I am a psycholinguist interested in the interplay between the form of language and the processes by which language is produced, understood, learned, and lost.
Most of the work in my lab is based on spontaneous speech produced under naturalistic conditions. Sometimes, we complement that evidence with "scripted" speech produced in the lab.
Why does pronunciation vary? For example, why are low-frequency words often spoken more slowly than high-frequency words? Intuitively, there are at least two mutually compatible explanations: production difficulty and intelligibility: For example, low-frequency words may be lengthened because they are relatively difficulty to retrieve from the mental lexicon and to produce, or because speakers anticipate that listeners will find these words difficulty to understand. In other words, pronunciation variation may reflect speaker-internal and listener-oriented processes.
Understanding the respective role of listeners and speakers is complicated by the fact that comprehension and production difficulty often go hand in hand: The same factors that make some words easier to say than others also make those words easier to understand. To understand the place of listeners’ and speakers’ needs in pronunciation variation better, we have been investigating a property of words that impedes comprehension, yet facilitates production: Phonological neighborhood density (Yao, Gahl & Johnson).