Recognition memory is better for vocal melodies than instrumental melodies. Here we examine whether this vocal advantage extends to recall. Thirty-one violinists learned four melodies (28 notes, 16 s), two produced by voice and two by violin. Their task was to listen to each melody and then immediately sing (for vocal stimuli) or play back on violin (for violin stimuli) the melody. Recall of the melody was tested in ten consecutive trials. After a brief delay (∼10 min), participants were asked to perform the four melodies from memory. Each performance was scored based on the accuracy of two measures: (1) intervals and (2) contour. The results revealed an advantage for vocal over violin melodies in immediate recall of the melodic contour and, after the delay, a reverse pattern with an advantage for violin over vocal melodies. The findings are consistent with the hypothesis that the voice facilitates learning of melodies and further show that the vocal advantage in recall is short-lived and based on contour.

You do not currently have access to this content.