Efference copy and context effects.
Mark Scott, Bryan Gick · The Journal of the Acoustical Society of America · 2011
One hypothesized component of speech production is “efference copy”; a signal carrying the predicted sensory-consequences of the motor-system’s actions [Wolpert and Ghahramani (2000)]. While brain-imaging studies [e.g., Aliu et al. (2009); Numminen et al. (1999)] have shown dampened auditory-cortex response to self-generated sounds (the predicted effect of efference copy), there are few behavioral demonstrations. This experiment will examine whether a context-effect (a common behavioral measure) is also dampened by efference copy. A context effect is a shift in categorization caused by surrounding sounds. For example, a syllable ambiguous between /da/ and /ga/ is perceived as more /da/-like when preceded by /ar/, but more /ga/-like when preceded by /al/ [Mann (1980)]. In this experiment, participants will silently mouth /ar/ or /al/ in time to a recording (it is assumed that mouthing engages efference copy). In one condition the recording will match what participants are mouthing; in another condition it will mismatch; in a third condition participants will hear the sounds without mouthing. After each mouthing, they will categorize a target syllable as /da/ or /ga/. The prediction is that the context effect will be dampened in the matching condition (due to efference copy), but not in the mismatching condition.