What’s wrong with the truth-conditional account of slurs

The aim of this paper is to provide arguments based on linguistic evidence that discard a truth-conditional analysis of slurs (TCA) and pave the way for more promising approaches. We consider Hom and May’s version of TCA, according to which the derogatory content of slurs is part of their truth-conditional meaning such that, when slurs are embedded under semantic operators such as negation, there is no derogatory content that projects out of the embedding. In order to support this view, Hom and May make two moves: (i) they point to cases where it looks like projection does not occur and (ii) they try to explain away cases where projection seems to occur by appealing to a pragmatic phenomenon that they call ‘offense’. Pace Hom and May, we argue that the derogatory content of slurs does in fact project and, in advocating for our claim, (i) we show that those cases where it looks like projection does not occur are in fact metalinguistic uses in which slurs are not really used, by relying on three linguistic tests (Section 3); and (ii) we refute Hom and May’s attempt to explain why speakers would entertain the supposedly wrong intuition that the derogatory content of slurs projects out of semantic embedding, by focusing on the case of slurs for fictional entities (Section 4). We conclude that Hom and May’s strategies to support TCA ultimately fail.