Haskins Publications

Number	545
Year	1984
Drawer	9
Entry Date	11/18/1999
Authors	Serafine, M. L., Crowder, R. G. & Repp, B. H.
Contact
Publication	Cognition, 16, 285-303.
url	http://www.haskins.yale.edu/Reprints/HL0545.pdf
Abstract	Conducted 2 experiments to determine the degree to which melody and text are independent, integrated, or nonseparable in memory for songs. In Exp I, 32 undergraduates heard a serial presentation of a male voice singing unfamiliar folksongs. Subjects then completed a recognition test requiring them to identify songs, melodies, or texts of (1) exact songs heard in the presentation, (2) new songs, (3) old tunes with new words, (4) new tunes with old words, and (5) old tunes with old words that had been sung to a different tune in the original presentation (mismatched songs). Results indicate that Subjects' recognition of components was higher in exact than in mismatched songs, supporting the integration hypothesis. Melody recognition was near chance unless the original words were presented. In Exp II, 48 undergraduates listened to excerpts by either a male or female voice, and 24 Ss were told only to attend to the melody. Findings show that the integration of melody and text still occurred across different performance renditions of a song and that it was not eliminated by voluntary attention to the melody.
Notes