Library and Archives Canada
Symbol of the Government of Canada

Institutional links

ARCHIVED - Initiative for Equitable Library Access

Archived Content

This archived Web page remains online for reference, research or recordkeeping purposes. This page will not be altered or updated. Web pages that are archived on the Internet are not subject to the Government of Canada Web Standards. As per the Communications Policy of the Government of Canada, you can request alternate formats of this page on the Contact Us page.

Strategic Research

FRBR and RDA: Advances in Resource Description for Multiple Format Resources

8. Conclusion: Impact of FRBR and RDA

8.1 Multiple formats issue

The cataloguing world has long wrestled with the problem of the book and its contents. Elaine Svenonius traces this distinction back to Anthony Panizzi, in the nineteenth century, and possibly as early as Thomas Hyde in 1674.99 Seymour Lubetzky's words, from 1956, identify this challenge:

The problem of cataloguing arises from the fact that ... cataloguing must concern itself not only with the book in hand but also with the work contained in it ... and ...with the fact that the reader's information about the name of the author and the title of the book are not infrequently imperfect.100

FRBR, the conceptual model, articulates a view of the bibliographic universe that is more precisely defined and nuanced than previous views. FRBR continues and expands on the understanding that there is a distinction between the book and its content. It clarifies the nature of the relationship between the book and its content.

The FRBR model brings a new understanding of bibliographic data by identifying the important entities in the bibliographic universe, the attributes of these entities and the realtionships between the entities. The problem of cataloguing is not simply the distinction between the book and its content. The categories of book and content have been made more precise and expanded into the four group 1 entities: work, expression, manifestation and item. The attributes of these entities and the relationships between them illuminate clearly the boundaries between content and carrier, and also open up the possibility of doing justice to both content and carrier when describing a resource.

The process of trying to align the existing cataloguing code, AACR2, with the concepts and modelling in FRBR resulted in a complete deconstruction of AACR2 and a rebuilding into a new standard, RDA. AACR2's approach, its organization and many of the individual rules were at odds with what had been learned from the FRBR model. In an effort to update AACR2 and prepare for the cataloguing of new types of publications, it became evident that amendments would not suffice and that a radically new approach was required.

There were several major issues that challenged the further development of AACR2. One of the most pressing problems was the multiple formats issue. The multiple formats issue, with its two aspects of alternative formats and multimedia resources, has its root in AACR2's inconsistent approach to content and carrier and inconsistent categorization of the classes of material. Alternative formats bring to the fore the unresolved problem of whether the content or the carrier should have primacy when describing a resource. Different approaches to the problem were unsatisfactory because they emphasized either the carrier or the content, to the detriment of the other. Resources consisting of multiple types of content and/or carriers were also not well served by AACR2 rules. AACR2 has a bias towards choosing one characteristic as having primacy. Such an approach may give the cataloguer a way to approach the description of the resource, but it does not necessarily allow for a full description of the resource, where all characteristics are equally well described.

The development of a new approach to content and carrier was a long and difficult task. The AACR cataloguing community went through years of working through problems associated with the primacy of the carrier and the boundaries between content and carrier, testing new ways to resolve issues and then going back to reframe solutions. These years of experimenting and debating did eventually lead to an approach around which consensus could be built. Achieving consensus was facilitated by the use of the FRBR conceptual model as the framework on which to build a new approach.

The FRBR model emerged out of the work of an international study group under the auspices of the IFLA (International Federation of Library Associations and Institutions). The model was drafted over many years and went through a period of world-wide review. Beginning immediately after its publication in 1998, the explanatory power of the model was acknowledged around the world as librarians and researchers began to apply the model and use it as the starting point for new research. Its broad base of acceptance was confirmed when it became a key part of the foundation for the new Statement of International Cataloguing Principles.

RDA was developed as a new metadata standard built on the theoretical framework expressed in the FRBR model (and the FRAD extension). Building on FRBR meant that there was a widely accepted theoretical framework to guide the development of the standard, and against which to test the standard and keep it logically consistent.

In terms of resolving the problems with alternative formats, FRBR provides the key in its modelling of the group 1 entities, work, expression, manifestation and item. By identifying four entities, analyzing their attributes and mapping their relationships to each other, FRBR offers a means to sort out the level of relatedness between resources. Elaine Svenonius' concise summary of the "defining objective" for the organization of information underlines that the key is the level of similarity and differences between resources: "to bring essentially like information together and to differentiate what is not exactly alike".101 The history of the description of alternative resources has been plagued by the tension of emphasizing either the similarity or the difference. When one has to choose whether to emphasize similarity or difference, then the other aspect remains less visible.

The problem of alternative formats can be approached more successfully by applying concepts in the FRBR model. With the modelling of the group 1 entities, there is a means to sort out levels of similarity and difference in more detail, and to record this information in unambiguous ways. Not all alternative formats differ from each other in the same way. Alternative formats can be divided into two types: 1) alternative formats where the content is expressed in a different form, i.e. different expressions of the same work; 2) alternative formats that are the same expression of the same work but are different manifestations. It is important to distinguish between the types of alternative formats, because the level of similarity or difference between the alternatives is information that can be of critical importance to the user when identifying and selecting the appropriate resource.

The RDA framework of content, media and carrier types clearly indicates the level of similarity and differences between resources. A difference in content type means a different expression. A difference in media and carrier type means a different manifestation. Content, media and carrier types are three among many attributes that distinguish between expressions and between manifestations. But they are especially significant and useful when looking at alternative formats. Alternative formats are resources that deliver the same content. Thus attributes such as author, title of the work, genre, etc., will be the same. Among the attributes that will differ, content, media and carrier types allow the user to find and select a version that they can use. If the user has difficulties with one of their senses, such as sight, then the user may be searching for a form of expression that uses hearing or touch. The difference in content type becomes of critical importance. If the user has access to a limited range of media options, then the media and carrier types become of critical importance.

In terms of content, the user needs to know the relationship of the resource to the original work. Thus, if the user searches for "Robinson Crusoe", the user needs to be able to grasp quickly the relationships between the resources in the result set. They need to be able to distinguish between the work, Defoe's Robinson Crusoe, and works that are related to it, but are different works. If the work is one that is embodied in many manifestations, such as Robinson Crusoe, then the user needs to be able to navigate through resources that are part of the same work family. It is here that we find alternative formats. Instead of facing a random accumulation of manifestations, a user should be able to navigate immediately to a form of expression that is accessible to them. Content type is an expression level attribute that permits the identification and selection of resources according to the form of expression. The media and carrier types, manifestation level attributes, may also be important for identification and selection, if the user needs a particular form of expression, such as spoken word, and also needs a particular carrier type, such as audiocassette.

RDA's solution for alternative formats is to move away from the content versus carrier issue to a new approach that respects both the content and the carrier, and gives scope for a full description of both aspects. The close mapping between FRBR and RDA means that RDA descriptions will record attributes of all the group 1 entities, permitting all levels of similarities and differences to be recorded.

FRBR's modelling of the group 1 entities also provides an answer to the problem of describing resources with multiple characteristics. AACR2 was not designed to support the description of resources with multiple, equally predominant characteristics, and it did not adapt well when the need arose. Rule 0.24 of AACR2 assumes that the cataloguer will determine one predominant "physical form" and then use the chapter that corresponds to the class of material to which the resource belongs. Part 1 of AACR2 consists of chapters organized according to the different classes of material. The categorization is flawed because the differences between the classes are not consistent; the classes of material represent different levels of generality, some are content types, some are carrier types. The GMDs are also logically inconsistent categories, reflecting content, expression or carrier; in addition, one must select a single GMD. Even with the revision of rule 0.24, it remained difficult to bring out multiple characteristics because there was no indication of precedence when following rules from different chapters. And one still had to select a single GMD.

When the AACR2 classes of material and GMDs are examined from a FRBR perspective, one problem is immediately evident: the classes of material and the GMDs are inconsistent because the categories map to different entities. In both cases, the differences between the categories are not differences at the same level of abstraction. Part 1 of AACR2 is riddled with logical inconsistency. RDA abandons the "class of material" organization used in AACR2 and bases its organizational structure on the FRBR conceptual model. RDA shifts to the principle of having general instructions that apply to all types of resources, followed, where needed, by supplementary instructions for specific types of resources.102 The possibility of conflicting instructions is further eliminated by the categorization of attributes according to the four group 1 entities. Each entity has its own logical attributes.

RDA's solution is to move away from the need to determine a predominant aspect. Instead, RDA opens up the possibility of describing a simple resource or a complex resource equally well. The description of the resource will include all relevant attributes and relationships. The cataloguer will include attributes at work, expression, manifestation and item level. All aspects of the resource can be recorded.

Using RDA, one can record more than one content, media and carrier type. This permits the full description of a resource consisting of many carriers, or combined content types. It also permits the recording of data about content, media and carrier type for new resources before the community has decided on terminology. The types, included in the lists for content, media and carrier, act as a framework so that one can record data in new combinations as required by new resources.

RDA brings a new approach to the description of content, media and carrier. RDA moves away from the AACR2 restrictions and limitations in dealing with content and carrier. RDA includes an extensible framework for content and technical description, a framework that rests on a rigorous and logically consistent conceptual model. With the FRBR model as the theoretical foundation, RDA offers a way out of the multiple formats impasse. It enables the recording of all aspects of content and carrier, and it improves the collocation of resources, with more precise definition of the similarities and differences between these resources.

8.2 Improved descriptions and improved access

The impact of RDA also extends to other areas beyond the multiple formats issue. Alignment with the conceptual model began a process of rethinking the organization of the cataloguing standard, and also of rethinking the cataloguing process. FRBR looks at bibliographic data from the user's perspective. FRBR changes the focus of the cataloguing process. The focus is no longer on the cataloguer creating a single record, but on the user seeking the record within the context of a large catalogue or database. Both activities continue to co-exist, but the defining viewpoint has changed.

The process of incorporating the FRBR model and the FRBR perspective on user tasks has resulted in a standard that aims to facilitate user access. This focus on the user appears in the functional objectives of each section of RDA but it is also evident throughout the standard, in the structure and in the instructions.

A user may approach a search in many different ways depending on their information need. A user will not necessarily always begin by identifying a work and then selecting the appropriate expression, and then the appropriate manifestation. To facilitate searching, RDA uses discrete data elements to record data, and identifies each data element uniquely and unambiguously. RDA moves away from long strings of data, especially from strings where information pertaining to different entities might be combined. The segmentation of data into discrete and unambiguous elements enables RDA metadata to be used in more versatile and flexible ways, for data retrieval, navigation and display.

Form of expression can be a vitally important characteristic for a user with a print disability. This attribute is an important part of the framework for technical and content description, with its prominent position as a content type. The content types reflect both the fundamental form of communication and the human sense through which it is intended to be perceived (6.10.1.1). Where necessary to maintain precision and to cover all possible content types, the list also includes some combined terms, such as text and tactile text, cartographic image and cartographic tactile image.

The possibility of describing all aspects of a resource is not limited to the content, media and carrier types. If a data element applies to the resource being described, then one can use it. By using separate data elements, any data element can also potentially be used to search and navigate. RDA has separate data elements for many attributes that can be helpful for the identification and selection of accessible resources for users with print disabilities, attributes such as font size, form of tactile notation, and encoding format.

Relationships between the entities play an important role in improving collocation and navigation. RDA has four sections devoted to the recording of attributes and six sections devoted to the recording of relationships. Not only does RDA encourage the recording of relationships, it also introduces additional means to improve the precision of information about the relationships. AACR2 had an option to add a designation of function to the access point for a person, and it offered a small list from which to choose. In contrast, RDA encourages the use of relationship designators and has developed extensive lists that are included in the appendices. Recording the relationships is the first step, but the ability to use information requires that the precise nature of the relationship is also recorded. In a card environment, a user was expected to read or infer about the relationship. In an online environment, there is the potential to develop new ways to search, navigate and display data but such improvements are predicated on the availability of unambiguous and consistent data about the nature of the relationship.

In the instructions on the construction of access points, and preferred access points, RDA addresses preferred access points for works, and it also addresses preferred access points for expressions. The preferred access points for expressions extend the preferred access point for the work, with the addition of an element identifying the expression. The first listed element is content type. Thus, not only can one record content type in the description, one can also use content type as part of the preferred access point for the expression. This brings the content type to prominent visibility and supports the collocation of expressions. A number of alternative formats are resources that differ in the form of expression, such as an audiobook of Hamlet or a tactile text of Robinson Crusoe. The possibility of creating a precise access point for the expression enables the user to identify and select the appropriate resource with greater precision and speed.

RDA has also opened up the option to capture metadata and re-use it as is. Thus one can harvest embedded metadata, download or automatically generate metadata, re-use what is there and enrich it with additional descriptive elements or access points, instead of having to start from scratch.

RDA has been developed as a metadata standard for the digital world, and therefore able to take advantage of new developments in data capture, storage, retrieval and display. RDA was designed so that it would not be tied to any one encoding standard. Work is underway to collaborate with encoding schema communities to ensure that RDA data can be fully encoded and supported when the standard is implemented.

It is important to remember that RDA is a content standard. It promotes the recording of well-formed metadata. RDA itself is silent on the encoding or display of the data. It creates the conditions for improved resource discovery and data display by supplying good metadata to support these tasks. It will provide some improvement using current encoding schema and current functionality for searching and data display. But to fully realize the benefits of recording data according to RDA, we should begin to envision a new generation of search engines and user interfaces that will thoroughly utilize all the RDA data elements.

It is now over ten years since the International Conference on the Principles and Future Development of AACR, and twenty years since the Multiple Versions Forum. The problems that were identified at these two events were fundamental and went right to the structure of AACR2. With the development of new types of resources, including resources that brought together content and media in new ways, and the growing number of works available in alternative formats, it became increasingly imperative to resolve the multiple formats issue.

The source of problems when looking at the multiple formats issue can be summarized by the tension between content and carrier: which one should be the defining factor for the description of a resource and access to it? RDA achieves the resolution of the multiple formats issue by moving away from this either/or question. RDA's answer is to affirm the importance and role of both the content and the carrier. RDA provides a solution to the multiple formats issue through its new approach to content, media and carrier. The new approach is based on the FRBR modelling of the group 1 entities.

The strength of RDA is that is built on the theoretical framework expressed in the FRBR conceptual model. Thus, RDA approaches description and access with a logically consistent framework underpinning it. RDA improves the description of resources and access to them, with its carefully defined data elements that record attributes and relationships. This improvement affects all resources. RDA adopts FRBR's focus on the user, and its instructions are given within the context of recording data in order to ensure that the user will find, identify, select and obtain the resource that meets his or her need.


99. Svenonius, Intellectual Foundation of Information Organization, 8.

100. Seymour Lubetzky. "Some observations on revision of the cataloguing code." In Seymour Lubetzky: writings on the classical art of cataloguing. (Englewood, Colorado: Librarie Unlimited, 2001): 184.

101. Svenonius, Intellectual Foundation of Information Organization, 11.

102. Joint Steering Committee for Development of RDA. RDA, Resource Description and Access: Objectives and Principles. Draft version. (5JSC/RDA/Objectives and Principles/Rev/2, 28 October 2008): 2. www.collectionscanada.gc.ca/jsc/docs/5RDA-objectivesrev2.pdf

Previous | Table of Contents | Next