With National Science Foundation support, Dr. Valerie Fridland will conduct eighteen months of linguistic research investigating how vowel perception is affected by dialect experience. In the past decade, sociolinguists have provided descriptive accounts of low level phonetic changes that are affecting the vowel systems of speakers in various regions of the U.S. Some of these changes are creating greater convergence in regional vowel production, but many are actually contributing to greater divergence. African-American groups in the U.S. appear to have only limited participation in many of these shifts, which are predominately associated with European-American speech. While such research illuminates how vowel sounds are produced across the country, little is known about how such changes are perceived by speakers within and among regions and ethnic groups. In addition, the impact of such changes on cross-dialectal intelligibility has not been explored. Building on previously supported research and groundwork laid by experimental speech perception research, this project examines how speech perception is shaped by both social and linguistic experience, addressing questions of change across the lifespan and the impact of social and regional variation on perception. To gain a better understanding of the impact such productive changes have on how we hear and understand speech, this research will investigate how shifts in vowel production shape speakers' recognition of phonetic and phonemic boundaries and whether such divergent trends across dialects and ethnic groups lead to greater cross-dialect confusion. Through the combination of instrumentally-synthesized perception tests, production data and cross-dialect comparison, the study hopes to gain insight into some of the prevailing claims about the origins and diffusion of sound change and the fluidity of perception and production across the lifespan.
Research in these areas should have both theoretical and applied benefits, from contributing to basic theories about the nature of sound change to the discussion surrounding the growth of national education standards for curricula and testing, the ability of adult language learners to adjust aspects of their phonological system, and issues of cross-dialectal comprehension in the development of educational materials. Beyond the educational sphere, comparative examination of both the productive and perceptual range of speakers affected by different regional and ethnic dialect experience can assist in the development of computer-based voice technology. Difficulties faced in the wide application of speech production and voice recognition software need not be insurmountable if research can calibrate productive and perceptual differences among major regional and social groups across the U.S.
