ALCTS - Association of Library Collections & Technical Services

Discussion Documents (continued)


Text of Rule 0.24
Modifying the Cardinal Principle in Rule 0.24
Cataloging from the Piece in Hand in an Electronic Age
AACR2 and Seriality
General Material Designations
Multiple Versions Issues
The Definitions of “Edition”
When to Input a New Record
CONSER Single Record Policy

Please note that the purpose of this document is to facilitate the work of the Committee and to provide a means for outreach to both library and non-library cataloging communities. This document is intended for the exclusive use of CC:DA and its cataloging constituencies, and is presented for discussion in the ongoing process of rule revision. Under no circumstances should the information here be copied or re-transmitted without prior consultation with the current Chair of CC:DA.

Cataloging from the piece in hand in an electronic age: a position paper

by Melinda Reagor Flannery and Sherry Kelley
with the assistance of the other members of the
PCC Task Group on AACR Revision

AACR2 and Seriality
[from the CONSER Web page]

Proposal to Adopt a Modified Model C
(Jean Hirons and Regina Reynolds)
This proposal revises the original Model C from the Hirons/Graham JSC paper in order to better accommodate new and existing publications that are “ongoing.” It was developed in the process of revising the term “serial” at the request of the JSC.

Graphical Model of the Bibliographic Universe
(Modified Model C)

Incorporating Entry: A New Concept for Cataloging Electronic Journals (Sara Shatford Layne)

This paper separates ongoing entities into those that are sequentially issued and those that are integrating and suggests entry conventions that might be applied to each. This proposal combines concepts from both successive entry and latest entry cataloging.

General Material Designations

[The following is an exchange of postings from the USMARC discussion list. It deals with cataloging issues not necessarily relevant to this Task Force, as well as issues of USMARC coding. However, it also raises the question of General Material Designations. Consider it a place-holder to remind us that the GMD may be relevant to our discussions.

The conversation was initiated by Crystal Graham. Here is my reply, which quotes here entire message. It is followed by some additional contributions. — JCA]

Some comments below. These are not terribly well thought out and I might not agree with them tomorrow, but what they're worth . . . At 1045 AM 5/2/98 -0700, Crystal Graham wrote:
>>My library is having difficulty identifying the proper Type code to select under 97-3R, particularly for Internet sites. As an example, please take a look at a site called The Visual front, http//

>>This site consists of Spanish civil war posters, which were first exhibited at UCSD and is now travelling around the country. It includes some additional posters in our collection, not actually in the physical exhibit. The text was written specifically for two purposes 1)as "caption text," printed and mounted on the wall in the actual gallery exhibition in order to explain each corresponding poster; 2) as publishable text for the "online exhibition," which in turn was designed to serve both as an electronic facsimile of the gallery exhibition and as an electronic exhibition catalog.

>>I have asked a few experts for their opinions, which I have excerpted below (without attribution). As you can see, we have votes for Computer File, Visual Materials, and Text. I'd like advice on this specific site, but more importantly I'd like to get a dialog going and perhaps a plan of action for coming up with better cataloger guidelines.

>>Computer File

>>As I have interpreted the LC "Draft Interim Guidelines" in conjunction with ISBD(ER), Web sites generally have been defined as "Computer online services," to use the field 256 terminology, and should be coded as Type m ... In your example, I'd say that the graphics and the text are both significant aspects, which leads me to lean toward Type m and File m (combination). You would not use a Computer File 006, but you could (optionally) include a Visual Materials 006 to bring out the graphical aspects (T006 k and TMat i would probably be the only coded values). The 007 field would be for the remote electronic resource 007 c $b r $d c $e n

I'm not sure that I'd use LDR/6=m simply because it's a Web site or simply because the text and graphics seem to be equally significant. Further thoughts on the other choices below.
>>Visual Materials

I think I'd go with Visual Materials and Ldr/06 = k. Among the examples in 97-3R, it would seem to be closest to the Microsoft art gallery from the National Gallery. I don't know if it's relevant but one can imagine it all printed out and it would be a sort of picture album with significant text.

>>I'm not sure if it's relevant that we call such sites "exhibits" when considering whether it's an exhibition catalog (aka text). It's more like the show itself than the catalog. And perhaps because it would be possible to print it out and have a reasonable piece of content, it wouldn't be a computer file Type.

I think the issue here is whether the images illustrate the text or whether the text explains the images. This is seldom a clear choice because obviously both are true in most cases. In this case, there is almost no text that isn't associated with and describing one of the images. The text seems to be there only because the images are there. Put another way, the images would form a coherent work without the text but the text wouldn't form a coherent work without the images. Therefore, I would probably say that the images predominate and use LDR/6=k.

My current best guess; five minutes ago I was leaning towards LDR/6=a. Go figure.


>>I think this site qualifies as digitized text and illustrations and should therefore use a text/books workform - Leader/06 "a", even though it is chiefly reproductions of the posters. (I am assuming that if this were a paper exhibition catalog, it would be treated in that fashion.) My justification is that my understanding of the motivation for the redefinition of code "m" was so that digitized text would be described the same as text - in this case, whether or not an exact textual equivalent exists.

If the framing text (discussions of wider issues not connected to specific images) were more extensive, I could make a case that the text predominates and code LDR/6=a. Again, this is a really close call and there will be room for disagreements in any given case.

>>>I think I agree with the computer file 007. I'm ambivalent about including an 006 for the visual material. Are people generally doing that for an illustrated work - for paper exhibition catalogs? If yes, then there's no reason not to use one here.

Since these elements are optional, I have no strong feelings, but I could see where there might be useful information in either the CF 007 or the VM 006 that one might want to code.
>>At the same time, I think that cataloging this as text does not do justice to the item and is not helpful to the potential user encountering such a record in a catalog.
Nor does coding this as either VM or CF. All of these aspects apply and having to choose one of them for LDR/6 misrepresents the mixed nature of the item. I think our recent decisions about narrowing code "m" apply here; this seems to the the kind of thing that should be coded for content and not format of carrier. But the choice between visual materials and text is arbitrary and has to be supplemented by things like 006 and 007 fields to make sure that the other aspects are clearly identified.
>>Of course, part of that problem is that there would, presumably, be no GMD for this record if treated as text. Is that right? Since CC:DA has not really discussed cataloging Websites (or other changes to "computer files"!!), I'm assuming that the GMD "computer file" would no longer be used for digitized text. Not that I think that particular GMD would be helpful, either. One wouldn't learn that this is a website and in some worlds directly accessible from the catalog record until somewhere down in the notes.
I think we shouldn't be too quick to move from a MARC coding decision to an AACR2 description decision. The choice of a GMD is governed by considerations internal to AACR2 (and the RI's) and may not be exactly parallel to the decision about coding LCR/6. I would argue that this is an issue that is still to be addressed by CC:DA or in the LCRI's. On the other hand, there is a building consensus that type of material labeling in OPAC displays should be generated from MARC codes and that the counterpart in the descriptive rules (the GMD) should be eliminated. At that point, the cataloging rules may need to care more about the MARC coding of this element.

>>Crystal Graham
>>Digital Information and Serials Cataloging Section
>>University of California, San Diego

======= Reply from Crystal ======= John Thanks so much for your thoughtful remarks on the choice of workform question. Although my original questions was limited to MARC coding, I want to follow up on your GMD comments.

I think we shouldn't be too quick to move from a MARC coding decision to an AACR2 description decision. The choice of a GMD is governed by considerations internal to AACR2 (and the RI's) and may not be exactly parallel to the decision about coding LCR/6. I would argue that this is an issue that is still to be addressed by CC:DA or in the LCRI's. On the other hand, there is a building consensus that type of material labeling in OPAC displays should be generated from MARC codes and that the counterpart in the descriptive rules (the GMD) should be eliminated. At that point, the cataloging rules may need to care more about the MARC coding of this element.

I also thought that GMDs were governed by AACR and not USMARC. I would have given the GMD [computer file] even if I put it on the text or vim workform. The cataloger who actually created the record put it on the vim workform and gave the GMD [graphic]. That seemed strange to me since a) I thought carrier took precedence; and b) graphic is not on the LC-approved list of GMDs. (We have a lot of them in our catalog for archival materials cataloged under the rules for graphic materials, so that could be where she got the idea to use it).

Coincidental with this discussion, we are working on changing the limit parameters in our OPAC. The limit criteria is locally-determined and coded, but when set up in the first place it paralleled the Type code (or bib lvl in the case of serials). Now that we've added a slide collection and a film/video library, we decided we should look at the general AV category and break it up. This raised the big question whether our categories should correspond to MARC codes or GMDs. Also, as often happens in these cases, we found our GMDs are a mess -- a mixture of AACR1 and AACR2 terms, some invalid terms like [poster], and terms valid in AACR but not used by LC, e.g., [art reproduction].

We were debating how far to go cleaning up the GMDs when I got your message about the growing consensus toward eliminating GMDs. Can you tell me more about that? Is there a group actively looking at that or just wishful thinking? Is this something that might fall to the 0.24 task force?

Thanks for your opinions/insights on any of this. As a serials cataloger, I'm out of my element with this AV stuff, but I like to learn new skills. And the lack of clear guidance for tagging and cataloging in light of 97-3R affects us all.

Thanks, Crystal

===== My reply to Crystal =====

The statement about eliminating GMD's came from conversations with Barbara Tillett.  At one point, she had announced to CC:DA that her office would be submitting a discussion paper on GMDs.  She later indicated that she saw a growing consensus for eliminating them altogether.  I think that part of this was the feeling that systems could or did in fact label records based on the MARC LDR codes.  I don't know who (if anybody) is following up on this, but I'm sure Barbara would have an opinion about whether continued work on the GMD rules is a fruitful exercise.

===== Another point of view from Robin Wendler =====

I started this response a few days ago -- and took it in a different direction than John did. As he says, go figure... My subject heading formulation is rusty, but you get the idea.

--------------- Crystal, In the spirit of the redefinition of rec type 'm', I believe this one is quite clearly the online equivalent of an exhibition catalog. This is *exactly* what I thought we made those changes to code 'm' for. Let me answer some of your points individually

-- Exhibition catalogs frequently include images for things which are not actually in the exhibition.

-- Even those exhibition catalogs which consist *overwhelmingly* of images are cataloged as rec type 'a', not 'k'.

-- Instead of "caption text", think catalog entry. (I have friends who write exhibition catalog entries for a living, so this is familiar territory.)

I would add a 655 for Computer network resources, a 650 for Posters..., a 651 for Spain--History--Civil War, 1936-1939--Posters--Exhibitions. I would not add a visual resources 006; I don't think it adds much, if anything. I don't think I'd add an 006 for computer files, either, but hey, it's optional.


Robin Wendler ........................ work (617) 495-3724
Office for Information Systems ....... fax (617) 495-0491
Harvard University Library ...........
Cambridge, MA, USA 02138 .............

Multiple Versions Issues: A Blast from the Past!

[Quite by accident, I ran across the following document yesterday. I was asked to purge some messages from an old e-mail system we no longer use. Among these was a folder of messages relating to the old Multiple Versions Task Force (anybody remember that — or were you trying to forget?). The last message in the folder was my last words (?!?) on the subject of multiple versions in AACR2, dating from 1993! The reference to rule 0.24 caught my eye and I decided that this might be worth contributing as background to our deliberations. Although time and much else has passed since this was written and I would undoubtedly say something rather different today, I think that these comments at the least lay out some issues that the Task Force needs to consider.

The context surrounding this issue includes an abortive proposal to revise 0.24 and 1.11; a Task Force on Editions; and a decision that my solution would not work until the definition of “edition” could be changed. Is now the time to re-open this question? — JCA]

Changes to AACR2

The cataloging of multiple versions in general and of reproductions in particular raises problems with three aspects of AACR2

  • Rule 0.24, the cardinal principle of AACR2, which calls for the description of the particular physical item in hand
  • Rule 1.11, which sets forth the general rules for cataloging reproductions
  • The definition of “edition” and the rules for describing editions

I will try to deal with each of these areas.

  1. Rule 0.24

    It is my perception that much of the Anglo-American world is willing (or even eager) to make some kind of exception to this rule in the case of reproductions. Certainly, in the United States, we have never followed this rule as written.

    The argument seems to be that reproductions are so closely related bibliographically to their originals, not only in terms of the intellectual content (which I will call the "text"), but even in terms of the physical nature of the original which is preserved to the extent allowed by the reproduction technology. By this I mean that, for example, a microfilm of a printed book is still organized in the same page images that are present in the original. It is also true that the intent of the reproduction is to serve as a substitute for the original. The work of this Task Force has greatly clarified these issues.

    Historically, there have been calls to abandon the cardinal principle entirely. This is usually expressed as a suggestion to catalog “works” or “texts” in the first instance, and then to describe all the various manifestations of these works in subordinate descriptions.

    While there is a superficial elegance about this concept, it has several problems. First, it would probably mean a total restructuring of our bibliographic databases and a reconceptualization of our user interfaces. At best, it will mean a significant amount of intellectual labor to identify “works” and to link all the manifestations to these works. At worst, it will mean starting over to create new bibliographic databases configured in the new manner.

    The other problem is that “works” are very slippery animals. One recent thought on the subject made the mistake of using as an example Shakespeare’s Hamlet. Obviously, there is such a work as Shakespeare’s Hamlet. Well, not really; there are two quite distinct textual traditions and most modern texts contain elements of both traditions. There is a vast body of textual scholarship that tries to deal precisely with this kind of problem.

    It is my feeling that it would be a mistake to abandon 0.24. It seems to me that the various service missions of libraries deal more with “books” than with “texts”; we are the custodians of books; we order and receive books; we lend, borrow, and inventory books. I think that it would be a mistake for catalogers to get into the business of textual scholars. We don’t really know what we would be getting ourselves into! In any case, I don’t think that we need to abandon or even significantly change 0.24 in order to describe reproductions in the manner proposed by the Task Force and approved by CC:DA. The rule says to describe the physical item in hand according to the rules in the chapter appropriate for that type of material. So, in describing a microform reproduction of a book, the description “should be based in the first instance” on chapter 11. According to our Guidelines, the description of the reproduction does indeed follow the rules in chapter 9. The only exceptional treatment required is to add the alternative of a multi-level description to rule 1.11. In rule 0.24, I would like to see a reference to rules 1.9, 1.10 and 1.11, which are the general rules for cases which are most likely to require use of more than one chapter.

  2. Rule 1.11.

    By adding the alternative of a multi-level description to this rule, it would be possible to describe reproductions according to our Guidelines in a manner that is compatible with rule 0.24.

    Multi-level description is a single description made up of two parts. (The fact that we hope to maintain and communicate these in separate MARC records is irrelevant.) In Chapter 13, the technique is used to turn normal analytic descriptions inside out, while still maintaining compatibility with rule 0.24. Rather than beginning the description with the item in hand (the analytic) the multi-level description begins with the larger bibliographic entity (“host”) and describes the analytic at the second level.

    The same alternative technique could be added to 1.11. Rather than beginning with the description of the reproduction, the multi-level description would begin with the description of the original; the reproduction would be described at the second level. The two levels together make up the description of the reproduction, and therefore rule 0.24 is not violated (at least, no more than the multi-level description technique in Chapter 13 violates rule 0.24).

  3. Edition

    AACR2 exhibits symptoms of schizophrenia with regard to editions. On the one hand, the definitions in the glossary stem from a long-standing tradition that concentrates on problems of the transmission of texts. The definition of “edition” seeks to identify groups of copies that are most likely to transmit a consistent text. They do this by going back to the technical processes that created the master from which copies are produced. In some cases, these masters are essentially stable (a photographic negative, the production master of a sound recording); in other cases, the textual stability is only approximate (in-press corrections to hand-printed type; revisions to engraved plates). However, the idea is to identify points at which the master has been remade and at which significant (even total) revision of the “text” becomes possible. The other approach to editions in AACR2 is the rules for Area 2 of the description. Here the emphasis is on transcription and the working principle seems to be that an edition is something that has a distinctive edition statement. Even if the “text” is produced from the same master and the ONLY thing new is the edition statement, it will be a new edition according to AACR2.

    There is a clear conflict between these two approaches. There are also some significant problems in choosing between them. The emphasis on the transmission of the text and on the identification of differing masters once again gets catalogers into the realm of the textual scholar, an environment where I suspect most catalogers would not be at home. It also requires that catalogers answer difficult questions with totally inadequate evidence; to identify differing masters without access to those masters, from the evidence of the copies alone, is virtually impossible. On the other hand, the tradition embodied in the definition of “edition” is a valuable one; our users, if they care about editions at all, want to know about the “text” they contain, not about the edition statements that they might or might not bear.

    The relevance of all of this to multiple versions is that the current definition of “edition” (even without the reference to issuing body, which itself compromises the usefulness of this definition in distinguishing truly different “texts”) is not compatible with treating reproductions as copies of (the same edition of) the original. We need to find a way out of this dilemma, other than admitting that it is an exception.

    One way to do this is to place complete emphasis on the “text”; if the reproduction is made from a different master, yet that the intellectual content of that master is the same as that of the original. If all other requirements were removed from the definitions, then the concept of copies of the same text would permit the application of our Guidelines. However, this choice gets us into the realm of textual criticism.

    Another way out is to rewrite the definition of “edition” in terms of explicit or supplied edition statements. (This would allow a cataloger who knew that the “text” was produced from a different master to supply an edition statement; but it would not require that the cataloger excavate the textual history of the work.) In the case of reproductions, the edition statement would be just one more feature of the original carried over to the reproduction; by definition, they would be identical (unless, of course, the “reproduction” did embody additions or changes, which according to our definition would not be a reproduction).

    This suggestion proposes that we abandon a bibliographic tradition that has been a part of cataloging from the very beginning. I do not suggest this lightly. In fact, its most attractive feature to me is that it solves the immediate problem; I’m not sure that I am as pleased with its broader implications. However, it seems the preferable alternative to solving what I have come to believe is a major pathology in the rules the schizophrenic and self-contradictory approaches taken to the concept of “edition”.

The Definitions of “Edition”

In 1992, CC:DA appointed a Task Force on the Definitions of “Edition” to consider the implications of the definitions of “edition” in the Glossary of AACR2 for the recommendations of the Task Force on Multiple Versions. The report of this Task Force noted that the proposed technique for describing reproductions was in fundamental conflict with rule 0.24 and suggested that a thorough review of rule 0.24 might be necessary.

When to Input a New Record

In a shared cataloging environment, one of the major impacts of the definition of the term “edition” is in determining when an item in hand matches (is a copy of the same edition of) an item described in an existing bibliographic record. AACR2 does not address this issue explicitly. The most extensive treatment is the document “When to Input a New Record” – Chapter 4 of OCLC’s Bibliographic Formats and Standards.

CONSER Single Record Policy

In 1996, CONSER developed an experimental policy for inclusion of online versions on records for the print. During the past few months a group within CONSER was asked to draft a policy that would suggest guidelines for determining when it is appropriate to use the single record approach. The paper was discussed at the April 1999 meeting of the CONSER Operations Committee but no conclusions were reached.

The topic will be discussed at three venues in the upcoming month: at the Cataloging Network Node at NASIG, at the Committee to Study Serials Cataloging meeting at ALA, and at the CONSER At Large meeting, also at ALA.

I invite you to read the paper as a background to these discussions. Valerie Bross of UCLA convened the group; Crystal Graham (UCSD) and Steve Shadle (U. Washington) were the principle writers of the report. It is available at

Jean Hirons
CONSER Coordinator
Library of Congress