Seminar: Peripherality and representations: A safari through the English low vowels, Thomas Kettig, 1 June
Seminar: Peripherality and representations: A safari through the English low vowels
Speaker: Thomas Kettig, University of Hawaiʻi at Mānoa
When: 1 June 2018, 3.30pm-5.00pm
Where: Basham Room, Baldessin Building, ANU
Focusing on the [æ~a~ɑ~ɒ] space, we trace how William Labov’s shifting framework accounts for vowel lengthenings, rises, and falls from Old English through to Early Modern English (EModE). Looking at contemporary shifts in New Zealand and Canadian English, we glimpse divergent pathways taken by the TRAP vowel that can each be reasoned to follow Labov’s principles. However, these explanations rely on definitions of peripherality that differ in their reliance on phonetic versus phonological criteria.
We take a closer look at the TRAP-BATH split, referred to here as primary /æ/-lengthening. Though peripheral BATH shifted in two different directions – i.e. [beəθ] in New York City English (NYCE) and [bɑːθ] in Southern Standard British English (SSBE) – the environments conditioning SSBE BATH appear form a subset of those in NYCE. This is suggestive of their common origin as durational lengthening of EModE TRAP at least three centuries ago.
Close inspection of contemporary SSBE reveals another phenomenon at work: secondary /æ/-lengthening. TRAP words like bad, sad, and bag have been intuited by 20th century SSBE-speaking linguists as longer than comparable words like lad and pad. Acoustic measurements confirm that though duration doesn’t seem to distinguish consistent minimal pairs, lexemes previously reported as especially long or short seem to have remained so for a century.
While treatments of primary /æ/-lengthening have been proposed within Lexical Phonology, generative frameworks do not account for the synchronic representation and diachronic transmission of the non-contrastive, lexically specific phonetic information we observe in secondary /æ/-lengthening. Instead, Exemplar Theory provides a better account for secondary /æ/-lengthening as the marginal residue of primary /æ/-lengthening. Freed of classic assumptions about the role of phonemic contrast in representations, we may be better able to conceive of peripherality in the low vowel space as flexible rather than strictly binary.