Monday, December 1, 2008

Perception of rhythmic grouping depends on auditory experience

Iverson, J.R., Patel, A.D., Ohgushi, K. (2008) Perception of rhythmic grouping depends on auditory experience. Acoustical Society of America 124(4), 2263-2271.

By Richard Burrows

“The perception of rhythm is central to how we find structure and meaning in speech and music.” The natural cognitive perception is to group rhythms into higher-level patterns. The fundamental question for the article asks, “if grouping is an innate building block of perception or if instead it is learned from the environment.”
Previous evidence has supported basic auditory perception as a bottom-up operation. Two principles are widely accepted 1) A louder sound tend to mark the beginning of a group, 2) a lengthened sound (or interval between sounds) tends to mark the end of a group.
The questions of grouping principles may be determined by experience has been argued by Jakobson in 1952. There are several reasons for experience playing a role in shaping perception. The first is a discovery by Kusumoto and Moreton in 1997 how found a difference in rhythmic grouping in simple tone sequences for American and Japanese listeners. There is further research to suggest that experience with language’s rhythm has a permanent influence on rhythmic interpretation. There is also research where infants show a remarkable sensitivity to the rhythms of their native language and music.
The experiment tests grouping preferences by presenting sequences of two alternating tones in which the second tone had either increased amplitude or duration. Participants were asked to report their perceptions of either long-short or short-long, and loud-soft or soft-loud. The results showed the Japanese having a preference for loud-soft but did not indicate any significant findings within the amplitude profile, however English participants showed a strong preference for short-long groupings. “A salient finding was that the most common Japanese preference was for long-short grouping, a choice made by virtually no English listeners.”
It is suggested, based on these findings, that one can predict the grouping preferences in other cultures based on the structure of their native language. However, when making predictions there are two factors that need to be addressed. First, “the temporal structure of common words may also be important in shaping of common words may also be important in shaping nonlinguistic grouping biases.” The second factor involves multilingualism. Exposure to a second language may have an influence on the preference of rhythmic grouping.

This experiment was very rigorous in its methodology. Multiple factors were addressed for variances in data analysis and contained a detailed general discussion on implications of their findings. They referenced many other researchers to back up their hypothesis and quoted where necessary, Overall, I feel this article was very succinct and thought provoking

The idea of a universal rhythmic perception is fascinating. I never really thought about language influencing our understanding of rhythm, but it does make sense. When I think about certain styles of music from certain cultures, it takes on a whole new level of understanding when referencing the native language to the rhythms. Dr. Patel had mentioned in his lecture, if you take a Mandarin lyric a place it to Western melody it does not make sense anymore. People who speak the language are unable to understand. Now this is more specifically looking at the pitch changes, but I would postulate this has the same relationship to language and rhythm.


Lee said...

Do you think this is also developmental and somewhat innate? or developmental only as in enculturation. For example, Mary Louise Serafine's research seems to indicate there is a developmental aspect to what we think is purely enculturation.


Adam Golding said...

It's worth noting that the article, as quoted, asks the question in a bit of an extreme form:

“if grouping is an innate building block of perception or if instead it is learned from the environment.”

The brain would require *some* innate grouping ability in order to learn new grouping facts. For instance, if it learned to group notes that more often follow each other in melodies, it would have to keep track of how often various groupings occur, which would require it to be able to make groupings.

That being said, they probably (hopefully) have the more the more moderate question in mind of asking to what degree experience shapes grouping preferences. Some of the classic 'rules' describing grouping preferences are in Lehrdahl and Jackendoff's "A Generative Theory of Tonal Music", not to mention the even older stuff from Gestalt psychology, and it would be interesting to test such rules for cross cultural validity, some of which is probably being done, from the sounds of this article.

The Rhythmic grouping rules commonly looked at are nicely summarized in the Grove articles on rhythm.