1. This paper, in its present form, is for the most part an annotated version of the one originally delivered at Brown University in January 1991; the notes include an indication of my responses to points raised on that occasion (particu- larly by my respondent, Professor Thomas Tuozzo), by audiences later in the year at the Universities of Amsterdam and Perugia, and by an anonymous com- mentator for the Proceedings. 2. That, for example, all the arguments somehow contribute to a single, com- plex conclusion (however we may define the "arguments": see below), or that the arguments and the dramatic action are connected, cannot reasonably be denied; even if this leaves out two significant elements, i.e. Socrates' "autobiog- raphy", and the "myth", the Phaedo already thereby possesses a rather higher degree of overt structural organisation than many or most other Platonic dia- logues. 3. My own primary concern will be to examine the relationships between the arguments, and between the characters who propose and respond to them, from which there will emerge a specific explanation of the combination in the dia- logue of argument and action. About Socrates' autobiography, and the "myth", I shall have little or nothing to say for reasons which will become clear; in any case, they deserve full treatment in their own right, for which I have no space in this paper, and whatever view is taken of them is unlikely to affect my conclu- sions. 4. As this already suggests, I shall be offering a highly determinate reading of the dialogue. In a preliminary seminar paper at Brown, I presented an argument against those who, for whatever reason, are inclined to deny the existence of any
enforceable limits on the interpretation of (Platonic) texts. I do not of course claim that my own interpretation of the Phaedo exhausts its meaning, or that it is invulnerable either to general arguments of a skeptical nature or to specific objections which I have failed to foresee or answer. I do claim that if the relevant parts of the dialogue are read in the way I shall suggest, they represent an inter- nally coherent and consistent ordering of elements; and that if it is at all legiti- mate to prefer one interpretation, of any work, over any other, the discovery of such an ordering is one plausible basis for making the choice. While any work, whether by Plato or by anyone else, may turn out to be internally incoherent, inconsistent, or ambiguous, the interpreter has no reason to begin by assuming it to be so; and there is strong evidence, as I hope to show, that the Phaedo is not like this. 5. Of the four arguments, according to the usual division, the first (the "cycli- cal" argument, or the argument from opposites) requires combination with the second (from recollection), if it is to fulfil Cebes' original requirements (70b2-4), and the second with the first (77cl-d5); only the last two (that from the alleged "affinity" of soul to the unchanging, and the final argument) appear as self- standing "arguments for immortality"- though if Cebes is prepared to attribute the same description separately to the first and the second too (as he implicitly does at 73a2-3), it will do no harm to go on doing the same. 6. Gallop 1975. 7. Bostock 1986. 8. Burger 1984. 9. Dorter 1982. 10. Op. cit. p. v.
11. Op. cit. p.3. 12. Op. cit. pp. 3-4. 13. What is it to write in a non-literary way? How do we draw the line between "literary elements" and "arguments?" However, a fair working exam- ple of the kind of division Dorter had in mind is probably provided by Bostock.
14. It is of course consistent with this to hold both that his ultimate concern would have been with finding persuasive arguments, whether about immortali- ty or anything else, and that this is the concern of the characters in the dialogue. (As Tuozzo rightly said, my later argument directly implies this latter point.) The question I raise is simply whether each and every piece of argumentation must be assumed to be presented as persuasive (as we shall be likely to imply, if we satisfy ourselves with establishing its strengths and weaknesses), or whether we should ask of any if it might not have some further purpose in its context within the conversation. Perhaps, from a dramatic point of view, any character who proposes an argument must be usually be supposed to be serious about it; but the author is not bound by the same limitations as his characters. 15. The difference I allege is not over the aims of philosophy if these might broadly, and no doubt naively, be characterised as the discovery or demonstra- tion of the truth by argument (see preceding note), but rather over the conditions under which someone may properly be said to be doing it. For Plato, or at least for the Socrates of the Phaedo, the requirements seems to be (i) that whoever is talking is talking to one or more other people, (ii) that they are talking about the right sort of subject (i.e. that it is one of those that Socrates thinks important), and (iii) that both sides are willing and able to treat what is said critically. (The justification for this list of requirements will, I hope, be provided cumulatively in what follows.) If so, any utterances or argument by itsel would not count as "philosophical", while a set of utterances which we might want to treat as "rhetorical" (i.e. merely persuasive, lacking adequate support of the appropriate kind) would count as such, provided requirements (i)-(iii) were fulfilled. Thus e.g. - as I shall argue - there would from Socrates' point of view be no differ- ence in kind between his "defence," on the one hand, and the arguments for immortality, to which they give rise, on the other; nor between the two parts of the argument from affinity. See esp. p.163-4 below. 16. See e.g. Stokes 1986.
17. Rowe 1992.
18. �aa,ia2a, ciic6; 78c7. 19. 84c6-7. 20. 91a-c.
21. 9laS-9. 22. 10764-9.
23. So forming a kind of false climax: compare Agathon's speech in the Symposium, and Socrates' second speech on love in the Phaedrus, both of which similarly tease us, momentarily, with a sense of closure. 24. Philosophy, then, is a dangerous game; we can never be quite sure of hav- ing played according to the rules (cf. Simmias' reference to "human weakness" at 107bl). 25. 88c-d. 26. Tuozzo objected to this that "neither of them ever in fact directs a dialecti- cal investigation by assuming the questioner's role," and that when they are
unhappy with a conclusion, they respond with speeches, which though insight- ful, "express that insight by means of an image." I doubt whether their objec- tions to the affinity argument can be written off so easily; and if they do not direct any "dialectical investigation" of the type familiar from the "Socratic" dia- logues, they certainly ask questions. The real issue, however, is once more about what constitutes "philosophy" (see n. 15 above), or dialectic. The philosopher/dialectician in Plato is not always the one who gets to ask the ques- tions ; in the Phaedrus, in fact, the position is reversed— it is his Myot which get to be questioned (Phaedrus 276a, 276e-277a, 278c-e). What matters is that Socrates, Simmias and Cebes are all involved in a particular type of conversation (i.e. in which each is properly critical of the others) and on a particular type of subject (whether or not the soul is immortal is, of course, a matter of crucial importance to Socrates). Socrates certainly still plays the leading part in that conversation; but Simmias and Cebes are not afraid to tackle him, and match argument with argument -which surely justifies my description of them. 27. See esp. Simmias' speech at 85b10-d10, which prepares the way for his and Cebes' objections. 28. 75c-d, 78d. 29. 59a, d. 30. 88c; d. also the intervention of the confused, anonymous objector at 103c5-6. 31. I believe it true to say that the only evidence for regarding Simmias and Cebes as Pythagoreans comes from the Phaedo itself: a) they have associated with (w�·yiyvea9a�) and heard Philolaus, who is undoubtedly a Pythagorean; b) Simmias sponsors the theory of soul as a kind of apiiovia which Echecrates, also certainly (?) a Pythagorean, is firmly attached to. But neither cruyyíyvoo9<1\ nor hearing someone necessarily implies adherence to his school; and if it does imply it, Simmias and Cebes are quite certainly (also) Socratics. Nor does the fact that someone who belongs to a particular school accepts an idea necessarily
imply that the idea belongs to the school; even if it did imply it, someone else's accepting it does not necessarily imply that the someone else belongs to that school, unless perhaps the idea is central to its doctrines. But the attunement theory is actually at odds with other central ideas associated with Pythagoreanism (immortality, transmigration), and Aristotle, for one, does not describe it as Pythagorean, even though he has plenty of opportunity to do so, having just referred to the Pythagorean doctrine of transmigration (De animal I. 3- 4). At least since Burnet, it has been usual to treat the whole dialogue as being set in a Pythagorean ambience, because Phaedo relates the conversation in Echecrates' home town of Phlius, where others are reputed to have shared his Pythagorean leanings. But the setting in Phlius is already sufficiently explained by the fact that Philus is on the way from Athens to Elis, Phaedo's native city that Echecrates and Phlius are a long way from Athens, and news of Socrates' last day in prison have not yet reached them, and that Echecrates, being philosophi- cally minded, is naturally interested in such news (see McQueen and Rowe 1989). Finally, if it is sufficient for being a Pythagorean that one accepts at least one Pythagorean doctrine, then the Socrates of the Phaedo is Pythagorean because he accepts transmigration; but he rejects the harmony-theory the accep- tance of which by Simmias is allegedly one of the grounds for treating him as a Pythagorean. (See further below, on 61d-e.) 32. See p. 167-8 below. 33. Phaedrus 242a-b; cf. below on 62e-63a.
34. Cebes' responses at 61d7 and e6 suggest that Philolaus endorses the first of the two propositions paradoxically paired by Socrates (the prohibition against suicide), but not the second (that philosophers will want to die as soon as possi- ble)— as, according to Socrates, any philosopher should (c6). There is no neces- sity to assume that the mention of Philolaus is in any way connected with his status as a Pythagorean; it may be enough simply that he is a prominent philoso- pher who had visited Thebes when Simmias and Cebes were there. There is equally nothing to suggest that the ČÍÂÎ..O\ nve� whom Cebes says he also heard speaking against suicide were Pythagoreans (so incidentally downplaying Philolaus as an influence on him), although they might have been (61e7-8); since Socrates too is against it, the doctrine was not exclusively Pythagorean. 35. 63d-e.
36. Cf. n. 15 above. 37. If, that is, he has been a regular member of Socrates' circle: see p. 160-1 above. 38. 69e-70b. 39. "cÎ>c; 8e8i6tojv;" or rather, not w5 n�wv ÔtÔlÓtcoV, but there is perhaps a child even in us who fears such things"- i.e., as Socrates has jokingly suggested, that the soul may literally be blown apart on its separation from the body, especially if death occurs in a high wind (77d-e). 40. See 90e-91c, with p.158-9 above. 41. Such a view emerges especially clearly from a combination of 90e-91c, 85c- d (Simmias' excursus on methodology), and 99-101 (Socrates on his "second sail- ing," where he implicitly endorses Simmias' position). 42. For which see Gorgias passim, and those parts of the Phaedrus which criticise contemporary rhetoric.
43. lliad 23. 100-104. 44. Suotv 9å'tEPOV. n ovSafiov amv Kpr\aaOai To ei8evaif\ n se7vev�njaaaw ... 45. 71d6ff. 46. 71e11 navsca5nov.
47. ov8e alia )i0t 8oxei (sc. gilxavh gh o\>%1 itdtvra xatava7uu6iwa� ei5 so Te6vavai ... , 6tw ¡¡,ol15oKeiç ... 48.77b5-7.
49.70b2�. 50. See 76c12-13. 51. 72e. 52. Cf. n. 40 above, and pp. 169-70 below. 53. See esp. Meno 86a-b. 54. 77a-b. 55. 72e4-5.
78. "Disappointingly ad hominem," because they are valid only against Simmias' version of the theory, not against other and more interesting versions of it. 79. 95c4-d4. Actually, he says only that you say that it is insufficient:" but the fact that he responds to Cebes' challenge by constructing a new argument strongly suggests that he accepts the point.
80. 103a4-10. 81. 103c5-6. This brief contribution from Cebes further lifts him above the level of the majority of those present (see p. 160-1 above): if the anonymous objector needs Socrates' clarification in 103all-c4 (as Plato evidently thought we might), he does not; at the same time it is hinted that he has other, perhaps more serious, problems with Socrates' argument.