Abstract

This dissertation is a rehabilitation of Locke's theory of language. I begin with a discussion of the relations of signification and reference, showing how Locke handles Mill's objection concerning the reference of words to ideas. Locke's theory allows that language may be used to describe the world by means of the composition of the relation of signification with the relation of representation that holds between ideas and things in the world. ;Next I deal with the objection, raised by Frege, that Locke's theory makes semantics, and language use itself, impossible because it places meaning in the mind of the language user, destroying any basis for intersubjectivity. This objection ignores the role of sense experience in Locke's theory. Some of our ideas are known to be similar because they are derived passively from sense. ;Next I consider general terms and Berkeley's objection that Locke's abstract ideas are either inconsistent or contentless. I consider two strands of thought concerning abstraction in the Essay. I conclude that Berkeley's objections gain much of their force from considering abstract ideas as images, but that Locke's theory does not demand such an interpretation. I also consider the underlying motivations for, and the true extent of, Locke's "nominalism". ;Finally, I argue that Locke's theory is akin to modern representational theories, such as those of Fodor and Kamp, both of whom argue that in order for mental representations to have a role in practical reasoning, they must admit of some syntactic structure. I consider the extent to which Locke's ideas admit of structure, and conclude by arguing that a theory of truth may be given for a Lockean account of language in a similar fashion to that of such modern theories, where one links the word to the mental representation, and then gives a semantics for ideas