[DCRB-L] WG6: Editions draft

[Deborah J. Leslie]

Herewith a link to the preliminary document by John Attig for Working Group 6: Editions, issues, and states, or, When to create a new record. The link is to its original MSWord format, as well as the text without formatting appended here.

http://www.folger.edu/bsc/dcrb/wg6.doc


Editions, Issues, and States, or, When to Create a New Record
For discussion in preparation for the
Descriptive Cataloging of Rare Materials Conference
to be held March 10-13, 2003 at Yale University

PRELIMINARY DRAFT
January 21, 2003 - John Attig


DCRM Working Group 6:  Editions, Issues, and States, or, When to Create a New Record

Many conditions influence this decision, both at the local and the utility level, which make it particularly thorny. This group will explore the feasibility of writing an equivalent to LCRI 1.0. In doing so, it will consider traditional bibliographical conditions of early printed books in context of the IFLA Functional Requirement for Bibliographic Records, in particular, in the definitions of work, expression, manifestation, and item. In addition, it may craft guidelines and considerations for individual agencies on choosing what level of cataloging is appropriate for different kinds of materials.

In this preliminary draft, I would like to briefly outline the following:
1.	The state of the Anglo-American Cataloguing Rules with regard to (a) the concepts and terminology introduced in Functional Requirements for Bibliographic Records, and (b) the question of when to make a new record.
2.	Some of the issues that make these questions particularly interesting for catalogers of rare materials.
3.	A proposed distinction between a default guideline for when to make a new record and a set of principles for diverging from that default.
4.	A further distinction between decisions to be made when creating records for a local catalog and decisions to be followed when entering a new record into a shared-cataloging, master-record union database.  This last discussion will revisit the perennial question of how a system for sharing bibliographic information about rare materials should be structured.

I'm going to spend most of my effort in this preliminary draft on the first point.  Because of my intensive involvement in the rule revision process, I feel that my major contribution to this discussion may lie in describing the dynamic state of discussions in the AACR revision process. These include initiatives being undertaken by ALA's Committee on Cataloging: Description and Access (CC:DA, for short) and by the Joint Steering Committee for Revision of AACR (JSC).

 
AACR2 and FRBR: Conceptual and Pragmatic Issues
The Joint Steering Committee has been working to introduce FRBR concepts into AACR2.  The first step has been to introduce the terms "expression" and "manifestation" in appropriate rules, and to restrict the use of the term "item" to its FRBR meaning as a synonym for "copy."  The term "work" is already present in AACR2 and is almost always used in its FRBR meaning.

One of the major conflicts between AACR2 and FRBR lies in the concept of "edition," which is a very important one in both the ISBD structure, in the AACR2 rules, and in the conventions of publication and research which these standards support.  "Edition" simply does not fit easily into the FRBR schema of work à expression à manifestation à item.  In FRBR terms, "edition" is an ambiguous concept: in most cases, an edition is a distinct expression of a work - but only if there is a change in the form of expression or in the intellectual or artistic content.  An unaltered reprint may be a distinct edition, and is definitely a distinct manifestation, but it is not a distinct expression.

To the extent that the "edition" concept is entrenched in conventions of publication and research, it is the "edition statement" which must be acknowledged as a significant identifying feature, to be carefully recorded in bibliographic descriptions.  It is possible, although somewhat difficult, to write rules of description in terms of "edition statements" and to ignore completely the underlying concept of "edition."  The Anglo-American cataloging tradition tends to accept edition statements at face value: if an item calls itself an edition, we tend not to ask whether it exhibits all the criteria included in the definition of the term "edition" (of which more later); we transcribe the statement and thus treat the item as an edition.
This works quite well (particularly if we ignore for the moment the use of the term "edition" in the names of several ISBD data elements).  However, the concept of "edition" is absolutely central to our practices about matching cataloging copy and determining when to create a new record.  Here the FRBR schema is no help whatsoever.

First, the expression entity in FRBR is not useful, because it is by definition abstract; only when an expression of a work is embodied in a manifestation is there a concrete entity that can be described.  An expression, to reduce the matter to absurd simplicity, may have a title, but it does not have a title page!

The manifestation entity in FRBR is no more useful, because any difference in any attribute results in a distinct manifestation.  Traditionally, we do not make separate bibliographic records for printings, and we often generalize minor differences in nonbook materials into a single record.
The AACR2 definition of "edition" contains two components which - in the FRBR schema - are contradictory:  an edition is defined in terms of its content ("embodying essentially the same content") and its issuance ("issued by the same entity").  Differences in content are different expressions, differences in issuance are different manifestations.  It is the combination of these two factors that traditionally determine whether a new record will be created.

The current text of AACR2 contains almost no guidance about when to make a new record - indeed, almost no recognition of the shared-cataloging environment in which we all work and which is perhaps the 20th-century's major contribution to the practice of cataloging.  The absence of rules in AACR2 has been remedied by an LC rule interpretation which offers some very general guidance and (most influentially) by the "When to input a new record" section in OCLC's Bibliographic Formats and Standards (various eds.).

OCLC's approach is both principled and pragmatic.  It references the concept of "edition" as the basis for separate bibliographic records, but also contains a detailed list of data elements which may differ "significantly" and therefore merit a new record.

When CC:DA appointed a Task Force to draft an appendix on when to create a new record for possible inclusion in AACR2, this OCLC approach was highly influential.  The draft appendix contained short "Basic Guidelines" and an extensive listing of differences and changes which were considered "major" and therefore required a new record.  These specific guidelines broke down in several ways.  First, it was noted that the guidelines had to deal with two distinct situations: differences between manifestations of a finite resource and changes within a continuing resource.  In addition, the rules had to cover single-part monographs, multipart items, integrating resources, and serials.  The result was a very complex set of specific guidelines.

Ultimately, the Joint Steering Committee decided not to include the appendix in AACR, although they plan to include the "Basic Guidelines" in the General Introduction.  They encouraged ALA to publish the guidelines as a stand-alone publication, and the Task Force has just produced the draft of such a publication, appropriately entitled Differences Between, Changes Within: Guidelines on When to Create a New Record, available on the Web at  http://www.ala.org/alcts/organization/ccs/ccda/tf-appx9.pdf.

For your consideration, here are the "Basic Guidelines":
Consider differences between manifestations or change(s) within a manifestation in terms of the content, the description, and other aspects of the manifestation (e.g., language, potential access points) in making a decision whether a difference or change is MAJOR or MINOR.
1.	Differences between the content of two or more manifestations or changes to the content of an existing manifestation require a re-evaluation of all areas of the description for the manifestation(s).
2.	Not all differences or changes are of equal importance.  To determine whether a new record is necessary for some differences or changes, a cataloger needs to look at all aspects of the manifestation and consider them in conjunction with the specific guidelines for the type of manifestation.
a.	A MAJOR difference between manifestations or a MAJOR change to an existing manifestation in any area of the description takes precedence over any MINOR difference(s) or change(s) in other area(s), and therefore requires a new record.
b.	A MINOR change, if considered important, may require adjustments to the existing bibliographic record and additional access point(s).  Adjustments for MINOR changes, depending on the type of issuance, may take the form of adding or changing notes in the record, or of modifying elements in the body of the description.  See the rules in the appropriate AACR2 chapter for guidance on how to record changes within the record.  See AACR2 chapter 21 for rules on access points.
3.	Decisions regarding the choice of main entry are made on the basis of AACR2 chapter 21 and, if applied by a cataloging agency, AACR2 chapter 25.  Since the main entry for a manifestation may or may not be reflected in its description, a cataloger should consider any differences between the main entry in the record and that on the manifestation separately from differences in the elements of the description.
4.	In comparing an item to an existing record, a cataloger must try to ascertain that the comparison is based on the same issue/part/iteration and the same prescribed source as was used in constructing the existing record.
5.	Any variations between printings or production runs that represent production errors will be considered minor, and will not result in creation of a new record, with the possible exception of rare book cataloging.

With regard to principle, the message here is that separate bibliographic records are made for distinct manifestations - but only if they contain major differences or changes.  With regard to pragmatics, the decision-making process eventually abandons the conceptual schemas and examines the individual data elements which comprise the record, seeking to identify those where differences or changes are sufficiently significant - sufficiently indicative of the underlying facts of content and issuance that make up the concept of "edition" - to warrant a separate bibliographic record in our catalogs, particularly in our shared-cataloging databases.

This is the current state-of-the-art with regard to AACR2 and FRBR and with regard to the question of when to create a new record.  These developments in general cataloging will need to be applied - with appropriate differences - to the cataloging of rare materials.

Application to Rare Materials

I'm not going to dwell at length about what makes rare materials different, and why they require special treatment.  We can all brainstorm our own lists.  In general, rare materials are rare and are valuable research materials because of their differences, even the unique features of individual copies.
At the same time, catalogers of rare materials work in the same shared-cataloging environment as catalogers of other materials, and they desire to obtain the same benefits from the systems which support both shared cataloging and (even more important, perhaps) sharing of bibliographic information on a global scale.  Therefore, standards and guidelines are important in cataloging rare materials - although they need to be unusually flexible in order to deal with the special characteristics of rare materials and their use.

Before moving on the some suggestions about how such standards and guidelines might be developed, I'd like to make a few points about what makes rare materials different.

We know more about them than about other materials.  We examine them more carefully, looking for significant features.  We describe them in more detail.  Collectively, we study them and compare them; bibliographic and textual scholars, curators, and collectors have accumulated a significant body of information about many of rare materials in our collections.  And as catalogers, we try to be aware of all that information when describing our own rare items.  We compare our copies with descriptions of other copies, we take account of the bibliographic scholarship and confirm its application to our copies (is that leaf a cancellans in our copy too? does our copy also have a particular catch-word on a particular page?).
Beyond that, we seek out and record the unique features of our rare items: binding, bookplates and other provenance evidence, manuscript annotations.

We do this because we understand that users of our collections may find these different or unique features of great significance.  We therefore want to share this information as widely as possible through the bibliographic records we create.  There are various ways in which this can be done, as will be discussed at the end of this document, but the need to share a great deal of very detailed information is the basis for our decision-making process in deciding when to create a new record and what to include in that record.

Recommendation #1:  Default Guidelines and Principles for Varying from the Default

It is impossible to agree upon a simple criterion that will be universally accepted by all catalogers of rare materials and applicable in every situation.  There are simply too many factors which are relevant to particular institutions and their collections, not to mention the specific features of individual titles and copies, that must be taken into account in deciding whether to create a new record in a given situation.  Therefore, I propose that Descriptive Cataloging of Rare Materials should contain a set of default guidelines, to be applied whenever there are no mitigating factors, and a set of principles for varying from the default based on the existence of mitigating factors.

Default guidelines.  As a starting point, I suggest that the ongoing work being done by CC:DA be examined to determine whether these guidelines, developed for general cataloging, can be applied to rare materials.  If for no other reason, consistency of treatment across all materials seems like a worthy goal.  Upon examination, this might turn out to be inadequate for rare materials, even as a default.  If so, I suggest that archival practices - particularly those developed for original graphic materials and for archival moving-image materials - be examined for relevant criteria and descriptive techniques.

Principles for varying from the default.  I'm not sure what such a set of principles would look like, but I hinted in the previous section at some of the factors that seem relevant to me:
ü	Context: Are there other copies of the same edition [I'm going to assume, for the purposes of this discussion, that the concept of "edition" is a significant feature of the default guidelines; this may not be true, but some sort of shorthand is needed to keep the discussion simple] in the collection that are not identical copies?  Are the differences such as would be identifiable from a standard description?  Are the differences or unique features significant either to the local institution or more broadly?
ü	Research potential:  Is the material likely to be used by researchers at a level of detail in which the differences or unique features may be significant?
ü	State of scholarship:  Have differences within the edition been identified by scholars?  Is there a standard bibliography that separately lists impressions, issues or states within the edition?  Can the item in hand be matched to a description in such a bibliography?
There are many additional factors that could be listed and formulated into a set of principles for making separate records, even if not mandated by the default guidelines.  This section will need to be significantly expanded, and I suggest this as an important topic for the Working Group.

Recommendation #2:  Local Practice vs. Shared Records

Catalogers should probably feel greater freedom in creating separate records within their own local catalogs than in contributing such records to shared-cataloging, master-record databases.

The use of the expression "shared-cataloging, master-record databases" is important, and I want to avoid using the term "utility."  Of the two main utilities, OCLC is a shared-cataloging, master-record database; RLIN is not.  Records in RLIN are essentially local records, and only the clustering makes them exemplars of any larger construct.  It is only in a master-record system where it becomes fairly important to have some consist guidelines on when to create a new record.  In a case in which there are two issues of an edition (to take a simple case), it seems unfortunate if one institution describes, and other institutions indicate their holdings of, the edition as a whole, while other institutions describe or indicate their holdings of the separate issues.  It then becomes difficult to determine what institutions hold what items.  If this is already the case, that is unfortunate, but reducing the confusion would seem to be a worthy goal.

This argument - at least in the case of a general shared-cataloging database like OCLC, i.e., one not limited to rare materials - would reinforce the tentative suggestion made above that the default guideline should be the same as that applied by general catalogers.  If nothing else, this is a result of the impossibility of determining what is and is not rare.  So perhaps the developing guidelines for when to make a new record should be applied to rare materials within a shared-cataloging, master-record database.

This same argument would also argue against varying from the default in such a database.  It is not clear that this is the best practice.  As indicated above, catalogers of rare materials record a wealth of information about the materials they describe, and they do this in order to share that information with other catalogers and with potential users of the materials, in the hopes that scholarship will be advanced by sharing such information.  Records contributed to shared databases are one of the most significant ways in which information is shared, and it would be unfortunate to limit the depth of information included.

However, shared databases are only one of the ways in which bibliographic information is shared.  The nature of a global information structure for communicating information about rare materials is something the Bibliographic Standards Committee discusses frequently.  We have developed the MARC formats to allow us to record additional types of information, particularly information unique to individual institutions or individual items.  The formats at the moment contain two sets of features that might be used for this purpose: the institutional/copy identifier ($5) in the bibliographic description itself, and the MARC holdings format.  Use of the first technique has been of limited use because of our heavy reliance on a master-record database which does not support much in the way of institution-specific data.  Use of the second has been little exploited because of the limited implementation of MARC holdings in shared databases, particularly the failure  to maintain and provide access to holdings records across institutions.  OCLC is undertaking a major redesign, and there are indications that MARC holdings will play a significant role in the new design, aiming at a virtual catalog of each member institution within the shared catalog.  If this goal is realized, then our task may be in providing guidelines on what level of granularity to describe in the bibliographic record (perhaps edition as a default) and what to describe in holdings records.

There is another technical development that offers an alternative model for constructing a virtual universal catalog.  Within the past ten years, most of our institutions have exposed our local catalogs to the Internet.  If we share our local catalogs with the world, then it becomes less critical to share all of the details in shared databases.  Perhaps it is sufficient to inform potential users that a title is available at a particular institution; a check of that institution's catalog with provide all of the information about any differences or unique features of the local copy or copies.

It is not yet clear which of these models (and there may be more) offer the best prospects for meeting our needs.  However, the discussion is relevant to the question of when to make separate records in what catalogs, and I would suggest that we think about these larger issues of bibliographic communication.

Conclusion

Tentatively, I would suggest that the best solution to the problem of when to create separate bibliographic records should involve:
ü	Relying on individual catalogers to use their best judgment and their knowledge of their collections and their craft to describe their rare materials at whatever level of detail they deem appropriate.  They should make informed judgments about the alternative use of separate bibliographic records and of holdings records to record information about differences and unique features.  They should make sure that these records are shared in some appropriate manner with the world.
ü	Contributing to shared databases records that are based on the highest level of consistency possible without concealing vital information.  We should make our best effort to come up with guidelines that can be applied as broadly and consistently as possible, while also making sure that the most detailed information we have created is made available to potential researchers in a clear and convenient manner.

But the devil is in the details, and there are a lot of details that still need to be filled in.  I welcome continued discussion of the issues raised in this paper.




___________________________
Deborah J. Leslie, M.A., M.L.S. 
Head of Cataloging
Folger Shakespeare Library
201 East Capitol St., S.E.
Washington, D.C. 20003
202.675-0369 (p)
202.675-0328 (f)
djleslie at folger.edu
www.folger.edu