Linking Discrete Phonological Representations to Continuous Phonetic Manifestations
There is a long history of discussing the relationship between discrete phonological representations and continuous phonetic manifestations, and opinions range from the relationship being very murky (Hockett, 1955) to isomorphic (Pierrehumbert 2003, Johnson 1997). In this talk, I point out that wherever one stands on the issue, there is a genuine benefit to stating the envisioned relationship precisely.
I look at two cases where there has been no clear consensus in the phonological literature: (a) ambisyllabic consonants, (b) segmenthood. In both cases, I will argue for a view that espouses specific consistent "mappings" between discrete phonological representations and their phonetic manifestations, which in turn allows us to identify/clarify the abstract representational analyses. Such cases suggest that developing clear linking hypotheses will help with the further understanding of discrete phonological representations through phonetic data.