Generalization of the imitation effect within a natural class

Abstract

A natural class refers to a set of segments which share acoustic or articulatory features. Support for this notion has traditionally been provided by phonological alternations and phonotactic constraints, and recently by experiments (e.g., Goldrick, 2004). This study investigates the psychological reality of natural class using the imitation paradigm (Goldinger, 1998) in which subjects speech is compared before and after they are exposed to target speech (= study phase). Although this paradigm has shown that subjects shift their production in the direction of the target, these results do not reveal the size of the linguistic unit(s) influenced by the effect. That is, when a subject shifts production of a particular sound in a word, it is uncertain whether the subject is picking up on the word, the segment, or the feature. In this study, to investigate whether phonetic imitation is generalized across members of a natural class, the study-phase word list includes words with initial /p/ and /t/ (with extended VOT), while the pre- and post-study production list includes (1) the modeled words, replicating Shockley et al. (2004), (2) the modeled segments /p/ and /t/ in new words, and (3) the modeled feature [+spread glottis] (aspiration) in a new segment /k/.