Search Abstracts | Symposia | Slide Sessions | Poster Sessions
When covert imitation becomes overt; uncovering neural basis of covert imitation in speech
Poster Session A, Wednesday, September 30, 11:00 am - 1:00 pm, Wangari Maathai
Patti Adank1, Takayuki Nagamine1, Chris Carignan1, Adamantios Gafos2; 1University College London, 2University of Potsdam
Introduction: We present an ongoing project and preliminary data examining the role of articulatory motor processes in speech perception. Motor theories of speech perception predict that listeners covertly imitate perceived speech actions based on their own speech production system using feedforward predictions that support speech perception. Previous studies, however, do not offer direct evidence of the motor engagement in speech perception due to the correlational nature of the neuroimaging studies and potential interactions with other articulatory processes in neurostimulation studies. In this project, we aim to establish whether perceptually driven motor commands interfere with goal-driven speech production and whether disruption of articulatory motor cortex reduces the interference from perceptually driven motor commands. Approach: We artificially created conflicts between perceptually driven and goal-driven articulatory commands using the Stimulus-Response Compatibility (SRC) task and performed a between-condition comparison of such conflicts with and without inhibition of the left tongue motor cortex (M1) using transcranial magnetic stimulation (TMS). We recorded tongue and lip movement with electromagnetic articulography (EMA) to capture detailed articulatory movements. Methods: We present data collected from ten British English speakers (6 female, 4 male, average age 27.4 years). They performed the SRC task before and after receiving 15 minutes of 1Hz rTMS to left tongue M1. In the SRC task, they were prompted to produce a target syllable (/ɹa/ or /va/) with a symbol cue (e.g., & or £), associated with either target syllable. Simultaneously, they were presented with congruent or incongruent audiovisual distractors, consisting of videos depicting a phonetically trained female British English speaker saying /ɹa/ or /va/ at three stimulus-onset asynchronies (SOAs). We analysed articulatory kinematics based on the velocity profile derived from changes in the tongue tip position recorded by EMA, following conventions in articulatory research. The velocity trajectory shape variations were converted into numerical values using Functional Principal Component Analysis (FPCA), which were then submitted to Bayesian linear mixed-effects modelling. Results: We have completed a pilot EMA-SRC study with ten listeners (6 female, 4 male, average 27.4 years). Analysis involved 2,638 tokens (/ɹa/: n = 1,320, /va/: n = 1,318) after excluding erroneous tokens. We compared the degree of tongue tip (TT) movement in /va/ production. The tongue tip is not usually involved in the production of /va/, yet in the incongruent condition, increased tongue tip movement was observed, especially at SOA3, thus demonstrating evidence of overt imitation during speech perception. Our ongoing EMA-SRC-TMS aims to affect the function of tongue M1, which we predict should modulate the degree of tongue tip movement in the incongruent /va/ condition. Conclusion: Our EMA results from pilot studies show that the tongue tip is actively involved during articulation of the incongruent /va/ condition. This result suggests that listeners’ show overt imitation in response to observing the distractor /ɹa/, which contains more tongue tip engagement than /va/. Our behavioural pilot study successfully captured conflicts between perceptually driven and goal-driven motor plans. Our EMA results therefore provide support for an active role of the entire speech production chain during speech perception.
Topic Areas: Speech Perception, Speech Motor Control