Header image

john hawks weblog

paleoanthropology, genetics and evolution

Photo Credit: Photo by rawpixel.com on Unsplash

Is it meaningful if geneticists find that marriage can be influenced by genes?

The power of gene-trait association studies has increased markedly over the last few years. Samples of hundreds of thousands of individuals with genotype and phenotype data are coming online.

This statistical power has opened the door to discovering genes that have very small but significant effects on traits. In the process, a good fraction of what was once called “missing heritability” has actually turned up. Rare genes of medium effect, and common genes of incredibly tiny effect both matter to some extent.

Geneticists have started to ask a simple question: If most genes that affect a trait have incredibly tiny effects, are we discovering anything biologically interesting?

For many years, human geneticists have argued for larger and larger studies of gene-trait associations based upon the idea that genes will highlight biological networks. Any one gene might have a small effect, but finding many genes will allow geneticists to understand the inner workings of biology.

For some traits, that has proven not to be the case. Hundreds of genes, each with very small effects, don’t fit into any coherent biological network. In some sense, these genetic associations reflect all the different aspects of physiology that can influence human growth, metabolism, and behavior.

Behavioral traits seem to fit into this model, which has been dubbed the omnigenic model, in a recent paper Evan Boyle, Yang Li, and Jonathan Pritchard. Today it’s commonplace to see studies finding genetic associations for traits that are most salient in social psychology, identifying a few genes of fairly small effect for traits like “educational attainment”, or “age at birth of first offspring”.

Eric Turkheimer writes in the Genetics and Human Agency Project blog: “Why marital status is heritable.

It is only in the weakest (softest!) possible sense that you can refer to these genes as “genes for” the phenotype, or as “causal variants” for a phenotype, because they don’t actually have anything to do with the phenotype, with divorce, per se. The are cellular level body size effects, or appearance effects, or impulsivity, or alcohol or ginger hair effects, and even listing them like that probably makes their effects much more discrete and concrete than they actually are. If I had to guess most of the genetic action is at the cellular level, so causally distant from how you feel after a fight with your spouse that there is no point in even talking about it.
Heritability in and of itself has no particular relevance to questions about whether things like intelligence or divorce are biological as opposed to social constructions. Social constructions are heritable.

The main idea of Turkheimer’s post is that heritability by itself does not tell us that biology (as opposed to social phenomena) is relevant to explaining a trait. That is, correlation is not causation, and a genetic explanation for a trait that involves tiny influences of hundreds of genes is not a meaningful explanation.

And yet, I’m not very satisfied by this example.

Turkheimer has chosen the example of marriage status precisely because we are meant to assume that any important causal explanations must be totally psychological and social. Surely it is absurd to suggest that genes have any causally important role in such a social institution.

In some sense Turkheimer appears obviously correct. “Marriage status” in the United States or Europe in the early 21st century is an absurdly specific phenotype. It didn’t exist in our evolutionary past, and our genes could not possibly have adapted to it. Right?

But widen the frame a bit. Long-term pair bonding is exactly the kind of thing that mattered to natural selection during human evolution.

Look at another trait that is a major component of fitness: “age at birth of first offspring”. Some individuals have children very early in their adult lives, and others only have children much later. The proximate causes of these behavioral outcomes are unquestionably social and psychological. Yet age at first birth is heritable in populations today, and a number of genes found to explain small fractions of the additive component of variation in large samples of people. The trait is also quite strongly selected within the human populations where it has been examined. On average, women who have given birth to their first child earlier in their lives also have more children over their life spans.

How can this be? It would seem like trait that is so strongly correlated with lifetime reproductive success should have very little additive genetic variance in a population that has been at equilibrium with its environment for a long time.

Humans, of course, have not been anywhere near equilibrium recently. Our environment, both cultural and physical, has changed radically during the last 50,000 years, and continues to change rapidly even now. “Marriage” is a social institution that means something different today than it did 50 years ago. Marriage has different participation rates by age in today’s United States than in many other countries, and different rates in comparison to the recent past. Sure, it is reputed as one of the most stable institutions in human societies, but its psychological and social dimensions nonetheless can change rapidly.

In that sense, heritability of marriage status as a trait does have at least one salient biological explanation. Some genetic factors might correlate with marriage status in today’s populations precisely because the genetic differences that exist today are those that persisted in past social and physical environments, which were in some ways different from today’s environment.

There’s nothing trivial about noting that today’s social environments are markedly different from the past, and that by itself is a biological fact that influences gene-trait associations.