Abstract

The term "language-game" recurs in Wittgenstein's later writings. Wittgenstein's overall concept of a language-game is not immediately clear, however, even upon a close examination of the Philosophical Investigations. This unclarity is made further manifest by the conflicting interpretations of what a language-game is which appear in the secondary literature. ;First I examine the historical and conceptual origins of Wittgenstein's use of "language-game". Tracing the development of Wittgenstein's philosophical interest in games, I find two chief sources of the term: the chess analogy as proposed by the game-formalists in the philosophy of mathematics and the apparent importance of play for a child's acquisition of language. ;Second, I use the extant confusion in the secondary literature as a guide to discovering what proves to be the genuine problem in understanding Wittgenstein's talk of language-games. By uncovering assumptions about what a theory of language should be and about what a philosophical heuristic must be, I discover an underlying order in those widely varying interpretations. Many interpreters share the assumption of an exclusive disjunction between language-games as theoretical constructs belonging to the philosophy of language and language-games as heuristical devices belonging to linguistic philosophy. When this is shown to be a false dilemma, a variety of apparently conflicting interpretations of what a language-game is can be synthesized into a new and more comprehensive interpretation. From this new point of view, many of the seeming unclarities in Wittgenstein's use of "language-game" can be resolved. ;Finally, I delineate the salient general characteristics of Wittgenstein's language-games and of his overall language-game approach to language and philosophy. The ubersichtlich character of language-games is claimed to be of central importance. I argue, contrary to many well-known critics, that Wittgenstein's concept of a language-game involves a form of relativism which, though radical, is not self-defeating