Doctoral Dissertation Research: Experimental priming of phonological variant choices

University Of Pennsylvania, Philadelphia PA

Investigators

Abstract

Language is highly variable such that words can have different pronunciations. For instance, in American English, the word WALKING can be pronounced as either walking or the more casual walkin'. Among the many social and linguistic factors shaping which option people choose, it is known that people tend to repeat the same pronunciation that they have recently used or heard. For example, an English speaker who just said or heard workin' becomes more likely to say playin' a few moments later, repeating the use of the casual -in' pronunciation even on a different word. In this doctoral dissertation project, the investigators use laboratory speech perception and production experiments to ask how basic properties of human cognition lead to this kind of repetitiveness in real-world language use. The results of the project have the long-term potential to help improve the capacity of language technology to handle the socially-meaningful variability that is pervasive in everyday language, to identify typical and atypical language use patterns in clinical settings, and generally to contribute to our scientific and cultural understanding of how humans communicate across a diverse range of language backgrounds. The repetitiveness effect described above, known as variant persistence, has often been attributed to a cognitive process called priming, where processing a particular linguistic form makes it easier to produce the same form again. This dissertation provides experimental evidence for priming variable pronunciations in perception and production through English words whose ending alternates between -ing and -in', connecting data from controlled laboratory experiments with data from informal conversational speech in order to understand the perception and production of socially-situated language variation. A series of laboratory experiments will be conducted to show that variable pronunciations can be primed in perception and to investigate how these priming effects may be weakened or strengthened under various circumstances. To tie more closely into previous work on variant persistence in large corpora of conversational speech data, the proposed experiments also examine priming of the variable pronunciations that individuals choose in speech 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.