On inclusivity in Less Wrong: a response to Scott Alexander


Thanks to Ruthie Byers for inspiring many of the points in this post and providing awesome feedback on a draft.

There was recently something of a brouhaha kerfuffle1 in the Less Wrong community over inclusivity to minorities. A person known as Apophemi wrote a blog post about why ze2 didn’t participate in Less Wrong, noting that ze felt threatened by dispassionate discussion of certain issues:

I… require adherence to these ideas or at least a lack of explicit challenges to them on the part of anyone speaking to me before I can entertain their arguments in good faith.

This raised the hackles of some Less Wrong folk, including Scott Alexander, at whom the essay was apparently primarily directed. So Scott wrote an excellent and detailed response, explaining that safe spaces were great, but there needed to be safe spaces for dispassionate discussion too:

This reminds me of the idea of safe spaces.

Safe spaces are places where members of disadvantaged groups can go, usually protected against people in other groups who tend to trigger them, and discuss things relevant to that group free from ridicule or attack. I know there are many for women, some for gays, and I recently heard of a college opening one up for atheists. They seem like good ideas.

I interpret Apophemi’s proposal to say that the rationalist community should endeavor to be a safe space for women, minorities, and other disadvantaged groups.

One important feature of safe spaces is that they can’t always be safe for two groups at the same time. Jews are a discriminated-against minority who need a safe space. Muslims are a discriminated-against minority who need a safe space. But the safe space for Jews should be very far way from the safe space for Muslims, or else neither space is safe for anybody.

The rationalist community is a safe space for people who obsessively focus on reason and argument even when it is socially unacceptable to do so.

I don’t think it’s unfair to say that these people need a safe space. I can’t even count the number of times I’ve been called “a nerd” or “a dork” or “autistic” for saying something rational is too high to count. Just recently commenters on Marginal Revolution – not exactly known for being a haunt for intellect-hating jocks – found an old post of mine and called me among many other things “aspie”, “a pansy”, “retarded”, and an “omega” (a PUA term for a man who’s so socially inept he will never date anyone).

As always, Scott’s post was a pleasure to read, and I think his broad point is right: Less Wrong shouldn’t go too far in the direction of becoming the kind of safe space that Apophemi wants it to be. As several commenters on the LW discussion of Apophemi’s post pointed out, there’s a trade-off between inclusiveness and freedom of speech, and there are places for communities on both sides of the trade-off.

However, I’m afraid that Apophemi’s extreme position inspired a too-extreme rebuttal, and the whole debate is distracting from a broader and more valuable point. Namely, there’s a tradeoff between inclusivity and free speech, but Less Wrong is not at a Pareto optimum in tradeoff-space. We can almost certainly improve on the dimension of inclusivity without sacrificing free discourse.

That is, Scott is too hasty to dismiss the suggestion “Less Wrong should become more inclusive” because Apophemi’s post anchored the discussion toward ways of becoming inclusive that also radically reduce freedom of discussion. The result of that is a temptation is to dismiss Less Wrong’s problems with inclusivity entirely. Which seems to be what led to Scott comparing Less Wrong to a yoga class:

I have been to several yoga classes. The last one I attended consisted of about thirty women, plus me (this was in Ireland; I don’t know if American yoga has a different gender balance).

We propose two different explanations for this obviously significant result.

First, these yoga classes are somehow driving men away. Maybe they say mean things about men (maybe without intending it! we’re not saying they’re intentionally misandrist!) or they talk about issues in a way exclusionary to male viewpoints. The yoga class should invite some men’s rights activists in to lecture the participants on what they can do to make men feel comfortable, and maybe spend some of every class discussing issues that matter deeply to men, like Truckasaurus.

Second, men just don’t like yoga as much as women. One could propose a probably hilarious evolutionary genetic explanation for this (how about women being gatherers in the ancestral environment, so they needed lots of flexibility so they could bend down and pick small plants?) but much more likely is just that men and women are socialized differently in a bunch of subtle ways and the interests and values they end up with are more pro-yoga in women and more anti-yoga in men. In this case a yoga class might still benefit by making it super-clear that men are welcome and removing a couple of things that might make men uncomfortable, but short of completely re-ordering society there’s not much they can do to get equal gender balance and it shouldn’t be held against them that they don’t.

The second explanation seems much more plausible for my yoga class, and honestly it seems much more plausible for the rationalist community as well.

So… women aren’t turned away by the rampant evo-bio-determinism? The discussions of pickup artistry? The generally ugh in the comment threads? Because it’s totally there. People find Less Wrong exclusionary. I’ve asked them and they said so, and as far as being discouraged from contributing goes, that’s the end of the story.3

My sense is that the site is getting better. But the atmosphere at some of the meetups (where there’s no danger of being immortalized on the internet) is still pretty dire. At the point at which there’s enough demand for the Bay Area mailing list to start a women’s-only group, we have a problem, no matter how hard we try to argue and excuse it away.

Scott goes on to try to show that Less Wrong’s demographics aren’t actually evidence that it’s exclusionary at all:

Most rationalists come from the computer science community, which is something like 80% male. A few come from hard science fields like math and physics, both of which are 80 – 90% male. There is zero need to invoke “discourse without limitations” as an explanation for why the rationalist community is heavily male-dominated…

Also economics (30% female) and psychology (seventy percent female). Rationalists are about 10% female. And computer science already suffers from unfriendly discourse.

We’re not actually missing all those groups you mention as minorities who might be driven away. In fact, in many cases, we have far more of them than would be expected by chance. For example, we contain transgender people at about five times the rate in the general population (1.5% vs. 0.3%), and gays/lesbians/bisexuals at about three times the rate in the general population (15% vs. 4%). People who Jewish by descent are four times the national average (8% vs. 2%), and people with mental disorders are either around equal to the general population or much much higher, depending on how one interprets the data I did a terrible job collecting (sorry). We have more people with English as a second language than almost any other online community I know (the country with most rationalists per capita continues to be Finland) and members from Kenya, Pakistan, Egypt, and Indonesia.

If we’re comparing to the base rate for computer science above, why compare to the general population here? We’re significantly worse than computer science on people of Jewish descent; foreigners (30% in grad schools, e.g.) and hence probably ESL speakers also; and probably even Asian people, though I can’t find any statistics. You can cherry-pick your indicators and reference classes any way you like.

And even against the general population, we’re worse than average for Hispanic and Black Less Wrongers, which Scott omitted—this argument doesn’t really present a balanced case.

Worse, the hypothesis fails in the other direction as well. There are lots of groups that are horribly offensive towards minorites yet nevertheless manage to have very many of them. Across nearly every denomination, far more women than men go to church – if you go to a Catholic Mass, you will see pews full of ladies at levels the atheist community can only dream of. The atheist community is so feminist that there has been a serious movement to replace it with “Atheism Plus” that excludes all non-feminists; the Catholic Church is so regressive that it won’t let women become priests and thinks they were created as a “helpmeet” for man. And yet women, in aggregate, love the one and hate the other.

It seems likely that Scott is right here, but with this specific example there’s some creative reinterpretation going on.

My understanding is that atheism is famous for having gender issues. The splinter movement is because atheists were so bad at being inclusive that atheist feminists decided they needed to start over. So I think “so feminist that…” is something of a misleading phrase. Atheism Plus seems roughly analogous to, e.g., Catholic feminist theology, which totally exists (not to the level of “exclude all non-feminists!” but it’s there).

We could always see whether it might help to inviting some feminists in, listen to them without protest, and agree to do whatever they say…

I think this line is a good place to sum up my issues with this post.

Basically, the rebuttal of Apophemi is fine. I agree, discourse without limitations is an important part of Less Wrong and shouldn’t go away, fine. But that’s no excuse for overreaching and ignoring the middle ground. Less Wrong does have problems with discourse that marginalizes people. We need to at least notice that we’re making a trade-off here, and look for ways to improve along both axes when we can.


  1. Apparently “a brouhaha has the connotation of undue excitement,” which I didn’t realize when I wrote this. Thanks to Nathan Cook for the correction! 

  2. I follow Scott Alexander in using the gender-neutral for Apophemi, as ze complained about being misgendered but didn’t give a clear indication of zir gender. 

  3. Examples commonly cited include Eliezer’s post on gender, filled with biodeterminism and heavy-handed tit-for-tat egalitarianism and completely neglecting socialization (of course nothing is allowed to be socially determined, because social scientists are bad at statistics); almost all discussions of PUA; and a general low-grade grossness emanating from the comment threads of most relevant articles. 

2 commentscomment


Let me start by saying this was a good response and thank you for phrasing it so nicely and reasonably. I think I agree with you on the practical policy proposal which follows from all of this, which I am charitably interpreting as “We need to better enforce taboos against talking politics and gender on Less Wrong” (if in fact you are arguing that we need to enforce a taboo on one side of political and gender debates while permitting encouraging the other, then we have a big problem, but I didn’t get that vibe). I don’t know about you but I think the best way to enforce such a taboo is downvotes, and I continue to downvote anything I find political or gender-y, most recently everything on this thread and in fact the thread itself because people should have known it would end up like it did. If this is your policy as well we don’t disagree on anything practical.

But we do disagree on a lot of the theory behind it, and the rest of this comment will be an explanation of why.

You start by saying that there are lots of ways we can improve inclusivity without sacrificing discourse. Then you don’t really explain what those ways are.

All you really mention are Eliezer’s (I thought extremely reasonable and bend-over-backwards-polite) thread on gender issues, make a variety of vague insults (“heavy-handed”, “gross”) and some obvious exaggerations (“nothing is allowed to be socially determined because social scientists are bad at statistics”).

This last seems especially unfair, since my conception of the debate you cite is between people who want to acknowledge both genetic and social issues in about the ratio found by experiment (eg intelligence is between 50%-85% genetic, but 15%-50% some other factor which is unaffected by shared environment and so probably not parenting, poverty, or schooling; if this is a surprise to you I can cite papers) versus people who want everything to be social and refuse to acknowledge genetic issues at all. If someone said something like “Racial differences in intelligence are probably part genetic and part social”, would anyone have any hesitation in pushing them all the way to the “super racist conservative reactionary” side of the scale?

(see also: points 2 and 3 here)

“So… women aren’t turned away by the rampant evo-bio-determinism? The discussions of pickup artistry? The generally ugh in the comment threads? Because it’s totally there. People find Less Wrong exclusionary. I’ve asked them and they said so, and as far as being discouraged from contributing goes, that’s the end of the story.”

You act above like you are trying to chart some middle path between my position and Apophemi’s position, but it’s hard for me to see here what it could be. You think women are “turned away” by “rampant evo-bio determinism” - which, as I said above, seems to be anyone claiming anything beyond the standard “all differences between people are purely social, evidence be damned, anyone who disagrees is racist”. But if you don’t believe we should ban (or at least downvote to hell) any discussion of genetics, how exactly do you propose to attract groups who are repelled by evolutionary biology (which, when you take away the scare-word “rampant”, is pretty much all you’re saying)?

At this point I’m not sure whether we disagree about a question of social norms or a question of scientific fact. Is your claim that the average Less Wrong discussion attributes much more power to genetics and evolution than the evidence supports? Or are you making a claim that, regardless of the truth about genetics, there is something wrong with the way Less Wrongers discuss it? If so, what exactly is wrong?

As for pickup artistry, at this point I’m ready to write the whole argument off as some sort of slander. The amount of discussion of pickup artistry on LW is so rare that I can’t imagine a random woman stumbling across it, and almost always comes up in the context of “Does anyone have any advice on how I can become more attractive”, someone else saying something like “You need to be more confident, women dig confidence”, and then some third party noncentral fallacy-ing this to “You’re a pickup artist! Therefore you think rape is okay and you ‘neg’ women!”

Possibly this discussion might be more productive if you can give me some examples of the comments you think aren’t okay? If I’m allowed to put conditions on this request, I’d like to see three or four fairly recent discussions (so I know you’re not just cherry-picking the One Worst Discussion In LW History) and preferably ones not started by social justice people where their opponents are just trying to defend themselves.

“Also economics (30% female) and psychology (seventy percent female). Rationalists are about 10% female. If we’re comparing to the base rate for computer science above, why compare to the general population here? We’re significantly worse than computer science on people of Jewish descent; foreigners (30% in grad schools, e.g.) and hence probably ESL speakers also; and probably even Asian people, though I can’t find any statistics. You can cherry-pick your indicators and reference classes any way you like. And even against the general population, we’re worse than average for Hispanic and Black Less Wrongers, which Scott omitted—this argument doesn’t really present a balanced case.”

Let me start by defending myself. You find that 40% of prestigious computer scientists who get into the NAS are Jewish, which is very very different from finding that 40% of the computer science field as a whole is Jewish. Jews are disproportionately represented among high-achievers; for example, 2% of the US population are Jewish but 20% of Ivy League students; the very site you cite makes this point with Nobel Prize winners. You’re also looking at the older generation of computer scientists (winning prizes) versus the newer generation (being on LW) which in a field growing as fast as CS are probably very different.

So fine. Let’s stick to one set of goalposts and avoid cherry-picking. Less Wrong versus comp sci. We already agreed women are about equal. We’re doing much better in “foreigners” (which is a silly word to apply to a multinational site) since 46% of Less Wrongers are non-American and you say 30% of American comp sci grads are (why are we comparing only to American comp sci grads again?). Nonwhites were 21% of American comp sci majors and 16% of Less Wrongers. Blacks and Hispanics are 8% of CS Bachelors and 4% of CS Masters, and we’re right in the middle with 6%. I don’t see anything on bisexuals and transsexuals, but we’re almost certainly much better than CS as a whole.

On the other hand, we could make LW look like a multi-racial diversity utopia by comparing it to mathematics instead of CS. Or we could make it look like the KKK by comparing it to psychology. We could also vary its comparative success does by comparing to different levels of the field (entrants vs. bachelors vs. masters vs. Ph. Ds vs. prestigious award-winners).

So the better question might be why we’re splitting hairs with these numbers at all. You said it yourself: Economics is 70% male, psychology is 70% female. I’ll throw in some other figures: Library Sciences are 95% female, Family Sciences are 92% female, Fashion is 90% female, Occupational Therapy is 82% female, Education is 78% female. On the other hand, Philosophy is about 66% male, Computer Science is 85% male (exactly the same as LW), Aerospace Engineering is 87% male, Physics is 89% male, and Construction Management is 93% male.

Unless we want to posit that library scientists and psychologists are just horrible people having all sorts of offensive anti-male conversations, it seems pretty obvious that different fields attract men and women at different rates (this is the same thing I was saying about yoga). The sites I got the above statistics from suggest that women prefer majors with more positive social impact, men prefer majors that are more about science-for-science’s-sake.

Of course, we at Less Wrong wouldn’t know anything about science-for-science’s-sake.

So my question for you is - at what gender balance will you agree that we’re probably as gender-balanced as we’re reasonably going to get in a society where men and women have very different preferences, and that we no longer need to engage in self-criticism over our offensive opinions in order to get better gender balance? Is it computer science’s 85%, in which case we’re already there? Is it philosophy’s 66%, in which case we have a little further to go? Is it perfect 50-50 gender balance? Or is it psychology’s 70% women, in which case we could be gender-even and still have a long way to go?

This wouldn’t matter if your promise that we could work on being more inclusive without having to silence people held true. But I really feel like you haven’t backed that statement up at all, and instead started giving a list of who needs to be silenced (the “rampant evo bio determinists” et cetera). And if we do have to silence people in order to improve the gender gap, then continuing to silence people when the gender gap realistically isn’t going to get any better is driving people away for zero gain.

Let me put this into perspective. There is a very large number of people, some of whom you have probably met, who think utilitarianism is totally beyond the pale. They describe it with the same words you use to describe the LW topics you disapprove of - words like “full of ugh” and “gross” and kinda Nazi-ish. They talk about how utilitarians think it’s totally okay to discuss torturing a bunch of people or pushing fat men to their deaths and don’t you think women and minorities are understandably squicked out by this sort of totally unnecessary banter. They talk about how Peter Singer says infanticide is okay and yet utilitarians talk about him as if he is some kind of perfectly reasonable person whose ideas should be taken seriously.

Once we have made unwelcome all the “pickup artists”, but our gender ratio is still not 50-50, somebody will think to say “Don’t you think women are driven away by people pushing the philosophical ideas of a guy who literally said we should kill babies? Women, as a gender, seem to be noticeably pro-baby!” There will be arguments that people shouldn’t mention Peter Singer, and that utilitarianism should only be discussed if it absolutely has to be, and that all utilitarians should make sure to say very clearly that they also think the ethics of care are also equally valid, and people will complain about the fanatical white male utilitarians who never acknowledge there could be an alternate point of view and can’t understand how offensive their assertions are to everyone else.

And once we get rid of the utilitarians, and yet are still not 50% female and 12% black, we will get the people who tell us that atheism is the highest level of white privilege.

I predict that no matter what we do, the gender balance will never change and the suggestions to get rid of this or that to improve the gender balance will never end.

“And computer science already suffers from unfriendly discourse.”

I worry you’re swallowing a narrative uncritically here. How do we know that computer science has unfriendly discourse? Because we hear lots of stories about the unfriendly discourse in computer science, and we know that there are few women in computer science.

But consider an alternative narrative. In 1920, women weren’t allowed pretty much anywhere except maybe nursing and teaching. There were stereotypes that women would be terrible doctors, terrible lawyers, terrible businesspeople, terrible politicians, terrible mathematicians, terrible philosophers, and all these fields were at least moderately unfriendly to the first women to enter.

But enter they did, and now women are at, near, or above parity in medicine, law, business, politics, and philosophy. Yet for some reason, they didn’t get near parity in math and its later descendent computer science.

And so everyone said “Aha! Computer science must have lots of unfriendly stereotypes about women!” And then every single incident of someone making a joke about the word “dongle” was televised to the world, and it was agreed that obviously computer scientists are unfriendly to women, with ample wringing of the “creepy nerd” stereotype for all it’s worth.

We compare this to a field like medicine, which is super-toxic and abusive to everyone, where seniors have pretty much absolute power over younger doctors and the extent to which they abuse it is famous, and which has an extremely tight-knit and masculine culture of working super-long hours all the time and making fun of anyone who complains. And in which 47% of beginning med students are now women, because women are interested in the field and people will totally ignore the odd joke in a field they are interested in.

(The abuse suffered by Jackie Robinson when he entered baseball is legendary, but fifty years later African-Americans were overrepresented in baseball at almost twice their rate in the general population. Yet a climate of subtle unconscious sexism is supposed to make women suddenly rush away from computing in droves?)

If women hadn’t flocked to medicine, every incident of someone in medicine making a slightly sexist comment would have gone viral, and it would now be a known fact that medicine “suffers from unfriendly discourse”. Since women in fact flocked to medicine, it was never necessary to deploy that argument.

If you think that computer science is unfriendly to women, you need an explanation of why much more macho fields that are much more subjective and therefore have much stronger ability to discriminate against people they don’t like - medicine, politics, business, law, etc - didn’t develop cultures unfriendly toward women - yet quiet, soft-spoken, pure-abstract-objective-mathematics computer science did.

I’ve never heard such an explanation and it seems much more likely to me that culture-of-unfriendliness-toward-group driving-group-away narrative looms a lot larger in discourse than in reality. This seems broadly consonant with the new research suggesting stereotype threat doesn’t really happen in the real world to any significant degree.

“My understanding is that atheism is famous for having gender issues. The splinter movement is because atheists were so bad at being inclusive that atheist feminists decided they needed to start over.”

Your evidence that atheism is bad depends on sexual harassment accusations. But these are (I admit counterintuitively) probably evidence in support of feminist values since I expect a fixed level of sexual harassment in every community (probably proportional to gender imbalance) but they become public and famous only in the communities that take them seriously. Compare the amount of discussion of sexual harassment, sexism, and racism in the social justice community versus the Catholic Church. Anyone who is remotely involved in the social justice community gets accused of sexism, racism, bullying, and harassment for real or imagined offenses on a near-daily basis (Ozy can tell you stories about this, and I suspect Zinnia can as well). On the other hand, the Catholic Church is super sexist and molested children for decades with almost complete radio silence until a force exterior to the community (the mainstream media) forced them to have a (still minisucle) level of internal discussion. The vastly increased amount of harassment claims per capita in the feminist movement as opposed to the Catholic Church isn’t because the Catholic Church is better-behaved than feminists, it’s because feminists feel more comfortable bringing it up.

There’s also some of what I was talking about above going on, where atheism has few women but many feminists, and so there is pressure to publicize every incident that might support a sexism-based explanation for this.

But we don’t really need to get into that. The point is whatever atheism’s gender problem, it is less of one than the Catholic Church, yet people who participate in Catholicism are majority female. Other organizations that are majority female include several large pro-life organizations and a few that focus on fighting contraception.

So although I feel I should register my disagreement with you on the object-level question, my real answer is that I feel like the broader point that harassment of women doesn’t predict percent women in a group at all remains valid regardless of this.

“Basically, the rebuttal of Apophemi is fine. I agree, discourse without limitations is an important part of Less Wrong and shouldn’t go away, fine. But that’s no excuse for overreaching and ignoring the middle ground. Less Wrong does have problems with discourse that marginalizes people. We need to at least notice that we’re making a trade-off here, and look for ways to improve along both axes when we can.”

I’m totally in favor of Pareto-optimal trade, I just haven’t seen any proposal on how to do it.


This is just a silly linguistic point, but a brouhaha has the connotation of undue excitement. Using it basically says that you think it’s a storm in a teacup, when you plainly don’t. What actually happened is better characterized as a kerfuffle.

comment