Music & Language Studies

SERAFINE, Mary Louise, Robert G. Crowder, and Bruno H. Repp. “Integration of Melody and Text in Memory for Songs.” Cognition 16 (1984): 285-303.

The authors conducted experiments to ascertain how memory for new songs is encoded in the mind. Their approach was to build on existing research regarding linguistic and musical memory, to address the novel question of memory for songs.

Indeed, research on hemispheric differentiation, especially that which suggests left-hemisphere dominance for language and right-hemisphere dominance for music… leaves entirely open how melody and text in songs might be processed. (286)

They concluded

that melody and text are integrated in memory to a considerable degree. …Moreover, we found no evidence that subjects can voluntarily reduce the degree of integration of melody and text. (300)

Further, they note the natural asymmetry between melodies and spoken texts, that melodies can exist without words, but speech always contains some form of intonation.

In the context of a song, the musical tune in large measure takes over the function of prosody and thus becomes an aspect of the suprasegmental properties of the words. Viewed in this way, it is quite conceivable that memory for tunes is more dependent on memory for words than vice versa; certainly, outside the realm of music the prosody of speech is remembered, if at all, only as an aspect of the words by which it is carried. We hope to investigate this interesting parallel between speech and music in future experiments. (301)

It is clear that further research needs to be done in teasing apart the cognitive differences between these two domains. Is it that, when anticipating speech, we process the melodic contour and details similarly or fundamentally differently than when in the context of singing? If similarly, then what is the difference between speaking and singing? If differently, then how do our minds prepare for the distinct expectations? Further, it remains to be experimentally shown what the authors assert, that the prosodic features of speech are not processed independently, but are fully integrated with the words. If so, it is an intriguing question just how we are able to suppress this natural encoding, in the context of singing, and replace the prosodic layer of speech with an artificial, musical melody.