RULES, CRITERIA, AND TRUTH CONDITIONS: AN ESSAY IN THE PHILOSOPHY OF LANGUAGE (WITTGENSTEIN)

Although there must be rules for language there is a philosophical problem about understanding them. That problem is that anything could be made out to accord with any expression of a rule. Kripke has wrongly concluded from this that one does not know which of an indefinite number of rules he is following and so what he means. This conclusion allows that there are rules independent of rule following behavior. I argue that such a rule is not one that speakers could either understand or follow. The problem is not that it can not be known what rule is being followed but that there seems to be nothing that correctly understanding a rule can come to. Whether it is followed one way rather than another seems quite arbitrary.;An initially more promising view is that there are criteria for the application of rules. However, I find that this can not be rendered coherent on any of the traditional accounts of criteria. Moreover, I find that on none of those accounts can the concept of a criterion have any role in an account of a speaker's understanding of his language. They lead either to an unacceptable reductionism; or to its being the case that the meaning of expressions inevitably escapes what they are understood to mean; or they provide us with something other than meaning (symptoms). Received views of criteria also fail to allow that they are for the truth of a claim as well as its sense. I argue that criteria must be for truth as well as sense. If they were not our understanding both of truth and of what it is to make an assertion would be undermined.;I suggest a view of criteria and their relation to rules which accommodates all these various strains and difficulties.;Current rationalist and empiricist views attempt to account for understanding by explaining the ability to utter and understand potentially infinitely many sentences. What understanding comes to is a function of what is needed to explain understanding. According to one what is needed is an innate mental structure, according to the other, knowledge of the truth conditions of sentences. I argue that each response fails to appreciate the reach of the original problem, and that explanation of the desired sort can not be had.