Wikipedia:Race and ethnicity

This is an essay.

It contains the advice or opinions of one or more Wikipedia contributors. This page is not an encyclopedia article, nor is it one of Wikipedia's policies or guidelines, as it has not been thoroughly vetted by the community. Some essays represent widespread norms; others only represent minority viewpoints.

Shortcuts

This page in a nutshell: "Race" and "ethnicity" (and ancestry in general) are usually screwy concepts to use as classifiers of people on Wikipedia.
☣ Wikipedia is a bad place to engage in labelling that isn't absolutely integral to international public perception of the subject.

The overall notion of ancestry or heritage – so many meanings

"Ancestry" is essentially too meaningless an idea (or, rather, one with too many subjective meanings) for consistent use in Wikipedia articles and categories, under any particular terms including race and ethnicity, and specific names and labels that qualify as classification by ancestry. Using them is a poor idea, whether it's done with categories, infobox parameters, or biased language in article leads and sections. It frequently involves both non-neutral viewpoint and original research, based on assumptions instead of verifiability in reliable sources. Even when it can be verified, for many if not most biography subjects it is indiscriminate trivia. Such labels can also have very different meanings that can directly mislead readers (and editors for that matter).

When thinking in terms of ancestry, some people (on Wikipedia and "out there") mean "bloodlines", vaguely speaking, as if humans were selectively bred livestock (an idea which can get downright nasty). Way back when, this idea of ancestry originally meant the largely unreliable genealogies and family histories people passed down from one generation to the next, often infused with mystical notions of God[s]-given rights or talents – or absence thereof – based on linear descent (ideas which can be bizarrely arbitrary and anal, as well as systematically oppressive). Later, it also came to mean the generalization that census-takers recorded on the rolls and what people like doctors and midwives wrote down on birth certificates, after those were invented.

From a modern scientific standpoint, ancestry comes down to haplogroups. But haplogroups, it turns out, do not correspond to things like "Italian", or even "African" – neighboring groups in Africa often have more diversity in their genes than is found between the Welsh and the Japanese. From a more cultural viewpoint, ancestry- or heritage-based labels refer to one's familial and meta-familial background – that sense in which 4th generation Americans or Australians say "I'm Irish" or "I'm German and English", by which they mean "my purported ancestors from the old country were these things, and my grandparents and parents kept alive some attitudes, beliefs, traditions, associations, etc. that were, or were believed to be, inherited from the old country, like Great-Granny's Moravian gingerbread 😉 recipe which I use and will pass on to my kids." When does this relate to encyclopedic coverage? Not so often.

Various heritage-related categorizations can also imply membership in a social group, which may have a cause, a lingo, a mode of dress, statistically more likely occupations, etc. – which has nothing to do with genes, nor in most cases with anything we'd write about in someone's biographical article. And there may be other, more nuanced implications of a heritage-related label in particular cultural contexts, such as implied obligations to other members of that group.

The original-research trap

Wikipedia has no reason to dwell on these things, in and of themselves, or habitually label people with them. The encyclopedia only cares when reliable sources or the subjects themselves consistently make a point of it – when it becomes part of the encyclopedic story of who they are. This is not AnthropometryPedia, and it is not our job to measure people's "breeding stock" and decide "how much" they are this group versus that one, then put them into labeled boxes.

To take a high-profile example: When someone like Meghan Markle, now Duchess of Sussex, says she is white- and African-American, repeatedly and in reputably published interviews, this is a socio-political statement about identifying with subcultural groups (and their heritage, pride, troubles, etc.), not a statement about genetics. "Biracial" is also a politicized socio-cultural label; if Markle doesn't use it and few if any reliable sources do with regard to her, then it's forbidden original research for Wikipedia editors to do so.

Another case was Bernie Sanders's self-identification as Jewish (in the familial background sense) leading to rancorous dispute about whether to label him as a religious practitioner of Judaism without reliably sourced information about his religious practices (if any). This Wikipedian battleground is what led to the removal of the |religion= and |denomination= parameters from most biographical infoboxes other than those for religious leaders. One also can be religiously Jewish without significant or any Jewish ancestry; for example, Sammy Davis Jr., an African-American from a Christian background, identified publicly as "a Jew" (not just "Jewish") after his conversion to Judaism.

Race and ethnicity – even more complicated, and often worse than useless

Next, even ethnicity is basically meaningless as a label to stick on someone in an encyclopedia just because you think they qualify for it. The term is generally confused with the notion of race, which is likewise unhelpful. For example, the average Westerner thinks of everyone in the world with a noticeable element of stereotypical African visual phenotype as "black", which they tend to conceive of as a single "ethnicity" or "race", and will use those two words interchangeably (or may resort to a label like mixed or multi-ethnic, if someone looks "part something else", too). This is woefully imprecise, confused, and subjective, yet it's what's usually in our readers' heads. This idea is radically confusing and nonsensical to many non-Westerners and non-English-speakers from different cultural backgrounds that do not think about people in these terms.

To an ethnologist or cultural anthropologist, the Oromo of Ethiopia are an ethnicity (definitely not a race). This use of ethnicity or ethnic group is much, much more specific, and this use is the dominant one found in reliable sources about ethnicity. Most anthropologists (across all fields of that broad discipline, the study of humanity) don't think "race" is a legitimate, science-based concept for classifying humans, but rather a social construct – something that, say, sociologists and politicians use for making generalizations about various blobs of the population in countries like the US and the UK. For the few anthropologists who still use the term race at all, there are only three or four races (Negroid, Caucasoid, Mongoloid, and sometimes Australoid, with various subdivisions and comminglings); the term has no relationship to ethnicity. The idea that all Africans are an "ethnicity" is downright stupid to anyone with even one undergraduate year in anthropological studies; the idea that they're all one "race" is not much higher up the chain, except in the peculiar and circumscribed vernacular sense of the word, tied to the socio-economics of populations in industrialized Western societies like North America, the UK, and France, all built in part on the backs of slave labor during the colonial era. (More on this below.)

Some examples, and what to learn from them

In everyday language (in the US anyway, and picked up in a few other places) Hispanic is an ethnicity. Yet it isn't one by any consistent definition; people just choose to treat it "as-if" – to put the glove in the sock drawer, so to speak – because our categorization systems are faulty. It has several more-or-less synonyms: Latino/a (not quite to be confused with Latin American), and Chicano in some circles. It's often called la Raza ('the Race') in Spanish, a language which doesn't split the same hairs we do in English; raza is even used for domestic animal breeds, without the capital R.

People from various ethnicities (under tighter definitions of that word) simultaneously identify as Hispanic/Latino and as something(s) else. They range from Afro-Cubans that many Americans would call "black"; to Americans who are "3/4 white" by old-school "bloodline" reckoning, but do not self-identify as white-/Anglo-/Euro-American at all; to rural Chileans who are about 90% Indigenous; to Northern New Mexico families that go all the way back to the Conquistador era and in some cases mostly descended not just from Spaniards but from Conversos – Spanish Jewish families that survived the pre-Inquisition pogroms in Spain by conversion to Roman Catholicism. Does this sound like "an ethnicity" (singular) to you? Even English-speakers outside North America are often confused by what Hispanic or Latino/a are supposed to imply.

Use of a socio-political term like "Hispanic" comes down to some combination of subject self-identification, and reliably sourced descriptions of the subject. Wikipedia has no business applying such a label based on who or "what" someone's ancestors were, even their immediate ones (and even if those earlier individuals did thus self-identify). Doing so is, again, patently original research. And if you don't think Hispanic is a political concept, you're sadly mistaken; most readers who are not Latin Americans don't realize how political it is, because they've never read The Cosmic Race and The Labyrinth of Solitude.

And "white" as an ethnicity or a "race"? Nuts. Not even people who are "white" can agree on who that does and doesn't include. At its broadest, it includes not only entire cultures as distinct as the Irish and the Kyrgyz, it spans multiple language families, and the haplogroup data shows that eastern "white people" are more closely related to "non-white" neighboring groups than they are to far-western "white" people like the Irish, in many ways.

White people as a sociological grouping is another thing entirely, at least in certain contexts, like the socio-economics of Western societies, the history of imperial colonialism, and many other topics of great encyclopedic importance. But that has nothing to do with whether to label a composer or a biologist as ethnically white. If we wouldn't willy-nilly apply that particular pseudo-ethnicity, then why label some other biography subject as being "Hispanic" or "Afro-Canadian" or "part-Asian, part-Arab", unless the subject and most critical literature about the subject also use the label and make a big deal out of it?

It makes much more sense to simply say where people are from, and leave ethnicity, race, ancestry, heritage completely out of it – unless and until we have proof from reliable sources or the subject's own material that such an identification is integral to who they are and what they do/did – i.e., it's directly tied to their notability.

"Some readers will want to know" isn't sufficient. Some readers out there also want to know about favorite color, clothing sizes, sexual proclivities, or what books the subject reads, but Wikipedia doesn't indiscriminately catalogue biographical trivia.

A counter-example

So when might ethnicity be appropriately tied directly to the notability of the subject? Aside from very obvious examples like campaigners for racial tolerance and justice, or victims of hate crimes, take John Taylor, Baron Taylor of Warwick, for example. He's only the third person of Afro-Caribbean descent to enter the UK House of Lords. His entry campaign into British politics in 1992 was controversially marred by racist opposition from members of the local Conservative Party constituency association, and people were fired (sacked) over it. He's also a prominent supporter of the West Indian Senior Citizen's Association and the Sickle Cell Anemia Relief organization, tied directly to Afro-Caribbeans and people of African descent, respectively. Given both his personal identification with such causes and the fact that his history as a public figure directly involves racial tensions in a country well known for them, it's appropriate for our article to identify him as Afro-Caribbean as well as British in the text (though just as British in the infobox), and as an English person of Jamaican descent (more specifically) in our category system and in the article's "Early life" section.

By contrast, a BBC article cited in our article just calls him "Britain's first black Conservative peer". This may be true, but it seems pandering to racialism – unnecessarily divisive and pigeonholing. Unfortunately, Wikipedia isn't immune to this; Taylor is also in our Category:Black British politicians. Why do we only have such a vague category, rather than something more specific and informative, like "British politicians of Afro-Caribbean descent"? After all, there are also "black" British politicians of South African and other more direct African extraction. After centuries of geographical and cultural separation, their social background has very little in common with that of Afro-Caribbeans, other than the lingering effects of Euro-centric racism. It's a strangely arbitrary and misleading category, at least on its own, making an overgeneralization that may satisfy racialistic curiosity (and lumping habits), but which isn't encyclopedically informative. It's rather like having a category named "Hovercraft and space ships" just because they're "the same" in being able to move above the ground.

Even if we decide such a category is worth retaining for research purposes – as a navigation tool, not a label – on questions of racism and its (hopeful) decline in British politics, shouldn't this be a container category, with everyone in it categorized in actual ethno-cultural subcategories instead of in this overbroad pseudo-ethnicity?

The "race" illusion

Race looks compelling phenotypically (i.e., with one's eyeballs), because some people are pale, some are really dark, some are a bit ruddy and have epicanthic folds, and so on, and these traits tend to be inherited. But genetics doesn't work along "race" lines. Any population will absorb and express a gene that is locally adaptive, while ones that are locally mal-adaptive will eventually die out or at least become rare in the same group of people across a substantially similar environment over a long period of time.

The same or a superficially similar adaptation is highly likely to arise in very distinct populations under the same sorts of environmental pressure. For example, large populations of South Asians have developed dark skin approaching that of many Africans, for the same reason of increased ultraviolet protection in the intense sunlight near the equator. Meanwhile, Indigenous peoples of the same latitude range in the Americas didn't change all that much from their ancestral East Asian complexion, because they needed another 10,000 years or so of evolutionary pressure to reach that level of melanin. But you will find it among many populations in the equatorial Pacific to Australia (where people have lived for much longer than the Americas) – despite them being the oldest surviving human groups to have left Africa and the African gene pool behind.

People's perceptions and assumptions are often simply wrong. "African features" of the stereotypical sort that Americans and Europeans have in mind are not pan-African, but primarily West African. Many north and east Africans have thin lips and pointy noses. Various features typically associated with West Africans show up elsewhere; e.g. broad, flat noses are also common among the Indigenous populations of Australia and the Pacific, as well as much of China. "Race" on our minds makes us see things that aren't real, and not see what is real when it doesn't agree with our assumptions. (More on this later, after some history.)

Where it came from, what came before, and how it happened

Ethnicity (in the narrow, precise sense) is a modern word, but of course the idea has its roots in prehistory: every tribe knew it was distinct from other tribes, and they had names for themselves and each other. A lot of writing that survives from Classical Antiquity is devoted to cataloguing various peoples and their alleged natures. This is hardly a new obsession.

However, the race concept is comparatively recent. "Race" arose toward the beginning of the Early Modern period in the West, primarily as a way to justify slavery on a massive, proto-industrial scale, taking off in earnest in the mid-15th century, and getting much worse during the "Age of Discovery", i.e. of colonialism and far-ranging conquest. The West at this time had various barely-permeable castes (sometimes finely sub-divided): royalty, nobility/gentry, merchant-class, peasant. But these had nothing to do with what people later called race. (They were, however, usually closely tied with "breeding" and inheritance, and there was common belief in the innate "ancestral stock" superiority of the upper class.)

Until well after the Renaissance, there wasn't even a concept of "European" (culturally, politically, or ethnically), nor did people think in terms of things like skin tone very much. In the West, you had Christendom and then you had heathens and pagans of various kinds. That's how the world-view of those we now retroactively call "Europeans" actually worked. You could be an icy-pale Scotsman or a swarthy Sicilian, and you were "one of us" as long as you were Christian, in their mindset. (Never mind the internecine but frequently shifting wars and struggles in Europe between various kingdoms and principalities, and ensuing prejudices. When it came time for a Crusade, they mostly pulled together.) If you were a Muslim, you could be a pale one of Umayyad Spain, a tan Arabian Saracen, or a "Blackamoor", and you'd be equally the enemy in Christians' eyes; you didn't get a free pass by looking more similar to most Christians. Just ask the Ashkenazi Jews, who were well interbred with other Europeans and hard to tell apart by features, but oppressed for their lack of Christian faith. Modern readers generally do not understand the importance, the centrality, of the concept of Christendom (and the mandate to spread it, by force if necessary) in the Medieval and post-Medieval motivations for conquest and expansion, well into the Industrial Age. It was all-consuming, the number-two motivation for any of it (after economics, of course).

Long before that era, the Roman Empire (which at its peak extended well into North Africa, the Middle East, and West Asia – as did the Alexandrian Greek empire before it) didn't break people down into "this one's white and this one's black and this one's brown", either, nor "you're Italian and that one's Spanish"; you were Roman or you were something else (labeled by empire or kingdom, by tribe or city-state, or by vague region when specificity didn't seem to matter). Nation-states like Italy and Spain didn't exist, though being from Rome itself (and a few other key cities, later) conveyed a prestige, at least among the upper classes, over being from the provinces like Hispania or Britannia.

Earlier still, Alexander the Great's idea was to make everyone he could conquer become Greek (in the Macedonian idiom in particular); today we'd call it a culturally imperialistic mission – not ethnic, much less racial. He and his army didn't care at all what you looked like; they just meant to make you Alexandrian and pay them tribute.

China had a not-dissimilar history, encompassing multiple ethnicities in the broad sense, and several entire language families, being periodically ruled by wildly different ethnic groups as (if and when they could manage it) a single big empire, with mostly a single written language (after one was imposed at sword-point). The Han, Zhuang, Mongol, Manchu, Turkic, etc. emperors and khans had little care for what your eyes looked like or how flat your nose was. You were either a productive part of their imperial machine, or you were an enemy.

However, one place that did have an ancient caste system which seems to have been at least partially what we'd call ethnically defined, in modern terms, was India and its various empires (sometimes commingled with the Persian ones, depending on whose army was better fed in one war or another). Interestingly enough, one of Europe's main imports from India, from the Crusader era leading up to the colonial period, was sugar. After European forces dabbled in sugar production in the Levant during the Late Middle Ages, sugar and the money derived from it rapidly became the primary impetus for both the slave trade and much of European overseas colonization (also motivated by Christianization, the hunt for gold and silver, and a trade route to East Asia). It began with the Canary Islands in the Eastern Hemisphere and resulted almost immediately in a genocide. The Spanish, French, English, and Dutch colonial Caribbean did little but produce sugar with slave labor, from about 30 years after Columbus's arrival (his first voyage brought Canarian sugar cane) for several more centuries. Only after colonization of the New World began did Spain suddenly come up with its intensely race-oriented casta system, despite millennia of frequent exposure to Africans, Arabs, Semitic Phoenicians, and other peoples of non-European stock. The subsequent colonization of continental North America by Europeans immediately set about more slave-driven production (mostly of other crops but following the same model), and already rigidly defined along racial lines – a concept that barely had existed in European minds a century earlier.

The notions of race and of racial superiority and inferiority were invented in Europe, on the Indic model, to justify African slavery (and used subsequently to justify seizure and exploitation of most of the Western Hemisphere and large tracts of Africa and Asia, ironically including India). The connections between Europe, India, sugar, race becoming "a thing" in the West, industrialized inter-continental slavery, and colonialism is no coincidence, but a direct causal chain. It led ultimately to both modern religious-like belief in race as innate and immutable, and the problematic position of race and racialist (and racist) thinking in Western societies, especially those with a slave-trade and colonialism legacy and substantial populations of descendants of formerly enslaved people and of natives of former colonies. The modern social construct of race is a direct consequence of colonialist self-delusion, propaganda, and exploitation, not an idea grounded in biology.

If you've been wondering why the very idea of "race" is controversial, you now have some idea, though this is hardly the whole picture. The dire socio-economic consequences of racialistic thinking – of treating people differently based on meaningless visual distinctions and illusory categorizations – have been a major political debate since the 1950s, with renewed intensity in 2020. Races may not be biologically real, but race as a concept (the perception and assumption of "races") and the human rights, societal equity, and other effects it has are very real.

Genes don't work the way the average person thinks they do

The idea that races are innate, either in human biology or psychology, is soundly rejected in modern science. What are innate are phenotypic distinctions, both obvious (skin color, average facial features) and not-so-obvious (e.g. different anti-malarial defenses), plus human tendencies to categorize, and to prefer to associate with familiar-looking people while distrusting "the other". The phenotypic differences vary almost ridiculously widely, across most of a continent in some cases and just across a few tribes in others. They do not correspond in most cases to traditional "races", but just show a frequency distribution bubble in them. If you made up entirely new pseudo-races by totally arbitrary criteria, you'd find the same thing (just different phenotypes would be more prevalent in your new "races" than in our current ones). It's just how statistics operates.

It's very similar to common mystified wonder (among those with little mathematics education) at the fact that the digits in multiples of 9 always add up to 9 (9×9=81; 8+1=9; etc.). This seems unmistakably significant at first, like a wink from God/the gods/the Multiverse, but it's not. It's just a property of the last digit in any numbering system (e.g. of numeral F in hexadecimal); if you make a huge table, you end up seeing that there's nothing mysterious or special about it; it literally could not be any other way, for the same reason that moiré patterns are not magical, and neither are snowflakes or other crystals. There is no unanswered question in the hexagonal honeycombs that bees make, or in the fractal patterns of plant growth (amazing as it may look sometimes). It's all just nature working along fixed lines of mathematics and physics and chemistry.

Back in the human genetics sphere, there's nothing special about, say, the anti-malarial and sickle-cell anemia gene that is common in most African populations. It doesn't prove a "race" exists like some kind of Platonic Ideal; it's simply passed around as effective enough against the regional kind of malaria to provide a survival-to-reproductive-age advantage. Other genes that are less effective against the disease don't survive there because their host persons tend to die before producing children. If a large population of Africans with their anti-malaria gene moved into southern China, and the displaced Chinese took their place in Africa, genetic tests done on them after a dozen generations would show a bunch of Afro-Chinese in China who still looked much like their ancestors but who now had the southern Asian anti-malarial gene; and a bunch of Sino-Africans in Africa who still had epicanthic folds but possessed the sickle-cell gene instead of the Asian anti-malarial one. Genes do not travel between humans as "racial packages". A single advantageous gene can sweep through a population in only a few generations and become prevalent, and another mal-adaptive one can be mostly lost, with little effect on the rest of the population's genetics, including appearance.

If you manufactured a new racial definition that consisted of everyone native to the tropics worldwide, you'd find some big correlations, e.g.: dark skin, hair, and eyes; and a variety of genetic defenses against malaria. You might be fooled into thinking this meant something deeply significant, that it proved your erstwhile "Sunbeltian" race was real. It wouldn't. It would just illustrate that a consistent environmental pressure produces similar adaptations in discrete populations. This is not a hypothesis or a novel idea, but a thoroughly understood fact about evolution and population genetics.

The psychological factors

On to the other part: We humans categorize like mad (and that word was chosen on purpose). We have such an in-built facility for labeling and classifying that it frequently leads us to perceive patterns that do not really exist, to make assumptions without (and even against) evidence, to exaggerate or invent similarities, and to ignore differences that don't suit the model we've built from our flawed perception and our ingrained "put everything in neat boxes" tendency.

Humans' instinctive xenophobia (inherited from our primate ancestors) is triggered by anything, and has nothing in particular to do with "race". Repeatable, simple experiments prove this with nearly zero effort to prove it. A classic one puts people dressed in, say, green jumpsuits on one side of a room and purple ones on the other, and doesn't let the groups talk to each other, only amongst the similarly dressed. Within the hour, the purples will be supposing and agreeing about all kinds of baseless, distrustful assumptions about the greens, and vice versa. First-year psych students repeat and re-confirm this experiment frequently.

However, we also have a strong instinct to gravitate toward (and eventually mate with) those who look like us, and to be suspicious of those who don't. Technically, and interestingly, it turns out actually to be familiarity, not self-similarity. We tend to prefer people who look like those with whom we grew up as children – our parents, siblings, and early school playmates – as has been shown by studies of adoptees. But for most of human history, it would have been our actual relatives who set our preferences in trust and in love, living in xenophobic populations geographically separated from most others.

An important non-psychological aside: The endogamy this might seem to imply, however, was very frequently broken by male war bands indiscriminately raiding neighboring and remote populations for women; humans have been massively "miscegenating" and gene-mixing since prehistory. Inter-"racial" marriage and procreation has, of course, increased by orders of magnitude after the invention of powered travel and the advent of modern political borders mostly becoming permeable. This means that discrete races are essentially genetically impossible. The reality is that there are dominant and recessive genes; what look like races are just statistically frequent assemblages of dominant genes in various broad populations – and they're becoming more mixed over time, with decreasing distinctiveness between populations.

Anyway, our distrust-the-unfamiliar instinct combines with our classifying instinct to generate a strong perception that races, based on looks alone, are real and defining, and for this "truth" to persist even when long-studied history and advances in modern science prove it to be an illusion.

It is thus probably inevitable that some idea like race will arise in any broad population, based on one thing or another, be it appearance, region of origin, or cultural features like religion. That doesn't somehow make it any less illusory.

Conclusion

Each of the topics and side topics above could easily cover several contemporary scholarly books. Guess what? They've already been written.

Editors of Wikipedia are often individually ignorant of what the last several decades of research and thought and re-thought are telling us. The editorial community has aggregate biases (mostly inherited from surrounding culture). We collectively ignore modern scientific and deep historical understanding only at our peril – or, rather, at Wikipedia's peril.

The community is making a mistake when it lets editors stick labels on biographical subjects and groups to suit prejudices and pseudoscience. And that's before we even get into more nuanced related topics, like cultures and communities and nations and peoples under various sharply conflicting definitions.

Wikipedia is a bad place to engage in labelling that isn't absolutely integral to international public perception of the subject. There's generally no logical basis for it, although it is commonly mentioned in many reliable sources, and what may look like a basis is often a completely cracked foundation.