Against a unified treatment of obligatory presupposition trigger effects

Abstract

This paper examines bi-sentential sequences where additive presupposition triggers (e.g. too, again) seem to be obligatory in the second sentence. We present linguistic and experimental evidence against treating these obligatory additivity effects as uniformly following from Maximize Presupposition! (Heim 1991). We propose that the environments giving rise to these effects involve a discourse move that corrects for over-restrictive assumptions about the domain in the immediately preceding move. Crucially, the second move must be compatible with the first. General considerations about how the discourse unfolds, in conjunction with a principle that sentences are interpreted exhaustively by default, make it so that two sentences in such sequences are mutually inconsistent in the absence of the additive.