[…] Environmental Hereditarianism The Son Becomes The Father More Behavioral Genetic Facts […]

Concerning your update “Edit, 6/5/14” it is interesting to see that self-report measures have higher shared-environmental effect compared to criminal report. I’m not surprised by this result. These measures surely reflect some shared familial experience. If you want studies on the heritability of income, see here. Try to do an CTRL+F and enter “Hyytinen”. One other thing I have noticed is that EEA may be violated for income. Try to enter “EEA appears to be violated”.

Reblogged this on Philosophies of a Disenchanted Scholar and commented:
reference

[…] More Behavioral Genetic Facts – The sequel to the previous post, I continue to tie up additional dangling points and affirm the high heritability and lack of “shared environment” impact on traits such as IQ, criminality, emotional/mental problems. I talk about the extended twin design and how it can clear up some dangling questions, like who do we choose our mates? Do spouses influence each other? I mention the key findings of behavioral genetics, namely: […]

Hm… concerning Molenaar (2013) you should take my earlier comment with pinch of salt. They said that GxE interaction could have diminished GWAS heritability. I thought I could believe them, but after reading the references they cite, it says the opposite. It’s curious they mis-understood it, or maybe their sentence was poorly phrased. After all, GWAS heritability is supposed to get only the additive portion, and GxE can’t be additive, by definition.

Honestly, GWAs seem more important than what the skeptics tend to believe. Since GCTA/GWAS sample only the nonrelated individuals, thus no genetic similarity (unlike twins) the necessary consequence is to remove (almost) entirely GE correlations of all types. This argument does not work anymore. It’s finished. And who is attempting to use that argument again is either a complete ignorant of GWAS or he is still locked up in his ideology.

[…] have recently updated two key posts, my post More Behavioral Genetic Facts and More Maps of the American […]

[…] the dad despite the clear folly of this as per my earlier posts The Son Becomes The Father and More Behavioral Genetic Facts), who has his own 8-factor causal proclamation. It doesn’t occur to many of these people that […]

When you say rGE is often negative, it depends on which one you are thinking. Passive ? Reactive ? Active ? For the last one, it’s impossible, because it means that people tend to seek environments contra their own genetic propensities. It’s extremely difficult to conceive that. But I can easily conceive that parents and teachers prefer to invest more on low IQ children, thus, in that case you have your passive/reactive negative rGE. Such outcome seems very likely in most of the modern (western) societies where the dominant political orientation is the supra-egalitarianism (just look at how the new book of the french economist Thomas Piketty encounters its success). Now that is said, I want to precise that it seems unlikely that behavioral researchers think about passive/reactive rGE when they use that argument for the explanation of the increase in h2 with age. No, they think about active rGE. It’s obvious when you think that shared environment (c2) don’t have much impact in adulthood, and that rGE shifts from passive to active from childhood to adolescence/adulthood. Nonetheless, you can have a rather strong critic of active rGE from Brant (2013) “The Nature and Nurture of High IQ: An Extended Sensitive Period for Intellectual Development”.

The most prominent theory of developmental increases in the heritability of IQ posits that across development, individuals gain more scope to shape their own environments on the basis of their genetic propensities (active gene-environment correlation), which causes an increase in genetic influence over time (Haworth et al., 2010; Plomin, DeFries, & Loehlin, 1977). Our results challenge this explanation, as they show a later increase in heritability for individuals of higher IQ. To explain our results in the context of active gene-environment correlations, one would need to posit, counterintuitively, that higher-IQ individuals seek out environments concordant with their genetic propensities later in development than do lower-IQ individuals.

The reason for developmental increases in the heritability of IQ thus remains unclear. Other possibilities include amplification of existing genetic influence by increasing population variance in cognitive ability and the simultaneous limiting of environmental influences and introduction of new genetic influences as a result of synaptic pruning processes and myelination at the end of the sensitive period (Plomin, 1986; Plomin et al., 1977; Tau & Peterson, 2010).

Like I used to say, it’s interesting that they dismiss rGE when at the same time they have opened another possibility. The genetic amplification. To be sure, it’s John Fuerst who suggested me this idea, first. Look here.
http://occidentalascent.wordpress.com/2011/01/30/genetic-amplification/

Now, just some ideas, when I was reading your article :

Lot of people expect heritability (h2) to be upwardly biased. In the case of twins, it’s always MZ correlation that is suspected to be biased upwardly. One reason is rGE effects. Suppose that’s true. In that case, I expect rMZ to be much higher than the double of full sibling correlation, because the double is just what an additive model surely expects to find. Because rMZ seems to be just the double of full sibling correlation, that makes me believe there is a high genetic additive component in the IQ. But another, better way to disentangle that difficult question is to look at the h2 of the trait in question in different countries, preferably in very different social environments. For example, you can expect h2 in poor countries such as Africa and India for IQ to be lower. A failure to find difference in h2 would surely dismiss any rGE hypotheses. That’s the best test of the “locality” of h2.

I mention IQ but evidently it goes the same for all behavior traits. It’s just that I did not find the data for them. But if h2 is similar for, say, happiness, trust, agression, openness, and some other things like those, among different countries, with different political regimes, cultures, and different regions, e.g., rural versus urban, then if the h2 are quite comparable under diverse condition, it would seem that neither rGE or GE interaction is likely to produce most of the h2. When researchers attempt to use model-fitting for choosing which hypothesis needs to be retained, it’s no sufficient enough. They must always be accompanied with experiments. Even if your data tells you that you model looks very likely, has the “best fit to the data” that’s meaningless if experiments lead you to reject your model anyway. It’s the same kind of guys who believed they can model financial behaviors, coming to the conclusion that the crisis won’t happen. Well, we see it’s not true. The subprimes reveal quite a lot of bad investments.

Concerning h2 interaction with SES, I don’t think we should be surprised by a possible lower h2 at lower SES. h2 may be expected to move closer to 100% when environments become stable and/or better because in this particular case you don’t have much environmental variation, but on the other hand, the high-risk environments will just add more environmental variation and thus will act to reduce h2. That seems to be common sense. For instance :

One comprehensive review of class and health surveyed mortality rates in Britain from 1921 to 1971 (Black, 1980; Townsend & Davidson, 1982). Everyone was living longer, but the professional classes gained more years than semiskilled and unskilled workers. In 1930, people in the lowest social class had a 23% higher chance of dying at every age than people in the highest social class. By 1970, this excess risk had grown to 61%. A decade later, it had jumped to 150%. In Britain, a National Health Service has long existed to minimize inequalities in access to medical care. The increasing correlation of health and social class makes sense when one realizes that removing environmental impediments makes individual-difference variables more dependent on innate characteristics. (Placing intelligence into an evolutionary framework or how g fits into the r–K matrix of life-history traits including longevity, Rushton 2004)

This aside, the topic is highly controversial, and the topic is fulled of dishonest claims. Researchers always cite Turkheimer, but rarely the studies contra such conclusions. That said, I just wanted to point out this document.

Genotype by Environment Interactions in Cognitive Ability: A Survey of 14 Studies from Four Countries Covering Four Age Groups (Molenaar 2013)

It’s a very important article. They say explicitly that lack of representativeness can distort the direction and magnitude of the GxE interaction. At some extremes, maybe the interaction disappears, or magnifies. IQ measurement is also a possible cause of the inconsistency. Sometimes, it’s verbal IQ that is measured, sometimes, full IQ, sometimes nonverbal IQ. Another problem, less known, is the error measurement in the IQ. In ACE model-fitting, the E component also includes measurement error, and the portion of E that is tied to error variance can give rise to spurious GxE effects. Thus you must use multiple IQ measurements (see how many subtests in your battery, how many items per subtests, etc.) and try to reduce error variance as much as possible. But, interestingly, age can make a difference. Indeed, Molenaar and his team found that in childhood, there is negative GxE, which means lower E at higher level of G. In adolescents, no effect at all. And in adulthood, E was stronger (not smaller) for higher level of G. That runs contra Turkheimer/Rowe/Sluis/Tucker-Drob. Unfortunately, the final result is not easily interpretable because you have lot of differences between the studies and the countries (look their figure 2). In other words, it’s not sure that aggregation makes sense at all. The Molenaar paper is infinitely more important than the Hanscombe. I don’t understand why no one else cited it. Because it’s by far the best one available on that topic. And everyone should read it.

There is also another feature worth noting in Molenaar paper. The authors explicitly stated that the unmodeled GxE interaction might be one reason for the “missing” heritability in GWAS estimates. Another factor of under-estimated GWAS h2 may be population stratification, as mentioned by :

A genome-wide association study for reading and language abilities in two population cohorts (Luciano 2013)

Among the last paragraphs they wrote that when they exclude non-white people in the analysis, the correlation was significant whereas it didn’t when non-whites were included. Stats stuff is really, really, highly complex.

There are dramatic difference in average outcomes between *groups* proportional to how sealed off those *groups* are from the dominant culture in the post-60s West and how much those *groups* promote different behaviors (Mormons, Muslims, Amish etc).

One explanation is boiling off over generations.

Another possibility is that as humans are a social animal then as well as hereditability of individual behaviors there may also be heritability of group conforming behaviors. This might not alter the frequency of inherited individual behaviors but it would alter the level of display of those behaviors.

The change in the dominant culture since the 60s and the change in average behavior among groups not sealed off from the dominant culture – mostly media driven – has been so sudden and dramatic that boiling off is not an option imo so we have a very clear experiment showing that the average difference in behavior between *groups* in the same dominant culture is proportional to how much their group rejected the post 60s culture *and* critically how much their group was also able to seal themselves off from the dominant culture into a sub-cultural bubble.

The post 60s cultural experiment shows that both are required: rejection of the dominant culture and relative isolation from it.

So individual hereditability of behaviors yes, individual parental powerlessness against the dominant culture yes but also the hereditability of group conformity behavior means *groups* who isolate themselves into a sub-cultural bubble can resist the dominant culture.

Great post,

I should say that Harris’ theory talks of peer pressure to differentiate rather than to conform so it’s hard to see how this could turn up as shared environment. While the company is shared the treatment will be differential and it’s the treatment that is the social environment. I suppose the evidence would be found in correlations between peers as adults, if they stabilize as more (inversely) correlated than control groups. That said, these increasing heritability estimates are making the mystery of unique environment a shrinking one.

[…] culture isn’t something they have within themselves? Something that affects behavior while being highly heritable and stable over the lifespan? And why are there so many Asian Americans holding on to face culture […]

As you mentioned before labelling the “25% or something” of human traits as unique enverionment is misleading, since this might not be environmental at all. Possibilities include among others:

– alterations in somatic genome (Mosaicism, Chimerism, X-inactivation for women etc.)
– “noise” from small auto-immune reactions based at random (receptor rearrangement) events during T and B cell maturation
– microbiome (pathogenes alter somatic genome as well)

What do you think are the most likely sources of this 25% non-inherited variability? What percentages would you assign to each possible source?

—

With the advent of surrogacy the effect of uterine environment could actually be deduced. It is tricky, because they cannot be seperated from effects directly caused by the artificial surrogacy (implementation, in vitro fertilisation). Do you know any studies looking at uterine environmental effects?

“There are two aspects where spouses are highly correlated – the things you don’t talk about in a bar: politics and religion”

But what is your theory to explain the high correlation on religiosity and political trends between spouses? it seems you are suggesting a homotypic preference.

First we should bear in mind that within-pair matching for a feature may arise even if people pay no attention to this feature in prospective partners. So marry each other depends partly on the structure of the marriage market and social stratification, the effect propinquity / nearness has on our pool of potential mates, and also involves cultural norms of endogamy, or the choice to marry within a group.

Owing to physical proximity and ease to contact, people in a romantic pair tend to come from the same rather than different social groups. This is why partners are usually similar in characteristics associated with a social group, such as age (due to attending to schools), education and social status, etc.

On the other hand, I doubt that there is a mate preference for these kind of personality traits (i.e. religiosity, politics) determined by a genetic locus.

Furthermore, from an overall genetic framework, we should not forget such mechanisms as inbreeding depression and heterozygous advantage, so I’d assume that same-type matings are less fertile than different-type mating. If all individuals have a disassortative mating preference a viability-reducing trait may be maintained even without the fertility cost of same-type matings; a disassortative mating preference can be established even if it is initially rare, when there is a fertility cost of same-type matings.

Moreover, an assortative mating preference is less likely to evolve than a disassortative mating preference. This may be applicable to the evolution of MHC-disassortative mating preferences documented in animals and humans.

As you know, although Handscombe’s study failed to find an SES-heritability interaction there are a number of papers other than Turkheimer’s 2003 paper that have found this effect. Given that at least some of these studies have reasonable sample sizes and utilize the normal MZA-DZA model. Why do you think some studies find an interaction affect and others don’t ?

I feel that everything has a genetic component, and part of that is the susceptibility to a triggering event or the possibility that a therapeutic intervention can make a difference or not. Personally, I’d rather see the strong evidence for heritability and work around that for exceptions and modifications than deal with the “Environmental Causation” folks who don’t narrow down the factors they’re looking at because they are so insightful that they “look at the big picture.”

Even in the strongest arguments for the genetic heritability of behaviors and disorders, I don’t see such a degree of absolute determinism that modification to environment or behavior is completely dismissed out of hand.

>One key question: how do they assess “trust”? Just how good was their measurement? Measurements in social science need to meet three basic criteria: they need to be reliable (that is multiple testing instances of the same individual should give roughly the same results), they need to be “valid” (that is, be predictive of some real-world outcome), and they should be heritable.

Doesn’t the third criterion make the “all behavioral traits are heritable” a tautology?

You have gradually and quite completely won me over to your “extreme” position on the issues you discuss. Still, it seems to me their is room for misunderstanding (or maybe I am misunderstanding?) in terms of the take-away that a lot of people will have to your position. For instance, a generation or two back African Americans were famous for their good manners, no doubt because of the way they were taught be their parents in combination with the fact that white society would not tolerate the kind of bad manners (ghetto manners) which are so common today. Maybe you cover this when you emphasize that your position does not hold when different generations are being discussed?

Or take the issue of child sexual abuse: obviously different children are affected in different ways by such abuse (I know a woman for whom it seems to have triggered a borderline personality disorder) and let us stipulate for the purposes of argument that you are absolutely correct (and I believe you are) when you say these different reactions are mediated by genetic differences. Still, it would be a shame if people came away with the idea that child sexual abuse didn’t matter. Somehow you need to emphasize this side of your argument more effectively if you want it to be more widely accepted. I think.