The Spanish Inquisition

Westermarck came up with it first..

Anything that can ever be said about moral relativity has probably already been anticipated by Westermarck.

Two volumes of The Origin and Development of Moral Ideas have:

  1. All the anthropological and ethnographical evidence you'll ever need to argue that moral norms vary across culture. Nothing upto 1905 is left out.

  2. An early argument for Haidt's social intuitionism, stating that all “morality” stems from moral emotion rather than reason.

  3. Repudiations of common arguments against moral relativity, such as relativity being overly permissive of “dangerous” notions.

    Ethical subjectivism is commonly held to be a dangerous doctrine, destructive to morality, opening the door to all sorts of libertinism. ... This inference was long ago drawn from the teaching of the Sophists and it will no doubt be still repeated as an argument against any theorist who dares to assert that nothing can be said to be truly right or wrong.

    Far from being a danger, ethical subjectivism seems to me more likely to be an acquisition for moral practice. Could it be brought home to people that there is no absolute standard in morality, they would perhaps be somewhat more tolerant in their judgments, and more apt to listen to the voice of reason.

  4. An explanation for why a “clearly subjective” experience is treated as having objective roots by appealing to language and history.

Westermarck is mesmerizing and original, yet nobody but anthropologists seem to recall him.

Moral relativity says that if cultural variations in morality exist, then objective morality does not exist.

This seems easy. If you look for a black cat, and find no black cat exists, then you are justified in believing no black cat is actually there.

Yet everybody seems to find flaws with this:

  1. People say you can have variations in moral rules, but not moral principles – that is, all moral rules are simply expressions of natural moral desires, like not wanting to harm people, thus making moral relativism a shallow explanation.

  2. People say moral relativism is permissive of harm, making moral relativism indefensible from a “saving the world” standpoint.

  3. People say that moral relativism doesn't permit enquiry into whether some belief systems have better reasons for holding beliefs than others – moral relativism is a descriptive, not a prescriptive treatment.

You just can't seem to please people! But the latter objections have merit in a practical sense: people look to ethicists to lay out guidelines for minimizing harm, and leave the business of the “truth” of ethics to philosophers. Applied ethicists can't use moral relativism, and so it languishes in the sidelines while effective altruism and co. attract attention.

When do you legitimately get to call someone racist anymore?

For example, can you be racist even if you don't intend to be? Can you be racist if you claim to “not see color”? Can you be racist if you unconsciously associate certain groups with stereotypes, but aren't aware of it? Can you be racist simply by virtue of belonging to a historically advantaged group? Can you be racist if you cite evidence indicating that there are differences between the races, even if you agree with racial equality?

In my personal life, as a liberal surrounded by the progressive American left, I'm often surprised by the sheer diversity of responses I get to these questions. All of the above questions are “racist” to some, yet “not racist” to others – often, people who answer one or the other clump heavily in terms of background and network.

When pressed, the criteria people use to qualify racism is often inconsistent – people disagree about what unconscious racism looks like, whether racism is individual or systematic, whether members of minorities that have experienced overt racism are better able to assess when something is racist than other participants, or even whether acts have to directly implicate race to be racist. This inconsistency not only muddles discussion and confuses people – it also leads to policy recommendations that spike controversy.

If you try to read the literature, the history of sociological research into racism offers many perspectives, but no concrete answers either. The term racism has undergone shifts in meaning, leading to popular confusion and more than one Internet argument. Moreover, the focus of academic inquiry into racism has bloomed into many disparate subtopics, introducing more terms (racialization, implicit bias, institutional racism) that all need to be kept track of to really address this question.

This post is part-research dump and part-essay. As someone who is decidedly not an expert on these issues, and as someone who prefers clear consistent criteria to understand when an accusation of racism is justified, this post serves as a sticky note for me to link to references. Yet, at the same time, the nagging conclusion I have been forced to make from this research is that the term “racism” itself is effectively a new word – one with a new meaning that cannot inherit any of its previous baggage. New standards of discourse need to be communicated out to ensure we're all on the same page.

This post summarizes why I'm led to that conclusion, as well as other findings that should help reduce conflict amongst allies on this question.


Merriam-Webster defines racism as “a belief that race is the primary determinant of human traits and capacities and that racial differences produce an inherent superiority of a particular race”.

This is the same view advocated by sociologist W. J. Wilson and Matthew Clair et. al, and is the standard definition on Wikipedia:

Racism is the belief in the superiority of one race over another. It may also include prejudice, discrimination, or antagonism directed against someone because they are of a different race or ethnicity, or the belief that members of different races or ethnicities should be treated differently.

The International Council on Human Rights Protection advocates the following criteria instead of an outright definition:

Racism thus has three elements: (i) it is a vision of society that is composed of inherently different groups; (ii) it includes an explicit or implicit belief that these different groups are unequal by nature – often enough based on a Darwinian interpretation of history; and (iii) it shapes and manipulates these ideas into a programme of political action. Combined, these three components give racism its force. (International Council on Human Rights Protection, 2000: 4–5)

The law also invokes concepts of institutional racism (defined in the next section) in labour law.

The Academic Standard

Steve Garner, in his exhaustive and thorough book Racisms: An Introduction, argues that the popular definition is obsolete, and has been since the 1960's:

Moreover, according to George Frederickson (1988: 189), the popular idea that racism comprised a set of beliefs of biological superiority, has gradually been replaced since the Second World War by ‘patterns of action which serve to create or preserve unequal relationships between racial groups’. This new understanding of the term is concomitant with the development of a so-called ‘new racism’.

We'll consider this the academic standard for the moment, but it's important to note that Garner doesn't stop there. He looks at several other definitions of racism that have appeared in the literature, only a few of which are reproduced here:

  • The “collective failure of an organisation to provide an appropriate and professional service to people because of their colour, culture or ethnic origin” (The Stephen Lawrence Enquiry [MacPherson Report], 1999: para. 6.34).

This can be understood as institutional racism.

  • “The attribution of social significance (meaning) to particular patterns of phenotypical and/or genetic difference” (M. Banton, 1996, ‘Racism’, in E. Cashmore (ed.) Dictionary of Race and Ethnic Relations, 4th edn. London: Routledge.).

This is, for want of a better word, “scientific” racism, in that one tries to explain social differences between racial groups in terms of biological factors.

These definitions strongly color the academic standard by adding significant context. While compacting both nicely into one sentence is difficult, the academic standard should be understood to allow for both under its umbrella.

So Which Should You Use?

If you're paying attention, you'll see right away that these definitions all have some drastic differences:

  1. The academic definition expands the focus from individual beliefs in the popular definition to include social and structural practice, history and philosophy. Racism ceases to be just a matter of overt interaction; it becomes rooted in systemic interaction, viewing specific interactions as consequences of larger forces.

  2. The academic definition includes outcomes for different ethnic groups as a component of its definition in addition to belief. This is opposed to the popular definition and the legal standard, where belief and/or intent are the sole factors. To quote Garner, a failure to do something to correct structural inequity between races is also racism under this definition.

At the same time, these definitions all have some things in common:

  1. None of these definitions directly address whether absence of intent is relevant to racism. Specifically, they don't touch on whether unconscious bias, such as in implicit association tests, constitutes racism. Whether intent should be factored into racism appears context-sensitive – for example, in this whitepaper by the Kirwan Institute, the institute calls out “institutional racism” as operating without racist intent.

  2. All of these definitions require persisting a racial hierarchy. The hierarchy can be ideological – you can argue some racial groups are superior than others, and act on that – or actual – structural inequalities that allow select racial groups to exercise power over other groups – but the hierarchy itself must be supported.

Because these definitions differ widely in prerequisites and scope, it's important to recognize that they technically don't describe the same things. The academic definition in particular is so sweeping that it is more appropriate to call it by a new term altogether, rather than re-use a term with so much history and connotation. I am not alone in this view – Michael Banton argues that 'racist' should only be used to describe actions, since “racism” can be weaponized in political contexts and thus add little to the discussion, and Garner advocates usage of the term “racisms” instead of “racism” to denote the plurality of racism.

Unfortunately, as of 2019, it doesn't seem like such pragmatism is like to catch on. There is already a sizable body of existing work that evolves the term “racism”, and it's a natural consequence of larger ideas to want to redefine terms to be more inclusive this way. Thus, for the moment, we're stuck with having to revise our mental dictionaries to engage in discussions with our peers, or insist on striking out on our own for the sake of having clear boundaries.

A satisfying middle ground is to identify a set of minimum criteria that bridges all definitions. This enables us to eschew any particular definition, but allows us to engage in dialogue inclusive of all.

Once again, Garner has already done our work for us by suggesting three such criteria:

  1. A historical power relationship in which, over time, groups are racialised (that is, treated as if specific characteristics were natural and innate to each member of the group).

  2. A set of ideas (ideology) in which the human race is divisible into distinct ‘races’, each with specific natural characteristics.

  3. Forms of discrimination flowing from this (practices) ranging from denial of access to resources through to mass murder.

The limiting power of this approach is that these criteria are so general that they do not allow us to make specific conclusions in all situations. For example, none of them would allow us to answer the “is it racist” questions this article began with.

However, I hope it is understood now that this is a false dichotomy – because the term “racism” itself has been evolved, one has to abandon all the connotations that the word carries from its previous incarnation. It's perfectly fair to replace “racist” with “zinkydoink” in all of the above questions. To be called racist in modern parlance is to be called something that should be interpreted as having no baggage attached – a conclusion that allows sentiments such as “most white people are racists” to be simultaneously true and entirely without tension.

(On a personal note, I should add that this particular type of sentiment is hard for me to accept, precisely because of the connotations I've built around the term “racism”. Language change is an undeniable aspect of our existence, yet it appears I am not yet quite ready to evolve with it – a sure sign that I'm getting older before my time.).

It's worth asking what benefit adopting any definition has before picking any. This Medium article, also linked above, makes the simple statement that the right definition to use depends on what the end goal of the conversation is:

  1. If you want to ensure racial harmony, use the popular definition
  2. If you want to ensure racial equity, use the academic definition

These are good guidelines to follow.

The New Lexicon: Terms Related to, but Distinct From, Racism

The academic definition above is an umbrella term. Because it includes attitudes, actions, systems, etc. it's worth exploring its ecosystem to understand the nuances.

New Racisms

The academic definition, as Clair et al. aptly summarize, was driven by a puzzle. Overtly racist attitudes, cited as the primary reason behind racial inequalities in Western countries, declined in the period following the sixties – yet racial inequality persisted. Sociologists struggled to answer the question, and largely took on a few separate approaches:

  1. That historically oppressed groups have had cultural problems preventing them from overcoming the impacts of racism. This has been a controversial opinion, even directly leading to the coining of the phrase “victim blaming”.

  2. That overtly racist attitudes continued to persist, but had changed character to continue discrimination while being superficially in favour of racial inequality. These “new” types of racist attitudes have been christened the “new racisms”, and it's worth examining them below.

It's important to note that the “new racisms” are all fundamentally attitudes. They are characterized by the following properties:

  1. They are subtle or implicit rather than overt. In other words, their racist character is often not obvious to the person perpetrating the attitude.

  2. They are highly context-sensitive. The same person holding these attitudes may not actually end up applying them in different contexts.

Examples of new racism include symbolic racism, modern racism and aversive racism. They include color-blind racism (not factoring race into decision-making leading to inadvertent discrimination against less-advantaged groups) and laissez-faire racism (the argument that minorities are responsible for their own poor socioeconomic advancement). I have been told, but cannot confirm, that an exhaustive list can be found in Lincoln Quillian's 2006 review.

A good rule of thumb is that most of the “new racisms” ultimately lead to people verbalizing policy positions that do not help minorities advance. Aversive racism is the one exception to this rule, being much closer to unconscious bias than subtle racism – yet it too falls under this class by virtue of being verbalizable.

It has been argued that some “new racisms” are indistinguishable from conservative policy positions. Nevertheless, they meet the criteria upheld by the academic definition, in that they persist disadvantages faced by minorities.

Racial Discrimination

Racial discrimination is a term that is technically distinct from racism. Specifically, it means to treat people differently on account of race. It need have nothing to do with racism per se – treating someone differently on account of race need have nothing to do with perpetuating racial inequality (per the academic definition) or believing in a racial hierarchy (per the academic definition).

Discrimination can come in many forms – for example, stereotyping community members is a great example.


In practice and in theory, I think there are a couple of major takeaways from this very quick sampling of the literature.

One is that “racism” has none of the connotations of the past. “Racism” as a word is very removed from the definitions of yesteryear, eschewing its focus on the individual or even his or her conscious beliefs. It is best to substitute a euphemism for it if you are having trouble adjusting to its modern connotations. If at all possible, go to pains to differentiate between the popular notion of racism and the contemporary academic definition of it.

Another is that not enabling socioeconomic equity is an implicit factor in contemporary academic definitions of racism. Any attitude, action, or belief system that does not address inequity, if perpetuated or held by groups that have held power over the other, including inaction, can be interpreted as racism. This is the most significant leap compared to the popular and legal definition (on a personal note, it is also my biggest grief with the definition, since it's such an ad-hoc requirement latched on to the definition).

Another is that unconscious/implicit bias does not meet the definition of racism. All the definitions cited above appear to rely on intent, which implicit bias does not possess, and implicit bias is treated as a separate field of study in all the review papers I can find. Subtle racism, which is “unconscious” in that its character is not immediately recognizable as perpetuating inequity but can be verbalized as specific policy positions, is not unconscious or implicit bias, and thus meets the definition of academic racism.

From all of this, we can see that call someone or something “racist” is ultimately either an exercise in tautology or a frustrating maze of labyrinthine partially-overlapping definitions that need to be teased out. It is much, much more productive to not use the word altogether – instead, identifying upfront whether the end goal of the discussion is to ensure racial harmony or to ensure racial equity will let you pick the right definition.