Socrates asks the young Theaetetus: what is knowledge? He says he really wants to know. Theaetetus hesitates a bit (he’s heard rumors about this guy), but answers in the end: Knowledge of geometry, of astronomy, knowledge of shoemaking and leatherworking—these, and things like them, are knowledge. And let’s be fair to Theaetetus: it’s not as though he’s wrong. Geometry is knowledge, or a knowledge anyway, in one extremely common and useful sense of the word “is.” It’s not an ignorance, certainly, nor is it for example a tree. It is knowledge. He knows his shapes.

Socrates is not satisfied, however. He raises two objections. First, when he said he didn’t know what knowledge was, Theaetetus evidently did not take him at his word: he keeps using the word “knowledge” in his answers, just as if Socrates knew what “knowledge” meant, when Socrates (unless he’s lying, which is likely) well and truly does not know what “knowledge” means. Second, Theaetetus offers a longwinded answer, or rather an indefinite one—presumably there is a knowledge of every fish in the sea, and we could spend our lives listing them off and never come to the end of it, but surely there’s a simple way to say what all the things share that we call knowledge. Surely in some sense they’re all alike. We don’t call them all knowledge by consulting some monstrous heterogeneous list.

Socrates gives the example of mud. You can do all sorts of things with mud—jump in it, wipe it off your feet, make bricks. But saying all the uses of mud would be neither necessary nor sufficient to define it. There’s a short way around: mix water and dirt, you’ve got mud. So (he seems to imply), what do you mix to get knowledge?

Comment 1: You can’t use a word in its own definition; everyone knows that! Theaetetus has made an amateur’s mistake. But why not, exactly? Well, the game would end very quickly if you could. “Those very things we call knowledge, those are knowledge”; that glaring tautology would be the final word. And yet it’s an open question (a question forced open by Socrates’ (feigned?) ignorance, a question which has never quite snapped shut since, the question which “concept” and “essence” and “definition” all seek to close) whether there’s anything more to say on the matter. Knowledge is what we call knowledge; maybe that’s it. This way of ceasing to read Plato comes in two very different forms and is uttered by two very different sorts of people.

First, there’s “that’s it” by fiat, the nominalist thesis: words are human inventions, we use them as we see fit and we’ll change their meanings, too, when and if we deem necessary. Words are our servants and they’ll mean what we want them to mean. This position, I think, is indefensible, an unwarranted extrapolation from the most trivial of cases. We can of course stipulate that any and all boats less than ten feet in length are “skiffs” and clear up any ambiguity there, but “knowledge,” say, or “justice”? Very quickly we’ll find ourselves using words like “truth” and “fairness” and “good” and “correct,” words that exceed our control to just the same degree as “knowledge” did. We can revise the meaning of any given word and agree upon it, but why bother? Our mastery of language will never go all the way down, indeed it barely scratches the surface. To define “knowledge” by referring to “truth” or “science” or “certainty” is to play a shell game. The act of making a word mean something is itself verbose, and you can’t get to the bottom of language by piling words upon words.

Second, there’s “that’s it” uttered by the mystic. Language transcends us absolutely: we cannot see the light by which we see. To recognize truth when it confronts us is a blessing; to ask it to prove its truth by signs (and those signs in turn by further signs) is foolish and perverse. In calling something knowledge, we listen to language and it speaks through us; we can mishear, or we can hear rightly, but we can’t comprehend the authority of the speaker any more than we can challenge it. To know the things that are and call them by their names—that’s it, that’s all.

Comment 2: But it’s not true, is it, that we can’t use a word in its definition: after all, we can define a word only if we know what it means. Consider the act of definition closely: one calls to mind all the many things which are called by that word to mind, and then one takes a stab at naming their sharing. The calling together presupposes knowing the word, knowing the very particular many it addresses. And that’s not the end of it either, because definition is precisely not by fiat: having taken a stab at pinning down the oneness, one returns to the many to check whether in fact they’re all there. “To face one’s death bravely is courage.” “Ah, but surely she too was courageous, though she didn’t face her own death?” Return to the drawing board, take another stab, get it “right” this time. Not that one’s instinct that she really was courageous possesses absolute authority: maybe it will turn out that that wasn’t quite the word after all. But the definition does test itself against the word. The word’s “meaning” multiplicity is the source of definition and the test of its adequacy; the circle is closed.

Definition isn’t only the business of philosophers, though it is their characteristic lifestyle choice. All deep gossip, for example, tends toward definition. “So she has friends you don’t know and you’re jealous?” “Not jealous, exactly, it’s more like...” We embark on definitions all the time, attempting to get clear. And that this is a road to clarity, that’s what’s strange. Because it’s a circle, after all—you define because you already know, you test your definition against this meaning that you already know. How can learning arise from this circle? How does the fixed point of tautology, “those very things that we call x, they are x,” expand into the circle of thought?

The question of definition is in fact the question of thought. It’s obvious that there are other ways to learn things, I mean other than thought: someone can tell you something you didn’t know, and then you know it. It’s much more surprising that you can learn by thinking, that the circular path of inward consideration can open onto unexpected vistas sometimes.

We know and yet we do not know what knowledge is. Plato’s name for this relation is forgetting: we know the forms because we have seen them, before, but in being born we forgot them. Birth is the plunge out of clarity, into unknowing, but we were there once. “There” being eternity. What we don’t know in the present we do know, eternally. It’s a delicate question what this changes—to say, instead of “we don’t know,” “we have forgotten.” Socrates says—as a man whose lover used to play the harp will remember the lover when he sees the harp. Like a sleeper awakening, the thinker comes back to himself.

Comment 3: How does the tautology expand into the circle of thought? For Theaetetus the answer is Socrates, Socrates’ innocent smile, his ignorance-game. Socrates blocks the way of tautology and forces Theaetetus to think. In one sense we know, in one sense we don’t: in one sense, Theaetetus, in the other Socrates. Thought is “speech with oneself,” an inward conversation, Socrates will later claim—but if we take the conversation before us as a model of thought, the partners in conversation are not symmetric in any sense. They don’t “exchange ideas.” Socrates is ignorance which knows itself, Theaetetus is knowledge not knowing itself. Socrates causes Theaetetus to remember that he has forgotten, but not what he knows; he makes the tautology strange; he is thought embodied, nothing but trouble, the troubled movement of thought.

Socrates supplements this with a myth: like his mother before him, Socrates says, he is a midwife. Because Artemis goddess of midwives is barren, only barren women, and those too old to bear, may be midwives. Socrates is barren of ideas. And yet, like a midwife, he can recognize when someone is pregnant with ideas; like a midwife, he can bring on the labor pains with potions and incantations, and soothe them too if need be. Another skill he has, beyond that of the ordinary midwife: unlike women bearing children, men bearing ideas sometimes produce phantoms, false ideas. Socrates, though himself without knowledge, can recognize these phantoms and cast them away. Often the proud fathers are angry with Socrates for this; no one likes to see his children killed. Nonetheless it must be done. Socrates is not empowered by the gods to countenance falsehood, any more than to bear truth.

Theaetetus-Socrates is the thinking-machine. Theaetetus pumps out babies, generates speeches; Socrates is their judge and executioner, the bringer of labor-pains and the killer of very small ghosts. The mechanism of the machine is a mystery: how can Socrates, though ignorant, recognize knowledge? Who but the one who knows could do that? And yet—if to test the truth of a claim required that we already knew its truth, learning would be impossible. The thought-process would begin and end in knowledge, indeed it would not advance a single step. To argue that the test of truth or falsehood is simply the way things really are is to pursue a vicious circle or a mystery: by what sign do we recognize the real?

The remainder of the dialogue pursues this question, seeking a concept of knowledge which grounds the possibility of thought. Thought presupposes that truth is neither everywhere present, after Protagoras’ claim (if we already knew we would not think) nor absolutely absent, as the skeptics might say (if we could never know we could never think). As to a definition of knowledge, the dialogue comes up empty: at the end of the Theaetetus, the ground is littered with the corpses of infants, and Socrates must go off to appear in court, to answer the charge of being nothing but trouble, to meet his death. And yet the possibility of aporia here, the impossibility of bringing the sense in which we know into harmony with the sense in which we are ignorant, demonstrates indirectly and inductively the actuality of thought. The conversation continues; there’s something more to say. Maybe that’s all there is to say, on thought and knowledge; maybe that’s it.

---