inheriting from FRBR, merrily smashing some things

This post over at Everybody’s Libraries captures, I think, the things I like about FRBR, while neatly throwing out the things I don’t. Go read it, it’s fun; I’m going to toy around with its concepts below.

So — instead of being keen on the “work, expression, manifestation, item” idea — Ockerbloom’s model seems to be pure inheritance. Objects {p, q, r} belonging to class A inherit metadata attached to A, and may then elaborate on it with their own specific metadata. So, for instance (and using his example), class A is “Hamlet” (including such metadata as “Hamlet” and “Shakespeare”, coded however you like), and p, q, and r each have that metadata as well as edition-level stuff like “published in 1966” or “translated into French” or what-have-you.

I made a graph!

As he goes on to note, this inheritance process can go on at as many levels as you like. For the Bible, it doesn’t make as much sense to go from that top-level idea to specific editions as it does to go from “the Bible” to “King James and Douai and whatall” to specific editions.

One thing I like about this is it breaks the idea that there is a level which is Work, and another level which is Expression, and so forth. The boundaries of these levels were always vague, and part of the reason was (as with the Bible) there’s no reason to think that any two documents have…ideational and textual traditions?…which can fit into the same sets of boxes. It simply takes more levels of information to situate some books in the bibliographic universe.

Another thing I like is that it’s presented simply as inheritance, and implicitly (albeit admittedly not necessarily) monohierarchical. FRBR always seemed to me to blithely assume that the entirety of textual and literary criticism could be carried out by catalogers. Trace all the influences of works on other works? Sure, why not! …no, actually, that is a hard problem, and people spend entire careers on it for tiny spheres of works. (My best friend got a Ph.D. for, like, prolegomena to this stuff. Seriously. From Oxford.) Encoding that kind of information, by default, in a library catalog, is….well, it makes sense in whatever world the Semantic Web makes sense in: you know, the one where we have utopian ideas without any reference to how we would implement them.

I was torn, in library school, between loving the cataloging theorists — far and away the most intellectually engaging part of the degree — and wanting to punch through some walls because catalogs must, at some point, be implemented, and if the discussion about theory doesn’t crosscut with practice, doesn’t iterate between principle and implementation, then even the principles are flawed. But here’s this pragmatic little model of inheritance, ripping from FRBR its most useful part (the ability to ignore distinctions among documents when not salient, and not ignore otherwise), and tossing away the most impractical. Lovely, lovely. And I got to make a graph.

in which FRBR clarifies my thinking on citation styles

Hey, I finally figured out that thing about citation styles that annoys me. Basically, it’s FRBR.

Let me back up a moment. Back when electronic content was starting to explode, lots of citation styles were getting all persnickety about how to cite the electronic vs. the paper version of different things, and which database it came from, and all this crud. And I was thinking, why? Do I care? Does it really matter where I found an article? What possible way does its provenance matter to my argument?

In other words, I’m really not interested in item- or manifestation-level citations. The kind of arguments I make — the kind of arguments people in most disciplines, I think, make — are expression-level, caring only about the content in question and not the particular form in which it’s realized.

It reminds me of some of the discussion at The Past’s Digital Presence about the Google Books digitization, which went off in the opposite direction — that, by treating books solely in terms of their intellectual content and treating physical distinctions among items as irrelevant or uninteresting, Google Books was stripping out a vital part of the historical record. And that’s true, too — there are kinds of scholarship for which you need to see how history has nicked and scratched a particular object. There are kinds of scholarship where subtle differences among versions are important. And for those kinds of scholarship, we need both access (one of those distinct advantages of libraries, by the way) and citation with fine levels of granularity. Even in everyday but monograph-heavy scholarship, where we’re going to be citing page numbers, we need enough edition-specific description to contextualize that (except where there are discipline-specific conventions for avoiding that — yay, classics!).

But most of the papers I’ve written? I’m reading journal articles, and it really does not matter where I accessed them. So, dear citation formats of the world, thank you for noticing, and chilling the heck out a bit about this.

(Why, yes, my entire life has been eaten lately by putting together a paper for the LITA/Ex Libris student writing contest…it’s a good thing I didn’t realize in advance that “3000-5000 words” meant I would be writing a 20-page paper in the scraps of time during the 2 weeks when my daughter, presently on spring break, was asleep! Because, I mean, that’s impossible, and if I’d known it was impossible…

…oh hell, I would’ve done it anyway.)