Cratylus (425)

Socrates. No, not we; I said that too hastily. For the ancients gave language its existing composite character; and we, if we are to examine all these matters with scientific ability, must take it to pieces as they put it together and see whether the words, both the earliest and the later, are given systematically or not; for if they are strung together at haphazard, it is a poor, unmethodical performance, my dear Hermogenes.

Hermogenes. By Zeus, Socrates, may be it is.

Socrates. Well, do you believe you could take them to pieces in that way? I do not believe I could.

Hermogenes. Then I am sure I could not.

Socrates. Shall we give up then? Or shall we do the best we can and try to see if we are able to understand even a little about them, and, just as we said to the gods a while ago that we knew nothing about the truth but were guessing at human opinion about them, so now, before we proceed, shall we say to ourselves that if anyone, whether we or someone else, is to make any analysis of names, he will have to analyze them in the way we have described, and we shall have to study them, as the saying is, with all our might? Do you agree, or not?

Hermogenes. Yes, I agree most heartily.

Socrates. It will, I imagine, seem ridiculous that things are made manifest through imitation in letters and syllables; nevertheless it cannot be otherwise. For there is no better theory upon which we can base the truth of the earliest names, unless you think we had better follow the example of the tragic poets, who, when they are in a dilemma, have recourse to the introduction of gods on machines. So we may get out of trouble by saying that the gods gave the earliest names, and therefore they are right. Is that the best theory for us? Or perhaps this one, that we got the earliest names from some foreign folk and the foreigners are more ancient than we are? Or that it is impossible to investigate them because of their antiquity, as is also the case with the foreign words?