A two dimensional analysis of the future: modal adverbs and speaker’s bias

Whether future morphemes in languages are temporal or modal operators is a central question in the semantics of the future. Most analyses (with the exception of [14]; see its rebuttal in [14]), agree that future morphemes do convey modality, and do not always make reference to future times. We study here the temporal (future time) reading of the future (John will arrive at 5pm), which we call ‘predictive’. The modality of the predictive reading is often assumed to be purely metaphysical (e.g. [13]). Based on novel Greek and Italian data with modal adverbs, we argue that prediction involves both a metaphysical and an epistemic dimension, and we offer an analysis that relies on knowledge of the speaker at the utterance time t0. This knowledge restricts the set of the futures (metaphysical branches) only to reasonable ones (in the sense of [18]). In these branches the prejacent comes out true. We also propose that when predicting the speaker has a degree of bias that the actual world-to-come at t0 will be reasonable. This bias, to our knowledge not observed before, is reflected in the use of adverbs, and although by default it expresses high confidence, it can be quite variable, from very strong to relatively weak– as is evidenced by the wide range of modal adverbs that can co-occur with the future morpheme.