Partial compensation to delayed auditory feedback: An analysis of syllable duration.

Abstract

Altered feedback perturbation studies have shown that speakers only partially compensate for shifts in vocal pitch, intensity, or spectral content. The partial compensation generally plateaus around 30%-40% of the non-shifted, baseline production and is proposed to reflect the integrated effect of altered (auditory) and non-altered (somatosensory) sensory feedback. The authors hypothesized that the same pattern should hold for changes in syllable duration with delayed auditory feedback. To test this, the authors calculated average syllable duration from 21 young adults who read a standard passage as auditory feedback was delayed by 0, 10, 50, 100, 200, 300, and 600 ms. Given the complex nature of speech errors elicited under delayed auditory feedback, the authors compared durational effects across two broad error categories: prolongations and repetitions. Average syllable duration increased non-linearly and reached a peak near 40% of baseline durations. Separated by error type, prolongations, and distortions made up 89% of perceived errors. Only 9% of perceived errors were identified as repetitions. Unlike prolongations, however, the duration of repeated tokens ranged from 74% to 110% of the target syllable and compensated fully for the delay. These results are consistent with the notion that increasing syllable duration compensates partially for delays in auditory feedback. The compensation pattern resembles that of responses to other forms of altered auditory feedback, which suggests that this may be a general attribute of low-level, sensorimotor compensation. The higher frequency of prolongation occurrences further suggests that these may be the primary means of compensation for delayed auditory feedback. Repetitions, on the other hand, likely reflect compensatory mechanisms of higher-level motor planning or sequencing processes.