Book cases.

Libraries, though they will be radically affected by electronic media, will continue to be needed and the buildings in which they are housed will keep on evolving so that they can cater for ever more diverse societies.

Some believe that libraries are becoming redundant. They argue that the electronic media are now so efficient, so ubiquitous and so economical that they will make the storage of paper unnecessary.

Yet the electronic media have had no very great effect on reducing the volume of books and other paper publications. More of these are printed now than ever before, and almost none of them has an intelligently recorded electronic version. And it will always be impossible economically to record electronically all the pages of all the books that have ever been made. 'Splendid', say the electronic enthusiasts, 'why on earth do we need so many dusty and crumbling files of paper held between boards?' They argue that many books go out of date within a short time of being published: the data they contain are superseded, and electronic media arc in every way a much more appropriate means of providing transient information than paper who wants an outdated reference book? And it is not just data that have become outmoded. For instance, who would now think of reading the huge and solemn volumes of sermons that Georgian and Victorian clerics published to establish their credibility? When were authors very popular in the nineteenth century like Samuel Smiles or Lord Macaulay last consulted? Is it worth keeping most of the products of pulp fiction publishing houses, no matter how sought after they once were?

Of course, the answer to all such questions is that anything ever published anywhere ought to be preserved somewhere. The society which destroys or wilfully forgets its published past is on the road to barbarism as the Nazis so clearly demonstrated. Each country, each community needs it great warehouse of books, because we cannot predict what future generations will want to refer to, nor can we usually afford to record books in any more compact form than the original. (Indeed, the original paper form is often of great importance, particularly in the case of illustrated publications like this and many others.) Hence the need for the great national repositories like the British Museum (p34), the Royal Library in Stockholm (p82) and the Deutsche Bibliothek in Frankfurt (p59). And for those great scholarly collections in our own discipline, those of the Royal Institute of British Architects and the Avery Library of Columbia University.

Having ensured that such great collections will continue to grow, and so safeguard the permanence of at least one example of every printed product, do we need other libraries? Electronic enthusiasts often argue that the Internet or various forms of CD will provide all data necessary for the efficient life.(1)

At any level above the everyday, this proposition is crass. We do not live in a world driven only by facts. While knowledge of such matters as the most up-to-date price of pork bellies, or the means of adding a new chip to a computer may be of vital importance at various times in our lives, it does not make us into civilized people. Civilization, from which all human progress has stemmed, depends on consciousness which is largely gained from understanding not just information, but the opinions of others about facts. Science is based on formulating and testing hypotheses. In rather less dramatic fashion, all other branches of human endeavour are based on critical analysis of past opinions. Human development has in a sense been enabled by extremely complex forms of textual criticism.(2) Because we cannot record most past publications electronically (and indeed fail to make electronic versions of all but a tiny percentage of contemporary paper products), to remain civilized, we need not just huge national repositories, but libraries at every level, from state down to school.

All living libraries have to grow. This means that they become full of books and papers, then either have to expand physically, or decide to locate parts of their collections in other places (what librarians call outhousing).(3) Both remedies for the problem are full of pain, are spasmodically applied and never complete cures for the condition. Now, problems are exacerbated (to a relatively small extent) by repeated explosions of new media: microfilm and fiche, and a whole spectrum of electronic recording and publishing devices, which do not of themselves take up large amounts of space but have special needs for storage and display.

Readers to books or books to readers?

In any serious reference library, outhousing will always be needed in the end, and whether it is provided near the existing site,(4) or elsewhere. It would of course be possible to locate great libraries on remote greenfield sites so that they could grow with little restriction. But a key principle