Music, Language And The Anthropic Principle

8 April, 2023
We can only observe the existence or non-existence of language if language exists, because our ability to comprehend the very concept of "language" is dependent on acquiring that concept from other people, communicated to us by language. Which means that we can't observe the non-existence of language.
But what about music?

The Anthropic Principle

The Anthropic Principle is a principle of observation and necessity.

We observe that something exists, and we ask ourselves "Why does it exist?"

But then we realise that this thing necessarily exists, because if it didn't exist, we would not be there to observe its non-existence.

For example, we observe that life exists, and we could ask "Why does life exist on our planet?".

And the Anthropic Principle reminds us that we are living beings, so if life didn't exist, we wouldn't be here to observe that it didn't exist.

An Anthropic Explanation of Language

We can apply the Anthropic Principle to language.

We can ask: "Why does language exist?"

And we can realise that "language" is a very abstract concept, and our ability to understand that abstract concept is very dependent on the use of language, because, in practice, we learn about such concepts from other people, which depends on those people using language to communicate those concepts to us.

In other words, we can only think about what "language" is because language itself is a thing that exists.

It follows that we cannot observe the non-existence of language – either language exists, or, if language doesn't exist, then it is not even possible to have a concept of "language", and we cannot observe the non-existence of something without having some concept of what it is that we are trying to observe.

What about Music?

It is not at all obvious that the Anthropic Principle can explain the existence of music.

Music does not seem to be necessary for anything.

It's this weird thing that just happens to exist, for some reason, but we don't know why.

And it doesn't seem to serve any purpose. It's just something that exists and makes us feel certain ways if we choose to listen to it, and typically we enjoy listening to it.

It is true that music is a very subjective concept, ie we only really know what music is because we hear it and we know what it feels like to hear it.

We might argue that the concept of "music" only exists in our minds because in fact music does exist.

This should be a dubious argument, because it would imply that the Anthropic Principle applies to anything which is described by a word.

However, there are words that describe things that do not actually exist.

For example, "unicorn".

The word "unicorn" has a definition, ie "a horse with a long pointy horn at the top of its head".

The word exists, even though we are fairly sure that the thing described by the word does not exist.

Similarly, we could imagine a world with people in it where music didn't exist, and someone in that world invented a concept of "music" as a hypothetical thing that might exist, defining it as a phenomenon where certain sounds are organized according to certain rules and those sounds cause emotional feelings which listeners enjoy listening to. It would be quite easy for the people in that world to observe that this "music" thing, as described, did not exist.

And it should be noted that the Anthropic argument for "language" does not depend on the word "language" existing because language exists – rather it depends on the fact that people are only able to imagine concepts like "language" because of their ability to communicate those concepts among each other, and that ability depends on the existence of language as a means of communication.

Music, as a Prerequisite for Language

But, what if the existence of language actually depended on the existence of music?

Human language had to evolve somehow.

We don't know any of the details of how it evolved.

What if the evolution of music was something that had to happen first?

What if the only way that language could evolve was to evolve in a world where music already existed?

Or more precisely, what if the only way that language could evolve was to evolve in a world where the ancestor of music already existed?

Hypothesis: Proto-Music was an Original Form of Language

"Proto-Music" is a term commonly used to describe a hypothetical ancestral form of music.

Normally when we talk about the origins of "language" and "music", the word "language" is referring specifically to human word-based language which consists of sequences of spoken sentences where sentences are constructed from sequences of words according to various syntactical rules.

There is also a more general concept of "language" which can refer to any system of communication where there are symbols that have meanings.

This more general concept of "language" can encompass a wide range of phenomena that include many different systems of animal communication, some of which involve the use of sounds that have specific meanings when animals of a given species use those sounds to communicate with each other.

Most or all of those systems are much simpler than human word-based language, but they can be considered to be "languages" in this more general sense.

It is entirely possible that modern human word-based language was preceded by other forms of language that may have been quite different from anything that we know about in the modern world.

And one of those earlier forms of language may have been proto-music, ie the ancestor of modern human music as we know it in the world today.

(Side note: By "modern music", I mean music as a form of mood alteration and/or entertainment. We don't know how long music has existed in this form, let alone how old any earlier form of "proto-music" might be that had other functions. One might consider the existence of musical instruments to be evidence for the existence of such modern music, and there are fossil flutes over 40,000 years old, which would indicate that "modern" goes back at least 40,000 years.)

Proto-Music as Prerequisite for the Evolution of Word-Based Language

Why would the existence of a proto-musical language be essential to the evolution of modern word-based language?

One could develop a number of specific hypotheses about the nature of a proto-musical language (and indeed I have – see other articles in this blog).

But for the purpose of explaining why the existence of a prior language might have been essential to the evolution of word-based language, we can consider the issue of the gradual evolution of word-based language.

Modern word-based language includes vocabulary made up of individual words, and syntactical rules for assembling words into meaningful sentences. It requires all people to know how to learn to listen to spoken language and understand it, and it also requires all people to know how to learn to speak that same language.

How could all of these inter-related aspects of word-based language have evolved gradually, step-by-step? What were the first words? Were the first words spoken individually, without any syntax? Why did people say those first words? How did other people learn to attach meaning to those spoken words?

The possible prior existence of an alternative form of language could provide part of the answer to these questions.

In particular, words could have evolved initially as an enhancement to the prior form of language, where, for example, even single isolated words could provide additional information to the listeners of that language.

In as much as music communicates anything, it communicates emotions, which are fairly abstract meanings.

So it's plausible that words evolved as a means of providing additional specific details explaining the reason for the emotions being communicated by the proto-musical language.

In this scenario, the prior form of language, not yet containing any words, allowed the speakers to express useful meanings. Then the addition of individual isolated words would have been immediately useful, because those words added to the existing meaning of the utterances of the earlier language form. (The initial "words" could have had minimal complexity, for example they might have consisted entirely of single vowel sounds, and there might have only been, for instance, three distinct vowel sounds, giving an initial vocabulary of just three words.)

Also, of course, we know that, in the modern world that we live in, many forms of music include words, ie songs.

The modern existence of songs, as a thing, is consistent with a four-step theory of evolution of language as follows:

  1. Evolution of proto-musical language to express emotions.
  2. Evolution of words as an enhancement to the proto-musical language.
  3. Further evolution of word-based language into more complex forms, including syntax.
  4. Disconnection of the more evolved complex form of word-based language from its original proto-musical progenitor progenitor "parent".

In particular, songs exist now, because they are a remnant of a earlier stage in human language evolution where human "language" actually consisted of musical phrases with words embedded in them.

One technical reason that the proto-musical language could easily evolve to include words is that most of the information in words is represented by consonant and vowel distinctions (at least this is the case for non-tonal languages), and at the same time, consonant and vowel distinctions do not play any major role in defining the musical quality of music (ie you can sing "doo doo doo" or "la la la", and it's still the same tune).

So from an information-theoretic point of view, we can understand that words evolved to use consonant and vowel distinctions to represent information, precisely because the proto-musical language wasn't making use of those distinctions, so they were there for the taking.

There may be other aspects of proto-musical language which enabled it to foster the evolution of word-based language, and at the same time those aspects may have constrained the proto-musical language in ways that could not be mitigated, so eventually the word-based language had to disconnect itself from its proto-musical "parent".

For example, music appears to have some kind of tree-based syntax, and this may have scaffolded the development of syntax in word-based language. We can observe that musical "syntax" differs from the syntax of word-based language in various ways – musical syntax is contained within melodies, and melodies are the approximate "units of meaning" in music, whereas in word-based language the units of meaning are (approximately) the individual words, and the syntax builds those individual units of meaning into the full meanings of sentences. Also, musical syntax trees are usually fairly symmetrical, whereas syntax trees for sentences can be quite asymmetrical.

Another feature of music is that it is very repetitive, where repetition includes both exact repetitions of extended musical phrases, and also partial repetitions, for example similar but different musical phrases in a melodic progression. The amount of repetition in word-based language is quite limited, and almost never involves exact repetition of extended phrases. Repetition is instrinsically wasteful, and if the proto-musical language was constrained to have repetition in it, then, in the long run, the only way to achieve greater efficiency would be to drop the musical component of language altogether.

The Principle of Least Miracle

Even if we can develop a plausible hypothesis about how word-based language initially developed as an enhancement of a pre-existing music-based language, we might still wonder if word-based language could have evolved some other way.

In general, the Anthropic Principle provides us with a way to explain the observation of an event E that should have a very low probability PE. The explanation is that the event itself is necessary, in order to attempt the observation in the first place.

A more complicated scenario is where we might have two events E1 and E2, where it's necessary that at least one of those events occurs, but it is sufficient for only one of them to occur, and where both possible events have very low probabilities PE1 and PE2.

If PE1 and PE2 are of similar magnitude, then we can't say anything much about whether we expect to observe E1 or E2.

But if, for example, PE2 is much much smaller than PE1, then the Principle of Least Miracle says that we must expect to observe E1, and that if we did observe E2, then the Anthropic Principle is not "explaining" how it is that we observed E2. In effect, the probability of observing E2 given that we had to observe at least one of E1 or E2 is PE2/(PE1 + PE2) ≈ PE2/PE1, which is, by our assumption, a very small number.

In the case of the evolution of word-based language, I am proposing that E1 is the evolution of word-based language via a prior proto-musical form of language, and E2 is any other scenario for the evolution or word-based language (or more precisely, E2 is the event consisting of the occurrence of any one of all the other possible scenarios).

Using the Anthropic Principle, combined with the Principle of Least Miracle, we can explain the existence of music as the descendant of an earlier music-based language, as follows: