Subphonemic and cross-phonemic priming in vowel shadowing: Evidence for the involvement of exemplars in production

Despite strong evidence that cognitive representations of speech targets rely upon a mapping between perceptual and motor memories, the nature of those representations—whether they are stored exclusively as abstract categories or can incorporate more detailed episodic memory—remains an open question. A primed vowel-shadowing experiment was conducted to investigate the extent to which the recent perception of subphonemic details of vowel quality can influence subsequent productions. If such effects are observed, they argue for an exemplar-based model of production in which the mapping between perceptual and articulatory representations can occur rapidly and can incorporate subphonemic detail. On experimental trials, subjects were primed with one vowel before they heard a second vowel, which they shadowed. On some trials, the formants of the prime were subtly manipulated. Significant subphonemic priming effects were observed in the F1 and F2 of responses. In addition, cross-phonemic priming tended to be dissimilatory, which may be related to an analogous phenomenon observed in studies of manual and oculomotor movement control. Accounts of these findings are discussed in the context of exemplar models of speech perception and production.