The perceptual distance between vowels : the effects of prototypicality and extremity

• Main Author: Nakai, Satsuki
• Published: University of Edinburgh 1999
• Subjects: 410
• Online Access: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.659862

The present dissertation is concerned with a question of how the perceptual distance between vowels in foreign languages (L2) and those in one's first language (L1) is determined. According to current models of L2 acquisition, the degree of difficulty that the learner faces in perceptually learning L2 sounds is determined by their perceived similarity to the L1 sounds. For instance, Flege's (1995) Speech Learning Model predicts that the learner would fail to establish a separate category for those L2 sounds which are similar to L1 sounds. Although empirical studies support a correlation between perceptual similarity between L1 and L2 sounds and the difficulty of perceptually acquiring L2 sounds, a measure for such similarity is yet to be established. In the first part of the dissertation I investigate possible locations of vowel prototypes, which current models of L2 phonetics see as reference points in determining the perceptual distance between L1 and L2 vowels. As far as vowels are concerned, evidence suggests that unlike visual prototypes vowel prototypes may be at a more extreme location than the centre of the geometrically represented category distribution (Bradlow, 1993; Johnson et al., 1993; Iverson and Kuhl, 1995; Lotto et al., 1996; Frieda, 1997; Lively and Pisoni, 1997). If vowel prototypes are more extreme than the typical production value for the category, how the language specificity observed in their phonetic realisations is reflected in the locations of prototypes becomes a crucial issue for the current models of L2 phonetics, which assume phonetic prototypes to be language specific. In the second part of the dissertation I re-examine Kuhl's Native Language Magnet theory which holds that phonetic prototypes established in the course of L1 acquisition attract nearby members of the category, causing the perceptual space to shrink towards these prototypes (Kuhl 1991, 1992, 1993). If L1 prototypes have such an assimilation effect, this should be taken into account when measuring the perpetual distance between L1 and L2 sounds. However, the validity of Native-Language Magnet Theory has been questioned by a number of researchers who replicated Kuhl's (1991) study, on which her Native Language Magnet Theory is based.