By happy accident the development of electronics in the transformation of a print-based environment for research and enquiry into the historical process exactly parallels my career in universities, commerce and librarianship. I say this because some of you might infer from some of what I have to say today that I represent a perspective slightly less enthusiastic than the views commonly put forward by the architects of the New Jerusalem.
I first became involved with electronics in 1956 when, at the University of Toronto, I persuaded the new Professor of Anglo Saxon, Jess Bessinger, that there might be benefits in converting Beowulf into electronic form using the university's Univac computer. Such an enterprise would, if successful, assist grammatical, syntactical and lexical analysis and obvious tools would be a concordance, evidence for internal assonance and an accurate word frequency index. After painful experiments conducted at Yorktown Heights grappling with the peculiar problems presented by both the alphabet and the structure of the poetic sentence solutions were eventually found but it took ten years to accomplish the task.
In 1964 I joined the staff at Leeds University where I started the Dictionary of Tudor English an enterprise that never succeeded because the proprietors of the original slips for the OED could not be persuaded to part with them. Nevertheless, in looking back at the proposals I can see the application of principles which have borne fruit in the OED on CDROM. 1964 was, in many respects, a year in which all the signs were there to be seen for the ways in which the computer would change not only manufacturing processes but the very objects on which knowledge depends and the way in which libraries handle those objects. The first printed book typeset on a computer; the IBM Selectric typewriter which proclaimed itself as a "word-processor"; the first report from a group of visionaries at the Library of Congress on the proposed standard for tagging bibliographical files christened MARC [two of the most important players were Henriette Avram and Fred Kilgour founder of OCLC]; the introduction of computers in controlling photographic processes so critical in offset lithography which was rapidly replacing hot metal processes in newspapers and printing companies. In 1964 Robert Willis saw technology somewhat differently from many of his colleagues in the NGA and wrote: “Unless we are prepared to recognise that film taking the place of type falls within the category of a compositor's work then other people will be recruited to do that work with consequent redundancy of our own members. It is folly in the extreme to attempt to fight the new innovations which are taking place." [Graphical Journal, June, 1964]
To the best of my knowledge only one person involved in library development in Britain understood what was happening in 1964: that person was John Jolliffe who began addressing the problem of how to convert the British Museum's General Catalogue to machine- readable form - a task that was to take twenty five years to mature. Jolliffe estimated that with adequate financial support the task could be done in five years. Less remote from ordinary life, but of long-term significance for the future of information transfer, 1964 saw the completion of NASA's Mercury programme and the start of Gemini. A key figure in the Apollo moon programme was Hank Epstein who went on to design the bibliographical system at Stanford known as BALLOTS, the precursor of RLIN. Hank came to the June Conference in 1976 which inaugurated the ESTC and it was from him that I learned what I needed in order to bring that project to fruition.
When, in 1983, the link between London and Stanford was established enabling cataloguers direct access to the electronic file communication between the editorial teams in London and America was accomplished using email on the Arpanet and Bitnet networks.
By 1986 it was clear to me that a wave was gathering momentum and I suggested to the new Director General of the British Library, Mike Smethurst, that staff in the Library should be made more positively aware of what the future held and started planning the 1987 Microcomputer Symposium which was held in Chancellor's Hall in London University. At that Symposium staff in the British Library witnessed for the first time the use of microcomputers to control management information; usage of the collections based on reader ticket analysis; the Hypercard system introduced by Apple and solutions to the intractable character set problem (which still frustrates libraries with large collections in non-roman alphabets); Hank Epstein's Mitinet, the first portable MARC-based system for libraries; the capture of digitised images of various sorts of library materials using a high resolution flatbed scanner developed by Ricoh in 1985 for the Air and Space Museum in Washington; and the prototype version of the British Library's OPAC using a BBC educational computer [the first OPACs introduced in the Reading Room were, you may remember, Acorn machines].
Between 1987 and 1990 I undertook the arduous task of training over 300 staff in Bloomsbury for the arrival of computers in all sections within the British Library including library assistants. The sessions, with small groups of four or five, lasted for a day and covered the essentials. I think it would have been quite pointless to start introducing staff to automation without these propaedeutic sessions. By 1990 my full-time responsibilities at the British Library were coming to an end, but there was still one challenge I wanted to undertake: to test the potential of the growing number of library catalogues and large cooperative databases that were becoming available on the Internet. In March 1990 I began a twice-weekly "surgery" for readers in a small room off the Round Reading Room. They became known as the Research Seminars and continued until September 1994. In total the seminars provided positive assistance to over 500 research projects. It remains, I believe, the only large-scale exercise documenting both the extraordinary variety of topics that people research and how they proceed. It also clarified my own perception of the very real difficulties which inexperienced users encounter when searching electronic catalogues.
In July 1990 I took up a quite new challenge: that of transforming the somewhat traditional curriculum in the three broad areas covered by the University College School: library studies; archives and records administration; and information management. Introducing change was painful, often resisted, and slow. What university administrators find it difficult to accept in disciplines which change rapidly is the absolute necessity to respond swiftly. After all, we have our students for one year only, and into that we must now accommodate not only the traditional but the new. How we are to continue to expand the basic curriculum and still teach basic principles of how information is handled is a question of some urgency. I will return to this.
It is often said that the problem with modern computers is that they represent solutions looking for problems - indeed, most software marketed on the Internet falls prey to this principle, being remarkable only for providing solutions to non-existent problems. The drift from analogue to digital in preserving the countless millions of books and documents in the world's libraries and archives has, in my view, reached a point where we need seriously to consider the alternatives. Do we either want or need to afford the expensive conversion of the materials on which history is based? The question is, indeed, a "tricky" one, suggesting perhaps why the organisers of this conference elected to give it the title "New Tricks?" In order to answer it we must first decide what it is we propose for the originals once they have been converted. There are, it seems to me, two possible scenarios. We could convert and discard; we could convert and retain, committing ourselves to a mixed-resource future in which both printed and manuscript sources continue to be stored but with an increasing proportion of exclusively electronic sources being made available both locally and remotely.
I doubt that the first choice is likely to be taken seriously by either librarians or those who use libraries. There are still too many questions about our paper heritage that cannot be answered by surrogates, however faithfully reproduced. But the second choice brings with it questions difficult to answer because we have a very imperfect understanding, as educators of librarians, of the needs and skills of those who use libraries. As a friend put it to me recently: we understand the problems associated with the providers of information; we understand the technology of information transfer; what we don't understand is the user of information. That, I suggest, should give us pause for thought.
Michael Lesk, who is Head of Computer Research at Bellcore in Morristown New Jersey, has been at the forefront of developments in communication technology for over a decade. He also happens to be someone keenly interested in research libraries and the ways in which they should respond to both the promises and the threats posed by automation. The electronic library is a concept which particularly intrigues him and he has just given me a copy of the final draft of his book which is engagingly titled Books, Bytes and Bucks: Practical Digital Libraries. If it is published by Oxford I suspect they will insist on a slightly less trendy title. Lesk covers all the important issues which must be faced if digital libraries are to be practical - that is to say effective and affordable - with one exception. It is an important exception since it is precisely the theme of this conference!
While Lesk asks a large number of pertinent questions - to which there few reliable answers - he is not without the historical perspective to observe that while we may make significant gains there are bound to be significant losses. One of the most depressing developments is, perhaps, the inexorable drift towards the search for knowledge being dominated by the entertainment industry. It is, after all, the industry which has the financial muscle and a clear understanding of marketing entertainment. We are not too far from this in the way distance learning is currently marketed as both "efficient" and "fun".
Unfortunately, "fun" electronic packages with multimedia effects are expensive to produce - much more expensive than the traditional book. The benefits derive, of course, from huge savings in distribution. But the price we may have to pay for Universal Access to Publications (one of IFLA's core programmes) is that all information will be sanitised, or worse still homogenised.
When I introduced digitisation into the British Library in 1993 there were few on the staff who had the lightest comprehension of either the technology or its potential. For a time there was a flurry of activity: there were a couple of terminals in the Reading Room which displayed a random collection of pretty images mostly from the Department of Manuscripts; there was extraordinary excitement about the Beowulf Project which promised a new life for one of our most important literary manuscripts; there was the project to digitise newspapers from the Burney collection. What has happened to these? The pretty images disappeared after a few months and have not been seen since. The digitised newspaper images scanned on a Mekel 400 camera have been parked on phase-change optical discs and the drives are no longer manufactured. The Beowulf Project has gone quiet, but there is some hope that the results will be available in some form on CDROM - which will be quite an achievement considering that each page of the manuscript was photographed five times and that each image is a 21MB JPEG. You can do your own arithmetic on the size of the image file! At a conservative estimate the project has cost £1000 per page to date. This is not to deny the relevance of forensic photography, using sophisticated cameras like the Kontron: but their use may be better employed in on-the-fly applications than in building massive image archives to which access is difficult and inevitably slow. Have you ever tried to retrieve a 21 MB file at prime time on the Net?
Let us suppose that enlightened government policy combined with the enthusiastic support of our schools and universities succeeds in bringing about electronic libraries, what will those involved in educating librarians have to consider as essential components in a curriculum? We should, in my opinion, provide our students with:
All of this, it goes without saying, in addition to the skills we have been teaching for some time: management theory; interpersonal skills; database theory; cataloguing; indexing; classification; and, if you are lucky, some understanding of library history and what used to be called historical bibliography. With semesterisation this all has to be accomplished in two 12 week terms.
If schools of librarianship cannot accomplish such demanding requirements, then it follows that libraries will be forced to carry out a systematic programme of in-house training designed to satisfy their particular needs. But who will do the training? Another cadre of elite managers? And what of the users who reasonably expect their library to assist them in finding what they need?
It has always seemed to me that preservation of the books and documents which are our memory represents more of an intellectual than a technological problem. Archivists have shown that retention of more than a fraction of the paper on which we record administrative history is impractical and unaffordable. Libraries may be forced to adopt a similar policy.
As it is, our society discards unwanted print at a frightening rate: in Britain some ten million newspapers and magazines every day; an unknown quantity of freely available catalogues and leaflets; newsletters and notices produced for organisations. In electronic form most of the daily inundation of junk mail accumulating on university servers gets deleted unread, and as the Net becomes more like a newspaper than a network supporting serious discussion the discard rate inevitably rises. The cost of all this unwanted "noise" is prodigious but is, at least in the United States, paid for without question. U.S. government policy is clear on this: by the year 2000 every university, college, school and public library will be connected to the Net, representing a population base of about thirty million. With such a vast potential market does anyone doubt that commerce will take advantage? You can withold your telephone number but email addresses are public and institutions like universities make them known. Publishers, for example, no longer need to target their potential customers using expensive printed publicity.
If we are serious about electronic libraries, not necessarily as alternatives to traditional libraries but rather complementing them, then the following issues must be addressed:
Twenty questions are quite enough, though I could have extended the list by raising other issues pertinent to our deliberations: legal issues concerning intellectual property; linear versus hypertext; policing fraud and illegal copying; protecting privacy.
I fear that we are going to need a bagful of "tricks" to deal with most of these questions. But there is one further question which I put to you: if electronic libraries are the way ahead, what are we going to do with all our institutions which have traditionally acquired, maintained and provided access to that rich memory represented in their collections of books; manuscripts; records; pictures; photographs;
recordings; videos? It has been estimated that the total world stock of printed, manuscript and electronic sources exceeds the number of souls that have lived on this planet since the age of Pericles - in spite of colossal losses caused by natural and man-made disasters. For the ancient world the merest fraction survives; for modern times, since the sixteenth century when the printing press multiplied the distribution of information by a factor of 500, the survival rate of what has been printed is less than 1%. Mass distribution of print was made possible by the development of mechanical printing, yet the survival rate stubbornly remains the same. Will the same fate be accorded to the vast number of electronic files we are currently building? Military and quasi-military agencies are collecting terabytes of data daily from satellites and space probes: will they still be readable when we find the time to analyse them?
Bio-diversity is seen by biologists as essential to the survival of life forms on this planet. Is it not possible that those who argue for the benefits of electronic information may well be guilty of making it more difficult for those with an enquiring mind to see new pathways ahead? What is being planned for our future by that fearsome duo - Bill Gates of Microsoft and Andy Grove of Intel? Is there hope for the small publisher with a limited but definable audience in an ocean dominated by conglomerate predators? Is the computer subtly re-fashioning our cognitive processes and capacities so that we shall lose the art of reading and thinking? Our concerns about the management of information might suggest to some that we have forgotten the purposes for which it exists.
The current crisis in libraries where ever-increasing sums of money must be spent on equipment and systems at the expense of acquiring print seems to me a dangerous gamble: we might end up with the worst of all possible worlds. Libraries are leaving their users behind, not out of malice, but because there are few clear voices to be heard in the debate on the issue of quality versus quantity. Whenever this is raised in Internet forums (on- line versions of the 24 hour radio phone-in shows) it becomes a debate about censorship versus freedom. But that is, perhaps, only because most of the predators are waiting and watching. As we debate these issues in Bournemouth Cable London is setting up the machinery to become an Internet Service Provider and I have agreed to participate in a pilot test. What the outcome will be is uncertain - but if you invite me back to Bournemouth ...