(1) Self-knowledge as a secondary sense of knowledge.
(2) Self-knowledge as a promissory sense of knowledge. (I take the term “promissory” from Cora Diamond’s papers on Riddles and Anselm’s Riddle.)
Both senses involve conceptual, logical, difficulties, and are thus different from the matter-of-course sense of self-knowledge—the sense in which I can know my own height, or blood-pressure.
(1) “Self-knowledge” is used as a secondary sense of “knowledge” when the paradoxes that characterize it are not taken to be resolvable—perhaps resolvable in the old language game, or in any new one.
Here are two of those paradoxes: a) To successfully get to know ourselves is, among other things, to establish our own agency. But since self-examination is an exercise of agency, there is something baffling about the possibility of such an activity: an activity which is aimed at establishing for the first time something that is a precondition for its exercise. b) There is a (related) paradox that can be formulated explicitly about an activity that is supposed to be both a kind of creation and a kind of discovery: For there to be discovery, what we discover must already exist; for there to be creation, what we create must not already exist. The use of this term—“self-knowledge”—is logically problematic, and yet, when used in this way, it is not as if we have a better term. ‘We need this term here.’ And the similarity between this and Wittgenstein’s saying that he needs the word “fat” to describe Wednesday relative to Thursday does not seem to me to be a coincidence.
(2) “Self-knowledge” is used as a promissory sense of “knowledge” when the paradoxes that characterize it are taken to be resolvable on a higher level, in a different (although perhaps related) language game which we do not yet have—or do not yet fully have.
We use promissory language elsewhere. We do this for instance in mathematics, when talking of “the next prime after 37 in the series of prime numbers”: Even though we have a strong sense that there must be such a series, we don’t have it; we only have a promise of one. Therefore we don’t have what it takes to refer to the series, and all our references to “it” rely on our sense that there has to be one, and the promise that there is. Likewise, when we talk of self-knowledge in this sense, the implication is that we may intentionally examine ourselves only to an extent: we may only act under the guidance of a promise that we are really doing something—a sense that we must be. Whether we are or not, while we are engaged in this would-be self-examination, remains to be seen. It may turn out, if things go right, that we have been doing something after all. And if it does, this means that we have thereby discovered a new language game—established a new stretch of grammar, a conceptual scheme for a new use of the term “knowledge.”
The reason why I suggest this distinction is that it seems to me that both kinds of uses are important, and both capture things that we care about. In particular, both capture something of the fact that when engaged in self-examination, we are not in full possession of the activity: We are building something, or we are relying on something, which we are not guaranteed to have (perhaps partly on others’ willingness to share the sense that the whole business is worthwhile), but all the while we are exposing ourselves. And in both cases the value of the self-examination goes hand in hand with this exposure: the exposure is what makes the examination humble, as it were. In religious terms, this exposure reveals our dependency on something like divine grace in the very use of our words. Nevertheless, the two uses of “self-knowledge” indicate two different types of actions, and two different ways in which we may be exposed. And this shows a possible ambiguity in our talk of self-knowledge.
Let me say and emphasize that I do not want to claim that if we are to self-examine, we have only the two options I mentioned, or even that these two options are mutually exclusive. Neither seems to me to capture in full all that we want to identify in the grammar of self-examination. Relative to that complicated activity, the two kinds of self-examination I distinguished have, I think, simpler grammar—perhaps more unified grammar. My hope is that they can be useful as objects of comparison. They are not meant to draw attention to themselves, but to help shed some light on the very complicated activity of self-examination.