¹ ‘Die arabische Version einer unbekannten Schrift des Alexander von Aphrodisias über die Differentia specifica’, Nachrichten der Akademie der Wissenschaften in Göttingen (1964): 85–148 (93–100); on the exclusion of Proclus (sections 15–17), see p. 95, under no. 8.Google Scholar

² The most comprehensive among recent lists are found in: Sharples, R.W., ‘Alexander of Aphrodisias: Scholasticism and Innovation’, Aufstieg und Niedergang der römischen Welt, 36.2 (Berlin and New York, 1987): 1176–243 (D23 and D27 are set aside at 1194, where they should be joined by D29 from 1192);Google Scholarand Goulet, R. and Aouad, M., art. ‘Alexandros d'Aphrodisias’, in Goulet, R. (ed.), Dictionnaire des philosophes antiques (= DPhA) (Paris, 1989), vol. I, pp. 125–39Google Scholar (D23 is listed among the spuria at 139 (62), where it should be joined by D29 from 137 (36)). For ease of reference, here is a checklist of the pieces of Proclus sive Alexander to be considered in this article.

P1–3: based on sections 1–3 of the Elements; text included in D27; separately transmitted as a piece of Alexander; no mention of translator; see under J below.

P15–17: based on sections 15–17 of the Elements; text included in D27; separately transmitted as a piece of Alexander translated by Dimashqi; see under I below.

D23: based on section 77 of the Elements; not included in D27; separately transmitted as a piece of Alexander; no mention of translator; see under K below.

*P77: lost original version of D23 (the asterisk serves to mark its conjectural nature).

D27: includes twenty sections of the Elements; obscure heading; no mention of translator; see beginning of section II below for the contents, and beginning of section IV for the heading.

D29: the piece here published; based on section 98 of the Elements; not included in D27; separately transmitted as a piece of Alexander; no mention of translator.

*P98: lost original version of D29 (inferred).

³ Endress, G., Proclus Arabus: Zwanzig Abschnitte aus der Institutio Theologica in arabischer Übersetzung (Beirut and Wiesbaden, 1973): contains an edition, translation, and extensive study of the Proclus of D27.Google Scholar

⁴ Dodds, E.R. (ed., tr.), Proclus, The Elements of Theology, 2nd edn (Oxford, 1963). To suit my comparative purpose, Dodds's English version has been adapted too drastically for it still to be called his.Google Scholar

⁵ In the Kindi-circle Alexander, Aristotle's sobriquet of ‘the Philosopher’ appears, now as al-faylasūf, now as al-ḥakīm. To mark the difference, I translate the latter as ‘the Sage’.

⁶ ‘X has come to be (ṣāra) P’ often means ‘One has come to think (claim, conclude) that X is P’. For further examples of the usage in Kindi-circle translations see Endress, Proclus Arabus, pp. 176–8 at (e).Google Scholar

⁷ I owe the conjecture to the kindness of Maroun Aouad, whose thorough reading of a complete draft of this article has led to several corrections and improvements.

⁸ As Aouad points out, the syntax of the transmitted text is sound, with the apodosis beginning at (n). If that is to be retained, the sentence will read: ‘(l) If so, … all things being posterior to it and secondary (thāniyatan), with the First Cause never too tight…, (n) then it has become clear and sound …’.

⁹ Manuscript C was described by Rosenthal, F., ‘From Arabic books and manuscripts V: A one-volume library of Arabic philosophical and scientific texts in Istanbul’, Journal of the American Oriental Society, 75 (1955): 14–23; on the scribe, see pp. 14–16.CrossRefGoogle Scholar

¹⁰ It so appeared to Rosenthal, ‘One-volume library’, p. 17n11. But it must be said that the scribe of C himself uses the ḥamdala of Sura 27:59 at least once (my microfilm covers only about a third of the codex), towards the end of a dated note to a text transcribed two or three years earlier: katabahu Muḥammad b. Ḥasan b. ‘Alī … at-Nihmī … al-Ṣa'dī … 'afā Atlāhu 'anhum wa at-muslimīn katabahu shahr Ṣafar sanat khams wa-thamānīn wa-thamānimi'a hijriya. at-ḥamdu li-ilāhi wa-satāmun 'alā 'ibādihi iladhīna ṣtafā wa-ḥasbunā llāhu wa-ni'ma i-wakīi. … (fol. 173^v, top margin). The same ḥmdala occurs in the subscription to Pseudo-Faiabi's R. fī at- 'ilm al-ilāhī at fol. 140^r26 (does that mean that the text derives from another manuscript by ‘Muṣatafā’?). It further occurs at the beginning of at least three other pieces included in our codex.Google Scholar Rosenthal (ibid.) mentions a work by Suhrawardī al-Maqtūl. The others are a K. al-hay'a by Qāsim b. Mattraf (sic) al-Qattān al-Andalusī al-Qurtubī (fol. 315^r2) and a K. jawāmī' 'ilm al-nujūm by A. b. M. b. Kathīr al-Farghānī (fol. 383^v32).

¹¹ Endress, Proclus Arabus, pp. 33–51. It is there that readers will find detailed information on manuscripts here simply referred to by means of Endress's sigla.Google Scholar

¹² Cf. Endress, Proclus Arabus, pp. 62 ff: ‘Die Sprache der Proklosversion und ihre Stellung in der griechisch-arabischen Übersetzungsliteratur’. For the istibāna, see pp. 180–2 at (a). In summing up his findings, Endress concludes that the Arabic pieces of Proclus's Elements were translated, not by Dimashqi (190), but by a member of the circle of Kindi (192); within the larger body of Kindi-circle translations, Plotinus, Liber de causis, Alexander, the Proclus of D27, and Aristotle's De caelo form a distinct group (186), within which the last three are closer to each other than to the rest (188f); as the translation of the De caelo is attributed to Ibn al-Bitrīq, the D27a: ‘On the First Cause’ and similar titles (= Proclus, Elements, sections Kindi-circle Alexander, including the Proclus of D27, may also be his work (191f). By extension, the same applies to D23 and D29.Google Scholar

¹³ The contents of D27 were sorted out by Ess, J. van, ‘Über einige neue Fragmente des Alexander von Aphrodisias und des Proklos in arabischer Übersetzung’, Der Islam, 42 (1966):148–68Google Scholar, who abolished the mark ‘D27’ (p. 152); catalogued separately the contents of what I have dubbed D27a and D27b (159–63); correctly identified D27d with D7 (150f, with list of variants) and, incorrectly, D27g with D16 (151); and added D27c, e, fon to Dietrich's list as nos 32–34 (153). Of the last five items, only D27d has hitherto been published (in the shape of D7). Cf. Sharpies, ‘Alexander’, p. 1194 (on D27) and 1190–2 under quaest. 1.8, 2.3, 2.11, 2.19, and D16; DPhA i.l32f, at 19(h) (= D27), (f), (k), (1), (m); and 137 (35). The last item in each case, D27g, should be distinguished from D16 in MS G (Escurial 798), which is not in fact another version of the same text.

¹⁴ Bruns, I. (ed.), Alexandri Aphrodisiensis praeter commentaria scripta minora, 2 pts (Berlin, 1887–1892)Google Scholar; Sharpies, R.W. (tr.), Alexander of Aphrodisias, Quaestiones 1.1–2.15 (London, 1992)Google Scholar; Badawi, A., Arisṭū 'ind al-'Arab (Cairo, 1947)Google Scholar. Here and under C-E, H below, I have freely adapted Sharples's English to my needs.

¹⁵ fim¯ C; limā Badawi.

¹⁶ See note 6 above.

¹⁷ ṣūra C; ṣuwar Badawi.

¹⁸ fi C, Badawi: read min. Min and fi are often mistaken for one another; an additional explanation for the present mistake is that the clause in which it occurs looks as though it should be related to ‘the forms’ rather than ‘the bodies’.

¹⁹ hiya anniyyatun min C; fa-hiya ityān Badawi: read fa-hiya ithnatān.

²⁰ al-ṣūra C, Badawi: read al-ṣuwar.

²¹ fa-in C; wa-kāna Badawi: read wa-in kāna.

²² mufāriq C; mufāriqan Badawi: read muqārinan.

²³ ṣūratihi C; ṣūratihiā Badawi.

²⁴ wa-aqūl C; aqūl Badawi.

²⁵ idhā mā C; idhā Badawi.

²⁶ wa-kānat 'illat 'adam wujūd ḍiddihā fihā ayḍtan C; wa-kānat 'illat wujūd ḍiddihā fihi ayḍan Badawi. The example of heat has previously been used in the authentic text (56.21ff Bruns = 287.11ff Badawi), where the Arabic for ‘e.g.’ is shibha. The present example is introduced by ka-qawlinā instead, which may be a sign that it was added after translation.

²⁷ wa al-'illa ḍidduhu wa-Ǡidd ḍiddihi Badawi: read wa al-'illa hiya wujūd ḍiddihi?

²⁸ The title is missing. The proposition that follows is not strictly a quotation; the original writer may in fact have been thinking of Plato rather than Aristotle. For the classical background see Sharples's note 151 to his translation.

²⁹ al-ladhdha Badawi: read kull ladhdha?

³⁰ The locus Aristotelicus in question is not in fact in the De gen. et corr.; see Sharples's note 387.

³¹ Ruland, H.-J. (ed., tr.), ‘Die arabische Übersetzung der Schrift des Alexander von Aphrodisias über das Wachstum (Quaestio I 5)’, Nachrichten der Akademie der Wissenschaften in Göttingen (1980): 51–74;Google ScholarBadawi, A., Commentaires sur Aristote perdus en grec et autres épîtres (Beirut, 1971).Google Scholar

³² For once the reference to the De gen. et corr. (i.5.321^b11ff) is apt; see Ruland's introduction, p. 55, and Sharples's note 85.Google Scholar

³³ fa-thumma qāla Badawi; fa-qāla Ruland (the fqym preceding in G and D, omitted by Ruland, suggests fa-kayfa as an alternative to fa-limā).

³⁴ The thesis is extrapolated from Aristotle's remarks that ‘motion may be thought of as a sort of actuality (ένέργεια), but an incomplete one’ (Phys. iii.2.201^b31f), and ‘motion is the incomplete entelechy (έντελέχεια) of a movable thing’ (Phys. viii.5.257^b8f), which imply the distinction mentioned at (d) below.

³⁵ ffym Badawi: read fa-kayfa?

³⁶ Phys. viii.5.257^b7f: When the movable moves, its ‘potentiality (τò δυνάμει) passes (παδίζει) into entelechy’.

³⁷ Phys. iii.1.201^a10f: ‘The entelechy of what is potentially, as such, is motion’.

³⁸ Dr Sharples points out to me that as the titles under which the original quaestiones are transmitted may well be secondary additions, it is of interest to know that the Arabic translations show them to have been in place by the early ninth century.

³⁹ The text of the note is reproduced from manuscript Z (fol. 115^v) by Badawi, Aristū 294, and Endress, Proclus Arabus 60. As it would take too long to discuss the ambiguities with which it is riddled, here is a translation: ‘The preceding [i.e. Kindicircle] pieces, which are attributed <by their inscriptions> to Alexander of Aphrodisias <with no mention of the translator> were all translated by Abū‘Uthmān Sa'īd al-Dimashqī, because <unlike the first exemplar, from which I transcribed the earlier part of my collection,> this second exemplar (hādhihi al-nuskha al-manqūla [understand minhā] al-thāniya) <from which I transcribed the last [i.e. Kindi-circle] pieces> is in Dimashqi's own hand’.

⁴⁰ The beginning of the reply is missing; cf. van Ess's German, J. translation, Archiv für Geschichte der Philosophie, 55 (1973): 188–90.Google Scholar

⁴¹ The inscription was evidently supplied by the transmitter who first copied, evidently from a defective exemplar, what is in fact a sequence of three pieces. Ours is the second, beginning in mid-sentence immediately after the unmarked end of the first. Its identity was established by Gätje, H., Zeitschrift der Deutschen Morgenländischen Gesellschaft, 116 (1966): 262. The identity of the parallel version in D8 remained unnoticed because it disagrees with the transmitted Greek at the beginning. See further Sharples, ‘Alexander’ 1191, at 1.21, where D2 should be joined by D8 from 1192; and DPhA i.133 (19p) and 137 (32), which should be linked.Google Scholar

⁴² The thesis seems to have been extrapolated from Phys. iii.1–3. Aristotle says that ‘the mover will always confer <on the moved> a form, <the form of a> this or such or this much’ (2.202^a9f). And he explains that since the bronze qua bronze is not the same as the bronze qua potential statue, i.e. the material without the form, ‘just as the colour is not the same as the visible, motion is clearly the entelechy of the possible qua possible’ (1.201^a29-^b5). That reference to colour and visibility may explain why the discussion turns to light and colour at (f) below. We understand that while the bronze is turning into a statue, its eventual form is no longer purely potential and not yet wholly actual. The movement from the one to the other is, so to speak, an ‘incomplete actuality’ (2.201^b31f). When the statue is finished, the new form has been accomplished and so has the movement that led to it. Forms can therefore be said to be the entelechy — in the sense of ‘accomplishment’ — of motion, but only if they are subject to change, and to gradual change at that (cf. F(d) above).

⁴³ The beginning does not tie up with the end, as Sharples observes in his note 220. The divergence of the Arabic makes it seem all the more likely that the Greek here is spurious.

⁴⁴ al-mutaḥarrik: read al-taḥarruk.

⁴⁵ thābita: read thālitha.

⁴⁶ A distinction between complete and incomplete actuality is implied at Phys. iii.2.201^b31f (see note 34 above). The substitution, here and later, of ‘motion’ for ‘actuality’ makes no sense in Aristotelian terms. It has been carried out too consistently to be explained as a mere scribal error. D9 preserves the original distinction at F(d) above.

⁴⁷ The same question receives a different answer at F(d) above.

⁴⁸ For once, Dimashqi's version agrees with D8 against the transmitted text of the Greek, which does not necessarily mean that the Greek version he had in front of him lacked the missing words. If it is true that he went to the trouble of making a copy of D8, he was presumably looking at it while re-translating the Greek. He may have simply overlooked the words skipped by D8. Note that the word ⋯ντελέχεια, omitted by him, is deferred in D8 to a gloss appended to the sentence.

⁴⁹ On the transliteration of Greek khi by means of Arabic shīn see Zimmermann, F., ‘The origins of the so-called Theology of Aristotle’, in Kraye, J. et al. (edd.), Pseudo-Aristotle in the Middle Ages, Warburg Institute Surveys and Texts, 11 (London, 1986), pp. 110–240, p. 114.Google Scholar

⁵⁰ harab: read dhahāb, translating the βαδíζειν of Phys. viii.5.257^b8? The parallel sentence at F(d) above has intiqāl.

⁵¹ See note 36 above.

⁵² The piece was first published from MS Z in Badawi, Arisḍū, 291f. The words in square brackets are not in Z, which refers to Dimashqi in a separate note (see note 39 above). The ascriptions to Alexander and Dimashqi are peculiar to manuscripts transmitting sections 15–17 as a separate unit. MS C, where the same text constitutes but three of the 26 sections of D27, adds a reference to Dimashqi in the margin of section 15, no doubt from a manuscript of the other family. MS R, where section 15 occurs as the first of four, adds ‘from a text by Aristotle’ to its title.

⁵³ The same type of inscription is further shared by D12 (= Quaestio 1.2) in MS G (ed., tr. Gätje, H., Nachrichten der Akademie der Wissenschaften in Göttingen (1967): 341–62)Google Scholar and D22 (= Quaestio 1.16) in MS Tk (ed. Badawi, , Commentaires 47–50). The inscription of D12 adds the words ‘Translated by (naql) Abū ‘Uthmān al-Dimashqī’, which could mean that D12 too was once included in the copy produced by Dimashqi according to MS Z (see note 39 above).Google Scholar

⁵⁴ The piece was first published from MS Tk in Badawi, Commentaires 24–6. The words in brackets occur only in Tk. The rest of the inscription closely resembles that of D27 in MS C, where the same text constitutes the first three of 26 sections.

⁵⁵ No version of section 77 exists in D27. According to Endress's table at Proclus Arabus 43, D23 occurs also in MSS G, L, H, M. For an additional copy in MS Lucknow Naziriya 841, see Daiber, H., ‘New manuscript findings from Indian Libraries’, Manuscripts of the Middle East, 1 (1986): 26–48, p. 37 (cf. DPhA i. 139 (62)).Google Scholar

⁵⁶ Cf. H(c) above; and contrast the Kindi-circle use of sharḥ in the sense of ‘order’: Endress, Proclus Arabus, pp. 131–3.

⁵⁷ Endress, Proclus Arabus, Arabic p. 3. 8f. For another attestation in the Cairo Genizah see Khan, G.A., ‘The Arabic fragments in the Cambridge Genizah collections’, Manuscripts of the Middle East, 1 (1986): 55–60, p. 60n23 (cf. DPhA i.133(h)): ‘T-S Ar.40.143 (Excerpts from Pseudo-Aristotle's Theology by Alexander of Aphrodisias)’.Google Scholar

⁵⁸ Dieterici, F. (ed.), Die sogenannte Theologie des Aristoteles (Leipzig, 1882), p. 1.3f;Google ScholarBadawi, A. (ed.), Aflūtīn 'ind al-'Arab (Cairo, 1955), p. 3.4–6. For further versions of the title see the table in Zimmermann, ‘Theology of Aristotle’, pp. 189f.Google Scholar

⁵⁹ For another attestation see Taylor, R.C., ‘Two manuscripts containing … the ‘Liber de causis’ and an anonymous Neoplatonic treatise on motion’, Mélanges de l'Institut Dominicain d'Etudes Orientales du Caire, 15 (1982): 251–64, p. 258: kitābuhu alladhī yud'ā thūlūjiyā wa-huwa kitāb al-rubūbiyya.Google Scholar

⁶⁰ Note that D8 uses alladhī yud'ā in connexion with titles, elsewhere sammā (see H(e) above). The dā'ī style is not always confined to titles of Aristotle: it occurs in references to Plato's Timaeus in the Kindi-circle version of De Caelo 280^a30, 293^b32, 300^a1, ^b17 (ed. Badawi, A., Cairo, 1961) and twice in the Kindi-circle PlotinusGoogle Scholar(see Zimmermann, , ‘Theology of Aristotle’, p. 149).Google Scholar

⁶¹ Zimmermann, ‘Theology of Aristotle’, p. 189.Google Scholar

⁶² Ibid., pp. 120–5 and passim.

⁶³ al-mabādi' al-kull C: read al-qawl (= λóγος?) fi mabādi' al-kull with Z (ed. Badawi, Aristū 253.3)?Google Scholar

⁶⁴ Badawi, Aristū 277.8. The subscription continues with the words: ‘and <translated> from Greek into Syriac by Abū Zayd Hunayn b. Isḥāq.’ That may explain the name of Hunayn in manuscript C, as Charles Genequand observes in his forthcoming critical edition, of which he has had the kindness to let me see a draft.+from+Greek+into+Syriac+by+Abū+Zayd+Hunayn+b.+Isḥāq.’+That+may+explain+the+name+of+Hunayn+in+manuscript+C,+as+Charles+Genequand+observes+in+his+forthcoming+critical+edition,+of+which+he+has+had+the+kindness+to+let+me+see+a+draft.>Google Scholar

⁶⁵ It is no longer possible, if ever it was, to make Dimashqi the original subject of the verb istakhraja, as I did in the past (‘Theology of Aristotle’, pp. 130, 185). The fact that D27 includes an earlier version (D27d) of a piece apparently transmitted in a later version (D7) by Dimashqi (see note 39 above) shows that D27 does not derive from Dimashqi's collection (l below).

⁶⁶ According to its inscription in the two manuscripts described by Taylor (‘Two manuscripts’, pp. 256, 259), the Liber de causis was sometimes attributed to Proclus. The same manuscripts include an unpublished compilation (pseudonymous rather than ‘anonymous’, as the title of Taylor's article has it) under the name of Aristotle, which draws, inter alia, on sections of the Elements not included in either D27 or the Liber de causis (see note 92 below). Sections 15, 17, 21 and 54 in MS R are likewise attributed to Aristotle (see note 52 above). The former cannot, the latter need not, be based on D27.

⁶⁷ A parallel (ikhrāj Ḥunayn b. Isḥāq) in a 15th-cent.(?) manuscript of a translation of Galen was explained as a Syriacism by Kraus, P. and Walzer, R., Galeni compendium Timaei Platonis (London, 1951), pp. 18, 21, 30. Rosenthal (“One-volume library’, p. 17n11a) notes it as a potential parallel to the equally odd use of istikhrāj discussed above.Google Scholar

⁶⁸ See note 39 above.

⁶⁹ See Endress's stemma in Proclus Arabus, p. 44.Google Scholar

⁷⁰ That Proclus was not completely forgotten in the Arabic tradition of the Liber de causis transpired some thirty years ago from the two manuscripts described more recently by Taylor (see note 66 above). One cannot help being impressed by the good sense of Otto Bardenhewer's remark, in his editio princeps of the Arabic Liber de causis (Freiburg i.Br., 1882, p. 51), that the Arabic version of the Elements underlying that compilation ‘doch wohl den Namen Proklos, und nicht Aristoteles, an der Stirne getragen haben wird’. D29 sheds little light on the question how the Liber de causis relates to our Proclus sive Alexander, except that it tends to add to one's dissatisfaction with Endress's conclusion (Proclus Arabus, pp. 55, 187f, 240) that two distinct adaptations of the Elements were separately translated by different members of the same group (see, in the volume cited in note 49 above, Taylor, R.C., ‘The Kalam fi mahd al-khair (Liber de causis) in the Islamic philosophical milieu’, pp.39f, 51f;Google Scholarand Zimmermann, ‘Theology of Aristotle’, pp. 184–8). Those adaptations would presumably have to be assigned to a pre-Arabic period, as Endress indeed inclined to do (Proclus Arabus, pp. 235ff) in accordance with a wide-spread tendency to take a similar view of the origins of the so-called ’Theology of Aristotle‘. In my study of the latter, I argued at length against assigning pre-Arabic origins to the peculiarities of the Kindi-circle Plotinus; I shall here argue more briefly against assigning pre-Arabic origins to the peculiarities of the Kindi-circle Alexander. Bardenhewer probably overrated the extent to which the ‘author’ of the Liber de causis (whom he located in 9th-cent. Iraq: p. 53) departed from the underlying (complete?) translation of the Elements. But we seem to be working our way back to his conviction (p. 38) that there was no pre-Arabic model other than the transmitted text of Proclus (for a lucid account of recent advancesGoogle Scholarsee Rowson, E.K., ‘An unpublished work by al-'Āmirī and the date of the Arabic De causis’, Journal of the American Oriental Society, 104 (1984): 193–9). That will almost inevitably lead us to suppose that the pieces of Proclus transmitted within and without the Liber de causis must be fragments of a single (though not necessarily complete) translation of the Elements — which will indeed have borne the name of Proclus.CrossRefGoogle Scholar

⁷¹ At De caelo 306^b7, al-shakl alladhī yud'ā birāmis renders φυραμíσ. D8, on a similar occasion, uses sammā (cf. note 60 above).

⁷² Cf. Endress, G., Die arabischen Übersetzungen von Aristoteles' Schrift ‘De Caelo’, Ph.D. thesis (Frankfurt a.M., 1966), p. 244. I owe example and reference to Hillary Wiesner.Google Scholar

⁷³ Ruland, H.-J. (ed., tr.), Die arabischen Fassungen von zwei Schriften des Alexander von Aphrodisias, Ph. D. thesis (Saarbrücken, 1976), p. 105.6–8.Google Scholar

⁷⁴ One each in D5 (see C(b) above), D14 (see next note), D27e, D27f, D29 (see A(b)); two in D9 (see F(b), (d)); three in D8 (see H(b), (d), (e)); four in D15 (see note 81 below).

⁷⁵ Cf. έκ δευτέρο … in the title of Quaestio 3.3 (ed. Bruns ii.82.21) with Arisṭū … fi (sic leg.) kitābihi alladhī yud'ā kitāb al-nafs in the opening sentence of D14 (ed. Ruland, H.-J., ‘Die arabische Übersetzung der Schrift des Alexander von Aphrodisias über die Sinneswahrnehmung’, Nachrichten der Akademie der Wissenschaften in Göttingen (1978), p. 168).Google Scholar

⁷⁶ It is of course to be expected that the Greek manuscript(s) of Alexander available to an early ninth-century Arab translator (or to an even earlier Syriac translator) were somewhat different from, and sometimes better than, those available to modern European editors. We saw that the beginning of D8 might well be closer to the original than the transmitted Greek (see note 43 above). It is all the less likely that the underlying Greek should have been riddled with spurious additions — additions, moreover, from which later re-translations are free (see note 80 below; and contrast D2 with D8 under H above).

⁷⁷ The phrase is borrowed, with the kind help of Patricia Crone, from the late SirGibb, Hamilton (‘The social significance of the Shuubiya’, in his Studies on the Civilization of Islam (London, 1962), p. 72).Google Scholar

⁷⁸ See Sharples's note 351 to his translation of Quaestio 2.11.Google Scholar

⁷⁹ D6 (two); D7; D15; the exception is in D19; see notes 30, 32, 78 above, and 80 below.

⁸⁰ Contrast the lower text at 81.7–10 with the upper text at 81.9f in Ruland's edn (note 73 above). I concur with Silvia Fazzo in failing to be convinced by Ruland's identification of the ‘quotation’ (p. 82n6).

⁸¹ Same edn, lower text: fi kitābihi alladhī yud'ā kitāb al-tadbīr (33.3, sim. 51.5); fi kitāb al-kawn wa al-fasād (81.7); fi kitāb al-samā' (89.7); fi kitābihi alladhī yud'ā kitāb mā ba'd al-ṭabī'a (93.1f); fi al-kitāb alladhī yud'ā asṭurlūjiyā (105.7).

⁸² Beginning of D14 (note 75 above); beginning of D17 (note 84 below); title and beginning of D20 (= Quaestio 1.24; ed. Badawi, Commentaires 44–6); beginning of D22 (note 53 above).Google Scholar

⁸³ D8 (see H(d) above); D14 (note 75 above); D15 (end of note 81 above).

⁸⁴ Contrast, e.g., the reference to the De anima in D14 (note 75 above) with another one at the beginning of D17: fi kitābihi fi al-nafs (ed. Ruland, H.-J., ‘Zwei arabische Fassungen der Abhandlung des Alexander von Aphrodisias über die Universalia (Quaestio I 11a)’, Nachrichten der Akademie der Wissenschaften in Göttingen (1979): 243–74, p. 254).Google Scholar

⁸⁵ Contrast, e.g., the references to the Physics at F(b) and (d) above: sam' al-kiyān is unusual (though not unique) and probably archaic.

⁸⁶ A likely case is the reference to the Physics at H(e) above, which occurs in what looks like a re-arrangement of the underlying Greek sentence by the translator himself.

⁸⁷ See note 12 above.

⁸⁸ Endress, in speaking of the ‘Zusammenarbeit einer Gruppe arabischer Übersetzer’ (Proclus Arabus, p. 189), seems to allow for the possibility that some texts were translated or revised collectively by a group of people sitting, so to speak, round a table.

⁸⁹ The istibāna need not always have been added in the first draft. In fact, it is almost easier to imagine that it was generalised, for the sake of uniformity, at a stage when the corpus was revised as a whole. (It may be significant that some sections of the Liber de causis lack an istibāna: see end of note 92 below). But if the reviser was not the translator himself, we shall have to add yet another second hand to our list, because all those listed so far are down for sporadic changes making for diversity rather than uniformity.

⁹⁰ Cf. note 70 above.

⁹¹ If you accept my argument at ‘Theology of Aristotle’, pp. 113ff.

⁹² See note 57 above. And the last text preserved in the two manuscripts described by Taylor includes lines (reproduced in his ‘Two manuscripts’, p. 258) apparently drawn from the translation of sections 15 and 17 that survives in D27 and P15–17. In the same text, the late Professor Pines additionally spotted four extracts from sections 20 (two), 41, and 66, which are not included in the Kindi-circle Alexander or the Liber de Causis (‘Hitherto unknown Arabic extracts from Proclus’ Stoicheiōsis theologikē and Stoicheiōsis physikē’, first published in The Collected Works of Shlomo Pines (Jerusalem and Leiden, 1986), vol. II, pp. 287–93). At the beginning of his account (p. 288), Pines enumerates section 48 among the sources; but there is no extract from that section among the passages reproduced in the sequel. Clarification is to be expected from a study promised by Professor Taylor (‘Two manuscripts’, p. 264n31). I am inclined to discount the possibility (left open by Pines, p. 293) that the Arabic text in question will prove on publication to be a translation of a lost Greek compilation. More probably, it will show that β included more sections of the Elements than survive in D23, D27, and D29. If it does prove to have had section 48 among its sources, it may even lend support to the view that the Proclus of β derived from a translation of the Elements which also included the Proclus of the Liber de causis.Google Scholar For the latter does include section 48, without istibāna but otherwise complete (ed. Badawi, A., Al-Aflāṭūniyya al-muḥdatha ‘md al-’Arab (Cairo, 1955), p. 27.2–6).Google Scholar

⁹³ This article is dedicated to Dr C.S.F. Burnett and his circle of enthusiasts, above all Hillary Wiesner and Silvia Fazzo, who invited me to a series of reading sessions at the Warburg Institute, London, in the spring of 1992 to pore over puzzling pieces like D29. Occasion and inspiration were thus provided by a group of people literally sitting round a table. But it would be a mistake to infer from any irregularities of style or substance that a plurality of hands had been at work.Google Scholar