I consider Plato’s argument, in the dialogue Cratylus, against both of two opposed views of the “correctness of names.” The first is a conventionalist view, according to which this relationship is arbitrary, the product of a free inaugural decision made at the moment of the first institution of names. The second is a naturalist view, according to which the correctness of names is initially fixed and subsequently maintained by some kind of natural assignment, rooted in the things themselves. I argue (...) that: 1) Plato’s critical challenge to both views anticipates considerations introduced by Wittgenstein in the Philosophical Investigations ’ consideration of rules and rule-following; 2) Understanding Plato’s appeal to the “form” [ eidos ] of a thing in resolving the problems of both views helps to explicate Wittgenstein’s own appeal to “forms of life” as the “given” ground of linguistic practice; and 3) We should not understand the grounding of language in form-of-life either as a basis in the plural practices of different communities, or as a biological/anthropological basis in the specific nature of the human organism. Rather, it points to an autonomous dimension of form, which articulates the relationship between language and life as it relates to the possibility of truth. (shrink)