Miriam Oschkinat

Spectral auditory feedback perturbations have provided evidence that speakers can flexibly adjust for erroneous feedback in the ongoing speech production process (online compensation) and in future productions (adaptation). In temporal auditory feedback perturbations, it was shown that speakers also adjust productions in segments following the perturbation (reactive feedback control) and can adapt for perturbed segment durations, but not in syllable onsets (Oschkinat and Hoole 2020, but see Karlin and Parrell 2022 for adaptation of VOT). We suggested that weaker responses to temporally perturbed onsets indicate less use of auditory feedback in syllable onsets for the temporal control of speech as compared to vowels and codas but a greater reliance on feedforward readout (or possibly somatosensory feedback). This finding is particularly intriguing given that certain speech timing disorders manifest mostly in syllable onsets. Previous research has suggested a different incorporation of auditory feedback into the speech production process in persons who stutter (Max et al. 2004, Civier et al. 2010, Hickok et al. 2011, Harrington 1988), especially in syllable onsets. In the study I am presenting, we apply focal temporal auditory stretches in real-time to syllable onsets, vowels, and codas and compare responses of a group of persons who stutter with a group of fluent speakers. The results so far indicate the following: (1) persons who stutter are more sensitive to perturbations of syllable onsets than to perturbations of vowels and codas in the online control of speech; (2) persons who stutter are also more sensitive to perturbations of onsets than fluent speakers, indicating a greater and indeed perhaps undue reliance on auditory feedback in onsets for persons who stutter (cf. Civier et al. 2010). The results of this study indicate the considerable influence of structural aspects of the speech material in shaping responses to perturbed speech timing.