Cratylus

The dialogue does not settle the nature-convention debate but performs a double refutation. Socrates first refutes Hermogenes' conventionalism by arguing that name-giving is a craft, like weaving, which must be performed by an expert with knowledge of the Forms . He then turns on Cratylus' extreme naturalism, using the portrait analogy to show that names, like pictures, can be imperfect likenesses and therefore false . The dialogue rejects both the view that names are mere arbitrary signs and the view that they perfectly embody essences. Its positive claim is methodological: names are tools for instruction, but they cannot be trusted to disclose truth. The dialectician must judge names, not serve them.

The dialogue introduces the distinction between names as tools (organa) and names as likenesses (eikones). The tool analogy, developed in Socrates' argument against Hermogenes, treats a name as an instrument that 'imparts instruction by separating being,' analogous to a shuttle separating threads . The likeness analogy, developed against Cratylus, treats a name as a portrait of its object—necessarily imperfect, since a perfectly accurate portrait would be a duplicate rather than a representation . Together these frameworks establish the terms for subsequent debates about reference, meaning, and the limits of linguistic representation. The Cratylus is the earliest sustained attempt in the Western tradition to think through what it means for language to be 'about' reality.

The long etymological section, comprising roughly two-thirds of the dialogue, is Socrates' demonstration of naturalism at work—an attempt to show that Greek words, properly analyzed, encode true descriptions of their referents . Whether Plato intends this seriously or ironically has been debated since antiquity. Most modern scholars treat the etymologies as playful and deliberately extravagant, designed to display the resources and limits of naturalism rather than to establish a genuine linguistic science. The etymologies consistently reveal a Heraclitean ontology of flux, which Cratylus accepts as confirmation of his position. But Socrates later undermines this acceptance by questioning whether the original name-givers were infallible. The section is thus performative rather than evidential: it shows what naturalism would look like if pushed to its limits, and why that push fails.

The Cratylus ends in aporia, having shown that names alone cannot guarantee truth. The Sophist revisits the problem of falsehood from a different angle, developing a theory of predication that explains how false statements are possible without positing false names. In the Cratylus, Socrates raises the possibility of false naming but does not resolve it; the Sophist supplies the conceptual machinery—the distinction between naming and saying, and the analysis of being and not-being—that the earlier dialogue lacks. Reading them together shows Plato moving from the critique of linguistic naturalism to a constructive account of how language can articulate reality. The Cratylus clears the ground; the Sophist builds on it.

The standard older translation is H. N. Fowler's Loeb edition (1926), which remained the only complete English version for over seventy years. In 1998, C. D. C. Reeve published a new translation with Hackett, described as 'the first new English translation since Fowler's Loeb edition of 1926' and praised for making the tortuous passages both faithful and readable . Benjamin Jowett's nineteenth-century translation remains widely available but tends to smooth over the dialogue's playfulness. The main translation challenges concern the etymologies—which involve puns and sound associations that rarely survive into English—and the precise rendering of key terms like onoma (name), orthotēs (correctness), and physis (nature). Reeve's edition includes an introduction that addresses these difficulties.