blood pressure, the slavery hypothesis, and social construction

My wife is a physician, and like many doctors was taught in medical schools that African Americans are susceptible to hypertension, and particularly salt-sensitive hypertension, as a result of genetic selection through conditions during the middle passage. I raised this possibility in chatting with Liana Richardson, a postdoc here at UNC, about her very interesting work  on hypertension as a biomarker for stress over the life course, and in particular as a marker for high stress among African Americans. Her response was very interesting, and illustrates an example of cross-disciplinary information flows.

The slavery hypothesis states that the physically brutal experience of the middle passage provided a strong genetic selection effect whereby those better able to store salt and use it to maintain blood pressure in the absence of hydration and nutrition were more likely to survive the passage. This is a plausible hypothesis. However, this article by Williams, Mohammed, Leavell, and Collins claims:

The slavery hypothesis is a final example of the misuse of genetics in contemporary health disparities research…. A critical review of this hypothesis shows that it is inconsistent with the processes of population genetics, historical data on the scarcity of salt in Africa and on the level of mortality during the slave trade as well as on the proportion of mortality attributed to diarrhea. (86-87)

The citation is to Kaufman and Hall, “The Slavery Hypertension Hypothesis: Dissemination and Appeal of a Modern Race Theory,” Epidemiology 14:1 (2003) . It’s a very good article, but the summary in Williams et al. is not really right. As I read it, Kaufman and Hall demonstrate that the slavery hypothesis is unproven, and that it is taught, particularly in medical schools, as far more settled science than it actually is. They then provide a very enlightening historical treatment of the tendency to reify race and view racial disparities as rooted in biology, implicating the ubiquity of the slavery hypothesis in that history. Thus, as I read it, they have not established that the theory is false or inconsistent with knowledge about mortality rates, salt availability and so on, but just that the theory remains only plausible, not demonstrated. But refracted through Williams et al.’s summary, it becomes “evidence” against a biological basis for racial disparities in hypertension–which, at least to my reading, is substantial overreach.

The Kaufman and Hall article is a good example of a type of approach in medical sociology I call the Social Construction Of Dot Dot Dot approach: using historical and documentary methods to demonstrate that some condition is socially constructed; therefore, not rooted in Absolute Scientific Proof ™ and, by implication, somehow not really real. Other high-quality examples of this approach include Elizabeth M. Armstrong’s “Diagnosing Moral Disorder” on Fetal Alcohol Syndrome (FAS) and Abigail Saguy and Rene Almeling’s “Fat in the Fire?” on obesity.

Both of these are excellent pieces of scholarship; I do not mean to malign them (or, for that matter, Kaufman and Hall) on their own terms. But there is a tendency to interpret these findings as evidence that the conditions in question are not fundamentally real. I have three concerns with this tendency:

First, it undermines the effectiveness of research like that of Armstrong and Saguy & Almeling by attributing undemonstrated findings to it;

Second, it suggests, if only by implication, that other conditions are not socially constructed, since the debunking style implies without specifying a type of evidence that would establish a condition as being ontologically different from the one(s) under consideration; and

Third, it implies that the opposite of “socially constructed” is “real.” This is a fool’s errand for sociologists, as it cedes the scientific high ground to other disciplines and relegates whatever we study to the margins. As scholars in the Foucault and Latour traditions (and elsewhere) have shown, essentially any reality has a specific, contingent, sometimes even contentious, pedigree in the practices and discourses of its discovery/development. This does not compromise their real-reality, only the inevitability of that real-reality.

Author: andrewperrin

University of North Carolina, Chapel Hill

15 thoughts on “blood pressure, the slavery hypothesis, and social construction”

  1. I think the payoff of work that illustrates how medical conditions are “socially constructed” is in demonstrating how powerful individuals/groups gain and maintain legitimate authority.
    My read of your post is that this is a different issue from that in the first half of your post, which seems to be one of overstating claims – or at least the degree of consensus around them – no?


  2. To your list of high quality examples, I would add Allan Horwitz’s work on mental illness — generally in _Creating Mental Illness_ and his more specific work with philosopher Jerome Wakefield on Depression in _The Loss of Sadness_. It’s especially enlightening because you can see the effects of different levels of evidence for the “reality” different mental disorders and how this affects the degree and character of the “social construction” that occurs.


  3. I’m not qualified to comment on the sociology of medicine, but will comment on two of the issues you raise. One is the tendency for published scholars to provide incorrect characterizations of other people’s research. This is remarkably common. I’m sure I’m not the only person who has observed a lot of mis-citations to my own work, but I have often looked up the citations in other people’s work only to find that the research that was actually published was not really what the person citing it said it was. I first learned this from a graduate school assignment (to look up all the citations in famous article X as part of evaluating its evidentiary base) and have replicated the finding since then. Typically 1/3 to 1/2 of all citations are erroneous in some way.

    As regards “socially constructed” versus “real” here’s what I say every semester when explaining the social construction of race: Education is entirely social, it is all social construction; it is clearly “real.” Language is all social construction; it is clearly “real.” Slavery is entirely a social construction; it is very “real.” Things that are socially constructed are just as real as physical objects or biological traits. I think the evidence is that there are biological traits that are correlated (quite imperfectly) with “race” as it is understood in the US, but the social consequences of race in the US are generally not coming from those genetic markers, they are coming from the socially constructed racial categories in the US that give people different life experiences.

    It may be the case that there are health factors that are directly tied to the genetic markers, not to the social consequences. But even clear-cut genetic differences get intertwined with social factors that are very “real.” I was just reading Mann’s 1493 that talks a lot about the genetic resistance to malaria in some African populations and how that played out in the organization of slave plantations in different latitudes.If Mann is right, you’d probably find that malaria resistance is correlated with poor health outcomes in the US today. But if it is, the causal path involves the mediating conditions of slavery and discrimination, not the direct consequence of the genetic resistance to malaria.

    There’s also a social construction of malaria to be done, I’m sure. But I’m not qualified to comment on that.


  4. Dan Hirschman began a discussion of this on Facebook: and commenters there are right that my post is a bit unclear. To recap what I wrote in response over there:

    I do think I was unclear in locating the blame for the fallacy I delineated, so let me be clearer here. Dan Hirschman is right in that I do not think any of the three “Social Construction Of…” papers I mentioned actually makes the ontological claims I critiqued. I think the problem arises in the interpretation of this research. However, I have heard people in this field discuss socially constructed medical conditions (at conferences, etc.) in ways that do make this mistake. So my sense is that this slippage is present in the worldview of (some of) these authors, but that they are careful in publication to avoid it.


    1. Chris, the Kaufman link you posted is the same article I referenced – as I read it, it does not provide a strong refutation of the plausibility o the slavery hypothesis, but rather a very good cultural history of the acceptance of that hypothesis.


      1. Whoops- I see that now. Let’s blame it on sleeplessness :)

        I heard some additional ramblings about the Roland Fryer paper from friends in the Econ program at Harvard who were echoing the Kaufmann and Hall criticism. Apparently he’s had a lot of trouble publishing the piece because of such criticism– too bad, because it’s a really interesting paper that compares hypertension among African Americans to other African diaspora.


  5. Thanks, Andy, for saying publicly what is clearly implied by so many who love to fly the constructionist banner but few have the cojones to say– i.e.,

    “My position is that there is no reality independent of of what we socially construct, and any concession that there are constraints on such constructions undermines my professional project. Therefore, research that claims that something is socially constructed is problematic because it implies that there are objective standards against which social constructions can be judged, and worse– that objective reality constrains social construction.”

    Some may remember that I made the same point a few years back [see here: when I said that the term “bubble” was problematic for pure constructionists because it implies an objective standard against which social constructions can be judged. But while in that debate, Kieran insisted on having his cake and eating it too [i.e., using the term bubble but denying the implication that there are objective standards against which prices can be judged; and refusing to offer an alternative term that would imply objective standards for prices], you are conceding that you can’t have it both ways. Of course, the difference between us is that you embrace this implication as your cause, while I regard it as demonstrating pure constructionism to be intellectually and practically bankrupt. But that is a trifling diffrence among friends, and it is the holiday season…

    Liked by 1 person

  6. Ezra, it may be the wafting smells of pumpkin pie and sweet potatoes, but I believe I am with you on the epistemological principle: that while people may construct things, they do not do so under circumstances of their own choosing (to borrow a phrase); and that at the bottom some of these constraining circumstances are likely received from the environment as opposed to completely endogenous to the social-construction system.

    The difference between us, I suspect, is two-fold. First, there is a difference of degree: I think, and I believe the surfeit of Social Construction Of Dot Dot Dot research suggests, that the vast majority of constraints upon construction are the result of prior constructions and not of exogenous limits. In other words, I think you’d need to pursue a chain of constructions constraining and enabling more constructions pretty far back to get to any exogenous limit. Second, there is a difference of kind: you assume, in prior posts, that that which is socially constructed is not real or true (to wit, your argument with Kieran over the true value of housing), a claim I believe to be false.

    Happy Thanksgiving!


  7. Andy:

    Thanks for accepting my reaction in the spirit with which it was intended. That said, I fear we are still talking past each other. First, I do not recognize my position in how you have characterized it (in particular, I never said that bubble prices were not real; I said they were *wrong*). Second, my main problem with what you wrote was not in the idea that social constructions are constrained by other (slower-to-change, and often taken-for-granted) social constructions, but what the implications of this are for research and our own (political) action. You seem to regard this point as a reason to avoid any research on the process of social construction because it might mislead people into not realizing that those constructions are build on top of longer-term constructions. I don’t see this as an important concern, especially since the moral justification for any (political) project we might have is rooted in the claim that existing institutions are *objectively* inferior to desired alternatives. E.b., I don’t know how anyone can argue for an end to discrimination without asserting that discriminatory valuations depart from objective standards of merit.

    For elaboration on this and related points, see this draft of my forthcoming Annual Review of Soc piece on social valuations [], esp p.3, and pp. 14-17.


  8. But while in that debate, Kieran insisted on having his cake and eating it too

    What? This is—to coin a phrase—an objectively false characterization of my view in that thread. In that case, the question at issue was, can there be a coherent theory of market bubbles does without the concept of objective value, and instead treats the phenomenon as a process endogenous to the beliefs of market actors. I argued that there can be. Your argument in that thread rested on an insistence that the term “bubble” must be defined as a phenomenon whereby asset prices diverge from their objective values, either because this was the usual meaning of the term in everyday use or simply because you were insisting that this was the only definition you would entertain. I argued that this was transparently begging the question, and that a constructivist would feel no compulsion to accept the folk definition of a term in order to explain it, even if very clever people at MIT insisted they had to. I went on to sketch what such an endogenous account might look like. This is a very long way, in my view, from insisting on having my cake and eating it. To the contrary, I think re-reading that thread shows Ezra was the one insisting everyone had to bake their cake by his recipe—because if you accepted his definition of “asset price bubble” then a constructivist account of bubbles would ipso facto be impossible.

    It’s probably worth saying that defending the possibility of such a theory does not commit me to believing in it. In other circumstances, I’m as likely as the next person to be skeptical of bog-standard constructivist arguments.


  9. Hi Kieran.

    Seems like we are pretty much where we were when we last engaged on this three years ago. But why not make like that “ant with the rubber tree plant,” and try again?

    My understanding of our exchange was as follows:

    1. I made the point that calling an event a “bubble” is problematic for pure constructionists because such a characterization implies that there are objective standards against which financial-market prices may be judged, and by which they are ultimately constrained.

    2. Your response was to say (as you do now again) that pure constructionists are perfectly entitled to use a different definition of “bubble” than the one I was using. According to the pure constructionist definition, “bubble” would denote an episode where collective beliefs led prices to increase significantly and then to decline markedly. This definition does not imply that these prices were high relative to an objective standard, and certainly not that they were constrained by objective standards. Rather, prices reflect nothing more or less than collective beliefs.

    3. My response was to concede your point for the sake of argument but to note that it is meaningful (and important) to distinguish between two different accounts of bubbles and what makes them pop. (And I quote):

    a. The reason that prices fell was because that they had gotten so high relative to their true underlying values that they could not be sustained.

    b. The reason that prices fell was because people came to believe that they had gotten so high relative to their true underlying values that they could not be sustained.

    4. Your response seems to be that a pure constructionist can stick with the latter account (or perhaps an even a “purer” account in which the myths that drive prices make no reference whatsoever to intrinsic values, and is just about a change in beliefs about what others would believe) and be perfectly consistent. This is the “cake” I had in mind.

    5. But as I pointed out back then, and elaborate upon in that ARS piece, there is a major price (!) to such purity in one’s constructionism—i.e. it means forswearing political action, or at least any political action that is justified in terms of a critique of social valuations and the institutions that support them. My example back in 2008 was that just it was foolishiness to have a pure realist for a financial-market regulator (Chairman Greenapan), it would be just as foolish to have a pure constructionist in such a role (“Chairman Ponzi”). And while this implication may not bother most sociologists, for whom financial-market regulation is not high on our ambitions, it should bother us when we realize that it undermines the political projects that we do hold dear. For instance, when we claim that there is discrimination in labor markets (and seek to eliminate it), this claim is necessarily founded on a belief that people of different genders and races are objectively equal (e.g., in their skill/potential for a given job). Back in 2008, you were not excited about embracing this implication; it was this lack of enthusiasm that I meant when I referred to you as “eating it too.” Sorry, but it seems to me that if you insist on renouncing objective standards, you lose any basis for criticizing existing social valuations. And if you want to criticize social valuations (e.g., fight against discrimination) then you cannot be a pure constructionist. (I realize that you personally have not endorsed a pure constructionist position. So then it is not you personally who is caught on the horns of a dilemma, but rather the pure constructionist position you are trying to defend. Such a position can retain its purity only at the price of an exit from politics—and really any kind of action that is intended to change social valuations [e.g., any attempt to use empirical evidence to make a case])

    Meanwhile, Andy seems perfectly happy to concede point 1. He expresses the very same discomfort with the claims of “social construction of…” that I said pure constructionists should experience with the term “bubble.” That is, by asserting that some valuations are constructed, the implication is that there are objective standards against which social constructions can be judged. And in my last comment, I pointed out to him that this means accepting the implication in point 5. I have a feeling he will also not be so excited about giving up the eating of cake.


  10. I should know better than to get into an argument with Ezra, particularly as the tryptophan-induced friendliness has dissipated, but…

    I believe there are three distinct kinds of pure-constructionism vs. pure-realism going on:

    1. Real vs. fake (existential judgment)
    2. True vs. false (factual judgment)
    3. Right vs. wrong (moral judgment)

    Ezra confounds these three. My claim about social constructionism in medical sociology is about type 1 (existential judgment): the implicit, or explicit, claim that medical conditions are unreal due to their demonstrably constructed character.

    Ezra’s claim about bubbles is about type 2 (factual judgment): bubble-driven prices are not unreal as in type 1; they are false, that is, inadequately tied to a (latent?) true valuation of the commodity being traded. (SIDE NOTE: I believe, with Kieran, that it is sufficient to note that market participants believe the prices are false to produce the observable effect labeled “bubble,” so the observation of price changes that look like bubbles is insufficient to demonstrate that underlying prices are really false.)

    Finally, claims about how things ought to be are type 3 (moral judgment). One does not need to demonstrate an absolute, non-constructed position, whether about the talents of groups of people or about recognition, redistribution, etc., in order to make and articulate moral claims. The critique of institutions and social valuations may easily rest upon an entirely, or nearly entirely, socially constructed scaffolding as long as it is widely shared or at least understood. This is why politics is largely deliberative or agonistic, not technocratic: most political (and, hence, moral) claims take place on a terrain of conflict in which multiple valid perspectives coexist in some relation, whether comfortable or not. That’s why a position that recognizes very extensive social construction in types 1 and 2 does not foreclose political action.


    1. Andy:

      Two clarifications, and then a response:

      Clarification 1: I have never disagreed that “it is sufficient for market participants to believe that prices are false to produce” sharp reductions in price. Nor I do I disagree that ” the observation of price changes that look like bubbles is insufficient to demonstrate that underlying prices are really [wrong].” I have never said nor implied this. If you want to know what I did say, you’d need to read what I wrote.

      Clarification 2: I have never said anything that is inconsistent with this line of yours: “The critique of institutions and social valuations may easily rest upon an entirely, or nearly entirely, socially constructed scaffolding as long as it is widely shared or at least understood.” In fact, I have said explicitly (in the draft of the ARS piece and in the earlier discussion) that it is useful to think of layers of social construction, with shorter-term, contingent practices and valuations built on top of longer-term constructions that confront us as if they are natural and objective. (Of course, even the longer term constructions are bound by the laws of physics and biology)

      But where we seem to part company is in what we do with that observation. You seem to use it as a reason to close off analysis of how those longer-term constructions constrain shorter-term constructions. You seem to be very worried that to engage in such analysis is to mislead people into a fallacy of misplaced concretness. I agree that this is a potential concern, but argue that it pales in comparison to the downsides of ignoring the constraints on (short-term) constructions. The main downside is that it undermines our most cherished political projects. And so to retain her purity, the pure constructionist must forswear politics.

      You responded by conceding my point. Oh, you didn’t? Well, yes you did. You tell me that politics can indeed be based on pure constructionism because politics is “agonistic”– i.e., it is just different people arguing with each other with no hope of resolution, and certainly no way of using objective evidence to support one position over the other. But of course, no one who actually engages in political action thinks that their position is as valid as anyone else’s. My whole point was to ask you to consider a political project that is important *to you*. When *you* make a political argument (e.g., against discrimination, against authoritarian rule, etc), you argue based on an appeal to that “scaffolding” you referred to (e.g., you use objective evidence to demonstrate that women can do the same work as men; you note that authoritarian leaders cause misery, etc.). To embrace “agonism” as an account of *one’s own* politics is precisely to forswear politics.


  11. Ezra – I believe we are actually on the same, or at least a very similar, page on this, the heated rhetoric notwithstanding. My concern with the social construction of… literature is that it is internally inconsistent; I am less committed than you are to establishing an Absolute Truth, but my objection here suggests either that analysts should pursue an everything-is-socially-constructed-and-therefore-showing-that-isn’t-very-exciting approach or a how-does-social-life-refract-presocial-constraints approach.

    I’m happy with the latter, and in fact part of my concern with Saguy’s work is exactly that it treats evidence about the effects of obesity as if the evidence is only floating signifiers without referent. But my argument is that if they choose to pursue the social-construction-of-everything approach then they need to confront the problem that they lack an available comparison case: something that is aptly understood as non-constructed and therefore different from the case(s) they study.

    I do not think agonistic politics implies no hope of resolution, just no hope of stable, absolute resolution based purely on facts. Political resolution tends to rest on more or less widely-shared values and preferences and the institutions that license these, not on external validity. I will be posting on this question separately soon.


Leave a Reply

Please log in using one of these methods to post your comment: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: