The Multiple Meanings of "Privilege"

I've been ruminating a lot about the concept of "privilege" recently. The inciting incident was a post by Matt Dillahunty in which he said:

I’m not even sure I can comment on that, as my own privilege might be preventing a clear understanding.

I noted at the time that this was an unusual, and perhaps problematic, comment to hear being uttered by a freethinker/skeptic, as it strongly implied that the concept of privilege is beyond the bounds of rational analysis. Then, last Thursday, there was a mini-debate among various factions of the blogosphere as to whether Conor Friedersdorf's reasons for not voting for Obama made him a privileged douchenozzle or whether the argument from privilege actually worked in his favor. All of which served as preamble to an article I read yesterday about the inherent racism of mentioning sandwiches in a multi-ethnic school.

No, seriously... here's a summary quote:

According to Gutierrez, using the example of a peanut butter sandwich in classroom lessons is technically a problematic and discriminatory move — one that was made by a teacher in her building last school year. While such a notion may bring out laughs among those who find it absurd, the principal explains her logic.

“What about Somali or Hispanic students, who might not eat sandwiches?,” she said. “Another way would be to say: ‘Americans eat peanut butter and jelly, do you have anything like that?’ Let them tell you. Maybe they eat torta. Or pita.”

On the face of it that seems to be a highly questionable claim. Surely there can't be anything wrong with mentioning sandwiches, right? And yet, at the same time, it may cause minority students to reflect on their own difference, triggering stereotype threat and thus depriving them of equal educational opportunity. Anyone who holds otherwise is blinded by privilege.

Am I just making shit up? I don't know!

The above example is pretty ridiculous on a number of counts, but it nevertheless highlights a serious issue: How do you go about testing claims of privilege? The problem is that reasonable people can agree that privilege exists in the abstract, but disagree about whether its a relevant factor in any particular instance. If the invocation of "privilege" isn't to automatically forestall discussion there needs to be some way to determine when privilege is and is not at work.

I did a non-trivial amount of googling yesterday and today and have been able to turn up precious little writing on this topic. Crommunist wrote a post awhile ago talking about the mechanism of privilege in some detail, and even goes so far as to say that it "bothers deniers because it seems like a non-falsifiable hypothesis ", but doesn't actually address that objection.

However, that foray offered me the chance to do a bunch of reading on the subject which is helping me to clarify things in my own mind. It seems that part of the problem is that the word "privilege" is used to convey multiple, somewhat distinct concepts:

1. An umbrella term describing the operation/effects of a bunch of -isms. This would be the context that Crommunist is addressing.
2. A type of epistemic closure which prevents members of favored classes from accurately perceiving reality. Dillahunty's statement would seem to be an example of this.
3. A worldview and set of advantages which adheres to members of favored classes by virtue of their membership in those classes. This is the type of "privilege" cited by Erik Loomis in his article on Friedersdorf.

Now it feels like we're getting somewhere. The sandwich example above would fall under definition 1; we can argue back and forth about what (if any) -isms are in operation, but at least we have a toe-hold for analysis. And, in theory at least, its possible to test for the operation of aforementioned -isms. Definitions 2 and 3 are the ones which have the characteristic of being non-testable.

And now, a grand realization: I suspect this is why "privilege" has caused such problems in the skeptical community, because it's a single word which encompasses both testable and non-testable concepts. Both sides can talk about "privilege" and have different ideas about what they're actually discussing.