Reading ooofia at 82d3, as Verdenius suggests. Cf. Timaeus 88c3.
Hackforth (92, n.l) translates (82e5-83al): 'the way in which the prisoner would be most likely to cooperate in his own incarceration is through his desire', taking the cjc clause as subject of 'eoTtv. This is very awkward and takes paXiara out of order. Verdenius would take eipjfxov to mean 'imprisonment'. However, it cannot mean that at 82e3, and if it is taken in that way at 82e5 there will be an awkward shift of sense.
Bracketing nai <po@cov (83b7) with Burnet, but retaining 7) Xvnrjdri (83b8—9) with Verdenius and Loriaux.
Omitting <j>aow (83e6) with the Arsinoe papyrus for the reasons given by Hackforth. With (j>aow the meaning will be: 'it is not for the reasons given by most people'.
The translation at 84a5—6 follows Burnet's text and Loriaux's explanation, eraw&o? is taken to mean 'in reverse' (to Penelope) rather than 'undoing her web' (Hackforth, Bluck). Penelope unpicked at night what she had woven during the day. The soul is thought of as reweaving at night, through sensual indulgence, the 'web' (rtw ... Iotov) that philosophy has unravelled during the day.
Hackforth translates 84c4 'were having a few words'. See, however, Lysis 211a and other passages cited by W. L. Lorimer, C.R. 1960, 7—8, supporting 'were talking in a low voice' (Burnet, Bluck).
Understanding the subject of Xex#i?wu (84dl) as ra \ex6evra from 84c5. This seems preferable to 'if you think it would be better that it should be stated' (Bluck).
The translation follows Burnet's text at 88a2, which involves taking 'you' here to refer to Cebes himself. He is voicing the argument of an imaginary objector, who is prepared to grant 'the speaker', i.e. the proponent of immortality, 'even more than what you say': even more, that is, than Cebes has himself conceded (87al—4)—that the soul's prenatal existence has been adequately proved. It is, however, difficult to take av to refer to Cebes himself, when the objection has been presented in oratio obliqua;
and it is awkward to understand raj Xeyom as meaning, without further explanation, the proponent of immortality. Some editors therefore bracket f} at 88a2: 'If one were to grant even more to someone who says what you (sc. Socrates) say'. On either reading, however, Cebes will be making the same concession to the same viewpoint. See Bluck, 157—9 for further difficulties. The passage may, as Hackforth suggests, be incapable of strict grammatical analysis.
Or perhaps (89b3—4) 'it was his way sometimes to play with my hair' (Bluck). If nafteiv means 'make fun of, Socrates may, as Robin suggests, be making fun of Phaedo for wearing his hair long, though he was past the age at which it was customary at Athens to do so.
\&)Kov r) yeXava (90a7) is usually translated 'white or black'. The words can, however, mean 'pale' and 'dark' (cf., e.gRepublic 474el—2), and this seems better suited to the notion of a range.
roof bvroiv (90d6) has been taken as if governed by the whole phrase ttj? aXrjtfere «aЈ emcm^T}*;. If it is governed only by rrj? aXrjfieta? (cf.99e6), the meaning will be 'deprived of the truth of the things that are and of knowledge'. For tlqp outoov see on 65c2—4.
Reading clvqwl at 9lb5 with Burnet, and translating 'ignorance' rather than 'folly' (Bluck). Socrates' ignorance will consist in his mistaken belief in the afterlife. With the reading 5iavota the meaning will be 'this opinion of mine will not persist'.
el kotos (92dl) has been translated 'likelihood' rather than 'analogy' (Burnet). Burnet cites Theaetetus 162e5 and Euthydemus 305el,but as Hackforth says, those passages tell strongly in favour of 'likelihood'.
L. Lorimer's proposal (C.R. 1938, 165-6) to read (92d2) 8 ok el <h SoKei> iwQpionovs would give the sense 'whence most people derive the opinions they hold'. However, what is at issue is not most people's opinions in general, but only their belief in the attunement hypothesis. For a convincing defence of Burnet's text, see J. Tate, C.R. 1939,2-3.
Placing a colon at the end of 92d9, and keeping the MSS. aurfj? with Burnet. 'Being' belongs to the soul, is 'of it' (aurrj<r), in the sense that it is the object of the soul's apprehension. Cf.76el, 'finding again what was formerly ours', and 75e5, 'knowledge belonging to us'. If, however, Mudge's aim? is adopted, the meaning will be: 'just as surely as the Being itself exists'. The soul's pre-existence will then be related only to the existence of the Forms, and not to its apprehension of them. Cf.76e2—4. For ova (a see note 7. o eanv has been translated 'what it is' in conformity with its interpretation elsewhere. See on 75c7—d6 (p. 130), and notes 21, 24-6, 28, and 31.
The difference, if any, at 93al4—b2 between iiaXhov and em nXeou, and between t^ttov and en' eXarrov is uncertain. Hackforth thinks there is no difference. Olympiodorus took the first member of each pair to refer to pitch and the second to intervals (see Burnet's note on 93al4). Verdenius plausibly takes em irXeov to mean 'extending over a greater part', i.e. covering a greater number of strings.
Verdenius says that the emep clause (93bl) should be taken with eni nXeov only, and expresses the fact that 'em irXeov (in the sense explained in note 51) is a rather theoretical case, since in practice all strings will have been tuned. But it is hard to understand how an attunement, as opposed to an instrument, could be tuned em rrXeov in this sense.
The translation follows Burnet's text at 93d4, but not his version of 93d2—3, 'this is just our admission'. The clause p.rjSev ... eivai (d3—4) has been taken as dependent upon to bptoX&yriiia: 'this is the admission that...' Hackforth (115-16, n.4) and others, finding contradiction between these lines and 93al4—b6, have bracketed apfioviac; in 93d4, taking apfxoviav in the same line as complement, and erepav ... erepae to refer to one soul as compared with another: 'one soul cannot be more or to a greater extent, or again less or to a smaller extent, an attunement than another'. With this reading, t-qp Se (93d6) and rj 5e (93d9) will refer to the soul and not to attunement, as they do in the translation adopted. However, there is no MS. support for the change, and it destroys what looks like a deliberate parallelism of language at 93d2 and 94d4. It also entails taking 93d6—7 in such a way as to anticipate the reasoning of 93d 12—e2.
Or perhaps (94b4—5) 'Is it a man's soul that controls every part of him?' (Hackforth).
(pQopa (95e9, 96b9) has been translated 'destruction', and oXedpos (e.g., 95d2) 'perishing'. The cognate verbs have been translated correspondingly throughout. There seems no significant difference between them, both being used as general terms for 'ceasing to exist'. At 106e aS idcpOopoq will be used as a variant for ai>co\e0po<r.