Tone of Voice and Mind

It is well established that pitch changes in the voice during normal speech carry information about the affective state of the speaker. Moreover, it is known that the right hemisphere is more sensitive to pitch information (melody and harmony) than the left, and is also more capable of perceiving emotions, while the left hemisphere is better at denotative and syntactic processing. These empirical facts form the backbone of a theory of interhemispheric communications that explains not only the highest-level division of labor of the cerebral hemispheres during language processing, but also why there are resolved (major and minor) and unresolved (augmented and diminished) harmonies and indeed why the "ring" of major chords is generally considered to be "bright" and that of minor chords "dark". These ancient problems in music perception can be neatly resolved on the basis of psychophysical principles if the relative size of the intervals contained within three-tone chords is brought into consideration. For a technical discussion, see these papers:

Eurospeech(2003), ICASSP(2004), SpeechProsody(2004), Interspeech(2005), EMR(2006),

EMR(2007), MusicPerception(2007), MusicPerception(2009)

Linguists have devised many notational schemes to describe the pitch changes in speech, but only recently have interval combinations been examined. Although the changes in pitch during normal speech do not of course have the musicality of song, it is of interest that the basic pitch phenomena underlying diatonic music can also be found in speech. For details, see "Tone of Voice and Mind: The connections between intonation, emotion, cognition and consciousness" (John Benjamins, Amsterdam, 2002) and this 2006 paper in IEEE Transactions on Speech and Audio Processing.

Return to Top Page