression ‘common nature’. If ‘common’ is taken
30 in the sense of ‘actually common’ an essence is actually common only in the
31 intellect, whereas it is actually non-common in the extra-mental world, as
32 Avicenna’s doctrine maintains. But ‘common’ can also be taken in another sense,
33 which can be referred to as ‘modally common’, since it is mainly revealed
34 through modal considerations about the different metaphysical constituents of a
35 substance. According to this second sense of ‘common’, something is common if
36 its individuality is not primitive but derivative, that is due to something else.
37 And in this sense an essence is extra-mentally common, as the metaphysical
38 analysis of material substances plainly shows, for it owes its individuality to
39 some extrinsic principle, i. e. the principle of individuation. But being common
40 in this second sense is compatible with being actually common. For the second
41 sense of ‘common’ indicates a counterfactual (and, as a matter of fact,
42 counterpossible) property of a thing and not an actual one. Therefore, Aquinas’s
Avicenna’s Doctrine of Essence and Aquinas’s View on Individuation 311
1 endorsement of Avicenna’s doctrine of essence is also compatible with his view

2 on individuation.
3
4
5 1. Essence as Indifferent: Aquinas’s Understanding
6 of Avicenna’s Doctrine of Essence
7
8 Throughout his career Aquinas holds to Avicenna’s view that an essence, when
9 considered in itself or absolutely, is indifferent.2 What Avicenna means is that,
10 when taken absolutely, an essence possesses only those properties which are
11 indicated by its definition. For instance: the human essence, humanity, possesses
12 only those properties which are spelt out in its definition, that is animality and
13 rationality. All the other characteristics, including some important ontological
14 properties such as being one or many, being universal or particular, do not
15 belong to an essence when considered in itself, but only as a result of its
16 acquiring some mode of existence or other. As is well known, the modes of
17 existence Avicenna takes into account are two, extra-mental and mental
18 existence. Thus, an essence is not in itself universal, but acquires universality as a
19 result of its existing in the intellect. Likewise, it is not in itself particular, but
20 acquires particularity as a result of its existence in matter, i. e. as a result of its
21 extra-mental existence. In the extra-mental world an essence exists as an
22 individual object of a certain kind or species, whereas in the mind it is nothing
23 but a universal concept which represents all the individuals of a certain kind in
24 the same way. Of course, since an essence does in fact acquire particularity in its
25 extra-mental existence and universality in its mental one, particularity and
26 universality are neither included in the definition of an essence taken in itself
27
nor excluded by it. In this sense too, therefore, an essence, when taken in itself
28
or absolutely, is indifferent.
29
Universality and particularity invariably accompany an essence in its mental
30
and extra-mental existence, respectively – and so are in this sense disjunctively
31
necessary properties of an essence. Therefore, they are not accidental if by
32
‘accidental’ we mean ‘purely accidental’ and we wish thereby to refer to
33
properties that something can, unqualifiedly, have or not have.3 For, when the
34
35
2 Cf. Aquinas, De ent. et ess., c. 3, pp. 374 – 75; Q. de quo., VIII, q. 1, a. 1, vol. I,
36 pp. 51 – 3; Sent. lib. de an., lib. II, c. 12, pp. 115 – 6, lines 95 – 151; Q. de pot., q. 5, a. 9,
37 ad 16, vol. II, p. 155; S. th., Ia, q. 85, a. 2, ad 2, pp. 412 – 3; a. 3, ad 1, p. 414 and ad 4,
38 p. 414; Exp. Metaph., lib. VII, lect. 13, nn. 1570 – 71. The doctrine of the indifference
39 of essence is presented in: Avicenna, Phil. pr., tr. V, c. 1. On Avicenna’s doctrine of
essence in general see: Marmura, Avicenna’s Chapters on Universals; id., Avicenna on
40
Primary Concepts; id., Quiddity and Universality; De Libera, La querelle, pp. 185 – 91;
41 id., L’art, pp. 576 – 607; Black, Mental Existence.
42 3 Cf. Avicenna, Phil. pr., tr. V, c. 1, pp. 232, 7 – 10; 234, 42 – 57; 235, 72 – 236, 91.
1 essence exists in the extra-mental world, it is necessarily particular; when it exists

2 in the mind, by contrast, it is necessarily universal. In another sense, however,
3 i. e. when ‘accidental’ is taken in the sense of ‘non-essential’, universality and
4 particularity can still be regarded as accidental properties, in that they fall
5 outside the definition of an essence when taken absolutely.4 The ambiguity of
6 the notion of accidentality is also relevant in the context of the debate over the
7 essence-existence distinction, on which the doctrine of the indifference of
8 essence is substantially based. When Aquinas in his late writings criticises
9 Avicenna for holding that the existence of a thing is accidental to its essence, he
10 is clearly taking ‘accidental’ in the sense of ‘purely accidental’ and crediting
11 Avicenna with the claim that existence is related to essence in a purely accidental
12 manner.5 But Avicenna maintains that mental and extra-mental existence – as
13 well as the different properties accompanying these two forms of existence – are
14
disjunctively necessary.6 So Thomas cannot be right in his critical remark.7
15
What is more, from the argument in the De ente et essentia, c. 3, it clearly
16
emerges that Thomas himself takes the different modes of existence of an
17
essence – and hence the different properties that accompany such modes of
18
existence – to be accidental only in the sense of ‘non-essential’, in that they fall
19
outside the definition of the essence taken absolutely. Thus, once the notion of
20
accidentality is kept straight, there seems to be no fundamental disagreement
21
between Avicenna and Aquinas with regards to the relation between the essence
22
23
in itself and the different modes of existence it happens to possess. Admittedly,
24
in so far as extra-mental particulars are concerned, Aquinas further maintains
25
that essence and existence should be understood as two principles or ontological
26
constituents of created things – a piece of doctrine which is hardly to be found
27 in Avicenna’s writings. However, the general model according to which the
28
29 4 Cf. Avicenna, Phil. pr., tr. V, c. 1, pp. 228, 24 – 32; 230, 58 – 232, 74; 234, 42 – 57; 235,
82 – 236, 91.
30
5 Cf. Aquinas, Exp. Metaph., lib. IV, lect. 2, nn. 556 – 60.
31 6 For Avicenna’s main texts on the distinction between essence and existence see: Avicenna,
32 Phil. pr., tr. I, c. 5, pp. 33, 25 – 40, 54; tr. V, c. 1, passim. In the light of Averroes’s
33 subsequent criticisms (see n. 7) the texts concerning the relation among being, one and
34 quiddity should also be added to the list: cf. Avicenna, Phil. pr., tr. III, c. 1, p. 106,
45 ff.; tr. III, c. 2, p. 114, 17 – 20; tr. III, c. 3, p. 117, 80 – 86; pp. 119, 19 – 122, 71;
35
tr. VII, c. 1, p. 349, 7 – 18.
36 7 In all likelihood, Thomas was led astray by Averroes’s partly inaccurate remarks in his
37 Commentary on the Metaphysics: Averroes, In Metaph., lib. IV, t.c. 3, p. 67B–E; lib. X,
38 t.c. 8, p. 257E–G; 257K (a further passage belonging to Averroes’ commentary on Met.,
39 Book V, is omitted in the Latin translation). Averroes does not seem to distinguish very
clearly between ‘accidental’ as ‘purely accidental’ and ‘accidental’ in the sense of ‘non-
40
essential’. On the fortune of Avicenna’s views on being and unity see Stephen Menn’s
41 contribution to the present volume (where the case of Averroes is especially taken into
42 consideration).
1 Dominican Master describes the relation between essence and existence seems to
2 be in line with Avicenna’s main intuition: the existence of a thing falls outside its
3 essence, i. e. is external to it without being accidental in the strict sense of the
4 term.8
5 My previous considerations already suggest that, despite what some texts
6 seem to suggest, when talking of an essence considered in itself Avicenna does
7 not wish to introduce a mysterious third realm of essences, enjoying some kind
8 of not further specified mind-independent existence.9 What Avicenna intends to
9 do, on the contrary, is simply to separate out, in accordance with his distinction
10 between essence and existence, the properties which belong to an essence in
11 virtue of what it is in itself from those following upon its existence, be it mental
12 or extra-mental existence. An essence always exists either mentally or extra-
13 mentally, it is just that the properties which accompany the mental or extra-
14
mental existence of an essence are not properties of an essence as such.10 This
15
distinction also enables Avicenna – and Aquinas, who follows Avicenna very
16
17
18 8 The sense in which the existence of a thing falls outside its essence without being
19 accidental to it in the strict sense of the term is clarified by Aquinas by having recourse to
the potentiality-actuality model. It is not built into the essence of a thing that it exists
20
and hence the existence of a thing falls outside its essence. The essence, however, fixes the
21 way in which a thing exists, if it exists. Thus, existing for a thing means to actualise a
22 certain essence, which is a possible way of existing. As Aquinas sometimes expresses
23 himself (De sub. sep., c. 8, p. D 55, lines 205 – 12), a thing receives the act of being, i. e.
24 existence, in accordance with the principles of its essence. Accordingly, even though the
existence of a thing is extrinsic with respect to its essence, the potentiality-actuality model
25
guarantees that the concrete existence of a thing does not exceed the boundaries fixed by
26 its essence. In this sense, essence and existence are not related in a purely accidental way.
27 9 For an analysis of Avicenna’s different texts concerning the status of the essence in itself
28 see Wisnovsky’s contribution to the present volume. Some texts in Avicenna (see for
29 instance: Phil. pr., tr. V, c. 1, pp. 233, 36 – 234, 42) seem to imply that the essence in
itself is prior to or precedes in being the different properties accompanying mental and
30
extra-mental existence and, consequently, mental and extra-mental existence as well.
31 However, I do not take these texts to entail, necessarily, that the essence in itself enjoys
32 some kind of ontological priority over its different ways of existence and so over existence
33 in general. Possibly, all that Avicenna means is that, since extra-mental individuals and
34 mental concepts are modes of existence of a certain essence, such an essence must be
logically presupposed by its different modes of existence. The claim that Avicenna’s
35
essence in itself enjoys some kind of being over and above the being it acquires in its
36 different modes of existence is outlined in Owens, Common Nature, p. 4 (even though
37 Owens confines his remarks to the Latin rendering of Avicenna’s text).
38 10 Cf. Avicenna, Phil. pr., tr. I, c. 5, p. 36, 78 – 82; 83 – 93 (but the whole discussion up to
39 p. 40, 54 of the absolutely non-existent is relevant); see also: tr. V, c. 1, pp. 234, 44 – 6;
235, 82 – 236, 91. Thus, even if an essence for Avicenna is indifferent to the properties
40
following upon both mental and extra-mental existence, it is not also indifferent to
41 existence taken generally (i. e. so as to include both mental and extra-mental existence).
42 For an essence always exists in some way or other, whether mentally or extra-mentally.
1 closely here – to explain how an essence can take on contradictory properties,

2 such as universality and particularity. Such properties are not constitutive of an
3 essence as such but are rather possessed by an essence only as a result of its
4 existing in different fashions. If universality and particularity were constitutive
5 properties of an essence they should accompany it, as it were, everywhere. But
6 universality does not belong to an essence when it exists in the extra-mental
7 world, just as particularity is not possessed by an essence when it exists in the
8 intellect. It is likely that Avicenna’s doctrine was also prompted by some
9 concerns about sceptical arguments: How do we bridge the gap between mind
10 and reality? How do we make it sure that our concepts reflect the actual
11 conditions of extra-mentally existing things?11 Avicenna’s doctrine of essence
12 solves the problem by maintaining that our concepts reflect real things in that it
13 is one and the same essence that exists as an individual in the extra-mental world
14 and as a universal concept in the mind and so things and concepts are in some
15 sense made of the same stuff.
16 Be that as it may, the fact that Avicenna does not introduce any mysterious
17 third realm of essences does not make his doctrine any less ambiguous and
18 difficult to interpret. As Giorgio Pini, for instance, has recently pointed out,
19 there are at least two different readings of Avicenna’s doctrine, i. e. an
20 ontological and an epistemological one.12 On the ontological reading, the
21 essence in itself is a real constituent of both extra-mental things and concepts,
22 i. e. something which both extra-mental things and concepts are made of in
23 addition to the other constituents which characterise more specifically extra-
24 mental and mental beings. According to this view, extra-mental individuals are
25 constituted by essence plus individual characteristics or conditions, whereas
26 concepts have as their constituents essence plus universal characteristics or
27 conditions. On the epistemological and more deflationary reading, by contrast,
28 the essence in itself is nothing but what we are left with when we mentally
29 separate out an essence from the characteristics it enjoys either in its mental or
30 extra-mental existence. According to this second interpretation, we get the
31 essence in itself when we consider extra-mental individuals without their
32 individual characteristics or when we consider concepts without their universal
33 characteristics. The essence in its absolute consideration, in other words, is just
34 the residue of our considering either extra-mental individuals or concepts
35 without the properties which characterise extra-mental and mental existence,
36 respectively.13 Pini further argues that Aquinas’s understanding of Avicenna’s
37
38
39 11 For this see: McGinnis, Logic and Science.
12 Cf. Pini, Absoluta consideratio.
40
13 It seems to me that the ontological reading has some solid ground in Avicenna’s text,
41 especially when the Arab philosopher employs the part-whole model and appeals to the
42 language of composition in order to describe the relation between the essence in itself
1 doctrine shifts from the ontological reading – which is characteristic of

2 Aquinas’s earlier works such as the De ente et essentia and the Quod., VIII.1.1 –
3 to a more epistemological reading, which marks out the Dominican Master’s
4 later presentations of Avicenna’s view. It is not my intent in this paper to go into
5 the details of Pini’s argument – which would require a detailed analysis of the
6 texts he presents in support of his general reconstruction. However, my
7 considerations in Sections 2 and 3 are meant to cast some doubts on the
8 chronological hypothesis and defend the ontological reading as Aquinas’s
9 prevailing interpretation of Avicenna’s doctrine of essence, at least in so far as the
10 extra-mental existence of an essence is concerned. As my analysis of the essence-
11 supposit problem will show, in fact, most of Aquinas’s late texts on essence seem
12 to presuppose that the essence should be understood as an ontological
13 constituent of extra-mental individuals. This suggests that it should be
14 understood as a constituent of concepts as well.
15 Before tackling Aquinas’s explanation of the individuation of sensible
16 substances, let me go back to the problem of whether Avicenna’s doctrine of
17 essence leaves room for any kind of extra-mental commonality of an essence. In
18 the detailed presentation of Avicenna’s doctrine of essence he offers in c. 3 of the
19
De ente et essentia, Aquinas explicitly defends the view that it is the essence in
20
itself – and not the essence as existing in the intellect – that is predicated of the
21
individuals existing in the extra-mental world.14 Since Aquinas standardly
22
describes predication as a mental operation and so connects it with the intellect,
23
it is likely that the passage in the De ente et essentia introduces a different sense of
24
‘predication’. Probably, Aquinas is echoing and re-elaborating here on some
25
remarks in Avicenna’s Philosophia prima, where the Arab philosopher insists that
26
the fact that the animal in itself exists in the extra-mental world as an individual
27
animal is compatible with saying that it also exists in the individual animals.15
28
Avicenna’s worry must be that his talking about the essence in itself could
29
mislead people into thinking that it is not the essence belonging to the extra-
30
mental individuals that he is talking about, but something different. On the
31
contrary, the essence in itself is not a thing other than the essence we say extra-
32
33
mental individuals possess. The same thought, I surmise, is expressed by
34
Aquinas’s observation that it is the essence in itself – and not the essence as
35
existing in the intellect – that is predicated of extra-mental individuals. What
36
Thomas means is that an essence is that which makes of the extra-mental
37 and the essence taken together with the properties it acquires in its different modes of
38 existence (cf. Avicenna, Phil. pr., tr. V, c. 1, pp. 233, 33 – 5; 233, 39 – 234, 42; 234, 50 –
39 57). I do not wish, however, to insist on this point here.
14 Cf. Aquinas, De ent. et ess., c. 3, p. 375, lines 120 – 22.
40
15 Avicenna, Phil. pr., tr. V, c. 1, p. 234, 50 – 57, but see also pp. 234, 58 – 237, 22, where
41 Avicenna criticises the Platonist view that the abstracted animal exists in reality (see in
42 particular p. 237, 16 – 21, where the idea of predication crops up as well).
1 individuals the kind of thing they are and hence is something somehow existing
2 in individuals as well as something individuals can be rightly said to possess. If
3 this is Aquinas’s main thought, it becomes clear why he does not connect, in this
4 instance, predication with mental existence: human beings are rational animals
5 and so possess a certain kind of essence whether or not our intellect applies to
6 them the general concept ‘human being’. Essentialism, in other words, is a view
7 about how things are and not about how we understand or conceptualise things.
8 In spite of his defence of the objectivity of our talk of essences, however, in
9 the De ente et essentia Aquinas is quite explicit that the extra-mental world is
10 populated only by individual entities. What is more, in each individual
11 everything is individuated.16 This suggests that there is no room for something
12 existing in the extra-mental world as a universal or a common entity, not even as
13 a common or universal constituent of particular entities. Universality and
14 commonality can only be found in the intellect, where an essence exists as a
15 concept representing all the individuals of the same kind in the same way.17 So,
16 Aquinas’s view seems to be that the essence that is predicated of extra-mental
17 individuals and is in some sense present in them is not something common or
18 universal, i. e. is not something all the individuals of the same kind share in any
19 ontologically significant sense of the term. Of course, we are still allowed to say
20 that individuals belonging to the same species possess the same essence, but this
21 should not be taken in the literal sense that there is one common component
22 identical in all the individuals of the same kind, but rather in the weaker sense
23 that all the individuals of the same kind possess objective and essential
24 characteristics which group them together and so allow the intellect to class
25 them non-arbitrarily under the same concept. This seems also the meaning of
26 Aquinas’s famous remark in his commentary on the first book of the Sentences to
27 the effect that universals possess certainly some extra-mental ground, but receive
28 their formal completeness, i. e. their full existence as universals, only in the
29 mind.18 This seems to amount to saying that universals have some kind of
30 potential or incomplete existence in the extra-mental world, which is made
31 actual and complete by the intellect. Things, in other words, have objective,
32 essential features which enable us to group them together in a non-arbitrary way.
33 But this does not imply that they share some actual, common constituent.19
34
35
16 Cf. Aquinas, De ent. et ess., c. 3, p. 374, lines 80 – 82.
36 17 Cf. Aquinas, De ent. et ess., c. 3, p. 375, lines 91 – 6.
37 18 Cf. Aquinas, In I Sent., d. 19, q. 5, a. 1, vol. I, p. 486.
38 19 In terms of the standard classification of Medieval views on universals, Aquinas’s
39 understanding of Avicenna’s doctrine of essence seems to suggest a nominalist position
concerning sameness in kind: things are not the same in kind because they somehow
40
share some common constituent, but only because they are similar in some relevant and
41 essential features. On this view, there is nothing common in reality, it is only our intellect
42 that is able to consider things only with respect to the aspects in which they are similar,
1 2. Essence as Common: Aquinas’s View on Individuation

2
3 1. When we turn to Aquinas’s view on individuation, we seem to get a rather
4 different picture of essence. For Aquinas consistently maintains that sensible
5 substances, i. e. material objects, should be analysed into two ontological
6 constituents, the essence or common nature,20 which accounts for the objective
7 characteristics all the individuals of a certain kind share, and the principle of
8 individuation, which explains, by contrast, why an individual is the individual it
9 is, i. e. why it is distinct from all the other individuals of the same kind. In the
10 light of what I have been saying about Aquinas’s understanding of Avicenna’s
11 doctrine of essence, my main task is explaining in which sense the essence which
12 is present in all the individuals of the same kind and which is contrasted with
13 the principle of individuation can be said to be common. For Avicenna’s
14 doctrine of essence seems to imply that there is no ontologically significant sense
15
in which an essence is common when it exists in extra-mental individuals.
16
Before directly addressing this problem, however, I wish to say in this section
17
some more words about Aquinas’s doctrine of individuation and common
18
nature. I shall tackle this topic by focusing on the question of the relation
19
between essence and supposit, which is often forgotten when discussing
20
Aquinas’s doctrine of essence in spite of its pertinence to the question at stake.
21
On Aquinas’s understanding, a supposit is a full-fledged individual of a
22
certain kind: a man or an angel are standard examples of supposita. Thus, a
23
24
supposit includes the principles or characteristics that make of a certain
25
individual the kind of individual it is as well as those that make of it the
26
particular individual it is. This means that in the case of sensible substances the
27
supposit includes the essence and the principle of inviduation (individual
28 matter), whilst in the case of angelic substances it includes only the essence,
29 since angels are self-subsisting essences and so are individuals in themselves or of
30
31
32 leaving aside those in which they are, instead, dissimilar. Aquinas’s view on
33 individuation, by contrast, seems to point toward a realist solution to the problem of
34 universals: things are similar in kind because they share some common constituent, even
though it may be difficult to specify what the commonality involved exactly amounts to.
35
Although my opinion is that Aquinas’s general position is realist in spirit, I wish to
36 present Thomas’s discussion of the problem of universals in abstraction from the
37 standard labels we are used to employing when putting some order in the Medieval
38 debate.
39 20 For Aquinas’s use of the expression ‘common nature’ to refer generally to the essence of
things see for instance: Aquinas, In I Sent., d. 8, q. 4, a. 1, ad 2, vol. I, p. 219; d. 19,
40
q. 4, a. 2, vol. I, p. 483; d. 34, q. 1, a. 1, vol. I, p. 788; C. gent., II, c. 52, n. 1275; IV,
41 40, n. 3780; S. th., Ia, q. 13, a. 9, pp. 68 – 9; q. 29, a. 2, pp. 151 – 2; q. 76, a. 3,
42 pp. 354 – 6; Q. de pot., q. 9, a. 1, p. 226; a. 5, arg. 13, pp. 234 – 5; ad 13, p. 237.
1 themselves.21 Usually, therefore, Aquinas does not include in the notion of

2 supposit the act of being, i. e. the ontological principle that unites with a certain
3 essence and enables it to exist.22 Occasionally, however, and in particular in
4 Quod., II.2.2, he seems to think that properly speaking a supposit should
5 comprise not only the essential properties and, when necessary, the individuat-
6 ing principles, but also the act of being which enters into composition with the
7 essence to make up the structure of actual, existing individuals.23 Aquinas’s
8 oscillation seems to conceal a deeper uncertainty regarding what counts as an
9 individual (and hence as a supposit). Not including the act of being in the
10 notion of supposit means to concede that something can be taken to be an
11 individual regardless of whether it exists or not. In other words, when we lay
12 down the conditions for something to be an individual, a full-fledged member
13 of a certain kind, we are not concerned with the individual’s actually existing:
14 some individuals exist, some others do not, but are still thinkable as individuals
15 in that, if they existed, they would be complete, individual objects. Including
16 the act of being in the notion of supposit amounts to saying that it makes sense
17 to talk about a complete individual only in the presence of an actually existing
18 individual. The things, in other words, about which we evaluate whether they
19 are identical with their essence or not are the actually existing individuals and
20
such individuals, in so far as they exist, evidently include the act of being.
21
Aquinas’s waverings concerning the notion of supposit do not affect the
22
question of the relation between essence and supposit in the case of material
23
substances. For on both understandings of supposit, i. e. whether we take the act
24
of being to be part of the supposit or not, supposit and essence turn out to be
25
non-identical in the case of material substances. How we understand the notion
26
of supposit is, by contrast, relevant to the case of angels. For, when Aquinas does
27
not include the act of being in the supposit, he consistently maintains that
28
supposit and essence are identical in the case of angels.24 Angels, in fact, are self-
29
subsisting essences, i. e. essences that do not need to be received in something
30
else in order to subsist. Accordingly, angels turn out to be individual essences,
31
i. e. essences that do not need something external to themselves to be
32
33
individuated. Thus, if the act of being is excluded from what it is to be a
34
21 When I say that angels are individual in themselves or of themselves I simply mean that,
35
since their essence does not need any additional, extrinsic constituent to subsist, it does
36 not need any additional, extrinsic constituent to be an individual, either. Angels, in other
37 words, are created as individuals, i. e. as individual essences. Thus, the individuality of
38 the essences of angelic substances is in this sense primitive. Of course, from a different
39 perspective the individuality of angels (or of their essences) is not absolutely primitive, in
that it depends on God’s creative act that brings them into being.
40
22 Cf. for instance: Aquinas, C. gent., IV, c. 40, n. 3779; Q. de pot., q. 9, a. 1, p. 226
41 23 Cf. Aquinas, Q. de quo., II, q. 2, a. 2, vol. II, pp. 215 – 8 (esp. p. 217, lines 85 – 102).
42 24 Cf. Aquinas, Q. de pot., q. 9, a. 1, p. 226.
1 supposit, angels will clearly be identical with their essence, because the essence
2 of each angel is precisely an individual self-subsisting essence. By contrast, when
3 he takes the act of being to be part of the supposit, as he does for instance in
4 Quod., II.2.2, Thomas denies that the identity between supposit and essence
5 holds for the case of separate substances, too.25 For – as Aquinas himself makes
6 it clear – essence and supposit are distinct if there is something in the supposit
7 that falls outside the essence. Thus, since for Aquinas the act of being falls
8 outside the essence and is external to it, if the supposit includes the act of being
9 essence and supposit will not be identical. All things considered, it is not
10 difficult to see why Thomas usually prefers to leave the act of being out of the
11 notion of supposit. For one thing, in fact, this understanding of supposit allows
12 one to talk of actual as well as possible individuals. And talking about possible
13 individuals presupposes that something is conceived of as a full-fledged
14 individual before its being brought into actual existence. For another thing,
15 Aquinas’s general view on the nature of angels fits in better with the claim that a
16 supposit does not include the act of being. For an angel is an individual on
17 account of its essence-principle alone and not also on account of its act of being.
18 An angel is individual because its essence is a self-subsisting essence and self-
19 subsistence implies individuality.
20 Be that as it may, it is clear that Avicenna’s doctrine of essence is mainly
21 concerned with material substances, where the identity between supposit and
22 essence does not hold regardless of what account of supposit one chooses to
23 endorse. Therefore, in what follows I shall mainly focus on the case of material
24 substances.
25
26 2. Aquinas discusses the question of the relation between supposit and essence
27 in a number of texts, within both philosophical and theological contexts. In
28 philosophical contexts, the question about essence and supposit is part of a more
29 general problem Aristotle raises in Book VII (Z), c. 6, of the Metaphysics,
30 namely the question as to whether a thing is identical with its own essence.
31 Admittedly, Aristotle’s notion of ‘thing’ includes more than Aquinas’s notion of
32 ‘supposit’. For a supposit is a full-fledged individual substance, whereas
33 Aristotle’s investigation into the identity claim concerns also things other than
34 substances, such as accidents and accidental composites (i. e. things like ‘white
35 man’, which are the result of the composition of a substance with an accidental
36 property). However, it is clear that the discussion about whether or not sensible
37 substances are identical with their essences is also one of Aristotle’s main
38 concerns. As a matter of fact, Aquinas’s commentary on Met., Book VII will be
39 of much help in Section 3 in determining the extra-mental status of essence.
40 Suffice it to say for the time being that Aquinas credits Aristotle with the view
41
42 25 Cf. Aquinas, Q. De quod., q. 2, a. 2, vol. II, p. 217, lines 85 – 102.
1 that a particular material substance is not identical with its essence26 – a view
2 which, as we have seen, is in agreement with Aquinas’s consistent doctrine
3 throughout his writings.
4 As to theological contexts, there are mainly three areas where Aquinas
5 discusses the problem of the identity between essence and supposit: (i) texts
6 concerning the nature of God, for instance His simplicity (S. th., Ia, q. 3, a. 3;
7 C. gent., I, c. 21) (ii) Trinitarian contexts (S. th., Ia, q. 29, a. 2; Q. de pot., q. 9,
8 a. 1); (iii) finally, texts concerned with the metaphysics of incarnation, with
9 particular reference to the problem as to whether the union of the divine word
10 with the human nature is made in persona or not (Q. verb. inc., a. 1; S. th., IIIa,
11 q. 2, a. 2). I have analysed Aquinas’s use of the supposit-essence problem in
12 theological contexts in another paper of mine27 and hence I shall not dwell on
13 the point in the present context. The only thing which I would like to recall here
14
and which is particularly relevant to our concern is that from all such texts it
15
clearly emerges that, when Aquinas claims that in the case of material substances
16
supposit and essence are not identical, what he means is that they are really, i. e.
17
mind-independently, distinct.28 This is easily realised if we turn our attention
18
for instance to the question of the simplicity of God: What would the point be
19
of insisting that in God supposit and essence are identical, if the non-identity
20
between supposit and essence introduced no mind-independent distinction but
21
a merely conceptual one? After all, supposit and essence can be distinguished
22
23
conceptually even in the case of God. Thus, to say that they are identical in the
24
case of God is supposed to exclude some kind of mind-independent distinction
25
and, consequently, composition in God’s nature. Similar considerations could
26
be advanced also for the case of incarnation.29 What Aquinas wishes to
27 conclude, therefore, is that in the case of material substances supposit and
28 essence are distinct independently of any activity of our mind.
29 When saying that essence and supposit are mind-independently distinct, I
30 want to exclude even the possibility that the distinction between essence and
31 supposit might be correctly described as a particular kind of conceptual
32
33 26 Cf. Aquinas, Exp. Metaph., lib. VII, lect. 11, nn. 1535 – 6.
34 27 See Galluzzo, Aquinas on Common Nature.
28 Cf. Aquinas In I Sent., d. 5, q. 1, a. 1, vol. I, p. 151; d. 34, q. 1, a. 1, vol. I, pp. 787 – 90;
35
Q. de pot., q. 9, a. 1, p. 226; Q. verb. inc., a. 1, p. 442; a. 2, ad 7, p. 428; S. th., IIIa,
36 q. 2, a. 2, pp. 1872 – 4. For texts where Aquinas employs more nuanced expressions
37 (which do not alter in any event his general point), see: C. gent., lib. IV, c. 40, n. 3781;
38 S. th., Ia, q. 3, a. 3, pp. 16 – 7; S. th., Ia, q. 29, a. 2, ad 3, p. 151; Q. de quo., II, q. 2, a. 2,
39 vol. II, p. 217, lines 107 – 8.
29 For an analysis of the metaphysical machinery behind Aquinas’s doctrine of incarnation,
40
see: Cross, The Metaphysics, pp. 51 – 64. Cross touches upon several issues that are
41 relevant to the understanding of the notions of essence and supposit which I am trying to
42 defend in this section (see in particular his explanation of the part-whole model).
1 distinction (distinction of reason), i. e. the kind of conceptual distinction which

2 depends not only on the characteristics of our mind but also on the objective
3 features of extra-mental things. In book I of his Commentary on the Sentences, in
4 fact, Aquinas implicitly distinguishes two different kinds of distinction of
5 reason, i. e. of conceptual distinction.30 According to one kind of distinction,
6 two things are distinct in the intellect only on account of an operation of the
7 intellect itself, only because, in other words, our intellect is capable of
8 introducing distinctions even if they have no objective ground in the extra-
9 mental world. According to another kind of distinction, by contrast, two things
10 are distinct in the intellect not only on account of an operation of the intellect
11 but also on account of some objective features of the world. Since also this
12 second kind of distinction is a conceptual distinction, Aquinas’s point is not that
13 the things that are distinct in the intellect are also distinct in reality. On the
14 contrary, such things are distinct only in the intellect. However, in the case of
15 the second kind of conceptual distinction our intellect’s conceiving things as
16 distinct or separate is somehow grounded on reality and is not just the product
17 of our mental activity. As an example of the second kind of conceptual
18 distinction Aquinas presents the case of the multiplicity of divine attributes. The
19 different divine attributes are in reality, i. e. in God, only one thing in that they
20 are all really identical with the divine essence. However, our attribution to God
21 of a multiplicity of attributes is grounded on the nature of God Himself, i. e. on
22 the plenitude or richness of God’s nature. In God there is no real multiplicity
23 and hence our conception of the divine attributes as multiple and distinct is due
24 only to the limitedness of our intellect, which is unable to capture the infinity of
25 God’s nature. Still, the conceptual distinction of different divine attributes is not
26 unrelated to God’s nature, in that God’s nature contains in its simplicity all the
27 perfections signified by the attributes we think of as distinct.
28 Now, when I say that, in the case of material things, the distinction between
29 essence and supposit is mind-independent, what I am suggesting is that such a
30 distinction should not be understood as an instance of the second kind of
31 conceptual distinction Aquinas mentions in his Commentary on the Sentences,
32 i. e. the conceptual distinction which is grounded on the nature of the things we
33 think of. For the distinction between essence and supposit does not depend on
34 the activity of our intellect and hence is there whether we think about things or
35 not. For if the distinction between essence and supposit in material creatures
36 were only a conceptual distinction, material objects would not be really
37 composed of essence and principle of individuation, just as the divine essence is
38 not really composed of the different perfections which the divine attributes
39 signify. Texts show, however, that Aquinas takes material objects to be really
40 composed of essence and principle of individuation. Therefore, the distinction
41
42 30 Cf. Aquinas, In I Sent., d. 2, q. 1, a. 3, vol. I, pp. 63 – 72 (esp. pp. 69 – 71).
1 between essence and supposit must be stronger than any kind of conceptual
2 distinction, including the one which is grounded on how things are in reality.
3 The sense in which supposit and essence are mind-independently distinct in
4 the case of material substances can be further refined. As I see things, Aquinas
5 conceives of the relation between supposit and essence according to a part-whole
6 model. In other words, in material substances the essence is a part, i. e. an
7 ontological constituent, of a supposit, i. e. of an individual. Thus, it is not
8 incorrect to describe an individual as ‘essence plus something else’, the
9 something else being the principle of individuation which marks off one
10 particular individual from all the other co-specific individuals. To my
11 knowledge, only in one text does Aquinas cast some doubts on the part-
12 whole model as an adequate explanation of the relation between essence and
13 supposit, namely Quod., II.2.2, which I have already mentioned in connection
14 with the problem of the identity between a thing and its essence. In this text,
15 Aquinas suggests that it is not unqualifiedly correct to describe the relation
16 between essence and supposit as one between a part and its whole, but it is
17 better to say that supposit and essence are signified, respectively, as whole and
18 part. In all likelihood, Aquinas’s occasional scepticism about the part-whole
19 model is due to a peculiar feature of his doctrine of essence. As is known,
20 according to Aquinas the essence of material substances includes not only the
21 form of such substances, but also their so-called common matter, i. e. the type of
22 matter all individuals of a certain species possess.31 The principle of
23 individuation for material substances, by contrast, is the so-called materia
24 signata or individual matter, i. e. the material characteristics which are proper to
25 one individual alone and so distinguish it from the other co-specific individuals.
26 Upon reflection, the couple common matter-individual matter is a real obstacle
27 to conceiving of the relation between essence and supposit in terms of part and
28 whole. For individual matter seems to be nothing but a further specification and
29 determination of common matter: if the common matter of human beings is
30 flesh and bones, for instance, their individual matter will be individual flesh and
31 bones, i. e. flesh and bones of certain, specified dimensions. Thus, it seems
32 difficult to view individual matter as a further, distinct component which is
33 added on to the essence in order to make up a whole, in that individual matter
34 relates to a part of the essence, i. e. common matter, not as something extrinsic
35 and distinct, but rather as the more determinate relates to the less determinate.
36 And in fact Aquinas quite understandably observes in the body of Quod., II.2.2
37 that the part-whole model would be more properly applicable if the essence of
38 material substances contained form alone and not also common matter. For in
39 such a case the essence would be purely formal and the principle of
40
41 31 Cf. for instance: Aquinas, De ent. et ess., c. 2, pp. 67 – 84, pp. 370 – 71; Exp. Metaph.,
42 lib. VII, lect. 9, nn. 1467 – 9.
1 individuation purely material. Matter and form, on the other hand, are totally
2 distinct components and hence also essence and principle of individuation
3 would be, in the hypothetical situation, two different entities which make up a
4 third, distinct entity, i. e. the supposit. In the situation considered, in other
5 words, both essence and principle of individuation would be easily conceived of
6 as parts of the supposit they jointly make up.
7 In spite of these difficulties, my impression is that Aquinas’s uneasiness in
8 the text at issue with the part-whole model does not undermine his general
9 intuition that the essence is an ontological constituent of material objects. For it
10 is known that Aquinas devises no intermediate distinction between real
11 distinction and distinction of reason, i. e. conceptual distinction. What is more,
12 while Aquinas recognises, as we have seen, that there are different kinds of
13 distinction of reason, i. e. conceptual distinction, he never explicitly admits of
14 different kinds of real distinction. His real distinction, however, seems to
15 include different degrees of mind-independent distinctness. Two things, in other
16 words, can be distinct from one another to a higher or lesser degree than two
17 others. For instance, Aquinas’s reasonable point in Quod., II.2.2 is that essence
18 and principle of individuation (and consequently essence and supposit) are
19 distinct to a lesser degree than matter and form. It is not difficult to see why
20 Aquinas thinks so. In the standard Aristotelian picture, as long as a material
21 substance exists, its matter and its form are only one object and not two. The
22 matter and form of a material substance are, as we say, coinciding entities. Still,
23 matter and form are distinct, i. e. are two distinct constituents of a material
24 substance. So, how can we set them apart? One line of thought is to appeal to
25 the idea that matter and form have or, at least, may have different life histories.
26 The one can exist when the other does not. A piece of matter, for instance, can
27 take on different forms over time. Presumably, matter and form will also have
28 different modal properties: there are things matter can do that form cannot do
29 or changes matter can undergo that form cannot. However, appealing to the
30 different life histories of matter and form seems to be enough to set them apart.
31 So, in so far as they may have different life histories, matter and form are in
32 some sense separable. The case of the relation between principle of
33 individuation and essence seems to be very different. The existence of
34 individual matter is always bound up with the essence it renders individual.
35 There is no time at which individual matter exists and the essence it makes
36 individual does not. Conversely, the essence of a sensible substance, in so far as it
37 is not a self-subsisting essence but always needs something else to subsist, cannot
38 exist without the principle of individuation which enables it to subsist. Thus,
39 there is no sense in which principle of individuation and essence can exist
40 separately. Therefore, Aquinas seems to be right in maintaining that matter and
41 form are more distinct than essence and principle of individuation. On the
42 other hand, Thomas seems to think that this difference in degree between the
1 matter-form and the essence-principle of individuation relations should not

2 prevent us from regarding essence and principle of individuation as mind-
3 independently distinct and from conceiving of them as two ontological
4 constituents of an individual thing.
5 Therefore, I shall assume that supposit and essence are mind-independently
6 distinct in the case of material substances and that the part-whole model works
7 after all as the best explanation of their relation. According to Aquinas an
8 individual substance can be correctly described as made of two ontological
9 constituents, namely the essence and the principle of individuation. More
10 important for our purposes, this understanding of the structure of material
11 beings strongly suggests that the essence-component is in some sense common
12 to all the individuals of the same species. For if one individual of a certain
13 natural kind is distinct from another on account of the principle of
14 individuation, i. e. of its different individual material characteristics, there
15 follows that co-specific individuals are not distinct on account of their essence
16 and so the essence-component of an individual must be somehow common or
17 shared. To see that this is actually the way Aquinas views things, I would like to
18 call the attention to his De potentia, q. 9, a. 1, a text devoted to introducing the
19 main terms involved in the discussion of the trinity, ‘person’, ‘essence’,
20 ‘subsistence’ and ‘hypostasis’. The thesis defended in the article is that the
21 distinction between essence and individual is real, i. e. mind-independent, in the
22
case of material substances, while it is only a distinction of reason in the case of
23
immaterial ones (i. e. God and angels). In so far as material substances are
24
concerned, Thomas justifies his claim as follows:
25
26 The reason for this distinction [i.e. the distinction between individual and essence]
27
is that several subjects may have a common nature; thus several men have in
common the nature of man. Hence the need of distinguishing what is one from
28 what is multiplied …
29 For if whatsoever is in the individual were to belong to the common nature,
30 there would be no possible distinction between individual substances of the same
31 nature. Now what is in the individual substance besides the common nature is
32 individual matter (which is the principle of singularity) …32
33
34 32 Cf. Aquinas, Q. de pot., q. 9, a. 1, p. 226: ‘Huius autem distinctionis ratio est, quia
inveniuntur plura subiecta in una natura convenire, sicut plures homines in una natura
35
hominis. Unde oportuit distingui quod est unum, ab eo quod multiplicatur: natura enim
36 communis est quam significat definitio indicans quid est res; unde ipsa natura
37 communis, essentia vel quidditas dicitur. Quidquid ergo est in re ad naturam
38 communem pertinens, sub significatione essentiae continetur, non autem quidquid est
39 in substantia particulari, est huiusmodi. Si enim quidquid est in substantia particulari ad
naturam communem pertineret, non posset esse distinctio inter substantias particulares
40
eiusdem naturae. Hoc autem quod est in substantia particulari praeter naturam
41 communem, est materia individualis quae est singularitatis principium’. English
42 translation: Aquinas, On the Power of God, vol. III, pp. 98 – 9 (slightly modified).
1 What is striking about this argument is its distinctively realist flavour. Aquinas
2 seems to assume right from the start that in order to explain how there can be
3 many individuals of the same species we need to posit in each of the co-specific
4 individuals a mind-independent distinction between two ontological constit-
5 uents. More particularly, we should distinguish between something which is in
6 itself common and one, i. e. the essence which gets individualised and
7 multiplied, and something which is responsible for the individualisation and
8 multiplication of the common component. It is crucial to Aquinas’s argument
9 that the Dominican Master speaks of ‘individuals of the same species’. For it is not
10 unconceivable to imagine a situation in which individual material substances
11 would possess individual essences and so would be distinct from one another not
12 on account of some extra component which particularises their common nature,
13 but on account of their individual essences themselves. However, the problem
14 with this suggestion is that, according to Aquinas, in the situation imagined
15 individual substances would not belong to the same species. For they would
16 differ on account of their essences, and each essential difference entails a
17 difference in species.33 The situation imagined is not dissimilar from the actual
18 case of angels. Angels are self-subsisting essences and hence individual
19 substances.34 However, since the difference between one angel and another is
20 a difference in essence, in that each angel is a different self-subsisting essence,
21 one angel differs from another in species and not in number. For difference in
22 number occurs only when two things have the same essence and differ on
23 account of characteristics external to the essence. This is why co-specific
24 material substances differ in number: their common essence must be received in
25 matter in order to subsist and so it gets numerically multiplied on being received
26 in different pieces of matter. In conclusion, for Aquinas there cannot be more
27 than one essence for each species of material substance and the plurification of
28 such an essence must be accounted for by a principle of individuation which is
29 external to the essence itself. In other words, if we could strip the principle of
30 individuation away from all co-specific material substances we would be left
31 with only one essence for each species.
32
33
34
35
36
37
38
39 33 Cf. Aquinas, In II Sent., d. 3, q. 1, a. 4, vol. II, p. 97; d. 17, q. 2, a. 1, ad 5, vol. II,
p. 430; C. gent., II, c. 81, n. 1621; c. 93, n. 1797; Q. de pot., q. 3, a. 10, p. 71.
40
34 Cf. for instance: Aquinas, De unit. intell., c. 5, p. 311, lines 63 – 7; 71 – 4; S. th., Ia,
41 q. 76, a. 2, ad 3, p. 354; Q. de an., q. 2, ad 5, pp. 19 – 20, lines 361 – 92; Q. de spir., a. 8,
42 ad 8, p. 63, lines 312 – 24; Exp. Peryer., lib. I, lect. 10, p. 50, lines 95 – 103.
1 3. A Possible Way out: Actual vs. Modal Properties35

2
3 If my analysis so far is correct, there seems to be some tension in Aquinas’s
4 conception of essence. On the one hand, according to Aquinas’s interpretation
5 of Avicenna’s doctrine of essence there is nothing common or universal in the
6 extra-mental world. In the extra-mental world everything is individual.
7 Moreover, the essence, when considered in itself or absolutely, is neither
8 particular nor universal. Commonality and unity, which are the distinguishing
9 marks of universality, accompany only the mental existence of an essence. It is
10 only in the intellect, which separates out the essence from the individuating
11 principles, that an essence can be regarded as common and one. On the other
12 hand, Aquinas’s views on the individuation of material substances suggest that
13 each material substance should be analysed into two ontological constituents,
14 the common essence and the principle of individuation which multiplies and
15
particularises such an essence. It might seem, therefore, that there is after all
16
something common existing in the extra-mental world, i. e. the common nature
17
which all the individuals of a certain species share.
18
Is there any genuine conflict between these two aspects of Aquinas’s
19
thought? If not, is there a way of dispelling the appearance of conflict? Finally:
20
Is there any connection between the sense in which an essence is common from
21
the point of view of individuation and the Avicennian claim that it is common
22
only in the intellect? These are the questions I shall try to give an answer to in
23
24
this section. As a step towards a possible solution, let us see first of all what
25
Aquinas means by ‘essence’ or ‘common nature’ when talking about
26
individuation. As I have already hinted, his point seems to be that if we could
27
strip the individuating principles away from each individual of a certain kind we
28 would be left with only one essence. However, since Aquinas believes that we
29 cannot in fact strip the individuating principles away from a sensible substance,
30 it is clear that in whatever sense an essence is common in the individuals
31 possessing it, it is not actually common. In actuality, an essence exists in the
32 extra-mental world as individuated and multiplied.
33 Probably, the text where Aquinas best explains the sense in which an essence
34 can be said to be common, even though not actually common, is his
35 commentary on Met. Z 11.36 In the text, Aquinas tries to make sense of
36 Aristotle’s claim that an individual sensible substance, i. e. a composite of matter
37
38 35 In this section I have taken inspiration from the remarks on coinciding entities contained
39 in Lowe, A Survey, pp. 59 – 76, where it is argued that coinciding entities (and
ontological constituents can be certainly thought of as coinciding entities) can be
40
distinguished on the basis of their different historical and/or modal properties. The
41 application of such conceptual tools to the case of Aquinas’s ontology is, of course, mine.
1 and form, is not identical with its own essence.37 Aquinas’s argument in favour
2 of Aristotle’s claim seems to have two crucial premisses. (P1) First, an essence is
3 what is expressed by a definition. Therefore, since a definition refers only to the
4 species an individual belongs to and not to the individual as such, an essence
5 pertains only to the species and not to the individual. (P2) Second, the essences
6 of material substances, of necessity, exist only in individuals. Such essences, in
7 other words, are not self-subsisting essences, but rather need something else, i. e.
8 some kind of material substratum, to subsist. From these two premisses,
9 Aquinas seems to draw two conclusions. (C1) First, since the essence belongs to
10 the species and the essences of material substances only exist in individuals, then
11 we must distinguish within each individual the essence from the principle of
12 individuation. The principle of individuation, besides being what individualises
13 and multiplies the essence, is also the principle enabling it to subsist. For the
14 essence of material substances cannot subsist by itself. (C2) The second
15 conclusion is that, since the essence of material substances contains matter, we
16 must also distinguish, within each material substance, between common matter,
17 which falls within the essence, and individual matter (materia signata), which
18 falls outside the essence. Individual matter is precisely the principle of
19 individuation.
20 As can be seen, (C2) is Aquinas’s standard distinction between common and
21 individual matter, while (C1) perfectly reflects his view on the relation between
22 essence and supposit I have tried to illustrate in Section 2, i. e. the claim that in
23 material substances the essence is distinct from the supposit and is an
24 ontological constituent of it. Some new element comes from the further remark
25 which Aquinas appends to his main argument.38 He observes that, even though
26 in the extra-mental world an essence cannot exist apart from individuals, i. e.
27 apart from the individuating principles enabling it to subsist, it can do so in the
28 intellect, which knows essences by separating them from the material and
29 individual conditions along with which they exist in the extra-mental world.
30
31 37 Aristotle discusses the question of the identity (or non-identity) between a thing and its
32 essence in Met. Z 6 and then goes back to it in Met. Z 11 (1037a33-b7), in the course of
33 his final summary of the results achieved by the enquiry into the notion of essence. It is
34 when commenting on the Z 11 passage (Exp. Metaph., lib. VII, lect. 11, nn. 1535 – 6)
that Aquinas presents his final interpretation of the identity claim. For, as he himself
35
observes (Exp. Metaph., lib. VII, lect. 11, n. 1535), Aristotle’s discussion in Z 6 leaves it
36 unclear whether sensible substances are in fact identical with their essence or not. In his
37 final reconsideration of the question in Z 11, however, Aristotle makes it clear that
38 things which can be analysed into a formal and a material constituent are not identical
39 with their own essence (1037b3 – 4). This clearly includes also the case of material
substances. For more on the identity thesis and for a general reconstruction of Aquinas’s
40
intepretation of Met., Book VII, I take the liberty of referring to Galluzzo, Aquinas’s
41 interpretation. See also: id., Averroes and Aquinas.
1 Therefore, even if an individual material substance is not identical with its own
2 essence in the extra-mental world, it is in some sense identical with its essence in
3 the intellect. For in the intellect an individual exists without material and
4 individuating conditions and so in its mental form of existence does not contain
5 anything in addition to its essence.
6 Aquinas’s remark in the commentary on Met. Z 11 might be taken to be
7 offering an easy way out of the contrast between Avicenna’s doctrine of essence
8 and the Dominican Master’s view on individuation. According to Avicenna’s
9 doctrine or at least to Aquinas’s understanding thereof, an essence is not
10 universal or common in the extra-mental world, for there it exists as an
11 individual of a certain kind. Neither is it common in itself, for in itself it is
12 indifferent. An essence is common, by contrast, in the intellect, where it exists as
13 a universal concept. Now, when Aquinas says that a material substance is
14 analysable into a common essence and a principle of individuation, what he
15 means, at least to stick to his own words, is that if we strip the principle of
16 individuation away from all the individuals of a certain kind we are left with
17 only one essence. However, it is only the intellect that can set apart essence and
18 principle of individuation and so it is only in the intellect that an essence exists
19 as one and common. Thus, the suggestion is that Aquinas’s talking of a material
20 substance’s being analysable into two different constituents is just another way
21 of saying that our intellect is capable of separating essence and principle of
22 individuation.39 Such an intellectual operation is not arbitrary, in that it is based
23 on the objective features of extra-mental things and on their essential
24 similarities. However, it does not force us to introduce any extra-mental
25 commonality or universality with regards to the essence of material substances.
26 Although this suggestion is very neat and economical, I find it, all things
27 considered, slightly misleading. Its main fault is that of not distinguishing
28 between the natural operation of abstraction and the philosophical consid-
29 erations that lead us to analyse sensible substances into essence and principle of
30 individuation. In other words, we arrive at the conclusion that sensible
31 substances are composed of essence and principle of individuation not as a result
32 of our natural cognitive processes, but rather through a series of philosophical
33 considerations concerning the metaphysical structure of sensible substances.
34 Therefore, the claim that essence is the common constituent of sensible
35 substances can hardly be reduced, as the suggestion just presented holds, to the
36
37 39 For texts that might be taken to support this suggestion see: Aquinas, In I Sent., d. 19,
38 q. 4, a. 2, vol. I, p. 483; q. 5, a. 1, vol. I, p. 486; C. gent., I, c. 26, n. 241. In all these
39 passages Aquinas puts emphasis on the idea that commonality and unity belong to an
essence only when it exists in the mind. In the second text in particular, he makes also
40
reference to the already mentioned claim that universals have only a potential or
41 incomplete existence in the extra-mental world, which is made actual and complete by
42 the intellect. This claim too goes in the direction of the suggestion I have just presented.
1 fact that the natural process of abstraction presents to the intellect the essence
2 without individuating conditions. Thus, I shall offer another way out of the
3 difficulty, which, I think, is closer to Aquinas’s main intuition. My view is that
4 Avicenna’s doctrine of essence and Aquinas’s view on individuation introduce
5 two different levels of analysis and hence appeal to different kinds of
6 consideration. Avicenna’s doctrine concerns what can be called the actual
7 existence of an essence. It explains in other words how it is possible for an
8 essence to actually exist in different manners and to enjoy incompatible
9 properties. The solution consists in distinguishing what an essence is in itself
10 from the properties it enjoys in its different modes of actual existence, be it
11 mental or extra-mental existence. Aquinas’s doctrine of individuation, by
12 contrast, does not have to do with the actual existence of an essence, but rather
13 with the metaphysical structure of extra-mental things, which can be unravelled
14
only by appealing to modal considerations, i. e. to situations which depart from
15
the actual existence of an essence. In other words, just as we can distinguish two
16
different constituents of a thing, such as for instance matter and form, on the
17
basis of their different life histories, so we can set apart essence and principle of
18
individuation on the basis of their different modal properties. The implicit
19
assumption in the argument is that two things which have different modal
20
properties are not identical. The very language Aquinas employs to characterise
21
the common constituent of a sensible substance reveals his appealing to the
22
23
modal properties of an essence. If we could – Aquinas seems to imply – strip the
24
principle of individuation away from the individuals of a certain kind we would
25
be left with only one essence; or alternatively: if the common nature of a certain
26
species could exist separately, i. e. in separation from the conditions along with
27 which it actually exists, it would be just one.40 More specifically, Aquinas is here
28 appealing to some sort of counterpossible considerations (per impossibile) – as is
29 signalled by the ‘could’ in my English rendering of his argument – since it is in
30 fact metaphysically impossible to separate essence and principle of individu-
31 ation. However, the fact that such considerations are based on metaphysical
32
33 40 Cf. Aquinas, Exp. Metaph., lib. VII, lect. 11, nn. 1535 – 6; C. gent., II, c. 42, n. 1275;
34 Q. de spir., a. 8, p. 80, lines 189 – 204. See also Exp. Peryer., lib. I, lect. 10, p. 50, lines
95 – 103, where Aquinas proves (as in Q. de spir., a. 8) that there does not exist more
35
than one separate substance for each species by appealing to the thought-experiment
36 that, if whiteness could exist separate from matter, it would be just one. In Q. de spir.,
37 a. 1, p. 13, lines 363 – 76 the example of whiteness crops up again, this time around to
38 support the claim that there can be only one thing that is its own being or existence: just
39 as it is impossible that there would be more than one whiteness, if whiteness could exist
in separation from any subject, so there cannot be more than one self-subsisting act of
40
being. The connection with the case of God and His unicity may suggest one more
41 reason why Aquinas is willing to put ontological weight on the modal status of an
42 essence, but I do not wish to pursue this issue at present.
1 counterpossibility does not exclude it that they may still reveal genuine,
2 metaphysical properties of an essence. Quite the contrary, modal considerations
3 – i. e. considerations about how an essence would exist if we could get rid of the
4 principle of individuation – enable us to separate the properties an essence
5 possesses as a result of its actual existence from its fundamental metaphysical
6 properties. The fundamental metaphysical property of an essence is the fact of
7 being sharable. The essence of sensible substances is not actually and literally
8 shared because it always exists together with the principle of individuation.
9 However, from the point of view of the metaphysical structure of extra-mental
10 things an essence is always sharable.
11 It is important to stress once again that the modal considerations which
12 underlie the metaphysical analysis of material substances should not be confused
13 with the process of abstraction by which we obtain universal concepts.
14 Abstraction is in fact an undeliberate and natural process, whereas the
15 metaphysical analysis of material objects is clearly based on voluntary and
16 deliberate considerations of the intellect, which reflects upon the modal
17 properties of things.41 Of course, the fact that our intellect is capable of
18 separating, through abstraction, an essence from the principle of individuation
19 may help us to carry through the metaphysical analysis of material objects.42 It
20
21
22 41 It might be observed that, in some cases, abstraction does not turn out to be such an easy
23 and natural process, which can be carried through unaided by philosophical
24 considerations. We may experience difficulties – it might be remarked – in forming
the concepts corresponding to certain peculiar common natures, such as for instance
25
those natures that have unique instantiations, like the sun and the moon. My response to
26 this observation would be that the difficulties we experience with the case of the sun and
27 the moon do not concern our concepts of what the sun or the moon is (which are
28 obtained rather naturally and spontaneously) but have rather to do with whether the
29 concept of the sun or that of the moon should be taken to be concepts of individuals or
of species. Unique instantiation might be wrongly taken to exclude a distinction between
30
individual and species. To see the distinction, of course, philosophical considerations are
31 required.
32 42 Cf. for instance: Aquinas, Q. de spir., q. 8, p. 80, lines 189 – 204; Exp. Peryer., lib. I,
33 lect. 10, p. 50, lines 95 – 103, where Aquinas always associates the thought that an
34 essence would be just one if it could exist separately with the claim that it does exist as
one in the intellect. I take it, however, that the meaning of these texts is not that the
35
extra-mental commonality of an essence reduces itself to its potentiality to exist as one
36 and common in the intellect, but rather that the kind of existence an essence enjoys in
37 the intellect can help us to push forward our investigation into the metaphysical structure
38 of material things. Amos Bertolacci rightly pointed out to me that also the modal
39 considerations by means of which we distinguish between common nature and principle
of individuation are mental in character and hence the extra-mental commonality of an
40
essence can be fully grasped only in the mind. I agree with him. I would like only to
41 insist that modal considerations reveal the existence of extra-mental natures, while
42 abstraction by itself does not tell us anything about the metaphysical structure of the
1 may help us to see, for instance, that conceiving of the essence without the
2 principle of individuation is at least logically, even though not metaphysically,
3 possible – a claim, incidentally, which seems to be implied by Aquinas’s remark,
4 in his commentary on Met. Z 11, to the effect that logic can treat essences as if
5 they existed in separation from individuating conditions.43 However, our
6 discovery of the modal properties of essences cannot be reduced to the ability of
7 our intellect to get universal concepts through abstraction or, to put it better, to
8 the things’ potentiality to be known by the intellect in a universal form.44
9
10
things the concepts of which we come to acquire (for further development of this idea,
11 see below, n. 44).
12 43 By ‘logically possible’ I simply mean that there is nothing contradictory in supposing that
13 an essence exists without individuating conditions, as is testified to by its so existing in
14 the intellect. Given the metaphysical structure of the extra-mental world, however, the
essence of a sensible substance cannot exist without individuating conditions and its so
15
existing, therefore, is metaphysically, although not logically, impossible. Alternatively,
16 one could distinguish between two kinds of metaphysical possibility, i. e. absolute
17 metaphysical possibility (which amounts to the same thing as logical possibility) and
18 conditional metaphysical possibility. Absolutely speaking, there is no metaphysical
19 impossibility in supposing the essence of a sensible substance to exist without
individuating conditions. Conditionally, however, i. e. given the general metaphysical
20
presuppositions that govern God’s creation of the world (such as, for instance, the need
21 of introducing a hierarchy of beings with different degrees of perfection and
22 metaphysical composition) it is metaphysically impossible for the essence of a sensible
23 substance to exist without individuating conditions. Since I find this second, Leibniz-
24 style solution slightly anachronistic, I prefer to distinguish straightaway between logical
and metaphysical possibility.
25
44 One might try to soften the sharp contrast I have introduced between abstraction and
26 metaphysical or modal considerations by talking, as an alternative, of two different
27 interpretations of the process of abstraction, which we can label for simplicity ‘realist’
28 and ‘nominalist’ interpretation. According to the realist interpretation, abstraction is a
29 process of separation by which the intellect is able to isolate the common constituents of
things and separate them out from the individuating conditions along with which they
30
exist in the extra-mental world. One crucial assumption of this way of construing
31 abstraction is that there are common constituents in the extra-mental world, which only
32 need an intellect to be isolated and so brought into light. According to the nominalist
33 interpretation, by contrast, there are no common constituents in the extra-mental world,
34 but only similarities among things. Abstraction, therefore, is simply a process of selective
attention by means of which the intellect is capable of spotting the respects in which
35
things are similar and leave out those in which they are dissimilar. On my reconstruction,
36 Aquinas is clearly a supporter of the realist interpretation of abstraction. I wish to pursue
37 this way of making my point in a future paper. For now, I shall confine myself to
38 observing that both the realist and the nominalist interpretation of abstraction are
39 philosophical reconstructions of the natural process by which we acquire general concepts.
Thus, some distinction between the natural process of abstraction and its metaphysical
40
interpretation seems to be in order even if one prefers to speak of two ways of conceiving
41 the process rather than distinguishing between abstraction and metaphysical
42 considerations.
1 My general point about the compatibility between Avicenna’s doctrine of

2 essence and Aquinas’s metaphysical views can be formulated also in a slightly
3 different way. According to Aquinas’s doctrine in the De ente et essentia, the
4 extra-mental sensible world is populated by individuals. All the individuals
5 belonging to a certain species are instances of a certain essence: all human beings
6 are instances of the human nature. Therefore, in the extra-mental world essences
7 are individual, i. e. exist as individuals. However, the individuality of an essence
8 can be primitive or derivative. It is primitive if an essence needs nothing else to
9 be individual, whereas it is derivative if an essence needs something else, i. e.
10 something extrinsic, to become individual. And it is here that the metaphysical
11 analysis enters the picture. Such an analysis does not call into question the claim
12 that an essence is individual in its actual existence, but rather reveals that the
13 individuality of an essence is not primitive, in that it is due to a principle
14 external to the essence itself. Aquinas’s argument for the claim that in material
15 substances supposit and essence are mind-independently distinct contributes
16 precisely to establish the point that the individuality of an essence is not
17 primitive, but rather due to a principle distinct from it. To the same effect is
18 Aquinas’s observation that, if an essence could exist separately from the principle
19 of individuation, it would be just one and not multiplied. Reflecting upon the
20 way in which an essence would exist, if it could exist separately, shows once
21 again that its actual individuality is not primitive but rather derivative. And, as I
22 see things, being non-primitively individual is compatible with being actually
23 individual. For being primitively individual and being derivatively individual are
24 precisely properties we attribute to essences which are in fact actually individual,
25 and so attributing such properties in some sense presupposes that an essence be
26 individual in its actual, extra-mental existence. Thus, there can be no possible
27 conflict between the two claims.
28
29
30 4. Two Senses of ‘Common’
31
32 It seems to me, therefore, that the notion of ‘common nature’, by which
33 Aquinas often refers to the essence of sensible things, is in some sense
34 ambiguous and that disambiguation in this case requires distinguishing between
35 two different senses of ‘common’. If by ‘common’ we mean ‘actually common’ it
36 is clear that, according to Avicenna’s doctrine of essence, an essence is actually
37 common only in the intellect, where it exists as a universal concept. This notion
38 is clear enough, but a couple of things should be clarified. The first is that, when
39 Aquinas says that a concept is common or universal, he means to refer to the
40 representational content of a concept and not to its ontological status.45 A
41
42 45 See for this distinction: Aquinas, Exp. Peryer, lib. I, lect. 3, p. 15, lines 51 – 78.
1 concept, in other words, is common or universal because for each species there is
2 in the intellect just one concept representing all the individuals belonging to the
3 species in the same way, i. e. with respect to their common aspects. From the
4 ontological point of view, however, i. e. from the point of view of the kind of
5 thing a concept is, a concept is just a quality of the mind and hence is in this
6 sense individual. In any event, it is clearly the representational content that
7 Aquinas has in mind when he describes a concept as common or universal. The
8 second thing to be clarified is that the fact that a concept is common or
9 universal because it represents all the individuals of a certain kind is not
10 equivalent to and so should not be confused with the intellect’s conscious
11 attribution of universality to a concept.46 Like many other medieval
12 philosophers, Aquinas distinguishes in fact between first and second intentions.
13 First intentions are concepts such as ‘man’, ‘horse’ and the like, i. e. first-order
14
concepts that directly represent extra-mental things. Second intentions, by
15
contrast, are concepts such as ‘species’, ‘genus’, ‘particular’ and ‘universal’. Such
16
concepts do not directly represent extra-mental things but rather reflect the way
17
in which things are known by our intellect and so exist therein. Second
18
intentions, in other words, are second-order concepts, i. e. attributes of first-
19
order concepts. It is the concept ‘man’ – and not the particular men represented
20
by the concept ‘man’ – that can be called both a species and a universal. On
21
Aquinas’s standard account, the intellect obtains second-order concepts by
22
23
reflecting upon first-order concepts and in particular upon the way they
24
represent the extra-mental things falling under them. The intellect, for instance,
25
attributes the property of being universal to the concept ‘man’ after realising that
26
such a concept is a unique concept representing all men in the same way. Thus,
27 when I say that, according to Aquinas, an essence is universal in the mind in that
28 it exists therein as a universal concept, I do not mean to refer to universality
29 taken as a second intention, but rather to the objective fact that first-order
30 concepts – such as ‘man’, ‘horse’ and the like – represent all the individuals
31 falling under them regardless of whether the intellect in fact recognises the
32 representational function of these concepts or not. It is, in other words, the
33 universality that goes along with first intentions, and not that connected with
34 second intentions, that I am thinking of. In conclusion, our first sense of
35 ‘common’ in the expression ‘common nature’ can be summarised as follows:
36 according to Avicenna’s doctrine of essence or at least to Aquinas’s interpretation
37 thereof, an essence is actually universal only in the mind where it exists as a first-
38 order concept representing all individuals falling under it in the same way.
39
40
46 For texts where Aquinas introduces, explicitly or implicitly, this distinction see: Aquinas,
41 In I Sent., d. 2, q. 1, a. 3, vol. I, p. 67; Q. de pot., q. 7, a. 9, p. 201; S. th., Ia, q. 85, a. 2,
42 ad 3, pp. 412 – 3; a. 3, ad 4, p. 414; Sent. lib. de an., lib. II, c. 12, p. 116, lines 139 – 43.
1 If my reconstruction is correct, however, Aquinas presents us with a second

2 sense of ‘common’, i. e. the sense emerging from the analysis of the metaphysical
3 structure of material objects. This second sense of ‘common’ concerns the extra-
4 mental existence of an essence. Since an essence, however, exists in the extra-
5 mental world as an individual, the second sense of ‘common’ cannot be
6 understood as ‘actually common’. Given the importance that modal consid-
7 erations play in introducing the second sense of ‘common’, I shall refer to it
8 simply as ‘modally common’. As we have seen, the modal commonality of an
9 essence seems to amount to the basic idea that the individuality of an essence is
10 not primitive but rather derivative, in that it depends on a principle falling
11 outside the essence, the principle of individuation. In other words, although an
12 essence is individual in the extra-mental world, it is not individual by itself or
13 primitively but is rather individuated by something else. Thus, commonality or
14 sharability in the sense specified turns out to be a fundamental modal property
15 of an essence, i. e. a property which we discover by reflecting upon how an
16 essence would exist if it could exist separately, i. e. in separation from the
17 principle of individuation. Considerations about the modal property of an
18 essence, in so far as they concern the (metaphysically) counterpossible existence
19 of an essence, do not bear on the way an essence exists in actuality. Thus, being
20 actually individual and being modally common seem to be compatible
21 properties. In this sense, Aquinas’s endorsement of Avicenna’s doctrine of
22 essence does not clash with his general analysis of the individuation of material
23 substances.
24
25
26
Conclusions
27
I suppose that my general reading of Aquinas’s doctrine of essence could be
28
classified as realist as opposed to a more nominalist and deflationary one, in that
29
I insist a good deal on the mind-independentness of the distinction between
30
essence and supposit as well as on the intrinsic commonality and sharability of
31
essence. In the history of philosophy, however, general labels are sometimes
32
insidious and may even turn out to be not sufficiently powerful to capture the
33
differences between one philosophical position and another. Thus, what is
34
important is not so much whether Aquinas is a realist or not as which
35
conceptual tools he employs in defending his particular version of realism. With
36
regard to this, my general view is that the reconciliation between Avicenna’s
37
doctrine of essence and Thomas’s view on individuation can be achieved by
38
means of a set of three related distinctions:
39
(i) The distinction between the level of the actual existence of an essence and
40
the analysis of the metaphysical structure of a sensible substance.
41
42
1 (ii) The consequent distinction between two senses of ‘common’: ‘actually

2 common’ and ‘modally common’, i. e. common in the sense of not being
3 primitively, but rather derivatively individual. Avicenna’s doctrine of essence
4 establishes only that an essence is actually common only in the intellect, but
5 does not prevent an essence from being extra-mentally common when
6 ‘common’ is taken in the sense of ‘modally common’ – the sense which
7 concerns the metaphysical analysis of material substances.
8 (iii) Finally, distinctions (i) and (ii) presuppose a further distinction between
9 abstraction and the modal considerations that guide us in the analysis of the
10 metaphysical structure of material things. Abstraction is an automatic and
11 natural process leading to the formation of universal concepts. Modal
12 considerations, by contrast, are part of a deliberate philosophical investigation
13 into the structure of things. The fact that abstraction and modal considerations
14 are mental processes of so different a kind suggests that there is more to the
15 extra-mental commonness of an essence than our intellect’s natural ability to
16 understand material things without their individuating conditions. Reflection
17 upon the process of abstraction and the way in which an essence exists in the
18 mind may certainly be of some help in carrying through the metaphysical
19 analysis of extra-mental objects. However, the modal considerations that shape
20 our metaphysical analysis of extra-mental objects should not be confused with
21 or reduced to the natural process by which we acquire concepts and so sort the
22 world into different kinds of things.
23
24
25 Bibliography
26
27 Primary Sources
28
29 Aristotle, Metaphysics. A Revised Text with Introduction and Commentary, 2 vols, ed. W.D.
30 Ross, Oxford: Clarendon Press, 1924.
31 Averroes, Averrois Cordubensis in Aristotelis Metaphysicorum libros commentarium, in
32 Aristotelis opera cum Averrois commentariis, vol. VIII, Venice: apud Junctas, 1562 –
33
74, reprinted Frankfurt a.M.: Minerva, 1962.
Avicenna, Liber de philosophia prima sive scientia divina, I–IV, ed. S. Van Riet, Louvain/
35 –– , Liber de philosophia prima sive scientia divina, V–X, ed. S. Van Riet, Louvain/
36 Leiden: Peeters/Brill, 1980
37 Thomas Aquinas, Scriptum super librum Sententiarum, 2 vols, ed. R.P. Mandonnet,
38 Paris: Lethellieux, 1929.
–– , Liber de veritate catholicae fidei seu Summa contra gentiles, 3 vols, ed. C. Pera,
39
Taurini-Romae: Marietti, 1961.
40 –– , Summa theologiae, Rome: Editiones Paulinae, 1962
41 –– , In duodecim libros Metaphysicorum expositio, eds M.-R. Cathala and R. Spiazzi,
42 Taurini-Romae: Marietti, 1964.
1 –– , Quaestiones disputate, 2 vols, eds P. Bazzi, M. Calcaterra, T.S. Centi, E. Odetto,

2 P.M. Pessiom and R. Spiazzi, Taurini-Romae: Marietti, 1965.
3
–– , De substantiis separatis, cura et studio fratrum Praedicatorum, Opera omnia, t. XL,
Opuscula I, Rome: Editori di S. Tommaso, 1969.
4
–– , De ente et essentia, c. 3, cura et studio fratrum Praedicatorum, Opera omnia,
5 t. XLIII, Opuscula IV, Rome: Editori di S. Tommaso, 1976.
6 –– , De unitate intellectus contra Averroistas, cura et studio fratrum Praedicatorum, Opera
7 omnia, t. XLIII, Opuscula IV, Rome: Editori di S. Tommaso, 1976.
8 –– , Sententia libri de anima, cura et studio fratrum Praedicatorum, Opera omnia,
9
t. XLV, Rome/Paris: Commissio Leonina/Vrin, 1984.
–– , Expositio libri Peryermenias, cura et studio fratrum Praedicatorum, Opera omnia,
10
t. I* 1, Rome/Paris: Commissio Leonina/Vrin, 1989.
11 –– , Quaestiones disputate de anima, ed. B.-C. Bazn, Opera omnia, t. XXIV.1, Rome/
12 Paris: Commissio Leonina/Les ditions du Cerf, 1996.
13 –– , Quaestiones de quolibet, 2 vols, cura et studio fratrum Praedicatorum, Opera omnia,
14 t. XXV, Rome/Paris: Commissio Leonina/Les ditions du Cerf, 1996.
15
–– , Quaestio de spiritualibus creaturis, ed. J. Cos, Opera omnia, t. XXIV 2, Rome/Paris:
Commissio Leonina/Les ditions du Cerf, 2000.
16 –– , On the Power of God, transl. and ed. English Dominican Fathers, 3 vols., London:
17 Burns, Oates & Washbourne 1932–4 (repr. in one vol., Westminster, Maryland:
18 Newman 1952).
19
20
Secondary Sources
21
22
D. Black, Mental Existence in Thomas Aquinas and Avicenna, Documenti e studi sulla
23 tradizione filosofica medievale, 10, 1999, pp. 45 – 79.
24 R. Cross, The Metaphysics of the Incarnation, Oxford: Oxford University Press, 2002.
25 A. De Libera, La querelle des universaux de Platon la fin du Moyen Age, Paris: ditions
26 du Seuil,1996.
27
–– , L’art des gnralit. Thories de l’abstraction, Paris: Aubier, 1999.
G. Galluzzo, Aquinas on Common Nature and Universals, Recherches de Thologie et
28
Philosophie mdivales, 70, 2004, pp. 131 – 71.
29 –– , Aquinas’s Interpretation of Aristotle’s Metaphysics, Book Zeta, Recherches de
30 Thologie et Philosophie mdivales, 74(2), 2007, pp. 423 – 81.
31 –– , Averroes and Aquinas on Aristotle’s Criterion of Substantiality, Arabic Sciences and
32 Philosophy, 19, 2009, pp. 157 – 87.
33
E.J. Lowe, A Survey of Metaphysics, Oxford: Oxford University Press, 2002.
M.E. Marmura, Avicenna’s Chapters on Universals in the Isagoge of his Shifa’, in Islam:
34
Past Influence and Present Challenge, eds. E.T. Welsh and P. Cacia, Albany, New
35 York: State University of New York Press, 1979, pp. 34 – 6.
36 –– , Avicenna on Primary Concepts in the Metaphysics of his al-Shifa’, in Logos Islamikos:
37 Studia Islamica in honorem Georgii Michaelis Wickens, eds. R.M. Savory and D.A.
38 Agius, Toronto: Pontifical Institute of Mediaeval Studies, 1984, pp. 219 – 39.
39
–– , Quiddity and Universality in Avicenna, in Neoplatonism and Islamic Thought, ed.
P. Morewedge, Albany, New York: SUNY Press, 1992, pp. 77 – 87.
40 J. McGinnis, Logic and Science: The Role of Genus and Difference in Avicenna’s Logic,
41 Science and Natural Philosophy, Documenti e studi sulla tradizione filosofica
42 medievale, 18, 2007, pp. 165 – 86.
1 J. Owens, Common Nature: A Point of Comparison Between Thomistic and Scotistic

2 Metaphysics, Mediaeval Studies, 19, 1957, pp. 1 – 14.
3
G. Pini, Absoluta consideratio naturae: Tommaso d’Aquino e la dottrina avicenniana
dell’essenza, Documenti e studi sulla tradizione filosofica medievale, 15, 2004,
4 pp. 387 – 438.
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
1
2
3
4 On the Latin Reception of Avicenna’s Theory of
5 Individuation
6
7 Martin Pickav
8
9
I
10
11
It is well known that medieval philosophers and theologians in the Latin West
12
debated vigorously over the principle of individuation, i. e. over the principle
13
that accounts for the individuality of individuals and for the multiplication of
14
things belonging to one species. The debate was especially lively at the end of
15
the thirteenth and the beginning of the fourteenth centuries, a development no
16
doubt fueled by the condemnation of certain aspects of Thomas Aquinas’s
17
theory of individuation in 1277. During the controversy many items were
18
proposed as candidates for the principle of individuation. These candidates
19
include matter, form, matter and form together, quantity, relations, a negation,
20
existence, haecceity, etc.1
21
One view that was often mentioned but never seriously defended during the
22
height of the debate is the view that objects are individuated by their accidents
23
or by the aggregation of their accidents. This theory, to which later medieval
24
authors sometimes refer as the ‘old opinion’ (antiqua opinio)2, can be found
25
prominently in Porphyry and Boethius. In his short treatise On the Trinity,
26
Boethius, for instance, explains:
27
28 The variety of accidents brings about numerical difference. For three men differ
neither by genus nor by species but by their accidents. For even if we were to
29
separate by the mind all their accidents, the locations of each of them are still
30 different, which we cannot in any way make into one, because two bodies will not
31 occupy one location, which is an accident. And therefore they are numerically many,
32 because they become many through their accidents.3
33
And in Boethius’s translation of Porphyry’s Isagoge we can read the following:
34
35
36
37 1 See my The Controversy over the Principle of Individuation.
38 2 See, e. g., text I.6 in the appendix.
39 3 Boethius, De trinitate, c. 1, p. 168: ‘Sed numero differentiam accidentium varietas facit.
Nam tres homines neque genere neque specie, sed suis accidentibus distant; nam vel si
40
animo cuncta ab his accidentia separemus, tamen locus cunctis diversus est, quem unum
41 fingere nullo modo possumus: duo enim corpora unum locum non obtinebunt, qui est
42 accidens. Atque ideo sunt numero plures, quoniam accidentibus plures fiunt.’
340 Martin Pickav
1 Things like this are called individuals because each consists of properties, the
2 combination of which will never be the same in something else. For the properties
3
of Socrates will never be found in any other particular thing. But the properties of a
human being (I mean those that belong to a human being as something common)
4 will be the same in many, or rather, they will be the same in all particular human
5 beings insofar as they are human beings.4
6
7
The motivation for such a theory seems to derive from the fact that we indeed
8
identify individuals by their different sets of accidents. I identify two human
9
beings as two different individuals because each of them is at a different location,
10
has a different size, accent, clothing, etc. But it is less clear that we are entitled to
11
move from this fact about how we discriminate between individuals to a deeper
12
ontological claim concerning their individuation. However, on the other hand it
13
is plausible to say that each individual object possesses some properties that it
14
shares with other objects of the same kind and some properties that belong
15
solely to itself; surely an object is not individuated by any of the common
16
properties.
17 The Porphyrio-Boethian account of individuation was very influential in
18 Late Antiquity and in the Early Middle Ages.5 But whether accidental
19 individuation is a promising solution to the problem of individuation depends
20 largely on the ontological framework in which it is presented. It also
21 depends, of course, on whether accidental individuation can be formulated in
22 an unambiguous way. In the Boethius passage quoted above, for instance, it
23 remains open whether individuation is caused by one accident alone (spatial
24 location) or by a combination or bundle of all of its accidents. Moreover, are the
25 individuating accidents particular properties or are they common properties
26 which themselves need to be individuated by something else? However this may
27 be, the ‘old opinion’ looks extremely vulnerable if it is understood on the
28 background of a strictly Aristotelian understanding of accidents as entities that
29 are ontologically dependent on the substances in which they inhere. With the
30 Aristotelian background in mind medieval philosophers frequently advance one
31 or more of the following six arguments against the idea of accidental
32 individuation:
33
34 4 Porphyry, Isagoge II § 15 (translatio Boethii, pp. 13 f.; cf. versio Graeca, p. 7): ‘Indiuidua
ergo dicuntur huiusmodi quoniam ex proprietatibus consistit unumquodque eorum
35
quorum collectio numquam in alio eadem erit; Socratis enim proprietates numquam in
36 alio quolibet erunt particularium; hae uero quae sunt hominis (dico autem eius qui est
37 communis) proprietates erunt eaedem in pluribus, magis autem et in omnibus
38 particularibus hominibus in eo quod homines sunt.’ See also Boethius’s commentary on
39 this passage in In Isagogen Porphyrii (ed. sec.) III, c. 11, pp. 253 f. and IV, c. 1, pp. 241 f.
5 See Chiaradonna, La teoria dell’individuo in Porfirio; Erismann, Collectio proprieta-
40
tum; id., L’individualit explique par les accidents; Gracia, The Problem of
41 Individuation in the Early Middle Ages; King, The Problem of Individuation,
42 pp. 163 – 7; Sorabji, Self, pp. 137 – 53.
On the Latin Reception of Avicenna’s Theory of Individuation 341
1 (1) Each individual substance is a per se being (ens per se). But if a substance
2 owes its individuality to the reception of an accident or an aggregation of
3 accidents, then the individual would essentially be a composite of a substance
4 and an accident and thus merely a being per accidens. In other words: accidental
5 individuation makes all individuals into composites and takes away the essential
6 unity individual substances seem to enjoy.6
7 (2) Substances are ontologically prior to their accidents. For under normal
8 circumstances accidents exist only insofar as – and because – they inhere in
9 substances. But if substances are ontologically prior to accidents, then it also
10 follows that an individual substance is ontologically prior to its particular
11 accident. And this rules out that accidents are capable of accounting for the
12 individuality of that on which they depend.7
13 (3) Individual (or numerical) unity is a property that belongs to every
14 existing thing by itself. For if numerical unity were to come about by something
15 added, say, by the addition of an accident, then one could ask about that added
16 thing whether or not it has numerical unity by itself or from something else. If
17
the latter is the case, we are at risk of an infinite regress, for then we have to ask
18
with regard to this other thing whether it has numerical unity by itself or from
19
something else. If, however, the former is true then one might wonder why the
20
accident is numerically one, but the object to which it is added is not.8
21
(4) In each category (or class of being) there are genera, species, and
22
individuals. Now it would destroy the division between the categories if a species
23
(say, of the category of substance) requires, in order to be a species (instead of
24
being a genus), some element from a different category. The same reasoning
25
applies to individuals belonging to each of the categories.9
26
(5) Since accidents are in themselves common, they are incapable of turning
27
28
something into a particular this. An aggregation of many different accidents will
29
6 See, e. g., Peter John Olivi, Quodlibet III, q. 4, pp. 177 f.; John Duns Scotus, Quaestiones
30
super libros Metaphysicorum Aristotelis, lib. VII, q. 13, n. 21, pp. 224 f.; id., Lectura II,
31 dist. 3, p. 1, q. 4, n. 65, p. 247; id., Ordinatio II, dist. 3, p. 1, q. 4, n. 70, pp. 422 f. Here
32 and in the following footnotes I only mention a few examples of authors presenting the
33 various arguments. Basically all the authors in part I of the appendix (and many of part
34 II) bring forward one or more of the following reasons against accidental individuation.
7 See, e. g., Godfrey of Fontaines, Quodlibet VII, q. 5, p. 320; Peter John Olivi, Quodlibet
35
III, q. 4, p. 177; John Duns Scotus, Quaestiones super libros Metaphysicorum Aristotelis,
36 lib. VII, q. 13, nn. 24 – 6, pp. 225 – 7; id., Lectura II, dist. 3, p. 1, q. 4, n. 79, pp. 252 f.;
37 id., Ordinatio II, dist. 3, p. 1, q. 4, nn. 82 – 8, pp. 429 – 33.
38 8 See, e. g., John Duns Scotus, Quaestiones super libros Metaphysicorum Aristotelis, lib. VII,
39 q. 13, n. 27, p. 227; id., Lectura II, dist. 3, p. 1, q. 4, nn. 73 – 8, pp. 250 – 52; id.,
Ordinatio II, dist. 3, p. 1, q. 4, nn. 76 – 81, pp. 426 – 9.
40
9 See, e. g., John Duns Scotus, Quaestiones super libros Metaphysicorum Aristotelis, lib. VII,
41 q. 13, nn. 28 – 30, pp. 227 f.; id., Lectura II, dist. 3, p. 1, q. 4, nn. 91 – 4, pp. 258 f.; id.,
42 Ordinatio II, dist. 3, p. 1, q. 3, n. 63, pp. 419 f. and ibid., q. 4, nn. 89 – 98, pp. 433 – 8.
342 Martin Pickav
1 of course restrict the commonality of the set of properties, but it will never be
2 able to rule out that there can, at least in principle, be another item instantiating
3 exactly the same set of properties. Individuals, on the other hand, can by
4 definition never be instantiated more than once.10
5 (6) Two or more individuals differ from each other as substances; in other
6 words: they differ substantially. Plato is substantially different from Socrates.
7 But accidents can only account for an accidental difference and therefore fail to
8 account for the robust difference between individuals.11
9 Some of these arguments against accidental individuation were first
10 formulated by Peter Abelard in the first half of the twelfth century.12 And
11 Abelard’s criticism seemed to have had the effect that ‘no medieval philosopher
12 argued seriously for accidental individuation after Abelard proposed his
13 objection.’13 All this is to say that in the thirteenth and the fourteenth centuries
14 the theory of accidental individuation was generally considered a bad idea and
15 something with which most people preferred not to be associated. The latter is
16 in particular true for defenders of the view according to which material objects
17 are individuated by quantity, i. e. the quantitative dimensions of the underlying
18 matter. Although proponents of quantitative individuation are often attacked by
19 their opponents using one or more of the arguments mentioned above, these
20 proponents are eager to disassociate themselves from the accidental individu-
21 ation view.14 The quantity involved in individuating material objects, so they
22 argue, is not an accident in the common sense.
23
24
25
10 See, e. g., Peter of Auvergne, Quodlibet II, q. 5, edited in Hocedez, Une Question indite
26 de Pierre d’Auvergne, p. 371; John Duns Scotus, Quaestiones super libros Metaphysicorum
27 Aristotelis, lib. VII, q. 13, n. 32, p. 229.
28 11 See, e. g., Bonaventure, In II Sententiarum, dist. 3, p. 1, a. 2, q. 2, p. 106a; Godfrey of
29 Fontaines, Quodlibet VII, q. 5, p. 321; Peter John Olivi, Quodlibet III, q. 4, p. 177.
12 See King, Metaphysics, pp. 72 – 5.
30
13 King, The Problem of Individuation, p. 167. The fact that no one positively endorsed
31 accidental individuation after Abelard is perhaps the true reason why later (13th- and
32 14th-century) authors refer to this position as the ‘old opinion.’
33 14 Thomas Aquinas is somewhat an exception. In his Super Boetium de trinitate he explicitly
34 argues that one accident, i. e. quantity, does indeed account for individuation (q. 4, a. 2,
pp. 122 – 6). But he also explains that the individuating quantity is not an accident in the
35
complete sense. For more detail on this see Wippel, The Metaphysical Thought of Thomas
36 Aquinas, pp. 351 – 75. Another exception is Albert the Great (at least in his earlier
37 works). For Albert, individuals are ‘individuated’ by their accidents, but since each
38 individual is also a substance and a subject, it is also true that the singularity of the
39 substance or subject is due to a more profound principle, which is not an accident nor an
aggregation of accidents. In other words, by using the notion ‘individual’ in a slightly
40
peculiar way (namely for something that is already composed of a substance and
41 accidents), Albert arrives at a harmonizing interpretation according to which both the
42 proponents and the opponents of accidental individuation come out to be correct. For
1 II
2
3 At the beginning of the fourteenth century it is common to find references to
4 Avicenna as one of the proponents of accidental individuation. Good examples
5 are the various texts in which John Duns Scotus deals with the problem of
6 individuation. In many of them Avicenna appears next to Porphyry and
7 Boethius (see texts I.7, I.8, and I.9 in the appendix). However, there is no reason
8 to believe that Duns Scotus was the first to consider Avicenna’s theory of
9 individuation as basically the same as the Porphyrio-Boethian theory. The
10 editors of Duns Scotus’s Quaestions on the Metaphysics notice that Scotus’s
11 arrangement of philosophical authorities is very likely based on the arrangement
12 Scotus found in the discussion of the principle of individuation in q. 5 of
13 Godfrey of Fontaines’s seventh Quodlibet. There too we find Avicenna next to
14 Porphyry and Boethius (see text I.5 in the appendix). But it is easy to find other
15 medieval authors who present Avicenna in the same way: Peter of Falco (I.1),
16 Henry of Ghent (I.2), Richard Middleton (I.3), Roger Marston (I.4), John of
17 Paris (I.10), and Peter of Auvergne (I.6). I have no reason to believe that this list
18 is complete.
19 But are these later medieval philosophers correct in considering Avicenna as
20 a proponent of accidental individuation? The Latin Avicenna no doubt uses the
21 language of accidents when he talks about individuation. One passage that is
22 often quoted by the authors mentioned above is from chapter 4 of the fifth
23 treatise of Avicenna’s Liber de philosophia prima:
24
And then the nature [of a thing] will be such that necessary concomitants befall it,
25 necessary concomitants consisting of properties and accidents through which the
26 nature is individuated and becomes designated.15
27
28
Another passage popular among Latin authors can be found a little bit earlier in
29
chapter 2:
30 A nature that requires matter does not possess existence unless the matter has been
31 prepared [for its reception]; accidents and dispositions thus befall its existence from
32
the outside, and through these accidents and dispositions the nature is
individuated.16
33
34
Albert’s account of individuation see Roland-Gosselin, Le ‘De ente et essentia’ de S.
35
Thomas d’Aquin, pp. 89 – 103.
36 15 Avicenna, Liber de philosophia prima, tr. 5, c. 5, p. 264: ‘Et erit tunc natura sic quod
37 accident ei concomitantia ex proprietatibus et accidentibus, per quae natura individuatur
38 et fit designata.’ For Latin authors who quote this passage see texts I.2, I.5, I.6 and I.10
39 in the appendix.
16 Ibid., c. 2, p. 240: ‘Quae vero ex istis naturis eget materia, non habet esse nisi cum
40
materia fuerit praeparata; unde ad eius esse adveniunt accidentia et dispositiones
41 extrinsecus per quae individuatur.’ Scotus refers to this passage in texts I.7, I.8 and I.9 in
42 the appendix. See also Liber de philosophia prima, tr. 3, c. 3, p. 117: ‘Dico igitur quod
344 Martin Pickav
1 Both texts are taken from Avicenna’s famous account of essences in which he
2 develops his idea that quiddities or essences can be considered in three ways:
3 (1) absolutely and in themselves, (2) insofar as they are in the mind, and (3)
4 insofar as they exist outside the mind.17 Insofar as essences exist in the mind they
5 have the property to be universal; insofar as they exist in things outside the
6 mind, an essence is singular. To be universal or to be singular are thus accidental
7 with respect to the essence insofar as the essence is considered in itself. But it is
8 one thing to say that to be universal and to be singular is accidental for an
9 essence and it is another thing to say that essences and the objects possessing
10 such essences are individuated by accidents.
11 There is yet another, albeit closely related, reason suggesting that Avicenna
12 endorses accidental individuation. This reason has to do with Averroes’s
13 criticism of Avicenna’s account of unity. In his Long commentary on Aristotle’s
14
Metaphysics, Averroes famously accuses Avicenna of teaching that unity is an
15
accidental feature. Or in medieval terms: Averroes reprimands Avicenna for
16
confusing transcendental unity with the unity which belongs to the category of
17
quantity, i. e. the unity which is the principle of number.18 But if unity is
18
supposed to be an accidental feature, so is individuality, because individuality is
19
apparently nothing else than a specific kind of unity. Latin authors discuss
20
Averroes’s criticism frequently, and although they usually defend Avicenna from
21
the accusations, they are by the same route familiar with this particular reading
22
23
of Avicenna’s account of unity and individuation.19
24
However, there are also serious reasons to be doubtful as to whether
25
Avicenna really thought that objects are individuated by accidents or by an
26
aggregation or bundle of accidents. In the following text from the Liber de
27 philosophia prima Avicenna anticipates one of the above-mentioned standard
28 objections to accidental individuation (objection 5):
29
unitas vel dicitur de accidentibus vel dicitur de substantia; cum autem dicitur de
30
accidentibus, non est substantia, et hoc est dubium; cum vero dicitur de substantiis, non
31 dicitur de eis sicut genus nec sicut differentia ullo modo: non enim recipitur in
32 certificatione quidditatis alicuius substantiarum, sed est quiddam comitans substantiam,
33 sicut iam nosti. Non ergo dicitur de eis sicut genus vel sicut differentia, sed sicut
34 accidens. Unde unum est substantia, unitas vero est intentio quae est accidens.’
17 For a detailed analysis of the argument of Liber de philosophia prima, tr. 5, chapters 1 – 4
35
see de Libera, L’art des gnralits, 499 – 607.
36 18 Averroes, Long Commentary on Aristotle’s Metaphysics, lib. IV, com. 3, in Aristotle, Opera
37 cum Averrois commentariis, vol. 8, fol. 67rB, D–E: ‘Avicenna autem peccavit multum in
38 hoc, quod existimavit, quod unum et ens significant dispositiones additas essentiae rei …
39 Et etiam, quia existimavit, quod unum dictum de omnibus praedicamentis, est illud
unum quod est principium numerorum. Numerus autem est accidens. Unde opinatus
40
fuit iste, quod hoc nomen unum significat accidens in entibus.’
41 19 For some aspects of the medieval discussion triggered by Averroes’s criticism see de
42 Libera, D’Avicenne Averros, et retour.
1 If A is a universal intention and B, another universal intention, is added to it, then

2 there results from this some restriction. But because this is the restriction of one
3
universal by some other universal, there remains a more restricted universal that can
nevertheless be shared by others. If, for example, this person Socrates is defined and
4 it is said that he is a philosopher, then there is sharing with others. And if it is said
5 that he is a chaste philosopher, there is still sharing with others, and if it is said that
6 he is a chaste philosopher who was unjustly killed, even this is still shared with
7 others. … And if it is added that he was killed in this city at this day, then this whole
8
enumeration with its specific uniqueness is still completely universal; for it is
possible to predicate this enumeration of many things, unless it is restricted to one
9
individual.20
10
11 Strictly speaking, Avicenna is here only concerned with the definability of
12 individuals. His point is that it is impossible to define an individual by simply
13 enumerating all of his properties, for although such an enumeration becomes
14 less and less universal with each property added, it never yields a description that
15 necessarily fits only one individual. But the point made can easily be translated
16 into a metaphysical claim: an aggregation of (common) accidents never renders
17 a substance individual. And some medieval readers read the above-mentioned
18 passage exactly in this metaphysical sense.21
19 In fact, when Avicenna explicitly addresses his own account of individuation
20 he clearly presents himself as a defender of the traditional Aristotelian view
21 according to which objects are essentially individuated through matter.22 This is
22
23 20 Avicenna, Liber de philosophia prima, tr. 5, c. 5, pp. 276 f.: ‘Cum enim a fuerit intentio
24 universalis et ei adiungitur b, alia scilicet similiter intentio universalis, poterit tunc hic
esse restrictio aliqua. Sed quia est restrictio unius universalis per aliud universale, remanet
25
postea propinquius universale in quo potest esse communio. Verbi gratia, cum definitur
26 hic Socrates et dicitur esse “philosophus”, in hoc etiam communio est; cum vero dicitur
27 “philosophus castus”, adhuc etiam communio est; si vero dicitur “philosophus castus qui
28 occisus fuit iniuste”, adhuc etiam hoc commune est. … Si vero additur “quem occiderunt
29 in civitate illa et in die illo”, haec etiam tota enumeratio cum sua individualitate omnino
universalis est; possibile est enim illam praedicari de multis, nisi restrinxeris circa
30
individuum.’ A similar argument can also be found in the Isagoge part of Avicenna’s
31 Shifa’, chapter I, 12; see Marmura, Avicenna’s Chapter on Universals in the Isagoge of his
32 Shifa’, pp. 50 f.
33 21 See Peter of Auvergne, Quaestiones in Metaphysicam, lib. VIII, q. 25, edited in Hocedez,
34 Une Question indite de Pierre d’Auvergne, p. 381: ‘… et quotcumque accidentia
accedant ad hoc, totum aggregatum commune est, sicut dicit Avicenna, quia si
35
componentia sunt communia, et compositum. Manifestum ergo quod accidens non est
36 causa individuationis.’ Peter of Auvergne is a strange case. Why does he consider
37 Avicenna as a proponent of accidental individuation in his later quodlibetal questions
38 (see text I.6 in the appendix) when he is aware in his earlier Metaphysics commentary that
39 this cannot be Avicenna’s position? The puzzle can maybe explained with the peculiar
character of the quodlibetal question, which is basically a paraphrasis of Godfrey of
40
Fontaines’s view on individuation.
41 22 For Avicenna’s theory of individuation see Bck, The Islamic Background; Black,
42 Individuation, Self-Awareness, and God’s Knowledge of Particulars.
346 Martin Pickav
1 true even for immaterial objects like souls. In the following passage from his
2 Liber de anima he writes:
3
The multiplicity of things either derives from the essence and the form; or it derives
4 from the relation to matter and to the multiple origin from places which surround
5 every matter with respect to something else; or the multiplicity derives from the
6 times proper to the each of the things that befall it with its accidents; or it derives
7 from the causes which divide matter. But souls are not different in essence and form,
8 for they have the same form. Their difference is therefore due to the recipient of
their essence to which the essence is related as its own, and this is the body. But if
9
the soul were to exist without the body, then it would be impossible that one soul
10 differs from another in number. And this applies generally to all things. For in the
11 case of things which are many and whose essences are mere intentions, their species
12 have been multiplied in their individuals, and their multiplicity exists only on
13 account of their subjects and their recipients and the things that are affected by
14
them or it exists on account of some relation to these or to their times.23
15 Although Avicenna here strictly speaking only says that souls are individuated
16 and multiplied by their relationships to bodies, it is obvious that bodies account
17 for multiplication qua being material and not qua possessing yet another form, a
18 forma corporeitatis. For then the question of individuation would just reappear at
19 a different level. Note also that Avicenna explicitly says that this account of
20 individuation applies ‘generally to all things’.
21 All this calls for a conciliatory interpretation of the respective roles matter
22 and accidents play in Avicenna’s theory of individuation. The picture supported
23 by the text seems to be the following : Matter is the primary principle of
24 individuation, because matter individuates forms and thus also form-matter
25
composites. But once the individual forms or the individual form-matter
26
composites exist they will immediately receive a variety of accidents, which
27
can in a broader sense also to be said to contribute to their individuation.24 For
28
it is through such accidents that we identify (and discriminate among) different
29
30
23 Avicenna, Liber de anima, tr. 5, c. 3, pp. 105 f.: ‘Multitudo enim rerum aut est ex
31 essentia et forma, aut est ex comparatione quae est ad materiam et originem
32 multiplicatam ex locis quae circumdant unamquamque materiam secundum aliquid
33 aut ex temporibus propriis uniuscuiusque illarum quae accidunt illis accidentibus, aut ex
34 causis dividentibus illam. Inter animas autem non est alteritas in essentia et forma: forma
enim earum una est. Ergo non est alteritas nisi secundum receptibile suae essentiae cui
35
comparatur essentia eius proprie, et hoc est corpus. Si autem anima esset tantum absque
36 corpore, una anima non posset esse alia ab alia numero. Et hoc generaliter est in
37 omnibus; ea enim quorum essentiae sunt intentiones tantum et sunt multa, quorum
38 multiplicatae sunt species in suis singularibus, non est eorum multitudo nisi ex
39 sustinentibus tantum et receptibilibus et patientibus ex eis, aut ex aliqua comparatione ad
illa aut ad tempora eorum.’ On this text see Druart, The Human Soul’s Individuation;
40
Sebti, Avicenne, pp. 25–36. For a list of medieval authors drawing on Avicenna’s theory
41 in Liber de anima, tr. 5, c. 3 see Hasse, Avicenna’s ‘De anima’ in the Latin West, p. 297.
42 24 This seems to be exactly what is said in the text in n. 16 above.
1 individuals in the first place. Yet strictly speaking, accidents – whether they
2 come alone or in bundles – are for Avicenna not responsible for individuation.25
3
4
5 III
6
7 If, however, Avicenna himself never held the view that objects are primarily
8 individuated by accidents or bundles of accidents, does this mean that all his
9 Latin readers in the thirteenth and fourteenth centuries got him completely
10 wrong? By no means. There are actually many authors who are quite aware that
11 Avicenna never defended accidental individuation. William of Auvergne and
12 John of La Rochelle, for example, correctly ascribe to Avicenna the doctrine that
13 objects are primarily individuated by matter (see texts II.1 and II.2 in the
14 appendix)26. Moreover, despite the impression created by the texts in the first
15 part of the appendix, most authors never mention Avicenna when they refer to
16 accidental individuation. It might be enough here to point at some examples:
17 Bonaventure (II.4), Roger Bacon (II.5), Robert Kilwardby (II.6), William de la
18
Mare (II.7), Giles of Rome (II.8), and Peter John Olivi (II.9). Bacon and Giles
19
are especially interesting, because they seem to be keen on adding names to the
20
list of the traditional bundle theorists Porphyry and Boethius; Bacon adds al-
21
Ġaza-lı̄’s name to his list, Giles mentions Simplicius. Why don’t they also add
22
Avicenna?
23
Yet the main question with which we are confronted is this: What moves
24
authors such as Henry of Ghent, Roger Marston, Godfrey of Fontaines, and
25
John Duns Scotus to regard Avicenna as an accidental individuation theorist?
26
27
And why do they seem to adopt an interpretation of Avicenna that is obviously
28
wrong? A possible solution to this puzzle may appear once we have discovered
29
the origin of the strange association of Avicenna, Porphyry and Boethius.
30
Avicenna, however, is not the only one whom the aforementioned authors
31 associate with Porphyry and Boethius; almost all the passages in question also
32 mention John Damascene. Now it is well known that both Avicenna and John
33 Damascene play a very prominent role in question 8 of Henry of Ghent’s
34 second Quodlibet (disputed in the Advent of 1277). In this quodlibetal question
35 Henry reacts to Aquinas’s recently condemned view according to which matter is
36 the sole principle of individuation.27 The text has a peculiar structure: In his
37
38 25 John of La Rochelle (text II.2 in the appendix) is a good medieval example for this
39 interpretation of Avicenna’s theory of individuation.
26 See also text II.3 from Peter of Tarantaise’s Sentences commentary.
40
27 Henry mentions the condemned articles explicitly. See Quodlibet II, q. 8, p. 45. On this
41 text see also Brown, Henry of Ghent; Aertsen, Die Thesen zur Individuation in der
42 Verurteilung von 1277; Suarez-Nani, Les anges et la philosophie, pp. 75 – 85.
348 Martin Pickav
1 reply Henry first turns to ‘the Philosopher’ (i. e., Aristotle) and rejects what he
2 has to say (or not to say) about individuation of immaterial substances, then
3 Henry moves on to ‘our philosophizers’ (nostri philosophantes), followers of
4 Aristotle’s account of individuation in Henry’s time, who are similarly criticized,
5 and finally he turns to ‘our saints’ (nostri sancti), who adhere, according to
6 Henry, to the correct theory of individuation.
7 Because they consider matter as the only principle of individuation, ‘the
8 Philosopher’ and ‘the philosophizers’ hold that without matter there cannot exist
9 multiple individuals belonging to one and the same species. ‘Our saints’, such as
10 John Damascene, teach on the other hand that there can be different immaterial
11 entities belonging to one species and that they are distinct from each other by
12 means of ‘characteristic properties’ (characteristicae proprietates). To illustrate
13 John’s point Henry quotes a longer passage from John’s Elementary Introduction
14 to Dogma:
15
Each thing through which a species differs from another species is called substantial
16 and natural and constitutive difference and natural quality and natural property and
17 property of nature. And so human being and ox differ from each other, because a
18 human being is rational, but an ox is irrational; and ‘rational’ is the substantial and
19 constitutive difference of human being, but ‘irrational’ is that of ox. The same
applies to the other species, substances and natures and forms. However, each thing
20
through which a hypostasis of the same species differs from a co-substantial
21 hypostasis is called adventitious difference and quality and hypostatic and
22 characteristic property. But this is an accident.28
23
To see John Damascene in this context is not surprising; authors of the
24
thirteenth and fourteenth centuries quote him frequently as an authority when
25
they deal with the nature of angels and immaterial substances in general.29 It is
26
27
28 28 Henry of Ghent, Quodlibet II, q. 8, pp. 47 f.: ‘Quomodo ergo et penes quid accipitur in
29 immaterialibus et incorporalibus individuatio et specierum distinctio, Damascenus iam
tetigit cum dixit: “characteristicis proprietatibus”. Quid autem appellet “characteristicam
30
proprietatem”, exposuit prius ubi dixit in eodem libro: “Omnis res qua differt species ab
31 altera specie, ‘substantialis’ et ‘naturalis’ et ‘constitutiva differentia’ dicitur et ‘qualitas
32 naturalis’ et ‘naturalis proprietas’ et ‘proprium naturae’, veluti differunt homo et bos ad
33 invicem, quoniam homo quidem rationalis est, bos autem irrationalis, et est rationale
34 substantialis differentia et constitutiva hominis, irrationale autem bovis. Similiter et in
reliquis speciebus, substantiis et naturis et formis. Omnis autem res, in qua differt
35
hypostasis ab eiusdem speciei et consubstantiali hypostasi, dicitur ‘adventitia differentia’
36 et ‘qualitas’ et ‘hypostatica’ et ‘characteristica proprietas’. Haec autem est accidens.”’ For
37 the Latin text of De institutione elementari (in Robert Grosseteste’s translation) see
38 H. Gravius’s edition of John’s works.
39 29 John’s appearance may also help to explain why Henry contrasts nostri sancti with the
philosophus and the philosophantes. In another text, his Dialectica, John refers to the
40
teaching of the ‘holy Fathers’ (sancti patres), when he explains that there is an angelic
41 species under which we can find a multitude of individuals. See Dialectica (translatio
42 Roberti Grosseteste), c. 11, p. 13. For other references to sancti patres in this context see
1 with the aim of providing an interpretation of John Damascene’s account of

2 individuation that Avicenna enters the picture at this point. One of the first
3 things Henry remarks is that John and Avicenna use the term ‘accident’ in the
4 same broad sense as referring to something that affects essences and is outside
5 the natures of the essences themselves.30 So to talk about accidental
6 individuation in this sense of accident means not that something is individuated
7 by an item belonging to some of the accidental categories (i. e., quantity, quality,
8 relation, location etc.), but only that the individuating factor is not part of the
9 essence of the individuandum.
10 But what is then the ‘accident’ causing the individuation of an immaterial
11 substance (and of substances in general), if it is not one or more items belonging
12 to the accidental categories? For Henry the principle of individuation is nothing
13 other than the subsistence (subsistentia) of the substance, i. e. the special mode of
14 being that a substance enjoys when it exists in reality. At the end of his response
15 he summarizes his view in the following way:
16
And so it has to be said that (if the essence of an angelic nature or the essence of
17 whatever creature does not oppose this) God can make that two individuals – also
18 two that are separated from any accident (if it is possible at all for a substance to
19 exist without any accident) – exist under one ultimate species, so that the two
20 individuals, in what is substantial to them alone, are distinct through diverse
21
subsistences in effect.31
22 Let me make two comments on this quotation and the picture emerging from
23 it: (1) As the talk of ‘subsistence in effect’ (subsistentia in effectu) on the one
24 hand and God on the other hand indicates, Henry distinguishes between remote
25 and proximate (or precise) causes of individuation. Subsistence is the proximate
26 cause, but the agent by whose action a substance exists in effect is also a cause or
27 principle of individuation.32 And because material substances exist or subsist in
28 matter, matter too can be called a principle of individuation. Aristotle and the
29
ibid., c. 30, pp. 27 f. The passages from the Dialectica will show up prominently in
30
William de la Mare’s Correctorium fratris Thomae (written between 1277 and 1282, i. e.,
31 at the same time or briefly after Henry’s quodlibetal question). See q. 11, edited in
32 Glorieux, Les premires polmiques thomistes, pp. 60 f. William ends his attack on Aquinas
33 with the line (ibid.): ‘Ecce quod angelus est species specialissima, tam secundum
34 doctrinam philosophorum quam secundum doctrinam Sanctorum Patrum.’ Although
Henry does not mention the Dialectia passage(s) in his quodlibetal question, this
35
common emphasis on John Damascene and the views of the ‘saints’ can be no
36 coincidence.
37 30 Henry of Ghent, Quodlibet II, q. 8, p. 48.
38 31 Ibid., p. 53: ‘Sic ergo dicendum quod, non repugnante essentia naturae angelicae vel
39 cuiuscumque creature essentia, duo individua, etiam absoluta ab omni accidente (si
tamen possibile sit substantiam aliquam creatam existere sine omni accidente), possunt
40
fieri a Deo sub una specie specialissima, in solis sibi substantialibus distincta per diversas
41 subsistentias in effectu.’
42 32 In this sense, Henry calls God the ‘first principle of individuation’, see ibid., p. 51.
350 Martin Pickav
1 ‘philosophizers’ therefore were wrong not in that they regarded matter as a

2 principle of individuation, but that they regarded matter as the only principle of
3 individuation.33 (2) Although Henry refers to subsistence as an ‘accident’ in the
4 broad sense (for to subsist as this or that individual is accidental to the essence),
5 he clearly distinguishes between different meanings of ‘accident’. Accidents in
6 the strict sense, i. e. really distinct properties inhering in substances, accidents to
7 which Henry refers as ‘posterior accidents’ (accidentia posteriora), are not
8 capable of individuating substances; they allow us only to identify individuals
9 and discriminate among them. Finally, Henry quotes Porphyry as someone who
10 really only talks about accidents and bundles of accidents as principles for
11 identifying individuals.34
12 In summary, according to Henry, the principle of individuation is
13 something outside of the essence or form of the individuated object. In this
14 sense, and in this sense alone, the principle of individuation is an accident. But
15 because immaterial substances consist of more than just their essences, a
16 multitude of individuals under the same species is possible. With respect to
17 John Damascene and Avicenna, Henry intends to create the impression that his
18 account of individuation and of the quasi-accidental nature of the individuating
19 principle is shared by the two authorities. Yet insofar as he completely suppresses
20 any indication of the crucial role matter plays in Avicenna’s account of
21 individuation, Henry somewhat ‘misinterprets’ Avicenna. Yet I think there is an
22 obvious explanation for this ‘misinterpretation’.
23 As it is not surprising to see John Damascene quoted in a text about angelic
24 individuation, so it is not surprising to see Avicenna used in the same context.
25 But usually the two authorities appear on opposing sides; Avicenna is normally
26 considered as one of those philosophers who deny that without matter there can
27 be a multiplicity of individuals belonging to one species, whereas for John such
28 a multitude is possible. In other words, Avicenna is commonly regarded as a
29 proponent of the very theory Henry rejects.35 One of Henry’s implicit goals in
30
31 33 See also ibid., pp. 47 and 51.
32 34 Ibid., ad arg., pp. 54 – 5 and 57. For more on Henry’s theory of individuation see my
33 Henry of Ghent on Individuation.
34 35 See, e. g., Peter of Tarentaise (Pope Innocent V), In II Sententiarum, dist. 3, q. 2, a. 1,
p. 34b: ‘Non est simile, quia animae ut dicit Avicenna individuantur per materias
35
corporum, quamvis ab eis separatae retineant individuationem, sicut cera impressionem
36 sigilli; vel saltem per relationem inclinationis naturalis ad corpus determinatum; angeli
37 vero non sic habent materiam per quam individuentur’; Thomas Aquinas, De ente et
38 essentia, c. 3, p. 376b: ‘Sed cum essentia simplicis non sit recepta in materia, non potest
39 ibi esse talis multiplicatio; et ideo oportet ut non inueniantur in illis substantiis plura
indiuidua eiusdem speciei, sed quot sunt ibi indiuidua, tot sunt ibi species, ut Avicenna
40
expresse dicit’; Thomas Sutton, Quaestiones ordinariae, q. 27, pp. 752 f.: ‘Ergo natura
41 specifica angeli nullo modo multiplicatur. Istam rationem ponit Avicenna sub his verbis:
42 “Ex his autem naturis, quae non eget materia ad permanendum vel ad incipiendum, si
1 Quodlibet II, q. 8 is to give what he considers the correct interpretation of

2 Avicenna’s account of individuation.
3 According to Henry of Ghent, ‘the Philosopher’ and ‘the philosophizers’
4 commit a simple mistake in claiming that the absence of matter precludes
5 multiplication of members under one and the same species. For their conclusion
6 would only follow if the essence of an immaterial angelic species were capable of
7 existing by itself. Then there would indeed be no room at all for something that
8 could cause their individuation. If on the other hand such an essence does not
9 exist by itself, but requires something further to exist actually, then there is no
10 reason to believe that the mere absence of matter renders such a multiplication
11 impossible. In Henry’s words, the proponents of the traditional Aristotelian
12 account of individuation are mistaken, for they consider immaterial substances
13 as identical with their essences and thus as necessary beings, ‘many gods’. But as
14 Avicenna has noticed correctly, the essences of immaterial substances are
15 indifferent with regard to their actual existence and they are therefore also
16
indifferent with respect to their existence as individuals. For this reason, Henry
17
holds Avicenna to be committed to the view that immaterial substances do allow
18
for multiple members of one and the same species.36 In his quodlibetal question
19
Henry never explicitly addresses how to interpret Avicenna. But this is what he
20
no doubt would have said if he had been asked about the correct way of
21
understanding Avicenna’s teaching on individuation: Although Avicenna
22
sometimes mentions the role matter plays in individuation, matter is not the
23
exact principle of individuation, because it doesn’t apply to all beings.
24
Immaterial beings are also capable of individuation because, as in all creatures,
25
their essences are not identical with their actual existence, and it’s the latter that
26
is responsible for their individuation.
27
28
29 IV
30
31 Whatever we may think about Henry of Ghent’s reading of Avicenna, we still
32 seem far away from the theory of accidental individuation. For one thing is
33 clear: Henry does not consider Avicenna as someone for whom objects are
34
est, impossibile est eam multiplicari, et species huiusmodi est una numero”, quia scilicet
35
non multiplicatur in pluribus suppositis secundum numerum …’.
36 36 Henry of Ghent, Quodlibet II, q. 8, p. 39: ‘Quod autem non ex se sed solum ab alio
37 agente singulare est in supposito subsistens, quia ex se nulli appropriatur et est essentia
38 tantum, quantum est ex se, indifferenter natum est esse singulare, subsistendo in unico
39 supposito, vel universale, subsistendo in pluribus. Quod etiam bene dicit Avicenna et
determinat in Vo Metaphysicae suae. Ex quo sequitur apertissime quod necesse est ut non
40
sit essentia creaturae, in quantum creatura est, quin possit, quantum est ex se, in plura
41 individua multiplicari, quantumcumque sit abstracta a materia.’ For the ‘many gods’ see
42 ibid., p. 41.
352 Martin Pickav
1 individuated by accidents in the proper sense. But a look at another of Henry’s

2 quodlibetal question shows us that we are not too far away. A couple of years
3 later Henry comes back to the topic of individuation in question 8 of his fifth
4 Quodlibet (disputed either Advent 1280 or Lent 1281). This time the setup of
5 his response is more traditional; after developing what the principle of
6 individuation is supposed to explain, he begins by presenting accidental
7 individuation as one attempt to account for individuation. From the criticism of
8 accidental individuation he moves on to the Aristotelian view and its
9 deficiencies before he finally presents his own account. The way in which
10 Henry presents accidental individuation merits our attention because he adds
11 John Damascene and Avicenna to the traditional authorities Porphyry and
12 Boethius (see text I.2 in the appendix).
13 However, there is no indication that Henry changed his mind regarding
14 Avicenna. Since he still explains that Avicenna’s use of the term ‘accident’ in the
15 context of individuation should not be understood in the narrow sense, there is
16 no reason to believe that he forgot his earlier interpretation.37 It must thus be
17 with full awareness that Henry places Avicenna next to Porphyry and Boethius
18 in the setup of his response. I can only think of three possible reasons why
19 Henry suddenly associates Avicenna with the defenders of accidental indi-
20
viduation. (1) Maybe he simply wanted to create a certain rhetorical effect.
21
Referring not only to Porphyry and Boethius but also to John Damascene and
22
Avicenna lends accidental individuation more credence and makes it more
23
reasonable to discuss this view first. (2) Associating Avicenna with accidental
24
25
37 In Quodlibet V, q. 8, Henry describes his own candidate for the principle of
26 individuation as an accident. And he mentions the specific meaning the term ‘accident’
27 can have in this context. See ibid., fol. 166rM: ‘Et quia talis negatio sive determinatio,
28 licet nihil rei apponit in separatis supra essentiam qua determinetur, inquantum tamen
29 est extra intentionem formae, ut forma est, large sumendo accidens potest dici accidens,
ut secundum hoc huiusmodi individuatio dicatur fieri per aliquod accidens formae
30
superveniens. … Unde Avicenna assignans differentiam inter accidens quo individuantur
31 formae separatae et formae materiales generabiles et corruptibiles, dicit quinto
32 Metaphysicae: Postquam determinata est species specialissima accidunt ei concomitantia
33 ex accidentibus et proprietatibus per quae natura individuatur, et fit designata. Ipsae vero
34 proprietates et accidentia aut erunt adventitia tantum, ita ut non sunt de essentia ullo
modo, et haec sunt accidentia quae accidunt individuis rerum simplicium, aut erunt
35
dispositiones superadditae …’ The same is true for his treatment of individuation in
36 Quodlibet XI, q. 1, fol. 438vQ: ‘Essentia sive natura sive forma specifica ex se non est nisi
37 ipsa … cetera autem omnia sunt ei aliquo modo accidentalia, cuiusmodi sunt esse in
38 anima sive in intellectu ut in cognoscente et esse in supposito existente extra animam.
39 Sed etiam accidit ei quodammodo esse in ipso supposito, secundum quod hoc alias
saepius exposuimus secundum Avicennam … Quod quidem esse determinatum non est
40
vere accidens, licet sit extra rationem speciei. Determinatio necessario fit per aliquid. Si
41 enim illud esset vere accidens, tunc suppositum sub specie non esset unum nisi per
42 accidens et nullum per se quod falsum est.’
1 individuation creates almost the strongest possible contrast with the traditional
2 interpretation of Avicenna as a defender of material individuation. (3) Henry’s
3 response begins with a scrutiny of two positions he will ultimately criticize. But
4 he does not reject them entirely. This is particularly obvious from the way he
5 treats the so-called Aristotelian account. Instead of simply dismissing it, he
6 points out that it does not explain the ‘precise and proximate cause of
7 individuation’. Moreover, once the precise cause of individuation has been
8 established it will become clear to what extent the previous positions are in a
9 sense correct.38 For someone with this pedagogical strategy it makes sense to
10 start by running together positions such as Porphyry’s, Boethius’s and Avicenna’s
11 on the sole ground that they all speak about accidents and their role in
12 individuation. For as it will turn out they are all, in some measure, expressing
13 the truth, albeit in different respects.
14 Henry’s treatment of individuation in Quodlibet V, q. 8 is notorious. Its
15 notoriety is due to the fact that he in the end identifies the ‘precise and
16 proximate’ cause of individuation with a ‘double negation’. Each individual is
17 characterized by two negations, for an individual is essentially what is not
18 further divided in itself and not identical with something else. This leads Henry
19 to identify the principium individuationis with the negations themselves.39
20 Although the text is best understood in the context of Henry’s other discussions
21 of individuation, uncharitable readers have often focused on it separately.40 Yet
22 it should not come as a surprise that Henry’s critics found not only a seemingly
23 strange theory in Quodlibet V, q. 8, but also other materials, for example,
24 Henry’s presentation of accidental individuation. This, I believe, explains why
25 we can find after Henry’s fifth Quodlibet a whole series of texts depicting
26 Avicenna as a proponent of accidental individuation. There is no reason to
27 believe that, for instance, Duns Scotus’s listing in texts I.7 – 9 in the appendix is
28 original. Scotus knew both Henry’s Quodlibet as well as Quodlibet VII, q. 5 of
29 Godfrey of Fontaines (who himself engages critically and extensively with
30 Henry’s views). Since all the other authors listed in part I of the appendix engage
31 in their respective questions with key aspects of Henry of Ghent’s account of
32
33
34
35
36 38 Quodlibet V, q. 8, fol. 165vL: ‘Haec ergo est opinio sua de causa individuationis et
37 determinationis formae ad suppositum in formis materialibus plurificabilibus per
38 supposita quantum est ex parte formae, et est vera in talibus formis ut secundum cursum
39 naturae, sed non explicat praecisam et proximam causam individuationis … et per ipsam
verificantur opiniones praedictae inquantum talium formarum individuatio fit per
40
materiam ut est sub quantitate quae est accidentalis.’
41 39 Ibid., fol. 166rM.
42 40 On this point see my Henry of Ghent on Individuation.
354 Martin Pickav
1 individuation, it is more than plausible that they owe their depiction of

2 Avicenna’s view also to him.41
3 So at the end there is good news. Medieval Latin authors of the thirteenth
4 and fourteenth centuries did not systematically misunderstand Avicenna’s
5 account of individuation. In the wake of Henry of Ghent, Avicenna’s view got
6 somewhat caricatured by association with Porphyry and Boethius. But this was
7 due to the heat of the debate and not to a serious misreading. Moreover, almost
8 all the authors depicting Avicenna in the way described above would later in
9 their texts exempt him from the misgivings of the accidental individuation view.
10 Still it is somewhat ironic that the one medieval author who spilled more ink
11 than anyone else over Avicenna’s account of individuation should be responsible
12 for the caricature.
13
14
15
Appendix
16
I Avicenna as a Proponent of Accidental Individuation (Selected Passages)
17
18
I.1 Peter of Falco, Quaestiones disputatae, q. 5, pp. 198 f. [ca. 1279 – 81]:
19
‘Nam essentia ut essentia non dicit quid in actu nec quid ut unum nec ut
20
multa, sicut docet Avicenna, V Metaphysicae; nam humanitas ut humanitas
21
dicit essentiam vel quidditatem hominis, humanitas ut humanitas non est haec
22
humanitas, quae est in Sorte, tunc omnis humanitas esset haec humanitas,
23
immo quod sit haec humanitas et in Sorte, hoc habet per aliquid additum ultra
24
essentiam, scilicet per respectum ad suppositum et per accidentia, unde
25
individuatio fit per accidentia, ut arguit Avicenna, V Metaphysicae.’
26
27
I.2 Henry of Ghent, Quodlibet V, q. 8, fol. 165rI [Advent 1280 or Lent 1281]:
28
29
‘Nulla ergo restat difficultas in videndo quomodo in creaturis fiat per
30
individuationem suppositum absolutum, nisi in modo individuationis quem
31
Porphyrius posuit per accidentales proprietates, quae ut dicit numquam simul in
32
eodem contingit reperiri. Et similiter Boethius dicit in commento super
33
Porphyrium de individuis sub eadem specie: Quantum ad substantiam unum
34
sunt non habentia substantialem differentiam, sed accidentibus efficiuntur, ut
35
accidentibus saltem distare videantur. Plures sunt de quibus species praedicatur,
36
non substantiae diversitate, sed accidentium multitudine. Et in principio de
37
Trinitate: In numero differentiam accidentium varietas facit. Et Ioannes
38
Damascenus in libro de duplici natura et una hypostasi in Christo dicit circa
39
41 Only Peter of Falco’s text (I.9) shows no traces of Henry’s fifth Quodlibet. But his text
40
does also not list Avicenna next to John Damascene, Porphyry and Boethius. I still
41 believe Peter is influenced by Henry, but by Quodlibet II, q. 8 rather than by Quodlibet V,
42 q. 8.
1 principium: Omnis res qua differt species ab altera specie substantialis et

2 naturalis, constitutiva differentia dicitur; omnis autem res in qua differt
3 hypostasis ab eiusdem speciei consubstantiali hypostasi dicitur adventitia
4 differentia et qualitas et hypostatica et characteristica proprietas; hoc autem
5 est accidens; velut differt homo ab homine altero, quoniam hic quidem est
6 longus, hic autem brevis, hic quidem albus, hic vero niger. Et Avicenna in
7 quinto Metaphysicae suae dicit: postquam determinata est species specialissima,
8 erit natura, sic quod accidunt ei concomitantia ex accidentibus et proprietatibus
9 per quae natura individuatur et fit designatum.
10 Quod non potest stare, quoniam accidentia omnia huiusmodi sunt
11 posterioris et adventitiae naturae, sicut et isti confitentur. Ergo naturaliter
12 praesupponunt, etsi non tempore, substantiam ut eis subiectam. Subiecta autem
13 substantia non est nisi individua …’
14
15 I.3 Richard of Middleton, In II Sententiarum, dist. 3, a. 4, q. 1, p. 58 [after
16 1281]:
17 ‘Primo ostendo, quod Angelus sit substantia una numero formaliter per
18 aliquid superadditum essentiae suae.
19 Damascenus lib. 3, cap. 11 vult, quod indiuiduatio sit per accidens.
20 Item Avicenna 3 Meta., cap. 3 dicit, quod unitas substantiae non recipitur
21 in certificatione quidditatis alicuius substantiarum, sed est quiddam commu-
22 nicans substantiam, dicitur ergo de eis sicut accidens.
23 Item Porphyrius ponit rationem indiuiduationis per accidentia. …
24 Dixerunt autem quidam, quod substantia creata est una numero formaliter,
25 per aliquid superadditum essentiae suae. De quorum opinione videtur fuisse
26 Avicenna. Dicunt enim, quod denominatiua significant accidens, unum autem
27 est denominatiuum.
28 Praeterea, quia unitas est principium numeri et numerus est quantitas
29 discreta, quae est accidens, et ita videtur eis, quod omnis unitas creata sit
30 accidens.
31 Praeterea, quia aliter esset nugatio, dicendo substantia una, nisi unitas
32 significet aliquid superadditum substantiae, quae est una.
33 Sed hoc opinio non videtur rationabilis. Unde et commentator super 4
34 Metaph. reprehendit Avicenna de hoc et dicit sic:42 Avicenna autem multum
35 peccavit in hoc quod existimavit quod unum et ens significant dispositiones
36 additas essentiae rei et mirum est de isto homine quomodo erravit tali errore. Et
37 arguit sic contra ipsum: Si res esset una per aliquam rem additam suae naturae,
38 sicut credidit Avicenna tunc nihil esset unum per se et per suam substantiam,
39 sed per rem additam suae substantiae, et illa res, quae est una si dicitur, quod est
40
41 42 See Averroes, Long Commentary on Aristotle’s Metaphysics, lib. IV, com. 3, in Aristotle,
42 Opera cum Averrois commentariis, vol. 8, fol. 67rB–vG.
356 Martin Pickav
1 una per intentionem additam suae essentiae, quaeretur etiam de illa re per quam
2 sit una. Si igitur sit una per intentionem additam illi iterum quaeretur, et sic
3 procedit in infinitum. …
4 Ad primum in oppositum dicendum, quod illa auctoritas Damasceni
5 intelligenda est quantum ad manifestationem indiuiduationis, manifestatur
6 enim Petrus esse aliud indiuiduum a Ioanne per diuersa accidentia, quae in eis
7 sunt.
8 Ad secundum dicendum, quod auctoritas Avicennae neganda est in
9 proposito. Ad tertium dicendum est sicut ad primum.’
10
11 I.4 Roger Marston, Quodlibet II, q. 30, p. 294 [1282 – 4]:
12 ‘Sapientes mundi plures, ut Boethius, Porphyrius et Avicenna, videntur velle
13 quod individua solo accidente differant, quibusdam proprietatibus accidenta-
14 libus in uno repertis quae non sunt in alio possibiles inveniri, ut sunt patria,
15 parentela, locus, tempus, et propria nominatio. Sed cum accidentia habeant esse
16 a subiecto, magis accipiunt individuationem ab ipso quam subiectum ab eis.’43
17
18 I.5 Godfrey of Fontaines, Quodlibet VII, q. 5, pp. 319 f. [1290/91 or 1291/92]:
19 ‘Videtur ergo quod individuatio fiat per accidentia, et hoc videntur dicere
20 philosophi et sancti doctores. Nam Porphyrius dicit quod individua differunt
21
per accidentales proprietates quas nunquam contingit simul in pluribus reperiri.
22
Et Boetius in commento dicit quod individua sunt eadem specie; quantum ad
23
substantiam unum sunt non habentia substantialem differentiam, sed acciden-
24
tibus videtur effici ut in accidentibus saltim differre videantur plures substantiae
25
de quibus species praedicatur et non substantiae diversitate, sed accidentium
26
multitudine; et in principio libri de Trinitate: in numero differentiam
27
accidentium varietas facit. Et Damascenus de Duabus naturis et una persona
28
Christi, circa principium: omnis res qua differt species ab altera specie
29
substantialis et naturalis et constitutiva differentia dicitur; omnis autem res in
30
qua differt hypostasis ab eiusdem speciei cum substantiali hypostasi dicitur
31
adventitia differentia et qualitas et hypostatica et caracteristica proprietas; hoc
32
autem est accidens; vel ut differt homo ab homine altero, quoniam hic quidem
33
est longus, hic est autem brevis, hic autem albus, hic vero niger. Et Avicenna,
34
quinto Metaphysicae: postquam determinata fuerit natura facta species special-
35
issima, erit natura determinata sic quod accidunt ei concomitantia ex
36
accidentibus et proprietatibus per quae natura individuatur et sic designata
37
38
videtur.’
39
43 For the examples of individualizing accidents see also Auctoritates Aristotelis, auctoritates
40
Porphyrii, ed. Hamesse, p. 300: ‘Et sunt septem proprietates, ut dicit Boethius, scilicet
41 forma, figura, locus, stirps, nomen, patria, tempus. Haec septem propria continet omnis
42 homo.’ However, such a list of seven properties cannot be found in Boethius’s works.
1 I.6 Peter of Auvergne, Quodlibet II, q. 5, edited in Hocedez, Une Question

2 indite de Pierre d’Auvergne, p. 371 [1297]:
3 ‘De hoc autem fuit antiqua opinio: quod individuationis principium est
4 aggregatio proprietatum accidentalium; quarum collectionem impossibile est in
5 alio reperire secundum Porphirium. Unde Avicenna dicit V Metaphysicae capite
6 3: nature speciei specialissime accidunt concomitantia ex proprietatibus et
7 accidentibus per que individuatur et sic designata est.’
8
9 I.7 John Duns Scotus, Quaestiones super libros Metaphysicorum Aristotelis, lib.
10 VII, q. 13, pp. 219 f.:
11 ‘Circa istam quaestionem quidam dicunt naturam individuari per aliquod
12 positivum aliquo modo aliud a natura; quidam non.
13 Prima pars habet quinque vias. Prima est de multis accidentibus aggregatis.
14 Porphyrius: individuum constat ex septem proprietatibus. Et Boethius ibi in
15
commento; idem De Trinitate. Et Damascenus; et Avicenna V Metaphysicae 2.’
16
17
I.8 John Duns Scotus, Lectura II, dist. 3, p. 1, q. 4, nn. 61 – 4, pp. 246 f.:
18
‘Quaeritur an quantitas sit illud positivum quo substantia materialis est haec
19
et singularis et indivisibilis in multas partes subiectivas.
20
Quod sic, videtur:
21
Per Boethium De Trinitate, in principio: “Differentiam in numero
22
accidentia faciunt, quae si omnia auferas, saltem locum auferre non potes;
23
quare diversa numero sunt in diversis locis”; sed quod sint in locis diversis, non
24
25
est nisi per quantitatem; ergo etc.
26
Praeterea, Damascenus in Elementario suo, cap. 5, hoc etiam dicit, quod
27
“omne per quod aliqua different hypostatice, accidens est”.
28
Praeterea, Avicenna V Metaphysicae cap. 2: “Naturae quae eget materia ad
29 esse adveniunt accidentia, per quae individuatur”.’
30
31 I.9 John Duns Scotus, Ordinatio II, dist. 3, p. 1, q. 4, n. 70, pp. 421 f.:
32 ‘Quarto quaeritur utrum substantia materialis per quantitatem sit individua
33 vel singularis.
34 Quod sic:
35 Boethius De Trinitate: “In numero differentiam, accidentium varietas facit,
36 nam tres homines neque specie neque genere sed suis accidentibus differunt;
37 nam si vel animo cuncta accidentia separamus, locum tamen cunctis diversus
38 est, quem unum fingere duobus nullo modo possumus: duo enim corpora
39 unum locum non obtinebunt, qui est accidens, et ideo sunt numero plures
40 quanto accidentia plura sunt”. Et inter omnia accidentia, primum accidens est
41 quantitas, quam etiam specialiter videtur exprimere in “loco” (dicendo quod
42 “eundem locum fingere non possumus”), qui competit eis in quantum quanta.
358 Martin Pickav
1 Praeterea, Damascenus in Elementario cap. 5 (non computando prooemi-

2 um): “Omnis res in qua differt hypostasis ab eiusdem speciei hypostasi, dicitur
3 adventicia differentia, et characteristica proprietas et qualitas hypostatica; hoc
4 autem accidens est, – velut ‘differt homo ab altero homine, quoniam hic quidem
5 longus, ille autem brevis’”.
6 Praeterea, Avicenna in V Metaphysicae 2: “Natura quae materia eget, – ad
7 eius esse adveniunt accidentia et dispositiones, extrinsece, per quae individu-
8 atur”.’
9
10 I.10 John of Paris, Quaestio de causa individuationis, edited in Mller, Eine
11 Qustion ber das Individuationsprinzip, p. 343 [1304 – 5?]:
12 ‘Quidam enim dicunt accidentia esse causam individuantionis, per quae
13 species communis ad determinatum contrahitur individuum. Et huic opinioni
14 alludit plurium auctoritas. Porphyrius enim dicit quod individuatio fit per
15 accidentales proprietates, quarum collectio eadem numero in alio reperiri non
16 potest. Boethius etiam in Commento dicit, ibidem, quod individua sunt sub
17 una specie unum quantum ad substantiam, non habentia substantiae differ-
18 entiam, sed accidentibus efficiuntur, ut accidentibus saltem distare videtur.
19 Plures enim sunt, de quibus species praedicantur, non substantiae diversitate,
20 sed accidentium diversitate vel multitudine.
21 Item, Boethius in principio De Trinitate …
22 Item, Ioannes Damascenus in libro De duplici natura et una hypostasi in
23 Christo, dicit circa principium: “Omnis res …”
24 Item, Avicenna in V libro Metaphysicae suae dicit: “postquam natura
25 determinata est species specialissima erit natura sic quod accidunt ei
26 concomitantia ex accidentibus et proprietatibus, per quae natura individuatur
27 et fit designata.”’
28
29
30 II Accidental Individuation Without Avicenna (Selected Passages)
31
32 II.1 William of Auvergne, De universo, II, c. 9, pp. 852bf.:
33 ‘Et de prima quidem quaestionum praenominatarum audivisti sermonem
34 Aristotelis, et sermones Alpharabii, et Avicennae, et aliorum, qui in parte ista
35 Aristoteli consenserunt. Qui omnes in hanc unam sententiam concordaverunt,
36 quod omnis diversitas, omnis numerus a materia est corporali.’
37
38 II.2 John of La Rochelle, Summa de anima, II, c. 113, pp. 272 f.:
39 ‘Notandum tamen est secundum Avicennam, quod formis materialibus
40 propter materiam accidunt disposiciones multe quae non habent ex sua essencia,
41 scilicet ex hoc quod sunt forme. … Relinquitur igitur quod accidit humanitati
42 ex materia iste modus multiplicationis et diuisionis que est singularibus.
1 Accidunt ei eciam alia preter hoc, quia cum fuerit in materia, acquiritur ei
2 aliquis modus quantitatis, qualitatis, situs et ubi, quae omnia abstracta sunt a
3 natura ipsius. Si enim humanitas, ex hoc quod humanitas determinaret sibi
4 accidencia huiusmodi, tunc oporteret unumquemque hominem conuenire cum
5 altero et in huiusmodi accidentibus, cum conueniant in ipsa humanitate in
6 quantum huiusmodi: essent igitur eedem qualitates et quantitates, etc.
7 Relinquitur ergo quod forma humana non habet ex sua essencia huiusmodi
8 accidencia; sed ista ei accidunt propter comparacionem ipsius ad naturam.’
9
10 II.3 Peter of Tarentaise (Pope Innocent V), In II Sententiarum, dist. 17, q. 1,
11 a. 2, p. 142a:
12 ‘Hoc aliquid dicitur individuum per se subsistens, hoc autem dupliciter: aut
13 habet principium individuationis intra essentiam propriam, et omne tale
14 compositum est ex materia et forma, quia principium individuationis est
15 materia, aut habens extra se, sic non omne, quod est individuum per se
16 subsistere valens, est compositum ex materia et forma, et tale est anima
17 rationalis quae secundum Avicennam individuatur per corpus.’
18
19 II.4 Bonaventure, In II Sententiarum, dist. 3, p. 1, a. 2, q. 2, p. 105a:
20 ‘Secundo quaeritur, utrum personalis proprietas sit in Angelis accidentalis,
21 vel substantialis. Et quod accidentalis, videtur sic.
22 1. Boethius dicit, quod “omne principium manat de genere accidentium,
23 non solum proprium individui, sed etiam speciei”; sed individuum magis
24 approximat accidentibus quam species: ergo multo fortius proprietas individ-
25 ualis de genere accidentium est. Sed discretio personalis est proprietas
26 individualis: ergo etc.
27 2. Item, Richardus de sancto Victore ait, “quod in divinis est personalis
28 discretio per originem, in Angelis per qualitatem, in hominibus utroque
29 modo.” …
30 3. Item, Porphyrius dicit, “quod individuum constat ex proprietatibus,
31 quarum collectionem impossibile est in altero reperire”; sed tales sunt
32 proprietates accidentales: ergo individuatio est per accidentia. Sed per eadem
33 est personalis discretio, per quae est individuatio: ergo etc.’
34
35 II.5 Roger Bacon, Questiones supra libros prime philosophie Aristotelis, pp. 226 f.:
36 ‘Queritur … de causa sue individuationis scilicet, et quid facit hujusmodi
37 individuum esse individuum. Et videtur quod non accidentia: quia nullum
38 accidens est perfectio substantie … Contrarium a multis dicitur, quia
39 Porphyrius, “individua solo numero differunt”; set numerus est accidens, ergo
40 cum accidens ex accidentibus fiat, ergo unitas numeralis fit ex accidentibus,
41 quare et individuatio. Item, Boethius in libro De Trinitate, “tres homines nec
42 genere nec specie discrepant set solo accidente”, quare accidens facit
360 Martin Pickav
1 individuationem. Istud idem dicit in Commento supra Porphyrium, ipse

2 Boethius loquitur ibi de illis 7 proprietatibus. Item, in Metaphysica Argazelis, in
3 causa prima, quia ibi non est compositio differentie vel accidentis, ideo non est
4 in eo dualitas ullo modo, quare omnis compositio, vel compositio differentie
5 cum alio, vel accidentis; ergo individuum cum sit compositum, aut vere erit per
6 additionem differentie, quod non est verum quia differentia antecedit speciem,
7 individuum sequitur, fiet ergo per compositionem accidentium, ergo etc.’44
8
9 II.6 Robert Kilwardby, Quaestiones in librum secundum Sententiarum, q. 17,
10 p. 63:
11 ‘Quod sit forma accidentalis, videtur per Porphyrium. Qui dicit quod
12 individuum dicitur, quoniam “consistit ex proprietatibus quarum collectio
13 numquam in alio eadem erit.” Item Boethius De Trinitate cap. 2 …’
14
15 II.7 William de la Mare, Scriptum in secundum librum Sententiarum, dist. 3,
16 q. 6, p. 62:
17 ‘Alii dicunt quod accidentia sunt causa individuationis et hoc videtur
18 sensisse Porphyrius qui dicit quod individuum constat ex proprietatibus quarum
19 collectio numquam potest in aliquo reperiri. Hoc idem videtur per Boethium,
20 De trinitate: “In numero”, inquit, “differentiam accidentium varietas facit.”’
21
22 II.8 Giles of Rome, In II Sententiarum, dist. 3, pars 1, q. 2, a. 1, p. 184:
23 ‘Tertiam difficultatem faciunt dicta Doctorum. Nam Boethius in libro de
24 Trinitate circa principium ait, quod in numero differentiam accidentium
25 varietas facit. Et subdit, quod tres homines, neque genere, neque specie, sed suis
26 accidentibus differunt. Et Porphyrius ait, quod individua dicimus huiusmodi,
27 quoniam ex proprietatibus consistit unumquodque eorum, quarum collectio
28 nunquam in alio eadem erit. Et Simplicius in libro super praedicamentis ait:
29 Numero autem differunt, quaecumque cum cursu accidentium proprietatem
30 suae substantiae determinaverunt …’
31
32 II.9 Peter John Olivi, Quaestiones in secundum librum Sententiarum, q. 12,
33 p. 212:
34 ‘Quidam enim voluerunt quod individuatio dicat collectionem accidenti-
35 um, de quorum numero Porphyrius et Boethius fuisse videntur.’
36
37
38
39
40
41
42 44 See al-Ġaza-lı̄, Metaphysica, p. 26.
1 Bibliography
2
3 J.A. Aertsen, Die Thesen zur Individuation in der Verurteilung von 1277, Heinrich von
4 Gent und Thomas von Aquin, in Individuum und Individualitt im Mittelalter
(Miscellanea Mediaevalia 24), eds J.A. Aertsen and A. Speer, Berlin/New York:
5
Walter de Gruyter, 1996, pp. 249 – 65.
6 al-Ġaza-lı̄, Metaphysica, ed. J.T. Muckle, Toronto: St. Michael’s College, 1933.
7 Aristotle, Opera cum Averrois commentariis, vol. 8, Venice, 1562, reprinted Frankfurt
8 a. M.: Minerva, 1962.
9 Avicenna, Liber de philosophia prima sive Scientia divina, ed. S. van Riet, 3 vols, Louvain:
Peeters, 1977 – 83.
10
A. Bck, The Islamic Background: Avicenna (b. 980; d. 1037) and Averroes (b. 1126;
11 d. 1198), in Individuation in Scholasticism: The Later Middle Ages and the Counter-
12 Reformation, 1150 – 1650, ed. J.J.E. Gracia, Albany, New York: SUNY Press, 1994,
13 pp. 39 – 67.
14 D. Black, Individuation, Self-Awareness, and God’s Knowledge of Particulars: The
Intersection of Three Avicennian Problems (forthcoming).
15
A.M.S. Boethius, De consolatione philosophiae et opuscula sacra, ed. C. Moreschini,
16 Munich/Leipzig: K.G. Saur, 2005.
17 –– , In Isagogen Porphyrii commenta (Corpus Scriptorum Ecclesiasticorum Latinorum
18 48), ed. S. Brandt, Vienna: Tempsky, 1906.
19 Bonaventure, Opera omnia, vol. 2: Commentarius in II Sententiarum, eds PP Collegii
S. Bonaventurae, Quarachi: Typographia Collegii S. Bonaventurae, 1885.
20
S.F. Brown, Henry of Ghent, in Individuation in Scholasticism: The Later Middle Ages
21 and the Counter-Reformation, 1150 – 1650, ed. J.J.E. Gracia, Albany, New York:
22 SUNY Press, 1994, pp. 195 – 219.
23 R. Chiaradonna, La teoria dell’individuo in Porfirio e l’IDIYS POIOM stoico, Elenchos
24 21, 2000, pp. 303 – 31.
T.-A. Druart, The Human Soul’s Individuation and Its Survival After the Body’s Death:
25
Avicenna on the Causal Relation Between Body and Soul, Arabic Sciences and
26 Philosophy 10, 2000, pp. 259 – 73.
27 C. Erismann, Collectio proprietatum. Anselme de Canterbury et le problme de
28 l’individuation, Mediaevalia. Textos e estudos 22, 2003, pp. 55 – 71.
29 –– , L’individualit explique par les accidents: Remarques sur la destine ‘chrtienne’ de
Porphyre, in Complments de substance: tudes sur les proprits accidentelles offertes
30
Alain de Libera, eds C. Erismann and A. Schniewind, Paris: Vrin, 2008, pp. 51 – 66.
31 Giles of Rome, In secundum librum Sententiarum, Venice, 1581, reprinted Frankfurt
32 a.M.: Minerva, 1968.
33 P. Glorieux, Les premires polmiques thomistes: 1. Le Correctorium corruptorii ‘Quare’,
34 Kain: Le Saulchoir, 1927.
Godfrey of Fontaines, Les Quodlibet cinq, six et sept (Les Philosophes Belges, vol. 3), eds
35
M. De Wulf and J. Hoffmans, Louvain: Institut suprieur de philosophie, 1914.
36 J.J.E. Gracia, Introduction to the Problem of Individuation in the Early Middle Ages,
37 Munich/Vienna: Philosophia Verlag, 1984.
38 J. Hamesse, ed., Les Auctoritates Aristotelis, Louvain: Publications universitaires, 1974.
39 D.N. Hasse, Avicenna’s ‘De anima’ in the Latin West: The Formation of a Peripatetic
40
41
42
362 Martin Pickav
1 Henry of Ghent, Quodlibeta, 2 vols, ed. J. Badius Ascensius, Paris, 1518; reprinted
2 Leuven: Bibliothque S. J., 1961.
3
–– , Opera omnia, vol. 6: Quodlibet II, ed. R. Wielockx, Leuven: Leuven University
Press, 1983.
4
E. Hocedez, Une Question indite de Pierre d’Auvergne sur l’individuation, Revue
5 noscolastique de philosophie 39, 1934, pp. 355 – 86.
6 John Damascene, Dialectica, version of Robert Grosseteste, ed. O.A. Colligan,
7 St. Bonaventure, New York: Franciscan Institute, 1953.
8 –– , De institutione elementari, version of Robert Grosseteste, ed. H. Gravius, Cologne,
9
1546.
John Duns Scotus, Opera omnia, vol. 7: Ordinatio II, dist. 1 – 3, ed. Commissio
10
Scotistica, Vatican City: Typis Polyglottis Vaticanis, 1973.
11 –– , Opera omnia, vol. 18: Lectura II, dist. 1 – 6, ed. Commissio Scotistica, Vatican City:
12 Typis Polyglottis Vaticanis, 1982.
13 –– , Opera philosophica, vols 3 – 4: Quaestiones super libros Metaphysicorum Aristotelis, eds
14 G. Etzkorn et al., St. Bonaventure, New York: Franciscan Institute, 1997.
15
John of La Rochelle, Summa de anima, ed. J.G. Bougerol, Paris: Vrin, 1995.
P. King, The Problem of Individuation in the Middle Ages, Theoria 66, 2000, pp. 159 –
16
84.
17 –– , Metaphysics, in The Cambridge Companion to Abelard, eds J.E. Brower and
18 K. Guilfoy, Cambridge: Cambridge University Press, 2004, pp. 65 – 125.
19 A. de Libera, L’art des gnralits. Thories de l’abstraction, Paris: Aubier, 1999.
20 –– , D’Avicenne Averros, et retour. Sur les sources arabes de la thorie scolastique de
21
l’un transcendantal, Arabic Sciences and Philosophy 4, 1994, pp. 141 – 79.
M.E. Marmura, Avicenna’s Chapter on Universals in the Isagoge of his Shifa’, in Islam:
22
Past Influence and Present Challenge, eds A.T. Welch and P. Cachia, Edinburgh:
23 Edinburgh University Press, 1979, pp. 34 – 56.
24 J.P. Mller, Eine Qustion ber das Individuationsprinzip des Johannes von Paris O.P.
25 (Quidort), in Virtus politica. Festgabe zum 70. Geburtstag von Alfons Hufnagel, eds
26 J. Mçller and H. Kohlenberger, Stuttgart-Bad Cannstatt: Fromann, 1974, pp. 335 –
27
56.
Peter of Falco, Quaestiones disputatae, 3 vols, ed. A.-J. Gondras, Louvain: ditions
28 Nauwelaerts, 1968.
29 Peter John Olivi, Quaestiones in secundum librum Sententiarum, 3 vols, ed. B. Jansen,
30 Quaracchi: Typographia Collegii S. Bonaventurae, 1922 – 6.
31 –– , Quodlibeta quinque, ed. S. Defraia, Grottaferrata: Editiones Collegii S.
32 Bonaventuae, 2002.
33
Peter of Tarentaise (Pope Innocent V), In IV libros Sententiarum, Toulouse, 1652;
reprinted Ridgewood, New Jersey: Gregg Press, 1964.
34 M. Pickav, The Controversy of the Principle of Individuation in Quodlibeta
35 (1277–ca. 1320): A Forest Map, in Theological Quodlibeta in the Middle Ages:
36 The Fourteenth Century, ed. C. Schabel, Leiden: Brill, 2007, pp. 17 – 79.
37 –– , Henry of Ghent on Individuation, Proceedings of the Society for Medieval Logic and
38 Metaphysics 5, 2005, pp. 38 – 49 [http://www.fordham.edu/gsas/phil/klima/SMLM/
PSMLM5/PSMLM5.pdf ].
39
Porphyry, Isagoge, translatio Boethii, ed. B.G. Dod, in Aristoteles Latinus, vol. I.6 – 7,
40 Bruges/Paris: Descle de Brower, 1966.
41 –– , Isagoge, ed. A. Busse, in Commentaria in Aristotelem Graeca, vol. IV.1, Berlin:
42 G. Reimer, 1887.
1 Richard of Middleton, In IV libros Sententiarum, 4 vols, Brescia, 1591, reprinted

2 Frankfurt a.M.: Minerva, 1963.
3
Robert Kilwardby, Quaestiones in librum secundum Sententiarum, ed. G. Leibold,
Munich: Bayerische Akademie der Wissenschaften, 1992.
4 Roger Bacon, Opera hactenus inedita, vol. 10: Questiones supra libros prime philosophie
5 Aristotelis, eds R.W. Steele and F.M. Delorme, Oxford: Clarendon Press, 1930.
6 Roger Marston, Quodlibeta quatuor, eds G. Etzkorn and I. Brady, Quaracchi:
7 Typographia Collegii S. Bonaventurae, 1968.
8
M.-D. Roland-Gosselin, Le ‘De ente et essentia’ de S. Thomas d’Aquin, Kain: Le Saulchoir,
1926.
9 M. Sebti, Avicenne: l’me humaine, Paris: Presses universitaires de France, 2000.
10 R. Sorabji, Self: Ancient and Modern Insights about Individuality, Life, and Death,
11 Chicago: Chicago University Press, 2006.
12 T. Suarez-Nani, Les anges et la philosophie, Paris: Vrin, 2002.
13
Thomas Aquinas, Super Boetium de trinitate, in id., Opera omnia, vol. 50, ed. Leonina,
Rome/Paris, 1992.
14 –– , De ente et essentia, in id., Opera omnia, vol. 43, ed. Leonina, Rome, 1976.
15 Thomas Sutton, Quaestiones ordinariae, ed. J. Schneider, Munich: Bayerische Akademie
16 der Wissenschaften, 1977.
17 William of Auvergne, De universo, in id., Opera omnia, vol. 1, Paris, 1674, reprinted
18
Frankfurt a.M.: Minerva, 1963.
William de la Mare, Scriptum in secundum librum Sententiarum, ed. H. Kraml, Munich:
19 Bayerische Akademie der Wissenschaften, 1995.
20 J.F. Wippel, The Metaphysical Thought of Thomas Aquinas, Washington, DC: Catholic
21 University of America Press, 2000.
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
1
2
3
4 Scotus and Avicenna on What it is to Be a Thing
5
6 Giorgio Pini
7
8 Scotus’s debt to Avicenna’s Metaphysics has been subjected to careful scrutiny in
9 the last few years. After a landmark article by tienne Gilson of some eighty
10 years ago, it became for some time customary to assume that Scotus had been
11 strongly influenced by Avicenna in a number of key issues.1 This claim is now
12 increasingly being regarded with some caution. This is mostly due to historical
13 and textual considerations. It is now recognized that Scotus’s discussion and
14 interpretation of Avicenna should be considered in the larger context of Scotus’s
15 complex relationship with the later medieval theologian, Henry of Ghent.
16 Scotus, it appears, focused on those passages by Avicenna that Henry of Ghent
17 had already taken into account before him. Admittedly, this does not entail that
18 Scotus had only second-hand knowledge of Avicenna’s text.2 But it is difficult to
19 resist the conclusion that Scotus discussed and quoted Avicenna because he
20 discussed and quoted Henry of Ghent. As to the supposedly exceptional
21 influence that Avicenna’s thought had on Scotus, a simple count of Scotus’s
22 explicit quotations from Avicenna’s Metaphysics suggests some caution. Scotus’s
23 references turn out to be more or less as many as those present in Thomas
24 Aquinas.3 Even though this rough comparison should not be taken as absolutely
25
26 1 Gilson, Avicenne et le point, p. 147: ‘Il est d’abord difficile de ne pas Þtre frapp de
27 l’influence exerce par Avicenne sur la formation du scotisme …’ In that article, Gilson
28 stressed Avicenna’s influence on Scotus with regard to the doctrine of the subject matter
29 of metaphysics, the object of the intellect, and the abstraction mechanism. Some forty
years later, Gilson repeated basically the same assessment and stressed Avicenna’s influ-
30
ence on Scotus’s teachings on the subject matter of metaphysics, the common nature, and
31 the univocity of the concept of being. See Gilson, Avicenne en Occident, pp. 109 – 12.
32 Specific aspects of the complex relationship between Avicenna and Scotus have recently
33 been considered by Sondag, La rception; Porro, Duns Scot; and Druart, Ibn Sina
34 (Avicenna).
2 Sondag, La rception, p. 591. On the influence of Avicenna’s metaphysics on Henry of
35
Ghent’s thought, see Teske, Henry of Ghent’s Debt.
36 3 I base this remark on a comparison between the first three books of Scotus’s Ordinatio
37 and of Thomas Aquinas’s commentary on the Sentences (I have left out the fourth book
38 because only the first thirteen distinctions of the fourth book of Scotus’s Ordinatio have
39 been critically edited at this date). In the first book of Aquinas’s commentary on the
Sentences, there are some 50 explicit references to Avicenna’s Metaphysics against the 67
40
references present in the first book of Scotus’s Ordinatio. Things are reversed in the
41 second book, where to 42 references to Avicenna’s Metaphysics in Aquinas’s commentary
42 there correspond 21 references in Scotus’s Ordinatio. As to the third book, there are 9
366 Giorgio Pini
1 conclusive evidence, it is a strong indication that Scotus’s debt towards Avicenna

2 is considerable but not exceptional among late thirteenth-century authors. It is
3 interesting to notice that things seemed to have changed after Scotus, at least as
4 far as English authors are concerned. Both William of Ockham and Adam
5 Wodeham virtually stopped referring to Avicenna’s Metaphysics, as attested by
6 the indexes of quotations in the critical editions of their works.
7 Setting aside these general concerns about Avicenna’s influence on Scotus, I
8 will consider a specific case, namely Scotus’s interpretation of a famous passage
9 taken from the fifth chapter of the first book of Avicenna’s Metaphysics. In that
10 chapter, Avicenna makes some famous claims about thing and being as the first
11 notions received in our minds.4
12 This paper has three parts. First, I consider Thomas Aquinas’s and Henry of
13 Ghent’s views about how to interpret Avicenna’s claims about what it is to be a
14 thing. Second, I turn to Scotus’s view, which may be seen as a development and
15 criticism of Henry of Ghent’s position. Third, I conclude with some remarks on
16 the role our concept of thing (or being or something) plays in cognition.
17 Specifically, I argue that Scotus’s distinction between our situation before and
18 after the fall – namely, the distinction between, on the one hand, what we were
19 originally supposed to cognize and, on the other hand, our current limitations –
20 is particularly relevant for his interpretation of the claim that being and thing are
21 the objects of the intellect, and more in general for his interpretation of
22 Avicenna’s Metaphysics. My conclusion is that Avicenna’s influence on Scotus is
23 undeniable (just as is the case with Thomas Aquinas and Henry of Ghent).
24 More importantly, however, even with regard to such a fundamental doctrine as
25 the primacy of the concepts of thing and being over other concepts, it makes
26 little sense to speak of Avicenna’s influence without specifying in which way
27 Avicenna was interpreted. In turn, the way Avicenna was interpreted had often
28 much to do with concerns that were extraneous to Avicenna’s original viewpoint.
29 Behind the interpretation of Avicenna, the main interest of these authors was to
30 understand how things really are in the world. Avicenna offered some important
31 conceptual tools to find an answer to this most fundamental question, but the
32 way Avicenna was interpreted in turn depended on the basic metaphysical
33 assumptions that each of these authors made.
34
35
36 references in Aquinas against 4 references in Scotus. I have taken the information about
37 Aquinas from Vansteenkiste, Avicenna-Citaten, pp. 457 – 507. The data about Scotus are
38 taken from the indexes to volumes 1 – 10 of the Vatican edition of Scotus’s Opera omnia.
39 On Avicenna’s influence on Aquinas, see Anawati, Saint Thomas d’Aquin, pp. 449 – 65;
Wippel, The Latin Avicenna, pp. 51 – 90.
40
4 In this paper, I follow the practice of putting a term between inverted commas (e. g.,
41 ‘being’) when referring to the term and of italicizing it (e. g., being) when referring to the
42 corresponding concept.
Scotus and Avicenna on What it is to Be a Thing 367
1 I
2
3 Avicenna’s claim in chapter I.5 that thing and being, as well as necessary, are the
4 first notions impressed in our minds is one of the most often quoted passages
5 from his Metaphysics. 5 After making that claim, Avicenna goes on to stress the
6 difficulty of providing a clarification of such general notions in terms of other
7 notions. It seems that every attempt to clarify what it is to be a thing and what it
8 is to be a being must be done by referring to concepts that already presuppose
9 the concepts of thing and being. Consequently, any description of thing and
10 being turns out to be circular. However, this does not end Avicenna’s discussion.
11 Specifically with regard to the concept of thing, Avicenna does actually suggests
12 two possible descriptions of what it is to be a thing. The first description is
13 extremely generic. A thing is described as something about which a true
14 statement can be made. Avicenna remarks that this description does not escape
15 the charge of circularity, because it uses notions that can only be explained by
16 presupposing what a thing is, and it basically amounts to saying that a thing is a
17 thing (again) about which a true statement is made. Avicenna, however, admits
18 that even such a circular description may be useful, as it directs one’s attention to
19 what we are trying to identify, namely things. In other terms, even though such
20 a description fails to explain what a thing is, it can be used to discriminate if
21 something is a thing or not. As long as I can make a true statement about
22 something, that something is a thing:
23
The case is similar with somebody’s statement: ‘The thing is that about which it is
24 valid [to give] an informative statement,’ for ‘is valid’ is less known than ‘the thing’;
25 and ‘informative statement’ is [likewise] less known than ‘the thing’. How, then, can
26 this be the definition of the thing? Indeed, ‘is valid’ and ‘information’ are known
27 only after one uses, in explaining what they are, [terms] indicating that each is either
a ‘thing,’ a ‘matter,’ a ‘whatever,’ or a ‘that which [is]’ – all of these being like
28
synonyms of the world ‘thing.’ How, then, can the thing be truly defined in terms of
29 what is known only through it? Yes, in this and in similar things there may be some
30 act of directing attention: but, in reality, if you say, ‘The thing is that about which it
31 is valid [to give] an informative statement,’ it is as if you have said, ‘The thing is the
32 thing about which it is valid [to give] an informative statement,’ because the
meaning of ‘whatever,’ ‘that which,’ and ‘the thing’ is one and the same. You would
33
have then included ‘the thing’ in the definition of ‘the thing.’ Still, we do not deny
34 that through this [statement] and its like, despite its vitiating starting point, there
35 occurs in some manner a directing of attention to the thing.6
36
37
38 5 Avicenna, The Metaphysics of ‘The Healing’, I, 5. See Aertsen, Avicenna’s Doctrine, p. 22.
39 Later in the same chapter, Avicenna lists thing, being, and one (rather than necessary) as
the notions that are common to everything. On the difference between the two lists, see
40
Aertsen, Avicenna’s Doctrine, pp. 25 – 6.
41 6 Avicenna, The Metaphysics of ‘The Healing’, I, 5, pp. 23 – 4. See the Latin translation in
42 Avicenna Latinus, Liber de philosophia prima, vol. 1, pp. 33 – 4: ‘Similiter est etiam hoc
368 Giorgio Pini
1 Second and more specifically, a thing can be described as something having a

2 certain reality (certitudo, in the Latin translation), and this is the proper being of
3 that thing:
4
‘The thing,’ or its equivalent, may be used in all languages to indicate some other
5 meaning. For, to everything there is a reality by virtue of which it is what it is. Thus,
6 the triangle has a reality in that it is a triangle, and whiteness has reality in that it is
7 whiteness. It is that which we should perhaps call ‘proper existence’ …7.
8
Avicenna’s readers in the Latin West made a real effort to clarify these two
9
descriptions. What exactly is Avicenna doing? Is he distinguishing between two
10
definitions of the term ‘thing’? Or is he rather distinguishing between two kinds
11
of things? Moreover, what is the relationship between these two descriptions? Is
12
the former just a first and generic attempt to clarify what a thing is, and in that
13
case, is it superseded by the second description? Or should these two
14
descriptions be considered as two equally valuable ways to pick out different
15
kinds of things? Finally, what is the ontological commitment of each of these
16
descriptions?8
17
18
By wondering about Avicenna’s intentions in this dense passage, Latin
19
medieval authors were trying to come to terms with an important if elusive
20
metaphysical point. For no matter how one interprets Avicenna, there seems to
21
be much merit in the attempt to distinguish between two descriptions or
22 quod dicitur quod res est id de quo potest aliquid vere enuntiari; certe potest aliquid
23 minus notum est quam res, et vere enuntiari minus notum est quam res. Igitur quomodo
24 potest hoc esse declaratio? Non enim potest cognosci quid sit potest aliquid vel vere
enuntiari, nisi in agendo de unoquoque eorum dicatur quod est res vel aliquid vel quid
25
vel illud; et haec omnia multivoca sunt nomina rei. Quomodo ergo vere potest sciri res
26 per aliquid quod non potest sciri per eam? Sed fortasse hoc et consimile erit innuitio
27 aliqua. Nam cum dicis quod res est id de quo vere potest aliquid enuntiari, idem est
28 quasi dicens quod res est res de qua vere potest aliquid enuntiari; nam id et illud et res
29 eiusdem sensus sunt. Iam igitur posuisti rem in definitione rei, quamvis nos non
negamus quod haec et consimilia, cum sint vitiosa, tamen aliqua designatio rei sunt’. See,
30
on this chapter, Marmura, Avicenna on Primary Concepts. On the background to
31 Avicenna’s distinction between thing and existent, and on its significance, see Jolivet, Aux
32 origines; Wisnovsky, Avicenna’s Concept of Thingness, which was revised and expanded
33 in Wisnovsky, Avicenna’s Metaphysics in Context, pp. 145 – 80. On some Latin
34 interpretations of Avicenna’s notions of being and thing (by Thomas Aquinas, Henry
of Ghent, and Francis of Marchia), see Aertsen, Avicenna’s Doctrine, pp. 27 – 33.
35
7 Avicenna, The Metaphysics of ‘The Healing’, I, 5, 9, p. 24. See the Latin translation in
36 Avicenna Latinus, Liber de philosophia prima, vol. 1, pp. 34 – 5: ‘Sed res et quicquid
37 aequipollet ei, significant etiam aliquid aliud in omnibus linguis; unaquaeque enim res
38 habet certitudinem qua est id quod est, sicut trangulus habet certitudinem qua est
39 triangulus, et albedo habet certitudinem qua est albedo. Et hoc est quod fortasse
appellamus esse proprium …’.
40
8 As far as I know, these questions have not been adequately considered in the literature on
41 Latin medieval authors, which focuses almost exclusively on the relationship between
42 essence and existence.
1 characterizations of what it is to be a thing. It is certainly not the case that we

2 can talk only about things with a certain essence, such as human beings, cats,
3 trees and the like. We can say of some things, for example of elves, that they do
4 not exist. In that case, we are still talking about something. But we certainly do
5 not want to attribute any ontological status to what we are talking about.
6 Rather, when we say that elves do not exist, our point is just to deny that elves
7 are real things. Similarly, we may describe blindness as an unfortunate situation
8 that some human beings and some animals experience. Thus, we do talk about
9 blindness. At the same time, we may be understandably unwilling to grant
10 blindness any ontological status. Avicenna’s distinction between two descriptions
11 of what it is to be a thing captures exactly this point. Sometimes we speak about
12 things to which we do not want to attribute any reality in the extramental world.
13 More precisely, sometimes we speak about things without taking any stance as to
14 whether those things have any ontological status and, if they do, as to what that
15 is. Rather than distinguishing between different senses or uses of ‘to exist’ or ‘to
16 be,’ Avicenna distinguishes between things as the objects of true statements and
17 things as real constituents of the world, i. e. essences.
18 Several Western readers of Avicenna tried to make the most of that
19 distinction. In his commentary on the Sentences, Thomas Aquinas first
20 distinguished between the term ‘thing’ and the term ‘being’ according to
21 whether they refer to what something is, i. e. its quiddity and essential account,
22
or to the fact that something exists, i. e. its existence. Aquinas goes on to claim
23
that a quiddity can exist in two ways, i. e. in a singular thing outside the mind
24
and in the mind itself as something apprehended by the intellect. For example,
25
suppose that I have a concept of what a cat is. In that case, the quiddity cat exists
26
both in individual cats and in my mind. Clearly, Aquinas is making use of
27
Avicenna’s famous view, which Avicenna presented in Metaphysics V and
28
elsewhere, that a certain essence can be considered in three ways, i. e. in itself,
29
according to the being it has in extramental individuals and according to the
30
being it has in the mind as something understood.9 What is particularly
31
interesting, however, is the way Aquinas connected the passage from Metaphysics
32
V, where Avicenna introduced the distinction between the three ways an essence
33
can be considered, with the passage from Metaphysics I.5, which we are currently
34
taking into account. For Aquinas went on to notice that to the two ways a
35
quiddity can exist, namely in the mind and outside the mind, there correspond
36
two meanings of the word ‘thing’. To the quiddity existing in extramental
37
38
individuals there corresponds the meaning of ‘thing’ as something determined
39
9 Avicenna, The Metaphysics of ‘The Healing’, V, 1, 4, p. 149. See Avicenna Latinus, Liber
40
de philosophia prima V, 1, vol. 1, pp. 228 – 9. See Marmura, Quiddity and Universality;
41 id., Avicenna’s Chapter on Universals; de Libera, La querelle, pp. 185 – 91; Black, Mental
42 Existence, pp. 47 – 51.
370 Giorgio Pini
1 and fixed in nature (res dicitur quasi aliquid ratum et firmum in natura). To the
2 quiddity existing in the soul there corresponds the meaning of ‘thing’ (in Latin,
3 res) taken from the verb reor-reris, namely ‘to opine’ or ‘to think’.10
4 By drawing this distinction between two meanings of res – as what is in the
5 mind when I think about something and as a real constituent of the world –
6 Aquinas tried to capture the distinction he found in Avicenna between, on the
7 one hand, a thing as that about which a true statement is made and, on the other
8 hand, a thing as an entity having a certain essence. Aquinas’s assumption is that,
9 when I think about a certain object, that object is both in the extramental world
10 and in my mind. Also, that object, say the essence of cats, can be called a ‘thing’
11 both when it is in the extramental world and when it is in my mind. But when I
12 say that the object I am thinking about is a thing in the world and is also a thing
13 in the mind, I am using ‘thing’ in two different senses. Thus, things in the mind
14
are not another kind of things over and above things in the world; rather, they
15
are things in a different sense of the word ‘thing’.
16
There is, however, a problem with this reading of Avicenna. Aquinas can
17
easily account for our thoughts about objects such as cats and human beings.
18
These objects are things in the world in the first meaning of ‘thing’ and, when I
19
think about them, they are also things in my mind in the second meaning of
20
‘thing’. Suppose, however, that I think about a nonexistent object, such as a
21
chimera or a goatstag (these are the standard medieval examples of imaginary
22
23
beings). Clearly, I can make several true statements about those objects; for
24
example, I can say that chimeras do not exist. In that case, there is something in
25
my mind but nothing corresponds to it in the extramental world, according to
26
Aquinas’s account in the passage we are considering. But then, what is this ‘thing
27 in the mind’? Surprisingly, Aquinas is committed to saying that it is a quiddity
28 or essence. For Aquinas, in the passage I am referring to, explicitly identified a
29 thing as that about which a true statement is made with a quiddity as existing in
30 the mind (quidditas … in anima secundum quod est apprehensa ab intellectu).
31 Admittedly, there is no individual exemplifying that essence and accordingly
32
33 10 Thomas Aquinas, In I Sent., d. 25, q. 1, a. 4: ‘Secundum Avicennam, ut supra dictum
34 est, hoc nomen “ens” et “res” differunt secundum quod est duo considerare in re, scilicet
quidditatem et rationem eius, et esse ipsius. Et a quidditate sumitur hoc nomen “res”. Et
35
quia quidditas postest habere esse et in singulari quod est extra animam et in anima
36 secundum quod est apprehensa ab intellectu, ideo nomen rei ad utrumque se habet: et ad
37 id quod est in anima, prout “res” dicitur a “reor, reris”, et ad id quod est extra animam,
38 prout “res” dicitur quasi aliquid ratum et firmum in natura. Sed nomen entis sumitur ab
39 esse rei.’ Aquinas distinguished between the two meanings of ‘thing’ also in In II Sent.,
d. 37, q. 1, a. 1. Before Aquinas, Bonaventure had drawn the same distinction in In II
40
Sent., d. 37, dub. I, p. 876. On Bonaventure’s and Aquinas’s passages, see Aertsen,
41 Transcendental Thought, pp. 11 – 14. On the claim that res comes from reor–reris, see
42 Hamesse, Res chez les auteurs philosophiques, pp. 91 – 104.
1 that essence is not something ‘determined and fixed in nature’. Nevertheless,

2 only because I can think about chimeras, there must be an essence of chimeras,
3 even though that essence is not instantiated. This claim, however, is extremely
4 costly from an ontological point of view. Elsewhere, Aquinas himself did not
5 seem to be willing to endorse it. Thus, we should conclude that Aquinas’s
6 interpretation of Avicenna’s distinction between two descriptions of what it is to
7 be a thing does not account for our thoughts about nonexistents, unless we are
8 willing to populate the world with uninstantiated essences. The same reasoning
9 can be extended from nonexistent things such as chimeras to privations such as
10 blindness, of which Aquinas explicitly denied that they possess any positive
11 ontological status.
12 The root of the problem is the correlation that Aquinas here posited
13 between the meaning of ‘thing’ as something in the mind and an essence’s way
14 of being in the mind. Accordingly, it does not come as a surprise that Aquinas
15 did not come back to this view when he considered the issue more carefully.
16 Elsewhere, he preferred to speak not of two meanings of ‘thing’ correlated to
17 two ways of being of an essence; rather, he introduced two main meanings of ‘to
18 be’ (esse). In its first meaning, ‘to be’ is said of an extramental essence (or its
19 actual existence). For example, we say that cats and human beings are or are
20 beings. In its second meaning, ‘to be’ is said of a truthmaker, i. e. of what a
21 certain statement is about and what makes a statement true. For example, we say
22 that Socrates is or is a being because Socrates is what the statement ‘Socrates is
23 white’ is about.11 Now Aquinas held that the second meaning of ‘to be’, i. e. as
24 that about which a statement is made, does not entail any direct reference to an
25 essence, whether extramental or merely in the mind. Thus, we can speak about
26 chimeras and goatstags without positing any quiddity of chimeras and goatstags.
27 Similarly, we can speak about blindness without positing any quiddity of
28 blindness.12
29 Some years after Aquinas, Henry of Ghent referred repeatedly to the
30 distinction between two meanings or, as he said, concepts of what it is to be a
31
32 11 Aquinas explicitly takes the second meaning of ‘to be’ from Aristotle, Met. V, 7,
33 1017a31 – 5.
34 12 See for example Aquinas, ST I, q. 3, a. 4; De pot., q. 7, a. 2, ad 1; In V Met., lect. 9,
nn. 889 – 96; see also In I Sent., d. 33, q. 1, a. 1, ad 1, ed. Mandonnet, pp. 765 – 6
35
(where Aquinas distinguishes between not two but three meanings of ‘to be’). See
36 Wippel, The Metaphysical Thought, pp. 24 – 5. It seems that, according to Aquinas,
37 chimeras, goatstags and the like count as things that are both not actual and not possible.
38 Such things have no essence and no ontological status whatsoever. By contrast,
39 nonexisting possible things, such as my nonexisting twin brother, do have some sort of
ontological status. Their ontological status as possible things is grounded on divine ideas.
40
It should be remembered, however, that Aquinas held that divine ideas are really identical
41 with God’s essence. As a consequence nonexisting possibles have no reality outside and
42 independently of God’s essence. See Wippel, The Reality of Nonexisting Possibles.
372 Giorgio Pini
1 thing.13 Henry was not always consistent in his interpretation of the two
2 concepts of what it is to be a thing. Here I will focus on his treatment in
3 Quodlibet VII, q. 1 – 2. In that work, Henry held that there are three main
4 concepts of what it is to be a thing. First, according to the most general concept,
5 a thing is ‘what is not nothing’, where ‘nothing’ means what neither is nor can
6 be, either in the mind or outside the mind. This most general concept of thing is
7 divided into two further concepts.14 Thus, second, a thing is what is called
8 ‘thing’ or ‘res’ from the verb reor-reris, i. e. something that is or can be only in the
9 mind as an object of thought. It is in this sense that fictitious beings, such as
10 golden mountains or goatstags, are things.15 Third, a thing is what is called
11 ‘thing’ or ‘res’ from the term ratitudo, which indicates the possession of an
12 essence. Only things that exist or can exist outside the mind are things in this
13 sense.16 These things are further divided into created and uncreated things, and
14
created things are further divided into the ten Aristotelian categories.
15
Thus, we can think and speak about golden mountains and goatstags even
16
though those objects do not have an essence and are called ‘things’ only in the
17
sense that they are objects of thought. Accordingly, Henry held that the scope of
18
what we can think and speak about is not limited to real being, i. e. what is or
19
can be in a natural kind by virtue of its essence. Nevertheless, Henry still linked
20
the concept of what it is to be a thing as an object of thought to the concept of
21
what it is to be a thing as a real essence. He actually made the former concept
22
23
dependent on the latter. Admittedly, golden mountains and goatstags are not
24
13 See for example Henry of Ghent, Quodl. V, q. 2, fol. 154rD; Summa (Quaestiones
25
ordinariae), art. 34, q. 2, pp. 174 – 5; Summa a. 21, q. 4, fols 127rO–127vO. See Aertsen,
26 Transcendental Thought, pp. 2 – 10; Porro, Possibilit ed esse essentiae, pp. 226 – 228. See
27 also Porro, Universaux et esse essentiae, pp. 9 – 51.
28 14 Henry of Ghent, Quodl. VII, q. 1 – 2, pp. 26 – 7: ‘… sciendum quod omnium
29 communissimum, omnia continens in quodam ambitu analogo, est res sive aliquid, sic
consideratum ut nihil sit ei oppositum nisi purum nihil, quod nec est nec natum est esse,
30
neque in re extra intellectum, neque etiam in conceptu alicuius intellectus, quia nihil est
31 natum movere intellectum nisi habens rationem alicuius realitatis. Res autem sive
32 “aliquid” sic communissime acceptum, non habet rationem praedicamenti, – sic enim
33 esset tantum unum praedicamentum continens Creatorem et creaturam –, sed
34 distinguitur distinctione analogica in id quod est aut natum est esse tantum in conceptu
intellectus sive in ipso intellectu, et in id quod cum hoc aut est aut natum est esse in re
35
extra intellectum.’
36 15 Ibid.: ‘Res primo modo est “res” secundum opinionem tantum, et dicitur “a reor, reris”,
37 quod idem est quod “opinor, opinaris” quae tantum res est secundum opinionem, quoad
38 modum quo ab intellectu concipitur, scilicet in ratione totius, ut est mons aureus, vel
39 hircocervus habens medietatem cervi, medietatem hirci.’
16 Ibid.: ‘Aliquid autem, sive res nata esse vel quae est aliquid extra intellectum, quae dicitur
40
“res a ratitudine”, adhuc non habet rationem generis aut praedicamenti sicut neque prius,
41 sed dividitur divisione analogica in id quod est aliquid quod est ipsum esse, et in id quod
42 est aliquid cui convenit aut natum est convernire esse.’
1 real things. According to Henry, however, such entities are the result of the
2 combination of real things. As Henry said in his Quodl. VII, the reason why a
3 golden mountain and a goatstag can be thought about is that, even though they
4 do not have a proper essence (and so they are not things in the second, stronger
5 sense of ‘thing’), their concepts nevertheless result from the combination of the
6 concepts of two or more real essences (the concept of a mountain and the
7 concept of gold, the concept of a goat and the concept of a stag):
8
[A thing as an object of opinion] is nevertheless a real thing with regard to what a
9
mountain is and what gold is and the like. For such a combination could not be in
10 the intellect and could not be a being according to opinion unless its parts were
11 something real, because the intellect cannot be moved by anything different.17
12 (Transl. mine)
13
It is our imagination, it seems, that carries out the combination of two distinct
14
essences resulting in entities such as golden mountains and goatstags. As to our
15
intellect, it thinks about such imaginary things as golden mountains and
16
goatstags as if each of them were a single essence, even though they are just the
17
result of a combination that cannot obtain in the extramental world.18 As a
18
matter of fact, it is not immediately clear why Henry thought that this is the
19
case. It may be admittedly difficult to gather enough gold to build up a
20
mountain, but there seems to be nothing physically impossible about this.
21
Probably, Henry would contend that it is part of the very concept of what it is to
22
23
be a mountain that a mountain is constituted not entirely by gold. Similarly, it
24
may be contended that it is part of the very concept of what it is to be a goat
25
that what is a goat cannot be combined with a stag, because goats and stags have
26
incompatible essential properties. Be that as it may, Henry’s conclusion is that
27
something can be an object of thought if either it is a real thing, i. e. an
28
instantiated or instantiable essence, or it is an imaginary combination of real
29 essences. Accordingly, the weaker meaning of what it is to be a thing (i. e., a
30 thing as an object of thought) is dependent on the stronger meaning of what it
31 is to be a thing (i. e., a thing as an essence). We could not have concepts of
32 things, in the weak sense of ‘thing’, if we had not been previously acquainted
33
34 17 Ibid., p. 27: ‘Est tamen res secundum veritatem quoad partes eius quae sunt mons et
aurum et huiusmodi; aliter enim non posset totum esse in intellectu et ens secundum
35
opinionem, nisi partes essent aliquid secundum veritatem, quia ab alio non potest moveri
36 intellectus.’ The importance of this passage was stressed by Aertsen, Transcendental
37 Thought, p. 5.
38 18 Henry of Ghent, Summa, a. 34, q. 2, pp. 174 – 5: ‘Ita quod ratio entis sit ratio primi
39 conceptus obiective in intellectu, quia “quod quid est, est proprium obiectum intellectus”
secundum Philosophum, ut etiam ratio rei a reor dictae non potest concipi ab intellectu
40
– licet possit ab imaginatione – nisi sub ratione entis quidditativi … Quia autem verum
41 est prima ratio qua aliquid est conceptibile ab intellectu, ut praedictum est, ideo tertia
42 ratio veri fundatur in ratione entis quidditativi, quod dicitur res a ratitudine …’
374 Giorgio Pini
1 with things in the strong sense. We could not think about goatstags if we did
2 not have the concept of what a goat is and of what a stag is. The reason for this
3 is clear, according to Henry. Only what has a real essence can act upon our mind
4 and ‘move’ our mind, i. e. cause a concept in our mind.
5 The general idea behind Henry’s position is that human beings are naturally
6 geared or predisposed to think about real essences because our minds are such
7 that they can be acted upon only by real essences.19 And Henry’s assumption is
8 that what we can think about is primarily what acts upon our minds. Thus, it is
9 true that we can think about imaginary beings such as golden mountains and
10 goatstags, but these are deviant cases that must be accounted for in terms of
11 imaginary combination of real essences. The scope of what our mind can think
12 about is primarily defined by what is real, i. e. (for Henry) what either is or can
13 become actual. Only secondarily and in a derivative sense can we think about
14 what is not real, i. e. what neither is nor can become actual, and we can do that
15 only to the extent to which what is not real is the result of a combination of
16 what is real. A combination such as golden mountain or goatstag neither is nor
17 can become actual, but its components are or can become actual.
18 In conclusion, Henry’s division is basically equivalent to the early Aquinas’s
19 distinction of the two meanings of ‘thing’, apart from an important difference.
20 Something, in order to be thought about, does not have to be a quiddity or
21 essence. Only its components do. Thus, Henry was not committed to the
22 implausible claim that when I think about chimeras I think about a certain
23 quiddity or essence. Still, conceivability is connected with the possession of an
24 essence, even though in a roundabout way, according to Henry. Admittedly,
25 there is no such an essence as the essence of a chimera or the essence of a golden
26 mountain. But what it is to be a chimera and what it is to be a golden mountain
27 must be analyzed into their components. And those components do have an
28 essence. Thus, a necessary condition for something to be thought about or
29 talked about is for it to have some sort of being – because, as Avicenna had said,
30 nothing affirmative can be stated of what is absolutely non-being. And the sort
31 of being that what is thought or talked about has is either essential being or
32 dependent on essential being (i. e. an imaginary combination of two essences).
33
34
35
36
37 19 Admittedly, extramental things ‘act upon’ our minds in a complicated way, which is
38 supposed to save our mind’s active role in the act of thinking. Henry, however, admitted
39 that our mind is passive with regard to its object in the first stage of cognition. See Henry
of Ghent, Quodlibet 9.2, p. 26: ‘Est igitur primo sciendum quod nihil intelligit aliquid
40
nisi id quod est per se obiectum virtutis qua cognoscit, vel id cuius illud per se obiectum
41 est ratio cognoscendi, quia virtus qua intellectus intelligit, quae est intellectus eius, passiva
42 est, quae non movetur ad actum intelligendi nisi a per se obiecto’ (italics mine).
1 II
2
3 In the third question of his Quodlibet, Scotus endorsed a distinction of the
4 meanings of ‘thing’ and ‘being’ that is very similar to that adopted by Henry of
5 Ghent, who was most probably Scotus’s immediate source.
6 Like Henry, Scotus started with the remark that ‘thing’, in its most general
7 meaning, is opposed to nothing. He then went on to notice that the term
8 ‘nothing’ has two meanings. First, ‘nothing’ means what includes a contra-
9 diction, i. e. what is constituted by two incompatible notions, such as ‘round
10 square’ (my example). In this sense of ‘nothing’, nothing is what cannot exist
11 either in the mind (as an object of thought) or outside the mind (as a real
12 thing). Accordingly, here we are not dealing with any object – either an object of
13 thought or a thing in the real world. Rather, we are just dealing with two
14 incompatible notions. When we say that a round square is nothing we are not
15 saying that there is a special kind of object that has the property of not existing
16 in reality; rather, we are saying that the notions round and square are such that
17 they are logically incompatible.20 Second, ‘nothing’ means what does not and
18 cannot exist extramentally.21
19 Since ‘thing’ (or ‘being’) means ‘not nothing’, corresponding to these two
20 meanings of ‘nothing’ there are two meanings of ‘thing’ (or ‘being’). The first
21 meaning of ‘thing’ and ‘being’ is ‘something that does not include a contradiction’.
22
23
24
20 Duns Scotus, Quodl., q. 3, n. 2, p. 113: ‘… hoc nomen res potest sumi communissime
25
et strictissime. Communissime, prout se extendit ad quodcumque quod non est nihil. Et
26 hoc potest esse dupliciter. Verissime enim illud est nihil quod includit contradictionem,
27 et solum illud, quia illud excludit omne esse extra intellectum et in intellectu. Quod
28 enim est sic includens contradictionem, sicut non potest esse extra animam, ita non
29 potest esse aliquid intelligibile ut aliquod ens in anima, quia nunquam contradictorium
cum contradictorio constituit unum intelligibile, neque sicut obiectum cum obiecto
30
neque sicut modus cum obiecto.’ See also Ord. I, d. 43, q. unica, n. 15, ed. Vat. VI, p.
31 359: ‘Et ex hoc apparet quod falsa est imaginatio quaerentium impossibilitatem
32 aliquorum quasi in aliquo uno, quasi aliquid unum – vel intelligibile vel qualecumque
33 ens – sit ex se formaliter impossibile sicut Deus ex se formaliter est necesse esse. Nihil
34 enim est tale primum in non-entitate, nec etiam entitatis oppositae tali non-entitati est
intellectus divinus ratio possibilitatis oppositae; nec etiam intellectus divinus est praecisa
35
ratio possibilitatis oppositae de nihilo, quia tunc teneret illud argumentum “de causis
36 praecisis in affirmatione et negatione”. Sed omne “simpliciter nihil” includit in se
37 rationes plurium, ita quod ipsum non est primo nihil ex ratione sui, sed ex rationibus
38 illorum quae intelligitur includere, propter formalem repugnatiam illorum inclusorum
39 plurium; et ista ratio repugnatiae est ex rationibus formalibus eorum, quam repugnatiam
primo habent per intellectum divinum’. It should be noticed that in these passages,
40
Scotus took ‘thing’ (res) and ‘being’ (ens) as synonymous.
41 21 Quodl, q. 3, n. 2, p. 113: ‘Alio modo dicitur nihil quod nec est nec esse potest aliquod
42 ens extra animam.’
376 Giorgio Pini
1 In this sense of ‘thing’ and ‘being’, both purely mental and extramental entities
2 are things and beings:
3
Being [ens] or thing [res] in the first of its very broad meanings, therefore, covers
4 anything that does not include a contradiction whether it be a conceptual being [ens
5 rationis], i. e., having being or existence only in the thinking intellect, or a real being
6 [ens reale], i. e., having some entity outside the consideration of the intellect.22
7 (Transl. Alluntis and Wolter)
8 By contrast, the second meaning of ‘thing’ (or ‘being’) is ‘something that has or
9 can have some entity independently of the intellect’s consideration’. In this sense
10 of ‘thing’ and ‘being’, only objects that exist or can exist in the extramental
11 world are said to be things or beings.23
12 Scotus held that there are some objects that are things in the first, more
13 general sense of the word ‘thing’ but not in the second, less general sense of the
14 word ‘thing’. His example is logical concepts and conceptual (or rational)
15 relations.24 Logical concepts are concepts such as species and genus, i. e. the so-
16 called ‘second intentions’. Conceptual (or rational) relations are relations that
17 hold among things only because of the way we describe them but are not
18 grounded on any real feature and as such can be acquired and lost without any
19 real change in either of the extremes (an example is the relation of being to the
20 left of something; depending on our position, a chair can be described as to the
21 left or to the right of a table, even though the chair has undergone no change of
22 position). Let us focus on logical concepts. There is no contradiction included
23 in the notion of what it is to be a species or what it is to be a genus. Accordingly,
24 we can think about what it is to be a species and what it is to be a genus.
25 Nevertheless, there is nothing in the extramental world whose essence is to be a
26 species or to be a genus. Rather, in the extramental world there are real things such
27 as cats and human beings, which can be classified in species and genera. But to
28 be a species or to be a genus is not something in the world; rather, it is a
29 conceptual (or rational) relation holding between concepts representing things
30
31
32
33 22 Ibid.: ‘Ens ergo vel res isto primo modo accipitur omnino communissime, et extendit se
34 ad quodcumque quod non includit contradictionem, sive sit ens rationis, hoc est praecise
habens esse in intellectu considerante, sive sit ens reale, habens aliquam entitatem extra
35
considerationem intellectus’. The English translation is taken from John Duns Scotus,
36 God and Creatures, p. 61.
37 23 Ibid.: ‘Et secundo accipitur in isto membro minus communiter pro ente quod habet vel
38 habere potest aliquam entitatem non ex consideratione intellectus’.
39 24 Ibid.: ‘Et istorum duorum membrorum … primum videtur valde extendere nomen rei,
et tamen, ex communi modo loquendi, satis probatur. Communiter enim dicimus
40
intentiones logicas esse res rationis, et relationes rationis esse res rationis, et tamen ista
41 non possunt esse extra intellectum.’ On second intentions, see G. Pini, Categories and
42 Logic, pp. 45 – 137.
1 in the world (i. e. objects that are said to be things in the second, less general
2 meaning of ‘thing’).
3 I think that Scotus would say that chimeras and goatstags are things in this
4 first, more general sense of ‘thing’. Chimeras and goatstags are nothing in the
5 sense that they are nothing actual. Scotus also agreed with Henry that chimeras
6 and goatstags cannot become actual. Scotus’s reason for this, however, is not that
7 there is no single essence corresponding to the term ‘chimera’ or the term
8 ‘goatstag’. Rather, Scotus’s claim is that the notion of a chimera and the notion
9 of a goatstag are such that they cannot be instantiated in the extramental world,
10 and apparently he takes this as a primitive fact. There is a contradiction between
11 what it is to be a chimera and to exist in the extramental world. The notion of
12 existing in the extramental world is ‘repugnant’ or logically incompatible with
13 what it is to be a chimera. It seems that the impossibility of existing in the
14 extramental world is just part of what it is to be a chimera. Nevertheless, even
15
though the notion of a chimera and the notion of extramental existence are in
16
contradiction with each other, there is no contradiction included in the notion
17
of what it is to be a chimera. Thus, a chimera can be an object of thought, even
18
though it cannot possibly exist.25
19
Scotus claimed that it is in the first meaning of ‘being’ and ‘thing’ that being
20
and thing are the first object of the intellect, because the intellect can think
21
about anything that does not include a contradiction. Scotus, however, was
22
hesitant to say whether Avicenna spoke of ‘thing’ in this most general sense in
23
his Metaphysics I.5. At first, Scotus stated that when Avicenna spoke of ‘thing’
24
25
and ‘being’, he spoke of both in the second, less general sense of those words,
26
i. e. as what has or can have existence outside the mind. Accordingly, when
27
Avicenna said that being and thing are the first concepts occurring to our mind
28
and that those concepts are common to everything, he should be interpreted as
29
saying that the first concepts our mind acquires are the concepts of things that
30 exist or can exist.26 Later in the same question, however, Scotus showed some
31 hesitation in construing Avicenna’s intention. He conceded that Avicenna might
32 be speaking of ‘being’ and ‘thing’ in the first, more general sense. Accordingly,
33
34 25 Scotus considered entities such as chimeras as ‘fictitious beings’ (figmenta) and he
defined what is not a fictitious being as that for which it is possible to exist. This sort of
35
existence is extramental existence. Accordingly, a fictitious being is an object of thought
36 such that real existence is repugnant to it, i. e. for which it is impossible to exist
37 extramentally. See Ord. I, d. 36, q. unica, nn. 48 – 50, ed. Vat. VI, pp. 290 – 91; nn. 60 –
38 63, pp. 296 – 7.
39 26 Quodl. q. 3, n. 2, p. 114: ‘In secundo autem membro istius primi membri, dicitur res
quod habere potest entitatem extra animam. Et isto modo videtur loqui Avicenna, 1
40
Metaphysicae, c. 5, quod ea quae sunt communia omnibus generibus sunt res et ens, nec
41 potest illud intelligi de vocabulis in una lingua, quia in unaquaeque lingua est unus
42 conceptus indifferens ad omnia illa quae [ed.: qua] sunt extra animam …’
378 Giorgio Pini
1 Avicenna should be interpreted as saying that the first concept common to

2 everything is the concept of what is non-contradictory. Scotus, however, added
3 that Avicenna might also be referring to both senses of ‘thing’ and ‘being’. In any
4 case, he was quite certain that Avicenna was referring at least to the second sense
5 of ‘thing’ and ‘being’, i. e. as what has or can have extramental existence.27
6 Scotus’s distinction between the two main meanings of ‘thing’ and ‘being’ is
7 clearly indebted to Henry of Ghent. Through Henry, Scotus was also giving his
8 interpretation of Avicenna’s teaching on ‘thing’ and ‘being’ and on Avicenna’s
9 distinction between two descriptions of what it is to be a thing.
10 There are, however, two important points with regard to which Scotus’s
11 account of the meaning of the word ‘thing’ (and, correspondingly, Scotus’s
12 interpretation of Avicenna) is different from Henry’s treatment. The first
13 difference concerns Scotus’s interpretation of the first meaning of ‘thing’. The
14 second difference concerns Scotus’s interpretation of the second meaning of
15 ‘thing’.
16 As to the first meaning of the word ‘thing’, Scotus’s description of a thing as
17
what does not include any contradiction does not make any reference to the
18
notion of essence. Henry of Ghent had maintained that a thing as a purely
19
mental item is not a single essence but is nevertheless an imaginary combination
20
of several real essences. As a consequence, Henry held that even items that are
21
things only in the sense of possible objects of thought can be thought about at
22
all because there are things that have real essences, and the latter things are the
23
elements out of which purely mental things are constituted. Thus, Henry of
24
Ghent’s assumption was still that the mind is naturally turned to grasping
25
essences, i. e. things that either are or can become actual constituents of the
26
world. By contrast, Scotus shifted the focus from essences to what is not
27
contradictory. In the question we are currently taking into consideration, he
28
29
explicitly stated that the object of the intellect is anything that does not include
30
a contradiction. Thus, he defined the object of the intellect by way of a merely
31
formal and logical requirement. Quite simply, in order for something to be
32
thinkable, it is sufficient for it not to include a contradiction. It is not necessary
33
for it to be an essence or a combination of essences.
34 Scotus’s method to find out what can be thought about and what cannot be
35 thought about has a considerable advantage. Scotus could decide if something is
36 thinkable by simply testing whether it satisfies a formal requirement. Thus,
37 logical intentions and conceptual or rational relations can be accounted for as
38
39 27 Quodl. q. 3, n. 3, pp. 114 – 5: ‘Habemus ergo primum membrum sic, scilicet
communissime, bipartitum in illud, scilicet quod non includit contradictionem,
40
qualecumque esse habeat, et in illud quod habet vel habere potest proprium esse extra
41 intellectum, et isto modo, vel utroque modo, vel saltem secundo modo accipit Avicenna
42 rem et ens, ut dictum est’.
1 objects of thought just because they do not include any contradiction. There is
2 no need to ground logical concepts and conceptual or rational relations on real
3 essences in order to account for their thinkability. There is no need to know
4 what there is out in the world to find out whether something is thinkable or not.
5 The test of thinkability is the satisfaction of the logical requirement of being
6 internally consistent, i. e. non-contradictory. Concepts do not have to be tested
7 against extramental essences in order to be thinkable.
8 Accordingly, Scotus’s approach, has a remarkable consequence. According to
9 Scotus, the mere fact that we can think about something leaves any question
10 concerning its ontological status undecided. Just because I am able to think
11 about something I am not in a position to say that that thing exists
12 extramentally or even that it can exist extramentally. Undeniably, logical
13 intentions, conceptual relations and chimeras can be thought about. Never-
14
theless, they cannot exist extramentally. Thus, there are things (in the first
15
meaning of ‘thing’) that God cannot bring about in extramental existence.28
16
Ontological questions concerning the real possibility of something cannot be
17
decided by virtue of merely logical criteria. In other words, Scotus distinguished
18
between, on the one hand, logical possibility as a formal relation holding
19
between concepts and, on the other hand, real or metaphysical possibility, if we
20
interpret ‘metaphysical possibility’ as the property of not being incompatible
21
with actual existence in at least one possible world (be it the actual world or
22
another world).29 There are things that satisfy the logical criterion of not
23
24
including a contradiction and there are things that satisfy the metaphysical
25
criterion of being compatible with actual existence. These, however, are not two
26
kinds of things that can be compared as more or less real; rather, these are things
27
in two different senses of the word ‘thing’. When we say that a chimera or a
28 logical concept is less real than a human being we are making what in
29 contemporary jargon is called a ‘category mistake’. Chimeras and logical
30 concepts are not less real than human beings; rather, they are not real at all. To
31 be real is a predicate that can be applied to what is a thing only in the second,
32 less general sense of thing, i. e. as what either is or can become actual.
33 As to the second meaning of ‘thing’, Scotus agreed with Henry of Ghent
34 that the word ‘thing’, in its second meaning, refers to all real essences, because,
35 necessarily, if x is a real essence, x is a thing in the second meaning of ‘thing’, i. e.
36 as that which has or can have extramental existence. From what Scotus said
37
38 28 Nevertheless, Scotus still endorsed the common claim that God can do whatever is non-
39 contradictory, for he held that it is contradictory for a logical concept or a chimera to
exist extramentally.
40
29 For a characterization of logical and metaphysical possibility, as well as of nomological
41 possibility, see Conceivability and Possibility, eds. T.S. Gendler and J. Hawthorne,
42 Oxford: Clarendon Press, 2002, pp. 4 – 5.
380 Giorgio Pini
1 elsewhere, however, it appears that he rejected the converse claim that,

2 necessarily, if x can have extramental existence, then x is an essence. It is not
3 difficult to see why this is the case if we consider that God could have created
4 some kinds of items that He actually did not create. The things that God could
5 have created but did not are things in the second sense of the word ‘thing’,
6 because it is possible for them to exist extramentally, or, as Scotus said,
7 extramental existence is not repugnant to (i. e., incompatible with) them.
8 Nevertheless, the things that God could have created but did not create are not
9 essences, according to Scotus. Something is an essence if and only if it was made
10 the object of God’s act of creation. As a consequence, it appears that Scotus,
11 unlike Henry of Ghent, held that there are no possible, i. e. non-actual, essences
12 (even though, given an actual essence, there are possible, i. e. non-actual
13 individuals that could but do not instantiate that essence).30 There are, however,
14 things that could have been created and could exist but do not. Such non-actual
15 things are called ‘things’ because existence is not repugnant to (i. e.,
16 incompatible with) them, but they are not essences. Scotus argued that the
17 same point holds even for essences, i. e. things that God actually created, when
18 they are considered as they were in themselves before God created them. By
19 contrast, Henry of Ghent had identified the second way something is said to be
20 a thing (as a res a ratitudine) with what it is to be an essence. According to
21 Henry, something can have extramental existence because it is a thing in the
22
sense of being or having an essence. Admittedly, an essence can fail to be
23
instantiated. In Henry’s jargon, an essence does not have to have existential
24
being (esse existentiae). Necessarily, however, if something can have extramental
25
existence, it is a thing in the sense of ‘essence’, i. e. it has essential being (esse
26
essentiae). Against Henry’s position, Scotus held that, before creation, a human
27
being and a chimera both lack an essence. Thus, before creation, the only
28
difference between a human being and a chimera is that it is not impossible for a
29
human being to be brought to existence, whereas it is impossible for a chimera
30
to be brought to existence. But there is an essence human being only once God
31
creates actual human beings in the world. Accordingly, Scotus held that to be
32
metaphysically possible (i. e., to be a thing in the second sense of ‘thing’ that
33
Scotus had distinguished in his Quodlibet, q. 3) and to be an essence are two
34
distinct notions. Whereas the latter entails the former, the former does not entail
35
the latter. Thus, what is metaphysically possible, i. e. what exists or can exist
36
extramentally, is not necessarily an essence. Rather, it is a contingent fact that
37
38
something metaphysically possible, for example the concept human being, is also
39
30 Scotus’s position is, in this respect, just the reverse of that of Henry of Ghent. For Henry,
40
there are non-actual essences but only actual individuals, because what accounts for an
41 essence’s being instantiated in a certain individual is that such an individual actually
42 exists. By contrast, for Scotus there are non-actual individuals but only actual essences.
1 an essence, because this is a fact that depends entirely on God’s will to create
2 human beings. There is nothing in the concept human being that entails its
3 being a certain essence. What is included in the concept human being is merely
4 the possibility of being a certain essence.31
5 To sum up, Scotus distinguished between (1) what is internally consistent,
6 i. e. logically possible; (2) what is not incompatible with extramental existence,
7 i. e. metaphysically possible; and (3) what has an essence, i. e. what belongs to a
8 natural kind. In order for a thing (or a being) to be thinkable, it is sufficient to
9 be a thing according to the first meaning of ‘thing’. What is a thing (or a being)
10 in this first meaning, however, is not necessarily compatible with extramental
11 existence. In turn, what is a thing (or a being) in the second meaning of ‘thing’,
12 i. e. as something with which extramental existence is not incompatible, does not
13 necessarily have an essence. Only an actual thing has an essence. Accordingly,
14 both actual entities and nonexistent possibles are things in the second sense of
15 ‘thing’, even though the former lack and the latter have an essence. To say of
16 something that it is a thing or a being – even in the second, stronger sense of
17 ‘thing’ or ‘being’ – does not tell us anything about its ontological status or even
18 whether it has some ontological status.
19
20
21 III
22
23 So far I have considered what Scotus took a thing to be and how he read
24 Avicenna accordingly. As I have indicated, Scotus held that the term ‘thing’ is
25 equivocal between something that does not include a contradiction and
26 something that does or can exist extramentally. And he maintained that
27 probably what Avicenna meant by ‘thing’ was the latter. Typical of Scotus’s
28 approach is also his view that the first sense of ‘thing’ as something internally
29 consistent is not dependent on the second sense of ‘thing’ as something that does
30 or can exist extramentally. Now I would like to turn to the third issue I
31 mentioned at the beginning of this paper, namely what role our concept thing or
32 being as something that is or can be extramental plays in cognition.
33
It turns out that the concept thing or being as something at least possibly
34
extramental plays an extremely important role in our cognition, both of the
35
material and of the immaterial world. In this respect, Scotus qualified what he
36
read in Avicenna. At the beginning of Metaphysics I.5, Avicenna made the
37
famous claim that thing and being are the first notions impressed in our souls.32
38
Scotus, however, was committed to the view that, in our current situation, our
39
40
31 Ord. I, d. 36, q. unica, nn. 60 – 63, ed. Vat. VI, pp. 296 – 7.
41 32 Avicenna, The Metaphysics of ‘The Healing’, I, 5, 1, p. 22. See the Latin translation in
42 Avicenna Latinus, Liber de philosophia prima, I, 5, vol. 1, p. 31.
382 Giorgio Pini
1 intellect is dependent on the senses for the information it receives about the
2 world.33 Now Aristotle had claimed that only sensible qualities such as colors,
3 sounds and the like are the per se objects of the senses.34 And Scotus fully agreed
4 with this claim. More precisely, Scotus argued that our senses are directly
5 acquainted not with individual instances of sensible qualities, such as this
6 singular instance of red or this singular instance of a certain sound. Rather,
7 Scotus held that our senses are directly acquainted with replicable features –
8 things such as this hue of red or this type of sound.35 These are what Scotus,
9 following a common usage, called species specialissimae. Since all cognitive
10 content comes to the intellect through the senses, it follows that the first thing
11 the intellect cognizes, in the order of origin, are replicable sensible features, such
12 as this kind of color or this kind of sound – the species specialissimae of sensible
13 qualities.36 The concepts being and thing become relevant only at the second
14 stage of the cognitive process, when we pass to what Scotus called ‘cognizing
15
something in a distinct way’, i. e. when we provide descriptions and definitions
16
of what we cognize.37 At first, however, we do not grasp the objects of our
17
cognition as beings and things, but as this sort of color, this sort of sound, and
18
so on–namely, as the least universal kinds of sensible qualities.
19
So how do we get from these sensible qualities to the concepts being and
20
thing? Scotus held that we arrive at the concepts being and thing, quite simply,
21
by abstraction from sensible qualities.38 Accordingly, Avicenna’s claim that thing
22
and being are the first notions impressed in our minds should be strongly
23
qualified. On the face of it, such a claim seems to be in contrast with Aristotle’s
24
25
view that the first objects of our cognition are sensible qualities. Thus, Scotus
26
reinterpreted Avicenna’s claim as referring to what is cognized in a distinct way,
27
i. e. when we set to provide descriptions and definitions of what we cognize.
28
Then and only then do being and thing have a primacy as the most general
29
concepts that are contained in the description of any other less general concept.
30 For example, if I have to give a definition or description of a human being, the
31 most general description I can provide is probably something of this sort:
32 ‘something that has a head, two legs, etc. and displays an intelligent behavior’.
33
34 33 Scotus, Ord. I, d. 3, p. 3, q. 1, n. 392, ed. Vat. III, p. 239. See also Ord. I, d. 3, p. 1,
q. 1 – 2, nn. 35 – 6, ed. Vat. III, pp. 21 – 2; Ord. II, d. 11, q. unica, n. 13, nn. 28 – 31,
35
ed. Vat. VIII, pp. 214 – 5, 222 – 4.
36 34 Aristotle, Cat. 8, 9a35-b7; De gen. et corr. II, 2, 329b19.
37 35 Scotus, Quaestiones super Metaphysicam, VII, q. 15, nn. 20 – 21, OPh IV, pp. 301 – 2;
38 Ord. II, d. 3, p. 1, q. 1, n. 21, ed. Vat. VII, pp. 399 – 400. See South, Scotus and the
39 Knowledge; Pini, Scotus on the Objects.
36 Ord. I, d. 3, p. 1, q. 1 – 2, nn. 72 – 3, ed. Vat. III, pp. 49 – 50.
40
37 Ord. I, d. 3, p. 1, q. 1 – 2, n. 80, ed. Vat. III, pp. 54 – 5.
41 38 Ord. I, d. 3, p. 1, q. 3, nn. 139 – 40, ed. Vat. III, pp. 86 – 8. See also Lect. I, d. 3, p. 1,
42 q. 1 – 2, n. 110, ed. Vat. XVI, p. 265.
1 In that description, there is no way to avoid such general labels as ‘something’,

2 ‘thing’, ‘being’. This is the less general meaning of ‘thing’ and ‘being’ Scotus had
3 distinguished in the passage we have considered above, i. e. ‘thing’ and ‘being’ as
4 something that has or can have extramental existence.
5 Scotus held that it is thanks to this concept thing and being that we can
6 extend our cognition beyond the species specialissimae of the sensible qualities we
7 are directly acquainted with. I see black and I touch soft. How do I pass from
8 these sensations to the perception of a cat? Scotus’s answer is that I infer the
9 presence of something underlying the sensible accidents. In this inference, the
10 notions of thing and being play a central role. After noticing that some sensible
11 qualities, say a certain color and a certain tactile impression, constantly occur
12 together, I infer that there must be something that accounts for their constantly
13 occurring together. Thus, I conclude that these sensible qualities occur together
14 because they inhere in the same thing, which I call a substance. With regard to
15 that underlying thing, I only know that it is something – a thing or being – that
16 accounts for the unity of certain sensible qualities, i. e. something in which
17 certain sensible qualities inhere.39 Accordingly, without the concept of being or
18 thing, it would be impossible for us to reconstruct the structure of the world.
19 Scotus admitted that the fact that we cannot grasp substances directly and
20 we rather have to infer that there are substances (if we want to account for the
21 regular conjoined occurrence of certain sensible qualities) is a lamentable
22
situation. He also thought that it is a contingent situation. Human beings were
23
not originally intended to cognize the world in this way. Rather, human beings
24
were supposed to get a direct grasp of substances – to know cats, dogs and other
25
human beings not on the basis of their sensible qualities but by inspecting their
26
very essences. To some extent, this will be the condition of the blessed in the
27
next life.40 The fact that we currently can cognize substances only by inference
28
from sensible qualities and through the concept thing and being is probably a
29
consequence of the fall of Adam. After the fall, our intellect became dependent
30
on the senses as its only source of information.41 Thus, Scotus thought that
31
Aristotle had given us an accurate description of the postlapsarian situation, but
32
not of our natural condition. Scotus, however, held that, even given the
33
cognitive limitations that characterize our current situation, we are still able to
34
get a fairly accurate picture of the world. What allows us to reach such a result is
35
our possession and use of the most general concepts being, thing, something and
36
the like. Thus, Avicenna had come across a very important point when he had
37
38
emphasized the role of thing and being as our most basic concepts, even though
39
39 Ord. I, d. 22, q. unica, n. 7, ed. Vat. V, pp. 344 – 5; Quaestiones super Metaphysicam, II,
40
q. 2 – 3, n. 115, OPh III, p. 233. See Pini, Scotus on Doing Metaphysics in statu isto.
41 40 Ord. II, d. 11, q. unica, n. 31, ed. Vat. VIII, p. 224.
42 41 See above, n. 33.
384 Giorgio Pini
1 he had not been aware of the reason why those concepts are so important in our
2 current situation, namely because they allow us to overcome our current
3 cognitive limitations and to reconstruct the structure of reality on the basis of
4 what we are directly acquainted with, i. e. sensible qualities. In that respect,
5 Avicenna went beyond Aristotle.
6 Avicenna’s views concerning thing and being also went beyond Aristotle with
7 regard to our current cognition of God. When Avicenna posited being and thing
8 as the objects of our intellect, taken in all their generality, he did not take into
9 account that in the current situation we are indeed limited to having cognitive
10 access only to sensible things. Thus, Avicenna was ignorant of the fact that there
11 is a sharp distinction between what our condition ought to be and what it is
12 now in the current, postlapsarian situation. We were not intended to rely on the
13 senses as we do now. Avicenna judged our cognitive powers according to their
14 own nature, not according to their current condition. According to Scotus,
15
Avicenna did so because he was being unwittingly influenced by his own
16
religious belief, which required that we should be able to see God in the next life
17
and thus that we should not be constitutionally limited to having cognitive
18
access only to sensible things.42 Scotus held that the influence of Islamic beliefs
19
on Avicenna’s philosophical positions, even though unconscious, was beneficial.
20
By ignoring our present cognitive limitations, Avicenna gave us a glimpse of
21
true nature, i. e. of what we were originally intended to know and what we shall
22
eventually know in the next life.
23
Thus, I think we should conclude that Scotus saw Avicenna as a salutary
24
25
complement to Aristotle. Both Aristotle and Avicenna were ignorant of the
26
difference between the present, postlapsarian condition and our original
27
condition. Accordingly, they both failed to appreciate the consequences that the
28
fall of Adam had on our cognitive powers. But their ignorance had opposite
29
results. Aristotle (and, to some extent, his follower Aquinas) took our present
30 cognitive limitation to sensible things as normal and constitutive of the human
31 condition. Accordingly, Aristotle and Aquinas thought that the proper object of
32 our intellect is the quiddity of sensible things, i. e. that our intellect is
33
34 42 Ord. prol., p. 1, q. unica, n. 33, ed. Vat. I, p. 19: ‘Ad aliud, negandum est illud quod
assumitur, quod scilicet naturaliter cognoscitur ens esse primum obiectum intellectus
35
nostri, et hoc secundum totam indifferentiam entis ad sensibilia et insensibilia, et quod
36 hoc dicit Avicenna quod sit naturaliter notum. Miscuit enim sectam suam – quae fuit
37 secta Machometi – philosophicis, et quaedam dixit ut philosophica et ratione probata,
38 alia ut consona sectae suae: unde expresse ponit libro IX Metaphysicae cap. 7 animam
39 separatam cognoscere substantiam immaterialem in se, et ideo sub obiecto primo
intellectus habuit ponere substantiam immaterialem contineri. Non sic Aristoteles; sed
40
secundum ipsum, primum obiectum intellectus nostri est vel videtur esse quiditas
41 sensibilis, et hoc vel in se sensibilis vel in suo inferiori; et haec est quiditas abstrahibilis a
42 sensibilibus’.
1 constitutionally limited to what is sensible. By contrast, Avicenna correctly

2 singled out being as the object of our cognitive powers, but he failed to consider
3 that in the current situation our grasp is, as a matter of fact, limited to sensible
4 things. Aristotle focused on what we are and neglected what we were and will
5 be. By contrast, Avicenna focused on what we should and will be, but he ignored
6 our present cognitive limitations. His neglect of our present state had the
7 fortunate effect of putting much emphasis on the concepts thing and being as the
8 objects of our intellect. It is by way of those concepts that we can measure up
9 what we lost, i. e. direct cognitive access to substances, and what we are
10 supposed to reach in the next life, i. e. direct cognitive access to God in His
11 individuality. Thanks to the concepts thing and being we can also reconstruct the
12 structure of reality, to the extent that we can do this in our present situation. By
13 the concepts thing and being, we can even get a sort of pale anticipation of the
14
knowledge of God’s essence that Scotus thought we shall experience in the next
15
life.
16
17
18 Bibliography
19
20 Primary Sources
21
22 Avicenna, Liber de philosophia prima sive de Scientia divina, 2 vols, ed. S. Van Riet,
23
Leuven/Leiden: Peeters/Brill, 1977 and 1980.
24
25 Bonaventure, In secundum librum Sententiarum (= In II Sent.), in id., Opera omnia II,
26 Quaracchi: Typographia Collegii S. Bonaventurae, 1885.
27 Henry of Ghent, Summa (Quaestiones ordinariae), a. 34, q. 2, in id., Opera omnia
28 XXVII, ed. R. Macken, Leuven: Leuven University Press, 1991.
29
–– , Summa quaestionum ordinariarum a. 21, q. 4, in id., Summa quaestionum
ordinariarum, Paris, 1520; reprinted St. Bonaventure, New York: Franciscan
30
Institute, 1953.
31 –– , Quodlibet V, q. 2, in id., Quodlibeta, Paris, 1518; reprinted Louvain: Publications
32 de l’Institut de Philosophie, 1961.
33 –– , Quodlibet VII, q. 1 – 2, in id., Opera omnia XI, ed. G. Wilson, Leuven: Leuven
35
–– , Quodlibet IX, q. 2, in id., Opera omnia XIII, ed. R. Macken, Leuven: Leuven
University Press, 1983.
36
John Duns Scotus, Opera omnia, Vatican City: Typis Polyglottis Vaticanis, 1950– (= ed.
37 Vat.).
38 –– , Quaestiones super libros Metaphysicorum Aristotelis, in id., Opera philosophica III
39 and IV (= OPh), St. Bonaventure, New York: Franciscan Institute, 1997 – 9.
40 –– , Quodlibet, in id., Opera omnia, 25 – 6, Paris: L. Vivs, 1895.
41 –– , God and Creatures. The Quodlibetal Questions, transls and eds F. Alluntis and
A.B. Wolter, Princeton/London: Princeton University Press, 1975.
42
386 Giorgio Pini
1 Thomas Aquinas, Scriptum super primum librum Sententiarum (= In I Sent.), ed.

2 P. Mandonnet, Paris: Lethielleux, 1929.
3
–– , Scriptum super secundum librum Sententiarum (= In II Sent.), ed. P. Madonnet,
Paris: Lethielleux, 1929.
4
–– , De potentia (= De pot.), ed. P.M. Pession, in id., Quaestiones disputatae, vol. 2,
5 Rome/Turin: Marietti, 1965, pp. 1 – 276.
6 –– , Expositio super duodecim libros Metaphysicorum Aristotelis (= In Met.) eds. M.-
7 R. Cathala and R.M. Spiazzi , Rome/Turin: Marietti, 1964.
8
9
10 Secondary Sources
11
12 J. Aertsen, Transcendental Thought in Henry of Ghent, in Henry of Ghent. Proceeding of
the International Colloquium on the Occasion of the 700th Anniversary of His Death
13
(1293), ed. W. Vanhamel, Leuven: Leuven University Press, 1996, pp. 1 – 18.
14 –– , Avicenna’s Doctrine of the Primary Notions and its Impact in Medieval Philosophy,
15 in Islamic Thought in the Middle Ages: Studies in Text, Transmission and Translation,
16 in Honour of Hans Daiber, eds A. Akasoy and W. Raven, Leiden: Brill, 2008,
17 pp. 21 – 42.
G.C. Anawati, Saint Thomas d’Aquin et la Mtaphysique d’Avicenne, in Saint Thomas
18
Aquinas, 1274 – 1974: Commemorative Studies, vol. 1, eds A. Maurer et al.,
19 Toronto: Pontifical Institute of Mediaeval Studies, pp. 449 – 65.
20 D. Black, Mental Existence in Thomas Aquinas and Avicenna, Mediaeval Studies, 61,
21 1999, pp. 45 – 79.
22 A. De Libera, La querelle des universaux. De Platon la fin du Moyen Age, Paris: Editions
23
du Seuil, 1996.
M.-T. Druart, Ibn Sina (Avicenna) and Duns Scotus, in Duns Scotus, Philosopher, eds
24
M.B. Ingham and O. Bychkov, Archa Verbi, Subsidia 3, Mnster: Aschendorff,
25 2010, pp. 13 – 27.
26 . Gilson, Avicenne et le point de depart de Duns Scot, Archives d’histoire doctrinale et
27 littraire du Moyen Age, 2, 1927, pp. 89 – 149, reprinted in . Gilson, Pourquoi saint
28 Thomas a critiqu saint Augustin, suivi de Avicenne et le point de depart de Duns Scot,
Paris: Vrin, 1986, pp. 129 – 89.
29
–– , Avicenne en Occident au Moyen Age, Archives d’historire doctrinale et littraire du
30 Moyen Age, 44, 1969, pp. 89 – 121.
31 J. Hamesse, Res chez les auteurs philosophiques des 12e et 13e sicles ou le passage de la
32 neutralit la spcificit, in Res. IIIo Colloquio Internazionale del Lessico Intellettuale
33 Europeo, eds M. Fattori and M. Bianchi, Rome: Edizioni dell’Ateneo, 1982,
pp. 91 – 104.
34
J. Jolivet, Aux origines de l’ontologie d’Ibn Sina, in tudes sur Avicenne, eds J. Jolivet and
35 R. Roshdi, Paris: Les Belles Lettres, 1984, pp. 11 – 18, reprinted in J. Jolivet,
36 Philosophie mdivale arabe et latine, Paris: Vrin, 1995, pp. 221 – 36.
37 M. Marmura, Avicenna on Primary Concepts in the Metaphysics of his al-Shifā’, in Logos
38 Islamikos: Studia Islamica in honorem Georgii Michaelis Wickens, eds R.M. Savory
39
and D.A. Agius, Toronto: Pontifical Institute of Mediaeval Studies, 1984, pp. 219 –
39.
40
–– , Avicenna’s Chapter on Universals in the Isagoge of his Shifa’, in Islam: Past Influence
41 and Present Challenge, eds A.T. Welch and P. Cacia, Edinburgh: Edinburgh
42 University Press, 1979, pp. 34 – 56.
1 –– , Quiddity and Universality in Avicenna, in Neoplatonism and Islamic Thought, ed.

2 P. Morewedge, Albany, New York: SUNY Press, 1992, pp. 77 – 87.
3
G. Pini, Categories and Logic in Duns Scotus: An Interpretation of Aristotle’s ‘Categories’ in
the Thirteenth Century, Leiden: Brill, 2002.
4 –– , Scotus on the Objects of Cognitive Acts, Franciscan Studies, 66, 2008, pp. 281 –
5 315.
6 –– , Scotus on Doing Metaphysics in statu isto, in Duns Scotus, Philosopher, eds
7 M.B. Ingham and O. Bychkov, Archa Verbi, Subsidia 3, Mnster: Aschendorff,
8
2010, pp. 29 – 55.
P. Porro, Possibilit ed esse essentiae in Enrico di Gand, in Henry of Ghent. Proceedings of
9 the International Colloquium on the Occasion of the 700th Anniversary of His Death
10 (1293), ed. W. Vanhamel, Leuven: Leuven University Press, 1996, pp. 211 – 53.
11 –– , Universaux et esse essentiae: Avicenne, Henri de Gand et le ‘Troisime Reich’,
12 Cahiers de philosophie de l’Universit de Caen, 38 – 9, 2002, pp. 9 – 51.
13
–– , Duns Scot et le point de rupture avec Avicenne, in Duns Scot Paris 1302 – 2002,
eds. O. Boulnois et al., Turnhout: Brepols, 2004, pp. 195 – 218.
14 G. Sondag, La rception de la Mtaphysique d’Avicenne par Duns Scot, in Wissen ber
15 Grenzen: Arabisches Wissen und Lateinisches Mittelalter, ed. A. Speer, Berlin/New
16 York: Walter de Gruyter, 2006, pp. 591 – 611.
17 J.B. South, Scotus and the Knowledge of the Singular Revisited, History of Philosophy
18
Quarterly, 19, 2002, pp. 125 – 47.
J.R. Teske, Henry of Ghent’s Debt to Avicenna’s Metaphysics, The Modern Schoolman,
19 85, 2007, pp. 51 – 70.
20 C. Vansteenkiste, Avicenna-Citaten bij S. Thomas, Tijdschrift voor Philosophie, 15, 1953,
21 pp. 457 – 507.
22 J.F. Wippel, The Latin Avicenna as a Source of Thomas Aquinas’s Metaphysics,
23
Freiburger Zeitschrift fr Philosophie und Theologie, 37, 1990, pp. 51 – 90.
–– , The Reality of Nonexisting Possibles According to Thomas Aquinas, Henry of
24 Ghent, and Godfrey of Fontaines, Review of Metaphysics, 34, 1981, pp. 729 – 58,
25 reprinted as Thomas Aquinas, Henry of Ghent and Godfrey of Fontaines on the
26 Reality of Nonexisting Possibles, in J.F. Wippel, Metaphysical Themes in Thomas
27 Aquinas, Washington, DC: Catholic University of America Press, 1984, pp. 163 –
28
89.
–– , The Metaphysical Thought of Thomas Aquinas: From Finite Being to Uncreated Being,
29 Washington, DC: Catholic University of America Press, 2000.
30 R. Wisnovsky, Avicenna’s Metaphysics in Context, Ithaca, New York: Cornell University
31 Press, 2003.
32
33
34
35
36
37
38
39
40
41
42
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
1
2
3
4 Index of Avicenna’s Works with Passages Cited
5
6
˘
7 Risāla l-Adhawiyya fı̄ l-ma ād (Adhawiyya al-Mabda wa-l-ma ād (Provenance and
˘
8 Treatise˙ on
˙ Destination) ˙˙ Destination)
9 p. 197: 15n I, 1, p. 1: 7n
10 pp. 2-3: 31n
11 al-Birr wa l-itm (Piety and Sin) II, 1-8: 13
12 pp. 360-68:¯ 22n III, 16-20: 13
13
14 al-Mašriqiyyūn (The Easterners)
˘
15 al-Hikma al- Arūdiyya (Philosophy for

˙ ˙ p. 8: 7n
˘
16 Arūd ı̄)
˙ v
fol. 2 : 31n
17
18 fols 3v-4r : 31n Mubāhatāt (Discussions)
¯
pp.˙87-8, §§ 150-54: 19
19
20 al-Išārāt wa-l-tanbı̄hāt (Pointers and Re- p. 94, §§ 181-2: 39n
21 minders) (ed. Dunyā) p. 112, § 261: 17n
22 vol. II, § 3.1, pp. 344-5: 106n pp. 140-41, §§ 386-90: 17n
23 vol. II, § 3.16, pp. 404-8: 102 pp. 155-6, §§ 427-8: 19
24 vol. II, § 3.17: 105n p. 226, § 674: 17n
25 vol. II, § 3.19-20: 111n pp. 271-2, § 787: 17n
26 vol. II, § 3.19: 107, 108, 112, 113, 115 pp. 301-2, § 844-6: 17
27 vol. II, § 3.19, pp. 415-20: 107 p. 316, § 888: 19
28 vol. II, § 3.20: 110 p. 318, §§ 892-3: 19
29 vol. II, § 3.20, p. 422: 100 p. 366, § 1141: 17n
30 vol. II, § 3.20, pp. 422-4: 100
31 vol. II, § 3.22, pp. 429–30: 105 al-Naǧāt (The Salvation) (ed. Fakhry)
32 vol. III, § 4.12-15, pp. 23-7: 12n pp. 54-5: 296n
33 vol. III, § 4.28, p. 53: 98, 99n p. 245: 52n
34 vol. III, § 7.13, pp. 275-6: 114 pp. 320-26: 161n
35 vol. III, § 7.14-16: 115
36 vol. III, § 7.14: 115 al-Naǧāt (ed. ‘Amı̄ra)
37 vol. III, § 7.15: 115 vol. II, pp. 153-5: 157n
38 vol. III, § 7.16, pp. 279-80: 115 vol. II, pp. 165-6: 157n
39 vol. III, § 7.17, pp. 281-5: 117
40 vol. IV, pp. 153-9: 19n
41 Šarh Kitāb Utūlūǧiyā (Commentary on the
˙
Theology) ¯
42
43 al-Išārāt wa-l-tanbı̄hāt (ed. Forget) pp. 35-74 (47-9): 14
44 IV: 30 pp. 46-7: 17n
45 IV, 29: 28 pp. 56-7: 17n
46 pp. 138-9: 31n pp. 49-50: 18
47 pp. 140-41: 31n
48 Fı̄ l-Siyāsa l-manziliyya (On Domestic Po-
49 al-Išārāt wa-l-tanbı̄hāt (transl. Inati) litics)
50 pp. 83-4: 295 pp. 232-60 (240): 21n
1 Šifā (The Cure) V, 6, p. 234: 182n
˘ ˘
2 Šifā : al-Mantiq (Logic): al-Madhal (Isago- V, 6, p. 245: 190n
3 ˙
ge) (ed. Anawati) ˘ V, 6, p. 248: 183n
4 I, 1, p. 10: 23n
5 I, 2, p. 15: 28n al-Nafs (Latin, ed. Van Riet)
6 V, 3: 346n
7 al-Madhal (tr. Marmura) V, 3, pp. 105f: 346n
8 ˘
pp. 50f: 345n
9 Šifā : al-Ilāhiyyāt (Metaphysics) (ed.
˘
Šifā : al-Mantiq: al-Maqūlāt (Categories) Marmura)
˘
10
11 I, 4: 8, 10˙ I: 289
12 I, 1-4: 8
13 Šifā : al-Mantiq: al-Safsata (Sophistici I, 2: 10
˘
14 Elenchi) ˙ ˙ I, 2, p. 7: 70n
15 I, 1, p. 7: 46n I, 2, p. 10: 296
16 II, 2, pp. 67-9: 46n I, 3, pp. 14-5: 156
17 I, 4: 289-90
˘
Šifā : al-Tabı̄ ı̄yāt (Natural Philosophy): al- I, 5: 29, 30, 52n, 53n, 366-7, 369, 377,
˘
18
Samā ˙ al-Tabı̄ ı̄ (Physics)
˘
19 381
20 I, 10: 262,˙ 263n, 264 I, 5, p. 22: 381n
21 I, 10, pp. 48-9: 262n I, 5, pp. 23-4: 367n
22 I, 10, p. 49: 259n, 264n I, 5, p. 24: 70, 368n
23 I, 13, pp. 63-6: 9n I, 5, pp. 24-5: 73
24 I, 14: 9 I, 5, pp. 25-9: 71
25 I, 6: 9
˘
Šifā : al-Tabı̄ ı̄yāt: al-Kawn wa-l-Fasād (On I, 6, p. 30: 156

˘
26
27 ˙
Generation and Corruption) (ed. Qas- I, 7: 9, 11
28 sem) I, 8: 8n, 70n
29 VI: 266n I, 8, pp. 38-9: 70
30 XIII, p. 187: 225n II, 1: 8, 256, 258n
31 XIV, p. 190: 225n, 232n, 259n, 266n II, 1, p. 45: 256n
32 II, 1, p. 46: 10n
33 al-Kawn wa-l-Fasād (Latin, ed. Van Riet) II, 1, pp. 46-8: 257n
34 XIV, p. 139: 232n II, 1, p. 48: 256n
35 II, 4: 9, 257, 259, 262
˘
Šifā : al-Tabı̄ ı̄yāt : Fı̄ l-af āl wa-l-infi ālāt II, 4, p. 65: 259n
˘
36
37 ˙
(On Actions and Passions) II, 4, p. 67: 259n
38 p. 256: 225n II, 4, p. 68: 260n
39 III: 302
˘
Šifā : al-Tabı̄ ı̄yāt: al-Nafs (On the Soul) III, 1: 8

˘
40
41 ˙
(ed. Rahman) III, 3: 9, 53
42 I, 1, p. 9: 46n III, 3-5: 302
43 I, 1, p. 16: 106n III, 3, p. 80: 83, 92
44 I, 5, p. 47: 190n III, 3, p. 81: 79, 80
45 II, 2, p. 62: 21n III, 3, pp. 82-3: 79, 80
46 IV, 4: 21n III, 3, pp. 82-4: 87
47 V: 193 III, 3, p. 84: 80
48 V, 2, pp. 209ff: 103n III, 5-6: 53
49 V, 5: 181 III, 5, p. 91: 80
50 V, 6, p. 239: 106n III, 5, pp. 91-3: 83n
Index of Avicenna’s Works with Passages Cited 391
1 III, 7-10: 8 VIII, 7, p. 295: 114n

2 IV: 298 VIII, 7, p. 296: 156
3 IV, 2, pp. 131-2: 156 IX, 1: 11
4 IV, 2, p. 134: 257n IX, 2, pp. 311-12: 156
5 IV, 2, p. 142: 264n IX, 2, p. 314: 156
6 IV, 3: 8, 28 IX, 4, p. 331: 225n
7 V: 289-90, 302, 369 IX, 5: 259, 265
8 V, 1: 52n IX, 5, pp. 334-6: 265n
9 V, 1, p. 149: 53, 74, 369n IX, 5, p. 335: 225n
10 V, 4: 9 IX, 5, p. 337: 225n
11 V, 5: 8 IX, 6: 14, 161
12 V, 6: 8, 9 IX, 6, p. 339: 161n, 162n
13 V, 8-9: 8, 9 IX, 6, p. 340: 162n, 163n, 165n, 166n
14 V, 9: 264n IX, 6, pp. 340-43: 163n
15 V, 9, p. 193: 265n IX, 6, p. 342: 169n
16 VI: 289 IX, 6, p. 343: 166n
17 VI, 1: 261 IX, 6, pp. 343-4: 162n
18 VI, 1, p. 198: 260n IX, 6, p. 345: 171n
19 VI, 1-2: 270 IX, 6, p. 346: 169n
20 VI, 1-4: 8 IX, 6, p. 347: 163n, 172n
21 VI, 2: 259-60, 262, 268, 272 IX, 7: 384n
22 VI, 2, p. 201: 261n IX, 7, pp. 350-51: 157
23 VI, 2, p. 202: 261n, 262n X, 1: 182
24 VI, 2, p. 203: 232n X, 1-2: 13, 19
25 VI, 2, p. 204: 232n, 267n X, 1, p. 359: 182n
26 VI, 3: 263, 270 X, 1, pp. 362-3: 157
27 VI, 3, p. 205: 263n X, 1, p. 363: 182n
28 VI, 3, p. 206: 264n X, 2, pp. 364-5: 157
29 VI, 3, p. 207-8: 263n X, 4-5: 16
30 VI, 3, p. 213: 264n
31 VI, 4, p. 219: 261n al-Ilāhiyyāt (ed. Anawati)
32 VI, 5, p. 220: 11n I, 5, pp. 29-30: 9
33 VII: 289-92, 302-303 I, 5, pp. 31-3: 31n
34 VII, 1: 52n, 290, 292 I, 5, p. 33: 46n
35 VII, 2-3: 8, 292, 302 I, 6, pp. 37-8: 31n
36 VII, 2, p. 243: 290 I, 6, pp. 38-9: 43n
37 VII, 2, pp. 243-4: 157, 291 II, 1, p. 57: 256n
38 VII, 2, p. 247: 293, 297 II, 1, p. 58: 10n
39 VII, 2, pp. 248-9: 294 II, 1, p. 59: 257n
40 VIII-IX: 225 II, 1, p. 60: 256n, 257n
41 VIII, 1: 11, 12 II, 4, pp. 83-5: 259n
42 VIII, 3-4: 11 II, 4, p. 87: 260n
43 VIII, 3, p. 272: 232n IV, 2, p. 175: 257n
44 VIII, 4, pp. 276-7: 298 IV, 2, p. 184: 264n
45 VIII, 5-7: 11 IV, 3, p. 188: 29n
46 VIII, 6-IX, 3: 161n V, 2, p. 211: 23
47 VIII, 6, p. 284: 185n V, 8, p. 247: 24
48 VIII, 6, pp. 284-5: 114n V, 9, p. 252: 265n
49 VIII, 6, p. 288: 157 VI, 1, p. 261: 260n
50 VIII, 7, pp. 291-2: 119n VI, 2, pp. 264-5: 261n
1 VI, 2, p. 266: 262n II, 3: 209n

2 VI, 2, p. 267: 267n II, 4: 209
3 VI, 4, p. 283: 261n II, 4, pp. 96-8: 259n
4 VI, 3, p. 268: 263n II, 4, p. 100: 260n
5 VI, 3, p. 269: 264n III, 1: 209
6 VI, 3, pp. 270-71: 263n III, 1, p. 106: 312n
7 VI, 3, pp. 274-8: 10 III, 2, p. 114: 312n
8 VI, 3, p. 276: 264n III, 3: 355
9 VI, 4, p. 283: 9n III, 3, p. 117: 312n, 343n
10 VI, 5, p. 284: 11 III, 3, pp. 119-22: 312n
11 VI, 5, pp. 288-300: 10n IV, 2, p. 200: 257n
12 VII, 3, p. 318: 24 IV, 2, p. 212: 264n
13 VIII, 4, pp. 343-7: 31n V: 351n, 352n, 354-6, 358
14 VIII, 7, pp. 368-9: 12 V, 1: 311n, 312n
15 IX, 2, pp. 381-4: 12 V, 1-4: 344n
16 IX, 2, p. 383: 13 V, 1, p. 228: 312n
17 IX, 2, p. 386: 12 V, 1, p. 228-9: 369n
18 IX, 2, pp. 389-92: 14 V, 1, pp. 230-32: 312n
19 IX, 2, pp. 392-3: 12 V, 1, p. 232: 311n
20 IX, 3, pp. 402-9: 13 V, 1, pp. 233-6: 313n
21 IX, 5, pp. 413-14: 12 V, 1, pp. 233-7: 315n
22 IX, 5, p. 410: 13 V, 1, pp. 234-6: 311n, 312n
23 IX, 5, pp. 410-11: 265n V, 2: 357-8
24 IX, 5, pp. 411-2: 13 V, 2, pp. 236-7: 297n
25 IX, 5, p. 413: 13 V, 2, p. 240: 343n
26 IX, 6, pp. 415-18: 12 V, 2, p. 243: 23n
27 IX, 7: 12 V, 3: 357
28 X, 1, p. 435: 14 V, 5, p. 276: 345n
29 X, 3: 12, 18 V, 5, p. 277: 24
30 V, 9, pp. 289-90: 265n
31 al-Ilāhiyyāt (Latin, ed. Van Riet) VI, 1, pp. 296-7: 260n
32 I, 1: 209, 212 VI, 2, p. 301-302: 261n
33 I, 1-3: 206 VI, 2, p. 303: 262n
34 I, 2, p. 10: 212n, 213n VI, 2, p. 305: 267n
35 I, 2, p. 13: 296n VI, 3, p. 307: 263n
36 I, 2, pp. 14-17: 211n VI, 3, p. 308: 264n
37 I, 5: 212, 218 VI, 3, p. 310: 263n
38 I, 5, p. 29: 212n VI, 3, pp. 317-8: 264n
39 I, 5, p. 31: 381 VII, 1, p. 349: 312n
40 I, 5, pp. 33-4: 367n VII, 2, pp. 363-4: 293n
41 I, 5, pp. 33-40: 312n VIII, 4, p. 402: 298n
42 I, 5, pp. 34-5: 368n VIII, 7: 209
43 I, 5, pp. 36-40: 313n IX, 2: 209n, 212
44 I, 6: 218 IX, 2, p. 387: 212n
45 I, 6, p. 44: 208n IX, 2–5: 209
46 I, 6-7: 208 IX, 5: 209n
47 II, 1: 212 IX, 5, p. 488: 265n
48 II, 1, p. 65: 256n IX, 5, pp. 489-90: 265n
49 II, 1, p. 67-9: 257n
˘
50 II, 1, pp. 68-9: 256n Ta lı̄qāt (Notes)

Index of Avicenna’s Works with Passages Cited 393
1 pp. 13-22: 17n pp. 69-70: 19

2 p. 23: 18 p. 77: 19
3 p. 37: 18 p. 79: 19
4 pp. 47-8: 19 p. 81-2: 19
5 p. 49: 18 p. 164: 28n
6 p. 52-4: 18 pp. 193-194: 19
7 p. 62-3: 18
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
1
2
3
4 Index of Names
5
6
Abd al-Ğabbār 34n, 40, 44 209-218, 228, 231n, 233, 235-6, 238-
˘
7
8 al-Abharı̄ 47, 123n, 125-34, 136, 142, 9, 241-2, 243n, 244, 251, 253n, 255n,
9 146 256n, 261n, 264n, 268, 271n, 276-81,
10 Abraham bar Hiyya 164, 166 286, 289-92, 295, 300-303, 319, 326-
11 Abraham Ibn Daud˙ 4, 153, 155, 159- 7, 348-9, 358, 371n, 382-5
12 61, 162n, 164-73 Asclepius 280
Abraham Ibn Ezra 164, 192n al-Aš arı̄ 30-31, 33n, 34n, 36, 40, 42-3,
˘
13
14 Abū al-‘Abbās Ahmad ibn ‘Alı̄ al-Isfa- 44n, 45, 127, 131-2
15 hānı̄ 13 ˙ ˙ Augustine 202, 206n, 208, 227-8
Abū Alı̄ al-Ǧubbā ı̄ 34n Avendauth (see also Abraham Ibn
˘
16
17 Abū l-Barakāt al-Baġdādı̄ 38 Daud) 160, 215n
Abū Hāšim al-Ǧubbā ı̄ 30, 35-6, 39-40, Averroes (Ibn Rušd) 3-4, 40, 42n, 47-8,
˘
18
19 44 51-6, 57n, 58-62, 63n, 64-9, 72-88,
20 Abū l-Hudayl 33, 36 90-94, 97, 153n, 191, 193, 200-203,
21 ¯
Abū l-Husayn al-Basrı̄ 42-3, 47, 127 205-206, 211n, 216, 229-31, 233, 235,
˙
Abū Ma šar 205n ˙ 238, 241, 245, 252n, 255n, 256n,
˘
22
23 Adam Wodeham 366 258n, 268, 312n, 344, 355n
Alā al-Dawla Farāmarz 38n Avner of Burgos 155, 179-85, 187-8,
˘
24
25 Albertus Magnus 198-99, 201-202, 190-93
26 227-9, 231, 232n, 238-45, 252n, 342n, Badawı̄ 14, 15n
27 343n Bahmanyār 7-12, 15, 18n, 19n, 23, 42n
28 Alexander of Aphrodisias 64, 66, 216n, al-Bāqillānı̄ 35-6, 39, 40, 44
29 239, 241, 280 al-Baydāwı̄ 47, 123n
30 Alexander of Hales 254n ˙
al-Bayhaqı̄ 7, 38n
31 Alfonso de Valladolid see Avner of al-Bazdawı̄ 33
32 Burgos C. Belo 172n
33 Alfred of Sareshel 216n A. Bertolacci 7n, 11n, 24, 302, 330n
34 M.-T. d’Alverny 200n, 208n, 209-210 al-Bı̄rūnı̄ 181, 184, 186, 189-90, 193
35 Amalricus of Bene 215 Boethius 199n, 202, 205, 206n, 207n,
36 al-Āmidı̄ 40 208, 277, 281-2, 284, 286, 288, 299-
37 Ammonius 280, 295n 300, 339-40, 343, 347, 352-4, 356-60
38 G. C. Anawati 10n, 11n Bonaventure 228, 231, 342n, 347, 359,
39 Anaxagoras 238n, 253 370n
40 Anonymous (d’Alverny) 202, 204, 209, M. Bouyges 57n
41 210n D. J. Brand 262n, 263n
42 Anonymous (de Vaux), author of De causis al-Buhārı̄ 186n
43 primis et secundis 202n, 208, 209n, ˘
Campanella, Tommaso 230
44 226, 239n J. Carrier 288
45 Anonymous (van Steenberghen) 226n Daniel of Morley 202n, 215n
46 Aristotle 1, 12, 42n, 46n, 51-3, 56, 58- David (Greek commentator) 281n
47 9, 60n, 62, 64-6, 70, 72, 74-7, 79-81, David of Dinant 215, 217n
48 82n, 83-6, 87n, 88-91, 94n, 98-100, H. Denifle 216n
49 102, 103n, 107-8, 119-21, 145, 169n, R. Descartes 198
50 178, 197, 200-204, 205n, 206, 207n, R. De Vaux 209n, 215n
396 Index of Names
1 Dexippus 60n Ibn Hazm 39

2 al-Dı̄bāǧı̄ 9n, 12n, 13n, 14n, 16 ˙
Ibn Kammūna 45, 47
3 Dominicus Gundisalvi (Gundissali- Ibn Kullāb 33-4, 36
4 nus) 2, 160, 197, 201-202, 204, Ibn Rušd s. Averroes
5 205n, 206-208, 210-11, 218, 226, 233 Ibn Tufayl 177, 181n, 183n, 185n,
6 M. Dunlop 20n ˙ 187n, 188n, 189n, 190n, 192-3
186n,
7 Eckhart of Hochheim (Meister Eck- Ibn Waqqar 155
8 hart) 55n al-Īǧı̄ 123n
9 Elias (Greek commentator) 281n Īlāqı̄ 39n
10 al-Fārābı̄ 16n, 18n, 20-23, 28, 51-3, 57, Isaac Polgar 179
11 59-62, 64-76, 79, 85-6, 88-92, 94, 108, Īśvarakŗşņa 184n
12 109n, 160, 206n, 207n, 209, 231, 268, A. Ivry 172n
13 288-9, 295, 301, 358 T. Izutsu 27n
14 Ficino, Marsilio 227, 233, 236 W. Jaeger 280
15 A. Fidora 206n, 207n James of Venice 204
16 Francis of Marchia 368n J. Janssens 172n
17 G. Galluzzo 309n John Blund 201-203, 212-3, 216-8
18 Gaudapāda 184n John Buridan 227, 233, 235-6, 238
19 ˙
al-Ġazālı̄ 3-4, 42n, 51, 54, 66-7, 72, John Damascene 347-50, 352, 354-8
20 153, 155, 161, 168-9, 200n, 202, John Duns Scotus 3, 198, 341n, 342n,
21 206n, 207, 210-11, 212n, 213n, 217, 343, 347, 353, 357, 365-6, 375-85
22 226, 236, 347, 360 John of Jandun 235
23 Gilbert of Poitiers 199n
John of La Rochelle 227, 254n, 347,
24 . Gilson 199n, 365
358
25 Gerald of Wales 215n
John of Paris 343, 358
26 Gerard of Cremona 202n
27 Giles of Rome 347, 360 John Pecham 227
28 al-Ǧı̄lı̄ 29-30, 40 John Scot Eriugena 202, 208
29 Godfrey of Fontaines 341n, 342n, 343, J. Jolivet 202n, 206n
30 345n, 347, 353, 356 Kapila 184n
31 A.-M. Goichon 115n al-Kātibı̄ al-Qazwı̄nı̄ see al-Qazwı̄nı̄ al-
32 V. Govind 184n Kātibı̄
33 M. Grabmann 214n, 215n al-Kindı̄ 52, 68-70, 85, 88, 90-92,
34 Gundissalinus see Dominicus Gundi- 103n, 206n
35 salvi C. Lafleur 288
36 D. Gutas 178-80, 188-9, 192n al-Lawkarı̄ 2, 7-16, 17n, 18n, 19-20,
37 al-Ğuwaynı̄ 30, 33-40, 44-5 22-4
38 Hastāmala 191n P. Lee 262n, 263n
39 M. Heidegger 277 G.W. Leibniz 198
40 Henry of Ghent 198, 227, 343, 347-54, A. de Libera 200n, 245n
41 365-6, 368n, 371-5, 377-80 O. Lizzini 10n, 260n
42 Hermann of Carinthia 205n, 208n M. Mahdi 52, 90
43 Hermann the German 216n Maimonides 4, 153-5, 159, 173
44 M. Horten 10n, 11n Mālik-Šāh 39n
45 G.F. Hourani 172n R.P. Mandonnet 215n
46 A. Hyman 258n J.-M. Mandosio 216n
47 Ibn al-Malāhimı̄ 44, 47 R. Marcotte 14n
Ibn al- Arabı̄˙ 48, 133 M. Marmura 10, 11n, 268n
˘
48
49 Ibn Falaquera 154, 156-7 C. Di Martino 216n
50 Ibn Gabirol 202, 206n, 208, 210-11 Mauricius Hispanus 215
Index of Names 397
1 Meister Eckhart see Eckhart of Hoch- Robert of CourÅon 214

2 heim Roger Bacon 201-203, 227, 347, 359
3 W. Mettmann 179 Roger Marston 227, 343, 347, 356
4 Michael of Ephesus 280 Roland of Cremona 202-203
5 Michael Scot 201-202, 204-206, 210- M.-D. Roland-Gosselin 199n
6 11, 216n A. Rucquoi 206n
7 Moses ben Judah 153 Russiliano, Tiberio 227, 230, 233, 237
8 Moses ha-Levi 154 A.I. Sabra 70n
9 Mullā Sadrā 11n, 28 al-Šahrastānı̄ 39-40, 44, 47
10 ˙
al-Nābulusı̄ 55n al-Šahrazūrı̄ 40, 45
11 F.M. Naǧǧār 19-20 al-Samarqandı̄ 44n, 47, 123n, 126-7
12 P. Natorp 280 Śamkara 190n
13 Nifo, Agostino 229, 230, 231n ˙
Scaliger, Julius Caesar 230
14 Olympiodorus 281n A. Schwegler 278-9
15 J.K. Otte 216n Sennert, Daniel 230
16 Parmenides 83, 84n Y. Shiffman 154
17 PataÇjali 184n, 186 A. Shihadeh 41
18 Peter Abelard 205, 342 Simplicius 281n, 347, 360
19 Peter John Olivi 341n, 342n, 347, 360 Socrates 291
20 Peter of Auvergne 342n, 343, 345n, 357 Speusippus 291
21 Peter of Corbeil 213 Spinoza 198
22 Peter of Falco 343, 354 A. Stone 258n, 266-7
23 Peter of Tarantaise 347n, 350n, 359 Suarez, Francisco 228, 229n
24 Philoponus 281n al-Suhrawardı̄ 27-30, 37, 40-41, 44-8,
25 Pietro d’Abano 229 125-31, 133, 141
26 G. Pini 314-5 Syrianus 280-81
27 Plato 84-6, 103, 189, 227, 228n, 230n, R. Szpiech 177, 192n
28 231-2, 235-6, 242n, 245, 254, 268, al-Taftāzānı̄ 123n, 133
29 271, 278, 291, 302 Themistius 231n
30 Plotinus 116, 173 Theophrastus 295n
31 Pomponazzi, Pietro 229, 230n Thomas Aquinas 3, 69, 198-9, 205n,
32 Porphyry 178, 339, 340n, 343, 347, 226n, 227n, 228-9, 232-3, 236, 238,
33 350, 352-60 251-7, 258n, 262, 264, 267n, 268-72,
34 Proclus 91n, 173 275, 277, 282-90, 298-301, 309-29,
35 Pseudo-Alexander of Aphrodisias 280- 330n, 331-4, 339, 342n, 349n, 350n,
36 81 365-6, 368n, 369-71, 374, 384
37 Pseudo-Dionysius 208, 242-3 Thomas Sutton 350n
38 Pseudo-Ptolemy 13 Tignosi, Niccol 229
39 Pythagoras 291, 299 Todros ben Meshullam ben David Tod-
40 al-Qazwı̄nı̄ al-Kātibı̄ 47, 123n, 125, rosi 154
41 128-30, 134, 147 al-Tūsı̄ 97, 98n, 104-9, 110n, 111n,
42 al-Qušayrı̄ 34n ˙
112-4, 115n, 116-21, 129
Qutbaddı̄n al-Šı̄rāzı̄ 45, 47 Umar Hayyām 30, 37-9, 41, 44-7
˘
43
44 ˙
al-Rāzı̄, Fahr al-Dı̄n 2, 29-30, 32, 40- ˘
al-Urmawı̄ 123n
45 47, 97-8,˘ 101, 103-106, 109, 110n, Ustāt 59n
46 111-21, 123-31, 134, 141-2, 155n ¯ Riet 10n, 11n, 23n
S. ˙Van
47 Richard of Middleton 343, 355 M.H. Vicaire 202n
48 Richard of St-Victor 359 Vincent of Beauvais 210, 227
49 Robert Grosseteste 201-203, 348n Vital du Four 227
50 Robert Kilwardby 347, 360 S. Weil 170n
398 Index of Names
1 William Brito 214n Xenocrates 291

William de la Mare 347, 349n, 360
˘
2 Yahya ibn Adı̄ 46n, 111n
3 William of Auvergne 198-9, 201-203, ˙ Zedler 258n
B.H.
4 218, 227, 233-5, 254n, 347, 358 Zimara, Marcantonio 229, 230n
5 William of Ockham 366 F.W. Zimmermann 60n, 263n
6 R. Wisnovsky 52n, 53n
7
8
9 This index of names lists all authors of primary studies. Authors of secondary studies are
10 included only if their views are discussed in some detail.
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50

(Scientia Graeco-Arabica, 7) Dag Nikolaus Hasse, Amos Bertolacci, (Eds.) - The Arabic, Hebrew and Latin Reception of Avicenna's Metaphysics (2011, Walter de Gruyter)

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

(Scientia Graeco-Arabica, 7) Dag Nikolaus Hasse, Amos Bertolacci, (Eds.) - The Arabic, Hebrew and Latin Reception of Avicenna's Metaphysics (2011, Walter de Gruyter)

Uploaded by

Copyright:

Available Formats

The Arabic, Hebrew and Latin Reception of Avicenna’s Metaphysics

Library of Congress Cataloging-in-Publication Data

The Arabic, Hebrew and Latin reception of Avicenna’s Metaphysics /

Bibliographic information published by the Deutsche Nationalbibliothek

쑔 2012 Walter de Gruyter GmbH & Co. KG, Berlin/Boston

1 sixteenth century. Even those thinkers who fundamentally disagreed with

1 occurred through Jewish philosophers reading Avicenna in Arabic, the partial

15 Ta lı̄qāt, but, above all, as the author of a major philosophical encyclopedia,

1 reference to Ilāhiyyāt, II, 4, with respect to chapter 12 (Tahsı̄l, II, 1, 12), or to V,

1 As to the introduction of two chapters of al-Samā‘ al-tabı̄‘ı̄, it once more has an

1 properties of prophecy are presented in an elaborated way, not just briefly

1 congruent with what precedes. They reflect moreover a genuinely Avicennian

1 proper object of this part is more precisely articulated as the establishment of

1 muntaza‘a, i. e. §§ 1 – 61 (except §§ 3, 15, 23 and 40) commonly attributed to

1 incomplete as indicated above, it might be worthwhile to note that it seems to

1 –– , Bayān al-haqq bi-dimāni al-sidq. Mantiq. al-Madhal, ed. I. Dı̄bāğı̄, Tehran:

wuğūduhā hāriğun an māhiyyatihā wa-zā idun alayhā); whereas the quiddity of

which the Ta lı̄qāt were composed, it would be rash to extrapolate a full-fledged

34 aqliyya as ‘intellectual fictions’, on

38 reality (haqı̄qa). On Suhrawardı̄’s role in Mullā Sadrā’s historiography of Islamic

42 5 Ibn Sı̄nā, at-Ta lı̄qāt, IV.32, p. 164,18–ult.

appearance in the Ta lı̄qāt, zā id is never used by Avicenna in this way.

11 Ibn Sı̄nā, al-Mabda wa-l-ma

39 and Māturı̄dite schools) from the two or three generations immediately

Early Mu tazilites, by contrast, tended to collapse all the divine attributes

34 motivated by the fear of violating their cardinal tenet of tawhı̄d, or divine

18 al-Aš arı̄, Maqālāt al-islāmiyyı̄n, p. 165,5.

42 tawhı̄d wa-l- adl, V, p. 233,5.

42 d’al-Ash arı̄ et de ses premiers grands disciples, p. 389.

accidental ( aradı̄) to it; that which is a necessary concomitant (lāzim) of the

(i tibārı̄) and those that

35 (wasf wuğūdı̄), whereas ‘blackness’ (i. e., corresponding to Ǧuwaynı̄’s ma nā) is

Risāla is reiterated in his Darūriyyat at-tadādd fı̄ l- ālam wa-l-ğabr wa-l-baqā .

19 amma l-qismu l-i tibāriyyu d-dātiyyu fa-huwa ka-wasfi s-sawādi bi-annahu

34 ˙ (admittedly,˘ ˘ one who is known for˘embellish-

42 hikma, pp. 110,7 – 111,11. ˙

existence is something conceptual (ma nan i tibāriyyun): Guftgū-yi Šahrastānı̄ va-Īlāqı̄,

al-Musāra a, p. 36,3 – 4 (Arabic) and p. 39 – 40 (English).

1 Avicenna’s ontology by holding that existence is something superadded to a thing’s

Malāhimı̄ (d. 532/1144). These Mu tazilites were openly opposed to Abū

39 ¯ ¯ ānin wa-lā ahwālin); and Tuhfat

annahu ta ālā qādirun

18 merely i tibārāt aqliyya, ‘mental constructs’ or ‘expressions of the intellect’.

al-anwār min matāli al-anzār I.1.2.2 (Fı̄ kawnihi [viz.,

28 Qazwı̄nı̄ al-Kātibı̄, Hikmat al- ayn I.1.1.2

¯ wa-huwa kawnu š-šay ¯i mawğūdan˘ ˙ šay un

37 dālika annahu ya taqidu anna l-anniyyata

–– , al- Aqı̄da an-nizāmiyya,

12 –– , Šarh Uyūn al-hikma, ed. A. as-Saqqā, Tehran, 1994.

Pažūhišı̄-i Hikmat wa Falsafa-i Īrān, 1385 H. ˙

34 Tehran:˙Mu assasa-i Pažūhišı̄-i .Hikmat

–– , Tafsı̄r mā ba da t- tabı̄ a, ed. M. Bouyges, 3 vols, Beirut: Imprimerie Catholique,

37 coextensive but different in meaning would be undermined if šay were more

1 I Averroes against Avicenna on Essence and Existence in the

42 hand knowledge or otherwise.

1 an accident – in reality8 it signifies a second, i. e. a logical, intelligible – he

29 ¯ ˙ Bouyges notes that the

33 a nı̄ ’l-mantiqiyyati, ‘but in reality it signifies only a second, i. e. logical, intelligible’.

1 translators signified, the signification which is said by priority and posteriority of

1 Averroes follows him without reservation – to ‘being as truth’. From Metaphysics

34 thoughts, or a predicate of other predicates. Wa d awwal/tānı̄ translate Greek pq¾tg/

deut´qg h´sir, but ma qūl awwal/tānı̄ and ma nā awwal/t

1 Tahāfut al-Tahāfut parallel, that Avicenna thought that ‘mawǧūd’ in the