Another thought-provoking Quora question: Are there any hereditary units in music? The question details give some context:
In his blog post “The Music Genome Project is no such thing,” David Morrison makes an edifying distinction between a genotype and a phenotype. He also makes the bold statement “there are no hereditary units in music.” Is this true?
Morrison’s post is a valuable read, because it’s so precisely wrong as to be quite useful in clarifying your thinking.
Morrison discusses The Music Genome Project, Pandora’s database of musical metadata. They have a team of musicologists analyzing all of the tracks in their library according to 450 different musical characteristics. Morrison doesn’t like the name, because he sees it as misleading.
There seems to be a major misunderstanding here, since the mere idea of atomizing something does not make the atoms genes. After all, the idea behind the Project is simply one of taking music apart and evaluating it by its acoustic elements.
But Morrison understands biology better than he understands music, and his criticism is incorrect. Pandora’s 450 musical characteristics aren’t all “acoustic.” They include such abstractions as harmony, structure and style.
The first problem is that the study of musical attributes is clearly a study of phenotype not genotype, as Gasser alludes in the quote above — there are no hereditary units in music.
Wrong again. Music consists almost entirely of hereditary units. The “genes” of music are things like:
- Rhythmic figures
- Tuning systems
- Scales
- Chord progressions
- Melodic figures, “licks“
- Vocal and instrumental techniques
- Lyrical conventions — rhyme schemes, cliche phrases and subject matter
- Basic structural forms: the twelve-bar blues, the rondo, the canon
- Samples and breakbeats
If I sit down to write a Baroque chorale, it’s going to inherit 98% of its content from existing Baroque chorale. Otherwise it wouldn’t be recognizable as a Baroque chorale! Like all musical forms, it’s defined by an elaborate system of rules and constraints that effectively add up to a generative algorithm. These rules evolve over time in a manner remarkably similar to genomes in organisms.
A genome is generative: there is a mapping from a genome to an organism. There is no reverse mapping. In the case of music, there is a reverse mapping from a piece of music to these 400 odd features, but there’s no forward mapping … Knowledge of a phenotype is not constructive, because there are many ways of constructing that phenotype; a genotype is unique, and is thus constructive.
Once again, not true. If you know the rules for producing a Baroque chorale, it’s quite easy to generate new ones. I had to do exactly that in grad school.
You can even write computer programs to compose music in particular styles, and this music can be surprisingly good. David Cope has infamously written software that generates “new” Bach chorales that fool experts into thinking they’re genuine. Bebop heads, minimal techno, gangsta rap, Hindustani ragas — all of these musical forms are generated by rule sets passed from one musician to another in just about exactly the same way that genes get passed from one generation of organism to the next. Mutations to the hereditary rule sets occur when someone learns the rules incorrectly, or willfully violates them.
I believe that we will eventually come to understand the way that ideas are physically encoded in the brain. Furthermore, once we do, I expect that we’ll be able to sequence memes the way we currently sequence genes. We’ll be able to map the “memome” of a song and construct rigorous “phylomemetic” trees of music. We can already do that with hip-hop samples, as the diagrams throughout this post demonstrate.
Mapping the human memome won’t happen anytime soon, but I expect the first halting steps within my lifetime.