Abstract to paper in submission to Cognitive Psychology:

​

The experiments reported here investigate possible correlations between participants’ ability to identify non-native speech sounds (phonological discrimination), and their discrimination of ‘non-native’ faces, also known as the Other Race Effect (ORE), (Kelly et al., 2007); also Spelke and Kinzler (2007). These two skills exhibit timing parallels in typical infant development: in both instances, an effect of perceptual narrowing is observed between 6 and 13 months of age; in the case of language, this generally leads to a life-long deficit in speech perception. The present study compared the behavior of two linguistically opposed groups of speakers—native Japanese learners of English and native English speakers with only limited experience of Asian language or cultures—in two complementary forced- choice (‘ABX’) discrimination tasks, one visual, the other auditory. As expected, the results of each task reveal clear between-item differences in respect of Western vs. Asian contrasts, as well as notable individual differences in perceptual ability. However, no reliable correlations were observed between the two abilities: above-average discrimination ability in the auditory task did not predict superior performance in visual discrimination, nor vice versa. More surprising than the absence of correlations were two other findings: first, that Asian faces, especially Japanese faces, were significantly less discriminable than Caucasian faces for all participants (including native Japanese participants); second, that English speakers failed to reliably discriminate a phonological contrast that was native to English—/d/ vs. /t/—whenever the carrier language was Japanese. I consider the implications of both of these findings for previous research on the ORE, as well as for standard assumptions regarding the phonological basis of narrowed speech perception.