Auditory feedback is an important aid in order to control the production of speech as well as of singing. It is assumed that auditory feedback beside other feedback mechanisms (e.g. somatosensory feedback and visual feedback) helps to verify whether the current production of a passage of speech or singing is in accord with the acoustic-auditory intention.
From the viewpoint of movement sciences and neurosciences the acoustic-auditory speech signal can be interpreted as the result of movements (skilled actions) of speech articulators (lower jaw, lips, tongue, etc.) and thus auditory feedback can be interpreted as a feedback mechanism controlling skilled actions in the same way as visual feedback controls limb movements (e.g. reaching movements).
Contents |
In speech production it is well known that auditory feedback is a very important aid during speech acquisition in order to control the learning of speech items which are produced by a communication partner (e.g. caretaker), heared by the toddler, and which subseqeuntly are tried to be imitated by the toddler (Perkell et al. 1997, Callan et al. 2000). After speech acquisition (i.e. in the case of adults), auditory feedback as well as other feedback mechanisms (e.g. somatosensory feedback in the case of speech) are assumed not to be used thus intensively.
But the well-known delayed auditory feedback experiment (Lee 1950, see also Fairbanks 1995) indicated that auditory feedback becomes important during speech production even for adults, if the auditory perception pathway is altered (see also alterations by permenent shift of a formant, Perkell et al. 1997). A further well-known effect, which underlines the importance of auditory feedback during whole lifetime is that deafened adults begin to detoriorate e.g. in the production of sibilant fricatives (like /s/).
Because auditory feedback needs more than 100 msec before a correction occurs on the production level (see e.g. Burnett et al. 1998), it is a slow correction mechanism in comparison with the duration (or production time) of speech sounds (vowels or consonants). Thus auditory feedback is too slow in order to correct the production of a speech sound online (i.e. in real time), but it has been shown, that this auditory feedback is capable of changing speech sound production over a series of trials (i.e. adaptation by relearning; see e.g. perturbation experiments done with the DIVA model: neurocomputational speech processing). Typically adaptation times of around 10 minutes are sufficient for a nearly full adaptation.
The importance of auditory feedback in the case of (human) singing is reviewed by Howell (1985).
The role of auditory feedback in during learning and production of bird songs is reviewed by Brainard and Doupe (2000).