Search Abstracts | Symposia | Slide Sessions | Poster Sessions
Speech Production–Perception Link: How individual Differences in Acoustic Adaptation are related to Neural Speech Tracking
Poster Session A, Wednesday, September 30, 11:00 am - 1:00 pm, Wangari Maathai
This poster is part of the Sandbox Series.
Giorgio Piazza1, Jaimy Hannah1, Emily Ip1, Giovanni M. Di Liberto1; 1Trinity College Dublin
Human communication relies on the interplay between speech production and perception, including the ability to adapt speech to noisy environments and maintain successful communication. An example is the Lombard effect, where speakers modify acoustic features to enhance intelligibility during communication in noise. On the perception side, the listener's brain can selectively process target speech against competing background noise. However, individuals vary in these abilities, and reduced adaptation can compromise communication. It remains unknown whether an individual’s own speech production adaptation relates to their perceptual skills. We hypothesize that individuals who adapt their speech more effectively will also show better neural tracking of others’ speech in noise, reflecting a close production–perception link. To test this, participants completed a behavioural session and a separate OPM-MEG session. In the behavioural session, participants verbally interacted with a confederate in quiet and babble noise conditions. Speech production adaptation was quantified using two measures: vowel space area (spectral distinctiveness) and word-onset acoustic energy (amplitude envelope marking of word boundaries). In the OPM MEG session, the same participants listened to 30 minutes of pre recorded speech while fixating on a cross. Speech stimulus consisted of a 1 minute description of an image, presented either in quiet or in babble noise (1 dB SNR). After each description, participants were asked to select the correct image from two alternatives as quickly and accurately as possible. Neural tracking of the speech envelope was assessed using backward modelling. Preliminary results on pilot participants (N = 10) revealed individual variability in both production measures, differentiating stronger from weaker adapters. Half of the participants showed better neural tracking in noise than in quiet, while the other half showed the opposite pattern (mean prediction correlation r = 0.21). This difference aligned with the distinction between strong and weak adapters, supporting our hypothesis. Moreover, we observed a strong correlation (r = 0.55) between neural tracking scores (perception) and participants’ own word onset energy (production). Analyses linking full acoustic adaptation profiles to neural tracking accuracy are ongoing. This work pioneers an integrated behavioural–neurophysiological approach to studying individual differences in real-world communication.
Topic Areas: Speech Perception, Language Production