Self-comprehension as a window on perception-production relationships

Imagine working for a major American company with branches around the world. You are on a conference call with speakers from multiple countries with different native languages. Even though everyone is speaking the same language, misunderstandings often happen.

Given that the common language is often not the native language for all speakers, the different accents may lead to communication issues. Many people assume that these misunderstandings happen because speakers hear sounds quite similarly to native speakers but have difficulty producing the speech sounds like a native speaker (i.e., their speech is “accented”), perhaps due to ingrained motor habits and/or less experience producing the sounds.

But another possibility is that nonnative speakers produce some sounds differently from most native speakers because they actually hear those sounds differently. This research project seeks to understand how a particular speaker’s perception of speech sounds (such as the vowel sound in “ship” vs. “shape”) is related to that speaker’s production of the sounds.

This is a particularly relevant question for second-language speakers, who often have difficulties with both perception and production. It also has ramifications for native listeners, who often have difficulties comprehending speech from second-language speakers or speakers of less-familiar dialects or with less-familiar accents. Better communication among speakers of a common language who have different language backgrounds and accents would benefit instructional contexts with international students, medical care settings, and extension of American businesses into new markets.

This project uses a novel research paradigm to test subtle aspects of how people perceive their own speech. Speakers for whom English is their first (or only) language and speakers who learned English as a second language will be recorded naming a set of pictures. Each participant will then hear their own or someone else’s recording, which they will need to identify.

For example, they may hear their own (or someone else’s) recording of “ship” and must select the intended word from one of four pictures: ship, shape, log, and lock. If perception is a source of communication difficulties, speakers of English as a second language should be relatively better at understanding their own speech (more accurate and faster at choosing the intended word) than the speech of others because it will match their mental representations.

If they speak inaccurately because of motor difficulty, then they will find their own speech difficult to understand, compared with their understanding of native speech. Indeed, this pattern may also hold for native speakers, albeit at a reduced level, providing more general insights into the links between representations for speech perception and production.

This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.