Effect of tune repetition on judgments of preference in nontonal music

Abstract

Previous studies have demonstrated a mere exposure effect (MEE) with unfamiliar tonal music. In this study, my goal was to investigate whether a MEE could be obtained with nontonal tunes. The fluency misattribution (Bornstein, 1989) and the affective precedence (Zajonc, 1980, 1984) accounts of MEE predict different outcomes for nontonal tunes in the MEE paradigm. Specifically, the affective precedence hypothesis predicts no relationship between increased fluency in processing these tunes and the extent of MEE and suggests that MEE should be observed for all tune types. On the contrary, the fluency misattribution hypothesis argues that MEE should be related to the extent of the increased fluency in processing the tunes. As a measure of the fluency I employed familiarity ratings. In Experiment I, nonmusicians gave familiarity ratings to 4 times repeated presentations of unfamiliar (tonal), semi-random (nontonal), or fully random (nontonal) tunes during the exposure phase. A subsequent forced-choice liking task measured their preference for old vs. new tunes. The second experiment was identical except that the way tunes were repeated during exposure was varied. The results showed a MEE for fully random and unfamiliar tunes in Experiment 1, and for semi-random and unfamiliar melodies in Experiment 2. Further, familiarity ratings for all tune types showed increases with more exposures, however, the extent of this effect differed for tune types between Experiment 1 and Experiment 2. These results clearly contradict the predictions of the fluency misattribution hypothesis, however, are also problematic for the affective precedence hypothesis at some points.