Bio: Bryan is a a professor in the Linguistics department at UBC, and director of the Interdisplinary Speech Research Lab (ISRL).

[ c l o s e ]

Abstract: Speech production models and phonetic description have traditionally been determinedly focused on the midline of the anterior tongue, largely downplaying the surrounding structures of the vocal tract. The lips have long been recognized as problematic for this view (e.g., Kelso & al 1986), being a muscular sphincter and having little to do with either the tongue or the midline. Close examination and modeling of the lips and other vocal tract structures reveal fatal flaws in dominant models of speech production, with solutions implicating more robust quantal relations in the biomechanics of >sphincters and other complex structures. Further examination and >modeling reveals similar sphincters and sphincter-like structures "all the way down" - and the principles that make sense of the lips are >shown to enable a more accurate and robust model for the whole vocal tract.

[ c l o s e ]

Abstract: One of the fundamental hypotheses that underlie the current model of phonology is that there is a systematic mapping between prosodic domains and their phonetic correlates (e.g. Pierrehumbert 1980; Beckman & Edwards 1990, 1994; Fougeron & Keating 1997; Pierrehumbert et al. 2000). While much research has been conducted on the phonology of prosodic domains in Salish languages, including Czaykowska-Higgins (1993, 1995, 1998, 2002, 2004a, b) and Urbanczyk (1996, 1999), the phonetic correlates of prosodic groupings in these languages have only recently begun to be investigated (Beck, 1999 Beck and Bennett, 2007; Barthmaier, 2004; Benner, 2006; Koch, 2008; Leonard, 2010). In this talk, I will present the results of my doctoral research, which examined the acoustic correlates of prominence and edge-marking in prosodic domains in St’át’imcets (Lillooet Salish). Results confirm the previously untested prediction that non-exhaustively parsed structures are acoustically distinct from exhaustively parsed structures and support a weak but constrained interpretation of the phonetics-phonology mapping.

Bio: Dr. Marion Caldecott completed her PhD at the University of British Columbia in 2009. Her research focus includes intonation, prosody and developing novel elicitation protocols in endangered languages. She has taught community-based language revitalization courses as well as a variety of courses at UBC, UVic and SFU. She is currently an instructor at Simon Fraser University

[ c l o s e ]

Abstract: Traditionally, sounds are considered phonologically related (considered allophones of the same phoneme) provided (1) they are in complementary distribution AND (2) they are phonetically similar (consider N and h in English: in complementary distribution, but phonetically dissimilar—they are not regarded as allophones). In this presentation I argue that neither (1) nor (2) is a reliable test for phonological relatedness among sounds: we cannot conclude that sounds are phonologically related, even when in complementary distribution AND phonetically similar. Rather, the only way sounds can be phonologically related (functionally equivalent) is if (3) they alternate with each other when morphemes are added (or subtracted), whether they are phonetically similar or not. Moreover, alternation need not invoke the segment as a phonological primitive. Instead, the composition of the alternating phonetic structures may be of any shape or size.

[ c l o s e ]