If you don't follow "nature", then the definition is kind of arbitrary.
What could "arbitrary" mean here? Paraconsistent logic is not "arbitrary," though it is hard to say in what sense it follows nature. As it happens, no human being employs a language that has been completely purified of all interests, all values, all pragmatic considerations, of everything but the Truth. But from this it does not follow that all the non-joint-carving terms in human language are completely arbitrary; they may even be universalizable, if the prudential or moral values they are predicated on happen to be shared among the relevant linguistic community.
If you complain about "gluing a harmful value to a virtuous one", I feel like you have already decided to dislike A and like B, and you are biased to think about definitions that will hurt A and/or help B.
Guilty as charged, I suppose. I do indeed dislike denialism and male supremacism, and I do indeed like supporters of men's and women's rights. Is this an unacceptable leap? Does intellectual seriousness demand that I maintain perfect neutrality at all times regarding the existence or moral character of systemic sexism? Absent an argument for taking denialism seriously as a factual claim, or for conferring respect upon supremacism as a scalable moral project, I see no reason to even consider actively linking these practices to productive social activism, any more than I see a reason to coin a catchy new term for 'environmentalists who deny the occurrence of anthropogenic climate change,' or 'white supremacists who regularly give to charity.' Certainly there are such people, but we have no responsibility to rhetorically fortify their position for them by gerrymandering a more respectable slice of peoplespace in their honor. That goes well beyond steel-manning.
Remember, I did not suggest inventing a term merely to promote gender egalitarianism or human well-being or what-have-you. All I noted was that the values in question are potentially hindered if we go out of our way to coin a new term linking denialism or supremacism to the general idea of the promotion of men's (or women's) liberties, rights, welfare, etc. As it happens, this also isn't a natural kind, isn't one of Nature's privileged Joints; but I thought that point was relatively obvious, so I stuck to the pragmatic question of the utility of the coinage.
My translation: "In my opinion, C pattern-matches P5."
It was already suggested earlier in the discussion that P5 is a kind of C (i.e., that 'MRA' is a specifically anti-feminist movement). My addition was to suggest that insofar as that's the case, it's clearer to regularly speak of 'C' in place of 'P5.' 'MRA' is already being used as a term of abuse; my addition just lets us note a more natural grouping of the intended targets of the abuse, while conferring the advantage of not giving up the meme of men's rights to the Dark Side, and the secondary advantage of allowing people who happen to identify as feminist 'Men's Rights Advocates' to clarify that they belong to a special P5* that isn't really C at all. So much the better.
My translation: "You could pick out other member of C, such as P1, P2, P3, P4, but this may not actually be a useful [for what purpose exactly?] category, since it glues P5 to P1".
Again, you aren't applying your own analogy schema to the case at hand. I already allowed P5 into the Big Feminist Tent alongside P1-4. I can't* allow anti-feminism into the feminism tent, on pain of logical contradiction. And I see no reason to define a new property D for an arbitrary and complex intersection of other properties. You give no principled reason to revise this policy; it's not as though we can define a new term for every possible intersection of properties in the Universe, so it is inevitable that our interests and desires will play a role in which intersections we pay heed to.
On meta level:
As it happens, no human being employs a language that has been completely purified of all interests, all values, all pragmatic considerations, of everything but the Truth.
Fallacy of gray. Just because we are not perfect, does not mean that some ways are not better than other ways. Humans are not perfectly unbiased, but we could still avoid the most obviously biased arguments.
Liking or disliking a group is not a problem per se. The problem for a rationalist would be if your liking or disliking motivated you to change your own perception of...
Robin Hanson has wondered why folks seem concerned about inequality based on some stuff, like race, gender, sexuality, ethnicity, and disability, but not other stuff, like height, appearance, intelligence, sleep, conscientiousness, and perhaps most importantly, happiness.
My explanation: Race, gender, sexuality, ethnicity, and disability are fairly discrete ways of classifying people. Most people (though not all) can be categorized fairly neatly in to a single race, birth gender, desired gender, and sexual orientation. By contrast, looks, smarts, and happiness all vary in a continuous fashion. For some/all of these characteristics, there's a bell curve--many people in the middle and fewer people at the extremes.
Why should this matter? One of the things that comes up a lot on this blog is how irrational people tend to get when talking about politics. For a really rousing political argument, you need group identification--hence the Greens and the Blues in Eliezer's original politics essay. How would that essay be different if everyone was a different shade of turquoise, somewhere on a continuum between blue and green?
Or here's another thought experiment: Instead of height varying in a continuous fashion, everyone in society is either a Tall or a Short. Just like in the real world, taller people tend to make more money and assume more leadership roles, but instead of saying "he got the job because he's taller", people would say "he got the job because he's a Tall". How well do you think society would handle this? (BTW, looks like shorter people live longer, so height inequality may actually favor shorter folks on balance.)
Speculation on factional conflict and social distance
The hypotheses I'm advancing here is that an argument between, say, a radical feminist and a men's rights activist, is not all that different in kind from an argument between a Democrat and a Republican, an Israeli and a Palestinian, or a Hutu and a Tutsi. (Regarding gender in particular: factional conflicts may be harder to spot when almost every human is a member of one relevant faction or another. There's no big, neutral third party to say stuff like "wow, this is getting kinda out of hand".)
Why do people think about competing factions this way? Well, humans are social animals that evolved to live in tribes. Much hunter-gatherer violence was inter-tribal.
The more iterated a prisoner's dilemma is, the more it makes sense to lean towards cooperation. Dilemmas between fellow tribe members would be highly iterated; dilemmas between tribes less so. Additionally, people of different factions would likely have lower genetic similarity--note family feuds, for example.
The below XKCD comic illustrates what I suspect may be a resulting fact about human nature: we're less likely to have friendly, empathetic feelings for those who seem distant or abstract.
More evidence for this idea:
Speculation on equality and fairness
Why does so much inter-factional conflict focus on issues of inequality? Robin Hanson's guess: "our distant ancestors got into the habit of complaining about inequality of transferable assets with a tribe, as a way to coordinate a veiled threat to take those assets if they were not offered freely." I'm not sure agree with this 100%. I suspect that inequality between tribes (e.g. over access to salt, as mentioned in War Before Civilization) may have been a bigger issue than within-tribe inequality. And "veiled threat" doesn't seem quite right--I suspect the feeling that things are unfair and I deserve more is just a signal for me to take stuff from others, in the same way hunger is a signal for me to eat stuff. (Yes, many of us modern humans are generally pretty good about suppressing this "things are unfair" instinct, but that doesn't mean our ancestors were.)
To replicate one's genes, it's useful to have certain things like food, water, and respect. But there's not always an unlimited supply of this stuff, and the only way to get more of it may be to get others to give it to you. You could forcefully demand all of it, but if there's more than one person doing that, you're liable to trigger an expensive fight (or lose your credibility for making forceful demands). Additionally, demanding all of a resource will trigger fiercer resistance from others, who really need at least some of it. Demanding that you get one Nth of some important resource, where N is the number of people fighting over the resource, is a strategy that won't cause you to come in to conflict with others doing the same.
Equitable distribution doesn't necessarily maximize collective utility, however. A world where the supply of disposable diapers was distributed equitably would have substantially less utility than the current world, where disposable diapers are concentrated in the hands of people who have young children.
From fairness to empathy
Pat and Jesse are roommates. Pat enjoys doing something that also happens to annoy Jesse. Consider the following two scenarios:
Scenario A: Jesse asserts a right to not be annoyed. Pat responds by accusing Jesse of being oversensitive and asserts a right to continue with the activity.
Scenario B: Jesse shares preferences. Pat shares preferences. They work together to find the solution that satisfies their collective preferences maximally. This solution could involve a behavioral change on Pat's part, a behavioral change on Jesse's part, some combination, or Jesse just learning to live with the activity. Throughout their discussion, they make it clear to each other that they respect and like one another and care about each others' preferences. They don't worry too much about who is "giving in" or "making a concession", and are careful to make requests rather than demands.
I think most people would agree that Scenario B is ideal. Unfortunately, many modern conversations about social justice look more like Scenario A. It's common, for instance, for members of different groups to argue about who has it worse. This seems like a failure mode for a number of reasons:
This dialogue may be better than nothing, just because preferences do end up getting communicated. But I think we can improve on it. In particular, listening to other peoples' preferences seems pretty key for priming cooperative, friendly behavior and getting them to consider yours. "Seek first to understand, then to be understood." (Note: This can be tough.)
To give a concrete example: Maybe women have a preference for not being objectified sexually that men don't share. Men can respect and work to achieve that preference even if they don't share it--empathy over fairness!
Of course, if priming cooperative behavior fails to work, continued loud complaining may be optimal. Sometimes cycles of defection can't be broken.
Non-ideal brains
Several years ago, I was taking a political science class and the professor was discussing feminism. At one point, the (male) professor said something like "But beyond all this, women are just better than men, and I think we all know it." This offended me, and I fumed to myself internally, thinking that if he'd made the opposite claim he'd probably get in trouble with the administration of the college, etc.
You may already have an opinion on this issue, thinking that either I should not have gotten offended or the professor should not have made his statement. But lets look through this issue through the Pat and Jesse lens. I have a preference for people to respect me, and for me not to feel angry and offended. There are a few different ways for these preferences to be achieved. There's not necessarily a "right" way. Preferences about my misfiring fairness neurons are preferences like any others. (I'm living in a first-world country that's the product of thousands of years of technological development. I've got all the food I want. In theory, grabbing even more resources shouldn't be a very high priority, but my brain still wants to do it. Hence the "misfiring" allegation against my neurons.)
Unfortunately, we all have these unwanted instincts. I think it makes sense to try to avoid activating the instincts in both yourself and others. I'm glad that saying "People of Group X suck" is considered worse than saying "You suck"--Western society has developed some useful memetic antibodies to nip inter-factional conflicts in the bud.
Thanks to HughRistik for offering feedback on this post.