¹ This paper emerged from work done in preparation for the New York Colloquium on Ancient Philosophy in the autumn of 2011. I thank Iakovos Vasiliou, NYCAP's organizer, for inviting me to chair a session, and the colloquium's participants for their interest and enthusiasm. Special thanks to Brad Inwood and Peter Simpson for vigorously resisting my proposal. Charles Brittain helped at several stages of its genesis. And as always, my deepest thanks go to Liz Karns.

² Meineke, A. (ed.), Ioannis Stobaei Eclogarum Physicarum et Ethicarum Libri Duo, Tomus II (Leipzig, 1864)Google Scholar, CLXIII ad p.28 line 22: ‘Sed totum de hoc triplici Stoicorum bono et malo locum nemo plenius illustrat quam Schol. Luciani p. 209. Iac. (vol. VII. p. 338 Lehm.), qui dubitari non potest quin sua ex antiquo fonte vel e pleniore Stobaeo, quam nunc habemus, derivaverit’ (‘But on the entire topic of the Stoics’ three-fold account of good and evil, no one sheds more light than the Schol. Luciani, who undoubtedly drew his material from an ancient source, or from a [version of] Stobaeus that was fuller than [the version] that we now have').

³ Wachsmuth, C. (ed.), Ioannis Stobaei Anthologii Librum Alterum (Berlin, 1884 )Google Scholar XXII: ‘magistellus Byzantinus … qui plerumque ne sententiam quidem Stoicorum assecutus sit’.

⁴ Hahm, D.E., ‘The ethical doxography of Arius Didymus’, ANRW 2.36.4 (Berlin, 1990), 2935–3055.Google Scholar

⁵ ‘Having made an adequate examination of goods and evils [2.7.5a–g] and things to be chosen and avoided [2.7.5h–o] and concerning the end and happiness [2.7.6a–f], we next thought it necessary to go over indifferents, too, in their proper place. Now they say that indifferents …’ (Stob. 2.7.7.1–5). The order of the discussions of goods and evils and things to be chosen and avoided seem to have been disrupted and confused in our texts of Stobaeus, and it is not clear to me that Wachsmuth made it better by the radical relocation of a long section – his ‘5b1–5b13’ – and his deletion of a heading at 5h which promises a section ‘On things to be chosen and things to be avoided’. One can have no faith in the current order of the paragraphs, either in the MSS, or in Wachsmuth's edition. But that the contents of 2.7.5 were a discussion of those two contrary pairs (sc. goods and evils, and things to be chosen and avoided) is clear, no matter how the internal contents are best arranged.

⁶ The βίος σπουδαῖος is a σύστημά τι λογικῶν πράξεων, SVF 3.293.10.

⁷ Pomeroy, Compare A.J., (ed.), Arius Didymus. Epitome of Stoic Ethics (Atlanta, GA, 1999)Google Scholar. This problem was not noticed by Pomeroy (p. 37), because he mistranslates the Greek: ‘… they also say that every virtue which is associated with man and the happy life is consistent with and in agreement with nature’. But εὐδαιμονίαν is not an adjective; ζωήν is the predicate, not part of a compound subject; and ἀκόλουθον and ὁμολογουμένην are attributes of ζωήν, not predicates of ἀρετήν and ζωήν. At least Pomeroy's English has the advantage of saying something true about virtue, sc. that it is consistent and in agreement with nature, instead of saying something false, as the Greek does.

⁸ Diog. Laert. 7.87 = LS 63C.

⁹ Stob. 2.7.5b5 = LS 61D1.

¹⁰ Τήν τ’ ἀρετὴν διάθεσιν εἶναι ὁμολογουμένην (Diog. Laert. 7.89 = LS 61A). Their translation of 61A2 as ‘Happiness consists in virtue …’ is misleading. The Greek (ἐν αὐτῇ τ’ εἶναι τὴν εὐδαιμονίαν) can simply mean e.g. ‘happiness depends on virtue’, and does not require confusing happiness as a diachronic process with virtue as a disposition.

¹¹ Our soul is an animal, so virtues are animals, since they are the same as our reason: Stob. 2.7.5b7 = 2.65. But our soul is not a life, either, in the sense in which happiness is a kind of life (our soul may be said to be a principle of life, but it is not an event or temporally extended sequence of events).

¹² Τῶν δὲ <ἀγαθῶν>ἐν σχέσει τὰ μὲν καὶ ἐν ἕξει εἶναι, οἷον τὰς ἀρετάς (Stob. 2.7.5k = 2.73 = LS 60 J).

¹³ διάθεσίν τινα καὶ δύναμιν (Plut. De virtute morali 441C1 = LS 61B8).

¹⁴ The particle μέν may not have been in the original text. It appears in the scholiast's text because he has run together several sentences from paragraphs corresponding to Stob. §§5b7, 6 and 6a, which he then coordinates with the μέν and δέ. Without the μέν, the line endings would look even more similar to the copyist who gave us our current text of Stobaeus, i.e. simply καὶ τὴν εὐδαιμονίαν at the end of each line.

¹⁵ Charles Brittain observed in correspondence that the argument is reminiscent of that in Diog. Laert. 7.85–9.

¹⁶ Meineke (n. 2), CLXX ad p. 38.29, quoted in n. 2.

¹⁷ Wachsmuth (n. 3), vol. II.75 in the apparatus quotes the material from the scholiast and then says ‘quod cave ex pleniore Stobaeo haustum esse putes (vide proleg.)’.

¹⁸ Wachsmuth (n. 3), vol. I.XXII. Some of his allegations misfire. For instance, among the ‘telic goods’ that Stobaeus lists are ‘prudent walking’, ϕρονίμη περιπάτησις. The scholiast also lists this among telic goods. But then among telic evils, the scholiast lists, as its counterpart, ἀϕρονικὴ περιπάτησις καὶ πλημμελὴς κίνησις (l. 38). Wachsmuth erupts: ‘audin hominem ineptum qui ϕρονίμην περιπάτησιν ad philosophorum disputationem spectare ignoravit et ei contrarium aliquid inter mala opponendum esse ratus “dementem ambulationem” et “improbam motionem” adscripsit?’ (‘Listen to this bungler, who didn't realize that “phronimê peripatêsis” refers to the discussions of the philosophers [i.e. who are walking], and who added “mad strolling” and “wicked motion,” to the [list of] bads, because he thought there should be something contrary to it!’).

The problem here is that the majority of contemporary scholars agree with the scholiast against Wachsmuth, that ‘prudent walking’ does not mean ‘the discussions of philosophers’, but rather, simply, ‘prudent walking’, i.e. doing a trivial and otherwise indifferent thing, walking, in a way that only a Sage can do it. That the example conforms to orthodoxy does not show that it is a faithful transcription from fuller doxography, as Meineke thought. But it also does not show the man's ineptitude, as Wachsmuth thought.

¹⁹ I set aside the very frequent differences of expression when they do not change the content – the reader will see a selection of them in the discussion below. They are interesting for independent reasons, as showing the scholiast's willingness to transcribe ad sensum rather than ad litteram. But they do not count as new content that the scholiast could either be right or wrong about.

²⁰ The following sample is roughly 16 lines out of the total 113 lines, or about one seventh of the total, and it is representative of the scholiast's treatment of the material.

²¹ See Appendix, Note A.

²² See Appendix, Note B.

²³ See Appendix, Note C.

²⁴ More substantive are the several places at which Stobaeus presents contrasting categories of ἀγαθά and κακά (i.e. evils), which the scholiast then reports as ἀγαθά and κακίαι (i.e. vices) (29.22.27, 29.22.35). He has been misled by the fact that the corresponding goods have included virtues.

²⁵ To see a clear exposition of the real Stoic view, with expanded examples that are error-free and illuminating, see Sext. Emp. Math. 11.22–6. He explains that the first sense of ‘good’ applies only to virtue itself, the second applies to virtuous actions (i.e. actions done κατ’ ἀρετήν), and the third sense applies to the items referred to by the first two senses (sc. virtues and virtuous actions) and adds to them such other things as virtuous people, virtuous friends and the gods themselves, all of which are ‘such as to benefit’. Similarly, in Diog. Laert. 7.94, the sense of ‘good’ in which it is said καθ’ ὅ is as follows: a πρᾶξις is called ‘good’ inasmuch as it is done in accordance with (κατά) something which is good in the first way, sc. virtue.

²⁶ Hahm (n. 4) spends several pages (2955–62) demonstrating that it is not impossible to illustrate Stoic distinctions about the good, using analogies drawn from heat. This is so. But he does nothing to explain how these particular examples are meant to work, nor does he investigate whether they reflect an understanding of Stoicism.

²⁷ Hahm (n. 4), 2950–1.

²⁸ We may add the further indictment that the scholiast seems to think that the Stoics would treat falling ill as a kind of being harmed, and illness as an evil, if we are right about how the heating and sleeping examples are supposed to be read. That would be such a fundamental error about Stoic ethics that the scholiast would thereby earn Wachsmuth's ire as someone ‘qui ne sententiam quidem Stoicorum assecutus sit’.