| |
Abstract:
Clinical and experimental evidence indicates that talkers are
influenced by acoustic feedback from their own speaking voice. In
postlingually deafened adults, for example, problems related to
intensity and pitch control are common. In this study we were
concerned with the role that acoustic feedback plays in the
control of vocal fundamental frequency. In particular we examined
the different roles played by moment-to-moment feedback and a
stored representation or internal model of habitual pitch.
Eighteen subjects were tested in three feedback conditions while
they produced vowels. In one condition fundamental frequency
feedback was slowly shifted up in frequency without the subjects'
awareness. In a second condition, the subjects' feedback was
slowly shifted down in frequency. In both of these conditions the
shift was carried out over 100 trials in one cent steps. After
this 100-trial training period, feedback was shifted back to
normal frequency. A control condition in which subjects produced
the same number of tokens with normal feedback was also tested.
Two effects were observed. Subjects compensated for the F0
transformation during the training sessions but also showed
aftereffects of the transformation. When F0 feedback was returned
to normal, the subjects modified their produced F0 in the
opposite direction to the training shift. The results suggest
that fundamental frequency is controlled through moment-to-moment
feedback and with reference to an internal pitch representation.
This is consistent with current work on internal models of speech
motor control.
|
