The Tritone Paradox and the Pitch Range of the Speaking Voice: A Dubious Connection
Deutsch and coworkers (Deutsch, 1991; Deutsch, North, & Ray, 1990) have proposed that individual differences in the perception of the "tritone paradox" derive from listeners' reference to a mental pitch template, acquired through experience with the pitch range of their own voice, as well as with the voice ranges typical of their language community. These authors have reported a correspondence between perceptual results and the upper limit of the individual voice range for a small group of selected subjects, as well as a striking difference in tritone perception between American and British listeners. The present study compared groups of Dutch, British, and American listeners on two tritone tests and also collected voice pitch data for the first two groups in a reading task. There was no within-group correlation of perceptual results with individual differences in voice range. Differences in tritone perception as a function of stimulus characteristics (spectral envelope) were much larger than reported by Deutsch, which casts doubt on the notion of stable individual pitch templates. A significant difference between British and American listeners, with the Dutch group in between, was found in one of the two tritone tests but not in the other. Although the origin of this difference remains unclear, it seems unlikely that it has anything to do with regional differences in voice pitch range.