Selective memory

I've been involved in a debate on wikipedia recently (Avoid Academic Boosterism), and it raised to me a longer-standing question about the nature of reference materials (including maps). As I said on the wiki,

I've been trying to figure out what's really going on here, and not just here but in general in the whole "selectivity" business in college guides and so on. Here's what I think: It is not seemly and polite to talk about a college's "mythos," but that's what's going on. One of the important things about Harvard is that "Harvard Aura" and the same is true of other "selective" institutions. You go there, you know you're hanging out with future Nobel Prize winners, or at least with people who can plausibly sound like future Nobel Prize winners. And this is known in the public at large.

So my question is, how can we talk about this in wikipedia? Some colleges have "Wobegon University in popular culture" sections,but these are mostly lists of mentions on TV. Seems to me this is the place to mention "aura", and in some cases there's specific examples to bring up: Robert Pirsig and the University of Chicago, Paper Chase and Yale. For smaller schools, not so much. I went to Carleton College, which has the reputation as the highest-caliber college in Minnesota. But there's no movie or popular book that backs this up, and no news outlet wants to tick off alumni of other places unnecessarily by saying things like "Minnesota's top college". Maybe reference here to less-rigorous college guides (like College Prowler or The Insider's Guide to the Colleges) is in order, under the rubric of "How Carleton College is talked about," separate from verifiable stats.

The point is, if we can find some way to talk about reputation that isn't the article defining that reputation, I think that will get at a lot of the underlying issues here.

Which gets us right back to "what kinds of things can you put in encyclopedias" or maps or other reference works? By classifying things and then only accepting those classes of information that can be part of a communication pidgin (verifiable, supportable, can-be-agreed upon), we leave out a whole lot. But by including those things, we lose the cross-community communication that reference material allows.

It's a dilemma. Well, its the same damned dilemma I keep talking about here, from another angle.