I was looking at pictures of Frank Sinatra’s home in Palm Springs, and I noticed that the kitchen looked very similar to a kitchen that I described in a book I wrote, VeGa. You can just take my word for it since it’s a long book, but the similarities are a bit tough to accept as mere coincidence. Now, it’s entirely possible I previously saw pictures of Sinatra’s home, since I’m a huge fan and have been for my entire adult life, but it doesn’t matter to the conclusion that followed, which is that art in particular and aesthetics generally, create an opportunity for information about a person’s genome to be expressed, and therefore, facilitate mating decisions.

I’ve seen some evidence that preferences are genetic, and this should surprise no one, since people clearly take aesthetic preferences seriously, often motivating people to work harder, just so they can live in a particular home, and wear particular clothes. The more formal conclusion I reached this morning is not on the empirical side, it is instead a theoretical observation, that the expression of preferences creates a language in which information about genomes can be conveyed. There is still of course the empirical question of whether or not this is actually happening, but as we’ll see, the opportunity to convey an enormous amount of information using preferences certainly exists.

Let’s begin by formalizing and limiting preferences as expressed with respect to a particular artifact (e.g., a piece of music, a painting, a pair of pants, etc.), as follows: I hate it, I don’t like it, it’s OK, I like it, and I love it. This creates an ordinal scale of 5 possibilities, ranging from hate to indifference to love, that we’ll use to express preferences. In reality, people have written entire books about single paintings, so the real world range of expression, is much wider, but as we’ll see, it doesn’t matter, you already have an incredibly expressive language using just these five possibilities.

Specifically, for any artifacts, there will be possible rankings. As a consequence, two people discussing, e.g., just 10 works of art, creates 9,765,625 possible rankings, and again, all they have to say is, I hate it, I don’t like it, it’s OK, I like it, or I love it, 10 times, and they’ll convey about 23 binary bits of information, which is genetic bases worth of information. As you can see, increasing the number of possible rankings simply increases the base of the exponent, whereas the exponent dominates the counting function.

Common sense says that people will primarily respond to appearance, smell, taste, and other visceral information when selecting mates, but there are still at least two potentially important roles that art and aesthetics could facilitate in mating decisions, and thereby explain its otherwise anomalous importance to humanity: (1) filtering large crowds of individuals on the basis of shared preferences and (2) marginal distinctions within homogenous populations.

Understanding point (1) is straightforward, if you e.g., put together an event that centers around aesthetic artifacts (e.g., a concert, a gallery exhibition, etc.), then generally speaking, people will attend that event only if they’re sufficiently interested in the full set of artifacts, and the particular combination of artifacts in question. That is, even if you like both Keith Haring and Caravaggio, it’s a bit weird to combine both in a single exhibit.

Therefore, we can express the same ordinal rankings over combinations of artifacts, and that drastically increases the amount of information conveyed. Specifically, given artifacts, there are possible combinations of artifacts (i.e., the cardinality of the set of all subsets of artifacts), each of which is capable of an ordinal ranking, and therefore, we have, for any artifacts, possible rankings over the set of all subsets of those artifacts. If we set as we did above, i.e., considering just 10 artifacts, we find that , which is approximately 1187 genetic bases worth of information, or approximately 7.0% of an entire human mtDNA genome. Now we’re talking about a significant amount of information, that can realistically be conveyed, since , and most people should be able to meaningfully consider around 1000 collections of artifacts, at least over a significant period of time. Just imagine selecting an outfit to wear to a particular gallery exhibition, this is exactly the kind of combination of preferences that we’re considering, which quickly start to cover large numbers of possible combinations. In fact, even having strong preferences of this sort could demonstrate genetic fitness, in addition to conveying genetic information. The net point being, because it’s possible, and it’s obviously something many people take very seriously, it should have some biological function, and I think it’s the obvious: you’re conveying information about your genome through your preferences.

Point (2) is not hard to understand given the above, which is that in a genetically and morphologically homogenous society (i.e., everyone is genetically and physically very similar), marginal differences in genomes could be important, in particular because you have a high risk of incest. It’s not clear to me that there’s any real evidence of heightened selection for aesthetics in homogenous societies, but the discussions above suggest it should happen, because you can convey marginal differences in genetics through your preferences, allowing people to maximize genetic diversity in an otherwise homogenous population.