Melodically accented tones are perceived as having higher intensity. Subjects judged whether or not all of the tones in a sequence were of equal intensity. Melodic accents were created by pitch skips, and the intensity of the tones that followed the skips was manipulated. Introduction of the melodic accents reduced detection of lower intensity tones more than it reduced the detection of higher intensity tones. This effect did not change as a result of regular or irregular timing of the tones whose intensity was manipulated. Contrary to results from an earlier experiment in which listeners tried to detect variations in timing, placement of the melodic accents at regular intervals did not lead to poorer detection than did irregular placement of the melodic accents. It is concluded that perception of the gaps that separate rhythmic groups may have a different nature different from the perception of elements within groups.

If one dimension of sound is manipulated in a way that suggests a particular rhythmic organization, does perception of other dimensions change in ways that are consistent with the same rhythmic organization? When subjects were asked to judge or adjust intensities of tones, rhythmic manipulations of pitch structure changed the perception of intensity. When subjects were asked to judge timing, rhythmic manipulations of intensity had a similar effect. Timing manipulations did not have an effect on judgments of pitch. The results indicate that temporal structure as a whole is more accessible than the individual physical manipulations that give rise to that structure. It may be concluded that the temporal structure itself, rather than pitches, intensities, and durations in isolation, is a perceptual object.

The effects of variations in three pairs of variables on identification and discrimination of rhythmic patterns in pure-tone sequences were investigated. It was found that adding a timing difference to two sequences that differed in terms of the pattern of tone intensities improved discrimination if variations of the two variables converged on the same rhythmic pattern but did not help if the variation in timing undermined the rhythmic pattern created by the intensity variation. Adding a difference in the intensities of tones to two sequences that differed in terms of the pitch structure produced a similar pattern of results. The effect of adding a difference in pitch structure to two sequences that differed in terms of timing was not reliably related to the way the differences were combined. The results showed that, at least for some variables, predicting discrimination performance from probabilities of detecting a rhythmic pattern is possible. The relationship of the results to cue trading in speech perception and research possibilities with similar methods are discussed.