Diehl's research focuses on the perception and production of speech sounds and on auditory category learning. Various methods are used including: acoustic analysis of natural speech, perceptual identification and discrimination of synthetic speech stimuli and of analogous non-speech stimuli, computational modeling of the representation of speech sounds in the auditory nerve and estimation of auditory distances among sounds, simulations of preferred speech sound inventories using a criterion of maximal auditory distance. We are currently applying Bayesian statistical decision theory in the analysis of tasks that involve learning of novel auditory categories (including both non-speech categories and second language sound categories) as well as tasks that involve recognition of sounds from one's first language.