Discussions: 1 session / week, 3 hours / session
Many details of phonetic realization cannot be predicted from standard phonological representations on a language-independent basis, so phonetic realization must be specified in grammar. In this seminar we will investigate phonetic realization as a component of grammar.
Besides its intrinsic interest, an understanding of phonetic realization is of considerable importance to phonological theory. There is evidence that the phonological patterning of sounds depends on their phonetic properties, but those phonetic properties are themselves the products of grammar, so a full analysis of phonological typology requires an understanding of phonetic realization. For example, Zhang (2001, 2004) argues that the phonological distribution of contour tones depends on the duration of the sonorous portion of syllable rimes, but the phonetic facts of rime duration are as much in need of analysis as the facts about contour tone distribution that they are intended to explain. That is, we need to analyze the grammar of duration as well as the grammar of tone.
Furthermore, given that patterns of segment duration cannot be determined from standard phonological representations on a language-independent basis, Zhang's analysis implies that phonological acceptability of contour tones depends on the output of phonetic realization. This is inconsistent with the common assumption that phonetics and phonology are separate modules and that the only interaction between the two is that the output of the phonology is passed to the phonetic realization component. Influences of phonetic realization on phonology can only be accommodated by incorporating the relevant aspects of phonetic realization into phonology (e.g. Flemming 2001, Zhang 2004) or by allowing some flow of information from the phonetic component to the phonology (e.g. Gordon 2004, Steriade 1997). We will review these and other proposals concerning the relationship between phonetic realization and phonology in light of several cases in which language-specific phonetic detail has been argued to affect phonological patterns (e.g. Gordon 1999, 2004, Jun 2002, Steriade 1997, Zhang 2001, 2004).
Adding phonetic detail into Optimality Theoretic phonology naturally leads to constraint-based analyses of phonetic realization (Flemming 1997, 2001, Gafos 2002, Zhang 2004). This approach to phonetic realization is independently interesting, not least because it offers a way to characterize the extent of cross-linguistic variation in phonetic realization (parallel to the OT approach to typology) - something that was lacking in previous models. We will look at analyses of consonant-vowel coarticulation, segment duration, gestural overlap and tone realization.
24.961 Introduction to Phonology
24.962 Advanced Phonology
|Does language-specific phonetics affect phonology?|
|4||Stop voicing and bursts|
|Models of phonetic realization|
|5||Rules and constraints|
|6||Implications of phonetic detail for phonological analyses|
|8||Releases and transitions|
|Stop bursts, consonant releases, etc.|
|10||Releases and transitions (cont.)|
|Too many solutions|
|11||Phonetic detail and the "too many solutions" problem|