Search Abstracts | Symposia | Slide Sessions | Poster Sessions
Exploring Cross-Categorical Pitch Shift Effects on Mandarin Tone Production
Poster Session A - Sandbox Series, Thursday, October 24, 10:00 - 11:30 am, Great Hall 3 and 4
This poster is part of the Sandbox Series.
Ya-Ning Wu1, Will Chih-Chao Chang1, Gregory Hickok1; 1University of California, Irvine
Introduction: The Dual Speech Coordination model proposed that laryngeal and supralaryngeal control of speech production are coordinated by distinct cortical pathways (Hickok et al., 2023). Based on the model, we hypothesize that speakers recruit a dorsal-lateral system to coordinate pitch vocalization, while a more ventral system is activated to coordinate phonetic/syllabic articulation. However, whether higher-level areas of linguistic planning in the dorsal-lateral system are involved in pitch coordination when the context is linguistically contrastive, such as producing lexical tones, is an open question. Previous research has explored the sensorimotor control of Mandarin tone production using behavioral Altered Auditory Feedback (AAF) tasks, where speech is recorded, manipulated, and present to participants in real time. However, these studies applied constant upward or downward shifts that alter only the pitch height but not the contour of the lexical tones (Tang, 2024). Therefore, the perturbed tone may sound less canonical but not necessarily from a different tone category. To address our research question regarding neural mechanisms underlying pitch control at higher linguistic and lower acoustic levels, we aim to conduct a variation of typical AAF task with fMRI design to investigate how cross-categorical pitch shifts affect Mandarin lexical tone production. Methodology: Participants will be asked to produce the Mandarin monosyllabic word /ma/ associated with three lexical tones: Tone1 (high level tone, 妈/ma1/), Tone2 (rising tone, 麻/ma2/) and Tone4 (falling tone, /ma4/骂) in both the baseline production task (session 1) and the main AAF experiment (session 2). In session 1, participants produce words shown on the screen for 2 seconds, while being recorded with a headset microphone with normal auditory feedback. Out of all recordings, we pick 5 for each target tone that are the closest to the mean pitch height (/ma1/) or slope (/ma2/, /ma4/) to become playback tokens for cross-categorical shift in the main experiment. In session 2, participants perform the same word production task with either normal feedback (Control condition) or cross-categorical pitch shift (Shift condition, e.g, producing /ma2/ but hearing a /ma1/ or /ma4/ token). Results & Discussion: Our behavioral pilot results (N=7) showed that participants exhibited compensation or following responses to cross-categorical shift during contour tone production (Tone2 and 4) but not level tone (Tone1); in addition, the magnitude of responses to shift increased over the time course within a trial. Moving forward to fMRI scanning, we predict that speakers will exhibit distinguishable activation patterns among the dorsal ROIs between conditions. Specifically, we expect to observe quantitative differences of dPCSA activation or qualitative differences of additional area 55b recruitment. Ultimately, our goal is to explore the dorsal-lateral substrates hierarchy underlying pitch control in linguistically contrastive context, which will further elucidate the sensorimotor integration and adaptation mechanisms of speech production in tonal language speakers.
Topic Areas: Language Production, Speech Motor Control