It is good to know that the time-consuming BFM work has been automated at UW. Are you using ALMA and this is part of the ALMA system? PUL is still using Voyager. Changes are made locally first and updates are exported to OCLC weekly—not an ideal situation obviously, since there’s always a time lag. Cooperative cataloging should be done at the master-record level and the revised master record should replace the local record (for the most part, I assume, local notes remain): truly, one for all, all for one. I hope the practice will be adopted widely.
Back to a couple of suggestions Adam made earlier. Obviously “Liver (Arias)” is good enough—if one does not read too much into it, taking “Arias” as a genre germ. “Liver (1983)” would be acceptable also, I suppose, but obscure and dicey. In order to identify this particular work “Liver” edited by Arias as a monograph or reference work, one could use its form as part of the qualifier so that library users may know what it. Form of a work is a core element—it’s an important work-level attribute.
Here are some examples. Suppose “Liver (2016)” in a catalog, without clicking on it and opening the record, one would not be able to know that it is a motion picture. If the AAP were “Liver (Motion picture : 2016)” (cf. IMDb), the meaning is immediately clear. In OCLC bib file, one may also find “Liver” as the title of a chart (1984), a video-recording (1978), a set of slices (1971), a computer file (1996?), etc. One could certainly use dates as qualifier, but they are not as useful as “form of work.”
As for using $f to distinguish different editions of a work, I tend to think “edition statement” would be more preferable, or, better still, one could use both:
Liver (Arias). [$l English] $s (6th edition)
Liver (Arias). [$l English] $s (6th edition. $f 2020)
Should there be a revised 6th edition in 2021, then
Liver (Arias). [$l English] $s (6th edition. $f 2021)
I put [$l English] in there, just to show that under RDA, Liver (Arias) itself has no language attribute. It may not necessary to add $l English to the expression-level AAP to indicate that it’s an English text.
But, theoretically speaking, it’s by far more logical than one would think.
Like us, more and more libraries are relying on master records in OCLC that automatically get loaded into our local system any time that they are changed, so we would hope that others do the extra step of fixing OCLC. We cannot maintain our records in our local system, we have to update the master records in OCLC because of our infrastructure. If we update just the local copy, it would get overlaid later on by any change made by anyone to the master copy in OCLC.
University of Washington Libraries
From: Program for Cooperative Cataloging <[log in to unmask]> on behalf of Benjamin A Abrahamse <[log in to unmask]>
Sent: Wednesday, March 4, 2020 6:56 AM
To: [log in to unmask] <[log in to unmask]>
Subject: Re: work-level vs expression-level access points
Thank you Steve for explaining your reasoning more fully.
I think if I had the older editions in my catalog I would add the UT to their records, but I don’t know if I’d do the extra step of fixing them in OCLC.
Yes, if we were cataloging this work anew today, we would need to qualify it. But if we used a qualified uniform title on this edition, would we not have to add it to the records for the previous editions? My thought was that the Work was catalogued 38 years ago, under different practices. Do we need to change it now, causing much work on older editions in hundreds of libraries? Hm. If we were cataloging a new Work with a title that conflicted, we clearly would have to qualify that title. But when we just have a new edition of an existing Work, do we not use the access point used on previous titles?
Doesn’t it need the UT to distinguish is from other works (e.g.: lccn: 93023602) that have the title “Liver” and no author Main Entry (per RDA 184.108.40.206)? Or am I misunderstanding something?
I agree that this title _could_ have a uniform title. By historical practice, it does not _need_ one. Liver is the only title that this work has been published under. If a first edition with no creator were catalogued today, it probably would get a uniform title. But the work has gone without a uniform title for 38 years.
If this record is given a uniform title, then all previous editions should also get the same uniform title. This would require a BFM to the Library of Congress for five previous editions. It is an awful lot of work to fix something that is perfectly acceptable by historical practice. As I suggested above, I expect that the Library of Congress will remove the 130 when the edition is fully catalogued.
Since the title proper is just Liver, it probably does need a 130.
1) Personally, I find Liver (Reference work : Arias) even more confusing than just Liver (Arias).
2) The 130 is the usually the uniform title of the Work rather than the Expression. There is no reason to include an AAP for the Expression in this record.
3) This record probably does not need to have a 130 at all. The first, second, third, fourth, and fifth editions, as catalogued by the Library of Congress, do not have uniform titles. This record is CIP. I expect that when fully catalogued, there will be no 130 field in this record.
2 quick questions about a work-level AAP used in this bib record (lccn 2019024961).
130 0 Liver (Arias)
1) The qualifier “Arias” sounds a bit funny and “untoward”, so to speak, since it happens to be a variant form of the lcgft term “Songs.” We all know, for machine indexing and data manipulation, it does not matter at all. But for a human to read and comprehend what it is, would it be better to see a more understandable heading like “Liver (Reference work : Arias)”?
2) This is the 6th edition of the work. At this point, would an expression level access point be more appropriate?
130 0 Live (Arias) $s (6th edition)
Per OCLC bib file, the first edition was published in 1983; Irwin M. Arias has been its chief editor since the 1st edition. I am not questioning the use of work-level AAP in this bib, seeing that it stands for “[a] distinct intellectual … creation.” I am just inquiring if there’s a best practice when we come to deal with reference books/textbooks that have gone through multiple editions.
If a work-level AAP is all we need, so be it. But if an expression-level AAP is more uniquely helpful to library users, then, why not?