Auditory Feedback Is Used for Adaptation and Compensation in Speech Timing

Abstract

Purpose Real-time altered feedback has demonstrated a key role for auditory feedback in both online feedback control and in updating feedforward control for future utterances. The aim of this study was to examine adaptation in response to temporal perturbation using real-time perturbation of ongoing speech. Method Twenty native English speakers with no reported history of speech or hearing disorders participated in this study. The study consisted of four word blocks, using the phrases “a capper,” “a gapper,” “a sapper,” and “a zapper” (due to issues with the implementation of perturbation, “gapper” was excluded from analysis). In each block, participants completed a baseline phase (30 trials of veridical feedback), a ramp phase (feedback perturbation increasing to maximum over 30 trials), a hold phase (60 trials with perturbation held at maximum), and a washout phase (30 trials, feedback abruptly returned to veridical feedback). Word-initial consonant targets (voice onset time for /k, g/ and fricative duration for /s, z/) were lengthened, and the following stressed vowel (/æ/) was shortened. Results Overall, speakers did not adapt the production of their consonants but did lengthen their vowel production in response to shortening. Vowel lengthening showed continued aftereffects during the early portion of the washout phase. Although speakers did not adapt absolute consonant durations, consonant duration was reduced as a proportion of the total syllable duration. This is consistent with previous research that suggests that speakers attend to proportional durations rather than absolute durations. Conclusion These results indicate that speakers actively monitor proportional durations and update the temporal dynamics of planning units extending beyond a single segment.

Abstracts of the 8th International Conference on Speech Motor Control Groningen, August 2022

“…Real-time auditory feedback perturbations typically trigger compensatory responses to an applied shift with productions opposed to the direction of the manipulation (see Caudrelier and Rochet-Capellan, 2019 for an overview). While this effect has been extensively studied for spectral alterations (e.g., first or second formant frequencies in vowels), recent studies also found compensatory responses in manipulations of temporal parameters in speech (Floegel et al, 2020;Oschkinat and Hoole, 2020;Karlin et al, 2021;Oschkinat and Hoole, 2022). In the current study, the duration of the German vowels /a/ and /a:/, a phoneme contrast realized in quantity without strong additional spectral cues, was manipulated in real-time.…”

“…Janine Schreen 1,3 , Tabea Thies (Yates, 1961). Karlin, Naber, and Parrell (2021), showed that individuals are sensitive to focal changes in feedback timing during vowel productions. The goal of this study is to quantify adaptation to gradual changes in latency of delayed auditory feedback (DAF).…”

An informal logic of feedback-based temporal control

“…The study in Oschkinat and Hoole (2020) found that post-vocalic intervals respond to temporal perturbations of feedback and that pre-vocalic intervals do not; specifically, subtle temporal delays of feedback imposed during a complex onset did not induce compensatory timing adjustments, while the same perturbations applied during a complex coda did. Another recent study (Karlin et al, 2021) found that temporal perturbations induced compensatory adjustments of vowel duration but not of onset consonant duration. Although the hybrid character of the model is a complication compared to purely feedforward or feedback control structure, it seems necessary to account for the dissociation in feedback sensitivity that was observed by these studies.…”

“…This dissociation in feedback sensitivity is a basic prediction of the hybrid model. Another recent study (Karlin et al, 2021) has found that temporal perturbations induced compensatory adjustments of vowel duration but not of onset consonant duration (codas were not examined). There may be other reasons why temporal feedback perturbations have differential effects on prevocalic and vocalic/post-vocalic intervals, and certainly there is much more to explore with this promising experimental paradigm.…”

Speakers monitor auditory feedback for temporal coincidence and linguistically relevant duration

“…Previous perceptual studies have indicated that listeners are less sensitive to non-contrastive duration differences in syllable onsets than in either syllable nuclei or codas (Goedemans & van Heuven, 1995;Huggins, 1972). Perturbations were of equal magnitude in syllable onsets and nuclei in both Oschkinat and Hoole (2020) and Karlin et al (2021)-i.e., if the syllable onset was perturbed by 40 ms, the vowel was also perturbed by 40 ms. Thus, the perceived error would not be as great for syllable onsets as for vowels, as the perturbation magnitude would be closer to the perceptual threshold.…”