The Origin of Vowel Length Neutralization in Final Position: Evidence from Finnish Speakers

An explanation for phonological word-final vowel shortening: Evidence from Tokyo Japanese

“…Like Myers and Hansen's (2007) account, the present account of phonological word-final vowel shortening can be classified as a listener-driven phonetic account of sound change, advocated most notably by Ohala (1981), with a number of followers (e.g., Beckman et al 1992;Guion 1998;Jacewicz, Fox, and Salmons 2006). In this view, some sound changes result from ambiguity in the speech signal arising from phonetic conditioning factors such as coarticulation.…”

“…Curiously, in many languages with a phonemic vowel length contrast, word-final vowel length neutralization is also commonly observed towards the short vowel phoneme (see, e.g., Buckley 1998;Myers and Hansen 2007). Myers and Hansen (2007) ask how such length neutralization often arises in the face of crosslinguistically prevalent phonetic final lengthening. Why do word-final vowels tend to be neutralized towards the short vowel phoneme if they are often in a position to be lengthened phonetically?…”

“…Why do word-final vowels tend to be neutralized towards the short vowel phoneme if they are often in a position to be lengthened phonetically? As a solution to the apparent paradox, Myers and Hansen (2007) propose that utterance-final devoicing, reported for various languages, may be responsible for the cross-linguistic tendency for phonological word-final vowel shortening. Myers and Hansen's (2007) proposal is based on their finding that the devoiced ending of a partially devoiced vowel is disregarded by the listener in phonemic length identification, producing perceptual shortening.…”

“…As a solution to the apparent paradox, Myers and Hansen (2007) propose that utterance-final devoicing, reported for various languages, may be responsible for the cross-linguistic tendency for phonological word-final vowel shortening. Myers and Hansen's (2007) proposal is based on their finding that the devoiced ending of a partially devoiced vowel is disregarded by the listener in phonemic length identification, producing perceptual shortening. This scenario may be a likely one for some languages that exhibit phonological word-final shortening, but not all, e.g., Tokyo Japanese, as we see below.…”

“…In this paper, I will present an alternative account to Myers and Hansen's (2007) for the apparent paradox between the cross-linguistic tendency for phonetic final lengthening and phonological word-final vowel shortening. I will argue that the paradox is resolved by postulating two hypotheses:…”