Bryn Mawr Classical Review

Byzantine Aristotelianism has shared in a recent growth in interest in Byzantine philosophy,¹ with several previously unedited works appearing since 1994 in Commentaria in Aristotelem Byzantina.² A central figure in this revival, Professor Linos Benakis, has now augmented this series with the editio princeps of a complete medium-length and avowedly pedagogical commentary on Aristotle's Physics by the eleventh-century polymath Michael Psellus (1018 - ca. 1078).³ It was previously available in printed form only in a Latin translation of 1554 (title page reproduced at 450). It can be dated to the 1060s, if the person addressed as "the noblest of my pupils" in the proem (1.7) and on points of detail throughout (see 25 n. 43; e.g. 34.8, 177.2, 177.23, 239.14, 395.26, 430.18-21), is, as seems certain, the future emperor Michael VII Doucas.

Byzantine Aristotelian commentaries are indisputably inferior to those of late antiquity and to the major ones from the Arabic and Western European Middle Ages and the Renaissance. A century ago they were represented in Commentaria in Aristotelem Graeca (see 10 with nn. 13 and 14) by a small and arbitrary selection. Benakis, whose labours on this edition extend back to a Cologne dissertation of 1960 (vii) and comprise several Vorstudien (12 n. 19), has made accessible one of the less justifiable omissions from that earlier series. He has ensured that important evidence for a philosophical and educational culture in which the Physics played a continuing role (9-10 and 61-63) can now be assessed for its intrinsic and historical merits.

The Introduction deals primarily with the work's authorship and its relation to Psellus' other writings (3-25), as well as its general character and major themes (26-46; including a case study of topics in Physics Book 1 at 29-40). Then an inventory of fourteen manuscripts (47-59) (six have facsimiles at 443-48) is followed immediately, and in that regard a little unexpectedly, by the stemma, and by further observations on the Latin translation and on the work's predictably limited influence (60-64). The bibliography (65-94) is extensive and analytical. The text (for which book- but regrettably not chapter-numbers are provided in headers) has an upper apparatus for both explicit and implicit references,⁴ and an apparatus criticus that seems to record all important variant readings. The choice between variants in the principal manuscripts is generally straightforward and, since both the main text and the lemmata can be emended from the Aristotelian text, the editor has to intervene relatively rarely.⁵ Within exegetical passages the wording of the Aristotelian text is signalled in expanded type both when quoted, as it frequently is, and when integrated into a paraphrase, where it can, however, sometimes be missed: e.g., 155.22-23 (208a15-16) and 375.24-25 (225a30). There are indices only for names and loci. An index verborum would have been helpful. A list of instances of sullogismos and sullogizesthai, for example, would have helped pinpoint formalisations of Aristotelian arguments (e.g., 4.13; 29.12; 164.20; 165.4) rather than leaving only a sub-set of them identifiable from citations of the Prior Analytics in the index locorum.

In this brief notice (and a major critical edition, especially a first edition, obviously merits a lengthy review) I can address only three important general topics: the authorship of the work, the status of its lemmata, and the genre of commentary represented here.

(1) On the authorship, Benakis (5-10; 49-50) defends the work's attribution to Psellus against Pantelis Golitsis' recent claim,⁶ made largely on the basis of evidence in the manuscripts, that it is by a later polymath, George Pachymeres (1242 - ca. 1310), the scribe of the earliest manuscript (Laur. 87.5). His counter-arguments, added when this edition was close to publication (5 n. 8), are three: (i) the content of the work complements material elsewhere in the Psellan corpus (5-9); (ii) Pachymeres' unedited paraphrase of the Physics depends on Psellus' (9-10) (Golitsis has to accept that Pachymeres engaged twice and in different ways with the Physics);⁷ and (iii) the commentary in Laur. 87.5 has errors of a kind associated with transcription, and also resembles one by Michael of Ephesus copied by Pachymeres in the margins of another manuscript (Vat. Gr. 261) (49-50). It is difficult to comment on a disagreement that crucially relies on manuscript evidence, but it seems safe to say that Psellus' authorship of this commentary has yet to be conclusively disproved.⁸

(2) On the lemmata, Golitsis 2007 (see note 6), 647-50, has rightly stressed that in ms. Laur. 87.5 these unusually brief quotations signal the page divisions of the Aristotelian text beside which Pachymeres copied (or, as Golitsis would argue, composed) the commentary. This situation is reflected in Benakis' edition, where the exegetical passages are of roughly equivalent length and the lemmata straddle unrelated material or function purely as guides: e.g., at 31.9, four words from a conclusion at 187b30 head an exegetical passage that applies to the following section, 187b30 - 188a6. I found this anomaly puzzling until I became acquainted with Golitsis' paper. Still, since this commentary follows the Aristotelian text closely (see below), lemmata do seem to be essential, and so perhaps earlier and better adjusted ones were replaced in the manuscripts with the inadequate substitutes derived from Laur. 87.5.

(3) As for this commentary's genre (31-33), the work is called "succinct" (suntomos) in its title, and it does include summaries of the source text, avoids digressions or polemics, and has few references to post-Aristotelian authorities (Cleomedes, Gregory Nazianzen, and Simplicius are exceptions)⁹ and no significant traces of Christianity (27-28) or of Byzantine culture (Anthemius, an architect of Hagia Sophia, serves at 83.10 to illustrate Aristotle's beloved "house-builder"). There are also just a few instances of original interpretive material (31-32 with nn. 50 and 51, and 37-40 on the methodology in Phys. 1.1), while the unusually large number of cross-references to other Aristotelian works are probably intended to serve didactic purposes rather than being indicative of any special emphasis on the unity of the Aristotelian corpus (pace Benakis, 37). What remains is a careful and fairly elementary explication de texte, roughly two to three times the length of its source texts, and Benakis (33) rightly distinguishes it from the kind of Aristotelian paraphrase associated with Themistius (4th century A.D.), whose more discursive and interventionist treatment of the Physics, which was also aimed at more advanced students, Psellus does not cite (43). Elsewhere, however, he does identify the usual Themistian procedure of adopting the persona of an Aristotle expressing himself more expansively,¹⁰ but in dealing with the Physics he maintains a consistently third-person stance towards the language and sequence of thought in the Aristotelian text. In this respect his commentary resembles the analyses that form the first parts of much lengthier comments on lemmata by Alexander, Philoponus or Simplicius. So in sum Psellus offers the restrained and reliable account to be expected from a mature scholar and experienced teacher faced with the chore of taking a pupil through a difficult text. With a few adjustments his commentary could probably serve beginners even today.¹¹ It is unlikely to be called on to fulfil that function, but its text, context and content are now, thanks to Professor Benakis' edition, finally available for further study.¹²

1.   See, for example, K. Ierodiakonou (ed.), Byzantine Philosophy and its Ancient Sources (Oxford, 2002), reviewed at BMCR 2002.10.37, though see also R.J. Hankinson at Notre Dame Philosophical Reviews 2003.05.08. On the efforts made since the 1960s to supplement the Commentaria in Aristotelem Graeca with Byzantine commentaries, see Benakis, vii-viii.
2.   See my review of the inaugural volume in this series, M. Share (ed.), Arethas of Caesarea's Scholia on Porphyry's "Isagôgê" and Aristotle's "Categories" (Athens, 1994) at BMCR 1995.12.14.
3.   A passage from a manuscript version of this edition, which I was kindly allowed to quote in my edition of Cleomedes, Caelestia (Leipzig, 1990), p. xxix (see Benakis, 6 n. 10), was the only part of this work that I had previously seen.
4.   This kind of apparatus can always be enlarged. For example, at 28.21-22 Anaxagoras' nickname Nous is noted and could have been referred to Diogenes Laertius 2.6, who quotes relevant verses by Timon (= H. Lloyd-Jones and P. Parsons (eds), Supplementum Hellenisticum, no. 798).
5.   This is not the place to examine Benakis' emendations, or the lack thereof, but I would suggest that at 38.9 if the comma after ἄν is removed the reading in the manuscripts that he emends can probably be retained, while at 191.26 ἐπεὶ might be emended to εἰ to complement the optative and create a clause that expands the elliptical conditional at Arist. 214b28.
6.   See his article "Un commentaire perpetuel de Georges Pachymère à la Physique d'Aristote faussement attributeé à Michael Psellos", Byzantinische Zeitschrift 100 (2007) 637-76, the thesis of which is presupposed in "Georges Pachymère comme didascle: essai pour une reconstitution de sa carrière et de son enseignement philosophique", Jahrbuch de Österreichischen Byzantinistik 58 (2008) 53-68 at 55, 57-8 and 66-7. In reporting only Golitsis' general view, I am not, I must stress, doing anything like justice to his arguments, and in particular to his analysis of the manuscript tradition as it bears on the question of authorship.
7.   Golitsis 2008 (see note 6 above), 58-60, speculates on how this duplication might have fitted into Pachymeres' curriculum.
8.   The article "Byzantine Philosophy" by K. Ierodiakonou and B. Bydén in the Stanford Encyclopedia of Philosophy has not yet taken account of Benakis' arguments: it currently (as of 8 September 2008) shows some sympathy for Golitsis' position.
9.   The influence of Simplicius need not, of course, be confined to explicit references (cf. the upper apparatus at 285.1-2, 286.21) but, that said, Psellus does not, here at least, seem to appropriate earlier exegetical material as slavishly as other Byzantine commentators: see, for example, P. L. Donini on Michael of Ephesus' use of Alexander of Aphrodisias at RFIC 96 (1968), 316-23.
10.   See K. Ierodiakonou, "Psellos' Paraphrase on Aristotle's De interpretatione", in Ierodiakonou (2002) (see note 1 above) 157-81 at 165 n. 31 for the description of paraphrasing by Psellus in his In De interpretatione, a work available only in the Aldine edition (first published 1503) but currently being re-edited (Benakis, 17 with n. 28). On Themistius' paraphrastic method see my discussion at Themistius on Aristotle "On the Soul" (London and Ithaca, 1996), 4-7, and for this commentator's rather different procedures in dealing with the Physics see the two volumes so far published of my translation of his paraphrase of that treatise: Themistius on "Physics" 4 and Themistius on "Physics" 5-8 in the Ancient Commentators on Aristotle series, London and Ithaca, 2003 and 2008 respectively.
11.   For example, see Psellus' 159-word response at 155.14 - 156.6 to the 55 words of Phys. 3.8, 208a14-19, where Aristotle rejects an Atomist (and later Epicurean) argument for the infinite that claims that we necessarily think of something existing beyond anything finite. The Byzantine commentator recreates and contextualizes this refutation rather more thoroughly than Themistius' brisker 149-word summary at In Phys. 100.23 - 101.5, and his explication stands up well against the 72-word summary (p. 370) and textual ruminations (ad 208a18) in W. D. Ross, Aristotle's Physics (Oxford, 1936).
12.   I noticed some errors: Introduction: p. 28: the reference should be to 430.16-21 not 430.19-24; text: 181.13: ἐν has obviously dropped out before τόπῳ where it is needed to complement οὐκ and thereby reflect οὔ που in the lemma at 212b14 (cf. also 212b9 and 22, and Psellus 180.10); 237.8: delete the rho in προσαχῶς; upper apparatus: obsolete editions are cited at 25.6 for Timaeus Locrus (G. Hermann 1852 rather than W. Marg [Leiden, 1972]), and at 27.12 for Proclus' commentary on the Platonic Alcibiades I (F. Creuzer 1820 rather than A. Segonds' Budé edition [Paris, 1985 and 1986]); 58.13-14 should be 58.14-15; 240.26: supply the reference to "in A3, 187a2" as 24.1-2 (Benakis too often inconveniently cites the text of the commentary by the Aristotelian reference instead of the page and line number of his edition). In the apparatus criticus ad 104.2-3: two entries are conflated; ad 153.16-17, the omission recorded there is not attributed to any manuscripts.