Scalar Implicatures, communication, and language evolution

This text deals with Scalar Implicatures (SIs). According to the main tenets of Grice, derivation of SIs is a pragmatic phenomenon that occurs at the root of the sentence, in a global fashion. This is the so-called “globalism”. Theorists like Chierchia argue that in order to get specific readings otherwise unavailable SIs can (and in fact must) occur in embedded contexts. According to Chierchia, SIs get computed in parallel with the computation of the semantic value of the sentence. This is the so-called “localism”. But there is also a third approach, supported by Recanati, that shares the account for embeddability with localism, and the pragmatic interpretation of implicatures with globalism. From an evolutionary point of view, I argue against the Recanati’s approach. Recent linguistic literature deals with language evolution. According to Hauser, Chomsky and Fitch, the assertion that language is an adaptation for communication is far too vague to be addressable, and moreover it fails to recognize the distinction between questions of computation and questions of evolution. I argue that Recanati fails to recognize the same distinction, or something very close to it. Therefore, his mixed approach to SIs needs to be modified. Finally, I present some evidence in order to reject the game-theoretical approach to language evolution. © 2010 ontos verlag.