My recent post on Richard Dawkins versus the science of sex and gender sparked a conversation in the comments, so I wanted to explore the subject in more detail.

For purposes of this post, let’s say that sex consists of the biological traits that relate to reproduction and childbearing, and that gender is the social roles we stack on top of that: things like what clothes you wear, what emotions you’re expected to express, what roles you’re expected to perform at home, in the workplace and in society in general.

Many people, including Dawkins, believe sex and gender are a straightforward binary. You’re either male or female, a man or a woman, end of story.

But if that were the case, it should be easy to come up with a rule that tells you which sex a person is. However, that turns out to be not nearly so simple. The more you look, the more you find that any such rule is fraught with complications, exceptions and judgment calls. No matter the criteria, there are cases that don’t fit neatly on either side of the line.

Genetics. The first place to look for an A-or-B rule is the genes. If you have XY chromosomes, you’re male; if you have XX, you’re female. Nothing could be simpler than that. Except it’s not so simple.

The “master” gene on the Y chromosome is called SRY. When present, it switches the fetus to the male development path – usually. But not always. Mutations in SRY can result in Swyer syndrome, a person who has XY chromosomes but a biologically female body.

People with Swyer syndrome are usually infertile, but not always. In one remarkable case, a woman with XY chromosomes got pregnant and gave birth… to a daughter who was also XY.

Even a functional SRY gene doesn’t guarantee a male body. A different group of mutations result in androgen insensitivity syndrome, in which cells fail to respond to the signal of SRY. A person with this condition always has testes (but they may be internal, so this may not be obvious), but in other respects, their bodies can be intersex or physiologically female.

It can happen the other way, too. The X and Y sex chromosomes are usually exempt from swapping of genetic material during meiosis, but not always. On occasion, SRY moves from the Y chromosome to the X. This results in XX male syndrome: a person with XX chromosomes and a biologically male body.

Thus, knowing a person’s chromosomes doesn’t necessarily tell you what sex they are. What else can we try?

Gametes. Another frequently heard suggestion is to determine sex on the basis of gametes. This has the advantage of having, seemingly, only two options. If your body produces eggs, you’re female; if sperm, you’re male. Nothing could be simpler than that!

However, this definition has some flaws. To name the most obvious, what sex are you if your body produces neither?

People with Klinefelter syndrome (XXY) have biologically male bodies, but smaller testicles that often produce no sperm. Are they male or not?

The female equivalent is Turner syndrome (XO). Like Klinefelter syndrome, people with this condition have biologically female bodies, but often lack functional ovaries. People with Swyer syndrome, mentioned above, also usually have nonfunctional “streak gonads” that don’t produce gametes.

That’s not even to mention people who’ve had testes or ovaries surgically removed. Using this as the sole definition would suggest that, if a person’s body is incapable of producing gametes, there’s no way to tell what sex they are. Obviously, this is absurd.

Anatomy. If genetics and gametes don’t yield a bright-line rule, the next place to look is a person’s genitals. A male has a penis and testicles; a woman has a clitoris, uterus and ovaries. Nothing could be simpler than that… except, again, it’s not so simple.

Male and female genitals develop from the same primordial structures in the fetus. Due to hormonal irregularities, some people are born with atypical genitalia that aren’t exactly one or the other. They may have genitals that resemble either a small penis or a large clitoris, or a partially fused labia similar to a scrotum. What should the rule be for people with these intersex conditions?

In days past, doctors often took it upon themselves to “fix” this “problem” with plastic surgery, assigning the infant to one sex or another before they were old enough to voice an opinion on the matter. Many of those people grew up to resent what had been done to them without their consent, and it’s now widely considered a human-rights violation.

Even beyond these cases, it’s easy to see why sex shouldn’t be defined solely on the basis of genitals. If a man lost his penis and testicles in a traumatic accident (say, a soldier who stepped on a land mine)… or if a woman had a full hysterectomy (say, to treat uterine cancer)… would they cease to be their former sex because they no longer had the equipment?

There are real-life stories that show it doesn’t work that way. One is the infamous case of David Reimer, who suffered a botched circumcision as an infant. On the advice of a psychologist, he was given sex reassignment surgery and raised as a girl. But he never accepted it, and his life ended tragically because of it.

An even more fascinating case is the Guevedoces: a community in the Dominican Republic where some children are born appearing female, but develop a penis and testicles at the onset of puberty. (How is this possible? Read the linked article for details.)

Hormones. Another popular proposal is that sex is determined by hormones. Men have higher levels of testosterone, while women have higher levels of estrogen.

Recently, this has been the preferred solution for professional sporting bodies. Some have ruled that women with naturally high testosterone levels wouldn’t be allowed to compete unless they take drugs to reduce them.

Whatever you think about the fairness of this rule in elite athletics, it would be infeasible for the general population. As with the other traits, hormone levels fall along a spectrum of variation. One study found that 16.5% of men had testosterone below the normal male reference range, while 13.7% of women had testosterone above the normal female reference range.

As men age, their testosterone levels naturally decline. Other medical conditions, like pituitary gland problems, can also cause low testosterone, as do conditions like Klinefelter syndrome. If a man’s testosterone level falls below the clinical standard, does he cease to be a man and become a woman?

Secondary sex characteristics. Unless you lead an unusual lifestyle, you probably don’t know what genes, genitals or hormone levels your friends have. Instead, we judge by people’s outward appearance, especially musculature, breasts, voice pitch, and facial and body hair.

However, these traits are even more clearly a spectrum. Some women are stronger than some men: for example, I’m never going to lift as much weight as Mary Theisen-Lappen, but I don’t think that makes her a man or me a woman.

Many people are androgynous, not easy to classify at a glance. Some women have hirsutism (excess facial and body hair) while some men have gynecomastia (enlarged breasts). We may consider these conditions unusual, but we don’t believe they make a person a sex other than the one they identify as.

When these conditions and others are taken into account, it’s virtually impossible to come up with an unambiguous rule that defines what sex someone is. You either have to resort to “I know it when I see it” vagueness, or write a rule that classifies some people as what seems clearly the wrong sex.

You might say that intersex conditions are rare anomalies, so we shouldn’t allow them to overturn an otherwise useful rule. It’s true that they’re rare, but that doesn’t make them irrelevant. No good scientist would say, “There are some exceptions my theory can’t explain, but those probably aren’t important, so I’m just going to ignore them.”

On the contrary: scientists know that anomalies are valuable, precisely because they show the incompleteness of our current models and point the way to a better understanding. Transitional fossils are rare, but that doesn’t mean they should be disregarded. Rare or not, they show that evolution is true and the presumed discontinuity of species is false. Just the same way, intersex conditions show that sex is more a spectrum than a binary.

I certainly don’t consider myself an expert on this topic. I’ve been learning a lot about it, especially in the last few years. My views may change further, but here’s where I’m at now: I believe that sex and gender are constellations.

What do I mean by that?

A constellation is a group of stars that form an image in the sky. In one sense, constellations are real: the individual stars that make them up are obviously real, and astronomers can agree on which stars belong to which constellations. They’re stable patterns whose boundaries are widely agreed upon and which haven’t changed over the millennia.

In another sense, constellations are arbitrary. They’re artifacts of our imagination; the patterns don’t have an objective existence of their own. We could divide the night sky up into different constellations, and it would work just as well. What’s more, the stars in a constellation are at different distances from Earth and from each other, and they’re all in motion over cosmological time scales. It’s only our vantage point in space and time that makes them appear to go together.

Sex and gender are the same category of thing. The “stars” are the facts on the ground – the biological traits a person either has or lacks and the cultural beliefs and roles a person either accepts or rejects. The “constellations” are the way we group them together, deciding what belongs with what.

But constellations are cultural constructs. There’s nothing sacrosanct about them. If we choose, we can group them in a different way – or we can just accept that our classifications don’t map onto any fundamental division of reality. Rather than insisting that everyone is either 100% male or 100% female, we can accept that some people have some traits that point one way and other traits that point another way. To argue otherwise is the same as demanding to know which constellation a star “really” belongs to.

The sky is full of stars, each one unique and beautiful. We don’t need each and every one to fall into a set of arbitrary boxes for us to appreciate them.