In a recent blog post on the web site Science Based Medicine, Steven Novella asks the question “What does the science actually say about biological sex?” His motivations are made quite clear in the opening paragraphs where he wants to clear up “misconceptions” about sex that mean we do not know how “best to approach people who identify as transgender or non-binary”.
But rather than clear up misconceptions, Novella wades in neck deep into specious arguments about sex and only adds to the confusion. In this post I wish to show where Novella goes wrong and suggest some reasons why that might be.
There are a lot of muddles. Strap in. We will cover this in two parts. Here is Part I.
Novella sums up his argument in a paragraph,
Biological Sex Is Not Binary
The notion that sex is not strictly binary is not even scientifically controversial. Among experts it is a given, an unavoidable conclusion derived from actually understanding the biology of sex. It is more accurate to describe biological sex in humans as bimodal, but not strictly binary. Bimodal means that there are essentially two dimensions to the continuum of biological sex. In order for sex to be binary there would need to be two non-overlapping and unambiguous ends to that continuum, but there clearly isn’t. There is every conceivable type of overlap in the middle – hence bimodal, but not binary.
This is quite an extraordinary claim for the simple reason that not a single peer reviewed biology paper, written by a biologist, has ever claimed that sex is best described as “bimodal”. There may be papers that characterise sex differences in various features (the amount of dimorphism etc) as being bimodal, but not sex itself. How can Novella be so confident in saying that the “bimodality” of sex is uncontroversial among experts when not a single expert has ever said it in their primary literature? This needs explaining.
The case that sex is categorical and a dichotomy of male and female
Novella starts by accepting that sex has two “poles” – male and female. The twoness of sex exists at least at some level for him.
What does the biology actually tell us? What case can we make for a “binary”? I will spend a little time on this as it is crucial to understand.
The first fact we must acknowledge is that sex is an evolved mechanism of reproduction. Everything must be grounded in this reproductive reality. That Novella talks so little about reproduction in an article that sets out to explain “the science of biological sex” should be a big red flag. Indeed, Novella tries to brush this foundation of reproduction away as “incredibly reductionist”, although he does not say why being reductive at this point is a bad thing, and chooses instead to retreat into an equivocation of sex also meaning “the act of sexual intercourse”.
So, sex is about reproduction. But not all life reproduces sexually. Some simple organisms simply split in two, or bud off new copies of themselves, taking a new copy of their chromosomes with them. At some point in evolution, perhaps over a billion years ago, a strange thing happened where reproduction required two individuals to share genetic material to create a new organism. On the face of it, this looks half as efficient as simply budding off a new copy of yourself, and has been the subject of much research. We will not look at this here – we just need to accept that almost all complex life goes through a strange two-stage lifecycle.
The first stage is the somatic stage where an organism consists of many complex and differentiated cells containing two copies of each chromosome – diploid cells. You are in this lifecycle stage right now. Through the process of meiosis the next life stage occurs where single cells are produced that contain only one copy of each chromosome – haploid cells. Two of these haploid cells, from different parents (typically) then combine to form a new diploid cell that reproduces by ordinary division to create a new somatic stage – a new multi-celled organism.
This is the fundamental feature of sexual reproduction: this two stage process of cell replication with one stage requiring two combining haploid cells. The number two is already deeply locked in here with a discrete functional aspect. But this is not in itself enough to give rise to the concept of male and female.
Male and female, as a biological concept, arise from another rather bizarre evolutionary development. Organisms started reproducing only by the combination of two very different types of haploid cells – the two gamete types. One gamete was massive and immobile and the other gamete much smaller and mobile. This is called anisogamy. One gamete type is required of each type for reproduction. Again, why this happened has been a puzzle for biologists but appears to be an inevitable result if different organisms tried to gain reproductive advantage with different gamete sizes. So, biological systems associated with small gametes are called male, and systems associated with large gametes, female. The extreme divergence in gamete size is associated with extreme divergence in function too. There is no gradient of gamete sizes in a species – it is a discrete, categorical, and countable system: and the count is exactly two. There are no evolved intermediate gamete size producers, or no mixed-function gamete types.
The last fact we must take note of is that evolution has produced a number of systems to support male or female function. In some species, one individual can play the role of either male or female , either simultaneously (like worms and slugs) or sequentially (like the famous clownfish). These are called hermaphrodites. But almost all species do another strange thing and have evolved two types of somatic stage individuals – one type supports only sperm, the other eggs. This is called gonochoric in amimals (or dioecious in plants.) This is what we mean when we say an individual organism is male or female. A male is an organism that has developed along the evolved somatic pathway associated with sperm, and a female is an organism that has developed along the evolved somatic pathway associated with eggs.
Now, I am sure Novella would agree with much I have written here – all up to this last sentence where he appears to reject this clean dichotomy of discrete and categorical male and female organisms, where we appear to be able to objectively talk about the two classes of male and female individuals. So, we need to look at the many reasons he claims this is not possible.
What is being claimed by saying “sex is bimodal”?
Steven Novella rejects idea that sex is “binary” and claims that is it ‘not even controversial’ that sex is “bimodal”. In doing so, he is saying that we can characterise an organism’s sex, not by a discrete classification, but by some degree along a continuum of maleness or femaleness. There is in essence no such thing as 100% male or female, but all organisms are some sort of amalgam of features and function from both ideals. It is though quite difficult to understand quite what Novella means by “bimodal” as his explanation is, at best , somewhat vague.
The thing is that a bimodal distribution is not a vague notion, but a well defined type of statistical distribution that can be well characterised. That characterisation is lacking in Novella’s account, and instead gives a number of handwaving arguments in its favour (which we shall look at below). Novella does not cite any biology literature that explains how sex is bimodal but instead points to a diagram taken from the web site of a real-estate and crypto “entrepreneur”, Cade Hildreth. And indeed, Novella’s arguments appear to very much parallel the muddled discusion on Hildreth’s website. Hildreth also offers many more “scientific” insights into the health benefits of protein powders, vitamin and mineral supplements, and also claims to be the founder of the “world’s largest stem cell industry news site”. Novella’s web site also has a keen interest in these subjects like stem cell clinics, although on all matters bar the nature of sex, Novella’s and Hildreth’s views diverge somewhat.
But saying a measurement follows a bimodal distribution is a very precise claim. A bimodal distribution is a probability distribution of some measurement of a population that has two “modes” or most common values.
The claim sex is bimodal suggests we can make a measurement on an individual and use that to plot them along a distribution. The most basic question you can ask about a bimodal distribution is “what is the measurement you are taking that leads to this bimodal distribution”? We are not told this in Novella’s blog. At least, not one that defines “sex”. If you are going to claim “sex is bimodal” you need to say what measurement characterises sex. No-one ever has.
We might, for example, be an ecologist and take measurements of fruit size from a tree in a population of similar trees. Typically, we would see a “most probable value” for the weight of a fruit – the average (or mode) fruit weight. If we saw two modes, say at 80g and 40g, we would have to ask why? Statisticians tell us that a bimodal distribution is telling is that we are not looking at a single population, but two distinct populations. In these case, we might have two very closely related, but different species of that fruit tree – a “dwarf” variety and a large variety. If we were an ecologist catching field voles and weighing them, we might again see a bimodal distribution. Again, we might be seeing two different species of vole. Or that voles are significantly sexually dimorphic – males are heavier than females on average: two distinct populations. It does not mean the sex of voles is bimodal, just that the weight of voles is bimodal because there are two different sex populations. The sex of voles is still very much categorical. Weight does not define what sex you are – heavy voles are not necessarily male, just less likely to be male; weight is a consequence of having a discrete sex – males tend to be heavier. This type of conceptual muddle is at the heart of much of Novella’s thinking. We shall explore this more.
The muddling of sex and karyotype
The first “killer” argument put forward is a common one found on twitter threads and blogs. It claims that it is false to think there is a one-to-one link between sex chromosome configurations and the sexes. That is that XX ‘means’ female and XY ‘means” male. This is a straw man as no biologist thinks this.
The XX/XY chromosome pairs in humans determine sex. Inherit two X chromosomes and, notwithstanding genetic errors, you will develop as female. Obtain an X and a Y and you will develop along the male pathway.
There are nuances we need to cover, of course.
Nature has provided an abundance of sex determining mechanisms that lead to males and females. In birds, different sets of chromosomes determine male and female – the ZZ/ZW mechanism where ZW leads to egg laying female birds. Crocodiles use environmental temperature to switch between male and female development. There is no necessary link between XX/XY and males and females. But of all these mechanism, only males and females result, for the reasons discussed above.
There are also chromosome anomalies where, for example, a male might get an extra copy of the X chromosome. Their karyotype is XXY. They are still male though. Such anomalies, or aneuploidies, occur not just on the sex chromosomes. Most are fatal to the embryo. The most common one is perhaps on chromosome 21 and causes Down Syndrome. An extra copy of chromosome 21 results in significant development issues. Sex chromosome aneuploidies, like XXY (Kleinfelter) Syndrome, can also result in a range of development problems, like learning delays, language learning difficulties and motor development. But, again, each aneuploidy only results in either males or females depending on the type. With XXY Syndrome, some males may even still be fertile as males. Their sex is unambiguously male.
Such chromosomal anomalies do not result in a “spectrum of sexes” and they most definitely do not lead us to conclude sex is “bimodal”. Data from such individuals cannot be used to show any such statistical distribution of sex. People are male or female. A few also have a rare aneuploidy. This notion again is a fundamental conceptual muddle about what a sex is.
The muddling of sex and dimorphism and morphological variation
The next muddle is about the nature of sexual dimorphism. It is often claimed that because so many sex-based features show a bimodal-like distribution that this means sex itself is bimodal. As we saw with our field vole example above, it is possible to measure a bimodal distribution in some sex-linked measurement (like the weight of our voles) and still have distinct categories of sex.
Sexual dimorphism describes the degree of difference in morphology between the two sexes. Some species are highly dimorphic – with incredible differences between males and females (like angler fish). Other species show little external clues to differentiate between males and females. In birds we see the extreme plumage differences between peacocks and peahens. Yet in other species, the ornithologist has real trouble telling if they are looking at a male or female. The Griffon vulture is near indistinguishable, but there are subtle population differences in wings span, bill length etc that give small clues if you look closely enough. Nonetheless, both peafowl and vultures have distinct and categorical males and females, despite our practical ignorance in sexing the latter. The vultures appear to know – which is what matters, of course. A male vulture is just as male as a male peacock, even he chooses not to make much of a song and dance about it.
Humans are significantly, but moderately, sexually dimorphic. Amongst our ape family, we can see the huge differences between male and female gorillas and orang-utans, and near identical features of the gibbon. Humans appear to sit in the middle.
Perfect dimorphism is rare though in any given feature. There are very tall women (I have worked a lot in the Netherlands). There are males with small hands. But being a small handed male does not make you a lesser male on some sort of spectrum. A male is a male regardless of the size of your things. Morphological variation does not create a spectrum of sex and bimodal distributions of sex related traits does not make sex bimodal. The idea that you are a lesser female human for being more flat chested is as offensive as it sounds.
The dichotomy of sex is not equivalent to dimorphism in sex. These are two different concepts. Just because dimorphism may be low in humans, does not mean the sex dichotomy is weakened.
The muddling of sex and mutations and intersex
Moving on from variations in chromosome number, Novella tell us that genetic mutations can create “intersex” individuals. He starts by saying “XY individuals with extra copies of the WNT4 gene can develop atypical genitals and gonads, and a rudimentary uterus and Fallopian tubes” and points to a paper describing this, but neglects to say these individuals are mice.
But it is true that genetic mutations, gene duplications, transpositions and deletions can seriously affect sex development. Perhaps a well known example is that of the athlete Caster Semenya.
Semenya has a genetic condition known as 5α-Reductase 2 deficiency which is caused by a mutation on the gene SRD5A2. This mutation can occur in both males and females but only adversely affects males. Semenya is male and has a so-called “intersex” condition of 5α-Reductase 2 deficiency.
What happens to a male with this condition is essentially that their penis is under-developed in the womb. At birth, their genitals then have a superficially female appearance. But nonetheless, the penis, albeit very small, has the associated structures of a male organ. As the boy grows though, they undergo a pretty normal male puberty and their penis can indeed then develop in size. (The male puberty is what gives Semenya sporting advantage.) Some males with this condition may even be fertile. In apartheid township South Africa, a baby boy with this condition was unlikely to get a proper medical diagnosis and so was raised as if they were female. They may though have a ‘sex identity’ as female because they have been misinformed and probably misled in later life.
This condition is exceptionally rare and only a few hundred cases have been recorded. Novella though gives us the impression that “intersex” births are very common.
Novella cites a figure of 2% for “intersex” from a paper “How Sexually Dimorphic Are We? Review and Synthesis” by M Blackless et al, 2000.
This is a curious paper that appears to base its claims on some dubious and unscientific assumptions: that there exists a “platonic ideal” of male and female and that humans display “absolute sexual dimorphism”. These are both absurd straw-man arguments.
The lead author, Melanie Blackness, appears to have left no trace in the world, with just this one citation in the literature to her name, and no mention on the staff pages of the cited affiliated university. However, the corresponding author is much more well known: Anne Fausto Sterling.
Fausto Sterling left an early career in biology to join the “Gender Studies” crowd. She has then published various essays and books with eccentric claims like there are ‘five sexes‘. (A claim she later appear to withdrew telling us she wrote her work with her “tongue in her cheek”).
She is responsible though for introducing to the world the idea that about 2% of the population are “intersex”. Fausto-Sterling, more than anyone, is responsible for introducing the fashionable nonsense that sex is a continuum and “bimodal”. She published these ideas in book form (not peer reviewed) and they were quickly demolished as nonsense. Her (more accurate) 1.7% “intersex” figure could only be obtained by including large numbers of medical conditions where there is absolutely no ambiguity about sex – the patients are unambiguously male or female. The vast majority of that 1.7% figures comes from late-onset congenital adrenal hyperplasia which accounts for 1.5% of that total. Included also are the many aneuploidies like Kleinfelter Syndrome where, again, no ambiguity in sex exists. Even rare mutations like Semenya’s do not result in ambiguous sex, but just potential for not recognising the sex correctly at birth.
Fausto Sterling has politicised the term “intersex”. She transformed it from a colloquial and archaic term for various exceptionally rare development conditions into an “identity”. It has now become part of the identity youth movement where people co-opt the term as an expression of so-called “gender identity” without having any medical diagnosis. It has got to the point online where if someone describes themselves as “intersex” it is likely they will have no medical diagnosis or reason to doubt they are not straightforwardly male or female.
What do we mean by “ambiguity in sex”?
But there may be some births where genuine ambiguity, or difficulty in recognising sex, exists. So what is the nature of this ambiguity? When can we say that a person does indeed have ambiguity in their sex?
We have seen that when we say an organism has a sex, what we mean is that it has developed along one of the two evolved pathways that support one of the two gamete types. This allows us to classify in principle any individual in a gonochoric species (one sex per individual) as either male or female. We must observe which reproductive anatomy has developed: is it associated with sperm or eggs? We have also seen that there are wide classes of development conditions, mutations and chromosome issues that still result in entirely unambiguous sexes because they still show development along a single reproductive anatomy pathway.
We might have genuine ambiguity if there is a mixture of development of male and female organs and tissues in an individual. And Novella does indeed point out that such developments can occur. He says there are individuals with ovotestes that produce both eggs and sperm (but neglects to say the individuals in the paper he points to are prawns and lobsters – there is some real Jordan Peterson energy here .
But ovotestes is also a name given to some very rare medical conditions in humans where gonads may have mixed development. Let us explore that. Whilst we have now seen that saying “sex is bimodal” has no formal definition with no formal parameter to define “sex”, and is therefore nonsense, we might allow Novella to say the bimodal idea is more metaphorical in nature, and people born with ovotestes “fall in the middle”.
We must first note that such “mixed development” is exceptionally rare with ovotesticular disorder only having ever been document in 500 cases or so. We are a long way from Fausto Sterling’s vastly over inflated figure of 2% of the population.
Such mixed sex development is exceptional rare because evolution has ensured development has mechanisms to make sure this is so. A growing embryo will be wasting resources if it develops organs and tissues that cannot contribute to future reproduction. Novella’s paper on mice (above) is actually about a gene that appears to be involved in cross-sex development suppression. Put simply, our development of reproductive anatomy is absolutely not a pick ‘n’ mix of organs and tissues from male and female parts that might just result in enough of one sex parts to enable an individual to be fertile and reproduce. Instead, it is a tightly regulated cascade of genetic events along a pathway that puts all development effort into male or female development. That is why pretty much everyone ends up as unambiguously male or female even when significant development conditions occur. Male and female development are mutually antagonistic.
Very rarely, and for not well understood reasons, the brakes may come off this cross-sex development and tissue development that ought to be suppressed starts to grow. It is a bit like a cancer where the normal growth regulating mechanisms fail. And indeed ovotesticular disorder is associated with malignancies of these tissues, so are often surgically removed soon after diagnosis to prevent lethal cancers.
What is not observed is an individual who is fertile both as a male and female. If fertile at all, it will be as one sex. The cross-sex tissue is typically under-developed. No human is a true hermaphrodite (in the biological sense as being able to reproduce as both a male and female). Unfortunately, medicine also uses the term “true hermaphrodite” to describe people with these very rare disorders. Do not be fooled by this equivocation.
So despite this cross-sex development, can we still say what sex a person is? That is a complex question as we are dealing with disorders that are so rare and with so many different causes and outcomes that a blanket statement is not easy. Doctors publish individual case reports where it may be clear a person has undergone predominately one sex development and in which case we may be confident in calling someone male or female. It is a matter of debate if there exist individuals where sex development is so mixed that such a classification is inherently meaningless.
Summary of Part I
Steven Novella sets out with the explicit political intention of showing how people with trans identities fall into the middle of a “bimodal distribution of sex”. He claims this characterisation of sex is settled and non-controversial.
What we have seen is that biology understands how sex is a strict dichotomy of male and female based on anisogamy (two distinct gamete types). No peer reviewed biology paper has ever characterised sex as bimodal and shown how to create this statistical distribution from measured data of sex. At best the bimodal idea is a metaphor. At worst, it is handwaving nonsense. The idea has not come from biological science but from academics in “gender studies” with explicit political agendas.
We have seen how in order to support the bimodal idea, various specious arguments about sex are made. We see muddles about sex determination and karyotypes. We see conflations of sex and development disorders. We see muddling of the continuous and varied nature of dimorphism in species with the categorical nature of sex. We see how exceptionally rare ambiguities of sex development are used to justify the idea we cannot classify any person with any rigour or objectivity.
In Part II, we will look at how Novella ups a gear and introduces new muddles and conflations between sex and sexuality, sex and gendered expressions, how the controversy over brain dimorphism is exploited, and how incoherent ideas of “gender identity” muddy the waters. Finally, I will address why this massive muddle exists. What is going on where so many people are believing things that is just not expressed in the actual primary biology? How did Novella come to write such a tangled web of nonsense?