The problem being, and this goes to Jim Weinheimer's comments as well, that in the case of non-collective titles, as the Thomale article identifies, the 245$a isn't enough. There are additional titles buried in the 245$b or worse in the 245$c. Yes, those titles should be traced in a 740 or 700$a$t. But how does one reliably program to account for this? How will a program know that an instance of 740 is for an additional title by the same author as the main entry in tag 100 vs. obsolete coding for a variant title? What would a machine do about instances when the transcribed title in the 245$b or $c does not match the authorized form as found in the 700$t? Further, reliable extraction of 700$t data for titles immediately applicable to the item in hand would be dependent on accurate coding of the second indicator as an analytic. If there is a collective title, then would one want to and how would one differentiate the treatment and extraction of title data for the contents, from both the instances of non-collective titles and "regular" listings of tables of contents? The preferences of Mac's clients aside, what is the place of uniform titles vs. transcribed titles in the entire enterprise?
In instances like these I am reminded of the advertising for a childhood game: moments to learn, a lifetime to master! This is all well and good for a game but less well so for an important enterprise. I am adept at and heretofore enamored of the previously workable compromises with which our cataloging practices produce a "descriptive bibliography-lite" enhanced by more robust access mechanisms. But having seen things from a programmer's viewpoint courtesy of Thomale, I would now tend to describe the results of our efforts as Kafka-esque -- Only this is the title, except when these other things are also titles, which are titles only when they aren't the something else they're normally supposed to be. John F. Myers, Catalog Librarian Schaffer Library, Union College 807 Union St. Schenectady NY 12308 518-388-6623 [email protected] -----Original Message----- J. McRee Elrod wrote: The discussion of extracting title from MARC has puzzled me. Since 1979 we have had no difficulty in extracting titles from MARC for new titles lists, circulation pocket and card labels, or whatever. Our clients all want the title as on the item, so we ignore 130/240. We use 245$a, on some occasions adding :$b, =$b, and ,$b (we code alternative titles as well as subsequent titles in a collection as ",$b"), depending on the product and space available.
