What will become of libraries when everyone has access to all the information in the world from their desk?

Time to shelve the library?

The library of Rabindranath Tagore University (Santiniketan) near Calcutta is an imposing institution. After several weeks completing many forms in multiple copies, students should be admitted to a grand Edwardian Raj hall piled high with priceless books and manuscripts--many crumbling to dust under an onslaught of worms, beetles and interesting fungi.

Might this library of the past be a foretaste of the future? Could existing libraries such as the British Library, with its 10 million books, 120 000 manuscripts, 100 000 seals of office and 3000 papyri, be superseded by the revolution in electronic communications, just as the monasteries dedicated to writing and copying manuscripts were by printing?

In fact, the process has already begun. Somewhere--let's say the University of Winnesota in the Midwest of the USA--there must be a collection of ancient Sanskrit literature held on computer. Perhaps many are the same as are held at Santiniketan. A London- based browser can access it in seconds via the Internet computer network running across the US Department of Defense's trans- Atlantic computer link.

Given a decent phone connection, it would even be possible to take a computer into the library at Santiniketan, sit amidst the decaying majesty of the books themselves, and electronically search and read many of their texts held as data half a world away. It is far more convenient on screen; so why, apart from sentiment and atmosphere, bother with a physical library at all?

Computer users can already consult the catalogues of 75 British university libraries, and hundreds of American and European institutions, without leaving their seats. Over half of Britain's university libraries accept orders for books electronically.

There is at least one fully-refereed scientific journal, Current Clinical Trials, which exists only in electronic form. It's published by the American Association for the Advancement of Science to demonstrate the technology in a field which needs rapid access to results. Papers are accessible through the Internet academic computer network 48 hours after acceptance. Similar projects are planned in the UK.

Tom Stonier, emeritus professor in science and society at Bradford University, imagines a totally electronic future for information. "The function for libraries as places which you physically visit to access information is going to be superseded. A future Karl Marx would write Capital at home, with the BL reading room on his desk... In my view everybody will have access to a personal computer with communications within 20 years." Librarians will become "theoreticians on ways of managing the wealth of information", more than custodians of physical objects.

Stonier thinks that "works of fiction, and libraries as a source of pleasure, will last." Local libraries will become primarily social centres, if they can embrace the new media. Technical libraries are likely to become a habitat for historians rather than researchers, with the growth of a world library available by telephone. Brian Perry, director of research at the British Library, talks about "libraries without walls": "from your computer terminal you should be able to access the riches of many collections."

This will be a major change in the way we handle information--so much so that it will cause a discontinuity in our culture's collective memory. For example, to consult the contents of almost any major US newspaper from 1980, you get your computer to dial a local computer network, and then type in commands to search for the subjects you are interested in. But if you want to search for a story from 1979, you have to take a plane to America; those papers' contents have not yet been -- and may never be -- typed into a database. Roughly the same cut-off date applies to university library catalogues. As Stonier says, "undoubtedly a good deal of information will get lost, as a lot of information must have been lost in the Middle Ages when the transition to print occurred."

He sees the transition to electronic media as similar to the advent of print: "Before the Renaissance you could read all the books that had ever been printed: later, you needed not only a repository but people who could guide you to the book you needed. Librarians now have to deal with global data-bases."

There is another similarity: just as early printing could not match the graphic quality of illuminated manuscripts, current computer technology displays graphic images coarsely and stores them in many incompatible forms. Remote access to higher-quality graphics will have to wait for high-speed communications networks such as Super JANET (see New Scientist November 21 1992) and, much later, domestic fibre-optic networks. Even then, as Stonier says, "the feel of a book and the smell of a book are very important -- to our generations they give a kind of companionship."

Lynne Brindley, in charge of the British Library of Political and Economic Science, agrees that "The future of libraries is the future of librarianship skills--helping people to learn what to discard." Technology is spreading rapidly: "In the past two years an awful lot more of the necessary technical building blocks have arrived." But as for paper libraries being replaced: "My answer is never, in some subjects, and certainly we have a ten- year transition" in those subjects which may become paperless.

That transition has already begun in earnest in France, where the National Library allows users of the country's six million Minitel telephone communications terminals to reserve books from their homes. When the new Library of France opens, in the Tolbiac district of Paris in 1995, 300 000 documents will be available in digital form, both as page images and as text. Researchers will be able to access them from work-stations within the library, provided with a set of software tools for analysis-- "Computer-Assisted Reader Stations".

These electronic documents will not, however, initially be accessible outside the building. One reason is publishers' concern over copyright. Another is inter-library politics: as the Library of France's policy paper says, making documents available remotely to all potential users would "short-circuit the network of university and public libraries".

The British Library is also due to move, to a site next to London's St Pancras station. Well-publicised problems with the movable shelving system--which requires extensive remedial work to fix problems with the drive mechanism and with rust--has delayed the start of stock transfer from March 1993 to an unknown date in 1994. The new library will be extensively automated, allowing readers to order books from computer terminals able to access a new electronic catalogue called the On-line Public Access Catalogue, or OPAC.

The Library's directors are still discussing whether to make OPAC publicly available. It is likely to be available to academics, through the JANET inter-university computer network. But funding is a serious problem: though 150 OPAC terminals will still be available in reading rooms, a planned 100 terminals behind the scenes have been cut. The existing information service, BLAISE (for British Library Automated Information Service) will keep operating; it gives remote access to 20 databases, including catalogues of all the stock the library keeps records on. The back catalogue, covering acquisitions the Library's opening in 1759 to 1975, will officially be made available this month [December 1992].

But in keeping with the distributed nature of the library of the future, it is operated from Boston Spa in Yorkshire and held on a computer belonging to Rank Hovis Mcdougall. No decision has yet been made on whether access to OPAC will be free; subscribing to BLAISE costs [UKpound]70 a year and [UKpound]10 an hour to use. Document digitisation is, says Perry, "purely in the experimental stage".

In January the Library will publish an ambitious strategic plan, setting out its aims for the rest of the century, which will place great emphasis on electronic documents: it says that "increasingly, end-users will access library services from their terminals, unfettered by national or geographic boundaries".

Although it can be seen as a scientific resource, the Library is funded by the Department of National Heritage. In the Autumn Statement, it gained a meagre 3 per cent increase in cash funding to [UKpound]72 million, a cut in real terms and 10 per cent below previous funding projections. Unsurprisingly, its strategic plan will announce an increasingly selective collection that will no longer aspire to being comprehensive in its coverage of the world's literature. New technology "will have to be paid for from savings in other areas," says Perry.

The Library will aim to be a "single gateway" to information around the world. Perry says this "does give you the ability to search with a common command language". Others are scathing. One senior academic librarian describes the draft strategic plan as "absolutely rubbish. The notion of a single gateway is absolutely misguided: we're talking about a world of distributed information."

Brindley points to another philosophical argument about how a library without walls should develop. At present, finding a really obscure title among the university libraries may involve searching dozens of catalogue databases. Will there be an on-line "union catalogue" covering all the universities' holdings? "The world is moving too fast for that," she says. "What we need is interfacing tools to help us use all these resources."

An example is the "knowbot", which is being discussed by users of the Internet. This will be a personalised programme that asks what you are interested in, then makes enquiries in a selection of likely databases in their own instruction dialects. Philip Bryant, director of the UK Office for Library and Information Networking, which assists the development of electronic communications by and between libraries, describes the current situation graphically: "[Using] Internet at present is like tipping the contents of the British Library and the American Library of Congress into Twickenham Rugby Ground, then saying 'go find'."

Technically, the problem is soluble. It depends, according to your taste, on standards or on intelligent search tools that can cope with the lack of standards. Other probable developments could transform a scientist's literature search from a week in the library to something the computer does while you have lunch.

Free-text databases allow users to scan the entire contents of a paper rather than a fallible list of key-words. Reverse reference searches--finding out what later papers a given publication informed--will become possible. That leads naturally to such facilities as automatic flagging of papers retracted by their authors, saving a lot of the tedium (and potential embarrassment) of a literature search on paper. It makes possible a real-time citation count (or even readership count) which changes before your eyes.

All this will be available on a personal computer--if you have one. For Perry, "the nightmare is where most material is available in machine-readable form, and a lot is available only in that form. People who don't have remote access will lose out." Consequently, we risk dividing the world into "the information- rich and the information-poor".

Brindley, too, sees "a danger of a particular cultural dominance- -the notion of the Third World getting involved in this when they haven't even decent telephones is a really big global problem." It seems unlikely, though, that people in industrial countries will relinquish the advantages of new technology just because colleagues in Bangladesh don't share them. After all, it hasn't happened so far.

And, as Perry observes, "theoretically, the new technology is an advantage for developing countries, because they will get information quicker. A scientific journal may take three months to reach Thailand through the post." This will only work, however, when developing countries have reliable phone systems-- or if universities can raise the money for satellite links to bypass their national telephone services.

Scientists in the UK will face more abstruse problems as publishing and libraries move onto the electronic networks. Brindley wonders whether electronic publication will count towards tenure: "It's too early to know".

And what is the legal status of an electronic "book"? Brian Lang, head of the Library, announced in October that he would press for legislation to require the deposit of electronic publications with a national archive, in the same way that the copyright libraries must receive a copy of every paper-and-ink publication. Moira Simpson, a researcher into computers and the law at Strathclyde University, believes that existing legislation "may stretch to cover this", but this interpretation is not being acted on. In her view it ought to apply even to computer games programmes, which claim the protection of the Copyright Acts as "publications".

Traditional libraries are a powerful symbol of our culture's regard for knowledge; it is no exaggeration to call them shrines to learning. They will at the very least remain, for all time, essential to historians of the pre-computer age. And they're historic places in their own right, comparable to the Coliseum or the Pyramids but much more fragile. One BL researcher believes that "in a thousand years, you'll be able to wander into the BL and still know it's a library." Let us hope so -- but it will take more than hope to stop the BL going the way of Santiniketan.