sorting 3898

Brian Inglis

Blame it on lack of staff. Normal card filing practice was to have the kiddies file "on top of the rod" and then have a senior cataloger (quite possibly the person who wrote the local filing rules) review their placement before the cards were permanently filed. Filers would also flag any errors of filing order they found while filing other cards, as would other library staff who came across them.

And strict alphanumerical order is not in general the correct order for library catalog filing anyway.

However, they did have rules. (In most American libraries, they were entirely or nearly standardized on the A.L.A. rules.) The annoying thing is that the rules -- very sensibly evolved over untold years of cataloging practice -- rarely (never?) were carried over into sorting of records in online-catalog display. Partly that's because some of the rules consbreastuted a pretty good Turing test that couldn't be gotten around even by the extensive markup used in machine-readable catalog records.

This might be the place to point out that libraries and library consortia have been dealing -- quite well -- with the problems of creating and maintaining multilingual machine-readable catalog records. These include non-Roman alphabet languages of all sorts. The work goes back to the early 1960s, and builds upon a hundred-year or more legacy of non-automated cataloging practice before that.

Surely most American readers here remember the beautiful printed Library of Congress catalog cards. These were, in their later incarnations, entirely "printed by steam" from the MARC records.

-- Roland HutchinsonÊÊÊÊÊÊÊÊÊÊÊÊÊÊWillÊplayÊviolaÊdaÊgambaÊforÊfood.

NB mail to my.spamtrap at verizon.net is heavily filtered to remove spam.ÊÊIfÊyourÊmessageÊlooksÊlikeÊspamÊIÊmayÊnotÊseeÊit.