Is race a social construct?

Written by AI. Help improve this answer by adding to the sources section. When the sources section is updated this article will regenerate.

Overview

“Race” serves two different functions in contemporary discussion.

1. A set of socially defined categories used for law, census, culture and personal identity.
2. A loose shorthand for patterns of human genetic ancestry and visible traits.

Whether race is “a social construct” therefore depends on which of these two senses one is emphasising. Most scholars and commentators now agree that the everyday race labels (Black, White, Asian, etc.) are created and maintained by societies, but they disagree about whether those labels map onto any biologically meaningful structure in human populations.

Arguments that race is mainly a social construct

• Historical scholarship shows that the meaning and number of races have shifted over time and across countries, shaped by politics, colonialism and changing scientific fashions. UNESCO’s mid-20th-century statements deliberately re-framed race as a social idea to fight racism, illustrating that the concept can be re-defined by policy rather than biology [4].
• Philosophers of biology argue that human genetic variation is continuous (clinal) rather than neatly partitioned; therefore any sharp racial boundary is imposed rather than discovered [9].
• The most cited empirical argument, Lewontin’s 1972 result, found that about 85 % of human genetic variation lies within populations and only 15 % between them, suggesting that racial groupings capture little biological information. (Critics dispute the interpretation, see below.)
• In practice, different countries classify the same individuals differently (e.g., “Coloured” in South Africa vs. “Hispanic” in the United States), reinforcing the view that race is context-dependent and constructed.

Arguments that race reflects underlying biological population structure

• Population geneticists using modern datasets consistently find that when you analyse thousands of genetic markers, individuals cluster roughly by geographic ancestry; these clusters often correspond to continental-level race labels used in the West [10][11].
• Edwards (2003) showed that although single genetic loci reveal little about group membership, information from many loci can assign individuals to regional ancestry groups with high accuracy, challenging Lewontin’s interpretation [10].
• Medical researchers have discovered that machine-learning systems can infer a patient’s self-identified race from X-rays or CT scans even when human experts cannot, implying that phenotypic signals correlated with ancestry exist in the data [2].
• Proponents of “race realism” argue that for some biomedical or forensic tasks, continental ancestry categories remain practically useful because genetic risk factors often vary in frequency between those broad groups [1][7].

Current scientific middle ground

A growing number of geneticists, exemplified by David Reich, hold an intermediate view: human populations are structured; genetic clusters exist; but historical mixing and continuous gene flow mean that boundaries are blurry and any classification scheme will be approximate [7]. The UCSC Science & Justice review describes this position as “clinal clusters” and notes that most researchers avoid the term “race” in favour of “ancestry” because of social baggage [6].

Public discourse and controversy

• Journalists and bloggers on the political right (e.g., Quillette, iSteve, Steve Sailer) accuse mainstream institutions of suppressing discussion of genetic differences for ideological reasons [1][5][8][13].
• Progressive commentators and some scientists warn that emphasising biological race risks reviving discredited hierarchies and can be misused in policy debates [4][9][14].
• Opinion pieces in Persuasion and Politico focus on a perceived “conformity problem”: scholars fear reputational costs if they publish results that contradict the social-construct narrative [3][12].

As a result, the phrase “race is a social construct” functions as both a descriptive claim about how categories are produced and a normative stance aimed at reducing discrimination. In empirical research, the trend is to replace “race” with “ancestry” and to treat population structure as real but complex.

Conclusion

Race, as commonly labelled, is socially constructed. Yet the labels correlate imperfectly with patterns of human genetic variation that are themselves real and measurable. Whether one says that race “exists” depends on whether the focus is on the social categories (constructed) or on the underlying ancestry patterns (biologically detectable). The scholarly consensus is moving toward acknowledging both facts while cautioning against reifying the categories.

Sources

1. The Case for Race Realism – Aporia Magazine. https://www.aporiamagazine.com/p/the-case-for-race-realism
2. “AI Recognition of Patient Race in Medical Imaging” (2022 pre-print PDF). https://www.thewikle.com/resources/AIrecognitionofpatientraceinmedicalimaging%282022%29.pdf
3. Discourse on Race Has a Conformity Problem – Persuasion. https://www.persuasion.community/p/discourse-on-race-has-a-conformity
4. Changing the Concept of Race: On UNESCO and Cultural Internationalism (2020). https://www.thewikle.com/resources/Changingtheconceptofrace-OnUNESCOandculturalinternationalism_%282020%29.pdf
5. David Reich: How to Talk About “Race” and Genetics – iSteve blog. https://www.unz.com/isteve/david-reich-how-to-talk-about-race-and-genetics/
6. Developing: Debate on “Race” and Genomics – UCSC Science & Justice. https://scijust.ucsc.edu/2019/05/30/developing-debate-on-race-and-genomics/
7. How Genetics Is Changing Our Understanding of “Race” – The New York Times. https://www.nytimes.com/2018/04/02/opinion/genes-race.html
8. No Voice at Vox: Sense and Nonsense About Discussing IQ and Race – Quillette. https://quillette.com/2017/06/11/no-voice-vox-sense-nonsense-discussing-iq-race/
9. Race: A Social Destruction of a Biological Concept – Biology & Philosophy. https://link.springer.com/article/10.1007/s10539-009-9193-7
10. Lewontin’s Fallacy – A. W. F. Edwards (2003). https://www.thewikle.com/resources/Edwards2003-LewontinFallacy.pdf
11. Current Status: It’s Complicated – Razib Khan’s Unsupervised Learning. https://www.razibkhan.com/p/current-status-its-complicated
12. Why Can’t We Talk About IQ? – Politico. https://www.politico.com/story/2013/08/opinion-jason-richwine-095353
13. Latest Rationalization: Race Doesn’t Exist, But Subraces Do – Steve Sailer Blog. https://www.stevesailer.net/p/latest-rationalization-race-doesnt
14. Trump “Annoyed” the Smithsonian Isn’t Promoting Discredited Racial Ideas – Ars Technica. https://arstechnica.com/science/2025/03/trump-annoyed-the-smithsonian-isnt-promoting-discredited-racial-ideas/

Suggested Sources

1. The Case for Race Realism – Aporia Magazine (Opinion / Essay)
2. “AI Recognition of Patient Race in Medical Imaging” (2022 pre-print PDF; Empirical research)
3. Discourse on Race Has a Conformity Problem – Persuasion (Opinion / Essay)
4. Changing the Concept of Race: On UNESCO and Cultural Internationalism (Historical scholarship)
5. David Reich: How to Talk About “Race” and Genetics – iSteve (Blog commentary)
6. Developing: Debate on “Race” and Genomics – UCSC Science & Justice (Research commentary / Blog post)
7. How Genetics Is Changing Our Understanding of “Race” – The New York Times (Opinion / Op-Ed)
8. No Voice at Vox: Sense and Nonsense About Discussing IQ and Race – Quillette (Opinion / Essay)
9. Race: A Social Destruction of a Biological Concept – Biology & Philosophy (Peer-reviewed journal article)
10. Lewontin’s Fallacy – A. W. F. Edwards (2003) (Peer-reviewed article)
11. Current Status: It’s Complicated – Razib Khan’s Unsupervised Learning (Newsletter essay / Blog post)
12. Why Can’t We Talk About IQ? – Politico (Opinion / Op-Ed)
13. Latest Rationalization: Race Doesn’t Exist, But Subraces Do – Steve Sailer Blog (Blog commentary)
14. Trump “Annoyed” the Smithsonian Isn’t Promoting Discredited Racial Ideas – Ars Technica (News article)