In the visual domain, cerebral hemispheric differences support the theory that relatively independent neural subsystems operat

The role of talker-specific information in spoken word recognition differs in the right and left hemispheres

Julio González* & Conor T. McLennan**

* Universitat Jaume I. Castellón de la Plana.

** University at Buffalo, NY. (USA)

In the visual domain, cerebral hemispheric differences support the theory that relatively independent neural subsystems operate in parallel to underlie abstract-category and specific-exemplar priming of word forms [Marsolek, 2004]. In particular, the right hemisphere (RH) is more sensitive than the left hemisphere (LH) to form-specific stimulus information (e.g., letter case). In the auditory domain, research has revealed the effects of indexical variation on spoken word processing and recognition [see Luce & McLennan, 2005, for review], but hemispheric differences during on-line processing of spoken words have yet to be examined.

In the present study, the authors investigated the role of talker-specific information in spoken word recognition in the LH and RH. In a long term repetition priming experiment, participants made lexical decisions to auditory stimuli in two blocks of trials. Stimuli in the 1st block served as primes and those in the 2nd as targets. Materials were speech stimuli recorded from two talkers (JG and LA) of different gender. Stimuli were presented monaurally over headphones with pink noise simultaneously presented to the opposite ear. Every participant participated in four sessions, each corresponding to one combination of hemisphere (ear) of presentation during the prime and the target blocks. Primes either matched, mismatched, or were unrelated to the targets. Matched primes and targets were identical on the talker dimension (e.g., foca_JG[seal]-foca_JG; foca_LA-foca_LA). Mismatched primes and targets differed on the talker dimension (e.g., foca_JG-foca_LA; foca_LA-foca_JG). Unrelated primes served as the baseline condition in which the experimental stimuli of interest presented during the 2nd block were preceded by unrelated stimuli in the 1st block (e.g., lirio[iris]-foca[seal]).

Orthogonal combination of the 3 levels of prime (match, mismatch, unrelated), 2 levels of target (talker JG, talker LA), 2 levels of hemisphere of presentation at prime block (left, right), and 2 levels of hemisphere of presentation at target block (left, right) resulted in 24 conditions. The results demonstrate that priming was attenuated in the mismatched condition, but only when targets were presented to the RH (left ear). The current findings provide evidence that talker-specific information affects processing of spoken words differently in the right and left hemispheres.

Marsolek, C. J. (2004). Abstractionist versus exemplar-based theories of visual word priming: A subsystems resolution. The Quarterly Journal of Experiemental Psychology: Setion A, 57,1233-1259.

Luce, P. A. y M^cLennan, C. T. (2005). Spoken word recognition: The challenge of variation. En D. B. Pisoni y R. E. Remez (Eds.), Handbook of Speech Perception. Malden, MA: Blackwell.