Abstract

Commonsense psychological explanation of action upon objects seems to require not only reference to agents’ demonstrative beliefs about the objects acted upon but also the de re ascription of these demonstrative beliefs. There is an influential objection, however, to the de re component: since de re ascriptions permit the attribution to agents of inconsistent attitudes about the objects acted upon, they cannot explain (or predict) agents’ actions upon those objects. This paper answers the objection by presenting a contextualist theory of de re action explanation according to which agents’ beliefs about objects are logically fragmented.