Matthias K. Franken, Robert J. Hartsuiker, Nicolas Bourguignon, Petter Johansson, Lars Hall, Andreas Lind

A popular technique to investigate the role of auditory feedback in speech production has been to alter auditory feedback (AAF) in real time so that speakers hear their productions perturbed. In response, speakers compensate by shifting their speech output in the opposite direction. Current theory suggests this is caused by a mismatch between expected and observed feedback. A methodological issue is the difficulty to fully isolate the speaker’s hearing so that only AAF is presented to their ears. Most research groups make use of commercial headphones that do not always offer much passive sound attenuation (PSA). As a result, participants may be presented with two simultaneous signals. If this is true, an alternative explanation for responses to AAF is that these may be due to a comparison between the two signals (AAF and the true feedback). The aim of the current study was to investigate this alternative hypothesis by varying the amount of PSA. An experiment (N=49) was carried out where participants vocalized while auditory feedback was unexpectedly pitch-shifted by either -25, +25, -100 or +100 cents. The feedback was received through three pairs of headphones, with varying amounts of PSA. The results suggest that participants compensated for the unexpected pitch shifts. In addition, their response scaled with the manipulation magnitude. However, the responses to unexpected pitch shifts were not affected by the different levels of PSA, suggesting that these results are in line with current theory’s hypothesis that responses are due to a mismatch between observed and expected feedback.