John Ashworth: Well, thank you very much for inviting me here. It's a great pleasure to be in Ann Arbor for the first time.

Let me begin by explaining to those of you who don't know, which I suspect is the majority, what the British Library is. The British Library is the national library of the United Kingdom, and it was created in 1972 by bringing together a disparate collection of books and artifacts which had very different histories. The core of the library are the historical collections, which date back to the collections, created largely by establishment figures in the 17th and 18th centuries, of medieval manuscripts and early published books. Then there were the royal collections. George III doesn't have a high reputation as a politician in this part of the world, I know, but he was a very eminent, perhaps one of the greatest, English bibliophiles, and his collection of 60,000 volumes is still an actively used collection and the core of the library.

But in addition to these historical collections, we have the library of the Patent Office; National Science and Technology reference and lending library; and the India office library, which contains all the records of the East India Company and then the British raj, and so effectively is the historical record of India for two or three hundred years.

And in addition, we don't only have books. We have the National Sound Archive, which includes great chunks of the BBC output; the map collection, not only the royal map collection but the map collection of the War Office, which, of course, since it was responsible for fighting wars across the British Empire, effectively is an early historical map record of the world. And then sheet music from autographs of Mozart to the Beatles and beyond and so on.

So it's an enormous corpus of artifacts, information containing artifacts of various kinds, about 15 million books, 150 million items all told. And it is of course a copyright library, which means that by law, one copy of anything published in the United Kingdom has to be deposited in the British Library. And the British Library has to keep it. The other copyright libraries in the country - universities of Oxford, Cambridge, Trinity College, Dublin, the national libraries of Scotland and Wales, they're the other ones - they don't have to keep everything, but we do. So we take in 8,000 items a day. And those of you who think the electronic revolution is just around the corner, just think of my colleagues back home who today will have to process 8,000 printed items. Of course the electronic world is coming, but as far as we're concerned, on our horizon, it's still less than 5 percent of our total intake.

I want to make two points illustrating the impact on us that the new technologies will have. The first relates to our historic collection. Let me give just one of many countless examples. We have the manuscript, the only surviving manuscript of an Anglo-Saxon poem called "Beowulf," which is one of the foundation myths of the English peoples. It relates to society in the sixth and seventh centuries, a pagan society in the sixth and seventh centuries, but the manuscript dates from the 11th century. So there is considerable scholarly doubt and interest about the relationship between the myth itself and its creators and those medieval copyists who almost certainly didn't fully understand what it was they were doing when they copied it. In addition, by an unfortunate accident in the 18th century, bits of it got burnt, and so it's charred round the edges.

What we did a few years ago as an experiment was to digitize it, using visible, infrared and ultraviolet light. So we have three images of this manuscript, and we put them on the Net. Now, there is no doubt that this means that up till that moment perhaps a handful of scholars has been allowed to handle that very precious and rather delicate thing. But of course, they only had visible light. By putting it on the Internet, millions of people now can see more of the Beowulf manuscript and see it in much more detail. They can see copyists' errors; you can see places where the scribe clearly didn't understand the word he was writing, and so guessed and guessed twice and had ink blots and then fiddled and all the rest of it. You can see that beautifully under UV and IR light, so that the information on the Net is incomparably richer and better than ever has been possible with the manuscript itself.

Our surprise was that the demands from people who wanted to see the original went up by an order of magnitude. So the belief that to digitize a text, particularly an iconic, historic text, will reduce demand from people to see the original is just false. What in fact you do when you put things on the Net is you advertise its existence. It's one of the most potent advertising media in the academic world that one could imagine. We have many other examples, by the way, besides "Beowulf," but what we have found without exception is that digitization creates a demand for access to the original.

This has one potentially very worrying aspect, and this relates to what is known as cultural restitution. I said our historic collections, thanks in part to the history of the British Empire, actually encompasses other people's history as well as our own. We have the Elgin marbles problem in spades. At the moment what we're doing to cope with that is providing free access to those who visit London and our fine new building in St. Pancras so that they could actually see it themselves. But, of course, this is inadequate, and we're digitizing those historic, iconic items. But my suspicion is that although that will be seen as a necessary part of sharing our access to our collection, it will fall far short of what is necessary. And I suspect that the very fact of digitizing will lead to enormous waves in demands for cultural restitution of various kinds, and how the academic community - who find consolidation of collections into libraries extremely useful - how the academic community is going to cope with political demands of nationalist groups who want to own the icons around which their nationalism has been focused, is going to be very interesting. It isn't only a problem relating to the nations as we know them. The library at the moment is under great pressure from people who live in Northumbria, northeastern part of England, to return to them the Lindisfarne Gospels, which were stolen, as they put it, by Henry VIII.
The final point I want to make, responding, Chairman, to your challenge, one of the other things we do with our collection is operate a document supply service. It relates largely to the scientific, technical and medical journals which we house, and we earn something like 20 million pounds sterling a year providing scholars with copies of our holdings - and I'm pleased to say the University of Michigan is a very good customer of ours and long may it continue. But that document supply service was set up in the 1960s. It's essentially paper-based in concept and indeed, still currently, most people, including those from the University of Michigan, want back from us faxes of journal articles, and that's what we provide. But clearly this is going to have to change. And clearly we're going to have to become much more attuned to the relatively few people at the moment who want electronic copies or electronic versions of printed material, and we're going to have to digitize it.

But if you think of a digitized document supply service, actually what you have is indistinguishable from an electronic publisher. Although it's no part of the British Library's intention, it could well be that some of the publishers of the future will stem from areas like document supply, which are currently peripheral to the publishing activity, because they have the marketing and access routes, unlike publishers, which tend to work with bookshops. The marketing outlets for publishers are quite different from the marketing outlets of document supply, and it seems to me at least as probable that the electronic publishers of the future are going to come out of the document suppliers rather than the book suppliers.

This is not uncommon in technological advances. I just close by reminding you that it wasn't those who made cotton shirts that gave us the delights of drip-dry nylon; it was the chemical companies who wanted to sell nylon, not those who wanted to sell shirts. Thank you very much.