Given the fact that the Library must address the complexities which the move to St. Pancras will involve, and that it will have to plan for the future with diminishing resources, I have been giving some thought to areas within the Division - in spite of ambiguities and uncertainties created by restructuring - where constructive effort, initiated by the Development & Systems Office (D & S O), seems essential if we are to balance resources and services. Certain of these areas I have covered in other papers, so some specific problems are not here dealt with in detail. I think it is important to establish the principle that the D & S O has a crucial role to play in drawing attention to the problems which must be solved if we are to continue to offer our users the services they have come to expect from a national library. We do not exist simply to satisfy ourselves that the institution functions efficiently: it is those who use our resources, and depend on them, that we should seek to satisfy. I think that one of the reasons for the depressed state of morale (evident in all grades) stems from a lack of conviction that whatever is said about "research" and the "needs of users" there is little evidence that those responsible for wide-ranging decisions affecting staff and users alike are being taken with any accurate understanding of what it is that users need and expect. If we wish to emphasize the importance of the Library as a research base for scholarly and imaginative enquiry, then we would be well advised to make an effort to comprehend the grammar of research. Apart from individuals within the institution engaged on research (an important but seldom-consulted community) which is of primary value for our users, corporate utterances about the importance of the "research" coefficient seem to me empty rhetoric.
The Library's centre is not 2 Sheraton Street:[i] it is the Reading Room. That is the place where most enquiries start, whether in person or by post. At the heart of this centre is the most heavily used catalogue in the world, which is why it is in constant need of physical maintenance. In terms of concentrated use the nearest equivalent I know is the card catalogue of the New York Public Library which managed to survive for a century before its "rehabilitation" became necessary.[ii] But it is no longer the definitive inventory of our collections, and researchers have had to learn strategies for finding books quite different from those needed to use GK. The provision, in recent years, of on-line services via the Enquiry Desk has added yet another complexity to the art of information retrieval. And, for reasons of space, it has been necessary to banish from public view important resources such as the printed catalogue of the New York Public Library (housed in Circle Ground East where are to be found numerous other reference works of fundamental importance).
Since the General Catalogue now has to be supplemented by microfiche catalogues and on-line catalogues for certain kinds of searches it is important that the Enquiry Desk staff should be thoroughly trained in on-line searching of various databases so that proper advantage can be taken of a service which is bound to grow as ever more information is converted to machine-readable form. When Mary Hurworth retires that will leave one member of staff in any way competent to assist readers and deal with postal enquiries involving on-line catalogues. That seems to me quite inadequate. If we envisage a situation within a mere six years which will require terminal access to information as the norm rather than the exception, then we must start planning for that now.
One suggestion I have mentioned to David Martin[iii] is that, with help from his staff, we should now be developing a microcomputer "front end" [interpreter] to enable readers to interrogate a machine-readable file with a simple unambiguous protocol.[iv] It should not be a costly exercise, but it seems to me essential that readers are eased into the use of computers, and not expected to cope with unfamiliar machines and procedures overnight. A converted GK available on-line would certainly require such an interpreter. The file which might form a very useful experiment for this is the DPB catalogue. I wonder how many readers even know it exists?
I think we should consider seriously the preparation of a video [no more than 30 minutes] which could be shown on a regular basis to readers who want to know more about the various sources of information now available in the Reading Room. It would also provide an opportunity to give readers a glimpse of what is being planned. It should include a few specific searches on various BLAISE LINE files as well as other relevant databases held on DIALOG.
In various documents I have read recently there is repeated reference to the needs of our users; yet they are, I believe, imperfectly understood. I suggest that the D & S O should appoint someone, preferably with experience of research methods, to investigate the needs and expectations of our readers. This should not be in the form of a questionnaire-survey, and should address the very different approaches adopted in different disciplines. I think that if this were done coincident with the suggested video we would find that readers would be prepared to cooperate. If we want their help, we must be willing to offer them help in return. The study could be carried out in about six months.
It seems to me obvious that the resources which are increasingly being devoted to processing the Library's acquisitions, whether by copyright or purchase, have reached an unacceptable level. By "processing" I include the following:
For a large number of books acquired by purchase I think investigation would reveal that where the cost of the book is less than £10 the processing costs are of the order of x3. The only way of reducing these costs is to determine priorities in (2) and (3), since there very few significant economies which can be made in (1).
It is probably fair to say that purchases reflect the Library's traditional policy of collecting materials of immediate relevance (research materials, primary sources, contemporary literature) and materials which might be defined as sources for posterity. Since the new press-marking scheme to be introduced in 1986 makes an attempt to accommodate the distinction between high- use and low-use items, then I think it is time to consider cataloguing in a similar light. What is achieved by devoting scarce resources to full cataloguing of an item which may sit on the shelves for ten years without being requested? And what is the point in adding further resources to giving it first binding if it is a paperback? In the last five years we have spent literally millions of pounds first-binding materials of very low use (some of it can be conveniently observed in NE1 and NE2), a large number of which reveal on inspection that they have never been called for. Waste of resources on this scale can not be allowed to continue, especially since the selection of materials for first binding is carried out by clerical staff with no knowledge of the languages or subject matter concerned.
Low use items should be given the minimum cataloguing necessary to find it on a machine-readable file. This would include minimal necessary subfields in 008 - year, language, country; 087 - pressmark; 245.1x [to get title listings]; 260 - $a/$b/$c. No 100/110 would be given since author statements (if present) are searchable in 245 and could be put in square brackets if occurring elsewhere in the work. No subject cataloguing would be included. The omission of headings and subject classification - the two elements in a bibliographical record which consume the largest amount of time - should make it possible to encode for keyboarding an "average" book in ten minutes. These records could be included on the DPB file with a flag indicating the status of the record as "provisional". Since every item is coded for title output, they could even be included in the fiche catalogue, providing that readers are warned that such records exist. If, for whatever reason, they are not found, then a simple on-line search could establish quickly whether or not it had been catalogued. Evidence subsequently revealing that the "low-use" item was, in fact, in demand would then justify upgrading the record.
Whether or not similar procedures could be adopted for copyright material should be investigated, even though this would involve alteration to the processes involved in the receipt, cataloguing, and press-marking. It should be possible to establish guidelines in BSD for certain categories of material to be given minimal cataloguing (as for purchases) and the book so marked before received for placing. This would involve alteration of the current validation for MARC records since a number of mandatory fields and subfields will be omitted in these records. A further complication would be the currency of the name authority file if an item by an author not current on the file were given the minimum standard. Ways of dealing with this problem should be investigated.
The one aspect of the Strategic Plan[v] which worries me is the notion of charging readers for value-added services, such as access to on-line catalogues. The danger is that, having established the notion that the Library is an "information store" rather than a collection of printed and manuscript materials, we shall be seen to be charging for information which our readers expect us to provide. The provision of access to a database seems to me not different in substance to providing access to the National Union Catalog or any other substantial catalogue. Providing that we can effect economies in processing, then it should be policy to provide computer services in exactly the same way that we provide a wide range of very expensive printed information-sources. The value-added principle could, for example, discriminate between the Encyclopedia Britannica [which we acquire on copyright deposit] provided "free", but a value- added charge being made for NUC [which we purchased]. The administration of such a policy would be ludicrously cumbersome and expensive. Are we prepared to face the criticism which would be levelled at us if we seriously pursue this policy?
It would mean, for example, that while permitting "free" access to the eighteen guard-books containing the First Line Index of English Verse in the Student's Room, which cost a great deal of staff time to compile and arrange, we would have to make a charge for its transformation into machine-readable form.[vi] I have elsewhere pointed out the fact that the General Catalogue has cost more to prepare, print, incorporate and maintain than all the books purchased since 1841. It is by far the most expensive "object" in the Library: yet we do not, I hope, contemplate charging for access to it. Even since 1973, when the British Library was created, it has cost more than BNB and the computer services provided by RHM.[vii] And if it is converted to machine-readable form by 1991, are we then proposing to charge for access to the information which its transformation can be expected to yield?
Libraries shifted in the 1970's from manual cataloguing [always provided free to users] to machine-readable cataloguing, encouraged by the belief that economies could be effected by so doing, and that computers would enable a new dawn in the complicated business of retrieving information. History has shown, however, that the promised economies have not resulted, and so the massively expensive transformation of bibliographic data is an "added-cost" factor, which somehow has to be paid for. OCLC is a splendid paradigm: for while it has succeeded in capturing the largest proportion of the "shared cataloguing" market in America [thereby making it a successful capitalist venture in an essentially non-capitalist market - no mean feat!] it has increased the burden on subscribers by its liberal policy of allowing them to input virtually any level of record. The result is that while the system holds over 200 million records, only about 12 million are bibliographically discrete. It is just this possibility of ambiguity which ESTC has sought, from the beginning, to avoid. Given the nature of the book [on which I have spent enough rhetoric in the past few years] it is no surprise to me that this should be so. The degressive principle has always applied in bibliography: the less information you provide, the easier it is to determine what you can "match"; the more information you provide, the less easy "matching" becomes. The Bibliographical Society's STC took about five years to compile; its meticulous revision has taken ten times that long.
Between 1875 and 1900 the senior staff of the three library Departments in the British Museum compiled and proof-read some three million bibliographic records for printed and manuscript materials. Included in that estimated total are the records which appeared in GK1, as well as special catalogues as follows:
Between 1875 and 1900 the average "curatorial" complement of the three Departments was supported by clerical and assistant staff. The contrast between the bibliographic information provided in that period and what has been achieved in the first ten years of the British Library is dramatic, if we consider the average number of curatorial and clerical staff involved in cataloguing since 1973. If the British Museum Library had to halt GK2 because productivity was falling behind intake, how can we possibly cope with the current intake given that AACR2 and PRECIS is more complicated and time- consuming by an order of magnitude? There must be an equation in which the cost of processing a book has an averaged relationship to the cost of its acquisition. Non-copyright libraries (i.e. those which purchase everything) are already moving in the direction of establishing a ratio between cost and processing which is affordable. Given the pressure to reduce costs and improve efficiency, it seems inevitable that we must do likewise. The alternative is to reduce acquisition, a course which hardly conforms to the tradition we have inherited, and one which our readers would neither expect nor welcome.
We will never please the fastidious bibliographer, whose habit of mind is to distrust all descriptions. The Library's tradition, on the other hand, has always been to demonstrate a concern for the "average" reader and student [Pollard's prefaces to the short-title catalogues], and to supplement information available in the General Catalogue by intelligently conceived sub-sets. Bullen's catalogue of pre-1641 English books is a historic landmark. Would we, today, be prepared to regard it as a value-added source, subject to charges for consultation? Surely the sensible compromise is to continue to acquire books on an encyclopedic scale, make freely available new methods of access to such books as we possess and acquire, and cut down the amount of information provided in the bibliographic record? Surely, as well, should we not seize the opportunities provided by the computer to simplify the structure of the bibliographic record and cease using it as a typesetting machine to produce a simulacrum of a printed catalogue?
There are a few people still left in the Division who share my concern about the Library's "future", and who have an informed understanding of its history. I think it would be an extremely useful exercise if a small group could be assembled - away from the Library - to discuss, in relaxed surroundings, some of the issues raised in this essay. Such a group might consist of the following: yourself, myself, Ian Gibb, Sarah Tyacke, Phil Harris, Richard Christophers, Mirjam Foot, Mervyn Jannetta, Brian Holt. It might be a nice gesture to invite an elder statesman like Don Richnell as an observer.[viii] The objective of such a gathering would not be to formulate a plan, but to give those of us who are deeply concerned about the Library's future an opportunity to clarify in our own minds possible, and practicable, solutions to a whole range of problems.
I will conclude with a detail which few of us have probably considered. In all the paperwork and staff effort which has been expended on the Strategic Plan [a very expensive exercise], has anyone counted the cost of the simplest of its implementations - printing the stationery? My own estimate is that for the two Divisions most concerned the cost will exceed £100,000. If we are prepared to spend that sum on stationery, can we not spare a modest sum to carry out some of the essential enquiries suggested above?
But even more worrying than the expenditure on stationery required to bestow legitimacy upon a re-organization of the Library is the creeping paralysis which seems to have overtaken us: since any decision, in a period of rapid change (to which a large institution finds it difficult to adapt) coupled with uncertainty of purpose (the ambiguities of the Strategic Plan), runs the risk of being wrong, it now seems to be accepted doctrine that doing nothing is a fairly safe course of action. As long as discussion is seen to be taking place, then not much harm can be done. I sometimes wonder where the world would be if man had always pursued the ablative principle.
I have repeatedly urged the analogy between the study of books, and the texts they contain, and the study of archaeological remains. Both are witnesses: but the validity of their testimony depends upon interpretative skills. The exercise of these skills is the province of the scholar: the provision of the testimony is the responsibility of the custodian/librarian. What is not preserved cannot be studied. Historically, the scholar and the librarian were little different: both understood the importance of intellectual husbandry and management of resources, and both understood the imperatives of enquiry into the nature of things, their whereby's and their wherefore's. I suspect that most users of the Library regard their quest for information as a struggle rather than an adventure. There is so much they are not told. Yet we support huge resources of information. This summer ESTC held a regular series of demonstrations to inform readers of the benefits which that project could bestow upon a wide variety of eighteenth-century historical disciplines. Are there comparable demonstrations for other files available on BLAISE LINE?
Libraries are part of a nation's "invisible assets". They are not subject to the cost-accountancy of importing bananas from a Caribbean island, for the simple reason that no one can predict the consequences of access to knowledge, whereas everyone knows the consequences of access to too many bananas.
[i] The administrative headquarters of the British Library.
[ii] The end result of this was the publication of the massive card catalogue (author/subject) in 800 volumes.
[iii] Head of computing services in the Bibliographical Services Division.
[iv] As the Z39 protocol in recent years seeks to do.
[v] By the 1980s it had become standard practice in Government controlled institutions to ‘publish’ strategic plans. These absorbed considerable human resources and were seldom (if ever) followed.
[vi] I wrote a simple program to accomplish this. Hilton Kelliher was the curator designated as responsible for creating the machine-readable records. My system was replaced by a far more complex one. The task has still not been completed.
[vii] Rank Hovis McDougall, the company which owns the computer system use by the British Library since 1974.
[viii] The first Director General of the Reference Division.