‘Homophobia’ looks like it’s a word for a kind of fear (likewise for ‘transphobia’, ‘fatphobia’, etc.); ‘misogyny’ looks like it’s a word for a kind of hatred (likewise for ‘misandry’). And ‘racism’ looks like it’s a word for some sort of system or theory (likewise for ‘sexism’, ‘ableism’, etc.). (‘Anti-semitism’ displays yet another pattern.) But it’s pretty clear that everyday usage doesn’t reflect any such distinction – whatever homophobia is, it seems like the same kind of thing as misogyny and racism and anti-semitism. When people want to talk about that kind of thing in general, they sometimes use words like ‘prejudice’ and ‘bigotry’, and when they want a short punchy word they use ‘hate’.
But it’s tricky to say what kind of thing this ‘hate’ is: it’s not simply ‘hatred’, the emotion, since it seems to gesture at something larger, more flexible, and more systematic than a single specific emotion. Debates about the definition of ‘hate’ generally, or particular types like ‘homophobia’, recur periodically, and recently I saw a couple of instances that prompted me to write something about them.
(This post is a long one: the tl;dr is that these are terms introduced to name real-world phenomena which are easier to recognise than to explain or define, and so they don’t admit of a clear definition. The best slogan is that they are ‘failures to see certain people as fully human’, but it takes me 3,000 words to explain that phrase.)
The first thing I saw was this post (which I think is good) from Scott Alexander about the difficulty finding a consistent definition of ‘racism’ that actually fits the way the term is used. Just after reading that post, I saw a Facebook discussion about this article (which I don’t think is good), which turns on related issues about whether prejudice against white men exists/is a problem/should be called ‘racism’.
Scott distinguishes three definitions of ‘racism’, and I noticed that none of them really matched the way I’ve been using the term in my posts here. So I thought it might be useful to outline what I think it’s best understood as meaning, and more broadly, what kind of thing ‘hate’ is, of the sort that’s involved in ‘hate crimes’, ‘hate speech’, etc.
Scott’s three candidate definitions of ‘racism’ are:
“1. Definition By Motives: An irrational feeling of hatred toward some race that causes someone to want to hurt or discriminate against them.
2. Definition By Belief: A belief that some race has negative qualities or is inferior, especially if this is innate/genetic.
3. Definition By Consequences: Anything whose consequence is harm to minorities or promotion of white supremacy, regardless of whether or not this is intentional.”
My facebook friend Robert distinguished two senses of ‘racism’/’sexism’:
In one sense, “racism and sexism require institutional privilege and power to carry out.” This isn’t a definition, just a proposed necessary condition, but it obviously has an affinity with Scott’s ‘definition by consequences’.
In another sense, “it’s racist or sexist to associate someone with, or blame someone for, the actions or attributes of people on the basis of shared skin colour, sex, or whatever else have you.” This is also not a definition, but a proposed sufficient condition, which conflicts with the above necessary condition.
Roland Merullo in the BostonGlobe article doesn’t actually use the term ‘racism’, but the focus of their discussion is on a form of “hatred” that involves “tossing [some individuals] into a group, and attributing to that group the most unattractive characteristics imaginable.”
The quotes from Merullo and Robert both suggest a fourth definition, additional to Scott’s three:
4. Definition by group-based thinking: A psychological habit of attributing character traits or responsibility for actions onto racial groups, rather than onto individuals.
Scott thinks that the definition-by-consequences is “the one that scholars in the area are most likely to unite around”, but that it doesn’t really fit the way the term is typically used. For one thing, it has the perverse implication that racist actions (his example is a KKK march) which prompt such a backlash as to discredit the people or ideas behind them would not count as racist – since their effect was, overall, the promotion of racial equality, contrary to the intentions of the agents.
For another, it changes racism, or any other sort of ‘hate’, from a cause into a description of effects: it renders it meaningless to ask whether, or postulate that, a particular fact is ‘the result of racism’ (compare: if ‘danger’ just means anything that could harm you, it makes no sense to ask whether a particular harm you suffered was ‘the result of danger’).
I think Scott’s right that the definition-by-consequences, as he’s formulated it, isn’t a very good one. But I don’t think that his formulation really captures what’s meant by people who define ‘racism’ (or another sort of hate) in terms of systematic disadvantage. At least, it doesn’t capture what I mean by the term.
Before attempting a better formulation, let me say that I do agree with Scott that you can get a lot of confusion, and a lot of bad rhetoric, by shifting between different meanings of a ‘hate’ word like this – particularly between very broad meanings that apply to almost everyone in contemporary societies, whatever their beliefs and actions, and narrower meanings that convey a very strong moral condemnation.
But I don’t think that the unclarity of ‘hate’ terms is just a matter of laziness, confusion or obscurantism. I think it’s also due to the fact that ‘hate’ terms are what philosophers sometimes call ‘natural kind terms’, like ‘gold’. People had words for gold long before they knew what defined gold (namely its atomic number). They recognised a certain kind of thing in the world, which had some noteworthy features (colour, shine, weight, resistance to corrosion), and was worth having a word for, even though they didn’t know exactly what it was. They recognised first, and defined later.
Likewise, I’d suggest that ‘hate’ is not best thought of in terms of trying to stipulate the right definition, but in terms of a salient feature of human societies which people recognised and introduced terms for prior to being able to specify its exact nature. So some uncertainty about the definition is to be expected.
Still, we might hope for an answer at least to the ‘what type of thing?’ question. This kind of thing, of which homophobia, misogyny, racism, etc., are examples, is it a motive? An emotion? An institution? A belief? Here’s my best current hypothesis:
‘Hate’ is a holistic pattern of de-valuation, which can be embodied in an individual mind, a social group, or a social system, insofar as all of those things can express or embody sets of values.
That’s a bit wordy, right? Sorry. I think the best way to bring out what I mean is to consider a different word, ‘love’. I don’t mean ‘love’ as a feeling, something that might overwhelm you or warm your heart at a particular moment. I mean ‘love’ as an achievement, as something demonstrated over a period of time. I mean the kind of thing that’s revealed to be missing by things that make us say ‘you don’t really love them’.
Suppose someone fills you with warm fuzzy feelings every time you think of them, but when you’re offered a way to enrich yourself at their expense, you take it without hesitation or regret. You don’t love them: if you did, that love would manifest not just in warm fuzzy feelings but in an unwillingness to hurt them, and guilt when you did.
Or suppose someone brings you joy whenever you’re around them, but when you hear about some success or accomplishment of theirs, you feel no pleasure but instead resentment. You don’t love them: that requires not just wanting them to be around, but wanting them to succeed.
Thirdly, suppose you’re spontaneously kind and affectionate to someone, and seek out their company, and support their endeavours, but only as long as they look exactly the way you want them to, say exactly the things you agree with, and act in the exact ways you recommend. Again, I’m inclined to say that you don’t love them, because love that’s conditional like that isn’t really love.
What the above examples are getting at is that, at least in one of the ways we use the word, ‘love’ isn’t a term for a feeling but for something that manifests in all sorts of feelings, and which sometimes doesn’t manifest in any very strong feeling but just in doing what someone needs you to do, however you’re feeling. And it’s something that has to have a certain kind of ‘unconditional-ness’ – perhaps not completely unconditional (I think it’s ok to stop loving someone in response to some kinds of change) but enough that the beloved doesn’t have to ‘earn’ your love by conforming to your specifications.
‘Holistic pattern of valuations’ is my best effort at formulating the kind of thing that love is. ‘Valuations’, not ‘feelings’, because valuing something implies different feelings in different situations – joy when it’s present, fear when it’s in danger, guilt when you know you’ve harmed it, etc. And ‘holistic’ because the test of love isn’t what you feel or do on a particular occasion, but the way you incorporate concern for the other person into your life, how you balance it against other things you value, how you hold onto it when conditions change. It may take many forms (just compare familial love and romantic love), and I’m not sure I could say what unites them all. There does seem to be a shared core of caring about another person for their own sake, appreciating and respecting their perspective and the richness of their life, but the point here is to get clear on what kind of thing it is that love is, not which things of that kind are love.
A milder version of love – a more lukewarm incorporation of other people’s value into our overall approach to life – is something most of us aspire to have for everyone. Call it ‘morality’, ‘solidarity’, ‘human decency’, whatever: I hope to have, and hope that those around me have, a robust unwillingness to directly harm or insult others, a respect for how others choose to live, a disposition to be pleased by other people’s happiness or at least to promote that happiness when I can easily do so, and so on.
Ok, so then what’s ‘hate’ – the kind of thing that racism, homophobia, fatphobia, misogyny, etc. are instances of? I think it’s something like the opposite of love: a holistic pattern of de-valuation. To hate people, in this sense, is to fail to incorporate their worth as people into your life and views and judgements, in a way that contrasts with how you do thus incorporate the worth of other sorts of people. That’s why it’s often said that ‘feminism is the radical notion that women are people’ – the ‘radicalism’ implying that they are routinely treated as something else. That ‘treating as something else’ is the holistic de-valuation I’m talking about here.
This de-valuation of some category of people can manifest in many different ways. Most simply, it can be a callous indifference to what happens to them – like a willingness to accept their deaths as collateral damage while treating the death of a single person in a preferred group as an outrage. More interpersonally, it can be a lack of interest in them as individuals – a lack of respect for, or enquiry about, their opinions, a lack of concern for how they do or might react to something, an unwillingness to update an opinion about them based on evidence.
It’s also often expressed in allowing a simple, one-dimensional, image to subsume all other aspects of them. This might happen without any kind of malice or feeling: it might just be hearing ‘a single story‘ about what life is like in some country, and never bothering to think ‘hey, that’s probably only one of a million different ways that people there live.’
(I don’t mean that ‘hate’ is just lazy generalisation: lazy generalisation is a pretty common thing, precisely because it’s the lazy, easy, effortless option for thinking about complex realities. But when we respect people’s personhood properly, we’re usually quick to recognise lazy thoughts about them as lazy. Imagine how you would react on hearing a good friend who happened to skateboard occasionally being dismissed as ‘just a skateboarder’: unless it was clear from context that this wasn’t meant seriously, you’d spontaneously find yourself thinking ‘well, not just a skateboarder – so many other things! For example…’ The failure to correct in that way, I think, is what manifests a lack of respect.)
Of course the most obvious manifestations of this sort of de-valuation involve literal emotions of fear, hatred, resentment, contempt, and so on. But of course it’s ok to feel such emotions sometimes – towards people who have individually done something to warrant them, or even towards people who haven’t, when you recognise their unwarrantedness. And I think if you respect someone’s humanity you’re usually only going to feel strong negative emotions towards them in those kinds of cases: if they haven’t done anything to warrant the emotions, your respect for them tends to spontaneously prompt you to stop and wonder why you feel that way, to remember other aspects of them, to ask if the problem is with you rather than with them. Lack of such respect – ‘hate’ in the sense here under analysis – allows you to feel consistent, strong, negative emotions towards someone (and to act on such emotions), independently of what they’ve done and without recognising it as an unwarranted emotion.
I think there are certainly respects in which ‘hate’/de-valuation that involves strong emotions is different from that which doesn’t, and poses different risks to its targets. There’s a difference between the misogyny of a mass shooter stewing in sexual resentment and the misogyny of a benign and friendly professor who never takes ideas seriously when they’re suggested by women. I don’t mean to deny the importance of that difference. But there’s also something in common, a lack of the respect for women’s humanity which would both let their ideas be heard fairly and would obstruct the growth, and festering, and acting on, of that resentment. I think ‘hate’ terms (in this case ‘misogyny’) are often, and appropriately, used to refer to this broader phenomenon which unites both cases, even though one case very obviously involves the emotion of ‘hatred’ and the other doesn’t.
Indeed, sometimes this de-valuation is even manifested, not in any display of disrespect, but in how respect is made conditional for some people more than others – in treating a gay person just as well as a straight person as long as they never discuss their romantic lives, never do anything reminiscent of a gay stereotype, never lapse in their display of acceptable normality. Indeed, this de-valuation may even come out as a special idolisation of an individual who’s seen as sufficiently ‘rising above’ their demographic, lauding them for being ‘not like the others’ as long as they work hard to distance themselves from those others.
That’s why I keep emphasising the holism of ‘hate’ terms. The overall pattern is what matters, including counterfactuals like ‘how would I see this person if they failed at things, or disagreed with me about something’. Particular actions, words, feelings, or beliefs just serve as symptoms.
So far I’ve been talking about ‘hate’ as a feature of individual psychology. But an advantage of this definition is that it extends fairly directly to social systems. Social systems aren’t exactly ‘people’, but they’re like people in that they have (express, enact, manifest) a worldview and set of values and priorities. They’re like a cloud of individual judgements and experiences thrown up in the air to settle and congeal. Laws express valuations in what they allow and prohibit, what they make exceptions for, what they guarantee. Media works express valuations in how they frame stories, who they place as protagonists, what they present as scary. Cities and businesses and bureaucracies express valuations in what they repair and what they don’t, what they invest in and what they don’t, who they employ and who they don’t, etc.
Of course, social systems are usually a lot less coherent in their value systems than individuals. At any given time there are competing valuations being fought for by opposed groups, and individual decisions may aggregate in unexpected ways to produce results that no individual wanted. But there are also mechanisms that tend to align different valuations: laws affect employment decisions, workplace environments affect people’s exposure to each other, media depictions get internalised, etc. In general, I think there’s enough coherence to talk meaningfully about the way that a given society values something.
One of the most basic things that social systems value, in this overall-coherence sort of way, is people. The idea that people deserve respect is baked into laws, customs, media narratives, rules of politeness, and lots of other things, in very basic ways. Insofar as ‘love’ is a holistic pattern of unconditionally valuing someone as a person, we could say that a good society embodies ‘love’ for its members.
It’s against this background that we can then talk about a society ‘hating’ some of its members – when the laws seem tailored to hurt them, or the legal system enforces laws more strictly against them, or the popular stories always make them villains or sidekicks, or they’re only hired for low-paying jobs and only rented sub-standard housing, or… well, the list could go on. But just as with person-level love and hate, the manifestations aren’t the definition. They’re signs that, in the complex web of subtle influences which hold together a society, some people aren’t cared about as much as others, if at all. They’re signs that people can recognise if they looking in the right way, even if they can’t explain or define the mechanisms that generate them.
So, that yields an account of what kind of thing systemic X-phobia (or X-ism, X-hate, miso-X, etc.) is, and what individual X-phobia is, and why they’re connected. It also yields secondary definitions of X-phobic acts, beliefs, institutions, etc. as being those which either manifest or contribute to X-phobia on either level.
Let me wrap up by bringing this back to Scott’s perplexity over different definitions of ‘racism’. If I’m right about ‘hate’ terms in general, and about ‘racism’ in particular, then there’s something true in both the definition-by-motives and the definition-by-consequences. While the former identifies racism as feature of individual minds, and the latter identifies it as a feature of social systems, my proposal allows it to be a feature of both.
In particular, the holism of my definition vindicates a key feature of the definition-by-consequences, namely that there is no symmetry or equivalence between racism and ‘reverse racism’. No single type of beliefs, feeling, action, or utterance can be judged racist or not outside its social context.
This is clearest for systemic, as opposed to individual, ‘hate’: if a society hates black people but not white people in the way it treats them, then individual disparagements of white people are neither manifestations of, nor contributors to, any society-wide de-valuation of people based on their race.
But a similar asymmetry applies (though less straightforwardly, I think) for individual ‘hate’, again because of that holism. What it means to show concern for someone, or to show a lack of concern for them, depends on their needs. If I fail to help someone when they really need it, that could reasonably be taken as evidence against my claim that I love them. But if I know they’ll be fine without my help, the same failure to help means much less. Likewise, if black people don’t want to listen to my opinion because I’m white, I might not like it but I won’t really suffer – I have no shortage of people who will listen to my opinion. Whereas if a white person doesn’t want to listen to someone’s opinion because they’re black, the same is often not true. Whatever criticisms there might be of the former attitude (black people dismissing white people’s opinions), I don’t think it’s evidence of ‘white-hate’, and isn’t symmetrical with the reverse situation.
On the other hand, my proposal isn’t simply Scott’s ‘definition-by-consequences’, and doesn’t have the perverse consequences he identifies. Systemic ‘hate’ is a causal factor, not just a summary of effects; things that harm people of a certain group aren’t automatically a consequence of hate against that group. And, to return to Scott’s example of the hypothetical KKK march which ends up discrediting white supremacists, that action can be racist because it’s an overt expression of hate against a racial group, even if its actual effect happens to be to benefit members of that group.
To put it another way, to use a ‘hate’ term is to implicitly presuppose certain things about how the world works: that the multiple disadvantages suffered by a particular group are not coincidences, but rather reflect something stable and coherent about the way a society values different people. That hypothesis isn’t trivial: it might be false, but in many cases it’s very plausible. And that hypothesis can be intuitively understood, and evidence for it identified, even in the absence of a precise definition.