Grasping At A Straw

Written By Dr. Gerard van den Aardweg

( JONAH has been given permission to reprint this article by Dr. van den Aardweg and NARTH, National Association for the Research and Therapy of Homosexuality . )

Fact Or Artifact?

July 17, 2006 - Evidently, one cannot infer from the information in Anthony Bogaert's PNAS article (2006) that the (slight) correlation between male homosexuality and having more older (biological) brothers, which the author had previously so well established,[1] also exists when homosexuals are raised outside their biological family. Too many ambiguities must be clarified first. The criticisms of Neal Whitehead are certainly correct: apart from statistical doubts, the volunteer effect is more likely than not to have contaminated the results[2]. Had this effect not been demonstrated for homosexual volunteers in twin research, many would probably not have believed it could make such a difference.

In this case, it is unlikely that Bogaert's older-brother phenomenon or the purpose of the study was not known by at least a portion of the respondents; and as many homosexuals are inclined to talk about their own case, chances are that those with more older brothers are overrepresented among the raised-apart respondents. At any rate, this sample of homosexuals raised apart cannot be considered as representative, not even of raised-apart "gays," i.e., practicing or self-committed homosexuals (That the homosexuals in this sample had relatively young mothers is another indication of its atypical composition. One may also have doubts about the representative qualities of the sample of the raised-apart heterosexuals, for that matter).

We should be skeptical, indeed. First, because the only way to demonstrate that some correlation is not determined by sample specificities is to repeat it in other samples, and even more than once. The history of personality psychology abounds with examples of promising or interesting correlations which could not be confirmed in subsequent replications and the history of supposed associations of biological factors and homosexuality is no exception (think of the failed replications of alleged correlations with hormonal, anatomical, neuroanatomical, or genetic factors[3]). This is going on for about eighty years now: a correlation between homosexuality and a presumably "biological" factor found in some sample is launched as if a great break-through in the search for a cause has been made. Over-attention and simplification in the media create the impression that homosexuality must be somehow inborn. After some time and a number of replications, it becomes clear that the original correlation cannot be generalized, that is, was tied to the original sample. Then a new star, a new biological "fact" appears at the firmament, to die out equally after it has run its cycle. And so on.

The net effect scientifically is that no solid homosexuality-correlated bodily factor has ever really been demonstrated, but on the other hand that public belief in its existence flourishes and even prevails. This, by itself, is reason enough to skeptically raise our brows when, after this long history of failed attempts to validate hormonal, genetic, and neuroanatomical correlates, it is now claimed (amidst the inevitable media euphoria and simplification) that the more-older-brothers factor must point to a biological cause - since it is alleged to be associated with homosexuality also in one sample of men reared outside their own family. And it does nothing to decrease our skepticism when we realize how very hard the political gay rights faction is up for some fresh "scientific" support for their born-that-way or biological-variant position. For the mythical hormonal, genetic, etc. showpieces (Levay, Hammer, etc.), their "findings" are still drilled into the public as if nothing has been learned since their first publication, are wearing thin. This "brothers-effect" is a welcome study at a time gay political battles must be won. Few arguments are more helpful for the gay rights movement than that of scientific evidence for biological "causes." Besides, these "arguments" are no less important for committed gays personally, for they need them to persuade themselves that their lifestyle is natural and normal, against some stubborn uneasy feeling they constantly must suppress.

The term gay science reflects a deplorable reality, namely, the tendency of investigators and authors who want to promote the gay cause and who are often activist gays themselves, to select, present and interpret data or observations in an entirely one-sided way. They turn their research reports into a kind of propagandistic half-science. Part of this reality is the sponsoring of their publications by the academic community and the editorial boards of most professional periodicals in the social sciences, which creates the false impression that "science" is on their side. Yet, in fact, it is merely the culturally dominating "church" of ideologized or politicized scientists which is on their side, not objective science.

Yes, this article of Bogaert has a touch of gay science about it. There is too much disproportion between the weaknesses of the design of his study, the lack of transparency of the data he reports, and the lack of modesty with which questionable results are presented as near-proof of far-fetched interpretations. There is the superficial, biased commentary of strong recommendation in the Proceedings of the NAS which has published Bogaert's article[4], not balanced by a more sobering one. There is the blind praise from gay circles, along with the familiar dogmatic attitude towards those who dare to express criticisms. Anyhow, as long as the more-older-brothers effect has not been unambiguously observed in at least various replications of studies with homosexuals who have been permanently reared from early age in adoption families, Bogaert's finding may not be called a fact and much more probably is a mere artifact.

More Older Brothers In Homosexuals Not Raised Apart

Why "more probably"? Because of what is known in relation to this more-older-brothers effect in homosexuals reared in their original families. Either implicitly or explicitly, both researchers and therapists have been familiar with the effect for a long time (which does not detract from Bogaert's merits to have given the phenomenon a firm statistical footing[5]). It was more implicit in the data collected by Lang in Germany (1936) and more explicit in Schultz' interpretation of them (1937; Schultz pointed to the psychological figure of the "nice little brother," "das liebe Brüderchen")[6]. Van Lennep (1954, University of Utrecht, Netherlands) found that so-called non-clinical (socially adapted) homosexuals were more often than to be expected on the basis of the national population, the younger brothers from families of more than four children, while they were equally distributed among the first four children of smaller families[7]. Reflecting on such results, we must not forget that sometimes the number of older siblings has also been found to correlate with male homosexuality, which is psychologically no surprise; Bogaert himself reports it for his "study 2" sample, 2003). Westwood (1960) found relatively more youngest children and youngest sons in his English sample[8], the male homosexuals in therapy in the study of Bieber c.s. (1962) were more often than other therapy clients younger than older children in larger families[9], Siegelman's (1973) homosexual subjects had more brothers who were at least 5 years older than brothers 1-4 years older[10], thus were more often a late-comer ("after thought") in the family. And this list is not complete.

Incidentally, the older-brothers factor may to some degree be correlated with maternal age (see for instance the above mentioned older-sibling effect of Bogaert's 2003 "study 2", and various older reports mentioning elevated maternal ages[11]), although the raised-apart homosexuals of Bogaert's sample had relatively young mothers. But the reality may be more complicated. The Sbardelini couple, for example, could distinguish two clusters of maternal age in a non-clinical male group of Brazilian homosexuals: one group of teenage mothers and one of relatively old mothers[12]. The computation of correlation coefficients between maternal age and homosexuality might obscure such clusters (Had relatively many of the mothers of the raised-apart homosexuals of Bogaert, or of the adoptees, been teenage mothers?). The São Paulo homosexuals had also considerably older fathers than their controls. Thus, if the age of the parents may have a certain influence on the genesis of homosexuality, no matter how many siblings or brothers there are, there is no reason why it should not have been of influence in the genesis of homosexuality in men with more older brothers as well.

In effect, the homosexual man with a relatively older mother and father or with an older father only, and who was one of the younger children or one of the younger brothers, or the "afterthought," is a familiar "type." If I generalize a little bit, his father was often somewhat distant, his mother was the real significant person in his childhood. The boy was brought up softly, overprotected, maybe feminized to a degree, and/or dominated by mother in other ways, and/or infantilized ("babied"). He had the psychological traits of Schultz' "nice little brother." His mother may have preferred a girl after having several boys, or after the death of a sister just above him or about the time he was born.

Undeniably, he has often been her favorite. Not only clinical experience shows this over and again, but it is also suggested by the repeatedly reported fact that a high percentage of homosexual men in general have been their mother's favorite or were closer to her than their brothers (75% of the homosexuals in the sample of Bieber cs., for example[13]). There are no indications at all that this frequent mother-relation would not apply to the specific subgroup of homosexuals with more older brothers. The same is true for the classical psychological "father factor": the disinterested and/or, to a lesser extent, negative or hostile father, the lack of father-son confidentiality. These parental factors invariably characterize male homosexual samples and are the highest correlations that have systematically been found (next to childhood "same-sex peer isolation").

In contrast, as I stated above, there is no hard, at best speculative, support for possible biological correlates. In my 1986 book I gave a detailed review of mother -- and father (and peer) items -- differentiating varying homosexual groups, clinical and non-clinical, and from various countries, from heterosexual controls[14]. The few studies conducted afterwards gave the same results. (It is sometimes falsely contended by gay activists that the 1981 study by Bell c.s.[15] from the controversial Kinsey Institute disproved this correlational pattern. But also these authors came up with the same pattern, although it did not fit their debatable statistical path analysis model[16]). If there is anything that may lay claim to the status of a fact in relation to male homosexuality, it certainly is these highly significant correlations.

The childhood background factors and their variations account for more than two-third of the samples and individual cases studied. They clearly favor a psychodynamic explanation of homosexuality over a biological, as they essentially concern psychological and not biological dimensions. Since a minor portion of these samples consist of homosexuals with more older brothers, they favor a psychodynamic explanation of homosexuality for this minority too, namely, the very same that can be given for homosexuals from different sibling constellations (With "psychodynamic" I do not suggest that more or less classical Freudian explanations are the best ones. Upon more precise inspection of the data, they are not; but this is not at issue here).

Psychology Of Pre-homosexuals With Older Brothers

Now, considering that parental-plus-peer correlates are the best established, actually, the only incontrovertible facts associated with male homosexuality (and that they doubtless indicate psychological or psychodynamic causation), why wouldn't they be equally powerful in explaining homosexuality in men reared apart from the family of their birth? Hence it is a priori unlikely that homosexuality in raised-apart men with more older brothers would have quite different roots. According to my experience, when the psychological childhood background factors of adopted homosexuals are examined more closely, the same kind of pathogenic (foster) mother, (foster) father, and same-sex peer factors emerge and satisfactorily account for this erotic disorientation.

I know of no systematic study comparing early and permanently adopted homosexuals with similarly adopted controls with respect to these factors -- only of case studies. New Zealand's former vehement gay activist Noel Mosen was a case in point: "As an adoptive child, I was very lonely and had always the impression that my adoptive father treated me differently from his 'real' children. I pined for affection and recognition from him but had never the feeling to really belong to him. Then a man from the neighborhood, a pastor, took me under his protection. I enjoyed being accepted and loved that much by an elderly man" (who after some time seduced him. After many years of obsessive homosexual acting out, Mosen eventually became fully heterosexual. Looking back, he rejected any biological causation)[17].

Central in the psychogenesis of his homosexuality is this traumatic not belonging: not to the male world his foster-father represented, not to his siblings; and the ensuing longing for male affection. Nothing special; nothing we have not encountered hundreds of times in the childhood and pre-adolescence of homosexual men. He does not refer to the possible role of his foster-mother. He describes the father factor and the factor of inner isolation (and inferiority) from his playmates, his adoptive siblings in the first place.

Compare this psychological life story with that, for instance, which the homosexual Dutch novelist Louis Couperus (1863-1923) wrote about his young years. He was no adoptee but grew up in his biological family, the late-comer with a number of older brothers and sisters, whose masculine, strong father did not pay much attention to "that kid," but shared interests and activities with his older brothers and left him under the care of his wife, the maids in the kitchen, and his older sisters (who pampered him). The world of his father and older brothers was inaccessible to the boy and in his loneliness and grief he dreamed of getting attention from a sporting, cheerful young uncle, etc.[18].

The early adopted homosexual men whose psychological stories I got to know better were all placed in child-less families. I do not believe one of them had more older brothers (the latter category will probably be found in Holland only among adoptees from Third World countries, mostly from poor or inadequate mothers with large families and from prostitutes who got pregnant by different partners. They often do not know how many biological siblings they have. Therefore, I wonder how many similar cases may have found themselves among Bogaert's Canadian adoptees). The men in my cases were overprotected and/or dominated by their foster-mother, the foster-father seemed not much involved or problematic. In one case, the boy had been placed with an unmarried aunt after his mother's death when he was still a toddler.

The common childhood and pre-adolescence frustration and grief of these men had been their not-belonging to their contemporaries, to be unable to get friends (the peer factor), (foster) maternal over-bonding and overprotection, weak (foster) fathering, regular conflicts with him and emotional isolation from the same-sex community: these are precisely the significant predisposing factors in the childhood of the vast majority of same-sex attracted men. So adoptive couples can predispose to a homosexual development in the same way as the biological parents; this must be the explanation why one study found that 11% of adoptive brothers of male homosexuals had a similar orientation[19].

The existence of the same set of psychological factors in all kinds of homosexual samples, inclusive of those who were adopted, raises the question of whether the very same factors will not have been the principal pathogenic influences in Bogaert's raised-apart homosexuals with more-older-brothers as well, even if adopted boys with this brother factor (the most interesting raised-apart subgroup) could be proved to have a slightly enhanced probability of developing homosexually (This is a big if, see further on).

Moreover, it is wrong to dissociate homosexuality-linked sibling factors from these basic principal childhood influences. The number of older brothers is not an autonomous (orthogonal) factor but a function of the much more powerful maternal factor, often in combination with the paternal (psychological) absence factor. For chances are that certain mothers, especially, tie one of their younger sons to them, or favor them, overprotect them, etc., while the father keeps aloof. The consequence is that the boy is and/or feels like the outsider among his older brothers and this feeling or self-view of not-belonging to them, is part of his feeling an outsider in the men's world outside the home.

The behavior of his brothers towards him may aggravate the boy's lack of male confidence, but once he feels unable to be part of his brothers' world inside the home and of the other boys' world outside the home, the road to homosexual interests is paved. By itself, the number of older brothers does nothing to further a homosexual development, nor the duration of the boy's living together with them, except for some critical period, either in his childhood or pre-adolescence, when he may have consciously suffered for not being accepted by them and by his father.

What Is "Raised Apart" In Bogaert's Homosexual Sample?

The preceding considerations on the psychological origins of homosexuality lead to a few additional critical questions regarding Bogaert's experimental design. In the first place, it must be known if the main homosexuality-predisposing influences, the rather specific interactions with the parents we just described, were really absent in the cases of raised-apart homosexuals with more older brothers. Specifically, were they placed in second homes before maternal preference or special bonding could already shape the boy's personality to a degree, say, before 5 or 6 years at the latest? Or did the mother stay together with the boy in her second family, after parting from the father and the older brothers? Or again, are there cases of homosexuals with more older brothers who stayed with their father in a second liaison of his while their older brothers started living on their own?

It is not unlikely at all that there were already a small number of such instances where one of the homosexuality-predisposing parents stayed with the boy while he parted from his older brothers. This might invalidate the reported correlation. This possibility cannot be ruled out. For 55% of the second families of raised-apart homosexuals were partly constituted after the parents divorced; normally, the younger children stay with their mothers in her new household. Secondly, as Whitehead remarked, to evaluate the correlation between homosexuality and more older brothers, Bogaert's measure of number of years lived together, irrespective of which years, is psychologically too clumsy to be valid. What we must know is how the boy thought about his brothers, how he experienced himself in relation to them -- his subjective views and feelings. Otherwise, the impact of older on younger brothers does not vanish when they do not live at home any more; and much-older brothers[15] are more likely to leave home when the young brother is still a child. Similarly, there is no psycho-anamnestic information as regards these boys' relations with their adoptive or step-brothers.

Relevant Psychological Dimensions Left Out

In short, all information on the relevant life-history data and developmental-psychological factors of these pivotal raised-apart boys is painfully missing. This unpardonable neglect makes any attempt at interpretation a mere groping in the dark. A comparison of the homosexuals and heterosexuals, with and without, more older brothers -- with respect to the relevant psychological dimensions -- would in all likelihood have provided the necessary data for the alternative psychological explanation of the results. It would simultaneously have provided the clues as to why homosexuals with more older brothers were overrepresented in the raised-apart group as a whole.

In addition, it would have made it transparent if they were overrepresented in the adoptee group, or perhaps in the second-family group which contains such a lumped-together assortment of cases that it arouses our curiosity about the meaning of the notion "blended family." For the exploration of the statistical link between homosexuality and the more-older-brothers factor, a comparison of the psychological data of all homosexuals with those of all heterosexuals with this factor -- and of their psychological anamnesis -- would be especially interesting and instructive.

The combined raised-apart sample (N=521) must have contained at most 333 homosexuals (bisexuals and exclusive homosexuals together, going by the data at the end of the article) and 188 heterosexuals. Hence, a detailed comparison of life history and psychological data would concern about 25 homosexuals with more older brothers of the adoptee category and 25 of the "blended family" category, versus about 15 heterosexuals with more brothers in both categories (under the assumption that about 15% of all homosexuals and a somewhat lower percentage of heterosexuals have more older brothers).


Bogaert's statistics may be due to his sampling method or:

  • to the wrong assignment of some homosexuals with more older brothers to the raised-apart group, either because they had already undergone the impact of their biological family constellation before entering a second household
  • because one of their parents, who in fact caused the correlation between homosexuality and the brothers-factor, accompanied them to the second household
  • because living in another household than their older brothers did not mean for some pre-homosexuals that their brothers' psychological influence was annihilated.

Bogaert's assessment of sexual orientation by averaging two Kinsey-scale scores may have been another debatable point, as it is not clear in how far it corresponds with a clinical diagnosis, which to me seems preferable. The question is about the contribution of overt sexual behavior to the final score, notably during adolescence.

Consequently, Bogaert's claim to have demonstrated an association between male homosexuality and number of older brothers in men brought up without psychological family influences has not been substantiated.

It is, therefore, not meaningful to try to interpret an effect that has such a high probability of being an artifact. Bogaert, however, has no such scruples. For him, his discovery is virtually proof that some mothers produce anti-boy antibodies each time when they are pregnant with a boy, the accumulating effect which somehow would affect the brain of a later-born son so that he becomes a homosexual. His argument? Because his finding is "consistent with" that theory (I would just as easily rely on an explanation based on a boy's horoscope). Consistency with a theory does not decide the validity of a theory, however.

But the less humorous thing is that, although they do not explicitly endorse this visionary theory, the advocates of Bogaert's article in the same Proceedings of the National Academy, nevertheless present it as compelling new evidence for the biological cause of homosexuality.

And why? Because "since LeVay's 1991 report that a brain nucleus that is larger in men than women is also smaller in gay men than in straight men, circumstantial evidence has accumulated suggesting that some people are really born to become gay,"[21] then follow references to twin studies and Hamer's X chromosome "discovery." So, these authors never took the trouble to critically examine the studies they quote, and are carelessly ignorant of the fact that subsequent research failed to corroborate them. In fact, to the contrary, the more studies done in these fields, the more obvious it has become that there is no hard evidence for any biological cause[22]. Evidently, this is not scientific evaluating, it is the selling of gay myths.

Suppose Bogaert's findings would, indeed, be replicated a couple of times in acceptable samples? Such an outcome would by no means automatically imply biological causation. It could, for example, mean that some younger sons are born with certain characteristics or even defects which elicit overprotective maternal attitudes, so that not these (various) physical or psychic features would cause homosexuality, but specific maternal ways of reacting to them. This is surely very speculative, but if far-fetched biological speculations are taken seriously, why not also psychological ones?

At the end of the day, a biological correlate of homosexuality would almost certainly prove to be a predisposition, not a cause, and the real effective causes would still turn out to be those psychological childhood factors, which have shown the highest correlation with SSA.

The "circumstantial evidence" which is believed to have "accumulated" and which would "suggest that some people are really born to become gay" is imaginary today. To proclaim that it is virtually certain that "some people are really born gay" is not only scientific nonsense. Coming from the pen of academics such language is irresponsibly cheap and unscientific. If this deceptive and inexpert kind of commentary is all the wisdom the U.S. Academy of Sciences has to offer on occasion of a scientifically weak publication, the conclusion must be that it has adopted a new member: gay science.

P.S. Let us not forget that this discussion is about an extinguishing subgroup of male homosexuals. Having more older brothers presupposes families of at least 3 children, but the more-older-brothers correlation is based on families larger than that. As the average family size in most Western countries has dived below the two-child level, a 10-15% decrease in male homosexuality must be in the making if "biology" is behind the older-brothers effect. It doesn't look like that. Anyhow, speculations on physical causes of homosexuality in men with more older brothers concern a rapidly dwindling minority.

Gerard van den Aardweg, Ph.D., studied psychology at Leiden University and received his Ph.D. in psychology at Amsterdam University (1967) with a dissertation on homosexuality and neurosis. He has a private psychotherapeutic practice since 1963 in Holland, specializing in the treatment of homosexuality and marriage problems. He has written for many publications in these fields, and he has authored several books on homosexuality, including On the Origin and Treatment of Homosexuality (N.Y., Praeger, 1986) and The Battle for Normality (Ignatius Press). Dr. van den Aardweg is a member of NARTH's Scientific Advisory Committee.

[1] Bogaert, A.F. Journal of Personality and Social Psychology, 2003, 84, 3, 644-652
[2] Whitehead, N.E., "Do Mothers Create Gay Sons In The Womb," NARTH web site article, June 29, 2006
[3] van den Aardweg, G.J.M. NARTH Bulletin, 2005, 13, 3, 19-28
[4] Puts D.A. c.s., PNAS, 2006, 103, 28, 10531-10532
[5] Bogaert 2003, see above (1)
[6] Lang, Th. Zeitschrift für die gesamte Neurologie und Psychiatrie, 1936, 155, 702-713; Schultz, I.H. Zeitschrift für die gesamte Neurologie und Psychiatrie, 1937, 157, 575-578
[7] van Lennep, D.J. c.s. Research Report. University of Utrecht, 1954
[8] Westwood, G.A. Minority. London, 1960; Slater, E. The Lancet, 1962, 1, 69-71
[9] Bieber, I. c.s. Homosexuality. New York, 1962
[10] Siegelman, M. Archives of Sexual Behavior, 1973, 7, 1-11
[11] Slater 162, see above (8); Moran, P.A.P. British Journal of Psychiatry, 1968, 114, 207-214; Moran, P.A.P & Abe, K. British Journal of Psychiatry, 1969, 115, 319-320
[12] Sbardelini, E.S. & Sbardelini, E.T. Research Report. Catholic University of Campinas, S.P., 1977
[13] Bieber c.s., 1962, see above (9)
[14] van den Aardweg, G.J.M. On the Origins and Treatment of Homosexuality. New York, 1986, p. 78
[15] Bell, A.P. c.s. Sexual Preference. Bloomington, 1981
[16] van den Aardweg, G.J.M. American Journal of Psychotherapy, 1984, 38, 180-189
[17] Mosen, N. (German version) Medizin und Ideologie, 1997, 19, 18-30
[18] van den Aardweg, G.J.M. Nederlands Tijdschrift voor de Psychologie, 1965, 20, 293-307
[19] Bailey, J.M. & Pillard, R.C. Annual Review of Sex Research, 1995, 6, 126-150
[20] Siegelman,1973, see above (10)
[21] Puts c.s., 2006, see above (4; p. 10531)
[22] van den Aardweg, 2005, see above (note 3)