William Bennett — you remember, the guy who used to be Secretary of Education before he went out and wrote The Book of Extremely Reactionary Virtues — made an even-more-ridiculous-than-usual comment on his radio show yesterday. To quote a Washington Post article on the incident:

“But I do know that it’s true that if you wanted to reduce crime, you could, if that were your sole purpose, you could abort every black baby in this country, and your crime rate would go down,” said Bennett, author of “The Book of Virtues.”
He went on to call that “an impossible, ridiculous and morally reprehensible thing to do, but your crime rate would go down. So these far-out, these far-reaching, extensive extrapolations are, I think, tricky.”

Hmm. He claims (of course) that he was quoted out of context, misunderstood, and so forth. but I’m not sure how one can misunderstand what he’s saying here: he’s linking “blackness,” whatever that is, with proclivity to commit crime, and then extrapolating a hypothetical policy based on it. The fact that he then calls that policy “morally reprehensible” doesn’t really address the central issue, which is the issue of essentialist reasoning. Call me crazy, but essentialist reasoning makes me angrier than just about anything, and opposition to such reasoning is a large part of why I’m proud to self-identify as a social constructionist and a progressive.

First, a little conceptual background. Essentialism is the philosophical orientation towards objects which begins by specifying a set of intrinsic, dispositional characteristics that an object has, and then reasons from those characteristics to some particular outcome. Quite common and relatively uncontroversial in much of the natural sciences — an object’s mass, for example, determines quite a bit about its behavior under various condition — essentialist reasoning also crops up in other realms. (I’m not really crazy about essentialism in the natural sciences either, but that’s material for another time.)

When we turn to the social realm, we can quite obviously see essentialism at work in various ethnic (“Isn’t that just like a Pole/German/Irishman/Japanses/etc.”) and gender (“Women are just nicer than men”) stereotypes. A bit less obviously, though, is the way that basic comparative/statistical reasoning is shot through with essentialist assumptions. The usual way that we make comparisons in an effort to lock down the impact of some given variable attribute (regime type, say) on a given outcome (like “proclivity to go to war with other states having the same regime type”) presumes a rather insidious kind of essentialism: the essentialism of conceptual categories.

“Being a democracy makes countries less likely to go to war with other democracies” is an essentialist claim about democracy-ness: something about being a member of the class called “democracies” makes or inclines countries to behave in a certain way. Andrew Abbott, in a brilliant 1988 paper in the journal Sociological Theory, called this “general linear reality” and pointed out that it implicitly invoked an ontology in which universals had causal efficacy that was last current in the age of Aquinas. But setting aside the historical anachronism of this essentialist reasoning, the main problem here is that essentialism denies agency. If the dispositional characteristic of democracy-ness is what causes a country to behave in a certain way, then there’s neither praise nor blame for the country doing any particular thing, because it couldn’t have done anything other than what it did. No responsibility; no contingency; no meaningful agency.

Now, consider Bennett’s absurd claim: aborting black babies would reduce the crime rate. Presumably what he’s drawing on here are the numerous statistical studies that demonstrate a significant correlation between “being a criminal” and “being black” in the United States — studies which are accurate, as far as they go — which is not very far at all. A correlation between “being black” and “being a criminal” tells us nothing whatsoever about any possible causal connection between the two factors, but in fact supports a number of causal stories: a story about police racism (there are more African American criminals because the police disproportionately arrest and prosecute African Americans), a story about poverty (African Americans are more likely to be poor, poverty breeds criminality, hence African Americans are more likely to be criminals), a story about educational networks (African Americans are disproportionately clustered in inner cities, where the schools are not as well networked with any number of opportunities; hence students in those environments are unable to take advantage of the same opportunities available to students in more densely networked environments, and turn to crime as a last resort), and so on.

And yes, the correlation also supports the essentialist claim that “blackness” is what causes criminal behavior. Actually, it does so in a rather profound philosophical way, since essentialist reasoning is basically hard-wired into standard statistical analysis. The very idea of testing a hypothesis based on correlations between categorical memberships is inextricably essentialist, even if the stories we then tell about that correlation involve context and nuance and an appreciation of the social environment. In this way, Bennett’s statement expresses not just simple racism (although it probably does express that as well), but also the purified character of statistical/comparative reasoning.

This is even clearer if we consider Bennett’s “solution” to the crime problem. Aborting black babies will only decrease crime if “blackness” causes crime. If other factors contribute to criminality, then Bennett’s “solution” probably wouldn’t make the slightest bit of difference — since aborting black babies would decrease “blackness,” but none of those other factors. And, even worse, if one did in fact believe that “blackness” caused criminality, Bennett’s “solution” would make sense, and would in fact be difficult to avoid: if X causes Y, and Y is undesirable, then eliminating X follows more or less inevitably. Here we see another ridiculous consequence of essentialist reasoning — rather than calling attention to social factors like poverty and network connectivity, we are led to more or less seriously consider eliminating whole categories of people. [At that point, elimination through abortion or elimination through deportation or whatever becomes a purely technical question.]

Putting these issues together we have support for Ian Hacking’s incisive observation that essentialism is effectively the strongest version of inevitability. An essentialist claim denies agency, since the outcome is derivable from the essence of the object; thus the only way to eliminate the outcome is to eliminate the object. But more importantly, the whole conceptual apparatus of essentialism is pretty much inherently conservative, since it denies the very possibility of fundamental change: person X can only do what her or his essence demands.

I don’t buy any of that. I have no idea what it would mean to identify the essence of a person, and absolutely no clue how one would ever know whether or not one had correctly identified such an essence. In fact, when I look around at social life what I see is a complex concatenation of social practices that continually produces and reproduces persons and organizations and the like — no essences needed, and none in evidence. To me, this is part of the essence of a progressive politics: the preservation of a meaningful notion of agency through the denial of (and indeed, the active effort to combat) essentialist reasoning. So Bennett’s comment is both substantively absurd and politically regressive.

[cross-posted at Progressive Commons