Gabriel J Cler1, Jackson C Lee2, Talia Mittelman3, Cara E Stepp4, Jason W Bohland5. 1. Graduate Program for Neuroscience-Computational Neuroscience, Boston University, MADepartment of Speech, Language, and Hearing Sciences, Boston University, MA. 2. Department of Health Sciences, Boston University, MA. 3. Department of Biomedical Engineering, Boston University, MA. 4. Graduate Program for Neuroscience-Computational Neuroscience, Boston University, MADepartment of Speech, Language, and Hearing Sciences, Boston University, MADepartment of Biomedical Engineering, Boston University, MASchool of Medicine, Department of Otolaryngology-Head and Neck Surgery, Boston University, MA. 5. Graduate Program for Neuroscience-Computational Neuroscience, Boston University, MADepartment of Speech, Language, and Hearing Sciences, Boston University, MADepartment of Health Sciences, Boston University, MA.
Abstract
Purpose: Delayed auditory feedback (DAF) causes speakers to become disfluent and make phonological errors. Methods for assessing the kinematics of speech errors are lacking, with most DAF studies relying on auditory perceptual analyses, which may be problematic, as errors judged to be categorical may actually represent blends of sounds or articulatory errors. Method: Eight typical speakers produced nonsense syllable sequences under normal and DAF (200 ms). Lip and tongue kinematics were captured with electromagnetic articulography. Time-locked acoustic recordings were transcribed, and the kinematics of utterances with and without perceived errors were analyzed with existing and novel quantitative methods. Results: New multivariate measures showed that for 5 participants, kinematic variability for productions perceived to be error free was significantly increased under delay; these results were validated by using the spatiotemporal index measure. Analysis of error trials revealed both typical productions of a nontarget syllable and productions with articulatory kinematics that incorporated aspects of both the target and the perceived utterance. Conclusions: This study is among the first to characterize articulatory changes under DAF and provides evidence for different classes of speech errors, which may not be perceptually salient. New methods were developed that may aid visualization and analysis of large kinematic data sets. Supplemental Material: https://doi.org/10.23641/asha.5103067.
Purpose: Delayed auditory feedback (DAF) causes speakers to become disfluent and make phonological errors. Methods for assessing the kinematics of speech errors are lacking, with most DAF studies relying on auditory perceptual analyses, which may be problematic, as errors judged to be categorical may actually represent blends of sounds or articulatory errors. Method: Eight typical speakers produced nonsense syllable sequences under normal and DAF (200 ms). Lip and tongue kinematics were captured with electromagnetic articulography. Time-locked acoustic recordings were transcribed, and the kinematics of utterances with and without perceived errors were analyzed with existing and novel quantitative methods. Results: New multivariate measures showed that for 5 participants, kinematic variability for productions perceived to be error free was significantly increased under delay; these results were validated by using the spatiotemporal index measure. Analysis of error trials revealed both typical productions of a nontarget syllable and productions with articulatory kinematics that incorporated aspects of both the target and the perceived utterance. Conclusions: This study is among the first to characterize articulatory changes under DAF and provides evidence for different classes of speech errors, which may not be perceptually salient. New methods were developed that may aid visualization and analysis of large kinematic data sets. Supplemental Material: https://doi.org/10.23641/asha.5103067.
Authors: HeeCheong Chon; Eric S Jackson; Shelly Jo Kraft; Nicoline G Ambrose; Torrey M Loucks Journal: J Speech Lang Hear Res Date: 2021-06-21 Impact factor: 2.297
Authors: Chris J D Hardy; Rebecca L Bond; Kankamol Jaisin; Charles R Marshall; Lucy L Russell; Katrina Dick; Sebastian J Crutch; Jonathan D Rohrer; Jason D Warren Journal: Front Neurol Date: 2018-10-29 Impact factor: 4.003