¹ Cooper, J.M., ‘Conclusion – and retrospect’, in Steel, C. (ed.), Aristotle's Metaphysics Alpha (Oxford, 2012), 335–64Google Scholar provides the most thorough analysis of A 10 available.

² On the transmission of the text, see Harlfinger, D., ‘Zur Überlieferungsgeschichte der Metaphysik’, in Aubenque, P. (ed.), Études sur la Métaphysique d'Aristote. Actes du VIe Symposium Aristotelicum (Paris, 1979), 7–33Google Scholar and Primavesi, O., ‘Introduction: the transmission of the text and the riddle of the two versions’, in Steel, C. (ed.), Aristotle's Metaphysics Alpha (Oxford, 2012), 388–464Google Scholar.

³ See the text in the recent edition of Metaphysics Book A: O. Primavesi, ‘Text of Metaphysics A’, in C. Steel (ed.), Aristotle's Metaphysics Alpha (Oxford, 2012), 465–516, at 515.

⁴ Cooper (n. 1), 336 translates ‘all [these people]’ and specifies ‘i.e. the people whose views on the causes of being are critically discussed in the immediately preceding chs. 8 and 9 … together with others, as discussed in chs. 3–6.’

⁵ The words ‘also … said before’ make clear that Aristotle here refers to his discussion in chapters 3–6, with the summary in chapter 7. See Cooper (n. 1), 337–9 and below.

⁶ Ross, W.D. (ed.), Aristotle's Metaphysics (Oxford, 1924), 212Google Scholar: ‘omitted by Al. … probably a gloss on καὶ κατ’ ἀρχὰς’.

⁷ Jaeger takes καὶ τὸ πρῶτον to be a variant reading to the preceding expression κατ’ ἀρχάς. See W. Jaeger (ed.), Aristotelis Metaphysica (Oxford, 1957), apparatus criticus on 993a16.

⁸ Diels, H., ‘Aristotelica’, Hermes 40 (1905), 301–16, at 303 n. 1Google Scholar.

⁹ Alex. Aphr. In Metaph. 63.31 Hayduck. On the relation of Alexander's commentary to the Metaphysics text, see Kotwick, M.E., Alexander of Aphrodisias and the Text of Aristotle's Metaphysics (Berkeley, 2016)Google Scholar.

¹⁰ H. Bonitz, Index Aristotelicum (Berlin, 1853), s.v. εἰπεῖν (pages 221b–222a) indicates that Aristotle uses the perfect passive forms of εἰπεῖν (i.e. εἴρημαι) standardly and predominantly to refer back to his own exposition.

¹¹ There is only one parallel in the corpus (Rh. 1419b27; see next note), where πρότερον is used within a ten-word proximity to εἴρηνται. And even there πρότερον refers to what was stated earlier in Aristotle's text.

¹² These are: An. pr. 46a11 (αἱ δ’ ἀρχαὶ τῶν συλλογισμῶν καθόλου μὲν εἴρηνται, ‘the principles of deductions have been stated in general terms’), Aristotle is the agent; De an. 403a2 (διαλεκτικῶς εἴρηνται καὶ κενῶς ἅπαντες, ‘they [i.e. the formulas] have been spoken only for the sake of the argument and vacuously’), the agent is Aristotle; Eth. Nic. 1117a28 (οἵ τε δὴ ἀνδρεῖοι εἴρηνται, ‘the brave have been described’), the agent is Aristotle; Eth. Nic. 1127a18–19 (οἱ μὲν πρὸς ἡδονὴν καὶ λύπην ὁμιλοῦντες εἴρηνται, ‘after those who are in company with people for pleasure and pain have been described’), the agent is Aristotle; Rh. 1419b18 (εἴρηνται οἱ τόποι, ‘the commonplaces have been stated’), the agent is Aristotle; Rh. 1419b27 (εἴρηνται δὲ καὶ τούτων οἱ τόποι πρότερον, ‘the commonplaces of these too have been stated before’), the agent is Aristotle.

¹³ In Rh. 1371b16 (ὅθεν καὶ αἱ παροιμίαι εἴρηνται, ‘hence the proverbs are said’) the agent is any Greek speaker.

¹⁴ Kotwick (n. 9), 99–177.

¹⁵ The alternative reading is not mentioned in the edition by Bekker (1831), nor in those by Schwegler (1847), Bonitz (1848) (which is especially surprising, given that Bonitz edited Alexander's commentary before he edited Aristotle's Metaphysics and had an intimate knowledge of Alexander's text), Christ (1895), Ross (1924), Jaeger (1957) or Primavesi (2012).

¹⁶ However, regarding the difference in line 993a16–17 (καὶ τὸ πρῶτον), which I have mentioned above, Jaeger and Ross made use of that passage.

¹⁷ Alex. Aphr. In Metaph. 63.26–31 Hayduck: ἐπὶ τέλει τοῦ βιβλίου τούτου αὐτὸς οὕτως λέγει ‘ὅτι μὲν οὖν τὰς εἰρημένας ἐν τοῖς Φυσικοῖς αἰτίας ζητεῖν ἐοίκασι πάντες, καὶ τούτων ἐκτὸς οὐδεμίαν ἔχομεν εἰπεῖν, δῆλον καὶ ἐκ τῶν πρότερον εἰρημένων· ἀλλ’ ἀμυδρῶς ταύτας, καὶ τρόπον μέν τινα ἁπάσας, ὡς καὶ πρότερον εἴρηται, τρόπον δέ τινα οὐδαμῶς· ψελλιζομένῃ γὰρ ἔοικεν ἡ πρώτη φιλοσοφία περὶ πάντων, ἅτε νέα καὶ κατ’ ἀρχὰς οὖσα.’ I have checked online (in the digital repository of the Biblioteca Medicea Laurenziana) MS Laurentianus Plut. 85.1 (= O), which is a crucial witness to Alexander's commentary that Hayduck did not consult; it confirms Hayduck's text.

¹⁸ The combination of πρότερον and εἴρηται (within six words) appears 290 times in Aristotle's corpus. Here are some examples of the combination introduced by a comparison: Eth. Nic. 1135a23 ὥσπερ καὶ πρότερον εἴρηται, Eth. Nic. 1177a11 καθάπερ καὶ πρότερον εἴρηται, Hist. an. 535a5 ὥσπερ καὶ πρότερον εἴρηται, Poet. 1450b13 ὥσπερ πρότερον εἴρηται, Top. 139a32 καθάπερ καὶ πρότερον εἴρηται.

¹⁹ See Cooper (n. 1), 337–40 for an extensive discussion of the implication of ‘also’ (καί).

²⁰ These lines follow seamlessly the lines to which the first back-reference in chapter 10 (line 993a13) refers.

²¹ See Primavesi, O., ‘Second thoughts on some Presocratics’, in Steel, C. (ed.), Aristotle's Metaphysics Alpha (Oxford, 2012), 225–63, at 226–7Google Scholar.

²² Jaeger, W., Studien zur Entstehungsgeschichte der Metaphysik des Aristoteles (Berlin, 1912), 14–21Google Scholar; cf. also Kotwick, M.E., ‘Die Entwicklungsgeschichte of a text: on Werner Jaeger's edition of Aristotle's Metaphysics’, in King, C.G. and Presti, R. Lo (edd.), Werner Jaeger: Wissenschaft, Bildung, Politik (Berlin and Boston, 2017), 171–208, at 175–7Google Scholar.

²³ Cf. Cooper (n. 1), 339, who also defends the status of A 10 against Jaeger, but does not build his argument on the fact that A 10 contains an explicit correction of A 7, something that becomes visible only through the new reading I propose here.