Several weeks ago, Ozy of Things of Things posted about three different ways of thinking about psychological gender differences:
- There may be no psychological differences between men and women.
- There may be population-level differences between men and women, but overlapping ranges.
- The may be differences between men and women that are so large that the ranges do not overlap.
The problem I see is that two of these three ways are trivially wrong and I think hardly anyone believes them. There are quite a few measurable psychological differences between men and women, but most of them are small, and the ranges overlap even for the large ones. I suspect that few conservative Christians or radical feminists would dispute that, and that their disagreements lie elsewhere.
If we want to list three different ways of looking at gender differences, and have those ways reflect a real-world continuum of political beliefs, I would suggest the following:
- Gender is a social construct: Innate psychological differences between the sexes either don’t exist or are very small. The differences we see in the world as it is are socially constructed. I recognize that Ozy suggested that innateness be treated as a different axis of disagreement than the existence of gender differences, but innateness is so fundamental to real-world debates about gender essentialism that I think we have to put it front and center.
- Mainstream center-left: A handful of reasonably large, innate, population-level psychological differences exist between men and women, but these have little individual or political significance and certainly don’t justify defining distinct social rules for men and women based on their essential natures.
- Gender essentialism: Like Ozy, I have a hard time wrapping my head around this one – in part, because I pretty much don’t believe anything has an essential nature.
As a statistician, my tendency is to view things in terms of distributions and populations, which is not the way most socially conservative gender essentialists tend to talk about things. However, I do think the essentialist position can be articulated in those terms, in a way that’s more plausible than the way Ozy described it – with the disclaimer that I don’t actually believe what I’m about to describe.
Let’s look at the parenting example – specifically, “discipline”, “play”, and “challenging.” I doubt it’s the case that the gender-essentialist author thinks playful women, disciplinarian women, or challenging women are as rare as women with eleven toes. Let’s look at some simulated-but-maybe-plausible bell curves:
This trait shows a fairly large difference between men and women, but there’s still a considerable area of overlap in the ranges. But now let’s assume those three traits are independently distributed, and create a “fatherfulness” index that sums discipline, play, and challenge:
The above bell curves are based on the same distribution as the first set, but averaged across three independent variables. The range only barely overlaps. There are vanishingly few women whose fatherfulness exceeds that of even the least fatherful men. If fatherfulness is a coherent concept, then suddenly “essential differences” – non-overlapping ranges – seem much more plausible.
In order for this story to work, two assumptions have to hold true:
- The traits have to be independently distributed. If the three traits are correlated – if they’re all associated with high testosterone levels, for example – then they don’t add up to non-overlapping ranges nearly so quickly, because a woman with high testosterone is likely to have all three traits necessary to be fatherful. The curves for the average of three variables that aren’t independent, but are rather normally distributed around a fourth variable, are barely distinguishable from the curves for individual traits:
- There has to be something special about the combination of these traits in a single person. They have combine into a cohesive role for an individual that can’t be played by – for example – one unusually disciplinarian woman alongside one unusually challenging woman.
Neither of these things seems especially plausible to me, but they don’t seem as wildly implausible as believing playful women are as rare as eleven-fingered women. As for the specific question of whether two partners of the same sex can effectively raise a child, it seems to me like the evidence is already in: Children of gay parents turn out just fine. Also, parenting doesn’t seem to matter very much in general.
And it seems like almost any other social role would be less gender-differentiated than parenthood. Assume, for the moment, that being good at mental object rotation is an important skill for designing bridges. Mental object rotation is one of the very few areas where men have a moderately large average advantage over women on standardized tests. That would lead us to expect that there will be somewhat more male than female bridge designers, and perhaps the very top echelon of bridge designers will be highly male. But there are many caveats:
- Obviously it doesn’t mean we should exclude women from being bridge designers, because many women have mental object rotation skills better than most men.
- Designing bridges is probably a team effort, and it probably requires a wide variety of skills other than mental object rotation – quite possibly including some for which the average woman scores better than the average man.
- I can’t think of any plausible “supertraits” that would require a bridge designer to combine multiple “masculine” traits.
Are there any social roles for which the “supertrait” model seems plausible? I can think of one, I suppose: Killing strangers. It’s possible that being an effective hit man might require a combination of aggression, upper body strength, and sociopathy that is colossally more common in men than in women. Whether this has any important implications for the legal status of female hit men is an exercise left for the reader.