Perceptual invariance or orientation specificity in American Sign Language? Evidence from repetition priming for signs and gestures

Repetition priming has been successfully employed to examine stages of processing in a wide variety of cognitive domains including language, object recognition, and memory. This study uses a novel repetition priming paradigm in the context of a categorisation task to explore early stages in the processing of American Sign Language signs and self-grooming gestures. Specifically, we investigated the degree to which deaf signers’ and hearing nonsigners’ perception of these linguistic or nonlinguistic actions might be differentially robust to changes in perceptual viewpoint. We conjectured that to the extent that signers were accessing language-specific representations in their performance of the task, they might show more similar priming effects under different viewing conditions than hearing subjects. In essence, this would provide evidence for a visually based “lack of invariance” phenomenon. However, if the early stages of visual-action processing are similar for deaf and hearing subjects, then no such difference should be found. In both groups, we observed robust effects of viewpoint, indicating that repetition priming for identical prime–target pairs was greater than in cases of categorisation in which the prime and target varied in viewpoint. However, we found little evidence of group-related differences that could be interpreted as effects of perceptual invariance. These outcomes indicate that initial stages of sign and gesture recognition required for the categorisation of action types do not differ as a function of experience with a signed language. Instead, our data are consistent with and extend previously described visual-perceptual studies that have reported evidence for orientation-specific representations of human actions.