Language Perception Laboratory

Research Areas

Spoken Word Recognition

Research in the speech lab addresses the following question: How does the mind of a listener translate the acoustic energy emanating from a talker's mouth into the words intended by the talker? Projects focus on how listeners' knowledge of their language (phonological, lexical) and the surrounding sentence context (semantic, prosodic) are used to aid recognition. Most recently, we have been examining how listeners recognize phonological variants (i.e., alternative pronunciations) of spoken words. Acoustic and phonological analyses of speech in the Buckeye Speech Corpus are used to inform these experiments.

Mathematical Modeling

With the development of mathematical models of language and other mental processes, there has arisen a need for tools with which to evaluate and understand the behaviors of these models. This work, done in collaboration with Jay Myung, focuses on developing and applying such techniques. Bayesian methods figure prominently in our approach. Most recently, the use of sampling methods such as MCMC have made it possible to analyze the performance of highly nonlinear models, including localist and distributed connectionist models.

Graduate Study in Psychology

Graduate study in the Language Sciences