It may seem odd that I should address an audience such as this on the Resources for Research. When, you might ask, have there not been resources for research? Has the past revealed all its secrets? Is there a text so thoroughly investigated that no further light can be shed on its meaning? Of course not. Libraries brim with excitement: and those one knows best tend to yield the most fascinating secrets. The resources have, quite simply, changed. What I wish to emphasize today is the subtle difference between "resources" and "opportunities". Discerning that difference requires the skill to catch the reflections of light upon water. "Resources" have always seemed to me endless and fugitive: not a library that I have ever worked in has failed to produce surprises: and the discovery of a surprise obeys no grammar I understand. A casual visit to the Folger Shakespeare Library in Washington last year yielded the only known copy of the first edition of a French grammar by Claude Mauger, printed at Blois. The Director was bemused at the happiness of the discovery: "Why don’t you stay a few more days: heaven knows what else you might discover." I am as fascinated by resources as I am by opportunities. I wish I could get them together. That is the irreducible grammar of research, to which no one has found the key; for which no one has written the manual. And yet, the opportunities, which I seized as a young researcher, eager to discover the resources hidden in the libraries of Europe, have disappeared. Travel is too expensive - of late it has shown signs of becoming too dangerous; libraries no longer can make the concessions one became accustomed to in the 1960's: sole occupant of a library, entrusted with the custody of its massive key; for a fleeting moment the "president" of a bibliographical republic. Such excesses of enthusiasm for knowledge as I was privileged to enjoy have vanished. If I revisited the Schlossbibliothek in Coburg I would be probably be suspected of terrorism. The resources for research now have a remoteness which is both necessary and in keeping with the notices in museums which forbid touching the objects on display. The art of printing, which led Europe from darkness to light, is in danger of becoming itself an artifacture, rather than a manufacture. Of artifactures we have quite enough; of manufactures we have only the resources preserved by our enlightened institutions. The pace of life has grown to the point where technology is not simply a convenience: it has become a necessity. My first electric typewriter - acquired in 1964 and still working - seemed an astonishing instrument. Today, with microelectronics a pervasive presence, affecting all aspects of our daily life, an IBM Selectric has all the archaeological glamour of the vacuum tube radio. What has changed?
As for many of you, I suspect, my opportunities for research lay less in contraptive accessories than in attempting to discern the boundary between the known and the undiscovered, and an acceptance of the fact that such a boundary might be illusory. Archaeologists find no difficulty in accepting these premises, since they accept the fundamental fact that the evidence is ambiguous and imperfect. History is not "bunk" - it is, more appositely, "rubble". Technology is incapacitated by the concept of "bunk" - it may help with "rubble".
In spite of the historian's complaint about the insufficiency - and unreliability - of the available evidence, there is, in fact, too much evidence. Our understanding of the past is less obscured by a lack of information, than it is by the overwhelming flood with which we must cope. Since Gesner and Morhof, the available information has multiplied to the point where the great research libraries of the world face paralysis; and researchers face bewilderment. And both parties in the adventure of research (the seekers of knowledge and the custodians of knowledge) struggle in the face of dwindling resources for any activity as non- industrial, non-commercial, non-profit-making and non cost-accountable as research in the humanities.
Yet, strangely, technology - which our undergraduates seem to find so unappealing by comparison with the "Arts" - has developed a device which could well change our perspective; our methods of research; indeed, our way of thinking. I wonder how many of those who read "books" in our research libraries reflect on the fact that the transmission of knowledge has as much to do with the goldsmith's art as that of the curer of sheep’s gut? That is why we are here, in Edinburgh, to share our experiences; to debate the "Battle of the Books"; to question tradition; to grasp, with enthusiasm, innovations which can liberate the strength of mind from the weakness of flesh. "Bold claim", I hear you mutter, as you face the twentieth-century's onslaught upon your senses; the un-endurable battle with bureaucracy; the hopelessness of succeeding in a world which prefers "enterprise" to "understanding".
I have, as you know, been associated for some years with the British Library: that estimable child of the British Museum. Of the many distinguished scholars who have figured in that institution's history, I would like to remind you today of Thomas Hartwell Horne. Not on account of the astonishing breadth of his learning and interests - from Biblical exegesis to land-management - but because in 1825, at the request of the Trustees, he submitted a detailed scheme (entitled modestly Outlines) for the classification of a library. Only a handful of copies were printed (four remain in the Library's collections), and it was never published. Little wonder, therefore, that it does not figure in histories of either librarianship or bibliography. Although Horne's scheme was never officially adopted, Thomas Watts, the Library's first "Placer" quietly adapted many of Horne's ideas to his own classification scheme, which was based on that of Brunet, with the result that, until well into this century, the books in the British Museum were placed according to an essentially Victorian view of knowledge:
Printed catalogues, while undoubtedly useful, have generally been helpful only in providing answers to questions already formulated, and susceptible to questions which recognise the constraints of their particular arrangements. The catalogues published by the British Museum since the middle of the nineteenth century have been almost exclusively alphabetical by author. For materials acquired since 1875, there is Fortescue's subject index and its successors, but there are only limited tools for items in the collections before that date.
It is obvious, therefore, that the conversion of the General Catalogue into machine-readable form, providing subject-access to the collections via pressmarks, keyword-access to titles, places and dates of publication, together with codings for country and language, will constitute a resource of immense importance for research in all subjects.
With the eighteenth century project [ESTC] now firmly established, I have recently embarked on a project which aims to republish, on microfiche, a comprehensive library of texts for the study of the nineteenth century. It will, I am told, take twenty-five years to complete - a fact about which I have understandably little enthusiasm - and is being planned by the publisher - Chadwyck-Healey - in close cooperation with the British Library. Separate from the obvious opportunities which access to the original texts will provide, scholars will have access to the machine-readable records which will be created for each text selected via the major bibliographical networks of Europe, North America, and Australasia, as is the case for eighteenth-century texts being published in microform by Research Publications. Full titles, and for the benefit of those concerned with book-trade history full imprints, will make it possible to explore nineteenth-century publishing in a way never before possible.
Awareness of the new technology has required those engaged in research to consider alternatives to the card-index; the familiar disorder of the pieces of paper which clutter the home and the office. And research has become less itinerant; less concerned with mislaid baggage than with the mislaid password. Information is available to all who can afford it, and it is available whenever needed. It has to be paid for, of course: but, as I sometimes remind those who grumble about the on-line charges for access to BLAISE-LINE [at £30 per hour] - £30 does not get you far on British Rail. With large computers storing ever-growing quantities of information and small computers able to communicate with them we are seeing the emergence of a totally new way of dealing with the materials of history. The most important benefit being, I think, that these new tools take us a step nearer the methodology of science. For the first time in the history of research it has become practicable to test assumptions.
ESTC is increasingly being used by scholars to produce statistical evidence which tools like printed catalogues cannot provide: the distribution of books in provincial towns, and the bookselling networks which only become apparent when imprints are analyzed; the rise of the popular book-formats (octavo and duodecimo) as an indication of the growth of literacy; the growing importance of provincial printers in educational publishing, for so long dominated by London; the relationship between contested elections and a vigorous local press (as in the case of Newcastle); the different kinds of books translated from European languages; the subterfuge of "Londres" for clandestine publication of works likely to get their publishers into trouble with the authorities in France (there are hundreds of instances of books popular in France, actually printed in Switzerland, Holland, or Belgium, but claiming to be published in London).
The accumulation of bibliographical evidence has, until now, required a great deal of patient sifting of rubble, not always with conclusive results: machine-readable records for a substantial proportion of the output of the printing press in eighteenth century Britain and North America will relieve the researcher of much tedious enquiry and the risks of speculation unsupported by evidence; and the distributed availability of the sources in microform alleviates the inconveniences and expense of travel. But these benefits have consequences: for one thing, the gathering of the evidence no longer requires the level of financial support we have come to expect, and the researcher will find it more difficult to disguise the subversion of interpretive skill in a welter of footnotes. Let me give you an example.
A few years ago, a study of the decisive role of a publisher like Jacob Tonson in the age of Pope and Swift would have involved the scanning of numerous library catalogues in Britain and America; a substantial subsidy from a foundation; much travel; and might well have absorbed five years of what we like to call "research". An interim result would almost certainly have been a file of several hundred cards. And cards, as we all know, have the disadvantage of usually being arranged in an order which necessitates the creation of indexes - i.e. more cards. By contrast, it is a simple matter to discover every book with Tonson's name in the imprint (so far described) and request a printed list of them; and the list will give full descriptions together with the location of copies. Better still, it is now possible to download the records into a microcomputer, and analyze the records with any one of a range of database programs currently available. The estimated cost of the exercise so far has not exceeded a hundred pounds. To the basic structure of the record - as developed for ESTC - can be added other fields more pertinent to the purposes of the project: subtleties of typography and design, for example. Providing that our researcher has a grasp of the possibilities provided by the program being used, it is possible to analyze the evidence in a variety of ways. The end-result ought to be an article of moderate length, executed in a matter of weeks.
The software now being made available to owners of professional (as opposed to hobby) microcomputers has brought the capabilities of the mainframe within the reach of everyman, nor is it necessary, as it used to be, to be an experienced programmer in order to take advantage of their power. One very healthy consequence of this has been the demystification of what computers are, and the tasks they can be made to perform. Another healthy consequence has been the realization that in building a database, of whatever size, it is essential to know where you want to finish before you start. For certain kinds of enquiry this is a disadvantage, but it has resulted (at least in my own experience in the British Library having developed numerous programs for internal use) in a greater clarity of purpose than might otherwise have been the case. Thus, the conversion of the First-Line Index of English Verse in the Department of Manuscripts into machine-readable form, has resulted in a tool of considerable importance for the student of English poetry, and has already identified the Department's substantial collection of common-place books as distinct from those manuscripts containing miscellaneous scraps and casual accumulations of transcripts. It has also made it possible to identify the relationship between manuscripts containing similar texts. The old manual index, now thumbed to the point of collapse, could only tell you where to find a text if you knew the first line.
A great research library, with a determination not to miss opportunities provided by the new technology, is quite a good vantage point from which to assess some of the remarkable developments taking place in Japan and America. One development, in particular, strikes me as likely to provide opportunities for research: the digitized image.
I have recently seen, at the Air and Space Museum in Washington, a prototype system which captures images (printed or manuscript) in perfect detail at a rate equivalent to microfilming; permits the labeling (or indexing) of each image so that its retrieval can be quickly retrievable; stores the data on a hard disk; transfers it to tape for conversion to optical disk format; and the entire system runs on an IBM AT. Furthermore, the images stored on disk can be converted from their bit-mapped form into ASCII strings by a new technique of optical pattern recognition and incorporated into a dBase III file. Given the immense storage capabilities of the new generation of optical disks, it is already possible to conceive of the entire ESTC file, and the necessary indexes, available on a single disk, and capable of manipulation and searching with microcomputer software. And the digitized image means that an end is in sight to the tedious task for bibliographers of transcribing title-page information.
The present use of large databases is still subject to the constraints imposed by the complexity, and cost, of telecommunications. In America, huge sums have been spent on trying to eliminate some of the problems created by bibliographical networks with different protocols and different telecommunication standards, so that these differences become transparent to the user. But I wonder if the future does not lie less in developing networks which link the user to vast stores of discretely held data than in distributing that data in a form which permits local, less expensive exploitation.
There are times when I look back on the traditional techniques of research (insofar as they concern bibliographical evidence) and regret their passing. Unexpected discoveries made in libraries throughout Europe, unfailingly providing me with a generosity of cooperation and privilege; the unstructured knowledge which comes from a lifetime of handling books. But the price we are now paying for that generosity is everywhere evident: books handled, microfilmed, to the point where their survival is threatened. And the burden of use is unequally distributed. The imminent availability of the Bodleian Library's resources for books printed before 1920 in machine-readable form will undoubtedly result in yet further deterioration of its wonderful collections.
Here in Edinburgh we have an opportunity, for two days, to learn more about and to debate the impact of technology on some of the disciplines taught in our universities. I am sure that there are surprises in store for all of us. But let us not forget that where the mind of man is concerned there are always surprises. The history of research, and the undocumented grammars which have sustained it, is a chapter in intellectual history still to be written; and here we are beginning a new chapter.