How to argue about whether something is racist

This class is about how to argue whether a text or action of some kind is racist; this isn’t about whether a person is racist. On the whole, that isn’t a productive argument (although you sometimes have to have it). The first step in a useful argument about whether something is racist is to try to figure out why we’re having the argument in the first place—what the determination of racist/not racist will do for us is what enables us to decide which definition of racism is the most relevant.

All of this may seem confusing to you, since you might be accustomed to thinking of “is this racist” as a straightforward question of right and wrong—if it’s racist, it’s morally wrong, and if it isn’t, then it’s morally right. And while I do think racism is morally wrong, I also think there is a continuum with some things being more racist than others (as you’ll find later, it’s even possible for something to be racist and anti-racist at the same time). Even if something is morally wrong, you’re likely to respond to it in different ways. For instance, your 90 year old not-quite-all-there grandpa might use a term we now consider racist but which was considered the polite term when he was young. You’d react to that differently than if someone your own age (who knows perfectly well it isn’t an okay term) uses it. You might not do anything at all with your grandpa, but drop like a hot rock the person your age.

If a person making hiring decisions for your company said something racist, you’d react differently than you would if some random person in line at the grocery store said the same thing. If you were HR, you might fire them—whether or not they intended to be racist, on the grounds that their mere presence on the hiring team jeopardized your company.

In this class, we’ll spend a lot of time talking about different definitions of “racist,” which ones are more useful than others, and under what circumstances.

On the whole, definitions of racism tend to emphasize one of several points on the rhetorical triangle: text, intent, consequence, relationship to context,or impact on audience.

For instance, for some people, as long as a text does not have racist epithets it isn’t racist (although those same people don’t usually immediately decide that a text with racist epithets is racist—more on that below). Others do decide that the use of racist epithet (by any character or in any context) is racist. Both those decisions rely purely on text.

This criterion—presence or absence of racist epithets—seems to me the least useful, in that there are the fewest instances in which it seems to me especially relevant. A person can, after all, argue for the expulsion or even extermination of another race without using racist epithets (in fact, that’s most commonly how it’s done). And some anti-racist texts can use racist epithets to persuade the audience that racism is harmful (the argument often made about Django Unchained).

Many people believe that the main problem with racism is that it is hostility against members of another “race”—that is, it is an issue of individuals’ feelings. If you define “racism” as “hostility toward members of another race,” then you will tend to look at texts for evidence of hostility—affective markers, boosters, and other linguistic signs of anger. That method also doesn’t work particularly well, as some of the most racist policies have been invoked in the name of kindness, with apparently calm tones, or by appealing to “facts” and “reality.” (White supremacists often call themselves “racial realists.”)

Emphasizing the feelings that individuals have is one example of how people imagine racism to be a problem of individual agency (rather than systems). In this model, racism exists because too many people choose to be racist, or allow themselves to slip into racist and ingrained habits. If enough individuals chose to stop responding in racist ways, then racism goes away. (That is a problematic assumption.)

Intent seems to me a slightly more usable criterion, but only for limited circumstances. It is important in social situations in which we’re trying to determine if a person should be forgiven. If a person says something racist, but didn’t mean to (didn’t realize it was a racist term, was thoughtlessly repeating a meme they didn’t understand to be racist), then you’re more likely to be willing to forgive them. If they keep saying that thing, although it’s been explained to them that they’re saying something racist, then we might conclude that they really do intend to be racist.

Intent matters in some legal situations (e.g., hate crimes) but not others (e.g., the question of disparate impact). “Disparate impact” is a kind of racism that doesn’t require any intent—if you have a policy with no intent of hurting a particular race (or religion), but that’s exactly what it will do, then you’ve got “disparate impact,” which has been ruled discrimination. If you ban hairstyles that you consider too casual, and they’re precisely and exclusively the ones worn by people of a particular race, then—whether or not it was your conscious intent—your policy has racist consequences. (A lot of school dress codes get challenged on exactly these grounds.)

Intent, like the question of feelings, assumes that racism is the consequence of individuals choosing or allowing themselves to be racist—that there is individual agency in racism. I think the reasoning works something like this: evil in the world is the consequence of individuals choosing to do evil things; racism is evil; therefore, racism must bet he consequence of individuals choosing to be racist. The assumption is that if we had a world in which no one intended to be racist, there would be no racism, but that isn’t the case.

Thus, intent matters for law and social castigation, but it’s of limited importance otherwise. For instance, google image search “beautiful hair.” You’ll see a very racist outcome—almost exclusively white women (and the nonwhite women usually have very, very high maintenance hair). But there was no one intending to create a racist cultural view of what is beautiful hair. There are people intending to sell products, and doing so within a racist culture.

One of the more straightforward ways to measure whether a text or action is racist is to look at whether it reinforces existing racist practices and structures. The most productive arguments, it seems to me, work within this framework. I think it’s helpful partially because it allows a more nuanced discussion—it’s possible to talk about how much harm something caused, what kind, and to whom (rather than a binary of harmful/not harmful). It’s also possible to talk more intelligently about texts that are both racist and anti-racist (South Pacific, To Kill a Mockingbird) if we think about harm; we can talk about the kind of harm the text or action tried to prevent or ameliorate and what kind of harm it caused.

Thinking about consequence also enables us to talk about the same act or text having different consequences in different era or with different audiences. Some critics of American Sniper argued that it was not seen as racist in its showings in Iraq because viewers saw it as demonizing a particular political party, but it had racist consequences in the US because viewers saw it as confirming demonized (and racist) views of Iraqis. It could be argued that To Kill a Mockingbird was progressive for its era, but now it’s actually regressive.

That last comment brings up the argument about relationship to context—what do we do about texts that are racist, but less racist than was the norm for their era or culture? If we think of “racist” as an absolute category—something is either racist or it isn’t—then we’re hopelessly entangled by these cases. If we can think of it as on a continuum, then we can talk about them more sensibly.

We have to be careful, however, not to assume that things have been getting steadily less racist as time goes on. Huckleberry Finn (1885) is much more racist than Uncle Tom’s Cabin (1852), and there was a lot of anti-racist being done in its era. Sometimes we excuse texts by overstating the dominance of racism in an era—even in eras in which it was common, there were people who spoke against it. While I don’t think that racism is subject to pure agency (people could simply choose to be or not to be racist), there are some choices. Being in a particular culture or time doesn’t force someone to be racist, after all. So, while I think it’s useful to put texts in contexts, it should be in service of a nuanced understanding of how the racism works in them, not as a “get out of racism free” card.

The criterion of impact on audience might be subsumed under consequence, but students have found it useful to separate them. The impact on the audience might cultural (the text problematizes or confirms common racist attitudes) but it might also be more individual (the text makes individuals uncomfortable in a good or bad way). For instance, while the word “niggardly” has nothing to do with the similar sounding racist term, it’s reasonable for some people to be made really uncomfortable by it, and so it’s reasonable to try to avoid using it, just on the grounds of how it makes people feel. On the other hand, while the phrase “welching on a bet” originally came from a racist stereotype about the Welsh, I don’t think anyone knows that anymore (or has that stereotype) so, at least in the US, it doesn’t seem to me helpful to call that a racist phrase.

My point in giving all these criteria is not to set out some kind of easy decision-tree on “is this racist.” Instead, I’m suggesting that students see these criteria as stases for arguments about racism. It’s hard to have a good argument on anything if interlocutors are on different stases, and a lot of people don’t realize that any argument can have multiple stases—and you can choose among them. Sometimes, for an argument about racism to become more useful, people have to agree first on the stasis, and that might mean that people might have to understand that their notion about how to determine “is this racist” isn’t a useful criterion. In this class, you’ll be looking at that question a lot–what is the stasis for this argument, and is it the most relevant and useful stasis?

A lot of definitions with racism describe it as a problem of individuals who have hostility based on irrational beliefs. Google gives us “prejudice, discrimination, or antagonism directed against someone of a different race based on the belief that one’s own race is superior.” And prejudice is “preconceived opinion that is not based on reason or actual experience.”

So, racism is an individual feeling antagonism against someone of a different race because she believes her race is superior, and that believe is not based on reason or actual experience.

Hubert Sumlin has accused Chester Burnette of being racist. If I agree with those dictionary definitions of racism, then I need to decide if Chester feels antagonism toward another race, if he thinks his race is superior, and if his beliefs have no basis in reason or experience.

One way to test whether that is a good definition would be to take a case we think a definition should catch. I think it’s reasonable to decide that the Nazis, since a major part of their political policy was the extermination of races, were racist. So, if this definition of racism doesn’t apply to them, then it isn’t a good definition.

Well, interestingly enough, if you apply those standards to major current defenders of the Nazis, you’ll find that they maintain they feel no antagonism to the other group but simply want separation. They insist that their support of Nazism is not unreasonable—they can cite experiences of the other races being bad, and one book advocating neo-Nazism has around 1k footnotes, so it gives reasons.

And, of course, the Nazis came to power on a policy of separation, not extermination, and the expulsion of illegal immigrants of that group, although the whole group was kind of suspect. So, that isn’t a good definition of racist.

And, let’s go back to Chester and Hubert. How would I would know what Chester feels and believes?

This is a surprisingly interesting question.

There is a period in cognitive development when children develop a theory of mind—that is, they understand that other people have ideas, feelings, and commitments that might be different from the ones they have (to put it crudely). Other people have other minds, in other words. (Not everyone develops that ability, and they go through life believing that everyone believes exactly what they do—people who disagree are just pretending to have different ideas.)

In some cases, developing a theory of other minds leads to the skill of perspective shifting—the ability to imagine what things would like from those other minds, and at least the effort to see things from that perspective. In an unhappy number of cases, however, it leads to the tendency to think that your mind is the one cleaved to reality, and those other minds are just unhinged or wrong. We also have a tendency to believe that our perception of those other minds is unmediated and perfect. This is such a profound problem that social psychologists call it the fundamental attribution error—it’s fundamental to a lot of other errors. We attribute views and feelings to other people, and then we believe we have perceived those views and feelings.

And that attribution is biased, especially by whether someone is in our ingroup or outgroup, what our feelings are toward that person (mad, afraid, attracted, sad). We tend to attribute bad motives to outgroup members and good motives to ingroup members, and can be astonishingly self-serving in our perceptions. For instance, if we’re feeling aggressive toward someone else, we’re tempted to attribute aggression to them, thereby making ourselves feel that our aggression is a justified response. Or, if we’re attracted to someone, we might interpret their inner state as attracted to us (that’s why stalkers never see themselves as stalking—they sincerely believe they know that the victim is or easily could be attracted to them).

So, imagine that we’ll use the standard dictionary definition of racist. It says “one’s own race,” so this is about what an individual does. And now we have to figure out if there is antagonism or prejudice, and if the person believes his/her race is superior.

If you believe that “Lithuanian” is a race, and you are racist about Lithuanians, believing them to be essentially stupid and criminal, you wouldn’t be able to use the common dictionary definition to recognize your racism (or the racism of anyone else in your ingroup who also believed Lithuanians are stupid and criminal). You would be able to ask yourself, “Is my belief about Lithuanians grounded in reason or actual experiences?” and answer, “Yes!” That’s because you would be able to think of a time that someone you thought was Lithuanian did some stupid. You might be able to think of an expert who also said they’re stupid. You’d be surprisingly likely to assume that every stupid person you met was Lithuanian, or to interpret smart things a Lithuanian did as stupid. If you were forced to acknowledge that there was a famous mathematician genius who was Lithuanian, you might decide she was probably adopted, illegitimate, or of mixed heritage. You might decide she wasn’t really a genius, but had just happened to get something right, or had stolen all her ideas from a non-Lithuanian. In fact, an infinite number of counter-examples would not change your mind about Lithuanians—more important, it wouldn’t even get you to see that your stance was unreasonable, because you could explain them all away.

You also might tell yourself that you don’t feel antagonism—you are just realistic about Lithuanians. And you don’t discriminate, you just don’t give them more than they deserve.

If a definition of racism is going to be helpful, it has to be one that enables racists to realize we’re being racist. And, that definition of “prejudice” isn’t helpful.

It isn’t any better if we’re trying to apply that definition to someone else. The inherent problem with believing that racism is a question of unreasonable feelings and bad intent is that we have to figure out the inner state of Chester, and we’re extremely likely to engage in the fundamental attribution error—if we think Chester is good, we’ll attribute good motives; if we don’t like Chester, we’ll attribute bad motives. There is a connected problem with this—we tend to think of racism as an issue of individual morality. Racism is immoral, and so people who are racist are immoral.

Therefore, if a person is not immoral, they can’t be racist. I am not immoral, therefore I can’t be racist. Ta effing dah.

It’s this false assumption that gets us into those weird cases of someone having said or done something racist and various people saying, “She can’t be racist, because she did these good things.”

There’s a second, and more complicated, problem with this way of thinking about racism—that it has to do with the intent, feelings, beliefs and/or prejudices of individuals. Take, for instance, this clip: It’s every single word said by a person of color in every single Nancy Meyers-directed movie (it lasts about five minutes). Does Meyers exclude people of color from her movies because of antagonism? Probably not. And probably not because of any conscious notion that her race is better. She might not intend to discriminate—maybe she just doesn’t think to include non-whites in her movies; or maybe she is trying to get the most “bankable” stars (and audiences tend to be racist). It doesn’t really matter if she intends to discriminate against non-whites; she does. Having twelve hours of movie with five minutes of non-white is racist. But it isn’t just her racism—it’s the racism of audiences, producers, and the systems of movie-making.

So, how should we think about racism?

First, a few terms that will make all this more straightforward:

  • A “taxonomy” is a method of categorizing. If you have your closet organized by shirts/pants/skirts/jackets, that’s your taxonomy. You might instead organize your closet by what’s useful to have for different outside temperatures (warm/cold/moderate), circumstances under which you might wear them (formal/business casual/casual/athletic), or perhaps by color, how often you wear them, or some other taxonomy.
  • socially constructed. Some people think that a belief is either subjective (random, arbitrary, entirely in your head) OR objective (a perception of a brute fact that is entirely external to your brain, and which exists regardless of whether anyone believes it—also called “ontologically grounded”). That isn’t actually a very useful division. Where, for instance, would you put money? It isn’t subjective—I can’t draw Millard Filmore’s face on green paper and get Starbucks to give me coffee—but it isn’t a brute fact. I can get coffee in the US with a five dollar bill, but not in England. Money is a socially-constructed fact. It is a fact—it is an inescapable condition of our culture, but only because we’ve implicitly agreed to give it that kind of power.
  • salience. That word simply means the condition of sticking out. You might have troops in a line, but with a bubble that sticks out—that’s your salient. In any situation, there are all sorts of things you might notice, but some of them stick out. Those things have more salience. Salience is context-dependent, idiosyncratic, and/or socially constructed. If religion is important to you, then you might notice how many people of various religions there are in a group (or class, or room)—you might notice there are a lot of people who mention being Wisconsin Synod Lutheran. You might also notice that the class is quieter than you think normal. We tend to confuse salience with importance, and attribute causality to salient conditions. So, you would be irrationally likely to assume that the two salient things are causally connected—it’s quiet because there are a lot of Wisconsin Synod Lutherans.
  • confirmation bias. We tend to notice things that confirm our beliefs more, and more easily than things that don’t. Even when we try to “test” our beliefs, we generally do so in ways that enable confirmation bias. If, for instance, you believe that Wisconsin Synod Lutherans tend to be quiet, then you’ll decide that the class is quiet because there are so many of them. If you believe that Wisconsin Synod Lutherans tend to intimidate others, you’ll decide the class being quiet is an example of their having intimidated everyone into silence. A person for whom religion isn’t important, but who has a hypothesis that morning classes are more quiet, will, if it’s a morning class, conclude that this class proves that hypothesis.
  • The ingroup is the group you’re in that is important to your sense of identity. You have lots of them, and they become more or less salient depending on context. Being a Texan (or whatever state you’re from) is important if you’re around people from various states, and would be especially important if someone said something nasty about your state, but isn’t something you’d mention introducing yourself to other people in Texas.
  • An ingroup is often defined (or made more salient) by the sense of there being an outgroup which it is not. In terms of race, the concept of “white” only makes any sense if there is a “not white” to define it. (The various groups that are called “white” don’t actually have very much in common, and used to be designated as different races. But they can claim white because they aren’t non-white.)
  • essentializing (or naturalizing). Ingroups and outgroups are socially constructed—they’re both as real and as arbitrary as money. They have tremendous power but only because a culture decides they do, and individuals can’t suddenly decide, “I don’t see money” and have any impact on how the world works. (And they’ll still use money to buy groceries.) That so many things are socially constructed (such as the boundaries of states and nations, or the boundaries of groups) makes many people extremely nervous—especially people who don’t like uncertainty, ambiguity, or nuance. There are people who get anxious and angry if they are made aware that their taxonomies are socially constructed and might change. Their inability to manage uncertainty and ambiguity cognitively means that they will INSIST that those taxonomies are Real. Race is not, they will insist, a social construct, but a biological one—they will insist those categories are appealing to essences of people, to identities grounded in nature. They will take a socially constructed category (such as nation) and insist that the members of that nation are essentially the same in that they all have certain temperaments, political tendencies, or identities.
  • social/cultural goods. This is a vague notion, but useful. Any culture has things that are culturally marked as goods—money, prestige, status symbols (living in that neighborhood, getting treated this way by police, being able to go to those restaurants), political power.
  • zero-sum relationships. There are some relationships in which the more one category gets, the less the other category gets. So, if I have a discretionary budget of $100 per week, the more I spend on coffee, the less I can spend on going to listen to music. There is a zero-sum relationship between those two categories. If you spend less on coffee, you have more to spend on music. Many people grow up in a highly competitive family situation in which there is a sort of zero-sum relationship in parental love or attention—the more that a sibling gets, the less there is for you. So, a person in that world might think, “if my parents spend less love on my sibling, there is more for me.”
  • Moral Foundations research. The research ( on this is pretty clear. People who self-identify as liberals tend to value fairness across groups as opposed to some notion of proportionality (fairness is you get what you deserve, which is what self-identified conservatives value). Self-identified conservatives value loyalty, authority, and sanctity more than liberals do.

Racism essentializes socially constructed taxonomies of ingroup/outgroup by relying on perceptions of salient characteristics of groups that we decide are true because of confirmation bias, and assumes those groups are in a zero/sum relationship as far as social/cultural goods; it therefore rejects any notion of fairness across groups.

Racism is a pernicious and toxic example of a relatively common phenomenon—that people have a tendency to categorize themselves and others in terms of groups, and to engage in “in-group favoritism.” It isn’t the consequence, let alone a necessary consequence, of that way of thinking—but a very nasty version of it. After all, while the tendency to think in terms of ingroups and outgroups is universal, making those groups racial categories isn’t—race is a relatively recent concept, not fully formed until the 18th or even 19th centuries, and different from other ways of thinking about groups in important ways.

In fact, “races” are socially constructed—what has counted as “white” has changed considerably, even in the last 100 years. In 1916, very powerful book was published that argued that there were three white races—not all whites, according to Madison Grant, were the same race (and race-mixing within whites was damaging to civilization). In the 19th century, racialist science sometimes claimed there were four races, sometimes three—that inability to agree is interesting, since racists claim that “race” is eternal and an obviously physiological category. If it’s obvious, why can’t they agree? In fact, even scientists promoting the notion of race as a scientific category couldn’t come up with a consistent definition of “race,” let alone one that fit the evidence they had.

Even in the early twentieth century, there were scientists who pointed out that the way eugenicists talked about race didn’t make any sense. They showed that people advocating racial purity used “race” in two very different meanings ways—there were the socially constructed categories, which were based on linguistic, political, and national boundaries (such as the notion of an “Irish” race); there were also possibilities of biological categories (such as Celtic) but 1) given the long history of human interaction, they were not discrete categories; and 2) those biological categories had nothing to do with the socially-constructed categories. Biologically, there was no “Irish” or “Italian” race, just nationalities. So, the notion that immigration quotas were backed by genetic data made no sense.

Racism doesn’t necessitate that people consciously think that their race is superior to others; it only requires ingroup favoritism (the unconscious tendency to perceive a situation as “fair” if our group gets slightly more).

It tends to get more antagonistic if we think our group and another (or several others, or all others) are in a “zero-sum” relationship. If we’re having a pie potluck, and each person who comes will bring more pie than he can eat, then more people who come means more pie for everyone. But, if we have a single pie and no one is bringing more, then the more people who come, the less pie there is for YOU. In that second circumstance, the more someone else gets, the less you get—someone’s gain is necessarily someone else’s loss. (There’s a complicated, but good, explanation of how it works in economics here:

There are some people who perceive every situation as zero-sum, even when it isn’t. Hyper-competitive people will often feel as though praise for someone else’s accomplishment injures them (as though it takes something away from their accomplishments), and so they can sincerely believe that the best response to someone else achieving something is to try to criticize or undermine that person. One of the characteristics of a toxic relationship is that one or both people believe that they are threatened by the other person being successful, and most of the bad behavior of bridezillas can be attributed to a sense that any attention paid to anyone else is taking something away from the bride to which she is entitled.

While not all racism requires zero-sum thinking, it is interesting that a lot of racism is justified through sloppy social Darwinism (that human interactions are inherently a contest which result in the survival of the fittest—not actually a Darwinian concept at all). White people (that is, people who think of themselves as “white”) who think about all social interactions as zero-sum competitions have trouble seeing the problem with Nancy Meyers’ movies having so few non-whites, and to condemn as “political correctness” any movie more diverse. It isn’t necessarily that they think to themselves that whites are superior, but they are likely to believe that it is “normal” to be more interested in white people (white people are “naturally” entitled to more attention, white people problems are more “interesting” or “universal” than the problems of non-whites, which are “particular” to those groups).

People drawn to racist explanations and assumptions tend not to be very good at perspective-shifting (they also get really uncomfortable with ambiguity and complexity). They not only see things only from their perspective, but the more racist they are, the less able they are to acknowledge that there are any other perspectives that might be legitimate. They might believe that everyone who disagrees is just pretending (but secretly agrees), or they might believe everyone who disagrees is an idiot, or they might believe that their way of looking at things is unbiased and universal and other ways are biased and particular. So, someone who looked at things that way would have trouble being interested in literature or films about people not like them—such art requires perspective-shifting, and they’re bad at it. They would be likely to think of their reaction as the one everyone like them has, so, if a white director made movies with a diverse cast, they would see that as deliberately pandering to people who don’t really have a valid perspective (wanting more diverse art).

But, again, they wouldn’t experience themselves as feeling any antagonism (except to “political correctness”), nor would they be aware of some sense of their race being superior—they would just want to focus on “people like them.”

People who tend to be racist tend to be drawn to thinking in black/white terms in general (they have trouble thinking in terms of continua, matrices, or probabilities). In fact, they often actively angry if you tell them a situation is complicated, as they think you’re being indecisive. They tend to believe that groups are meaningful (you can get most of what you need to know about someone by deductions that you can make from their group membership); they tend to think in terms of identities and motives (they essentialize people); they tend to dislike new music, new food, new genres (of literature, movies, TV), and new places (they get agitated by difference). They tend to be naïve realists. They have trouble admitting error, let alone learning from mistakes.

One more point about outgroups. It seems to me common for people to have two kinds of outgroups—a group that is threatening because it is cunning and scheming, and another that is threatening because it is animalistic. The first group is often represented as controlling the animalistic group, and the second is often thought of as in a binary of either submissive (like a domesticated pet) or in rebellion. When racists talking about that first group as “cunning” people can think it isn’t racism, since it seems to be a compliment; when racists talk about that second group as childlike and happy, people can also miss the racism (since they seem to be saying “nice” things about the group). Sometimes racists will praise submissive members of the outgroup, as though that shows they aren’t racist—but they’re only praising members who “know their place.” It’s still racism.

As you’ll see in the class, there is a lot of disagreement as to whether racism is a new phenomenon—some people categorize it as a kind of hostility to outgroup that is inherent to the human condition. I’m dubious, since other kinds of hostility to outgroups allowed conversion or assimilation. Because racism “naturalizes” the differences (that is, puts them into nature) there is no possibility of being treated as equal to the ingroup—the Other will always be Other.

7 thoughts on “How to argue about whether something is racist”

  1. This is the first thing I’ve read that has a decent starter discusision about racism. Can you go further and define the term race? That would really help for me and my students.

    1. This might sound funny, but I don’t define race. It’s really what Burke calls an “ultimate term,” meaning that it’s all connotation and no denotation. Even racialist scientists never managed to define the term consistently (either consistent with the definitions of other “scientists” or consistent with how they use the term in their own writings). Legal definitions generally depended on common usage, while acknowledging that usage was muddled and inconsistent.

      It’s one of the important things to point out about racist discourse–that “race” is a social construct, and politically shifting.

      (Sorry to have taken so long to reply–I just saw this comment now.)

Leave a Reply

Your email address will not be published. Required fields are marked *