Introduction
In a world where increasing amounts of information are available in a multitude of physical and digital formats and through a range of access routes and conditions, metadata remains a crucial element in matching the resources to user need. As library catalogues evolve to meet new user needs, more detailed metadata is required. The 1998 report on Functional Requirements of Bibliographic Records (FRBR)Footnote 1 has changed the way we think about cataloguing and bibliographic metadata. The past few years have seen a gradual move from multiple, often nationally-based, bibliographic formats towards a single format (MARC 21Footnote 2) accepted in many different countries, while other information-based communities have developed their own forms of metadata, such as Dublin Core (DC)Footnote 3 and Encoded Archival Description (EAD)Footnote 4.
But the model (FRBR) and the metadata schemas and formats are not in themselves enough. The third component is guidance on what should go into bibliographic records, and in what detail, and how to ensure consistency of referencing, collocation of variant forms of names and terms and linkage of related resources. Since 1967 guidance for the English-based cataloguing community has been the Anglo-American Cataloguing Rules (AACR(UK), then AACR2)Footnote 5 but things are about to change. Enter Resource Description and Access (RDA)Footnote 6, the new cataloguing code. So what is it and how will it change things for you?
The case against AACR2
The second edition of AACR (AACR2) was first published in 1978. Since then it has been widely used in the creation of millions of bibliographic records and undergone a continuous revision process built on the principles of consultation with the cataloguing community, via a number of committees and consensus decision-making.
But revisions built an ever-more complex text and from 2002 it became increasingly evident that tweaking the text would not solve the issues posed by new resource types and publication practices. An initial attempt to draft a third edition met with criticism for not adequately addressing the perceived flaws in AACR2 and community consensus favoured a new text to reflect current theoretical models and practical issues. So what exactly was wrong with AACR2?
Perhaps the most obvious issue was that the text had a distinct Anglo-American bias due to its original inception as a tool for English-language-based catalogues, though this was paralleled by biases to other languages and cultures in the cataloguing codes developed in other communities. For everyone, the cost of maintaining a cataloguing code has become an increasingly expensive operation but the inherent biases worked against any one code of practice becoming an international standard.
AACR2 is based on the concept of single item cataloguing, although ‘single item’ covers multi-piece resources such as multi-volume works, serials and kits, while only a few relationships, e.g. editions and title changes were recognised. In contrast, the FRBR model maps a whole range of relationships between works, expressions, manifestations and items and with agents, people and corporate bodies, and subjects and events.
As new digital formats were created, the editorial committees were faced with almost continuous revision. Revisions typically addressed a specific issue at a single point in time. This approach meant that often cataloguers needed to flip back and forth between the general rules and the format specific rules. The new thinking places value on more generalised instructions that can be applied even to resource types not yet created.
The newer resource types highlighted a long-standing issue over content and carrier data. Within AACR2 this data was conflated, where now the need is for separate data elements to enable new catalogue functionality. Life was simpler when books, pamphlets, serials, maps and music were all we had to worry about. The pace of development and take-up of new information carriers is sometimes so fast that some of the earlier new formats become virtually obsolete in a matter of years. Think how music recordings have gone from vinyl to audio cassette, then to CD and now MPEG3 files, in a short space of time.
Today resources may not be located physically in the library or even held on a library computer or server. E-books and e-journals are accessed by subscription and regularly updated web sites are today's equivalent of loose-leaf binders. Stopping a subscription can mean not only no new items but also losing the back-run. Some e-textbooks are licensed only for students following a specific course, so catalogues need to provide information on who can access a resource and under what conditions. This is a very different world from when libraries permanently acquired, by purchase or subscription, physical items to which they controlled the access.
Further, although libraries often used both AACR2 and a bibliographic format, often MARC, to create catalogues, there was no formal relationship between the rules and the formats, although some mappings had been done, and crucially the digital community created Dublin Core and other metadata schemas without reference to content guidance such as AACR2.
The evidence
Everyone knows of Romeo and Juliet – but what exactly does that mean to you? Perhaps it is Shakespeare's play that comes to mind, but this exists in many forms: the play-script, full, abridged or edited in English or a translation to another language and recordings, audio, film and video of performances with various casts and versions of the text.
There are relationships with other works. Shakespeare's play was inspired by English translations of an Italian story: a 1562 verse translation by Arthur Brookes and a 1582 prose translation by William Painter. Other people used Shakespeare's play as the basis of ballet, having both musical scores and choreography, opera and musical theatre productions, each of which may have been recorded in different media and with varying casts, as well as graphic novel and animated versions.
Other relationships may be relevant to a user. Creators of these works often created other works. Shakespeare wrote many other plays and poetry. Prokofiev composed scores for several other ballets as well as other musical compositions. Directors and performers of Romeo and Juliet in play, ballet, opera and musical theatre work on other performances, films or broadcasts. And just to add to the mix, there can be variant versions of the names of any of these works or people.
Where does the user fit into this? What do they want? For a drama student it might be a specific version of the play-script, perhaps including critical comment. A musician needs the score for their instrument, but for performance it must be in the same edition as the rest of the orchestra. A dancer might want both choreographic notation and a musical sound recording for practice and maybe also a recorded performance. Sensory impairments may restrict usable format choices: a visually-impaired person might want an audio recording of the play or a Braille version of the script, while a hearing-impaired person might want a sub-titled version of a recorded performance.
Thus so much information is potentially relevant to the user, but to produce effective catalogues this needs to be split into small, discrete pieces of data. The conclusion was that AACR2 needed urgent revision, but that producing a new edition would not address basic problems inherent in the current structure and presentation of the rules.
The case for RDA
RDA had a definite focus from the beginning. It would have a logical structure based on internationally agreed principles, hence the use of FRBR terminology in the text. The rules should be easy to use and apply, especially in an online, networked environment and provide effective bibliographic control for all types of media. Significantly, in recognition of the new information landscape, the new code should be compatible with other standards and both relevant to, and capable of, use outside the library community. Indeed, the development process became increasingly open as time went on, with draft texts being made publicly available and responses received from national cataloguing communities outside the original AACR partnership and from the digital and archival communities. This has built an international community of interest centred on, but not restricted to, the library domain.
It was not necessary to abandon everything in AACR2, as the principles enshrined in many of the rules have stood the test of time both within and without the library community, but it was time to re-think and re-phrase. Where possible, instead of separate general and media specific rules for the same type of data, single rules have been created that can be used for any type of resource. Sometimes the generalised instruction is supplemented by alternatives e.g. if cataloguing rare books and there are many more examples than in AACR2.
The Anglo-American bias has been largely removed and RDA mandates the use of the language and script preferred by the agency creating the record. A controversial decision was to replace abbreviations with the spelt out words or phrases as it was felt that these could be more accurate.
For example s.l. could be used for either ‘place not recorded’ or ‘place not known’ which are two entirely different statements.
There are other changes in the language. AACR2's authorised form of a name is RDA's “preferred” name access point, while added entries or references are “variant” access points. The word “resource” is used frequently for whatever is being catalogued, so where you see “resource” think “insert here the type of resource I am cataloguing just now”. This future proofing approach should limit the instances where the text must be updated, as instead an example of a new type of resource can be easily added.
There are differences, of course. Statements of responsibility are no longer restricted by the “rule of three” and there are more options for entry under compiler. The intervening sub-headings of old and new testaments, which were based on Christian Bibles, have been removed so that Bible. Exodus not Bible. OT. Exodus, in a Hebrew Bible, which does not include the New Testament, Exodus is part of the Torah. Another difference is that treaties are to be entered under the first named party, rather than giving priority to specific countries or bodies.
A significant change is the separation of content and carrier data. Content is the cultural and/or intellectual information contained in the resource, while the carrier data describes how that content is stored and accessed. Content types are text, still and moving images, cartography, notations for music and movement, sounds and tactile forms, etc. RDA groups carriers by the media needed to access them. Resources that can be accessed directly by users belong to the unmediated carrier type: sheet, card, volume and object. Other carrier types include projected media films and filmstrips and computer/digital media CD, DVD and computer files. Separating this data in records means that new ways of limiting searches and presenting lengthy results lists would be possible in catalogues.
The “victim” statement
What will be the impact on cataloguers and catalogues? Since much of RDA contains the essence of rules in AACR2, albeit rephrased and re-grouped, the need for retrospective changes to records created according to AACR2 is likely to be small. Specific areas will be affected of course, as with changing access points for the Bible and first named entities for treaties. The scale of changes will depend as much on the cataloguing system as the actual number of records for which changes are needed and might be achieved relatively easily through global edits of authority records.
RDA looks huge if you print out all of the ten sections, 12 appendices, glossary and index and this linear, static version of the text can be hard to read. But it will be a different experience in the online resource, which will be its primary format. Hypertext links will enable cataloguers to move easily from one instruction to the next in a logical sequence, to a related instruction, from the index to the point in the text, or from the text to the glossary for a definition.
As an online product, it has the potential to be integrated into library management system modules, providing access to rules throughout the cataloguing process and enabling cataloguers to construct “MyRDA” with local examples and omission, or hiding, of unused sections and options. Other possibilities are tailored digital products, a concise text, texts for specialist areas e.g. law, music, rare books or a training resource. While RDA could be developed in all these ways, the final range of products will depend on collaborations and partnerships with system designers and vendors.
There will be a closer relationship with the MARC 21 Format and RDA Appendix D contains a mapping of format data elements to RDA rulesFootnote 7 and to accommodate RDA data elements some changes have already been made to the format and others are proposed. One accepted change is the creation of new fields for content and carrier information: 336 Content Type, 337 Media Type and 338 Carrier Type.
Judge's summary
Inevitably there have been problems and delays along the way. There was the decision to abandon work on AACR3 and then the change to developing text with an online product. The lock-down on working on the text while the online product is being developed has meant that some text has yet to be agreed or written and these text sections are detailed in the Deferred Issues documentFootnote 8. The text is the result of the efforts of many people, in committees e.g. CILIP/BL Committee on RDAFootnote 9 and working groups e.g. the two Examples Groups, the GMD/SMD Group, the RDA/ONIX Group, the Appendices Group, etc. However, the end is in sight.
The online product now has a name, the RDA ToolkitFootnote 10 and will be available by subscription in September 2010 following a free trial period from late June to the end of August. Subscription levels are based on the number of concurrent users and there are special arrangements for training and instructional usage. To ensure RDA reaches the widest possible audience, a full text loose-leaf print version in English will also be published in 2010 and the Co-Publishers will work with publishing partners on translations. More information can be found on options and pricing on the Pricing pageFootnote 11 of the RDA Toolkit Web site.
For the first few months use of RDA will be experimental. A number of libraries have signed up to carry out a formal test of RDA as a working tool and UK libraries testing RDA include the British Library. During the free access period, individual cataloguers will be able to assess RDA as a cataloguing code as well as how the online product works.
Finally, as RDA influences development of the MARC Format, there are now opportunities for system vendors to develop new features in OPACs, though these may take some time to emerge.
