Jump to ContentJump to Main Navigation
Virtue in the CaveMoral Inquiry in Plato's Meno$

Roslyn Weiss

Print publication date: 2001

Print ISBN-13: 9780195140767

Published to Oxford Scholarship Online: November 2003

DOI: 10.1093/0195140761.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2017. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a monograph in OSO for personal use (for details see http://www.oxfordscholarship.com/page/privacy-policy). Subscriber: null; date: 27 February 2017

(p.185) Appendix I

(p.185) Appendix I

Recollection in the Phaedo

Virtue in the Cave

Roslyn Weiss

Oxford University Press

Abstract and Keywords

In the Phaedo, Socrates pulls away from and repudiates recollection as it is portrayed in the Meno. In the Phaedo, it is Cebes who champions the Meno's form of recollection; Socrates presents recollection in a new and distinct way. In the Meno, recollection of all knowledge is said to come by way of questions or diagrams; in the Phaedo, recollection is of pure Forms by way of deficient sensibles. In the Meno, the soul remembers what it learned in previous lives here and in Hades; in the Phaedo, the soul recollects what it has seen in the realm of the nonvisible intelligibles, the Forms. The version of recollection that appears in the Meno is designed specifically for the misologist Meno; the version of recollection that appears in the Phaedo is designed for Simmias and Cebes who love argument only too well: what they lack is the philosophical yearning for the otherworldly life of communing with transcendent Forms.

(p.185) Appendix I

Recollection in the Phaedo

Appendix I Appendix I

One reason for taking seriously the recollection thesis in the Meno is that the Meno is not the only place in the Platonic corpus where Socrates advances this thesis. It emerges in full force in the Phaedo (primarily at 72e–77a, though it is mentioned again later on in the dialogue); and it turns up again in the Phaedrus (at 250) as well. Although some scholars have imagined it present or “implied” elsewhere (Tim. 42b;1 Rep. 7.518b6–c2;2 Symposium,3 and even Statesman4), it is, in fact, rather conspicuously absent from these dialogues.5 Indeed, it seems that what lies behind the sightings of recollection in dialogues besides the Phaedo and Phaedrus is little more than that it strikes scholars that recollection would fit nicely there. That recollection is missing just where it would fit nicely, however, makes its absence even more profoundly felt.

Recollection is, we may note, glaringly absent as well from other

An earlier version of this appendix, entitled “The Phaedo's Rejection of the Meno's Theory of Recollection,” appeared in Scripta Classica Israelica 19 (2000): 51–70. I gratefully acknowledge permission to reprint. (p.186) Platonic dialogues where one would quite reasonably expect to find it, dialogues in which questions of what knowledge is and how it is acquired are prominent: the Symposium and, especially, the Theaetetus.6 Moreover, of the two places besides the Meno where recollection is found explicitly,7 the Phaedrus passage is so heavily mythic that perhaps one need not see in it anything literally intended. That leaves, in all of Plato, but one dialogue besides the Meno where recollection might qualify as a Socratic or Platonic “doctrine”: the Phaedo. If the Phaedo is, in fact, the only place outside the Meno where the recollection thesis is developed with some sustained effort and is proposed more than just mythically, then the strength of the case for taking seriously the Meno's recollection thesis is much diminished. Nevertheless, it is worthwhile to look at the recollection thesis in the Phaedo to see whether it sufficiently resembles the theory found in the Meno to serve as support and confirmation of it.

I will argue here that the Phaedo's discussion of recollection consciously draws itself away from the Meno's, making reference to the Meno for the express and sole purpose of severing all connection to it.8 I will contend that the Phaedo is interested in recollection only insofar as recollection and Forms provide mutual support for one another and, together, help make the case for the immortality of the soul.9 Yet recollection is not indispensable even in the Phaedo to (p.187) the case for the soul's immortality: the dialogue offers other arguments for the nonbodily existence of the soul—both in its preexistence and in its postexistence—that are independent of recollection. Nor is recollection regarded in the Phaedo as proved: its truth rests on the imperfectly established existence of Forms.10 Indeed, when Socrates notes that recollection is “out of tune” with the attunement theory of the soul, he does not presume that it is a foregone conclusion that the theory to be adopted is recollection theory—instead, he allows Simmias to choose the theory he prefers (Phdo. 92c). Moreover, even when Simmias chooses recollection because, as he says, it is grounded in a “hypothesis,” the Theory of Forms, that he is convinced he holds rightly (Phdo. 92d—e), Socrates, not content to let matters be, shores up the first argument he offered against attunement theory with two additional ones.11

It is probably fair to say, then, that Socrates is less than wedded to recollection even in the Phaedo. One imagines that, were it to turn out that people have no preexistent souls that might once have known but since forgotten the Forms, Socrates would simply seek some way other than recollection to account for how human beings come to posit realities that transcend the sensibles that share their names. Indeed, such an alternative explanation is suggested later on in the Phaedo, where Socrates, in describing his second‐best quest for “causes,” arrives at Forms to provide a “safe” solution to logical puzzles that are generated by experience (Phdo. 100–102).12

(p.188) Socrates in the Phaedo, then, avoids dogmatism both about the Forms and about recollection. And this is so despite his meeting no resistance in Simmias and Cebes to either thesis: the same Simmias and Cebes who stubbornly resist virtually everything else Socrates proposes become surprisingly deferential when he speaks either of Forms or of recollection.13 Indeed, against the backdrop of Simmias's and Cebes' ready endorsement of these Socratic offerings, Socrates' own restraint and circumspection are all the more striking.

Regardless of the degree to which Socrates is committed to the version of recollection he promotes in the Phaedo, he leaves no doubt that this version of recollection is incompatible with the version in the Meno—and that he means to have nothing to do with recollection as it is presented there. He signals in the Phaedo his break with the Meno in several ways: (a) dramatically, by challenging rather than defending the kind of recollection he espoused in the Meno (and leaving its defense to Cebes); (b) philosophically, by correcting the Meno's version of recollection; and (c) by making parts of the Phaedo's argument incomprehensible unless read as countering the Meno's account. Let us turn now to the Phaedo's discussion of recollection.

After Socrates does his best to establish that the souls of the dead exist (for if they died, eventually everything would be dead) (Phdo. 72a—e), Cebes chimes in with a supplementary argument: “And besides, Socrates, [the existence of the souls of the dead may be proved] . . . according to that thesis (logos) that you are always accustomed to spouting,14 that for us learning turns out to be nothing (p.189) but recollection” (Phdo. 72e3–6). In Cebes' understanding of what he calls “Socrates' ” thesis, what one is now reminded of one must have learned at some former time, and such “being reminded” would be impossible unless the soul existed somewhere before being born in human form. Unless our souls were immortal, Cebes concludes, we could not learn.

This first stage of Cebes' presentation of recollection theory corresponds roughly to the myth presented in the Meno at M. 81a—e. Indeed, two of the difficulties that plague the Meno's recollection myth recur in Cebes' account of Socrates' recollection thesis. First, it is not clear in Cebes' account how the “learning” done in the “somewhere” where we presumably existed before we were born is accomplished: if it, too, is recollection, does that not create a regress that cannot be stopped‐and, consequently, a chain of learning that cannot get started? It is true that in the Meno myth the soul is said to have “seen” (heōrakuia) all things; yet insofar as the myth also proclaims that all learning is recollection, it is not clear that there can be a first seeing that unproblematically constitutes the first learning. Second, the nature of the “somewhere” where our souls presumably existed before being born in human form is not clear, although Cebes, like the Meno myth, gives us no reason to think it is a place radically different from those places familiar from life and legend. We may observe, however, that whereas the Meno myth begins with the assumption that the soul is immortal, Cebes uses recollection to prove the immortality of the soul.15

(p.190) When Simmias cannot “recall” the “proofs,” apodeixeis, for the thesis that learning is recollection, Cebes rehearses for him the proof in the Meno's slave‐boy‐demonstration. There is one excellent argument, Cebes says, namely, that when people are questioned, they are able, if someone asks the questions well, to say by themselves all that is; yet they surely would be unable to do so unless knowledge and a “right account,” orthos logos, happened to be present within them. Thus, he continues, if one leads people to diagrams or anything else of that sort, there is proof positive that this is so (Phdo. 73a7–b2).16

The reference to the Meno is unmistakable.17 The proof Cebes offers is but a summary of what takes place in the Meno's slave‐boy‐demonstration. Socrates, there, leads the boy to a diagram and asks his questions well—too well.18 The boy, in turn, comes up with the (p.191) correct answer, presumably on his own. As we saw in Chapter 3, the slave‐boy does not really produce the correct answer on his own. Nevertheless, Meno in the Meno concurs with Socrates' assessment that he does; and Cebes in the Phaedo appears to reach the same conclusion. Furthermore, although Socrates, in his recapitulation of the slave‐boy‐demonstration, determines that what the slave‐boy has in his soul is not knowledge and a correct logos but only true opinions,19 Cebes assumes, just as the Meno's discussion predisposes Meno to assume, that what those questioned must have inside them is knowledge.20

Perhaps the most striking feature of the introduction of the Meno's recollection thesis into the Phaedo is that the character who both introduces and defends it is not Socrates but Cebes. Not only is Socrates not the one to advance what Cebes presents as a customary Socratic view, but he will propose, momentarily, in his own name, a very different theory, for which he will adduce a very different proof. Moreover, Socrates immediately disparages Cebes' proofs: “But if you are not persuaded by that, Simmias . . . then see whether you might agree by looking at it this way” (Phdo. 73b3–4). Socrates even plants doubt in a Simmias who, apparently, has none: Socrates says, “For, you are indeed distrustful as to how what is called learning is recollection?” (Phd. 73b4–5),21 to which Simmias says, “I am not (p.192) distrustful,” apistō ou (Phdo. 73b6). Simmias in fact declares himself “nearly convinced,” schedon peithomai, by Cebes' way of putting the matter; he is nearly able to “recall,” schedon memnēmai, that learning is recollection (Phdo. 73b6–9). It is clear even to Simmias, however, that So crates wants him to resist Cebes' account (which is, of course, the Meno's account) of recollection and to hear Socrates' new account instead. He is willing to oblige: he would be pleased, he says, to hear how Socrates would put it (Phdo. 73b9–10).

What follows is Socrates' statement of the theory that learning is recollection, accompanied by a new proof. Neither the content nor the demonstration that Socrates presents resembles Cebes' account at all.

Socrates begins his account by distinguishing it from that of Cebes. “I, for my part (egōge),” he says, “put it this way” (Phdo. 73c1).22 The way Socrates puts it is, in other words, not Cebes' way. In what follows, Socrates departs both from Cebes' opening account of recollection, that is, from Cebes' paraphrase of the Meno's myth, and from the proofs Cebes subsequently provides for it, that is, from his distillation of the slave‐boy‐demonstration.

In the Meno, as we saw in Chapter 3, the slave‐boy‐demonstration is hardly the defense or proof that it purports to be of recollection as it appears in the recollection myth. On the contrary, the picture of the process of recollection that emerges from the myth is very different from the one that derives from the slave‐boy‐demonstration. The learner, according to the myth, is a solitary inquirer; no mention is made of his having need of another. Having lived numerous lives before the present one and having seen and learned all things, namely, the things both here and in Hades, he is able to recollect what he knew previously. Moreover, once he recollects a single thing, the recollection of all the others requires but courage and perseverance since all things have a natural kinship to one another. The process of recollection as depicted in the slave‐boy‐demonstration, (p.193) by contrast, resembles nothing so much as elenchus, a procedure that requires the participation of both a questioner and an answerer, a procedure that relies on questioning to arouse opinions held by the answerer but hidden from his view.23 The Phaedo's recollection thesis needs to defeat both the Meno's myth‐related brand of recollection and the sort of recollection found in the slave‐boy‐demonstration. To that end, it replaces both (a) the myth's notion that the sole way in which each of us learns is by recapturing by ourselves, by trying hard, things we saw in a previous lifetime, since “all nature is akin,” and (b) the slave‐boy‐demonstration's notion that all learning takes place only by “recollecting,” that is, by being asked questions and being shown diagrams, with (c) the distinct notion that sometimes when we learn we do so by recollecting Forms with which we were acquainted before birth.

Two divergences from Cebes' statement of the recollection thesis thus appear immediately: first, whereas according to Cebes, whatever was learned previously must have been learned before one was in human form, according to Socrates in the Phaedo, what was previously known could have been learned at any previous time; second, whereas according to Cebes, the proof of the soul's immortality lies in that one can only be reminded of what one has learned before this lifetime, according to Socrates in the Phaedo, the soul's immortality is not derivable solely from the phenomenon of “being reminded,” since one can be reminded, as Socrates says, as long as one knew the thing one is reminded of “before,” proteron (Phdo. 73c2)—that is, at any previous time.

Socrates is not, of course, correcting only Cebes. What he is correcting is recollection in the Meno. In affirming that one's being reminded requires only that one have known “before,” he challenges both the implication of the Meno's myth that all kinds of knowledge are “seen” prenatally and its assumption that the soul is therefore immortal. In the Phaedo, the immortality of the soul will stand or fall with the recollection specifically of Forms—entities which, as the Phaedo argues, cannot be “seen” by embodied souls, but only by disembodied ones.

As Socrates continues, he makes his position even clearer. It is not that all learning is recollection (as in the Meno); it is only when (p.194) knowledge comes to one by one's being reminded that there is recollection (Phdo. 73c4–5).24 Any other way of learning, then, is not recollection. And how is one reminded? Is it, as the Meno would have us believe, by being asked a question or by being shown the answer in a diagram? Is it true, in other words, that the slave‐boy was “being reminded” of something?25 Not according to Socrates in the Phaedo. In the Phaedo, being reminded occurs when someone sees, hears, or in some other way senses something and, upon doing so, recognizes not only the thing perceived but also, as a result of that perception, something else (Phdo. 73c6–8). Moreover, the knowledge by which a thing of which one is reminded is known must be, Socrates insists, not the same as, but different from, the knowledge by which the original thing is recognized (Phdo. 73c8).

Many subtle and ingenious suggestions have been made with respect to what Socrates intends by this last qualification, that is, by the stipulation that the knowledge by which the reminding thing is known must differ from the knowledge by which the thing of which one is reminded is known.26 Hackforth (1955), 67, n. 4, for example, thinks that the qualification seeks to rule out as bona fide recollection one's being reminded, when one perceives x, of the characteristics of x that one does not perceive at the moment—since the knowledge by which one knows x's characteristics would be the same knowledge as that by which one knows x. For Burnet (1911), 54, it is knowledge of opposites that does not count for Socrates as genuine recollection—since the knowledge of one of the opposites is not distinct from the knowledge of the other. For Ackrill (1974), 184, what Socrates wishes to exclude as a true case of being reminded is the case in which “thinking of y is already involved in (p.195) perceiving and recognizing x. One would not want to say that something brings so‐and‐so to mind if so‐and‐so is necessarily in mind when that something is.” According to Ackrill, if one, on seeing a picture of Simmias, thinks of Simmias, then one has not been reminded—for in recognizing that this is a picture of Simmias, one has already, by the same knowledge, as it were, thought of Simmias.27

Since the case of one's being reminded of Simmias by seeing his picture serves as Socrates' prime example of what being reminded is, it certainly seems that Ackrill must be misunderstanding Socrates' point. And neither Burnet nor Hackforth explains why it is so important to Socrates that the knowledge with which the reminding thing is known differ from the knowledge with which the thing of which the reminding thing reminds one is known. Yet if one reads this passage of the Phaedo as a correction of the Meno's recollection thesis, it becomes immediately evident what is at stake for Socrates in this otherwise obscure point. In the slave‐boy‐demonstration, the diagonal that is drawn reminds the slave‐boy of no other diagonal. First, if, as was argued in Chapter 3, Section ii, the diagonal was in place in the diagram from the start, the boy's seeing it certainly reminded him of nothing. Second, even if the diagonal was not drawn until the demonstration's end, nevertheless the boy's answer consisted of nothing more than his pointing to the drawn diagonal: there is no other diagonal that this diagonal reminds him of.28 Moreover, the slave‐boy is surely not reminded by the sight of one thing of another thing that is the object of a different knowledge. Only if the slave‐boy were reminded, by seeing the drawn diagonal, of, say, the geometer's perfect and purely conceptual diagonal,29 which surely is the object of a separate and distinct knowledge, then and only then would his learning qualify as “recollection.”

(p.196) Through this caveat, Socrates sets the stage for what is for him the most important kind of reminding and being reminded, namely, when a thing of “lower” ontological status serves as a reminder of a similar thing of “higher” ontological status. In a case of this kind, the one who knows surely knows the things in question with two different knowledges. When one is reminded of Simmias by seeing a “drawn” Simmias, one is reminded by an ontologically inferior Simmias of an ontologically superior one.30 Although it is true that one knows Simmias and Cebes by different knowledges, and that one knows Cebes' lyre and Cebes by different knowledges, the more interesting and significant instance of knowing by different knowledges, the instance toward which the discussion progresses and in whose discovery it culminates, is that of knowing the real Simmias from knowing the drawn one. It is by way of the Simmias example that Socrates is able to make the transition to what most concerns him: how one is reminded by sensibles of their corresponding Forms.

Having reviewed several kinds of reminding in which things that are different from one another serve, respectively, as the reminding and reminded things—seeing lyres and cloaks reminds one of their owners; seeing Simmias reminds one of Cebes; seeing a drawn horse or lyre reminds one of a person; seeing a drawn Simmias reminds one of Cebes—Socrates goes on to conclude that recollection occurs especially when the thing recollected has been forgotten, either through lapse of time or through inattention (Phdo. 73e).31 He then proceeds to discuss the last of the cases reviewed, the case in which one is reminded of something by a similar thing—for example, when a drawn Simmias reminds one of the real Simmias. It is only in this (p.197) case, says Socrates, that the question arises of whether the reminding thing is deficient with respect to that of which it reminds.32

Having broached the issue of deficiency, Socrates is able to take his first step in the direction of what will shortly be his proof for immortality. Socrates speaks of the relationship between equal logs or stones, on the one hand, and the Equal Itself, on the other.33 Although it is the sensible equals that “remind” one of the Equal Itself, they do so by way of their deficiency:34 unlike the Equal Itself, they can appear unequal and can change from being equal to being unequal.35 How, Socrates wonders, could one recognize the deficiency of equal logs and stones unless one were formerly acquainted with the perfection of which these fall short?36 It is only when things are similar, let us note, that from seeing the one that is deficient, a person comes to think of the other that is perfect.37

(p.198) The perfect realities, the Forms, that are known by way of the perception of their similar sensibles, Socrates now contends, must have been known by us before such time as we perceived the sensibles that are similar to them and recognized their deficiency. Since we begin sensing at birth, Socrates reasons, we must have had knowledge of the Forms before birth.38 We apparently do not hold on to our knowledge of the Forms when we are born and during our lives, because, as Socrates points out to a confused Simmias (at Phdo. 76b), we cannot give an account of them.39 That means we must forget them when we are born. And if we forget them at the moment of birth, we surely could not also get them at the moment of birth. Therefore, we must get them before we are born. By perceiving sensibles, we are reminded of the perfect realities that we once knew—before birth. Therefore, our souls must have existed before birth.40

(p.199) The recollection in the Phaedo, then, that proves the soul's immortality is radically different from the recollection in the Meno's slave‐boy‐demonstration. The recollection in the Phaedo is not “recollection” by way of questions or diagrams; it is recollection of pure Forms by way of deficient sensibles. The recollection in the Meno's slave‐boy‐demonstration does not even require being reminded; and the drawn diagonal in the Meno is never recognized to be in any way deficient. Moreover, being reminded in the Phaedo is clearly not the whole of reacquiring knowledge formerly possessed; rather, being reminded is but the first step: it makes the recollector aware of what he does not know by reminding him that he once knew it. According to the Phaedo, one recollects, that is, is put in mind of, via sense perception, things that one formerly knew; but insofar as one is not able, simply by virtue of having been reminded, to give an account of the things one is put in mind of, it seems that the recollection by which one is put in mind of something previously known does not suffice for coming to know again what one knew before. (On this point, see Phdo. 76b5.)

Recollection in the Phaedo departs significantly, too, from recollection as it is depicted in the Meno's myth. What the soul in the Phaedo remembers is what it cannot know when it is embodied; in the Meno, however, it remembers what it learned in previous lives—both here and in Hades. In the Meno myth, one is able to recollect what one knew in previous lives because “all nature is akin,” sungenous (M. 81d1);41 in the Phaedo, one is able to recollect what one knew because one's soul is “akin,” sungenēs, to the Forms (Phdo. 79d3, 84b2). In the Meno, one has knowledge of the very things of which one has opinions; in the Phaedo, one's soul, by following reason, beholds what is true and divine “and not the object of opinion,” adoxaston (Phdo. 84a8). In the Meno, one's soul recollects what it has seen both here and in Hades; in the Phaedo, one's soul recollects what it has seen in “ ‘Hades’ in the true sense,” Haidou hōs alēthōs (Phdo. 80d6–7), that is, in the realm of the nonvisible (aides) intelligibles, the Forms.42 In the Meno, no release from the cycle of rebirth (p.200) is envisioned: the soul goes from here to Hades and back, over and over again; in the Phaedo, the soul of a man who has lived philosophically is returned to its natural home to dwell in the rarefied atmosphere of the pure, the unseen, the intelligible: the Forms.

I have argued that (a) the Phaedo is the only dialogue within the Platonic corpus besides the Meno where recollection is set forth as a Socratic or Platonic account of human learning, yet (b) the Phaedo's account departs radically from the Meno's, and (c) the Phaedo's account repudiates the Meno's. In light of these conclusions, it seems that if, for reasons internal to the Meno, one is inclined to believe that Socrates does not seriously endorse its version of recollection, one need not, for reasons external to it, suppress that inclination.

In closing, it is perhaps worth noting that Socrates is an unqualified fan of recollection in neither the Meno nor the Phaedo. Just as he limits his endorsement of recollection in the Meno, so he does, as we have seen, in the Phaedo as well: he raises the possibility that there might be no Forms, in which case there would be nothing for the immortal soul to be remembering when it notes the deficiency of sensibles (Phdo. 76d–e). If Socrates is unhappy with the recollection thesis in the Meno and, therefore, provides a different recollection thesis in the Phaedo, why, we may wonder, does Socrates raise doubts about recollection in the Phaedo as well?

Socrates' introduction of recollection in both the Meno and the Phaedo—in their respective versions—reflects the great care he takes for his interlocutors. Meno is a misologist of the very type described in the Phaedo at 90c–e.43 When arguments conflict or disappoint, he is quick to bail out. For the sake of securing Meno's continued participation in what he, Socrates, regards as all‐important virtue inquiry, Socrates tailors a recollection thesis to suit Meno—one that recalls Pythagorean ideas familiar to Meno, plays on Meno's pride in his ability to remember, and capitalizes on Meno's love of the esoteric and ostentatious, the tragikē. Unlike Meno, however, Simmias and Cebes are avid arguers. In that sense, they are already (p.201) deeply philosophical.44 We note in this regard Simmias's speech at Phdo. 85c, a speech worthy of Socrates in the Meno: “I think, Socrates, as perhaps you do too, that in these matters certain knowledge is either impossible or very hard to come by in this life; but that even so, not to test what is said about them in every possible way, without leaving off till one has examined them exhaustively from every aspect, shows a very feeble spirit” (trans. Gallop [1993]). The recollection thesis, then, that Socrates proposes in the Phaedo has the goal of turning Simmias and Cebes, not to philosophy construed as the life of argument, but to a different kind of philosophical life, the peculiarly otherworldly life of communing with transcendent Forms. Indeed, it is this latter type of philosophy, the sort that separates one from one's body and frees one's intellect to consort with truth, that is repeatedly called in the Phaedo not simply “philosophy” but philosophy “in the proper manner,” orthōs (64a4–8, 67b4, d8, 69d2). It is also called “real philosophy,” tōi onti, (68b2–3). When Socrates, then, prevents Simmias from being satisfied with Cebes' presentation of recollection, he intimates, in effect, not simply that that version is defective in certain ways but that it is not suitable for Simmias. For the recollection thesis that Socrates fashions for the sake of Simmias and Cebes is not one that urges the ordinary soul, as the Meno's version of recollection does, to seek to remember all that it has learned here and in Hades but one that inspires the philosophical soul to yearn to recover Forms. Yet Socrates succeeds no better with Simmias and Cebes than he does with Meno. Meno does not turn to the life of argument; Simmias and Cebes do not turn away from it.45 Despite his failures, however, the valiance of Socrates' efforts and the lengths to which he is willing to go for the sake of his interlocutors attest to the boundlessness of the benefaction he seeks to bestow on all those he encounters—old or young, Athenian or stranger. (p.202)


(1.) Vlastos (1991b), 54.

(2.) See, for example, Hackforth (1955), 77; Gulley (1954), 195; Adam (1969), II, 98.

(3.) Bluck (1961a), 50.

(4.) Skemp (1952), 76, sees in the Statesman's “sensible likenesses” (aisthētai homoiotētes) at 285e “the later form of the earlier doctrine of Recollection.”

(5.) Although there is talk of reincarnation in the Timaeus, recollection does not appear there. As for the Republic passage, see Klein (1965), 158, who rightly insists that although Socrates does indeed maintain in this passage that there is within each of us the power to know, so that education is not the pouring of knowledge into ignorant souls, “still, there is no mention of ‘recollection’ in this passage.”

(6.) See Klein (1965), 157–72. Hackforth (1955), 77, approvingly quotes Cornford (1934), 28, who explains recollection's absence from the Theaetetus by saying that the Theaetetus “presupposes that we know the answer to the question here to be raised afresh: what is the nature of knowledge and of its objects.”

(7.) One might say that recollection is implicitly parodied in the Euthydemus at 293–96, where Euthydemus and Dionysodorus seek to demonstrate to Socrates that he knows everything, maintaining that he knew “even when you were a child and when you were being conceived. And before you yourself came into being and before the foundation of heaven and earth, you knew absolutely everything, if it is true that you always know” (trans. Sprague [1965]).

(8.) It is the contention of this appendix, not only that the two versions of recollection are different, but that the Phaedo's deliberately pulls away from the Meno's. Many commentators recognize that the two versions differ, but they are reluctant to assert that Plato rejects in the Phaedo the version he proffers in the Meno. See, for example, Anderson (1993), 125; Bostock (1986), 63; Ackrill (1974), 177; and Hackforth (1955), 74. Bostock is typical. He says, on the one hand: “As Socrates indicates at 73b3–4, the version now to be presented is not meant to be the same as the Meno's version.” But he insists, on the other, that what the Meno contains is “an earlier version of this argument.”

(9.) Scott (1995), 56–73, argues forcefully against what I, too, regard as an ill‐conceived view of recollection in the Phaedo, namely, the view that it is a theory of concept formation.

(10.) Although the Phaedo has its own “method of hypothesis,” it instantiates the Meno's hypothetical method in the way it proceeds, at Phdo. 72e–77a, to prove the soul's immortality by way of recollection. It asks what would have to be true if the soul is to be immortal. It answers: learning would have to be recollection. And what would have to be true if learning is to be recollection? Answer: there would have to be Forms. And do we know that there are Forms? Answer: not quite. So we do not really know that the soul is immortal. (Although Socrates seems at first to recognize that only the argument for immortality founders if there are no Forms—“this argument will have been in vain” [Phdo. 76e4–5]—he goes on to speak as if immortality itself is on the line: “and if not the former [that is, if the Forms do not exist], then not the latter [that is, our souls do not exist before birth] either” [Phdo. 76e]).

(11.) The first is the rather peculiar argument that since an instrument can be more or less in tune, it would follow from the assumption that the soul is an attunement that a soul could be more or less a soul; but since a soul cannot be more or less a soul, then, if it were an attunement, all souls would have to be, contrary to fact, equally good (Phdo. 93a–94b). The second is the argument that whereas an attunement, qua compound, follows and indeed cannot oppose its components, the soul often opposes what, on attunement theory, would be its components, namely, the body (Phdo. 94b–95a).

(12.) See also Rep. 7.524c: “the intellect was compelled to see big and little, too, not mixed up together but distinguished, doing the opposite of what sight did. . . . Isn't it from here that it first occurs to us to ask what the big and the little are? . . . and so, it was on this ground that we called the one intelligible and the other visible” (trans. Bloom [1968]).

(13.) Burger (1984), 70, notes how readily Cebes and Simmias accept recollection: “the recollection argument will turn out to be the only one in the entire conversation that both Cebes and Simmias wholeheartedly endorse. It is, therefore, the one too that Socrates shamelessly exploits: he need only appeal to the recollection thesis, and his interlocutors will immediately give up any opinions they believe conflict with it. Socrates takes advantage of their acceptance, despite the fact that it is based upon unexamined, and even preposterous, assumptions about the psyche, knowledge, and the objects of knowledge.”

(14.) What is the best way to understand Cebes' characterization of the recollection thesis as something that Socrates is “always accustomed to spouting” (Phdo. 72e4–5)? Are we to think that Socrates regularly put forward the thesis that learning is recollection? Burnet (1911), 51, for one, thinks that “it is very difficult to regard this definite statement as a fiction.” But Burnet also believes that the Theory of Forms must be attributed to the historical Socrates, inasmuch as Plato would not, in his depiction of Socrates' dying day, attribute to Socrates views that are not really Socrates' but are “novel doctrines” of Plato himself (xi–xii). I admit that I do not find as repugnant as Burnet does the notion that Plato uses Socrates as a mouthpiece for his own views: if Plato believes that his views are in some way a natural extension of Socrates', then it might well seem to Plato an act of devotion to ascribe his own views to Socrates (as, say, Pythagoreans ascribe their views to Pythagoras). Nevertheless, it is rather puzzling that Cebes speaks of recollection as a familiar Socratic thesis. It is true that Socrates, referring to the Forms, says the following in the Phaedo: “Well . . . this is what I mean: it is nothing new, but is what I have spoken of incessantly both at other times and in our earlier conversation” (Phdo. 100b1–3). The two cases are hardly the same, however. For in the case of Forms we might at least be able to say that the character Socrates talks incessantly about them, whereas not even the character Socrates is always talking about recollection. Perhaps it is best to suppose that Cebes, in his youthful exuberance, simply exaggerates: it is not, after all, Socrates but Cebes who says of Socrates that he is accustomed to speaking always of learning as recollection.

(15.) As was argued in Chapter 3, although it may appear that the slave‐boy‐demonstration ends by deriving the soul's immortality from recollection by way of the idea that the truth about the things that are is in our soul always (M. 86b1–2), the fact is that immortality is simply revived and reaffirmed without argument and without any real connection to recollection. See Chap. 3, n. 90.

(16.) Burnet (1911), 52, following Bury (1906), 13, argues that the matter of leading one to diagrams “is opposed to, rather than included in,” the process of asking questions well. This reading, however, is a forced one. It requires bypassing the most straightforward reading of the preposition that introduces the matter of leading one to diagrams: rather than as “thus,” Burnet renders epeita as “secondly”—and this despite the absence of a prior prōton men, “firstly.” In the absence of a prōton men, Burnet's reading requires that heni logōi, “by one argument,” be read “firstly,” even though he notes that this phrase has a very different sense (“in a word”) at Phdo. 65d13. None of this is necessary. In the Meno, the process of putting questions well to the slave‐boy, that is, of asking him leading questions, is not separate from the use of diagrams: the questioning does not occur apart from the diagrams. Moreover, Cebes himself regards the argument he presents as “one excellent argument” (Phdo. 73a7)—and this despite Simmias's request for “proofs,” in the plural. We note, too, that Socrates subsequently refers to Cebes' proof in the singular, tautēi.

(17.) Somewhat remarkably, commentators have been hesitant to make the connection definitively. Gallop (1993), 88–89, says that this passage contains what is “possibly an allusion to Meno 81e–86b.” Burnet (1911), says: “This seems a fairly certain reference to Meno 82b9 sqq. . . . No doubt, if we hold this doctrine and its proof to be genuinely Socratic, the reference to the Meno is less certain.” I am not sure exactly why Burnet thinks that the reference to the Meno becomes suspect if one takes the view and proof as genuinely Socratic, unless he takes the Meno to be clearly un‐Socratic. But whatever he may mean, it is hard to see how there could be anything here but a direct allusion to the Meno. See also Hackforth (1955), 74, n. 1, who thinks that with respect to the leading of people to diagrams, there is a clear reference in this passage to the Meno, but that with respect to the “proper questioning,” there may be only “a quite general reference to that Socratic ‘midwifery’ which is abundantly illustrated in the early dialogues.” I think, however, that, considering how many times Socrates in the Meno emphasizes that he is only asking questions and not teaching the slave‐boy, it is highly unlikely that there is anything here but a reference to the Meno.

(18.) Cornford (1952), 51, recognizes that the phrase “if one asks well” at Phdo. 73a might point to “some uneasiness” in Plato's mind with respect to the slave‐boy‐demonstration in the Meno, where Socrates' questions are leading ones.

(19.) We may compare Socrates' expression in the Meno, enesontai autōi alētheis doxai, “there are going to be present within him true opinions” (M. 86a7), with Cebes' kaitoi ei mē etungchanein autois epistēmē enousa kai orthos logos, “yet unless knowledge and right account were present within them” (Phdo. 73a9–10). Although Socrates speaks at one point in the Meno of the knowledge the slave‐boy has now (M. 85d9), he proceeds, as was argued in Chapter 3, Section iv, to reduce to absurdity the possibility that the slave‐boy has knowledge now, concluding that all he has now are true opinions.

(20.) Whereas Cebes is surely right to connect having knowledge with possessing a correct account, he is wrong to conclude that when those questioned are able to produce the right answer, that in itself demonstrates that they have knowledge and a correct account within. In the Meno, although Socrates contends that the slave‐boy's arriving at the right answer indicates that he has true opinions within does he really have, with respect to the diagonal, either knowledge or true opinions within? Surely, what the slave‐boy has within is the ability to follow the compelling proof for, and hence to learn, the new bit of geometry that Socrates teaches him.

(21.) See Hackforth (1955), 74, who says, rightly, that “the description of the Meno argument as ‘excellent’ (kallistos) is partly offset by Socrates' doubt whether Simmias finds it convincing.” Hackforth resists, however, drawing the conclusion for which I argue, namely, that Socrates is “repudiating the earlier argument for recollection and immortality.” The most Hackforth will concede is that Socrates regards the argument he will currently expound as “far superior” (see n. 8). See also Gulley (1954), 197: “In 73c Plato introduces a new aspect of the theory, hinting at its novelty in his suggestion that if Simmias is not convinced by it in its presentation so far, then perhaps he will agree if it is presented in another way (73b).” Let us observe, however, that it is not Socrates but Cebes who calls the Meno's argument kallistos (at Phdo. 73a7); all Socrates does is immediately offer to replace it.

(22.) See Burger (1984), 71: “Dissatisfied, apparently, with Cebes' enactment of recollection, Socrates takes over.” We shall see at the end of this appendix that Socrates wishes to substitute in the Phaedo for the version of recollection in the Meno a version suitable to Simmias and Cebes: just as the Meno's version is fashioned for Meno's sake, so is the Phaedo's designed for Simmias and Cebes.

(23.) The slave‐boy‐demonstration is not a genuine elenchus, of course, because Socrates in the demonstration is a knower who, in fact, teaches the slave‐boy something the slave‐boy never knew before.

(24.) The term epistēmē in this passage refers not to the agent's cognitive state but to the thing known: it is the just‐mentioned necessarily previously known “thing,” ti (Phdo. 73c1), that “comes,” paragignētai, to one (Phdo. 73c4–5). See Phdo. 75e4, where the “knowledges” (pl.), tas epistēmas, are clearly not the agent's many knowings but the many (previously) known things. The terms “knowledge” and “know” are used loosely in this passage, sometimes meaning no more than “coming to think of” and sometimes “having full understanding.” It has the former sense in many of the examples Socrates uses of reminding and being put in mind of, as well as at Phdo. 74b2–4 and 74c9.

(25.) As was noted in Chapter 3, Section iii, Socrates virtually never asks the slave‐boy if he remembers or is reminded of anything; the one time that he does so (M. 84e), he asks him if he remembers what the question is!

(26.) See Bostock (1986), 64, who notes Plato's “obscure condition that the knowledge of the reminding thing, and of the thing it reminds us of, should not be ‘the same knowledge.’ ”

(27.) See also Burger (1984), 73, who says: “Of course, one would ‘know’ the image only if one knows Simmias, and just for that reason, it is unclear how it could satisfy the condition that knowledge of what is recollected be other than knowledge of what causes the recollection.”

(28.) Even if the slave‐boy can generalize from this diagonal to all diagonals, he is not “reminded” of anything: generalizing is not the same as being reminded. One can—indeed, one must—generalize whenever one learns; otherwise, one would have to relearn everything each time anew. This rule holds for anything, from math to tying shoelaces.

(29.) Cf. Rep. 6.511d, where Socrates notes that although geometers use visible figures and make claims about them, their thought is directed not to the figures but to the thing they resemble: the Square Itself, the Diagonal Itself, “not the diagonal they draw.”

(30.) It has been asked why Socrates speaks of a “drawn” or “pictured” Simmias rather than of a picture of Simmias. The answer, I think, is that Socrates wishes to contrast Simmiases of different ontological orders. By speaking of a drawn Simmias and of a real one, Socrates speaks of different kinds of Simmiases; it is not as clear that a picture of Simmias is a kind of Simmias. In addition, a “drawn,” gegrammenon, Simmias, horse, and so on calls to mind the drawn diagonal, the grammē that reaches from one corner of the square to the other.

(31.) “Reminded” and “forgotten” are obviously being used rather broadly here to mean, respectively, “being put in mind of” and “not having in mind at the moment”; otherwise, seeing Cebes could not remind one of Simmias without Simmias's being quite forgotten. Thus, when Gosling (1965), 154, says: “Normally, when I see my wife's handbag it does not remind me of my wife, even if it makes me think of her: I am not that forgetful,” he is pressing “remind” and “forget” too hard or reading them too narrowly.

(32.) Gosling (1965), 160, is surely right to note that when we ask whether or not a pictured Simmias is deficient with respect to Simmias, we are not asking whether it is a good likeness but whether it is actually Simmias or just a representation of him. If the pictured Simmias lacks not a single feature of Simmias, then it does not remind one of Simmias; one thinks it is Simmias. To speak in terms of being reminded is to recognize an ontological falling short.

(33.) We should note that the Phaedo does not limit the Forms to those of mathematics and virtue. It puts all Forms on an equal footing, mentioning, in particular, largeness, health, and strength (at Phdo. 65d). Interestingly, it is just these three that Socrates uses in the Meno (M. 72d‐e) to illustrate his point that in defining a term one looks, not to the variable individual instances of the term, but to what they have in common. Yet in the Meno, there are no Forms that correspond to these qualities.

(34.) Scott (1995), 63, n. 12, asks why the senses are necessary as a catalyst for recollection: “Another possibility, one that Plato ignores, is that we grasp the forms by rational intuition without any need for the senses.” What Scott's question fails to take into account is that grasping the Forms “by rational intuition without any need for the senses” is not recollection. For Socrates, it is only learning by recollection that requires a sensible reminder. There are, however, for him, other ways of learning—perhaps even, as suggested at the beginning of this appendix and in n. 12, other ways of grasping the Forms.

(35.) Here, too, I agree with Gosling (1965), 160, that the sensible equals are not deficient in being less equal than the Equal Itself. To be deficient equals is to fail to be “eternally and immutably equal.”

(36.) The assumption here is that unless one already had the notion of perfection, one could not recognize things as deficient, yet one could (obviously) have the notion of perfection without having perceived deficiency. Perhaps it is because the prisoners in the Cave in Rep. 7 do not recognize the deficiency of their perceptions that they never “recollect” the real things of which their perceptions are mere images.

(37.) The sensible equals are other (heteroi) than the Equal Itself. But they are similar (homoioi) rather than dissimilar (anhomoioi) to one another. When Cebes is asked whether it is in being similar or dissimilar to the many equals that the Equal Itself comes to be known from them, Cebes answers, as if in a comedy routine, “Certainly.” The right answer is “from being similar.” Socrates lets the point go. (Later, however, Socrates is careful not to make the same mistake. When he wants Simmias to say whether we are born knowing or are later reminded of the things of which we acquired knowledge before, he says: “Then which do you choose, Simmias?” [Phdo. 76a9].)

(38.) This Socratic argument has caused scholars considerable consternation. See, for example, Cornford (1952), 51. It seems at first that Socrates must be confusing (a) the idea that we start sensing at birth with (b) the idea that at birth we make the determination that our sensibles fall short of the corresponding Forms. It is hard to see how, without this confusion, Socrates could derive from (c) the notion that our knowledge of the Forms must precede our judgment that sensibles fall short of them (d) the conclusion that we must know the Forms before birth, that is, before the moment at which our sensing begins. Surely, however, there is a better way of understanding Socrates' point: he is, after all, hardly so foolish as to think that as soon as we begin perceiving we begin making the determination that sensibles fall short of their corresponding Forms. Let us consider, then, the following alternative construal of Socrates' argument. Since, on the one hand, it is from sensibles that we are put in mind of the Forms, insofar as we judge sensibles deficient, and since, on the other hand, we are unable to get original knowledge of the Forms once we begin using our senses, it follows that we had to have gotten our original knowledge of the Forms before we were born—since we begin using our senses at the moment of birth. On this construal of the argument, Socrates relies, not on the notion that knowledge of the Forms necessarily precedes the judgment that sensibles are deficient with respect to them, but on a point made earlier, namely, that sentient beings are precluded from knowing the Forms insofar as the senses hinder reason's ability to know them (see Phdo. 65b–66a). Indeed, that Socrates' argument depends on this earlier point is confirmed by Simmias's remark: “That must follow from what has been said before, Socrates” (Phdo. 75b9).

(39.) Socrates here confirms a point made earlier in Cebes' statement of the Theory of Recollection, that there is a link between having knowledge and being able to give an account. A similar notion is found as well in the Meno's discussion of knowledge as what results from “working out the reason,” aitias logismos (M. 98a).

(40.) The Phaedo highlights, then, specifically that form of recollection that the Meno conspicuously avoids in its review of the slave‐boy‐demonstration (see Chap. 3, Sec. iv), namely, acquiring knowledge at some point before birth and recovering it later, in this lifetime. The only two possibilities the Meno considers are possession always and acquiring knowledge in the present lifetime.

(41.) As was noted in Chapter 2, Section iv, since in the Meno all nature is akin, there are no ontological levels that might clear a space for the kind of recollection we find in the Phaedo.

(42.) Plato's fondness for plays on words is in evidence also at Phdo. 92d, where Simmias prefers recollection theory to attunement theory because although recollection theory, too, has not been “proved,” apodeixasthai, it, unlike attunement theory, is worthy of “acceptance,” apodexasthai.

(43.) This passage of the Phaedo is very much in the spirit of M. 81d‐e and M. 86b‐c, where Socrates diagnoses as lazy, soft, and cowardly those who refuse to inquire, taking refuge in eristic argument. Socrates warns in the Phaedo that when arguments go awry, it is not the arguments but those who present them who are to be blamed. “But we must be courageous and be eager to be sound” (Phdo. 90e3). Let us note that Socrates, even in the Phaedo, does not promise that there will be knowledge at the end of rigorous argument.

(44.) When the others present show signs of frustration and require a pep talk by Socrates, Simmias and Cebes are the only ones who are not in danger of lapsing into misology. The only thing that might keep them from pursuing an argument is their concern for Socrates—not any unwillingness to continue arguing. See Phdo. 88c–91c. See also Phdo. 63a1–3: “There goes Cebes again, always hunting down arguments and not at all willing to accept at once what anyone may say.” Simmias, too, even near the dialogue's end (107a—b), still admits to having misgivings about the argument; he remains prepared to argue further.

(45.) Cebes and Simmias endorse a philosophical life they neither live nor understand. Neither Simmias nor Cebes can give an “account” of the Forms—indeed, Simmias believes that once Socrates dies, there will be no one who can. Furthermore, it is not at all clear to them that death is not fearful. And before Socrates suggests otherwise, Simmias counts even Evenus as a philosopher.