FRBR and Fundamental Cataloguing Rules

Last modified: 22 January 2022

Do not use this old essay. Please read this book chapter instead.

I wrote this in library school in 2003. In 2007 I was able to turn it into something much better: a chapter in Understanding FRBR: What It Is and How It Will Affect Our Retrieval Tools, edited by Arlene Taylor (ISBN 9781591585091).

The chapter is available at York University’s institutional repository: FRBR and the History of Cataloging (211 KB PDF). Please read it instead of what’s below.

Denton, William. “FRBR and Fundamental Cataloguing Rules.” May 2003. https://www.miskatonic.org/library/frbr.html.

This is another essay I wrote in the spring of 2003 for Clare Beghtol’s course on the theory of classification and knowledge organization, at what was then the Faculty of Information Studies at the University of Toronto. It was a fascinating course. We didn’t talk about FRBR but I had become interested in it after reading a mention in American Libraries (reference below), and it tied in to topic maps and ontologies and all that, so I made it the subject of my final essay, this twenty-two-pager. I’m glad I did. It’s a great topic, and reading Lubetzky was a treat. (He died, aged 103, while I was researching this.)

I had three diagrams in the table, lifted from the FRBR spec, but here you’ll have to look at the PDF yourself. There are links.

Whoever embarks on a study of the development of our cataloging rules, from their spectacular trial before a royal commission and their brilliant defense by Panizzi in 1849 to the appearance of the A.L.A. Cataloging Rules for Author and Title Entries in 1949, a round century later, cannot fail to be impressed with the broad knowledge, keen thinking, and fruitful imagination which the founders of the rules have brought to the profession of cataloging. At the same time, one could hardly view with equanimity the continuous proliferation of the rules, their growing complexity, and the obscurement of the objectives and design of the code as a whole. One is impelled to ask: Are all these rules necessary? Are all the complexities inevitable? Is there an underlying design which gives our code unity and purpose?

— Seymour Lubetzky, Cataloging Rules and Principles (83)

Abstract: IFLA’s proposed data model, Functional Requirements for Bibliographic Records (FRBR), arranges bibliographic entities in a new way, using an entity-relationship model that unites the products of intellectual and artistic endeavour, their creators, and their subjects. A catalogue that uses FRBR principles would give the user more helpful information and power over the collection than current catalogues do. This essay looks back at the fundamental laws and objectives of cataloguing and librarianship — as set down by Panizzi, Cutter, Lubetzky, and Ranganathan — and shows how FRBR embodies them. The FRBR entity-relationship model is described, and a possible web-based FRBR-aware catalogue imagined. Current work on FRBR, including experiments done with WorldCat, is examined. FRBR offers great hope for the future of catalogues.

0. Looking for a book

Brigid wanted to read Raymond Chandler's The Big Sleep. She remembered seeing the movie a few years ago, the one with Humphrey Bogart, and when some friends talked about the book the other day and she decided it was time to finally read it. She went to the library, but their copy was not on the shelf. Someone had already borrowed it. Interlibrary loan within the city was very fast, so she went to the computer, pulled up the online catalogue, searched for the title "big sleep" ... and got a list ten books, all named The Big Sleep or some slight variation. The top book on the list was by Chandler and had 73 copies in the system. Brigid choose it from the menu ... and got a list of 46 libraries, all holding at least one copy of the book. Her local library was there, but when she chose it from this new menu it confirmed what she knew: their one copy was already checked out. She went back to the previous list, of all the libraries. The Alexandria Street branch had three copies: a 1965 hardcover was checked out, but a hardcover from 1973 and a paperback from 2000 were there. Brigid didn't care which one she got, it made absolutely no difference, but she picked the hardcover at random and placed an order for it. Three days later the book arrived and she enjoyed it immensely.

Library catalogues are supposed to make things easy, but this turned out to be a little annoying. Brigid just wanted The Big Sleep, any copy (as long as it was in English, and not large print — just a regular edition of the original book, nothing special), but the system did not let her find the book that way. It made her pick a library, and from that library's collection, choose a particular copy. Some librarian in that other library will see the order, find the book's call number, go to a precise location on a shelf, pull down the exact copy she ordered, and send it to her library. The catalog forced her to make a direct connection to a particular copy of the book in another library: the 1973 hardcover owned by the Alexandria St. branch. But she did not care which library she got the book from, and she certainly did not care which copy they take from their shelves. She just wanted to read the book!

1. Introduction to FRBR

There is a vast body of thought, starting in earnest 150 years ago, about how to organize and catalogue books so they are easy to find. Through the works of the major figures in the field — all practitioners and theoreticians both — there are common, basic ideas. A recent development are the Functional Requirements for Bibliographic Records (FRBR), a proposed data model put out by the International Federation of Library Associations and Institutions (IFLA). FRBR is a new way of defining relationships between books, their creators, and their subjects. The details of how it works are new, and it can only be done with computers, but the ideas it uses are old. FRBR would do away with all those problems finding a copy of The Big Sleep, but it also makes it easy for someone who does care about minor details to get what they need. It embodies the fundamental laws of cataloguing and librarianship. This essay will examine some of those laws, explain how FRBR works and how it obeys the laws, and look at some current work on enriching existing catalogues with FRBR.

Before going back into history, let us first take a brief look at FRBR. Its core idea is the "work," which is defined as "a distinct intellectual or artistic creation" (IFLA 16). Examples given are Henry Gray's Anatomy of the Human Body, Bach's The Art of the Fugue, and Franco Zeffirelli's film Romeo and Juliet. "A work is an abstract entity; there is no single material object one can point to as the work. We recognize the work through individual realizations or expressions of the work, but the work itself only exists in the commonality of content between and among the various expressions of the work." (Italics are used by IFLA to highlight entity names.)

The work is at the top of a hierarchy. Immediately below it is the "expression," defined as "the intellectual realization of a work in the form of alpha-numeric, musical, or choreographic notation, sound, image, object, movement, etc., or in any combination of such forms" (IFLA 18). Expressions are also abstract. "An expression is the specific intellectual or artistic form that a work takes each time it is 'realized.' Expression encompasses, for example, the specific words, sentences, paragraphs, etc., that result from the realization of a work in the form of a text, or the particular notes, phrasing, etc., resulting from the realization of a musical work." Physical details such as page size or recording do not matter here. Given a work such as Schubert's Trout Quintet, example expressions include the composer's score, a particular performance by the Amadeus Quartet and Hephzibah Menuhin on piano, and a particular performance by the Cleveland Quartet with Yo-Yo Ma on the cello.

The third element in the hierarchy is the manifestation, and it is concrete. It is "the physical embodiment of an expression of a work. The entity defined as manifestation encompasses a wide range of materials, including manuscripts, books, periodicals, maps, posters, sound recordings, films, video recordings, CD-ROMs, multimedia kits, etc. As an entity, manifestation represents all the physical objects that bear the same characteristics, in respect to both intellectual content and physical form" (IFLA 20). An example given is for the work Bach's Six Suites for Unaccompanied Cello. One of the many expressions is the performance by Yo-Yo Ma recorded in 1983. One manifestation of that is the recordings released by CBS Records on LP in 1983; another manifestation is the CD re-release in 1992

The fourth and last layer in the hierarchy can be held in one's hands: the item, "a single exemplar of a manifestation" (IFLA 23). For books, each copy of the edition that makes a manifestation is an item. For music, each copy of that CD re-release of Yo-Yo Ma doing the cello suites is an item. Items can vary from each other "where those variations are the result of actions external to the intent of the producer of the manifestation (e.g., damage occurring after the item was produced, binding performed by a library, etc.)."

Figure 1 [not available, see p. 13, FRBR PDF] illustrates how all the entities relate. The single arrow from expression to work means that each expression is the result of only one work. The double arrow from work to expression means that one work can have many expressions. Similar relationships hold for the other entities, resulting in a sort of pyramid, with one work at the top and many items at the base. Works can also be related to other works: Shakespeare's Romeo and Juliet and its movie adaptations are all works of their own.

When Brigid went looking for The Big Sleep, she wanted a particular expression: the author's edited text for publication. She did not want such manifestations as a large print edition or an audio book, but any "normal" print manifestation would have done. The catalogue did not let her browse that way, however. It showed her the digital equivalent of a stack of catalogue cards, with no help about how they relate to each other. It forced her to choose a particular item, the lowest layer of the hierarchy with the most precision and level of detail, well beyond her level of interest. She could not choose a manifestation, even though such information is latent in the catalogue. Brigid was forced to become an expert, distinguishing items, when to her even what made manifestations different from each other was mere trivia.

A catalogue that used FRBR would have made Brigid's search much easier: she would have started by finding the work, then looked at a list of expressions of the work (in this case, there is probably just one, so this step could be skipped), and finally, from a list of all manifestations in the system, been able to indicate that any of them (except large print and audio books) would be fine. The system would have located a particular item on its own and arranged for it to be shipped.

Although the catalogue made Brigid do a bit of extra work, nevertheless she did get what she wanted just by doing a search and looking at a few computer screens. Catalogues do work very well as they are now. They are the result of centuries of hard work, research, and thought, guided by some fundamental principles of librarianship. How did the catalogue become what it is now? What are the principles behind it? How does history lead up to FRBR, and how does FRBR continue it?

2. A brief history of cataloguing

The history of cataloguing can be summed up (briefly and rather rudely, leaving out a few important people and publications) in a short chain of people and rule sets: Sir Anthony Panizzi's "Rules for the Compilation of the Catalogue" (1841), Charles Jewett's Smithsonian Report on the Construction of Catalogs of Libraries, and Their Publication by Means of Separate, Stereotyped Titles, with Rules and Examples (1853), Charles Cutter's Rules for a Printed Dictionary Catalog (1876), the Prussian Instructions (1908), the American Library Association's Catalog Rules: Author and Title Entries (1908) and ALA Cataloguing Rules for Author and Title Entries (1949) (the latter closely criticized by Seymour Lubetzky), the Paris Principles (1961), and Anglo-American Cataloguing Rules (1967), which is used internationally and has been revised several times (Chan 33-41, Lehnus 2-3). Donald J. Lehnus explained, "Each successive code has brought certain refinements, clarifications, simplifications, and in general, many improvements. Many major changes have taken place, but at the same time there are several areas where little or no change has taken place, and some have even reverted to rules used earlier" (Lehnus 3).

Most of these cataloguing milestones consist of extremely detailed rules: pages and pages on how to determine the actual title of a book or how to arrange names in alphabetical order. All possible situations need to be covered so that any book, periodical, recording, movie, map, or object of any sort, can be described, put in a catalogue, and found there. Classification by systems such as the Library of Congress or Dewey Decimal Classification schemes will of course help, as will subject headings that tell what the thing is about, but these are beyond our range of discussion.

3. Sir Anthony Panizzi and his 91 rules

Sir Anthony Panizzi (1787-1879) was a pioneering English librarian whose 91 rules for the British Museum's catalogue, "Rules for the Compilation of the Catalogue", are "considered to be the first major cataloguing code" (Chan 34). There is not room to go into all of Panizzi's work or the Royal Commission that examined it, but a statement Panizzi made to the Commission is perfectly valid 160 years later: "A reader may know the work he requires; he cannot be expected to know all the peculiarities of different editions; and this information he has a right to expect from the catalogues" (Freedman 93). Our reader Brigid would certainly agree. On a larger scale, Panizzi's work "remains generally valid and relevant to the present discussion of the question of whether the catalog should be a 'finding list' or a 'reference tool.' For in this argument Panizzi has laid the cornerstone for the modern catalog" (Lubetzky 1956 175).

Maurice Freedman (104) summarized Panizzi's core ideas about the purpose of a catalogue:

1. To relate the works of an author so that the user can know all of that author's works.

2. To identify and distinguish particular editions, translations, etc., so that those different editions, translations, etc., of a given work are not confused with each other.

3. To assemble all of the editions, etc., of a work so that a user seeking a given publication will not just find it, but also will have presented with that given publication, all of the editions, etc., of the given work represented by it, as well as works related to it.

Some of that terminology looks familiar. Let us restate it using words from FRBR:

To relate the works of an author so that the user can know all of that author's works.
To identify and distinguish particular expressions and manifestations of a given work so they are not confused with each other.
To assemble all works and their expressions and manifestations so that a user seeking any one will see all the other related entities in the hierarchy, and other works related to the work in question.

We will see that these ideas continue on, almost unchanged, to today.

4. Charles Cutter and his rules

Charles Cutter (1837-1903) was another pioneering librarian. One part of his large body of work was Rules for a Printed Dictionary Catalog (1876, with several revisions), which "became the basis for the dictionary catalog, which was to become the predominant form of catalogs in general libraries in the United States" (Chan 34). Cutter (8) elegantly set out what a catalogue should do, and how:

			       OBJECTS
1. To enable a person to find a book of which either
	A) the author	}
	B) the title	} is known
	C) the subject	}
2. To show what a library has
	D) by a given author
	E) on a given subject
	F) in a given kind of literature
3. To assist in the choice of a book
	H) as to its edition (bibliographically)
	I) as to its character (literary or topical)
				MEANS
1. Author-entry with the necessary references (for A and D).
2. Title-entry or title-reference (for B).
3. Subject-entry, cross-references, and classed subject table (for C and E).
4. Form-entry (for F).
5. Giving edition and imprint, with notes when necessary (for G).
6. Notes (for H).
			  REASONS FOR CHOICE
among the several possible methods of attaining the OBJECTS.
   Other things being equal, choose that entry
   (1) That will probably be first looked under by the class of
       people who use the library;
   (2) That is consistent with other entries, so that one
       principle can cover all;
   (3) That will mass entries least in places where it is difficult to
       so arrange them that they can be readily found, as under names of
       nations and cities.

These could be reworded with FRBR terms, but there are some distinctions that the next cataloguer will clarify. We will leave Panizzi and Cutter, noting that their rules underly all modern catalogues. To the newcomer their ideas seem so natural that they do not describe how a catalogue should work, but rather how it does work, and how all catalogues work, and one that did not would be a poor and strange one indeed.

5. Seymour Lubetzky and "the work"

Seymour Lubetzky (1898-2003) was another giant in the world of cataloguing. He worked at UCLA, then spent seventeen years at the Library of Congress. While there, he wrote important criticisms of existing practices for descriptive cataloguing and author and title entries. Cataloguing rules — the finicky details not discussed here — had become a quagmire of complicated, sometimes contradictory, often confusing rules, made up over the years to patch over problems as they appeared. Lubetzky worked to simplify the rules, to build them rigorously from fundamental principles, and he had a great influence on the Anglo-American Cataloguing Rules (Svenonius and McGarry xii-xx). Lubetzky is important in this discussion because he drew from Panizzi and reworked Cutter, and his ideas directly influenced FRBR. Compare the above purposes of catalogues with Lubetzky's in this introduction to his discussion of the function of the main entry:

The question of the function of the main entry presupposes recognition of the facts (a) that the materials of a library — books, manuscripts, phono-records, etc. — are representations of the works of authors, not the works themselves; (b) that a given work may be represented in a library in different forms or editions, under different names of the author or under different titles; and (c) that the catalogue of a library must be designed not only (1) to show whether or not the library has a particular item or publication, issued under a certain name of the author or under a certain title, but also (2) to identify the author and the work represented by the item or publication, and to relate the various works of the author and the various editions and translations of the work. Although this second objective is the source of most difficulties in cataloguing, it has always been recognized as essential to the basic purposes of the catalogue — to enable a user of the catalogue (member of the library staff or reader) to determine with certainty whether or not the library has a particular work, under whatever name or title, and to select the edition or translation which will best serve his purpose. (Lubetzky 1961 231)

So the purposes of a catalogue are:

(1) to show whether or not the library has a particular item or publication, issued under a certain name of the author or under a certain title

(2) to identify the author and the work represented by the item or publication, and to relate the various works of the author and the various editions and translations of the work.

Item and work — FRBR terms. A few years later, Lubetzky wrote more on the objectives of a catalogue:

The book, it should be noted, comes into being as a dichotomic product — as a material object or medium used to convey the intellectual work of an author. Because the material book embodies and represents the intellectual work, the two have come to be confused, and the terms are synonymously used not only by the layman but also by the cataloger himself. Thus catalogers refer to the author and title of a book instead of, more accurately, to the author of the work and the title of the book embodying it, and the inquirer searching the catalog for a particular book is more often than not after the work embodied in it, although he is very likely unaware of the distinction between the two.... The question that must then be faced at the outset — and that has been faced since Panizzi, though beclouded by the failure to distinguish clearly and consistently between the book and the work — is whether the objective of the catalog should be merely to tell an inquirer whether or not the library has the particular book he is looking for, or whether it should go beyond that and tell him also what other editions and translations — or other representations — of the work the library has so as to help him more effectively determine whether the library has what he needs and to select what might best serve his purposes. (Lubetzky 1969 270-271)

Later in the same essay, Lubetzky compares Cutter's "objects" with the function of the catalogue as defined in the 1961 Paris Principles, which "drew heavily" (Chan 40) on Lubetzky's work :

The catalogue should be an efficient instrument for ascertaining

2.1 whether the library contains a particular book specified by
    (a)  its author and title, or
    (b)  if the author is not named in the book, its title alone, or
    (c)  if author and title are inappropriate or insufficient
         for identification, a suitable substitute for the title; and
2.2 (a) which works by a particular author and
    (b) which editions of a particular work are in the library
                 (International Conference on Cataloguing Principles 26)

Lubetzky comments:

The second objective, however, is significantly different in specifying the editions of a work and the works of an author for Cutter's vague what the library has by a given author. Cutter's unqualified what is expressive of the failure to distinguish clearly and consistently between the book and the work in his rules, and characterizes also the old Anglo-American rules which were based on them. (1969 273)

This new idea of what a catalogue should do, how it should work, and how it should arrange and relate all the bibliographic information it contains, fulfills Cutter's rules and then extends them, making the catalogue a much more powerful tool. Another quote from Lubetzky, the concluding "Epitome" from the same essay, will put this best:

The "information" found in the materials of a library is not in the form of nuggets that can be collected, sorted, and labeled in isolation. Rather, it is part of the collective work and thought of men, and of the fabric of the particular works in which it is found. It is pertinent, therefore, that in organizing this information cognizance should be taken of the authors who created it, the particular works of which it is an organic part, and the materials embodying these works. Together, the authors, the works, and the materials, may be said to constitute the determinant dimensions of the information found in bibliographical sources. (Lubetzky 1969 334)

The chain of Panizzi — Cutter — Lubetzky shows how the purpose of the catalogue has been understood, adapted, and improved over the last 150 years. All three based their thought on printed or card catalogues, though Lubetzky's later work did considered the computer and he lived to see the growth of the World Wide Web and instant online access to library catalogues around the world — none of which fully employ his ideas. The means by which the objectives of the catalogue are met are not as important as their fulfillment, but a complete arrangement as imagined by Lubetzky is practically impossible to do with a typewriter and paper cards. Computers and databases make the management of the information and relationships easy, once a model has been made and rules set out for how works, creators, and subject matter will relate. FRBR does that.

6. What is a work?

Lubetzky's discussion of works inspired many others, and FRBR's definition of a work is only one among many. Richard Smiraglia's The Nature of "A Work": Implications for the Organization of Knowledge (2001) looks at all the different treatments the idea has received. Where FRBR has work-expression-manifestation-item, others have work-version-adaptation, work-text-edition printing-book, work-document-text, or work-derivations-item (51, 145). Smiraglia gives his own definition: "A work is a signifying, concrete set of ideational conceptions that finds realization through semantic or symbolic expression" (129) which is on the surface more philosophically profound than "a distinct intellectual or artistic creation" but probably needlessly complicated. Smiraglia adds an appendix that compares various definitions of work, document, and text, from Panizzi, Cutter, and Lubetzky, through contemporary philosophers and librarians, up to FRBR.

Elaine Svenonius, in The Intellectual Foundations of Information Organization (2000), uses set theory in her definitions. The most basic entity is the document, "information-bearing messages in recorded form," such as a copy of a book. A work W_i is "the set of all documents that are copies of (equivalent to) a particular document a_w (an individual document chosen as emblematic of the work, normally its first instance) or related to this individual by revision, update, abridgment, enlargement, or translation." In mathematical notation, this is W_i = def {x: x is a copy of a_{w_i} or x is a revision, update, abridgment, enlargement, or translation of a_{w_i}}. Her entire hierarchy is work-edition-subedition-version-document (34-43).

There are differences in all of these definitions — Smiraglia identifies the biggest problem as "the degree to which change in ideational or semantic content represents a new work" (52) — but they are beyond the scope of this essay. In all of them, the idea of a work is more or less similar to FRBR's, and the other levels of the hierarchy can be mapped to FRBR's. We concentrate on how FRBR works and how it obeys basic cataloguing principles. Indeed, we shall see later that it is not the work entity that could be a problem for FRBR, but the expression.

7. Ranganathan's five laws of library science

Before the more detailed explanation of FRBR, however, there is another, more fundamental, more abstract, and more profound, set of rules to consider. S.R. Ranganathan (1892-1972) was an Indian librarian who delved into every possible aspect of librarianship, from the theory of classification to what duties must be performed before closing up a library for the night. His body of work, and influence on the field, is enormous. One of his most important books, setting out core beliefs that helped guide his life and work, is The Five Laws of Library Science, published early in his career in 1931. The five laws are (Ranganathan 9):

Books are for use.

Every reader his book.

Every book its reader.

Save the time of the reader.

Library is a growing organism.

Panizzi's, Cutter's, and Lubetzky's cataloguing rules above are detailed and concrete in comparison. Ranganathan's seem vague, but only because they are so axiomatic and Delphic. Ranganathan expounds on them in inspired prose, examining each, showing how common library practices of the time violated the laws, and building on the laws alone and in combination to show what libraries should do and how they can do it, from lofty ideals to experienced, practical advice.

Ranganathan uses the word "book," not "work." We do not need to worry about this. He wanted users to have complete mastery over the collection, and FRBR-style work/manifestation distinctions would help. Remember how much work Brigid had to do: a better catalogue design would save the time of the reader and help a reader find her book. Where he says "book" we can read "work" or "item" or whatever else, as need be. The rules hold for all entities and user demands.

There is not much discussion of catalogues in The Five Laws of Library Science, but there is a nice example given of their importance in the discussion of "every book its reader":

A scholar asked one day for material on migration of population. He was taken to the shelf containing the books on the subject. There were at most two dozen volumes. He turned them through and a little later he was helped to examine the catalogue cards relating to that class — particularly the red cross-reference cards. But, his attention was arrested by couple of titles on the white main cards themselves and the eagerly said, "Where are these two books? I did not find them on the shelf." But the books were actually there.... They were thin unprepossessing pamphlets published by the National Research Council. After looking into them he found them so interesting that he felt thankful that he was taken to the catalogue, as he would have otherwise missed those two useful publications on account of their thinness and unimposing appearance. (Ranganathan 264-265)

This is a good example of how helpful and informative the catalogue can be compared to browsing the shelves. The shelves hold books: the catalogue holds information about the books. The richer the relationships made in the catalogue between books, works, authors, and subjects, the more useful it will be.

8. FRBR in detail

An introduction to part of FRBR was made in §1. Now that we have seen how a catalogue should work, and how it should serve users, we will take a closer look at FRBR. We will first see how it works, then how it fulfills the goals of a catalogue.

The work-expression-manifestation-item hierarchy is actually only one of three sets of entities described in the FRBR specification. To summarize, "The entities defined as work (a distinct intellectual or artistic creation) and expression (the intellectual or artistic realization of a work) reflect intellectual or artistic content. The entities defined as manifestation (the physical embodiment of an expression of a work) and item (a single exemplar of a manifestation), on the other hand, reflect physical form" (IFLA 12).

The second group of entities are persons and corporate bodies, "responsible for the intellectual or artistic content, the physical production and dissemination, or the custodianship of the entities in the first group" (IFLA 13) or the subject of a work. A person is defined as encompassing "individuals that are deceased as well as those that are living" (IFLA 23). Corporate bodies encompass "organizations and groups of individuals and/or organizations that are identified by a particular name, including ... conferences, congresses, expeditions, exhibitions, festivals, fairs," governments, and more, either defunct or still operating (IFLA 24). Figure 2 [not available, see p. 14, FRBR PDF] shows how these entities relate to the first set with "responsibility" relationships. We see that one or more works can be created by one or more persons and/or corporate bodies, one or more expressions may be realized by one or more persons and/or corporate bodies, and so on.

The third set of entities represent subjects of works. There are four: concept, object, event, and place. These are all defined very broadly. A concept "encompasses a comprehensive range of abstractions that may be the subject of a work: fields of knowledge, disciplines, schools of thought (philosophies, religions, political ideologies, etc.), theories, processes, techniques, practices, etc." (IFLA 25). Examples include hydroponics and supply-side economics. Concepts are abstract, but objects are material things. They encompass "a comprehensive range of material things that may be the subject of a work: animate and inanimate objects occurring in nature; fixed, movable, and moving objects that are the product of human creation; objects that no longer exist" (IFLA 26). Events are "actions or occurrences that may be the subject of a work: historical events, epochs, periods of time, etc." (IFLA 27). Events need not be brief or accurately defined: examples include the Age of Enlightenment. The tenth and last entity is location, which "encompasses a comprehensive range of locations: terrestrial and extra-terrestrial; historical and contemporary; geographic features and geo-political jurisdictions" (IFLA 27) such as Toronto, Baffin Island, or the Andromeda Galaxy. Figure 3 [not available, see p. 15, FRBR PDF] shows how these last four entities relate to the earlier ones.

We can now imagine a better web-based FRBR-aware catalogue. Brigid would search for a work called "the big sleep" and be shown a list of the 1939 book and its two movie adaptations, from 1946 and 1978. The author of the book is a person, Raymond Chandler. She could choose his name, a hyperlink, and see a list of the works he created. Chandler is also a subject: there are several biographies about him, and he would be included in any book looking at American (location) mysteries (concept) or hardboiled fiction (concept) or twentieth century (event) writing (concept), as would Dashiell Hammett. Hammett is the creator of many works (e.g. The Maltese Falcon) some of which became related works (movies or radio or television series), and he is also the subject of books and movies (e.g. Julia). If Brigid did not become distracted by Hammett, she could choose from the many manifestations of The Big Sleep. The first was published by Knopf, and she could browse what other books Knopf published, and books written about the company. All the entities — works, expressions, subjects, locations, concepts — weave into an intricately linked, overlapping web of relationships.

All of the FRBR entities have attributes that give detailed information about them. A work has these:

title of the work

form of the work

date of the work

other distinguishing work

intended termination

intended audience

context for the work

medium of performance; numeric designation; key (musical work)

coordinates; equinox (cartographic work) (IFLA 33-33)

Expressions, manifestations, and items have many more attributes, e.g. "revisability of expression" and "access restrictions on the manifestation." Persons have names, dates, titles and "other designations" (IFLA 51), and corporate bodies have similar attributes. Concepts, objects, locations, and events only have one attribute: the term describing them (IFLA 53-54). All these are subject to authority control. Subjects and persons and corporate bodies do exist in catalogues now, but without FRBR's structured relationships.

The FRBR specification goes into great detail explaining all the possible relationships between all the entities. They discuss work-to-work relationships such as translations, musical settings for texts, and parodies; whole-part relationships where one work is made up of many other works; how abridgments, expurgations, and revisions relate expressions; manifestation-item relationships when something is bound with something else, or split up, or reproduced; etc. It is an exhaustive discussion.

9. How well does FRBR embody fundamental principles?

Now that a more detailed description of FRBR has been given, and we have in mind how a FRBR-aware web-based catalogue could work, we can consider how FRBR helps fulfill Panizzi's, Cutter's, Lubetzky's, and Ranganathan's rules. The FRBR specification lists four simple tasks that users will perform, which helped guide their arrangement of entities and relationships:

* to find entities that correspond to the user's stated search criteria (i.e., to locate either a single entity or a set of entities in a file or database as the result of a search using an attribute or relationship of the entity);

* to identify an entity (i.e. to confirm that the entity described corresponds to the entity sought, or to distinguish between two or more entities with similar characteristics);

* to select an entity that is appropriate to the user's needs (i.e., to choose an entity that meets the user's requirements with respect to content, physical format, etc., or to reject an entity as being inappropriate to the user's needs);

* to acquire or obtain access to the entity described (i.e., to acquire an entity through purchase, loan, etc., or to access an entity electronically through an online connection to a remote computer). (IFLA 82)

Brigid's ideal search for a copy of The Big Sleep follows this exactly: she searches for a work entity, distinguishes the book from the movies, casts out large-print or audio book versions, then places an interlibrary loan for any of the remaining manifestations.

How well does FRBR embody Panizzi's cataloguing functions as given in §3? Perfectly. The work-author relationships will let a user see all works by an author. All translations, editions, and so on (expressions and manifestations) are distinguished and clearly separate. All the expressions and manifestations of a work are grouped under it so a user seeking one in particular will see all the others. Related works are connected.

FRBR is far richer than Cutter's rules (see §4) demand. A person can certainly find a book (manifestation or item) knowing the author, title, or subject. A user can see the library's holdings by author, subject, or kind of literature (an attribute of the work or expression). Choosing a book based on details about the edition is easy using manifestation attributes, and choosing based on its character can be done with work or expression attributes. Cutter's means (author and title entry, etc.) are primitive compared to the rich set of attributes and relationships FRBR can define. A FRBR-aware electronic catalogue would satisfy Cutter's rules as an almost trivial side effect.

Lubetzky's principles (see §5) are completely satisfied — unsurprising, since his work inspired FRBR. Authority control and the author entity handle showing if the library has something issued under a certain name. Finding material by title is taken care of by the work entity and authority control. "To identify the author and the work represented by the item or publication and to relate the various works of the author and the various editions and translations of the work" (Lubetzky 1961 231) is precisely what FRBR is meant to do. Similarly, the Paris Principles are fulfilled. A user can find material knowing the author or title or (with authority control and a good search facility) a variation on the title. All authors are related to their works, and the catalogue will show which editions (expressions or manifestations) of a work are in the collection.

Ranganathan's five laws of library science (see §7) are also obeyed by FRBR. A FRBR catalogue will make a library's collection more open to users, by increasing the numbers of ways in which people can use the catalogue. Readers will have more help in finding the book they need, and books will be exposed to more interested readers. We have seen how the fourth law, "save the time of the reader," is obeyed. The fifth law, "library is a growing organism," holds not because FRBR can not only help librarians manage local collections far larger than decades ago, but also union catalogues containing tens of millions of records from all over the world. Libraries grow not only by adding more shelves and buildings, but by making use of new things in the world around them, for example adapting new technologies to serve their purposes, as networked databases and entity-relationship models are used here.

We can see, therefore, that FRBR observes all of the principles and objectives that cataloguing and library theoreticians have conceived.

10. Current work on FRBR

Why are there no FRBR-aware catalogues in wide use? The potential is immense, but it will take a lot of work to open up catalogues, examine their records, construct the proper work and expression entities from the existing manifestation and item entities (which are what catalogues hold now), clarify the person, subject, concept, object, event, and location entities, fix all the attributes, and then build all of the relationships. National libraries and union catalogues hold tens of millions of records, and it is impossible to do the work by hand. As well, we have assumed that authority control mechanisms were available to handle names of people, titles, events, etc. FRBR does not handle authority records, however, and there is a related project that is doing for them what FRBR is doing for bibliographic records: Functional Requirements of Authority Numbering and Records (FRANAR). FRBR and FRANAR together will, eventually, "cover the whole bibliographic universe" (Working Group on FRBR).

Three FRBR catalogues do exist, but they are not easy to see or use. The Danish software company Portia developed VisualCat, which uses FRBR, but no online sample is available. AustLit, the Australian Literature Gateway, used FRBR and topic maps when constructing its hypertext bibliography of Australian literature, which contains over 385,000 works and 60,000 persons or corporate bodies (authors and publishers) (Fitch). Their web site is, unfortunately, restricted to subscribers, but one page gives a glimpse of what is inside: one can see a view of Patrick White's Voss as a work, "showing a work with a range of Expressions (versions) and numerous Manifestations (publications). The Expressions include translations of Voss into languages other than English" (AustLit Sample Pages).

VTLS Inc., a library software company, last year debuted a new version of its integrated library system, Virtua, that uses FRBR. They have on their web site what claims to be an example catalogue that can be searched but precisely how it uses FRBR is unclear even after several tests. However, Virtua did receive an excellent review after being shown at the 2002 ALA conference:

In the VTLS demonstration of how FRBR works, I was able to place a hold on the Expression level record of Beau Geste by P.C. Wren, rather than being forced to choose among a half-dozen different editions, each of which represented exactly the same novel. Music librarians in particular think they've died and gone to music metadata heaven when they see how FRBR can organize musical works....

[FRBR] is a long-overdue and dramatic improvement in OPAC [online public access catalogue] functionality. (Dorman)

International work is going on testing how FRBR can be applied to large catalogues and identifying problems that only emerge when the model is put into practice. OCLC has been doing research by examining sets of records in their WorldCat union catalogue, and found that the expression entity is hard to identify and can interfere with the catalogue's usefulness. Hickey, O'Neill, and Toves looked at Tobias Smollett's epistolary novel Humphry Clinker. They found that the existing bibliographic records were not reliable enough to make it easy to group all the expressions, that identifying works gave the users most of what they needed, and that "below works, dynamic division of records into sets based on a particular user's needs, such as by illustrator or translator, would be more appropriate." They decided "to abandon our experiments on identification of expressions for now."

Not only is it difficult to automatically distinguish between expressions, but sometimes the differences between them are almost trivial. Most new expressions of Humphry Clinker were found to come from the addition of "material such as acknowledgments, bibliography, biographical note, chapter titles, chronological tables, dedication, glossary, illustrations, introduction and/or forward [sic], list of illustrations, maps, notes, publisher's note, table of contents, textual notes, reproduction of original title page, and reviews" (O'Neill). It is doubtful whether it is useful to create a new high-level entity for what could be a very minor change. O'Neill complained that FRBR "provides an unambiguous definition for expression but then proceeds to allow for flexible interpretations. The IFLA report does not adequately consider the impact of such flexibility in a shared cataloging environment where consistency can be more important than flexibility." His experience with WorldCat experiments leads him to question whether expressions should be used at all, and he proposes dropping them and using an expanded manifestation entity, with added entries for contributors (illustrators, editors, writers of forewords, etc.). For Humphry Clinker, "replacing the expression in the FRBR model with additional manifestation attributes simplifies the model without any loss of functionality."

Identifying works has proven to be fairly simple and very beneficial. In another OCLC experiment, Bennett, Lavoie, and O'Neill took a sample of 996 records (manifestations) in WorldCat, then used algorithms to match each one with related manifestations in order to identify the common work. They found that most works in WorldCat, 25 million out of 32 million, or 78%, have but a single manifestation. The average work has 1.5 manifestations. 99% of works in WorldCat have seven or fewer manifestations, and only 30,000 have more than 20. The 78% of the works that have only one manifestation are trivial cases. Of the remaining roughly 20%, 1% of the whole, with eight or more manifestations, is the most beneficial to work on and also the easiest. The works that are of the most interest will have the most manifestations — for example, consider the Bible, Shakespeare, Homer, Dickens, Marx, and Freud, and how many different editions of their works exist. The more manifestations that exist, the more chance there is that a piece of information missing in one record will be in another. For example, one crucial piece of information about a manifestation is its class number. If a work has one manifestation, there is a 50% chance the Library of Congress number is known, but works with nine or more manifestations are almost certain to have an LC number. The number can be copied to the other manifestations that are missing it. Bennett, Lavoie, and O'Neill conclude that computer algorithms can handle much of the job of identifying works, and that

the sample data suggests that the task of applying FRBR may not be as burdensome as a priori estimates might suggest: FRBR can be applied non-trivially to only a small percentage of works in WorldCat. At a maximum, 20% of the works would be candidates (i.e., works with two or more manifestations); in practice, however, the percentage is likely to be much lower. Analysis suggests that concentrating on relatively large works, in particular those works whose content has been augmented, revised, or consists of collections of other works (a relatively small portion of the catalog) might be sufficient to capture the lion's share of benefits potentially available from implementing FRBR.

Aside from these experiments, FRBR receives continuing attention in journals and at conferences, and a number of suggestions have been made: addition of "manage" and "navigate" tasks, addition a level about work, splitting the expression into four levels, addition an "action" group of entities, separation of subject and object relations, and more (Le Boeuf 31-32). Le Boeuf looks at all of the suggestions made regarding FRBR (as of 2001) and also discusses FRBR in relation to the goals of the catalogue, referring, as here, to Cutter and Ranganathan (33-35). There is international interest in FRBR, both as a practical data model and as an object of theoretical interest, and work continues towards making a reliable final specification and integrating it with other standards.

11. Conclusion

We have examined Panizzi's, Cutter's, and Lubetzky's cataloguing principles, Ranganathan's five laws of library science, and IFLA's proposed data model FRBR. In its current conception, FRBR perfectly embodies the long-standing fundamental objectives of a catalogue and a library service. It goes beyond merely listing details of a particular copy or edition of a book (or other bibliographic entity), instead using a work-expression-manifestation-item hierarchy of entities and relationships. The creators of these entities, and their subjects, are also entities, and also related, making a rich, informative web of information and knowledge that can be searched or browsed.

The FRBR specification seems ideal on paper. In practice it may prove to be unclear or imperfect, particularly regarding the expression entity. Converting existing catalogues with normal MARC records will be a huge undertaking, but intelligent computer analysis will simplify identification of most works, which on its own would be of great benefit. More work is needed so that problems can be found, the FRBR specification clarified, data formats standardized, and software shared. Soon, when the problems have been worked out and national libraries and WorldCat start using the FRBR model, the cataloguing world will have taken a huge step forward by using available technology to make improvements in the organization of bibliographic knowledge and in making that knowledge easily accessible to users.

Works Cited

"AustLit Sample Pages." 13 Feb. 2003. AustLit: Australian Literature Gateway. 20 Apr. 2003. http://www.austlit.edu.au/subscribe/samples.

Bennett, Rick, Brian F. Lavoie, and Edward T. O'Neill. "The Concept of a Work in WorldCat: An Application of FRBR." Library Collections, Acquisitions, and Technical Services 27.1 (Spring 2003): [pagination unavailable].16 Apr. 2003. http://www.oclc.org/research/publications/archive/2003/lavoie_frbr.pdf.

Chan, Lois Mai. Cataloging and Classification: An Introduction. 2nd ed. New York: McGraw-Hill, 1994.

Cutter, Charles A. Rules for a Dictionary Catalogue. 2nd ed. Washington, DC: Government Printing Office, 1889.

Dorman, David. "'Furber' Debuts at ALA Exhibits." American Libraries. Aug. 2002: 84.

Freedman, Maurice. "The Functions of the Catalog and the Main Entry as Found in the Work of Panizzi, Jewett, Cutter, and Lubetzky." Diss. Rutgers, The State University of New Jersey, 1983.

Fitch, Kent. "Taking RDF and Topic Maps Seriously — What Happens When You Drink the Kool Aid." 2002. AusWeb02. 10 Apr. 2003. http://ausweb.scu.edu.au/aw02/papers/refereed/fitch2/paper.html.

Hickey, Thomas B., Edward T. O'Neill, and Jenny Toves. "Experiments with IFLA Functional Requirements for Bibliographic Records (FRBR)." D-Lib Magazine 8.2 (Sep. 2002). 12 March 2003. http://www.dlib.org/dlib/september02/hickey/09hickey.html.

IFLA Study Group on the Functional Requirements for Bibliographic Records. Functional Requirements for Bibliographic Records. Munich: K.G. Saur, 1998. 20 Mar. 2003 http://www.ifla.org/VII/s13/frbr/frbr.pdf.

International Conference on Cataloguing Principles (1961: Paris, France). Report. Ed. A.H. Chaplin and Dorothy Anderson. London: Organizing Committee of the International Conference on Cataloguing Principles, 1963.

Le Boeuf, Patrick. "FRBR and Further." Cataloging & Classification Quarterly 32.4 (2001): 15-52.

Lehnus, Donald J. A Comparison of Panizzi's 91 Rules and the AACR of 1967. Urbana, IL: University of Illinois Graduate School of Library Science, 1972.

Lubetzky, Seymour. Cataloguing Rules and Principles. 1953. Seymour Lubetzky: Writings on the Classical Art of Cataloging. Comp. and ed. Elaine Svenonius, Dorothy McGarry. Englewood, CO: Libraries Unlimited, 2001. 78-139.

Lubetzky, Seymour. "Panizzi vs. the 'Finding Catalog'." 1956. Seymour Lubetzky: Writings on the Classical Art of Cataloging. Comp. and ed. Elaine Svenonius, Dorothy McGarry. Englewood, CO: Libraries Unlimited, 2001. 174-179.

Lubetzky, Seymour. "The Function of the Main Entry in the Alphabetical Catalogue — One Approach." 1961. Seymour Lubetzky: Writings on the Classical Art of Cataloging. Comp. and ed. Elaine Svenonius, Dorothy McGarry. Englewood, CO: Libraries Unlimited, 2001. 231-237.

Lubetzky, Seymour. Principles of Cataloging. Final Report. Phase I: Descriptive Cataloging. 1969. Seymour Lubetzky: Writings on the Classical Art of Cataloging. Comp. and ed. Elaine Svenonius, Dorothy McGarry. Englewood, CO: Libraries Unlimited, 2001. 259-341.

O'Neill, Edward. "FRBR: Application of the Entity-Relationship Model to Humphry Clinker." 15 June 2002. Notes taken by Judith Hopkins. 16 Apr 2003. http://www.acsu.buffalo.edu/~ulcjh/FRBRoneill.html.

Ranganathan, S.R. The Five Laws of Library Science. 2nd ed. with amendments. Bombay, India: Asia Publishing House, 1963.

Smiraglia, Richard P. The Nature of "A Work": Implications for the Organization of Knowledge. Lanham, MD: Scarecrow Press, 2001.

Svenonius, Elaine. The Intellectual Foundations of Information Organization. Cambridge, MA: MIT Press, 2000.

Svenonius, Elaine, and Dorothy McGarry. Introduction. Seymour Lubetzky: Writings on the Classical Art of Cataloging. Comp. and ed. Svenonius and McGarry. Englewood, CO: Libraries Unlimited, 2001. xi-xxi.

Working Group on FRBR. "Frequently Asked Questions About FRBR." 12 Feb. 2003. International Federation of Library Associations and Institutions. 18 Apr. 2003. http://www.ifla.org/VII/s13/wgfrbr/faq.htm.