Is self-reported attentional control associated with second language speech perception?

Abstract

Although attention control has been proposed as a potentially relevant source of individual variation in speech perception, its association with nonnative segmental discrimination remains underexplored. The present study investigates whether individual differences in self-reported trait attentional control are associated with the perception of a difficult nonnative consonantal contrast by examining how adult Mandarin Chinese listeners discriminate the Cypriot Greek / m b/–/p/ contrast. Forty native speakers of Mandarin Chinese and 20 native speakers of Cypriot Greek completed an AXB discrimination task targeting the contrast in 3 phonetic contexts: initial singleton, intervocalic singleton, and intervocalic cluster. Mandarin listeners also completed the Attentional-Control Scale, which was used as a continuous measure of trait-like attentional control. The binomial logistic mixed-effects models showed that Mandarin listeners performed significantly worse than Cypriot Greek controls across all contexts. Within the Mandarin group, discrimination was significantly better in the intervocalic singleton condition than in the initial singleton and intervocalic cluster conditions, indicating that phonetic context modulates perceptual difficulty. Furthermore, higher attention-control scores were positively associated with discrimination accuracy, with the association reaching statistical significance in the intervocalic cluster condition but not in the other two contexts. However, the attention slopes did not differ significantly across contexts, providing no clear evidence that the association was context-dependent. These findings are consistent with current models of nonnative speech perception while suggesting that trait-like attentional control may help explain additional individual variation beyond broad cross-linguistic and contextual factors. More broadly, the results identify self-reported attentional control as a promising correlate of nonnative speech-perception performance.