Abstract

How does language represent the world it can be used to talk about? Or does it? A negative answer is maintained by one of the main traditions in language theory that includes Frege, Wittgenstein, Heidegger, Quine and Rorty. A test case is offered by the question whether the critical ''mirroring'' relations, especially the notion of truth, are themselves expressible in language. Tarski's negative thesis seemed to close the issue, but dramatic recent developments have decided the issue in favour of the expressibility of truth. At the same time, the "mirroring" relations are not natural ones, but constituted by rule-governed human activities a la Wittgenstein's language games. These relations are nevertheless objective, because they depend only on the rules of these "games", not on the idiosyncrasies of the players. It also turns out that the "truth games" for a language are the same as the language games that give it its meaning in the first place. Thus truth and meaning are intrinsically intertwined.