Paradoxically enough, and perhaps in the same way that attending a tragic play is supposed to be both disturbing and delightful, Aristotle's Poetics has been a fascinating, yet nightmarish, affair for editors – and editors have been exceptionally numerous (and translators too). It is not as if the material at our disposal were too scanty, with four main witnesses (two Greek MSS, as well as Moerbeke's Latin translation and an Arabic translation from a Syriac one). But these witnesses offer quite different readings at some (sometimes crucial) points, and there are also (evident or possible) mistakes even in cases when all of them concur on their readings. From the Renaissance on, scholars have gone to great pains to offer an unusually high number of (more or less plausible) emendations – but this has in a way caused even more trouble for editors trying to recreate what Aristotle might have written.

When R. Kassel's OCT edition came out in 1965, it was widely heralded as being the most careful and balanced edition of the text, with the best and most complete apparatus that had ever been produced; and not without a sort of relief among puzzled readers who were at a loss between various, and variously flawed, editions, it was immediately considered the standard. But even a standard edition of such a difficult and disputed text can hardly be trusted in all of its choices, and indeed no subsequent translator has followed Kassel's text blindly. It is, thus, not a complete surprise to find yet another attempt to try to cope afresh with the text of the Poetics.

As the title indicates, this is an editio maior – first of all in length (538 pages), as it comprises not only a long introduction to the history of the text by T. and to the Arabic translation by G. before the edition of the text itself, but also two substantial sets of notes after it, one by T., wherein he justifies his choices for the text he has edited, and one by G. on the Arabic readings. Perhaps the most innovative part of the book is G.'s enterprise, which explains in minute detail (but very clearly, so that non-Arabic readers too can follow) all the issues linked to the Arabic which had remained largely dependent on Tkatsch's sometimes unsatisfactory edition. But for Greek classicists, their best bet will be T.'s apparatus which, thanks to a new recollectio, reports all the lectiones one can find in our four primary sources, being thus an editio maior in this specialised sense. Especially in the (in fact, relatively few) cases of imprecisions or mistakes due to Kassel (and Tkatsch), T. and G. have accomplished a very useful task, and every future translator or commentator of the Poetics will benefit immensely from it.

If this apparatus seems to be more reliable and more comprehensive as to the primary sources, one might wonder about the policy T. adopts as to the modern emendations: they are only mentioned in cases where T. believes they need to be integrated into the text he has chosen to print. This is, in principle, what editors do for brevity's sake, but it is obviously a policy that needs to be adapted somewhat given the difficulty of the text – and indeed, Kassel and other editors provide at least a sample of emendations for the most contentious passages, even in cases where they accept the MSS text. Admittedly, T. does from time to time report and discuss some such emendations in his notes. But that leaves many cases where he does not. For instance, there is 1456a20 which is not even discussed in the notes, where many scholars have preferred Castelvetro's most plausible emendation τῷ θαυμαστῷ or Tyrwhitt's θαυμαστῶν instead of the rather bizarre θαυμαστῶς (in fact, a hapax in Aristotle). T. believes, we may suppose, that the received text should not cause any philological or interpretative trouble, but given that this passage has been one famous conundrum in the literature, I cannot help but find his silence here (and in other similar places) rather awkward, almost cavalier, for an editio maior.

What about the text itself? Two features struck me: in many cases, but without saying so explicitly, T. does not hesitate to hearken back to Bywater's 1909 edition and – which is less surprising given his working in tandem with G. – he has adopted several readings from the Arabic that Kassel did not.

On the positive side, T. is right to restore some of Bywater's readings contra Kassel, especially when the latter ends up with hopeless cruces while there are likely readings available thanks to one light emendation: see for example 1448a23–4 where T. endorses Casaubonus' emendation πάντα instead of the MSS πάντας that put τοὺς μιμουμένους in jeopardy – which Kassel logically put into cruces. As to readings from the Arabic, T. accepts (in fact, after Gudeman, followed by some translators) μᾶλλον ἢ instead of ἢ in the famous sentence διὸ εὐφυοῦς ἡ ποιητική ἐστιν μᾶλλον ἢ μανικοῦ at 1455a32–3 – a both philologically and philosophically better choice (and curiously absent from Kassel's apparatus). Another such case is at 1449a19 where T. prints ἔτι δὲ τὸ μέγεθος ἐκ μικρῶν μύθων, καὶ < ἡ λέξις ἐκ > λέξεως γελοίας διὰ τὸ ἐκ σατυρικοῦ μεταβαλεῖν ὀψὲ ἀπεσεμνύνθη (instead of Kassel ἔτι δὲ τὸ μέγεθος· ἐκ μικρῶν μύθων καὶ λέξεως γελοίας κτλ): < ἡ λέξις ἐκ > was one conjecture made by Christ (and duly reported by Kassel) which is now at least half confirmed by the Arabic (which reads ἡ *** ἐκ), making λέξις the subject of ἀπεσεμνύνθη, while on the Greek MSS reading, μέγεθος is supposed to mean ‘grandeur’ in the metaphorical sense, which is unlikely given what Aristotle argues for in P. 7. In such cases, there is perhaps nothing totally new for seasoned readers, but these (re)endorsements are most welcome, and they have gained in plausibility thanks to T.'s excellent notes ad loc.

It is a pity, though, that in several places, T.'s return to Bywater seems to have been much less fortunate. One very infelicitous choice is at 1448b13 where he reprints αἴτιον δὲ καὶ τοῦτο (instead of Kassel and most editors' αἴτιον δὲ καὶ τούτου): philologically, this may appear the most authoritative reading (τούου only comes from Moerbeke's translation, which Bywater could not have known as it was discovered after his edition), but it would be very strange were Aristotle to have confused a simple σημεῖον with a forceful αἴτιον, as one must suppose he did if he wrote τοῦτο – in fact, I am afraid to add, it is the note where T. interprets the whole, famous passage on the natural causes of poetry which is strangely confused (see pp. 239–40). In addition, there is at least one case where Kassel offered a completely new and brilliant reading which T. has not followed, reverting to the MSS text which was printed by Bywater too: at 1454a22, Kassel wrote ἔστιν γὰρ ἀνδρείαν μὲν τὸ ἦθος, ἀλλ' οὐχ ἁρμόττον γυναικὶ οὕτως ἀνδρείαν ἢ δεινὴν εἶναι, instead of ἔστιν γὰρ ἀνδρείον μὲν τὸ ἦθος which would imply, quite implausibly, that ἦθος means ‘personage’ here rather than ‘moral character’, and also that women were not supposed to be courageous (which flatly contradicts Pol. 1260a20–4 and 1277b20–5). On the other hand, there are also places where endorsing the Arabic could prove disputable: at 1455b31, αἰτίασις might have been read (see the cautious note by G., pp. 405–7) in the sequence δέσις μὲν τά τε προπεπραγμένα καὶ ἡ τοῦ παιδίου λῆψις καὶ πάλιν ἡ αὐτῶν αἰτίασις [the Greek MSS have δὴ; Kassel prints **], λύσις δ' ἡ ἀπὸ τῆς αἰτιάσεως τοῦ θανάτου, but in the context, this bare repetition seems to be less probable than the emendation proposed by Gomperz, δήλωσις (= the disclosure of the parents).

This new edition is an important work that committed readers of the Poetics must have on their desk together with their copy of Kassel, but as has been the case with the latter, as standard as it is, this is an edition that should not be read at face value.