Phonological and phonetic biases in speech perception

This dissertation investigates how knowledge of phonological generalizations influences speech perception, with a particular focus on evidence that phonological processing is autonomous from (rather than interactive with) auditory processing. A model is proposed in which auditory cue constraints and markedness constraints interact to determine a surface representation, which is taken to be isomorphic to the listener's perceptual response under some psychophysical conditions. Constraint ranking is argued to be stochastic in this model on the basis that the probability of computing the least marked surface representation (and perceptual response) is greater when the input auditory representation is ambiguous between two alternative categories than when it strongly favors a category that completes a more marked surface representation (and perceptual response). Experimental evidence is presented to demonstrate that (1) native listeners of languages with assimilation processes confuse unassimilated and assimilated sequences when discrimination is category-based (but not when discrimination is based on auditory representations), (2) German listeners use phonological context to anticipate the presence of a following allophone iff it is the allophone with broader distribution, and (3) that non-rhotic English listeners perceptually epenthesize and delete /r/ and they also may perceptually undo /r/ deletion. (1) suggests that knowledge of a phonological generalization may be applied only after auditory processing, which is a result consistent with the predictions of 'autonomous theory' and inconsistent with the predictions of 'interactive theory'. (2) and (3) show that phonological effects in speech perception go beyond biases against illicit sequences and lead to the novel proposals that positive constraints (2) and opposite faithfulness constraints (3) exist in the perceptual grammar.