People often talk about music as if it were language. Leonard Bernstein even recorded a series of lectures applying Chomsky’s theory of generative grammar to music. Chomsky himself answered a question on this in a not very satisfying manner. Some people can get very exercised over this.
But it seems to me that a playing around with strengths and weakness of the music = language metaphor can help us come to grips with the question a bit better. We can find a number of mappings between music and language but an equal number of mis-mappings. We do not need to go very far into it to see where they are. That does not mean that a deeper investigation into musical properties of language and linguistic properties of music cannot be fruitful. And obviously they are both universal human faculties — but looking for a musical essence in language or a linguistic essence in music is what a metaphor-aware approach to this question is hoping to warn against.
Here are some of the obvious similarities and dissimilarities. (Note: After, I finished my list, I came across a Chomskean comparison by Jackendoff which has a slightly different focus but comes to the same general conclusion.)
Music is like (a) language in that:
- It can be described through a system of rules that operate on a limited vocabulary. There are 12 notes (on the Western chromatic scale) that can produce an infinite variety of melodies just purely in their combination further enhanced by their combination with rhythms, tempos and harmonies. (Although I have argued elsewhere that language is not actually much like this, at all.)
- It combines small building blocks into larger components that are like words, phrases, sentences and text. In fact, we talk about phrasing in music. But we have things like bars, stanzas, movements, etc.
- It is recursively expressive. I can embed little segments of music in others indefinitely. Bach’s variations are an example of this as is jazz improvisation.
- It has dual articulation in that smaller segments like scales are organized independently of large segments. This is well-known about language (in certain circles). We articulate sounds into words and words into statements at the same time but also seemingly independently of each other — we know this because we can be good at one and bad at the other (thus dual articulation). In music, producing individual notes (e.g. fingering on piano or guitar, or breathing and embouchure on a trumpet or trombone) is a skill independent of expressing the musical ideas contained in the notes.
- It has phraseology and idioms: We speak of musical phrasing but that is more a question of production. But music also has set ways of expressing certain things. There are things like chord progressions or minor or major modes that combine together to express musical meaning. They are more than the sum of their parts and form their own building blocks.
- It can cross-reference between compositions (texts): We can hear echoes of folk songs in classical music or we see direct quotations of melodies in jazz. (But it should be noted that this is much less pervasive than in language where co-reference is one of the core components of language.)
- It can communicate emotion both segmentally (sequences of notes) and suprasegmentally (expression, emphasis, etc.) In the same way, any phrase (such as Please, sit down.) can be pronounced in many different ways (e.g. in a welcoming, quizzical or threatening manner; deliberately, offhand, formally or casually, etc.).
- It has styles, genres and dialects or even accents. We can instantly recognise music recorded in different time periods or in different styles. Even individual artists have particular ways of expressing themselves musically that can be imitated or parodied. YouTube is full of videos of people playing X in the style of Y.
- It can be acquired and learned. To reach a ‘fluency’ in music requires an effort that is not dissimilar to the acquisition or learning of language. Part of it happens naturally, simply through exposure to music and part of it is formal — such as learning words to name parts of music, such as notes, chords, harmony, etc. or learning to ‘read sheet music’.
- It is culturally conditioned. Different cultures have developed very different takes on what music sounds like. Chinese music does not sound very musical to Western ears due to the very different approach to tonality.
This list makes it seem like music and language are very similar. But the next list of dissimilarities shows that they are also different in fundamental ways.
Music is NOT like (a) language in that:
- It cannot be used to directly communicate propositional meaning. I can say, ‘my house is right at the end of the street’ or ‘that will be 50 cents’. But there’s no way to express this kind of content in music. Sometimes music tries to imitate language (Janáček is often cited doing that) but without the words, nobody would know what an opera is about.
- It has a radically smaller set of building blocks and rules for their combination than language. There are only 12 intervals/tones in Western Music. But this in itself would not be a problem. There are languages with similarly low numbers of phonemes (distinct sounds). However, there is no intermediary unit of expression equivalent to the word. The rules of melodic composition operate directly on these 12 (or more if we include microtones) tones.
- It does not have internal instruments of disambiguation. Being able to repair a conversation that is broken with phrases like ‘what did you say’ or ‘can you say this again’ is a fundamental part of the communicative process. Without them, language would not be nearly as useful. There’s nothing like that to be found within the ‘communicative’ inventory of music that does not rely on verbal or written language in one way or another.
- It can only be universally acquired in the most rudimentary sense (i.e. everybody can hum a tune but very few people can play an instrument). Everybody can and does learn their first language. And everybody acquires some musicality as part of their socialisation into their culture. But most of what we would consider musical fluency is learned through some means of instruction. Music (in this strict sense) is more like written than spoken language.
- There is a much greater difference between receptive and productive competence. Everybody knows more of their language receptively (passively) than productively. Depending on context, this gap is very small or very large. A person speaking a language of a small community without a lot of specialisation will have a smaller gap between what they can say and what they can understand (although there are many specialised languages even in these contexts). But this difference will be greater when it comes to technical language between a first-year university student and their professor. But in music, everybody can listen passively and receive most of the intended effect while their ability to produce the same music will be severely limited. Many people can hum back a simple tune but they cannot reproduce a full musical performance. Even professional musicians will vary in their ability here.
- There is much greater variability in individuals’ ability to produce music beyond the most trivial. It requires effort and study to produce music in a way that we think of as music. We can say that everyone is musical to some degree but most people cannot actually produce anything beyond the simplest of tunes. However, in language, everybody can communicate (even people with impairments) to a significant degree. The differences in competence only appear at the higher levels.
- Much more of the production process requires cooperation among individuals. While not a requirement, most music we consume is produced by groups of people. Or if not another person, it requires an instrument made by another person. Language production (at least in spoken form) is primarily by individuals. (Although, there are instances of group production — theatre, speeches, etc.)
- It is much more limited in its dialogic potential (i.e. it is most often used for one-way communication between few producers and more recipients or joint co-production of producer/recipients). Language is fundamentally dialogic. Anything that is said can be responded to. Questions can be answered, propositions can be countered or elaborated. Music, on the other hand, is primarily declarative. Of course, metaphorically, we can talk about dialogic elements in music. In jazz or blues, we have call and response, in classical music we have things like counterpoint. We can also talk about members of an orchestra communicating and responding to each others’ musical ideas. But there is no such thing in music as saying: ‘No, I disagree with you about X, instead, I believe Y and this is why.’
Ultimately, this is not all that important. We know what language is and we know what music is. Saying one is or is not like the other in some way won’t change any of that. However, it can help us think more clearly about them and avoid ignoring important aspects and unique properties of both.
Note: What is music
You often hear popularising musicians saying things like ‘Everybody is musical’ or ‘everything is music’. But that is not what most people’s intuitions about music tell them. For most people, music is the sort of thing they hear on the radio. It is the result of composition and production. There are instruments involved and skills and ability to play those instruments. That is the sense in which I’m comparing language to music. When I sing in the shower, I may be engaging in a musical activity but it is not the prototypical meaning of the word. And if we try to build our case for music around that, we’d be leaving other important aspects out.
Language of X
Lists similar to this one could be constructed for other cases where people talk about the ‘language of X’. Programming, gestures, art, or architecture. These would probably end up with lists that overlap with this one in many ways.
There are different cliches in use about many of these domains. So people tend to overestimate the extent to which facial expressions or architecture or art are like language but underestimate the degree to which programming languages are like natural language.
So in some contexts, I’d want to stress the dissimilarities. When people say things like ‘He can express all he needs through dance.’ or ‘90% of all language is nonverbal’, we would want to point to the propositional and dialogic aspects of language that are lacking in these domains.
But in other contexts, we’d want to point to the parallels. For instance, we may want to remind ourselves that programming languages are more like natural language in many (but not all) of the ways, music is like language. They have dialects, phrases and idioms, multiple levels of articulation, recursiveness (duh!), etc.
This post started life as an answer on Stack Exchange which I then cleaned up for Tumblr. This is an expanded version. It is also posted on Medium. This version has been slightly amended for spelling and punctuation.