Music Perception: What Is It That We Perceive When We Perceive Music?

15 October, 2006
One of the things that music scientists study about music is music perception. But "music perception" is the perception of music, and music is something that we don't know what it is. In fact, music appears to be something created entirely so that we can perceive it. Which is all very circular.

Defining "Music Perception"

What is music perception? This might sound like a very silly question. Surely "music perception" is just the perception of music. Which is true enough. But the problem with defining "music perception" as being the perception of music is that we don't know what music is. Furthermore, music does not appear to have any existence independent of the fact that we perceive it – we create music so that we can perceive it.

Thus music perception is the perception of music, music is something which is created in order to be perceived as music, and neither concept can be defined independently of the other.

Perceiving the Features of Music

Music does have a number of objectively identifiable features, and these include such things as melody, rhythm, harmony and repetition. We could say that music perception is the perception of these features, and often "music perception" is measured and studied as though it is the ability to perceive these features of music. For example, the melody component of music perception can be measured according to a subject's ability to identify a melody and distinguish it from other melodies.

The only problem with this understanding of music perception is that properties like melody and rhythm can also be perceived when listening to sounds which are not music. For example, what is melody? We could define it as a sequence of notes, or, if we don't want to restrict ourselves to sounds with constant pitch values, we could define it in terms of pitch as a function of time. Given this definition, there are plenty of non-musical sounds which have a "melody" which can be perceived, identified and distinguished from other "melodies" (a major example is speech "melody", which, depending on your language, might be determined by intonation, pitch accent or tonality). But if we are listening to a non-musical melody, presumably we are not engaging in "music perception", even though we are engaged in "melody perception".

The same applies to all other known features of music, which leads us to conclude that music perception cannot just be the sum total of the perceptions of all its component features, because each and every one of those features can be observed in some form in non-musical sound, and when a person listens to a non-musical sound, there is presumably no "music perception" going on.

The Perception of "Musicality"

There is one thing which we do perceive when we are listening to music, which we don't perceive when we are listening to non-musical sound, which is precisely that it is music.

It is easier to discuss this property of being music if we give it a name, so I will give it a name and I will call it musicality.

This leads to the hypothesis that the perception of music is the perception of musicality. As far as I can tell, this hypothesis is not given much consideration in the world of music science, because it implies the question "What makes some music better than other music?", and that question seems to be treated as a question you are not allowed to ask, perhaps because it is politically incorrect if applied to music from different cultures, or perhaps because it is perceived to ignore personal variation in musical taste.

(One exception: the question of what makes music good music is taken seriously in the world of commercial music science, where "musical intelligence" companies promise to use secret methods to predict whether or not your composition will become a big hit. Unfortunately commercial music science research takes place behind closed walls, so the researchers are not engaged in the public conversation about how to explain music scientifically. Their research also seems to rely on brute force statistical approaches disconnected from any attempt to explain music within a biological framework.)

An Analogy: The Perception of Beauty

Beauty is the quality that we perceive when we perceive something that is beautiful, for example the face of a beautiful woman. We know that facial beauty is a function of things like skin complexion, cheek bones, shape of nose and general facial proportions (this is not a complete list), and the perceptions of these features are a necessary prerequisite to the perception of beauty. But they are not, either singly or all together, sufficient for the perception of beauty. We can imagine the existence of someone who can perceive skin complexion, cheek bones, nose shape and facial proportions, as measured by their ability to describe those features consistently and accurately, and at the same time imagine that this person cannot perceive beauty, and cannot tell whether or not a given face is beautiful. It is quite possible that such a perceptual deficiency might result from damage to a specific brain area, i.e. a "beauty-detecting" brain module, although I do not know if a neurological deficit of this kind has ever been observed and recorded by the medical profession.

We can also have some idea of the meaning of beauty, at least as far as the faces of beautiful women are concerned. The most plausible explanation of the meaning of beauty in this context (and one that I think is widely accepted scientifically) is that beauty is one component of sexual attractiveness, and it is an estimate of the suitability or desirability of a person as a short-term or long-term sexual partner. This explanation has to be an over-simplification, because most of us can perceive the beauty of a member of the sex that we are not sexually attracted to, so we might have to qualify the explanation by adding that perceived beauty represents our perception of how much other people (as well as ourselves) find a person to be a desirable partner.

In this respect, our scientific understanding of the meaning of facial beauty is miles ahead of our understanding of the meaning of musicality, because we do not have any plausible explanation for what musicality means. (Well actually I think I do, but for the sake of this discussion I'll ignore my own theories on the subject.)

Every Theory About Music is a Theory about Music Perception

Every theory about music is a theory about what it is that we perceive when we perceive music, and what the meaning is of whatever it is that we do perceive. Or at least it should be. Many theoretical discussions about music somehow end up avoiding these questions, and many theories are presented which don't appear to answer them at all.

In the following sections I consider some "popular" scientific theories of music, and I state what I think they say about what is perceived when we perceive music and what it must mean according to each theory.

Sexual Selection Theory of Music

According to this theory, music perception is selected for when a woman perceives that a man is good at making music, so therefore he is a desirable sexual partner. The man makes music so that the woman will perceive him as attractive, and he also needs to be able to perceive music in order to get feedback about his own performance (i.e. when he practices).

Thus, according to this theory, "music perception" is the perception of how good a potential partner is at performing music.

As you might notice, this theory says nothing at all about what the features of music would be.

It can be claimed that this is not necessarily a problem – the theory of sexual selection can explain the evolution of sexual characteristics that are somewhat arbitrary, such as brightly coloured plumage on male birds. But, given the many different aspects of music that exist, I think that an explanation of music based on sexual selection needs to go into more detail about why music has the kinds of features that it has. Even though sexual selection can explain the evolution of arbitrary new features, there are constraints on how these features and the perceptions of them are genetically determined, and these constraints have some effect on what kinds of features can evolve this way.

Reference: Evolution of human music through sexual selection by Geoffrey F Miller.

Group Selection

Group selection theories are weak and unconvincing just because they involve group selection. (Group selection is a very weak process, because evolution by natural selection is based on reproductive success, and it is almost always individuals that do the reproducing, successful or otherwise.) A theory based on group selection can become even more unconvincing when you get to the unconvincing details of how the thing being selected for, in this case music, contributes to the survival of the group.

According to the group selection theory of music, music is selected for because it requires and/or encourages group activity, i.e. singing and dancing together. Just like the sexual selection theory, this theory says nothing about why music has the features that it has (except perhaps that the features should require cooperation if group performance is to be successful). Presumably, according to this theory, music "means" something like: "you can hear music, so you should be joining in and singing along and/or dancing".

It is unclear how an instinct that encourages everyone to waste time and effort on an otherwise pointless and meaningless activity is really worth whatever level of cooperation and "togetherness" it might cause to occur. If cooperation and togetherness are so important, why can't there just be an instinct for cooperation and togetherness? Why do these social characteristics have to be encouraged by participation in a pointless activity? Why not have an instinct that encourages people in a society to cooperate in activities that aren't pointless?

Reference: I can't find any good online references that don't require use of a credit card, but The nature of music from a biological perspective (a review article by Isabelle Peretz published in Cognition) does include a discussion of this theory as a one of the major theories (page 24), and gives a few references (all of them off-line).

Music Evolved from Something Else, Or Vice Versa

Some theories that have been proposed to explain music are based on the observation that something else is a bit like music, and music might have evolved from that something, or vice versa, or they both evolved from some common ancestor. Since none of these theories asserts that music actually is the something that it resembles, they say very little about what music itself means. Two major candidates for a similar "something" are language (in general), and particular kinds of language, such as motherese.

Because they start off with something resembling music, these theories explain something about the features of music. But, unless they can be specific about what selective pressures caused music or the other thing to evolve, they say nothing about the meaning of music, other than perhaps that it used to have the same meaning as the thing that it evolved from. Or that the thing that evolved from music used to have the same meaning as whatever the meaning of music was. Or that music and the other thing used to have the same meaning before they evolved from their unknown common ancestor.

References: The "Musilanguage" Model of Music Evolution by Steven Brown, chapter 16 of The Origins of Music, and, from the same book, Antecedents of the Temporal Arts in Early Mother-Infant Interaction by Ellen Dissanayake. A more recent exploration of the "musilanguage" idea can be found in Steven Mithen's book The Singing Neanderthals: The Origins of Music, Language, Mind and Body.

Music is a "Universal Language"

This is a popular assertion, although it's hard to say if it counts as a scientific theory. It sounds good, in a warm fuzzy kind of way, because it paints a picture of people from all around the world all listening to the same music and hearing the same meaning, even though they all speak different languages.

The only problem is, that subjective or objectively, music doesn't actually have much meaning. When we listen to music, we are not receiving any useful information about anything other than the music itself. And a language without meaning isn't really a language. (Although we can consider that music has an emotional meaning, which is the subject of the next theory I consider.)

Reference: "Music is the universal language of mankind." is a quotation from the poet Henry Wadsworth Longfellow, apparently from "Outre-Mer. A Pilgrimage Beyond the Sea", but I have not been able to find it in any online version of that book.

Music Expresses Emotion

This is a theory that has some connection with our own personal subjective experience of music. Music has an emotional effect, and some music consistently expresses particular emotions. It seems plausible that emotion might be the "meaning" of music, and that when we perceive music we are perceiving emotion.

There are, however, a number of difficulties with this theory:

Thus, although the connection between music and emotion cannot be denied, and music certainly has an emotional effect, this is not enough to convince us that the perception of music is the perception of emotion.

References: because the relationship between emotion and music has been known for a long time, and discussed both historically and more recently by many, many people, it is difficult to given any definitive reference for this theory.

Conclusion: It's All Still Just a Mystery

"Music perception" will remain a popular catchphrase, and researchers will continue to research it. But we must remain aware that since we don't know what music is, or what it means (if anything), the concept of "music perception" has a very uncertain meaning. We must be careful not to think that just because we can give it a name, that we actually know what it is.

Hopefully we can keep on studying "music perception", without fooling ourselves that we really know what we mean when we say that phrase. But the danger of being fooled is always present, because the alternative to fooling ourselves is to insist on admitting our profound ignorance, and nobody wants to admit that.