Phonetic recalibration of visual and auditory speech by visual sentences

Abstract

Phonetic recalibration is a form of perceptual learning in which experience with an ambiguous speech segment accompanied by disambiguating context biases subsequent perception of that segment. For example, presenting an auditory "f" final word (i.e., “Witlof”; Dutch for “Chicory”) with the final fricative replaced with a sound between /s/ and /f/ results in participants subsequently categorizing more items from a /f/-/s/ continuum as /f/ (Norris, McQueen, Cutler 2003). A similar effect is found for ambiguous visual speech that is accompanied by clear auditory speech (Baart & Vroomen 2010), and for ambiguous auditory speech that is accompanied by clear visual speech (e.g., Vroomen & Baart, 2009). It is unknown whether visual speech alone can recalibrate visual speech segments or whether recalibration in one modality can be transferred to another modality (but see, Dias & Rosenblum 2016). In Experiment 1 of this study we found that sentence context disambiguated a silent visual ambiguous /s/f/ fricative and resulted in visual recalibration. Experiment 2 used these same silent sentences but measured recalibration on an auditory continuum. Preliminary results suggest that recalibration effects can be transferred across modalities in this way. These experiments have implications for theories of amodal perceptual learning.