ASP 2012 Vol.22 Issue.2

09574239_22-2_09574239_22-2 08/08/12 12:53 PM Page 1
ISSN 0957-4239
VOLUME 22 NUMBER 2 SEPTEMBER 2012
Arabic Sciences
Arabic Sciences and Philosophy

and Philosophy
CONTENTS
Al-Miklātı̄, a Twelfth Century Ašʿarite Reader of Averroes 155

YAMINA ADOUHANE
Regards d’Ibn Rushd sur al-Juwaynı̄. Questions de méthode

MOKDAD ARFA MENSIA
199 Arabic Sciences
Avicenna among Medieval Jews. The Reception of Avicenna’s Philosophical,
Scientific and Medical Writings in Jewish Cultures, East and West 217
and Philosophy
VOL 22 NO 2 SEPTEMBER 2012

GAD FREUDENTHAL AND MAURO ZONTA
A HISTORICAL JOURNAL
pp 155–287
Cambridge Journals Online

For further information about this journal
22:2 SEPTEMBER 2012
please go to the journal web site at:
journals.cambridge.org/asp
Volume 22 Number 2 September 2012
CONTENTS
YAMINA ADOUHANE Al-Miklātī, a Twelfth Century Ašʿarite

Reader of Averroes 155
MOKDAD ARFA MENSIA Regards d’Ibn Rushd sur al-Juwaynī

Questions de méthode 199
GAD FREUDENTHAL and MAURO ZONTA Avicenna among

Medieval Jews. The Reception of Avicenna’s Philosophical,
Arabic Sciences and Philosophy, vol. 22 (2012) pp. 155–197
doi:10.1017/S095742391200001X © 2012 Cambridge University Press
AL-MIKLĀTĪ, A TWELFTH CENTURY AŠʿARITE

READER OF AVERROES
YAMINA ADOUHANE
École transdisciplinaire “lettres/sciences” de l’ENS
45 rue d’Ulm, 75230 Paris Cedex 05
Email: adouhane_yamina@yahoo.fr
Abstract. The aim of this article is to present a new witness of Averroes’ reception in
the Muslim world, in the years that immediately followed his death. Indeed Abū al-
Hağğāğ al-Miklātī (d. 1237) is an Ašʿarite theologian, who was born in Fez. He is
˙ author of a Quintessence of the Intellects in Response to Philosophers on the
the
Science of Principles in which he aims at refuting the Peripatetic philosophers in
their own field, using their own weapons. This article will first attempt to draw the
portrait of this atypical theologian. It will then focus on showing that al-Miklātī –
although he never mentions his name – is a reader of Averroes and in particular, of
his Tahāfut al-Tahāfut, of which he makes various and unexpected uses. A close
look at these uses will enable us to better define the nature of al-Miklātī’s work.
More importantly, this article will try to prove that al-Miklātī provides us with a
key passage of Averroes’ lost treatise On the Prime Mover. At the heart of the
Rushdian criticism of Avicenna’s “metaphysical” proof, this passage should throw
new light on Averroes’ precise understanding of this proof.
Résumé. L’objectif de cet article est de présenter un nouveau témoin de la

réception d’Averroès dans le monde musulman, dans les années qui suivirent
immédiatement sa mort. En effet, Abū al-Hağğāğ al-Miklātī (m. 1237) est un
théologien ašʿarite, né à Fès. Il est l’auteur ˙ d’une Quintessence des intellects en
vue de répondre aux philosophes en matière de science des fondements, où il entre-
prend de réfuter les philosophes – les Péripatéticiens en premier lieu – sur leur ter-
rain et avec leurs propres armes. Cet article cherchera dans un premier temps à
dresser le profil de ce théologien atypique. Il s’attachera ensuite à montrer qu’al-
Miklātī – bien qu’il ne mentionne pas une seule fois son nom – est un lecteur
d’Averroès et en particulier de son Tahāfut al-Tahāfut, dont il fait des usages
variés et inattendus. L’examen de ces usages permettra de mieux définir le statut
de l’œuvre d’al-Miklātī. Enfin, et surtout, cet article tentera de prouver qu’al-
Miklātī nous fournit un passage clé d’un traité perdu d’Averroès: le Traité sur le
Premier Moteur. Au cœur de la critique rushdienne de la preuve “métaphysique”
d’Avicenne, ce passage devrait apporter un éclairage nouveau à la compréhension
qu’Averroès pouvait avoir de cette preuve.
INTRODUCTION
It is a common assumption that, although Averroes has had an impor-

tant influence on Jews and Christians, he had no heir among his
156 YAMINA ADOUHANE
fellow worshippers and that, with Averroes’ death, philosophy itself

disappeared from the Muslim West.1 There have been attempts to
explain such a phenomenon:2 the reasons appear to be both political
and theological, and the particular historical context led to a con-
fusion of these two fields which was prejudicial for philosophy and
philosophers. The truth is that we know little about the theological
environment3 in which Averroes evolved and even less about the
reception his work received in the Muslim world.4 In this context of
1
E. Renan, Averroès et l’averroïsme (Paris, 2002): “Quand Averroès mourut, en 1198, la phi-
losophie arabe perdit en lui son dernier représentant, et le triomphe du Coran sur la libre
pensée fut assuré pour au moins six cents ans.” Such a picture appears today as a mere car-
icature and has been discredited notably by studies that defend the idea of a “second ‘for-
mative’ period” of philosophy in the Islamic world “after Avicenna and beginning with
influential reaction he provoked from al-Ġ azālī”. Cf. the introduction of P. Adamson
(ed.), In the Age of Averroes: Arabic Philosophy in the Sixth/Twelfth Century (London
and Turin, 2011), pp. 167, p. 2. Still, it seems that such studies rarely challenge this con-
ception when it comes to the Muslim West, at the exception perhaps of the Andalusian mys-
tic Ibn ‘Arabī.
2
M. Benchérifa, Ibn Rušd al-hafīd. Sīrat waṯāʾiqiyya (Casablanca, 1999); M. Cruz
Hernandez, Abū-l-Walīd Muhammad ˙ ibn Rušd (Averroes). Vida, obra, pensamiento, influ-
˙
encia (Cordoba, 1997); C. D’Ancona (ed.), Storia della filosofia nell’Islam medievale. Vol. II,
“Averroè” di M. Geoffroy (Torino, 2005); G. Endress, “Le projet d’Averroès”, in Averroes and
the Aristotelian Tradition: Sources, Constitution and Reception of the Philosophy of Ibn
Rushd (1126-1198), Proceedings of the Fourth Symposium Averroicum, Cologne, 1996
(Leiden, 1999); O. Leaman, Averroes and his Philosophy (Oxford, 1988); D. Urvoy, Ibn
Rushd (Averroes) (London and New York, 1991); D. Urvoy (sous la dir. de), La philosophie
andalouse. Auteurs et œuvres, Introduction par D. Urvoy (Casablanca, 2006).
3
The information and anecdotes we have about the theologians that opposed Averroes and
other philosophers does not draw a glorifying portrait of them. The attacks they orche-
strated seem to have been mainly sophistic and opportunist. I am thinking of someone
like Muhammad ibn Zarqūn and Abū ʿĀmir ibn Rabīʿ who was said “[to insult Averroes]
in public˙ and with obscenities” and “[to accuse] him of plagiarism” (cf. J. Puig Montada,
“Materials on Averroes’ circle”, Journal of Near Eastern Studies (JNES), 51 [1992]: 241–
60, pp. 253–5). As M. Benchérifa explains in his biography of Averroes, the motives
which led to Abū ʿĀmir’s attacks were an old and profound rivalry between the two families
and no doubt his lusting for the function of Qādī of Cordoba. Still, his opposition was also
doctrinal and he wrote several works that all seem˙ from their titles to be directed against
Ibn Rušd, the grandson mainly but also the grandfather. However, only the titles of these
refutations have reached us (thanks to one of his disciples, Abū al-Hasan al-Raʿīnī, who
wrote a bibliography of his masters) and so we are left with no information ˙ on the nature
and content of the theologians’ doctrinal attack against Averroes (cf. Benchérifa, Ibn Rušd
al-hafīd, pp. 180–3).
4
“It ˙is generally accepted that the philosophical works of Averroes were not read in the
Islamic world after Averroes’ fall from grace in Córdoba in 1195, and until the early twen-
tieth century, when they were taken up again by Egyptian philosophers. There are, how-
ever, traces of the presence of his works amongst Western Muslims.” (C. Burnett, “The
‘sons of Averroes with the emperor Frederick’ and the transmission of the philosophical
works by Ibn Rushd”, in Averroes and the Aristotelian tradition [Leiden, 1999], pp. 259–
99, p. 275.) Burnett mentions the notorious example of Ibn Ḫ aldūn. One can also think
of Ibn Taymiyya: according to Jon Hoover, “Ibn Taymiyya’s view of God’s perpetual creativ-
ity is remarkably similar to that of Ibn Rushd” (“Perpetual creativity in the perfection of
God: Ibn Taymiyya’s hadith commentary on God’s creation of this world”, Journal of
Islamic Studies, 15.3 (2004): 287–329, p. 295). Elsewhere, he even calls him “the nearest
of the philosophers to Islam” (Mağmūʿ Fatāwā Šayḫ al-Islām Ahmad Ibn Taymiyya, ed.
ʿA. Ibn Qasim and M. Ibn Muhammad, 37 vols. [Cairo, n.d.], vol. ˙ 17, p. 295, quoted by
AL-MIKLĀTĪ, A TWELFTH CENTURY AŠʿARITE 157
scarce information, every new source is precious. In the present
article, I would like to present such a new witness: Yūsuf Ibn
Muhammad Abū al-Hağğāğ al-Miklātī al-Fāsī.
˙
Al-Miklātī, ˙
5 who died in 1237, was probably born in Fez, around
1155. There he received a first-class education in theology and
seems to have also acquired a good knowledge of philosophical texts.
In 11986 he accompanied the Emir al-Mansūr Ibn ʿAbd al-Muʾmin
˙
on a trip to Andalusia. Because of his intelligence and erudition, he
was chosen to defend the Ašʿarite credo against philosophers. This
is precisely what al-Miklātī does in his only conserved work: The
Quintessence of the Intellects in Response to Philosophers on the
Science of Principles (Kitāb Lubāb al-ʿuqūl fī al-radd ʿalā al-falāsifa
fī ʿilm al-usūl). It was edited in Cairo, in 1977, by Fawqiyya Husayn
Mahmūd,7˙ from one unique manuscript kept in the library˙ of the
˙
Qarawiyyīn University in Fez. It seems to have gone unnoticed
among specialists.8 It is however referenced by Daiber9 who refers
to Anawati10 for the enumeration of the chapters. As can be guessed
from the title, the scope of this work is polemical: al-Miklātī intends
to criticize the eternalist cosmology inherited from the Greeks and
adopted by Arab philosophers. As an Ašʿarite theologian, he aims to
prove that the world is adventice,11 that it therefore has a beginning
J. Hoover in p. 295). Burnett then adds: “What has not been sufficiently appreciated, how-
ever, is the apparent survival of interest in his philosophical works for at least one gener-
ation after his death, among scholars associated with the court of al-Nāsir in
Marrakesh.” (The underlining is mine.) In this article, Burnett examines the hypothesis ˙
of the presence of Averroes’ sons at the court of the Emperor Frederick and of their role
in the transmission of their father’s works. He also mentions one of Averroes’ disciples
who held an ambiguous position towards his former master: Ibn Tumlūs. Cf. A.
Elamrani-Jamal, “Éléments nouveaux pour l’étude de l’Introduction à l’art ˙ de la logique
d’Ibn Tumlūs (m. 620 H./1223)”, in A. Hasnawi, A. Elamrani-Jamal and M. Aouad (eds.),
˙
Perspectives arabes et médiévales sur la tradition scientifique et philosophique grecque
(Leuven and Paris, 1997), pp. 465–83. If we admit Elamrani-Jamal’s new datation, Ibn
Tumlūs (c. 1150/5–1223) is the exact contemporary of al-Miklātī (c. 1155–1237).
5 ˙ most likely owes this nickname to his affiliation to the tribe of Miklata. Cf. p. 11 of the
He
long introduction to his work by the editor presented in note 7.
6
The year of Averroes’ death.
7
Professor at the University of ʿAyn Shams, she published the same year in Cairo Ašʿarī’s
al-Ibāna ‘an usūl al-diyāna. Cf. G. C. Anawati, “Textes arabes anciens édités en Egypte
˙
au cours des années 1976–1977–1978”, MIDEO, 14 (1980): 211–62, pp. 214–15.
8
I am immensely grateful to Marwan Rashed for bringing my attention to this work and giv-
ing me the opportunity to study it under his direction.
9
H. Daiber, Bibliography of Islamic Philosophy, vol. 1 (Leiden, 1999). He describes it as “a
dogmatic work, refuting Ibn Sīna, Fārābī and theological schools of Islam, esp. the
Ašʿarites”.
10
Anawati, “Textes arabes anciens”, pp. 219–23.
11
This is how I translate the terms derived from hadaṯa: hādiṯ = adventice, hadaṯ = adventi-
˙
city, hudūṯ/ihdāṯ = advention, muhdaṯ = advented, muh˙ diṯ = adventor. Although
˙ it is not
˙
the most ˙
intuitive ˙ of terms is the one
one, this family ˙ that better enables to translate
the range of Arabic terms. For instance it renders the difference between hādiṯ and
muhdaṯ, and hadaṯ and ihdāṯ, although it fails to distinguish between hudūṯ and˙ ihdāṯ.
˙ ˙ ˙ ˙ ˙
158 YAMINA ADOUHANE
and that God is its creator (this is discussed in the first part of the
book). But his work also treats a number of other topics including
God’s unity, His attributes, whether He knows Himself, if He wants
through an eternal will and speaks through an eternal speech. It
deals as well with questions such as the Prophets, miracles, or punish-
ment and reward.
In this article, I will focus on the first part which goes up to page
173. My aim is to present al-Miklātī and his work. To this end, I
will first draw elements from the text, that will enable me to sketch
some of the aspects of the intellectual personality of our Ašʿarite theo-
logian. Then, in the second part of the article, I will show that he is a
close reader of Averroes’ Tahāfut al-Tahāfut,12 of which he makes var-
ious – sometimes even surprising – uses. This will help us to grasp the
exact nature and aim of al-Miklātī’s work. Finally, the third part of
the article will focus on a different type of issue that will nevertheless
bring to light another aspect in which al-Miklātī can be of interest to
theoreticians: I will try to prove that he is quoting a passage from
Averroes’ lost treatise On the Prime Mover.
I. AL-MIKLĀTĪ: REFUTING THE PHILOSOPHERS IN THEIR

OWN FIELD
After the traditional tributes to God and His prophet, al-Miklātī

evokes the context in which he was asked to write this book: “O unique
scholar, you mentioned to me that philosophical doctrines are, in your
region, an argument extremely widespread, boldly and freely pro-
fessed [mašhūra al-bayʿ wa-al-ibtiyā ʾ], discussed in meetings, and
that their advocates are openly glorified. You asked me to write a
book in response to philosophers, which would contain the healing
of this disease, would be composed according to rational science,
and whose ambition would aspire to the eminence of its Supreme
end”.13 This is indeed what al-Miklātī aimed at in his book
12
However, Averroes is not once named in al-Miklātī’s entire work.
13
Al-Miklātī, The Quintessence of Intellects in Response to Philosophers on the Science of
Principles (Cairo, 1977), pp. 2–3. All the translations are mine. Around 1140, in the
Muslim East, the popularity of philosophical ideas among scholars brought the
Muʿtazilite theologian, Rukn al-Dīn ibn al-Malāhimī, to write a Tuhfat al-mutakallimīn
fī al-radd ʿalā al-falāsifa (The Unique Gift˙ of/for Theologians ˙ in Response to
Philosophers). He was afraid that these Muslim scholars would follow the path of the
Christians whose “leading proponents were inclined towards the Greeks in philosophy,
to the point that they modelled the religion of Jesus upon [the doctrines of] the philoso-
phers” and produced such absurdities as the three hypostases and the incarnation. Cf.
H. Ansari and W. Madelung’s introduction to their edition of the text: Tuhfat
˙
al-mutakallimīn fī al-radd ʿalā al-falāsifa (Tehran, 2008), and G. Schwarb, “Muʿtazilism
in the age of Averroes”, in Adamson (ed.), In the Age of Averroes, pp. 251–82, at pp. 259–61.
The Quintessence of Intellects in Response to Philosophers on the
Science of Principles.14 He then gives an additional precision regard-
ing his target: the philosophers he confronts are Aristotle and his fol-
lowers, the Peripatetics. He did not endeavour to show the fallacies of
the doctrines and argumentations of the Stoics for they are too
obvious. Consequently he concentrated his attacks on the “main phi-
losophers” (ruʾasāʾ al-falāsifa), drew the ultimate consequences of
their concepts and intentions, and revealed the fallacies of their doc-
trines. Al-Miklātī’s goal in this book is to confront the philosophers
on their own ground, using their own weapons.15
1. Establishment of a glossary
To this end, he prefaces the discussion on whether the world has a
beginning or not with an introductory part16 in which he gives the defi-
nitions of several words used by the theoricians. To constitute his glos-
sary, al-Miklātī uses two sources: al-Fārābī’s Book of Letters (Kitāb
al-Hurūf) on the one hand, and Averroes’ Epitome on Metaphysics on
the˙ other. Many questions will arise from the way al-Miklātī uses
these two sources. In order to render these choices more visible, I
will reproduce the lists of the terms defined in each of these three
works:
Al-Fārābī,The Book of Al-Miklātī, Averroes, Epitome on

Letters,17(chap.1) Quintessence, Metaphysics18(chap.1)
(section 3)
That (harf inna) Existent Existent19
When ˙ Thing Entity (huwiyya)
Category Essence Substance
Continued
14
From now on, will be referred to as Quintessence.
15
In the fourth introduction of his Tahāfut al-Falāsifa, al-Ġ azālī writes that, in this book, he
will address the philosophers “in their own language, i.e. with the expressions [they use] in
logic” (Algazel, Tahāfot al-Falāsifat, ed. M. Bouyges, Bibliotheca Arabica Scholasticorum,
Série Arabe, tome II (Beirut, 1927), p. 16, l. 24).
16
There are actually three preliminary parts. The first one aims at determining the relation
of this science – fundamental science – to the other religious sciences. The second one
exposes the subject-matter of this science. In my Phd thesis, I will study more closely
the place ascribed by him to kalam and its consequences as to his conception of the role
of the mutakallim.
17
Al-Fārābī’s Book of Letters (Kitāb al-Hurūf). Commentary on Aristotle’s Metaphysics, ed.
M. Mahdi (Beirut, 1990). ˙
18
When no contrary indication is made, I will be quoting Arnzen’s translation: R. Arnzen,
Averroes on Aristotle’s Metaphysics. An Annotated Translation of the So-called “Epitome”
(Berlin, 2010).
19
Arnzen translates “mawğūd” by “being”.
160 YAMINA ADOUHANE
Continued
Al-Fārābī,The Book of Al-Miklātī, Averroes, Epitome on
Letters,17(chap.1) Quintessence, Metaphysics18(chap.1)
(section 3)
Secondary categories Substance Accident
The first categories Accident Quantity
of arts and sciences One Quality
The names of The same, the Relation
categories opposite,
The forms of terms and the other, difference Essence
their declensions Potency and actuality Thing
Connection (al-nisba) The complete, the One
deficient
Relation (al-idāfa) The whole, the part, The same, the opposite,
˙ the total the other, difference
Relation and Prior, posterior
connection
Connection and Cause and reason On potency and actuality
the number of
categories Matter (al-hayūlā) The complete, the
Accident Form deficient, the whole,
Substance Principle the part, the total
Essence Element (al-ustuquss) Prior, posterior
Existent ˙
Necessity (idtirār) Cause and reason
Thing Nature ˙˙ Matter
For the sake of Form
(allaḏ ī min ağlihi) Principle
From (ʿan) Element
Necessity (idtirār)
Nature ˙˙
Three remarks may be made. First, regarding the choice of the

terms defined, al-Miklātī follows Averroes’ list with very few excep-
tions.20 Second, for the first five terms he defines, which are the
only ones that figure in both al-Fārābī’s and Averroes’ lists,
al-Miklātī chooses to copy al-Fārābī’s Book of Letters rather than
Averroes’ Epitome. Third, the order in which the terms are defined
is much closer to Averroes’ than to al-Fārābī’s, though again some
differences can be observed concerning the first five terms. A more
thorough study of the details of these differences will be pursued
20
“Quantity”, “quality” and “relation” are not defined by al-Miklātī, most likely because they
are subdivisions of the concept of “accident” and defined immediately after the definition of
“accident” in Averroes’ Epitome, whereas here al-Miklātī gives al-Fārābī’s definition of
‘accident’. “Huwiyya” is also dropped, for, in the Epitome, its definition is joined to that
of “existent” and al-Miklātī follows al-Fārābī’s definition of “existent”, not Averroes’.
in my PhD thesis, but I can already raise a few hypotheses to start
drawing a portrait of al-Miklātī and defining the exact nature of his
project.
2. Between al-Fārābī and Averroes: the meaning of

al-Miklātī’s choices
That al-Miklātī’s list of terms is almost identical to Averroes’ may
be accounted for by the fact that he comes only one generation
after Averroes21 and is – as shall be shown in the second part of
this article – a close reader of the Tahāfut al-Tahāfut.22 It does not
therefore seem surprising that the terms of the debate should be
those of Averroes. Yet, when it comes to choosing the definitions of
these terms, al-Fārābī is sometimes preferred to Averroes. What
can explain such a preference? It may be due to al-Fārābī’s perspec-
tive: if al-Miklātī is not interested in the linguistic complexities of
the translation from Greek into Syriac and again into Arabic,23 he
might however agree with the project of deriving the technical sense
of words from their prosaic meaning. For, although al-Miklātī is refut-
ing philosophers, his book does not solely address the philosophers
but is intended for a larger readership which must nevertheless be
of fine intelligence and high education.24 In his Epitome, Averroes
also gives the meaning of the word for the masses but, whereas it is
the starting point of al-Fārābī’s definitions, it appears only seconda-
rily in the Epitome, after the word has been defined in its technical
sense or senses, and Averroes does not seem to attach much impor-
tance to this perspective.25 This does not prevent al-Miklātī from
resorting to Averroes when he finds the latter’s formulation more
satisfying: he does so for the definition of the accident, inserting a sen-
tence from the Epitome in al-Fārābī’s definition.26 Al-Miklātī seems
21
Averroes was born in 1126 and died in 1198 whereas al-Miklātī was born around 1155 and
died in 1237.
22
Averroes, Tahāfot at-Tahāfot, ed. M. Bouyges, Bibliotheca Arabica Scholasticorum, Série
Arabe, vol. 3 (Beirut, 1992). Will be noted “TT” from now on. When no contrary indication
is made, I will be quoting Van Den Bergh’s translation with the Arabic pagination: S. Van
den Bergh, Averroes’ Tahāfut al-Tahāfut (London, 1954).
23
In addition to the numerous homoioteleuta which corrupt al-Miklātī’s text and which my
translation will make sure to indicate and correct, al-Miklātī skips whole paragraphs
from the definitions of the Book of Letters that go into linguistic considerations of this sort.
24
Indeed, before entering the discussion and before the three preliminary sections, al-Miklātī
enunciates four conditions one must fulfil in order to be qualified to read his work. Cf.
pp. 4–6.
25
About the meaning of mawğūd for “the masses”, Averroes writes: “There is no need to take
this [meaning further] into account” (chap. 1, p. 28).
26
Al-Miklātī (in Quintessence, p. 35) reproduces al-Fārābī’s presentation of the common
meaning of the term, then he inserts Averroes’ definition (“As for the theoreticians, ‘acci-
dent’ is predicated – according to them – of that which does not make known the quiddity
162 YAMINA ADOUHANE
therefore to be an active reader who does not merely and passively

copy the definitions that are offered to him but compares various defi-
nitions and chooses the ones he finds more accurate and clear. This is
confirmed by his reorganisation of the order of the terms. If the term
“existent” (mawğūd) is – both in Averroes and in al-Miklātī – the first
term to be explained, it is caught in two completely different concep-
tual complexes. In the first introductory section, al-Miklātī explains
that:
the mutakallim is the one who examines the most general [aʿamm] of things,
i.e. the existent [al-mawğūd]. He divides it into ‘eternal’ and ‘advented’
[muhdaṯ], then examines which acts He must necessarily [do], which are
˙
impossible for Him, and which are possible, and [shows] that the sending
out [baʿṯat] of prophets is part of His possible acts and that this possible
has occurred.27
Whereas in the Epitome the term “existent” is followed by and
thought in connection with the terms “substance” and “accident”, it
is immediately followed in the Quintessence by the terms “thing”
(šayʾ) and “essence” (ḏ āt). We can find an explanation for this at the
beginning of the definitions of each of these terms. Before giving
their meaning for the philosophers, al-Miklātī reminds the reader
of the disagreement there is between the Ašʿarites and the
Muʿtazilites on this matter:
You must know that “thing”28 in the vocabulary of the Ašʿarites among the
theologians is synonymous with “existent”, whereas for the Muʿtazilites it
is not. Indeed the non-existent and the possible [al-ğāʾiz] are, for the
Muʿtazilites, a thing and an essence but not an existent.
With this reorganisation of the order of the terms, we go from an
Aristotelian framework to a theological one. One may also notice
that if this order moves al-Miklātī away from Averroes, it moves
him nearer to Avicenna. In the first book of the Ilāhiyyāt of his
Šifāʾ, Avicenna notoriously devotes the fifth chapter to the study of
“the existent, the thing and their first division” and concludes by say-
ing that it is now clear in which way “‘the thing’ differs from what is
understood by ‘the existent’ and ‘the realised’ [al-hāsil] and that,
˙ ˙‘the existent’]
despite this difference, the two [that is, ‘the thing’ and
of the concrete thing [al-mušār ilayhi] that is not in a substrate. It is of two types: one which
does not make known the essence of a thing, this is [the accident] qua individual and
another which makes known the essence of the individual [accident], this is the [accident]
qua universal”), he then returns to al-Fārābī’s definition.
27
Quintessence, pp. 8–9. The mutakallim is defined by the subject-matter of his science that
is nothing less than the ‘existent’.
28
The same is true of “essence” (al-ḏ āt), p. 19.
are necessary concomitants”.29 It seems therefore that we may have
been a little too prompt in asserting that the terms of the debate
were those of Averroes. If the latter is clearly present in this debate
and most certainly plays a part in it, the exact nature of this part is
yet to be determined.
3. A logical refutation that makes no compromises

By first establishing30 this glossary, al-Miklātī makes sure his refu-
tation will attain its target and not be merely rhetorical. He does
not want the discussion to be about words but about concepts. In
accordance with his intention of confronting the philosophers with
their own weapons, al-Miklātī submits both his argumentation and
his refutation to the strictest logical rigour. Following in the steps of
al-Ġ azālī, who recommended the use of logic to defeat the philoso-
phers, he goes even further by casting all his arguments into the
form of syllogisms, specifying the type of syllogism (hypothetical or
predicative, continuous or not) and the premise on which there is con-
tention (the major, the minor, or the inference). Though, as a theolo-
gian, he does not share many of the philosophers’ premises, he
admits them for the sake of discussion (ğadalan) and refutes them
on a logical level – by pointing out a logical mistake in the argumenta-
tion – or by showing their contradiction with another of the philoso-
phers’ doctrines.31 If he merely rejected the premises, there would
be no possible discussion with the philosophers and his refutation
would not affect their argumentations. Moreover he just as vehe-
mently rejects rhetorical arguments presented by theologians, be
they Ašʿarites. An example can be found in the section where he estab-
lishes the impossibility of an infinite number of adventice things
(hawādiṯ). He starts by saying that “theologians have [written] a lot
˙ this matter” and he intends to “present the fallacies [that corrupt]
on
their proofs and show how to refute them” before giving his demon-
stration of this impossibility.32 The refutation here is aimed at the
Qādī Abū Bakr al-Baqillānī and his followers. He shows that their
˙ are erroneous because based either on a confusion between
proofs
29
Avicenna, The Metaphysics of the Healing. Aš-Šifāʾ: Ilāhiyyāt, ed. M. E. Marmura (Provo,
Utah, 2005), Book I, chapter 5, p. 27, ll. 7–8. When no contrary indication is made, I will
be quoting Marmura’s translation.
30
Though al-Miklātī does not produce the definitions himself but merely collects them, he
can be said to establish the glossary for, as I have tried to show, he is far from being a neu-
tral transmitter: by choosing the terms he wishes to define, the definitions he prefers and
the order of the definitions, he is truly active and, in this preliminary section, what takes
place is already a re-framing of the discussion.
31
Cf. for example p. 133, ll. 14 sq.
32
Quintessence, pp. 86–9.
164 YAMINA ADOUHANE
judgements that apply to the whole and judgements that apply to

the parts33 or on a petitio principii. One last element can be noted,
al-Miklātī appears to have some knowledge – whether directly or
indirectly has yet to be established34 – of Aristotle’s Organon. In
the second introductory section, he explains that every speculative
science includes subject-matters, questions and principles, which he
defines by referring to the Posterior Analytics, and he recalls the
Aristotelian rule that the first principles of a science cannot be demon-
strated in this science.35
The picture we get from these first indications is that of an uncom-
mon theologian, with no equivalent in the Muslim West that we know
of: a highly demanding refuter, who will not be content with a sophis-
tic or rhetorically brilliant refutation, and a sharp logical mind, which
makes him a true disciple of al-Ġ azālī.
II. AL-MIKLĀTĪ, A CLOSE READER OF THE TT
His debt towards al-Ġ azālī36 will be confirmed by the rest of the text.
Indeed his refutation often follows that of al-Ġ azālī’s Tahāfut
al-Falāsifa37: when satisfied by the argumentation, al-Miklātī merely
copies the TF. But what is interesting in these passages where he is
faithful to al-Ġ azālī’s words is not so much the similarities as the
divergences. These differences – as I aim to show in this second
part – are due to al-Miklātī’s reading of the TT.
1. Al-Miklātī’s Quintessence, a Tahāfut “Tahāfut al-Tahāfut”?38

One would indeed expect that a work of the nature of al-Miklātī’s
would only be justified because something (or one should probably
33
He reminds them that “what is true of the whole is not necessarily true of each one [hukm
al-ğumla ġayr hukm al-āhād]”, p. 87, l. 15, and again p. 89, l. 14. ˙
34 ˙
The principles presented ˙ the Posterior Analytics have played an important part in the con-
in
struction of Avicenna’s “metaphysical” proof of the existence of God and in Averroes’ criticism
of this proof. Cf. H. A. Davidson, Proofs for Eternity, Creation and the Existence of God in
Medieval Islamic and Jewish Philosophy, (New York and Oxford, 1987), pp. 284–8,
pp. 312–18. As A. Bertolacci notes in his article “Avicenna and Averroes on the proof of
God’s existence and the subject-matter of metaphysics”, Medioevo. Rivista di storia della
filosofia medievale, 32 (2007): 61–97, p. 62, “the common background of the overall discus-
sion is given by Avicenna’s and Averroes’ attempt of adjusting the epistemological profiles
of metaphysics and physics to the canons of Aristotle’s Posterior Analytics”. Therefore, it is
not unlikely that al-Miklātī’s knowledge of Aristotle’s Organon derives from such a context.
35
Aristotle, Posterior Analytics, I, 9–10.
36
Al-Miklātī was also influenced by other theologians, notably al-Šahrastānī whom he quotes
several times (without necessarily naming him).
37
Will be noted “TF ” from now on.
38
Will be noted “TTT” from now on.
say someone) has happened since the TF that renders it insufficient,
and that would be Averroes’ TT. In this perspective, al-Miklātī’s
Quintessence would appear to be a TTT where Averroes has replaced
Avicenna as the major representative of philosophers and the main
target. Some elements seem in fact to corroborate this expectation.
When copying al-Ġ azālī’s presentation of Avicenna’s arguments
and his objections to them, al-Miklātī inserts objections made by
Averroes and undertakes to refute them. Many examples can be
found but I will limit myself to one. In the chapter pages 126 to 153,
al-Miklātī criticises the philosophers’ conception of the human soul
as “a self-subsistent39 substance, that is not situated [in any
place],40 is neither a body [ğism] nor impressed on a body, is neither
continuous with the body [al-badan] nor separated from it, just as
God the Blessed and Sublime is neither outside the world nor inside
the world”. This corresponds to the eighteenth discussion of the TF
(and the second discussion about the natural sciences in the TT,
pp. 543 sq.). Al-Miklātī is discussing the validity of the following infer-
ence: “everything that inheres in something divisible is necessarily
[ yalzimu an yakūna] divisible”. He refers to “some of the latest
Peripatetic philosophers”, i.e. Averroes, who justifies this inference.
He quotes two paragraphs of the TT41 where Averroes distinguishes
two types of division, and affirms that “everything that is susceptible
to division in [either of] these two kinds of division inheres in a body of
some sort”.42 He then adds that the converse is also evident, i.e. that
“everything that is in a body is susceptible to division in one of these
two kinds of division”. If this is verified, then – he concludes – the con-
verse of its opposite, i.e. “what is not susceptible to division in any of
these two sorts does not inhere in a body”, is also true. Al-Miklātī
admits that, if a premise is true, then the converse of its opposite is
also true. But he denies the validity of the premise for it has not
been demonstrated. Indeed the converse of a universal premise is
true only particularly and not universally so that it is not valid to
admit that, because this premise, i.e. “everything that is susceptible
to division in [either of] these two kinds of division inheres in a
body of some sort”, is evident,43 its converse, i.e. “everything that is
39
I replace “qadīm bi-nafsihi” in our text by “qāʾim bi-nafsihi” as can be read in the TT. I will
have to check the manuscript to determine whether this is a mistake of the editor who mis-
read the word or if it is present as such in the manuscript.
40
In the same way, I replace “la yatağazza’” by “la yatahayyaz”, which is graphically very
close in Arabic. ˙
41
TT, p. 550, l. 12–p. 551, l. 2, and p. 551, l. 12–p. 552, l. 8.
42
The translation is mine.
43
The validity of this premise is admitted by al-Miklātī only for the sake of discussion
(ğadalan).
166 YAMINA ADOUHANE
in a body is susceptible to division in one of these two kinds of div-

ision”, is evident too. This is what al-Miklātī writes:
All theoreticians [al-nuzzār] agree that it is not necessarily implied from a
universal premise that˙ its˙ converse is true universally [laysa yalzimu an
tanʿakisa kulliyya], but it is only implied that its converse is true particu-
larly. For if we say that every man is an animal, it is not necessarily implied
that every animal is a man, but what is necessarily implied is that some ani-
mals are men. The claim in this way [al-mutālaba bi-ḏ ālika] is sophistic.44
˙
Thus, he rejects Averroes’ justification on logical grounds.
Another example of al-Miklātī’s use of Averroes could also lead to
the interpretation mentioned above. At the beginning of the chapter
I have just quoted, al-Miklātī, who is once again quoting al-Ġ azālī,
surreptitiously replaces Avicenna’s doctrine by Averroes’. Al-Ġ azālī
starts by saying that “the discussion of this question demands the
exposition of [the philosophers’] theory about the animal and
human faculties”.45 Al-Miklātī expresses the same demand and copies
al-Ġ azālī’s exposition. Averroes remarks that the doctrine al-Ġ azālī
presents is that of Avicenna “who distinguished himself from the
rest of the philosophers” on a number of points which Averroes enun-
ciates. As we learn from al-Ġ azālī’s account, Avicenna distinguishes
three internal senses: the representative faculty, the estimative fac-
ulty and the imaginative/cogitative faculty.46 But because “one
thing does not retain another through the faculty by which it receives
it”,47 each of the first two faculties has a corresponding faculty that
retains the forms and intentions that they respectively receive,
these are the retentive faculty and the memorative faculty, which
brings the number of internal senses to five. Al-Miklātī appears to
faithfully copy al-Ġ azālī for he literally transcribes the latter’s
account of Avicenna’s doctrine, but while doing so, he introduces
emendations and actually replaces Avicenna’s conception of internal
senses with that of Averroes. Indeed, Averroes rejects Avicenna’s esti-
mative for he considers it superfluous: animals do not need another
faculty in addition to the imaginative, it is the imaginative faculty
in the animal that “determines that the wolf should be an enemy of
the sheep and that the sheep should be a friend of the lamb”.48
Al-Miklātī replaces the term “estimative” in al-Ġ azālī’s account by
“imaginative” and adds this precision: “Avicenna said that the percep-
tion [idrāk] of concord [muwāfaqa] and discord [muḫ ālafa] happens
through another faculty, additional to the imaginative, and he
44
45
TT, p. 543, l. 7.
46
It is called ‘imaginative’ in animals, ‘cogitative’ in men.
47
TT, p. 544, l. 21.
48
TT, p. 547, ll. 7–8.
names it the estimative. But, on this point, he is in disagreement with
all Peripatetic philosophers.”49 Moreover, Averroes considers that the
retentive and the memorative faculties can both be located in the pos-
terior part of the brain for “retaining and memory are two in function
[bi-al-fiʿl], but one in their substratum”.50 Therefore, there are,
according to him, four internal senses: the common sense, the imagi-
native or formative (al-mutasawwira), the cogitative (al-mufakkira)
and the retentive or memorative.˙ Consequently, where al-Ġ azālī
wrote that there were first three then five internal senses according
to philosophers, al-Miklātī writes that there are four. Averroes does
not expose his complete theory of the internal senses and refers to
his De Sensu et sensato for more detail. He does not mention here
the cogitative faculty so that it is not clear if al-Miklātī enumerates
four faculties because he knows the theory of the De Sensu et sensato
or if he distinguishes the retentive and memorative as two distinct
faculties.
But though these two elements seem to corroborate the hypothesis
of a TTT, a closer look at the text will show that this is not the case.
Indeed, the refutation is always limited and selective and neither
Averroes’ doctrine as a whole nor any major doctrinal points are
refuted or even discussed. Moreover, the rectification I have presented
concerns an aspect which is minor from al-Miklātī’s point of view and
represents no major issue for him. Thus, the status of al-Miklātī’s
work still needs to be determined.
2. Averroes: an ally in spite of himself

Let us examine the other ways in which al-Miklātī uses the TT to try
to better understand his relation to Averroes. More than once,
al-Miklātī rectifies al-Ġ azālī’s presentation of Avicenna’s doctrine by
taking into consideration Averroes’ remarks or criticisms. I will
focus on one example.51 In the second discussion about the natural
sciences page 552 of his TT, Averroes criticises al-Ġ azālī’s restitution
of Avicenna: “al-Ġ azālī has not adduced the argument in the manner
in which Avicenna brought it out”.52 He then rectifies it by giving what
he considers a more faithful account of the argument:
Avicenna built his argument only on the following: If the intelligibles inhered
in a body, they would have to be either in an indivisible part of it, or in a divis-
ible part. Then he refuted the possibility of their being in an indivisible part
of the body, and after this refutation he denied that, if the intellect inhered in
49
Quintessence, p. 128.
50
TT, p. 547, ll.5–6.
51
52
TT, p. 552, l. 12.
168 YAMINA ADOUHANE
a body, it could inhere in an indivisible part of it. Then he denied that it could
inhere in a divisible part of it and so he denied that it could inhere in body at
all.53
In his presentation of the philosophers’ proof of the indivisibility of
the human soul, al-Miklātī begins by giving (a) his own formulation
of the philosophers’ proof, then he integrates (b) Averroes’ formulation
of Avicenna’s argument to prove the major premise of his syllogism,
and finally invokes (c) al-Ġ azālī’s reformulation of the argument in
support of Averroes’ restitution. I quote:
The philosophers say [that] (a) the proof that the human soul is indivisible is
[as follows]: The human soul is the place of inherence of knowledge. It is
impossible that the place of inherence of knowledge be a divisible body.
Therefore it is impossible that the human soul be a divisible body. For (b)
if the intelligibles inhered in a body, they would have to be either in an indi-
visible part [of it], or in a divisible part. And it is false according to their prin-
ciples [ʿalā usūlihim] that they are in an indivisible part of the body, for they
˙
reject the doctrine of the separate part [al-ğuzʾ al-fard], and it is [also] false
that they inhere in a divisible [part], because (c)54 if the place of inherence of
knowledge is a divisible body, then the knowledge which inheres in it must be
divisible too; but the knowledge which inheres in it is not divisible, and there-
fore the place of inherence is not a body.
This is representative of the way al-Miklātī proceeds: he rarely
innovates in the arguments he uses in the sense that he takes the
material of his criticism from different sources, but he reorganises
this material, this stock of arguments that are at his disposal, in
such a way that the argumentation he construes is new and original,
and often more effective, for his selective mind leaves no place for
digressions and reduces the argumentations to the bare essential. It
also gives one first example of how the TT almost serves as a commen-
tary of the TF that helps al-Miklātī correct some of its deficiencies.
Al-Miklātī makes another use of Averroes which seems to confirm
this point: when criticising a common opponent, he uses Averroes’
arguments and objections. In the second discussion, in the course of
the refutation of the incorruptibility of the world, al-Ġ azālī first dis-
misses the solutions adopted before him by various groups of theolo-
gians, beginning with the Muʿtazilites. Following al-Ġ azālī,
al-Miklātī presents the philosophers’ argumentation.55 The world,
according to them, cannot be annihilated for, if its annihilation was
possible, it could only happen through an instantiator (muḫ assis)56
˙˙ ˙
53
Id., p. 552, ll. 12–17.
54
It is a litteral quotation of al-Ġ azālī. See TT, p. 548, ll. 5–6.
55
56
Al-Miklātī defines the “muḫ assis” as “that which realises [al-muṯbit] existence rather than
non-existence” (Quintessence,˙ ˙p.˙ 67).
that would be the cause of the passage of the world from existence to
non-existence. This instantiator would have to be either an [agent]
endowed with choice (muḫ tār) or the occurrence of a contrary
(tarayān didd) or the absence of a condition (intifāʾ šart). But, contrary
˙ what theologians
to ˙ claim, it cannot be a choosing˙ agent without
implying a change in the will and consequently in the agent posses-
sing the will. To avoid this inadmissible consequence, the
Muʿtazilites said that the instantiator was in fact the occurrence of
a contrary (tarayān didd): in this case, nothingness (al-fanāʾ). They
believe that˙ God creates
˙ it in no place of inherence ( fī ġayr mahall)
and that it provokes the non-existence of the world at one and˙ the
same moment ( fī dufʿatin wāhidatin). They add that nothingness dis-
appears by itself, thus avoiding ˙ a regression ad infinitum, each noth-
ingness needing another nothingness in order to stop existing. To
criticise this position, al-Miklātī interlocks al-Ġ azālī’s objections
and Averroes’, constructing thereby his own refutation.57 He begins
by saying that what the Muʿtazilites mention is false and weak
because “nothingness and non-existence are synonymous” which is a
strict quotation of Averroes. He then raises a possible counter-
objection of the Muʿtazilites who would say that nothingness is not
non-existence but an existent which is an accident contrary to sub-
stance and which does not persist for two moments. To refute it,
he uses one of al-Ġ azālī’s arguments58 which he reformulates as
follows: “the Sublime Creator either creates [nothingness] in the
world’s essence or creates it in no place of inherence. And if he creates
it in the world’s essence, then it is not a contrary for it co-exists
[li-iğtimāʿihi] with the world”. To reject the second branch of the
alternative, this time he uses one of Averroes’ objections (“the exist-
ence of an accident in no place of inherence is impossible”) and writes:
“if he creates it in no place of inherence, its reality would be trans-
formed [inqalabat] because it belongs to the accident’s reality [min
haqīqat al-ʿarad] to be in need of a place of inherence”. Finally, he
˙
invokes ˙
one last argument ad absurdum which he finds in
̇
al-Gazālī: “It follows from this that God the Sublime does not have
the power to annihilate part of the world without [annihilating] the
whole”. But this is absurd for it contradicts God’s omnipotence.
This use of Averroes, in favour of al-Miklātī’s own goal, making
him almost an ally, is even more spectacular when directed against
57
58
“It is impossible for [nothingness] to be created and to inhere in the world’s essence for then
the inherent meets its place of inherence and co-exists with it if only for an instant, but if
they could co-exist, then [nothingness] would not be the contrary [of the world] and would
not annihilate it.” (TT, II, p. 134, ll. 10–12, the translation is mine).
170 YAMINA ADOUHANE
Avicenna. One example is particularly interesting for it appears

twice in the text and each time at key moments of the argumenta-
tion: first, in pages 67 to 69, in the course of the consolidation of
al-Miklātī’s proof of the adventicity of the world,59 then, in pages
162 and 163, following the proof by which he establishes the exist-
ence of a producer.60 In both cases, Avicenna’s objection rejects the
claim that positing eternity prevents from positing an instantiator
and therefore does not allow philosophers respectively to prove the
adventicity of the world in the first occurrence and to establish a pro-
ducer in the second. This is how – according to al-Ġ azālī’s account –
Avicenna’s argument goes: when the agent advents, what proceeds
from Him and is connected to Him is either pure existence or pure
non-existence or both. It cannot be the precedent non-existence, for
the agent has no influence whatsoever on non-existence. It cannot
either be both. It must then be pure existence. If so, the more perma-
nent the existence, the more permanent the connection, and conse-
quently the more influential and the better the agent. In other
words, and this is indeed what can be read in the Ilāhiyyāt of
Avicenna’s Šifāʾ,61 the conception of an eternal world not only does
not contradict the idea of a producer but is even a better conception
of it, a conception that implies a more powerful God, i.e. a truly
powerful one. Al-Miklātī literally quotes al-Ġ azālī’s presentation of
the Avicennian argument and adds one sentence by Averroes that
synthesises Avicenna’s argument: “advention is nothing more than
the connection of the act to existence, I mean that the act of the
59
Unlike al-Ġ azālī in his TF, al-Miklātī begins by giving a positive proof of the adventicity of
the world before criticising the philosophers’ arguments in favour of eternity. His proof is
as follows: “All existents except God the Sublime are possible [ğāʾiz] in view of themselves;
all possibles in view of themselves are adventice; thus all existents except God the Sublime
are adventice.” (Quintessence, p. 64). He then proves each of the minor and major premises
and the validity of the inference of the consequent from the antecedent by refuting possible
objections.
60
Quintessence, p. 159: “The proof [establishing the producer] is to say: if it is established that
the world is adventice, then it is unavoidably established that it is possible; all possibles
need an instantiator; therefore the world needs an instantiator. Then to say: its instantia-
tor is either endowed with choice or not. [But it cannot be devoid of choice]. It has therefore
been established that the producer of the world is an agent that instantiates it through
power and will, and this is what we wanted to show.” It can be noted that al-Miklātī
deduces the world’s possibility from its adventicity whereas, in the proof of the adventicity
of the world, he deduced its adventicity from its possibility. It enables him to tie together
the proof establishing a producer and the affirmation of the adventicity of the world, as
al-Miklātī’s goal is to show that, contrary to what they claim, philosophers who admit
the eternity of the world cannot establish a producer. This does not invalidate
al-Miklātī’s proof for, as he states in page 168, the affirmation of the eternity of the
world not only prevents the philosophers from establishing a producer of the world but
also from affirming the latter’s possibility.
61
Avicenna, Šifāʾ: Ilāhiyyāt, Book VI, chapter 2, §9, p. 203.
agent is an existentiation [īğād], and in this matter, existence pre-
ceded by non-existence and existence not preceded by non-existence
are equivalent”.62 To refute this argument, al-Miklātī uses Averroes’
objection63 rather than al-Ġ azālī’s. This is what he writes:
What he says is sophistic for the division is not exhaustive [al-taqsīm ġayr
hāsir]. Indeed, advention is connected with an existence combined with
˙ ˙
[bi-wuğūd iqtarana bihi] non-existence, i.e. existence in potentiality
[al-wuğūd bi-al-quwwa]. It is not connected with absolute existence, for
that whose existence is at its ultimate perfection [mā kāna min al-wuğūd
ʿalā kamālihi al-aḫ īr] does not need an existentiation; and it is not [con-
nected] with absolute non-existence, nor with both.64
So, on two major questions (the adventicity of the world and the
establishment of a producer of the world) and at an absolute key
moment of his argumentation (the impossibility of holding together
the belief in eternity and the belief in the existence of an instantia-
tor65), al-Miklātī invokes an objection that he finds in Averroes’ TT.66
Thus the TT appears to be a tool used by al-Miklātī to improve the
TF and Averroes seems to be almost as much an ally as an opponent.
Therefore the Quintessence cannot be considered as a TTT, which
would imply a change in the target, but is rather a sort of update of
the TF, which takes into account some of the objections and new
data supplied by the TT.
62
TT, p. 164, ll. 2–5.
63
TT, pp. 163–4: “[This argument] is sophistic, because Avicenna leaves out one of the factors
which a complete division [al-taqsīm al-hāsir] would have to state. [. . .] This argument is
faulty, because the act of the agent is only˙ ˙ connected with existence in a state of non-
existence, i.e. existence in potentiality [al-wuğūd allaḏ ī bi-al-quwwa], and is not connected
with actual existence, in so far as it is actual, nor with non-existence, in so far as it is non-
existence. It is only connected with imperfect existence in which non-existence inheres. The
act of the agent is not connected with non-existence, because non-existence is not an act,
nor is it connected with existence which is not linked together with non-existence
[al-wuğūd allaḏ ī lā yuqārinuhu ʿadam], for whatever has reached its extreme perfection
of existence [mā kāna min al-wuğūd ʿalā kamālihi al-āḫ ir] needs neither existentiation
nor an existentiator.” I modified the translation slightly.
64
65
The existence of an instantiator is indeed a necessary step in both proofs, which actually
imply a common preliminary syllogism: the world is possible, possibility implies an instan-
tiator, hence the world needs an instantiator. From this need of an instantiator, the first
proof deduces the adventicity of the world – as eternity contradicts the idea of an instan-
tiator –, and the second proof deduces the existence of a producer of the world by showing
that this instantiator is endowed with power and will.
66
One might be surprised that a theologian defending creation ex nihilo may admit, with
Averroes, that adventing is bringing existence in potentiality into actuality (while this cor-
responds to the process of generation – and corruption – which implies a precedent matter).
But – as I will show in my PhD thesis – al-Miklātī is actually using the philosopher’s voca-
bulary and stripping it of its meaning, in a way that is similar to the way Averroes, in the
TT, sometimes feigns to agree with al-Ġ azālī and to share his conception when he is actu-
ally professing the contrary.
172 YAMINA ADOUHANE
3. Averroes does not replace Avicenna as the principal target of the

criticism of philosophers
I would like to bring the reader’s attention to one more passage where
it appears that Avicenna remains the true target of the criticism
against philosophers, and that, although Averroes must also be
ousted, it is only as an additional representative of Peripatetism
and not as the author of a specific philosophical system. On the con-
trary, the criticism directed against Avicenna aims at destroying his
whole doctrine.
In the chapter establishing the existence of a producer of the world
(pp. 154–68), al-Miklātī exposes the views of the philosophers on this
question: all philosophers admit the possibility that an existent pro-
ceeds from something other than God on the condition that it can be
assigned to another existent and this one to another and so on until
it ends in a necessary existent by itself, which is God. On the other
hand, they do not agree on whether it is possible or not that more
than one proceed from Him. Al-Miklātī enumerates the different pos-
itions, but the only one he presents thoroughly is Avicenna’s emana-
tionist conception.67 Moreover, in the course of the refutation,
Avicenna receives the same kind of special treatment: as I have said
in the previous part, al-Miklātī first presents his proof establishing
a producer, then consolidates it by proving separately each of the pre-
mises and inferences. He then concludes that the philosophers,
because they believe in the eternity of the world, cannot establish a
producer and, according to their principles, the world cannot be said
to be the act of God, for “it is a condition of the act that it be preceded
by non-existence and that it brings [a thing] from non-existence to
existence, but this is impossible in the case of the eternal”.68 He
then evokes two ways in which the philosophers have tried to get
around this consequence. He begins with that of “some philosophers”,
which designates Averroes most of the time. Indeed he quotes a pas-
sage from the TT:
If the world were eternal by itself and existent not in so far as it is moved – for
each movement is composed of parts which are adventice [hādiṯa] –, then,
˙
indeed, the world would not have an agent at all. But if the meaning of ‘eter-
nal’ is that it is in perpetual advention [ fī hudūṯ dāʾim] and that this adven-
tion has neither beginning nor end, certainly ˙ the term ‘advention’ [al-ihdāṯ]
is more truly applied to him who brings about a perpetual advention than ˙ to
him who procures a discontinuous advention [al-ihdāṯ al-munqatiʿ]. In this
way, the world is indeed advented by God the Praised ˙ and Sublime, and
the name ‘advention’ is even more suitable [awlā] for it than the word
67
Cf. Quintessence, pp. 155–6.
68
Quintessence, p. 161, ll. 6–9.
‘eternity’, and the philosophers [al-hukamāʾ] only call the world eternal to
safeguard themselves against the word˙ ‘advented’ in the sense of ‘what is
from something, in time and after non-existence’.69
Al-Miklātī quite abruptly rejects this distinction: Averroes, accord-
ing to him, shifts the problem to the question of movement, which is a
merely sophistic move “for the talk is not about the movement that is
composed of parts which are adventice, but is about the celestial
bodies and the separate intellects and the totalities of the four
elements according to their principles, for all these existents have
been characterised by eternity and the negation of a beginning [nafy
al-awwaliyya] according to their principles. It is therefore impossible
according to their principles that they are the act of God the Sublime,
as we have shown.”70 Al-Miklātī accuses Averroes of feigning to
resolve the problem of the contradiction between the affirmation of
the eternity of the world and the establishment of a producer of this
same world by slipping from the question of the existence of the
world to that of the movement of the world, i.e. from the eternity of
its existence to that of its movement.71 Al-Miklātī then exposes
Avicenna’s way of resolving the contradiction72 and this corresponds
to the second occurrence of the argument I have presented above:
advention is nothing more than the relation of the act of the agent
to existence, so that the supposition of a perpetual advention does
not contradict the idea of an agent but rather implies an eternal
and perfect agent. Against this argument, al-Miklātī opposes the
objection we have already examined, which he finds in Averroes’ TT,
and, just as in the case of Averroes, Avicenna’s argument is said of
be sophistic. Al-Miklātī then concludes: “it has been made clear by
this that it is impossible for those who profess eternity to establish
a producer, according to their principle, and this is necessarily implied
for all Peripatetic philosophers who profess eternity”. At this stage it
seems that al-Miklātī has reached his goal, and Averroes and
69
TT, p. 162, I modified the translation slightly. Quintessence, p. 161.
70
71
This is a true problem for Averroes: as I shall show in the last part of this article, Averroes
admits only Aristotle’s physical proof from motion. He will argue that the First Mover and
the First Principle which Avicenna distinguishes are one and the same thing. Therefore the
proof leading to the First Mover is a proof leading to the First Principle. Nevertheless this
First Principle as a First Mover is only proved to be the principle of the motion of the world
not of its existence. According to Wolfson (“Averroes’ lost treatise”, in note 90), Averroes’
answer to this difficulty is to say that, as its formal and final cause, God is the cause of
the unity, order and motion of the world, and because the real existence of the world con-
sists precisely in its unity, order and motion, it can be said that God in this sense is the
cause of its existence. Cf. TT, III, pp. 167–8; p.172; pp. 180–1 + Long Commentary of
Metaphysics, Lām, C44, pp. 1650–2.
72
One may be surprised that he starts by refuting Averroes before criticising Avicenna, but
al-Miklātī is actually following the order of the TT. This suggests that al-Miklātī is dealing
with arguments rather than their authors, at least in the case of Averroes.
174 YAMINA ADOUHANE
Avicenna have received a similar treatment, both dismissed for being

sophistic. But he does not stop here: al-Miklātī adds one paragraph
where he characterises Avicenna’s method and compares it with
Aristotle’s. He then “succinctly” presents the whole of Avicenna’s
argumentation, and criticises it. I will take a closer look at this resti-
tution and criticism in the next part but what I wanted to show here is
the discrepancy there is between the treatment of Avicenna and
Averroes. If some of the latter’s objections are no doubt refuted, this
refutation bears no comparison with the systematic refutation
aimed at the former. The reasons behind such a difference can only
be the object of speculation and so this does not need to concern us
here. But what now appears clearly is that Avicenna remains the
main target.73
There is still a lot to say concerning al-Miklātī’s relation to the TT
and the various uses he makes of it. But this is something I will
treat in the commentary of my translation of this work. In the rest
of the present article, I would like to focus on a different aspect
which shows another way in which this work can be a precious source
of information.
III. AL-MIKLĀTĪ AS A NEW SOURCE OF AVERROES’ WORK
1. Aristotle and Avicenna face to face: the nature and purpose

of this confrontation
As announced in the previous part of this article, I would like to draw
to the reader’s attention one particular passage of al-Miklātī’s work
which corresponds to the end of the first section of the chapter estab-
lishing a producer (page 163 to page 168). To begin with, let us focus
on the first paragraph. This is what al-Miklātī writes:
‫ﻭﻗﺪ ﺍﺳﺘﺪﻝ ﺃﺑﻮ ﻋﻠﻲ ﺑﻦ ﺳﻴﻨﺎ ﻋﻠﻰ ﺇﺛﺒﺎﺕ ﺍﻟﺼﺎﻧﻊ ﲟﻘﺪﻣﺎﺕ ﻋﺎﻣﺔ ﻻﺋﻘﺔ ﺑﺎﻟﻨﻈﺮ‬
‫ ﺇﳕﺎ ﺟﻌﻞ ﻃﺮﻳﻘﻪ ﰲ‬،‫ﰲ ﺍﳌﻮﺟﻮﺩ ﲟﺎ ﻫﻮ ﻣﻮﺟﻮﺩ ﲞﻼﻑ ﻣﺎ ﻓﻌﻞ ﺃﺭﺳﻄﻮ؛ ﻷﻥ ﺃﺭﺳﻄﻮ‬
‫ﺫﻟﻚ ﻣﻦ ﺃﻣﺮ ﺍﳊﺮﻛﺔ ﻭﻟﺬﻟﻚ ﱂ ﻳﺼﺢ ﻟﻪ ﺑﻴﺎﻥ ﻭﺟﻮﺩﻩ ﺇﻻ ﰲ ﺁﺧﺮ ﺍﻟﺜﺎﻣﻨﺔ ﻣﻦ‬
.“‫”ﺍﻟﺴﻤﺎﻉ ﺍﻟﻄﺒﻴﻌﻲ‬
73
This corroborates the idea developed by Peter Adamson in his introduction to In the Age of
Averroes (pp. 5–6), in which he states that, in the period extending from al-Ġ azālī’s death to
the mid-seventh/thirteenth century, Avicenna is taken to be “the main representative of
falsafa, and thus as a target of extensive criticism” and that, during this period, “most
authors think of falsafa as being synonymous with Avicennism, rather than as synonymous
with Aristotle and his commentators”, or in other words, that “in this period, falsafa means
Avicenna”.
‫ﻭﳌﺎ ﻧﻈﺮ ﰲ ﺍﳌﻘﺎﻟﺔ ﺍﳌﺮﺳﻮﻡ ﻋﻠﻴﻬﺎ ﺣﺮﻑ ﺍﻟﻼﻡ ﻣﻦ ﺍﻟﻌﻠﻢ ﺍﻹﳍﻲ ﰲ ﺫﺍﺗﻪ‬
.‫ﻭﺧﻮﺍﺻﻪ؛ ﻓﺴﻠﻢ ﻫﻨﺎﻙ ﻭﺟﻮﺩﻩ ﻣﻦ ﺍﻟﻌﻠﻢ ﺍﻟﻄﺒﻴﻌﻲ‬
.‫ ﲟﻘﺪﻣﺎﺕ ﻋﺎﻣﺔ‬،‫ ﻓﺈﻧﻪ ﺭﺍﻡ ﺑﻴﺎﻥ ﻭﺟﻮﺩﻩ ﰲ ﺍﻟﻌﻠﻢ ﺍﻹﳍﻲ‬،‫ﻭﺃﻣﺎ ﺍﺑﻦ ﺳﻴﻨﺎ‬
Avicenna has proved the establishment of the producer through common
[ʿāmma] premises, suitable [lāʾiqa] in view of existent qua existent, as
opposed to what Aristotle did, because Aristotle found his way in this matter
from the nature of movement [min amr al-haraka] and that is why he was
˙ Book VIII of his Physics. And
able to show its existence only at the end of
when he examined its essence and proprieties in the book named Lām of
his Metaphysics, he admitted there its existence from physics; whereas
Avicenna wanted [rāma] to show its existence in metaphysics through com-
mon premises.74
What exactly is the point of this comparison? This paragraph is

composed of four elements which go in pairs, each pair made up of
an opposition between Avicenna and Aristotle. First, the method is
compared, then the science in which the proof must take place.
Avicenna is said to base his proof on common or general premises.
What this means seems to be made explicit by the end of the sentence:
these premises are “ suitable in view of existent qua existent”. They do
not concern a particular kind of existent, nor existent considered from
one particular aspect – existent qua moved or existent qua quantity –,
but rather concern existent qua existent. In other words, they are
metaphysical premises. At first sight, this may appear as an objective
characterisation of Avicenna’s proof, for he himself writes in his Išārāt
that his proof consists in “examining nothing but existence itself”, by
considering “the state [hāl] of existence”, and that it has “existence
qua existence testify to ˙ the [first cause]”.75 This “metaphysical”
method which deals with existent qua existent and uses general pre-
mises is contrasted with Aristotle’s which, because it is based on the
phenomenon of motion, is physical and concerns existent qua
moved, i.e. the physical existent. This too could be endorsed by
Avicenna as, when thus describing his proof in the Išārāt, he opposes
it precisely to a proof that would be established on the examination,
not of existent qua existent, but of existent insofar as it is created,
in the case of the mutakallimūn, and moved, in the case of the
74
75
Avicenna, Kitāb al-Ishārāt wa-’l-tanbīhāt, ed. J. Forget (Leiden, 1892), p. 146. French
translation (where Forget’s pagination is given): Livre des directives et remarques, ed. A.
Goichon (Beirut and Paris, 1951). Quoted by Davidson, Proofs for Eternity, p. 287.
176 YAMINA ADOUHANE
physicists. Indeed, in both cases, the proof proceeds from the effect to
the cause, which Avicenna considers a less noble path.76 In The
Provenance and Destination,77 he presents the two kinds of proofs,
and in the Appendices,78 he distinguishes between the way the
Physicists establish the First Mover and the way the Theologians
establish it. In the Fair Judgement,79 he even goes so far as to reject
the method from motion as nonsensical and unworthy of its object.
Then the consequences of this divergence in the method are stated
by al-Miklātī: Aristotle could only show the existence of a producer
in physics so that, when he examined it in metaphysics, he did not
prove its existence there but accepted it from physics. Avicenna, on
the contrary, claimed to show its existence in metaphysics. Now, his
opposition to Aristotle on this point is never asserted in such an expli-
cit way. In the first chapter of the Ilāhiyyāt of the Šifāʾ, he shows that
the proof of God’s existence can take place in no other science than
metaphysics, and that what appears in physics aims only at giving
“a fleeting idea” of His existence.80 However, it seems that Avicenna
did not believe that it was Aristotle’s true belief that the proof from
motion was the appropriate proof of the existence of God. According
to Gutas, Avicenna considered this to be “only the apparent sense of
the Metaphysics; its innermost meaning and implicit doctrine was,
in fact, to establish the First Truth by means of the Theological
way”.81 This appears in the Memoirs from a Disciple from Rayy and
in the Discussions.82
From this perspective, the opposition made between Avicenna and
Aristotle looks less like an objective comparison and more like a criti-
cism of Avicenna’s non Aristotelianism. By construing a proof from the
necessary, Avicenna contravenes the Aristotelian method. But
because he is a Peripatetic – and considered so by al-Miklātī83 –,
when he contravenes Aristotle’s rules, he in a way contradicts himself.
If this is so, then the adjective ʿāmma must be understood as having a
pejorative connotation: the premises are not said to be general merely
76
Id., p. 146.
77
Cf. Avicenna, Al-Mabdaʾ wa-al-maʿād, fol. 141r = Nurani 33, quoted by D. Gutas in his
Avicenna and the Aristotelian Tradition. Introduction to Reading Avicenna’s
Philosophical Works (Leiden, 1988), p. 263 (L54).
78
Quoted in Gutas, Avicenna and the Aristotelian Tradition, p. 263 (L55).
79
Quoted in Gutas, Avicenna and the Aristotelian Tradition, p. 264 (L56).
80
Cf. Šifāʾ: Ilāhiyyāt, I, 1, §12, p. 4, and Davidson, Proofs for Eternity, p. 285.
81
Gutas, Avicenna and the Aristotelian Tradition, p. 264.
82
Quoted in Gutas, Avicenna and the Aristotelian Tradition, T13, §8, p. 71 and L80, p. 309. It
is however not unlikely that Avicenna was led to develop such a rhetoric as a defence strat-
egy against attacks similar to the ones addressed to him by al-Miklātī, questioning the
compatibility of his proof with his Aristotelianism.
83
“This is necessarily implied for all Peripatetic philosophers who profess eternity”,
Quintessence, p. 163, l. 7. I underline.
because of the generality of their subject (existence qua existence) but
are said to be common premises as opposed to appropriate premises,
i.e. premises that are proper to their subject. But, in the Posterior
Analytics,84 Aristotle writes that there can be a demonstration only
from principles proper to the subject, not from common ones.
Because his premises are common, Avicenna’s proof of the existence
of God cannot be demonstrative.
If we reinsert this paragraph into al-Miklātī’s general argumenta-
tion – which I have presented at the end of the second part of this
article –, what al-Miklātī seems to be doing is actually playing off
the Peripatetics. Indeed al-Miklātī reproaches Avicenna with the
exact opposite of what he was reproaching Averroes: Averroes was
said to be sophistic because, instead of talking about being and exist-
ence, he talked about motion; Avicenna, on the other hand, deals with
existence, existence qua existence, but by doing so, his proof is said to
be logically deficient because based on common and inappropriate
premises. This however is not a contradiction on al-Miklātī’s part.
For it is precisely al-Miklātī’s aim to show that philosophers –
Peripatetic philosophers – can in no way establish the existence of
the producer as long as they profess the eternity of the world: for
either they remain faithful to their Master and use a proof from
motion, but then, all they can establish is a mover not a producer,
or they build a proof from existence itself, i.e. qua existence, but
then, their premises are common and their proof therefore cannot
be demonstrative.
2. The origin of this criticism

The next question that comes to mind is where al-Miklātī found this
criticism. The characterisation of the proof as a proof from existence
qua existence, the accusation raised against it for not being
Aristotelian, the assertion that the science where the proof must
take place is physics and not metaphysics where it is merely accepted
from physics, all this reminds us of Averroes’ criticism of Avicenna’s
proof. But even if we admit that al-Miklātī’s source is Averroes, we
still need to determine in which work or works he finds it. I have
shown in the second part that al-Miklātī has a direct knowledge of
the TT: what could he read there?
In the fourth discussion, page 276, Averroes writes:
The first man to bring into philosophy the proof [al-burhān] which al-Ġ azālī
gives here as a philosophical one, was Avicenna, who regarded this method
[tarīq] as superior to those given by the ancients, since he claimed it to be
˙
84
Cf. the beginning of I, 9.
178 YAMINA ADOUHANE
based on the substance of the existent [ğawhar al-mawğūd], whereas the

older methods [turuq al-qawm] are based on accidents consequent on the
˙
First Principle.85
Here we have the opposition between a method based on existent
itself and a method based on its accidents. These accidents that are
not named here are motion and time, as appears from another pas-
sage, where the expressions are even closer to those found in
al-Miklātī’s text. Indeed, in the tenth discussion, page 419, Averroes
writes almost exactly the same remarks as in the fourth, only he is
more explicit:
He said that it [the method of a necessary existent] was superior to the proof
of the ancients, because the ancients arrived only at an immaterial being, the
principle of the universe, through derivative things, namely motion and time;
whereas this proof, according to Avicenna, arrives at the assertion of such a
principle as the ancients established, through the investigation of the nature
of the existent in so far as it is an existent [tabī ʿat al-mawğūd bi-mā huwa
mawğūd].86 ˙
Like in our text, two kinds of proofs are opposed, although we can
note two differences: Aristotle is not directly named but designated
by “the ancients” and time is mentioned as well as motion. Finally,
Averroes’ criticism can be found in another passage of the TT, in the
eighth discussion. After having formulated Avicenna’s proof in such
a way that it is “the nearest to demonstrative method [al-tarīqat
˙
al-burhāniyya]”, Averroes concludes, in pages 395–6, that “the demon-
stration [al-burhān] which Avicenna uses in dealing with the necess-
ary existent, when this distinction and this indication are not made, is
of the type of common dialectical statements [min tabīʿat al-aqāwīl
al-ʿāmma al-ğadaliya]; when, however, the distinction ˙ is made, it is
of the type of demonstrative statements [min tabīʿat al-aqāwīl
al-burhāniyya]”.87 Averroes uses the same adjective ˙ as in our text
here, although it is accompanied by the adjective “dialectical” which
leaves no doubt as to the pejorative meaning of ʿāmma in this context,
and more generally characterises Avicenna’s statements (aqāwīl) and
not his premises (muqaddimāt). Though these similarities confirm
the idea that Averroes is al-Miklātī’s source, the differences make it
unlikely that he is quoting the TT. Indeed al-Miklātī – as appears
from the glossary and from his use of the TT – tends to quote literally.
More importantly, there is no trace in the TT of the last two elements
of our paragraph: nowhere is it said that the science where Aristotle’s
proof takes place is physical science and that the metaphysician must
85
I underline and modify the translation slightly.
86
The underlining is mine.
87
The underlining and the translation are mine.
admit it from there, and there are a fortiori no precise references to
“Book VIII” and “Book Lām”, nor is it said that Avicenna on the con-
trary develops his proof in metaphysical science. The examination of
another of Averroes’ works we know was read by al-Miklātī, the
Epitome on Metaphysics, turns out to be just as unfruitful: there is a
criticism of the propositions Avicenna uses but no trace of the adjec-
tive ʿāmma;88 and we do find the idea – absent from the TT – that
the metaphysician accepts the existence of the first principle from
physics, but without the precise references to Physics VIII and
Metaphysics Λ.89 In one passage of the Long Commentary of
Physics, the opposition between Aristotle and Avicenna on this
point is explicit but again the double reference is missing:
For the metaphysician accepts the first moving principles from the physicist,
and he has no way to demonstrate the existence of a first mover unless he
accepts it as something well-known from the physicist. As for the opinion
of Avicenna who thought that the metaphysician ought to demonstrate the
existence of the First Principle, it is false and his method of proof [. . .] is a fee-
ble method and is in no way demonstrative.90
In the same way, similar elements can be found in other passages of
the Long Commentary of Physics91 and in the Long Commentary of the
Posterior Analytics,92 but they are never to be found together in the
same text and the precise references are always absent. In the Long
Commentary of Metaphysics,93 we find the reference to Book VIII of
Physics as the place where physical science shows the existence of
the eternal substance (al-ğawhar al-sarmadī)94 and the affirmation
that “there is no other way to show the existence of the separate sub-
stance than through movement [min qibali al-haraka]”.95 But here, in
˙
88
Chap. I, p. 24, ll. 10–13: “The demonstrations [al-bayānāt] employed by Ibn Sīnā in this
science [of metaphysics] in order to show [the existence] of the first principle are, on the
other hand, altogether dialectical and untrue propositions [aqāwīl], which do not state any-
thing in an appropriate manner”.
89
Chap. I, p. 24, ll. 16–18: “Therefore, he who practises this science [of metaphysics] takes for
granted the existence of the [first mover] from physics, as said before”.
90
Averroes, Commentaria Magna in Libros Physicorum, VIII, C3 (in Aristotelis opera, 10
vols. [Venice, 1573–76]). Quoted by H. A. Wolfson in “Averroes’ lost treatise on the
Prime Mover”, Studies in the History of Philosophy and Religion, 2 vols. (Cambridge,
Massachusetts, 1973), vol. 1, pp. 402–29, p. 413. The translation is his.
91
I, C83 and II, C26 (quoted in Wolfson, “Averroes’ lost treatise”, pp. 410–11 and pp. 412–13).
92
Averroes, Šarh al-Burhān li-Aristū wa-Talḫ īs al-Burhān, ed. ʿA. Badawī (Kuwait, 1984),
C70: “[. . .] the˙ only way by which˙ he [Aristotle]
˙ could demonstrate the existence of the
prime mover was through a sign in that science, namely, physical science, and not as it
was thought by Avicenna. Whence we have composed a special treatise to show the falsity
of the universal [kullī] method whereby Avicenna thought the metaphysician can prove the
existence of a First Principle”. Quoted in Wolfson, “Averroes’ lost treatise”, p. 411.
93
Averroes, Tafsīr mā baʿd at-tabīʿat, ed. M. Bouyges, Bibliotheca Arabica Scholasticorum,
Série Arabe, tome VII (Beirut, ˙ 1990).
94
Lām, C5, p. 1422, ll. 5–7.
95
Lām, C5, p. 1423, l. 11.
180 YAMINA ADOUHANE
C5, he is criticising Alexander’s interpretation and not Avicenna’s,

although he believes that the former’s interpretation is the cause of
the latter’s error. When talking about Avicenna in the following
lines, he writes that according to him “the metaphysician is the one
who demonstrates their existence [i.e. the existence of the principles
of sensible substances, both eternal and non eternal ones]”.96
However I do not believe that this is al-Miklātī’s source for, as I
already said, the confrontation is between Alexander and Aristotle –
or rather Averroes in the name of Aristotle – so that Avicenna is not
directly addressed. Furthermore the fact that it takes place in the
commentary on Book Λ does not seem enough to explain the double
reference in our text. And we do not find a characterisation of
Avicenna’s premises. It is moreover unlikely that al-Miklātī is a
reader of Averroes’ long commentaries.
Among his extant works, the text that gathers together the highest
number of elements from our passage is the treatise On the
Separation of the First Principle.97 Let us quote the beginning of the
treatise:
Qādī (which means Sheriff or Judge) Abū-l-Walīd-ibn-Rušd says that those
who˙ have pursued science in the manner of the Peripatetics, that is,
Aristotle and his followers, knew no other way to demonstrate the existence
of a first principle that is separate and incorporeal except the method estab-
lished by Aristotle at the end of Book VIII of his Physics; and the path along
which they proceeded is that of motion. However, when the books of
Abū-ʿAlī-ibn-Sīnā reached us, we found that he used in these books another
method, one based on the division of being itself [ex respectu divisionis ipsius
entis], and that he believed this way to be superior to that of Aristotle: for
Aristotle’s method proceeds from posterior things and from the nature of
motion, whereas this way (namely, Avicenna’s) proceeds from a consideration
of being as being [entis in quantum ens] to arrive at the first principle; and
therefore this method must be proper to the philosophy that speculates
about being as being. Hence it must be better to follow this method than to
grant the existence of a separate principle on the basis of the authority of
another science, as we find Aristotle doing; for he says that the divine science
accepts the existence of the principles of being from physics, and that it accepts
the number of those principles from a mathematical science, namely, astron-
omy. However, having frequently examined the methods followed by
Ibn-Sīnā, we discovered his mistakes, which we have written about else-
where [in aliis locis].98
96
Lām, C5, p. 1424, l. 2.
97
C. Steel, G. Guldentops, “An unknown treatise of Averroes against the Avicennians on the
first cause. Edition and translation”, Recherches de théologie et philosophie médiévales, 64
(1997): 86–135. The first title given by the translator, “Master Alfonso”, was: “Against some
Avicennians, to prove that the First Necessary Being, i.e., the first principle, namely God,
exists separated from matter, i.e., subsists in virtue of itself ”.
98
Id., p. 97. The underlining is mine.
In this text, we find almost all the elements present in our text with
two exceptions however: the premises are not said to be common, and
there is no specific reference to Book Λ.
After this review of the texts that present similarities with our pas-
sage, there seems to be one last work where al-Miklātī could have found
this criticism, but this work is unfortunately not extant: the lost trea-
tise On the First Mover which Wolfson reconstituted from a refutation
of the treatise by Moses ben Joseph al-Lāwī. This is not the same work
as the precedent one, although they seem to share a same subject, and
even though Wolfson, and Steel and Guldentops use the same refer-
ences,99 as allusions to their respective treatise. It has been said that
Averroes could not have written more than one treatise on the same
subject100 so that either the two treatises are in fact one and the
same or one must not be authentic. Yet, as Steel and Guldentops
note, the two works do not deal with the same questions: “In our
text, for instance, Averroes does not claim that Avicenna’s proof is
not a scientific demonstration, nor does he discuss Avicenna’s assump-
tion that there is a First Cause beyond the Prime Mover. Also, the text
reconstructed by Wolfson is far less cosmological and contains no refer-
ence to the so-called Oriental Philosophy.”101 Moreover it seems to me
that Averroes’ motivations and intentions are completely different
from one treatise to another. From what can be gathered from al-
Lāwī’s refutation, the treatise On the First Mover is clearly directed
against Avicenna’s proof in his claim both that it is metaphysical and
demonstrative and against the conception, which motivates this
“new” proof, that the First Principle is distinct from (and beyond) the
First Mover. Avicenna is the direct target of this work. On the other
hand, the treatise On the Separation of the First Principle is motivated
by Averroes’ encounter with an Avicennian who claimed that
99
Long Commentary of Physics, VIII, C3: “We, moreover, have composed a special treatise
concerning this, and he who would like to learn the difficulties which occur in this method,
let him consult the work of Algazali, for many things which he inveighs against others are
true.”; Long Commentary of Posterior Analytics, C70 quoted in note 91. Cf. Wolfson,
“Averroes’ lost treatise”, p. 413, n. 29; p. 411, n. 25; Steel and Guldentops, “An unknown
treatise of Averroes”, p. 92, n. 18, n. 19.
100
Cf. the translator’s preface in Steel and Guldentops, “An unknown treatise of Averroes”,
p. 95: “it is reasonable that there should be no other such treatise, for another one would
have been superfluous”. A. Elamrani-Jamal (in his article “Ibn Rushd et les Premiers
Analytiques d’Aristote: aperçu sur un problème de syllogistique modale”, ASP, 5 [1995],
pp. 51–74) provides us with a convincing counter-example: on such a technical question
as that of the mode of conclusion of mixed syllogisms, Averroes wrote several short
works to justify the truth and coherence of his Master’s statements. “La conscience de ce
problème conduit Ibn Rushd à lui consacrer plusieurs petits traités ou Opuscules, après
en avoir traité dans son Commentaire moyen des APr.” (p. 62) and A. Elamrani-Jamal con-
cludes his article by saying: “Ses efforts dans ce domaine aride et difficile pourraient nous
servir d’indicateur pour apprécier la rigueur de ses recherches sur d’autres grands
problèmes qui l’ont occupé, relatifs à la psychologie ou à la métaphysique d’Aristote.”(p. 74.)
101
Steel and Guldentops, “An unknown treatise of Averroes”, p. 93.
182 YAMINA ADOUHANE
“Avicenna had never meant by the ‘necessary being’ something

abstracted from the universe or separated from it, but rather he
meant the whole universe” and that he consigned his true doctrine in
his Oriental Philosophy. Averroes does not seem to believe that this
is what Avicenna defended but he does see how his erroneous proof
could lead to such an absurd theory.102 Avicenna’s theory is therefore
criticised, not directly because of what Avicenna himself holds to be
true – indeed Averroes writes that the “mistakes” Avicenna made in
the methods he followed are dealt with “elsewhere” – but because of
the erroneous and blasphemous theory it could lead to. Up until now,
I have proceeded by elimination, but are there any positive elements
that indicate that the lost treatise is al-Miklātī’s source?
The presentation of Averroes’ treatise by al-Lāwī103 and by
Averroes himself in his long commentaries on the Physics104 and
the Posterior Analytics105 indicates that its object is the refutation
of Avicenna’s method and his claim to prove the existence of the
first principle in metaphysics. We also learn from al-Lāwī that:
by all this, Averroes tries to prove that the metaphysician is not able to estab-
lish [the existence] of the Necessary Existent, but must accept it from the
physicist, or the demonstration must be done by using both sciences together,
physics and metaphysics. [Averroes acts thus] only because the Philosopher
had applied this method, I mean that he did not prove the existence [of the
Necessary Existent] in Metaphysics, where he argued, having admitted
that its existence had been proved in Physics.106
Moreover, al-Lāwī writes that:
in the course of that argument Averroes mentioned that the method followed
by Avicenna in the proof of the Necessary Being is similar to the method of
the mutakallimūn, by which he means that the propositions used therein
are common and inappropriate propositions.107
102
Cf. also TT, X, p. 421.
103
Wolfson, “Averroes’ lost treatise”, p. 403: “Averroes has composed a treatise for the purpose
of refuting this method, in which treatise he censures its inventor, namely, Avicenna, and
ridicules the method, and argues in favour of the view that the first mover is identical with
the Necessary Being.”
104
Quoted above in note 99.
105
Quoted above in note 92.
106
I translate from the French translation of the text in G. Vajda, “Un champion de l’avicen-
nisme. Le problème de l’identité de Dieu et du Premier Moteur d’après un opuscule
judéo-arabe inédit du XIIe siècle”, Revue thomiste, III (1948): 480–508, pp. 486–7. Vajda
translated both from the Arabic original, which he found had been quoted by Joseph Ibn
Waqar in one of his treatises (“al-maqāla al-ǧāmiʿa bayn al-falsafa wa-al-šarīʿa”) and
from the Hebrew translation, whereas Wolfson followed the Pamplona manuscript of the
Hebrew translation. Wolfson’s quotation is slightly different and is only partial: “Hence
it necessarily follows that the metaphysician cannot prove the existence of the Necessary
Being but must accept it from the physicist as granted or else he must compose a proof
for it out of the combination of the two sciences.” (p. 409.)
107
Wolfson, “Averroes’ lost treatise”, fragment 4, p. 417. The underlining is mine.
We can assume that these adjectives108 are not a simple gloss of
al-Lāwī but were in the original text. Averroes’ attribution of the ori-
gin of the proof to theologians is absent from al-Miklātī’s text, which is
not surprising as he intends to dismiss Avicenna’s proof.
There is another place where we seem to find all the elements of our
text, and that is the beginning of al-Lāwī’s dissertation. At this point,
he is not presenting Averroes’ treatise but talking in his own name. He
explains the two motives of his study: first the excellency of its object,
second, the divergences amongst philosophers on these questions. To
illustrate these divergences, he presents Aristotle’s proof and method
and compares them to those of later philosophers such as Alexander,
Themistius, al-Fārābī and Avicenna. This presentation, which is too
long for us to reproduce in its entirety here, begins thus: “Indeed,
when, in Book VIII of his Physics, the Philosopher wanted to establish
the existence of the First Mover [. . .]”.109 Follows the development of
Aristotle’s proof which ends in the establishment of a First Mover
that is eternal and is neither a body nor a bodily force. This he estab-
lishes in his Physics and does not go any further. According to al-Lāwī,
this would have been excusable had it been due to the science in
which the examination took place, i.e. physical science. But, al-Lāwī
adds, “however we note that not even in Book Λ of his Metaphysics,
which was the most appropriate place to establish the existence of
the First Cause, did [Aristotle] rise above the First Mover. Indeed,
in this treatise, he considers as given the existence of the First
Mover, demonstrated in Physics”.110 Further on, about Avicenna, he
writes: “When discussing the Necessary Being, i.e. the First Cause,
Avicenna demonstrates, using general premises, drawn from the
nature of being, that it is an existent through itself, not through
another”.111 Almost all the elements of our text are present here:
the general/common premises, the divergence between Aristotle and
Avicenna on the science in which the proof must take place
and most importantly the double reference to Book VIII of Physics
and Book Λ of Metaphysics which was absent from Averroes’ other
works. Now al-Lāwī (second third of the thirteenth century) cannot
be al-Miklātī’s source and it is very unlikely that al-Miklātī is
al-Lāwī’s source, so al-Miklātī and al-Lāwī must have a common
108
Id., p. 418: Wolfson comments on these adjectives by saying that it is “an indirect way of
saying that it is not a true scientific demonstration, for a true scientific demonstration,
according to Aristotle, must be based upon premises which are appropriate (ἀρχαὶ
οἰκεῖαι) and not something common (κοινόν τε)”.
109
Vajda, “Un champion de l’avicennisme”, p. 482. The translation from the French and the
underlining are mine.
110
Id., p. 483. The translation from the French and the underlining are mine.
111
Id., p. 483. The translation from the French and the underlining are mine. G. Vajda writes
in a note that the Arabic text is not certain here. I was not able to have access to the Arabic
manuscript to verify the text nor to see what adjective he translates by ‘general’.
184 YAMINA ADOUHANE
source which we do not dispose of: the treatise in question. Al-Lāwī

holds Averroes in high esteem as appears from the laudatory remarks
he makes of him112 and considers him as a genuine defender of Aristotle
and therefore probably a faithful source on the Philosopher,113 hence it
is not surprising that he would use Averroes’ words to describe
Aristotle’s method. Concerning Avicenna, as I have already noted, the
characterisation of the premises as general can be taken as an objective
description and is only a criticism for those who refuse the metaphysical
nature of the proof.114
Thus, although the treatise is not extant and therefore there can be
no absolute certainty that this treatise is al-Miklātī’s source, the simi-
larities we have noticed strongly support this hypothesis. Moreover it
is not unlikely that al-Miklātī would have been interested in the trea-
tise On the Prime Mover both because of its subject – the proof of the
existence of God – and because of its nature and size.
3. The restitution of Avicenna’s doctrine: content and origin

Let us now examine what directly follows the paragraph where, if I am
right, al-Miklātī – following Averroes’ lost treatise – opposes Avicenna
to Aristotle. It is a restitution of Avicenna’s doctrine, which is then cri-
ticised by al-Miklātī. Al-Lāwī informs us that Averroes started his
treatise by such a restitution: “In that treatise of his Averroes opens
the discussion by reproducing some of Avicenna’s words on this pro-
blem”.115 Could this restitution be the one Averroes makes in the
lost treatise? This is what al-Miklātī writes:
We shall now present, succinctly, what Avicenna says and show how to refute
it, with God the Sublime’s help. Avicenna says: the proof establishing a pro-
ducer is that the conception of every existent that can be conceived as an
existent outside the soul is inevitably one of two things: either it is conceived
112
Id., pp. 483–4: “cet homme, dont nous ne nions pas les grandes qualités et l’autorité en
matière de sciences spéculatives”.
113
Id., p. 483: “Quant à Abū al-Walīd Ibn Rošd, il prend, selon son habitude, la défense de la
doctrine du Philosophe et cherche à réfuter ses adversaires”; “Dans son désir de secourir le
Philosophe [. . .]”.
114
Id., p. 488: “Abū al-Walīd dit dans sa dissertation que la méthode suivie par Abū ʿAlī dans
sa démonstration de l’Être Nécessaire ressemble à celle des adeptes du Kalām, autrement
dit que les prémisses qui y sont employées sont de caractère général et ne sont point perti-
nentes. Mais cette affirmation n’est pas juste. En effet les prémisses générales qu’on
nomme “logiques”, c’est-à-dire celles qu’on emploie dans les disciplines particulières qui
spéculent sur l’être en une chose, [ces prémisses, dis-je] ne sont point pertinentes; elles
sont dès lors inférieures aux prémisses démonstratives. Lorsque cependant elles sont
employées dans les disciplines générales, c’est-à-dire celles qui spéculent sur l’être pris au
sens absolu comme la métaphysique, elles sont pertinentes. Cela est clair pour quiconque
possède la moindre formation logique”. The underlining is mine.
115
Wolfson, “Averroes’ lost treatise”, fragment 1, p. 406.
as an existent through no cause [lā bi-ʿilla] or it is conceived as an existent
through a cause. That which is conceived through no cause is called “a
necessary existent” and that which is conceived through a cause [min qibali
ʿilla] is called “a possible existent”. This latter is then divided, according to
Avicenna, into two parts: if its existence is perpetual, he calls it a necessary
existent through another, [possible] through itself; and if its existence is not
perpetual, it must necessarily be a possible existent through itself and
through another. He said: and it is this latter that must necessarily be pre-
ceded by temporal non-existence and [must]116 be preceded in time by mat-
ter. He then posits [ yadaʿu] that it is a propriety of the possible existent
through itself that it can˙ only exist through another, whether it is, through
this other, possible or necessary. He then posits that it is part of the propri-
eties of the necessary existent through itself to be one and simple, for if it was
composed, it would necessarily be a possible existent through itself, a necess-
ary existent through another, but it was said117 to be a necessary existent
through itself, [and] this is a contradiction that cannot be. He also posits
as part of its proprieties that there cannot exist two of its kind, I mean
that there cannot exist two existents each of which is a necessary existent
through itself, for duality requires a difference [muġāyara] in at least one
meaning, and the fact that they both are necessary existents by themselves
requires that they share the meaning of ‘necessary existent’, so that there
will be in each of them two meanings: one which they share and another
by which [they diverge].118 But what corresponds to this description is a poss-
ible existent through itself, a necessary existent through another. He says: If
this is valid, then we say: each of the four things, I mean the matter
[al-hayūlā], the agent, the form and the end, each one of them, regresses to
a first cause, I mean that the efficient causes regress to a first agent accord-
ing to him, the matter [al-mādda] to a first matter, the end to a first end, the
form to a first form. He then posits that for each of these first causes, except
the agent, it is not possible for them to be a necessary existent through itself,
rather [bal] the first of each is a possible existent through itself, necessary
through another. That is because the first form of the first body – it is as if
you said the body of the first heaven –, as it can only subsist, according to
him, in matter and [conversely] matter exists only combined with the form,
each one of the two – I mean the form and the matter – belongs to the
genus of the possible through itself, necessary through another. It may
well appear to be the same in the case of the end insofar as it is a form of
that which possesses an end and a perfection of it. For it to exist, it is there-
fore necessary that what precedes the end exists. Now if all things are either
composites or causes, and each of these is a possible existent through itself,
although it is necessary through another, and as it has been established that
the possible existent through itself – even if it is necessary through another –
needs a necessary existent through itself in order to exist, or else there will be
a regression ad infinitum, and as this latter [the necessary existent through
116
I have replaced “allā” by “an”.
117
According to the editor, the word cannot be read clearly in the manuscript.
118
I have replaced “yaʿtarifāni” by “yaftariqāni”.
186 YAMINA ADOUHANE
itself] cannot be a composite, nor a form, nor a matter, nor an end, therefore it
can only be an efficient cause. And if all this is so, then the all is originated
[muḫ taraʿ] and created [mubtadaʿ] by the necessary existent, even matter,
and the form of the celestial body, or rather the entire119 celestial body.
This is how Avicenna’s words end.”120
What is al-Miklātī’s source for this account? That it is Averroes and
not Avicenna himself or another intermediary is suggested by some
differences between this account of Avicenna’s doctrine and what he
actually professes in his works, differences which correspond to
Averroes’ “misunderstanding”121 of Avicenna’s proof as it appears in
his works.
One first difference lies in the fact that, according to what al-Miklātī
writes, Avicenna considers the primary distinction of beings to be a
distinction between the uncaused and the caused. The distinction
between ‘necessary’ and ‘possible’ comes second, and this division is
presented as an equivalent, a different nomination, of that primary
division. Two things can be noted: first, the order of the distinctions,
second, the equivalence drawn between them.
Indeed, in most of Avicenna’s works, things are presented the other
way round: first, existents are divided into necessary existents and
possible existents, or – as we shall see shortly – into a necessary exist-
ent through itself and necessary existents through another, and then,
the necessary existent through itself is shown to have no cause,
whereas the possible existent through itself, necessary through
another, has a cause. For instance this is what can be read in his
al-Hikma al-ʿarūdiyya:
˙ ˙
‘The Necessary’ is that which is compulsory of existence [darūrī al-wuğūd] in
˙ principle of exis-
respect to the way it is, this being either in itself (such as the
tents) or through another (such as the fact that two and two are four). [. . .] All
that is necessary of existence in itself has no cause. Whatever has a cause will
be [. . .] possible of existence, while being necessary of existence through its
cause.122
and in the Ilāhiyyāt of his Šifāʾ:
Things which fall into [the category of] existence bear dividing in the
mind into two subdivisions. One is that which, when considered in itself,
its existence does not follow necessarily [. . .] and this thing is found to be
in the domain of possibility. The other is that which, when considered
in itself, its existence follows necessarily. We say that the necessary of
119
I have replaced “bi-amrihi” with “bi-asrihi”.
120
Quintessence, pp. 163–6. For the Arabic text, see below Appendix p. 195.
121
Davidson, Proofs for Eternity, pp. 319, 320, 334–5.
122
Al-Hikma al-ʿarūdiyya, 3v16–4r12, quoted by R. Wisnovsky, in Avicenna’s Metaphysics in
˙
Context ˙
(Ithaca, N.Y., 2003), p. 247. I modified his translation slightly.
existence in itself has no cause, and that the possible of existence in itself has
a cause.123
Moreover, what appears in our text to be obvious, i.e. the equival-
ence between the terms of the two distinctions, is not always pre-
sented in this way by Avicenna. As Wisnovsky remarks, “in some
texts, such as the Mabdaʾ wa-maʿād, Avicenna felt that the link
between being possible of existence and being caused was so intuitive
that he did not need to appeal to an argument for compositeness to
forge that link. In other texts, such as the Hikma ʿarūdiyya and the
˙
Ilāhiyyāt of the Šifāʾ, he felt the need to buttress ˙
that intuition with
an argument for compositeness.” 124
One might argue that the distinction between the uncaused and the
caused is the actual intention and meaning of Avicenna’s division of
beings into necessary and possible. But, even if this were true, the pri-
ority in the presentation is nevertheless paramount, for – as shown by
the references to the mind and to the permissibility in the mind of
making the distinction – the distinction between necessary and poss-
ible is for Avicenna a distinction between primary concepts that “are
imprinted in the soul in a primary way”125 so that the distinction itself
may seem primary to him, whereas this is not true of causality. That
causedness and uncausedness can be the only clear meanings of possi-
bility and necessity is however affirmed by al-Ġ azālī: “The terms
‘possible’ and ‘necessary’ are obscure, unless one understands by
‘necessary’ that which has no cause for its existence and by ‘possible’
that which has a cause for its existence”126 and this is taken up by
Averroes in his criticism of the body of the proof.127 Averroes will
indeed object to the claim of evidence of the distinction between
necessary and possible when understood as a distinction between
what has no cause and what has a cause.128
There is however one text written by Avicenna which apparently
contradicts what I have just said as the first distinction made in it
is that between what has a cause and what has none. This is a
short epistle on the affirmation of God’s unity (al-tawhīd), known as
˙
123
Šifāʾ: Ilāhiyyāt, I, 6, 37, 7–11, text quoted by Wisnovsky, Avicenna’s Metaphysics, p. 256.
The translation is his.
124
Wisnovsky, Avicenna’s Metaphysics, p. 263.
125
Šifāʾ: Ilāhiyyāt, I, 5, pp. 22, 27–28, quoted in Davidson, Proofs for Eternity, p. 289.
126
TT, IV, p. 277, ll. 12–13.
127
TT, IV, p. 276, ll. 10–11: “Avicenna wanted to give a general sense to this statement, and he
gave to the ‘possible’ the meaning of ‘what has a cause’, as al-Ġ azālī relates”.
128
TT, X, p. 417, l. 13–p. 418, l. 1: “We have already said that if by ‘necessary existent’ is under-
stood the causeless and by ‘possible existent’ is understood that which has a cause, the div-
ision of being into these two sections is not acknowledged [. . .]”; TT, IV, p.279, ll. 10–12: “To
divide existents primarily into what is possible and what is not possible is not valid, I mean
that it is not a division that comprehends [tahsuru] existent qua existent” (the translation
is mine). ˙˙
188 YAMINA ADOUHANE
al-Risāla al-ʿaršiyya.129 Its first section establishes a Necessary

Existent. This is how it begins:
You must know that the existent either has a cause of its existence or has no
cause. If it does have a cause, it is a possible [existent] [al-mumkin], whether
before existence when we suppose it in the mind, or when it exists [ fī hālati
al-wuğūd], for the coming into existence (duḫ ūluhu fī al-wuğūd) of the˙ poss-
ible of existence does not remove from it its possibility. If [on the other hand]
it does not have a cause of its existence, in any way whatsoever, it is a necess-
ary existent.130
One may wonder to what extent this epistle questions what has
been deduced from the analysis of Avicenna’s major works. The inver-
sion of the order of the distinctions can perhaps be accounted for by
the nature of the epistle and of the addressee. But this goes beyond
the object of this article. What we need to establish is whether this
text can be al-Miklātī’s direct source. This seems very unlikely, for
the rest of the epistle does not correspond to our text. Moreover,
there is yet another difference between al-Miklātī’s account and
Avicenna’s theory which points less ambiguously towards Averroes.
That it might be one of Averroes’ sources is a completely different
question which again cannot be addressed here.
The second difference concerns Averroes’ understanding of Avicenna’s
division of beings. Indeed Avicenna’s distinction between necessary
and possible when considered independently of concrete existence is
a conceptual distinction between what cannot be thought not to
exist without implying a contradiction and what can be thought to
exist or not to exist without any contradiction following either of the
suppositions.131 When applied to existents outside the soul, the dis-
tinction is made more precise in two different ways: either the necess-
ary is subdivided into necessary through itself and necessary through
another, this latter is then shown to be possible through itself132; or
the precision “through itself” is added and the division is then
between the necessary through itself and the possible through itself,
whereupon it is specified that the possible through itself can exist in
actual fact only when rendered necessary through another, namely
its cause.133 In both cases, the result is a two-fold division of exis-
tents: the Necessary Existent through Himself which is God and the
existents necessary through another, possible through themselves.
The latter group includes all existents except God, i.e. both sublunary
existents submitted to generation and corruption and eternal
129
Avicenna, Rasāʾil al-šayḫ al-raʾīs Abī ʿAlī al-Husayn ibn ʿAbd Allāh ibn Sīnā (Iran, 1980).
130
The translation is mine. ˙
131
Avicenna, al-Nağāt min al-ġaraq fī bahr al-dalālāt (Tehran, 1985), p. 546.
132
Id., p. 546, l. 11; p. 547, l. 12. ˙ ˙
133
Avicenna, Išārāt, p. 141; Šifāʾ: Ilāhiyyāt, I, 6, §2, p. 30.
supralunary existents. Avicenna distinguishes inside this group
between existents that are necessary through another perpetually
(dāʾiman) and those that are necessary through another at one time
and not at another.134 Thus, we end up with a three-fold division of
existents: necessary through itself (God); possible through itself,
necessary through another perpetually (supralunary beings); possible
through itself, necessary through another at one time (sublunary
beings). But the main separation is between God and the rest of exis-
tents. Averroes admits this three-fold division but, as an Aristotelian,
cannot accept to place the main separation between God and the rest,
for the main separation, for Averroes, is between the supralunary and
the sublunary world.135 Averroes has a clear conception of Avicenna’s
division and of its implications but misunderstands or chooses to mis-
understand Avicenna’s formulations. He believes that, when
Avicenna talks of the “necessary through another possible through
itself”, he means the supralunary existents, as opposed to the possible
existents which are not necessary through another: the sublunary
existents.136 In the TT, the three-fold division is not stated in an expli-
cit way137 but it is clear that the necessary through another is under-
stood by him as designating the heavens. In the eighth discussion,
p. 394, he writes: “And the thing moved by this movement [i.e. the
movement which is partly eternal, partly temporal] is what
Avicenna calls ‘the existence necessary through another’, and this
‘necessary through another’ must be a body everlastingly moved”
and a little further, p. 395: “What led Avicenna to this division was
that he believed that the body of the heavens was essentially
134
Šifāʾ: Ilāhiyyāt, Book I, chap.7, §14: “As regards the possible existent, from this its specific
property has become evident – namely, that it necessarily needs some other thing to render
it existing in actuality. Whatever is a possible existent is always, considered in itself, a
possible existent; but it may happen that its existence becomes necessary through another.
This may either occur to it always, or else its necessary existence through another may not be
permanent – occurring, rather, at one time and not another.” The underlining is mine.
135
At first sight, Avicenna and Averroes’ divisions appear to be similar, but both the two-fold
and the three-fold divisions correspond to two diametrically opposed conceptions of the uni-
verse:
Avicenna: Necessary through itself (uncaused) // Possible through itself (caused)
↔ God // Supralunary / Sublunary
Averroes: Necessary (i.e. eternal) // Possible (i.e. contingent)
↔ God / Supralunary // Sublunary
136
Cf. Davidson, Proofs for Eternity, pp. 318–20. In one work, namely his Long Commentary of
Physics (VIII, C79), the division is understood slightly differently by Averroes: the incorpor-
eal movers of the spheres are included in the first group with the first principle, and the
second group contains only the celestial spheres.
137
In the fourth discussion, pp. 276–7, he understands, in accordance with al-Ġ azālī, the dis-
tinction between what is possible and what is not possible to signify the distinction between
what has a cause and what has none, and writes that what has a cause can be divided into
what is possible and what is necessary (darūrī) but it is not clear whether or not Averroes
˙
attributes this subdivision and the expressions “truly possible” and “necessary-possible” to
Avicenna.
190 YAMINA ADOUHANE
necessary through another, possible by itself”. We find the same idea

in the tenth discussion, p. 418:
Avicenna wanted by this division only to conform to the opinion of the philo-
sophers concerning existents, for all philosophers agree that the body of the
heavens is necessary [darūrī] through something else.
˙
There is one text where the division of possible existents into those
that are necessary through another and those that are possible
through another, namely their agent, is stated by Averroes and attrib-
uted to Avicenna:
He [Avicenna] believes that every existent, except God, when it is considered
in itself is possible and contingent and that these contingent beings are of
two kinds: one kind that is contingent in view of its agent, and the other is
necessary in view of its agent, possible in view of itself, and that the necessary
in all respects is the first agent.138
Let us come back to our text. What we have here is exactly this
“erroneous” restitution of Avicenna’s division of beings, exposed
here in the most explicit way:
That which is conceived through no cause is called “a necessary existent” and
that which is conceived through a cause [min qibali ʿilla] is called “a possible
existent”. This latter is then divided, according to Avicenna, into two parts: if
its existence is perpetual, he calls it a necessary existent through another,
[possible] through itself; and if its existence is not perpetual, it must necess-
arily be a possible existent through itself and through another.139
It makes no sense for Avicenna to talk of a concrete existent that is
possible through another. That an existent is possible through itself
means that, when it is considered in itself, independently from its
relation to another, it can be thought either to exist or not to exist.
When considered in relation to another – its cause –, the existent is
either necessary when this other is present, or impossible when it is
absent.
All these elements are significant clues in favour of the hypothesis
that Averroes is al-Miklātī’s source. However, Averroes and al-Miklātī
could also have a common source, either direct – Avicenna himself – or
indirect – an Avicennian or anti-Avicennian intermediary. The ques-
tion of the reception of Avicenna’s works and doctrines in the
Andalusian world is very complex and calls for further study. In the
present state of knowledge on the matter, such an alternative theory
cannot be totally overruled. My intention in what precedes has been to
produce some arguments in favour of the first explanation. If we
138
Averroes, Al-Kašf ʿan manāhiğ al-adilla fī aqāʿid al-milla, ed. M. A. al-Jābirī (Beirut, 1997),
pp. 113–14, §55. The translation and underlining are mine.
139
admit – at least as a hypothesis – that Averroes is indeed al-Miklātī’s
source, we still need to determine if the work al-Miklātī is reading and
copying is the lost treatise.
One first argument in favour of this hypothesis is a negative one:
this restitution can be found nowhere in Averroes’ extant works. We
find nothing of the sort in the TT where Avicenna’s proof is more
alluded to or characterised140 than exposed and developed at length
and for itself. The same can be said of the Epitome on Metaphysics.
Moreover, al-Miklātī introduces his presentation of Avicenna’s doc-
trine with these words: “We shall now present, succinctly [ʿalā īğāz],
what Avicenna says”. And, as quoted above, this is what, according
to Wolfson’s translation from the Hebrew, al-Lāwī writes about
Averroes’ treatise: “In that treatise of his Averroes opens the discus-
sion by reproducing some of Avicenna’s words on this problem”.141
The French translation from the Arabic reads: “Abū al-Walīd com-
mence son opuscule en résumant certains textes d’Abū ʿAlī sur la ques-
tion”.142 I was not able to check the Hebrew and the Arabic behind the
translations “some of Avicenna’s words” and “en résumant” but it is
not unlikely that ʿalā īğāz refers to the same idea. Another sign can
be seen in a remark made by al-Lāwī: “We do not consider that we
can be content here with the usual speculative examination concern-
ing this object of research. It is indeed [commonly] said that if these
two existents have necessary existence in common, and yet differ at
the same time on another point, [the result] is necessarily compo-
sition, which is absurd”.143 For this reason, he gives a different
proof of the impossibility of two necessary existents. This “usual
speculative examination” is very close to the argumentation attribu-
ted to Avicenna by al-Miklātī:
He also posits as part of his proprieties that there cannot exist two of its kind,
I mean that there cannot exist two existents each of which is a necessary
existent through itself, for duality requires a difference [muġāyara] in at
least one meaning, and the fact that they are both necessary existents through
themselves requires that they share the meaning of necessary existent, so that
there will be in each of them two meanings: one which they share and another
by which [they diverge]. But what corresponds to this description is a poss-
ible existent by itself, a necessary existent by another.144
140
Averroes evokes Avicenna’s conception of his own proof: “superior to those given by the
Ancients”, “based on the essence of the existent” (IV, p. 276; same idea in X, p. 419); he
exposes briefly some aspects of Avicenna’s argumentation that he criticises – such as the
idea that the existent necessary through another is in itself a possible existent – (VIII,
p. 395) and gives his interpretation of the motives that led Avicenna to conceive such a
proof (VIII, p. 395; X, p. 418).
141
Wolfson, “Averroes’ lost treatise”, fragment 1, p. 406. The underlining is mine.
142
Vajda, “Un champion de l’avicennisme”, p. 486. The underlining is mine.
143
Idem, p. 492. The translation from the French is mine.
144
192 YAMINA ADOUHANE
Al-Lāwī’s remark might be an allusion to Averroes’ restitution of

Avicenna’s proof. But this raises the following question: where does
Averroes find such a formulation of the proof? For the proof of the
uniqueness of God is not expressed in such a concise way in
Avicenna’s works.145 The closest to our text is its formulation in the
Nağāt, although the reference to discourse does not appear in our text:
[. . .] If they have one thing in common, and if each of them has in addition to
[this thing] its own distinct meaning by which its quiddity is completed and
which is part of it [its quiddity], then each one of the two is divisible in dis-
course [bi-al-qawl]. But it has been said that the necessary existent is not
divisible in discourse, thus not even one of the two is a necessary existent.146
Other parts of our text also seem to find no strict equivalent in
Avicenna’s works though they do not contradict his doctrine.147 This
seems to indicate either that Averroes possessed a text that has
since been lost, or, on the contrary, that Averroes did not have the
whole of Avicenna’s works and therefore attempted to reconstruct a
coherent argumentation from the different – possibly incomplete –
texts or accounts he had of Avicenna’s proof, supplying it with argu-
ments when it seemed to lack consistency.148 What al-Lāwī says –
from what can be read in the French translation – gives support to
the second hypothesis: Averroes, according to him, summarized
Avicenna’s doctrine from various texts. Averroes might have wanted
to give a unified and purified version of Avicenna’s proof, regardless
of the divergences he might have noticed between his different
works, so as to enable him to make a univocal and consistent criticism.
This criticism can also offer an additional sign in favour of the
identification of this text with Averroes’ restitution in the treatise
On the Prime Mover, for as al-Lāwī notes: this treatise is made up of
two parts, the first one is devoted to the criticism of the proof, the
second one to the refutation of Avicenna’s motive for this special
proof: his belief in the separation between the First Mover and the
First Principle. To achieve this refutation, Averroes will show that
the first principle – which he identifies with the first mover – is not
an efficient cause but a formal and final cause: “In his endeavor to con-
tradict this view and to prove that the First Principle is identical with
the First Mover, Averroes starts out by showing that the First
Principle is not the efficient cause of anything inferior to it but is its
cause only by way of form”.149 In order to criticise this precise point
one would first need to present Avicenna’s doctrine which states
145
Cf. for instance Šifāʾ: Ilāhiyyāt, Book I, chap.7, §§12–13.
146
Avicenna, al-Nağāt, p. 559.
147
This is the case of the development on the four causes.
148
Cf. Davidson, Proofs for Eternity, p. 334.
149
Wolfson, “Averroes’ lost treatise”, fragment 6, p. 421.
that the Necessary Existent is an efficient cause, as is the case in our
text. Al-Miklātī on the other hand does not deny the causal efficiency
of God – although, according to him, God is a voluntary and not a
necessary agent. For the purpose of his own refutation, al-Miklātī
could have presented Avicenna’s proof independently from the discus-
sion of the causes. This point is not crucial in al-Miklātī’s criticism,
and he will simply deny that Avicenna is able to demonstrate such
a point: “his assertion that the Necessary Existent through Himself
cannot be a matter nor a form nor an end is false [kalām bātil]
because, to deny each of these [eventualities], [Avicenna] confined ˙
himself to allegations devoid of demonstration”. 150
About al-Miklātī’s criticism, one final remark can be made. Al-Lāwī

informs us that Averroes addresses two kinds of criticisms: a material
one against the premisses of Avicenna’s proof and a formal one which
shows that, even if the premisses were admitted, the proof would not
attain its goal for it is not demonstrative.151 In a similar way,
al-Miklātī criticises both the content and the form of Avicenna’s
proof. First he rejects Avicenna’s premises: he denies the validity of
the division of possible existents into existents that are possible
through themselves, necessary through another, and existents that
are possible both through themselves and through another. Indeed,
this division enables Avicenna to reserve temporal adventicity to
the second group of possible existents, the first group being eternal
and therefore perpetually advented, whereas al-Miklātī believes he
has demonstrated that all possible existents are adventice, which
means – for the theologian that he is – that all possible existents
come to existence from non-existence and in time. Then he restates
some of Avicenna’s affirmations and denies that he has produced a
demonstration of them. Finally, in conclusion, he repeats what he
had already concluded a few pages earlier (p.163): that even if, for
the sake of discussion (ğadalan), he admitted that Avicenna had suc-
ceeded in establishing an agent necessarily existent through itself,
this would still be incompatible with the philosopher’s affirmation of
the eternity of the world and of some of the existents that constitute
it. Not surprisingly, Averroes and al-Miklātī’s material criticisms
have nothing in common for their aims are opposed. It is nevertheless
interesting to note that Averroes and al-Miklātī are both encumbered
by the group of existents that are said to be “possible through them-
selves, necessary through another” and which correspond – according
to Averroes’ interpretation – to the intermediary group of the eternal
caused existents. But, whereas Averroes rejects the possible nature
150
151
Wolfson, “Averroes’ lost treatise”, p. 405.
194 YAMINA ADOUHANE
of these necessary existents,152 al-Miklātī refuses the necessity – in the

sense of eternity – of these possible existents, and shows that as long
as they are held to be so, it cannot be proved that God is their producer.
All this put together seems to provide sufficient evidence to justify if
not the certainty, at least the high probability of the hypothesis that
Averroes’ treatise On the Prime Mover is the source from which
al-Miklātī derives this passage.
CONCLUSION
In this article I intended to present al-Miklātī and his work, and I

hope that this first and partial presentation is sufficient to convince
the reader of the importance and interest of a further study. I have
tried to give a first glimpse of the variety and complexity of his
relations to Averroes by showing some of the numerous uses he
makes of the latter’s works, and in particular of the Tahāfut
al-Tahāfut. No direct conclusions can be made from the omnipresence
of Averroes who is however not named even once. But it does give
some clues as to the reception his work received in the years that
immediately followed his death, and offers some keys – if not to how
we must understand it – at least to how we must not understand it.
In a different way, al-Miklātī gives us direct access to Averroes by
revealing parts of a treatise now lost. If my hypothesis is true,
al-Miklātī provides us with an important passage of the treatise On
the Prime Mover, which is crucial to our understanding of Averroes’
comprehension of Avicenna’s proof and might help to establish exactly
what the Andalusian philosopher knew of his Persian predecessor’s
doctrine and which texts or parts of texts he must have had at his dis-
posal. All this remains a desideratum.
I also hope to have shown that all of al-Miklātī’s interest does not lie
in what he can teach us about Averroes, but that he is worth studying
in himself. By the nature and aim of his work, as well as by the
method he adopts, al-Miklātī appears to be an heir worthy of
al-Ġ azālī. His mastery of the art of logic enables him to confront phi-
losophers as an equal.
If I have shown some aspects of our theologian’s relation to philoso-
phers, I have not yet tackled his relation to the other mutakallimūn.
Indeed, al-Miklātī’s position on various doctrinal points within
Ašʿarism remains to be examined. Whether his long and close associ-
ation with philosophical texts influenced his Ašʿarism is something I
intend to investigate in my PhD thesis.
152
What he rejects is the essential possibility that Avicenna attributes to them but he admits
that they are possible in their local movement.
‫‪AL-MIKLĀTĪ, A TWELFTH CENTURY AŠʿARITE‬‬ ‫‪195‬‬
‫‪APPENDIX: A QUOTATION OF “AVICENNA” IN AL-MIKLĀTĪ’S‬‬
‫‪QUINTESSENCE, PP. 163–6‬‬
‫ﻭﳓﻦ ﺍﻵﻥ ﻧﻮﺭﺩ ﻛﻼﻡ ﺃﰊ ﻋﻠﻲ ﺑﻦ ﺳﻴﻨﺎ‪ ،‬ﻋﻠﻰ ﺇﳚﺎﺯ‪ ،‬ﻭﻧﺒﻴﻦ ﺍﻻﻋﱰﺍﺽ ﻋﻠﻴﻪ‪،‬‬
‫ﻣﺴﺘﻌﻴﻨﲔ ﺑﺎﻟﻠﻪ ﺗﻌﺎﱃ‪ .‬ﻗﺎﻝ ﺃﺑﻮ ﻋﻠﻲ ﺑﻦ ﺳﻴﻨﺎ‪:‬‬
‫ﺍﻟﺪﻟﻴﻞ ﻋﻠﻰ ﺇﺛﺒﺎﺕ ﺍﻟﺼﺎﻧﻊ ﺃﻥ ﻛﻞ ﻣﻮﺟﻮﺩ ﻳﺘﺼﻮﺭ ﺃﻧﻪ ﻣﻮﺟﻮﺩ ]ﻻ ﺑﻌﻠﺔ[ ﺧﺎﺭﺝ‬
‫ﺍﻟﻨﻔﺲ‪ ،‬ﻓﻼ ﳜﻠﻮ ﺗﺼﻮﺭﻩ ﻣﻦ ﺃﺣﺪ ﺃﻣﺮﻳﻦ‪:‬‬
‫ﺇﻣﺎ ﺃﻥ ﻳﺘﺼﻮﺭ ﺃﻧﻪ ﻣﻮﺟﻮﺩ ﻻ ﺑﻌﻠﺔ‪ ،‬ﻭﺇﻣﺎ ﺃﻥ ﻳﺘﺼﻮﺭ ﺃﻧﻪ ﻣﻮﺟﻮﺩ ﺑﻌﻠﺔ‪ .‬ﻭﻳﺴﻤﻰ‬
‫ﺍﻟﺬﻱ ﻳﺘﺼﻮﺭ ﻭﺟﻮﺩﻩ ﺑﻐﲑ ﻋﻠﺔ‪ :‬ﻭﺍﺟﺐ ﺍﻟﻮﺟﻮﺩ؛ ﻭﺍﻟﺬﻱ ﻳﺘﺼﻮﺭ ﻭﺟﻮﺩﻩ ﻣﻦ ﻗﺒﻞ ﻋﻠﺔ‪:‬‬
‫ﳑﻜﻦ ﺍﻟﻮﺟﻮﺩ‪ .‬ﰒ ﻫﺬﺍ ﻳﻨﻘﺴﻢ ﻋﻨﺪ ﺃﰊ ﻋﻠﻲ ﺑﻦ ﺳﻴﻨﺎ ﻗﺴﻤﲔ‪:‬‬
‫ﻓﺈﻥ ﻛﺎﻥ ﻭﺟﻮﺩﻩ ﺩﺍﺋ ًﻤﺎ‪ ،‬ﲰﺎﻩ ﻭﺍﺟﺐ ﺍﻟﻮﺟﻮﺩ ﻣﻦ ﻗﺒﻞ ﻏﲑﻩ >ﳑﻜﻨًﺎ< ﻣﻦ ﻗﺒﻞ‬
‫ﺫﺍﺗﻪ؛ ﻭﺇﻥ ﻛﺎﻥ ﻏﲑ ﺩﺍﺋﻢ ﺍﻟﻮﺟﻮﺩ‪ ،‬ﻭﺟﺐ ﺃﻥ ﻳﻜﻮﻥ ﳑﻜﻦ ﺍﻟﻮﺟﻮﺩ ﻣﻦ ﺫﺍﺗﻪ‪ ،‬ﻭﻣﻦ‬
‫ﻏﲑﻩ‪.‬‬
‫ﻗﺎﻝ‪ :‬ﻭﻫﺬﺍ ﻫﻮ ﺍﻟﺬﻱ ﳚﺐ ﺃﻥ ﻳﺘﻘﺪﻣﻪ ﺍﻟﻌﺪﻡ ﺍﻟﺰﻣﺎﱐ ﻭﺃﻥ‪ 1‬ﺗﻜﻮﻥ ﻟﻪ ﻣﺎﺩﺓ‬
‫ﻣﺘﻘﺪﻣﺔ ﻋﻠﻴﻪ ﺑﺎﻟﺰﻣﺎﻥ‪.‬‬
‫ﰒ ﻳﻀﻊ ﺃﻥ ﻣﻦ ﺧﺎﺻﺔ ﳑﻜﻦ ﺍﻟﻮﺟﻮﺩ ﻣﻦ ﺫﺍﺗﻪ ﺃﻻ ﻳﻮﺟﺪ ﺇﻻ ﻣﻦ ﻏﲑﻩ‪ ،‬ﻭﺳﻮﺍﺀ‬
‫ﻛﺎﻥ ﻣﻦ ﺍﻟﻐﲑ ﳑﻜﻨًﺎ ﺃﻭ ﻭﺍﺟﺒًﺎ‪ .‬ﰒ ﻳﻀﻊ ﺃﻥ ﻣﻦ ﺧﻮﺍﺹ ﻭﺍﺟﺐ ﺍﻟﻮﺟﻮﺩ ﻣﻦ ﺫﺍﺗﻪ ﺃﻥ‬
‫ﻳﻜﻮﻥ ﻭﺍﺣ ًﺪﺍ ﺑﺴﻴ ًﻄﺎ‪ ،‬ﻭﺫﻟﻚ ﺃﻧﻪ ﺇﻥ ﻛﺎﻥ ﻣﺮﻛﺒًﺎ‪ ،‬ﻭﺟﺐ ﺃﻥ ﻳﻜﻮﻥ ﳑﻜﻦ ﺍﻟﻮﺟﻮﺩ ﻣﻦ‬
‫ﺫﺍﺗﻪ‪ ،‬ﻭﺍﺟﺐ ﺍﻟﻮﺟﻮﺩ ﻟﻐﲑﻩ‪ ،‬ﻭﻗﺪ ﺍﺩﻋﻰ ﻭﺍﺟﺐ ﺍﻟﻮﺟﻮﺩ ﻣﻦ ﺫﺍﺗﻪ‪ .‬ﻭﺫﻟﻚ ﺧﻠﻒ ﻻ‬
‫ﳝﻜﻦ‪.‬‬
‫ﻭﻛﺬﻟﻚ ﻳﻀﻊ ﻣﻦ ﺧﻮﺍﺻﻪ ﺃﻻ‪ 2‬ﻳﻮﺟﺪ ﻣﻨﻪ ﺍﺛﻨﺎﻥ‪ ،‬ﺃﻋﲏ ﺃﻧﻪ ﻻ ﻳﻮﺟﺪ ﻣﻮﺟﻮﺩﺍﻥ‬
‫ﻛﻞ ﻭﺍﺣﺪ ﻣﻨﻬﻤﺎ ﻭﺍﺟﺐ ﺍﻟﻮﺟﻮﺩ ﺑﺬﺍﺗﻪ؛ ﻓﺈﻥ ﺍﻹﺛﻨﻴﻨﻴﺔ ﺗﻘﺘﻀﻲ ﻣﻐﺎﻳﺮﺓ ﰲ ﻣﻌﲎ ﻣﻦ‬
‫ﰲ ﺍﳌﻄﺒﻮﻉ‪ :‬ﺃﻻ‪.‬‬ ‫‪1‬‬
‫ﰲ ﺍﳌﻄﺒﻮﻉ‪ :‬ﺃﻥ ﻻ‪.‬‬ ‫‪2‬‬

‫‪196‬‬ ‫‪YAMINA ADOUHANE‬‬
‫ﺍﳌﻌﺎﱐ‪ .‬ﻭﻛﻮﻥ ﻛﻞ ﻭﺍﺣﺪ ﻣﻨﻬﻤﺎ ﻭﺍﺟﺐ ﺍﻟﻮﺟﻮﺩ ﺑﺬﺍﺗﻪ ﻳﻘﺘﻀﻲ ﺍﺷﱰﺍ ًﻛﺎ ﺑﻴﻨﻬﻤﺎ ﰲ‬
‫ﻣﻌﲎ ﻭﺍﺟﺐ ﺍﻟﻮﺟﻮﺩ‪.‬‬
‫ﻓﻴﻜﻮﻥ ﰲ ﻛﻞ ﻣﻨﻬﻤﺎ ﻣﻌﻨﻴﺎﻥ‪ :‬ﺃﺣﺪﳘﺎ ﻳﺸﱰﻛﺎﻥ ﻓﻴﻪ ﻭﺍﻟﺜﺎﱐ ﻳﻔﱰﻗﺎﻥ‪ 3‬ﺑﻪ‪ .‬ﻭﻣﺎ‬
‫ﻛﺎﻥ ﺑﻬﺬﻩ‪ 4‬ﺍﻟﺼﻔﺔ ﻓﻬﻮ ﳑﻜﻦ ﺍﻟﻮﺟﻮﺩ ﻣﻦ ﺫﺍﺗﻪ‪ ،‬ﻭﺍﺟﺐ ﺍﻟﻮﺟﻮﺩ ﻣﻦ ﻏﲑﻩ‪.‬‬
‫ﻗﺎﻝ‪ :‬ﻓﺈﺫﺍ ﺻﺢ ﻫﺬﺍ‪ ،‬ﻓﻨﻘﻮﻝ‪:‬‬
‫ﺇﻥ ﻛﻞ ﻭﺍﺣﺪ ﻣﻦ ﺍﻷﺷﻴﺎﺀ ﺍﻷﺭﺑﻌﺔ ﺃﻋﲏ‪ :‬ﺍﳍﻴﻮﱃ‪ ،‬ﻭﺍﻟﻔﺎﻋﻞ‪ ،‬ﻭﺍﻟﺼﻮﺭﺓ‪ ،‬ﻭﺍﻟﻐﺎﻳﺔ‪،‬‬
‫ﻳﺮﺗﻘﻲ ﻛﻞ ﻭﺍﺣﺪ ﻣﻨﻬﺎ ﺇﱃ ﺳﺒﺐ ﺃﻭﻝ‪ ،‬ﺃﻋﲏ ﺃﻥ ﺍﻷﺳﺒﺎﺏ ﺍﻟﻔﺎﻋﻠﺔ ﺗﺮﺗﻘﻲ ﺇﱃ ﻓﺎﻋﻞ‬
‫ﺃﻭﻝ ﻋﻨﺪﻩ‪ ،‬ﻭﺍﳌﺎﺩﺓ ﺇﱃ ﻣﺎﺩﺓ ﺃﻭﱃ‪ ،‬ﻭﺍﻟﻐﺎﻳﺔ ﺇﱃ ﻏﺎﻳﺔ ﺃﻭﱃ‪ ،‬ﻭﺍﻟﺼﻮﺭﺓ ﺇﱃ ﺻﻮﺭﺓ ﺃﻭﱃ‪.‬‬
‫ﰒ ﻳﻀﻊ ﺃﻥ ﻛﻞ ﻭﺍﺣﺪ ﻣﻦ ﻫﺬﻩ ﺍﻷﺳﺒﺎﺏ ﺍﻷﻭﻝ‪ ،‬ﻣﺎ ﻋﺪﺍ ﺍﻟﻔﺎﻋﻞ ﻟﻴﺲ ﳝﻜﻦ ﰲ‬
‫ﺷﻲء ﻣﻨﻬﺎ ﺃﻥ ﻳﻜﻮﻥ ﻭﺍﺟﺐ ﺍﻟﻮﺟﻮﺩ ﺑﺬﺍﺗﻪ‪ ،‬ﺑﻞ ﻛﻞ ﺃﻭﻝ ﻣﻨﻬﺎ ﻫﻮ ﳑﻜﻦ ﺍﻟﻮﺟﻮﺩ ﻣﻦ‬
‫ﺫﺍﺗﻪ‪ ،‬ﻭﺍﺟﺐ ﻣﻦ ﻏﲑﻩ‪.‬‬
‫ﻭﺫﻟﻚ ﻷﻥ ﺍﻟﺼﻮﺭﺓ ﺍﻷﻭﱃ ﻟﻠﺠﺮﻡ ﺍﻷﻭﻝ‪ ،‬ﻛﺄﻧﻚ ﻗﻠﺖ ﺟﺮﻡ ﺍﻟﺴﻤﺎﺀ ﺍﻷﻭﻝ‪ ،‬ﳌﺎ‬
‫ﻛﺎﻧﺖ ﻻ ﺗﻘﻮﻡ ﺇﻻ ﰲ ﻣﺎﺩﺓ ﻋﻨﺪﻩ‪ ،‬ﻭﺍﳌﺎﺩﺓ ﻻ ﺗﻮﺟﺪ ﺇﻻ ﻣﻘﱰﻧﺔ ﻣﻊ ﺍﻟﺼﻮﺭﺓ ﻛﺎﻥ ﻛﻞ‬
‫ﻭﺍﺣﺪ ﻣﻨﻬﻤﺎ‪ ،‬ﺃﻋﲏ ﺍﻟﺼﻮﺭﺓ ﻭﺍﳌﺎﺩﺓ‪ ،‬ﺩﺍﺧ ًﻼ ﰲ ﺟﻨﺲ ﺍﳌﻤﻜﻦ ﻣﻦ ﺫﺍﺗﻪ‪ ،‬ﺍﻟﻮﺍﺟﺐ ﻣﻦ‬
‫ﻏﲑﻩ‪.‬‬
‫ﻭﻛﺬﻟﻚ ﻳﺸﺒﻪ ﺃﻥ ﻳﻈﻬﺮ ﺍﻷﻣﺮ ﰲ ﺍﻟﻐﺎﻳﺔ ﻣﻦ ﻗﺒﻞ ﺃﻧﻬﺎ ﺻﻮﺭﺓ ﻟﺬﻱ ﻏﺎﻳﺔ ﻭﻛﻤﺎﻝ‬
‫ﻟﻪ‪ .‬ﻓﻤﻦ ﺿﺮﻭﺭﺓ ﻭﺟﻮﺩﻫﺎ‪ ،‬ﻭﺟﻮﺩ ﻣﺎ ﻗﺒﻞ ﺍﻟﻐﺎﻳﺔ‪.‬‬
‫ﻭﺇﺫﺍ ﻛﺎﻧﺖ ﺍﻷﺷﻴﺎﺀ ﻛﻠﻬﺎ‪ ،‬ﺇﻣﺎ ﻣﺮﻛﺒﺎﺕ‪ ،‬ﻭﺇﻣﺎ ﺃﺳﺒﺎﺑًﺎ‪ ،‬ﻭﻛﺎﻥ ﻛﻞ ﻭﺍﺣﺪ ﻣﻦ‬
‫ﻫﺬﻩ ﳑﻜﻦ ﺍﻟﻮﺟﻮﺩ ﺑﺬﺍﺗﻪ ﻭﺇﻥ ﻛﺎﻥ ﻭﺍﺟﺒًﺎ ﺑﻐﲑﻩ‪ ،5‬ﻭﻗﺪ ﺗﻘﺮﺭ ﺃﻥ ﳑﻜﻦ ﺍﻟﻮﺟﻮﺩ ﺑﺬﺍﺗﻪ‪،‬‬
‫ﻭﺇﻥ ﻛﺎﻥ ﻭﺍﺟﺒًﺎ ﺑﻐﲑﻩ‪ ،‬ﳛﺘﺎﺝ ﰲ ﻭﺟﻮﺩﻩ ﺇﱃ ﻭﺍﺟﺐ ﺍﻟﻮﺟﻮﺩ ﻣﻦ ﺫﺍﺗﻪ‪ ،‬ﻭﺇﻻ ﳝﺮ ﺍﻷﻣﺮ‬
‫ﰲ ﺍﳌﻄﺒﻮﻉ‪ :‬ﻳﻌﱰﻓﺎﻥ‪.‬‬ ‫‪3‬‬
‫ﰲ ﺍﳌﻄﺒﻮﻉ‪ :‬ﳍﺬﻩ‪.‬‬ ‫‪4‬‬
‫ﰲ ﺍﳌﻄﺒﻮﻉ‪ :‬ﺑﻐﲑ‪.‬‬ ‫‪5‬‬

‫‪AL-MIKLĀTĪ, A TWELFTH CENTURY AŠʿARITE‬‬ ‫‪197‬‬
‫ﺇﱃ ﻏﲑ ﻧﻬﺎﻳﺔ‪ ،‬ﻭﳌﺎ ﻛﺎﻥ ﻫﺬﺍ ﻟﻴﺲ ﳝﻜﻦ ﺃﻥ ﻳﻜﻮﻥ ﻣﺮﻛﺒًﺎ‪ ،‬ﻭﻻ ﺻﻮﺭﺓ‪ ،‬ﻭﻻ ﻣﺎﺩﺓ ﻭﻻ‬
‫ﻏﺎﻳﺔ؛ ﻓﻘﺪ ﺑﻘﻲ ﺃﻥ ﻳﻜﻮﻥ ﺳﺒﺒًﺎ ﻓﺎﻋ ًﻼ‪.‬‬
‫ﻭﺇﺫﺍ ﻛﺎﻥ ﺫﻟﻚ ﻛﺬﻟﻚ‪ ،‬ﻓﺎﻟﻜﻞ ﳐﱰﻉ ﻣﺒﺘ َﺪﻉ ﻟﻮﺍﺟﺐ ﺍﻟﻮﺟﻮﺩ‪ ،‬ﺣﱴ ﺍﳍﻴﻮﱃ‬
‫ﻭﺻﻮﺭﺓ ﺍﳉﺮﻡ ﺍﻟﺴﻤﺎﻭﻱ‪ ،‬ﺑﻞ ﺍﳉﺮﻡ ﺍﻟﺴﻤﺎﻭﻱ ﺑﺄﺳﺮﻩ‪.6‬‬
‫ﺑﻬﺬﺍ ﺍﻧﺘﻬﻰ ﻛﻼﻡ ﺃﰊ ﻋﻠﻲ ﺑﻦ ﺳﻴﻨﺎ‪.‬‬
‫ﰲ ﺍﳌﻄﺒﻮﻉ‪ :‬ﺑﺄﻣﺮﻩ‪.‬‬ ‫‪6‬‬

doi:10.1017/S0957423912000021 © 2012 Cambridge University Press
REGARDS D’IBN RUSHD SUR AL-JUWAYNĪ

QUESTIONS DE MÉTHODE
MOKDAD ARFA MENSIA

Université de Tunis, Faculté des Sciences Humaines et Sociales de Tunis,
94 boulevard du 9 avril 1938, 1007 Tunis
Email: mokdadarfa@hotmail.com
Abstract. This essay is concerned with the complex relationships between falsafa
and kalām. As regards the history of the latter, it has been generally agreed that
al-Juwaynī played a decisive role at a moment when Avicennism became intrusive.
It is mainly in his al-ʿAqīda al-nizāmiyya that al-Juwaynī initiated a doctrinal and
methodic evolution of Ashʿarism. ˙One necessarily invokes here Ibn Rushd, who, by
exposing the dogmas in their literal manifestation in his al-Kashf ʿan manāhij al-
adilla fī ʿaqāʾid al-milla, actually sought to operate a systematic refutation of the
Ashʿarites’ theses. It should be noted, however, that, despite his critical distance,
Ibn Rushd nonetheless remained attentive to al-Juwaynī’s evolutionary study. In
this contribution, some aspects of the major issues in kalām will be examined: the
existence of God and the argument of takhsīs (particularization), His unicity
(wahdāniyya) and the argument of mumānaʿa ˙ ˙ (mutual hindering), as well as
human˙ agency with the acknowledgement of man’s capacity (qudra) and its role.
Albeit limited in scope, the essay attempts to make a more profound evaluation of
al-Juwaynī’s role at a crucial moment of the history of kalām in its specific develop-
ment during which the Great Masters have not always been faithful to the classical
tenets of their school, but also in its relations with falsafa and its influence on al-
Juwaynī, and then and mainly on his disciple al-Ghazālī’s theological thought.
Résumé. Le présent travail concerne les rapports complexes qui ont existé entre la
falsafa et le kalām. Pour l’histoire propre de ce dernier, l’on s’accorde sur le rôle de
premier ordre joué par al-Juwaynī à un moment décisif où l’avicennisme devenait
envahissant. C’est surtout dans al-ʿAqīda al-nizāmiyya qu’il fit évoluer l’ashʿarisme
aussi bien dans la doctrine que dans la méthode. ˙ Le témoignage d’un faylasūf est
ici invoqué. Ibn Rushd, en voulant exposer dans al-Kashf ʿan manāhij al-adilla fī
ʿaqāʾid al-milla les dogmes dans leur sens manifeste, a cherché à opérer une
réfutation systématique des thèses ashʿarites. Ce faisant, il a été attentif au renouvel-
lement opéré par al-Juwaynī. Tout en restant critique à son égard, il a trouvé chez lui
un réconfort à sa polémique anti-ashʿarite. Sont évoqués ici certains aspects des ques-
tions majeures dans le kalām: l’existence de Dieu avec l’argument de takhsīs (parti-
cularisation), Son unicité avec l’argument de l’empêchement mutuel (mumānaʿa), ˙ ˙
et les actes humains avec l’affirmation de la capacité (qudra) de l’homme et le rôle
de celle-ci. Même si le travail présenté ici peut être continué et élargi, il cherche à
avoir une appréciation plus exacte du rôle joué par ce grand maître de l’ashʿarisme
à travers son petit traité, à une époque jugée cruciale dans l’histoire du kalām
selon son développement propre au cours duquel les grands maîtres n’ont pas été tou-
jours fidèles aux doctrines classiques de l’École, mais aussi dans ses rapports avec la
200 MOKDAD ARFA MENSIA
falsafa et l’influence de celle-ci sur la pensée théologique d’al-Juwaynī et puis et sur-

tout sur celle de son disciple al-Ghazālī.
L’histoire des rapports complexes qui existent entre la falsafa et le

kalām présente encore des aspects qui demanderont longtemps à
être discutés et explicités.1 Et pour l’histoire propre du kalām, l’on
connaît déjà le fameux tournant décrit par Ibn Khaldūn dans sa
Muqaddima quand il reprend la distinction depuis longtemps
consacrée de deux méthodes: celle des “Anciens” (al-aqdamūn) et
celle des “Nouveaux” (al-mutaʾakhkhirūn). En passant sous silence
le rôle du muʿtazilite Abū al-Hudhayl al-ʿAllāf et après avoir montré
comment Abū Bakr al-Bāqillānī (m. 403/1013)2 instaura la première,
en la fondant sur certaines prémisses et en la soumettant à la
règle logique de la réciprocité qui lie la validité de l’argument (dalīl)
à celle de la réalité qu’il établit (madlūl), Ibn Khaldūn place
immédiatement al-Juwaynī (m. 478/1085) comme maître incontesté
de l’ashʿarisme. Après lui, l’usage de la logique se répandit, cette
dernière étant désormais isolée des sciences philosophiques et
considérée dans une neutre instrumentalité mise à la disposition de
tout raisonnement. Elle n’engage plus à accorder foi nécessairement
aux vérités philosophiques et métaphysiques jugées incompatibles
avec le dogme et dont elle était jusque-là solidaire. Ce nouvel usage
aboutit à d’autres fondements et prémisses logiques différents de
ceux de la première méthode. Ce fut là le point de départ de la seconde
méthode, celle des “Nouveaux”, caractérisée par l’abandon et des
prémisses de l’ancienne méthode et de sa règle de réciprocité du
1
Dans ses travaux philosophiques, Richard. M. Frank, à qui nous devons une meilleure con-
naissance de l’ashʿarisme, mais aussi du muʿtazilisme, a pu comparer d’une manière pru-
dente, fine et profonde certains de ces aspects, en essayant de saisir le kalām dans son
essence. Voir par exemple “Kalām and philosophy, a perspective from one problem”, dans
P. Morewedge (éd.), Islamic Philosophical Theology (Albany, 1979), pp. 71–95; “The science
of Kalām”, Arabic Sciences and Philosophy, 2 (1992): 7–37.
2
Voici ce qu’en dit Ibn Khaldūn: “Il est le chef de file de la première méthode, son rôle fut capi-
tal. Il affina la méthode et posa les prémisses rationnelles sur lesquelles reposeront
désormais arguments et raisonnements: démontrer que l’atome et le vide existent, que l’ac-
cident ne peut avoir pour substrat un autre accident, qu’il ne subsiste pas deux moments
successifs, etc. Étant donné que ce sont là les fondements sur lesquels reposent les argu-
ments établissant la vérité des dogmes, et la foi en ceux-ci étant obligatoire, al-Bāqillānī
étendit l’obligation de croire à ces prémisses, les joignant ainsi aux articles de foi. Il
définit ainsi la règle essentielle caractéristique de cette méthode, celle de la réciprocité de
l’argument (dalīl) et de la vérité qu’il établit (madlūl), la validité de l’un est tributaire de
celle de l’autre, le rejet du premier équivaut au rejet du second” (Ibn Khaldūn,
al-Muqaddima, 3e éd. [Beyrouth, 1967], pp. 834–5; voir L. Gardet et M.-M. Anawati,
Introduction à la théologie musulmane. Essai de théologie comparée [Paris, 1970], p. 72).
Les formes de l’argumentation, explique encore Ibn Khaldūn, n’étaient pas toujours techni-
quement satisfaisantes à cause de la naïveté des Mutakallimūn, et parce que la science de la
logique n’était pas encore acceptée ni répandue (id., pp. 835 et 914, où il mentionne d’autres
prémisses: l’instant indivisible, la négation de la nature et de la composition intelligible des
essences, la vérité du hāl (état ou mode). Il mentionne comme partisans de la réciprocité du
dalīl et du madlūl Abū ˙ al-Hasan [al-Ashʿarī] et al-Ustādh Abū Ishāq [al-Isfarāʾ īnī]).
˙ ˙
REGARDS D’IBN RUSHD SUR AL-JUWAYNĪ 201
dalīl et du madlūl.3 Elle fut initiée par al-Ghazālī, continuée et
consacrée par Fakhr al-Dīn al-Rāzī.4
Dans cette histoire de l’ashʿarisme, Ibn Khaldūn, bien que conscient
du rôle essentiel joué par al-Juwaynī, ne s’y attarde pourtant pas,
étant intéressé surtout par le clivage existant entre les deux
méthodes. Il se contente d’évoquer les deux œuvres maîtresses
d’al-Juwaynī: Kitāb al-Shāmil et son résumé al-Irshād. Pour lui,
Imām al-Haramayn suivait encore la méthode des Anciens, et son
˙ le traité type de l’ashʿarisme classique.5 Ce qui est
Irshād reste
6
vrai. Mais ce qu’il omet de dire c’est qu’al-Juwaynī, dans son autre
œuvre appelée al-ʿAqīda al-nizāmiyya,7 fit évoluer considérablement
les positions doctrinales de˙ l’École. Pourtant, bien avant Ibn
Khaldūn, Ibn Rushd, dans sa discussion des thèses ashʿarites,
menée avec le souci acharné de les réfuter méthodiquement, a attiré
l’attention sur ce renouvellement de l’École opéré par al-Juwaynī.
L’auteur de la Nizāmiyya souligne lui-même l’originalité et la valeur
de sa méthode dans˙ l’ashʿarisme, méthode qui est, dit-il, indispensa-
ble aux fondements de l’islam et dispense de trop longues méthodes
traditionnelles.8
L’apport d’al-Juwaynī dans son œuvre en général et dans sa
Nizāmiyya en particulier a été perçu et apprécié par les chercheurs
˙ manière inégale. Dans leur Introduction à la théologie musul-
d’une
mane, L. Gardet et M.-M. Anawati, par exemple, tout en donnant
une appréciation détaillée du rôle décisif joué par al-Juwaynī dans
le renouvellement de l’ashʿarisme,9 n’ont pourtant pas analysé cette
3
Voir dans al-Muqaddima (pp. 914–15) les raisons profondes qui expliquent l’existence d’un
rapport nécessaire entre, d’une part, l’adoption par les “Nouveaux” de la logique et leur rejet
aussi bien des prémisses mentionnées plus haut que de la règle de la réciprocité du dalīl et
du madlūl, et, d’autre part, entre l’adoption par les “Anciens” de cette règle de réciprocité et
leur hostilité à l’égard de la logique. Voir aussi Gardet et Anawati, Introduction à la
théologie musulmane, pp. 72–3.
4
Al-Muqaddima, pp. 835–6, 913 et 915.
5
Ibid., p. 837.
6
Voir al-Irshād, éd. M. Y. Mūsā et ʿA. ʿA. ʿAbd al-Hamīd (Le Caire, 1950), pp. 17 sq.
7
Édité par M. Z. al-Kawtharī (Le Caire, 1948). Il en ˙ existe une traduction allemande par H.
Klopfer, Das Dogma des Imām al-Haramain al-Djuwainī und sein Werk al-ʿAqīda
an-nizāmiya (Le Caire/Wiesbaden, 1958).
8
Pp. 8˙et 12–13.
9
Le rôle capital qu’al-Juwaynī a joué dans l’histoire de l’ashʿarisme est bien décrit par les
deux auteurs. Ils parlent de l’apport propre d’al-Juwaynī (Introduction, p. 67), dans une
période où le kalām se trouve à un tournant décisif de son évolution (p. 181), où la
théologie est moins attachée au littéralisme du texte coranique, elle devient plus
“libérale” (p. 66). Chez al-Ghazālī et son maître, expliquent-ils, le kalām est saisi comme
en son point de démarcation entre la voie des anciens, encore suivie, bien qu’avec de forts
aménagements, par l’Imām al-Haramayn, et la voie des modernes, déjà inaugurée par
al-Ghazālī. La nécessité d’autres˙ procédés logiques, et d’une systématisation logique plus
poussée, se fera rapidement jour dans la ligne même du kalām orthodoxe. Cette nécessité
n’apparaît que timidement encore en cet ashʿarite quelque peu évolué que fut l’Imām
al-Haramayn. Ashʿarite évolué, mais qui se fait le protagoniste des “modes”, et reste
˙
œuvre qui marque d’une manière spécifique ce renouvellement. Harry

A. Wolfson, au contraire, s’intéressa dans The Philosophy of the
Kalam10 à cette épître d’al-Juwaynī et à son appréciation par Ibn
Rushd. Citons aussi H. A. Davidson11 et Mohammad M. A. Saflo12
qui ont souligné l’apport d’al-Juwaynī. Ulrich Rudolh, en analysant
la grande influence d’Ibn Sīnā sur le kalām, évoque, mais très rapide-
ment, al-Juwaynī et son épître.13 La question étant assez étendue,
nous nous contenterons d’en évoquer quelques aspects qui se rappor-
tent aux questions majeures (mabāhith) du kalām: l’existence de
Dieu, Son unicité et les actes humains.˙
C’est surtout dans al-Kashf ʿan manāhij al-adilla fī ʿaqāʾid
al-milla14 que ces questions sont discutées. Rappelons qu’Ibn Rushd
s’y propose “d’examiner. . . les dogmes qui sont manifestes et que la
Loi religieuse a voulu imposer au commun des gens. En tout cela,
nous prenons comme unique critère la finalité fixée par le
Législateur. . .”.15
Il s’agit en somme de rétablir les dogmes dans leur littéralité
initiale et de réfuter leurs interprétations élaborées par les
Mutakallimūn, spécialement ashʿarites, et dans lesquelles ils se
sont écartés de cette signification primordiale. Pour ce faire, Ibn
Rushd suit, pour l’essentiel, la répartition classique des traités du
fidèle à l’ontologie générale que recouvrait la représentation atomistique occasionaliste du

monde, en particulier aux relations par elle précisées entre substance et accidents (p. 367).
Les auteurs trouvent chez al-Juwaynī la tendance à mêler dans le plan de ses ouvrages
l’exposé des thèses philosophiques et théologiques (p. 368). Enfin, en se référant (pp. 73
et 449) à la thèse du shaykh ʿAlī M. Jabr, qui montre comment al-Juwaynī se distingue
de ses devanciers, al-Ashʿarī et al-Bāqillānī, par l’introduction de la logique aristotéli-
cienne dans la méthode d’argumentation, ils commentent en disant: “Juwaynī nous semble
représenter non le chef de file des ‘modernes’ mais plutôt se situer dans l’entre-deux: en
même temps qu’il commençait à employer le syllogisme aristotélicien, il gardait encore
les procédés de l’ancienne logique et une utilisation des atomes et des modes qui se situait
selon l’évolution normale préparée par ses devanciers. D’où sa grande importance dans l’his-
toire du kalām” (p. 73).
10
(Cambridge, Mass. et Londres, 1976), pp. 436–44.
11
Proofs for Eternity, Creation and the Existence of God in Medieval Islamic and Jewish
Philosophy (Oxford, 1987), pp. 144 sq., signalé par Wilferd Madelung dans “Abū l-Husayn
˙
al-Basrī’s proof for the existence of God”, dans J. E. Montgomery (éd.), Arabic Theology,
Arabic Philosophy, from the Many to the One. Essays in Celebration of Richard M. Frank
(Leuven, 2006), pp. 273–80, p. 273.
12
Al-Juwayni’s Thought and Methodology: with a Translation and Commentary on Lumaʿ
al-Adillah (Berlin, 2000) (signalé par. W. Madelung dans “Abū l-Husayn al-Basrī’s proof
for the existence of God”, p. 273, malheureusement nous n’avons pas ˙ pu le consulter).
˙
13
“La preuve de l’existence de Dieu chez Avicenne et dans la théologie musulmane”, dans A. de
Libera, A. Elamrani-Jamal et A. Galonnier (éd.), Langages et philosophie, Hommage à Jean
Jolivet (Paris, 1997), pp. 339–46, p. 344.
14
In Falsafat Ibn Rushd, 2e éd. (Le Caire, 1353/1935), pp. 40–158 (dorénavant Kashf ).
D’autres références à al-Juwaynī dans l’œuvre d’Ibn Rushd, qu’elles soient explicites ou
non, devraient être prises en compte dans une étude plus détaillée que la présente (ex:
Tahāfut al-Tahāfut, éd. M. Bouyges, 2e éd. [Beyrouth, 1987], pp. 541–2).
15
Kashf, pp. 40–1.
kalām,16 cherchant à infirmer une à une les thèses ashʿarites, en mon-
trant qu’elles ne sont conformes ni à la méthode légale imposée par le
Législateur ni à la méthode démonstrative suivie par les philosophes.
Tout en élaborant ces critiques, il prend soin de rétablir chaque dogme
en sa vérité conformément à la voie légale.
I. MÉTHODES POUR DÉMONTRER L’EXISTENCE DE DIEU
Ibn Rushd résume la méthode rationnelle suivie par les Ashʿarites

pour prouver l’existence de Dieu. Elle repose sur la démonstration
du caractère contingent du monde, lequel se fonde sur la composition
des corps d’atomes indivisibles. La contingence de ceux-ci implique
celle des corps, et celle des corps implique celle du monde.17
C’est en réfutant cette méthode qui a prévalu dans l’ashʿarisme
qu’Ibn Rushd y distingue deux voies. Celle, la plus répandue, qui
repose sur trois prémisses faisant fonction de principes de raisonne-
ment, à savoir: 1) qu’il est impossible aux atomes d’exister en étant
dépourvus d’accidents, 2) que les accidents sont créés, 3) que ce qui
ne peut être dépourvu de choses créées est lui-même créé. Ibn
Rushd discute18 ces prémisses une à une avant d’exposer19 la seconde
voie, inventée par al-Juwaynī dans son épître al-Nizāmiyya. Elle
˙
repose sur les deux prémisses suivantes: 1) Il était possible ( jāʾiz)20
que le monde, avec tout ce qu’il contient, fût à l’opposé de ce qu’il
est présentement (il aurait pu, par exemple, être plus petit ou plus
grand), 2) ce qui est possible ( jāʾiz) est créé (muhdath), il a besoin
d’un agent ( fāʿil) qui lui affecte l’un des deux ˙possibles opposés
plutôt que l’autre.21
Ibn Rushd discute ces deux prémisses et les réfute.22 Il qualifie la
première de rhétorique, elle vient immédiatement à l’esprit du com-
mun. Pour lui, toute chose produite ne contient que ce qui est obliga-
toire et nécessaire, ou ce qui la rend plus parfaite et meilleure. C’est là
d’ailleurs, dit-il, le sens même du mot sināʿa (production). Or, cette
première prémisse d’al-Juwaynī dénie au ˙ Créateur toute sagesse, la
sagesse n’étant autre chose que la connaissance des causes.
Quant à la seconde prémisse, à savoir que tout contingent est créé,
elle est loin d’être évidente par soi. Elle constitue d’ailleurs un point
de litige entre les philosophes eux-mêmes: Platon admet qu’il y a un
16
Exception faite du chapitre politique sur l’imamat, introduit auparavant par al-Ashʿarī, en
réaction et en réponse aux traités shīʿites (Ibn Khaldūn, Muqaddima, pp. 833–4).
17
Kashf, p. 43. Cf. Wolfson, Philosophy, pp. 392–402.
18
Kashf, pp. 46–54.
19
Ibid., pp. 54–6.
20
Wolfson traduit jāʾiz par admissible (Philosophy, p. 434).
21
Voir Wolfson, ibid., pp. 434–8.
22
Kashf, pp. 56–9.
possible éternel, Aristote le nie.23 Al-Juwaynī cherche à fonder la

prémisse en question sur d’autres propositions:
1°) Ce qui est possible a besoin d’un déterminant (mukhassis) qui le rende
plus apte à recevoir l’un des deux attributs possibles plutôt˙ ˙ que
˙ l’autre.
2°) Ce mukhassis est nécessairement doté de volonté (murīd).
˙ ˙ ˙ par la volonté est créé dans le temps (hādith).
3°) Ce qui existe
˙
C’est le fameux argument dit de takhsīs (particularisation)24 qui
˙ ˙ qu’agent qui choisit libre-
sert à démontrer l’existence de Dieu en tant
ment ( fāʿil mukhtār). À propos de cet argument, W. Madelung
montre, dans son étude mentionnée plus haut, qu’al-Juwaynī est
devancé par Abū al-Husayn al-Basrī (m. 436/1044) dans son Kitāb
Tasaffuh al-adilla25 ˙tel qu’il est exposé
˙ par son disciple Mahmūd
˙ ˙
al-Malāhimī (m. 536/1141) dans son Kitāb al-Muʿtamad fī ˙ usūl
al-dīn.26˙ ˙
Cette rencontre d’al-Juwaynī et d’al-Basrī est-elle le fruit de deux
développements parallèles avec une influence˙ directe d’Ibn Sīnā
subie également par l’un et par l’autre? Ou bien est-elle le fruit
d’une influence du maître muʿtazilite sur le maître ashʿarite? La
première éventualité ne peut être raisonnablement écartée, avec la
certitude d’une communication des idées dans les deux sens à une
époque où se répand la pensée d’Ibn Sīnā comme le tuteur de la fal-
safa. Ayman Shihadeh montre que l’argument par la particularisa-
tion est utilisé par les premiers Ashʿarites, par les Muʿtazilites et
par les Ashʿarites postérieurs. Dans sa forme simple, il se trouve
chez al-Bāqillānī, lié à l’occasionalisme, à l’atomisme et au rejet de
la causalité naturelle. Il sera repris par al-Juwaynī sous une forme
plus sophistiquée.27 La seconde éventualité, admise par Madelung,28
reste fortement probable, à une époque ou se répandent les deux
23
Distinction entre les deux philosophes grecs déjà signalée dans Fasl al-maqāl, dans Falsafat
Ibn Rushd, p. 21. ˙
24
Kashf, p. 60. Voir aussi Wolfson, Philosophy, pp. 434 ( particularization), 438 et 440 qui cite,
parmi ceux qui mentionnent cet argument, Ghazālī (Tahāfut I, 8, p. 26), Shahrastānī
(Nihāyat al-aqdām, p. 12) et Maïmonide qui fait allusion à al-Juwaynī (Guide, I, 74,
p. 152); Madelung, “Abū l-Husayn al-Basrī’s proof for the existence of God”, p. 274.
25
Abū al-Husayn al-Basrī, Kitāb ˙ Tasaffuh ˙al-adilla. The extant parts introduced and edited
˙
by W. Madelung ˙
and ˙S. Schmidtke˙ (Wiesbaden, 2006).
26
Ed. M. McDermott et W. Madelung (London, 1991).
27
A. Shihadeh, “The existence of God”, dans T. Winter (éd.), The Cambridge Companion to
Classical Islamic Theology (Cambridge et New York, 2008), p. 209 sq.
28
Madelung écrit: “ . . . the Ashʿarī theologian al-Juwaynī must have been well acquainted with
Abū l-Husayn al-Basrī’s Muʿtazilī thought. . . it is evident that much of the apparent orig-
inality ˙ of his theological
˙ thought had been broadly anticipated by Abū l-Husayn.
Al-Juwaynī adopted his views selectively and developed them independently in line ˙ with
his own Ashʿarī creed” (“Abū l-Husayn al-Basrī’s proof for the existence of God”, pp. 278–
9) et “Al-Juwaynī probably was˙the first Ashʿarī ˙ theologian to provide a stringent proof of
particularisation for the existence of God. He certainly received the idea, however, from
the Muʿtazilī Abū l-Husayn al-Basrī” (ibid., p. 280).
˙ ˙
branches rivales: les Bahshamiyya partisans d’Abū Hāshim al-Jubbāʾī
(m. 321/933) dont ʿAbd al-Jabbār (m. 415/1025) et les Husayniyya par-
tisans d’Abū al-Husayn, à qui on reconnaît des affinités ˙ avec les
Falāsifa. 29 ˙
Ceci se passe au Mashriq. Certainement la circulation limitée du
muʿtazilisme au Maghrib, selon le fameux témoignage d’Ibn
Rushd,30 ne peut laisser voir ses développements différenciés.31
Quand, souvent, il expose les thèses muʿtazilites en opposition aux
thèses ashʿarites, il le fait certainement à partir de manuels ashʿarites
qui reproduisent les thèses adverses pour les réfuter: exposé de la
thèse, sa réfutation, les objections et les réponses à ces objections.
Sans disposer de manuels originaux, Ibn Rushd peut donc avoir une
certaine connaissance des thèses muʿtazilites, connaissance qui ne
peut certainement pas couvrir les développements détaillés du muʿta-
zilisme au Mashriq. Elle permet à son apport d’être appréciable en ce
qu’il offre l’intelligence des options doctrinales de base qui comman-
dent les conceptions opposées. Et quand parfois il reconnaît aux
Muʿtazilites une proximité des Falāsifa sur certaines questions32 et
à leurs procédés un certain avantage méthodologique ou logique sur
ceux des Ashʿarites, mais aussi le mérite de ceux-ci d’user moins
d’interprétation que leurs rivaux muʿtazilites,33 les deux démarches
relèvent à ses yeux du même genre ( jins). Ce qui revient à rejoindre
les autres Falāsifa dans la disqualification classique des doctrines
et des méthodes des Mutakallimūn quant à la valeur logique et
épistémologique. De toute façon il porte son attention sur les thèses
dont il est convaincu que la paternité revient à al-Juwaynī,
conformément à la prétention même de celui-ci, et récuse le tajwīz
ashʿarite qui, en niant les causes, nie la sagesse de Dieu.34
Une fois qu’il aboutit à la conclusion que les méthodes pour prouver
l’existence de Dieu répandues dans l’ashʿarisme ne bénéficient ni de la
certitude discursive (nazariyya yaqīniyya) ni de la certitude légale
( yaqīniyya sharʿiyya),35 ˙il peut procéder à l’exposition de sa propre
méthode qu’il dit avoir déduite du Coran et qu’il limite à ses deux
29
Sur ces deux branches, voir Shahrastani, Livre des religions et des sectes, vol. I, trad. avec
introd. et notes par D. Gimaret et G. Monnot (Louvain et Paris, 1986), pp. 265–89. Pour
un état de l’extension du muʿtazilisme, voir l’article bien informé de G. Schwarb,
“Muʿtazilism in the age of Averroes”, dans P. Adamson (éd.), The Age of Averroes: Arabic
Philosophy in the 6th/12th Century, Warburg Institute Colloquia (London, 2011), pp.
251–82, p. 257.
30
Kashf, pp. 64–5.
31
Voir, pour l’état du muʿtazilisme (juif) en Andalousie, Schwarb, “Muʿtazilism in the age of
Averroes”, pp. 280–2.
32
Tahāfut al-Tahāfut, pp. 106, 218, 225, 242, 276.
33
Fasl al-maqāl, p. 33.
34 ˙
Kashf, pp. 113–19.
35
Ibid., pp. 62–3.
fameux arguments: celui de la providence (dalīl al-ʿināya) et celui de

la création sans modèle précédent (dalīl al-ikhtirāʿ).
2. MÉTHODES POUR DÉMONTRER L’UNICITÉ DE DIEU
Sur cette question aussi le témoignage d’al-Juwaynī est invoqué. Le

problème que pose Ibn Rushd est celui de la méthode prônée par la
Loi religieuse pour établir l’unicité divine (wahdāniyya), c’est-à-dire
pour “savoir qu’il n’y a de Dieu que Lui” (lā ˙ilāha illā huwa). Son
expression reste assez énigmatique:
Cette négation, dit-il, est une idée (maʿnā) qui s’ajoute (zāʾid) à l’affirmation
(ījāb) contenue dans cette parole. Or l’affirmation a été établie (thabata)
dans le discours précédent. Que cherche-t-on alors à montrer en établissant
la négation ( famā yutlabu bi-thubūti al-nafyi)?36
˙
Ibn Rushd s’explique en disant que la parole qui affirme l’unicité de
Dieu (kalimat al-tawhīd) contient deux idées, à savoir: l’affirmation de
˙
l’existence de Dieu, démontrée dans le propos précédent, et la négation
de la divinité à tout autre que Lui (nafyu al-ulūhiyya ʿamman siwāhu),
démontrée ici.37 La voie légale pour celle-ci est exposée dans trois ver-
sets du Coran, qui sont autant d’arguments ayant, à ses yeux, le double
avantage d’être énoncés par la Loi religieuse et d’être enracinés dans
la nature humaine. Les versets sont les suivants:
– S’il y avait [dans le ciel et la terre] d’autres divinités que Dieu,
tous deux auraient péri.38
– Dieu n’a pas de fils. Il n’y a avec Lui aucun autre Dieu. S’il en
était ainsi, chacun d’eux se serait approprié sa création et se ser-
ait élevé au-dessus des autres.39
– Dis: “S’il était avec Lui des divinités, comme ils disent, celles-ci
chercheraient un chemin jusqu’au possesseur du Trône”.40
Ibn Rushd procède ensuite à la réfutation de l’argument que, selon
lui, les Ashʿarites “s’évertuent” (tatakallafu) à tirer du verset XVII: 42,
le fameux argument de l’empêchement mutuel (mumānaʿa),41 et
36
Ibid., p. 71.
37
Ibid., p. 76.
38
‫( ﻟَ ْﻮ َﻛﺎ َﻥ ِﻓﻴ ِﻬ َﻤﺎ ﺁ َِﳍ ٌﺔ ﺇﻻ ﺍﻟﻠ ُﻪ ﻟَ َﻔ َﺴ َﺪﺗَﺎ‬Coran, XXI: 22).
39
‫ﺾ ُﺳْﺒ َﺤﺎ َﻥ‬ ٍ ‫َﻣﺎ ﺍ َﲣ َﺬ ﺍﻟﻠ ُﻪ ِﻣ ْﻦ َﻭﻟَ ٍﺪ َﻭ َﻣﺎ َﻛﺎ َﻥ َﻣ َﻌ ُﻪ ِﻣ ْﻦ ﺇِﻟَ ٍﻪ ﺇِ ًﺫﺍ ﻟَ َﺬ َﻫ َﺐ ُﻛ ّﻞ ﺇِﻟَ ٍﻪ َِﲟﺎ َﺧَﻠ َﻖ َﻭﻟَ َﻌ َﻼ ﺑَ ْﻌ ُﻀ ُﻬ ْﻢ َﻋَﻠﻰ ﺑَ ْﻌ‬
.(Coran, XXIII: 91) ‫ﺼ ُﻔﻮ َﻥ‬ ِ ‫ﺍﻟﻠ ِﻪ َﻋﻤﺎ ﻳ‬
َ
40 ِ ِ
‫( ُﻗ ْﻞ ﻟَ ْﻮ َﻛﺎ َﻥ َﻣ َﻌ ُﻪ ﺁ َﳍٌﺔ َﻛ َﻤﺎ ﻳَ ُﻘﻮﻟُﻮ َﻥ ﺇِ ًﺫﺍ َﻻْﺑﺘَ َﻐ ْﻮﺍ ﺇِ َﱃ ﺫﻱ ﺍْﻟ َﻌ ْﺮ ِﺵ َﺳﺒِﻴ ًﻼ‬Coran, XVII: 42).
41
Argument qui, sous le terme de tamānuʿ, semble fonctionner encore chez Ibn Khaldūn
(Muqaddima, p. 830). Pour cet argument de tamānuʿ et mumānaʿa, voir Wolfson
montre qu’un tel argument ne relève ni de ceux qui sont conformes à
la nature, ni de ceux indiqués par la Loi. En gros, cet argument est le
suivant: s’il existait deux dieux, il serait possible qu’ils divergent
( yakhtalifān) [dans leurs actions]. Si tel était le cas, nous nous trou-
verions devant trois alternatives: 1) chacun des deux réaliserait ce
qu’il voudrait, 2) aucun des deux ne réaliserait ce qu’il voudrait, 3)
seul un des deux réaliserait ce qu’il voudrait à l’exclusion du second.
Or, il est impossible qu’aucun des deux ne réalise ce qu’il voudrait,
le monde ne serait alors ni existant ni inexistant. Et il est aussi
impossible que les deux réalisent également ce qu’ils voudraient, le
monde serait alors existant et inexistant à la fois. Seule subsisterait
la possibilité suivante: ce qui serait voulu par l’un se réaliserait, à l’ex-
clusion de ce qui serait voulu par l’autre. Celui qui serait dans le
second cas serait dans l’incapacité de créer, et, par conséquent, ne
saurait être Dieu.42 Cet argument, déjà utilisé par al-Ashʿarī,43 est
consacré par al-Bāqillānī qui le qualifie d’argument de la raison tiré
d’un texte explicite de la Tradition (dalīl maʿqūl mustanbat min
nass manqūl).44 ˙
˙Pour
˙ Ibn Rushd, le point faible dans l’argumentation ashʿarite est le
suivant: en appliquant l’analogie avec les sujets voulants dans le
monde apparent (shāhid), la raison humaine comme elle admet que
les deux dieux puissent agir en divergence ( yakhtalifān), elle admet
également qu’ils puissent agir de concert ( yattafiqān), chose plus
digne des Dieux. S’ils sont d’accord sur la production (sināʿa) du
˙
monde, ils seraient semblables à deux artisans qui fabriqueraient
ensemble une même chose, en concordance et en commun accord.
S’il en est ainsi, il faudrait alors dire que leurs actions se conjugue-
raient, parce qu’elles convergeraient et se rencontreraient dans le
même lieu (mahall).
˙
À ce raisonnement Ibn Rushd substitue un autre qu’il trouve plus
vraisemblable (ashbah) et qu’il formule en syllogisme: s’il y avait
deux dieux, il y aurait deux mondes (car l’action une existe à partir
d’un seul agent). Or, le monde est un. Donc, l’agent est un.
Il y a là deux interprétations différentes faites par Ibn Rushd et les
Ashʿarites du même verset (XVII: 42). Pour ceux-ci, ce verset ne peut
être compris que dans le sens de la divergence (ikhtilāf) et donc de la
concurrence entre les actes respectifs de divinités rivales. C’est pour-
quoi l’empêchement mutuel fonctionne comme un argument pour
(Philosophy, p. 49, note 39) qui le traduit par mutual hindering et hindrance, et renvoie
aussi à Maïmonide (Guide I, 75, 1) et à Shahrastānī (Nihāyat al-aqdām, pp. 91–2).
42
Kashf, p. 74.
43
Al-Luma, éd. R. J. McCarthy (Beyrouth, 1952), p. 8.
44
Al-Insāf, éd. M. Z. al-Kawtharī (Le Caire, 1950), p. 30; al-Tamhīd, éd. M. M. al-Khudayrī et
M. A.˙ Abū Rīda (Le Caire, 1947), p. 46. ˙
l’unicité de Dieu. Pour Ibn Rushd, rien n’empêche que cette parole soit
comprise également dans le sens d’une concordance (ittifāq) entre ces
actes. Car, dit-il, les actes concordants (muttafiqa) se conjuguent pour
aboutir (wurūd) au même lieu d’inhérence (mahall), exactement
comme les actes divergents (mukhtalif) qui confluent ˙ au même lieu
d’inhérence.
Même si, relativement à cet argument, al-Juwaynī semble garder la
position de l’École,45 Ibn Rushd fait encore appel à lui et le cite pour le
mérite d’avoir pu entrevoir cette deuxième hypothèse comme objec-
tion possible à l’argument de l’École, objection jusque-là rejetée par
les Ashʿarites et qu’Ibn Rushd reprendra à al-Juwaynī et leur oppo-
sera de nouveau.
3. MÉTHODES POUR LE PROBLÈME DES ACTES HUMAINS
Ibn Rushd traite de ce problème dans le chapitre: Fī al-qadāʾ wa-

al-qadar (Décret et prédétermination).46 L’ayant considéré comme-˙
l’un des problèmes légaux (sharʿiyya) les plus ardus, il montre
comment il y a opposition (taʿārud) interne aussi bien aux arguments
˙
de la tradition qu’à ceux de la raison. 47 C’est, dit-il, la raison pour
laquelle les Musulmans se sont divisés en deux sectes: les

Muʿtazilites, qui soutiennent que ce que l’homme acquiert (iktisāb)
est la cause des bonnes et des mauvaises actions, et que le
châtiment et la rétribution sont déterminés en conséquence, et les
Jabriyya (déterministes) qui soutiennent tout le contraire (naqīd) et
prétendent que l’homme est déterminé dans ses actions et y est ˙con-
traint. Les Ashʿarites, cherchant à apporter une solution médiane,
ont soutenu que l’homme a bien une acquisition, mais ce par quoi il
acquiert et ce qui est acquis par lui (al-muktasab bihi wa-al-kasb)
sont tous deux la création de Dieu.48
C’est en montrant l’opposition inhérente aux arguments de la rai-
son, comme étant le deuxième facteur de divergence opposant les
sectes, qu’Ibn Rushd expose la position d’al-Juwaynī. Il y a donc
bien une aporie, c’est la suivante: si, d’une part, on admet que
l’homme était créateur de ses actes, il y aurait nécessairement des
actes qui ne se dérouleraient pas conformément à la volonté et au
45
Je pense qu’Ibn Rushd renvoie à al-Irshād (pp. 53–4) où l’auteur rejette cette objection.
46
Kashf, pp. 134 sq.
47
Pour les arguments de Tradition, il y a opposition et non pas contradiction. Cette opposition
subsiste tant que les versets sont compris dans leur signification générale (tadullu
bi-ʿumūmihā); la solution à cette opposition est de leur assigner un sens précis (takhsīs)
(ibid., p. 139). ˙ ˙
48
Position intenable aux yeux d’Ibn Rushd, elle n’a aucun sens, car si l’acquisition et l’acquis
(al-iktisāb wa-al-muktasab) sont tous deux créés par Dieu, l’homme serait nécessairement
contraint dans son acquisition (majbūr ʿalā iktisābihi) (ibid., pp. 136 et 143).
libre choix de Dieu, et il y aurait de la sorte un autre créateur que
Lui.49 Or, sur ce point, les Musulmans sont unanimes à penser qu’il
n’y a de créateur que Dieu. Mais si, en revanche, l’on nie que l’homme
soit acquéreur (muktasib) de ses actes, il serait nécessairement con-
traint (majbūr), toute autre position intermédiaire entre contrainte
( jabr) et acquisition (iktisāb) étant inexistante. Ibn Rushd dégage
une série de conséquences auxquelles mène la seconde position,
conséquences jugées absurdes. Elles sont reprises aux objections clas-
siques formulées par les Muʿtazilites. C’est dans ce contexte qu’il mon-
tre comment la position d’al-Juwaynī se démarque de la position
ashʿarite classique. Si l’homme n’était donc pas acquéreur de ses
actes, il y serait nécessairement contraint, l’obligation légale qui lui
serait imposée par Dieu (taklīf) relèverait alors de l’imposition de
“ce qui serait au-dessus de sa capacité” (mā lā yutāq), et il n’y aurait
de la sorte aucune différence entre cette imposition˙ et celle qu’on
adresserait aux choses inertes qui, elles, ne sont évidemment pas
dotées de capacité (istitāʿa). C’est la raison pour laquelle la majorité
( jumhūr) a admis que la˙ capacité est une condition requise dans l’im-
position religieuse (taklīf), au même titre que la raison.50
C’est pourquoi, dit Ibn Rushd, nous trouvons Abū al-Maʿālī dire dans
al-Nizāmiyya que l’homme a bien une acquisition de ses actes, et une
˙
capacité de les faire. En cela, il s’appuie sur l’impossibilité d’imposer
légalement [à l’homme] ce qu’il serait incapable d’accomplir, mais
[al-Juwaynī affirme cette impossibilité en se plaçant] d’un point de vue
( jiha) autre que celui des Muʿtazila quand ceux-ci nient cette même impo-
sition. Quant aux anciens Ashʿarites, ils ont admis la possibilité ( jawwazū)
de l’imposition légale [à l’homme] de ce qu’il est incapable de faire, cherchant
ainsi à éviter le principe (asl) sur la base duquel les Muʿtazila avaient rejeté
˙
cette imposition, à savoir que c’est mauvais selon la raison. Les Ashʿarites
tardifs ont soutenu une position différente (khālafahum) [de celles des
anciens].51
49
Arguments qui se trouvent dans al-Nizāmiyya, pp. 33 et 37.
50
Ibid., p. 43. ˙
51
Kashf, p. 136–137:
.‫ ﻓﺈﻧﻪ ﻻ ﻭﺳﻂ ﺑﻴﻦ ﺍﳉﱪ ﻭﺍﻻﻛﺘﺴﺎﺏ‬،‫ ﻭﺟﺐ ﺃﻥ ﻳﻜﻮﻥ ﳎﺒﻮﺭﺍ ﻋﻠﻴﻬﺎ‬،‫ﻭﺇﻥ ﻓﺮﺿﻨﺎﻩ ﺃﻳﻀﺎ ﻏﲑ ﻣﻜﺘﺴﺐ ﻷﻓﻌﺎﻟﻪ‬
‫ ﻭﺇﺫﺍ ﻛﻠﻒ ﺍﻹﻧﺴﺎﻥ ﻣﺎ ﻻ ﻳﻄﻴﻖ‬.‫ﻭﺇﺫﺍ ﻛﺎﻥ ﺍﻹﻧﺴﺎﻥ ﳎﺒﻮﺭﺍ ﻋﻠﻰ ﺃﻓﻌﺎﻟﻪ ﻓﺎﻟﺘﻜﻠﻴﻒ ﻫﻮ ﻣﻦ ﺑﺎﺏ ]ﻣﺎ[ ﻻ ﻳﻄﺎﻕ‬
‫ ﻭﻛﺬﻟﻚ ﺍﻹﻧﺴﺎﻥ ﻟﻴﺲ ﻟﻪ ﻓﻴﻤﺎ ﻻ‬.‫ ﻷﻥ ﺍﳉﻤﺎﺩ ﻟﻴﺲ ﻟﻪ ﺍﺳﺘﻄﺎﻋﺔ‬،‫ﱂ ﻳﻜﻦ ﻓﺮﻕ ﺑﻴﻦ ﺗﻜﻠﻴﻔﻪ ﻭﺗﻜﻠﻴﻒ ﺍﳉﻤﺎﺩ‬
‫ ﻭﳍﺬﺍ ﳒﺪ‬.‫ ﻭﳍﺬﺍ ﺻﺎﺭ ﺍﳉﻤﻬﻮﺭ ﺇﱃ ﺃﻥ ﺍﻻﺳﺘﻄﺎﻋﺔ ﺷﺮﻁ ﻣﻦ ﺷﺮﻭﻁ ﺍﻟﺘﻜﻠﻴﻒ ﻛﺎﻟﻌﻘﻞ ﺳﻮﺍﺀ‬.‫ﻳﻄﻴﻖ ﺍﺳﺘﻄﺎﻋﺔ‬
‫ ﻭﺑﻨﺎﻩ ﻋﻠﻰ ﺍﻣﺘﻨﺎﻉ ﺗﻜﻠﻴﻒ ﻣﺎ‬،‫ﺃﺑﺎ ﺍﳌﻌﺎﱄ ﻗﺪ ﻗﺎﻝ ﰲ ﺍﻟﻨﻈﺎﻣﻴﺔ ﺇﻥ ﻟﻺﻧﺴﺎﻥ ﺍﻛﺘﺴﺎﺑﺎ ﻷﻓﻌﺎﻟﻪ ﻭﺍﺳﺘﻄﺎﻋﺔ ﻋﻠﻰ ﺍﻟﻔﻌﻞ‬
‫ ﻭﺃﻣﺎ ﻗﺪﻣﺎﺀ ﺍﻷﺷﻌﺮﻳﺔ ﻓﺠﻮﺯﻭﺍ ﺗﻜﻠﻴﻒ ﻣﺎ ﻻ ﻳُﻄﺎﻕ ﻫﺮﺑﺎ ﻣﻦ‬.‫ ﻟﻜﻦ ﻣﻦ ﻏﲑ ﺍﳉﻬﺔ ﺍﻟﱵ ﻣﻨﻌﺘﻪ ﺍﳌﻌﺘﺰﻟﺔ‬،‫ﻻ ﻳﻄﺎﻕ‬
.‫ ﻭﺧﺎﻟﻔﻬﻢ ﺍﳌﺘﺄﺧﺮﻭﻥ ﻣﻨﻬﻢ‬.‫ﺍﻷﺻﻞ ﺍﻟﺬﻱ ﻣﻦ ﻗﺒﻠﻪ ﻧﻔﺘﻪ ﺍﳌﻌﺘﺰﻟﺔ ﻭﻫﻮ ﻛﻮﻧﻪ ﻗﺒﻴﺤﺎ ﰲ ﺍﻟﻌﻘﻞ‬
Comment comprendre ce témoignage d’Ibn Rushd sur al-Juwaynī?

Tout en donnant les raisons qui rendent compte de l’attitude
d’al-Juwaynī, il est clair qu’il l’apprécie parce qu’elle confirme à ses
yeux sa critique des thèses des anciens Ashʿarites.
Il est vrai que dans al-ʿAqīda al-nizāmiyya, l’affirmation de la
capacité (qudra) de l’homme et du rôle ˙joué par cette capacité dans
l’acte qui s’y rapporte (maqdūr) constitue le premier pilier sur lequel
repose la possibilité même de l’imposition légale. Pour al-Juwaynī,
Dieu a imposé aux hommes des actes qu’ils ont à accomplir dans
cette vie-ci et pour lesquels ils seront sanctionnés dans l’autre vie. Il
les a dotés de la capacité (aqdarahum) d’accomplir ce qu’Il leur a
demandé. Quiconque “met encore en doute le fait que les actes des
hommes se déroulent conformément à leur préférences, choix et apti-
tudes, est quelqu’un qui est atteint dans sa raison, ou qui persiste
dans son mimétisme et s’entête dans son ignorance.”52
Peut-on être plus dur à l’égard des maîtres de l’École?
Ibn Rushd explique53 que l’idée d’al-Juwaynī, à savoir que l’homme
a bien une acquisition de ses actes et une capacité de les accomplir, se
fonde chez lui sur l’impossibilité d’imposer légalement à l’homme ce
dont il est incapable. Al-Juwaynī dit effectivement dans l’épître en
question que quiconque prétend que la puissance créée n’intervient
nullement dans les actes relatifs à elle (maqdūruhā), doit admettre
en conséquence que si l’on exigeait de l’homme qu’il accomplisse des
actes, ce serait comme si on exigeait de lui de créer en lui-même des
couleurs et des perceptions. Ce qui est, on le sait, chose inadmissible.54
L’impossibilité d’imposer légalement à l’homme ce dont il est incapable
est donc proclamée hautement et clairement par al-Juwaynī.55 Et c’est
sur cette base qu’il reconnaît et la raison et la capacité comme con-
ditions nécessaires à toute imposition religieuse.56
Mais, dit Ibn Rushd, si al-Juwaynī affirme cette impossibilité, il le
fait néanmoins selon un point de vue ( jiha) autre que celui des
Muʿtazila quand ceux-ci nient cette même imposition. Pour comprendre
cette différence relevée par Ibn Rushd, rappelons que pour les
Muʿtazilites, la raison humaine est généralement apte à saisir ce qui
est bien et ce qui est mal. Mais si elle peut le faire c’est que les choses
et les actes sont axiologiquement déterminés: en eux-mêmes ils sont
bons ou mauvais. Axiologie et ontologie sont pour ainsi dire solidaires.
En possession de ces critères, la raison humaine, avant même la venue
de la Loi religieuse (al-ʿaql qabla wurūd al-samʿ), peut de la sorte
52
Al-Nizāmiyya, p. 30.
53
Kashf,˙ p. 136.
54
Objection muʿtazilite encore rejetée par al-Juwaynī dans al-Irshād, p. 203.
55
Al-Nizāmiyya, p. 32. Cette position est confirmée dans son autre livre: al-Burhān fī usūl
˙ éd. ʿAbd al-ʿAzīm Mahmūd al-Dīb (al-Mansūra, 1997), p. 89.
al-fiqh, ˙
56
Al-Nizāmiyya, p. 42. ˙ ˙ ˙
˙
qualifier moralement les actions, qu’elles soient d’ailleurs humaines ou
divines. C’est à ce titre que la raison exclut l’imposition légale de ce que
l’homme est incapable de faire comme étant quelque chose de mauvais,
et donc incompatible avec la conduite du Dieu sage. Celui-ci fait plutôt
figure d’un juge qui se plie au code universel du bien et du mal. Les
Ashʿarites, en rejetant la qualification inhérente aux choses en bien
et en mal, soustraient l’ontologie à l’axiologie; par ailleurs, en rejetant
les jugements moraux objectifs et rationnels et en reconnaissant les
préceptes de la Loi comme source de ces jugements, ils admettent la
possibilité ( jawwazū) de l’imposition légale à l’homme de ce qu’il est
incapable d’accomplir, puisqu’il n’y a point de norme morale rationnelle
pour soutenir l’incompatibilité de cette imposition et de la capacité
humaine. Il est vrai qu’il s’agit là de la simple possibilité rationnelle
( jawāz ʿaqlī) et non pas de l’imposition effective: la volonté divine
n’étant soumise à aucune norme extérieure, on ne peut plus exclure
logiquement la possibilité qu’elle impose à l’homme ce dont il serait
incapable. Même si elle le faisait, elle ne serait pas pour autant injuste.
Il y va du statut de la divinité: volonté du Maître, possesseur absolu et
transcendant toute réglementation.
Dans Kitāb al-Irshād,57 al-Juwaynī défendait encore la doctrine
d’al-Ashʿarī58 de la possibilité rationnelle d’une telle imposition. C’est
seulement dans sa Nizāmiyya qu’il change de doctrine en optant pour
˙
l’impossibilité en question, se rapprochant ainsi de la position adverse
des Muʿtazila. Mais, ne l’oublions pas, il le fait, note Ibn Rushd, selon
un autre point de vue ( jiha) que celui des Muʿtazila, c’est-à-dire en par-
tant d’autres prémisses que celles qui posent des critères objectifs et des
jugements moraux rationnels (tahsīn wa-taqbīh ʿaqliyyān). En effet,
al-Juwaynī continue dans cette ˙épître à nier˙ l’objectivisme moral.
Tous les jugements moraux sont déterminés par des situations
bénéfiques ou nuisibles à l’homme, et se ramènent donc à des senti-
ments de confiance ou de peur. Tout ce que l’on prend pour des vertus
morales a le plus souvent pour origine la satisfaction d’être flatté
(al-ihtizāz bi-husni al-thanāʾ = la sensibilité à l’éloge) ou, parfois, une
˙
certaine tendresse propre au genre humain (riqqat al-jinsiyya). Or,
tout cela est complètement étranger à Dieu dont les actes ne contien-
nent au fond ni bien ni mal.59 Imposer à l’homme ce qui est au-dessus
de ses capacités ne serait donc en rien un mal de la part de Dieu.
C’est pourquoi al-Juwaynī, contrairement aux Muʿtazila, ne fonde pas
sur les jugements de bien et de mal rationnels l’impossibilité d’imposer
57
Pp. 226–8.
58
Sur la possibilité d’imposer à l’homme ce qui est au-dessus de sa capacité, voir al-Ashʿarī,
al-Ibāna ʿan usūl al-diyāna (Le Caire, 1929), pp. 52–3; Ibn Fūrak, Mujarrad Maqālāt
˙ Gimaret (Beyrouth, 1987), p. 111.
al-Ashʿarī, éd. D.
59
Al-Nizāmiyya, pp. 26–7.
˙
à l’homme ce qu’il n’est pas en son pouvoir de faire. Sur quelles autres
prémisses assoit-il alors son raisonnement? Ibn Rushd ne se donne
pas la peine de le dire.
Pour répondre à cette question, revenons au problème auquel
al-Juwaynī tente d’apporter une nouvelle solution: selon quelle
modalité et dans quelle mesure la capacité propre à l’homme
intervient-elle dans ses actes? À ce problème trois solutions sont
entrevues, mais pour être immédiatement écartées: 1) soit l’homme
prétend à la création exclusive de ses actes par lui-même, 2) soit il
se soustrait aux obligations des Lois religieuses, ce qui revient à
nier à sa capacité toute efficace, 3) soit il se proclame associé de
Dieu dans la création du même acte.
Al-Juwaynī critique en des termes très sévères les thèses ashʿarites
typiques, comme celle de l’acquisition (kasb), celle de la volonté et de
la liberté absolue de Dieu.60 Pour lui, nier que la capacité de l’homme
intervienne réellement dans ses actes, tout en continuant à lui impo-
ser légalement de les accomplir, reviendrait à lui imposer ce qu’il n’est
pas en son pouvoir de faire. Une telle attitude aboutirait tout simple-
ment à abolir les Lois religieuses et rendre caduc tout message des
prophètes.61 L’imposition ne peut se faire sans poser au préalable cer-
tains attributs, dont la capacité, qui qualifient l’homme, sujet de l’im-
position. Il y va de la validité de l’imposition légale et de la véracité
des prophètes. En somme, c’est une exigence de cohérence du
Législateur, qui ne saurait se contredire. C’est bien de ce point
qu’al-Juwaynī part pour soutenir l’impossibilité en question. Pour
lui, force est d’admettre que la puissance créée (qudra hāditha) inter-
˙
vient dans l’acte qui lui est relatif (son maqdūr), mais qu’il est impos-
sible de soutenir de manière absolue et indéterminée (itlāqu al-qawl)
que l’homme soit créateur de ses actes, et qu’il est tout˙ aussi impos-
sible de soutenir que l’acte de l’homme se réalise conjointement par
sa puissance créée et la puissance éternelle de Dieu.62
La solution apportée par al-Juwaynī dans sa Nizāmiyya63 consiste à
dire que la capacité de l’homme est créée par Dieu˙ et que l’acte qui lui
est relatif se produit (wāqiʿ) par cette capacité, mais que ce dernier est
rapporté à Dieu en tant que déterminé (taqdīr) et créé (khalq) par Lui,
puisqu’il est produit par cette capacité qui est un acte de Dieu et non
pas de l’homme. Dieu a doté l’homme du libre choix par lequel il utilise
( yasrif) sa capacité.
˙ réalité, les principes mêmes sur lesquels repose l’imposition
En
légale reposent à leur tour sur un principe supérieur, considéré
60
Ibid., pp. 30–3.
61
Ibid., pp. 30–1.
62
Ibid., p. 32.
63
Ibid., pp. 34–5.
comme fondement ultime du dogme (qāʿidat al-ʿaqīda) où le concept
du possible est central ici aussi: ce qui est exigé de l’homme c’est tou-
jours le possible ( jāʾiz) et jamais l’impossible (mustahīl). On lui
demande soit d’accomplir un acte, soit de s’en abstenir,˙ et les deux
sont également des possibles.64
REMARQUES
Ces brèves remarques peuvent être continuées à propos d’autres ques-

tions discutées par Ibn Rushd et pour lesquelles al-Juwaynī a
renouvelé la méthode ashʿarite en se démarquant des positions classi-
ques de l’École.65 Ces quelques passages de al-Kashf ʿan manāhij
al-adilla d’Ibn Rushd, quoique de nature polémique, sont de la
première importance: ils contribuent à une meilleure connaissance
de la pensée d’al-Juwaynī et de son évolution, que ce soit d’ailleurs
dans les options doctrinales ou dans les procédés logiques, et permet-
tent donc une appréciation plus juste du rôle joué par ce grand maître
de l’ashʿarisme à travers sa petite épître al-Nizāmiyya, à une époque
jugée cruciale dans l’histoire du kalām et celle ˙de ses rapports avec la
falsafa.
Ibn Khaldūn a agi comme historien des disciplines scientifiques,
mais aussi en tant que partisan de telle ou telle doctrine. Il était
d’obédience ashʿarite, mais une obédience difficile à tenir après
la synthèse philosophico-théologique dont il décrit lui-même d’une
manière magistrale les étapes, les tournants et les raisons
épistémologiques, obédience rendue encore plus difficile après la
généralisation du dispositif philosophique logique et doctrinal. Ce
grand historien du kalām ne nous a pas renseignés sur ce rôle
d’al-Juwaynī, pour s’être tenu à ses traités classiques. Un grand
adversaire du kalām ashʿarite, comme Ibn Rushd, qui était
intéressé surtout par sa réfutation systématique, a pu repérer ce tour-
nant décisif dans l’histoire de l’ashʿarisme tout en continuant à mon-
trer son insatisfaction devant certains aspects de cet apport.
D’un autre côté, les précédentes remarques cherchent à jeter encore
quelque lumière sur cette œuvre dans laquelle un faylasūf cherche à
exposer un système de dogme dont il veut retrouver la première
expression littéraliste élaborée selon l’intention première du
Législateur. Il s’agit d’Ibn Rushd et de son livre al-Kashf. Certes,
son entreprise reste conforme à la philosophie de la religion
d’al-Fārābī et d’Ibn Sīnā, puisque dans une téléologie du bonheur et
64
Ibid., p. 42.
65
Exemples: la théorie d’al-Juwaynī sur la possibilité de la vision (ruʾya) de Dieu et sur la
direction ( jiha) (al-Nizāmiyya, p.45) discutée par Ibn Rushd (Kashf, p. 105) qui renvoie à
l’Irshād (pp. 166–86). ˙
dans le cadre politique indispensable, le législateur, conscient de

l’inégalité parmi les hommes quant à la saisie des vérités, élabore
une version imagée et imitative (tamthīl et muhākāt) de ces vérités
(maʿānī bātina), version qui est l’expression ˙ littérale (zāhir)
˙
nécessaire au discours religieux. Elle est donc telle quelle voulue ˙ par
le Législateur et il faut absolument la préserver de tout exposé
interprétatif non autorisé (taʾawwul). Le littéralisme dont il s’agit ici
est seulement théorétique (zāhiriyya nazariyya), il porte sur le
dogme (ʿaqāʾid) et les principes ˙ ˙
de la religion (usūl al-dīn) et le
littéralisme pratique (zāhiriyya ‘amaliyya), qui rejette ˙ le raisonne-
˙
ment par analogie (qiyās) comme source de jurisprudence dans les fon-
dements du Droit (usūl al-fiqh) n’est pas concerné ici.66 Littéralisme à
l’usage du jumhūr et˙ interprétation démonstrative (taʾwīl) réservée au
faylasūf vont de pair. Ce en quoi Ibn Rushd est en parfait accord avec
Ibn Sīnā.67 Si le littéralisme est préconisé, c’est parce qu’il renvoie à un
niveau intelligible de vérités (haqāʾiq) comme à son garant. Ibn Sīnā
était déjà là avec une théologie ˙ philosophique, bien assisse sur ces
vérités, bien construite et opposée au kalām.68
Ce genre de recherche nous renseigne certainement sur le type de
solutions proposées par Ibn Rushd pour chaque problème classique:
existence de Dieu, liberté humaine, l’Au-delà, etc., telles qu’elles
sont dictées par les engagements philosophiques ou telles qu’elles
sont traitées dans une théologie littéraliste. Il serait assez instructif
de comparer par exemple la solution au problème des actes humains
proposée par al-Juwaynī, bon connaisseur de la falsafa, et celle
apportée par Ibn Rushd,69 bon connaisseur du kalām. Tous les deux
ont au moins ceci de commun: le rejet de la théorie du kasb
d’al-Ashʿarī. Ont-ils apporté une solution véritable au problème? On
ne saurait y répondre facilement ici. En offrant à l’homme la
disponibilité de deux possibles (accomplir l’acte ou s’en abstenir),
66
Tahāfut al-Tahāfut, pp. 428–30. L’analyse, la méthode d’approche et l’appréciation du
zahirisme pratique par Ibn Rushd varient selon ses écrits: al-Darūrī fī usūl al-fiqh aw
˙
Mukhtas ˙
ar al-Mustasfā [de Ghazālī] écrit en 522/1157 (éd. J. al-ʿAlawī ˙
[Beyrouth, 1994],
˙
pp. 115–17, 125–6, ˙130–1), Bidāyat al-mujtahid wa-nihāyat al-muqtasid écrit en 563/
1168 ([Le Caire, s.d.], t. I, pp. 2–3) et Tahāfut al-Tahāfut écrit ˙ en 576/1180.
Cf. J. al-ʿAlawī, al-Matn al-rushdī (Casablanca, 1986), pp. 66, 100. Voir sur la question
notre article, “Ibn Rushd et le zāhirisme pratique”, dans A. Hasnawi (éd.), La lumière de l’in-
tellect. La pensée scientifique ˙ et philosophique d’Averroès dans son temps. Actes du
IVecolloque international de la SIHSPAI, Cordoue, 9-12 décembre 1998 (Leuven, 2011),
pp. 457–70.
67
Al-Shifā’, al-Ilāhiyyāt, ed. M. Y. Moussa, S. Dunya et S. Zayed (Le Caire, 1960), pp. 441–3;
Ibn Sīnā, al-Adhawiyya fī al-maʿād, éd. H. ʿĀsī (Beyrouth, 1404/1984), pp. 97–103.
68 ˙
Voir l’exposé ˙ concentré et succinct de˙ cette
˙ théologie dans sa Risāla al-ʿArshiyya
(Hyderabad, 1353 H).
69
Kashf, pp. 137–43. La solution est médiane entre déterminisme et muʿtazilisme: affirmer la
volonté humaine mais en rapport avec les causes disposées par Dieu, solution médiane
différente de celle du kasb ashʿarite, qualifié de totalement inexistant.
al-Juwaynī semble pouvoir rejeter la doctrine d’al-Ashʿarī de la
possibilité d’imposer à l’homme ce dont il est incapable avec son argu-
ment classique que les choses n’arrivent pas toujours conformément à
notre volonté. Il est possible qu’il ait réussi à avoir sa propre doctrine
différente de celle du kasb d’al-Ashʿarī, comme il est possible qu’Ibn
Rushd se soit démarqué du déterminisme d’Ibn Sīnā et de son parti
pris anti-muʿtazilite.70
Nous serions peut-être en mesure d’avoir une meilleure
appréciation des doctrines suivant deux axes. Le premier est celui
du développement propre au kalām dont les grands maîtres ne sont
pas toujours fidèles aux doctrines classiques de l’École, ayant parfois
l’air de dire qu’au fond l’adversaire n’est pas totalement et toujours
dans l’erreur et que les amis ne sont pas totalement et toujours
dans le vrai. Il arrive qu’on assume le risque de contredire les compa-
gnons de l’École (mukhālafat al-ashāb). Cela arrive dans les deux
camps. Il suffit de citer le fameux ˙ ˙ cas de la doctrine des ahwāl
(modes ou états) développée par Abū Hāshim al-Jubbāʾī, refusée˙ par
son père Abū ʿAlī et par al-Ashʿarī et adoptée par des maîtres ashʿa-
rites comme al-Bāqillānī (après hésitation) et al-Juwaynī (qui a fini
par la refuser),71 ou celui non moins fameux de la théorie du kasb
d’al-Ashʿarī rejetée aussi bien par certains Ashʿarites que par les
Muʿtazilites en vue d’une reconnaissance d’une part d’efficace à
l’agent humain. C’est certainement le fruit de la longue activité dia-
lectique de controverses, subtiles et ardues qui poussent parfois à se
rendre à l’évidence. C’est dans ce sens qu’al-Juwaynī, pour sortir
des apories et dépasser les insuffisances des solutions de l’École, a
peut-être recours à des théories proposées dans le muʿtazilisme:
Abū al-Husayn al-Basrī ou autres. Le deuxième axe est celui de l’influ-
ence de ˙la falsafa sur˙ la pensée théologique d’al-Juwaynī et d’autres
où ces recherches rendent une meilleure description des théories et
des concepts, de leur circulation entre le kalām et la falsafa, et de
leur aménagement respectif. L’on sait que l’un des griefs majeurs,
d’Ibn Rushd à l’égard d’Ibn Sīnā est une certaine perméabilité aux
doctrines des Mutakallimūn.72 En revanche, c’est déjà avec Abū
70
Ibn Sīnā, Risālat al-Qadar, éd. M. A. F. Mehren (Leyde, 1899).
71
Al-Shahrastānī, Nihāyat al-aqdām fī ʿilm al-kalām, éd. A. Guillaume (Oxford, 1934), pp. 131
sq.; al-Juwaynī, Irshād, pp. 80–2.
72
Ibn Rushd remarque qu’Ibn Sīnā admet, d’une certaine manière, la prémisse (discutée plus
haut) qui sera admise par al-Juwaynī, à savoir qu’il était possible ( jāʾiz) que le monde fût à
l’opposé de ce qu’il est présentement (Kashf, pp. 57–8). En parlant d’“éléments familiers” et
de “concepts communs à Ibn Sīnā et au kalām”, Ulrich Rudoplh écrit que: “. . . la rencontre
entre Avicenne et les théologiens est un événement à plusieurs facettes. C’est un échange au
cours duquel chaque partie a donné et a reçu – même s’il faut bien reconnaître que ce fut
plutôt le kalām qu’Avicenne qui en tira profit” (“La preuve de l’existence”, p. 340).
L’auteur renvoie à Ibn Rushd (Tahāfut al-Tahāfut, éd. M. Bouyges [Beyrouth, 1930],
p. 276) qui indique cet emprunt d’Ibn Sīnā au kalām, emprunt qu’il faut limiter parce
al-Husayn al-Basrī et puis avec al-Juwaynī, maître de Ghazālī73 et qui

˙
continuera certes˙ à réfuter les thèses des Falāsifa (Dieu comme cause
nécessitante et éternité du monde),74 que le kalām commence à céder
aux méthodes et aux idées philosophiques d’Ibn Sīnā qui a élaboré
une puissante théologie philosophique renvoyant avec mépris dos à
dos Ashʿarites et Muʿtazilites sur les topiques classiques du
kalām.75 La séduction du Shaykh al-Raʾīs devenait irrésistible, elle
sera consacrée par al-Ghazālī comme par un disciple qui ne veut
pas l’avouer.76 Les auteurs ne cesseront de dévoiler cette relation:
Abū Bakr Ibn al-ʿArabī, Ibn Rushd et Ibn Taymiyya.*
que pour le philosophe le monde est éternel et coexiste avec sa cause. Voir aussi J. Jolivet,
“Aux origines de l’ontologie d’Ibn Sīnā”, dans J. Jolivet et R. Rashed (éd.), Études sur
Avicenne (Paris, 1984), pp. 11–28.
73
Anawati et Gardet (Introduction, p. 369) montrent comment il y a dans al-Iqtisād de Ghazālī
abandon quasi total de l’atomisme des “anciens” et franche adoption du ˙ raisonnement
déductif. Les caractéristiques mêmes de la “voie des modernes” s’affirment.
74
Al-Juwaynī les qualifie de matérialistes (mulhida) en ce qu’ils contredisent l’un des prin-
cipes fondamentaux, à savoir que les atomes (˙jawāhir, qu’ils appellent hayūlā) ne peuvent
exister sans les accidents (qu’ils appellent sūra). Pour al-Juwaynī, il y a impossibilité qu’il y
ait des choses qui adviennent (hawādith)˙ et qui n’aient pas de commencement. (Irshād,
˙
pp. 23 et 25). Il est clair qu’il expose ici la doctrine des philosophes sur l’éternité du monde.
75
Exemple sa critique du hudūth des “Mutakallimūn à l’esprit limité” dans al-Najāt, éd. M. S.
e ˙
al-Kurdī, 2 éd. (Le Caire, 1357/1938), pp. 213–14; al-Shifāʾ, al-Ilāhiyyāt, pp. 259–68. ˙
76
Voir Richard M. Frank, Creation and the Cosmic System: Al-Ghazâlî & Avicenna
(Heidelberg, 1992), ex. pp. 12–21. En parlant d’al-Ghazālī, l’auteur conclut en disant:
“what he has to say concerning God’s relation to the cosmos as its creator, however, reveals
that from a theological standpoint most of the theses which he rejected are relatively tame
and inconsequential compared to some of those in which he follows the philosopher” (ibid.,
p. 86).
*
Une première version du présent travail a été présentée lors du colloque “Identité culturelle
des sciences et des philosophies arabes: auteurs, œuvres et transmissions” organisé par la
Société Internationale d’Histoire des Sciences et des Philosophies Arabes et Islamiques
(S.I.H.S.P.A.I.) à Namur et Bruxelles du 15 au 18 janvier 2003. Mes remerciements vont
au Prof. Marwan Rashed pour m’avoir communiqué des remarques et indiqué un
complément de documentation qui m’ont permis de proposer une meilleure version. Je
remercie également Prof. Sabine Schmidtke pour m’avoir fourni une partie de cette
documentation.
doi:10.1017/S0957423912000033 © 2012 Cambridge University Press
AVICENNA AMONG MEDIEVAL JEWS

THE RECEPTION OF AVICENNA’S PHILOSOPHICAL,
SCIENTIFIC AND MEDICAL WRITINGS IN JEWISH
CULTURES, EAST AND WEST
GAD FREUDENTHAL
Université Paris Diderot, Sorbonne Paris Cité, SPHERE, UMR 7219, CNRS,
5 rue Thomas Mann, Bâtiment Condorcet, Case 7093,
F-75205 Paris Cedex 13, France
Email: gad.freudenthal@gmail.com
MAURO ZONTA
“Sapienza” Università di Roma, Dipartimento di Filosofia,
Via Carlo Fea 2, 00161 Roma, Italy
Email: maurozonta@libero.it
Abstract. The reception of Avicenna by medieval Jewish readers presents an

underappreciated enigma. Despite the philosophical and scientific stature of Avicenna,
his philosophical writings were relatively little studied in Jewish milieus, be it in
Arabic or in Hebrew. In particular, Avicenna’s philosophical writings are not among
the “Hebräische Übersetzungen des Mittelalters” – only very few of them were translated
into Hebrew. As an author associated with a definite corpus of writings, Avicenna hardly
existed in Jewish philosophy in Hebrew (contrary to Averroes). Paradoxically, however,
some of Avicenna’s most distinctive ideas were widely known and embraced by Jewish
philosophers. This is the phenomenon that we dub Avicennian knowledge without
Avicenna. In contrast with the philosophical treatises, Avicenna’s medical writings
were widely and intensively studied by Jews, especially in Hebrew, and remained influ-
ential until at least the seventeenth century. The present article presents a comprehen-
sive picture of Avicenna’s reception within medieval Jewish cultures in both Arabic and
Hebrew and tries to explain the Jews’ complex attitude to Avicenna.
Résumé. La réception d’Avicenne par les érudits juifs médiévaux présente une énigme
dont on n’a pas encore pris toute la mesure. Malgré la grande stature scientifique et phi-
losophique d’Avicenne, ses écrits philosophiques ont été peu connus des savants juifs,
que ce soit en arabe ou en hébreu. Ils n’ont guère fait partie des “Hebräische
Übersetzungen des Mittelalters” – peu seulement ont été traduits en hébreu. En tant
qu’auteur associé à un corpus de textes, Avicenne n’existe presque pas dans la philoso-
phie juive en hébreu (contrairement à Averroès). Paradoxalement cependant, certaines
des idées les plus caractéristiques d’Avicenne étaient bien connues et acceptées par
des philosophes juifs. Nous appelons ce phénomène savoir avicennien sans Avicenne.
Contrairement aux écrits philosophiques, les ouvrages médicaux d’Avicenne, eux,
étaient lus et utilisés par les juifs, notamment en traductions hébraïques, et ce
218 GAD FREUDENTHAL AND MAURO ZONTA
e
jusqu’au XVII siècle. Cet article présente un tableau général de la réception
d’Avicenne, en arabe et en hébreu, dans les différentes cultures juives et il tente d’ex-
pliquer l’attitude complexe des savants juifs vis-à-vis d’Avicenne.
The reception of Avicenna by medieval Jewish readers – philosophers,

scientists, and physicians – presents an underappreciated enigma.
Despite his philosophical and scientific stature, and the fact that
Avicenna inaugurated a “golden age” of Arabic science and philos-
ophy,1 his philosophical writings were relatively little known in
Jewish milieus, arabophone or other. The direct familiarity with
Avicenna’s works is strikingly less than that with Averroes, the
Commentator par excellence. However, and perhaps paradoxically,
despite the slight diffusion of Avicenna’s philosophical oeuvre
among Jews, some of his ideas were widely known and discussed in
scholarly Jewish circles. In contrast with the philosophical treatises,
Avicenna’s medical writings had a broad distribution among Jews,
especially in Hebrew, and remained influential until at least the
seventeenth century. The present article presents a preliminary but
comprehensive picture of Avicenna’s reception by medieval Jewish
cultures and tries to explain the Jews’ complex attitude to his work
in a comparative perspective.2
An important methodological point should be stated from the out-
set. When discussing the reception of a philosopher like Avicenna
by later generations of scholars, three different things need to be
clearly distinguished. The first is simple acquaintance with that phi-
losopher’s works: this acquaintance is attested to (in our case) by
1
Y. Tzvi Langermann (ed.), Avicenna and his Legacy. A Golden Age of Science and
Philosophy (Turnhout, 2009). See also Dimitri Gutas, “The heritage of Avicenna: The
golden age of Arabic philosophy, 1000–ca. 1350,” in Jules Janssens and Daniel De Smet
(eds.), Avicenna and his Heritage (Leuven, 2002), pp. 81–97.
2
Much valuable material on Avicenna in Hebrew can be found in Moritz Steinschneider, Die
Hebraeischen Übersetzungen des Mittelalters und die Juden als Dolmetscher (Berlin, 1893;
repr. Graz, 1956), hereafter: HÜ. Only partial overviews of the subject have been attempted
to date. See the useful but brief account: Shlomo Pines and Bernard Suler, “Avicenna,” in
Encyclopaedia Judaica (Jerusalem, 1971), 3, pp. 955–9. See also: Mauro Zonta, “Avicenna
in medieval Jewish philosophy,” in Janssens and De Smet (eds.), Avicenna and his
Heritage, pp. 267–79 (a slightly different version of this article appeared as: “The role of
Avicenna and of Islamic ‘Avicennism’ in the 14th-century Jewish debate around philosophy
and religion,” Oriente Moderno, 59, n.s. 19 [2000]: 647–60); Steven Harvey, “Avicenna’s
influence on Jewish thought: some reflections,” in Langermann (ed.), Avicenna and his
Legacy, pp. 327–40; id., “Editor’s introduction: Avicenna – his thought and influence,” in
id., Anthology of Writings by Avicenna (Tel-Aviv, 2009), pp. 11–34, on pp. 23–32 (Heb.).
See also the sections on Avicenna in Mauro Zonta, “Linee del pensiero ebraico nella storia
della filosofia ebraica medievale,” Annali dell’Istituto Universitario Orientale di Napoli, 57
(1997): 101–44, 450–83, on pp. 450–63, and in id., “Influence of Arabic and Islamic philos-
ophy on Judaic thought,” in Edward N. Zalta (ed.), Stanford Encyclopedia of Philosophy
(plato.stanford.edu/entries/arabic-islamic-judaic/). Some still useful material can also be
found in Steinschneider’s three articles entitled “Anzeigen,” in Hebräische Bibliographie,
10 (1870), no. 55, pp. 16–23, no. 56, pp. 53–9, no. 57, pp. 72–8.
AVICENNA AMONG MEDIEVAL JEWS 219
explicit references to Avicenna, by quotations from his works, or, more
subtly, by the presence of distinctively Avicennian ideas. Here we
refer only to the information on Avicenna’s thought that Jewish philo-
sophers had at their disposal, but without judging whether a given
thinker also adhered to any Avicennian views. The second level of
reception includes, in addition, a commitment to at least some of his
ideas: the writer not only signals his familiarity with (in our case)
Avicenna’s ideas, but also accepts some of them as true and integrates
them into his own thought. On a third level, his reading of Avicenna
may have stimulated a thinker to ponder certain ideas or inquiries
and to react to Avicenna’s positions, without necessarily accepting
any of them. In what follows, we will focus mainly on the historical
questions of the dissemination of and acquaintance with Avicennian
texts and ideas. To the extent possible, however, we will comment on
whether a given author who evinces familiarity with Avicenna also
accepted some of his ideas or drew on them in developing his own
views. We will also distinguish between direct reception (reading
Avicenna’s own works), and indirect reception (becoming acquainted
with Avicenna through another author’s work; for example, via
al-Ghazālī’s Maqāsid al-falāsifa [Intentions of the Philosophers], in
its original Arabic ˙version or in one of its three Hebrew translations).
We will advisedly avoid using the popular term “influence”: in line
with its origin in astrology, this term construes thinkers of the past
as passive, their minds having been affected by “influences” emanating
from other thinkers. We prefer to think of past philosophers as active
agents, who thoughtfully and carefully decided to embrace certain of
the ideas they encountered, rather than others.
As is well known, medieval Jews came into contact with Greco-Arabic
science and philosophy in two phases: first in Arabic (beginning in the
late ninth century), in both East and West; and later in Hebrew (start-
ing towards the middle of the twelfth century), in southern Europe
(Christian Spain, southern France, and Italy). We will therefore have
to consider the attitude to and reception of Avicenna in these two dis-
tinct cultural spheres separately. We will include all of Avicenna’s
oeuvre in our purview – in philosophy and science and in medicine.
We will also briefly mention treatises now known to be pseudonymous
but which are part of the story of the reception of Avicenna because at
least some medievals held them to be authentic.
PART A: AVICENNA AMONG ARABOPHONE JEWS
A.0. Introduction
Jews in Islamic lands read Arabic and also wrote in Arabic (usually in
Hebrew characters). In principle, then, they had direct access to
Avicenna’s Arabic writings.3 Although we would expect the works

of arabophone Jewish scholars to evince a strong presence of
Avicenna, one of the leading Arabic philosophers, and of
post-Avicennian philosophies,4 this expectation is not borne out.
This will be shown by our review of the works of arabophone Jewish
writers in East (A.1.1) and West (A.1.2), followed by a consideration
of the literary evidence of Arabic manuscripts of Avicenna’s works
copied and used by Jews (A.2).
A.1.1. Avicenna among Arabophone Jews: The East

The reception of Avicenna in the East essentially boils down to that
by four arabophone Jewish thinkers: (i) Hibat Allāh ʿAlī ibn Malka
Abū al-Barakāt al-Baghdādī, (ii) Moses Maimonides, (iii) David II
Maimonides, and (iv) ʿIzz al-Dawla Saʿd ibn Mansūr Ibn al-Kammūna.
(i) As was shown by the late Shlomo Pines, ˙ Abū al-Barakāt
al-Baghdādī (c. 1080–after 1165), a Jew who converted to Islam late
in life, seriously engaged with Avicenna’s al-Shifāʾ in his Kitāb
al-Muʿtabar (“The book of what has been established by personal
reflection”), although he may also have adopted ideas from the
Isharāt wa-al-tanbihāt. Abū al-Barakāt’s magnum opus is a purely
philosophical work addressed to the philosophical community, in
which the author’s Jewish identity plays no role. This work includes
large excerpts from the Shifāʾ, which it follows closely; Abū
al-Barakāt’s innovations resulted from a critical dialogue with that
work.5 Thus Abū al-Barakāt’s psychology builds on Avicenna’s, from
which he borrows, in particular, the idea that an individual’s self-
awareness offers apodictic proof of the existence and activity of the
soul. Unlike Avicenna, though, Abū al-Barakāt does not try to inte-
grate this idea into Aristotelian psychology. In physics, Abū
al-Barakāt begins from Avicenna’s notion of “violent inclination”
(mayl qasrī) to explain the continued motion of projectiles. On the
other hand, rejecting Avicenna’s notion that motion would continue
indefinitely in the absence of an obstacle, Abū al-Barakāt holds that
the “violent inclination” is “consumed” by the violent motion, which
consequently terminates.6
3
As far as we know, there is no evidence of Jewish scholars reading any of the few Persian
texts by Avicenna, notably the Dānesh-Nāmeh.
4
On these trends see Langermann (ed.), Avicenna and his Legacy.
5
The book is divided into three main parts: logic, physics (including psychology), and meta-
physics. This division clearly corresponds to that of Avicenna’s Najāt.
6
Shlomo Pines, Studies in Abu’l-Barakāt al-Baghdādī: Physics and Metaphysics (= The
Collected Works of Shlomo Pines, vol. 1) (Jerusalem and Leiden, 1979); id.,
“Abū’l-Barakāt al-Baghdādī, Hibat Allah,” Dictionary of Scientific Biography, vol. 1
(New York, 1970), pp. 26–8; id., “Abu’l-Barakāt, Hibat Allah b. Malkā al-Baghdādī,”
Encyclopédie de l’Islam, 2nd edn, vol. 1, pp. 114–16.
Abū al-Barakāt’s second work is an Arabic translation of
Ecclesiastes and commentary on that book. This work, composed in
Judeo-Arabic and obviously addressed to a Jewish audience, has yet
to be edited and has hardly been studied. Its contents seem to be
quite unrelated to the doctrines presented in Kitāb al-Muʿtabar, and
hence to Avicenna.7 Abū al-Barakāt dictated this work to his student,
Isaac Ibn Ezra, the son of the famous polymath Abraham Ibn Ezra;
the amanuensis composed a panegyric on his master when the work
was completed in 1143.8
Despite his apostasy, there was no animosity to Abū al-Barakāt
among Jews;9 he played a role, albeit minor, in later Jewish thought.
The Kitāb al-Muʿtabar was copied over in Hebrew characters and used
by Samuel b. Eli, head of the Baghdad academy, in the controversy
over Maimonides’ position on resurrection.10 As noted by Y. Tzvi
Langermann, “al-Baghdādī’s works continued to be studied at the
yeshivah of Baghdad, then the centre of Jewish conservatism, into
the thirteenth century. Al-Baghdādī’s commentary to the Book of
Ecclesiastes continued to be copied at the same yeshivah, with full
acknowledgement of its authorship.”11 Abū al-Barakāt is thus a
thinker who was a Jew not only by birth but also (until his conversion)
by social identity and who engaged seriously with Avicenna. He left
his marks on both the Muslim and Jewish intellectual worlds.
(ii) Maimonides (1138–1204) was, like Avicenna himself, both a phi-
losopher and a physician. Accordingly we will consider Avicenna’s
presence in Maimonides’ philosophy and then in Maimonides’ medical
works. We assign Maimonides to the “East,” where he wrote the great
majority of his works; it should be kept in mind, however, that he was
7
Shlomo Pines, “A study of Abu’l-Barakāt’s commentary on the Ecclesiastes” (1964), repr. in
his Studies in the History of Jewish Philosophy. The Transmission of Texts and Ideas
(Jerusalem, 1977), pp. 68–83 (Heb.).
8
The poem is published in Menahem H. Schmelzer (ed.), Yitzhak ben Avraham Ibn Ezra:
Shirim (New York, 1980), pp. 44–5. On Isaac Ibn Ezra and his relationship with Abū
al-Barakāt see Jefim Schirman, The History of Hebrew Poetry in Christian Spain and
Southern France, edited, supplemented and annotated by Ezra Fleischer (Jerusalem,
1997), pp. 71–3 (Heb.). See also Pines, Studies in Abu’l-Barakāt al-Baghdādī, p. 101, n. 2.
9
On his conversion see the analysis in Sarah Stroumsa, “On intellectual converts to Islam in
the early Middle Ages,” Peʿamim, 42 (1990): 61–75, on pp. 66–8 (Heb.).
10
Sarah Stroumsa, “On the Maimonidean controversy in the East: the role of Abu’l-Barakat
al-Baghdadi,” in Haggai Ben-Shammai (ed.), Hebrew and Arabic Studies in Honour of
Joshua Blau (Tel-Aviv and Jerusalem, 1993), pp. 415–22 (Heb.); ead. (ed. and trans.),
The Beginnings of the Maimonidean Controversy in the East. Yosef Ibn Shimon’s
Silencing Epistle Concerning the Resurrection of the Dead. Arabic and Hebrew Texts of
the Risālat al-Iskāt fī hashr al-anwāt, with Introduction and Annotated Hebrew
˙
Translation (Jerusalem, 1999), §§ 46–47, 54–58, 146 of the texts and the discussions in
the notes pp. 135–41 (Heb.).
11
Y. Tzvi Langermann, “Al-Baghdadi, Abu’l-Barakat,” in http://www.muslimphilosophy.com/
ip/rep/J008.htm (accessed Oct. 10, 2010).
educated in Andalusia, whose intellectual climate reverberated in his

thought throughout his life.12
Whether or not Maimonides drew on Avicenna’s philosophy is a
question of great importance, given his momentous influence on
Jewish philosophy throughout the ages. The topic has been repeatedly
debated by scholars; what follows is a summary account of the “state
of the art.”13 To begin with, Maimonides himself informs us of his
knowledge and appreciation of Avicenna. In a celebrated letter
(dated 8 Tishrei 1511 Seleucid era [= AM 4960], i.e., September 30,
1199), addressed to Samuel Ibn Tibbon, the future Hebrew translator
of the Guide, he commented on the relative merits of several Arabic
philosophers. Avicenna is the last to be assessed. This passage of
the letter is extant in two slightly different Hebrew translations
(except for two short passages, the letter is not preserved in the
Arabic original):14
[Version 1] The books of ʿAlī Ibn [Version 2] The books of Ibn Sīnā,
Sīnā, although they evince fine although it is appropriate to take
carefulness [or thoroughness; diyyuq issue with them [le-haqshot
tov] and subtle inquiry [ʿiyyun daq], ʿaleihem] and [although] they are
˙are not like those of Abū Nasr not like al-Fārābī’s utterances –
˙
al-Fārābī. Nonetheless, his books are there is usefulness in his books and
useful and he, too, is an author one should study his utterances
whose utterances you should and probe into their [lit. its] ideas.
examine and into whose works you
should probe.
This short passage, in its two versions, has been diversely inter-
preted, but two conclusions seem to be warranted: first, Maimonides
was acquainted with works he took to be by Avicenna; second, he had
a fairly high opinion of Avicenna’s philosophy and deemed it worth
studying, although he did not esteem Avicenna as highly as al-Fārābī
and (according to Version [2]) he may have believed that some matters
were subject to reconsideration. With this in mind, let us consider the
presence of Avicenna in Maimonides’ philosophical works.
12
See infra, n. 31.
13
A full review of the discussion (with bibliography) is given in Mauro Zonta, “Maimonides’
knowledge of Avicenna. Some tentative conclusions about a debated question,” in Georges
Tamer (ed.), The Trias of Maimonides/Die Trias des Maimonides. Jewish, Arabic, and
Ancient Culture of Knowledge/Jüdische, arabische und antike Wissenskultur (Berlin
and New York, 2005), pp. 211–22.
14
Maimonides, “Letter to R. Shmuel Ibn Tibbon,” in Alexander Marx, “Texts by and about
Maimonides,” The Jewish Quarterly Review, 25 (1934–1935): 371–428, on p. 380; the full
text of one of the translations (with the extant fragments of the Arabic original) is also
in Letters and Essays of Moses Maimonides (Heb.), ed. Isaac Shailat (Maʿaleh Adumim,
5748 [1988]), 2, pp. 511–54, on pp. 553–4.
We begin with Maimonides’ The Guide of the Perplexed, written in
Fustāt and completed by 1191, in which Avicenna is never named.
˙ ˙
Nevertheless, medieval Jewish scholars, most notably the very eru-
dite arabophone Shem-Tov Ibn Falaquera (c. 1280), followed by
Moses Narboni (d. after˙ 1362), identified distinctively Avicennian
ideas in the Guide.15 More recently, Pines showed that although
Avicenna is never named in the Guide, a number of Avicennian doc-
trines are an integral part of its philosophy. Pines summarizes his
thesis, which also accounts for Maimonides’ slightly reserved assess-
ment of Avicenna, quoted above, as follows:
at a certain level of philosophic thought Avicenna had considerable influence
upon Maimonides; this is indicated by the latter’s adoption of negative theol-
ogy, of the distinction between essence and existence, and of various particu-
lars of Avicenna’s prophetology and theory of the worship of God. This
influence, however, [. . .] did not essentially modify Maimonides’ fundamental
position, which he inherited from al-Fārābī, and the Spanish Aristotelian
school.16
W. Z. Harvey has confirmed and deepened Pines’ suggestions in a
number of studies,17 as has H. A. Davidson.18 Dov Schwartz has
convincingly argued that Maimonides’ ideas about the afterlife, too,
follow Avicenna.19 Steven Harvey has suggested that Maimonides
borrowed some of his ideas on the notions of ʿishq and prayer from
15
In his commentary on the Guide, Falaquera repeatedly compares Maimonides’ statements
with Avicenna’s views; for an overview see Yair Shiffman, “Again on Avicenna and
Maimonides,” Tarbiz, 64 (1995): 523–34 (Heb.); see also n. 179 below. In his Commentary
˙
on the Guide, ed. Jacob Goldenthal (Vienna, 1852), Narboni writes that Maimonides “fol-
lowed the opinion of Avicenna and his statements” (p. 1r) and that “in my view the Master
borrows from Avicenna in al-Shifāʾ and al-Najāt, not from what is offered by Aristotle”
(p. 14v). Narboni’s statements must be treated with circumspection, however. For example,
he charged Maimonides with more than one plagiarism from Avicenna’s Najāt, but, as it
seems, without justification. See Gad Freudenthal, “Maimonides on the scope of metaphysics
alias Maʿaseh Merkavah: the evolution of his views,” in Carlos del Valle, Santiago
García-Jalón and Juan Pedro Monferrer (eds.), Maimónides y su época (Madrid, 2007), pp.
221–30; id. “Four observations on Maimonides’ four celestial globes (Guide 2:9–10),” in
Aviezer Ravitzky (ed.), Maimonides: Conservatism, Originality and Revolution (Jerusalem,
2008), pp. 499–527 [Heb.] (expanded version of the former item).
16
Moses Maimonides, The Guide of the Perplexed, translated by Shlomo Pines (Chicago,
1963), “Translator’s introduction,” p. cii.
17
Notably in “Why Maimonides was not a Mutakallim,” in Joel L. Kraemer (ed.), Perspectives
on Maimonides. Philosophical and Historical Studies (Oxford, 1991), pp. 105–14;
“Maimonides’ Avicennism,” Maimonidean Studies, 5 (2008): 107–19.
18
Herbert A. Davidson, Alfarabi, Avicenna, and Averroes on Intellect: Their Cosmologies,
Theories of the Active intellect, and Theories of the Human Intellect (New York, 1992),
pp. 197–207.
19
Dov Schwartz, “Avicenna and Maimonides on immortality: a comparative study,” in Ronald
L. Nettler (ed.), Medieval and Modern Perspectives on Muslim-Jewish Relationships
(Oxford, 1995), pp. 185–97. This position is endorsed in Sarah Stroumsa, Maimonides in
his World. Portrait of a Mediterranean Thinker (Princeton, 2009), pp. 154–61.
Avicenna.20 In sum, then, as Davidson put it, “despite Maimonides’

judging him [Avicenna] inferior to Alfarabi, he contributed consider-
ably to the formation of Maimonides’ philosophic thinking: What
Maimonides calls Aristotle’s Metaphysics is, we saw, not
Aristotelian but Avicennian, and Maimonides’ picture of the structure
of the universe, while it draws from both Alfarabi and Avicenna, owes
more to the latter than to the former, though Maimonides was uncon-
scious of the debt.”21 There thus is no question that the major philoso-
phical work of the most influential Jewish philosopher of all time
incorporated fundamental ideas going back to Avicenna (while main-
taining more traditional Aristotelian positions on other matters).
Maimonides also expounded some of these Avicennian ideas in
the short metaphysical section of Sefer ha-Maddaʿ (The Book of
Knowledge), the first part of his code of Jewish law (Mishneh
Torah), written in Hebrew.
The ideas Maimonides borrowed from Avicenna are relatively inde-
pendent of one another: Maimonides did not take over a systematical
and coherent doctrine that would have compelled him to follow
his predecessor’s positions consistently. This is precisely why
Maimonides could follow Avicenna on some issues but maintain
more traditional Aristotelian positions on others. To describe this
situation, which we will repeatedly encounter in what follows, we
introduce the term “doctrinal item”: using it, we can say that
Maimonides appropriated a number of Avicennian doctrinal items,
i.e. discrete, unconnected, Avicennian ideas, which he integrated in
his thought individually.
Maimonides’ works, both the Guide and even more so The Book of
Knowledge, have been widely read by Jews in many cultural contexts
over the generations. As a result, Avicenna’s ideas (partly simplified)
had a particularly wide dissemination: all students of Jewish philos-
ophy from the early thirteenth century onward were familiar with
them, although without the help of suitable commentaries they
were unable to know that these ideas derived from Avicenna.22
Another question is how Maimonides became acquainted with
Avicenna’s doctrines. Davidson has drawn attention to the fact
that in the Guide, whenever Maimonides refers to “Aristotle’s
Metaphysics” he in fact has in mind Avicennian doctrines: “What he
20
Harvey, “Editor’s introduction,” pp. 27–8, referring to two as yet unpublished articles.
21
Herbert A. Davidson, Moses Maimonides: The Man and his Works (New York, 2004), p. 115.
Similarly: “[Avicenna’s] influence on him [Maimonides] was considerable, but he was not
conscious of it because of his failure to distinguish Avicenna’s views from Aristotle’s”
(p. 121). Similarly in id., “Maimonides, Aristotle, and Avicenna,” in Régis Morelon and
Ahmad Hasnawi (eds.), De Zénon d’Élée à Poincaré. Recueil d’études en hommage à
Roshdi Rashed (Louvain and Paris, 2004), pp. 719–34
22
See below, p. 252.
[Maimonides] calls Aristotle’s Metaphysics is Avicenna’s metaphys-
ical philosophy, which he confused with Aristotle’s or at best
Aristotle’s metaphysics as refracted through an Avicennian
prism.”23 But how did Maimonides become acquainted with these
ideas? Davidson has forcefully argued that Maimonides did not
know Avicenna first-hand: had he actually read a text by Avicenna
he would have been aware of the difference between Aristotle’s and
Avicenna’s doctrines and would not have ascribed Avicennian views
to Aristotle. But if Maimonides did not learn Avicennian ideas from
a work by their author, what was his source?
“Virtually everything of metaphysical character attributed by
Maimonides to Aristotle but actually deriving from Avicenna can be
found in Ghazālī’s summary of Avicenna’s philosophy, the Maqāsid
˙
al-falāsifa,” Davidson states.24 In a recent article, one of the present
authors confirmed this thesis: a detailed examination of the evidence
gathered by scholars in the past decades suggests that Maimonides
was not directly acquainted with any of Avicenna’s philosophical
works and that he obtained his knowledge of Avicenna from
al-Ghazālī’s Maqāsid.25 Al-Ghazālī, who expounded philosophy alleg-
edly in order to ˙ refute it,26 thus emerges as a crucial conduit
of Avicennian ideas to Maimonides and through him to Jewish
thought – an apparent paradox that we will encounter again in
what follows. (Note that Maimonides does not even mention
al-Ghazālī in his letter to Ibn Tibbon.) Nonetheless, the possibility –
maintained by some recent authors – that Maimonides read
Avicenna’s own al-Najāt (The Salvation) cannot be completely ruled
out.27 In fact, both al-Najāt and the Maqāsid, which is based on
Avicenna’s Dānesh-Nāmeh,28 summarize Avicenna’s˙ ideas and are
accordingly similar: since Maimonides’ discussions are quite general
23
Davidson, Moses Maimonides, p. 105; see also pp. 102–4, 364. On the relationship of
Avicenna’s metaphysics to Aristotle’s, see Amos Bertolacci, The Reception of Aristotle’s
Metaphysics in Avicenna’s Kitāb al-Šifāʾ: A Milestone of Western Metaphysical Thought
(Leiden, 2006).
24
Davidson, Moses Maimonides, p. 104, n. 146. Davidson continues: “There are, moreover,
striking similarities between what Maimonides writes, for example, in Guide for the
Perplexed 2.4, and Ghazālī’s Maqāsid al-falāsifa, pp. 209–21.” See also pp. 115, 121. An
analogy used by Maimonides in his ˙ commentary on the Mishnah which goes back (in
part) to Avicenna’s writings, is also traceable to al-Ghazālī (Davidson, Moses
Maimonides, pp. 94–5).
25
Zonta, “Maimonides’ knowledge of Avicenna.”
26
See below n. 100 on modern appreciations of al-Ghazālī’s sincerely.
27
Schwartz, “Avicenna and Maimonides on immortality”; Amira Eran, “Al-Ghazali and
Maimonides on the world to come and spiritual pleasures,” Jewish Studies Quarterly, 8
(2001): 137–66.
28
On the relationship between al-Ghazālī’s Maqāsid and Avicenna’s Dānesh-Nāmeh, see
Jules Janssens, “The Dānesh-Nāmeh d’Ibn Sina:˙ un texte à revoir?,” Bulletin de philoso-
phie médiévale, 28 (1986): 163–77, repr. in id., Ibn Sīnā and his Influence on the Arabic
and Latin World (Aldershot, 2006), Essay VII.
and devoid of details, it may be impossible to decide the issue defi-

nitely. In any event, Maimonides does not seem to have known
Avicenna’s major philosophical work, al-Shifāʾ.
But why should Maimonides ascribe to Aristotle the doctrines he
apparently appropriated from al-Ghazālī’s Maqāsid (or, for that
matter, Avicenna’s al-Najāt)? We conjecture that ˙many readers of
Ghazālī’s Maqāsid thought of it as a convenient summary of the “stan-
dard position,” ˙as it were, of the worldview of the falāsifa: it was not
taken to present a specific Avicennian doctrine, but the typical,
“mean” stance shared by adherents of the philosophical school (see
below, p. 248). Since in Andalusia – to whose intellectual preferences
Maimonides largely subscribed – Aristotle continued to be viewed as
the leader and fountainhead of the philosophical mode of thought,29
Maimonides (and his milieu) may have associated the doctrines
found in the Maqāsid with Aristotle’s emblematic name.30 This may
be why Maimonides ˙ attributed typically Avicennian positions to
Aristotle. A last mystery remains: seeing that Maimonides was prob-
ably not acquainted with Avicenna’s own works (especially not with
Avicenna’s major philosophical work), and, moreover, that he ascribed
the Avicennian doctrines that he apparently knew from al-Ghazālī to
“Aristotle,” what was the basis for his evaluation of Avicenna’s philos-
ophy in the letter to Samuel Ibn Tibbon? Our surmise is that
Maimonides’ vague remark (it mentions no title) may have simply
expressed one of the mashhurāt – generally accepted ideas – current
and accepted in his basically Andalusian cultural milieu.31
Let us now consider Avicenna’s presence in other works by
Maimonides. Maimonides’ only explicit reference to a work of
29
Gutas, “The heritage of Avicenna,” p. 90.
30
This hypothesis is consistent with Charles Burnett’s observation that al-Ghazālī was read
principally among Arabic (Muslim and Jewish) theologians; see his “The coherence of the
Arabic-Latin translation program in Toledo in the twelfth century,” Science in Context, 14
(2001): 249–88, on p. 265.
31
On the lasting sway of Maimonides’ Andalusian cultural context, see Joshua Blau,
“Maimonides’ ‘At our place in al-Andalus’ revisited,” in del Valle, García-Jalón, and
Monferrer (eds.), Maimónides y su época, pp. 327–40. This is the full text of a paper
whose abstract had previously been published under a very similar title: “‘At our place
in al-Andalus’, ‘At our place in the Maghreb’,” in Kraemer (ed.), Perspectives on
Maimonides, pp. 293–4. On the negative attitude to Avicenna in al-Andalus see e.g.
Gutas, “The heritage of Avicenna,” pp. 90–1, 97. Abdelhamid I. Sabra, followed by
Y. Tzvi Langermann, has argued for the existence of an “Andalusian self-assertiveness
vis-à-vis the rest of the Islamic world” and suggested that scholars in Andalusia sought
to create an alternative to the science and philosophy of the East in all domains of learning;
see, respectively, “The Andalusian revolt against Ptolemaic astronomy. Averroes and
al-Bitrūjī,” in Everett Mendelsohn (ed.), Transformation and Tradition in the Sciences.
Essays˙ in Honor of I. Bernard Cohen (Cambridge, 1984), pp. 133–53; “Another
Andalusian revolt? Ibn Rushd’s critique of the pharmacological computus of al-Kindi,” in
Abdelhamid I. Sabra and Jan P. Hogendijk (eds.), The Enterprise of Science in Islam
(Cambridge, 2003), pp. 351–72, esp. pp. 366–8.
Avicenna is found in his Treatise on Resurrection of the Dead (1191/2),
in which he affirms that some ideas of his rival (Samuel ben ʿEli)
“derive from Avicenna’s al-Maʿād.”32 It has proved impossible to
trace the ideas in question in any of Avicenna’s works whose title
includes the word maʿād.33 Remarkably, Maimonides’ favorite stu-
dent, Joseph Ibn Shimʿon, in the treatise he composed in c. 1192 to
defend his Master and “silence” Samuel ben ʿEli, drew on Avicenna,
albeit silently. He incorporated quotations from Avicenna’s Kitāb
al-Nafs to the effect that the soul’s persistence after death is eternal,
rather than of limited duration.34 It has recently been suggested that
Maimonides’ erroneous statement that Samuel’s treatise drew on
Avicenna was intentional and was concealed criticism of his student’s
(silent) reliance on Avicenna.35
We can tentatively conclude that although Maimonides derived
essential building blocks of his philosophy from Avicenna, he appar-
ently did not have direct access to this philosopher’s works (except,
perhaps, al-Najāt). For unknown reasons, he contented himself, as
it seems, with the “digest” of Avicenna’s doctrines offered by
al-Ghazālī. The main point, however, is not how Maimonides came
into contact with Avicenna’s thought, but the fact that he embraced
some Avicennian ideas. Maimonides’ appropriation of significant
Avicennian doctrinal items, albeit not derived from Avicenna’s own
writings, is an example of a phenomenon that we dub Avicennian
knowledge without Avicenna. We will repeatedly encounter it later.
The reception of Avicenna in Maimonides’ medical works presents
another enigma. Although Avicenna was doubtless by far the most
famous Arabic doctor in East and West throughout the Middle Ages,
Maimonides nowhere refers to his magnum opus, al-Qānūn fī
al-tibb (The Canon). Maimonides’ only explicit reference to a medical
˙ by Avicenna seems to be in his On the Causes of Symptoms,
work
where he quotes prescriptions that turn out to derive from
Avicenna’s al-Adwiya al-qalbiyya (The Cardiac Drugs).36 Whether
Maimonides silently borrowed from Avicenna’s medical magnum
opus requires further study.
32
Maimonides, “Treatise on Resurrection,” in Y. Shailat, ed., Letters and Essays of Moses
Maimonides, pp. 319–38 (Arabic text), on p. 325.15; in Samuel Ibn Tibbon’s Hebrew trans-
lation: Maʾamar ha-gemul (ibid., pp. 339–74 on p. 351.13).
33
The source of Maimonides’ statement was already discussed in detail in Moritz
Steinschneider, Al-Farabi (Alpharabius). Des arabischen Philosophen Leben und
Schriften mit besonderer Rücksicht auf die Geschichte der Griechichen Wissenschaft
unter den Arabern (St. Petersburg, 1869; repr. Amsterdam, 1966), pp. 35–7 (n. 44). See
also Davidson, Moses Maimonides, p. 528, n. 194.
34
Stroumsa, The Beginnings of the Maimonidean Controversy in the East, p. 139 (note to §§
51–52) (Hebrew).
35
Stroumsa, Maimonides in his World, p. 174.
36
Zonta, “Maimonides’ knowledge of Avicenna,” p. 220.
Maimonides’ attitude towards Avicenna thus is a complex riddle.

Although his philosophy certainly incorporates essential Avicennian
ideas, Maimonides apparently did not read any work by him (or at
least no major work). Maimonides rarely mentions Avicenna’s
name, except, paradoxically, to comment on the quality of his works
in the letter to Samuel Ibn Tibbon. In medicine, too, where
Avicenna had been the medical authority par excellence for more
than a century, Maimonides almost totally ignored him. Why
Maimonides did not read Avicenna himself, especially the Shifāʾ
and the Canon, remains a mystery. The cause is probably more pro-
found than the unavailability of manuscripts. Perhaps it was a conse-
quence of the already noted Andalusian disdain for Avicenna. But,
again, the uncertainties surrounding the reasons for Maimonides’
attitude to Avicenna should not lead us to lose sight of the historically
important fact: via Maimonides’ Guide and Book of Knowledge,
characteristically Avicennian ideas reached an immense Jewish read-
ership over many centuries; Maimonides’ readers appropriated these
ideas, even though they were rarely aware of their source.
The phenomenon of Avicennian knowledge without Avicenna had
momentous historical repercussions. Because Maimonides in the
Guide identified ideas that we know to be Avicennian (such as the
Necessary Existent) as deriving from Aristotle, Jewish readers who
encountered them in other writings by various authors (for example,
in al-Ghazālī’s Maqāsid) were not liable to think of them as specifi-
cally Avicennian. In˙ other words, through Maimonides’ Guide
Jewish readers became acquainted with and embraced key ideas of
Avicenna’s, which they took to be Aristotle’s. One result was to blur
the opposition between Avicennism and orthodox Aristotelianism
(as represented by Averroes) for many Jewish readers, especially
those without access to Arabic literature. This may help explain the
lack of any consolidated Avicennian movement in Jewish philosophy
and the almost total absence of Hebrew translations of Avicenna’s
works (more on this below).
(iii) Maimonides’ descendents can hardly be described as continuing
their ancestor’s rationalist philosophy, but some of them nonetheless
drew on Avicenna. Both his son Abraham (1186–1237) and especially
his great-great-great-grandson, David II Maimonides (1335–c. 1415),
were interested in the transmigration of souls (apparently an issue in
their respective milieus) and wrote to refute metempsychosis. As Paul
B. Fenton has shown, their rationalist arguments drew on a passage
of Avicenna’s Najāt.37 A Hebrew translation of David Maimonides’
37
Paul B. Fenton, “New light on Maimonidean writings on metempsychosis and the influence
of Avicenna,” in Langermann (ed.), Avicenna and his Legacy, pp. 341–68. This study super-
sedes the author’s conclusions in the earlier “The literary legacy of David ben Joshua, last
text was one of the sources used in a celebrated discussion of transmi-
gration by two Jewish scholars, which took place in Candia in 1468.38
(iv) Finally, Ibn Kammūna (d. 1284), too, needs to be considered
here briefly. “Briefly” only, because while he was both Jewish and a
philosopher, he wrote exclusively as a philosopher for philosophers,
never as a Jew on Jewish philosophy (in this he differs from Abū
al-Barakāt al-Baghdādī discussed above). Ibn Kammūna wrote in
Arabic, not Judeo-Arabic, for the entire Arabic philosophical commu-
nity of his time; hence he has no direct bearing on our overview.
Ibn Kammūna received a thorough Jewish education, as his knowl-
edge of Hebrew and familiarity with rabbinic literature show.39
Notwithstanding his commitment to philosophy, late in life he pur-
sued his interest in Jewish subjects and composed a treatise on the
differences between the Rabbanites and the Karaites.40 In his cele-
brated Examination of the Inquiries into the Three Faiths (1280)
he draws on two Judeo-Arabic writers (Judah Hallevi and
Maimonides).41 Ibn Kammūna’s intellectual horizons thus included
some writings on Jewish religious philosophy.
Avicenna looms large in Ibn Kammūna’s thought. Pourjavaday and
Schmidke observe: “As a philosopher, Ibn Kammūna was a follower of
Ibn Sīnā. With some modifications, his cosmology is Avicennian. His
distinction between the One as the Necessary Existent (wājib
al-wujūd) and other existents as contingents (mumkināt) is also
Avicennian, as is his notion of prophecy, explained in detail in the
first chapter of his Tanqih. [. . .] The main sources of his psychology
˙
are again Ibn Sīnā’s writings.” 42 Ibn Kammūna wrote a commentary
on Avicenna’s Ishārat wa-al-tanbihāt;43 Langermann has pointed out

the importance of Avicenna’s concept of hads for Ibn Kammūna.44 On
the other hand, Ibn Kammūna parted ˙ company from Avicenna
with regard to the eternity a parte ante of the human rational
of the Maimonidean Negidim,” The Jewish Quarterly Review, 75 (1984–1985): 1–56; see
however p. 34 for texts by Avicenna used by R. David Maimonides.
38
Fenton, “New light,” pp. 364–8.
39
Reza Pourjavaday and Sabine Schmidke, A Jewish Philosopher of Baghdad. ʿIzz al-Dawla
Ibn Kammūna (d. 683/ 1284) and his Writings (Leiden, 2006), pp. 8–9.
40
Leon Nemoy, “Ibn Kammunah’s Treatise on the differences between the Rabbanites and
the Karaites,” Jewish Quarterly Review, NS 63 (1972–73): 97–135, 222–46.
41
Moshe Perelman (trans.), Ibn Kammūna’s Examination of the Three Faiths (Berkley, 1971),
“Introduction,” pp. 1–9.
42
Pourjavaday and Schmidke, A Jewish Philosopher of Baghdad, p. 23; Lukas Muehlethaler,
“Ibn Kammūna (d. 683/1284) on the argument of the Flying Man in Avicenna’s Ishārāt and
Suhrawardī’s Talwīhāt,” in Langermann (ed.), Avicenna and his Legacy, pp. 179–203. On
Avicenna’s presence˙ in the Examination, see Perelman, Ibn Kammūna’s Examination of
the Three Faiths, pp. 4, 9, 28 (n. 21).
43
Pourjavaday and Schmidke, A Jewish Philosopher of Baghdad, pp. 59–63; see also p. 129.
44
Y. Tzvi Langermann, “Ibn Kammūna and the ‘New Wisdom’ of the thirteenth century,”
Arabic Sciences and Philosophy, 15 (2005): 277–327.
soul,45 although his interest in soul itself “can be understood only in

the context of the Avicennian legacy.”46
Ibn Kammūna was read by only a few arabophone Jewish scholars
in the East: he is cited by David Maimonides (just mentioned); a num-
ber of his writings were copied in Hebrew characters.47 It seems unli-
kely that any of his writings were ever translated into Hebrew.
A.1.2. Avicenna Among Arabophone Jews: The West

The cultural language of the Jews in Muslim Spain, especially
Andalusia, was Arabic. So they, too, had direct access to works in
Arabic, which deeply informed their culture, including their works com-
posed in Hebrew (notably poetry). Some arabophone Jews continued to
write scientific and philosophical works in Arabic even after the rise of
Hebrew-language Jewish philosophy in northern Spain and the Midi.
Between the twelfth and fourteenth centuries we find no fewer than
four arabophone Jewish authors who were familiar with and drew on
Avicenna’s works: (i) Judah Hallevi, (ii) Abraham Ibn Daʾud, (iii)
Moses ha-Levi, and (iv) Joseph Ibn Waqār.48 Their use of Avicenna in
their Judeo-Arabic writings is important in its own right; so too is the
fact that, via Hebrew translations, these works contributed to the
transmission of Avicennian ideas to the Hebrew philosophical culture.
(i) As early as 1876, S. Landauer recognized that Judah Hallevi
(c. 1080–1141) incorporated entire sections of Avicenna’s Risāla fī
al-Nafs (Compendium [lit. Epistle] on the Soul) into his magnum
opus, the Kuzari (completed 1140).49 Harry A. Wolfson and especially
David H. Baneth confirmed this thesis through a detailed comparison
of Hallevi’s and Avicenna’s texts.50 Hallevi’s dependence on Avicenna
establishes that at least the Risāla fī al-Nafs was available in
45
Pourjavaday and Schmidke, A Jewish Philosopher of Baghdad, p. 24; Lukas Muehlethaler,
“Ibn Kammūna (d. 683/1284) on the eternity of the human soul: The three treatises on the
soul and related texts,” Ph.D. Dissertation, Yale University, 2010.
46
Y. Tzvi Langermann, “Ibn Kammuna,” in Edward N. Zalta (ed.), The Stanford
Encyclopedia of Philosophy (Fall 2008 Edition), URL = http://plato.stanford.edu/archives/
fall2008/entries/ibn-kammuna/ (last accessed Oct. 27, 2010).
47
Pourjavaday and Schmidke, A Jewish Philosopher of Baghdad, pp. 54–7.
48
For other arabophone thinkers who may have used Avicenna, see Harvey, “Editor’s
Introduction,” p. 25.
49
Samuel Landauer, “Die Psychologie des Ibn Sīnā,” Zeitschrift der Deutschen
Morgenländischen Geselleschaft, 29 (1876): 335–418, on pp. 335–6.
50
See Harry A. Wolfson, “The internal senses in Latin, Arabic and Hebrew philosophic texts,”
Harvard Theological Review, 28 (1935): 69–133; Judah Hallevi, Kitāb al-Radd wa-al-dalīl
fī al-dīn al-dhalīl (al-Kitāb al-Khazarī), ed. David H. Baneth and Haggai Ben-Shammai
(Jerusalem, 1977), “Introduction,” p. 10 and the notes to the text, pp. 200–8. In Book V,
Chapter 12 of his work, Hallevi describes the five external senses that correspond precisely
with what Avicenna calls the “five internal senses” of the human mind: the common sense,
the “formative faculty” (which preserves the external forms of the sensible objects), the
memory, the imagination (which recalls what has not been preserved in it), and the “esti-
mative faculty.”
Andalusia before 1140.51 In other parts of the Kuzari, too, Hallevi bor-
rows various doctrinal items from Avicenna (and al-Fārābī);52 but he
never names his philosophical forebear.53
(ii) In the case of Abraham Ibn Daʾud (1110–c. 1180), the acquain-
tance with and borrowing from Avicenna is no less clear. Ibn Daʾud
wrote his The Exalted Faith in Toledo (then the capital of Castile)
in 1160–1161. The Arabic original of the book (al-ʿAqīda al-rafīʿa) is
lost, but two late-fourteenth-century Hebrew translations survive.
Once again, Avicenna’s name is not mentioned, but Ibn Daʾud’s direct
or indirect borrowings from Avicenna have been noted by scholars,
notably J. Guttmann, R. Fontaine, and H. A. Davidson.54 More
recently, A. Eran has detailed Ibn Daʾud’s direct dependence on
Avicenna’s Najāt.55 For example, in Book 1, Chapter 6, Ibn Daʾud’s
discussion of the five “internal senses” follows al-Najāt, which on
this topic deviates from the Risāla fī al-Nafs (excluding a dependence
of Ibn Daʾud on Judah Hallevi).56 Moreover, a general correspondence
between the order of the discussions in the first part of The Exalted
Faith – on logic, physics, and metaphysics – and the corresponding
part of al-Shifāʾ has been pointed out.57
More than half a century ago, M.-T. d’Alverny suggested that
Abraham Ibn Daʾud is to be identified with the noted Jewish translator
Avendauth, who, between 1154 and 1166, cooperated with the Christian
philosopher and translator Domingo Gundisalvi in producing a complete
Latin version of the section On the Soul of Avicenna’s Shifāʾ.58
51
Pines has argued that Avicenna’s writings reached Judah Hallevi in the time between the
composition of Book I and Book V of the Kuzari. See Shlomo Pines, “Shiʿite terms and con-
ceptions in Judah Halevi’s Kuzari,” Jerusalem Studies in Arabic and Islam, 2 (1980): 165–
251 on pp. 210–19; repr. with original pagination indicated in The Collected Works of
Shlomo Pines, vol. V: Studies in the History of Jewish Thought, ed. by Warren Z. Harvey
and Moshe Idel (Jerusalem, 1997).
52
Davidson, Alfarabi, Avicenna, and Averroes on Intellect, pp. 180–95.
53
See on this p. 252, n. 142 below.
54
Jacob Guttmann, Die Religionsphilosophie des Abraham ibn Daʾud aus Toledo. Ein Beitrag
zur Geschichte der jüdischen Religionsphilosophie und der Philosophie der Araber
(Göttingen, 1879); T.A.M. (Resianne) Fontaine, In Defence of Judaism. Abraham ibn
Daʾud. Sources and Structures of ha-Emunah ha-Ramah (Assen, 1990), see index s.v.
al-Ghazali, Ibn Sina; Davidson, Alfarabi, Avicenna, and Averroes on Intellect, pp. 195–7.
55
See Amira Eran, From Simple Faith to Sublime Faith. Ibn Daud’s Pre-Maimonidean
Thought (Tel Aviv, 1998), esp. p. 27 (Heb.).
56
See this passage in Abraham Ibn Daʾud, Das Buch Emunah Ramah, oder: der Erhabene
Glaube verfasst von Abraham Ben David Halevi aus Toledo (im Jahr 1160), ed. and
trans. Shimshon Weil (Frankfurt, 1852), pp. 20–30.
57
See Zonta, “Linee del pensiero islamico,” pp. 453–4.
58
Marie-Thérèse d’Alverny, “Avendauth?,” in Homenaje a Millás-Vallicrosa, 2 vols.
(Barcelona, 1954), vol. I, pp. 20–43; Burnett, “The coherence of the Arabic-Latin translation
program,” pp. 251–2, 264; Alexander Fidora, “Ein philosophischer Dialog der Religionen im
Toledo des 12. Jahrhunderts: Abraham Ibn Daud und Dominicus Gundissalinus,” in Yossef
Schwartz, Volkhard Krech (eds.), Religious Apologetics – Philosophical Argumentation
(Tübingen, 2004), pp. 251–66.
Avendauth himself rendered the introduction and some chapters of the

logical section of al-Shifāʾ into Latin. If (as it seems) d’Alverny was cor-
rect, then Ibn Daʾud alias Avendauth was indeed familiar with a large
corpus of genuine Avicennian texts. We then face a remarkable and
interesting situation, inasmuch as Avendauth the translator’s attitude
towards Avicenna is different from that of Abraham Ibn Daʾud the
Jewish theologian: the former, far from being a mere translator, played
an active role in initiating the Latin translations of Avicenna; the latter,
by contrast, concealed his borrowings from Avicenna and never men-
tioned Avicenna by name. Unless Ibn Daʾud wrote The Exalted Faith
before his translation activity made him familiar with additional
texts by Avicenna, he censored the Avicennian materials on which he
drew in his work addressed to Jewish readers. In any event, the
Arabic-into-Latin translation activity of the Jew Avendauth (whether
or not identical with Ibn Daʾud) had no impact on subsequent Jewish
intellectual life.
(iii) In Andalusia, probably around the middle of the thirteenth cen-
tury, Moses ben Joseph ha-Levi of Seville (Mūsā Ibn al-Lāwī
al-Ishbīlī)59 wrote a number of works drawing on Avicenna, of which
the only survivor is the short Maqāla ilāhiyya (Metaphysical
Treatise), preserved both in the original Arabic and in a Hebrew
translation. Steinschneider, followed by Georges Vajda, showed that
Moses ha-Levi defended Avicenna’s metaphysical ideas about the
difference between God and the First Mover against those of
Averroes (who identified the two).60 In particular, one passage in
the Maqāla closely paraphrases Avicenna’s doctrine of the nature of
the Necessary Existent (i.e., God), as found in the metaphysical sec-
tion of Avicenna’s Najāt.61 Here we have a first and rare use of
Avicenna in a direct criticism of Averroes by a Jewish author, a cir-
cumstance that justifies viewing Moses ha-Levi as a committed
Avicennian. It is also noteworthy – and no coincidence – that whereas
Judah Hallevi and Abraham Ibn Daʾud never mention Avicenna by
name (even in passages directly depending upon him), this committed
Avicennian refers to “Abū ʿAlī Ibn Sīnā” several times.
59
See on him Encyclopedia Judaica (Jerusalem, 1972), vol. 12, col. 421–422.
60
Steinschneider, Al-Farabi, p. 151; Georges Vajda, “Un champion de l’avicennisme. Le
problème de l’identité de Dieu et du Premier Moteur d’après un opuscule judéo-arabe
inédit du XIIIe siècle,” Revue thomiste, 48 (1948): 480–508, repr. in Georges Vajda, Études
de théologie et de philosophie arabo-islamique à l’époque classique, ed. by Daniel
Gimaret, Maurice Hayoun and Jean Jolivet, Variorum Reprints (London, 1986), Essay
IX. See also Harry A. Wolfson, “Averroes’ lost treatise on the Prime Mover,” Hebrew
Union College Annual, 23 (1950–51): 683–710. See also Georges Vajda, Recherches sur la
philosophie et la kabbale dans la pensée juive du Moyen Âge (Paris and The Hague,
1962), pp. 133, 213–15; id., “R. Moses ha-Levi’s view on divine providence,” Melila, 5
(1955): 163–8 (Heb.).
61
See Vajda, “Un champion de l’avicennisme,” p. 484.
(iv) Our fourth and last case is that of Joseph Ibn Waqār, who was
active in Toledo around 1360. His main work, al-Maqāla al-jāmiʿa
bayna al-falsafa wa-al-sharīʿa (Treatise Reconciling Philosophy and
Religious Law), extant in the original Arabic and in a fragmentary
Hebrew version, was studied by Vajda.62 Ibn Waqār’s goal is to “recon-
cile” philosophy and the superior science of Kabbalah. He devotes
some space to a discussion of the relationship between God and the
First Mover, incorporating the complete text of Moses ha-Levi’s
Avicennian treatise just considered.63 Ibn Waqār’s Avicennian sympa-
thies are further confirmed by his explicit references (including quota-
tions) to Avicenna himself and to two other philosophers with
Avicennian tendencies – al-Ghazālī and Fakhr al-Dīn al-Rāzī (a fol-
lower of Abū al-Barakāt al-Baghdādī).64 Ibn Waqār mentions
Avicenna by name more than once, especially in connection with the
metaphysical doctrines of the “giver of forms,” the nature of divine
providence, and the difference between the First Cause and the
First Mover. He also refers to the account of the activities of the five
senses as found in al-Shifāʾ.65 According to Vajda, Ibn Waqār, unlike
other arabophone Spanish philosophers (notably Ibn Tufayl and
Averroes), did not have first-hand knowledge of all of ˙Avicenna’s
oeuvre. Although he was clearly familiar with many doctrines pro-
pounded by Avicenna and his followers, he had access only to some
parts of al-Shifāʾ and al-Najāt; certain Avicennian doctrines he
knew only through their summary presentations by other authors.
His Avicennian leanings notwithstanding, Ibn Waqār does not follow
Avicenna uncritically and distances himself from some of his ideas.
The study of other texts in Arabic composed by Jews is apt to dis-
cover additional authors who drew on works by Avicenna. For
example, according to Steinschneider, Solomon Ibn Yaʿish, a physician
in Seville (d. 1345), drew heavily on al-Shifāʾ in his as yet unstudied
and mainly philosophical Arabic commentary on Avicenna’s Canon.66
A.2. Avicenna among Arabophone Jews: The Manuscript Evidence

Although arabophone Jewish scholars could read Arabic in Arabic
script, when they copied texts for their own or their coreligionists’
use they usually preferred to employ the Hebrew alphabet. (Arabic
transcribed in Hebrew letters is often designated “Judeo-Arabic”; we
62
Vajda, Recherches, pp. 115–297.
63
Ibid., pp. 201–6.
64
Ibid., pp. 125–6, 128, 132.
65
Ibid., p. 132.
66
See Moritz Steinschneider, Die arabische Literatur der Juden (Frankfurt, 1902), p. 167 (§ 129,
no. 2); id., HÜ, pp. 686–7; and id., “Anzeigen,” p. 23; Y. Tzvi Langermann, “Solomon Ibn
Yaʿish’s Commentary on Avicenna’s Canon,” Kiryat Sefer, 63 (1990–1991), pp. 1331–3 (Heb.).
will use this somewhat imprecise term.67) The existence or absence of

Arabic texts in Hebrew characters is thus an excellent indicator
of whether arabophone Jews were interested in a given work: a
Judeo-Arabic exemplar testifies to real, direct knowledge of a
work among arabophone Jews. In 1893, the great historian and bibli-
ographer Moritz Steinschneider compiled a list of all Hebrew-alphabet
manuscripts of non-Jewish works in Arabic known to him.68 A century
later, Y. Tzvi Langermann prepared a new list to complement
Steinschneider’s.69 Drawing on both lists, supplemented by data
from the online catalogue of the Institute for Microfilmed Hebrew
Manuscripts (IMHM) at the National Library of Israel in Jerusalem,
we offer some rough data on the existence of Avicenna’s works in
Judeo-Arabic manuscripts. (The information could not be checked
first-hand and depends on that supplied by the IMHM catalogue.)
The discussion here excludes evidence from the Geniza, which is pre-
sented in the Appendix, contributed by Renate Smithuis and Sagit
Butbul (below, pp. 284–7). Our purpose is not to offer a comprehensive
list of available manuscripts, but only to provide an idea of arabophone
Jewish readers’ use of Avicenna. As we will see, his philosophical
works and medical ones have very different histories.
A.2.1. Avicenna’s Philosophical Works

We begin by noting a glaring absence. There is no trace of any com-
plete or partial Judeo-Arabic copy of the key texts, al-Shifāʾ and
al-Najāt. This seems consistent with the observations above that
the philosophical works written by arabophone Jewish scholars,
with the notable exceptions of Ibn Daʾūd and Judah Hallevi, show
no evidence of familiarity with these Avicennian texts.
Manuscripts in Hebrew script of the following works are extant:
■ A section on geometry from Avicenna’s Shifāʾ, a single fifteenth-
century manuscript (Paris, Bibliothèque Nationale de France,
héb. 1099, fols. 25v–26r).70
67
The term is misleading inasmuch as the term “Judeo-Arabic” denotes a diglossia, namely
that used by Jews speaking Arabic, characterized by the inclusion of numerous Hebrew
words and phrases. When non-Jewish Arabic texts are copied in Hebrew script, the
language remains standard literary Arabic and is not, strictly speaking, Judeo-Arabic.
68
Moritz Steinschneider, “Schriften der Araber in Hebräischen Handschriften, ein Beitrag
zur arabischen Bibliographie,” Zeitschrift der Deutschen Morgenländischen Gesellschaft,
47 (1893): 335–84, on pp. 343–5.
69
Y. Tzvi Langermann, “Arabic writings in Hebrew manuscripts: a preliminary relisting,”
Arabic Sciences and Philosophy, 6 (1996): 137–60, on pp. 156–8. See also id.,
“Transcriptions of Arabic treatises into the Hebrew alphabet: an underappreciated mode
of transmission,” in F. Jamil Ragep and Sally P. Ragep (eds.), Tradition, Transmission,
Transformation (Leiden, 1996), pp. 247–60.
70
Tony Lévy, “Une version hébraïque inédite des Éléments d’Euclide,” in Danielle Jacquart
(ed.), Les voies de la science grecque. Études sur la transmission des textes de l’Antiquité au
■ The section on physics of Avicenna’s encyclopedia, al-Hikma
˙
al-mashriqiyya (The Oriental Philosophy), a single thirteenth-
century manuscript (Oxford, Bodleian Library, Pococke 181/4,
fols. 63b–154a).71
■ Risāla fī Aqsām al-ʿulūm (Epistle on the Divisions of the
Sciences), in the same Oxford manuscript (Pococke 181/3, fols.
54a–61b).
■ Kitāb al-Hudūd (Book of Definitions), in a single seventeenth-
˙
century Yemenite manuscript, showing the work was still read
and copied by Yemenite Jews at that period (London, British
Library, Add. 27542).
■ Risāla fī Ahwāl al-nafs (Epistle on the States of the Soul), in a
˙
single fifteenth-century manuscript (St. Petersburg, Russian
National Library, Evr. Ar. II 2373).
■ A philosophic-theological work ascribed to Avicenna, Kalimat
al-tawhīd (Discourse on the Unity [of God]), St. Petersburg,
˙ II 1831 (a single folio).
Evr. Ar.
We also take note of the following, apparently spurious work, which
is, however, imbued with Avicennian thought72:
■ Risāla fī Maʿrifa al-nafs al-nātiqa wa-ahwālihā (Epistle on the
Knowledge of the Rational Soul ˙ ˙ States), in a single
and its
fourteenth-century manuscript (St. Petersburg, Evr. Ar. I 3048).
A.2.2. Avicenna’s Medical Works

First and foremost, as could be expected, is the Qanūn fī al-tibb (The
Canon). Langermann lists no fewer than eighteen copies, though ˙ none
of them is complete. His list can be supplemented by the following:
■ Munich, Bayerische Staatsbibliothek, ar. 816: the oldest extant
Judeo-Arabic copy of the complete text of the Canon, dating to
the fourteenth century.
■ Paris, Bibliothèque Nationale de France, héb. 1130: a sixteenth-
century copy of the complete Arabic text of the Canon.
■ Oxford, Bodleian Library, Hunt. Don. 7: another sixteenth-
century copy of the complete Arabic text of the Canon.
e
XIX siècle (Genève, 1997), pp. 181–239, on p. 208, n. 37. Lévy notes that this section is iden-
tical with the one translated into Hebrew (below, at n. 86).
71
See also Y. Tzvi Langermann, “Arabic writings in Hebrew manuscripts: Suhrawardi, Ibn
Sīnā, and Ibn al-Tayyib,” Alei Sefer, 21 (2010): 21–33, on pp. 25–7 (Heb).
72 ˙
See Jean Michot, “L’épître sur la connaissance de l’âme rationelle et de ses états attribuée à
Avicenne,” Revue philosophique de Louvain, 82 (1984): 479–99; Dimitri Gutas, Avicenna
and the Aristotelian Tradition. Introduction to Reading Avicenna’s Philosophical Works
(Leiden, 1988), p. 305.
■ Cambridge University Library, Add. 551,1: sixteenth-century

copy, Book II only.
The following Judeo-Arabic copies of three minor medical works by
Avicenna can also be added to Langermann’s list:
Three copies of the medical poem Urjūza (Cantics) are extant:
■ St. Petersburg, Evr. Ar. I 2245 (excerpt with commentary; fif-
teenth century).
■ Paris, Alliance Israélite Universelle, H 157 A: late Oriental
manuscript, seventeenth-eighteenth centuries.
■ Berlin, Staatsbibliothek, OR Qu 751/6 (Karaite, dated 1666).
There is one copy of a commentary on the Mudkhal fī al-tibb
(Introduction to Medicine): ˙
■ St. Petersburg, Evr. Ar. I 4813: date unknown.

Finally, the Risāla fī Dafʿ al-mudār al kulliyya (Letter on the
Protection against Damage in General)73 was copied twice, in
Oriental hands:
■ Moscow, Russian State Library, Guenzburg 724/15: sixteenth
century.
■ Cambridge, University Library, Add. 1491: dated 1720.
What can be concluded from these manuscripts? A first observation
is the striking absence of Avicenna’s major philosophical works in the
medieval Judeo-Arabic world. The contrast with Averroes (seven phi-
losophical treatises and two medical works) is particularly striking.74
Clearly, Avicenna’s philosophy was low on the priority list of arabo-
phone Jewish scholars. Data from the Cairo Geniza confirm this
appreciation: there too, fragments of works by Avicenna are rare
(see the Appendix). By the same token, philosophical works by
Avicenna seldom appear in book lists found in the Geniza.75
We saw above that Maimonides probably derived all or most of his
knowledge of Avicenna’s philosophy from al-Ghazālī’s Maqāsid. This
raises the question: did also other medieval arabophone Jews ˙ use
that work instead of works by Avicenna himself? Below we will see
that this was precisely the case for Jews who were ignorant of
Arabic. Insofar as arabophone Jewish scholars are concerned,
however, the answer to our question is negative: only a single
73
Steinschneider, “Schriften der Araber,” pp. 344–5.
74
Ibid., pp. 342–3.
75
See Nehemya Aloni, The Jewish Library in the Middle Ages. Book Lists from the Cairo
Genizah,˙ edited by Miriam Frenkel and Haggai Ben-Shammai (Heb.) (Jerusalem, 2006),
index, s.v. Ibn Sīnā.
Judeo-Arabic copy of the Maqāsid is extant (some additional frag-
ments were found in the Geniza). ˙ 76
As against this shunning of Avicenna the philosopher, arabophone
Jews were highly appreciative of Avicenna as a physician. Between c.
1300 and 1600, Avicenna’s medical magnum opus was copied over, in
whole or part, many times; even some minor medical works were disse-
minated among Judeo-Arabic physicians. “Clearly Ibn Sīnā’s book was
the most important text used by medieval Jewish doctors in medieval
and Renaissance times,”77 Langermann comments. He further observes
that the marginal notes in the manuscripts provide valuable indications
about how Jewish physicians read the work. By contrast, although most
of the Avicenna material in the Cairo Geniza is medical, it amounts to
very little (see again the Appendix). Nor do Avicenna’s medical works
appear frequently on booklists preserved in the Geniza.78
The prominent role of Avicenna’s medical works for arabophone Jewish
physicians has been confirmed by P. Sj. van Koningsveld’s exhaustive
inspection of nearly all the extant medieval Arabic manuscripts (in
Arabic script) produced in Christian Spain.79 Of 104 such manuscripts,
24 could be identified as having circulated among Jews in those realms.
Three of them contain works by Avicenna: al-Shifāʾ (one copy) and
extracts from The Canon (in two manuscripts).80 While this is obviously
not much, it should be noted that the manuscripts that circulated
among Muslims (as well as those that could not be ascribed to a particular
religious milieu) contain no material by Avicenna at all.
A.2.3. Conclusion
Avicenna the philosopher was little read by arabophone Jews.
Maimonides’ attitude towards him is enigmatic. Although he commen-
ted on the quality of Avicenna’s philosophical works, he does not seem
to have read them at first-hand, but to have extracted their content
mainly through al-Ghazālī. Whatever the case, the historically signifi-
cant fact here is that Maimonides used some Avicennian doctrinal
items as important building blocks of his thought, thereby contributing
76
Langermann, “Arabic writings in Hebrew manuscripts,” pp. 148–9. To this, two copies of
Zakkah al-nufūs, published by Saʿīd b. Daūd al-ʿAdeni (Syria, c. 1451–1485) under his
own name, should be added: Steinschneider (HÜ, p. 298) showed that this is a plagiarism
of the Maqāsid.
77
Langermann, ˙ “Arabic writings in Hebrew manuscripts,” p. 157.
78
See n. 75 above.
79
Pieter Sj. van Koningsveld, “Andalusian-Arabic manuscripts from Christian Spain: a com-
parative intercultural approach,” Israel Oriental Studies, 12 (1992): 75–110; id.,
“Andalusian-Arabic manuscripts from medieval Christian Spain: some supplementary
notes,” in M. Forstner (ed.), Festgabe für Hans-Rudolph Singer (Frankfurt, 1991), 2, pp.
811–23.
80
Van Koningsveld, “Andalusian-Arabic manuscripts from Christian Spain,” pp. 100–3, nos.
75, 88, 89.
to their diffusion (albeit as “Aristotle’s”) among Jewish readers (both

Arabic-speaking and Hebrew-speaking) in the following centuries.
Maimonides’ descendants appropriated only Avicenna’s refutation of
metempsychosis. In the West, arabophone Jewish philosophers turned
to Avicenna mostly for specific topics. The extant manuscripts, too,
suggest that only a few minor works, bearing on specific themes, notably
the nature of the soul, were read and copied. This state of affairs is con-
sistent with our observations concerning the presence of Avicennian
doctrinal items in Judeo-Arabic works on philosophy: two authors
(Judah Hallevi and Joseph Ibn Waqār) drew on Avicenna for their dis-
cussions of the human soul. Moses ha-Levi is exceptional in his clear
self-identification as an Avicennian. Avicenna the physician, by con-
trast, although all but ignored by Maimonides, was very popular
among his arabophone Jewish confreres.
PART B: AVICENNA AMONG NON-ARABOPHONE JEWS
B.0. Introduction
Arabophone Jews usually wrote their theoretical works (linguistics, reli-
gious philosophy, biblical exegesis, etc.) in Arabic, while producing
belles-lettres (mainly poetry, but occasionally prose) in Hebrew,
especially in Spain. In the early twelfth century, a movement of
Arabic-to-Hebrew cultural transmission (whose roots lay in the previous
century) gathered momentum. The channels of transmission were two-
fold: (i) arabophone Jewish scholars wrote works in Hebrew that drew on
knowledge they acquired from Arabic sources, thereby transmitting it to
their coreligionists who had no Arabic; (ii) bilingual scholars produced
Hebrew translations of works by Jews, Muslims, and pagans available
to them in Arabic. This translation project, which spanned some four
centuries (early twelfth–end of 15th century), laid the foundations for
medieval Hebrew philosophy and science.81 Philosophical tracts
were translated into Hebrew from Latin, too.82
81
The summa on this subject is still Steinschneider, HÜ. See also Mauro Zonta, La filosofia
antica nel Medioevo ebraico (Brescia, 1996). A sociologically informed account is offered in
Gad Freudenthal, “Les sciences dans les communautés juives médiévales de Provence: leur
appropriation, leur rôle,” Revue des études juives, 152 (1993): 29–136; id., “Science in the
medieval Jewish culture of Southern France,” History of Science, 33 (1995): 23–58; repr.
in Science in the Medieval Hebrew and Arabic Traditions, Variorum Collected Studies
Series, 803 (Aldershot, 2005), Essay I. For a chronological listing and bibliography see
Mauro Zonta, “Chronological table of the medieval Hebrew translations of philosophical
and scientific texts,” in Gad Freudenthal (ed.), Science in Medieval Jewish Cultures
(Cambridge, 2011), pp. 17–73.
82
See Alexander Fidora, Resianne Fontaine, Gad Freudenthal, Harvey Hames, and Yossef
Schwartz (eds.), Latin-into-Hebrew – Studies and Texts. 2 vols (Leiden, forthcoming in
2012).
We will now consider how, when, and to what extent Jewish scholars
whose cultural tongue was Hebrew and who, as a rule, read only
Hebrew, could become aware of Avicenna and Avicennian thought,
whether identified as such or not. We begin with philosophy (B.1), look-
ing at translations and adaptations of works by Avicenna as well as
works written in Hebrew by arabophone Jewish scholars who drew on
Avicenna. We will treat them by literary genre, more or less chronologi-
cally within each genre. The contribution of the Latin-to-Hebrew chan-
nel will also be addressed. A short concluding section will mention
pseudo-Avicennian scientific works thought to be genuine by medieval
scholars. Finally, we will turn to Avicenna’s medical works (B.2).
B.1. Avicenna’s Philosophy in Hebrew

B.1.1. Hebrew Translations and Adaptations of Philosophical
Works by Avicenna
Remarkably little of Avicenna’s philosophical oeuvre was trans-
lated into Hebrew.83 This very significant datum of medieval Jewish
cultural history calls for explanation, which we will attempt in the
Conclusion. Specifically, almost nothing of Avicenna’s great philoso-
phical work, al-Shifāʾ, was translated; only his shorter encyclopedia
al-Najāt (The Salvation) was translated into Hebrew, albeit partially
and late. However, two so-called “visionary” texts by Avicenna were
adapted in Hebrew.
(i) As noted, only a few short passages of al-Shifāʾ were translated
into Hebrew:
■ One brief text on the division of the sciences (preserved in six
manuscripts), includes the explicit notation that it derives
from the Shifāʾ.84 Steinschneider has plausibly suggested that
the anonymous translator is either Samuel Ibn Tibbon (whose
fondness for Avicenna will be discussed below) or his son
Moses;85 this would place the date of the translation in the
second quarter of the thirteenth century.
83
Already noted by Steinschneider, HÜ, p. 279.
84
Published by Leopold Dukes in Ozar Nechmad. Briefe und Abhandlungen jüdische
Literatur betreffend 2 (1857), pp. 114–15; repr. in Harry A. Wolfson, “The classification
of sciences in medieval Jewish philosophy” (1925), in id., Studies in the History of
Philosophy and Religion, ed. by Isadore Twersky and George H. Williams (Cambridge,
Mass., 1973), 1, pp. 493–545, on p. 495, n. 8. Preserved in MS Leipzig, University
Library, B. H. Fol. 14/9, fol. 205a–206a. The original Arabic text of this passage is found
in the second chapter of treatise 1 of Avicenna’s Eisagoge in al-Shifāʾ: see Ibn Sīnā,
al-Shifāʾ, al-Mantīq, vol. I, ed. by George C. Anawati, Mahmūd Khudayrī and Fuʾād
˙
Ahwānī (Cairo, 1952), pp. 12–14. ˙ ˙
85
Steinschneider, HÜ, p. 285 (§ 154).
■ The sections of al-Shifāʾ on arithmetic and on geometry were

translated separately into Hebrew. Both of these circulated
anonymously, so readers were unaware of their Avicennian ori-
gin. The place and time of these (unpublished) translations, pre-
served in only one and two manuscripts (respectively), are
unknown.86
■ Several passages from al-Shifāʾ, on logic and physics, were
included in Todros Todrosi’s Liqqutim (“Philosophical
˙
Anthology”), which ˙
had ˙a negligible circulation (see below,
B.1.4 [iii]).
So Avicenna’s most ambitious philosophical work, al-Shifāʾ, had

hardly any presence in Hebrew. As a result, it did not even have a
Hebrew title; Hebrew authors usually refer to it by its Arabic title,
al-Shifāʾ, which they knew first- or secondhand.87
(ii) In fact, the only philosophical work by Avicenna of which sub-
stantial parts were translated into Hebrew is al-Najāt; sections two
and three (on physics and metaphysics) were translated by Todros
˙
Todrosi in Arles, c. 1334–1340, under the title Hassalat ha-nefesh. 88
˙ ˙ by the transla-
˙The section on logic was probably deliberately skipped
tor rather than having been lost in the course of transmission. The
reason for its omission may be that by 1337 Todrosi had already trans-
lated (or was about to translate) Averroes’ ˙ commentary on the
Rhetoric and had given an account of logic in his “Philosophical
89
Anthology.” In that work, however, the section on logic drew on

additional sources (Themistius, al-Fārābī – including his lost Long
Commentary on Book 8 of the Topics – and Averroes).90 Hassalat
˙˙
86
The translation of arithmetical part of the Shifāʾ is embedded in the translation of a work
ascribed to Abū al-Salt al-Andalusī (1068–1134). See Tony Lévy, “L’histoire des nombres
˙
amiables: Le témoignage des textes hébreux médiévaux,” Arabic Sciences and
Philosophy, 6 (1996): 63–87, on p. 65. On the geometrical text, a Hebrew version of the
usūl al-handasa in the Shifāʾ, see id., “Les Eléments d’Euclide en hébreu (XIIIe-XVIe
˙
siècles),” in Ahmad Hasnawi, Abdelali Elamrani-Jamal, and Maroun Aouad (eds.),
Perspectives arabes et médiévales sur la tradition scientifique et philosophique grecque
(Leuven, 1997), pp. 79–94, on p. 80; this text is extant in a manuscript in Hebrew charac-
ters (above, near n. 70). See also below (B.1.4 [v]) on the “paraphrase” of these sections by
Judah ben Solomon ha-Kohen in his Midrash ha-Hoḵmah.
87
Pointed out in Steinschneider, HÜ, p. 280. ˙
88
Ibid., p. 285; Ernest Renan [and Adolf Neubauer], Les Écrivains juifs français du XIVe
siècle (= Histoire littéraire de la France, vol. XXXI) (Paris, 1893), p. 571. The translation
of Book II, Part VI of this work (On the Soul) has been edited in Gabriella Berzin, “The
medieval Hebrew version of psychology in Avicenna’s Salvation (Al-Najāt),” Ph.D. disser-
tation, Harvard University, 2010. We thank Dr. Berzin for having put a copy of her disser-
tation at our disposal.
89
Steinschneider, HÜ, pp. 64–5 (§ 21).
90
More on this below (B.1.4 [iii]); see Zonta, “Fonti antiche e medievali,” pp. 555–62.
ha-nefesh, which survives in only two manuscripts, did not play a
notable role in the reception of Avicenna in Hebrew or in medieval
Jewish thought in general.91
Strikingly, the only works by Avicenna that were rendered into
Hebrew in toto are two of the three compositions that Henry Corbin
has described as “visionary.”92 Also remarkable is that they were
not translated, but rather freely adapted in the form of maqāma.
(iii) Avicenna’s Hayy Ibn Yaqzān was adapted into Hebrew by the
˙
great polymath Abraham ˙ (1089–1167). Ibn Ezra presumably
Ibn Ezra
composed this work after he began writing theoretical works in
Hebrew, following his migration from Spain to Italy in 1140.93 The
fact that Ibn Ezra’s Hay ben Meqis is a Hebrew adaptation of
Avicenna’s Hayy Ibn Yaqz˙ ˙
ān was noted by the medieval editor of
˙ ˙
Ibn Ezra’s diwān; so some medieval readers of the maqāma were pre-
sumably aware of it.94 A very creative thinker and poet, Ibn Ezra, who
may have used Ibn Zayla’s Arabic commentary on Avicenna’s Hayy
˙
Ibn Yaqzān, did not follow his model slavishly. The study of the differ-
˙
ences between the two sheds interesting light on Ibn Ezra’s own
worldview and intentions.95
At an unknown date, an otherwise unidentified “Moses” translated
Avicenna’s text into Hebrew, this time together with Ibn Zayla’s
commentary.96
(iv) The second of Avicenna’s “visionary” works to be adapted in
Hebrew is the lesser-known Risālat al-Tayr (Treatise on Birds). Its
Hebraicization, too, was written in the ˙form of a maqāma – in fact,
three different ones. Each of the adapters “transformed” Avicenna’s
prose work into a maqāma, producing an independent poetical work
in a literary form conforming to the philosophical content. Although
91
London, British Library, Add. 27559/2, and Paris BNF, MS héb. 1023/4.
92
Henry Corbin, Avicenna and the Visionary Recital (Princeton, 1990).
93
For a chronological listing of Ibn Ezra’s works, see Shlomo Sela and Gad Freudenthal,
“Abraham Ibn Ezra’s scholarly writings: a chronological listing,” Aleph, 6 (2006): 13–55.
94
That fact that Hayy Ibn Yaqzān is an adaptation of Avicenna’s work was confirmed by
Steinschneider ˙in his “Hai ben˙ Mekiz,” in Jacob Egers (ed.), Diwān des Abraham Ibn
Esra, mit einer Allegorie Hai ben Mekiz (Berlin, 1886), pp. 177–82; see also id., HÜ,
pp. 285–6. For a critical edition of Ibn Ezra’s adaptation, see Israel Levin (ed.), Hay ben
˙
Meqis (Heb.) (Tel Aviv, 1983); the “Editor’s introduction” (pp. 11–45) offers a detailed
˙
analysis of Ibn Ezra’s work and comments on its differences with respect to Avicenna’s.
95
Aaron Hughes, “The three worlds of Ibn Ezra’s Hay ben Meqis,” Journal of Jewish Thought
and Philosophy, 11 (2002): 1–24; id., “A case of˙ twelfth-century
˙ plagiarism? Abraham Ibn
Ezra’s Hay ben Meqis and Avicenna’s Hayy Ibn Yaqzān,” Journal of Jewish Studies, 55
(2004): ˙306–11; id., ˙The Texture of the˙ Divine: Imagination
˙ in Medieval Islamic and
Jewish Thought (Bloomington, IN., 2004).
96
The two texts have been edited by David Kaufmann: Iggeret Hayy ben Meqis le-Ibn Sina ‘im
perush talmido Ibn Zayla,” Qoves ʿal-Yad, 2 (1886) (separate ˙ numbering, VI+
˙ 29 pp.). See
also Steinschneider, HÜ, p. 286.˙
the Hebrew adaptations included the notation that they derived from
an Arabic original, Avicenna’s authorship was discovered only by
modern scholarship.97
None of Avicenna’s great philosophical works was available in
Hebrew, then: only a few passages (two unattributed) of al-Shifāʾ;
al-Najāt was translated partially and late, and was very little disse-
minated; two “visionary” works were reworked into maqāmas, one
of them without attribution. (The Risāla fī al-Nafs [Epistle on Soul]
was also translated in part, albeit incorporated, unidentified, into
two works that will be discussed below.98 ) This was certainly an insuf-
ficient basis for Hebrew readers to become familiar with Avicenna’s
thought. Nonetheless, and this must be re-emphasized, despite the
absence of translations, Avicenna was undoubtedly a household
name among Hebrew-reading intellectuals. They derived information
on Avicenna’s ideas from a number of sources: Hebrew translations of
works by Muslim scholars other than Avicenna, but containing
accounts of his thought (B.1.2); Avicennian doctrinal items in
Hebrew translations of works by arabophone Jewish authors (B.1.3);
and Avicennian doctrinal items in works by arabophone Jewish scho-
lars who wrote in Hebrew (B.1.4). Let us consider these three chan-
nels in turn.
B.1.2. Avicenna in Hebrew Translations of Philosophical Works by

Other Muslim Authors
The Muslim authors whose works contained substantial material
on Avicenna and whose works reached the Hebrew-reading public
are mainly al-Ghazālī and Averroes; some additional information
came via Fakhr al-Dīn al-Rāzī.
97
The first identification of the Arabic Vorlage of the text is due to Israel Levin, “The gazelle
and the birds: On Megillat ha-Ofer of Rabbi Elijah ha-Cohen and the Treatise on Birds of
Avicenna,” in Menachem Brinker, Yosef Yahalom, and Jonah Fraenkel (eds.), Essays in
Memory of Dan Pagis (= Jerusalem Studies in Hebrew Literature, vols. 10–11 [1988]),
pp. 577–611. Text published in Zvi Malachi, “Megillat ha-ʿOfer by Rabbi Eliyahu
ha-Kohen, an allegorical maqāmah from Spain,” in id. (ed.), Aharon Mirsky Jubilee
Volume (Lod, 1986), pp. 317–41 (Heb.). The translation was made by Eliyahu ben
Moses ben Nissim ha-Cohen on Sunday, 26 January 1276. Nothing is known about this
translator – Steinschneider tried in vain to identify him with individuals known from
other sources; see his “Devarim ‘atiqim. 3. Megillat ha-‘Ofer,” in Ha-Karmel, 6 (1867):
319–21 (Heb.); he lists 10 individuals named “Eliya ha-Kohen” or the sons of Eliya
ha-Kohen, in the hope that this would be helpful for future research. See also HÜ,
p. 884. Yoseph Yahalom identified two further adaptations, which are not extant in full.
See his “Ibn Sīnā’s Iggeret ha-Sipporim in three Hebrew Maqāma-like adaptations,” in
Nahem Ilan (ed.), The Interwined ˙ Worlds of Islam. Essays in Memory of Hava
˙
Lazarus-Yafeh (Jerusalem, 2002), pp. 282–314 (Heb.). This article also includes interesting
reflections on the genre of the maqāma and its use in philosophy.
98
It was incorporated in Judah Hallevi’s Kuzari, translated by Judah Ibn Tibbon; it was also
translated into Hebrew in Falaquera’s Deʿot ha-pilosofim.
(i) Al-Ghazālī
As is now well known, al-Ghazālī’s philosophical work Maqāsid
al-falāsifa (The Intentions of the Philosophers) is largely based ˙on
Avicenna’s Dānesh-Nāmeh (Book of Knowledge).99 In his introduction,
al-Ghazālī states that his intention is to describe the philosophers’
ideas in order to refute them – a passage included in all Hebrew trans-
lations (but absent from most Latin manuscripts).100 Nowhere in his
work does al-Ghazālī mention Avicenna. Jewish scholars understood
al-Ghazālī’s stated goal in various ways: some (like Moses Narboni in
the fourteenth century) believed that al-Ghazālī’s real intention was
to teach philosophy and that it was only the fear of his anti-
philosophical contemporaries that led him to present his philosophical
work as a criticism of philosophy.101 Others (like Isaac Albalag and
Judah ben Solomon Nathan, also in the fourteenth century) took
al-Ghazālī at his word and either applauded his anti-philosophic inten-
tions (Judah Nathan) or scolded him for them (Albalag).102 Between
99
See above, n. 28.
100
Steinschneider, HÜ, p. 310. Translated in Steven Harvey, “Why did fourteenth-century
Jews turn to Alghazali’s account of natural science?,” The Jewish Quarterly Review, 91
(2001): 359–76 on p. 361. Recent scholarship (following some medieval predecessors) has
challenged the sincerity of this self-description. See, e.g., Frank Griffel, “Al-Ghazālī,”
The Stanford Encyclopedia of Philosophy (Fall 2008 Edition), Edward N. Zalta (ed.),
URL = <http://plato.stanford.edu/archives/fall2008/entries/al-ghazali/> id., Al-Ghazālī’s
Philosophical Theology (Oxford, 2009); Alexander Treiger, “The science of divine disclos-
ure: al-Ghazālī’s higher theology and its philosophical underpinnings,” Ph.D.
Dissertation, Yale University, 2008.
101
Narboni’s statement is paraphrased in Steinschneider, HÜ, pp. 316–17, published in
Gershon B. Chertoff, “The logical part of al-Ghazālī’s Maqāsid al-falāsifa, in an anonymous
Hebrew translation with the Hebrew commentary of˙ Moses of Narbonne” (Ph.D.
Dissertation Columbia University, New York, 1952; available online at http://www.gha-
zali.org/books/chertoff.pdf), Part II, p. 6 (= p. B3), translated in Harvey, “Why did
fourteenth-century Jews turn to Alghazali’s account,” pp. 366–7 (n. 24). Similarly, in
1492, an anonymous copyist wrote: “al-Ghazālī was a disciple of Avicenna and this book
is a compendium of Avicenna’s logic, as [Walter] Burley wrote in many places”; Oxford,
Bodleian Library, MS Pococke 343, fol. 160r; see A. Neubauer, Catalogue of the Hebrew
Manuscripts in the Bodleian Library (Oxford, 1886), c. 470. Averroes, too, saw the
Maqāsid as a work whose intention was merely to teach science; see Harvey, “Why did
˙
fourteenth-century Jews turn to Alghazali’s account,” pp. 362–3.
102
Steinschneider, HÜ, 301–4, 307–9, 316–17. Simeon Duran, writing in 1423, opined that “this
man [al-Ghazālī] learnt wisdom only from the books of Avicenna, as Averroes has remarked
in the Destruction of the Destruction”; see Moritz Steinschneider (ed.), Simeon Duran, Qeshet
u-magen, in Magazin für die Wissenschaft des Judentums, 9 (1882), Hebrew Part: Osar Tov,
˙ ˙
1–36, on p. 26; = Sefer Magen ʾavot (Jerusalem, 1997), p. 241. Translation in Steinschneider,
“Islam und Judenthum. Kritik des Islam von Simon Duran (1423),” Magazin für die
Wissenschaft des Judentums, 7 (1880): 1–48, on p. 36. The passage to which Duran refers
seems to be in Averroes, Tahafut at-tahafut, ed. Maurice Bouyges, Bibliotheca Arabica
Scholasticorum, Série Arabe 3 (Beirut, 1930), p. 403.11–12. In Simon Van Den Bergh’s
English translation (Cambridge, 1954) it reads: “But all the arguments which al-Ghazali
gives in this book either against or on behalf of the philosophers or against Avicenna are dia-
lectical through the equivocation of the terms used, and therefore it is not necessary to expati-
ate on this.” In a note ad loc Bouyges observes that in the Hebrew translation of this passage
there is a variant reading here, according to which Averroes writes that “all what is in this
roughly 1290 and 1330, the Maqāsid was translated into Hebrew, per-
haps even three times,103 and was ˙ the object of quite a few commen-
taries in Hebrew. It was immensely popular for centuries, as
indicated by the great number of extant manuscripts of all versions.
The Intentions of the Philosophers in Hebrew garb (entitled
Kawwanot [sometimes Deʿot] ha-pilosofim) brought the Avicennian
ideas it presented to Hebrew-reading scholars. It is important to rea-
lize, however, that the text did not reveal its Avicennian pedigree; read-
ers could know this only if they had access to external sources of
information, notably in Arabic, as the translators did.104
The first Hebrew version of the Maqāsid was made in the late thir-
teenth century by the philosopher Isaac˙ Albalag, who may have been
from Narbonne.105 He accompanied the translated text with a run-
ning commentary and named the entire work Sefer Tiqqun ha-deʿot
(Rectification of the Opinions [or Ideas]).106 Albalag explains what
motivated him to translate the work. The Torah and philosophy
express the same truth, albeit in different languages, adapted to the
capacities of different groups of people. He himself belongs to the phi-
losophers and would therefore translate strictly philosophical books,
namely those of Aristotle. “Nonetheless, I saw fit to begin by translat-
ing this book by Abū Hāmid al-Ghazālī, because it narrates most of
˙
those [philosophical] ideas, treading a mean route that is common
to philosophy and the vulgar belief. Its method is close to that of nar-
ration [sippur], so that it is easily comprehensible [even] to one who is
not trained in philosophy.”107 In his commentary, Albalag criticizes
book by Abu Hamid about [or: against] philosophy [comes] from Avicenna.” It seems that this
is what Duran˙ read. Among Muslim and Christian authors, al-Ghazālī’s intentions were also
diversely perceived; see Harvey, “Why did fourteenth-century Jews turn to Alghazali’s
account?,” pp. 362–3.
103
Steinschneider, HÜ, p. 309, distinguishes three translations, but Wolfson casts doubt on
this; see Harry A. Wolfson, Crescas’ Critique of Aristotle. Problems of Aristotle’s Physics
in Jewish and Arabic Philosophy (Cambridge, 1929), p. 10, n. 44.
104
Charles H. Manekin, “The logic of the Hebrew encyclopedias,” in Steven Harvey (ed.),
Medieval Hebrew Encyclopedias of Science and Philosophy (Dordrecht, 2000), pp. 277–99
discusses the intentions of the translators of Maqāsid and the success of the work on
pp. 287–92. ˙
105
Steinschneider, HÜ, pp. 299–306 (§§ 166–171). About this author and his possible place of
origin, see also Mauro Zonta, “Due note sulle fonti ebraiche di Giovanni Pico e Giordano
Bruno,” Rinascimento, 40 (2000): 143–53, on pp. 143–6.
106
Albalag’s comments, but not his translation of the Maqāsid, have been edited in Georges
˙
Vajda (ed.), Isaac Albalag, Sefer Tiqqun ha-deʿot (Jerusalem, 1973). Of the translation,
only the introduction has been edited and published: Heimann Auerbach, Albalag und
seine Übersetzung des Makāsid al-Gazzalis (Breslau, 1907). Vajda translated most of
Albalag’s commentary into French˙ in his Isaac Albalag, averroïste juif, traducteur et anno-
tateur d’al-Ġ azālī (Paris, 1960).
107
Sefer Tiqqun ha-deʿot, ed. Vajda, p. 4.22–26. The term sippur here translates the term
hikāya employed by al-Ghazālī himself in the Introduction to the Maqāsid (for this obser-
˙
vation we are indebted to Alexander Treiger). ˙
the opinions expressed in the text, siding with Averroes against
Avicenna and al-Ghazālī; in this way, his readers became aware of
the discrepancies between their views. Albalag had no qualms about
voicing “radical” opinions openly; for this reason he was often deemed
heretical by later orthodox scholars.108 Nonetheless, to judge from the
number of surviving manuscripts (at least thirteen), Albalag’s book
was widely read. A number of mainly critical supercommentaries on
Albalag’s commentary were also composed.109
A second, anonymous version was produced in the first decades of
the fourteenth century. It too is extant in a number of manuscripts
(seven).110 The section on logic has been published.111
Finally, a third translation was executed by Judah ben Solomon
Nathan (extant in a dozen manuscripts), a physician who translated
a number of medical works as well.112 We have some information on
Judah Nathan’s motives for undertaking the translation. In his long
and informative introduction to the translation,113 he says that
al-Ghazālī presented the philosophers’ ideas in logic, natural science,
and philosophy most exactly and comprehensively.114 The compo-
sition therefore suited Judah Nathan’s purpose, which was to allow
a scholar who is engaged in Talmud study – which requires much
time – to learn the sciences “in a short time, so as not to have to inter-
rupt his [Talmud] study.” Everyone, Judah writes, desires to know the
sciences “by nature,” an allusion to the opening sentence of Aristotle’s
Metaphysics that bespeaks the author’s philosophical leanings. Any
such student, Judah Nathan says, will “find what he seeks in this
book which contains the totality of the intention[s] of the philosophers
in the greatest brevity, without needing any further book.” Thus, by
offering Jewish scholars a digest of all philosophical science in a single
volume, Judah Nathan would allow them to avoid losing precious time
with philosophy; they can and should put their time to better use by
studying the Talmud. Judah Nathan was a conscientious and meticu-
lous scholar: realizing that he did not always understand al-Ghazālī’s
text, and recognizing moreover that it expressed ideas deriving from
Avicenna, he also consulted works by the latter: he writes that he
“perused [. . .] different books which follow this author [Avicenna]: a
108
Isaac Albalag, ed. Vajda, pp. 267–74.
109
Steinschneider, HÜ, p. 306 (§171); Isaac Albalag, ed. Vajda, p. 270.
110
111
Chertoff, “The logical part of al-Ghazālī’s Maqāsid al-falāsifa.”
112 ˙
Steinschneider, HÜ, pp. 306–9 (§ 172); Renan-Neubauer, Écrivains, pp. 574–80.
113
Published by Steinschneider in his Die Handschriften-Verzeichnisse der Königlichen
Bibliothek zu Berlin. 2. Band: Verzeichniss der Hebraeischen Handschriften (Berlin,
1878), pp. 130–2; summarized in HÜ, pp. 307–8; Harvey, “Why did fourteenth-century
Jews turn to Alghazali’s account?,” p. 372.
114
“We-kalal bo [kol] kawwanotehem ba-yoter nakon she-efshar” (Steinschneider, Die
Handschriften-Verzeichnisse, p. 131).
part of the book al-Shifāʾ, namely [the parts] on physics and metaphy-
sics, by Avicenna; [and his] book al-Najāt, which covers all the seven
sciences in brief.”115 He notes that he added to his translation expla-
natory glosses from these works by Avicenna (and others); but these
glosses are lost.116 Still, Judah’s introduction made it clear to users
of his translation that al-Ghazālī’s work was Avicennian in its
tendencies.
Judah next makes an apologetic move. His intention is to counter
the “philosophizing” individuals who “shriek and utter” and to refute
their positions. Why then should he translate al-Ghazālī’s précis of
the philosophers’ views? Judah argues that al-Ghazālī’s book in
truth (and felicitously!) does not follow its declared goal. The name
“philosopher” refers “to Aristotle and to the Greek philosophers,
who have no religion, and to their counterparts of the Ishmaelite phi-
losophers and even to some extent [be-sippuq] to individuals of our
own faith.” In al-Ghazālī’s summary, however, the philosophers’
views are mingled with those of Avicenna, who “deviates from
Aristotle’s way.” This book, Judah Nathan implies, is therefore not
an account of the – dangerous – ideas of “the philosophers.”
Kawwanot ha-pilosofim will rather allow its readers to defend “our
holy Torah” against its critics, just as “was the intention of the author
of this book [al-Ghazālī], namely to refute in a [further] dedicated
work [namely, the Tahāfut al-falāsifa] everything that contradicts
any of the cornerstones of religion.” Judah Nathan apologetically
remarks that he is aware that the book he translated still “includes
some matters that contradict the cornerstones of our religion, by
which we live”; as a sort of penitence he therefore inserted
Maimonides’ Thirteen Principles in the beginning of the book.
Judah seems to be ambivalent. He does not state that it is specifi-
cally the Avicennian bent of the book that makes it particularly useful
for the defense of the Torah. What seems to have made the book com-
mendable are rather two other features: first, al-Ghazālī’s declared
intention to refute the radical philosophical positions (those Judah
identified with Aristotle’s); second, the book’s brevity, allowing its
readers to learn the foundations of philosophy rapidly, so as to be in
a position to engage in discussions with the holders of radical philoso-
phical positions. The fact that Judah’s annotations to the translation
did not accompany the work implies that the readers of the trans-
lation presumably viewed the book as faithfully presenting a mild
115
For the complete text of this passage, see Steinschneider Die Handschriften-Verzeichnisse,
p. 132; Zonta, “The role of Avicenna and of Islamic ‘Avicennism’,” p. 656; id., “Avicenna in
medieval Jewish philosophy,” pp. 275–6
116
See Steinschneider, Die Handschriften-Verzeichnisse, p. 132, n. 5. Judah says he handed
them out only to “wise men” (ibid., p. 132), a remark indicating that they were transmitted
in separate manuscripts. This easily accounts for their loss.
version of the views of “the philosophers.” We will return to Judah
Nathan’s role in diffusing Avicenna’s thought in fourteenth-century
Jewish philosophy in Hebrew.
As noted by Steinschneider, the three translations of the Maqāsid
˙
are extant in a great number of copies, an indication of the popularity
of al-Ghazālī’s work. Many copies bear marginalia witnessing to
intensive study.117 The work also appears on numerous Jewish book
lists in many cultural contexts,118 including Karaite.119 Moreover,
al-Ghazālī’s composition in Hebrew garb was the object of numerous
commentaries: “few Hebrew works were commented as often as the
Intentions,” Steinschneider observed.120 The first of these commen-
taries, written by the noted Averroist scholar Moses Narboni in
1344–1349, is extant in more than 30 copies.121 This Averroist com-
mentary in turn was a source of inspiration for an encyclopedia
entitled Ahavah be-taʿanugim (Love with Enjoyment), composed in
1354 by the Spanish philosopher Moses ben Judah Nogah.122
Another supercommentary was composed in Italy by Moses Rieti
(1389–after 1452).123 All in all, no fewer than twenty commentaries,
written in southwest Europe as well as in the East, are known
(some of them not extant in full).124 In 1367, the work was versified
by the Abraham Avigdor b. Meshullam (then aged 17), in order to
make the science laid out in al-Ghazālī’s work as accessible as
possible.125
117
Steinschneider, HÜ, p. 310.
118
E.g., the analysis of 430 lists of books owned by Jewish families, compiled in Mantua at the
very end of the sixteenth century, reveals that five (out of 430) families owned manuscripts
of Kawwanot ha-pilosofim; see Shifra Baruchson, Books and Readers. The Reading
Interests of Italian Jews at the Close of the Renaissance (Ramat Gan, 1993), p. 149
(Heb.). A manuscript of the work was also owned by an amateur of philosophy in the
Midi in the sixteenth century; see Jean-Pierre Rothschild, “Quelques listes de livres
hébreux dans des manuscrits de la Bibliothèque nationale de Paris,” Revue d’histoire des
textes, 17 (1987): 291–346, on pp. 322–3 (no. 19).
119
See Ofer Elior, “Ruah hen as a looking glass: the study of science in different Jewish cul-
tures as reflected in ˙the
˙ history of a medieval introduction to Aristotelian science,” Ph.D.
dissertation, Ben-Gurion University, Beer Sheva, Israel, 2010. Elior studies the cultural
history of this short thirteenth-century introduction to science and found that
al-Ghazālī’s Maqāsid fulfilled a similar role.
120
Steinschneider, HÜ, ˙ p. 311.
121
Ibid., p. 315; the commentaries are described ibid., pp. 311–26 (§§ 175–183).
122
Zonta, “The role of Avicenna and of Islamic ‘Avicennism’,” pp. 657–9. On this work see Esti
Eisenmann, “Ahavah be-taʿanugim: a fourteenth-century encyclopedia of science and theol-
ogy,” in Harvey (ed.), The Medieval Hebrew Encyclopedias, pp. 429–40.
123
124
Ibid., pp. 311–25 (§§ 175–182).
125
Ibid., HÜ, pp. 325–6 (§183). The translator’s introduction was published by Steinschneider
as “Miscelle 21,” Monatsschrift für Geschichte und Wissenschaft des Judentums, 37 (1893):
407–9. See also Harvey, “Why did fourteenth-century Jews turn to Alghazali’s account?,”
pp. 373–4.
Al-Ghazālī’s work was certainly a remarkable success. It became a

major conduit of Avicennian ideas into Hebrew philosophical thought.
It is doubtful, however, that the readers of the Hebrew versions of the
Maqāsid embraced specifically Avicennian modes of thought. No
study ˙ dedicated to this question has been conducted, but the odds
are that they did not. The Maqāsid seems to have been perceived
and used simply as a convenient short˙ manual on science, not as an
exposition of anti-Aristotelian (or anti-Averroist) science.126 In this
respect it is comparable, say, to Ruah hen, Midrash ha-Hokhmah,
and Shaʿar ha-shamayim, the popular ˙ ˙thirteenth-century˙ Hebrew
encyclopedias or exposés of science that afforded readers a quick intro-
duction to the basics of science, required for the perusal of works in
religious philosophy. More often than not, authors “lifted” from
al-Ghazālī’s composition separate doctrinal items, disconnected
from the global Avicennian outlook. Moreover, the Maqāsid was
often, perhaps mostly, studied with one of its commentaries, ˙ of
which the most disseminated ones were Albalag’s and Narboni’s,
both of which were Averroist in tendency. (The famous study program
of Yohanan Alemano [1435–after 1504], for instance, includes the
study ˙of the Maqāsid accompanied with these two commentaries.127)
Thus, at least until ˙ more detailed studies show otherwise, the
Maqāsid cannot be construed as a fountainhead of a “Hebrew
˙
Avicennism.” The first scholar to use the Maqāsid to criticize science
may have been Hasdai Crescas at the very end ˙ of the fourteenth
century. 128 ˙
Al-Ghazālī’s Tahāfut al-falāsifa (The Incoherence of the
Philosophers), which also presents – and refutes – Avicennian doc-
trines in great detail, was translated into Hebrew, too. But this trans-
lation was late (first decade of the fifteenth century) and had little
influence.129 It seems to have played only a minor role in the reception
of Avicenna in Hebrew (but a fuller examination of this question may
be useful).
(ii) Averroes
The Spaniard Averroes was Avicenna’s staunch critic and sought
to reestablish the authority of “orthodox” Aristotelianism (as he
126
Harvey, “Why did fourteenth-century Jews turn to Alghazali’s account?”.
127
Moshe Idel, “The study program of R. Yohanan Alemano,” Tarbiz, 48 (1978–1979): 303–31,
˙
on p. 308 (Heb.): “whoever wishes to investigate ˙ ha-dat] will read in the
religion [haqirat
˙
morning The Intentions of the Philosophers with the commentaries of Narboni and Isaac
Albalag and the Incoherence of the Incoherence by Ibn Rushd, and the Kuzari and
Emunah Ramah and the Guide with its commentaries.”
128
As suggested in Harvey, “Why did fourteenth-century Jews turn to Alghazali’s account?,”
p. 376. See also Wolfson, Crescas’ Critique, p. 10.
129
understood it). In his earlier works, notably the epitomes of Aristotle’s
works, his positions on many subjects were not far removed from
Avicenna’s: readers of these works, most of which were translated
into Hebrew, encountered in them positions fairly close to
Avicenna’s, though not labeled as such.130 In his later works, by con-
trast, Averroes takes Avicenna to task, more often than not by name.
These later works by Averroes were also very heavily translated into
Hebrew – both the commentaries and the monographs.131
Particularly important in this context is Averroes’ Tahāfut
al-tahāfut (Incoherence of the Incoherence), the refutation of
al-Ghazālī’s critique of the philosophies of al-Fārābī, and, especially,
Avicenna. This work, translated into Hebrew twice in the first half
of the fourteenth century,132 contained much information on
Avicenna’s ideas, albeit biased: Averroes reported only those of
Avicenna’s ideas he took to task; hence readers of this work could
become aware only of these – Avicenna’s more controversial views.
As a result of the substantial availability of Averroes’ compositions
in Hebrew, Hebrew-reading scholars were aware of Avicenna’s pos-
itions on a great number of issues, albeit through their refutations.
To draw the “image of Avicenna” as it emerges from the works of
Averroes available in Hebrew would possibly be a worthy contribution
to the history of Jewish philosophy in Hebrew; but this cannot be
attempted here.133 We must content ourselves with drawing the read-
er’s attention to Averroes’ translated writings as an important conduit
for Hebrew Jewish cultures’ acquaintance with Avicenna.
(iii) Fakhr al-Dīn al-Rāzī

Although critical of him, some works by Fakhr al-Dīn al-Rāzī (1149–
1209) were openly inspired by Avicenna. Traces of one of them
are found in a fourteenth-century Hebrew translation of ʿUyūn
al-masāʾil (The Fountains of Questions, or: The Principal
130
Pointed out in Davidson, Alfarabi, Avicenna, and Averroes on Intellect, p. 181.
131
See Steinschneider, HÜ, index, s.v. Averroes; Harry A. Wolfson, “Plan for the publication of
a Corpus commentariorum Averrois in Aristotelem,” Speculum, 36 (1961): 88–104, repr. in
id., Studies in the History of Philosophy and Religion, ed. by Isadore Twersky and George
H. Williams (Cambridge, Mass., 1973), vol. I, pp. 430–44.
132
Steinschneider, HÜ, pp. 332–4 (§§ 187, 188–189). There is also a fragment of a third trans-
lation. In the introduction to his translation, Qalonymos ben David proclaims his
Ghazalian sympathies and lauds Ghazālī for his successful critique of philosophy. He
explains that he agreed to translate the defense of philosophy by “the heretic Averroes”
only because friends who admire al-Ghazālī’s work pleaded with him to do so. See the
summary in Steinschneider, HÜ, pp. 332–3 and the text in Steinschneider, Die
Handschriften-Verzeichnisse der Königlichen Bibliothek zu Berlin, pp. 133–5, on p. 133.
133
A pioneering but small step in this direction was taken by Steinschneider, HÜ, p. 280 and
p. 998, n. 26. Miguel Cruz Hernández gathered passages in which Averroes criticizes
Avicenna in his Abu-l-Walid ibn Rušd (Averroes). Vida, obra, pensamiento, influencia 2nd
edn (Córdoba, 1997), pp. 371–5 (we are indebted to Jules Janssens for this reference).
Questions), ascribed to al-Fārābī,134 by Todros Todrosi of Arles (active

˙
1330–1340), whom we have already encountered ˙ as the translator of
al-Najāt. In his translation of this work, Todrosi incorporated two
˙
passages from al-Rāzī’s al-Mabāhith al-mashriqiyya (Oriental
Researches). 135 ˙
These passages bear on the logical concept of “defi-
nition” (taʿrīf) and the Avicennian idea of the chain of emanations of
the intellects.136
It is quite likely that additional Arabic texts containing Avicennian
materials were translated into Hebrew. They await identification by
scholars.137
B.1.3. Avicenna in Hebrew Translations of Judeo-Arabic

Philosophical Works
The Arabic-to-Hebrew translation movement, whose beginnings go
back to the eleventh century, gathered momentum after the Almohad
takeover of Muslim Spain, which compelled Jews (as also Christians)
to leave the country. Many reached the Midi, among them the Ibn
Tibbons, who hailed from Granada. Judah Ibn Tibbon (c. 1120–
1190) settled in Lunel around 1155 and initiated a systematic trans-
lation project under the patronage of Meshullam ben Jacob, a power-
ful and erudite local head of a yeshiva. This enterprise was to be
continued by members of his family for at least three generations.
The first works Judah Ibn Tibbon rendered into Hebrew were in reli-
gious philosophy. The translation of Judeo-Arabic works of this kind
134
See on this work Miguel Cruz Hernández, El “Fontes quaestionum” (ʿuyun al-masaʾil) de
Abu Nasr Al-Farabi (Paris, 1951); Davidson, Alfarabi, Avicenna, and Averroes on
Intellect, pp. 128–9.
135
136
On these two passages and their source, see Mauro Zonta, “Fonti antiche e medievali della
logica ebraica nella Provenza del Trecento,” Medioevo, 23 (1997): 515–94, on p. 567 n. 151;
id., “The role of Avicenna and of Islamic ‘Avicennism’,” p. 655. One of Fakhr al-Dīn’s works
survives in a transcription in Hebrew letters; see Langermann, “Transcriptions of Arabic
treatises into the Hebrew alphabet.”
137
Commenting on an early draft of this study, Prof. Y. Tzvi Langermann kindly informed us
that his study (in progress) of al-Tabrīzī’s commentary on the 25 Propositions with which
Maimonides opens Part Two of the Guide led him to conclude that it “draws heavily upon
Avicenna, especially some Avicennian innovations, such as the notion of mayl (or inclinatio)
in dynamics.” Al-Tabrīzī’s commentary was translated into Hebrew twice: by Isaac ben
Nathan of Majorca in 1347 and by an anonymous translator; see Steinschneider, HÜ,
p. 362–3. It was used by a number of Jewish scholars, most notably Hasdai Crescas.
However, as Langermann has also shown, although “al-Tabrīzī’s book has˙ a great deal of
information about the Avicennan and post-Avicennan developments in the physics of
motion, [. . .] Crescas made no use of this at all; it may just as well have not been trans-
mitted. Neither Crescas nor anyone else in the West – including, it seems now, the Latin
scholastics whose work he may have known, and who famously developed their own the-
ories of impetus – reacted in any significant way to the new physics that reached them
from the East.” See Y. Tzvi Langermann, “No reagent, no reaction: the barren transmission
of Avicennan dynamics to Hasdai Crescas,” Aleph, 12 (2012) (in Honor of Ruth Glasner):
161–88. ˙
continued throughout the thirteenth and fourteenth centuries, in par-
allel to the translation of profane works by non-Jewish authors. The
four Judeo-Arabic compositions discussed above that incorporated
Avicennian doctrinal items indirectly introduced Avicennian ideas –
not earmarked as such – into Jewish thought in Hebrew.
(i) The Kuzari (Book of the Khazar) by Judah Hallevi was one of the
first books translated into Hebrew by Judah Ibn Tibbon (1167).138
This work quickly became – and remains to this very day – a “must”
in most Hebrew-reading intellectual milieus, both among those
inclined towards philosophy and those more circumspect about
“foreign” wisdom.139 Correspondingly, it has also been the subject of
multiple commentaries. Inasmuch as the Kuzari incorporated
Avicenna’s treatise on the soul, its Hebrew version acquainted a
very wide audience with Avicenna’s psychological doctrine, albeit
without any reference to its author.140
(ii) Abraham Ibn Daʾud’s The Exalted Faith, no longer extant in the
original Arabic, was twice translated into Hebrew in Spain: around
1370 by Salomon Ibn Labi, probably a member of the circle of
Jewish scholars around Hasdai Crescas (1340–1412), and in 1391–
1392 by Samuel Ibn Mot˙ot, who apparently revised Ibn Labi’s ver-
sion.141 While Ibn Motot˙’s˙ version had no success (only one manu-
˙ ˙
script survives), Ibn Labi’s circulated widely (eleven manuscripts
extant). In this way, the Avicennian elements in The Exalted Faith
reached a fairly wide readership, though again without attribution
to Avicenna.
(iii) As noted above, Maimonides’ Guide of the Perplexed contains
essential Avicennian building blocks. This foundational text of Jewish
philosophy was translated into Hebrew twice, almost simultaneously:
by Samuel Ibn Tibbon (first edition 1204), whose version became cano-
nical, and by Judah al-Harizi. Other texts by Maimonides, less relevant
in the present context,˙ were also translated into Hebrew during the
thirteenth century. Needless to say, the Guide was the most widely
read philosophical text in Hebrew, and its doctrines were known to a
very large readership throughout the ages. However, as already
explained, the numerous distinctively Avicennian doctrines in the
Guide (and in other philosophical writings by Maimonides available
in Hebrew) could not be recognized as such by the vast majority of the
138
For this and for what follows, see Steinschneider, HÜ, pp. 403–4 (§ 234).
139
On the reception and transmission of the Kuzari, see Adam Shear, The Kuzari and the
Shaping of Jewish Identity, 1167–1900 (New York, 2008).
140
Judah Hallevi’s work was translated a second time, by Judah Ibn Cardinal, but this trans-
lation had no impact whatsoever and is thus of no relevance in the present context; see
141
About these two versions of The Exalted Faith, see Steinschneider, HÜ, pp. 368–71 (§ 211);
Eran, From Simple Faith to Sublime Faith, pp. 22–5.
readers, except when enlightened by arabophone commentators (see

below). (Maimonides’ Hebrew composition Sefer ha-Maddaʿ will be dis-
cussed in B.1.4.)
The Guide of the Perplexed, the Kuzari, and to a lesser degree The
Exalted Faith all belong to the canon of medieval Jewish philosophy.
They introduced Avicennian doctrines – concerning especially the
notion of the Necessary Existent (Maimonides) and the theory of soul
(Kuzari and Exalted Faith) – into the Jewish philosophical discourse.
For the majority of their readers, we must remember, these works
were authoritative works by respected Jewish masters, so adherence
to their doctrines was almost de rigueur. It should be emphasized
again that none of these works ever referred to Avicenna by name;142
the Avicennian ideas introduced into Jewish philosophy through
these works were not labeled as going back to Avicenna. Only learned
arabophone readers who were acquainted with Avicenna through their
readings in Arabic could identify the Avicennian doctrinal items
encountered in these works. The Guide is to some extent a special
case: in their erudite commentaries, Shem-Tov Ibn Falaquera and
Moses Narboni did report the Avicennian origin˙ of certain ideas.143
Presumably, however, their observations did not essentially alter
most readers’ perception that the doctrines in question were
“Aristotle’s,” as stated by Maimonides. So the translations of these
influential Judeo-Arabic works introduced Avicennian ideas into
Jewish philosophical thought, even as Avicenna himself remained
mostly unknown.
Two further translations need to be mentioned, although they had a
very small circulation and impact:
(iv) Moses ha-Levi’s Metaphysical Treatise, which defends some key
ideas of Avicenna’s metaphysics (above, A.1.2 [iii]), was translated
into Hebrew (probably in the late thirteenth century), but only a
single manuscript survives.144 It may have been translated into
Hebrew another time, as part of Joseph Ibn Waqār’s Treatise
Reconciling Philosophy and Religious Law, to which we now turn.
(v) Joseph Ibn Waqār’s Arabic Treatise Reconciling Philosophy and
Religious Law (supra, A.1.2 [iv]) was apparently translated into
142
The significance of this fact is not clear. Yossef Schwartz has pointed out a general ten-
dency among Jewish Andalusian writers to refer to their Jewish sources but not to the
non-Jewish ones, and an opposite practice in Maimonides. See Yossef Schwartz,
“Zwischen Philosophie und Theologie im 12. Jahrhundert: Halevi, Ibn Daud und
Maimonides,” in Alexander Fidora and Andreas Niederberger (eds.), The Relation between
Metaphysics and Theology in the Philosophical Discussion of the Twelfth Century
(Turnhout, 2004), pp. 113–35, on pp. 122–5. Thus, the fact that these authors do not men-
tion Avicenna is not necessarily a consequence of a particular attitude toward him. We are
grateful to Prof. Yossef Schwartz for an exchange on this point.
143
See above, n. 15.
144
Oxford, Bodleian Library, Opp. Add. 4° 10, fols. 115r–125r.
Hebrew, although only a fragment of this translation survives.145 If
the translation was indeed complete, it must have included also
Moses ha-Levi’s Avicennian treatise. The same applies to two explicit
quotations from Fakhr al-Dīn al-Rāzī that Joseph Ibn Waqār inserted
in his work: from al-Rāzī’s commentary on Avicenna’s al-Ishārāt
wa-al-tanbīhāt (Pointers and Reminders), and from the Lubāb
al-Ishārāt (The Pith of [Avicenna’s] Ishārāt).146
Given the small circulation of the last two works, their contribution
to the knowledge of Avicenna among Hebrew-reading students of phil-
osophy was nil. Indeed, one can wonder whether the near disappear-
ance of these two works was not related to the general lack of interest
in Avicenna’s thought in Jewish circles.
B.1.4. Avicenna in Philosophical Works by Arabophone Jewish

Scholars Writing in Hebrew
The first major composition to expound philosophical matters in
Hebrew was Sefer ha-Maddaʿ (Book of Knowledge), the opening part
of Maimonides’ great code of Jewish Law, the Mishneh Torah.
Maimonides composed this work in Hebrew in Egypt (concluded
1178–1180); it reached Europe in the last decade of the twelfth cen-
tury. As already explained, the first chapters of Sefer ha-Maddaʿ pre-
sent the basics of the Avicennian notion of the Necessary Existent and
the contingent world in dogmatic fashion (meaning without argumen-
tation). This was the first substantial exposition of these Avicennian
ideas in Hebrew, although, again, they were not labeled as such.
Inasmuch as Sefer ha-Maddaʿ is part of Maimonides’ authoritative
legal code, written in Hebrew, his statement of fundamental
Avicennian tenets made them (albeit in a very simplified form) obliga-
tory articles of belief for numerous Jews throughout the ages.
(Maimonides also incorporated basic Avicennian tenets in a succinct
formulation in his Thirteen Principles of Faith.) The importance of
this channel of transmission cannot be overemphasized: it was cer-
tainly the most significant conduit of Avicennian ideas to Jewish
minds from the late twelfth century on. Indeed, these ideas became
so well known as to appear trite, rather than as one way of construing
the relationship between the deity and the world.
Even after most Jewish authors began writing on philosophy in
Hebrew, many scholars (though by no means the majority) could
still read Arabic: in some families, especially of physicians, Arabic
was passed down from generation to generation. Some of these
bilingual and bicultural intellectuals became Arabic-to-Hebrew
145
Rome, Vatican Library, vat. ebr. 384, fols. 125r–139v.
146
On these quotations whose origin has not yet been identified, see Vajda, Recherches,
pp. 125–6.
translators; others wrote books in Hebrew that drew on knowledge

available only in Arabic and not accessible to scholars who knew
Hebrew only. Here we will examine information about Avicenna
included in the Hebrew works of this small number of arabophone
Jewish scholars.
(i) Abraham Ibn Ezra (1089–1167) has already been mentioned
above as the adapter of Hayy ben Yaqzan. But he also adhered to
and expressed some Avicennian˙ views.˙ In his biblical commentary
he shows, en passant, that he is aware of and accepts Avicenna’s
notion of God as the Necessary Existent.147 More important, accord-
ing to one of its possible interpretations, his obscure and elusive exeg-
esis of the account of creation in Genesis follows Avicenna, much along
the lines of the interpretation put forward by Samuel Ibn Tibbon a few
decades later (see immediately below).148 However, Ibn Ezra does not
mention Avicenna; only commentators familiar with the latter’s the-
ory could identify it in Ibn Ezra’s exegesis. We will come back to
this shortly. It has been suggested in the past that Abraham Ibn
Ezra referred to Avicenna (as “Abencine”) in his Liber de rationibus
tabularum (preserved in Latin), but J. L. Mancha has recently argued
persuasively that “Abencine” in fact refers to Ibrāhīm ibn Sinān ibn
Thābit ibn Qurra (908–946).149
(ii) Samuel Ibn Tibbon (c. 1160–1232), the well-known translator of
Maimonides’ Guide, was an accomplished scholar and original thinker
who introduced Avicenna into Hebrew thought in two of his works. The
first, which does so in a minor way, is his meticulous Arabic-to-Hebrew
version of Aristotle’s Meteorology, finished in 1210. Here Ibn Tibbon
complements and elucidates the difficult text with glosses based on
information culled from various sources, among them Avicenna.
(However, their brevity does not allow a precise identification of the
works to which Ibn Tibbon had access then.)150
More consequential is the use Ibn Tibbon made of Avicenna in his
most important original composition, Maʾamar Yiqqawu ha-mayim
(Treatise on “Let the Water Gather,” concluded c. 1231), which
addresses the question of why the element water does not entirely
cover the surface of the terrestrial globe (as it should according to
147
See Warren Z. Harvey, “The first commandment and the God of history: Halevi and Crescas
vs. Ibn Ezra and Maimonides,” Tarbiz, 57 (1988): 203–16, on p. 208 (noted in S. Harvey,
˙
“Avicenna’s influence on Jewish thought,” p. 329).
148
For a review of all the interpretations of Ibn Ezra’s obscure statements, see Avraham
Lifschitz, “Le-torat ha-beri’ah šel R. Avraham Ibn Ezra,” Sinai, 84 (1979): 105–25 (Heb.).
149
José Luis Mancha, “The theory of access and recess in Levi ben Gerson’s astronomy and its
sources,” Aleph, 12 (2012) (in Honor of Ruth Glasner): 37–64, on p. 55, n. 21.
150
See Otot ha-Shamayim. Samuel Ibn Tibbon’s Hebrew version of Aristotle’s Meteorology: A
Critical Edition, with Introduction, Translation, and Index by Resianne Fontaine
(Leiden, 1995), Introduction, pp. XVI, LVII–LIX. See in particular passages II.438, 480–
1, 489–91, III.96–101. We thank R. Fontaine for her kind assistance.
the four-element theory). Ibn Tibbon recounts that he had been
engaged with this problem for twenty years but failed to find a satis-
factory solution. (He rejected notably Averroes’ account as not natur-
alistic.) The breakthrough came, he tells his readers, through
Avicenna’s “great book al-Shifāʾ,” in which he finally discovered a sat-
isfactory naturalistic solution to the worrisome problem. Ibn Tibbon
next inserts in his text his own Hebrew translation of a lengthy pas-
sage from the section on meteorology in al-Shifāʾ and proceeds to
show that Avicenna’s naturalistic theory – dry land repeatedly
emerges from under the water and subsequently is covered by it
again – accords with the biblical texts.151 An essential premise of
Avicenna’s theory is that after each flooding of the land humans are
generated again “not from man,” that is, through “spontaneous gener-
ation,” like any other animal species. This bold, indeed blasphemous
statement is included in the passage quoted from Avicenna, and Ibn
Tibbon was audacious enough to endorse it explicitly.152 Ibn Tibbon
incorporated in Yiqqawu ha-mayim another passage from al-Shifāʾ,
from the section on “Generation and Corruption” (but without indicat-
ing its source): it gives a concise account of the theory of the mutual
transformation of the four elements, albeit with an original and con-
sequential twist that provides the basis for the theory of the infinite
succession of sublunary worlds.153 It would therefore seem that he
had access to the entire work, or at least to large parts of it.
Maʾamar Yiqqawu ha-mayim is thus based on two distinctively
Avicennian theses; this makes Ibn Tibbon one of the foremost
Avicennian thinkers in medieval Hebrew culture. Indeed – and it is
important to emphasize this point – unlike most other authors who
integrated Avicennian ideas into their writings, Samuel Ibn Tibbon
proclaims his indebtedness to al-Shifāʾ loudly and explicitly.
Ibn Tibbon’s Maʾamar Yiqqawu ha-mayim was quite successful; at
least fourteen manuscripts survive. But the passage translated from
Avicenna, including the quotation about man’s possible generation
151
As just noted, according to some interpreters, this Avicennian theory was held by Abraham
Ibn Ezra as well.
152
Samuel Ibn Tibbon, Maʾamar Yiqqawu ha-mayim, ed. Mordecai Bisliches (Pressburg,
1837), p. 7. The passage is translated in Gad Freudenthal, “(Al-)chemical foundations for
cosmological ideas: Ibn Sīnā on the geology of an eternal world,” in Sabetai Unguru
(ed.), Physics, Cosmology and Astronomy, 1300–1700: Tension and Accommodation,
Boston Studies in the Philosophy of Science 126 (Dordrecht/Boston/London, 1991),
pp. 47–73; repr. in id., Science in the Medieval Hebrew and Arabic Traditions, Variorum
Collected Studies Series, 803 (Aldershot, 2005), Essay XII. Samuel Ibn Tibbon’s theory
is studied in detail in id., “Samuel Ibn Tibbon’s Avicennian theory of an eternal world,”
Aleph, 8 (2008): 41–129.
153
Ibn Tibbon, Maʾamar Yiqqawu ha-mayim, pp. 3-5; Freudenthal, “Samuel Ibn Tibbon’s
Avicennian theory of an eternal world,” Appendix B, esp. pp. 118–19 (with n. 144, in
which M. Ahmed Hasnaoui of Paris is thanked for his help in the identification of the pas-
sage in al-Shifāʾ).
“not from man,” had an incomparably larger diffusion owing to its

unaltered incorporation into Gershon ben Solomon’s Shaʿar
ha-Shamayim (The Gate of Heaven),154 one of the most popular med-
ieval Hebrew works of science; more than thirty manuscripts survive
and it was printed several times. Through its inclusion in these two
works, this account became one of those most commonly associated
with Avicenna’s name.155 Some supercommentaries on Abraham
Ibn Ezra’s commentary on Genesis argued that his explanation is
identical to Avicenna’s theory, which they summarized with explicit
attribution.156 Whether they knew it from Samuel Ibn Tibbon or
from other sources remains to be seen.
The fact that this Avicennian theory was publicized so early (1231)
and was so widely disseminated may have played a role in the later
phases of the reception of Avicenna in Hebrew. Ibn Tibbon had a repu-
tation as a radical rationalist;157 his endorsement of Avicenna may
have backfired and deterred translators and potential patrons from
any attempt to render works by Avicenna into Hebrew. It thus
seems possible that, through the widely circulated Maʾamar
154
See Shaʿar ha-shamayim 2:1 (Warsaw, 1876), pp. 13–14; already noted in Steinschneider,
HÜ, p. 14.
155
It is quoted e.g. by the kabbalistically minded Samuel Motot (end of fourteenth century), in
his supercommentary on Abraham Ibn Ezra’s commentary ˙ ˙ on Genesis: “You know that
there is a controversy among the scientists on whether the formation of a human without
the copulation of a male and a female is possible. And the wise Avicenna wrote that it is not
impossible. Among the recent [scholars] there are some who decided in favor of Avicenna
and said that at the equator, where the air is balanced with respect to cold and heat and
the other qualities, it occasionally occurs.” Samuel Motot, Megillat setarim (Venice,
1554), fol. 7b-c (noted in Steinschneider, “Anzeigen,” p. 22).˙The
˙ first part of the statement
comes either from Ibn Tibbon or from Shaʿar ha-shamayim; the second refers to Hayyim
Israeli’s Maʾamar Gan Eden, discussed below. ˙
156
See, e.g., Joseph Bonfils (Tob Elem), Sophnat Paneʾah. Ein Beitrag zur Pentateuchexegese
des Mittelalters, ed. David ˙
˙ Herzog (Heidelberg, ˙ 1930), vol. 1, p. 36 (composed in
1911,
Jerusalem around 1385); we are grateful to R. Leicht for this reference. On the supercom-
mentaries on Ibn Ezra’s biblical commentaries, see Uriel Simon, “R. Abraham Ibn Ezra –
the exegete who became the object of exegesis” (Heb.), in Sara Yafet (ed.), Ha-miqraʾ bi-reʾi
mefarshaw. Sara Kamin Memorial Volume (Jerusalem, 1994), pp. 367–411 [translated as
“Interpreting the interpreter: supercommentaries on Ibn Ezra’s commentaries,” in Isadore
Twersky and Jay M. Harris (eds.), Rabbi Abraham Ibn Ezra: Studies in the Writings of a
Twelfth-Century Jewish Polymath (Cambridge, Mass., 1993), pp. 86–121]; see also Dov
Schwartz, Yashan be-qanqan hadash (Old Thought in New Vessels) (Jerusalem, 1996),
about the supercommentaries˙ on Ibn Ezra written by a northern Spanish circle of
Jewish authors in the second half of the fourteenth century.
157
On the perception of Samuel Ibn Tibbon as an impious radical, see, e.g., the criticism by
Jacob ben Sheshet in his Sefer Meshiv devarim nekohim, ed. Georges Vajda,
Introduction by Georges Vajda and Efraim Gottlieb (Jerusalem, ˙ 1968); Jacob ben
Sheshet’s criticism is studied by Gad Freudenthal, “The Kabbalist R. Jacob ben Sheshet
of Gerona: the ambivalences of a moderate critique of science (ca. 1240),” in Proceedings
of the Symposium “Times and Places of Jewish Girona”, Girona Judaica, vol. 5 (Girona
2011), pp. 275–89. Ibn Tibbon was persona non grata among traditionalist and even mildly
rationalist circles; indeed, in the decades following the publication of Maʾamar Yiqqawu
ha-mayim, parts of it were hidden by its supporters so as to avoid putting ammunition
in the hands of Ibn Tibbon’s detractors (Sefer Meshiv devarim nekohim, 11: 75–76.).
˙
Yiqqawu ha-mayim and Shaʿar ha-Shamayim, Avicenna became
associated not only with the heterodox thesis of the eternity of the
world (which was bad enough), but also with the provocative thesis
that man can be “spontaneously” generated. Avicenna’s identification
with these blasphemous doctrines, we thus suggest, may have contrib-
uted to his being discredited for the Hebrew-reading public and
thereby to the fact that his works were not selected for translation
into Hebrew. We will return to the reasons for the paucity of
Hebrew translations of Avicenna in the Conclusion.
(iii) Moses ben Samuel Ibn Tibbon ( fl. 1244–1283) was far more con-
servative and less bold than his father. This is immediately evident
from his commentaries, which constitute the great bulk of his own
writing.158 But he wrote also an original work on science, which con-
sists of a series of responses to questions addressed to him by a young
relative. This Answers to Queries on Physics is preserved in a single
manuscript.159 The questioner takes Samuel Ibn Tibbon’s Maʾamar
Yiqqawu ha-mayim as the point of departure of his queries. In his
answers, Moses Ibn Tibbon distances himself from most of his father’s
positions. He is clearly acquainted with various parts of the section on
physics in the Kitāb al-Shifāʾ, but more often than not rejects
Avicenna’s positions in favor of Averroes’. One important issue
discussed is that of elemental transformations. Moses Ibn Tibbon dis-
misses Avicenna’s notion of “intermediate transformation,” which was
the very foundation of Avicenna’s (and his father’s) innovative theory
of the eternity of the world, and instead accepts Averroes’ theory.
Consequently, Moses Ibn Tibbon also rejects Avicenna’s idea of
a “dynamic stability” of the sublunar world, which was the core of
his father’s Maʾamar Yiqqawu ha-mayim. Similarly, he sides with
Averroes against Avicenna on the causes of the elemental forms.
Moses Ibn Tibbon’s Answers to Queries on Physics was little read.
This is understandable, given that the confrontation between
Avicenna and Averroes was passé soon after it was composed. Moses
Ibn Tibbon indeed exemplifies Jewish philosophers’ almost wholesale
adherence to Averroes from the middle of the thirteenth century
onward. He is also one of its promoters: as the translator (between
1244 and 1258) of all of Averroes’ epitomes of Aristotle’s works on phy-
sics and metaphysics and of one middle commentary (of De anima), he
helped spread Averroes’ philosophy in Hebrew, to the exclusion of
158
Hayyim Kreisel, Colette Sirat, and Abraham Israel (eds.), The Writings of R. Moshe Ibn
˙
Tibbon (Beer Sheva, 2010) (Heb.).
159
This work (preserved in Parma, Biblioteca Palatina, MS parmense 2620, ff. 91v–100v) has
not been published. A critical edition is being prepared by Hagar Kahana-Smilansky. The
following brief remarks are all based on her study: “Moses Ibn Tibbon’s Answers to Queries
on Physics: sources and problems,” Aleph, 12 (2012) (in Honor of Ruth Glasner), which in
turn is based on her forthcoming critical edition.
alternatives. Moses Ibn Tibbon, and specifically his rejection of his

father’s Avicennian ideas (and ideals), epitomizes the Averroist turn
of Jewish philosophy in Hebrew.
(iv) As noted, Todros Todrosi of Arles translated Avicenna’s
al-Najāt (1334–1340). ˙ We should
˙ now take notice of another under-
taking with an Avicennian component, namely his Hebrew “philoso-
phical anthology” (Liqqutim, i.e. collectanea, compilation), dated
1334, which includes a number˙ of passages from Avicenna.160 Todrosi
begins his work with logic: this section includes a translation ˙ of
almost the full text of the first four chapters of this section of
al-Shifāʾ (recapitulating Porphyry’s Eisagoge) and some passages
from the section on the Prior Analytics, including Avicenna’s own
classification of Aristotelian syllogisms. Todrosi complements these
long quotations with a few short passages˙ taken from the logical sec-
tion of Avicenna’s Remarks and Admonitions and from al-Ghazālī’s
and Fakhr al-Dīn al-Rāzī’s interpretations of Avicennian ideas
about Aristotelian logic. The inclusion of quotations from authors
with Avicennian tendencies suggests that when compiling his anthol-
ogy Todros was guided by an interest in Avicennian ideas running
against ˙ the dominant Averroist consensus.161 However, the anthol-
ogy, now preserved in a single manuscript,162 probably incomplete,
clearly had a very limited circulation; it therefore bespeaks the inter-
est of an individual rather than a widespread intellectual tendency.
In another work, his Arabic-to-Hebrew version of pseudo-Fārābī’s
ʿUyūn al-masāʾil (The Fountains of Questions, or: The Principal
Questions), Todros added in the margin of a certain passage two
“Avicennian”˙ quotations – one from Fakhr al-Dīn al-Rāzī and the
other from the metaphysical section of Avicenna’s al-Najāt.163
Unless Todros found these quotations already added to his Arabic
manuscript˙ of ʿUyūn al-masāʾil, he must have been well informed
about Arabic-Islamic Avicennism and deemed it relevant to Jewish
readers of his time.
Todros Todrosi undoubtedly marks an interest in Avicennian
˙
thought. ˙ only is he the only Hebrew translator of a scientific trea-
Not
tise by Avicenna; in his anthology he also gathered and translated
other texts of the Avicennian tradition. Much the same holds
of Todros’ friend Judah Nathan,164 who, we saw, translated
˙
al-Ghazālī’s Maqāsid into Hebrew, taking care to fill out unclear
˙
160
See Zonta, “Fonti antiche e medievali della logica ebraica,” pp. 568–72, 575; id., “The role of
Avicenna and of Islamic ‘Avicennism’,” pp. 653–5.
161
For details, see Zonta, “The role of Avicenna and of Islamic ‘Avicennism’,” pp. 654–5.
162
London, British Library, Add. 27559.
163
See above, § B 1.2. (iii), and Steinschneider, HÜ, pp. 293–294; Zonta, “The role of Avicenna
and of Islamic ‘Avicennism’,” p. 655.
164
Zonta, “The role of Avicenna and of Islamic ‘Avicennism’,” p. 655.
passages by drawing on works by Avicenna. The two scholars in fact
acted in tandem. We know something of Judah Nathan’s motives for
appropriating Avicennian bodies of knowledge into Hebrew; perhaps
he shared them with Todros. As already noted, Judah Nathan says
that al-Ghazālī’s work,˙ whose affinity with Avicenna he recognized,
would allow one to rebuke the stance of “the philosophers” and
thereby support that of the Torah. He observed that al-Ghazālī “devi-
ates” from Aristotle, meaning from the Averroean philosophy that was
the focus of the controversies over the legitimacy of the study of phil-
osophy in the early fourteenth century. Judah’s remark followed the
intention to get the translated work of philosophy across to the adver-
saries of the contested “philosophy” identified with “Aristotle.” Put
differently, Judah Nathan – and presumably Todros Todrosi – seem
to have had a clear philosophical agenda, ˙ namely, ˙ to promote
Avicenna and al-Ghazālī as a philosophical alternative to radical
Averroism.
Does this make Todros Todrosi and Judah Nathan into major
Avicennians, on a par ˙ with ˙ Samuel Ibn Tibbon or Moses ha-Levi?
Probably not. Considering their choices of works for translation
(and assuming that these choices reflect the personal preferences of
the translators, not those of some patrons), both appear to be eclectic:
Todros translated not only a work by Avicenna, but also works by
˙
Averroes, 165 whereas Judah Nathan translated medical works,
too.166 Since neither author composed an original work that would

have allowed us access to his own thought, it seems preferable not
to describe them as Avicennians.167
Instead, the two scholars can be described as champions of moder-
ate philosophical study. The controversy over the legitimacy of the
study of philosophy of 1305 revealed that the early split between
arch-conservative anti-philosophical fundamentalists and radical
Maimonidean philosophers had given way to a more stratified picture.
After almost a century of controversies about the value of and danger
to faith of the study of philosophy, most intellectuals of a conservative,
anti-philosophical temper ended up persuaded that the study of phil-
osophy was indispensable, even for themselves, and that a radical
rejection of the study of philosophy was an untenable position.168
165
Steinschneider, HÜ, pp. 62–5 (§ 21), 182, 197; Renan-Neubauer, Écrivains, pp. 224–7.
166
Steinschneider, HÜ, pp. 735 (§ 472), 738–9 (§ 475), 781.
167
In his introduction to the translation of Averroes’ commentary on the Rhetoric, Todros does
not mention any author except Averroes and says that he did the translation for˙the benefit
of “our brothers who seek out philosophy” (ahenu dorshei ha-pilosofiah); see Berzin, “The
˙
medieval Hebrew version of psychology in Avicenna’s Salvation (Al-Najāt),” p. 188. This
is not how a committed Avicennian would write.
168
See, e.g., Gad Freudenthal, “Holiness and defilement: the ambivalent perception of philos-
ophy by its opponents in the early fourteenth century,” Micrologus, IX (2001): Gli Ebrei e le
Scienze. The Jews and the Sciences (2001): 169–93 [repr. in Science in the Medieval Hebrew
These “moderate conservatives” thus shared some interests with the

committed students of philosophy, although their objective in study-
ing philosophy was not philosophy per se, but rather a philosophical
defense of Torah study. The translation of the Maqāsid was appar-
˙
ently intended for these circles of intellectuals who wished to tread
a mean path between the radical philosophers and the staunch anti-
philosophical conservatives and who needed adequate intellectual
tools, notably works of science that presented an alternative to
Aristotle.
(v) Ruah hen is a small summary of the so-called Aristotelian
˙ ˙
sciences, composed in the Midi or southern Italy in the second third
of the thirteenth century.169 Its anonymous author produced an ele-
gant synthesis of materials deriving from numerous unidentified
sources, presumably in Arabic. As Ofer Elior has recently shown,
the discussions in Ruah hen include some distinctively Avicennian
˙ ˙
doctrinal items: for example, the notion that the qualities derive
from the elements’ forms, rather than define them; the definitions of
“priority”; the idea that the change of form of a substance occurs in
no time; the definition of change in the category of position. Ruah
hen was read and studied throughout many centuries in different cir- ˙
˙cles, but its numerous readers could not be aware of the Avicennian
character of the doctrinal items presented in it.
(vi) Judah ben Solomon ha-Kohen Ibn Matqah composed his
philosophical-scientific encyclopedia Midrash ha-hoḵ mah (The
Study of Science) in Arabic (now apparently lost) in ˙Toledo around
1235 and translated it himself into Hebrew in 1247, in Sicily, under
the patronage of Frederick II.170 In his very dense and summary expo-
sitions of Aristotelian science, Judah as a rule follows Averroes, but
occasionally complements his accounts with Avicenna’s ideas (with
or without explicit attribution); it has not yet been established
whether he took them directly from Avicenna’s works or indirectly
from Averroes. For example, in the section on meteorology, Judah
interpolates a remark to the effect that Avicenna (explicitly named)
holds that the region of the equator is more moderate than the
and Arabic Traditions, Essay II]; Dov Schwartz, “Changing fronts in the controversies over
philosophy in medieval Spain and Provence,” Journal of Jewish Thought and Philosophy, 7
(1997): 61–82, revised version in id., Central Problems of Medieval Jewish Philosophy
(Leiden, 2005), pp. 117–36; Gregg Stern, Philosophy and Rabbinic Culture: Jewish
Interpretation and Controversy in Medieval Languedoc (London, 2009).
169
Steinschneider, HÜ, pp. 426–8 (§ 246). The paragraph that follows entirely depends on the
findings of Ofer Elior’s Ph.D. dissertation, “Ruah hen as a Looking Glass.” We are indebted
˙ findings
to Dr. Elior for permission to describe a few of his ˙ prior to their publication by him.
170
Steinschneider, HÜ, pp. 1–4 (§ 1); Resianne Fontaine, “Judah ben Solomon ha-Cohen’s
Midrash ha-Hokmah: its sources and use of sources,” in Harvey (ed.), Medieval Hebrew
Encyclopedia, pp. 191–210.
other climates, whereas Averroes holds the opposite view; the section
on metaphysics contains a reference (again explicit) to Avicenna’s the-
ory of the necessary and contingent beings and its refutation by
Averroes. There are as well borrowings from Avicenna not identified
as such.171 The mathematical section of the work also draws on
al-Shifāʾ.172
(vii) Shem Tov Ibn Falaquera (c. 1225–1295) was an erudite and
˙
extremely well-read author who lived in Provence or northern Spain
and played a pivotal role in the Arabic-to-Hebrew cultural transfer,
notably by composing in Hebrew works grounded in his deep knowl-
edge of Arabic science and philosophy.173 He read a substantial num-
ber of works by Avicenna, including al-Shifāʾ, al-Najāt, the Epistle on
the Soul, and the Epistle on the Parts of Sciences, as well as at least
one medical work (albeit with much philosophical content), The
Cardiac Drugs.174 Falaquera, who pursued the goal of acquainting
the Hebrew reading public with Arabic philosophical lore, was cer-
tainly a most important agent of transmission of Avicennian ideas
in Hebrew. Here we cannot survey the entire range of quotations
from Avicenna found in Ibn Falaquera’s voluminous writings and
must content ourselves with a few representative examples.
In his Reʾshit Hoḵ mah (The Beginning of Science, c. 1250), Ibn
Falaquera devotes˙ a section to the division of the sciences, in which
he included a partial translation (unattributed) of Avicenna’s
Epistle on this subject.175 We already noted that manuscripts of this
Arabic text in Hebrew letters circulated among arabophone Jews, tes-
tifying to their interest in it. Ibn Falaquera’s short popular treatise for
laymen, Sefer ha-Mevaqqesh (The Book of the Seeker), includes an
adaptation of a passage on logic derived from the beginning of
al-Najāt.176 In these synthetic works the Avicennian ideas (as well
as those derived from other authors) are incorporated into the text
without crediting their originators.
171
We are indebted to Resianne Fontaine, who is editing parts of Midrash ha-hokhmah, for
this information. ˙
172
Tony Lévy, “Mathematics in the Midrash ha-Hokhmah of Judah ben Solomon ha-Cohen,”
in Harvey (ed.), Medieval Hebrew Encyclopedias,˙ pp. 300–12, on p. 309.
173
Steinschneider, HÜ, pp. 5–9 (§ 2); Raphael Jospe, Torah and Sophia: The Life and Thought
of Shem Tov Ibn Falaquera (Cincinnati, 1988).
174 ˙
For an overview, see Zonta, “Linee del pensiero ebraico,” pp. 456–7; “Avicenna in medieval
Jewish philosophy,” pp. 269–270; “The role of Avicenna and of Islamic ‘Avicennism’,”
pp. 647–8. Cf. also Mauro Zonta, “Hebrew transmission of Arabic philosophy and science:
a reconstruction of Shem Tov Ibn Falaquera’s ‘Arabic library’,” in Mauro Perani (ed.),
˙
L’interculturalità dell’Ebraismo (Ravenna, 2004), pp. 121–7.
175
See Mauro Zonta, “The reception of al-Fārābī’s and Ibn Sīnā’s classifications of the math-
ematical and natural science in the Hebrew medieval philosophical literature,” Medieval
Encounters, 1 (1995): 358–92.
176
Zonta, “Linee del pensiero ebraico,” pp. 456–7; the passage concerns the two Avicennian
philosophical key terms tasawwur and tasdīq (conceiving and verifying).
˙ ˙
Ibn Falaquera’s most wide-ranging work is Deʿot ha-pilosofim (The

Opinions of the Philosophers; c. 1270), a comprehensive encyclopedia
summarizing what “the philosophers” had to say about various pro-
blems. Avicenna is cited frequently, usually by name. For example,
one of the main sources of Book 8, Section 1, on the nature of the
human intellect, is Avicenna’s Epistle on the Soul (which, we saw,
was also incorporated into the Kuzari); similarly, his discussion of
hads quotes from al-Shifāʾ.177 Elsewhere, the work draws on some
˙
passages of the zoological section of al-Shifāʾ as well as of The
Cardiac Drugs. Similarly, Falaquera’s Sefer ha-Nefesh (Book on the
Soul) is full of Avicennian doctrines, borrowed from al-Najāt and
al-Shifāʾ.178 While Deʿot ha-pilosofim impresses for its author’s erudi-
tion and synthetic capacity, it was little read (only two incomplete
manuscripts are extant). More important and influential was Ibn
Falaquera’s Moreh ha-Moreh (Guide to the Guide, 1280), an ambitious
detailed commentary on Maimonides’ Guide that meticulously com-
ments on Maimonides’ text from the stance of an expert in the history
of philosophy. Avicenna (along with Aristotle and Averroes) is one of
the most quoted authors. Ibn Falaquera was the first scholar to
point out Maimonides’ dependence on Avicenna and to corroborate
this statement with explicit quotations, notably from al-Shifāʾ and
al-Najāt.179
The knowledgeable and prolific Shem Tov Ibn Falaquera was cer-
tainly one of the most important conduits˙ of knowledge of Avicenna
into Hebrew culture. He was acquainted with Maimonides’ letter to
Samuel Ibn Tibbon, but, like the addressee himself, was not
negatively impressed by Maimonides’ evaluation of Avicenna; quite
the contrary, he presumably understood Maimonides as recommend-
ing Avicenna (even if less than al-Fārābī).180 Falaquera discussed
177
Steven Harvey, “Shem-Tov Ibn Falaquera’s Deʿot ha-filosofim: its sources and use of
˙
sources,” in id. (ed.), Medieval Hebrew Encyclopedias, pp. 211–37, on p. 232. This notion
was of great importance to Avicenna; see, e.g., Dimitri Gutas, Avicenna and the
Aristotelian Tradition. Introduction to Reading Avicenna’s Philosophical Works (Leiden,
1988), pp. 159–77.
178
Jospe, Torah and Sophia, pp. 181–9; Berzin, “The medieval Hebrew version of psychology
in Avicenna’s Salvation (Al-Najāt),” pp. 138–49, juxtaposes parallel sections from al-Najāt
in Todros’ translation and from Falaquera’s Sefer ha-Nefesh.
179
For˙ a detailed analysis of the use of Avicenna in Moreh ha-Moreh see Shem-Tov ben Joseph
Falaquera, Moreh ha-Moreh. Critical Edition, Introduction and Commentary ˙ by Yair
Shiffman (Jerusalem, 2001), pp. 48–60 (Heb.); see also n. 15 above. The work includes
some passages on metaphysics from al-Shifā’; for a complete list see Mauro Zonta,
“Avicenna’s Metaphysics in the medieval Hebrew philosophical tradition. A short historical
sketch of its evident traces,” in Dag Nikolaus Hasse and Amos Bertolacci (eds.), The Arabic,
Hebrew and Latin Reception of Avicenna’s “Metaphysics” (Berlin-Boston, 2012), pp. 153–8.
180
We are grateful to Hagar Kahana-Smilansky for this suggestion. For Falaquera’s acquain-
tance with Maimonides’ letter to Samuel Ibn Tibbon, see Steven Harvey, “Did Maimonides’
Letter to Samuel Ibn Tibbon determine which philosophers would be studied by later
Jewish thinkers?,” The Jewish Quarterly Review, 83 (1992): 51–70, on p. 63.
Avicenna at great length and repeatedly pointed out that Maimonides
himself was indebted to him. It is important to note that in his works
for the intellectual elite (Deʿot ha-pilosofim and Moreh ha-Moreh)
Falaquera refers to Avicenna explicitly. This is significant, inasmuch
as Judeo-Arabic writers usually incorporated Avicennian ideas or
quotations only implicitly. Consequently, Falaquera not only injected
Avicennian ideas into Hebrew philosophical thought, but made his
readers aware of their origin. In this way he deepened their acquain-
tance with Avicenna’s thought despite the paucity of Hebrew trans-
lations of his works. Specifically, his Moreh ha-Moreh identified the
Avicennian origins of many of the ideas expounded in the Guide.
Undoubtedly Moreh ha-Moreh (preserved in more than ten manu-
scripts) and Falaquera’s other numerous and informed writings
were a major source of information about Avicenna’s thought for
Hebrew-reading scholars and were probably more important in this
respect than any of the other works considered in this section.
(viii) Abner of Burgos (c. 1270–c. 1345) (known as Alfonso of
Valladolid after his baptism [c. 1320]), refers to Avicenna a number
of times. Specifically, his Teshuvot la-meharef (Replies to the
˙
Blasphemer), which is extant in both the original Hebrew and a
fourteenth-century Castilian translation, as well as in three works
extant only in medieval Castilian versions – namely, Offering of
Zeal (Ofrenda de Zelos), Book of the Law (Libro de la Ley), and
Teacher of Righteousness (Mostrador de Justicia) – include references
to Avicenna’s ideas. Especially intriguing are numerous citations
attributed to Avicenna’s so-called ‘Oriental Philosophy.’181 It turns
out, however, that they are nearly identical to passages in the
Hebrew translation of Ibn Tufayl’s Hayy Ibn Yaqzān, which would
˙
thus seem to be Abner/Alfonso’s ˙ 182
source. ˙
181
See Mauro Zonta, “Possible Hebrew quotations of the metaphysical section of Avicenna’s
Oriental Philosophy and their historical meaning,” in Hasse and Bertolacci (eds.), The
Arabic, Hebrew and Latin Reception of Avicenna’s “Metaphysics,” pp. 177–9.
182
This discovery is described in Ryan Szpiech, “In search of Ibn Sīnā’s ‘oriental philosophy’ in
medieval Castile,” Arabic Sciences and Philosophy, 20 (2010): 185–206. Szpiech writes
(p. 194): “Although Abner specifically cites the source of his statements as Ibn Sīnā’s
‘Oriental Philosophy,’ these citations are all actually direct quotes from Ibn Tufayl’s
Hayy Ibn Yaqzān.” See, as well, the following encyclopedia articles by Szpiech,˙ which
˙
offer a similar˙ summary of Abner’s references to Avicenna and to other Islamic philoso-
phers: “Alfonso of Valladolid/Abner of Burgos”; “Teshuvot la-Meharef”; “Sermonas contra
los moros y judíos”; “Moreh Ẓedek”; “Libro de las tres creencias”;˙ all in David Thomas
and Alex Mallett (eds.), Christian-Muslim Relations: A Bibliographical History, IV:
1200–1350 C.E. (Leiden, 2012), pp. 941–62. It has been argued that Abner/Alfonso’s deter-
minist philosophy draws on Avicenna; but following Szpiech’s study of the Avicennian quo-
tations in Abner/Alfonso’s works this seems to be less likely. See Zonta, “The role of
Avicenna and of Islamic ‘Avicennism’,” p. 652, referring to information by Charles
H. Manekin, according to whom “there was a strong influence [on Abner/Alfonso] of
Avicennism on such points as the doctrine of God’s knowledge and providence, which
passed from Mūsa Ibn al-Lāwī through Abner of Burgos, arriving to the Spanish Jewish
Abner/Alfonso’s Hebrew philosophico-mathematical work Meyasher

ʿaqov (Straightening the Crooked; after Is. 40:4) contains a number of
explicit references to Avicenna,183 but it seems dubious that they
derive directly from Avicenna.
(ix) Hayyim Israeli ( fl. 1320), a member of a well-known Toledan
family ˙that produced a number of distinguished scientists, wrote at
least three scientific treatises, only one of which is extant.184 In
this treatise, which discusses the physical location of the Garden of
Eden, Hayyim Israeli explicitly refers to and draws on Avicenna’s
˙ He says that the philosophers identify the deity with
Shifāʾ.185
the ninth sphere and explains the proof from motion of God’s exist-
ence. Some Muslim philosophers, however, “first and foremost
among them the wonderful sage, Abū ʿAlī Ibn Sīnā,” hold that “the
Creator, blessed be He, is the cause of the First Mover, not the
First Mover itself.”186 Referring to the meteorological section of
al-Shifāʾ, he paraphrases Avicenna’s account that the region of the
equator is the most balanced clime,187 so that the temperaments of
its inhabitants are best balanced, and adds that Avicenna reiterated
this view in “his medical work.” He observes that Avicenna mentions
a composition proving this point that “has not reached us.”188
(Consequently, Hayyim Israeli saw himself obliged to prove the
point himself.) He˙ notes that Averroes held the opposite view and
comments:
Abū-l-Wālid Ibn Rushd is a great scientist in the eyes of the wise men of our
generation, especially among the Christian philosophers. In his epitomes of
Aristotle’s books he wrote that he wished to take exception to Avicenna’s
view, [affirming] that the zone under the [celestial] equator is not balanced.
I wish to probe into the claims of this scientist [Averroes] – who is wise in his
own eyes – that are directed against Avicenna. [. . .] I wish to show to truly
philosopher Hasdai Crescas (d. 1412): a sort of ‘Spanish Jewish Avicennism’” (ibid. p. 653).
˙
See also: Charles H. Manekin and Oliver Leaman (eds.), The Jewish Philosophy Reader
(London, 2000), p. 246; Charles H. Manekin (ed.), Medieval Jewish Philosophical
Writings (Cambridge, 2007), p. xxi. We are indebted to Ryan Szpiech for his advice on
this section.
183
Alfonso, Meyashsher ʿAqov, ed. and trans. Gita M. Gluskina (into Russian) (Moscow, 1983),
see index p. 130, s.v.
184
Y. Tzvi Langermann, “‘The making of the Firmament’: R. Hayyim Israeli, R. Isaac Israeli
and Maimonides,” in Shlomo Pines Jubilee Volume on the ˙ Occasion of His Eightieth
Birthday, Part I, Jerusalem Studies in Jewish Thought IX (Jerusalem, 1988), pp. 461–76
(Heb.). He is also the author of one of the Hebrew versions of Avicenna’s Urjuza; see
below, B.2 [ii].
185
“Trattato del Paradiso di Hajjim Israel,” ed. Pietro Perreau in Jubelschrift zum neunzigsten
Geburtstag des Dr. L. Zunz (Berlin, 1884), text: pp. 20–42 (Hebrew section), introduction:
pp. 141–2 (non-Hebrew section).
186
“Trattato,” ed. Perreau, p. 21.32–34.
187
See above, near n. 171; this statement is also reported by Samuel Motot (supra, n. 155).
188
“Trattato,” ed. Perreau, p. 25.17–27. ˙ ˙
wise men that the matter is not as Averroes holds. It is the same thing wher-
ever Averroes diverges from Avicenna in other philosophical matters, where
his claims are false [too].189
Hayyim Israeli emphasizes that only scholars trained in both natu-
ral˙ philosophy and astronomy are able to investigate this issue
appropriately and that Averroes lacked Avicenna’s astronomical com-
petence and therefore failed to understand his position.
Hayyim Israeli also alludes to Avicenna’s theory of the formation
of ˙dry land, which, as we saw, was enthusiastically embraced by
Samuel Ibn Tibbon. He does not mention Ibn Tibbon’s treatise on pre-
cisely this topic,190 but believes that Abraham Ibn Ezra supported
Avicenna’s theory in his biblical commentary.191 Hayyim Israeli
audaciously affirms that Avicenna’s naturalistic account ˙ of the for-
mation of dry land is identical with the belief in creation (hiddush)
as “affirmed by our perfect Torah”: regrettably he does not˙ explain
this thesis, to which Samuel Ibn Tibbon had devoted an entire
volume. “It is extraordinary,” Hayyim Israeli comments, “that a
man [Avicenna] would arrive through ˙ his own intellect at the foun-
dations of the Torah as transmitted by Tradition.”192 “I embraced
the opinion of the sage Avicenna,” he adds, “because I found that
his belief[s] on any philosophical subject, which he reached by the
power of his intellect, are close to the belief of the tradition of our per-
fect Torah.”193
Hayyim Israeli was neither prolific nor influential: two of his three
˙
treatises are lost, and the third is extant in only three manuscripts.
He is interesting as an instance of a dedicated Avicennian who does
not hesitate to go against what he knows to be the Averroist consensus
of the day. (Noteworthy, too, is his awareness of the climate of opinion
among Christian philosophers.) What brought him to this Avicennian
commitment is unclear, but it may be mentioned that Hayyim Israeli
translated Avicenna’s Urjūza fī al-tibb into Hebrew (B.2˙ [ii]). Together
with Samuel Ibn Tibbon and Moses ˙ ha-Levi, he is one of the few
Jewish scholars who strongly and consistently side with Avicenna
against Averroes.
(x) Judah Kohen, active in Provence around 1350, was apparently
inspired by Avicenna’s philosophic-mystical Risāla fī al-ʿIshq
189
Ibid., pp. 26.36–27.1; similarly p. 30.6–7. Averroes makes this statement in Epitome on
Meteorology; see Risāla al-Āthār al-ʿulwiyya li-Ibn Rushd, ed. Rafiq al-ʿAjam and Gerard
Gehami (Beirut, 1994), p. 58.6–8.
190
Ibid., p. 30:9–14. Hayyim Israeli mentions Ibn Tibbon’s commentary on Ecclesiastes (ibid.,
˙
p. 39.11–12), but does not seem to have seen Maʾamar Yiqqawu ha-mayim.
191
“Trattato,” ed. Perreau, p. 30.9, 18. See above, nn. 148, 156.
192
Ibid., p. 30.14–15.
193
Ibid., p. 30.7–8.
(Epistle on Love), which he seems to say he had in his hands while he

was in “Melgueir” in Provence.194
(xi) Moses Narboni (Provence, c. 1300–1360), the well-known com-
mentator on Maimonides’ Guide and on Averroes, also commented
on al-Ghazālī’s Intentions of the Philosophers. The commentary is
Averroist in its tendency, but often refers to al-Shifāʾ and al-Najāt,
as pointed out by Steinschneider.195 The same holds of Narboni’s com-
mentary on Maimonides’ Guide.
(xii) Simeon ben Semah Duran (Majorca, 1361 – Algiers, 1444) was
a communal leader˙ and˙ prolific writer. In his philosophical work
Magen ʾavot (Shield of the Forefathers) he often draws on Avicenna,
whom he must have read in Arabic. Avicenna is particularly present
in the chapter in which Duran criticizes the Qurʾan and Islamic theol-
ogy. The sources of Duran’s voluminous work have not yet been
studied; we have to content ourselves with pointing out what Duran
describes as a literal quotation from Avicenna:
In his Hayy ben Yaqzān, Avicenna has left this question [of the world being
˙
newly created ˙
or eternal] open and controversial, but his view is that of the
philosophers [who uphold] the eternity of the world. When questioned
about what is said in the Book of their religion [= the Qurʾan], viz. that the
world has come to be after not having been and that it will be corrupted
after having come to be, he gave the following answer: “There is a demonstra-
tive proof, which cannot be dismissed, that the spheres are eternal and are
not susceptible to corruption. Concerning what is said in the Book, namely
that they are newly [created] and have come to be after not having been,
this means that they are not the cause of their own existence; rather, the
cause of their existence preceded them, viz. their separate intellects which
precede them temporally. The Book’s reference to temporal priority is a
hint for those who are learned and who understand.
The Book refers to the Throne of Glory as having existed prior to all existents.
This means: it is their material cause and more honorable than they are. And
what the Book says concerning the heaven’s destructibility, this concerns . . .
[the last five words are unintelligible].” So far his words.196
Magen ʾavot repeatedly refers to Avicenna and many other thinkers;
the study of its sources may well be rewarding.
(xiii) Aaron ben Elijah the Younger (1328 or 1329–1369), active
mainly in Constantinople, is the only Karaite author included in
194
See Geoges Vajda, “La question disputée de l’essence et de l’existence vue par Judah
Cohen, philosophe juif de Provence,” Archives d’histoire doctrinale et littéraire du Moyen
Age, 43 (1977): 127–47, on p. 136 and note 58.
195
See Steinschneider, HÜ, p. 318. Some of Narboni’s statements are problematic, however;
see above, n. 15.
196
Qeshet u-magen, ed. Steinschneider, p. 21.25–22.8; = Magen ʾavot, p. 238. Translation in
Steinschneider, “Islam und Judentum,” pp. 29–30. The origin of this quotation remains
to be determined.
this survey. The absence of Karaites is understandable, inasmuch as
they remained faithful to the atomism of their Muʿtazilite forebears
until very late and thus were not interested in philosophical systems
in Aristotelian vein.197 In his great theological-philosophical work, ʿEs
Hayyim (Tree of Life, 1346),198 intended as a Karaite counterpart to˙
˙
Maimonides’ Guide, Aaron ben Elijah refers to Avicenna, whom he
read in Arabic, in his discussion of creation. He reports the contro-
versy between Avicenna and Averroes concerning whether the exist-
ence of the heavens is necessary or contingent and seems to side
with the latter.199
(xiv) Judah Nathan, we recall, translated al-Ghazālī’s Maqāsid and
realized that the work was related to Avicenna. He composed ˙ a
Hebrew work consisting of glosses that constituted a running com-
mentary on his translation. Unfortunately this works is lost (above,
B.1.2 [i]).
These thinkers read some of Avicenna’s works in Arabic and incor-
porated Avicennian doctrinal items into their works composed in
Hebrew. The list has no pretensions to be exhaustive; texts that
have been little studied may add new and valuable information.
B.1.5. Avicenna in Hebrew Translations of Latin Philosophical Works

Latin-to-Hebrew translations were far less numerous than those
from Arabic. Still, here and there (especially in Italy), Latin-to
Hebrew translations were produced. As it happens, one of them is of
immediate interest to us: the Hebrew translation of the Tractatus
de anima by Dominicus Gundissalinus (c. 1110–c. 1190), which offers
a systematic presentation of Avicenna’s psychology as found in the De
anima, the Latin version of Avicenna’s Kitāb al-Nafs.200 The anon-
ymous translator, possibly in Catalonia around the middle of the thir-
teenth century, called his work Sefer ha-Nefesh. This work did not
circulate widely (extant in only a unique late and incomplete manu-
script201); but a long excerpt (from its eighth chapter) was included
197
See Daniel J. Lasker, “Medieval Karaism and science,” in Freudenthal (ed.), Science in
Medieval Jewish Cultures, pp. 427–37.
198
Etz Chayyim, Ahron ben Elias aus Nikomedien des Karäer’s System der
Religionsphilosophie, ed. Moritz Steinschneider and Franz Delitzsch (Leipzig, 1841).
199
Ibid., pp. 26–9, 39. We are indebted to Daniel J. Lasker for his advice on this section.
200
See Dag Nikolaus Hasse, Avicenna’s ‘De anima’ in the Latin West. The Formation of a
Peripatetic Philosophy of the Soul 1160–1300 (London and Turin, 2000). (Note that Kitāb
al-Nafs is the section on the soul in Avicenna’s al-Shifaʾ and is different from the previously
mentioned Risāla fī al-Nafs.) The discussion that follows draws on Zonta, La filosofia
antica nel Medioevo ebraico, pp. 193–5 and Yossef Schwartz, “The medieval Hebrew trans-
lations of Dominicus Gundissalinus,” in Alexander Fidora, Harvey Hames, Yossef
Schwartz (eds.), Latin-into-Hebrew–Studies and Texts. Volume 2: Texts (Leiden, forthcom-
ing in 2012). We are grateful to the author for his kind permission to draw on his study
prior to its publication.
201
Cambridge, University Library, Add. 1858, ff. 183v–230v.
in the widely disseminated Shaʿar ha-Shamayim by Gershon ben

Solomon, albeit ascribing it to Aristotle.202 While the discussion
leads up to the conclusion that the soul subsists after the death of
the body as a separate substance, the aim of the inclusion of the pas-
sage from Sefer ha-Nefesh was to show how the discussion can be
grounded in medical or biological theory.203
B.1.6. Avicenna’s Philosophy in Hebrew Works by Latin-Reading

Jewish Scholars
Contrary to the situation in the Midi, a number of Jewish scholars
in Italy were in close contact with the Latin majority culture and
wrote treatises drawing on Latin works, effecting a Latin-to-Hebrew
transfer of knowledge. For example, in his Sefer Tagmulei ha-nefesh
(Book of the Rewards of the Soul, c. 1290), Hillel ben Samuel of
Verona (c. 1220–1295) incorporated long passages translated from
Latin sources. As Joseph Sermoneta, the editor of Hillel’s work, has
pointed out, this includes “several fragments of the Latin version of
Avicenna’s book on the soul, Liber sextus naturalium [i.e., the treatise
on the soul from al-Shifāʾ], to which are added expanded sections from
the Liber de Anima written by the Christian scholar and translator,
Dominicus Gundissalvi.”204 Hillel drew on a work by Avicenna avail-
able to him in Latin translation and on the very same Avicennian text
in Latin that had been translated into Hebrew about half a century
earlier (a translation with which he was unacquainted).205 His view
of soul follows Avicenna on several points,206 but with Averroist
ideas mixed in.
B.1.7. Latin-to-Hebrew Translations of Pseudo-Avicennian

Philosophical Texts
The medieval European philosophical and scientific tradition
falsely ascribed to Avicenna two works that circulated in Latin and
were translated into Hebrew. Neither has any connection with
Avicenna’s authentic thought. Nonetheless, inasmuch as some medie-
val scholars took them to be genuine, it seems appropriate to consider
them here, albeit only briefly.
202
Gershon b. Shlomo, Sefer Shaʿar ha-shamayim 2.11 (Warsaw, 1875), p. 75 f.; (= Chapter 12
of the Rödelheim, 1801 edition, fol. 78r f.).
203
Schwartz, “The medieval Hebrew translations of Dominicus Gundissalinus.”
204
Hillel ben Shmuʾel of Verona, Sefer Tagmulé ha-Nefesh (Book of the Rewards of the Soul),
ed. Joseph Sermoneta (Jerusalem, 1981), Editor’s English Introduction, p. vi; see also
Index, p. 268, s.v.
205
Yossef Schwartz, “Einleitung,” in Hillel von Verona, Über die Vollendung der Seele.
Hebräisch-Deutsch. Eingeleitet und mit Anmerkungen herausgegeben von Yossef
Schwartz, übersetzt von Yossef Schwartz in Verbindung mit Alexander Fidora (Freiburg,
2009), pp. 13–14.
206
Schwartz, “Einleitung,” pp. 14, 23, 26–8, 30–1 and index, s.v. “Avicenna.”
(i) Liber de celi et mundi (On the Heaven and the World) was
ascribed to Avicenna in the Latin tradition and printed in the 1508
edition of Avicenna’s works in Latin. Alonso showed that this text is
not by Avicenna, a conclusion that was subsequently substantiated
by Ruth Glasner.207 Alonso considered it to be a paraphrase of
Themistius’ commentary on De caelo; but according to Oliver
Gutman, its modern editor, it is a translation from Arabic of a para-
phrase (perhaps by Hunayn Ibn Ishāq) of Aristotle’s De caelo, trans-
˙
lated into Latin in Spain ˙ twelfth century.208 It was then
in the late
rendered into Hebrew by Solomon ben Moses Melguiri (c. 1225–
1306) in the third quarter of the thirteenth century.209 The Hebrew
text (unpublished) was much studied in the Middle Ages (preserved
in eighteen manuscripts). On several issues (such as the concept of
the ray), the views expressed in this text not only differ from, but out-
right contradict those of Avicenna, a circumstance that may have con-
fused some readers.210
(ii) Another treatise needs to be mentioned here, although it was
ascribed to Avicenna by only some Hebrew writers. This is Aristotle
on Sleep and Wakefulness: it was translated into Hebrew, again by
Solomon Melguiri, certainly from Latin, although the Latin Vorlage
has yet to be discovered. Melguiri himself thought the text was by
Aristotle and most of the readers, too, accepted Aristotle’s authorship.
But Gershon ben Solomon, in Shaʿar ha-shamayim, ascribed the text
to Avicenna. Hagar Kahana-Smilansky has compellingly shown that
the core of the text (whose Latin version probably goes back to an
unknown Arabic original) consists of Aristotle’s three treatises on
sleep, dreams, and dream-divination, but “updated” (by Melguiri,
possibly also by the Arabic and/or the Latin redactors) in the light
of later theories, especially Galen’s.211 This Hebrew text, too was
very popular (sixteen manuscripts).
B.2. Avicenna’s Medical Thought in Hebrew

Contrary to Avicenna’s philosophical works, which, as we have seen,
were hardly translated into Hebrew, Avicenna’s medical works,
207
Manuel Alonso, “Hunayn traducido al Latin por Ibn Dawud y Domingo Gundisalvo,”
˙
Al-Andalus, 16 (1951): 37–47; Ruth Glasner, “The Hebrew version of De celo et mundo
attributed to Ibn Sīnā,” Arabic Sciences and Philosophy, 6 (1996): 89–112.
208
Pseudo-Avicenna, Liber Celi et mundi. A Critical Edition with Introduction by Oliver
Gutman (Leiden, 2003).
209
On Solomon b. Moses, see Glasner, “The Hebrew version of De celo et mundo,” on pp. 95–6,
and Hagar Kahana-Smilansky, “Aristotle on sleep and wakefulness: a medieval Hebrew
adaptation of an unknown Latin treatise,” Aleph, 10 (2010): 67–118, on pp. 70–82.
210
For an in-depth study see Glasner, “The Hebrew version of De celo et mundo.”
211
Kahana-Smilansky, “Aristotle on sleep and wakefulness.”
especially The Canon, were the object of enthusiastic reception,

reflected in a multitude of translations and extant manuscripts and
an early printing history.
(i) The Canon, Avicenna’s major medical work, was translated into
Hebrew, partially or totally, at least seven times.212 Its translation
history begins in Italy, and this is no mere chance. As the late
Hayyim Rabin explained, most Jewish physicians in the Midi knew
˙
Arabic and could read The Canon in the original. In Italy, by contrast,
very few Jews knew Arabic, or even Latin. When in the thirteenth cen-
tury Christian medicine in Italy came to rely increasingly on
Avicenna, Jewish physicians were in desperate need of a Hebrew ver-
sion of the work in order not to fall behind their confreres.213 Two ver-
sions of The Canon originated in Italy almost simultaneously.
A first translation was produced in Rome in 1279 by Nathan
Hameʾati, who had acquired his mastery of Arabic during his tribu-
lations in many countries.214 This translation is now preserved in at
least 77 manuscripts and thus seems to be the most widely dissemi-
nated version.
A second, incomplete, version was made around 1280 by the
Catalan Jewish scholar Zerahiah ben Isaac Hen after he had migrated
˙ two books of˙the work: Steinschneider
to Rome. It covers only the first
thought that Zerahiah suspended the project when he became aware
˙
that a complete Arabic-to-Hebrew version already existed,215 but
Rabin convincingly demonstrated that Zerahiah merely revised
˙
Nathan’s translation, limiting his work to the first two books, which
are theoretical and were therefore more difficult to translate.216
Zerahiah’s version survives in ten manuscripts.217
˙
212
On the Hebrew versions, commentaries, and summaries of The Canon, see Steinschneider,
HÜ, pp. 678–695 (§§430–442). For an updated listing of the manuscripts, see Benyamin
Richler, “Manuscripts of Avicenna’s Kanon in Hebrew translation. A revised and up-to-date
list,” Koroth, 8/3–4 (1982): 145–68. See also Lola Ferre, “Avicena Hebraico: la traducción
del Canon de Medicina,” Miscelanea de estudios árabes y hebraicos, sección hebrea, 52
(2003): 163–82; ead., “Tras las huellas del Canon ebraico,” in Canon medicinae Avicena
(Barcelona, 2002), pp. 243–87; Giuliano Tamani, ed., Il “Canon medicinae” di Avicenna
nella tradizione ebraica. Le miniature del manoscritto 2197 della Biblioteca
Universitaria di Bologna (Padua, 1988). The data given below on the number of manu-
scripts of each translation are based upon the online catalogue of Institute of
Microfilmed Hebrew Manuscripts, National Library of Israel, Jerusalem.
213
Chaim Rabin, “The history of the translation of the Canon into Hebrew,” Melilah, 3–4
(1950): 132–42 (Heb.); repr. in Hayyim Rabin, Linguistic Studies. Collected Papers in
˙
Hebrew and Semitic Languages (Jerusalem, 1999), pp. 324–38 (Heb.).
214
See HÜ, pp. 679–80 and Hameʾati’s introduction to his translation, published in
Steinschneider, “Miscelle 29,” Monatsschrift für Geschichte und Wissenschaft des
Judentums, 38 (1894): 179–80.
215
216
Rabin, “The history,” p. 134.
217
Oxford, Bodleian Library, Opp. Add. fol. 12, includes what is probably the archetype of the
entire manuscript tradition, from which all the other copies derive. See Mauro Zonta,
A third version, produced by the Spaniard Joseph b. Joshua
ha-Lorqi before 1402, includes only Book 1 and the first chapter of
Book 2. It was executed either in Spain or the Midi, where the knowl-
edge of Arabic had meanwhile declined. Like Zerahiah, ha-Lorqi did
not translate from scratch, but revised the first˙ part of Nathan
Hameʾati’s translation.218 Inasmuch as this version is incomplete, it
is not surprising that in some manuscripts it is transmitted together
with Nathan Hameʾati’s translation, while in others it is transmitted
independently (this version is extant in 23 manuscripts).
There are, in addition, no fewer than four more versions (full or par-
tial) of The Canon, all anonymous, transmitted in 32 manuscripts.
All in all, there are more than 150 manuscripts with partial or com-
plete versions of The Canon, making this work into the most widely
disseminated medieval Hebrew work of science. In addition, about
thirty commentaries were written.219
The Canon in Hebrew was printed very early, in Naples in
November 1491. It is the largest Hebrew incunabulum volume and
was printed in a great number of copies, testifying to the work’s impor-
tance for its users. Scholars have debated the question of which trans-
lation served as the basis for the print edition.220 The most thorough
examination of the question, by Hayyim Rabin, concluded that the
print text was eclectic. The printers˙ took Hameʾati’s version as their
copy text, but sometimes preferred Zerahiah’s or ha-Lorqi’s, though
their criteria for doing so are not obvious.˙ The printers seem to have
occasionally corrected the Hebrew text on the basis of a comparison
with the Arabic. The print edition is thus a third revision of Nathan
Hameʾati’s translation.221 (It is possible, however, that the decisions
were not made the printers, but that they employed a manuscript
with a contaminated version of the text and with marginalia.)
(ii) The Urjūza fī al-tibb (Cantics), Avicenna’s medical poem, was
translated into Hebrew ˙ four times, always accompanied by
Averroes’ commentary:222 by Moses Ibn Tibbon in 1260; by Solomon
b. Joseph Ibn Ayyub in 1261; by Hayyim Israeli, whom we have
already encountered above (B.1.4 [ix]),˙ in the 1330s;223 and in an
“Hebraica Veritas. Temistio, Parafrasi del De coelo. Tradizione e critica del testo,”
Athenaeum, n.s. 82 (1994): 403–28, on pp. 412–13, n. 47.
218
Steinschneider, HÜ, p. 682; Rabin, “The history,” pp. 134–5; Neubauer-Renan, Écrivains,
pp. 770 (bottom) and 772.
219
Steinschneider, HÜ, pp. 686–94.
220
Ibid., pp. 683–4; Rabin, “The history,” p. 135.
221
Rabin, “The history,” pp. 137–8.
222
See Steinschneider, HÜ, pp. 699–700 and the exhaustive treatment in Maud Kozodoy,
“Medieval Hebrew medical poetry: uses and contexts,” Aleph, 11 (2011): 213–88.
223
Steinschneider (HÜ, p. 700) lost the trace of the single surviving manuscript of this trans-
lation; it is now London, British Library, MS Add. 27562, Cat. Margoliouth 1032.
anonymous version. Together, the four versions are preserved in ten

manuscripts.
(iii) Al-Adwiya al-qalbiyya (The Cardiac Drugs), a work that, as we
observed, attracted the attention of a number of authors, because
of its medical and philosophical content,224 was translated into
Hebrew twice by two anonymous translators: once from Arabic, and
once from one of the Latin versions (probably that by Arnaldus of
Villanova).225
This work was little read in Hebrew: only five manuscripts are
extant. Nonetheless, inasmuch as it contains some interesting psycho-
logical ideas, mainly about the relationship of soul to body, it attracted
some attention, notably in the fifteenth century. Thus David Ibn
Shoshan (straddling the fifteenth and sixteenth centuries) drew on
it for his interpretation of Ecclesiastes 3:21.226 At roughly the same
time it was also deemed important enough to be the object of two com-
mentaries:227 one by Baruch Ibn Yaʿish, a “Jewish Schoolman”
imbued in Latin culture, who migrated from Spain to Italy, where
he wrote his partly philosophical commentary around 1485;228 and
by an anonymous translator. These commentaries did not circulate
widely; each is extant in a unique manuscript.
(iv) The Short Canon (al-Qānūn al-saghīr) is included here inas-
much as it circulated under Avicenna’s ˙name, although it is pseudepi-
graphical.229 It was rendered into Hebrew twice: by Moses Ibn Tibbon
in 1272 (extant in a single manuscript) and later by an anonymous
translator (more than ten manuscripts are extant).230
Doubtless, Avicenna was a major figure in Hebrew medicine. Even
before the heyday of Avicennian medicine, Shem-Tov Ibn Falaquera,
˙
in a sort of roadmap for the lay amateurs of philosophy, recommends
the medical books of Hippocrates, Galen, al-Rāzī, and Isaac Israeli. He
continues: “The books of Avicenna are very meticulous (meduqdaqim)
and enlighten [even] the unsophisticated, [but] they can be [truly]
224
Text: “Edriyei Kalbiye,” ed. Kilisli Rifat Bilge in Büyük Türk Filozof ve Tib Ustadi Ibni Sina
(Istanbul, 1937), vol. 3, pp. 1–55. See also the translation in Hakeem Abdul Hameed,
Avicenna’s Tract on Cardiac Drugs and Essays on Arab Cardiotherapy (New Delhi,
1983), pp. 11–75.
225
Steinschneider, HÜ, pp. 700–1. On the three Latin translations of this work, see Simone
van Riet, “Trois traductions latines d’un texte d’Avicenne: ‘Al-adwiya al-qalbiyya’,” in
Actas IV Congreso de estudios árabes e islámicos (Leiden, 1971), pp. 339–44.
226
Y. Tzvi Langermann, “David Ibn Shoshan on spirit and soul,” European Journal of Jewish
Studies, 1 (2007): 63–86.
227
228
On Baruch Ibn Yaʿish, see Mauro Zonta, Hebrew Scholasticism in the Fifteenth Century.
A History and Source Book (Dordrecht, 2006), pp. 109–15.
229
See Manfred Ullmann, Die Medizin im Islam, Handbuch der Orientalistik,
Ergänzungsband VI/1 (Leiden, 1970), pp. 152–4.
230
understood only by the experts in this science who are proficient in the
science of logic.”231
CONCLUSION: AVICENNA AMONG THE JEWS
We begin with the reception of Avicenna’s medical writings by Jewish

physicians, a topic that can be dealt with in a few sentences. The pic-
ture is clear and unequivocal: whether they read it in Arabic or in
Hebrew, Jewish physicians used Avicenna’s Canon extensively, just
like their Muslim and Christian confreres. Maimonides is a single,
baffling, exception to this rule. Other medical writings by Avicenna
also found wide reception. Occasionally, the Canon, especially the
theoretical Treatise One, was taken into account by philosophers in
discussions of natural philosophy.
The question of the reception of Avicenna’s philosophical writings is
much more complex. In Andalusia, several arabophone Jewish
authors drew on Avicenna (some of them even inserted extracts
from his works in their writings), albeit silently.232 Maimonides, an
heir to the Andalusian tradition, continued this tradition, but stran-
gely ascribed many Avicennian doctrinal items to Aristotle. In the
East, the interest in Avicenna among Jewish scholars was nearly
nil. Avicenna was thus absent from the overt, public agenda of
Arabophone Jewish intellectuals in both East and West. Indeed, as
noticed, Jews rarely copied Arabic works by Avicenna in Hebrew
letters. This absence is prolonged in Hebrew-writing cultures,
where works by Avicenna were rarely translated (only the Najāt
was partly translated, but hardly circulated). As an author associated
with a definite corpus of writings, Avicenna hardly existed in Jewish
philosophy in Hebrew.233 In this respect, the reception of Avicenna
by Hebrew-reading Jewish philosophers stands in stark contrast
with that of the other two major Arabic philosophers – al-Fārābī
and Averroes.
As we saw, this absence of a clear textual presence by no means
implies that Avicenna did not leave his imprint on Jewish philosophi-
cal thought. On the contrary: Avicenna is definitively present, both in
231
Falaquera, Sefer ha-Mevaqqesh (Bene Beraq, 1990), p. 57. Noted in Steinschneider, HÜ,
p. 38.
232
The centrality of Avicenna among philosophers in Andalusia is underscored in Burnett,
“The coherence of the Arabic-Latin translation program,” see esp. p. 265 and infra,
pp. 280–1.
233
This state of affairs is rendered even more complex by the fact that what medieval Hebrew
readers took to be Avicenna’s authentic corpus included a Hebrew version of Liber de celi et
mundi, a well-disseminated ps.-Avicennian text some of whose doctrines contradict those of
Avicenna. Some authors also ascribed a second spurious text, Aristotle on Sleep and
Wakefulness, to Avicenna.
Arabic and in Hebrew. This was made possible by works containing

accounts of Avicenna’s thought. We will now attend to this theme
and then raise the key question: why so little Avicenna in Hebrew?
I. Avicenna’s Presence in Hebrew Philosophical Texts

We have distinguished two types of thinkers in their relationship to
Avicenna: (i) The first category are the committed Avicennians, who
saw Avicenna’s writings as a source of philosophical truth and
accepted the essentials of Avicenna’s thought, or who explicitly
sided with Avicenna against other philosophers (notably al-Ghazālī
and Averroes). These committed Avicennians explicitly ascribe to
Avicenna certain doctrinal items, which they also embrace.
Inasmuch as Avicenna was hardly available in Hebrew, committed
Avicennians had to be scholars who read his works in Arabic. (ii)
The second category are the eclectic Avicennians, that is, thinkers
who chose to appropriate one or more Avicennian doctrinal items,
that is, isolated views on some subject, which they integrated in
their works. These thinkers characteristically did not have direct
access to Avicenna’s works and did not adhere to an overall
Avicennian scheme or defend Avicenna’s positions. In most cases
they avoided explicit references to Avicenna as their source.
Occasionally some thinkers made “an ‘ideological’ use of Avicenna,”
notably by drawing on him “as a polemical tool against Averroes.”234
There were few committed Avicennians in the history of Jewish
thought. In fact, only three seem to qualify as such: Samuel Ibn
Tibbon, Moses ha-Levi, and Hayyim Israeli. All of them overtly
˙
endorsed Avicenna against Averroes. (To be sure, it remains possible
that additional committed Avicennians will be identified in the
future.) By contrast, eclectic Avicennians are legion. In the
Judeo-Arabic culture, we saw that Abraham Ibn Daʾud’s Exalted
Faith, Judah Hallevi’s Kuzari, and Maimonides’ Guide all contain sig-
nificant Avicennian elements. But none of these authors can be said to
be an Avicennian. Instead, they borrowed distinct doctrinal items
from Avicenna, such as the distinction between essence and existence,
ideas regarding the soul, theory of prophecy, the doctrine of the “giver
of forms,” accounts of divine providence and divine knowledge, etc.
In this sense, Avicennism can be said to be “present” in these
Judeo-Arabic works, but their authors are not committed to
Avicenna’s philosophy and their works themselves are not globally
Avicennian. Usually the Avicennian doctrinal items were not ident-
ified as such. While it can be assumed that those who read these
234
Zonta, “The role of Avicenna and of Islamic ‘Avicennism’,” p. 651.
works in the Judeo-Arabic originals were sufficiently immersed in
Arabic culture to recognize doctrinal items as Avicennian, this does
not usually hold for readers of the Hebrew translations.
The bulk of medieval Jewish philosophy was written in Hebrew.
We should now consider Avicenna’s presence in it. Through their
translations into Hebrew, the Judeo-Arabic works incorporating
Avicennian doctrinal items introduced them – again without being
marked as Avicennian – into the bloodstream of Hebrew philosophy.
In the thirteenth century, however, works became available that
allowed Hebrew-reading scholars to identify Avicennian doctrinal
items when they came across them. These works fall into at least
three categories. First are the commentaries on Maimonides’ Guide,
written in Hebrew by arabophone scholars who identified many
ideas as Avicennian: this is notably the case of Shem-Tov Ibn
Falaquera and Moses Narboni. Second come treatises in Hebrew ˙ by
arabophone scholars discussing Avicennian theories, with full
acknowledgement of their authorship. Naturally, these are mostly
works of the committed Avicennians discussed above (Samuel Ibn
Tibbon, Hayyim Israeli, and the Hebrew translation of Moses
ha-Levi’s ˙treatise). Their explicit references to Avicenna are not due
to ethical considerations concerning intellectual property (in the rab-
binic tradition, the principle that “whoever repeats a statement in the
name of the one who said it brings redemption to the world” [M Avot
6:6]) but to their endorsement of Avicenna. Moses Ibn Tibbon is the
rare instance of an arabophone scholar writing in Hebrew who went
to the trouble of writing against Avicenna. Third, we have Hebrew
translations of writings by Muslim authors: al-Ghazālī gave a general
account of Avicennian ideas without referring to Avicenna, but
Averroes explicitly referred to Avicenna and discussed some of his
ideas, notably those he disagreed with, at great length. Thanks to
these texts, Hebrew-reading scholars of subsequent generations
could recognize at least some Avicennian doctrinal items when they
saw them. Taken together, all these writings conveyed a considerable
amount of information about Avicenna to readers of Hebrew philoso-
phical literature, although how deeply they understood it needs to be
investigated. (The apocryphal writings ascribed to Avicenna may have
blurred the picture somewhat.) This is how Avicenna became a house-
hold name in Hebrew-language Jewish philosophy, even though
almost none of his philosophical work was available in Hebrew, and
how Jewish philosophers writing in Hebrew often integrated
Avicennian doctrinal items into their works. This is the phenomenon
that we dub Avicennian knowledge without Avicenna.
In order to gain a more comprehensive view of “Avicenna among
Medieval Jews” beyond the mere enumeration of Avicennian texts
available to Jewish scholars, two further kinds of research now
seem to be necessary: First, the presence of Avicennian doctrinal

items in Hebrew works needs to be “mapped.” Second, we should
identify the thinkers who appropriated any combination of these doc-
trinal items. This task is far beyond what we can endeavor to do here
at present.235
II. Why so Little of Avicenna in Hebrew?

We finally come to the intriguing question, already asked above: why
were none of Avicenna’s major philosophical writings translated into
Hebrew, with the (late) exception of part of al-Najāt? In particular:
why was no significant part of al-Shifāʾ translated into Hebrew
(with the exception of isolated short passages)? This problem is heigh-
tened by two considerations: First, other Avicennian works were
translated into Hebrew (the Najāt, the Canon, and the two “visionary”
235
We will content ourselves with a few examples, drawn from the scholarly literature. The
first concerns Maimonides: as noted earlier, generations of scholars identified a number
of distinctly Avicennian doctrinal items in the Guide. Here we have a paradigmatic
example of a philosopher who silently incorporated Avicennian ideas in his thought, but
only erudite arabophone readers could identify them. Much the same can be said about
the identification of Avicennian doctrinal items in the Kuzari and in the Exalted Faith.
These works are discussed by Davidson in his Alfarabi, Avicenna, and Averroes on
Intellect; he devotes a paragraph to what he calls “reverberations” in Jewish philosophy
of the themes studied in his work (pp. 180–209; similarly pp. 298–300). After a detailed dis-
cussion of the ideas of al-Fārābī, Avicenna, and Averroes on various subjects related to
intellect, Davidson systematically examines Jewish philosophers who appropriated some
of them. His survey shows that the appropriation was selective; that is, each thinker
chose this or that doctrinal item from one of the Islamic philosophers and integrated it
into his work. As far as our subject is concerned, Davidson offers a full account of the pres-
ence of Avicennian views on a number of topics in the thought of several arabophone Jewish
philosophers. This is a model for the kind of research needed for an adequate treatment of
the presence of Avicenna (or any other philosopher) in the thought of Jewish philosophers.
A second example is given by Dov Schwartz’s Yashan be-qanqan hadash, which studies the
˙
thought of a fourteenth-century Neoplatonic circle of Jewish thinkers. Schwartz analyses
their doctrines on several issues and points out that on some topics these thinkers followed
Avicenna rather than Averroes. For example, the majority appropriated the Avicennian
idea of emanation, sometimes with modifications (e.g. pp. 63, 78–9, 98–9 [n. 77], 111 [n.
106].), and explicitly preferred Avicenna’s notion of the deity to Averroes’ (pp. 119–21,
125–38). This did not prevent them from rejecting decidedly Avicennian theses, such as
the naturalistic cyclical coming-to-be of the sublunar world advanced by Samuel Ibn
Tibbon (pp. 74, 107–10; implicitly also in the commentary on the Kuzari by R. Solomon
b. Judah of Lunel [1424], see Hesheq Shelomo by R. Shelomo ben Yehuda of Lunel, ed.
Dov Schwartz [Ramat Gan, 2007], ˙ pp. 165–6). Although there is an inner coherence in
the thought of each member of this group, the borrowings from Avicenna are eclectic;
each thinker appropriated discrete doctrinal items as he saw fit. And a last example:
Avicenna substantially influenced the classifications of the sciences in Hebrew medieval
writings, as has been shown, e.g., by Mauro Zonta, “The reception of al-Fārābī’s and Ibn
Sīnā’s classiﬁcations of the mathematical and natural sciences in the Hebrew medieval phi-
losophical literature,” Medieval Encounters, 1 (1995): 358–82. On the special case of psy-
chology see Hagar Kahana-Smilansky, “The mental faculties and the psychology of sleep
and dreams,” in Freudenthal (ed.), Science in Medieval Jewish Cultures, pp. 230–54, on
pp. 248–50.
compositions). One would expect their existence, together with the
general awareness of Avicenna and his philosophy, to have stimulated
curiosity and generated a demand for a Hebrew translation of
Avicenna’s magnum opus, of whose existence Jewish scholars were
well aware. Second, the Latin philosophical culture in Andalusia
evinced an early and intense interest in Avicenna. As early as
1150–1175, substantial portions of the Shifāʾ were translated into
Latin in Toledo (with the likely participation of Abraham Ibn
Daʾud); these are extant in a large number of manuscripts.236 What
motives stimulated the Latins’ interest but were absent among the
Jews?
It has been suggested that Avicenna was little translated into
Hebrew as a consequence of Maimonides’ evaluation of him in his
letter to Samuel Ibn Tibbon (above, p. 222). Although the remark
(in both versions) is in fact appreciative – Maimonides explicitly rec-
ommends the study of Avicenna’s works, the only reservation being
that they are somewhat inferior to al-Fārābī’s – it has been presented
as reserved or even negative. Based on this (mis-)reading, it has been
claimed that the agenda of the Arabic-to-Hebrew translation move-
ment was influenced or even “determined” by Maimonides’ letter to
Samuel Ibn Tibbon. Maimonides’ allegedly unfavorable remark was
taken to be the reason why translators shunned Avicenna’s writ-
ings.237 Inasmuch as this account offers a simple and straightforward
explanation for a major cultural phenomenon of medieval Jewish
letters, it has been accepted by some writers.
We beg to differ. Only rarely can a simple explanation come to grips
with a complicated historical reality. Consider the following. Each of
the more than 500 Hebrew translations listed in Steinschneider’s
Die Hebraeischen Übersetzungen resulted from a separate decision
by a translator, perhaps negotiated with a patron. Between 150 and
200 translators spanning more than three centuries were involved.238
236
Amos Bertolacci, “A community of translators: the Latin medieval versions of Avicenna’s
book of The Cure,” in Constant J. Mews and John N. Crossley (eds.), Communities of
Learning: Networks and the Shaping of Intellectual Identity in Europe 1100-1450
(Turnhout, 2011), pp. 37–55. We are grateful to Amos Bertolacci for having allowed us
access to his paper prior to its publication. See further his “Avicenna’s Christian reception”
in Peter Adamson (ed.), Interpreting Avicenna: Critical Essays (Cambridge, forthcoming in
2013).
237
The idea was very briefly and prudently entertained by Steinschneider (HÜ, p. 280) and
advanced as a full explanation by Harvey, “Did Maimonides’ Letter to Samuel Ibn
Tibbon determine which philosophers would be studied?”. Harvey aims “to show that the
letter was dramatically influential and indeed determined [!] those philosophers who
would be studied by Jewish thinkers in the centuries after Maimonides” (p. 42). Harvey
(perhaps circularly) considers the “surprising disinterest in translating the writings of
Avicenna, especially, Al-Šifāʾ,” as an indication of “the influence of Maimonides’ letter on
the choice of scientific texts to be translated” (p. 66).
238
It seems quite unreasonable to suppose that a single sentence in a

letter, even by the Eagle of the Synagogue, none of whose Hebrew ver-
sions is completely preserved today, could possibly have played a sig-
nificant role in (let alone “determined”) the decisions taken by so
many individuals in so many locations over such a long period of
time. Maimonides’ Epistle on [or rather: against] Astrology, a long,
strongly worded and unambiguous admonition to reject astrology as
idolatrous futility (preserved in over twenty manuscripts), did not
keep generations of Jews, laymen and scholars, from engaging in
this art. All the more so, a few ambivalent words in a poorly diffused
letter can hardly have had a notable impact on the direction of a
highly decentralized intellectual movement.
Another strong indication that this explanation does not hold water
has already been noted: Samuel Ibn Tibbon himself, the addressee of
the letter (who also translated it into Hebrew), openly turned to
al-Shifāʾ for the foundations of his cosmology and cosmogony.
Moreover, he emphatically proclaimed his indebtedness to
Avicenna. In all likelihood he understood that Maimonides rec-
ommended Avicenna; in any event, he did not consider Maimonides’
opinions as binding on him. Thus, in the same letter, Maimonides
characterized the translations of the Yahyā Ibn al-Bitrīq as “doubtful”
˙
(mesuppeqet) and “very corrupt” (meshubeshet); ˙ al-Bitrīq’s own
239 Ibn
works he qualified as “lost books” (sefarim avudim; probably˙ meaning

useless books), adding that “whoever occupies himself with them loses
his time.”240 Yet one of Samuel Ibn Tibbon’s earliest substantial
translations (1210) was of a translation by Ibn al-Bitrīq, namely his
version of Aristotle’s Meteorology. Although this was ˙ a work by
Aristotle, not by Ibn al-Bitrīq, it is clear that Maimonides’ appraisal
of Ibn al-Bitrīq, with which˙ Ibn Tibbon fully concurred,241 did not
˙
deter him from turning to the latter’s writings when he deemed
them useful. Ibn Tibbon had an independent mind and did not rely
on Maimonides for his literary agenda.242
239
Letters and Essays of Moses Maimonides, ed. Shailat, p. 532.
240
Marx, “Texts by and about Maimonides,” p. 380; the quoted sentence is found in only one of
the two versions, but there is no reason to assume it is a later interpolation; see Letters and
Essays of Moses Maimonides, ed. Shailat, p. 553.
241
“[Ibn Al-Bitrīq’s] translation is very confused [mevulbelet] as the great Rabbi, the divine
philosopher˙ [. . .] Moses [Maimonides] wrote to me in his first letter that came to me”; see
Otot ha-Shamayim, ed. and trans. Fontaine, p. 2.18–19 (text), pp. 3–4 (translation).
242
Maimonides’ comment also refers to two other Christian philosophers: Abū al-Faraj Ibn
al-Tayyib and Yahyā Ibn ʿAdī. Three Ibn al-Bitrīqs are known: chronologically first is the
˙
translator Yahyā ˙Ibn al-Bitrīq, c. 877–940; the˙ second is Saʿīd Ibn al-Bitrīq, c. 877–940,
who was Melchite˙ ˙
patriarch of Alexandria (known in his ecclesiastical ˙ capacity as
Eutychius) from 933; and the third is the latter’s brother ʿĪsā. On the first, see Douglas
M. Dunlop, “The translations of al-Bitrīq and Yahyā (Yuhannā) b. al-Bitrīq,” Journal of
˙
the Royal Asiatic Society of Great Britain ˙
and Ireland, 3/4˙ (Oct., 1959): 140–50;
˙ for a list
Nor did Maimonides’ letter keep Shem-Tov Ibn Falaquera (who was
acquainted with the letter) from studying ˙Avicenna: he, too, chose his
reading list for himself, and, moreover, probably also understood
Maimonides as recommending Avicenna.243 Moreover, Falaquera
amply and explicitly draws on Avicenna’s writings and meticulously
points out where Maimonides himself had borrowed his ideas. In
addition, contrary to the proposed explanation, the great interest in
al-Fārābī among the Jewish scholars of Provence and the translation
of his works into Hebrew had begun around 1150–1160: this phenom-
enon thus predates Maimonides’ letter to Ibn Tibbon and cannot be
attributed to it.244
In sum, there is no evidence whatsoever that Maimonides’ letter to
Samuel Ibn Tibbon played any role in determining the intellectual
agenda of medieval Jewish scholars and in shaping the corpus of med-
ieval Arabic-to-Hebrew translations; least of all did it keep scholars
from translating Avicenna’s works. In our view, this letter should be
regarded as expressing the cultural preferences of its author, one
document among others supplying evidence about the preferences of
the Andalusian cultural milieu, with its Aristotelian tendencies, in
which Maimonides was educated. It reflects preferences, but it did
not create them.
Why, then, was Avicenna so little translated into Hebrew? The
question is paradoxically underscored by the fact that Avicenna was
so present in the minds of Hebrew authors. Avicenna was, as we put
it, “a household name” among medieval Hebrew-writing Jewish scho-
lars, many of whom integrated Avicennian doctrinal items into their
works. Why, we wonder, did this presence of Avicenna not stimulate
curiosity to know more about him? Why did no arabophone scholar
of his translations, see Hans Daiber, Bibliography of Islamic Philosophy (Leiden, 1999),
p. 201. On Saʿīd and ʿĪsā, see Aziz S. Atiya (ed.), The Coptic Encyclopedia, 8 vols.
(New York, 1991), 4, pp. 1265–6 (Aziz Atiyya); “Saʿīd b. al-Bitrīq,” Encyclopédie de
˙
l’Islam, 2nd edn, vol. 8, pp. 883–5 (Françoise Micheau). The comparison of their respective
profiles shows that Maimonides must have had Yahyā Ibn al-Bitrīq in mind. This con-
˙
clusion is confirmed by the fact that Maimonides expresses ˙ opinion of al-Bitrīq –
his low
˙
clearly identified as the translator – in his Aphorisms, too (24:44); cf. the Hebrew trans-
lation in Maimonides, Ketavim refuʾiyim, ed. Sussmann Muntner (Jerusalem, 1961), vol.
3, p. 311. Maimonides’ judgment of the three philosophers mentioned in the letter reflects
his generally negative attitude to Christian philosophy; see Sarah Stroumsa, “Al-Farabi
and Maimonides on the Christian philosophical tradition: a re-evaluation,” Der Islam, 68
(1991): 263–87.
243
See n. 180.
244
Noted by Charles H. Manekin in “Logic in medieval Jewish culture,” pp. 113–35, in
Freudenthal (ed.), Science in Medieval Jewish Cultures, on pp. 117–18; for another early
translation of al-Fārābī, see Gad Freudenthal, “Ketav ha-daʿat or Sefer ha-Sekhel
we-ha-muskalot: the medieval Hebrew translations of al-Fārābī’s Risālah fī’l-ʿaql. A
study in text history and in the evolution of medieval Hebrew philosophical terminology,”
Jewish Quarterly Review, 93 (2003): 29–115.
take the initiative to translate al-Shifāʾ, a title that was well known to
Hebrew scholars, as was the name of its author?
The following facts seem to us germane for an answer.245 (1) In mid-
to late 12th-century, two philosophical programs were running in par-
allel in Andalusia. One – in tune with the tendencies in the East –
treaded notably in the footsteps of Avicenna (and al-Ghazālī): it was
followed by most scholars, notably the Judeo-Arabic philosophers,
who are in fact the earliest witnesses of the reception of Avicenna in
the Iberian Peninsula (Ibn Daʾud, Judah Hallevi, etc.). The other,
specifically Andalusian tradition consisted in a literal, philologically-
grounded reception of Aristotle and his commentators and it culmi-
nated in Averroes246; these Andalusian anti-Avicennian and
pro-Aristotelian tendencies can be viewed as a component of a wider
“attempt [. . .] of the Andalusians to construct an alternative to the
syntheses which were produced in the East.”247 The two parallel
and partly competing traditions gave rise to the two translation pro-
grams in Toledo as described by Charles Burnett: Dominicus
Gundissalinus favored Avicenna’s philosophy, whereas Gerard of
Cremona preferred “the authentic works of the Greeks and their
Arabic commentators.”248 (2) The reception of Avicenna was quite pro-
blematic even among arabophone Jewish scholars: almost no manu-
scripts in Hebrew letters of his philosophical works exist and his
works do not appear in Jewish book lists. By the same token, when
Judeo-Arabic works of religious philosophy drew on Avicenna – and
some of the more important ones did so quite substantially – they
did not mention Avicenna’s name; even quotations or paraphrases
remained unattributed. (Maimonides is a baffling special case, inas-
much as he presents Avicennian ideas that he apparently drew from
the Maqāsid al-falāsifa but ascribes them to Aristotle.) Only the few
committed˙ Avicennians (Samuel Ibn Tibbon, Hayyim Israeli, Moses
˙
ha-Levi) discuss Avicenna’s ideas with full attribution, as do arabo-
phone commentators on Maimonides (Ibn Falaquera, Narboni). (3)
Beginning in the third decade of the thirteenth century, Averroes’
commentaries on Aristotle began to be systematically translated
into Hebrew, a trend that started with Jacob Anatoli and became a
torrent with Moses Ibn Tibbon. This Averroean trend culminated in
the work of Gersonides (1288–1344), who greatly furthered it by writ-
ing supercommentaries on most of Averroes’ commentaries and by
245
The following section somewhat differs from (and hopefully improves upon) its counterpart
in the short version of this paper (see infra, p. 283); it owes much to David Wirmer’s insight-
ful suggestions.
246
Burnett, “The coherence of the Arabic-Latin translation program,” p. 265.
247
Langermann, “Another Andalusian revolt?,” p. 366. See also Pines, Studies in
Abu’l-Barakat al-Baghdadi, p. 77; see also above, n. 31.
248
Burnett, “The coherence of the Arabic-Latin translation program,” p. 269.
teaching them to his students.249 (4) Some of Avicenna’s views were
rather controversial: in addition to the tenet of the eternity of the
world (shared by all philosophers) also the particularly audacious the-
sis that the surface of the earth is repeatedly flooded and that humans
are (“spontaneously”) generated without a human parent. These are
precisely the ideas embraced by two of the committed Avicennians
(Samuel Ibn Tibbon and Hayyim Israeli), who emphatically ascribed
them to Avicenna. ˙
If we bring these threads together, the following account may be
submitted. Of the two parallel trends active in late 12th-century
Andalusia – the Avicennian and the more “purist” Aristotelian
one – the early Andalusian Jewish scholars followed the first, heavily
borrowing from Avicenna. But they did so without attribution (Judah
Hallevi, Abraham Ibn Daʾud, Maimonides) and thus did not contrib-
ute to Avicenna becoming an “authority” for Jewish intellectuals.
Now the intellectual climate in Provence, the region where the first
Arabic-into-Hebrew translations were written, naturally depended
to a considerable extent on that in Andalusia, from where the first
translators originated. The “father of the translators,” Judah Ibn
Tibbon, did not translate any work by a non-Jewish author. His son,
Samuel, the translator of the Guide, was the first to translate such
works, but only one or two and only in the beginning of his career;
although he embraced Avicenna’s cosmology, he did not translate
any of Avicenna’s works, probably because he was more interested
in his own research program and also because he discovered them
only late in life. With Samuel Ibn Tibbon’s death in 1231 or 1232, it
seems, the last chance to have a Hebrew translation of parts of the
Shifāʾ vanished. For Samuel’s son, Moses Ibn Tibbon, turned away
from his father’s Avicennian commitments to become a leader of
Jewish Averroism – he dramatically epitomizes the Averroean land-
slide among Jewish thinkers in the Midi. In fact, as of the 1230s all
Arabic-to-Hebrew cultural transmitters already followed the
Averroean agenda: this is exemplified notably by the translation
activity of Jacob Anatoli, and by Judah ben Solomon Ibn Matqah’s
Midrash ha-Hoḵ mah, a précis of Averroes’ philosophy composed in
˙
Arabic and subsequently (1247) translated by the author himself
into Hebrew (see p. 260).
Thus, whereas the early Judeo-Arabic scholars in Andalusia
silently followed the Avicennian tradition, their Hebrew-writing suc-
cessors in 13th-century Provence switched allegiance to the competing
philosophical program. This shift seems to owe much to chronology.
By the early 13th century, Avicenna already appeared to be outdated
249
Ruth Glasner, “Levi ben Gershom and the study of Ibn Rushd in the fourteenth century: a
new historical reconstruction,” Jewish Quarterly Review, 86 (1995): 51–90.
and was superseded by Averroes. The Cordovan had refuted

Avicenna’s positions systematically in many of his treatises, which
were translated into Hebrew en masse. Potential Arabic-to-Hebrew
translators and patrons had good reason to think that Averroes
made Avicenna’s philosophy obsolete and replaced it. Here it should
not be forgotten that Maimonides identified his main philosophical
authority in the Guide as “Aristotle”: Maimonides’ followers in
Provence could plausibly believe that Averroes’ writings, which
exposed Aristotle’s philosophy so minutely, gave them the key to
interpreting the Master’s writings. Thus, whereas Latin scholars
(aided by Jewish ones) had translated the Shifāʾ in the second half
of the twelfth century with the intention “to provide Western scholars
with a commentary on Aristotle’s works,”250 the Shifāʾ could no longer
fulfill this function in the post-Averroean Hebrew philosophical cul-
ture a century later. Moreover, as Averroes’ star rose, manuscripts
of Avicenna’s works must have become increasingly difficult to obtain,
especially since (as it seems) Avicenna’s works had never been copied
in Hebrew letters. In addition, contacts with Islamic culture loosened
in the fourteenth century, making access to manuscripts held by
Muslim scholars more difficult.
We thus suggest the hypothesis that the Arabic-into-Hebrew trans-
lation activity got momentum at a time when it was too late to trans-
late Avicenna. From the second third of the 13th century onward,
Jewish scholars generally believed that Averroes’ thought rep-
resented the true philosophy, namely that developed by Aristotle:
the study of philosophy became largely coextensive with the study
of the works of the Philosopher par excellence, namely Aristotle as
presented by Averroes. Accordingly, Hebrew-reading Jewish scholars
assumed that Avicenna’s thought had been replaced by the Averroean
system. This conviction, we suggest, curbed any potential motivation
of translators and patrons to produce Hebrew versions of works by
Avicenna.
Finally, one may surmise that the reception of Avicenna in Hebrew
was also hampered by his image, which to some extent was associated
with the controversial naturalistic theories mentioned above: the the-
ory of the repeated flooding of the surface of the earth (itself part of the
doctrine of the eternity of the world) and the theory that a human can
be generated “not from a human.” Avicenna came to be associated
with both theses, which were acceptable only to radical philosophers.
We have an eloquent statement about this by the noted Hebrew poet
Immanuel of Rome. Writing in Italy in the third decade of the
250
Marie-Thérèse d’Alverny, “Translations and translators,” in Robert L. Benson and Gilles
Constable (eds.), Renaissance and Renewal in the Twelfth Century (Oxford, 1982), pp.
421–62, on p. 451; see also Bertolacci, “A community of translators.”
fourteenth century, he suggested that Avicenna’s philosophical pos-
itions had earned him eternal damnation. Following in Dante’s foot-
steps, Immanuel describes his voyage through Hades and reports
having sighted there, among others, Aristotle, Galen, al-Fārābī,
Plato, and Hippocrates; Avicenna is the last on this list of illustrious
freethinkers:251
‫ ָהיָה ְלַלַעג וּ ְשׂחוֹק‬,‫ָשׁם ִא ְבּן ִסינָא‬
‫יַַען ֲא ֶשׁר ָאַמר ִכּי ִהוֵָּלד ָאָדם לא ֵמָאָדם ֶאְפ ָשׁר ִלזְַמן ָרחוֹק‬
,‫וְָאַמר ִכּי ֵלַדת ָהָהִרים ָהיָה ֶדֶרֹךּ ִטְבִעי‬
‫ִמי יִ ֵתּן וְנֱֶאָלם‬
.‫נְִמ ָשׁךּ ַאַחר ֱאמוּנַת ַקְדמוּת ָהעוָֹלם‬
What translator would volunteer to be the mouthpiece of a thinker
condemned to hell?
Acknowledgements
A shorter version of this study is appearing under the title “The Reception of Avicenna
in Jewish Cultures, East and West,” in Peter Adamson (ed.), Interpreting Avicenna:
Critical Essays (Cambridge, forthcoming in 2012). We are grateful to the respective
editors for welcoming this parallel appearance of a short and a long version of our
study, addressed to different audiences. For their kind reading of an early version
of this paper and their helpful suggestions, we are very grateful to Peter Adamson,
Ofer Elior, Ruth Glasner, Jules Janssens, Hagar Kahana-Smilansky, Y. Tzvi
Langermann, Lukas Mühlethaler, Dov Schwartz, Yossef Schwartz, Alexander
Treiger and David Wirmer. To Lenn Schramm we are yet again indebted for his
close reading of the text and for his helpful suggestions.
251
Mahhbarot Immanuel ha-Romi, ed. Dov Yarden (Jerusalem, 1957), 28.90–98. In all likeli-
˙ ˙ Immanuel derived his knowledge of Avicenna’s geology from Samuel Ibn Tibbon’s
hood,
summary, perhaps via Shaʿar ha-shamayim. The following translation has absolutely no
pretensions of doing justice to the poetical qualities of the original:
Ibn Sīnā is there,

an object of mockery and ridicule.
Because he said that the generation of man not from man is possible over a long time,
And that the mountains were born in a natural process.
If only he had kept his mouth shut!
For he followed the belief in the eternity of the world.
Interestingly, Dante does not criticize Avicenna: he refers to him as a physician (not a phi-
losopher) and lists him, together with Hippocrates and Galen, among the so-called spiriti
magni or “high souls” (Inferno IV, 143). This highlights that the reticence about Avicenna
was specific to Jewish intellectuals.
APPENDIX – AVICENNA IN THE CAIRO GENIZA
RENATE SMITHUIS
University of Manchester, Department of Religions and Theology, Oxford
Road, Manchester, M13 9PL, United Kingdom
Email: renate.smithuis@manchester.ac.uk
SAGIT BUTBUL
Bar-Ilan University, Department of Arabic, Ramat Gan, 52900, Israel
Email: sagitbutbul@yahoo.com
What follows is a preliminary list of Avicenna fragments identified

to date in the Taylor-Schechter collection in Cambridge and the
Genizah collection of The John Rylands University Library of
Manchester. The Rylands fragments were identified by Dr Sagit
Butbul during her five months in Manchester in 2010; the
catalogues of the fourteen-volume Cambridge University Library
Genizah Series provided the information about the Taylor-
Schechter fragments. All fragments were included, including
those in Arabic script, because their location in the Genizah indi-
cates that they were used by Jews and are thus germane to the sub-
ject of this article.
Arabic Fragments in Hebrew Script

1. al-Qānūn fī al-Tibb (The Canon of Medicine)
T-S Ar.11.19 ˙ leaves. Cf. ed. Būlāq II:57–8 and 67–8.
Two
T-S Ar.11.31 Pharmacopoeia. Excerpt with some variations. Two leaves
conjoined. Cf. ed. Būlāq III:342, 345, 405.
T-S Ar.43.92 Excerpt. Two leaves conjoined. Cf. ed. Būlāq I:14–5.
T-S Ar.43.165 Excerpt. Two leaves conjoined. Cf. ed. Būlāq I:19, 25.
T-S Ar.44.99 Cf. ed. Būlāq I:104–5.
T-S AS 166.149 Synopsis. Two leaves conjoined. Cf. ed. Būlāq II:245–7.
2. Urjūza fī al-Tibb (Poem on Medicine)
T-S AS 173.52 ˙ Verses 299–313. Ed. H. Jahier and A. Noureddine (Paris,
1956), pp. 31–32.
3. al-Kalima al-Ilāhiyya (The Divine Word)
T-S Ar.37.78 The beginning. Two conjoined pages.
4. Kitāb al-Najāt (The Book of Deliverance)
T-S Ar.43.189 2 × 2 conjoined pages. Ed. Cairo (1938), pp. 180–5.
5. Kitāb al-Ishārāt wa-al-Tanbīhāt (The Book of Directives and
Remarks)
T-S Ar.43.190 Ed. J. Forget, Le Livre des théorèmes et des avertissements
(Leiden, 1892), pp. 204–6.
T-S Ar.43.309 Precise section unidentified
Hebrew Fragments
T-S K14.13 ˙
Fann 4, Chapter 4 (end) and Chapter 5 (beginning). Cf. ed.
Būlāq I:196–7.
T-S NS 90.75 Chapters 1–27. 2 × 2 leaves conjoined. See also T-S NS
222.63. Cf. ed. Būlāq I:43–51.
T-S NS 172.87 Cf. ed. Būlāq I:112.
T-S NS 181.54 Part of Chapters 9, 10 and 30. Two leaves conjoined. Cf.
ed. Būlāq I:42–3, 53.
T-S NS 222.63 Chapter 7, part of Chapter 8, part of Chapters 28 and 29.
Two leaves conjoined. See also T-S NS 90.75. Cf. ed.
Būlāq I:41, 51–2.
T-S NS 329.994 Two leaves conjoined. Cf. ed. Būlāq II:404.
T-S AS 142.137 Excerpt. Cf. ed. Būlāq II:4.
2. Passages from unidentified commentaries on the Canon of
Medicine
T-S K21.42 6 × 2 conjoined pages. Cf. ed. Būlāq I:3–7.
T-S AS 202.306 Commentary? See also T-S AS 202.360.
T-S AS 202.360 Commentary? See also T-S 202.306. Cf. ed. Būlāq I:13.
Fragments in Arabic Script

In the Rylands: ˙
Ar. 188
Two conjoined pages:
Ar. 188–1/Left Book III, Section ( fann) 16, end of Maqāla 4.
page: Ed. Baghdad252 (1970), vol. 2, pp. 472.25–473.9.
Ar. 188–2/ Book III, Section 16, Maqāla 4–5.
Right page: Ed. Baghdad (1970), vol. 2, p. 473.9–25.
Ar. 188–2/Left Book III, Section 17, Maqāla 1.
page: Ed. Baghdad (1970), vol. 2, pp. 479.29–480.14.
Ar. 188–1/ Book III, Section 17, Maqāla 1.
Right page: Ed. Baghdad (1970), vol. 2, p. 480.14–30.
Ar. 479
From same manuscript as Ar. 481, perhaps Ar. 188 from same manuscript as
well. Texts difficult to identify precisely in the edition. Could this be part of
a table of contents?
Ar. 479–1: Book III, Section 6.
Ed. Baghdad (1970), vol. 2, cf. pp. 180.21–2 (lines 1–2),
181.7 (line 3).
Ar. 479–2: Book III, Section 6, last chapter ( fasl).
˙
Ed. Baghdad (1970), vol. 2, see p. 183.33 for the title of the
chapter ( fasl) only.
˙
252
The Baghdad edition is a reprint of the Būlāq 1877 edition mentioned above.
Ar. 481
From same manuscript as Ar. 479, perhaps Ar. 188 from same manuscript as
well. Texts difficult to track down in the edition. Could this be part of a
table of contents?
Ar. 481–1: Book III, Section 5, Maqāla 2.
Ed. Baghdad (1970), vol. 2, see pp. 173.30 (line 2) and
174.32 (line 5) for the titles of the chapters ( fasls) only.
Ar. 481–2: Book III, Section 6. ˙
Ed. Baghdad (1970), vol. 2, see pp. 176.32 (line 2) and
177.33 (line 4) for the titles of the chapters ( fasls) only.
In Cambridge: ˙
T-S Ar.11.1 Excerpt with slight variations. Two leaves conjoined. Ed.
Būlāq I:58 and 61.
T-S Ar.11.3 Synopsis. Five leaves of which 2 × 2 conjoined.
Ed. Būlāq II:137–55.
T-S Ar.35.36 The beginning. Ed. Būlāq I:3.
T-S Ar.35.193 The beginning. Ed. Būlāq I:3.
T-S Ar.39.95 Excerpt. Two leaves conjoined. Cf. ed. Būlāq III:130.
T-S Ar.39.473 Excerpt. Two leaves conjoined. Cf. ed. Būlāq III:333.
T-S Ar.40.74 Excerpt. Two leaves conjoined. Cf. ed. Būlāq I:13, 17.
T-S Ar.40.114 End of Section ( fasl) 16 and beginning of 17. Ed. Būlāq
I:46. ˙
T-S Ar.40.146 Ed. Būlāq I:4
T-S Ar.40.168 Ed. Būlāq I:4
T-S Ar.41.96 Excerpt. Cf. ed. Būlāq III:302–3.
T-S Ar.41.128 Excerpt. Cf. ed. Būlāq III:2–3.
T-S Ar.42.29 Two leaves. Ed. Būlāq I:129–30.
T-S Ar.42.71 Excerpt. Cf. ed. Būlāq I:74.
T-S Ar.42.154 Ed. Būlāq I:109.
T-S Ar.42.175 Two leaves conjoined. Ed. Būlāq I:149–50, 152–3.
T-S Ar.42.182 Excerpt. Cf. ed. Būlāq I:6–11.
T-S Ar.42.199 Recipes mostly extracted from the Canon with some
variations. Cf. ed. Būlāq III:325, 317, 341, 332, 331.
T-S Ar.44.130 Two leaves conjoined. Ed. Būlāq II:37–9.
T-S NS 327.55 Ed. Būlāq I:248–9.
T-S AS 144.291 Title-page of Book II in Arabic and Judeo-Arabic.
T-S AS 178.301 Excerpt. Two leaves conjoined. Cf. ed. Būlāq II:172, 217–8.
T-S AS 182.170 Chapters 17–20. Ed. Būlāq I:204.
T-S AS 184.152 Excerpt. Cf. ed. Būlāq I:288 ff.
T-S AS 187.242 End of chapter 18–beginning of chapter 19. Ed. Būlāq
I:47–8.
2. Passages from unidentified commentaries on the Canon of
Medicine
T-S Ar.39.282 Two leaves conjoined. Cf. ed. Būlāq I:73.
T-S Ar.39.461 2 × 2 leaves conjoined. See also T-S Ar.39.465. Cf. ed.
Būlāq I:203–4, 216, 222.
T-S Ar.39.465 See also T-S Ar.39.461. Cf. ed. Būlāq I:196, 201.
T-S Ar.42.167 Cf. ed. Būlāq II:32–5.
T-S NS 327.22 Jumla 1, Fasl 1 and Fasl 9. Cf. ed. Būlāq I:123, 121.
T-S AS 183.163 Commentary˙ in the form˙ of questions and answers. Cf. ed.
Būlāq I:79–80.
Or.1080 4.10 First part, with title-page.
09574239_22-2_09574239_22-2 08/08/12 12:53 PM Page 1
ISSN 0957-4239
VOLUME 22 NUMBER 2 SEPTEMBER 2012
Arabic Sciences

and Philosophy
CONTENTS
Al-Miklātı̄, a Twelfth Century Ašʿarite Reader of Averroes 155

YAMINA ADOUHANE
Regards d’Ibn Rushd sur al-Juwaynı̄. Questions de méthode

MOKDAD ARFA MENSIA
199 Arabic Sciences
Avicenna among Medieval Jews. The Reception of Avicenna’s Philosophical,
and Philosophy
VOL 22 NO 2 SEPTEMBER 2012

GAD FREUDENTHAL AND MAURO ZONTA
A HISTORICAL JOURNAL
pp 155–287
Cambridge Journals Online

For further information about this journal
22:2 SEPTEMBER 2012
please go to the journal web site at:
journals.cambridge.org/asp

ASP 2012 Vol.22 Issue.2

Uploaded by

Document Information

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

ASP 2012 Vol.22 Issue.2

Uploaded by

Copyright:

Available Formats

09574239_22-2_09574239_22-2 08/08/12 12:53 PM Page 1

VOLUME 22 NUMBER 2 SEPTEMBER 2012

Arabic Sciences and Philosophy

Al-Miklātı̄, a Twelfth Century Ašʿarite Reader of Averroes 155

Regards d’Ibn Rushd sur al-Juwaynı̄. Questions de méthode

VOL 22 NO 2 SEPTEMBER 2012

Cambridge Journals Online

YAMINA ADOUHANE Al-Miklātī, a Twelfth Century Ašʿarite

MOKDAD ARFA MENSIA Regards d’Ibn Rushd sur al-Juwaynī

GAD FREUDENTHAL and MAURO ZONTA Avicenna among

AL-MIKLĀTĪ, A TWELFTH CENTURY AŠʿARITE

Résumé. L’objectif de cet article est de présenter un nouveau témoin de la

It is a common assumption that, although Averroes has had an impor-

fellow worshippers and that, with Averroes’ death, philosophy itself

I. AL-MIKLĀTĪ: REFUTING THE PHILOSOPHERS IN THEIR

After the traditional tributes to God and His prophet, al-Miklātī

Al-Fārābī,The Book of Al-Miklātī, Averroes, Epitome on

Three remarks may be made. First, regarding the choice of the

2. Between al-Fārābī and Averroes: the meaning of

therefore to be an active reader who does not merely and passively

3. A logical refutation that makes no compromises

judgements that apply to the whole and judgements that apply to

II. AL-MIKLĀTĪ, A CLOSE READER OF THE TT

1. Al-Miklātī’s Quintessence, a Tahāfut “Tahāfut al-Tahāfut”?38

in a body is susceptible to division in one of these two kinds of div-

2. Averroes: an ally in spite of himself

Avicenna. One example is particularly interesting for it appears

3. Averroes does not replace Avicenna as the principal target of the

Avicenna have received a similar treatment, both dismissed for being

III. AL-MIKLĀTĪ AS A NEW SOURCE OF AVERROES’ WORK

1. Aristotle and Avicenna face to face: the nature and purpose

What exactly is the point of this comparison? This paragraph is

2. The origin of this criticism

based on the substance of the existent [ğawhar al-mawğūd], whereas the

C5, he is criticising Alexander’s interpretation and not Avicenna’s,

“Avicenna had never meant by the ‘necessary being’ something

source which we do not dispose of: the treatise in question. Al-Lāwī

3. The restitution of Avicenna’s doctrine: content and origin

al-Risāla al-ʿaršiyya.129 Its first section establishes a Necessary

necessary through another, possible by itself”. We find the same idea

Al-Lāwī’s remark might be an allusion to Averroes’ restitution of

About al-Miklātī’s criticism, one final remark can be made. Al-Lāwī

of these necessary existents,152 al-Miklātī refuses the necessity – in the

In this article I intended to present al-Miklātī and his work, and I

‫ﰲ ﺍﳌﻄﺒﻮﻉ‪ :‬ﺃﻻ‪.‬‬ ‫‪1‬‬

‫ﰲ ﺍﳌﻄﺒﻮﻉ‪ :‬ﺃﻥ ﻻ‪.‬‬ ‫‪2‬‬

‫ﰲ ﺍﳌﻄﺒﻮﻉ‪ :‬ﻳﻌﱰﻓﺎﻥ‪.‬‬ ‫‪3‬‬

‫ﰲ ﺍﳌﻄﺒﻮﻉ‪ :‬ﳍﺬﻩ‪.‬‬ ‫‪4‬‬

‫ﰲ ﺍﳌﻄﺒﻮﻉ‪ :‬ﺑﻐﲑ‪.‬‬ ‫‪5‬‬

‫ﰲ ﺍﳌﻄﺒﻮﻉ‪ :‬ﺑﺄﻣﺮﻩ‪.‬‬ ‫‪6‬‬

REGARDS D’IBN RUSHD SUR AL-JUWAYNĪ

MOKDAD ARFA MENSIA

falsafa et l’influence de celle-ci sur la pensée théologique d’al-Juwaynī et puis et sur-

L’histoire des rapports complexes qui existent entre la falsafa et le

œuvre qui marque d’une manière spécifique ce renouvellement. Harry

fidèle à l’ontologie générale que recouvrait la représentation atomistique occasionaliste du

I. MÉTHODES POUR DÉMONTRER L’EXISTENCE DE DIEU

Ibn Rushd résume la méthode rationnelle suivie par les Ashʿarites

possible éternel, Aristote le nie.23 Al-Juwaynī cherche à fonder la

fameux arguments: celui de la providence (dalīl al-ʿināya) et celui de

2. MÉTHODES POUR DÉMONTRER L’UNICITÉ DE DIEU

Sur cette question aussi le témoignage d’al-Juwaynī est invoqué. Le

3. MÉTHODES POUR LE PROBLÈME DES ACTES HUMAINS

Ibn Rushd traite de ce problème dans le chapitre: Fī al-qadāʾ wa-