For reference (some god perhaps knows whose) here are what I determine to be the seven definitions of sophrosune (plus three transitional forms) which fall to elenchos in the Charmides. I follow Van Der Ben’s outline of the dialogue when possible, adhering to his selection of the Greek text where available. I add definitions a-d and 7, the numbering scheme, as well as all translations, analysis, and identification of spoilers.
First, a bare list:
(1) ἡσυχιότης τις / whatever is quiet (159b)
(2) αἰδώς / shame (160e)
(3) τὸ τὰ ἑαυτοῦ πράττειν / doing one’s own things (161b)
(4) ἡ τῶν ἀγαθῶν πρᾶξις / doing [not making] good things (163e)
(5) τὸ γιγνώσκειν ἑαυτόν / to know oneself (164d)
(a) ἐπιστήμη ἑαυτοῦ / knowledge of oneself (165c)
(b) ἐπιστήμη ἑαυτῆς / knowledge of itself (166c)
(c) …τῶν τε ἄλλων ἐπιστημῶν / …and the other knowledges (166c)
(d) …καὶ ἀνεπιστημοσύνης / …and non-knowledge (166e)
(6) τὸ εἰδέναι ἅ τε οἶδεν καὶ ἃ μὴ οἶδεν / to know [that one knows] what one knows and [that one does not know] what one does not know (167a)
(7) ἧι τὸ ἀγαθόν καὶ τὸ κακόν / [that knowledge by which are known] the good and the bad (174b)
Now some minimal remarks, rudiments of a more complete metalogical analysis, mostly identification of why the definitions fail (spoilers), except in the most important case of (6), for which, see my recent talk at sjc and a forthcoming post:
(1) ἡσυχιότης τις / whatever is quiet (159b)
Spoiler: original elenctic dilemma: a definition can be good (as a definition) or of a good (by definition), but not both. ἡσυχιότης τις is a good definition but not necessarily of a definition of a virtue (of something good), if a virtue is καλὸν, or relatively κάλλιστον. To learn the import and finality of this dilemma, it must be played out on multiple conceptual levels, to exhaust all avenues of escape.
(2) αἰδώς / shame (160e)
Spoiler: likewise with terminological shift to ἀγαθόν
(3) τὸ τὰ ἑαυτοῦ πράττειν / doing one’s own things (161b)
Spoiler: τὰ ἑαυτοῦ dilemma (1): If the definition is well-founded it picks out only wrong objects. (well-founded → wrong objects)
(4) ἡ τῶν ἀγαθῶν πρᾶξις / doing [not making] good things (163e)
Note: (G.I.) Goodness is built into the definition, at the cost of effectiveness / foundation.
Spoiler: τὰ ἑαυτοῦ dilemma (2):If the rightness of the object is built into the definition, we cannot say what it is of. (right object → non-well-founded.) But acknowledgment of the dilemma is postponed by the reflective reprise in which knowledge is introduced into its own scope (K.I.) and remains an open question here.
(5) τὸ γιγνώσκειν ἑαυτόν / to know oneself (164d)
Note: (K.I.) Here, the reflexive gambit opens. Knowledge is introduced into its own scope, in order, Critias hopes, to become its own foundation. This is not to “drop” the previous definition, as Van Der Ben frames it, but to marshal every means to realize it in the face of the impending dilemma. The dialectical richness of gambit is shown in the rapid development of the transitional formulations (a-d) which finally transform (5) into (6). As the opening statement of the reflexive gambit, even (5) can reasonably be regarded as a first draft of (6), but this is apparently not how Socrates is counting, since he identifies (6) as Critias’ third definition.
Spoiler: Explicit appearance of the τίνος-question: ἡ σωφροσύνη τίνος ἐστὶν ἐπιστήμη, ὃ τυγχάνει ἕτερον ὂν αὐτῆς τῆς σωφροσύνης; / Of what is sophrosune the episteme, which would be different from sophrosune itself? (166b)
(a) ἐπιστήμη ἑαυτοῦ / knowledge of oneself (165c)
Note: An hybrid expression, like the provocation “idea of the Good”, composed of halves we do not know how to fit together. The top-down ethical concept of “oneself” cannot be appended directly directly to the bottom-up formal “knowledge/science”. At least one of these concepts – self or knowledge – must leave its certainty and be fitted to play by the rules of a different game. The choice is neither neutral nor arbitrary. Since it is selfhood and not knowledge which is precisely on the side of certainty rather than of truth, it is this concept which must be led out of itself in order to find a meaning for itself. Now, in the rule-bound contexts in which we know how to speak of knowledge, we do not know how to speak of subjects or selves, while in the thick, ethical, (rather: complete) ordinary-language contexts in which self-identifications are used, the limitation of scope which makes possible scientific rules like “the knower must be other than the known” is absent. (Cf. the question between Althusser and Lacan, between Badiou and Badiou: Is there a subject of science?) In an apparently-very-different setting, we have the treatment of the same problem in the Tractatus in perhaps its most telling analysis – there is no relation between a person and a proposition. (I suggest reading this is in the sense of il n’ya pas de rapport…) Later the clarification of this nonrelation will lead us to have to reunite the problems of reflexivity and of incommensurability. For now, if Critias wants to claim episteme, he has to translate the self into epistemic terms, which means clarifying the form of self-reference itself. This great experiment is not, as many commentators have suggested, a simple fallacy which can be pinned on the first step from episteme heautou to episteme heautes, but is rather the taking up of a task obviously of great interest to Plato, who is perhaps the first to recognize its extreme difficulty.
(b) ἐπιστήμη ἑαυτῆς / knowledge of itself (166c)
Note: An incomplete expression, transitional under analytic pressure, but by no means simply fallacious. If there is to be a subject, that is, if there is to be a self in two places with respect to knowledge, something that knows and something that is known, then knowledge must in turn be in two places relative to itself.
Contra Brentano, this would seem to be the real discovery of the problem of intentionality and of the for-itself.
The transition is easily accomplished from the side of structure. The structural requirements for episteme being episteme – its self-sufficiency, power, and effectiveness require the expansion of its formulation and of its scope, just as in the apparently-much-simpler Ion. Ion would like his knowledge to be knowledge, i.e. something meeting structural requirements, but he would also like it to be of one object only, Homer, who happens unfortunately to individuate himself by being not just a paradigmatic member of a class of objects, but also a subject who knows and fails to know. That Critias recognizes that the problem is no easier in the first person case is enormously to his credit.
(c) …τῶν τε ἄλλων ἐπιστημῶν / …and the other knowledges (166c)
Likewise. Accepting that the input of sophrosune is knowledge per se does not have the effect of narrowing its domain but of expanding it.
(d) …καὶ ἀνεπιστημοσύνης / …and non-knowledge (166e)
Likewise. In fact knowledge of ignorance or non-knowledge (anepistemosune) is easily seen to be the only real advance (apart from the dubious advance of its centralization) that metaknowledge would claim to make over first-order knowledge.
(6) τὸ εἰδέναι ἅ τε οἶδεν καὶ ἃ μὴ οἶδεν / to know [that one knows] what one knows and [that one does not know] what one does not know (167a)
Note: Critias’ “third” attempt, by Socrates’ count, with 4 & 5. The complete expression for reflexive knowledge has been arrived at with inclusion of the relations to other knowledges, non-knowledge, and finally both intentional objects. The dilemma of reflection can now be posed. (q.v.)
(7) ἧι τὸ ἀγαθόν καὶ τὸ κακόν / [that knowledge by which are known] the good and the bad (174b)
Responds to the reiteration of the τίνος-question at 173d as “one little thing” (smikron): Proximately, responds to the alternative form ᾗ τί; (lit. “by which [instrumental dative] [implied: knowledge] what [interrogative accusative] [implied: is known]?”)
Spoiler: The response is circular: ὦ μιαρέ…πάλαι με περιέλκεις κύκλῳ / Wretch!…you have all this time been dragging me around in a circle. Note the pun now on the circular sense of peri in peri-elkeis – instead of telling me what it is about, providing the referent, you have been dragging me about, that is, leading me pointlessly around a circle of sense, or of nonsense, which is more likely when the aboutness relation itself goes in a circle! (The verb appears just one other time in the Platonic corpus in the Protagoras, in the context of a related ethical problem pertaining to episteme, namely the problem of akrasia: Many, perhaps most, people feel about knowledge (peri tes epistemes) the same way as “about a slave, that it may be dragged about (perielkomenes) by any other force.” (352c) The accusation of circularity is just the one repeated at Republic 505: The Good is knowledge of (knowledge of (knowledge of (knowledge of….))) the Good.*
Closes elenctic sequence: the reflective gambit has not evaded the τίνος-question. We have attempted to avoid the contingency or exteriority that this question indexes through the assertion of an absolute knowing, and have failed.
*9/24/2022: The circle will be left. In the absence of a rational exit, any force will do the job.