Alexander of Aphrodisias and the Text of Aristotle’s Metaphysics by Mirjam E. Kotwick (review)

This is not a book for the ordinary historian of philosophy. It consists almost exclusively of detailed analyses of the manuscript readings at a few scores of places in Metaphysics A–Δ and Λ, confronting the transmitted readings each time with Alexander of Aphrodisias’s comments on the relevant passage. The reason why only those books are studied is simple: Alexander’s commentary (produced about AD 200) on books E–N was lost before the end of the Byzantine era, but Averroes preserved information about the contents of an Arabic translation of the commentary on book Λ.

If you are brazen-bowelled (χαλκέντερος) enough to stomach so much philological detail, this is a rich book, which teaches the reader a lot about the complexities of the tradition of a text as complex as the Metaphysics. Kotwick is very methodical, very good at analyzing both the linguistic and the philosophical consequences of chosing one or another variant at a particular place, and she makes it perfectly clear which theses she argues for and believes to have corroborated. They are notably five:

1. 1. Alexander used only one copy of the Metaphysics, ω^AL. His occasional mentions of variant readings do not reflect an independent collation of manuscripts. The lemmata, which introduce the separate scholia that make up his commentary, were always part of the work, and—together with the quotations and paraphrases that occur in his interpretation of the text—the lemmata can be used to reconstruct ω^AL.

2. 2. ω^AL was independent of ω^αβ, the archetype from which the two Byzantine ms. branches, α and β, are descended.

3. 3. ω^αβ was produced some time between AD 250 and 400. It postdates Alexander, because some of its readings must have come from his commentary.

4. 4. ω^αβ had an ancestor, some of whose erroneous readings were known to Alexander, quite possibly thanks to information he found in Aspasius’s commentary.

5. 5. Readings from Alexander’s commentary were introduced both in the α and, more massively, in the β branch after their separation.

On the whole, I find Kotwick’s argumentation convincing, though I do have some worries.

Thus, re (1–2), the basis for reconstructing ω^AL is not quite as strong as she would like it to be. There is an unpleasant number of disagreements between the Aristotelian text of Alexander’s lemmata and that of his quotations and paraphrases. Kotwick tries to downplay this inconvenient fact by saying, for example, that “Of all 296 lemmata . . . only 28 (about 10%) display a reading that visibly disagrees with” that of a quotation or paraphrase (49). I do not think her ‘only’ is warranted. Fortunately, in her examination of the single variant places she is very careful not to jump to conclusions about what ω^AL must have read.

Re (3) and (5) I do not find all alleged instances of influence from the commentary on the text equally convincing, but her hypotheses are prima facie probable. An authoritative commentary like Alexander’s is likely to have influenced all branches of the text from shortly after its appearance.

A minor detail: Kotwick deals with the question whether the Ps.-Alexander responsible for the commentary on books E–N transmitted together with Alexander’s on A–Δ is Michael of Ephesus (early twelfth century, 22–23). She omits one piece of evidence, which to my mind settles the question: the commentary known as Ps.-Philoponus, which is beyond reasonable doubt due to Georgios Pachymeres (1217–1282), contains a reference to the ‘Ephesian’ that can be matched in Ps.-Alexander on book VI (E). I see no reason to distrust Pachymeres’s attribution. Paul Moraux had drawn attention to the reference in 1942; I did so again in my Commentators and Commentaries on Aristotle’s Sophistical Refutations, 1981. For the identification of Ps.-Philoponus with Pachymeres, see Pantelis Golitsis, Jahrbuch der Österreichischen Byzantinistik 58 (2007).

We may still have to wait some time for a new Metaphysics to replace the editions by W...