Donald Davidson's account of interpretation purports to be a priori , though I argue that the empirical facts about interpretation, theory of mind, and autism must be considered when examining the merits of Davidson's view. Developmental psychologists have made plausible claims about the existence of some people with autism who use language but who are unable to interpret the minds of others. This empirical claim undermines Davidson's theoretical claims that all speakers must be interpreters of other speakers and that one need not be a speaker in order to be a thinker. The falsity of these theses has consequences for other parts of Davidson's world-view; for example, it undermines his argument against animal thought.