Mike Stern presents on a topic in phonetics.
“Perceptual magnets” in bilingual perception
Learning one’s first language corresponds with a decrease in one’s ability to discriminate sounds near the “best exemplars” or “prototypes” of the phonological categories of the language (Grieser & Kuhl, 1989). This inverse relationship between category goodness and discriminability has been called the “perceptual magnet effect” (PME), as the prototype functions like a magnet by drawing similar sounds toward it in the perceptual space (Kuhl, 1991; Iverson & Kuhl, 1995). This phenomenon has interesting implications for phonological theory, i.e. that phonological categories are structured around prototypes, rather than clear boundaries; that phonological acquisition involves acquiring category prototypes; and that speech perception is facilitated by selective inattention to phonetic information near prototypes.
To date, the PME has only been studied in monolinguals, leaving open a number of questions:
1. Can PMEs develop for multiple languages within an individual?
2. Can PMEs strengthen, weaken, or disappear over time based on changes in language dominance?
The present study explores these questions by utilizing a perceptual experiment with Turkish-English bilinguals (target N=20) and American English (AE) monolinguals (target N=20). These groups were chosen because the Turkish phonological inventory contains /y/, which is absent in AE; and the AE inventory contains /æ/, which is absent in Turkish. By examining only these two vowels in these two groups, therefore, the linguistic source of any observed PMEs can be confidently designated as one or the other language.
In this talk, I will present preliminary results from the AE monolingual group (n=10), which suggest the existence of a phonetic prototype and a PME for /æ/. I will discuss the implications of this result, as well as possible results from the bilingual group. Specifically, the existence of PMEs in bilinguals would be consistent with the Areal Sound Pattern Hypothesis, in which PMEs in bilinguals account for contact-induced sound change (Blevins, 2017).
All are welcome!