Do working memory and autistic traits predict L2 prosody perception?

This research examines the role of working memory and autistic traits in predicting the prosodic prominence and the perceptual strength of the prosodic boundaries in a second language. University students in Taiwan completed two questionnaires for cognitive and personality assessments – an autism-spectrum quotient questionnaire and a working memory questionnaire. Each subject participated in one of the two auditory tasks – Rapid Prosody Transcription Task and Boundary Detection Task. In the Rapid Prosody Transcription Task, the subjects marked the prominent words and the prosodic boundaries according to what they heard. In the Boundary Detection Task, the subjects heard fragments of English, Swedish, and Taiwanese varying in length (two levels: 2-second or one-word long) and size of the actual following boundary (three levels: word boundary, intermediate phrase boundary, and intonational phrase boundary). The subjects were asked to predict the upcoming boundary size using an onscreen slider. Previous studies have shown that native speakers and L2 learners used different acoustic measures as predictors of their perception of phrasal prominence and prosodic phrasing. This research takes intrinsic individual differences into account and discusses whether individual differences are influential factors for perceptual results. The overall results will also speak to issues regarding the development of L2 prosody acquisition and the role of attention in prosodic processing.

Publications

Presentations

Funding

Follow My Lead: Prosodic Entrainment
Prosodic entrainment is the tendency for individuals to modify their intonation and rhythm to converge with the behavior of their interlocutors. This research examines the prosodic entrainment in spoken dialogues, a natural speech corpus is constructed by collecting spontaneous conversations among young people. The participants were recorded speaking their native language and second language in different sessions at the game settings (computer game and board game). The entrainment is measured by the prosodic features and social variables. A follow-up perception experiment examines the similarity between the entrained speech and the model speech as well as between speakers’ native language and their second language.

Publication

Funding

Prosodic Turn-Taking Cues in Face-Masked Speech

This study aims to investigate the impact of wearing a mask when speakers are having a conversation with their interlocutor and how they signal turn exchanges using prosodic cues in two different languages. Ten speakers (native: Mandarin; L2: English) will be recruited for a series of board gameplays that require cooperation and extensive conversations. The speakers will form five dyads and attend four recording sessions varied by the following conditions: 2 Face Mask conditions (i.e., with or without the mask) × 2 Face-to-Face conditions (i.e., whether or not the interlocutors can see each other) × 2 Language conditions (i.e., whether the language in use is Mandarin or English). Their conversations will be recorded, annotated, and serve as stimuli for the auditory experiment. The prosodic cues (including loudness, pitch, voice quality,  and speaking rate) speakers used in their recordings will be normalized and analyzed. In the auditory experiment, listeners of the same language background and native speakers of English will listen to the stimuli and recognize words missing from the text. The accuracy of the listeners’ word recognition responses will reveal to us whether the attenuated speech, the lack of visual cues, and the use of non-native language will affect the intelligibility of speech in turn-taking to listeners.

Funding

Post-focus compression in code-switching

English and Mandarin Chinese are both known to show on-focus pitch range expansion. However, post-focus compression (PFC) was found in English but not in Taiwan Mandarin, a dialect of Mandarin Chinese. Given that code-switching is often used to examine what happens when two linguistic systems come into contact and interact, we would like to see how PFC is realized and processed in code-switching sentences in these two languages. In this proposal, production and perception experiments are to be conducted to see (a) Is PFC observed in simultaneous/early bilinguals’ code-switching production? (b) Does the occurrence of PFC in production correspond to late bilinguals’ (aka language learners’) language proficiency; in other words, is there a transfer of the lack of PFC from Taiwan Mandarin to English? (c) How well can listeners identify the focused element in code-switching sentences given that one of the languages lacks PFC in unilingual production? The findings are expected to provide evidence for language transfer in code-switching sentences produced and perceived by bilingual speakers of different language backgrounds and to carry the implication for using code-switching in an English as a Medium of Instruction (EMI) classroom in Taiwan.

Funding