The Naturalistic Fallacy, which most of you surely know, it the erroneous equation of what does exist with what should exist. Discussed extensively by Hume, it is the false equation of “is” with “ought”. In biology, it takes the form of observing some behavior in animals that is similar to a behavior in humans, and then justifying or saying the human behavior “natural” or “good” because we see it in other species.
But this is a bad argument, for it cuts both ways. After all, animals show a lot of behavior that would be considered reprehensible or even immoral in humans. In fact, Joan Roughgarden wrote a book, Evolution’s Rainbow, which describes sex and gender diversity in nature as an explicit way of justifying similar behaviors in humans as good—because they are natural. I reviewed the book for TLS and wrote this bit (review no longer online but I can send a copy).
Coyne, J. A. 2004. Charm schools. (Review of Evolution’s Rainbow, by Joan Roughgarden). Times Literary Supplement, London. July 30, 2004 (No. 5287), p. 5.
But regardless of the truth of Darwin’s theory, should we consult nature to determine which of our behaviours are to be considered normal or moral? Homosexuality may indeed occur in species other than our own, but so do infanticide, robbery and extra-pair copulation. If the gay cause is somehow boosted by parallels from nature, then so are the causes of child-killers, thieves and adulterers. And given the cultural milieu in which human sexuality and gender are expressed, how closely can we compare ourselves to other species? In what sense does a fish who changes sex resemble a transgendered person? The fish presumably experiences neither distressing feelings about inhabiting the wrong body, nor ostracism by other fish. In some baboons, the only males who show homosexual behaviour are those denied access to females by more dominant males. How can this possibly be equated to human homosexuality?
The step from “natural” to “ethical” is even riskier. As the philosopher G. E. Moore argued, identifying what is good or right by using any natural property is committing the “naturalistic fallacy”: there is no valid way to deduce “ought” from “is”. If no animals showed homosexual behaviour, would discrimination against gay humans be more justified? Certainly not. Roughgarden’s philosophical strategy is as problematic as her biological one.
Now a 2022 paper in Nature Communications had the potential to demonstrate the same fallacy, but fortunately the authors went to great lengths to avoid that The same, however, is not true of a new take on this paper in a new article in ZME Science, which gave a précis of the paper and stepped on the Fallacy’s tail.
First the Nature paper itself, which you can access by clicking on the article below, or by reading the pdf here.
It’s a good paper on the evolution and phylogeny of “same-sex sexual behavior” in mammals, which they define as “transient courtship or mating interactions between members of the same sex“.
Note that it’s “transient,” which explicitly excludes homosexuality, most notably in humans, which is a persistent sexual attraction to members of one’s own biological sex. This form of transient sexual interaction is surprisingly common—a conservative estimate is 4% of all animal species, and, as the authors say, [includes] “all main groups from invertebrates such as insects, spiders, echinoderms, and nematodes, to vertebrates such as fish, amphibians, reptiles, birds, and mammals.”
Now there are two ways to explain a behavior that seems on its face maladaptive. Why would you engage in sexual behavior that doesn’t involve passing on your genes? One hypothesis is that it’s just a nonadaptive byproduct of other behaviors: a general drive to mate when the appropriate mates aren’t available, or simply mistaken identity. But the authors investigate two hypotheses that it is adaptive, and give some tentative evidence for that.
First, the results:
- The authors did a comprehensive survey of same-sex behavior (defined above) in 2546 species of mammals, and superimposed species with and without such behavior on their phylogenetic tree. The object was to see how many times the behavior evolved independently, and whether it was present in the common ancestor of a group (and thus could be passed along to its descendants). Here’s one of those phylogenies with the caption. (You needn’t worry about the details or summary, as I’ll give it below).
A summary:
- The behavior was reported in 261 mammalian species
- Same-sex sexual behavior appears to be equally common in males and females, and the behavior in both sexes tends to be correlated across groups. That is, male and female same-sex behavior is more likely to both appear in the same species than if it either were distributed randomly among groups.
- It was not possible tell, using phylogenetic analysis, whether same-sex behavior was likely to be a trait in the ancestor of all mammals, but was NOT likely to be a trait in the ancestor of all placental mammals.
- The behavior seems to have evolved independently in many lineages, so same-sex sexual behavior seems to be a case of “convergent evolution.”
- The behavior is correlated with whether or not a species is social. If it is social, there’s a significantly higher probability of same-sex sexual behavior. (Remember that this is a correlation and doesn’t imply that sociality prompts the evolution of such behavior. The behavior could simply result from iondividuals in social species being closer to other individuals than those in non-social species.)
- The common ancestor of all primates does seem to have possessed same-sex sexual behavior.
The association of same-sex sexual behavior with sociality leads the authors to conclude that the behavior evolved by natural selection as a way to enforce inter-individual harmony required by sociality. They mention two such advantages:
1.) Same-sex sexual behavior is a way of creating and maintaining social bonds between individuals in a group; it’s a bonding mechanism.
2.) The behavior could also help prevent or resolve conflicts between members of a group, allowing a hierarchy to develop without injury of death to group members.
The authors mention that these effects have been demonstrated in some species like bottlenose dolphins and American bison, but I’m not familiar with this work, and such conclusions seem to me to be extraordinarily difficult to arrive at. However, I’ll take the authors’ word for it.
The authors are, to be sure, careful in their conclusions. First, they note that nonadaptive hypotheses, like “mistaken identity” could also contribute to the behavior.
Second, and the big one, they note that the behavior they studied is not the same as homosexual behavior like we find in humans. They do add, however, that it humans do show same-sex sexual behavior in humans (I presume they’re referring to “bisexual” people who have sex with both males and females). From the paper:
However, same-sex sexual behaviour is operationally defined here as any temporary sexual contact between members of the same sex2. This behaviour should be distinguished from homosexuality as a more permanent same sex preference, as found in humans. For this reason, our findings cannot be used to infer the evolution of sexual orientation, identity, and preference or the prevalence of homosexuality as categories of sexual beings Nevertheless, even taking into account this cautionary note, by using phylogenetic inference, our study may provide a potential explanation on the evolutionary history of the occurrence of same-sex sexual behaviour in humans.
They may be right, but I think they should have added that even if same-sex sexual behavior was rare or nonexistent in mammals, its existence in humans is not made “ethical” or “natural” in our species. That would be an example of the naturalistic fallacy, and I emphasize that they do not commit it. I’d would also emphasize, as I did above, that any sexual behavior between consenting human adults is not for us to judge, regardless of whether or not other species show it, and that such behaviors are fine so long as they’re legal. We don’t need to justify same-sex sexual behavior in humans by seeing it elsewhere in nature. But perhaps this stuff doesn’t belong in a scientific paper. But I want to emphasize it here, as I did in my review of Roughgarden’s book.
As I said, the authors don’t commit the naturalistic fallacy, but the new ZME Science paper below comes close to it. Click headline to read:
Up until the end, this article is okay, but then it can’t resist diving into our own species (bolding is mine).
However, the researchers distinguish between SSSB and sexual orientation. While SSSB involves occasional same-sex interactions, sexual orientation encompasses consistent patterns of attraction and identity, particularly prominent in humans.
While SSSB in animals supports the naturalness of such behaviors, human experiences of sexuality include layers of identity, culture, and personal meaning that go beyond biological explanations. Homosexuality in humans often involves stable sexual orientations and relationships, distinct from the transient or context-dependent SSSB observed in some animal species.
Ultimately, the widespread occurrence of SSSB in mammals, especially primates, strongly suggests that such behaviors are natural and adaptive. Normalizing same-sex behavior as a part of this spectrum aligns with both biological evidence and a broader understanding of human social and emotional complexity.
The last paragraph explicitly says that the results show that homosexuality (one of “such behaviors”) is “natural and adaptive”, as are all “same-sex behaviors” in humans. The Nature paper says nothing of the sort. The authors of the Nature paper explicitly exclude homosexuality as not a behavior they studied, but ZME Science lumps it in with other same-sex sexual behaviors, dwspite homosexuality being very different from SSSB.
Again, you do NOT need to justify same-sex sexual behavior, whether it be transient or permanent, by finding examples in the natural world. If we didn’t find any other species with homosexual behavior, would that make it wrong or bad in humans? Of course not! “Is” does not equal “ought,” and I’ll add the corollary that “not is” does not equal “not ought”. The Nature paper is valuable it looking at the evolution of a behavior and testing hypotheses about its adaptiveness, but of course adaptiveness or evolution has nothing to do with the ethics of behaviors between consenting human adults.
Comment by Greg Mayer
I should, of course, read the paper to see if they address the issues, but the first thing that came to my mind is that mammalian species have been very unevenly studied from the point of view of how often a rare sexual behavior occurs. This makes it very difficult to do a comparative phylogenetic analysis, because you don’t know where on the tree it really occurs– you only know for well-studied species.
And as for the association with sociality, rather than thinking it’s an evolved consequence of sociality, I would want to exclude not only “social equals more opportunities” as an explanation (which they apparently do consider), but also “social equals better studied”: the apparent association may be generated by biased attention towards social species. Many animal behaviorists will sit for hours observing social interactions, but it’s much harder to observe numerous interactions among individuals of a solitary species to see if there’s occasional same-sex sexual behavior.
GCM
Yes, it is a great paper. We discussed it in a journal group in my department when it first came out almost a couple of years ago.
With regards to the first point above, I completely agree. But one would be amazed at the variety of species with documented observations included here.
A note on the source of the vast majority of the observational data for the paper: it is from a book called “Biological exuberance (1999) by Bruce Bagemihl. I have it and it is a very fun book to open randomly and read about one of the hundreds of species there. The weird thing is that as far as I can tell the identity of the author is a big enigma. Last I checked we basically only know “from Wikipedia”:
“He completed his BA at the University of Wisconsin–Milwaukee in 1981, and served on the faculty of the University of British Columbia, where he taught linguistics and cognitive science. He earned a Ph.D. in linguistics from UBC in 1988, with a dissertation entitled Alternate phonologies and morphologies.”
I am fascinated by this: someone who studied linguistics and cognitive science, compiled evidence of same sex sexual behaviour in hundreds of species, wrote a book and disappeared. In case anyone is interested, I heard a good podcast about it a while ago (https://www.canadaland.com/podcast/998-nature-is-so-gay/).
When a behavior is found to be more common in social species, I can’t help but wonder if it is just easier for humans to observe in social species. Edit: Greg Mayer already said it, and more. So let me say instead, I agree with Greg!
I wonder what the adaptive function of this behavior is:
https://www.cbc.ca/radio/asithappens/as-it-happens-the-tuesday-edition-1.4962804/horny-confused-cane-toads-are-humping-a-python-says-biologist-1.4962815
Comment by Greg Mayer
Toads and frogs will mate with anything– it’s a minor theme in the “gee whiz” literature of herpetology. Attempting to mate with pythons isn’t adaptive. But having a very low bar for the stimuli needed to evoke mating behavior may be adaptive. Female anurans are attracted to calls of conspecific males, and will approach them. If something other than a receptive female gets close to a calling male, he may grab it and amplex with it. (“Amplexus” is what mating is called in anurans.) In at least some anuran species, males have a release call which signals to a same-sex suitor that they have grabbed the wrong guy, and should let go.
GCM
I speculate that some species have stronger and more frequent sexual urges that are independent of mating. And that horny feeling leads to seeking physical relief. There are also parallels with grooming behavior as a bonding mechanism as the authors suggest.
In humans, there are so many variations of sexual behaviors, the biological imperative is only one factor. Individual biological and psychological differences and cultural influences are at least as important.
There is a pretty good hypothesis for the evolutionary maintenance of same-sex sexual behaviour in humans. The same genetic variants associated with SSB in some individuals are also associated with larger numbers of sexual partners in other individuals who show only opposite-sex sexual behaviour (OSB). Huge study, genome-wide association analysis. The mechanism could explain how SSB persists as a heritable human trait in spite of its fitness costs (many fewer offspring than OSB individuals).
Zietsch et al. (2021) Genomic evidence consistent with antagonistic pleiotropy may help explain the evolutionary maintenance of same-sex sexual behaviour in humans. Nature Human Behaviour. DOI10.1038/s41562-021-01168-8
I don’t understand the confidence given in the third bullet point that SSSB was “NOT likely to be a trait in the ancestor of all placental mammals”, and the fourth bullet that SSSB “seems to have evolved independently in many lineages”. Is that because there are a couple ‘islands’ of mostly Opposite-SSB? I see the cats and a couple other places are pretty much “straight”, with white pie charts. It seems to me that SSSB could be a primitive trait among mammals, as it is very widespread among all 3 of the mammalian clades, but was secondarily lost in some parts of the tree.
And the acquisition or secondary loss of SSSB might not be targeted by selection at all. SSSB (and for that matter wanting to hump just about anything of approximately the right size) could just be something that is neutral to natural selection as an unintended but selectively neutral byproduct of strong selection to mate with conspecifics. That makes SSSB more like a spandrel, I guess.
While same-sex sexual behavior (SSSB) has been observed in a range of animal species, its interpretation through the lens of genetic or adaptive value remains contentious. From a genetic inheritance perspective, behaviors that do not contribute directly to reproductive success pose an evolutionary paradox. The Darwinian model of selection suggests that traits which consistently reduce reproductive fitness are unlikely to be strongly selected for or maintained across generations. In this view, SSSB—if not contributing to inclusive fitness or kin selection—raises questions about its persistence across evolutionary timescales.
From a genetic standpoint, there is limited evidence for specific alleles or genetic structures that consistently promote same-sex preferences across species. Studies in humans, such as the large-scale GWAS by Ganna et al. (2019), indicate that no single “gay gene” exists, and instead, multiple loci of very small effect may be involved—none of which explain a significant fraction of the variance. Extrapolating this to non-human animals, it becomes difficult to establish a consistent genetic architecture for SSSB.
Epigenetically, transient environmental influences or developmental plasticity could play a role. Some researchers propose that hormonal imbalances or disruptions during critical periods of neurodevelopment might contribute to atypical sexual behavior, including SSSB. However, epigenetic mechanisms (e.g., DNA methylation or histone modification) often operate in highly specific, tissue- and time-dependent manners, making it unlikely that stable same-sex orientation patterns could be widely inherited epigenetically without clear selective benefit.
From a cognitive science standpoint, animal behavior—especially in more neurologically complex species—may include play, dominance assertion, or social bonding mechanisms that mimic sexual behavior but are not necessarily indicative of fixed orientation. In bonobos, for instance, SSSB is often understood more as a social bonding mechanism than a true sexual preference. Similarly, behavioral controls in higher mammals are influenced by environmental cues, stressors, and social hierarchies, which can modify sexual behaviors in context-specific ways.
In sum, while SSSB is undeniably observed, its genetic and epigenetic basis remains insufficiently understood and likely non-adaptive in most contexts. It may be better interpreted as a byproduct of broader neurobehavioral flexibility rather than a trait shaped by directional selection.
Very good points! Saying that SSB is non-adaptive is not the same as saying it is unaffected by selection. And saying the architecture of the trait is likely to be different in humans vs. other animals isn’t really a criticism of this area of research because the behaviour is mostly of interest in humans.
Understanding of the genetic architecture of SSB in humans is incomplete but non-zero. Ganna et al. (2019) showed only that the heritability of same-sex sexual behaviour is polygenic like almost all behaviours. The same authors (Zietsch et al. 2021 Nature Human Behaviour) used similar data and methods to show that alleles that are significantly over-represented in individuals with same-sex sexual behaviour also have pleiotropic effects on number of sexual partners in individuals with only opposite-sex sexual behaviour. They suggest that these alleles confer a mating advantage in individuals with opposite-sex sexually behaviour, and this could account for the evolutionary persistence of same-sex sexual behaviour in other individuals. This is I think what Mark @ 6 means by a spandrel.
I don’t mean to claim that any of this evidence is conclusive (in some ways the data are not very good because of limitations of the survey methods and questions), but this kind of pleiotropy is one good way to account for the evolutionary persistence of a trait that has an obvious fitness cost for individuals with the trait.
Absolutely, well said. You’re right to highlight that describing SSB as non-adaptive doesn’t mean it’s evolutionarily inert. The distinction is critical: non-adaptive traits can still persist due to indirect selective pressures or linkage to adaptive traits, and the pleiotropy described in Zietsch et al. (2021) is a compelling illustration of this. The idea that alleles associated with SSB could boost reproductive success in heterosexual individuals offers a plausible resolution to the apparent evolutionary paradox.
Also, the point about species differences is important. While genetic architecture may differ across species, that doesn’t undermine the validity or relevance of studying SSB in humans—especially since the focus is precisely on human-specific patterns of behaviour and evolutionary dynamics.
And yes, data limitations are real, but that’s par for the course in large-scale human behavioural genetics. The fact that we can identify meaningful patterns despite these constraints suggests there’s something genuinely interesting going on.
I think the concept of divvying up bits of the universe into the concepts of natural, unnatural and I suppose supernatural is a little bit fallacious.
This seems contradictory. If such behaviors are not for us to judge, then how did they legitimately become legal or illegal in the first place?
Are there any sexual behaviours between two consenting adults that are illegal these days? Asking for a friend.
Is eating a penis a sexual act?
https://www.theguardian.com/world/2003/dec/04/germany.lukeharding
During much of the gay and lesbian movement, the Naturalistic Fallacy was, well, -irrelevant to the anti-gay animus. One (of many) arguments used to justify laws and social policies against gay people was the idea that homosexuality was unnatural — that it didn’t occur in nature.
Sure, you could argue on philosophical grounds… or you could — and most of us did — point out that homosexuality was pretty damn common. It wasn’t an aberration. It was a “normal” thing that happened in the world, including with humans. It was a pushback against that argument as well as the numerous religious ones in play.
I know Jerry is right on the philosophy. In practical terms, the naturalness of homosexuality was key for both self acceptance and in larger society.
I would point out that the theory “if animals do it it’s natural and ok” would certainly imply that bestiality, at least at the level of sexual play rather than copulation, is quite ok. Animals can certainly get sexually frisky across species. I myself have been victimized and sexually molested by horny dogs who tried to hump my leg. I could name names. It’s okay dog friends, I forgive you, rest in peace.
I recall a common argument against gay and lesbian rights was that if society normalises this then bestiality could be next (shock, horror, clutching at pearls). But, seriously, other than the yuck factor and possible animal-rights concerns, what would be socially objectionable about it? (Of course we don’t want another SIV -> HIV, but that seems an unlikely prospect in the “developed world”.)
The difference, I think, is that an animal seems like more than just a “thing”. I do not know how much of a glimmer of consciousness a sheep has, but I see a difference between copulating with a sheep than with a synthetic sex doll. Certainty the animal cannot meaningfully give consent. The sex doll doesn’t need to since it’s a mere thing.
I can’t make a great theory that the animal is harmed. Domestic sheep, as opposed to wild sheep, are dumb animals. I don’t believe a sheep is psychologically complicated enough to emotionally suffer from the event. And I doubt there is physical discomfort since the size proportion of man and sheep is right. Yet it seems very wrong to me. I guess the feeling is that a sheep is a fellow living being that deserves dignity.
Yet I eat sheep (and goats and cattle and pigs and birds). If I was a hunter gatherer or a farmer who raised livestock, I would personally slaughter the animals I eat. I can’t rationalize that sex with an animal is worse than killing the animal.
The person who commits bestiality seems disordered to us. In a way, an obese person who consumes thousands of calories of junk food and then has half a cheesecake for desert seems disordered to me. I would try to be sympathetic, but surely the natural appetite for food has become unhealthy and disordered. The person who commits bestiality similarly has a natural appetite for sex that has become disordered and unhealthy.
Thanks for the thoughtful reply.
I once saw a video of a woman copulating with a horse. I found it difficult to watch but I did notice the horse was enthusiastic.
Is this relevant to the discussion?
There was prior discussion about bestiality and it’s impact on the animals. I shared an observation that I thought was both relevant and a bit humorous.
Sorry if I went too far.
If same-sex sexual behaviour has a genetic base, then some cases of transsexualism can be genetic. And that is why many people intuit, I think, that transsexualism and transracialism are not comparable.
But there is no such thing as transsexualism. No one can in fact change his sex. He can imagine himself to be the sex he is not. He can demand that everyone else must, under penalty of law, regard him as the sex he is not. (The sleight of hand here is to compel us to regard self-perceived and elastically imagined gender as more definitive than sex when we talk about sexual matters: “If you are a woman with a penis….”. “This mentorship program is open by competition to musicians who identify as women.” Etc.) But none of that weaponized kindness changes the truth. It can’t be meaningful to talk of a genetic basis for a condition that doesn’t exist.
It could be a mental disorder with a genetic base.
Individuals who believe they are the sex they aren’t might well be suffering from a mental disorder (which could, like anything else, have a partially genetic basis.) It seems obvious to us they have to be, but those affected resist the notion that they have any kind of disorder, other than their body doesn’t match who they really are and must be changed with medicine and surgery to look they way they believe it should. (And the rest of us must buy into the idea.) The ideological evolution has been to de-pathologize gender incongruence through successive iterations of the DSM until it becomes medically necessary body-identity affirmation in people who are mentally healthy because they say they are.
This just came into my Inbox this morning. I offer it for background information, not to prolong a tit-for-tat:
https://genspect.substack.com/p/attention-insurers-false-diagnostics
Homosexual behavior isn’t “evolved” and it isn’t an adaptation – it’s a spandrel (sensu Gould), an accident of the fact that sex feels good and animals have a strong impetus to satisfy sexual craving generated by hormones. Other animals don’t care and there are no repercussions for those animals that practiced it. Humans are just complicated and we hold grudges, judge and ostracise those who practice things that are not considered “normal” by the majority. In our society, we can assign costs to homosexuality, but it didn’t evolve “in our society” and may not have evolved at all. That is, who knows what Neanderthals, Denisovans, and even Homo erectus did, but if they were like other mammals, they sometimes had their fun with same-sex partners, and that may have had no cost at all. And, we all know the satisfaction felt after sex is transient, so maybe there were no benefits either (other than the momentary satisfaction of urges forced on them by their hormones). Finally, nobody had MORE offspring because of homosexual behaviors, so, unless kin selection was extremely strong, AND those that practiced homosexual behavior were very good care givers for their kin, there is simply no natural selection to speak off associated with it. I think that the urge to have sex was what was strongly selected by natural selection, and because of that, to paraphrase a song, if you can’t have sex with the one you want, have sex with the one you’re with (at least, to just relieve the urges).
+1
You speak with great certainty. You may well be right, but simply declaring that a behavior that does have a partial genetic basis is a spandrel evinces a degree of knowledgte that we certainly don’t have.
If i were writing this, I’d say that “I am speculating here”. After all, you don’t give any sources are are expressing your opinion. Others have also said it’s a pleiotopic effect of some other adaptation, but not one involving “sex feels good.”