Positional Neutralization - Linguistics - University of California ...

syllable 2 of disyllables in Turkmen, but the closeness of the relationship of the latter to
Modern Turkish does not allow generalizations to be expanded on this point.
It might be objected at this point that English does in fact use vowel duration to
signal things other than placement of stress. Final lengthening is a well-known
example 123 . However, English final lengthening is consistent and robust only in the higher
prosodic domains. At the word-level, it is sporadic and weak (cf. Beckman and Edwards
1990), which would tend to minimize its effect on perception of lexical stress placement
(The phonetic differences between final lengthening and stress are discussed at length in
chapter 3). Turkish initial strengthening, unlike the final lengthening found in English, is
realized consistently in both lower and higher prosodic domains, and would therefore be
expected to be more disruptive to the perception of stress placement than English final
lengthening. Final lengthening is also frequently realized alongside a severe amplitude
and F0 drop, both of which would minimize the likelihood of additional duration on the
final syllable being mistaken for lexical stress. The crosslinguistic propensity for stress to
avoid final syllables is also discussed in chapter 3.
123
I do not consider durational variations induced by local segmental or syllabic environment in this
connection, as these lack the culminativity associated with stress placement. In such cases duration is not
being deployed in the language as the primary cue for a uniquely-identifiable prosodic position (such as
primary stress or edge). That additional duration supplied by local segmental environment can nonetheless
dramatically impact the structure of a stress system is of course clear, as the attraction of otherwise-edgebased
stress to internal heavy syllables in many languages clearly demonstrates.
324