A week ago I was at a workshop Rules as Genotypes in Cultural Evolution (check out the Focus Article by Elinor Ostrom that set the stage for the meeting). One major topic of discussion was what might be the cultural analog of genotype.

In biology phenotype is the observable traits and characteristics of an organism, including morphology, coloration, behavior, etc. Phenotypic traits are determined jointly by the organism’s environment and its genotype, or genetically encoded information. Multicellular organisms like us store genetic information in the DNA (although things are somewhat complicated by the possibility of epigenetic transmission of acquired traits).

The distinction between the phenotype and the genotype has been enormously productive in evolutionary biology, so folks studying human cultural evolution have proposed that we need to find cultural analogs of the genotype and phenotype. One such scheme that I find fairly coherent (I actually teach it in my class on cultural evolution) is the one formulated by Richerson and Boyd (see their Not By Genes Alone; Rob Boyd participated in the workshop and argued in favor of this view). Richerson and Boyd define culture very broadly, as socially transmitted information. The cultural phenotype is pretty clear – it is the behavioral traits of humans, understood broadly (includes collective behaviors such as dance and rituals; knowledge, philosophy, and science; tools, books, clothing, tattoos, domesticated animals, technology, etc). Unlike biological traits in most organisms, human behaviors are affected not only by genes and the environment, but also by culture. Because both genetic and cultural information is transmitted across generations, this theory is also known as the ‘dual inheritance theory.’

So what’s cultural genotype? Boyd and Richerson argue that humans had culture before there were any technological means, such as memory chips of computers or written books, to store cultural information. The only place where cultural information could be stored in prehistoric times was people’s brains. So cultural genotype is the information stored in human brains.

Fine so far, but other participants in the workshop had different views. Some objected to the idea that any information is ‘stored’ in the brain (I never figured out why, though). Others, like David Sloan Wilson, proposed very different views of cultural genotypes. Wilson, together with his graduate student Yasha Hartberg, argued that a sacred text can be thought of as a cultural genotype, because it “consists of many ‘genes’ in the form of stories, commandments, and other texts. A sacred text such as the Christian Bible is replicated with high fidelity and has a potent effect on behavior, which are two requirements of a cultural genotype.”

This view also sounds reasonable, but can cultural ‘genes’ be both neural circuits in the brain and words inked on a parchment? After all, biological genes come in only one variety, the DNA (let’s ignore viruses and prions for simplicity). This leads me to the question whether the whole idea of ‘cultural genotype’ is a useful concept.

After all, what gets transmitted is not the ‘cultural genotype,’ whatever that is, but the cultural phenotype. Dawkins’ phrase of memes jumping from brain to brain is a striking metaphor. On further thought, however, I think it is a silly, and certainly not a useful idea. We are not telepathic! The way cultural information is transmitted is by people observing the behaviors of others and then attempting to imitate them, with greater or lesser degree of success. We actually don’t even know whether the observer/learner encodes the cultural information with precisely the same configuration of neural circuits (if that’s how we store information in our brains) as the one in the brain of the person being imitated. (I believe that Richerson and Boyd made this point before me.) In fact, most likely the same behavior can be encoded by a multitude of very different circuitry configurations. Cultural evolution is Lamarckian, but the distinction between the genotype and phenotype is really useful only in a Mendelian framework.

So what really matters is the actual observed behaviors, not how they are encoded in brains. That’s a relief, because we really don’t know how information is stored in the human brain. As Rob Boyd stressed during the workshop, cultural evolution is currently in its pre-Mendelian phase. But I would argue that while it would certainly be interesting to know how brains work, this knowledge is rather academic for the scientific study of cultural evolution. Yes, we need to know about various biases affecting learning and transmission of cultural information, but psychologists are doing a pretty good job investigating such mechanisms experimentally. I am not against brain research, I am just saying that we don’t need to wait for new great insights from neuroscience to study cultural evolution productively.

In any case, in this day and age we have an alternative cultural genotype, whose physical characteristics are completely understood – digital information: books, technical manuals, audiotapes, videos, etc. Any human behaviors can be recorded and transmitted to others. You can now learn how to fix a leaky faucet or study an esoteric martial art on the Youtube.

The genotype/phenotype distinction is not a useful way to think about cultural evolution because cultural evolution is too different from genetic evolution. Cultural evolution is Lamarckian, while genetic evolution is Mendelian (but both are Darwinian). Cultural traits can be both discrete and continuous, while genetic traits are discrete. Cultural information is transmitted ‘asexually.’ Finally, in cultural evolution what ultimately matters is not what an individual person does, but what groups of people do.