Kipnis on Epistemology

Laura Kipnis's Unwanted Advances is, in important ways, a work of epistemology. It's not an academic monograph—don't file it in the philosophy section. But many of its most central questions are epistemic: given the murky circumstances, psychological complexity, and private details that characterize sexual misconduct allegations in academia, how are investigators or members of the public to know, or to come to reasonable beliefs about, what has happened?

Kipnis’s view is that the prevalent practices among student activists and campus administrators are epistemically faulty: we are, Kipnis thinks, far too quick to believe students' allegations of sexual harassment and assault. On p. 1 of the book she describes the campus status quo as "officially sanctioned hysteria" and a "sexual paranoia" akin to McCarthyism and the Salem witch trials. ("As in Salem," Kipnis quips on pp. 66–7, "the accusations of post-adolescent girls still factor heavily.")

In this post I’d like to express disagreement with some elements of Kipnis’s epistemic outlook. All the usual content warnings.

Kipnis starts making epistemological assumptions on p. 1, when she writes that irony "helps you think because it gives you critical distance on a thing". The idea that critical distance is an epistemic virtue tempts many, but it faces important limits that are too often overlooked. Some things are nearly impossible to understand from a critical distance—lived experience is epistemically critical in some areas; experiences of sexual harassment and institutional betrayal are very plausible candidates. Taking a detached and ironic view of them is not a better way to understand them. But claiming the rhetoric of objectivity is a classic technique for brushing aside one's opponents' perspectives. (And "hysteria" is textbook misogynistic propaganda.)

pic.twitter.com/ewZzj85vH6

— Alex Forrest ☝🏻️ (@380kmh) May 24, 2017

Kipnis also often uses the rhetoric of "honesty" in a potentially question-begging and propagandistic way:

• "the educational system is failing to educate anyone, largely because speaking honestly about sexual realities has become taboo." (17)
• "Campus life has gotten so ludicrous and censorious that it hardly seems worth caring. It’s far more impossible to have an intellectually honest conversation about sex on campus on an American campus than off these days, which is nothing short of bizarre if you’re someone who was drawn to academia in the first place because talking about difficult stuff was supposedly what went on there." (33–4)

Kipnis thus takes on the mantle of "telling it like it is." Like many who frame their ideas in this way, she gains credibility in many readers' eyes for saying what they want to hear, whether or not it really amounts to honesty.

When she starts talking about specific cases, Kipnis brings in more explicit epistemological assumptions. In a passage spanning pp. 66–71, Kipnis questions the methods of Joan Slavin, a Northwestern University Title IX officer who investigated allegations that Professor Peter Ludlow had gotten an undergraduate student (“Cho” is her pseudonym) drunk, pressured her to come to his apartment, and groped her. Kipnis goes over competing descriptions of several elements of the night, expressing disagreement with Slavin's judgments of Cho’s credibility. E.g.:

For instance, according to Slavin, Ludlow told Cho that he thought she was attractive, "discussed his desire to have a romantic and sexual relationship" with her, and shared sexual information, all of which was unwelcome to her. 

I’m dying to know how Slavin came to the conclusion that Ludlow wanted to have a romantic relationship with Cho. Because an evening spent drinking and going to galleries indicates a man’s desire for a relationship? If so, single women of America, your problems are over. (71)

Kipnis wonders how Slavin came to the conclusion that Ludlow said he wanted to have a romantic relationship with Cho? The obvious answer is: Cho told her. Kipnis often reaches for strange explanations for simple conclusions when she disagrees with them. Here, Slavin listened to Cho tell her that Ludlow said he wanted to have a relationship with her, and concluded that Ludlow probably told Cho that he wanted to have a relationship with her. Kipnis hypothesizes that Slavin is employing the implicit premise that an evening spent drinking and going to galleries indicates a man's desire for a romantic relationship. This is the kind of premise one might need if young women's testimony carries no epistemic significance. But if we don't assume that, there's no need to reach for such bizarrities.

Kipnis's move here is similar to one I discussed a few weeks ago, where she attributes to campus officials and activists the assumption that, e.g., female students have no sexual desires. Kipnis was assuming then that sexual assault is inconsistent with any role for the victim's agency. I think her similar maneuver here is a denial of victims' epistemic significance. She describes investigators who are speaking with the people involved as making decisions "based on nothing but their own suppositions" (p. 162). This is something that would make sense only if people's word carried no evidential weight at all.

On p. 85 Kipnis writes that "[a]bsent Slavin's prejudices about male behavior, there simply isn't a preponderance of evidence supporting Cho's story." The "prejudices" Kipnis is referring to are manifested by Slavin's taking Cho's testimony to be credible. Kipnis goes on to offer an alternate theory of the evening in question:

Let's say, just as a thought experiment, that Cho came on to Ludlow, and he told her to chill—that he was willing to spend an evening having drinks and going to galleries, but had no interest in taking it further. Is it inconceivable that a professionally ambitious young woman, after spending months emailing her fa­mous professor about this and that, finds herself in his company and decides to test the waters? Let's go further: if we're cynical enough to think Ludlow offered to promote Cho's career in exchange for sex or romance, why aren't we cynical enough to think that Cho spotted a potential gravy train and decided to play it for what she could get? After all, the gold digger and the predatory professor inhabit the same universe of crude stereotypes. 

Kipnis is writing as if the only way to figure out what happened is to imagine a hypothetical scenario and think about whether it is "inconceivable". She concocts a competing stereotype and assumes it's thereby equally credible, thus undermining the finding that Cho's story was more credible than alternatives. She assumes the testimony carries no epistemic weight.

(And talking of embracing stereotypes as evidence, let's not forget Kipnis's intial reaction to Cho's story. She started out skeptical because "single non-hideous men with good jobs (or, in this case, an international reputation and not without charm) don't have to work that hard to get women to go to bed with them in our century." On pp. 67–8 & 71 Kipnis also casts Cho's testimony as incredible because Cho describes Ludlow as drinking "disgusting student drinks" instead of proper professor beverages, and using language that, in Kipnis's opinion, sounds more like the vernacular of a "rapper" than that of a "mid-fifties philosophy professor".)

On p. 87 Kipnis writes: "In Joan Slavin's slippery-slope evidentiary standards, if he'd paid for drinks, it meant he'd groped Cho; if he was male, he came on to her." But Slavin thought he groped her because she said so, not because he'd paid for drinks. If you ignore the relevance of her testimony, then Slavin's decisions do look rather odd. Perhaps this is why Kipnis so often interprets conjunctions as entailments and inferences, as in this case. She did the same thing in an email to me in April. She'd read my blog post pointing out that, contrary to her claim, attractive and charming men certainly do sometimes rape people, and citing Bill Cosby as an example. In her email to me, she described the argument of my post as "because Cosby, then Ludlow". But obviously I'm making no such argument. To give one more example: on p. 117, Kipnis describes "Hartley" (a pseudonym for another student accuser of Ludlow) as telling Slavin that "Ludlow was 'a serial sexual predator who targeted vulnerable young women'"; on p. 122 Kipnis asks, "Did dating younger women make Ludlow a serial sexual predator, as Nola Hartley assured Joan Slavin?" Hartley says "he's F and G" and Kipnis retorts, "does being F really make one G?"

A similar dismissal of complainants' testimony characterizes Kipnis's discussion of "retroactive withdrawal of consent." She takes pains to argue (as if anyone denies this) that sometimes, sexual agency can lead to sexual choices one regrets. (One representative passage spans pp. 95–6.) On these grounds she says we should reject later reports that earlier sexual activity was nonconsensual. But the fact that something sometimes happens is no reason at all to discount testimony to the effect that it didn't happen on a given occasion. I sometimes forgetfully leave my phone in public washrooms; that doesn't mean you shouldn't believe me if I tell you someone robbed me of my phone in the washroom today.

At the end of the book there is a chapter called "Coda: Eyewitness to a Witch Trial" (also published in a slightly edited version in the Chronicle of Higher Education). There, Kipnis glowingly recounts Jessica Wilson's testimony as a character witness at Ludlow's Northwestern termination hearing. As she does throughout the book, Kipnis places evidential weight on assumptions about what "kind" of person would commit sexual violence. At the hearing, Wilson described Ludlow officiating her wedding. Kipnis writes that "[h]earing about Ludlow presiding over a marriage ceremony came as a small shock, I think, to a roomful of people who'd been told he was virtually a predator." (227) I don't want to sound overly inflammatory, but I also don't want to understate this point: the idea that respectable people who do things like preside over weddings couldn't be sexual predators isn't merely contrary to the evidence—it's actively harmful. It has contributed to the rape of tens of thousands of children.

More plausibly relevant is Wilson's testimony that her personal knowledge of Peter Ludlow is inconsistent with the behaviour he was accused of. Herself a survivor of sexual assault, Wilson described herself as "very sensitive to these issues. If Peter had a predator bone in his body, I would know it." (229) Kipnis describes her saying that "she never heard a single negative comment or even a whisper about Ludlow in fifteen years (and she would have, she said, because people came to her about this sort of thing)." (227)

I'm willing to believe that Wilson went fifteen years without hearing negative rumours about her friend's behaviour. But epistemically speaking, I think that tells a lot less than she suggests. As Sandy Goldberg lays out nicely in his book on testimony, for the fact that one hasn't heard something to be evidence that it hasn't happened, two conditions need to be met: first, you need the people around you to have reasonable access to the facts, and second, you need to be in a social environment such that, if it had happened, and people around you did know about it, they'd be likely to tell you right away.

I think it's very implausible that Wilson satisfies the second condition here. If someone knew that Peter Ludlow was sexually harassing students, whether firsthand or because someone chose to trust them with that knowledge, I'd imagine that Peter Ludlow's intimate friend who asked him to officiate her wedding would be the last person they'd choose to divulge to. They'd guess—rightly, as it turns out—that Wilson would be an extremely unsympathetic confidante.

Kipnis herself did hear about earlier rumours about Ludlow's behaviour. On p. 80 she mentions that the Northwestern Philosophy department chair had heard uncomfortable rumours about Ludlow's behaviour at a conference in South America in (I think) 2009. But it seems these rumours escaped Wilson's notice—refuting her assertion that if people had had concerns about Ludlow, she would have heard about it.

Kipnis describes Wilson as a specialist on inference to the best explanation, which she glosses as "how do you know what happened when you yourself didn't see it happen?" (229) Wilson's research focuses mostly on metaphysics, but I think Kipnis is referring here to a series of recent papers Wilson has co-authored with Stephen Briggs, exploring the role of non-deductive reasoning in theory selection and the determination of content, and emphasizing that some forms of inductive reasoning can be a priori. I'm pretty sympathetic to this line—the central thoughts there are actually extremely similar to some things Ben Jarvis and I published in our 2013 book—but I don't think it's accurate to say that it represents any particular research expertise relevant to the question of how to decide whether to believe someone's report. Their work concerns how we should theorize about inference to the best explanation—not which inferences are justified. (For one explicit statement to that effect, see the last paragraph of §1 of this paper.)

So I don't see anything in Wilson's research that puts pressure one way or the other as to whether, in a given instance, we should consider an allegation of sexual misconduct to be credible. (It's actually a controversial matter whether inference to the best explanation is even relevant to this question at all—non-reductionists about testimony might well think it's not.) Nothing in her work, to my knowledge, supports the idea that if two students say a professor assaulted them, even though some of the professor's friends can't imagine him doing that, the best explanation is that the students are deluded or lying. So I don't see Wilson's discussion here as anything more than a statement of her opinion as Ludlow's friend:

Another panelist quickly countered, "It seems very clear the students feel they have been harmed. That is definitely the message that the university is presenting, at least from the interviews with the investigators: that they feel harmed." 

Wilson's solution, again, was inference to the best explanation. "What's the best explanation of all of the data? It's a delicate position. But, I mean—we're aware of human nature. Sometimes a person's memories can be distorted. There are psychological and other motivations that need to be brought to bear." (232)

Nor does Wilson's research—nor any other plausible story that isn't radically skeptical—support the assumption that one can't know things on others' testimony. But that's what's implicit in Wilson's rhetorical question about even her own story of sexual harassment. Kipnis describes her as asking: "Yet didn't the panelists have to ask whether she was telling the truth? They hadn't been there, so how would they know?" (231)

This is also the kind of skeptical intuition implicit in the rhetoric of a "he-said/she-said" situation. I think this phrase is tied up in a kind of skeptical propaganda—it pulls us, non-rationally, towards the assumption that there's no way to know what really happened. It suggests a kind of epistemic proportionality: one piece of testimony on one side, one perfectly balancing it on the other. It's a call to give up on the project of figuring out what happened.

But not all testimony is equal—many potentially tie-breaking features can be considered. For example, what motivations would be served by lying, and who might have them? Is there relevant history? (Is there a history of false allegations? Is there a history of complaints?) Many such factors are epistemically relevant. Especially when questions matter, we should resist the skeptical temptation to just throw up our hands and suspend judgment. Sometimes that's going to be necessary, because sometimes there's really just going to be no way to figure out what happened. But the idea that this is so whenever there is competing testimony with no other direct evidence is both mistaken and harmful.