As it is currently known, the extant corpus of the Arabic Plotinus (AP) is constituted by three distinct groups of fragments: the Theology of Aristotle (ThA),Footnote 1 the Letter on Divine Science attributed to al-Fārābī (DS),Footnote 2 and the fragments transmitted as the Sayings of the Greek Sage (GS); this last group of fragments is scattered over three different textual sources: the Oxford Anthology Footnote 3 (preserved in a manuscript in the Bodleian Library in Oxford, MS Bodl. Or. Marsh 539), the Kitāb al-Milal wa-al-niḥal by al-Shahrastānī, and the Muntakhab Ṣiwān al-Ḥikma. Footnote 4 These fragments all represent material from Plotinus' Enneads IV-VI in a paraphrastic translation that is interwoven with material original to the Arabic version, ranging from single sentences to whole paragraphs.Footnote 5 Since Franz Rosenthal first postulated a common source for all these fragments in 1952,Footnote 6 this thesis has been firmly established and elaborated upon by subsequent scholarship, even though the exact character, origin and scope of this original Arabic Plotinus Source (*AP) are still a matter of dispute and speculation.
There is now broad scholarly consensus that this translation-cum-adaptation was produced within the circle of al-Kindī,Footnote 7 and that the translation was made by the Kindī-circle translator ‘Abdalmasīḥ b. ‘Abdallah b. Nāʿima al-Ḥimsī directly from the Greek rather than via a Syriac intermediary.Footnote 8 However, there are diverging views on the question whether al-Ḥimsī was responsible for the translation only, or for the adaptation as a whole – the position taken by Fritz Zimmermann, and, following him, Peter AdamsonFootnote 9 – and to what degree other agents, in particular al-Kindī himself, may have been involved. Thus Cristina D'Ancona has argued for a more active role for al-Kindī, as author of at least some of the original material of the adaptation, and as the person responsible for the re-ordering of material that is visible in ThA. Footnote 10 The different opinions on this issue are in some respects linked to the way in which scholars interpret the Prologue to ThA (which names Aristotle as author, Porphyry as commentator, al-Kindī as editor, and al-Ḥimsī as translator of the text), as well as the so-called Headings, a catalogue of questions based on Ennead IV.4 that precedes the first chapter of ThA.Footnote 11
A further point of contention is the question whether *AP preserved the order of the treatises devised by Porphyry, or whether the translator / adaptor of the text rearranged the material according to his own purposes. This is closely connected to the further question whether *AP would have covered all of Enneads IV-VI (or at least all treatises of Enneads IV-VI, even if in excerpts only), or whether it included only selected passages from the start.Footnote 12 The textual evidence is inconclusive. Neither DS nor GS can on their own be expected to tell us much about the original structure of *AP. With ThA, the largest text of the extant Arabic Plotinus corpus (AP), the case is different. On the one hand it provides evidence that the translation was made from a Greek manuscript containing the Enneads in the Porphyrian order,Footnote 13 naturally a conditio sine qua non for any possibility of its preserving this order. On the other hand, in the state in which ThA has come down to us it clearly does not preserve this order. Whether the structure of ThA is entirely the result of the labours of a later editor who – after a ‘physical catastrophe’Footnote 14 during which chunks of *AP were lost and the rest of it thrown into disorder – joined (parts of?) the surviving material together as best as he could, or whether (and to what extent) the present arrangement of ThA reflects decisions of the Kindī-circle adaptor, and hence possibly of the translator, remains a matter of dispute. Fritz Zimmermann argues for the former scenario; he assumes that al-Ḥimsī's translation-cum-adaptation preserved the order of the Enneads and included ‘rather more of Enn. IV-VI than is covered by the extant fragments’,Footnote 15 perhaps even covering Enneads IV-VI ‘without major gaps’.Footnote 16 Peter Adamson, on the other hand, submits that the scope and structure of ThA are not entirely fortuitous, but reflect a conscious selection and reordering by the translator/adaptor (al-Ḥimsī) that occurred already at the level of *AP. Footnote 17 He thinks it plausible that the material was ‘divided into thematic sections’, and that the translator / adaptor was ‘at least to some extent selective in choosing what to paraphrase’,Footnote 18i.e. that at least some of the gaps in the text may be his responsibility. Cristina D'Ancona likewise argues that the arrangement of material in ThA is at least partly due to a ‘process of rational editing’,Footnote 19 which in her view, however, was performed by al-Kindī ‘on the basis of a running translation already made’.Footnote 20
The treatises that are missing from AP Footnote 21 contribute significantly to the blurring of the overall picture. Naturally, the negative evidence they offer is unreliable and subject to potential revision (of which this article shall be proof). For the moment, however, it does not sit badly with Peter Adamson's supposition that ThA reflects a section of *AP devoted to the topic of ‘soul’.Footnote 22 This would be consistent with the absence of, e.g., Enneads VI.1–3 (On the Kinds of Being), or Ennead VI.6 (On Numbers). However, under the additional assumption (in which Peter Adamson seems to follow Fritz Zimmermann) that *AP, whatever shape it may have taken, was part of a larger Kindī-circle collection of texts related to metaphysics,Footnote 23 there is much less reason to assume that these treatises were entirely left aside by the original translator/adaptor.
Ennead IV.6 is a different case. The treatise On Sense-Perception and Memory clearly falls within the domain of psychology and thus may have had a good claim to being included within a section on the soul – and of course it may well have been. On the other hand, the treatise does not, perhaps, fit the metaphysical turn of ThA, or indeed any ‘metaphysics file’Footnote 24*AP may have belonged to, focussing, as it does, on two functions of the soul connected solely with its attachment to the sensible, material world: sense-perception and memory.Footnote 25 It is therefore all the more intriguing that we can find fragments of Ennead IV.6.1 and IV.6.3 in an Arabic text that is devoted specifically to such functions: Kitāb al-Ḥiss wa-al-maḥsūs, the Arabic version of Aristotle's Parva Naturalia.
I. THE SOURCE TEXT: MAQĀLA I OF KITĀB AL-ḤISS WA-AL-MAḤSŪS
Kitāb al-Ḥiss wa-al-maḥsūs has come down to us in a 17th century manuscript held at the Raza Library, Rampur (India), MS Rampur 1752.Footnote 26 The acephalous text comprises three maqālāt, Maqāla I being equivalent to De sensu et sensato, Maqāla II to De memoria et reminiscentia, De somno et vigilia, De insomniis, and De divinatione per somnum (the latter three compounded in one chapter, Bāb al-Nawm wa-al-yaqaẓa), and Maqāla III to De longitudine et brevitate vitae. However, the Arabic text contains but little of a translation of the Aristotelian text. Most of it is made up of added material that appears to be original to Kitāb al-Ḥiss wa-al-maḥsūs. Footnote 27 This material is also Greek in character, drawing on Neoplatonic philosophy and Galenic medicine. Its most remarkable feature is a theory of the three ‘spiritual faculties’ of imagination (or ‘formative faculty’, muṣawwir), thought (fikr), and memory (dhikr), which also encompasses a theory of divinatory dreams.Footnote 28
We do not know who exactly was responsible for the translation or the adaptation of Kitāb al-Ḥiss wa-al-maḥsūs; but there is evidence that its production was linked to the Kindī-circle.Footnote 29 The translation of the Aristotelian passages – as far as it can be assessed – was probably produced prior to the process of adaptation. It shows terminological similarities to texts like the early De anima translation by Ps.-Isḥāq and the doxography of Ps.-Ammonius, which has been associated with the Kindī circle.Footnote 30 It is the added material and the overall adaptation however, that in particular bear the hallmarks of Kindī-circle texts.Footnote 31 This is especially true of Maqāla II. Not only does it abound in the introductory, summarising and connecting phrases typical for Kindī-circle texts; more significantly (in particular for our present purposes), it shares terminological preferences and certain doctrinal propensities with the Arabic Procliana and Plotiniana in particular. This concerns first and foremost its concept of spirituality (rūḥāniyya), and a dualistic tendency that is accompanied by a certain disdain for the corporeal as opposed to the ‘spiritual’ world, with the human soul precariously posed in-between the two.
It is not Maqāla II however, where borrowings from Ennead IV.6 can be detected, but Maqāla I. This is unfortunate in so far as Maqāla I is a much more problematic text to study: its larger part is missing in the Rampur manuscript; only its last 6 1/2 pages are extant.Footnote 32 These pages do not contain any text that could be linked to Aristotle's De sensu. As far as it is extant, Maqāla I consists entirely of additional material that does not have a counterpart, however tentative, in the Greek Parva Naturalia. Instead, major parts of the text can be traced back to chapters 1 and 3 of Plotinus' Ennead IV.6.
The absence of any Aristotelian text in the extant pages of Maqāla I does not indicate that they do not originally belong to Kitāb al-Ḥiss wa-al-maḥsūs: we find extensive passages consisting solely of additional material in other parts of Kitāb al-Ḥiss wa-al-maḥsūs as well, and especially so in the later sections of its treatises or chapters.Footnote 33 The lack of any trace of De sensu in the last pages of Maqāla I is not, therefore, out of character.
More intriguing is perhaps the fact that the extant pages of Maqāla I are not very well represented in Ibn Rushd's Talkhīṣ Kitāb al-Ḥiss wa-al-maḥsūs. As even a cursory comparison of the Talkhīṣ with Maqāla II and III of Kitāb al-Ḥiss wa-al-maḥsūs can reveal, Ibn Rushd's epitome does indeed rely on the version of the Arabic Parva Naturalia that is transmitted in the Rampur manuscript. Given that the first part of Maqāla I is missing in the manuscript it is hence still our most important witness for the Arabic version of De sensu.Footnote 34 Granted, the Talkhīṣ generally does not follow its source text particularly slavishly. All the same it is bewildering how hard it is to detect any traces of the extant part of Maqāla I in it: they are restricted to a few rather brief references to the question, discussed on fols 7a–8a of the Rampur manuscript, whether sense-perception involves impressions in the soul (see below).Footnote 35 The remainder of the text (fols 8b–10a) is not referred to explicitly at all. That these last pages of Maqāla I were nevertheless present in Ibn Rushd's copy of Kitāb al-Ḥiss wa-al-maḥsūs is, however, indicated by a faint but distinct echo that they have left at the end of the first maqāla of the Talkhīṣ, which Ibn Rushd concludes with the following words:Footnote 36
وأمّا ما يذكر في آخر هذه المقالة من إعطاء السبب في جيدة قوّة الذكر وضعفها فالموضع اللائق به هو عند التكلّم في المقالة الثانية في القوّة الذاكرة
What he (i.e. Aristotle) says at the end of this maqāla in terms of giving the causes for excellence or weakness of the faculty of memory has its appropriate place where he discusses the memorative faculty in the second maqāla.
Ibn Rushd's curt remark appears to point directly to the extant pages of Maqāla I of Kitāb al-Ḥiss wa-al-maḥsūs, which indeed focus more on memory than on sense-perception. The misplaced discussion of the topic may have been enough justification for Ibn Rushd to pass it over in this manner; however, additional motivation may have been provided by the fact that what is said about memory at the end of Maqāla I of Kitāb al-Ḥiss wa-al-maḥsūs does not square entirely with the presentation of the concept of memory in Maqāla II.Footnote 37 Both factors are due to the same cause (which will not have been discernible to Ibn Rushd): the presence, in Maqāla I, of material from a Plotinian treatise entitled Περὶ αἰσθήσεως κaὶ μνήμης – On Sense-Perception and Memory.
II. FRAGMENTS FROM ENNEAD IV.6
Ennead IV.6 is divided into three chapters. In the first chapter Plotinus takes issue with the Aristotelian theory that sense-perceptions are impressions received by the senses as the impression of a signet ring is received in wax, an example that Aristotle uses in De memoria and De anima. Footnote 38 This is not the only treatise where Plotinus discusses the topic of sense-perception, or, indeed, mentions the image of signet ring and wax.Footnote 39 What he is particularly concerned with in Ennead IV.6 are the implications such an understanding of sense-perception would have for a concept of memory: as he rejects the Aristotelian idea that sense-perception leaves impressions in the soul, he will also have to reject the understanding of memory as the retention of such impressions. In the second chapter, Plotinus provides an alternative account emphasising the active role of the soul in sense-perception. In the third chapter he turns to his account of memory, again emphasising that memory is an activity of the soul rather than a passive affection.
Most key passages in the extant part of Maqāla I of Kitāb al-Ḥiss wa-al-maḥsūs – more than half of the text – have a parallel in Ennead IV.6.1 or IV.6.3. They will be presented here together with the corresponding passages from Ennead IV.6.Footnote 40
1. Perception and physicality
In the second paragraph of the extant part of Maqāla I, the text argues against the idea that sense-perceptions are literally ‘formed’ in the soul, by pointing to the impossibility of accommodating such physical impressions of sense objects (which would have a certain size and would exist in space) within the limits of our imagination:
i) MS Rampur 1752, fols 7a, 20 – 7b, 5:Footnote 41
وذلك 
توهّموا الصور المحمولة بالمتوسّطات جسمانيّة فإنّا نقول ألا ترى أنّ الصور الجسمانيّة تأخذ من المكان قدراً ما والصور الروحانيّة لا تمسك مكاناً والدليل على ذلك ما نقول ألا ترى أنّ العالم كلّه 
في وهمك فإن أنكرت أن يكون هذا هكذا قلنا ألا ترى أنّ في وهمك مائة صورة قد نظر إليها بصرك 
فإذا 
في أخذ البصر فهي في الوهم بقدر نظرك إليها في حدّ الجرميّة فإن كانت في الوهم على قدر نظرك إليها في حدّ الجرميّة فمحال لأنّه ليس في الوهم من المكان ما تتصوّر صورة واحدة مكانيّة فقط فكيف مائة صورة
For they fancy the forms that are carried by the intermediaries to be corporeal. So we shall say: Do you not see that corporeal forms take up some measure of space, whereas spiritual forms do not lay hold of any space? Proof of that is what we shall say now: Do you not see that the whole world is formed in your imagination? If you deny that this is so, we shall say: Do you not see that there are hundred[s of] forms in your imagination which your gaze has met and witnessed? Once you do not have them within your sight [any more], they will be in [your] imagination in the [same] measure in which your gaze [has taken] them [in] in the realm of corporeality. But if they are [supposed to be] in [your] imagination in the measure in which your gaze [has taken] them [in] in the realm of corporeality, that will be absurd. For there is no space in the imagination where even a single spatial form could be formed – so how [a fortiori] could hundred[s of] forms [be formed there]?
A similar point is made in Ennead IV.6.1. As one of several difficulties associated with the theory that understands sense-perceptions as impressions in the soul, Plotinus mentions that it could not account for our ability to judge the size of objects we perceive.
Ennead IV.6.1, 26–28 (pp. 322f.):
τό τε μέγεθος αὐτοῦ, ὅσον ἐστὶν ἔξω, πῶς ἂν ὅσον ἐστὶ λέγοι, ἢ ὅτι μέγα, οἷον τὸ τοῦ οὐρανοῦ, τοῦ ἐν αὐτῇ τύπον τοσοῦδε εἶναι οὐ δυναμένου;
And how would it [i.e. the soul, RH] be able to state the size as it is outside or say that it is large, to state for instance the size of the sky, when it would be impossible for as large an impression to be inside it?
2. Focus and repetition as prerequisites for memory
A closer correspondence between Maqāla I and Ennead IV.6 can be observed in the following passage, where the text explains how memory is supposed to come about. (The phrases have been numbered to facilitate comparison with the Greek.)
ii) MS Rampur 1752, fols 7b, 24 – 8b, 2:
فنقول (1) إنّ النفس إذا حرّكت الحسّ المشترك نحو شيء ما وألقى الحسّ المشترك بصره على ذلك الشيء الأرضيّ وأطال الحسّ 
بصره على ذلك الشيء وصورته حتّى ينظر إليه البصر نظراً طويلاً في زمان طويل كانت صورة ذلك الشيء أثبت وأتقن في الوهم وكانت النفس أشدّ له ذكراً وكان 
أحسن إتقاناً لصورته 
لم تشغل وهمها بشيء آخر سوى تلك الصورة (2) والدليل على هذا القول الصبيان 
أذكر للأشياء من أجل 
لا يشغلون فكرهم وأوهامهم بأشياء كثيرة بل إنّما 
بذلك الشيء الذي يشهونه ويلتذّون به فقط وكذلك البصر إذا شغلته بالنظر إلى صورة ما ولم تصرفه إلى صورة غيرها حتّى ينظر إليها نظراً دائماً ويتفكّر التصوّر لإتقان تلك الصورة ويحفظها الوهم ويذكرها الذكر ثبتت تلك الصورة ورسخت فإذا طلبت أحضرت من كلّ ناحية يحضرها المصوِّر والوهم والذكر فهي أبداً موجودة عند الحاسّ المشترك (3) فأمّا الذين يشتهون ويلتذّون إلقاء أبصارهم على صور كثيرة وإذا نظروا إلى صورة ما لم يدعوا النظر إليها حتّى يتقنها المصوِّر ويحفظها الذكر فإنّ تلك لا يقدر صاحبها على إحضارها في كلّ وقت بل تضمحلّ وتذهب ولن يقدر على إحضارها إن طلبها إلاّ كالخيال وذلك أيضاً تعب وعسر ومشقة ونصب وأقول أيضاً (4) لولا أنّ الصورة إنّما تؤدّيها المتوسّطات 
كانت تضمحلّ أبداً من النفس ولا كانت تنسى صورة أبداً ولو كانت النفس هي التي تظهر حتّى تلقي بصرها على المحسوس وينطبع المحسوس في النفس احتاجت النفس إلى هذه القوى الثلاث أعني الفكر والذكر 
وذلك أنّ الصورة كانت فيها منطبعة ولكن لما صارت النفس إنّما تقبل من الحسّ المشترك وكان الحسّ المشترك إنّما يقبل الصور من المتوسّط احتاجت النفس إلى هذه القوى الثلاث لتحفظ وتذكر وتتقن الصور كيما إذا احتاجت إلى أن تحضر صورة ما 
بهذه القوى الثلاث
We shall say: (1) When the soul moves the common sense towards something, and the common sense casts its gaze on that earthly thing, and the sense casts its gaze on that thing and its form for such a long time that [its] gaze will rest on it extensively for a long while, then the form of that thing will be more firmly established and more perfect in the imagination (wahm), the soul will be more effective in remembering it, and the formative [faculty] (muṣawwir) will be better at perfecting its form. For the soul will not have been keeping its imagination occupied with anything apart from that form. (2) Evidence for this claim are the young, for they are best at remembering things, due to the fact that they do not occupy their thoughts and their imaginations with many things, but occupy them exclusively with one particular thing they desire and take pleasure in. Likewise, if you keep your gaze occupied by looking at a form, and do not avert it towards another form, so that it will look at it continuously, and so that fancy (taṣawwur)Footnote 42 will be deliberating in order to perfect that form, imagination will retain it and memory will remember it, then that form will be firmly established and well-rooted. Hence when it is sought, it will be made present in every aspect which the formative [faculty], imagination and memory can make present. Thus [the form] will always be found within the common sense. (3) However, persons who desire and take pleasure in casting their gaze on many forms, and who do not, when they look at a form, fix their gaze on it long enough until the formative [faculty] has perfected it and memory has retained it, will not be able to make such a form present at any and every moment, but it will vanish and go away. If they seek it, they will not be able to make it present, except as a phantasm (khayāl), and even that will be tiring, difficult, troublesome and laborious. I shall also say: (4) If it were not the intermediaries that conveyed the form, it would never vanish from the soul, and [the soul] would never forget a form. If it were the soul itself that emerged to cast its gaze on the perceptible, and the perceptible were [thus] imprinted in the soul, the soul would not need these three faculties – I mean [the faculty of] thought, [the faculty of] memory and the formative [faculty] – because the form would be imprinted in it. However, since it is from the common sense that the soul receives [forms], whereas the common sense receives the forms from the intermediary, the soul needs these three faculties in order to retain, remember and perfect the forms; so that when it needs to present a form it will present it by means of these three faculties.
This is paralleled in a (much briefer) passage of Ennead IV.6.3.
Ennead IV.6.3, 19–29 (pp. 328f.):
(1) ὅταν τοίνυν ῥωσθῇ πρὸς ὁτιοῦν τῶν φανέντων, ὥσπερ πρὸς παρὸν διάκειται ἐπὶ πολὺν χρόνον καὶ ὅσῳ μᾶλλον, τόσῳ ἀεί. (2) διὸ καὶ τὰ παιδία μνημονεύειν λέγεται μᾶλλον, ὅτι μὴ ἀφίστανται, ἀλλὰ κεῖται αὐτοῖς πρὸ ὀμμάτων ὡς ἂν ὁρῶσιν οὔπω εἰς πλῆθος, ἀλλὰ πρὸς ὀλίγα· (3) οἷς δὲ ἐπὶ πολλὰ ἡ διάνοια καὶ ἡ δύναμις, ὥσπερ παραθέουσι καὶ οὐ μένουσιν. εἰ δέ γε ἔμενον οἱ τύποι, οὐκ ἂν ἐποίησε τὸ πλῆθος ἧττον μνήμας. (4) ἔτι, εἰ τύποι μένοντες, οὐδὲν ἔδει σκοπεῖν, ἵνα ἀναμνησθῶμεν, οὐδὲ πρότερον ἐπιλαθομένους ὕστερον ἀναμιμνήσκεσθαι κειμένων.
(1) When, therefore, the soul is strongly moved to anything that appears to it, it is for a long time in a state as if the object was present to it, and the more strongly it is moved, the more lasting the presence. (2) This is why children are said to be better at remembering, because they do not go away from things, but they remain before their eyes, since they do not look at a multiplicity of things, but at few; (3) but those whose thought and soul-power are directed to many things, as it were, rush past them and do not linger on them. But if the impressions remained, their multiplicity would not make memories less. (4) Further, if impressions remained, there would be no need for us to consider in order to remember, nor should we forget before and remember afterwards if the impressions were lying ready to hand.
The passage that immediately follows in Maqāla I again has a parallel in Ennead IV.6.3.
iii) MS Rampur 1752, fol. 8b, 2–5:
والدليل على ما ذكرنا وأنّ النفس قد تحتاج إلى هذه القوى الثلاث 
إذا سمعت شيئاً ما مرّة أو مرّتين لم تقدر على حفظه حتّى تسمعه مراراً كثيرة حتّى يبقيه الحفظ فيرسخ عنده فإذا رسخ ذلك الشيء عند الحفظ ذكره متى شاء وإذا شاءت النفس أن تحضره أحضرته
What we have just said, i.e. that the soul needs these three faculties, is evident from [the following]: When [the soul] hears something once or twice it will not be able to retain it until it will have heard it so many times that retentive memory makes it remain, so that [the form] will be [firmly] rooted within it. Once that thing is [firmly] rooted in retentive memory, [retentive memory] will remember it whenever it wishes, and whenever the soul wishes to make it present it will do so.
Ennead IV.6.3, 33–38 (pp. 328ff.):
διὰ τί γὰρ ἅπαξ μὲν ἀκούσας ἢ δεύτερον οὐ μέμνηται, ὅταν δὲ πολλάκις, καὶ ὃ πρότερον ἀκούσας οὐκ ἔσχε, πολλῷ ὕστερον χρόνῳ μέμνηται ἤδη; οὐ γὰρ δὴ τῷ μέρη ἐσχηκέναι πρότερον τοῦ τύπου· ἔδει γὰρ τούτων μεμνῆσθαι· ἀλλ᾿ οἷον ἐξαίφνης γίγνεται τοῦτο ἔκ τινος [τῆς] ὑστέρας ἀκροάσεως ἢ μελέτης.
For why, when one has heard something once or twice, does one not remember it, but [only] when one has heard it many times, and, when one has heard something before and not retained it, why does one remember it long afterwards? It is certainly not because one previously had the parts of the impression; for then one would have to remember these; but this happens with a kind of suddenness, as the result of some later hearing or exercise.
3. Retentive memory and quickness of mind
In De memoria (449b6–8), Aristotle observes that ‘it is not the same people who are good at remembering and at recollecting’ and that ‘slow people are better at remembering, while those who are quick and learn well are better at recollecting’.Footnote 43 This point is taken up in our next passage.
iv) MS Rampur 1752, fol. 8b, 12f.:
ومن الواجب أن يكون جيّد الحفظ جيّد الذكر ولأنْ يحضر الأشياء إذا أحبّ أن يكون عارفاً بالإشارة
It is a necessary rule that a person of good retentive memory must also have good recollective memory, and that in order to make things present whenever he wishes, he must be acquainted with the [relevant] pointer.
The same point is referred to by Plotinus in Ennead IV.6.3:
Ennead IV.6.3, 63–69 (pp. 332f.):
γίγνονται δὲ ὡς ἐπὶ τὸ πολὺ οὐχ οἱ αὐτοὶ μνήμονες καὶ ἀγχίνοι [πολλάκις], ὅτι οὐχ ἡ αὐτὴ δύναμις ἐκατέρου, ὥσπερ οὐδ᾿ ὁ αὐτὸς πυκτικὸς καὶ δρομικός· ἐπικρατοῦσι γὰρ ἄλλαι ἐν ἄλλῳ ἰδέαι. καίτοι οὐκ ἐκώλυε τὸν ἁστινασοῦν ἔχοντα πλεονεξίας ψυχῆς ἀναγινώσκειν τὰ κείμενα, οὐδὲ τὸν ταύτῃ 
υέντα τὴν τοῦ πάσχειν καὶ ἔχειν τὸ πάθος ἀδυναμίαν κεκτῆσθαι.
But the same people do not as a general rule have both good memories and quick minds, because it is not the same power in each case, just as the same man is not a good boxer and a good runner: for different characteristics dominate in different people. And yet there would be nothing to prevent a man who had any kind of superiority of soul from reading off the deposited impressions, nor would a man who had a strong inclination this way have to possess an inability to be affected and to retain the affection.
While Plotinus agrees with the factual claim, but turns it into an argument against Aristotle's theory of memory as the retaining of impressions, Kitāb al-Ḥiss wa-al-maḥsūs here seems to claim that people with good retentive memory do have good recollective memory – in contrast to Plotinus and Aristotle, and so also to the second maqāla of Kitāb al-Ḥiss wa-al-maḥsūs, which reproduces the point Aristotle makes in De memoria.Footnote 44
4. Critique of the concept of memory as retention of impressions
Immediately afterwards, the Arabic text rephrases the leading topic of Ennead IV.6 in argumentative style:
v) MS Rampur 1752, fol. 9a, 5–19
فإن قال قائل إنّكم قد نفيتم أن تكون الحسائس تقبل رسوم الصور قبولاً جسمانيّاً فكيف يمكن أن نقول كما قال الأوّلون إنّ الذكر إنّما حفظ رسوم الأشياء وقد نفينا الرسوم فثبّتنا أنّ الحسائس إنّما تقبل رسوم الأشياء بمتوسّطات من غير أن تكون الرسوم تنطبع في الحسائس انطباعاً جسمانيّاً
If someone asks: You have denied that the sensory organs receive the pictures of forms in a corporeal manner; if so, how can we say, as the ancients did, that memory is the retention of the pictures of things, given that we have denied the [existence of such] pictures? [The answer is that] we have already established that the sensory organs receive the pictures of things through intermediaries only, without the pictures being imprinted in the sensory organs in a corporeal manner.
This takes up the introductory statement Plotinus sets down at the beginning of Ennead IV.6.1 (1–5, pp. 320f):
Tὰς αἰσθήσεις οὐ τυπώσεις οὐδ᾿ ἐνσφραγίσεις λέγοντες ἐν ψυχῇ γίγνεσθαι, οὐδὲ τὰς μνήμας πάντως τε καὶ ἀκολούθως ἐροῦμεν κατοχὰς μαθημάτων καὶ αἰσθήσεων εἶναι τοῦ τύπου μείναντος ἐν τῇ ψυχῇ, ὃς μηδὲ τὸ πρῶτον ἐγένετο.
Since we say that sense-perceptions are not impressions or seal-stamps on the soul, we cannot say either with any sort of consistency that memories are retentions of what we have learnt or perceived while the impression remains on the soul which was not there to begin with.
Given that the problem of sense-perception, intermediaries, and the retention of pictures has been discussed in previous paragraphs already whereas now the text is turning to memory, the second part of the Arabic passage may seem somewhat out of place or redundant.Footnote 45 However, the question whether memory has anything to do with the retention of pictures in the soul is indeed touched upon again in the immediately following passage (quoted below), which goes back to Ennead IV.6.3. This link could be the reason why the adaptor thought it a convenient piece of text to create the transition from the topic of sense-perception to that of memory.
5. Lapse of time between perception and memory
vi) MS Rampur 1752, fols. 9a, 19 – 9b, 6:
وإن قال قائل إن كان الذكر يذكر الأشياء بمتوسّطات من غير أن تكون رسوم الصور الجسمانيّة عنده مرسومة فما باله لا يذكر جميع الأشياء بل إنّما يذكر واحداً بعد واحد وشيئاً بعد شيء قلنا إنّ هذه المسألة واجبة عليكم وذلك أنّكم لو قلتم إنّ صور الأشياء إنّما تدركها النفس بأن تظهر فتلقي بصرها على الشيء المحسوس فينطبع رسم تلك الصورة في النفس وقواها فإن كانت تنطبع رسوم صور الأشياء في النفس وقواها فما بالها تحتاج إلى أن تذكر شيئاً بعد شيء ولا تحضر الشيء كلّما احتاجت إلى إحضاره على الصواب لأنّ الرسوم فيها
If someone says: If memory remembers things through intermediaries, without any pictures of corporeal forms being engraved in it, then why does it not remember all things [at once] but can only remember them one by one, one thing after another?, we shall say: It is you who are obliged [to answer] this question, [not we]. For if you were saying that it was the soul that perceived the forms of things by emerging to cast its gaze on the perceptible thing, so that the picture of that form was imprinted in the soul and its faculties, why, then, if the pictures of the forms of the things were imprinted in the soul and its faculties, would [the soul] need to remember one thing after another, and could not make a thing present correctly whenever it needed to do so, given that the pictures would be inside it?
In Ennead IV.6.3, Plotinus asks why, on his own understanding of memory, we will only remember some time after having had the relevant perception.
Ennead IV.6.3, 55–59 (pp. 330f.):
ἔτι τῶν αἰσθήσεων τυπώσεων οὐκ οὐσῶν, πῶς οἷόν τε τὰς μνήμας κατοχὰς τῶν οὐκ ἐντεθέντων οὐδὲ τὴν ἀρχὴν εἶναι; ἀλλ᾿ εἰ δύναμίς τις καὶ παρασκευὴ πρὸς τὸ ἕτοιμον, πῶς οὐχ ἅμα, ἀλλ᾿ ὕστερον εἰς ἀναπόλησιν τῶν αὐτῶν ἐρχόμεθα;
Further, when sense-perceptions are not impressions, how could memories be retentions of imprints which were never made [in the soul] at all? But if memory is a power and a preparation for readiness, why do we not come to recall the same things at once, but only later?
There is a notable difference between the two texts. The problem envisaged by Plotinus is how to explain that we remember only after time has lapsed, rather than immediately after perception. According to the text of Maqāla I, on the other hand, the challenge for a viable theory of memory would be to explain why we remember different forms (or impressions, according to the fictional opponents' theory) one after the other rather than all at once. This deviation could well be the result of a mistranslation of ἅμα, rendered by the translator as ‘together at the same time’ rather than as ‘immediately’.
6. The example of signet ring and wax
As if recapitulating, the last sentence of Maqāla I once again turns against the proponents of the doctrine that conceives of sense-perceptions as impressions in the soul, encapsulated by Aristotle's metaphor of the imprint of a signet ring in wax.
vii) MS Rampur 1752, fol. 10a, 3–7:
وقد أخطأ أيضاً مَن زعم أنّ رسوم الأشياء قد تنطبع في النفس كانطباع نقش الخاتم في الشمع
Again, anybody who claims that the pictures of things are imprinted in the soul like the engraving of a signet ring is imprinted in wax is mistaken as well…
Aristotle's example is mentioned and criticised by Plotinus twice in Ennead IV.6.1:Footnote 46
Ennead IV.6.1, 1–2 (pp. 320f.):
Tὰς αἰσθήσεις οὐ τυπώσεις οὐδ᾿ ἐνσφραγίσεις λέγοντες ἐν ψυχῇ γίγνεσθαι
Since we say that sense-perceptions are not impressions or seal-stamps on the soul…
Ennead IV.6.1, 18–20 (pp. 320ff.):
ἅτε μηδενός, οἶμαι, τύπου ἐν αὐτῇ γενομένου ἢ γιγνομένου, οὐδέπω σφραγῖδα ὥσπερ ἐν κηρῷ δακτυλίου λαμβανούσηςFootnote 47
… since, I think, no impression has been or is being imprinted on it, nor has it yet received a seal-stamp, like the mark of a seal-ring on wax.
Either phrase could have been the basis for the Arabic, although the second one is a touch more likely, since it mentions the wax explicitly.
The parallels between the Greek and Arabic passages quoted in this section establish without doubt that Plotinus' Ennead IV.6 in one way or other served as the ultimate source for these parts of Maqāla I of Kitāb al-Ḥiss wa-al-maḥsūs. This leads to a further question: how and in what form did the Plotinian material enter Kitāb al-Ḥiss wa-al-maḥsūs, and what, if anything, do these fragments have to do with AP?
III. MAQĀLA I OF KITĀB AL-ḤISS WA-AL-MAḤSŪS AND THE ARABIC PLOTINUS
The fact that we find borrowings from Ennead IV.6 in Maqāla I of Kitāb al-Ḥiss wa-al-maḥsūs could mean one of two things: either the passages drawing on Plotinus were already present in the Greek Vorlage of the Arabic translation of the Parva Naturalia and were translated into Arabic as part of the Greek treatise, or else they were added to the text at the Arabic end; in which case the adaptor responsible for the inclusion of these quotations must have had access to an Arabic translation of Ennead IV.6.
This is not the place to discuss whether Kitāb al-Ḥiss wa-al-maḥsūs represents the Arabic translation of a Greek (or Syriac) adaptation of the Parva Naturalia, or whether its adaptation was effected in Arabic, i.e. after or at the same time as the translation of the Aristotelian text. A detailed analysis of Maqāla II that I have undertaken elsewhereFootnote 48 has shown that the text is in all likelihood the work of an Arabic adaptor who was working with an already existing translation of (parts of) the Parva Naturalia – whatever the exact shape and state of the Greek Vorlage of that translation may have been.Footnote 49 This result is further supported by the circumstantial evidence of the text's relationship with several Kindī-circle texts,Footnote 50 which suggests that Kitāb al-Ḥiss wa-al-maḥsūs is indeed another adaptation created by one or several members of this group. This very fact at the same time renders it likely that the adaptor of Kitāb al-Ḥiss wa-al-maḥsūs had access to the Kindī-circle translation of Plotinus' Enneads IV–VI. Conversely, the presence of quotations from that translation in Kitāb al-Ḥiss wa-al-maḥsūs would once more re-enforce the thesis of its probable origin in the Kindī circle.
As mentioned before, Ennead IV.6 is not represented in the extant Arabic Plotinus corpus (AP) as it is currently known. If the Plotinian borrowings in Kitāb al-Ḥiss wa-al-maḥsūs could be shown to be hitherto unknown fragments not just of any rendering of Plotinus, but of the Kindī-circle translation of Enneads IV-VI, this would indicate that the original Plotinus source *AP contained a translation of IV.6 (at least of chapters 1 and 3), supporting the assumption that the extant AP does not reflect the full coverage of the treatises in the original *AP.
Is there anything to show that the borrowings from the Enneads found in Kitāb al-Ḥiss wa-al-maḥsūs go back to the supposed Arabic Plotinus source? In order to find out, their language and style will have to be compared with that of the extant Arabic Plotinus. One difficulty that arises in this context is that it is impossible to tell whether the adaptor of Maqāla I quoted literally from an Arabic translation of the Enneads or not. The ‘Plotinian’ passages in Maqāla I render the Greek text rather freely, and show a tendency to elaborate on certain points in a somewhat repetitive manner. This happens to be a stylistic feature of the Arabic Plotinus, too;Footnote 51 and it could, therefore, simply go back to the original Kindī-circle translation. The paraphrastic features of the quotations could thus even be counted as a point in favour of their being taken from *AP. However, they could also have been the work of the adaptor, or even of successive adaptors, of Kitāb al-Ḥiss wa-al-maḥsūs: liberties with the Greek, frequent elaborations and repetitiveness are characteristics of the Arabic Parva Naturalia no less than of AP. Hence these features may not with ultimate certainty be attributed to the translation. Such uncertainty applies in equal measure where omissions from Plotinus' text are concerned: they could have occurred in the translation, but could just as well have been the result of the adaptor's selection of material.
Notwithstanding these caveats, even a brief comparison indicates a likely relation between the Plotinian passages in Maqāla I of Kitāb al-Ḥiss wa-al-maḥsūs and AP, as I hope to show. For the comparison I have chosen a passage from Sayings of the Greek Sage V, corresponding to Ennead IV.5[29].1, 3–17 and 2, 49–50. Should the original Arabic translation indeed have followed the order of the Enneads (as suggested by Fritz Zimmermann), the passage would have been in close vicinity to the translation of IV.6. Being devoted to sense-perception, it is furthermore related thematically, and hence seems a suitable text to compare with the Plotinian borrowings in Maqāla I of Kitāb al-Ḥiss wa-al-maḥsūs.
Sayings of the Greek Sage V (Badawī, Aflūṭīn, pp. 189f., 191):
قال الشيخ اليوناني: البصر وسائر الحسائس لا تنال المحسوس إلا أن يكون بينها وبينه جرم متوسّط. وذلك أنّ الحسّ إنّما هو من حيّز النفس. والنفس إذا كانت في المكان العقلي علمت الأشياء العقلية من غير أن تحتاج إلى شيء من الأجرام وإذا كانت في العالم الحسّي احتاجت في معرفة المحسوسات إلى الجسم. والنفس إذا اتّحدت بالأشياء المحسوسة فإنما تنال معرفتها بالأشياء الشبيهة 
القوّية على قبول الآثار المحسوسة فتحسّ النفس الأشياء المحسوسة وتعرفها بتوسّط. والبصر يدرك الألوان بمتوسّط يقبل أثر اللون أوّلاً ثم يؤديه إلى البصر […] فإن قال: إن صور الأشياء المحسوسة إن لم تكن معلّقة محمولة في الهواء اضمحلّت وبادت قبل أن تأتي البصر فلا يقع عليها البصر البتّة فلذلك صار الهواء يقبل صور الأشياء المحسوسة أوّلاً ثمّ يقبلها بعد ذلك البصر بتوسّط الهواء … .
Al-Shaykh al-Yūnānī said: Vision and the other senses attain the objects of sense-perception only if [senses and objects] are connected through the existence of an intermediary bodi[ly substance]. Sense-perception belongs to the sphere of the soul. While the soul is in the realm of the intellect, it knows the intelligibles and has no need of any bodi[ly substance], [in order to know them]. However, while it is in the world of the senses, it needs the body, in order to know the objects of sense-perception. United with the perceptible objects, the soul attains the knowledge of them only through things which are similar to them and able to receive sensual impressions. Thus, the soul perceives and knows the objects of sense-perception through mediation. Vision achieves the perception of colours through a medium which first receives the impression of colour, and then transmits it to vision […] He might say: If the forms of the objects of sense-perception are not attached to the substratum of air, they would dissolve and disappear before they come to vision, and they could not be seen at all. It is for this reason that the air first receives the forms of the objects of sense-perception and that, then, vision receives them through the mediation of the air. (Translation based on F. Rosenthal, ‘Aš-Šayk al-Yūnānī’ II, pp. 371, 373, 377, with some alterations.)
The passage contains several expressions that we have encountered in the excerpts of Kitāb al-Ḥiss wa-al-maḥsūs quoted above, in particular texts ii) and iii): baṣar (‘gaze’), nāla / qabila ṣuwar al-ashyāʾ (‘to attain / receive the forms of things’); mutawassiṭ (‘medium’ / ‘intermediary’); iḍmaḥalla (‘to vanish, dissolve’), addā (‘to convey’, i.e. the perception or form). This list can be supplemented with further expressions that the ‘Plotinian’ passages of Maqāla I share with other passages of AP:
– The characteristic expression alqā baṣarahu (‘to cast / direct one's gaze’, see text ii above) is used time and again in AP. Here are a few selected examples:
ThA II (Badawī, Aflūṭīn, p. 33, l. 13):
وأما النفس 
تستحيل إذا أرادت علم الأشياء ، وذلك تلقي بصرها على الأشياء كلها بحركتها المائلة.
ThA II (Badawī, Aflūṭīn, p. 35, l. 10f.):
ثمّ اطّلعت على هذا العالم وألقت بصرها على شيء من الأشياء دون العقل فاستفادت الذكر هي وصارت ذات ذكر.
ThA VIII (Badawī, Aflūṭīn, p. 120, l. 9f.):
لأنّ النفس إنّما هي صنم، إلاّ إذا ألقت بصرها على العالم العقلي ازدادت حسناً
– The expression aṭāla ilqāʾ al-baṣar (‘to prolong one's gaze’) in Maqāla I (see text ii above) is related to the synonymous expression aṭāla al-naẓar which is used in AP:
ThA VIII (Badawī, Aflūṭīn, p. 117, l. 15f.):
والبصر كلّما أطال النظر إلى الشيء المحسوس أضرّ به المحسوس حتى يصيّره خارجاً من الحسّ أي لا يحسّ شيئاً
– Also notable is the choice of the word jirmiyya (‘corporeality’) in text i above; Kitāb al-Ḥiss wa-al-maḥsūs normally uses the term jusmāniyya. The Arabic Plotinus uses the cognate adjective jirmānī (Badawī, Aflūṭīn, e.g. p. 36), alongside jusmānī / jusmāniyya (Badawī, Aflūṭīn, e.g. pp. 34, 41, 42) and jusdānī (e.g. p. 41).
– Finally, the word wahm, likewise an irregular term within Kitāb al-Ḥiss wa-al-maḥsūs that is only found in Maqāla I, is used in the Arabic Plotinus for (the faculty of) imagination (Badawī, Aflūṭīn, pp. 38, 193).
All in all, the evidence does suggest a dependence of Maqāla I on the Arabic Plotinus source, even if it cannot ultimately prove it. Such an assumption has the further advantage of being able to explain the differences in doctrinal and linguistic details that exist between Maqāla I and II of Kitāb al-Ḥiss wa-al-maḥsūs. The idea that the adaptor made use of the Kindī-circle Plotinus also suits the fact that a more general affinity to AP can be found in Kitāb al-Ḥiss wa-al-maḥsūs, especially in Maqāla II. At the beginning of Bāb al-Nawm wa-al-yaqaẓa, for instance, the concepts of ‘potentiality’ and ‘actuality’ are introduced in a way that is reminiscent of the discussion of potentiality and actuality in Mīmar VIII of ThA. Footnote 52 This passage of the Theology has no equivalent in the Greek Enneads. The affinity therefore points to a connection between the two Arabic adaptations. Footnote 53
IV. THE PLOTINIAN FRAGMENTS WITHIN KITĀB AL-ḤISS WA-AL-MAḤSŪS
When did these fragments of the Arabic Plotinus become part of Kitāb al-Ḥiss wa-al-maḥsūs? Had they always been an integral part of the adaptation from the start, or were they added at a later date? Since so little of the first maqāla of Kitāb al-Ḥiss wa-al-maḥsūs is extant, it is hard to tell how intimately the Plotinian borrowings were connected to its other, more ‘Aristotelian’ parts. Ibn Rushd's testimony does not give us much to go by, for the reasons mentioned above. From what he says we can nevertheless surmise that it was a discussion of the role of the medium in sense-perception that may have occasioned the inclusion of the Plotinian passages. According to the Talkhīṣ, Kitāb al-Ḥiss wa-al-maḥsūs seems to have contained an account of four different views held by ‘the ancients’ on this matter.Footnote 54 The central topic in the extant pages of Maqāla I is the question whether sense-perceptions are ‘impressed’ directly on the soul, or whether they are transmitted to the soul via ‘intermediaries’. Dominated by the Plotinian borrowings, the text attacks the Aristotelian view that compares sense-perceptions to impressions of a seal in wax, and argues for the more Plotinian stance that the soul cannot be directly affected by objects of sense-perception but needs an intermediary organ or faculty to mediate the affection. In Plotinus' doctrine (cf. Ennead IV.4.23) this is so because the human soul, belonging to the intelligible realm, cannot be assimilated to the sensually perceptible and hence needs an instrument to bridge the essential difference between itself and the perceptible.Footnote 55 In Maqāla I, these considerations do not really transpire. In fact, there may even exist some confusion between the Plotinian ‘intermediaries’ and the external ‘medium’ (air, water) which according to Aristotelian doctrine is necessary for the transmission of sense-perception.Footnote 56 The discussion in Maqāla I may hence have been occasioned by a passage in the Aristotelian text referring to the external medium, with the Plotinian text being brought in by the adaptor to supply further (and different) points.
The transition to the topic of memory would have been brought about by the Plotinian borrowings: as we have seen, Ennead IV.6 combines the two topics of sense-perception and memory, whereas Kitāb al-Ḥiss wa-al-maḥsūs, following the layout of the Parva Naturalia, in principle reserves the latter for the first part of Maqāla II. In pursuing the question how memory can be said to retain the pictures (rusūm) of things if they are not imprinted in the soul in a physical manner, Maqāla I is following the lead of the Plotinian material.
What remains unclear is the motivation behind the inclusion of this material, especially as Plotinus' account of memory (and sense-perception) is not quite consistent with that of Maqāla II of Kitāb al-Ḥiss wa-al-maḥsūs: in Maqāla II the Aristotelian comparison to the signet ring and wax is maintained as valid, and there is talk of forms being imprinted in the soul;Footnote 57 moreover, the theory of the three spiritual faculties,Footnote 58 pervasive in Maqāla II, does not have an equivalent in the Plotinian text. As one can see from texts ii) and iii) quoted above, such discrepancies are smoothed over by additions and alterations which adjust the Plotinian passages to the theory of the three faculties promoted in Maqāla II of Kitāb al-Ḥiss wa-al-maḥsūs. In effect, Plotinus' stance against the idea that sense-perceptions are to be conceived of as impressions in the soul is turned into an argument for the necessity of the ‘mediating’ faculties – identified not as the sense-organs (as in Plotinus), but as the ‘spiritual’ faculties of imagination, thought, and memory. The text furthermore shifts the emphasis towards another issue close to the heart of the adaptor: the dichotomy between corporeality and spirituality.Footnote 59 Focussing on the point that there can be no physical impressions in the soul, the adaptor of our text stresses that it is, instead, spiritual forms that are imprinted in the soul's faculties.Footnote 60
This manner of picking out isolated points from the ‘original’ text and distorting them in order to convey a different sense altogether is absolutely characteristic of the adaptor we find at work in Maqāla II of Kitāb al-Ḥiss wa-al-maḥsūs. This is one reason to assume that the Plotinian borrowings were added during the ‘main’ stage of the adaptation process that shaped Kitāb al-Ḥiss wa-al-maḥsūs, and that most of the alterations to the Plotinian passages can be laid at the door of the principal adaptor – even though some remarks are likely to have been penned by later glossators or subsequent adaptors.Footnote 61 Hence the method by which the Plotinian borrowings are made to fit the adaptor's theory does not need to surprise us. Plotinus' text is treated just like Aristotle's: it is excerpted, altered and re-interpreted to support the theory of the three spiritual faculties. However, there is of course a crucial difference: given that the whole project of Kitāb al-Ḥiss wa-al-maḥsūs starts out with Aristotle's Parva Naturalia, it is no wonder that even an adaptor intent on promoting a quite different theory nevertheless could not ignore this text entirely, but had to make use of it one way or another. The same would not, prima facie, apply to the Plotinian borrowings. Why choose to integrate yet another text that again needed a good deal of ‘editorial’ work on the part of the adaptor in order to offer support to his theory? Were the points gained by including the Plotinian material really worth the adaptor's efforts in adjusting it?
Here we are entering the field of speculation. It will not be possible to say with certainty why, how and under what circumstances the Plotinian fragments entered Kitāb al-Ḥiss wa-al-maḥsūs. Nevertheless it is worth exploring various possible hypotheses, not least to gain a sense of how our questions concerning Kitāb al-Ḥiss wa-al-maḥsūs are interwoven with those relating to the Arabic Plotinus.
In principle, there could have been two possible motivations for the adaptor of Kitāb al-Ḥiss wa-al-maḥsūs to include an additional text: either because it offered desirable theoretical points, or else because the adaptor believed that the text in question was also authored by Aristotle. If it is questionable whether the Plotinian borrowings were in themselves useful enough for the adaptor to justify their inclusion, we may want to ask whether he may have known them as Aristotle's work. In this case it may have been quite natural not to ignore a second text on sense-perception and memory supposedly written by the Stagirite.
This again leads us to two possibilities. The source text may have been ascribed to Aristotle independently of the production of Kitāb al-Ḥiss wa-al-maḥsūs; or else the adaptor could have been led to assume that it was by Aristotle because he found the Arabic version of Ennead IV.6 in some way associated with the translation of the Parva Naturalia underlying Kitāb al-Ḥiss wa-al-maḥsūs.
Given that ThA is transmitted under the name of Aristotle, it is an obvious question to ask whether the presence of Plotinian fragments in ‘Aristotle's’ Kitāb al-Ḥiss wa-al-maḥsūs is in any way related to that attribution. However, to assume that an earlier version of ThA had included parts of Ennead IV.6 which then dropped out in the course of transmission seems otiose given the lack of textual evidence for such a version. Furthermore, if we want to accept the suggestion that ThA is not an entirely fortuitous assembly of parts of *AP but reflects a thematic grouping of passages, as argued for by Peter Adamson and Cristina D'Ancona, it is somewhat doubtful whether the rather ‘prosaic’ aspects of sense-perception and memory discussed in Ennead IV.6 would have fitted the generally somewhat loftier metaphysical outlook of ThA.
It is perhaps more pertinent to ask whether the fact that we find fragments of *AP in Kitāb al-Ḥiss wa-al-maḥsūs can tell us anything about the point at which the attribution to Aristotle was consolidated within the transmission. Fritz Zimmermann has argued that Aristotle's name might have been mentioned in the title of *AP itself (or even in that of the ‘metaphysics file’), though the actual attribution would have been the result of a misunderstanding on the part of the ‘editor’ of ThA.Footnote 62 Could the fragments in Kitāb al-Ḥiss wa-al-maḥsūs be a similar case? When excerpting *AP, could the adaptor of Kitāb al-Ḥiss wa-al-maḥsūs have misread its title so as to attribute the text to Aristotle? Somehow this does not strike one as being too likely. The adaptor would not have faced the same difficult textual situation that Zimmermann envisages for the scribe who compiled ThA after the disintegration of *AP. Moreover, working with the translation of Aristotle's Parva Naturalia he should have been aware of the crucial theoretical differences between the two texts. And if he was indeed linked to the Kindī circle we should furthermore expect him to have had ‘insider knowledge’ about any Kindī-circle texts he was using.
Do the new fragments, then, indicate that the attribution to Aristotle had in fact taken place before any disintegration of *AP and / or the compilation of ThA? Did copies of *AP exist which stated that Aristotle was its author? Even if this were the case, it seems unlikely to have occurred within the Kindī circle itself.Footnote 63 Hence this scenario would cast doubts on our assumption that Kitāb al-Ḥiss wa-al-maḥsūs was produced in the Kindī circle in temporal proximity to *AP, suggesting a later date for the adaptation – or at least for the inclusion of the Plotinian fragments. Due to the general uncertainty surrounding Kitāb al-Ḥiss wa-al-maḥsūs this cannot be categorically excluded; but it does not appear likely.Footnote 64 And there are further questions to be asked: assuming that the adaptor did take the Plotinian fragments from *AP, why would he only have chosen passages from Ennead IV.6.1 and IV.6.3,Footnote 65 if he thought that text to be by Aristotle? Why did he not use other parts of *AP, for instance, to bolster the second maqāla of Kitāb al-Ḥiss wa-al-maḥsūs? Such uncertainties weigh all the heavier given that Kitāb al-Ḥiss wa-al-maḥsūs assembles a host of other material under Aristotle's name as well. Hence it does not seem entirely necessary to explain the presence of the Plotinian fragments by the misattribution of *AP to Aristotle: there are other possible scenarios that would explain their inclusion in the text.
One of these scenarios is in some ways analogous to Fritz Zimmermann's idea of a ‘metaphysics file’: the adaptor may have found the Arabic translation of Ennead IV.6 grouped together with that of De sensu, and perhaps with other thematically relevant texts which subsequently found their way into Kitāb al-Ḥiss wa-al-maḥsūs. There might have been, in other words, a ‘file’ similar to the stipulated Kindī-circle ‘metaphysics file’, which assembled different texts on sense-perception, or, if we consider Kitāb al-Ḥiss wa-al-maḥsūs as a whole, perhaps on the ‘faculties’ or powers of the soul.Footnote 66 Due to some accident of transmission the adaptor may have believed it to be part of the Aristotelian text, or perhaps a further text by Aristotle on the same topic. Leaving aside the fact that this would still require us to ask further where the compiler of that file had found his texts originally, the option is not unproblematic for other reasons, too. One may again wonder whether the adaptor of Kitāb al-Ḥiss wa-al-maḥsūs would not have recognised the crucial differences between the two texts, especially if we take into account that he used other sources as well: would he have assumed them all to have been produced by Aristotle?
The alternative assumption is that the adaptor of Kitāb al-Ḥiss wa-al-maḥsūs knew that the passages in question were not by Aristotle. Given his prevailing practice of using several different sources which he blends together to form a single new text, this is perhaps more likely. In that case there are again two options, running parallel, as it were, to those already mentioned. The adaptor may have found Ennead IV.6 as a pre-selected excerpt together with the translation of the Parva Naturalia and other texts in a pre-arranged collection of texts; or he may himself have actively searched for related texts in additional sources, in this case probably *AP.
Above, I have suggested that Ennead IV.6 may not have suited the adaptor's theory well enough prima facie to be considered a ‘must have’ for his work (whatever valuable points he manages to extract from it in the end). There might, however, have been other motivations to use the Plotinian text, even for an adaptor who knew it did not have anything to do with Aristotle. For example, he might have wanted to combine several ‘Greek’ philosophical texts on one and the same topic, to provide an overview of some sort. With respect to the adaptor of Kitāb al-Ḥiss wa-al-maḥsūs, though, this cannot be the case: what he produces is not an overview or compendium. He blends the texts together, and moreover does not provide any names apart from Aristotle's. If, however, we assume that the adaptor worked from a ‘file’ of assembled texts on the psychic faculties, such texts would automatically have recommended themselves for use, by virtue of their presence in such a file. But we do not have to assume that the adaptor worked only with texts that had been assembled beforehand. He may well have consciously and intentionally been in the business of creating a ‘syncretistic’ account of what ‘the Greeks’ had to say on the topics represented by Aristotle's Parva Naturalia, and so may have done the collecting of suitable texts himself. In both cases a Plotinian treatise called On sense-perception and memory would certainly have recommended itself for inclusion. Another text for which this would hold true is Ennead IV.5 (On Sight), which is transmitted in Arabic among the fragments of GS.Footnote 67 The extant part of Maqāla I does not contain any parallels to this text, as far as I can make out. However, one or two lines in the first maqāla of Ibn Rushd's Talkhīṣ suggest that the translation of Ennead IV.5 may have been used by the adaptor in other passages of Maqāla I.Footnote 68
Given the overall character of Kitāb al-Ḥiss wa-al-maḥsūs this is perhaps the most plausible account of what spurred the adaptor: an attempt to create a ‘digest’, starting out from the most prominent work, i.e. Aristotle's Parva Naturalia, and integrating sources from the later tradition – in a manner which, in the end, allowed him to promote a theory that appealed to him. This way of proceeding would certainly explain his ‘picking and choosing’, which seems to be the most likely cause, for instance, of the absence of Ennead IV.6.2 from the text. Whether the adaptor was relying on a collection of texts compiled by someone else, or whether he selected his sources entirely by himself, and whether he had access to *AP as a whole, cannot be determined at this point.
The adaptor's very active role in the construction of Kitāb al-Ḥiss wa-al-maḥsūs means that we cannot identify accurately the sources of his Plotinian borrowings. Therefore we cannot say with any certainty what exactly his source text looked like, how close a translation or paraphrase of the Greek it was, and how much of Ennead IV.6 (and IV.5?) it covered. Given the level of interference that the text has suffered (probably from our adaptor, but possibly not just from him), it is impossible to determine exactly what happened to it at which stage.Footnote 69
However, the presence of new fragments from a translation of Ennead IV.6 in Kitāb al-Ḥiss wa-al-maḥsūs does, I believe, give some measure of support to the assumption that *AP was a translation-cum-adaptation of Enneads IV-VI which followed the treatises in chronological order, and which contained more of the text than is extant in AP; and that this text was used by people of the Kindī circle (and beyond) as a kind of philosophical ‘quarry’: single treatises or even smaller parts of this translation-cum-adaptation could be copied out, ‘filed’ and used according to topic. Whether this text as a whole would have been the one included in the Kindī-circle ‘metaphysics file’ envisaged by Fritz Zimmermann is a question that is still worth discussing.
As with so many things to do with Kitāb al-Ḥiss wa-al-maḥsūs the actual evidence it offers is rather tentative and inconclusive. While the fragments from Ennead IV.6 do present a new piece of the Arabic Plotinus puzzle, this new piece does not yet give us a much clearer view of what the overall, complete picture should look like in the end. In fact, the most palpable result of its appearance might well be the realisation that this puzzle may be even larger and more complicated than we have assumed so far. Like his colleague Proclus, Plotinus Arabus has still some way to go.Footnote 70
