I was updating one of the many bibliographies I maintain (in BibTeX, of course!) the other day. One of the tasks when making a new entry in the bibliography is to choose keywords for that entry. This allows the bibliography management software I use (BibDesk on the Apple, JabRef on Linux; although this may change once the Zotero standalone application comes out of alpha.) to organize the entries into meaningful groupings. The list of keywords has grown in an ad hoc fashion. I never really gave much thought to it, thinking the choice of terms used self-evident.
As I was choosing a keyword for a certain entry, I found myself unable to do so. None of my lists of keywords really fit; yes, I could add more than one keyword, but that really did not work, either. It needed…more than one word. In a sudden inspiration, I looked up the book’s Library of Congress subject classification. Ah! Just what I needed. “Hebrew language — discourse analysis — Congresses”. Here are three terms, associated together in a hierarchical manner. The moral of this tale is to abandon keywords and use LOC subjects, no? Of course, if my entry had been a journal article, I would have to choose the subject myself; I could not rely upon a library science professional to do the classification for me. Nevertheless, a definite step forward. Or so it seems.
I proudly mentioned my brilliant idea to my friend, Patrick Durusau, who, among other hats, is the convener of one of the ISO Topic Maps working groups. His reaction: “Well, yes, that works, sort of. Why not create a topic map of your discipline and use that instead of keywords for bibliography subject association? It’s more general, can be more accurate and flexible than LOC subjects, and is useful for a lot more than just bibliography.”
Duh. I knew that. It would have come to me eventually. But how to bell the cat? I’m not sure yet. I’m working on it. More anon.



