should we ban cognitive group differences research?

Having been raised a good child of modernity, Science(!), and the Enlightenment, my instinctive reaction to the question “should we ban research on X?” is “of course not!” Much as we know (hope, believe) that more speech is the right response to harmful speech, we know that the solution to bad, racist, sexist research is better, emancipatory research. But just as critical scholars of race and law have recognized that free speech and equality offer “conflicting promises“, philosopher Janet Kourany argues in a forthcoming Philosophy of Science piece that freedom of research may also conflict with the principle of equality.

Kourany briefly rehearses how 3+ centuries of biological differences research has been constantly put to racist and sexist ends, despite offering evidence that (in retrospect) sucks. Modern iterations of this paradigm focus on cognitive and brain differences:

And now, in the beginning of the twenty-first century, the claims continue: that women’s brains are smaller than men’s brains, even correcting for differences of body mass; that women’s brains have less white matter; that women’s brains have less focused cortical activity (lower “neural efficiency”); that women’s brains have lower cortical processing speed (lower conduction velocity in their white matter’s axons); and so on.

As shoddy as this research has historically been, and likely continues to be, the effect of this research is chilling: “For example, studies have documented the harm done to women and girls by the publication of scientific claims suggesting an innate female deficit in mathematics.” Given the proven harms of publishing and promoting (likely to be overturned) race and gender cognitive differences research, should we just go ahead and ban that research?

Kourany turns to other recent debates about scientific ethics for insight. She notes that the freedom to research has been restricted when it comes to scientists method of inquiry, as in the rise of rules for the ethical treatment of human subjects. But she also draws on recent examples from synthetic biology, where scholars and policymakers have argued that some knowledge is in fact too dangerous. Researchers in the past 20 years have recreated deadly viruses like the 1918 Spanish Flu, and designed synthetic smallpox that seems to resist existing vaccines. Given how devastating the release of one of these viruses would be, should we ban or restrict this research? Some moves have been made in just this direction. In other words, the freedom to research is not absolute.

If we’re willing to restrict the creation of vaccine-resistant smallpox, should we also restrict cognitive differences research? Kourany proposes that we should:

So, on the one hand, synthetic genomics research unconstrained may cause serious harm to various groups of people. On the other hand, cognitive differences research unconstrained, as previously stated, has already been shown to cause significant harm, though lesser harm than a serious illness, but for much longer periods than the length of an illness, to lots more people—to all the people whose self esteem and self-efficacy and ambitions and successes are lessened as a result of direct or indirect exposure to the research or aspects of the research (e.g., its results or even just its questions), to all the people whose self esteem and self-efficacy and ambitions and successes are lessened as a result of the treatment they receive from others who have been directly or indirectly exposed to the research or aspects of the research, and so on; in short, to all or most women and, in the United States at least, to most minority men…

Kourany ends by arguing that if we spent less time trying to find innate differences between men and women, or white people and black people, we might spend more time on research that recognizes power and social locations as a springboard for improving education, self-esteem, and so on:

When we look at some of the newer work in child development in minority populations in the United States, how the old genetic deficit models and cultural deficit models of minority child development are being replaced by models that foreground and seek to build upon minority competencies and strengths and resourcefulness in the context of racism and poverty; or when we look at some of the work in feminist science studies over the last 30 years, how the old doubts whether women can make the same contributions to science as men have been replaced by studies showing the importantly different critical and constructive contributions to the sciences women have made—when we look at such work, we get a hint of what this new discourse and these new benefits can be…

I’m not sure I’m 100% convinced by the argument, but I wonder if there might be a useful moderate version of it. Given that any individual cognitive differences study cannot have the same world-shattering impact as a supervirus, there’s less of a pressing need to completely ban or restrict research. But there’s also very little need to fund such research, it seems. So what if we (the collective, political we) simply stopped funding it? What if the NIH and etc. said, “we’re no longer going to fund research like this because it hasn’t panned out for a few hundred years and also seems to be doing active harm to vulnerable groups”? Would you support or oppose such a policy? Why or why not? Is Kourany persuasive in her claim that the freedom to research must be constrained by the right to equality – and that this constraint may take the form of prohibitions on producing certain knowledge, not just on methods for producing it?

Side note: if you want to be a mad scientist, apparently synthetic biology is where it’s at!

 

Author: Dan Hirschman

I am a sociologist interested in the use of numbers in organizations, markets, and policy. For more info, see here.

15 thoughts on “should we ban cognitive group differences research?”

  1. I think an issue here is identifying the proper counter-factual. With research on antibiotic-resistant small pox, we have good reason to believe that if we DON’T create a super-charged version, then one won’t emerge in the wild. However, in the domain of erroneous beliefs about social groups, for example males and females, there are already a wealth of beliefs loose in the world arguing that these groups ARE different and that one is BETTER. I’ve more often found research useful for dispelling these issues in class than in reinforcing them. So, if banning research means that we stand on the status quo, I’d say banning it is pretty silly.

    I actually have a pretty recent paper pointing to possible differences in how males and females encode social networks (http://www.sciencedirect.com/science/article/pii/S0378873315000519), so perhaps I’m biased, but I should note that my colleagues and I included the following in the discussion:

    “In summary, we find that females are markedly superior to males in recalling social network information and that this superiority cannot be accounted for using priming effects, differences in the use of compression heuristics, working memory capacity, personality traits, or cognitive flexibility. Somewhat reluctantly, we are forced to suggest that this may reflect an underlying neurological difference between the sexes. There is a growing awareness that social science must take biology into account (e.g., Freese et al., 2003, Hopcroft, 2005, Hopcroft, 2006 and Yamagishi et al., 2003) and we agree that biological and neurological explanations for social phenomena should receive more attention (e.g., Todorov et al., 2011). Nevertheless, it seems more likely to us that circumstances shape males and females such that females develop a relatively greater ability to encode and recall social networks. In general, psychological research on gender difference is more consistent with overall similarities than distinctiveness (see Hyde, 2005, Hyde, 2014 and Joel et al., 2014). Moreover, neurological research shows that mean differences in brain structure between the sexes are smaller than the variations within-sex (Joel, 2012). Previous research showing that low power actors have superior network knowledge (Simpson and Borch, 2005 and Simpson et al., 2011a) is consistent with, and may explain, our results without the need to assume fundamental differences between males and females, and this possibility is worthy of further study.”

    When I talk about this work I almost always have the chance to point out that the main evolutionary advantage that humans seem to enjoy is tremendous cognitive flexibility. So, when you have a species whose primary “hat” is being flexible and adaptable, you should be really, REALLY reluctant to explain differences between subgroups within that species as being hardwired. But people in general seem to believe that these differences are biological by default, so I don’t really see research as making things much worse.

    Liked by 3 people

    1. “So, when you have a species whose primary “hat” is being flexible and adaptable, you should be really, REALLY reluctant to explain differences between subgroups within that species as being hardwired.”
      This is a *fantastic* way of putting it and I’ll have to remember it for future use.

      Liked by 1 person

    1. How about a trade – for the next 20 years, we put all the money we spent on cognitive differences research into gun violence research and then reevaluate where we’re at? Call it a moratorium and a new set of priorities rather than a perma-ban.

      Like

  2. While we’re at it, let’s ban research on implicit associations. Small magnitude differences in people’s milliseconds reaction time to words and pictures of black people are being used to claim that institutional bias is buried in everyone’s subconscious. In what other field would people find an effect observed over the course of a half a second prove a durable effect that lasts hundreds of years? And yet this sort of research is being used to justify the reeducation of men and whites at the threat of expulsion and other administrative sanction.

    Liked by 1 person

  3. The author of the Phil. of Science piece, and the earlier feminist scholars on the differential effects of free speech, need to reread Mill. That speech, or research, may generate negative outcomes is in no way an argument for preventing it. This is true even if we have a very strong prior that the research cannot learn anything interesting (since, e.g., we have been able to gather evidence that the math gap does not appear to be driven by genetic differences across sexes). But it is even more true when we have a reasonable prior that the research can learn something interesting, but we just don’t like what it says.

    As an example, who would deny that genetic differences between men and women drive the gender gap in violence? We have tons of theoretical, biological and epidemiological evidence on this point. Devising a strategy to lower violence rates while pretending such a difference doesn’t exist is absurd. When it comes to other cognitive differences – and there are some group-level cognitive differences, though less than the uniformed or the bigoted might suppose – how are we to make policy without understanding which are real and which are a phantom?

    Liked by 2 people

    1. Regarding the “chilling” “proven harms”:

      Based on page 159 of the relevant chapter in the Routledge International Handbook of Ignorance Studies, three stereotype threat publications are cited as evidence for the claim that “…studies have documented the harm done to women and girls by the publication of scientific claims suggesting an innate female deficit in mathematics.”

      First, stereotype threat studies are priming studies that depend on participants receiving a treatment, not just the publication of a treatment. More importantly, though, the stereotype threat literature cannot be trusted.

      See the discussion of publication bias in the Flore and Wicherts (2015) meta-analysis in which the trim-and-fill estimate of stereotype threat on girls “did not differ significantly from zero” (p. 38).

      See the asymmetric funnel plot for the 116 stereotype threat studies in Nguyen and Ryan (2008). Check in particular where the studies from the original Steele and Aronson (1995) stereotype threat article appear on the plot, along with the studies involving Diederik Stapel.

      See the Finnegan and Corker (2016) preregistered replication study testing for stereotype threat for women in mathematics that provided “no evidence suggesting the presence of significant stereotype threat main effects, nor any moderation by performance avoidance goals, in spite of the fact that the current replication study had a much larger sample size than the original study” (p. 40).

      There might be real nontrivial lasting harms here, but it’s difficult if not impossible to estimate the size of these harms based on a literature of studies conducted by an ideologically-imbalanced set of researchers that have had the flexibility to selectively report short-term results from small, unrepresentative, and cheap-to-obtain samples.

      Liked by 1 person

      1. Priming is a fancy word for demand characteristics. To the degree that priming studies aren’t reporting noise, they’re reporting the degree to which intelligent agents decipher the “rules of the game” that authoritative seeming researchers set up, and give them what they want.

        Liked by 1 person

      2. They’re also evidence that left sociologists are perfectly willing to cite biological and individualist/psychological research, as long as they’re reasonably confident they can pull their subjects’ pants down, exposing subjects’ hidden -isms.

        Like

    2. “As an example, who would deny that genetic differences between men and women drive the gender gap in violence?”

      I’d be very curious to get more details on the literature you have in mind, because I think a lot of people would deny this, and there’s at least some scholarship that does so. For an older example, see Anne Fausto-Sterling’s work. Fausto-Sterling (and others, I think) argues – among other things – that hormone levels may be caused by aggressive behavior, rather than causing it .

      For (somewhat) newer work (h/t Matt Brashears), there are experimental studies that find very little difference. For example, Lightdale and Prentice find that with some experimental manipulation to “deindividuate” (heighten anonymity) women and men behave equally aggressively. Hyde (2014) summarizes some of this kind of work in a review piece:

      “Findings such as these have led some researchers to conclude that the emphasis on gender differences in aggression research is misplaced and that, instead, context plays a far greater role (Richardson & Hammock 2007). Consistent with this view, a meta-analysis by Bettencourt & Kernahan (1997) of experimental studies indicated that, in the context of violent cues but no provocation, a moderate gender difference is found (d = 0.41), whereas when violent cues and provocation are both present, the gender difference is eliminated.”

      I certainly have my skepticism about soc psych studies like these in general, and I don’t want to cherry pick, but you asked “who would deny” and the answer is, plenty of people it seems.

      Part of the problem, I think, is that most of the “theoretical, biological, and epidemiological” evidence you allude to is very tricky to interpret to get at some kind of underlying genetic causation, or more specifically, the sort of genetic causation which would lead some sex difference to be a relatively immutable fact of nature rather than a contingent production of genetics plus current social arrangements.

      Even the data themselves have to be interpreted carefully. Given the differential stigma of being a victim of violence, for example, there’s strong reason to think that men would differentially underreport intimate partner violence which would in turn affect the observed levels of violence in public health surveys (which is not at all to say that men are actually abused more than women!). See some discussion here, for example.

      I wonder if any anthropologist might also want to chime in with examples of societies that observe a very different relationship between gender and violence. We’re far from my field, but I’d be curious to know what the observed range of variation is here!

      For a rather literary take on how women’s aggression and violence has been systematically written out of understanding of our own history, see the brilliant essay “We Have Always Fought.”

      Like

      1. “We have Always Fought” is a wonderful essay. Well worth my time to read it. Thanks for the link.

        I think my bias is in favor of research and against banning research, mostly on the grounds that powerful people with agendas will find alternate paths to an end if one is blocked. I was reminded recently that well into the 1970s, sociologists interpreted high rates of civic/organizational participation for Whites as civic virtue, while the even higher rates of participation for Blacks (after class is controlled) were interpreted as pathological. If you are determined to disparage another group, you can always find a way to do it.

        At the same time, there are always political pressures to make certain kinds of research more fundable and acceptable than other kinds of research. I do find myself attacking the politics of people who ban or under-fund research I’m in favor of, while thinking it is legitimate to lobby for more funding for research on topics I think are important.Worth attending to the asymmetries.

        And research cuts multiple ways. When I started researching racial disparities in criminal justice, a lot of activists (White and Black) were not sure they wanted to publicize them,because they thought it would just play into White biases about Black criminality. And there is research that suggests that being told about the disparities makes some Whites support mass incarceration. Even as the same data makes other people sure the system is unjust and has to be reformed. Or education: the manifestly worse performance of boys in school is rarely interpreted [at least by men] as evidence of the biological inferiority of male cognition, and almost always as proof that schools are too feminized and failing to educate boys properly. For that matter, how often is the higher academic performance of girls over boys even lauded or noticed? How often have you been reminded that Black juveniles and young adults use illegal drugs at lower rates than Whites? My point is the same as in “We Have Always Fought”–there is lots of evidence challenging stereotypes out there, but we tend to ignore it.

        On gender and violence, I’m pretty sure the domestic assault literature says that women are as likely to hit partners as to be hit, but women get hurt more, which is attributed to size differences. This average size difference is a biological trait that I think most people believe arises from some sort of selective pressure among our ancestors that is probably correlated with other things, even as we keep in mind that our sexual dimorphism is only moderate, as our size distributions overlap. And apparently testosterone really does make your muscles stronger, pound for pound. The physicality of birthing and feeding babies also tends to tip us into certain kinds of social arrangements, although these can obviously be overridden. Mothers of course are often violent, both toward and in defense of their children, but the survival of the species probably tips us toward some sort of nurturance for babies and children, at least our own.

        Liked by 2 people

  4. Given that social research does not require the creation of a thing like a drug-resistant virus, I can’t fathom a reason to ban research that otherwise manages to pass human subjects review. It (such a ban) seems like an excellent tool for blocking work we don’t like, rather than work that is actually harmful (and how can we define harmful in a way that does not also include much other inequality work?).

    Also, “cognitive group differences” could easily apply to group attitude differences. Banning that seems absurd. Really, it seems to rely on the assumption that some cognitive traits are immutable – are we certain that is the case? Or that measured levels of cognitive traits in adulthood are perfectly heritable, given all the contextual environmental influences that gave rise to them? Of course not. If it requires we accept the premise of a simple biological deterministic view of human cognition to ban such work – why on earth are we doing it?

    Like

Leave a Reply

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

WordPress.com Logo

You are commenting using your WordPress.com 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 )

Google+ photo

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

Connecting to %s