Phonetic imitation of multidimensional acoustic variation of the nasal split short-a system

Abstract

The current study investigates phonetic imitation of multiple acoustic features of pre-nasal /æ/ in California English. There is a great deal of cross-speaker heterogeneity: Many speakers show a raised /æN/ variant, in tandem with a backing and lowering of /æC/ (i.e., “split” short-a nasal system). This innovative split can also be realized with enhanced coarticulatory acoustic features on the pre-nasal allophone, namely extensive vowel nasality and increased diphthongization over the vowel duration. The present study compares phonetic imitation of three model talkers who produce distinct pre-nasal /æ/ variants: 1) a non-split talker who also does not produce enhanced nasal coarticulation and diphthongization, 2) a talker who produced the split, with a raised pre-nasal /æ/, but does not produce enhanced nasal coarticulation and diphthongization, and 3) a talker who produces both a pre-nasal /æ/ that is both raised and hyper-nasalized and heavily diphthongized. Participants who shadowed these model talkers could also be categorized as being advanced or conservative in their baseline productions. Results show that imitators who were more advanced in the sound change were more likely to imitate both vowel positioning and nasalization patterns of the conservative model talkers. Only diphthongization patterns of the most advanced model talker were imitated by the advanced participants. The results are discussed in terms of the multidimensionality of phonological contrast, theoretical proposals about phonetic imitation, and models of sound change and propagation.