ATTEMPTS TO EXPLAIN HARMONY PERCEPTION SINCE Helmholtz (1877/1954) have relied primarily on psychoacoustical models of the dissonance among the partials of chord tones. Those models are successful in explaining interval perception and the interval structure of common scales, but do not account for even the basics of triadic harmony. By introducing a 3-tone "tension" factor, I show how the sonority of the triads of diatonic music can be explained.Moreover, the relative size of the intervals among the partials in triads determines the major/minor modality of chords: major chords have a predominance of larger lower intervals, while minor chords have a predominance of smaller lower intervals. Finally, by invoking the "frequency code" known from linguistics and ethology, the positive/negative valence of the major/minor chords is shown to have an acoustical basis. I conclude that the perception of harmony can be explained by the acoustical structure of triads, without invoking cultural factors.