On August 25, Adam Keiper, editor of The New Atlantis emailed to give me his reaction to my initial critique of the new article by Lawrence Mayer and Paul McHugh on sexual orientation and gender. Please review that post before you read his remarks.
I appreciate Keiper’s professionalism in his remarks. I also appreciated a cordial phone call we had on Thursday. Keiper gave permission for me to publish his remarks. They are reproduced in full from the email. I plan to add my reactions over the next day or two within this post. I wanted to post his thoughts while the issue is current. My reactions to the part of the article on sexual orientation (I have yet to really examine the section on gender identity) are interspersed below.
In what follows, I would like to offer a few responses to some of the points you raise. I invite you to post this e-mail as an addendum to your piece on Patheos. I must note at the outset that I am not here writing on behalf of the authors of the report, nor as a scientist or physician (as I am neither), but rather as an interested reader of your piece and as the editor of The New Atlantis who worked closely with the authors on the report over the course of several months.
You begin by pointing out that “Sexuality and Gender: Findings from the Biological, Psychological, and Social Sciences” is not a study. You are correct. It is, as you rightly remark, a scientific review of the literature.
You are also correct in noting that The New Atlantis is not a peer-reviewed scientific publication. It is, rather, editorially reviewed — like many other journals and magazines intended for a wide public audience (such as Democracy Journal, National Affairs, The American Interest, The New Yorker, The Atlantic, etc.). When we publish essays and articles on technical subjects, our fact-checking process is especially rigorous, and in such cases we often ask experts to help our editorial team in its work. In the case of “Sexuality and Gender,” both our editorial team and the authors consulted with a range of experts in different fields. Peer review can be a very important part of the scientific publishing process. Our aim, however, was not to publish an original research study but rather to translate into accessible prose the scientific findings that were already published in peer-reviewed publications.
You next say:
Even calling the paper a new study isn’t accurate, there are no new studies in the paper. A bunch of old ones are missing as well.
I disagree with your characterization; there are some recent studies and papers that are mentioned in the report, including a handful that were published in 2015 and 2016. All told, I believe the report is quite up-to-date. I also would not say that old papers are “missing” per se; that seems to imply cherry-picking, which is an unsupported charge. Of course the authors of the report could not have discussed every paper in the vast scientific literature, but they selected the papers that they discussed on the grounds of their quality and scientific significance — emphasizing literature reviews and meta-analyses, pointing out when other significant papers contradict or criticize the literature reviews and meta-analyses, and then discussing more recent papers and studies that fill in gaps or further advance knowledge. Some older papers in the literature were deemed to be neither sufficiently important nor sufficiently rigorous to warrant discussion.
While The New Atlantis article was thorough, I still don’t think Mayer and McHugh captured everything necessary to make the claims they make. More on that below.
We are gratified to learn that Professor Bailey agrees with some (even if not all) of the report’s major findings; namely, that the social stress model does not alone account for all of the mental health difficulties experienced by LGBT people, that the empirical evidence does not support the idea that gender identity is innate and fixed, and that all of these issues should be studied more openly and rigorously by scientists. (Professor Bailey also mentions that “Sexuality and Gender” does not discuss his review of the literature on sexual orientation. Professor Bailey’s paper is very recent, and so our authors were unable to include a discussion of it before “Sexuality and Gender” went to press.)
Bailey’s paper was published online on April 25, 2016. Given the quality of Bailey et al’s work, I think the paper should have been included in their review or they should have waited to publish until they were able to include it.
Moving on to some of the topics that you suggest ought to have been included in the report but were not:
Mayer and McHugh’s paper is missing any serious discussion of epigenetics.
It is true that, other than a passing reference, this report does not discuss epigenetics. There is a good reason for that: the literature on epigenetics and sexual orientation remains inconclusive with regard to the question of whether homosexuality is innate and fixed. It is in the nature of epigenetic markers that they are (for the most part) acquired rather than inherited, and so, without a well-supported theory about why (for instance) gay men have distinctive patterns of DNA methylation, the fact that they have such markers does little to explain the origins of homosexual attractions, behaviors, or identity. (For what it’s worth, the Bailey paper only dedicates a single short paragraph to epigenetics [on page 77], and that paragraph basically says that the evidence doesn’t amount to much.)
This is exactly why the recent work on epigenetics should have been included in both papers. One would not need to go into it much in order to say that the line of research is intriguing and may yield answers after more studies are done. Mayer and McHugh’s paper needed to do this because they made a very definite claim: “The understanding of sexual orientation as an innate, biologically fixed property of human beings — the idea that people are “born that way” — is not supported by scientific evidence.”
Since epigenetic research may indeed be relevant to that claim, I believe it was an oversight not to discuss the where that research program could lead. Keiper’s statement, ” without a well-supported theory about why (for instance) gay men have distinctive patterns of DNA methylation, the fact that they have such markers does little to explain the origins of homosexual attractions, behaviors, or identity.” I agree that such a theory doesn’t exist. However, I am not making strong claims against innateness. Given the state of research at present, I don’t think making such a strong claim is warranted. More on this point at the end of the post.
You go on to claim that
[Mayer and McHugh] overlook the new genetic linkage paper involving gay brothers
I believe you are incorrect. “Sexuality and Gender” does mention the 2015 genetic linkage paper involving gay brothers (Sanders, et al., including Bailey). If this is not the paper to which you are referring, please let me know.
You are correct. I am sorry for that oversight and have corrected that error in my original post.
Having said that, I believe Bailey et al’s discussion is more thorough. They note the need for very large samples in genome studies and tell us that findings approaching significance were found in the large 23andMe study (see Bailey on page 77 and Mayer and McHugh on page 32). While 23,000 sounds like a large sample, there were only a few over 1,000 exclusively gay males in the study. We learn that from Bailey et al, not Mayer and McHugh. In other words, the methods we have available at present may not be sensitive enough to find the very fine biological differences which may move one individual toward same sex attraction and another toward opposite attraction unless very large numbers of people are involved.
You also note that the report ignores “work on ‘gay rams.’” It is true that “Sexuality and Gender” does not discuss homosexuality in non-human animals. While there is evidence of apparently exclusive homosexuality among domestic rams, this is not very convincing evidence that homosexuality is innate and fixed in human beings. And as Bailey and his colleagues note in their review (pages 68–69), although homosexual behaviors are frequently observed among wild animals, exclusive homosexuality has never been documented for animals in the wild. (It is worth acknowledging that documenting exclusive homosexuality among wild animals would be a difficult task, and so it is possible that there are exclusively homosexual animals that have not been discovered — but this only shows the need for more research on these questions.)
The gay ram research is important because of the parallels with research in humans. One would not posit weak fathers or child abuse to explain rams who prefer other rams. Apparently, something in the biology of the rams is involved. We are not rams but there are similar brain structures involved, some of which show up in studies of human brains.
You claim that
[t]he TNA authors minimize the neural differences between gays and straights, calling them ‘minor differences in brain structures.’ How do these authors know what differences are minor and which are not? In fact, the differences in symmetry and brain activity are quite provocative and have not been accounted for by any environmental theory.
The neurological differences between homosexuals and heterosexuals represent some interesting correlations, but they cannot be used to accurately predict whether a person is heterosexual or homosexual. (Moreover, as anyone familiar with the fMRI literature can tell you, it is very easy to find brain differences in brain-scan studies; e.g. this study found that when people were aware that they were drinking Pepsi or Coke, this “brand knowledge for one of the drinks had a dramatic influence … on the measured brain responses.”) I don’t see much distance between the ending of your last sentence quoted above and what the report says, as in the executive summary: “… [S]uch neurobiological findings do not demonstrate whether these differences are innate or are the result of environmental and psychological factors.”
Here is another example of rushing to a judgment that apparently the authors of TNA paper prefer. No one knows for sure what these brain differences mean. However, the authors seem to lean toward the conclusion that they are minor. They don’t have any reason to say that for the trait of sexual attraction. Soft drink preferences probably don’t matter much in the furtherance of the species. Reproductive preferences do.
Furthermore, brain responses are different than brain structures. I am not surprised that the brain responds differently to Pepsi and Coke (alternatively, any brain would count it all joy if it could get some Dr. Pepper). However, I am intrigued that gay male hemispheric symmetry is more like a straight female than a straight male in more than one study.
With respect to the elevated incidence of childhood sexual abuse suffered by LGBT people, you write that
For the most part, [Mayer and McHugh] report the relevant details but they failed to catch the mistakes in the Tomeo study and report it incorrectly.
I believe “Sexuality and Gender” covers the Tomeo et al. study well, and mentions important limitations of the study. You link to a 2009 post in which you describe an interesting error you and Gary Welton uncovered in the Tomeo study, in which contradictory figures were reported in a table. My editorial team will discuss the error in the Tomeo paper with the authors of “Sexuality and Gender,” but while the 68% figure may be wrong, my sense is that this does not really make a difference for the analysis in “Sexuality and Gender,” which does not make causal inferences about the relationship between sexual abuse and homosexuality. (As you wrote in your 2009 post, “the main results — gays report more abuse than straights — may indeed be correct, given the similarity to past studies. However, I do not believe any inferences about causation should be made.”)
First, I encourage editor Keiper to create a page of reactions to the paper and make readers aware that the Tomeo paper is flawed.
Then, speaking to abuse and causal inferences, Mayer and McHugh write on page 43: “The results presented below raise the question whether there is an association between sexual abuse, particularly in childhood, and later expressions of sexual attraction, behavior, or identity. If so, might child abuse increase the probability of having a non-heterosexual orientation?”
Then, on page 50 they conclude: “In short, while this study suggests that sexual abuse may sometimes be a causal contributor to having a non-heterosexual orientation, more research is needed to elucidate the biological or psychological mechanisms. Without such research, the idea that sexual abuse may be a causal factor in sexual orientation remains speculative.”
I will grant that these statements are measured and tentative but the inferences are made nonetheless. In fact, most gays were not abused.
Finally, you write that you
find it contradictory that the authors express uncertainty about the causes of orientation but then say with great certainty that the “born that way” theory isn’t supported by scientific evidence.
I don’t see the contradiction. One of the purposes the authors of this report had in writing it was to show the limits of what is scientifically known about sexual orientation and gender identity. The fact that there are no robustly supported alternative theories for explaining sexual orientation does not, in itself, provide evidence that the “born that way” theory is correct. More and better research that is open to both environmental and biological factors would be necessary to draw more definitive conclusions about the causes of sexual orientation.
Failing to find proof for a biological theory at this stage of knowledge and research doesn’t mean that prenatal causes aren’t important or even causative. It is a rush to judgment to word their conclusions as a lawyerly effort to mount a case against prenatal influences rather than to present a objective assessment of what we know and don’t know. Their statement comes across as an indictment against one viewpoint (“born that way”) which is a popular expression of belief in prenatal factors as causative. On the other hand, Bailey et al say on page 87: “No specific theory of what causes people to be attracted to men, to women, or to both has received enough support to win the backing of all reasonable scientists, most of whom remain open-minded to a large extent.”
Resorting the language of the culture wars (“born that way”) detracts from the value of Mayer and McHugh’s paper. Most personality theorists don’t think of the nature-nurture issue as disagreement over how people are born. For instance, I am moderately extraverted. Was I born an extravert? Since at the time, I wasn’t capable of preferring anything socially, I wasn’t born introverted or extraverted. I do think the evidence points to a role for genes in helping push me toward extraversion, my environment no doubt played a role in the way I am extraverted. I suspect there are some circumstances which could have made becoming an extravert improbable. So which is it? Was I “born that way?”
I would say there is a sense in which I was born with a high likelihood that I would become that way. If society told me that being an extravert was awful, I might even want to defend myself by saying that how I am is strongly influenced by factors outside of my control. And I would be at least partly correct. Saying a person is “born that way” is a way of communicating a lack of personal agency in a life trajectory that begins with a particular genotype within a particular environment. As an organism, my biology has to adapt to my environment. However, the likelihood that any environment will push me out of that trajectory is low. Some behavioral traits may be more malleable than others. Sexual orientation for men seems not to be terribly malleable whereas for women as a group (not every woman of course), the picture seems to include more fluidity. I think Mayer and McHugh, perhaps without intending this result, used scientific work to address a poorly framed question leading to a poorly framed, tendentious conclusion.
Scientists used to think cold, aloof mothers caused autism. At one point in history, scientists thought all obsessive-compulsive symptoms were psychologically based. Now we know mothers don’t cause autism and that sometimes strep anti-bodies cause OCD symptoms in children. While I agree with Bailey that the evidence has not caused consensus around a specific biological theory, I don’t think one can say that prenatal factors have been ruled out.
Having said all of that, I think research on specific social factors (parenting, abuse, early sexual experiences, pornography) is sufficient to cast serious doubt on any theory involving those factors as causative in a general sense. In short, some persons might owe their same sex attractions to a set of life experiences. However, according to the research on these factors, those persons are in the minority. What that means for the causes of homosexual attractions for the majority is not clear.
I want to thank Adam for his responses and I invite him to reply to me in another post. I think the give and take may help bring additional clarity and awareness to the important questions raised by the Mayer and McHugh and Bailey et al papers.