Obviously you choose the ‘boundary of the whole’ according to the location of the instance of hate speech. E.g., Jews were a minority in 1930s Germany; LGBTQ individuals are a minority more or less everywhere; Muslims are a minority in London, etc. etc. There’s nothing about this that’s difficult to understand, and you really just seem to be trolling at this point.

As to your ‘so what’, you now seem to accept that there is a connection between minorities and hate speech, which was the point at issue.

There's no connection. It's an observation at best, based on what you define as a minority. For example, billionaires are also a minority group and frequent recipients of hate speech. Do you disagree?

I don't disagree that they are a minority group. I haven't seen examples of hate speech directed at billionaires per se. In this context, people are usually thinking of ethnic, religious and sexual minorities - but of course you know that, right?

To pick a concrete (if inflammatory) example of the subjectivity of "minority", perhaps it is worth considering the example of "Trump voters". Within the context of the US, they are less than 50% of the population, so would it be hate speech to insult his supporters?

To pick a smaller minority, what about "Baby Boomers". Should the "OK, boomer" meme be banned as hate speech? At 22% of the US population, they constitute a smaller minority than the proportion of non-white Americans (28%, excluding White Hispanics).

No-one thinks that anything insulting said to any minority automatically qualifies as hate speech. This is a straw man. The point is that e.g. ethnic, religious and sexual minorities have historically been some of the primary victims of hate speech.

Thanks for clearing that up. Maybe the problem then is the misleading nature of the phrase "hate speech". The speech that is being banned isn't distinguished by being "hateful", but because it is targeted at certain subjectively chosen minorities.

Don't get me wrong, I'm not saying that we shouldn't give extra protections to groups of people who have historically faced disproportionate amounts of violence (and other harms), just that we should maybe call it "selected minority endangering speech" instead of "hate speech", and be clear about the specific cost-benefit trade-off we are making by how we choose and delineate those categories of people and how much speech is covered.

It's pretty easy to find out what people mean by hate speech. What you're doing seems a bit like people who derail discussions about homophobia by pointing out that homophobes aren't literally afraid of homosexuals. In both cases, the terminology is well established. Fussing over it just serves as an excuse to avoid addressing the problem.

I'm not trying to derail the discussion by pointless complaints about etymology, I'm saying that part of the problem with the concept of hate speech is that the name given to it obscures (accidentally) the nuances of how it is applied in practice.

A better analogy would be if the critics of homophobes genuinely thought that homophobia was literally a fear of homosexuals, causing the homophobes to complain that this framing of their position made it hard for them to explain their objection to homosexuality.

By hiding the subjectivity of "hate speech", people then get surprised or angry when it does or doesn't get applied to terms like "communist bandits" or "OK, boomer", or "eat the rich". The real debate isn't about whether the terms are hateful (as the name suggests), but whether the specific groups that are targeted need the specific protections being implemented.

So you were genuinely and not merely rhetorically confused when you asked whether insulting Trump supporters would qualify as hate speech?

Sorry, it just seems like you are deliberately trying to introduce confusion about what hate speech is into this discussion.

There's nothing particularly 'subjective' about the definition of hate speech. At least, it's no more subjective than the definition of 'free speech' or 'censorship' or any of the other relevant concepts in this domain. There's a perfectly objective history of persecution targeting certain minority groups.

I was genuinely looking for a logically consistent framework for excluding "insulting Trump supporters" from being an example of hate speech. I'm sorry if it seemed like I was labouring the point too much by asking where the lines around hate speech should be drawn.

I accept that there are objective historical examples of majority groups persecuting minority groups, and I'll ignore the difficulties of constructing well-defined subsets of a population (e.g. "working class") or whether a given group is numerically a minority (e.g. "females" in many countries). What I still think is subjective, though, is how much (and what sort of) persecution is necessary before a group becomes entitled to claim that hateful language used against them is "hate speech".

Imagine a hypothetical African country that had, say, France as its colonial occupier, under an apartheid system, but then allowed free elections, leading to the native population gaining political power. If the native population had talked about "getting rid of" their French occupiers, while the apartheid system was in place, presumably your definition of "hate speech" wouldn't have applied to that speech. But would your definition also not apply to similar speech (targeted at the same French people) after the occupying minority population lost their power? Would some amount of time (and violence) have to pass before the minority was entitled to point out that the hateful speech directed towards them was this special kind of "hate speech"?

Again, I apologise if this seems like a contrived example (and it's very hard to come up with an example that people don't have instinctive pre-conceptions and biases around), but I'm trying to explore if your definition really is as neutral as you think it is. You're right, though, that terms like "free speech" can be very nebulous, while still being useful concepts.

Yes, the history matters.

If your point is that you can contrive edge cases then, well, duh. There are also edge cases involving free speech and just about every other legal/moral/political concept.

My point isn't that edge cases exist, but that the edge cases force us to examine the process by which we decide whether something is or isn't hate speech.

I think that for a lot of people, they are exposed to a few clear examples of hate speech, and unconsciously build a heuristic that says "Anything that makes me feel the same sense of disgust towards the speaker or sympathy towards the target, is hate speech". Fortunately that heuristic works quite well most of the time for people, but I think it can work so well that the people using it don't question it, and don't realise that their definition has some blind spots in some areas, or scope-creep in others.

So, regarding my contrived edge case, when you say "the history matters", do you mean that the majority can continue to talk about "getting rid" of the minority without it being classed as hate speech, because the minority were historically privileged?

Alternatively, perhaps you mean "speech can change from being allowed to being hate speech (and vice versa) over the course of history". I don't disagree that the meaning of (and people's sensitivities to) words can change over time, but in my example, the change in circumstances happens in a single day. If that is significant, then it means the definition of hate speech depends not solely on the words themselves, or the size of the target group relative to that of the speaker's group, but rather on some sort of determination of whether the target "deserves" to be a subject of hate because of their membership of a group that you (the arbiter of hate speech) deems to be currently or historically over-represented politically.

I really am trying not to put words into your mouth, and I appreciate you taking the time to understand my concerns. Hopefully we'll both be more clear about what we mean when we use the term "hate speech" in future.