Relationships among individual differences in speech perception, speech production, and cognitive functions : a case study of Cantonese tone merger

Abstract

Studies of speech processing have generally made the implicit assumption that typically developed speakers can distinguish all sounds of their mother tongue in perception and production. As such, individual differences in speech processing is usually studied with speakers differing in training/experience (Strait & Kraus, 2011), or populations with developmental disorders (Facoetti et al., 2010), and few investigations have been conducted without the effects from training/experience among typically-developed individuals. However, sociolinguists have long recognized that native speakers vary in their ability to discriminate speech sounds in their language, and enormous variability exists especially during a sound change in progress.

Taking the opportunity of an on-going tone merging in Hong Kong Cantonese, this thesis aims to systematically investigate individual differences of native speech perception, production, and their relationships with cognitive functions among typically-developed speakers. Three participant groups were recruited, who presented respectively the pattern of good perception and good production of all Cantonese tones [+Per+Pro], that of good perception of all tones but poor production of specifically the T2/T5 distinction [+Per-Pro], and that of poor perception and production of specifically the T2/T5 distinction [-Per-Pro]. Behavioral and neural measures of tone perception included reaction time, discrimination sensitivity index, and components of event-related potentials (ERPs) – the mismatch negativity (MMN), P3a, and rise time of amplitude envelope. Acoustic measurements were used to evaluate tone production in terms of both pitch and amplitude rise time. Components of attention and working memory in auditory and visual modalities are assessed with published cognitive test batteries.

The results show that, apart from the expected differences in accuracy and discrimination sensitivity of tone perception, both [+Per-Pro] and [-Per-Pro] took significantly longer to discriminate between tones than [+Per+Pro]. As for the performance in production, besides the differences in pitch offset, both [+Per-Pro] and [-Per-Pro] showed decreased differentiation in rise time between the two rising tones in production, compared with [+Per+Pro].With respect to the brain responses reflected in the MMN and P3a to pitch deviations among tones, [-Per-Pro] showed smaller and slower responses than one or both of the other two groups, but [+Per+Pro] and [+Per-Pro] did not differ from each other. However, both [+Per-Pro] and [-Per-Pro] showed weaker neural responses compared with [+Per+Pro] to the rise time of T5. In addition, [-Per-Pro] was poorer in tasks pertaining to the ability of attention switching/shifting regardless of modality than one or both of the other two groups, but [+Per+Pro] and [+Per-Pro] did not differ from each other.

Further correlation and regression analyses reveal that both pitch contour/height and rise time contributed to distinctive perception and production of rising tones, measures of perception (behavioral and neural) and production were correlated with each other, and attentional shifting in visual and auditory modalities significantly predicted performances of discrimination and production. Taken together, the findings of the present study suggest that attentional switching influences the quality of phonological representations, resulting in different degrees of distinctiveness in perception and production among typically-developed individuals.