[Just a quick note that the giveaway in anticipation of The Cancel Culture Panic is still ongoing! I’ve mailed the first 15 books (a few more are coming, sorry to those who are waiting) — but I have a few more left. So hurry and get yours while supplies last.]
Okay, this is my (super tame) experiment with a provocative title. I don’t have any business giving anyone lessons in statistics. A fun fact about me: Statistics is the ONLY class I have ever failed (it was at 9 am my sophomore year and I tried to not go and work just from the class handouts. It … did not work). But in diving into the cancel culture panic (for The Cancel Culture Panic), I’ve spent a lot of time thinking about how people use statistics. I’m of course not alone in this — taking apart or debunking the kinds of statistics that subtend cancel culture narratives is something of a cottage industry. And in doing that you come up against the almost totemic way in which people wield these statistics. But today I want to focus on something slightly different. It’s about what we’re taught not by what the statistics say (40 disinvitations in one year!), but how they teach us to pay attention and where.
There’s a classic Simpson’s joke about statistics:
I take this joke to be about something well all know: that people use, yes, even gesture with statistics. They drop in a figure to sound authoritative, to stump their interlocutor. But they do other things too: they use figures to substantiate certain intuitions and dismiss others; they use them to indicate who in society deserves being a statistic, or in other situations is important enough to be a statistic. One of the things I spend a lot of time on in The Cancel Culture Panic is that it’s likely no accident that cancel culture arrived in public discourse within one year of #MeToo and more or less simultaneously with the most dramatic wave of #BlackLivesMatter protests. “Fear of cancel culture is #MeToo for people who are afraid of #MeToo.”
What I meant by that is threefold: (1) Especially in its explicitly right-wing version, cancel culture discourse substitutes its own anecdotes for the (much more numerous) anecdotes that largely drove #MeToo/#BLM discourse. It “floods the zone with shit”, whataboutism and gaslighting. It’s just there to clog discursive drains. Oh, you’re upset about Breonna Taylor? Well, here’s a thing that happened to a conservative judge at Berkeley, kind of. (2) Cancel culture anecdotes in many ways resemble #MeToo/#BLM anecdotes. They substantiate more systemic fears just like George Floyd’s death did — with the important difference that the systemic fears cancel culture anecdotes claim to substantiate are usually vastly overblown. You know the fears I mean: that as a “cis hetero white guy” you can’t say anything anymore without everyone getting offended, etc. etc. The relationship of anecdote and the systemic issue it’s supposed to prove is roughly analogous in both cases — except that in the case of #MeToo/#BLM that relationship is based on decades of reporting and research. And in the case of cancel culture it’s based on decades of just vibes.
Finally, though, (3) anti–cancel culture discourses ultimately don’t care that much about anecdotes. As I write in the book, they “poison the well” where anecdotes are concerned. “The fear of cancel culture is a kind of malicious parody of the #MeToo narrative: slapdash where the #MeToo story needs to be airtight and interchangeable where the #MeToo story centers the individual woman.” The point is: cancel culture anecdotes and statistics teach us about whose experiences are considered valid, are granted universal (or indeed simply broader) validy — and, just as importantly, whose do not.
This is something that came up for me when listening to an excellent recent episode about Steven Pinker’s The Better Angels of Our Nature by the If Books Could Kill-duo Michael Hobbes and Peter Shamshiri. They discussed a passage from Pinker’s chapter on racial violence, which I had frankly forgotten about. Here’s the passage.
There are, as Michael notes on If Books Could Kill, some problems with that assertion just on its own. Hate crime statistics are extremely dependent on local law enforcement, on reporting infrastructure, and prosecutorial choices. Measuring something as notoriously difficult to measure as hate crimes against a form of crime that we have actually fairly good measurements for (after all the room for interpretation is fairly minimal if you have a dead body on the street), certainly feels like it’s suggesting that perhaps we’re making too much of a fuss about these supposed 5 cases per year. And even though Pinker goes on to emphasize that obviously even one hate crime is one too many, it’s hard not to hear more than a little dismissiveness in the “statistical noise”. Pinker goes on:
You get the idea. Again, I don’t want to belabor this point. I instead want to compare these paragraphs to the one this passage reminded me of. It’s one I’ve written about before (and that’s in my book):
This is from an op-ed that appeared last year in the Boston Globe. It’s by — you’ll be shocked to learn — Steven Pinker (and Bertha Madras). Again, I went over the passage and why its way of using statistics is bad in my newsletter last year. But seeing the two passages back to back I think you get a sense that there’s also a lesson being imparted by the way he gestures with statistics. You can tell that everything that Pinker so spectacularly does with the hate crime statistics (exploit relative versus absolute frequency of a phenomenon, for instance), he just as spectacularly declines to do when it comes to cancel culture numbers. The number of US colleges almost quintupled between the beginning of the McCarthy era, the number of college professors that could be “punished for expression” increased tenfold. Seems like the kind of thing a guy like Pinker might want to sprinkle into his argument. Pinker almost certainly knows that he could. He might even assume that you sense it too, and notice him not doing it.
But that would only matter if that were a genuine inconsistency. But it probably isn’t. These statistics are meaningful on a meta-level: they teach us how to treat very similar-looking types of statistics differently. Some events deserve relative frequency and contextualization; others do not. Some deserve to become drops in a bucket of similar events (things that happen to minority groups Steven Pinker happens not to belong to), others get to stand on their own (things that happen to “tenured professors” — a group Steven Pinker very much happens to be part of). These statistics aren’t really statistics at all, in a way. At least not in the way they function. They are little lessons in empathy, empathy granted and withdrawn. Lessons in how to deal with the pain and discomfort visited on others. And how to decide whether their pain and discomfort bears dealing with at all.
I have described statistics — the one math-y thing I’m really good at doing — as narratives told using numbers. If the narrative is wrong, the numbers will be useless. Pinker refuses to analyze and discuss the underlying facts — the historical narrative — that he summarizes using statistics. Since he has a number thrown in there, most people assume he’s Smart and will refuse to think any further.