Some Comments on "The Three Laws of Behavior Genetics" (and the two other laws)

Twenty years ago, Eric Turkheimer wrote an often cited paper titled:

Three Laws of Behavior Genetics and What They Mean

This paper is still often cited today and perhaps has taken on a life of its own, with a more deterministic interpretation than Turkheimer apparently intended and for which he recently clarified in a blog post his original intent. Nonetheless, much of the criticism of his paper comes from those in the genetic determinism camp, with the extremes being the "race scientist" crowd. So, since I sit on the other end of the see saw from the genetic determinists with Turkheimer poised somewhere in the middle, I will weigh in with my own thoughts about his three laws, as well as the two additional non-Turkheimer laws added into the soup. In the process of this, I will posit my own Four Laws of the Behavioral Genetics Fallacy. First, let's lay out the three laws that Turkheimer posits:

First Law. All human behavioral traits are heritable.

Second Law. The effect of being raised in the same family is smaller than the effect of genes.

Third Law. A substantial portion of the variation in complex human behavioral traits is not accounted for by the effects of genes or families.

I agree with perhaps one and a half of these laws. I’ll start with my half agreement with the First Law. For starters, I take issue with the use of the term “heritable,” because the term predates genetics and has had many different meanings and interpretations over the years, as this article points out:

The term ‘heritability,’ as it is used today in human behavioral genetics, is one of the most misleading in the history of science. Contrary to popular belief, the measurable heritability of a trait does not tell us how ‘genetically inheritable’ that trait is. Further, it does not inform us about what causes a trait, the relative influence of genes in the development of a trait, or the relative influence of the environment in the development of a trait.

 

While Turkheimer is cautious in his meaning, most behavioral geneticists would take his First Law to mean that all human behavioral traits have a genetic component and, as his quote of Gilbert Gottlieb included in his paper suggests, and for which I assume he is in agreement:

“The present . . . viewpoint holds that genes are an inextricable component of any developmental system, and thus genes are involved in all traits” .

At face value, how could one argue with that? Certainly a bumble bee is genetically distinct from a human being and the genetics of each will play a role in their development and differing behaviors. Their differing genetics and chromosomal structure have contributed to the creation of two very different species, with vastly different neurobiology, physical characteristics and abilities. It would be ludicrous to suggest on that level that their behavioral differences don’t have a strong genetic component.

Even with such obvious genetic differences, though, it would be difficult to determine which specific genes make us behaviorally different and we would have difficulty quantifying such differences in a meaningful way. Can we say with confidence, for example, whether a bumble bee, a parrot or a dog is more intelligent, despite being very genetically distinct, with entirely different genetic, neurobiological and physical constitutions? Could we find a measure to adequately quantify their intelligence for comparison purposes? Moreover, could we then identify which genes among these species would account for these differences? Clearly, we could not do so in any real way, and the development of such an assessment would be little more than a Rohrshach test of those developing the assessment tool.

Thus, in this context, when we say that genes are involved in all behavioral traits, that is true in a broad sense, even though in the example given, we are not really able to identify causal genes to explain those differences and perhaps have little ability to even define those differences.

However, the topic here is human behavioral genetics, not the behaviors that make us human in comparison to other species, but rather the differences in behaviors among members of our species, individually or as groups. The actual genetic differences between individual humans is obviously quite minuscule by comparison. Rather than an entirely different genetic makeup and chromosomal structure, we are looking at single base pair variations of a functional gene or some small area of difference somewhere in a relatively uniform genome. Moreover, such small variations can be found in different ethnic or geographic groups with no relation to any behavioral mechanism and, as has been seen, it is rather difficult to discern through genetic studies, whether these variations are simply markers for population stratification or something actually related to the behavior in question. We are, arguably, after decades of this research, almost completely unable to identify any genetic variations in human beings that we can definitively link to a behavior, intelligence or a personality trait. In short, the entire field of behavioral genetics has failed to achieve any real advancement in definitively identifying genetic mechanisms for any trait.

Proponents of Behavioral Genetics and genetic studies will argue that we just need larger sample sizes or improved methods and technology to get meaningful results. Or they will say, “Look at this new study that came out,” though generally such studies never replicate and fade into obscurity, only to be replaced by another “new study that came out.” I refer to this asymptotic optimism as “The Shell Game.” At a certain point, and we have arguably reached that point, we need to recognize that we are not going to win this shell game and need to move onto other games. We have hunted for the genes and didn’t find them. This, at the very least, suggests the possibility that they don’t exist.

That said, I would point out a few exceptions to the above. The first would be chromosomal abnormalities, such as are found in Down Syndrome. Such an abnormality affects brain development and a host of other physical issues that I make no argument against. The second would be single gene mutations that have a deleterious effect (Huntington’s Chorea, Fragile X Syndrome, Tay Sachs, etc.). Again, these are genetic abnormalities that cause specific diseases or disorders that can also have an effect on behavior or normal intelligence. A more subtle possibility are minor physical differences related to a variant. An example commonly cited are variants of the alleles coding for the aldehyde dehydrogenase enzyme (seen more commonly in Asian populations) delaying the breakdown of that product of alcohol metabolism and causing an unpleasant enough reaction when imbibing alcohol that it might be preventative for alcohol abuse. This is a plausible mechanism. Lastly, would be genetic variations that can be traced to obvious physical characteristics, such as height, skin color, obesity or other physical traits that might alter someone’s station in society.

These are the exceptions, however. Implied in behavioral genetics, is the idea that a genetic variant or some combination of genetic variants (or other variations in the genome) somehow cause you to be more intelligent, become depressed, be prone to criminal activity, go farther in school, etc. This would presumably be due to neurological development related to particular genes or changes in neuroreceptors or neuropeptides or hormones that would cause a person to present with a trait or over or under exhibit said trait from what would be expected by the “standard” variant. Turkheimer was prescient in predicting that such mechanistic thinking would not be likely to meet with any real success. In short, other than the above noted exceptions, we have not established that genetic mechanisms exist for any behavioral trait, human intelligence, or human personality.

To do so, we would first have to disentangle the population stratification issues that confound these genetic studies. As anyone has seen results from commercial genetic databases, such as Ancestry.com and 23andMe, it is possible to determine from a person’s genetic code, many aspects of their ethnic, cultural and geographic ancestry. Genetic markers of this nature will of course correlate to traits that are more predominant in one group than another, even if the genes in question have no causal relation to the trait, as the oft-used example of chopstick preference among those of Asian background suggests. The question then, is whether it’s all just chopsticks?

Even if one assumes that hidden among this miasma of noncausal correlations, are true genetic signals, we have had no real success in finding them. At least some such signals, I believe, would have been definitively discovered at this point if they existed, so I am inclined to posit that they do not, in fact, exist. I would say that the evidence points in that direction. Therefore, I will restate the First Law as Follows:

Any behavioral trait studied within a society will be correlated genetically to specific subpopulations, regardless of whether these genetic correlations are directly related to the trait.

Let’s move on to the Second Law:

The effect of being raised in the same family is smaller than the effect of genes.

On its face, it is not hard to see why I disagree with this, since I have just laid out why I do not think that genetic variations are responsible for behavior differences in humans. So even if being raised in the same family had no effect, it would be a tie. That said, let’s look at the argument for this. As already noted, genetic studies have not demonstrated any concrete effect of specific genes on behavior, so this law rides on twin study assumptions. There are plenty of reasons to question the validity of twin studies and some work has been done in that regard. However, I see little benefit in rehashing twin studies. The vast majority of the human population are not twins, and identical twins have reasons for their similarity that are not causally genetic.  Moreover, identical twins are far from identical in their expression of traits. This is conveniently ignored or chalked up to some vague concept of "nonshared environment," where no clear environmental factor can even be recognized. Thus, we have identical twins, where only one has symptoms consistent with schizophrenia. How is that possible, when they are genetically identical and no clear environmental factor can be established? It is far-fetched to think that families won't have some effect, yet which side of the crib or some such factor would have such a far-reaching effect on development. Yet there is little recourse if you wish to remain in a genetic determinist camp. It flies in the face of any kind of common sense, which might not definitively rule it out, but should give pause to anyone proposing this as an end-around to this conundrum.

Even disregarding such issues and accepting twin studies as valid, why am I comfortable disregarding genetic claims based on these studies? Because we have already assessed this with decades of genetic studies and we are not able to confirm the high genetic heritability assumptions made by these studies. Thus (and this is where I break with both the genetic and nurture sides of this debate), neither genes, nor family environment, nor a combination of these can adequately explain the differences between individual human beings. It’s worth noting that I am referring to what might be considered “adequate” family environments. Abusive or neglectful situations will most certainly damage individuals and affect their behavior, intelligence and personality in profound ways, as I observed repeatedly in my practice as a clinical psychiatrist (it is, in fact, a substantial portion of any psychiatrist's practice). So, leaving that aside, I will rephrase the Second Law as follows:

A substantial portion of the variation in complex human behavioral traits is not accounted for by the effects of genes or families.

The astute reader will notice that my restatement of the Second Law is actually identical to Turkheimer’s Third Law. Admittedly, I dispute the effect of genes, altogether, but I also do not attribute the rest to family or environment and make no apologies for the paradox. The problem is that most of the science on this subject, whether under the guise of behavioral genetics or human psychology, involves a turf war between gene hunters and environment proponents battling for this tiny sliver of the pie.

For behavioral geneticists, as Turkheimer noted, this is indeed a gloomy prospect. Likewise, for those looking for environmental factors to explain human behavior, this is also a gloomy prospect. I believe however, that I also need to posit a Third Law to address behavioral neuroscience, another gloomy prospect.

In addition to genetic studies, there have been a number of studies, generally using MRI or functional MRI’s and other brain imaging techniques, to try to delineate differences in the brain’s structure or function that might cause differences in behavior, intelligence and personality. These studies can often border on glorified versions of phrenology and are sometimes mixed with genetic studies. In any case, these studies have also been largely unsuccessful and, as with genetic studies, the proponents rely on an asymptotic shell game of false optimism, suggesting at some point in the future, with better technology, they will be able to make such delineations. The primary rationale, however, is that since genetic studies and environmental studies have been unsuccessful, there must be something in the brain that explains the differences between human beings. In other words, this is an assumption based not on solid evidence, but on the failure of other types of studies. I believe that this is yet another mechanistic dead end and will add my own (again with the above previously noted exceptions) Third Law:

Differences in human behavior, intelligence and personality will not be accounted for by structural or functional differences in the brain.

If these laws are valid, then what are we left with, you might ask? Before trying to answer that, it’s important to accept that psychology, or the study of human behavior cannot realistically be treated as a pure, hard science and looking for nice, little mechanistic rocks is, for the most part, a dead end. That doesn’t mean that behavioral genetics and neuroscience is a failure. On the contrary, these ideas were explored and not found to explain human behavior in any adequate way. Now we know this, so take a bow and move on.

Where do we go from here? I think the answer to that involves those in the field getting their hands a bit dirty. We can’t just be pilots in white gloves and expect to understand the workings of the mind. Psychology is experiential before anything else. Recording the stated experiences of others in some sort of passive, uninvolved way will not lead to an understanding of the human mind. It is not the place to start. You can no more understand the mind in that fashion than you can understand a song by a description or classification (Note: I expand on that idea in this blog post). One needs to delve into one’s own psychology in order to posit some theories or predictions. No amount of data collection will provide such answers. Thus, this brings me to a Fourth Law:

Advancements in understanding human behavior and psychology require inner exploration from the scientist, the subject or both.

I will refer to this as “The Exciting Prospect.” Will the same individuals currently researching these subjects be a part of such a prospect? I think, for the most part, they will not, although some might consider it out of professional frustration or scientific curiosity, and perhaps a new crop of “researchers” could take the field in some very different directions, even if the majority still stick with collecting the same rocks.

Some might accuse me here of proposing metaphysical or even mystical explanations to the human mind. While I have some of my own ideas on these subjects, I don’t say that’s necessarily the case. I do, however, leave that door open without apology and challenge the researcher to disprove it rather than dismiss it out of hand.

Before closing, and while I’ve indirectly addressed them in this piece, I’ll say some things about the Fourth and Fifth Laws of Behavioral Genetics added to Turkheimer’s original three for completion sake. First, let’s take a look at the Fourth Law from Christopher Chabris, stated as follows by Turkheimer in the link above:

Genetic effects are broken up into the minute effects of many genes.

This is the idea of polygenic inheritance, of course, which is accepted by most in the Behavioral Genetics field. In my view, it’s based largely on the failure to find single genetic variants or a small number of variants that have a large effect on a trait. In other words, it is an assumption, whose only successful correlate is for things like cattle breeding. We are now even getting rumblings of an “omnigenic” model in which all genes are said to affect a trait in some way. This leads us further and further into an unquantifiable, unfalsifiable theory where some kind of interactive, alchemical magic happens and we have our trait (Lest I be the one accused of mystical explanations).

Apparently, a Fifth Law was proposed by Emil Kirkegaard. Kirkegaard is a controversial figure with an arguably eugenic bent and is often focused on racial genetic superiority, so I hesitate to spend time on this, but also don’t want to shy away from it. His Fifth Law was also noted in the above-linked Turkheimer blog entry as follows:

Associations between phenotypes are confounded by genetic associations

I am suspicious of the motivation for this kind of idea, as it often  leads to race science types of explanations for traits. I will simply restate what I think is really the case (consistent with my reimagining of Turkheimer’s First Law):

Phenotypes with cross genetic associations are a good indication that you are working with population stratification.

So to recap my 4 Laws of the Behavioral Genetics Fallacy (with my second being the same as Turkheimer’s third):

1. Any behavioral trait studied within a society will be correlated genetically to specific subpopulations, regardless of whether these genetic correlations are directly related to the trait.

2. A substantial portion of the variation in complex human behavioral traits is not accounted for by the effects of genes or families.

3. Differences in human behavior, intelligence and personality are not accounted for by structural or functional differences in the brain.

4. Advancements in understanding human behavior and psychology require inner exploration from the scientist, the subject or both.