The Mafa people, who live in the far north of Cameroon in the Mandara mountains, are one of the most culturally isolated groups in the world. Since many of their settlements lack electricity, there are some individuals who have never been exposed to western movies, art, or music.
But the Mafa do have their own musical tradition. Many of their ceremonies are accompanied by a unique chorus of flutes of varying sizes, which can produce different pitches by covering and uncovering a small hole at their tip. The music they produce is quite different from Western-style music. Here's a sample:
Because of their isolation and very different musical tradition, they can help answer a question that has perplexed music scholars and psychologists for generations: are there musical "universals"? In other words, do the emotions conveyed by music depend on what we've learned through our culture, or can anyone perceive the emotion intended by a composer of a given musical work? Does "good" or pleasant music have cultural boundaries?
A team led by Thomas Fritz visited the Mafa people and played excerpts from Western music intended to evoke one of three emotions: happiness, sadness, or fear. The listeners were pre-screened to make sure they had never been exposed to Western music. The experimenters showed the listeners images of faces expressing each of these emotions, and asked the listeners to point to the face that best represented the emotion conveyed by each excerpt. Here are the results, as compared to the results for the same clips as identified by German listeners:
Both Western and non-Western listeners recognized the intended emotion at rates significantly higher than chance (the dotted line on the chart). So even people who've never been exposed to Western music can understand the intended emotion.
The reverse experiment, playing Mafa music to Westerners, would have been impossible, because Mafa don't traditionally associate particular emotions with their music. However, the researchers could test whether Westerners could distinguish between "good" and "bad" Mafa music. The clip you heard above is an example of traditional Mafa music. Here's another clip that's been digitally modified to sound dissonant:
And here's an unmodified Western music clip:
Here's a dissonant Western clip:
Both the German and Mafa listeners heard dozens of clips like this -- some as they were originally played, and some modified to be dissonant. They then rated each clip using a slider marked with a smiley face (good) and a frustrated face (bad). Here are the results:
Although the differences were small for Mafa listeners, they rated the unmodified clips significantly higher than the modified clips, both for Western and Mafa music. For Western listeners, a similar, though stronger result occurred.
The researchers say this shows that at least some portion of music appreciation may be universal. The more dramatic effects among Western listeners suggest that cultural influences may also contribute, but some basic part of the the way we understand music may be shared by everyone, no matter what we have learned from our culture.
Fritz, T., Jentschke, S., Gosselin, N., Sammler, D., Peretz, I., Turner, R., Friederici, A., & Koelsch, S. (2009). Universal Recognition of Three Basic Emotions in Music Current Biology DOI: 10.1016/j.cub.2009.02.058
- Log in to post comments
That is interesting--and I especially liked being able to listen to the clips.
That's a very important result if it holds up - someone needs to replicate it with another isolated tribe (before they all start drinking Coke and listening to Britney...)
Something seems to have gone wrong with the clips though - I can't see the links.
They're not links, but embedded... you need a music-player plugin
Very interesting. Was the music in the first (happy/sad/scary) experiment similar in tempo, instrumentation and so on?
Hm. In both cases I prefer the dissonant clip to the consonant. Perhaps I'm an alien.
I like their music. Interesting sound to it.
The fact that the Westerners have a greater range of likes and dislikes to the "dissonant" music could show that the researcher's idea of good and bad music was not universal. In western music, which is based very much on melody and harmony, dissonant notes stand out. But I noticed that the rhythm was not changed. What would be a better experiment would be using two full-fledged songs, a good and a bad (may I suggest Mary-Kate and Ashley?). There is always going to be an element of the subjective here...
Doesn't this study also brings forward a point : how educatio shape our critical abilities.
A mafa listener is less prone to actually rate as good or bad a music clip compared to a westerner that, as the study seems to suggest, will not hesitate in giving rating.
A westerner has, for obvious reason, be submitted more regularly to music than a mafa what with radio, TV, IPOd and the likes that are constant source of musical education.
Couldn't this say something about the advent of democracy ? When the level of education is raised, follows naturally a desire for democracy in the sense that people are acquiring the faculty to criticize their society's political regime.
Now, a question that is not raised in this study who is getting the better kick out of listening to music. Is the music virgin mafa more happy upon hearing a tune or, on contrary, is it the westerner.
In short are our "pleasure center" directly correlated to our ability to reflect on the "meaning of life" or not at all ?
Please update it to include the sample size.
Really interesting post. Thanks! I've linked to it in "Neurartic", along with an image of an extraordinary contemporary Wafa earthenware vessel. The making of art is a universally human cultural activity. Fascinating to read research getting at the underlying commonalities in how we humans perceive, and think about art.
ARF: Sample size was around 20 Westerners and 20 Mafa for each of the experiments.
Makes natural sense as the brain/mind is wired to interpet sounds this way.
Its the same for animals. For example a growl is universally aggressive for humans and mammal animals alike. The brain interprets these sounds in the way it does, as its a survival mechanism that determines behavior, such as flight, fight, freeze or ignore. Music is just a sophisticated extension of this concept.
It may be that the bigger takeaway from this study is the dramatic difference in Westerners' readiness to label the music as "good" and "bad" relative to the Mafa - even judging the Mafa's own music much more definitively than the Mafa themselves!
What does this say about Western propensity for labelling? Why are we so quick to tag something as "good" or "bad" compared to the Mafa? Is it because we're so bombarded with inputs that we learn to identify patterns quickly and assign values? Or is there a deeper dynamic... that we're taught at an early age to fit what we see into constructs, and if they don't fit, to view them as "bad"?
I think there's a lot of food for thought in these findings, and possibly, for further research.
Did anyone else think this was talking about a cell culture and was VERY confused and amazed to begin with?
Now I will be humming Mafa music all day. That was catchy. Is there a longer version somewhere?
What strikes me is that the Westerners claimed to like/dislike the two music styles equally, whereas the Mafa had much stronger reactions to the Mafa music.
I can think of a couple of scenarios which would explain this:
(1) The Westerners are making a conscious effort to avoid the appearance of ethnocentrism and are therefore choosing to rank consonant music and dissonant music equally regardless of source, choosing the midway "0.5" for every answer. Meanwhile, the Mafa naturally prefer their own music and are honest about liking original Mafa music more than original Western music (and about disliking more strongly Mafa music that's been tampered with).
(2) The opposite. The Mafa want to seem polite, and so, not being familiar with Western music, do not want to offend by being adamant about liking/disliking Western music (and potentially getting the answer "wrong", thinking perhaps the Westerners prefer the dissonance and we'll look naive) and so they only move the sliders a little bit. But with their own music, being more familiar with the original, they can tell when it's off. Meanwhile, "Western music" encompasses so many different genres, that the Westerners have been exposed to a wide enough range of styles that they are better equipped to analyze a new musical style abstractly, focusing on its harmonics and not, say, the instrumentation.
(2b) Western music is so broad, that some of the responders genuinely like that sort of 60s/70s R&B sound (or is it the SNL band?), while others really don't, and the answers just happened to average out. Meanwhile Mafa music is much more limited in its variety and so everyone was equally accustomed to enjoying the Mafa clip they played (but, as Mafa music is not intended to evoke particular emotions, they were not so moved by it either way, only making it up to the 0.2 mark).
It would be interesting to see the individual responses, or at least the variances and covariances on the Westerner results. Were the responses to the two originals positively correlated, so some Westerners really liked both, or did it go the other way, where some really liked the Mafa music but not the Western music, with others the reverse? If someone really liked the original, was it necessarily the case that he really disliked the modified version?
I'll say that I follow the basic trends of the graphs and like the originals and dislike the modified samples, but not equally. I like the original Western clip more than I like the original Mafa clip, but I dislike the dissonant Mafa clip MORE than I dislike the dissonant Western clip.
Er, I kinda liked all of them. My top choice would probably be the "dissonant Western" (but then, I'm a dissident Western).
"Paging Pierre Bourdieu. Pierre Bourdieu to the Sb courtesy phone, please."
But the comparison wasn't between consonant and dissonant music, it was between music and digitally-modified sounds designed to sound dissonant. Of course the real music was preferred.
A tripartite taxonomy of emotions is far too coarse for these results to mean anything major. "Happy," "Sad" and "Scared" are conflations of many different emotional states. If the study had done some investigation of Mafa emotional taxonomy and had attempted to cross-associate aspects of Western music with whatever structure the Mafa used, the results would have greater significance. (A cross-cultural study of emotional states and their associated taxonomies was done in the 70s by Manfred Clynes, who wrote about it in a book called "Sentics.")
I agree with what eff said up there about the criteria for telling good music from bad. WarrenS also has a valid point.
On a related note, regarding culturally conditioned approaches to the criterion involved in this research, 'likeability' may not necessarily be the best criterion with which to judge universality, considering the role played by 'music' within a traditional Mafa society.
Western music's primary function has arguably been shifted to focus on the factor of likeability - music is generally thought to define an individual's aesthetic taste. In community-oriented societies like the Mafa, music tends to take on more pronouncedly social and overtly functional overtones. Those two approaches could be seen to be in stark contrast with one another. The very IDEA of music is radically different in those two cases.
Hence it is conceivable that the Mafa would, upon hearing music meant for a particular occasion and being asked to rate for likeability, refer back to the occasion and state their emotions towards the occasion evoked, instead of responding to the music for the music's own sake. On the other hand, I assume their Western counterparts would simply state their emotional responses towards what they perceive to be the purely aesthetic value of the music involved.
(Of course, the Westerner could be responding to memories evoked by the music in a similar train of reflection. But even then the nature of the images evoked would be different, in that the function for which music in the Western sense is expressly employed differs drastically from that for which music in the Mafa sense is. -> socially shared and inseparable from everyday life in the case of the Mafa vs. personal, emotional, and individually separate in the case of the Westerner.)
Control of such factors would require either a) exposure to a series of cognitive steps meant to simulate musical paradigm shifts of a fundamental degree for the control groups involved, which would soundly defeat the point, or b) lots and lots and LOTS of cross-cultural fieldwork carried out with multiple hypotheses concerning music's various social functions/ social judgment criteria/ etc. It would take no small amount of intricacy to work out the kinks, but there probably should be a way.
Overall, this seems to be an interesting study that throws out many important questions, though I'm afraid it asks far more than it answers.
And to be perfectly honest, I'm rather biased against studies that focus on some 'isolated human populations' and work under the impression that there are elements of a universal humanity to be 'extracted' from them through research somehow. First of all, you don't NEED to go that far to find cultural difference, it's all over the place; secondly, isolated groups are just as unique as they are universal insofar as they are human, I suppose. Although it's all a matter of degree, and I'd readily concur that gradational differences lead up to stark oppositions at extremes. Dunno. Anyhow, this is my immediate response to the findings. Cheers.
The western music clip is the intro from the song "The Chicken" by the Jaco Pastorius Big Band.
I'm curious whether the happy/sad/scared songs attempted to equalize variables like tempo and volume, and only differentiate on melody. I'd think that "slow and quiet" would be obvious universals that correspond to "sad".
As for the dissonance test, that doesn't surprise me either; there are mathematics behind harmonic pitch relationships, and that should be universal. I think a more interesting test would be to alter a Western song such that it had rapid key changes; most Westerners would find that unpleasant, but someone not accustomed to our equal-tempered scale might not notice.
...or it could show that we are just able to detect horrific digital manipulation of sound clips. It is patently obvious from the provided clips that the "dissonant" music clips are "good clips" horrendously processed. The instruments and tonal separation is no longer clean, and it sounds like phased mush - which is itself unpleasant. Unless the sound clips exhibit exactly the same acoustic (as opposed to musical/melodic) characteristics, then how can you be really sure what reaction you are measuring?