Aristotleʼs ›Physics‹ VIII, Translated into Arabic by Ishaq ibn Hunayn (9th c.): Introduction, Edition, and Glossaries [Translation ed.] 3110576996, 9783110576993, 9783110582086, 2020945249

Table of contents :
Preface
Table of Contents
Sigla and Abbreviations
INTRODUCTION
CRITICAL EDITION OF PHYSICS BOOK VIII
GLOSSARIES
BIBLIOGRAPHY

Citation preview

Aristotle’s Physics VIII, Translated into Arabic by Isḥāq ibn Ḥunayn (9th c.)

Scientia Graeco-Arabica herausgegeben von Marwan Rashed

Band 30

De Gruyter

Aristotle’s Physics VIII, Translated into Arabic by Isḥāq ibn Ḥunayn (9th c.) Edited with Introduction and Glossaries by

Rüdiger Arnzen

With a Contribution by Pieter Sjoerd Hasper

De Gruyter

ISBN 978-3-11-057699-3 e-ISBN (PDF) 978-3-11-058208-6 ISSN 1868-7172

Library of Congress Control Number: 2020945249 Bibliographic information published by the Deutsche Nationalbibliothek The Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliografie; detailed bibliographic data are available on the Internet at http://dnb.dnb.de. © 2021 Walter de Gruyter GmbH, Berlin/Boston Printing: CPI books GmbH, Leck www.degruyter.com

Preface Let’s not sugar-coat it: Graeco-Arabic textual criticism is a weird passion, shared and cultivated by a small and marginalised obsessed group, which keeps causing furrowed brows among colleagues in neighbouring disciplines and blunt perplexity or xenophobia among those who —understandably— don’t give two figs about philology. I am afraid the present piece of philological microscopy will not ameliorate this situation: on the contrary! Even so, Aristotle’s Physics and its Arabic and Latin versions is perhaps one of the most paradigmatic and typical challenges for Graeco-Arabic and AraboLatin Studies, embracing in a single stroke a multiplicity of exciting scholarly problems and questions. I cannot think of a better example for Harold Bloom’s oft-quoted dictum that “reading a text is necessarily the reading of a whole system of texts, and meaning is always wandering around between texts” (Kabbalah and Criticism, New York 1975, p. 107). To mention but four points: Barely any Aristotelian work is fraught with so many and so weighty uncertainties of the Greek wording and its redaction in general as the collection of treatises known as Physica. The only modern edition of the outstanding ninth century Arabic translation, one of the most accurate and congenial of all the medieval scientific translations from the Greek, is absolutely inadequate and unreliable. The Arabic manuscript preserving this translation is a precious treasure worthy of being unearthed even apart from the translation itself, because it contains thousands of 10th century scholia and comments of considerable evidence and value both in terms of textual criticism and for the history of philosophy. During the 12th and 13th centuries the Arabic version enjoyed (at least) three translations into Latin, only a very small portion of which is accessible in a modern critical edition, while none of them has yet been exploited systematically and completely with regard to the establishment of the Arabic or Greek texts. The present book addresses those who appreciate and accept the above challenges as veritable scholarly desiderata, although it will probably make disappointing and stodgy reading to anybody else. Be that as it may, it is my pleasant duty to announce here that my work was supported by a research grant of the Fritz Thyssen Stiftung für Wissenschaftsförderung (Cologne, Germany). I wish to express my gratitude to this foundation and to the two anonymous referees who recommended the funding.

VI

Preface

Equally essential and constitutive was the continuous cooperation of two learned colleagues: Pieter Sjoerd Hasper readily put at my disposal his complete collation of 14 Greek manuscripts of Aristotle’s Physics and the Latin vetus translatio and made me aware of important variant readings in the manuscripts of Simplicius’ commentary. Furthermore, he kindly accepted my invitation to contribute a chapter of the introduction below, which —for the first time since Ross’s 1936 edition— attempts to offer a fresh and more comprehensive systematic approach to the Greek manuscript tradition of the Physics (cf. below, chapter VI). Horst Schmieja collaborated at all stages of the project in two distinct ways, first on the technical side by adapting and consistently refining the Tubingen System of Text Processing Tools (TUSTEP) for the needs of the edition and glossaries below. Secondly, he kindly granted me access to his unpublished critical edition of Michael Scot’s Latin translation of Ibn Rušd’s “Long Commentary” on the Physics and was always willing to share his knowledge about the Latin manuscripts in general and difficult passages of this text in particular. The present study could not have been realised without the collaboration of these two scholars, to whom I am deeply indebted. Furthermore, a number of colleagues and friends provided help and support in various ways. Geoffrey J. Moseley patiently took on the time-consuming task of correcting the English of my “Introduction” and made substantial suggestions for the improvement of its contents. Nicolás Bamballi offered keen and perspicacious comments and critique on matters of editorial technique and stemmatics, and shared his profound knowledge of Arabic Galenica. At the initial phase of the project, Gerhard Endress and Dimitri Gutas were so kind as to grant basic advice. Last but not least, Amos Bertolacci, Charles Burnett, Cristina D’Ancona, Gerhard Endress, and Paul Lettinck obligingly put digital and printed copies of various medieval manuscripts at my disposal. My sincere thanks go to them all. Finally, a word on ʿAbdurraḥmān Badawī (1917—2002), the 1964 editor of the Arabic translation of Aristotle’s Physics. When I started working on the present study, almost six years ago, I hoped to finish it by 2017 and planned to dedicate its publication to the memory of Badawī’s contribution to GraecoArabic Studies on the occasion of his centenary. As things fell out, other obligetions impeded me from meeting this self-imposed deadline. However, I wish to commemorate here the outstanding achievements of this, to say the least, controversial scholar, if only because without Badawī’s incredible industriousness and editorial zeal (more than 40 editions of medieval Arabic translations of Greek philosophical works), Graeco-Arabic Studies of the past 50 years would have been deprived of a large part of its working basis and could not have evolved into what it is today. Ottersberg, Summer 2020

R. A.

Table of Contents Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Sigla and Abbreviations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

V IX

I NTRODUCTION I. “Aristotle’s Physics” and Aristotle’s Physics, the post-modern confusion of textual criticism and the aims of the present book . . .

XIX

II. The Arabic manuscripts of Aristotle’s Physics . . . . . . . . . . . . . . . . II.1 Ms. Leiden Or. 583 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . II.2 The “prehistory” of ms. Leiden Or. 583 . . . . . . . . . . . . . . . . . . II.3 The marginal and interlinear notes and glosses in ms. Leiden Or. 583 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . II.3.1 Excursion: Notes drawing on the commentary by Alexander of Aphrodisias . . . . . . . . . . . . . . . . . . . . . . . . II.4 Ms. Escorial ár. 896 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

XXXVII

III. The author of the Arabic translation: Isḥāq ibn Ḥunayn . . . . . . .

LXXXI

IV. Contaminations of Isḥāq ibn Ḥunayn’s translation in ms. Leiden Or. 583 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

XCIV

V. Isḥāq ibn Ḥunayn’s Greek exemplar(s) . . . . . . . . . . . . . . . . . . . . . .

C

VI. The Greek manuscript tradition of Aristotle’s Physics (by Pieter Sjoerd Hasper) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VI.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VI.2 The three main groups of manuscripts and their stemmatic relations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VI.3 The β group: Vat. Barb. gr. 136, Vat. gr. 1025 and Erlangen A 4 VI.4 The γ group: Vind. Phil. gr. 100, Vossius Q3, Laur. 87.07, Vat. gr. 1027, Vat. gr. 241, Par. Suppl. gr. 643, the old Latin translation VI.5 The δ group: Par. gr. 1859 and Vind. Phil. gr 64 . . . . . . . . . . VI.6 Ms. Laur. gr. 87.24 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

XXXVII XLV LV LXXII LXXVI

CXIII CXIII CXIX CXXVIII CXXXII CXLV CLIII

VIII

Table of Contents

VI.7 Pachymeres’ paraphrase of the Physics . . . . . . . . . . . . . . . . . . . VI.8 Some remarks on the stemmatic relation between Simplicius and the direct tradition . . . . . . . . . . . . . . VI.9 The α group: Par. gr. 1853 and the main exemplar of the Arabic translation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VI.10 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

CLXI CLXVI CLXXVII CLXXXVI

VII. The indirect tradition of Isḥāq ibn Ḥunayn’s translation . . . . . . . CLXXXVIII VII.1 Quotations in Arabic sources . . . . . . . . . . . . . . . . . . . . . . . . . CLXXXVIII VII.2 Arabo-Latin versions of Isḥāq ibn Ḥunayn’s translation . . . . CC VII.3 The stemmatic relationship between the Latin translations, CCIX ms. Leiden Or. 583 and ms. Esc. ár. 896 . . . . . . . . . . . . . . . . VIII. Results and materials for a tentative stemma codicum . . . . . . . . .

CCXXVII

IX. Quotations from Aristotle’s Physics in two treatises by Alexander of Aphrodisias . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

CCXXXI

X. Principles of the present edition . . . . . . . . . . . . . . . . . . . . . . . . . . . . X.1 Establishment of the Arabic text and documentation of the underlying Greek readings . . . . . . . . . . . . . . . . . . . . . . . X.2 Structure, layout and apparatuses . . . . . . . . . . . . . . . . . . . . . . . .

C RITICAL E DITION OF P HYSICS BOOK VIII . . . . . . . G LOSSARIES

CCXXXIV CCXXXIV CCXL

1

............... ..........................

121

Greek—Arabic glossary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Arabic—Greek glossary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

125 179

Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

263

Indices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Index nominum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Index locorum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

279 279 285

Sigla and Abbreviations 1. Greek manuscripts, editions, commentaries and translations of the Physics

α A As

Am

Ar

β b Bek γ Camotiana

Car δ e E E1/E2 F H HB HaGa

Branch α represented by mss. E Ψ Alexander’s Commentary as quoted by Simplicius Fragments of Alexander’s Commentary (extant in the form of scholia), edited by M. Rashed, Alexandre d’Aphrodise: Commentaire perdu à la Physique d’Aristote, Berlin 2011 Quotations of the Physics in Alexander’s treatise Fī mabādiʾ alkull, edited by C. Genequand, Alexander of Aphrodisias, On the Cosmos, Leiden 2001 Quotations of the Physics in Alexander’s Radd ʿalā Ǧālīnūs fīmā ṭaʿana bihī ʿalā Arisṭū fī anna kulla mā yataḥarraku fa-innamā yataḥarraku ʿan muḥarrik (Against Galen’s Εἰς τὸ πρῶτον κινοῦν ἀκίνητον), edited by N. Rescher & M. E. Marmura, Islamabad 1965 Branch β represented by mss. L N R ms. Parisinus gr. 1859, ca. 1300 Aristoteles, Opera. Ex rec. I. Bekkeri. Vol. 1. Berlin 1831 Branch γ represented by mss. F H I J Q w Johannes Baptista Camotius (Giovanni Battista Camozzi, ed.), Aristotelis De Physica Auscultatione, De Coelo, De Mundo [...] Tomus II, apud Aldi Filios, Venetiis 1551 Aristote, Physique. Texte établi et traduit par H. Carteron, 31961 Branch δ represented by mss. b e ms. Vindobonensis phil. gr. 64, middle of the 15th c. ms. Parisinus 1853, beginning/middle of the 10th c. ms. Parisinus 1853, first/second hand ms. Laurentianus 87.7, 14th c. ms. Vaticanus gr. 1027, end of the 12th / beginning of the 13th c. H as read by Bekker Aristotle, Physica. Transl. R. P. Hardie, R. K. Gaye, Oxford 1930

X

Sigla and Abbreviations

I J K Λ L M N Π [(praeter #)] P

Pc Pl Pp Pa Pacius Pel Pra Q R Ross S

Sc Sl Sp

ms. Vaticanus gr. 241, 13th c. ms. Vindobonensis gr. 100, 9th c. ms. Laurentianus 87.24, 12th c. The Greek exemplar of the Latin translatio vetus of the Physics (by James of Venice, before ca. 1150) [not extant] ms. Erlangen A4, 15th c. Parallels of the Physics in Aristotle’s Metaphysics ms. Vat. Barb. gr. 136, 12th c. Agreement of the manuscript branches β, γ, δ and mss. E and K [except for ms. #] Philoponus’ Commentary (in general), Ioannis Philoponi in Aristotelis Physicorum libros tres priores commentaria. Ioannis Philoponi in Aristotelis Physicorum libros quinque posteriores commentaria. Edidit H. Vitelli, Berlin 1887-1888 (CAG XVI-XVII), cf. infra § 5: “Rashed 2016”, “Rescigno”. Citation in Philoponus’ Commentary Lemma in Philoponus’ Commentary Paraphrase in Philoponus’ Commentary [Pseudo-]Michael Psellos, Kommentar zur Physik des Aristoteles, ed. L. Benakis, Athens 2008 [rather by George Pachymeres] Iulius Pacius (Giulio Pace, ed.), Aristotelis Naturalis auscultationis libri VIII, Francoforti 1596 Aristote, Physique. Traduction et présentation par Pierre Pellegrin, Paris 22002 Aristoteles, Acht Bücher Physik. Griechisch und deutsch hrsg. von Carl Prantl, Leipzig 1854 ms. Leidensis Voss. Q3, 12th c. ms. Vaticanus gr. 1025, 13th c. Aristotle’s Physics. Ed. with introduction and commentary, Oxford 1936 Simplicius’ Commentary (in general), Simplicii in Aristotelis Physicorum libros quattuor priores commentaria. Simplicii in Aristotelis Physicorum libros quattuor posteriores commentaria. Edidit H. Diels, Berlin 1882-1895 (CAG IX-X). Citation in Simplicius’ Commentary Lemma in Simplicius’ Commentary Paraphrase in Simplicius’ Commentary

Sigla and Abbreviations

T w Wag WiCo ξ

Ψ ΨΩ Ψ*

XI

Themistii in Aristotelis Physica paraphrasis. Edidit H. Schenkl, Berlin 1900 (CAG V.2) ms. Parisinus Suppl. gr. 643, 13th c. Aristoteles, Physikvorlesung. Übersetzt von H. Wagner, Darmstadt 1967 Aristotle, Physics. With an English translation by P. H. Wicksteed & F. Cornford, London 1929 Greek readings reconstructed on the basis of marginal or interlinear readings in Φ (cf. below, § 2) explicitly labelled as variae lectiones; such var. lectiones may trace back to Isḥāq ibn Ḥunayn’s translation of γράφεται-notes in Ψ, or to another Greek ms. consulted by Isḥāq ibn Ḥunayn, or to later collation notes based on another Arabic translation of the Physics (viz. a commentary) Isḥāq ibn Ḥunayn’s main Greek exemplar (probably 8th/early 9th c.) reconstructed on the basis of the Arabic text preserved in Φ Isḥāq ibn Ḥunayn’s Greek exemplar reconstructed on the basis of an Arabic text preserved defectively in Φ, yet correctable by means of the indirect Arabic and/or Arabo-Latin tradition Isḥāq ibn Ḥunayn’s Greek exemplar reconstructed on the basis of an Arabic text emended with the help of the received Greek text or divined by the present editor due to a corruption in Φ and the whole indirect Arabic and Arabo-Latin tradition

2. Arabic manuscripts, readings and editions of the Physics and related commentaries Ar.

Bad car Ǧc

Isḥāq ibn Ḥunayn’s Arabic interpretation reflecting a certain disparity between the Greek and Arabic texts without implying necessarily a Greek Vorlage different from the received Greek text (cf. below, § 4) Arisṭūṭālīs, al-Ṭabīʿa. Ed. ʿAbdurraḥmān Badawī, 2 vols., Cairo 1384/1964—1385/1965 Copy of Isḥāq ibn Ḥunayn’s translation used by Gerard of Cremona for his Latin translation (= C), before 1180 [not extant] Citation of Isḥāq ibn Ḥunayn’s translation in Ibn Bāǧǧa’s Commentary (Shurūḥāt al-Samāʿ al-ṭabīʿī, ed. Maʿn Ziyāda, Beirut 1398/1978)

XII

Sigla and Abbreviations

Ǧp

Paraphrase of Isḥāq ibn Ḥunayn’s translation in Ibn Bāǧǧa’s Commentary Θ Copy of Isḥāq ibn Ḥunayn’s translation that served as an exemplar of θ1, θ2, etc. and was a third generation ancestor of Φ, ca. 930 [not extant] Copies of Θ prepared (and annotated) by scholars of the Arisθ1, θ2 ... totelian school in Baghdad between ca. 930 and 1000 [not extant] Ibn ʿAdī Yaḥyā ibn ʿAdī’s comments on the Physics as edited in “Bad” Ibn al-Ṭayyib Abū l-Faraǧ ibn al-Ṭayyib’s comments on the Physics as edited in “Bad” Σ Common ancestor of car, vzar (and other copies of Isḥāq ibn Ḥunayn’s translation circulating in 12th and 13th c. Andalusia), probably late 11th / early 12th c. [not extant] Copy of Isḥāq ibn Ḥunayn’s translation quoted in Ibn Rušd’s vzar “Long Commentary”, before ca. 1185 [not extant] Y Copy of Isḥāq ibn Ḥunayn’s translation prepared and/or owned by Yaḥyā ibn ʿAdī (d. 974) [not extant] Φ ms. Leidensis or. 583, dated 540/1129-30, reproducing the copy of Abū l-Ḥusayn al-Baṣrī, dated 395/1004 Φ1/Φ2 Readings in Φ ante correctionem and post correctionem (probably by one and the same the hand) Ω Common ancestor of Θ, Σ, Φ and ms. Escorial ár. 896, ca. 920—930 [not extant] 3. Latin manuscripts and witnesses of the Arabo-Latin translations of the Physics C Ca Cp Cv V

Arabo-Latin version by Gerard of Cremona (consensus of Ca, Cp and Cv) based on the Arabic ms. car, preserved in: ms. Aosta, Biblioteca del Seminario Maggiore 71, late 12th/ early 13th c. ms. Parisinus lat. 16141, 13th c. ms. Vindobonensis lat. 234A, 13th c. Arabo-Latin version by Michael Scot as transmitted in the lemmata of Ibn Rušd’s “Long Commentary”, based on a descendant of vzar (unpublished critical ed. by Horst Schmieja, Thomas Institute Cologne)

Sigla and Abbreviations

Z

XIII

Anonymous Arabo-Latin version (by Hermannus Alemannus?) of Book VIII, 265a27-b16 and 266b6-267b26, as transmitted in the lemmata of Ibn Rušd’s “Long Commentary”, based on a descendant of vzar (unpublished critical ed. by Horst Schmieja, Thomas Institute Cologne)

4. Abbreviations used in the apparatuses acc. add. adnot. Alex. App. 1 App. 2 App. 3 c. comm. corr. γρ. del. dittogr. err. typ. exp. fort. gen. gloss haplogr. i.l. i.m. illeg. ind.

accusative addition, added by annotation(s), footnote(s) or commentary (by modern scholar[s]) Alexander of Aphrodisias the first critical apparatus (recording Greek variant readings) the second critical apparatus (informing about the constitution of the Arabic text) the third apparatus (dealing with scholia and annotations in Φ) cum; used together with commentary (both ancient Greek and Arabic comments in Φ) correction, corrected by the scribe “γράφεται”, variant reading mentioned in one of the ancient commentaries deletion, deleted by erasure or crossing out dittography typographical error expunxit; deleted by use of superscript points by fortasse; perhaps genitive marginal or interlinear gloss or scholium next to Isḥāq ibn Ḥunayn’s translation in Φ haplography infra lineam; below the line in margine; in the margin illegible indicating, indicated by

XIV

intell. Ar.

Sigla and Abbreviations

Indicates a disparity between the received Greek text and Isḥāq ibn Ḥunayn’s translation which may originate from an elaboration or addition by Isḥāq ibn Ḥunayn and cannot be attributed unambiguously to Ψ intell. C/V/Z Indicates a disparity between the edited Arabic text and the translation by Gerard of Cremona/Michael Scot/Anonymous which may originate from an elaboration or addition by the Latin translator and cannot be attributed unambiguously to his Arabic exemplar interpol. interpolation, interpolated by interpr. Ar. Indicates a disparity between the received Greek text and the Ar. translation caused by the fact that Isḥāq ibn Ḥunayn’s Greek exemplar —although basically consistent with the received Greek text (or one of its branches)— was crabbed, ambiguous (e.g. missing accents) or slightly damaged, thus requiring some interpretative decision interpr. C/V/Z Indicates a disparity between the edited Arabic text and the translation by Gerard of Cremona/Michael Scot/Anonymous caused by the fact that an Arabic word —although basically consistent with Φ (viz. Ω)— appeared in the latter’s Arabic exemplar in an ambiguous way (e.g. missing diacritics) requiring some interpretative decision inv. inverted, (two words) written in reverse order by i’punxit Ar. Indicates a disparity between the received Greek text and Isḥāq ibn Ḥunayn’s translation caused by the latter’s idiosyncratic parsing or punctuation of the Greek text without necessarily implying a problem or variant reading in Ψ iter. iterated by l. line lac. lacuna leg. legendum; to be read litt. litterae; letters loc. locus; the place in question marg. margin non vert. Ar. Indicates the absence of an Arabic equivalent of a Greek word in Isḥāq ibn Ḥunayn’s translation which does not necessarily imply its omission in Ψ, but may have been caused otherwise

Sigla and Abbreviations

XV

non vert. C/V/Z Indicates the absence of a Latin equivalent of an Arabic word in the translation by Gerard of Cremona/Michael Scot/Anonymous which does not necessarily imply its omission in the translator’s Arabic exemplar, but may have been caused otherwise om. omitted by or not present in om. hom. omitted due to homoioteleuton by Philop. John Philoponus propos. proposed by ras. in rasura; written over an erasure by s.l. supra lineam; above the line s.p. sine punctatione; without diacritical points s.r. sub radice; (listed) under the root s.v. sub voce; (listed) under the word scil. scilicet; i. e. scr. scripsit; as read/modified by the modern editor sec. secundum; according to secl. secluded or bracketed by sim. similar(ly) Simpl. Simplicius suppl. supplied by Them. Themistius tit. title, subheading transl. translation(s), translated by transp. transposed to (the place indicated) by vid. ut videtur; as appearing in the manuscript v. vide; see 5. Modern works referred to in the apparatuses Bonitz II/III

Denniston

Hermann Bonitz, Aristotelische Studien II—III. Wien 1863 [= Sitzungsberichte der phil.-hist. Classe der kaiserlichen Akademie der Wissenschaften XLI, Bd. 8, p. 379—434, XLII, Bd. 8, p. 25—109]. John Dewar Denniston, The Greek Particles. Second ed. revised by K. J. Dover. Oxford 1950, repr. London 1996.

XVI

Diels

Sigla and Abbreviations

Hermann Diels, “Zur Textgeschichte der Aristotelischen Physik”, Philosophische und historische Abhandlungen der Königlichen Akademie der Wissenschaften zu Berlin 1882, Abhandlung 1, Berlin 1883, p. 3—42. Dozy Reinhart Pieter Anne Dozy, Supplément aux dictionnaires arabes. 2 vols., Leiden 1881, 3ème éd., Leiden 1967. Freytag Georg Wilhelm Freytag, Lexicon Arabico-Latinum. 4 vols., Halle 1830-1837. GALex I/II A Greek and Arabic Lexicon (GALex). Materials for a Dictionary of the Mediæval Translations from Greek into Arabic. Vol. I, second revised ed. by R. Arnzen, G. Endress, D. Gutas with the assistance of G. J. Moseley, Leiden 2020; Vol. II, ed. by G. Endress & D. Gutas, Leiden 2017. Giannakis Elias Giannakis, Philoponus in the Arabic Tradition of Aristotle’s Physics. Unpublished doctoral thesis, University of Oxford, St. Antony’s College, 1992. Glossary The Greek—Arabic and Arabic—Greek glossaries at the end of the present book. Gutas Theophr. Dimitri Gutas, Theophrastus On First Principles (known as his Metaphysics). Greek Text and Medieval Arabic Translation, Edited and Translated with Introduction, Commentaries and Glossaries, as Well as the Medieval Latin Translation, and with an Excursus on Graeco-Arabic Editorial Technique. (Philosophia Antiqua; vol. 119), Leiden 2010. Hayduck Michael Hayduck, Bemerkungen zur Physik des Aristoteles, Greifswald 1871. Lane Edward William Lane, An Arabic—English Lexicon. 8 vols. London 1863—1893, repr. Beirut 1968. Rashed Marwan Rashed, Alexandre d’Aphrodise, Commentaire perdu à la Physique d’Aristote (Livres IV—VIII). Les scholies byzantines. Édition, traduction et commentaire. (Commentaria in Aristotelem Graeca et Byzantina. Quellen und Studien; vol. 1.) Berlin 2011. Rashed 2016 Marwan Rashed, “Fragments inconnus du commentaire à la Physique de Jean Philopon”, in M. Rashed, L’Héritage aristotélicien: Textes inédits de l’Antiquité. Nouvelle édition revue et augmentée. Paris 2016, p. 751—777.

Sigla and Abbreviations

Rescigno

XVII

Andrea Rescigno, “Nuovi frammenti del Commento di Filopono ai libri V—VIII della Fisica”, Studia Graeco-arabica 7 (2017), p. 75—104. Torstrik Adolf Torstrik, “Die authentica der Berliner ausgabe des Aristoteles”, Philologus. Zeitschrift für das Klassische Altertum 12 (1857), p. 494—530. Ullmann NE I/II Manfred Ullmann, Die Nikomachische Ethik des Aristoteles in arabischer Übersetzung. Teil 1: Wortschatz. Wiesbaden 2011; Teil 2: Überlieferung, Textkritik, Grammatik. Wiesbaden 2012. WGAÜ (Suppl.) Manfred Ullmann, Wörterbuch zu den griechisch-arabischen Übersetzungen des 9. Jahrhunderts. Wiesbaden 2002; Supplement vols. 1-2. Wiesbaden 2006—2007. WKAS Manfred Ullmann, Wörterbuch der Klassischen Arabischen Sprache. Herausgegeben durch die Deutsche Morgenländische Gesellschaft. Vol. I, vol. II in 4 parts, Wiesbaden 1970— 2009.

INTRODUCTION I. “Aristotle’s Physics” and Aristotle’s Physics, the postmodern confusion of textual criticism and the aims of the present book This book is not about Aristotelian doctrines or their history, no matter whether relating to what is known as Aristotle’s “Physics” in general or its most influential part, the eighth and last book of this work. Nonetheless, it has been written for those interested in these doctrines or their transmission and reception in view of the fact that such philosophical studies are faced —often unnoticed by their protagonists— by obscurities and problems of their textual substratum. Rather, this book is about signs conveying meaning and (mis-) carriers of doctrines: to be more precise, about Greek, Arabic and Latin words and texts of “Aristotle’s Physics”, and their interrelation. It contains a critical edition of a 9th c. Arabic translation, which serves, among other purposes, to remove some of these problems1. By implication, its main methods, concepts and tools are those commonly used in the fields of literary and textual criticism, which in some way or another proved to be beneficial for the investigation of a cluster of texts demonstrably interrelated by a certain historical kinship or genealogical dependence, rather than the techniques of philosophical analysis or descriptive history of philosophy2. The main dilemma of genealogical-reconstructive textual criticism, which simultaneously constitutes in a way its programme, has been aptly encapsulated by Richard Tarrant in his recently published, enchanting overview of the history, methods and problems of textual criticism as follows: _____________ 1

It goes without saying that this translation —by itself and through its medieval Latin and Hebrew translations— is of crucial interest for the history of medieval and early modern philosophical doctrines among Arabic, Latin or Hebrew-writing philosophers, which is not among the topics of the present book. 2 In view of the numerous modern translations of the Aristotelian text and the fact that the extant Arabic translation is very close to the Greek text, I also refrained from providing a translation of the Arabic text.

XX

Introduction

Editors of classical texts have no difficulty in defining their aim as that of reconstructing the author’s original version, while at the same time recognizing that, given the evidence available, that aim can never be fully achieved3. However pellucid this statement may sound from one perspective, almost each of its conceptual components is fraught with ambiguities and endless scholarly debates from another, which indeed may cause difficulties in defining one’s aim as a critical editor. To name but a few, in the order of the above quotation: (i) To what extent do “classical texts” differ with respect to the editor’s aim from nonclassical texts? Is the category of “classical texts” defined by a certain timeless quality, or by the (objective and culturally independent?) limits of a particular period of time, or else by a specific group of original languages? In the latter case, are translations of “classical texts” classical texts? If time is the main criterion and “classical” means something like pre-medieval, is a literary work composed in ancient times, the oldest extant testimony (physical document) of which dates from medieval or early modern times, still a classical text? More specifically: “Aristotle’s Physics” is commonly ranked among the “classics” of ancient Greek literature. The present book deals with its medieval Arabic translation. So, does it deal with a “classical text”, and if so, does it deal with the same classical text? On the other hand, if we banish the Arabic translation for linguistic or historico-cultural reasons from our list of classics, how can we appropriately account for the fact that the meaning conveyed in it often seems to be closer to what Tarrant calls “the author̕s original version” than that transmitted in many Greek manuscripts? (ii) The definition of the editor’s aim as a process of “reconstruction” is a matter of perennial dispute, and this on two distinct levels. On a first, fundamental level, the idea of reconstruction, which is not self-explanatory and in its strict sense always implies a genealogical approach to the witnesses of a work, was and is repudiated wholesale by many philologists and textual critics4. Among the most influential trends of this kind was Joseph Bédier’s critique, which reproached Lachmannian reconstruction for fostering subjectivity and arbitraryness due to the fact that the overwhelming majority of genealogical recensions of the witnesses result in two-branched stemmata, thus allegedly leaving the editor with an arbitrary or subjective 50/50 chance of choosing the correct _____________ 3

Richard Tarrant, Texts, Editors, and Readers. Methods and problems in Latin textual criticism, Cambridge 2016, p. 2. 4 For a critical assessment of the various kinds and meanings of “reconstruction” in the field of critical editing cf. George Thomas Tanselle, “Textual Criticism at the Millennium”, Studies in Bibliography 54 (2001), p. 1—80, esp. chapters I—II and p. 74—76.

I. “Aristotle’s Physics” and Aristotle’s Physics

XXI

reading5. Another attack came from the so-called New Philology. Bernard Cerquiglini, one of its protagonists, rejects the “desire for reconstruction” as an idea based on a “hygienist system of thought about the family” and a literary “theory of the genius”, which “by magnifying a transcendent author [...] tied an authoritarian theory of the subject (the master of the sense as well as of the signifier expressing it) to the notions of origin and textual stability”6 that ought to be replaced by appropriately appreciating the divergence of versions and accepting variance as a veritable object of research and documentation in its own right7. Still another recent trend that challenges in general “reconstruction” in the above sense is formed by the so-called Material Philology, which focuses on individual synchronic aspects rather than diachronic elements of a textual tradition, such as the interrelation between textual and paratextual features or the question how individual material characteristics of a manuscript or printed book interact with social and cultural circumstances of its production8. On another, less fundamental level, we are faced with the enduring controversy over the correct method of reconstruction held by all those who generally would (have) accept(ed) Tarrant’s above-quoted dictum. Recurrent points of disagreement include the notions “error / innovation / variant reading” and _____________ 5

On Bédier cf. Alain Corbellari, Joseph Bédier. Écrivain et philologue (Publications romanes et françaises; vol. 220). Genève 1997. For reactions to Bédier’s critique and his alternative, the choice of the codex optimus (edition mainly based on the best manuscript), cf. Sebastiano Timpanaro, The Genesis of Lachmann’s Method. Edited and translated by Glenn W. Most, Chicago 2005 (original Italian edition: Florence 1963), p. 157—187. 6 Bernard Cerquiglini, In Praise of the Variant. A Critical History of Philology. Translated by Betsy Wing, Baltimore 1999 (original French edition: Paris 1989), p. 49, 61. 7 The idea as such is not as new as the label “New Philology” might suggest. For a precursor regarding authorial variance cf. Gerhard Seidel, Die Funktions- und Gegenstandsbedingtheit der Edition untersucht an poetischen Werken Bertolt Brechts (Veröffentlichungen des Instituts für Deutsche Sprache. Deutsche Akademie der Wissenschaften zu Berlin; vol. 46), Berlin 1970, esp. p. 62ff. For an example of New Philology (usually concerned with medieval and contemporary literature) applied to an ancient “classical text” cf. Mark de Kreij, “Transmission and Textual Variants: Divergent Fragments of Sappho’s Songs Examined”, in André Lardinois et al. (eds.), Texts, Transmissions, Receptions. Modern Approaches to Narratives (Radboud Studies in Humanities; vol. 1), Leiden 2015, p. 17—34. 8 On Material Philology cf. Guglielmo Gorni, “Material Philology, Conjectural Philology, Philology without Adjectives”, in Teodolina Barolini, H. Wayne Storey (eds.), Dante for the New Millennium, New York 2003, p. 44—55; H. Wayne Storey, “Method, History, and Theory in Material Philology”, in Marc van der Poel (ed.), Neo-Latin Philology: Old Tradition, New Approaches. Proceedings of a Conference Held at the Radboud University, Nijmegen, 26-27 October 2010 (Supplementa Humanistica Lovaniensia; vol. 35), Leuven 2014, p. 25—47.

XXII

Introduction

“archetype”, the role and/or effectiveness of stemmatics, the (ir)relevance of codices descripti for the reconstruction, the criterion of lectio difficilior, the question whether recentiores non deteriores, and many other issues9. (iii) The concept of “the author’s original version” and the closely related notion of “authorial intention”, the final aim of reconstruction, are likewise highly controversial. Issues up for debate are, for example, the relationship between the (reconstructible) archetype and the “original version”10, the lack of any objective or mechanical criterion for the determination of the original reading11, the problem of authorial variants (or changing authorial intentions)12 and cases of multiple authorial versions as well as versions the attribution of which is uncertain13, or, more generally, the questionable hypothesis of the original as a perfect and flawless version14. Another fundamental objection against the concept of “authorial intentions” as a methodically sound criterion of textual criticism is based on the idea that the final product of literary composition is always a social event determined not only by authorial intention but also by non-authorial (e.g., editorial, audience-related, contextual, historical, local, and —typically, but not exclusively, in modern literature— institutional) factors15. (iv) Finally, Tarrant’s equally true and bitter observation that “that aim can never be fully achieved” leaves the critical editor with the unsettling and unsettled question which editorial method to choose, once the recensio has been completed. The answers to this question provided throughout the history of textual criticism are no less manifold and controversial than the above broached subjects; they even seem to have moved further apart in recent times since editors of “classical texts” have begun to provide digital editions and, for that purpose, _____________ 9

A masterly presentation and discussion of these problems and the diverse approaches to their solution is now available in Paolo Trovato, Everything You Always Wanted to Know about Lachmann’s Method. A Non-Standard Handbook of Genealogical Textual Criticism in the Age of Post-Structuralism, Cladistics, and Copy-Text. Revised Edition. Foreword by Michael D. Reeve (Storie e linguaggi; vol. 7), Padova 2017. 10 Cf. Timpanaro, The Genesis, p. 50; Trovato, Lachmann’s Method, p. 138—154. 11 Cf. Walter Wilson Greg, The Calculus of Variants. An Essay on Textual Criticism, Oxford 1927, esp. p. 30—43; Timpanaro, The Genesis, p. 75—81. 12 Cf. George Thomas Tanselle, “The Editorial Problem of Final Authorial Intention”, Studies in Bibliography 29 (1976), p. 167—211. 13 Timpanaro, The Genesis, p. 133—136; Trovato, Lachmann’s Method, p. 158. 14 Cf. Cerquiglini, In Praise of the Variant, p. 61f., 69. 15 Most programmatically framed by Jerome J. McGann, A Critique of Modern Textual Criticism, Chicago 1983. Revised Edition. Foreword by David C. Greetham, Charlottesville 1992, esp. p. 65—94, 121—123.

I. “Aristotle’s Physics” and Aristotle’s Physics

XXIII

adapt their critical methodology to the capability of the software and tools at hand. To sum up, there is not much of a firm and well-established ground to build on. Defining one’s aim as a critical editor emerges as a complex process determined by a variety of conceptual and methodical issues, their individual compatibility with and effectiveness for the particular circumstances of the text in question and its unique ways of transmission, and the editor’s degree of willingness to interrogate theoretical assumptions and to consider and incorporate —if not as general rules, at least as tactics— insights from various subfields of text linguistics and literary criticism. On that note, the threefold aim of the present book might be described as follows16: (1) A critical edition and, if needed, reconstruction (to the extent possible) of the Linguistic Text of Isḥāq ibn Ḥunayn’s Arabic translation of what he conceived as the meanings of the Greek text of a particular Version of the work known as “Aristotle’s Physics” in due consideration of the fact that these meanings were in large part generated from an individual physical instantiation of this work17, one (or more) Greek manuscript(s) accessible to the translator (but no longer extant), but also modelled on or imported from epistemic entities other than the individual text itself, such as the translator’s prior knowledge about Aristotle’s philosophy or his preconceptions about the work in question. (2) (a) A minute comparison of this reconstructed Arabic text with the extant Greek copies of “Aristotle’s Physics”, in order to derive as much information as possible about the non-extant Greek manuscript(s) used by the translator. (b) Inferring from this derived information clues to the genealogical relationship of the extant Greek copies of “Aristotle’s Physics” and possible consequences for the reconstruction of the Greek text. (3) A (con)text-based contribution to Graeco-Arabic lexicography and by extension to early Arabic lexicography in general by means of Greek-Arabic and Arabic-Greek glossaries18. _____________ 16

Here and in what follows I borrow shamelessly a number of notions from the disciplines of text linguistics and literary criticism (more on that in the course of the subsequent explanations). 17 Or, in post-structuralist terminology, the meanings created in the translator’s interaction with the text. 18 The need for bilingual word indices as a reference tool for the establishment of the text is obvious. For the Arabic reception of the language of scientific and philosophical translations from the Greek and the benefits of more elaborate glossaries for Arabic lexicography in general, cf. Manfred Ullmann, Wörterbuch der Klassischen Arabischen Sprache.

XXIV

Introduction

In section (1), the term Linguistic Text refers to the series of signs (words and marks) the translator intended to note down as a result of his efforts to translate the Greek text into Arabic19. The Linguistic Text must be distinguished from the text of the translator’s autograph. Whether the translator ever produced a “final” fair-copy, we don’t know. Judging from the high quality and completeness of the preserved text of his translation, his activities must have been very close to that stage. However, the autograph possibly (if not even very likely) contained certain touches of revision, additions, deletions, etc., and (taking the length of the text into consideration) presumably also this or that slip of the pen by the translator himself. More important, it was a Material Text the meaning of which was not solely determined by the translator’s intention, but also by other, formal and physical aspects (cf. below). The Linguistic Text is also distinct from the reading-translating-text in the translator’s mind, which I call Conceptual Text. The latter —being the cognitive interface between comprehending the source text and re-thinking it in the language of the (still to be formulated) target text— is a sort of polyglot or interlingual proto-text made up of Graeco-Arabic approximations and equivalences (probably accompanied by provisional alternative phrasings) and certainly also non-verbal units such as mental symbols or codes representing words or phrases of the Greek text which are immediately clear to the translator and thus not (yet) in need of precise formulation20. The Linguistic Text, on the other hand, is a self-contained entity composed of Arabic _____________ Herausgegeben durch die Deutsche Morgenländische Gesellschaft. Vol. I, vol. II in 4 parts, Wiesbaden 1970—2009, vol. II, part 2 (‫ﻂ‬ ّ ‫ ﻟ‬bis ‫)ﻟﻜﻰ‬, p. IX—XI, vol. II, part 4 (‫ﻟﻮى‬ bis ‫)ﻟﻴﻮان‬, p. X—XXII; idem, Wörterbuch zu den griechisch-arabischen Übersetzungen des 9. Jahrhunderts, Wiesbaden 2002, p. 60f.; idem, Wörterbuch zu den griechisch-arabischen Übersetzungen des 9. Jahrhunderts. Supplement Band I: A—O, Wiesbaden 2006, p. 11; idem, Die Nikomachische Ethik des Aristoteles in arabischer Übersetzung. Teil 1: Wortschatz, Wiesbaden 2011, p. 23—29; idem, Theorie und Praxis der arabischen Lexikographie, Wiesbaden 2016, p. 32—35, 252–278 (v. also p. 287, s.v. “Griechische Materialien”); Gerhard Endress, “The Language of Demonstration: Translating Science and the Formation of Terminology in Arabic Philosophy and Science”, Early Science and Medicine 7 (2002), p. 231—253. 19 I adopt the terms Linguistic Text, Conceptual Text and Material Text from Peter L. Shillingsburg, “Text as Matter, Concept, and Action”, Studies in Bibliography 44 (1991), p. 31—82. However, I use them in slightly different meanings than those defined by Shillingsburg. 20 This meaning of “Conceptual Text” must not to be confused with (a) what Ricoeur calls “un troisième texte inexistant”, which contains a supposed ideal meaning (le sens même) as opposed to its linguistic instantiations in the source and target texts, cf. Paul Ricoeur, Sur la traduction, Paris 2004, p. 14; nor (b) an entirely “deverbalized” tertium comparationis that replaces the source text in the translator’s mind, on which see Anthony Pym, Exploring Translation Theories. Second Edition, London 2014, p. 17f.

I. “Aristotle’s Physics” and Aristotle’s Physics

XXV

words and thus finally abstracts or breaks away from this intrinsic GraecoArabic matching of the Conceptual Text21. Furthermore, the Linguistic Text is a text created by the translator for readers unfamiliar with the Greek language. It is, so to speak, exoteric and, thus, may contain elements “added” to the (“esoteric”) Conceptual Text for the sake of clarity with an eye toward this audience. Both Linguistic and Conceptual Texts differ again from their physical manifestation, the Material Text, in at least two respects: First, the meaning conveyed through the Material Text is directed, in some cases even manipulated, by material aspects such as the writing surface, (colours of) ink, page layout, the segmentation of the text into paragraphs, and many other things. It thus may differ from the meaning the author of the Linguistic Text originally intended to convey, although the signs proper are more or less the same22. Secondly, the Material Text may misrepresent the Linguistic Text or preserve it in an ambiguous manner due to codicological damages, scribal errors or modifications etc., most conspicuously in later copies, but as indicated above even in its first physical manifestation, the translator’s autograph. While the translator’s Conceptual Text is not accessible to us, the Linguistic Text is preserved in the form of its (deficient) individual material instantiations and their indirect tradition (cf. below, chapters II, VII). For the most part, there is no reason to assume that the signs of this Material Text differ from those of the Linguistic Text; hence no “reconstruction” is required (as mentioned above, this identity of the signs does not necessarily entail the identity of meanings). In all other cases, my editorial approach was guided by the idea or principle of an anterior superiority of the Linguistic Text, the restoration or reconstruction of which has priority over respecting and visualising the dynamics and variance of the Material Text23. Although this variance is concededly of undeniable and crucial importance for the historiography of Arabic philosophy and Begriffsgeschichte (especially for the reception of Isḥāq ibn Ḥunayn’s translation in the Aristotelian school of Baghdad), this methodology is justifiable for the following three reasons. First, any attempt to accomplish goals (2) and (3) summarized above presupposes a thoroughly established wording of the translator’s Linguistic _____________ 21

For a collection of recent studies of the cognitive aspects of translation processes cf. John W. Schwieter, Aline Ferreira (eds.), The Handbook of Translation and Cognition, Hoboken 2017. 22 Needless to add, the same disparity may be caused by socio-cultural, regional and/or historical remoteness between the author of the Linguistic Text and the user of the Material Text. 23 For a contrary approach cf. Cerquiglini, In Praise of the Variant, p. 33—45, and Gabler’s “Introduction” in Hans Walter Gabler, George Bornstein, Gillian Borland Pierce (eds.), Contemporary German Editorial Theory, Ann Arbor 1995, p. 1—16.

XXVI

Introduction

Text. However interesting a textual intervention (e.g. a brief explanatory addition or the replacement of a terminus technicus) by a tenth century scholar from Baghdad may be for the history of physical doctrines in medieval Arabic philosophy, it is absolutely irrelevant (even misleading) for making inferences from the text of the translation about the text of the Greek manuscript used by the translator (unless we assume that this later textual intervention was prompted by a collation with the same Greek manuscript) as well as for text-based studies of the influence of Graeco-Arabic translations on the medieval Arabic lexicon. Secondly, this methodology is by no means suffused with a hidden or vague intentionalist approach to the Linguistic Text. Rather each reconstructive intervention is substantiated and verified through the comparison of the Material Text ready at hand with the extant Greek manuscripts and the Arabic and Latin indirect traditions. Thirdly, variance is neither disregarded nor suppressed by this approach, but appreciated and recorded in as much detail as possible in the critical apparatuses of the edition below. It goes without saying that anyone attempting a reconstructive intervention must keep in mind that despite the valid and reasonable assumption that between the meaning of the translator’s Linguistic Text and the meaning of the Greek text there is a relation of convergence, ideally one of coincidence, this intended convergence is not universal. Apart from blunt misinterpretations or variant readings, the semantic relationship may appear odd or implausible to us, because we ascribe to the same Greek or Arabic word or phrase today another meaning than the translator in 9th c. Baghdad did, just as the meaning conveyed by the Greek text to Aristotle’s contemporaries certainly differed from the meaning inferred almost one thousand years later by its famous commentators John Philoponus and Simplicius. As mentioned above, the process of translation is to a certain extent modelled by the translator’s preconceptions about the text/work and its author. Whether the creator of the below edited translation conceived the individual Greek text which formed the object of his translation activity as a complete self-contained literary unit or “work”, is not quite clear, but very likely24. Such a conception of the Aristotelian text(s) is ascertainable in scholarly Syriac literature from the second half of the eighth century on and in Arabic literature from the early ninth century25 and had already, at that time, a long Greek tradition. I refer to this _____________ 24

The preserved text of the translation contains no title or author’s name from the translator’s hand. However, neither the transmission of his translation nor the extant bibliographical references to it provide any support for the assumption that the translator conceived the object of his translation as a collection of separate treatises. 25 Cf. Yury Arzhanov & Rüdiger Arnzen, “Die Glossen in Ms. Leyden Or. 583 und die syrische Rezeption der aristotelischen Physik”, in Elisa Coda, Cecilia Martini Bonadeo

I. “Aristotle’s Physics” and Aristotle’s Physics

XXVII

concept by the phrase “Aristotle’s Physics” and to the group of texts conceived today (but probably not as early as Aristotle’s time) as a physical manifestation of this concept by the phrase Aristotle’s Physics. The conception itself is a universal mental construct which occupies an indifferent position as regards the meaning or meanings ascribed to or conveyed by the text (e.g., although Alexander of Aphrodisias, Thomas Aquinas and Martin Heidegger certainly differed substantially as regards the meanings of the text, they applied the term “Aristotle’s Physics” unanimously to a mental entity characterised by a certain diachronic identity the ongoing validity of which enables us to understand what they were talking about). It is even possible to use this concept in a meaningful manner without any knowledge or opinion about these meanings (if you ask a well-trained bookseller for a copy of “Aristotle’s Physics”, she or he will immediately understand what you are looking for without ever having read a single line of the work)26. It is a universal concept in so far as it retains a certain sameness or stability while admitting a variety of features and manifestations, more precisely (a) multiple spatial and temporal instantiations of the same text (e.g., several copies of Aristotle’s Physics, revised edition by W. D. Ross, Oxford 1936, in different libraries or exactly the same text in this edition and its reprint Oxford 1998), (b) a variety of versions (e.g., “Aristotle’s Physics” containing the seventh book in version α and “Aristotle’s Physics” containing this book in version β), (c) variance of the actual wording (e.g., different readings in two medieval manuscripts or two different editions), (d) instantiations in different languages and with different signs or letters, (e) a great variety of physical media of instantiation (e.g., manuscript, printed book, audio book, internet, etc.), and (f) even the aggregation of all above features and dispositions27. Thus, the translator’s Linguistic Text, the main subject of the present book, is based on the translator’s interpretation of an individual physical manifestation of a particular version of what he conceived universally as “Aristotle’s Physics”. The term “versions” refers here and below to multiple sets of meaningful units (rang_____________ (eds.), De l’Antiquité tardive au Moyen Âge. Études de logique aristotélicienne et de philosophie grecque, syriaque, arabe et latine offertes à Henri Hugonnard-Roche, Paris 2014, p. 415—463, esp. p. 419—426. 26 Which raises the question addressed by Jacques Brunschwig: “quand on entre dans une librairie et qu’on achète la Physique d’Aristote, qu’est-ce qu’on achète exactement?”, cf. J. Brunschwig, “Qu’est-ce que ‘la Physique’ d’Aristote”, in François De Gandt, Pierre Souffrin (eds.), La Physique d’Aristote et les conditions d’une science de la nature (Bibliothèque d’histoire de la philosophie; Nouvelle Série), Paris 1991, p. 11—40. 27 Here and in what follows I draw on Jorge Gracia’s reflections on the nature of texts; cf. Jorge J. E. Gracia, Texts. Ontological Status, Identity, Author, Audience (SUNY Series in Philosophy), Albany 1996.

XXVIII

Introduction

ing from clauses or syntagmata over paragraphs up to whole chapters, independent from the languages or signs of transmission) each whole of which resembles any other whole of sets to such a degree that it is recognised as a particular instantiation of the universal concept of one and the same work. Versions differ from one another in that one or several meaningful units of one version have been modified or replaced through revised or newly composed meaningful units, or in that such meaningful units have been deleted from an earlier version or originally added or interpolated from another source, or in that a certain series of meaningful units has been arranged in another order (while preserving the contents of each unit as such) in the other version in such way that the modification, revision, deletion, addition or re-structuring is the result of an intentional editorial intervention rather than of “infelicities” of the transmission28. Such editorial interventions may be performed by the original author (authorial versions/ revisions), editors, scholarly scribes of manuscripts or translators. So far, we have moved away quite a distance from the genealogical editorial ideal of “reconstructing the author’s original version” (Tarrant). Not only has the term “original version” turned out to be inappropriate and in need of further scrutinising, but also our subject is a translation, and the translator is not the author of “Aristotle’s Physics”, but, if anything, the author (i.e. creator) of its Arabic translation (he is not even the author of a peculiar version in the above specified sense of “version”, as he did not introduce any substantial intentional modification of the meaning, but faithfully reproduced one of the known and otherwise attested versions of the text)29. However, matters get further complicated by the following two facts: _____________ 28

It goes without saying that the boundaries between versions and texts differing from one another through mere unintentional modifications in the transmission or what seem to be insignificant variations are fluid and often indeterminate. For a discussion of how “versions” relate to authority and the different stages in the process of publishing cf. Shillingsburg, Text as Matter, p. 49—53, 65—72. For other views which bring “versions” close to “copies” cf. Greg, The Calculus of Variants; Gracia, Texts, p. 101f., Cerquiglini, In Praise of the Variant, p. 37—40. 29 For the debate about the author-translator relationship cf. Anthony Pym, “The Translator as Non-Author, and I Am Sorry about That”, in Claudia Buffagni, Beatrice Garzelli, Serenella Zanotti (eds.), The Translator as Author. Perspectives on Literary Translation (Proceedings of the International Conference, Università per Stranieri of Siena, 28-29 May 2009), Münster 2011, p. 31—44; on the distinction between author and translator also Gracia, Texts, p. 108, 114. (N.B.: A translator may be called “author” in the sense that she or he creates through the dissemination of the translation new meanings of particular words or expressions in the lexicon of the language of the target text or enriches it with neologisms. However, this is not the meaning of “author” we are concerned with here.)

I. “Aristotle’s Physics” and Aristotle’s Physics

XXIX

(1) The main textual witness at our disposal is not a separate copy of Isḥāq ibn Ḥunayn̕s translation, but a copy of an 11th c. “edition”, which includes apart from the translation marginal and interlinear glosses and notes as well as commentary sections proper (cf. below, chapter II). Evidently, the creator of this “edition” had access not only to different copies of Isḥāq ibn Ḥunayn’s translation, but also to (partial) translations of “Aristotle’s Physics” prepared by other translators. Thus, the task of the present day critical editor of Isḥāq’s translation includes two procedures that must be carefully kept apart. On the one hand, one must determine the mode and degree of potential editorial interventions into the text of the translation on the part of this 11th c. editor and decide how to deal with them appropriately in terms of editorial technique. On the other hand, one must remove errors and clarify other problems that occurred in the manuscript tradition of this 11th c. “edition” itself (cf. below, chapter IV). (2) Furthermore, the commonly accepted meaning of “author” is not only unsuitable for the creator of the Arabic translation, but even questionable with respect to the universal concept of “Aristotle’s Physics” as a self-contained literary work. The latter is a conventional term applied from Late Antiquity onwards to a cluster of texts marked by certain inconsistencies and occasionally vague coherence. Although there can be little doubt that each of these texts (or at least one version of them) was composed by Aristotle, it is rather unlikely that he intended them to be combined or united as a self-contained work30. On the contrary, internal references within Aristotle’s Physics as well as cross-references in other works by Aristotle point to the fact that Aristotle drafted the single books of the Physics or groups of several books as separate treatises to which he referred by different titles31. Both the chronological order of these (groups of) treatises and the question whether the intervals of time (hence, the concomitant philosophical development of Aristotle) between their composition were long or short are matters of dispute32. _____________ 30

Cf. William David Ross (ed.), Aristotle’s Physics. A revised text with introduction and commentary, Oxford 1936, p. 1—4; Brunschwig, Qu’est-ce que ‘la Physique’, p. 20f.; Pierre Pellegrin (ed.), Aristote: Physique. Traduction, présentation, notes, bibliographie et index. 2e édition revue, Paris 2002, p. 27. 31 Cf. Werner Wilhelm Jaeger, Studien zur Entstehungsgeschichte der Metaphysik des Aristoteles, Berlin 1912, p. 150—151; Ross, Aristotle’s Physics, p. 2—3; Friedrich Solmsen, Aristotle’s System of the Physical World. A Comparison with His Predecessors (Cornell Studies in Classical Philology; vol. 33), Ithaca 1960, p. 71—73; Brunschwig, Qu’est-ce que ‘la Physique’, p. 28f.; Pellegrin, Aristote: Physique, p. 22—27. 32 Cf. Paul Tannery, “Sur la Composition de la Physique d’Aristote”, Archiv für Geschichte der Philosophie 7 (1893), p. 224—229; idem, “Sur la Composition de la Physique d’Aristote. Second article”, ibid. 9 (1896), p. 115—118; Georges Rodier, “Sur la composition de la Physique d’Aristote”, ibid. 8 (1895), p. 455—460; idem, “Sur la composition de la

XXX

Introduction

Our knowledge about the history of these texts in the Peripatetic school and the Hellenistic period is rather vague. Aristotle’s disciple Eudemus of Rhodes composed his own Physics, the structure of which —judging from the surviving fragments— resembled more or less that of “Aristotle’s Physics”33. However, this does by no means prove that he actually conceived the cluster of texts that he took as a model as a complete self-contained writing, because we do not know in which way he embedded its structure in his overall conception of Aristotle’s natural philosophy and its branches. Furthermore, both Eudemus and his younger colleague Theophrastus apparently referred to the first four books of what would later be known as “Aristotle’s Physics” by the title “Physics” (Τὰ φυσικά) and to the second group of texts as “On Motion” (Περὶ κινήσεως), which does not suggest a holistic conception of “Aristotle’s Physics” in the Peripatetic school34. In any case, we are relatively safe in assuming that the unitary “work” conception does not trace back to Andronicus of Rhodes (fl. end of the first c. BC35), who apparently conceived of the two groups of books I—V and books VI—VIII as connected, yet discrete parts and did not think of what is known today as “Aristotle’s Physics” as a self-contained single whole36. Although Andronicus is often referred to as the one who produced a kind of canonical edition of the Aristotelian corpus, his role in the construction of such a self-contained _____________ Physique d’Aristote. Second article”, ibid. 9 (1896), p. 185—189; Ross, Aristotle’s Physics, p. 7—11; Helen S. Lang, Aristotle’s Physics and Its Medieval Varieties (SUNY Series in Ancient Greek Philosophy), Albany 1992, p. 2—13; H. Wagner’s “Einleitung” in Aristoteles. Physikvorlesung. Übersetzt von Hans Wagner. Fünfte, durchgesehene Auflage. (Aristoteles. Werke in deutscher Übersetzung; vol. 11), Berlin 1995, esp. p. 275—278. 33 Cf. Jonathan Barnes, “Roman Aristotle”, in J. Barnes, Miriam Griffin (eds.), Philosophia Togata II: Plato and Aristotle at Rome, Oxford 1997, p. 1—69, esp. p. 60, Brunschwig, Qu’est-ce que ‘la Physique’, p. 27, also Simplicius’ remarks on Eudemus, Simplicii in Aristotelis Physicorum libros quattuor posteriores commentaria, edidit Hermannus Diels (Commentaria in Aristotelem Graeca, vol. 10), Berlin 1895 (henceforth: Simpl., Phys. II), p. 924.18, 1036.13f. 34 Cf. Pantelis Golitsis, Les Commentaires de Simplicius et de Jean Philopon à la Physique d’Aristote. Tradition et Innovation (Commentaria in Aristotelem Graeca et Byzantina. Quellen und Studien; vol. 3), Berlin 2008, p. 72—74; Barnes, Roman Aristotle, p. 35f., 60. It is also unclear which position (if any) in the second ensemble of texts Eudemus ascribed to what is known today as book VII of “Aristotle’s Physics”; cf. Ross, Aristotle’s Physics, p. 18, Barnes, Roman Aristotle, p. 61. 35 Cf. Barnes, Roman Aristotle, p. 22f. 36 On Andronicus̕ conception of the Physics cf. Simpl., Phys. II, p. 923f., Barnes, Roman Aristotle, p. 34—36, 60 and 67—69, and Paul Moraux, Les listes anciennes des ouvrages d’Aristote (Aristote: Traductions et Études), Louvain 1951, p. 235f.

I. “Aristotle’s Physics” and Aristotle’s Physics

XXXI

Aristotelian writing entitled Physics remains dubious and was in all likelihood rather insignificant37. The earliest vague evidence for a conception of “Aristotle’s Physics” qua selfcontained work composed of eight books traces back to Adrastus of Aphrodisias (middle of the second c. CE)38. According to Simplicius, Adrastus reports in his (lost) work On the Order of Aristotle’s Writings (Περὶ τῆς τάξεως τῶν Ἀριστοτέλους συγγραμμάτων) that some refer to Aristotle’s “discourse” (πραγματεία) by the title “On Principles” (Περὶ ἀρχῶν), others by “Lecture on Physics” (Φυσικὴ ἀκρόασις), while yet others call the first five books (τὰ πρῶτα πέντε) “On Principles” and the remaining three ones (τὰ λοιπὰ τρία) “On Motion” (Περὶ κινήσεως)39. Provided the terms “the first five” and “the remaining three books” indeed originate from Adrastus (or one of his sources) and are not owed to Simplicius̕ own conception of “Aristotle’s Physics”, Adrastus must have regarded Aristotle’s πραγματεία as a complete work in its own right composed of eight books40. However, the strength of this evidence is weakened by the fact that Simplicius uses almost the same terminology when referring to Aristotle himself: τὰ μὲν πρῶτα πέντε Περὶ ἀρχῶν εἴωθε καλεῖν ὁ Ἀριστοτέλης, τὰ δὲ ἐφεξῆς Περὶ κινήσεως (“Aristotle usually calls the first five [books] ‘On Principles’ and the following ones ‘On Motion’)41. Thus, the first irrefutable evidence for a unitary literary conception of “Aristotle’s Physics” is the commentary by Alexander of Aphrodisias42 which, _____________ 37

Cf. Barnes, Roman Aristotle, p. 28—31, 60; Brunschwig, Qu’est-ce que ‘la Physique’, p. 29—34. 38 On Adrastus cf. Paul Moraux, Der Aristotelismus bei den Griechen. Von Andronikos bis Alexander von Aphrodisias. Zweiter Band: Der Aristotelismus im I. und II. Jh. n. Chr. (Peripatoi; vol. 6), Berlin 1984, p. 294—332; Richard Goulet, “Adraste d̕Aphrodise”, in R. Goulet (ed.), Dictionnaire des philosophes antiques. Vol. 1: Abam(m)on à Axiothéa, Paris 1989, p. 56—57. 39 Simplicii in Aristotelis Physicorum libros quattuor priores commentaria, edidit Hermannus Diels (Commentaria in Aristotelem Graeca, vol. 9), Berlin 1882 (henceforth: Simpl., Phys. I), p. 4.11-15. 40 Cf. Marwan Rashed, Aristote: De la génération et la corruption, Paris 2005, p. ccxvii— ccxviii; Myrto Hatzimichali, “Andronicus of Rhodes and the Construction of the Aristotelian Corpus”, in Andrea Falcon (ed.), Brill’s Companion to the Reception of Aristotle in Antiquity (Brill’s Companions to Classical Reception; vol. 7), Leiden 2016, p. 81—100, esp. 94—96. 41 Simpl., Phys. I, p. 6.9-10. Similar bipartite conceptions are reported for Nicolaus of Damascus and Porphyry, cf. Barnes, Roman Aristotle, p. 35f. 42 Cf. Marwan Rashed, Alexandre d’Aphrodise, Commentaire perdu à la Physique d’Aristote (Livres IV—VIII). Les scholies byzantines. Édition, traduction et commentaire. (Commen-

XXXII

Introduction

by the mere fact of commenting on the contents of these eight books in the form of a continuous self-contained literal commentary, attributes implicitly to what had been conceived either as a patchwork of separate treatises or as a bipartite composite the same formal rank and literary genre as to other, undoubtedly selfcontained Aristotelian writings that formed the object of separate commentaries in the Peripatetic school of Athens or the school of Alexandria43. Strictly speaking, it is here (or perhaps slightly earlier) where the history of “Aristotle’s Physics” as an integral piece of literature begins and where the “archetype” of this tradition must be sought44. Of course, this does not preclude any serious research on the wording of the Aristotelian texts prior to this turning point. However, whenever this research is faced with questions involving or presupposing a unitary holistic view of these texts (such as the wording of introductions and conclusions of particular books or of the transition from one book to another, subtitles, the numbering of the books, etc.), it comes here to its limits. Evidently, this applies also to the second aim of the present book. As the Greek manuscript(s) used by the translator stood in the tradition of this “archetype”, we cannot expect to gain from our reconstruction on the basis of the Arabic text insights about any particular conception of “Aristotle’s Physics” (no matter whether conglomerate or unitary) antedating this tradition. In general, this second aim, which consists of (a) deriving as much information as possible about the non-extant Greek manuscript(s) used by the translator and (b) inferring from this information ideally substantial clues for the establishment of the Greek text, is marked by a certain imbalance between the incontestable evidence that it is justified or desirable for the sake of scholarship and the inherent uncertainty that whatever results are achieved will indeed be useful for these purposes. _____________ taria in Aristotelem Graeca et Byzantina. Quellen und Studien; vol. 1.) Berlin 2011; Golitsis, Les Commentaires, p. 58f. 43 Alexander is followed in this respect by John Philoponus; cf. Barnes, Roman Aristotle, p. 34, note 155. 44 “Archetype” means here the first manifestation of a particular literary form assigned to the cluster of texts known as “Aristotle’s Physics” rather than a particular manuscript from which all extant and reconstructable witnesses derived or the author’s final version intended for publication. For the different meanings of “archetype” in textual criticism cf. Trovato, Lachmann’s Method, p. 63—67, and the literature quoted there. Rashed, De la génération, p. ccix—ccx, points to the role of Aphrodisias (in the historic region of Caria, western Anatolia) in the formation of the “Aristotelian corpus” in the second half of the second c. CE. We may add the observation that it was exactly in that period that our two earliest witnesses for a conception of “Aristotle’s Physics” as a unitary work, Adrastus and Alexander, were active in that place (as Alexander’s teacher, Herminus, likely also was).

I. “Aristotle’s Physics” and Aristotle’s Physics

XXXIII

As for its justification, it doesn’t take a genius of classical philology to become immediately aware of the fact that we are still far from a reliable standard edition of the Greek text. Not only is the 1936 edition by Sir David Ross at hundreds of places in conflict with the preceding editions by Immanuel Bekker, Carl Prantl, Philip H. Wicksteed and Henri Carteron, but also the more recent translators and revisors of the text such as Francis Cornford, Hans Wagner or Pierre Pellegrin in turn often felt urged to follow selectively and inconsistently this or that edition or even ventured to propose again different modifications or alternative readings, partly with scanty support from the Greek manuscripts. Furthermore, the need for a new edition of the Greek text is plainly shown by a couple of severe shortcomings from which all available editions suffer. Perhaps the most conspicuous one is the blunt disregard of a number of important independent manuscripts or manuscript families. Another serious problem is raised by the over-simplified conception of the stemmatic relationship of the manuscripts underlying all editions. As shown by Marwan Rashed’s study on the transmission of Aristotle’s De generatione et corruptione, which is closely related to that of the Physics, we are not faced with a clear-cut bifurcation, but with a much more complicated structure including important witnesses which occupy certain middle positions between what were previously treated as the two main branches45. During the past two decades, Aristotle’s natural philosophy in general as well as its key text, the Physics, in particular attracted —after a period of rather limited activity during the second half of the 20th century— a constantly increasing attention among historians of philosophy and historians of science. Apart from hundreds of recent monographs and articles, which deal with almost all major doctrines and conceptions of Aristotle’s Physics, both in terms of philosophical historiography and in pursuit of leading questions of contemporary physics, aspects of methodology and didactic purposes, the medieval or modern reception of Aristotle’s natural philosophy and other issues, a remarkable number of new translations into western languages have been published and international research projects have been launched46. However, all translators relied more or _____________ 45 Marwan Rashed, Die Überlieferungsgeschichte der aristotelischen Schrift De generatione et corruptione (Serta Graeca. Beiträge zur Erforschung griechischer Texte, vol. 12), Wiesbaden 2001. For more details, see P. S. Hasper’s contribution below, chapter VI. 46 To mention but the most recent translations in chronological order: Aristóteles. Física. Introducción, traducción y notas de Guillermo R. de Echandía, Madrid 1995; Aristóteles. Física, libros III—IV. Traducción, introducción y comentario Alejandro Vigo, Buenos Aires 1995; Aristotele. Fisica, libri I e II. Traduzione e cura di Ferruccio Franco Repellini, Milano 1996; Aristóteles. Física. Texto revisado y traducido por José Luis Calvo Mar-

XXXIV

Introduction

less on the previous editions; and none of these studies addresses the pending philological questions47. Only the 1996 edition of the Greek text by J. L. Calvo Martínez tries to push the process a step forward by taking into consideration three manuscripts ignored by the previous editors. However, as Calvo Martínez himself has to admit in his introduction, this attempt was rather fruitless in so far as these manuscripts, the Escorialenses Σ.II.15, Σ.III.2 and Φ.III.9, turned out to be (partly rather late) descendants of ms. Vat. gr. 241 (I), which of course had been considered by Ross48. Their sole benefit thus consists in providing a corrective for individual readings, errors and innovations, in the famous ms. Vindobonensis 100 (J). That the consideration of medieval Graeco-Arabic translations may contribute —to a lesser or greater degree— both to recovering otherwise lost readings and to determining in a more reliable way the relationship of the extant Greek manuscripts has been illustrated in general by a number of recent studies. It will suffice here to point exemplarily to Uwe Vagelpohl’s study on Aristotle’s Rheto_____________ tínez, Madrid 1996; Aristotle. Physics. Translated by Robin Waterfield with an introduction and notes by David Bostock (Oxford World’s Classics), Oxford 1996; Aristote. Physique. Introduction de Lambros Couloubaritsis, traduction de Annick Stevens, Paris 1999; Fisica di Aristotele. A cura di Marcello Zanatta, Torino 1999; Aristotle. Physics, Book VIII. Translated with a commentary by Daniel W. Graham, Oxford 1999; Aristote. Physique. Traduction, présentation, notes, bibliographie et index par Pierre Pellegrin, Paris 2000, 2e édition revue, Paris 2002; Aristóteles. Física. Traducción y notas de Ute Schmidt Osmanczik, introducción de Antonio Marino López, México 2001; Aristotle. Physics or Natural Hearing. Translated and edited by Glen Coughlin (William of Morebeke Translation Series), South Bend, Indiana, 2005; Aristotele. Fisica. Saggio introduttivo, traduzione, note e apparati di Luigi Ruggiu, Milano 2007; Aristóteles. Física I e II. Prefácio, introdução, tradução e comentários Lucas Angioni, Campinas 2009; Aristotele. Fisica. Testo greco a fronte. Introduzione, traduzione, note e apparati di Roberto Radice, Milano 2011; Aristotele. Fisica, libro III. Introduzione, traduzione e commento di Monica Ugaglia, Roma 2012; Aristotele. Fisica, libro IV. Introduzione, traduzione e commento di Laura M. Castelli, Roma 2012; Farah Cherif Zahar, Le traité d’Aristote sur l’éternité du mouvement. Traduction et commentaire de Physique VIII. Thèse de doctorat en Philosophie. Université Paris IV (Paris-Sorbonne), 2016 (http://www.theses.fr/ 2016PA040186); Aristotle. Physics. Translated with an introduction and notes by C. D. C. Reeve, Indianapolis 2018. 47 With the exception of Olshewsky’s new approach to the problem of the authenticity of book VII of the Physics: Thomas M. Olshewsky, “The Bastard Book of Aristotle’s Physics”, Classical Quarterly 64 (2014), 58-74. 48 Cf. Calvo Martínez (ed.), Aristóteles. Física, p. lxxiii—lxxvi. As a consequence, the critical apparatus of Calvo Martínez’ edition does not register a single variant reading of the three Escorialenses.

I. “Aristotle’s Physics” and Aristotle’s Physics

XXXV

ric 49, Dimitri Gutas’ edition and study of Theophrastus’ Metaphysics50, the edition and study of Aristotle’s Poetics by Leonardo Tarán and Dimitri Gutas51, the study of the Arabic translation of Aristotle’s Nicomachean Ethics by Ernst A. Schmidt and Manfred Ullmann52, Oliver Overwien’s edition of Hippocrates’ De humoribus 53, not to mention the “Index of Variant Greek Passages” covering more than (!) fifty ancient Greek texts (with hundreds of variant readings reconstructed from the Arabic translations) published in the Greek and Arabic Lexicon (GALex)54. As in most of the above cases, the Arabic translation of the Physics antedates the oldest extant Greek manuscripts, thus presupposing a still earlier date of the translator’s Greek exemplar(s). It gains an even greater importance through the fact that the latter evidently formed part of the so-called α-branch of the tradition (cf. above), which is otherwise attested by only one, and at that rather faulty, witness, the famous Parisinus gr. 1853 (E). Neither the textual nor the lexicographic Graeco-Arabic significance of the translation has been taken into consideration by its previous editors55. That _____________ 49

Uwe Vagelpohl, Aristotle’s Rhetoric in the East. The Syriac and Arabic translation and commentary tradition. (Islamic Philosophy, Theology and Science. Texts and Studies; vol. 76), Leiden 2008. 50 Dimitri Gutas, Theophrastus On First Principles (known as his Metaphysics). Greek Text and Medieval Arabic Translation, Edited and Translated with Introduction, Commentaries and Glossaries, as Well as the Medieval Latin Translation, and with an Excursus on Graeco-Arabic Editorial Technique. (Philosophia Antiqua; vol. 119), Leiden 2010. 51 Aristotle’s Poetics. Editio Maior of the Greek Text with Historical Introductions and Philological Commentaries by Leonardo Tarán (Greek and Latin, edition of the Greek Text) and Dimitri Gutas (Arabic and Syriac). (Mnemosyne. Supplements. Monographs on Greek and Latin Language and Literature; vol. 338), Leiden 2012. 52 Ernst A. Schmidt, Manfred Ullmann, Aristoteles in Fes. Zum Wert der arabischen Überlieferung der Nikomachischen Ethik für die Kritik des griechischen Textes. (Schriften der Philosophisch-historischen Klasse der Heidelberger Akademie der Wissenschaften; vol. 49). Heidelberg 2012. 53 Hippokrates. Über die Säfte. Herausgegeben, übersetzt und erläutert von Oliver Overwien. (Corpus Medicorum Graecorum; vol. I 3,1). Berlin 2014. 54 Gerhard Endress, Dimitri Gutas (eds.), A Greek and Arabic Lexicon (GALex). Materials for a Dictionary of the Mediæval Translations from Greek into Arabic. (Handbook of Oriental Studies. Section One: The Near and Middle East; vol. 11). Vol. 1: ‫ أ‬to ‫أي‬, Leiden 2002, Part C (cf. second, revised ed. by R. Arnzen, G. Endress, D. Gutas with the assistance of G. J. Moseley, Leiden 2020); Vol. 2: ‫ ب‬to ‫ﺑﲔ‬, Leiden 2017, Part C. 55 Book I was edited together with the commentaries extant in the Arabic codex unicus by Wilhelm Kutsch and Khalil Georr, “al-Maqāla al-ūlā min Kitāb al-Samāʿ al-ṭabīʿī liArisṭūṭālīs”, Mélanges de l’Université Saint-Joseph 39 (1963), p. 266—312 [= p. ١—٤٧]; the whole text of the codex was published in Arisṭūṭālīs: al-Ṭabī ʿa. Tarǧamat Isḥāq ibn

XXXVI

Introduction

being said, one must realise —without intending to belittle Badawī’s enormous achievements for (past and future) research on the Graeco-Arabic translation movement— that the edition by ʿAbd ar-Raḥmān Badawī, the most relevant one for the present study, has so many other severe deficiencies that a new edition is certainly in order. To name but a few such shortcomings: Badawī misread (or misreproduced) the text of the Arabic codex unicus at numerous places (roughly 2-3 cases per page). At hundreds of places, he modified the text (added, deleted or altered words) without indicating this in the printed text. His edition omits many marginal and interlinear glosses and variant readings or relates them to wrong sections of the main text. At numerous places, Badawī reproduced evident errors of the manuscript, although they could have been corrected easily by a comparison with the Greek text. Finally, Badawī made no use of the extensive Arabo-Latin indirect tradition, which is of crucial importance for the establishment of the text. Hence, the present edition also aims at replacing Badawī’s publication. However, for lack of time this could be realised, if at all, only for book VIII of Isḥāq ibn Ḥunayn’s translation of Aristotle’s Physics. The following considerations seem to justify a separate edition of this book: (i) Although Aristotelian scholars disagree about the chronology of the single books of the Physics, there is a broad consensus that book VIII forms within the whole a self-contained module, which may even have been conceived of by Aristotle as a separate treatise in its own right56. (ii) Book VIII is distinctive in its content. Aristotle turns here from sublunar physics and cosmology to two central issues of the whole universe: the prime mover and the eternity, continuity and unity of supralunar motion, issues, that is, with fundamental metaphysical and theological implications. Hardly surprising this book attracted an immense attention in medieval Arabic and Latin philosophy which may be illustrated representatively by the mere fact that the two most important medieval commentators, Ibn Rušd (Averroes) and Thomas Aquinas, devoted no less than one quarter of their commentaries on the _____________ Ḥunayn maʿa Šurūḥ Ibn al-Samḥ wa-Ibn ʿAdī wa-Mattā ibn Yūnus wa-Abī l-Faraǧ ibn alṬayyib. Ḥaqqaqahū wa-qaddama lahū ʿAbd ar-Raḥmān Badawī. Vols. 1—2, al-Qāhira 1384/1964—1385/1965 (henceforth: Badawī, al-Ṭabī ʿa). 56 Cf. Werner Jaeger, Aristoteles. Grundlegung einer Geschichte seiner Entwicklung. Berlin 1923, p. 314f.; Ross, Aristotle’s Physics, p. 10f.; Gérard Verbeke, “La structure logique de la preuve du Premier Moteur chez Aristote”, Revue philosophique de Louvain 46 (1948), p. 137—160, esp. p. 151; Friedrich Solmsen, Aristotle’s System of the Physical World. A Comparison with His Predecessors (Cornell Studies in Classical Philology; vol. 33), Ithaca 1960, p. 222—224, 244; Wagner, Aristoteles. Physikvorlesung, p. 275; Graham, Aristotle. Physics, Book VIII, p. xi, xiii—xiv.

I. “Aristotle’s Physics” and Aristotle’s Physics

XXXVII

Physics to this book57. (iii) The below edition avoids any idiosyncracies by scrutinising, prior to any attempt at establishing the text proper, the complete manuscript tradition —Greek, Arabic and Latin— as well as the most critical questions concerning the genesis and transmission of the Arabic translation as a whole. It thus may serve as a model or starting point for future editions of the remaining books of the Physics.

II. The Arabic manuscripts of Aristotle’s Physics The Arabic translation of Aristotle’s Physics ascribed to Isḥāq ibn Ḥunayn (d. 910) survives in almost complete form in the manuscript Leiden Or. 583 and in a short fragment in ms. Escorial ár. 896. The latter is of minor importance for the present edition of Physics bk. VIII, because it covers only fragments of the section IV 9, 217b27 — V 1, 224b10. However, as it provides a valuable instrument of comparison for the assessment of the reliability and structure of the Leiden manuscript, it deserves to be taken into consideration here as well (cf. below, chapter II.4).

II.1 Ms. Leiden Or. 583 (olim Cod. 583 Warner)58 The manuscript is stored among the Oriental Collections of the Leiden University Library. In all likelihood, it was acquired by Levinus Warner (d. 1665) between 1655 and 1665 in Istanbul. It is made of 233 folios of beige (slightly pale pink) woven paper with visible (undissolved) fibres59. Before and after the textblock proper we find two (beginning) or three (end) flyleaves of another, lighter sort of paper with visible laid lines. The codex has a golden brown wavy patterned binding with envelope flap and colourful decorated doublure. Binding and flyleaves are of a much more recent date than the textblock proper. The _____________ 57

Not to mention numerous self-contained treatises by Arabic, Latin and Hebrewwriting scholars on particular topics or sections of Physics, book VIII. 58 For previous descriptions cf. Pieter De Jong, Michael Jan De Goeje, Catalogus codicum orientalium bibliothecae Academiae Lugduno-Batavae, vol. 3, Leiden 1865, p. 310—312; Samuel M. Stern, “Ibn as-Samḥ”, Journal of the Royal Asiatic Society 88 (1956), p. 31—44; Badawī, al-Ṭabī ʿa, vol. 1, p. 26—27. 59 All information provided here is based on colour photos of the manuscript kindly put at my disposal by Professor Amos Bertolacci (Pisa). My deepest thanks go to him as well as to Professor Adam Gacek (Montreal), who kindly offered advice on some particular questions concerning this manuscript.

XXXVIII

Introduction

whole codex has been rebound and restored at an unknown point of time (certainly after its acquisition by Warner), as becomes clear from traces of mending and filling with paper strips (possibly also cellulose), cutting of the outer margins, and partly illegible notes and glosses in the inner margins, which disappear behind the gutter seam. In the middle of fol. 1a we read the title Šarḥ al-Samāʿ al-ṭabīʿī (“Commentary on the Physics”). Next to it is a seal, probably of Ottoman origin, with the text Allāhu ḥasbī waḥdahū wa-kafyī — al-Faqīr ʿUṯmān (?) * * * and the date 1021 (= 1612). The same seal is found on fol. 1a of ms. Chester Beatty Ar. 438860 (copied 910/1504) and in a number of manuscripts from Berlin, Gotha and Leipzig61. Underneath we read the handwritten ownership statement by a certain Yaʿqūb ibn al-Fahmī (or: al-Fahrī?) al-Azadī (or: al-Azrī?) dated 660/1262, and further below, possibly written by the same hand, two verses of a poem attributed to Muḥammad ibn Idrīs al-Šāfiʿī (d. 204/820)62. On the lower part of fol. 1a, which has been restored with modern paper, we read in printed letters Ex Legato Viri Ampliss. LEVINI WARNERI. Another circular seal is found at the bottom of fol. 115a. The owner of the seal calls himself ʿAbd al-Bāqī ibn ʿAlī al-ʿArabī. A man with this name is mentioned in Muḥammad Rāġib al-Ṭabbāḫ (1877—1951), Iʿlām al-nubalāʾ bitaʾrīḫ Ḥalab al-šahbāʾ 63. According to this source, ʿAbd al-Bāqī was born in 899/ 1493-4 and passed away in 971/1563-4 in Istanbul. He was appointed as Ḥanafī Qāḍī of Aleppo in the year 951/1543-4 and later occupied positions as Qāḍī in Cairo and Mekka. Perhaps it was this man who brought the codex from some central Middle Eastern place to Istanbul. _____________ 60

As Dr Moya Carey, Curator of the Islamic Collections of the Chester Beatty Library, kindly informed me, the Islamic Seals Database has been taken offline for the purpose of restructuring. The seal is stored there as seal no. 409. 61 Cf. http://www.manuscripts-gotha.uni-jena.de/receive/GothaMSSecentry_secentry_ 00000329. An image of ms. Petermann II 406 (Ahlwardt 8625), fol. 1r, which contains the same seal, is retrievable at the digital manuscript database of the Staatsbibliothek zu Berlin – Preußischer Kulturbesitz. 62 ِ ‫أﻧﺸﺪت ﺑﻴﺘﺎً وﻫﻮ ﻣﻦ‬ The verses run as follows: ‫أﻟﻄﻒ اﻟﺸﻌﺮ‬ ‫و‬ ‫أﺳﻠﺒﺖ َﻋﱪﰐ‬ ‫إذا ﻫﺠﻊ اﻟﻨﱡﻮام‬ ُ ُ

‫وﲢﺴﺐ ﻣﻦ ﻋُﻤﺮي‬ ُ ‫ﲤﱡﺮ ﺑﻼ ﻧﻔﻊ‬

ً‫أﻟﻴﺲ ﻣﻦ اﳋُﺴﺮ ِان أ ّن ﻟﻴﺎﻟِﻴﺎ‬ َ

Cf. Dīwān al-Šāfiʿī. Ǧamʿ wa-taḥqīq wa-dirāsat al-duktūr Muǧāhid Muṣṭafā Bahǧat. Damascus: Dār al-Qalam, 1420/1999, p. 67. 63 Muḥammad Rāġib ibn Maḥmūd ibn Hāšim al-Ṭabbāḫ al-Ḥalabī, Iʿlām al-nubalāʾ bitaʾrīḫ Ḥalab al-šahbāʾ, 8 vols., Aleppo: Dār al-Qalam al-ʿArabī, 21988—92, vol. 6, p. 55— 56.

II.1 Ms. Leiden Or. 583

XXXIX

The textblock proper begins on fol. 1b and ends on fol. 233b as follows: Incipit: al-ḥamdu li-llāhi l-wāḥidi l-ḥaqqi ʿalā ǧamīli naʿmāʾihī wa-l-ṣalātu ʿalā ǧamīʿi malāʾikatihī wa-anbiyāʾihī. qāla Arisṭūṭālīs lammā kānat ḥālu l-ʿilmi wal-yaqīni fī ǧamīʿi l-subuli Explicit: wa-qad bayyannā anna l-ḥarakata wāḥidatun muttaṣilatun lā waqfata fīhā Colophon (fol. 233b): tamma kitābu l-Samāʿi l-ṭabīʿiyyi maʿa taʿlīqihī. wa-lḥamdu li-llāhi waḥdahū wa-ṣalawātuhū ʿalā sayyidinā Muḥammad wa-ālihī wa-sallama taslīman. waqaʿa l-firāġu minhu awwala ḏī l-qaʿdati min sanati arbaʿin wa-ʿišrīna wa-ḫamsi miʾatin li-l-hiǧrati l-ḥanīfiyyati bi-madīnati lsalāmi. wa-kataba † ... † li-nafsihī ḥāmidan li-llāhi wa-muṣalliyan ʿalā ḫayrihī ḫalaqahū (sic cod. : ʿalā ḫayri ḫalqihī Badawī) min rusulihī According to the colophon, the copy was completed on Ḏū l-Qaʿda 1, 524 (= October 5, 1130) in Baghdad. The name of the scribe, who says that he prepared the copy “for himself”, is covered by a glued strip of paper and illegible to me. Badawī likewise left it blank in his edition64, whereas Giannakis claims to be able to read “Abū l-Ḥakam”65. The name given by Giannakis corresponds with what we read in the colophons of books I and V of the Physics, where it says indeed Abū l-Ḥakam kātibuhū (“Abū l-Ḥakam [is] its scribe”, fol. 15b) and kataba Abū l-Ḥakam al-Maġribī li-nafsihī (“Abū l-Ḥakam al-Maġribī wrote [it] for himself”, fol. 150a). The colophons on fols. 233b and 150a are definitely written by the same hand. Stern identified the scribe with the physician, geometer and poet Abū l-Ḥakam ʿUbaydallāh ibn al-Muẓaffar ibn ʿAbdallāh al-Bāhilī, born in 486/1093-4 in Yemen and called “al-Maġribī”, because his family was originally from Murcia (Spain)66. Under the Seljuq Sultan Maḥmūd II (reigned 1118—31), Abū l-Ḥakam was in charge of a travelling hospital of the Seljuq army. He died on Ḏū l-Qaʿda 4, 549 (= January 10, 1155) in Damascus. The textblock consisted originally of 17 quires each of which was made up of seven bifolios (septenions) or 14 folios. However, the first and the last quires comprise today only 13 folios. In the original binding, the first folio of the first quire, which in all likelihood had been left blank by the scribe, was used as a pastedown leaf glued onto the inner front cover. In the process of rebinding, this folio got lost together with the original bookcover. As for the last folio of _____________ 64

Badawī, al-Ṭabīʿa, p. 937.10. Elias Giannakis, Philoponus in the Arabic Tradition of Aristotle’s Physics. Unpublished doctoral thesis, University of Oxford, St. Antony’s College, 1992, p. 19. 66 Stern, Ibn as-Samḥ, p. 34. 65

XL

Introduction

the last quire, the case may be analogous, unless it was cut out immediately, because the text came to an end on the 13th folio of this quire. The beginning of a new quire is indicated by quire signatures in Eastern Arabic numerals at the top left-hand corner of the first recto of each quire. There are no catchwords, yet all quires close with a collation note on the lower right margin of the last verso. In most cases we read the formula qūbilat bi-ḥamdi llāh; quires V, XI and XII are signed qūbilat bi-ḥamdi llāhi wa-ʿawnihī; at the end of quires IV and XIV it simply says qūbilat; and in the first quire the scribe noted qūbilat bi-l-umm 67. The middle of each septenion (i.e. the double-page which prior to folding the whole quire lies on top of the pile) is additionally marked by a sign (similar to this: ᒯᒧ) at the bottom of the outer right margin of the right-hand page (i.e. the verso of the seventh folio) and at the top left-hand corner of the facing page (i.e. the recto of the eighth folio). Both marking systems date from a period prior to the rebinding, probably from the original process of copying and binding. The numerical quire signatures appear in the manuscript as follows: fol. 25a: “3” fol. 39a: “4” fol. 53a: “5” fol. 67a: “6” fol. 81a: “7”

fol. 95a: “8” fol. 109a: “9” fol. 124a: “10” fol. 138a: “11” fol. 152a: “12”

fol. 166a: “13” fol. 180a: “14” fol. 194a: “15” fol. 208a: “16” fol. 222a: “17”

Obviously, the quire signature of the second quire (to be expected on fol. 14a) is missing. Furthermore, one would expect the third quire to begin on the 28th folio of the actual binding instead of fol. 25a, the first quire comprising 13 counted folios (+ 1 pastedown leaf) and the second one 14 folios. The reason for the two phenomena is as follows: The quire signature of the second quire is missing, because the first two folios of this quire (originally the two folios following what is today fol. 13b) got lost. These two folios contained the Aristotelian text from Physics I 6, 189b28 (τούτων) up to I 8, 191a29 (ἐξ ὄντος ἢ) and the commentary related to this section68. Due to this loss, the two half sheets, which together with the two lost folios formed one bifolium each (i.e. fols. *27—*28 of the original binding), came off and sat loosely in the codex. Fortunately, they are not lost; rather, the bookbinder who completed the new binding inserted the two sheets of paper in the wrong places. What was originally fol. *27 thus became fol. 38 in the new binding, and what was originally fol. *28 is now fol. 123. However, the bookbinder somehow realised _____________ 67

For al-umm = “the exemplar” cf. below, p. XLV–XLVIII. The corresponding Aristotelian text printed in Badawī, al-Ṭabīʿa, p. 58.1—66.10 is Badawī’s own translation. 68

II.1 Ms. Leiden Or. 583

XLI

that the second quire must be incomplete. In order to fill the gap, he took another loose sheet of paper, originally the last folio of the third quire, and put it at the end of the second quire (i.e. as the 24th fol. of the new binding). The fate of the first three quires can be illustrated as follows: Original binding

Today’s binding

Septenion I

fol. *1 (originally pastedown leaf) fols. *2—*14

lost fols. 1—13

Septenion II

fols. *15—*16 (= 189b28—191a29 + comm.) fols. *17—*26 fol. *27 fol. *28

lost fols. 14—23 † fol. 38 † fol. 123

Septenion III fols. *29—*41 fol. *42

fols. 25—37 fol. 24

The following quires IV—IX, covering today fols. 39—122, are in the right order. They are followed, as mentioned above, by what was originally fol. *28, which is why the first folio of the tenth quire forms today fol. 124 (originally *123). Another confusion occurs in the eleventh quire (today fols. 138—151), the correct order of which was as follows: fols. 138—139, 141—150, 140, 151. To sum up, the original order of the textblock was as follows: Fols. 1—13; two fols. lost; 14—23; 38; 123; 25—37; 24; 39—122; 124—139; 141—150; 140; 151—233. While the above-mentioned confusion has been caused by the loss of folios and the misplacement of single sheets in the process of rebinding, we encounter another type of confusion within quire III. Here the disorder concerns sections of text which begin and end in the middle of a page. Apparently, no text got lost, yet the order of the original text must be restored, as was done correctly by Badawī69 as follows: fol. 25a—26a11 (al-ḥarakatu) ǁ 26b19 (wa-ḏālika)—27a21 (al-ǧawāb) ǁ 26a12 (ʿan li-mā)—26b19 (fī ṣūratihī) ǁ 27a21 (ʿalā qawlihī)—etc. It goes without saying that the cause of this disorder must be sought in one of the antecedents of the Leiden manuscript (probably the “Karḫ-copy”, on which see below, section II.2). In this copy, two folios were bound in reversed order, a fact that went unnoticed by the scribe of the Leiden manuscript. This antecedent _____________ 69

Badawī, al-Ṭabīʿa, p. 138—143.

XLII

Introduction

copy must have been much more comprehensive as regards the quantity of folios, as each of the two confused folios covers only slightly more text than one page of the Leiden manuscript. Ms. Leiden Or. 583 has an inconsistent Western foliation partly written with black ink, partly with blue ink, partly with a pencil. Some folio numbers are struck out and corrected. However, the struck out folio numbers do not correspond to the various cases of disorder discussed above. The foliation is consecutive up to fol. 115. From that point on, the following folios bear no folio number: 116—118, 121—122, 126—129, 131—133, 136—139, 142—144, 146—149, 156—159, 161—163, 168—169, 171, 173, 176—179, 182—183, 187—189, 191—192, 196—199, 201—204, 206—209, 211—219, 221—229, 231—232. As for the page-layout in our manuscript, we can generally distinguish between the main text and anything else written in the margins or between the lines (for the latter, cf. below, chapter II.3). There are no frames or rule-borders. The main text is written in one column of 22—23 lines. It contains —in accord with the title of the manuscript (“Commentary on the Physics”)— not only the Arabic translation of the Aristotelian text, but also comments on this work introduced by the names of their authors. The commentators most frequently mentioned are John Philoponus, Abū Bišr Mattā ibn Yūnus70, Yaḥyā ibn ʿAdī71, Abū ʿAlī ibn al-Samḥ72, and Abū l-Faraǧ ibn al-Ṭayyib73. Occasional commentary sections are ascribed to Alexander of Aphrodisias, Themistius, or a certain _____________ 70

Cf. Gerhard Endress, “Die Bagdader Aristoteliker”, in Grundriss der Geschichte der Philosophie. Die Philosophie in der islamischen Welt. Band 1: 8.—10. Jahrhundert. Herausgegeben von U. Rudolph unter Mitarbeit von R. Würsch. Basel 2012, p. 297—301; idem, “The Arabic Aristotle and the Transmission of Aristotelian Philosophy in Baghdad: Abū Bišr Mattā b. Yūnus”, in U. Rudolph, R. Hansberger, P. Adamson (eds.), Philosophy in the Islamic World. Vol. 1: 8th—10th Centuries, Leiden 2017, p. 421—434. 71 Cf. Gerhard Endress, The Works of Yaḥyā Ibn ʿAdī. An analytical inventory. Wiesbaden 1977; idem, Die Bagdader Aristoteliker, p. 301—324; idem, “Yaḥyā Ibn ʿAdī”, in U. Rudolph, R. Hansberger, P. Adamson (eds.), Philosophy in the Islamic World, p. 434— 468; Robert Wisnovsky, “New Philosophical Texts of Yaḥyā Ibn ʿAdī. A supplement to Endress̕ analytical inventory”, in Felicitas Opwis, David Reisman (eds.), Islamic Philosophy, Science, Culture, and Religion. Studies in honor of Dimitri Gutas. Leiden 2012, p. 307—326. 72 Cf. Endress, Die Bagdader Aristoteliker, p. 342—345; idem, “Ibn al-Samḥ”, in U. Rudolph, R. Hansberger, P. Adamson (eds.), Philosophy in the Islamic World, p. 490— 496. 73 Cf. Cleophea Ferrari, “Die Bagdader Aristoteliker — 6. Abū l-Faraǧ Ibn aṭ-Ṭayyib”, in U. Rudolph (ed.), Die Philosophie in der islamischen Welt. Band 1, p. 346—352; ead., “Abū l-Faraǧ Ibn al-Ṭayyib”, in U. Rudolph, R. Hansberger, P. Adamson (eds.), Philosophy in the Islamic World, p. 496—506; Gerhard Endress, “Ibn al-Ṭayyib̕s Arabic

II.1 Ms. Leiden Or. 583

XLIII

Abū ʿAmr, probably Abū ʿAmr al-Ṭabarī, a disciple of Mattā ibn Yūnus, who is known as a commentator on Alexander’s treatise on the differentia specifica74. The Aristotelian text is divided (as already in the Greek commentary tradition) into eight books. In the Arabic manuscript each book begins on a new page, independently of whether or not the text area on the preceding page has been filled with text. Within each book, the text is further divided into portions of Aristotelian text (lemmata) of varying length (from two lines up to two Bekker pages of the Greek text) followed by a commentary section which, in turn, may by composed of comments by different authors. For the transition from an Aristotelian lemma to the following comment or from a commentary section to the next Aristotelian lemma different strategies have been pursued: 1. The use of the following textual dividers: ∴ , ☉ , ○ , ዋ, or ዋዋ 75 2. Blank space (2—3 letters) before and/or after the phrase qāla Arisṭūṭālīs (“Aristotle says”, introducing the next lemma) 3. Blank space (2—3 letters) before and/or after the name of the (first) commentator. 4. The use of bold Nasḫ script for the word qāla (before Arisṭūṭālīs) with a disproportional elongation of the final horizontal part of the letter lām. _____________ Version and Commentary of Aristotle̕s De Caelo”, Studia graeco-arabica 7 (2017), p. 213—275, esp. p. 229—230. 74 Cf. Ahmad Hasnawi, “Un élève d’Abū Bišr Mattā b. Yūnus: Abū ʿAmr al-Ṭabarī”, Bulletin d’études orientales 48 (1996), p. 35–56. 75 The use of three dots arranged in the form of a triangle for the separation of single verses is already attested for manuscripts of the Qurʾān dating from the Umayyad period; the same holds true for the use of a circle (with or without central dot) for the indication of a certain cluster of verses (mostly six or ten) of the Qurʾān; cf. François Déroche, La transmission écrite du Coran dans les débuts de l’islam (Texts and Studies on the Qurʾān, vol. 5), Leiden 2009, p. 122–123; idem, Qurʾans of the Ummayads. A First Overview (Leiden Studies in Islam and Society, vol. 1), Leiden 2014, p. 46, 51, and Figures 11, 14, 32. Both dotted triangles and circles with central dot occur also in the second famous Aristotelian manuscript from the school of Baghdad, Paris BN, ar. 2436 dating from the first third of the 11th c. For dotted triangles as textual dividers in non-Qurʾanic manuscripts see also Adam Gacek, Arabic Manucripts. A Vademecum for Readers (Handbook of Oriental Studies. Section 1, The Near and Middle East, vol. 98), Leiden 2009, p. 269. A slightly different use of three dots in triangular form is found in a manuscript containing another Aristotelian text, now De Caelo, together with Ibn alṬayyib̕s commentary. The marker serves here not as divider between Aristotelian lemma and following comment, but indicates the end of each Aristotelian lemma commented upon in a parallel column which is clearly set apart from the main column containing Aristotle̕s work alone; cf. Endress, Ibn al-Ṭayyib’s Arabic Version, p. 236—237.

XLIV

Introduction

5. The use of red ink. Evidently, this has been inserted after the original process of copying, as the red ink was brushed on the dots of the marker ∴ or on the central dot in the marker ☉ after the fact. In many cases, the red-ink-hand added a horizontal stroke over the name Arisṭūṭālīs. Less frequently, it used the blank space (cf. nos. 2—3) to add one of the textual dividers (cf. no. 1) in red. 6. All kinds of combinations of the above five methods. The entire codex, main text as well as notes and glosses, is written in black ink in an old Nasḫ script with sparse (for many parts completely lacking) punctuation and even more sketchy vocalization. On the whole, it is difficult to assess with certainty whether it is written throughout by one and the same hand. Roughly from fols. 95—97 onwards the script is slightly more sloppy, twisted, and less punctuated than before, in other words inclined to what is known as Nasḫ muʿtād 76. However, I do not venture to suggest a change of hands, for the following reasons: From the subscriptions which follow after each of the eight books of the Physics with the exception of book VII, we learn that the copy was made during a travel from Khuzestan to Baghdad. The colophons mention the following places and dates: 1. Ḫūzistān bi-l-Qaṣr77, Ṣafar 1, 524 = January 13, 1130 2. Ǧundīšābūr (Gondēšāpūr), Ṣafar 22, 524 = February 3, 1130 3. ʿAskar Mukram, Rabīʿ al-awwal (no day), 524 = February/March 1130 4. Baġdād, Raǧab 28, 524 = July 6, 1130 5. Baġdād, Šaʿbān 20, 524 = July 28, 1130 6. [No place], Šawwāl (no day), 524 = September 1130 8. Baġdād, Ḏū l-Qaʿda 1, 524 = October 5, 1130 The change of the script occurs gradually in the transcription of bk. IV, thus coinciding conspicuously with two other facts, first the long pause (3—4 months) between the completion of bk. III and bk. IV, and secondly the transition from copying at stopovers of the journey to copying while being domiciled in Baghdad. Hence, we cannot rule out that this change was caused by external factors such as health impairments (therefore the long pause?), varying writing _____________ 76

For examples, cf. King Faisal Center for Research and Islamic Studies (ed.), Arabic Calligraphy in Manuscripts, Riyadh 1406/1986, p. 211—212; as well as http://makhtota. ksu.edu.sa/makhtota/6013/16#.WYQjPlFpyUk; and http://makhtota.ksu.edu.sa/ makhtota/2415/29#.WYQjuVFpyUk. 77 Stern, Ibn as-Samḥ, p. 361, identifies this place with Qaṣr al-Rūnaš in Dezfūl. One might also think of Qalʿat al-Salāsil, the fortress of Tustar (Šuštar).

II.1 Ms. Leiden Or. 583

XLV

postures, or increasing time pressure during the stay in Baghdad. What speaks in favour of changing modes of handwriting by one and the same hand are the facts that the transition is gradually and, more important, that we also encounter in the first, more neatly written part short sections copied more hastily and in a more sloppy style similar to that of the pages copied in Baghdad (e.g., fol. 14, parts of fols. 47-52) and, conversely, in the later part passages written in a style very close to the first 90 fols. (e.g., fols. 99a, 144b, 162b, 173b). Furthermore, there are no significant disparities in particular letter forms or other characteristics that might be unambiguously attributed to different hands. And finally, the two colophons on fols. 15b and 150a, that is before and after the gradual change of the script, mention one and the same person as the scribe of the manuscript. To sum up, the question of the number of hands is difficult. There are certain indicators pointing to the involvement of more than one hand, while others point to the opposite. In what follows, references to the scribe of our manuscript must be taken as implying this uncertainty. Closely related to the above problem is the fact that the text of the manuscript has been corrected at a number of places. In some cases, the corrections are written immediately in the text, i.e. without erasure in bold script onto the word or letter to be corrected, but in other cases they are superscript over the relevant word. Deletions are either marked by superscript horizontal strokes or by crossed out words. Corrections are not restricted to the main text, but also found in the marginal scholia, i.e. correcting the text of the marginal note itself. It is not possible to determine whether they originate (partly or altogether) from the hand of the main scribe.

II.2 The “prehistory” of ms. Leiden Or. 583 The subscriptions, collation notes and other details of the Leiden manuscript reveal important information about its antecedents and the sources of the compilation composed of Aristotelian text and commentary sections. About the immediate ancestor of the Leiden manuscript we learn from the colophon attached by the scribe, Abū l-Ḥakam al-Maġribī, to bk. I of the Physics 78: “I have reproduced all this in the way it is [written] in the manuscript which was copied from ‘the original’ (al-aṣl) in al-Karḫ in Ǧumādā al-aḫīr (sic) of the year 470. The only thing I changed is the date [given] in accordance with the present copy. Not a single letter has been added or omitted. Hence, if anybody reads the present [copy], it is as if he were reading that one, I mean _____________ 78

Fol. 15b, ed. Badawī, al-Ṭabīʿa, p. 77.

XLVI

Introduction

‘the exemplar’ (al-umm) which was copied from ‘the author’s original’ (aṣl almuʾallif ).” Hence, ms. Leiden Or. 583 was copied from a manuscript prepared in December 1077 or January 1078 in al-Karḫ. “Al-Karḫ” means “fortress”, “fortified town”, and as to be expected, there is a number of places called “al-Karḫ” in the eleventh century, most notably in or near Baṣra, Sāmarrā and Baghdad79. There is good reason to assume that it is al-Karḫ in Baghdad that Abū l-Ḥakam is referring to here. We know that “the author’s original” (aṣl al-muʾallif ) which served as the exemplar of what might be called the Karḫ-copy, is nothing other than the authoritative compilation prepared by the young Abū l-Ḥusayn (or: Abū lḤasan) Muḥammad ibn ʿAlī ibn al-Ṭayyib al-Baṣrī80, who spent his life in Baghdad (where he died —having achieved some fame as a Muʿtazilite theologian— in October 1044). In all likelihood, the first copy of his autograph was prepared in the milieu of his disciples. Two of his first generation disciples, Abū ʿAlī Muḥammad ibn al-Walīd al-Karḫī and Abū l-Qāsim ibn al-Tabbān, worked as teachers of logic, medicine and Muʿtazilite theology during the second half of the 11th century in Baghdad81, and Ibn al-Walīd (d. 478/1086) is even reported to have spent the last five decades of his life there in the district called “alKarḫ” 82. Apart from the date and place of copying we have no further information about the ancestor of the Leiden manuscript, except that Abū l-Ḥakam checked his copy once again against his exemplar as becomes clear from collation notes and corrections referring explicitly to al-umm 83. However, at least one marginal _____________ 79

Cf. Yāqūt al-Ḥamawī, Muʿǧam al-buldān, 5 vols., Bayrūt: Dār Ṣādir 1397/1977, vol. 4, p. 447—449, al-Ṭabarī, History, Index, vol. 10, p. 269. 80 On him cf. Stern, Ibn as-Samḥ, p. 36—38, Wilferd Madelung, “Abū l-Ḥusayn al-Baṣrī”, in: Madelung, Wilferd, “Abū l-Ḥusayn al-Baṣrī”, in: Encyclopaedia of Islam, THREE, ; Sabine Schmidtke, “The Muʿtazilite Movement (III): The Scholastic Phase”, in S. Schmidtke (ed.), The Oxford Handbook of Islamic Theology, Oxford 2016, p. 159—180, esp. p. 169—175. 81 Among their disciples were the physician Ibn Ǧazla and the theologian Ibn ʿAqīl; cf. Šams al-Dīn al-Ḏahabī, Sīrat aʿlām al-nubalāʾ, ed. Šuʿayb al-Arnaʾūṭ, Muḥammad Nuʿaym al-ʿArqasūsī, 23 vols., Bayrūt 111417/1996, vol. 17, p. 587—588; ʿAbd al-Raḥmān ibn Raǧab, K. al-Ḏayl ʿalā ṭabaqāt al-ḥanābila, ed. ʿAbd al-Raḥmān ibn Sulaymān alʿUṯaymīn, 5 vols., Mekka 1425/2005, vol. 1, p. 319, 322; William Montgomery Watt, Muslim Intellectual: A Study of al-Ghazzali, Edinburgh 1963, p. 32, 107. 82 Cf. Schmidtke, Muʿtazilite Movement (III), p. 172. 83 Cf. Badawī, al-Ṭabīʿa, p. 992, 2092, further notes referring to al-umm occur in the margin of fol. 6b, at the bottom of fol. 50a, and in the margin of fols. 64b, 95b, 144b (all omitted by Badawī).

II.2 The “prehistory” of ms. Leiden Or. 583

XLVII

note (related to Physics VIII 9, 265b24-28) suggests that Abū l-Ḥakam had access to a further copy of the Baghdad compilation. This note —originating almost certainly from Abū l-Ḥakam’s hand— refers explicitly to another manuscript (nusḫa uḫrā) and reports a phrase of the Aristotelian text omitted by homoioteleuton within the Arabic transmission (for more details see below, p. L—LI). Unless Abū l-Ḥakam uses nusḫa uḫrā here as a synonym for al-umm, there must have been another copy at his disposal. That Abū l-Ḥusayn al-Baṣrī is the author/editor of “the original” (al-aṣl ), i.e. the one who arranged the compilation made up of lemmata of the Aristotelian text together with comments of the Baghdad Aristotelians, follows from the colophons of bks. I and IV of the Physics: Fol. 15b (end of bk. I): [a] “At this place, there is [the following] note in the handwriting of the Šayḫ Abū l-Ḥasan [scil. Abū l-Ḥusayn al-Baṣrī], may God have mercy on him: ‘I finished copying and annotating (taʿlīq) it [scil. bk. I] in Ṣafar of the year 395 [= November—December 1004], may God bless the Prophet Muḥammad and his family.’” “And on the back of the first and the second quires [it reads] in his [scil. Abū l-Ḥusayn al-Baṣrī’s] handwriting, may God have mercy on him: ‘I have collated Yaḥyā ibn ʿAdī’s copy with the text [contained] in this quire. [Yaḥyā ibn ʿAdī] mentioned (ḏakara) that he had transcribed it [scil. his copy] from the dastūr 84 of Isḥāq[̕s translation] and collated it with this three times, and a fourth time with the Syriac. Corrections and annotations (taʿlīq) found in the margins of this quire together with the mark ‫ ح‬are taken from Yaḥyā [ibn ʿAdī’s] copy.’” [b] “And on the back of the first quire [it reads]: ‘The first quire of Aristotle’s Physics translated by Isḥāq ibn Ḥunayn. Muḥammad ibn ʿAlī al-Baṣrī studied it under Abū ʿAlī al-Ḥasan ibn al-Samḥ. It contains his [scil. Abū l-Ḥusayn al-Baṣrī’s] annotation (taʿlīquhū ms. : taʿlīqatun ed. Badawī) [as taken] from him (ʿanhu, scil. Ibn al-Samḥ).’” “And on the back of the second [quire it reads]: ‘The second [quire] of Aristotle’s Physics translated by Isḥāq ibn Ḥunayn, including annotations (taʿlīq) from Abū ʿAlī al-Ḥasan ibn _____________ 84

On this term see below, p. LXVIII f.

XLVIII

Introduction

al-Samḥ. Muḥammad ibn ʿAlī al-Baṣrī took down the annotations (ʿallaqahū) from him.’” “On the back of the third [quire], he adds: ‘[Annotated] also from the words (kalām) of Mattā [ibn Yūnus].’ “And in the fourth [quire], he adds: ‘[Annotated] also from the words (kalām) of Abū Bišr Mattā [ibn Yūnus] and the words of Yaḥyā.’” “And in the fifth [quire]: ‘[Annotated] also from the words (kalām) of Yaḥyā and Abū Bišr Mattā [ibn Yūnus].’”85 Evidently, the indented passages in section [a] quote Abū l-Ḥusayn al-Baṣrī speaking in the first person. The one who took down these statements by Abū l-Ḥusayn al-Baṣrī indicating the places of the latter’s notes in his exemplar must be the scribe of the Karḫ-copy (al-umm). Whether the same holds true for the indented passages in section [b] is not quite clear, as they refer to Abū l-Ḥusayn al-Baṣrī in the third person. Possibly these notes were phrased by the scribe of the Karḫ-copy, while the information proper contained in them must trace back to Abū l-Ḥusayn al-Baṣrī himself86. The honorific title Šayḫ supports my assumption that the Karḫ-copy was prepared by a pupil or follower of Abū lḤusayn al-Baṣrī. Similarly, we read at the end of bk. IV (fol. 113b): “[Here] ends the annotation (taʿlīq) of the fourth book of the Physics by the Šayḫ and learned Imām Abū l-Ḥusayn Muḥammad ibn ʿAlī al-Baṣrī.” Thus, we learn about Abū l-Ḥusayn al-Baṣrī’s autograph that it was undertaken in the year 1004 and that it was divided into separate quires. The Šayḫ himself depicts his work as “copying and annotating” (nasḫ wa-taʿlīq), where it is a matter _____________ 85

English translations of Arabic, Greek or Latin quotations are mine unless other translators are indicated. 86 Paul Lettinck, Aristotle’s Physics and its Reception in the Arabic World, with an Edition of the Unpublished Parts of Ibn Bājja’s Commentary on the Physics (Aristoteles SemiticoLatinus, vol. 7), Leiden 1994, p. 4, interprets the references to Ibn al-Samḥ in this section of the colophon in the sense that Abū l-Ḥusayn al-Baṣrī “had a copy of Isḥāq’s translation by Ibn as-Samḥ” at his disposal. In my view, this is by no means deducible from the wording of the colophon. If anything, we may speculate whether this might apply to the first two quires, i.e. up to Physics II 2, 194a12 (see below). However, even for this part “taking down Ibn as-Samḥ’s annotations” does not necessarily presuppose that Abū l-Ḥusayn drew on Ibn as-Samḥ’s own annotated copy (see what follows).

II.2 The “prehistory” of ms. Leiden Or. 583

XLIX

of speculation in which way the two terms relate to his handling of (written and oral?) sources. Interestingly, the annotations drawing on the teaching of Abū Bišr Mattā ibn Yūnus, who had passed away long before Abū l-Ḥusayn alBaṣrī’s student career in Baghdad, are said to be taken from his words (kalām), whereas the annotations of his teacher Abū ʿAlī ibn al-Samḥ are simply said to be taken “from him” (ʿanhu). As already observed by Giannakis87 this may point to Ibn al-Samḥ as an oral source as opposed to written sources by the former. As regards the written sources by Abū Bišr Mattā ibn Yūnus and Yaḥyā ibn ʿAdī, we can infer indirectly from Abū l-Ḥusayn al-Baṣrī’s notes that “Yaḥyā ibn ʿAdī’s copy” (henceforth: Y), from which Abū l-Ḥusayn al-Baṣrī transcribed corrections and annotations which he marked with the letter ‫ح‬, was a copy of Isḥāq ibn Ḥunayn’s translation with rather brief remarks and corrections by Yaḥyā ibn ʿAdī. Since the Leiden manuscript also contains comprehensive doctrinal comments by Yaḥyā ibn ʿAdī, Abū l-Ḥusayn al-Baṣrī must have had access to other written materials of his. In any case, Y did not serve as Abū lḤusayn al-Baṣrī’s primary source for the transcription of the Aristotelian text. This is clear not only from the fact that Abū l-Ḥusayn mentions explicitly having collated his copy with Y, but even more conclusively from a number of marginal notes which obviously result from this collation and point to variant readings in Y which were nonetheless rejected by Abū l-Ḥusayn al-Baṣrī88. Fortunately, the scribe of the Karḫ-copy recorded (with three exceptions) the beginning or end of all quires of Abū l-Ḥusayn al-Baṣrī’s autograph89, and even the scribe of the Leiden manuscript took these notes over into his copy. That we are dealing with the quires of Abū l-Ḥusayn al-Baṣrī’s autograph and not, as one might suspect, with those of the Karḫ-copy becomes clear from a quire mark referring explicitly to “the quires of the Šayḫ Abū l-Ḥasan”90, while the other quire marks refer to “the quires in his [scil. Abū l-Ḥusayn al-Baṣrī’s] handwriting”. Due to these notes we know that Abū l-Ḥusayn al-Baṣrī’s autograph was composed of twenty-two quires (each covering about 40—50 pages of Badawī’s edition) corresponding to the following parts of the compilation: _____________ 87

Cf. Giannakis, Philoponus, p. 30. Cf. Badawī, al-Ṭabī ʿa, p. 185, 301, 7732, 7743, etc. 89 For a list of all notes indicating the end or beginning of the quires cf. Giannakis, Philoponus, p. 243. The missing quire marks concern the points of transition from quire IV to V, IX to X, and XVI to XVII. It is not clear whether they were omitted by the scribe of the Leiden manuscript or were already absent from his exemplar. 90 Cf. Badawī, al-Ṭabī ʿa, p. 3232. 88

L

Quire

Introduction

Aristotelian text + Commentary

1 2 3 4

I 1, 184a10 — I 4, 188a18 (post comm.91) I 5, 188a19 — II 2, 194a12 (post comm.) II 2, 194a12 — II 8, 199a13 (post comm.) II 8, 199a13 — ? [probably III 2, 202a12]

5

? [probably III 3, 202a13] — III 5, 205a7 (intra comm.) III 5, 205a7 — III 8, 208a23 (post comm.) IV 1, 208a27 — IV 4, 212a7 (post comm.) IV 4, 212a7 — IV 8, 216a11 (intra comm.) IV 8, 216a11 — ? [probably IV 9, 217b28]

6 7 8 9

11

? [probably IV 10, 217b29] — IV 12, 220b32 (intra comm.) IV 12, 220b32 — V 1, 224b35 (intra comm.)

12 13 14 15 16

V 1, 225a1 — V 3, 227a13 (post comm.) V 3, 227a13 — V 6, 230a18 (intra comm.) V 6, 230a18 — VI 2, 233a12 (post comm.) VI 2, 233a13 — VI 4, 235b5 (post comm.) VI 5, 235b6 — ? [probably VI 8, 239b4]

17 18 19

? [probably VI 9, 239b5] — VII 2, 245b18 (intra comm.) VII 3, 245b19 — VII 4, 249b19 (intra lemma) VII 4, 249b19 — VIII 2, 253a21 (post comm.)

20 21 22

VIII 3, 253a22 — VIII 5, 258b9 (intra comm.) VIII 6, 258b10 — VIII 8, 264a8 (post lemma) VIII 8, 264a8 — end.

10

Ed. Badawī p. 1.1—42 ult. p. 43.1—93.11 p. 93.12—148 ult. p. 149.1—? [probably p. 190 ult.] ? [probably p. 191.1]— p. 232.19 p. 232.19—270.20 p. 271.1—323 ult. p. 324.1—372.3 p. 372.4—? [probably 403 ult.] ? [probably p. 404.1]— p. 444.8 p. 444.9—501.20 (om. Bad.) p. 501.21—546.2 p. 546.3—587 ult. p. 588.1—625 ult. p. 626.1—665 ult. p. 666.3—? [probably 710 ult.] ? [probably p. 711.3]— p. 756.11 (om. Bad.) p. 756.12—790.2 p. 790.3—822 ult. (om. Bad.) p. 823.1—864.10 p. 864.11—904 ult. p. 905.1—end.

_____________ 91

In this overview “post comm.” indicates that the end of a quire coincides with the end of a commentary section (with the next quire starting with an Aristotelian lemma), “intra comm.” marks a quire ending in the middle of a commentary section (with the next quire resuming the commentary before moving on to the next Aristotelian lemma), “intra lemma” points to a quire ending in the middle of an Aristotelian lemma, “post lemma” to a quire ending after an Aristotelian lemma and before a commentary section. “Om. Bad.” indicates that the quire mark is extant in the manuscript, yet not reproduced in Badawī’s edition. The precise wording of each quire mark is also quoted in Elias Giannakis, “The Structure of Abū l-Ḥusayn al-Baṣrī’s Copy of Aristotle’s Physics”, Zeitschrift für Geschichte der Arabisch-Islamischen Wissenschaften 8 (1993), p. 251—258, esp. 253—254.

II.2 The “prehistory” of ms. Leiden Or. 583

LI

Evidently, Abū l-Ḥusayn al-Baṣrī set no great value on arranging his compilation in such a way that each quire started with a new Aristotelian lemma (only ten of the twenty-two quires do so). Fourteen quires either began or ended in the middle of a commentary section; two even began or ended in the middle of an Aristotelian lemma. It is a matter of speculation why Abū l-Ḥusayn did not choose Y as his main model. The rejected variant readings of Y may indicate that he did not deem it reliable enough, unless the reason is to be sought in the methodology of his compilation, such that he preferred to work with sources comprising both the Aristotelian text and comprehensive commentary sections. Similarly, we do not know whether it was Mattā ibn Yūnus’ or Ibn al-Samḥ’s materials that served as Abū l-Ḥusayn’s basic source for the Aristotelian text or whether in general he had at his disposal a separate copy of the Arabic translation apart from Y and other written materials of the Baghdad Aristotelians. Various indications suggest that Abū l-Ḥusayn al-Baṣrī worked with different groups of sources or switched his exemplars during the course of his compilation. The most conspicuous shift concerns the transition from Physics VI 5 to VI 6 (in the middle of quire XVI). Up to bk. VI 5 the main bulk of the commentary sections draws (apart from John Philoponus) on materials of Abū Bišr Mattā ibn Yūnus, Yaḥyā ibn ʿAdī and Abū ʿAlī ibn al-Samḥ, but from bk. VI 6 on Abū l-Ḥusayn al-Baṣrī quotes almost exclusively Abū l-Faraǧ ʿAbdallāh ibn al-Ṭayyib (d. 435/1043)92. The famous Nestorian physician and philosopher was a pupil of the Baghdad Aristotelians ʿĪsā ibn Zurʿa and al-Ḥasan ibn Suwār ibn al-Ḫammār and, like Ibn al-Samḥ, a teacher of Abū l-Ḥusayn al-Baṣrī93. The reason for this shift is not known. Since the compilation related to Physics VI 6 — VIII 10 draws not only on Ibn al-Ṭayyib’s explanations, but also on earlier materials inaccessible to Abū l-Ḥusayn al-Baṣrī otherwise than by means of written sources, there can be little doubt that he switched between his sources, and this independently of the arrangement of his materials into separate quires94. Yet, the Leiden manuscript provides further signs pointing in this direction: (i) The manuscript contains altogether 18 marginal and interlinear notes beginning with “Isḥāq” 95. There can be little doubt that this labelling points to _____________ 92

Cf. Giannakis, Philoponus, p. 25—27, 34—46, 65—66, 132—151. Cf. Ibn Abī Uṣaybiʿa, ʿUyūn al-anbāʾ fī ṭabaqāt al-aṭibbāʾ, ed. August Müller, Cairo 1299/1882, 2 vols., vol. 1, p. 239—241. 94 That it was indeed Abū l-Ḥusayn al-Baṣrī and not the scribe of the Karḫ-copy who switched exemplars at this place has been shown by Giannakis, The Structure of Abū lḤusayn al-Baṣrī’s Copy, p. 255—257. 95 On which cf. below, ch. III. 93

LII

Introduction

Isḥāq ibn Ḥunayn as the author of the following note. Conspicuously, 17 out of the total of 18 notes labelled with “Isḥāq” are related to sections pertaining to books V 1 — VI 5 of the Physics. That Isḥāq supplied his translation with his own brief annotations only in this section of the Aristotelian text to the exclusion of the other books is highly improbable. The reason for the present phenomenon must rather be sought in the circumstances of the transmission of his translation. That the record of these notes comes to an end with Physics VI 5 can be easily related to the above-mentioned switch from materials derived from Abū Bišr Mattā ibn Yūnus, Yaḥyā ibn ʿAdī and Abū ʿAlī ibn al-Samḥ to those based mainly on the authority of Abū l-Faraǧ ibn al-Ṭayyib. Since the notes labelled with “Isḥāq” occur almost exclusively in the margins or between the lines of the Aristotelian text, we are quite safe in assuming that Abū l-Ḥusayn al-Baṣrī switched at this point not only his commentary sources, but together with them also his exemplars of Isḥāq’s translation of Aristotle’s text. The exemplar he used from bk. VI 6 onwards together with the commentary by his teacher Ibn al-Ṭayyib evidently no longer reproduced Isḥāq’s marginal remarks. For an alternative explanation one would have to suppose that the inconsistent occurrence of the notes labelled with “Isḥāq” is due to Abū l-Ḥusayn al-Baṣrī’s commentary sources in such a way that he had extracted these notes from separate commentary materials and placed them (always correctly!) in the margin or between the lines of the Aristotelian text. In view of the fact that most of these notes consist of Isḥāq ibn Ḥunayn’s explanation alone, i.e. without repetition of the Aristotelian word or phrase they are related to, this assumption is extremely unlikely96. As regards the fact that (with a single exception) no note labelled with “Isḥāq” has been recorded in bks. I—IV of the Physics, see what follows. (ii) Some parts of the compilation are divided and headed by numbered “lectures” (taʿālīm), others are not. The distribution of these two layouts coincides conspicuously partly with the occurrence of notes labelled with “Isḥāq”, partly with the transition from Physics VI 5 to VI 6. The parts divided into “lectures” are bks. III—IV and VI 6 — VIII 1097; in other words, the section preserving Isḥāq’s notes lies exactly between two parts divided into “lectures”. In _____________ 96

For a third hypothesis according to which Abū l-Ḥusayn took the notes from Y, see below. 97 The division of chapters 6—10 of bk. VI into “lectures” is not marked by separate headings. However, there can be no doubt that Abū l-Faraǧ ibn al-Ṭayyib had such a division at least in mind, if not noted down in his exemplar, as he refers to the sections he comments upon as taʿlīm (cf. Badawī, al-Ṭabī ʿa, p. 686.4, 718 ult., 728.5). On the division of text + commentary into taʿālīm as “a hallmark of his method” cf. Endress, Ibn al-Ṭayyib’s Arabic Version, p. 230.

II.2 The “prehistory” of ms. Leiden Or. 583

LIII

all likelihood, the part of the compilation related to Physics V 1 — VI 5 is based on written materials tracing back to another Baghdad authority than the preceding and following parts. (iii) Regarding the comments attributed to “Yaḥyā” (Yaḥyā ibn ʿAdī or John Philoponus, occasionally Ibn ʿAdī summarizing Philoponus) and Abū ʿAlī ibn al-Samḥ we observe a significant change of their arrangement in the commentary sections in bks. I—II as opposed to bks. III—VI 5. Ibn al-Samḥ’s comments on bks. I—II are generally rather comprehensive and self-contained discussions of the preceding section of the Aristotelian texts, while this part contains only one comment attributed to “Yaḥyā” (ed. Badawī, p. 148). On the other hand, comments attributed to “Yaḥyā” are ubiquitous in bks. III—VI 598, whereas Ibn al-Samḥ’s comments are here rather short and in many cases refer not to the Aristotelian text, but rather (as a sort of super-commentary) to “Yaḥyā’s” comment quoted previously. With few exceptions (ed. Badawī, p. 428—431, 441) bks. III—VI 5 do not contain any comment by Ibn al-Samḥ that is not accompanied (in most cases preceded) by a comment attributed to “Yaḥyā”. (iv) The Leiden manuscript contains also minutes taken down during class meetings in Baghdad99. Although Ibn al-Samḥ evidently concerned himself with all parts of the Physics from the beginning up to bk. VI 5, it is only in bks. III— VI 5 that we encounter such notes of classroom conversations (probably noted down by Abū l-Ḥusayn al-Baṣrī) of the pattern qultu li-Abī ʿAlī (in one case saʾaltu Abā ʿAlī) ... fa-qāla ..., i.e. “I said to/asked Abū ʿAlī [ibn al-Samḥ] ... and he replied ...”100. Taking all this evidence together we may draw the conclusion that Abū l-Ḥusayn al-Baṣrī switched more than once the exemplars (henceforth: θ1, θ2, etc.) for his compilation. (θ1) In all likelihood he used for bks. I—II a compilation that originated from a period prior to Ibn al-Samḥ’s and Ibn al-Ṭayyib’s teaching activities, as this part contains no classroom minutes of questions directed to them, but only those addressed to Abū Bišr Mattā. (θ2) The source he used for bks. III—IV was divided, unlike the preceding part, into lectures. The intense consideration of Yaḥyā ibn ʿAdī’s thoughts on place (Physics IV 4) in this part of the compilation (ed. Badawī, p. 316—334101) suggests that its formation is closely _____________ 98

Cf. Giannakis, Philoponus, p. 152—178. Cf. Giannakis, Philoponus, p. 44—46. 100 Badawī, al-Ṭabī ʿa, p. 236, 277, 314, 432, 447, 521, 525, 588, 624, 650. 101 Cf. Giannakis, Philoponus, p. 38, notes 1 and 3; Endress, The Works of Yaḥyā Ibn ʿAdī, p. 38. 16 out of the total of 17 comments which can be attributed beyond doubt to Yaḥyā ibn ʿAdī occur in Physics bk. IV. 99

LIV

Introduction

linked to Yaḥyā ibn ʿAdī and his teaching and writings, and that the latter might be responsible for its division into lectures. On the other hand, it is worth noticing that this part contains not a single comment by Yaḥyā ibn ʿAdī related to bk. III, although he evidently dealt with this book in separate treatises102. Hence, we must take into consideration the possibility that part θ2 derives ultimately from two distinct partial compilations θ2.a (= bk. III) and θ2.b (= bk. IV), especially as two other phenomena underline the exceptional status of bk. IV: (i) it is only in this book that we encounter marginal notes referring to „some/many manuscripts“ (baʿḍ al-nusaḫ / kaṯīr min al-nusaḫ, cf. below, chapter II.3); (ii) from bk. IV onwards (up to the end of bk. VII) the commentary sections are frequently interspersed with quotations (lemmata) of the Aristotelian text introduced by the formula qāla Arisṭū or qāla Arisṭūṭālīs (cf. below, chapter VII.1). This phenomenon is entirely absent from bks. I-II and VIII and occurs only twice in bk. III. (θ3) Bks. V 1 — VI 5 of Abū l-Ḥusayn al-Baṣrī’s compilation must trace back to yet another source. This becomes clear from (i) the careful reproduction of Isḥāq ibn Ḥunayn’s notes in this part alone, (ii) the absence of any division into lectures as provided in the preceding and in the following parts, and (iii) the fact that this part contains significantly more commentary sections introduced by (qāla) Yaḥyā wa-Abū ʿAlī (“John [Philoponus] and Abū ʿAlī [ibn al-Samḥ] say”) than the preceding part of the compilation103. Unlike part θ2, this section presumably did not derive from Yaḥyā ibn ʿAdī’s materials. As we know from the colophon of bk. I of the Physics (cf. above), Abū l-Ḥusayn al-Baṣrī collated his transcription with Yaḥyā ibn ʿAdī’s copy and noted down “corrections and annotations” (iṣlāḥ wa-taʿlīq) taken from the latter in the margins of his copy with the sign ‫ح‬. Notes of this type are found in all parts of the compilation and for all books of the Aristotelian text, which means that Abū l-Ḥusayn must have had access to the whole translation as preserved in Y. Hence, if this copy had served as Abū l-Ḥusayn’s source for the transcription of Isḥāq ibn Ḥunayn’s own marginal notes, one would expect to encounter notes labelled with “Isḥāq” not only in part θ3, but equally distributed over all parts of the compilation. However, parts θ2 and θ3 share the peculiarity of incorporated classroom minutes with Abū ʿAlī ibn al-Samḥ as reference person and respondent. (θ4) Finally, we have from bk. VI 6 onwards the switch to the commentary by Abū l-Faraǧ ibn al-Ṭayyib described above. _____________ 102

Cf. Endress, Die Bagdader Aristoteliker, p. 310f. Cf. Giannakis, Philoponus, p. 152—178. Parts θ1 and θ4 do not contain any comment of this type.

103

II.2 The “prehistory” of ms. Leiden Or. 583

LV

Thus, the sources of Abū l-Ḥusayn al-Baṣrī’s compilation reflect in a certain way the different stages of the reception and commentation of Aristotle’s Physics among the subsequent generations of Baghdad Aristotelians, which he unified and “updated” as matters stood in the year 395/1004. The evidence above suggests that Abū l-Ḥusayn probably relied for the inclusion of the Aristotelian text translated by Isḥāq in his compilation not on a separate copy, but rather on these different sources which already included the Aristotelian text together with other materials. That all parts of his compilation reproduce Isḥāq ibn Ḥunayn’s translation in an equally meticulous way without imbrications or substantial lacunae is certainly to Abū l-Ḥusayn’s credit. However, he could not do better than his sources. Unfortunately, three of the four sources at his disposal did not reproduce the translator’s own notes and bits of advice for the reader.

II.3 The marginal and interlinear notes and glosses in ms. Leiden Or. 583 The Leiden manuscript contains more than (!) 1.000 marginal and interlinear notes which for the analysis of the composition and history of the codex must be carefully kept apart from the integral commentary sections104. Since most notes seem to be written by the hand of the main copyist, we must take into consideration that many of them derive from an early stage of transmission of Isḥāq ibn Ḥunayn’s translation. Roughly 30—40% of the marginal notes are marked with various kinds of signs corresponding to superscript signs in the main text at the place to which the note refers. The marks most frequently used are as follows (roughly in order of frequency)105:

+ ∴ ⟇

￮—

⊣

⊢

⚯

⩊

⩛

ዋ

♊

￮

‫ﻌ‬

ᓀ

Some of these signes-de-renvoi also occur in the sections of ms. Paris BnF ar. 2346 covering Porphyry’s Isagoge and Aristotle’s Categories, De interpretatione, Prior and Posterior Analytics (significantly less frequently in the other texts of this codex). However, in proportion to the great variety of marks the conformity is rather limited. More interestingly, the use of some of these marks is also attested in Greek manuscripts contemporary with the Graeco-Arabic translation movement (such as mss. Vind. gr. 100, Pal. Soc. I 231, Vat. gr. 2066 [all 9th c.], _____________ 104

The notes certainly deserve an in-depth study with regard to the prevalent philological ambitions and teaching practices among the Baghdad Aristotelian which goes far beyond the scope of the present edition. 105 Only three of these marks are registered in A. Gacek, Arabic Manuscripts, p. 250f.

LVI

Introduction

Ambros. L 93 sup., Ven. Marc. 201 [both 10th c.])106. E.g., ∴ (Shiel107, p. 328f. [IIIc]) , ￮— (Shiel, p. 328, 333 [XIIIf], Atsalos, p. 229 [2.2]), ⚯ (Atsalos, p. 226 [25], p. 228 [53a], ⩊ (Shiel, p. 328, 334 [XIVe], Atsalos, p. 226 [10]), ዋ (Atsalos, p. 228 [56a]), ♊ (Shiel, p. 328, 334 [XVi]), and ᓀ (Atsalos, p. 229 [2.1]). Other notes are connected to the word or line they relate to by solid or dotted lines, while the remaining notes (about 50%) bear no marks at all as to their object of reference. As mentioned above, the codex has been rebound and the margins of the pages have been cut at some time or another. As a consequence, some of the notes at the outer margins and many notes in the inner gutter margins are partly illegible. The notes are written in horizontal, vertical and diagonal directions. They occur both next to the Aristotelian text and in the margins or between the lines of commentary sections. The contents and comprehensiveness of the notes range from single words (i.e. words omitted in the main text or words supposed to explain or replace another word in the main text) through brief explanations (“glosses” in the narrow sense of the word, often introduced by yaʿnī or ay) up to elaborate remarks on doctrinal questions and problems of the textual transmission (“scholia”)108. Regarding the provenance or sources of the notes we can distinguish generally between four different types, three of which provide certain information about their origin: (i) Notes referring explicitly to somebody’s translation. The key term of these notes is naql, which obviously refers to some form of translation from one language into another (no matter whether [complete or partial] translation proper, paraphrase, epitome, or whatever it may be)109. Nothing is said about _____________ 106

This phenomenon has already been noted by Marwan Rashed (Alexandre d’Aphrodise, p. 22) for glosses on De interpretatione in the Paris manuscript. 107 For signes-de-renvoi in early Byzantine manuscripts see James Shiel, “A Set of Greek Reference Signs in the Florentine MS. of Boethius’ Translation of the Prior Analytics (B.N. Conv. Soppr. J. VI. 34)”, Scriptorium 38 (1984), p. 327—342; Basile Atsalos, “Les signes de renvoi dans les manuscrits grecs”, in D. Harlfinger, G. Prato, M. D’Agostino, A. Doda (eds.), Paleografia e codicologia greca. Atti del II. Colloquio internazionale (BerlinoWolfenbüttel, 17-21 ottobre 1983), 2 vols., Alessandria 1991, vol. 1, p. 211—231. N.B.: Atsalos reproduces tables of signes-de-renvoi compiled by S. Lambros and G. Charitakis (v. Atsalos, p. 213, note 3). 108 For a general appreciation and systematisation of marginal and interlinear notes in manuscripts see the “Introduction” in René Nünlist, The Ancient Critic at Work. Terms and Concepts of Literary Criticism in Greek Scholia. Cambridge 2009, p. 1—19. 109 Naql can also mean “transcription”. However, all notes of the Leiden manuscript employing this term seem to refer to a translation in the proper sense, with one exception: At the end of bk. I of the Physics, we read in the margin next to the colophon the note nuqila min ḫaṭṭ al-Šayḫ raḥimahu llāh (“transcribed from the handwriting of the Šayḫ, may God have mercy on him”). This notes repeats the information contained in

II.3 Marginal notes and glosses in ms. Leiden Or. 583

LVII

the exact contents of the translated text referred to, whether this was Aristotle’s Physics itself or some sort of commentary or paraphrase comprising quotations of the Aristotelian text, nor about the source and target languages of the translation. Notes of this type either mention the author of the version, to be more precise “Isḥāq” [ibn Ḥunayn], “Qusṭā” [ibn Lūqā] (d. 912), [Abū ʿUthmān] “al-Dimašqī” (d. after 914) and, in one case, “Yūḥannā”, or they refer anonymously to “another translation” (naql āḫar). The notes referring to “Isḥāq’s translation” (naql Isḥāq) evidently relate to the translation transmitted in full in ms. Leiden Or. 583 (cf. below, chapter III). Qusṭā ibn Lūqā and Abū ʿUthmān al-Dimašqī are mentioned by the bibliographers in the context of Aristotle’s Physics. However, the information provided is rather vague and fragmentary. Apparently, both translators were involved in translations of parts of the commentary on the Physics by Alexander of Aphrodisias110, while Qusṭā ibn Lūqā is also mentioned as translator of (parts of) John Philoponus’ commentary111. Neither of them is said to have translated the Aristotelian text proper. In all likelihood, the notes in ms. Leiden Or. 583 referring to “Qusṭā” or “al-Dimašqī” are based on a collation of the Aristotelian lemmata in these commentaries. The unique note beginning with fī naql Yūḥannā seems to refer to an alternative translation (or rather paraphrase) of Physics III 5, 205a7-8: δεῖ δὲ κατὰ παντὸς ἐκ τῶνδε σκοπεῖν, εἰ ἐνδέχεται ἢ οὐκ ἐνδέχεται εἶναι This sentence is difficult due to Aristotle’s brachylogy and prompted the Greek scribes to replace the absolute εἶναι with different elaborations, e.g. εἶναι σῶμα ἄπειρον αἰσθητόν and the like in the Greek manuscripts FIJ or εἶναι ἄπειρον as quoted by Simplicius112. In the Arabic manuscript it stands at the end of a long lemma comprising the section 204b4—205a8113. The note in question is written in the margin of a commentary to this lemma by Abū ʿAlī ibn al-Samḥ114, and runs as follows (ms. Leiden Or. 583, fol. 48b, ed. Badawī, p. 2301): fī naqli Yūḥannā qubaylahū: wa-bi-l-ǧumlati yanbaġī an nabḥaṯa hal fī (fī ms. : an min leg.?) l-arbaʿati mā yaqbalu an yakūna aw lā yaqbalu an yakūna _____________ the colophon proper according to which the colophon of the Leiden manuscript quotes from the subscription of Abū l-Ḥusayn al-Baṣrī’s autograph (cf. above). 110 Riḍā Taǧaddud (ed.), Kitāb al-Fihrist li-l-Nadīm, sec. ed., Tihrān [1973], p. 310. 111 Ibid. 112 In Arist. Phys. 482.6-7. 113 Badawī, al-Ṭabī ʿa, p. 227.3—230.11. 114 Badawī, al-Ṭabī ʿa (p. 2301), gives the correct text of the note, but relates it to the wrong place.

LVIII

Introduction

“In the translation by/of Yūḥannā shortly before this: In general, we must investigate whether among the four [elements] (for ἐκ τῶνδε?) there is what admits existence or does not admit existence.” The expression qubaylahū (“shortly before this”, i.e. before the comment by Abū ʿAlī ibn al-Samḥ) indicates that the note must have been added after the segmentation of the Aristotelian text and the arrangement of the commentary section. As Isḥāq’s translation of the sentence does not reflect any of the explanatory augmentations mentioned above, its understanding is as problematic as the original Aristotelian phrasing. Since the sentence is not taken into consideration in the subsequent commentary section, the note was added here immediately after Isḥāq’s translation and at the beginning of the commentary, in order to address this difficulty and to seek help from another source. There is good reason to assume that the author of the note refers with “Yūḥannā” not to the author, but to the object of the translation, i.e. John Philoponus’ commentary. The only appropriate candidate for Yūḥannā as translator would be Yūḥannā/ Yaḥyā ibn ʿAdī, who is indeed said to have translated Ḥunayn ibn Isḥāq’s Syriac version of Alexander’s commentary on Physics II from Syriac into Arabic. However, exactly the same source reports that Alexander’s commentary on Physics III, from which our sentence is taken, was not extant115. On the other hand, we find immediately after the note quoted another marginal note, which contains Abū ʿAlī ibn al-Samḥ̕s explanation: Abū ʿAlī: yaʿnī bi-ḏālika yaqbalu an yakūna mabdaʾan “Abū ʿAlī: he means with this that it admits to be a principle.” That this note relates to the preceding one is clear not only from its place116, but also from the fact that Ibn al-Samḥ adopts the same (and rather odd) terminology with yaqbalu an corresponding to ἐνδέχεται. The explanation offered by Ibn alSamḥ consists in reading Aristotle’s absolute use of εἶναι not in the sense of εἶναι ἄπειρον, but in the sense of “being a principle” (εἶναι ἀρχήν); and this is exactly what we read in Philoponus’ commentary on the sentence in question (In Arist. Phys., p. 436.3-7): Ἐπειδὴ ὡς ἐπὶ πυρὸς καὶ τοῦ μεταξὺ ἐποιήσατο τὸν λόγον (ἐμνήσθη γὰρ Ἡρακλείτου τὸ πῦρ ἀρχὴν τιθεμένου [...]), διὰ τοῦτο προσέθηκεν, ὅτι τοῖς αὐτοῖς δείξεις ὅτι οὐδὲ τῶν ἄλλων οὐδὲν δυνατὸν εἶναι ἀρχήν. _____________ 115

Ibn al-Nadīm, Kitāb al-Fihrist, ed. Taǧaddud, p. 310. That the two notes follow each other immediately is not visible in Badawī’s edition due to his error described in footnote 114.

116

II.3 Marginal notes and glosses in ms. Leiden Or. 583

LIX

“Since he has formulated his argument with regard to fire and the intermediary (for he has mentioned Heraclitus, who posited fire as origin [...]), for this reason he has added that by the same means you will show that it is not possible for any of the others to be a principle either.”117 Notes introduced by the formula fī naqlin āḫara (“in another translation”) occur once in bk. I (fol. 4a, ed. Badawī, p. 132), twice in bk. III (fol. 35b, ed. Badawī, p. 177+...+, and fol. 48b, ed. Badawī, p. 2304), and once in bk. VII (fol. 192a, ed. Badawī, p. 7595). Two of them supply a clue to the source referred to: The note on bk. VII quotes a phrase to be found only in the so-called version α of this book (φανερὸν ὅτι οὐκ ἂν εἶεν αἱ γενέσεις ἀλλοιώσεις, VII 3, 246a3-4). Why it is exactly this clause that is quoted, while the discrepancy between version α and version β, the one translated by Isḥāq ibn Ḥunayn, concerns the whole of bk. VII 1—3, is not clear. The other note occurs in bk. III and relates to the following passage: Physics III 1, 201a27-29: ἡ δὲ τοῦ δυνάμει ὄντος, ὅταν ἐντελεχείᾳ τι ὂν ἐνεργῇ ἢ αὐτὸ ἢ ἄλλο ᾗ κινητόν, κίνησίς ἐστιν Thus the wording as reconstructible from Isḥāq ibn Ḥunayn’s translation, which runs as follows118: fa-ammā ḥarakatu mā huwa bi-l-quwwati fa-hiya mā takūnu matā kāna l-šayʾu llaḏī huwa bi-l-kamāli yafʿalu immā bi-ḏātihī wa-immā bi-ġayrihī min ǧihati mā huwa mutaḥarrikun “As for the motion of what is potential (= ἡ δὲ τοῦ δυνάμει ὄντος, scil. κίνησις), this is what [motion] is (= κίνησίς ἐστιν), when that which is in actuality acts either by itself or through something else insofar it is movable (= ὅταν ἐντελεχείᾳ τι ὂν ἐνεργῇ ἢ [or ἤτοι] αὐτὸ ἢ ἄλλο ᾗ κινητόν)” Whether Isḥaq read indeed τι ὂν or only ὂν is difficult to assess. However, his otherwise attested use of šayʾun suggests the former. The reading τι ὂν is attested in manuscripts be2 and reported by Simplicius as the reading found in the commentaries by Alexander and Porphyry119. In any case, this is only a minor problem of the passage quoted. The correct reading of this difficult sentence has been a matter of dispute from late antiquity onwards. As for the beginning of _____________ 117

Philoponus: On Aristotle Physics 3. Translated by Mark Edwards (Ancient Commentators on Aristotle). London 1994, p. 101 (slightly modified). 118 Badawī, al-Ṭabī ʿa, p. 177.13—178.2. 119 In Arist. Phys., p. 422.22-24.

LX

Introduction

the sentence, we read in Aldus Manutius’ Venice edition (the so-called Aldina) and in the manuscripts LN2Q2R120 instead of ὄντος : ὄντος ἐντελέχεια, based on the interpretation of ἡ as implying ἡ ἐντελέχεια. Isḥāq ibn Ḥunayn interprets ἡ unambiguously in the sense of ἡ κίνησις. According to Bonitz121 and Ross122 the addition of ἐντελέχεια (or ἐνέργεια) is necessary, while Prantl points out that the difficulty proper is not eliminated by this addition123. For the second half of the parenthetical ὅταν-clause we encounter the following variant readings124: 1. The reading ἢ αὐτὸ is attested in: veFIJQΛdfgh MAb Alex. apud Philop. [356.8-10 ἐν τοῖς παλαιοτέροις τῶν ἀντιγράφων] Philop.l[356.2 ms. t, cf. 356.12, 15]. Other sources display the following variant readings: ᾗ αὐτὸ E : ἤτοι αὐτὸ R Alex. apud Simpl.[422.22-24] Porph. apud Simpl. [ibid.] : οὐχ ᾗ αὐτὸ LNHw Mαε Asp. apud Simpl.[422.20-21] Alex.γρ apud Simpl. [422.24-25] Them. apud Simpl.[422.20-21] Simpl.γρ[τὰ πολλὰ τῶν ἀντιγράφων] Philop.γρ[356.5] : οὐκέτι αὐτὸ b 2. After αὐτὸ we read ἢ ἄλλο in the following manuscripts: EbveFIJQRΛdfgh MAb Alex. apud Philop.[356.8-10 ἐν τοῖς παλαιοτέροις τῶν ἀντιγράφων] Alex. apud Simpl.[422.22-24] Porph. apud Simpl. [ibid.] Philop.l[356.2, cf. 356.12, 15]. Other sources display the following variant readings: ἀλλ' N1H Mαε Asp. apud Simpl.[422.20-21] Alex.γρ apud Simpl. [422.24-25] Them. apud Simpl.[422.20-21] Simpl.γρ[τὰ πολλὰ τῶν ἀντιγράφων]: ἢ ἄλλο ἀλλ' LN2 : ἀλλ' ἢ ἄλλο w : ἀλλ' ᾗ ἄλλο Philop.γρ[356.5] : om. I The reading of Isḥāq ibn Ḥunayn’s Greek exemplar agrees with that of the witnesses veFIJQRΛdfgh MAb, Philoponus, and Alexander’s reading as reported (with the negligible variance of ἢ vs. ἤτοι) by Philoponus and Simplicius. This being the situation of the Greek transmission and Isḥāq’s translation, we read in the margin of ms. Leiden Or. 583 the following note (fol. 35b, ed. Badawī, p. 177+...+): _____________ 120

Not taken into consideration by Ross. Cf. Hermann Bonitz, “Aristotelische Studien I”, Sitzungsberichte der phil.-hist. Classe der kaiserlichen Akademie der Wissenschaften XXXIX, Bd. 8, Wien 1862, p. 183—280, here p. 241. 122 Cf. Ross, p. 537 ad loc. 123 Aristoteles’ Acht Bücher Physik. Griechisch und deutsch mit sacherklärenden Anmerkungen herausgegeben von Dr. Carl Prantl, Leipzig 1854, repr. Leipzig 1935, p. 486—487. 124 Taking additionally into consideration the following manuscripts: d = Marc. gr. 219; f = Marc. gr. 205; g = Mosc. Sinod. gr. 7; h = Ambros. gr. 268; v = Vindob. phil. gr. 75. 121

II.3 Marginal notes and glosses in ms. Leiden Or. 583

LXI

wa-fī naqlin āḫara —wa-huwa llaḏī fassarahū Yaḥyā— wa-llaḏī bi-l-quwwati iḏā ṣāra kamālan wa-faʿala immā (faʿala immā scripsi : fiʿlan mā ms.) huwa wa-immā ġayruhū min ǧihati mā huwa mutaḥarrikun fa-ḏālika huwa l-ḥarakatu “In another translation —and this is the one commented on by Yaḥyā— [we read:] and that which is potential, when it turns into actuality and acts either125 by itself or [through] something else insofar as it is movable, this is motion.” What is surprising about this note is that it does not deviate from Isḥāq’s translation as far as the major problems of the Greek sentence are concerned. On the contrary, immā huwa wa-immā ġayruhū min ǧihati mā huwa mutaḥarrikun is a rather faithful translation of ἤτοι [or ἢ] αὐτὸ ἢ ἄλλο ᾗ κινητόν, the wording transmitted in the majority of the Greek manuscripts and translated by Isḥāq ibn Ḥunayn. Obviously the scholar who added the note did not aim at drawing attention to the above-mentioned alternative readings which modify the meaning of this passage significantly. Thus, the reason for its addition must be sought in the alternative (and wrong) interpretation of the beginning of the sentence, which simply ignores the article ἡ and takes what follows in the sense of τὸ δὲ δυνάμει ὂν (unless it is based on a defective Greek manuscript). Taking the evidence of these two notes together, we may suggest the assumption that references of the type fī naqlin āḫara (“in another translation”) go to a translation of Philoponus’ commentary. Unlike Isḥāq ibn Ḥunayn’s exemplar, this text followed for the most part version α of bk. VII (as suggested by the note on VII 3, 246a3-4 [above, p. LXII] and other notes). Next we have the information that this other translation “is the one commented on by Yaḥyā”. No matter whether this refers to “the translation commented on by Yaḥyā ibn ʿAdī” or to “the translation of the text of the Physics as commented on by John Philoponus”, either interpretation points to Philoponus’ commentary, as the latter formed the focus of Ibn ʿAdī’s comments on the Physics 126. Finally, the last note, linked to Yaḥyā’s commentary, draws —as far as the main textual problems are concerned— on a wording which is supported both by the lemma and the quotations in Philoponus’ commentary. Furthermore, ms. Leiden Or. 583 contains nine notes reporting variant readings in Syriac materials, thus referring likewise to an alternative translation. _____________ 125

The emendation wa-faʿala immā (instead of wa-fiʿlan mā) is required by the syntax. Otherwise the sentence is syntactically broken and unintelligible. The mistaken fiʿlan mā is caused by an erroneous analogy to the preceding kamālan. 126 As shown by Endress, The Works of Yaḥyā Ibn ʿAdī, p. 38; Giannakis, Philoponus, p. 38—43.

LXII

Introduction

Most of these notes refer simply to what is found “in Syriac” (fī l-suryānī), one note mentions “an old Syriac manuscript” (nusḫa qadīma suryāniyya). The notes relate to Physics I 8, IV 3, IV 8, VI 2, VI 5, VII 2 (version α), and VII 3 (version α)127. Obviously, they provide further evidence, in addition to Abū l-Ḥusayn alBaṣrī’s report about Yaḥyā ibn ʿAdī’s collation with the Syriac (cf. p. XLVII), that at least parts of Aristotle’s text or a commentary on the Physics circulated in a Syriac version in 10th c. Baghdad. I have argued elsewhere in favour of Alexander’s commentary as a potential source for such a Syriac adaptation128. (ii) Notes referring explicitly to a manuscript. The key term of these notes is nusḫatun (“copy, manuscript”), but in one case we read ḫaṭṭun (“handwriting”). In some cases, the author of the note also adds a name, while it is not clear at first sight whether this name refers to the scribe of the manuscript, its owner, or the author of the text contained in it. The names mentioned are: “Isḥāq” (Physics IV 8, fol. 80a, ed. Badawī, p. 3643). The reference draws on a note originating from Yaḥyā ibn ʿAdī who indicates that he reproduced an interlinear addition (mulḥaq bayna l-saṭrayni) “he found in the handwriting of Isḥāq” (waǧadtu bi-ḫaṭṭi Isḥāq) in his own copy with a slight modification; “Yaḥyā” (twice Physics I 3, fol. 5a, ed. Badawī, p. 185; fol. 8a, ed. Badawī, p. 301; once V 1 or V 2 [the point of reference is not clear], fol. 120b, ed. Badawī, p. 5122; VI 1, fol. 140b, ed. Badawī, p. 6101; twice VII 4, fol. 196b, ed. Badawī, p. 7732, 7743; once VIII 1, fol. 207b, below, p. 13.4; VIII 2, fol. 208b, below, p. 17.8). From the contents of these notes it becomes clear that they refer without exception to an Arabic copy of Isḥāq ibn Ḥunayn’s translation prepared and/or owned by Yaḥyā ibn ʿAdī; “Ibn ʿAdī” (VII 1, fol. 185b, ed. Badawī, p. 7342; VII 3, fol. 193a, ed. Badawī, p. 7622; VII 5, fol. 201b, ed. Badawī, p. 7921; VIII 1, fol. 205a, below, p. 2.5; VIII 3, fol. 210a, below, p. 22.9; VIII 8, fol. 223b, below, p. 81.10). “Yaḥyā ibn ʿAdī” (collation note at the end of bk. I, fol. 15b, ed. Badawī, p. 762; Physics VI 2, fol. 153a, ed. Badawī, p. 618*); “Yaḥyā al-Naḥwī” (VII 5, fol. 201b, ed. Badawī, p. 7921). The author of this note says that the reference to “John Philoponus’ manuscript” (nusḫat Yaḥyā _____________ 127

For a detailed analysis of these notes cf. Arzhanov & Arnzen, Die Glossen in Ms. Leyden Or. 583. 128 Cf. ibid., p. 461—462; see also below, chapter II.3.1.

II.3 Marginal notes and glosses in ms. Leiden Or. 583

LXIII

al-Naḥwī ) is found in “Ibn ʿAdī’s manuscript” (fī nusḫat Ibn ʿAdī ). This note is discussed below, chapter IV129. Apart from those notes referring to manuscripts together with the names of their owners or authors, a number of notes mention anonymously variae lectiones occurring: — “in another manuscript” (fī nusḫatin uḫrā, or simply fī uḫrā)130, — “in other manuscripts” (fī nusaḫin uḫrā)131, — “in some/many manuscripts” (fī baʿḍi l-nusaḫ / fī kaṯīrin mina l-nusaḫ)132, or — “in the Greek manuscripts” (fī l-nusaḫi l-yūnāniyya)133. As for the notes referring to “some/many manuscripts” (fī baʿḍi l-nusaḫ/fī kaṯīrin mina l-nusaḫ), these must originate from Yaḥyā ibn ʿAdī’s examination of John Philoponus’ commentary in the translation of Qusṭā ibn Lūqā. All of them occur in bk. IV of the Physics, the book most intensely dealt with by Yaḥyā ibn ʿAdī. One of the notes links explicitly the expression “in some manuscripts” to Qusṭā’s translation (naql Qusṭā), i.e. Qusṭā’s translation of Philoponus’ commentary, and two notes trace back to γράφεται-notes by Philoponus. Hence, the fact that these notes speak about “some” or “many” manuscripts must not be taken as indicating that Yaḥyā ibn ʿAdī himself worked with a plurality of Greek manuscripts. Rather this applies either to Qusṭā ibn Lūqā, who may have used for his translation of Philoponus’ commentary more than one manuscript134, or _____________ 129

Cf. p. XCVII f. Fol. 4b, ed. Badawī, p. 152, on Physics I 2, 185b25-26; fol. 32b, ed. Badawī, p. 1651, on III 1, 200b15-16; fol. 56a, ed. Badawī, p. 259+ (fī uḫrā at the beginning of the note is omitted by Badawī), on III 6, 207a6-7; fol. 63a, ed. Badawī, p. 2952, on IV 3, 210b10; fol. 63a, omitted by Badawī (cf. below), on IV 3, 210b11; fol. 108b, ed. Badawī, p. 4663, on IV 13, 222b16; fol. 150b, ed. Badawī, p. 6061, on VI 1, 231a29; fol. 155b, ed. Badawī, p. 6261, on VI 2, 233a17; fol. 159b, ed. Badawī, p. 6391, on VI 3, 234a13-14; fol. 229b, ed. below, p. 106.1-2, on VIII 9, 265b26. 131 Fol. 19a, ed. Badawī, p. 951, on Physics II 2, 194a24-25. 132 Fol. 63b, ed. Badawī, p. 2971, on Physics IV 3, 210b23-24; fol. 65b, ed. Badawī, p. 306.21, in a commentary section on IV 4, 211a2-3, introduced by „Yaḥyā“ (on which cf. Giannakis, p. 158); fol. 66b, ed. Badawī, p. 3121 (! fī baʿḍi l-nusaḫi ziyādatun fī naqli Qusṭā), on IV 4, 212a5-7; fol. 78a, ed. Badawī, p. 3532, on IV 7, 214b5-6; fol. 107b, ed. Badawī, p. 460.15, in a commentary section on IV 12, 221b7-8, introduced by “Yaḥyā” (on which cf. Giannakis, p. 167). 133 Fol. 42a, ed. Badawī, p. 2033, on Physics III 4, 203a9-10; fol. 74b, omitted by Badawī, on IV 6, 213a13; fol. 112a, ed. Badawī, p. 4805, on IV 14, 224a2-4; fol. 132a, ed. Badawī, p. 5475, on V 3, 227a31. For these four notes cf. below, chapter V. 134 On this point cf. below, chapter V. 130

LXIV

Introduction

to Philoponus’ own citation of the manuscripts at his disposal (cf. “ἰστέον ὅτι τὰ πολλὰ τῶν βιβλίων ...”, In Arist. Phys., p. 756.8, on 221b7-8). The source(s) of the notes referring to “another [other] manuscript[s]” is not clear and requires in-depth investigation. In some cases it is possible to relate the variant reading reported in the note to a reading attested in the Greek manuscript tradition, e.g. Physics I 2, 185b25-26: [a] ἐθορυβοῦντο δὲ καὶ οἱ ὕστεροι τῶν ἀρχαίων (ὕστεροι τῶν ἀρχαίων E Philop. Simpl. : ὕστεροι135 καθάπερ καὶ οἱ ἀρχαῖοι γρ. E i.m. : ὕστερον καθάπερ καὶ οἱ ἀρχαῖοι post ras. J : ὕστερον καθάπερ οἱ ἀρχαῖοι FI) [b] Transl. Isḥāq, ed. Badawī, p. 15.18: wa-qad ḫāḍa ayḍani l-mutaʾaḫḫirūna mina l-qudamāʾi [c] Marginal note, with marks over al-mutaʾaḫḫirūna and at the beginning of the note: waǧadnā fī nusḫatin uḫrā “al-mutaʾaḫḫirūna ʿalā miṯāli l-qudamāʾi” “In another manuscript we read: the more recent [philosophers] just as the ancient ones” [d] The variant reading reported conforms with the alternative reading reported in the margin of ms. E, whereas Isḥāq’s translation gives the text read in ms. E in textu and by Philoponus and Simplicius. Physics IV 3, 210b10: [a] δεήσει γὰρ ἀμφότερα ἑκάτερον (ἑκάτερον codd. praeter J : ἑκατέρῳ J) ὑπάρχειν [b] Transl. Isḥāq, ed. Badawī, p. 295.7-8: wa-ḏālika anna hāḏā l-qawla yaqtaḍī an yakūnā ǧamīʿan kulla wāḥidin minhumā [c] Marginal note: fī uḫrā : “li-kulli wāḥidin” “In another [manuscript]: for both” [d] The reading li-kulli wāḥidin minhumā instead of kulla wāḥidin minhumā reflects the variant reading ἑκατέρῳ transmitted in ms. J _____________ 135

Sic pace Ross.

II.3 Marginal notes and glosses in ms. Leiden Or. 583

LXV

Physics VI 1, 231a28-29: [a] ἕτερον γὰρ τὸ ἔσχατον καὶ οὗ ἔσχατον (οὗ ἔσχατον E2 K FHIJ Simpl. : οὐκ ἔσχατον fort. E1) [b] Transl. Isḥāq, ed. Badawī, p. 606.3-4: li-anna l-āḫira ġayru mā huwa lahū āḫirun [c] Marginal note, with marks over al-āḫira and at the beginning of the note: wuǧida fī nusḫatin uḫrā “li-anna l-āḫira ġayru mā laysa bi-āḫirin” “In another manuscript it reads: because the extremity and that which has no extremity are distinct” [d] Isḥāq’s translation follows the majority of the manuscripts, whereas the alternative reading mentioned in the marginal note corresponds with ms. E 1. However, this evidence is not indicative as to the question whether the author of the note had access to a Greek manuscript with the particular reading he reports or to another Arabic text translated from a Greek text (either Aristotle or a commentary quoting Aristotle) which transmitted this reading. Due to this uncertainty, we cannot determine without further evidence whether such notes were taken down by Isḥāq ibn Ḥunayn himself during the process of translation or by any later scholar of the school of Baghdad. At least three of these notes seem to trace back to Arabic rather than Greek manuscripts: Physics IV 3, 210b11: [a] οἷον ... τὸν οἶνον οἶνόν τε καὶ ἀμφορέα (scil. εἶναι) [b] Transl. Isḥāq, ed. Badawī, p. 295.9: miṯālu ḏālika an ... yakūna l-šarābu šarāban wa-qinnīnan (qinnīnan Badawī : fīnā vid. ms.) [c] Marginal note, with mark over wa-fīnā: [nusḫatun] uḫrā: “wa-ināʾan” “Another [manuscript]: and jar” [d] The correction of the mistake wa-fīnā is either based on another translation which rendered ἀμφορεύς by ināʾ, or on an Arabic manuscript the scribe of which had emended the text by conjecture. Isḥāq never translates ἀμφορεύς with ināʾ, but consistently with qinnīn (although in most cases miswritten in the Leiden manuscript as q-n-y, cf. 210a31-b16 = ed. Badawī, p. 294.1— 296.1)

LXVI

Introduction

Physics IV 13, 222b16-17: [a] μεταβολὴ δὲ πᾶσα φύσει ἐκστατικόν. ἐν δὲ τῷ χρόνῳ πάντα γίγνεται καὶ φθείρεται [b] Transl. Isḥāq, ed. Badawī, p. 466.10-11: wa-kullu taġayyurin fa-huwa bi-l-ṭabʿi muzīlun wa-fī zamānin yatakawwanu wa-yafsudu kullu mā yatakawwanu wa-yafsudu [c] Marginal note, with marks between muzīlun and wa-fī zamānin and at the beginning of the note: wuǧida fī nusḫatin uḫrā: “ʿalā zamānin” “In another manuscript it reads: over/during [a certain period] time” [d] Probably, some scribe (prompted by Isḥāq’s absolute use of muzīl ?) had replaced wa-fī by ʿalā, unless the note is supposed to point to the reading wa-ʿalā zamānin instead of wa-fī zamānin, that is to another translation of the preposition ἐν. Either interpretation points to an Arabic manuscript as the object of reference. Physics VIII 9, 265b24-28: [a] καὶ γὰρ οὗτοι τὴν κατὰ τόπον κίνησιν κινεῖσθαι τὴν φύσιν λέγουσιν — ἡ γὰρ διὰ τὸ κενὸν κίνησις φορά ἐστιν ἐν τόπῳ (ἐστιν ἐν τόπῳ Simpl. 1319.7 : ἐστιν ὡς ἐν τόπῳ EK : ἐστιν καὶ ὡς ἐν τόπῳ cett.) —, τῶν δ' ἄλλων οὐδεμίαν ὑπάρχειν τοῖς πρώτοις [...] οἴονται· [b] Transl. Isḥāq, fol. 229b19-20 (cf. below, p. 105.10–106.2): fa-inna hāʾulāʾi ayḍan yaqūlūna inna l-ṭabīʿata tataḥarraku ḥarakata lmakāni wa-innahū laysa fī l-makāni yumkinu an takūna wāḥidatun mina lḥarakāti l-bāqiyati fī l-ūlā [c] Marginal note: waǧadnā fī nusḫatin uḫrā ziyādatan fī hāḏā l-mawḍiʿi: “wa-ḏālika anna lḥarakata min qibali l-ḫalāʾi hiya nuqlatun fī l-makāni” “At this place we find in another manuscript the [following] addition: because the motion which is accounted for by void is locomotion” [d] In this case, we are faced with a scribal error and its correction within the Arabic tradition. The “addition” mentioned in the marginal note corresponds exactly with the clause ἡ γὰρ διὰ τὸ κενὸν κίνησις φορά ἐστιν ἐν τόπῳ as transmitted by Simplicius, and its terminology is doubtless the one used by Isḥāq ibn Ḥunayn. The omission of the phrase in the Arabic tradition is easily explained through the homoioteleuton fī l-makān (added at the wrong place, then deleted by the copyist of the lacunary manuscript), yet rather

II.3 Marginal notes and glosses in ms. Leiden Or. 583

LXVII

unlikely within the Greek manuscript transmission. The correct text run as follows: fa-inna hāʾulāʾi ayḍan yaqūlūna inna l-ṭabīʿata tataḥarraku ḥarakata lmakāni / wa-ḏālika anna l-ḥarakata min qibali l-ḫalāʾi hiya nuqlatun fī lmakāni / wa-innahū laysa yumkinu an takūna wāḥidatun mina l-ḥarakāti lbāqiyati fī l-ūlā The case is of course different with the four notes referring explicitly to “Greek manuscripts” (nusaḫ yūnāniyya). Here we can indeed expect that their author(s) had recourse to Greek sources. As a consequence, these notes are of crucial importance with regard to the question whether the translator of Aristotle’s Physics prepared his translation on the basis of a single or more than one Greek manuscript, which is why they are scrutinised in this context below, chapter V. Yet a different class of marginal notes referring to another manuscript is formed by notes with the mark ‫ح‬. From Abū l-Ḥusayn al-Baṣrī’s colophon of bk. I, reproduced by Abū l-Ḥakam al-Maġribī in ms. Leiden Or. 583, we learn that Abū l-Ḥusayn used this mark to indicate corrections and annotations transcribed from Yaḥyā ibn ʿAdī’s copy. Although this information concerns only the first two quires of Abū l-Ḥusayn’s compilation (containing bk. I 1, 184a10 — II 2, 194a12 of the Physics), it seems reasonable and plausible to assume that this holds true for the remaining notes marked by ‫ ح‬as well136. The ‫­ح‬notes contain observations on competing readings/interpretations of the Aristotelian text, remarks on readings and potential errors in Yaḥyā ibn ʿAdī’s copy of Isḥāq’s translation, as well as doctrinal explanations. Giannakis (p. 31—32) argues that some of them “were taken from marginal notes of a Greek manuscript”, because they display competing readings attested in different branches of the Greek manuscript tradition. However, such notes may just as well derive from Yaḥyā ibn ʿAdī’s comparison of Isḥāq’s translation with another translation of the Physics or with an Aristotelian lemma in the Arabic translation of a commentary (especially as Giannakis himself points to three notes marked by ‫ ح‬which evidently draw upon Philoponus’ commentary137). It is also worth noticing that the mark ‫ ح‬indicates only the source of transcription, yet not necessarily its original creator. As Yaḥyā ibn ʿAdī studied the Aristotelian corpus under Mattā ibn Yūnus (and probably under Abū Naṣr al-Fārābī), we cannot rule out that the contents of some ‫­ح‬notes may derive ultimately from his teacher(s). Finally, we have to mention here the references to antecedent exemplars of the Leiden manuscript transcribed or originally introduced by its scribe, Abū l_____________ 136 137

A complete list of all 57 notes marked ‫ ح‬is given by Giannakis, Philoponus, p. 301. Cf. Giannakis, Philoponus, p. 33—34.

LXVIII

Introduction

Ḥakam al-Maġribī. In the colophon of bk. I, Abū l-Ḥakam refers to Abū l‐ Ḥusayn’s compilation as “the original” (al-aṣl)138 and to the copy of this “original”, which in turn served as his exemplar, as “the exemplar” (al-umm)139. Furthermore, he adopts from Abū l-Ḥusayn al-Baṣrī the term al-dastūr 140, which seems to trace back to Yaḥyā ibn ʿAdī, who is reported to have collated his copy with the dastūr of Isḥāq ibn Ḥunayn’s translation. Yaḥyā ibn ʿAdī̕s use of the term dastūr is confirmed by similar notes in the Paris manuscript of the Aristotelian Organon 141. In the scholarly literature, these notes are often taken as referring to the translators̕ autographs. However, in view of the below examination of the relationship between our direct and indirect text witnesses, we have good reason to doubt that Yaḥyā ibn ʿAdī had access to Isḥāq’s autograph of the Physics. More likely, dastūr was used in the Baghdad school of Aristotelians as a generic term that served as a scholastic label for the authoritative reference exemplar or “master copy” of an Aristotelian work that had been certificated for the purposes of teaching, copying (viz. collating) and commentary exercises _____________ 138

Cf. Giannakis, Philoponus, p. 20—23. Ibid. 140 Or al-dustūr, “guiding principle”, “exemplar”, “collection of principles/rules” [i.e. canon], cf. Vullers I, 876a; Freytag II, 29a; Lane 879a. N.B.: Ḥunayn ibn Isḥāq uses the term in the sense of “methodical principle”, when he paraphrases αἱ μέθοδοι (Galen, De simplicium medicamentorum temperamentis ac facultatibus VI, p. 803.9 Kühn) through alfarāʾiḍ wa-l-qawānīn allatī ǧaʿalnāhā dastūrāt naʿmalu ʿalayhā, ms. Florence, Bibl. Medicea Laurenziana Orient. 193, fol. 104a10. 141 Such a note occurs in the colophon of Aristotle’s Categories, ms. Paris BnF ar. 2346, fol. 178b; cf. Khalil Georr, Les Catégories d’Aristote dans leurs versions syro-arabes, Beyrouth 1948, p. 358. However, the reference is anything but clear, for the term dastūr appears twice, first in the phrase al-dastūr allaḏī bi-ḫaṭṭi Isḥāq (“the dastūr in Isḥāq’s handwriting”), then three lines below in the phrase nusḫat Yaḥyā Ibn ʿAdī al-manqūla min dastūr al-aṣl allaḏī bi-ḫaṭṭi Isḥāq Ibn Ḥunayn (“Yaḥyā Ibn ʿAdī’s manuscript which was copied from the dastūr of the original in Isḥāq’s handwriting”). If dastūr meant unambiguously “autograph”, the addition “in Isḥāq’s handwriting” would be pleonastic and the expression dastūr al-aṣl allaḏī bi-ḫaṭṭi Isḥāq Ibn Ḥunayn, which reappears in identical form in the colophon of De Interpretatione (ms. Paris BnF ar. 2346, fol. 191b ult.), would mean “the autograph of the original written by Isḥāq Ibn Ḥunayn” (unless one takes dastūr al-aṣl as qualifying genitive construction [iḍāfat al-ʿayn ilā l-maʿnā], but then again the qualification would be pleonastic). An almost identical phrase occurs in a note by Ibn Suwār reported at the end of Sophistici elenchi (fol. 379b11f. of the same ms.): wa-ḏakara l-Ḥasan Ibn Suwār anna nusḫatahu llatī nusiḫat minhā hāḏihi l-nusḫatu naqalahā min nusḫatin kutibat min dastūri Yaḥyā Ibn ʿAdī allatī (sic!) bi-ḫaṭṭihī (“al-Ḥasan Ibn Suwār mentioned that he had copied his manuscript, from which the present manuscript is copied, from a manuscript that had been transcribed from Yaḥyā Ibn ʿAdī’s dastūr written in the latter’s hand”). 139

II.3 Marginal notes and glosses in ms. Leiden Or. 583

LXIX

among the school members. References to this “master copy” may relate to the translator’s autograph, if they are qualified by the extension allaḏī bi ḫaṭṭihī (“in his handwriting”)142. References to these three manuscripts, “the original”, “the exemplar”, and “the master copy”, are found in the Leiden manuscript not only in the colophons or straightforward collation notes at the end of a book or quire, but also in the margins together with other information. In some cases they give the source of the text added in the margin; in others they report a fenestra (e.g. bayāḍ fī l-aṣl ) or indicate the verification of a particular correction (e.g. ṣaḥḥa mina l-umm). Obviously, the notes referring to the “master copy”143 trace back to a very early copy of Isḥāq’s translation. However, at least one of these references suggests that this early “master copy” already contained glosses or scholia. The reference in question (ms. Leiden Or. 583, fol. 205a) is difficult to read and probably lacunose. According to the attached superscript signs it is related to Physics VIII 1, 250b23: “If then it is possible that at any time nothing should be in motion” (transl. Hardie-Gaye). The text of the note runs as follows:

‫( ﺣﺪان ﰲ اﻟﺘﺒ ّﺪل ﻣﻦ ﺣﻴﺚ ﺳﻜﻮن أي ﻣﻦ ﺑﻌﺪ ﺳﻜﻮن ﻗﺎل ﳛﲕ ﻛﺬا ﰲ اﻟﺪﺳﺘﻮر‬or ‫أي )? ﻟـﺎ‬

“I.e., †...† in change insofar as [there is] rest, that is, after [the state of] rest. Yaḥyā says: thus in the dastūr”

Badawī (al-Ṭabīʿa, p. 802, note 4) reads ‫“( ﺧﺬان‬both begin” [to change]) instead of ‫ ﺣﺪان‬which can be ruled out, because neither the preceding text nor what follows contains anything to which the dual form might refer. The most plausible reading might be ay li-aḥadin an (“I.e., somebody may that”) or ay yaʾḫuḏu an(na) (“I.e., he [Aristotle] takes [the assumption] that”). In any case, it is clear that before or after fī l-tabaddul something must have been omitted by the author or scribe of the note. Be that as it may, there is little doubt that the note does not report any variant reading of the Aristotelian text proper, but some interpretative explanation quoted by Yaḥyā ibn ʿAdī from the “master copy”. Whether such references originate altogether from collation notes by Yaḥyā ibn ʿAdī or whether other persons involved in the genesis of our manuscript after Yaḥyā ibn ʿAdī had access to this “master copy”, we don’t know. Notes referring to “the original”144 must have been transcribed from corresponding _____________ 142

The term occurs once again in collation notes of another manuscript originating from the Baghdad school, ms. Paris BnF ar. 2281; cf. Endress, Ibn al-Ṭayyib̕s Arabic Version, p. 234—236. 143 Such notes occur in ed. Badawī, p. 4112, 4642, 6231, 660*, below, p. 3.1 (ad 250b23). 144 Cf. ed. Badawī, p. 5661, 590*, 6101.

LXX

Introduction

notes in Abū l-Ḥakam al-Maġribī’s exemplar, while notes referring to this “exemplar” (al-umm) originate directly from Abū l-Ḥakam’s pen. (iii) Personalised notes, i.e. notes indicating —prior to the content proper of the note— their authors, or notes introduced by a verbum dicendi in the first pers. sg. (e.g. qultu, saʾaltu). The latter obviously originate from classroom conversations between the compiler of the present corpus and his teachers Abū ʿAlī ibn alSamḥ and Abū l-Faraǧ ibn al-Ṭayyib, and occur in bks. III—VIII (cf. above)145. The majority of what I call personalised notes are introduced by “Abū ʿAlī” [ibn al-Samḥ] and relate to bks. I 5 — VI 3 of the Physics 146. One note each is introduced by “Abū Bišr” [Mattā ibn Yūnus]147 or by “Mattā” [probably likewise Mattā ibn Yūnus]148, four notes are introduced by “Abū l-Faraǧ” [ibn alṬayyib]149, and nine by “Yaḥyā”. Of the latter, two originate clearly from Yaḥyā _____________ 145

One further note, introduced by qultu ana (“I said”), occurs in bk. I, fol. 2a, ed. Badawī, p. 41. It is not clear whether this note originates from Abū l-Ḥusayn al-Baṣrī (probably it does not, as it seems to draw on Philoponus’ commentary; cf. Giannakis, p. 136), nor who the referenced person is. 146 Altogether 31 notes: bk. I 5, fol. 11a, ed. Badawī, p. 441, 443; fol. 12a, ed. Badawī, p. 1 49 ; bk. I 6, fol. 13a, ed. Badawī, p. 541 (probably another note by Abū ʿAlī ibn al-Samḥ is found at the top of this folio, partly cut off at the beginning and omitted by Badawī); bk. I 8, fol. 14a, ed. Badawī, p. 682 and 691; fol. 14b, ed. Badawī, p. 704; bk. I 9, fol. 15a, ed. Badawī, p. 734; bk. II 3, fol. 20b, ed. Badawī, p. 1013; fol. 21a, ed. Badawī, p. 1081 (the mark at the beginning of the note indicates that it belong to the previous commentary, ed. Badawī, p. 107.12-19, not to p. 108.3 as printed by Badawī); bk. III 1, fol. 35b, ed. Badawī, p. 1771; bk. III 5, fol. 46b, ed. Badawī, p. 2232; fol. 47a, ed. Badawī, p. 2242; fol. 48a-b, ed. Badawī, p. 2302, 2305; fol. 53a, ed. Badawī, p. 2461 (the note is introduced by “Abū ʿAlī”, which is omitted by Badawī), ed. Badawī, p. 2471 (2 notes), p. 2481; bk. III 6, fol. 54b (omitted by Badawī); bk. IV 4, fol. 65a, ed. Badawī, p. 3046, p. 3052; bk. IV 5, fol. 72a, ed. Badawī, p. 3303; bk. IV 11, fol. 95b, ed. Badawī, p. 4181, 4182, 4202, fol. 98a, ed. Badawī, p. 4282 (this note does not relate to the place indicated by the scribe [and reproduced by Badawī], but to the following lemma, i.e. 220a3-4); fol. 98b, ed. Badawī, p. 4303; bk. IV 14, fol. 109b, ed. Badawī, p. 4713; fol. 111b, ed. Badawī, p. 4781; bk. VI 3, fol. 159b, ed. Badawī, p. 6392. 147 Bk. II 3, fol. 21a, ed. Badawī, p. 1081. The comment introduced by “Abū Bišr” as printed by Badawī, p. 1621, is not a marginal note, but part of a regular commentary section at the end of fol. 31a. 148 Bk. VIII 6, fol. 218b (cf. below, p. 58, App. 3). Badawī, p. 8663, reproduced the note proper, but failed to notice that it is introduced interlinearily by “Mattā”. 149 Bk. VI 10, fol. 184b, ed. Badawī, p. 731 note; VIII 7, fol. 221a (cf. below, p. 69, App. 3); fol. 221b (cf. below, p. 73, App. 3); VIII 9, fol. 229a (cf. below, p. 102, App. 3).

II.3 Marginal notes and glosses in ms. Leiden Or. 583

LXXI

ibn ʿAdī150, whereas seven derive from Philoponus’ commentary151. Of special interest for the question of the authenticity of Isḥāq ibn Ḥunayn’s translation as well as for its transmission are the above-mentioned 18 notes introduced by “Isḥāq”, to be discussed in chapter III. (iv) Notes which are entirely silent about their sources or authors. This is by far the largest and most heterogeneous class of notes containing simple substitutions of single words (mostly technical terminology), portions of text supposed to be inserted in the main text (in most cases omission by homoioteleuton), explanations of varying comprehensiveness, and even diagrams. Unlike the previous kinds of notes, which occur mostly in the margins of the manuscript, notes of this type are frequently added between the lines. As we saw above, most of the notes pertaining to categories (i)—(iii) originate from the early period of copying and annotating Aristotle’s Physics up to Abū lḤusayn al-Baṣrī’s compilation in the year 395/1004. We do not know whether this applies likewise to the notes which —on the surface— do not reveal anything about their authors or sources. However, as the latter do not exhibit any stylistic characteristics that might set them apart from the other notes, nor do they contain any doctrinal or cultural element pointing to a later period of reception, there is no reason to doubt that they originate from the same period of time. As a matter of fact, this assumption is corroborated by the fact that a remarkable number of type-iv-notes imply that their author(s) was / were acquainted with and built on the commentary tradition of late antiquity, which is a distinctive feature of the 10th c. school of Baghdad152. On the other hand, this result raises the question why so many notes and glosses were added in the margins and between the lines of the compilation. This concerns particularly the doctrinal and exegetical notes accompanying the Aristotelian text and the notes written in the margins of commentary sections. _____________ 150

Bk. IV 8, fol. 80a, ed. Badawī, p. 3643, bk. VIII 1, fol. 205a, below, p. 3, App. 3. Bk. III 8, fol. 57b, ed. Badawī, p. 2671 (cf. Philop., p. 493.23-25, Giannakis, p. 156); bk. VI 7, fol. 175b, ed. Badawī, p. 6961 (cf. Philop., p. 870.30f., Giannakis, p. 347); bk. VI 9, fol. 179a, ed. Badawī 7111 (cf. Philop., p. 817.5f., Giannakis, p. 350) and 7113 (cf. Philop., p. 816.30—817.4, Giannakis, p. 350); fol. 180a, ed. Badawī, p. 7152 (cf. Philop., p. 817.20-27, Giannakis, p. 351f.); bk. VI 10, fol. 184a, ed. Badawī, p. 7301 (cf. Philop., p. 820.17-19, Giannakis, p. 354); bk. VII 4, fol. 197b, ed. Badawī, p. 7791. 152 A detailed index of parallels between Arabic scholia and commentary sections and the extant parts of Philoponus’ commentary is found in Giannakis, Philoponus, Appendices I—IV. For examples of type-iv-notes drawing on the commentary by Alexander of Aphrodisias, see below, chapter II.3.1. 151

LXXII

Introduction

While it seems natural and easily understandable that notes concerning philological problems of the Aristotelian text sensu stricto were written immediately in the margins of the relevant passages, it is a matter of speculation why Abū lḤusayn al-Baṣrī did not incorporate such notes into his arrangement of the commentary sections, especially since many (if not all) of them trace back to exactly the same authorities, the Baghdad Aristotelians and their late antique sources. Any hypothesis drawing on chronological reasons can be ruled out, as many notes and glosses originate from scholars who had died long before the date of Abū l-Ḥusayn’s compilation. For the same reason, we can rule out the possibility that those marginal notes derive from oral communication or teaching as opposed to written materials used for the regular commentary sections, not to mention the fact that the latter evidently also include passages derived from oral sources. Considering the disparity and multiple origins of the materials at Abū l-Ḥusayn’s disposal, would it be possible that all that we read in the main text of the Leiden manuscript, no matter whether Aristotelian lemmata or commentary sections, originates from more or less elaborate and well structured documents used in the Baghdad school classes, whereas anything reproduced in the form of marginal notes and glosses derives from random notes and scattered minutes or leaflets stored in the Baghdad “school archive” and collected and incorporated by Abū l-Ḥusayn al-Baṣrī after he had completed the main compilation? Or are we faced with a testimony of Abū l-Ḥusayn’s exaggerated respect for the documents of his teachers and their predecessors such that he even borrowed or imitated the structure of their materials by placing a comment underneath an Aristotelian lemma, when he found it written there, but in the margins or between the lines, whenever it was written there in his documents? — The answer to these questions must be reserved for a systematic study of all extant manuscripts originating from or depending on the Baghdad school tradition153.

II.3.1 Excursion: Notes drawing on the commentary by Alexander of Aphrodisias In an article published in 1996, Elias Giannakis provides a survey and edition of the marginal scholia preserved in the Leiden manuscript and attributed explicitly _____________ 153 In this context it might be worth mentioning that Abū l-Ḥakam al-Maġribī, the scribe of the Leiden manuscript, at a number of places divided —in anticipation of more comprehensive annotations— the writing area for a certain number of lines (never for a whole page) into two columns using the right column (roughly half or two thirds of the usual line length) for Aristotelian text and the left column for commentary sections proper or annotations, with slightly different leading between the lines of each column, e.g. fols. 148b, 201a, 217a, 221b; cf. also above, footnote 75.

II.3.1 Excursion: Notes drawing on Alexander of Aphrodisias

LXXIII

to Alexander of Aphrodisias154. In view of the fact that most of these scholia occur in commentary sections ascribed to “Yaḥyā” (John Philoponus), Giannakis argues that “it may be concluded that the majority of the passages quoting Alexander’s views derive indirectly from John Philoponus’ commentary on the Physics”155. This conclusion is especially supported by the fact that most of the explicit references to “Alexander” (al-Iskandar) indeed relate to corresponding references in Philoponus’ commentary. However, with regard to the question whether the Baghdad scholiasts had access to other sources transmitting (parts of) Alexander’s commentary —a question that I am unable to answer here— it may be worth noting that the Leiden manuscript contains scholia (ultimately) deriving from Alexander’s commentary, yet without being explicitly labelled as such. What is more, since such unspecified (type-iv) scholia constitute the vast majority of the marginal materials compiled and preserved in this manuscript, we may infer that this imbalance applies to scholia inspired by Alexander’s commentary too. As for book VIII of the Physics, this assumption can be fleshed out as follows. While ms. Leiden Or. 583 contains for this book only one scholium explicitly attributed to Alexander of Aphrodisias156, at least thirty further anonymous marginal notes can be related to extant fragments of his commentary (cf. Index locorum). Below is a list of the most conspicuous cases: Aristotle, Physics VIII157

Alexander, Commentaire perdu ... (ed. Rashed)158

Scholia & glosses in ms. Leiden Or. 583

[1] 252a10] καὶ οἱ μίαν frag. 550] ἢ χρονικήν [...] ἢ fol. 207a] yaʿnī l-ʿaqla ἀρχὴν ποιοῦντες ποιητικήν — τὸν νοῦν [2] 253b7] παρὰ τὴν frag. 566] τὴν προκειμένην fol. 210a] yaʿnī l-maḏhaba l-ṭabīʿiyya μέθοδον μέθοδον τὴν φυσικήν _____________ 154

E. Giannakis, “Fragments from Alexander’s Lost Commentary on Aristotle’s Physics”, Zeitschrift für Geschichte der arabisch-islamischen Wissenschaften 10 (1995/1996), p. 157—187. 155 Ibid., p. 158. 156 Cf. ibid., p. 178, 185. 157 The left column of the table displays the lemmata or key-phrases (partly reconstructed by M. Rashed) to which the Greek fragments of Alexander’s commentary relate. In most cases, it is only one of the corresponding Arabic words of Isḥāq ibn Ḥunayn’s translation that is marked as point of reference of the relevant Arabic scholium displayed in the right column. For more details on the particular modes of reference see App. 3 of the Arabic edition below. 158 Cf. supra, footnote 42.

LXXIV

Aristotle, Physics VIII [3] 255a19] ταῦτα

Introduction

Alexander, Commentaire perdu ... (ed. Rashed)

Scholia & glosses in ms. Leiden Or. 583

frag. 587] τὰ φυσικὰ καὶ fol. 212b] yaʿnī ġayra lmutanaffisati wa-an taἄφυτα σώματα ḥarrakat ṭabʿan

[4] 255a30] τὰς αὑ- frag. 590] τὰ fol. 213a] yaʿnī amākinahā φύσιν τόπους τῶν ἐνεργείας [5] 255b25] ὡς κινεῖ

frag. 597] [...] κατὰ συμ- fol. 213a] ay bi-l-ʿaraḍi βεβηκὸς δὲ κινεῖται

[6] 256b3-4] ταὐτὰ συμβήσεται

frag. 606] [...] ἔστι δὲ τοῦτο τὸ ἐν τοῖς κινουμένοις [...] εἶναι πρῶτόν τι κινοῦν αὐτοκίνητον

fol. 214b] yalzamu l-wuqūfu ʿinda lmutaḥarriki mabdaʾa ḥarakatihī li-nafsihī

[7] 257b2-4] αὑτὸ κι- frag. 625-626] αὐτοκίνηνοῦν [...] ἄτομον τον νῦν λέγει τὸ ὅλον ὑπὸ ὅλου κινούμενου ἑαυτοῦ. τῷ εἴδει ἄτομον τῷ εἴδει καθὸ κινεῖται μίαν κίνησιν

fol. 216b] fa-li-hāḏā kāna wāḥidan ġayra munqasimin fī nawʿihī li-annahū bi-asrihī muḥarrikun wabi-asrihī mutaḥarrikun

frag. 636] καὶ τοῦτο ἄτοπον [...] ὡς ἑκάστου τῶν μερῶν αὐτοῦ αὐτοκινήτου ὄντος

fol. 217a] taqdīru l-kalāmi huwa anna kulla wāḥidin min aǧzāʾihī (scil. aǧzāʾi lmuḥarriki min ḏātihī) yuḥarriku nafsahū

[9] 258a20] ἐξ ἀνάγ- frag. 640] εἰ δὲ ἀσώματον (scil. τὸ κινούμενον), θακης ἁπτόμενα τέρου θατέρον ἅψεται οἷον τὸ ἀσώματον τοῦ σώματος

fol. 217a] [...] li-anna lmutaḥarrika qad yaǧūzu an yakūna lā ǧismun [lahū] fa-yakūnu yamassu l-ǧisma [...] wa-lā yamassuhu l-ǧismu

[8] 258a3] ἔτι εἴπερ

[10] 258b15-16] ἐκτὸς frag. 653] τοιοῦτον τὸ fol. 218b] [...] li-annahū laysa tuḥaṣṣilu l-ṣūratu lἄϋλον εἶδος μεταβολῆς hayūlā τῆς frag. 680] [...] ἀρχὴν δὲ fol. 219b] al-mabdaʾ al[11] 259b13-14] πρώτην τὸν αἰθέρα λέγει awwal + yaʿnī l-hawāʾa πρώτης ἀρχῆς supra lin.

II.3.1 Excursion: Notes drawing on Alexander of Aphrodisias

Aristotle, Physics VIII

Alexander, Commentaire perdu ... (ed. Rashed)

[12] 259b30-31] ἐνίαις frag. 689] μέναις ἀρχαῖς

LXXV

Scholia & glosses in ms. Leiden Or. 583

ταῖς πλανω- fol. 219b] yušīru bi-ḏālika ilā l-aflāki l-mutaḥayyirati

[13] 260a8-9] διὰ τὸ frag. 698] οἷον βορείοις ἢ fol. 220a] ay al-ǧanūbu wa-l-šamālu ἐν ἐναντίοις εἶναι νοτίοις τόποις [14] 260b8] τῶν παθη- frag. 708] τῶν ἀλλοίωσιν μάτων

κατὰ fol. 221a] al-taġayyurāt

[15] 260b26] τὴν φο- frag. 713] τῇ φύσει πρό- fol. 221b] yurīdu l-ūlā (scil. al-nuqlata l-ūlā) bi-l-ṭabʿi τερα ἡ φορά ρὰν ... πρώτην [16] 264a14] ἐπὶ τῶν frag. 759] κινήσεων, οὐ fol. 227b] yaʿnī sāʾira lμόνον τῆς φορᾶς ὡς ḥarakāti l-bāqiyati baʿda lἄλλων nuqlati llatī ḏakarnāhā ὑπέθετο [17] 264b2] ἅμα

frag. 763] ἐν ἀμερεῖ τῷ fol. 227b] ay ānan wāḥidan νῦν

[18] 264b6] χρόνος

frag. 766] χρόνον νῦν κοι- fol. 227b] istaʿmala sma lzamāni fī hāḏā l-mawḍiʿi νότερον λέγει τὸ νῦν makāna l-āni

Taking into consideration that the extant comments by John Philoponus related to sections no. [4], [7], [8], [12], [14] and [18] are relatively comprehensive, if not completely preserved159, but apparently could not have served as sources of the Arabic quotations from Alexander’s commmentary, the above-noted examples provide another clue to the fact that parts of Alexander’s commmentary other than quotations in Philoponus’ commentary must have been available to the Aristotelian scholars of Baghdad.

_____________ 159

In addition to Vitelli’s edition see now also Andrea Rescigno, “Nuovi frammenti del Commento di Filopono ai libri V-VIII della Fisica”, Studia graeco-arabica 7 (2017), p. 75—104.

LXXVI

Introduction

II.4 Ms. Escorial ár. 896 A fragment of Isḥāq ibn Ḥunayn’s Arabic translation of the Physics is preserved in ms. Escorial ár. 896 (henceforth: Esc. 896)160. This codex is composed of 63 folios and contains five fragmentary texts written by different hands, two of them in Maġribī script. The manuscript catalogue is silent about whether the fragments originate from different independent codices. My access to this manuscript is restricted to monochrome pictures of the folios which contain the fragment of the Physics 161 and the first page of the following fragment. Judging from a comparison of these folios, the paper of the Physics fragment, which is in good condition and shows vertical chain lines, is different from the following one (which is lighter and without visible chain lines). In both texts the margins of the folios have been heavily cut down. The latest identified author (frag. 2, fols. 30a—38b) is the grammarian ʿAbd al-Raḥmān ibn ʿAlī ibn Ṣāliḥ al-Makkūdī (d. 807/1405 in Fes). However, other parts of the codex may be much older. The fact that two texts are written in Maġribī script and a third one was composed by a Moroccan author may point to Morocco or al-Andalus as place of compiling and binding. Again, this does not necessarly hold true for the place of copying of the other two fragments. The manuscript has a Western foliation entered after (re-)binding and cutting of the fragments. The script of the Physics fragment is a bold and large Nasḫ with sparse punctuation and almost no vocalization signs at all. The text is written in one column of 15 lines per page. It contains marginal and interlinear corrections and glosses, which partly seem to be written by the same hand, partly originate evidently from another hand. Four notes are marked with the sign ‫خ‬ (= nusḫa) and report variant readings from another manuscript. Whether the scribe of ms. Esc. 896 himself collated his exemplar with another copy or simply transcribed these notes from his only exemplar is a matter for speculation162. Our fragment occurs on fols. 48a1—60b ult. covering the following sections of the Physics: _____________ 160

Cf. Les manuscrits arabes de l’Escurial. Décrits d’après les notes de Hartwig Derenbourg, revues et complétées par le Dr. H.-P.-J. Renaud. Tome II, fasc. 2, Paris 1941, p. 112—114. 161 I am grateful to Professor Gerhard Endress, who kindly placed these photos at my disposal. 162 The ‫خ‬-notes occur on fols. 50b14, 53b4, and 58a8(bis). The last two notes offer the correct reading as opposed to erroneous wordings in the main text. At both places, the Leiden manuscript preserves the correct reading. The first two notes report evidentially valueless variant readings, while the main text of ms. Esc. 896 and ms. Leiden Or. 583 has the correct wording.

II.4 Ms. Escorial ár. 896

LXXVII

fol. 48a-b: IV 13, 222a27 — 222b17 fols. 49a—55b: IV 9, 217b27 — IV 12, 221a17 fol. 56a-b: V 1, 224a21 — 224b10 fols. 57a—60b: IV 13, 222b17 — IV 14, 224a17 Thus, the original order was as follows: fols. 49—55; four fols. lost (covering Physics IV 12, 221a17 — IV 13, 222a27); fol. 48; fols. 57—60; fol. 56. Ms. Esc. 896 evidently contains the same translation as the Leiden manuscript (including, of course, minor variant readings). The most striking difference from the Leiden manuscript is that the Escorial manuscript contains no commentary section at all. As a natural consequence, the introductory markers of the Aristotelian lemmata (qāla Arisṭūṭālīs) are omitted here. Furthermore, it lacks the subheadings dividing the text into “lectures” (taʿālīm) as found in the corresponding section of ms. Leiden Or. 583. It does, however, indicate the end of bk. IV and the beginning of bk. V of the Physics, and this in an even more complete form than the Leiden manuscript (al-Maqālatu l-ḫāmisatu min Kitāb al-Samāʿi lṭabīʿiyyi li-Arisṭū, fol. 56a, as opposed to al-Maqālatu l-ḫāmisatu qāla Arisṭūṭālīs in ms. Leiden Or. 583). All of this doubtless points to a separate circulation of the Arabic translation of the Aristotelian Physics without any accompanying commentaries 163. However, whether this separate circulation is the result of a continuous independent transmission of the translation or, conversely, based on a “post-Baghdad” extraction of the Aristotelian lemmata from the Baghdad compilation is difficult to assess. I shall argue below (chapter VII.3) in favour of the former without being able to put forward incontrovertible proofs. As for the stemmatic relationship between ms. Esc. 896 and ms. Leiden Or. 583, each witness displays separative errors against the other, as becomes clear from the following examples. 1. Separative errors in ms. Esc. 896 against ms. Leiden Or. 583: ka-mā lā ms. Esc., fol. 50a1 : ka-mā ms. Leiden, fol. 92b11 (ed. Badawī, p. 406.9) ṯumma kāna mā qablu wa-mā baʿdahumā fī l-āni bi-ʿaynihī om. hom. ms. Esc., fol. 50a9 : habet ms. Leiden, fol. 92b17 (ed. Badawī, p. 407.3-4) fa-inna bihī ms. Esc., fol. 48b3 : fa-innahū ms. Leiden, fol. 108a ult. (ed. Badawī, p. 465.9) _____________ 163

This is in accordance with the fact that neither of the two major Arab Aristotelians of al-Andalus, Ibn Bāǧǧa and Ibn Rušd, exhibits the vaguest acquaintance with the commentaries of the Baghdad Aristotelians on the Physics.

LXXVIII

Introduction

wa-huwa fī zamānin wa-yatakawwanu ms. Esc., fol. 48b ult. : wa-fī zamānin yatakawwanu ms. Leiden, fol. 108b7-8 (ed. Badawī, p. 466.10) 2. Separative errors in ms. Leiden Or. 583 against ms. Esc. 896: fī hāḏā l-waqti wa-qad kāna fī-mā taqaddama mawǧūdan om. hom. ms. Leiden (post mawǧūdan ed. Badawī, p. 406.5) : habet ms. Esc., fol. 49b1112 ḥarakatu l-kulli ms. Leiden, fol. 94a15 (ed. Badawī, p. 411.13) : kuratu l-kulli ms. Esc., fol. 50b6 taqaḍḍin ms. Leiden, fol. 108a20 (ed. Badawī, p. 465.4) : inqiḍāʾun ms. Esc., fol. 48a14 yūḍaʿu ms. Leiden, fol. 115a7 (ed. Badawī, p. 492.15) : wa-maʿa ms. Esc., fol. 56b3 3. On the other hand, mss. Esc. 896 and Leiden Or. 583 share some errors and peculiarities which may point to a common 10th century ancestor (henceforth: Θ): Physics IV 12, 220b12-13: Both ms. Esc. 896 (fol. 54b) and ms. Leiden Or. 583 (fol. 102a ult., ed. Badawī, p. 442.7) read al-baqar (a corruption of al-bašar?) for the Greek ἀνθρώπων, while the two Arabo-Latin translations have the correct “hominum”. Physics IV 13, 222b23-24: Both ms. Esc. 896 (fol. 57a) and ms. Leiden Or. 583 (fol. 108b, ed. Badawī, p. 467.5-6) display in textu the same saut du même au même with a corresponding addition in margine: γίγνεται μὲν οὐδὲν ἄνευ τοῦ κινεῖσθαί πως αὐτὸ καὶ πράττειν, φθείρεται δὲ καὶ μηδὲν κινούμενον laysa min šayʾin yatakawwanu dūna an yataḥarraka ḍarban mina l-ḥarakati [wa-yaʿmala wa-qad yafsudu l-šayʾu wa-in lam yataḥarrak aṣlan ḍarban mina l-ḥarakati] (transl. Isḥāq) In both manuscripts the text printed between square brackets (corresponding to καὶ πράττειν ... κινούμενον) is omitted in the main text and added — apparently by the main scribe of each manuscript— in the margins164. Physics IV 14, 223b1 ἀλλ’ ἔστι νῦν κινεῖσθαι (κινεῖσθαι ms. G Philop. Arab. : καὶ κινεῖσθαι ms. H Simpl. : κεκινῆσθαι codd. cett.) καὶ ἄλλο _____________ 164

Nevertheless ignored by Badawī.

II.4 Ms. Escorial ár. 896

LXXIX

The meaning of this clause is ambiguous, depending on whether κινεῖσθαι is taken (together with νῦν) as (i) the subject of ἔστι or as (ii) subordinate to impers. ἔστι c. inf. It can mean: (i) “But being now in motion is (or: means) also something else”, or: (ii) “But it is possible for another thing as well to be moved now” In our two Arabic manuscripts the translation of this clause runs as follows (ms. Esc. 896, fol. 58b12; ms. Leiden Or. 583, fol. 111b2, ed. Badawī, p. 477.10-11): ġayra annahū qad yakūnu an yataḥarraka l-āna šayʾun āḫaru ayḍan This was certainly not what Isḥāq had noted down. However, it is difficult to assess in which of the above-mentioned ways he interpreted the clause. In accordance with the first interpretation, one would expect the following text: (i) ġayra annahū qad yakūnu an yataḥarraka l-āna šayʾan āḫara ayḍan Since Isḥāq renders impers. ἔστι c. inf. (“it is possible”) usually through qad yumkinu an (cf. Glossary), his translation according to the second mode of interpretation would have run as follows: (ii) ġayra annahū qad yumkinu an yataḥarraka l-āna šayʾun āḫaru ayḍan 165 In all likelihood, Isḥāq decided in favour of the first option. This is not only the more complicated syntactic construction (with an yataḥarraka for the nominalized infinitive), which might have caused the error in Θ, but also the reading confirmed by the Arabo-Latin translations by Gerard of Cremona and Michael Scot (on which see below, chapter VII): “Verumtamen est ut moueatur nunc aliquid aliud” (transl. Gerard of Cremona) “Sed est nunc motum esse aliquid aliud” (transl. Michael Scot) Hence, this error points to Θ as the common ancestor of the two extant Arabic manuscripts and simultaneously shows that the manuscripts used by Gerard of Cremona and Ibn Rušd (viz. Michael Scot) probably do not derive from Θ166. _____________ 165

The scribe or a later reader of ms. Esc. 896 realized that something is wrong with the clause as written in the main text and proposed (‫ )ظ‬the reading yumkinu between the lines over yakūnu. A similar interpretation is found in the margin of the Leiden manuscript: yaqūlu innahū qad yaǧūzu an yataḥarraka fī waqtin wāḥidin šayʾāni muḫtalifāni (“he says that two different things may be moved in one [and the same point of] time”). 166 Interestingly, both Latin translations omit ayḍan.

LXXX

Introduction

Physics V 1, 224b8-9 διὸ καὶ ἡ φθορὰ εἰς τὸ μὴ ὂν μεταβολή ἐστιν Ms. Esc. 896 (fol. 56b13) and ms. Leiden Or. 583 (fol. 116a3, ed. Badawī, p. 495.13-14) read: wa-ka-ḏālika fa-inna l-fasāda innamā huwa l-taġayyuru ilā ġayri mā huwa mawǧūdun Evidently, wa-ka-ḏālika is a mistranscription of wa-li-ḏālika (= διὸ or διὸ καὶ167) that occurred already in Θ. Again, this error does not appear in the two Arabo-Latin translations: “Et propter illud (= wa-li-ḏālika) corruptio non est mutatio nisi ad illud quod non est ens” (transl. Gerard of Cremona) “Quapropter et (= wa-li-ḏālika) corruptio in id quod non est mutatio est” (transl. Michael Scot) The cumulative evidence of the above cases points to a common ancestor Θ of our two Arabic manuscripts. This assumption is corroborated by the fact that we encounter in mss. Esc. 896 and Leiden Or. 583 at corresponding places of the Aristotelian text marginal and interlinear notes which evidently trace back to one and the same source. In all likelihood, these glosses formed already part of Θ. Three examples: (i) παρὰ πᾶσιν, Physics IV 10, 218b13, is rendered through ʿinda kulli šayʾin in the main text (ed. Badawī, p. 412.9-10 = ms. Esc. 896, fol. 50b14). In the Leiden ms. (fol. 94a20) we read above ʿinda the words bi-l-qiyāsi lanā, where lanā is a scribal error for ilā. In other words, the interlinear note proposes bi-l-qiyāsi ilā as alternative translation of παρὰ (instead of ʿinda). Almost the same note, now with the complete phrase bi-l-qiyāsi ilā kulli šayʾin, is found at this place in the margin of ms. Esc. 896. (ii) Isḥāq renders ἐν Σαρδοῖ, Physics IV 11, 218b24, through bi-l-baladi l-musammā Sardū (“in the land/town called Sardō”, ed. Badawī, p. 414.6-7168 = ms. Esc. 896, fol. 51a10). In the margin of ms. Leiden Or. 583 we read yuẓannu anna hāḏā l-balada Ḥarrān (“it is said that this town is Ḥarrān”), and similarly in the margin of ms. Esc. 896 ẓ [= ẓāhirun] Sard Ḥarrān (“obviously Sard[ō] [refers to] Ḥarrān”). (iii) In the margin of Physics IV 11, 220a23-24, which is rendered faithfully by Isḥāq, we read in ms. Leiden Or. 583 (fol. 99a, upper margin, ed. Badawī, p. _____________ 167 168

καὶ is omitted in a number of manuscripts, among others in ms. E. Badawī misinterpreted the final Wāw of Sardō as the conjunction wa-.

II.4 Ms. Escorial ár. 896

LXXXI

4322) the note: ka-mā tuqālu mina l-wāḥidi waḥdatun qulnā mina l-ʿašarati ʿašriyyatan (sic leg. pro ʿašaratan) (“as the state of being one is predicated of the one [thing], [so] we predicate the state of being ten of ten [things]”). This note is marked by the letter ‫ح‬. Almost the same note (without the mark ‫ )ح‬is found at the same place in ms. Esc., fol. 53b9: ka-mā naqūlu mina l-wāḥidi waḥdatan wamina l-ʿašarati ʿašriyyatan (“as we predicate the state of being one of the one [thing] and the state of being ten of ten [things]”). The first two notes doubtless point to an Arabic background of their formation. The third one has no parallel in any extant Greek commentary. None of the glosses is reflected in the Arabo-Latin transmission of the text. However, whether Θ comprised only the Aristotelian text or also the Baghdadian commentaries, we don’t know. In general, ms. Esc. 869 is an important witness for the passage in question. It seems to be slightly more reliable than the Leiden manuscript, notably as regards omitted or wrong case endings or a certain carelessness with small conjunctions and particles in the latter. For an attempt at locating the stemmatic position of ms. Esc. 896 (and Θ), see below, chapter VII.3.

III. The author of the Arabic translation: Isḥāq ibn Ḥunayn Aristotelian physics has attracted the attention of Arabic writing Muslim and Christian scholars since the earliest period of the Graeco-Arabic translation movement169. The only extant complete medieval Arabic translation of Aristotle’s Physics is unanimously ascribed to Isḥāq ibn Ḥunayn (d. 910)170. The _____________ 169

For the beginning of the reception and the earliest (nonextant) Arabic translations see Francis E. Peters, Aristoteles Arabus. The Oriental Translations and Commentaries on the Aristotelian Corpus. Leiden 1968, p. 32; Dimitri Gutas, Greek Thought, Arabic Culture, London 1998, p. 69–73. Although peripheral in the present context, it may be worth mentioning that the Ottoman “revival” of the translation movement caused another, hitherto relatively unknown (partial) Arabic translation of Aristotle’s Physics together with the commentary of Ioannis Kottounios (d. 1657) by Yanyalı (also Yanyavı) Esad Efendi (d. 1143/1731); cf. Mehmet Sait Özervarlı, “Yanyalı Esad Efendi’s Works on Philosophical Texts as Part of the Ottoman Translation Movement in the Early Eighteenth Century”, in Barbara Schmidt-Haberkamp (ed.), Europa und die Türken im 18. Jahrhundert / Europe and Turkey in the 18th Century, Göttingen 2011, p. 457–472. 170 With the exception of erroneous ascriptions to his famous father, Ḥunayn ibn Isḥāq, such as in Bernard Carra de Vaux, Avicenne (Les grands philosophes, vol. 3), Paris 1900, p. 58, Auguste Mansion, “Étude critique sur le texte de la Physique d’Aristote (L. I-IV)”, Revue de philologie. Nouvelle série 47 (1923), p. 5—41, there p. 9, 14, etc., or W. D. Ross,

LXXXII

Introduction

earliest implicit evidence for this ascription comes from no less a figure than Isḥāq ibn Ḥunayn himself. As mentioned above, the Leiden manuscript contains eighteen marginal and interlinear notes beginning with “Isḥāq”. Since some of these notes are self-referential in such a way that its author refers to and/or quotes his own translation, we can rule out the possibility that Isḥāq annotated here somebody else’s translation. The label “Isḥāq” thus refers to Isḥāq ibn Ḥunayn as the author of the note and the author of the translation. This evidence is additionally supported by a number of notes —likewise labelled with “Isḥāq”— which seem to trace back to corresponding notes in the translator’s Greek exemplar(s). As it is not very likely that such notes originated from the later members of the Baghdad school of Aristotelians who mainly worked with Syriac sources (other than Arabic171) or from the scribe of our manuscript, it is highly probable that such notes were immediately taken down in the process of translation by the translator, Isḥāq ibn Ḥunayn, himself. Obviously, Isḥāq ibn Ḥunayn ventured to supplement his translation here and there with brief comments172. An example of such self-referential notes is _____________ Aristotle’s Physics, p. 114. That the translation preserved in ms. Leiden Or. 583 does not originate from Ḥunayn’s hand is also clear from a comparison of Ḥunayn’s authentic translation of Galen’s De elementis secundum Hippocratem, which contains —as kindly brought to my attention by Nicolás Bamballi (Österreichische Akademie der Wissenschaften, Vienna)— two quotations of Aristotle’s Physics translated in a way considerably different from what we read in the Leiden manuscript (cf. the translations of Physics I 2, 184b25—185a1 = ed. Badawī, p. 7.7-8 and De elem. 5.2, ed. De Lacy, p. 92.16-17 = ed. Sālim, p. 61.4-5, and I 2, 185a1-3 = ed. Badawī, p. 7.8-11 and De elem. 5.12, ed. De Lacy, p. 94.21-23 = ed. Sālim, p. 64.-3—ult.). 171 Cf. Richard Walzer, “New Light on the Arabic Translations of Aristotle”, Oriens 6 (1953), p. 91—142, esp. p. 96. 172 In all likelihood, he had acquired this habit from his father, Ḥunayn ibn Isḥāq, whose explanatory notes of most akin style accompany for example his translations of Galen’s Commentary on Hippocrates’ Aphorisms, cf. Taro Mimura (ed., together with P. Pormann, K. Karimullah et al.), Book I, (https://www.research.manchester.ac.uk/portal/files/ 57968806/Galen_commentaries_book_1.pdf, p. 45ult.—46.1, 51.5-6, 60ult.), Book VI (https://www.research.manchester.ac.uk/portal/files/57968950/Galen_commentaries_book_ 6.pdf, p. 23.3-6, 63.7ff.); as well as his translations of Galen’s De partibus artis medicativae and De causis contentivis; cf. Malcolm Lyons (ed.), Galen. On the Parts of Medicine. On Cohesive Causes. On Regimen in Acute Diseases in Accordance with the Theories of Hippocrates (Corpus Medicorum Graecorum. Supplementum Orientale; vol. 2). Berlin 1969, p. 26.5-6, 18-21; 30.3-4; 32.11-13; 34.11-13, 15-16; 62.23-24. For other examples of notes jotted down by Isḥāq ibn Ḥunayn at the margins of his own translations cf. Gutas, Theophrastus, p. 123, note 44, p. 294, and Georr, Les Catégories d’Aristote, p. 181 (No 130). Cf. also below, footnotes 179, 220.

III. The author of the Arabic translation

LXXXIII

found at fol. 146a10-13. We read there in the main text Isḥāq’s translation of Physics V 6, 230a29-32: [1] ἢ εἰ ἔστιν τὸ βίᾳ παρὰ φύσιν, καὶ φθορὰ ἂν εἴη φθορᾷ ἐναντία ἡ βίαιος ὡς παρὰ φύσιν οὖσα τῇ κατὰ φύσιν; ἆρ' οὖν καὶ γενέσεις εἰσὶν ἔνιαι βίαιοι ...; fa-naqūlu fī ḏālika innahū in kāna mā yakūnu qasran fa-huwa ḫāriǧun ʿani lṭabīʿati fa-qad yakūnu fasādun ḍiddan li-fasādin aʿnī l-fasāda qasran ʿalā annahū amrun ḫāriǧun ʿani l-ṭabīʿati li-l-fasādi l-ṭabīʿiyyi fa-qad yakūnu ayḍan aṣnāfun mina l-takawwuni qasran ... (ed. Badawī, p. 591.8-11) “We reply to this that if what happens violently is unnatural then there might be perishing contrary to perishing, I mean violent perishing insofar it is something unnatural [contrary] to natural perishing. But then there might also be cases of violent becoming ...” The beginning of the last sentence (fa-qad yakūnu) is marked with a superscript sign, and in the margin we read (with a corresponding mark) the following: Isḥāq: yaǧūzu an yunqala hāḏā l-lafẓu makāna “fa-qad yakūnu” “ḥattā yakūna” – innamā qāla ʿalā sabīli baḥṯin [or, in Middle Arabic usage, baḥṯi] wa-naqdi šayʾin [sic? 173] lā an yuḥkama bihī “Isḥāq: this phrase [scil. ἆρ' οὖν ... εἰσὶν] might be rendered not by ‘but then there might be’, but by ‘such that there will be’ – however, he [Aristotle] speaks in the mode of investigating and (?) questioning something rather than establishing it [as a fact].” In addition to providing a nice example of Isḥāq’s painstaking sensitivity to the varying use and meaning of Greek particles (here the conclusive and interrogative/expectant aspects of ἆρ' οὖν174), the phrasing of the note suggests that Isḥāq quotes here his own translation in order to provide the reader with a slightly different, yet less likely interpretation. Another example occurs in a commentary section by Abū Bišr Mattā ibn Yūnus (ms. Leiden Or. 583, fol. 149a, ed. Badawī, p. 600). Abū Bišr first quotes Isḥāq’s translation of Physics V 6, 230b26-28: _____________ 173

The phrase wa-naqdi šayʾin is not clear in the manuscript; Badawī reads instead watafsīrin. (I am grateful to Geoffrey Moseley for his comment on this passage.) 174 On which see J. E. Harry, “Indicative Questions with μή and ἆρα μή”, in Studies in Honor of Basil L. Gildersleeve, Baltimore 1902, p. 427—434; John Dewar Denniston, The Greek Particles, 2nd ed., Oxford 1950, p. 46—48.

LXXXIV

Introduction

[2] ἔτι δοκεῖ τὸ ἵστασθαι ἢ ὅλως εἶναι τὸ εἰς τὸν αὑτοῦ τόπον φέρεσθαι ἢ συμβαίνειν ἅμα wa-qad yuẓannu ayḍan anna qawlanā fī l-šayʾi yatawaqqafu immā an yakūna maʿnāhu ay yantaqilu bi-l-ǧumlati ilā l-mawḍiʿi llaḏī huwa lahū wa-immā an yalzama maʿan (cf. ed. Badawī, p. 598.10—599.3) Before going on with his own rephrasing of the Aristotelian statement (ed. Badawī, p. 600.8-10), Abū Bišr intersperses the following remarks: Isḥāq: hāḏā ḥallu l-šakki wa-qawluhū “ayḍan” ḥašwun “Isḥāq: this is the solution of the aporia and [Aristotle’s] expression ‘furthermore’ [ayḍan = ἔτι] is redundant.” Evidently, Isḥāq ibn Ḥunayn had supplied his translation with a remark on “ayḍan” [= ἔτι] pointing to the fact that the sentence (quoted here by Abū Bišr Mattā) is part of the solution of the aporia introduced at 230b21175. This solution starts one line earlier with the words οὐ γενόμενον ἄρα ἠρεμοῦν ἔσται ἠρεμοῦν (230b25-26). The following ἔτι seems to be out of place, because it draws on a position which is part of the aporia, not part of its solution (Carl Prantl was aware of this difficulty and proposed reading εἰ instead of ἔτι; ms. E1 has τι). Being a reliable translator, Isḥāq nevertheless rendered ἔτι faithfully by “ayḍan” and warned the reader in a marginal note referring to his own translation. Most of the remaining notes labelled with “Isḥāq” are of commentarial or exegetical character. At least some of them possibly draw on similar notes or glosses in Isḥāq’s Greek exemplar(s). In the following three cases, the first of which is explicitly adduced as an annotation or scholium (taʿlīq) by Isḥāq, we encounter striking parallels in John Philoponus’ commentary on the Physics. [3] The translation of Physics II 8, 199b7-9 “Further, seed must have come into being first, and not straightaway the animals: what was undifferentiated first (“οὐλοφυὲς μὲν πρῶτα”, Empedocles, frag. 62 Diels-Kranz) was seed.” (transl. Hardie-Gaye), is followed by a note from Abū ʿAlī ibn al-Samḥ’s hand (ms. Leiden Or. 583, fol. 29b, ed. Badawī, p. 153.16-17 and note 3176): Abū ʿAlī: hāḏā l-taʿlīqu li-Isḥāq: hāḏā lafẓu Anbāduqlis wa-huwa yaʿnī bihi llaḏī huwa min ṭabīʿatin wāḥidatin kulluhū lā min ḥayawānayni muḫtalifayni miṯla waladi l-baqarati llaḏī wuṣifa _____________ 175 176

On which see Ross, Aristotle’s Physics, p. 637 ad loc. Badawī did not recognize that the note begins in textu and is resumed in the margin.

III. The author of the Arabic translation

LXXXV

“Abū ʿAlī: Isḥāq [made] the following annotation: this [i.e. οὐλοφυής] is an expression [coined] by Empedocles and he means by it that which is entirely of one and the same nature, not from two different living beings, like the cow’s offspring which was mentioned [above] (i.e. τὰ βουγενῆ, 198b32).” No matter whether Isḥāq ibn Ḥunayn knew or had access to Empedocles’ poems (presumably he did not), it is clear that it must have been a scholar brought up in the environment of Greek paideia in the (late) ancient Greekspeaking world, who felt compelled to comment upon Aristotle’s allusion in this way. As a matter of fact, we find a similar note explaining the term οὐλοφυής in the sense of ὁλοφυής in Philoponus’ commentary on these lines (In Arist. Phys., p. 319.29—320.3): ὁ Ἐμπεδοκλῆς τὸ “οὐλοφυὲς“ πρῶτόν φησι γενέσθαι ...· παρὰ γὰρ τὴν τοῦ ὅλου φύσιν περιέχειν τὸ ὁλοφυὲς αὐτῷ κέκληται. “Empedocles says that ‘the undifferentiated’ came first ... for ‘the undifferentiated’ is so called by him from its containing the nature of the whole.” [4] In the margin of the Arabic translation of Physics VI 4, 235a13-16 “And since everything that is in motion is in motion in a certain sphere and for a certain time and has a motion belonging to it, it follows that the time, the motion (τῆς κινήσεως), the being-in-motion (τοῦ κινεῖσθαι), the thing that is in motion, and the sphere of the motion must all be susceptible of the same divisions.” (transl. Hardie-Gaye), we read (ms. Leiden Or. 583, fol. 165b): Isḥāq: al-farqu bayna l-ḥarakati wa-l-taḥarruki anna l-ḥarakata ka-annahū aqāmahā maqāma mā qad ḥaṣala wa-ṣāra fīhi wa-l-taḥarruka ka-annahū šayʾun dāʾibun fī l-fiʿli “Isḥāq: the difference between motion and being-in-motion is that he [Aristotle] takes motion as if it is something that might happen and become, whereas [he takes] being-in-motion as if it is something that is carried out in actuality.” In all likelihood, Isḥāq translated or summarized here a marginal gloss of his Greek exemplar explaining the difference between motion and being-inmotion. The author of this gloss must have been acquainted with Philoponus’ commentary on this passage, where we read that Aristotle takes motion as a sort of ἕξις and being-in-motion as actuality (τὸ μὲν ὡς ἕξιν τινὰ λαμβάνει

LXXXVI

Introduction

τὴν κίνησιν, τὸ δὲ κινεῖσθαι ἐνέργειαν, In Arist. Phys., p. 865.26-27)177. A comparable remark to this effect is not found in the commentaries on this section by Themistius or Simplicius. [5] Another example is found in the context of Physics VI 5, 236b1-2 αὐτὸ δὲ ὃ μεταβάλλει ἢ καθ' ὃ μεταβάλλει, οὐκέθ' ὁμοίως ἕξει “With regard, however, to the actual subject of change —that is to say [ἢ] that in respect of which a thing changes— there is a difference to be observed.” (transl. Hardie-Gaye) Above the Arabic translation of αὐτὸ ... ὃ μεταβάλλει we read (ms. Leiden Or. 583, fol. 171a): Isḥāq: yaʿnī l-ṣūrata Due to lack of space this note is extended into the margin as follows (ed. Badawī, p. 6792): wa-qawluhū “aw” makāna an yaqūla “aʿnī” “Isḥāq: he [Aristotle] means [with ‘the actual subject of change’] the form [...] and he says ‘or’ [ἢ] instead of saying ‘that is to say’.” Again, this note seems to trace back to interlinear or marginal Greek glosses derived from Philoponus’ commentary, which says at the present place (In Arist. Phys., p. 868.22-24): εἰπὼν δὲ αὐτὸ ὃ μεταβάλλει, ἐπειδὴ αὐτὸ τὸ μεταβάλλον τὸ πρᾶγμά ἐστι, διὰ τοῦτο προσέθηκεν ἢ καθ' ὃ μεταβάλλει, ἵνα σημάνῃ τὸ εἶδος.178 The most plausible way to explain the fact that these glosses are labelled with “Isḥāq” instead of “Yaḥyā” (i.e. John Philoponus) or any of the Arab commentators, who had access to at least large parts of Philoponus’ commentary, is to assume that they derive not from an Arabic or Syriac adaptation of this commentary, but rather from notes taken down by Isḥāq ibn Ḥunayn in the process of translation and on the basis of corresponding Greek glosses in his exemplar(s) of Aristotle’s Physics. Together with the notes discussed previously, they supply further evidence for the identity of their author and the author of the translation transmitted in ms. Leiden Or. 583, namely Isḥāq ibn Ḥunayn. _____________ 177

Philoponus, in turn, draws probably on Alexander; cf. Rashed, Alexandre d’Aphrodise, p. 353, frag. 310. 178 A similar remark on the meaning of ἢ καθ' ὃ μεταβάλλει is found in Simplicius (In Arist. Phys., p. 988.9-10). However, Simplicius does not equate αὐτὸ ὃ μεταβάλλει with “form”.

III. The author of the Arabic translation

LXXXVII

Twelve other notes labelled with “Isḥāq” are found in the margins or between the lines of the translation of the Aristotelian text on fols. 121b (on 225b21-22), 137a (on 228a28-29), 142a (on 229a27-28), 149a (on 230b32—231a1), 156b (on 233 b6-7), 159a (on 233b35—234a3), 159b (on 234a21-22), 161a (on 234a34-35) of this manuscript; one further note (fol. 160b) occurs within another commentary section by Abū Bišr Mattā immediately after the lemma quoting 234a13. The majority of these 13 notes consist of short explanations of single words in the Arabic translation according to the pattern Isḥāq: qawluhū X yaʿnī bihī Y (“Isḥāq: by X [Aristotle] means Y”) or, without mention of the keyword, simply Isḥāq: ay/yaʿnī ... (“Isḥāq: i.e./he means ...”)179. Isḥāq’s authorship is not only confirmed by the marginal notes discussed above. It is also acknowledged by the Baghdad Aristotelians involved in the corpus transmitted in ms. Leiden Or. 583. (i) Yaḥyā ibn ʿAdī: At the end of Physics bk. I the scribe of the Leiden manuscript reproduces the colophon written by Abū l-Ḥusayn al-Baṣrī180. Among other things, Abū l-Ḥusayn mentions Yaḥyā ibn ʿAdī as a witness of Isḥāq ibn Ḥunayn’s authorship of the present translation: “I have collated Yaḥyā ibn ʿAdī’s copy with the text [contained] in this quire. [Yaḥyā ibn ʿAdī] mentioned (ḏakara) that he had transcribed his copy from the ‘master copy’ (dastūr) of Isḥāq[̕s translation] and collated it with this three times, and a fourth time with the Syriac.” No matter what kind of manuscript the term dastūr refers to, there can be little doubt that Yaḥyā ibn ʿAdī was convinced to transcribe Isḥāq’s translation of the Physics. This information is absolutely in accordance with the fact that Yaḥyā ibn ʿAdī quotes or paraphrases in a number of his genuine treatises Aristotle’s Physics in the translation ascribed to Isḥāq ibn Ḥunayn181. _____________ 179

Another note following this pattern, yet introduced by the formula “fī naqli Isḥāq” is mentioned below, p. LXXXVIII, section (iii.2). Additionally, the manuscript contains a large number of marginal and interlinear notes introduced simply by ay/yaʿnī ... (“i.e./he means ...”) without any reference to their author(s). There is good reason to believe that at least some of them trace back to Isḥāq ibn Ḥunayn, too (on these and other marginal notes cf. also below, chapter IV). 180 Arabic text Badawī, al-Ṭabī ʿa, p. 76. 181 Cf. Endress, The Works of Yaḥyā Ibn ʿAdī, p. 53—62, 69; idem, Die Bagdader Aristoteliker, p. 310—311; Wisnovsky, New Philosophical Texts, p. 307—326, and below, chapter VII.1.

LXXXVIII

Introduction

(ii) Abū l-Ḥusayn al-Baṣrī: The editor of the Baghdad compilation himself mentions Isḥāq as the author of the translation incorporated into his compilation at least two times, namely in the headings of bk. III (naql Isḥāq ibn Ḥunayn, fol. 32b2, ed. Badawī, p. 165) and bk. VI (naql Isḥāq, fol. 150b2, instead of binaql Isḥāq, ed. Badawī, p. 605). In addition, we have the testimony in the colophon at the end of bk. I which either derives directly from Abū l-Ḥusayn alBaṣrī or from the scribe of the Karḫ-copy, probably a disciple of Abū l-Ḥusayn alBaṣrī182. (iii) Anonymous references: The Leiden manuscript contains five anonymous references to “Isḥāq’s translation” (naql Isḥāq), which in all likelihood originate from one (or more) of the main commentators, Abū Bišr Mattā ibn Yūnus, Yaḥyā ibn ʿAdī, Abū ʿAlī ibn al-Samḥ, or Abū l-Faraǧ ibn al-Ṭayyib. It becomes clear from the contents of these notes that their author(s) also had access to two other sources containing Aristotle’s Physics ascribed to Qusṭā ibn Lūqā and Abū ʿUthmān Saʿīd ibn Yaʿqūb al-Dimašqī, as well as to (excerpts of ?) Philoponus’ commentary. In the following four cases the notes are based on collation with other Arabic versions and call attention to a wording or interpretation deviating from Isḥāq’s translation: (iii.1) fol. 7a in marg. (on Physics I 3, 186b17; cf. Badawī, al-Ṭabīʿa, p. 271): fī naqli Isḥāq “wa-immā” wa-fī naqli Qusṭā “immā” (note: to be expected faimmā = ἢ οὖν). (iii.2) fol. 114a in marg. (on V 1, 224a29: κατ' ἄλλην δὲ κίνησιν ἕτερον = wa-qad yakūnu ġayran min qibali ḥarakatin uḫrā, transl. Isḥāq, fol. 114a9-10, Badawī, al-Ṭabīʿa, p. 4891): fī naqli Isḥāq “ay ġayra mutaḥarrakin bi-l-ḏāti” – fī naqli Qusṭā badalan min qawlihī “wa-qad yakūnu ġayran” “wa-qad yataḥarraku šayʾun āḫaru ʿalā ṭarīqin āḫara mina l-ḥarakati” (in addition to quoting Isḥāq’s translation this note reports also another marginal piece of advice by Isḥāq on how to understand the absolute use of ἕτερον = ġayran). (iii.3) fol. 133b22-23 (Badawī, al-Ṭabīʿa, p. 553.2-3) at the end of a commentary section (on V 4, 227b13-14: ἐπιστήμη εἶδος μὲν ὑπολήψεως = ʿilman fainnahū nawʿun li-l-ẓanni, transl. Isḥāq, fol. 133a10, Badawī, al-Ṭabīʿa, p. 551.5-6): _____________ 182

Cf. above, p. XLVII.

III. The author of the Arabic translation

LXXXIX

fī naqli l-Dimašqī anna l-ʿilma taḥta l-fāʾidati [ὑπόληψις = “subsidy”, “benefit”] wa-huwa aǧwadu min naqli Isḥāq anna l-ʿilma taḥta l-ẓanni [ὑπόληψις = “assumption”, “opinion”]. (iii.4) fol. 190b14-17 (Badawī, al-Ṭabīʿa, p. 754.13-14, on VII 2, [version β] 243b29: καὶ πᾶσα δὴ κίνησις ἡ κατὰ τόπον σύγκρισις καὶ διάκρισίς ἐστιν = wa-kullu ḥarakatin fī l-makāni fa-hiya ǧamʿun wa-tafrīqun, transl. Isḥāq, fol. 189a21, Badawī, al-Ṭabīʿa, p. 748.9-10): fī naqli Isḥāq wa-l-Dimašqī “wa-kullu ḥarakatin fī l-makāni fa-hiya ǧamʿun wa-tafrīqun”, wa-fī naqli l-Dimašqī ziyādatun hiya “ḫalā mā kāna minhā fī l-kawni wa-l-fasādi” (the author of this note reports an addition in alDimašqī’s version which corresponds exactly to the sentence VII 2, [version α] 243b9-10 : πλὴν ὅσαι ἐν γενέσει καὶ φθορᾷ εἰσιν. This sentence is without an equivalent in version β of bk. VII, the version translated by Isḥāq ibn Ḥunayn)183. The fifth reference to Isḥāq’s translation occurs in parenthesis in a comment by John Philoponus. The reason for this unusual placement is the incoherent Arabic rendering of Aristotle’s concept τὸ ἄρτι (the “just now”) in the Arabic version of Philoponus’ commentary and Isḥāq’s text. Isḥāq renders τὸ ἄρτι through qubaylu (cf. Badawī, al-Ṭabīʿa, p. 466.3, 467.12 = IV 13, 222b12, b28); the paraphrase of Philoponus’ commentary has instead al-sāʿata. Without this additional hint, the reader of Isḥāq’s translation would have difficulty identifying the exact passages Philoponus is talking about. The relevant section, which draws on Philoponus’ In Arist. Phys., p. 760.19—761.30, begins as follows (ms. Leiden Or. 583, fol. 108b16-18, ed. Badawī, p. 467.13-16): (iii.5) qāla Yaḥyā: ġaraḍu Arisṭūṭālīs fī hāḏā l-taʿlīmi an yuʿarrifanā maʿāniya alfāẓin zamāniyyatin miṯla l-āna wa-matā wa-huwa ḏā wa-l-sāʿata —fī naqli Isḥāq “qubaylu” badalan min “al-sāʿata”— wa-qabla an yufīdanā ḏālika yatakallamu fī l-āna fa-yaqsimuhū qismayni ... (cf. πρόκειται νῦν αὐτῷ περὶ τῶν χρονικῶν προσρημάτων διαλαβεῖν, τί ἕκαστον σημαίνει, οἷον τί ἐστι τὸ νῦν, τί τὸ ποτέ, τί τὸ ἤδη, τί τὸ ἄρτι, [τί τὸ πάλαι, τί τὸ ἐξαίφνης]. πρῶτον δὲ περὶ τοῦ νῦν διαλέγεται, καί φησι διττὸν εἶναι τὸ νῦν ..., Philop. In Arist. Phys., p. 760.19-22) The structure of these five references shows that their author(s) simultaneously employed an Arabic version of Philoponus’ commentary, two works (yet _____________ 183

For more details on this note cf. Arzhanov & Arnzen, Die Glossen in Ms. Leyden Or. 583, p. 455—457.

XC

Introduction

possibly only one authoritative text translated by one of them and revised by the other, or translated independently twice) by Qusṭā ibn Lūqā and Abū ʿUthmān al-Dimašqī, which included at least parts of the Aristotelian text, and a translation of the Aristotelian text unanimously ascribed to Isḥāq ibn Ḥunayn, which evidently coincides with the translation transmitted in the lemmata of the Baghdad compilation. Furthermore, they rule out any speculation as to whether “Isḥāq’s translation” (naql Isḥāq) might refer to a Syriac translation of the Aristotelian text. Additional evidence for Isḥāq’s authorship is provided by linguistic and terminological peculiarities of the translation. Two valuable and convincing studies on this issue have been published recently. Dimitri Gutas shows in his edition of Theophrastus’ On first principles the consistent and characteristic use of a number of functional words and phrases in the Physics and Theophrastus’ essay, the translation of which is likewise attributed to Isḥāq ibn Ḥunayn184. Manfred Ullmann’s study of the Arabic text and transmission of the Nicomachean Ethics contains a detailed comparison between the Arabic equivalents of technical terms, adverbs and functional words in translations attributed to Isḥāq ibn Ḥunayn (especially Aristotle, Categoriae, De Interpretatione, Ethica Nicomachea bks. I—IV, Physica, Metaphysica bk. α, and Themistius, In de anima paraphr.) and their counterparts in translations attributed to Usṭāṯ (Eustathios) (especially Aristotle, Historia animalium, De partibus animalium, De generatione animalium, Ethica Nicomachea bks. V—X, and Metaphysica bks. B—I)185. Since both studies confirm and even reinforce the identity of the translator of the Physics on linguistic grounds, I can confine myself here to some corroborative additions to the lexical peculiarities highlighted there (in alphabetical order, ref. to Greek texts): ἄνευ = ḫilwan min 186 also Physics 256a11, 256a12, 256a25, 258a16, 260b18 (but also twice = min ġayri Physics 258b17, 258b19) ἁπλῶς = ʿalā l-iṭlāq 187 also Physics 258b15, 263b7 (but also = basīṭun 260a17) ἀπορία = šakkun 188 also Physics 255a11, 262b9(bis), 266b28, 267b9 _____________ 184

Cf. Gutas, Theophrastus, p. 85—89. Isḥāq ibn Ḥunayn’s use of discourse markers is also investigated by Kerstin Eksell, “Pragmatic Markers from Greek into Arabic. A Case Study on Translations by Isḥāq ibn Ḥunayn”, Studia graeco-arabica 5 (2015), p. 321—344. 185 Manfred Ullmann, Die Nikomachische Ethik des Aristoteles in arabischer Übersetzung. Teil 2: Überlieferung, Textkritik, Grammatik, Wiesbaden 2012 (henceforth: Ullmann, NE II), p. 15—53. 186 Ullmann, NE II, p. 29. 187 Ullmann, NE II, p. 43. 188 Ullmann, NE II, p. 28—30.

III. The author of the Arabic translation

XCI

ἑκών = ṭawʿan 189 also Physics 251a33 ᾗ = min ǧihati mā /anna190 also Physics 250b30, 251a2, 251a4, 255a13, 255a14, 258a26(bis) ἴσως = aḫliq bi-191 also Them. In de an. 26.31 = 15.12, 82.34 = 142.14, 104.17 = 189.17 καθόλου = ʿalā l-ǧumlati 192 also Physics 264a21, 265a8 κατὰ συμβεβηκός = bi-ṭarīqi l-ʿaraḍi 193 also Physics 254b7-8, 255a26, 256b5-6 + 9 times (cf. Glossary) μάλιστα = ḫāṣṣatan 194 also Physics 253a7, 254b24-25, 254b33, 261a24 (but also = awlā [...] bi-ḏālika 265a5, 267b4) μᾶλλον (in conclusions a minori ad maius) = fa-bi-l-ḥarīyi 195 also Physics 259a12, 261a18, 265a17 οἷον = miṯālu ḏālika196 also Physics 251a14, 251a29, 251b8, 251b32 + 29 times (cf. Glossary) (but also = ka-annaka qulta Physics 255b9, 257a19, 266a15, 266a27, 266b31) ὅλως = bi-l-ǧumlati 197 also Arist. Metaph. 994a19, Physics 252a32, 254a27, 257b 10, 261a13, 263b25, 266a24, 266a28 ὀρθῶς = ṣawābun 198 also Physics 252a33, 252b31 (but also = ṣaḥīḥun 252b2) πλεοναχῶς, πολλαχῶς = ʿalā anḥāʾin šattā 199 also Physics 255a31, 255b17, 260b17 τέλος = ġāyatun 200 also Arist. Metaph. 993b21, 994b9, 994b16, Physics 262b30-31 φαντασία = taḫayyulun 201 also Physics 254a28, 254a29 aʿnī introducing parenthetical clauses202 or bridging Greek anacoluthon also Physics 257a31-32, 261a25-26 _____________ 189

Ullmann, NE II, p. 35. Ullmann, NE II, p. 45. 191 Gutas, Theophrastus, p. 86, Ullmann, NE 2, p. 38–39; cf. Lisān al-ʿArab, ed. Dār alMaʿārif, Cairo, p. 1248a. 192 Ullmann, NE II, p. 44. 193 Ullmann, NE II, p. 41. 194 Gutas, Theophrastus, p. 86–87, Ullmann, NE II, p. 40. 195 Ullmann, NE II, p. 50. 196 Ullmann, NE II, p. 40. 197 Ullmann, NE II, p. 44. 198 Ullmann, NE II, p. 42–43. 199 Ullmann, NE II, p. 43–44. 200 Ullmann, NE II, p. 25. 201 Ullmann, NE II, p. 24. 202 Gutas, Theophrastus, p. 88. 190

XCII

Introduction

layta šiʿrī in indirect/conditional questions203 also Physics 250b11, 252b25, 258a 28-29 mā /allaḏī naḥnu bi-sabīlihī referring to an ongoing discourse/investigation204 also Physics 258b13, Them. In de an. 102.32 = 186.17f. To this list we may add with reservation the following observations which possibly point to further idiosyncracies of Isḥāq ibn Ḥunayn’s translation technique (ref. to Greek texts): ἄπειρος / μυριάκις = ḥ-ṣ-y IV c. neg. („countless“) Arist. Eth. Nic. 1103a22, Metaph. 994b28, Physics 258b32, Theophr.205 11a17 διαιρέω / διορίζω = l-ḫ-ṣ II206 Arist. Eth. Nic. 1101b10, 1109b33, Int. 21a34, Physics 239b13, 251a8, 254a17, 257b6, 262a1, 266a11-12, 267b17-18, Them. In de an. 23.25, 51.27, 108.13, Theophr. 9a18 (cf. 9a25) ἔχω („to be in a certain state“) = (kāna) bi-ḥāli ... / ḥāluhū ... / al-ḥālu fīhi ... Arist. Eth. Nic. 1095b7, 1105a33, 1105b27, Int. 19a3, 19a13, 19b23, Metaph. 993b30, Physics 236b2, 251a5, 251b2, 251b6, Them. In de an. 65.25, 71.19, Theophr. 5b19, 12a2 μόνον = dūna ġayrihī 207 Arist. Cat. 13a1, Physics 261b16, Theophr. 5b11 ὁμοίως = ʿalā miṯālin wāḥidin208 Arist. Cat. 11a8, Eth. Nic. 1094b13, 1098a27, 1098b1, 1101b32, 1114b13, etc.209, Int. 16b15, 18a13, 18b8, 23b33, Metaph. 994a18, Physics 253a27, 260b3, 261b25, 263a14, 265a33, Them. In de an. 22.22, 36.35, 48.9, 48.27, 54.15, Theophr. 5a26, 6b8, 8a3, 9b3. ὑπολαμβάνω = iʿtaqada210 Arist. Eth. Nic. 1098b32, Physics 208a29, 260b22, Them. In de an. 89.6, Theophr. 9b24b In view of the evidence collected, there can be little doubt that Isḥāq ibn Ḥunayn is indeed the translator of the Arabic version of Aristotle’s Physics preserved in _____________ 203

Gutas, Theophrastus, p. 87f., Ullmann, NE II, p. 47. Ullmann, NE II, p. 47–48. 205 Ed. Gutas, v. note 19. 206 As opposed to the more common roots ǧ-z-ʾ, f-ṣ-l and q-s-m for διαιρέω and ḥ‑d‑d, f‑ṣ‑l and m-y-z for διορίζω. 207 Only rarely; Isḥāq uses more frequently faqaṭ. 208 As opposed to ʿalā hāḏā l-miṯāli and various phrases formed with ḥālun, qiyāsun, ka‑, or miṯlun. 209 Cf. Manfred Ullmann, Die Nikomachische Ethik des Aristoteles in arabischer Übersetzung. Teil 1: Wortschatz, Wiesbaden 2011 (henceforth: Ullmann, NE I), p. 370. 210 As opposed to ẓanna, raʾā and others equivalents. 204

III. The author of the Arabic translation

XCIII

ms. Leiden Or. 583, notwithstanding that none of the medieval bibliographers and biographers wastes any words on this fact. Whether his Greek Vorlage consisted in a manuscript or manuscripts containing only the Aristotelian text or a literal commentary quoting the latter in the lemmata, is difficult to assess. Based on the extant medieval biographical reports, Francis E. Peters comes to the bold conclusion “that the Arabs had no freestanding text of the Physics but used one taken from one or other of the Greek commentators”211. In all likelihood, this does not hold true for Isḥāq ibn Ḥunayn’s Vorlage. Judging by the extant commentaries from late antiquity, a complete reconstruction of the Aristotelian text from the lemmata is rather unlikely212, whereas the Arabic translation covers the whole text. Furthermore, the Aristotelian text retrievable in the quotations and the lemmata of the extant commentaries belongs in large part or entirely to another branch of the textual tradition than the one translated by Isḥāq ibn Ḥunayn (see below, chapter VI, and the Greek critical apparatus underneath the Arabic text edition). Finally, we can exclude with certainty the commentaries by Philoponus and Simplicius as potential sources of extraction. The latter, because it was not accessible to the 9th century Arabic-speaking scientific community, the former on grounds of textual peculiarities. To name but the two most conspicuous disparities: As mentioned above, the exemplar of Isḥāq ibn Ḥunayn’s translation contained bk. VII 1—3 of the Physics in the so-called version β, while Philoponus oscillates in the first two chapters (up to 244b17) between versions α and β, and from there on up to the end of chapter 3 follows rather strictly version α. The final section of Physics V (231a5-17) is a sort of “alternative version of 230b10-28” (Ross) omitted in a number of manuscripts. Alexander and Simplicius were familiar with this section, but Philoponus neither quotes nor comments on it. However, Isḥāq ibn Ḥunayn’s translation contains the whole section at its correct place (ed. Badawī, p. 602.11—603.7, notably without any comment attributed to “Yaḥyā”, whereas all preceding and following sections are intensely annotated by “Yaḥyā”). _____________ 211

Peters, Aristoteles Arabus, p. 31. The lemmata are often truncated and quote no more than the first part or the beginning and the end of the passage to be explained subsequently, while other sections of the Aristotelian text were not even fragmentarily quoted. On the (in)completeness of the Aristotelian lemmata see also James O. Urmson’s “Introduction” in: Simplicius, On Aristotle. On the Soul 1.1—2.4 (Ancient Commentators on Aristotle). London, New York 2014, p. 3; Mirjam E. Kotwick, Alexander of Aphrodisias and the Text of Aristotle’s Metaphysics (California Classical Studies, vol. 4). Berkeley 2016, p. 38—50. 212

XCIV

Introduction

IV. Contaminations of Isḥāq ibn Ḥunayn’s translation in ms. Leiden Or. 583 As we have seen in the preceding chapters, ms. Leiden Or. 583 is the result of roughly one century of learned activities revolving around Aristotle’s Physics in the translation by Isḥāq ibn Ḥunayn. A remarkable number of Aristotelian scholars in Baghdad were involved in these activities, and we can be sure that neither their attitude towards the written word in general nor their particular interest in this text were dominated by the ambition to foster an immaculate transmission of Isḥāq ibn Ḥunayn’s translation. Rather, their primary aim consisted in understanding, adaptating and incorporating Aristotle’s doctrines into the contemporary Christian-Islamic settings of 10th century ʿAbbāsid philosophy and cosmology. Hence, they certainly did not hesitate to switch between different translations of an authoritative work, replace technical terms or phrases, or even interpolate materials taken from the exegetical literature, whenever they deemed such methods appropriate and useful for this programme. We therefore have to ask if and to what extent these philosophical activities may have given rise to “philological” interventions in the transmission of Isḥāq ibn Ḥunayn’s translation, in other words, to what extent contaminations of Isḥāq’s text on the part of the Baghdad Aristotelians (and their scribes) are to be expected. Such contaminations or modifications may appear in form of interpolations or by changes of Isḥāq’s text on the basis of glosses reporting variant readings or alternative technical terms found in another translation. In what follows I shall discuss six paradigmatic cases in detail. That the question of contamination is by no means far-fetched has already become clear from the preceding examples of marginal collation notes referring to other versions of the Aristotelian text. One of these notes (v. p. LXXXVIII f., text [iii.3]) even includes the unambiguous assessment that a particular word in al-Dimašqī’s version is more adequate, lit. “superior to” (aǧwadu min) the relevant word in Isḥāq’s version, which indeed comes close to advising or at least inviting any copyist to substitute one for the other. In the following case, we see that what was originally a composite marginal note drawing on another translation and on a commentary has been slotted into a passage containing Isḥāq ibn Ḥunayn’s translation by the scribe of the Leiden manuscript or its ancestor. [1] The section in question is Physics IV 4, 212a5-7, which runs in all manuscripts of the direct tradition as follows: ἀνάγκη τὸν τόπον εἶναι τὸ λοιπὸν τῶν τεττάρων, τὸ πέρας τοῦ περιέχοντος σώματος. λέγω δὲ τὸ περιεχόμενον σῶμα τὸ κινητὸν κατὰ φοράν.

IV. Contaminations of Isḥāq ibn Ḥunayn’s translation

XCV

Isḥāq’s translation renders the text faithfully: fa-wāǧibun an yakūna l-makānu ... al-bāqiya mina l-arbaʿati wa-huwa nihāyatu l-ǧismi l-muḥīṭi aʿnī l-ǧisma l-muḥāṭa bihi llaḏī šaʾnuhū an yataḥarraka ḥarakata nuqlatin However, this is not how it appears in the Leiden manuscript, where it is interrupted by a contaminated note which seems to originate from three different sources, separated by a circle and three triangular dots. The text runs there (fol. 66b17ff.) as follows (Isḥāq’s text in italics, the interspersed note in Roman characters): fa-wāǧibun an yakūna l-makānu ... al-bāqiya mina l-arbaʿati wa-huwa nihāyatu l-ǧismi l-muḥīṭi ∴ wuǧida fī baʿḍi l-nusaḫi ziyādatun fī naqli Qusṭā “wa-huwa nihāyatu l-ǧismi l-muḥtawī yumāssu ʿalayhā mā yaḥtawī ʿalayhi aʿnī l-ǧisma llaḏī yaḥtawī ʿalayhi l-mutaḥarriku ḥarakata ntiqālin” ○ taraka “llatī ʿalayhā yalqā l-ǧisma l-muḥāṭa bihī” ∴ ay anna kulla ǧismin qad yataḥarraku ḥarakata nuqlatin ay ʿalā l-istiqāmati wa-qad yaskunu fa-huwa fī makānin aʿnī l-ǧisma l-muḥāṭa bihi llaḏī šaʾnuhū an yataḥarraka ḥarakata nuqlatin The first part of the note (from the first triangle of dots to the circle) mentions an addition in Qusṭā ibn Lūqā’s translation which is based on a Greek text similar to what we read in the commentaries by Themistius and Simplicius, who add after τοῦ περιέχοντος σώματος the phrase καθ’ ὃ συνάπτει τῷ περιεχομένῳ σώματι (Them. 276.13f.) or καθ’ ὃ συνάπτει τῷ περιεχομένῳ (Simpl. 580.2f.)213. The second part of the note (from the circle to the second triangle of dots) states that Isḥāq omitted (taraka) this phrase. However, it differs from Qusṭā’s terminology in using laqiya for συνάπτω and muḥāṭun bihī for περιεχόμενος and must have been supplied by somebody who had access to both translations. The third part is of exegetical nature and draws presumably on Philoponus, In Arist. Phys. p. 556.29—557.2. In other words, we are faced with (!)four intermingled text levels, the authoritative translation by Isḥāq ibn Ḥunayn, a note pointing to an (alleged) omission in the latter and quoting the more comprehensive version found in manuscripts of Qusṭā ibn Lūqā’s translation, another note drawing attention to this omission on the basis of yet another translation, and a doctrinal comment on the passage in question. _____________ 213

On this addition cf. Hermann Diels, “Zur Textgeschichte der Aristotelischen Physik”, Philosophische und historische Abhandlungen der Königlichen Akademie der Wissenschaften zu Berlin 1882, Abhandlung 1, Berlin 1883, p. 3—42, esp. p. 4—5; Ross, Physics, p. 575, and Pieter Sjoerd Hasper, below, p. CLXVI–CLXVII.

XCVI

Introduction

The acuteness of the question of contaminations is additionally emphasised by the fact that the Leiden manuscript contains numerous marginal and interlinear notes, the contents and structure of which do not clearly reveal whether we are faced with variant readings of Isḥāq’s version or extrinsic proposals for alternative interpretations (cf. above, chapter II.3). We may even take into consideration that some of them trace back to alternative translations noted down by Isḥāq himself. That some or most of them date from an early period of transmission has already become clear from the fact that the same notes occur not only in the margin of ms. Leiden Or. 583, but also in the margin of ms. Esc. 896 which belongs to another branch of the tradition214. While contaminations of the above-mentioned type are quite readily identifiable, others are in need of elaborate excavation. Here are some examples: [2] Physics IV 10, 218b5-6 ἡ δὲ τοῦ ὅλου σφαῖρα ἔδοξε μὲν τοῖς εἰποῦσιν εἶναι ὁ χρόνος, ὅτι ... “Those who said that time is the sphere of the whole thought so, no doubt, on the ground that ...” (transl. Hardie-Gaye) Ms. Escorial 896, fol. 50b5-6: wa-qad qāla qawmun inna l-zamāna kuratu l-kulli wa-innamā ẓannū ḏālika min qibali anna ... “Some said that time is the sphere of the whole. As a matter of fact, they thought so on the ground that ...” Ms. Leiden Or. 583, fol. 94a15 in textu (ed. Badawī, p. 411.12-13): wa-qad qāla qawmun inna l-zamāna huwa ḥarakatu l-kulli wa-innamā ẓannū ḏālika min qibali anna ... “Some said that time is the motion of the whole. As a matter of fact, they thought so on the ground that ...” Ms. Leiden Or. 583, fol. 94a in marg. with corresponding marks over ḥaraka (“motion”) and at the beginning of the marginal note: hākaḏā fī l-dastūri wa-innamā huwa anna qawman zaʿamū annahu l-kullu nafsuhū “Sic in the ‘master copy’. It [means] that some [philosophers] maintained that it [masc., scil. time] is the whole itself” At a first glance, the marginal note seems to indicate that huwa ḥarakatu lkulli (“is the motion of the whole”) is the reading of the “master copy”. However, the following explanation reveals that the text of the “master copy” implied the identity of time and the whole, which must be based on a text _____________ 214

Cf. above, chapter II.4.

IV. Contaminations of Isḥāq ibn Ḥunayn’s translation

XCVII

which read either kuratu l-kulli (“the sphere of the whole”) or simply al-kull (“the whole”). Since ms. Esc. 896 preserves the correct reading kuratu l-kulli (= ἡ τοῦ ὅλου σφαῖρα), it is clear that this is the original wording of Isḥāq’s translation. This assumption is supported by the subsequent commentary in ms. Leiden Or. 583, which is attributed to “Yaḥyā wa-Abū ʿAlī” and contains twice the phrase al-zamānu huwa l-kuratu nafsuhā (“time is the sphere itself”)215, thus leaving no doubt that Abū ʿAlī ibn al-Samḥ had access to a version equating time and the sphere. Hence, the reason for the implementation of the marginal note must be sought in an antecedent copy of Isḥāq’s translation which contained at the present place two competing readings, first kuratu l-kulli or huwa kuratu l-kulli (“is the sphere of the whole”) and secondly huwa ḥarakatu l-kulli (“is the motion of the whole”). The note was supposed to point to kuratu l-kulli as the correct reading supported by the “master copy” and to add for the sceptical reader an explanation of its meaning. Nevertheless, the scribe who used this antecedent copy as his exemplar opted for the wrong alternative, yet he was thoughtful (or insecure) enough to transcribe also the marginal note he found there. [3] Physics VII 5, 250a9-11 καὶ εἰ τὸ Ε τὸ Ζ κινεῖ ἐν τῷ Δ τὴν Γ, οὐκ (E2beLNRFHIJQ γρ.Simpl. [1106.2] Philop.[880.29] : om. E1K Simpl.) ἀνάγκη (ἀνάγκη E2beLNRI γρ.Simpl.[1106.2] : ἀναγκαῖον E1KFHJQ, fort. Simpl.[1105.12-17]) ... τὸ ἐφ' οὗ Ε “And if E moves Z in the time D [over] the distance G, it follows (E1K Simpl.) / it does not follow (codd. cett.) necessarily that that upon which is E ...” Ms. Leiden Or. 583, fol. 201b12-13 in textu (ed. Badawī, p. 792.5-7): wain kāna H yuḥarriku Z fī zamāni D masāfata Ǧ fa-laysa yaǧibu ḍarūratan an yakūna llaḏī ʿalayhi H ... “And if H moves Z in the time D [over] the distance Ǧ, it does not necessarily follow that that upon which is H ...” This is a faithful translation of the version containing the negation οὐκ (rendering the letter Epsilon by Hāʾ is common practice). Over the negation laysa (“does not”) we find a superscript mark, and in the margin we read the following note, introduced by a corresponding mark: fī nusḫat Ibn ʿAdī: “fa-yaǧibu ḍarūratan” wa-fīhā ayḍan mā hāḏihī ḥikāyatuhū: waǧadnā fī nusḫati Yaḥyā al-Naḥwī “fa-laysa yaǧibu ḍarūratan” _____________ 215

Badawī, al-Ṭabī ʿa, p. 416.5, 416.16.

XCVIII

Introduction

“In [Yaḥyā] ibn ʿAdī’s copy [it reads] ‘it follows necessarily’; also in this [manuscript is] a statement to the following effect: In the copy of John Philoponus[’ commentary] we found [the reading] ‘it does not necessarily follow’.” From the second part of the note we learn that Yaḥyā ibn ʿAdī checked a copy of John Philoponus’ commentary, probably not a Greek manuscript, but rather an Arabic (or Syriac?) translation. There he read “it does not necessarily follow”, which he noted in his copy of the Aristotelian text, which for its part read “it follows necessarily”. Since we know that the expression “Ibn ʿAdī’s manuscript” (nusḫat Ibn ʿAdī ) means Yaḥyā ibn ʿAdī’s copy of Isḥāq ibn Ḥunayn’s translation, we are quite safe in assuming that Isḥāq had been translating from a manuscript omitting the negation οὐκ (as mss. E K and the manuscript[s] quoted by Simplicius do). Otherwise Yaḥyā ibn ʿAdī would not have been prompted to add this remark. While Yaḥyā ibn ʿAdī confined himself to noting this variant reading in his copy, another scholar or scribe of the Baghdad school ventured to modify Isḥāq’s translation by adopting the Philoponan reading either on the basis of Yaḥyā ibn ʿAdī’s note or by direct recourse to the Arabic version of Philoponus’ commentary. [4] Physics VIII 1, 250b18-20 ἀλλ’ ὅσοι μὲν ἀπείρους τε κόσμους εἶναί φασιν, καὶ τοὺς μὲν γίγνεσθαι τοὺς δὲ φθείρεσθαι τῶν κόσμων, ἀεί φασιν εἶναι κίνησιν “But those who say that there is an infinite number of worlds, some of which are in process of becoming while others are in process of perishing, say that there is always motion” (transl. Hardie-Gaye, slightly modified) Ms. Leiden Or. 583, fol. 205a9-10 (cf. below, p. 3.3-5): lākinna llaḏīna qālū bi-ʿawālima kaṯīratin bi-lā nihāyata wa-anna baʿḍahā yatakawwanu wa-baʿḍahā yafsudu fa-hum yaqūlūna bi-ḥarakatin taḥduṯu wa-tafsudu wa-ǧamīʿuhum aṯbatū l-ḥarakata dāʾiman. “But those who say that there is an infinite plurality of worlds and that some of them are in process of becoming while others are in process of perishing, say [implicitly] that motion [is something that] becomes and perishes, while they all assert that there is always motion.” The phrase printed in spaced letters is not found in the Greek text of Aristotle. In all likehood we are faced with an interpolated gloss, the original purpose of which consisted in supplying the additional argument that the assumption that an infinite number of worlds come to be and perish necessarily implies that motion comes to be and perishes, which contradicts

IV. Contaminations of Isḥāq ibn Ḥunayn’s translation

XCIX

a claim made by those who hold the above opinion, namely that motion is eternal. It reveals itself as being a foreign intruder in three ways: (a) it does not occur in the Latin translations of Isḥāq’s version; (b) it uses the verb ḥadaṯa for „coming to be“, while Isḥāq both in the preceding and in the following lines uses the fifth form of k-w-n for γίγνομαι and γένεσις; (c) in the manuscript it is tagged with superscript marks and the corresponding marginal note: “the words marked are omitted in Ibn ʿAdī’s copy”. [5] Physics VIII 4, 255a33-34 ἔστι δὲ δυνάμει ἄλλως ὁ μανθάνων ἐπιστήμων καὶ ὁ ἔχων ἤδη καὶ μὴ ἐνεργῶν (ἐνεργῶν E Simpl.[1213.32] (?)Them.[219.12] : θεωρῶν KbeLNR FHIJQ) “The learner possesses [potentially] knowledge in another way of potentiality than he who has already acquired [knowledge] but is not [actually] exercising [it] / thinking [about it]” Ms. Leiden Or. 583, fol. 213a6-7 (cf. below, p. 34.5-7): fa-l-mutaʿallimu huwa bi-l-quwwati ʿālimun bi-waǧhin mā ġayri l-waǧhi llaḏī bihi l-ʿilmu wa-huwa lā yuʿmalu (lā yuʿmalu ante corr. : lā yaʿlamu post corr. + bihī s.l.) ʿālimun bi-l-quwwati “The learner possesses potentially knowledge in a certain way which is different from the way in which he who has the knowledge [already] acquired, while it is not [actually] practiced (ante corr.) / while he is not thinking about it (post corr.), possesses potentially knowledge” The two conflated readings in the Leiden manuscript probably reflect the variant readings ἐνεργῶν and θεωρῶν in the Greek manuscripts216. Both readings were known to the Baghdad Aristotelians, as is shown not only by the fact that the one is written over the other, but also by their separate later transmission. The Arabic exemplar used by Gerard of Cremona obviously read lā yuʿmalu (= “non operatur”), whereas Michael Scot translated lā yaʿlamu bihī (= “non scit”). Isḥāq’s translation probably read lā yuʿmalu as it usually follows the Greek manuscript E and this seems to be written prior to lā yaʿlamu in the Leiden manuscript. What we see in the Leiden manuscript is the result of a contamination. _____________ 216

For ʿalima = θεωρεῖν see the immediately following passage, Physics VIII 4, 255b2-4 (twice), and Aristotle, Prior Analytics 43a23 = ed. Badawī, p. 214.6, Posterior Analytics 81b2 = ed. Badawī, p. 385.10, Topica 149b27 = ed. Badawī, p. 687.6, Sophistici elenchi 169a32 = ed. Badawī, p. 842.12, Historia animalium 566a13 = ed. Badawī, p. 266.15-16 (together with arāda l-maʿrifata), etc.

C

Introduction

[6] Physics VIII 8, 263b24-26 ἢ ἅμα λευκὸν καὶ οὐ λευκὸν καὶ ὅλως ὂν καὶ μὴ ὂν ἀνάγκη εἶναι “or else things must at the same time be white and not white and in general be and not be” (transl. Hardie-Gaye) Ms. Leiden Or. 583, fol. 225b ult.—226a1 (cf. below, p. 92.4-6): aw waǧaba ḍarūratan an yakūna [scil. al-šayʾu] maʿan abyaḍa wa-lā abyaḍa wa-bi-lǧumlati mawǧūdan k a ḏ ā wa-ġayra mawǧūdin The word kaḏā (“sic”) is evidently out of place here. In an antecedent copy, it must have served as a marginal note pointing to a correction in the preceding line (i.e. “sic legendum”)217, which at some point of the transmission was interpolated into the main text at the wrong place. The interpolation even found its way into the Latin translations by Gerard of Cremona (“tale”) and Michael Scot (“hoc”, i.e. hoc modo). Hence, the Leiden manuscript is not untainted by contaminations caused by various kinds of notes and glosses added to Isḥāq’s translation in the school of Baghdad. However, taking into consideration (1) the overall accuracy and consistency of Isḥāq ibn Ḥunayn’s translation as transmitted in ms. Leiden Or. 583, (2) the high diligence and precision in the transcription of the marginal and interlinear notes exhibited both by Abū l-Ḥusayn al-Baṣrī and the scribe of the Leiden manuscript, and (3) the fact that the two Arabo-Latin translations only very occasionally reflect contaminations of the Aristotelian text caused by notes and glosses in the Arabic transmission, we are quite safe in assuming that contaminations of this type happened only rarely.

V. Isḥāq ibn Ḥunayn’s Greek exemplar(s) One of the three primary aims of the present edition consists in bringing some glimmers of light into the darkness of competing variant readings among the Greek manuscripts of the Physics and their intricate stemmatic relations. Two things give reason to hope that there may be some prospect of success for this undertaking at all, first the fact that the Greek exemplar from which Isḥāq ibn Ḥunayn prepared his Arabic translation of Aristotle’s Physics antedates all extant Greek manuscripts, and secondly the circumstance that —judging from a comparison with the sections of the Greek text transmitted in the extant Greek _____________ 217

The relevant scribal error and its correction by second hand are still present in the Leiden ms., cf. the critical apparatus ad loc.

V. Isḥāq ibn Ḥunayn’s Greek exemplar(s)

CI

manuscripts without substantial variation and in accordance with the Arabic text— Isḥāq’s translation is in general very reliable and meets the high standards of Aristotle’s difficult style and terminology with an accuracy, sophistication and linguistic empathy scarcely to be found among modern translators of the text218. Before turning to the question what general information about Isḥāq’s Greek exemplar we can derive from his translation, we must address another important issue: Did Isḥāq ibn Ḥunayn prepare his Arabic translation on the basis of a single manuscript, or did he use a multiplicity of codices219? It goes without saying that the answer to this question is of vital importance for any kind of conclusion supposed to be drawn with regard to the stemmatic position and the relationship between the reconstructed Greek exemplar(s) of the Arabic translation and the extant Greek manuscripts. All the more surprising, this general problem has not been addressed in any recent Graeco-Arabic edition or study, no matter whether concerned with a translation by Isḥāq ibn Ḥunayn or any other translator. That it is anything but far-fetched or pseudo-scientific quibbling is shown by the simple fact that Isḥāq’s father Ḥunayn mentions in the autobiographical notes of his epistle on the translations of Galen into Syriac and Arabic more than once having used, whenever available to him, more than one Greek manuscript for his translations. As a matter of fact, the Leiden manuscript contains several serious hints seemingly indicating that Isḥāq ibn Ḥunayn had a multiplicity of Greek manuscripts of Aristotle’s Physics in front of him. As mentioned above, four marginal notes refer explicitly to Greek manuscripts (nusaḫ yūnāniyya, in the plural); even more conspicuously, two of them qualify the Greek manuscripts as those “from which we translated” (allatī minhā naqalnā /tarǧamnā)220. In order to clarify whether these notes trace back to Isḥāq ibn Ḥunayn or to somebody else, we must investigate them in detail. _____________ 218

Especially noteworthy is the fact that Isḥāq ibn Ḥunayn —unlike many other translators from Greek into Arabic— refrains from any syntactic banalization and tries to retain or mirror the complex structure of long compound sentences. 219 In what follows I narrow this question down to the Greek tradition of the Aristotelian text. That Isḥāq may have used additional Syriac or Arabic sources cannot be excluded with certainty. However, the text contains no specific hints pointing in this direction. 220 Again, a similar phenomenon is known from translations by Isḥāq’s father Ḥunayn. Two manuscripts of Ḥunayn ibn Isḥāq’s translation of Galen’s De simplicium medicamentorum temperamentis ac facultatibus report a text-critical note introduced as follows: qāla Ḥunayn: waǧadnā fī kaṯīrin mina l-nusaḫi l-yūnāniyyati ... (“Ḥunayn says: We found in many Greek manuscripts ...”); cf. Manfred Ullmann, Wörterbuch zu den griechisch-arabischen Übersetzungen des 9. Jahrhunderts, Wiesbaden 2002, p. 32—33; v. also above, note 172.

CII

Introduction

[1] Ms. Leiden Or. 583, fol. 42a, ed. Badawī, p. 2033, on Physics III 4, 203a9-10 [a] Greek Text and Arabic translation τὸ [...] ἄπειρον καὶ ἐν τοῖς αἰσθητοῖς καὶ ἐν ἐκείναις (ἐκείναις [scil. ταῖς ἰδέαις] beLNRFHIJQw Simpl.[452.20] : ἐκείνοις E) εἶναι “The infinite [...] exists both in the objects of sense and in those.” inna lā nihāyata mawǧudun (sic) fī l-maḥsūsāti wa-fī ḏālika (ḏālika ms. Leiden : tilka scr. Badawī; transl. Isḥāq, ed. Badawī, p. 203.11-12) “The infinite exists in the objects of sense and in that (masc. sg.).” [b] Marginal note The word ḏālika (“that”) is marked with a superscript sign referring to a marginal note with the same sign. The note, which bears additionally the mark ‫ ح‬pointing to Yaḥyā ibn ʿAdī’s copy as source of transcription, runs as follows: fa-qad yūǧibu nasqu l-kalāmi wa-l-maʿnā anna qawlahū “ḏālika” (ḏālika vel tilka ms. Leiden : tilka Badawī) raddahū (sic ms., an raǧaʿa leg.?) ʿalā lṣūrati illā anna ḏālika lā yarǧaʿu fī l-luġati l-yūnāniyyati ʿalayhā fa-immā (fa-immā ms. : mā Badawī) an yakūna hāḏā ḫaṭaʾan fī l-nusaḫi lyūnāniyyati wa-immā an yakūna ʿalā ḥasabi l-yūnāniyyi qawluhū “ḏālika” (ḏālika vel tilka ms. : tilka Badawī) rāǧiʿan ʿalā l-aʿdādi wa-in kāna innamā qāla l-ʿadadu wa-lam yaquli l-aʿdādu “The word order and the meaning require that he [scil. Aristotle] refers by saying ‘[in] that’ back to ‘the Form’ (al-ṣūra). However, in the Greek language it does not refer to it. Hence, either this is a mistake in the Greek manuscripts or his saying ‘[in] that’ refers in accordance with the Greek [syntax] to the numbers, even though what he had said [203a7] is ‘the number’ [in the singular], he did not say ‘the numbers’ [in the plural].” [c] Analysis The note concerns the competing readings ἐκείναις and ἐκείνοις (cf. [a]). Isḥāq’s translation corresponds with ἐν ἐκείνοις as far as the grammatical gender is concerned. Obviously he did not read ἐκείναις whose antecedent, τὰς ἰδέας (203a8), would have been easily recognisable. τὰς ἰδέας is correctly translated in the main text by al-ṣuwar (sic leg. cum ms., alṣūratu ed. Badawī, p. 203.10) and requires a feminine pronoun. On the other hand, there is no masc. pl. antecedent in the preceding lines to which ἐν ἐκείνοις could reasonably be related. The most plausible object of reference is τὸν ἀριθμόν (203a7), which is presumably what Isḥāq had in mind when deciding for the masc. sg. pronoun ḏālika. Taking into

V. Isḥāq ibn Ḥunayn’s Greek exemplar(s)

CIII

consideration that Simplicius read ἐκείναις, whereas Themistius and Philoponus did not quote the passage in question, this evidence points to Isḥāq ibn Ḥunayn as the author of the note. As the note is additionally marked with the sign ‫ح‬, we may assume that Yaḥyā ibn ʿAdī copied it from his exemplar of Isḥāq’s translation. [2] Ms. Leiden Or. 583, fol. 74b, omitted by Badawī, on Physics IV 6, 213a13 [a] Greek Text and Arabic translation περὶ κενοῦ, εἰ ἔστιν ἢ μή, καὶ πῶς ἔστι (καὶ πῶς ἔστι om. E), καὶ (καὶ Them., Simpl. : ἢ all mss., Philop.[610.23]) τί ἐστιν “About the void, whether it exists or not, and how it exists, and what it is.” (transl. Hardie-Gaye, slightly modified) fī amri l-ḫalāʾi ... hal huwa aw laysa huwa wa-kayfa ayḍan huwa wa-mā huwa (transl. Isḥāq, ed. Badawī, p. 338.8-9) “About the void, whether it exists or not, and also how it exists, and what it is.” [b] Marginal note The last huwa (“it is”) is marked with a superscript sign. In the margin we read the following note (without corresponding mark): waǧadnā fī l-nusaḫi l-yūnāniyyati llatī minhā naqalnā “aw mā huwa” waaḥsabu l-ḥarfa llaḏī aḫraǧnāhu “aw” kāna fī l-yūnāniyyi “wa” fa-ḥarrafahū (addidi) baʿḍa l-taḥrīfi ʿalā mā ẓannahū kāna maʿnāhu bi-lʿarabiyyati “in kāna šayʾan mā” wa-hāḏā ašbahu bi-nasqi l-kalāmi bi-an yaǧʿala qisma “kayfa huwa” “mā huwa” “In the Greek manuscripts from which we translated [it reads] ‘or what it is’. I assume that the word we rendered [here] by ‘or’ [i.e., ἢ] [originally] was ‘and’ in Greek [i.e., καὶ] and [that] altered it somehow based on the opinion that its [original] meaning was (in Arabic words): ‘if it is anything [at all]’. Τhis fits better with the word order in so far as the latter makes [the question] ‘what is it’ a [subordinate] part of [the question] ‘how is it’.” [c] Analysis This note addresses three different readings/meanings for the third question concerning the void: (α) “or what it is”, (β) “and what it is”, (γ) “if it is anything [at all]”.

CIV

Introduction

Reading (α) corresponds to ἢ τί ἐστιν as attested in the extant manuscripts of the direct tradition. This is the reading the author of the note found in his Greek exemplars, as explicitly stated. Reading (β) corresponds to καὶ τί ἐστιν, the wording quoted by Themistius and Simplicius. Obviously, the author of the note had no access to a manuscript confirming this reading, but conjectured it by himself (“I assume ... the word [originally] was ...”). Reading (γ), in kāna šayʾan mā, corresponds to a Greek text reading εἴ τι ἔστιν (less probable ᾗ τι ἔστιν221), for which there is no manuscript evidence. The author of the note considers “καὶ” as the original reading changed by some scribe or scholar into “ἢ”. While this part of the marginal note is quite clear, the second half is rather enigmatic. Apparently, the author of the note takes reading (γ), “if [the void] is anything at all”, as the scribe’s rationale behind the alteration of reading (β) into reading (α). However, he does not explain why the scribe did not straightaway replace reading (β) with reading (γ). Rather, he supplements another remark, which seems to refer again to the two readings (α) and (β). This remark draws on the idea that the series of questions raised about the void should move from the more general to the more specific, which is not the case, if the question “what it is” comes at the end of the sentence, coordinated by “καὶ”222. Since the author of the note explicitly considers “καὶ” as the original reading, the conundrum may be solved by assuming that he conceived the combination of “καὶ” and reading (γ), i.e. καὶ εἴ τι ἔστιν, as the correct one. Or he took ἢ τί ἐστιν as reflecting more likely a corruption of εἴ τι ἔστιν than of καὶ τί ἐστιν. Either explanation is plausible. The confusion of ει and η is well attested and certainly more frequent than that of και and η (while the interpretation of τι as interrogative or indefinite pronoun is dependent on the preceding word). As a matter of fact, the relation of the three questions —and especially of the last two— about the void raised by Aristotle was a matter of bewilderment and dispute among the commentators of late antiquity223. _____________ 221

For ᾗ (“in so far as”) = in cf. GALex, vol. I, in 8.3.4. However, Isḥāq ibn Ḥunayn renders ᾗ consistently through min ǧihati anna/mā or bi-mā huwa. 222 Interestingly, similar doubts concerning the order of the questions are raised in Simplicius’ commentary on the beginning of Aristotle’s investigation of place, Physics IV 1. There, Aristotle asks in the same order with regard to τόπος: “εἰ ἔστιν ἢ μή, καὶ πῶς ἔστι, καὶ τί ἐστιν” (208a28-29), which prompts Simplicius to ask: “ἀλλὰ διὰ τί τὸ πῶς ἐστι προέταξεν;” (In Phys., p. 520.34). 223 Cf. Rashed, Alexandre d’Aphrodise. Commentaire perdu, p. 225—226, frag. 85; Philoponus: On Aristotle Physics 5—8, with Simplicius: On Aristotle On the Void. Translated by

V. Isḥāq ibn Ḥunayn’s Greek exemplar(s)

CV

How could Aristotle, who generally denied the existence of the void, reasonably undertake an investigation of its How and What? Simplicius reports that Aspasius maintained that the latter two questions have the same meaning, or rather are both meaningless, because the void is nonexistent224, while Philoponus has on the question “what it is” only a brief remark which actually summarises what he said before in extenso about the question “how it is”225. Hence, it is not too far fetched to assume a gloss or scholium in a late ancient manuscript pointing out that the basic and essential question is not What-it-is, but rather whether the void exists at all. Be that as it may, we are still faced with the problem that the author of the note explicitly mentions that his Greek exemplars provided reading (α), whereas the main text of the Leiden manuscript displays reading (β). Assuming that the author of the note was none other than Isḥāq ibn Ḥunayn himself, we can explain this phenomenon in two different ways: (i) As we have seen above, Isḥāq rather tends to restrict himself to marginal proposals for emendation while adhering in the translation proper to the Greek wording in front of him, even if he is in doubt about its meaning or its correct transmission (cf. p. LXXXII–LXXXIV, also the immediately following case [3]). Thus, if he remained true to this principle in the present case and translated ἢ τί ἐστιν as attested in his manuscripts (to which he then supplied the above-quoted note), the original version of his translation must have read aw mā huwa, which was then changed by some scribe according to the emendation proposed in the note. This alteration must have happened at an early stage in the transmission, because the intervention is confirmed by the Arabo-Latin translations226, which belong to another branch of the tradition. Or (ii) Isḥāq ibn Ḥunayn deviated in the present case from his usual reluctance and implemented his emendation ad hoc. The style and diction of the note clearly speak in favour of option (i), especially if we consider the fact that Isḥāq ibn Ḥunayn may have conceived καὶ εἴ τι ἔστιν (wa-in kāna šayʾan mā) as the correct reading. _____________ Paul Lettinck & J.O. Urmson (Ancient Commentators on Aristotle), London 2013, p. 223—224, note 4. 224 Cf. Simpl., In Phys. 646.3-8 225 Philop., In Phys., p. 610.29-31. 226 Cf. “et quid est” in Gerard of Cremona’s translation (Cv, fol. 32a4). The phrase is omitted in the textus of Michael Scot’s translation, but quoted (“et quid [sit]”) by Ibn Rušd in the commentary, Aristotelis De Physico auditu libri octo cum Averrois Cordubensis [...] commentariis. Venice 1562 apud Iunctas, p. 147E, 147G.

CVI

Introduction

On the other hand, we should also consider the possibility that the note derives not from Isḥāq’s hand, but from the margin of a translation of Philoponus’ commentary on this section. As mentioned above, Philoponus grants no more than the following two lines to the last of the three questions (In Arist. Phys., p. 610.29-31): τὸ δὲ τί ἐστι, πότερον διάστημά τι καθ̕ αὑτὸ μηδὲν ἔχον τῶν φυσικῶν σωμάτων, ἢ ὕλη ἐστὶ τῶν σωμάτων, ἢ ἕτερόν τι. Significantly, the literal Arabic translation of these lines, which is quoted in the commentary section preserved in ms. Leiden Or. 583, runs as follows: wa-ammā qawluhū “wa-mā huwa” fa-arāda bihī hal huwa buʿdun mā fī ḏātihī laysa fīhi šayʾun mina l-aǧsāmi l-ṭabīʿiyyati aw huwa ʿunṣurun li-laǧsāmi aw huwa šayʾun āḫaru 227 “As to [Aristotle’s] words ʽand what it is’, he means whether it is a certain extension in itself without encompassing any natural body, or [whether] it is matter for bodies or something else.” This translation reflects the reading τὸ δὲ καὶ τί ἐστι instead of τὸ δὲ τί ἐστι. Judging from the critical apparatus of Vitelli’s edition, there is no manuscript evidence for this reading. However, the translator of Philoponus’ text may have read τὸ δ̕ ἢ τί ἐστι instead of τὸ δὲ τί ἐστι. Hence, the note under discussion may derive from his hand. As we have seen, the Baghdad scholars collated Isḥāq’s translation with the Arabic translation of Philoponus’ commentary and noted variant readings in its margins. We, thus, cannot rule out in advance that the present note ultimately traces back not to Isḥāq ibn Ḥunayn, but to the translator of Philoponus’ commentary. In any case, this hypothesis leads to the same conclusion as the preceding one, namely that in all likelihood the original version of Isḥāq’s translation read aw mā huwa. Otherwise, there would have been no discrepancy between the two Arabic texts, hence no reason for any collator to transcribe the relevant note. [3] Ms. Leiden Or. 583, fol. 112a, ed. Badawī, p. 4805, on Physics IV 14, 224a2-4 [a] Greek Text and Arabic translation λέγεται δὲ ὀρθῶς καὶ ὅτι ἀριθμὸς μὲν ὁ αὐτὸς ὁ τῶν προβάτων καὶ τῶν κυνῶν, εἰ ἴσος ἑκάτερος, δεκὰς δὲ οὐχ ἡ αὐτὴ οὐδὲ δέκα τὰ αὐτά (οὐδὲ δέκα τὰ αὐτά om. F) _____________ 227

Badawī, al-Ṭabī ʿa, p. 345.13-15.

V. Isḥāq ibn Ḥunayn’s Greek exemplar(s)

CVII

“It is said rightly, too, that the number of the sheep and of the dogs is the same number if the two numbers are equal, but not the same decade or the same ten.” (transl. Hardie-Gaye) niʿimmā yuqālu ayḍan inna ʿadada l-ġanami wa-ʿadada l-kilābi wāḥidun biʿaynihī iḏ kāna kullu wāḥidin mina l-ʿadadayni sawāʾan wa-yuqālu inna l-ʿašarata mina l-ʿadadi laysat wāḥidatan bi-ʿaynihā wa-lā l-ʿašaratu l-ašyāʾi wāḥidun bi-ʿaynihā (transl. Isḥāq, ed. Badawī, p. 480.10— 481.1) “It is said rightly, too, that the number of the sheep and of the dogs is one and the same number if the two numbers are equal; and it is said that the number ten is not one and the same, nor are the ten things one and the same.” [b] Marginal note The words wa-lā l-ʿašaratu (“nor are the ten [things]”) are marked with a superscript sign referring to the following marginal note marked with the same sign: hākaḏā waǧadnā fī l-yūnāniyyi fī l-nusaḫi llatī tarǧamnā minhā wa-l-maʿnā yaqtaḍī fī-mā aḥsabu an yuzāda hāḏā l-qawlu “iḏā kānat” ḥattā yakūna (without points ms. : takūna Badawī) hākaḏā: “iḏā kānat wa-lā l-ʿašaratu l-ašyāʾi wāḥidatun bi-ʿaynihā” wa-aẓunnu hāḏā saqaṭa “This is what we found in Greek in the manuscripts from which we translated. In my estimation, the [correct] meaning requires the addition of ‘if it exists’ to these words, so that it runs as follows: ‘[...,] if it exists, nor are the ten things one and the same’ – I think this was omitted.” [c] Analysis Aristotle illustrates in the section quoted the relationship between time and the different modes of its predication by adducing the relationship between number and the objects numbered. The number of sheep and dogs is one and the same, if we count ten sheep and ten dogs. However, neither is it the case that the unit of ten sheep and the unit of ten dogs are one and the same, nor that the objects counted are one and the same. Apparently, the author of the note assumes that the mention of the unit of ten (δεκάς) was originally accompanied by the proviso “if it exists”, which dropped out in the transmission of the Greek text. The basis of this assumption is not clear to me. There is no hint to this effect in the direct or indirect traditions. Perhaps it was prompted by the author’s awareness of the perennial controversy about the separate existence of numbers. Once again, the author refers to the Greek manuscripts “from

CVIII

Introduction

which he translated”. In the present case this may relate to separate copies of the Aristotelian text or the commentaries by Philoponus and Simplicius, both of which quote and comment on this passage228. [4] Ms. Leiden Or. 583, fol. 132a, ed. Badawī, p. 5475, on Physics V 3, 227a31 [a] Greek Text and Arabic translation πᾶσα γὰρ γραμμὴ μεταξὺ στιγμῶν (στιγμῶν EbeLNRFHIJQ Simpl. [880.25] : δύο στιγμῶν H) : πασῆς γὰρ γραμμῆς μεταξὺ στιγμή γρ.Simpl. [880.25-26]) “[F]or all lines are intermediate between [two ms. H] points.” (transl. Hardie-Gaye) wa-ḏālika anna kulla ḫaṭṭin fa-fī-mā bayna nuqṭatayni (transl. Isḥāq, ed. Badawī, p. 547.13-14) “For all lines are intermediate between two points.” [b] Marginal note Next to this sentence, yet without any mark indicating the place of reference, we read in the margin the following note: hākaḏā waǧadnā fī l-nusaḫi l-yūnāniyyati wa-aḥsabu l-ḥaqqa an yuqāla inna kulla nuqṭatayni fa-fī-mā baynahumā ḫaṭṭun “This is what we found in the Greek manuscripts. In my estimation it would be [more] correct to say that intermediate between each pair of points there is a line.” [c] Analysis Isḥāq’s translation corresponds with the majority of the Greek manuscripts. Whether his use of the dual for the translation of στιγμῶν is based on his own interpretation or on the variant reading δύο στιγμῶν as found in ms. H is difficult to assess and not relevant for the evaluation of the marginal note. Simplicius points to a different transmission of the whole sentence, which says that “intermediate (or: in the midst) of each line there is a point”. However, this is not what the author of the note proposes to read. In all likelihood, the reason it was added lies in Aristotle’s lax diction. Of course, it is only the finite line which is intermediate between to points, but not the (potentially) infinite line, and the point Aristotle wishes to make is not a general statement about all lines, but rather the _____________ 228

Presumably this holds true also for Alexander of Aphrodisias, whose remarks are taken into consideration by Simplicius (In Arist. Phys., p. 770—771)

V. Isḥāq ibn Ḥunayn’s Greek exemplar(s)

CIX

fact that between two points there is (always) a line. As in the preceding two cases, the author of the note presents his remark as a result of his own reflection. Apparently, he was not acquainted with the commentaries by Philoponus and Simplicius, where he could have found support for the re-phrasing he proposes. In the Arabic translation of Philoponus’ commentary on this section229 we read: wa-ayḍan laysa yumkinu an yakūna bayna l-waḥdatayni šayʾun wa-qad yumkinu an yakūna bayna kulli nuqṭatayni ḫaṭṭun (ed. Badawī, alṬabīʿa, p. 549.6-7) “Furthermore, it is impossible that anything is intermediate between two units, whereas it is possible there is a line intermediate between each pair of points.” Similarly, in Simplicius’ commentary (p. 881.11-12): μεταξὺ τῶν στιγμῶν ἐνδέχεται εἶναι γραμμάς “Intermediate between two points there can be lines.” However, the proposed (and correct) interpretation is found only few pages later, where Aristotle says: στιγμῶν δ̕ αἰεὶ [τὸ] μεταξὺ γραμμή (Physics VI 1, 231b9), faithfully rendered by Isḥāq as follows: wa-fī-mā bayna kulli nuqṭatayni abadan ḫaṭṭun (ed. Badawī, p. 608.13). Let us try to sort the disparate data compiled in the above discussions of the four notes: (i) In all probability, all notes derive from the same hand. Three notes introduce the observation on the Greek text by the formula aḥsabu with accusative + noun phrase (“I assume”, “in my estimation”); the first two coincide furthermore in using the expression nasqu l-kalāmi in order to refer to the Greek word order. Note [2] taken by itself may originate from the translator of Philoponus’ commentary. However, its ascription to Isḥāq ibn Ḥunayn is equally plausible and probable. (ii) Notes [1]—[3] quote the text of Isḥāq ibn Ḥunayn’s translation. Hence, they must be contemporary or from a later date. (iii) In case [1], the reference to “the Greek manuscripts” cannot relate to Greek manuscripts of Themistius’ or Philoponus’ commentaries, because these texts do not quote the passage in question. Neither can it refer to Simplicius’ commentary, as the author of the note criticises a mistake which is not found there, _____________ 229

The Greek version of which is lost.

CX

Introduction

and proposes in turn an interpretation corresponding to the wording taken for granted by Simplicius230. No less unlikely is a reference to Alexander’s commentary, as one might expect Simplicius to have noted —as he usually does— a variant reading encountered in Alexander’s commentary. “Greek manuscripts” in note [1] must therefore refer to manuscripts of the Aristotelian text. (iv) In case [2], we can rule out that the author of the note refers to the commentaries by Themistius or Simplicius (as he reports a reading “in the Greek manuscripts” which is not attested there), whereas the reference may indeed be to Philoponus’ commentary. Note [3], on the other hand, refers to a wording which is confirmed by Philoponus’ and Simplicius’ commentaries. Assuming that the phrase “the Greek manuscripts from which we translated” refers in both cases to the same group of manuscripts, resp. to the same text translated, this narrows down the possible candidates for their identification either to Philoponus’ commentary or the Aristotelian text itself. (v) The author of note [4] presents his remark as a result of his own reflection, although the proposed reading is found in almost identical form in the commentaries by Philoponus and Simplicius. Obviously, he had no access to Philoponus’ commentary. (vi) Supposing that all four notes were authored by the same scholar contemporary with or later than Isḥāq ibn Ḥunayn, we are thus quite safe in assuming that both kinds of references, those to “Greek manuscripts” and the more explicit ones to “the Greek manuscripts from which we translated” relate to one and the same group of manuscripts of the Aristotelian text, to the exclusion of the commentaries by Alexander, Themistius, Philoponus and Simplicius. (vii) As far as we know, the only Graeco-Arabic translator to whom all the above criteria apply is Isḥāq ibn Ḥunayn, who —as shown above— was not principally averse to supplementing his translations with marginal glosses and notes. Needless to say, this result entails the consequence that some of the above mentioned notes referring more generally to “other manuscripts” may originate from Isḥāq’s hand, too. We are thus faced with the unfavourable situation that any reconstruction of the Greek text as translated by Isḥāq ibn Ḥunayn must take into account that it starts from an Arabic text which (a) might be the product of an eclectic use of a multiplicity of Greek manuscripts, and (b) is not entirely free from contaminations, to be more precise two different types of contamination: (i) seeming _____________ 230

Not to mention the fact that any circulation of Greek manuscripts of Simplicius’ commentary in 9th or 10th c. Baghdad is rather doubtful.

V. Isḥāq ibn Ḥunayn’s Greek exemplar(s)

CXI

contaminations of the wording of Isḥāq’s main Greek exemplar (henceforth: Ψ) caused by the translator’s eclectic use of other Greek manuscripts which do not belong to the family or branch of this exemplar, and (ii) contaminations of Ishāq’s translation which, although itself originally derived from an uncontaminated Greek text, was modified on the basis of other Arabic translations prepared from Greek manuscripts which do not belong to the family or branch of Ψ. As for type-ii contaminations, we are not utterly defenceless. Most of them can be identified through comparative linguistic and lexicographical studies of Isḥāq ibn Ḥunayn’s translations and/or the collation of the Arabic text of his translation with the indirect Arabic and Latin traditions. The determination of type-i contaminations is less likely to produce verifiable results. It depends largely on the question if or to what extent the contamination can be relegated to Isḥāq’s (non-extant) main Greek exemplar instead of attributing its origin to Isḥāq’s use of additional Greek manuscripts. The only means at our disposal to answer this questions is the Parisinus gr. 1853 (E), which belongs to the same branch as Ψ. If E turns out to be contaminated from other parts of the tradition, there is a certain probability that other representatives of the same branch such as Isḥāq’s main Greek exemplar were contaminated, too231. However, even in that case we cannot rule out that apparent cases of contamination in the Arabic translation do not reflect real contamination in Ψ, but originate from Isḥāq’s recourse to a multiplicity of witnesses. However, this drawback is mitigated to a certain extent by the following facts: (1) The extremely small number of notes pointing explicitly to the use of a multiplicity of manuscripts as compared to the comprehensiveness of the text suggests either that the Greek manuscripts at Isḥāq ibn Ḥunayn’s disposal were very closely related, or that Isḥāq only rarely had recourse to additional manuscripts. The latter assumption is supported by the fact that the whole direct and indirect tradition of Isḥāq’s translation of book VIII contains four omissions which can be easily explained by a saut du même au même in the Greek text, yet not by an omission in the Arabic text. Had Isḥāq collated his main exemplar of translation continuously with the other manuscripts at his disposal, he certainly would have spotted and filled that gap with their help. An example of such a saut du même au même is Physics VIII 1, 252a12-16: ἡ γὰρ φύσις αἰτία πᾶσιν τάξεως. τὸ δ̕ ἄπειρον πρὸς τὸ ἄπειρον οὐδένα λόγον ἔχει· τάξις δὲ πᾶσα λόγος. τὸ δ̕ ἄπειρον χρόνον ἠρεμεῖν, εἶτα κινηθῆναί ποτε, ... οὐκέτι φύσεως ἔργον _____________ 231

On contaminations of the Greek manuscript branch α (represented by E and Ψ) see below, p. CLXXIX—CLXXXV.

CXII

Introduction

Transl. Isḥāq, p. 12.5–13.1: wa-ḏālika anna l-ṭabīʿata hiya sababu l-niẓāmi fī kulli mā hiya lahū * * * walaysa min fiʿli l-ṭabīʿati bi-waǧhin mina l-wuǧūhi 232 an takūna l-ašyāʾu kānat sākinatan zamānan bi-lā nihāyatin ṯumma taḥarrakat ḥīnan Evidently, Isḥāq’s exemplar omitted πρὸς τὸ ἄπειρον — τὸ δ̕ ἄπειρον (252a1314) by homoioteleuton. No matter how this passage might be translated into Arabic, the translation would not provide any occasion for a saut du même au même within the Arabic transmission. The other three omissions caused by saut du même au même in Isḥāq’s main exemplar are as follows: 255b23-24: καὶ τὸ ποσὸν — κωλύῃ (p. 37.3) 262b23-24: πρὸς τὸ Δ — φέροιτο (omitted also in E1) (p. 86.4) 266b15-16: ὑπερβάλλει δυνάμεως — πεπερασμένης (omitted also in E1) (p. 111.8) Since the Arabic text as preserved in the Leiden manuscript is confirmed in each case by all witnesses of the indirect Arabic and Latin tradition, we can be sure that we are faced with omissions by homoioteleuton in Isḥāq’s exemplar and, by inference, that Isḥāq did not consult his additional manuscripts systematically (or, if so, that these must have displayed the same omissions and, by implication, were very closely related to each other). (2) Many notes referring to (an)other manuscript(s) can be traced back either to (Arabic) copies of Isḥāq’s translation or to other Arabic translations. As far as we can exclude in these cases any contamination, such notes do not distort our inferences about Isḥāq’s Greek exemplars. (3) The overwhelming part of Isḥāq’s translation can be assigned beyond doubt to one of the known branches of the Greek manuscripts (branch α, which displays also two of the above-mentioned sauts du même au même). Even if Isḥāq’s translation was perhaps contaminated by a more frequent use of additional manuscripts, most of the manuscripts at his disposal must have belonged to this branch of manuscripts (cf. point 1). So far the branch in question is represented by only one extant manuscript (E). It thus won’t make a great difference for the relevance of our inferences about this branch, whether Isḥāq ibn Ḥunayn used one or more than one of its representatives. (4) None of the variant readings reported in the notes points to a Greek manuscript entirely independent of or unrelated to ms. E. Rather some of them reflect different readings corresponding to the two hands of ms. E or γράφεται_____________ 232

The phrase laysa ... bi-waǧhin mina l-wuǧūhi points to οὐδέν τι instead of οὐκέτι (252a16)

V. Isḥāq ibn Ḥunayn’s Greek exemplar(s)

CXIII

notes in its margin, while others raise doubts about the reading of Isḥāq’s Greek exemplars, yet precisely by quoting the doubtful reading confirm that they preserved a text identical with or similar to ms. E, that is the text of the α-branch. The only verifiable exception is formed by the note on Physics IV 3, 210b10233, which supports the reading of ms. J, which is shown to be the source of contaminations in ms. E in the following chapter. In any case, it is not clear whether this note originates from Isḥāq’s hand.

VI. The Greek manuscripts of Aristotle’s Physics, by Pieter Sjoerd Hasper VI.1 Introduction The analysis of the preceding sections allows us to adopt the hypothesis that Isḥāq ibn Ḥunayn relied for his translation basically on a single manuscript and only occasionally consulted additional Greek manuscripts, which, the same analysis has shown, must have been closely related to his principal source. The question to be addressed in this section is how to situate this main exemplar Ψ within the Greek tradition as we know it through extant Greek manuscripts and ancient commentators, in the case of book VIII notably Simplicius – for book VIII there is not enough evidence to relate the extant manuscript tradition to the paraphrase of Themistius or to the commentaries or scholia from the commentaries by Philoponus and Alexander of Aphrodisias. In order to answer this question, however, it is mandatory first to get a clear view of the Greek tradition itself, for only on the basis of such a view will it be possible, in so far as there is evidence, to determine what readings found in Ψ are peculiar to itself, what readings may constitute readings shared with other manuscripts in that they share a common ancestor, and what readings are emendations, corrections and contaminations on the basis of other manuscripts. To some extent we find ourselves here in an evidential circle, for since, as will become clear below, Ψ is closely related to only a single extant manuscript, namely the famous Par. gr. 1853, generally known by its siglum E, the readings we find in Ψ must be used to determine the position of the common ancestor of E and Ψ within the Greek manuscript tradition. By itself this does not seem to constitute a problem, as long as we have sufficient evidence, provided by other (groups of) extant manuscripts, to bring to bear on the evaluation of the readings in E and Ψ. However, the evidence as provided by the apparatus of the presently _____________ 233

Cf. above, p. LXIV.

CXIV

Introduction

authoritative edition of Aristotle’s Physics, that by W. D. Ross, can easily be improved upon, since Ross’ collations (which he partly took over from Bekker) are far from flawless, and since Ross only based his edition on a very limited set of manuscripts. What is more, not only can the evidential base on which to assess the Greek tradition be easily improved upon, there is also room for methodological improvement as to the evaluation of the individual manuscripts. For although Ross used one important additional manuscript (namely the ninth century Vind. Phil. gr. 100, known by its siglum J), and also some occasional readings from the Arabic translation234, he claims to rely for the most part on the evaluations of the manuscripts put forward by Diels, in his booklet Zur Textgeschichte der aristotelischen Physik 235 . As it is, however, Diels’ evaluations are methodologically questionable. First, he argues on the basis of in fact a single passage, namely 212a6 (discussed above, p. XCIV–XCV, and again to be discussed below, p. CLXVI–CLXVII), that the extant Greek manuscripts go back to a common ancestor which featured erroneous readings not present in or presupposed by Simplicius’ commentary236. Thus the text as read by Simplicius would be completely independent from the whole extant Greek manuscript tradition. Second, since his estimation of the readings featured by the different manuscripts (the ones used by Bekker for books I–IV: Par. gr. 1853 [E], Laur. 87.07 [F], Vat. gr. 1027 [H] [for the last five chapters of book IV], Laur. 87.06 [G] and Vat. gr. 241 [I]) is that the readings to be adopted are distributed over all of them, and, more importantly, that readings to be rejected are also distributed over all of them, and that some of them also contain readings to be adopted which we do not find in Simplicius, he comes to the hypothesis that their common ancestor was a manuscript featuring many marginal and interlinear interpollations as well as that each manuscript also draws on other sources, be they manuscripts whose successors did not survive or late ancient commentaries. Still, he believes that the different manuscripts chose between all these different readings available to _____________ 234

As provided to him by A. Mansion, Étude critique (cf. footnote 170). To be more precise, Mansion reports the readings of the Arabic translation as known from Averroes’ commentary, which derive from, but are not in all cases identical to, the original Arabic translation edited in the present volume. 235 Cf. footnote 213. Ross, Aristotle’s Physics, p. 103, expresses his admiration for Diels’ argument. 236 On p. 5–6 Diels has an impressive list of 25 passages where, he claims, the commentators, usually only Simplicius, have the better reading against the Greek manuscripts. However, it is clear that also according to Diels none of these have the status of clear Trennfehler between the extant manuscripts and the commentators, in the way the passage at Physics IV 4, 212a6, according to Diels, would have; and indeed, most of these places are rejected by Ross.

VI.1 The Greek mss. of the Physics. Introduction (P.S. Hasper)

CXV

them in different ways. With regard to one particular case he even comes to a very strict verdict: “Wenn man nun das Schwanken der Lesarten betrachtet, so bietet sich als leichteste Erklärung dafür ein mit der Doppellesung ausgestatteter Archetypus dar, den I geschickt, F G ungeschickt, E stupid benutzt hat.”237 To rely on a single case for such far-reaching conclusions about the complete stemmatic separation between the whole extant Greek manuscript tradition and Simplicius is quite audacious. Moreover, Diels’ evaluations of the different readings in those manuscripts are often rather subjective238. Thus the hypothesis that there was an archetype with many interlinear and marginal interpollations rests on rather shaky grounds, especially in light of the small number of manuscripts used. It seems as if Ross, despite his praise of Diels, is partly aware of these problems, for in addition to taking over Diels’ arguments and examples, he sets up a stemmatic problem based on Trenn- and Bindefehler for which Diels’ hypothesis would be a solution. The problem he highlights is that there are both Bindefehler joining all the Greek manuscripts used which are Trennfehler over and against the commentators, notably Simplicius, and that there are for each Greek manuscript used Bindefehler joining them with the commentators, which are then Trennfehler with regard to the other Greek manuscripts used239. This is methodologically a sounder approach, but the evidence cited by Ross is far from convincing. In fact, he merely states that there are Bindefehler between the commentators and each of the manuscripts used, and only gives four passages which allegedly show that there are Bindefehler between E and Simplicius, the most important of the commentators in this respect. And there are problems about these four passages as well240. _____________ 237

Diels, Zur Textgeschichte, p. 20, a verdict also literally adopted by Ross, Aristotle’s Physics, p. 106, without clear acknowledgement: “Here it is most natural to suppose that the archetype was furnished with the two readings … and that I used it skillfully, FG unskillfully, and E stupidly.” 238 Indeed, Ross sensibly accepts only a few of them. 239 Ross, Aristotle’s Physics, p. 105, sets up the problem. To be sure, Diels, Zur Textgeschichte, p. 17–19, does point to readings shared by Simplicius and each of the other manuscripts used, but they concern only allegedly better readings, as they serve to argue against taking E as the only reliable manuscript. 240 The four allegedly erroneous readings mentioned by Ross are: (1) 249a23 ὁμωνύμων; (2) 255a35 τὸ1 om.; (3) 259b28 τὸ2 om.; (4) 262b2 καὶ ἄλλῳ om. Now (1) is based on incomplete collation, since Simplicius cites the correct ὁμωνυμιῶν at p. 1086.19, whereas the passage Ross has in mind (p. 1097.26) merely concerns a paraphrase; (2) need not necessarily be a mistake, since Aristotle can be quite liberal in omitting the definite article; and (3) and (4) may well be cases of coincident error due to haplography, for no

CXVI

Introduction

Thus Diels’ strong hypothesis, accepted by Ross, that the common ancestor of the manuscripts used by them featured many interpollations suffers from a fatal lack of evidence. And with this hypothesis also Ross’ conclusion, that in the case of Aristotle’s Physics we should give up on establishing a stemma based on the method of Trenn- and Bindefehler (which stemma should, at least to a considerable extent, determine the editorial choices)241, should go as well. There is no other option but to improve and enlarge the evidential base and, by drawing up lists of Trenn- and Bindefehler, to set up a stemma and then try to sort out the undoubtedly numerous contaminations. Fortunately it is easy to enlarge the evidential base, for Ross’ edition is for book VIII based on the following rather limited list of manuscripts and sources: E F H I J K

Par. gr. 1853 – 10th century Laur. 87.07 – 1125-1150 Vat. gr. 1027 – 12th century242 Vat. gr. 241 – 13th century Vind. Phil. gr. 100 – 9th century Laur. 87.24 – 12th century

T S P A

Themistius’ paraphrase – 4th century Simplicius’ commentary, with lemmata – before 549 Philoponus’ commentary, with lemmata – before 570 Fragments of Alexander of Aphrodisias’ commentary in Simplicius – ca. 200

Moreover, in his apparatus for the first four books Ross occasionally mentions, as stated above, readings underlying the Arabic translation with the siglum V. Of course, through the present edition of book VIII of the Physics in the Arabic translation, we get for the first time access to the readings of its exemplar, to be dated to the 9th century or earlier, in a systematic and comprehensive way. _____________ other manuscript of the many I have collated features the same mistakes. The fact that E is littered with small mistakes is relevant in this respect. 241 Ross, Aristotle’s Physics, p. 107, does not state this conclusion in precisely these words, though he comes quite close to it, but from his completely eclectic practice as he argues for it in the subsequent pages, it is obvious that he has precisely the conclusion as phrased thus in mind. 242 Ross relied on the collations of H by Bekker, who collated the manuscript before its restauration. The manuscript was damaged, however, in the restauration: for some folia the top line disappeared or became illegible. In that case I report H’s readings under the siglum HB, because it is only through Bekker that we have access to them.

VI.1 The Greek mss. of the Physics. Introduction (P.S. Hasper)

CXVII

However, for the purposes of this edition and of this chapter of the introduction, not only have all the manuscripts and sources used by Ross been collated again, so as to correct inevitable collating mistakes, but also a number of new manuscripts and sources have been collated, so as to get a better view of the readings available in the Greek tradition at the time of the Arabic translation, and to have a more accurate assessment of the stemmatic relations between all the manuscripts and sources. Thus in addition to the manuscripts and sources listed above, the following ones have been used as well: L N R Q w b e

Erlangen A 4 – 15th century Vat. Barb. gr. 136 – 12th century Vat. gr. 1025 – 1250-1325 Leiden Voss. Q3 – 12th century Par. Suppl. gr. 643 – 13th century Par. gr. 1859 – ca. 1300 Vind. Phil. gr. 64 – ca. 1450

Λ Ψ

Exemplar of the old Latin translation by James of Venice – 12th century243 Exemplar of the Arabic translation – 9th century or earlier

As P

In addition: scholia derived from Alexander’s commentary on the Physics244 In addition: scholia only found in Arabic manuscripts, newly discovered scholia from Vat. gr. 2208 and Par. Coisl. 166, and citations from Philoponus’ Contra Proclum245 Pachymeres’ paraphrase of the Physics – around 1300246.

Pa

The choice of the manuscripts has been dictated by a number of considerations. First, all the available manuscripts copied before the important watershed _____________ 243

See Fernand Bossier, Jozef Brams, Physica. Translatio Vetus. (Aristoteles Latinus; vol. VII,1, Fasc. 1.) Leiden 1990, p. xxi–xxvii. 244 They have been edited and translated in Rashed, Alexandre d’Aphrodise. 245 See P. Lettinck (transl.), Philoponus, On Aristotle Physics 5-8 from the Arabic Summary (London, 1994); Giannakis, Philoponus; M. Rashed, “Fragments inconnus du commentaire à la Physique de Jean Philopon”, in M. Rashed, L’Héritage aristotélicien: Textes inédits de l’Antiquité. Nouvelle édition revue et augmentée. Paris 2016, p. 751–777; A. Rescigno, “Nuovi frammenti del Commento di Filipono ai libri V-VIII della Fisica”, Studia graecoarabica 7 (2017), p. 75–104; and H. Rabe (ed.), Ioannes Philoponus: De aeternitate mundi. Leipzig 1899. 246 Edited by L. G. Benakis, Michael Psellos: Kommentar zur Physik des Aristoteles. Athens 2008 – for a discussion as to its authorship, see below.

CXVIII

Introduction

of 1204 should be taken into account – thus N, Q and Λ have been added247. Further, the results of Marwan Rashed’s research into the manuscript tradition of De generatione et corruptione should be taken into account and thus all the independent and non-contaminated manuscripts containing the Physics have been collated as well – but surprisingly enough this only leads to one further addition248, namely Par. Suppl. gr. 643249. Finally, Par. gr. 1859 and Vind. Phil. gr. 64 are the only two independent manuscripts containing the complete first version of the text of book VII250, and thus may be expected to have had access to an independent part of the tradition for the other books as well; these expectations are at least partially borne out by the collations, as we shall see. In addition, _____________ 247

According to G. Prato, “I manoscritti greci dei secoli XIII e XIV: note paleografiche”, in D. Harlfinger & G. Prato (eds.), Paleografia e Codicologia greca. Atti del II Colloquio internazionale (Berlino-Wolfenbüttel, 17–21 ottobre 1983), Alessandria 1991, volume I, p. 136, the manuscript Ambr. M 46 sup. is also to be dated to the end of the 12th century. However, the available scans do not allow for complete collations, so that it has not been used here. The indications are that the manuscript belongs to the group later to be called γ. 248 One of the reasons is that none of the three sources grouped together by Rashed as closely related to Par. gr. 1853 is of interest for the constitution of the text in Physics VIII: Marc. gr. 210 does not contain the Physics; Matr. 4563 is for the Physics clearly related to the vulgate tradition; Trapezuntios, whose translation of De generatione et corruptione is based on a manuscript related to E, also translated the Physics, but from inspection of two of the manuscripts of that translation (Vat. lat. 2988 and Vat. lat. 4534 – for the list of manuscripts, see J. Monfasani [ed.], Collectanea Trapezuntiana. Texts, Documents, and Bibliographies of George of Trebizond. Binghamton, NY, 1984, p. 702) – it appears that though there are shared errors with EΨ, they are very few. 249 Marc. gr. 214 (commonly known under the siglum Ha) has not been considered, because it should be dated to the thirteenth century and is for De generatione et corruptione an indirect copy of J. 250 All other manuscripts are direct or indirect copies of Vind. Phil. gr. 64: Laur. gr. 87.11, Riccard. 14, Mon. gr. 200, Oxford Auct. T.3.21, Par. gr. 1861, Par. gr. 2033, as well as Esc. Σ.III.02. Vat. gr. 1027 (H) contains only part of this first version, from 244b5b, but is for that part independent from Par. gr. 1859 and Vind. Phil. gr. 64 – see also Rashed, Alexandre d’Aphrodise, p. 442. It is, however, not the only manuscript featuring the first version from 244b5b; M.-L. Boureau, “La version α du livre VII de la Physique d’Aristote et son rapport aux familles byzantines a et b”, Elenchos 39 (2018), p. 99–148, has traced another twelve, but her list is far from complete, as without having checked all manuscripts yet, I have found an additional seven. Also Vat. gr. 241 (I) contains a part of the first version of book VII, from about the beginning of VII 3 onwards, but here again there are at least another six manuscripts which have the first version from there, among them three manuscripts which are important for book VIII as well: L, N and R (see also Boureau, La version α).

VI.1 The Greek mss. of the Physics. Introduction (P.S. Hasper)

CXIX

the paraphrase by Pachymeres has been used as a source, because it turned out to feature readings or interpretations which on a significant number of occasions can be connected to readings (almost) uniquely present in Ψ. The remaining manuscripts were added on the basis of clear resemblances to those already taken into account. All manuscripts available online, as well as quite a few more, were checked. Most were established to belong to or to be heavily contaminated by the vulgate tradition, whose clearest early representatives are J (Vind. Phil. gr. 100), Q (Voss. Q3) and w (Par. Suppl. gr. 643), but to which also F (Laur. gr. 87.07), H (Vat. gr. 1027) and I (Vat. gr. 241) are closely related. Thus it seems not necessary to take them into account when assessing the relations between the Arabic translation and the Greek tradition, nor, presumably, for the purposes of the constitution of the text. But L (Erlangen A 4) and R (Vat. gr. 1025) turned out to be very closely related to N (Vat. Barb. gr. 136)251, and, more importantly, together constitute a group independent from both the vulgate tradition (FHIJQw) and from the tradition to which E (Par. gr. 1853) belongs. It should be stressed that there is not yet an overview over the whole Greek manuscript tradition, and thus that it is possible that there are further manuscripts which are of greater interest to the history of the text or of importance to the constitution of the text. On the basis of this considerably enlarged and improved evidential base I shall try to construct a stemma for all these sources for book VIII of Aristotle’s Physics (with the exception of those which are too fragmentary, that is, the commentators or paraphrasts except Simplicius and Pachymeres) and then try to sort out the contaminations and corrections in some of these sources, especially in E and Ψ. It will turn out that the direct tradition divides into two branches, that the vulgate tradition constitutes a part of the one branch, but is heavily corrected and smoothed, that none of the manuscripts belonging to the other branch, among them E and Ψ, is free from contamination from the vulgate tradition, but that, fortunately, with the help of the readings we find in Simplicius, we can get a good view of the extent of this contamination.

VI.2 The three main groups of manuscripts and their stemmatic relations Now if we take this evidential base into account, it becomes immediately clear that we may discern four groups of manuscripts, and one isolated manuscript: _____________ 251

Perhaps the manuscripts Vind. Phil. gr. 75 (1445) and Mon. gr. 336 (15th century) should have been added as well, but since both contain clear amounts of contamination, unlike Erlangen A 4 and Vat. gr. 1025, and for the rest hardly provide readings deviating from LNR, their full collations would not have contributed significantly to the evidential base. See also footnote 262.

CXX

Introduction

the α group, consisting of E and Ψ, the β group, consisting of L, N and R, the γ group, consisting of FHIJQw and, at least partly, Λ, the δ group, consisting of b and e, and K. In this section first the basic outline of the stemma as far as the first three groups are concerned is discussed; since group δ and the single manuscript K can be shown to be heavily contaminated, their relations to the main stemmatic groups can only be determined at the next stage. I will identify in these first three groups the passages which betray their ancestral affiliations: what errors are shared by any two of these three groups? Subsequently all four groups and the one isolated manuscript are analysed for their internal structure and relations of contamination; since I want to finish with some observations about the exemplar of the Arabic translation, the discussion of the α group will conclude this chapter, after a discussion of the relations of the indirect sources, in the case of book VIII the commentary by Simplicius and the paraphrase by (?) Michael Psellus, to the extant manuscript tradition.

VI.2.1 Bindefehler between the β and γ groups That the β and γ groups shared a common ancestor can be argued for, first, on the basis of a list of errors or at least clearly less preferable readings common to them. I have identified the following list of passages: 251a22 προάγουσιν EΨK1 T[211.2] : προϊοῦσι K2beLNRFHIJQwΛ Sp[1128.2]

As Ross remarks, προϊοῦσι is clearly the lectio facilior, and unlikely to have been changed into προάγουσιν, while the reverse change is much easier.252 ὡς ἦν EΨw1 : ὡς εἶναι Ke2LNRFHIQΛ : ὡς εἴη? J2 : ὡς .. J1 : ὥστε ἦν be1?w2 [S][T] 251b6 ἦν EΨKbe : om. LNRHJ2QΛ : μὴ FIJ1?w [S][T] 251b6-7 ἀλλ’ ἐδεῖ μεταβάλλειν beLNRFHIJQwΛ : ἀλλὰ μετέβαλεν EΨ T[211.14C, Laur.85.14] : ἀλλὰ μετέβαλλεν K T[211.14-MSL] [S] 251b4

As Ross remarks, at 251b4 the reading of the β and γ groups does not make sense253. At b6 ἦν was probably corrupted into μὴ, and then struck in the other manuscripts of the β and γ groups. I would also argue that at b6-7 the reading ἐδεῖ μεταβάλλειν constitutes an unnecessary correction into an easier reading, for it is quite possible to make sense of the reading provided by EΨ: with the sentence εἰ τοίνυν μὴ ἀεὶ ἐκινεῖτο, δῆλον ὡς οὐχ οὕτως εἶχον ὡς ἦν δυνάμενα τὸ μὲν κινεῖσθαι τὸ δὲ κινεῖν, ἀλλ’ μετέβαλεν (vel μετέβαλλεν) θάτερον αὐτῶν Aristotle would then correctly be saying that things which in the past were not moving _____________ 252 253

Ross, Aristotle’s Physics, p. 688. Ross, Aristotle’s Physics, p. 688.

VI.2 The three main groups of Greek manuscripts (P.S. Hasper)

CXXI

all the time, but only started moving at some stage, one or the other of a pair of mover – moved thing changed (or: was in the process of changing) from not being capable of causing motion or being moved to being thus capable. 252a1

φθαρτικὸν EΨKbL Sc[1170.26] : φθαρτὸν eNRFHIJQwΛ [T]

It is likely that the reading of L is a correction of the obviously incorrect reading of one of its predecessors. Thus the β and γ groups shared the erroneous reading φθαρτὸν. 253a3-5 τοῦτο – δὲ μή EΨKbeLN2R2F2IJQ2w Sp[1189.34] : om. N1R1F1HQ1Λ

Not all manuscripts belonging to the β and γ groups omit these lines, but from the distribution it is clear that both β and γ left them out; in the ancestors of L and of IJw they must have been reinserted again, just as they have been by a second hand in N, R, F and Q (or alternatively these manuscripts report the state of their exemplar more accurately, there being a marginal or supralinear addition there). Clearly the omission is an error. 253b30 ὁ δὲ EKbe?J2w : ὅ τε LNRFHJ1QΛ : ἔτι δὲ ὁ Ψ

Having τε seems awkward in the context, but it is a common enough error for δὲ. 254a3

οὐδὲ EKbeJ2w Sl[1200.27] : οὔτε L1NRFHIJ1Q : οὐ L0 [Λ]

The sequence ἀλλὰ μὴν οὔτε without an οὔτε ... οὔτε construction seems very awkward and is exceedingly rare. 254a28 κἂν1 EbeFIJ2Qw : καὶ KLNRHJ1Λ? [S]

In the context (a κἂν εἰ ... κἂν εἰ ... construction) κἂν is required, but from the distrubition of καὶ it appears that both the β and γ groups must have featured it. 255b7

ἐμποδίζῃ EK Sp[1215.10]? : ἐμποδίσῃ beLNRFHIJQw [Ψ][Λ]

By itself ἐμποδίσῃ is not incorrect, but from the context it becomes clear that the aorist form is a corruption of the present form: τὸ γὰρ ψυχρὸν δυνάμει θερμόν, ὅταν δὲ μεταβάλῃ, ἤδη πῦρ, καίει δέ, ἂν μή τι κωλύῃ καὶ ἐμποδίζῃ (with κωλύῃ E2LNHJ Sp[1215.18] : κωλύει E1 : κωλύσῃ beRFIQw Sp[1215.12] : κωλύσε K [Λ] – thus the present κωλύῃ/κωλύει is found in all three groups, with the aorist κωλύσῃ therefore being a correction analogous to ἐμποδίσῃ; the same conclusion follows from consideration of the three parallel cases shortly after, at 255b11, 23 and 24). 256a3

ἂν … κινοῖτο EΨ?Kbe : ἄρα … κινοῖτο L[κινεῖται]NR : ἄρα … ἂν κινοῖτο FHJQwΛ

It is clear how the different readings came about: the correct text with ἂν, featured in EΨ, was corrupted into ἄρα, as LNR have it, and then corrected into

CXXII

Introduction

ἄρα … ἂν, as in the γ group. The reading of the β group, being grammatically incorrect, cannot have been the original reading, and if the reading of the γ group had been the original reading, it is quite difficult to explain the appearance of the other two readings, especially since paleographically the reading of the γ group is not conducive to the disappearance of ἂν. 256b22-23 μέν οὐχ ὑπ’ ἄλλου δὲ ἀλλ’ EΨbe1 As[617]?Sp[1227.13-20] : μέν ὑπ’ ἄλλου δὲ ἀλλ’ οὐχ e2LNRFHIJQwΛ : οὐχ ὑπ’ ἄλλου μὲν ἀλλ’ K

The two different readings lead to two very different arguments. According to the reading of the α group, Aristotle’s argument is that (1) there is something which can be moved (by another thing), but does not have a principle of motion within itself; (2) there is something which is moved by itself, and not by another thing; therefore (c) there is something which moves while being immovable. According to the reading of the β and γ groups, premiss (2) says that there is something which is moved by another thing, and not by itself. It is clear that the latter argument is incorrect, especially since there does not seem to be a real difference between premisses (1) and (2). The change to the βγ reading must have been brought about deliberately, perhaps because the argument on the α reading was perceived to be unclear? 257b16 αὐτὸ ἑαυτὸ K P[CP 89.22] : ἑκάτερον LNRFHIJQwΛ : αὐτὸ αὐτὸ E1 : τὸ αὐτὸ αὐτὸ E2 : αὐτὸ b : αὐτὸ αὑτὸ e Sp[1238.1] : αὑτὸ Ψ? Sp[1238.8]

The construction εἴ γε ἑκάτερον κινήσει ἑκάτερον is clearly an attempt to bring this phrase in line with 257b14 ἑκάτερον ὑφ’ ἑκατέρου κινεῖσθαι, and thus to simplify the text254. Given the readings we find in EΨKbe and Simplicius, the correct reading is probably αὐτὸ αὑτὸ. 258a2

μόνως – εἶναι E2beLNRFHJQ2wΛ : om. E1ΨK Sl[1240.22-1241.27]As[635] [Q1]

From a scholium from Alexander of Aphrodisias’ commentary on the Physics found in Par. Suppl. gr. 643 (scholium number 635 in Rashed’s edition) it appears that this sentence originates from his commentary; further support for this hypothesis can be derived from the fact that it is left out in the lemmata to Simplicius’ commentary (1240.22 and 1241.27). This means that EΨ feature the original text255. 258a21 μὲν2 EΨK As[641]Sc[1244.14]Pc[836.26] : om. beLNRFHJQwΛ

_____________ 254 255

See also Ross, Aristotle’s Physics, p. 439 and 701. As already suggested by Rashed, Alexandre d’Aphrodise, p. 556.

VI.2 The three main groups of Greek manuscripts (P.S. Hasper)

CXXIII

Again from a scholium from Alexander’s commentary (number 641 in Rashed’s edition), as well as from Simplicius’ and Philoponus’ commentaries we know that the omission of μὲν is a reading only introduced later into the manuscript tradition. 258a27 τὸ δὲ – γάρ KbeLNRFHJQwΛ Sγρ[1245.2-3: ἔν τισι] : om. EΨ A[apud S(1245.3-5)]Sγρ[1245.3-5: τὰ πλείονα τῶν ἀντιγράφων]

Simplicius reports that Alexander of Aphrodisias did not know this ‘addition’, and that most manuscripts of his time do not feature it either; only some have it. In the extant manuscripts the proportions have been reversed: only the α group retain the original text without the addition, while all other manuscripts canvassed here have it. It is, moreover, very difficult to make sense of the addition at this place. 258b3

φύσιν EΨKbe Sc[1246.25]Sp[1246.16]Pp[R-97v, 837.3]A[apud S(1246.25)] : δύναμιν LNRFHIJQwΛ

Not only is the reading φύσιν unanimously attested in the commentators’ tradition, including Alexander, it is also the lectio difficilior; δύναμιν may well have started its life as an elucidating supralinear addition. 258b11 εἴτε1 EΨ?KbeI As[648] : εἴτε οὖν LNRFHJQwΛ

The combination of εἴτε … εἴτε and οὖν after the first εἴτε is confined to late ancient authors; apart from the spuriously Aristotelian Problemata (935b10), we only find it in authors not earlier than the fourth century AD. The scholium from Alexander confirms that Aristotle did not write οὖν here. 262a9

τὸ EΨ?KbeF1wΛ Sp[1280.2] : τῷ LNRF2HIJQ

The context requires an accusative. Not all manuscripts of the γ group have the dative, but F and w may have corrected the mistake themselves (correction on the basis of other manuscripts seems less likely in such a case, but cannot be ruled out, especially not for F). 264b4

ἐκ EΨKI2Q2 Sp[1306.1] : μὴ ἐκ beLNRFHI1? : ἐκ μὴ JwΛ : ἐκ τοῦ Q1

The addition of μὴ amounts to a Verschlimmbesserung, stemming from the idea that it is impossible to have a change from and to white at the same time; that, however, is exactly the point of Aristotle’s argument. Given the distribution over the manuscripts, there are two possibilities: either the β and γ groups had μὴ ἐκ, which was then emended to ἐκ μὴ, in the way it features in J and w, presumably the ancestor of Q, and in Λ; or the common ancestor of the two groups had a supralinear μὴ which ended up at two different places.

CXXIV

Introduction

266a30 ἡ ἄπειρος EΨ[τὸ ἄπειρον?]K A[apud Sc(1325.20)] : ἡ ἄπειρος δύναμις beLNRFHIJQwΛ

In order to understand what Aristotle claims, it is not necessary to have δύναμις; it thus seems an addition. 266a32 ἐθέρμανεν E1[αι θερμανεν] : ἐθέρμηνεν E2KbeLNRFHIJQw [Λ]

The form with η is common in late ancient times, but cannot be found in Aristotle (nor in Plato and some other contemporaries). 266a33 τῷ E1Ψ Sp[1324.31] : ᾧ E2KbeLNRFHIJQwΛ

The scribe who changed τῷ into ᾧ, if it was a deliberate change, obviously did not understand the structure of the sentence. 266b12 ΖΘ EΨK As[802] : Ζ καὶ Θ beLNRFHIJQwΛ : ΘΖ Sc[1342.39]

The fact that Alexander’s commentary seems to presuppose a reading without καὶ makes it almost certain that the archetype was without it as well. 267a7

κινεῖ τι ἄλλο E2Kbe : κινεῖται ἄλλου LNRFHIJQΛ : κινεῖ ἄλλο E1Ψ* [Sp(1346.11-12) κινεῖ vel κινεῖ τι]

The reading κινεῖται is wrong in the context, and can easily be explained as a corruption of κινεῖ τι. But even more wrong is the genitive ἄλλου, which cannot be made sense of in any proper way. Thus there are quite a few errors or infelicitous readings common to the β and γ groups, while the α group features at those places the correct or preferable reading. Though it is possible to explain some of the infelicitous readings common to the β and γ groups as due to contamination, it is certainly not possible for all of the errors or infelicitous readings.

VI.2.2 Bindefehler between α and either β or γ? The list of errors and infelicitous readings common to the β and γ groups constitutes very strong evidence that these two groups descend from a common ancestor – call this common ancestor χ2. However, we are not yet allowed to conclude that both groups descend from χ2 in a straightforward way, unless there are no or very few errors or infelicitous readings common to the α and γ groups or the α and β groups. For if there are, then we must conclude that the picture is more complex, with at least one group being ‘mixed’ from several traditions.

VI.2 The three main groups of Greek manuscripts (P.S. Hasper)

CXXV

Now there is cause of concern in this respect, since in chapters 1 and 2 of book VII256 there seem to be shared errors between the α and β groups: 241b31 242a20 242b6 244b21

ΔΕΖ LR2FHIJQΛ – K : ΕΖ EΨNR1? ὃ FHIJQΛ – ΨK : om. ELNR τοῦ FHJQΛ – K : τοῦ αὐτοῦ EΨLNR – I τὸ FJQ – K : om. ELNR – I [Ψ]

It is immaterial that Ψ does not join E in error at 242a20, and that it is unclear whether it does so at 244b21. However one is to account for these Bindefehler (book VII, with its two different versions, might be a rather special case), this list does make it more pressing to check in the first place those places where the α and β groups might be joint in error against the γ group. Now there are quite a few places where the α and β groups share a reading against the γ group, but we need to find common errors. By using the text as constituted by Ross as a criterion for acceptability, it is possible to draw up a rather short list of the readings common to the α and β groups which, since they have not been adopted by Ross, might be unacceptable257: 250b21 251a5 251b1 251b26 252a6-7 252a35 253a35 254a7 254a32 256a31 256a31 *256a32 257b7

τὰς FHIJQw – Kbe T[209.10] : om. ELNR [Ψ] δὴ FHIJQwΛ – Kbe Sl[1125.25-aAM] : δὲ EΨLNR – Sl[1125.25F]T[209.22] καὶ FHIJQwΛ – be : ἢ EΨLNR – K [S][T] ἀεὶ FHIJQwΛ – Pp[823.20]? : om. EΨLNR – Kbe T[212.6] νομίζειν εἶναι ἀρχήν FHIJQwΛ – K : εἶναι νομίζειν ἀρχήν ELNR : νομίζειν ἀρχὴν εἶναι be [Ψ] [S][T] ὡς FHIJQw – Kbe : ὅτι ELNR – Sp[1186.20] [Ψ][Λ] [T] πάσας FHIJQw – L – be : ἁπάσας EKNR – Sc[1195.10-11] [Ψ][Λ] γὰρ FHIJQwΛ – Kbe T[216.23]Pc[827.17] : γὰρ καὶ EΨLNR [S] τὸ2 FHIJQw – K : om. ELNR – be [Ψ] [S] τὴν βακτηρίαν FHIJQwΛ - Kbe : τῇ βακτηρίᾳ EΨLNR ἄλλο FHIJQwΛ – K : ἀλλο τι EΨ?LNR – be τινι F2IJQ1wΛ – Kbe As[605]Sc[1224.3] [F1 τινὰ] : τι2 EΨLNR – HQ2 δυνάμει2 FHIJQwΛ : δυνάμει κινούμενον be : κινούμενον EΨLNR – K As[627]Sp[1235.14-15]Pp[834.25]

_____________ 256

There is one in chapter VII 4 as well: 250a12 τὸ2 FHJQΛ – be Sl[1106.6] : om. EKLNR – I, but this may be a coinciding error, since the context is conducive to haplography. 257 That this list is not much longer is due to the fact that Ross has a strong preference for readings found in E.

CXXVI

Introduction

ἐνεῖναι FHIJQΛ – e : ἐνεῖναι/εἶναι τί EΨNRL – Kb – w [Sp(1246.11: ἑνεῖναι vel ἐνεῖναι τι)] 260a4 τὴν αὐτὴν ἀεὶ FHIJQwΛ : ἀεὶ EΨLN – Kbe [R: ἀεὶ τὴν αὐτὴν] 260a6 ὑπὸ τοῦ κινουμένου μὲν FH1IJQwΛ – e Pγρ[894.10] : om. EΨLNR – Kb – H0 Sc[1263.16]Pl[894.5] 262b15 οὐ γὰρ ἅμα Ross : οὐκ ἄρα ἅμα E1?FHIJQwΛ – be : οὐχ ἅμα ἄρα E2Ψ?LNR – K Sc[1285.31] 262b18 ἅμα κινεῖσθαι FHIJQw – K : κινεῖσθαι ΨΛ : κινεῖσθαι ἅμα ELNR – be [Sp(1286.9): ἅμα hab.] 263b12-13 κοινόν καὶ τοῦ προτέρου FHIJQwΛ – b : καὶ τοῦ προτέρου κοινὸν EΨLNR – Ke [S] 263b13 ταὐτὸν FHIJQw – be : τὸ αὐτὸ ELNR – K [Ψ] 263b16 χρόνῳ FHIJQwΛ – b Sp[1295.14]? : ὅλῳ χρόνῳ EΨLNR – Ke A[751] 263b19 οὐ FHIJQw : μὴ ELNR – Kbe [Λ] [Ψ][S] *264a32 τῷ F2HIJQΛ – Ke Sc[1304.28] : τὸ EΨ?LNR – b – F1w 264b15 ταύτῃ FHIJQwΛ – be : ταῦτα EΨLNR – K [S] *264b19 αὑτὸ : ἑαυτὸ FQΛ Sl[1308.33-A] : αὐτὸ HIJw T[442.25]Sp[1309.15] Pp[847.30] : τὸ αὐτὸ EΨ?LNR – be Sl[1308.33-a] [K] *264b26 ταὐτὰ vel τὰ αὐτὰ FHIJΛwQ – E2be1 : τὰς αὐτὰς E1ΨLNR – Ke2 264b30 εἶναι FHIJQwΛ : γίγνεσθαι EKLNR – be Sp[1311.27]? [Ψ] 265a28 ἐπ’2 FHIJ2QwΛ – Sp[1315.9] : om. EΨLNR – Kbe T[232.28] [J1] 265b30 ἢ FHIJQwΛ : καὶ E2ΨKbeLNR – be Sp[1319.22, 25] [E1 ἰδια] 266a2 τὴν κατὰ τόπον κίνησιν FHIJQwΛ – be : κατὰ τόπον EΨLNR – K Sc[1320.12] 266a8 ἀΐδιον ἐνδέχεται FHIJQwΛ – be – NR : ἐνδέχεται ἀΐδιον EL Sp[1320.3334] [Ψ] 266b33 κινεῖσθαι EKb2FJwΛ : κεκινῆσθαι HIQ – e – L Scp[1345.1-aA, 1345.2Exc.]Sγρ[1345.9] : καὶ κινεῖσθαι ΨNR – Scp[1345.1-Exc., 1345.2-aA, 1345.12, 1345.21]Sγρ[1345.10] [b1] 258b4

Of all these places, the reading shared by the α and β groups can only be thought to be problematic for the places indicated with an asterisk, for intrinsic reasons for 264a32, 264b19 and 264b26, while it is suspect for 256a32 because Alexander of Aphrodisias clearly had the same reading as the γ group. Now, I do not think that at 264b26 τὰς αὐτὰς is impossible, since it may pick up on περιφερείᾳ from the beginning of the sentence. Further, at 264b19 τὸ αὐτὸ seems possible, to form a contrast with ἄλλο in the next clause. Finally, at 264a32 the whole phrase ἀνάγκη πρότερον ἠρεμῆσαι πρὸς τὸ Γ clearly has to be understood in the sense of arriving and then resting at C, so that the accusative τὸ is appropriate258. This _____________ 258

In the so-called ‘pregnant sense’ as discussed in R. Kühner & B. Gerth, Ausführliche Grammatik der griechischen Sprache. Vol. II Satzlehre 1 (3rd ed; Hannover & Leipzig, 1898),

VI.2 The three main groups of Greek manuscripts (P.S. Hasper)

CXXVII

leaves 256a32, where Alexander provides strong evidence against the reading of the α and β groups. However, there are several ways of explaining away this case, for example by postulating contamination from Alexander in the γ group, after an error very early in the tradition from ΤΙΝΙΚΙΝΗΙ to ΤΙΚΙΝΗΙ, or by hypothesising the same error occurring twice. Thus there is no substantive evidence based on clear errors that the α and β groups share a common ancestor. In order to find out whether there are any clear errors joining the α and γ groups, I use the same procedure, and first list the places for which Ross, of course without knowing the readings of the β group, decides against the reading shared by the α and γ groups at least partly in favour of that offered by, among others, the β group (except for orthographical variants). This list is very short and only features two places, both of them are rather complicated: 258b11 ἀΐδιον ὃ πρῶτον κινεῖ Sp[1250.35, 1251.3] – Ross : ἀΐδιον ὃ κινεῖ LNR – E2be As[648, 649, 650]?T[223.14] : ὃ πρῶτον κινεῖ E1?ΨFHIJwΛ – K : ὃ πρῶτον κινεῖ .. Q1 : ὃ πρῶτον κινεῖ ἀΐδιον Q2

The reading shared by the α and γ groups is clearly wrong, since the passage does not make much sense in the context without ἀΐδιον. Now Ross only had E2 and Themistius as sources for the reading which we find in the β group, so it is small wonder he went with the reading suggested by Simplicius’ paraphrase, which can be described either as an insertion of ἀΐδιον into the text of the α and γ groups, or as an insertion of πρῶτον into the text of the β group. However, especially the fact that Alexander’s scholia suggest that he also read the text as provided by the β group (as well as be), underpinned by the consideration that either the β reading or the αγ reading is the result of deliberate intervention, leads one to the conclusion that the β reading is probably the original one. 265b7

τὸ1 μένειν E2K[τῷ]beLNR T[233.4, 6] : μένειν HΛ : μένει E1ΨFHIJQw Sp[1316.19] || ἀεί E2beLNRHIJQwΛ : αἰεὶ Kbe : ἀεὶ ἀεὶ E1? Sp[1316.22]? T[233.4, 6]? : ἀεὶ εἰ Ψ?

The second passage is even more complicated, since there are grammatically two very different readings on offer. On the one hand, according to the β group, as well Kbe and Themistius, Aristotle would have said: διὰ δὲ τὸ τοῦτο μένειν ἀεί τε ἠρεμεῖ πως τὸ ὅλον καὶ κινεῖται συνεχῶς – ‘and because this [scil. the centre] _____________ p. 453. Moreover, even if the accusative were problematic, it would still be probable that it would be the reading of the archetype, since it can also be found F and w, and thus in part of the γ group as well. As it seems difficult to imagine the accusative being the result of correction, one should assume that the dative τῷ is the result of correction for all manuscripts which have it. (In EΨ we also find an attempt at correction, with the insertion of τὴν before πρὸς, presumably so as to let κίνησιν be understood.)

CXXVIII

Introduction

stays put, the whole in a way is always at rest and also moves continuously’. On the αγ version (the dissent of H and perhaps Λ is a matter of contamination), on the other hand, Aristotle would have said: διὰ δὲ τοῦτο μένει ἀεί τε ἠρεμεῖ πως τὸ ὅλον καὶ κινεῖται συνεχῶς – ‘and because of this [scil. because the centre is the goal without being reached], the whole stays put and in a way is always at rest and moves continuously’ (where there might be a variant reading suggested by the α group as well as Simplicius of μένει ἀεὶ ἀεί τε ἠρεμεῖ). There are three reasons why the reading of the β group (as well as Kbe) is probably the correct one here as well. First, the addition of τὸ is more difficult to explain than its omission. Second, the function of τε changes between the two versions: on the β version it combines with the next καὶ to state, quite appropriately, that the whole is both at rest and in motion, while on the αγ version it merely serves to combine the parts of the sentence. Third, the argument on the β version is better, since it explicates the crucial premiss that the centre stays put. Thus there are two places where the β group offers a better text against the α and γ groups. Obviously that is not enough to cancel out the solid evidence that the β and γ groups go back to a common ancestor; rather, they should be explained in another way, most probably by postulating contamination in the α group – as we shall see below, there are such cases. Thus we may conclude that the basic stemmatic relations between the α, β and γ groups can be represented as follows:

ω χ2 α

β

γ

VI.3 The β group: Vat. Barb. gr. 136, Vat. gr. 1025 and Erlangen A 4 None of these manuscripts have been used before for an edition of the Physics, though Vat. Barb. gr. 136 (N) is already mentioned by Diels in the apparatus to Simplicius’ commentary on the Physics. Erlangen A 4 (L) is an important witness to the text of De motu animalium, as well as De longitudine et brevitate vitae, De juventute and De respiratione, in which works it is closely related to Berlin 1507, forming a completely independent branch of the stemma, featuring many readings

VI.3 The β group (P.S. Hasper)

CXXIX

clearly superior to all other manuscripts259. It is, however, also a rather odd collection of texts that are assembled in this manuscript, which suggests that it or one of its ancestors may have been drawn from several exemplars, and that thus the importance of L for different Aristotelian texts may well be a matter of accident. Finally, Vat. gr. 1025 (R) is a manuscript which is not even well described260. That LNR constitute a group is evident from the following list of shared clear errors unique to them: 253a13 254a5 255a22 255b16 256a22 260b28 261a21 262b25 263a10 263b26 264a17 265a2 265a8 267b3

ἴσως EΨKbeFHIJQwΛ Sc[1190.25] : οὕτως LNR δ’ EΨ?K[δὲ]beFHIJQwΛ : om. LNR [S] κινητικός EΨKbeFHIJQwΛ : κινητικὸν LNR [S] το μὲν τῷ .. τὸ δὲ τῷ EΨKbeFHIJQwΛ : τῷ μὲν τὸ .. τῷ δὲ τὸ LNR ἔστιν ἐπελθεῖν EΨKbeFHIJQwΛ Sl[1222.21] : ἐπελθεῖν ἐστι LNR τῆς EΨKbeFHIJQw Sc[1269.26] : om. LNR [Λ] κινήσεων EΨKbeFHIJQwΛ Pc[901.1] : κινουμένων LNR στῆναι EΨKbeFHIJQwΛ : στῆναι τε LNR ὁμολογουμένως EΨKbeFHIJQwΛ : ὁμολογούμενόν LNR ᾖ EΨKbeFHIJQwΛ? Sl[1296.36] : om. LNR φέρεται E2Ψ?KbeFHIJQwΛ : φέρετε E1 : ἐφέρετο LNR [S] ταὐτὰ E1beFHIJQwΛ Sl[1311.14] : τὰ αὐτὰ E2K : ταὐτὸ Ψ? : ταῦτα LNR εἴρηκε E2ΨbeFHIQwΛ : εἴρηκεν KJ : εἴρη E1 : εἴρηται LNR τὸ οὕτω EΨKbe[οὕτως]FHIJ2[οὕτως]Q2[οὕτως]w : om. Λ : τοῦτο τῷ LNR [J1Q1]

_____________ 259

The importance of the Erlangen manuscript, in combination with ms. Berlin 1507, was noticed for the first time by Pieter De Leemans, [Aristotelis] De progressu animalium. De motu animalium. Translatio Guillelmi de Morbeka. (Aristoteles Latinus, vol. XVII 2.II-III.) Turnhout 2011, e.g. p. ccxiv, identifying Erlangen A 4 and Berlin 1507 as part of the independent manuscript tradition postulated by Nussbaum, but presumed by her to be lost. These two manuscripts provide the most important new manuscript evidence on which the new edition by Oliver Primavesi is based – see: O. Primavesi, “Introduction Part II: The Text of De Motu Animalium”, in C. Rapp and O. Primavesi (eds.), Aristotle’s De Motu Animalium. Symposium Aristotelicum. Oxford 2020 (with a new critical edition of the Greek text by Oliver Primavesi), p. 67–156, at p. 103ff., especially p. 122. For the other three works the assessment of the importance of ms. Erlangen A 4 is based on my own collations (though for the De respiratione De Leemans, p. ccxvi, reports that B. Proietti, La Tradizione Manoscritta del Trattato Aristotelico De respiratione. Tesi di diploma in Paleografia Greca, Università degli Studi di Cassino, 1999 [non vidi], had already established its importance), but it was to be expected given the situation for De motu. 260 For example, on the Pinakes website there is no information to be found, nor on the website of the Commentaria in Aristotelem Graeca et Byzantina. It is mentioned, though, by Bekker as containing part of the α version of book VII of the Physics (ad 243a11).

CXXX

Introduction

This is confirmed by quite a few places261 where LNR share a reading against all other manuscripts: ἀναγκαῖον EΨ?be Sp[1128.4]? : δεῖ F[ἀεὶ]HIJQwΛ T[211.5]? : ἀναγκαῖον δεῖ K : ἀναγκαῖον δὴ LNR 253a31 τε πάντων τῶν ἀπορουμένων EΨ?K : πάντων τε τῶν ἀπορουμένων FHIJQwΛ : τε τῶν ἀπορουμένων ἁπάντων be : τε τῶν ἀπορουμένων LNR [S: πάντων hab.] 254a21-22 ἠρεμεῖν … κινεῖσθαι1 EKbeFHIJQwΛ : κινεῖσθαι … ἠρεμεῖν LNR [Ψ] [S] 256a14 ὑπὸ2 E1ΨKF2HIJQwΛ : ὑπὸ τοῦ E2beF1 : ὑπό τινος LNR 259a21-22 ἐπιβλέψας ἐπὶ τὰς ἀρχὰς EKbeFHIJQwΛ Sc[1257.8]Pl[888.17] : ἐπὶ τὰς ἀρχὰς ἐπιβλέψας LNR 259b2 τὸ2 EΨKbeFHIJQw Pl[889.31] : om. LNR [Λ] 259b17 ἑαυτὸ EΨKbeFHIJQwΛ Sc[1259.9]Pl[891.14] : om. LNR 260a26 δ’/ δὲ EKbeFHIJQwΛ Sl[1265.8] : δὴ LNR [Ψ] 263b12 μὲν EΨKbeFHIJQwΛ : om. LNR 263b21 τὸ ὕστερον EΨbeFHIJQwΛ : τοῦ ὑστέρου Ross : τοῦ Α ὕστερον K : ὕστερον LNR [S] 264a28 μὲν EΨKbeFHIJQwΛ : αἱ LNR 265a34 μέσον καὶ τέλος ΨKbeI Sp[1315.19, 24]T[232.31-32] : μέσον EFHJQwΛ : τέλος καὶ μέσον ἐστὶν LNR 266a31 γε EKbeFHIJQw Sc[1324.21, 1325.29, 1339.28] : om. LNR [Λ][Ψ] 267a7 ἔτι ἐστίν EKbeFHIJQwΛ Sc[1348.9, 1349.13] : ἔστιν ἔτι LNR [Ψ] 267b24 τὸ δέ γε E2beFHIJQw : τὸ γε K : δὲ τὸ γε E1 : τὸ δὲ LNR [ΨΛ: δὲ hab.] [S] 251a25

Moreover, these three manuscripts share, uniquely among the manuscripts considered here, an overview of topics at the beginning of the book262. In addition, N and R share, uniquely, after the text of the Physics, the text of De mundo. The importance of this group for the textual tradition of the Physics also appears from the places at which it uniquely among the extant manuscripts _____________ 261

I have not listed places which concern merely orthographical differences (including elision), even though there are many. 262 I have now found the same overview (kephalaia) in quite a few further manuscripts: Ambr. B 96 sup. (= 118) (15th cent.), Ambr. Q 1 sup. (= 657) (15th cent.), Marc. gr. IV.4 (end 15th cent.), Mon. gr. 336 (15th cent.), Vat. gr. 251[partim] (13th/14th cent.), Vat. gr. 256 (1320), Vind. Phil. gr. 75 (1445), as well as the second hand in Par. gr. 1853 (E) (10 or 11th cent.), but I have not been able to check all manuscripts yet. Of these only LNR and Vat. gr. 251 contain in Physics VII the α-version from VII 3 onwards, and this close relation is borne out by the number of variants in the kephalaia they share against the other manuscripts, as well as by the fact that they seem to be the only four manuscripts (in addition to the second hand in E again) which have similar kephalaia for book VII (see also Boureau, La version α, p. 110–111).

VI.3 The β group (P.S. Hasper)

CXXXI

testifies to the correct reading. There are no such places in Physics VIII, but elsewhere in the Physics there are, for example in book III: 201a10-11 ἡ … ἐντελέχεια … κίνησίς ἐστιν EΨbeFHIJQΛR : τὴν … ἐνεργείαν … λέγω κίνησιν εἶναι LNw[ἐντελέχειαν] M Sl[413.13-14]Sc[413.21, 1137.3] T[69.6-7: ἐντελέχειαν] 201a11

ἐντελεχεία EΨwFHIJ[ἐντελεχία]QΛR Aγρ[apud S(414.19) ἔν τισιν ἀντιγράφοις] Porph.[apud S(414.19) ἔν τισιν ἀντιγράφοις] Sp[412.21] Pc[350.10]T[69.7], cf. Theophr.[apud S(413.7)] : ἐνεργείαν Ψs.l.LN M Sl[413.13]Sc[413.21, 414.16, 1137.4]Sp[408.16, 417.31], cf. Theophr.[apud S(413.2)]

202a15

δὲ EbeFHIJQΛR : om. LN M Pc[375.7]

Here the siglum M refers to the excerpt from Physics III 1-5 and V 1-3 in Metaphysics K. This excerpt was already known to Alexander, and thus constitutes an independent tradition. At each of these three places LN (R does not belong to this group for this part of the Physics), once joined by w (presumably because of contamination), share a reading with at least one late ancient commentator and M, it thus making very likely that these readings are the correct ones. There is not enough evidence to assess definitely the internal structure of the β group, for there are not many places where they diverge. Only by reviewing the evidence for books V–VIII as a whole will it be possible to come to a definite verdict (for books I–IV R features a text which is not clearly akin to LN anymore). However, the first indications are that N and R are closer to each other than either is to L, without R being a copy of N263. It will be very difficult to identify contamination at the level of the β group as a whole: any potential case of contamination in the β group from the γ group could be interpreted as a shared reading, and any potential case of contamination in β from the other branch of the stemma could also be taken to be the correct reading (since clear errors usually do not lend themselves for contamination). The only exception will be if the process of contamination leaves a trace. In Physics VIII there is one place where there might be such a trace in β: 251a25 ἀναγκαῖον EΨ?be Sp[1128.4]? : δεῖ F[ἀεὶ]HIJQwΛ T[211.5]? : ἀναγκαῖον δεῖ K : ἀναγκαῖον δὴ LNR

It could be that the reading δή of LNR derives from the δεῖ of the γ group – then we have here a case of contamination in the β group, just as we have in K. _____________ 263

Thus I doubt the claim of Boureau, La version α, p. 131, that R is a copy of N for the part of book VII where they both give the text of the first version.

CXXXII

Introduction

It is, however, equally possible that the δεῖ of the γ group was the result of corruption of the δή of LNR (it is possible to have an apodosis with δή after the protasis264, though in this case it might be a later addition), after which in the γ group the superfluous ἀναγκαῖον was struck. Small numbers of Bindefehler could also be a sign of contamination, but as I have argued above, there is no real evidence for such joint errors shared by the α and β groups. As we shall see below, however, the δ group as well as manuscript K are closely related to the α group, so that we will have to check whether there is real evidence for Bindefehler between the β group and the δ group and/or K as well. As we shall see, there is no such evidence in either case.

VI.4 The γ group: Vind. Phil. gr. 100, Vossius Q3, Laur. 87.07, Vat. gr. 1027, Vat. gr. 241 and Par. Suppl. gr. 643 as well as the old Latin translation With the exception of Vossius Q3 (Q), Par. Suppl. gr. 643 (w) and the exemplar of the old Latin translation (Λ), this group was already recognised by Ross, who abbreviates its common readings with Λ (not to be confused with the siglum Λ as used in the present work). There are not many shared errors unique to them which make them out being one group: 251b16 γε EKbeLNR : τε FHIJQw [Λ] 252a18 καὶ EΨbeLNRI2 Sp[1185.1]? : om. KFHI1JQwΛ [T] 257b5 μανθάνοι EKbeLNR Sc[1234.14, 1249.29, 1250.6] : διδάσκοιτο FHIJQwΛ [Ψ] 257b33 ἂν ταῦτα κινοῖ EKbeLNR : κινοῖ ταῦτα FH[κινεῖ]IJQ : ἂν κινοῖ ταῦτα wΛ [Ψ] 259a16 ἀεὶ1 EΨKLNRI2Q2 Sp[1254.34]Pp[838.29] : αἰεὶ be : δ.. J1 : δεῖ FHII1?J2 Q1?w [Λ]

However, there are very many readings uniquely shared by them: 250b18

ὅσοι EΨKbNR : ὅσους eL : ὁποσοι FHIJQwΛ

_____________ 264

With ἀναγκαῖον in the apodosis there might be an example in Magna moralia 1189a23 (Susemihl entertains it seriously), and there is a clear example in Galen, Ars medica I 394.18. With ἀνάγκη in the apodosis there are in Aristotle examples in De caelo I 8, 276a15 (accepted by Moraux, against E), Ethica Eudemia II 4, 1221b32 (see also H. Bonitz, Observationes criticae in Aristotelis quae feruntur Magna Moralia et Ethica Eudemia, Berlin 1844, p. 14, scolding Bekker for not recognizing that the apodosis starts with the particle δή), De generatione et corruptione I 6, 322b5 (rejected by Rashed, with E, against most manuscripts) and Rhetorica II 11, 1388a36, and further for example in Xenophon, Anabasis 1.3.5, Plato, Euthyphro 6b1, and Plato, Sophistes 264b2.

VI.4 The γ group (P.S. Hasper) 251b19 251b26 252b20 253a31 256a3 256b1 256b2 256b3-4 256b5 256b16 256b25-26 256b26 257a9-10 257a17-18 257b7 257b14 257b25 257b34 258a22 258a26-27 258b11 258b24 258b24-25 258b30 259b5 259b5 259b33 260a4

CXXXIII

καὶ EΨKbeLNR Sl[1156.4-a]T[212.2] : om. FHIJQwΛ Sl[1156.4-AFM] Sp[1166.33]? εἶναι EΨKbeLNR T[212.6] : εἶναι ἀεὶ FHIJQwΛ P[823.20]? [S] κινήσεως EKbeLNR : κινήσεως ἐνιότε FHIJQwΛ [S] τε πάντων τῶν ἀπορουμένων EΨ?K : τε τῶν ἀπορουμένων ἁπάντων be : τε τῶν ἀπορουμένων LNR : πάντων τε τῶν ἀπορουμένων FHIJQwΛ ἂν … κινοῖτο EΨ?Kbe Sl[1219.12] : ἄρα … κινοῖτο L[κινεῖται]NR : ἄρα … ἂν κινοῖτο FHIJQwΛ ἤτοι εὐθὺς τὸ κινούμενον EKbeLNR : τὸ κινούμενον ἤτοι εὐθὺς FHIJ QwΛ κινεῖται EKbeLNR : κινήσεται FHIJQwΛ ταὐτὰ vel τὰ αὐτὰ συμβήσεται EKbeLNR : συμβήσεται τὰ αὐτὰ FHIJ QwΛ ὑπάρχει τοῖς πράγμασιν EKbeLNR : τοῖς πράγμασιν ὑπάρχει FHIJQ wΛ ἀνάγκη EKbeLNR : ἀνάγκη μὲν FHIJQwΛ ἐπειδή γε EbeLNR : ἐπειδὴ K : ἐπειδήπερ FHJQwΛ [I] εἶναι ποιεῖ EKbeLNR : ποιεῖ εἰ J1? : ποιεῖ εἶναι FH{I}J2QwΛ διδάσκεσθαι τὸ διδάσκον EKbeLNR : τὸ διδάσκον διδάσκεσθαι FHIJ QwΛ πᾶν .. εἶναι EΨKLNR : πᾶν … ἐστὶ be : ὅτι .. ἔσται FHIJQwΛ κινούμενον EΨKLNR As[627]Sp[1235.14-15]Pp[834.25] : δυνάμει κινούμενον be : δυνάμει2 FHIJQwΛ κινοῦν EΨKbeLNR Sl[1237.3] : κινεῖν FHIJQwΛ κίνησιν EΨKbeLNR Sl[1239.33-AFM] : κίνησιν καὶ FHIJQwΛ Sl[1239.33-a] ἀναγκαῖον EΨKbeLNR : ἀναγκαῖα FHIJQwΛ ἀναγκαῖον εἶναι συνεχές EKbeNR : εἶναι ἀναγκαῖον συνεχές L : συνεχὲς ἀναγκαῖον εἶναι FHIJQwΛ κινεῖται δὲ ἡ τὸ Β EKbeLNR : τὸ δὲ Β κινεῖται FHIJQwΛ εἶναί τι EΨKbeLNR Sc[1251.15] : τι εἶναι FHIJQwΛ μή EΨ?KbeLNR : μὴ εἶναι FHIJQwΛ ἅπαν ἔχειν ἀνάγκη μέγεθος EKbeLNR : ἀνάγκη ἅπαν μέγεθος ἔχειν FHIJQwΛ [Ψ] οὔτε πάντα αἴτια EΨKbe[αἴτιον]LNR Sc[1252.32] : αἴτιον οὔτε πάντα FHIJQwΛ ὁρᾶν ἡμᾶς EKbeLNR : ἡμᾶς ὁρᾶν FHIJQwΛ [Ψ] ποτε ὄντα EKbeLNR : ὄντα ποτε FHIJQwΛ [Ψ] ὑπὸ τούτου κινούμενον EKbeLNR Pl[893.4] : κινούμενον ὑπὸ τούτου HIJQwΛ : κινούμενον F ἀεὶ EΨKbeLN : ἀεὶ τὴν αὐτὴν R : τὴν αὐτὴν ἀεὶ FHIJQwΛ

CXXXIV

Introduction

κινεῖται ἢ ἠρεμεῖ EΨKbeLNR Sc[1263.37-1264.1] : ἠρεμεῖ ἢ κινεῖται FHIJ[ἠρεμεῖν]QwΛ 260a13 κινεῖται ἀεὶ EKbeLNR : ἀεὶ κινεῖται FHIJQwΛ Sc[1264.1] [Ψ] 260a15 κινεῖται2 EΨKbeLNR Sc[1264.4] : μεταβάλλει FHIJQwΛ 260a21-22 ἐνδέχεταί … εἶναι EKbeLNR Sc[1264.29]As[703] : εἶναι … ἐνδέχεται FHIJQwΛ [Ψ] 260a31-32 τροφὴ γὰρ λέγεται τῷ ἐναντίῳ τὸ ἐναντίον EKbeLR : τροφὴ γὰρ λέγεται τὸν ἐναντίον τῷ ἐναντίῳ N : om. w : τὸ γὰρ ἐναντίον τροφὴ λέγεται τῷ ἐναντίῳ FHIJQΛ, cf. Sp[1265.24] [Ψ] 260b1 δεῖ τι EΨKbeLNR : δεήσει FHIJQwΛ 260b16 ἐφ’ ἑτέρων EΨKbeLNR : καὶ ἐπὶ τῶν ἄλλων FHIJQwΛ 263b18 λευκὸν ἀληθὲς εἰπεῖν EKbeLNR : ἀληθὲς εἰπεῖν λευκὸν FHIJQwΛ 263b23 ἐν ἐκείνῳ πρῶτον EΨKbeLNR : πρῶτον ἐν ἐκείνῳ FHIJQwΛ 264a16 φέρεται πρὸς τὸ Γ EΨKbe : ἐφέρετο πρὸς τὸ Γ LNR : εἰς τὸ Γ ἐφέρετο FHIJQwΛ 264a23 τῶν εἰρημένων τινὰ κινεῖται E1[κινῆται]E2Kb2eLNR Sc[1304.11] : τῶν εἰρημένων κινεῖται b1 : κινεῖται τινὰ τῶν εἰρημένων FHIJQwΛ [Ψ] 264a27 στέρησις E1Ψ? : στέρησις τῆς E2Kbe[ἐστι τῆς]LNR : στέρησις ἐστὶ HIJ[ἐστιν]QwΛ : ἐστὶ στέρησις F 264b14 κατ’ εὐθεῖαν EKbLNR : ἐπ’ εὐθείας FHIJQwΛ 264b29 οὐδὲ τὰς ἄλλας ἐνδέχεται KbeLNR Sp[1311.26] : οὐ τὰς ἄλλας ἐνδέχεται EΨ : οὐκ ἐνδέχεται οὐδὲ τὰς ἄλλας FHIJQwΛ 264b30 γίγνεσθαι E[γείνεσθαι]KbeLNR Sp[1311.27]? : εἶναι FHIJQwΛ [Ψ] 265a5 τούτων τινὰ EKbeLNR : τινὰ τούτων HIJQwΛ : τούτων F 265a18 ἅμα δ’ EΨ?KbeLNR : ἀλλ’ FHIJQwΛ 265b12-13 ἀνωμαλῶς ἀπὸ τῆς ἀρχῆς EKbeLNR Sc[1317.17] : ἀπὸ τῆς ἀρχῆς ἀνωμαλῶς FHIJQwΛ [Ψ] 265b23 πρῶτον EΨKbeLNR Sp[1318.29 ἐξ ἀρχῆς] : πρότερον FHIJQwΛ 265b30 καὶ E2ΨKbeLNR Sp[1319.22, 25] : ἰδια E1 : ἢ FHIJQwΛ 265b34 κινεῖ EKbeLNR Sp[1320.4]? : κινήσει FHIJQ?wΛ 260a12

And here is the list of places where FHIJQw share a reading against all the other manuscripts where the reading of Λ cannot be determined: 250b12 μηδέν EKbeLNR : μηθὲν FHIJQw 251b22 παρελθόντος EKbeLNR Sp[1156.15, 1166.35]T[212.3]P[823.18] : παρεληλυθότος FHIJQw 254b10 αὑτά EbeLNR : ἑαυτά FHIJQw 255b35 τὰ2 EKbeLNR Sp[1219.23] : τά τε FHIJQw 256b6 οὐ EKbeLNR : μὴ FHIJQw [S] 256b23 αὑτοῦ EKbeLNR : ἑαυτοῦ FHIJQw 256b34 αὐξάνεσθαι EKbeLNR Sc[1228.20] : αὔξεσθαι FHIJQw

VI.4 The γ group (P.S. Hasper) 257a4 257b31 258a17 259b12 259b32 260b23 261a34 262a15 263b19 263b21 264a9 265b10

CXXXV

αὐξάνεσθαι EKbeLNR : αὔξεσθαι FHIJQw αὑτοῦ EKbeLNR : ἑαυτοῦ FHIJQw αὑτοῦ EKbeLNR : ἑαυτοῦ FHIJQw ἡ EΨKbeLNR Sp[1258.19] : om. FHIJQw γε / γ’ E1KbeLNR Sl[1262.14]Pl[893.3] : om. FHIJQw [Ψ] ἂν EKbeLNR Sc[1269.32] : ἐὰν FHIJQw τὸ1 … τὸ2 EKbeLNR : om. Ψ? : τὸ1 … FHIJQw Sc[1274.9] ταὐτὸν E1Kbe : ταὐτὸ E2LNR : τὸ αὐτὸ FHIJQw μὴ EKbeLNR : οὐ FHIJQw [S] τὸ1 EKbeLNR : om. FHIJQw [S] ταὐτὸ EKbeLNR : ταὐτὸν Sl[1301.2] : τὸ αὐτὸ FHIJQw ἅπαντα EKbeLNR : πάντα FHIJQw

Together these two lists are impressive, and sufficient to mark out these manuscripts as a group. The fact that there are hardly any errors which unify this group against other groups suggests that at least many of these characteristic readings were the result of deliberate changes to the text, sometimes in order to correct the text, as one might, for example, surmise for 265a18, or also for 253a31 (where LNR probably have the uncorrected reading), 256a3 (where LNR clearly have the uncorrected reading) and 257a17-18 (where be may have the uncorrected reading), sometimes in order to improve the text or make it easier to read. Further, it is clear that J and Q, on the one hand, and F and H, on the other, form two sub-groups within the γ group. Since we know that both JQ and FH belong to the γ group and this group forms a larger group together with the β group, we do not need to list Trenn- and Bindefehler anymore; it suffices to list the places where JQ or FH go together against the other manuscripts of the β and γ groups. Here is the list for JQ: 253b13 ἀπαντῆσαι EKbeLNRFHJ2w Sp[1196.15-a] : ὑπαντῆσαι Sp[1196.14-AFM] : ἀπαντᾶν J1Q [Λ] 254b4 θεωρητέον beLNRFHIJ2Q2wΛ : θεωρῆσαι EΨK – J1Q1 Sl[1206.8] 255b4 ἢ EKbeLNRFHIJ2Q2wΛ : .. J1Q1 : καὶ Sc[1214.34] 255b5 καὶ ἐν EΨbeLNRHQ2 AγρSγρ[1214.35] : καὶ KFIw : om. Λ : καὶ οὐχ ἁπλῶς ἐν As[593]A[apud S[1214.26-27]] : καὶ οὐκ ἐν JQ1 255b17 διωρισμένον EKbeLNRFHIwΛ Pc[831.17] : διωρισμένων JQ 256a33 αὑτῷ I2J2Q2 : αὑτὸ ΨeLNRFHI1Λ : αὐτὸ Ebw? : om. J1Q1 257a25 ἀεὶ EΨKbeLN[αἰεὶ]RFHIJ2Q2wΛ Sp[1231.37, 1232.3] : om. Sc[1231.36] : δεῖν J1?Q1 258b13 οὐδὲν EKbeLNRFHIw : οὐθὲν JQ [Λ] 259b4 ἐγγίγνεσθαι LNR : ἐγγίνεσθαι beFHIw : γίνεσθαι EΨ?K Sp[1258.1]? : ἐγγένεσθαι JQ [Λ] 259b16 κινεῖται EKbeLNRFHIJ2Q2w Plc[891.13, 18] : κινεῖν τε J1Q1

CXXXVI 259b17 260a25 262a29 262b28 263a8 263a20 263b25 266a20 266a26 266b13 266b22 267b3

Introduction EKbeLNRFHIJ2Q2w Sc[1259.9]Pl[891.14] : τὸ J1Q1 [Λ] ταύτην EΨKbeLNRFHIwΛ Sp[1265.5]A[704]?Pl[896.16] : αὐτὴν JQ κατὰ E2ΨKbeLNRFHIJ2Q2wΛ : καὶ E1 – J1Q1? μὴν EKbeLNRFHIwΛ Sl[1287.17]Pl[844.8] : δὴ JQ τῷ EKebLNRFHIJ2Q1wΛ : τὸ J1Q2 χρόνου EKbeLNRFHIJ2Q2wΛ Sc[1289.38] : χρόνου καὶ J1Q1? ὂν2 cum diagrammate JQ ἀεὶ EΨKbeLNRFHIwΛ : om. JQ ἐν E2ΨKbeLNRFHIwΛ Sp[1324.9] : om. E1JQ1 ΑΒ EKbeLNRFHIwΛ : ΑΓ J1?Q1? [S] ἀπείρῳ EKbeLNRFHIwΛ Sc[1343.18] : ἀπείρῳ μεγέθει JQ ἄτοπον … κινεῖν E2ΨKbeLNRFHIJ2Q2wΛ : bis E1 : om. J1Q1 [S]

Places where FH go against the other manuscripts of the β and γ groups are: 251a11 251a14-15 251b22 251b24

252a28 252b5 252b18 252b22 254b12 254b13 254b31 255b16 255b17 255b21 257a17 257a21 258a10

καὶ EΨ?KbeLNRIJQwΛ T[210.5] : om. FH τὸ κατὰ τόπον EKbeLNRIJQwΛ T[210.8] : κατὰ τόπον τὸ FH [S] ἀεὶ EKbeLNRIJQwΛ Sp[1156.24, 1166.35] : om. FH [T][P] λαβεῖν ἐν τῷ χρόνῳ EKbeLNR Sp[1166.33]T[212.4] : λαμβάνειν ἐν τῷ χρόνῳ IJQw : ἐν τῷ χρόνῳ λαμβάνειν F : ἐν τῷ χρόνῳ λαβεῖν H [Λ: λ.. ἐν τῷ χρόνῳ] [T][P]265 ὃ συνάγει EKbeLNRIJQw : τὸ συνάγον FH [Λ] [S][T] οὖν EKbeLNRIJQΛ Sl[1186.36] : om. w : τοίνυν FH ἐν ἡμῖν ἐνούσης E2KbeLNRIJQwΛ : ἐν ἡμῖν μεούσης E1 : ἐνούσης H : οὔσης F [S] κινεῖ τι αὐτὰ EKLNRIJQwΛ : κινεῖ τι I : τι αὐτὰ F : τι κινεῖ αὐτὰ be – HB 266 αὑτὰ EKbeLNRIJQw : ἑαυτὰ FH αὑτοῦ E2beLNRIJQw : αυτοῦ K : ἑαυτοῦ FH [E1] οὕτω EKLNRIJQw : οὕτως be – FH ἄνω … κάτω EKbeLNRIJQwΛ Sp[1216.31] : κάτω … ἄνω FH πολλαχῶς EKbeLNRIJQwΛ : πλεοναχῶς FH γίγνεται EKbeLNR : γίνεται IJQwΛ : φέρεται FH λέγοι EKbeLNRIJQwΛ : λέγει FH μαθητικόν EKLNRF2IJQwΛ : μαθηματικὸν F1 : μαθητόν be – HB 267 ᾧ EKbeLNRIJQw : οὗ FH [Λ]

_____________ 265

That H is for λαβεῖν here contaminated from EKbeLNR also appears from the fact that other manuscripts which share with H the α version of book VII from 244b5b do have the same reading as F. 266 The reading of H is the result of the insertion of κινεῖ into the text. 267 The reading of F is a correction of the reading featured in H.

VI.4 The γ group (P.S. Hasper)

CXXXVII

om. EKbeJQΛ A[apud S(1244.15-20)]As[641]? : δὲ Ψ[vel δὴ]?LNR : δὴ FH 259a18 δ’ ἡ EΨbeF2IJQwΛ : δὲ ἡ KLNR : δ’ F1HB 259b9 αὑτῶν EKbeLNRJQ : αὐτῶν Iw : ἑαυτῶν FH 261a16 ζώντων EKbeLNRF2IJQΛ Sc[1272.9] : ζῴωντῶν w : ζῴων F1H Pc[900.26] 261a26 αὐτὸ αὑτὸ beLNRIJΛ Sl[1272.27-a] : αὐτὸ αὐτὸ Q1?w? : αὐτὸ ἑαυτὸ EKQ2 Sl[1272.27-A] : αὑτὸ Ψ – H : αὐτὸ F 261b24 ὥστε EKbeLNRIJwQ : ὥστ’ FH 261b25 γίγνοιτο E2beLNRI[γίνοιτο]JQw[γίνοιτο] : γένοιτο E1K – FH [Ψ][Λ] 261b34 κατὰ τόπον EKbeLNRIJQwΛ Sp[1278.25]T[228.5] : om. H : post κάτω F : post ἄνω Ψ? 262a4 ἐστὶν E1Ψ?J : ἐστὶ E2KbeLNRIQwΛ : ἐστι οἷον Sc[1278.38] : ἐστὶν ἢ FH 262a15-16 κύκλῳ … κύκλον EΨ*KbeLNRIJQwΛ Sc[1280.29]Pp[908.5] : κύκλον … κύκλῳ FH 263a6 ἀδύνατον διεξελθεῖν EKbeLNRJQwΛ : ἀδύνατον ἐξελθεῖν I : διεξελθεῖν ἀδύνατον FH 263a17 διεξελθεῖν EKLNRIJQw : διελθεῖν be – FH [Λ] 264b19 αὑτοῦ EbeL2NRIJQ2wΛ : αὐτοῦ L1 : ἑαυτοῦ FH [K][Q1] 264b26-27 μεταβάλλειν μεταβολάς EKbeLNRFIJQwΛ : μεταβολὰς μεταβάλλειν FH 265a23 τέλειον μὲν τοῦ EKbeLNRIJQwΛ : τέλειον οὐ H : μὲν τέλειον τοῦ F 265b23 καὶ EKbeLNRIJQwΛ Sp[1318.31] : om. FH 265b24 οὐδεμίαν αἰτίαν E2beLNR : μηδεμίαν αἰτίαν E1KIJQwΛ : αἰτίαν μηδεμίαν FH [S] 265b30 δὲ EKbeLNRIJQwΛ : δὴ FH [S] 266b11 κινήσει χρόνῳ E2KbeLNR[κινήσει bis]IJQwΛ Sc[1342.34] : χρόνῳ κινήσει F : χρόνῳ E1Ψ – H268 267a4 τι ἄλλο E2KLNRIJQwΛ Sp[1345.30] : τι τε ἄλλο E1 : ἄλλο τι be – FH 267a22 τέ EΨNRIJQ : om. Lw : γε Kbe – FH [S] 258a22

This is sufficient evidence to establish that FH and JQ form two separate branches of the γ group. What about, then, the three remaining manuscripts belonging to the γ group, I, w and Λ? To take w first, the evidence suggests that it forms a sub-group together with JQ. This appears from the following lists of joint readings: 252b27 καὶ2 EKbeLNRFHIΛ : κἂν JQw Pa[367.11] [S] 253a3 κινῆσαν EΨbeLNRFHIΛ? : κινῆσον K – JQw [S]

_____________ 268

Here it is clear that the reading of F is the result of the insertion of κινήσει into the text.

CXXXVIII 253b6 254a22 254b10 258a9 258a23 259a14 259b6 259b26 260b22 262a10 262a12 262b13 263b9 263b15 264a11 267a9

Introduction τῆς EKbeLNRHIQ : om. F : τις Jw [Λ: non τις] ἀεὶ EKbeLNRFHIQ2Λ : om. JQ1w [S] τὰ1 eLNRFHIJ2Λ Pp[829.6]? : om. Eb – J1Qw δὲ2 Ψ?KbeLNRFHIJ2QΛ : om. E – J1w τοιοῦτον EKbeLNRFHIQ : τοιοῦτο Jw δὲ EKbeLNRFHIΛ Sl[1254.29]Pl[888.3] : οὖν JQw δὴ δεῖ LF0HI : τε δεῖ E2 : .. δεῖ E1 : δεῖ b : δεῖ δὴ NR : δὲ δεῖ ΨKeF1J2Q2Λ Pl[890.8] : δὲ δὴ J1?Q1?w αὐτὸ EKbeLNRFHIQ2Λ Sc[1260.17] : αὐτῷ JQ1?w δὲ EKbeLNRFHIJ2QΛ Sc[1269.32] : om. J1w κἂν E1K Sc[1280.6] : καὶ ἐὰν E2beLNRFHI : καὶ ἂν JQw [Λ] οὐχ ἡ E2ΨKbeLNRHIJ2QΛ Sl[1279.18] : οὐχ ἧς E1 : οὐχὶ FJ1 : οὐχὶ ἡ w τῷ Α τὸ Δ KLN2RF2HIΛ : τὸ Α τὸ Δ e2N1 : τὸ Δ EΨbe1Q : τὸ Α F1 : τῷ Α Jw ἐὰν ELNRFHI Sl[1294.3-a] : κἂν K : ἂν JQw – be Sl[1294.3-A] χρόνος EKbeLNRFHIJ2QΛ : χρόνος δ’ J1?w εἰ E2Ψ[cum τὸ Α]KbeLN1RFHIJ2Q2Λ : ἡ E1 : om. N0 – J1w [Q1] ἐγγίνηται E2Ψ?KbNFHΛ : ἐγγένηται E1?eLR[ἐγγένεται a.c.]IJ2Q : ἐν ᾗ γένηται J1w

It is not a very long list, and there are quite a few places on it for which only J and w share a reading. Now the latter point can be explained by appeal to contamination or correction in Q, for which we also have independent evidence, but because of the former point we need to make sure that there are no such connections between w and FH. There are only two places where w shares a deviant reading with them: 251b25 τε EKbeNRFIJQ : om. L – FHw [Ψ][Λ] [S][T][P] 267b7 om. beLNRIJQ : αἱ EΨK – FHw [Λ] [S]

That is not enough to dislodge the evidence that w shares a common ancestor with JQ. There are several ways of explaining these two places, but contamination seems the best possibility, certainly if we take the following place into account: 261a16 ζώντων EKbeLNRF2IJQΛ Sc[1272.9] : ζῴωντῶν w : ζῴων F1H Pc[900.26]

Clearly, the reading of w is a combination of the two alternatives, and thus probably the result of a supralinear ζῴων above ζώντων, given that w shares an immediate ancestor with JQ and only further away with FH. The stemmatic position of I is already more difficult to determine, for it shares both readings with FH and with JQw. Here are the readings shared by I and at least one of JQw:

VI.4 The γ group (P.S. Hasper)

CXXXIX

251b24 λαβεῖν ἐν τῷ χρόνῳ EKbeLNR Sp[1166.33]T[212.4] : ἐν τῷ χρόνῳ λαμβάνειν F : ἐν τῷ χρόνῳ λαβεῖν H : λαμβάνειν ἐν τῷ χρόνῳ IJQw [Λ: λ.. ἐν τῷ χρόνῳ] [T][P] 252a22 δεῖ EKbeLNRFHI2QwΛ : δὴ I1J [T] 252b30 τὰ ἀντικείμενα EKb2eLNRFHQ : τὴν ἀντικείμενα b1 : τ’ἀντικείμενα IJw 253b18 τοσοίδε τοσόνδε KLNRFH : τοσοίδε τοσονδὶ be : τόσοιδ. τοσόνδε E1 : τόσοι δὴ τοσόνδε E2 : τοσοιδὶ τοσόνδε IJQw [Λ] [Sp 1198.14: τοσοίδε] 254b14 αὑτοῦ EKbeLNRFH : ἑαυτοῦ IJQw 255a9 ὥστ’ EKbeLNRFH : ὥστε IJQw 255a12 ἑαυτὸ EKbeLNRFHJQ : αὑτὸ Iw 255a30 αὑτῶν EFHJQ : ἑαυτῶν KbeLNR Sc[1212.30] : αὐτῶν Iw 256a4 τὸ EKbeLNRFHQ1? : om. IJQ2w [Λ] 257b31 αὑτὸ KeLNRFHJwΛ : αὐτὸ EΨb : ἑαυτὸ IQ 259a25 ἀεὶ1 EKLNRFHJ2QwΛ Sc[1257.18] : αἰεὶ be : om. IJ1 259b9 αὑτῶν EKbeLNRJQ : ἑαυτῶν FH : αὐτῶν Iw 260a12 ἢ κινεῖται ἢ ἠρεμεῖ EΨKbeLNR Sc[1263.37-1264.1] : ἠρεμεῖ ἢ κινεῖται FHJ2Q : ἢ ἠρεμεῖ ἢ κινεῖται IJ[ἠρεμεῖν]wΛ 261b9 τοῦτο EKbeLNRFHI2JQ2Λ : τούτῳ I1Q1w 262a31 τῷ2 beLNRFHJQ2wΛ : τὸ IQ1? [EK] 263a7 λόγον τινὲς EΨKbeLNRFHΛ : τινες pro τὸν As[746] : τινες λόγον IJQw 263a12 αὑτῷ EbLNRFHw[Λ] : αὐτῷ ΨKe : ἑαυτῷ IJQ 263b21 οὐ E2ΨKbeLNRFHΛ : καὶ οὐ E1 : om. IJQw 264a1 ἐγίγνετο EbeLNR : ἐγίνετο KFHw : γέγονε Sp[1297.28-38] : ἐγένετο IJQ Ap[apud S(1299.8)] [Λ] 264a3 γέγονε EKbeLNRFHw : γέγονεν IJQ 266b22 τῇ EΨKbeLNRFQ2wΛ Sc[1343.17] : τὴν IJQ1

And here are the readings shared by I and at least one of FH, against the other manuscripts of the γ group: 253a22 253a35 253b9 253b34 254a28 254a35 255a3 257b15 259b6

ἥπερ KbeLNRJ2QwΛ Sl[1193.1-AF] : εἴπερ E : ἢπερ J1 : ἔσται ἥπερ Ψ – FHI Sl[1193.1-a?] τὸν EKbeLNRFH2JQwΛ : τὸ H1I [S] ὁμοίως EΨLNRHBJQwΛ : ὅμως Kbe – FI [S][T] κινοῦνται EKbeLNRJQwΛ : κινεῖται FI : κινοῖται H [S] δοκεῖ E1KJQw : δοκῇ E2beLNRH : δοκεῖ εἶναι FI [Λ sine εἶναι] [S] ἱκανὴ μία EKbeLNRJQwΛ : μία ἱκανὴ FHI [S] μὲν EKbeLNRFJQwΛ Sp[1209.17] : om. HI οὐδὲν EKbeLNRHJQw : οὐθὲν FI τε δεῖ E2 : .. δεῖ E1 : δεῖ b : δεῖ δὴ NR : δὲ δὴ J1?Q1?w : δὲ δεῖ ΨKe F1J2Q2Λ Pl[890.8] : δὴ δεῖ L – F0HI

CXL

Introduction

259b12 260b25 260b26 261a25 261b16 261b28 261b35

μὲν EKLNRJQwΛ : om. be – FHI [Ψ] ἄλλην EKbeLNRF2JQΛ Sp[1269.34]Pp[897.23] : ἄλλο w : om. F1HI αὔξεσθαι ELNRJQw Pl[899.14] : αὔξάνεσθαι Kbe – FHI [Ψ][Λ] καὶ EKbeLNRIJQwΛ Pp[901.8, 10] : καὶ τῶν FHI [Ψ] ἡ EΨKbeLNRFJ2QΛ : ?? J1 : εἰ HI μὲν EΨ?KLNRJQΛ : om. be – FHI T[227.31] [S] ἀριστερὰ … δεξιὰ EΨKb[τὰ … τὰ]e[τὰ … τὰ]LNRΛ T[228.6] : δεξιὰ … ἀριστερὰ J[τὰ ἀριστερὰ]Qw : τὰ δεξιὰ … τὰ ἀριστερὰ FHI Sp[1278.2526]Pp[842.8] 262b17 ὑστεριεῖ EK2beLNRF2HJQwΛ Sc[1285.34] : ὑστερεῖ K1 – F1I 262b19 τοῦ EKbeLNRJwΛ : τὸ Q : τῷ FHI 262b30 οὐ EΨKbeLNR2JQw : οὐδὲ F : οὔτε R1? – HI 263a19-20 ἄπειρα διεξελθεῖν EKbeLNRJQwΛ Sc[1289.38] : διεξελθεῖν ἄπειρα FHI 263a26 ἔσται EKbeLNRJQwΛ : ἔστι FHI 263b26 γίγνεσθαι EKbeLNRJQ : γίνεσθαι FHI 265a6 φασιν ἀεὶ Eb[αἰεὶ]e[αἰεὶ]LNRJQΛ : ἀεὶ φασὶ K – FHI 265b9 περιφορά EΨKbeLNRJQw Sp[1316.26]Aγρ[apud Sc(1317.6-7) ἔν τισιν] : φορά Λ A[apud S(1317.4)] : περιφορά ἐστι FHI 266a29 τὸ EΨKLNRHJQw A[apud Sc(1325.20)] : om. be – FI [Λ] 266b10 ᾧ EeLNRJQw : οὗ Kb – FHI [Ψ][Λ]

It may seem that the list of readings shared with the FH group is longer, and thus that I is closer to FH than to JQ. However, if one looks at the kind of readings shared with either group, it appears that only at 253b34, 261b16, 262b30 and 266a29 the reading shared with F and/or H is clearly incorrect, and that of these readings only 253b34 is shared with both of FH, whereas the readings shared with JQ are more often than not clearly incorrect: of those which are not orthographical variants only 251b24, 253b18, 260a12, 263a7 and perhaps 264a1 could be accepted. Moreover, quite a few of the readings shared with FH are also found outside the γ group, whereas that does not apply for the readings shared with JQw. Therefore I is probably closer to JQ than to FH. The readings shared with FH would then be contaminations in I or occasionally constitute places where JQw have been (for some) corrected. Finally, the stemmatic relations of the exemplar of the translatio vetus within the γ group are most difficult to determine. On the one hand, there are seven places where Λ shares unique readings with JQw269: _____________ 269

This is the kind of evidence which prompts Bossier and Brams, Physica. Translatio Vetus, p. XCIII, to claim that the Greek exemplar underlying the Translatio vetus is to be situated in the γ group and is especially closely related to J.

VI.4 The γ group (P.S. Hasper)

CXLI

250b13 ἀλλ’ ἀεὶ EΨKeLFI Sp[1118.23-24]T[209.4] : ἀλλὰ ἀεὶ NRH : ἀλλὰ εἰ bw : ἀλλ’ εἰ JΛ : ἀλλ’ Q 256b11 συμβήσεται EKbeLNRFHIJ2 : συμβαίνει J1QwΛ [S] 257b33 ἂν ταῦτα κινοῖ EKbeLNR : κινοῖ ταῦτα FH[κινεῖ]IJQ : ἂν κινοῖ ταῦτα wΛ [Ψ] 259a6 ἀΐδιος EKbeLNRFHIJ2Q Sc[1254.3] : ἀΐδιον J1wΛ 259a14 τοῦδε EΨ?KbeLNRFHIJ2 Sl[1254.29]Pl[888.3] : τούτων J1QwΛ 260a12 ἢ κινεῖται ἢ ἠρεμεῖ EΨKbeLNR Sc[1263.37-1264.1] : ἠρεμεῖ ἢ κινεῖται FH J2Q : ἢ ἠρεμεῖ ἢ κινεῖται IJ[ἠρεμεῖν]wΛ 262b20 ἔσται EKbeLNRFHIJ2Qw : ἐστιν J1Λ 264b4 ἐκ EΨKI2Q2 Sp[1306.1] : μὴ ἐκ beLNRFHI1? : ἐκ μὴ JwΛ : ἐκ τοῦ Q1

whereas Λ shares only two readings with FH: 254b11 om. ELNRIJQw : τῷ2 KbeFH – Λ 262b22 τῆς EΨbeLNRFJQw Sp[1286.17-18]Pp[844.2, R-103r] : τοῦ KHIΛ Sl[1286.5]

Since these two places may perhaps be discounted as being a matter of interpretation in translating, it seems that Λ is more closely related to JQw, especially because among the readings shared with JQw there are at least two clear errors, at 250b13 and 264b4. However, that cannot be the whole story, as there are many places at which Λ goes against the rest of the γ group and sides with E(Ψ)LNR: 251b21 252a10 252b26 253b16 254a25 255a9 257a17 258a17 259a19 259b25 260b25 261b23 261b34 261b35

καὶ1 EΨKbeLNR – Λ : om. FHIJQw ὥσπερ EΨKbeLNR – Λ : ὥσπερ καὶ FHIJQw γίγνεται vel γίνεται EΨbKbeLNR – Λ : γίνεσθαι FHIJQw ἐξέωσεν EΨbeLNR – Λ : ἔωσεν FHIJQw – K φασί τινες ELNR – Λ : τινες φασὶν FHIJQw – Kbe εἰ EKbeLNR – Λ : ἐπεὶ FHJQw – Ψ ὑγιάζον EΨ[vel ὑγιάζειν]KbLNR – Λ: ὑγίαζον καὶ FHIJQ[τὸ ὑγίαζον]w – e κινεῖσθαι ὑπ’ ἄλλου EKbeLNR – Λ : κινεῖσθαι ὑπ’ ἄλλου κινεῖσθαι J : ὑπ’ ἄλλου κινεῖσθαι HIFQw ἄλλο EΨKbeLNR : τι ἄλλο HIJQwΛ : τι καὶ ἄλλο F ἔσεσθαι ἐν τοῖς οὖσιν EKbeLNR – Λ : ἐν τοῖς οὖσιν ἔσεσθαι FHIJQw ἀλλ’ ἢ φοράν EΨKbeLNR – Λ : om. F1 : ἀλλὰ φοράν F2HIJQw εὐθὺς ἀνάγκη EKbeLNR – Λ : ἀνάγκη εὐθὺς FHIJQw [Ψ] ἡ ἄνω τῇ EΨKbeLNR – Λ : τῇ ἄνω ἡ FHIJQw ἀριστερὰ … δεξιὰ EΨKbe[τὰ … τὰ]LNR – Λ : τὰ δεξιὰ … τὰ ἀριστερὰ FHIJ[τὰ1 om.]Q[τὰ1 et τὰ2 om.]w

CXLII

Introduction

262a5-6

τόπος ἢ πάθος ἢ εἶδος ἢ μέγεθος EΨKbeLNR – Λ : πάθὸς ἢ εἶδος ἢ τόπος ἢ μέγεθος FHIJQw 262a9-10 Β EΨK2e2 – FHIJw Sp[1280.3] : Γ K1be1LNR – Λ [Q] || Γ E1K2e2 FHIJQw Sp[1280.3]Pa[403.4, 9] : Β {E0}K1be1LNRΛ : A Ψ270 262a30 ἐστι(ν) διαίρεσις EKbeLNR – Λ : διαίρεσις ἐστιν FHIJQw [Ψ] 262b5-6 τελευτῇ καὶ ἀρχῇ EΨKbeLNR – Λ : ἀρχῇ καὶ τελευτῇ F[ἀρχὴ καὶ τελευτὴ]HIJQw 262b24 ἐφ’ οὗ EΨK[ἐφ’ οὗ τὸ]beLNR – Λ : τῷ FHIJQw 262b25 καὶ ἀρχῇ κέχρηται EΨK[ἁρχὴ]beLNR – Λ : κέχρηται καὶ ἀρχῇ FHIJQw 262b26 τῷ EKLNR – Λ : τὸ be : τοῦ FHIJQw [Ψ] 262b31 τὰ EΨKLNR – Λ : τὸ FHIJQw – be 263a14-15 ὑπάρχει καὶ ἐν τῷ χρόνῳ EΨ?K[ὑπάρχειν]beLNR – Λ : καὶ ἐν τῷ χρόνῳ ὑπάρχει FIJQw : καὶ ἐν τῷ χρόνῳ H 263b1-2 ἑτέρου τελευτὴ ἡμίσεος τοῦ δ’ ἑτέρου ἀρχὴ E[ἡμίσεο]ΨKbeLNR – Λ : ἑτέρου ἡμίσεος ἀρχὴ τοῦ δ’ ἑτέρου τελευτὴ HIJQw : ἡμίσεος τοῦ ἑτέρου ἀρχὴ τοῦ δ’ ἑτέρου τελευτὴ F 263b28 Α EKbeLNR – Λ : Α χρόνῳ F[πρώτῳ]HI[χρόνῳ eras.]JQw – Ψ 264a1 χρόνον EKbeLNR – Λ : χρόνος ἦν FHIJQw – Ψ 264a14 om. EΨKbeLNR – Λ : ἐπὶ τὸ Γ FHI[Γ?]JQ[τ.]w 264a15 ὅταν … ἔλθῃ EΨKbeLNR – Λ : om. FHIJQw 264b28 δὲ vel δ’ EΨKLNR – Λ : om. be : δὲ καὶ FHBIJQw 264b30 ταύταις συμβαίνει κινεῖσθαι E1[συμβαιν.ν]Ψ : τὰ αὐτὰ συμβαίνει κινεῖσθαι E2K[ταὐτὰ]beLNR – Λ : συμβαίνει ταὐτὰ κινεῖσθαι FHI[κινεῖσθαι συμβαίνει ταὐτὰ]JQw

The only possible scenario which could explain these facts is that Λ is a manuscript which draws on at least two sources (or, alternatively, the translator used at least two different manuscripts), one close to JQw and at least one _____________ 270

The most likely scenario to explain this distribution of readings is that the α and β groups originally both read at 262a9-10 οἷον ἡ ἀπὸ τοῦ Α ἐπὶ τὸ Γ τῇ ἀπὸ τοῦ Α ἐπὶ τὸ Β, but that the α group then was contaminated from the γ group for the first slot, reading B instead of Γ, after which the Β in the second slot was adjusted in Ψ, reading A instead of B (cf. the adjustment of A to B in Ψ immediately after: 262a10 A EKbeLNR FHIJQwΛ : B Ψ), whereas the whole phrase τῇ ἀπὸ τοῦ Α ἐπὶ τὸ Β dropped out in E because of saut-du-même on the basis of ἀπὸ τοῦ Α ἐπὶ τὸ Β (262a9-10 τῇ – τὸ Γ E1Ψ KbeLNRFHIJQwΛ Sp[1280.3] : om. E0). In cases for which we may surmise that the predecessor of a manuscript featured a certain reading which is necessary to explain the dropping out of a phrase by saut-du-même-au-même, I ascribe this reading to that manuscript with the help of braces { }, in this case {E0}.

VI.4 The γ group (P.S. Hasper)

CXLIII

related to the α group. There must at least be for Λ a source related to the α group, for two reasons. First, there are very few readings in Λ which are for the rest (almost) unique to the β group: 251b30 συνέβαινε(ν) EKbeLNFHIJQw T[212.11] : συμβαίνει RΛ [S] 253b21 ἀλλ’ ἅμα E1ΨKFIJQw Sc[1198.17] : ἀλλ’ ἅμα ὅλον H : ἀλλ’ ὅλον ἅμα E2beLNR – Λ 254a19 τῶν ὄντων EKbe – FHIJQw Sp[1203.13] : om. LNR – Λ 261b1 ὂν EKbeLNFHIJQw Pl[903.6]Pp[903.12]Pc[841.12] : ὃν RΛ 264a14 δ’ E1Ψ?be – FHIJQw : δὴ E2KLNR – Λ [S]

Though 251b30 and 261b1 are probably accidental, the other three do indicate that there is some slight contamination from the β group in Λ, but not more than that. Second, there is stronger evidence for contamination in Λ from the α group (I leave out those places which might be a matter of translation): 251b32 252a7 253b20 258a12 258a17 258b4 259b22 260b11 261a10 261b2 262a31 263a1 263a3

263b15 266a16 266a21

παύσεται e2LNRFHIJQw Sc[1169.26] : παύεται EKbe1 – Λ [Ψ] ὅπερ E2KbeLNRFHIJQw : ὥσπερ E1Ψ – Λ μὲν οὖν EKbeLNRFHIJQw : γὰρ Ψ – Λ [S] ἅπαν KbLNRFHIJQw : ἅπαν τὸ EeΛ Sp[1242.38]Pa[385.22]? [Ψ] ἄρα Ke2LNRFHIJQw Sp[1243.6]Pa[386.3] : γὰρ EΨbe1?Λ πρώτως2 e2LNRFIJQw : πρῶτον EΨ?Kbe1 – Λ? κινεῖν E2KbeLNRFHIJQw Sc[1259.36]Pp[891.24, Ra-34.2]Pa[393.6] : om. E1Ψ : κινεῖσθαι Λ271 διάκρισις E2ΨKbeLNRFHIJQw As[709] : διάκρισεις E1? – Λ προτέρα Ψ?KbeLNRFHIJQw : προτέρον E – Λ οὖν E2KbeLNR1FHIJQw : om. R0 : δ’ E1Ψ – Λ ἐν … τῷ ΑΒΓ LNRFHIJQw : om. EΨKbe – Λ ἄρα LNRFHIJQ : om. EΨKbe – Λ εἶναι ἀΐδιον ἐπὶ τῆς εὔθείας E3Ψ?KbeLNR Sl[1287.18-a]Sp[1288.29]? : ἀΐδιον εἶναι ἐπὶ τῆς εὐθείας FIJQw Sl[1287.18-A] : ἁΐδιον ἐπὶ τῆς εὐθείας εἶναι H : εἶναι ἐπὶ τῆς εὐθείας ἀΐδιον E2E1[ἴδιον] – Λ ΑΓΒ E2ΨKbeLNFHIJQ : ΑΒΓ E1?R – Λ κινοῦν beLNRFHIJQw Pal[422.8] : τὸ κινοῦν EK – Λ [S] Β ὅλην LNRFHIJQ : ὅλην Β EKbe – Λ [Ψ][S]

_____________ 271

I would venture the hypothesis that the reading of Λ is meant to correct the reading we find in the α group.

CXLIV

Introduction

The shared readings at 260b11 and 261a10, and even 263b15 may be coincidental, but the other places must involve contamination. And these contaminations also explain all the readings Λ shares with the α and β groups together272. However, one has to go further, for among the readings shared with the α group against the β and γ groups there are also errors: 258a17, 259b22, 260b11, 261a10 and 261b2. Their number seems to exclude the possibility that Λ was a manuscript belonging to the γ group which was ‘corrected’ on the basis of a manuscript related to the α group, for then one would not expect errors in such amounts; rather, Λ seems to be a genuinely mixed manuscript, sharing in errors belonging to the γ group as well as to the α group. There is even dramatic evidence for this ‘merger’: 257b32 γὰρ τοῦτο μὲν KLNRFHIJQw Sp[1240.32] : μὲν γὰρ τοῦτο μὲν Λ : μὲν γὰρ τοῦτο Ebe [Ψ]

Thus the stemma for the γ group looks as follows:

γ α

Λ

J

Q w I

F

H

For the most part this stemma accords with the findings of Rashed for De generatione et corruptione, where we find the same stemmatic pattern for the group constituted by FH and JQ273. Apart from Λ and I, which we do not have _____________ 272

The contaminated nature of Λ also appears from: 265b9 περιφορά EΨKbeLNRJQw Sp[1316.26]A[apud Sc(1317.6-7) ἔν τισιν] : περιφορά ἐστι FHI : φορά Λ A[apud Sc(1317.4)]. Thus Λ is the only source among those collated which has picked up the reading reported by Alexander to be the dominant reading of his days. 273 See Rashed, Überlieferungsgeschichte, Stemma Codicum, p. 383.

VI.4 The γ group (P.S. Hasper)

CXLV

for De generatione et corruptione, the only real difference is the position of w in the stemma: Rashed posits J, Q and w as stemming from a common ancestor immediately, whereas in Physics VIII w seems one more knot removed from JQ. On the basis of this stemma, it will also be possible to determine the amount of contamination in each of the sub-groups and manuscripts. For present purposes it will go too far to present the evidence, but it seems that there is a considerable amount of contamination in individual manuscripts, especially in F, H, and to a lesser extent Q, relatively little at the level of sub-groups, and some clear cases of contamination at the level of the γ group as a whole.

VI.5 The δ group: Par. gr. 1859 and Vind. Phil. gr 64 These two manuscripts, with the sigla b and e, are the only two surviving independent manuscripts containing the extended version of book VII, thus betraying privileged access to the late ancient tradition. It is certainly true that be constitute a group which is really different from all other manuscripts. This appears from the following list of uniquely274 shared errors: τάδ’ ἀλλάσσοντα E1[ἀλλάσοντα]E2KL[ἀλάσσοντα]NRFHIJQw Sc[34.2, 158.11, 160.20, 1125.1]Pp[823.10] : διαλλάσσονται Λ : τάδ’ ἀλλάσσονται be 251a4-5 τὸ γὰρ ᾗ δὲ τάδ’ ἀλλάσσοντα ἐνθένδε ἐκεῖσε λέγειν αὐτὸν ὑποληπτέον LNRFHIJQwΛ : δεῖ γὰρ ὑπολαβεῖν λέγειν αὐτὸν ᾗ δὲ ταδ’ ἐνθένδε τὰ ἀλλάσσοντα EΨ[τὰ om.?]K[ἀλάσσοντα] Slp[1123.8, 1125.5 – τὰ om.] : δεῖ γὰρ ὑπολαβεῖν λέγειν αὐτὸν τῇ δὲ ταδ’ ἄλλασονται ἐνθένδε ἐκεῖσε be [T] 251b1 ὅσα γε EKLNRFHIJQw : γε ὅσα be [Λ] [S][T] 251b21 ἅμα KLNRFHIJQwΛ T[212.2-3] : ἀλλ’ E1Ψ : ἅμα ἀλλ’ E2be [S][P] 252a1 φθαρῆναι EKLNRFHIJQw Sc[1170.27] : φθείρεσθαι be [Λ] [T] 253b19 διαιρεῖται EKLNRFHJQwΛ : διαιρεῖ διαιρεῖται be [S] 253b29 ἐστὶ E2KLNRFHJQw Sp[1199.32] : ἐστὶν E1 : ἐν Λ : ἐστὶν ἐν be 254a8 ὅτι EKLNRFHJQwΛ : ἔτι be [S] 254b8 καὶ EKLNRFHJQwΛ Sp[1207.4] : τὰ δὲ be 254b24 τῆς κινήσεως EKLNRFHJQwΛ : τῶν κινήσεων be 255b11 κωλύῃ E2KLNRFHJQw : κωλύει E1 : κωλύσῃ be [Λ] 255b14 ποτε EKLNRFHJQwΛ : ποτε οὐ be 255b15 τὸ EKLNRFHJQwΛ : τῷ be 257a17-18 πᾶν .. εἶναι EΨKLNR Sp[1231.10-11] : πᾶν … ἐστὶ be : ὅτι .. ἔσται FHJQ wΛ 251a2

_____________ 274

Some of the readings of be were already known to Ross, namely through one of the second hands in E. I will also add these readings as unique to be, since it is clear that the tradition of the δ group is the source of correction here.

CXLVI 257b18 258a7 260a22 261a1 261a13 261b10 261b32 262a25 262a30 264b9 264b15 264b28 264b31 265a13 265a34 266a3 266a12 267a2

Introduction τὸ2 Ψ?LNRHJQw : τῶ. E1 : τῷ E2be : om. KF [Λ: non τῷ] [S] ἐκ EKb2e1LNRFHJQwΛ : om. b1e2 A[apud S(1242.26)]? εἰ EΨKLNRFHJQwΛ : om. be ἀλλ’ ἕτερον EΨKLNRFHJQwΛ Pl[900.1] : om. b1 : ἀλλ’ ἕτερον ἐστιν b2e ἀρχὴν EKLNRFHJQwΛ Plc[900.16, 19, 840.30] : ἀρχῇ be [S] μὴ ἀνάγκη EΨKLNRFHJQwΛ Sc[1275.36] : ἀνάγκη be ἀνακάμπτει EΨKLNRFHIJKQwΛ Sp[1278.18] : ἀνακαλύπτει be ἀρχὴ EΨKLNRFHIJQwΛ Sp[1281.23] : ἀρχὴ γὰρ be τὸ EΨKLNRFHIJQwΛ Sc[1282.33] : τῷ be ἡ EΨKLNRFHIJQwΛ Sl[1307.13] : εἰ be τὰ KLNR Sp[1307.33]A[apud S(1308.16)] : τὸ EΨ[vel ἡ] : om. be : ἡ FHJQw : εἰ I [Λ] δὲ EΨLNΛ Sl[1311.13-A] : δ’ KR : δὲ καὶ FHBIJQw Sl[1311.13-a] : om. be ἐν1 EΨ[vel ἐν τῇ]KLNRFHIJQwΛ : ἔν τε be κυκλοφορία EKLNRFHIJQwΛ Sl[1313.20]p[1311.24] : κυκλωφορὰ be τε EΨKLNRI T[232.32-33] : τε τινα FHJQwΛ : om. Sp[1315.35]? : τινὰ be αὐξάνηται E2ΨKLNRFHIJQwΛ : αὐξάνεται be : αὐξάνη… E1 [S] ὅτι EΨKLNRFHIJQwΛ : om. be [S] ὃ EΨKLNRFHIJQwΛ Sc[1345.14] : om. be

In addition there are many places where the δ group provides readings which are not found in the other manuscripts. What is more, some of these find support in one or several of the commentators, thus suggesting that the δ group is an independent part of the direct manuscript tradition: ἔχει E1KLNRFHIJQwΛ Sp[1184.32] : ἔχειν δεῖ E2be T[212.31], cf. Sl[1183.20] ἔχει E1KLNRFHIJQwΛ : ἔχειν E2be Sl[1183.20] [T] τῆς κινήσεως EKLNRFHIJQwΛ : τῶν κινήσεων be Sc[1208.18] ἅπαν EKLNRFHIJQw : πᾶν be Sl[1206.33] ἔχειν EKLNRFHIJQwΛ Sc[1253.5] : ἔχον be Sc[1252.33] οὐκ ἀεὶ E1ΨKe2LNRFHJQwΛ : οὐ E2be1 Sp[1258.30]T[224.15] κινήσει EKe2LNRFHIJQwΛ Sl[1263.31] : κινεῖ be1 Sp[1264.7] λέγεται δὲ EKLNRFHJQwΛ : λέγοιτο δὲ ἂν Ψ?, cf. A[apud S(1268.3)] : λέγεται γὰρ be Sc[1267.36] [T] 260b30 μόνον ἐνδέχεται EΨKLNR Sc[603.10]Plc[899.19. 20, 840.17] : ἐνδέχεται μόνον FH[μόνην]IJQwΛ : μόνην ἐνδέχεται be T[226.7], cf. P[899.20, 840.17] 262a12 τὸ πλάγιον EΨKLNRFHIJQwΛ Sl[1279.18-a]Pl[842.19] : τὰ πλάγια be Sl[1279.18-A]Sp[1280.13]Pp[842.20]T[228.14]

252a17 252a19 254b24 254b32 258b30 259b14 260a19 260b17

VI.5 The δ group (P.S. Hasper)

CXLVII

264a21 οὐ μία EKLNR Sl[1301.31] : οὐκ ἕστι μία FHIJ[ἔστιν]QwΛ : οὐ μία ἔστιν be T[230.28] [Ψ] [S] 265a21 συνθετὴ EKLNRFHIJQw : σύνθετος be Sp[1314.13] [Ψ][Λ] [T]

To this list should be added seven further places where the δ group together with K offers a reading found only in the commentators: δὲ ELNRFHIJQwΛ : δὲ καὶ Kbe Sp[1208.17] [T] τι μὴ ELNRFHIJQwΛ : μὴ τι Kbe Sc[1214.16, 25]Pl[830.31]T[219.5] δὲ EΨLNRFHIJQwΛ : τε Kbe Sp[1244.34] δὲ κινουμένου ἤδη FHIJQw Pl[894.5] : δὴ κινουμένου ἤδη LNR : κινουμένου ἤδη Λ : δὲ κινουμένου ἢ E : om. Ψe : ἢ κινουμένου ἤδη Kb Sc[1263.16-17] 261a16 τοῦ ὀργάνου E1E3Ψe2LNRFHIJQwΛ : om. E2Kbe1 Pc[900.26] Sp[1271.30]? T[226.22]? || 262a6 om. EΨ?LNRHIJQw Sl[1279.17-a]T[228.12] : ἡ KbeF Sl[1279.17-A] [Λ] 265a26-27 στάσιν … κίνησις E1ΨK2e2LNRFHIJ[partim]QwΛ, cf. Sp[1314.36] : om. E2K1be1 Sl[1313.21-1315.5]T? 254b23 255b4 258a24 260a6-7

For some of these places the evidence is rather tenuous, because it could concern coincidental deviations or because the commentator might be just paraphrasing. Thus 265a21 may concern a coincidental adherence to more normal usage on the part of both be and Simplicius. Also 258a24, 260b30 and 262a6 are by themselves not really significant. On the other hand, 252a17 and 19, 255b4, 260b17, 262a12 and 265a26-27 are striking, and 261a16 seems significant as well275. Should we therefore take seriously also other interesting readings only found in the δ group, whether or not in conjunction with K276? _____________ 275

Ross brackets τοῦ ὀργάνου as well. Apart from some further places to be mentioned below it would concern the following places: 250b27 ἠρεμεῖν E1ΨKLNRFHIJQwΛ : ἠρεμεῖν καὶ E2be [S][T] || 251b9 εἰ EΨ KLNRFHJwQ1Λ : καὶ εἰ IQ2 : ἢ εἰ be [S][T] || 251b25 ἐπεί ELNRFHIJQwΛ : ἐπείπερ Kbe [S][T][P] || 252a27 προσοριεῖται EΨbe2LNRFHIJQwΛ : προσδιοριεῖται Ke1 [S][T] || 252a34 ἐφ’ Ee2LNRFHIJwQ : εἰς Kbe1 [T] || 252b24 γί(γ)νοιτο EKLNRFHIJQwΛ : γένοιτο be [S] || 252b32 αὐτὴν EΨKLNRFHIJQwΛ : αὐτὴν τῷ ἀριθμῷ be [S] || 253a2 ὕστερον EKLNRFHIJQwΛ Sl[1188.18] : ὑστέρων be || 253a11 ψεῦδος EKLNR FHIJQw : ψευδὲς be [Λ] [S] || 253a28 καὶ EKLNRFHIJQw Sc[1193.19] : καὶ τὸ be [Λ] || 253b3 εἰσιν E1[vel ἔστιν]J : εἰσὶ LNRFHQwΛ : ἔτι E2K : ἐστι be [Ψ] [S] || 253b32 ἢ EKLNRFHJQwΛ : ἢ καὶ be [S] || 254a10 φύσιν EKLNRFHJQwΛ Sc[1201.9] : φύσιν τὸ be Sp[1201.15] || 254a23 μὲν τοίνυν EKLNRFJQw Sl[1203.18] : μὲν οὖν τοίνυν H : μὲν οὖν be [Λ] || 254b9 τε EKLNRFHJQwΛ : om. be [S] || 254b22 καὶ τὸ EKLNR FHJQwΛ : τὸ δὲ be || 254b23 δὲ ELNRFHJQwΛ : δὲ καὶ Kbe [S][T] || 255a1 παράσχοι 276

CXLVIII

Introduction

To be in a position to answer this question, we need to determine what the place of the δ group is within the stemma. Now on the one hand it is clear that be share many readings with the γ group: 251a9 251b1 251b11 251b32 251b33 252a2 253b6-7 253b24 255a26 255a27 258b14 259a16 259a26

ἐνέργειαν EΨKLNR Sc[923.18]Sp[1126.12]As[544] : ἐντελεχείαν beFHIJ QwΛ T[210.3] ἢ EΨKLNR : καὶ beFHIJQwΛ [S][T] ἢ KLNR : ᾗ EI1 : ἢ ὁ beFHI2JQw [Ψ][Λ] [S] καίομενον EKLNR – F Sp[1170.10] : κάομενον beHIJQw καίομενον EKLNRF Sp[1170.10] : κάομενον beHIJQw φθείρῃ E1ΨKLN : φθαρῃ E2R : φθείρηται beFHIJQwΛ [S][T] κινεῖσθαι πάντα EKLNR Sl[1195.26] : πάντα κινεῖσθαι beFHIJQwΛ Pp[825.19-20. 23. 25, 826.2] ἄπειρα EKLNRJ2 Sc[1199.11] : ἄπειρον beFHIJ1QwΛ Sc[107.10] [Ψ] αὑτῷ : αὐτῷ EΨKLNRF : ἑαυτῷ beHIJQwΛ Sc[1212.2]Pp[830.10] ποιὸν καὶ ποσόν E2ΨKLNRHQ Sp[1202.3] : πο.ὸν καὶ ποσόν E1 : ποσὸν καὶ ποιὸν beFIJwΛ πάσης EΨKLNR Sc[1251.17]Pp[837.35] : πάσης τῆς beFHIJQw [Λ] ἀνάγκη EΨKLNRQ Sc[1255.1, 1256.12]Pp[838.29] : ἀνάγκη καὶ beFIJw : καὶ HΛ κινεῖσθαι καὶ EΨKLNR : ὁτὲ μὲν κινεῖσθαι ὁτὲ δὲ beFHIJQ[καὶ ὁτὲ2]wΛ

_____________ ἂν ELNRFHJQwΛ : ἂν παράσχοι Kbe [S][T] || 255b28 βάλλοντος EKLNRFHJQw : βάλοντος be [Λ] || 255b29 τοίνυν EKLNRFHJQw : οὖν be [Λ] || 255b32 καὶ3 EΨ LNRFHJQw Sp[1219.15] : καὶ τὰ Kbe [Λ] [T] || 256a19 τὸ EKLNRFHJQwΛ As[601] : om. be || 256b23 εἴπωμεν EΨKLNRFHJQwΛ : εἴπω be [S] || 257b8 ἔστιν … ἀτελής EΨKLNRFHJQwΛ Sp[1235.13-17] : om. be || 258a5 Α ΨKLNRFHJQwΛ Sl[1241.29] : ΑΒ E : Α πρώτως be || 258a27 μόνον EKLNRFHJQwΛ Sc[1245.2-3]A? : om. be || 258b30 ἔχειν EKLNRFHJQwΛ Sc[1253.5] : ἔχον be || 259a4 ἕκαστον EKLNRFHJQwΛ Sp[1253.23]Pc[838.22] : ἕκαστα ὂν be || 260a30 αὐξάνεται EKLNRFHJQwΛ : αὔξεται be || 260b13 ἀνάγκη E1ΨKLNRFHJQwΛ T[225.30] : ἀνάγκη καὶ E2be || 260b17 καὶ EKLNRFHJQw : κἂν be || 261a3-4 εἶναι πρώτη EKLNRFHJQwΛ Sc[1270.27] : πρώτη εἶναι be || 261a33 καὶ EKLNRFHJQw : καὶ αἱ be [Λ] [S] || 261a34 ὅροι EΨ KLNRFHJQwΛ Sc[1274.9] : om. be || 261b2-3 ἠρεμήσει EΨKLNRFHJQΛ : ἠρεμήσειε w : ἠρεμήσει πρότερον be [S] || 262b8 τοῦ E1Ψ?KLNRFHIJQwΛ Sl[1280.17] : τὸ E2be || 262b16 τῷ EKLNRFHIJQwΛ Sc[1285.32] : τὸ be || 263a24 αὐτὸ EKLNRFHIJQwΛ : αὐτὸ καὶ be [S] || 265a24 ἔτι EKLNRFHIJQwΛ : ἔτι δὲ be [S] || 265a30 τέλος καὶ μέσον EΨKLNRFHIJQwΛ Sp[1315.10-11] : μέσον καὶ τέλος be || 265b9 αὐτὴν E1KLNR HIJQwΛ Sp[1316.38]A[apud Sc(1317.4)] : ταυτην E2be || 266a1 τὴν EΨKLNRFHIJQw Λ S : καὶ be? To be sure, there are many variants among these which should not be adopted, but there are also some which deserve serious consideration, most strikingly perhaps at 257b8.

VI.5 The δ group (P.S. Hasper)

CXLIX

259b16 ἑαυτά/αὑτὰ E1KLNRFJ2 Sc[1258.32]P[890.26-35] : ἑαυτό be1HIJ1Qw [αὐτὸ Ψ?] 260a4 τὸν αὐτὸν κινήσει E?KLNR Sc[1262.31-33] : κινήσει τὸν αὐτὸν beFHIJ QwΛ [Ψ] 260b33 τετελειωμένων EΨKLNR Sp[1270.19] : τελουμένων F1J : τελειουμένων beF2HIQw [Λ] 262b31 τὰ EΨKLNRΛ : τὸ beFHIJQw 263b23 ὥστε ἢ KLNR : ὥστε εἰ ἦν EΨ : ὥστε beFHIJQwΛ Sp[1295.28]? 264b15 ταῦτα EΨKLNR : ταύτῃ beFHIJQwΛ [S] 266a2 κατὰ τόπον EΨKLNR Sc[1320.12] : τὴν κατὰ τόπον κίνησιν beFHIJQwΛ

Since these readings in be go against the consensus of the α and β groups, it is unlikely that they are what Aristotle wrote. Still, none of them is obviously wrong – indeed, quite a few of these readings are adopted by Ross. Thus without counter-evidence, one might hypothesise that be somehow share an ancestor with the γ group. That hypothesis, however, is false, for there are two related kinds of indications that be share an ancestor with the α group and were subsequently contaminated from the γ group. The first indication is a list of readings be share with the α group: 251a25 ἠρεμεῖ δεῖ F[ἀεὶ]HI[ἠρεμεῖν]JQwΛ T[211.5]? : ἠρεμεῖν ἀναγκαῖον δὴ LNR : ἠρεμεῖν ἀναγκαῖον δεῖ K : ἠρεμεῖ ἀναγκαῖον EΨ[ἠρεμεῖν]? be Sp[1128.4]? 251a27 στέρησις KLNRFHIJQw Sc[1128.17] : στέρησις τῆς Ebe [Ψ][Λ] [T] 251b6 om. LNRHQΛ : μὴ FIw : ἦν EΨKbe – J [S][T] 251b10 τὸ KLNRFHIJQw Sl[1152.20] : om. Ebe [Ψ][Λ] [T] 251b17 δ’/δὲ αὐτὸν LNRFHIJQwΛ T[211.28]? : δὲ EKbe Sp[1155.35] [Ψ] 251b32 παύσεται LNRFHIJQw Sc[1169.26] : παύεται EKbeΛ [Ψ] [T] 252a35 οὕτω KLNRFHIJQw : οὕτως Ebe [Ψ] 252b25 τὸ αὐτὸ KLNRFHIJQΛ Sp[1188.3] : τοῦτο τὸ αὐτὸ EΨ[vel τὸ αὐτὸ τοῦτο] : τὸ αὐτὸ τοῦτο be 253a16 τῷ σώματι NRHIJQw : om. L : τῷ σώματι post κινήσεις F : ἐν μὲν τῷ σώματι EΨ?KbeΛ Sp[1190.34]T[214.18] 253b7 τοῦτο KFIJQwΛ Pc[825.24] : om. LNRH : τούτου EΨbe [S][T] 253b23 ὁποιασοῦν KLNRFHIJQw : ὁποιασποτοῦν Ebe [Ψ] [S][T] 253b27 ἀνάγκη δὲ KLNRFHIJQwΛ : καὶ EΨb [e][F] [S][T] 253b28 om. KLNRFHIJQwΛ : μὴ EΨbe [S][T] 254a20 κινεῖται τῶν ὄντων LNRFHIJ{Q}wΛ : κινεῖται EΨKbe [S][T] 254b19 om. KLNRFHIJQ : ἂν pro τε w : ἂν Ebe [Ψ] [S][T] 255b14 αὑτῶν KLNRHJΛ Sp[1216.29] : αὐτῶν Iw : αὐτοῦ FQ? : αὐτὸν EΨbe [T] 256a3 ἄρα … κινοῖτο L[κινεῖται]NR : ἄρα … ἂν κινοῖτο FHIJQwΛ : ἂν … κινοῖτο EΨ?Kbe Sl[1219.12]T[220.2-3]

CL

Introduction 256b22-23 μέν ὑπ’ ἄλλου δὲ ἀλλ’ οὐχ LNRFHIJQwΛ : οὐχ ὑπ’ ἄλλου μὲν ἀλλ’ K : μέν οὐχ ὑπ’ ἄλλου δὲ ἀλλ’ EΨbe As[617]? [S][T] 256b30 ὥς γε Ross : ὥστε KLNRFHIJQw : ὥς τε Λ : ὡς τὸ Ψ? : ὥστε τὸ Ebe [S][T] 257b3 φέροιτο … φέροι KLNRFHIJQwΛ : φέροι … φέροιτο EΨbe Sc[1234.1213]T[221.29] 257b16 ἑκάτερον LNRFHIJQwΛ : αὐτὸ αὐτὸ E : αὑτὸ Ψ? Sp[1238.8] : αὐτὸ b : αὐτὸ αὑτὸ K[ἑαυτὸ]e 257b28 αὑτὸ KeLNRFHIJQ[ἑαυτὸ]wΛ : αὐτὸ EΨb [S] 257b31 αὑτὸ KeLNRFHIJQ[ἑαυτὸ]wΛ : αὐτὸ EΨb [S] 257b32 γὰρ τοῦτο μὲν KLNRFHIJQw Sp[1240.32] : μὲν γὰρ τοῦτο μὲν Λ : μὲν γὰρ τοῦτο Ebe [Ψ] 258a4 κινηθήσεται1 LNRFIJ{Q}w : κινήσεται EKbeHB Sl[1241.28]Sc[1224.32] [Ψ][Λ] 258a4 κινηθήσεται2 KbLNRFHIJQw Sl[1241.28] : κινήσεται Ee Sc[1224.32] [Ψ][Λ] 258a7 ἔτι LNRFHIJQwΛ : om. EΨKbe1 A[apud S(1242.26)]? [S] 258a12 ἅπαν KbLNRFHIJQw : ἅπαν τὸ EeΛ Sp[1242.38] [Ψ] 258b3 δύναμιν LNRFHIJQwΛ : φύσιν EΨKbe Sc[1246.25]Sp[1246.16] Pp[837.3]? A[apud S(1246.25)] 258b11 εἴτε οὖν LNRFHJQwΛ : εἴτε1 EΨ?KbeI As[648] [S] 260a26 καὶ2 LNRFHIJQwΛ Sp[1265.6] : om. EΨKbe [T] 260b23 δ’ ἐστὶ LNRFHIJQwΛ Sp[1269.33]? : δὲ EKbe [Ψ] 260b29 πρώτην KLNRFHIJQw Sl[1270.1] : πρότεραν vel πρότερα Λ : πρώτη EΨ?be [T] 262a31 ἐν … τῷ ΑΒΓ LNRFHIJQw : om. EΨKbeΛ [S] 262b13 τῷ Α τὸ Δ KLRHIΛ : τὸ Α τὸ Δ N : τῷ Α Jw : τὸ Α F : τὸ Δ EΨbeQ [S][T] 263a1 ἄρα LNRFHIJQw : om. EΨKbeΛ Sc[1288.25]Sp[1288.13-16] [T] 263a29 ἐὰν E2LNRFIJQw : ἂν E1KbeH 263b16 μὲν KLNRFHIJQwΛ : om. Ebe [Ψ] [S] 264a16 ἐφέρετο eLNRFHIJQwΛ, cf. T[230 – de Alexandro?] : φέρεται EΨKbe [S] 264b7 δ’ οὖν ἂν LN : ἂν FI? : οὖν ἂν RHQΛ : ἂν οὖν Jw : δ’ ἂν EKbe Sp[1307.4]? [Ψ: non solum ἂν] 265a31 οὗ KLNRFHIJQwΛ Sp[1315.14]T[232.29] : οὐ EΨbe 266a7 κίνησις πρώτη LNRFHIJQwΛ : πρώτη κίνησις EKbe Sp[1320.30] [Ψ] 266a10 ἀναγκαῖον ἀμερὲς εἶναι LNRHIJQwΛ : ἀναγκαῖον εἶναι ἀμερὲς F : ἀμερὲς ἀναγκαῖον εἶναι EKbe Sl[1321.1] [Ψ] 266a21 Β ὅλην LNRFHIJQw Sp[1322.10] : ὅλην Β EKbeΛ [Ψ] [T] 266a24-25 ἐν πεπερασμένῳ μεγέθει οὐκ ἐνδέχεται LNRFHIJQwΛ : οὐκ ἐνδέχεται ἐν πεπερασμένῳ μεγέθει EKbe Sl[1323.34]Sp[1324.6-7]T[234.5-6]? [Ψ]

VI.5 The δ group (P.S. Hasper)

CLI

And unlike the list of readings shared with the γ group, there are also some errors among these readings: 251b10, 252b25 (where the reading τοῦτο has been corrected in both EΨ and be, but in not quite the same way), 253b23, 255b14, 263b16 and 265a31. Also among the readings be share with only E there are errors: 256a28 257b18 264a8 266a10

αὑτῷ Ψ?KLNRFHJQΛ Sp[1223.20] : αὐτῷ Ebe [T] τὸ2 Ψ?LNRHJQ : om. KF : τῷ E1?E2be [Λ: non τῷ] [S] κἂν E2Ψ?KLNRHIJQΛ : καὶ E1beF [S][T] ἔχειν E2ΨKb2eLNRFHIJQΛ Sp[1321.13]? : ἔχ… E1b1

(though at 256a28 Ψ may have been without breathing mark and at 257b18 the reading of Ψ cannot be reconstructed with certainty). The second indication that be have been contaminated from the γ group is that there are a number of places where we can observe the process of contamination itself: 251a4-5 τὸ γὰρ ᾗ δὲ τάδ’ ἀλλάσσοντα ἐνθένδε ἐκεῖσε λέγειν αὐτὸν ὑποληπτέον LNRH : τὸ γὰρ τῇ δὲ τάδ’ ἀλλάσσοντα ἐνθένδε ἐκεῖσε λέγειν αὐτὸν ὑποληπτέον F[τὰ?]IJQwΛ[διαλλάσσονται] : δεῖ γὰρ ὑπολαβεῖν λέγειν αὐτὸν ᾗ δὲ ταδ’ ἐνθένδε τὰ ἀλλάσσοντα EΨ[τὰ om.?]K[ἀλάσσοντα] Slp[1123.8, 1125.5 – τὰ om.] : δεῖ γὰρ ὑπολαβεῖν λέγειν αὐτὸν τῇ δὲ ταδ’ ἄλλασονται ἐνθένδε ἐκεῖσε be [T] 251b21 ἅμα KLNRFHIJQwΛ T[212.2-3] : ἀλλ’ E1Ψ : ἅμα ἀλλ’ E2be [S][P] 257b7 δυνάμει2 FHIJQwΛ : κινούμενον EΨKLNR As[627] Sp[1235.14-15] Pp[834.25] : δυνάμει κινούμενον be 264a21 οὐ μία EKLNR Sl[1301.31] : οὐκ ἕστι μία FHIJ[ἔστιν]QwΛ : οὐ μία ἔστιν be T[230.28] [Ψ] [S]

As to the more specific source of these contaminations in be from the γ group, it may be that the following list is relevant: κινεῖ τι αὐτὰ EKLNRIJQwΛ : τι αὐτὰ F : κινεῖ τι I : τι κινεῖ αὐτὰ beHB μαθητικόν EKLNRF2IJQwΛ : μαθηματικὸν F1 : μαθητόν beHB ἀεὶ εἶναι EKLNRFIJQwΛ Sc[1266.4] : εἶναι αἰεὶ beH[ἀεὶ] μὲν EΨ?KLNRJQwΛ : om. beFHI T[227.31] [S] διεξελθεῖν EKLNRIJQw : διελθεῖν beFH [Λ] πιστεύσει λόγοις EKLNRFIJQwΛ Sl[1301.1]Sc[1301.16] : λόγοις πιστεύσειεν beHB 266a29 τὸ EΨKLNRHJQw A[apud Sc(1325.20)] : om. beFI [Λ] 267a4 τι ἄλλο E2KLNRIJQwΛ Sp[1345.30] : τι τε ἄλλο E1 : ἄλλο τι beFH

252b22 257a21 260b6 261b28 263a17 264a7

CLII

Introduction

Contamination from a manuscript related to H or FH could explain these shared readings. On the other hand, there are also some striking shared readings between be and Λ: 254a18-19 ἢ τὰ μὲν … κινεῖται EKe2LNRFHJQw : om. be1Λ [S] 256a29 κινούμενον τι E1[.τι]E2ΨKLNRFHJQw : κινούμενον Λ : τι κινούμενον be [S][T] 257b2 αὑτὸ κινοῦν ΨeLNRI Sl[1233.34-a] : ἑαυτὸ κινοῦν FHJQ : αὐτὸ κινοῦν Eb2w : κινοῦν Kb1 Sl[1233.34-AFM]As[625]? : κινοῦν αὑτὸ Λ 263a16-17 ἄπειρα ἐνδέχεται διεξελθεῖν ἢ ἀριθμῆσαι EKLNRFHIJQw : ἄπειρα διεξελθεῖν ἢ ἀριθμῆσαι ἐνδέχεται Λ : ἐνδέχεται ἄπειρα διεξελθεῖν ἢ ἀριθμῆσαι be [S] 263b4 ἔστιν EKLNRFHIJQw : ἔστι μὲν beΛ [Ψ]

A further consideration, however, is that consensus in the case of be and Λ does not say anything about direction. We had, therefore, best leave the matter open. The fact that the δ group thus seems to be closely related to the α group, though contaminated from the γ group, opens up the possibility that the places where the δ group agrees with one of the commentators, notably Simplicius, against the consensus of the α and γ groups, are places we have to take very seriously for the constitution of the text. They would then also be places where the α group has been contaminated from the γ group – we will look into this matter when discussing the possible contaminations in the α group. What this fact also suggests is that a consensus between the δ and β groups, against the consensus of the α and γ groups, is to be taken very seriously for the constitution of the text as well. It concerns the following places: 255a15 συμφυῆ KFHIJQwΛ : συμφυὴς EΨ : συμφυὲς beLNR [S] 257b32 αὑτὸ beLNR : ἑαυτὸ KFHIJQwΛ : αὐτὸ EΨ277 259a13 ἀρχὴ τοῖς ἄλλοις EKFHIJQwΛ Sl[1254.10]Sc[1254.28] : τοῖς ἄλλοις ἀρχὴ beLNR [Ψ] 259b23 συνεχῶς EFHIJQwΛ : συνεχῆ KbeLNR [Ψ] 261a5 πρῶτον EΨFHIJQwΛ : προτέρον KbeLNR Sc[1270.28] 261b36 ἐστὶν ἡ EΨKFHIJQw Sp[1279.2] : ἐστι beLNR [Λ] 262b1 τὸ Α ἅμα γεγονέναι EΛ : ἅμα τὸ Α γεγονέναι KbeLNR : ἅμα γεγονέναι τὸ Α HIJQw [Ψ][F] [S] 263a26 ἡμίση EFHIJQw : ἡμίσεα KbeLNR [Λ] 263b1 ἀριθμεῖν δύο EFHIJQwΛ : δύο ἀριθμεῖν KbeLNR Sp[1291.17-18] [Ψ]

_____________ 277

Often the addition of ἑ to turn αὑτός into ἑαυτός is the result of a reading with either the wrong breathing sign (αὐτός) or without breathing sign (αυτός).

VI.5 The δ group (P.S. Hasper)

CLIII

264b3-4 ἀλλοίωσις εἰς E2ΨFHIJQwΛ : ἀλλοίωσις ἡ εἰς KbeLNR : ἀλλοίωσεις E1 [S] 264b14 ἐναντία μὲν ἡ EΨ?KFHIQ2wΛ : ἐναντίαι μὲν αἱ beLNR : ἡ ἐναντία μὲν Ψ?JQ1 [S]

In all of these passages the reading of beLNR seems possible, and in quite a few even preferable. What is more, for 257b32 it is adopted by Ross, even without any direct support in the manuscripts he uses.

VI.6 Ms. Laur. gr. 87.24 Ross already recognises that Laur. gr. 87.24, known as K, a manuscript which only contains the final three books of the Physics, features readings relating it to the α group and readings relating it to the γ group. However, he takes K to belong to an independent tradition, and thus grants it the authority to decide whether to go with the α reading or the γ reading278. Against this hypothesis, I want to argue here that just as be, K is a manuscript which is stemmatically related to the α group, but has been contaminated or corrected from at least the γ group. First I list the evidence that K is related to the α group: 251a4-5 τὸ γὰρ ᾗ δὲ τάδ’ ἀλλάσσοντα ἐνθένδε ἐκεῖσε λέγειν αὐτὸν ὑποληπτέον LNRFHIJQwΛ : δεῖ γὰρ ὑπολαβεῖν λέγειν αὐτὸν ᾗ δὲ ταδ’ ἐνθένδε τὰ ἀλλάσσοντα EΨ[τὰ om.?]K[ἀλάσσοντα]b[τῇ, ἄλλασονται ἐνθένδε ἐκεῖσε] e[τῇ, ἄλλασονται ἐνθένδε ἐκεῖσε] Slp[1123.8, 1125.5 – τὰ om.] [T] 252a33 ὅτι ΨbeLNRFHIJQwΛ P[824.10]? : εἴ τι EK [S][T] 252a33 ἢ1 beLNRFHIJQwΛ P[824.10] : om. EΨ?K T[212.34] [S] 252b29 ἀεὶ beLNRFHIJQw : αἰεὶ EK 252b31 ἀεὶ E2beLNRFHIJQw : αἰεὶ E1K 252b34 ἀεὶ E2beLNRFHJQwΛ Sp[1189.2] : αἰεὶ E1K : om. I 253a31 τε πάντων τῶν ἀπορουμένων EΨ?K : τε τῶν ἀπορουμένων LNR : πάντων τε τῶν ἀπορουμένων FHIJQwΛ : τε τῶν ἀπορουμένων ἁπάντων be [S: πάντων hab.] 253b35 ἔνι’ E1K : ἔνια E2beLNRFIJQw [HB] 254a23 πάντ’ EK : πάντα beLNRFHIJQw 254a26 κινεῖσθαι E2ΨK : κεινεῖσθαι E1 : κινεῖται beLNRFHBIJQwΛ [S] 254a28 οὕτως beLNRFHIJQw : οὕτω EK 254b32 αὑτὸ ΨLNRFHIJQwΛ Sl[1206.33] : αὐτὸ EK : ἑαυτὸ be 254b35 ἀντιθεῖναι beLNRFHIJQwΛ : ἀντιτιθεῖναι K : τιθείναι E1? : τιθέναι E2Ψ [S]

_____________ 278

Ross, Aristotle’s Physics, p. 111.

CLIV

Introduction

255b7 ἐμποδίζῃ EK Sp[1215.10]? : ἐμποδίσῃ beLNRFHIJQw [Ψ][Λ] 257a29 τὸ αὐτὸ αὑτὸ LNRFHIJQ[ἑαυτὸ]wΛ : τὸ αὐτὸ αὐτὸ E1K : τὸ αὐτὸ E2Ψbe : τὸ αὑτὸ Sc[1247.31] 257b1 τοῖς EK Sp[1233.30] : om. Ψ?beLNRFHIJQwΛ? 257b4 καὶ1 E2ΨbeLNRFHIJQwΛ Sc[1249.28] : ἢ E1 Sc[1234.14] : ἢ καὶ K 258a2 μόνως … εἶναι E2beLNRFHIJQ2wΛ Sl[1240.22-a]Sp[1241.24-25], cf. As[635] : om. E1ΨK Sl[1240.22-AFM]Sp[1242.15] [Q1] 258a18 ἑαυτὸ Ψ[vel αὑτὸ]beLNRHIJQw : αὐτὸ EK : αὑτὸ F 258a21 μὲν2 EΨK As[641]Sc[1244.14]Pc[836.26] : om. beLNRFHIJQwΛ 258b8 πρώτως κινοῦν beLNRFHIJQwΛ Sl[1247.11-a]Pp[887.6] : πρῶτον EΨK Sl[1247.11-AFM] 258b10 διαλείπειν E2Ψ?beLNRFHIJQw Sl[1250.32] : διαλίπειν E1K [Λ] 259a33 κινουμένων2 EbeLNRFHIJQwΛ Sp[1257.31]? : κινούμενον Ψ?K 259b4 ἐγγίγνεσθαι LNR : ἐγγένεσθαι JQ : ἐγγίνεσθαι beFHIw : γίνεσθαι E1Ψ?K Sp[1258.1]? : γίγνεσθαι E2 [Λ] 260b21 συνεχὴς οὖσα ἢ ἐφεξῆς EΨK : ἡ συνεχὴς οὖσα ἢ ἡ ἐφεξῆς LNR : ἡ συνεχὴς ἢ ἡ ἐφεξῆς b[ἡ1 p.c.]eFHIJQwΛ 261a29 καὶ2 EΨK Sc[1256.21, 1273.21] : καὶ τὸ beLNRFHIJQw [Λ] 262a10 κἂν E1K Sc[1280.6] : καὶ ἐὰν E2beLNRFHI : καὶ ἂν JQw [Λ] 262a10 γίγνηται LNR : γίνηται beFHIJQwΛ : γίγνεται EK [Ψ] 262b29 ἡ EbeLNRFHIJQwΛ : ἢ ΨK 263a5 εἰ EbeLNRFHIJQwΛ : om. ΨK [S][T] 263a16 ἠρωτᾶτο E2ΨbeLNRFHIJQw : ἠρωτᾶτω E1K : ἠρωτᾶ Λ 263b2 ἂν E1K : ἐὰν E2beLNRFHIJQw Sl[1288.31] 263b3 Β E2ΨbeLNRFHIJQ2wΛ Sc[1298.1]Sp[1297.27] : ?? Q1 : Ω K : ΩΒ E1? 263b31 Β E2ΨbeLNRFHIJQ2wΛ : ?? Q1 : Ω E1K 264a1 τοῖς E2ΨbeLNRFHIJQw Pl[846.1] : om. E1K [Λ] 264a3 οὗ E2beLNRFHIJQΛ : ᾧ K : om. E1Ψ 264b9 οὐθὲν E1K : οὐδὲν E2beLNRFHI?JQw 265b26 καὶ beLNRFHIJQwΛ : om. EΨK Sc[1319.7] 266a17 τῆς Β κινείτω τι μέρος EK : κινείτω τι μέρος τῆς Β beLNRFHIJQwΛ Sp[1322.8]? [Ψ] 266a26 γὰρ ἡ … ἀεὶ ἡ beLNRFHIJQwΛ : γὰρ αἰεὶ … E1K : γὰρ ἀεὶ ἡ … E2Ψ : γὰρ ἡ … Sp[1324.8-9]? 266a30 ἡ ἄπειρος EK A[apud Sc(1325.20)] : τὸ ἄπειρον Ψ? : ἡ ἄπειρος δύναμις beLNRFHIJQwΛ 266a32 οὗ E1Ψ?K : ᾧ E2beLNRFHIJQw [Λ] 266b1 λαμβάνων ἀεὶ EK : ἀεὶ λαμβάνων beLNRFIJQwΛ Sp[1324.33]? : λαμβάνων H [Ψ: ἀεὶ hab.] 266b3 ἀφαιρῶν EbeLNRFHIJQwΛ Pp[850.28] : ἀφαιρῶν ἀεὶ Ψ*K [S][T]

VI.6 Ms. Laur. gr. 87.24 (P.S. Hasper) 266b4 266b7 266b9 266b12 266b13 266b26 266b27 267a4 267a15 267a19 267a20 267b5 267b12 267b22 267b24

CLV

τῇ ἀπείρῳ ἡ πεπερασμένη EK : ἡ πεπερασμένη τῇ ἀπείρῳ b[ἢ]eLNR FHIJQwΛ Sp[1324.38] [Ψ] πλείω E1K Sp[1340.24] : πλείονα E2beLNRFHIJQwΛ [Ψ] ἣ ERFHIJQΛ : ἢ beLNw : ᾗ ΨK Sp[1341.24] ΖΘ EΨK As[802] : Ζ καὶ Θ beLNRFHIJQwΛ : ΘΖ Sc[1342.39] τοῦ EΨK Sc[1342.4] : om. beLNRFHIJQw [Λ] ἐν ἀπείρῳ πεπερασμένην EΨ*K : πεπερασμένην ἐν ἀπείρῳ beLNRF[ἐν om.]HIJQwΛ Sp[1344.3]? ἔχει καλῶς EK Sl[1344.4-A] : καλῶς ἔχει beLNRFHIJQwΛ Sl[1344.4-a] [Ψ] τοιοῦτον2 beLNRFHIJQwΛ : om. EΨK Sp[1345.30]? διὸ E1ΨK Sc[1350.19, 28] : διὸ καὶ E2beLNRFHIJQwΛ παύεσθαι E1?Ke2J2w : παύεται E2Ψbe1LNRFHIJ1QΛ [S] τι ἓν E1K : ἓν τι E2beLNRFHIJQwΛ Sp[1351.19] [Ψ] τὸ κινοῦν οὐδεμίαν EK T[235.32] : οὐδεμίαν τὸ κινοῦν b2eLNRFHIJQwΛ [Ψ][b1] [S] ἀεὶ ὠθεῖν EΨK : ὠθεῖν ἀεὶ Q1 : ὠθεῖν beLNRFHIJQ2wΛ Sp[1356.7-8]? δύναμιν ἄπειρον EK : ἄπειρον δύναμιν beLNRFHBIJQwΛ Sp[1358.11] [Ψ] τὸ δέ γε E2Ψ[vel γε om.]beFHIJQw : τὸ δὲ LNR : τὸ γε K : δὲ τὸ γε E1 [Λ: δὲ hab.] [S]

The reading of Ψ cannot be ascertained for all of these places, but it is clear that there are both places for which K only shares a reading with E (thirteen), and places for which it only shares a reading with Ψ (five). What is more, there are errors among these readings, notably 251a4-5, 252a33, 254b35, 258a18, 259a33, 262b29, 263a5, 263a16, 263b31, 266a26, and perhaps 258b10 and 266b3. At 254b35, 257b4 and 264a3 we can still witness the process of correction in K’s exemplar. The conclusion to draw is therefore that K is related to the common ancestor of E and Ψ. This conclusion is confirmed if we consider the places where K shares a reading with both the α and the δ groups, for, as shown above, the δ group itself is related to the common ancestor of E and Ψ: 251b6 ἦν EΨKbeJ1? : om. LNRHJ2QΛ : μὴ FIw [S][T] 251b17 δὲ EKbe Sp[1155.35] : δ’ αὐτὸν LNR[δὲ]FH[δὲ]I[δὲ]JQwΛ T[211.28]? [Ψ] 253b26 τι νοσήσῃ E1 : νοσήσῃ τί LNRFHI2J0Qw2Λ : νόήσῃ τι w1 : νοσήσῃ τις KbeI1J1 : τις νοσήσῃ E2 : νοσήσῃ τις vel τις νοσήσῃ Ψ [S][P] 253b30 ὁ δὲ EKbe?J2w : ὅ τε LNRFHIJ1QΛ : ἔτι δὲ ὁ Ψ [Sp 1200.4: ὁ] 254a20 κινεῖται EΨKbe1 : κινεῖται τῶν ὄντων e2LNRFHIJ{Q1}Q2wΛ [S] 255b19 ὡς EΨKbe1 Sp[1216.37] : om. e2LNRFHIJQwΛ 256b29 ᾖ LNRFHBJQ2 : ἢ Q1?w : εἰ EΨKbeQ3Λ [S]

CLVI

Introduction

οὐκ ἔστιν E2ΨKbe Sl[1237.3] : ἔστιν E1 : οὐκ ἐνδέχεται LNRFHIJQwΛ, cf. As[629] 257b16 αὐτὸ ἑαυτὸ K : αὐτὸ αὐτὸ E1 : τὸ αὐτὸ αὐτὸ E2 : αὐτὸ b : αὐτὸ αὑτὸ e Sp[1238.1] : αὑτὸ Ψ? Sp[1238.8] : ἑκάτερον LNRFHIJQwΛ 258a4 κινήσεται E2KbeHBJ2 Sl[1241.28]Sc[1224.32] : κεινήσεται E1 : κινηθήσεται LNRFIJ1{Q1}Q2w [Ψ][Λ] 258a7 ἔτι LNRFHIJQwΛ : om. EΨKbe1 A[apud S(1242.26)]? 259a10 αἰεὶ EKbe : ἀεὶ LNRFHIJQw 260a26 καὶ2 e2LNRFHIJQwΛ : om. EΨKbe1 260b2 ἐνεργείᾳ EKbe : τὸ ἐνεργείᾳ ΨLNRFHIJQw Sp[1265.31] [Λ] 260b23 δὲ EKbe : δ’ ἐστὶ LNRFHIQwΛ : δ’ ἐστιν J [Ψ] 262a9 τὸ EΨ?KbeF1wΛ Sp[1280.2] : τῷ LNRF2HIJQ 262b1 τε EKbeF2 : om. LNRHIJQwΛ [Ψ][F1] 263a1 ἄρα LNRFHIJQw : om. EΨKbeΛ Sc[1288.25]Sp[1288.13-16] 264a16 φέρεται πρὸς τὸ Γ EΨKbe1 : ἐφέρετο πρὸς τὸ Γ e2LNR : εἰς τὸ Γ ἐφέρετο FHIJQwΛ [S] 264b7 δ’ ἂν EKbe Sp[1307.4]? : δ’ οὖν ἂν LN : ἂν FI1? : οὖν ἂν RHI2QΛ : ἂν οὖν Jw [Ψ: non solum ἂν] 266a7 πρώτη κίνησις EKbe Sp[1320.30] : κίνησις πρώτη LNRFHIJQwΛ [Ψ] 266a10 ἀμερὲς ἀναγκαῖον εἶναι EKbe Sl[1321.1] : ἀναγκαῖον ἀμερὲς εἶναι LNR HIJQwΛ : ἀναγκαῖον εἶναι ἀμερὲς F [Ψ] 266a13 κινοῦν τὸ κινούμενον EΨKbe[καὶ τὸ]Λ Sp[1322.1] : κινούμενον τὸ κινοῦν LNRFHIJQw 266a21 ὅλην Β EKbeΛ : Β ὅλην LNRFHIJQw Sp[1322.10] [Ψ] [T] 266a24-25 οὐκ ἐνδέχεται ἐν πεπερασμένῳ μεγέθει EKbe Sl[1323.34]Sp[1324.6-7] T[234.5-6]? : ἐν πεπερασμένῳ μεγέθει οὐκ ἐνδέχεται LNRFHIJQwΛ [Ψ] 257b14

The very few cases in which Kbe here agree with E or Ψ alone, and disagrees with the other, must be cases where either Ψ or E was corrected (253b30, 260b2) or simply made a mistake (253b26, 257b14, 16). On the other hand, the readings K shares with other groups, notably the γ group, do not comprise many errors, and are thus more likely the product of contamination and correction: 252a6-7 νομίζειν εἶναι ἀρχήν KFHIJQwΛ : εἶναι νομίζειν ἀρχήν ELNR : νομίζειν ἀρχὴν εἶναι be [Ψ] [S][T] 252a18 καὶ EΨbeLNRI2 : om. KFHI1JQwΛ Sp[1185.1]? [T] 253a3 κινῆσον KJQw : κινῆσαν EΨbeLNRFHIΛ? [S] 253b7 τούτου EΨbe : τοῦτο KFIJQwΛ Pc[825.24] : om. LNRH [A][S] 253b16 ἐξέωσεν EbeLNRΛ : ἔωσεν KFHIJQw Pc[826.17]

VI.6 Ms. Laur. gr. 87.24 (P.S. Hasper)

CLVII

254a32 τὸ2 KFHIJQw : om. EbeLNR [S] 255b5 καὶ ἐν EΨbeLNRHQ2 AγρSγρ[1214.35] : καὶ KFIw : καὶ οὐχ ἁπλῶς ἐν As[593]A[apud S(1214.26-27)] : καὶ οὐκ ἐν JQ1 : om. Λ 256a25 αὑτῷ Sp[1223.16] : αὐτῷ LN2HI2? : αὐτὸ EbN1RFw : αὑτὸ Ψ?eI1?J2Q2Λ? : om. KJ1Q1 256a31 ἄλλο KFHIJQwΛ : ἀλλο τι EΨ?beLNR 262b18 ἅμα κινεῖσθαι KFHIJQw : κινεῖσθαι ἅμα EbeLNR : κινεῖσθαι ΨΛ [Sp(1286.9): ἅμα hab.] 264a33 ἦν EΨbeLNR Sc[1304.29] : om. KFHIJQwΛ 264b5 γέγονε λευκὸν καὶ γέγονεν οὐ λευκόν E2Ψbe1LNRHQ2Λ Sp[1306.1-4] Pp[847.11-18] : γέγονεν οὐ λευκόν KFIJQ1w Pl[847.10]? : γέγονε λευκὸν E1 : γέγονε λευκὸν λευκόν e2 265a6 φασιν ἀεὶ EbeLNRJQΛ : ἀεὶ φασὶ KFHIw

Of these places only 256a25 and 264b5 concern errors, but they could well be coincidental ones.279 The picture does not change if one takes all the places at which K and the δ group agree with the γ group: 250b21 251a5 251b16 251b30 252a6 252a35 253b9 254a7 254a25 254b11 256a31 258a26 258a26 258b22 260b26 263b21 267a22

τὰς KbeFHIJQw T[209.10] : om. ELNR δὴ KbeFHIJQwΛ Sl[1125.25-aAM] : δὲ EΨLNR Sl[1125.25-F]T[209.22] τούτου EΨLNRHJQ1w : τοῦτο KbeFIQ2Λ T[211.31]? [S] εἶναι τινα ELNR T[212.11-M] : τινὰ εἶναι KbeFHIJQwΛ : εἶναι T[212.11CSL] [Ψ] [S] ταύτην δεῖ KbeHIJQwΛ : ταύτην δεῖν F : δεῖ ταύτην ELNR : ταύτην Ψ Sp[1183.23] [T] ὡς KbeFHIJQw : ὅτι ELNR Sp[1186.20] [Λ] [T] ὅμως KbeFI : ὁμοίως EΨLNRHBJQwΛ [S][T] γὰρ KbeFHIJQwΛ T[216.23]Pc[827.17] : γὰρ καὶ EΨLNR [S] φασί τινες ELNRΛ : τινες φασὶν KbeFHIJQw [S] τῷ2 KbeFHΛ : om. ELNRIJQw τὴν βακτηρίαν KbeFHIJQwΛ : τῇ βακτηρίᾳ EΨLNR ἡ1 E2LNRQ : ἢ E1 : ᾗ Ψ : om. KbeFHIJw [Λ] ἡ2 E2LNR : ἢ E1 : ᾗ Ψ : om. KbeFHIJQw [Λ] γε πάσας EΨ?LNR T[223.21] : πάσας γε KbeFHIJQw [Λ] αὔξεσθαι ELNRJQw Pl[899.14] : αὔξάνεσθαι KbeFHI [Ψ][Λ] δ’ ἤδη KbeFHIJQwΛ : γέγονεν L : δὴ Ψ[vel δὲ]?NR : δὲ δὴ E1 [S] τέ EΨNRIJQ : γε KbeFH : om. Lw [S]

_____________ 279

Indeed, Ross adopts the reading found in K at five of these places, though not as often as one would have expected from his ascribing to K the authority to decide between the α and γ groups.

CLVIII

Introduction

There are no clear errors among these places280. Thus there is no reason to assume that K belongs to the γ group as well, or that it constitutes a tradition which is independent from the α and γ groups. This is confirmed by three passages for which we can catch a glimpse of the process of contamination from the γ group: 251a25

ἠρεμεῖ δεῖ F[ἀεὶ]HI[ἠρεμεῖν]JQwΛ T[211.5]? : ἠρεμεῖν ἀναγκαῖον δὴ LNR : ἠρεμεῖν ἀναγκαῖον δεῖ K : ἠρεμεῖ ἀναγκαῖον EΨ[ἠρεμεῖν]?be Sp[1128.4]? 258a26-27 κινεῖται δὲ ἡ τὸ Β E[ἢ]Ψ[ᾗ in loco ἡ]beLNR : κινεῖται δὲ τὸ Β K : τὸ δὲ Β κινεῖται FHIJQwΛ 266a33 τῷ δὲ E1Ψ Sp[1324.31] : ᾧ δὲ E2K : ᾧ δὲ ὁ beLNFHIJQ1?wΛ : ᾧ δὲ ἡ RQ2

At 251a25 K inserts δεῖ, just as in the γ group, without striking the ἀναγκαῖον which we find in EΨbe. At 258a26-27 K still follows the α, β and δ groups as far as the word order is concerned, but is the only manuscript to strike, with the γ group, the article ἡ. The passage at 266a33 is even clearer: if one reads ᾧ, as K does with the β and γ groups, then one must add the article ὁ in order to make even grammatical sense of the passage (philosophical sense it still does not make) – but only K does not do that. That K is susceptible to contamination also appears from the fact that there are places at which one can recognise an influence from Simplicius’ commentary: 251b18 γεννᾷ EΨbeLNRFHIJQwΛ Sl[1154.1-a]P[PC 222.23]T[211.28] : γεγονέναι K Sl[1154.1-AFM] 252a19 βέλτιον EK2beLNRFHIJQw : ἄμεινον K1 Sl[1185.16] [Λ] [T] 257b1 καθ’ αὑτὸ κινούμενον EbeLNRFHIJQwΛ : κινούμενον καθ’ αὑτὸ K Sl[1233.24] 265a32 ἢ ὅθεν ἢ οὗ E1?Ψ*be1LNRFHIJQwΛ : τὸ πόθεν ποῖ E2Ke2, cf. Sp[1315.16] 267b9 ὅλου EΨbeLNRFHIJQwΛ T[236.3] : κύκλου K Sp[1354.8] Sc[1357.16]

Given that both K and the δ group are related to the α group, one might wonder what the stemmatic relations between these three are. This is a difficult question to answer, because of the strong contaminations from the γ group in both K and the δ group, but also because of the contamination, be it less significant, from the γ group in the α group281. Here is a list of places at which Kbe probably deviate uniquely from the archetype: _____________ 280

Here Ross adopts the reading found in K at nine of these places, still not as often as one would have expected.

VI.6 Ms. Laur. gr. 87.24 (P.S. Hasper) 252a27 252a34 252b25 254b21 254b23 255b32 266a30

CLIX

προσοριεῖται EΨbe2LNRFHIJQwΛ : προσδιοριεῖται Ke1 [S][T] 281 ἐφ’ Ee2LNRFHIJQw Pa[365.10] : εἰς Kbe1 [Ψ] [T] ἐπεί ELNRFHIJQwΛ : ἐπείπερ Kbe [Ψ] [S][T][P] κινεῖται EK2LNRFHIJQwΛ Sc[1207.12] : κινεῖσθαι K1be [Ψ] δὲ ELNRFHIJQwΛ : δὲ καὶ Kbe Sp[1208.17] : καὶ Pa[373.25] [Ψ][T] καὶ3 EΨLNRFHIJQw Sp[1219.15] : καὶ τὰ Kbe [Λ] [T] πλείων EΨeLNRFHIJQw : πλείω Kb

There are also some orthographical variants which are (almost) exclusive to Kbe: 251a25 οὐθέν ELNRFIJQ : οὐδὲν KbeH 259a32 ἀεὶ ELNRFHIJQw Sc[1257.26] : αἰεὶ Kbe : om. Pa[391.8] 265b7 ἀεί EbeLNRHIJQwΛ Sl[1315.6]Pa[419.12] : αἰεὶ Kbe

and: 250b22-23 δ’ ἕνα ἢ μὴ FHIJQwΛ : δὲ ἕνα ἢ μὴ Kbe : δ’ ἓνα καὶ μὴ Ψ Pa[358.2-3] : δὲ ἢ ἕνα ἢ μὴ E1LNR Sc[1122.26-28] : δὲ εἶεν ἀεὶ μὴ E2 [T(209.11-13) interpr. ἕνα ἢ ἀεὶ ἢ μὴ ἀεὶ, sed S quoque] 252b25 ὥστ’ ELNRFHIJQw : ὥστε Kbe 253a1 δ’ ELNRFHIJQw : δὲ Kbe 257b34 ὥστε ELNRFHIJQw : ὥστ’ Kbe 267a13 ἐγγίγνεται Ke : ἐγγίνεται EbLNRFHIJQw

Unfortunately there are only two places for which we have the evidence of Ψ, so that these lists do not provide solid evidence, though it cannot be excluded that collations of Physics VI would provide more. Also a comparison between the places at which only K or only the δ group agree with the α group would in principle not be of much help, because one might just be comparing the amounts of contamination and correction in K and in be. Thus it seems best to leave the matter open. Thus we have the following stemma for the other branch of the stemma – let us call it the χ1 branch: _____________ 281

As can be ascertained from the places at which Kbe probably have the correct reading: 255a1 παράσχοι ἂν ELNRFHIJQwΛ : ἂν παράσχοι Kbe [S][T] || 257a3 ἢ E1ΨK2 LNRFHIJQwΛ : εἰ δ’ E2[δὲ]K1be Sl[1229.6: δὲ] || 260a6-7 δὲ κινουμένου ἤδη E2e2 FHIJQw Pl[894.5] : δὴ κινουμένου ἤδη LNR : ἢ κινουμένου ἤδη Kbe1 Sc[1263.16-17] : κινουμένου ἤδη Λ : δὲ κινουμένου ἢ E1 : om. Ψ || 261a16 τοῦ ὀργάνου E1E3Ψ e2LNRF[cum ὑπάρχει]HIJQwΛ : om. Kbe1 Pc[900.26] : exp. E2 || 265a26-27 στάσιν … κίνησις E1ΨK2e2LNRFHIJ[partim]QwΛ, cf. Sp[1314.36] : om. E2K1be1 Sl[1313.211315.5]T[232.26-27 non reddit] || 267a2 κινεῖ τι ἄλλο E2Kbe : κινεῖται ἄλλου LNR FHIJQwΛ: κινεῖ ἄλλο E1Ψ* Sp[1346.11-12(vel κινεῖ τι)].

CLX

Introduction

χ1 γ α

E

Ψ

K

be

Just as in the case of consensus between the δ and β groups, this stemma suggests one should take very seriously those readings which are shared by K and the β group: 251b11 ἢ KLNR : ἢ ὁ beFHI2JQw : ᾗ EI1 [Ψ hab. ἢ][Λ] [S] 254a18 ἢ γάρ τοι EbeFHJQw : ἤτοι γὰρ KLNR : ἤτοι ἄρα Ψ? [Λ: γάρ] [S] 260b32-33 ἀλλοίωσις καὶ αὔξησις EΨbHIJ1?Q2wΛ : αὔξησις καὶ ἀλλοίωσις KLNR : ἀλλοίωσιν καὶ αὔξησιν eF1[ἀφεξιν]F2J2Q : ἀλλοίωσις καὶ ἀφεξιν F0? [S] 261a24 μάλιστα ταύτην EbeFHIJQwΛ : ταύτην μάλιστα KLNR Pl[901.5-6] [S] 263b23 ὥστε beFHIJQwΛ Sp[1295.28]? : ὥστε ἢ KLNR : ὥστε εἰ ἦν EΨ 264a14 δὴ E2KLNRΛ : δ’ E1Ψ?beFHIJQw [S]

One could adopt the reading of KLNR at any of these places; indeed, Ross accepts it at 264a14. Equally seriously one should a fortiori take the readings shared by K and be and the β group: 259b23 261a5 261a33 261b1 263a26 263b1 264b3-4 264b29

συνεχῶς EFHIJQwΛ : συνεχῆ KbeLNR [Ψ] πρῶτον EΨFHIJQwΛ : προτέρον KbeLNR Sc[1270.28] εἰσιν αἱ EΨFHIJQw : εἰσὶ καὶ KLNR : εἰσὶ καὶ αἱ be : εἰσὶ T[227.8] [S]282 τὸ Α ἅμα γεγονέναι EF2Λ : ἅμα τὸ Α γεγονέναι KbeLNR : ἅμα γεγονέναι τὸ Α HIJQw [Ψ][F1] [S] ἡμίση EFHIJQw : ἡμίσεα KbeLNR [Λ] ἀριθμεῖν δύο EFHIJQwΛ : δύο ἀριθμεῖν KbeLNR Sp[1291.17-18] [Ψ] ἀλλοίωσις εἰς E2ΨFHIJQwΛ : ἀλλοίωσις ἡ εἰς KbeLNR : ἀλλοίωσεις E1 [S] οὐδὲ τὰς ἄλλας ἐνδέχεται KbeLNR Sp[1311.26] : οὐ τὰς ἄλλας ἐνδέχεται EΨ : οὐκ ἐνδέχεται οὐδὲ τὰς ἄλλας FHIJQwΛ [T]

_____________ 282

Clearly the reading of be is the result of the reading of KLNR with an extra αἱ inserted.

VI.6 Ms. Laur. gr. 87.24 (P.S. Hasper)

CLXI

I would adopt all of the readings supported by KbeLNR, also because they have the further support of Simplicius (261a5, 263b1 and 264b29) or Themistius (261a33), because they are paleographically most plausible (261b1) or because they constitute the lectio difficilior (263b3-4)283.

VI.7 Pachymeres’ paraphrase of the Physics In several manuscripts we find a paraphrase of the Physics which in most of those manuscripts is attributed to Michael Psellos. If this attribution were correct, the paraphrase would have been written around the middle of the 11th century and thus belong to the sources to be covered in this introduction. Now Psellos’ authorship has been vehemently denied by Golitsis, who has argued that instead Pachymeres is to be recognised as its author284. The main ground for Golitsis’ claim is that the version of the paraphrase we find in Laur. gr. 87.05 is the original work by Pachymeres, because the paraphrase has been written by Pachymeres himself, just as most of the text of the Physics which the paraphrase surrounds, and because the lemmata in the paraphrase clearly have been taken from the text of the Physics in this very manuscript, structuring the paraphrase in a somewhat arbitrary way, not dividing the text into units which necessarily have some unity in terms of content, but rather into units which more or less coincide with the division into pages of the text of the Physics in this manuscript285. By themselves these considerations do not decide the issue, because it still needs to be shown that the lemmata-structure has not been imposed onto an existing paraphrase in a completely arbitrary way. Indeed, there are places at which the paraphrase discusses certain issues before the lemma, while the issue itself in Aristotle’s text only comes after the lemma286. However, it can be shown that the paraphrase is indeed based on the text of the Physics as we find it in Laur. gr. 87.05. A first indication is that the text of the _____________ 283

The form ἡμίση for ἡμίσεα tends to be later. Pantelis Golitsis, “Un commentaire perpétual de Georges Pachymère à la Physique d’Aristote, faussement attribué à Michel Psellos”, Byzantische Zeitschrift 100 (2007), p. 637–676. 285 Subsidarily Golitsis argues that all the other manuscripts containing the paraphrase are (indirect) copies of Laur. gr. 87.05, but as he presents it, this argument seems to presuppose that Laur. gr. 87.05 provides the author’s text of the paraphrase, for he does not argue any more that the other manuscripts can only have been copies of that manuscript. 286 For example, at Pa[349.22] words from 249b7 appear, while the lemma only follows later; likewise at Pa[417.6-7] with the lemma from 265a18. 284

CLXII

Introduction

paraphrase after each lemma starts without a transitory particle, like καί and δέ, while up to the next lemma every sentence starts with such a particle. Decisive, however, is the point that the readings of the Physics featuring in the paraphrase almost always agree with those in the text of the Physics in Laur. gr. 87.05 (which I have given the siglum o), including readings which are not known from any other of the manuscripts collated for this edition: 259a21-22 ἐπιβλέψας ἐπὶ τὰς ἀρχὰς EKbeFHIJQwΛ Sc[1257.8]Pl[888.17] : ἀποβλέψας ἐπὶ τὰς ἀρχὰς o Pa[390.19-20] : ἐπὶ τὰς ἀρχὰς ἐπιβλέψας LNR 265b7 πως EKbeLNRFHIJQwΛ o2 : om. o1 Pa[419.12] 266b18 ἀντιστροφὴν EKbeLNRFHIJQwΛ o2 Pa[425.4-cett.]? : ἀναστροφὴν o1 Pa[425.4-L]287 266b21 ληψόμεθα KbeLNRFHIJQwΛ o1 Pa[426.7-cett]? : λη.ψόμεθα E : ληψώμεθα o2 Pa[426.7-L] 267b3 γὰρ EKbeLNRFHIJQwΛ : δὲ καὶ o Pa[428.20: δὲ]

That this list is not longer is solely due to the fact that for the Physics Laur. gr. 87.05 is very closely related to Q, though it is certainly not a direct copy of it, and may not even be an indirect copy288. I have found ten places at which the paraphrase shares a reading with Q and o alone among the manuscripts collated here, and about another ten at which merely one manuscript joins Q and o in this respect289. _____________ 287

The siglum L is the siglum given by Benakis to Laur. gr. 87.05 for the text of the paraphrase (cf. Benakis, Michael Psellos. Kommentar zur Physik, p. 48*–50*). For the two places on this list for which I mention the reading of the paraphrase in Laur. gr. 87.05, and leave open the possibility that the other manuscripts containing the paraphrase do not share that reading, Benakis’ collation of Laur. gr. 87.05 turned out to be incorrect. It must be noted that, even though I did not systematically check the text of the paraphrase in Laur. gr. 87.05 against Benakis’ collations, I have found numerous incorrect reports in the apparatus to Benakis’ edition of the paraphrase. 288 Strikingly, according to Rashed, Überlieferungsgeschichte, p. 116, also the paraphrase by Pachymeres of De generatione et corruptione (in the still unedited fourth book of his Philosophia) is based on Q or a manuscript closely related to Q. 289 There are also a few places at which the paraphrase goes against Q, but with Laur. gr. 87.05 and only one other manuscript collated here: 252b20 μηθὲν EKLNRFHIJQw : μηδὲν be o Pa[367.6] || 255b15 ποι EKbeLNRFHIJQ wΛ? o2? : που Ψ? o1? Sc[1216.30]Pa[376.21] || 256a24 αὐτὸς1 EKbeLNRF1HIJQwΛ Sc[1222.36] : αὑτῷ F2 o Pa[379.4] || 256a34 αὑτὸ EK1beLNRFHIJQwΛ o1? : αὑτῷ K2 o2 Pa[379.18] || 257a25 ἀεὶ EΨKbeLRFIJ2Q2wΛ Sp[1231.37, 1232.3] : αἰεὶ N : δεῖν J1?Q1 o : εἶναι ἀεὶ H : om. Sc[1231.36]Pa[382.13] || 257a25 κινούμενον EΨbeLNRFHIJ2QwΛ o1 : κινούμενον κινεῖσθαι J1 : κινοῦν K o2 Pa[382.13].

VI.7 Pachymeres’ paraphrase of the Physics (P.S. Hasper)

CLXIII

This settles the issue as to the authorship of the paraphrase. However, it can also be shown that when he wrote the paraphrase, Pachymeres did not merely rely on Laur. gr. 87.05. This appears from the many readings which seem to be presupposed by the paraphrase which do not feature in Laur. gr. 87.05, or generally in the γ group, but rather in other sources collated for the present edition. As it must have been natural for Pachymeres to use the available commentaries, we may surmise that quite a few of these readings derive from Simplicius: ταῦτα ἀδύνατα E[ταῦτ’]beLNR Sl[1170.21]Pa[363.19] : τοῦτ’ ἀδύνατον Ψ?KFHBIJQ[τοῦτο]wΛ o[τοῦτο] [T] 252a6 ταύτην δεῖ KbeHIJQwΛ : ταύτην δεῖν F ο : δεῖ ταύτην ELNR : ταύτην Ψ Sp[1183.23]Pa[363.23] [T] 252a16 οὐκέτι EbeLNRFHIJQwΛ o : οὐκ ἔστι K Sp[1184.31]Pa[364.18] : οὐδέν τι Ψ [T] 254b23 δὲ ELNRFHIJQwΛ o : δὲ καὶ Kbe Sp[1208.17] : καὶ Pa[373.25] [Ψ][T] 255a34 ἐνεργῶν EΨ1 Sp[1213.30-1214.8]T[219.12]?Pa[375.27] : θεωρῶν Ψ2Kbe LNRFHIJQwΛ o 255b4 τι μὴ ELNRFHIJQwΛ o : μὴ τι Ψ?Kbe Sc[1214.16, 25]Pl[830.31] T[219.5]Pa[376.6] 256a25 αὑτῷ Ψ? o2 Sp[1223.16]Pa[379.8] : αὐτῷ LN2HI2? : αὐτὸ EbN1RFw : αὑτὸ eI1?J2Q2Λ? o1 : om. KJ1Q1 Pp[Ra-16.5] 258a2 μόνως – εἶναι E2beLNRFHIJQ2wΛ o Sl[1240.22-a]Sp[1241.24-25], cf. As[635] : om. E1ΨK Sl[1240.22-AFM]Sp[1242.15]Pp[R-96v]Pa[385.3: non reddit] [Q1] 258b14 πάσης EΨLNR Sc[1251.17]Pp[837.35]Pa[388.1-2] : πάσης τῆς beFHIJ Qw o : καὶ πάσης K [Λ: sine καὶ] 260a6 ὑπὸ τοῦ κινουμένου μὲν e2FH1IJQwΛ o Pγρ[894.10] : om. EΨKbe1 LNRH0 Sc[1262.33, 1263.16]Pl[894.5]Pa[394.10-11] 264a21 οὐ μία EKLNR Sl[1301.31]Pa[411.26] : οὐκ ἕστι μία FHIJ[ἔστιν]QwΛ o : οὐ μία ἔστιν be T[230.28] [Ψ] [S] 265a21 συνθετὴ EKLNRFHIJQw o : σύνθετος be Sp[1314.13]Pa[417.18] [Ψ][Λ] [T] 265a25 ἐνδέχεται ἀΐδιον εἶναι EKbeLNR Sc[1314.38]As[776]Pa[418.8-9] : ἀΐδιον ἐνδέχεται εἶναι FHJQwΛ o : ἀΐδιον εἶναι ἐνδέχεται I [Ψ] [T] 265a26-27 στάσιν … κίνησις E1ΨK2e2LNRFHIJQwΛ o, cf. Sp[1314.36] : om. E2K1 be1 Sl[1313.21-1315.5]T[232.26-27 non reddit]Pa[418.10-11] 267b13-14 εἰ διαιρετὸς KH2Q ο : εἰ δὲ διαιρετὸς E2beLNRFIJΛ : εἰ δ’ ἀδιαίρετος w : αιρετὸς γὰρ E1 : εὐδιαίρετος γὰρ Ψ, cf. Sp[1356.12]As[824]Pp[851.16], cf. Pa[429.25] : .. διαιρετοῦ H1 [Pa(429.24): διαιρετὸς] 252a3

CLXIV

Introduction

However, many of the places listed here could just as well derive from a manuscript belonging to the χ1 branch of the stemma. There are other places as well, for which a joint reading between the paraphrase and one or several of the manuscripts EΨKbe can be established: 250b22-23 δ’ ἕνα ἢ μὴ FHIJQwΛ o : δὲ ἕνα ἢ μὴ Kbe : δ’ ἓνα καὶ μὴ Ψ Pa[358.2-3] : δὲ ἢ ἕνα ἢ μὴ E1LNR Sc[1122.26-28] : δὲ εἶεν ἀεὶ μὴ E2 [T(209.11-13) interpr. ἕνα ἢ ἀεὶ ἢ μὴ ἀεὶ, sed S quoque] 251a5 τούτων EbeLNR Sl[1125.25-a]T[209.22]Pa[359.8] : τούτου Ψ?KFHIJQ wΛ o Sl[1125.25-AFM] 251a16 ταῦτα EKbeLNRFHBIJQw o Sl[1127.16]T[210.10]? : ταῦτα τὰ πράγματα Ψ? Pa[360.7] : τοῦτο Λ 251b26 ἀεὶ FHIJQwΛ o P[823.20]? : om. EΨKbeLNR T[212.6] Pa[362.25] [S] 253a17 ἢ E2KbeLNRFHIJQwΛ o Sp[1190.35]?T[214.19] : om. E1 : καὶ Ψ? Pa[368.21] 253a33 τίς EKbeLNRFHIJQwΛ o : om. Ψ Pa[369.12] | 253b33 ἐξ ἀνάγκης EKbeLNRFHIJQwΛ o : om. Ψ Pa[378.16] [S] 255b8 καὶ2 EbeLNRFHIJQw o : καὶ τὸ K Pa[376.12] [Λ] 255b18 τε EK2LNRFIJQwΛ o : om. ΨK1beH Pa[376.23] 255b19 γὰρ KbeLNRFHIJQwΛ o : om. E : δὲ Ψ? Pa[376.25] 256b2 κινεῖται EKbeLNR Pa[379.19] : κινήσεται FHIJQwΛ o 256b13 εἶναι EKbeLNRFHIJQwΛ o : εἶναι ὥστε ἀναιρεῖται τὸ κινεῖσθαι Ψ? cf. S[1225.38-39: οὐδὲ περιλείπεται], cf. Pa[380.9-10: ὥστε ἀναιρεῖται] 257b6 ὅτι EKbeLNRFHIJQwΛ : τί vel τί τὸ Ψ Pa[383.13] 258a10 πλειόνων ΨFHIJQwΛ o Pp[836.9] : πολλῶν EKbeLNR Pl[836.6] Pa[385.18] [S] 258b11 ἀΐδιον ὃ πρῶτον κινεῖ o Sp[1250.35, 1251.3, 1251.28, 1254.32] : ἀΐδιον ὃ κινεῖ E2beLNR As[648, 649, 650]?Sp[1251.14]?T[223.14]Pa[387.26] : ὃ πρῶτον κινεῖ E1?ΨKFHIJwΛ : ὃ πρῶτον κινεῖ .. Q1 : ὃ πρῶτον κινεῖ ἀΐδιον Q2 258b26 δὴ E2KbeLNRFHIJQwΛ o : δὲ E1 Pa[388.26] [Ψ] 258b28 ταδί E2beLNRFHIJQwΛ o Sp[1252.26] : αὐτὰ E1Ψ[vel ταῦτα]K Pa[389.2] 259a19 ἄλλο EΨKbe1LNR Pa[390.16] : τι ἄλλο e2HIJQwΛ o : τι καὶ ἄλλο F 259b4 ἐγγίγνεσθαι LNR : ἐγγένεσθαι JQ o : ἐγγίνεσθαι beFHIw : γίνεσθαι E1Ψ?K Sp[1258.1]?Pa[391.14] : γίγνεσθαι E2 [Λ] 259b19 καὶ2 KbeLNRFHIJQwΛ o Sl[1256.32] : καὶ τὸ ἐν E1?Ψ Pa[393.2] : καὶ τὸ E2

VI.7 Pachymeres’ paraphrase of the Physics (P.S. Hasper) 259b23 260a6-7

261a16 261b2 261b25 262b8 262b11 263a11 263b11 263b16 265a1 265b15 266b12 266b17

CLXV

συνεχῶς EFHIJQwΛ o : συνεχῆ KbeLNR Pa[393.7] [Ψ] δὲ κινουμένου ἤδη E2e2FHIJQw o Pl[894.5] : δὴ κινουμένου ἤδη LNR : ἢ κινουμένου ἤδη Kbe1 Sc[1263.16-17] : κινουμένου ἤδη Λ : δὲ κινουμένου ἢ E1 : om. Ψ Pa[394.11] τοῦ ὀργάνου E1E3Ψe2LNRFHIJQwΛ : om. E2Kbe1 Pc[900.26] Pa[399.16] οὖν E2KbeLNR1FHIJQw o : δ’ E1ΨΛ Pa[400.12] : om. R0 γένοιτο E1KFH Pa[401.22] : γίγνοιτο E2beLNRJQ o : γίνοιτο Iw [Ψ][Λ] τοῦ E1Ψ?KLNRFHIJQwΛ o Sl[1280.18] : τὸ E2be Pa[404.28, 405.1] τοῦ ἄκρου EKLNRFHIJQwΛ : τοῦ Ε ἄκρου Ψbe Pa[405.5] τοῖς περὶ κινήσεως E2K2beLNRFHIJQw o : περὶ κινήσεως K1 : ταύτην τὴν ἀπορίαν Ψ add., cf. Sp[1290.20-21]Pa[407.3] : om. Λ [E1] ἅμα EKbeFIJQwΛ ο Pl[845.16] : om. ΨLNRH Pa[410.23-24] [S] μὲν Ke2LNRFHIJQwΛ ο : om. Ebe1 Pa[409.15]? [Ψ] θεῖναι E2bLNRFHIJQwΛ o Pl[848.12] : θεῖ.ναι E1 : προσθεῖναί K Pa[415.23] : εἶναι e μόνης EKbeLNRFHIJQwΛ o : om. Ψ Pa[419.26] ΖΘ EΨK o2 As[802]Pa[424.19] : Ζ καὶ Θ beLNRFHIJQwΛ o1 : ΘΖ Sc[1342.39] τὸν E2KbeLNRFHIJQw : om. E1 Pa[424.25] [Ψ][Λ] [S]

Now it would be foolish to claim that at all these places there is a shared reading between the paraphrase and the χ1 side of the stemma; many may be due to coincidence. However, it is difficult to believe that all of them are accidental; especially 250b22-23, 251a16, 253a33, 253b33, 255b19, 256b13, 257b6, 258b11, 258b28, 259b4, 259b19, 260a6-7, 261a16, 261b2, 261b25, 262b11, 263b11 and 265b15 seem certainly for the most part non-accidental. Particularly striking are the places at which the paraphrase and the Arabic translation share a reading uniquely: 250b22-23, 251a16, 253a33, 253b33, 255b19 (perhaps), 256b13, 257b6, 260a6-7 and 265b15. The number of places, but also the kind of parallels, seem to suggest a significant connection. We can only guess at what might explain this connection, but the odds would seem to be some kind of commentary which has influenced both the exemplar of Arabic translation and Pachymeres while composing the paraphrase. It would perhaps be worth our while to check scholia to Aristotle’s Physics as to whether there are any similar traces to be found in them.

CLXVI

Introduction

VI.8 Some remarks on the stemmatic relation between Simplicius and the direct tradition It is difficult to determine the place within the stemma of the primary text read by Simplicius for his commentary, primarily because of the paucity of evidence, but also because of the nature of a commentary, both involving an unclear boundary between citation and paraphrase and making the correction of small errors along the way very natural. Moreover, since commentaries often were transmitted along with the text they are commentaries on, there is always the danger of interference between the two. Finally, in the case of Simplicius’ commentary, it is not completely clear whether the lemmata are a later addition or were written by Simplicius himself – though if they are a later addition, they must have been drawn from a text which is very close to the text cited, or presupposed by Simplicius in the actual commentary290. Still, it is important to present the available evidence here, because Simplicius’ text is at least three centuries older than that of the oldest extant manuscripts. Thus there has been less time for contamination to enter. It is Diels’ claim, accepted by Ross, that Simplicius’ manuscript is independent from the whole extant manuscript tradition. If this were true, one should in principle adopt any consensual reading shared by Simplicius and any other branch of the stemma, or, alternatively, try to explain such a consensual reading by positing contamination in either direction. As already pointed out, Diels’ claim is based on the fact that at Physics IV 4, 212a2-7 Simplicius (as well as Themistius and perhaps Philoponus291) reads: ὁ τόπος ἐστίν, μήτε τὸ εἶδος μήτε ἡ ὕλη μήτε διάστημά τι ἀεὶ ὑπάρχον ἕτερον παρὰ τὸ τοῦ πράγματος τοῦ μεθισταμένου, ἀνάγκη τὸν τόπον εἶναι τὸ λοιπὸν τῶν τεττάρων, τὸ πέρας τοῦ περιέχοντος σώματος καθ’ ὃ συνάπτει τῷ περιεχομένῳ. λέγω δὲ τὸ περιεχόμενον σῶμα τὸ κινητὸν κατὰ φοράν. whereas all the extant manuscripts leave out the clause in bold. There is no doubt about the fact that Simplicius read this clause in his manuscript, for at 584.19-20 he says: …, ὅπερ πρότερον ἐσήμανε διὰ τοῦ καθ’ ὃ συνάπτει τῷ περιεχομένῳ. _____________ 290

A possible indication that the lemmata were added later is the fact that both the lemma at 1264.12-13 and that at 1265.9 feature the text of 260b7. Diels, in the apparatus to 1264.12-13, claims not to know whether this mistake is Simplicius’ or a scribe’s; see also above, p. XCIII, footnote 212. 291 There is certainly room for doubt whether Philoponus, In Phys., p. 567.5 paraphrases this clause, as Ross does claim.

VI.8 Simplicius and the direct tradition (P.S. Hasper)

CLXVII

Now Diels thought that this clause is absolutely required, for two reasons. First, without the clause it is not stated that place is the innermost boundary of the surrounding body, and that there should not be any space between this innermost boundary and the boundary of the surrounded body. Second, the phrase τὸ περιεχόμενον σῶμα in the elucidating phrase would, according to Diels, presuppose an earlier reference to a surrounded body292. Neither reason convinces, however. With the phrase τῶν τεττάρων Aristotle refers back to an earlier list of four candidates for the definition of place at 211b6-9, where he does mention the condition that there not be any space between, and also there it is understood that it concerns the innermost boundary; he thus can refer to this candidate definition in a brief way, just as he refers to the other three in an abbreviated way at the beginning of the quoted passage. Moreover, if there is a surrounding body, there must be a surrounded body; thus a reference to a surrounded body is quite appropriate. Diels seems to assume, incorrectly, that the phrase τὸ περιεχόμενον σῶμα in λέγω δὲ τὸ περιεχόμενον σῶμα τὸ κινητὸν κατὰ φοράν functions exclusively as a quotation, as if it were some entry in a dictionary. A possible explanation for the fact that Simplicius read the clause in his manuscript is that it derives from Alexander’s commentary, just as at 258a2 the addition μόνως γὰρ οὕτως οἷόν τέ τι αὐτοκίνητον εἶναι, which also appears in Simplicius’ paraphrase at 1241.24-25, stems from Alexander’s commentary293. The fact that also Themistius has the clause may equally be explained by reference to Alexander, because Themistius as well is strongly influenced by Alexander’s commentary. Does this mean that there is no evidence at all suggesting that Simplicius’ exemplar is independent from the extant manuscript tradition? Such a conclusion would be rash as well, for at Physics III 1, 201a32 all manuscripts read: ἔστι γὰρ ὁ χαλκὸς δυνάμει ἀνδριάς, ἀλλ’ ὅμως οὐχ ἡ τοῦ χαλκοῦ ἐντελέχεια, ᾗ χαλκός, κίνησίς ἐστιν· οὐ γὰρ τὸ αὐτὸ χαλκῷ εἶναι καὶ δυνάμει τινὶ κινητῷ, ἐπεὶ εἰ ταὐτὸν ἦν ἁπλῶς καὶ κατὰ τὸν λόγον, ἦν ἂν ἡ τοῦ χαλκοῦ, ᾗ χαλκός, ἐντελέχεια κίνησις· but both the verbatim identical version in Metaphysics K 9, 1065b26 and Simplicius’ quote (at 424.17) omit κινητῷ. Now κινητῷ seems clearly out of place, and the result of the incorporation of a gloss on δυνάμει τινί, which must have been present in the common ancestor of all extant manuscripts, but probably not in the manuscripts as excerpted in Metaphysics K and as used by Simplicius. It is not to be excluded that the excerpter would have corrected its source along the _____________ 292 293

Diels, Zur Textgeschichte, p. 4–5. See above, p. CXXII.

CLXVIII

Introduction

way, but with Simplicius this seems less likely given the frequency with which he comments on the readings he finds in the manuscripts294. Thus this passage provides at least indicative evidence that Simplicius’ main manuscript is independent from the whole extant manuscript tradition. One would, however, like to have more of such passages – we need a thorough examination of the whole of the Physics in order to establish such an independence definitely. There is some reason even to expect that Simplicius’ commentary shows a certain independence from the extant manuscript tradition. This is constituted by the fact that Simplicius frequently reports on the different readings he finds in the manuscripts at his disposal. On the one hand, this evidence shows that the main groups of manuscripts in the extant tradition, or at least readings demarcating these groups, were known to him, thus suggesting that the process of differentiation between the different groups of manuscripts in the extant tradition was at least already well on its way in the sixth century. On the other hand, Simplicius thus has the opportunity to choose between the different readings he is familiar with. Sometimes he informs us about his choice, but it is not to be excluded that he sometimes does not. In his commentary on Physics VIII there are three such reports: 255b5

καὶ ἐν EΨbeLNRHQ2 Aγρ[apud S(1214.38)]Sγρ[1214.35: ὡς ἐν τοῖς πλείστοις ἀντιγράφοις] : καὶ KFIw : καὶ οὐχ ἁπλῶς ἐν As[593] A[apud S(1214.26-27)]Sp[1214.30] : καὶ οὐκ ἐν JQ1 : om. Λ

_____________ 294

Rashed, Alexandre d’Aphrodise, p. 389–390, suggests in fact that at Physics VI 6, 236b 34-37 we find another indication of the independence of Simplicius’ commentary from the whole extant manuscript tradition. Ross reads there: εἰ γὰρ ἐν τῷ ΧΡ πρώτῳ χρόνῳ τὸ ΚΛ κεκίνηται μέγεθος, ἐν τῷ ἡμίσει τὸ ὁμοταχῶς κινούμενον καὶ ἅμα ἀρξάμενον τὸ ἥμισυ ἔσται κεκινημένον.

All the manuscripts used in the present work feature μέγεθος, which is indeed most easily interpreted as specifying that τὸ ΚΛ is the magnitude the moving object has moved over. However, Simplicius (In Phys., p. 993.19-20) assumes that τὸ ΚΛ refers to the moving object, while Alexander (Rashed’s scholium 366), and Philoponus (In Phys., p. 810.15-16), assume that τὸ ΚΛ need not refer to the magnitude moved over. Rashed’s conclusion is that none of them read μέγεθος and that μέγεθος started its presence in the extant manuscript tradition as a gloss. I do not think this conclusion is warranted: even in the presence of μέγεθος one could wonder whether τὸ ΚΛ refers to the moving object or to the magnitude moved over, even if one thinks, as we nowadays do, that it is quite obvious that τὸ ΚΛ refers to the magnitude moved over. Presumably the most important ground for our understanding is that in the apodosis there is a subject to the verb, τὸ ὁμοταχῶς κινούμενον καὶ ἅμα ἀρξάμενον, which makes a subject τὸ ΚΛ in the protasis rather awkward.

VI.8 Simplicius and the direct tradition (P.S. Hasper)

CLXIX

258a27 τὸ δὲ … γάρ KbeLNRFHIJQwΛ Sγρ[1245.2-3: ἔν τισι]Pa[386.16-18]? : om. EΨ Sγρ[1245.3-5: τὰ πλείονα τῶν ἀντιγράφων]A[apud S(1245.3-5)] 266b33 κινεῖσθαι EKb2FJwΛ : κεκινῆσθαι eLHIQ Scp[1345.1-aA, 1345.2-Exc., γρ. 1345.9]Pa[426.15] : καὶ κινεῖσθαι ΨNR Scp[1345.1-Exc., 1345.2-aA, 1345.12, 1345.21, γρ. 1345.10] [b1]

This evidence for 255b5 is somewhat ambiguous: it seems as if Simplicius prefers the reading he finds in Alexander, and he implies that at least one manuscript available to him features this reading; he is, however, aware that it has become the minority reading. It could mean that his main manuscript has this reading and is thus part of a manuscript tradition of Aristotle’s Physics which was already present at the time of Alexander, but which was being crowded out by the manuscript tradition available to us. It is, however, also possible that his main manuscript is, regrettably, among ‘the many’. Generally Simplicius does not betray in such cases which reading he finds in his main manuscript295. The third passage referred to here shows that Simplicius’ primary reading was καὶ κινεῖσθαι296 and that it is retained in part of the α and β groups, to wit Ψ and NR. Simplicius knows the alternative κεκινῆσθαι, which can be understood as a corruption of καὶ κινεῖσθαι (> καὶ κινῆσθαι > κεκινῆσθαι). The chances are that this reading is the original reading of the γ group, for were it not, we would have to postulate contaminations in eLHIQ without having a source among the extant manuscripts, and in addition a separate disappearance of καὶ elsewhere. A scenario according to which the γ group featured originally the corruption κεκινῆσθαι, which then was corrected into κινεῖσθαι in some of these manuscripts (and in E), seems more plausible. According to this scenario Simplicius’ report about κεκινῆσθαι would in fact be a report about the γ group. Also from his comments on Physics VI 1, 232a20, it appears that Simplicius knew a reading which is distinctive of the γ group, though it later must also have contaminated the α group: 232a20 πᾶσα/ἅπασα διαιρετός E2KbeF1H2I2J2Q Sc[936.4] : πᾶσα διαιρετὴ LNRF2 : πᾶς/ἅπας διαίρετος E1Ψ?H1I1?J1w A[apud S(936.22)]Sγρ[936.1-ἔν τισι] : πᾶσα ἀδιαίρετος Λ : πᾶς ἀδιαίρετος Iγρ Asp[apud S(936.30)]

The reading πᾶς/ἅπας must have been the reading of γ, whereas the fact that Kbe have πᾶσα/ἅπασα suggests it used to be representative of the whole χ1 branch. _____________ 295

A nice case one can find in In Phys., p. 422.10–423.23. The manuscripts of Simplicius’ commentary are divided at 1345.1-2, but from 1345.12 and 1345.21 it appears that he read καὶ κινεῖσθαι. The variation at 1345.1 and 1345.2 does show that the manuscripts containing Simplicius’ commentary are not impervious to influence from the manuscript tradition of the Physics.

296

CLXX

Introduction

What we learn from the evidence at 258a27, it seems, is that the χ1 branch in Simplicius’ time was the numerically dominant branch, and that the χ2 branch was in the minority, a state of affairs which must at a certain moment have turned around, though we do not know exactly when. That the χ1 branch was numerically dominant is compatible with the α group being a small group – as we have seen, the δ group, K and to some extent even Λ show that the χ1 branch is or at least used to be broader than just the α group. That the α group was a minority group known to Simplicius is suggested by another report by him: ἔν τισι δὲ καὶ τοῦτο πρόσκειται τῆς γραφῆς ἐχούσης οὕτως· ἢ γὰρ ἄμφω ἐν τῷ πάσχοντι καὶ ποιουμένῳ ἢ ἐν τῷ ποιοῦντι καὶ διατιθέντι, ἢ ἡ μὲν ποίησις ἐν τῷ ποιοῦντι ἡ δὲ πάθησις ἐν τῷ πάσχοντι. (In Phys., p. 441.30-33) This addition we find at 202a26 in the α group alone. On the other hand, there are also places for which Simplicius reports that the reading we only find in the α group are present in at least many manuscripts; here is an example: 226b34 ἀρχὴν EΨ M Sl[874.21-ACM]Sc[876.23-πολλῶν ἀντιγράφων] : ἀρχὴν μόνον beLNRFHIJQΛ A[apud S(876.24)]Sl[874.21-aF]

This apparent inconsistency can be explained by reference to the contamination we have established in the other manuscripts belonging to the χ1 branch. A third way, finally, to determine how Simplicius’ exemplar is related to the extant manuscript tradition is to check whether there are any joint errors or otherwise rejectable readings in Simplicius shared with parts of the extant tradition. Next I will go through lists of possible candidates Simplicius shares with the manuscripts from the χ1 branch (the α group, the δ group and K), the β and γ groups together, and finally the β group and the γ group separately. It is immediately clear that in his quotes and paraphrases as well as in the lemmata Simplicius shares many readings with the χ1 branch of the stemma, without support from manuscripts belonging to the β or γ groups. For the most part these readings are unproblematic, as also appears from the fact that they are adopted by Ross. However, there is an impressive list of places297 where Ross decides against the consensus of Simplicius and manuscripts from the χ1 branch (though for quite a few of these places Ross is either not aware of the readings in Ψ or be or has not checked Simplicius’ commentary so carefully, for otherwise _____________ 297

For obvious reasons I have left out most of the places at which the manuscripts of Simplicius are divided or at which Simplicius himself is inconsistent; I have only mentioned a few at which the Aldine (siglum a, ed. Diels) is set against the other manuscripts featured in the apparatus to Diels’ edition of Simplicius’ commentary, because of the likelihood that the Aldine was there contaminated.

VI.8 Simplicius and the direct tradition (P.S. Hasper)

CLXXI

he might even have adopted these readings). Most of them concern places at which I would happily accept the reading shared by Simplicius and one of these manuscripts298: 251a4-5

τὸ γὰρ ᾗ δὲ τάδ’ ἀλλάσσοντα ἐνθένδε ἐκεῖσε λέγειν αὐτὸν ὑποληπτέον LNRFHIJQwΛ : δεῖ γὰρ ὑπολαβεῖν λέγειν αὐτὸν ᾗ δὲ ταδ’ ἐνθένδε τὰ ἀλλάσσοντα EΨ[τὰ om.?]K[ἀλάσσοντα]b[τῇ, ἄλλασονται ἐνθένδε ἐκεῖσε] e[τῇ, ἄλλασονται ἐνθένδε ἐκεῖσε] Slp[1123.8, 1125.5 – τὰ om.] [T] 251b18 μὲν EeLNRFHJQw : om. KbIΛ Sl[1154.1]Sc[1155.36] [Ψ] [T] 252a6 ταύτην δεῖ KbeF[δεῖν]HIJQwΛ : δεῖ ταύτην ELNR : ταύτην Ψ Sp[1183.23]Pa[362.23] [T] 252a26 τοῦτ’ LNR : τοῦτο KbeFHIJQwΛ : ταὐτ’ E1 : ταὐτὸ E2 : ταὐτὸ τοῦτο Ψ : ταὐτὸν Sp[1185.31] [T] 254b23 δὲ ELNRFHIJQwΛ : δὲ καὶ Kbe Sp[1208.17] : καὶ Pa[373.25] [Ψ] [T] 255b4 τι μὴ ELNRFHIJQwΛ : μὴ τι Ψ?Kbe Sc[1214.16, 25]Pl[830.31]T[219.5] Pa[376.6] 257a3 ἢ E1ΨK2LNRFHIJQwΛ : εἰ δ’ E2[δὲ]K1be Sl[1229.6] 257b3 φέροιτο … φέροι KLNRFHIJQwΛ : φέροι … φέροιτο EΨbe Sc[1234.1213, 1249.27-28]T[221.29] 257b4 καὶ1 E2ΨbeLNRFHIJQwΛ Sc[1249.28] : ἢ E1 Sc[1234.14] : ἢ καὶ K 258a12 ἅπαν KbLNRFHIJQw : ἅπαν τὸ EeΛ Sp[1242.38]Pa[385.22]? [Ψ] 258b4 πρώτως2 e2LNRFHIJQw : πρῶτον EΨ?Kbe1Λ? Sl[1247.11] 258b8 πρώτως κινοῦν beLNRFHIJQwΛ Sl[1247.11-a]Pp[887.6] : πρῶτον EΨK Sl[1247.11-AFM] 263a1 ἄρα LNRFHIJQw : om. EΨKbeΛ Sc[1288.25]Sp[1288.13-16] 263b26 ὄν2 E2KbeLNRFHIJQwΛ : om. E1Ψ Sp[1297.16]PA[907.24]? 265a21 συνθετὴ EKLNRFHIJQw : σύνθετος be Sp[1314.13]Pa[417.18] [Ψ][Λ] 265a26-27 στάσιν – κίνησις E1ΨK2e2LNRFHIJ[partim]QwΛ, cf. Sp[1314.36] : om. E2K1be1 Sl[1313.21-1315.5]T[232.26-27 non reddit] 265b26 καὶ beLNRFHIJQwΛ : om. EΨK Sc[1319.7] 266b9 ἣ ERFHIJQΛ : ἢ beLNw : ᾗ ΨK Sp[1341.24] 267a4 τοιοῦτον2 beLNRFHIJQwΛ : om. EΨK Sp[1345.30]?Pa[426.20] 267b14 ἄλλος F2 : ἄλλως E1?Ψ Sp[1356.13]Pc[851.16] : ἀλλ’ ὡς beLNRF1HIJ QwΛ : ἄλλον E2K299

_____________ 298

Here I ignore the five places mentioned above (section VI.6) at which Simplicius only agrees with K, as the best hypothesis seems to be that they concern cases of influence from Simplicius on K. 299 πῶς δὲ κινεῖ ὁ ἀήρ; ἄλλος ἀεὶ καὶ ἄλλος κινούμενος καὶ οὕτως κινῶν at Simplicius, In Phys., p. 1356.13-14 strongly suggests that Simplicius also read ἄλλως.

CLXXII

Introduction

There are also a number of places on this list for which we have only Simplicius’ paraphrastic reference as evidence and this paraphrase is only weakly indicative as to what he read300: 252a16 οὐκέτι EbeLNRFHIJQwΛ : οὐκ ἔστι K Sp[1184.31]Pa[364.18] : οὐδέν τι Ψ [T] 256b23 ἵνα EKbeLNRFHIJQΛ : εἰ καὶ Ψ? Sp[1227.24-25, 32] 259b4 ἐγγί(γ)νεσθαι beLNRFHIw : ἐγγένεσθαι JQ : γίνεσθαι E1Ψ?K Sp[1258.1]? Pa[391.14] : γίγνεσθαι E2 [Λ] 260a6-7 δὲ κινουμένου ἤδη E2e2FHIJQw Pl[894.5] : δὴ κινουμένου ἤδη LNR : ἢ κινουμένου ἤδη Kbe1 Sc[1263.16-17] : κινουμένου ἤδη Λ : δὲ κινουμένου ἢ E1 : om. Ψ Pa[394.11] 260b17 λέγεται δὲ EKLNRFHIJQwΛ : λέγοιτο δὲ ἂν Ψ? Ap[apud S(1268.3)] : λέγεται γὰρ be Sp[1267.36] 265b13 καὶ EKbeLNRFHIJQw : om. Ψ Sp[1317.17] 267b14 ἢ E2KbeLNRFHIJQwΛ Pp[851.16] : καὶ E1Ψ Sp[1356.12]

Then there are five places at which Simplicius merely agrees with be, against most of the further extant tradition: ἔχει E1ΨKLNRFHIJQwΛ : ἔχειν E2be Sl[1183.20] [T] τῆς κινήσεως EΨKLNRFHIJQwΛ : τῶν κινήσεων be Sc[1208.18] οὐκ ἀεὶ E1ΨKe2LNRFHIJQwΛ : οὐ E2be1 Sp[1258.29-30]T[224.15] τὸ πλάγιον EΨKLNRFHIJQwΛ Sl[1279.18-a]Pl[842.19] : τὰ πλάγια be Sl[1279.18-A]Sp[1280.13]Pp[842.20]T[228.14] 267b18 ἀκίνητον EΨKFIJQwΛ : ἀκίνητόν ὂν beLNRH Sp[1358.4]

252a19 254b24 259b14 262a12

Strikingly, at three of these places this consensus between Simplicius and be is supported by further late ancient commentators, Themistius and/or Philoponus301. Moreover, at 259b14 and 267b18 this consensus gives a better text than the alternative302, though at 252a19 it is the other way round (though it is not impossible). _____________ 300

This is also the case for a place at which Ross does accept the consensus of E and Simplicius (though he does not mention Simplicius in the apparatus, and apparently collated J incorrectly): 263b31 τινὰ εἶναι E Sp[1297.27] : εἶναι τινὰ KbeLNRFHIJQw Ap[apud S(1299.8)] : εἶναι τι Λ [Ψ]. The support for the order εἶναι τινὰ is much better, and it is also the lectio difficilior. 301 For 252a19 this support is indirect, but still significant. For be read at 252a17-19: ἢ γὰρ ἁπλῶς ἔχειν δεῖ τὸ φύσει, καὶ οὐχ ὁτὲ μὲν οὕτως ὁτὲ δ’ ἄλλως, ...· ἢ λόγον ἔχειν τὸ μὴ ἁπλοῦν, where the alternative readings are ἔχει and ἔχει. We find ἔχειν δεῖ in Themistius’ paraphrase as well, while Simplicius’ lemma reads λόγον ἔχειν, a reading only possible with the earlier ἔχειν δεῖ.

VI.8 Simplicius and the direct tradition (P.S. Hasper)

CLXXIII

Finally, there are a number of places which each individually would probably betray carelessness or constitute merely coinciding errors:302 257a29 τὸ αὐτὸ αὑτὸ LNRFHIJQwΛ : τὸ αὐτὸ αὐτὸ E1K : τὸ αὐτὸ E2Ψbe : τὸ αὑτὸ Sc[1247.31] 257b2 αὑτὸ κινοῦν ΨeLNRI Sl[1233.34-a] : ἑαυτὸ κινοῦν FHJQ : αὐτὸ κινοῦν Eb2w : κινοῦν Kb1 Sl[1233.34-AFM] : κινοῦν αὑτὸ Λ303 258a4 κινηθήσεται Kbe2LNRFHIJQw Sl[1241.28] : κινήσεται Ee1 Sc[1224.32] [Ψ][Λ] 258b30 ἔχειν EKLNRFHIJQwΛ Sc[1253.5] : ἔχον be Sc[1252.33] 259a16 καὶ γὰρ τὸ EKbeLNRFHIJQwΛ : τὸ γὰρ Ψ Sc[1256.13] 260b18 τε EKbeLNRFHIJQwΛ : om. Ψ? Sc[1267.36]Pa[397.3] 262b2 καὶ ἄλλῳ ΨKbeLNRFHIJQwΛ Sp[1283.13]Pp[843.16] : om. E Sc[1283.19] 265b26 ὡς EKbeLNRFHIJQwΛ : om. Ψ* Sc[1319.7]

For most of these places there is only one manuscript involved, and for the remaining places the mistake is common. Thus this list does not constitute evidence of joint errors between Simplicius and the manuscripts from the χ1 branch304. Most places at which Simplicius shares a reading with the other side of the stemma, the consensus of the β and γ groups, concern errors or less preferable readings in the manuscripts from the χ1 branch: 251b10 τὸ KLNRFHIJQw Sl[1152.20] : τε E2 : om. E1be [Ψ][Λ] [T] 252a7 ὅπερ E2KbeLNRFHIJQw Sp[1183.24]? : ὥσπερ E1ΨΛ [T] 252b25 τὸ αὐτὸ KLNRFHIJQwΛ Sp[1188.3] : τοῦτο τὸ αὐτὸ EΨ : τὸ αὐτὸ τοῦτο be

_____________ 302

At 259b14 it is rather awkward to say διὸ οὐκ ἀεὶ κινοῦνται συνεχῶς ὑφ’ αὑτῶν, as in the context ἀεί and συνεχῶς convey exactly the same point. At the same time one could explain how ἀεί was added to the text, for example as a gloss on the less common συνεχῶς. At 267b18 the disappearance of ὂν in φανερὸν ὅτι ἀδύνατον τὸ πρῶτον κινοῦν καὶ ἀκίνητον ὂν ἔχειν τι μέγεθος, as beLNRH and Simplicius read it, is easier to explain than its addition. 303 Rashed, in his comments to scholium 625, speculates that Alexander’s αὐτοκίνητον is indicative of τὸ αὑτὸ κινοῦν rather than τὸ αὐτὸ αὑτὸ κινοῦν, also because it would be better Greek. It does not seem to be Aristotle’s Greek, however, for the phrase αὐτὸ αὑτὸ with caseforms of the participle κινοῦν are very common in Physics VIII: it can be found at 254b32, 257b13, 257b14, 257b27, 257b31, 261a26 and 265b33. 304 There is also the following place: 255b15 ποι EKbeLNRFHIJQwΛ? : που Ψ? o1? c S [1216.30]Pa[376.21], which probably provides evidence for influence from Simplicius on the Arabic translation or its exemplar.

CLXXIV 254b21 255b14 258a17 258b10 259a18 260b2 260b29 264a3 265a31 266b11

Introduction κινεῖται EK2LNRFHIJQwΛ Sc[1207.12] : κινεῖσθαι K1be [Ψ] αὑτῶν KLNRF2HJQ2Λ Sp[1216.29] : αὐτῶν Iw : αὐτοῦ F1Q1? : αὐτὸν EΨbe [T] ἄρα Ke2LNRFHIJQw Sp[1243.6] : γὰρ EΨbe1?Λ διαλείπειν E2Ψ?beLNRFHIJQw Sl[1250.32] : διαλίπειν E1K [Λ] καὶ KbeLNRFHIJQw Sc[1256.30] : καὶ ὑφ’ EΨΛ ἐνεργείᾳ EKbe : τὸ ἐνεργείᾳ ΨLNRFHIJQw Sp[1265.31] [Λ] πρώτην KLNRFHIJQw Sl[1270.1] : πρώτη EΨ?be : πρότεραν vel πρότερα Λ οὖ E2beLNRFHIJQwΛ Sc[1298.14, 35] : ᾦ K : om. E1Ψ οὗ KLNRFHIJQwΛ Sp[1315.14]T[232.29] : οὐ EΨbe χρόνῳ E1ΨH : κινήσει χρόνῳ E2KbeLNR[κινήσει bis]IJQwΛ Sc[1342.34]: χρόνῳ κινήσει F305

There are also places where both alternatives are equally possible (even though the χ1 reading is perhaps preferable): προάγουσιν EΨK1 T[211.2] : προϊοῦσι K2beLNRFHIJQwΛ Sp[1128.2] στέρησις KLNRFHIJQw Sc[1128.17] : στέρησις τῆς Ebe [Ψ][Λ] [T] παύεται EKbe1Λ : παύσεται e2LNRFHIJQw Sc[1169.26] [Ψ] [T] χρόνῳ κινήσει EK : χρόνῳ R : κινήσει χρόνῳ beLNFHIJQwΛ Sp [1343.11] [Ψ κινήσει hab.] 266b23-24 ἐν τῷ ἀπείρῳ πεπερασμένην δύναμιν beLNRFIJQwΛ Sp[1343.23-24] : πεπερασμένην δύναμιν τὴν ἐν τῷ ἀπείρῳ E[τὴν om.]Ψ*[τὴν om.]?K 266b26 ἐν ἀπείρῳ πεπερασμένην EΨ*K : πεπερασμένην ἐν ἀπείρῳ beLNRF[ἐν om.]HIJQwΛ Sp[1344.3] 267a20 τι ἓν E1K : ἓν τι E2beLNRFHIJQwΛ Sp[1351.19] [Ψ] 267b7 ἐν E1Ψ?FIΛ Sl[1353.35-a]T[236.2]Pa[429.9] : om. E2KbeLNRHJQw Sl[1353.35-A] 251a22 251a27 251b32 266b18

There are no places where the consensus of Simplicius and the β and γ groups features errors. There are not many places at which there is a reading shared by Simplicius and the β group alone or virtually alone: 258a31 τὸ EbeFHIJw : om. KLNRQ Sc[1245.25] [Ψ] 258b10 ἀεὶ εἶναι E1beFIJQwΛ Sl[1250.32-a]T[223.14] : αἰεὶ εἶναι E2 : εἶναι ἀεὶ KLNRH Sl[1250.32-AFM] [Ψ] 261b15 χρήσιμον EKbeFIJQwΛ : χρήσιμος LNRH Sc[1275.41]

_____________ 305

This seems to be a case where saut du même au même in ἐν ἡμίσει κινήσει χρόνῳ occurred twice, both in α and in the common ancestor of FH.

VI.8 Simplicius and the direct tradition (P.S. Hasper)

CLXXV

For 258a31 this consensus seems to constitute the better reading, for 258b10 there is, apart, perhaps, from the support of Themistius for ἀεὶ εἶναι, no real ground for a preference either way, and for 261b15 one would surmise the reading χρήσιμος to be a case of smoothing. There are more places at which there is a reading shared by Simplicius and the γ group alone or virtually alone. Again, for some of them we have only Simplicius’ paraphrase so that it may be only weakly indicative of what he read: 258a17 κινεῖσθαι ὑπ’ ἄλλου EKbeLNRΛ : κινεῖσθαι ὑπ’ ἄλλου κινεῖσθαι J : ὑπ’ ἄλλου κινεῖσθαι HIFQw Sp[1243.6] 260a13 κινεῖται ἀεὶ EK[αἰεὶ]beLNR : ἀεὶ κινεῖται FHIJQwΛ Sp?[1264.1] [Ψ] 260a13 δ’ ἀεὶ ΨKFHIJQ[δὲ]wΛ Sp[1264.1] : δὲ EbeLNR 262b5-6 τελευτῇ καὶ ἀρχῇ EΨKbeLNRΛ Plp[843.20, 25-26]? : ἀρχῇ καὶ τελευτῇ F [-ὴ,-ὴ] HIJQw Sp[1283.25]Pp[843.22-23]? 265a29 ἐπ’2 FHIJ2QwΛ Sp[1315.9] : om. EΨKbeLNR T[232.28] [J1]

However, there are also five places where the reading shared by Simplicius and (part of) the γ group for which the evidence is far more solid: 251b30 εἶναι τινα ELNR T[212.11-M] : τινὰ εἶναι KbeFHIJQwΛ Sc[1174.11] : εἶναι T[212.11-CSL] [Ψ] 261a34 τὸ1 … τὸ2 EKbeLNR : τὸ1 … FHIJQw Sc[1274.9] : om. Ψ? [Λ] 262a14 ἐπ’ EKbeLNR : ἐπὶ τῆς Ψ?FHIJQwΛ Sc[1280.27]Sp[1280.23] 263b21 οὐ E2ΨKbeLNRFHΛ : om. IJQw Sp[1295.26] : καὶ οὐ E1 264a32 τῷ KeF2HIJQΛ Sc[1304.28] : τὸ EΨbLNRF1w

To these one may add three places where the manuscripts for Simplicius’ commentary are divided, but the majority reading is shared by the γ group, against the consensus of the α and β groups as well as Themistius: δὴ KbeFHIJQwΛ Sl[1125.25-aAM] : δὲ EΨLNR Sl[1125.25-F]T[209.22] τούτων EbeLNR Sl[1125.25-a]T[209.22]Pa[359.8] : τούτου Ψ?KFHIJQwΛ Sl[1125.25-AFM] 251b19 καὶ EΨKbeLNR Sl[1156.4-a]T[212.2] : om. FHIJQwΛ Sl[1156.4-AFM] Sp[1166.33] 251a5 251a5

There are even errors among them: 263b21 is really wrong, and 264a32 can, once one understands the purpose of the accusative (see footnote 258), be seen as a mistake. There is even one passage where Simplicius must have agreed in error with the γ group on purpose, a passage which has already been discussed above (p. CXXVII–CXXVIII):

CLXXVI 265b7

Introduction τὸ μένειν E2K[τῷ]beLNR T[233.4, 6] : μένειν HΛ : μένει E1ΨFHIJQw Sp[1316.19]

This is a striking passage, for given the argument that the reading τὸ μένειν is to be adopted, both the α group and Simplicius are wrong, and since Simplicius knew Themistius’ paraphrase, he must have decided against τὸ μένειν. Thus it seems that Simplicius, while agreeing most of the time with manuscripts from the χ1 branch of the stemma, shared some dubious readings with the γ group. Now the places at which Simplicius shares its reading either with the manuscripts from the χ1 branch or with the β and γ groups together may serve as evidence that Simplicius’ exemplar is not completely independent from the whole extant manuscript tradition, but they do not undermine the basic stemma set up at the beginning of this section. The places where Simplicius shares a reading with the γ group alone, however, do pose a threat to that stemma. On the one hand, Simplicius’ exemplar does not share any of the errors which unite the β and γ groups and is thus to be placed above the split between the two groups. On the other, with these readings, which are not shared by both the χ1 branch and the other sub-group of the χ2 branch, and which thus on stemmatic grounds should not be accepted, it is to be placed below the split between the two groups. There are in principle two ways of dealing with this problem. One way is to explain the shared readings as coincidental (whether as error or as correction), or, if they are not to be considered coincidental, as contamination from Simplicius in the manuscripts at issue or, alternatively, as having a common source which influenced both the text as read by Simplicius and the group which shares these readings. Another way is to explain away for the places involved the readings at the χ2 side of the stemma as the result of contamination. As we shall see in the next section, the second way is a feasible strategy if the β group agrees with Simplicius against the consensus of the α and γ groups (be and K have been shown to be contaminated from the γ group anyway), but it is more difficult in the case the γ group agrees with Simplicius against the consensus of the α and β groups, since there are no real signs of contamination from the β group in the α group. In that case, however, the first strategy may be more effective, as the γ group is full of changes to the text anyway, and one source of these changes may be Simplicius’ commentary, or the same source may be responsible for all these changes in the γ group and the readings in Simplicius.

VI.9 The α group (P.S. Hasper)

CLXXVII

VI.9 The α group: Par. gr. 1853 and the main exemplar of the Arabic translation That E and Ψ are to be grouped together appears from the following list of errors unique to them: ποτε E2KbeLNRFHIJQΛ : om. E1Ψ ἅμα KLNRFHIJQwΛ T[212.2-3] : ἀλλ’ E1Ψ : ἅμα ἀλλ’ E2be [S][P] δὲ KbeLNRFHIJQwΛ Sp[1184.29]T[212.30] : δὴ EΨ? ὀρθῶς – πάντων οὐκ E2KbeLNRFHIJQwΛ Sp[1186.22-24]As[556] : om. E1 : οὐκ Ψ [T] 254b35 ἀντιθεῖναι beLNRFHJQΛ : ἀντιτιθεῖναι K : τιθείναι E1? : τιθέναι E2 : τιθέναι vel θεῖναι Ψ 255a15 συμφυῆ Ke2FHJQΛ : συμφυὲς be1LNR : συμφυὴς EΨ 255b8 τὸ γὰρ κοῦφον E2KbeLNRHJQΛ : om. E1Ψ [F] 257b2 πάντῃ KbeLNRFHIJQwΛ : πάντι EΨ? 257b24 κινεῖν E2KbeLNRFHIJQwΛ : κεινοῦν E1?[κειν…] : κινοῦν Ψ 257b32 αὑτὸ beLNR : ἑαυτὸ KFHIJQwΛ : αὐτὸ EΨ 258b29 τούτων E2KbeLNRFHIJ1QwΛ : τῶν J2 Sp[1252.26] : διὰ τούτων E1Ψ 259a13 ὂν KbeLNRFHIJQwΛ Sc[1254.28] : om. EΨ? 259b19 καὶ2 KbeLNRFHIJQwΛ Sl[1256.32] : καὶ τὸ ἐν E1?Ψ : καὶ ἐν Pa[393.2] : καὶ τὸ E2 259b22 κινεῖν E2KbeLNRFHJQ : κινεῖσθαι Λ : om. E1Ψ 260b20 εἴη … συνεχῶς KbeLNRFHIJQwΛ : om. EΨ[sed adest in tr.lat.] 261b28 ἡ E2KbeLNRFHIJQwΛ Sp[1278.2]Pl[905.9] : ἡ τεχνη ἡ E1 : ἢ τέχνῃ ἢ Ψ 262b23-24 πρὸς … φέροιτο E2KbeLNRFHIJQΛ : om. E1Ψ 263b23 ὥστε beFHIJQwΛ : ὥστε ἢ KLNR : ὥστε εἰ ἦν EΨ 263b27 καὶ ὅτε γίγνεται E2KbeLNRFHIJQwΛ Sp[1297.16-17]PA[907.24]Pp[R104r] : καὶ οὔτε γίγνεται καὶ ὁτὲ γίγνεται E1 : καὶ ὃ (τε) γίγνεται ὅτε γίγνεται Ψ? 264a3 οὗ E2beLNRFHIJQΛ : ᾧ K : om. E1Ψ 264a32 πρὸς KbeLNRFHIJQwΛ Sc[1304.28] : τὴν πρὸς EΨ 264b30 ταὐτὰ KFHIJQ : τὰ αὐτὰ E2beLNRΛ : ταύταις E1Ψ 265a26 ἄλλην E2KbeLNRFHIJQwΛ Sl[1313.21]Sp[1314.38-39] : ἄλλην στάσιν E1Ψ 265a27 εἰ … ἔφθαρται ἡ K1/2beLNRFHIJQΛ : ἡ … ἐφθαρμενη E1Ψ [E2][K1 partim] 266b15-16 ὑπερβάλλει … πεπερασμένης E2KbeLNRFHIJQΛ : om. E1Ψ 250b11 251b21 252a15 252b1-2

This is reinforced by the following list of shared readings unique to them:

CLXXVIII

Introduction

251b6-7

ἀλλ’ ἐδεῖ μεταβάλλειν beLNRFHIJQwΛ : ἀλλὰ μετέβαλεν EΨ T[211.14-C, Laur.85.14] : ἀλλὰ μετέβαλλεν K T[211.14-MSL] [S] 252b6 τοσαῦτα KbeLNRFHIJQwΛ : ταῦτα EΨ [S] 253a18 οἷον EΨ : ὁποῖον KbeLNRFHIJQw [Λ] [S][T] 253a20 τὰ … πάλιν KbeLNRFHIJQwΛ T[214.25]? : πάντα τὰ … EΨ [S] 254b28 τὰ KbeLNRFHIJQw Sc[1208.28] : om. EΨ [Λ] 257a34 ἀεὶ KbeLNRFHIJQwΛ : om. EΨ [S] 258a27 τὸ δὲ … γάρ KbeLNRFHIJQwΛ Sγρ[1245.2-3: ἔν τισι] : om. EΨ Sγρ[1245.3-5: τὰ πλείονα τῶν ἀντιγράφων]A[apud S(1245.3-5)] 259a32 ἢ1 E2KbeLNRFHIJQwΛ Sc[1257.26] : om. E1Ψ 260a30 προϋπαρχούσης E2KbeLNRFHIJQwΛ Sp[1265.20] : ὑπαρχούσης E1Ψ 263b26 ὄν2 E2KbeLNRFHIJQwΛ : om. E1Ψ Sp[1297.16]PA[907.24]? 264b29 οὐδὲ τὰς ἄλλας ἐνδέχεται KbeLNR Sp[1311.26] : οὐ τὰς ἄλλας ἐνδέχεται EΨ : οὐκ ἐνδέχεται οὐδὲ τὰς ἄλλας FHIJQwΛ [T] 265b28 οἴονται E2KbeLNRFHJQwΛ Sp[1319.3] : οἴοντα E1 : οἷόν τε Ψ – I 266a33 τῷ δὲ E1Ψ Sp[1324.31] : ᾧ δὲ E2K : ᾧ δὲ ὁ beLNFHIJQ1?wΛ : ᾧ δὲ ἡ RQ2 267b13-14 εἰ διαιρετὸς KH2Q : εἰ δὲ διαιρετὸς E2beLNRFIJΛ : εἰ δ’ ἀδιαίρετος w : εὐδιαίρετος γὰρ E1[αιρετὸς]Ψ, cf. Sp[1356.12]As[824]Pp[851.16], cf. Pa [429.25] : .. διαιρετοῦ H1 267b14 ἢ E2KbeLNRFHIJQwΛ Pp[851.16]Pa[429.24] : καὶ E1Ψ Sp[1356.12] 267b14 ἄλλος F2 : ἄλλως E1?Ψ Sp[1356.13]Pc[851.16] : ἀλλ’ ὡς beLNR F1HIJ QwΛ : ἄλλον E2K

Diels, and in his wake Ross, valued E very highly, despite the fact that E is a manuscript which is riddled with silly mistakes. On the basis of the comparison with the Arabic translation it has become even easier to see through these mistakes as peculiar to E alone, and to focus on the evaluation of the readings provided by E and Ψ together. There are three places where E’s reading is accepted by Ross, but now appears to be a mistake: 253b26 τι νοσήσῃ E1 : νοσήσῃ τί LNRFHI2J0Qw2Λ : νοήσῃ τι w1 : νοσήσῃ τις Ψ(vel τις νοσήσῃ)KbeI1J1 : τις νοσήσῃ E2 [S][P] 255a18 ἐμψύχων E : ἐμψύχων αὐτὰ ΨKbeLNRFHIJQwΛ Sp[1210.35-36]As[587] 256b31 δ’ : δὲ E : δ’ ὅτι ΨKb[δὲ]e[δὲ]LNRFHIJQ[δὲ]wΛ

And there is one place where Ψ could be cited in support of the decisions by Ross to adopt a reading which he only could find in E: 253a18 οἷον EΨ : ὁποῖον KbeLNRFHIJQw [Λ] [S][T]

even though I would favour the alternative reading ὁποῖον, because it is more likely that ὁποῖον was simplified into οἷον than the other way round.

VI.9 The α group (P.S. Hasper)

CLXXIX

On the basis of the results obtained in this introduction on the stemma for Physics VIII, it is also possible to give an estimate to what extent the α group as a whole or the manuscripts E and Ψ individually are contaminated from other parts of the tradition. Because K and be, and to a lesser extent Λ, are, despite their massive contamination, related to the α group, it is possible to argue from the consensus of K and/or be (and perhaps occasionally of Λ) with either Simplicius or the β group (or both) that the α group is contaminated or corrected: κινητῶν EΨK2FHI1JQwΛ : κινητῶν καὶ κινητικῶν beLNRI2 Sp [1127.22] : κινητικῶν K1 [T] 252b25 τὸ αὐτὸ KLNRFHIJQwΛ Sp[1188.3] : τοῦτο τὸ αὐτὸ EΨ[vel τὸ αὐτὸ τοῦτο] : τὸ αὐτὸ τοῦτο be 253b9 ὅμως KbeFI : ὁμοίως EΨLNRHBJQwΛ [S][T] 257a3 ἢ E1ΨK2LNRFHIJQwΛ : εἰ δ’ E2[δὲ]K1be Sl[1229.6: δὲ] 258b11 ἀΐδιον ὃ πρῶτον κινεῖ Sp[1250.35, 1251.3, 1251.28, 1254.32] : ἀΐδιον ὃ κινεῖ E2beLNR As[648, 649, 650]?Sp[1251.14]?T[223.14]Pa[387.26] : ὃ πρῶτον κινεῖ E1?ΨKFHIJwΛ : ὃ πρῶτον κινεῖ .. Q1 : ὃ πρῶτον κινεῖ ἀΐδιον Q2 258b14 τὸ E1ΨFHIJQwΛ : τὸ ἀεὶ E2[αἰεὶ]Kb[αἰεὶ]eLNR Sp[1251.16], cf. As[651] 261a5 πρῶτον EΨFHIJQwΛ : προτέρον KbeLNR Sc[1270.28] 261a16 τοῦ ὀργάνου E1E3Ψe2LNRF[cum ὑπάρχει]HIJQwΛ : om. Kbe1 Pc[900.26]Pa[399.16] : exp. E2 261a33 εἰσιν αἱ EΨFHIJQw : εἰσὶ καὶ KLNR : εἰσὶ καὶ αἱ be : εἰσὶ T[227.8] [S] 261b10 οὐδὲν EΨHIJQw : om. KbeLNRFΛ 262a9 Β EΨK2e2 – FHIJw Sp[1280.3] : Γ K1be1LNR – Λ [Q]306 262a12 τὸ πλάγιον EΨKLNRFHIJQwΛ Sl[1279.18-a]Pl[842.19] : τὰ πλάγια be Sl[1279.18-A]Sp[1280.13]Pp[842.20]T[228.14] 262b1 τὸ Α ἅμα γεγονέναι EF2Λ : ἅμα τὸ Α γεγονέναι KbeLNR : ἅμα γεγονέναι τὸ Α HIJQw [Ψ][F1] [S] 264a14 δὴ E2KLNRΛ : δ’ E1Ψ?beFHIJQw [S] 264b3-4 ἀλλοίωσις εἰς E2ΨFHIJQwΛ : ἀλλοίωσις ἡ εἰς KbeLNR : ἀλλοίωσεις E1 [S] 265a26-27 στάσιν … κίνησις E1ΨK2e2LNRFHIJ[partim]QwΛ, cf. Sp[1314.36] : om. E2K1be1 Sl[1313.21-1315.5]T[232.26-27 non reddit]Pa[418.10-11] 265b7 τὸ1 E2beLNR T[233.4, 5] : τῷ K : om. E1ΨFHIJQw Sp[1316.19] [Λ] 265b7 μένειν E2KbeLNRHΛ T[233.4, 5] : μένει E1ΨFIJQw Sp[1316.19] 265b12 ἀπὸ E1ΨFHIJQ1w Sc[1317.17] : ἀπό τε E2KbeLNRQ2 265b15 οὔτ’ : οὔτε E1ΨKFIJQw : οὔτε ἡ E2beLNRHB Pa[419.26], cf. Sp [1317.31] 251a18

_____________ 306

See footnote 270.

CLXXX

Introduction

οὔτε E1ΨKFIJQwΛ : οὔτε τὸ E2beLNRHB Pa[419.26], cf. Sp[1317.31] λέγωμεν EΨ?KIJQwΛ : λέγομεν beLNRFH T[233.15]? [S] κινεῖν E2KbeLNRHQ2Λ Sc[1345.14] : κινεῖ E1Ψ*?FIJ : κινῇ w : κινοῦν Q1? ἐν E1Ψ?FIΛ Sl[1353.35-a]T[236.2]Pa[429.9] : om. E2KbeLNRHJQw Sl [1353.35-A] 267b18 ἀκίνητον EΨKFIJQw Pa[430.5] : ἀκίνητόν ὂν beLNRH Sp[1358.4]

265b15 266a11 267a2 267b7

It is not necessary to accept every single instance on this list as a case of contamination in the α group, but the case for most of them seems pretty solid, notably for 252b25, 258b11, 261a5, 261a33, 261b10, 262a9, 265a26-27, 265b7 (both), 267a2 and 267b7. Also striking is that (possible) contamination in the α group as a whole can be almost exclusively be found in the second half of the book. I do not really have an explanation for this fact; thus the uneven distribution of these possible contaminations enhances the chances that they do constitute contaminations, and not accidentally shared readings. Not for all places can the reading of Ψ be determined, but it seems probable that at least E has been contaminated – and thus perhaps Ψ as well: 252a6-7 νομίζειν ἀρχὴν Pa[363.23-24] : νομίζειν εἶναι ἀρχήν KFHIJQwΛ : εἶναι νομίζειν ἀρχήν ELNR : νομίζειν ἀρχὴν εἶναι be [Ψ] [S][T]307 255a1 παράσχοι ἂν ELNRFHIJQwΛ : ἂν παράσχοι Kbe [Ψ] [S][T] 259b23 συνεχῶς EFHIJQwΛ : συνεχῆ KbeLNR Pa[393.7] [Ψ] 263a26 ἡμίση EFHIJQw : ἡμίσεα KbeLNR [Ψ][Λ] 263a28 ἄπειρα ἡμίση E1KFHIJQwΛ : ἡμίσεα ἄπειρα LNR : ἄπειρα ἡμίσεα E2be [Ψ] 263b1 ἀριθμεῖν δύο EFHIJQwΛ : δύο ἀριθμεῖν KbeLNR Sp[1291.17-18] [Ψ]

Again the distribution over the book is somewhat uneven. This is different if we look at the probable cases of contamination in E alone: 250b11 πότερον ΨKLNR Sl[1118.12] : πότερον δὲ E2beFHIJQwΛ : πότερον δὲ τὲ? E1 [T]? 250b18 τε E – FHIJQwΛ : om. Ψ?KbeLNR Sl[1120.28]T[209.8] 251a15 ὥστε δεῖ ΨKbeNRFHIQwΛ T[210.6-9] : ὥστ’ εἴδει L : ὥστε δὴ E – J [S] 251b3 τὸ μὲν κινεῖ Ψbe, cf. Sp[1129.17-18], cf. T[211.13] : κινεῖ E – KLNRFHIJ QwΛ 252a1 δὴ E – beF2IJQwΛ : δὲ ΨKLNRF1H Sc[1170.26] [T] 252a6 ταύτην δεῖ KbeHIJQwΛ : ταύτην δεῖν F : δεῖ ταύτην E – LNR : ταύτην Ψ Sp[1183.23]Pa[362.23] [T] 252a25 οὐκ αἴτια τὰ ὑποτεθέντα ΨKbeLNRFHIJ2w : οὐκ αἴτια ὑποτεθέντα E – Q1[..τια]Q2 : οὐκέτι ἀποτεθέντα J1 [Λ: οὐκ αἴτια] [S][T]

_____________ 307

The variation and distribution of the readings indicates that the original reading is νομίζειν ἀρχὴν.

VI.9 The α group (P.S. Hasper) 254a18 254a28 255a8 255b4 255b18 256a28 256b26 257b10 257b14 257b22 258a9 259a1 259b24 260a25 262a29 263b11 265a34 266a26 267a3

CLXXXI

ἤτοι γὰρ KLNR : ἤτοι ἄρα Ψ? : ἢ γάρ τοι E – beFHJQw [Λ: γάρ] [S] καὶ E – beLNRFHIJQw A[apud S(1202.31)] : om. ΨKΛ αὐτὸ bLNRH : αὐτὸ αὑτῷ Ψ : αὑτῷ FIJ : αὐτῷ E – KeQwΛ [S] τι μὴ E – LNRFHIJQwΛ : μὴ τι Ψ?Kbe Sc[1214.16, 25]Pl[830.31]T[219.5] Pa[376.6] τε E – LNRFIJQwΛ : om. ΨKbeH Pa[376.23] τι καὶ ΨbeLNR : om. Λ : τι E – KFHIJQw μόνως Ψ?Q Sl[1227.11-a] : ἂν μόνως FIwΛ Pc[833.9] : μόνως cum ἂν post b27 κρατοίη Sl[1227.11-AFM] : μόνως ἂν E – KbeLNRHJ308 αὐτὸ καὶ Ψ?KbeLNR Sc[1235.23] : αὐτὸ E – FHIJQwΛ τὸ αὐτὸ E2Ψ?eLNR Sl[1237.3] : αὐτὸ E1 – KbFIJQw Pa[383.16] : illeg. H ἔσται Ψ?Kbe1LNRFIQwΛ : ἔστιν E2e2J : ἔτι E1 : illeg. H δὲ2 Ψ?KbeLNRFHIJ2QΛ : om. E – J1w κινουσῶν E2ΨKbLNRFHIQw Sc[1253.19] : κινούντων E1 – J [Λ] καὶ Ψ*K1LNRH: om. Sc[1260.11] : καὶ μὴ E1K2beFIJQwΛ ἣν E2ΨKb2eLNRFHIJ2QwΛ Sp[1265.5] : ὂν b1 : om. E1 – J1 [P] κατὰ E2ΨKbeLNRFHIJ2Q2wΛ : καὶ E1 – J1Q1? ἅμα E – KbeFIJQwΛ Pl[845.16] : om. ΨLNRH Pa[410.23-24] [S] μέσον καὶ τέλος ΨKbeI Sp[1315.19, 24]Pp[R-106r]T[232.31-32] : τέλος καὶ μέσον ἐστὶν LNR : μέσον E – FHJQwΛ ἐν E2ΨKbeLNRFHIwΛ Sp[1324.9]Pa[423.9] : om. E1 – JQ1 τε κινεῖν E2KbeLNRHIQ2?Λ Sc[1345.28] : τε καὶ κινεῖν beF1Q1?w : τε καὶ καὶ κινεῖν J : κινεῖν Ψ : τι καὶ κινεῖν F2 : καὶ κινεῖν E1

What is also striking is that quite a few of these contaminations in E can be explained as from J alone309 or from JQw: at 251a15, 257b22 and 260a25 E’s reading is only found in J, at 252a25, 258a9, 259b24, 262a29 and 266a26 E’s reading is found in J together with Q and/or w. The only places for which E’s reading cannot be explained as from J (or a manuscript closely related to it) are 252a6 and 252a6-7. That E was copied from a manuscript which contained supralinear corrections can be seen from the following three places: κινούμενον E2ΨKbeLNRFHIJQwΛ Slp[1189.22, 27]Pa[367.23] : κινουμενουμενον E1 256b17 τοῦτο ΨKbeLNRFHIJQwΛ Sc[1226.35]Pp[885.17] : τούτου τὸ E

253a2

_____________ 308

The variation in the place of ἄν, together with the fact that some sources do not feature it at all, seems enough of an indication that the original reading is without. 309 The same phenomenon, attested in a variant reading in the margin of ms. Leiden Or. 583, is described above, p. LXIV.

CLXXXII 259b3

Introduction παρεῖχε K[παρεῖχεν]beFHIQΛ : παρέχει E2ΨLNRJw Sp[1258.1]? Pc[889.34] Pa[391.17] : παρειχει E1

Similarly the exemplar of the Arabic translation can be shown to be in all likelihood contaminated310: 251a5 252a3 252a23 253a22 255a9 255b15 256a26 257b18 258a11 258b27 259b16 260b2 260b17 261b13 262a14 263a11 263b9 263b28 264a1

τούτων EbeLNR Sl[1125.25-a]T[209.22]Pa[359.8] : τούτου Ψ? – KFHIJQ wΛ Sl[1125.25-AFM] ταῦτα ἀδύνατα E[ταῦτ’]beLNR Sl[1170.21: ταῦτ’]Pa[363.19] : τοῦτ’ ἀδύνατον Ψ? – KFHBIJQ[τοῦτο]wΛ [T] φάναι EK1be1 Sp[1185.25]? : ἀποφᾶναι NRFIJQΛ : ἀποφάναι Ψ – K2e2 LHw [T] ἥπερ E[εἴπερ]Kbe1LNRJ[ἢπερ]QwΛ Sl[1193.1] : ἔσται ἥπερ Ψ – e2FHI εἰ EKbeLNRΛ Sp[1210.7] : ἐπεὶ Ψ – FHIJQw ποι EKbeLNRFHIJQwΛ? : που Ψ? o1? Sc[1216.30]Pa[376.21] αὐτὸ αὑτῷ LNRFHJ2Q2 : αὐτὸ αὐτῷ E1be2w : αὑτῷ K : αὐτὸ αὐτὸ E2 J1?Q1? : αὐτὸ αὑτὸ Ψ? – e1 [I] τῶ. E1 : τῷ E2be : om. KF : τὸ2 Ψ? – LNRHIJQw P[CP 90.1] [Λ: non τῷ] πολλῶν EKbeLNR Pl[836.6]Pa[385.18] : πλειόνων Ψ – FHIJQwΛ Pp [836.9] [S] οὐδ’ αὖ τῶν ἀεὶ EKeLNRFHIJQΛ : οὐδ’ αὐτῶν ἀεὶ bw : ὥστ’ αὑτῶν ἀεὶ Ψ, cf. Sp[1252.26] ἑαυτά E1Ke2FJ2 Sc[1258.32] : eras. E2 : αὑτὰ LNR : αυτά P[890.26-35] : ἑαυτό be1HIJ1Qw : αὐτὸ Ψ? ἐνεργείᾳ EKbe : τὸ ἐνεργείᾳ Ψ – LNRFHIJQw Sp[1265.31] [Λ] λέγεται δὲ EKLNRFHIJQwΛ : λέγεται γὰρ be Sc[1267.36] : λέγοιτο δὲ ἂν Ψ? Ap[apud S(1268.3)] ἔστιν Ebe : ἔσται Ψ – KLNRFHIJQΛ [w] ἐπ’ EKbeLNR : ἐπὶ τῆς Ψ? – FHIJQwΛ Sc[1280.27]Sp[1280.23] τοῖς περὶ κινήσεως E2K2beLNRFHIJQw : περὶ κινήσεως K1 : ταύτην τὴν ἀπορίαν Ψ, cf. S[1289.3-4]Pa[407.3] : om. Λ [E1] δὲ καὶ ὅτι EbeLNRF2JwΛ Sl[1294.3-A] : δὴ ὅτι H : δὲ ὅτι καὶ Q Sl[1294.3a] : γὰρ I1 : γὰρ δὴ I2 : δὲ ὅτι Ψ – KF1 Α EKbeLNRΛ Sp[1297.22-23]?P[41.9] : Α cum ras. χρόνῳ I : πρώτῳ χρόνῳ F1 : Α χρόνῳ Ψ – F2HJQw χρόνον EKbeLNRΛ Sp[1297.27] : χρόνος ἦν Ψ – FHIJQw

_____________ 310

I have omitted two cases in which the Arabic translation does not translate a certain word, but this may be due to the translation, not to its exemplar Ψ: 253a23 ποτε EbeLNRHIJQw : ποτ’ K : om. Ψ? – FΛ [S] || 260b27 δὴ EKbeLNRFHIJ2Qw : ἤδη Λ : om. Ψ? – J1.

VI.9 The α group (P.S. Hasper)

CLXXXIII

265a14-15

εὐθεία EKLNR T[232.17] : ἐπ’ εὐθείας Ψ – beFHI[ἐπὶ]JQwΛ, cf. Sp[1313.26, sed cf. 1313.28-29, 33-34, 34 ad a16] 265a20 πεπερασμένης EKLNR T[232.22] : πεπερασμένης εὐθείας Ψ – be FHIJQwΛ T[232.19], cf. Sp[1314.10-11] 267b5 ἐκεῖνο EKHI1Q2 Sp[1353.26, 29]T[236.1-CS] : ἐκείνου Ψ – b2eLNR FI2JQ1?w T[236.1-M] [b1] 267b13-14 εἰ διαιρετὸς KH2Q : εἰ δὲ διαιρετὸς E2beLNRFIJΛ : .. διαιρετοῦ H1 : εἰ δ’ ἀδιαίρετος w : αιρετὸς γὰρ E1 : εὐδιαίρετος γὰρ Ψ?, cf. Sp[1356.12] As[824]Pp[851.16], cf. Pa[429.25] [Pa(429.24): διαιρετὸς]

Since we only have access to Ψ through the Arabic translation, it may sometimes be a matter of translation rather than contamination, even though its reading agrees counter-stemmatically, for example at 262a14 and 265a20. There seem to be two or three sources of contamination in Ψ. Most places can be explained with contamination from (a manuscript related to) FH, notably 253a22, 253a23 and 263b9. There may be some contamination from JQ, as perhaps in 256a26 and 260b27, but the evidence is less conclusive. Clearly there is some influence from the commentary tradition, as is shown by 258b27, 260b17, 263a11, and 267b13-14. The same influence of the commentary tradition on the Arabic translation also appears from two places at which it deviates from the correct understanding. At 256b3 the Arabic translation fails to render πρὸς δὲ τοῖς εἰρημένοις. This may well be ultimately due to the suggestion mentioned by Alexander (according to Simplicius, In Phys., p. 1224.26-27) and accepted by Themistius, In Phys. paraphr., p. 222.23–223.13, that the section 256b3-27 is to be moved down to after 258b9. At 266a18 the Arabic translator renders the phrase ἐν πλείονι γὰρ τὸ μεῖζον not with ‘For in a longer period of time it will cause to move over a larger [stretch]’, as it should be, but rather with ‘For being the greater fraction it must cause to move in a longer time’, with which he follows a suggestion by Simplicius, In Phys., p. 1322.8-14311. The process of contamination through presumably supralinear additions left also some traces in Ψ: 252a26 τοῦτ’ LNR : τοῦτο KbeFHIJQwΛ : ταὐτ’ E1 : ταὐτὸ E2 : ταὐτὸν Sp [1185.31] : ταὐτὸ τοῦτο Ψ [T] 255a8 αὐτὸ bLNRH : αὑτῷ FJ : αὐτῷ EKeQwΛ : αὐτὸ αὑτῷ Ψ [S]

Further we can find in Ψ a correction which is quite dramatic, but which probably goes back to a discussion in the late ancient commentary tradition, given that we find the same reading in Pachymeres’ paraphrase: _____________ 311

See also Ross, Aristotle’s Physics, p. 722–723.

CLXXXIV

Introduction

250b22-23 δ’ ἕνα ἢ μὴ FHIJQwΛ : δὲ ἕνα ἢ μὴ Kbe : δ’ ἓνα καὶ μὴ Ψ Pa[358.2-3] : δὲ ἢ ἕνα ἢ μὴ E1LNR Sc[1122.26-28]As[541]? : δὲ εἶεν ἀεὶ μὴ E2 [T(209.11-13) interpr. ἕνα ἢ ἀεὶ ἢ μὴ ἀεὶ, sed S quoque]

This is a difficult passage, which is actually emended by Ross to read: ὅσοι δ’ ἕνα ἢ μὴ ἀεί, καὶ περὶ τῆς κινήσεως ὑποτίθενται κατὰ λόγον (250b2224), so as to ensure that Aristotle claims that those predecessors who posit there to be one universe fall into two camps, those who hold this one universe to be eternal and those who deny this, and have corresponding views about motion, eternal in case the universe is eternal, not eternal in case the universe is not eternal. Though both Themistius and Simplicius interpret the passage in this way, they do not claim that the text should be read in this way. It seems likely that the original reading was ἢ ἕνα ἢ μὴ ἀεί, which is difficult enough to understand to warrant Alexander’s exclamation θαυμαστὸν τὸ τῆς συντομίας (in scholium 541 in Rashed’s edition)312. Presumably for that reason the emendation ἓνα καὶ μὴ ἀεὶ was proposed – and this emendation found its way into the Arabic translation and Pachymeres’ paraphrase. On this emendation Aristotle would first limit the philosophers whose views he reports to those who claim that there is motion also because of ‘creating a world’ (κοσμοποιεῖν) and distinguish between two groups: those who claim that an infinite series of worlds are created, and thus that there is always motion (250b18-21); and those who claim that there is a single world which has not always been there and thus that there is not always motion (250b21-23). This interpretation does make sense of Aristotle’s subsequent reference to Anaxagoras, but it has difficulties accommodating Empedocles’ position, which does not concern a single world which has not always been there. A first trace of such an interpretation may be found in Alexander’s scholium, who says that the complete formulation should be ὅσοι δὲ ἕνα καὶ μὴ ἀεὶ ἢ ἕνα καὶ ἀεί: ‘Those who claim that there is one universe which is not all the time there or there is one which is all the time there.’ This is in fact the interpretation put forward by Simplicius and Themistius, but since it is formulated differently, Alexander’s comment may have given rise to the reading found in Ψ. Alternatively, that _____________ 312

This should then be understood as saying: ‘Those who [claim that] either one [universe comes into being] or [the universes] do not all the time [come into being], have also according assumptions about motion.’ The first category comprises Anaxagoras, who believes that there is one universe, which came into being at a certain moment, the second Empedocles, who believes that there is a series of universes who come and go interspersed with periods of rest. Interpretations which supply γίγνεσθαι (the second verb at 250b19) fare much better than those which supply εἶναι, even though the latter make more sense given the order between the verbs at 250b19.

VI.9 The α group (P.S. Hasper)

CLXXXV

reading is the result of a supralinear addition derived from Alexander’s comment which ended up in the text313. As stated above, there are a striking number of further readings uniquely shared by the Arabic translation and Pachymeres’ paraphrase: 251a16 ταῦτα EKbeLNRFHBIJQw o Sl[1127.16]T[210.10]? : ταῦτα τὰ πράγματα Ψ? Pa[360.7] : τοῦτο Λ 253a17 ἢ E2KbeLNRFHIJQwΛ o Sp[1190.35]?T[214.19] : om. E1 : καὶ Ψ? Pa[368.21] 253a33 τίς EKbeLNRFHIJQwΛ o : om. Ψ Pa[369.12] 253b33 ἐξ ἀνάγκης EKbeLNRFHIJQwΛ o : om. Ψ Pa[378.16] [S] 255b19 γὰρ KbeLNRFHIJQwΛ o : om. E : δὲ Ψ? Pa[376.25] 256b13 εἶναι EKbeLNRFHIJQwΛ o : εἶναι ὥστε ἀναιρεῖται τὸ κινεῖσθαι (vel sim.) Ψ cf. S[1225.38-39: οὐδὲ περιλείπεται], cf. Pa[380.9-10: ὥστε ἀναιρεῖται] 257b6 ὅτι EKbeLNRFHIJQwΛ : τί vel τί τὸ Ψ Pa[383.13] 260a6-7 δὲ κινουμένου ἤδη E2e2FHIJQw o Pl[894.5] : δὴ κινουμένου ἤδη LNR : ἢ κινουμένου ἤδη Kbe1 Sc[1263.16-17] : κινουμένου ἤδη Λ : δὲ κινουμένου ἢ E1 : om. Ψ Pa[394.11] 265b15 μόνης EKbeLNRFHIJQwΛ o : om. Ψ Pa[419.26]

The only explanation for these shared readings, it seems, is a common source in the form of a commentary or scholia used by Pachymeres and the scribe of Ψ (or one of its predecessors). Ψ may have undergone all these influences, but in addition to the places listed above where it confirms Ross’ choice for a reading which otherwise is only found in E, it also features a few readings (almost) unique to it which are to be adopted as correct: 252a6

ταύτην δεῖ KbeHIJQwΛ : ταύτην δεῖν F : δεῖ ταύτην ELNR : ταύτην Ψ Sp[1183.23]Pa[362.23] [T] 256b26 μόνως ἂν EKbeLNRHJQ2 : ἂν μόνως FIwΛ Pc[833.9] : μόνως Ψ?Q1 Sl[1227.11-a]Pa[380.25-27] : μόνως cum ἂν post b27 κρατοίη Sl[1227.11AFM] 259a28 ἑκάτερον E : ἑκατέραν KbeLNRFHIJQwΛ : ἑκατέρου Ψ, Ross [S] 259b26 μενεῖν Ψ T[431.27]?, Ross : μένειν EKbeLNRFHIJQw [Λ]

Thus, together with E, Ψ is the most important source for the constitution of the text of Physics VIII. _____________ 313

Rashed, Alexandre d’Aphrodise, p. 492, speculates that Alexander actually found only ὅσοι δὲ ἕνα καὶ μὴ ἀεί in his manuscript, but felt the need to add ἢ ἕνα καὶ ἀεί. But this cannot make sense of Alexander’s exclamation about the marvelous conciseness of Aristotle’s phrase, which is echoed by Simplicius (p. 1122.26–1123.1).

CLXXXVI

Introduction

VI.10 Conclusion The manuscript tradition for the eighth book of Aristotle’s Physics turns out to be quite complicated, in particular because of the influence of what became later the vulgate (group γ) on other parts of the tradition. This influence can be detected in every part of one of the two main groups, namely in the one constituted by EΨKbe and to some extent Λ, in K and be extensively, and in EΨ (both together and each individually) to a lesser degree. This makes it difficult to assess the authority of each of the individual manuscripts of this group, though clearly E and Ψ are the most important ones. These claims about the extent of contamination from group γ in each part of the group constituted by EΨKbe cannot be made without the evidence of two further sources: Simplicius’ commentary and the β group. It cannot be established whether the main manuscript used by Simplicius is completely independent from the extant manuscript tradition, but that may also be because the evidence is almost exclusively drawn from just one book of the Physics. It seems as if Simplicius shares a small number of errors or rejectable readings with the γ group, but this cannot be taken to imply that Simplicius is to be located in the stemma as most closely related to that group. It seems also possible. As there is no real evidence in Physics VIII that Simplicius’ manuscript shares errors with parts of the direct tradition, we may for the time being assume that it is independent from the direct tradition, and thus that for most cases the consensus between Simplicius and a substantial part of the direct tradition provides the reading to be adopted. But since the information provided by a commentary is by its nature rather patchy and does not lend itself for passing on insignificant errors, even more important for it is the position of the β group within the stemma. This group clearly shares a substantial list of errors with the γ group, and thus constitutes together with that group the other half of the stemma. On the other hand, it agrees often with (parts of) the EΨKbe group, in that it does not feature many of what turn out to be changes to the text which are found in the γ group. Thus stemmatically inappropriate constellations of consensus between parts of the EΨKbe group and the γ group can be identified as contaminations. The main exemplar of the Arabic translation is of similar importance for drawing these conclusions, since knowledge of its readings allows us to see the structure of the EΨKbe group far more clearly, and to filter out all the many singular mistakes in E. It often joins E in providing the clearly superior reading, and occasionally offers alone the correct reading. Thus the stemma for the whole tradition for Physics VIII presumably looks as follows:

VI.10 Conclusion (P.S. Hasper)

CLXXXVII

ω χ1

χ2 γ

α

β

* * *

Ψ

δ *

J

900 E 1000 1100

K

Λ

N

1200

Q

w 1300

b

R

1400 e 1500

L

I

F

H

CLXXXVIII

Introduction

VII. The indirect tradition of Isḥāq ibn Ḥunayn’s translation Apart from Isḥāq ibn Ḥunayn’s translation as preserved in the two manuscripts described above and the extant Greek codices of the Physics, additional materials for the critical constitution of the Arabic text are provided by its indirect tradition. On the one hand, this concerns quotations of Isḥāq’s translation in Arabic sources; on the other hand, we have access to two (and for certain parts of the text even three) medieval Arabo-Latin translations.

VII.1 Quotations in Arabic sources The most copious indirect Arabic transmission of Isḥāq’s translation is found in the very manuscript which serves simultaneously as our sole complete witness to its direct tradition, ms. Leiden Or. 583. As mentioned above, this codex contains comments by the Baghdad Aristotelians Abū Bišr Mattā ibn Yūnus, Yaḥyā ibn ʿAdī, Abū ʿAlī ibn al-Samḥ, and Abū l-Faraǧ ibn al-Ṭayyib, who often quote the Aristotelian text in their discussions as well, occasionally in a rather paraphrastic way. Such quotations —although preserved in the same manuscript and together with the complete copy of Isḥāq ibn Ḥunayn’s text— constitute in a way an indirect tradition, as many of them ultimately trace back to copies of this translation other than the one that served as the exemplar of the scribe of ms. Leiden Or. 583. In general, we can distinguish between (mostly rather short) quotations that occur en passant within the exegetical discourse (“implicit citations”) and lemmata proper prefixed to commentary sections and often explicitly introduced by the formula qāla Arisṭū or qāla Arisṭūṭālīs. On the whole, we are thus faced in ms. Leiden Or. 583 with three levels of transmission: (i) The main lemmata of the compilation reproducing Isḥāq ibn Ḥunayn’s translation in large portions; (ii) excerpts of the Aristotelian text resumed at the beginning of a commentary section related to the preceding main lemma in the form of a separate partial lemma; and (iii) implicit citations. Not all quotations of type (ii) and (iii) are taken from Isḥāq ibn Ḥunayn’s translation, as the Baghdad Aristotelians also used translations by Abū ʿUṯmān al-Dimašqī, Qusṭā ibn Lūqā, and other sources (cf. above, chapter II.3). However, even in the case of paraphrastic citations of the Aristotelian text it is often easy to distinguish between citations of Isḥāq’s version and those extracted from other materials. The exact sources of type-(ii) and type-(iii) quotations are not known. As for the latter, it is safe to assume that they formed a unity together

VII.1 Quotations from Isḥāq’s translation

CLXXXIX

with the surrounding commentary section, which was as a whole incorporated into the Baghdad compilation, no matter whether the relevant commentary section originates from a self-contained commentary or scholia by Arabic-writing scholars or an Arabic translation of a late ancient Greek commentary. The same may hold true for type-(ii) quotations, some of which correspond with lemmata found in Philoponus’ commentary, while others evidently do not originate from this source. However, as far as these partial lemmata give Isḥāq ibn Ḥunayn’s text, there is no need to assume that the author of the following comment was responsible for prefixing the relevant lemma. These lemmata may just as well be added by any scholar responsible for one of the Baghdad school compilations (θ1, θ2, etc.), who had access to a separate copy of Isḥāq ibn Ḥunayn’s translation; they may even have been incorporated much later, e.g. by Abū l-Ḥusayn al-Baṣrī or by the scribe of the Karḫ-copy314. On the other hand, we can certainly rule out the possibility that Abū l-Ḥakam al-Maġribī, the scribe of ms. Leiden Or. 583, incorporated them himself on the basis of his transcription of Isḥāq ibn Ḥunayn’s text. For some type-(ii) quotations provide the correct or complete text, whereas the corresponding sections of the main lemmata are faulty or incomplete and cannot have served as the source of these quotations. While type-(iii) quotations are scattered through all parts of the text (as far as taken into consideration in the commentary sections), type-(ii) quotations occur exclusively in bks. III—VII of the Physics, more precisely at the following places315: Bk. III: 198.12; 217.16 Bk. IV: 336.13; 337.3; 345.8.16.22; 346.8; 372.16; 375.16.20; 391.3; 409.20; 410.1.5; 417.10.13; 424.11.18; 425.9.15.20.26; 426.14; 434.9.16; 435.6; 436.1.4.15.20.24; 437.9; 438.3; 446.4.13.21; 460.13; 461.7; 475.12.20.23; 476.3.14 Bk. V: 494.23; 495.4; 496.17; 501.21; 504.15.25; 507.1.3.13.16; 509.9.17; 517.8.16; 518.6.13; 519.11; 525.15; 526.1.10.15.24; 527.7; 535.21.24; 536.11.13.16.18; 552.18; 555.3.10.13; 557.11; 558.3.8; 560.17; 561.11; 561.18.22.26; 562.3; 565.23; 566.7; 568.4.12.20; 572.5.16.22; 573.8; _____________ 314

The varying introductory formulas “qāla”, “qāla Arisṭū”, and “qāla Arisṭūṭālīs” may point to different compilators responsible for interspersing separate partial lemmata. However, the distribution of the different formulas reveals no significant pattern or regularity. 315 References are to page and line number of the beginning of the quotation in Badawī’s edition. Underlined references point to quotations drawing unambiguously on Isḥāq ibn Ḥunayn’s translation.

CXC

Introduction

581.11.15.19; 582.3.10; 586.16.23; 587.3.14.17; 588.8.21; 589.16; 590.1.6.8; 595.13.24; 596.2.5.9.15.22; 600.5; 601.13; 602.3.7 Bk. VI: 614.14; 621.10; 625.14.19; 629.19; 632.20; 633.4.13.17.23; 634.10; 642.17.21.23; 643.3; 659.3.6.9.18.22; 662.11.13; 664.20; 665.3.8; 668.14; 670.7.12; 672.14.21; 673.20; 677.3.7.12; 681.19; 691.11.18.22; 692.8; 698.25; 699.7.12.17; 722.6; 729.5 Bk. VII: 742.12; 743.14; 744.9.18; 745.3.11.18.22; 755.6.10.17.26; 757.3.10; 766.21; 768.1.6.20; 769.14; 770.9; 771.4.15; 772.3; 782.22; 784.11; 785.5.13.21; 786.1.5.11.16; 787.4.11.18; 796.11.18; 797.7.18; 798.16; 799.4.13; 800.17 In many cases, such quotations (both partial lemmata and implicit citations) preserve Isḥāq’s translation in a more accurate state than the corresponding passages of the main lemmata of ms. Leiden Or. 583. It goes without saying that they must be taken into consideration for the critical edition of the Aristotelian text. Other sources of the indirect tradition are supplied by quotations in surviving self-contained works of Arabic-writing philosophers, which demonstrate the dissemination of Isḥāq ibn Ḥunayn’s translation in various parts of the Near Eastern world and Muslim Spain between the 10th and 13th centuries. Since Ibn Rušd’s Šarḥ or “Long Commentary” on the Physics is not extant in the Arabic original (see the following section on the Latin testimonies of the indirect tradition), our main sources consist in the works of Yaḥyā ibn ʿAdī, Ibn Sīnā (Avicenna, d. 1037), Ibn Bāǧǧa (Avempace, d. 1139) and ʿAbd al-Laṭīf al-Baġdādī (d. 1231). Yaḥyā ibn ʿAdī. The relationship between Yaḥyā ibn ʿAdī’s comments transmitted in ms. Leiden Or. 583 and his separate treatises on particular questions raised in Aristotle’s Physics has not been studied so far. However, what can be assessed with certainty is the fact that Yaḥyā ibn ʿAdī draws in these treatises on Isḥāq ibn Ḥunayn’s translation. This becomes clear chiefly from his technical terminology, but also from a number of literal correspondences, as illustrated by the following examples316: _____________ 316

Unless otherwise indicated the texts quoted are published in: Maqālāt Yaḥyā ibn ʿAdī al-falsafiyya. Dirāsat wa-taḥqīq Sahbān Ḫalīfāt. ʿAmmān 1988 (= ed. Ḫalīfāt). Most of the below correspondences were already pointed out by G. Endress, The Works of Yaḥyā ibn ʿAdī (cf. Index of names, p. 139, “Aristotle, Physica”), and in the footnotes of Ḫalīfāt’s edition. Further paraphrastic borrowings, not quoted here, are found in another short treatise edited and studied by Peter Adamson & Robert Wisnovsky, “Yaḥyā ibn ʿAdī on the Location of God”, Oxford Studies in Medieval Philosophy 1 (2013), p. 205—228, cf. notes 39—41.

VII.1 Quotations from Isḥāq’s translation

CXCI

[1] Physics I 2, 185a33-34: τὸ γὰρ ἄπειρον ἐν τῷ ποσῷ Transl. Isḥāq, ed. Badawī, p. 11 = Yaḥyā ibn ʿAdī, “Nusḫat mā kataba bihī Yaḥyā ibn ʿAdī fī l-ḥukūma bayna Ibrāhīm ibn ʿAlī [...] wa-munāqiḍihī fīmā ḫtalafā fīhi min anna l-ǧism ǧawhar wa-ʿaraḍ ...”, ed. Menn/ Wisnovsky317, p. 27: [Arisṭūṭālis ... qad ṣarraḥa bi-ḏālika fī qawlihī fī lmaqālati l-ūlā mina l-Samāʿ al-ṭabīʿiyyi:] wa-ḏālika anna ġayra l-mutanāhī innamā (innamā ms. Leiden : om. Ibn ʿAdī) huwa šayʾun dāḫilun fī l-kammi [2] Physics I 2, 185b32-34: πολλὰ δὲ τὰ ὄντα ἢ λόγῳ οἷον ἄλλο τὸ λευκῷ εἶναι καὶ μουσικῷ, τὸ δ’ αὐτὸ (τὸ δ’ αὐτὸ] τῷ δ’ αὐτῷ FI) ἄμφω318 Transl. Isḥāq, ed. Badawī, p. 16 = Yaḥyā ibn ʿAdī, “Maqāla fī tabyīn wuǧūd al-umūr al-ʿāmmiyya ...”, ed. Ḫalīfāt, p. 158: illā anna l-mawǧūdāti kaṯīratun immā bi-l-qawli miṯla anna l-wuǧūda li-l-abyaḍi ġayru l-wuǧūdi li-l-mūsīqār wa-l-šayʾu l-wāḥidu bi-kawnihimā (sic leg. pro yakūnu humā) ǧamīʿan [3] Physics I 3, 186a28-29: οὔτε γὰρ τῇ συνεχείᾳ ἓν ἔσται τὸ λευκὸν οὔτε τῷ λόγῳ. ἄλλο γὰρ ἔσται τὸ εἶναι λευκῷ καὶ τῷ δεδεγμένῳ Transl. Isḥāq, ed. Badawī, p. 21 = Yaḥyā ibn ʿAdī, “Maqāla fī tabyīn wuǧūd al-umūr al-ʿāmmiyya ...”, ed. Ḫalīfāt, p. 159: wa-ḏālika anna l-abyaḍa laysa huwa wāḥidan bi-l-ittiṣāli wa-lā bi-l-qawli li-anna l-wuǧūda li-l-abyaḍi ġayruhū li-l-qābili lahū [4] Physics I 3, 186b12-14: οὐ τοίνυν οὐδὲ μέγεθος ἔσται (ἔσται ms. J, Isḥāq, Ibn ʿAdī : ἕξει codd. cett.) τὸ ὄν, εἴπερ ὅπερ ὂν τὸ ὄν· ἑκατέρῳ γὰρ ἕτερον τὸ εἶναι τῶν μορίων Transl. Isḥāq, ed. Badawī, p. 26 ≈ Yaḥyā ibn ʿAdī, “Maqāla fī tabyīn wuǧūd al-umūr al-ʿāmmiyya ...”, ed. Ḫalīfāt, p. 159: wa-ayḍan fa-inna l-mawǧūda lā yakūnu ʿiẓaman in (in ms. Leiden : iḏ Ibn ʿAdī) kāna l-mawǧūdu innamā huwa (huwa ms. Leiden : yadullu ʿalā Ibn ʿAdī) llaḏī huwa l-mawǧūdu waḏālika anna maʿnā mawǧūdin li-kulli wāḥidin min ǧuzʾayhi ġayruhū (sic leg. cum Ibn ʿAdī pro ġayr ms. Leiden) li-l-āḫari [5] Physics I 8, 191b19-23: οἷον εἰ ἐκ ζῴου ζῷον γίγνοιτο καὶ ἐκ τινὸς ζῴου τι ζῷον· οἷον εἰ κύων (κύων codd., transl. Isḥāq : κύων Ross sec. Laas) ἐξ ἵππου γίγνοιτο. γίγνοιτο μὲν γὰρ ἂν οὐ μόνον ἐκ τινὸς ζῴου ὁ κύων, ἀλλὰ καὶ ἐκ ζῴου, ἀλλ' οὐχ ᾗ ζῷον· ὑπάρχει γὰρ ἤδη τοῦτο. _____________ 317

Stephen Menn, Robert Wisnovsky, “Yaḥyā ibn ʿAdī and Ibrāhīm ibn ʿAdī: On Whether Body Is a Substance or a Quantity. Introduction, Editio Princeps and Translation”, Arabic Sciences and Philosophy 27 (2017), p. 1—74. 318 For the Greek text, cf. Hermann Bonitz, “Aristotelische Studien I”, p. 186.

CXCII

Introduction

Transl. Isḥāq, ed. Badawī, p. 69 = Yaḥyā ibn ʿAdī, “Maqāla fī tabyīn wuǧūd al-umūr al-ʿāmmiyya ...”, ed. Ḫalīfāt, p. 157: miṯālu ḏālika annahū qad yatakawwanu ʿan ḥayawānin ḥayawānun wa-ʿan ḥayawānin mā ḥayawānun mā fa-in takawwana maṯalan kalbun ʿan farasin kāna takawwunu (without points ms. Leiden : yukawwinu Badawī) l-kalbi ʿan ḥayawānin mā wa-ʿan ḥayawānin muṭlaqin illā anna (anna Ibn ʿAdī : annahū ms. Leiden) takawwunahū minhu laysa min ǧihati mā huwa ḥayawānun iḏ kāna maʿnā l-ḥayawāni ḥāṣilan lahū [6] Physics I 9, 192a31-32: λέγω γὰρ ὕλην τὸ πρῶτον ὑποκείμενον ἑκάστῳ, ἐξ οὗ γίγνεταί τι ἐνυπάρχοντος μὴ κατὰ συμβεβηκός Transl. Isḥāq, ed. Badawī, p. 75: fa-innī aʿnī bi-qawlī hayūlā l-mawḍūʿa lawwala li-šayʾin šayʾini llaḏī ʿanhu yukawwanu l-šayʾu wa-huwa mawǧūdun fīhi lā bi-ṭarīqi l-ʿaraḍi Cit. Yaḥyā ibn ʿAdī, “Maqāla fī l-Mawǧūdāt”, ed. Ḫalīfāt, p. 270: fa-ammā l-hayūlā fa-hiya bi-ḥasabi mā rasamahā l-ḥakīmu Arisṭūṭālīs al-mawḍūʿu lawwalu li-šayʾin šayʾini llaḏī ʿanhu yakūnu takawwunu l-šayʾi wa-huwa mawǧūdun fīhi lā bi-ṭarīqi l-ʿaraḍi [7] Physics II 1, 192b20-23: ὡς οὔσης τῆς φύσεως ἀρχῆς τινὸς [...] τοῦ κινεῖσθαι καὶ ἠρεμεῖν ἐν ᾧ ὑπάρχει πρώτως καθ' αὑτὸ καὶ μὴ κατὰ συμβεβηκός Transl. Isḥāq, ed. Badawī, p. 79: fa-takūnu l-ṭabīʿatu mabdaʾan mā [...] li-an yataḥarraka wa-yaskuna l-šayʾu llaḏī hiya fīhi awwalan bi-l-ḏāti lā bi-ṭarīqi lʿaraḍi 319 Cit. Yaḥyā ibn ʿAdī, “Maqāla fī l-Mawǧūdāt”, ed. Ḫalīfāt, p. 269: fa-hiya (sc. al-ṭabīʿatu) mabdaʾu ḥarakatin wa-sukūnin fī l-šayʾi llaḏī hiya fīhi awwalan wa-bi-l-ḏāti lā bi-ṭarīqi l-ʿaraḍi [8] Physics II 1, 193b3-4: ἡ φύσις ἂν εἴη τῶν ἐχόντων ἐν αὑτοῖς κινήσεως ἀρχὴν ἡ μορφὴ καὶ τὸ εἶδος Transl. Isḥāq, ed. Badawī, p. 86: [inna] l-ṭabīʿata hiya ḫilqatun wa-ṣūratun limā fīhi bi-nafsihī mabdaʾun li-l-ḥarakati Cit. Yaḥyā ibn ʿAdī, “Maqāla fī l-Mawǧūdāt”, ed. Ḫalīfāt, p. 270: hiya (sc. al-ṭabīʿatu) bi-ḥasabi mā rasamahā l-ḥakīmu Arisṭūṭālīs ḫilqatun wa-ṣūratun li-mā fīhi bi-nafsihī mabdaʾu l-ḥarakati _____________ 319

A different Arabic translation of this passage is attested in the 11th c. “Philosophy reader” preserved in ms. Oxford, Bodl. Marsh 539; cf. Elvira Wakelnig, A Philosophy Reader from the Circle of Miskawayh. Edited and translated by E. W., Cambridge 2014, p. 144.18-19. For Miskawayh’s reception of Aristotle’s Physics see also ʿUbayd, ʿAlī Imām, Falsafat Miskawayh al-ṭabīʿiyya wa-l-ilāhiyya. Dirāsa wa-naqd. Al-Manṣūra 1431/ 2010, esp. p. 67–74, 176–203.

VII.1 Quotations from Isḥāq’s translation

CXCIII

[9] Physics III 5, 204b5-8: εἰ γάρ ἐστι σώματος λόγος τὸ ἐπιπέδῳ ὡρισμένον, οὐκ ἂν εἴη σῶμα ἄπειρον, οὔτε νοητὸν οὔτε αἰσθητόν, ἀλλὰ μὴν οὐδ' ἀριθμὸς οὕτως ὡς (ὡς om. Arab.) κεχωρισμένος καὶ ἄπειρος· ἀριθμητὸν γὰρ ἀριθμὸς ἢ τὸ ἔχον ἀριθμόν Transl. Isḥāq, ed. Badawī, p. 227 = Yaḥyā ibn ʿAdī, “Maqāla fī ġayri lmutanāhī”, ed. Ḫalīfāt, p. 137: in kāna maʿnā l-ǧismi huwa (huwa Ibn ʿAdī : s.l. ms. Leiden) annahū maḥdūdun bi-saṭḥin fa-laysa yumkinu an yakūna ǧismun ġayru mutanāhin wa-lā maʿqūlun wa-lā maḥsūsun (maʿqūlan ... maḥsūsan scr. Ḫalīfāt) wa-lā ʿadadun ayḍan yumkinu an yakūna mufāriqan (mufāriqan ms. Leiden, Ibn ʿAdī : mufāriqun Badawī) ġayra mutanāhin wa-ḏālika anna l-ʿadada wa-mā lahū ʿadadun fa-maʿdūdun [10] Physics VI 1, 231b16: ὅτι πᾶν συνεχὲς διαιρετόν Transl. Isḥāq, ed. Badawī, p. 609: anna kulla muttaṣilin fa-innahū munqasimun Cit. Yaḥyā ibn ʿAdī, “Maqāla fī tabyīn anna kulla muttaṣilin ...”, ed. Ḫalīfāt, p. 141: anna kulla muttaṣilin munqasimun [11] Physics VII 1, 242a6: πᾶν γὰρ τὸ κινούμενον διαιρετὸν ἦν Transl. Isḥāq, ed. Badawī, p. 734: li-annā qad bayyannā anna kulla mutaḥarrikin fa-munqasimun Cit. Yaḥyā ibn ʿAdī, “Maqāla fī tabyīn anna kulla muttaṣilin ...”, ed. Ḫalīfāt, p. 142: fa-yaǧibu min ḏālika an yakūna kullu mutaḥarrikin munqasiman320 Ibn Sīnā. Andreas Lammer correctly points out in his voluminous recent study on Ibn Sīnā’s Physics that in “many passages [. . .] Avicenna’s diction differs from that of Isḥāq ibn Ḥunayn”321. However, this by no means disproves that Ibn Sīnā was acquainted with Isḥāq’s translation and attests to his well-known modus operandi of paraphrasing or reworking his Arabic Aristotelian sources. While Ibn Sīnā may have relied in his early period on other versions322, there is good reason to assume that the (infrequent) allusions to and paraphrases of Aristotle’s Physics in his later writings ultimately refer to the translation by Isḥāq ibn Ḥunayn. Here are some examples taken from his most com_____________ 320

For further literal quotations in Ibn ʿAdī’s hitherto unedited Maqāla fī anna l-ʿadad laysa huwa ḏā nihāya cf. Endress, The Works of Yaḥyā ibn ʿAdī, p. 60. 321 Andreas Lammer, The Elements of Avicenna̕s Physics: Greek Sources and Arabic Innovations (Scientia Graeco-Arabica; vol. 20). Berlin 2018, p. 39; cf. also the examples adduced ibid., p. 39—40. 322 Cf. ibid., p. 38—39.

CXCIV

Introduction

prehensive discussion of the Physics, the first Book of the Ṭabīʿiyyāt of his Kitāb al-Šifāʾ 323: [1] Physics II 1, 192b20-23: ὡς οὔσης τῆς φύσεως ἀρχῆς τινὸς [...] καὶ αἰτίας τοῦ κινεῖσθαι καὶ ἠρεμεῖν ἐν ᾧ ὑπάρχει πρώτως καθ' αὑτὸ καὶ μὴ κατὰ συμβεβηκός Transl. Isḥāq, ed. Badawī, p. 79: fa-takūnu l-ṭabīʿatu mabdaʾan mā [...] li-an yataḥarraka wa-yaskuna l-šayʾu llaḏī hiya fīhi awwalan bi-l-ḏāti lā bi-ṭarīqi lʿaraḍi Ibn Sīnā, “al-Samāʿ al-ṭabīʿī”, p. 31.6-7: ḥuddat al-ṭabīʿatu bi-annahā mabdaʾun awwalu li-ḥarakati mā yakūnu fīhi wa-sukūnihī bi-l-ḏāti lā bi-l-ʿaraḍi sim. p. 33.2-3: [inna] l-ṭabīʿata hiya mabdaʾun mawǧūdun fī l-aǧsāmi litaḥrīkihā [...] wa-taskīnihā [...] huwa mabdaʾun awwalu li-ḥarakati mā huwa fīhi [...] bi-l-ḏāti lā bi-l-ʿaraḍi [2] Physics II 4, 196b5-7: εἰσὶ δέ τινες οἷς δοκεῖ εἶναι μὲν αἰτία ἡ τύχη, ἄδηλος δὲ ἀνθρωπίνῃ διανοίᾳ ὡς θεῖόν τι οὖσα καὶ δαιμονιώτερον Transl. Isḥāq, ed. Badawī, p. 116: wa-qad ẓanna qawmun āḫarūna anna lbaḫta huwa sababun mina l-asbābi illā annahū alṭafu min an yudrikahū ḏihnu l-insāni li-annahū amrun ilāhiyyun šarīfun Ibn Sīnā, “al-Samāʿ al-ṭabīʿī”, p. 61.5-6: wa-qad qāma [...] ṭāʾifatun uḫrā [...] fa-qāla qāʾilun minhum inna l-baḫta sababun ilāhiyyun mastūrun yartafiʿu ʿan an tudrikahā l-ʿuqūlu [3] Physics II 8, 198b10-11: Λεκτέον δὴ [...] διότι ἡ φύσις τῶν ἕνεκά του αἰτίων Transl. Isḥāq, ed. Badawī, p. 142: fa-yanbaġī l-āna an naǧʿala kalāmanā [...] fī anna l-ṭabīʿata mina l-asbābi llatī innamā tafʿalu min aǧli šayʾin Ibn Sīnā, “al-Samāʿ al-ṭabīʿī”, p. 69.16: qālū wa-kayfa takūnu l-ṭabīʿatu fāʿilatan li-aǧli šayʾin [4] Physics IV 1, 208b29-31: Ἡσίοδος [...] λέγει γοῦν “πάντων μὲν πρώτιστα χάος γένετ', αὐτὰρ ἔπειτα γαῖ’ εὐρύστερνος“ (Hesiod, Theogony 116) Transl. Isḥāq, ed. Badawī, p. 274: Īsiyūdus [...] qāla fī šiʿrihī [...] awwalu mā kāna l-faḍāʾu ṯumma baʿdahu l-arḍu l-wāsiʿatu l-ruḥbi (Hesiod, Theogony 116) Ibn Sīnā, “al-Samāʿ al-ṭabīʿī”, p. 114.7-8: Īsiyūdūs al-šāʿiru [...] qāla inna awwala mā ḫalaqa llāhu taʿālā l-makānu ṯumma l-arḍu l-wāsiʿatu _____________ 323

Ibn Sīnā, K. al-Šifāʾ. al-Ṭabīʿiyyāt 1, al-Samāʿ al-ṭabīʿī. Ed. Saʿīd Zāyid, al-Qāhira 1983. Cf. also Ibn Sīnā, al-Mubāḥaṯāt. Ed. Muḥsin Bīdārfar, Qum 1992, p. 76, with a paraphrase of Physics VI 5, 236a27-28 together with VI 8, 239a10-11.

VII.1 Quotations from Isḥāq’s translation

CXCV

[5] Physics IV 11, 219b1-2: ὁ χρόνος ἀριθμὸς κινήσεως κατὰ τὸ πρότερον καὶ ὕστερον Transl. Isḥāq, ed. Badawī, p. 420: al-zamānu huwa ʿadadu l-ḥarakati min qibali l-mutaqaddimi wa-l-mutaʾaḫḫiri Ibn Sīnā, “al-Samāʿ al-ṭabīʿī”, p. 157.6: fa-l-zamānu huwa ʿadadu l-ḥarakati iḏā nfaṣalat ilā mutaqaddimin wa-mutaʾaḫḫirin [6] Physics IV 12, 221a14-15: τὸ μὲν νῦν καὶ τὸ πρότερον καὶ ὅσα τοιαῦτα οὕτως ἐν χρόνῳ ὡς ἐν ἀριθμῷ μονὰς καὶ τὸ περιττὸν καὶ ἄρτιον Transl. Isḥāq, ed. Badawī, p. 449: fa-inna l-āna wa-l-mutaqaddima wa-sāʾira mā ašbaha ḏālika hiya fī l-zamāni ka-l-waḥdati fī l-ʿadadi wa-l-fardu wa-lzawǧu Ibn Sīnā, “al-Samāʿ al-ṭabīʿī”, p. 170.18: fa-l-āna fī l-zamāni ka-l-waḥdati fī lʿadadi wa-l-mutaqaddimu wa-l-mutaʾaḫḫiru ka-l-zawǧi wa-l-fardi [7] Physics V 6, 230b16: ἡ γὰρ ἄνω μονὴ τῇ ἄνωθεν [...] κινήσει ἐναντία Transl. Isḥāq, ed. Badawī, p. 593: wa-ḏālika anna l-labṯa fawqu ḍiddu lḥarakati min fawqu Ibn Sīnā, “al-Samāʿ al-ṭabīʿī”, p. 290.9-10: wa-qad qīla inna l-sukūna fawqu ḍiddu l-ḥarakati min fawqu [8] Physics VIII 10, 266b5-6: οὐδὲν ἄρα πεπερασμένον ἐνδέχεται ἄπειρον δύναμιν ἔχειν Transl. Isḥāq, below, p. 110: fa-laysa yumkinu iḏan an yakūna mutanāhin aṣlan lahū quwwatun ġayru mutanāhiyatin Ibn Sīnā, “al-Samāʿ al-ṭabīʿī”, p. 228.17: fa-bayyinun min hāḏā annahū lā yaǧūzu an yakūna fī ǧismin mutanāhin quwwatun ġayru mutanāhiyatin Ibn Bāǧǧa. The Andalusian philosopher Abū Bakr Muḥammad ibn Bāǧǧa dealt in a number of disparate and unhomogeneous works with Aristotle’s Physics; on the one hand, he composed a group of rather short miscellaneous notes and what seem to be unfinished treatises on particular questions or sections of the Physics, on the other hand he composed a more or less full-fledged commentary. Possibly, some of the short writings are either drafts of or supplements to related sections of the long commentary324. According to Paul Lettinck, Ibn _____________ 324

On these treatises and their manuscript transmission cf. Ǧamāl al-Dīn al-ʿAlawī, Muʾallafāt Ibn Bāǧǧa. al-Dār al-Bayḍāʾ 1983, p. 43—49, 61f., 89—93, 111f., 156—165; Paul Lettinck, Aristotle’s Physics and its Reception, p. 6—13; for Ibn Bāǧǧa on Physics VIII 1-2, cf. Farah Cherif Zahar, “L’Éternité du mouvement chez Ibn Bāǧǧa (Avempace):

CXCVI

Introduction

Bāǧǧa “did not (only) use Isḥāq’s translation”325. It is, however, possible to show that Isḥāq’s translation served as Ibn Bāǧǧa’s main source for his references to the Aristotelian text, as illustrated by the examples below326. Furthermore, his works contain dozens of rather paraphrastic references to Aristotle’s Physics, most of which coincide as regards the technical terminology with Isḥāq ibn Ḥunayn’s version. [1] Physics II 1, 192b20-22: ὡς οὔσης τῆς φύσεως ἀρχῆς τινὸς καὶ αἰτίας τοῦ κινεῖσθαι καὶ ἠρεμεῖν ἐν ᾧ ὑπάρχει Transl. Isḥāq, ed. Badawī, p. 79: fa-takūnu l-ṭabīʿatu mabdaʾan mā wa-sababan li-an yataḥarraka wa-yaskuna l-šayʾu llaḏī hiya fīhi Cit. Ibn Bāǧǧa, “Šarḥ al-Samāʿ al-ṭabīʿī”, ed. Ziyāda, p. 22.4-5: fa-l-ṭabīʿatu iḏan hiya sababu l-ḥarakati wa-l-sukūni li-mā hiya fīhi wa-l-sababu huwa mabdaʾun fa-l-ṭabīʿatu mabdaʾu l-ḥarakati wa-l-sukūni [2] Physics III 1, 201a10-11: ἡ τοῦ δυνάμει ὄντος ἐντελέχεια, ᾗ τοιοῦτον, κίνησίς ἐστιν Transl. Isḥāq, ed. Badawī, p. 171: al-ḥarakatu hiya kamālu mā bi-l-quwwati bi-mā huwa ka-ḏālika Cit. Ibn Bāǧǧa, “Šarḥ al-Samāʿ al-ṭabīʿī”, ed. Ziyāda, p. 31.3-4: Arisṭū ḥadda l-ḥarakata ḥīna qāla innahā kamālu bi-l-quwwati min ǧihati mā huwa bi-l-quwwati ka-ḏālika327 [3] Physics V 1, 225a3-4: μεταβάλλοι ἂν τὸ μεταβάλλον [...] ἢ [...] ἐξ ὑποκειμένου εἰς ὑποκείμενον Transl. Isḥāq, ed. Badawī, p. 502: wa-kāna mā yataġayyaru fa-innamā yataġayyaru [...] immā min mawḍūʿin ilā mawḍūʿin Cit. Ibn Bāǧǧa, “Šarḥ al-Samāʿ al-ṭabīʿī”, ed. Ziyāda, p. 55.3: inna l-taġayyura immā an yakūna min mawḍūʿin ilā mawḍūʿin [4] Physics V 2, 226a26-27: ἡ μὲν οὖν κατὰ τὸ ποιὸν κίνησις ἀλλοίωσις ἔστω Transl. Isḥāq, ed. Badawī, p. 529: fa-l-ḥarakatu fī l-kayfi la-takun istiḥālatan De la définition générique à la définition numérique. Le commentaire aux chapitres 1 et 2 du livre VIII de la Physique”, Les Études philosophiques 162 (2016), p. 161—216. 325 Lettinck, Aristotle’s Physics and its Reception, p. 12. 326 Quotations from Ibn Bāǧǧa’s commentary on the Physics are taken from: Šuruḥāt al-Samāʿ al-ṭabīʿī li-bn Bāǧǧa al-Andalusī. Taḥqīq wa-taqdīm Maʿn Ziyāda. Bayrūt 1978; and Lettinck, Aristotle’s Physics and its Reception, p. 681—747. 327 Where the second bi-l-quwwati is in all likelihood an interpolated explanation of kaḏālika; cf. ed. Ziyāda, p. 52.5, Lettinck, Aristotle’s Physics and its Reception, p. 737.22.

VII.1 Quotations from Isḥāq’s translation

CXCVII

Cit. Ibn Bāǧǧa, “Šarḥ al-Samāʿ al-ṭabīʿī”, ed. Ziyāda, p. 51.8: wa-l-ḥarakatu fī l-kayfi yuqālu lahā stiḥālatun [5] Physics V 2, 226a29-31: ἡ δὲ κατὰ τὸ ποσὸν τὸ μὲν κοινὸν ἀνώνυμος, καθ’ ἑκάτερον δ’ αὔξησις καὶ φθίσις Transl. Isḥāq, ed. Badawī, p. 530: wa-l-ḥarakatu fī l-kammi ammā l-ʿāmmu minhā fa-ġayru musamman wa-ammā wāḥidun wāḥidun minhā fa-yusammā numuwwan wa-naqṣan Cit. Ibn Bāǧǧa, “Šarḥ al-Samāʿ al-ṭabīʿī”, ed. Ziyāda, p. 54.18-19: wa-l-ḥarakatu fī l-kammi fa-laysa lahā smun yaǧmaʿu ṭarafayhā bal ṭarafun minhā yusammā numuwwan wa-l-āḫaru yusammā naqṣan [6] Physics V 2, 226b10-11: τό τε ὅλως ἀδύνατον κινηθῆναι, ὥσπερ ὁ ψόφος ἀόρατος Transl. Isḥāq, ed. Badawī, p. 533—534: allaḏī laysa yumkinu fīhi l-ḥarakatu aṣlan miṯla anna l-ṣawta ġayru marʾiyyin Cit. Ibn Bāǧǧa, “Šarḥ al-Samāʿ al-ṭabīʿī”, ed. Ziyāda, p. 55.17: mā laysa šaʾnuhū an yataḥarraka ka-mā yuqālu fī l-ṣawti annahū ġayru marʾiyyin [7] Physics VI 2, 233a7-8: διαιρήσει γὰρ τὸ μὲν θᾶττον τὸν χρόνον, τὸ δὲ βραδύτερον τὸ μῆκος Transl. Isḥāq, ed. Badawī, p. 623: wa-ḏālika anna l-asraʿa yaqsimu l-zamāna wa-l-abṭaʾa yaqsimu l-ṭūla Cit. Ibn Bāǧǧa, “Šarḥ al-Samāʿ al-ṭabīʿī”, ed. Ziyāda, p. 80.3: li-anna l-abṭaʾa yaqsimu l-masāfata wa-l-asraʿa yaqsimu l-zamāna [8] Physics VI 4, 234b11: ἔκ τινος εἴς τι πᾶσα μεταβολή. Transl. Isḥāq, ed. Badawī, p. 648: kullu taġayyurin innamā yakūnu min šayʾin ilā šayʾin Cit. Ibn Bāǧǧa, “Šarḥ al-Samāʿ al-ṭabīʿī”, ed. Ziyāda, p. 96.2: [Arisṭū ... qāla:] kullu taġayyurin (sic leg.) fa-huwa min šayʾin ilā šayʾin (sic leg.) [9] Physics VI 4, 234b17-18: λέγω δ’ εἰς ὃ μεταβάλλει τὸ πρῶτον κατὰ τὴν μεταβολήν, οἷον [...] τὸ φαιόν, οὐ τὸ μέλαν Transl. Isḥāq, ed. Badawī, p. 648: wa-aʿnī bi-qawlī llaḏī ilayhi yataġayyaru lawwala fī l-taġayyuri miṯālu ḏālika [...] al-adkanu lā l-aswadu Cit. Ibn Bāǧǧa, “Šarḥ al-Samāʿ al-ṭabīʿī”, ed. Lettinck, p. 742 ult. — 743.1: wa-aʿnī bi-qawlī mā ilayhi yataġayyaru l-awwala mina l-taġayyuri miṯālu ḏālika l-adkanu (addidi) l-aswadu

CXCVIII

Introduction

[10] Physics VI 5, 236a29-30: πᾶν ... δέδεικται διαιρετὸν τὸ μεταβάλλον (cf. 234b10) Transl. Isḥāq, ed. Badawī, p. 677: qad bayyannā anna kulla mutaġayyirin fa-huwa munqasimun Cit. Ibn Bāǧǧa, “Šarḥ al-Samāʿ al-ṭabīʿī”, ed. Ziyāda, p. 96.1: [fa-aqūlu inna] kulla mutaġayyirin fa-huwa munqasimun [11] Physics VII 3, 246a24-25: φανερὸν ὡς ἐν μόνοις τοῖς αἰσθητοῖς ἡ ἀλλοίωσις Transl. Isḥāq, ed. Badawī, p. 759: fa-ẓāhirun anna fī l-maḥsūsāti waḥdahā takūnu l-istiḥālatu Cit. Ibn Bāǧǧa, “Min kalāmihī fī-mā yataʿallaqu bi-l-nuzūʿiyya”328, p. 132.9: wa-qad tabayyana fī l-sābiʿati mina l-Samāʿi anna l-istiḥālata innamā takūnu fī l-maḥsūsāti [12] Physics VIII 1, 250b11-12: Πότερον γέγονέ ποτε κίνησις οὐκ οὖσα πρότερον, καὶ φθείρεται πάλιν οὕτως ὥστε κινεῖσθαι μηδέν Transl. Isḥāq, below, p. 1: layta šiʿrī hal ḥadaṯati l-ḥarakatu wa-lam takun qablu wa-hal tafsudu ayḍan fa-iḏan lā yakūnu maʿahū šayʾun aṣlan yataḥarraku Cit. Ibn Bāǧǧa, “Šarḥ al-Samāʿ al-ṭabīʿī”, ed. Ziyāda, p. 193 ult. — 194.2: layta šiʿrī hal al-ḥarakatu ḥadaṯat baʿda an lam takun wa-hal tafnā fanāʾan lā yakūnu maʿahū šayʾun yataḥarraku aṣlan (cf. ed. Ziyāda, p. 150 ult.) [13] Physics VIII 3, 253a29-30: τὰ δ’ ἀεὶ κινούμενα, τὰ δ’ ἀμφοτέρων μεταλαμβάνειν Transl. Isḥāq, below, p. 21: baʿḍuhā yataḥarraku abadan wa-baʿḍuhā yatadāwalu l-amrayni ǧamīʿan Cit. Ibn Bāǧǧa, “Šarḥ al-Samāʿ al-ṭabīʿī”, ed. Ziyāda, p. 163.19-20: baʿḍu lašyāʾi yataḥarraku ḍarūratan wa-baʿḍuhā yatadāwalu l-amrayni [14] Physics VIII 4, 254b12-13: τῶν δὲ καθ’ αὑτὰ τὰ μὲν ὑφ’ ἑαυτοῦ Transl. Isḥāq, below, p. 29 = Ibn Bāǧǧa, “Šarḥ al-Samāʿ al-ṭabīʿī”, ed. Ziyāda, p. 167.10: wa-l-mutaḥarrikatu bi-ḏawātihā baʿḍuhā min tilqāʾihā [15] Physics VIII 6, 259a12: ἱκανὸν δὲ καὶ ἕν Transl. Isḥāq, below, p. 61 = Ibn Bāǧǧa, “Šarḥ al-Samāʿ al-ṭabīʿī”, ed. Ziyāda, p. 216.1: wa-fī wāḥidin kifāyatun _____________ 328

Ed. Ǧamāl al-Dīn al-ʿAlawī, Rasāʾil falsafiyya li-Abī Bakr ibn Bāǧǧa. Nuṣūṣ falsafiyya ġayr manšūra. al-Dār al-Bayḍāʾ 1983, p. 121—134.

VII.1 Quotations from Isḥāq’s translation

CXCIX

ʿAbd al-Laṭīf ibn Yūsuf al-Baġdādī. According to Ibn Abī Uṣaybiʿa, this polymath composed at least three works dealing with Aristotle’s Physics329, none of which is extant today. He was also very much interested in the epistemological relationship between physics and metaphysics (criticising, among other things, Ibn Sīnā’s approach to the specific subject matters of these sciences330). From his extant book on Aristotle’s Metaphysics (Kitāb fī ʿIlm mā baʿd alṭabīʿa331) it becomes clear that he had access to Isḥāq ibn Ḥunayn’s translation of the Physics, as illustrated by the following paraphrastic excerpts: [1] Physics II 5, 197a31-32: οὔτε γὰρ ἀεὶ οὔθ' ὡς ἐπὶ τὸ πολὺ οἷόν τ' εἶναι τῶν ἀπὸ τύχης οὐθέν Transl. Isḥāq, ed. Badawī, p. 125: li-anna mā yakūnu bi-l-baḫti ... laysa minhu šayʾun yumkinu an yakūna dāʾiman wa-lā fī akṯari l-amri Cit. al-Baġdādī, “Mā baʿda l-ṭabīʿa”, ed. Y. Aǧʿūn332, p. 129.17-19: [wa-qad tabayyana fī l-Samāʿ anna] l-ašyāʾa ... llatī bi-l-ʿaraḍi lā yumkinu an takūna dāʾimatan wa-lā ʿalā l-akṯari [2] Physics III 1, 201a10-11: ἡ τοῦ δυνάμει ὄντος ἐντελέχεια, ᾗ τοιοῦτον, κίνησίς ἐστιν Transl. Isḥāq, ed. Badawī, p. 171: al-ḥarakatu hiya kamālu mā bi-l-quwwati bi-mā huwa ka-ḏālika Cit. al-Baġdādī, “Mā baʿda l-ṭabīʿa”, p. 166.10-11: [qīla inna] l-ḥarakata hiya kamālu mā bi-l-quwwati min ǧihati mā huwa ka-ḏālika [3] Physics III 7, 207b10-11: ἐπὶ δὲ τὸ πλεῖον ἀεὶ ἔστι νοῆσαι· ἄπειροι γὰρ αἱ διχοτομίαι τοῦ μεγέθους _____________ 329

(1) On “Natural things [al-ṭabīʿiyyāt], from the Physics up to the end of De sensu et sensibilis, in three fascicles [muǧalladāt]”; (2) “Physics [al-Samāʿ al-ṭabīʿī]”, in two fascicles; (3) On “Natural things, from the Physics up to De anima”; cf. Ibn Abī Uṣaybiʿa, ʿUyūn al-anbāʾ fī ṭabaqāt al-aṭibbāʾ, ed. August Müller, Cairo 1299/1882, 2 vols., vol. 2, p. 212—213. ʿAbd al-Laṭīf al-Baġdādī is also mentioned (ibid., p. 213) as the author of a comprehensive summa (al-ǧāmiʿ al-kabīr) of Logic, Natural science (al-ʿilm al-ṭabīʿī) and Divine science (al-ʿilm al-ilāhī). 330 Cf. Cecilia Martini Bonadeo, ʿAbd al-Laṭīf al-Baġdādī’s Philosophical Journey. From Aristotle’s Metaphysics to the ‘Metaphysical Science’. Islamic Philosophy, Theology and Science. Texts and Studies, vol. 88. Leiden, Boston 2013, p. 183—186. 331 On which see Martini Bonadeo, ʿAbd al-Laṭīf al-Baġdādī’s Philosophical Journey, p. 209ff., and the literature mentioned there. 332 ʿAbd al-Laṭīf ibn Yūsuf al-Baġdādī, Mā baʿda l-ṭabīʿa. Qaddama lahū wa-ḥaqqaqahū Yūnus Aǧʿūn. Bayrūt: Dār al-kutub al-ʿilmiyya, 1438/2017.

CC

Introduction

Transl. Isḥāq, ed. Badawī, p. 264.2-4: wa-ammā ḏahābuhū ilā l-kaṯrati faqad yumkinu tawahhumuhū dāʾiman fa-inna qismata l-miqdāri bi-niṣfayni waniṣfihī bi-niṣfayni tamurru (sic leg.) bi-lā nihāya Cit. al-Baġdādī, “Mā baʿda l-ṭabīʿa”, p. 190.5-6: [qad ṣarraḥa bi-ḏālika Arisṭū fī l-ṯāliṯati mina l-Samāʿi fa-qāla:] inna qismata l-muttaṣili (fort. al-miqdāri vel l-muttaṣili legendum) bi-niṣfayni wa-niṣfihī bi-niṣfayni qad yumkinu tawahhumuhū dāʾiman [4] Physics VIII 3, 253b8-9: ἡ φύσις ... ἀρχή, καθάπερ κινήσεως, καὶ ἠρεμίας Transl. Isḥāq, below p. 22: al-ṭabīʿatu ... mabdaʾun ka-miṯli mā li-l-ḥarakati li-l-sukūni ayḍan Paraphr. al-Baġdādī, “Mā baʿda l-ṭabīʿa”, p. 72.16: al-ṭabīʿatu mabdaʾu lḥarakati wa-l-sukūni

VII.2 Arabo-Latin versions of Isḥāq ibn Ḥunayn’s translation Even more important for the critical edition than the indirect Arabic transmission is the Latin reception of Isḥāq ibn Ḥunayn’s Arabic translation333. Two complete Arabo-Latin translations are extant today. The earlier one was prepared by Gerard of Cremona (d. 1187), probably in Toledo334. The later one forms part of Michael Scot’s (d. ca. 1234) translation of Averroes’ (Ibn Rušd’s) so-called “Long Commenary” (Šarḥ) on the Physics, which quotes in the textus _____________ 333

This takes place partly contemporaneously with the medieval Graeco-Latin reception of the Physics. For a general overview of the three Graeco-Latin versions known as translatio vetus, translatio nova and translatio Vaticana, and their different recensions cf. Jozef Brams, Gudrun Vuillemin-Diem, “Physica Nova und Recensio Matritensis. Wilhelm von Moerbekes doppelte Rezension der Physica Vetus”, in A. Zimmermann (ed.), Aristotelisches Erbe im arabisch-lateinischen Mittelalter: Übersetzungen, Kommentare, Interpretationen (Miscellanea mediaevalia; vol. 18). Berlin 1986, p. 215—288; Fernand Bossier, Jozef Brams, Physica. Translatio Vetus — Praefatio (Aristoteles Latinus; vol. VII,1, Fasc. 1). Leiden 1990, p. xi—lxxix. 334 For Gerard of Cremona’s works and translations see: Charles Burnett, “The Coherence of the Arabic-Latin Translation Program in Toledo in the Twelfth Century”, Science in Context 14 (2001), p. 249—288; Art. “Gherardo da Cremona”, in Dizionario Biografico degli Italiani, vol. 53 (2000), accessible via http://www.treccani.it/enciclopedia/ gherardoda-cremona_(Dizionario-Biografico); for further literature cf. Cristina D’Ancona, “Alexander of Aphrodisias, De unitate: A Pseudepigraphical Testimony of the De unitate et uno by Dominicus Gundissalinus”, in A. Akasoy, W. Raven (eds.), Islamic Thought in the Middle Ages. Studies in Text, Transmission and Translation, in Honour of Hans Daiber (Islamic Philosophy, Theology and Science. Texts and Studies; vol. 75). Leiden 2008, p. 459—488, p. 461, note 7.

VII.2 Arabo-Latin versions of Isḥāq’s translation

CCI

preceding the comments proper Isḥāq’s text335. Finally, a number of manuscripts transmit together with Scot’s translation of Ibn Rušd’s work fragments of a third, anonymous version, which cover the complete seventh book of the Physics and the sections 265a27—b16 and 266b6—267b26 of bk. VIII together with Ibn Rušd’s comments. An in-depth recensio of the Latin text witnesses goes far beyond the scope of the present work. The following brief notes provide no more than some basic information on the sources I used for the comparison of the three translations with the Arabic text as well as some succint samples of style and terminology. An attempt at clarifying the relationships between the Latin translations and the Arabic ms. Leiden Or. 583 is made in the following section (chapter VII.3). C Gerard of Cremona’s translation, which has not yet been edited, is preserved in seven or eight partially incomplete manuscripts336. Unlike the other two translations, it contains only the Aristotelian text (without Ibn Rušd’s commentary or, for that matter, any comments by the Baghdad Aristotelians). For the present edition I collated in detail Gerard’s translation of Physics bk. VIII as preserved in the following three manuscripts337: _____________ 335

On Michael Scot’s life and works in general see the “Preface” by Gerhard Endress in Averrois Cordubensis commentum magnum super De celo et mundo Aristotelis. Ex recognitione Francis James Carmody † in lucem edidit Rüdiger Arnzen (Recherches de Théologie et Philosophie médiévales. Bibliotheca, vol. 4), 2 vols. Leuven 2003, vol. 1, p. 16*—23*, and the literature mentioned there; for Michael Scot’s translation technique see the numerous studies on Aristotle’s zoological books by Aafke van Oppenraay. That the textus of Ibn Rušd’s commentary quote Isḥāq ibn Ḥunayn’s translation has been shown by Paul Lettinck, Aristotle’s Physics and its Reception, p. 180—184. 336 For a general introduction to Gerard’s translation of the Physics cf. Auguste Mansion, “Note sur les traductions arabo-latines de la Physique d’Aristote dans la tradition manuscrite”, Revue néo-scolastique de philosophie 37 (1934), p. 202—218; Bossier & Brams, Physica. Translatio Vetus, p. li—liv. For the eight manuscripts see Aristoteles Latinus. Codices descripsit G. Lacombe in societatem operis adsumptis A. Birkenmajer, M. Dulong, Aet. Franceschini (Corpus Philosophorum Medii Aevi). Pars 1. Rome 1939, repr. Bruges—Paris 1957, p. 13, 51—52, 125—126, 263—265, 283, 286—287, 496, 565; Aristoteles Latinus. Codices descripsit G. Lacombe in societatem operis adsumptis A. Birkenmajer, M. Dulong, Aet. Franceschini, supplementis indicibusque instruxit L. Minio-Paluello (Corpus Philosophorum Medii Aevi). Pars 2. Cambridge 1955, p. 881, 1015—1016, 1095; Aristoteles Latinus. Codices. Supplementa altera ed. L. Minio-Paluello (Corpus Philosophorum Medii Aevi). Bruges—Paris 1961, p. 140—141. 337 I am most grateful to Paul Lettinck (University of Malaya, Kuala Lumpur), Charles Burnett (Warburg Institute, London) and Cristina D’Ancona (University of Pisa) for putting digital copies of the three manuscripts at my disposal.

CCII

Introduction

Ca Aosta, Biblioteca del Seminario Maggiore, cod. 71. The manuscript is written in two columns and dates from the end of the 12th or beginning of the 13th c. Apart from Aristotle’s Physics it contains the Liber De causis, ps.Aristotle’s De plantis and Aristotle’s De Caelo and Meteorology. The codex was first described by A. Franceschini, who realised that the folios containing Physics VIII and the beginning of the Liber De causis were mixed up338. In August 1967, H. J. Drossaart Lulofs inspected the manuscript and made a handwritten note with further details of this disorder and a proposal for its restoration (a sheet of paper with his note is today glued on the inner surface of the upper cover of the codex). According to Drossaart Lulofs the correct order of the folios containing Physics VIII is as follows: 23*, 24*, 27*, 26*, 32*, 28*, 29*, 25*. In the meantime his proposal has been put into action. After the restoration these folios received a second foliation following the new order, i.e. what was fol. 27* is now fol. 25, 26* = 26, 32* became 27, 28* = 28, 29* = 29, 25* became 30. In the below edition I refer to the new foliation after restoration, that is, fols. 23vb—30ra for Physics VIII339. Cp Paris, Bibliothèque nationale de France, lat. 16141 (olim Sorbonne 936)340. The manuscript, which dates from the 13th c., became famous for the fact that it transmits in three parallel columns Aristotle’s Physics in the translations by James of Venice (the vetus translatio), Michael Scot and Gerard of Cremona. Gerard’s translation is copied in the right column. Due to the fact that it is sometimes more verbose than the other two versions, the scribe used here and there blank space in the middle column (containing Scot’s translation) or even the overall breadth of all three columns for the transcription of two or three lines of Gerard’s translation, in order to keep the three translations aligned. Bk. VIII of the Physics is found on fols. 171r— 217v. v C Vienna, Österreichische Nationalbibliothek, lat. 234341. The manuscript dates likewise from the 13th c. and contains a number of Aristotelian and _____________ 338

Aristoteles Latinus. Codices. Pars 2, p. 881. For further details on this manuscript cf. Charles Burnett, “The Arabo-Latin Aristotle”, in Aafke M. I. van Oppenraay, with the collaboration of R. Fontaine (ed.), The Letter before the Spirit: The Importance of Text Editions for the Study of the Reception of Aristotle (Aristoteles semitico-latinus; vol. 22). Leiden 2013, p. 95—105. 340 The manuscript is accessible online via http://gallica.bnf.fr/ark:/12148/btv1b9076633j/. 341 Cf. also Academia Caesarea Vindobonensis (ed.), Manuscripta mediaevalia. Tabulae codicum manu scriptorum praeter graecos et orientales in Bibliotheca Palatina Vindobonensi asservatorum. 8 vols., Wien: Gerold, 1864-1899, vol. 1, p. 33 (accessible also via https:// archive.org/). 339

VII.2 Arabo-Latin versions of Isḥāq’s translation

CCIII

pseudo-Aristotelian works, written in one column. Bk. VIII of Gerard’s translation of the Physics is found on fols. 66v—84r. V Michael Scot’s translation of Ibn Rušd’s “Long Commentary” on the Physics gives Isḥāq’s version divided into lemmata as quoted by Ibn Rušd (d. 1198). The text is extant in more than ninety (partially incomplete) manuscripts and a number of early prints (15th-16th c.)342. For the comparison of Scot’s translation of Physics VIII with the Arabic text of Isḥāq ibn Ḥunayn, I relied on the critical edition of the Latin text prepared by Horst Schmieja343. Having prepared sample collations of all extant manuscripts, Schmieja chose nine representative codices as the basis of his edition, accompanied by systematic collations with the early print of Venice 1562 apud Iunctas and desultory collations with further manuscripts. All discrepancies are reported in a comprehensive critical apparatus. In the margins of his edition, Schmieja gives (among other information) a concordance to the Bekker lines of the Aristotelian text, which is very helpful for the comparison of the Arabic translation. In an article published in 1923 Auguste Mansion studied the relationship between V and the extant Greek witnesses of the direct and indirect traditions344. To this article, which focuses on bks. I—IV of the Physics, we owe the pivotal finding that V (viz. Isḥāq ibn Ḥunayn’s translation) (a) is closely related to the Greek ms. E, (b) occasionally agrees with the indirect tradition against all witnesses of the direct tradition, and (c) in rare cases is contaminated by readings of Greek manuscripts belonging to branch γ. However, as far as individual textual problems are concerned, Mansion’s analysis must be treated with due caution, as it relies solely on the Latin text as accessible in the 1562 print _____________ 342

A descriptive inventory of all 97 Latin manuscripts is accessible in the “Digital Averroes Research Environment” (DARE) of the Thomas Institute of the University of Cologne (cf. http://dare.uni-koeln.de/). The same database also provides online access to most of the manuscripts as well as to two early prints (Venice 1483, Andreas Torresanus, and Venice 1562, Apud Iunctas [Lucas Antonius Iuncta]). For other early prints cf. Ferdinand Edward Cranz, A Bibliography of Aristotle Editions 1501 - 1600. Second ed. with addenda and revisions by Charles B. Schmitt (Bibliotheca Bibliographica Aureliana; vol. 38). Baden-Baden 1984. For some Arabic fragments of the commentary extant in ms. Madrid, Biblioteca Nacional, ár. 5000, fols. 4a-7b (related to Physics 190ab, 192a-b, and 194b) cf. Josep Puig Montada, “Fragmentos del gran comentario de Averroes a la Física”, Al-Qanṭara 30 (2009), p. 69—81. 343 My deepest thanks go to Horst Schmieja who kindly granted me access to his unpublished materials and generously shared his knowledge about the Latin manuscripts in general and difficult passages of the text in particular. Schmieja’s edition shall be published by the Institute for the History of Arabic-Islamic Science in the series Islamic Philosophy at the end of the year 2020. 344 Cf. above, footnote 170.

CCIV

Introduction

of Venice. This procedure evidently entails a number of methodological problems: (a) The Arabic text quoted by Ibn Rušd in the lemmata of his commentary often contains internal defects of the Arabic transmission from the 9th century translation up to Ibn Rušd’s times. (b) Even where Ibn Rušd did rely on a correct text of the Graeco-Arabic translation, this text may have been corrupted within the Arabic transmission from Ibn Rušd’s autograph up to the exemplar used by Michael Scot for his Latin translation. (c) At many places, Michael Scot did not understand the Arabic text adequately. Hence, the 1562 print of Venice cannot be taken to be a reliable witness to the original Arabic version and is —taken by itself— of no use for the clarification of intricate questions concerning the relationship between Isḥāq ibn Ḥunayn’s Greek exemplar and the extant Greek manuscripts345. (d) What is more, the Venice edition apud Iunctas (or even Michael Scot’s Latin text itself) is contaminated from the Graeco-Latin tradition of the Physics. As a consequence of these methodological shortcomings, Ross relied in his Greek edition at a remarkable number of places on false conclusions drawn by Mansion on the basis of a corrupt or contaminated Latin text. Some three decades later, Mansion revisited the issue, now extended by the consideration of Gerard of Cremona’s translation (C) as compared to Michael Scot’s one346. Apart from proving once again the utility and exploitability of C and V for the evaluation of the Greek testimonies, Mansion tries there for the first time to show that C and V derive from different copies of the Arabic translation. Of the 21 cases he discusses with this intent, 19 are of minor conclusiveness, because they consist of omissions or inversions. The former may have occurred as late as during the Latin transmission and do not admit inferences to the Arabic exemplars, while the latter may be caused by the style or interpretational license of the translators. The remark on Physics VII 2, 243a3 (Mansion, p. 83) concerns the confusion of muḥarrik and mutaḥarrik, which is likewise not convincing, as the two words may just as well have been misread or confounded by one of the translators as by the scribe of an Arabic manuscript. _____________ 345

To give just one example: Concerning Physics II 4, 196a25, Mansion (Étude critique, p. 36) concludes from the word mundanorum in Scotus’ translation as printed in Venice 1562 apud Iunctas (textus 44) that V (viz. Isḥāq’s translation) is based on the reading κοσμικῶν (mss. FIJ) as opposed to κόσμων in ms. E and the indirect tradition. A comparison with the Arabic manuscript reveals that this reading must be based on a contamination from the Graeco-Latin vetus translatio, which reads likewise mundanorum, whereas Isḥāq’s translation (al-ʿawālim, ed. Badawī, p. 114.14) corresponds exactly with ms. E. 346 A. Mansion, „Le texte d’Aristote Physique H, 1—3 dans les versions arabo-latines“, The Journal of Hellenic Studies 77 (1957), p. 81—86.

VII.2 Arabo-Latin versions of Isḥāq’s translation

CCV

What remains is one moderately conclusive case with Michael Scot’s translation reading “in respectu nobilissimi animalium”, where the Arabic has bi-lqiyāsi ilā afḍali l-aḥwāli (πρὸς τὸ ἄριστον, 246b23 = “per comparationem ad meliores dispositiones” C). Obviously, Michael Scot’s Arabic exemplar read here al-ḥayawān instead of al-aḥwāl. Nevertheless Mansion’s conclusion claiming distinct Arabic exemplars of C and V is tentatively confirmed by the comparison below347. Z The existence of fragments of a third Latin version of Isḥāq’s Arabic translation was for the first time brought to our attention by Horst Schmieja348. So far, three manuscripts are known to preserve different parts of this anonymous translation349. The Vindobonensis lat. 2334 contains the complete seventh book of the Physics and the sections 265a27—b16 and 266b6—267b26 of bk. VIII together with the related sections of Ibn Rušd’s “Long Commentary” in this translation. Two further manuscripts, Erfurt, Bibliotheca Amploniana Fol. 350 and Vatican, Palatini lat. 1035, contain only the sections of bk. VIII together with Ibn Rušd’s commentary. Whether only these parts were translated once again in order to fill gaps in the transmission of Michael Scot’s translation or whether we are faced with extracts excerpted from another complete translation of Ibn Rušd’s work is not known. _____________ 347

Nonetheless, the above-mentioned methodological qualms apply also to this second study by Mansion. To give another example: Misled by the Latin translations Mansion lists as an example of “added specification lacking in the Greek” (la traduction arabe [...] paraît avoir ajouté des précisions absentes du grec) the following: “246b24: περὶ τὴν φύσιν secundum rem naturalem puram G [= Gerard of Cremona] quae est naturalis pura M [Michael Scot]” (Le texte d’Aristote Physique H, p. 83). However, pura/puram in the translations by Gerard of Cremona and Michael Scot is by no means an addition absent from the Greek text, but the result of a (semantically and syntactically) wrong interpretation of the Arabic word ḫāliṣan in Isḥāq’s translation (ed. Badawī, p. 760.15). The Arabic word can indeed mean “pure, mere”, but renders at the present place faithfully τὸ σῶζον in the phrase immediately before the one quoted by Mansion (λέγω δὲ τὸ βέλτιστον τὸ σῶζον καὶ διατιθὲν περὶ τὴν φύσιν, 246b23-24). 348 Horst Schmieja, “Secundum aliam translationem — Ein Beitrag zur arabisch-lateinischen Übersetzung des großen Physikkommentars von Averroes”, in G. Endress, J. A. Aertsen, with the assistance of K. Braun (eds.), Averroes and the Aristotelian Tradition: Sources, Constitution and Reception of the Philosophy of Ibn Rushd (1126-1198). Proceedings of the Fourth Symposium Averroicum (Cologne, 1996). (Islamic Philosophy Theology and Science. Texts and Studies, vol. 31.) Leiden 1999, p. 316—336. 349 The manuscripts and their contents are described in Horst Schmieja (ed.), [Averrois] Commentarium magnum in Aristotelis Physicorum librum septimum (Vindobonensis, lat. 2334) (Averrois Opera. Series B. Averroes Latinus = Abhandlungen der NordrheinWestfälischen Akademie der Wissenschaften, vol. 115). Paderborn 2007, p. xviii—xxvii.

CCVI

Introduction

In his edition of bk. VII of this translation, Schmieja shows through a detailed comparison of the style and terminology in the two translations first that the anonymous translation is not a revision of an earlier Latin version, but an independent translation prepared on the basis of the Arabic text, and secondly that it displays occasionally influences from Michael Scot’s translation and, hence, must be later than this350. According to Schmieja, the translation technique and terminology point to Hermannus Alemannus (Herman the German, d. 1272) as the potential author of this version351. For the present edition, I took into consideration the above-mentioned sections of bk. VIII of this third Latin translation, a critical edition of which is found in Schmieja’s upcoming edition of Michael Scot’s translation (cf. above). Unless otherwise indicated, all quotations from these two translations (V & Z) draw on the texts as edited by Schmieja. The three Latin translations differ remarkably regarding style, terminology and degree of literalness, as might be illustrated by the below table. However, the section chosen for this purpose provides simultaneously one out of numerous examples of their usefulness for the establishment of the correct Arabic wording. I give first Isḥāq’s translation of Physics VIII 9, 265a27-32 (cf. below, p. 102.6– 103.2), then, in tabular form, the three corresponding Arabo-Latin versions: wa-bi-l-wāǧibi lazima an takūna l-ḥarakatu dawran wāḥidatan muttaṣilatan llatī ʿalā l-istiqāmati wa-ḏālika anna l-mustaqīma fa-inna l-ibtidāʾa maḥdūdun fīhi wa-l-inqiḍāʾa wa-l-wasaṭa wa-kullu ḏālika mawǧūdun fīhi (reading with mss. EK1beLNRH ἐν αὐτῇ instead of ἐν αὑτῇ or ἐν ἑαυτῇ in the remaining mss.) wa-li-ḏālika qad yabtadiʾu l-mutaḥarriku fīhi min mawḍiʿin wa-lā tatimmu (reading with mss. Ebe οὐ τελευτήσει instead of οὗ τελευτήσει in the remaining mss.) ḥarakatuhū wa-ḏālika anna kulla mā yataḥarraku ʿalayhi fahuwa yaskunu ʿinda ṭarafihī immā min ḥayṯu btadaʾa wa-immā ḥayṯu huwa.

Gerard of Cremona

Michael Scot

Anonymous

Et necessario quidem sequitur ut motus circularis sit unus continuus non qui est secundum

Et recte contingit ut motus circularis sit unus et continuus, non motus rectus, in recto

Necesse autem fuit motus circularis unus esse et continuus, non qui sit recte, in recto

_____________ 350

Cf. Schmieja, [Averrois] Commentarium magnum, p. 85—135. On Hermann the German cf. Frédérique Woerther, Le plaisir, le bonheur, et l’acquisition des vertus. Édition du Livre X du Commentaire moyen d’Averroès à l’Éthique à Nicomaque d’Aristote (Islamic Philosophy, Theology and Science. Texts and Studies, vol. 108), p. 8–16, and the literature mentioned there. 351

VII.2 Arabo-Latin versions of Isḥāq’s translation

CCVII

Gerard of Crem. (cont.)

Michael Scot (cont.)

Anonymous (cont.)

rectitudinem, quod est quia in recto principium est determinatum et finis et medium et totum illud existit in ipso. Et propterea illud incepit quod movetur in eo ex loco et non completur motus eius quod est quia apud extremitatem omnis super quod movetur est quiescens aut unde incepit aut ubi (ubi CaCp : non Cv) est.

enim principium est terminatus et finis et medium, et omnia ista sunt existentia in eo. Et ideo possibile est ut motum incipiat in eo ab aliquo loco et non compleatur suus motus, omne enim quod movetur super ipsum quiescit apud extrema aut ubi incoepit aut ex quo est.

enim principium distinctum et medium et finis, et omnia haec insunt ei. Et ideo incipit quod movetur per ipsum ex loco et non perficit motus, omne enim quod per ipsum movetur, quiescit in extremis aut non aut quo est.

Apart from terminological differences, we observe that Gerard imitates quite slavishly the Arabic word order, while the other two translators try to achieve a more fluid Latin and deviate from it. Gerard even adopts the asyndeton wāḥidatan muttaṣilatan (= unus continuus, without et) and renders (here and consistently throughout the Physics) the conjunctional phrase wa-ḏālika anna intricately through quod est quia352. Michael Scot alone is sensitive to the modal force of qad in qad yabtadiʾu (rendering fut. ἄρξεται = possibile est ut ... incipiat)353. While Gerard misinterprets ʿalayhi in the last sentence of the quoted section as referring back to mā (the correct antecedent of which is al-mustaqīm, sc. al-ḫaṭṭ al-mustaqīm), he is the only one who renders correctly the distinction min ḥayṯu ... ḥayṯu at the end of the section (possibly aut non in the anonymous translation is based on a failed resolution of an abbreviated form of ubi or unde, coinciding with an omission of incepit). What is important for the Arabic text is the fact that all three translations support the addition of lā before allatī (corresponding to μὴ at 265a28). The lexical diversity of the three translations may be illustrated by the following examples taken from the above-mentioned sections of Physics bk. _____________ 352

On Gerard’s translation technique cf. Ilona Opelt, “Zur Übersetzungstechnik des Gerhard von Cremona”, Glotta 38 (1960), p. 135—170. 353 However, that qad is not reflected in the other two translations does not license the conclusion that it was omitted in their Arabic exemplars.

CCVIII

Introduction

VIII. Note also the gradually decreasing degree of literalness from Gerard of Cremona through Michael Scot to the Anonymous:

Greek

Arabic

Gerard

ἀμφοτέρως c. neg.

lā ʿalā wāḥidin mina l-waǧhayni wāǧibun ḍarūratan iḫtalafa izdāda buʿdan bayyana

aliquorum duorum modorum c. neg. oportet necessario diversificare addere elongationem ostendere

ἐγγύτατος

aqrabu

propinquior

ἐκδεχόμενον ἄλλο παρ’ ἄλλου ἐλάττων γίγνεται ἐχόμενος καλῶς

tadāwala wāḥidun ʿan āḫara intaqaṣa

permutatur unum ab alio

mutašāfiʿun lā baʾsa an

κατέχω c. acc.

lāzimun c. li-

λέγω c. dupl. acc. ὁ αὐτὸς λόγος

nasaba c. acc. et ilā ǧarā hāḏā lmaǧrā biʿaynihī ʿalā miṯālin wāḥidin

ἀνάγκη ἀνωμαλῶς ἀφίστημι πλεῖον δείκνυμι

ὁμοίως πάλαι παύομαι ῥίπτω συνεχής substantivized participle

ānifan kaffa ʿan ramā tawālin šayʾ + verb

Michael Scot Anonymous secundum utrumque modum c. neg. debet de necessitate diversari magis removeri declarare

neutro modo

minui

dicere, probare magis appro- vicinus pinquare succedit vicissim unum post accipiunt aliud singuli diminui deficio

continuus non est malum ut adicere c. dat.

successivus bonum est c. inf. tenere c. acc.

proportionare c. acc. et ad currere eodem cursu

consecutus nihil nocet c. inf. observari c. acc. ascribere c. dat. et acc. sic se habere

attribuo c. dat. et acc. procedere eodem processu eodem modo similiter

secundum similitudinem unam nuper stare ab proicere consequens res + adj.

superius quiescere ab sagittare successio qui + verb

necesse est differe distare

paulo ante cessare ab eicere consecutio aliquid + adj.

VII.3 The relations btw. the Latin versions and the Arabic MSS

CCIX

On the other hand, these examples should not obscure the fact that the three translators agree in large part in their choice of the technical teminology.

VII.3 The stemmatic relationship between the Latin translations, ms. Leiden Or. 583 and ms. Esc. ár. 896 The following observations are mainly based on my collation of the abovementioned Latin sources of Physics bk. VIII with the corresponding section of the Arabic ms. Leiden Or. 583. To put it bluntly, they link in a rather uncritical way stemmatic observations proper with glimpses of the real history of the tradition, for which Jean Fourquet coined the term arbre réel 354. The below results should therefore not be taken to aim at a stemma codicum in the strict sense, but rather as a schematic summary of selective insights into this “real tree” enriched now and then with truely stemmatic data. Taking into consideration that according to Schmieja’s research on Michael Scot’s translation of Ibn Rušd’s “Long Commentary” the manuscript tradition of this (very comprehensive) work is marked by recurrent changes of exemplars of one and the same extant manuscript as well as by changing scribes, any statement below about the relationships between the Arabic and Latin witnesses must be taken as being limited to bk. VIII of the Physics and does not necessarily hold true for the other books. Accordingly, the combination of our above findings concerning the relation between ms. Leiden Or. 583 and the fragments in ms. Esc. 896 (chapter II.4) with the below observations on Physics VIII is, for the same reasons as well as in view of the fact that Abū l-Ḥusayn al-Baṣrī changed his exemplars, rather speculative. Finally, the results concerning the Anonymous translation are based exclusively on the limited number of rather short sections of bk. VIII mentioned above and should be used with due caution. In what follows, I shall use the above introduced sigla for the Latin translations, that is, C for Gerard of Cremona’s translation, V for Michael Scot’s one and Z for the Anonymous. As mentioned earlier, C contains only the Aristotelian text, whereas V and Z transmit this together with Ibn Rušd’s commentary. Taking additionally into consideration the fact that the commentaries by the Baghdad Aristotelians have left no traces at all in Ibn Rušd’s work, we may assume that both Arabic manuscripts, the one used by Gerard of Cremona (henceforth: car) and that quoted by Ibn Rušd (henceforth: vzar) contained (apart from glosses and marginal or interlinear notes) only Isḥāq’s translation of the Aristotelian text _____________ 354

J. Fourquet, “Le paradoxe de Bédier”, Mélanges 1945. Tome 2: Études littéraires, Strasbourg 1946, p. 1–16. For the distinction between arbre réel and stemma codicum see also Trovato, Lachmann’s Method, p. 65—67, 80, 92, 144—146.

CCX

Introduction

to the exclusion of any full-fledged commentary sections derived from the Baghdad school compilation. By implication, the same must pertain to the text of the Physics as quoted in the exemplars used by Michael Scot and the Anonymous, both of which descended ultimately from Ibn Rušd’s text. The exact dates of Ibn Rušd’s work and Gerard’s translation are not known. As an approximate terminus ante quem for car and vzar we can determine the middle of the 1180’s355, that is, about 180 years after the compilation prepared by Abū l-Ḥusayn al-Baṣrī and about 300 years after Isḥāq’s translation. As shown above, Abū l-Ḥusayn al-Baṣrī drew on heterogeneous precursors, all of which contained Isḥāq’s translation together with comments and/or scholia by different members of the Baghdad school. Such precursors may date back to the late 930’s, the period of Abū Bišr Mattā’s teaching in Baghdad, or the middle of the 10th century. As for the delicate question whether car and/or vzar were representatives of a consistently independent transmission of the Aristotelian text only (and thus might originate from this early period of transmission) or derive from an Arabic copy whose scribe had extracted Isḥāq ibn Ḥunayn’s version from one or the other version of these 10th c. Baghdad school compilations, the extant witnesses provide pros and cons for both positions. Interestingly, vzar reflects here and there the influence of marginal notes or scholia also transmitted in ms. Leiden Or. 583. Two short examples have already been mentioned above (cf. chapter IV, examples [5] and [6], p. XCIX–C). Two more comprehensive contaminations are found in the passages Physics VIII 3, 253b13-14 and VIII 10, 267a1-2. [1] Physics VIII 3, 253b13-14: οὔτε γὰρ αὐξάνεσθαι οὔτε φθίνειν οἷόν τε συνεχῶς, ἀλλ’ ἔστι καὶ τὸ μέσον A. Arabic transmission (with superscript marks in textu and in margine) Transl. Isḥāq, below, p. 23

Ms. Leiden, fol. 210a, in the margin of Isḥāq’s translation

fa-innahū laysa yumkinu an yakūna l-šayʾu yanmī wa-lā an yaḍmaḥilla dāʾiman bal baynahumā ayḍani l-amru lwasaṭu*.

* ay bayna l-numuwwi wa-l-iḍmiḥlāli amrun awsaṭu huwa llaḏī ilayhi yantahī l-numuwwu waminhu yabtadiʾu l-iḍmiḥlālu.

* i.e. between growth and decay is something intermediate, namely that in which growth comes to an end and from which decay begins.

_____________ 355

Gerard of Cremona died in 1187. Ibn Rušd’s commentary possibly dates from the early 1180’s; cf. Ǧamāl al-Dīn al-ʿAlawī, Al-Matn al-rušdī: Madḫal li-qirāʾa ǧadīda. alDār al-bayḍāʾ 1983, p. 21—22; Ruth Glasner, Averroes’ Physics. A Turning Point in Medieval Natural Philosophy. Oxford 2009, p. 19—21.

VII.3 The relations btw. the Latin versions and the Arabic MSS

CCXI

B. Latin transmission Gerard of Cremona

Michael Scot

Non enim est possibile ut aliquid augeatur nisi ut minuatur semper immo inter utraque est res media.

Impossibile est enim aliquid augeri aut diminui semper sed inter ea est medium, scilicet illud ad quod pervenit augmentum et a quo incipit diminutio.

The addition in Scotus’ translation corresponds exactly to the second half of the marginal gloss in ms. Leiden Or. 583. For Physics VIII 3, the translation of the Anonymous is not extant. The source of this gloss is, in all likelihood, Philoponus, who explains the Aristotelian sentence as follows (In Arist. Phys. 826.8-10): ἔστι γὰρ καὶ τὸ τούτων , ὃ μήτε αὔξησίς ἐστι μήτε φθίσις, εἰς ὃ ἡ αὔξησις φθάσασα ἵσταται, λέγω δὲ τὸ κατὰ φύσιν μέτρον ἐξ οὗ καὶ ἡ φθίσις ἄρχεται. [2] Physics VIII 10, 267a1-2: καὶ εἰ ποιεῖ (sc. ὁ κινήσας) ὥσπερ ἡ λίθος, οἷον κινεῖ (οἷον κινεῖ mss. E1FIJ : οἷον κινεῖν mss. be1LNRHQ2 : οἷόν τε κινεῖν mss. E2Ke2, Ross) ὃ ἐκίνησεν: A. Arabic transmission (with superscript marks in textu and in margine) Transl. Isḥāq, below, p. 113

wa-ka-ḏālika in kāna (sc. al-muḥarriku) ayḍan yafʿalu ka-mā yafʿalu lḥaǧaru* miṯālu ḏālika anna llaḏī ḥarrakahū + yuḥarriku.

Ms. Leiden, fol. 231b, in the margin of Isḥāq’s translation

* yaʿnī l-maġnīṭīs

* he means the magnet

+ ay

+ i.e.

that which has allaḏī ḥarrakahu lḥaǧaru wa-huwa l-ḥadī- been moved by the du yuḥarriku bi-an yaǧ- stone, that is the iron, moves by attracting anḏiba l-ḥadīda l-āḫara. other [piece of] iron.

B. Latin transmission Gerard of Cremona

Michael Scot

Anonymous

Et similiter etiam est [...] si agat (sc. movens) sicut agit lapis sicut si illud quod movet illud

Similiter [...] si etiam facit (sc. motor) sicut facit lapis scilicet magnetes verbi gratia

Similiter etiam si agit (sc. motor) sicut agit magnes, verbi gratia quod illud quod movet

CCXII

Gerard of Crem. (cont.)

Introduction

Michael Scot (cont.)

movetur (reading ta- quoniam illud quod ḥarraka instead of yu- movet lapis quod est ḥarriku). ferrum idest attrahit ferrum a quo movetur et sic movet.

Anonymous (cont.) lapis scilicet ferrum movet, idest attrahit ferrum quod movet, unde movet.

Again, the marginal note in ms. Leiden Or. 583 derives in all likelihood from a late antique commentary. Similar explanations are found in the commentaries by Philoponus (In Arist. Phys., p. 851) and Simplicius (In Arist. Phys., p. 1345). Michael Scot’s Arabic exemplar and, by implication, the exemplar of Isḥāq’s translation used by Ibn Rušd must ultimately derive from an Arabic copy containing these scholia. Whether they had already been interpolated in the text of Isḥāq’s translation in this copy, that is prior to Ibn Rušd’s commenting on his Arabic exemplar, or were incorporated by Ibn Rušd himself, we do not know. On the other hand, the fact that the interpolation occurs at exactly the same place in Michael Scot’s and the Anonymous’ Latin translations suggests that it must have been completed in the Arabic version of Ibn Rušd’s commentary, that is, prior to any Latin translation. In both cases the Latin translations correspond almost literally to the Arabic notes and are incorporated exactly at the two places marked in the Leiden ms. by superscript signs. Gerard of Cremona’s translation, on the other hand, displays no traces of the scholia. Either they did not figure in the exemplar of car, or the scribe of car neglected to copy them, or Gerard refrained from incorporating them into his translation. Two further contaminations of single words deserve to be mentioned. Isḥāq translates ἐν ἀδιαφόρῳ κατ’ εἶδος, Physics VIII 8, 262a2, with fī-mā lā yaḫtalifu bi-l-ṣūrati (“in what admits no differentiation with respect to the form”). In the Leiden manuscript (fol. 223b13) we read above bi-l-ṣūrati between the lines al-nawʿ (“the species”), evidently an alternative rendering of εἶδος. Whether this alternative was noted down by Isḥāq himself356 or by a later learned scribe, is difficult to assess. In any case, it must have been present in vzar, as we read in Michael Scot’s translation of this phrase “in eo qui non diversatur secundum formam, scilicet secundum speciem” (Averroes, textus 64)357. _____________ 356

Who demonstrably used both terms, depending on the context, in order to render εἶδος (cf. below, Glossary; also Gutas, Theophrastus, p. 418b; Ullmann, NE 1, p. 405). 357 Nothing of the like is found in Gerard’s translation, while the Anonymous is not extant.

VII.3 The relations btw. the Latin versions and the Arabic MSS

CCXIII

A similar case is found at Physics VIII 8, 262b21-23: ἐνταῦθα μὲν οὖν ἀδύνατον οὕτως λέγειν ἐπὶ τῆς συνεχοῦς· ἐπὶ δὲ τοῦ ἀνακάμπτοντος ἀνάγκη λέγειν οὕτως. “In this case, then, where [we are faced with] the continuous [motion], it is not possible to say something of the like, whereas in the case of that which turns back, it is necessary to say something of the like.” fa-qad bāna fī l-muttaṣilati annahū lā yumkinu an yuqāla ḏālika wa-ammā fī llaḏī yaʿṭifu fa-yarǧiʿu fa-qad yaǧibu an yuqāla ḏālika (transl. Isḥāq, below, p. 86) “It is, then, clear with respect to the continuous [motion] that it is not possible to say this, whereas in the case of that which turns and goes back, it is necessary to say this.” Above the first ḏālika (“this”, rendering οὕτως, 262b21) the Leiden manuscript (fol. 225a1) has yaʿnī l-wuqūf (“i.e., coming to a halt”). The source of this note –added by some Baghdad scholar– seems to be Alexander of Aphrodisias, who explains regarding the first part of Aristotle’s dichotomy: ἡ προκειμένη ἀπορία τοῦτο συνάγειν βούλεται, ὅτι οὐκ ἀνάγκη τὸ γενόμενον ἔν τινι καὶ ἀπογενόμενον ἀπ’ αὐτοῦ ἠρεμηκέναι “La présente aporie veut conclure qu’il n’est pas nécessaire que ce qui est survenu en quelque chose et qui quitte cette chose se soit trouvé en repos.”358 The commentaries by Philoponus (In Arist. Phys., p. 844.1-3) and Simplicius (In Arist. Phys., p. 1286.15ff.) interpret οὕτως, 262b21, as referring to coming to and departing from an (intermediate) thing / point, yet do not mention “coming to a standstill”. Michael Scot’s translation of the passage (Averroes, textus 66) runs as follows: Manifestum est igitur quod in continuo impossibile est ut hoc dicatur. In reflexo vero necesse est ut hoc dicatur, scilicet quies. Evidently, scilicet quies reflects the Arabic note yaʿnī l-wuqūf. That it has been related to the second ḏālika (“this” in “it is necessary to say this”, rendering οὕτως, 262b23) can be explained either through the assumption that the superscript mark attached to the first occurrence of ḏālika was moved erroneously by the scribe of vzar to the second ḏālika, or through Michael Scot’s (or even Ibn Rušd’s) interpretative licence, as it makes no difference in terms of content or doctrine to which of the two it is added. _____________ 358

Text and translation Rashed, Alexandre d’Aphrodise, frag. 741, p. 600—601.

CCXIV

Introduction

As for Gerard of Cremona’s translation, we find fewer traces of interpolated glosses. However, it contains in bk. VIII of the Physics two diagrams transmitted also in the Leiden manuscript. The first case occurs in the margin of Isḥāq’s translation of Physics VIII 8, 263b15-16, χρόνος ἐφ’ ᾧ ΑΓΒ, πρᾶγμα ἐφ’ ᾧ Δ. In the margin of the Arabic version (ms. Leiden Or. 583, fol. 225b16‫ﺍ ﺟ ﺏ‬ 17), the scribe reproduced the op‫ﺩ‬ posite diagram: Exactly the same diagram, mirrorimaged due to the reverse writing directions of Arabic and Latin, is found in the margin of ms. Ca, fol. 28va, next to Gerard’s translation:

a

c d

b

A similar case, slightly more complex, is Physics VIII 10, 266b9-12. Isḥāq’s translation of this section replaces Aristotle’s Greek letters with corresponding Arabic letters and runs as follows (for the Arabic text see p. 111): “Let AB be infinite. Then BǦ has a certain power through which (reading like ms. K ᾗ instead of ἣ or ἦ in the other mss.) it moves D in a certain time. Let this time correspond to HZ. Now, if we take BǦ twice, it will move it in half of the time HZ. Let this go on analogically, and let the time be (probably reading ἔστω instead of ὥστε) ZṬ. Now, by continually taking [i.e., a greater magnitude] in this way, we shall never ever arrive at AB.” In the margin of this section, we find in the Leiden manuscript (fol. 231a) the opposite diagram (based on the letters used by Isḥāq): Again, we find the same diagram mirror-imaged in Gerard’s translation, now in the margin of two manuscripts (Ca, fol. 29va, Cv, fol. 83r [partly cut off]):

‫ﺍ ﺟ ﺏ‬ ‫ﺩ‬ ‫ﻫ ﻃ ﺯ‬ a e

c d t

b z

The above cases may be taken as evidencing that car and vzar ultimately descend from an Arabic manuscript whose scribe had excerpted the Aristotelian text together with a number of marginal notes and diagrams from a more com-

VII.3 The relations btw. the Latin versions and the Arabic MSS

CCXV

prehensive compilation containing text, comments and scholia by the Baghdad Aristotelians. However, such a conclusion is not compelling. What is clear from the above cases is that vzar and the Baghdad compilation as preserved in the Leiden manuscript drew on a common source containing scholia which were then interpolated into Isḥāq’s text at an unknown stage of the transmission. Analogically, car and ms. Leiden Or. 583 must go back to a common source which contained the two above diagrams in its margins. Yet this common source may just as well have consisted in an early copy of Isḥāq ibn Ḥunayn’s translation without the comprehensive commentary corpus of the Baghdad school. As we have seen above, here and there Isḥāq himself added marginal notes to his translation, both those derived from scholia in his Greek exemplar and those conceived by himself as a note or a piece of advice for the Arabic reader. Other scholia may have been added during the course of the separate transmission of Isḥāq’s translation prior to the first stages of reception among the Baghdad Aristotelians, i.e. between the date of Isḥāq’s work and the 930’s. That the above mentioned diagrams were already present in Isḥāq’s Greek exemplar and transcribed by the latter is suggested by the fact that the passage Physics VIII 10, 266b9-12 in the 10th c. Parisinus gr. 1853 (ms. E), fol. 67r, is likewise accompanied by a diagram very similar to the one just mentioned359:

Furthermore, the following considerations speak against the hypothesis that a compilation of text, scholia and commentaries similar to the one preserved in ms. Leiden Or. 583 served as a source of car and/or vzar: Due to the frequent and in many cases unmarked bidirectional changes between (a) separate Aristotelian lemmata, (b) comments, (c) repeated quotations of Aristotle within the comments, and (d) separate sections providing alternative versions of an Aristo_____________ 359

As most of the scholia on the Physics preserved in ms. E originate from the hand of the 10th c. scholiast E2, this concurrence points to the fact that not only E1 and Ψ derive ultimately from a common source of the α-branch, but E2 must also have had access to a manuscript or manuscripts belonging to this branch. For the date of E2 and its sources cf. Paul Moraux, “Le Parisinus graecus 1853 (ms. E) d'Aristote”, Scriptorium 21 (1967), p. 17—41; Myriam Hecquet-Devienne, “Les mains du Parisinus graecus 1853: Une nouvelle collation des quatre premiers livres de la Métaphysique d'Aristote (folios 225v—247v)”, Scrittura e civiltà 24 (2000), p. 103—171, esp. p. 126—132.

CCXVI

Introduction

telian passage in the Baghdad compilation, one would expect at least occasional confusions of these different types of materials in car and vzar, resp. in C, V and Z. To my knowledge, cases of such confusion are not attested, in any case not in bk. VIII of the Physics. A similar argument concerns contaminations of Aristotelian sections and glosses in the Baghdad compilation. As shown above (cf. chapter IV, example 1, p. XCIV–XCV), the structure of such contaminations is at some places so intricate and tightly interwoven that it is almost impossible to filter out the original wording of Isḥāq’s translation without recourse to the Greek text. Again, one would expect to find traces of such contamination in car and vzar, which is not the case. Finally, we must take into account the frequency of marginal glosses and scholia. In the Leiden manuscript, the Arabic translation of Physics VIII is accompanied by more than 180 such notes (not counting marginal notes on the commentary sections!). Only a handful of them made their way down to car and vzar. The most plausible explanation for this phenomenon is that the scholia attested in one way or the other in both traditions originate from an exemplar prior to the activity of the Baghdad Aristotelians, while the bulk of scholia were added in a copy of Isḥāq’s translation which belonged to another branch of witnesses than the ancestors or exemplars of either car or vzar. That such a “pre-Baghdadian” exemplar (or exemplars) indeed existed has become clear above (chapter II.4) from our examination of the fragments preserved in ms. Esc. 896. Hence, both car and vzar descend probably from manuscripts containing Isḥāq ibn Ḥunayn’s translation together with a rather limited number of glosses and scholia, yet not an abundance of notes and glosses, as found in the Baghdad compilation, nor any integral commentary sections. It goes without saying that this does not necessarily imply that the manuscripts from which car and vzar derive date from a period prior to the copy used by the Baghdad Aristotelians. However, as shown above (chapter II.4), car and vzar occasionally give the correct reading, where the two extant Arabic manuscripts, Esc. 896 and Leiden Or. 583 display common errors. This may point to the fact that their ancestor(s) date(s) from a period prior to Θ, the common ancestor of these two Arabic manuscripts (on Θ, cf. above, p. LXXVIII–LXXXI). The stemmatic relationship between the three Latin versions C, V and Z and the copy of Isḥāq’s translation used by the Baghdad Aristotelians is, to say the least, complicated. The situation is additionally made worse by the methodological problem that any conclusion to be drawn on the basis of particular readings (errors) in C, V and Z with respect to their Arabic exemplars car and vzar suffers from the uncertainty about whether this reading reflects a corresponding reading in these exemplars or is the result of a faulty or failed translation (e.g., a misread or misinterpreted yet correct Arabic text, a line acciden-

VII.3 The relations btw. the Latin versions and the Arabic MSS

CCXVII

tally skipped by the translator, or a word or phrase deliberately not translated360). The only exception to this uncertainty is formed by those errors in C, V and Z that can be verified also in the Arabic codex unicus. Thus, most of the below observations on car and vzar offer no more than tendential statements of probability corroborated by the total of the collation results, yet uncertain as regards the individual textual disparities. In what follows, I refer to the text of Isḥāq ibn Ḥunayn’s translation as preserved in ms. Leiden Or. 583 by the siglum Φ. V and Z depend ultimately on the same Arabic text vzar quoted by Ibn Rušd, but may have used different copies of Ibn Rušd’s commentary. As mentioned above, Z is extant only for the short sections 265a27—b16 and 266b6—267b26. These sections provide no hints for significant discrepancies between the Arabic exemplars of V and Z, while the two translations display remarkable differences as regards stylistic and terminological variations361. Other disparities may have been caused by mishaps within the Latin transmission and cannot be attributed beyond doubt to different Arabic texts. We can therefore suppose that anything stated about the relation between Φ and V holds also true —in general— for other sections of bk. VIII in Z (assuming that the extant sections of the Anonymous translation were extracted from a complete version of this book). 1. Φ shares a number of mistakes with CV (resp. CVZ): p. 19.2362 ḥīnan ... ḫāriǧan ʿanhu (= ὁτὲ μὲν ᾖ τὸ κινῆσον ἔξωθεν, ὁτὲ δὲ μή 253a3) om. hom. Φ, om. CV p. 19.4 wa-huwa ... mawǧūdun Φ = et est CV instead of huwa ... mawǧūdun (= ὄντος 253a5) p. 19.11 ḥ-r-k-h Φ = motus (ḥarakatun) CV instead of ḥarakatihī (= τούτου ... τῆς κινήσεως 253a12-13) p. 72.9 bi-l-azaliyyati ... yaqdiru/taqdiru ʿalā Φ = cum perpetuitate potest C = secundum ... aeternitatem ... potest V instead of al-azaliyyatu ... taqdiru ʿalā (= τοῖς ... ἀϊδίοις ... ἐνδέχεται c. inf. 260b29-30) p. 83.9 fa-yakūnu ... huwa l-zamānu (= χρόνος ἄρα ἔσται ὁ ἐν μέσῳ 262b2-3) om. hom. Φ, om. CV _____________ 360

Not to mention mistakes occurring after a complete and correct Latin translation in the course of its manuscript tradition. 361 In one case, we encounter a saut du même au même in V, while Z reflects a complete Arabic text; cf. p. 112.6-7, where Michael Scot’s translation shows the omission of fī ʿiẓamin mutanāhin fa-takūnu by homoioteleuton. Whether this omission is due to Michael Scot himself or to his Arabic exemplar, we do not know. In any case, it cannot be attributed to vzar. 362 Page and line references are to the edition below.

CCXVIII

Introduction

p. 85.3 ya-/taḫtalifu Φ = diuersificaretur C = diuersaretur V instead of yabṭuʾu (= ὑστερίζει 262b16) p. 94.2 ṣāra Φ = fuit CV instead of wa-ṣāra (= καὶ ἐγίγνετο 264a6) p. 94.9 innamā Φ = non ... nisi C : in principio (= awwalan ?) V instead of ayḍan (= καὶ1 264a12)363 p. 95.3-4 al-muntaqilu min A in kāna ... ʿāda fa-ṣāra Φ = quod permutatur ex A si ... redit C = quod transfertur ex A si ... revertatur V instead of almuntaqilu min A kāna ... ʿāda fa-ṣāra (= τὸ ... ἀπὸ τοῦ Α ... φερόμενον ... πάλιν ἥξει 264a14-15) p. 107.6 al-ḥarakatu l-azaliyyatu (without diacritical points) Φ = motus perpetuus C = motus aeternus V instead of al-ḥarakatu l-awwaliyyatu (= πρώτη κίνησις 266a7-8) p. 119.1 immā hāḏā Φ = hoc aut C = aut quod V = hunc aut Z instead of hāḏā (= αὐτὸ 267b11, immā was erroneously inserted here due to its omission in the preceding line). Some of the above mistakes can be regarded as conjunctive errors. Since neither Φ served as an exemplar of car or vzar nor the other way around, Φ, car and vzar must ultimately derive from a common source (henceforth: Ω). On the assumption that car and vzar were not the results of an extraction of the Aristotelian text from a more comprehensive Baghdad school compilation, Ω must be prior to the Baghdad scholars’ activities of compiling and commenting, i.e. before ca. 930. The frequency and types of mistakes suggest that Ω was a rather faulty copy of Isḥāq ibn Ḥunayn’s text, certainly not his autograph. Considering the results of our comparison of Φ and ms. Esc. 896, the common ancestor of the latter two (Θ) must be later than Ω, while car and vzar probably derive either directly from Ω or by means of a descendant of Ω that was independent from Θ. This general picture is confirmed by at least two errors shared by both Φ and ms. Esc. 896 (viz. their common source Θ) on the one hand, and car and vzar on the other hand, errors, that is, which in all likelihood trace back to the faulty manuscript Ω: Physics IV 10, 218a1-2 ἐκ δὲ τούτων καὶ ὁ ἄπειρος καὶ ὁ ἀεὶ λαμβανόμενος χρόνος σύγκειται The scribe of Θ copied Isḥāq’s translation as follows: wa-min hāḏayni yatarakkabu l-zamānu ġayru (ġayru Esc. 896 : ʿalā Φ) l-mutanāhī l-maʾḫūḏu dāʾiman (Esc. 896 fol. 49a11; Φ fol. 92a22) _____________ 363

For innamā as scribal error for ayḍan v. also GALex, vol. I, p. 492.

VII.3 The relations btw. the Latin versions and the Arabic MSS

CCXIX

This reading is confirmed by Gerard’s and Michael Scot’s Latin translations: car “Et ex his duobus componitur tempus infinitum acceptum semper” (Cv fol. 41a9) vzar “Et ex his duobus componitur tempus infinitum quod invenitur semper” (transl. Scotus, liber IV, textus 88) All versions point to the fact that at least the conjunction wa- (“and”), if not a more comprehensive phrase such as wa-l-zamānu between al-zamānu ġayru l-mutanāhī and al-maʾḫūḏu dāʾiman had been omitted already in Ω. Assuming an omission of ὁ2 (or καὶ ὁ) in the translator’s exemplar does not provide a satisfying alternative explanation, for in that case Isḥāq ibn Ḥunayn would have rendered the first καὶ almost certainly by ayḍan. Physics IV 10, 218b9-10 δοκεῖ μάλιστα κίνησις εἶναι καὶ μεταβολή τις ὁ χρόνος Θ

qad yuẓannu aḥrā l-umūri bi-l-zamāni an yakūna ḥarakatan wa-taġayyuran (taġayyuran Esc. 896 : taġayyur Φ) (Esc. 896 fol. 50b9-10; Φ fol. 94a17)

car

“putatur quod dignior res temporis est ut sit motus et mutatio” (Cv fol. 41b15-16)

vzar “existimatur quod dignius est ut sit (scil. tempus) motus et transmutatio” (transl. Scotus, liber IV, textus 95) τις lacks an equivalent in all versions. The omission of mā after the accusative ending -an in taġayyuran is a well-known phenomenon and much more likely than the omission of τις, which is preserved in all extant Greek manuscripts, in the translator’s exemplar. A third potential candidate is Physics VI 10, 218b12, where the whole AraboLatin tradition omits an equivalent of καὶ μεταβάλλον, which is attested in all Greek manuscripts, while its omission in the Arabic tradition is easily explained through the syntactic vagueness and complexity of the phrase ‫ﻳﻜﻮن ذﻟﻚ‬

‫اﳌﺘﻐﲑ< ﻧﻔﺴﻪ‬ ّ > ‫اﳌﺘﺤﺮك‬ ّ .

2. That Φ derived from Ω in such a way that it cannot have served as an exemplar of either car or vzar is shown by many separative errors. I mention here only the most severe omissions in Φ. All readings in CVZ in the following list are supported by the Greek witnesses, some of them additionally by the indirect Arabic tradition (references are to the edition below): p. 2.1 mawǧūdatun om. Φ : habent CV p. 2.2 yumkinuhum an om. Φ : habent CV

CCXX

Introduction

p. 3.4 ǧaʿala om. Φ : habent CV p. 3.5 tataḥarraku om. Φ : habent CV p. 7.7 wāḥidun ... šayʾun om. hom. Φ : habent CV p. 16.3-4 hiya ... aṣlan om. hom. Φ : habent CV p. 21.3 minhā sākinatun om. hom. Φ : habent CV p. 25.3-4 fī mawāḍiʿihā ... qasran om. Φ : habent CV p. 33.7 li-l-ḥārri om. Φ : habent CV p. 34.6 al-ʿālimu llaḏī lahū om. Φ : habent CV p. 36.8–37.1 ʿāʾiqun ... yaʿūquhū om. hom. Φ : habent CV p. 42.8 an takūna ... yaǧibu om. hom. Φ : habent CV p. 44.2 in lam om. Φ : habent CV p. 44.9–45.1 bal kullu wāḥidin ... miṯālu ḏālika anna om. Φ : habent CV p. 45.10 iḏ ... yataḥarraku om. hom. Φ : habent CV p. 46.2 al-taḥrīku fa-šaʾnuhū om. hom. Φ : habent CV p. 47.2 ayḍan om. Φ : habent CV p. 51.6-7 ḏātihī ... min2 om. hom. Φ : habent CV p. 58.4-5 mutaḥarrikatin ... ġayru om. hom. Φ : habent CV p. 58.11 mutaḥarrikatan om. Φ : habent CV p. 60.8 mutanāhiyatin om. Φ : habent CV p. 65.6–66.2 hāhunā ... an yakūna om. hom. Φ : habent CV p. 68.1 min sāʾiri l-ašyāʾi ... mutaḥarrikan om. hom. Φ : habent CV p. 68.6 azaliyyin ... mutaḥarrikin om. hom. Φ : habent CV p. 69.8 innamā ... yanmī om. hom. Φ : habent CV p. 71.1 wa-ammā ... allaḏī om. Φ : habent CV p. 71.9 wa-innamā ... sarmadan om. hom. Φ : habent CV p. 78.9 wa- om. hom. Φ : habent CV p. 82.1 ḥarakata l-dāʾirati ... annahū om. Φ : habent CV p. 82.2 wāḥidan ... ʿalā nafsihī om. Φ : habent CV p. 86.1-2 lā fī l-zamāni om. hom. Φ : habent CV p. 99.4 yataḥarraku om. Φ : habent CV p. 100.8-9 wa-taḍmaḥillu ... istiḥālatan om. Φ : habent CV p. 102.3 al-uḫaru om. Φ : habent CV p. 102.6 lā om. Φ : habent CV p. 106.3-4 mina l-ūlā ... innamā takūnu om. hom. Φ : habent CV p. 107.5 al-ḥarakata om. Φ : habent CV p. 108.8 Z fa-in om. hom. Φ : habent CV p. 110.5-6 kulli maḥdūdin ... naqaṣa ʿalā om. hom. Φ : habent CV p. 110.6-7 fa-tuḥarraku ... ġayri l-mutanāhiyati om. Φ : habent CV p. 114.4 wa-li-ḏālika ... yuḥarriku om. hom. Φ : habent CVZ p. 115.9 ašyāʾu om. Φ : habent CVZ

VII.3 The relations btw. the Latin versions and the Arabic MSS

CCXXI

The above list could be easily expanded by more than one hundred individual errors of other types in Φ as opposed to correct readings in CVZ. Neither car nor vzar can derive from Φ. Apart from proving the independent descent of Φ this list shows also the high frequency of omissions, especially of sauts du même au même, in the Leiden manuscript364. 3. That neither Φ nor vzar (resp. VZ) are descendants of car is suggested by the following selection of individual mistakes in C/car: (a) Words or phrases omitted in car (unless overlooked/ignored by the translator) p. 15.2 abadan1 om. C : habent ΦV p. 39.2-3 fa-ammā ... mina l-aḫīri om. C : habent ΦV p. 55.7 B̅ C instead of A̅B̅ ΦV p. 57.3 wāḥidun (unum) C instead of kullu wāḥidin ΦV (b) Sauts du même au même p. 30.2 ammā ... ṭabʿīhā om. hom. C : habent ΦV p. 71.8 mutaqaddiman ... fa-lammā kāna om. hom. C : habent ΦV p. 106.1 wa-ḏālika ... fī l-makāni om. hom. C : habent ΦV (c) Words mistranscribed in car (unless misread by the translator) p. 26.1 mawǧūdun awwalu/awwalan (existens primum) C instead of numuwwun wa-lā ΦV p. 26.1 ḥarrakahū qaṣdan (moueat ipsum secundum intentionem) C instead of ḥarakatun qasran ΦV p. 27.1 aw ḥīnan (aut quandoque) C instead of wa-ḥīnan ΦV p. 32.4 wa-aqūlu (et dico) C instead of wa-l-qawlu ΦV p. 35.5 anna l-nāra (quia ignis) C instead of anna l-bārida ΦV p. 63.7 an naqbala vel an yuqbala (ut accipiatur) C instead of an naʿmala ΦV (d) Inversion p. 8.5 hāḏā šaʾnuhu l-taḥarruku wa-hāḏā šaʾnuhu l-taḥrīku C instead of hāḏā šaʾnuhu l-taḥrīku wa-hāḏā šaʾnuhu l-taḥarruku ΦV _____________ 364

It goes without saying that the non-Aristotelian contents of the manuscript (scholia, commentary sections, etc.) are no less affected by the scribe’s carelessness than the text of Isḥāq ibn Ḥunayn’s translation. In at least one case (fol. 20b12 = p. 105.1-2, ed. Badawī) we can be quite sure that even a whole commentary section (on Physics II 3, 195a3-27) has been omitted, because we find between two consecutive Aristotelian passages the textual divider ☉ followed by the formula qāla Arisṭūṭālīs, which elsewhere indicates the transition from Isḥāq’s translation to a commentary section and back to the Aristotelian text.

CCXXII

Introduction

(e) Additions p. 6.10 al-mutaḥarriku l-awwalu (hoc quod mouetur primum) C instead of al-mutaḥarriku ΦV p. 10.7 ʿan ǧanbatay ḥarakatihī (ab utroque latere motus eius) C instead of ʿan ǧanbatayhi ΦV p. 61.9 ʿan ġayri mutaḥarrikin (ab eo quod non mouetur) C instead of ʿan mutaḥarrikin ΦV For most of the errors above, we can rule out that they were caused by mishaps within the Latin tradition. On the other hand, it is difficult to distinguish between genuine mistakes in car, merely illegible/damaged passages in car, and omissions and mistranslations by Gerard of Cremona. Considering the sheer quantity of individual mistakes in C/car, which is about three times as much as the above selection, we can quite safely assume that not all of them are to be imputed to the (otherwise very careful) translator. 4. With a similar degree of probability it can be ruled out that vzar is a common ancestor of Φ and car. Again, I give only a selective list of individual errors in VZ (resp. vzar): (a) Words or phrases omitted in vzar (unless overlooked/ignored by the translator) p. 1.3 qablu om. V : habent ΦC p. 4.2 abadan om. V : habent ΦC p. 7.5 al-mutaḍāddatayni om. V : habent ΦC p. 8.1 wa-taṣarrama om. V : habent ΦC p. 9.11 waḥdahū om. V : habent ΦC p. 13.5-6 aw ġayrihī mimman qāla miṯla qawlihī om. V : habent ΦC p. 38.11–39.1 mina l-muḥarrikāti om. V : habent ΦC p. 111.7 mimmā aḫaḏnāhu om. VZ : habent ΦC p. 117.6 ʿan om. VZ : habent ΦC p. 120.5 ʿan om. VZ : habent ΦC (b) Sauts du même au même p. 10.7 wāǧibun ḍarūratan an om. hom. V : habent ΦC p. 66.9-10 wa-in kāna ... azaliyyun2 om. hom. V : habent ΦC (c) Words mistranscribed in vzar (unless misread by the translator) p. 8.1 waqtan [mā?] (quandoque) V instead of ṭawʿan ΦC p. 9.2 wa-huwa ʿalā ǧihatin (? et est modo) V instead of wa-huwa l-āna ḍiʿfuhū ΦC p. 9.11–10.1 fī l-samāʾ (in celo) V instead of maʿa l-samāʾ ΦC p. 38.8 al-āḫar (aliud) V instead of al-aḫīr ΦC

VII.3 The relations btw. the Latin versions and the Arabic MSS

CCXXIII

p. 40.5 rafaʿathu (elevat) V instead of dafaʿathu ΦC p. 72.8 al-ḥarakatu l-muntaqilatu or ḥarakatu l-nuqlati (motus localis) V instead of al-ḥarakatu l-muttaṣilatu ΦC p. 78.1 al-takawwun (generatio) V instead of al-sukūn ΦC (d) Inversions p. 8.7 muḥarrikan ... mutaḥarrikan (mouere ... moueri) V instead of mutaḥarrikan ... muḥarrikan ΦC p. 25.9 yataḥarraku ... yaskunu (moueatur ... quiescat) V instead of yaskunu ... yataḥarraku ΦC p. 26.6 tataḥarraku ... taskunu (mouentur ... quiescunt) V instead of taskunu ... tataḥarraku ΦC p. 29.3 al-mutaḥarrikāt ... al-muḥarrikāt (motis ... mouentibus) V instead of al-muḥarrikāt ... al-mutaḥarrikāt ΦC p. 32.8-9 an yaqbala ... an yafʿala (recipere ... agere) V instead of an yafʿala ... an yaqbala ΦC (e) Additions p. 27.4 anna l-mawǧūda kullahū (quod totum ens) V instead of anna l-mawǧūda ΦC p. 28.8 takūna kulluhā sākinatan abadan (omnia quiescant semper) V instead of takūna kulluhā sākinatan ΦC p. 32.3 ilā fawqu ṭabʿan (ad superius naturaliter) V instead of ilā fawqu ΦC p. 105.3 bal ḫāriǧun ʿanhā fī-mā yataḥarraku ʿalā l-istiqāma (sed extra ipsum in quo est motus recte) V (= sed extra [om. ʿanhā?] in quo recte mouetur) Z instead of bal ḫāriǧun ʿanhā ΦC (f) Interpolated glosses (interpolated text in italics) p. 5.4 ex praedictis in quibus praediximus diffinitionem et determinationem in artibus, et nos incipiemus a rebus determinatis prius V : mina l-ašyāʾi llatī laḫḫaṣnāhā fī-mā taqaddama ΦC p. 11.10 Et sermo dicentium hoc magis assimilatur stulto, et similiter eius qui dixit V : wa-ka-ḏālika qawlu man qāla ΦC p. 23.6-7 Et iste sermo est similis ei quod dicitur de ... divisione lapidis a planta, posita enim est quaestio similis praedictae. Si igitur nos dissolverimus istam quaestionem, statim dissolvemus hanc quaestionem in qua sumus, et bonus verificator dixit quod distillatio continua perforat lapidem. Quaerendum est igitur si acciderit ut lapis perforetur hoc modo a centum guttis distillatis quod non necesse est V : wa-hāḏā l-qawlu šabīhun bi-mā yuqālu min amri ... falqi l-nabti li-l-ḥiǧārati wa-ḏālika annahū laysa yaǧibu ΦC

CCXXIV

Introduction

The last three errors mentioned in section (a) and the erroneous addition of fīmā yataḥarraku ʿalā l-istiqāma (section [e], last entry) prove that at least some of the above errors must trace back to vzar, because they occur both in V and in Z365. The reason for the comprehensive interpolations found in all books of the Physics in Michael Scot’s translation is that it was this version which was widely disseminated among the Latin scholastics. Having shown that none of our three witnesses Φ, car and vzar likely served as the exemplar of another and that all of them derive ultimately from the faulty pre-Baghdad-school codex Ω, we can now try to determine their internal relationships. Keeping in mind the above-mentioned uncertainty as to whether a certain peculiarity should be attributed to the Latin translator or his Arabic exemplar, we can first identify a number of common errors in car and vzar that may count as conjunctive errors and serve —at least partly— simultaneously as separative errors. 5. Shared errors in car and vzar: (a) Omissions p. 51.1 yataḥarraku huwa Φ : yataḥarraku (moueatur) CV p. 51.4 kullu wāḥidin Φ : kullun (totum, i.e. om. wāḥidin) C : om. V p. 75.6 aysara ḏāka Φ : om. CV p. 110.3 mā Φ : om. CV (b) Mistranscribed or transposed words p. 41.2 bi-l-ʿukkāzi Φ : al-ʿukkāza (baculum) CV p. 61.2 yūǧada Φ : yuʾḫaḏa (accipiatur) CV p. 72.4 wa-kāna Φ : quia C : quod V, i.e. car and vzar read li-annahū or wa-liannahū p. 79.8 bi-ḥīlati l-muntaqili Φ : totaliter C : totius translati V, i.e. car and vzar read bi-ǧumlati l-muntaqili p. 86.3 ḥ̅ Φ : A̅ CV p. 86.8 d̅ Φ : C̅ CV p. 91.1 al-amr Φ : al-aṯar (passione) CV p. 91.21 al-amr Φ : al-aṯar (passione) CV _____________ 365

In view of the above-mentioned evidence it seems rather unlikely that Michael Scot may have had access to Ibn Rušd’s autograph (as has been taken into consideration by Ruth Glasner, Averroes’ Physics, p. 37 and note 52); for the earliest circulation of Averroes manuscripts see also Charles Burnett, “The ‘Sons of Averroes with the Emperor Frederick’ and the Transmission of the Philosophical Works by Ibn Rushd”, in G. Endress & J. A. Aertsen (eds.), Averroes and the Aristotelian Tradition, p. 259—299.

VII.3 The relations btw. the Latin versions and the Arabic MSS

CCXXV

p. 107.2 ḏātiyyan (unpointed rasm) Φ : dāʾiman (semper) CV p. 110.3 dafaʿat Φ : eleuans C : eleuat V, i.e. car and vzar read rafaʿat (c) Additions p. 6.4 post mutaḥarrikan add. postquam erat in potentia eius ut moueatur C : add. postquam in sua potentia erat ut moueretur V (= baʿda an yakūna fī quwwatihī an yataḥarraka vel sim.) : om. Φ p. 43.6-7 al-ʿaql ... ġayru qābilin li-l-taʾṯīri Φ : intellectu ... est immobilis sed suscipiens impassionem C : intelligentia non mouetur neque recipit passionem V, i.e. car and vzar added ġayru mutaḥarrikin before or after ġayru qābilin li-l-taʾṯīri p. 60.7 post kaṯīrun add. illud enim quo completur illud sunt tres res C : add. hoc enim complebitur per tria V (= fa-innahū yatimmu hāḏā biṯalāṯati ašyāʾa vel sim.) : om. Φ p. 80.1 post rāǧiʿan add. secundum rectitudinem C : add. recte V (= ʿalā listiqāmati) : om. Φ The additions common to car and vzar may have been caused by contamination. However, their common source was apparently independent of Φ, where the additions left no trace at all. The remaining errors point clearly to a common ancestor of car and vzar (henceforth: Σ) independent of Φ. Unfortunately, the clarity of this picture is blurred by the following five mistakes that occur in Φ and car, yet not in vzar. 6. Shared errors in Φ and car: p. 49.5 mutaḥarrikan Φ car (moueatur) : muḥarrikan V (motor) p. 76.8 post al-taġayyur add. hāḏihī Φ car : om. V (recte) p. 97.2 post wa-ṣāra add. al-šayʾu Φ car (aliquid) : om. V (recte) p. 106.1 wa-ḏālika anna ... fī l-makāni om. hom. Φ in textu, sed suppl. i.m. (waǧadnā fī nusḫatin uḫrā ziyādatan...) : om. hom. car : habet V (motus enim propter vacuum est motus localis) p. 112.4 kull om. Φ car : habent VZ (omne) However, the nature of these errors is not such that they provide evidence for any closer relationship between Φ and car. The omission by homoioteleuton certainly does not constitute a conjunctive error. The second and fifth verbal stems of the root ḥ-r-k are consistently confused throughout the whole text. The additions may trace back to interlinear or marginal admixtures in Ω correctly ignored by the scribe of vzar. Finally, the last case did not pose a problem too difficult to be removed by a scholarly scribe. It concerns Physics VIII 10, 266 b19-20:

CCXXVI

Introduction

ἄπειρος δ’ ἔσται (δ’ ἔσται ut vid. Arab. : δὲ πᾶσα mss.) δύναμις ὥσπερ καὶ πλῆθος καὶ μέγεθος τὸ ὑπερβάλλον παντὸς ὡρισμένου Isḥāq’s translation runs in Φ and car as follows: wa-l-quwwatu takūnu bi-lā nihāyatin ka-mā takūnu l-ʿiddatu wa-yakūnu lʿiẓamu mā yafḍulu ʿalā maḥdūdin Transl. Gerard of Cremona: Potentia autem est infinita sicut est numerus et magnitudo aliqua (= ʿiẓamun mā instead of al- ʿiẓamu mā) superfluit super determinatum. Obviously before maḥdūdin/determinatum an equivalent of παντὸς is missing in Φ and car. As Aristotle is talking about an infinite power, number or magnitude, it is clear that this must exceed not only a certain limit, as it says in Φ and C, but all limits. A scholarly scribe could have realized this inconsistency and added kull (“all”). The correction was apparently already present in vzar, because both Michael Scot and the Anonymous render kull faithfully with omne. That both translators relied on Isḥāq’s translation is clear from the fact that they adhere to the reading wa-l-quwwatu takūnu bi-lā nihāyatin (= virtus [V] / potentia [Z] autem est infinita), which is probably based on the Greek reading ἄπειρος δ’ ἔσται δύναμις etc. Had they drawn on the translatio vetus, which is based on the Greek reading known from the extant manuscripts, i.e. ἄπειρος δὲ πᾶσα δύναμις etc. = infinita autem potentia omnis etc., one would expect traces of πᾶσα/omnis in their versions. Thus, I assume that the last error in the list above, which in a way may count as conjunctive error, was already present in Ω, yet was corrected by conjectural criticism of the scribe of vzar or its intermediate ancestor. However, we cannot rule out the possibility that the latter had access to another witness of unknown origin. While Physics VIII is free from anything that might look like a common mistake in Φ and vzar against the correct reading in car, we encounter one such case in book VII: 7. Shared error in Φ and vzar 366: Physics VII 2, version β, 244b18-20: ὁμοίως δὲ καὶ τὰ ἄλλα τὰ ὑπὸ τὰς αἰσθήσεις, ὧν ἐστι καὶ ἡ θερμότης καὶ ἡ ψυχρότης, καὶ ἡ λειότης καὶ ἡ τραχύτης. ταῦτα γάρ ἐστι πάθη τῆς ὑποκειμένης ποιότητος. These three lines are omitted in Φ and in the Latin translations by the Anonymous and Michael Scot, yet they are preserved in Gerard’s translation367. _____________ 366

The case was already noted by Mansion, Le texte d’Aristote Physique H, p. 83.

VII.3 The relations btw. the Latin versions and the Arabic MSS CCXXVII

What makes the case even worse is the fact that they are also omitted in the most relevant Greek source, namely the first hand of ms. E (cf. ed. Ross, app. crit. ad loc.), thus suggesting that the omission traces back as far as to Ψ. This would necessarily lead to the assumption that Gerard of Cremona used for his translation additional, Greek or Arabo-Latin materials, an assumption which is by no means supported by any other evidence. However, perhaps it is exactly the fact that only the first hand of E omitted the lines, whereas they were added in margine by E2, which provides the clue for solving this conundrum. As we have seen above, Isḥāq ibn Ḥunayn almost certainly supplied his translation with marginal notes and variant readings taken from the margins of his Greek exemplar. Hence, the following scenario seems to be worth considering: The lines 244b18-20 were omitted in the main text of Ψ, yet added in its margin (similar to the actual evidence in ms. E). Isḥāq ibn Ḥunayn followed this model and added a translation of these lines, perhaps accompanied by an appropriate note, in the margin of his autograph. While this marginal information got lost (as presumably did many other notes by Isḥāq) at a certain point in the Near Eastern transmission of the translation, it was still in place in the Andalusian manuscript Σ. In the aftermath, it shared in vzar the fate of its Baghdad cousin (which is why we find no equivalents in V and Z), whereas Gerard of Cremona still found it in his exemplar car. 367

VIII. Results and materials for a tentative stemma codicum of Isḥāq ibn Ḥunayn’s translation The above observations about the relationship between the Leiden manuscript, the fragments in ms. Esc. 896 (chapter II.4) and the Arabo-Latin witnesses can be summarized as follows. Isḥāq ibn Ḥunayn’s translation of the Physics as preserved almost completely in ms. Leiden Or. 583 and partially in ms. Esc. 896, as well as the Arabic exemplars of the three Latin translations derive ultimately from an early 10th c. manuscript Ω. This manuscript was a rather faulty copy of Isḥāq’s translation which contained already a limited number of glosses and marginal or interlinear notes, at least some of which originated from Isḥāq ibn Ḥunayn’s hand, yet not the abundance of glosses that have come down to us in Φ nor the separate commentary sections transmitted in the latter. The descendants of Ω split into two branches. _____________ 367

Mansion (Le texte d’Aristote Physique H, p. 81) collated five manuscripts. The only manuscript of book VII at my disposal, ms. Vindobonensis Bibl. Nat. 234A, contains the relevant phrase at fol. 66a2-4.

CCXXVIII

Introduction

(i) On the one hand, there is good reason to postulate a manuscript Θ, which was copied from Ω around the year 318/930, probably in Baghdad. Θ is a (possibly rather distant) ancestor of ms. Esc. 896, the unique extant Arabic witness of a separate transmission of Isḥāq ibn Ḥunayn’s translation (probably 11th c.). But Θ served also as the main source of the scholastic activities of the Baghdad Aristotelians. In all likelihood, the scribe of Θ added further scholia and notes to those already present in Ω. These scholia and notes are also found in the descendants of Θ, yet not in the common ancestor Σ of the Latin witnesses (resp. car or vzar). On the basis of Θ —possibly by means of intermediate copies— the Baghdad Aristotelians produced between ca. 930 and 1004 various school corpora (θ1, θ2, etc.) including self-contained commentary sections (both excerpts from late ancient commentaries translated into Arabic and original Arabic comments), glosses, marginal and interlinear variant readings, and minutes of the Baghdad teaching practice. While the Aristotelian text itself was in all corpora that of ms. Θ, the way in which its subdivision is structured and the style and organisation of the accompanying materials varied depending on the individual methodology of the scholars responsible for their composition or the preparation of their teaching materials (among them Ibn al-Samḥ and Ibn al-Ṭayyib). Whether each of these corpora contained a complete copy of Θ covering the whole Aristotelian text is a matter of speculation. They certainly differed with respect to the comprehensiveness and structure of the comments and notes. In the year 1004, Abū l-Ḥusayn al-Baṣrī collected these materials and prepared his compilation composed of Aristotle’s text and the (partial) accompanying materials (commentaries, glosses, etc.) as accessible to him through θ1, θ2, etc. This compilation is extant in the unique Leiden manuscript, separated from the latter by one intermediate copy prepared in the year 1077/78 in al-Karḫ. Taking into consideration that Abū l-Ḥusayn al-Baṣrī claims to have collated his materials with Yaḥyā ibn ʿAdī’s copy of Isḥāq’s translation (Y), while the Aristotelian text preserved in his compilation displays numerous substantial mistakes that obviously could not be remedied by means of Y, we may furthermore assume that either Y was a descendant οf Ω, or Y and Ω derived from the same error-ridden ancestor368. This latter may be identical with what Yaḥyā ibn ʿAdī calls al-dastūr or “master copy”, but certainly not Isḥāq ibn Ḥunayn’s autograph. _____________ 368

This assumption is additionally supported by the fact that ms. Esc. 896 and Φ contain one almost identical marginal note which in Φ is marked by the letter ‫( ح‬cf. above, p. LXXX–LXXXI). The note must have been reproduced by Yaḥyā ibn ʿAdī from Ω or from the common ancestor of Y and Ω.

VIII. Results and tentative stemma codicum

CCXXIX

(ii) On the other hand, Ω served (directly or by means of further copies) as the exemplar of a manuscript Σ which by some route or another made its way to Andalusia and constituted the Western tradition of Isḥāq ibn Ḥunayn’s translation. Σ had (at least) two descendants: vzar, the copy of Isḥāq’s translation used by Ibn Rušd (Averroes), and car, Gerard of Cremona’s Arabic exemplar. The legacy of vzar is connected with Ibn Rušd’s “Long Commentary” on the Physics. In all likelihood, Michael Scot and the Anonymous used different Arabic copies for their translations of Ibn Rušd’s work. Thus, the main stages, nodes and ramifications of the Arabic and Latin transmission and reception of Isḥāq ibn Ḥunayn’s translation might be illustrated as follows (p. CCXXX).

CCXXX

Introduction

Tentative Stemma of the Arabic Transmission and Some MSS ca. 870-900

Isḥāq ibn Ḥunayn’s translation

ca. 900-920

●

ca. 920-930

Ω

(? “al-dastūr”)

Θ

ca. 930 (?) (?)

betw. 930-1000

(Baghdad

θ1

school corpora)

θ2

θ3

Y

ca. 950

θ4...(?)

(Yaḥyā ibn ʿAdī’s copy)

1004

(Abū l-Ḥusayn

11th c. (?)

Σ (Andalusian ancestor)

ms. Esc. ● ár. 896

1077/78

●

ca. 1180

ca. 1180-1220 (?) ca. 1220-1230 (?) ca. 1250 (?)

al-Karḫ-copy (“al-umm”)

Φ ms. Leiden Or. 583

1130 before ca. 1180

●

al-Baṣrī’s compilation)

car ☉

vzar

Transl. Lat. Cremonensis Michael Scot’s Ar. exemplar

●

●

Ar. Šarḥ Ibn Rušd ●

indicates: exemplar(s) → copy

Anonymous’ Ar. exemplar

Trans. Lat. ☉ Michael Scot ☉

Transl. Lat. Anonymous

indicates: exemplar → translation indicates: collation or use of additional source

IX. Quotations from the Physics in two treatises by Alexander

CCXXXI

IX. Quotations from Aristotle’s Physics in two treatises by Alexander of Aphrodosias The works of Alexander of Aphrodisias are among the earliest and most important external witnesses of the textual transmission of Aristotle’s œuvre. As for Aristotle’s Physics, literal quotations occur not only in the fragments of Alexander’s commentary preserved in scholia and through Philoponus’ and Simplicius’ references, but also in Alexander’s treatises On the Principles of the Cosmos and Against Galen on Motion, the Greek originals of which have been lost. So far, the relationship between these quotations, preserved in medieval Arabic translations, and the Greek manuscripts of the Physics has not been scrutinised. Furthermore, a comparison of the quotation extant in On the Principles of the Cosmos with Isḥāq ibn Ḥunayn’s translation of the Physics may help to clarify the pending question, if Isḥāq is the translator of this treatise. Alexander’s On the Principles of the Cosmos is extant in two Arabic versions, both entitled Fī mabādiʾ al-kull. One version, incomplete and lacking the quotation from the Physics, has been prepared by Isḥāq ibn Ḥunayn’s contemporary Abū ʿUṯmān al-Dimašqī, while the authorship of the other one is ambiguous. The extant manuscripts mention no less than three different translators, Ḥunayn ibn Isḥāq, Isḥāq ibn Ḥunayn, and Ibrāhīm ibn ʿAbdallāh alKātib369. This second version contains Alexander’s quotation of Physics VIII 6, 259a6-19. The following conspectus of this quotation along with the Greek text and Isḥāq ibn Ḥunayn’s translation of the corresponding section will show that the latter can be excluded from the list of contemplable translators of Alexander’s treatise: Physics VIII 6, 259a6-19

Alexander, On the Principles of the Cosmos, ed. Genequand, p. 88–90

Physics transl. Isḥāq ibn Ḥunayn (cf. below, p. 60–61)

εἴπερ οὖν ἀΐδιος ἡ κίνησις, ἀΐδιον καὶ τὸ κινοῦν ἔσται πρῶτον, εἰ ἕν· εἰ δὲ πλείω, πλείω τὰ ἀΐδια. ἓν δὲ μᾶλλον ἢ πολλά, καὶ πεπερασμένα ἢ ἄπειρα, δεῖ νομίζειν. τῶν αὐτῶν γὰρ συμβαινόν-

‫اﶈﺮك‬ ّ ‫ﻓﻤﱴ ﻛﺎﻧﺖ اﳊﺮﻛﺔ أزﻟﻴّﺔً ﻓﺈ ّن‬ ‫اﻷول إن ﻛﺎن واﺣﺪاً ﻓﻬﻮ أﻳﻀ ًﺎ‬ ّ ‫ﻛﺎن‬ ‫وإن‬ ‫أزﱄ‬ ً‫اﶈﺮﻛﻮن ﻛﺜﲑا‬ ّ ّ ‫ﻓﺎﻷﺷﻴﺎء اﻷزﻟﻴّﺔ أﻳﻀﺎً ﻛﺜﲑة وﻗﺪ‬ ‫اﻷول‬ ّ ‫اﶈﺮك‬ ّ ‫ﻳﻨﺒﻐﻲ أن‬ ّ ‫ﻳﻈﻦ أ ّن‬ ‫واﺣﺪ ﻻ ﻛﺜﲑ وأ ّ ﺎ ﻣﺘﻨﺎﻫﻴﺔ ﻻ ﻏﲑ‬ ‫اﳌﺘﻨﺎﻫﻴﺔ وذﻟﻚ أﻧّﻪ ﳌﺎ ﻛﺎﻧﺖ‬

‫اﶈﺮك‬ ّ ‫ﻓﺈن ﻛﺎﻧﺖ اﳊﺮﻛﺔ أزﻟﻴّﺔً ﻓﺈ ّن‬ ‫اﻷول ﻳﻜﻮن أﻳﻀﺎً أزﻟﻴّﺎً إن ﻛﺎن‬ ّ ‫أﻛﺜﺮ ﻣﻦ واﺣﺪ‬ ‫ﻛﺎن‬ ‫وإن‬ ‫ا‬ ‫اﺣﺪ‬ ‫و‬ ً َ ‫ﻓﺎﻷزﻟﻴّﺎت أﻛﺜﺮ ﻣﻦ واﺣﺪ‬ ‫واﻷﺣﺮى ن ﻳﻌﺘﻘﺪ أﻧّﻪ واﺣﺪ ﻻ‬ ‫ﻛﺜﲑ وإن ﻛﺎن ﻛﺜﲑاً ﻓﺎﻷﺣﺮى أن‬ ‫ﻳﻌﺘﻘﺪ أ ّ ﺎ ﻣﺘﻨﺎﻫﻴﺔ ﻻ ﻏﲑ‬

_____________ 369

The two versions are edited, translated and analysed in Charles Genequand (ed.), Alexander of Aphrodisias, On the Cosmos (Islamic Philosophy, Theology and Science. Texts and Studies, vol. XLIV), Leiden 2001.

CCXXXII

Introduction

Physics VIII 6, 259a6-19 (cont.)

Alexander, On the Principles of the Cosmos (cont.)

Physics transl. Isḥāq ibn Ḥunayn (cont.)

των αἰεὶ τὰ πεπερασμένα μᾶλλον ληπτέον· ἐν γὰρ τοῖς φύσει δεῖ τὸ πεπερασμένον καὶ τὸ βέλτιον, ἂν ἐνδέχηται, ὑπάρχειν μᾶλλον. ἱκανὸν δὲ καὶ ἕν, ὃ πρῶτον τῶν ἀκινήτων ἀΐδιον ὂν ἔσται ἀρχὴ τοῖς ἄλλοις κινήσεως. φανερὸν δὲ καὶ ἐκ τοῦδε ὅτι ἀνάγκη εἶναί τι ἓν καὶ ἀΐδιον τὸ πρῶτον κινοῦν. δέδεικται γὰρ ὅτι ἀνάγκη ἀεὶ κίνησιν εἶναι. εἰ δὲ ἀεί, ἀνάγκη συνεχῆ εἶναι· καὶ γὰρ τὸ ἀεὶ συνεχές, τὸ δ’ ἐφεξῆς οὐ συνεχές. ἀλλὰ μὴν εἴ γε συνεχής, μία. μία δ’ ἡ ὑφ’ ἑνός τε τοῦ κινοῦντος καὶ ἑνὸς τοῦ κινουμένου

‫اﻷﺷﻴﺎء اﻟﱵ ﺗﻌﺮض أﺷﻴﺎء واﺣﺪة‬ ‫ﻋﻴﺎ ﺎ داﺋﻤﺎً ﻓﺈ ّن اﻷوﱃ واﻷوﺟﺐ‬ ‫أن ﻳﺮى أﻧّﻪ ﻣﺘﻨﺎﻩٍ ﻷﻧّﻪ ﻳﻨﺒﻐﻲ أن‬ ‫ﻳﻜﻮن اﻟﺸﻲء اﳌﺘﻨﺎﻫﻲ ﰲ اﻷﻣﻮر‬ ‫اﻟﱵ ﻟطﺒﻊ أوﱃ ﻟﺘﻔﻀﻴﻞ إن ﻛﺎن‬ ‫ وﻗﺪ‬370ٍ‫ذﻟﻚ ﳑﻜﻨﺎً ﳑّﺎ ﻟﻴﺲ ﲟﺘﻨﺎﻩ‬ ‫ﻳﻜﺘﻔﻲ ن ﻳﻜﻮن ذﻟﻚ اﻟﻮاﺣﺪ‬ ‫اﻷزﱄ اﻟﺬي ﻫﻮ أﻗﺪم ﻣﻦ ﲨﻴﻊ‬ ّ ‫ ﻣﺒﺪأ ﺣﺮﻛﺔ‬371‫اﳌﺘﺤﺮﻛﺔ‬ ّ ‫اﻷﺷﻴﺎء ﻏﲑ‬ ‫ﺗﺒﲔ ﳑّﺎ‬ ّ ‫[ وﻗﺪ‬. . .] ‫ﺳﺎﺋﺮ اﻷﺷﻴﺎء‬ ‫أﻗﻮل أﻧّﻪ ﻳﻠﺰم ﻣﻦ اﻻﺿطﺮار أن‬ ً‫اﻷول ﺷﻴﺌﺎً واﺣﺪا‬ ّ ‫اﶈﺮك‬ ّ ‫ﻳﻜﻮن‬ ‫ﺗﺒﲔ أﻧّﻪ ﳚﺐ‬ ‫ﻗﺪ‬ ‫ﻪ‬ ‫ﻧ‬ ‫أ‬ ‫أزﻟﻴّﺎً وذﻟﻚ‬ ّ ّ ً‫ﺿﺮورةً أن ﺗﻜﻮن اﳊﺮﻛﺔ ﺳﺮﻣﺪﻳّﺔ‬ ‫وﻣﱴ ﻛﺎﻧﺖ ﺳﺮﻣﺪﻳّﺔً ﻓﻘﺪ ﻳﻠﺰم‬ ‫ﺿﺮورةً أن ﺗﻜﻮن ﻣﺘّﺼﻠﺔً ﻷ ّن‬ ‫اﻟﺴﺮﻣﺪﻳّﺔ ﻫﻲ ﻣﺘّﺼﻠﺔ و ّأﻣﺎ اﻟﱵ‬ ‫ﺗﻜﻮن ﺷﻴﺌﺎً ﺑﻌﺪ ﺷﻲء ﻓﻠﻴﺴﺖ‬ ً‫ﻣﺘّﺼﻠﺔً ّإﻻ أ ّ ﺎ إن ﻛﺎﻧﺖ ﻣﺘّﺼﻠﺔ‬ 372 ﻓﺎﳊﺮﻛﺔ واﺣﺪة‬ ‫اﶈﺮك واﻟﻮاﺣﺪ‬ ّ ‫ﺿﺎﻓﺘﻬﺎ إﱃ اﻟﻮاﺣﺪ‬ ‫اﳌﺘﺤﺮك‬ ّ

‫>ﻣﺘﻨﺎﻫﻴﺔ< وذﻟﻚ أ ّن اﻟﻠﻮاﺣﻖ‬ ‫إذا ﻛﺎﻧﺖ واﺣﺪةً ﻋﻴﺎ ﺎ ﻓﻴﻨﺒﻐﻲ‬ ‫اﻟﺘﻤﺴﻚ‬ ّ ‫أﺑﺪاً أن ﻳﻜﻮن اﻷﺣﺮى‬ ‫ﳌﺘﻨﺎﻫﻴﺔ وذﻟﻚ أ ّن اﻷﺷﻴﺎء اﻟﱵ‬ ‫ﺗﻜﻮن ﻟطﺒﻴﻌﺔ ﻳﻨﺒﻐﻲ أن ﻳﻮﺟﺪ‬ ‫ﳊﺮي اﳌﺘﻨﺎﻫﻲ واﻷﻓﻀﻞ‬ ‫ﻓﻴﻬﺎ‬ ّ ‫ﳑّﺎ ﻫﻮ ﳑﻜﻦ وﰲ واﺣﺪ ﻛﻔﺎﻳﺔ‬ ‫ﻳﻜﻮن ّأو َل اﻷﺷﻴﺎء اﻟﱵ ﻻ‬ ‫ﺗﺘﺤﺮك أزﻟﻴّﺎً وﻣﺒﺪأ ﻟﻠﺤﺮﻛﺔ ﰲ‬ ّ ‫ﺳﺎﺋﺮ اﻷُﺧﺮ وﻗﺪ ﻳﻈﻬﺮ ﳑّﺎ ﳓﻦ‬ ‫ﻗﺎﺋﻠﻮﻩ أﻳﻀﺎً أﻧّﻪ واﺟﺐ ﺿﺮورةً أن‬ ً‫اﻷول ﺷﻴﺌﺎً واﺣﺪا‬ ّ ‫اﶈﺮك‬ ّ ‫ﻳﻜﻮن‬ ‫أزﻟﻴّﺎً وذﻟﻚ أ ّ ﻗﺪ ﺑﻴّـﻨّﺎ أﻧّﻪ‬ ‫واﺟﺐ ﺿﺮورةً أن ﺗﻜﻮن اﳊﺮﻛﺔ‬ ‫داﺋﻤﺎً وإن ﻛﺎﻧﺖ داﺋﻤﺎً ﻓﻮاﺟﺐ‬ ً‫ﺿﺮورةً أن >ﺗﻜﻮن< ﻣﺘّﺼﻠﺔ‬ ‫ﻓﺄﻣﺎ‬ ّ ‫وذﻟﻚ أ ّن اﻟﺪاﺋﻢ ﻣﺘّﺼﻞ‬ ‫اﳌﺘﻮاﱄ ﻓﻠﻴﺲ ﻣﺘّﺼﻼً ﻟﻜﻦ إن‬ ‫ﻛﺎﻧﺖ ﻣﺘّﺼﻠﺔً ﻓﻬﻲ واﺣﺪة‬ ‫واﳊﺮﻛﺔ اﻟﻮاﺣﺪة ﻫﻲ اﻟﱵ ﺗﻜﻮن‬ ‫ﻣﺘﺤﺮك‬ ‫ﳏﺮك واﺣﺪ وﻋﻦ‬ ّ ّ ‫ﻋﻦ‬ ‫واﺣﺪ‬

That the two translations do not originate from one and the same translator becomes clear not only from the differences in terminology and translation technique, but also from the fact that they interpret the same Greek text differently. E. g., at Physics 259a9-10, τῶν αὐτῶν γὰρ συμβαινόντων αἰεὶ τὰ πεπερασμένα μᾶλλον ληπτέον, Isḥāq ibn Ḥunayn relates αἰεὶ to μᾶλλον ληπτέον, whereas the translator of Alexander’s treatise takes it as qualifying τὰ αὐτὰ συμβαίνοντα. At 259a16-17, Isḥāq interprets Aristotle as speaking generally about “that which is always” and “that which is in succession”, while the translator of Alexander’s treatise takes τὸ ἀεὶ and τὸ ἐφεξῆς as referring to types of motion. More important, the comparison shows that Alexander quotes at some places another Greek text than the one at Isḥāq ibn Ḥunayn’s disposal. The Arabic _____________

Read thus for ٍ‫ ﻣﺘﻨﺎﻩ‬ed. Genequand. Sic ms. Damascus, ‫اﳌﺘﺤﺮك‬ ّ codd. cett., ed. Genequand. 372 ‫ =( ﻓﺎﳊﺮﻛﺔ واﺣﺪة‬μία δ’ ἡ [sc. κίνησις]) omitted in the extant mss. by homoioteleuton and omitted by Genequand. 370 371

IX. Quotations from the Physics in two treatises by Alexander CCXXXIII

version of Alexander’s quotation of the sentence ἐν γὰρ τοῖς φύσει δεῖ τὸ πεπερασμένον καὶ τὸ βέλτιον […] ὑπάρχειν μᾶλλον (259a10-12) takes βέλτιον not as a noun coordinated to τὸ πεπερασμένον, but rather as qualifying, together with μᾶλλον, the predicate δεῖ … ὑπάρχειν (“the finite thing must be present in things [constituted] by nature in a way more appropriate to be preferred [awlā bi-l-tafḍīl] than that which is infinite”). This suggests the omission of τὸ (or even καὶ τὸ) before βέλτιον, which is worth mentioning, because ms. E with Aristotle’s Physics likewise omits τὸ. However, Isḥāq clearly read τὸ πεπερασμένον καὶ τὸ βέλτιον (= al-mutanāhī wa-l-afḍalu). The phrase φανερὸν δὲ καὶ ἐκ τοῦδε … (259a14) is transmitted in the lemma of Philoponus’ commentary (p. 888.3) and in the Latin vetus translatio without καὶ. This reading is confirmed by the Arabic translation of Alexander’s treatise, whereas Isḥāq’s Greek Vorlage read καὶ (= ayḍan) as the majority of the extant Greek manuscripts and Simplicius (p. 1254.29). Further quotations from the Physics are found in Alexander’s Refutation of Galen’s Εἰς τὸ πρῶτον κινοῦν ἀκίνητον, translated into Arabic by Abū ʿUṯmān al-Dimašqī373. The main topic of the treatise is Aristotle’s doctrine that everything that is in motion must be moved by something and its subsequent proofs in book VII of the Physics. In terms of textual criticism it thus may shed some light on Galen’s and Alexander’s access to book VII of the Physics in as much as some of the quotations clearly match with the so-called version α, others with version β of this book. In addition, the treatise contains the following quotation from Physics VIII 4, relevant to our present concern: Physics VIII 4, 254b7-14

Τῶν δὴ κινούντων καὶ κινουμένων τὰ μὲν κατὰ συμβεβηκὸς κινεῖ καὶ κινεῖται, τὰ δὲ

Alexander, Refutation of Galen, ed. Rescher & Marmura, p. 103

Physics transl. Isḥāq ibn Ḥunayn (cf. below, p. 29)

‫وﺗﺘﺤﺮك ﻓﺈ ّن ﻣﻨﻬﺎ‬ ّ ‫ﲢﺮك‬ ّ ‫ﲢﺮك و ّأﻣﺎ اﻷﺷﻴﺎء اﻟﱵ‬ ّ ‫اﳌﺘﺤﺮﻛﺎت‬ ّ ‫اﶈﺮﻛﺎت و‬ ّ ‫إ ّن‬ ‫ط‬ ‫وﻳﺘﺤﺮك ﻋﻠﻰ ﻃﺮﻳﻖ اﻟﻌﺮض‬ ‫ك‬ ‫ﳛﺮ‬ ‫ﻣﺎ‬ ‫اﻟﻌﺮض‬ ‫ﻳﻖ‬ ‫ﺮ‬ ‫ﺑ‬ ‫ﺑﻌﻀﻬﺎ‬ ‫وﺗﺘﺤﺮك‬ ّ ّ ّ 374 ‫ﺑﻂ‬ ‫وﻳﺘﺤﺮك‬ ‫ك‬ ‫ﳛﺮ‬ ‫اﻟﺬي‬ ‫و‬ ‫ﺑﺬاﺗﻪ‬ ‫وﻣﻨﻬﺎ‬ ‫اﻟﻌﺮض‬ ‫ﻳﻖ‬ ‫ﺮ‬ ‫ﺎ‬ ‫أﻣ‬ ‫ﺎ‬ ‫ا‬ ‫و‬ ‫ﺑﺬ‬ ‫وﺑﻌﻀﻬﺎ‬ ّ ّ ّ

_____________ 373

Nicholas Rescher, Michael E. Marmura (ed.), The Refutation by Alexander of Aphrodisias of Galen’s Treatise on the Theory of Motion. Translated from the medieval Arabic version, with an introduction, notes, and an edition of the Arabic text. Islamabad 1965. Neither Galen’s treatise (mentioned in his auto-bibliography) nor Alexander’s refutation are extant in Greek. The Arabic manuscripts are difficult to read, and the text is definitely in need of a new critical edition (v. also the reviews by J. van Ess, Erasmus 24 [1972], and M. Ullmann, Der Islam 49 [1972]). 374 After ‫وﻳﺘﺤﺮك‬ ms. Escorial adds erroneously ‫وﻳﺘﺤﺮك‬ ‫ﳛﺮك‬ ّ ّ ّ ‫ﻓﺈّﳕﺎ‬, adopted by R[escher] & ‫و‬ ‫ﻛﺔ‬ ‫اﶈﺮ‬ ‫اﻷﺷﻴﺎء‬ ‫ﰲ‬ ‫ﺗﻮﺟﺪ‬ ‫ﺎ‬ ‫ﳕ‬ M[armura]. Ms. Carullah adds ‫اﳌﺘﺤﺮﻛﺔ‬ ّ‫ﻓﺈ‬, cf. below, footnote 376. ّ ّ

CCXXXIV

Introduction

Physics VIII 4, 254b7-14 (cont.)

Alexander, Refutation of Galen (cont.)

Physics transl. Isḥāq ibn Ḥunayn (cont.)

καθ' αὑτά, κατὰ συμβεβηκὸς μὲν οἷον ὅσα τε τῷ ὑπάρχειν τοῖς κινοῦσιν ἢ κινουμένοις καὶ τὰ κατὰ μόριον, τὰ δὲ καθ' αὑτά, ὅσα μὴ τῷ ὑπάρχειν τῷ κινοῦντι ἢ τῷ κινουμένῳ, μηδὲ τῷ μόριόν τι αὐτῶν κινεῖν ἢ κινεῖσθαι. τῶν δὲ καθ' αὑτὰ τὰ μὲν ὑφ' ἑαυτοῦ τὰ δ' ὑπ' ἄλλου, καὶ τὰ μὲν φύσει τὰ δὲ βίᾳ καὶ παρὰ φύσιν

‫ﻋﻠﻰ ﻃﺮﻳﻖ اﻟﻌﺮض ﻫﻮ ﲟﻨﺰﻟﺔ ﲨﻴﻊ‬ ‫اﳌﺘﺤﺮﻛﺔ واﻷﺷﻴﺎء اﻟﱵ‬ ّ ‫اﶈﺮﻛﺔ و‬ ّ ‫اﻷﺷﻴﺎء‬ ‫وﺗﺘﺤﺮك ﳉﺰء و ّأﻣﺎ اﻷﺷﻴﺎء‬ ّ 375‫ﲢﺮك‬ ّ ‫اﻟﱵ‬ ‫ ﻓﻬﻲ‬376‫وﺗﺘﺤﺮك ﺑﺬا ﺎ‬ ‫ك‬ ‫ﲢﺮ‬ ّ ّ ‫ﲨﻴﻊ اﻷﺷﻴﺎء اﻟﱵ ﻟﻴﺲ إّﳕﺎ ﺗﻜﻮن ﺬﻩ‬ ‫اﳊﺎل ﻣﻦ ﻗِﺒَ ِﻞ أ ّ ﺎ ﺗﻮﺟﺪ ﰲ اﻷﺷﻴﺎء‬ 377 ‫اﳌﺘﺤﺮﻛﺔ وﻻ ﻣﻦ ﻗِﺒَ ِﻞ أ ّن‬ ‫اﶈﺮﻛﺔ أو‬ ّ ّ ‫ﻳﺘﺤﺮك وﻣﻦ‬ ‫أو‬ ‫ك‬ ‫ﳛﺮ‬ ‫ﻣﻨﻬﺎ‬ ‫ﺟﺰءاً ﻣﺎ‬ ّ ّ ‫ﳛﺮك‬ ّ ‫ﺗﺘﺤﺮك ﺑﺬا ﺎ ﻣﺎ‬ ّ ‫اﻷﺷﻴﺎء اﻟﱵ‬ ‫ﻳﺘﺤﺮك ﻋﻦ ﻏﲑﻩ وﻣﻨﻬﺎ‬ ّ ‫ﻧﻔﺴﻪ وﻣﻨﻬﺎ ﻣﺎ‬ ً‫ﻳﺘﺤﺮك ﻟطﺒﻊ وﻣﻨﻬﺎ ﻗﺴﺮاً وﺧﺎرﺟﺎ‬ ّ ‫ﻣﺎ‬ ‫ﻋﻦ اﻟطﺒﻊ‬

‫ﻓﻤﺜﻞ‬ ً‫>ﻛﻞ< ﻣﺎ ﻛﺎن ﻣﻮﺟﻮدا‬ ّ ‫اﳌﺘﺤﺮﻛﺎت وﻣﺎ‬ ‫أو‬ ‫ﻛﺎت‬ ‫اﶈﺮ‬ ّ ّ ‫ﰲ‬ ‫ﻛﺎن ﳉﺰء و ّأﻣﺎ ﺑﺬوا ﺎ ﻓﺎﻷﺷﻴﺎء‬ ‫ﺗﺘﺤﺮك ﻻ ّ ﺎ‬ ‫ﻛﻠّﻬﺎ اﻟﱵ‬ ّ ‫اﳌﺘﺤﺮك‬ ‫ﰲ‬ ‫أو‬ ‫ك‬ ‫اﶈﺮ‬ ّ ّ ‫ﻣﻮﺟﻮدة ﰲ‬ ‫ﺣﺮك أو‬ ‫ﻣﻨﻬﺎ‬ ‫ﻣﺎ‬ ً‫وﻻ ّن ﺟﺰءا‬ ّ ‫اﳌﺘﺤﺮﻛﺔ ﺑﺬوا ﺎ ﺑﻌﻀﻬﺎ‬ ّ ّ ‫ﲢﺮك و‬ ‫ﻣﻦ ﺗﻠﻘﺎﺋﻬﺎ وﺑﻌﻀﻬﺎ ﻣﻦ ﻏﲑﻫﺎ‬ ً‫وﺑﻌﺾ ﻫﺬﻩ ﻃﺒﻌﺎً وﺑﻌﻀﻬﺎ ﻗﺴﺮا‬ ‫وﺧﺎرﺟﺎً ﻋﻦ ﻃﺒﻌﻬﺎ‬

Obviously Isḥāq ibn Ḥunayn’s translation is superior and more precise. As for the underlying Greek text, Alexander’s sources must have been very close to the above text, reprinted from Ross’ edition and confirmed by Isḥāq’s translation, with two minor exceptions: At the beginning of the passage quoted, Alexander’s text points to τῶν δὲ κινούντων rather than τῶν δὴ κινούντων. And at 254b11, where Isḥāq’s version supports τῷ κινοῦντι ἢ τῷ κινουμένῳ (= mss. K be FH), the Arabic version of Alexander’s treatise is probably based on τῷ κινοῦντι ἢ κινουμένῳ as transmitted in mss. E LNR IJQw.

X. Principles of the present edition X.1 Establishment of the Arabic text and documentation of the underlying Greek readings As mentioned previously, the edition of the Arabic text had to be limited to bk. VIII of Aristotle’s Physics, in order to accomplish the required philological and codicological tasks within a reasonable period of time. For the same reason, a new edition of the Arabic commentary sections accompanying the Aristotelian text in the Leiden manuscript, although certainly a veritable scholarly deside_____________ 375 ‫ و ّأﻣﺎ اﻷﺷﻴﺎء اﻟﱵ‬scripsi : ‫ و ّأﻣﺎ اﻟﱵ‬ms. Escorial, ed. R. & M. : ‫ واﻷﺷﻴﺎء اﻟﱵ‬ms. Carullah. 376 After ‫ ﺑﺬا ﺎ‬ms. Escorial adds erroneously ‫اﳌﺘﺤﺮﻛﺔ‬ ّ ‫اﶈﺮﻛﺔ و‬ ّ ‫ ﺗﻮﺟﺪ ﰲ اﻷﺷﻴﺎء‬from the following

line, adopted by R. & M. The present text follows ms. Carullah. 377 ‫ أو‬ms. Carullah (= ἢ) : ‫ و‬ms. Escorial, ed. R. & M.

X.1 Editorial principles: Establishment of the Arabic text

CCXXXV

ratum, could not be accomplished here. Considering the above historical and codicological results, the present edition rests upon the following editorial techniques and principles. I. The establishment of the Arabic text, aiming at the nearest possible approximation to Isḥāq ibn Ḥunayn’s Linguistic Text378, is principally based on two distinct types of sources: (a) Testimonies of Isḥāq ibn Ḥunayn’s translation, and (b) the extant Greek witnesses of Aristotle’s text. Both groups comprise witnesses of the direct as well as of the indirect traditions. (a.1) The extant direct witnesses of Isḥāq’s translation are formed by ms. Esc. 896 and the lemmata of the Aristotelian text in ms. Leiden Or. 583 (Φ). Since the former contains only parts of Physics IV 9—V 1, it cannot serve as a source for the present edition. (a.2) The indirect testimonies of the Arabic tradition consist of: (α) quotations of Isḥāq’s translation (including variant readings thereof) in the scholia and commentaries by the Baghdad Aristotelians transmitted in Φ, (β) quotations of Isḥāq’s translation in independent Arabic works such as the treatises on physics by Yaḥyā ibn ʿAdī or Ibn Bāǧǧa, (γ) Gerard of Cremona’s Latin translation (= C) based on the 12th c. copy car of Isḥāq’s text, and (δ) the lemmata (textus) of the Latin translations of Ibn Rušd’s Šarḥ by Michael Scot (= V) and Anonymous (= Z), which ultimately trace back to the Arabic exemplar vzar of Isḥāq’s translation used by Ibn Rušd. (b.1) The direct Greek tradition is represented by the 14 manuscripts of Aristotle’s Physics analysed in chapter VI. For the relevant readings of these manuscripts registered in the apparatus of the present edition I relied in large part on the painstaking collations by Pieter Sjoerd Hasper. (b.2) The indirect tradition consists of quotations of the Aristotelian text in: (α) the extant fragments of Alexander of Aphrodisias’ commentary as quoted by Philoponus or Simplicius (= A) or preserved in the form of scholia in ms. Paris. suppl. gr. 643 (= As), (β) the reconstructed Greek text of Alexander’s quotations from the Physics as witnessed in the Arabic translations of his treatises On the Principles of the Cosmos (= Am) and Against Galen on Motion (= Ar, cf. chapter IX), _____________ 378

Cf. above, p. XXIII–XXVII.

CCXXXVI

Introduction

(γ) Themistius’ paraphrase (= T), (δ) the commentary by John Philoponus (= P, from bk. V on only fragmentarily preserved), (ε) scholia extracted from Philoponus’ commentary and transmitted separately379, (ζ) the commentary by Simplicius (= S), (η) the commentary attributed to Michael Psellos (= Pa), and (θ) the Graeco-Latin translatio vetus (= Λ). As for the readings of Λ, I depend again entirely on Hasper’s meticulous notes kindly put at my disposal. Most witnesses of the indirect traditions are of crucial importance for the constitutio textus, the Arabic and Arabo-Latin ones in view of the sparse and defective direct Arabic tradition, the Greek ones in light of the facts that they trace back to a period prior to any extant witness of the direct Greek tradition and that they are closely related to a branch of transmission scarcely represented by any other manuscript or even constitute a rather independent branch by itself (as in the case of Simplicius’ quotations). Even the commentary/paraphrase attributed to Michael Psellos (d. ca. 1078), although probably not by Psellos, but rather by Georgios Pachymeres (d. 1310)380 and in any case later than Isḥāq ibn Ḥunayn’s translation, turned out to be worth being considered for the apparatus. For it contains a number of readings (of uncertain origin) otherwise attested only through Isḥāq’s Vorlage or in branch α of the manuscripts of the direct tradition381. Bearing in mind the preceding analysis of the stemmatic relationship, we are able, as a first step, to reconstruct on the basis of Φ (viz. Θ) and the witnesses mentioned under (a.2) the readings of the common ancestor of the entire surviving tradition, ms. Ω (ca. 920—930, not extant)382. Any reconstruction based solely on the extant Arabic and Arabo-Latin witnesses comes to a halt with this “archetype”. However, as became likewise clear above, Ω was a quite unreliable copy of Isḥāq’s translation. Hence, whenever it is possible to determine that what seems to be an error in Ω must be due to a corruption within the Arabic transmission, it must be examined whether the problem can be settled by looking beyond this horizon, that is, with the help of the extant Greek witnesses. _____________ 379

As edited by M. Rashed and A. Rescigno; v. above, footnote 159. Cf. Pieter Sjoerd Hasper, chapter VI.7 above. 381 For the comparison I relied not only on the edition by L. Benakis, Athens 2008, but additionally on Hasper’s collation of Pachymeres’ autograph, ms. Laur. gr. 87,05. 382 As for Physics IV 9, 217b27 — V 1, 224b10, the reconstruction of Ω can of course additionally benefit from the reconstruction of Θ by means of ms. Esc. 896. 380

X.1 Editorial principles: Establishment of the Arabic text

CCXXXVII

It goes without saying that the causes of apparent or ostensible mistakes and obscurities in Ω often cannot be determined with certainty. E.g., frequently we cannot ascertain whether what looks like an omission is due to the scribe of Ω, the oversight or deliberate neglect of the word/passage in question by the translator, or its omission in the translator’s Greek exemplar(s). Analogous considerations may apply to what looks like an addition. Yet other cases of disparity may equally be caused by a corruption within the Arabic transmission, or by the translator’s misreading or misunderstanding of the Greek text, or by various potential kinds of problems in the Greek manuscripts383. Thus, it is evident that we enter here the realm of probability, and this even in a twofold sense or on two different levels of heuristics, first on the objectlevel or level of events, i.e. the material or intellectual occurrences and processes that may have caused a mismatch, and secondly on the level of logical or propositional probability, as far as our (subjective) rational conclusions about the underlying Greek text are concerned. The first level may be illustrated in a general way through the following scenario: The extant Greek manuscripts display two competing readings, the one consisting in some verb form in the optative mood, the other by the same verb in the indicative, both of which equally acceptable in the relevant context, while the Arabic translation of the verb indicates no hypothetical or potential aspect at all. In other words, there is a certain mismatch between the Arabic evidence and those Greek manuscripts which support the optative reading. However, Classical Arabic verbal inflection does not provide a particular mood equivalent to Greek optative, which is why we are faced mostly with indicative or subjunctive verbal forms rendering a Greek optative, while some especially reliable and sensitive translators try to solve the problem at least in cases of a very strong hypothetical or potential aspect by the use of conjunctional paraphrases or particles like qad. Hence, the uncertainty described above is located (a) on the level of the de facto given incompatibility of the Greek and Arabic verbal systems, and (b) in the uncertainty (and the lack of appropriate frequency analyses) regarding the reliability, sensitivity and consistency of the relevant translator. It precedes any kind of propositional probability eventually interfering with an editor’s statement about the Greek text underlying the actually attested Arabic translation. Things are different in cases such as these: (a) The Arabic translation suggests a reading not attested in any extant Greek witness; (b) the Arabic translation suggests a reading attested in only one extant Greek manuscript _____________ 383

For a more detailed discussion of the various categories of obscurities whose causes cannot be determined unambiguously see Gutas, Theophrastus, p. 80—82, 95—97.

CCXXXVIII

Introduction

which otherwise evidently belongs to another branch than the translator’s Greek exemplar; (c) the Arabic translation suggests a reading attested in only one extant Greek manuscript which evidently belongs to the same branch as the translator’s Greek exemplar; (d) the Arabic translation suggests a reading attested in most/all extant Greek manuscripts of a branch different from the translator’s Greek exemplar; (e) the Arabic translation suggests a reading attested in most/all extant Greek manuscripts of the same branch; (f) the Arabic translation agrees with all extant Greek witnesses. Such cases reflect various degrees of (roughly) increasing probability, not on the event level, but rather on the level of their interpretation and propositional depiction by the editor. Here the state of facts on both the Greek and the Arabic sides is more or less certain, but probability and (un)certainty creep in by (subjectively) interrelating the facts. Obviously, relating the Arabic translation to a single Greek witness belonging to the same branch as the translator’s Greek exemplar (case [c]) is more likely valid than relating it to a single Greek witness belonging to a different branch (case [b]). However, as soon as such propositions are abridged according to the requirements of a critical apparatus, they cushion the different degrees of probability by simply recording (without further differentiation) the correspondence between the Arabic translation and the relevant unique Greek reading. Even more problematic is the relationship between cases (a) and (b): Does the reconstruction of a Greek reading on the basis of the Arabic translation contain a higher degree of certainty in case (b) as compared to case (a)? — Probably not so. Nevertheless this might be suggested by the critical apparatus which in case (a) nolens volens emphasizes the uniqueness of the reconstructed reading as opposed to any other witnessed readings, whereas in case (b) this reading will be registered under the siglum of the extant Greek manuscript together with that of the translator’s exemplar. To cut a long story short, no matter which of the two levels of probability is concerned, the information derivable from the Arabic as regards disparities between Ω and parts or the whole of the Greek tradition is bound to various degrees on the scale of probability ranging from perfect certainty over unconditioned probability down to conditioned probability and tendential statements, the likelihood of which cannot be measured exactly, let alone be stated explicitly, and this even “ex negativo”, that is, if the critical apparatus is silent in cases such as the above mentioned optative/indicative problem. On the other hand, we can —in a remarkable number of cases indeed— determine the Arabic transmission as the unique cause of such an obscurity. To mention but one example:

X.1 Editorial principles: Establishment of the Arabic text

CCXXXIX

Physics VIII 2, 253a4-5 πῶς ἂν εἴη, [...] λέγω δὲ ὥστε τὸ αὐτὸ ὑπὸ τοῦ αὐτοῦ κινητικοῦ ὄντος ὁτὲ μὲν κινεῖσθαι ὁτὲ δὲ μή “How comes it about, [...] I mean, that the same thing is at one time moved and at another time not by the same [continuously] existent motive [force]” Isḥāq’s translation was transmitted in Ω as follows: aʿnī kayfa yakūnu l-šayʾu bi-ʿaynihī wa-llaḏī šaʾnuhū an yuḥarrikahū wahuwa wāḥidun bi-ʿaynihī mawǧūdan ḥīnan yataḥarraku wa-ḥīnan lā yataḥarraku “I mean, how comes it about that the same thing and its motive [force], which is one and the same, are something existent which at one time is moved and at another time is not moved.” The cause of this nonsense, which is found in Φ as well as in the two Arabo-Latin branches, was certainly not any misread or mistranscribed Greek text, but the ignorant scribe of Ω, who failed to realise that, in the original text, wa-llaḏī šaʾnuhū an yuḥarrikahū huwa wāḥidun bi-ʿaynihī mawǧūdun was a circumstantial (ḥāl) clause interposed between the subject and the predicate of the main clause. Isḥāq’s original translation ran as follows: aʿnī kayfa yakūnu l-šayʾu bi-ʿaynihī wa-llaḏī šaʾnuhū an yuḥarrikahū huwa wāḥidun bi-ʿaynihī mawǧūdun ḥīnan yataḥarraku wa-ḥīnan lā yataḥarraku “I mean, how comes it about that the same thing at one time is moved and at another time is not moved while its motive [force] is one and the same [continuously] existent [thing].” The scribe of Ω took wa-llaḏī šaʾnuhū an yuḥarrikahū as a second subject and added, in order to preserve a halfway sound syntax, wa- before huwa wāḥidun bi-ʿaynihī. Probably, he is also responsible for reading mawǧūdan instead of mawǧūdun, or this may be an after-effect of the interpolation of wa-. Thus, we can pin down the transmission of Isḥāq ibn Ḥunayn’s translation prior to Ω as the source of the present obscurity, which can be removed through recourse to the Greek tradition. To sum up, we must principally distinguish between (α) obscurities whose causes cannot be determined with sufficient certainty and (β) those attributable either (β.1) to the translator’s Greek exemplar(s), or (β.2) to the process of translation, or (β.3) to the Arabic (or Arabo-Latin) tradition arising from the translator’s autograph. Only in the latter case (β.3) may it be possible to “remedy” an obscurity that occurred already in Ω by means of the received Greek text and to present the word or passage in question in a reliable way as it ran origi-

CCXL

Introduction

nally in the Arabic translation, whereas for all other problems tracing back to Ω the edited text remains fraught with a certain indeterminacy, which has to be indicated in the critical apparatuses as precise as possible, bearing in mind the above described limits. II. After the reconstruction of the Arabic text as closely as possible to Isḥāq ibn Ḥunayn’s Linguistic Text, the second main task of the present edition consists in deriving from the Arabic text information about the underlying readings of the translator’s main Greek exemplar Ψ. As far as this information affects the relationship of the Greek manuscripts in general, it is summarized above (chapter VI). The huge mass of particular variant readings of Ψ, on the other hand, is reported in a separate critical apparatus below the Arabic text. What is recorded there is all disparities between Ψ and the extant witnesses of the direct and indirect Greek traditions which bear upon the general stemmatics of the Greek tradition and/or may be of relevance to a future critical establishment of the Greek text (for details of the editorial principles followed in this apparatus cf. below, chapter X.2, section 3.1). III. In addition to these two main purposes, the present edition aims also to shed some light on the contents, origins and structure of the marginal and interlinear notes and scholia accompanying the Aristotelian text in the Baghdad school compilation (cf. above, chapter II.3). As mentioned above, some of these notes quote, correct or paraphrase the Aristotelian text and thus serve as an indirect source for the establishment of the Arabic text worth being documented as an independent tool of textual criticism. Furthermore, they provide insight into the sources and methodology of the Baghdad Aristotelians. And finally, they provide important information for the historian of philosophy as many of them derive from the Greek commentary tradition of late antiquity. On this basis, it seemed advisable to report all notes and scholia, albeit in various ways (cf. below, chapter X.2, sections 3.2 and 3.3).

X.2 Structure, layout and apparatuses The pages of the Arabic edition are composed of (1) the main text, (2) marginal line numbers and concordances, and (3) three apparatuses. (1) The main body of the text contains Isḥāq ibn Ḥunayn’s Arabic translation as preserved in Φ (ms. Leiden Or. 583) or reconstructed on the basis of the testimonies and according to the methods described above. Since Φ serves as the main source of the edition, folio numbers of this manuscript are given within the main text between square brackets. Whenever the Aristotelian text is interrupted in this manuscript by separate commentary sections, the present

X.2 Editorial principles: Structure, layout, apparatuses

CCXLI

edition, which does not reproduce these comments, displays two folio numbers in immediate succession. The first folio number refers to the end of the preceding Aristotelian lemma, the second one to the beginning of the following lemma. Whenever this following lemma is preceded in the Leiden manuscript by a separate subheading, the latter —being secluded from the main text— is quoted in the second critical apparatus. Such subheadings are of interest for the study of the teaching practice and the segmentation of the Aristotelian text within the Baghdad school, which for its part may be influenced by the division of the text in the Alexandrian commentaries, especially the one by John Philoponus384. Τhe edition aims at reproducing, to the extent possible, the original version of Isḥāq ibn Ḥunayn’s translation. Hence, the main text does not include words or parts of words transmitted in ms. Leiden Or. 583, yet secluded by the present editor. This method is additionally justified by the fact that the identification and elimination of interpolated words or signs is never guided or inspired by the editor’s subjective iudicium, but in each case supported by some or all witnesses of the indirect tradition. The secluded element is registered in the second critical apparatus together with a specification of the reason for its seclusion. Conjectural additions, on the other hand, are indicated according to the usual practice by angle brackets, no matter whether they are taken from witnesses of the indirect Arabic or Arabo-Latin tradition or derived from the editor’s divination. A series of three asterisks (* * *) in the main text indicates a lacuna that can be attributed with a high degree of probability to Isḥāq’s Greek exemplar(s), either because the same lacuna is evidence in extant Greek manuscripts or because we can determine a Greek saut du même au même as its cause, while a corresponding Arabic saut du même au même can be ruled out. (2) In the right margin of the text body one finds (a) the Bekker lines of the Aristotelian text as edited by Immanuel Bekker (Aristotelis opera, Berlin 1831)385 and (b) the chapter numbers of the ten chapters of Physics bk. VIII. The left margin provides (a) line numbers of the main text and (b) concordances to the Arabo-Latin witnesses used for the establishment of the text, to be more precise to the folio numbers of the three manuscripts of Gerard of Cremona’s translation taken into consideration (= Ca, Cp, Cv, cf. above p. CCII–CCIII) and to the lemmata or textus of Averroes’ “Long Commentary” (= Av. t. #), _____________ 384

Cf. Giannakis, Philoponus, p. 44, 54, 74. Note that due to the different syntax of Greek and Arabic or through transpositions of syntagmata or clauses by the translator, the marginal Bekker numbers often refers to the middle or the end of the Arabic line next to which it is printed and provides no more than a rough orientation.

385

CCXLII

Introduction

which are commonly used for references to Latin versions of Averroes’ ad litteram commentaries. (3) The main text is accompanied by three apparatuses which serve different purposes. In what follows I refer in top-down order to these apparatuses as “App. 1”, “App. 2” and “App. 3” (the same abbreviations are also used for crossreferences within the apparatuses themselves). As a general rule, I favoured in the apparatuses clarity and legibility over economy. Each apparatus entry is preceded by a bold number referring to the relevant line(s) of the main text. Whenever more than one note refers to the same line(s), the bold number is not repeated. Rather, the subsequent notes are separated from each other by an extra-wide space. Within the apparatuses I use various types of sigla and abbreviations, some of which are explained in detail in what follows. For a tabular overview of all sigla and abbreviations, cf. above, p. IX–XVII. The parlance within the apparatuses is a Salvatore-of-Montferrat-like odd pidgin (a titbit for any hostile critic), partly due to my inability to find an alternative, partly enforced by constantly changing writing directions or the diktat of brevity. (3.1) App. 1 is the main tool of information concerning the Greek manuscript(s) used by the translator Isḥāq ibn Ḥunayn and its relation to the extant Greek manuscripts of Aristotle’s Physics. As far as the latter are already taken into consideration in the editions by Bekker and Ross, they are represented in the present apparatuses by the same sigla386. In order to facilitate the application of App. 1 to the available Greek editions and their apparatuses, each entry indicates in parentheses, after the line number referring to the Arabic text (viz. after the extra-wide space before each subsequent item), the column and line number of the so-called Bekker pagination, e.g. “3 (b25)” introduces a note on a word or words found in line b25 of the relevant Bekker page the Arabic equivalent of which stands in line 3 of the actual page of the present edition. The data provided in App. 1 are derived mainly from (α) a complete collation of the Greek text and its variant readings in 14 Greek manuscripts of the direct tradition, in the editions and critical apparatuses of the commentaries by Alexander of Aphrodisias, Themistius, John Philoponus, Simplicius, and (Pseudo-)Psellos, and —indirectly— in the Graeco-Latin translatio vetus. (β) A complete comparison of the reconstructed Arabic text with all materials mentioned under (α). Additionally the following tools are taken into consideration and referred to in App. 1: (γ) information provided in the critical apparatuses of the editions by I. Bekker, H. Carteron, and W. D. Ross; (δ) conjectural emenda_____________ 386

With the exception of the siglum Λ used by Ross for consensus among the manuscripts F, H, I and J. In the present edition Λ refers to the Graeco-Latin vetus translatio of the Physics (analogous to its use in Ross’ edition of the Topics) .

X.2 Editorial principles: Structure, layout, apparatuses

CCXLIII

tions of the Aristotelian text (including deletions) proposed by the above editors or by modern translators (e.g., C. Prantl, J. Wicksteed & F. M. Cornford, R. P. Hardie & R. K. Gaye, H. Wagner, P. Pellegrin, D. W. Graham), or in philological studies on the Aristotelian text (e.g. by H. Bonitz, M. Hayduck, H. Diels, A. Torstrik); (ε) various lexicographical tools, among others the Greek & Arabic Lexicon (eds. G. Endreß, D. Gutas), the Wörterbuch zu den arabischen Übersetzungen des 9. Jahrhunderts and its supplements (ed. M. Ullmann), and of course the Greek-Arabic and Arabic-Greek glossaries of the present edition. The major purpose of App. 1 consists in calling attention to disparities between the Greek exemplar(s) of the translator and the received Greek text of the direct and indirect traditions387. No matter whether we are faced with a reading attested in one branch of the Greek manuscripts and challenged in another or with a reading not attested in any extant Greek witness, its value for any future edition or revision of the Greek text is self-evident. However, the preconditions, methods of reconstruction and degrees of preciseness and certainty underlying any statement about disparities and correspondences between the Greek and Arabic texts show considerable differences and require various modes of notation (on certainty and probability see also above, chapter X.1). Each type of relationship between the Arabic evidence and the Greek text witnessed by or reconstructed through it is indicated in the apparatus by a particular siglum: (Ψ) In general, the Greek reading reconstructed on the basis of the Arabic main text is given under the siglum Ψ, representing the translator’s main exemplar. For the most part, the Greek text can be reconstructed (given it is reconstructable at all) on the basis of Isḥāq ibn Ḥunayn’s translation as transmitted correctly and unambiguously in ms. Leiden Or. 583 (Φ). In a number of instances, the basis of reconstruction is more complicated or requires undertaking intermediate investigations and/or conjectures. In detail, we can distinguish between three types of evidence and reconstruction, visualised in App. 1 by specific sigla as follows: (ΨΩ) As we have seen above, the precious value of Φ is marred by numerous omissions and scribal errors. In many cases it is possible to recover the original Arabic reading by means of Σ, the common Andalusian ancestor of the AraboLatin translations reconstructed on the basis of the extant Latin witnesses. Less frequently Φ can be corrected by readings of Y (reported in the margin of Φ) and/or by the indirect Arabic tradition, i.e. quotations of Isḥāq ibn Ḥunayn’s _____________ 387

For the following explanations on App. 1 I greatly benefited from Dimitri Gutas’ “Principles of Graeco-Arabic Textual Criticism and Editorial Technique”, in Gutas, Theophrastus, p. 93—101.

CCXLIV

Introduction

translation in the commentary sections preserved in Φ. In all these instances the common ancestor of these branches, Ω, must still have preserved the correct Arabic text. Thus, any Greek reading recovered on the basis of an Arabic text that has been reconstructed by means of the above witnesses of the indirect Arabic and Arabo-Latin tradition, because Φ is corrupt or lacunose at that place, is registered in App. 1 under the siglum ΨΩ. In that case, the related Arabic and/or Arabo-Latin evidence justifying and explaining the correction of Φ (viz. the reconstruction of Ω) is documented in App. 2. (Ψ*) In other cases we are faced with obvious corruptions of Ω which — although unambiguously attributable to the faulty transmission of Isḥāq ibn Ḥunayn’s original version— cannot be resolved with the help of the indirect Arabic and Arabo-Latin traditions. However, some of these corruptions can be removed with the help of the received Greek text. It goes without saying that such an emendation is reliable and straightforward only (a) in cases of consensus among all Greek witnesses, or (b) if the corrupted Arabic text (and its immediately surrounding context) is such that it can be related unambiguously to one Greek branch, to the exclusion of any other. Although the reasoning in these cases is to a certain extent circular, the result deserves to be registered in the apparatus, if the conjectured Arabic reading corresponds with a Greek reading which may affect the constitution of the Greek text or is controversial among its modern editors. (c) Other corruptions concern issues of Arabic grammar or syntax that cannot be removed on the basis of the Greek text, but —if at all— by divination only. Following the general principles outlined below (cf. Focus and exhaustiveness), App. 1 is silent in most cases of type-(a) emendations. Likewise, it is silent whenever an obvious corruption of Ω has been resolved by divination such that the restored Arabic text does not admit an unambiguous reconstruction of the translator’s exemplar. On the other hand, if the cause of the corruption can be easily identified and the original Arabic text can be restored in an unambiguous way, the underlying Greek reading can be recovered with reasonable certainty. In order to keep such conclusions based on type-(b) and type-(c) emendations apart from straightforward correspondences between Ω (viz. Φ) and Ψ, the relevant reconstructed Greek readings are registered in App. 1 under the siglum Ψ*. (ξ) Again another level of reconstruction is formed by those cases where the Arabic manuscript offers in the margin or between the lines an alternative reading —explicitly labelled as such— which evidently reflects a Greek reading different from the one reconstructable on the basis of the manuscript’s main text. Such variae lectiones may originate from the translator’s consideration (and translation) of γράφεται-notes in his main Greek exemplar (Ψ). However, they may just as well trace back to (an)other Greek manuscript(s). To be more precise,

X.2 Editorial principles: Structure, layout, apparatuses

CCXLV

we may be faced with the outcome of the translator’s collation of additional Greek manuscripts or with a collation note by a later Arabic-writing scholar who had access to another Arabic translation of the Physics (or a commentary quoting the Physics), which for its part drew on a Greek manuscript other than Ψ. No matter how the alternative readings came about, they may provide valuable information for the establishment of the Greek text or about the circulation of Greek manuscripts belonging to different branches, etc. In order not to jump to conclusions and to set the Greek text reconstructed on the basis of or attested by such explicit alternative readings apart from the aforementioned types of evidence, the relevant Greek readings are registered in App. 1 under the siglum ξ (notwithstanding that ξ may coincide with Ψ or γράφεται-notes in Ψ). It goes without saying that variant readings registered under the siglum ξ must not be confused with marginal variae lectiones in the Arabic or AraboLatin tradition arguably tracing back to one and the same Greek text. Such notes may have been caused by a problem within the Arabic manuscript tradition or by some scholar’s collation of another Arabic translation which displays another Arabic phrasing or terminology (variae translationes) than the manuscript’s main text, but nevertheless reflects the same Greek reading. These notes do not provide any significant information about the underlying Greek exemplar and are not taken into consideration in App. 1, while they are evidently relevant for the constitution of the Arabic text. Hence, they are registered in App. 2 (cf. below). Focus and exhaustiveness. In principle, App. 1 is silent if the information inferable from the Arabic translation coincides with a reading that is unanimously confirmed by all extant direct and indirect Greek witnesses. In the majority of cases, this information is valueless; but there is a single exception to this general rule: In the case of a unanimously transmitted reading that has been rejected by the conjecture of one or more modern editors of the Greek text, it is worth noting that this reading is likewise confirmed by the Arabic tradition, because the latter traces back to a stage of transmission prior to the extant Greek witnesses and thus may help to determine the terminus ante quem of the rejected reading. Similarly, App. 1 is silent in cases of variant readings indiscernible on the basis of the Arabic translation such as mere orthographic disparities without semantic implications (e.g. γίνεται vs. γίγνεται, πᾶν vs. ἅπαν, αὑτὸ vs. ἑαυτὸ, etc.) or cases which are indistinguishable due to (i) the lack of preciseness in the Arabic (as often the case with Greek particles), (ii) the different requirements of grammar, syntax and word order in Greek and Arabic, or (iii) the synonymy of the competing Greek readings (e.g. ἐπ’ εὐθείας vs. κατ’ εὐθεῖαν, both = ʿalā l-istiqāmati, ἐξ ἀνάγκης vs. ἀναγκαῖον, both =

CCXLVI

Introduction

ḍarūratan, etc.). Accordingly, slight changes in the word order or the omission or addition of the definite article or of certain particles in the Greek witnesses are taken into consideration in App. 1 only if they are unambiguously discernible on the basis of the Arabic text. Thus, the absence of any note in App. 1 by no means indicates that the Greek transmission of the section in question is coherent and devoid of textual problems, nor in case of a unanimous Greek transmission that the translator’s exemplar must have preserved precisely this text ! As a general rule, App. 1 pays more attention to how Ψ (viz. ΨΩ and Ψ*) relates to ms. E and the manuscripts which show a certain affinity to branch α, especially mss. K, b and e, than to the witnesses of the branches β (LNR) and γ (FHIJQw, probably also Λ). The reason for this procedure is the fact that Ψ is of special benefit for a more reliable assessment of the two hands of E, establishing the readings of the common ancestor of branch α on a broader and more secure basis, and determining the relationship between this branch and the manuscripts stematically “intermediate” between α and γ. As a consequence, all determinable disparities between Ψ and E are registered in the apparatus, whereas individual variant readings of one manuscript of either branch β or branch γ are not registered, provided that Ψ agrees with the remaining witnesses against this reading. That is not to say that such readings are irrelevant for the constitution of the Greek text. Recording them simply would have blown the apparatus out of proportion, while the Arabic translation cannot contribute substantially to their evaluation. However, if the relevance of such an individual reading of the direct tradition is further reinforced by the indirect tradition or through the fact that it has been adopted by one of the modern editors, or if it coincides with further disparities in any other group, the apparatus is explicative. Order of notation. Sigla of Greek manuscripts belonging to the same branch are written next to each other, those representing manuscripts of different branches are separated by blanks. If the recovered Greek reading also appears in one or more of the extant Greek witnesses, it is of course registered in the apparatus only once (such that the siglum Ψ is grouped together with the sigla of the relevant manuscript[s]). In view of the close relationship between Ψ and the Parisinus gr. 1853 (E) and its importance for the establishment of the Greek text, I regularly strived to register a common reading of E (viz. E1 or E2) and Ψ (= α) challenged by other witnesses at the beginning of each entry in App. 1, then those of the “intermediate” manuscripts of branches K and δ, to be followed by readings of the branches β and γ. However, for the sake of clarity this rule had to be broken occasionally, in order to ensure an unambiguous reference of the keyword in the apparatus to the relevant word(s) in the modern

X.2 Editorial principles: Structure, layout, apparatuses

CCXLVII

editions. Readings attested in the works by the Aristotelian commentators are registered after the witnesses of the direct tradition in chronological order of the commentators. Particular types of notation; abbreviations. As explained above, it is often impossible to determine whether a certain disparity (i) emerged in the process of translation or (ii) is due to an error or variation in a Greek or Arabic manuscript. Despite this uncertainty, it is in some of these cases at least possible to ascertain which of the latter two —the translator’s Greek exemplar(s) or the Arabic manuscript tradition— must have caused the obscurity, given that it is not the translator who is to be blamed for it. E.g. Physics VIII 10, 267b9, where the translation of ἔχει δ’ ἀπορίαν εἰ ἐνδέχεται κτλ. runs as follows: wa-hāhunā mawḍiʿu šakkin wa-ʿasā qad yumkinu ilḫ. The translation of εἰ through wa-ʿasā is not quite to the point. On the other hand, it is not so remote from the Greek that it points unambiguously to a variant reading in the Greek or an error in the Arabic tradition. In other words, the case is vague and indifferent as far as alternative (i) is concerned. However, what we can indeed ascertain is the fact that, if the disparity is not due to the translator, its cause is not to be sought in the Arabic transmission, since Isḥāq translates εἰ introducing indirect questions otherwise consistently through hal, a scribal confusion of which with wa-ʿasā can be ruled out. Hence, if the cause of the disharmony between εἰ and wa-ʿasā is extrinsic to the translation process proper, it must be sought in a Greek variant reading such as εἴ ποτε instead of εἰ (as Isḥāq uses ʿasā at other places to translate ἄν ποτε, 209a14, and μή ποτε, 259b3). In cases like this the potential variant reading is noted together with a question mark in App. 1. Conversely, whenever it is possible to determine that an apparent disparity may be due either to the process of translation or to the transmission of the Arabic text, yet surely not to a variant reading in the Greek, this is registered with a question mark in the “Arabic” apparatus (App. 2). Other indeterminable or vague cases require different strategies of notation. In case of a Greek word or phrase attested in all extant Greek witnesses, yet without any correspondence in the Arabic translation it is often difficult to decide whether we are faced with an omission in the translator’s Greek exemplar(s), an oversight or deliberate ellipsis by the translator, or a post-translation phenomenon of the Arabic tradition. In order to avoid unjustifiable prejudgements, such cases are noted in App. 1 not with “om. Ψ”, but with “non vert. Ar.” indicating that the Arabic translation as extant today in the direct and indirect traditions lacks any equivalent for one of the three potential reasons, yet without necessarily implying the omission of the relevant word(s) in the translator’s exemplar.

CCXLVIII

Introduction

A similar problem is raised by words or phrases added in the Arabic translation as compared to the extant Greek witnesses. In a rare number of cases it is fairly clear that Isḥāq ibn Ḥunayn must have found a corresponding addition in his exemplar(s), because (a) his translations are generally devoid of arbitrary or redundant additions and (b) there is no intrinsic requirement for the addition, no matter whether in terms of syntax or semantics388. E.g. at Physics VIII 3, 253b30 the sentence begins in all extant manuscripts with ὁ δὲ λίθος or ὅ τε λίθος. Although the chain of thought is clear and the phrase could have been easily rendered by wa-ammā l-ḥaṣā fa- (= ὁ δὲ λίθος) or wa-l-ḥaṣā ayḍan (= ὅ τε λίθος) or the like as in many similar cases, Isḥāq’s translation reads waayḍan fa-inna l-ḥaṣā. Since Isḥāq never renders simple δέ or τε with wa-ayḍan fa-inna, but consistently uses this expression for ἔτι or ἔτι δέ, there can be little doubt that his exemplar read ἔτι δ’ ὁ λίθος or the like, which is accordingly registered in App. 1 under the siglum Ψ. However, if the Arabic text fleshes out a syntagma which in the received Greek text is elliptical due to Aristotle’s brachylogy or for any other reason difficult to understand, we do not know whether the translator found himself constrained by the requirements of the Arabic syntax or by his aim to prepare an elaborate and easily readable translation to add what he thought to be implied by Aristotle, or whether this elaboration traces back to a Greek scribe389. Semantic disparities of this kind are therefore registered in App. 1 not under the siglum Ψ, but with the neutral specification “intell. Ar.”, affording either possibility, sometimes accompanied for the sake of clarity by an English translation of the Arabic version. Again, the reading noted together with this specification actually may have been present in Ψ. Some disparities seem to be caused by the translator’s idiosyncratic parsing or punctuation of the Greek text, no matter whether or not this may have been indicated in his Greek exemplar(s)390. The result is an Arabic text which at first _____________ 388

Such as words or phrases in need of explanation for the addressee of the translation, who is unfamiliar with the Greek language. For such cases see my distinction between Conceptual Text and Linguistic Text above, p. XXIII–XXV. 389 For examples of this phenomenon in the Greek tradition of Aristotelian manuscripts cf. Diels, Zur Textgeschichte der Aristotelischen Physik, p. 15—17; Mansion, Étude critique, p. 27. 390 On punctuation in Greek manuscripts of Late Antiquity cf. David L. Blank, “Remarks on Nicanor, the Stoics and the Ancient Theory of Punctuation”, Glotta 61 (1983), p. 48—67; Kyriakos Tsantsanoglou, “Punctuation”, in A.-F. Christidis, with the assistance of M. Arapopoulou, M. Chriti (eds.), A History of Ancient Greek. From the Beginnings to Late Antiquity, Cambridge 2007, p. 1326—1333. For punctuation marks in the so-called Philosophical Collection (9th c. Byzantium) see Lidia Perria, “L’Inter-

X.2 Editorial principles: Structure, layout, apparatuses

CCXLIX

glance seems to be based on a wording different from the received text, although the obscurity can be explained without this assumption. The relevant phrases, which are surely of interest for the history of Aristotelian commentaries, are quoted in App. 1 together with the note “i’punxit Ar.”. Somehow intermediate between these latter cases, which are plausibly deducible without the assumption of a substantial variation of the Greek text, and those which definitely point to a particular determinable Greek wording, we encounter a third class of disparities. Such disparities are characterized by the fact that the Greek text reflected in the Arabic translation was basically the same (or at least supposed to be so by the relevant Greek scribe) as the one attested in all witnesses or in a particular branch of manuscripts, yet was received by the translator in an unclear or slightly modified form or in a way that left some room for interpretation. Most likely the translator was faced here with an ambiguity due to missing accents, Iota subscriptum missing or mistaken for Iota adscriptum and v.v., partially illegible or otherwise (due to physical damage) dubious words, or sketchy and unclear corrections in his exemplar. E.g., if the Arabic translation reflects the word αὐτῷ, while all extant Greek witnesses read αὑτῷ, we cannot ascertain that the translator’s exemplar(s) indeed read αὐτῷ. It may equally have read ΑΥΤΩ(I), αυτῷ, αυτῳ or αυτω understood by the translator as αὐτῷ. At Physics VIII 2, 252b14, all manuscripts read ὧν οὔτε μέρος οὐδὲν, while the Arabic translation reflects ὧν οὐδὲν οὔτε μέρος. For a plausible explanation of this phenomenon it is sufficient to assume that the scribe of Ψ had added οὐδὲν (possibly after a careless omission) above οὔτε μέρος, which was then understood by the translator in the above word order. Other cases of a slightly changed word order that may have come about in this way occur at Physics VIII, 253b29-30, 255a34 or 261a15. Similarly at Physics VIII 5, 257b22-23, the Arabic translation (mutaḥarrikan faqaṭ) seems to be based on the Greek phrase κινούμενον μόνον, while all extant witnesses read only κινούμενον. Again, this discrepancy may be caused either by an unclear correction of the ending of κινούμενον written above κινού... or by a dittography of the word ending interpreted by the translator as meaning κινούμενον μόνον, while the scribe of Ψ essentially intended nothing else than κινούμενον391. Such cases, which seem to be based on a certain interpretative decision by the translator required by a minor obscurity in his exemplar, are noted in App. 1 with “interpr. Ar.” instead of providing guesswork about the underlying reading of Ψ. In most instances, they can be clearly kept apart from the category of disparities punzione nei manoscritti della «Collezione Filosofica»”, in Harlfinger, Prato et al. (eds.), Paleografia, p. 199—209. 391 Cf. above, p. CLXXXI, ad 253a2.

CCL

Introduction

subsumed under “intell. Ar.”, although the boundaries between the categories are at some places rather fluid. As a rule of thumb, Greek readings mentioned in the apparatus together with “interpr. Ar.” are more likely to have been actually present in the translator’s exemplar than those registered under “intell. Ar.”. In order not to overload the apparatus, I used the following abbreviated modes of notation in App. 1. (i) If Ψ displays a unique variant reading where all witnesses of the direct and indirect Greek traditions have one and the same different reading, the note gives only this latter reading (without reference to its witnesses) followed by a closing square bracket and the reading suggested by Ψ. (ii) If the disparity concerns only the indirect Greek tradition (late antique commentaries and Λ), while there is agreement among all witnesses of the direct tradition taken into consideration, the latter are subsumed under the siglum Π. (iii) Often the Greek witnesses display three or more competing readings, say δεῖ, δὲ and δὴ. In some of these cases it is not possible to associate Ψ unambiguously with one of these variant readings, while it is indeed possible to exclude one of them as the source of the Arabic reading, for instance Ψ read either δὲ or δὴ, yet certainly not δεῖ. In order to save space, such constellations are occasionally registered in App. 1 by means of the “either—or” symbol (∨) without repetition of the relevant text as follows: “δεῖ {ms. X} : δὲ {ms. Y} : δὴ {ms. Z} : Ψ = Y ∨ Z” (i.e. Ψ = (Y ∨ Z) ∧ ¬ X). A preventive admonition. App. 1 documents the Greek readings evidenced by or reconstructable from the Arabic by means of the above procedures. It should not be read as reflecting the editor’s proposals for how to read or modify the Aristotelian text, no matter whether the received Greek text is absolutely confirmed by the Arabic tradition (in which case the apparatus is silent), only partly supported or entirely disproved through unique readings suggested by the Arabic translation. (3.2) App. 2 is the main tool of information concerning the establishment of the Arabic text. The data provided in this critical apparatus are derived from various sources and procedures, namely (α) a complete collation of Physics VIII in ms. Leiden Or. 583 with the extant witnesses of the indirect Arabic tradition and the editions and manuscripts of the three Arabo-Latin translations; (β) the collation and consideration of terminological and phraseological peculiarities of the remaining books of the Arabic Physics; (γ) conjectural emendations of the Arabic text proposed partly by ʿAbdurraḥmān Badawī in his edition and partly in the “Index of Variant Arabic Passages” of the Greek & Arabic Lexicon (GALex); (δ) the direct and indirect witnesses of the Greek text of Physics VIII, wherever required for the constitution of the Arabic text; and (ε) various lexico-

X.2 Editorial principles: Structure, layout, apparatuses

CCLI

graphical tools, among others the Greek & Arabic Lexicon, the Wörterbuch zu den arabischen Übersetzungen des 9. Jahrhunderts and its supplements, and the glossaries of the present edition. Since ms. Leiden Or. 583 (Φ) is the only complete extant Arabic witness, App. 2 gives full documentation of all disparities between Φ and the edited Arabic text as well as of all kinds of ambiguities displayed by Φ (with the exception of missing diacritical point, provided the reading is unambiguous given the context or in light of the received Greek text and the indirect tradition). For the same reason, all diverging readings in Badawī’s editio princeps and all emendations by Badawī (no matter whether these were adopted or rejected in the present edition) are registered in App. 2 (excepting obvious mistakes and misprints). Due to the deficient and corrupt character of Φ, the Arabic text edited here (= Ed.) has been established at many places with the help of the indirect tradition or the received Greek text or both together. In all these cases App. 2 reports not only the reading of Φ, but together with the edited reading also the rationale, that is, the reading(s) and source(s) on the basis of which the change was made, added in parentheses. The advantage of this method (as compared to presenting the parallels and sources in a separate appendix) is obvious, although admittedly it burdens the user with the task of disentangling data of evidenced and conjectured readings from information about the editor’s establishment of the relation between these readings and other material. E.g. in the translation of Physics VIII 1, 250b26 the word ǧaʿala, omitted in Φ, had to be added. The corresponding note in App. 2 runs as follows:

‫ ﺟﻌﻞ‬addidi (= ἐμποιῆσαι Π = fecit CV) : om. ex haplogr. Φ : ﻃﺒﻊ‬Bad thus indicating that and how the addition of ǧaʿala is supported by all Greek manuscripts of the direct tradition (= Π) and by the two Arabo-Latin translations (CV), as well as that Badawī proposed another reading and why this has been rejected. Similarly, the apparatus entry on VIII 2, 253a16 registers the editor’s emendation as follows:

‫ ﻋﻦ اﶈﻴﻂ‬scripsi (= ὑπὸ τοῦ περιέχοντος Π = a continente CV) : ‫ ﻋﯩﺮ اﶈﻴﻂ‬Φ : ‫ﻏﲑ‬ ‫ اﶈﻴﻂ‬Bad In both instances, the emendation is evidently supported by the received Greek text and the Arabo-Latin witnesses. Despite this evidence, or rather because the constitution of the Arabic text is based on non-Arabic evidence, I deemed it helpful to specify also the exact pathway and agent of textual change instead of simply giving the readings warranting it. For that purpose not only conjectures provided by the present editor without support from Greek, Arabo-

CCLII

Introduction

Latin or other Arabic sources (“divinations” or emendationes ope ingenii) are indicated in App. 2 by terms like “addidi”, “scripsi”, etc., but also conjectures of the above type (emendationes ope codicum Graecorum vel Latinorum). Any conjecture provided by ʿA. Badawī (who did not take into consideration the AraboLatin translations) is indicated in App. 2 with the abbreviation “Bad”. Dubious, yet favourably tolerable readings of Φ which are supported by the indirect Arabic and/or Arabo-Latin tradition have been retained in Ed. In such cases, App. 2 offers proposals for possible conjectural emendations in the form “an ... legendum/addendum/delendum ?”. In general, disparities between Ed. and a particular witness (or group of witnesses) of the indirect Arabic and/or Arabo-Latin tradition are not registered in App. 2, if Ed. is in accordance with Φ and the received Greek text and is supported by the remaining witnesses of the indirect tradition. However, if the whole indirect tradition differs from Ed.=Φ, this is made explicit in App. 2, no matter whether its witnesses attest one and the same variant reading or differ among each other. Furthermore, the following disparities are reported in App. 2, even if Ed.=Φ is partially supported by the indirect tradition, because they may have a bearing on the establishment of the Arabic text or, more important, may be of interest for the editor of the Arabo-Latin translation or for diachronic studies of the Arabic reception of Aristotle’s Physics: (1) Two different readings of the indirect Arabic and/or Arabo-Latin tradition correspond with two different attested Greek readings, while eo ipso only one of them agrees with Ed.=Φ (such cases may point to variae lectiones in the Arabic or Arabo-Latin tradition derived from alternative Greek sources392). (2) The correspondence between Ed.=Φ and the received Greek text rests on a particular interpretation of the Arabic rasm (skeleton without diacritical points and orthographic signs), while parts of the indirect Arabic and/or Arabo-Latin tradition attest another interpretation of the same rasm. (3) The disparity coincides with a conjectural modification by ʿA. Badawī. As for the evidence provided by the three Arabo-Latin translations, App. 2 refers consistently to C (Gerard of Cremona) and V (Michael Scot). Since C and V are complete translations, App. 2 indicates cases where one of the two translations cannot contribute to the establishment of the Arabic text (e.g. due to an omission or because the Latin wording does not admit an unambiguous reconstruction of the translator’s Arabic exemplar) by bracketing the relevant siglum, e.g. “C[V]”. The third, anonymous translation (Z) is extant only for _____________ 392

Cf. above, p. LXIII–LXVII, CCXLIV–CCXLV.

X.2 Editorial principles: Structure, layout, apparatuses

CCLIII

the sections 265a27-b16 and 266b6-267b26. However, the fact that Z is of no avail for the establishment of the Arabic text corresponding to 250b10–265a26 and 265b17–266b5 is tacitly taken for granted and not indicated by “[Z]” in the apparatus notes on these sections. In a rare number of cases slight grammatical changes seemed appropriate, although neither the received Greek text nor the indirect Arabo-Latin tradition could reasonably be adduced as evidence due to the different grammatical structures of the languages. E.g. if Φ gives the predicate of kāna (or its sisters) in the nominative, because it is separated from the latter by a long sequence of other text, or if Φ ignores dual agreement, although the agreement is observed immediately before and after, this has been corrected in the established text for the sake of consistency and noted in App. 2 without reference to the Greek or Latin sources. By analogy with App. 1 the specification “interpr.” is used in App. 2 to indicate that the reading of an Arabo-Latin source is based on a specific interpretation of the Arabic text by the translator. E.g. if Michael Scot rendered an Arabic word with the rasm (skeleton) ‫ ﯨﻜﻮن‬with generatio (i.e. in the sense of takawwun), while the correct reading is yakūnu, this is registered as “interpr. V ”, because we do not know exactly what Scotus’ exemplar looked like at this place and there is no need to presuppose a different Arabic text. Similarly, what looks like an omission in e.g. Gerard of Cremona’s translation tracing back to his exemplar car, while it may equally be a deliberate ellipsis or an oversight by the translator, is noted not as “om. car”, but as “non vert. C ”. It goes without saying that for discrepancies of this type it is impossible to rule out that they were caused not by the translators (or, for that matter, within the post-translation Latin tradition), but by the scribes of their Arabic exemplars (viz. car or vzar). App. 2 also reports marginal glosses accompanying the Aristotelian text in Φ (cf. above, chapter II.3) as far as these directly concern the constitution of the text, e.g. by providing a variant reading from another manuscript or supplying a word or phrase omitted in the main body of text, etc. On the other hand, marginal notes and scholia which have no bearing on the constitution of the text are reproduced in App. 3, below. (3.3) App. 3 is not a critical apparatus sensu stricto, although it refers like the preceding apparatuses to the line numbers of the main text. This apparatus contains all glosses and scholia accompanying the Aristotelian text in Φ other than those registered in App. 2. Some of them are entirely omitted in Badawī’s 1964 edition, others are reproduced there incompletely or with mistakes or by indicating wrong sections of the Aristotelian text as their object of reference.

CCLIV

Introduction

Basically we can distinguish between three types of reference between scholia and main text with an increasing degree of ambiguity: (i) Both scholium and Aristotelian text referred to are marked by the same signe-de-renvoi 393 or are interconnected by a solid or dotted line in Φ. (ii) The Aristotelian text referred to is marked by a sign which has not been reproduced at the beginning of the marginal note. This is unproblematic as far as the scholium in question is not accompanied by other marginal notes, yet may cause confusion if other notes are written next to it. (iii) Neither the Aristotelian text nor the marginal note displays any graphical hint as to which word or phrase the author of the note refers to. In this case, the relationship must be established on the basis of the contents of the note and its position in relation to the main text of Φ. In order to make this latent ambiguity visible in App. 3, the type of reference is explicitly specified (at the risk of redundancy) in each single case. Entries concerning notes of type (i) or (ii) give after the line number the word or phrase marked by a superscript sign in Φ. Type-(iii) scholia are reported without any keyword with the exception of rare cases where this could be unambiguously deduced from the content of the scholium. In any other case they are registered under the line number of the main text next to which they begin, although they may actually relate to a preceding or following passage of the Aristotelian text. Due to the fact that many glosses are difficult to read and/or partly cut off at the outer margins of the Leiden manuscript, App. 3 also contains a lot of critical annotations concerning the state of transmission, actual emendations and proposals for variant readings. Finally, whenever I was able to identify literal or doctrinal parallels in the commentaries on the Physics, I noted these at the end of each apparatus entry. For the most part this concerns the commentaries by John Philoponus and Abū l-Faraǧ ibn al-Ṭayyib. More important, a remarkable number of scholia are closely related (partly literally corresponding) to the extant fragments of the commentary by Alexander of Aphrodisias (otherwise not extant in Arabic; cf. above, chapter II.3.1). Furthermore, it was in many cases possible to point to related sections in Simplicius’ commentary, while the parallel sections of the commentaries by Alexander and Philoponus could not be taken into consideration due to their fragmentary preservation. As far as we know, Simplicius’ commentary was not available to the Baghdad Aristotelians in a Syriac or Arabic adaptation. Accordingly, the actual sources of Arabic glosses mentioned in App. 3 together with Simplicius’ commentary are to be sought not in the latter, but in related sections of Alexander’s and Philoponus’ commentaries adopted or paraphrased by Simplicius, but also accessible in some 10th century Arabic adaptations. _____________ 393

On which see p. LV f.

‫كتاب السماع الطبيعيّ لأرسطوطاليس‬ ‫المقالة الثامنة‬

‫‪Critical Edition of the Arabic Text‬‬

Sigla and abbreviations See above, p. IX–XVII

Signs ***

In the main text indicates an omission in Ψ

*/**

In the critical apparatuses indicates 1/2 illegible letters or parts of a word cut off at the margin

+

The letter(s) / word(s) after + are added above or below the line or in the margin of the manuscript, as specified)

< ... >

Editorial addition to the text as transmitted in Φ

† ... †

Crux, a word or textual passage corrupted in Φ

[٧‫ﺏ‬٢٠٩]

Fol. 209v, line 7 of ms. Leiden Or. 583

(iLÑPÕpSri  fÏ yibpLA û4LA êÑtK nMÑoLA LÑqmLA

Ca 23vb Cv 66v

Av. t. 1 Cp 171r

5

~gIÒ  ÙsfT øEÖ ø³bQ \kT GÖ KaA OÙH øE cyW iL ÑenÏkL ÙsfT iÖ ÔÙW G KaA ‡Ò  …ÏhtI ÆCÒ  a ñyM YÕkI i DÑËF „ £fT siLÖ ñL dÒÕB i MÒ  ÒÚEÖ ÙBÒ  è҂T iÖ ,M .F è‚T G ?yibpLÑB ñMÒÕQ ÑM õim` ÑM ÑiH ÑeNÏÑ  K øB ÒdÕJÕmLA 1–2 Tit. Α Æ ριστοτε λους ΠεριÁ κινη σεως τω Ä ν ειÆ ς Γ τοÁ Γ. Θ E : Α Æ ριστοτε λους ΠεριÁ κινη σεως τω Äν ειÆ ς τρι α τοÁ Γ Α Æ ριστοτε λους ΦυσικηÄ ς αÆ κροασε ως Θ b : Α Æ ριστοτε λους ΦυσικηÄ ς αÆ κροασε ως ÍΟγδοον e : Α Æ ριστοτε λους ΦυσικηÄ ς αÆ κροασε ως Θ NR : Ψ fort. = e ∨ NR [nisi βιβλι ον Θ/οÍ γδοον intell. Ar.] : Α Æ ριστοτε λους ΠεριÁ κινη σεως Η K : Θ w : Α Æ ριστοτε λους ΦυσικηÄ ς αÆ κροασε ως Θ ηà ωë ς ΠεριÁ κινη σεως Γ L : αÆ ρχηÁ τουÄ Θ F : Α Æ ριστοτε λους ΦυσικηÄ ς αÆ κροασε ως I : Α Æ ριστοτε λους Φυσικω Ä ν Θ J1 : Α Æ ριστοτε λους Φυσικω Ä ν Η J2 : Α Æ ριστοτε λους Φυσικω Ä ν ÍΟγδοον

3 (b11) πο τερον δεÁ E2 γ δ Λ : πο τερον δε τε (?, ras. post δε) E1 : πο τερον Ψ K β ˇ c(150.1, 194.1) S (1118.12) (b11) ποτε E2 K β γ δ Λ T(209.2) Sl(1118.12) : om. E1Ψ G Q : om. H l

4 (b12) ωÏ στε E2ΨΩ K β γ Λ T(209.3) Sp(1118.17) : ωë ς δ : om. E1

5 (b13) αÆ λλ’ αÆ ειÁ α K e L

FI T(209.4), fort. Sp(1118.23) : αÆ λλαÁ αÆ ειÁ NR H : αÆ λλαÁ ειÆ bw : αÆ λλ’ ειÆ J Λ : αÆ λλ’ Q 1–2 hyibpLA û4LA † (iLÑPÕpSrÒ êÑtK \M nMÑoLA LÑqmLA Φ : incipit tractatus octavus Ca : incipit

2 post nMÑoLA add. èÑQ èÖÏi   Ò ylytLA (iLÑPÕpSrÒ Φ : om. ΠCV : seclusi : (iLÑPÕpSrÒ  èÑQ 〈KaA ‡ÙQ〉 1 èÖÏi Ò ylytLA Bad 3 øEÖ] + ÙQ s.l. Φ : om. ÙQ CVGˇp(150.16, 194.1) : secl. Bad 4 DÑËF scripsi (= ut C) : dÑF Φ + D s.l. : DÑËF Bad : in qua V (= Y{B?); (for DÑËF = ωÏ στε cf. GALex I, idan 2; Ullmann NE I, 41; ¯ Glossary) 5 ñL dÒÕB i scripsi (= non habet corruptionem V, cf. la¯ ba¯ idun, la¯ yabı¯du = αÆ θα νατος 203b13 = Bad 212.14, 259b25) : ñLrҁBi vel ñLdҁBi Φ : ñL è҂I i Bad : res que non remouetur ab eo C (= ñnM è҂I i MÒ  ?) siLÖ] + KaA yI s.l. Φ : om. ΠCV oct. *** (illeg.) Cv : sine tit. CpV : lac. ind. Bad

250b10

Chapter 1

250b15

(iLÑPÕpSrt fÏ yibpLA û4LA êÑtK \M nMÑoLA LÑqmLA

2

dÕJÕM 〉 Í KH YÏ Ñ  B èÑQ ÙQ yibpLA MÒ  „ ~miW èÑQ \M õi? YÏ ÑËF .F ÒցznIÖ GÑyLA ynC „ ÒÕmlÏktI 〈 YÒ  ˆŽkm 〉 G ~yi? ˆeNÏÑËF ÒÕLÑQ \IÚLA \Ï kL KH \kT G ÒDÒË dÑsfLÒÖ YÕkLA „ Mi Ò ñilY cO YÕLÕqI ˆeF ÙsfI љyBÖ YÕÏktI љyB YÏ zÖ Iш jB ¡oK GÒÕyB ÙQ Ñ= 〈 LA 〉 dÑsfLÒÖ YÕÏktLA ŸÑnCÒ  YÏ Ò  †LDÖ qÒd KT ñNÏÒ  Ö ÙHÒÖ GÑyLA YÏ Ñ  B ÒÕLÑQ \IÚLA ÑMÏÑ  F . ÑKH õM YÕkT YÒ  Í rÖ² úO .†LD úsH ‹Y ~gIÒ  KaA † ˆEÙqY YÏ ÑËF ÑI­ ÙBÒ  (iL 〈

Cp 171v 3

6

1 (b15) µεÁν Π(praeter E) Ψ Λ Sl(1120.4) : µηÁ E

2 (b17) παÄ σιν Π(praeter L) Ψ Λ

S (1120.17, mss. AM) : παÄ σαν L S (ms. F) : om. Pa(357.16) (cf. Torstrik 528f.) c

c

(b17) ηÊν]

2

ηË ν (?) E : non vert. Ar. (τοÁ κοσµοποιειÄν καιÁ ... παÄ σιν αυÆ τοιÄς αÆ δυ νατον υë πα ρχειν intell. Ar.)

3 (b17) καιÁ Π(praeter KQ) Ψ Λ T(209.7) Sc(1120.17) Pa(357.16) : καιÁ περιÁ K : καιÁ cum ras. Q

(b18) οÏσοι α K b NR T(209.8) Sl(1120.28) : οÏσους e L : οë πο σοι γ Λ Pa(357.19)

(b18) τε E γ Λ : om. K β δ T(209.8) Sl(1120.28) Pa(357.19) : non vert. Ar.

4 (b18–19)

κο σµους Π(praeter E ) Ψ T(209.8) S (1120.28) Pa(357.19) : κο σµου E

(b19) ειËναιÂ

1

l

Π(praeter K) Λ : om. K : non vert. Ar.

1

6–7 (b22) ηà εÏ να ηà µηÁ E β S (1122.26–27, mss. 1

c

AM) : ειËεν αÆ ειÁ µηÁ E : εÏ να καιÁ µηÁ Ψ Pa(358.2–3), fort. As(frag. 541) : εÏ να ηà µηÁ K γ δ Λ 2

Sc(ms. F) : εÏ να 〈 ηà αÆ ειÁ 〉 ηà µηÁ interpr. T, scr. Ross, Wag

1

dÕJÕM

suppl. Bad (= ειÄÆ ναι Π = est C = esse V)

2

ˆeNÏÑËF scr. Bad (= διαÁ Π = propterea

ˆˆÒ Φ YÒ  ˆŽkm addidi (= αÆ δυ νατον υë πα ρχειν Π = non est possibile C = non ... possunt V) 5 post KT add. KaA ÒÕtbOÒ  ˆeyi?Ö ÙsfTÖ ÔÙW (with deletion marks s.l.) Φ Bad : om. ΠCV : seclusi, cf. gloss i.m. Φ: \BÒ xsN „ siL эlY ˆlÏytmLA ‰ÑfLi   Ò áÚE cÙY (»these words with [superscript] marks are omitted in Ibn Adı¯’s copy«) qÒd scripsi (= motum aeternum V) : ~qÒd Φ Bad : motum sempiternum esse C LA addidi (= eorum C) : om. vel Ñ= (LA) post ÑKH õM transp. V (cum motu ipsorum) quod C = enim V) :

250b15

250b20

3

VIII 1, 250b15–31

Av. t. 2

4 Cv 67r

8 Cp 172r

ÙqF ÆCÒ  a …ÏhtI ÇÒ  ÑQÖi Ò \M QÖ „ \km ÙQ YÑK YÑËF èÕqI ÑM ‹Y ÑMÏÒË ›eJÖ ÙHÒ  ‹Y ¿yI ;NÏÒË †LD YÕkI YÒ  Í rÖ² úO ~NÑMR Í nKÑS NÑKÖ ~yM ÑelÏK NÑK aÑiWi Ò õi? YÏ ÒË èÕqI ñNÏÑËF %rÕVÑskNÒ  èÕqI ÑM ‹Y ÑMÏÒËÖ ÑE‚iÏB Í KH эF 〈 øyJ 〉 øqyLA YÏ ÒËÖ Iш jB 〈 …htT 〉 ÑeNÏÒ  Ö gXÒª \ksT v 〈 Í M 〉 …htT ÑeNÏÒ  \M (lQdÑbNÒ  Ï Ï Ï \ksTÖ ÙHÒÖ \M ¡oK blvLA lmY ÖÒ  ÙHÒÖ ¡okLA \M bÏdA lmY ÒDÒË :èÑQ ›H †LD ›B .F LA nMRi Ò „ ¡oK \Y Ñ  wnI YÒ  ñNÑ  W ÙHÒÖ YÏ Ò  eJ \M ÑMÏÒ  ¡oK ñnM úyÏwT ÙHÒÖ ~gIÒ  ‡Ñ  tLAÇ ÒDÒËÖ 1 (b23) δηÁ E, fort. Ψ, β δ FHIQw Λ Slc(1123.7.11) : δειÄ J1 : δ’ J2 : δεÁ K (praeter K) Ψ S (1123.7.12, mss. AM) : µηÁ K S (1123.12, ms. F) lc

ter H) Ψ : om. H, fort. S

c

(b23) µηδεÁν Π

2 (b24) τουÄ το Π(prae-

5 (b27) κινειÄσθαι1 Π(praeter E) Ψ Λ T(209.17) : om. E

(b27) πα λιν ηÆ ρεµειÄν E1Ψ K NR γ Λ : πα λιν ηÆ ρεµειÄν καιÁ E2 δ : εÆ ν µε ρει ηÆ ρεµειÄν πα λιν L, fort. S

6 (b28) νειÄκος Π(praeter E1) Pa(358.25) : νικος E1Ψ

7 (b29–30) λε γων ουÏ τως.

Π(praeter E) Ψ : sine interpunct. E : λε γων, ουÏ τως Sc(33.26) HaGa Ross cum Diels (cf. Empedocles frag. 26.10)

8 (b30) ηÎì µεÁν E2Ψ K b2e β FHIJQ Λ : ηÆ µεÁν E1 : ηÎì ... b1 : ηë µεÁν

S (33.26, mss. DEF) : ηà µεÁν w : ηÎ µεÁν Sc(33.26, ms. a) c

9 (b31) ηÆ δεÁ E1Ψ K HI1J Sc(33.27,

mss. DE) : ηÎì δεÁ E2 β δ FI2Q Λ Sc(33.27, ms. a) : ηë δεÁ Sc(33.27, ms. F) : ηà δεÁ w 4 YÏ ÒËÖ scripsi :

YÒÖ Φ : YÒ  Ö Bad øyJ addidi (= εÆ µποιηÄ σαι Π = fecit CV) : om. ex haplogr. 5 Í ÏM add. Bad (= quandoque ... iterum C = quandoque ... quandoque V) ÑeNÏÒ  Ö scripsi (= et quod ipse C) : YÒÖ Φ : шҠ Ö Bad : enim (= ÑeNÏÑËF) V 2…ÏhtT addidi (= κινειÄσθαι Π = mouentur CV) 6 2\M vix legitur : de V : ex C (= εÆ ξ Π) : „ Bad Φ:

õbP〉 Bad

〈

QÖ with corresp. marks s.l. and above the following gloss i.m.: (sic cod.) † YÒÙH øL † cz èÑQ YÕkS ÙyB \M cz YÕkS ÓiH \M èÙÏbtLA „ 〈* * *〉 (Bad YÒÚX{I : 〈YÏ x èÕqI〉 Yz ÙHt legendum ?) rÕtSÙLA „ ÒÚK U; probably an explanatory note on Physics 250b23, tracingÍ back to the ms. referred to as al-dastu¯r by Yahya¯ ibn Adı¯; before or after èÙÏbtLA „ some words must have 1

˙ dropped out; see also above, p. LXIX.

250b25

250b30

4

3

(iLÑPÕpSrt fÏ yibpLA û4LA êÑtK \M nMÑoLA LÑqmLA

Ñ= ~iÏLRÒ  EÙLA YÕkI iÖ ÑeN³Õ° kT YÕkI e`A áÚE \B †LD õM YÑinI iÖ ÙBÒ  YiÙÏbtI \IÚE YÏ Ò  eJ \M ÑMÏÒ  Ö .rÖd KÏhtM ¡V ÙBÒ  fE e`A áÚE \M ÓiH \M cÒ  \IÚE YÏ Ò  eJ \M ÑMÏÒ  Ö ñLÕQ \M ˆefN YÒ  fvbnI ÙQ ñNÏÑËF 2 (a2) ηë E : ηÎì Ψ K (?)e1 β HQ Sc(34.2, 158.11, 160.20, 1125.1) : τηÄì b FIJw Λ

(a2) ουÆ δαµαÁ

E Qw Sc(34.2, ms. F, 160.20, mss. DF, 1125.1, mss. Ma) : ουÆ δ’ αÏ µα Ψ K be1 Λ Sc(34.2, mss. DE, 1125.1, mss. AF) : ουÆ ηÎ δ’ αÏ µα e2 : ουÆ δαµαÄ β FHI2 Sc(34.2, ms. a, 158.11, mss. aE, 160.20, ms. a) : ουÆ δαµον I1 : ουÆ δ’ αë µοιÄ [vel αë µαÄ ?] J, cf. Pp(823.11) : ουÆ δαµηÄ Sc(160.20, ms. E)

3 (a3) δ’ E K b FIJQw Λ Sc(34.3, mss. DE, 158.12, 160.20, 1125.2) : om. e β H

Sc(34.3, mss. aF) : non vert. Ar.

4-p. 5.1 (a4–5) δειÄ γαÁ ρ υë πολαβειÄν λε γειν αυÆ τοÁ ν ηÎì δεÁ τα δ’

εÆ νθε νδε ταÁ αÆ λλα σσοντα E K[αÆ λα σσοντα] δ : δειÄ γαÁ ρ υë πολαβειÄν λε γειν αυÆ τοÁ ν ηÎì ταÂ δ’ εÆ νθε νδε αÆ λλα σσοντα fort. Ψ(ταÁ om. etiam Sl[1123.8, mss. AMF] Sp[1125.5]) : τοÁ γαÁ ρ ηÎì δεÁ (τηÄ δε vel τηÄì δεÁ F2IJQw Λ) ταÂ δ’ αÆ λλα σσοντα (δια[λ]λα σσοντα FIJQw Λ) εÆ νθε νδε εÆ κειÄσε λε γειν αυÆ τοÁ ν υë ποληπτε ον β γ Λ Sl(1123.8, ms. a)

Ñ= ~iÏLRÒ  EÙLA s.p. Φ1 (= ουÍ σϕισιν εÍ µπεδος αιÆ ω ν Π = non est eis aeternitas saecularis tempus perpetuum C = non erit saeculum eis aeternum V) : ÑnL Ñ= s.l. add. Φ2 (caused by misreading Ñ= ÑnL YÒ EÙLA ?) : Ñ=ÑnI YÒ  EÙlL scr. Bad 1

ÑeN³Õ° kT with corresp. marks s.l. and i.m.: aÑiWtÒ YÕÏkT cz Φ i.m. 2 \IÚE with corresp. marks s.l. and i.m.: ŠÒË ¡okLA \M Í ÏMÖ ¤kLA ŠÒË ÙHÒÕLA \M èÒÙtBiÒ Í ÏM ÙBÒ  Y;lÑQ ›inymLA \IÚE YÏ z cz .F YÑK (Bad gXtÒ : Φ gXz) gXz ÑEÙyB (Bad Φ ÙHÒÕLA : scripsi) ÙHÒÖ ŠÒË ÑNzN ÒDÒË ÑMÏ{F ÙHÒÕLA YÕkS ;ŽiB (Bad Φ ÑM : scripsi .F) Φ i.m.; cf. Them. 209.17–20, Simpl. 1125.19–24 4 ñNÏÑËF and ÓiH with the same marks s.l.; this mark again i.m. with: Yz fvbnI ÙQ ñNÏÑËF Φ i.m. (relating ˆefN ... ñNÏÑËF = δειÄ γαÁ ρ υë πολαβειÄν λε γειν to the foll. ... ÓiH \M?) 1

251a

5

VIII 1, 251a1–16

Av. t. 3

3 Av. t. 4 Ca 24ra

6

Cp 172v 9

ñiF èÑaA žiK ˆlyN htÏH †LD „ znN YÒ  úO ÙQÖ .YÒÚE èÙÏbtI fÏ yibpLA ˆlyLA „ Mi Ò qiqH ‹Y ŸÕQÕLA „ ñB õftnI³ ;NÏÒË (iL †LD YÏ ÑËF .èÖÏi Ò zÙbmLA „ znLA ŠÒË IdÏÕ  mLA øb³s³LA „ ~gIÒ  ñB õftnI³ ÙQ øB ×qF ‡ÙÏqT .F ÑEÑncbّ LA aÑiWi Ò \M ÅÖÏÒ  YÕlÙtbM \VÖ [ê205] ;B …ÏhtI YÒ  ñNÑ  W \M ÑM ø³yF fE KaA YÏ ÒË èÕqnF iÏyibpLA rÕMiªÒ „ YÒ  щÕÏQ „ LA rÕMiªÒ ÙJÕT YÒ  Í rÖ² DÒË úJÖ ÙqF …ÏhtI YÒ  ñNÑ  W %³ ÑnLA ÏqI ÙQ ÙIÙhtLA ÒÚE KD ¡V \MÖ . ÑKaA \M Í ÙHÒÖ Í ÙHÒÖ …ÏhtT øÏK † úJÒÖ †LDÖ …ÏhtI YÒ  \km ÑM …ÏhtI YÒ  úO ;NÏÒË ñNÏÑ  B ~yi? øihtsI YÒ  ñN{W \M ÑM øihtsI ;NÏÒË ñNÏÒ  †LD èÑoM ÑKaA \M ÙHÒÖ YÕkI YÒ  DÒË Í rÖ² úO ÙqF \KÑMi Ò èÙÏbI YÒ  ñNÑ  W \M ÑM øqtnIÖ ñNÑ  W aLA YÕkI YÒ  Ö ƒ£U YÒ  øbQ \M iÍ ÖÏÒ  ƒÒ£HiÒ ñNÑ  W aLA 1 (a5) δεÁ α β T(209.22) Sl(1125.25, ms. F) : δηÁ K γ δ Λ Sl(mss. aAM) Pa(359.8) του των E β δ Sl(1125.25, ms. a) Pa(359.8) : του του K γ Λ Sl(mss. AFM), fort. Ψ τηÁ ν2 α K β δ H T(210.1) Sl(1125.25) : om. FIJQw

(a5) 3 (a7)

5 (a9) εÆ νε ργειαν α K β Sp(1126.12); cf.

As(frag. 544) : εÆ ντελε χειαν γ δ Λ T(210.3) Pa(359.16–17)

6 (a11) ταÁ Π(praeter K) Ψ

Sc(1127.1) : om. K T(210.5)

10 (a14–15) τοÁ καταÁ το πον α K be1 β F2IJQw Λ T(210.8) :

καταÁ το πον τοÁ e2 FH, fort. S

(a15) ωÏ στε δειÄ Ψ K δ FHIQw Λ : ωÏ στ’ ειÍδει L : ωÏ στε δηÁ E J;

cf. T(210.8)

KD (= rememoratione C) : om. ΠV, an delendum? Ö …ÏhtI YÒ  sed del. Φ : om. CV 7

9 post

ÑKaA iter. ... KD ¡V \MÖ

†LD „ with corresp. marks s.l. and i.m.: (Bad ñMÙY : s.p. Φ) OÙL Öz iÏLRz KaA YÕK „ cz i.m. Φ; cf. Philop. 882.11–16 3 gloss without mark i.m.: yibpLA YÏ i   fÏ yibpLA ˆÂlyLA ~gIÒ  õfnIÖ fÏ yibpLA MtÒ „ zN KaA „ znLÒÖ KH zÙbM Φ i.m., fort. ex comm. Ibn al-Tayyib 808.14–16 ˙ 1

251a5

251a10

251a15

(iLÑPÕpSrt fÏ yibpLA û4LA êÑtK \M nMÑoLA LÑqmLA

6

Av. t. 5

Av. t. 6 4

Av. t. 7 Cp 173r 8

YÕkT YÒ  ÑMÏÒË Í rÖ² úO ~gIÒ  rÕMiªÒ áÚeF .ƒhI³ YÒ  øbQ \M ƒÒHiË Ò ÒÚkE NÑK ÙQ YÕkT YÒ  ¡V \M ÑQÖi Ò \M  QÖ „ ÒÚkE ¡cT ;NÏÒË ÑKÏhtmLA \M ÙHÒÖ øÏK 〈 YÑK 〉 YÊ ÑËF .ÒÚkE Í iÏLRÒ  YÕkT YÒ  ÑMÏÒËÖ ÙQ cÚLA iÏvtLA øbÊQ \M YÕkI YÒ  Í rÖ² úO ÙqF ~KÏhtM ¡cI ;NÏÒË .…® U YÒ  ÖÒ  …ÏhtI YÒ  ~nkJ rÑC ñB cÚLA ÕEÖ KHÖ X| i° vT ÔÙH ÒÚE YÏ ÑËF KH \kT G ñNÏÒ  ¡V fcLA áچ è‚T G ,M .F NÑK YÒËÖ \M †LD ezI ÙQÖ .ñL ñJÖ i ñNÏÒ  ñsfnB ñnM eZ øM® Õª TÖ B® ÙT³ ÒDÒË èÕqLA ñNÏÒ  †LDÖ .ùitfT øgF ù  t® F³ ÒDÒË ¤KÒ  ñiF YÑnwLA ‡‚lTÖ ›Bz rÕeZ áMÒ  …® W YÒ  шѠ W LA aÑiWi ÒÖ …ÏhtT YÒ  ш{W LA aÑiWi Ò NÑK YÒË ~niHÖ …ÏhtmLA ÒÚEÖ èÂÖÏi Ò …ÏdA ÒÚE YÕkI ÑM ~niH YÑK vÏ Í dÕJÕM 1 (a16) ταυÄ τα] + ταÁ πρα γµατα (?)Ψ; cf. Pa(360.7)

3 (a18) κινητωÄ ν α K2 FHI1JQw Λ :

Ä ν K1 κινητω Ä ν καιÁ κινητικω Ä ν β δ I S (vel inv., 1127.22) : κινητικω

4 (a18) αÆ ναγκαιÄον E2 K

be1 β FIJQw Λ : αÆ να γκη e2 H : αÆ ναγκαιÄον vel αÆ να γκη Ψ : om. E1

5 (a20) τοÁ ] om. Ψ, fort.

2

om. Sp(1128.14–15)

p

8 (a22) προα γουσι(ν) α K1 T(211.2) : προιÈουÄ σι K2 β γ δ Λ Sp(1128.2)

10 (a24) οë τεÁ Π(praeter E)Ψ : οÏτι E

(a24) τι] τοÁ F1, fort. Ψ(»when this is a first mover and

that is a [first] moved« intell. Ar., reflecting τοÁ µεÁν ... τοÁ δεÁ ?)

YÑK〉 YÊ ÑËF scripsi (= ειÆ µεÁν τοι νυν Π = nam si C = si igitur V) : YÑF Φ : YÑËF (fa-inna) Bad 4 øbÊQ \M (= προ τερον Π = ante C)] : øÂbQ \M Bad 5 rÑC scr. Bad : rÑC Φ …® U scr.  Bad : …hbB Φ1 : …W Φ2

3

〈

áÚeF with corresp. marks s.l. and i.m.: …ÏhtmLÒÖ …ÏdA yI Φ i.m.; cf. Simpl. 1127.18–21 2 ÒÚkE ¡cT with corresp. marks s.l. and i.m.: …ÏhtLÒÖ †IhtLA ш{W \M YÕkI cz Φ i.m. 10-p. 7.1 gloss without mark i.m.: øyfLÑB i ÕÏqLÑB õGÕmLA ÒÚE „ ;emefT Yz fvbnI …ÏhtmLÒÖ …ÏdA Φ i.m., 1

fort. ex comm. Ibn al-Tayyib 809.1–8 ˙

251a20

251a25

7

VIII 1, 251a16–32

Cv 67v 3

Av. t. 8 6

YÕkI YÒ  úO ÙqF YÕkS YÕkI øB \IMi Ò \M ÙHÒÖ iÖ YÕkI i ÑM †LDÖ ÑM úÈbS YÕkslL YÑK ÙQ ñNÏÒ  †LDÖ i° vT †LD øbÊQ \M èÑaA áچ i° vT èÖÏi Ò iÏvtLA øbQ YÕkI YÒ  †LÚL újiF .KaA ‡ÙY ÕE YÕksLA YÏ Ò  .ñMÙÏqtI ›TdÏÑgtmLA ›tKaA …ÏU љyBÖ dfM …ÏU ;NÏÒË aÑiWi Ò ¡yB YÏ ÑËF ÑeNÏÒ  YÈ ÕnzB FymLA ÑMÏÒ  Ö d Ï T iÖ \xÏsT ÑeNÏÑËF rÑnLA ÑMÏÒ  †LD èÑoM .~yi? 〈 a ~gIÒ  …ÑnE YÕkI YÒ  ezI ÙQ ñNÏÒ  ÇÒË ÙHÒÖ 〉 a \IÙÏglL êÖ³LA \M ê³B õJr ÒDÒË \xÏsI ÙQ d LA YÏ Ò  †LDÖ ê³LA ÒÚE ñbwI 1 (a25) ηÆ ρεµειÄ E δ FHJQw Λ : ηÆ ρεµειÄν Ψ K β I

(a25) αÆ ναγκαιÄον E δ Sp(1128.4) : δειÄ

HIJQw Λ T(211.5) (?)Pa(360.20) : αÆ ναγκαιÄον δηÁ β (?)Ψ : αÆ ναγκαιÄον δειÄ K : αÆ ειÄ F ηË ν Π(praeter E1)Ψ Sp(1128.4) : om. E1

Pa(360.25) : πρω της E1

7 (a31) τι ειËναι E2, fort. Ψ, K β γ Λ : om. E1 : ειËναι τι δ

YÕkT Bad 2 áچ scripsi (= sit in hac dispositione C = in hac dispositione ... sit V) : áÚE Φ Bad 5 1…ÏU scripsi (= κινειÄ Π = moueant [-ent Cp] C = mouent V) : …htI Φ Bad 7 a ... ÙHÒÖ addidi (= µι α ϕαι νεται µεÁν ουË ν καÆ κειÄ τι ειËναι [τι

1

YÕkI scripsi

2 (a26)

3 (a27) προÁ τηÄ ς Π(praeter E1)Ψ Sl(1128.26)

: s.p. Φ :

ειËναι om. E1 : ειËναι τι δ] Π = una verumptamen apparet illic etiam aliquod C = eadem sed

apparet etiam illic aliquod V) : om. hom. Φ Bad 236f., Ullmann NE I, 218) :

ê³L Φ Bad

8 ê³B scripsi (cf. WGAÜ Suppl. II,

FymLA with corresp. marks s.l. and i.m.: ˆlyLA ‹Y (fnLA „ dÕJÕmLA ÕÏqLA yI Φ i.m. 7 \IÙÏglL with corresp. marks s.l. and i.m.: XvÒÖ ¿yLÑB ;EÙHÒ  \kL \IÙÏgLA ÕÏqLA øyfT ÙQ cz (Bad ñlyfI : s.p. Φ) ñlyfT ÒÚLÑB Φ i.m. 8 êÖ³LA] + ¿yLÑB cz s.l. Φ; cf. Simpl. 1138.3ff.

6

251a30

(iLÑPÕpSrt fÏ yibpLA û4LA êÑtK \M nMÑoLA LÑqmLA

8

Cp 173v 4

Av. t. 9

\kL .(kyLÑB ñmÂlY ømytS ÒDÒË ~YÕP hªp[ ÙQ GÑyLA YÏ Ò  ;K ‡±Ï TÖ feF …ÏhtT ÖÒ  …ÏhÂT³ ÖÒ  øyfnT ÖÒ  øyfT YÒ  \km LA ÑelÏK aÑiWi Ò (iL  NÑK ÒDÒË øB †LD ‹Y [Ò 206] LÑL i rdÑQ êrÑQ ÒDÒËÖ ÒÚK èÑT ÒDÒËÖ ÒÚE …ÏWÖ ÒÚE …ÏH BrÑqT ÒDÒË YÕkI htÏH ~gyB љyB .…ÏhtLA ñNÑ  W ÒÚEÖ †IhtLA ñNÑ  W ÒÚE YÕkI эF LA èÑaÑB rÑC èÑaA †ltB \kT G ÑeNÏÒ  \iÏbF KÏhtM ~qÒd NÑK ÙQ DÒË \kI G YÑËF iÏvT ÙQ øB ~KÏL ÒÚEÖ ~KÏhtM ÒÚE YÕkI YÒ  ÑeyM \km YÑK LA 2 (b1) καιÁ γ δ Λ : ηà α K β

3 (b2) αÆ λλ’ Π(praeter E2)Ψ : ras. E2

Π(praeter E2)Ψ : ωë ς διε χοντα E2

(b2) ωë διÁ εÍ χοντα

4 (b3) πλησια σηì E, fort. Ψ, K be1 β : πλησια ζηì e2 γ

(b3) κινειÄ E K β γ Λ : τοÁ µεÁν κινειÄ Ψ δ; cf. T(211.13) Sp(1129.17–18)

5 (b4) ωë ς ηË ν E, fort.

Ψ, w : ωë ς ειËναι K e β FHIQ Λ Pa(361.24) : ωë ς ειÍη (?)J : ωë ς... [ras.] J1 : ωÏ στε ηË ν be1 w2 1

2

2

6 (b5) εÆ κινειÄτο Π(praeter E1)Ψ : εÆ κειτο E1

7 (b6) ηË ν α K δ J1 : om. β HJ2Q Λ : µηÁ FIw

(b7) µετε βαλεν α T(211.14, mss. C Laur.85.14) : µετε βαλλεν K T(211.14, mss. MSL) : εÍ δει/εÆ δειÄ µεταβα λλειν β γ δ Λ (cf. App. 1 ad 251b9)

3 ÒDÒËÖ Φ (= καιÁ Π 251b3) :

Φ:

Bad

ÒDÒË ÖÒ  Bad : quando Cp : aut quando CaCvV

4 YÕkI scripsi : s.p.

YÕkT Bad …ÏWÖ Φ (= τοÁ δεÁ κινειÄται Π = et mouetur C = et ... mouebitur V) : …W ÖÒ  5 …ÏhtLA ñNÑ  W ÒÚEÖ om. hom. in textu, suppl. i.m. Φ, with signs ind. its omission in the

main text (= τοÁ δεÁ κινητο ν Π[sed κινητικοÁ ν ... κινητο ν inv. F] = et hoc innatum moueri V : est hoc aptitudo moueri [et hoc aptitudo est mouere] inv. C)

6 post

ÑeNÏÒ  add. YÒË Φ Bad :

seclusi (= ωë ς ουÆ χ ... ειËχον Π = manifestum est quod ipse [sic] non sunt C = manifestum est quod non erant V)

KaA ‡ÙÏqtI Yz êrÑqmLA dÕJÖ \M ‡‚lI, gloss 2: ‡‚lI FÑGiË Ò „ i YÑkmLA „ XvÒ ŠÒË ;EÙHÒ  …ÏW ÒDÒË bsnLA áچ XvÒ õM ;EÙHÒ  YÕkI Yz Φ i.m. 4–5 two glosses without mark i.m., gloss 1:

251b

251b5

9

VIII 1, 251a32-b18

4 Av. t. 10

8 Cp 174r

Ca 24rb

„ øXÒd ÕE .F ~MRi ÒÚE YÕkI YÒ  Í rÖ² úO ÙQ ñNÏÒ  †LDÖ ;EÙHÒ  ñ³f³yÊG YvÒ ÕEÖ ÒÚE žÂyÊG \kI G YÒË ÒÚE YÏ Ò  †LD èÑoM .ŸÑgmLA êÑB  YÒ  †LÚL újiF ;EÙHÑ  F ;EjK \kI G YÒË iÏvT ÙQ YÕkI YÒ  úJÒÕF [7Ò206] .ñlbQ iÏvT èÂÖÏi Ò ÂiÏvtLA ‡ÙÏqT ÙQ YÕkI ‡ÙÏqtmLA YÕkI cyW iL ñJÖ cÏ Ò  ‹Y †LD õM vÏ [21ê206] \kT G ÒDÒË 〈 YÑM‚LA 〉 YÕkI žiKÖ [Ò 207] YÑMR \kI G ÒDÒË XÏÑ  tmLÒÖ ÙM¬ YÑK YÒË YÑM‚LA YÏ ÑËF ÑM Í KH ÖÒ  KH dÂÙY YÑM‚LA YÑK YÑËF ?KH ÙHÒÖ jX ÑM õim`A ÙP ÑnÏkL .iÏLRÒ  Í KH YÕkT YÒ  Í rÖ² úJÒÕF †LÚBÖ .YÕÏkM ¡V ñNÏÒË YÕLÕqI ˆeNÏÒ  †LDÖ YÑM‚LA MÒ  \M gÒ  LA ft¬ M³ YÑM‚LA YÏ i  èÑQ NÕÏkM ÑelÏK ÒÑiWi Ò YÕkT YÒ  \km i ñNÏÒ  (pIqmd \iÏbI õM YÕÏkT ñNÏÒË èÕqI ñNÏÒ  †LDÖ ñNÕÏkI áÙHÖ ñNÏÑËF \PjFÒ  ÑMÏÑ  F .YÕÏkM ¡V 3 (b9) µεταβα λλει E2 K β H : µεταβα λλειν E1 δ F1IJQw Λ : εÍ δει µεταβα λλειν fort. Ψ (cf. App. 1 ad 251b7) : µεταβαλειÄν F2

(b9) ειÆ α K β FHJQ1w Λ : καιÁ ειÆ IQ2, fort. Pa(362.4) : ηà ειÆ δ

6 (b11) ηà K β : ηà οë δ FHI2JQw : ηà vel ηà οë Ψ : ηÎì E I1 του του α β HJQ1w : τουÄ το K δ FIQ2 Λ

8 (b13) καιÁ ] non vert. Ar.

9 (b16)

10 (b16) αÏ παντα Π(praeter E1)Ψ : αÏ παν E1

11 (b17) αÆ γε ν(ν)ητον ειËναι α K β δ Sl(1152.21) : αÆ δυ νατον γεγονε ναι γ Λ Sp(1155.35) : δ’ (vel δεÁ) αυÆ τοÁ ν Ψ β γ Λ; cf. T(211.28)

(b17) δεÁ E K δ

(b18) γεγονε ναι Π(praeter Q1)Ψ

Sc(1155.36) : om. Q1 : secl. Ross

vÏ add.(iLÑPÕpSrÒ  èÑQ ÑoLA ylytLA Φ Bad : om. ΠCV : seclusi 6 YÑM‚LA addidi (= 8 ÙHÒÖ scr. Bad : ÙHÒÖ Φ + \PjFÒ s.l. (cf. Sl 1153.19f.) 9 Mz \M Φ (= de CV) : Mz „ Bad (for Mz \M = περι c. gen. cf. GALex I, amrun 27) 10 (pIqmd scr. Bad : (ipIqmd Φ 11 YÕÏkT] s.p. Φ : est factum C : generari V

5 ante

χρο νος Π = tempus CV)

9 YÑM‚LA with mark s.l. and note without mark : (vel Ôd)Ôr†..†  Y vid. Φ i.m. (= ÔÙL ¡V ?); cf. transl. Scotus »ipsum (scil. tempus) esse non generatum, idest non novum«

251b10

251b15

(iLÑPÕpSrt fÏ yibpLA û4LA êÑtK \M nMÑoLA LÑqmLA

10

Av. t. 11

4

8 Av. t. 12 Cp 174v

YÒ  ÖÒ  YÑM‚LA YÕkI YÒ  \km i YÑK YÒË ~gIÒ  Ö .NÕÏkM a4LA YÏ ÒËÖ a4LA zÙbM \kL aÑgqNÖ zÙbM ñiFÖ ÑM Í pSÒÖ YvÒ YÑKÖ YvÒ \M ÕlX øÂqÂyI³ YÕkI YÒ  Í rÖ² úJÒÕF ;LA YÑM‚lL 〈 aÑgqNÖ 〉 øbqtsmLA YÑM‚lL ÚÂXÒª ÒDÒË YÈ ÑMR áÙyB (iL cÚLA ¡Xi Ò YÑM‚LA ÂX| YÏ Ò  †LDÖ .ÙM¬ YÑM‚LA  YÑM‚LA „ ÚXqI YÒ  \km i ñNÏÑËF ÑNvÒ Â ¡yB „ YÕkI YÒ  úJÒÕF a YÕkI YÒ  Í rÖ² úJÒÕF ÐÑgqNÖ Ò¡ ÙbM YvÒ YÑK DÑËF .YvÒ gÕS jÍ CÒ  YÒ  Í rÖ² úJÒÖ ñNÏÒ  EÑzF YÈ ÑMR ñitbnJ \Y YÑK YÒËÖ .YÑMR ñitbnJ \Y .KaA ¿ÒYÒ  \M ÑM ¿ È Y ÕE ;NÏÒË YÑM‚LA YÑK DÒË KH ~gIÒ  YÕkT ñNÏÒ  †LDÖ Í ÙSÑF siL KaA 〈 YÏ Ò  〉 „ èÑqI ñniyB èÕqLA ÒÚEÖ èÖÏi Ò iÏvtlL ~MÙÏqtM ÑM Èi° vT YÕkI YÒ  NÕÏktM KaA YÏ Ñ  B èÂÕqLA ‡‚L ;K 1 (b19) καιÁ α K β δ T(212.2) Sl(1156.4, ms. a) : om. γ Sl(mss. AFM) Sp(1166.33) 1

καιÁ α β δ T(212.2) : om. γ

(b21) αÏ µα K β γ T(212.3) : αÆ λλ’ E Ψ : αÏ µα αÆ λλ’ E2 δ

6 (b25) τοÁ 1 Π(praeter E)Ψ Sp(1166.34) : τοÁ ν E

(b25) τε] om. L FHw : non vert. Ar.

7 (b26) αÆ µϕο τερα E, fort. Ψ, δ R γ Λ : αÆ µϕο τερον K LN T(212.6) Pa(362.26)

2 (b21)

1

(b26) αÆ ειÁ γ Λ : om. α K β δ

9 (b29) γαÁ ρ α K β δ HQ Λ T(212.11) : om. FIJQ1w 2

κι νησιν Ψ K β γ δ : καιÁ κι νησιν E : και νησιν (?)E 2

10 (b30)

1

1 a4LA YÏ ÒËÖ scr. Bad (= τοÁ ν δ’ ουÆ ρανοÁ ν Π[om. Q1] = et quod coelum CV) :

YÑM‚LA YÑF Φ 2 øÂqÂyI³ scripsi (= νοηÄ σαι Π = intelligatur CV) : øyqB Φ + ñybB(?) with correction mark ‰ s.l. : øyfI Bad 3 aÑgqNÖ (= τελευτηÁ ν δεÁ Π = et finis CV) suppl. Bad 4 ÂX| scripsi (= τοÁ ... εÍ σχατον Â Π = quando accipitur C Π = ultimum CV) : ҂JÒ Φ : Ï҂JÒ  Bad ÚÂXÒª ÒDÒË scripsi (= ληϕθε ντος  663.3, 238a22 = 696.6) : ÔÙHÒdÒ Φ : ÔÙH ÒDÒË = si fuerit acceptum V, cf. also 235a33 = Bad Bad 5 ÚXqI scripsi (= λαβειÄν E K β δ H [λαµβα νειν FIJQw] = accipiatur C) : ÙHÖB Φ : ÙJÕI Bad (= sit V) 6 2YvÒ Φ : om. Bad 9 YÏ Ò  addidi (= in hoc quod CV) 10-p. 11.1 ... èÖÏi  Ò ¡Xi Ò om. in textu, suppl. i.m. with corresp. marks after iÏvtlL and i.m. Φ : habent CV

251b20

251b25

251b30

VIII 1, 251b19 – 252a6

3 Cv 68r Av. t. 13 6

9 Av. t. 14 Cp 175r

11

aLA ž° K (iL ñNÏÒ  †LDÖ ¡Xi Ò iÏvtLA ÙyB iÏvT YÕkI YÒ  ÑnEÑE ‡‚lI †LÚK èÑoM .~yM YÑNÕkI …ÏhtLA ñNÑ  W YÕkI YÒ  \Y ñf° KÖ ~KÏhtM YÕkI YÒ  \Y ñNÑ  W YÕkI YÒ  \km ÙQ ñNÏi  ƒÒ£HiÒ ñNÑ  W YÕkI YÒ  Ö ƒ£U YÕkI YÒ  †LD ñNÑ  W YÕkI YÒ  \Y aLA žÏK iÖ ƒ£U YÕkI YÒ  ¡V \M ƒÒ£HiÒ YÕkI YÒ  úO ÙQ ~gIÒ  Ö .~yM YÑNÕkI …ÏU YÕkI YÒ  \Y ñfÏKÖ †IhtLA ~gIÒ  Ùs³fI ÒÚ= ÙsfmLA YÕkI YÒ  Ö ÙsF〈 z 〉 ÒDÒË Ùs³fI ÒÚ= ÙsfmLA YÏ Ò  \iÏbLA \B ÆihtsM †LD YÑK DÑËF .ÑM iÏvT dÑsfLA YÏ ÑËF ñL ádÑsFÒÂË ÙyB †LÚB èÑQ \M YÏ ÑËF ~niH \kI GÖ ~niH YÑK ÑJÏ siLÖ iÏLRÒ  KH ÑnEÑE .ñbWÒ   IxtLÑB ñLÕqF YÏ Ò  \¬Z \M \°ZÖ cQ YÒ  rÕMiªÒ õbpL ÒÚkE YÏ ÒË èÑQ \M èÕQ †LÚKÖ 2 (b31) αÏ µα Π(praeter E1)Ψ : αÏ µα αÆ λλαÁ E1

3 (b33) οÍ ν α γ δ N Λ Sp(1170.10) : om. K LR

6 (a1) ϕθαρτικοÁ ν α K b L Sc(1170.26), fort. Philop. (Rescigno, p. 92.12–13) : ϕθαρτοÁ ν e NR

ÙSÑF, v. 192b1 = Bad 76.1, 201a14 = 171.6, 221b28 = 458.2, 261a9) (a1) δηÁ E δ F IJQw Λ : δεÁ fort. Ψ, K β F1H Sc(1170.26) (for καιÁ ... δηÁ Ar. has usually ... Ÿ/Ö ~gIÒ  , v. 206b16 = Bad 253.7, 221a30 = 450.11, 224a12 = 481.14, or ;NÏÒË ... Ö 212b30 = 332.4)

γ Λ (for ϕθαρτο ς = 2

(a2) ϕθει ρηì E1ΨΩ K LN, fort. Sp(1170.30–32) : ϕθαρηì E2 R : ϕθει ρηται γ δ Λ (v. also notes Hayduck, Wag and Ross)

7 (a3) τουÄ τ’ αÆ δυ νατον fort. Ψ, K γ Λ : ταυÄ τ’/ταυÄ τα αÆ δυ νατα E β

l

δ S (1170.21) Pa(363.19)

2 ~yM (= αÏ µα Π[praeter E1] = simul C = insimul V) Φ :

ÙHÒÖ Gˇp(157.12) 5 …ÏU scripsi (= κινουÄ ν Π = mou¯ C = mouere V) : s.p. Φ : moueatur C Cp : …W Bad 6 ÙsFz ÒDÒË scripsi (= 1 οÏταν ϕθει ρηì E KLN = quando corrumpet V) : ÙsF ÒdÒ Φ : ÙsF ÒDÒË Bad (= quando corrumpitur C) 7 DÑËF Φ (= cum igitur C) : et si (= YÑËF) V †LD scr. Bad (= illud C = hoc V) : †LÚK Φ 9  IxtLÑB scripsi (= πλα σµατι Π [praeter L] = stulto V, v. Glossary s.r. H-R-S) :  BhbLÑB ˘ ˙ Φ : žIxtLÑB Bad : figm ¯¯ to vel sigm ¯¯ to (i.e. sigmento ?) C 10 õbpL Φ : õ³bP Bad v

10

a

YÏ ÒË èÑQ with corresp. marks s.l. and i.m.: (lQdÑbNÒ  yI Φ i.m.

252a

252a5

(iLÑPÕpSrt fÏ yibpLA û4LA êÑtK \M nMÑoLA LÑqmLA

12

Av. t. 15 4

YÏ Ò  \M (lQdÑbNz ñbwI ÑM ‹Y ñB èÕqI cÚLA 〈 ÒÚE 〉 ÕE zÙbmLA rÕMjªL YÒdÕJÕM ;ekÂIWÖ Í ÏM áÚEÖ Í ÏM áÚE blvLÒÖ bÏdA ÏjitS Ò  ÙbmLA ÒÕlyJ \IÚLA ~gIÒ  øÏyLÖ .YÑM‚LA \M †LD ›B .F YÕksLÒÖ Í rÖ² Ó (iL ñNÏÒ  ¡V jÏaA áÚE øol YÕjÏtU %rÕVÑskNz øoM ÙHÒÖ 207] YÏ Ò  †LDÖ .ñL ‡ÑzN i fÏ yibpLA gcA cOÖ õbpLÑB ÕE ÑJÏ ÆCÒ  øyF \M (iLÖ *** .ñL fE ÑM øÏK „ ‡ÑznLA úbS fE yibpLA [ê vÏ Iш jB ~NÑMR Í nKÑS NÑK aÑiWi Ò YÕkT YÒ  áÕJÕLA \M ñJÕB yibpLA 1 (a6) ταυ την δειÄ K δ HIJQw Λ : ταυ την δειÄν F : δειÄ ταυ την E β : ταυ την fort. ΨΩ, (a7) οÏπερ Π(praeter E1) : ωÏ σπερ E1 Ψ Λ (for ÑM ‹Y = ωÏ σπερ

Sp(1183.23), fort. Pa(363.23)

cf. 184b20 = Bad 5.13, 189b11,18 = 54.7, 55.11; Glossary; Gutas Theophr. 459b) νειÄκος] νιÄκος Ψ (cf. App. 1 250b28) 2

K δF hom. Ψ

2 (a8)

(a8) υë πα ρχει α K β HIJQw Λ S (1183.26) : υë πα ρχειν 1

4 (a10) ωÏ σπερ α K be β Λ : ωÏ σπερ καιÁ e2 γ 1

p

6 (a13–14) προÁ ς ... τοÁ δ’ αÍ πειρον] om.

7 (a16) ουÆ κε τι E β γ δ : ουÆ κ εÍ στι K S (1184.31) Pa(364.18) : an ουÆ δε ν τι vel p

ουÆ δαµω Ä ς Ψ? (»not at all« Ar.)

ÒÚE addidi (= ταυ την Π = est illud quod dicitur C = est illud de quo diximus... V) èÕqI (scil. (lQdÑbNÒ  ) scripsi : s.p. Φ : dicitur C : diximus (= èÕqN) V ñbwI scripsi (= εÍ οικεν Π = secundum quod videtur C = secundum ... similitudinem V) : s.p. Φ : ñbsN Bad 1–2 YÏ Ò  ÏjitS scripsi (= ωë ς τοÁ κρατειÄν Π = quod dominium C [V]) : jitSiÒ Φ with the sign C above the La¯m of the article : ÍjitSÒ Bad 6 ***] Φ = CV cf. App. 1 7 NÑK Φ : omnes (sint) C = omnia (sint) V (= ÑelÏK ?, nisi NÑK aÑiWi   Ò øÏK)

1

4 ÙHÒÖ with corresp. marks s.l. and i.m.: øqyLA yI Φ i.m.; cf. Alex. frag. 550 (v. p. LXXIII), Simpl. 1184.19f.

5 gloss without mark next to ñL ‡ÑzN i i.m.: FցyM lcÏ L lÏY ñL YÕkT i cz Φ

i.m.; cf. Simpl. 1184.22f.

252a10

252a15

13

VIII 1, 252a7–25

Cp 175v 4

8

YÑ  B gHÒ  YvÒ KaA YÏ Ò  úÂJÂÖÒ  ÆCÒ  †LD „ ƒF i vÏ ~niH KÏW ÑM YÏ Ò  †LDÖ .ÆCÒ  ‡ÈÑzN ~gIÒ  †LD „ YÕkI iÖ ‡ÙÏqT .F юM YÕkT øoM ÒÚK ~niHÖ ÒÚK ~niH YÕkI i htÏH ~JDÑS YÕkI YÒ  ÑMÏÑËF yibpLÑB cO ÑMÏÒËÖ i ~niHÖ ÕmsT ~niH YÕkT Yz ¡V \M õbpLÑB ƒÕF ŠÒË ÕmsT rÑnLA YÏ Ò  ÖÒ  (lQdÑbNz èÕQ YÏ ÑËF †LÚLÖ .ÑM bsnL ~zFÑH YÑK ~JDÑS \kI G YÒË ÒÚE YÏ ÑËF …ÏhtI vÏ Í M \ksI øÏkLA YÏ Ò  YÒ  d³ÕJÒ  ñLÕQ øoM èÑQ \J¬ á¡V YÑK ÙQ ñNÏÒ  ¡V ÑM ~MÑzN fcLA áچ YÑK ÒDÒË èÑH ‹Y ñL úJÕI èÕqLA Í Òچ ølÑqLA ‹Y úO øB ×qF rÑbXÒË ñLÂÕqI YÒ  ‹Y ±tqI i YÒ  èÕqLA øB jÏH ¡V \M ÆCÒ  igqB qI iÖ ~yGÖ õgI iÖ ñbÂbÂS ñyM ÂKÚI .YÑE B ÑMÏÒËÖ ÎҁqtS€B ñiF Ñ  I YÒ  ÑMÏÒË 1 (a15) δηÁ α (cf. Denniston 237) : δεÁ K β γ δ T(212.30) Sp(1184.29)

Λ S (1184.32) : εÍ χειν δειÄ E δ T(212.31) p

2

Sl(1183.20)

3 (a17) εÍ χει E1Ψ K β γ

5 (a19) εÍ χει E Ψ K β γ Λ Pa(364.20) : εÍ χειν E2 δ 1

7 (a22) τιν’ (vel τιναÁ ) Π(praeter E1)Ψ : τηÁ ν E1

8 (a23) ϕα ναι α (for

αÆ ποϕαι νω Isha¯q uses the root H-K-M, v. De Int. 17a19ff, Phys. 209b16 = Bad 286.1) K1 be1 : ˙ ˙ αÆ ποϕα ναι K2 e2 L Hw : αÆ ποϕαÄ ναι NR FIJQ Λ 9 (a24) µηδ’ (vel µηδεÁ ) Π(praeter E1)Ψ : ηë δ’ E1

9–10 (a24–25) αÆ λλ’ (vel αÆ λλαÁ ) ηà εÆ παγωγηÁ ν ηà α β δ FHJQw Λ : αÆ λλαÁ ειÆ εÆ παγωγηÁ ν ηÃ

I : αÆ λλαÁ εÆ παγωγικηÁ ν (vel εÆ πα γουσαν) fort. ξ (v. App. 2 ad loc.) 1 YÑ  B Φ : Yz Bad 4 post i add. ÕmsT Φ Bad : om. ΠCV : seclusi, cf. gloss i.m. Φ: „ (iL õGÕmLA ÒÚE „ ÕmsT zfL U xsN (»at this place, the word tasmu¯ is missing in Yahya¯’s ˙ copy«) 5 †LÚLÖ scr. Bad : †LDÖ Φ 8 i YÒ  Φ : iÒ  Bad 9 õgI s.l. Φ with the sign ° (= non ponit positione C [V]) : ñygI in textu Φ Bad 9–10 YÑE B ... ÑMÏÒË øB Φ in textu (= Π[v. App. 1] CV) : ñiLAË ÙlÑQ YÑE B Ñ  I øB (= αÆ λλαÁ εÆ παγωγικηÁ ν [vel εÆ πα γουσαν] αÆ πο δειξιν ϕε ρειν ?) preceded by ° „ (»in ms./version H«) with corresp. marks i.m. and above YÑE B Φ i.m. ˙

2

ÆCÒ  with corresp. marks s.l. and i.m.: (Φ árbBÖ : sic?) ¡TÖ ‹Y cz Φ i.m.

252a20

252a25

(iLÑPÕpSrt fÏ yibpLA û4LA êÑtK \M nMÑoLA LÑqmLA

14

3 Cp 176r

6 Ca 24va

YÏ Ò  YÒ  ~gIÒ  ÒÚE iÖ ~BÑbSÒ  siL yGÖ LA aÑiWi Ò áÚE YÏ ÑËF YÑK YÑËF .ƒÏfT áÚEÖ õmQ †ltF blvLÑB ÖÒ  bÏdÑB YÕkT ;NÏÒË rÕMiªÒ lÑS ›miwLA KÚI YÒ  úO ÙqF ›fltM ›miW „ ›lcÏL YҁMi Ò YÒÚE bÏdA ÕE 〈 ˆ= 〉 õMÑ`A YÏ Ò  %ÑnLA „ øcÏU ;K ;F ;E \IÚlLA øÏkLA „ YÏ Ò  õGÂÕI³ ñNÏÑËF .¡yB \M ˆ™yB ꁇ \IÚLA ˆE aÒÙYi Ò YÏ Ò  Ö ¡yB „ dÕJÕM MÒ  ÒÚE YÏ Ò  †LDÖ ÒÕLÑQ .gcA ÒÚE Mi Ò cO ~gIÒ  jÏH ŠÒË ÑtU ~gIÒ  Î|ÕSÊ〈 z 〉 YÑMRz „ †LD dÕJÖ YÏ Ò  ÇÒË aÑiWi Ò Í .ñhhÏcT 1 (a25) αυÆ ταÁ ] fort. ταυÄ τα Ψ

(a25) γαÁ ρ Π(praeter E1)Ψ : om. E1

(a25) ουÆ κ αιÍτια ταÁ

υë ποτεθε ντα Ψ K β δ FHIJ w : ουÆ κ αιÍτια υë ποτεθε ντα E Q : ουÆ κε τι αÆ ποτεθε ντα J1 2

(a26)

τουÄ τ’ β : ταυÆ τ’ E : ταυÆ τοÁ E : τουÄ το K γ δ Λ : ταυÆ τοÁ ν S (1185.3) : ταυÆ το , τουÄ τ’ fort. Ψ (»nor 1

2

p

is this [the cause], I mean [the fact] that...« Ar.) αÆ λλαÁ ] ειËναι ταË λλα fort. Ψ

2 (a26) νει κει] νικει fort. Ψ

(a26) ειËναι,

3 (a28) τοÁ Π(praeter w ) : om. w : ταυÄ τα fort. Ψ (»if these two 1

1

issues are realised/determined in two different things«, Ar.); cf. Pa(365.3) δ NR γ Λ Pa(365.4) : τι E : om. L

(a28) τι] non vert. Ar.

4 (a28) οÏτι Ψ K

6 (a31) ουÏ τως] an ουÎ τος Ψ ?

7 (a31) δι’ Π(praeter E1)Ψ : om. E1 3 ›miW „ scripsi (= εÆ ν µε ρει Π = in duobus rebus C = in aliquibus duobus V) :

šbbS „ Φ : ›bbS „ Bad ›miwLA scripsi (= ωÎ ν Π = duas res C = illa duo quae V) : šbbsLA Φ : ›bbsLA Bad 4 ˆ= addidi (= aggregans eos C = quod congregat illos V) 5 õG  ÕI³ scripsi (= υë ποτι θεται 1 Π[praeter e ] = ponitur C) : õCÕI s.l. Φ with the sign ° : õGÖ in textu Φ Bad (= ponit V) øÏkLA Φ (= τωÄì οÏλωì Π) : øK Bad 7 Î|ÕSÊz YÑMRz scripsi (= ιÍσων χρο νων Π = temporibus aequalibus CV) : ÒÕS YÑMRÒ Φ : aÒÕS YÑMRÒ  BadÍ

LA aÑiWi Ò with corresp. marks s.l. and i.m.: úJÕT rÖ³LA YÏ ÒË Öz ÒÚkE rÕMtÒ õbP YÏ ÒË ˆ=ÕQ yI †LD Φ i.m., fort. ex comm. Ibn Adı¯ 816.19–22 3 ›fltM ›miW] + „IftLÒÖ õm`A yI Φ s.l. (probably ref. to YҁMi   Ò YÒÚE) ›miwLA with corresp. marks s.l. and i.m.; the note is hidden in the gutter margin 7 †LD with corresp. marks s.l. and i.m.: blvLÒÖ bÏdA cz Φ i.m. 1

252a25

252a30

15

VIII 1, 252a25-b6

4

Cv 68v

8

ÒÚE cO Mi Ò YÏ Ò  ÕE †LD „ „ÑkLA Ò  ÙbmLA YÏ Ñ  B \¬zLA YÏ ÑËF lm`ÑBÖ Mi Ò ÕE ÒÚEÖ .êÒÕcB (iL ÙBÒ  YÕkI ÒÚK ñNÏÒ  Ö〈 z 〉 ÙBÒ  gcA ~gIÒ  ‡ÙÏqtmLA YÏ Ò  YÒ  yibpLA „ êÑbSi Ò (pIqmd d¬ r ñiLAË cÚLA ÒÚEÖ .Ò« ÙbM ÙBÒ  YÑK ;L úlpI³ YÒ  ˆYR úO iÖ ñN³Õ° kT YÑK ÒÚkE  ÙBÒ  áÑIÒÖR ÓlÏomLA YÏ Ò  †LDÖ ¯ihcB (iL aÑiWi Ò ¡yB „ ˆkaA ÑMÏÑ  F ˆÏ Ri Ò Mi Ò ÒÚ= ÑM ~bbS ÑnEÑE YÏ Ò  ¡V ›tqÑQ ›tIÖ҂L idÑyM ñNÏÒ  „ ñLÕqN ÑM ÒÚE \kilF .iÏLRÒ  fEÖ X| úbS Ñ= (ilF ndÑbmLA .ÆCÒ  YÑMR YÕkI iÖ \kI G KH YÕkT i ÖÒ  \kT G "M [21ê207] 1 (a32) ταυ την] fort. ταυ τηì Ψ (»the adequate principle for this«, Ar.)

(a33) ηà β γ δ Λ P(824.10) : om. E K Pa(365.8) : non

Ψ β γ δ Pa(365.8); cf. S(1186.10) vert. Ar.

2 (a33) εÍ χει Π(praeter E)Ψ : εÍ χειν E

3 (a35) τοÁ α δ FHIJ2 Λ : ταÁ K β J1Qw

5 (b1–2) οÆ ρθωÄ ς ... ουÆ κ Π(praeter E ) : om. E : ουÆ κ Ψ 1

L : om. K

1

6 (b3) τι Π(praeter LK)Ψ : τι αÆ ειÁ

(b3) αÆ ιÈδιο τητος Π(praeter E )Ψ : ιÆ διο τητος E1 1

NR γ : post αιÍτιον L : om. Ψ

(a33) ειÍ τι E K : οÏτι

1

(b4) εÏ τερον1 ante αιÍτιον E K δ

7 (b6) τοσαυÄ τα Π(praeter α) : ταυÄ τα α

8 (b6) ουÆ κ1 E1,

fort. Ψ, K β γ Λ Pa(366.2) : ηà ουÆ κ E2 δ

2 ñNÏÒ  Öz scripsi (= ηÃ Π = aut quia C) : scripsi : s.p. Φ :

ñNÒÖ Φ (= et quod V)

3 d¬ r scr. Bad :

ádr Φ

4 ñN³Õ° kT

ñNÕkI Bad Ò« ÙbM scripsi (= τουÄ ... αÆ ειÁ ... αÆ ρχηÁ ν ζητειÄν Π = quare est semper

principium C = quare semper fuit principium V, where CV misinterpreted the preceding lima¯ ka¯na) :

ÒÙbM Φ : ÒaÙbM Bad

6 ˆÏ Ri   Ò Mi Ò ÒÚ= scripsi (= τηÄ ς αÆ ιÈδιο τητος ταυ της Π [praeter

ŠrÒ MiÒ ÒÙ= Φ : Š¦ÖÊÒ  Mi Ò Òچ Bad 7 (ilF scr. Bad (= ουÆ κ εÍ στιν εÏ τερον αιÍτιον Π = non est causa altera C) : silF Φ (cf. non habent alias causas V = XÒª êÑbSz Ñ= silF) 8 \kI scripsi : s.p. Φ : \kT Bad 1

E ] = huic rei perpetuae C = ad hanc rem aeternam V) :

252a35 252b

252b5

(iLÑPÕpSrt fÏ yibpLA û4LA êÑtK \M nMÑoLA LÑqmLA

16 Av. t. 16 Cp 176v

4

Av. t. 17

8

ÒDÒË LA aÑiWi ÑF .†LD ñB ¿rÑyI ÑM ø° H ­yI (iLÖ [18Ò208] ¡V \M ~niH KH YÕkT YÒ  \km ÙQ ñNÏÒ  Í CÏÑX юM \Ï Z³ KaA MÒ  „ \Ï zI³ ˆÏ RÒ  iÏvT YÕkI i ñNÏj F iÍ ÖÏÒ  ÑMÏÒ  .áÚE fE 〉 ÆCÒ  NÑK YÕkT YÒ  †LÚLÖ Î ŠÒË Î \M YÕkI YÒ  ñN{wF iÏvT øÏK YÏ Ò  †LDÖ .〈 ÆCÒ  iÖ iÏvT øÏkL ›tIш iÏvtLA YÕkI ;F YÒÚlLA YÒÙÏgLA YÕkI YÒ  úO ÙQ (iL ÑM …ÏhtI YÒ  \km ñNÏÒ  gN ÙQ ~gIÒ  Ö .Iш jB ÆCÒ  a …ÏhtI áÚE ÙP ÑNÏÑËF ñL (fN i .F †LD èÑoM .ÆCÒ  KH ñiF iÖ …Ïhtl vÏ ~nKÑS YÕkI øB øÏkLA iÖ …ÏhtI юM aLA \M a‚`A i aÑiWi Ò ¡V ÑMÏÒËÖ ÙBÒ  KÏhtM [ê208] ÑMÏÒË YÕkT YÒ  úJÒÕLA YÑKÖ .~niH …ÏhtI G YÒ  ÙyB ÔÙW siL KaA NÑK YÒË ÑQÖi Ò \M QÖ „ KÏhtM 3 (b8) τοιωÄ νδε Π(praeter E1), fort. ΨΩ : τοιουτων E1 E FHIJQ Λ S (1187.21) Pa(366.11) 1

p

7 (b13) αυÆ τωÄì E2Ψ K β δ w : εë αυτωÄì

8 (b14) ουÍ τε ... ουÆ δεÁν] ουÆ δεÁν ουÍ τε µε ρος interpr. Ar.

10 (b16) ουË σα α K γ δ Λ : ουË σα αÆ λλ’ εÍ στι[ν] αÆ ιÈÂδιον β [αÆ ιÈÂδιος R]

(iLÖ add. (iLÑPÕpSrÒ èÑQ ÓLÑoLA ylytLA Φ : (iLÑPÕpSrÒ  èÑQ 〈KaA ‡ÙQ ÙG ÑGÒ£YiÒ ‹Y Bad 1 ante

om. ΠCV : seclusi : 3–4

dLA〉 2 ÓLÑoLA ylytLA

ÆCÒ  ... fE addidi (= τωÄ ν τοιωÄ νδε

[τοιουτων E ] ..., πρω Ä τον µεÁν [µεÁν ουË ν Q] οÏτι ουÆ δεµι α αÆ ι διος [ιÍδιος L] µεταβοληÂ Π = sunt iste. 1

In primis quidem quia non est mutatio perpetua omnino C = sunt ista. Primo quia non erit transmutatio aeterna omnino V) : om. hom. Φ, Bad

YÑK ex ras. Φ

4

9 YÑKÖ (= et est C = et esset V) Φ :

ÕLÖ

2

Î ŠÒË Î \M] + ÙÏG ... ÙÏG cz Φ s.l. ex comm. Philop. 824.15f. (cf. Simpl. 1187.8f.)

Chapter 2

252b10

252b15

17

VIII 2, 252b7–27

Cp 177r 3

6

9

YÕkN ÙQ ÑNÏÒ  †LDÖ Í CÏ ÑX (fNtÒ ÒÖD „ ¡oK eZÒ  !ymLA ÒÚEÖ .\kT ÑnsfNÒ  aÑqlT \M ÑniF ÔÙUÖ ~niH …ÏhtN vÏ ÆCÒ  ÑniF KH i \IrÏÑQ ÑnsL !ymLA ÒÚE YÏ ÑËF .ÑnKÏU rÑX \M a \kI G YÒËÖ KH zÙbM ÑM a \Y …ÏhtT ;NÏÒË ÙBÒ  fE øB ñL (fN i .F èÑomLA ÒÚE ‹Y áÙP YÑK DÒË †LÚLÖ .ñsfN …ÏU ÕE ñNÏÒË èÕqN ÑNÏÑËF fÏ aA ÑMÏÑ  F ÑE¡V rÑX \M ñTÒD \M KH …ÏhtI i .F ÔÙU ÙqF ‡0LA ‹Y ~nKÑS ~niH YÕkI ÙQ iL ÒD;F YÒÕiaA „ ÔÙU YÒ  \km ÙQ †LD YÑK YÒËÖ .rÑX ÑJÏ i YÕkI ÙQ YÑK YÒË ñNÏÒ  †LDÖ ?~gIÒ  øÏkLA „ ñniyB †LD ‡‚lI YÒ  õnm cyW GÑyLA „ YÕkI ÙQ YÑK YÒËÖ .~gIÒ  ¡bkLA „ YÕkI ÙqF ¡vcLA GÑyLA „ 2 (b18) εÆ ν ηë µιÄν εÆ νου σης E2Ψ K β δ IJQw Λ : εÆ ν ηë µιÄν [ras.]εου σης E1 : εÆ νου σης H : ουÍ σης F (b19) οϵως Π(praeter E)Ψ : οë µοι ως E

(b20) αυÆ τωÄ ν Π(praeter E1)Ψ : om. E1

κινη σεως α K be β : κινη σεως εÆ νιο τε e γ Λ 1

2

F : κινειÄ τι I : τι κινειÄ αυÆ ταÁ δ H ζω Äì ον F

4 (b22) κινειÄ τι αυÆ ταÁ α K β RIJQw Λ : τι αυÆ ταÁ

5 (b22) δεÁ ζωÄì ον E1Ψ K β δ HIJQw Λ : δ’ εÍ ξω οÃν E2 : τε

6 (b23) πα µπαν Π(praeter K)Ψ : om. K Sp(1187.38)

Sp(1187.40) : αυÆ τουÄ E β δ J1 Q(?) w Pa(367.8) E

3 (b20)

(b24) αυë τουÄ Ψ K FHIJ2

7 (b25) εÆ ν Π(praeter E)Ψ Sp(1188.1) : om.

8 (b25) τοÁ αυÆ τοÁ K β γ Λ Sp(1188.3) : τουÄ το τοÁ αυÆ τοÁ E : τοÁ αυÆ τοÁ τουÄ το δ : Ψ = E ∨ δ

(b26) γι (γ)νεται α K be1 β Λ : γι νεσθαι e2 γ

9 (b27) καιÁ 1 Ψ K γ Λ Pa(367.10) : καà ν E β δ

(b27) καιÁ 2 α K β δ FHI Λ : καà ν JQw Pa(367.11)

1 (fNtÒ Φ :

(fnLA Bad

5 ÕE scr. Bad : om. in textu + ÖÙE vid. s.l. Φ : om. CV

scripsi (= ωÏ στ’ [vel ωÏ στε] Π = et propterea C = et ideo V) :

†LÚKÖ Φ Bad

†LÚLÖ

7 rÑX ÑJÏ scripsi

rÑX \M (= extrinsecus? V) Φ in textu Bad : ÑJÏ i U xsN † 〈r〉ÑX Φ i.m. (»in Yahya¯’s copy: la¯ mimma¯ kha¯〈ri〉jun«) 8 †LD ‡‚lI s.l. Φ2 (= τοÁ αυÆ τοÁ ˙ συµβηÄ ναι [v. App. 1] = sequatur illud C) : †LD YÕkI Φ in textu (= hoc sit V) Bad (for ‡‚L =

(= ex eo quod est extra C) :

συµβαι νω cf. 220a18 = Bad 430.15, 233b18 = 635.16, 234a22.b4 = 640.9, 645.12; Glossary)

252b20

252b25

18

Av. t. 18 4

Cp 177v

8

(iLÑPÕpSrt fÏ yibpLA û4LA êÑtK \M nMÑoLA LÑqmLA

i ÑM á¬Ñ  B \ksIÖ …ÏhtI YÒ  \km ÙQ YÑK YÒË ñL Iш i .F YÕkI ÙqF .ñL Iш \IÙÏgLA ŠÒË KaA YÏ Ò  \M iÍ ÖÏÒ  øiQ cÚLA èÕqLA ÑMÏÒ  †LD „ èÕqnF YÕkI YÒ  †LÚB „ÊlXÒ  ñNÏÑËF .êÈÒÕcF dÙyLA „ юiyB Í ÙHÒÖ ÙBÒ  siL YÕkT YÒ  \km ÙBÒ   (iL ñniyB ÙHÒÕLA aLA KH NÑK YÒË Í rÖ² ~bJÒÖ ÙHÒÕLA TÂÕÂLA mvN øE èÕqN YÒ  øoM †LD ˆÕqB YÒ  Ö .юiyB Í ÙHÒÖ LÑH ‹Y ÂTÂÕÂLA YÏ Ò  ‹Y gXiªÒ ¡V ÙHÒÖ ÙBÒ  fE ÖÒ  юiyB ÙHÒÖ (iL cO Mi Ò YÑK ›LÕqLA cÏ Ò  ‹Y \kLÖ .…ÏhtI ñNÏÒ  ‹YÖ ÙHÒÖ .iÏLRÒ  lctÏM ÑeNÏÒ  øbÂQ \M юiyB Í ÙHÒÖ ÑM KH YÕkT YÒ  \M õnm õÈNÑM   .ÙyB .F ¤Â KÒ  ~NÑiB †LD ›bIÖ 1 (b27) καà ν E N : καÆ ν F : καιÁ εÆ ν fort. Ψ, K δ LR HIJQw Λ Sc(1188.5) Pa(367.11) Äì αÆ ριθµω Äì δ αυÆ τηÁ ν Π(praeter δ)Ψ : αυÆ τηÁ ν τω

6 (b32)

9 (b35) κωλυ ει E K β γ Λ : κωλυ ει µι αν καιÁ

fort. Ψ, δ, fort. Pa(367.20)

4 ñNÏÑËF scripsi (= µεÁν γαÁ ρ Π = quod igitur C = ergo V) :

ÑF Φ : ÑËF Bad 6 †LD ] Φ parum clare : †LÚB scr. Bad 7 юiyB scripsi (= οë αυÆ τοÁ ς Π = idem CV) : їyB Φ (post corr.?) : љyB Bad gXiªÒ ... fE ÖÒ  = ηà αÆ ειÁ [αÆ ειÁ om. I] εÏτερος Π = aut semper sit unus praeter alterum C = aut semper mutatur una post aliam V : gXi ª Ò ¡V ÙHÒÖ ÖÒ  ÙBÒ  iTÒD fE ÖÒ  Φ 2 Bad : iTÒD et ÖÒ  seclusi 8 (iL scripsi (= ΠCV) : (iLÖ Φ Bad

8 gloss without mark i.m.:

dÒÙGtÒ „ siL ÑnlQ Öz dÒÙGtÒ „ KaA YÏ ÒË ÑnlQ aÒÕS cz Φ i.m. ex

comm. Ibn al-Tayyib 822.4f. ˙

252b30

252b35

253a

19

VIII 2, 252b27 – 253a13

Av. t. 19 Cv 69r

4

Av. t. 20 8

Cp 178r Ca 24vb

õGÖ "M knl (ilF …ÏhtI \kI G YÒ  ÙyB aLA …ÏhtI YÒ  ÑMÏÒ  Ö .〈 ñnY ~JrÑX …ÏL ÙJÕI i ~niHÖ dÕJÕM ~niH 〉 ñnY ~JrÑX ñL …ÏdA žiK YÒ  †LD YÕkI žiK ÓhbN YÒ  õGÕmLA ÒÚE „ ÑNÙcQ (iLÖ ~niH dÕJÕM ñniyB ÙHÒÖ ÕE ñKÏU YÒ  ñNÑ  W cÚLÒÖ ñniyB aLA YÕkI áÙÂHÖ !ymLA ÒÚE „ ñk° W ;NÏÒË †LÚB ølÑqLA YÏ ÑËF .…ÏhtI i ~niHÖ …ÏhtI .ÙBÒ  Í KÏhtM љyBÖ ÙBÒ  Í nKÑS љyB ÒdÕJÕmLA \kT 〈 GÊ 〉 ˆÂL ÕEÖ ÔÖÙH YÒ  †ÏW õGÕM ñNÏÒ  \Ï zI Í CÏÑX ÕE ñNÏÑËF ÓLÑoLA èÕqLA ÑMÏÑ  F ÙQ YÒÕiaA YÏ ÑËF .(fNi Ò ÒÖD „ ¿yI ÑM ÕEÖ ø³bQ \M \kT G KH ÑJÏ ÆCÒ  Ó YÕkI YÒ  ¡V \M m †LD ÙyB \M vÏ ~nKÑS iÍ ÖÏÒ  YÕkI ~miW ÙBÒ  ÙP ÑNÏÒ  †LDÖ øPÑB ÒÚE YÏ Ò  ¡V .\Ï zI .F ñKÏH rÑX ÒÚE ñtKH úbS (iLÖ ñiF cÏ ‚IV ÕE ÑJÏ YÒÕiaA „ …ÏhtI [Ò 209] 2 (a3) κινηÄ σαν E, fort. Ψ(cf. 198a19.24 = Bad 136.10, 137.19), β δ FHI : κινηÄ σον K JQw Pa(367.24)

3–5 (a3–5) τουÄ το ... µη E K δ LN2R2 F2IJQ2w Sp(1189.34–35) : om. N1R1

F1HQ1 Λ : habet Ψ, sed fort. µηÁ νυÄ ν pro µε ντοι interpr. Ar. (for

õGÕmLA ÒÚE „ = νυÄ ν cf. ÒÚE „ ... i

196b32 = Bad 120.7, 207b25 = 265.5–6, 213a6 = 333.2–3, etc., also µηÁ νυÄ ν =

õGÕmLA

Galen, Anat. admin. 227.6[Kühn] = 16.2[Garofalo])

F2JQ2w : ουÏ τως LN2

4 (a5) οÍ ντος α K δ R2

*

11 (a12) του του Π(praeter E)Ψ : του των E Pa(368.10)

2 ñnY ... ~niH (= οë τεÁ µεÁν ... οë τεÁ δεÁ µη [253a3] E b d FHI) addidi : om. hom. ΦCV

3 ÒÚE „ õGÕmLA (= hic CV) Φ i.m. (partly illeg.) with insertion mark after ÑNÙcQ : om. in textu Φ : secl. Bad 4 ÕE scripsi : ÕEÖ Φ (= et CV) Bad dÕJÕM scripsi (= οÍ ντος Π[praeter β]) : dÕJÕM Φ Bad : om. CV 5 †LÚB scripsi : †LÚL Φ Bad 6 \kT GÊ ˆÂL scripsi (= διαÁ τι ουÆ κ [...  ηÆ ρεµειÄ] Π = quare ... non sunt [... quiescentia] V) : an non [... quiescunt] C (= ... \kT GÊ YÒË ?) : \kT G (om. G ex haplogr.) Φ Bad 10 ñKÏH scripsi (= κινη σαντος Π = moueat ipsum C = mouente ipsum V) : ñKH Φ : KH Bad 11 ñtKH scripsi (= του του ... τηÄ ς κινη σεως ... αιÍτιον Π) : ñKH Φ : KH Bad, fort. CV (v. App. 1 ad loc.)

253a5

253a10

(iLÑPÕpSrt fÏ yibpLA û4LA êÑtK \M nMÑoLA LÑqmLA

20

4

8 Av. t. 21 Cp 178v

…ÏU ñNÏÒË ñiF ÑnLÕQÖ .úsHÒ  .F YÒÕiaÑB ×idA \Ï kL ñsfN Y ÒÕiaA õNÑM (ilF .YÑkmLA „ KaA yN ;NÏÒË øB KH øÏK ñB yN (iL ñsÂfN ÑKH YÙbLA „ ÔÙW ÙQ YÕkT YÒ  Í rÖ² úO ñNÏÒ  „ÈilX øB õnm ÒÚE …ÏUÖ ՕLÒÖ czLA …ÏU áÚE ¡yB YÕkIÖ ×idA \Y ¡oK YÑK YÒËÖ ñNÏÒ  †LDÖ ‡ÕnLA ÙnY ¿yI ÑM †LD èÑoM .á¬Ñ  B YÒÕiaA ÚmniH ñlÏK YÒÕiaA ñMÕN \M ñbtnI ;NÏÒË ñNÏÑËF tÏbLA ( Ï H KH wÑnlL ÙJÕT (iL .èÕqLA ÒÚE ÕltI ÑJÏ ~gIÒ  †LD „ Mi Ò eziS \kL .bT KH ñiF YÏ i  [7Ò209] ‡ÙÏqT cÚLA †ÏwLA „ ÂKD³ ÑM ÕE  hfLA ÒÚE zÙbB [9ê209] èÕqnF ?\ksiF dÕyI ~niHÖ …ÏhtI ~niH ÒdÕJÕmLA ¡yB rÑC ˆÂL ñfCÖ Â 1 (a13) ιÍσως α K γ δ Λ Pa(368.11) : ουÏ τως β

3 (a16) εÆ ν µεÁν τωÄì σω µατι α K δ T(214.18) Sp 4 (a17) ηÃ

(1190.33f.) : τωÄì σω µατι NR HIJQ : om. L : τωÄì σω µατι post κινη σεις F

Π(praeter E ) T(214.18) : om. E : om. vel καιÁ Ψ (cf. Pa[368.21]) 1

1

5 (a18) οιÎον α : οë ποιÄον

Kβγδ

6 (a20) πα ντα ταÁ ζωÄì α α : ταÁ ζωÄì α πα λιν K β γ δ Λ

fort. Ψ

9 (a22) εÍ σται ηÏ περ Ψ e2 FHI : ειÍπερ E : ηÏ περ K be1 β J2Q : ηÃπερ J1

K δ FHIw Λ Sl(1193.1) : om. β JQ

7 (a20) εÆ νου σης Π : µενου σης (a22) περιÁ α

10 (a23) ποτε] om. F, fort. Pa(368.28) : non vert. Ar.

(a23) οë τεÁ Π(praeter E1)Ψ : ουÍ τε (?)E1 2 yN ... yN scripsi (= significamus ... significamus C = intendimus V) : s.p. Φ :

ÔÙU ... YÕkI Bad 4 ×idA \Y scripsi (= υë ποÁ τουÄ ×idA  Y Φ : ×idA ¡V Bad 5 ÚmniH scripsi (= ηÍ δη Π = quandoque C = tunc V) : s.p. Φ : ÙsJ Bad 7 bT scr. Bad : ÑqbT Φ ÑJÏ Φ (= ΠCV) : .F Bad 9 ante zÙbB add. (iLÑPÕpSrÒ èÑQ õBҁLA ylytLA Φ in textu + ñl҂JÒ  \M ¯Y õSÑtLA a‚`A X| Φ i.m. : om. ΠCV : seclusi : (iLÑPÕpSrÒ  èÑQ 〈YÕkLA † YÕksLÒÖ KaA õIRÕT YÑkMÒË〉 3 õBҁLA ylytLA Bad zÙbB scripsi (= αÆ ρχηÁ δεÁ Π = et principium quidem C = et principium V; cf. also Gˇp 162.7:  hfLA „ zÙbI) : øoMÖ Φ Bad

Bad

3 ÔÙW ... YÕkT scripsi : s.p. Φ :

!yI ...!yI

περιε χοντος Π = a continente CV) :

253a15

253a20

Chapter 3

21

VIII 2, 252a13 – 3, 253a34

4

8 Av. t. 22 Cp 179r

ÑelÏK ÑMÏÒËÖ ÙBÒ  nKÑS ÑelÏK 〈 ÑMÏÒË 〉 aÑiWi Ò YÕkT YÒ  Í rÖ² úJÒÖ ñNÏÒË YÕkT YÒ  ÑMÏÒË ~gIÒ  áÚEÖ nKÑS љyBÖ KÏhtM љyB ÑMÏÒËÖ ÙBÒ  KÏhtM YÕkT YÒ  ÑMÏÒËÖ ÙBÒ  〈 nKÑS юM 〉 nKÑsLÒÖ ÙBÒ  …ÏhtT юM KÏhtmLA Mi Ò YÕkI YÒ  ÑMÏÒËÖ \ksT YÒ  Ö …ÏhtT YÒ  ÙHÒÖ èÑoM ‹Y шѠ W \M ÑelÏK …ÏhtI i љyB ÒdÕJÕmLA YÕkT YÒ  \kmÍ ÙQ Í ñNÏÒ  †LDÖ
ÑbLA ÓLÑoLA ñB cÚLA ÒÚEÖ .~yi? \IMi Ò èÖÒÙtI љyBÖ ÙBÒ  …ÏhtI љyBÖ ÙBÒ  èÈ;KÖ †LD \M †ÏwLA ñiF õQÖ ÑM øÏK ø° H YÕkI †LÚB YÏ ÑËF èÕqN YÒ  úO .ñlibsB \V cÚLA ˆlyLA ÒÚ= …³TÖ †LD „ jÏH %³ 0LÖ nKÑS ÑelÏK aÑiWi Ò YÏ Ñ  B èÕqLA ÑMÏÑ  F Í .czLA \M žÈygF ( ;NÏÒË †LD „ †° wLÒÖ Ï aA ñB ٕI ÑM ŠÒË ÑftLiÒ 1 (a24) δηÁ ηÍ τοι E1ΨΩ K β Q : δ’ ηÍ τοι E2 δ FIJw Λ : δη τοι ηà H

3 (a27) ηÆ ρεµειÄν Π(praeter

e1Q) : ηÆ ρεµειÄν αÆ ειÁ /αιÆ ειÁ (aut inv.) ΨΩ e1 Q2 As(frag. 563) Sp(1193.17) Pa(369.2–3) εÍ τι καιÁ E K β γ : εÍ τι καιÁ τοÁ δ : an εÍ στι [καιÁ ] τοÁ Ψ? (καιÁ non vert. Ar.)

K γ, δ (post τωÄ ν αÆ πορουµε νων) : om. β (cf. HaGa note ad loc.)

1

5 (a28)

7 (a31) (αë )πα ντων α

9 (a32) τοÁ µεÁν ουË ν πα ντ’ ηÆ ρεµειÄν λε γειν interpr. Ar.

10 (a33) τι ς] om. Pa(369.12) : non vert. Ar.

ÑMÏÒË addidi (= ηÍ τοι Π[praeter H] = aut CV)

2–3

…ÏhtT ... KÏhtmLA YÕkT scripsi (= ταÁ ...

κινου µενα κινειÄσθαι Π[om. Q ] = que sunt mobiles mouentur C = illa quae mouentur moue1

antur V) :

…hbB ... ñKhbmLA YÕkB Φ : …htI ... …htmLA YÕkI Bad

3 nKÑS юM addidi (= ταÁ δ’

ηÆ ρεµουÄ ντα [...] ηÆ ρεµειÄν Π[v. App. 1] = que sunt quiescentes quiescunt C = quae quiescunt

…ÏhtT s.p. Φ (= ΠCV, v. also Gˇp 204.12) : …W Bad ÒÚE YÒ;KÖ Φ : ÒÚE è³;KÖ scr. Bad 9 èÕqLA scr. Bad : èÕqmLÒÖ with

quiescant V) : om. hom. Φ

7–8 ÒÚ=

èÈ;KÖ scripsi :

4

deletion signs over Wa¯w and Mı¯m Φ

253a25

253a30

(iLÑPÕpSrt fÏ yibpLA û4LA êÑtK \M nMÑoLA LÑqmLA

22

Cv 69v

4

Av. t. 23 8

úHÑC ‹Y dÏr ÕE ;NÏÒË †LD (iLÖ .ñnM a‚J „ i ñlÏK Mi Ò „ ÕE YÏ Ò  †LDÖ ÑelÏK aÒrvÒÖ ÆoM ÑelÏK ŸrÑymLA dÏr ñnÏkL áÙHÖ fÏ yibpLA ˆlyLA iÏgÑytLA ÑymLA „ ndÑbmLA dÏr YÏ Ò  ;K ~gIÒ  Ö .KaA ømytsT ÑelÏK áÚE †LÚK ~gIÒ  ‡ÕlyLA lÑS „ èÑomLA ÒÚE ‹YÖ fÏ gÑytLA ‹Y Òd­ r ÕE (iL øCi Ò YÏ Ò  †LDÖ fÏ yibpLA ‹Y Òd­ r YvÒ áÑNKD cÚLA d° r ÕE (iL .KhlL Ò« ÙbM yibpLA YÏ Ò  ñL ûÕGÕmLA ÇÒË [Ò210] jÍ PÑB …ÏhtT ÑelÏK aÑiWi Ò YÏ Ñ  B èÕqLA ~gIÒ  YÕkI YÒ  dÑkIÖ „ yÂGÖ³ yibpLA YÏ Ò  †LDÖ .úEÚmLA \Y ցbA „ †LD YÖd ñNÏÒ   ‹YÖ ~gIÒ  YÕkslL KhlL ÑM øomK zÙbM ÑeNÏÒ  iyibpLA rÕMiªÒ èÑomLA ÒÚE Ï 4 (b3) ειÆ σιν vel ειÆ σιÁ E1, fort. Ψ, β γ Λ : εÆ στι δ : εÍ τι E2 K

(b3–4) τοÁ ν µαθηµατικο ν α β

FHJQw Λ : τοÁ µαθητικο ν I : ταÁ µαθηµατικα K : τοÁ ν µαθητικο ν b (cf. 257a21)

6 (b6) τηÄ ς

Π(praeter FJw)Ψ : τις Jw : om. F Pa(369.21)

2

7 (b6) δεÁ E Ψ K β FH : δε τι E δ JQw Λ

8 (b7) του του α δ : τουÄ το K FIJQw Λ : om. β H

1

(b8) µεÁν] exp. E2, non vert. Ar.

9 (b9)

οë µοι ως α β HJQw Λ : οϵως K δ FI Camotiana (〈ουÆ χ〉 οë µοι ως propos. HaGa, v. also notes

HaGa, Ross and WiCo)

2 ŸrÑymLA scripsi :

ŸrÑymlL Φ Bad

3 igÑytLA Φ :

imilytLA Bad 5 d° r scripsi (= [αιë ] περιÁ τουÄ νυÄ ν ρë ηθε ντος [εÆ νστα σεις] Π = negare hoc quod diximus V) : dρLA ΦC Bad fyibpLA scripsi (= τοÁ ν ϕυσικο ν Π = naturali CV) : õibpLA Φ : õbpLA Bad 9 YÕkslL Φ (= ΠCV) : YÕksLA Bad post èÑomLA add. øiQ Φ Bad : om. ΠCV : seclusi, cf. gloss øiQ zfL cÙY \BÒ xsN „ (iL i.m. Φ (»the word qı¯la is missing in Ibn Adı¯’s copy«)

úEÚmLA] + fÏ yibpLA úEÚmLA yI Φ i.m. = Alex. frag. 566 (v. p. LXXIII); cf. Philop. 825.25, Simpl. 1195.32 9 øomK with corresp. marks s.l. and i.m.: øilytLA ÒÚE YÏ ÒË lqF fLA ÑBz L{S †LÚK (vel ÖÙbI ?) ÕE èÑqF êÒÕcLA \Y ցbA „ ›LÕqLA cÖÑsT tqI Φ i.m.

8

253a35 253b

253b5

23

VIII 3, 253a34-b20

Cp 179v 3

6

9

siL ÒdÕJÕmLA \M KÏhtmLA YÏ ÒË ÒÕLÑQ ~MÕQ YÏ Ò  ÇÒË fÏ yibP MÒ  ÑeNÏÒ  †LD YÏ Ò  ¡V qÒd яKHÖ KÏhtM ÑelÏK fE øB …ÏhtM ¡V ¡yBÖ ¡ybLA ÖÒ  YÕnyI KH cÏ Ò  ÒÕclÏ[ G ÒÕNÑK YÊ ÒËÖ aiqEÖ .ñB ywN jF ÑnsÏH \Y Ú° wI YÒ  \km (iL ñNÏÑËF .ˆlY dρLA úycI (ilF ÑelÏK ÑKaA YÕnyI ˆeNÏÒ  .×SÕLA Mi Ò ~gIÒ  ;ŽiB øB ~qÒd øÏhmgI YÒ  iÖ fmnI aLA YÕkI rÑjhlL bnLA „lÊFÂÖ pÊqÂLA Ï H MÒ  \M èÑqI³ ;B ñibW èÕqLA ÒÚEÖ YÒ  ÑM rÒÙqM t¬ H ÖÒ  yFd ÙQ pqLA NÑK YÒË úO (iL ñNÏÒ  †LDÖ „ ñloM pqLA „ Mi Ò øB YÑM‚LA †LD žcN „ ~MÙÏqtM †LD žcN YÕkI юM a‚`A ÑMÏÑ  F ÑM rÒÙqM …ÏU ҁpqLA \M ÑM rÒÙqM YÏ Ò  YÒ  nifsLA ÙÏM Ï aÑB øcfN cÚLA YÏ Ò  †LDÖ .rÒÙqmLA †LD ÆCÒ  YÑMR „ …ÏU (ilF Í a҂JÑ  B qnI ÙQ ádҁfN ‹Y a҂Ji Ò †lT \M ÙHÒÖ …ÏhtI G ñNÏÒ  ¡V ¡oK 1 (b9) καιÁ 2 Π(praeter E) : om. E : »however/except that some say...« intell. Ar. καιÁ ... εÍ στι2 Πpraeter E1)Ψ : om. hom. E1

(b15) τουÄ τοÁ ν α β δ F2HQw Λ : τουÄ τον K (?)F1 J Pa(370.7) : τοÁ ν I K δ FIJQw Λ : om. β H

5 (b14)

6 (b14) οë Ψ K e β FHIJQ Pa(370.6) : om. E b w 8 (b17) προ τερον EΨ*

10 (b19) τοσουÄ τον Π(praeter δ)Ψ : τοσουÄ τον διαιρειÄ δ

(b20)

µεÁν ουË ν Π : µεÁν γαÁ ρ fort. Ψ(cf. Glossary and Ullmann NE II, 330), fort. Λ

5 fmnI s.p. Φ :

h¦mnÊI Bad øhmgI s.p. Φ (= ΠCV) : øhgI Bad 6 ñibW scripsi (= οϵοιος Π = similis CV) : s.p. Φ : ñbbS Bad 8 ~MÙ® qÂtÂM scripsi (= προ τερον [v. App. 1] : ÆoM Φ (= verbi gratia C) Bad : om. V

×SÕLA Mi Ò with corresp. marks s.l. and i.m.: ñiLAË cÚLA ÕE ×SÖz Mz èjhmGiÒÖ ÕÏhA ›B cz èjhmGiÒ nÙtbI ñnMÖ ÕÏhA fnI Φ i.m., fort. ex comm. Philop. 826.8–10; the last part of this 5

gloss is interpolated in Scotus’ transl. : »scilicet illud ad quod pervenit augmentum et a quo incipit diminutio«.

253b10

253b15

253b20

(iLÑPÕpSrt fÏ yibpLA û4LA êÑtK \M nMÑoLA LÑqmLA

24

Cp 180r

4

Ca 25ra

8

qnI ÙQ èjhmGiÒ YÏ Ò  øbÂQ \M úO (iL ñNÏÒ  EÑzF . KÏW ~yM øB úED ;BÏr øB ~qÒd ÑM a øÏ hmgmLA \M úEÚI ÙQ YÕkI YÒ  Iш jB .á¬Ñ  B ÕE NÑK LÑhtS cÏ Ò  ~gIÒ  LÑhtSiÒ „ Mi Ò cO èÑomLA ÒÚE ‹YÖ \M YÕkT YÒ  Iш jB qnI ÙQ øihtsmLA YÑK YÒË úO (iL ñNÏÒ  †LDÖ Í yFd YÕkT ¡oK rҁM \kL Iш jB LÑhtSiÒ ~gIÒ  qnT †LD øbÂQ hÏcLA ŠÒË á¡cM YÕkI YÒ  úO ¡ImLA YÏ ÑËF ~gIÒ  Ö .dÕm`A YÕkI  ;K i hÏcLA ŠÒË YÕkI ;NÏÒË áiÏvTÖ YÑMR ŸP „ áiÏvT YÕkI (iLÖ YÑMR „ Í ŠÒË ~qÒd øihtsI ñNÏÒË ñL³ÑH áÚE \MÂ Í „ èÕqI \M èÕqF .jÍ CÒ  X| a ŠÒË YÕkT ;NÏÒË LÑhtSiÒ YÏ Ò  †LDÖ rÕezLA IÑV „ ÕE .F †ÏW ÕE ;NÏÒË .ÕE ÑJÏ ›Â LA  iÖ úÂlCÒ  ¡cI i *aA YÏ ÑËF ~gIÒ  Ö ÙÏgLA 1 (b21) αÏ µα E1Ψ K FIJQw Sc(1198.17) Pa(370.12) : οÏλον αÏ µα E2 β δ Λ : αÏ µα οÏλον H 2 (b21) αÆ ει Π(praeter KF)Ψ : ειÆ K F

7 (b26) τι νοση σηì E1 : τις νοση σηì E2 : νοση σηì τις K

δ I J : Ψ = E ∨ K δ I J : νοση σηì τι β FHI2J1Qw2 Λ Pa(370.15) 1 2

2

1 2

8 (b27–28) αÆ να γκη δεÁ ειÆ ς

K e β F HIJQw Λ : καιÁ E : καιÁ ειÆ ς E Ψ b : αÆ να γκη δεÁ καιÁ ειÆ ς Pa(370.22–23, mss. LC) 2

2

1

2

[αÆ να γκη ... µεταβα λλειν om. hom. e1 F1]

9 (b28) µηÁ E, fort. Ψ, δ : om. K β γ Λ

10 (b29–30) λι αν εÆ στιÁ (λει αν εÆ στιÁ ν E : λι αν εÆ στιÁ ν εÆ ν δ) τοιÄς ϕανεροιÄς αÆ µϕισβητειÄν Π : εÆ στιÁ 1

τοιÄς λι αν ϕανεροιÄς αÆ µϕισβητειÄν intell. Ar.

11 (b30) οë δεÁ E K b (?)e J2w : οÏ τε β FHIJ1Q Λ :

εÍ τι δ’ οë fort. Ψ

5 YÕkT s.p. Φ :

YÕkI Bad

8 ŸP s.p. Φ (= πε ρατι Π = extremitate C = extremo V) :

ŸZ

Bad 6 gloss without mark i.m.: dÕJÕLA „ yI Φ i.m. (the point of reference is not clear, cf. comm. Ibn al-Tayyib 832.4–7) ˙

8

YÑMR ŸP] + YvÒ yI Φ s.l., fort. ex comm. Philop. 826.8–11

253b25

253b30

25

VIII 3, 253b21 – 254a10

Av. t. 24 Cp 180v 3

Cv 70r 6

9

12

øfSÒ  ŠÒË ñLւN ÑnilY úEÚI ñtlqN³ „ jaA YÕkI YÒ  ~gIÒ  újyLA \MÖ ÙHÒÖ øÏKÖ ¿ri Ò YÏ ÑËF ~gIÒ  Ö .ñB ywN i htÏH ¿ri Ò ‹Y ñobL ÖÒ  юM KÏhtMÖ Ñec° ] LA ÑeyGÒÕM „ 〉 oBi XÂiªÒ aÑiWi Ò lÑS \M úO ÙqF Ñec° ] LA ÑeyGÒÕM „ fE aÑiWi Ò ¡yB YÑK YÑËF 〈 ­Q .Í KÏhtM ÑelÏK aÑiWi Ò YÕkT i YÒ  Í rÖ² i ñNÏÑ  B [ê 210] „IÙctLA {iϏI ÑeŒWÒ  ÑJÏ ÑE¡VÖ aÑiWi Ò áÚE \B ~gIÒ  \km iÖ .ÙBÒ  \ksT ÑelÏK ÖÒ  ÙBÒ  …ÏhtT ÑelÏK aÑiWi Ò YÕkT YÒ  \km \M a (iLÖ Í KÏhtM ÙBÒ  љyBÖ ÙBÒ  Í nKÑS aÑiWi Ò ¡yB YÕkT YÒ  iÖ áÑnlQ ÑM øoM †LD LÑHÒË „ èÕqNÖ .~niH …ÏhtIÖ ~niH \ksI jÍ CÒ  aÑiWi Ò ÔÙW ÙQ ÑEÑNKD LA ¡ivtLA ŸÑnCÒ  ÙP ÑNÏÒ  †LDÖ ñlbQ øiQ ÑM LÑHÒË „ rÕMtÒ ÙNÑyI ÒÚE „ …ÏÑwLA YÏ ÑËF †LD õMÖ .~gIÒ  шÑiYÑ  B ÙHÒÖ aÑiWÒ  „ YÑK ÙQÖ ñybP \Y ~JrÑX …ÏhtI ÙQ aLA \kI G YÒË ñNÏÒ  †LDÖ EÑzLA 2 (b32) ηà Π(praeter δ)Ψ : ηà καιÁ δ

3 (b33) εÆ ξ αÆ να γκης] non vert. Ar., om.(?) Pa(371.1)

(b34) κινουÄ νται Π(praeter FHI)Ψ Pa(371.2) : κινειÄται FI : κινοιÄται H Ω

J w S (1200.27) : ουÍ τε L NR FHJ Q : ουÆ L 2

l

2

1

1

8 (a3) ουÆ δεÁ α K δ

9 (a5) δ’ [δεÁ K] Π(praeter β)Ψ : om. β

11 (a7) καιÁ εÆ πιÁ α β : εÆ πιÁ K γ δ Λ T(216.23) Pc(827.17) Pa(371.12f.) 1 ñtlqN s.p. Φ (= τοÁ ϕε ρεσθαι Π = permutatione sua C = suo motu V) :

ñblqT Bad (for nuqlatun

= ϕορα and ϕε ρεσθαι v. 215a18 = Bad 363.2, 217b22 = 401.2, 226a33–34 = 531.8–9, 243a39 = 746ult., etc.) clare Φ

2 ÖÒ  Φ (= Π[praeter δ]CV) : Ö Bad

3 oBi scr. Bad :

ñbBÒi parum

3–4 ­Q ... „ addidi (= εÆ ν τοιÄς οιÆ κει οις το ποις, κινουÄ νται [v. App. 1] δεÁ βιαι ως εÆ κ

του των Π = in locis suis que sunt sibi proprii et mouentur ab eis violente C = in suis locis

propriis et mobiles ab eis violente V) : om. Φ CV) :

YÒËÖ Φ Bad

5

i (= ΠCV) om. Bad

4

YÑËF scripsi (= ειÍπερ ουË ν Π = si igitur

253b35 254a

254a5

254a10

(iLÑPÕpSrt fÏ yibpLA û4LA êÑtK \M nMÑoLA LÑqmLA

26

Cp 181r

3 Av. t. 25

6

Av. t. 26 9

ø³pbI³ YvÒ èÕqLA ÒÚeF .­Q KH iÖ Õ± n ÑnEÑE \kI G ~nKÑS ø³bQ  %ÑnLA YÏ Ò  †LDÖ ~gIÒ  KaA ø³pbI³ YÕkI YÒ  dÑkIÖ dÑsfLÒÖ YÕkLA ~yi? áÑIÏÑËF iÏvtI ñiLAË cÚLA aLA YÏ Ò  †LDÖ Â.KH ádÑsFÖ aLA YÕK YÏ Ò  YցI †LD \M YÑB ÙqF .…ÑnE \M ÖÒ  ÙsfI ÕeF iÏvtI ñnY cÚLÒÖ ñiF ÖÒ  YÕÏktI . nkS ;BÏr љyBÖ KÏW ;BÏr aÑiWi Ò ¡yB YÏ Ò  YÒ  YvÒ fvbnI ÙqF …ÏhtT ~niHÖ \ksT ~niH ÑelÏK aÑiWi Ò YÏ Ñ  B èÕqLA ÑMÏÑ  F LA ÑymLA †lT \M ~gIÒ  †LD Ò  ÙbM øyPÖ .MÙÏqtmLA øIÖÑQi ÑB ñNqN ّ .‡ÙÏqT .F ÑnMjK ÑnhttF ñB cÚLA ñniyB zÙbmLA †LD Yz YvÒ ÑEÑncb љyB ÑMÏÒËÖ Í KÏhtM ÑelÏK ÑMÏÒËÖ Í nKÑS ÑelÏK ÒdÕJÕmLA YÕkT YÒ  ÑMÏÒË èÕqnF ÙqF Í KÏhtM љyBÖ Í nKÑS љyB NÑK YÑËF .Í KÏhtM љyBÖ Í nKÑS YÕkT YÒ  ÑMÏÒËÖ ÙBÒ  Í KÏhtM љyBÖ ÙBÒ  Í nKÑS љyB YÕkT YÒ  ÑMÏÒË úO 1 (a9) ηë αυÍ ξησις E, fort. Ψ, Q Pa(371.16) : αυÍ ξησις K β δ FHIJw Pa(371.19) : νυÄ ν fort. Ψ

(a10) ουË ν Π Λ

2–3 (a11–12) οë λο γος, σχεδοÁ ν καιÁ τοÁ κινειÄσθαι. i’punxit Ar.,

deinde γι γνεσθαι τι γαÁ ρ καιÁ ϕθει ρεσθαι δοκειÄ παÄ σιν [κι νησιν ειËναι] intell. Ar. τοι E γ δ : ηÍ τοι γαÁ ρ K β : ηÍ τοι αÍ ρα fort. Ψ

9 (a18) ηà γα ρ

(a19) τωÄ ν οÍ ντων α K γ δ : om. β Λ

ηÆ ρεµειÄ1 Ψ K e2 β γ : ηÆ ρεµειÄν αÆ ειÁ E [ηà ταÁ µεÁν ... κινειÄται (a18–19) om. hom. be1 Λ] κινειÄται α K be : κινειÄται τω Ä ν οÍ ντων e β γ Λ Pa(372.9) 1

2

10 (a19) (a20)

11-p. 27.2 (a20–22) ηÍ τοι πα ντα

οë τεÁ µεÁν ηÆ ρεµειÄν οë τεÁ δεÁ κινειÄσθαι ηà ταÁ µεÁν αÆ ειÁ ηÆ ρεµειÄν ταÁ δεÁ αÆ ειÁ (αÆ ειÁ om. JQ1w) κινειÄσθαι Ä ν οÍ ντων οë τεÁ µεÁν ηÆ ρεµειÄν οë τεÁ δεÁ αυÆ τω Ä ν E K e γ : ηÍ τοι πα ντα ηÆ ρεµειÄν ηà πα ντα κινειÄται τω κινειÄσθαι ηà ταÁ µεÁν ηÆ ρεµειÄν ταÁ δεÁ αÆ ειÁ κινειÄσθαι αυÆ τω Ä ν b : ηÍ τοι πα ντα οë τεÁ µεÁν ηÆ ρεµειÄν οë τεÁ δεÁ →

1

ÑnEÑE Φ : ÑnE Bad

3

áÑIÏÑËF with corresp. marks s.l. and i.m.: ¿yLA „ iÏvtLA yI Φ i.m., fort. ex comm. Philop.

4

828.8 (ad γι γνεται τουÄ το)

…ÑnE Φ : †LÑnE Bad

11

1et2

YÕkT s.p. Φ : YÕkI Bad

254a10

254a15

254a20

27

VIII 3, 254a10–31

Cp 181v

3

6

9

љyBÖ ÙBÒ  Í nKÑS љyB YÕkT YÒ  ÑMÏÒËÖ …ÏhtT ~niHÖ \ksT ~niH ÑelÏK YÕkT YÒ  \km i ñNÏÒ  ÑMÏÑ  F .Í KÏhtM ~niHÖ Í nKÑS ~niH љyBÖ ÙBÒ  Í KÏhtM .~gIÒ  YvÒ ñLÕqN \VÖ ‡ÙÏqT .F áÑnlQ ÙQ aF Í nKÑS ÑelÏK ÒdÕJÕmLA iÖ áÑntM ¡V dÕJÕmLA YÏ Ò  \M ‡ÕQ ñLÕqI ÑM ‹Y qiqaÑB Mi Ò YÑK YÒË ñNÏÑËF ¡oKÍ YÏ Ò  ~niH áÙP cÚLA øB ( Ï aÑB ezI (iL a ÒÚE YÏ Ò  ¡V …ÏhtM ÑnEÑeF lm`ÑB \Ï Z ÖÒ  øPÑB \Ï Z ÑnEÑE YÑK YÑËF …ÏhtI dÕJÕmLA aÑiWi Ò \M aLA YÏ Ò  ~niH \Ï Z ÑnEÑE YÑK YÒËÖ øiÏ] ÑnEÑE YÑK YÒË †LÚKÖ .KH ;† \Ï zI³ ÙQ \Ï zLÒÖ øiÏxtLA YÏ Ò  †LDÖ †LD ŸjZ ñNÏÒ  ~niHÖ ÒÚK èÑT ÑNÙnY ñmlY .F jÏH %0LÖ †LD \Y  hfLA \Ï kL .ÑM ÑKH ;eNÏÒ   κινειÄσθαι ηà ταÁ µεÁν αÆ ειÁ κινειÄσθαι ταÁ δεÁ αÆ ειÁ ηÆ ρεµειÄν αυÆ τωÄ ν β : ηÍ τοι ταÁ µεÁν αÆ ειÁ ηÆ ρεµειÄν ταÁ δεÁ αÆ ειÁ κινειÄσθαι ηà πα ντα οë τεÁ µεÁν ηÆ ρεµειÄν οë τεÁ δεÁ κινειÄσθαι ηà ταÁ µεÁν αÆ ειÁ ηÆ ρεµειÄν ταÁ δεÁ αÆ ειÁ κινειÄσθαι αυÆ τω Ä ν Ψ : ηÍ τοι πα ντα οë τεÁ µεÁν ηÆ ρεµειÄν οë τεÁ δεÁ κινειÄσθαι ηà ταÁ µεÁν αÆ ειÁ ηÆ ρεµειÄν ταÁ δ’ αÆ ειÁ κινειÄσθαι 〈ηà ταÁ µεÁν αÆ ειÁ ηÆ ρεµειÄν ταÁ δ’ αÆ ειÁ κινειÄσθαι〉 αυÆ τω Ä ν scr. Pra (and so HaGa WiCo) : ηÍ τοι πα ντα οë τεÁ µεÁν ηÆ ρεµειÄν οë τεÁ δεÁ κινειÄσθαι 〈ηà ταÁ µεÁν αÆ ειÁ ηÆ ρεµειÄν ταÁ δεÁ αÆ ειÁ κινειÄσθαι〉 ηà ταÁ µεÁν αÆ ειÁ ηÆ ρεµειÄν ταÁ δεÁ αÆ ειÁ κινειÄσθαι αυÆ τω Ä ν scr. Ross (v. also note Ross)

(praeter E )Ψ : E )Ψ : ειÍπων E 1

1

1

4 (a24) γαÁ ρ καιÁ Π(praeter K) : γαÁ ρ fort. Ψ, K Λ

6 (a26) κινειÄσθαι E , fort. Ψ, K : κεινειÄσθαι E1 : κινειÄται β γ δ Λ 2

FHIJ1Qw Λ : ηë E2 : ηËì J2

Π(praeter E )Ψ : om. E 1

1 …ÏhtT ... \ksT s.p. Φ :

3 (a24) ειÍπωµεν Π 7 (a28) ειÆ 1 E1Ψ K β δ

9 (a29) κινη σεις Π(praeter E1)Ψ : κινη σις (!)E1

(a31) ωÎ ν

1

ÑbbÍbH Φ : ~sÏH Bad 8 ÒÚK èÑT scripsi (= ουÏ τως Π = est cum dispositione hac C = est talis dispositionis V) : ÒÚK èÑL Φ Bad ~niHÖ Φ (= ΠV) : ~niH ÖÒ  C Bad 9 ÑNÙnY scripsi (= εÍ χοµεν Π = apud nos CV) : ÑB Y Φ : ÑN¡V Bad

…htI ... \ksI Bad

5 ~niH scripsi :

254a25

254a30

(iLÑPÕpSrt fÏ yibpLA û4LA êÑtK \M nMÑoLA LÑqmLA

28

Cp 182r

4 Cv 70v Ca 25rb

8 Av. t. 27

Cp 182v

QftLÑB ñL ÂcÂB i \M øyÊF \M ÕE ;NÏÒË jÏH ŠÒË ñiF ÑtV YÒ  \M ³ bOÒ  ÕE ÑM ›BÖ ñB ƒÕOÕmLA ¡VÖ ñB ƒÕOÕmLA ›BÖ ( Ï Xi ÒÖ Ÿ®i Ò ›B YÕkT YÒ  \km i ~gIÒ  èÑomLA ÒÚE ‹YÖ .zÙbl (iL ÑMÖ zÙbM [Ò211] Í nKÑS љyBÖ ÙBÒ  Í KÏhtM љyB YÕkT YÒ  iÖ Í KÏhtM ÑelÏK ÒdÕJÕmLA ÑNÏÒ  ÕEÖ ñB ƒÕOÕM ÙHÒÖ ÈMÒ  õ?Ò  †LD ¡qN „ …kI cÚLA YÏ ÑËF .ÙBÒ  \km i ñNÏÒ  eZ ÙqF .\ksT ~niHÖ …ÏhtT ~niH ÒdÕJÕmLA ¡yB ÙP \km i ñNÏÒ  eZ ÑM èÑoM ‹Y Í KÏhtM љyBÖ ÙBÒ  Í nKÑS љyB YÕkT YÒ  [5Ò211] .~qÒd Í KÏhtM ÑelÏK YÕkT YÒ  iÖ Í nKÑS ÑelÏK YÕkT YÒ  èÑT ÑelÏK ÒdÕJÕmLA øE znN YÒ  YvÒ B cÚLÑF [6Ò212] Í nKÑS љyBÖ èÑaA áچ љyB 〈 Öz 〉 \ksT YÒ  Ö …ÏhtT YÒ  ÑeyM \km .ñniÏbN YÒ  YvÒ ÑnL fvbnI cÚLA ÕE ÒÚE YÏ ÑËF .ÙBÒ  KÏhtM љyBÖ ÙBÒ  2 (a32) τοÁ 2 fort. Ψ, K γ : om. E β δ ... ηÆ ρεµειÄν] inv. Ψ

(a32) τοÁ 3 Π(praeter E1)Ψ : om. E1

(b4) αÆ ει ante ηÆ ρεµειÄν [b3] transp. Q : non vert. Ar.

K J Q S (1206.8) Pa(372.24) : θεωρητε ον β δ FHIJ Q w Λ 1

1

l

2

2

7 (b3) κινειÄσθαι2 9 (b4) θεωρηÄ σαι α

(b4) πα ντα Π(praeter K) : ταÁ

οÍ ντα πα ντα K, an Ψ? (sed v. πα ντα = »all beings«, Ar.[254a34], εÍ νια = »some beings«,

Ar.[254b1])

1 ÑtV scripsi : s.p. Φ :

ÑtU Bad ÂcÂB scripsi (= visus C) : c¦ Ä I vel sim. vid. Φ : [qui non] potest V : ±bT Bad 4 YÕkT scripsi : s.p. Φ : YÕkI Bad љyBÖ Φ (= ΠCV) : љyB Bad 5 ¡qN scripsi (= ad contradicendum C = in contradicendo V) :  qB vel  yB Φ : ¡yB Bad 6 \ksT ... …ÏhtT s.p. Φ : \ksI ... …htI Bad 7 YÕkT s.p. Φ : YÕkI Bad 9 ante cÚLÑF add. (iLÑPÕpSrÒ èÑQ (MÑbA ylytLA Φ Bad : om. ΠCV : seclusi 10 љyB Öz scripsi (= ηà εÍ νια Π = aut quaedam eorum C = aut quaedam V) : љyB in textu + Ö i.l. Φ : љyBÖ Bad

254a35 254b

254b5

VIII 3, 254a31 – 4, 254b20

3

6

9 Cp 183r

29

„IpB љyB …ÏhtTÖ …ÏW ÑKÏhtmLÒÖ ÑKÏdA YÏ ÒË èÕqnF dÕJÕM YÑK ÑM 〈 øÏK 〉 øoB ¿yLA „IpB ÑMÏÒ  .щÒÖÚB љyBÖ ¿yLA ÑelÏK aÑiWi ÑF щÒÖÚB ÑMÏÒ  Ö a‚`ÑB YÑK ÑMÖ ÑKÏhtmLA ÖÒ  ÑKÏdA „ юM ÑM a‚J Y¬Ñ  B iÖ …ÏhtmLA „ ÖÒ  …ÏdA „ dÕJÕM ÑeNÏÑ  B i …ÏhtT LA ÑE¡V \M љyBÖ Ñ‹ÑqlT \M љyB щÒÖÚB KÏhtmLÒÖ .…ÏW ÖÒ  …ÏH \M …ÏhtI cÚLA YÏ ÑËF .ÑeybP \Y ~JrÑXÖ ­Q љyBÖ ~ybP áÚE ¡yBÖ ñlÑqlT \M ÕE YÒÕiaÑF YÒÕiaA \M ÙHÒÖ øÏK øoM ~ybP …ÏhtI ;NÏÑËF ñlÑqlT .~ybP …ÏhtI ñNÏÒË †LD „ èÕqN ÑNÏÑËF ñiF ñtKH Òª ÙbM YÑK ÑM øÏKÖ …ÏhtI YÒ  \km ÙqF ñmsJ ÑMÏÑ  F ñTÒD ÕE …ÏU ~ybP á¬Ñ  B YÒÕiaA YÏ ÑËF †LÚLÖ LA KaA úsT žlt[ †LD YÏ ÑËF ñybP \Y ~JrÑXÖ ~ybP …ÏhtI YÒ  „fTρ ñnY cÚLA ( Ï qpStÒ úsTÖ fE KH cÏ Ò  ÑeKÏhtI YÒ  „ftÏI 1 (b8) καιÁ Π(praeter δ)Ψ : ταÁ δεÁ δ

3 (b10) ταÁ 1 Ψ e β FHIJ2 Λ Ar(103.7) : om. E b J1Qw

4 (b11) τωÄì 2 fort. Ψ, K δ FH Λ : om. E β IJQw, fort. Ar(103.11) εë αυτουÄ Π(praeter E )Ψ : om. E 1

1

5 (b12–13) ταÁ µεÁν υë ϕ’

10 (b19) τε αà ν E δ : τε fort. Ψ, K β FHIJ1Q : αà ν J2w

1 ante èÕqnF add. 〈…Ïhl …htB …htM øK〉 4 Bad

2 ÑM øÏK scripsi (= οÏσα Π = omnia CV) : ÑM …ÏW scripsi (= κινειÄσθαι Π = mouetur CV) : s.p. Φ : …U Bad KÏhtmLÒÖ scr. ˇ c 167.10) : …htmLÒÖ Φ 7 \M ÙHÒÖ øÏK scripsi (= εÏ καστον Π = unumquodque CV) : Bad (= G Ï \M ÙHÒÖ \M ÙHÒÖ øK Φ : \M ÙHÒÖ ÙHÒÖ øK Bad 10 úsT ... ~ybP supplevi ex marg. Φ (with correction mark ¯C = ΠCV) : om. Φ in textu

Φ Bad

5

2–3 gloss without mark i.m.:

…ÏW Öz …ÏH ñnM a‚J YÏ i  cz Φ i.m., fort. ex comm. Philop. (v.

Rashed 2016, p. 754 [Scholie 2, l. 5–6]); cf. also Simpl. 1207.4–6

Chapter 4

254b10

254b15

254b20

(iLÑPÕpSrt fÏ yibpLA û4LA êÑtK \M nMÑoLA LÑqmLA

30

3 Av. t. 28

6

9

~ybP …ÏhtT љyB ÑE¡V \M KÏhtmLÒÖ .ÕE ( Ï qpSz cÏ Ò  ñMÒÕQ YÕkI ‡ÒJi Ò KH øoB ÑeybP \Y ~JrÑX ÑMÏÒ  .ÑeybP \Y ~JrÑX …ÏhtT љyBÖ …ÏhtT ÑM ¡oK ~gIÒ  YÒÕiaA aÑgYÒ  Ö øfSÒ  ŠÒË rÑnLA KHÖ ƒÕF ŠÒË iÏGri Ò „Ö .яKH ÑeJ øbÂQ \MÖ ÑeyGÖ ŸÑnCÒ  øbÂQ \M ÑeybP \Y ~JrÑX  …htT LA aÑiWi Ò …ÏhtI ;NÏÒË …ÏhtmLA YÏ Â Ò   Í CÏÑX ezI ÑeybP \Y  ~JrÑX Ï ~JrÑX KÏhtmLA ÙyB \M vÏ .áMÒ  \È i® B á¡V \Y …ÏhtI ÑM YÏ i  ÑM a \Y  \M fE  …htT ;NÏÒË õbpLÑB KhtmLA \M YÑK ÑM ÑeybP \Y øoM ыÑqlT Ï Ï øE [ê212] áÚE MÒ  \M ‡[ cÚLA (iL ñNÏÒ  †LDÖ YÒÕiaA ŸÑnCÒ  \M …ÏdA YÒÕiaA „  lÏ\ YÒ  fvbnI žiK øB …ÏhtT a \Y 1 (b20) καιÁ 2 Π(praeter E1)ΨΩ : om. E1

(b21) κινειÄται E, fort. Ψ, K2 β γ : κινειÄσθαι K1 δ

3 (b22) καιÁ τοÁ Π(praeter δ)Ψ : τοÁ δεÁ δ τρο πους Π(praeter E )Ψ : τρο πος E 1

S (1208.18) c

E1)Ψ : om. E1 Sp(1208.28f.)

1

4 (b24) θε σεις Π(praeter E1)Ψ : θε σις E1

(b24)

(b24) τηÄ ς κινη σεως Π(praeter δ)Ψ : τωÄ ν κινη σεων δ

6 (b26) ειËναι Π : ειËναι τοÁ fort. Ψ

7 (b27) τωÄ ν καταÁ ϕυ σιν Π(praeter

8 (b28) ζωÄì α E, fort. Ψ(cf. Gutas Theophr. 7b4, 8b27) : ταÁ ζωÄì α K β γ δ 9 (b29) κινειÄται Π(praeter E1)Ψ : κινειÄ E1

1 KÏhtmLÒÖ scripsi (= καιÁ [καιÁ om. E1] τωÄ ν ... κινουµε νων Π = et eorum quae mouentur C, v. ˇ c 168.20) : KhtmLÑF Φ (= eorum vero quae mouentur V) Bad also G …htT scripsi : s.p.

Ï …htI Bad 4 ÑeJ øbÂQ \MÖ scripsi (= καιÁ [scil.   partes C (= øbÂQ \MÖ παραÁ ] τουÁ ς τρο πους Π[praeter E1]) : ÑeJ øcF \MÖ Φ Bad : et propter Ñil‚J) : aut propter partes V (= Ñil‚J øbÂQ \M Öz) 6 \M vÏ Φ (= ΠCV) : \Ê M v \M v  Bad    Φ:

…htI Bad

Ï

2 …ÏhtT scripsi : s.p. Φ :

ñiF úLÑvLA (vid. Φ ÑB : scripsi ÑM) ÑM cz; gloss 2: ÒÚE YÒÕiH ÑnEÑE YÏ z ¿F Φ i.m., fort ex comm. Philop.

1 two glosses without mark i.m., gloss 1:

ñilY rÑnLA blvL ~ybP ƒÕF ŠÒË …ÏhtI

(sic!)

829.15–18 (v. Rashed 2016, p. 756 [Scholie 4])

254b20

254b25

31

VIII 4, 254b20 – 255a7

Cp 183v Cv 71r 4

8 Av. t. 29

ñMÒÕQ ÑM lÑS „Ö \fsLA „ ;K YÕkI YÒ  ñbwI ñNÏÒ  †LDÖ .…ÏhtmLA ‹YÖ …ÏhtmLA \M ÆcfnM …ÏdA ~gIÒ  YÒÕiaA „ †LÚK ~iÏyibP (iL
ÑbLA Í CÏÑX †ÏwLA ñiF ÑM YÏ Ò  ÇÒË ñTÂÒD ÕE …ÏU á¬Ñ  B YÕkI e`A áÚE \M ÑKÏhtmLA YÏ ÒË ÑnlQ 〈 ÑNÏz〉 †LDÖ ¡XÒ  ÑEÑNKD LA msqLA \M …ÏhtT љyB YÏ Ò  õGÕI YÒ  biF ÑeybP \Y ~JrÑX …ÏhtT љyB ÑE¡V †LD èÑoM ÑM a \Y …ÏhtT NÑK YÊ ÒË †¬ wLA õQÕT áÚEÖ .~ybP ­Q ›lBÑqtmLA ›yGÕmLA ŠÒË …ÏhtT áÚE YÏ Ò  †LDÖ .liqoLÒÖ fifbA ŠÑËF øiqoLA ÑMÏÒ  Ö ƒÕF ŠÑËF žifbA ÑMÏÒ  ÑecÏ] LA ÑeyGÒÕM ŠÒË …ÏhtTÖ EÑZ ÕE ;K эF EÑzB (ilF …ÏhtT ÑM a \Y YÒ  ÑMÏÑ  F .~ybP øfSÒ  \M fE …ÏhtT ;NÏÒË ÑeNÏÑ  B èÕqLA YÏ Ò  †LDÖ .ÑeybP \Y ~JrÑX KÏW ÒDÒË ÒÖD   Ï [ a ÕEÖ YÒÕihlL ÕE ;NÏÒË !ymLA ÒÚE YÏ ÑËF èÑL ыÑqlT 1 (b30) γαÁ ρ Π(praeter EF)Ψ : γαÁ ρ cum ras. E : τα χ’ F1 : γαÁ ρ τα χ’ F2 εë αυτοÁ δ : Ψ = βγΛ ∨ δ : αυÆ τοÁ E K

E2)Ψ : exp. E2

Π(praeter E )Ψ : ειÆ E

2

Bad

5

øcfnM Φ Bad

3 ÕE s.l. Φ (= αυÆ τοÁ Π) : secl. Bad

…ÏhtT scripsi : s.p. Φ : …htI Bad YÑ  B Bad

enim V) : s.p. Φ :

1 gloss without mark i.m.:

YÑncLA yI

11

4 ÑnlQ 〈ÑNÏz〉 †LDÖ scripsi

ÑnlQ †LDÖ Φ : ÑnlQ †LD

YÏ ÑËF scripsi (= τε γαÁ ρ Π = quoniam C =

Φ i.m.; cf. Alex. frag. 583, Simpl. 1208.34–37

11 YÒÕihlL with corresp. marks s.l. and i.m.: i.m.

8 (a3) ειÆ ς

9 (a4–5) υë ποÁ τι νος Π : υë πο τινος interpr. Ar.

(= γαÁ ρ ... εÆ θη καµεν Π = quod est quia diximus C = diximus enim V) : 1et2

(a2) οιÎον Π(praeter

7 (a2) ταυÄ τα γαÁ ρ α K δ NR HIJw Λ : om. L : ταυÄ τα γαÁ ρ αà ν FQ

2

2 ÆcfnM scripsi :

5 (b35) τιθε ναι E Ψ : τιθειναι (?)E1 : αÆ ντιτιθειÄναι K :

6 (a1–2) υë ποÁ τι νος Π : υë πο τινος interpr. Ar.

αÆ ντιθειÄναι β γ δ

3 (b32) αυë τοÁ β γ Λ :

2

\Ï zI .F ×qF YÒÕiaA   Ï ] èÕqLA ÒÚE LÑHÒË YÏ z yI Φ

254b30

254b35 255a

255a5

(iLÑPÕpSrt fÏ yibpLA û4LA êÑtK \M nMÑoLA LÑqmLA

32

Cp 184r 5 Av. t. 30

Ca 25va

10

YÒ  Ö ,ыÑqlT \M žqT YÒ  юkm YÑK ÙqL †LÚK MÒ  YÑK ÕLÖ (fNi Ò úbS ~gIÒ  ÕeF mLA „ ñTÒÚL ÕE ~bbS aLA YÑK YÊ ÒË ñNÏÒ  øoM †LÚB rÑnLA ŠÒË ƒÕF ŠÒË rÑnLA ÕÏE „ Mi Ò YÑK YÊ ÒË †LÚlF mLA …T „ ñTÒÚL aÑiWÒ  YÏ Ñ  B ~gIÒ  èÕqLÒÖ .øfSÒ  ŠÒË êÕSLA ~gIÒ  rÑnLA ŠÒË YÏ Ò  \iÏbLA \B …ÏW fE NÑK YÒË %ÑqnM ¡V ÙqF ÙHÒÖ KH …ÏhtT ыÑqlT \M KÏhtM ñTÒD ÕE …ÏU øKÑwtM øctÏM a YÕkI YÒ  \km žiK ~gIÒ  Ö .щÒÖD ¡V ÕE e`A áÚE \B aÑqlLÑB i øctÏM ÙHÒÖ ÕE ÑM eJ \M YÏ Ò  †LDÖ YÒ  ñN{W \M ñgyB YÕkI YÒ  ƒ£fM ñNÏÒ  eJ \M 〈 ÕE 〉Ö ¡O{tlL øBÑQ …ÏU ÆCÒ  DÒË áÚE \M a (ilF .øbqI YÒ  ñN{W \M ñgyBÖ øyfI úO ÙQ øB øctÏM ÕE ÑJÏ ÆCÒ  ÑE¡V a iÖ ÆKÑwtM YÑK DÒË ñTÒD ÕE 1 (a7) αυÆ τουÄ αυÆ τοÁ E1 : αυÆ τωÄì αυÆ τοÁ E2 : εë αυταÁ e2 Pa(374.20) : αυë ταÁ I : Ψ = e2 ∨ I : αυÆ τοÁ εë αυτοÁ K b3 : αυÆ τοÁ εë αυτω Äì b2 [om. b1] : αυÆ τω Äì εë αυτα e1 : αυÆ ταÁ γ FHJQw : αυÆ ταÁ αυë τα scr. Ross

2 (a8) αυë τωÄì FIJ : αυÆ τωÄì E Ke Qw Λ : αυÆ τοÁ b β H : αυÆ τοÁ αυë τωÄì Ψ E2)Ψ : τοÁ E2

(a9) ειÆ α K β δ Λ Sp(1210.7) Pa(374.21) : εÆ πειÁ γ

αυë τω Äì K : αυÆ τοÁ J

3 (a8) τουÄ 2 Π(praeter (a9) αυÆ τωÄì α β δ FHIQw :

4 (a10) καιÁ τοÁ Ψ K β γ δ : καιÁ E1 : τοÁ καιÁ E2 : τοÁ Λ

αÆ παθε ς ... ταυ τηì Π(praeter E )Ψ : om. hom. E 1

1

7–8 (a13–14)

10 (a15) συµϕυηÁ ς E : συµϕυεÁς be1 β :

συµϕυηÁ ς vel συµϕυεÁς Ψ : συµϕυηÄ K e γ Λ 2

2 ~bbS scripsi (= αιÍτιον Π = causa CV) : s.p. Φ :

ÑmÍ iW Bad ÕeF scripsi (cf. est etiam CV = ÕeF vel ÕE) : ÕEÖ Φ Bad 3 †LÚlF Φ (= ΠCV) : †LÚkF Bad 8 eJ \M 〈ÕE〉Ö scripsi (= αÆ λλ’ ηÎì Π [om. E1] = sed inquantum ... est C = et secundum ... erit V) : ñJÖ \MÖ Φ Bad ƒ£fM Φ : ƒ£fI Bad 1YÒ  scripsi (= ταυ τηì Π [om. E1] = est C = tale quod V) : ÒDÒ Φ : ÒDÒË Bad 9 ñgyBÖ scr. Bad : љyBÖ Φ DÒË (= αÍ ρα Π) secl. Bad …ÏU Φ (= ΠCV) : ~KL Bad 6 øKÑwtM with corresp. marks s.l. and i.m.: ûÑbpLA „ftÏM Φ i.m., fort. ex comm. Alex., v. Simpl. 1210.14–34 and Alex. frag. 586

10

øctÏM ÕE with mark s.l. and (without corresp. mark):

a҂Ji Ò ñBÑwtM cz Φ i.m., fort. ex comm. Ibn al-Tayyib 842.20f. ˙

255a10

255a15

33

VIII 4, 255a7–27

Av. t. 31 3

6 Cp 184v

9

ÑM øoM …ÏhtmLA \M ‚iÏmtM юM ÙHÒÖ øÏK „ …ÏdA YÕkI YÒ  Í rÖ² ÙQ ñNÏÒ  ¡Â V .(fNi Ò ÒÖD \M a ÑeKÏH "M Ñ= (fNÒ  i .F áÙP †LD ezI ÙQÖ .ÑM a \Y ÙBÒ  …ÏhtT ;NÏÒË YÕkT YÒ  ~gIÒ  áÚE „ ‡‚lI .êÑbSi Ò qT YÑ  B YÏ Ò  †LDÖ ÑEÑNKD LA ÑymLA †lT ~gIÒ  ÑKÏdA „ ÚX{N YÒ  ÑnLÖ „injÂnÊmÂLA ˆeÊS †IW YÏ Ò  †LD èÑoM õbpLA \Y Í JrÑX KÏL fE љyB 〈 rÑhlL 〉 …L øyfLÑB rÑaA Y Ï Ï ¬ Ï Ò  †LD èÑoM ~ybP љyBÖ ~iÏyibP (iL øqolL ~gIÒ  †LÚKÖ .ÑeŒWÒ  ÑJÏ áÚE ¡V „ Mi Ò cO èÑomLA ÒÚE ‹YÖ ÕÏqLÑB ÒDÒË [Ò213] ÓiT ÖÒ  ˆÏK ÖÒ  žiK ÕE ÕÏqLÑB cÚLA ÕE ~ybP …ÏhtmLA YÑËF ÙQ ñNÏÑËF .¿yLA „IpB i ñiF gcA ÒÚE cO cÚLA zÙbmLA ñL YÑK „IpB XwL YÕkI ;NÏÒË ;EÙHÒ  YÏ Ò  i¬ ÒË Ñm­ KÖ ~fiK ÙHÒÕLA aLA YÕkI 2 (a18) εÆ µψυ χων αυÆ ταÁ Π(praeter E)Ψ : εÆ µψυ χων E1, sed αυÆ ταÁ add. s.l. E2

10 (a26) αυÆ τωÄì α

K β F : εë αυτωÄì δ HIJQw Λ Pp(830.10) Sc(1212.2) Pa(375.18) : αυë τωÄì scr. Ross non vert. Ar.

(a26) καιÁ ]

11 (a27) ποιοÁ ν ... ποσο ν α K β HQ : inv. δ FIJw Λ

1 …ÏhtmLA scripsi (= τοÁ κινου µενον Π = eo quod mouetur C = moto V) :

…dA Φ Bad 2 i Ñ= (fNÒ  scripsi (= τωÄ ν αÆ ψυ χων Π = inanimatis C = carentibus anima V) : ñL (fNÒ  i Φ : i ñL (  fN Bad 7 rÏÑhlL addidi (= τουÄ δυνα µει θερµουÄ Π = calidum potentia C = calidum in potentia V) : om. Φ, Bad

3 áÚE „ with corresp. marks s.l. and i.m.: ~ybP KÏW YzÖ (Φ msqnmLA : scripsi) sfÏntmLA ¡V yI

Φ i.m., fort. ex comm. Alex., v. Alex. frag. 587 (cf. p. LXXIV), Simpl. 1210.38–1211.2

5 ÑEÑNKD

LA

1211.15–18

with corresp. marks s.l. and i.m.:

MÙÏqtmLA msqLA yI

Φ i.m.; cf. Simpl.

255a20

255a25

(iLÑPÕpSrt fÏ yibpLA û4LA êÑtK \M nMÑoLA LÑqmLA

34

Cv 71v Av. t. 32 4

Cp 185r 8

ÑtKÏW ÒDÒË ­Q ÑMÏÒË ÑM a \Y YÑKÏhtT ¿ri ÒÖ rÑnLÑF .ñTÒÚB i ¿yLA .ÕÏqLÑB ;= LA ;=ÑyFÒ  ŠÒË ÑtKÏW ÒDÒË ~ybP ÑMÏÒËÖ ;eybP \Y ~JrÑX YÕkI Yi  ~bbS †LD rÑC htÏW aÑVÒ  ‹Y èÑqI³ ÕÏqLÑB ÑM YÑK ;LÖ èÑoM EÑZ ¡V ÕE a cÏ Ò  ;eŒWÒ  ÑMÖ áÚE …ÏhtT ñnY cÚLA aLA GÑY ÕÏqLÑB ÕE ˆlÏytmLÑF .øfSÒ  ŠÒË ¿ri ÒÖ ƒÕF ŠÒË …ÏhtT rÑnLA YÏ Ò  †LD ø³mÂyÊI³ i ÕEÖ ˆlÊyLA 〈 ñL cÚLA GÑyLA 〉 ñB cÚLA ñJÕLA ¡V ÑM ñJÕB øYÑfLA ÙBÒ  õmtJ ÒDÒË øyfLÑB rÑC ;BÏr ÕÏqLÑB cÚLÒÖ .ÕÏqLÑB GÈ ÑY ÕÏqLÑB X| ~miW ÕÏqLÑB ñilY YÑK ÑmÏY ¡cI ˆlÏytmLA YÏ Ò  †LD èÑoM øBÑqLÒÖ 1 (a28) καιÁ 1] non vert. Ar. c

S (1212.30)

2 (a30) αυë τωÄ ν E FHJQ : αυÆ τωÄ ν Iw, interpr. Ar. : εë αυτωÄ ν K β δ

4 (a32) ταÁ τοιαυÄ τα] ταυÄ τα καιÁ τοιαυÄ τα fort. Ψ

6 (a34) εÆ νεργωÄ ν E, fort. Ψ(=

Φ [v. App. 2]), fort. S (1213.32), Pa(375.27) : θεωρωÄ ν K β γ δ (= Φ2[v. App. 2]); cf. T(app. 1

p

crit. ad 219.12)

7 (a35) γι (γ)νεται εÆ νι οτε α K β δ HJQw As(frag. 592) Sc(1214.10) :

γι γνεται F γρ.A(apud S 1214.10f.), Ross Wag

(a34) αÆ ειÁ δ’ οÏταν Π : ειÆ δ’ οÏταν propos.

Hayduck, p. 6 : οÏταν (nisi ειÆ ) δ’ αÆ ειÁ intell. Ar. (cf. WiCo 312, Ross 696 ad loc.)

8 (a35)

τοÁ E , fort. Ψ, K β γ : om. E δ S (1214.9) 1

2

1 ÑM Φ (= ΠCV) : om. Bad

1

c

4 ;eŒWÒ  Φ : ista CV :

ÑeŒWÒ  Bad ÕE a cÏ Ò  scripsi (= τι νος 6 ñL cÚLA GÑyLA addidi (= οë εÍ χων ηÍ δη Π = Π) : ÕE a ŠÒË Φ Bad : sit ad rem C [om. V] sciens cui est scientia C = illud quod habet scientiam V) : om. hom. Φ post ˆlyLA suppl. †lm \M Bad ømyI i Φ (ante corr.) (= µηÁ εÆ νεργωÄ ν [v. App. 1] = non operatur C) : ˆlyI i + ñB s.l. Φ (post corr.) (= µηÁ θεωρωÄ ν [v. App. 1] = non scit V)

;=ÑyFÒ  with corresp. marks s.l. and i.m.: юKÑMz yI Φ i.m. = Alex. frag. 590 (cf. Simpl. 1213.3f. and Rashed p. 524f.) 8 gloss with mark i.m. (without corresp. mark in textu): yI iNÑoLA ÕÏqLÑB Φ i.m.; cf. Them. 219.23f., Simpl. 1213.29–1214.6, Ibn al-Tayyib 843.2–6 ˙ 2

255a30

255a35 255b

35

VIII 4, 255a28-b12

3

6

9

(iLÖ ÑM ñJÖ ‹Y ÕÏqLÑB ˆlyI ÕE ;NÏÑËF ˆlyI i ÕEÖ ˆlyLA ñL cÚLA YÏ Ò  †LDÖ rÑC ÒDÑËF .ˆlÏytI YÒ  øbÊQ \M ñilY YÑK cÚLA ñJÕLA ÕE ÕÏqLA \M ñJÕLA ÒÚE YÒ  ‡‚L ÇÒËÖ õNÑM ñynm G ÑM ÚmniH ˆlyIÖ ñlÂyF øyfI ñNÏÑËF èÑaA áچ „ MtÒ cO èÑomLA ÒÚE ‹YÖ .øe`A ÕEÖ ˆlyLA ¡iqN „ YÕkI rÑN ÚmniH rÑC iÏvT ÒDÑËF rÏÑH ÕÏqLÑB ÕE drÑbLA YÏ Ò  †LDÖ ~gIÒ  iÏyibpLA rÕMtÒ žifbA „ Mi Ò cO †LÚKÖ .„lÑY ñqyI ÖÒ  õNÑM ñynm G ÑM ƒHÒ  Ö ñNÏÑËF aÒÕE a;LA \M lQ †NÏÑ  K øiqoLA \M YÕÏktI ÙQ ñNÏÒ  †LDÖ ~gIÒ  øiqoLÒÖ G ÑM ñlÂyF YÑkmLA ‹Y øyfIÖ ~fifX ÚmniH ¡cIÖ ÕÏqLÑB ÒÚE iÍ ÖÏÒ  YÑK õnmLA ñL õqI ;NÏÒËÖ ƒ³ÕFÂÖ ÑM ÓiT YÕkI YÒ  žifbA øyFÖ .õNÑM ñynm 3 (b4) θεωρειÄ] an θεωρειÄ ηÍ δη Ψ?, cf. Sc(1214.25) l

c

T(219.5) P (830.31) S (1214.16.25) Pa(376.6)

(b4) τι µηÁ E β γ Λ : µηÁ τι fort. Ψ, K δ 4 (b5) καιÁ αÆ γνοι αì E1, fort. Ψ, K FIw (cf.

Pa[376.7–8]) : καιÁ εÆ ν αÆ γνοι αì E2 β δ HQ2 γρ.S(1214.35) : καιÁ ουÆ κ εÆ ν αÆ γνοι αì JQ1 : καιÁ ουÆ χ αë πλω Ä ς εÆ ν αÆ γνοι αì A(S[1214.27]) (cf. Rashed, p. 528f.)

E1 : ηÍ δη cum lac. F

5 (b7) ηÍ δη Π(praeter E1F)Ψ : δη

6 (b7) και ει Π(praeter E1)Ψ : καιÁ ηËì E1

γι (γ)νεται Π(praeter E) : γι γνεται E : γι γνεται γαÁ ρ fort. Ψ

7 (b8–9) τοÁ γαÁ ρ κουÄ ϕον

8 (b9–10) τουÄ το γαÁ ρ δυνα µει

πρω Ä τον καιÁ ηÍ δη κουÄ ϕον i’punxit Ar. (cf. As[frag. 594] et Rashed, p. 530)

Π(praeter EN)Ψ : τοπου E : το πον E : τοÁ πουÁ cum lac. N 1

2

9 (b11) τοÁ πουÁ

(b12) δ’ vel δε Π(praeter

E)Ψ : om. E 1 ÕÏqLÑB ˆlyI ÕE scripsi (= δυνα µει [δυνα µει om. F] εÆ στιÁ ν εÆ πιστη µων Π = est sciens in potentia

ÕqLÑB ÕE ;NÏÑËF ˆlyI i ÕE Φ Bad : est sciens C 3–6 „lÑY ... õNÑM Φ (= ΠCV) : „lÑY ñqy³I ÖÒ  drÑbLÑF iyibpLA rÕMi Ò † Mi Ò gO †LÚKÖ øe`A èÑH † YÕkI YÒ  ñilY rÙqI ÑM ¡QÑnT èÑH † YÑkL iÒËÖ〉 〈„lÑY ÖÒ  õNÑM Ñeynm G ÑM ƒW áÚEÖ rÑnLA † ÕqLÑB YÕkI ¡vtLA ÙyBÖ rÑaA † ÕqLÑB dÕJÕM scr. Bad 9 ƒ³ÕFÂ Ö scripsi (= ΠCV) : ÑQÕQÖ Φ : ~FÕQÖ Bad V) :

9 gloss without mark: øyfLÑB fE áÚEÖ Φ i.m., fort. ex comm. Philop. 831.8–11 (NB: this gloss has been blended by Bad with the foll. one, which relates to 255b18 (v. p. 36, App. 3)

255b5

255b10

36

Cp 185v 3

6

(iLÑPÕpSrt fÏ yibpLA û4LA êÑtK \M nMÑoLA LÑqmLA

ˆÏkLA „ èÑomLA ÒÚE ‹Y cO !ymLA ÒÚEÖ .ñyGÕM ÙÏG „ YÑK ÒDÒË .žikLA „Ö ~gIÒ  ŠÒË …ÏhtT ;NÏÒË fifbA rÑC ˆÂL ñtFyM úlÂpT³ ÑJÏ ÒÚE YÏ Ò  ‹Y  .liqoLA †LÚKÖ ñniyB ÙHÒÖ õGÕM YÒ  ‹Y YÕbpM ÑeNÏÒ  †LD „ úbsLÒÖ ƒÕfB †LD RÑiV YÒ  ÒÚE fE øiqoLÒÖ žifbA iNÏÒ  Ö ÑM ÓiT YÕkT htÏW aÑVz ‹Y YÑNÕkI ÕÏqLÑB øiqoLÒÖ žifbA YÏ Ò  i¬ ÒË øfSÑ  B ÒÚE RÑiVÖ ÒDÒËÖ ÑM ñJÖ ‹Y žifX ÕÏqLÑB ÕeF ÐÑM ‡Òd ÑM a;LA YÏ Ò  †LDÖ .ÑnlQ ;K i ñNÏÒË htÍ Ï H „lÑY 〉 ñQÕyI YÒ  \km ÙqF .ÕÏqLÑB Ù³yB ñNÏÑ  kF ÐÒÕE rÑC 1 (b12) εÆ ν Π(praeter KF)Ψ : ηËì K : om. F

3 (b14) κινειÄται Π(praeter δ)Ψ : ουÆ κινειÄται δ

4 (b14) αυë τωÄ ν K β F2HJQ2 Λ : αυÆ τωÄ ν Iw : αυÆ τοÁ ν α δ : αυÆ τουÄ F1 (?)Q1 Pa(376.21, mss. LCD), intell. Ar., cf. S(1216.30)

(b15) τοÁ Π(praeter δ)Ψ : τωÄì δ

(b16) τοÁ µεÁν τωÄì ... τοÁ δεÁ τωÄì Π(praeter β)Ψ : τωÄì µεÁν τοÁ ... τωÄì δεÁ τοÁ β

Λ : om. Ψ K1 δ H Pa(376.23)

*

5 (b15) ποι] που

7 (b18) τε E K2 β FIJQw

8 (b19) ωë ς EΨ K be1 : om. e2 β γ Λ

(b19) γαÁ ρ Π(praeter

E) : om. E : an δεÁ Ψ? (cf. Pa[376.25]) 1 cO s.p. Φ (= currit C = est cursus V) :

gÕU Bad 3 úlÂpT³] s.p. Φ, an (ñtÂFyM) úlpN leg.? 5 iNÏÒ  Ö] ñiBÒÖ Φ : iN|Ö Bad 6 YÑNÕkI scripsi : YÕkI Φ 8 ñNÏÑ  kF scripsi (= ωë ς) : ñNjF Φ : tunc dum C [ÕqLÑB ... ñNÏÑ  kF om./non vert. V] : ñNÏj F Bad (for Isha¯q’s use of ñNÏÑ  K = ωë ς cf. Gutas ˙ Theophr. 8a18, Eth. Nic. 1125a29, Phys. 228b16, 18 = Bad 566.17, 567.2) 8-p. 37.1 „lÑY ñQÕyI ... addidi (= εÆ µποδιζο µενον µηÁ αÍ νω ειËναι αÆ λλ’ εÆ αÁ ν [αà ν K] αÆ ϕαιρεθηÄì τοÁ εÆ µποδι ζον Π = impediat ipsum aliquod quare [quare CpCa : quin Cv] non est sursum at vero si remoueatur ab eo quod impedit ipsum [impedit ipsum Ca : impedit Cv : efficit ipsum Cp] C = impediatur quin sit superius sed si illud impedimentum auferetur V) : om. hom. Φ 7 gloss without mark i.m.: ÒDÒË ÑLAË %ÑiqLÑB ÕÏqLÑB fEÖ qiϏLA ŠÒË %ÑiqLÑB rÕcLA fEÖ iNÑoLA ÕÏqLÑB ÙII

ÑelyF юY rÙC Φ i.m.; cf. Simpl. 1216.33–39

255b15

255b20

37

VIII 4, 255b12–33

3 Ca 25vb Cp 186r 6

Cv 72r 9 Av. t. 33

‹YÖ .‹YÒ  ÙBÒ  YÑKÖ ñlÂyÊF øÂyÂF 〈 ñQÕyI ÑM ñnY õFr³ YÒË ñNÏÒ  i¬ ÒË ƒÕF YÕkI ‹Y ˆLÑyLA YÏ Ò  †LDÖ øyfLÑB YÕkI ÑM ŠÒË ~gIÒ  žikLA iÏvI ÙQ èÑomLA ÒÚE  .*** õNÑM ñynm G ÑM ˆlyI YÕkI YÑkmLA (iL eJ \MÖ …³ÏdA eJ \M ñNÏÑËF õNÑM ÖÒ  øMÑa øI‚m³LA ÑMÏÑ  F Í \M Í NÒÕpSz hV Í Ï [ê 213] Í ~NÑsNÒË YÏ Ò  Í ÕL †LD èÑoM …dA aÑnB W Ï ÕE ;NÏÒË ÒÚE †IW YÑkL a;LA „ ƒÏR ‹Y ~YÕGÕM jH õFr ÖÒ  эlY ×lÑaA (iL ×lÑaA ц f Mr³ ÒDÒË Í yJÒr hqÏltT LA kLA YÏ Ò  ;K ¿yLÑB øB ñTÒD …ÏU ÕE a áÚE \M (iL ñNÏÒ  YÑB ÙqF .ц fMҁLA øB ÑeKÏH [4ê213].øyfLA øbqI Yi  øB øyfI ÖÒ  …ÏU Yi  i KH zÙbM ñiF ~ybP …ÏhtT YÒ  ÑMÏÒË ÑelÏK KÏhtmLA aÑiWi Ò NÑK DÑËF [4 Ò 214] ÑeybP \Y ~JrÑXÖ ­Q ÑKÏhtmLA NÑKÖ ­QÖ ÑeybP \Y ~JrÑX ÑMÏÒËÖ 3 (b23–24) καιÁ τοÁ ποσοÁ ν ... κωλυ ηì ] om. hom. Ψ 256a2)

6 (b26) εÆ ν Π(praeter E)Ψ : εÆ ν δεÁ E

Π(praeter E )Ψ : οë E 1

1

9 (b30) ουÆ

11 (b32) καιÁ α β γ : καιÁ ταÁ K δ, fort. Pa(377.18) 3

DÑËF add. (iLÑPÕpSrÒ èÑQ %dÑsLA ylytLA Φ Bad DÑËF Φ (= cum C = et cum V) : ÒDÑËF Bad

3 ***] Φ = CV v. App. 1

ΠCV : seclusi

4 (b24–25) κινη σας] fort. λυ σας Ψ (cf. 7 (b27) καιÁ non vert. Ar.

10 ante

: om.

…³ÏdA eJ \M (without mark): ¿yLÑB cz Φ i.m.; cf. Alex. frag. 597, Simpl. …ÏdA (iL (without mark): ~KÏL ÕE (iL ÒÚLÑB ÑM eJ \M cz Φ i.m.; cf. Simpl. 1217.19f. 5–7 gloss without mark i.m.: (iL ƒÏ ‚LA ûÑfTrÖ aÑnbLA ØÕqS YÏ ÒË cz ;eybP gK ‹Y ÑKÏW ;ŽY „ÂlÒÕyLA …ÒD èÒRÒ  ;L øB (leg. ÒÚLÑB?) …ÒD \M KH Φ i.m.; cf. Simpl. 1217.19–23 6 gloss next to ƒÏR ‹Y (without mark): ~XÕfnM ƒÏ‚LA ˆefI Yz fvbnI Φ i.m.; cf. Simpl. 1217.14 (τωÄì πεϕυσηµε νωì αÆ σκωÄì ) 8 áÚE \M] + sfÏntmLA ¡V \M cz Φ s.l., fort. ex comm. 4 gloss next to 1217.19f.

Í next to 4–5 gloss

Alex.; cf. Alex. frag. 598, Simpl. 1218.20ff., and Rashed, p. 532f.

255b25

255b30

38

5 Cp 186v

10

(iLÑPÕpSrt fÏ yibpLA û4LA êÑtK \M nMÑoLA LÑqmLA

ÑM ~ybP ~gIÒ  ÑKÏhtmLA NÑKÖ á¡V \YÖ ÑM a \Y …ÏhtT ;NÏÒË ÑelÏK \M \kI G ÑM †LÚKÖ …ÏhtI ÑM a \yF ñlÑqlT \M ~KÏhtM юM YÑK YÕÏkmLA \Y ÑMÏÒË …ÏhtT ÑeNÏÒ  †LDÖ) liqoLÒÖ fifbA øÂoM …ÏhtI ñlÑqlT YÒ  úJÒÕF (õNÒÕmLÒÖ „lÒÕylL øI‚m³LA \Y ÖÒ  ÆiqO ÖÒ  ~fifX øYÑfLA ÖÒ  .ÑM a \Y …ÏhtT ;NÏÒË ÑelÏK ÑKÏhtmLA YÕkT øbÂQ \M ÕE (iL YÕkI YÒ  ÑMÏÒË …ÏdA YÏ Ò  †LDÖ YÑB² ñL ÒÚEÖ Â Â YÒ  ÑMÏÒËÖ ÕE …U ñKU á¡Â V YÏ Ò  øbÂQ \M øB …htmLA …U ñsfN YÕkI Ï Ï Ï Ï Â Â ¡Xi Ò ÙyB \M èÖÏi Ò ÕE YÕkI YÒ  ÑMÏÒË …ÏdA ÒÚEÖ .ñsfN øbÂQ \M …ÏU   YÒ  ÑMÏÒËÖ ÂjaA …Â Ï U RÂÑkÏy³LA YÏ Ò  †LD èÑoM ÙHÒÖ \M ¤³ KÒ  ×SÏÕtI YÕkI ñtKH (ilF YÑsNiË Ò ÑMÏ{F Y³ ÑsNiË Ò ÑeKÏU ÙiLÒÖ ÙiLA \Y …ÏhtI R³ÑkÏy³LÒÖ \M èÖÏi ÒÖ ¡Xi Ò YÒ  YÑKÏU ;eNÏÒË ~yi? ;F èÕqN \VÖ .á¡V \Y 2 (b34) υë ϕ’ αυë τωÄ ν Π(praeter E)Ψ : υë π’ αυÆ τωÄ ν E

4 (a2) ηà Π(praeter E1)Ψ : om. E1

4–5 (a3) αà ν ... κινοιÄτο E K δ be T(220.2–3) S (1219.12–13) : αÍ ρα ... κινοιÄτο fort. Ψ, L[κιl

νειÄται]NR : αÍ ρα ... αà ν κινοιÄτο γ αυÆ τοÁ b Λ

7 (a4) αυë τοÁ E e β FH : εë αυτοÁ K IJQw : αυë τοÁ vel εë αυτοÁ Ψ :

8 (a5) αυë τοÁ E β γ δ : εë αυτοÁ K Sc(1222.31) : αυë τοÁ vel εë αυτοÁ Ψ : αυÆ το Λ, scr.

Ross

ÑelÏK scripsi (= πα ντα Π = omnes C = omnia V) : ÑelÏKÖ Φ : ÑelÏkF Bad (cf. Gp 174.16f.) 6 ante ÒÚEÖ add. 〈ñTÑbO ; èÖtÒ …dA rÖ²〉 Bad : om. ΠCV YÑB² ... ÒÚEÖ scripsi (= τουÄ το δεÁ διχω Ä ς Π) : et hiis quidem sunt duo modi C : et hoc duobus modis V : YÑB³LA YÒÚEÖ vid. Φ, 2 Bad 7 1…ÏU] …U Φ (= mouet CV) : …htI Bad …ÏU] s.p. Φ : mouet Ca post corr. : p v a mouetur C C , C ante corr. : om. V : …W Bad 8 …ÏU scripsi (= mouet C = moueat V) : …W Φ Bad 11 post á¡V ñKÏU addendum (nisi …ÏhtI pro ñtKH legendum)? (cf. τωÆì υë π’ αÍ λλου 1

κινειÄσθαι Π; homo autem non movetur ab alio V)

4 gloss without mark i.m.:

øqO KH Öz fÏX KH яKH NÑK aÒÕS cz Φ i.m.

255b35 256a

Chapter 5 256a5

39

VIII 4, 255b33 – 5, 256a21

3 Av. t. 34

6

Cp 187r 9

¡Xi Ò i ¡Â Xi Ò …ÏU ÕE ñNÏÒ  †LDÖ èÖÏi Ò †LÚB „¬ Hi Ò \Ï kL ÑKÏdA ÕlX …ÏU ñNÏÑËF èÖÏi Ò ÑMÏÑ  F èÖÏi Ò \M ÕlX ¡³ Xi Ò …ÏU (iLÖ èÂÖÏi Ò …ÏU .Y³ ÑsNiË Ò ñKÏU G ÑM …ÏU i RÂÑkÏy³LA YÏ Ò  †LD èÑoM .¡Xi Ò \M YÒ  Ö …ÏhtI ÑM a \yF …ÏhtM øÏK YÕkI YÒ  Í rÖ² ~bJÒÖ YÑK YÑËF NÑK YÒËÖ …ÏhtM ¡V ÕEÖ ÑMÏÒËÖ …ÏhtI ÕEÖ ÑMÏÒË á¡V \M ñtKH YÕkT …ÏhtI i èÖÏÒ  …ÏL ÑnEÑE YÕkI YÒ  úJÒÕF …ÏhtM ÕEÖ á¡V \M ñtKH Mi Ò YÕkI YÒ  fcLA áچ èÖÏi Ò YÑK YÒË úO (iL YÑKÖ á¡V \Y á¡V \M ÙBÒ  ~KÏhtM …ÏdA Iш jB Ïm YÒ  \km i ñNÏÑËF ~bJÒÖ XvÒ …ÏhtM øÏK YÑK YÑËF .ÆCÒ  èÖÏÒ  Ñ= (iL Iш jB LA aÑiWi Ò YÏ Ò  †LDÖ \Y …ÏhtI (iL ñNÏÒ  i¬ ÒË …ÏhtI èÖÏi Ò …ÏdA YÑKÖ …ÏhtI ÑM a \yF .ñlÑqlT \M ÕE …ÏhtI ;NÒË YÕkI YÒ  Í rÖ² úO ÙqF á¡V 1 (a10) εÆ κειÄνο Π(praeter E1)Ψ : εÆ κει νωì E1 κινουµε νη υë ποÁ FIJQw Λ

Ar.

3 (a13) κινουÄ ντος E, fort. Ψ, K β δ HB :

4 (a14) τε E K β δ HIJw : αυÆ τοÁ FQ : om. Pa(378.16) : non vert.

4–5 (a14–15) καιÁ ηà υë ποÁ κινου µενου υë π’ αÍ λλου E1 K F2HIJQw Λ : καιÁ ηà υë ποÁ τουÄ

κινου µενου υë π’ αÍ λλου E2 δ F1 : καιÁ ηà υë πο τινος κινου µενου υë π’ αÍ λλου β : καιÁ υë π’ αÍ λλου ηà κινουµε νου fort. Ψ (cf. App. 3 ad loc.)

Pa[378.17])Ψ : secl. Ross Pel

6 (a15) κινουµε νου Π(κινουµε νον F1, cf.

7–8 (a17) θα τερον Π(praeter δ) : θα τερον ειËναι δ, an Ψ?

8 (a18) καιÁ E K β S (1222.10) : καιÁ τοÁ E2γ δ : αÆ ειÁ Ψ 1

1

p

„¬ Hi Ò scr. Bad : „HjLA Φ

Bad

5 YÒËÖ Φ (= ΠCV) :

YÒË Bad

7 YÑK YÒË Φ (= ΠCV) :

YÕkI YÒ 

RÑkÏyLÑK ÑpSÏÕtmLA ¡Xi ÑB yI Φ i.m., fort. ex comm. Philop. …ÏhtM ¡V with corresp. marks s.l. and i.m.: …ÏhtM á¡V \M ˆeFÒ Φ i.m.; probably erroneously relating a corrective note on …ÏhtI ÕEÖ ... YÕkT YÒ  Ö (l. 4–5 = καιÁ υë π’

1 gloss without mark i.m.: 831.35–832.1

5

αÍ λλου ηà κινουµε νου instead of καιÁ ηà υë ποÁ κινουµε νου υë π’ αÍ λλου [v. App. 1]) to the present

transl. of ηà µη [256a15]; cf. Alex. frag. 600, and Rashed ad loc.

256a10

256a15

256a20

(iLÑPÕpSrt fÏ yibpLA û4LA êÑtK \M nMÑoLA LÑqmLA

40

Av. t. 35

3

6

Cp 187v 9

…ÏL øÏK ñJÕLA ÒÚE ‹Y ñniyB è³ÕqLA ÒÚE \ i¬ bÂI³ YÒ  ~gIÒ  \km ÙQÖ [ê 214] YÕkI YÒ  ÑMÏÒË …ÏdA YÏ Ò  †LDÖ .aB …ÏUÖ ~miW …ÏU ñNÏÑËF ÑMÏÒË YÑsNiË Ò YÏ Ò  †LD èÑoM .á¡vB …ÏU YÕkI YÒ  ÑMÏÒËÖ ñs³fN ÏLA …ÏU ¯ILA ÑMÏÒË ÏLA fqlÊI³ ÙQÖ RÑkÏ y³B …ÏU YÕkI YÒ  ÑMÏÒËÖ …ÏU ÕE YÕkI YÒ  …ÏU ñB ÑM …ÏU YÒ   \km (iLÖ .ñtyFd fE cÚLA jaA ÑMÏÒËÖ Ñ“fN YÒ  ~bJÒÖ (ilF ñTÒD …ÏU ÕE a YÑK YÒË øB ñTÒD …ÏL ÕE ÑJÏ ÕlX LÑL i …ÑneF …ÏU ñB X| a YÑK YÒËÖ .…ÏU ñB X| a ñyM YÕkI YÑËF .Iш jB †LD gd;T i¬ ÒËÖ ñsfnB øB aB i ~gIÒ  …ÏL a jB gd0I iÖ †LD žqI YÒ  úO ÙqF ÑM a ñKÏU …ÏhtmLA DÒË YÑK 2 (a22) γαÁ ρ Π(praeter K)Ψ : om. K Pa(379.2)

(a22) καιÁ Π(praeter E1)Ψ : καιÁ κεινει καιÁ

3 (a23) αυÆ τωÄì E δ L (?)R JQ1 : αυë τωÄì fort. Ψ(v. App. 2 et 3), N FHIQ2 Λ : αυÆ τοÁ w

E1

6 (a25) τουÄ αυÆ τοÁ E2Ψ K β δ FHJQ Λ : του του E1 : τουÄ αυÆ τοÁ ς I : αυÆ το ς w

(a25) αυë τωÄì fort.

Ψ, S (1223.16) Pa(379.8, mss. BLD) : αυÆ τωÄì LN HI Pa(379.8, mss. AC) : αυÆ τοÁ E b N1R p

2

(a26) αυÆ τοÁ αυë τωÄì fort. Ψ, β HJ2Q2, fort. Pa(379.9) : αυÆ τοÁ

Fw : αυë τοÁ e I1J2Q2 : om. K J1Q1 1

2

2

2

αυÆ τω Äì E be w : αυÆ τοÁ αυÆ τοÁ E (?)J (?)Q1 : αυÆ τοÁ αυë τοÁ e1 : αυë τω Äì K

E2)Ψ : οÊ E2

1

8 (a28) τι E K γ : τι καιÁ Ψ β δ

7 (a27) ωÎì 1et2 Π(praeter

(a28) αυÆ τωÄì E δ : αυë τωÄì fort. Ψ, K β γ

9 (a29) κινου µενο ν τι κινειÄ Π(praeter δ) : τι κινου µενον κινειÄ δ : [τοÁ ] κινου µενο ν (acc.) τι (nom.) κινειÄ intell. Ar. (cf. App. 2 ad loc.) 3 ñs³fN] s.p. Φ :

…¬ hÂI³ (!)Bad

ñsÂfN Bad : an ñsfnB leg. (v. l. 8 = 256a28)?, sed v. App. 3 5 2…ÏU] s.p. Φ : 7 2…ÏU] s.p. Φ : …³¬ hÂI³ (!)Bad 9 …ÏhtmLA ΦCV : …ÏdA scr. Bad (cf. comm. Ibn

al-Tayyib 853.13, and App. 1 ad loc.) ˙

3 gloss without mark i.m.: aB …ÏU ñNÏÒË ñsfN …ÏU .F èÑqI (2–3 litt. illeg., an ÙQ?) i[**] Φ i.m.; cf. Simpl. 1223.18f. and 1222.24–32 (ad Alex.)

256a25

41

VIII 5, 256a21-b6

Cv 72v

3

Ca 26ra 6

Av. t. 36 9 Cp 188r

KÏH ÙiLÑF ÙiLA \Y …ÏW ñNÏÑ  B …ÏU YÑK YÒË RÑkÏy³LA YÏ Ò  †LDÖ Iш a ~gIÒ  RÑkÏy³lL …ÏdA YÏ ÑËF ÑE¡V Ó ÙÂiLA …ÏU YÑK YÒËÖ RÑkÏy³LÑB ‡ÙÏqtM ÑnEÑE YÕkI YÒ  úJÒÕF …ÏU X| a ÙBÒ  ÑnEÑE YÑK ÒDÒËÖ .X| úJÒÕF ñKÏU á¡V a \kI GÖ …ÏhtI ÒÚE YÑK YÑËF .ñTÒD …ÏU ÕE YÕkI YÒ  úJÖ ÙQ ~gIÒ  èÕqLA ÒڌF .ñTÒD …ÏU ÕE YÕkI YÒ  Í rÖ² YÒ  ÑMÏÒËÖ Mi Ò èÖÏÒ  ÚnM ñsfN ñL …ÏdA \Y …ÏhtI YÕkI YÒ  ÑMÏÒË …ÏhtmLA .ñlibS áÚE ÑM ŠÒË èÒÕHi Ò \M èÑH „ èÖqI áMÒ  YÕkI YÒË ñNÏÒ  †LDÖ ñJÕLA ÒÚE ‹Y znLA ÑnlyJ YÒËÖ ñniyB †LD ‡‚lI ÙQÖ „a ~miW ÒÚE YÕkI YÒ  ÑMÏÑËF …ÏhtM \Y …ÏW ;NÏÑËF …ÏhtM øÏK YÑK …ÏhtI ÕEÖ …ÏU ;NÏÒË YÑK YÒËÖ aLA YÕkI htÏH ¿yLA „IpB rÂÕMiªÒ 2 (a31) τηÄì βακτηρι αì α β, fort. P(v. Rashed 2016, p. 762 [Scholie 17]) : τηÁ ν βακτηρι αν K γ δ

Λ

(a31) ταυ την ΠΨ : ταυ τηì Pacius HaGa Pel Ross

KγΛ

(a31) αÍ λλο τι E, fort. Ψ, β δ : αÍ λλο

3 (a32) τινι K δ F IJQ w Λ A (frag. 605) : τι α β HQ2 : τιναÁ F1 2

1

s

E Ψ K β δ HI J Q w Λ : τοÁ αυ E : τοÁ τοÁ E : τοÁ αυÆ τωÄì FI : τοÁ εë αυτοÁ J Q 2

2 1

1

1

3

1

fort. Ψ, I2J2Q2 : αυÆ τοÁ E b w : αυë τοÁ e β FHI1 Λ : om. J1Q1

2

4 (a33) αυë τωÄì

(a34) αυÆ το 1 α β δ HJQw Λ

Pa(379.17, ms. B) : αυÆ τωÄì K F : αυë τωÄì I, fort. Pa(379.17, mss. ACDL)

Ψ, K β FHI1J : αυÆ τοÁ δ R w Λ : αυÆ τοÁ αυÆ τωÄì I2Q

(a33) τοÁ αυÆ τοÁ 2

6 (b2) αυë τοÁ E, fort.

8 (b3) προÁ ς δεÁ τοιÄς ειÆ ρηµε νοις non vert. Ar.

10 (b5) καταÁ Π(praeter E2)Ψ : ηà καταÁ E2

RÑkyLÑB clare Φ (= τηÄì βακτηρι αì , v. App. 1) : RÑkyLA scr. Bad (= τηÁ ν βακτηρι αν, v. App. 1 = RÑkÏy³lL Φ (= CV) : ÙilL scr. Bad 6 YÕkI om. Bad Mi Ò èÖÏÒ  ÚnM scripsi (= ευÆ θυÁ ς, cf. GALex I, awwalu 8.4[a]) : MiÒ èÖÒ joM Φ : in principio rei C : sine medio V 9 …ÏW] s.p. Φ : …¬ hÂI³ Bad 2

baculum CV)

8 gloss without mark i.m.:

ñsfnL ñtKH zÙbM …ÏhtM ÙnY (addidi) 〈YÏ z ‹Y〉 ŸÕQÕLA ‡‚lI Φ i.m., fort.

ex comm. Alex., v. Alex. frag. 606, and above, p. LXXIV

256a30

256b

256b5

42

3

6

Av. t. 37 9

Cp 188v

(iLÑPÕpSrt fÏ yibpLA û4LA êÑtK \M nMÑoLA LÑqmLA

ÒÚE øB †LÚK Mi Ò YÕkI Çz ÑMÏÒËÖ …ÏhtI ñNÏÒ  øbÂQ \M (iL †LD YÏ Ò  ÇÒË YÒ  Í rÖ² úO (ilF ¿yLA „IpB †LD YÑK YÑÂ Ë F iÍ ÖÏÒ  ÑMÏÑ  F . ÒÚLÑB a „ \km ÙQ ñNÏÒ  \iÏbLA \B †LÚK †LD YÑK YÒËÖ .…ÏhtI …ÏhtmLA YÕkI YÕkI ÑM YÏ Ò  †LDÖ …ÏhtI ÒdÕJÕmLA \M a YÕkI Çz èÒÕHi Ò \M èÑH ÕE ÑM ÑnL‚NÒ  \V YÊ ÑËF .YÕkI Çz \km ÙQ øB ~IÏrÖ² ÕE (ilF ¿yLÑB KaA ‡ÙY YÏ Ò  ¡V êÚK & ¦ Y øB jÍ CÒ  èÈÑL ÑnM‚lI G YÕkI Çz \kJ úO ÙQ ñNÏÒ  ø³bQ \M ÑnÏiÏB ÙQ ÑNÏÒ  †LDÖ .èÑL jÍ CÒ  KH YÕkT i htÏH YÒ  〈 úO ÙQ ñNÏÒ  †LDÖ ÒÚE ‡‚lI úJÒÕLÑBÖ .~qÒd KaA YÕkT YÒ  〉 …ÏhtmLA ÑMÏÑ  F .…ÏU ñB ÑMÖ …ÏdÒÖ …ÏhtmLA aÑiWÒ  OjO ÑnEÑE YÕkI úJÒÕF …³ÏdA …ÏU ñB ÑM ÑMÏÒ  Ö .…ÏU YÒ  ~bJÒÖ (iLÖ …ÏhtI YÒ  úJÒÕF …ÏhtmlL „FÒÕM ñNÏÑËF †LD õMÖ Âi¬ vÂT ~gIÒ  ÒÚE YÏ ÑËF …ÏhtIÖ …ÏU YÕkI YÒ  1 (b7) αυÆ το E K β δ JQ1w Λ, fort. P(v. Rashed 2016, p. 763 [Scholie 20]) : αυÆ τοÁ αÆ ει FHIQ2 : fort. om. Ψ As(frag. 608) Pa(379.26) κινου µενον α e2 LN1R γ Λ

3 (b8) κινουÄ ν K be1 N2 As(frag. 608) Sc(1225.25) :

6 (b11) µηÁ ειËναι E1Ψ F : ειËναι E2 K β δ HIJQw Sp(1225.27)

7 (b12) µηÁ ειËναι] µηÁ ειËναι ωÏ στε µηδεÁν κινειÄσθαι /ωÏ στε αÆ ναιρειÄται τοÁ κινειÄσθαι vel sim. Ψ (nisi intell. Ar., sed v. Sp[1225.38–39] Pa[380.9–10]) γαÁ ρ ηÍ δη τουÄ το H

(b13) γαÁ ρ α K β δ : γαÁ ρ τουÄ το FIJQw Λ :

8-p. 43.9 (b13–27) καιÁ ευÆ λο γως ... αÆ µιγηÁ ς ωÍ ν hic ΠΨ Pa(380.11ff.) : post

αÆ κι νητον 258b9 transp. T(220.29, 223.12) A(apud S[1224.26])

E)Ψ : του του τοÁ E

8

11 (b17) τουÄ το Π(praeter

(b17) συµµεταβα λλει γαÁ ρ τουÄ το, αÏ µα καιÁ ... i’punxit Ar.

úO ... YÕkT YÒ  addidi (= κι νησιν αÆ ειÁ ειËναι. καιÁ ευÆ λο γως δεÁ [δεÁ om. HJ :

δηÁ K] τουÄ το

συµβε βηκε[ν]. τρι α γαÁ ρ αÆ να γκη Π = ut semper sit motus et necessario sequitur illud quod est

quia oportet C = ut semper sit motus et de necessitate sequitur hoc necesse est enim V; cf. also 220b24 = 443.9) : om. hom. Φ

11 ÒÚE Φ1 :

áÚE Φ2 „FÒÕM scripsi (= καταÁ τοÁ αυÆ τοÁ τωÄì

κινουµε νωì οÍ ν [τοÁ οÍ ν L] Π = est compar ei quod movetur C = est coniunctum cum moto V) :

ƒrÑfM Φ Bad

256b10

256b15

43

VIII 5, 256b6–27

3

6

Cv 73r 9

.ÑM Í FÑsM iQjtM Í rÖ² ÑeNÏi  YÑkmLA „ ÑKÏdA „ áMÒ  \M \È iÏB †LDÖ ÙQ ÑnÏK DÊÑËF .…ÏhtM ¡vF …° htLA YÕkI ñB ÑM (iL ñNÏÒ  ‹Y …ÏU ÑM ÑMÏÒ  Ö ÑM ÙPÖ KH zÙbM ñiF (iL ñNÏÒ  ¡V …ÏhtI YÒ  \km cÚLA ÕEÖ ¡Xi Ò ÙP èÕqN i- ŠÖÊi ÑF ñlÑqlT \M øB á¡V \M (iL ñNÏÒ  i¬ ÒË …ÏW [Ò 215] .…ÏhtM ¡V ÕEÖ …ÏU cÚLA ÕEÖ dÕJÕM ~gIÒ  ÓLÑoLA YÕkI YÒ  -úJÒÕLÑF øBÑQ ¡V ñNÏÒË øqyLA „ ñLÕQ „ êÑCÒ  ÙQ ~gIÒ  %rÕVÑskNÒ  YÏ ÑËF †LÚLÖ YÕkI áÙHÖ ñJÕLA Òچ YÏ ÑËF .KhlL Ò¡ ÙbM ñlyJ ÙQ YÑK DÊÒË ×LÑM iÖ ¡OÑ  tlL YÑ  B YÒ  ~bLÑV YÕkI áÙHÖ ñJÕLA ÒÚ†Ö …ÏhtM ¡V YÕkI YÑ  B YÒ  ~KÏL .×LÑM ¡V YÕkI 1 (b19) αÆ λλη λων αÆ να γκη α K be1 β : αÆ να γκη αÆ λλη λων ουÏ τω[ς] γ Λ : αÆ λλη λων αÆ να γκη ουÏ τως 2 (b20) ωÎì Π(praeter E2)Ψ : οÊ E2

e2

4 (b22–23) κινειÄται µε ν, ουÆ χ υë π’ αÍ λλου δεÁ αÆ λλ’ υë ϕ’ α

be , fort. A (frag. 617) : κινειÄται ουÆ χ υë π’ αÍ λλου µεÁν αÆ λλ’ υë ϕ’ K : κινειÄται µεÁν υë π’ αÍ λλου δεÁ 1

s

αÆ λλ’ ουÆ χ υë ϕ’ e2 β γ, fort. Pa(380.21–22) : κινειÄ µε ν, υë π’ αÍ λλου δεÁ 〈 κινειÄται〉 αÆ λλ’ ουÆ χ υë ϕ’ Pra

sec. T(223.5ff.) S(1227.10ff.), sed v. Ross 699 c

S (1227.24f.,32)

(b23) ιÏνα] ειÆ καιÁ (nisi om.) Ψ, v.

(b23) ειÍπωµεν Π(praeter δ)Ψ : ειÍπω δ

6–7 ¡OÑ  tlL øBÑQ ¡V Φ(= Π) : est immobilis sed suscipiens impassionem C (= øBÑQ øB …ÏhtM ¡V

¡OÑ  tlL ?) : non mouetur neque recipit passionem V (= ¡OÑ  tlL øBÑQ ¡VÖ …ÏhtM ¡V ?) YÒ  Bad

8 2YÑ  B Φ :

2 ¡vF with mark s.l. and two glosses with the same mark, gloss 1: \Y rÑX (sic) ;emefI Yz fvbnI

á¡V; gloss 2: á¡V (Φ …T : scripsi) ñKÏU (iL cz Φ i.m., fort. ex Philop. 832.32f. 4 ñlÑqlT \M with the following gloss next to it: á¡V \Y …ÏhtM …ÏhtM øÏK YÏ ÒË èÕqI i cz Φ i.m.; cf. Alex. (frag. 617)

256b20

256b25

(iLÑPÕpSrt fÏ yibpLA û4LA êÑtK \M nMÑoLA LÑqmLA

44

Av. t. 38

Cp 189r 4

8

øB ¿yLA „IpB …ÏhtI ;NÒË …ÏdA \kI G YÊ ÒË Ù³yB \M vÏ …ÏdA YÑK YÒË Í rÖ² úJÒÕF …ÏU G …ÏhtI 〈 G YÒË 〉 YÑKÖ Í rÖ² .á¡vB ÑMÏÒËÖ …ÏU ñB cÚLA ñniyB ûÕnLA †LÚB ÑMÏÒË …ÏhtI YÕkI YÒ  …ÏhtI ÕE gªbÊm³LÒÖ \x³sÊI ~gIÒ  ñsfN ÕE \x® sÂm³LA YÕkI YÒ  ÑMÏÒË ñNÏÒ  †LÚB YÒ  Ö Â .ÕmnI øQÑnLÒÖ øqtnI gª mLA YÕkI YÒ  ÑMÏÒËÖ øqtnI ñsfN ÕE øQÑnLÒÖ z I ñsfN „ msqLÑB DlbN YÒ  úJÒÕLA \M YÏ Ò  †LDÖ èÑL ÒÚE YÏ Ò  \iÏbLA \M \kL ˆlÏytM ÕeF ÑM Î  HÑsM ˆÂl¬ Y \Ê M Y¬Ò  †LD èÑoM ®ÑxWi Ò ŠÒË ÑnLÕQ \M ê³LA †LD °ÖpM ñNÏÑËF ÑM ~miW °P \MÖ ñniyB aLA †LD HÑsM юM ÙHÒÖ øÏK øB 〉 øibsLA áÚE ‹Y cO Mi Ò YÕkI i ÖÒ  .°pLA 2 (b29) ηÎì β FHJQ2 : ηà Q1 (?)w : ειÆ α K δ Q3 Λ τοÁ E δ : ωë ς τοÁ fort. Ψ, propos. HaGa WiCo

3 (b30) ωÏ ς γε scr. Ross : ωÏ στε K β γ : ωÏ στε

4 (b31) δ’ οÏτι (vel δεÁ οÏτι) Π(praeter E)Ψ : δεÁ E

5 (b33) ηà ... ϕε ρεσθαι Π(praeter E )Ψ : exp. E2 2

8 (a2) ηà Π(praeter E2)Ψ : om. E2

διαϕε ρον be1

(b33) δεÁ ϕε ρον Π(praeter be1)Ψ : (a2–3) ρë ιπτειÄ] ρë ι πτει τι intell. Ar.

9 (a3) ηà E Ψ K β γ : ειÆ δεÁ /δ’ E K δ 1

2

2

1

YÊ ÒË scripsi (= ειÆ Π = si V [non vert. C]) : YÒ Φ : YÒ  Bad …ÏdA scripsi (= τοÁ κινουÄ ν Π = motor V = illud quod mouet C) : …htmLA Φ Bad …ÏhtI ΦΠC : mouet V 2 G YÒË addidi (= ειÆ δεÁ µηÁ Π = et si non V = et quando non C) : om. Φ Bad …U scripsi (= κινοι η [κινοιÄ FJQ] Ï Π = mouet CV) : …htI Φ Bad 3 …ÏU scripsi (= mouet CV) : …W Φ Bad 5 ÕmnI ... ÑMÏÒËÖ] 1

aut ut transferatur sanans et sanetur transferens aut ut augmentetur V : aut sanans permutatur (aut ... permutatur om. Cv) et permutans ipsum (ipsum om. CaCv) sanatur aut augmentetur (augmentetur CaCv : augetur Cp) C 9-p. 45.1

YÏ Ò  ... øB

7

ˆÂl¬ YÂ] ˆlY Φ

: docet C : scit V :

addidi (= αÍ λλο δ’ εÆ ξ αÍ λλου γε νους [γε νους om. F ] οιÎον Π 1

ˆlY Bad =  immo

unumquodque mouetur ex genere generum [generum om. CaCp] sicut C = sed unumquodque mouetur secundum unumquodque generum verbi gratia V) : om. Φ Bad

256b30

257a

VIII 5, 256b27 – 257a18

5 p

C 189v Ca 26rb

10

45

ÒÚ= fmÏnÂm³LÒÖ hmÏnI³ øQÑnLA 〈 YÏ Ò  †LD èÑoM ûÒÕNi Ò \M ûÕN ‹Y …ÏhtI Mi Ò gJ YÒË \kL .gXÒª ÑM KH …ÏhtI ÒÚ= øidÒÖ á¡V \Y øihtsI .iEÑntM KaA ŸÑnCÒ  YÏ Ò  †LDÖ KaA žqT YÒ  Í rÖ² úJÖ ÒÚE ‹Y ñLÕQ YÑK jÍ qntM øidA YÕkI htÏH žpyiF dÕyI ÒÚE YÏ ÒË ølÑQ èÑQ YÒË ÑMÏÑ  F .ˆlÏytM ˆlÏymLÒ〈 Ö 〉 øqtnM øQÑnLA YÏ ÒË Â Mi Ò èÖÏÒ  ÚnM èÑQ ÕL ;K ÒÚE õFri Ò …ÏdA \Y ~gIÒ  …ÏhtI ñNÏÑËF …ÏhtM øÏK YÏ Ò  \iÏbLA \M ñNÏÒ  †LDÖ YÕkI YÒ  ‡‚lI ñNÏÒ  †LDÖ èÈÑL ÒÚE YÒ  ¡V ÑKÏdA ‡ÙQÒ  \Y -i¬ S iÖ .ñtFyM áÙnY ñgyBÖ ñtFyM áÙnY (iL LÑL i ñgyB ÑM ˆlÏytI ˆlÏymLA ñN{wF †IhtLA ñN{W ÑM ø° K YÕkI YÒ  ‡‚lI ñNÏÒ  ~gIÒ  †LD \M õnWÒ  Ö YÒ  ‡‚lI ñNÏÒ  †LDÖ .〈 …ÏhtI …ÏhtM \yF …ÏhtM øÏK YÑK DÒË 〉 …ÏhtI YÒ  nbÊI³ YÒ  ñNÑ  W \M ÑM ø¬ K YÏ ÒË èÑQ jÍ lÑQ YÏ Ò  ÕL ;K …ÏhtI YÒ  ñNÑ  W YÕkI ÚnM ÑMÏÒË !¦ bI³ YÒ  ñN{wF bÊI YÒ  ñNÑ  W ÑM øÏKÖ Ò  ÂbÊI YÒ  ñNÑ  W \M YÏ ÑËF Ò|BiË ÒÖ 1 (a4) ϕε ρον µεÁν Π(praeter E1 K) : ϕε ροµεν οÃν E1 : ϕε ρον K : ϕε ρον µεÁν vel ϕε ρον Ψ 3 (a7) γαÁ ρ α K be1 β F : γαÁ ρ ειÆ σιν e2 HIJQw Λ

11 (a17) παÄ ν α K β δ : οÏτι FHIJ1Qw Λ : οÏτι

12 (a17) καιÁ υë για ζον α (sed v. App. 2 ad loc.) K b β : καιÁ υë για ζον καιÁ e FHIJw

παÄ ν J2

Pa(382.3) : καιÁ τοÁ υë για ζον καιÁ Q : om. (?)Sp(1231.9–10) : secl. HaGa Ross

(a18) καιÁ τοÁ

Π(praeter E ) : τοÁ E : καιÁ παÄ ν τοÁ intell. Ar. 1

1

1 ÒÚ=] non vert. V : habet C

Φ

CV :

5 øqtnM] s.p. Φ :

øqntM Bad Ö addidi (= καιÁ Π = et CV) : om. …htM Φ Bad 8 ˆlÏytI] s.p. Φ : addiscat 10 …ÏhtI ... DÒË addidi (= ειÍπερ αÏ παν [αÏ παν : αÍ ρα

6 …ÏhtI scripsi (= κινειÄται Π = mouetur CV) :

ˆl° ytB Bad

9

ñNÏÒ  Φ : YÒ  scr. Bad

παÄ ν Q] υë ποÁ κινουµε νου κινειÄται τοÁ κινου µενον Π = si omne quod mouetur ab eo quod

mouetur mouetur C = si omne quod mouetur moueatur a motore moto V) : om. hom. Φ, om. Bad

6

12

Ò|BiË ÒÖ] om. V : habet C

õFri Ò] with corresp. marks s.l. and i.m.: èÖiÒ ˆeFÒ Φ i.m.

257a5

257a10

257a15

(iLÑPÕpSrt fÏ yibpLA û4LA êÑtK \M nMÑoLA LÑqmLA

46

3

Cp 190r Cv 73v 6 Av. t. 39

9

12

ø¬ K YÏ ÒË jÍ oM lQ †NÏÑ  K †LÚB YÒ  Ö .ÙHÒÖ \M ¤KÒ  ×SÏÕtB ÑMÏÒËÖ Mi Ò èÖÏÒ  KaA siL \kL á¡V \Y …ÏhtI YÒ  〈 ñN{wF †IhtLA 〉 ñNÑ  W ÑM .gXÒª KH øB ñbÂIQ ц …ÏU LA KaA fE ц …ÏhtI YÒ  ñN{W LA ÒDÒË †nÏkL .ˆl° ytLA [ê 215] ñNÑ  W nbÊI³ YÒ  ñNÑ  W cÚLA YÏ Ò  †LD èÑoM ñniyB ûÕnLA †LD ŠÒË èÒÕHi Ò \M èÑH  „ ³Mi Ò †B è| ÒÚE \Y iqTr ~ynW YÒ  †LDÖ ®ÏxtM ñgyBÖ èÑLÍ †LD \M áÑnlQ ÑM ¡ybF .~fN| ÑnlQ ;K  .ÕhA ñN{W YÕkI YÒ  Í rÖ² úJÒÕF LÑHiË Ò ñNÑ  W cÚLA YÏ ÒË èÑqI YÒ  (ilF Ï ÒÚE YÕkIÖ á¡V \M ÙBÒ  …ÏhtI …ÏhtmLA YÕkI YÒ  Í rÖ² DÒË úO …ÏhtI YÕkI YÒ  ÑMÏÒË èÖÏtÒ …ÏhtmLA YÕkiF .†LD DÒË žqI ÙqF ~KÏhtM ¡vLA .ñTÒD …ÏU ÕE YÕkI YÒ  ÑMÏÒËÖ \KÑS \Y KaA úbS ÕE ›nOiÒ cÏ Ò  zÂnI³ YÒ  ŠÒË ÑtU ~gIÒ  YÑK YÑËF YÒ  ŠÒË ûrÑsI ÙHÒÖ ø¬ K YÏ ÑËF á¡V \M …ÏhtmLA ÖÒ  ñTÒD …ÏdA øE ÑEÖª ÙbMÖ Í 3 (a20) ηÊν Π(praeter E1)Ψ : ηÊν κεινειτον ηÊν E1

8 (a25) αÆ ειÁ α K δ LN[αιÆ ειÁ ]R FIJ2Q2w :

δειÄν J1 (?)Q1 : ειËναι αÆ ειÁ H : om. Sc(1231.36) Pa(382.13)

J K)Ψ : κινουÄ ν K Pa(382.13) : κινου µενον κινειÄσθαι J 1

Ar.

(a28) ειÍ Π(praeter E1)Ψ : ηÍ E1

1

(a25) κινου µενον Π(praeter 11 (a27–28) αÆ λλαÁ µηÁ ν] non vert.

12 (a29) τοÁ αυÆ τοÁ αυë τοÁ β γ Λ : τοÁ αυÆ τοÁ αυÆ τοÁ E1K : τοÁ

αυÆ τοÁ E2, fort. Ψ, δ, fort. P(v. Rashed 2016, p. 767 [Scholie 26])

2

ñN{wF †IhtLA

addidi (= κινητοÁ ν µεÁν ... τοÁ κινητικο ν Π = quod aptum est moueri est aptum

... C = quod innatum est mouere innatum est ... V) : om. hom. Φ, om. Bad

ц …ÏhtI] LÑHiË Ò Φ =

3

ьKhbB Φ (ut moueatur ipso C : moueri secundum ipsum V) : ÑeKhtI Bad 7 alterare V : resolui (= èjHiË Ò ?) C 10 …ÏU s.p. Φ (= mouebit CvV) : …ÏH (= mouet CaCp) Bad 11 ante YÑËF add. (iLÑPÕpSrÒ èÑQ Φ : om. CV : secl. Bad zÂnI³ scripsi : znI Φ1 : \Ï znI Φ2

257a20

257a25

VIII 5, 257a18-b3

Av. t. 40

5 Cp 190v

47

ÙBÒ  ÕeF ÑM èÑT ñTÒÚB ÕE cÚLA aLA YÏ Ò  〈 †LDÖ 〉 …ÒD ñygI Í ] .〈 ~gIÒ  〉 èÑaA †ltB ÕE á¡vB YÑK ;L ‡ÙQi Ò úbsLA [9ê215 g ÙtbN YÑ  B ÒÚE † ÑNzN øyP YÒ  YvÒ fvbnI ÙqF [20 Ò 216] …ÏU žikF ñTÒD …ÏU ÕE a YÑK YÒË ñNÏÒ  !ymLÑB YÒ  Ö X| ÐÒÙtB ~msqnM …ÏhtM øÏK [ê216] YÕkI YÒ  rÖ² ñNÏÒË èÕqnF .ñJÖ cÏ Ò  ‹YÖ MÒ  ømÂJ³ „ ÑnMjK „ ‡ÙÏqT .F \iÏbT ÙQ a ÒÚE YÏ ÑËF Í 4qnM ŠÒË ÕE cÚLA YÕkI YÒ  \km (iLÖ .øctÏM ÕeF ñTÒÚB …ÏhtM ø¬ K YÏ i  yibpLA Í lqÊN³ øqtnIÖ øqnI á¬Ñ  B YÕkI ñNÏi  ñTÒD ÕE ñtlmN …ÏU ñTÒD …ÏU 1–2 (a30) αÆ ειÁ προ τερον αιÍτιον α K : αιÍτιον αÆ ειÁ προ τερον β γ δ Λ : προ τερον αÆ ειÁ καιÁ αιÍτιον T(221.5, cf. Pel 4111)

4 (a32) κινειÄ αυÆ τοÁ αυÆ τοÁ E K b : κινειÄ αυÆ τοÁ αυë τοÁ e HI Λ Pa(382.20) :

αυÆ τοÁ αυë τοÁ κινειÄ β JQ : Ψ = eHI ∨ βJQ : κινειÄ αυë τοÁ αυÆ τοÁ F : κινειÄ αυÆ τοÁ w

Π(praeter E) : om. α, fort. Sp(1233.23ff.)

6 (a34) αÆ ειÁ

7 (b2) δηÁ Π(praeter e1) : δεÁ fort. Ψ (?)e1

8 (b2) αυÆ τοÁ αυë τοÁ 1 Ψ e β I : αυÆ τοÁ αυÆ τοÁ E b2 w : αυÆ τοÁ εë αυτοÁ FHJQ : αυÆ τοÁ K b1 (cf. As[frag. 625] Sl[1233.34 + app. crit.]) εë αυτοÁ w

(b2) αυë τοÁ 2 Ψ e β FHIJQ, fort. As(frag. 625) : αυÆ τοÁ E K b :

(b3) ϕε ροιτο ... ϕε ροι K β γ Λ : inv. α δ T(221.29) Sp(1234.12f.)

1 ñygI scripsi (= θει η Π = ut ponat C = ponendum V) :

ñcyB Φ : Ùg® yI Bad (for õGÖ c. dupl. …ÒD Φ (= εÆ κειÄνο Π) : †LD Bad †LDÖ addidi (= γαÁ ρ Π = illud enim quia C = illud enim quod V) : om. hom. Φ, om. Bad 2 á¡vB scripsi (= καθ’ εÏ τερον Π [v. also 210a27 = 293.10 Bad, 256b31] = per aliud C; cf. cuius transmutatio [= ái° vT pro á¡vB] ~gIÒ  addidi (= καιÁ Π = iterum C = etiam V) 3 ante ÙqF add. V) : á bbB Φ : áiÏsI Bad (iLÑPÕpSrÒ èÑQ õBÑsLA ylytLA Φ Bad : om. ΠCV : seclusi 4 !ymLÑB secl. Bad : habet C [non vert. V] 6 \iÏbT scripsi : šbB Φ : \iÏB Bad 7 YÏ i   Φ (= οÏτι Π = quoniam C = enim V) : Yz 2 Bad 8 …ÏU (= κινειÄν) s.p. Φ : …¬ hÂI³ (!) Bad acc. v. Ullmann NE I, 430)

3 ÒÚE „ with corresp. marks s.l. and i.m.: ñTÒD ñtKH zÙbM cÚLA èÖÏtÒ …ÏdA yI Φ i.m.; cf. Alex. frag. 623

g ÙtbN YÑ  B with corresp. marks s.l. and i.m.: X| êÕlpM ŠÒË øqtnN Y{B cz Φ i.m.; cf.

Alex. frag. 623 and Rashed ad loc.

257a30

257b

(iLÑPÕpSrt fÏ yibpLA û4LA êÑtK \M nMÑoLA LÑqmLA

48

3

6 Cp 191r

øiUÖ øihtsI YÕkIÖ ñYÕN „ qnM ¡V ÙHÒÖ ÕEÖ ÑŽiyB Í ÙHÒÖ Ñncbّ ÙQÖ .ñniyB ÙHÒÖ ÒЁÊB³ zÂbÊIÂÖ gªbÊI³Ö ~yM ˆlÏytIÖ ˆlÏyI DÒË YÕkiF  .F …htI ÑnLÕQ !yM ÑM ~gIÒ  i …ÏhtM ÕÏqLÑB cÚLA ÕEÖ …ÏhtI YÒ  ñN{W Ï è;ktS fE KaÒÖ è;ktSiÒ ŠÒË hysI ~blÒd …ÏhtmLÒÖ .è;ktSiÑB YÏ Ò  †LD èÑoM øyfLA ŠÒË X ÙqF …ÏdÒÖ .…ÏhtI YÒ  ñN{W ;L ‡ÏÑT ¡V .rÕcLA ñL lcH ÙQ ÑM YÕÏkI ñNÏz lm`ÑBÖ rÏÑaA ÕE \³ x® sÂI³ cÚLA ‹YÖ .rÏÑH (iLÖ rÏÑH юiyB ÙHÒÖ eJ \MÖ ~yM ñniyB aLA YÕkiF YÕkI LA ÑKÏhtmLA lÑS \M ÙHÒÖ øÏK „ Mi Ò cO èÑomLA ÒÚE Í 1 (b3) εÊν οÃν Π(praeter E)Ψ : εÆ νον E

(b4) καιÁ 1 Π(praeter E1K)Ψ β γ δ : ηà E1 : ηà καιÁ K

(b4) αÆ λλοιοιÄτο Πpraeter E)Ψ : om. E

2 (b5) αÏ µα Π(praeter E2)Ψ : exp. E2

υë για ζοι Π(praeter E)Ψ : om. hom. E

(b5) καιÁ

2–3 (b6) διω ρισται οÏτι ...] an διω ρισται οÏ τι (vel τοÂ

τι) ... Ψ ? (»we have determined what it means to say about the movable that it is moved«

Ar., sim. Pa[383.12–13])

4 (b7) δυνα µει2 γ Λ : κινου µενον α K β : δυνα µει κινου µενον δ

4–5 (b8–9) εÍ στιν ... αÆ τελη ς Π(praeter δ)Ψ : om. δ 7 (b10) αυÆ τοÁ E γ : αυÆ τοÁ καιÁ fort. Ψ, K β δ S (1235.23) c

E )Ψ : om. E 1

6 (b10) τοÁ 2 α K δ FIJQw : om. β H (b10–11) καταÁ τοÁ αυÆ τοÁ Π(praeter

1

1 øiUÖ Φ :

øiU ÖÒ  scr. Bad 4 ~blÒd] ÑbBÒd Φ : quaerendo C : semper (= ~qÒd ?) V è;ktSÒ] perfectio V : actus (= øyÊF) C 6 ñNÏz] ñNÒ Φ : ñNÒË Bad ñYÕN „ qnM ¡V ÙHÒÖ YÑK ÒÚelF …ÏhtM á¬{BÖ …ÏL á¬{B ñNÏt Φ i.m.; cf. Philop. 834.10, Alex. frag. 625–626 (v. p. LXXIV) 7 gloss without mark: rÏÑH ~yM YÕkI cz ÙHÒÖ eJ \M ñL †LD YÕkIÖ ñniyB ÙHÒÖ ÕEÖ rÏÑH (iLÖ Φ i.m., fort. ex comm. Ibn al-Tayyib ˙

1 gloss without mark:

863.9–11

257b5

257b10

49

VIII 5, 257b3–18

Av. t. 41

3 Ca 26va

6

…ÏU ÕE cÚLA YÕkI YÒ  DÒË úJÒÕF .Í rÖ² Ñ= ~mPÒÕM Ñ= …ÏdA .…ÏhtI ñgyBÖ …ÏU ñgyB ñTÒD …ÏU ;NÏÒË ñTÒD …ÏU ÕE cÚLA (iL ñNÏÒ  áÕllÑQ \V ÑJÏ ezI ÙQÖ i ñNÏÒ  †LDÖ .ñil ‚ÊJ³ \M ÙHÒÖ øÏK \Y …ÏhtI ñil ‚ÊJ³ \M ÙHÒÖ ø¬ K YÏ Ò  ‹Y …ÏU ;ŽM ÙHÒÖ øÏK YÑK DÒË iÍ ÖÏÒ  ~KÏL ÚmniH ;ŽM ÙHÒÖ iÖ YÕkI úÂbS Y ÕkI YÒ  \M ŠÖÊÒ  ‡³ÙQi Ò †IhtLA Í úÂbS Y ÕkI YÑ  B ñNÏi  ñTÒD    †LÚL ñkIWÖ õBÑtLA †IhtLA ›B² ‹Y †IhtLA YÑK ÙqF .¤KÒ ³  2 (b13) αυë τοÁ (vel εë αυτοÁ ) E2Ψ K e β FHBIJQ Λ : αυÆ τοÁ b w : om. E1 E )Ψ : om. E 1

1

3 (b14) ουÆ κ Π(praeter

(b14) εÍ στι[ν] α K δ S (1237.3) : εÆ νδε χεται β γ Pa(383.26) l

K b FIJQw Pa(383.26) : τοÁ αυÆ τοÁ E2, fort. Ψ, e β Sl(1237.3) : illeg. H εë αυτοÁ ) Ψ K e β FHIJQ Sl(1237.3) Pa(383.21.26) : αυÆ τοÁ E b w

Sl(1237.3) : κινειÄν γ Pa(383.26)

(b14) αυÆ τοÁ E1 (b14) αυë τοÁ (vel

(b14) κινουÄ ν α K β δ

4 (b14) ωÏ σθ’ (vel ωÏ στε) Π(praeter E)Ψ : ωë ς καθ’ E

5 (b15–16) ουÆ δε ν, ειÍ Π(praeter E) : ουÆ δενι E : an ουÆ δεÁν εÍ τι, ειÆ Ψ?

(b16) αυÆ τοÁ εë αυτοÁ K,

fort. S (1238.1,6) : αυÆ τοÁ αυÆ τοÁ E : τοÁ αυÆ τοÁ αυÆ τοÁ E : αυÆ τοÁ b : αυÆ τοÁ αυë τοÁ e : εë κα τερον β γ : p

fort. αυë τοÁ Ψ

1

2

6 (b16–17) αιÆ τιω τερον Π(praeter E)Ψ : om. E

(b17) τουÄ κινειÄσθαι] τουÄ

κινειÄν (vel τουÄ κινηÄ σαι) fort. Ψ (cf. αιÆ τιω τερον τηÄ ς κινη σεως Pa[384.5] and App. 2 ad loc.)

1 1Ñ= om. CV

…ÏU scripsi (= mouet CV) : …L vid. Φ, Bad (cf. Glossary) 5 ~KÏL scripsi 2 (= κινουÄ ν Π = motor V) : ~KÏhtM Φ (= que moueatur C) Bad ;ŽM Φ : ;ŽiB scr. Bad 6 †IhtLA clare Φ (= motionis CV) 7 ÙqF scripsi (= γαÁ ρ Π = autem CV) : ÙQÖ Φ Bad

1 ~mPÒÕM] with corresp. marks s.l. and i.m.:

!ymLÒÖ ~iÒ „ ~KrÑwM cz Φ i.m. 4 2ñil ‚ÊJ³] with corresp. marks s.l. and i.m.: øibQ ц ñtmsQ cÚLA yI Φ i.m.; cf. Simpl. 1237.6f. 6 1úbS] with corresp. marks s.l. and i.m.: gHzÖ ŠÖz ñfsnL YÕkI Y{B ÕeF KaA úbS á¡vL ÕE cÚLA cz Φ i.m. =˜ Philop. 835.14f. 7 õBÑtLA] with corresp. marks s.l. and i.m.: XvÒ a‚`A yI á¡vL cz Φ i.m., fort. ex comm. Ibn al-Tayyib 863.20f. ˙

257b15

(iLÑPÕpSrt fÏ yibpLA û4LA êÑtK \M nMÑoLA LÑqmLA

50

Cv 74r 3 Cp 191v

6

…ÏW ÕE YÕkI YÒ  XvÒÖ á¡V \Y ÕE …ÏhtI …ÏdA YÕkI YÒ  ;EÙHÒ  ñNÏÑËF ~gIÒ  Ö .;ŽiB ÑJÏ zÙbmLA ŠÒË êQÒ  …ÏhtmLA \M ÙyBi ÒÖ .ñsfnB .ñlÑqlT \M …ÏdA \kI G YÒË …ÏhtI …ÏdA YÕkI YÒ  Í rÖ² úO (iL YÒ  DÒË \km ÙqF .ñbÂHÑC ;ŽM ÙHÒÖ øÏK …ÏU DÒË ¿yLA „IpbF Í   F .…U i ;ŽM ÙHÒÖ øÏK YÕkI XvÒÖ ×qF ~KÏhtM YÕkI DÒË ;EÙHÑ Ï dÕyI …ÏdA YÕkI YÒ  Í rÖ² ~bJÒÖ (iL ñNÏÑËF ~gIÒ  Ö .…Í Ï htM ¡V ~KÏL ÖÒ  …ÏhtM ¡V …ÏL a ÑnEÑE YÕkI YÒ  Í rÖ² úO ÙQ øB …ÏhtiF 1 (b18) τοÁ 2 β HIJQw Pa(384.9) : τω c. ras. E1 : τωÄì E2, fort. Ψ, δ : om. K F

3 (b20) τοÁ

κινουÄ ν κινειÄσθαι ΠΨ Pa(384.14) : τοÁ κινου µενον κινειÄν A(apud S[1239.11–12]) As(frag.

630, v. also Rashed, p. 553) ξ (v. App. 2 ad loc.)

4 (b22) εÍ λαβον τοι νυν E β δ FHIJ1Qw Λ

Sc(1239.13), fort. Pa(384.15) : inv. K : εÍ λαττον τοι νυν J2 : εë κα τερον (nisi εÏ καστον?) τοι νυν fort. Ψ (et εÆ νδε χεται pro εÆ νδε χεσθαι interpr. Ar.; on εÍ λαβον cf. Rashed, p. 553) κινου µενον] κινου µενον µο νον interpr. Ar. (cf. T[222.24–26])

Pa(384.19) : om. E : ειÆ E : ειÍη (nisi ηËì ) Ψ 1

5 (b22–23)

7 (b24) ηà Π(praeter E) 1

(b24) τι Π(praeter KF)Ψ : om. K F

2

(b24)

κινειÄν Π(praeter E1) : κειν c. ras. (an κεινουÄ ν?) E1 : an κινουÄ ν Ψ?

1 …ÏdA] quod mouetur C : motum V (= …ÏhtmLA)

ÕE om. CV 3 …ÏhtI …ÏdA (= τοÁ κινουÄ ν …ÏU …ÏhtmLA with the sign X i.m. 5 post DÒË add. i Φ Bad : om. ΠCV : seclusi 1

κινειÄσθαι Π = quod mouet moueatur C) Φ (with mark s.l.) :

Φ (= A As[v. App. 1] = motum moueat V)

2 ~gIÒ  Ö] with corresp. marks s.l. and i.m. : G ñTÒD \MÖ ñlÑqlT \M ~KÏhtM …ÏdA YÕkI Yz úO G YÒË èÕqI

;EÙHz YÕkI Yz \km ÙqF ¿yLÑBÖ úJÒÖ ñL ñkIW øB ñbHÑC ›l‚`A \IÚE \M ÙHÒÖ øÏkL …ÏU Yz úO ñIzr ÕE ÒÚEÖ …ÏL ¡V ~KÏhtM XvÒÖ ×qF ~KÏhtM YÕkI ÙQ YDÒË ;EÙH{F èÑQ v …ÏL ¡V Φ i.m. (ex comm. Ibn Adı¯?, the last sentence quotes Isha¯q’s transl., for the first part cf. Simpl. 1238.36– ˙ 1239.8) 5 ;ŽM] with corresp. marks s.l. and i.m.: ñTÒÚB …ÏhtmLA (leg.? ‚J : Φ sic) eJ \M cz

Φ i.m., fort. ex comm. Ibn al-Tayyib 863.22ff. ˙

257b20

51

VIII 5, 257b18–32

Av. t. 42 4

8

ÇÒËÖ .KH ÙBÒ  YÕkT YÒ  Í rÖ² ~bJÒÖ YÑK YÒË ñlÑqlT \M ÕE …ÏhtI a .\x³sÊI \x® smLA YÕkiF …ÏU LA  KaA …ÏhtI YÑK ÕE 〈 ÑJÏ 〉 ÙÈHÒÖ Ó‚J i〈 Ö 〉 YÕkI YÒ  ~gIÒ  [Ò217] \km iÖ 〈 ñNÏz 〉 †LDÖ .ñsfN ÒD ÙHÒÖ ø° K …U ÙHÒÖ \M ¤ ³ KÒ  iÖ ñTÒD …ÏU ÅÖÏÒ  Ï Í YÒ  ÑMÏÑËF ñlÑqlT \M …htI ÕE YÑK YÒË á¬Ñ  B \M ÑM a‚J \Y …ÏhtI YÕkI Ï …ÏhtI YÑK YÊ ÑËF .ÑE¬Ñ  B ñTÒD 〉 \M …ÏhtI á¬Ñ  B ÕE YÕkI YÒ  ÑMÏÒËÖ ñl҂JÒ  cÚLA èÖÏi Ò ÕE a‚`A †LÚF ñlÑqlT 〈 \M …ÏhtI ÕE ñl҂JÒ  \M ÑM ÒЂJ YÏ {B ÒD 〈 ÕE …ÏU 〉 dFÒ  ÒDÒË Ò‚`A ÒÚE YÑK ñNÏÑËF ñsfN ÒD ÕE …ÏU 1 (b25) αυë τουÄ (vel εë αυτουÄ ) α β δ FHIJQ Λ : αυÆ τουÄ K F w

(b25) εÍ τι] οÏτι γρ.A(apud

S[1240.18]) : an εÆ πειÁ (v. GALex I, wa-illa¯ 3) vel ειÆ [δεÁ] µηÁ (v. An. post. 71a29 = 331.7 Bad, 99a6 = 478.9, GALex I, wa-illa¯ 1) Ψ? γρ.A(apud S[1240.18]), fort. Pa(384.24)

2 (b25) κινειÄ E1Ψ K γ δ Sl(1239.33) : κινειÄται E2 β (b26) κι νησιν α K β δ : κι νησιν καιÁ γ Sl(ms. a)

(b26) κινοιÄτ’ Ψ K β γ δ : κινοι c. ras. E1 : κινοι η E2 A(apud S[1240.19]) Pa(384.24) 2 \x³sÊIÂ] \x¬ sÂI³ Bad

Bad (for

3 a‚J iÖ scripsi (= ουÍ τε ... µο ριον Π = neque ... pars C) :

iÖ ... ~gIÒ  ... iÖ

Hi Φ : Xv

= αÆ λλαÁ µηÁ ν ουÆ δεÁ v. Glossary, and 238b13 = 701.7 Bad)

ÑJÏ addidi (= τουÄ [... κινουÄ ντος] Π = eius quod [... mouet] V) : om. Φ : ex s’o Ca : ex sub’o Cp : ex scd’o Cv 4 2ÙHÒÖ om. C : ÙHÒÖ øÏK om. V ñNÏz addidi (cf. γαÁ ρ [b28] = quod est quia C = quoniam V) 6–7 \M …ÏhtI ... ñTÒD addidi (= υë ϕ’ οÏλου ... υë ϕ’ [b30] = a seipso toto. si igitur mouetur quia [quia CaCv : quando Cp] aliqua pars eius mouetur ex [ex CpCv : a Ca] C

= a seipso toto. si igitur mouetur ita quod aliqua pars ipsius mouetur ex V) : om. hom. Φ

ñsfN ... ñNÏÑËF om. hom. Φ in textu : [damaged] * * * dFÒ ÒDÒ a‚`A ÒÚE YÑK (sic) YÑF [damaged] * * * N ÒD suppl. Φ i.m. with corresp. marks i.m. and after ñsfN : hec enim pars

8-p. 52.1

quando separatur [quando separatur Cv : quando separatur in textu + sit singularis i.m. Ca : sit singularis separatum Cp] ipsa est quod mouet seipsam C : quoniam ista pars singulariter mouet seipsam V

257b25

257b30

(iLÑPÕpSrt fÏ yibpLA û4LA êÑtK \M nMÑoLA LÑqmLA

52

Cp 192r

4

Av. t. 43 8

ÑE¬Ñ  B ñTÒD \M …ÏhtI á¬Ñ  B ÕE YÑK YÒËÖ .jF á¬Ñ  B ÕE ÑMÏÑ  F 〈 ñsfN〉 (iL †LD YÑK DÑËF .¿yLA „IpB щÒÖD fE …ÏW ;NÏÒË áÚE NÑK …ÏU ñgyB YÕkI DÒË øÏkLÑF .ыÑqlT \M …ÏhtT (iL ÑeNÏÒ  èʂnÊn³lÊF ~IÏrÖ² ÕE á¬Ñ  B aLA YÑK YÒË ~gIÒ  〈 Ö 〉 .…ÏhtI ñgyBÖ …ÏhtM ¡V ÕEÖ ñlÑqlT \M …ÏhtI DÒË êÑF …ÏhtI ñgyBÖ …ÏU ñgyB ñNÏÑËF ñsfN …ÏU .Ò \MÖ …ÏhtM ¡V ÕEÖ ÑMÏÒËÖ á¡V \Y …ÏhtI ÕEÖ ÑMÏÒË …ÏU aLA YÑK ;LÖ ÕE cÚLA YÏ ÑËF jÍ CÒ  ~miW …ÏU i ÕEÖ ÑMÏÒËÖ …ÏU ÕEÖ ÑMÏÒË …ÏhtI aLÒÖ ÇÒË …ÏhtM \MÖ …ÏL ñNÏÒ  ÇÒË …ÏhtM ¡V \M YÕkI YÒ  úO ñTÒD …ÏL 1 (b32) αυë τοÁ vel εë αυτοÁ Π(praeter Eb) : αυÆ τοÁ E, fort.ΨΩ, b, cf. Philop. (Rashed 2016, p. 768 [Scholie 28, l. 14–15]) E )Ψ : om. E 1

2 (b34) αÆ ναγκαιÄον α K β δ : αÆ ναγκαιÄα γ Λ 1

l

Pa(385.3), cf. A (frag. 635), Rashed p. 556, et Pel 4131 ad loc. s

αυÆ τηÁ ν E, fort. Ψ, K δ NR : εë αυτηÁ ν Q Pa(385.4) 2

ΑΒ K

2

5 (a3) αυë τηÁ ν L FHIJw :

(a4) ΑΒ Π(praeter E2K)Ψ : τοÁ Β E2 : τοÁ

(a4) τε Π(praeter E β) : om. E β : non vert. Ar.

E : Α πρω τως δ

3 (a1) οÃν Π(praeter

4 (a2) µο νως ... ειËναι] om. E Ψ K S (1240.22 mss. AFM) Sp(1242.15)

1

2

6 (a5) Α Π(praeter Eδ)Ψ : ΑΒ

7 (a5) κινειÄ τοÁ µεÁν Π(praeter Lδ)Ψ : τοÁ µεÁν κινειÄ L δ

εë αυτοÁ Π(praeter Eb)Ψ : αυÆ τοÁ E b

9 (a7) αυë τοÁ vel

(a7) εÍ τι β γ Λ : om. α K δ Pa(385.11), fort. As(frag.

637) 2 ;NÏÒË clare Φ : iterum C (= ~gIz) : etiam V (= ~gIz)

DÑËF Φ (= ωÏ στε ειÆ Π) : si autem C : et si ÒDÑËF scr. Bad 3 …ÏU] …U Φ : …W Bad 4 ~gIÒ  Ö scripsi (= εÍ τι Π = et rursum C = et etiam V) : ~gIÒ  post ras. (ex ;NÒË ?) Φ2 5 ñNÏÑËF scripsi : tunc C (non vert. V) : árÑQ (nisi árÑB vel ñBÑB) parum clare Φ : rÑT Bad V:

4 ñgyBÖ] with corresp. marks s.l. and i.m.: ñsfN …ÏU ñl҂Jz \M ÙHÒÖ øÏK YÏ z ÕE ‡jkLA IÙqT Φ i.m.,

fort. ex comm. Alex., v. Alex. frag. 636, and p. LXXIV i.m.:

ñnM …ÏdA a‚`A yI Φ i.m.; cf. Simpl. 1241.33–35

9 2…ÏL] with corresp. marks s.l. and

258a

258a5

VIII 5, 257b32 – 258a18

Cp 192v 4

8

53

~KÏL Ò \kilF .ñL „fTρ \IMi Ò cÏ Ò  ‹Y ~IrÑJ øB ~KÏL Í rÖ² (iL ñNÏÒ  ~KÏhtM JÖ J ñilY ;L ~KÏLÖ Ò \Y ~KÏhtM êÖ …ÏhtM ¡V ñNÏÒ  ÇÒË †LD ŠÒË èÖqI ÙQ YÑK DÒË Mi Ò YÏ ÑËF .jÍ CÒ  aL …ÏL ¡VÖ ê \Y ñlÏK DÒË jBÑF .×qF ÙHÒÖ ×SÏÕtl ñL¨nlF ÙHÒÖ \M ¤KÒ  ÑpSÏÕtm〈 B 〉 Ò YÏ Ò  †LDÖ ñTÒD ÕE ~KÏL êÒ YÑK J ÑnpqSÒ  YÒË Ñn¬ kL .ñTÒD …ÏU ÕE jÍ CÒ  YÕkI (iL øB ñTÒD ÕE …Ïhl (iL J YÑKÖ …ÏhtM êÖ …ÏL ñNÏÑ  B …ÏU ;NÏÒË ê YÏ Ò  †LDÖ .Ò \M ÕlX ñTÒD …ÏU ~gIÒ  jB iÖ ~KÏhtM cÚLA ÕE áÙHÖ êÒ YÏ i  ñsfN „ ñnM ÑM a‚J \Y i á¡V \Y …ÏhtM 2 (a9) δεÁ 2 fort. Ψ, K β δ FHIJ2Q : om. E J1w E1

(a12) τοÁ Γ Π(praeter E1) : τουÄ το fort. Ψ

E K β δ Pl(836.6) Pa(385.18) ουÆ Π(praeter E2)Ψ : exp. E2

post ras. b

3 (a11–12) ηÏ ξει ... τοÁ Γ Π(praeter E1)Ψ : om. 4 (a11) πλειο νων Ψ γ Λ Pp(836.9) : πολλωÄ ν

5 (a14) µεÁν γαÁ ρ Α E2Ψ β γ δ : µεÁν γαÁ ρ E1 : µεÁν Α K

6 (a14)

7 (a16) ΒΓ Π(praeter EB)Ψ : Γ cum ras. E1 : ΑΒΓ E2 : ΒΓ

(a16) αυÆ τηÁ ] non vert. Ar. (cf. Sp 1243.5)

γαÁ ρ F : Ψ = EK(etc.) ∨ F : τοÁ δεÁ γαÁ ρ N : τοÁ δεÁ R H

(a16) τοÁ γαÁ ρ E K δ L IJQw : τοÁ µεÁν 8 (a17) αÍ ρα K e2 β γ Sp(1243.6)

Pa(386.3) : γαÁ ρ α be1

JÖ J scripsi (= Γ τουÄ το δεÁ Π = c¯ et c¯ CV) : HÖ d vel HÖ H Φ : H g Bad 3 aL …ÏL a(B …¬ hÂM³ scr. Bad Mi Ò YÏ ÑËF scripsi (= res igitur C) : MjL YÑF Φ : MuL YÑËF Bad 4 ÑpSÏÕtl scripsi (= διαÁ Π = per media C = cum mediis V) : ÑpSÕtM Φ Bad 5 J Φ (= c¯ CV) : H Bad 6 J Φ (= c¯ CV) : g Bad 7 jB scripsi (= b¯c¯ C) : hB Φ Bad : b¯ V 8 ÑM a‚J] scr. Bad : ÑMH Φ : + ñsfN „ cz Φ s.l. 2

Φ(=ΠCV) :

iÖ …ÏW (fnLÑF ÙHÒÖ ypqB áÚE è‚NÖ êÕoLÒÖ YÙbLÒÖ (fnLA L¨M ° ê Ò è‚N ñKÏU jF ñilY êÕoLA YÕkI i ÙQ ñNÏt êÕoLA Yz ~miW …ÏU Yz úO (iLÖ YÑKÏhtI êÕoLÒÖ YÙbLA〈Ö〉 …ÏhtT 4 gloss without mark:

Φ i.m. ex comm. Philop. 835.37–836.5

258a10

258a15

(iLÑPÕpSrt fÏ yibpLA û4LA êÑtK \M nMÑoLA LÑqmLA

54

3

Cv 74v Cp 193r

ñL øcH ÙQ ñTÒD …ÏU ÕE cÚLA YÕkI YÒ  Í rÖ² úO ÙqF .ñTÒD …ÏU ÑM Í rÖ² ÎL ~KÏL (iL ñNÏÒ  ÇÒË …ÏhtmLÒÖ …ÏhtM ¡V ñNÏÒ  ÇÒË …ÏdA …ÏdA YÑK YÑËF .XvÒ ;EÙHÒ  %Ï ÑM ÖÒ  ñbÂHÑC ;ŽM ÙÈHÒÖ øÂKÂÑW aLA YÏ Ò  \iÏbLA \B (øctÏM Í rÖ² ñNÏÑËF …ÏhtmLA ÑMÏÑ  F) jÍ ctÏM ñTÒD …ÏU ñNÏÒ  YÒ  fcLA áچ ñgÂyB YÏ Ñ  B i ñTÒD ~KÏL YÕkI ñlÏK 1 (a18) εë αυτοÁ 2 (vel αυë τοÁ ) Π(praeter EK)Ψ : αυÆ τοÁ E K

2 (a19) καιÁ Π(praeter E)Ψ : om. E

4 (a21) µεÁν α K A (frag. 641) S (1244.14) : om. β γ δ Λ c

(a22) αÏ ψεται εë κα τερον εë κατε ρου

LN I : om. α δ K N R FHJQw Λ

(a22) δηÁ FHIJ : δεÁ β : δηÁ vel δεÁ Ψ : om. E K δ J1Qw Λ

2

2

s

1

5 (a23) αυÆ τοÁ 1 non vert. Ar.

2

(a23) αυÆ τοÁ 2] non vert. Ar.

(a24) αυë τοÁ 1 (vel εë αυτοÁ ) Ψ β δ

FHIJQ : αυÆ τοÁ E K w : αυë το τι Λ

3 øÂKÂÑW ÑM Φ, cf. dum conuenit C = dum ... conuenit V :

øÏK ÏÑW ÑM Bad %Ï ÑM ÖÒ  scripsi : ÑSÑM ÖÒ Φ : ÏÑW ÑM ÖÒ  scr. Bad XvÒ scripsi : XjL Φ Bad 4 …ÏhtmLA scripsi (= τοÁ ... κινου µενον Π = illud quod mouetur C = motum V) : …ÏdA Φ Bad 5 YÏ Ñ  B i scripsi (= ουÆ τωÄì Π = non quia   Bad V : non quia ... non quod dittogr. C) : YÑB Yi Φ : YÑK Yi ñNÏt Ñ= %Ï ;l (iL `ÒÖ ñiF øyfT ÑeNÏt jlL SÏ;M (fnLA (leg. cz ?) YÒ ÑeNÏt aÑgYuL (Bad ÑE¬{B : Φ ÑEH}B : scripsi) ÂXÂ{B ÕE `A \M KhlL (fnLA øyFÖ ÑF øyfI i ÙiLÑB jaA (Bad O : sic Φ) …ÏW ;K Í qiqH ñiFÍ øyfT Φ i.m. (cf. Philop. 836.15–17, Simpl. 1243.22–28) 3 ;EÙHÒ  ] with corresp. marks s.l. and i.m.: (Φ lbSÒ : scripsi) !otS ;NÏÒË ñNÏÒ  zN ³ `A ñsÏm iÖ ñKÏU cÚLA  `A (Ï m YÕkiF 〈ñL〉 J i YÕkI Yz RÕO ÙQ …ÏdA YÏ t †LÚB (Bad tntSÒ) Φ i.m., cf. As(frag. 640) et Rashed ad loc., Philop. 836.16f., Simpl. 1243.25–27 4 øctÏM] with corresp. marks s.l. and i.m.: Yz \kMz ÒDÒË ›SÏ0M …ÏhtmLÒÖ ÕE YÑK ÆctÏM …ÏdA YÑK YÒË ñNÏz yI ;EÙHz øB ñbHÑcL ÑS­ ;M ;ŽM ÙHÒÖ øÏK \kI G (fnLÑK ÆctÏM øYÑfLA \kI G YÒËÖ XvÒ „ ;ŽM ÙHÒÖ øÏK øyfI ÑnlQ ;K XvÒ %Ï Ñm Φ i.m., fort. ex comm. Philop. (v. Rashed 2016, p. 770 [Scholie 30, l. 4–6]) 3 gloss without mark:

258a20

55

VIII 5, 258a18-b1

Av. t. 44 5 Ca 26vb

ÕE ñnM ~miW YÏ Ñ  B …ÏLÖ …ÏhtM ñNÏÒ  ¡V ñTÒD …ÏU ÕE á¬Ñ  B aLA øB iÖ …ÏU [ê217] á¬Ñ  B (iL ñNÏÒ  †LDÖ .…ÏhtmLA ÕE ñnM ~miWÖ …ÏdA .×qF ê eJ \M …ÏhtIÖ Ò eJ \M …ÏU ñnÏkL …ÏhtI á¬Ñ  B YÒË) Ò \M  qN YÊ ÒË cyW iL !ymLA ÒÚE „ ñiF †kÏwtI ÑJÏ Ö (…ÏhtmLA ÕEÖ) ê \M Öz (…ÏhtmLA ¡V …ÏdA ÕEÖ jÍ ctÏM YÑK YÒË ñNÏÑËF ?~KÏhtM ê \M
ÑbLA Öz ~KÏL Ò \M
ÑbLA YÕkI øE ÑM Ó ñNÒ  †LDÖ ñlÑqlT \M ÅÖÏ〈 z 〉 …ÏhtmLA êÒ ñilY cÚLA \kI G †LD YÑK (iL ñNÏÒË †LD „ èÕqnF .ñTÒD ~KÏL
ÑbLA êÒ Ù³yB BÖ êÒ \M  qN ÙQ (…ÏhtmLA ÕEÖ) ;EÙHÒ  ÖÒ  ;ŽM ÙHÒÖ øÏK ÕÏqLÑB YÕkI YÒ  knM 1 (a24) κινειÄ Π(praeter βH) : κινειÄν β H

(a24) τε K δ Sp(1244.34) : δεÁ α β FHIJQw Λ

3 (a26) κινειÄται αÆ λλαÁ Π(praeter E1)Ψ : om. E1 ad l. 2–3) : om. K δ FHIJw

(a26) ηë 1 E2 β Q : ηà E1 : ηÎì Ψ(v. also App. 3

(a26) ηë 2 E2 β : ηà E1 : ηÎì Ψ(v. also App. 3 ad l. 2–3) : om. K γ δ

(a27) post µο νον add. (ex gloss.?) τοÁ δεÁ Γ ... αÆ δυ νατον γα ρ Π(praeter α), fort. Pa(386.16) : om. α γρ.A(apud S[1245.3–5]) (cf. notes Ross 703, WiCo 336f., Wag 682, Rashed 556) 5 (a28) κινουÄ ν µεÁν Π(praeter E2)Ψ : κινου µενον E2 (cf. Sc[1244.24, 1245.18])

YÒË scripsi (= εÆ αÁ ν [αà ν bHB] Π = si CV) : gÏ Ò Φ : gÏ Ò  Bad 5 \M Öz scripsi (= ηà τηÄ ς Π = aut ex CV) : \MÖ Ò Φ Bad 6 Öz scripsi (= ηÃ Π = aut C) : Ö Φ (= et V) Bad 7 ÅÖÏz scripsi (= πρω τως Π = in primis CaCv : a primis Cp = primum [sc. motum] V) : iÖ Φ Bad 9 ;EÙHÒ  Φ : ÑEÙHÒ  scr. Bad 1

ÕE] ÕEÖ scr. Bad

1

4

1

2–3 gloss without mark: …ÏhtIÖ Ò Yz eJ \M …ÏU ñnÏkL ~gIz ÕhnLA ÒÚE ‹Y èÕqLA ÒÚE øqnI Yz yqtsI

ê Yz eJ \M Φ i.m. (supporting ηÎì ... ηÎì [a26] with an implied form of ειËναι) corresp. marks s.l. and i.m.: ÆctÏM …ÏdA ¿fN Φ i.m.; cf. Simpl. 1245.18

5 ÆctÏM] with

258a25

258a30

258b

(iLÑPÕpSrt fÏ yibpLA û4LA êÑtK \M nMÑoLA LÑqmLA

56

Cp 193v

5

ÚmniH ñybP \kI G ˆÂsQ³ YÒË ñnÏkL .qnl (ilF è;ktSiÑB ÑMÏÑ  F ~msqnM Â Í 4qnmLA 〈 „ 〉 YÕkI YÒ  knM (iL †LÚlF .ñniyB õ bpLA †LD …ÏhtM ¡V èÖÏi Ò …ÏdA YÏ Ò  †LD \M ezI ÙqF .~iÏLÖÏÒ  ÑM a ÕÏqLÑB YÏ ÑËF ÑM a \Y …ÏhtI YÑKÖ êQ³ \M žqI YÑK YÒË …ÏhtmLA YÏ Ò  †LDÖ Í YÒËÖ …htM ¡V èÖÒ  ŠÒË ñB èÖqI Mi Ò ÒÚE YÏ Ò  i¬ ÒË …ÏhtM ŠÒË fI³ YÑK Ï Ï YÒ  ~yi? e`A áÚE \MÖ e`A †lT \M ‡‚lI ÙqF Ñef³QÕI³Ö ñTÒD …ÏU ÕE [11ê217] .…ÏhtM ¡V ÑelÏK ÑKÏhtmLA „ èÖÏi Ò YÕkI 1 (b3) ϕυ σιν α K δ Sp(1246.16.25) : δυ ναµιν β γ Pa(387.4)

3 (b4) εÆ νειÄναι e FQ1 : εÊν ειËναι

HIJ1 Λ : εÊν ειËναι τι K J2w : εÆ νειÄναι τι E1 NR Q2 Sl(1244.12) : ειËναι τι E2 b L : Ψ = E1NRQ2 ∨ E2bL : τοÁ ειËναι Pa(387.6–7)

5 (b6) τοÁ ] om. Ψ

7 (b8) πρω τως κινουÄ ν β γ δ Pp(887.6)

Sl(1247.11, ms. a) : πρωÄ τον α K Sl(1247.11, mss. AFM) β : εÆ ν αÏ πασιν e γ : Ψ = EKβ ∨ e γ 2

2

(b8) αÏ πασιν be1 : εÆ ν παÄ σιν E K

(b9 fin.) v. App. 1 ad 256b13

YÕkI] s.p. Φ : YÕkT Bad „ addidi (= εÆ ν [διαιρετοιÄς] Π = in [divisibilibus] CV, v. also 4 êQ³ \M žqI scripsi (= ευÆ θυÁ ς ιÏσταται Π = stat ex proximo C [cessabit 6 Ñef³QÕI³Ö scripsi : ÑefFÕBÖ Φ : ÑefqIÖ V]) : êQ \M žfB \MÍ žfB Φ : êQ \M ñtbTM žqT Bad  Bad 2

App. 3 ad loc.)

2 4qnmLA] with two different marks s.l. and two glosses i.m., gloss 1: ñLÕQ ‹Y žqI rÙnkSiÒ

iÍ ÖÏÒ  ) iÍ ÖÏÒ  ÕE a cÒ  ÕÏqLÑB ˆÏ ÖÏÒ  ÕE a 4qnmLA „ YÕkI YÒ  õnm (ilF ‡jkLA IÙqtF 4qnmLA (iLÖ ÕgyLA õpQ øbQ iÍ ÖÏz YÕkI YÒ  \km ÕE ÒÚE YÏ i  ÕgyLA õpQ ÙyB ÕEÖ ñTÒD \M …ÏhtM (Φ èÖÒ : scripsi iÍ ÖÏz øyfLÑB ÕE Φ i.m. (cf. Alex. apud Simpl. 1246.16–26); gloss 2 without corresp. mark: U (Φ ÑF : scripsi ñiF) ñiF ÕÏqLÑB (Bad mgIÖ : sic ?) \mÏgIÖ ÕÏqLÑB ~msqnM cÒ  ÕÏqLÑB ñLÕQ ‹Y žqI ÕÏqLÑB Ñi­ LÖÏz ÕÏqLÑB 4qnmLA Φ i.m., fort. ex comm. Philop. 837.25–27(apud Ibn Adı¯?), v. also Alex. frag. 646, Giannakis p. 79f. ad loc.) 3 †LD \M] with corresp. marks s.l. and i.m.: yI ñlÏHÖ †ÏwLA Mi  MÙÏqtmLA èÕCtÒ \M Φ i.m., cf. Alex. frag. 647 and Rashed ad loc.

258b5

VIII 5, 258b1 – 6, 258b17

Av. t. 45

Av. t. 46

5 Cp 194r

57

øÏxÂT³ iÖ ÙBÒ  YÕkT YÒ  ~bJÒÖ KaA NÑK ;LÖ [3ê218] YÑK ÙHÒÖ …ÏU cÚLA èÖÏi Ò ÕE a ÑnEÑE YÕkI YÒ  Í rÖ² úO ÙqF \M ÙHÒÖ øÏK YÑK YÒ  ÑMÏÑ  F .…ÏhtM ¡V èÖÏi Ò …ÏdÒÖ ¡oK ÖÒ  ÒÚE ñNÏÒ  ÑMÍÏ Ñ  F .ñlibsB \V .F øXÙI ÑJÏ (ilF ~iÏLRÒ  …ÏhtT i LA ÑKÏdA øÏK \M rÑX …ÏhtM ¡V ñsfN „ ÕE a ÑnEÑE YÕkI YÒ  Í rÖ² úJÒÖ 〈 ‹Y 〉 \iÏbT ÙqF X   | ~miW …ÏU ¿yLA „IpBÖ ƒjPiË Ò ‹Y iÏvT YÒ  љyB „ \km †LD YÏ Ò  mW YÒË èʂnÊn³L .znLA \M 〈 ñJÕLA 〉 ÒÚE   2 (b11) αÆ ΅ιδιον οÊ πρωÄ τον κινειÄ scr. Ross (cf. Sp[1250.35, 1251.3]) : οÊ πρωÄ τον κινειÄ E1Ψ K FHIJw (?)Q1 Λ : αÆ ΅ιδιον οÊ κινειÄ E2 β δ : οÊ πρωÄ τον κινειÄ αÆ ΅ιδιον Q2 : αÆ ΅ιδιον habent As(frag. 648, 650) Pa(387.26); cf. Torstrik, p. 511 ειÍτε ουË ν β γ Λ

(a11) ειÍτε1 E, fort. Ψ, K δ As(frag. 648) Pa(387.26) :

4 (b12) αÆ ΅ιδιον Π(praeter E1)Ψ : ιÍδιον E1

Pa(388.2) : τοÁ αÆ ειÁ (vel αιÆ ειÁ ) E K β δ 2

5 (b14) τοÁ E1Ψ γ, fort.

(b14) πα σης α β P (837.35) Sc(1251.17) : πα σης τηÄ ς

γ δ : καιÁ πα σης K Pa(388.2) (cf. notes HaGa, Ross)

c

6 (b15) καιÁ 2 Π(praeter E1)Ψ : om. E1

7 (b16) δε (vel δ’) Π, fort. Ψ(cf. 234a26.b24, 266a15.b8 with fa-l + Jussive = εÍ στω δη ) : δη scr. Ross

;LÖ add. ylytLA Φ : 〈ÓLÑoLA øcfLA † rÑomLA †wLA øÏH èÖi Ò …dA ‡³ÙÂQ〉 6 〈\MÑoLA〉 ylytLA (iLÑPÕpSrÒ  èÑQ〉 Bad : om. ΠCV : seclusi øÏxÂT³] s.p. Φ : øÏhÂT³ Bad; cf. 220a8, 226b28, 228b4 2 post èÖÏi   Ò add. † Φ Bad : om. ΠCV : seclusi 3 YÒ  scripsi (v. GALex I, an 1.1) : YÒ Φ : YÒË Bad 4 ÑKÏdA scripsi (= κινου ντων Π(praeter J1) = mouentibus C = mouentium V) : ÑKhtmLA Φ (= κινουµε νων J1) Bad 6 …ÏU scripsi (= moueat C = mouet V) : …L Φ Bad 6–7 znLA ... ‹Y scripsi (= ωÎ δε σκοπουÄ σιν Π = hoc modo considerationis C [hoc modo V]) : èÖiÒ znLA \M ÒÚE Φ Bad (cf. 256b3) 1 ante

〈

6 gloss without mark:

YÕkI iÖ YÑMR „ (Φ ŠÕielL : scripsi ŠÕi=A) ŠÕi=A ³ rÕcLA øcÏhT³ (iL ñNÏi 

Ñ=ÕcH èÑH „ dÑsF Φ i.m., fort. ex comm. Alex., v. Alex. frag. 653, and above, p. LXXIV

Chapter 6 258b10

258b15

(iLÑPÕpSrt fÏ yibpLA û4LA êÑtK \M nMÑoLA LÑqmLA

58

4

Cv 75r

8

Cp 194v

„ÊlXÊÒ  ñNÏÑËF dÑsF iÖ YÕ° kT ¡V \M ~niH dÕJÕM ¡VÖ ~niH dÕJÕM YÕkI YÒ   YÒ  ~niH dÕJÕM ¡VÖ ~niH dÕJÕM YÑK "M n‚® jtl (iL ;B YÕkI .iÏvT ¡V \M YÕkI YÒ  ~bJÒÖ ñ³M³ÙÂY ÖÒ  gcA ÒÚE gJ ÑM ø® K dÕJÖ KÏhtM 〉 ¡V fE LA ~gIz ndÑbmLA ¡yB YÕkI YÒ  \km ÙQ ñNÏÒ  èʂnÊn³lÊF YÒËÖ ~nkJ (iL ñNÏÒ  ÇÒË dÕJÕM 〈 ¡V ~niHÖ ~niH Í dÕJÕM KÏL ÑeNÏÒ   ÇÒË ÑM ~bbS ÑnEÑE YÏ Ò  \iÏbLA \M ñNÏÒ  †LDÖ .ÒÚkE ÑelÏK YÕkT YÒ  †LD ÑnL‚NÒ  ñTÒD …ÏL øÏK YÏ ÑËF .~niH ÑEÙqFÖ ~niH ÑEdÕJÖ „ щÒÖD ÑKÏhmlL ÑMÏÒ  Ö …ÏhtI (ilF z‚¬ jtI i ÑM YÑK DÒË ˆÈzÂY ñL YÕkI YÒ  Í rÖ² úJÒÕF ÑMÏÑ  F .ˆzÂY ÒD YÕkI YÒ  áÕJÕLA \M ñJÕB ÑnlQ ÑJÏ úO (ilF …ÏdA  ÕE †LD (ilF ÙM¬ †LD YÏ Ò  Ö ÙsfI љyBÖ YÕÏktI љyB YÏ Ò  „ úbsLA dÕJÕLA qÒd fE iÖ 〈 KÏhtM 〉 siL ÑeNÏÒ  ¡Â V fE LA \M ÙHÒÖ 1 (b18) ϕθοραÄ ς Π(praeter E1)Ψ : ϕοραÄ ς E1

Π(praeter E b )Ψ : οÊ E : om. b 1

1

1

δ (cf. Denniston 140–44)

1

2 (b18) οë τεÁ Π(praeter E1)Ψ : οÊ E1

6 (b23) αιÍτιον τι Π(praeter E1)Ψ : αιÍτιον E1

E δ, fort. S (1252.13f.) : εÆ στι E2Ψ K β γ Λ 1

(b19) οë τεÁ

5 (b22) γε πα σας E, fort. Ψ, β T(223.21) : πα σας γε K γ 7 (b23) εÆ στι [ν] τι

(b23) τουÄ Π(praeter E1)Ψ : το E1

p

(b24)

εë αυτοÁ (vel αυë τοÁ ) Π(praeter E Λ)Ψ : αυÆ τοÁ E : om. Λ 2

2

1 ñNÏÑËF scr. Bad (= γαÁ ρ Π = nam C = enim V) :

Îg‚jtl Bad Ö secl. Bad

4 ~gIz secl. Bad

ÑQ vel ÑB Φ 2 n‚® jtl scripsi : hbl Φ : 4–5 ¡V ... KÏhtM addidi (= αÆ κινη των ... οë τεÁ

δεÁ µηÁ Π = immobilia sed sunt mouentia, quandoque existant et quandoque non C = non mota

sed sunt mouentia, sint quandoque et quandoque non V) : om. hom. Φ

6 ~bbS s.p. Φ (= αιÍτιον Π = causa V) : res (~miW) C

5 2dÕJÕM] dÕJÕM Φ

11 ÙHÒÖ] ÙHÒÖ scr. Bad

KÏhtM addidi

(= τωÄ ν αÆ κινη των Π = quae sunt immobilia C : quae sunt non mota V) : om. Φ Bad

11 ÙHÒÖ] + "M (i.e. Matta¯ ibn Yu¯nus?) s.l. + note i.m. without mark: YÕÏkT „ úbsLA (iL ñNÏÒ  yI

X| úbS øB юM ÙHÒÖ ÕE ¡yB dÑsFÖ %ÕfnLA ¡yB Φ i.m., cf. Alex. frag. 656, 658 ad 258b27

258b20

258b25

59

VIII 6, 258b17 – 259a2

3 Av. t. 47

6

áÚE ÑM YÏ Ò  †LDÖ .ÑE¡Â V …ÏW †LÚBÖ ÙBÒ  щÒÖD …ÏW áÚE YÕkT htÏH wÒÙLA alL ~BÑbSÒ  YÕkI ÑelÏK iÖ ÑŽM ÙHÒÖ ÙHÒÖ 〈 (iL 〉 ñlibS áÚE ÑMÏÑ  F rÖ² MRi iÏLRÒ  áÚE эlY LA èÑaA áÚE YÏ Ò  †LDÖ øctÏmLA ÑnEÑE NÑK ÕL ñNÏz DÒË YÑB ÙqF .~yM ÑelÏ K d³ÕJÖ (iLÖ Iш jbF ÑelÏK \M ¡oKÖ …ÏhtT iÖ …ÏW LA ndÑbmLA \M Í ¤K *W i  ndÑbM …ÏU …ÏhtM ¡V ÕEÖ ÒÚE YÑKÖ ÔÙWÖ dÕyT vÏ ÙsfT щÒÖD ÑKÏdA 1 (b28) ουÆ δ’] ωÏ στ’ fort. Ψ (cf. ουÏ τως ωë ς ... Sp[1252.26])

(b28) αυË τωÄ ν αÆ ειÁ E K e β FHIJQ Λ,

fort. Pa(389.1–2) : αυë τωÄ ν αÆ ειÁ Ψ : αυÆ τωÄ ν αÆ ειÁ b w : αυË τωνδιÁ scr. Ross (cf. Ross 705f., Pel 417 ad loc.)

(b29) ταδι E2 β γ δ Λ : αυÆ ταÁ E1 K Pa(389.2) : αυÆ ταÁ vel ταυÄ τα Ψ

κινου ντων ΠΨ S (1256.26) : fort. om. P (838.15) : secl. Ross p

p

FHIJ Qw Λ : τωÄ ν J S (1252.26) : διαÁ του των E Ψ 1

2

p

(b29) του των E K β δ

2 (b30) εÏ καστον αυÆ τωÄ ν Π(praeter

1

(b30) ουÍ τε πα ντα αιÍτια α K b

E ) : εÏ καστα αυÆ τωÄ ν E : εÏ καστα τοιου των interpr. Ar. 2

2

e[αιÍτιον] β Sc(1252.32) : αιÍτιον ουÍ τε πα ντα γ, fort. Pa(389.4–5) c

S (1253.6) : non vert. Ar.

(b29) 2

3 (b31) καιÁ 1] om.

4–5 (b32) εÍ νιαι αÆ ρχαιÁ E K be β γ Λ : εÏ νι αÆ ρχαιÁ E1 : εÍ νιαι ωë ς 2

2

(?)e1 : ειËεν (vel ειÆ σιÁ ν) αÆ ρχαιÁ fort. Ψ : [τοιαυÄ τα] ειÆ σιÁ ν αÆ ρχαιÁ Pa(389.9) : εÍ νια (secl. αÆ ρχαιÁ ) Ross

5 (a1) µεÁν Π(praeter E1) : om. E1, fort. Ψ

κινου ντων E1 J

(a1) κινουσωÄ ν Π(praeter E1J)Ψ :

6 (a2) ταÁ Π(praeter H) : τιναÁ H : ειËτα fort. Ψ

2 (iL addidi (= ουÍ τε [... ουÍ τε] Π = non [... neque] CV) : om. Φ Bad 4 jbF ] jB scr. Bad dÕJÖ scripsi : ÑEÕHÖ Φ with superscr. d + sign ind. del. of final Alif (and Ha¯’?) : ÑEdÕJÖ Bad 6 YÑKÖ] YÑKÖ vel YÑkF Φ : καιÁ Π = et CV : YÑkF Bad

1

щÒÖD] with corresp. marks s.l. and i.m.:

(v. ταÁ µερικα , Alex.[frag. 658]) Φ i.m.

3

шrÑqT LA ‡ÑsJtÒ Ñ‰ÒÖÚB yI áÚE] with mark s.l. and gloss without mark: yI (vel

ÑeQrÑfT

?)

3

ÙsfTÖ YÕÏkT LA (vel Ñil‚`A ?) ÑKaA Φ i.m., cf. Simpl. 1253.12–14

258b30

259a

60

Cp 195r 4 Av. t. 48

8

(iLÑPÕpSrt fÏ yibpLA û4LA êÑtK \M nMÑoLA LÑqmLA

YÕkI YÊ i  áÕJÕLA \M ñJÕB jÍ I‚M †LD \kI G X| ~miW …ÏU ᳡³ VÖ ÒÚE „ úbsLA ÕEÖ aÑiWi Ò \M ÙHÒÖ øÏK „ †LD ‹Y ~gIÒ  ømtwM a ÑnEÑE ~bbS ÒÚE YÕkI htÏH øctÏÍ mLA iÏvtLA ÒÚE „Ö ¡ybLA ÙqÊFÖ ¡ybLA dÕJÖ .XÂiªÒ lÑS Ka ~BÑbSÒ  áÚE YÕkTÖ áÚE Ka ~iÏLRÒ  ~gIÒ  YÕkI èÖÏi Ò …ÏdA YÏ ÑËF Í iÏLRÒ  KaA NÑK [Ò 219] YÊ ÑËF ÙHÒÖ \M ¤KÒ  ÑiÏLRi ÑF ÙHÒÖ \M ¤Â KÒ  YÑK YÊ ÒËÖ .ÙHÒÖ YÑK YÊ ÒË Í ÙqtyI YÊ Ò  gHi ÑF ¡oK YÑKÍ YÒËÖ ¡oK i ÙHÒÖ ñNÏÒ  ÙqtyI YÊ Ñ  B gHtÒÖ Í ÙHÒÖ NÑK ÒDÒË „HÒÕlLA YÏ Ò  †LDÖ .〈 iEÑntM 〉 ¡V i iEÑntM ÑeNÏÒ  1 (a3) το δε Π(praeter E1)Ψ : το δε δεÁ E1 : τοÁ δεÁ Λ

2 (a3) οÊ Π(praeter E1)Ψ : οÊ καιÁ E1

(a3–4) οÊ περιε χει καιÁ τουÄ το παρ’ εÏ καστον i’punxit Ar.

(a4) τουÄ Π(praeter E1)Ψ : τοÁ E1

5 (a6–7) αÆ ΅ιδιος Π(praeter J1w)Ψ Am(88.15) : αÆ ΅ιδιον J1w Λ Am(88.15) : εë νιÁ E1

6 (a7) εÏ ν. ειÆ Π(praeter E1)Ψ

7 (a8–9) καιÁ πεπερασµε να] και , ειÆ πολλα , πεπερασµε να intell. Ar.

8 (a9–10) συµβαινο ντων Π(praeter E1)Ψ Am(88.17–18) : συµβαινο ν E1 1 jÍ I‚M] s.p. Φ : sin˜ CaCv : su˜ vel sıı˜ Cp : faceret V

4 ~BÑbSÒ  ] causa CV (= ~bbS)

6 post 1ÙHÒÖ

ÙHÒÖ ñNÏÒ  ÙqtyI YÑ  B gHi ÒÖ Φ Bad : om. ΠCV : seclusi ÑiÏLRi ÑF] ÑiLÖÏi ÑF Bad 7 YÊ Ñ  B] YÒ  Bad (erroneously annotating: ÙqtyI Yz 〈gz〉 : é; for this marginal note v. App. 3 ad p. 61.1) post ¡oK add. illud enim quo completur illud sunt tres res C : hoc enim complebitur per tria V : om. ΠΦ YÒËÖ scripsi (= καιÁ Π = et si V; v. also Gˇp 215.19) : YÑËF Φ Bad (= si autem C) 8 iEÑntM addidi (= αÍ πειρα Π = infinita C = infinitos V) : om. Φ Bad

add.

úbsLA] with corresp. marks s.l. and i.m.: ÑelÏK Ñ= úbS cz Φ i.m. (ex Philop. 838.20–22?) щÒÖD ÑKÏdA yI Φ i.m. (cf. Alex. frag. 662 and Rashed ad loc.) 8 „HÒÕlLA] with mark s.l. and ‡RÒÖ〈lLA〉 g〈Ñ〉 (partly illeg.) Φ i.m. without 2

4 1áÚE] with a vertical dash linked to the note:

mark, fort. ex comm. Ibn al-Tayyib 875.22 ˙

259a5

61

VIII 6, 259a3–20

Ca 27ra 4 Av. t. 49

7 Cp 195v Cv 75v 10

YÏ Ò  †LDÖ iEÑntmLÑB †sÏiA gHi Ò YÕkI YÒ  ÙBÒ  fvbniF шÑiYÑ  B fEÑntmLA cÏ aÑB эF ÙJÕI YÒ  fvbnI yibpLÑB YÕkT LA aÑiWi Ò i LA aÑiWi Ò èÂÖÏÒ  YÕkI È IÑfK ÙHÒÖ „Ö .\kJ ÕE ÑJÏ øgFi ÒÖ .XÂiªÒ lÑS „ KhlL Ò¡ ÙbMÖ ~iÏLRÒ  …ÏhtT …ÏdA YÕkI YÒ  Í rÖ² úJÒÖ ñNÏÒ  ~gIÒ  áÕllÑQ \V ÑJÏ ezI ÙQÖ KaA YÕkT YÒ  Í rÖ² úJÒÖ ñNÏÒ  ÑnÏiÏB ÙQ ÑNÏÒ  †LDÖ ~iÏLRÒ  ÙHÒÖ ~miW èÖÏi Ò YÏ Ò  †LDÖ Í lctÏM 〈 YÕkT 〉 YÒ  Í rÖ² úJÒÕF ~qÒd NÑK YÊ ÒËÖ .~qÒd feF Í lctÏM NÑK YÊ ÒË \kL .ÆctÏM (ilF ˆÒÕtmLA ÑMÏÑ  F øÈctÏM w ÒÙLA …ÏhtM \YÖ ÙHÒÖ …ÏL \Y YÕkT LA fE ÙHÒÕLA KaÒÖ .ÙHÒÖ ÑE¬Ñ  B KaA †lT (ilF X| ÙyB ÙHÒÖ …ÏU YÑK YÒË ñNÏÒ  †LDÖ ÙHÒÖ .iLÒÕtM fE øB Í lctÏM 2 (a11) δειÄ Π(praeter E)Ψ Am(90.1) : δηÁ E Am(90.2) οÃν E2

3 (a11) τοÁ 2 Π(praeter E), fort. Ψ : om. E, fort.

(a12) εÏ ν α K β Am(90.3) Sc(1254.27) : ειÆ εÏ ν γ δ Λ

4 (a13) οÃν Π(praeter E) : om. E, fort. Ψ, fort. Am(90.3)

(a12) οÊ Π(praeter E2)Ψ : ωÎì 7 (a16) αÆ ειÁ 1 (vel αιÆ ειÁ ) α K

β δ I2Q2 Am(90.8) Sp(1254.34) Sc(1256.12) Pa(390.14) : δειÄ FHI1J2 (?)Q1 w [om. Λ]

(a16) αÆ να γκη α K β Q Am(90.8–9) Pp(838.29) Sc(1256.12) : αÆ να γκη καιÁ δ FIJw : καιÁ H Λ, fort. Pa(390.14)

(a16) καιÁ γαÁ ρ τοÁ ] τοÁ γαÁ ρ Ψ, fort. Am(90.9), Sc(1256.13)

Π(praeter E), fort. Am(90.11) : καιÁ υë ϕ’ α Λ

9 (a18) καιÁ

10 (a19) αÍ λλο1 α K be1 β Pa(390.16) : τι αÍ λλο

e2 HIJQw Λ : τι καιÁ αÍ λλο F 2 ÙJÕI Φ :

ÚXqI (accipiatur) CV fEÑntmLA scripsi (= τοÁ πεπερασµε νον Π = finitum C = pars finita V) : fEÑntLA Φ Bad 7 YÕkT add. Bad (= sit CV) 11 Í lctÏM] scr. Bad (= continuus CV) : ÑcbM Φ 1 †sÏiA] with similar marks s.l. and i.m.:

ÙqtyI YÊ Ò  Φ i.m.

259a10

259a15

259a20

(iLÑPÕpSrt fÏ yibpLA û4LA êÑtK \M nMÑoLA LÑqmLA

62

Av. t. 50

4

Cp 196r

8

ÙQÖ .…ÏhtM ¡V ÅÖÏÒ  ~miW ÑnEÑE Y¬Ñ  B YÑsNiË Ò „oI aÑiWi Ò áÚE \B ›B \M YÏ Ò  EÑZ YÏ Ò  †LDÖ ÑKaA g dÑbM øMÏÑ  T ÕE ÒDÒË ~gIÒ  †LÚB „oI  ÑeNÏÒ  Y ÑB †LÚBÖ .\ksT ~niHÖ …htT ~niH ÑM aÑiWÒ  ÒdÕJÕmLA ÑelÂ Ï K siL Ï †LDÖ .ÙBÒ  …ÏhtM љyBÖ ÙBÒ  \KÑS љyB iÖ \ksT ÑelÏK iÖ …ÏhtT \ksT YÒ  ‹YÖ …ÏhtT YÒ  ‹Y ÕÏQ Ñ=Ö ~yi? \IMi ÑB Ÿ±Ï tT LA YÏ Ò  \i® bN YÒ  ÑNÙcQ YÑKÖ ÙHÒ  øÏkL ~niÏB ñLÑH áÚE ÑM YÑK ;LÖ .áÚE ÂMÒ  \³ i® bT³ ÙBÒ  fE aÑiWÒ  ÒdÕJÕmLA \M YÏ Ò  YÒ  ›fncLA \M ÙHÒÖ øÏK yibP ~gIÒ  øÏK YÏ Ò  ÑnyGÖÖ †LD „ ÑnY®Ö KÏhtM ÙBÒ  fE aÑiWÒ  Ö KÏhtM ¡V …ÏhtM ÑMÏÒËÖ …ÏhtM ¡V ÑMÏÒË aLA ÒÚE YÏ Ò  Ö …ÏhtI ÑM Î \yF …ÏhtM 2 (a21) καιÁ Π(praeter E2)Ψ : exp. E2

(a22) τωÄ ν κινου ντων Π(praeter Q1) Pl(888.17)

Sc(1256.34) Pa(390.20) : τουÄ κινουÄ ντος (?)Q1 : τωÄ ν κινη σεων (nisi τωÄ ν κινου ντων) Ψ : secl. Ross Wag β δ Iw

(a22) µεÁν Π : an γαÁ ρ Ψ? (cf. As[frag. 667])

(a22) αÍ ττα EΨΩ K FHJQ : αÏ ττα

3 (a23) του του γε γονε[ν] Π(praeter E )Ψ : τουÄ το γε γονεναι E1 1

καιÄ α K β : οë τεÁ µεÁν κινειÄσθαι οë τεÁ δ’ γ δ, fort. P (889.15) p

5 (a26) κινειÄσθαι

7 (a28) εë κατε ρου Ψ, scr. Ross

Pel : εë κατε ραν Π(praeter E) : εë κα τερον E

2–3 ÑM aÑiWÒ  ... ›B \M scripsi (= αÍ ττα [v. App. 1] τωÄ ν οÍ ντων [οÍ ντων om. K] = entium alia C = entium quaedam V)  :

ÑM ÏÑiWÒ  ... ¡yB \M Φ Bad 5 Ñ=Ö scripsi (= εÍ χοντα Π = habent C = habet V) : Ñ†Ö Φ Bad 8 fE secl. Bad ÑnY®Ö scripsi (= προιÈο ντες δ’ Π = et intendimus C = et incoepimus V) : ÑnY®Ö vel ÑnY¯F Φ : ÑnY¯F Bad

3–4 gloss without mark: ×qF Y4qLA YÒÚE fE (an ;NÒË leg. pro ;B ?) ;B ÑelÏK ÒdÕJÕmLA (iL cz Φ i.m.; cf. Simpl. 1257.19–23

259a20

259a25

259a30

63

VIII 6, 259a20-b6

3

6 Av. t. 51 Cp 196v

YÏ Ò  ÚXÑ  N YÒ  ŠÒË ³Mi Ò ÑnB ,FÒ  〈 Ö 〉 ~qÒd á¡V \M ÖÒ  …ÏhtM ñlÑqlT \MÖ ¡V ÕE ÑM õim`A Òª ÙbMÖ ñTÒD …ÏU ÕE …ÏhtM ÑEÒ  ÙÂbÊM ÑKÏhtmLA \M ÑKÏdA fEÖ fcLA áچ ÑM ÏÑiWÒ  〈 EÑZ 〉 ÙP ÙqF …ÏhtM áÚeF YÒÕiaA ŸÑnCÒ  (nJÖ (fNi Ò ÒÖD (nJ †LD èÑoM щÒÖD NÑK YÕkT YÒ  ¡V \M KaA ÔÙW YÒ  \km ÙQ & ¦ Y ~gIÒ  ˆ³EÕT³ YvÒ ÑeNÏÒ  †LDÖ [ê 219] áÚE „ ¿yI ÙQ †LD ÙP ÑNÏÒ  øbÂQ \M jÍ CÒ  ‹Y ømyN YÒ  fvbnI ;NÏÒËÖ .\Ï zI³ .F …ÏhtT vÏ KÏhtM ¡V  ~niH YÕkT 1 (a32) ηà 1 Π(praeter E1) Sc(1257.26) Pa(391.7) : om. E1 : non vert. Ar.

Π(praeter E ) : om. E : non vert. Ar. 1

fort. Ψ β

1

(a33) κινουµε νων Π(praeter K) : κινου µενον K,

(a33) µεÁν Π(praeter E ) : om. E : non vert. Ar. 1

2 (a33) εÆ στιÁ ν

2

1

4 (b2) τοÁ 2 Π(praeter β)Ψ : om.

(b3) δεÁ Pc(889.34) Ross : δηÁ Π(praeterE1)ΨΩ Pa(391.13) : om. E1

β Jw Pc(889.34) : παρειχει E1 : παρειÄχε K δ FHIQ (cf. Wag 685 ad loc.)

5 (b3) παρε χει E2Ψ (b4) εÆ γγι [γ]νεσθαι

β δ FHIw : εÆ γγε νεσθαι JQ : γι [γ]νεσθαι E, fort. Ψ, K S (1258.1–3) Pa(391.14) p

7 (b6) δεÁ δειÄ

fort. Ψ, K e F J Q Λ P (890.8) : [ras.] δειÄ E : τε δειÄ E : δηÁ δειÄ LF HI : δειÄ δηÁ NR : δεÁ δηÁ 2 2

2

l

1

2

1

J1 (?)Q1 w : δειÄ b

,FÒ  〈Ö〉 scripsi (= et pervenimus V) : ,FÒ  Φ (= perduxit C [sine et]) (for the apodosis of εÆ πει = ;L [259a27 = p. 62.6] see below, p. 63.4ff., and notes WiCo, Ross and HaGa ad loc.) 1–2 ÑKÏhtmLA \M YÏ Ò  scripsi (= οÏτι τωÄ ν κινουµε νων Π = quod eorum que mouentur ... est C = quod mota habent V) : ÑKhtmLA ÑB vid. Φ Bad 2 ÑEÒ  ÙÂbÊM scripsi : ÑEÖÙbM Φ : ÑEpÙbM Bad …ÏhtM Φ (= mobile C = motum V) : …htI Bad 3 EÑZ addidi (= ϕανερωÄ ς Π = visibiliter CV) : om. Φ Bad ÑM ÏÑiWÒ  scripsi (= οÍ ντα Π = quedam C = aliqua V) : ÑBÑbbbBÒ Φ : ~NÑsNÒË Bad ÑKÏdA Φ2 (= ΠCV) : ÑKÏhtmLA Φ1 4 áÚeF scripsi (= ista enim C = ista igitur V) : áÚEÖ Φ Bad (cf. App. 1 ad 259b3, App. 2 ad p. 63.1) 6 øbÂQ \M] øbQ \B Φ Bad : F non vert. CV,   om. Π : seclusi 7 \Ï zI³ s.p. Φ : \zN Bad 1

259b

259b5

64

3

6

9

(iLÑPÕpSrt fÏ yibpLA û4LA êÑtK \M nMÑoLA LÑqmLA

(iL KaA áÚE ñ³kÂIW YÏ Ò  Ö Í ÙHÒÖ Í KH …ÏU ;NÏÒË ÒÚLÑB …ÏdA YÏ Ò  YÒÕiaA „ øB ñlÑqlT \M úbsLA (iL ñNÏÒ  †LDÖ .èÖÏi Ò Mi Ò ÕE ñNÏÒ  ‹Y èjhmGiÒÖ ÕÏhA †LD èÑoM ,ñlÑqlT \M YÕkT siL iÏyibP XÂÒª ÑKH YÒ  ¡V \M YÒÕiaA ŸÑnCÒ  \M ÙHÒÖ øÏK ÑeKÏhtI ÑKH áÚEÖ ,(fÏntLÒÖ ÑKaA áÚE úbSÖ .~nKÑS øBÍ ñlÑqlT \M LA KaA ~KÏhtM YÕkI ñNÏÒ  †LDÖ YÒÕiaA ¡yB „ aÒÚvLA †LD èÑoM .ádI ÑJÏ ¡È oKÖ ñB ×idA èÖÏi Ò zÙbmLÒÖ ñTÒD …ÏUÖ ñbtnI ñNÏÑËF ‚iÏo ÒDÑËF wÑN ÕeF …ŽI ‡Òd ÑM …ÏdA YÏ Ò  †LDÖ .jÍ ctÏM ñlÑqlT \M ÙBÒ  …ÏhtI i rÑC †LÚLÖ ñnM rÑX áÚE „Ö .ñKÏU ÑJÏ ÙHÒÖ øÏK úsT iÏvtIÖ …ÏhtI ÑMÙnY á¡Â V YÕkI ñL Í 1 (b7) αυÆ ταÁ E K β δ F2HIJw Pl(890.8) : αυë ταÁ F1 Ross : εë αυταÁ fort. Ψ, Q fort. Sp(1258.13) 3 (b9) κινουÄ νται] γι γνονται fort. Ψ ηë α K β δ Sp(1258.19) : om. γ εë αυτουÁ ς vel εë αυταÁ Ψ : αυÆ τουÁ ς β

5 (b11) του του HIJQw : του των α K β δ F Λ

6 (b12)

7 (b13) εë αυτουÁ ς E K δ FHJ1Qw Pa(392.7) : εë αυταÁ J2 : (b13) πρω της Π(praeter E2)Ψ : exp. E2

8 (b14) ουÆ κ αÆ ειÁ

E1Ψ K e2 β γ Pa(392.13) : ουÆ E2 be1, fort. T(224.15), fort. Sp(1258.29–30) κινουÄ ν αυÆ το , κινου µενον i’punxit Ar.

1

2

2

9 (b15) τοÁ

c

(b16) εë αυτα E K e FJ S (1258.32) : exp. E2 :

εë αυτο be1 HIJ1Qw : εë αυτ\ Pl(890.27 αυτα ed. Vitelli, cf. Pp[890.32ff.]) : αυÆ το fort. Ψ

†LD add. YÒ Φ : om. ΠCV : secl. Bad 6 ÑJÏ ¡È oKÖ scr. Bad (= καιÁ πολλαÁ τωÄ ν Π = et multa eorum que C = et multa V) : ;MÒ bKÖ Φ 7 ñbtnI] s.p. Φ : ñbntI Bad …ÏUÖ] dittogr. Φ 9 á¡Â V scripsi (= αÍ λλο Π = aliud CV) : ñnY Φ Bad 3 post

2 èÖÏi   Ò Mi Ò with mark s.l. and gloss without mark: úbsLA fE siL ÑeNÏÒË èÕqIÖ %ÕfnLA †LÚB yI

ÑKaA øÏkB YÒÕiaA †IW „ Φ i.m. (cf. Alex. frag. 677, Philop. 889.32–890.7, Simpl. 1257.39–1258.17) ñlÑqlT with corresp. marks s.l. and i.m.: ÑKaA øÏK …ÏhtI YÒ  „ ñsfN „ cÒ  Φ i.m., fort. ex comm. Philop. 890.9–10 7 èÖÏi   Ò zÙbmLÒÖ] + aÒÕ=A yI s.l. Φ, fort. ex comm. Alex., v. Alex. frag. 680 (above, p. LXXIV), v. also Philop. 890.33 9 áÚE with corresp. marks s.l. and i.m.: YÒÕiaA ŸÑnCzÖ (fNtÒ ÒÖD ŠÒË ‡jkLÑB õJr Φ i.m. (cf. Alex. frag. 683,

Philop. 891.16–18, Simpl. 1259.8f.)

259b10

259b15

65

VIII 6, 259b6–23

Cp 197r

3 Av. t. 52 Cv 76r

6

YÏ Ò  ÇÒË ñlÑqlT \M ñTÒD ñkIW úbS ÕE cÚLA èÖÏi Ò …ÏdA …ÏhtI ÑelÏK ~gIÒ  øqtnI †LÚLÖ ñNÂÑkM èÙÏbI YÙbLA YÏ Ò  †LDÖ .¿yLA „IpB YÕkT ñtKH .ñTÒD …ÏL lmÊÂaA „ ÕE ÑMÖ YÙbLA „ ÕE ÑM ÑeNÏÒ  ÇÒË …ÏhtT siL LA \M YÑK YÒË ~miW YÏ Ñ  B „OÕI ÙQ †LD \B †LÚlF .ÙM¬ KaA YÕkT YÒ  \km (iLÖ ¿yLA „IpB ~gIÒ  ÕE …ÏW ÑnEÑE 〉 YÕkI YÒ  úO ÙqF ÙM¬ KaA YÕkT YÒ  Í rÖ² ~bJÒÖ YÑK YÒË 1 (b17) τοÁ 2 E2 K β δ FHIQw : om. E1, fort. Ψ, J Sc(1259.9), fort. Pa(392.26)

Π(praeter E) : καιÁ τοÁ E2 : καιÁ τοÁ εÆ ν E1Ψ, fort. Pa(393.1–2) i’punxit Ar.

3 (b19) καιÁ 2

4 (b21) κινου ντων δε , καιÁ ...

5 (b21) αυë ταÁ scr. Ross : αυτοÁ E Ψ (αυÆ τοÁ interpr. Ar.) : αυÆ τωÄ ν E2 : αυÆ τωÄ ν 1

κινουµε νων K2 β γ δ : εë αυταÁ K1 Sc(1259.35)

(b22) κινειÄν Π(praeter E1) : om. E1, fort.

Ψ(nisi ειËναι leg.)

…ÏdA Φ2 (= ΠCV) : …ÏhtmLA Φ1 3 lmÊÂaA scripsi : ñlbaA vel ñlmaA Φ (for τηÄì µοχλει αì ; an ilxÊm³LA vel ‹xÊm³LA [〈 syr. muhla¯] leg.?, cf. S. Gerö, New Test. Stud. 27, 411–14; R. M. Frank, ˘

1

The Wisdom of Jesus ben Sirach, 28.25; for muhlun = µοχλο ς and ibu¯muhliyu¯n = υë ποµο χ˘ ˘

λιον cf. al-Khwa¯rizmı¯, Mafa¯tı¯h, ed. van Vloten, p. 247.6–248.1; Freytag IV, 158b13; Dozy

˙

lm`A Bad (= toto C = universo V), v. also App. 3 ad loc. 4–5 ~gIÒ  ... ÇÒË] sed mouet ipsum etiam (= ~gIÒ  ÕE …ÏU ñNÏÒ  ÇÒË) C : sed mouent etiam (= ~gIÒ  …ÏW ÑeNÏÒ  ÇÒË) V 6-p.66.2 YÕkI ... ÑnEÑE addidi (= ειËναι τι ... εÍ σεσθαι [v. App. 1 ad 259b24–25] = ut sit primum II, 572b12) :

mouens immobile non per uiam accidentis et si est sicut diximus in eis que sunt praemissa oportet ut C = ut primus motor sit qui non mouetur neque per accidens sed [fort. si leg.] necesse est sicut diximus superius ut V) : om. hom. Φ Bad (for ειÆ µε λλει =

~yM‚M YÑK YÒË v.

Phys. 198b7 = 141.2f.; An. post. 80b36 = 383.4, also Hipp. Aphor. V 23, transl. Hunayn) ˙

lmaA (liaA?) with corresp. marks s.l. and i.m.: ñTÒÚB …ÏhtmLA †IW cz Φ i.m. (cf. Simpl. ñTÒD with corresp. marks s.l. and i.m.: *** aA three or four words illeg. in the gutter margin of Φ (probably another gloss on lmaA) 3

1259.18–20, and App. 2 ad loc.)

259b20

(iLÑPÕpSrt fÏ yibpLA û4LA êÑtK \M nMÑoLA LÑqmLA

66

4 Ca 27rb

Cp 197v 8 Av. t. 53

‡ÙÏqT .F ÑnlQ ;K ~yM‚M YÑK YÒË ¿yLA „IpB iÖ …ÏhtI i èÖÏÒ  …ÏL ÕE dÕJÕmLA bI YÒ  Ö ÙibT iÖ £fT i ÑM KH ÒdÕJÕmLA „ 〈 YÕkI YÒ  ~gIÒ  øÏkLÑF ~iQÑB YÑK ÒDÒË Ò  ÙbmLA YÏ Ò  †LDÖ .юiyB ÙHÒÖ èÑT ñsfN † ¿yLA „IpB aLA …ÏhtI YÒ  (iLÖ .zÙbmLÑB ÆctÏM YÑK DÒË ƒÑB cÏ rÖ² ÙQ ¡vLA \Y …ÏhtLA YÏ Ò  †LDÖ ñniyB ÙHÒÖ !Í yM á¡V \M ÖzÍ ñlÑqlT \M øqtnI юM YÑK ÑM YÒ  a4LA „ ÑJÏ ÏÑiWÒ  ndÑbM ¡yB „ ~gIÒ  ÙJÕI ÒÙSÑfLA „ ÙJÕI ;NÏÑËF XÂvÒ Mi Ò ÑMÏÒ  Ö ÙHÒÖ \M ¤Â KÒ  lqnLA \M ~FÑnCÒ  .×qF ÕEÖ ÑM ~miW …ÏU ñNÏÒ  YÒ  ÒÚE ‹Y ÙBÒ  ÕE ÑM a ÑnEÑE YÑK YÒËÖ .ˆÏ RÒ  〈 ñnY 〉 …ÏhtM èÖÏÒ  ~gIÒ  YÕkI YÒ  Í rÖ² úJÒÕF ˆÏ RÒ  …ÏhtM ¡V 1 (b24) καιÁ fort. ΨΩ(»not even per accidens« Ar.; cf. WKAS I, 31b), K1 β H : καιÁ µηÁ K2 δ FIJw Λ : καιÁ c. ras. E Q : καιÁ µηδεÁ Pa(393.7) p. 65.6) : om. E1

(b24) ειÆ µε λλει Π(praeter E1)ΨΩ(v. App. 2

(b25) ειÍποµεν Π(praeter E1) : ειÍποµεν εÍ µπροσθεν E1ΨΩ

Ψ, scr. Ross (sec. διατελε σειν T[224.21]) : µε νειν Π FHI1JQw : αυÆ τωÄì fort. Ψ, K b β I2 Λ

4 (b28) τοÁ 2 E2Ψ K δ L2NR FHI2J1Qw Sc(1261.20,

1262.11) Pa(393.15) : τωÄì I1 : dittogr. L1 : om. E1 J2 Sl(1261.12) om. be1 : an τινωÄ ν τωÄ ν Ψ?(cf. εÆ πι τισι τωÄ ν Pa[393.18–19]) E1 : exp. E2

6 (b30) τωÄ ν E K e2 β γ :

(b30) τωÄì Π(praeter E)Ψ : illeg.

9 (b32) κινουÄ ν µεÁν E2Ψ β δ FJQw2 Λ : κινουÄ ν µε ντοι I(sed om. τι) : τοÁ κινουÄ ν

K Sl(1262.14) : κινου µεν E1

10 (b33) τοÁ ] fort. om. Ψ

ˆÏ RÒ  ... YÒËÖ om. V : habet C ˆÏ RÒ  ] ~iLRÒ  scr. Bad

9–10

2 (b26) µενειÄν

3 (b26) αυë τωÄì (vel εë αυτωÄì ) E e

2

10

ñnY addidi (= υë ποÁ του του Π[sed om. F] = ab eo C)

2

7

ÙHÒÖ with corresp. marks (!)in textu and i.m.: iÏhtmLA …jFi Ò ŠÒË †LÚB ¡wI Φ i.m. = Alex.

frag. 689 (v. above, p. LXXV)

259b25

259b30

67

VIII 6, 259b24 – 260a9

3

6

YÈ ÕÏkT YÕkI YÒ  áÕJÕLA \M ñJÕB \km i ñNÏÑ  bF ÅÖÏÒ  ÑMÏÒ  †LD \iÏbT ÙQÖ 〈 …L 〉 a ÑnEÑE \kI G ÑM X  iªÒ aÑiWi Ò lÑsL ÈiÏvTÖ dÈÑsF iÖ Ï Í KH ÙHÒÖ eN ÙBÒ  …ÏU ;NÏÑËF …ÏhtI i ÑM YÏ Ò  †LDÖ .…ÏhtI [Ò220] …ÏhtmLA ŠÒË %ÑiqLÑB ÆCÒ  ÕE iÏvtI (iL ñNÏÒ  øbÂQ \M 〈 Í ÙHÒÖ〉  ÑMÑ  F .ñnY (iQ ÒDÒË fltM èÒÕH{B ¡cI ñNÏj F …ÏhtmLA ¡V \Y …Â Ï htmLA Ï „ YÕkI ñNÏÒ  øJÒ  \M øB юiyB ÙHÒÖ Ka ~bbS YÕkI i ñNÏÑËF rÕMtÑB ø¬ K ñ³k³IW áÙifI³ ÑM YÕkI †LD øbÂQ \M dÏÑgtmLA rÕcLÒÖ õGÒÕmLA   3 (a3) γαÁ ρ Π(praeter E1)Ψ : αρ E1

1 (a1) καιÁ ] non vert. Ar.

αυÆ τηÁ ν αÆ ειÁ e2 γ : αÆ ειÁ τηÁ ν αυÆ τηÁ ν R

(a4) αÆ ειÁ α K be1 LN : τηÁ ν

(a4) τοÁ ν αυÆ τοÁ ν κινη σει τρο πον K β : κινη σει τοÁ ν αυÆ τοÁ ν

τρο πον γ δ : Ψ = Kβ ∨ γδ : ras. E : τηÁ ν αυÆ τηÁ ν κινη σει E2

5 (a5–6) τοÁ δεÁ κινου µενον

1

Π(praeter E )Ψ : om. hom. E 1

(a6) υë ποÁ τουÄ κινουµε νου µεÁν E2(i.m.) e2 FH2JQw Λ

1

γρ.P(894.10), fort. Pa(394.10) : om. E1Ψ K be1 β H1 Pl(894.5) Sc(1262.33) (cf. Ross 708)

(a6–7) δεÁ κινουµε νου ηÍ δη E2 e2 γ Pl(894.5) : δεÁ κινουµε νου ηà E1 : ηà κινουµε νου ηÍ δη K be1 Sc(1263.16–17) : δηÁ κινουµε νου ηÍ δη β : om. Ψ, fort. Pa(394.11) E )Ψ : εÍ χειν ηÍ δη E 1

1

\iÏbT]

s.p. Φ :

1

›bI

Bad

ÅÖÏÒ  ÑMÏÒ 

ñNÏÑ  bF Φ : ñNÑËF Bad

om. Φ Bad [Í ÙHÒÖ Í KH om. V]

c

~gIÒ  ... vÏ ), 260a20 (cf. Alex. frag. 693,

2 …ÏL addidi (= κινη σει [κινη σι E1] Π = mouens C =

quod mouet V) : om. ex haplogr. Φ Bad et Gutas Theophr. p. 416 s.v. γα ρ)

1

(= in primis C = primo autem V, for µεÁν 260a1)

foreshadowing the transl. of ουÆ µηÁ ν αÆ λλαÁ καιÁ (= Philop. 890.30ff.)

(a7) εÍ χειν Π(praeter

6 (a8) εÆ ν Π(praeter E b)Ψ : om. E b P (894.14)

1

3 YÏ Ò  †LDÖ scripsi :

Yi †LdÖ vid. Φ (cf. Glossary

4 Í ÙHÒÖ addidi (= µι αν κι νησιν Π = motum unum C) :

7 rÕcLÒÖ scripsi (= ηà ειÍδεσιν Π = aut formis CV) :

Bad

7

õGÒÕmLA] + è6LÒÖ êÕn`A Φ s.l. = Alex. frag. 698 (v. above, p. LXXV)

rÕcLÒÖ Φ

260a

260a5

(iLÑPÕpSrt fÏ yibpLA û4LA êÑtK \M nMÑoLA LÑqmLA

68

Cp 198r

3

6

9 Av. t. 54

.〈 ~KÏhtM ~niHÖ ~nKÑS ~niH YÕkiF Òd­ ÑgtM aÑiWi Ò lÑS \M 〉 …ÏhtM Í rÑC ˆÂL Mi Ò zÙbM „ ñiF \i ÏKÑW ÑnÏK ÑM áÑnlQ ÑJÏ ~gIÒ  eZ ÙqF    Y4Q fE iÖ nKÑS ÑMÏÒËÖ KhtM ÑMÏÒË ÑelÏK (iL aÑiWi Ò …ÏhtM ;EÙHÒ Ï YÏ ÑËF .~niH \ksTÖ ~niH …ÏhtT aÑiWÒ  ÑnEÑE øB ÙBÒ  \KÑS XvÒÖ ÙBÒ  …ÏhtM ¡V \Y …ÏhtT aÑiWi Ò ¡yB YÏ Ò  ÕEÖ YvÒ \È i® B †LD „ úbsLA iÏvtM 〈 …ÏhtM \Y …ÏhtT ;NÏÒË Ñ™yBÖ ÙBÒ  …ÏhtT ÑeNÏÑËF †LÚLÖ ˆÏ RÒ  〉 ñNÏj F …ÏhtM ¡V ÕE cÚLA ÑMÏÑ  F .iÏvtM ~gIÒ  fE YÕkT YÒ  úO †LÚLÖ Í KH …ÏU ;NÏÑËF ñniyB ÙHÒÖ aBÖ ÙHÒÖ MÒ  ‹Y ƒÑB ×isB ÑnlQ ;K Í ] .Í pisB Í ÙHÒÖ [9Ò220 ÚX{M ÚX{N YÑ  B ¤Â KÒ  rÕeZ †LD „ Mi Ò ezI ÙQ vÏ [3 Ò 221] ‡Ò  È lctÏM ÑM KH YÕkT YÒ  \km øE znN YÒ  fvbnI ÙQ ñNÏÒ  †LDÖ ~gIÒ  ÂX| 2 (a11) καιÁ Π(praeter be1)Ψ : om. be1 ηÆ ρεµειÄ Π(praeter γ)Ψ : inv. γ

4 (a13) αÆ ειÁ 2 Ψ K γ : om. E β δ

γ)ΨΩ Sp(1264.4) : µεταβα λλει γ Λ

1 ~KÏhtM

3 (a12) ηà 1 α β δ Iw : om. FHJQ

(a12) κινειÄται ...

6 (a15) κινειÄται2 Π(praeter

7 (a17) αÏ τε Π(praeter E1)Ψ : om. E1

... \M addidi (= εÆ ναντι ως ... κινου µενον 260a9–10 Π = unumquodque reliquorum

aliorum contrarium quare sit quandoque quiescens et quandoque mouetur C = res contrarias et quandoque erit quiescens et quandoque motum [om. 4

…ÏhtT] dittogr. Φ

5

…ÏhtT s.p. Φ : …htI Bad

\M …ÏhtM ø¬ K] V) : om. hom. Φ Bad Í Ï htM ... ˆÏ RÒ  addidi (= αÆ ιÈδι ου ... 6 …

κινουµε νου 260a15–16 = perpetuo et propterea illa mouentur semper et quedam non mou-

entur nisi ab eo quod mouetur C = aeterno et ideo mouentur semper et quedam non mouentur nisi a moto V) : om. hom. Φ Bad unum eundem C : om. V :

11 gloss without mark:

œSÂÖ Bad

8 aBÖ scripsi (= καιÁ εÆ ν τωÆì αυÆ τωÄì Π) :

10

ÙQ vÏ Φ : ÙQÖ scr. Bad

!SÖ Φ : et

¿yLA „IpB iÖ ÒÚLÑB i …ÏhtM i …ÏLÖ ñsfN ñtKH zÙbM …ÏL ÑnEÑE cz Φ

i.m. ex comm. Ibn al-Tayyib 884.6–7 ˙

260a10

260a15

Chapter 7 260a20

69

VIII 6, 260a10 – 7, 260a33

Cv 76v Cp 198v 3

Av. t. 55 6

9

. ÑKaA èÖÏÒ  fE KH cÏ Ò  Ö áÚE fE KH cÏ Ñ  F ~nkJ †LD YÑK ÒDÒËÖ i KH NÑKÖ qÒd KH YÕkT YÒ  Í rÖ² ~bJÒÖ YÑK YÒË ñNÏÒ  \iÏbLA \M ñNÏÑËF áÚE …ÏU ;NÏÒË èÖÏi Ò …ÏdA YÏ ÑËF lctÏmLA ŠÖiªÒ fE эLAË rÑwM ÑM .ŠÖÒª lctÏM fE юiyB ÙHÒÖ YÕkT YÒ  Í rÖ² úO ÙQ LA KaA ¡O{tLA „ KaÒÖ ˆzÂyLA „ KaA ~OjO NÑK ;L ÑKaA YÏ ÒË èÕQÑ  F áÚE YÕkT YÒ  Í rÖ²  úO ÙqF Í lqN³ эmÏsN LA fEÖ YÑkmLA „ KaÒÖ LÑhtS YÕkT YÒ  ¡V \M ÕÏn YÕkI YÒ  \km i ñNÏÒ  †LDÖ . ÑKaA èÂÖÏÒ  ;NÏÒË ñNÏÒË èÕqT YÒ  †LÖ ñibwLÑB 〈 fmnI ñNÏÒË èÕqT YÒ  †L ;NÏÒË 〉 fMÑnLA YÏ i  øÏK „ øctÏI ;NÏÒËÖ .áÙÏgL ÙÏgLA ñNÏÒË èÑqI³ aÒÚvLA YÏ Ò  †LDÖ ñibwLA ¡vB fmnI ÙÏgLA ŠÒË ÙÏgLA \M iÏvtLA YÕkI YÒ  úO ÙqF .ñbwB ñ³ibwLA YÕÏktM 1 (a22) ειÆ Π(praeter δ)Ψ : om. δ

4 (a25) ηÊν E2Ψ K b2e β FHIJ2Qw Λ : om. E1 J1 : οÃν b1

(a26) καιÁ 2 e2 β γ Λ Sp(1265.6) : om. α K be1 (cf. ουË ν Pa[395.9])

3 эLAË rÑwM scr. Bad :

5 (a26) δ’ FHIJw : δεÁ E K δ Q : δηÁ fort. Ψ, β

7 (a30) προυÈ παρχου σης Π(praeter E1) : υë παρχου σης E1Ψ

эLÒÒrÑsM Φ

7 post ÕÏn add. KaA Φ : om. ΠCV : seclusi

8 fmnI ... ;NÏÒË

addidi (= τοÁ ... αυÆ ξανο µενον ... αυÆ ξα νεται [αυÍ ξεται δ] Π = quod augmentatur tibi inest ut dicas quia non augmentatur nisi per simile C = augmentabile ... potes dicere ipsum augeri per simile V) : om. hom. Φ

10

ÙqF scripsi (= ουË ν Π = igitur V) : ÙQÖ Φ (= et C) Bad

lctÏmLA] + IÙM¬ gz Φ s.l. (cf. Alex. frag. 704, Simpl. 1265.2–7) 5 ¡O{tLA with corresp. marks s.l. and i.m.: LÑhtSiÒ cz Φ i.m., fort. ex comm. Ibn al-Tayyib 884.19f. 9–10 gloss ˙ without mark: ÏMÖ áX}B rÑC ÒDÑËF áÙ¬ G øihI³ øidA YÏ t øihtsI ÒÚ=Ö ÒÙ­ G iÍ ÖÏz YÕkI aÒÚvLA fLA ÕBz ˆhlLÒÖ ‡ÙLA ÕV ~bW rÑC iÑhtSiÒ ñilY Φ i.m.; cf. Philop. 896.17–19 3

260a25

260a30

(iLÑPÕpSrt fÏ yibpLA û4LA êÑtK \M nMÑoLA LÑqmLA

70

Cp 199r 3

6

9 Ca 27va

ÆihM³ ÑM Ó YÕkI YÒ  úO ÙqF LÑhtS NÑK "M \kL .Í LÑhtS ÙnY …ÏdA YÏ Ò  \iÏbLA \MÖ .rÏÑH øyfLÑB ÑM ŠÒË rÏÑH ÕÏqLÑB ÑJÏ ~JxM³ øihtsmLA ŠÒË êQÒ  YÑK ;BÏr \kL ÙHÒÖ èÑoM ‹Y áMÒ  cO (iL †LD DÒË YÑK YÑËF .lqnLA YÖd YÒÚE \km (iL〈 Ö 〉 ñnM ÙyBÒ  YÑK ;BÏrÖ qÒd ~gIÒ  lqN YÕkT YÒ  Í rÖ² úO ÙqF qÒd KH YÕkT YÒ  Í rÖ² YÕkT YÒ  XÏÑ  tM lqNÖ MÙÏqtM lqN lqnLA \M YÑK YÒËÖ ÑKaA èÖÏÒ  .qÒÙLA fE MÙÏqtmLA øqoLA YÏ Ò  †LDÖ .øxlxtLÒÖ žOÑktLA rÑOvÒ õi? Ò  ÙbM YÏ ÑËF ~gIÒ  Ö FÑokLA \M ŸÑnCÒ  ÑeNÏÒ  \Ï zI³ ÙQ dÖ LÒÖ rҁaÒÖ BjcLÒÖ ›lLÒÖ fÏbÒÖ ;= YÒÚlLA ;EÖ ƒÒ£FÖ û0J ;E øxlxtLÒÖ žOÑktLÒÖ Â .FÑxsLÒÖ 1 (b1) δειÄ τι Π(praeter γ)Ψ : δεη σει γ Λ Kδ

2 (b2) τοÁ εÆ νεργει αì Ψ β γ Sp(1265.31) : εÆ νεργει αì E

4 (b5) ειÆ Π(praeter E1)Ψ : om. E1

6 (b6–7) ϕοραÄ ς Π(praeter E1)Ψ : ϕορα ν E1

10 (b11) αÊ ς E1Ψ K HIJQw Λ Sc(1266.31), fort. Pa(396.20) : αÊ ς καιÁ E2 β δ F

ÑiW Φ 2 1rÏÑH Φ : Òr­ ÑH Bad 2–3 †LD ÙnY] hapax leg., pro ουË ν (b2)? : tunc CV (sed †LD ÙnY = ουÏ τω Metaph. 995a18 = Bouyges 50.3, transl. Isha¯q) 4 Ö suppl. ˙ Bad (= δ’ Π = et CV) YÒÚE] with corresp. marks s.l. and i.m.: ÙH* vel J* Φ i.m. (partly illeg. in the gutter margin, an ÙH〈ÒÖ〉 »singular« [i.e., leg. ÒÚE]?) : hoc CV 6 XÏÑ  tM scripsi (= υë στε ρα Π = posterior CV) : á BÑbM Φ : ®ÑbM Bad ante YÒ  add. sed oportet Cv in textu, a p C s.l. [non C ] : necesse est add. V 8 øxlxtLÒÖ scr. Bad (= καιÁ µα νωσις Π = et raritas CV) : øxlxtL iÏvtLÒÖ, sed del. ut vid. iÏvtLA Φ 9 ÙQ scr. Bad : ÙQÖ Φ 1 Ó' scr. Bad :

8

rÑOvÒ with corresp. marks s.l. and i.m.: ҁiÏvtLA (pace Bad.) Φ i.m. = Alex. frag. 708 (v.

above, p. LXXV)

260b

260b5

260b10

71

VIII 7, 260a33-b21

3 Av. t. 56 Cp 199v 6

9

ƒ£fI ÖÒ  õmtO 〈 cÚLA aLA ÑMÏÒ  Ö 〉 .ÑEdÑsFÖ EÒÕ`A YÕÏkT YiÑqI³ ñm³zÂY èÙ® bÂI³ øÏhmgmLÒÖ fMÑnLA YÏ ÑËF ~gIÒ  Ö .ñNÑkM èÙ® bÂI³ YÒ  úJÒÕF  .Y ÑkmLA

„IpLA ÒÚE znLA † ÒÕklS ÒDÒË ÑKaA èÖÏÒ  lqnLA YÏ Ò  ~gIÒ  ezI ÙQÖ ~gIÒ  †LÚK XÂÒª aÑiWÒ  „ htÏW ÎÑVÒ  ‹Y èÑqI ñNÏÒ  ;K èÖÏi Ò YÏ Ò  †LDÖ dÕJÕM \kI G ÒDÒË ÑM ‡ÙÏqtmLA [ê 221] èÑqI ÙQÖ .KaA „ èÑqI èÑqIÖ XÂiªÒ \M ÕlX dÕJÕM ÕE YÕkI ÙQÖ Í dÕJÕM XÂiªÒ lÑS \kT G YÕkT YÒ  Í rÖ² ~bJÒÖ YÑK ;lF . ÒÚLA „ ÖÒ  YÑM‚LA „ ~MÙÏqtM YÑK .F iLÑttM ÑeNÏi  ÑMÏÒËÖ lctÏM ÑeNÏi  ÑMÏÒË 〈 ÙM¬ YÕkT ÙQ ;NÏÒËÖ 〉 ÙM¬ KH 2 (b13) αÆ να γκη Π(praeter E2δ)Ψ : αÆ να γκη καιÁ E2 δ

6 (b17) αà ν λε γοιτο E2 δ : λε γοιτο E1 :

λε γεται fort. Ψ, K β : αà ν λε γοιτο πρω Ä τον FIJQw : λε γοιτο πρω Ä τον H

(b17) λε γεται δεÁ

Π(praeter δ) : λε γεται γαÁ ρ δ : λε γοιτο δεÁ αà ν fort. Ψ, fort. Alex.(apud S[1268.3]) om. S (1267.36) Pa(397.3) : non vert. Ar. c

(b18) τε]

8 (b19) τοÁ Π(praeter E ) : om. E , fort. Ψ 2

1

1

9 (b20) ειÍη δ’ αà ν συνεχωÄ ς] om. hom. E1, fort. Pa(397.11) : suppl. i.m. E2 : om. hom. (?)Ψ(nisi suppl. i.m.?), vel om. hom. Φ? (habent transl. Gerardi et Scoti, an ex alia transl.?; cf. App. 2 ad loc.)

(b20–21) ηà συνεχηÁ ς ουË σα ηà εÆ ϕεξηÄ ς α K, fort. Sp(1269.30) : ηà ηë συνεχηÁ ς ουË σα ηÃ

ηë εÆ ϕεξηÄ ς β : ηà ηë συνεχηÁ ς ηà ηë εÆ ϕεξηÄ ς γ δ Λ

ÑEdÑsFÖ ... YiÑqI³ scripsi (= [καθ’ αÊ ς] γε νεσις καιÁ ϕθοραÁ λε γεται τωÄ ν ουÆ σιωÄ ν, nisi ;eNÏÑ  B èÑqI³  ... YÕÏkT leg.?) : YÕkB ÑEdÑsFÖ EÒÕ`A YÕkB YÒ èÑqI Φ : [secundum quas] dicitur generatio

1

substantiarum et earum corruptio fiunt C : [per quas] dicitur substantias generari et corrum-

cÚLA ... ÑMÏÒ  Ö addidi (= συγκρινο µενα δεÁ Π = ƒ£fI 2 1 Φ : ƒfI Bad 2 øhmgmLÒÖ corr. Φ with mark i.l. + øhmgmLA i.m. : øhgmLÒÖ Φ Bad 6 post ÑM add. YÑK Φ Bad : non vert. CV : seclusi 7 2XÂtÒ] XiÒ Φ : XvÒ Bad 8 ~MÙÏqtM 2 YÑK ... om. hom. C : habet V 9 ÙM¬ ... ;NÏÒËÖ addidi (= ειÍη δ’ αà ν συνεχωÄ ς Π[sed om. E1] =

pi V :

YÕkI ÑEdÑsFÖ EÒÕ`A YÕÏkT YÒË èÑqI Bad

quod autem aggregatur C = et quod congregatum V) : om. (deinde õmtOÖ scr.) Φ Bad

et non est semper nisi C = et non est aeternus nisi V) : om. hom.(?) Φ Bad (cf. App. 1 ad loc.)

260b15

260b20

(iLÑPÕpSrt fÏ yibpLA û4LA êÑtK \M nMÑoLA LÑqmLA

72

Cv 77r 4

8 Av. t. 57 Cp 200r

Í iLÑttM YÕkT YÒ  i Í lctÏM YÕkT YÒ  øgFi ÒÖ Í lctÏM YÕkT YÒ  g¦HÊi Ò YÑKÖ \km ÙQÖ ~nkJ YÑK ÑM øgFi Ò Mi Ò ñNÏÒ  yibpLÑB YÕkI .F ÙqtyN ÙBÒ  ÑnÏKÖ ñL¨nlF øJÑyLA „ ÑMÏÒ  Ö XÑ  B ñniÏbnS a ÒÚEÖ) Í lctÏM YÕkT YÒ  lqnLA g¦ÕS Í lctÏM gXÒª È KH YÕkT YÒ  \km i YÑKÖ (Å҂NÒË úO (iL øqtnmLA YÏ Ò  †LDÖ .ŠÖiªÒ fE lqnLA YÕkT YÒ  Í rÖ² úJÒÕF YÕÏktI YÒ  iÖ LÑhtSiÑB iÖ ÕÏhÑB i ÆCÒ  gXÒª KH …ÏhtI YÕkI YÒ  G ÑM юM ÙHÒÖ iÖ YÕkT YÒ  \km (ilF ÑKaA áÚE ÑMÏÑ  F .ÙsfI YÒ  iÖ .èÖÏi Ò …ÏdA …ÏU ÑEÑIÏÒË LA fEÖ lctÏmLA KaA \kT ‹Y rÙqT ×qF iÏLRi Ò YÏ Ò  †LDÖ YÑM‚LA „ ŠÖÒª fE ÑeNÏÑËF ~gIÒ  Ö YÕ° kT ñL ÑJÏ YÑK  xW c¬Ò  ÙHÒÕLA  xwLA YÏ Ò  ¡Â V KaA áÚE …ÏhtLA 4 (b25) αÆ λλ’ ηà ϕορα ν α K β δ Λ : αÆ λλαÁ ϕορα ν F2HIJQw : om. F1

6 (b27) δηÁ Π(praeter

J ) : om. J : ηÍ δη Λ : non vert. Ar.

l

1

1

9 (b30) µο νον εÆ νδε χεται α K β P (899.19) Pc(899.20) :

µο νην εÆ νδε χεται δ Tc(226.7) Pp(899.20) Sp(1270.12) : εÆ νδε χεται µο νον γ

4 YÑKÖ] quia C : quod V (= ñNi   vel ñNi Ö ?)

"LA ÑMÑF Φ : et est ille quem C C : et est ille n¯ quem C : secundum quem V : an LA leg.? 9 ŠÖÒª Φ : èÖÒ  Bad iÏLRi Ò rÙqT ... scripsi (= τοιÄς ... αÆ ιÈδι οις ... εÆ νδε χεται) : rÙqB ... iLRi ÑB Φ : rÙqI ... iLRi ÑB Bad : cum perpetuitate ... potest C : secundum ... aeternitatem ... potest V 10 …ÏhtLA scripsi (= κινειÄσθαι Π = moueri V [moue˜ vel sim. C]) : †IhtLA Φ Bad p

5

v

a

8 LA fEÖ scr. Bad :

ŠÖiªÒ with corresp. marks s.l. and i.m.: õbpLÑB ŠÖiªÒ ÙII Φ i.m., fort. ex comm. Alex. (frag.

713), cf. above, p. LXXV

260b25

260b30

VIII 7, 260b21 – 261a9

3

6

9

Cp 200v

73

\M YÕkT ;NÏÒË ÑeNÏÒ  †LDÖ LÑL i ÑelÏK ÑKaA ÂX| lqnLA ñL ¡cT ;NÏÒË .ømÂkÊtsmLA KH fE ;NÏÑËF lqnLA ÑMÏÑ  F Õ± nÖ È LÑhtSÂ ÅÖÏÒ  YÕÏkI YÒ  ÙyB ÕEÖ lqN  KH ÆbQ …ÏhtI á¡V XÂ| a YÕkI YÒ  Í rÖ² úO ÙQ \kL YÕÏktIÍ ÕE YÕkI YÒ  ¡V \M ÑNÕÏktmLA YÕÏkT „ úbsLA ~gIÒ  YÕkI cÚLA Y¯ÑzL YÑK ÙqF †LÚK Mi Ò \kI G ÕLÖ ÙLÏÕtmLA úbS ÙLÏÕmLA YÏ Ò  †LD èÑoM YÕÏktI YÒ  ÅÖÏÒ  úO ÙQ ñNÏÒ  øbÂQ \M ÑKaA èÖÏÒ  YÕÏktLA YÏ Ò  \Ï zI YÒ  c¬Ò  ÙHÒÕLA  xwLA „ ÒÚE  ‹Y cO YÑK YÒËÖ Mi Ò YÏ Ò  ¡V aLA ÑM a ÑnEÑE YÕkI YÒ  Í rÖ² úO ÙQ \kL ÑNÕÏktmLA \M YÑK  xW 〈 X  | 〉 a ÒÚ=Ö YÕÏktM i dÕJÕM ñsfN „ ÕEÖ …ÏhtI ÑNÕÏktmlL ‡ÙÏqtM YÏ Ò  †LDÖ èÖÏi Ò ÕE YÕÏktLA YÕkI YÒ  \kJ ¡V YÑK ;LÖ .ñnM ‡ÙQÒ  \M a (iL ñNÏÒ  \iÏbLA \B Í ÙSÑF YÕktS NÑK ÑelÏK ÑKÏhtmLA 2 (b32–33) αÆ λλοι ωσις καιÁ αυÍ ξησις α b HIJ1 (?)Q2 w Λ : αÆ λλοι ωσιν καιÁ αυÍ ξησιν e F2[F1 αÆ ϕεξιν]J2Q1 : αυÍ ξησις καιÁ αÆ λλοι ωσις K β εÆ στιν b e : om. b 2

1

3 (a1) αÆ λλ’ εÏ τερον Π(praeter δ)Ψ : αÆ λλ’ εÏ τερον

4 (a2) αιÍτιον Π(praeter b1F1)Ψ : om. b1 F1

6 (a5) πρωÄ τον α γ :

c

προ τερον K β δ S (1270.28)

2 YÕÏkI (scil. ÙHÒÕLA  xwLA) scripsi :

YÕkB Φ : YÕkT Bad

9 XÂ| addidi (= εÏ τερον Π = aliud

11 \iÏbLA \B scripsi (= δηÄ λον Π = tunc manifestum est C = manifestum \iÏbLA †LD \B Φ : ›btI †LD \B Bad a scripsi (jÍ Cz ... a (iL = ουÆ δεµι α Π = non est aliquis ... omnino C = nullus est V) : jÍ Cz Φ1 Bad : øCz Φ2

CV) : om. Φ Bad est V) :

2

ømÂkÊtsmLA

1270.14)

with corresp. marks s.l. and i.m.:

ømK ÙQ cÚLA YÒÕiaA cÒ 

Φ i.m. (cf. Simpl.

5 †LÚK with corresp. marks s.l. and i.m. and a note written in a separate column

øÏKÖ fÏ yibP J øÏK „ õiwI Yz úJÕL èÖÏi Ò ÕE YÕÏktLA YÑK ÕL fLA ÕBz ÑelÏK aÑiWtÒ dÑsF RÕO Yz †LD \M ‡‚liF ÙSÑF ÕeF \lÑK Φ i.m., fort. ex comm. Philop. 840.27f., 900.8–10 6 YÕÏktI with corresp. marks s.l. and i.m.: øqtnI vÏ cz Φ i.m., fort. ex comm. Ibn

(cf. p. LXXII, note 153):

al-Tayyib 885.14f. ˙

261a

261a5

74

3

Av. t. 58 6

(iLÑPÕpSrt fÏ yibpLA û4LA êÑtK \M nMÑoLA LÑqmLA

áÕltT LA ÑKaÑB YÒ  Ö .‡ÙQÒ  ÕE jÍ CÒ  áÕltT LA ~gIÒ  ÑKaA .YÕÏktLA ÙyB \M ÑelÏK ÑeNÏÒ  †LDÖ dÑsfLÒÖ èjhmGiÒÖ LÑhtSiÒ vÏ ÕÏhA lÑS \M ÙHÒÖ iÖ (ilF lqnLA \M ‡ÙQÒ  YÕÏktLA (iL 〈 DÒË 〉 †LÚlF .юM ‡ÙQÒ  ~gIÒ  iÏvtLA [Ò222] ŸÑnCÒ  ÒË ÙbM ŠÒË cO ~qÒd ÕEÖ  QÑN YÕÏktmLA YÒ  EÑzF lm`ÑBÖ Í ÑM õi? X| lqnLÒÖ yibpLA „ ~MÙqtM YÕktLA „ XÏÑ  tmLA YÕkiF ÙJÕI Ï Â 5 (a13) τε E2, fort. Ψ, β δ J1Qw Pa(398.25) : om. E1 F : τε ειÆ K Sl(1271.19) : δεÁ HIJ2 Λ (a13) καιÁ ] αÆ ειÁ fort. Ψ

6 (a14) τηÄì γενε σει Ψ K γ δ LN : τι γενε σει E : τηÄ ς γενε σεως R

6-p. 75.1 (a15) παÄ σιν ... γενε σει] παÄ σιν τοιÄς εÆ ν γενε σει υë πα ρχουσιν intell. Ar.

áÕltT LA scripsi (= τωÄ ν εÆ ϕεξηÄ ς [κινη σεων] Π = qui sequuntur eam C = [motuum] consequentium V; v. also the immediately foll. clause) : áÕlbB gÙLA Φ : áÕltI cÚLA Bad jÍ CÒ  scripsi : øCÒ  Φ Bad; v. App. 2 ad p. 73.11 2 vÏ scripsi (= ειËτ’ beNRF vel ειËτα EKLHJQ = postea C) : Ö Φ (= et V) Bad post YÕÏktLA iter. †LDÖ dÑsfLA Φ, sed del. 3 DÒË addidi (= ειÆ Π = cum C) : om. ΦV Bad 5–6 ~MÙÏqtM ...  QÑN scripsi (sim. diminutum et hoc 1

1

semper currit ad principium quare posterius in eo quod generatur est prius C / diminutum et

YÕÏktmLA pro YÕÏktLA leg. CV]) : ~clqbM  QÑN Φ in textu, with mark btw. the words and ŠÒ gT (ÑbBÒd vel) ;Bd ÕEÖ ÑgbqnM YÕkbLA † HÑbmLA YÕkbQ ÒÙbM Φ i.m. : ÑMÍÙqtM \IÕktLA † XÑ  tmLA YÕkiÂF ÒËÙbM ŠÒË gO ~blÒd ÕEÖ  QÑN scr. Bad semper currit ad principium et sic posterius in generato est prius V [sed

lÑS YÏ z õM ҁiÏvtLA èÖÏz (Φ YÕkLA : scripsi YÕÏktLA) YÕÏktLA YÕkI Yz ‚O G ÒDÒËÖ ÑKaA èÖÏz YÕkT i Yz gHz áÚeF áÕltT áÚE LÑhtSiÒÖ  qnLÒÖ ÕÏhÑK ÑKaA Φ i.m.; cf. Simpl. 1271.9–16 ad 261a11f. 6 lqnLÒÖ with dots connected to the note: lqnLA KH cz 3 gloss without mark:

Φ i.m.

261a10

261a15

75

VIII 7, 261a9–25

3 Av. t. 59 v

C 77v Cp 201r 6 Av. t. 60 Ca 27vb

†LD èÑoM LvÒ ‡ÙyL ÆCÒ  …ÏhtI i fÏ aA ¡yB rÑC †LÚLÖ .YÕÏktLA „ lqnLA NÑK DÑËF .ñL ÑeNÏÑËF øMÑkLA ÑMÏÑ  F YÒÕiaA \M ¡oK %ÑnJÒ  Ö ÑbnLA KaA áÚE YÕkT YÒ  úO ÙqF †¦ÖÊÒ  yibpLA \M ñzÏH YÑK .F cÏ aÑB ;NÏÒË ŠÖiªÒ fE ×qF êÑbSi Ò †ltL (iLÖ ÑKaA lÑS ŠÖÒª ÒÚLÑB ~gIÒ  lqnLÑB 〈 …ÏhtI ;L 〉 ÑEÙiybT ÑKaA lÑS ›B \M ~gIÒ  ñNÏi  øB iÏvI ;K ñtiNÏz „ ÆCÒ  〈 á³ 〉³iÏvT³ i ÑEÙHÖ ÑeNÏÑËF …ÒD ­IÒ  áEÕJ \M YÏ Ò  \iÏbLA \MÖ .èjhmGiÒÖ Â ÕÏhA ÙnY ˆ° kLA iÏvIÖ LÑhtSiÒ ÙnY ž³ikLA ÙQÖ .KaA áÚE èÖÏi Ò ÙcqLA ‹Y …UÏ ;NÏÒË Í CÏ ÑX ñTÒD ÕE …ÏdA 1 (a16) αÆ κι νητα Π(praeter E1)Ψ : α***κι νητα [ras., an αÆ ειÁ κι νητα?] E1

(a16) τουÄ οÆ ργα νου

E1Ψ e2 β γ : exp. E2 : om. K be1 Pc(900.26) Pa(399.15–16) : secl. Ross

5 (a21) κινη σεων

Π(praeter β)Ψ : κινουµε νων β

7 (a23) µα λιστα] non vert. Ar.

8 (a24–25) καταÁ το πον]

κι νησιν fort. Ψ (nisi fı¯ l-maka¯n post al-haraka addendum ?)

˙

øMÑkLA corr. Φ2 ex øMÑaA (? parum clare) DÑËF (= ωÏ στ’ [ωÏ στε LNR] ειÆ Π = cum igitur CV) Φ : ÒDÑËF Bad 4 ŠÖÒª scripsi : parum clare Φ : èÖÒ  Bad †ltL scripsi (= δια ... ταυÄ τα Π = propter has C = propter ista V) : lÑsB Φ Bad 5 ñNÏi   scripsi (= διο τι Π = quoniam C = propter hoc ... quia V) : ÑˆÒ Φ : шÒË Bad …ÏhtI ;L ÑEÙiybT scripsi (= εÆ ξι σταται τοÁ κινου µενον Π = removetur illud quod movetur V; cf. εÆ ξι σταται = ÙYÑbtI 235b9 = Bad 666.7 : quod mouetur ... permutat (= …ÏhtI ÑM iÏvtI) C : ÑEÙbybB vel ÑE ybB Φ (om. …ÏhtI ;L) : ÑE¡vtB Bad 6 áiÏvT³ scripsi (= mutat ipsum C = transmutat ipsum V) : ¡vT Φ Bad ñtiNÏz ΦCV : яiN| Bad 7 iÏvI scripsi : á¡vI Φ, fort. CV, Bad 8 ;NÏÒË scripsi (= non [mouet] nisi C; non vert. V) : ;NÏÑËF Φ Bad post KaA add. (iLÑPÕpSrÒ èÑQ Φ : om. ΠCV : secl. 2

ÑMÏÑ  F] ÑMÒ  Bad

Bad

2 øMÑkLA with mark s.l. and gloss without mark: (Φ ÑbnLÒÖ : scr. Bad ÑbnLA) ÑbnLA fÏ aÑB yI

gÕqLA áÚE ñL YÒÕiaÒÖ øomlL ÙLÏÕM ÕÏQÖ iMÑN ÕÏQÖ IDÑV ÕÏQ ñL ÑM ÕE fÏ aA YÏ i  YÒÕiaA ›BÖ fÏ aA ›B ƒÏFÖ KaÒÖ ( Ï aA ÕÏQ †LD õM ñLÖ Φ i.m.; cf. Philop. 900.29

261a20

261a25

76

Av. t. 61 4

Av. t. 62

8

Cp 201v

(iLÑPÕpSrt fÏ yibpLA û4LA êÑtK \M nMÑoLA LÑqmLA

;NÏÒË ÑKÏhmlLÖ ÑKÏhtmlL ÑKÏhtmLA „ èÖÏi Ò〈 Ö 〉 Ò  ÙbmLA YÏ ÒË èÕqN .ñTÒD …ÏdA YÒ  ÒÚE ÕE ÑMÏÒ  Ö [1Ò223/10Ò222] ÑKaA èÖÏÒ  lqnLA YÏ Ò  †LD \M eZ ÙqF øibsLA áچ †LD õM eziSÖ .øbÂQ cD \M áÕniÏbM ÑN¬ ÑËF ŠÖiªÒ fE lqN cÏ Ò  ÙQ ñNÏÒ  \M ‡ÙÏqT .FÖ õGÕmLAÍ ÒÚE „ ~yGÖ áÑnyGÖ cÚLA Mi Ò юiyB \km (iL ñNÏÒ  ezI ÙQ ñNÏÒË èÕqnF .È iÏLRÒ  È lctÏM ÑM È KH YÕkT YÒ  \km YÏ Ò  †LDÖ èÕqLA ÒÚE \M Í lctÏM ÑKaA lÑS \M ÙHÒÖ iÖ YÕkT YÒ  èÑoM ñL øBÑqmLA ŠÒË øBÑqmLA \M YÕkT ;NÏÒË ÑelÏK iÏvtLA ŸÑnCÒ  Ö ÑKaA LÑhtSjL \IÙÏaÒÖ dÕJÕM iÖ dÕJÕM dÑsfLÒÖ YÕklL \IÙÏaA YÒ  †LD è;K ÖÒ  vÂcLÒÖ ˆzÂyLA èjhmGiÒÖ ÕÏmnlL \IÙÏaÒÖ YÒdÏÑgtmLA YҁOi Ò  Ñe`A ŠÒË ÑŽM YÕkT LÒÖ ñNÑcqNÖ ˆzyLA .dÏÑgtM feF dÏÑgtmLA 2 (a26) αυÆ τοÁ αυë τοÁ β δ IJ (?)Q1 Λ : αυÆ τοÁ εë αυτοÁ E K Q2 : αυë τοÁ Ψ H : αυÆ τοÁ F fort. Ψ, K Sc(1273.21) : καιÁ τοÁ β γ δ

8 (a33) αιë α γ : καιÁ K β : καιÁ αιë δ : om. T(227.8)

9 (a34) ϕθοραÄì Π(praeter E )Ψ : ϕθοραÄ ν E2 2

5 (a29) καιÁ 2 E,

(a34) τοÁ οÃν καιÁ τοÁ µηÁ οÃν E K β δ Tp(227.8–9),

fort. Pp(901.27–28) : τοÁ οÃν καιÁ µηÁ οÃν γ Sc(1274.9) : οÃν καιÁ µηÁ οÃν fort. Ψ

11 (a36) αÆ τε λεια

Π(praeter E1)Ψ : αÏ τε αÆ ειÁ E1

Ö addidi (= καιÁ Π [a26] = et C) 3 post ÑKaA add. ylytLA X| Φ Bad : om. ΠCV : ante ÑMÏÒ  Ö add. (iLÑPÕpSrÒ èÑQ ®ÑyLA ylytLA Φ Bad : om. ΠCV : seclusi 7 YÕkT] s.p. Φ : YÕkI Bad 8 post iÏvtLA add. áÚE Φ Bad (= iste C) : om. ΠV : seclusi 11 Ñe`A] partes CV (= a|‚JtÒ vel Ñil‚`A ?) 1

1

seclusi

10 gloss without mark:

Y z RÕO i ñiLAË hnI 〈~m〉zY fyibP J 〈øÏk〉— L YÏ i  †LD èÑQ (vid. ñNi) ñNÏÒË

〈 〉

áRÖÑO Φ i.m. (partly cut off at the margin); cf. Alex. frag. 721

261a25

261a30

261a35 261b

VIII 7, 261a25-b11

4

8 Cp 202r

77

YÑK ÙqF dÕJÕM ø³bQ \M ÕEÖ ÑLAË rÑwM Í KH ÙBÒ  …ÏhtI \kI G ÑMÖ ÒÚE ‹YÖ ÙÏgLA „ rÑC ÒDÒË \ksI iÏvtmLA YÏ Ò  EÑZÖ .~nKÑS ø³bQ \M Í rÖ² ‹Y YjBÑqtM YÕÏktLÒÖ dÑsfLA YÏ Ò  †LDÖ iÏvtLA ŸÑnCÒ  „ Mi Ò cO èÑomLA „ ñnM YÕkI ;L øÈBÑqM  xW  xW „ ñnM YÕkI ÑMÖ ƒjPiË Ò …nC ~yM iÏvtI aLA YÕkI YÒ  \km \kI G YÊ ÒË †LÚlF . xW  xW YÕkI YÒ  úJÒÖ øB ÆctÏM iÏvtLA YÕkI YÒ  \km (ilF ›lBÑqtmLA iÏvtLA YÑgQÑntmLA iÏvtLA ÑfnC YÕkI YÒ  ›B ƒF i ñNÏÒ  †LDÖ .YÑMR ;ŽiB ;F \kJ ¡V áÙHÖ ÒÚE ‡Òd ÑM \IdÏÑgtM ÑNÕkI i YÒ  ›BÖ \IdÏÑgtM ñiLAË ÑtU ÑJÏ (iL †LD YÏ ÑËF ñniyB ÙHÒÖ a „ ~yM ÒÙJÕI YÒ  YÒ  YÕkI YÒ  Í rÖ² úO (iL ñNÏÒ  ŠÒË iÖ ~gIÒ  ñiF ÑtU iÖ èÕqLA ÒÚE „ 2 (b2) ουË ν Π(praeter ER1) : δ’ E1Ψ Pa(400.12) : del. δ’ et ουË ν s.l. scr. E2 : om. R1 µεταβα λλειν Π(praeter E )Ψ : µεταβα λλει E 1

om. K β δ F Λ

1

5 (b6)

9 (b10) ουÆ δεÁν α HIJQw, fort. Pa(400.24) :

10 (b10) λο γωì ΠΨ Ross : οÏλωì Bek Pra (err. typ., v. Torstrik, p. 529)

(b10) µηÁ αÆ να γκη α K e2 β γ : αÆ να γκη µηÁ b2 : αÆ να γκη b1e1

эLAË rÑwM Φ : эLAË rÑwM Bad (cf. Galeni De partibus artis medicativae [...], ed. Lyons, p. 66.15–16; Averroe`s. Tahafot at-tahafot, ed. Bouyges, p. 358.14–15) 2 EÑZÖ Φ : ҁÍEÑZÖ scr. Bad 5 †LÚlF Φ : †LÚLÖ scr. Bad 6 ›lBÑqtmLA scr. Bad : YjBÑqtmLA Φ úJÒÖ scr. Bad : ÑbJÒÖ Φ 7 YÑgQÑntmLA scr. Bad : ›gQÑntmLA Φ 8 i YÒ  ] iÒ  scr. Bad 10 ñNÏÒ  ŠÒË iÖ scr. Bad : ñbL hlLiÖ vid. Φ : neque quia (= ñNÏi  iÖ ?) C : scilicet quod (= ñNÏÒ  YÒ  Ö ?) V 1

4–5 gloss without mark: YÕkLA yI  xwLA †LD „ YÕkI ;L jÍ BÑqM YÕkI ñNÏÑËF  xwLA „ YÕkI ÑM cz

dÑ〉sfLÒÖ

〈

904.4–6

Φ i.m.

9 ñniyB with corresp. marks s.l. and i.m.: ÙHÒÖ YÑMR „ ˆeFÒ Φ i.m.; cf. Philop.

261b5

261b10

(iLÑPÕpSrt fÏ yibpLA û4LA êÑtK \M nMÑoLA LÑqmLA

78

3 Cv 78r

6

9

YÏ ÑËF) YÕkslL ÒÙ­ G ÕE (iL iÏvtLA YÏ Ò  ŠÒË iÖ ›gQÑntmLA „ YÕksLA ŠÒË cdÏqI ;NÏÒË dÑsfLÒÖ \ksI YÒ  RÕO i YÕkI YÒ  „ÈilhF dÕJÕl (iL ÑM Mi Ò ÒÚE ;NÏÒË èÕqLA ÒÚE „ ñiLAË ÑtU cÚLA øB (dÕJÕl (iL ÑM YÒ  \km G †LD YÑK YÒË ñNÏÑËF .YÑMR ;ŽiB .F YÕkI øE YÒ  áÙHÖ øB dÏÑgtLA ŠÒË \kT G ›LÖÏi Ò \IMi Ò „ ÑntJÑH YÏ i  ÆctÏM iÏvtLA YÕkI .~yM ÒÙJÕI YÒ  \km i ñNÏÒ  ŠÒË èÑoM ¡oK aÑiWi  ÒÙ­ G YÕkI ÙHÒÖ ~miW YÒ  ÑnpÂlvI³ YÒ  fvbnI (iLÖ znN YÒ  fvbnI ;NÏÒË øB ÙÏgLA ŠÒË KhlL ÙÏGÖ ŸÕQÕlL ÙÏG KaA YÏ Ò  †LD Y ÕksLA〈 Ö 〉  KaA øBÑqT³ ÙQ ñJÖ YÖd ñJÖ \M YÏ Ò  á¡V YÖd †LD „ Í Í Â 1 (b11) µεταβοληÁ ηÆ ρεµι αì α K β γ F Λ Pl(904.7) Sc(1276.6–7, ms. A; 1276.16, ms. a) : µεταβοληÄì ηÆ ρεµι α HIJQw Pp(904.11) : µεταβοληÁ ηÆ ρεµι α Sc(1276.6–7, ms. a)

Π(praeter E )Ψ : exp. E 2

2

(b13) εÍ στιν E δ : εÍ σται Ψ K β γ Λ Pa(400.30)

Π(praeter E ), fort. Ψ : om. E 1

1

4 (b13) ουÆ κ 5 (b14) ηë

8 (b18) λαµβα νειν Π(praeter E K)Ψ : λαµβα νει E1 : 1

λαβειÄν K

1 YÏ Ò  ŠÒË corr. Φ2 (= CV) :

ŠÒË ŠÒË Φ1 2 ;NÏÒË] semper (~qÒd) C : autem V 5 \kT] s.p. Φ : \kI Bad 8 ŸÕQÕlL scripsi : ŸÕQÕLA Φ Bad znN scripsi (= nos oportet ut consideretur C = oportet nos considerare V) : s.p. Φ : znI Bad 9 YÏ Ò  (= οÏτι) non vert. C : scilicet V (= cz ?) Y ÕksLÒÖ  KaA scripsi (= τηÄì κινη σει καιÁ τηìÆ ηÆ ρεµι αì = [opponitur] motui ... et quieti C : motus ... et quies [Y³ ÕksLÒÖ ³ KaA] interpr. V) : YÕksLA KaA Φ : Y ÕksLA ³ KaA Bad

1 ÕE (iL with corresp. marks s.l. and i.m.: ›lBÑqtmLA ›B ƒfLA \iÏbN Yz õGÕmLA áÚE „ ÑnB JÑH i cz

Yz ¡QÑntmLA „ ÙJÕI (Φ Yi : scripsi i) i YÕksLA YÏ z \iÏbN Yz iÖ ÑsiL \IÚlLA ›lBÑqtmLÒÖ \IÙÏgLA YÕksLA ñiF ÙJÕI i dÕJÕl (iL ÑM YÏ t ;ŽM dÕJÕM i „ Φ i.m.; cf. Philop. 904.9–12 7–8 gloss without mark i.m.: ›teJ \M YÑK ÒDÒË knI i ÒÚE Φ i.m.; cf. Philop. 904.17–19

261b15

79

VIII 7, 261b11 – 8, 261b32

Av. t. 63 Cp 202v 5

Av. t. 64

10

ñNÏÒ  Ö ¡Â cqtLÒÖ ØÂҁFiË Ò øBÑqI ÙcÊqÂLA ÖÒ  cÖÑstLA YÏ Ò  øoM ³ dÏÑgtmLA ³ KaA YÑfnC iÖ YÑtlBÑqtM YÑtKH i ~yM [ê 223] 〈 Ò 〉ÙJÕI YÒ  \km i YÕkLA MÒ  „ \Ï zI³ YÒ  ¯ibqLA ø¬ K ¯ibqLA \M ñNÏÑËF ~gIÒ  Ö .YjBÑqtM iÏvtLA \M iÖ YÑkmLA ‹Y ÙsfI YÒ  Í rÖ² úJÒÕF YÕÏktI ›H aLA YÏ Ò  dÑsfLÒÖ XÂiªÒ lÑS „ !ymLA Òچ ³ qoLA †LD øbÂQ \M YÕktF ÆCÒ  ~NÑMR bI Â Ï K эF ÙHÒÖ Â Â èÑoM ‹Y yibpLA gK YÏ i  .Ñel Iш jB ÑM KH ÑnEÑE YÕkT YÒ  \km ÙQ ñNÏÒ  YvÒ YÕtboM \VÖ øqtnM øÏK YÏ ÒË èÕqnF .rÖd ÑMÏÒËÖ øqtnmLA liT ÑMÏÒË YÕkT ÑeNÏÒ  Ö lctÏM ÙHÒÖ \kT G YÊ ÑËF Í pltM Í KH ÑMÏÒËÖ Í miqtsM Í KH ÑMÏÒËÖ rÖd ÑMÏÒË …ÏhtI ñNÏÑËF ~yi? ;ŽM fLÏqmLA iÖ YÕkT YÒ  \km (ilF Í lctÏM †ÂniÊT \M ÙHÒÕLA ÆctÏM …ÏhtI (iL áÑntM yqtsM ×ÏX ‹Y øqtnmLA YÏ Ò  \iÏb LA \MÖ .Í lctÏM Í 4 (b22) καιÁ 2 E K be1 FIJQw Λ : om. e2 β H : non vert. Ar.

7 (b27) ειËναι τινα Π(praeter

Ew) : τινα ειËναι E : ειËναι τινα vel τινα ειËναι Ψ : ειËναι τι w : τινα Λ τεχνη η κυ κλωì E , sed exp. η τεχνη E : ηà τε χνηì ηà κυ κλωì fort. Ψ 1

2

8 (b28) ηë κυ κλωì ] η

(b28) µεÁν E, fort. Ψ, K β

JQw Λ : om. δ FHI T(227.31) 1 ³ KaA Φ :

³ KaÒÖ scr. Bad post dÏÑgtmLA add. dÏÑgmLA (interpol. var. reading) Φ Bad : om. ΠCV : seclusi ÖÒ  ] καιÁ Π : aut C : et V 2 ÒÙJÕI] ÙJÕI Φ Bad 7 ante \VÖ add. lqnLA〉 8 〈lctmLA Bad !ymLA Òچ with mark s.l. and gloss without mark: ÑM ~NÑMR ñLÑT bI iÏvtmLA YÏ z cz Φ i.m. XÂiªÒ with corresp. marks s.l. and i.m.: aÑbvLA \M ñNÏz †LDÖ dÑsfLÒÖ YÕkLA ¡V ҁiÏvtLA XÂiªÑB yI rÑC ÒDÒË vÏ ñiLAË …ÏhtT a „ yibpLA (Bad û­T : Φ s.p. : scripsi) û¯T Yz (Φ ÑbyLA : Bad aÑbvLA) YÑkmLA ‹Y ñtQrÑF ñiLAË Φ i.m.; cf. Philop. 905.4–7

5

261b20

261b25

Chapter 8

261b30

(iLÑPÕpSrt fÏ yibpLA û4LA êÑtK \M nMÑoLA LÑqmLA

80

a

C 28ra Cp 203r 3

6

9

›tKH …ÏhtI MÑqtSiÒ ‹Y ~yJÒr ‡ªknI cÚLÒÖ ~yJÒr ‡ªknI ñNÏÒ  †LDÖ øfSÒ  ŠÒË KhlL dÏÑgM ƒÕF ŠÒË YÑkmLA „ KaA YÏ ÑËF .›TdÏÑgtM KhlL dÏÑgM Å5 KaÒÖ žlX ŠÒË KhlL dÏÑgM ‡ÒÙÏQ³ ŠÒË KaÒÖ .YÑkmLA „ dÏÑgtLA Ÿ³ÑnCÒ  áÚE YÏ Ò  †LDÖ ~nim ÑnlqF lctÏmLA ÙHÒÕLA KaA fE KH cÏ Ò  ‡ÙÏqT .F Ñncbّ ÙQÖ .rÕcLÑB žlt[ i .FÖ ÙHÒÖ YÑMR „ ÙHÒÖ aL YÕkT LA KaA ÑeNÏÒË (áiÒË ÖÒ  YÑsNÒË †LD èÑoM) …³ÏhtmLA aÑiWÒ  OjO †LD ñB {Ï I cÚLA YÏ ÑËF ÒÚEÖ KaA YÕkT ñiF cÚLA ÕE ÓLÑoLÒÖ (YÑM‚LA †LD èÑoM) "MÖ žlt] ÑeNÏÑËF dÏÑgtmLA ÑMÏÑ  F .ˆzÂyLA ÖÒ  rÕcLA ÖÒ  Oi Ò Ö〈 Ò  〉 YÑkmLA ÕE  1 (b32) αÆ νακα µπτει Π(praeter δ)Ψ : αÆ νακαλυ πτει δ κινη σεις ειËναι E1

(b34) κινη σεις Π(praeter E1)Ψ :

2 (b34) καταÁ το πον ηë αÍ νω τηÄì κα τω E K β δ Sp(1278.25) : τηÄì αÍ νω ηë κα τω

καταÁ το πον F : τηÄì αÍ νω ηë κα τω H : καταÁ το πον τηÄì αÍ νωì ηë κα τω IJQw : ηë αÍ νω καταÁ το πον (nisi ηë καταÁ το πον αÍ νω) τηÄì κα τωì interpr. Ar.

3–4 (b35–36) αÆ ριστεραÁ ... δεξια α K β δ T(228.6) :

5 (b36) ηë α K γ Sp(1279.2) : om. β δ

inv. γ Pp(842.8) Sp(1278.25–26)

7 (a2) ηË ν E K β δ

(?)F J Λ : εÆ στι[ν] HIJ Qw P (907.22) : εÆ στιν, εÆ ξ ωÎ ν συµπληρουÄ ται ηë κι νησις Sp(1278.33) : 2

1

l

εÆ στιν, εÆ ξ ω Î ν συµπληρουÄ ται τουÄ το fort. Ψ

8 (a4) τρι τον τοÁ δ FHIw Pl(907.23) Sc(1278.37) :

τοÁ τρι τον E JQ : τρι τον E K β : τοÁ τρι τον τοÁ vel τρι τον τοÁ Ψ 1

2

fort. Ψ : εÆ στιÁ ν ηà FH : εÆ στι οιÎον S (1278.38) c

9 (a4) εÆ στιÁ [ν] Π(praeter FH),

(a4–5) το πος ηà πα θος ηà ειËδος ηà µε γεθος α K β

δ : πα θος ηà ειËδος ηà το πος ηà µε γεθος γ, fort. As(frag. 723) : οë το πος ... ηà πα θος ... ηà µε γεθος ... ηà ειËδος Sp(1278.38–1279.1)

1

1et2

‡ªknI scr. Bad (= ΠCV) : ‡tkI Φ

scilicet secundum speciem V) vert. C) : seclusi

7

9

6

rÕcLÑB] + ûÕnLA s.l. Φ (cf. secundum formam

7 post áiÒË add. èÑoM ÒÚE Φ Bad : om. ΠCV (áiÒË ÖÒ  om./non

ÖÒ  scr. Bad (= ηÃ Π = aut CV) : Ö Φ

1

áiÒË with corresp. marks s.l. and i.m.: il4LA ‡ÒJtÒ \M ~MJ yI Φ i.m. ex comm. Philop.

907.25

261b35

262a

262a5

81

VIII 8, 261b32 – 262a15

3

Cv 78v Cp 203v 6

Av. t. 65 9

.YÑkmLÑB ÑeFjtX€F ÑEÑNKD LA ŸÑnCi Ò †lTÖ ÙHÒÖ siLÖ rÕcLÑB ;eNÏÒ  Ò ŠÒË ê \M KhlL dÏÑgM ê ŠÒË Ò \M KaA YÏ Ò  ‹Y øiLÙLÒÖ YÒËÖ Mi Ò cO ~gIÒ  †LÚKÖ .~yM ÑtNÑK "M YÑyN0TÖ YÑMÖÑqtTÖ YÑfQÒÕtT Ò ÕV ê \M KaÒÖ ê ÕV Ò \M KaA †LD èÑoM lÒd ‹Y ÑtNÑK iÖ ûÕJr ;= YÕkI iÖ ›tlctÏM ÑtNÑK YÒËÖ YÑMÖÑqtTÖ YÑfQÒÕtT ;eNÏÑËF KaA siLÖ .YÑyN0mLA YÒÙSÑftmLA ;E \IÙÏgLA YÏ Ò  øJÒ  \M žpY .ƒÕF ŠÒË KhlL MÖÑql úNÑJ ŠÒË ×ÏX ‹Y KaA YÕkT YÒ  \km i ñNÏÒ  ¤Â KÒ  rÕeZ ñB ezIÍ ÑJÏ Ö .F (iL žqI YÒ  Í rÖ² úO ÙqF ~yJÒr žpY ÑM YÏ Ò  Í lctÏM yqtsM ñNÏÑËF [Ò224] .lÒd ‹Y øqtnI YÑK ÕLÖ øB ×qF yqtsM ×ÏX ‹Y øqtnI 1 (a6) αιë ειÆ ρηµε ναι Π(praeter E1)Ψ Sl(1277.35) : inv. E1 καà ν ειÆ ] εÆ πιÁ Ψ (v. GALex I, wa-in 1.1)

Sp(1280.3) : τοÁ Β K1 be1 β Λ : τοÁ Α Ψ 1

1

FJ : ουÆ χ ηÎ ς E : ουÆ χιÁ ηë w

3 (a8–9) καιÁ εÆ πιÁ ] fort. καà ν [vel

4 (a10) τουÄ Α] τουÄ Β Ψ

(a10) τοÁ Γ E K2 e2 γ, fort.

6 (a12) ουÆ χ ηë E2Ψ K β δ HIJ2Q Sl(1279.18) : ουÆ χιÁ

(a12) τοÁ πλα γιον E1 K β HIJQw Sl(ms. a) Pa(403.10–11) : ταÁ

πλα για δ T(228.14) Sl(ms. A) Sp(1280.13) : τι πλα γιον F1 : τηÄì πλα γιον F2 : πλα γιον E2, fort.

Ψ, Pl(842.19)

10 (a14) εÆ π’ ευÆ θει ας E K β δ : εÆ πιÁ τηÄ ς ευÆ θει ας fort. Ψ(v. Glossary s.v. ευÆ θειÄα)

γ S (1280.23.27) p

3 YÑfQÒÕtT scripsi (= ιë σταÄ σι[ν] Π[om. Q] = resistunt C : contradicunt V) : s.p. Φ :

YÑqFÒÕtT Bad

;eNÏÑËF iter. Ò ÕV ... YÑfQÒÕtT (l. 3–4) Φ : seclusi YÑMÖÑqtTÖ YÑfQÒÕtT scripsi (= ΠCV) : YÑMÖÑqtIÖ YÑqFÒÕtI Bad YÒËÖ scripsi (= καà ν E1KSc[1280.6] vel καιÁ εÆ αÁ ν E2βδ FHI = nam quamvis C : si igitur [= YÑËF?] V) : YiÖ Φ : YiË Ö Bad 9 .F scripsi (= in eo quod C) : illud quod (= ÑM ?) V : ;M Φ Bad 10 1‹Y] + note without mark i.l.: ‹Y zfL cÙY \BÒ xsN „ 2 1 (эlY Φ : ñilY Φ ) эlY 〈êÖ〉³M (»in Ibn Adı¯’s copy the word ala¯ [= εÆ π’/εÆ πιÁ a14] is erased«)

5 post

s.p. Φ :

Φ i.m., partly cut off (?cf. App. 1 ad loc.) 3 gloss without mark:

MÖÑqmLA \M i ‡ÑiqLA \M cz Φ i.m.

262a10

262a15

(iLÑPÕpSrt fÏ yibpLA û4LA êÑtK \M nMÑoLA LÑqmLA

82

3

6 Cp 204r

эlY …ÏhtI 〈 ñNÏÒ  !yMÖ lÒÙLA  KH 〉 …ÏhtI aLA YÏ Ò  !yM (iL ÒDÒË YÑK ;BÏrÖ 〈 ñsfN ‹Y õJI ÕEÖ …ÏhtI ;BÏr ñNÏÒ  †LDÖ ñniyB ÙHÒÖ〉 „oI ÙQÖ .~yJÒr Ñ  fkN ñtKT Ò  ÙtB ñnM cÚLA õGÕmLA ŠÒË ÑŽM DlB (iL žqT YÒ  Í rÖ² úJÒÕF gcA ÒÚE cQ LA KaA YÏ Ñ  B YÑsNiË Ò aÑiWÒ  OjO ÑnEÑE NÑK ;L †LD zÙbM ÒÚEÖ .~gIÒ  %ÑiqLÑB øB ×qF ( Ï aÑB ÕEÖ ~yi? ;E ÕE ›nOiÒ \M ÙHÒÖ øÏK ÙnY ×SÕLA YÏ ÑËF X|Ö ×SÖÖ zÙbM øyfLÑB ÑM ¡V ÕÏqLÑB ÑM YÏ ÑËF ~gIÒ  Ö .YÑnOÒ %ÑiqLA „Ö ÙHÒÖ dÙyLA „ YÖd \M fE LA ×qnLA \M ÑemEÏÕtN pqN cÏ Ò  yqtsmLA ×ÏbA YÏ ÑËF †LÚLÖ 1 (a15–16) κυ κλωì ... κυ κλον Π(praeter FH)ΨΩ : inv. FH

ΨΩ : om. E1

3 (a17) πα λιν] non vert. Ar.

κι νησιν ιÏστασθαι intell. Ar.

(a16) καιÁ Π(praeter E1), fort.

4 (a18) αÆ να γκη ιÏστασθαι] αÆ να γκη τηÁ ν τοιαυ την

6 (a20) τελευτηÄ ς Π(praeter E1)Ψ : τελευτη E1

7 (a22) τοÁ 2

Π(praeter be1)Ψ : om. be1 1–2 ñsfN ...  KH scripsi (= ταυÆ τον [ταυÆ τοÁ E2β : τοÁ αυÆ τοÁ γ ] ... κινου µενον 262a15–16 (v. App. 1) = non est intentio quod aliquid mouetur circulariter ipsa intentio quia mouetur [mouetur CpCa1 : mouet CvCa2] circulus una eadem quod est quia quandoque mouetur et ipsum redit super ipsum C = non est ... eadem intentio aliquid moueri circulariter et aliquid moueri per circulum forte enim mouetur eundo super ipsum V) : Bad

8

Ñe>ÕtN scr. Bad : Ñe>ÕB Φ

эlY …U ÕEÖ Φ Bad

2

;BÏrÖ Φ : ;BρF

ѓfN lÒÙLA KhK lÒÙLA ‹Y ;LA KH siL Φ i.m. 7 %ÑiqLA with corresp. marks s.l. and i.m.: Í IÑVÖ Ò¡ ÙbM YÕkI Yz ŠÒË %ÑiqLÑB cz Φ i.m. = Alex. frag. 730 8 cÏ Ò  pqN with corresp. marks s.l. and i.m.: ÕÏqLÑB B ñtmsQ øbQ ñNÏÑËF yqtsmLA ×ÏbA \M ÑEÑnGF pqN cÏ Ò  …ÏhtmLA …ÏW ÒDÑËF Ò¡ ÙbM rÑC юM ×ÏbA Q³ ÒDÒË htÏH юM KaA nÙtbT (iLÖ Y| „ эlY Ïm …ÏhtmLÒÖ (after YÑK partly cut off and ambiguous) юM/zÙbM øyF *** aÒÙtBiÒ YÑK юM Φ i.m.; cf. Philop. 1 gloss without mark:

842.32–843.6

262a15

262a20

83

VIII 8, 262a15-b3

3

6

9

žQÖÖ ÒÚE Q ÒDÒË ÇÒË øyfLÑB Í dÕJÕM siLÖ ÕÏqLÑB hpSÖ³ feF ñiFP zÙbM ÑMÏÒ  ÐÑgqNÖ Ò¡ ÙbM hpSÕLA ÚmniH YÕktF .…ÏhtI Ò  ÙtB€F dÑY vÏ ÑEÙnY žQÕF øqtN Ò YÏ ÒË lQ YÑ  K †LD Í èÑoM .ŠÖulF aÑgqN ÑMÏÒ  Ö ¡Xi Ò Kh