Definitions, part 1 (racism)

What is “racism”? This question is really at the heart of much of the current debate on race in the US. My take on it is that the preexistent definition(s) of “racism” gave rise to the massively negative connotation associated with the word, and then activists redefined the word, hoping to change the denotation while retaining the attached connotation. In this way, they hoped to attach opprobrium to something that, regarded in isolation from the connotation, might otherwise be regarded as inevitable or even benign.

TL;DR

From the time the term came into common use, “racism” has had a definition that was widely agreed-upon in broad terms. The concept to which the word refers is now the subject of near-universal opprobrium. Ibram X. Kendi and Robin DiAngelo, in their controversial but influential bodies of work, revise the definition of “racism” in a couple of different ways.

DiAngelo defines “racism” so that only White people are capable of racism, and so that it requires no real action on the part of White people to be racist. She then posits the notion of “White Fragility” in such a way that if White people react at all to being called racist (one of the worst epithets in society at the moment), then it proves that she is right and they are, in fact, racist. This is a well-understood rhetorical device called a Kafka Trap.

Kendi defines “racism” so that it occurs systemically whenever there are any statistically cognizable disparities between races. Since the presence of individual freedom and cultural differences guarantees that these disparities will exist at some level, this means that “racism” in his sense is ubiquitous and permanent. He then adds the feature that anyone who disagrees with his definitions or (tyrannical) policy prescriptions (which include a totalitarian Department of Antiracism and expansive racial discrimination for the foreseeable future) is personally racist (also a Kafka Trap).

Under the Classic Definition, racism is morally repugnant. Under the DiAngelo New Definition, racism is more-or-less unavoidable (and her proposed actions to mitigate racism are themselves racist under the Classic Definition). Under the Kendi New Definition, racism just means disagreeing with Kendi (whose “anti-racist” proposals are also racist using the Classic Definition).

If racial progress is to continue in the US, we simply can’t allow the New Definition denotations to stand, while retaining the Classic Definition connotation of moral repugnance. Civil discussion and debate demand the rejection of these redefinitions.

Classic Racism Definitions

Here are three definitions that roughly encapsulate the state of play among people of approximately my age –

Dictionary (Webster’s Third New International, 1971): “the assumption that psychocultural traits and capacities are determined by biological race and that races differ decisively from one another, which is usually coupled with a belief in the inherent superiority of a particular race and its right to domination over others.” There are also variant definitions, dealing with political/social systems founded on this idea.

Victor Davis Hanson: “Racism is the deductive bias against, and often hatred of, an entire racial group. It is often birthed by dislike of particular individuals of a given group that supposedly justifies, by extension, disliking or indeed hating all of them.”

My own definition (slightly different, though I wouldn’t wish to quibble too much with Dr. Hanson): a “racist act” is bad treatment of an individual because of perceived biological group membership. “Racism” is what motivates someone to do or desire to do a racist act.

For historical clarity, the first widespread use of the word “racism” is around 1928, so that the Webster definition of the term is about halfway between the inception of the word and today (the time of the other two definitions).

The first (dictionary) definition is a bit different from the other two (note that it comes from an age before racism had really acquired the full horrific connotation it has today) in that it has two parts: the first part is wrong (a false assumption), but not evil (no one is harmed by the indicated mistaken belief) – it is only when coupled with the second half, and, even then, only when acted on that it causes harm. In what follows, I will be referring to the full dictionary definition (“coupled with…”) since that more closely corresponds to the other two.

All of these have several elements in common:

  • They all describe individual flaws, not group flaws. It certainly might be the case that there is a group consisting entirely of racists, but they are racist or not individually; the group may consist of the guilty, but the group is not guilty itself.
  • They do not reference power. The powerless may be seized as much by racism as the powerful, even though less able to act on it and cause harm to others.
  • The concepts defined are not necessarily detectable by an outside observer, though persistent conversation would likely reveal it.
  • All require animus, not simply ignorance. A belief that all Black men are good at basketball is ignorant, but not racist by these definitions (unless you also happen to dislike basketball players for some reason).
  • All are rooted in the “sample size of one” or “arguing from whole to part” statistical fallacy – knowing any measure of central tendency of a group (even if accurate) tells you precisely nothing about any member of that group (or even the average of a nonrandom sample).
  • All are also rooted in the “arguing from sample to population” fallacy – knowing anything about an individual member (or even a nonrandom sample) of a group tells you nothing about the group.

These last two fallacies together provide what I think is the bizarre, fallacious core of racism: an individual is disliked for some (possibly valid) reason; this dislike is then applied (fallaciously) to an entire group to which the individual belongs; in turn, this dislike of the entire group is then applied (fallaciously) to another individual belonging to the same group.*

There are at most minor differences in these definitions. From here on, I will be referring to either Hanson’s definition or mine interchangeably as the “Classic Definition” of racism. The two are close enough that it should make no difference. I mainly included the dictionary definition so as to make clear whence we have come – the Webster definition was essentially the definition that Dr. Hanson and I both learned in school (he and I are fairly close to the same age). As you can see, we have both refined our understanding of the subject, but there is clear continuity between the definition of our youth and our definitions today.

New Racism Definitions

Robin DiAngelo: Racism is “White racial and cultural prejudice and discrimination, supported by institutional power and authority, used to the advantage of Whites and the disadvantage of people of Color.”

Ibram X. Kendi: “A ‘racist policy’ is any measure that produces or sustains racial inequity between groups. “

“‘Racial inequity’ occurs when two or more racial groups are not standing on approximately equal footing…Every policy in every institution in every community in every nation is producing or sustaining either racial inequity or equity between racial groups.”

By frequent example in his writing, the word “approximately” in the preceding definition has no real force: any inequality of outcome is prima facie evidence of inequity for him. He has also said, “As an anti-racist, when I see racial disparities, I see racism.”

One more definition: “the opposite of ‘racist’ isn’t ‘not racist.’ The true opposite of ‘racist’ is ‘anti-racist.'” Then, ‘anti-racism’ is further defined as support of his policy prescriptions for leveling all racial disparities.

I should also mention that this pairing (racist policy/racial inequity) is one of his more lucid and helpful definitions. He has also stated, in writing, that racism is “a collection of racist policies that lead to racial inequity that are substantiated by racist ideas.” It’s hard to know where to start with that one, as it’s pretty well impervious to comprehension.

Kendi’s proposed solution(s): “the remedy for past discrimination is present discrimination; the remedy for present discrimination is future discrimination.” More specifically, he has proposed a Department of Antiracism, empowered by Constitutional Amendment, to “Preclear all local, state and federal public policies to ensure they won’t yield racial inequity, monitor those policies, investigate private racist policies when racial inequity surfaces, and monitor public officials for expressions of racist ideas. The DOA would be empowered with disciplinary tools to wield over and against policymakers and public officials who do not voluntarily change their racist policy and ideas.”

Compare and Contrast: DiAngelo

Where to begin? Let’s start with Dr. DiAngelo, since her redefinition has the smallest edit distance from the original.

As is fairly frequently done these days, she adds the “power and authority” clause. Others have more concisely redefined racism as “prejudice plus power.” For my part, I don’t mind having a word referring to such a concept, but not at the expense of overloading a word that has a pre-existing definition – again, the Classic Definitions do not have “power” as a qualification – power enables racism to be more harmful, but racism can certainly exist without it.

DiAngelo goes further, though, by making it abundantly clear that, for her, racism is a Whites-only club. So, for example, the treatment of White farmers in Zimbabwe around the turn of the millennium (though it was done by those in power) couldn’t possibly be racist because Mugabe was not White. Again, for those of us of a certain age, a redefinition of terms. How handy, though, to have a term of opprobrium that by definition only applies to the people you particularly wish to derogate!

In addition, DiAngelo’s definition requires no specific action or attitude on the part of a White person to be racist (and we will see this again in Kendi’s work, though he does allow Whites to “opt out” of racism by agreeing with him and working for his preferred policies). Merely existing as a White person benefited in some ineffable way by racist institutions is sufficient to make one a racist.

Here’s the difficulty with making race-specific definitions for racism: at this point, pretty much everyone wants to eliminate the last vestiges of racism from society. With the Classic Definition of racism, this can be done in ways that actually help people get along and work together better, ways that support human flourishing of all groups (more on this later). With DiAngelo’s definition of racism, though, the solution is entirely “make the White people to be less White” or possibly “get rid of the White people and give their jobs to people of Color.” Since White people can’t completely rid themselves of their Whiteness (and thus their racism), they can only ramp up their self-loathing or get out of the way. It’s very difficult for me to see how anyone is benefited by this sort of hostile work/classroom/social environment.

It’s obvious how Whites are intentionally and explicitly harmed by this (and thus, how her proposals are themselves racist using the Classic Definition). Less obvious (but still considerable) is the harm to People of Color. For a thorough discussion of this, see John McWhorter’s review of DiAngelo’s work.

I should also mention that DiAngelo’s signature work, White Fragility, (as well as, reportedly, her in-person lectures) is in essence a book-length Kafka Trap. This rhetorical device (which I believe was actually used earlier by Freud) is an unfalsifiable argument that can be adduced in discourse about human subjects. Essentially, the Kafka trapper makes a pejorative statement about some category of people, then adds that a characteristic of that category is that they disagree with the statement just made (“you White people are always alcoholics AND you deny that this is true”). The trappee then either agrees with the statement (in which case the trapper is right) or they disagree with the statement (in which case the trapper is also right but for different reasons, and the trappee has been self-outed as being in the pejorative category). To disagree with Dr. DiAngelo is to exhibit White Fragility and thus prove her correct.

Should you find yourself in conversation at this point with Dr. DiAngelo (or a similar Kafka trapper), I suggest one of the following responses:

  • “You are a maker of false accusations, and such people always deny it.” (not original with me).
  • Wholehearted laughter – she predicts defensiveness on the part of White people, not amusement and mockery.

Compare and Contrast: Kendi

We move on now to Dr. Kendi, who makes more extensive and inscrutable revisions to the definition of racism.

It’s a bit difficult to really make sense of what he means in his definitions – I think he’s intentionally unclear – but, by frequent example, he seems to mean that systemic racism is anything that produces a statistically cognizable difference between racial groups (I hesitate to say “significant difference” since that is actually a statistical term of art, and I’m unconvinced of Dr. Kendi’s depth of statistical understanding). More precisely, he seems to mean that the existence of any socially significant sample group with racial demographics that differ from those of the population as a whole is evidence of racism in the social system related to the selection of the sample.

Then, again by frequent example, he seems to mean that personal racism (which he sometimes calls, “holding racist ideas”) consists of disagreeing with him about either his definition of systemic racism or any of his “anti-racist” policy prescriptions.

This is truly unrecognizable to anyone who holds to any approximation of the Classic Definition of racism.

  • Classic racism is an individual flaw; Kendi’s racism is primarily a systemic characteristic, and secondarily an individual flaw (that seems to consist entirely of disagreeing with him).
  • Classic racism is not necessarily detectable to an outside observer; Kendi’s racism is only detectable to an outside observer – many racists aren’t aware they are racists until they read his work on the subject.
  • Classic racism requires animus against a group which is then directed against an individual; Kendi’s racism requires no animus, only failure to be sufficiently “anti-racist” (as he defines it).
  • Classic racism is based on the fundamental statistical fallacies described above; Kendi’s “anti-racism” is based on those same fallacies (plus a few others). Though he has not articulated this verbatim, it would appear that a reasonable grasp of statistical reasoning would be “racist” in Kendi’s view, since it would lead one to deny some of his favorite examples of racism; he has said (video link, no transcript available that I can find) that “to be racist is to constantly, consistently deny, deny, deny.”

To be fair, there are a few areas in which Kendi’s definition is closer to the Classic Definition than DiAngelo’s. Specifically, he doesn’t require power in order to be racist. He also says that racism is something you do (or fail to do), not something you are; for him, White people are not inherently racist. He also doesn’t support hatred of White people (though some of his adherents certainly do).

On the other hand, his racial disparity notion is FAR too broad and simplistic. I will be discussing some of his statistical shortcomings here – suffice for now to say that there are many possible reasons for statistical racial disparities, with racism (Classic Definition) being only one. Also, see John McWhorter’s book review here for a thorough, careful review. Incidentally, this review led Kendi to charge (predictably, and in remarkable bad faith) that McWhorter is racist (an episode discussed here). All this, despite the fact that he said in the review that he thought Kendi’s work was considerably better than DiAngelo’s.

Kendi’s definitions, then, lead to the use of “racist,” a word with overwhelmingly negative connotations, to be used for someone who has simply disagreed with Kendi about either the cause or the prescription for statistical racial disparities. Once he’s made this set of definitions, any particular prescription stands as a Kafka Trap – you either agree (in which case he’s correct) or you disagree (in which case you’re a racist by definition).

Stunningly, these prescriptions include many proposals (including his egregiously totalitarian Department of Antiracism suggestion) that are explicitly based in racial discrimination. Racial discrimination, under the Classic Definition, is racism, but under the Kendi Definition, it’s racist to oppose it. When the definitions are changed to this degree, debate and discussion are simply impossible. Kendi’s definitions must be rejected completely in order to even discuss his policy proposals in a sane manner.


Addendum: after writing this essay, an event occurred illustrating the key point clearly. Credible reports claim that the Anti-Defamation League (now satirically referred to as the “Anti-Definition League”) changed their website definition of “racism” from “Racism is the belief that a particular race is superior or inferior to another, that a person’s social and moral traits are predetermined by his or her inborn biological characteristics. Racial separatism is the belief, most of the time based on racism, that different races should remain segregated and apart from one another” to “[Racism is] the marginalization and/or oppression of people of color based on a socially constructed racial hierarchy that privileges white people.” This is especially fascinating for the ADL to do this since the Holocaust was racist under the old definition but not the new one. Apparently the definition was subsequently changed to an “interim” definition. Who knows what it is now?


* It is worth remarking that this is also the source of the difference between what I have called “Woodrow Wilson racism” and “Teddy Roosevelt racism” – both of these men, based on their writing, seem to have believed that other races, as groups, were inferior to their own in some way. Thus, they likely fell prey to the first fallacy in generalizing from some personal experience with individuals to the whole group. However, Roosevelt appears to have pretty much taken people as he found them (no evidence that his opinions on race affected his treatment of individuals he encountered), while Wilson was, frankly, a virulent racist who resegregated the Civil Service during his presidency, reportedly so that he and his Southern cabinet members would not experience the shock of encountering black workers in the offices of the federal bureaucracy they oversaw. Thus, Wilson also fell prey to the second fallacy of applying his broad opinions about race to the actual individuals he encountered. Both forms are wrong, but the harm (at least the direct harm) of the “Wilson” form is clearly greater than that of the “Roosevelt” form.