The literature of identity politics and social justice, with or without capitals, is full of assertions that this or that system, conception, or object is a “social construct.” This is nearly always claimed without defining “social construct,” though most of us have a vague idea that the term means something that lacks an objective reality independent of human social agreement. And it’s usually used dismissively—not to deny something like gender identity or racism—but to deny that they exist independently of human thought. That is, the claim that “race is a social construct” is taken to mean that “there is no objective reality to the concept of race, which was simply created by humans” (the usual reason is to give groups power over other groups), but nevertheless race is seen and treated as real in the same way that the idea of a monarchy (see below) is treated as real.
I decided to look up various definitions of “social construct” to see if my notion was true. It turns out that, by and large, it is, and most definitions are pretty similar. Below, for example, are four definitions with links (given definitions and glosses are indented).
Oxford English Dictionary: A concept or perception of something based on the collective views developed and maintained within a society or social group; a social phenomenon or convention originating within and cultivated by society or a particular social group, as opposed to existing inherently or naturally.
Merriam-Webster: an idea that has been created and accepted by the people in a society. Class distinctions are a social construct.
Macmillan Dictionary: a concept or belief that is based on the collective views of a society rather than existing naturally
yourdictionary.com: Social constructs develop within a society or group. They don’t represent objective reality but instead are meaningful only because people within the society or group accept that they have meaning. Simply put, social constructs do not have inherent meaning. The only meaning they have is the meaning given to them by people.
For example, the idea that pink is for girls and blue is for boys is an example of a social construct related to gender and the color of items. The collective perception that a particular color can be associated with a certain gender is not an objective representation of truth or fact. Instead, it a social convention that came to have meaning within the context of society.
So I was correct: “social constructs” are ideas or objects or notions that do not exist independently of human decision making and social agreement. They are not “real” in the sense that without social agreement about what they mean, they would have no objective reality. Or so it is claimed.
Now the term “reality”, of course, is slippery. Certainly money is real, in terms of paper currency, but the agreement that it can be used to purchase goods and has ascribed value is a social construct. Even Martian sociologists could observe this, but the value of a dollar bill would have to be ascertained by observing how it’s used. And the British monarchy is real, though it wouldn’t exist without social agreement. I won’t go on in this vein, as it leads into psychological hinterlands where I would be criticized by some no matter what I said. I simply present the definitions I’ve seen above.
Now, here is a list of examples of “social constructs” along with my rough take on whether I think they really do adhere to the definitions above. You can find more examples here.
gender. Gender and gender roles are multifarious, and more are devised each day. The behaviors associated with these (e.g., “genderfluid”) do describe real behaviors, but “genderfluid” as a given category seems to me a social construct.
gender roles. Same as above, though the behaviors may stem from biology. I would have trouble, for example, with the idea that being bisexual is “just” a social construct, for it does describe people who are attracted to members of both sexes. And there may be a biological basis for this.
sex. As I’ve argued at length, sex is a biological and objective reality, in nearly all cases of animals a binary category with a strict basis resting on gamete size. So while gender may be a social construct, sex, as in “biological sex”, is not.
sex roles. This is a mixture of both an objective reality and a social construct. That is, the view that men are generally attracted to women and vice versa, a feature that has an evolutionary basis, is not something agreed on by society, despite numerous exceptions like homosexuality. And some behavioral differences between the sexes, like aggression and risk-taking, are, I think, not social constructs but partly encoded in our DNA by natural selection. Other “roles”, like guys should like blue and girls pink, are clearly social constructs.
religion. Despite the claim that people have an inborn desire to apprehend and worship divine deities or concepts, I see religion as a social construct. It may have a biological origin, as some claim (ascribing mysterious events to specific causes), but religion in the form we know it is clearly something devised by humans. I also don’t think that if we wiped out all religious sentiment from the planet, it would return with nearly the ubiquity it has today. We simply know too much about what makes things happen, and we still have no evidence for gods.
social class system. It’s an objective fact that some people are smarter than others and some make more money than others. But the idea that this makes some people superior to others is clearly a social construct, and a maladaptive one. Indian castes are similar, but have been genetically separated for so long via historical origins as well as prohibitions on intermarriage that now there are correlations between one’s caste and one’s genes.
monarchy. A social construct and, I think, another maladaptive one.
marriage. A social construct; many societies don’t have marriage in the way we know it. The rules, rituals, and laws about marriage have all been made up by society.
countries. Clearly social constructs based on human history and either warfare or general agreement among different groups of people.
money (see above).
biological species. Not a social construct in general, but a reality existing independent of humans, most obvious in sexually reproducing animals but also in many plants (animals, after all, chose to mate with members of their own species, and that choice has nothing to do with human consensus). For a full-scale justification of species as real groups, independent of human conception, see Chapter 1 of my book with Allen Orr, Speciation.
disability. Another slippery one. Clearly if someone has lost their sight or their limbs, they are not as “able” to do some stuff than people who are relatively intact, though they may develop compensatory skills (like more acute hearing in the deaf) that make them “super able” in other ways. Ergo the term “differently abled.” In general the idea that people with such losses should have interventions to compensate for them and enable them to participate more fully in society, and should have such interventions, is both an objective reality (e.g., for the blind) but also a social convention (our moral view that the disabled deserve to be accommodated).
I should add here that I see morality is perhaps the most prominent social construct, for while it’s a fact that societies have moral systems, the specific actions viewed as “good” or “bad” have no objective justification or even a label independent of human agreement.
race. This is the most hot-button of the topics, so I’ve saved it for last. Clearly race is a “social construct” if by the term you mean that “races are absolutely distinguishable groups of individuals with substantial and diagnostic genetic differences.” The old Carleton Coon-ian races of “Caucasoid, Mongoloid, Capoids, Congoids, and Australoids” have gone down the drain.
On the other hand, multi-site genetic analysis shows, in general, that humans do fall into groups largely distinguishable from their DNA, though those groups are overlapping and show gene mixing, so that many individuals cannot be said to fall into a given group. But the grouping of humans can, with fair accuracy, give an idea of someone’s geographic origins and ethnicity, because it reflects an ancient geographic separation of populations that led to their genetic differentiation. As we know, the amount of diversity within any given group exceeds the diversity between groups, but that doesn’t mean that you can’t use multiple segments of DNA, combined, to diagnose someone’s ancestry and ethnicity.
Multilocus groupings of humans, for example, show that they can be divided into various fairly distinct genetic clusters, ranging from 4-7, and which correspond roughly to areas where humans were genetically isolated (Africa, Oceania, East Asia, the Americas, etc.) In the U.S., multi-site cluster analysis identifies four clusters, corresponding to whites, Hispanics, African-Americans, and East Asians (Chinese and Japanese). Further, when you look at someone’s genetic profile and put it into one of those four clusters, and then ask them, without that knowledge, what their self-identified “race” is, the match between genetics and self-identified “race” is remarkable. As the paper of Tang et al. notes:
“Of 3,636 subjects of varying race/ethnicity, only 5 (0.14%) showed genetic cluster membership different from their self-identified race/ethnicity.”
I won’t cite other studies showing that you can identify the location of one’s genetic ancestors with remarkable accuracy. The point is that this correspondence between genes and ancestry, and between phenotype (correlated with ancestry) and genes means that “race”, while a loaded term—I use “ethnic groups” as a substitute—has some basis in biological reality and therefore is not a social construct. If the concept of “race” (or “ethnicity”, as I prefer to say) were purely an agreement of people within society having nothing to do with objective reality, you wouldn’t see the correspondence between how one identifies themselves and the code in their DNA. I hasten to add that these biological identifiers of races say nothing about hierarchies, but they are biologically and evolutionarily meaningful.
All this discussion goes to show several things. First, the concept of a “social construct” is bandied about widely, but often used either inaccurately or is not defined at all. Some things seen as social constructs, like sex and race—or species, for that matter, as some misguided biologists have asserted that species in nature are purely human-defined segments of a biological continuum—actually have an objective reality independent of human consensus. Others, like a monarchy or Mormonism, are purely the results of a human consensus. Thus you need to explain what you mean when you claim that something is a “social construct”, and explain why that concept has no objective reality but is purely the result of social agreement.