Over at the National Humanities Center’s On the Human website, Paula Moya has posted a fascinating piece on cultural neuroscience, science reporting, and race. Go check out the discussion going on there and then, if you wish, consider these thoughts on cultural comparison.

This is an extended version of my comment on Moya’s short essay A Story in Two Parts, With An Ending Yet To Be Written. The NHC invited me (along with others) to write a brief response, but I was so enchanted by the subjects Moya discussed that I spent a chunk of time this week reading some of her references and trying to wrap my brain around the psychology involved. Important lesson: psychology articles are not entirely transparent for a lay person! But not unreadable for all that, except where the statistical analyses get beyond the remains of my never-more-than elementary background in stats.

Anyway: Writing in a witty fairy-tale mode, Moya discusses the work of the social psychologists Shinobu Kitayama and Hazel Rose Markus. Kitayama has done considerable work on the mismatch between theories of the self derived from the study of Americans (and Western Europeans) and the self-concepts found in other societies, like Japan and China. In comparative studies, he and his collaborators have shown how concepts of the self as “independent” or “interdependent” influence cognition and behavior even at the most basic level. Moya discusses some recent work by Jinkyung Na and Kitayama on spontaneous trait inference, that is, the practice of inferring a person’s traits from that person’s behavior. (The essay, “Spontaneous Trait Inference is Culture Specific: Behavioral and Neural Evidence,” is forthcoming in Psychological Science. Moya’s post directs us to a press release from APS on the research, and I was able to look at a preprint thanks to APS’s public relations office.) This inference, when it ignores the constraints context can place on behavior, has been described as a “fundamental attribution error” endemic to human cognition. But it turns out that Na and Kitayama’s subjects, drawn from Michigan undergraduates, do not all infer traits to the same degree. And Na and Kitayama show that some of the variation in trait inference is explained by varying self-concepts: those students with more interdependent self-schemas were less prone to spontaneous trait inference. And this variation was apparent even in a subtle test of “spontaneous” or automatic trait inference, a lexical decision task that never explicitly asked subjects to make judgments or infer personality traits. The same variation was even apparent, to some degree, on an EEG (though I don’t know enough to understand the underlying physiology of this part of the research).

But the main part of Moya’s, and Na and Kitayama’s, story is that the students were members of two ethnic groups, European-American and Asian-American. The former were more likely to favor the independent self-schema and to infer traits. Moya documents the way in which reporters have taken even advance press releases about this study and made stories implying that scientists have “proven” that East and West think differently, that Europeans can’t help their prejudices, and that Asians and whites have different brains. Moya and Hazel Markus, in their recent anthology of essays on race in contemporary American society, have developed a very useful conceptual model for talking about moments like this. Those science reporters are doing race: that is, they are doing the cultural maintenance work of reproducing racial categories and the stereotypes and inequalities that go with them. Moya and Markus argue that this work is, indeed, what race is: not a biological essence or a bundle of phenotypic characteristics but “a dynamic set of historically derived and institutionalized ideas and practices” (Doing Race, 21) that produce social hierarchies. They are taking a result which is, for Moya and for Kitayama et al., about a cultural variation in and influence on cognition and turning that result into a spurious justification for racial categorization.

All the way down

In other words, the point should be that culture goes all the way down into brain physiology, not that a genetically fixed brain physiology goes all the way up into individual attitudes. In the case of the work Moya is discussing, the irony of the misconstruction is particularly acute: the subject of the research is whether everyone regularly attributes a person’s behavior to that person’s fixed characteristics—which is a part of the cognitive apparatus underlying racism (via judgments of the form “She’s doing that because she’s Chinese, and that’s how Chinese people are”). This mode of attribution is rooted in the brain, but it is culturally specific and correlated with an independent view of the self.

It strikes me that this work shows that questions of ethnicity and cultural difference are great matches for cognitive-scientifically influenced humanities research. It has sometimes seemed that cognitive approaches in the humanities were opting for a strongly universalist version of “human nature” that plugged its ears to the many critiques of the idea that “our” cultural products contain straightforward, decontextualized lessons about the essence of the human. But of course few questions could be closer to the heart of the ways in which culture and human biology interact than those of identity, ethnicity, and race. Those questions are ideal domains in which humanists interested in neuroscience and psychology could show the power of those disciplines in clarifying the breadth and complexity of cultural variation. I think Moya’s response to Linda Martín Alcoff’s comment on her post is an eloquent elucidation of this point.

Seeing Moya’s discussion also reminded me of a recent Behavioral and Brain Sciences article I had seen: Joseph Henrich et al., The weirdest people in the world? Henrich et al. discuss how an overuse of experimental subjects from WEIRD [“Western, Educated, Industrialized, Democratic, and Rich”] societies, especially American university students, has distorted psychologists’ accounts of human nature. It isn’t just that the subject pool has been only one very specific slice of humanity; it’s that, from the perspective of globally comparative work that has been done, WEIRD subjects are frequently some of the oddest of the lot. One of Henrich et al.’s specific examples is, in fact, comparative research on the “fundamental attribution error,” which turns out, as they say, to be “less fundamental elsewhere.”

In the work of Kitayama et al., weirdness comes home: the fundamental attribution isn’t even fundamental here, in the US, if one draws from an ethnically or culturally mixed pool of subjects. One doesn’t even have to leave the class and age bracket of University of Michigan students to see the variation. This leads to a point which is supplementary to Moya’s central, and crucial, argument that “human difference really matters—but not in the way most people think it does” because cultural difference is not a product of the supposedly biological difference of race. This second point is that culture is not like race: cultural categories can’t be handled the way racial identities get handled. Otherwise we will “do culture” in the way that we still, in many harmful ways, do race.

As Moya says, in contemporary discourse too often culture is just used as a “proxy” for race. This struck a chord for me with a well-known argument in my field of twentieth-century literature: Walter Benn Michaels’s claim in Our America (1995) that US discourse (and not just US discourse) has been dominated by a racialized concept of culture at least since the rise of anti-immigrant nativism in the first decades of the twentieth century. (Michaels’s work is certainly not without problems, but I think the basic genealogy and critique of some of the most common ways of speaking of “culture” are still telling.) But if culture refers to institutions, patterns of behavior, habits of thought, customs, etc., it is clearly an error to think of it as a fixed attribute of a person.

Moya wryly tells the story of Kitayama’s career as that of a man from “the land of Interdependence” who is schooled in social science in “the land of “Independence.” Her fairy-tale style helps to underline a difference between that simplified mythical world and the one we inhabit. Work like Na and Kitayama’s shows that the United States is not exactly the land of Independence: after all, their study focused exclusively on Americans but discovered a wide variation in the use of independent or interdependent schemas of self. Can we then say that “American” culture is a culture of independence and expect all Americans to believe in the ideology of the independent self? Or that all Japanese regard the self as fully defined by relationships? It wasn’t simply the case that European-Americans were Independents and Asian-Americans were Interdependents.

The necessity of further fuzziness

After years of education at the hands of Language Log, I have learned to try to find out whether (what reporters describe as) apparently categorical differences are really average differences between variables with overlapping distributions.* Though the European-Americans in Na and Kitayama’s subject pool were more likely to score higher on a measure of independent self-conception than the Asian-Americans, both groups showed considerable variation on the measure. And the “effect size” (I believe that’s difference in means divided by joint standard deviation) for the experiments in the study are all moderately big but less than 1.0. That is to say, there is at best only a fuzzy connection between ethnicity, inter/independent self-concept, and spontaneous trait attribution as measured in Na and Kitayama’s work. The cultural variable of independent vs. interdependent self concept is not neatly, categorically explained by ethnicity.

The important result is at a tangent to the attention-grabbing question of ethnicity and the insidious cultural work of doing race. It has to do with the relationship between self-schema and trait attribution. That is the “cultural specificity” Na and Kitayama are trying to demonstrate, though again the association is tendential and noisy rather than definitive. But that’s to be expected; as Hazel Markus and Kitayama explain in a recent theoretical essay Moya cites, independent and interdependent schemas of the self are probably available in every culture, but the frequency with which people use them varies widely. These uses are then consequential for many kinds of behavior. It is this relationship among variably distributed cultural representations and behaviors which is the real object of study and the empirical core of human cultural difference, a relationship which is shaped only in part by the processes that yield what we call “race” and “ethnicity.”

To put it in a series of rhetorical questions: Do Asian-Americans have “a” culture and European-Americans another? Are those respectively Asian and European? Eastern and Western? Are they both nested within a larger American culture? What are the boundaries of “a” culture, and what is the relationship of a given individual to a culture, whether that individual has race or ethnicity done to her or not?

For me, these issues resonate with some of the most important open questions in literary studies today. First question: what is the relationship of an individual writer or reader to a widely distributed cultural representation or ideology? E.g., what does the realization that America is not 100% the Land of Independence but that it is p the Land of Independence, 0 < p < 1, tell us about particular engagements (by an individual or subgroup or a genre) with the idea of the independent self?

Second question: what are we going to do with the ethnic, racial, and national boundaries that literary and cultural studies use to divide up the cultural field? If we are not going to use those divides to “do race,” what else will we do with them? Or will we be able to invent new frameworks? The national framework, in particular, gets less and less comfortable by the day (witness the fervor of attempts to think about world literature, world history), but we’re still living with it and reproducing it.

How can we begin a conversation about cultural difference that helps all of us see complex variation rather than categorical oppositions? Perhaps it’s time to develop a culture-concept that does not fall so easily into national and racial categories. But frankly I still do not know what such a concept would be. Suggestions welcome.

But first make sure you read Moya’s post. Great contributions in the comments by Linda Martín Alcoff, Ramón Saldívar, and Lupe Carrillo, too [edit 5/9/11: and many others!]. Skip the long one by whatshisname.**

*For examples on a topic related to that of Kitayama’s work, see Mark Liberman’s posts David Brooks, Social Psychologist and How to turn Americans into Asians (or vice versa), or, even more powerfully, his takedowns of pop-psych stereotypes about gender and language use (e.g.: David Brooks, Neuroendocrinologist and Gabby guys: the effect size). I never finish one of those posts without kicking myself for not taking statistics and probability in college.

**And despite all that Arcade discussion about the values of the informality of the blog, I’ve decided to refer to everyone involved, even people I know personally, by scholarly last names. However! First names in comments, please. This reminds me of the first day of every semester.