TACIT REFERENCE Some remarks on unarticulated parameters

John Perry introduced the idea of thought without representation. At first sight, this idea seems odd: we are surely used to consider that a thought about an object is a representation of that object. However, Perry's idea describes an important and pervasive phenomenon in language and thought. The study of this phenomenon highlights unexplored ways in which context contributes to the interpretation of our statements and beliefs, and points to central facts about our cognitive architecture. The present paper contains only preliminary remarks on Perry's important paper "Thought Without Representation" (1993), and certainly not a definite assessment of his complex view. It is structured as follows. In the first section, I distinguish (following Perry himself) two cases of thought without representation. In the second section, I explore the possibility that some supposed instances of thought without representation really involve something like simplified views of the world. This possibility suggests that thought without representation is in fact simplified thought, which does not rely on unarticulated parameters. In section 3, I argue that this possibility is not obviously wrong, although it does not imply that the notion of thought without representation is useless. In section 4, I describe in more details the move from tacit to articulated reference, which is nothing like an ordinary inference. In the last section, I briefly compare my account with Recanati's.