Assessing the role of melodic and rhythmic factors in structuring musical salience.

In a series of previous studies (Olivetti Belardinelli 2005; Olivetti Belardinelli et al. 1999,2000) we demonstrated a pre-eminence of salience over tonality as an anchor point for musical memory, as well as a strong link between salience and episodic memory, and tonality and semantic memory. According to the operational definition we provided, salience was characterized by the redundancy of melodic and/or rhythmic parameters. Nevertheless it is not clear what is the contribution of each single factor (and their interaction) to the mnestic performance, and which one is the most effective as an anchor point for recognition memory. Therefore the aim of this study was the assessment of the influence of melodic, rhythmic and melodic-rhythmic redundancies on musical memory. 60 healthy (30 males) right-handed non-musicians participated in this study. Stimuli were 48 short musical themes made up of modified versions of those originally employed in our previous studies. Half of the new stimuli were tonal and half non-tonal, 16 were characterized by only melodic salience, 16 by only rhythmic salience, and 16 by the presence of both. Subjects were asked to listen two times to a “study list” containing 24 musical excerpts. After 15-20 min, they were administered a “test list” containing 48 stimuli (24 previously heard plus 24 new ones), and they should press a different button whether, according to Tulving’s model (1972): 1) they recognized the melody as previously heard (Remember response); 2) the melody evoked in them a sense of familiarity (Know response); 3) they could not recognize the melody at all (X response). Results showed a significant influence of all kinds of salience on episodic memory. The best memory performance occurred when both melodic and rhythmic salience were present and tonal information was absent. The worst performance was obtained when melodic salience was present and tonal information was absent, while rhythmic salience in association with nontonality scored a significantly higher performance. In conclusion, the association of both kind of salience seems to improve recognition memory, while the influence of each of them on memory performance seems to differ according to the concurrent availability of tonal information.