Cataloguing principles and rules ensure that bibliographic / catalogue records contain structured data about information resources and are created in a consistent manner within the various catalogue and metadata formats. Today 'catalogues' (in the widest sense) need to provide access to a wider range of information carriers, with a greater depth and complexity of content.
While building on the Anglo-American Cataloguing Rules (AACR), the work on Resource Description and Access (RDA) is going back to basic principles and aiming to develop a resource that can be used internationally by a wide range of personnel working in different areas.
RDA is the working title of a new cataloguing code based on AACR, which is currently the world's most used content standard for bibliographic description and access. It is being developed by the Joint Steering Committee for Revision of AACR (JSC) [1].
The origins of AACR go back to previous cataloguing codes. In 1841, Panizzi's cataloguing rules for the British Museum [2] set out the concept that each book should have one principle entry in the catalogue, while Cutter's Rules for a Printed Dictionary Catalogue [3], published in 1876, contained 369 rules for descriptive cataloguing, subject headings and filing. Separate UK and US rules were then in use up till the middle of the twentieth century.
Two sets of internationally agreed principles were drawn up in the 1960s. The Paris Principles were the work of the International Conference on Cataloging held in Paris in 1961 [4], which had representation from 53 countries, and drew on the work of Seymour Lubetzky. The Copenhagen Principles were agreed at the International Meeting of Catalogue Experts in Copenhagen in 1969 [5], and were the impetus for the International Federation of Library Associations (IFLA) [6] to start work on the International Standard Bibliographic Description (ISBD) [7], first published in 1971. The ISBD organises the description of an item into distinct areas: title, statement of responsibility, edition, resource specific information, publication, physical description, series, notes and standard number identifiers.
The first edition of AACR was published in 1967, in separate British and North American texts. The second edition (AACR2) - published in 1978 - unified the two sets of rules, and also made them consistent with the ISBD. Key principles of AACR include cataloguing from the item 'in hand' rather than inferring information from external sources and the concept of the 'chief source of information', which is the preferred source where conflicts exist.
Since publication in 1978, a number of amendments and revisions have been made to the text of AACR2. In 1997 the JSC invited experts world-wide to an International Conference on the Principles and Future Developments of AACR in Toronto [8]. Participants identified a number of issues of concern - principles, content vs carrier, logical structure of the Rules, seriality and internationalisation. Revisions of the text post-1997 addressed some issues in a limited way, but there was a growing recognition that a new resource was needed, not simply a third edition of AACR.
This view was supported by the 1998 IFLA study on Functional Requirements for Bibliographic Records (FRBR) [9], which reinforced the basic objectives of catalogues and the importance of relationships for users to carry out the basic tasks of 'find', 'identify', 'select' and 'obtain'. The study produced a conceptual model of entities, relationships and attributes that are independent of communications formats or data structure. The model identified information content as having Work, Expression, Manifestation and Item level characteristics, which were related to a number of entities (persons, corporate bodies, concept, object, event and place). The structure of this theoretical model allows collocation at Work / Expression level. The challenge of incorporating this model and its concepts into the new cataloguing rules also contributed to the decision to create RDA in preference to a revision of AACR2, since it is not a simple matter of replacing an old term in AACR with a new term but usually requires complete rewriting of text sections.
So what are the issues and problems with AACR2 that demonstrate the need for RDA?
Over time, the rules had become more complex, in an attempt to cover increasingly diverse and complex resource formats. General rules were followed by chapters for specific rules covering books and other printed materials, cartographic materials, manuscripts, music, sound recordings, motion pictures and video recordings, graphic materials, electronic materials, 3D items, microforms and continuing resources (e.g. serials). RDA aims to phrase rules more simply so that they can easily be applied to a variety of resources with the minimum of specific instruction, and with the aid of suitably chosen examples.
The current structure of AACR2, noted above, is in itself the source of some problems. The arrangement of the descriptive chapters by resource type means that the cataloguer must first decide on the format or material type of the resource to be described in order to locate the appropriate chapters to use. However, there is no clear guidance as to how the decision should be made. While this may be obvious for some resources (e.g. a single textual manuscript), in others it will be less so - is a music serial primarily 'music' or a 'continuing resource'? RDA plans to address this by placing identification of the resource type as the first step in the cataloguing process, and therefore at the beginning of the general instructions.
The AACR2 structure of general and specific rules led to a situation where some resources were catalogued in slightly different ways to other resources. Alternative rules in some areas and optional additions to rules further complicated things, to the extent that two catalogue records for the same item could be created using AACR and yet have substantial differences. RDA aims to simplify rules so that they are more easily applied to a range of resources and to limit alternative variations in treatment.
Today's information environment is one in which individual information resources are available in a range of formats and often with a number of derivative works. For example, a novel may exist in hardback and paperback, standard and large print, Braille and Moon tactile formats, digital files and spoken word recordings on audio tape and CD. Additionally, it may be the source work for films, musicals, operas and ballets, or the inspiration for prequels, sequels and related works by the same or different authors. There is a whole range of relationships between these entities, which have been described in FRBR, but AACR does not directly reflect this theoretical model and the terms that it uses. These relationships are important in enabling catalogues to collocate appropriately records that are retrieved from searches to reflect these relationships. RDA will include a chapter on relationships in addition to incorporating FRBR terminology throughout the text.
The general and specific rule structure and the provision of alternative instructions, led to a situation where rule interpretations were developed to assist cataloguers. RDA aspires to develop a set of rules that is based on clearly stated principles, with limited alternatives, that will enable cataloguers to build their own judgement and expertise.
Bringing AACR into line with ISBD resulted in the use of 'general material designation' (GMD) and 'specific material designation' (SMD) terms. Potentially these terms can be used in systems as filtering parameters (find only DVD version(s) of a film and not video), display sequence parameters (for a single work with 8 text versions, 3 spoken word recordings, 1 digital file and 1 Braille text) and content information (musical notation and recorded performances of a musical work).
To date, use of GMD and SMD terms has been limited and patchy, partly due to problems inherent in the lists of terms. There are 2 GMD lists, one British and a slightly longer North American/Australian one. Confusingly, GMD terms cover both content (e.g. music, cartographic material) and format (e.g. Braille, filmstrip). Additionally, as the range of information carriers increased rapidly, SMD terms became increasingly out of touch with user practice: authorised terms include 'sound cassette' and 'videodisc' and not the more commonly used audio tape and DVD. In an effort to address this, recent revisions of AACR2 have included some options to use such product-related terms. RDA will replace the GMDs and SMDs with a more flexible approach, within which existing content and carrier terms can be defined without precluding the use of other terms that may be required in the future.
AACR evolved in the English-speaking cultures of Britain and North America and certain aspects of it have an Anglo-American bias. Meanwhile, other countries had developed their own sets of cataloguing rules, which had both common ground with, and divergence from, AACR.
One example is the Regeln für die alphabetische Katalogisierung (RAK) [10] used in Germany. RAK is mainly based on the Copenhagen and Paris Principles, and replaced the Prussian Instructions (PI) [11] as the standard cataloguing rules in Germany in the 1970s and early 1980s. There are major differences between RAK and AACR, although both are based on the Copenhagen and Paris Principles.
The library world was quick to recognise the economic advantage of the 'create an electronic catalogue record once, and reuse it many times' principle, with the result that there is globally a great deal of bibliographic data exchange. As this exchange increases, it has highlighted the obstacles to exchanging data created to different sets of rules. The Anglo-American bias in the AACR2 is seen as a barrier to wider adoption in some areas of the world, where the information community is increasingly international. The aim with RDA is therefore to base it on internationally agreed cataloguing principles, and to remove instances of Anglo-American bias in the new rules.
While earlier forms of catalogue still exist, the current environment is of online catalogues and searching across different metadata formats. There are many new metadata formats and packaging (communications) formats and increasingly these are used in on-the-fly 'mash-up' connections.
Today, a wider range of personnel than previously - authors, administrators, cataloguers, etc. - create bibliographic metadata, with varying levels of skill and ability. Some metadata is even created without direct human input, using computer software.
RDA aims to be independent of any communication formats, the number of which has increased greatly in recent years. The library community is familiar with ISBD and with machine-readable cataloguing, though even this now encompasses a range of formats: UNIMARC [12], MARC 21 [13], MODS [14], MADS [15] and MARCXML [16]. Other communities have developed Dublin Core [17] for Internet resources, EAD [18] for archival descriptions, the VRA Core Element Set [19] for works and images of visual culture, and MPEG-7 [20], the multimedia content description standard.
A structure of committees and organisations underpins AACR. The Committee of Principals oversees the work of the AACR Fund Trustees/Publishers and the Joint Steering Committee for Revision of AACR (JSC). The JSC is made up of representatives of the American Library Association (ALA) [21], the Australian Committee on Cataloguing (ACOC) [22], the British Library (BL) [23], the Canadian Committee on Cataloguing (CCC) [24], the Chartered Institute of Library and Information Professionals (CILIP) [25] and the Library of Congress (LC) [26].
In the UK, the JSC representatives from CILIP and the BL are members of the CILIP/BL Committee on AACR [27], along with four members nominated by CILIP and three members by the BL. There are also several invited members with backgrounds in specific areas such as digital or art materials or in the national libraries of Scotland and Wales and the committee contacts specialist groups when considering specific proposals and issues.
Creating the text of RDA is a collaborative process that is timed to work in the context of what are currently twice-yearly meetings of the JSC. The JSC and/or the Editor (Tom Delsey) create drafts of the text, while any of the constituent committees can put forward proposals for changes to rules or for new rules. These drafts are issued for constituency review, with JSC members responding on behalf of their respective committees or organisations. Where there is consensus, the Editor will take the next steps in preparing new drafts. In other cases, working groups will be set up to review specific areas and make further proposals or recommendations. A further round of drafts and proposals will then be issued for comment. This process continues until a final approved text has been agreed.
The JSC has also made the process more open by publishing most documentation online. It has also engaged with other communities during the review process in order to ensure that RDA can be applied widely.
The JSC has a number of strategic goals for RDA. They are to:
RDA will contain new introductions, content rules and updated examples, will cover authority control, use FRBR terminology and simplify the text to improve consistency. It will also reach out to other communities to achieve greater alignments with other standards.
It will be offered as a Web-based product as well as a loose-leaf printed version. The Web-based version will have added functionality (e.g. internal and external links) and the option of customisable versions is being explored. Potentially there could be full, concise and tailored (e.g. serials, cartographic) versions. There is also the opportunity for RDA (or parts thereof) to be incorporated into data input templates and task-oriented work-flows by system vendors.
Part A covers description, relationships and related resources. General guidelines will cover identification of the resource type, technical description, content description, and sourcing and item-specific information. New categories of carrier and content will be introduced.
The section on relationships will cover those between FRBR bibliographic entities and agents (persons, etc.). There will be a simplified choice of primary access points for citations of works, and special rules will be simplified or eliminated.
Part B on authority control will cover purpose, scope, and authorised and un-authorised forms. Appendices will cover display standards, ISBD, capitalisation, abbreviations, numbers and a glossary.
JSC is aiming for an initial release of RDA in 2008. This will be the culmination of a three-year period of intense activity. The initial proposals envisaged a three-part structure (covering resource description, relationships and access point control). The draft of Part 1 issued in December 2004 received much negative comment on the approach being taken, with the result that the JSC decided to abandon the draft and go back to the drawing board.
The prospectus for RDA was issued to the committees in July 2005, followed by a draft of most of Part 1 in December 2005. Responses were more favourable to the new approach, although a number of issues were identified as requiring further work. 2005 saw additional work on the issue of content and carrier. JSC set up a working group to review this area; its report recommended that content and carrier information was needed for both search and display purposes and, since current listings of GMDs and SMDs were not appropriate, proposed new sets of terms. In a parallel initiative, JSC worked with the creators of ONIX [28], the book trade XML message format, on harmonising the ONIX material categories with RDA content and carrier terms. These two reports were used by the Editor to produce a draft for the relevant sections of RDA.
In April 2006, drafts of the remaining chapters of Part 1, now designated Part A in a new two-part structure, were issued for review. Responses to these documents will be considered at the October 2006 JSC meeting. The period between October 2006 and April 2007 will focus on Part B, while the period May to September 2007 will consider the general introduction, appendices and glossary.
Work is already being undertaken in some of these areas. Two working groups have been set up to look at examples required for inclusion in RDA. Their remit is to make a thorough review of existing examples, with a view to recommending the omission or replacement of some existing examples and proposing new examples where these are needed. Work is also underway on revision of the appendices. At present the following appendices are envisaged: capitalisation, abbreviations, initial articles, presentation of data and a glossary. There is a move for RDA to permit less use of abbreviations in bibliographic records (for example: s.l. and c.). However, this raises issues about the language of full text replacements: 'date not known' and 'date not stated' are not equivalents. A counter-argument suggests retaining abbreviations, such that systems could be designed to display full text in the language of the specific catalogue.
Proposals have been put forward to extend the lists of initial articles with lists for several additional languages, with the result that this is now becoming an extensive document. This increase in size and the question of future additions and maintenance have prompted suggestions that this information should not be part of RDA but a supplementary work.
During early discussions about RDA, the CCC had proposed that there be a mapping of specific RDA rules and the appropriate / relevant field(s) in MARC 21. The JSC agreed that this would be useful and the task of carrying out this work was devolved to CCC and ACOC. The work will identify where RDA will affect MARC 21 and also where existing data provision in MARC 21 may indicate a need for inclusion within the rules in RDA. The JSC plans to submit a discussion paper on these areas to the January 2007 MARBI [29] meeting.
RDA is a new standard for resource description and access, designed for the digital environment, aimed at all who need to find, identify, select, obtain, use, manage and organise information. It is a multinational content description standard covering all media, that is independent of technical communication formats. Input from a wider range of communities has been sought, and progress can be followed on the JSC and CILIP/BL Web sites.
Ariadne is published every three months by UKOLN. UKOLN is funded by MLA the Museums, Libraries and Archives Council, the Joint Information Systems Committee (JISC) of the Higher Education Funding Councils, as well as by project funding from the JISC and the European Union. UKOLN also receives support from the University of Bath where it is based. Material referred to on this page is copyright Ariadne (University of Bath) and original authors.