A fundamental insight of Leonard Meyer’s approach to melody perception is
his notion of reversal. Defined as a marked deflection in an ongoing
pattern, reversal was a syntactic necessity for Meyer, without which
ensuing resting points would feel incomplete (Meyer, 1980). In this
exploratory study, we frame Meyer’s theory as an empirical question: do
listeners judge musical phrases with marked reversals to be more complete
than those without? In a forced-choice paradigm, music students listened to
pairs of synthesized phrases with matched beginnings and endings but
different kinds of reversal in their continuations, selecting which phrase
felt more complete. A conjoint analysis modeling selection behavior as a
function of reversal type (pitch, rhythm, and both pitch and rhythm)
compared to the baseline case of no reversal revealed a significant
preference for reversal melodies, F(3,2192)=6.05,p<.01. Average marginal
component effects demonstrated that phrases with pitch reversals, as well
as phrases with both pitch and rhythm reversals, significantly increased
likelihood of selection relative to no reversal (p<.05 and p<.01,
respectively). The effect of solely rhythm reversal did not reach, but
approached, significance. These preliminary results are consistent with
Meyer’s insight that melodic reversals result in feelings of phrase-ending
completeness, and future directions are considered.