Abstract
We frequently rely on the circumstances of our utterance when making recognisable to others what object our utterance is about. We rely, for instance, on spatial proximity to an object. A suitably close object can be referred to in a way that exploits that spatial relation: it can be referred to with the help of demonstratives. The purpose of such expressions is precisely to accomplish that feat: they allow the speaker to exploit the circumstances of her utterance when making recognisable to others what she is talking about. Our ways of expressing content thus draw - in such, and other, ways - on the circumstances in which that expressing occurs. But does that mean that what one thereby expresses also draws on, and has a similar connection to, those circumstances? Is it only we speakers who point to objects when making recognisable what we are talking about - or does what we make recognisable also point to an object?
Affirmative answers to this last question have been particularly popular in recent decades. And, yet, there are reasons to think that the answer had better not be affirmative. This will, at any rate, be my first contention. If truth is to be public, thoughts cannot be the kind of thing that has a location - and cannot therefore be the kind of thing that points. They can, that is, contain no 'essentially indexical' components. My second contention will be about one of the main reasons for appealing to such thought components. This appeal has widely been taken to be required to account for certain transitions between thought and action. Nothing short of such thought components can - it is widely held - account for the way in which recognition of particular situations bear on a subject's conduct. Nothing short of a certain 'I-thought' can, for instance, account for the realization that not just anyone, but I am spilling sugar - and nothing short of such a thought can thereby move me to inspect the sugar package in my trolley. But appeal to such kinds of thoughts is only required on a certain view of the individuation of thought. It is required on a view according to which the mere conceivability of a difference between thoughts is sufficient to (always) establish such a difference. This is, however, a mistake. Or so I shall argue.
