The Cratylus is a dialogue about names, and starts with the question of whether names have an essential, or simply a conventional, relationship to the things they name. Cratylus is arguing for the former and Hermogenes for the latter proposition; they put the question to Socrates.
As you might predict, Socrates comes down (at first, anyway), on the side of Cratylus. That names are purely arbitrary or conventional smacks of Protagoras’ position that “man is the measure of all things” (386a); following out the implications of this proposition, there can be no such thing as a wise or foolish man. As all three are pretty sure there are such things, they reject Protagoras and thus, it seems, must accept the competing notion that names do, in fact, have some connection to the essence of things.
In fact Socrates is somewhat cagey about his own views. Once they’ve gone through the preliminary argument and Hermogenes has seen that it leads to the opposite conclusion that he started with, he says “it is not easy to change my conviction so suddenly. I think you would be more likely to convince me, if you were to show me just what it is that you say is the natural correctness of names” and Socrates responds “I, my dear Hermogenes, do not say that there is any. You forget what I said a while ago, that I did not know, but would join you in looking for the truth” (391a).
A lengthy and sometimes quite entertaining discussion follows during which Socrates proposes etymologies for dozens of words, connecting them to their essences. We learn that those who gave the first names thought, like Heracleitus, that everything was in flux, and therefore that motion (phoras) and flowing (rhou or rhein) are ideas that enter into many words: “Wisdom (phronesis)… is perception (noesis) of motion (phoras) and flowing (rhou) or it might be understood as benefit (onesis) of motion (phoras)” (411d).
Hermogenes naturally asks, after a number of these, where the orginial elements come from, and we’re treated to another entertaining discussion of the values of the phonemes themselves — the letter rho, for instance, is connected with flowing and movement because “the tongue is least at rest and most agitated in pronouncing this letter” (426e).
Throughout all of this we’re cued not to take any of it very seriously in two ways: first, because Socrates is clearly being so flippant in much of it (with many expressions of amazement at his own wisdom: his analysis of phonemes is prefaced with his own sense that his ideas on the subject are “outrageous and ridiculous” 426b). A little more substantively there is that explanation at the outset about the “lawgivers” who made the names being convinced that everything is in flux. This is a doctrine which is rejected elsewhere, and which Socrates does indeed come back to at the end of the dialogue and argue is mistaken.
The lesson seems to be that we should look not to names, but to the truth, as our source of learning (439a-b). This is of course comfortingly Socratic, but also familiarly frustrating — clearly we have no tool other than language to use as we investigate, so even if the direction of our investigation shifts from trying to determine how the word “virtue” is like virtue, we are still ultimately bound to words as we stumble along toward understanding.