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

60th IFLA General Conference - Conference Proceedings - August 21-27, 1994

Quality of a national bibliographic service:
in the steps of John Whytefeld - an "admirable cataloguer"

Philip Bryant
Director UKOLN: The Office for Library and Information Networking,
University of Bath, United Kingdom


This paper focuses primarily on the bibliographic data created for, and used in, a national bibliographic service rather than on the technology used for the manipulation and distribution of the data. Selected aspects of the UK experience are described to provide examples of problems which, it is believed, face any national bibliographic service. 'Quality' of bibliographic records should be conce rned with 'accuracy', 'consistency', 'functionality' and 'timeliness'. Some performance measurement activities in the UK are described and the paper stresses that education and training of those who create bibliographic records must be adequate if national bibliographic needs are to be satisfactorily met. It is at the level of the individual cataloguer that the foundations of a high quality ser vice are laid.


"... an admirable cataloguer in John Whytefeld, a monk who lived during the second half of the fourteenth century ... He gives not only the title of each book and its shelf reference, but also such details as the number of folios in each volume, and the opening words of each book and of each text in each volume that contains ... more than one composition"
Frederick Harris on A Book about books1 published in 1943.

The art or science of cataloguing was practiced long before the birth of the humble John Whytefeld; however, the idea of the catalogue record as a 'packet' of information which could be made available from a central source, and shared, did not really develop until the nineteenth century. There had been little reason for it to do so, but the growth in public education and the early stages of the 'publications explosion' was to change the situation rapidly:

Paul Otlet writing in 1896 stated that:

"The greatest bibliographical effort of the 19th Century ... has been made by the Americans. Voluntary co operation of librarians on the one hand and of booksellers on the other, has given birth to a series of admirable works to which are attached the names of Poole, Fletcher, Cutter, Dewey, Windsor and others. In 1850 in the United States ... there were about 100 libraries containing 5000 volu mes or more. The whole of their collections were judged to be 1 million volumes. In 1890, forty years later, four thousand libraries contained 27 million books."2
This proliferation of libraries and the exponential growth in publications, and hence of library stocks, made it seem obvious to the Americans that the most cost effective approach to the work involved was centralised cataloguing and in 1901 the Library of Congress (LC) centralised catalogue card service was started. I doubt if those who set up this service realised to the full what the long ter m effects of their initiative would be. An embryonic national bibliographic service led to a national cataloguing code. The development of co operation with the United Kingdom was a natural progression and in 1908 the first Anglo American Cataloguing Code3 was published; but then little further of real significance internationally was to happen for over fifty years. Joint revision of the Code did start in the late 1930s, but World War II was soon to interrupt this process.

After the war the Americans went ahead on their own with the production of the ALA/LC rules in 19494, and during the 1950s the British and the Americans continued to revise the Code independently of each other. In 1950 the British National Bibliography (BNB) was launched and also began to provide a card service, though on a more limited scale than that of the Library of Congress. The next most significant international initiative was to occur in 1961 when the International Conference on Cataloguing Principles5 was held in Paris. At the conference Sir Frank Francis spoke of the acceptance and use of catalogue entries made in different countries as a most promising route to improve bibliographic control. This initiative, taken in 1961, together with the development of the Machine Reada ble Cataloguing (MARC) programme, resulted in another key event the IFLA International Meeting of Cataloguing Experts held in Copenhagen in 1969. This meeting was to mark the birth of the IFLA Programme of Universal Bibliographic Control (UBC):

"The purpose of the system is to make universally and promptly available, in a form which is internationally acceptable, basic bibliographic data on all publications issued in all countries. The concept of UBC presupposes the creation of a network made up of constituent national parts, each of which covers a wide range of publishing and library activities all integrated at the international level to form the total system"6
The idea of UBC appears to be a very admirable one, but as Dr. Liebaers, who first gave the name and initials to the programme, wrote in 1980:
"The implementation, however, is another story. The simple aim poses quite different problems depending on whether you work with a computer produced MARC tape of bibliographic descriptions, or try to cope with cataloguing problems in a country which has not yet been able to produce any kind of a national bibliography"7
In this paper I am concerning myself primarily with the production of definitive records for use by other libraries, and not with the technical processes of manipulating and distributing these records, nor with the various forms in which a national bibliography can be made available. While good management of the service and effective computing and telecommunication skills are vital ingredients i n ensuring an excellence of performance, I believe strongly that the basic foundation of 'quality' in any national bibliographic service is dependent on the quality of records, either created by its own cataloguers, or those in libraries and agencies which may share the burden at a national level.

The difficulty lies in arriving at an adequately shared view of what constitutes 'quality', and then in both achieving and sustaining it. Even though a main aim of UBC is to enable economies by encouraging each national bibliographic service, or agency, to be solely responsible for the production of definitive records for its own country's publishing output, a major drawback of centralised catal oguing has been the delay which all too often occurs in creating and delivering records to libraries. At a forum organised by the UK Library Association and the MARC Users' Group (MUG now UBIS: Users of Bibliographic Standards), which I attended in 1988, entitled The future of a national database, a small group of senior members of the UK library and bibliographic community were invited to di scuss "what bibliographic data does the United Kingdom library community require, and how best is it to be provided?" After a very lively debate an attempt was made to reach a consensus with the result that the following statement was issued:

"We recognise that the national bibliographic effort exists on two levels: (a) the creation of the national archival record for which the focus and responsibility lies with the British Library, as advised by the library community; (b) the development of a network of databases from which users can be satisfied for various functions, the co ordination of which may lie with the British Library, but the responsibility for which should be more widely shared"8
The need for the forum, and the issue of this statement, arose from frustration felt by many in the UK at the considerable delay by the then British Library Bibliographic Services (BLBS) in producing records for the national imprint. Delay which impeded good management for libraries and caused many problems for the bibliographic community as a whole. It is good to be able to report that the bac klog of over 40000 uncatalogued items which had built up at the BLBS in the mid 1980s has now been eliminated. A point I wish to make here is that the problems of providing a cost effective service, that meets both national and international needs for high quality records in a timely fashion, have by no means been restricted solely to the UK. In 1990 Hope Clement of the National Library of Canada and Chair of the International MARC Network Committee (IMNC) wrote an article which made this abundantly clea r.9 She was reporting the conclusions of an IMNC survey of the record creation processes of its members within the context of UBC. The survey started with the acquisitions process and then proceeded through the creation of bibliographic products and services, including "costs and measures to combat backlogs and increase timeliness". The conclusions very much reflected the situation in the UK t hat I have just described:

"Human resources and budgetary constraints are seen as the greatest problem in the production and delivery of national bibliographic products and services. These must be coupled with increasing publishing volumes. All national libraries surveyed have had to function in an environment of restraint ... at the same time as the demand for library services has increased, both in volume and complexit y ..."
Problems commonly identified were increasing backlogs, similar to those experienced by the BLBS, due to the growth in publishing and the production of new forms of material; the length and complexity of the bibliographic record creation process; the need to use several complex bibliographic standards and the need to keep up with their constant revisions, and another problem was the delay often experienced in the receipt especially at the legal deposit stage of the national publishing output. This problem in fact often contributing to time delays in record production which will be unfairly attributed to the national bibliographic service when the real fault lies with the publishers concerned. In relation to the UK Cataloguing in Publication (CIP) programme, research with which I have been involved at Bath for over a decade, has demonstrated that it is far more important to encourage the major publishers to provide preliminary details of all their publications than to spend too much effort endeavouring to make sure that every small publisher joins the programme. Although the UK situation has now improved greatly, a few years ago we discovered that some of the biggest pu blishers were only providing data for some 50% 60% of their total imprint.

The importance of 'timeliness' in bibliographic record creation became especially obvious at the end of the 1970s with the development of mini computer based integrated library systems (ILS) and their new modules for acquisitions which meant that libraries required records, not only at the time of 'cataloguing' but also at the 'ordering' stage.

The British Library has taken a number of major initiatives over the past decade, in order to maintain and improve the national bibliographic service during a period of financial restraint. Probably one of the most important of these has been to recognise the importance of performance measurement and I am going to devote most of the second half of this paper to describing some of the measurement s which the unit I direct at Bath University has either introduced, or been involved with, over the past fifteen years. I will also look at their success, or otherwise, in improving the quality of bibliographic records provided by the now British Library National Bibliographic Service (BLBS). Lack of time only allows me to give brief descriptions of projects, and I will not be discussing the na ture of performance measurement. If you want more detail I wrote an article 'Performance measures for national bibliographic services' which appeared in the second issue of the journal Alexandria10 in 1989. I said in that article:

"The question which always has to be addressed is: Are the majority of users of the service obtaining the maximum benefit from the way the resources are being allocated? The problem for many librarians is that they believe that the nature of librarianship is such that one can never tell when a shortfall in service to a minority, or to an individual, may have serious adverse ... consequences for learning or human achievement. Such idealism and altruism of intent have to be commended, but it is not a sound basis for solving the practical, economic and political problems that face library administrators and policy makers, whether at the local, national, or international level."
I happen to believe that the main imperative for a national bibliographic service is to produce bibliographic records so that these records are available when they are most needed by users whether these be librarians, export booksellers or others. Nevertheless, this is not a view that everyone agrees with; there are those who believe that fulness of bibliographic description, heading structur e, and strict adherence to internationally agreed cataloguing conventions should be recognised by the national bibliographic service as having greater priority than 'timeliness'. It will be stated that the national service has the responsibility for providing the definitive archival records for the national imprint and that posterity demands that this unique role should not be sacrificed for sho rt term advantage. However, as I hope I have already made clear, we have to live in the real world and, in any case, are the hypotheses on which such arguments are based really justified by the evidence? I would suggest that several studies with which I have been involved tend to demonstrate the opposite! The most recent project I completed was one I undertook at Cambridge University Library i n 1992 of the use and understanding of the library catalogues by postgraduate students, academic and research staff and external researchers.11 I was prompted to undertake this survey because of the findings of a project I had undertaken in 1986 in the British Library's main reading room at the British Museum.12

That study had been specifically concerned with users' understanding of catalogue entries for serials; however the findings gave clear evidence of users' general lack of understanding of the information provided in the entries and was indeed a cause for concern. The results demonstrated the need for producers of catalogues and cataloguing codes to be far more aware of the people who use what th ey provide. The Cambridge survey allowed the opportunity, not only to study understanding of catalogues by academics and researchers in a major academic research library, but also to examine the 'posterity' argument that more detail is required as users look for older items. In the British Library study, 225 in depth interviews were undertaken and there were nearly 2000 instances of confusion, or lack of understanding, of cataloguing conventions i.e. abbreviations, jargon and ISBD punctuation (e.g.v., 'cover title', . , etc.). At Cambridge University 209 interviews resulted in 1501 instances of users uncertainty or misunderstanding of catalogue entry details, plus a further 131 cases of uncertainty about the headings or 'labels' used in the on line public access catalogue (OPAC) to d istinguish the various areas of catalogue data (e.g. 'uniform title'). Finally no relationship was demonstrated between the wish for more data in catalogue entries and the requirement for items published in different centuries; although, as I pointed out in my report, the needs of many bibliographers, historians and literature scholars may well demand more data when one goes back to the period of hand printing.

'Quality has all too often been equated with fulness of bibliographic description, whereas the development of OPACs and networked access to them nationally and internationally has resulted in a demand for far more information about the content of books and documents to be provided. Whatever views exist about the necessary level of detail in a bibliographic record many would agree that the 'quali ty' of records should primarily be concerned with: 'accuracy', 'consistency', 'timeliness' or 'currency', and 'functionality' i.e. is the record appropriate for the tasks it is expected to fulfil for the user. I would like to deal with 'timeliness' first. In 1980 at Bath we started monitoring the currency and coverage of UKMARC records on a continuing basis and are still doing so today. Once a month, on a randomly chosen day and at a randomly chosen time, 12 libraries in the UK (randomly chosen from the public library and academic library sectors) record the titles of 10 newly acquired titles which are within the scope of the BNB MARC record service, i.e. items with a UK imprint published since 1974. The titles are selected just as books proceed to cataloguing from the acquisitions section. (For the last few years a second sample is taken at the time that items are ordered). The libraries send the details to Bath, where they are checked against the BNB MARC database to see if a record would have been available on the day the sample was taken. At the end of twelve months the percentage of 'hits' (i.e. records that would have been available) are conflated to produce the 'h it rate'. A fuller description of this survey is given by my colleague Ann Chapman in the April 1992 issue of the Library Association Record.13

The BLBS (now the BLNBS) has taken a keen interest in this survey since the earliest days and it is, to my knowledge, the only external performance measurement of its type of a national bibliographic service anywhere. In 1982, in an attempt to improve the 'hit rate', the British Library greatly developed its existing CIP programme, but this failed to achieve any marked improvement. For several years the 'hit rate' remained at around 62% 64% and by the mid 1980s the situation regarding the backlog I mentioned earlier had reached crisis proportions. In 1987 a consultation paper Currency with Coverage14 was issued by the British Library which, among other initiatives, provided for a lower level of bibliographic description for certain categories of material in order to allow resources to be reallocated to speed up production of records. This policy was adopted and achieved a marked improvement in 'timeliness', but at the expense of attracting criticism from many quarters. Great concern was expressed, not only by librarians, but also by many in the book trade.

National bibliographic services should always be aware of the need to study the impact of changes of policy; however, it was the UK MARC Users Group which commissioned the CBM in 1989 to undertake a study of15 of the impact of the changes in BNB MARC records introduced as a result of the Currency with Coverage policy. Nevertheless, despite the criticism directed at the BLNBS, previous research projects undertaken by the Bath team16 have shown that, for end users of catalogues, records with considerably less data than AACR2 Level 3 will perform very effectively. As was demonstrated by both the Cambridge and British Library surveys already referred to, of far greater importance to the user than fulness of bibliographic description is the visually clear and unambiguous presentation of da ta, whether on cards, pages, microfilm or computer screens. 'Accuracy' and 'consistency' are of primary importance and are obvious candidates for internal performance measurement undertaken by national bibliographic services themselves; however, they can also very easily be monitored externally. UKOLN is just about to undertake its second continuous performance measurement exercise on behalf o f the BLNBS by regular checking of the bibliographic records contributed by the libraries participating in the Copyright Libraries Shared Cataloguing Programme.

Even though the expanded CIP programme and the Currency with Coverage initiatives failed to achieve all that had been expected of them, the BLNBS was not to be deterred in its efforts to improve performance. Late in 1988 it was recognised that the British Library would need assistance if the National Bibliographic Service was to be able to provide the records for the national imprint which were required by the library and bibliographic community. In February 1989 the British Library announced that: "The Librarians of the six UK copyright deposit libraries are planning a co operative programme in the creation of bibliographic records of British books for the British Library National Bibliographic Service ..."17 The aims of the Programme were "to speed up the flow of records to the Nat ional Bibliographic Service, thus improving the cost effective use of the service in all libraries, and the currency of cataloguing information ..." A report on the first 'pilot project was published in 1993.18 A full programme is now in operation. The BLNBS contributes 70% of the records for the national imprint and the five Copyright Library contribute 6% each. The programme had an influen ce on the BLNBS's previous policies in a number of ways. One example has been the decision to reintroduce Library of Congress Subject Headings (LCSH) which had been dropped as a result of Currency with Coverage. The Copyright Libraries pressed hard for a reversal of the decision.

Before I finish I must make a brief reference to a project undertaken by the Bath Centre in association with BLCMP (the UK's largest cataloguing co operative) to examine the fourth aspect of quality in bibliographic records 'functionality'. In 1992 the CBM approached BLCMP to see if it would be prepared to co operate in a study which would extend the 'performance measurement' work of the Centr e. The aim of the project was to examine why changes were made to records by member libraries and how important were the reasons! In fact CBM's approach coincided with BLCMP's own wish to evaluate the quality of records in its own 10,000,000 record database, which includes the complete BNB MARC files. The study was undertaken by my colleague Ann Chapman and her report was submitted in 1993.19 It certainly demonstrated the need for further research in this area!

I said earlier that the foundation of 'quality' in any national bibliographic service is based on the quality of the records created for it by its cataloguers; therefore, I deplore the fact that 'cataloguing' no longer figures prominently in the curricula of departments of library and information science. I have found that it is not only the UK which has found it increasingly difficult to attra ct applicants with the necessary training and experience! In addition, we must educate the 'standards' makers and the trainers to recognise that 'quality' does not lie in the strict observance of arcane principles, but in accurate, consistent and understandable practice.


1. HARRISON, Frederick. A book about books. London: John Murray, 1943

2. OTLET, Paul. Creation of a Universal Bibliographic Repository: a preliminary note. In: International organisation and dissemination of knowledge: selected essays of Paul Otlet. Translated and edited with an introduction by W. Boyd Rayward. Amsterdam: Elsevier, 1990.

3. Cataloguing rules, author and title entries. Compiled by committees of the Library Association and of the American Library Association. English edition. London: Library Association, 1908.

4. American Library Association. Cataloguing rules for author and title entries. 2nd edition. Edited by Clara Beetle, Chicago : ALA, 1949.

5. International Conference on Cataloguing Principles, Paris 1961. Statement of principles. Annotated edition, with commentary and examples by Eva Verona (et al.). London: IFLA, 1971.

6. ANDERSON, Dorothy. UBC : a survey of Universal Bibliographic Control. London: IFLA International Office for UBC, 1982. (IFLA International Office for UBC Occasional Papers No. 10)

7. LIEBAERS, Herman. Mostly in the line of duty: thirty years with books. The Hague : Martin Nijhoff, 1980.

8. MARC Users' Group/Library Association Forum on the future of a national database. [A report]. London: MARC Users' Group/Library Association, 1988.

9. CLEMENT, Hope. National Bibliographic Agencies cataloguing survey. International Cataloguing and Bibliographic Control, 19(1) January/March 1990, 6 10

10. BRYANT, Philip. Performance measures for national bibliographic services. Alexandria, 1(2) 1989, 27 35

11. BRYANT, Philip. Use and understanding of the library catalogues in Cambridge University Library : a survey. Bath: UKOLN: The Office for Library and Information Networking, October 1993. (British Library R& D Report 6124)

12. BRYANT, Philip. Bibliographic access to serials: a study for the British Library. Serials 1(3) November 1988, 41 46

13. CHAPMAN, Ann. Why MARC surveys are still a hot bibliographic currency. Library Association Record, 94(4) April 1992, 247 254

14. Currency with coverage. Consultative paper. London: British Library Bibliographic Services, 1987.

15. Dempsey, Lorcan. Currency with coverage a survey report. London: MARC Users' Group, 1989

16. Seal, Alan, et al. Full and short entry catalogues: library needs and uses. Aldershot: Gower Publishing Company, 1982

17. British Library Bibliographic Services Newsletter, 48, February 1989, 2.

18. Shared cataloguing. Report to the principals of the Copyright Library Shared Cataloguing Project Steering Group. Boston Spa: British Library. 1993. (NBS Occasional Publications 1)

19. CHAPMAN, Ann. Quality of bibliographic records in a shared cataloguing database : a case study using the BLCMP database. Bath : Centre for Bibliographic Management, March 1993. (British Library R&D Report 6120)