Children often think in terms of black and white. This encourages them to make certain mistakes, which the Internet then turns into memes.
As adults, we’re encouraged to move beyond this and think in terms of grey scales, which is fine. Lots of things aren’t dichotomous, and treating them as if they were can get you into trouble.1 But even though you can usually treat just about anything as a spectrum, there are times when thinking in terms of types, groups, classes, or clusters really does make sense.
I’ve wanted to discuss this is for a while—it leads to other important places. But this post is especially a response to Steersman, who likes to write about gender:
When you think "gender", think sexually dimorphic personalities, behaviours, roles and expressions; think two halves of a gender spectrum -- the masculine and feminine halves -- each of which comprises a myriad of entirely different masculine and feminine genders. Like the bluish and reddish halves of the colour spectrum.
I told him I didn’t think that way. But I didn’t have time to really explain what I do think, or why.
Would you rather be a Computer Consultant, or a Florist?
Let me go on record saying you can absolutely treat gender as a spectrum. That was Richard Lippa’s research paradigm from the 1990s, which I can explain most easily by asking you a question:
Would you rather be a Computer Consultant, or a Florist?
Think about it for a second. Answer the question with a number in your head; let 5 be “Computer consultant all the way,” let 1 be “Flowers forever,” and 3 be for if you don’t care.
Unsurprisingly, this question is an excellent way of determining your sex: In a survey I gave in university, the mean answer of males was 3.8, and females averaged 2.1.
This survey I made matched Richard Lippa’s method, which was to ask many, many of these questions about hobbies and job preferences. Doing this, he obtained a factor that described the degree to which any given item relates to gender. He called this factor gender diagnosticity (GD); it’s essentially just all the preferences a person have that, taken together, allow a psychologist to predict whether somebody is male or female. And he found that GD was independent of the Big Five personality traits:2
Preferences for various occupations, school subjects, everyday activities, and hobbies and amusements were rated by 119 male and 145 female Ss. Discriminant analyses were conducted to compute gender diagnostic probabilities… gender diagnosticity measures were independent of the Big Five…
I thought this was pretty interesting. Could there be a meaningful portion of gender interests that are unrelated to personality? So I did the thing I always did: I carried out a small survey (n ≈ 60) to check, but I used the HEXACO instead of the Big Five.
Like many findings in psychology, Lippa’s failed to replicate. I did find a very clear GD factor, but the feminine pole of Gender Diagnosticity was pretty well captured by Emotionality, Honesty-Humility, and Extraversion, and for me, that was that.
Gen Z Pro Tip: Can’t decide on your career? If you’re introverted or misanthropic, you might consider a traditionally masculine profession like engineering, carpentry, or programming, even if you’re female. If you’re extraverted or ethically-minded, try traditionally feminine vocations like counseling, human resources, or working with children, even if you’re male.
But even though gender isn’t independent of personality, it can still be described on a spectrum. I had a couple of guys in that survey who scored more than half a standard deviation toward the female side, and one lady who scored almost a full standard deviation toward the male side. These are people whose psychology matches more people of the opposite sex than their own.
So the question is not whether gender can be conceptualized as a spectrum. It can. Rather, I think we should ask whether it can really only be conceptualized as a spectrum, or whether it really ought to be conceptualized as a spectrum.
But the issue is both philosophical and highly politicized. So I want to circle around a bit and investigate similar topics first. What about something that isn’t really philosophical or politicized?
What about something like climate?
Do Climates Just Run on a Spectrum from Equator to Pole?
People talk about desert and tundra as distinct climate types. And visually, they do look pretty distinct:
But are they really different types? Maybe we should rather talk about them as just lying at different positions of a spectrum.
When I was young, I definitely thought of climate on a spectrum. I thought that, as you went further south toward the equator, things got hotter and dryer; going north, things got colder and wetter. This matched my experience driving in the car, and it made sense according to the way heat dried things out. Logically, then, I presumed the poles were wet and frozen, and the equator was abjectly dry.
This may have been eminently rational (yes this is another dig against rationalists). But it was totally wrong; the world’s biggest rainforest is found right at the equator in Brazil:
In fact, everywhere on Earth there’s an overall positive relationship between heat, rainfall, and proximity to the equator. Put another way, both annual temperature and precipitation are extremely high at the equator, and extremely low at the poles. But precipitation also sees local minima around 25° away from the equator, at the Horse Latitudes, where evaporation is extremely high. This combination of high evaporation with low precipitation results in a moisture deficit, creating deserts overland, and regions of high salinity in the ocean:3
So climates are not simply found on a spectrum from north to south. Rather, they form discrete types along the way. Yes, these types do blend into one another geographically, and they do shift from year to year, but they can’t be described in terms of a simple linear relationship with latitude, or even a quadratic one. To get a graph that looks anything like this at all requires a pair of exponential functions:
In defense of my younger self who understood none of this, I lived around latitude 35°, where the local precipitation function sloped upward with latitude. Look at the graph and you’ll see it made sense to think that further North = more rain. But the conclusion is still that
Reason is not a good substitute for information, and
Not everything just lies on a spectrum.
Are North & South America Just a Spectrum?
A related question is whether it makes sense to divide the Americas up into two continents. This isn’t (completely) trivial; some systems classify “America” as one big continent:
But if you treat North and South America as just one thing blended from north to south, you might have trouble remembering that:
They’re on separate tectonic plates, the North American and South American plate. This means that…
They had a separate and distinct past, where North America was part of Laurasia and South America was originally part of Gondwana, so that the vegetation of North America is closer to that of Eurasia and South America closer to that of Africa. If you go back around 135 million years ago, that’s the way they were. But also…
They have a separate future, where around 150 million years from now, they’ll break apart completely. But also…
Mankind is too impatient to wait millions of years for something to break. So we just went ahead and built the Panama Canal like nature had always intended. So now anyone who wants North and South America to be one continent has to explain how they are the kind of continent you can sail clean through.
So to me, it really looks like splitting North and South America into two continents makes more sense.
Are Millennials Just on a Spectrum from Gen X to Gen Z?
I thought this for a while. After all, technology has been steadily increasing, individualism has been steadily rising, anxiety has been steadily rising, mental health problems have been steadily rising, and the next thing you know everyone’s clinically depressed and addicted to their phones. Right? Millennials are more anxious, depressed, and addicted to devices than Gen X, but less anxious, depressed, and addicted to devices than Gen Z. No need for generations at all; just a spectrum from earlier to later born.
Jean Twenge says this is wrong—the generations aren’t just on a spectrum.4
Now I like Jean Twenge, but I don’t just roll over when people try to sell me on their personal perspectives. So she points out that anxiety, depression, and generally poor mental health have been rising over the past half-century, and this did seem legitimate—but couldn’t it just be explained by treating the generations as a blend from totally cool Gen X to totally sad Gen Z?
But as it turns out: No! Not even remotely! Twenge crushed that seemingly clever little suspicion with a 2021 meta analysis showing that Millennials are, well… special:5
That study also has other graphs for other samples, but they tell the same story: Narcissism rises over time for young adult Americans starting in 1994, when the first Millennials were entering adolescence, and then falls again 15-20 years later, when the first members of Gen Z were entering adolescence. American Millennials came of age in a Narcissism bubble, and that bubble will never pop, because for many of them, it isn’t a bubble—it’s Just The Way Things Are.
And there’s at least some reason to think there are a lot of other little ways Millennials look different from either Gen X or Gen Z. Although some of the stats I’m looking at here come from limited single-surveys, Millennials do show up as more trusting of big corporations, and more concerned about work-life balance, global warming and fair trade, than either Gen X or Gen Z. And it looks like Gen Z expects to work harder than Millennials did, valuing more face-to-face communication as well.
There’s always a certain risk of wandering off the deep end when describing some group of people as being a certain way. Americans are notoriously obsessed with guns; I’ve fired a gun on two occasions, and had a gun pointed at me on one. And granted, this may be a lot of gun-related life events compared to my readers across the world. But years go by without my seeing a gun.
Similarly, I’ve met American Millennials who don’t fit the mold. And really, I have no idea how much these generations are even really a thing outside the US. Still, at least in America, at least on average, Millennials are clearly their own group, not just a blend from Gen X to Z.
So with all of this established, I’m now ready to answer the question:
Is Gender Just on a Spectrum from Male to Female?
Not even remotely.
Gender is ultimately where culture and society make sense of sex. And those who live in between the two primary sexes definitely do not just live in between. In fact, the intersex are generally more different from women and men than women and men are different from each other. By itself, their sheer rarity makes it hard for intersex people to relate to the majority who do fit in one of the two primary genders:6
But more than this, it turns out that many of the intersex have one or more disabilities. A 2020 study of intersex people found:
9.7% were either deaf or had serious hearing difficulties,
8.7% were blind or had serious difficulties seeing,
22.8% had serious difficulty walking or climbing stairs,
8.3% had difficulty dressing or bathing,
30.9% had difficulty doing errands alone, and
56.6% had serious difficulty concentrating, remembering, or making decisions.7
These numbers say a lot on their own, but it’s hard to know how much of this is because of social stigma. One report used a clever methodology which compared intersex respondents to LGBTQ+ nonintersex respondents. I like this a lot, because it helps to compare intersex with non-intersex, controlling for attraction patterns. This report corroborated that intersex individuals have much worse physical health:
According to the same study, intersex people also have much worse overall mood:
Life for the genuinely intersex can be pretty rough, even compared to run-of-the-mill LGBTQ+. I’ve had a few friends and relatives who were L or G, and they were a bit unusual, but they played sports and had good financial prospects. I’ve also known half a dozen people with gender dysphoria (presumably biological girls and women who would have just been called tomboys once upon a time) and none of them were blind, deaf, or had serious difficulty walking or climbing stairs.
So the intersex are not just “on a spectrum.” Some of them are intersex due to a genetic anomaly with severe health consequences. Some are intersex due to anomalous hormone exposure in utero. Some are probably intersex because their parents tried to hormonally stabilize their bodies along the wrong path from birth. But however they got there, they’re in a place that most of us won’t experience or be able to fully empathize with, because most of us have a clear and easily assigned sex that we took for granted as long as we can remember.
Conclusion
We can meaningfully talk about spectra for climate, geography, generations, and genders. But considering these things in terms of spectra is not the most sophisticated way to look at them.
And those weren’t the only examples I could have given. Space doesn’t merely exist on a spectrum. In between vast areas of nearly zero density, matter clumps into planets, stars, black holes, and entire galaxies, like the Milky Way:
Species don’t merely exist on a spectrum. There’s a lot of variation among cats, and among dogs, but you may have noticed there are not so many dogcats out there. (Though mathematically inclined Millennials will know that there are surprising numbers of slugcats)
Even numbers don’t merely exist on a spectrum. Count integers from 49 to 55 and all of a sudden, 53 is prime. Feeling intrepid enough to brave the ocean of real numbers? OK, sail from √5 to π, and amidst that controversial, infinite sea, all of a sudden you’ll hit dry land at 3, an integer.
And that number, three, is where we are with gender. We can smooth away the people who don’t fit by pretending that two genders are enough. We can magnify their significance by trying to talk about dozens of genders. Or we even can say “It’s all on a spectrum.” But none of those are the best way to think about it.
That’s basically why the MBTI fell out of favor, because it erroneously treated dimensions of smooth, bell-curve personality variation as dichotomous types.
Lippa, R. (1991). Some psychometric characteristics of gender diagnosticity measures: reliability, validity, consistency across domains, and relationship to the big five. Journal of Personality and Social Psychology, 61(6), 1000.
Schmitt, R. W. (2008). Salinity and the global water cycle. Oceanography, 21(1), 12-19.
Twenge, J. M. (2023). Generations: The Real Differences Between Gen Z, Millennials, Gen X, Boomers, and Silents—and What They Mean for America's Future. Simon and Schuster.
Twenge, J. M., Konrath, S. H., Cooper, A. B., Foster, J. D., Campbell, W. K., & McAllister, C. (2021). Egos deflating with the Great Recession: A cross-temporal meta-analysis and within-campus analysis of the Narcissistic Personality Inventory, 1982–2016. Personality and Individual Differences, 179, 110947.
Blackless, M., Charuvastra, A., Derryck, A., Fausto‐Sterling, A., Lauzanne, K., & Lee, E. (2000). How sexually dimorphic are we? Review and synthesis. American Journal of Human Biology: The Official Journal of the Human Biology Association, 12(2), 151-166.
Rosenwohl-Mack A, Tamar-Mattis S, Baratz AB, Dalke KB, Ittelson A, Zieselman K, et al. (2020) A national study on the physical and mental health of intersex adults in the U.S. PLoS ONE 15(10): e0240088. https://doi.org/10.1371/journal.pone.0240088
Thanks for this response to my earlier comments on your “How many genders” post. And my apologies for the delay in getting back to you on this response of yours. The whole topic is a rather complex issue and I have been somewhat restricted to a smartphone which really wasn’t sufficient to deal with it.
However, while I think you make a number of good points – which I’ll get into in a bit and which may justify something in the way of a compromise position – I still think you’re conflating sex and gender. Seems to me that either you’re being careless or sloppy, or you’re doing so intentionally, or you’re engaging in “motivated reasoning”. For instance, your “conclusion” section says:
AP: “Is Gender Just on a Spectrum from Male to Female?
Not even remotely.”
And your concluding paragraph here says, “And that number, three, is where we are with gender.” Which links to your earlier post which says, “The number of genders is three: Male, Female, and Other.”
But saying “three genders” is what makes it a spectrum: two end points and something in between, however you may want to order them. For example, see the Google/OxfordLanguages definition:
“spectrum (noun): used to classify something, or suggest that it can be classified, in terms of its position on a scale between two extreme or opposite points.”
Nothing in between the end points is a binary; some thing(s) there is a spectrum.
So you may want to do what Hippiesq suggested in the comments here: use “gender” “as a description of femininity versus masculinity in personalities”. And you may also wish to deep six the use of “male” and “female” as genders, and use “masculine” and “feminine” as both she and the late great US Justice Scalia suggest:
Scalia: “The word 'gender' has acquired the new and useful connotation of cultural or attitudinal characteristics (as opposed to physical characteristics) distinctive to the sexes. That is to say, gender is to sex as feminine is to female and masculine is to male.”
https://tile.loc.gov/storage-services/service/ll/usrep/usrep511/usrep511127/usrep511127.pdf
In addition to which, you might also want to pay close attention to what philosopher Will Durant has to say about Voltaire’s, “if you wish to converse with me then define your terms”:
https://quotefancy.com/quote/3001527/Will-Durant-If-you-wish-to-converse-with-me-said-Voltaire-define-your-terms-How-many-a
Although I don’t see that you picked up on that idea when I’d broached it in your “how many genders” post.
In any case, we might start on that course with a definition for “sex”. While I’m glad to see that you more or less endorse “Either of the two divisions, designated female and male, by which most organisms are classified on the basis of their reproductive organs and functions” for “sex”, those really are not the standard biological definitions promulgated in reputable biological journals, encyclopedias, and dictionaries. See:
"Female: Biologically, the female sex is defined as the adult phenotype that produces [present tense indefinite] the larger gametes in anisogamous systems.
Male: Biologically, the male sex is defined as the adult phenotype that produces [present tense indefinite] the smaller gametes in anisogamous systems."
"Gamete competition, gamete limitation, and the evolution of the two sexes" https://academic.oup.com/molehr/article/20/12/1161/1062990 (see the Glossary)
https://web.archive.org/web/20181020204521/https://en.oxforddictionaries.com/definition/female
https://web.archive.org/web/20190608135422/https://en.oxforddictionaries.com/definition/male
https://twitter.com/pwkilleen/status/1039879009407037441 (Oxford Dictionary of Biology)
In which case we say, biologists say, that to have a sex is to have functional gonads of either of two types, those with neither being sexless. Which knocks into a cocked hat your assertion, and that of many others, that less than 2% of us are “biologically indeterminate”. I guestimate that some third of us sexless at any one time, primarily the prepubescent.
But a big part of the problem – your “protracted, angry dialogue about how the binning is supposed to work” – is largely due to the fact that many if not most people are too emotionally attached to, if not psychologically and intellectually crippled by, the quite risibly unscientific “idea” that everyone has to have a sex.
But that then brings us around to the question of defining gender. In which case, I think that even you would agree -- as you've illustrated -- that there are many traits that are sexually dimorphic, that show significant degrees of correlation with our sexes, but more with one sex than the other. Which many people – including Hippiesq, Scalia, WHO, the BMJ, and, arguably, the progenitor of the concept (Stoller) – all accept, more or less, as different genders:
https://www.who.int/health-topics/gender#tab=tab_1
https://www.bmj.com/content/372/bmj.n735
https://www.perlego.com/book/1505189/sex-and-gender-the-development-of-masculinity-and-femininity-pdf
However, I think you have a valid point in arguing or suggesting – even if imperfectly, more on which later – that there are many traits or trait values that no male and no female exhibits: the correlation between those traits and “male” or “female” is zero. In which case one might endorse your view, at least a modification of it, that there are three genders: masculine, feminine, and “genderless”. Even if that latter is something of a contradiction in terms – something can’t be both A and Not-A.
But for example and to address that “imperfectly”, I think that that was probably what you were getting at with your “Bimodal Continuum” graph, your “joint probability distribution”. However, you show an overlap which means that there are some trait values – the spectrum or range of values on the X axis – that are exhibited or manifested by BOTH male AND female SEXES. I think what you want to say is that, at least for some traits and trait values, the intersex will have some trait values that no males (sex) and no females (sex) exhibit or possess.
More particularly and for a detailed example of that latter case, IF all males (sex) had heights between 60 and 66 inches (5’ and 5’ 6”), and IF all females (sex) had heights between 48 and 54 inches (4’ and 4’ 6”) and IF all sexless people (including the intersex) had heights between 54.1 and 59.9 inches (4’ 6.1” and 4’ 11.9”) THEN there are some trait values that no male and no female exhibits. Ergo that height range is neither a masculine height (60 to 66 inches) nor a feminine height (48 to 54 inches).
Though not quite sure, to elaborate on that idea of a contradiction in terms, whether it’s more reasonable to say that there are three genders (each a sub-spectrum) – i.e., masculine, feminine, and neither. Or whether it is better to say – for example, as you’re apparently suggesting, at least in your more credible or intended arguments – that there are those who have a masculine gender on some dimensions (traits) of a multidimensional gender spectrum, and who have a feminine gender on other dimensions (traits) of that multidimensional gender spectrum, and who might be genderless on other dimensions (traits).
However, part of the problem, as the hypothetical heights example suggests, is that many don’t realize that there is a great deal of variation in the overlap between those joint probability distributions. Many – like intelligence – show next to none whereas many others are substantially greater to the point of being no overlap at all. As I say, bit of a murky concept but it helps to be clear on the mathematics involved, particularly the statistics. And likewise with defining one’s terms – kind of like anchoring oneself to solid ground before going down any rabbit holes. Largely Voltaire’s point and Durant’s elaborations thereon.
This is interesting, but I have a question. Are you discussing "gender" as a synonym for sex, which it used to be (back in the day, a survey or even a document you might fill out at the doctor's office might use the word "gender" with the choice of male or female), or as a description of femininity versus masculinity in personalities, which seems to be the way "gender" is used more often today? Assuming it's the former, this essay makes sense to me. I would agree that sex is not on a spectrum, but rather there are males, females, and a rare group of people with dsds that make it rather impossible to determine whether they are male or female so one could say they are neither or both, or give them the right to choose their sex. (The vast majority of people with dsds are clearly male or female, but, according to the scientific articles I've seen, about .02% of people have dsds that render their sex unclear). There is no spectrum, but rather different types of males (more or less effeminate or masculine), different types of females (more or less effeminate or masculine), and a much smaller group of those who cannot readily be said to be male or female.