Thursday, January 26, 2006

An online catalog revolution?

Librarian Karen Schneider of the Free Range Librarian and the Librarians' Internet Index has some interesting things to say about North Carolina State University (NCSU) Libraries' new online catalog. The catalog is fantastic. However, I don't agree with the idea of pitching the controlled vocabulary used in our online catalogs. Since our catalogs don't contain the full texts of books, there is no real way a Google-type search could work. Indeed, that is one of the faults of Karen's argument: the online library catalog is not the same thing as "Google, Amazon, A9, AskJeeves, Technorati, [or] Google Book Search." (Google indexes billions of Web pages; my library catalog contains descriptions of a relatively small number of books that sit on shelves.) The controlled vocabulary used in the catalog, in the form of Library of Congress Subject Headings, provides additional searchable metadata and a greater degree of precision to our searches. Sure, the "aboutness" of any information resource is disputable; but generally, LCSH has served me well. The revelation occurred to me when librarian Thomas Mann, in his book on library research, pointed out the problem of a catalog search without standardized subject metadata: Imagine performing a keyword search using the phrase "death penalty. Without controlled vocabulary, you would miss resources that use different terminology, such as "capital punishment."

Yes, catalogs aren't perfect, and I want NCSU's catalog at my library. Subject-level cataloging is horribly expensive. But in a world with varying terminology and catalogs that contain records that are meant to represent physical objects, I think subject cataloging is necessary and will always be with us.


Barbara said...

One thing that struck me, browsing through Flickr the other day, was that people often assign eight or ten tags to an image. Why is it so expensive to tag books? Is it because it's too hard to look closely at the book? (Most books spell out the contents pretty thoroughly in the intro material. You don't have to read the whole thing.) Is it because LCSH makes it hard to find a heading that makes sense? Is it lack of nerve? Of imagination? I don't get it.

Here's my proposal: spend no time describing the book as a physical object and instead describe its content. I'd trade centimeters and page numbers and however many roman numeral pages there are for an extra subject heading any day.

Finding books by subject is the hardest challenge for searchers. Why spend money on books, and then hide them because it'll take an additional 15 minutes to add a few more headings?

Full disclosure: I'm not a cataloger. But still..... sheesh...

Alec said...

Frustrated with catalogers, Barbara? ;-) Deanna Marcum, a librarian at the Library of Congress, makes a similar argument in an article in the most recent issue of Library Resources & Technical Services.

Amazingly, she notes that the LOC invests $44 million in cataloging each year!

Barbara said...

I think I'm more frustrated by whoever sets the rules and says "two subject headings is enough." Amazing, that 44K figure - but there is such an increase in the numbers of books being published, maybe it's the cost of having enough staff to do even minimal cataloging... it does make me scratch my head. And with more people tagging their own stuff, there may be increasing dissatisfaction with the stingy work we do with headings.

Maybe I should sit down with a few books myself (ones I haven't read) and see how long it takes me to tag them with LC headings.... my last cataloging class was in the days of typewriters and AACRI!