Encoding speaker bias: The semantics of 'wh'-polar questions

Abstract

This study examines the semantics of wh-polar questions (wh-PQs, henceforth), interrogatives where the wh-phrase is used non-canonically to convey speaker’s negative bias. Based on novel data from Korean, I first show that wh-PQs do not require contradictory or adversative licensing conditions, diverging from previous cross-linguistic observations. It will then be shown that wh-PQs convey speaker’s negative bias toward the prejacent p, rather than asserting ¬p. I propose that the wh-phrase in a wh-PQ is not associated with a sentential constituent and that wh-PQs have the semantics of biased polar questions, where the wh-phrase functions as a polar question modifier (i.e. wh + p?). The wh-phrase is argued to encode the two hallmarks of the wh-PQs: (i) obligatory reactiveness and (ii) speaker’s negative bias toward the prejacent p