I defend semantic holism, the view that the meaning of an expression is determined by its relations to every other expression in the language of individual competent users. I argue that, once properly understood, most disadvantages attributed to holism can be dissolved and suggest that the core division between the holist and the non-holist is on the question whether invariant meanings shared across all possible occasions where the corresponded expressions are uttered are necessary for the explanation of meaning sharing. I give reason why the answer is negative and demonstrate how to explain our linguistic interaction without such invariant meanings. |