Plato’s (c. 427–347 BC) Cratylus is a dialogue about one question: is there a “correctness” to the names of things? In Greek the question is called the “correctness of a name” (ὀνόματος ὀρθότης). It comes about eighty years before Aristotle, some 2,200 years before Humboldt, 2,300 before Saussure. If the West has a sustained philosophy of language, this is where it starts.
There are three speakers. Hermogenes holds that a name is nothing but a convention. Cratylus holds that a name is fixed by nature. And Socrates, standing between them, signs up to neither position wholesale. The dialogue opens as Socrates is drawn into an argument that Hermogenes and Cratylus have already run into the ground.
1. Two positions — physis or nomos
Cratylus’s claim is reported at the opening by Hermogenes:
ὀνόματος ὀρθότητα εἶναι ἑκάστῳ τῶν ὄντων φύσει πεφυκυῖαν
There is a correctness of name that belongs to each thing by nature. (Greek quotations throughout; Japanese and English are L/LAB working translations.)
Hermogenes cannot accept it. For him the correctness of a name is nothing but “convention and agreement” (συνθήκη καὶ ὁμολογία).
οὐ γὰρ φύσει ἑκάστῳ πεφυκέναι ὄνομα οὐδὲν οὐδενί, ἀλλὰ νόμῳ καὶ ἔθει τῶν ἐθισάντων τε καὶ καλούντων.
No name belongs to anything by nature, but by law and by the habit of those who fix the usage and do the calling. (384d)
2. Three parts of the dialogue
Socrates first dismantles Hermogenes’s relativism. Things have a being of their own (οὐσία), and actions have a right way to be done. So “naming” too must have a right way (article 037). From there he builds the view that a name is a tool, and separates the “lawgiver (νομοθέτης)” who makes the tool from the “dialectician (διαλεκτικός)” who uses and judges it (article 038).
In the central section Socrates takes name after name — of gods, virtues, natural things — and breaks them into etymologies. Most of them carry “flow” and “motion” in their sense: the first name-givers, he suggests, built a Heraclitean worldview of universal flux into the names (article 039). And the “primary names,” the ones that cannot be broken down further, are said to be vocal imitations of their objects. ρ is motion, λ is smoothness — the first theory of sound symbolism in the West (article 040).
3. The ending — a name is not the doorway to truth
At the last, Socrates turns on the nature view he had been leaning toward. If the first name-giver got the world wrong, then we who follow the names are deceived. And before the primary names existed, how did the name-giver come to know things at all? So “things must be learned and sought out not from names but far rather from themselves” — knowledge has to sit prior to names (article 041).
οὐκ ἐξ ὀνομάτων ἀλλὰ πολὺ μᾶλλον αὐτὰ ἐξ αὑτῶν καὶ μαθητέον καὶ ζητητέον
Not from names but far rather from the things themselves must one both learn and inquire. (439b)
The dialogue closes by arguing that if everything flows, knowledge cannot stand either. “The beautiful itself” could not even be called “that” if it became something else while we were still speaking. Here is Plato’s step toward the unchanging Forms.
In sum
| Question | How it appears in the Cratylus |
|---|---|
| Is the correctness of names natural or conventional? | Cratylus = physis, Hermogenes = nomos; Socrates demands both |
| What is a name? | a tool (ὄργανον) for “dividing” the being of things — likened to a weaver’s shuttle |
| Who makes names? | the lawgiver (νομοθέτης), supervised by the dialectician |
| What the etymologies say | the worldview of the first name-givers (mostly “flux”) is fossilised in the names |
| How primary names work | imitation of the object through the sound of letters (ρ = motion, λ = smoothness) |
| The dialogue’s conclusion | no truth is learned from names; knowledge lies in “the things themselves,” and they do not change |
Why it is still read
- The prototype of the arbitrary sign — the physis-versus-nomos axis passes straight into Saussure’s “arbitrariness of the sign”. Hermogenes’s position became a founding assumption of modern linguistics, 2,300 years later.
- The starting point of sound symbolism — Socrates’s argument that “ρ expresses motion” is a distant ancestor of today’s research on sound symbolism (the bouba/kiki effect and the like) — and also the opponent Saussure set out to refute.
- Language and worldview — the idea that names carry a theory anticipates Humboldt’s “language = worldview” and the Sapir–Whorf relativity thesis.
- A long prehistory of “meaning is use” — this dialogue’s question about the relation of name to object keeps recurring, in changed form, until Wittgenstein says that the meaning of a word is not the object it points at but its use.
