Thread by @magi_jay

1/ This Op-Ed by Bret Stephens is intellectually flaccid, at best, and morally reprehensible, at worst.

It’s written as a superficial appeal to our “reason,” while, in fact, Stephens repeatedly argues that we behave in a fundamentally irrational manner. nyti.ms/2zngCL0

2/ First, let’s just talk about how flawed Stephens’ “argument” is. . . . .

3/ Basically, Stephens is arguing that, in the absence of *specific* corroborating evidence as to an event, we must give *equal* weight to two people’s claims regarding the nature of that event: in this case, an alleged sexual assault.

4/ No person who was seriously interested in the pursuit of truth would make such an argument.

5/ In weighing the relative truth of two statements, from two different parties, we can *very* reasonably draw on general information, such as probability, motivation, and indeed, a person’s credibility.

Referring to such information is, in fact, *rational*.

6/ In the case of sexual assault, the probability of a false accusation is low. Research indicates that, at least in cases reported to police, “false” accusations comprise about 2-8% of all accusations, which is similar to the rate of “false” accusations of other criminal events.

7/ (As an aside, there have been issues w/ how “false” is coded by police or researchers. There are cases where an accusation may be coded as “false,” simply due to lack of evidence.)

8/ So, *overall*, our best evidence points to a low probability of false allegations.

9/ Conversely, we *also* have research indicating that men (and women) who are accused of sexual assault often *lie* about having committed the assault.

The probability of *denial* of a crime is actually rather high.

10/ In fact, we even have research indicating that a subset of sex-offenders will admit to every specific behavioral action they committed during an assault, but will still not *label* these behavioral actions as “assault,” “rape,” etc.

11/ So, overall, there is an *observed* low probability for false reporting on the part of alleged victims. There is a much higher probability of false denial on the part of perpetrators.

12/ It would be foolish for us to disregard these known probabilities in a general pursuit of truth.

Remember, we’re not talking about a court of law here—or a possible criminal conviction.

We’re talking about what we believe is likely to be true.

13/ We're not talking about the evidence we use to decide whether someone should go to jail.

We *are* talking about why we say #BelieveWomen & #BelieveSurvivors

14/ Now, again, contrary to Stephens' "argument," we *also* rationally take other things into account when we weigh disparate claims from two sources, such as the possible motivations underlying each claim.

15/ There are few motivations to claim to be a victim of assault. Little is gained. Much is lost.

There is all the motivation in the world to deny having perpetrated an assault.

16/ Please note, again, that I am not saying these relative considerations *entail* the alleged victim is telling the truth & the alleged perpetrator is lying.

I am saying that, in a general attempt to make sense of the world, it is rational to consider this information.

17/ In a much more concrete sense, it is also rational to consider the relative *credibility* of the two people making opposing claims. . . .

18/ Does Dr. Ford have a history of lying? In particular, does Dr. Ford have a history of lying about bad acts committed against her?

We have seen no evidence that she does.

19/ In this same vein: Does Kavanaugh have a history of lying? In particular, does Kavanaugh have a history of lying about having *committed* bad acts?

We have seen evidence of this. Including, his claim, under oath, that he did not receive stolen documents.

20/ Now, even if Dr. Ford had a history of dishonesty, this would not *entail* she was lying about having been assaulted. People who regularly tell lies can be assaulted, just like anyone else.

21/ And, similarly, Kavanaugh’s *documented* history of lying does not entail he is lying about this event.

Liars can lie about some things and tell the truth about other things.

22/ But Bret Stephens would have us outright reject this evidence in our evaluation of the relative claims made by two individuals.

Ignoring relative credibility is an *irrational* behavior if you are *actually* interested in truth.

23/ Now, notably, Stephens advanced his "there is no way we can make a judgment here!" argument by relying on a fallacious conflation between 1. an acquittal in a criminal trial and 2. a lifetime appointment to the Supreme Court.

24/ B/c Stephen's relies so heavily on this conflation, I will say, that my judgments above *do* not and *should* not apply to criminal proceedings.

I am talking about how we make *real* world judgments in the presence of conflicting information from two sources.

25/ Would I argue, for example, that, as a juror on a rape trial, you should say, “Well, generally, there is a 92% chance that what she is saying is true based on known probabilities, so let’s convict!”

Hell no.

26/ And people need to *stop* acting as if, when we say #BelieveWomen, we are making this argument.

27/ My belief re: what is *likely* true is not the same as my decision to acquit or convict.

And, importantly, the consequences of Kavanaugh *not* being appointed to the Supreme Court are in no way equivalent to the consequences of a criminal conviction for sexual assault.

28/ So, in short, Stephens’ piece is idiotic. He espouses very juvenile beliefs about epistemic judgments.

And he puffs up his naivety through multiple conflations between how we judge what we believe is likely true & how we operationalize this knowledge in a criminal trial.

29/ Finally, I also said this piece was morally repugnant and, indeed, quite dangerous.

Why?

Because it places undue stress on the idea that a victim of assault must produce a bevy of corroborating evidence to even come forward.

30/ Stephens even suggests that providing such evidence is, in some way, a moral duty.

Let this paragraph sink in:

31/ The kind of evidence Stephens demands is not common.

Sexual assaults usually occur w/out witnesses.

And, even in the presence of physical evidence (e.g. semen), evidence is often not proof of an assault, as perpetrators can argue the encounter was consensual.

32/ In this case, Stephens is not just arguing, “W/out absolute proof, we must appoint Kavanaugh,” he is explicitly saying that if a victim is to make an accusation, the victim themselves has a MORAL responsibility to provide evidence that the assault occurred.

33/ Stephens then goes on to claim that a person only "deserves" to be accused of assault if he can disprove that he did it, not just deny it.

34/ Thank GOD that there are few teenage boys or girls who are reading the NYTimes Op-Ed page. But their parents are reading it. Their teachers are reading it.

And all of these messages about the *moral obligations* of accusers get transmitted to past AND future victims.

35/ So, overall, Stephens’ piece is intellectual malpractice at best, & insidiously dangerous at worst.

It is, for example, absolutely rational for us to draw on information such as probability, motivation, & credibility in a pursuit of truth.

36/ And, most importantly, no victim of sexual assault should be held accountable—morally or otherwise—for the evidence that they can or cannot produce.

The idea that they should be is one of many factors underlying why victims are afraid to report in the first place.