Why quotation is not a semantic phenomenon, and why it calls for a pragmatic theory

In this paper, I argue that quotation is not primarily a linguistic phenomenon. Linguistic productions are essentially symbolic (in Peirce’s sense), whereas quotations, in essence, are iconic. They are therefore best treated as ‘demonstrations’ (Clark & Gerrig 1990; Clark 1996; Recanati 2001). As a consequence, any purely semantic account of the meaning of quotations is bound to be not only incomplete but flawed in key respects. Any sound theory of quotation must be a pragmatic one at heart, with semantics playing only an ancillary role. Since most existing accounts of quotation are fundamentally semantic, they are also necessarily deficient, and therefore unsuitable as general, comprehensive theories of quotation.