As recently as a couple years ago, “political correctness” was one of those phrases that triggered a reflexive eye-roll from me: Sure, it was a thing in the early 90s, but had long since morphed into a tedious gripe typically voiced by folks chagrined that nobody finds their racist jokes funny anymore. Lately, though, even many progressives in good standing have begun wondering whether our present culture of “calling out” is, perhaps, not entirely healthy. The latest comes via Jonathan Chait, and has been greeted with a mix of mockery and (more quietly) “oh God, someone finally said it.”  Two thoughts:

First: The mockery makes a certain amount of sense if you read Chait’s essay as a straight white male’s “help, I’m being oppressed!”—a reading for which Chait, alas, provides some textual support whether or not it’s what he intended. For people accustomed to seeing their opinions greeted with everything from dismissive condescension to harassment and death threats, a successful writer complaining from a perch at New York magazine about his friends being “bludgeoned… into despondent silence”—because people are mean on social media—simply sounded whiny.  Chait also moves a bit too seamlessly from real, honest-to-God censorship by public institutions to more informal social pressure in a way that makes it sound like he’s conflating them—claiming that criticism is somehow tantamount to censorship or repression.

Let’s take it as given that this is not the case, and that any subjective chill experienced by folks like Jon Chait is not usefully understood as a problem of social justice. We can grant all that and still ask: Is there something unhealthy about discursive norms that lead to substantive self-censorship, even by the obviously privileged?  Not because these norms constitute an injury to the reticent privileged, or interfere with anyone’s sacred right to unfettered self-expression, but because they yield less robust, less interesting conversations?

Think of Solomon Asch’s famous experiments in group conformity, or the broader social psychology literature on information cascades.  The problem isn’t so much that some precious snowflake’s project of expressive self-realization has been constrained, but that constraints deprive groups of deliberative input that can help them make better decisions. When the constraints are on the order of “don’t use sexist or racist language,” probably nothing of value is lost.  When the constraints include “under no circumstances express any skepticism about any claim of sexual assault,” to pick a salient recent example, you may end up with bad journalism that hinders the ultimate goal of getting society at large to treat survivors’ stories more seriously and respectfully.

Second:  The dynamic of self-censorship Chait describes bears some resemblance to one I wrote about a few years back in a post on the conservative movement—a post which, quite accidentally, ended up introducing the phrase “epistemic closure” to the political lexicon. (I know, I know: mea maxima culpa.)  Over time, it creates a feedback loop that looks something like this:  Within any political or ideological group, however reasonable and noble its aims might be on the whole, some percentage at the margin are going to take a good idea further than is reasonable, whether out of authentic zeal or because being the most hardcore is an easy way to distinguish yourself in a crowded intellectual marketplace. The more these moves prove effective in shutting down an argument, and the less the relevant audience seems to care how well they fit the specific facts of the case, the more tempting it becomes to deploy them whether they do or not.

It doesn’t take any very fearsome campaign of intimidation for a group’s self-correcting mechanisms to break down. Imagine an argument where someone invokes a spurious or unfair accusation of (let’s say) racism or sexism as a cudgel to close down a conversation. Maybe many other members of the accuser’s ideological in-group (“allies”) themselves perceive it as unfair, but life is short and people are busy—what’s the incentive to chime in and say “hey, wait a minute”? Pretty weak even in the absence of negative feedback. And as anyone who’s watched these arguments play out is well aware, questioning whether such a claim is fair or reasonable in a particular instance is going to be read by some observers as denying that sexism or racism are problems at all.  (I recently, rather gently, questioned whether one specific document from the Snowden cache should have been published, then had to expend a whole lot more words insisting that I am not, in fact, a shill for the surveillance state.  Anyone who knows my privacy writing understands why this was slightly surreal.) You end up having to explain and justify yourself to all these folks whose good opinion you care about, and who needs the hassle?

Iterated over time, though, that means the people who do object in particular cases are increasingly from out-groups: People who really don’t care about racism or sexism or think they’re serious problems.  Now the incentives are even worse for in-group members. Because now being the one to say “hey, wait a minute” in a particular instance doesn’t just mean conflict with an ally, it means associating yourself with those assholes.  Increasingly the objections are coming from people who just don’t care about the good opinion of the in-group, many of whom are expressing those objections in actively racist or sexist terms.  So now anyone voicing reservations has to do all sorts of throat clearing (I did it instinctively at the start of this post) to avoid the ever more statistically reasonable heuristic inference that any pushback is coming from those repulsive quarters. If you’re the only ally pushing back, hey, maybe you’re not an ally at all, but secretly one of those assholes.

You end up with team “x is the problem” and team “x is not a problem,” and ever fewer people prepared to say “x is a problem, but maybe not the most useful lens through which to view this particular disagreement.” When teetotalers are the only ones willing to say “maybe you’ve had one too many,” because your friends are worried about sounding like abstemious scolds, the advice is a lot easier to dismiss. Which is fine until it’s time to drive home.

You see this dynamic, in fact, with the response to Chait’s essay: Progressives who think maybe he’s kinda-sorta got a point quickly move on, ceding the field to those who want to revoke his ally card and conservatives eager to welcome him, at least for the next ten seconds, to “their” side. (I wrote a tweet suggesting I thought some of the vituperative online reactions to Chait’s essay showed he was on to something. Conservatives and libertarians retweeted it; progressives favorited.) And this makes it still easier to conclude that nothing interesting or valuable is lost by any self-censorship that may be occurring. We know what the counterargument looks like, after all: It’s the garbage those assholes are spouting. Discourse gets increasingly polarized and, in the process, stupider. Which, again, seems like a bad outcome even if you don’t particularly care whether Jon Chait gets his feelings hurt.