Kitdb al-fJiss wa-l-maJ;sus

Aristotle's Parva naturalia in Arabic Guise

ROTRAUD HANSBERGER
King's College, Cambridge
By the end of the lOch century, most of rhe texts that constitute the
Aristotelian corpus had been translated into Arabic, 1 and on the face of it
the Parva naturalia are no exception to that: at least their first six treatises'
were transmitted in Arabic, known by the collective title of Kiliib al-Ifiss wa-I-
mabslis, 'Book of Sense-Perception and the Perceived'. However, this 'trans-
mission' was of a peculiar kind: Kztiib aI-Hiss wa-I-mabslis does not constiture
what we would call a 'translation' of the Parva naturalia. Instead it is an adap-
tation that is, in fact, characterised far more by Neoplatonic and Galenic
than by Aristotelian ideas.
Given that the text ofKztiibal-Hisswa-I-mabsiis itself has only recently been
rediscovered, this circumstance has long put constraints on any enquiry
into the transmission of the Parva naturalia, since not much was known
about the text Arabic authors were referring to by the title Kiliib ai-HIss wa-I-
mahsus - that it was not identical with the version of the Parva naturalia
we know, however, was obvious. Thanks to Hans Daiber's discovery, now
some twenry years ago, of an acephalous copy of Kiliib ai-Hiss wa-I-mahslis in
a 17
ch
century manuscript in the Raza Library in Rampur (India),' we are
now in a posirion to examine the text itself. While this presents a big srep
1. For details, see R. GOULET ed., Dictionnaire des philosophes antiques, Paris, 1989, S.Y.
"Aristote de Sragire"; for an (incomplete) ovenriew, see A. BADA'WI, La transmission de la
philosophie grecque flU monde drabe, Paris, Vrin, J 9G8.
2. I.e. from De sensu et sensihilibus to De longitudine et brevitate vitae.
3. Ms. Rampur 1752, fa!' 7a-S4b; see I.'A. 'ARSJ-Il, Catalogue ofthe Arabic Manuscripts
in Raza Lihrar}, Rampur, Rampur University Press, 6 vols., Rampur, 1963-1977, vol. 4,
p. 534 f., H. DAIBER, "Salient Trends of the Arabic Aristotle", in G. EN DRESS, R. KRUK
cds., The Ancient Tradition in Christian and Islamic Hellenism: Studies on the Transmission of
Greek Philosophy and Sciences, dedicated to H.]. Drossaart Lulofs on his ninetieth birthday,
Leyde, Brill, 1997, p. 29-41, p. 36 ff.; R. HANSBERGER, The Transmission o/Aristotle's Parva
naturalia in Arabic, D.Phil. Thesis, Oxford, 2006, p. 4 if. The thesis is scheduled to be
published, in modified form, in the series Aristoteles Semitico-Latinus (ed. by H. DAlBER,
Brill, Leyde). 1he Rampur manuscript also contains an anonymous commentary that is
inserted into the text of Kitdbal-fjiss
Kitdb al-fJiss wa-l-maJ;sus
Aristotle's Parva naturalia in Arabic Guise
ROTRAUD HANSBERGER
King's College, Cambridge
By the end of the lOch century, most of rhe texts that constitute the
Aristotelian corpus had been translated into Arabic, 1 and on the face of it
the Parva naturalia are no exception to that: at least their first six treatises'
were transmitted in Arabic, known by the collective title of Kiliib al-Ifiss wa-I-
mabslis, 'Book of Sense-Perception and the Perceived'. However, this 'trans-
mission' was of a peculiar kind: Kztiib aI-Hiss wa-I-mabslis does not constiture
what we would call a 'translation' of the Parva naturalia. Instead it is an adap-
tation that is, in fact, characterised far more by Neoplatonic and Galenic
than by Aristotelian ideas.
Given that the text ofKztiibal-Hisswa-I-mabsiis itself has only recently been
rediscovered, this circumstance has long put constraints on any enquiry
into the transmission of the Parva naturalia, since not much was known
about the text Arabic authors were referring to by the title Kiliib ai-HIss wa-I-
mahsus - that it was not identical with the version of the Parva naturalia
we know, however, was obvious. Thanks to Hans Daiber's discovery, now
some twenry years ago, of an acephalous copy of Kiliib ai-Hiss wa-I-mahslis in
a 17
ch
century manuscript in the Raza Library in Rampur (India),' we are
now in a posirion to examine the text itself. While this presents a big srep
1. For details, see R. GOULET ed., Dictionnaire des philosophes antiques, Paris, 1989, S.Y.
"Aristote de Sragire"; for an (incomplete) ovenriew, see A. BADA'WI, La transmission de la
philosophie grecque flU monde drabe, Paris, Vrin, J 9G8.
2. I.e. from De sensu et sensihilibus to De longitudine et brevitate vitae.
3. Ms. Rampur 1752, fa!' 7a-S4b; see I.'A. 'ARSJ-Il, Catalogue ofthe Arabic Manuscripts
in Raza Lihrar}, Rampur, Rampur University Press, 6 vols., Rampur, 1963-1977, vol. 4,
p. 534 f., H. DAIBER, "Salient Trends of the Arabic Aristotle", in G. EN DRESS, R. KRUK
cds., The Ancient Tradition in Christian and Islamic Hellenism: Studies on the Transmission of
Greek Philosophy and Sciences, dedicated to H.]. Drossaart Lulofs on his ninetieth birthday,
Leyde, Brill, 1997, p. 29-41, p. 36 ff.; R. HANSBERGER, The Transmission o/Aristotle's Parva
naturalia in Arabic, D.Phil. Thesis, Oxford, 2006, p. 4 if. The thesis is scheduled to be
published, in modified form, in the series Aristoteles Semitico-Latinus (ed. by H. DAlBER,
Brill, Leyde). 1he Rampur manuscript also contains an anonymous commentary that is
inserted into the text of Kitdbal-fjiss
FOn,,\\ III IIANIBFHCEH
forward in our understanding of the transmission ofAristotle's Parva natu-
ralia in Arabic, the text raises many questions of its own, and does not allow
us to reach many definitive answers regarding its origins. A lack of reliable
external data, for instance on date and authorship of
contributes to the problem.
In this article, I will present, in summarised fashion, some of the results
of my doctoral thesis, the first edition and study of Kitab al-/jiss
4
After a general description of Kitiib ai-Hiss wa-I-mahsus and its relation to the
Greek Parva naturalia (1) I will sketch the main theories that dominate the
adaptation (2), and outline the tentative conclusions regarding the origins of
Kitiib al-/jiss wa-I-ma!lsus (3); I will then illustrate the character of the adapted
text, and the modus operandi of the adaptor, with textual examples (4), before
concluding with a brief survey of the reception of Kitab al-/jiss in
the mediaeval Arab world (5).
1. Kitiib al-/jiss is divided into three sections (maqalas). The first
maqala contains the equivalent to De sensu et sensibilibus; its beginning is
missing in the Rampur manuscript. The second maqala is again subdivided;
the first part represents De memoria et reminiscentia, the second one, enti-
tled Bab al-Nawm wa-I-yaqa?a (Chapter on Sleep and Waking), comprises the
equivalents of De Jomno et vigilia, De insomniis, and De divinatione per
somnum. The third maqala contains the Arabic counterpart of De longitu-
dine et brevitate vitae. That said, the Arabic text does by no means consti-
tute a faithful rendering of the Greek Parva naturalia. First of all, only parts
of the Aristotelian treatises are represented in the Arabic text; conversely,
those passages in rhe Arabic text that do reflect text from the Parva natu-
ralia do not themselves make up even half of the text of Kitab ai-Hiss wa-I-
mahsus. In the surviving parts of the first maqala we find nothing that would
amount to a translation or paraphrase of any passage in De Jensu;5 in the
case of De memoria and De somno, roughly 30% of the Greek text can be
said to be represented (however vaguely) in the Arabic, accounting for about
35 and 30% of the equivalent parts of the second maqala respectively. As for
De insomniis and De divinatione, only traces of the Greek text are left in the
Arabic version. The third and final maqala of Kitab al-/jiss reflects
about 40% of the Greek De longitudine, the Aristotelian passages making
up about the same percentage of the Arabic text.
Apart from the issue of quantity, even where the Arabic represents
Aristotelian text it rarely does this in a faithful manner. This may partly be
4. R. ibid
5. cr. below p. 145, n. G, p. 150.
[(ITAIl AI,-l! ISS WA -L-MAI/SliS
due to shortcomings of the translation itself, but for the most part it is J
result of the adaptation, which fuses its additional, non-Aristotelian mate-
rial together with passages taken from the Parva naturalia. In those cases
where more substantial amounts of the Aristotelian text are identifiable
in the Arabic, i.e. De memoria, De somno, and De longitudine, a recurring
pattern can be observed: the beginning of the Greek treatise is translated
comparatively exactly and completely; further on, however, the translation
becomes more patchy and less precise. More and more of the Aristotelian
text is omitted; the last chapters of the Greek treatises are more or less left
out altogether. GMeanwhile the passages ofadded material grow longer and
longer; in the last third or so of the Arabic counterparts of De memoria and
De somno, added material makes up the bulk of the text, with only occa-
sional appearances ofshort phrases or keywords relating to the Greek Parva
naturalia.
Consequently, not much of the original philosophical content of the
Parva naturalia can be detected in Kitiib al-/jiss The Arabic Parva
naturalia are dominated by ideas expressed in the added material. At the
same time, the text does not reveal the author or the source of this addi-
tional material, maintaining Aristotle, and Aristotle alone, to be its author.
As a result, the theories proposed by the adaptor (or adaptors) were intro-
duced into later Arabic philosophy as those of the Stagirite.
2. The most prominent, and at the same time most influential, doctrine of
the adaptation concerns divination through dreams - the issue in which
the Arabic Parva naturalia most strikingly deviate from Aristotle's text.! Its
most important psychological component, the 'three-faculty theory' (as I
call it), is based on the Galenic concept of the mental faculties of imagina-
tion, cogitation, and memory which are located in the brain ventricles and
are operated by the 'animal spiri t'. 8 In Kitiib al-/jiss these facul ties
G. This is the case to a lesser degree in the third maqala, where all chapters of De longitudine
are represented in some way or other. - The recurrence of this pattern in all three cases
supports the assumption that the lost beginning of the first maqala did in fact contain a
translation or paraphrase of (parts of) De sensu. This is further supported by the evidence
of Ibn Rushd's Epitome (Talkhis) ofKitrib aI-Hiss wa-I-mahsiis.
7. For a more detailed account, see R. HANSBERGER, "How Aristotle Came to Believe in
God-given Dreams: The Arabic Version of De divinatione per somnum", in L. MARLOW
cd., DreamingAcross Boundaries: lhe Interpretation a/Dreams in Islamic Lands, Cambridge,
Mass., Harvard University Press, 2008, p. 50-77.
8. The adaptor may have drawn upon translations ofworks by GaIen himself, or texts from
the later medical tradition. Possible sources are mentioned in G. STROHMAIER, "Avicennas
Lehre van den 'inneren Sinnen' und ihre Voraussetzungen bei Galen", in ID., Van Demokrit
bis Dante Die Bewahrung antiken Erbes in der arabischen Kultur, Hildesheim-New York,
Ohm, 19%, p. 330-341: 331: 337; cf also H. MHANA-SMILANSKY, "The Mental Faculties
IH)TRAUD l!ANSBERCER
---------
play an all-imponanr role in dreaming: in conrrasr to rhe senses, they stay
active during sleep, producing dreams by occupying themselves with prior
perceptions. However, through them the human soul may become the recip-
ient of veridical dreams as well, which are not related to prior perception
but come to the dreamer from 'outside'. Here the second component of the
adaptor's theory comes into play, a metaphysical framework that berrays
Neoplatonic leanings. The ultimate cause of veridical dreams is God, their
direct source, however, the 'universal intellect'. In one act of creation, God
has created everything that is ever going to exist in this intellect; there it
exists as 'intellectual form' before eventually coming to be in the world as
'corporeal form'. 111erefore the universal intellect can reveal to human beings
in veridical dreams what will come to pass in the future. What is revealed
in a dream is not, however, the 'intellectual form' of a thing or event, but
its 'spiritual form' which, having the same qualities as the stored images of
past perceptions, is accessible to the aforementioned three (at times called
'spiritual') faculties of the dreamer. These faculties are not only respon-
sible for the dream itself but also for assigning it the right interpretation,
and for remembering both after awakening. The veridical dream owes its
prophetic character to the fact that it as well as the prophesied event are just
two different representations of one and the same 'intellectual form'. 111is
is also why it is sometimes possible for third persons to interpret divinatory
dreams correctly: they are subject to the same revelation from the universal
intellect, though they do not receive it in the form of a dream, but as 'spir-
itual words' (a concept not specified any further).
Although in this pan of the theory we find a three-tier hierarchy of
intellectual, spiritual and corporeal forms, suggesting that the 'spiritual'
holds an intermediate position between the material and the immaterial
realm (perhaps even retaining the ultimately material aspect that 'spirit'
carried in Greek medicine), the adaptor generally favours a clear dichotomy
between 'spiritual' and 'corporeal', where spiritualiry is practically equated
with immateriality, and represents the realm of the pure, perfect incorpo-
real 'higher world', as opposed to the impure, imperfect corporeal world.
Against the background of this dichotomy, the mental faculties are ordered
hierarchically in terms of their 'spirituality', with perception belonging to
the corporeal world, and imagination, cogitation, and memory increasingly
partaking of spirituality. 9 The ambivalence with regard to the concept of
spirituality may suggest that several adaptors were at work - generally a very
and the Psychology of Sleep and Dreams", in G. FREUDE:-:THAL ed., Science in i>1edieval
Jewish Cultures, Leyde, Brill, forthcoming.
9. 111is order reflects the location of the faculties in the ventricles of the btain as the adaptor
will have found it described in medical texts. ef. R. HANSBERGER, op. cit., n. 3, p. 138 If.
K[TAn A [ ~ - / : I [ S S WA-L-MAI-{SOS
real possibility in Kitab ai-Hiss wa-I-mahsus -, but in the present case it more
likely indicates that there are passages in Kitab al-fiiss wa-I-mabsus where the
adaptor stays more closely to his source texts, and passages where he shapes
the text more actively.
While the added material introduces and discusses various other themes
as well, many of them of medical nature, the three-faculty theory and the
theory of divinatory dreams are the issues closest to the adaptor's heart, as
we can see from the prominence and the space they are given in Kitabal-Hiss
wa-I-mabsUs. The three-faculty theory plays a role in both parrs of the second
maqala and also appears in the extant pages of the first maqala. (It is absent
from the third maqala, which of course deals with a rather unrelated topic,
and which is the section of Kitab al-fiiss wa-I-mabsus to have received the least
attention from the adaptor.) Nevertheless it is the theory of divinatory
dreams that seems to be the central concern of the adaptor: due mainly
to the added material, Bab al-Nawm wa-I-yaqO-.?a is the longest section of Kitab
al-fiiss wa-I-mabsus (covering fifty-three of the surviving ninety-six manuscript
pages), and it is effectively permeated by the topic of divinatory dreams
from beginning to end.
In any case the theory of divinatory dreams is certainly the most eye-
catching and defining feature ofKitabal-fiisswa-I-mabsus. Being diametrically
opposed to everything Aristotle has to say on the topic in De divinatione
per somnum, it furthermore adds a particular edge to the question of how
the Arabic adaptation of the Parva naturalia came into being.
3. Unfortunately we have but little information on the origins ofKltiib al-fiiss
wa-I-mahsus. The situation is complicated by the fact that we have to account
for both a translation process and an adaptation process, which may (or
may not) have been linked to each other. The Rampur manuscript does not
reveal any dates or names that could help us identity the translator or the
adaptor(s); and references in Arabic bibliographical works, although they
do exist, do not provide any reliable data, either. 10 The enquiry into genesis
and provenance of Kitab al-fiiss wa-I-mabsus therefore has to rely entirely on
clues provided by the text itself.
In my doctoral dissertation I have approached the problem through a
detailed analysis of major parts of Kltiib al-fiiss wa-I-mabsus, with particular
emphasis on the second maqala. In what follows I will present the main
results of this study without, however, being able to broach the underlying
arguments in any detail.
10. See below p. 160 f.
I ROTRAUD HANSBERGER
play an all-important role in dreaming: in contrast to the senses, they stay
active during sleep, producing dreams by occupying themselves with prior
perceptions. However, through them the human soul may become the recip-
ient of veridical dreams as well, which are not related to prior perception
but come to the dreamer from 'outside'. Here the second component ofthe
adaptor's theory comes into play, a metaphysical framework that betrays
Neoplatonic leanings. The ultimate cause of veridical dreams is God, their
direct source, however, the 'universal intellect'. In one act of creation, God
has created everything that is ever going to exist in this intellect; there it
exists as 'intellectual form' before eventually coming to be in the world as
'corporeal form'. Therefore the universal intellect can reveal to human beings
in veridical dreams what will come to pass in the future. What is revealed
in a dream is not, however, the 'intellectual form' of a thing or event, but
its 'spiritual form' which, having the same qualities as the stored images of
past perceptions, is accessible to the aforementioned three (at times called
'spiritual') faculties of the dreamer. These faculties are not only respon-
sible for the dream itself but also for assigning it the right interpretation,
and for remembering both after awakening. The veridical dream owes its
prophetic character to the fact that it as well as the prophesied event are just
two different representations of one and the same 'intellectual form'. This
is also why it is sometimes possible for third persons to interpret divinatory
dreams correctly: they are subject to the same revelation from the universal
intellect, though they do not receive it in the form of a dream, but as 'spir-
itual words' (a concept not specified any further).
Although in this part of the theory we find a three-tier hierarchy of
intellectual, spiritual and corporeal forms, suggesting that the 'spiritual'
holds an intermediate position between the material and the immaterial
realm (perhaps even retaining the ultimately material aspect that 'spirit'
carried in Greek medicine), the adaptor generally favours a clear dichotomy
between 'spiritual' and 'corporeal', where spirituality is practically equated
with immateriality, and represents the realm of the pure, perfect incorpo-
real 'higher world', as opposed to the impure, imperfect corporeal world.
Against the background of this dichotomy, the mental faculties are ordered
hierarchically in terms of their 'spirituality', with perception belonging to
the corporeal world, and imagination, cogitation, and memory increasingly
partaking of spirituality. 9 The ambivalence with regard to the concept of
spirituality may suggest that several adaptors were at work - generally a very
and the Psychology of Sleep and Dreams", in G. FREUDENTHAL ed., Science in Medieval
Jewish Cultures, Leyde, Brill, forthcoming.
9. This order reflects the location of the faculties in the ventricles of the brain as the adaptot
will have found it desctibed in medical texts. Cf. R. HANSBERGER, op. cit., n. 3, p. 138 If.
KITAB AL-IjISS WA-L-MAJjSUS
real possibility in Kitab al-fjiss wa-I-mabsiis -, but in the present case it more
likely indicates that there are passages in Kitlib al-fjiss wa-I-mabsiis where the
adaptor stays more closely to his source texts, and passages where he shapes
the text more actively.
While the added material introduces and discusses various other themes
as well, many of them of medical nature, the three-faculty theory and the
theory of divinatory dreams are the issues closest to the adaptor's heart, as
we can see from the prominence and the space they are given in Kitlib al-fjiss
wa-I-mabslls. The three-faculty theory plays a role in both parts of the second
maqala and also appears in the extant pages of the first maqala. (It is absent
from the third maqala, which of course deals with a rather unrelated topic,
and which is the section of Kitlib al-fjiss wa-I-mabsiis to have received the least
attention from the adaptor.) Nevertheless it is the theory of divinatory
dreams that seems to be the central concern of the adaptor: due mainly
to the added material, Bab al-Nawm wa-I-yaqa;a is the longest section of Kitlib
al-fjiss wa-I-mabsiis (covering fifty-three ofthe surviving ninety-six manuscript
pages), and it is effectively permeated by the topic of divinatory dreams
from beginning to end.
In any case the theory of divinatory dreams is certainly the most eye-
catching and defining feature ofKitab al-fjiss wa-I-mabsiis. Being diametrically
opposed to everything Aristotle has to say on the topic in De divinatione
per somnum, it furthermore adds a particular edge to the question of how
the Arabic adaptation of the Parva naturalia came into being.
3. Unfortunately we have but little information on the origins ofKitlibal-fjiss
wa-I-mabslis. The situation is complicated by the fact that we have to account
for both a translation process and an adaptation process, which may (or
may not) have been linked to each other. The Rampur manuscript does not
reveal any dates or names that could help us identify the translator or the
adaptor(s); and references in Arabic bibliographical works, although they
do exist, do not provide any reliable data, either. ID The enquiry inro genesis
and provenance of Kitlib al-fjiss wa-I-mabsiis therefore has to rely entirely on
clues provided by the text itself
In my doctoral dissertation I have approached the problem through a
detailed analysis of major parts of Kitab al-fjiss wa-I-mabsiis, with particular
emphasis on the second maqala. In what follows I will present the main
results of this study without, however, being able to broach the underlying
arguments in any detail.
10. See below p. 160 f.
I
ROTRAUD HANSBERGER
As mentioned earlier, the Aristotelian text is not separated clearly from
the added material in Kitiib al-Ijiss wa-I-mabsiis; we are not looking at the case
of a text accompanied by a clearly demarcated commentary. Aristotelian
and added material are fused together to constitute one text; the translated
Aristotelian text in Kitiib al-Ijiss wa-Imabsiis can be identified as such solely
in comparison with the Greek text of the Parva naturalia. The often rather
fragmentary character of the Aristotelian text further reduces the number
of passages that can actually be used to assess the style and quality of the
translation.
As far as one can tell from the Aristotelian passages that survive in Kitiib
al-Ijiss wa-I-mabsus, the text of the Parva naturalia was in all likelihood trans-
lated directly from the Greek; there is no evidence for a Syriac interme-
diary. The translation seems to be comparatively early, produced perhaps in
the early 9
th
century A.D. It is generally of rather poor quality. Time and
again the translator appears to struggle with Greek syntax, and even in those
passages that come closest to the Greek text the Aristotelian doctrines are
conveyed in a simplified, if not distorted fashion. The poor quality of the
translation supports the assumption that it is early, an assumption other-
wise endorsed by terminological characteristics that place the text in vicinity
to other early Graeco-Arabica.
ll
The regular presence of certain introduc-
tory and summarising formulae in particular links the text to the style of
the so-called 'Kind! circle', 12 a group of translators working for the 9th
century philosopher al-Kind! in Baghdad. However, these formulae are at
least equally prominent in the additional material, and hence could be a
feature of the adaptation rather than the original translation of the Parva
naturalia. The same goes for the use of certain technical terms.
The distinction between features of translation and adaptation is in fact
generally problematic in Kitiib al-Ijiss wa-I-mabsus. This is directly related to
the poor representation of Aristotelian text in Kitiib al-Ijiss wa-I-mabsus (in
terms of quality as well as quantity). It means that we cannot quite be sure
whether what we identifY as 'Aristotelian' text in Kitiib al-Ijiss wa-I-mabsiis
represents the original Arabic translation of the Parva naturalia, or a subse-
quently altered version, all the more so as - not surprisingly, perhaps - the
additions of the adaptation are frequently linked to instances of'misrepre-
sentation' ofAristotle's text. Where does a flawed translation or paraphrase
11. As, e.g., the doxographical work ofPs.-Ammonius, or Ps.-Ishaq's translation of De anima.
See R. HANSBERGER, op. cit., n. 3, p. 212 If.
12. For a comprehensive description of this group, their extant works and stylistic char-
acteristics, see G. ENDRESS, "The Circle of al-Kindi", in G. ENDRESS, R. KRUK eds., The
Ancient Tradition in Christian and Islamic Hellenism: Studies on the Transmission of Greek
Philosophy andSciences, dedicated to H.]. Drossaart Lutofs on his ninetieth birthday, Leyde,
Research School CNWS, 1997, p. 43-76.
raTAB AL-lfISS WA-L-MA/fSUS
end and adaptation begin? What should count as an innocent mistake and
what as deliberate distortion? Do deviations point to the translator's limi-
tations, the poor condition of the Greek manuscript he used, or a deviant
Greek archetype? To what extent are they the responsibility of the translator,
the adaptor(s), or glossators? Again, was the Aristotelian text in Kitiib al-Ijiss
wa-I-mabsiis excerpted from a more complete Arabic translation of the Parva
naturalia? Or was the Greek text available to the translator already truncated
and altered in comparison to the extant Greek version we know?
Too little has survived of the Aristotelian text for us to determine to
what degree the Greek text used by the Arabic translator was identical with
the extant text of the Parva naturalia. This does not mean, however, that
we have to assume that Kitiib al-Ijiss wa-I-mabsiis is, as it stands, a more or less
accurate translation ofan altered Greek version of the Parva naturalia.
13
On
the contrary, there are indications that the bulk of the additions and alter-
ations that are now visible in Kitiib al-Ijiss wa-I-mabsiis were carried out in the
Arabic text, i.e. after the translation process, or perhaps also in the course
of it. The adaptation further seems to have been effected in several - two
or more - stages (not counting later glosses).
To begin with, the Arabic text does not give the impression of having
been penned by one single author. In particular, there are stylistic differ-
ences between passages devoted to the aforementioned three-faculty theory,
and passages that contain mainly Aristotelian text, even if the translation
may in some instances have undergone alterations under the influence of
the adaptation.
Furthermore, the added material is not homogeneous in itself and its
various themes and ideas are occasionally at odds with each other. On
the one hand, there is the most distinctive theme of the adaptation, the
three-faculty theory and the theory ofveridical dreams. Some of the added
passages, on the other hand, seem more closely connected to, or moti-
vated by, the translated Aristotelian text. If the adaptation was carried out
in several stages, the translator could thus have been the first (though not
the main) adaptor of the text.
The adaptation frequently takes its cues from, and builds on misrepre-
sentations of the Aristotelian text. Primafacie this may appear to speak for
deliberate distortions of the text at the hands of the adaptor. But whether
13. That the Arabic tradition of the Parva naturalia points to the existence of a second
Greek version, adapted perhaps under Stoic influence, is a hypothesis suggested by S. Pines
in the context of a comparative study ofIbn Rushd's Talkhis Kitrib al-Ijiss wa-l-mabsus and Ibn
Sina's al-Risrila al-Manrimiyya. See S. PINES, "The Arabic Recension ofParva naturalia according
to al-Risrilaal-Manrimiyya and other sources", Israel Oriental Studies 4 (1974), p. 154-163;
cf. R. HANSBERGER, op. cit., n. 3, p. 181 If., and R. HANSBERGER, op. cit., n. 7.
deliberate or not, in many cases the nature of those misrepresentations
is such as to be most easily explained within the scenario of a transla-
tion process. This not only concerns the general reduction and simplifica-
tion of Aristotelian doctrine prevalent in Kitlib a/-Ijiss wa-/-maf.zsiis, but also
features such as the rendering ofa sentence from the Parva naturalia where
every Greek word is accounted for in the translation, but which neverthe-
less conveys a meaning entirely different from that of the original; or the
development of a thought depending on the misunderstanding (and conse-
quently mistranslation) of a word that is equivocal in Greek. This does not
mean that all misrepresentations of the Aristotelian text are due to simple
translation errors, but it does show that the translator experienced difficul-
ties with the text, a circumstance which, in turn, may have made him more
prone to add to, and elaborate on his text, and which will have provided
opportunities for a later adaptor to do the same.
While the translator may have been the first adaptor of the text, the
most prominent additions, i.e. the passages containing the three-faculty
theory, seem to have been introduced at a different stage of adaptation, by
the 'main' adaptor of Kitlib a/-Ijiss He will have worked with a
translation of the Parva naturalia, possibly excerpting the text (if indeed
it was originally more substantial), and combining it with material taken
from other Graeco-Arabic sources. That he had ties to the KindI circle, a
group renowned for producing translations with 'a tendency towards inter-
pretation with a markedly Neoplatonic preference', 14 is suggested not only
by stylistic and linguistic characteristics, but also by certain theoretical and
doctrinal features, as for example the notion of 'spirituality' which we find
in Kitlib a/-Ijiss wa-/-maf.zsiis as well as in other KindI-circle texts.
15
The link to the KindI circle is further corroborated by the extant pages of
the first maqli/a, which contains what appear to be so far unknown excerpts
from the Arabic translation of Plotinus' Enneads IV-VI, a work which was
produced in the KindI circle.
1G
The fragments we find in the first maqli/a
stem, appropriately enough, from Ennead IY.6: On sense-perception and
memory. I?
14. G. ENDRESS, op. cit., n. 12, p. 59.
15. See e.g. G. ENDRESS, ProclusArabus, "BeiruterTexte und Studien" 10, Beirut-Wiesbaden,
E Steinet, 1973, p. 127 ff.
16. See G. ENDRESS, ibid., and further E W. ZIMMERMANN, "The Origins of the so-called
Theology ofAristotle", in J. KRAYE, W.E RYAN, C.B. SCHMITT eds., Pseudo-Aristotle in the
Middle Ages. The Theology andother Texts, Londres, The Warburg Institute, 1986, p. 110-240;
P. ADAMSON, The Arabic Plotinus. A philosophical Study ofthe Theology ofAristotle, Londres,
Duckworth, 2002.
17. See R. HANSBERGER, op. cit., n. 3, p. 190 ff.
Kitaba/-fjisswa-l-maf.zsiis is not, however, in all aspects a typical KindI-circle
text. It is unusual first of all in its general appearance: it is badly structured,
inhomogeneous, repetitive, not always intelligible, and at times self-contra-
dictory. This may partly be due to the incorporation of later glosses, or to
other accidents of transmission; but nevertheless the text does leave the
impression of never having been subjected to a final revision.
4. Despite this somewhat untidy appearance of the text, the adaptation
does not seem to have been produced in a rushed or superficial manner. In
the following, I will discuss three passages from the second maqli/a where
we can see the adaptor employ skill and diligence in fusing Aristotelian text
and added material together. The examples are also meant to give a more
tangible idea of the character of the text.
Clauses are numberedfor easy identification afparallelphrases in theArabic
andthe Greek version. In addition, words orphrases that represent Greek text are
printed bold in the English translation ofthe Arabic text. For editorial reasons
the Arabic text (reproduced without critical apparatus) is printed on a separate
page at the end ofthis article.
i) Ms. Rampur l752,fols. lOb, 25 -Ha, 12
(1) I also say that whatever a man sees, hears, senses, looks at or touches,
when he acts [on it] with [his] memory, he will not sayin his soul: 'I have
sensed [I la] this' or 'I have heard it' or 'I have seen it' (2) bur he will
search for it in another way which is more noble than finding it by sense-
perception. (3) For memory is not sense-perception, (4) by which things
are perceived, because sense-perception can perceive things through a body
only. (5) As for memory, it is an innate faculty which perceives things [in]
abstracted [form], (6) and it perceives them only after time [has passed].
(7) As for perceiving them at the time of their existence or in the time to come,
this is not the function of memory, nor can it be attributed to memory.
Ar., De mem., 1, 449b22-25
18
(1) a.d yap (hav tvepyfj KaTa TO IlVllIlOVeUelv, tv Tn 'I'uxn Asyel,
OTt npoTepov TOiho nKou<Jev nn<JeeTO ntvOTl<JEV.
(3) E<JTlIlV oDv l] oihe
oihe
18. Greek passages of the Parva naturalia are quoted according to Ross' edition, WD. Ross,
Aristotle. Parva naturalia, a Revised Text with Introduction and Commentary, Oxford,
Clarendon Press, 1955.
I
ROTRAUD HANSBERGER
(5?) aUu ,OUTCllV 'ClVOe; e ~ l e ; nn6.8oe;,
(6) amv yivllml xpovoe;.
(1) For whenever someone is actively engaged in remembering, he always
says in his soul in this way that he heard, or perceived, or thought this before.
(3) Therefore memory is not perception or conception, (5?) but a state or
affection connected with one of these, (6) when time has elapsed. 19
This is a typical example of a passage where we can see the Arabic text
following the Greek words, while at the same time diverging from its sense
and spirit. In several instances elements of the three-faculty theory have found
their way into the paragraph: in the description of memory as a 'faculty',20
and in the contention that memory is 'nobler' than sense-perception, that it
'perceives in abstracted form' and is not associated with corporeality in the
same way as sense-perception. These references to the adaptor's theory are
all found in the additional material ofclauses (2), (4), and (5). Nevertheless
they are tightly linked to the 'Aristotelian' sentence (1) - to be exact: to
the negation (Id) that has somehow slipped in, converting the meaning of
the translated sentence into the exact opposite of its original Greek coun-
terpart. This negation could be the result ofa misreading of the Greek orig-
inal: hout6s (which is in fact not explicitly translated in the Arabic) could
easily have been mistaken for the Greek negation oup6s, or perhaps for ou
or ouk, if the Greek manuscript was illegible or defective. Alternatively, the
negation could have been inserted deliberately by someone wanting to alter
the meaning of (1), the most likely candidate being, of course, the main
adaptor of Kitrib al-fjiss wa-l-mabsiis.
What is remarkable about this negation is that it seems extremely implau-
sible. Up to this point, the one topic that the Arabic text has focused on
more than anything else (and much more so than Aristotle's De memoria) is
memory's being solely concerned with things that have happened, and have
been perceived, in the past (c clauses 6 and 7). What would be more natural
in such a context than to say that memory is accompanied by the awareness
that one has perceived before what one is now remembering? In the Arabic
adaptation, however, the additional sentences (2), (4), and (5) let the matter
appear in a new light, alter the emphasis within (1) and in effect provide
a coherent interpretation. (Needless to say, Aristotle's point is completely
lost in the process!) The focus is shifted away from the thought that we can
19. English translation by R. SORABJI, Aristotle on Memory, second edition, Londres,
Duckworth, 2004, p. 48.
20. The term 'innate' (ghariziyya) appears, rather unusually, to be derived from Greek hexis;
this is suggested by a parallel case in another passage of the second maqiila (fol. 1Gb) where
hexis is translated as ghariza, 'natural disposition'.
KIYAB AL-ijISS WA-L-MAIfSOS
remember only what we have perceived (thought, known) before; what the
Arabic text says is rather that even if that is so, we do not remember through
recourse to sense-perception, but by means ofa higher, 'nobler' faculty that
is removed from the corporeal objects of sense-perception, dealing instead
with 'abstract', immaterial objects. From the perspective of the adaptation
this is a masterful reinterpretation of the passage: building on Aristotle's own
words, it promotes a concept of memory quite different from that of the
Stagirite. Most importantly, the adaptor avoids calling memory a 'state or
affection connected with perception'; according to the Arabic text, memory
is distinct from perception and of a very different quality.
The at first sight implausible negation in clause (1) hence turns out to be
very well suited to the adaptor's purposes, and this may tempt one to think
he inserted the negation deliberately, especially as this negation is implausible
enough in the context of the original text as to make one wonder whether
the translator should not have realised the mistake. On the other hand, there
would have been no particular need for the adaptor to take this step which,
after all, resulted in the statement of (1) - a statement that even within the
adaptation strikes one as baffling and uncalled for, and that moreover is at
odds with the paragraph directly preceding it (cf. De mem., 1, 449b9-2l).
Without the crucial negation, clause (1) would have been perfectly compat-
ible not only with the preceding, but also with the following lines, where
the text comes back again to the point that memory is to do with the past
(6, 7). There would have been little need or motive to change (1), and there
would have been easier ways for the adaptor to avoid committing himself
to theoretical points he did not want to endorse. AB there is a convincing
textual explanation for the appearance of the negation it is, on the whole,
more plausible to assume that it was present already in the translation of
the Aristotelian text, and that the theoretically charged parts of the passage
(2, the second part of4, 5) were inserted by the adaptor in order to do away
with the inconsistencies it created. That he should do so in the terms of his
own favoured theory comes as no surprise.
The hand of the adaptor is even more visible in Brib al-Nawm wa-l-yaqa;a,
where we can see him reworking the Aristotelian text in an even bolder,
and more deliberate way than in the part corresponding to De memoria.
The following example stems from the first part ofBrib al-Nawm wa-l-yaqa;a, a
paragraph marking the transition from a passage of comparatively faithful
translation ofAristotelian text (De som., 1, 453bll-31) to the first passage
of mainly additional material in the chapter.
ROTRAUD HANSBERGER
ii) Ms. Rampur 1752, foL 21b, 6 - 22a, 24
(1) This may [be shown to] prove true, and can be known, from [consid-
ering] the waking and the sleeping person. For a sleeper may petceive
many things without doubting that those things that he is perceiving in his
sleep are there while he is awake. (2) The difference between the percep-
tion of the waking and that of the sleeping person consists in that the
sleeper perceives from inside only, whereby that [kind of] perception of
his [takes place] without any movement on his part, whereas the waking
person perceives from outside, that [kind of] perception [taking place]
through movement. (3) Let us therefore say: the difference between sense-
perception of the sleeping and of the waking person has become clear,
[i.e.] which one of the two [types of] sense-perception is more apt and
more correct; and this ought to be recognised and known. (4) The sense-
perception of the sleeper, we shall then say, is sense-perception potentially,
whereas the sense-perception of the waking person is sense-perception actu-
ally. (5) Whatever is potential is hard to perceive, whereas what is actual can
be perceived and known. (6) However, the sense-perception of the sleeper,
even though being potential, may emerge into actuality; although some of
it will emerge in a clear and plain manner, while some of it will be difficult
[to perceive] and unclear. (7) As for [the question which one is] the most
perfect and the noblest ofthe two: the spiritual is nobler than the corporeal.
However, [22a] the spiritual is not [considered] nobler than the corporeal
by the corporeal, nor is the corporeal [considered] nobler than the spir-
itual by the spiritual; rather, the spiritual is [considered] nobler than the
corporeal by the spiritual, whereas the corporeal is [considered] nobler than
the spiritual by the corporeal; but it is not at all possible that the spiritual
should be [considered] nobler by the corporeal, whereas it may indeed be
possible that the spiritual, which we have said to be potential, is [consid-
ered] nobler by man than the corporeal, which we have said to be actual.
(8) Evidence for the spiritual being nobler than the corporeal is that the
spiritual indicates what will come to be in the future, whereas the corpo-
real indicates what has come to exist at the present time only. (9) When a
person unifies his faculties through the most subtle of things and makes
them one, he will see that thing which he sees potentially just as someone
who sees it actually. It is because his faculties are separated that a man is
prevented from seeing things potentially in the same manner as when he
sees them actually...
Ar., De som., I, 453b31 - 454a4; 454a7-11
(1) En bE Kat EK Twvbs bf]AOV" 4J yap TOV rPTlYOpOW wu'tC(l
Kat TOV Ka8unvouVTa'
KITAB AL-lfISS WA-L-MAlfSOS
(2) TClV bE ata8avo).!svov EYPTlYOPEVUl KUt TOV EYPTlYOPOTU
reuna 11 TWV nvoe; uia8uvw8ar11 TWV EV ulJ't('p
[ ... ] End b oihs Tf]e; IjfUxf]e; '{blOV TO uta8Uvsa8ar oihs TOU aw).!uwe;
(4) (OD yap r, MVuflle;, TOUTOU Kat r, EVEPYS1U'
r, bE ASYOflEVTl u'(a8TlGle; we; EVEPYStU KtVTlcrte; ne; bta TOU aWflUTOe; Tile;
Ijfuxf]e; Ean), <puvspov we; oihs Tfje; IJfUxf]e; TO nu8oe; '(blOV, oih' UljfuXOV
aWflu bUVUTOV uta8uvsa8ar.
(1) Again, the point is clear from the following. We recognize a person
as sleeping by the same mark as rhat by which we recognize someone as
waking. (2) It is the person who is perceiving thar we consider to be awake;
and we take every waking person to be perceiving either somerhing external
or some movement wirhin himself [... ] But given that perceiving belongs
neither to the soul nor to the body solely (4) (for what owns any capacity
also owns its exercise; and what is called perception, in the sense of exer-
cise, is a certain movement of the soul by means of the body), it is plain
that the affection is not peculiar to the soul, nor is a soul-less body capable
of perceiving.
21
The most prominent (and crucial) feature ofthis paragraph is the dichotomy
of spirituality and corporeality which, as I have mentioned above, is one of
the favourite themes of the adaptor. We can see here how certain concepts
stemming from the Aristotelian text, i.e. internal and external perception,
movement, potentiality and actuality, are fitted into this dichotomy, which is
at the same time equated with the distinction between waking and sleeping.
Aristotle's description of the waking state - always to be perceiving some-
thing, be it external objects or events within ourselves - is split up, resulting
in a comparison of two opposite states: while external perception remains a
characteristic ofthe waking state, internal perception is now made a feature
of sleep. As a result, sleep and waking are not distinguished any longer by
presence and absence of perception, but by two different kinds of percep-
tion. Again, in Aristotle's text, a 'movement within oneself' constitutes an
object ofperception during the waking state. In the Arabic text, 'movement
within' remains a feature of the waking state, but no longer as an object of
perception: it represents the mode in which perception takes place, having
become a criterion for the distinction between external and internal percep-
tion. The former is associated with movement, the latter with rest. The two
keywords 'potentiality' and 'actuality' (dunamis/energeia, clause 4), taken up
21. English translation by D. GALLOP, Aristotle On Sleep andDreams. AText and Translation
with Introduction, Notes and Glossary, second edition, Warminster, Aris & Philipps,
1996, p. 61 ff.

8' 1:ponov nvo: 1:J1V aK1VTj(J{UV KUt olov /iE(JJlQV 1:QV

UTl:VOV dvu{ <pUJlEV, 1:J1V 86 Matv KUt 1:J1V aVEatV ypnyopatv.
And we maintain that sleep is, in a certain way, an immobilization or
'fettering' of perception; whereas its liberation or release is waking. 23
(That the expressions 'bond of the faculties' and 'union of the faculties'
refer to the same thing is evident from a further passage (fol. 23b-24a),
where both terms are used.)
The quoted examples stand for many more passages of Kitab al-fjiss wa-I-
mal;sUs in which we can see the adaptor actively interfering with the Aristotelian
text, reshaping it for his purposes. Such passages are particularly instructive
in that they can tell us something about the adaptor's very own interests and
concerns. The main theories that the adaptation of Kitlib al-fjiss wa-I-ma!zsus
puts forward, such as the three-faculty theory, are presented most clearly
in the longer, continuous passages of additional material. These theories,
however, are not the adaptor's free invention; based as they are on Greek
medical theory and philosophy, he will have drawn on Greek, or Graeco-
Arabic, sources. So far, we do not know the specific sources he used; hence we
cannot tell whether he quoted literally or freely, or how much of the added
material in Kitlib al-fjiss wa-I-mabsiis he actually composed entirely himself.
In those places, however, where he joins Aristotelian text and added mate-
rial together, we can be certain to see himself at work, and the fruits of his
creativity tell us something about his own preferences - as, for example,
iii) Ms. Rampur 1752, fol 27b, 20j
If this is so, then waking will inevitably [equal] the unfastening of the
bond of the faculties, whereas sleep will [equal] the [fastened] bond of
the faculties. Thus it has been established clearly and soundly that waking
is the opposite of sleep.
Ar., De som., I, 454b25-27
. Sleep, spirituality, and perception of the future are thus associated with
each other. 22 The other aspects mentioned in the table have supportive char-
acter. The integration of rest, potentiality, and internal perception into the
system is, as we have seen, motivated by keywords in the Aristotelian text.
The same is true of union versus separation of the faculties, as transpires
further down in the text:
KITAB AL-IfISS WA-L-MAIjSrJS
22. The adaptor seems to ignore an inconsisrency thar rhis dichotomy inrroduces inro his
rext rhrough suggesring thar dreams generally are endowed wirh divinatory power, a posi-
rion rhar rhe rexr dearly does not hold.
23. English translation by D. GALLOP, op. cit., n. 21, p. 67.
ROTRAUD HANSBERGER
actuality
external perception
separation of faculties
waking
corporeality
perception of present
motion
potentiality
internal perception
union of faculties
sleep
spirituality
perception of future
rest
isolation from their immediate context, are integrated into the system
ith a similar manoeuvre. Potential perception is assigned to sleep, actual
perception to waking. In (7) and (8), the 'potential' is finally identified as
'spiritual', perception during sleep hence declared spiritual; the 'actual' is iden-
tified as 'corporeal', and perception during waking accordingly pronounced
corporeal. Spiritual things are then declared to be 'nobler' and 'more perfect'
than corporeal things: while the corporeal (object of perception, we have
to understand) can indicate only what exists at present, the spiritual points
to things that will come to exist in the future.
Obviously it has been the adaptor's aim, in redesigning the paragraph,
to link the perception of future events not only to the realm of the spir-
itual, but also to sleep, thus setting the stage for the theory of divinatory
dreams. The adaptor achieves this by associating sleep with 'internal' and
'potential' sense-perception - both of them notions that are inspired by
keywords in the Aristotelian text. This is then relatively easily linked to
the spiritual realm, by contrasting it to 'external', 'actual' sense-perception,
which belongs to waking, informs us of the present state of things and is
associated with the material world.
How 'potential', 'spiritual' perception is supposed to work at a psycho-
logical level, so that one will indeed perceive things that are potential as if
they were actual (i.e. one will perceive future events as if they were reality
already) is indicated in the immediately following passage. This kind of
perception, we learn, requires the 'union' of the 'three faculties'. Once they
are separated, and perform each their proper function only, the faculties
will not be able to carry out the function of internal perception properly.
There will still be perception during sleep, but it will consist in dreams that
do not portray things correctly, i.e. that are not foretelling the future.
Even before the discussion of divinatory dreams gets under way, Bab
al-Nawm wa-Iyaqr1fa thus presents its reader with a clear-cut dichotomy that
is obviously developed with this topic in mind. Collecting a number of
aspects, the adaptor assigns them neatly to two sides:
I
ROTRAUD HANSBERGER
his predilection for dividing things up along the lines of the dichotomy
between spirituality and corporeality.
The careful work that the adaptor has performed in adjusting the
Aristotelian text to his favoured theory demonstrates that the text trans-
mitted in Arabic as Aristotle's Kitiib al-Ijiss wa-I-mal;siis was never supposed to
be a mere collection of several authors' proclamations on a certain range
of topics. It was meant to be an adaptation. Whether its author (I mean
the adaptor) meant it to be transmitted under Aristotle's name is a ques-
tion that one may speculate about; so far, however, there is no evidence to
the contrary.
5. Transmitted under Aristotle's name, Kitiib al-Ijiss wa-I-mabsiis was read
and drawn upon by several mediaeval Arabic scholars. How big its impact
really was is yet to be determined by future research, but it is already clear
that the text had considerable influence. However, it is first and foremost
the ideas contributed by the adaptation, namely the three-faculty theory
and the theory of veridical dreams, that commanded the scholars' interest,
inspiring philosophical theories of prophecy and divination (naturally a
salient topic in Islamic philosophy). Elements of the theories ofKitiibal-Ijiss
wa-I-mal;siis are reflected in theories of dreaming and ofprophecy in the work
ofal-Parabi (d. _950),24 even though he does not explicitly quote Kitiibal-ljiSs
wa-I-mabsiis as his source; and the Arabic Parva naturalia have most certainly
been a source of inspiration for important parts ofIbn Sinas philosophical
psychology, including, most significantly, his famous theory of the 'internal
senses'.25 Furthermore the text may have inspired the theory of veridical
dreams in the Book ofthe Elements by the Jewish philosopher and physi-
cian Isaac Israeli (-855 - -955).
26
1he 'Brethren of Purity' Ukhwiinal-Safii:
10
th
century) could also have been influenced by the Arabic Parva natu-
ralia, a point that is yet to be studied. Kitiib al-ljiSs wa-I-mal;siis was, however,
ofinterest not only to philosophers: it was, for example, also quoted by Abu
Sa'd al-Dinawari (fl. -1010), the author of a dream manual.l7
Kitiib al-Ijiss wa-I-mal;siis received a keen response also in mediaeval Spain,
where it found the interest of both Muslim and Jewish thinkers, and from
where, through Ibn Rushd's (d. 1198) Epitome ofthe Parva naturalia (TalkhIs
24. E.g. in his Mabiidi'arii'ahlal-madinaal-fiil/zla; see R. HANSBERGER, op. cit., n. 7, p. 73 f.
25. See R. HANSBERGER, op. cit., n. 3, p. 221 If. Ibn Sinas authorship of al-Risiilaal-Maniimiyya,
which mentions rhe text explicitly, is still under discussion; orher texts that reflect Kitiib
al-fliss include al-Shlfii' and al-Nojiit.
26. See R. HANSBERGER, op. cit., n. 7, p. 72 f.
27. Kitiibal-Ta'bii-jl/-ru'yiiawal-Qiidirifil-ta'bir, see R. HANSBERGER, op. cit., n. 7, p. 69 If.
KITAB AL-ijISS WA-L-MAl;iSUS
Kitiib al-Ijiss wa-I-mal;siis) its influence was to reach the Latin West.
28
Until
Kitiib al-Ijiss wa-I-mal;siis was rediscovered, Ibn Rushd's Epitome was the most
important witness to the text. As we know now, it is based on a version of
Kitiib al-Ijiss wa-I-mal;siis that must have been the same as, or very similar to,
the one preserved in the Rampur manuscript.
29
Before Ibn Rushd, it is in
particular the philosopher Ibn Bajja (d. 1138) who, with his theory of'spir-
itual forms', enthusiastically embraced some of the characteristic ideas of
Kitiib al-Ijiss wa-l-mabsiis.
30
The reception of the Arabic Parva naturalia, in particular its theories
concerning sleep and dreams, within the Jewish scholarly community is
currently being studied by Hagar Kahana-Smilansky.31 Direct quotations
from Kitiib aI-Hiss wa-I-mahsiis are found in the work Kitiib al-Muhiidara wa-I-
mudhiikara by Moshe ibn ;Ezra (-1060 - -1140)32 in the context'ofpoetical
inspiration during sleep; and quotations in Hebrew translation, again from
passages devoted to the theory of divination in dreams in Biib al-Nawm wa-I-
yaqCl?a, are attested in the 13
th
century Commentary on Maimonides' Guide
ofthe Perplexed by R. Zera.1:Iya b. Isaac b. She'alt!' el J:Ien.
33
Scholars that used
the Arabic Parva naturalia as a source further include Maimonides (I 135-
1204), Joseph ibn (d. 1149), and Joseph ibn Waqar (d. -1360).34
While the reception of Kitiib al-Ijiss wa-I-mal;siis is well attested from the
late 10th and particularly the 11 th century onwards, its early fate is much
more in the dark: in the case ofearlier authors like the philosopher al-Kindi
(d. - 870) or the physician Qusta ibn Luqa (d. 912), who wrote on the topic
ofsleep and dreams as well as on that oflength and shortness oflife, it so far
has not been possible to detect any particular influence of Kitiib al-ljiSs wa-I-
ma/;siis, and it is not easy to ascertain whether they were actually familiar
with the text itself and, if they were, whether that version ofKitiib al-Ijiss wa-I-
mal;siis would have been the same as the one extant in the Rampur manu-
script. 35 This is a particularly intriguing question with regard to al-Kindi,
given that the adaptation ofKitiib al-ljiSs wa-I-mal;siis shows so many affinities
28. See C. DI MARTINO, "Les Parva naturalia dans la rradition arabe", in R. GOULET, Dictionnaire
des Philosophes Antiques, supp!. 1, Paris; Editions du CNRS, 2003, p. 375-378, p. 378.
29. See R. HANSBERGER, op. cit., n. 3, p. 230 If.; and R. HANSBERGER, op. cit., n. 7, p. 67 f.
30. See R. HANSBERGER, op. cit., n. 7, p. 67.
31. See her forthcoming article for details, op. cit., n. 8. I am indebted to Dr. Kahana-
Smilansky for allowing me to consult a draft of her article.
32. See R. HANSBERGER, op. cit., n. 7, p. 66, and cf. S. PINES, op. cit. n. 13, p. 147 If.
33. See A. RAVITZKY, "Hebrew Quotations from the lost Arabic Recension of Parva natu-
ralia", Jerusalem Studies in Arabic and Islam, 3 (1981-82), p. 191-202 (1981-82), and
R. HANSBERGER, op. cit., n. 7, p. 68 f.
34. As documented in H. MHANA-SMILANSKY, op. cit., n. 8.
35. See R. HANSBERGER, op. cit., n. 7, p. 241 If.; for al-Kindi see also P. ADAMsoN,AI-Ki'ndi,
New York-Oxford, Oxford University Press, 2007, p. 135 If.
I
ROTRAUD HANSBERGER
to texts produced in the 'Kindi circle'. In other cases, e.g. al-Parabi and Isaac
Israeli, who take up some of the characteristic ideas of the adaptation of
Kitab al-fjiss wa-l-ma(zsTis, one still has to take into account the possibility that
they may have encountered these ideas in other sources, perhaps the same
that were used by the adaptor of the Parva naturalia.
This uncertainty is reflected in reports about Aristotle's Parva naturalia
in Arabic bibliographical works, starting with the Fihrist of Ibn al-Nadim
(d. 990), which record information about Kztab al-fjiss wa-l-mabsUs that does
not square with the extant text. Ibn al-Nadim states that Kitab al-fjiss wa-l-
mabsUs consists of two maqalas, and that no reliable translation of the work is
known. 36 While the latter claim may arguably be compatible with what we
find in Kitab al-fjiss wa-l-ma(zsUs, the first one is clearly wrong. Ibn al-Nadim
goes on to say: "What has been mentioned is that it is a small piece which
al-Tabari has commented on after Abu Bishr Matta ibn Yunus".37 Apart from
the question of how Kitab al-fjiss wa-l-mabsTis could be described as a 'small
piece' - could this perhaps refer to passages of Aristotelian text only? Or
is Ibn al-Nadim speaking of a different text altogether? -, it has not been
possible so far to link the translator Abu Bishr Matta ibn Yiinus (d. 940)
to the extant text, and the same goes for al-Tabari.
38
The bibliographical
sources
39
hence raise similar questions about the Arabic Parva naturalia to
those brought up by the analysis ofKitabal-fjisswa-l-ma!zszis: Was there ever a
(more) complete Arabic translation of the Parva naturalia? At what point
of time did the extant version of Kitab al-fjiss wa-l-mabsUs begin to be trans-
mitted, and to be accepted, as the translation of the Aristotelian work?
And who are the people responsible for the text? Despite the new insights
the text of Kitab al-fjiss wa-l-mabsUs offers, there are still many questions that
surround this work, and the transmission ofAristotle's Parva naturalia in
Arabic certainly needs and deserves further study.
36. Cf. M.R. TAJADDUD ed., Ibna/-Nadim:Kltiiba/-fihrist, Tehran, 1971, p. 312.
37. For the Arabic text, see M.R. TAJADDUD, ibid., r,. 312.
38. On the identity of this person, see A. HASNAWl, 'Un eleve d'Abu BiSt Matta b. Yunus:
Abu 'Amr al-TabatI", Bulletin d'itudes orientales, 48 (1996), p. 35-55, p. 35 If., 40 If.
39. Cf. R. HANSBERGER, op. cit., n. 3, p. 253 If., where further bibliographical sources
(partly dependent on Ibn al-Nadirn) are discussed as well; see also H. DAIBER, op. cit.,
n. 4, p. 39 f.
KITAB AL-IfISS WA-L-MAlfSUS
Textual Examples from Kitab al-fjiss wa-l-mabsiis:
i) Ms. Rampur 1752, fols. 10b, 25 - lla, 12
Ij! ct.......l Ji 0Y'""':'i J1 .L.....:..1 J1 J1.).1 01.) L, 4i (1)
J1 Ji Il.b [lIa] J! .G,li J.U
(3) JJ-;>oJ L1"" ..Jyl.1 y>).,1 JJ-;>oJ t.i>-! Lw! (2) cGY'""':'1
(4)
(6) fljJl Cai..Q (5)
J.U .:,L,)I Ji !..AlJJ-;>oJ .:,L,j 4S.)J; .:,i (.'tJ (7) .:,L,j .u.:
... fljJl fljJl
ii) Ms. Rampur 1752, fols. 21 b, 6 - 22a, 24
.lQ .:J!jJ L1"" Il.b F .lQJ (1)
(2) ......". d.;, Y>J
.:J!j.L.....:..J Lt-:' \?jJl J.-.QJI
J
.lQ L! (3) .:J!jJ L1"" L1""
L1""J &J1 Lt-:' L,
:.,...,.. (4)
.:,1S L,J 4.5:1.).11 y.=. .G,li Y> L, JsJ (5)
T-:"..9.p- (..P= L, cC..o -..1! (..P= .ill .:,1S.:,lJ :""""J (6)
Cajj (7) y.=. Y> L, cC..oJ 6.....:,I
J
1.;
.L..c L1"" ...Jyl.i :;bJ)1 [22a] L1"" ..Jyl.1
- L1"" L1"" .jyl.1

-'1 c .lQ J.> ' .L..c""...s 1'b)l .<.. i __ .
.. ..r. 'Y""..r '-'
.G1 l:.flj csjJl .).1 .L..c ...Jyl.i l:.flj csjJl
-..k JIJ L1"" ...Jy1.1 -..k Oi)
.:,t.:j L,--..k JIJ Y> Lw1 .:,L,)I .)1Sy> L,
.:J!j cs.>: 0l,.Q .).i 6) .b.iQ
cs.>: .:,i L1"" .).1 LwlJ cs.>: csjJlS 01.>:
-- -' 1--'1.., I I. WI.'I
'" ul UO'-"''7 .>: l..4 0-""" . _ l
4

I
ROTRAUD HANSBERGER
iii) Ms. Rampur 1752, fo!' 27b, 20
'11 J)l:...:,1 d' illJl"j djb..,)\j I.AA l.:u 'IS Ij!..Q
(5.Y'" .) i"-J-'"' J (5.Y'" .) ..r-. U U
...
Le feu agit-il en tant que feu?
Causalite et synonymie dans les Quaestiones
sur le De sensu et sensato de Jean de JandunI
]EAN-BAPTISTE BRENET
Universite de Paris-Ouest Nanterre-La Defense
Num ignis agat in quantum ignis? Cette question est extraite des Quaestiones
sur le De sensu et sensato du maitre es arts parisien du XIV" siecle Jean de
Jandun, prince des averro'istes ,,2. Loin de se reduire au probleme technique
du mode d'action d'un element, dIe lance, dans une ample dispute ou figure
l' essentid du peripatetisme greco-arabe, une reflexion sur la causalite effi-
ciente al'a:uvre dans l' ensemble des generations substantielles du monde
sublunaire. On voudrait en prendre la mesure et en saisir I'issue. Deux choses
importent dans les commentaires medievaux d'Aristote: le decoupage du
texte qu'ils proposent (ce qu'ils retiennent, isolent, negligent, rdient - la
maniere, done, qu'ont les auteurs d'investir le texte de base); et les boule-
versements theoriques, explicites, maltrises, deliberes ou non, qu'ils lui font
subir, i.e. leur charge proprement conceptuelle. Guide par Averroes, Jean
de Jandun mise sur la synonymie pour sauver Aristote contre Platon et ses
successeurs. Sous ce double rapport, son analyse merite examen.
Jean pose cinq questions apropos de la saveur, qui occupe le chapitre 4
du traite De la sensation et des sensibles d'Aristote
3
. La troisieme, sur le feu,
1. Je tiens it remercier chaleureusement Cristina CERAMI pour son aide dans I' elaboration
de cet article. J' ai pu beneficier de son livre, Generation et substance. Aristote etAverroes entre
physique et metaphysique, avant sa parution prochaine (chez De Gruyter, coil. "Scientia
Graeco-Arabica). Sans ses conseHs, ses critiques, ce travail n'aurait jamais trouve sa forme
actuelle.
2. Sur]ean de Jandun (m. 1328), voirJ.-B. BRENET, Tramfertsdusujet. La noetiqued'Averroes
selon Jean de Jandun, Paris, Vrin, 2003, et les references bibliographiques donnees en fin
de volume. Pour une presentation succincte de l'"averrolsme latin, voir ID., Averroes,
commentateur ou depravat.eur? ", dans M. A1u<:OUN dir., Histoire de l'islam et des musul-
mans en France, du MoyenAge anosjours, Paris, AIbin Michel, 2006, p. 223-235. Comme
texte de base, nous utilisons Ioannis de Iandunophilosophi acutissimi Quaestiones super Paruis
Naturalibus, Venise, apud Haeredem Hieronymi Scoti, 1589.
3. Elles suivent le mouvement du texte, comme le voulait l'usage. La premiere (q. 21):
l'humide est-Hie sujet propre de la saveur (utrum humidumsitpropriumsubiectum saporis) ?
La deuxieme (q. 22): le chaud et le sec sont-ils les causes actives de la saveur (num calida
/ ~ L fJ'-"'-- <;....'./ ~ - - - - --v
Serie Philosophie - 28
Universite Paris 1 - Pantheon-Sorbonne
Les Parva naturalia d'Aristote
Fortune antique et medievale
Sous la direction de
CHRISTOPHE GRELLARD et PIERRE-MARIE MOREL
Ouvrage publit avec le concours
du Conseil scientifique de Paris 1
Publications de la Sorbonne
2010
Illustration de couverture:
Arisrote, Traittfs et opuscules, 6, De sensu et sensato, xme-Xlv" siecle.
Bibliotheque municipale de Tours, ms 679, f. 166v. CNRS-IRHT.
Composition rypographique: Benedicte Chantalou/Yuruga
Publications de la Sorbonne, 2010
212, rue Saint-]acques, 75005 Paris
www.univ-paris1.fr
Loi du 11 mars 1957
ISBN 978-2-85944-635-2
ISSN 1255-183X
Presentation
CHRISTOPHE GRELLARD
Universite Paris 1 Pantheon-Sorbonne
PIERRE-MARIE MOREL
ENS de Lyon
Les Petits traites d'histoire naturelle d'Aristote -les Parva naturalia pour
reprendre le titre latin en usage depuis le xm
e
siecle - posent une ques-
tion essentielle pour toute enquete sur l'ame: quelles sont les operations
ou les affections que nous pouvons dire communes aI'ame et au corps?
Ces opuscules formulent et developpent en efIet, au coeur de la psycho-
logie entendue en son sens etymologique, une problematique que l' on
dirait aujourd'hui physiologique . Qu'ils enquetent sur la sensation et
les sensibles, la memoire, le sommeil et les reves, la longevite du vivant, la
respiration ou encore la vie et la mort, ils convergent tous pour confronter
la definition generale de I'ame aux proprietes du corps. Sous ces dernieres,
comprenons aussi bien: les proprietes elementaires, comme le chaud et le
froid; les proprietes organiques, comme la resistance des tissus, les mouve-
ments du souffle; les accidents du vivant, comme les impressions sensibles,
les maladies ou l' epaisseur du sang.
Cette maniere d'aborder la question de l'ame est incontestablement
originale. La psychologie du Timee de Platon constitue, il est vrai, un
precedent remarquable, qu'Aristote n'ignore pas et qu'il doit affronter en
plusieurs lieux de nos traites. Il n'est pas certain, cependant, qu'elle exerce
sur la philosophie de l'ame aproprement parler une influence comparable
acelle d'Aristote. Il est vrai que le Timee sera la reference obligee en ce qui
concerne le theme de l'animation cosmique pour toute la tradition plato-
nicienne posterieure. Toutefois, Aristote refuse d' envisager serieusement
qu'il y ait une ame du monde)} et sa philosophie de !'ame ouvre une tout
autre perspective. Appliquant les grands principes formules dans le traite
De i'ame, les Parva naturalia etablissent un lien nouveau, avec une precision
encore inegalee dans le corpus philosophique, entre la traditionnelle et vene-
rable conception de l'ame comme principe de mouvement et de connais-
sance, et cette partie de la philosophie naturelle qui prend avec le Stagirite
un essor considerable: la science du vivant.
I
LES PARVA NATURALIA D'ARISTOTE
YACK B., "AReinterpretation ofAristotle's Political Teleology", History ofPolitical
Thought, 12 (1991), p. 15-33.
ZIMMERMANN A., Verzeichnis ungedruckter Kommentare zur Metaphysik undPhysik
des Aristoteles aus der Zeit von etwa 1250-1350, Leyde-Cologne, Brill, 1971.
ZIMMERMANN EW, "The Origins of the so-called Theology of Aristotle , dans
]. KRAYE, WE RYAN, C.B. SCHMITT dir., Pseudo-Aristotle in the Middle Ages.
The Theology and other Texts, Londres, Warburg Institute, 1986, p. 110-240.
ZUPKO ]., John Buridan. Portrait ofa Fourteenth Century Master ofArts, Notre-
Dame, Notre-Dame University Press, 2003.
Table des matieres
Presentation , 5
Christophe Grellard, Pierre-Marie More!
Anomalies de l'intelligence, intelligence de l'anomalie.
Note sur la representation de l'organisation du corps vivant
chez Aristote entre les Parva naturalia et les Problemes 11
Andrea L. Carbone
Lejardin des Parva naturalia: les plantes chez Aristote et apres lui 31
Luciana Repici
Stages in the reception ofAristotle's works on sleep anddreams
in Hellenistic and Imperialphilosophical and medical thought 47
Philip]. Van der Eijk, Maithe Hulskamp
Le Commentaire du De sensu par Alexandre d'Aphrodise 77
Carla Di Martino
Aristotle's De memoria and Plotinus on memory 101
Richard A.H. King
La premiere reception du De memoria et reminiscentia
au Moyen Age Latin: le commentaire d'Adam de Buckjield 121
]ulie Brumberg-Chaumont
Kitab al-Ijiss wa-l-mabsus. Aristotle's Parva naturalia in Arabic Guise 143
Rotraud Hansberger
Lefeu agit-il en tant quefeu? Causalid et synonymie
dans les Quaestiones sur le De sensu et sensato de Jean de Jandun 163
Jean-Baptiste Brenet
Secundum viam naturae et docrrinae. Lire le De motu animalium
et les Parva naturalia d'Aristote au Moyen Age 197
Pieter De Leemans
La reception medievale du De somno et vigilia. Approche anthropologique
et epistemologique du reve, d'Albert le GrandaJean Buridan 221
Christophe Grellard
PostJace
Apropos de l'enluminure de couverture: une incitation aLa recherche 239
Max Lejbowicz
Bibliographie 249