Replies

“Is there a scientific study that makes this observation? If so, can you direct us to a source? Thanks.”

Many, many, many. I’ll see if I can’t quickly dig up a couple from different ends of the ideological spectrum, as I don’t currently have access to the last couple of large meta-analyses I’ve read recently. I recall a couple decent-enough smaller studies that have published working papers or published the full papers outside of a pay-wall.

First though, it might help to define a few useful terms and ideas. As a general summary of the basis of twin studies - nature gives us a wonderful tool to pry apart the “nature vs. nurture” question: twins. Naturally twins share the same gestational environment, be they monozygotic twins (commonly called identical twins) or dizygotic twins (commonly called fraternal twins) they are still twins and a pair of twins are in the same womb at the same time experiencing the same conditions. Monozygotic twins arise from a single fertilized egg and hence share the same genome, whereas fraternal twins arise from separate eggs and are only as closely related on average as any other pair of siblings.

This useful facet of nature gives us a ready-made laboratory to study how genes and nature interact. That is to say, if a given trait is entirely genetic we’d expect identical twins to display that given trait identically (concordance), and we’d expect fraternal twins to display that trait discordantly (non-identically).

However, reality is far more complex than that simplistic definition. The majority of traits we could look at, such as cancer risk, heart disease rates, height, IQ, schizophrenia, are a product of both the underlying biology of an organism as well as the sometimes unique developmental track that organism takes. Some human traits, like gender or aneuploidy, are purely genetic (despite the social dogmas of our age), whereas many other things are a product of both nature AND nurture, and finally a very few things are entirely the result of non-genetic factors - a mixed bag of nurture, experience, social influence, environmental exposure and everything else that isn’t adenine, cytesine, guanine and thymine. Further, gene expression is not constant and this variation can both hide the role of genetics in the development of a trait and at times serve as a part of complex gene-environment interactions. Finally, fraternal twins can be opposite-sexed and this can either illustrate or obscure sex-linked and hormonal issues.

But twins can allow us to quantify the relationships between the genome and “everything else”. If the trait or “thing” we are trying to measure is predominantly a product of “nature”, then by definition genetically identical individuals will show a high concordance (that is, the majority of identical twins will be identical for the trait).

On the other hand, if the “thing” we are trying to measure is predominantly a product of “everything BUT genetics”, then we’re going to see that identical twins are often NOT identical for that given trait. You’d see a high level of discordance, with one identical twin sometimes picking up the “thing” and the other identical twin not doing so, and only rarely would both identical twins have the “thing” you are trying to measure. This would likely be especially visible when those identical twins are raised apart from one-another in different environments, which is a rare circumstance that really illustrates nature vs. nuture.

Fraternal twins can serve as a very useful statistical tool to pry apart the influence of gestational environment, maternal nutrition, parental age, and birth order. Just as siblings can illustrate the differential roles of familial environment, birth order or general parentage. Comparing the rates of a given trait among and between identical twins, and between identical and fraternal twins, and between twins and non-twins, can give you hard numbers on nature vs. nurture.

So, as a thought experiment consider the statement that “People are just born with THIS-TRAIT”. This is actually an eminently testable statement, regardless of what “THIS-TRAIT” actually is. If people are born with “THIS-TRAIT” we would expect that a pair of monozygotic twins would both be born with “THIS-TRAIT” or both would not have “THIS-TRAIT”. If the trait in question is entirely genetic, rather than partially influenced by gestational development, then we’d expect identical twins to be totally identical when considering the trait and conversely we’d expect fraternal twins to show high discordance.

If something besides genetics is the major driver of the trait, but people are still “born that way”, then we’d expect both identical AND fraternal twins (at least fraternal twins of the same sex) to frequently be identical for that trait, all other things being equal. This is where we’d see things like a birth-order effect, maternal nutrition, and things like gestational exposure to outside hormones (like the freemartin effect you see in cattle from a differently sexed fraternal twin) or gestational exposure to contaminants influencing what traits someone is born with.

However, if we find that identical twins are rarely identical for the trait, then ipso facto the trait is not predominantly genetic and is likely not predominantly the result of gestation - there are non-genetic and non-gestational factors at play that have more influence. Said another way, if there’s high discordance among twins, that means that by definition the most influential cause of the trait is something that happens AFTER birth. This doesn’t mean that there’s NOT some genetic predisposing factor that increases the rate of what you’re measuring when combined with a specific environmental trigger(s), it just means that what’s actually CAUSING the trait to appear is something that happens after birth and is largely the product of non-genetic factors.

Unfortunately there are a few confounding factors that can obfuscate the truth for a while. Firstly twins, especially identical twins, are not all that common; it takes a lot of time and a lot of sampling of a large population to get large sample of twins, let alone a statistically valid cross-section of the wider population that accurately reflects the totality of human variation and humanity’s circumstances. As an aside to that point, you’ll often see studies which rely heavily on self-reporting from western, English-speaking, educated, and affluent individuals - often college students - simply due to the fact that this is the type of person a researcher can most easily collect data on. Secondly, it is exceptionally difficult in either academia or society at large to unemotionally and clinically evaluate issues that touch upon highly personal topics. Issues of politics, ideology, identity and personal expression are particularly thorny creatures; the personal bias of the study participants, the goals and ideals of the researchers, the politics and financing policies of grant committees and academic institutions, the orthodoxy and ideologies of the journals which may or may not publish the findings, and the motivations of media conglomerations that disseminate highly abbreviated and often wildly inaccurate summaries of studies, all combine into a thick miasma of personal interest and misdirection that clouds accurate inquiry. The culture of “publish or perish” doesn’t necessarily support detached, well-supported clinical inquiry either.

I encourage you the reader to not blindly swallow my conclusions and opinions or the conclusions and opinions of anyone else. I’d further advise you to treat your OWN opinions and conclusions as suspect, and constantly re-evaluate them based on objective and empirical information. Take what you read and hear with a grain of salt, demand to see the numbers before you accept someone’s conclusion, understand the limitations of mathematical models, and be ready to revise the conclusions you have accepted if better information comes along. At the end of the day demand the same rigor of your own ideas as you would from your political opponents, and be most careful about accepting the things that you WANT to hear.

On to the studies.

Bailey, J. Michael., Dunne, Michael P., Martin, Nicholas G. (2000) “Genetic and Environmental Influences on Sexual Orientation and Its Correlates in an Australian Twin Sample” published in the Journal of Personality and Social Psychology

A PDF copy of the study can be found on Bailey’s faculty page at Northwestern U: http://faculty.wcas.northwestern.edu/JMichael-Bailey/Publications/Bailey%20et%20al.%20twins,2000.pdf

If the %20 messes up your web-browser’s attempt to access the URL, simply replace each %20 with a space and it should work fine.

Bailey(et al.) found a very low concordance among their sample of Australian monozygotic twins, and the study was even utilizing a somewhat fuzzy Kinsey “psychological rather than behavioral” definition of homosexuality - in other words, such a metric would count participants that have never engaged in a homosexual act as homosexual, vastly inflating concordance.

This is utterly aside from the, ahem, “overall validity” of Kinsey’s work (Alfred Kinsey, of Kinsey report on human sexuality fame, was all the talk around enlightened coffee tables in the 60’s). Given that Kinsey thought a sample consisting of some 20%+ prison convicts is statistically representative of the overall U.S. population in terms of sexuality (he’s refused to release his sample data for his reports on human sexuality, so we can’t judge for certain how screwed up the Kinsey reports really are), given that background of scientific integrity you would expect researchers to attempt to distance themselves from his work. But, even flawed metrics do have some use. The Kinsey scale is used in the social sciences and serves as an easy tool to compare different studies (even a broken ruler can be useful if you care to standardize it).

As a summary of the Bailey study, out of 387 male Australian monozygotic twins, Bailey(et al.) only reported 3 that were concordant for Kinsey scale strict homosexuality, compared to 9 that were discordant.

Bailey(et al.) reported an even greater level of discordance among female monozygotic twins, with 3 pairs concordant and 19!! discordant for strict Kinsey scale homosexuality out of 561 monozygotic female twins.

In regards to that last note, we tend to see that gender difference in concordance in quite a number of twin studies. The implication would be that male and female homosexuality may have different etiologies and/or different expressions. As an anecdote, this supposition is supported by wider cultural trends in the GLBTwhatever communities, with lesbian women (especially radical feminists) far more likely than gay men to describe their sexual orientation as a form of social or political ideology rather than some deeply in-born sexual attraction. It would appear from some studies that development of homosexual tendencies in women also take place later in life than they do for men. If I recall correctly off the top of my head, the Minnesota study of twins reared apart (again, IIRC) found a single pair of female monozygotic twins in which at least one of the women was homosexual - if memory serves one of the twins was a lesbian, one was straight, and it would appear that the difference first manifested in late adolescence, with one twin putting on some extra weight and feeling like a bit of a social outcast in junior high/high school and the other twin experiencing a more normal early social and dating scene in her late-teens. If I recall, the socially awkward and overweight teenage girl who never quite caught the eyes of the opposite sex went on to identify as homosexual.

Back to the study though. Keep in mind, Bailey has made statements to the press that he believes even with exceptionally low concordance in twin studies that these studies indicate a primarily genetic basis for homosexuality, citing a higher overall rate of homosexuality among families that have homosexual members compared to the general population as support for his belief. Such an assertion is naturally fraught with assumptions and pitfalls; there are many non-genetic factors conflated into familial environments that could reasonably be the drivers of the phenomenon. Without adjusting for all possible environmental and social factors involved blindly assuming that higher familial incidence of a trait is purely genetic is outright dangerous; as an example increased rates of cancer primarily due to environmental or behavioral exposure to carcinogens would likely show up as higher cancer rates among families with a cancer-striken member as well. But, some researchers don’t let statistical validity or sound reasoning get in the way of publishing what they (or their grant committees) want published.

Moving on to the next convenient study.

Bearman, P. S. & Bruckner, H. (2002) “Opposite-sex twins and adolescent same-sex attraction.” published in the American Journal of Sociology

A working paper, similar to the published paper but lacking charts and tables, can be found here: http://iserp.columbia.edu/sites/default/files/working_papers/2001_04.pdf

Bearman and Bruckner approach the topic from a different direction. Whereas Bailey(et al.) have championed a genetically-deterministic explanation of sexual orientation, in spite of strong data to the contrary such as high discordance among twins, Bearman and Bruckner focus chiefly on designing adequate methodology to objectively test for social, genetic, or hormonal influences on same-sex attraction. Both studies show the same thing, very high discordance among twins for same-sex attraction, but Bearman and Bruckner either don’t let their ideology detract from the results or the results fit their ideology.

As a summary, Bearman and Bruckner briefly discuss and highlight the findings of numerous other studies, and briefly discuss the problems inherent in constructing valid samples to study same-sex attraction. Using a self-reporting sample derived from the National Longitudinal Study of Adolescent Health, a study of 7th to 12th grade students in the United States that one of the researchers helped collect, Bearman and Bruckner assembled a sample of 527 pairs of twins to judge a genetic-basis hypothesis. The researches asked the participants if they had ever had a romantic attraction to a member of the same sex. Bearman and Bruckner found an even lower concordance than Bailey(et al), stating that there was only 6.7% concordance for homosexual attraction among monozygotic twins, and 7.2% concordance for homosexual attraction among dizygotic twins.

Both the Bailey and Bearman studies indicate a VERY low rate of concordance among their respective participants. The studies are not directly comparable on a fine scale due to different sampling methods and different breakdown & presentation of their data sets, and both studies face some methodological problems due to the inherent difficulties in collecting an ideally representative sample, but in spite of these differences and shortcomings some broad and widely applicable themes are apparent. There is a reproducibly high discordance among monozygotic twins when measuring self-reported instances of homosexual attraction, which is exactly the opposite result we would expect to see if either a genetic explanation or gestational explanation of homosexuality were correct. These findings would only support either a socialization or a developmental environment hypothesis.

However, don’t expect these findings to be widely accepted any time soon. It took three decades before the Soviet Union abandoned Lysenko’s neo-Lamarckian bullshit and admitted that heredity existed. Politically appealing fallacies die hard.

As a side note, both of those studies I linked have several citations of other studies that you could look up for more data sets.