Bibliographic description

A primary function of the national bibliography is the accurate bibliographic description of the works it contains. Rules for the creation of bibliographic descriptions have evolved in sophistication over the years and have developed from institutional, to national and ultimately global standards. NBAs should ensure that their metadata is created according to a recognised bibliographic description standard in order to ensure the utility, interoperability and efficiency of their bibliographic services.

Cataloguing rules compatibility

Compatibility of cataloguing rules is as essential as compatibility of formats. While formats govern the structure of metadata, cataloguing rules govern its content.

Two events marked the start of international initiatives on the definition of compatible cataloguing rules:

  • 1961, the Paris Principles made recommendations on the choice, form and structure of headings for names and for titles
  • 1969 the IFLA International Meeting of Cataloguing Experts held in Copenhagen (Chaplin, Anderson & HonoreĢ, 1970) produced a resolution that proposed creation of standards to regularise the form and content of bibliographic descriptions.

Core national bibliographic elements

The starting point for selecting core metadata requirements is the analysis of uses of the national bibliography since use is only possible if appropriate metadata are available. Users will inevitably have different needs relating to standard functions, such as find, and may even have unique requirements. National bibliographic agencies should avoid creating metadata of specific use solely to individual users although they may offer a facility for users to assign their own tags or annotations.

Analysis of user requirements should go beyond requirements for resource discovery as articulated by FRBR but should also take account of requirements for:

Publishing activities, such as:

  • Rights management
  • Market research

Library functions, such as:

  • Collection management
  • Derived cataloguing
  • Data management

Some data elements required to support these functions are not normally considered to be in scope for resource description. It is not recommended that NBAs should create this information but instead, where possible, NBA’s are encouraged to link to trade sources and to reuse metadata created by other agencies.

Granularity of description

A basic decision upon metadata requirements has to be made concerning the unit described. Does the metadata of a national bibliography pertain to a collection, to an item, or even to parts of items? User needs as well as the national bibliographic agency’s capacity, together with the amount of the national output and the material collected will affect this decision.

Collection-level description

Collection-level description (sometimes also: archival description) consists of metadata describing a collection as a whole, in contrast to item-level description (manifestation description in terms of Functional Requirements for Bibliographic Data (FRBR)). The application of collection-level description for national bibliographic agencies may lie in the national bibliography coverage of special material, such as learning objects, Web sites or some types of archival collections like ephemera. Collection-level description for library collections has not been subject to standardisation.

Title description

Library metadata most commonly describes items – documents in any physical form, treated as an "entity”. The item forms the basis of the bibliographic description. If a national bibliographic agency also decides to cover electronic resources from the World Wide Web it may at the same time find it impossible to cover all of the increased national output on a title-level description.

Component description

An NBA may, in some cases, decide for the description of items such as journal articles, chapters and other parts of monographs, individual tracks on sound recordings, and other types of “works within works”. This lowermost level description may be useful for special collections as “analytic” finding aid. A framework for this level of description exists with the Guidelines for the application of the ISBDs to the description of component parts.

Level of description detail

The NBA’s responsibility to create extensive records of its national imprint as recommended by the 1977 Paris conference and affirmed by the 1998 Copenhagen International Conference on National Bibliographic Services was predicated on predominantly print resources and is no longer sustainable.

The national output now includes an increasing proportion of electronic resources which may soon become the predominant form of publication. The expansion of the national output may therefore exceed the capacity of the NBA to comprehensively process resources. A graduated approach will therefore be required in which the level of descriptive cataloguing appropriate to different types of resource will be determined in relation to:

  • The level of metadata already associated with the resource
  • The significance of the resource for the national bibliography
  • Its content, not its carrier

The first recommendation represents a significant change of approach to the creation of the national bibliography and implementation of cataloguing levels will have a direct impact on users which NBAs must therefore consider. In addition processes must be simplified and, where possible, automated. Additional efficiencies should also be gained via cross sector collaboration and by elimination of barriers to data exchange.

The IFLA Bibliography Standing Committee Working Group recommended four potential levels of description:

  • Authoritative
  • Comprehensive
  • Enhanced
  • Basic

Authoritative

Authoritative level description:

  • Denotes the highest level of metadata assigned to a resource
  • Offers the most flexible support for resource discovery
  • Includes (specified) access points that are controlled by authority records
  • Conforms to explicit content standards
  • Is the most expensive category to assign but NBAs can have confidence that the record can be reused with minimum intervention

Authority control:

  • Over name, title and subject access points supports accurate identification of resources
  • Of names and titles supports clustering for economy of display
  • Reference structure supports navigation to related resources, identities or topics

It is recognised that creating authoritative metadata for all potential resources is likely to be beyond the capacity of most NBAs. Thus such agencies must develop criteria to determine which parts of the national output should receive this level of description. The guiding principle should be the value of the content of the resource rather than its format or carrier.

Comprehensive

Comprehensive level:

  • Denotes assignment of extensive metadata for description and access
  • Explicitly conforms to published content standards but there is no commitment to support any access points with authority records.
  • Does not fully support user tasks

Failure to control access points with authority records inhibits clustering in displays, resulting in multiple sequences for individual entities, such as names of authors, works or topics and a degradation of service to the user. The find task is not fully supported, navigation between resources is compromised and unambiguous identification of resources will be constrained.

Examples of comprehensive metadata may include MARC records, qualified Dublin Core, ONIX or other book seller records.

Enhanced

Enhanced level description:

  • Indicates that the basic resource has been supplemented by assignment of some metadata
  • May not explicitly indicate the content standards used nor is there any commitment to authority control of access points. Note: The absence of clear identification of content standards means the NBA cannot be certain of the consequences to the users
  • Is not recommended for use in national bibliographies and should be upgraded over time if it is to be incorporated into the national bibliography
  • Does not adequately support resource discovery tasks

Examples of enhanced metadata which could be re-used by the NBA include unqualified Dublin Core and TEI headers.

Basic

Basic level description:

  • Assumes no metadata have been assigned to the resource, either by the bibliographic agency or by the source
  • Is likely to include web resources harvested directly with no metadata or metadata produced by automatic processing
  • Is most appropriate to textual resources of relatively low value harvested from the web and is not appropriate for graphic or other non-textual resources
  • May be enhanced over time if it is to be incorporated into a national bibliography
  • Is the most economical for the NBA to create, but the most expensive for the agency to upgrade

Resource discovery is entirely dependent on natural language searching with results returned corresponding to the query terms used. In a standard web interface results will be ranked for display and this approach is most effective when the resource being sought is readily distinguished, (e.g. by its terminology), or when the user has a general information requirement that may be satisfied from a range of similar resources.

Provision of resource discovery at basic level is achieved by transferring costs to the end user who will not be able to navigate to related resources other than by native links present. Huge result lists may however make it difficult to refine results sufficiently to select a specific resource.

The FRBR Basic Level record:

  • Is basic in name only and is actually sophisticated metadata which satisfies requirements necessary to support the user tasks: find, identify, select and obtain defined by FRBR as fundamental to resource discovery.
  • Specifies relationships between the resource described and other entities. The identification and description of such relationships support subsidiary functions such as exploration and database clustering.
  • Is expensive to produce and extending use of such records to match increased output of traditional and electronic media would be a challenge for most NBAs.
  • Is rooted in traditional concepts of bibliographic control. If these concepts are still valid in the digital age the library community must find ways to streamline the processes required to produce this level of metadata more efficiently.