Functional Requirements for Bibliographic Records (FRBR): Hype or Cure-All?

Transcription

1

2 Functional Requirements for Bibliographic Records (FRBR): Hype or Cure-All? Functional Requirementsfor Bibliographic Records (FRBR): Hype or Cure-All? has been co-published simultaneously as Cataloging & Classification Quarterly, Volume 39, Numbers 3/42005.

3 This page intentionally left blank

4 Functional Requirements for Bibliographic Records (FRBR): Hype or Cure-All? Patrick Le Boeuf Editor Functional Requirements for Bibliographic Records (FRBR): Hype or Cure-All? has been co-published simultaneously as Cataloging & Classification Quarterly, Volume 39, Numbers 3/

5 First published by The Haworth Information Press, Inc., 10 Alice Street, Binghamton, NY USA The Haworth Information Press is an imprint of The Haworth Press, Inc., 10 Alice Street, Binghamton, NY USA This edition published 2013 by Routledge Routledge Taylor & Francis Group 711 Third Avenue New York, NY Routledge Taylor & Francis Group 2 Park Square, Milton Park Abingdon, Oxon OX14 4RN Routledge is an imprint o f the Taylor & Francis Group, an informa business Functional Requirements for Bibliographic Records (FRBR): Hype or Cure-All? has been co-published simultaneously as Cataloging & Classification Quarterly, Volume 39, Numbers 3/ by The Haworth Press, Inc. All rights reserved. No part of this work may be reproduced or utilized in any form or by any means, electronic or mechanical, including photocopying, microfilm and recording, or by any information storage and retrieval system, without permission in writing from the publisher. The development, preparation, and publication of this work has been undertaken with great care. However, the publisher, employees, editors, and agents of The Haworth Press and all imprints of The Haworth Press, Inc., including The Haworth Medical Press and Pharmaceutical Products Press, are not responsible for any errors contained herein or for consequences that may ensue from use of materials or information contained in this work. Opinions expressed by the author(s) are not necessarily those of The Haworth Press, Inc. With regard to case studies, identities and circumstances of individuals discussed herein have been changed to protect confidentiality. Any resemblance to actual persons, living or dead, is entirely coincidental. Cover design by Kerry E. Mack. Library of Congress Cataloging-in-Publication Data Functional requirements for bibliographic records (FRBR): hype or cure-all? / Patrick Le Boeuf, editor. p. cm. Co-published simultaneously as Cataloging & Classification Quarterly, Volume 39, Numbers 3/ T.p. verso. Includes bibliographical references and index. ISBN-13: (hc. : alk. paper) ISBN-10: (hc. : alk. paper) ISBN-13: (pbk. : alk. paper) ISBN-10: (pbk. : alk. paper) 1. Cataloging. 2. Bibliography-Methodology. 3. Machine-readable bibliographic data. 4. Information retrieval. 5. Information organization. I. Le Boeuf, Patrick. II. Cataloging & classification quarterly. Z693.F dc

6 Functional Requirements for Bibliographic Records (FRBR): Hype or Cure-All? CONTENTS Dedication Maja turner xv FRBR: Hype or Cure-All? Introduction 1 Patrick Le Boeuf The Origins of the IFLA Study on Functional Requirements for Bibliographic Records 15 Olivia M. A. Madison Extending FRBR to Authorities 39 Glenn E. Patton Modeling Subject Access: Extending the FRBR and FRANAR Conceptual Models 49 Torn Delsey rdfs:frbr-towards an Implementation Model for Library Catalogs Using Semantic Web Technology 63 Stefan Gradrnann Cataloguing of Hand Press Materials and the Concept of Expression in FRBR 77 Gunilla Jonsson The AustLit Gateway and Scholarly Bibliography: A Specialist Implementation of the FRBR 87 Kerry Kilner

7 Musical Works in the FRBR Model or "Quasi la Stessa Cosa": Variations on a Theme by Umberto Eco 103 Patrick Le Boeuf Paradigma: FRBR and Digital Documents 125 Ketil Albertsen Carol van Nuys "Such Stuff as Dreams Are Made On": How Does FRBR Fit Performing Arts? David Miller Patrick Le Boeuf Folklore Requirements for Bibliographic Records: Oral Traditions and FRBR Yann Nicolas FRBR and Cataloging for the Future 197 Barbara B. Tillett Sloven ian Cataloguing Practice and Functional Requirements for Bibliographic Records: A Comparative Analysis 207 Zlata Dimec Maja tumer Gerhard 1. A. Riesthuis Implementation of FRBR: European Research Initiative 229 Majatumer FRBRizing OCLC's WorldCat 239 Thomas B. Hickey Edward T. 0 'Neill Implementing the FRBR Conceptual Approach in the ISIS Software Environment: IFPA (ISIS FRBR Prototype Application) 253 Roberto Sturman

8 FRBR Display Tool 271 Jacqueline Radebaugh Corey Keith XOBIS-An Experimental Schema for Unifying Bibliographic and Authority Records 285 Dick R. Miller Index 305

9 This page intentionally left blank

10 ABOUT THE EDITOR Patrick Le Boeuf obtained the "archiviste-paleographe" degree delivered by the Ecole nationale des chartes (Paris, France) in 1986, and he is currently Library Curator in the Department for Standardization at Bibliotheque nationale de France. Mr. Le Boeuf has been a member of the IFLA Cataloguing Section's Standing Committee since Both his talk on FRBR, given in Boston, Massachusetts, on the occasion of the 2001 IFLA Conference, and his FRBR and Further article published in CCQ 32(4) in 2001, established his reputation as a "FRBR commentator," and he was asked to chair the IFLA Working Group on FRBR (now renamed the FRBR Review Group) in 2002, a position he is still holding as of Mr. Le Boeuf also chairs the IFLA Working Group on FRBRJCRM Dialogue, a working group that aims at interoperability between the FRBR model for bibliographic information and the CIDOC CRM conceptual model for information about museum objects. He is frequently asked to give talks and lectures about FRBR, in France and abroad, and in October, 2003, at Richard P. Smiraglia's invitation, he gave a talk on Musical Works in the FRBR Model at the ASIS&T (American Society for Information Science & Technology) Annual Conference in Long Beach, California.

11 This page intentionally left blank

12 Dedication MajaZumer This special volume is dedicated to Zlata Dimec ( ), Slovenian librarian, cataloguer, researcher, colleague, and friend. Zlata Dimec spent most of her career at the National and University Library in Ljubljana. Six years ago she transferred from her position in the Cataloging department to Research and Development, the department I was heading at that time, so there was ample opportunity to discuss with her the potential, possibilities, and challenges of the FRBR model. Zlata gave me my first introduction to FRBR. She was very enthusiastic about the model, so much so that she prepared, at least to my knowledge, the first translation of the study into a language other than English. During the following years, FRBR was very much the focus of her workand our discussions. She concentrated on the cataloguing aspects; I was giving more emphasis to the potential application aspects. We worked together on the use of FRBR for the IFLA study 'Guidelines for OPAC Displays.' We planned our future research: Ziata as a seasoned librarian, I as a relative newcomer. We were both intrigued by the consequences offrbr implementation, both for cataloguers and for end-users. At the same time, we were also engaged in a retrospective conversion project bridging the past and the future. I still remember how we analyzed the implementation of very old cataloguing rules and compared them to current theory and practice-beautiful, old handwritten Maja Zumer, PhD, is affiliated with the Department of Library and Information Science and Book Studies, University of Ljubljana, Faculty of Arts, Askerceva 2, 1000 LjUbljana, Slovenia ( maja.zumer@ff.uni-lj.si). She is a member of the IFLA FRBR Review Group, the IFLA WG on FRBR Teaching & Training, and the IFLA WG on the FRBRlCRM Dialogue. [Haworth co-indexing entry note]: "Dedication." Zumer. Maja. Co-published simultaneously in Calaloging & Classification Quarterlv (The Haworth Information Press, an imprint of The Haworth Press. Inc.) Vol. 39. No. 3/4, pp. xxi-xxii; and: Functional Requirements for Bibliographic Records (FRBR): Hype or Cure-All? (ed: Patrick Le Boeuf) The Haworth Information Press, an imprint of The Haworth Press, Inc., 2005, pp. xv-xvi. Single or multiple copies of this article are available for a fee from The Hawol1h Document Delivery Service [1-800-HAWORTH, 9:00 a.m. - 5:00 p.m. (EST). address: doc deli very@hawol1hpress.com] by The Haworth Press, Inc. All rights reserved. xv

13 xvi Functional Requirements for Bibliographic Records (FRBR): Hype or Cure-All? cards and the FRBR computer catalog, two technologies aiming at the same goal. By that time, Zlata was already recognized as the most important cataloguing expert in Slovenia. She was working on an overview of Slovenian cataloguing rules and practice as a basis for the development of new cataloging rules. She was a member of the IFLA Cataloguing Section Standing Committee and several working groups: 'Guidelines for OPAC Displays,' ISBD(S) Revision, GARE Revision, and Multilingual Dictionary of Cataloguing Terms. It was an extremely productive period for Zlata. She published several papers and, finally, also decided to formalize her knowledge and experience by acquiring an M.L.S. degree. It was a strange quirk of fate that she enrolled exactly on the day she learned of her illness. But she did not give up. She continued to work even during her therapy and lengthy stays in the hospital. She finished her course assignments, and we were still making plans for future projects. And then, on a winter morning two years ago, we learned that Zlata had passed away. We were suddenly facing what we stubbornly refused to believe possible. Zlata, with all her strength and energy, was gone. We lost a colleague and a friend. Too many things were left undone and unsaid, so many plans unfulfilled. We never wrote the article on bibliographic relationships in UNIMARC and we never went to see The Lord of the Rings... I am deeply grateful for the time we spent together.

14 FRBR: Hype or Cure-All? Introduction Patrick Le Boeuf SUMMARY. FRBR does not account for "reality," but for a "conceptualization" of reality. It certainly shows innovative features, particularly with regard to activities related to the "Semantic Web," but also elements of conservatism in its approach. The "logical flaws" that are sometimes denounced in the analysis it embodies actually reflect logical flaws in cataloging practice itself, showing the value of FRBR as a tool for assessing such practice. As to future evolutions in cataloging, alternatives to FRBR are possible. [Article copies available for a fee from The Haworth Document Delivery Service: J-800-HA WORTH. address: <docdelivery@haworthpress.com> Website: < by The Haworth Press, Inc. All rights reserved.} KEYWORDS. Functional Requirements for Bibliographic Records, semantics of bibliographic records, conceptual modeling for cataloging practice Patrick Le Boeul', archiviste-paleographe, is Library Curator, Bibliotheque nationale de France, Department for Standardization, Quai Fran~ois Mauriac, Paris cedex 13, France ( patrick.le-boeuf@bnffr). He is a member of the IFLA Cataloguing Section's Standing Committee, chair of the IFLA FRBR Review Group, chair of the IFLA Working Group on the dialog between FRBR and the CIDOC CRM, and co-chair (with Martin Doerr) of the FRBRlCRM Harmonization Group. [Haworth co-indexing entry note]: "FRBR: Hype or Cure-All? Introduction." Le Boeuf, Patrick. Co-published simultaneously in Cataloging & Classification Quarterlv (The Haworth Information Press, an imprint of The Haworth Press, Inc.) Vol. 39, No. 3/ , pp. 1-13; and: Functional Requirements jar Bibliographic Records (FRBR): Hype or Cure-All? (cd: Patrick Le Boeuf) The Haworth Information Press, an imprint of The Haworth Press, Inc , pp Single or multiple copies of this article are available for a fee from The Haworth Document Delivery Service [1-800-HA WORTH, 9:00 a.m. - 5:00 p.m. (EST). address:docdelivery@haworthpress.com] by The Haworth Press, Inc. All rights reserved. Digital Object Identifier: /11 04v39n03_01 1

15 2 Functional Requirements for Bibliographic Records (FRBR); Hype or Cure-All? The title of a study on Functional Requirements for Bibliographic Records (FRBR Final Report 1998) has now come to be known as the name of the very model it describes, and is more often referred to through the acronym FRBR than through its complete title. The acronym "FRBR" is now understood as a singular, at least in English (in other languages, such as French or Spanish, it is still perceived as a plural noun). It is more and more often pronounced as two syllables ("FeR-BeR") rather than four initials. In librarians' jargon, it has even become the root of a family of new terms: FRBR tree, to FRBRize, FRBRizing, FRBRization. Perhaps the capitalization itself will eventually disappear some day, making the acronym look like any other word (except for the vocalic R that reads more like Czech than like English): "Your catalog has not been frbrized yet? Oh, how old-fashioned... " FRBR has been included in training courses; FRBR is the main topic of meetings, symposia, workshops, and their associated proceedings; FRBR is even mentioned outside libraries... FRBR is trendy. And now, this special volume is devoted to it. New words do not necessarily indicate that new needs or new concepts have emerged, but they sometimes do. Is it the case with the whole bunch of FRBRish terms, or is FRBR just hype? A JOURNEY FROM "REAL WORW ENTITIES" TO THE LIES OF LANGUAGE Aren't we all guilty of the arrogant belief that our words describe real things rather than just describe our own weak internal images of reality? (Hofstadter 1997, 100) Imagine you are not a librarian, and you are sitting in a comfortable armchair (i.e., definitely not in a library's reading room). You are holding a book in your hands-say, Edna Ferber's So Big translated into Spanish as ias! de grande! (Your intention is to improve your Spanish.) This is a physical object; you can feel its weight. The ink on the paper is physical as well, but it is arranged in precise patterns that you can recognize: letters (although this is a new font you had never seen before), words, sentences. Even though you may not understand each single sentence (you have not practiced your Spanish for a long time, perhaps), you can still recognize that this is not merely physical ink on physical paper, but that the ink on paper gives you access to something that is not physical, a conceptual object-a text. As you are neither a librarian nor a bibliographer, you are only interested in those two real things: the object you are holding in your hands, and the text you are reading, and (hopefully) understanding, and (optionally) enjoying.

16 Patrick Le Baeuf J Now, imagine you are a librarian, and you are striving to catalog iasf de grande! Of course you are aware that this is a physical object, but you are mainly interested in its title page: you are eager to find information there that does not pertain to the specific object you are holding in your hands, but to the general class it belongs to-the "edition." What is an edition? Perhaps you could not tell. You cannot hold one in your hands, you cannot read and understand and enjoy one; but this is precisely what you want to describe in a bibliographic record: an "edition." Warning! Warning! The title page reads: "traducido en castellano por... " A translation? "Oh my! I have to make a uniform title, so that the work can be collocated." What is a work? You are unable to say, but you know it has to be collocated. As a librarian-nay, as a catalogeryou are mainly interested in two "things" that you are unable to define: "work" and "edition." A cataloger's world is, therefore, amazingly blurred: "Critical as it is in organizing information, the concept of work has never been satisfactorily defined" (Svenonius 2000, 35); "Like work, the concept of edition has fuzzy boundaries" (Svenonius 2000, 39). All you can say is that a "real" physical book belongs to an "edition," however fuzzy the concept is, and that a "work," again however fuzzy the concept is, is collocated across a number of "real" texts. You have a picture with four objects: two distinct objects, each of which is related to an indistinct cloud of points. This is roughly the FRBR picture. But since FRBR was not meant just for books, but for any kind of material that is likely to be found in a library, the FRBR words are more generic. In FRBR parlance, the physical carrier of a text is said to be an "Item." Well, this is just a matter of taste. An "edition" (that first cloud of points) is called a "Manifestation"-something has to manifest itself, then? Yes, indeed: the "text," which is not named "text" any longer, but "Expression." "Text" is a word that stands on its own, whereas "Expression" makes you expect something else-the expression of what? And here comes the second cloud of points-surprise! in the FRBR dialect, a work is still called a "Work." At last a word you are used to using-that seems encouraging. As a matter of fact, it is only the beginning of even more serious trouble, as the FRBR term "Work" may well not have the same meaning as your "work." But that is another story. The FRBR picture is not entirely new. It has forerunners. In a sense, FRBR can be seen as Seymour Lubetzky' s victory, since Lubetzky "lamented that the material book and the intellectual work were often confused, a serious confusion in that it implied that the objectives of the catalog were not understood" (Svenonius 2000, 35). Ranganathan had a picture with three objects instead of four: he "viewed a document as consisting of three aspects: (1) a Soul, that is,

17 4 Functional Requirements for Bibliographic Records (FRBR): Hype or Cure-All? the thought(s) or ideas embodied, (2) a Subtle body, that is, the mode of exposition [... ], and (3) a Gross body, that is, the medium or carrier for storing and distribution" (Raghavan & Neelameghan 2002, 194). (Interestingly enough, the "missing" object seems to be the edition, i.e., precisely what catalogers deem most important to describe in bibliographic records.) In 1995, Michael Heaney proposed an object-oriented model for cataloging that has much in common with FRBR, again with three objects only in the picture: abstract work, publication, and copy (Heaney 1995). And Alfredo Serrai complains that the FRBR study does not pay credit to him, although he suspects that FRBR's originators plagiarized ideas expressed by himself well before IFLA even thought of providing itself with an entity-relationship model for cataloging (Serrai 2002). FRBR had one serious advantage on its forerunners, though: it was promoted by IFLA, and could thus meet worldwide recognition. Besides, FRBR had the potential to appeal to many people, because it is a unique blend of nostalgia and innovation. PARADISE LOST Technological advance has brought with it a steady deterioration in the integrity of bibliographic structures since the time of Panizzi and, with it, an undermining of bibliographic objectives. (Svenonius 2000, 64) The "steady deterioration" that Elaine Svenonius denounces consists in "a loss of bibliographic structure," the absence of a hierarchical display of entries, the lack of sense for the context in which each new translation, edition, and adaptation of a work inevitably takes place. The wonderful syndetic structure of printed catalogs has yielded to databases that are barely more than collections of unrelated monads. The so-called "FRBR tree," such as implemented for instance in the Virtu a system by VTLS Inc., revives that syndetic structure, and gives a graphic representation of the context for each new member of a "bibliographic family." The FRBR model could therefore be seen as a (last chance?) attempt at restoring both bibliographic structures and bibliographic objectives, at regaining the Paradise we have lost by computerizing our catalogs, and at giving the correct answer to the spidery Sphinx who keeps facing us with the Digital Era Dilemma: "Bibliographic Control, or Chaos'?" In that sense, FRBR can be labeled conservative-a device intended to make our computerized catalogs look more like nineteenth-century printed catalogs than like those undecipherable screens full of ill-related, erratic bibliographic

18 Patrick Le Boeu;{ 5 descriptions. The entire bibliographic history of Hamlet at a glance! All translations, editions, and copies of the Bible on only one screen, and sorted so as to make sense! A wonderful dream indeed, a dream that has come true in the Virtua system, in the AustLit Gateway, in the Library of Congress' FRBR Display Tool-hut it does not take two facts into account (of course, this is certainly not to criticize Virtua, nor AustLit Gateway, nor the FRBR Display Tool): (a) a "FRBR tree" may not he what a user is interested in (a user can look for all translations by a given translator, for instance, and only that translator, rather than for all translations of a given work); (b) some creators (e.g., Emily Dickinson I) resist FRBRization, at least in that acceptation of the word. THE DAY AFTER TOMORROW The transition from card to online catalogs, though ongoing for over thirty years now, is still in its initial stages. (Svenonius 2000, 63) Memory Organizations are in a unique position to lead the Semantic Web to its full potential. (Miller 2002, 32) FRBR cannot be reduced to just the "FRBR tree" and its implicit longing for nineteenth-century "bibliographic objectives"-although this is, of course, a significant feature of FRBR, the benefits of which I certainly do not reject. Perhaps, though, the mere fact that FRBR strives to explicate a great deal of the relationships that lie dormant in ISBD records and MARC formats, and what they relate at all, is a more innovative aspect of the model, and shows more potential for the future. It makes it possible to map our "flat" catalogs to an "ontology," to transcend practice into vision, and to connect our librarian world to the broader domain of information management on the Web. As Elaine Svenonius puts it, online catalogs are only at a beginning. We developed them with the card catalog paradigm in mind (not even the printed catalog paradigm, which would have been better) and under the implicit assumption that they could only serve to locate a "known item" or to answer a precise subject query. Current trends in information management (outside lihraries) are more demanding. Ontologies are being developed in order to feed the (already) controversial "Semantic Weh" (or, as Mark Butler would rather call it, the "Symbolic Web" (Butler 2003, 40), as the verbal association between semantics and machines is deemed by him "unhelpful and confusing"-indeed, the Semantic Web is as controversial as its predecessor Artificial Intelligence, with which it is often compared). In combination with other ontologies, and

19 6 Functional Requirements for Bibliographic Records (FRBR): Hype or Cure-All? thanks to tools borrowed from the Semantic Web techniques, the incredibly rich but largely unexploited information dug into our catalogs could allow for such deductive reasoning or "inferences" as: If Person X was female and spent most of her lifetime in Victorian England, then the works written by that person can be regarded as a testimony of women's condition and state of mind during the Victorian Era; or: If Person Y was friends with Person Z, and Z's correspondence has been published, perhaps I will find information about Yin Z's published correspondence;2 or: If a given author has written a text titled Memoir on... and that author was a member of a learned society that published a serial titled Memoirs..., perhaps that specific text titled Memoir on... was published in the serial titled Memoirs All the information necessary for such deductive reasoning need not be taken from our catalogs only (of course, we do not record information about friendship relationships in our authority records), but the combination of ontologies, resources, and semantic agents will (hopefully) allow future systems, as Patrick Sinclair puts it, to "automatically obtain missing information from the semantic web" (Sinclair 2004,12). Unfortunately, for the time being, as Patrick Sinclair admits immediately after expressing that possibility, "there is not much semantic web around, especially containing the type of information we are looking for" (Sinclair 2004, 12). Stefan Gradmann, in his article included in the present collection, says more about the FRBR potential in the context of the Semantic Web, and gives further examples of "bibliographic inferences," if I may say so. Unfortunately again, the form under which FRBR is currently available may not be sufficient for such purposes. As a matter of fact, the FRBR model is not finished, it is only a first step, a transitory stage. FRBR is a beginning, not an end. Further work is still ongoing, or planned. Two directions for such work are described in the present collection: Glenn Patton exposes the FRANAR (or FRAR, as you will notice some other authors call it) model, which extends the IFLA modeling effort to authority files, and Tom Delsey lays the basis for a future, more encompassing model that will deal with subject relationships, indexing, and classifying issues. A third, major project could not be handled in the present collection, due to time constraints. There is a very, very promising collaboration with the museum community that is developing the CIDOC CRM semantic model for cultural heritage information in museums and assimilated institutions (Gill 2004). The IFLA Cataloguing Section has formed a working group on the dialog between FRBR and CIDOC CRM. That working group has merged with part of the CIDOC CRM Special Interest Group in order to form the FRBRlCRM Harmonization Group, which is currently "translating" FRBR

20 Patrick Le Boeul 7 into object-oriented formalism, so that it can be more easily compared to CIDOC CRM, and so that all of its implicit semantics can be explicated. The intellectual rigor and demands of Martin Doerr, the main originator of the CIDOC CRM, and his expertise in modeling techniques, are very much appreciated and help us chase out every assumption that had been left implicit in the entity-relationship formalization of the model. FRBR will never look the same again. Indeed, the two articles (outside this introduction) which I authored or co-authored in the present collection would not have been written the same way at all, if I had written them after the FRBRlCRM Harmonization Group's meetings that took place at the end of March The library community has to reckon with the CIDOC CRM model; and the resulting object-oriented model will be even more helpful in the Semantic Web context. FIA WS IN FRBR?... the arrogant belief that our words describe real things rather than... our own weak internal images of reality. (Hofstadter 1997, 100) I am repeating that Hofstadter quotation because I think it is nowhere as true as in the field of ontology-making. It matches Tom Gruber's definition of an ontology as "a specification of a conceptualization" (Gruber 1993), i.e., not a description of what exists, as one could easily believe, because of the literal meaning of the etymons that constitute the word, but a description of what one thinks exists. Just like any other human statement, an ontology can be wrong. The FRBR model has often been criticized, and several contributors to the present collection do not express a blind, unconditional acceptance and support of the model-so much the better. In particular, it is a frequently expressed opinion that the FRBR Expression entity is "problematic." This always surprises me, because for me Expression is the most self-evident entity in FRBR, the closest approximation in words, together with Item and Object, to something real. It is much clearer to me than the puzzling Manifestation (the precise nature of which I have not elucidated yet), or the ethereal Work. Is it due to my background in textual criticism,4 or to my background as sound recording cataloger,5 or is there another explanation? Much of the "Expression debate" is devoted to existential questioning: "Is this a work or an expression? Is that an expression or a manifestation?" There also is frequent complaining about the alleged connict between the need for collocation and the claim that even a single comma is sufficient to define a

21 8 Functional Requirements for Bibliographic Records (FRBR): Hype or Cure-All? new expression (which is supposed to ruin the collocation purposes), and about the fact that catalogers have neither time nor competence to check whether two manifestations carry "exactly" the same expression or two distinct expressions. Actually, I think that the "Expression problem" is not inherent to the FRBR model, but to our cataloging practice. FRBR models what we do, not what we should do. If we are inconsistent and illogical in our practice, the mirrored image of our practice cannot be consistent and logical. A picture of a cat cannot depict a dog. The FRBR Final Report does not clearly state whether an illustrated text, or a text issued together with an introduction, is "an Expression" or not, because the conclusion that illustrations or forewords are just parts of an expression of a work would be totally illogical, and yet this is what we do in our everyday practice: better to hide the corpse in a closet than to explain how it is here. Kerry Kilner makes that point perfectly clear in her contribution to the present collection: "the Introduction is a separate work with its own expression and manifestation and is contained in the embodied publication"; but unlike other kinds of separate works, there is no separate entry for that work as such, and the semantic value of the ISBD punctuation explicitly says that the author of the introduction or of the illustrations is just a secondary kind of author for the main work. Compare the orthodox use of ISBD punctuation, as found in an example from ISBD(G), Revised Edition: Hard times [GMD] ; Hunted down; Holiday romance; & George Silverman's explanation / by Charles Dickens; with seven illustrations / by F. Walker and Maurice Greiffenhagen with the "heretic" use of that same punctuation, with the same semantic value: Hard times [GMD] ; Hunted down; Holiday romance; & George Silverman's explanation / by Charles Dickens. With, Seven illustrations / by F. Walker and Maurice Greiffenhagen The orthodox punctuation indicates that F. Walker and Maurice Greiffenhagen are not the authors of a separate work, distinct from Hard Times, etc., but of an aspect (an expression?) of -of which of the four works, by the way, since there is no collective title proper? The "heretic" punctuation, by contrast, indicates that the illustrations (graphic works by nature, as opposed to the textual works otherwise embodied in the same manifestation) are regarded as separate works, just like in the "modified FRBR model with enhanced manifestation options" developed by the AustLit Gateway team according to Kerry Kilner (actually, this is not a "modified

22 Patrick Le Boeuf 9 FRBR model," but a FRBR model in which the logical quality that is missing in ISBDs, AACR, MARC, and other usual cataloging tools, has been restored). FRBR is a cruel looking-glass. It could be argued that the frequent absence of any title (proper) for illustrations, forewords, etc., forbids that such contributions should be regarded as works on their own. But most relator codes (not all of them, though), which are often said to be dramatically important for the accurate automated identification of expressions, could be analyzed as segments of uniform titles for separate works. For instance, would it not be possible to interpret: Dickens, Charles Hard times / by Charles Dickens; ill. by Maurice Greiffenhagen Greiffenhagen, Maurice. Ill. as actually meaning: Work #1: Dickens, Charles.-[Hard times] Work #2: Greiffenhagen, Maurice.-lIlIustrations to Hard times (Dickens, Charles)] Manifestation embodying both works: Hard times / by Charles Dickens. [With,] Ill. / by Maurice Greiffenhagen with both artistic creations being regarded as works? Dick Miller demonstrates, in the present collection, that XML makes it quite possible to have such "cascading" constructs for uniform titles in a library catalog, with full authority control and enough flexibility so as to serve many purposes. By contrast, the relator code for "translator" clearly can serve for a uniform title for an expression: Shakespeare, William Hamlet / Shakespeare; trad. de Franc;ois-Victor Hugo Hugo, Franc;ois-Victor. Transl. actually means: Work: Shakespeare, William.-[Hamlet] Expression: Shakespeare, William.-[Hamlet (French (Hugo, Franc;ois-Victor»]

23 10 Functional Requirements fur Bibliographic Records (FRBR): Hype or Cure-AIL? Manifestation: Hamlet / Shakespeare; trad. de Fran<;ois-Victor Hugo All in all, FRBR is neither hype nor a cure-all; it is just a helpful tool. THE PRESENT COLLECTION I deeply regret that, due to obvious constraints, it was not possible to investigate the implications of the FRBR model for all kinds of materials and to cover all of the topics that can be thought of when mentioning FRBR. It would have taken at least another volume to cover: continuing resources; manuscripts; engravings; photographs; movies; the SMART model for Chinese calligraphy (Could it serve for Arabic calligraphy as well? How significant is it that the language attribute has been displaced in SMART from the Expression entity to the Work entity? These are issues I would have liked to see addressed.); and publishers' and rights managers' concerns, etc. The first section of the present collection deals successively with the past and with the future (or rather, with some aspects of the future) offrbr: Olivia Madison, who chaired the FRBR Study Group, tells us all the truth about why and how FRBR was developed; Glenn Patton addresses FRANAR and highlights its "usefulness as a model for implementations that break the traditional authority file structure"; Tom Delsey, one ofthe main designers of FRBR and FRANAR, opens the way to revisions and extensions of FRBRIFRANAR so that subject relationships can be more precisely modeled; Stefan Gradmann highlights the potential of FRBR in the international context of the Semantic Web. The second section investigates the applicability (or non-applicability, or conditional applicability) of FRBR to a range of diverse kinds of human creations, from the most physical to the most "impalpable": hand press materials (Gunilla Jonsson), literary texts (Kerry Kilner), musical works (Patrick Le Boeui), digital documents (Ketil Albertsen and Carol van Nuys), stage productions (David Miller and Patrick Le Boeui), and oral literature (Yann Nicolas). The third section addresses implementation issues, and is introduced by Barbara Tillett who, in a short but far-reaching and comprehensive article, sketches the possible impact of FRBR on our everyday cataloging practice in the future. Then FRBR is mapped to a national cataloging code by Zlata Dimec, Maja Zumer, and Gerhard Riesthuis. Maja Zumer exposes the major currents in FRBR implementation research in Europe, and the experiments toward some FRBRization of OCLC's WorIdCat are presented by Thomas Hickey and Edward O'Neill. Roberto Sturman describes an amazing system

24 Patrick Le Boeuf II that he developed by himself, and for which he created a format that seems to have the structure of a MARC format, but that is entirely articulated according to the FRBR entities. Jacqueline Radebaugh and Corey Keith give us a demonstration of the FRBR Display Tool developed by the Library of Congress. The fourth and last section contains only one article, by way of conclusion. Dick Miller shows us everything that XML can do for libraries-and their users. XOBIS is a revolutionary alternative to the comparatively conservative and "traditional" approach that FRBR represents. FRBR has a futuristic potential, but in this introduction I also strove to highlight its underlying nostalgia for the nineteenth-century "bibliographic ideal," and its limitations as an ontology designed exclusively to model extant practices. XOBIS resolutely belongs to the twenty-first century. In inviting Dick Miller to conclude this collection, my intention was not to arouse vain comparisons or to create futile contrasts, but just to show that other directions can be envisioned for the future of cataloging. This collection of FRBR articles is dedicated to the memory of Zlata Dimec. I never met her personally-she was already seriously ill when I first attended an IFLA Conference in 200 I, and she was not able to make the trip to Boston-but I know how appreciated she was within the Cataloguing Section, how enthusiastic she was about FRBR, and how eagerly she strove to promote the model developed by IFLA. Let her name remain associated with that landmark in the history of bibliographic thought that FRBR is. NOTES I. In a future article, I intend to demonstrate that Emily Dickinson's writings are not "FRBRizable" (i.e., in the context of the present paragraph, not likely to be graphically represented as a "FRBR tree" in a satisfactory way). 2. A colleague of mine, who works on the Gallica Digital Library Project at Bibliotheque nationale de France, once said to me, triumphantly: "You catalogers, all your cataloging rules and authority records are now totally useless! For instance, there is an online database devoted to Proust's archives; if you are interested in information about Montesquiou, you can assume that Montesquiou is represented in the database since he was friends with Proust, and you just have to type "Montesquiou" as a Google-like, full-text search query, and you retrieve all hits!" I gaped at him, somewhat puzzled, and dared not ask him: "How am I supposed to know that Montesquiou was friends with Proust? And should I look for all databases about all of Montesquiou's friends, if I want to retrieve something about Montesquiou?" 3. This is a real-life example, however strange it may read. A few months ago, a user asked me to retrieve in the catalog of Bibliotheque nationale de France a Memoire sur... (I do not remember what topic) written by... (I do not remember the author's name) in the nineteenth century. There was no hit for the title in the catalog, and only

25 12 Functional Requirements for Bibliographic Records (FRBR): Hype or Cure-All? one hit for the author's name, a discourse pronounced on the occasion of a meeting of a given Socihe savante from Marseilles. A further search on the name of that Societe savante showed that it used to publish a serial titled Memoires... ; I advised the user to look at that serial: the requested Memoire was there and the user was happy. All I did was try to apply the Semantic Web way of "thinking." 4. My dissertation was on the textual tradition of St. Augustine's treatise on music throughout the Middle Ages. It is an incredibly, indescribably thrilling experience to discover that a fifteenth-century manuscript that seems to present itself as a whole, was actually copied after a ninth-century manuscript, the last leaflet of which was missing and replaced with the last leaflet of an eleventh-century manuscript (two fragmentary expressions of one work reassembled in one manifestation/item as a third, complete expression of that work!). 5. It is often stated that music catalogers are more responsive to the Expression entity than printed materials catalogers-because it is a concept they are more used to dealing with. WORKS CITED Butler, Mark H Is the Semantic Web hype? Available online at <hup:llwwwuk.hpl.hp.com/people/marbutlisthesemanticwebhype.pdf> (accessed on 23 December 2003). (FRBR Final Report 1998) IFLA Study Group on the Functional Requirements for Bibliographic Records Functional requirements for bibliographic records: Final report. Miinchen: K. G. Sauro Also available online at <hup:// orglviiis 13/frbr/frbr.pdf> and at <hup:// Gill, Tony Building semantic bridges between museums, libraries and archives: The CIDOC Conceptual Reference Model. First Monday 9(5). Available online at < Gruber, Tom R A translation approach to portable ontologies. Knowledge Acquisition 5(2): Also available online at <ftp://ftp.ksl.stanford.edu/pub/ KSL_Reports/KSL ps.gz>. Heaney, Michael Object-oriented cataloguing. Information technology and libraries 14(3): Hofstadter, Douglas R Le ton beau de Marat: In praise of the musicoflallguage. New York, NY: Basic Books. Miller, Eric Weaving meaning: The Semantic Web and the potential for library leadership. Available online at < (accessed on 10 May 2004). Raghavan, K. S. & A. Neelameghan Composite multimedia works on CD: Catalog entry according to ISBD(ER) and AACR2 Revision In Smiraglia, Richard P., ed. Works as entitiesfor information retrieval. New York: London: Oxford: The Haworth Information Press: Also issued as Cataloging & Classification Quarterly Vol. 33, No. 3/4: Serrai, Alfredo Critica dei Functional Requirements for Bibliographic Records. Bibliotheca: rivista di studi bibliografici 2:

26 Patrick Le Boeuf 13 Sinclair, Patrick Mapping museum metadata to the CRM in Sculpteur. In Symposium on Practice of know/edge sharing, Heraklion, April Available online at < Svenonius, Elaine The intellectual foundation of information organization. Cambridge, MA: MIT Press.

27 This page intentionally left blank