Are not many thoughts essentially indexical, in a way that adds a distinctive mode of presentation to their referential content?
We need to be careful, however, not to infer the indexicality of thoughts too quickly from the indexicality of their linguistic expression. While linguistic indexicality is common, for understandable reasons, nearly of the thoughts so expressed are name-like rather than character-like. I shall argue that even "perceptually, and spatiotemporally, demonstrative" thoughts are best modelled by the (often temporary) creation of name-like files.
The only definite role for character-like thoughts is at the interface between cognition and basic actions. The action-production system identifies objects and places using repeatable character-like referring terms. But precisely because these terms are repeatable they play no lasting role in thought.
By properly understanding the guidance of basic actions, we can both dissolve the so-called "interface problem" and cast light on the nature of conscious sensory experience.
