One of the interesting features of the Polanski conversations around the internet is the way the director’s defenders emphasize that he pled guilty to statutory rape while the prosecutorial-minded among us emphasize that he is guilty of rape. And if the victim’s grand jury testimony is accurate, this was a case of rape regardless of her age. (That she kept saying no was one clue.) This leads to some thinking about statutory rape and why people tend to find it…if not exculpatory, at least a mitigating factor, not-quite-real-rape. I should note that I’m not endorsing this, just wondering why it happens.

First, it’s worth noting that “statutory rape” covers a lot of moral ground. Clearly seriously wrong case of SR: sexual activity with a person who is obviously not old enough to give consent, e.g., a ten-year-old. A morally innocuous case of SR: two competent people who want to have sex with each other and do so, though one partner is just over, and the other just under, the relevant legal lines. (Supposing there’s no “Romeo and Juliette” exception to the law– or you can move the ages to be just out of reach of that exception.) My hunch is that if you asked people to describe a generic case of statutory rape, they’d come up with something like this.

This, in turn, I suspect, leads people to think of SR as wink-nudge rape, i.e., not really rape. And this is true of some SR instances, e.g. the second of my two examples above, though it’s clearly not true of all of them. (Though I wonder if the harm done to the really young victim and the harm done to an adult victim are different in kind if not in seriousness because of the way these two people are likely to conceptualize sex and violation.)

So let’s take a detour through the concepts-as-prototypes literature, one of the few psychological research programs inspired by philosophy, in this case, Wittgenstein’s remarks about family resemblance in the Philosophical Investigations. Wittgenstein noted that with some concepts, e.g. game, there’s no one set of necessary and sufficient conditions for something to count as an instance of that concept. Games are a really diverse group, and Wittgenstein suggested that their similarities are more like a family resemblance– one feature shared by some, another by others– than like a group united by a single feature or defining condition. Another common metaphor is that of a rope no one strand of which continues from one end to the other. Maybe our psychological representations have the same structure: more like a cluster of features and less like a list of conditions. I think the picture is that we’d have a prototype– a clear and obvious instance of something that falls under a concept– and then classify things by thinking about how they relate to that prototype.

There are some suggestive empirical results. People are quicker to classify clear cases of Fs, slower to classify unusual cases, for example, and they identify clear cases as “better examples.” I think–please keep in mind that this is not my area and I refuse to look things up in books when blogging– that the same prototype concept structure would explain conjunction-fallacy results like subjects’ willingness to judge the probability of “Linda is a bank teller and she is active in the feminist movement” as, impossibly, higher than the probability of “Linda is a bank teller.”

So is this what’s going on in the rape case? Thinking of “statutory rape” brings to mind the most salient examples– people having sex with almost-old-enough consenting partners. And those cases are far from the prototype of rape that most people carry around in their heads (the violent stranger cases). Hence it doesn’t seem as wrong. As a result, describing it as “statutory rape” is a rhetorical move that reduces the judged degree of wrongness. Of course, this description is true but misleading, as if I killed a person and said “I just killed a mammal.” (It violates a coversational implicature, in other words.) It brings to mind (I’m guessing) the most-permissible case of SR, which this case is clearly not.

This would also make sense of the mistake (I think) people are making when thinking of SR as not-real-rape. Here’s a big problem for the concepts-as-prototypes view: it doesn’t distinguish between features typical of Fs and those that are constitutive of Fs. A penguin is just as much a bird as an eagle, though a penguin is an unusual bird. A large prime number is just as prime as 3 is. (Yet the concept of prime number exhibits the same prototype effects of classification time, ranking as a “good example” and so on.)

So I suspect something like this is happening: a prototypical rape involves a stranger and physical coercion. Other cases– acquaintances, underage partners, and so on– are often quite a bit of representative distance from that prototype, and so people tend to take them less seriously, or think that degree of distance translates to degree of permissibility, or something like this. But the typical features and the wrong-making features are different, so representational distance from the prototype doesn’t track degrees of wrongness.

This might be all wrong on a number of levels, of course.