I think I need to explain more clearly what a digital registry record is.
The digital registry record actually IS the WorldCat record. An 042 of dlr
is what "puts" the WorldCat record into the Registry of Digital Masters.
There is just one record. It can be found in WorldCat using a regular
search or it can be found in the digital registry by limiting a WorldCat
search using authentication code of dlr. More details are available at:
It is similar to the CONSER database. The CONSER record IS the WorldCat
record and it's the authentication code which makes it a CONSER record. The
CONSER database is a subset of WorldCat. Same for the digital registry.
So, Robert's comments about the 533 in print version digital registry
records causing problems were, in fact, talking about the WorldCat records.
As I mentioned in the last message, however, those records were created
before the guidelines that we are now discussing were developed. The
library which added those 533s uses a single record approach, so they added
the 533 to the print version record. In order to avoid the problems that
Robert mentioned, our guidelines recommend using different wording in the
533 when using the single record approach.
We are recommending the following: If you are using the single record
approach, adding information about the electronic version onto the print
version record, you should use "Also available as electronic reproduction"
in subfield a of the 533 in the print version record. If you are using the
separate record approach, creating a separate record for the electronic
version, you should use "Electronic reproduction" in subfield a of the 533
in the electronic version record, which is the way we have been doing it
all along for reproductions.
I hope this makes sense. I know it's not ideal to use the same field in two
different ways like this, but so far we haven't been able to come up with
anything better. My first response to Celine explained some of the other
things that we tried and why we ultimately decided on 533 for both the
single and separate record approach.
By the way, there is an interesting discussion about single vs. separate
records occurring on the informal digital registry list, which is basically
just an email list that I have been putting together to coordinate the work
on the guidelines. Originally it consisted of the CIC Digital Resource
Cataloging Task Group and the CONSER group that was looking at
compatibility issues between the aggregator neutral record and the digital
It has grown to include people from the PARS Intellectual Access to
Preservation Data Interest Group, the DLF Registry of Digital Masters
Working Group, the Library of Congress, OCLC, and other people who have
expressed an interest in the topic. If anybody from this PCC list wants to
be on that list, please let me know. I would be happy to forward messages
back and forth between lists if everybody on this list wants to know what's
happening on that list, but after awhile that would get to be unwieldy.
At 09:05 AM 1/11/2007, you wrote:
>Thanks, Renette for your further comments. I now see why I misunderstood
>the comments from Robert Bremer that you mentioned: I thought his
>comments referred to reproduction records being added to Worldcat when
>what he meant was reproduction records were being added to the Digital
>Registry on the same assumptions/guidelines as those in Worldcat, hence
>the duplicate records!
>Also, thanks for pointing to the RDA FAQ. I'm having a hard time keeping
>up with RDA developments but I had heard that there was opposition to
>single records for FRBR reasons. Locally, we too are evaluating single
>records again. Our current campus policy is to use single records
>whenever possible, but we've had to create more and more exceptions
>recently and are using fewer single records than we used to.
>Still, the savings in time can be very persuasive and public services
>staff like single records, so I think you are right in trying to
>accommodate single records in your guidelines.
>Renette Davis wrote:
>>Thanks again, Celine, for your comments. I'm going to continue the
>>discussion by interspersing some comments below, preceded by RD. I'm
>>really glad to have some input from the PCC monographic cataloging
>>perspective. We've had some PCC participation from the serials side with
>>people from CONSER on the group, and I imagine some of the CIC cataloging
>>people are PCC members, but I think it's good to have the issues
>>discussed on this list as well.
>>I don't think this is going to work unless we have broad participation in
>>the discussion and come up with something that everybody can live with.
>>If anybody thinks we should take the discussion off this list, let me
>>know, and I'll add Celine (and anybody else who is interested) to the
>>smaller group that has been discussing these guidelines. The goal of that
>>group has been to come up with some guidelines that other people can then
>>react to. Hopefully we will eventually come up with something that
>>everyone can live with.
>>At 11:09 AM 1/10/2007, you wrote:
>>>Thanks, Renette. I can see that you all have given this a lot of
>>>thought and that your needs are somewhat different from other
>>>Because of this, I'm not sure that we are understanding each other so
>>>I've added a few more comments below and then will wait to hear the
>>>results of the meeting. Think of this as fodder for discussion.
>>>Renette Davis wrote:
>>>>Thanks, Celine, for your comments. I'm going to try to give some
>>>>background on how and why it was decided to use the 533 on print
>>>>version records when using the single record approach, for anyone who
>>>>might be interested.
>>>>First of all, there are already a number of print version records in
>>>>the Registry of Digital Masters with 533, and I think you're right, it
>>>>is very difficult to interpret what version these records are. For
>>>>examples see OCLC #642567, 17604932, and 19823978.
>>>These records are taking the single record approach to new lengths!
>>>As one of my colleagues put it: "it looks like a poor man's FRBR record!"
>>>It's hard to know what a public user would make of these, but maybe as
>>>long as there's a URL, it won't matter.
>>RD - Yes, "Poor man's FRBR record" is probably a good description! :-) As
>>you might guess, we had a lot of problems coming up with guidelines for
>>the single record approach, and there was a time when I was ready to send
>>a message to the group asking if we REALLY want to be promoting the
>>single record approach. Before doing that, though, I ran the idea by
>>colleagues here at the University of Chicago (we use the single record
>>approach) and there was strong objection to moving to a separate record
>>approach here, so I didn't send out the message. The fact is that there
>>are a number of institutions, including the Library of Congress, who use
>>a single record approach by policy, so what we have tried to do is come
>>up with guidelines for those who use a single record approach as well as
>>guidelines for those who use a separate record approach. By the way, the
>>records that I pointed out above were created before we started working
>>on these guidelines, and the 533s are not worded in the manner that we
>>are recommending they be worded in the guidelines. We are recommending
>>using different wording in subfield a in the 533 for the single record
>>approach, which will hopefully make it more clear that this is a print
>>version record to which information about the electronic version has been
>>added rather than a reproduction record itself.
>>>>Early on in our discussions, Robert Bremer from OCLC said that
>>>>catalogers are seeing a 533 reproduction note in print version records
>>>>and assuming that the record really represents the electronic
>>>>reproduction itself. As a result, they are inputting duplicate print
>>>We seem to be talking at cross-purposes here. Outside of the
>>>preservation world, records with 533 fields have always represented only
>>>the reproduction. Someone cataloging the original is not supposed to
>>>use a record with a 533 field, but input a new record for the print.
>>RD - This has been true in the past, which is why Robert said that the
>>records which had been added to the digital registry that were print
>>version records to which a 533 had been added were causing problems. I
>>had a hard time with those records myself when I first started looking at
>>digital registry records. That's the reason we are suggesting that if you
>>are using a single record approach, adding digital registry information
>>to the print version record, you should use something like "Also
>>available as electronic reproduction" instead of "Electronic
>>reproduction" in subfield a of the 533. We are hoping that "Also
>>available as electronic reproduction" would be a clue that this is a
>>print version record, and "Electronic reproduction" could continue to be
>>a clue that it is an electronic version record.
>>>Of course this way of defining and describing a reproduction predates
>>>the current "single record for multiple formats" environment so it used
>>>to be understood that one record represented one format, and in records
>>>with a 533, that format was the reproduction.
>>>Now that single records can represent more than one format and also
>>>happen to look a lot like the old reproduction records, the waters are
>>>very muddied, indeed.
>>RD - I agree. I don't know if we can ever make the water clear, but
>>hopefully we can un-muddy it somewhat.
>>>I'm not against redefining the 533 field in the single record
>>>environment, but just pointing out that while a revised 533 note looks
>>>like a minor and harmless change, it actually represents a MAJOR change
>>>for many of us.
>>>What you're really saying is: from now on, any record with a 533
>>>represents both the print and the reproduction.
>>RD - Not really. A record with a 533 could represent both the print and
>>reproduction or it could represent only the reproduction. If the subfield
>>a says something like "Also available as electronic reproduction" the
>>record represents both the print and the reproduction. If it says
>>"Electronic reproduction" then it represents only the reproduction.
>>(Actually, because of LCRI 1.11A, the record describes the print with
>>reproduction information in the 533, but the record itself is an
>>electronic version record.) Maybe I should restate this to make it more
>>clear. If the subfield a says something like "Also available as
>>electronic reproduction", it is a print version record, using the single
>>record approach to add information about the electronic version to the
>>print version record. If the subfield a says "Electronic reproduction" it
>>is an electronic version record, using the separate record approach and
>>representing the reproduction. (If it is a serial, it might also
>>represent other electronic versions because we will be adding information
>>about these reproductions to the aggregator neutral record when using a
>>separate record approach.)
>>>In the preservation environment, this makes sense. In the wider world
>>>of cataloging, libraries cataloging a reproduction are often doing it
>>>because they don't have the original. What would a record for a
>>>reproduction-only look like?
>>RD - A record for only the reproduction would look the same as it does
>>now, i.e., a library that doesn't have the original would only have an
>>electronic version record for the reproduction. It would either describe
>>the original with reproduction information in 533 if the library is
>>following LCRI 1.11A, or it would describe the reproduction with
>>information about the original in 534 if the library is following AACR2.
>>If this is a resource that was digitized by one of our libraries, and if
>>the library which digitized the item used a separate record approach,
>>there would be an electronic version record that the library which
>>doesn't have the original could use. If the library which digitized the
>>item used a single record approach, then the library which doesn't have
>>the original would have to create an electronic version record. I
>>personally think this is an argument for always using a separate record
>>approach for things that we digitize locally, but I don't think I have
>>the energy to pursue that argument right now.
>>> Ironically, we might be forced to go back to following AACR2, basing
>>> the description on the reproduction alone and adding a note like a 534
>>> to describe the original! Do we know yet what RDA is going to say
>>> about reproductions and single records?
>>RD - The RDA FAQ says:
>>4.12 How will multiple manifestations of the same title be handled in RDA
>>(the "multiple versions" issue)? Will RDA recommend that we use one
>>bibliographic record for all manifestations of the same title, will it
>>recommend that we use a separate bibliographic record for each
>>manifestation of a title, or both?
>>These questions address the current practices of how we catalogue
>>multiple manifestations of the same title. There are two primary
>>techniques in use for doing this. The first is to create a separate
>>bibliographic record for each manifestation of a work (as in AACR2). The
>>second involves using the same bibliographic record (most often that of
>>the print manifestation) for all manifestations of a title.
>>RDA is heavily based on the FRBR (Functional Requirements for
>>Bibliographic Records) model. There are four entity levels in FRBR's
>>Group 1 - work, expression, manifestation, and item - that point to the
>>creation of a separate record for each manifestation of a work or
>>expression. Therefore, RDA will follow the principle of creating a record
>>for each manifestation of a title.
>>The technique of using a single record for all manifestations of a work
>>or expression was developed before the FRBR model was created. Because of
>>this, a cataloguer could not easily show the relationships between
>>multiple manifestations of the same title using separate bibliographic
>>records. In using the FRBR model, RDA will provide guidance on
>>collocation at the expression and work levels through citations (i.e.,
>>uniform titles). In large part, this is precisely what the single record
>>approach has attempted to accomplish. The solution in taking advantage of
>>the FRBR model, however, is in database structures rather than in the
>>Although the practice of describing multiple manifestations on a single
>>record is widely used, the JSC sees it as one that prevents agencies from
>>sharing data effectively and from taking advantage of FRBR. While RDA
>>will recommend the use of one record for each manifestation, the JSC has
>>agreed that the technique of using a single record for multiple
>>manifestations of the same title will be mentioned in RDA's General
>>Introduction. The new code will not provide any specific guidance for its
>>use in the body of the rules, however.