My research examines how we identify our own voices from the sounds around us, and how that identification affects our ongoing speech. I think the vestibular system could play an important role, because vestibular mechanoreceptors are deflected by the vibrational energy produced during vocalisation. I focus on the following questions:
- How do cochlear and vestibular auditory streams interact in the brainstem?
- How is own voice identified as an ascending auditory stream?
- What are the developmental and phenomenological consequences of us hearing our own speech?
Much of my research appraises persistent developmental stuttering. I am funded by the Economic and Social Research Council, and based at the University of Manchester.
Weak Vestibular Response in Persistent Developmental Stuttering
The first publication from my PhD measures the vestibular evoked myogenic potential (VEMP) in a stutter group paired on age and sex with a non-stutter control group. VEMP amplitude was 8.5 dB smaller in the stutter group than the non-stutter group (p = 0.035, 95% CI [-17 0.9, -16.1], t = -2.1, d = -0.8, conditional R² = 0.88). You can read a Twitter summary here.
The pre-registration gives the disruptive rhythm hypothesis (Howell et al., 1983; Howell, 2004) as a rationale. This proposes that sensory inputs additional to own speech audition are maximally fluency-enhancing when they coordinate with ongoing speech. The disruptive rhythm hypothesis is supported by this study. Vestibular input which coordinates with ongoing speech is fluency enhancing in ordinarily fluent controls, whereas the smaller vestibular input in people who stutter results in less fluency enhancement, accounting for the observed stuttering behavior.
The study was motivated by a compatible hypothesis. The basis of the hypothesis is that coincidence detection between deflection of cochlear and vestibular mechanoreceptors during vocalisation is fundamental to own voice identification, and that own voice identification differs between stutter and non-stutter groups.
A theory paper (Gattie, Lieven & Kluk, in preparation) explores the own voice identification hypothesis in more detail. It covers implications for auditory scene analysis, speech-motor control, and theories of mental content. These implications are not specific to stuttering. An application specific to stuttering aims to tie together neurological and psycholinguistic explanation with the phenomenology of the moment of stuttering. It offers a synthesis of well-established perspectives on stuttering, and would bridge medical and social models.
Our follow-up work will assess the auditory brainstem and the cortex (e.g. the temporo-parietal junction) with stimuli above and below vestibular threshold. We expect a morphological change upon crossing vestibular threshold, with differences between stutter and non-stutter groups.
In this 14 minute captioned video I describe the physiology of the inner ear, explain the hypotheses behind my research, and give a summary of findings so far.
Get in touch
© Max Gattie, 2021