This morning I finally got around to reading Thomas Mann’s most recent essay, “The Peloponnesian War and the Future of Reference, Cataloging, and Scholarship in Research Libraries” [.pdf].
HIGHLY recommended for everyone who has anything to do with research libraries.
Recommended for anyone because just in reading it, I learned one bibliographic research trick I didn’t know before (which could have been saving me LOADS of time recently).
Basically, this essay is a passionate yet reasoned, articulate argument that the direction many researchers/theoreticians in our field and library managers seem convinced is the right one for research libraries is an assault on the culture of scholarship and the ability to conduct scholarly research.
Keyword searching, relevance ranking, folksonomies, federated search, etc are useful additions to our systems, and are obvious good solutions for the Web. But they are not acceptable substitutes for professional subject cataloging and all of the structure it brings to the catalog and the library’s carefully built collections.
It is dangerous to conflate the the needs of person doing a quick information search with the scholar doing scholarly research, who engages in an intensive process of iterative information seeking and knowledge building.
Libraries are based on principles that serve the needs of scholars. Are we ready to admit that scholarship is archaic, unnecessary, and not worth supporting in today’s world? In today’s market? It’s not sexy. It’s not quick and easy. The cash value of it isn’t readily apparent.
I think LIS educators should definitely read this, and not just those who teach subject cataloging. It is highly relevant to reference and bibliographic instruction as well.
We need to continue to teach and champion the power and relevance of the principles on which libraries are based, without either clinging to the way things have been done in the past or claiming that everything needs to change.
(and this post is an example of why i don’t blog more. i’ve spent far too much time on it and it is still all over the place and doesn’t make my point well. practice?)
It’s a great article; I agree with most of it, but there are a few points he makes that I can’t quite buy in to.
For example, I’m not convinced that pre-coordination can only be thought of in terms of strings; it’s the encoded relationships that are really important. It’s possible to make headings more human compatible, without losing that structure.
The subject heading he ultimately identifies is “Finance, public–Greece–Athens”, which is about as unfriendly a heading as you can get; distinguish “Finance, public–Athens (Greece)”.
The chronological relationships aren’t explicit (Greece–Athens puts some limits, since the empire expired, but the war itself covered a much smaller timeframe.
my strategy for this question would have been to do a keyword search on “Peloponnesian War” (expecting and finding Thucydides and Kagan). From there, I’d look for subdivisions of that books main heading (“Greece — History — Peloponnesian War, 431-404 B.C. — Finance.”), and indeed discover several works classified under
“Greece — History — Peloponnesian War, 431-404 B.C. — Finance.”
All of which supports his main point as to the importance of subject analysis, but which doesn’t necessarily support keeping the subject headings unchanged.
It’s possible to make headings more human compatible, without losing the all important structure. I’m not sure I know *exactly* what you mean, but I would agree. Relationships between the subdivisions could be represented in clear, human ways with the machine readable stuff behind the scenes, etc. You are the nuts and bolts expert on this kind of thing.
It is POSSIBLE but we don’t have it now. And LC seems to be moving swiftly (relative to normal library speed anyway) to make subject analysis cheaper, easier, etc without a better solution in place. Which is what scares me.
What I’ve read is not “Let’s do away with subject headings and come up with an advanced ontology based thing that would make headings more human compatible, without losing the structure and relationships.” It is more “all these structures and relationships are archaic and too complicated and expensive. Besides, we can just keyword search and relevance rank and have tags and everything will be sufficient.” No no no no no.
Or maybe I just have a skewed and paranoid view of the discussion?
I don’t think Mann ever said that LCSH was Truly Good the way it is, or that the headings shouldn’t be changed. Anyone who works with the headings is beat over the head with inconsistency and problems. But having them with all their warts is better than getting rid of them if there is nothing better to replace them.
I think that’s the crux of it. By all means, develop better ways to meet the demands of scholarship, but don’t stop meeting those needs because you are conflating scholarly research with the quick queries people are doing with Google. Yeah. anyway…