Hello, fellow BIBCO catalogers,
As part of a maintenance project to add OCLC numbers to older, vendor-provided bib records, we're running into records where the title on the book-in-hand doesn't quite match the DLC/PCC record it "belongs" to, or they have minor pagination differences.  We're also finding "duplicate" records (often batch-loaded to OCLC) for the "same" book with the spelling/pagination corrected, and they are often created by a PCC library.  We surmise this could be due to CIP cataloging not being fully corrected when the actual book is received, the result of "fast cat" processes for LC/PCC records, and/or catalogers not realizing minor variation in pagination is not justification for a new record (barring the presence of other differences).

We've been debating whether we should correct the LC/PCC record or not, given the presence of the "duplicate" record and the fact the incorrect record is held by many, many libraries, apparently without correction (we've searched a number of catalogs to determine if it's only incorrect in OCLC, but found the OCLC record used as-is in all we checked).  Correcting the PCC record should allow OCLC's duplicate detection program to match and merge them, but the fact so many libraries have holdings on the "incorrect" record gives us pause.

Some examples:
# 51756087, Themes in Neoplatonic and Aristotelian logic : order, negotiation, and abstraction / John N. Martin (LC record, held by 141 libraries).
The subtitle on the book-in-hand is actually "order, negation, and abstraction," and is reflected in the batch-loaded record #316733623 (held by 15 libraries).  Everything else appears to match (LCCN, ISBN, etc.).  Negotiation and negation are two entirely different concepts and the former isn't related to logic; someone looking for a book on negation in logic would NOT (necessarily) find this title in many, many libraries, including LC.

#45002174 and #50661578, The righteous of Switzerland : heroes of the Holocaust / by Meir Wagner ; edited by Andreas C. Fischer, Graham Buik (held by 181 and 69 libraries respectively).
Both are PCC records, one from LC and one from a member library, both coded as level 4.  The LC record has a publication date of 2000 (no copyright date), the other only a copyright date of 2001.  The LCCN and ISBN are identical on the records, as is the rest of the description.  The only difference we can spot is the 260$c, which is a minor difference and does not justify a new record, according to OCLC documentation.

#43286432 and #122980339, Feminism and the family : politics and society in the UK and the USA / Jennifer Somerville ; consultant editor, Jo Campling (held by 250 and 16 libraries, respectively).
The records are identical, both coded level 4 and PCC, one from LC, one from a member library.  In this case, the member record has DLC in 040$a with the member in $c, and the records are identical, except for the extra "the" ("the USA") in the subtitle on the LC record, which doesn't exist on the book in hand.

As a BIBCO library, we CAN make the corrections in OCLC, but we're debating whether we SHOULD make them, because there's an outside possibility (way outside) the "incorrect" versions actually DO exist.  We would like to know what you think - to fix or not to fix?  (apologies to the Bard)  Have we overlooked some documentation indicating PCC libraries should be creating new records if the existing one doesn't match exactly?  (I've looked but maybe not in the right place(s).)  Should we report them as duplicates to OCLC? Or should we just ignore them (very hard to do)?

Thank you!

Rebecca Uhl
Principal Cataloger
Cataloging & Metadata Management
Acquisitions & Metadata Services
Arizona State University
[log in to unmask]<mailto:[log in to unmask]>