Abstract

A new theory of thought is introduced based on a distinction between thought-tokens and thoughts; thought-tokens map many-one to the sentences that express them. What an agent is thinking on a given occasion constitutes her thought-token. Thought-tokens are given expression via a sentence uttered in a public language. Such sentences have determinate standard contents but the thought-tokens they express frequently do not. Moreover, the contents of thought-tokens of various agents may differ significantly, yet our common linguistic practices of thought attribution warrant the use of the same sentence to express them. Consequently, there is a many-one relation between subjective thought-tokens and public sentences which express them. Agents “having the same thought” amounts to no more than that the same sentence may be used to express thought-tokens with different content. We have thought-tokens; we do not have thoughts. The expression ‘thought’ is a useful facon de parler. The implementation of this new theory allows for novel solutions to several problems. I sketch one such application here (several others in my 2015).