File Download
Supplementary
-
Citations:
- Appears in Collections:
postgraduate thesis: Expanding the native and non-native architectures of speech perception : behavioral and neural perception of Cantonese lexical tone and English lexical stress in Cantonese-English bilinguals
Title | Expanding the native and non-native architectures of speech perception : behavioral and neural perception of Cantonese lexical tone and English lexical stress in Cantonese-English bilinguals |
---|---|
Authors | |
Advisors | |
Issue Date | 2018 |
Publisher | The University of Hong Kong (Pokfulam, Hong Kong) |
Citation | Choi, T. W. [蔡浚文]. (2018). Expanding the native and non-native architectures of speech perception : behavioral and neural perception of Cantonese lexical tone and English lexical stress in Cantonese-English bilinguals. (Thesis). University of Hong Kong, Pokfulam, Hong Kong SAR. |
Abstract | Human speech is constituted by both segmental and suprasegmental information. In many languages, suprasegmental information carries a functional load for word distinction. Cantonese and English are two prominent examples as they involve the lexical use of suprasegmental information i.e., lexical tones and lexical stress respectively. Despite the high functional load of suprasegmental information at least in tone languages, theories and models for native and non-native speech perception were largely devised to account for segmental speech perception only. To expand the native and non-native architectures of speech perception, this thesis investigates three important issues in native and non-native suprasegmental speech perception. For native suprasegmental speech perception, Study 1 investigates whether Cantonese lexical tones and vowels are processed integrally at the pre-attentive cortical level. For non-native suprasegmental speech perception, Study 2 explores whether tone language experience induces behavioral perceptual advantage with English lexical stress in Cantonese speakers relative to native English speakers. Study 3 further examines whether the perceptual advantage is influenced by segmental variation, and its cortical underpinnings.
In Study 1, Cantonese speakers were tested with a passive oddball paradigm where mismatch negativities (MMNs) to tonal, vowel and tonal-and-vowel deviations were elicited in the speech condition. Similarly, MMNs to fundamental frequency, formant frequency and fundamental frequency-and-formant frequency deviations were elicited in the non-speech condition. In both conditions, the MMNs to double deviants were significantly smaller in amplitude than the sum of single feature MMNs. Morphological comparisons only showed marginally significant latency differences between formant frequency-MMN and fundamental frequency-MMN. Collectively, results reflect the perceptual integration of Cantonese lexical tones and vowels at the phonological level, and partial integration of also the perceptual integration of fundamental frequency and formant frequency at the auditory level. The TTRACE+ model is proposed based on these findings.
In Study 2, Cantonese speakers and native English speakers discriminated English lexical stress patterns under various phonotactic/lexical and acoustic conditions. Cantonese speakers outperformed native English speakers when attending to natural speech stimuli and acoustically manipulated stimuli that preserved only the fundamental frequency cue. The outperformance was amplified when the lexical stress patterns were carried in English real words but became absent when the fundamental frequency cue was taken out with the duration and intensity cues preserved. The findings highlight the facilitative effect of Cantonese tone language experience on English lexical stress perception, which is presumably due to Cantonese speakers’ enhanced sensitivity to fundamental frequency.
In Study 3, Cantonese speakers and native English speakers were tested with both active and passive oddball paradigms where event-related potential (ERP) responses to English lexical stress deviations were elicited. Cantonese speakers exhibited behavioral perceptual advantage relative to native English speakers, but the perceptual advantage disappeared when segmental variations were introduced. In terms of ERP responses, Cantonese speakers exhibited diminished P3b amplitudes relative to native English speakers under the conditions with segmental variations, implying the increased cognitive demand for Cantonese speakers when segmental variation was introduced. Theoretical implications of Study 1, 2 and 3 to native and non-native models of speech perception are further discussed. |
Degree | Doctor of Philosophy |
Subject | Speech perception |
Dept/Program | Speech and Hearing Sciences |
Persistent Identifier | http://hdl.handle.net/10722/266335 |
DC Field | Value | Language |
---|---|---|
dc.contributor.advisor | Tong, X | - |
dc.contributor.advisor | Wong, LLN | - |
dc.contributor.author | Choi, Tsun-man, William | - |
dc.contributor.author | 蔡浚文 | - |
dc.date.accessioned | 2019-01-18T01:52:05Z | - |
dc.date.available | 2019-01-18T01:52:05Z | - |
dc.date.issued | 2018 | - |
dc.identifier.citation | Choi, T. W. [蔡浚文]. (2018). Expanding the native and non-native architectures of speech perception : behavioral and neural perception of Cantonese lexical tone and English lexical stress in Cantonese-English bilinguals. (Thesis). University of Hong Kong, Pokfulam, Hong Kong SAR. | - |
dc.identifier.uri | http://hdl.handle.net/10722/266335 | - |
dc.description.abstract | Human speech is constituted by both segmental and suprasegmental information. In many languages, suprasegmental information carries a functional load for word distinction. Cantonese and English are two prominent examples as they involve the lexical use of suprasegmental information i.e., lexical tones and lexical stress respectively. Despite the high functional load of suprasegmental information at least in tone languages, theories and models for native and non-native speech perception were largely devised to account for segmental speech perception only. To expand the native and non-native architectures of speech perception, this thesis investigates three important issues in native and non-native suprasegmental speech perception. For native suprasegmental speech perception, Study 1 investigates whether Cantonese lexical tones and vowels are processed integrally at the pre-attentive cortical level. For non-native suprasegmental speech perception, Study 2 explores whether tone language experience induces behavioral perceptual advantage with English lexical stress in Cantonese speakers relative to native English speakers. Study 3 further examines whether the perceptual advantage is influenced by segmental variation, and its cortical underpinnings. In Study 1, Cantonese speakers were tested with a passive oddball paradigm where mismatch negativities (MMNs) to tonal, vowel and tonal-and-vowel deviations were elicited in the speech condition. Similarly, MMNs to fundamental frequency, formant frequency and fundamental frequency-and-formant frequency deviations were elicited in the non-speech condition. In both conditions, the MMNs to double deviants were significantly smaller in amplitude than the sum of single feature MMNs. Morphological comparisons only showed marginally significant latency differences between formant frequency-MMN and fundamental frequency-MMN. Collectively, results reflect the perceptual integration of Cantonese lexical tones and vowels at the phonological level, and partial integration of also the perceptual integration of fundamental frequency and formant frequency at the auditory level. The TTRACE+ model is proposed based on these findings. In Study 2, Cantonese speakers and native English speakers discriminated English lexical stress patterns under various phonotactic/lexical and acoustic conditions. Cantonese speakers outperformed native English speakers when attending to natural speech stimuli and acoustically manipulated stimuli that preserved only the fundamental frequency cue. The outperformance was amplified when the lexical stress patterns were carried in English real words but became absent when the fundamental frequency cue was taken out with the duration and intensity cues preserved. The findings highlight the facilitative effect of Cantonese tone language experience on English lexical stress perception, which is presumably due to Cantonese speakers’ enhanced sensitivity to fundamental frequency. In Study 3, Cantonese speakers and native English speakers were tested with both active and passive oddball paradigms where event-related potential (ERP) responses to English lexical stress deviations were elicited. Cantonese speakers exhibited behavioral perceptual advantage relative to native English speakers, but the perceptual advantage disappeared when segmental variations were introduced. In terms of ERP responses, Cantonese speakers exhibited diminished P3b amplitudes relative to native English speakers under the conditions with segmental variations, implying the increased cognitive demand for Cantonese speakers when segmental variation was introduced. Theoretical implications of Study 1, 2 and 3 to native and non-native models of speech perception are further discussed. | - |
dc.language | eng | - |
dc.publisher | The University of Hong Kong (Pokfulam, Hong Kong) | - |
dc.relation.ispartof | HKU Theses Online (HKUTO) | - |
dc.rights | The author retains all proprietary rights, (such as patent rights) and the right to use in future works. | - |
dc.rights | This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License. | - |
dc.subject.lcsh | Speech perception | - |
dc.title | Expanding the native and non-native architectures of speech perception : behavioral and neural perception of Cantonese lexical tone and English lexical stress in Cantonese-English bilinguals | - |
dc.type | PG_Thesis | - |
dc.description.thesisname | Doctor of Philosophy | - |
dc.description.thesislevel | Doctoral | - |
dc.description.thesisdiscipline | Speech and Hearing Sciences | - |
dc.description.nature | published_or_final_version | - |
dc.identifier.doi | 10.5353/th_991044069409003414 | - |
dc.date.hkucongregation | 2018 | - |
dc.identifier.mmsid | 991044069409003414 | - |