Laura Dilley is an Associate Professor in the Department of Communicative Sciences and Disorders. She received her B.S. in Brain and Cognitive Sciences in 1997 from MIT and her Ph.D. in Speech and Hearing Biosciences and Technology from MIT and Harvard in 2005. Her research has covered a variety of areas, including: the role of prosody and prediction in speech perception and linguistic understanding; language and speech development in neurotypical and atypical populations, such as adults and children with typical hearing and hearing loss/disability; and social identity and language use. In 2011 she won the Outstanding Faculty/Staff Award and Michigan State University. Further, her work was recently highlighted in the introduction to a special issue of Topics in Cognitive Science honoring Michael K. Tanenhaus for his receipt of the 2018 Rumelhart prize; a paper in the special issue, along with several other collaborative papers with Dr. Tanenhaus since around 2015, have all featured her research paradigms.
Language understanding, prosody, and segments: Role of the predictive brain
The past half century of linguistic research has seen dramatic changes in the way researchers frame and conceptualize language as a human capacity. In this talk I will present a synthesis of insights from recent decades and argue that language is the outgrowth of perception, action, and cognition. Language perception does not entail mere recovery of abstract linguistic units; instead, social and ecological contexts shape how linguistic units are understood to be composed, and how meanings are apprehended.
Prosody has long been held to be a mere overlay on implicitly foundational segmental underpinnings of sentences. Contrasting with this view, I will present an overview of experimental work from my lab which shows that, perhaps surprisingly, distal context prosody can alter perception of lexical and segmental content, as well as meaning. For example, changing only prosody in initial distal portions of utterances may cause listeners to hear a word string as ending, alternatively, with timer derby or tie murder bee, or a sentence to contain either saw a raccoon or else saw raccoons. Such traditionally unexpected findings highlight the role of predictive brain mechanisms tuned to language-relevant timescales in apprehending – emergently – the meaning, form, and content. Arriving at new insights regarding such effects requires challenging some long-held assumptions about the synchronic relationship of prosody to words and meanings.
I argue that viewing linguistic capacities as grounded in temporal dynamics of social brains permits new insights into some of the most historically challenging linguistic problems, including perception of speech, apprehension of meaning, and ontology of language, while also fostering novel connections among linguistic sub-disciplines.