Epistemic Status: Somewhat speculative, mostly descriptive.

What is gender?

That is an excellent question, but it seems to be a very hard one to answer well. Instead I’m going to ignore it. This post should work for most “reasonable” notions of gender.

Instead this is a post about how categorising people into genders affects how we conceptualise them, and how this leads to the creation of gender norms that we then enforce.

I’ll mostly be focusing on binary genders (male and female) in this post, but I’m not making any assumption that those are the only ones, only that they feed heavily into how we reason about gender.


Given a group, we tend to form inferences based on group membership. This is a perfectly reasonable thing to do – if someone is from France, we tend to assume they speak French. When someone votes for a particular party, we tend to assume they support many of that party’s polices (or at least reject other parties’ policies more).

Unfortunately what starts as a set of perfectly reasonable inferences often then plays out very badly in practice – reasonable inferences get exaggerated, and feed in to how we construct the social norms we enforce, often harming the people we stereotyped.

We do this in particular with genders. If a trait is particularly prominent in people we gender a particular way, we form stereotypes around it, and the trait itself becomes gendered.

For example, consider strength. It is simply true that men are typically stronger than women. That’s not to say that any given man is taller or stronger than any given woman (many men are short and/or weak), but looking at group averages the link between gender and height and strength is fairly clear.

We then reverse this stereotype. If it’s true that men are typically stronger, then it’s true that if someone is stronger then they are more likely to be a man. Thus strength becomes gendered – the trait becomes used as a marker of masculinity.

In and of itself this is a perfectly reasonable inference procedure. It’s literally true that if someone is stronger they are more likely to be a man. The problem is that we now erase the underlying data and simply treat strength as intrinsically manly, labelling strength as more masculine even once you have surpassed the typical strength of a man.

These social expectations then lead to enforcement. Men are shamed for being weak and women are shamed for having visible muscles because they look too manly. What started with a reasonable inference about differences between groups has turned into a social norm where everyone is forced to construct their gender to exaggerate the differences.

This enforcement in turn means that the group differences are larger than those we started with – if most people are expending effort to seem more masculine or feminine, the observed difference between them on that gendered trait will be larger than they would be in the absence of enforcement.

Thus we engage in a sort of “gender inflation”, where we take our initial notions of gender and expand them out into a kind of social halo around our original gender categorisation. This inflated gender manifests both in our social expectations and in the actual data we observe.

Small genderings become large

Because of this gender inflation, it is extremely normal to have gendering for traits which is more or less invented out of thin air, because a small gendering occurs which we then inflate it into a large one.

These small genderings can come up in all sorts of ways, but the easiest way is just chance. Culture is formed mostly out of memetic evolution (that is, people copy behaviours from others, and retain behaviours that in some sense work well), and as a result is highly contingent – often the reason why people behave in a particular way is the result of some random variation years back. There’s no intrinsic difference that leads to, say, the distinction between English and French, we just made different choices generations back which have been built on over time.

This contingency of culture can often lead to genderings because of some degree of homosociality – the tendency to prefer same-sex friendships (which sometimes may be strongly enforced by culture). The result is that there are opportunities for different contingent developments to occur between men and women, and that difference then becomes gendered, and gender inflation exaggerates those differences.

Genderings can also just be made up of course. There’s a long history of men theorising major gendered differences where none exist, and often that theorising is all that’s needed to create a runaway gender inflation where that difference becomes real solely because it is enforced. Because access to power is gendered, it is often easy to reshape gendering in ways that serve power.

What to do about it?

If gendering was purely descriptive and there were widespread acceptance that the posession of masculine or feminine traits didn’t necessarily imply much about other masculine or feminie traits, that would be one thing, but unfortunately it goes further than that in at least two ways:

  • People treat gender as predictive. If you have some gendered traits, you are expected to also have other gendered traits. This isn’t intrinsically incorrect, but leads to significant access problems where your gendered traits may open or close certain doors. I’ve e.g. written about this previously in the context of interviewing.
  • People enforce gendering. If you are perceived as a particular gender, you will be punished for not conforming to expectations of that gender. This actually doesn’t work well for anyone, because so many traits get gendered that even if we tick the right boxes on most of them it’s very unlikely we tick the right boxes on all of them. This is similar to some of the issues I talked about in On Not Quite Fitting.

As a result of these two factors, gendering tends to feed in to a lot of systems of control, where we reward people for gendering themselves “correctly” (by adhering to a consistent set of gendered traits) and punish them for mixed gendering.

Figuring out how to solve all of these issues is a rather big task, and I don’t propose to do that in this blog post.

I’ve previously described my utopian position as somewhat gender abolitionist. I no longer think that’s a good idea, because fundamenally regardless of whether we regard people as having genders, we will still regard many traits as gendered because of underlying biological differences, and I think most of the dynamics described above will continue to hold.

I think the current increasingly diverse range of non-binary genders we recognise is very helpful, both for letting people find the “points in gender space” that work for them, and allowing us to have a richer understanding of gender and gendered traits, but I don’t yet know what that richer understanding looks like.

My more modest short run suggestions are:

  • Gender inflation seems like a big deal, and I don’t think the extent of it is widely appreciated or understood. Be aware of its effects and try to damp them down rather than enforce it.
  • This feels like it shouldn’t need saying if you’ve read this far, but stop enforcing gendered traits. If someone exhibits a mix of masculine and feminine traits, that is a perfectly reasonable thing for them to do, regardless of whether that’s because they have a non-binary gender or are just breaking out of stereotypes within their binary gender.
  • In “Rewriting the Rules”, Meg-John Barker suggests that once you get to know someone as an individual you have much higher quality sources of information about their traits than relying on their gender as predictive. I strongly endorse this.
This entry was posted in Uncategorized on by .