The harmonious ape

August 1, 2009 • 11:48 am

Speaking of animals and dissonance, an upcoming article in the journal Primates investigates the question of whether our closest living relative, the chimp, shows a preference for consonant over dissonant music.  The answer is yes, suggesting that the human preference for consonant music lies in our genes.  But be aware that this result is based on just a single chimp.

Background:  Humans prefer consonant over dissonant music, as consonance “evokes a pleasant feeling.”  The fact that the human preference is seen in infants as young as two days old suggests that this preference is inborn rather than learned.  Animals as distantly related as birds can distinguish between the two types of music, but studies of another primate, the cotton-top tamarin (Saguinus oedipus) showed no preference for consonance.  The authors of this study looked at a closer relative to see if perhaps the trait “preference for consonance” might not only be genetically based, but also have appeared in the primate lineage more recently.

Methods and materials:  The authors studied, over six weeks, a single five-month-old female chimpanzee (Pan troglodytes) named Sakura. She had been reared in a zoo by humans, as her mother had rejected her.  The authors claim that she had never been “exposed to any particular music source such as a radio, TV, or CD player throughout her development. . ”

Sakura was strapped onto a bed with a soft belt, and was forced to listen to either consonant or dissonant music.  There were six sets of computer-generated music, each set consisting of a consonant piece and a nearly-identical dissonant piece, in which some of the notes had been swapped for dissonant ones (see paper for further details).  A string was attached to her right arm (see Fig. 1), and by pulling this string she could affect whether the music stayed the same or changed to the alternative version (i.e., consonant to dissonant or vice versa).  The recorded music began. If the chimp pulled the string within seven seconds, the same music would continue playing for up to two minutes.  If she pulled the string between 7 and 14 seconds, the music would also continue, but with a pause of a few seconds.  If she waited longer than 14 seconds to pull the string, the music would change to the alternative version. Sakura was tested once per week for the six week period.


Fig. 1. Sakura in bed, choosing the music that appeals (from original paper).

Results:  The mean duration of consonant music sessions for Sakura was 24.6 seconds, but only 15.9 seconds for the dissonant music. Statistical analysis of the sessions (each including about 10 bouts of each type of music) showed that this difference was significant.  That is, there was a significant preference for the consonant tunes over their dissonant alternatives. When Schoenberg was played, Sakura became very agitated and flung feces at the observers (kidding!)

What it means:  The authors interpret this result as suggesting that “one of the major factors that constitute musical appreciation might not be unique to humans; instead it might be something that we share with our phylogenetically closest relatives.” That is, dissonant tones may affect the nervous systems of chimps and humans in similar ways, and lead to similar subjective sensations.  That seems to be a reasonable conclusion, although of course we need studies where the number of primates exceeds one.

The authors note one potential flaw in the work: suppose that the dissonant music caused the chimp to relax more.  Then she would be less inclined to pull the string when that music was playing, and according to the experimental protocol the music would then change over to the consonant form. This could give the impression that she preferred the consonant over dissonant music as a simple experimental artifact.  The authors say that they consider this possibility “unlikely,” but they didn’t control for it.  One way would be to change the experimental protocol in a separate experiment so that the music would change over when the string was pulled quickly rather than after a longer interval.

One question the authors don’t bring up is whether animals of any sort produce consonant rather than dissonant music.  Do birds, for example, tend to sing consonantly rather than dissonantly?  I am a music tyro and don’t know the answer.  However, the efficacy of animal communication may rest on things other than whether it’s pleasant for them to hear.

T. Sugimoto, H. Kobayashi, N. Nobuyoshi, Y. Kiriyama, H. Takeshita, T. Nakamura, and K. Hashiya.  2009. Preference for consonant music over dissonant music by an infant chimpanzee. Primates, in press.

12 thoughts on “The harmonious ape

  1. To me the funniest part is the “soft belt” of Sakura! Look at it! It has bees on it! (I think)

    I don’t know why, but I find that just hilarious

  2. I’m glad you found this interesting study and hopefully the research will be expanded. I wonder had older chimps would react.

    What would have happened if the music of Charles Ives had been used?

  3. Fascinating. I think it would be interesting to see if by sequencing the genomes of people who have severe amusia we could find some of the genes responsible for musical appreciation and talent.

  4. Fantastic – thanks for posting this Jerry. I love reading about scientific research on music and musical perception.

    I wonder if there’s a site where we can hear for ourselves what the researchers considered to be consonant or dissonant music? If you take the seven-note diatonic scale (do, re, mi, etc.), all the intervals can be ranked from most consonant to most dissonant (there’s pretty broad, but not complete, agreement on those judgments). Everyone agrees that two notes a half-step apart played simultaneously is extremely dissonant, though if they’re played in succession as part of a familiar chord progression or cadence (the most familiar of these in Western music are pervasive and everywhere, from classical to rock, pop, and blues), they can sound perfectly consonant to our ears.

    Did the researchers play single notes in succession (melody), or chords and chord progressions (harmony), or both? What were the parameters given to the computer program to determine consonant vs. dissonant intervals? What musical scales were used? The Western musical scales most of us are familiar with (major and minor), are different from Arabic scales, pentatonic scales common in Chinese music, and so on.

    I suspect there is some innate human preference for consonance over dissonance that transcends cross-cultural differences in musical scales and harmonic progressions. But the variation in these around the world brings up the question of how elastic the concepts of consonance and dissonance are. I’m very curious about the methodology the researchers used. Guess I’d better read the paper!

    1. Good point, my partner shows every sign of enjoying hours of Cantonese opera at a stretch, whereas I find it . . . challenging. But then he’s Chinese.

      1. These are all very good points. I have to say that I’m not at all convinced this experiment controlled for these variables. I’ve been a musician and composer for years, and have seen my own definitions of consonance and dissonance (as well as my preferences for one or the other) change in myriad ways. I accept that people who know little about the technical aspects of music use the terms consonance and dissonance fluidly and subjectively, but that’s simply unacceptable in a scientific experiment.

        For instance, it seems highly problematic to me that they apparently aren’t controlling for degree of dissonance. As the original poster points out, although there are some intervals or harmonies which we more or less agree about (and even here I’m not at all sure that this crosses cultural lines), there are certainly plenty of intervals and harmonies which are only slightly dissonant or fairly consonant. For instance, the tritone (aka the devil’s interval) has long been considered a strong dissonance in western culture, but is actually a primary component of the music of Bach, Mozart, Beethoven, etc. Such composers, I assume, would be called “consonant” composers by musical amateurs, but dissonance is a primary component without which the music would lose all teleological drive.

        Finally, it seems very problematic to me that they simply took “consonant” music and then tweaked it for the dissonant version. This probably just made for one version that sounded right and one version with wrong notes. This is generally not how composers write highly dissonant music, and it ignores all context. Perhaps the experimental subject was simply reacting to out-of-context notes rather than consonance and dissonance specifically. This might still be an interesting result, but it’s far from applying a sheen of objectivity to the question the experimenters claim to be investigating.

  5. In other news, a Techno Chimp prefers Techno and a Heavy Metal Chimp prefers hard rock thus proving, regardless of chimp choice in music, that enjoyment of sounds is a genetic thing.

Leave a Reply