Music Is A Vestigial Trait

18 August, 2019
Music evolved to provide a major biological function. Then it "unevolved", and it no longer provides that function.
Update (5 Oct 2019): Since writing this article I have developed a second theory about music as a vestigial trait, in which I propose that music originally evolved as a language extension. The newer theory assumes less difference between the form of music now and the form of music when it originally evolved and served a major biological function.

Vestigial trait

A vestigial trait is a trait of a species that has evolved to serve a particular function, and then, at a later time, it has evolved (or "unevolved") not to provide that function.

What makes a trait "vestigial" is that we cannot really understand its current state without knowing what it used to be.

Ostriches, and their wings

A classic example of a vestigial trait is the ostrich wing.

Ostriches have wings, but they do not use them to fly.

Ostriches do use their wings for other things. They use their wings to help them when running, to stabilize themselves, particularly when turning sharp corners or braking (LiveScience). Other examples are courtship displays, displays of dominance and submission, and sheltering chicks from the weather (San Diego Zoo).

If the ostrich was the only bird in the world, and we were trying to understand the evolution of their wings, we might look at the current functions of those wings, and try to formulate a theory of how those wings evolved to serve the purposes that they currently serve.

But in doing this, we would be missing the main point, which is that ostrich wings used to be wings, and wings are for flying.

If ostriches were the only birds in the world (and we didn't yet know about fossils), then we wouldn't know about birds having wings and using them to fly.

In that hypothetical world, where ostriches didn't fly, and we didn't know about any flying relatives of the ostrich, it would seem crazy to think that the primary reason ostriches have wings is because of what those wings used to do. Scientists would be reluctant to abandon their detailed studies of the current functions of ostrich wings and instead spend their time on wild and uncertain speculations about the hypothetical functionality that those wings had millions of years ago.

But, in the end, speculating about the past functionality of those mysterious wings would be the only way to solve the mystery of why ostrich wings are the way they are now.

The Vestigiality of Music

I propose that human music is a vestigial trait.

In other words, music is like ostrich wings.

Although there are some differences.

The main difference is that I propose that music evolved to serve a certain biological function, and then it was replaced by a new trait, which served the same biological function, but a whole lot better.

It was as if ostriches evolved wings, and instead of just going back to running around on their legs, they evolved rocket engines, and that's why they didn't need wings any more.

To fill out the details of this hypothesis, I need to answer two main questions:

  1. What biological function did music evolve to provide?
  2. What trait evolved to replace music, and why was it so much better?

The original evolved function of music: language

I propose that music originally evolved as a form of language.

It replaced, or evolved from, prior systems of communication that used relatively fixed and genetically determined sets of vocalisations to communicate fixed and genetically determined meanings.

Music evolved as a means to provide an open set of symbols, to which meanings could be culturally assigned, and it operated as follows:

  1. The rules of music defined a criterion of musicality, or musical "strength", such that a local optimum of musicality within the space of possible musical vocalisations constituted a melody.
  2. Because a melody was a local optimum for the criterion of musicality, each melody was attractive to listeners, easily learned and easily transmitted from one generation to the next.
  3. Each melody defined what was effectively a symbol, to which could be assigned a meaning. That is, in the music-based language, the unit of meaning was the melody.
  4. Music created a state of mind in the listener such that the listener was interested in paying attention to the melody, and very interested in determining what was the assigned meaning of that particular musical item.
  5. The rules of music determined an emotional quality associated with each melody, and this emotional quality amounted to a partial predetermination of the meaning of the music. Thus the culturally assigned meaning of a melody had to be consistent with the emotional quality of the melody, just like today we expect the meaning of the lyrics of a song to be consistent with the emotional quality of the melody.
  6. The music-based language did not contain any mechanism for the combination of simpler meanings into more complex meanings. (Or, if such a mechanism did exist, its operation was very limited, and this would correspond to the observed chorus/verse structure of modern popular songs.)

The replacement of music-based language by word-based language

The thing that replaced music, the "rocket engine" so-to-speak, was word-based language.

Word-based language was superior to music-based language for two main reasons:

  1. Words are much shorter and simpler than melodies
  2. Word-based languages allow the combination of smaller meanings into more complex meanings

Word-based language was so superior to music-based language, that once it got past a certain point of development, there was no reason for music-based language to continue existing, and music ceased to be a form of language.

However, music did not completely disappear. Or at least, it has not yet fully disappeared.

What actually happened is that some of the operational components of the music-based language system were disabled, or altered in a manner that music no longer operated as a system of communication.

In particular:

But, apart from this one change, most of the other components of music, as a biological system, continue to exist and operate:

  1. The rules of music define a criterion of musicality, or musical "strength", such that a local optimum of musicality within the space of possible musical vocalisations constitutes a melody.
  2. Because a melody is a local optimum for the criterion of musicality, each melody is attractive to listeners, easily learned and easily transmitted from one generation to the next.
  3. Music creates a state of mind in the listener such that the listener is interested in paying attention to the melody. The prior instinct to assign a meaning to the music itself has been replaced by a less specific instinct to consider the possibility that some things have meanings or significance that is not immediately obvious.
  4. The rules of music predetermine an emotional quality associated with each melody, and this emotional quality amounts to a partial predetermination of the meaning of the music.

Music strongly constrained the evolution of its replacement

Given the obvious superiority of word-based language over the music-based language, one might suppose it was a simple story of (1) good thing, (2) better thing, (3) good thing no longer needed. For example, to recap the hypothetical rocket-powered ostrich example: (1) wings, (2) rocket engine, (3) give up on the wings.

However, there are a number of relations and interactions between music and word-based language (as these things exist in the present day), which suggest, when considered together, that the prior existence of music-based language strongly constrained, and to some extent prevented, the evolution of any replacement system of language.

In particular:

Assuming that these constraints applied, we can determine the following likely history of how word-based language evolved, subject to the prior existence of music-based language:


The replacement of music-based language by word-based language was probably the single most important step in the evolution of "behavioural modernity" in the human species.

That is, humans became fully beviourally modern when and only when they started using word-based language in lieu of music-based language.

Given this hypothesis, we need to be able to explain why music has not yet completely disappeared.

There are two main reasons why a vestigial trait can still be existing:

  1. The primary cause of vestigiality happened only "recently", and the vestigial trait hasn't yet had time to fully unevolve.
  2. The vestigial trait has secondary functions, which reduce the selective pressue against the existence of the trait.

Exactly what counts as "recent" depends on many factors that are not necessarily straightforward to compute, such as how a given trait is represented genetically, and which genetic mutations could lead to the disappearance of that trait without causing undue disruption to other traits of the organism, and how frequently any of those genetic mutations is likely to occur.

Vyshedskiy: An alternative theory of the transition to modern recursive language

In Language evolution to revolution ... Andrey Vyshedskiy proposes a hypothesis that human language suddenly transitioned from non-recursive to recursive language about 70,000 years ago.

Vyshedskiy's hypothesis is rather similar to my own. The most important difference is that Vyshedskiy assumes that the transition was from a non-recursive word-based language to a recursive word-based language, whereas I assume that the non-recursive language was not word-based (becaused it was music-based). Also, in my hypothesis, the music-based language was less than non-recursive, indeed it was almost entirely non-compositional, ie there was no composition at all of smaller units of meaning into larger units of meaning. Vyshedskiy does seem to allow for some composition in the earlier non-recursive language, such as use of modifiers, and his proposal is that the critical step was the addition of recursive forms of composition, in particular those involving the use of prepositions.

In Vyshedskiy's scenario, the recursive language evolved from the prior non-recursive language. In my scenario, the recursive word-based language replaced the prior non-recursive music-based language.

Vyshedskiy has to explain what difficulties prevented the non-recursive language to easily evolve into the recursive language.

Whereas, I have to explain how the non-recursive language constrained and prevented the easy evolution of the recursive word-based language as a replacement.

However, the arguments that Vyshedskiy gives for the timeline of the transition are both convincing and plausible.

As a result, my current best estimate of when the transition occurred between music-based language and word-based language is ~70,000 years ago. (Some authors date the transition to full behavioural modernity happening slight later, eg perhaps ~50,000 years according to Richard Klein.)

The continued existence of a vestigal trait

The hypothesis of vestigiality solves the problem of explaining why music exists even though it doesn't seem to be for anything, and why, considering all the things that music does do, like expressing meaning or emotion, it seems to do them in a very vague and ineffectual manner.

Given an estimate for when the transition to vestigiality occurred, the problem we then have is to explain how and why music has not yet already disappeared entirely.

Is it a sufficient explanation that 70,000-50,000 years ago is very "recent", and we wouldn't expect anything to fully unevolve in that period of time?

One difficulty is that if we believe that something as complex as word-based language can appear in just a few thousand years, then surely it is possible for music-based language to fully disappear in a longer period of tens of thousands of years.

However, it is entirely possible that the potential for the development of word-based language existed all along, and that it only took one single mutation to "release" the hold that music had over language acquisition.

Given these uncertainties, I don't have any clear idea of how much I need to explain the not yet complete disappearance of music in the timescale proposed.

However, it is worth trying to understand all the factors that may be relevant to explain why music had not disappeared yet, or even to explain why some aspects of music might be expected to persist indefinitely.

Secondary Functions

The primary explanation for the slower-then-otherwise-expected demise of any vestigial trait is secondary functions.

In other words, the trait is no longer any use for the thing it used to do, but it has found other things to do, so we keep it around.

It might still be the case that these secondary functions don't really justify the full cost of maintaining even the vestigial remains of the trait, but they do somewhat reduce the selective pressure against the trait, enough to slow down its demise.

In the case of human music, there may be additional societal factors, where the secondary function of music for the individual relates to the fact that everyone else happens to like music, and it is a rather circular "function", because if music didn't already exist, there would be no reason for it to evolve to satisfy that function.

In the case of ostrich wings, I have already listed some secondary functions:

These functions come into the category of "if we have some no-longer-functional wings, then we might as well use them to do stuff, just because they happen to be attached to the body and they can be moved around".

If the vestigial wings did not already exist, and there was a requirement for any of these biological functions, a hypothetical non-winged ostrich would probably evolve some traits to provide those functions, but the result would not necessarily be anything like the vestigial wings that real ostriches actually have.

With music, there have been many attempts to identify the supposed functions of music.

All of these attempts suffer from being unconvincing. There may be some benefit to each of these proposed functions, but in each case the alleged benefit doesn't seem to be that great, and it's very unclear why all the specific complex features of music need to come into existence in order to provide those functions.

However, once we assume that music is a vestigial trait, then it is entirely plausible that some of these identified functions do provide some benefit, and the sum total of these benefits might be enough to somewhat slow down the "unevolution" of music, even if they do not (and will not) prevent its eventual disappearance.

Some of the proposed functions are societal (in the sense that they result form an interaction between the individual and society as a whole), and I will consider them separately, because these factors have their own special somewhat circular evolutionary logic.

Non-societal factors

Societal factors

Societal factors are factors where the individual has to like music, because everyone else in society likes music.

If you, the individual, don't like music, and everyone else does like music, then you will be less enthusiastic about joining activities and situations where music is involved, and you will end up being socially excluded.

Also, given the nature of music, and given that people like music, and that it has a certain effect on the mood, music will tend to get inserted into many social situations.

Some specific examples of situations where there's lots of music, and you might decided not to go if you didn't like music, are:

What Societal Factors can explain, and what they can't explain

Societal factors can explain why there is selective pressure on the individual to like music, in accordance with how everyone else likes music, and this can explain why music continues to exist (or at least, why it disappears more slowly).

But, societal factors cannot explain how music comes into existence in the first place.

For example, if music is in the early stages of its supposed evolution from nothing, and you are one of the 0.1% who likes music, and you only go to parties where there's music, then you will be going to parties where hardly anyone else turns up. In which case you will be excluding yourself from socializing with the other 99.9% of people who are only going to the non-musical parties. So in this case, the societal factors will be strongly selecting against the appreciation of music by the individual.

It's interesting to note that there is a similar societal argument in favour of drinking alcohol: