All representations have contents. Even representations of things that donâ€™t exist are meaningfully described as â€œrepresentations of somethingâ€. But not all representations are representations of themselves. Compare, for instance, the sentences â€œThe cat is on the matâ€ and â€œThis sentence has seven words in itâ€. Compare also, the thought that â€œCherries grow on treesâ€ and â€œI like thinking about thinkingâ€.
While there may be a sense in which a representation â€˜s content is a property of the representation, representing the content doesnâ€™t suffice for representing the representation. All representations represent their contents, since their contents just are what they represent. But not all representations are representations of themselves. So whatever does suffice for representing a representation, it cannot simply be representing its content. If all properties of representations that arenâ€™t content properties are vehicular properties, then whatever does suffice for representing a representation, it must include representing its vehicular properties.
States in which we are conscious of something bear sufficient similarities to representation to warrant postulating that such states are implemented by mental representations. Whatâ€™s postulated, then, is that being conscious of something is just a certain kind of mentally representing something â€“that the content of consciousness just is a kind of representational content. It wonâ€™t follow from this implementation story without further argument, though, that all states in which we are conscious of something are automatically states in virtue of which we are conscious of those states. Nor will it follow without further argument that simply by being conscious of the stateâ€™s content are we thereby conscious of the state itself.