Below is the response from Margaretta Yarborough on behalf of University of North Carolina, Chapel Hill. Margaretta would like to share the message on the BIBCO listserv but her message could not go through for some reason. Thanks. -- Jessalyn
Cynthia: If we have to choose one, UNC would go with option C. We divided pretty evenly along the lines of general and special collections cataloging between options B and C, respectively. We all recognize that this is not much of an issue with e-resources.
Option B, we think, would be greatly improved by retaining all the location data in the final merged record. As one cataloger put it, "Worst case, I’d rather live with multiple, repetitive 26X fields than lose what may be an important clue to publisher info." And generalist catalogers register the concern that not all rare materials are coded dcrm(b) in the 040 field.
Catalogers with a greater stake in rare materials cataloging strongly prefer C. Several comments:
--I think different places of publication should qualify as a major difference, even if we haven’t cataloged the item as “rare.”
--In rare books we only code the 040 $e dcrm (b) when it is a pre-1801 piece. We do put in every possible detail about that piece that we can in the publication area and in the notes to distinguish our title from other variants. I am not sure that I am okay with choice B due to the fact that many of our titles are variants, are after the cut off date for rare materials, and are not coded as dcrm (b) in the 040. A great example of this is the Wordsworth Collection, which has so many rare variants, is late 19th century and is difficult to protect in OCLC.
--I’m having trouble thinking of examples where the publisher would be the same but the place of publication would be different, though I’m sure there are some. But definitely in special collections we have plenty of materials that are very similar but have meaningful differences that we need separate records for, and in most cases they don’t fall into OCLC’s guidelines for what constitutes a rare book. I’m wary of any attempt by OCLC to do more merging of records, since I’ve already had plenty of problems with records being merged that shouldn’t have been.
One suggestion for a way to proceed:
“Perhaps OCLC can look to the notes fields with copy specific information? More and more rare book catalogers, including LC, add in their 500 copy specific with $5 NcU to let researchers know what stands out about their specific copies. Perhaps this could be used to prevent merging?”
And I’d also like to second Robert Bratton’s comments, especially greater emphasis on reporting duplicates. He is a master of understatement: “The printing and publishing industries evolved very differently in different parts of the world.”
Margaretta J. Yarborough
Head, Resource Description & Management
CB# 3914, Davis Library
University of North Carolina at Chapel Hill
Chapel Hill, NC 27515-8890
Thanks Robert, and everyone else for your input. Input is welcome throughout the rest of today, but none after today please, so I can compile all the responses for consideration.
We definitely have plans to update the chapter in BFAS about identifying and reporting duplicates. We are concentrating on the updating all the fields themselves right now, but hope to update the five introductory chapters in 2014.
While I feel the pain of dealing with duplicates with the same publisher info. recorded 10 different ways, I have to agree with Naun and Bruce that I do *not* think OCLC should make their bib. record merging parameters any more liberal.
There is no good cut off date for rare books. We see some publishers doing bizarre things in their print runs from the 19th and even into the 20th century. The printing and publishing industries evolved very differently in different parts of the world.
Perhaps OCLC should place more emphasis on users reporting duplicates. Also, since OCLC is completely updating Bibliographic Formats and Standards, the section on identifying and reporting duplicate records should be updated as well.
Thank you for seeking our input,
George Washington University Law Library
Washington, DC 20052
On Tue, Nov 12, 2013 at 5:26 PM, Whitacre,Cynthia <[log in to unmask]> wrote:
Thanks Bruce, for sending this input on behalf of your colleagues. OCLC would welcome additional input on this topic from anyone on the
BIBCO list through this Friday, November 15.
I’m forwarding this message on behalf of some Harvard Colleagues.
Cynthia – Thank you for sending out to various listservs the announcement about proposed changes in OCLC’s policy relating to place of publication and matching records. We catalogers at Houghton Library (the main special collections library at Harvard University) feel strongly that policy should not be changed. The place of printing is far too important to records treated by special collections catalogers to chance losing information or creating misinformation through computer matching. When it comes to special collections records, there are just too many variations in records to create uniform algorithms. As we are sure you have heard from other special collections catalogers, great numbers of records for “rare” material are not coded in the 040 as being cataloged using any particular set of rules. To use a record’s dates as the determining factor is impossible for this reason as well as the fact that many newer materials are now cataloged using rare book cataloging standards. Additionally, what is not considered rare today, may well be considered so in the future. We cannot chance the fact that a book with the imprint "Cheltenham, UK: Edward Elgar, " is the same as one with "Cheltenham: Edward Elgar, 2013" and "Northampton, Mass.: E. Elgar Pub., 2013" when the possibility is just as likely that they are different. In your email, you suggest that "end users really don’t care about these distinctions when it comes to obtaining the content and find the multiple records confusing" – that is simply not true of the many scholars doing historical research on a wide array of topics, many of which extend into the 21st century.
Thank you for giving us the opportunity to voice our opinion.
Head, Rare Book Team
Cambridge, MA 02138
Hello BIBCO Colleagues:
This is an admittedly long message, but please read all of it, as your opinion and thoughts are requested.
OCLC is contemplating making a change in our “when to input a new record” standards as published in Bibliographic Formats and Standards Chapter 4, as well as in our internal merging standards for what is considered a duplicate. We’d like your help in reaching a decision. The element in question is the place of publication.
Currently in Chapter 4 of BFAS (http://www.oclc.org/bibformats/en/input.html) under 260 subfield a (we know we need to add 264 to this) we list some instances where minor differences in place are acceptable for considering the records to be the same. However, the basic statement says that “differences in the place of publication justify a new record.”
One of the reasons we have traditionally NOT merged “duplicate” records when everything else is the same is when place of publication differs. Here is what our current internal merge instructions say:
Subfield $a guidelines
· Records may be considered duplicates for merge even with the absence or presence of the subfield a.
· Always match 1st place of publication.
· New York matches New York, Toronto
· London, Orlando matches London, Toronto
· New York does not match Toronto, New York
· New York, Bombay does not match Bombay, New York
Note: Places of publication within the same country are considered a match and justify a merge.
· New York, Chicago matches Chicago, New York
For example, look at these 3 publication statements from 3 WorldCat records for the same title:
264 1 Cheltenham, UK : ǂb Edward Elgar, ǂc 
260 Cheltenham : ǂb Edward Elgar, ǂc 2013.
260 Northampton, Mass. : ǂb E. Elgar Pub., ǂc 2013.
Under our current OCLC criteria for place, we would merge the first two but not the third (assuming everything else matched).
Looking at the cataloging codes, here’s what AACR2 says:
1.4C5. If two or more places in which a publisher, distributor, etc., has offices are named in the item, give the first named place. Give any subsequently named place that is given prominence by the layout or typography of the source of information. If the first named place and any place given prominence are not in the home country of the cataloguing agency, give also the first of any subsequently named places that is in the home country. Omit all other places.
[log in to unmask]" alt="cid:[log in to unmask]">
And, RDA says:
More Than One Place of Publication
If more than one place of publication is named on the source of information, record the place names in the order indicated by the sequence, layout, or typography of the names on the source of information.
Our thinking, in random order:
· It has always seemed that cataloger’s judgment is at play regarding what gets recorded, no matter which code is in use.
· RDA gives a bit more leeway than AACR2. In our discussions, we are of two minds regarding this.
· In the past, when the same content was published in two countries (UK publication and US publication for example), it often came out at different times and may have had bibliographic significance. We are not so certain that is the case today.
· Many reported duplicates and perceived duplicates exist in WorldCat because of the policy of not matching places of publication in different countries when everything else matches.
· End users really don’t care about these distinctions when it comes to obtaining the content and find the multiple records confusing
· Records for remote electronic resources, are often created by content providers through automated means, and careful checking of the actual publication is often not done; thus the accuracy of place of publication may be meaningless in many records for electronic resources.
· Distinctions like this will continue to matter for rare books.
So, we are left with some choices:
A) Continue with the current policy, and require different WorldCat records for different places of publication
B) Modify the current policy to allow the merging of records (and matching of records) with different places of publication for non-rare materials if everything else matches.
C) Stick with the current policy for tangible resources, but allow matching/merging of records for different places of publication for records for remotely-accessed resources only.
D) Another alternative entirely; suggestions welcome!
We would welcome discussion and input on this matter. Which of the choices, A, B, or C do you prefer? If you prefer D, please suggest the alternative that you have in mind. We have asked BIBCO colleagues, since we believe this is a larger issue with records for monographic materials rather than with records for continuing resources. Please share your thoughts on the BIBCO list. If you prefer to send a message directly to OCLC, send it to [log in to unmask], as that will go to the appropriate people. We welcome comments through November 15, as we will be discussing this again at OCLC the week of November 18 to try to reach a decision.
Cynthia M. Whitacre
Manager, WorldCat Quality & Partner Content Dept.
800-848-5878, ext. 6183