As of 22 April 2009 this website is 'frozen' in time — see the current IFLA websites

This old website and all of its content will stay on as archive – http://archive.ifla.org

IFLANET home - International Federation of Library Associations and InstitutionsAnnual ConferenceSearchContacts

62nd IFLA General Conference - Conference Proceedings - August 25-31, 1996

Cooperative cataloguing: supply and exchange of data through a European Project and a European Union Feasibility study: the italian experience

Isa de Pinedo


Cristina Magliano


This paper is dealings, for what concerns Italian experience, with two aspects of cooperative cataloguing. Such activity has in fact been developed and funded in our country both at national and international levels, in the last case through an European project, issues and results of which, even if not completed at the moment, could be of common interest for an analysis within the framework prov ided by this Session. Before describing the European cooperative project, we hope by now well known also outside Europe but which could still be interesting hearing of for our eastern friends, we will briefly introduce the two different sides of action. We will subsequently try to point out major issues encountered and compromises and solutions adopted for reaching a common ground which has mad e, in these years, possible to cooperate as well as to operate, without compressing, beyond an unbearable and unacceptable pressure, the single national traditions of all participating members. We are realistically convinced that at this stage a global evaluation cannot be made, but we are on the other hand positively inclined to draw up some comments which we hope could be of some help in a wi der field of cooperation.


Servizio Bibliotecario Nazionale - SBNSBN Servizio Bibliotecario Nazionale, the Italian library network, started more than ten years ago and is alive with on line centralised functions since June 1992. Scope of SBN, which is funded by the Ministry for Cultural and Environmental Assets together with Regional and University bodies responsible for local and academic libraries, is that of building, maintaining and making available, together with a series of other special databases, the National Union Catalogue, through on line shared cataloguing. Istituto Centrale per il Catalogo Unico delle Biblioteche Italiane e per le Informazioni Bibliografiche - ICCU, an institution whose task, among others, is that of coordinating state libraries activities in the country, is responsible for running t he network and for managing an editing team, in charge of catalogue and of SBN Authority File maintainance. This last file will be soon available on line.

Libraries, participating to SBN and connected to the central system, the so called Index, through two main functions: shared cataloguing and interlibrary loan, are of various typologies (State, University, Local and Public Libraries etc.), with different tasks, depending for their fundings on different parent bodies and having a varied kind of end users. Up to now more than 500 libraries are co nnected on line, 424 of which are actively operating in the database. Input of records is provided, by each single participating library, according to its personnel professional training and availability, at different cataloguing levels. Records, where necessary, can be subsequently enhanced and edited by other libraries, through the cooperation and the editing team activities. According to cata loguing differences required by the two kind of material, the Union Catalogue has been conceived separated in two files one for current and the other for older publications, both having in common the Authors' File. Current and older, as well as Authority File records are available for downloading, on magnetic supports, in UNIMARC format. ICCU and five SBN libraries are CERL full members. Conso rtium of European Research Libraries

The environment where the international project has been developed is that of the Consortium of European Research Libraries-CERL. The Consortium's origins can be dated since early 1992, when it started in a temporary form, with a well defined scope, in term of time and money, which was that of funding a feasibility study directed to evaluate the establishing of a database for records of European Printing of the Hand Press Period (1450-1830).The Consortium is now operating, since two years, as a non profit organisation yearly funded by its members for a three year period, reneawable after each term. The company, limited by guarantee,has been incorporated in the United Kingdom and is now associating 28 Full and 19 Associate Members, between the more relevant and prestigious library instit utions in Western and Eastern Europe.

The Consortium's primary objective is that of establishing a database, in which will be recorded all European printing of the handpress period (from 1450 to 1830), named the Hand Press Book (HPB) Database. The project intends to promote, in Europe and outside Europe, activities directed to the documentation, preservation and dissemination of the invaluable European cultural heritage. Scope of th e project is stimulating retrocon activities as well as research, by means of making high quality records available, firstly through Europe and subsequently all over the world, to librarians and to end users.

CERL Project: its developments, parallel activities and issues


The Consortium, in order to attain its primary objective and according to the European Union requirements, made in 1993 a call for tender, open to European and non European participants, with the scope of finding a system where to develop the project and host the future Hand Press Book (HPB) Database. The basis of the call for tender were the specifications as defined and agreed upon in the feas ibility study. The offer submitted by the Research Library Group in the U.S. was considered by an "ad hoc" Technical Committee as beeing the most suitable to the project requirements.The Research Libraries Group Inc., Mountainview, California, is then the host of the HPB database for a period of three years. Since 23 January 1996, after a due period of tests, access to the first file is availab le on RLIN to full Consortium members. The file has been supplied by the Bayerische Staatsbibliothek Mnich and contains approx. 520.000 records of the period 1501-1830. Files to follow soon thereafter will include records from the Royal Library in Stockholm (approx. 40,000); ICCU, Rome (approx. 40,000); the BibliothŠque Nationale de France, Paris (approx. 40,000); the Royal Library in the Hague; The National and University Library in Zagreb; the National Library of Scotland and the British Library.The HPB Database will comprise individual members' files contributed by libraries or other organizations and will subsequently build up a union file with records at a higher cataloguing level. The union file will have an index function to the members' file records. The format agreed for loadin g and downloading of data is UNIMARC.The Consortium has worked during these past two years in a very tight cooperation with IFLA PUC, in order to adapt the format and to make it hospitable to old and rare material cataloguing needs. Downloading and editing procedures will be available in the second phase of the project, which will be starting next October.Parallel activitiesCERL's project, wher e participating members are all European institutions and which is born, thanks to voluntary cooperation, but outside the Line of Action provided for Libraries in the several "Framework Programmes for Research, Technological Development and Telematics" of the Commission of the European Communities, has been in this last year also the natural cradle of a separate EC funded feasibility study: the P ROLIB/CoBRA-UNIMARC.

In fact, in September 1995, after some initial approach to better define the extent and limits of the study, the final proposal was agreed upon between the Consortium and the EC. While only a subset of the Consortium Members was proposed and accepted by the European Union as contracting partners, the Bayerische Staatsbibliothek, Mnich took the lead in the project as main contractor. The Task has been completed at the end of January and the final report accepted at the end of March 1996.

Participant to the project have been:
- Bayerische Staatsbibliothek, Germany (BS)
- Bibliotheque Nationale de France, France (BNF)
- Koninklijke Bibliotheek Albert 1er, Belgium (KBB)
- Kungliga Biblioteket, Sweden (KBS)
- Istituto Centrale per il Catalogo Unico, Italy (ICCU)
- The British Library, United Kingdom (BL)

Scope of the CoBRA/UNIMARC feasibility study has been that of identifying and testing thepossible divergences in using bibliographic description components of the UNIMARC record and to identify problems arising from these and other issues when merging records from a number of different sources. The feasibility project has set out the development of a software tool for the analysis of UNIMARC data . The analysis is providing profiles, characteristics and trends of the different databases monitored in view of determining what divergences, if any, are in practice and in the application of UNIMARC, based on the data themselves. Results of the analysis will be used to make recommendations on the conformity and compatibility of the datasets and the practices analised with the recommendations o f the PUC for the minimum content record. Data will be disseminated as soon as EC will make them publicly available.

While CERL project, considered the widespreaded cooperation levels between different participating libraries institutions, has represented the natural ground on which the feasibility study could bedeveloped, on the other hand, the same study has valuably contributed to integrate activities directed to a better use of the exchange format for this kind of material, in the establishment of the HPB database. UNIMARC format issues and our direct experience.

We must keep in mind that the practice of cataloguing machine readable records or of converting in a m/r form records previously produced in other forms is highly varied in Europe as all over the world: technical applications, cataloguing and formats used are the best demonstration, if ever needed, of whatlibrarians can do helping phantasy and individuality to grow and diversify. We can't blame t hem for this. Economic, social and cultural developments have not always progressed at the same level everywhere.

For what concerns ICCU contribution, we are quite used to deal with problems arising from a cooperative scenario and concerning cataloguing rules, standards and formats. We know very well that notwithstanding efforts made in this direction, not only for standardisingbut also for monitoring, editing and enhancing the quality of the database, errors and misinterpretations are and will be the unavoi dable causes of duplications, in our as well in similar systems, in all union catalogues where one record is supposed to describe all the same editions of an item with all holding information linked to it. ICCU should perhaps have been favoured in participating to the two activities, since SBN records are available in UNIMARC and the same format has been chosen for uploading and downloading data , both in CERL project and in the UE feasibility study. But things have not worked out so easily as we might have thought at the beginning.First of all the only UNIMARC format, intensively used for the exchange of data all over Europe up to now, is that related to current publications. The availability of a standard is often at some extent a pure theoretical building not immediately comprehensiv e of all issues involved with its applicability.

As for all standards, solutions prompted and adopted in some peculiar cases, will need a further thorough assessment through a period of sound daily use and application. Apart from this consideration we can easily say that UNIMARC (A), at its first start, has found his best test bed in CERL/CoBRA project, for the benefit of all parts involved, and that the Consortium members have been the volu ntary guinea-pigs of this massive test. The UNIMARC format has become hospitable to the more sophisticated needs of the descriptive cataloguing of older materials without altering its scope and structure, and the very tight cooperation with IFLA PUC, through CERL consultants, has prooved very helpful during all these months in defining and trying to find out solutions, of mutual satisfaction to a ll parts involved, for the major problems arising often from national/local cataloguing practices and choices. This has meant for us a continous reporting, back and forth, and checking all possible decisions and agreed upon solutions against our national cooperative environment. It is not within the scope of this paper to go through a deep examination of all problems encountered in practice but we can make a rapid survey of some of them with which we have been dealing both on a national and an international basis. The already existing format was expected to be dealing in someway with the differences related to the special material. This has consequently brought up, as an immediate result, the definition of new fields, subfields, tags etc., adapted to support the different needs of thi s kind of material in the identification, description and notes blocks. We have been compelled to deal with all aspects and elements required for a correct identification of the edition described, being the hand press typographic technics quite different from the machine ones. This is happened for example for the fingerprint. And again, in order to ensure the possibility of individuating differ ences between variants of one editions, for which is necessary to give a new bibliographic description, from those for which is sufficient to supply the information in a note, to assess that the note field available could be comprehensive also for that kind of information related to the item described, and so on.Moreover, being the environment where the project and the study have been both develo ped, that of an international multilingual union catalogue, also the coded information and the international use blocks have presented some relevant problems.

On the other hand other major issues involved have been those related to the use of character sets and those connected with the need of linking holdings to libraries locations, through thevarious fields where it's requested, or where it has been identified as necessary, to do so.

We must mention that, in addition to the problems of finding out and agree on a common path to be followed, not only by all CERL members participating to the project, but in general by all users of the format in the future, each single participant was also bound, on the side of the HPB database establishment, which was developing in parallel, to match solutions, often not very easy to take, with the different format used in the host database, which is USMARC. This for what concerns international cooperation, and we would dare to say that it has been a quite hard task.

We have in our country the usual acquaintances with the use of UNIMARC as the average of Europeanlibrarians. In fact BNI, our national bibliography, and SBN records are available in UNIMARC. Inboth cases the format is related to current publications. Since the beginning of SBN project ICCU provided, also for older material, a conversion format of SBN records into UNIMARC. We can easily consider that conversion as having been for a long time just a "draft" one, since at the time when it was projected nothing was already decided or existed about the definition of a standard UNIMARC (A) format. The software needed then to be reshaped and redesigned, if not from scratch, at least for all those problematics parts for which it has been found a definite solution in those last months. It has been made a very sound and accurate work, in progress and in line, with parallel developments of CERL and CoBRA projects.In some cases, as for example for what concerns special characters se ts, for which in SBN is provided a very thorough and extended use, we have been bound, considered ISO sets provided for by UNIMARC, to undergoa reduction of functionalities, comparing the possibilities available on our system.

In other cases,as for example for what concerns libraries codes, stricly related with holding locations, we have used the national standard codes which have been developed in a special database, the Libraries Register Directory available on line on SBN system. Taking into account the peculiar requirements arising in CERL cooperative union catalogue, the project has been the input for submitting to the ISO SC46 "ad hoc" Subcommission the proposal of adopting a new work item for the definition of an international standard for libraries codes. Considerations on the first stage of the cooperative projectWe do not want to draw any conclusion at this stage but we have to make some considerations especially for what concerns our experience. The establishment of a database of this kind, where contributions are the results of different cataloguing practices, as well as ofdifferent levels of development of programmes and technologies, in the individual participating institutions, has required the development of a particular comprehensive attitude. All parts involved at different levels of participation, nationally and internationally, have been andstill are sometimes overwhelmed by th e need of finding out the more direct path to reach the ultimate scope of a project in which the end user is the last but not the least target of a chain. We worked very hard for maintaining a continous, fair and equitable balance between realisation of objectives, according both to relieable technical solutions, fundings availability and the need of meeting, if not all, at least the most releva nt requirements of a multilingual and multicultural environment. All this, in order to preserve the peculiarities of the European heritage and recognizing the particular and different traditions of European institutions. The database architecture, considered the special material with which we are dealing and the peculiar multilingual environment where it is developed, if at the very beginning ca used some perplexities, especially since we were running a national union catalogue where the description of the same edition of a work is supposed to be represented by a sole record to which are linked all locations of libraries holding that edition, it is now for us source of sound reconsideration. In fact, a database of this kind, where records produced from differents sources and through the most varied cataloguing practices are not merged at the moment, but kept separately, through the mechanism of "clustering", it is important not for the maintainance and the survival of theindividual cataloguing tradition, which is anyway well kept and preserved at each single national level, but because it is providing the possibility of downloading, for retrocon purposes, all those records whic h are more similar to one's own rules. Further editing activities to make records acceptable according to different national rules will be contained in this way as well as cost will kept at a lower level.

It sounds quite a heavy task but it will gives and has already given in turn some results, at least in developing a very good level of cooperation, in verifying products and projects cooperatively developed at national level and in offering a wide possibility of exchange of experiences. This has proved worthwhile also if in some case we have been compelled to make some renounce. At this stage w e are well aware that agreements reached on an exchange format will not solve all remaining problems connected with different cataloguing practices and related to headings choices and representations, but the work done up to now could be a good start for comparing differences and evaluating the possibilities of future common solutions also in this field. We are confident that new advanced techn ologies will be working on our side to this end and that we will probably be all able in the very near future of easily interfacing our possible differences, that we hope will progressively be reduced in number and variety.

We can then say that the project, to which we are happy of having contributed, has reached its first objective. The HPB database is live and accessible to its members, thanks to the cooperation of all parts involved: 28 libraries from 13 European countries, with at least 11 different languages.

Rome 15 aprile,