Evidence and Bias: The Case of the Evidential Future in Italian

Evidential markers encode the source of information that an individual (the evidential Origo) has for a proposition. In root declaratives, the Origo is always the speaker (see Korotkova 2016 and references therein). Instead, questions often display interrogative flip: the Origo shifts to the hearer (Garrett 2001; Speas & Tenny 2003, a.o.). While interrogative flip is widely attested across languages, some evidentials have been reported not to flip in questions (see, e.g., San Roque, Floyd & Norcliffe 2017; Bhadra 2017). What determines whether evidentials flip or not? Recent work (Korotkova 2016; Bhadra 2017) has proposed that there is a correlation between lack of flip and bias in questions. This paper contributes to our understanding of the interaction of evidentials and bias by investigating the behaviour of questions with the Italian non-predictive future. We characterize the non-predictive future as an inferential evidential marker (see also Mari 2009; Eckardt & Beltrama forthcoming), and show that lack of flip for the future correlates only with a particular type of bias: a reversal of the default bias associated with negative polar questions (Frana & Rawlins forthcoming). We trace back this pattern to an interaction between the evidential component of the future and the operator that triggers bias reversal.