Google books have no shelf life

"This is the day the world changes," said John Wilkin, a librarian at the University of Michigan, commenting on the announcement of the Google Print programme, a plan to digitise millions of books and make them available to all on the internet.

Well, librarians live famously sheltered lives, and Mr Wilkin can be forgiven for getting excited - but cultural change rarely pivots on so fine a point. In this respect, at least, the world has been changing for years, and will continue to do so for years yet. The Google announcement - which included the fact that a huge chunk of the Bodleian Library would be put online - was just the latest step in an enterprise that has been under way for decades.

Project Gutenberg, one of the internet's main sources of free online texts, was first conceived in 1971. Since then, it's been joined by a long list of other sites, all of them dedicated to making literature available to readers at the click of a mouse. The Digital Book Index, an online catalogue, notes that there are now 63,000 books, texts and documents available online without charge.

Still, Google's act of calculated philanthropy does represent a step change in the digitisation of human knowledge. It prompts the question of what that change might mean in terms of the feel of a culture - the way in which ideas are transmitted and arguments constructed. Is Google knowledge the same as book knowledge, only faster? Or is it so different in texture that it effectively alters how we understand the world?

This isn't exactly a new question; indeed, it's nicely dramatised in a scene in Middlemarch in which Will Ladislaw is stung by Dorothea's praise of that 19th-century search engine, Edward Casaubon. "Such power of persevering devoted labour as Mr Casaubon's is not common," she says defensively, making an unfavourable contrast between her husband's toil in the bookstacks and a flightier kind of intellect. Will retaliates with an attack on the parochialism of English scholarship. "The Germans have taken the lead in historical inquiries," he says, "and they laugh at results which are got by groping about in woods with a pocket-compass while they have made good roads."

The implication is clear; it's time Casaubon got on the information superhighway, downloaded theses from Heidelberg and signed up to the Key to All Mythologies discussion forum. But he can't, because the internet hasn't been invented - so he must plod on, unable to see the wood for the trees. Casaubon could be taken as a model for unwired scholarship, while Ladislaw is plugged in, aware of global developments. And, as Casaubon is a cautionary figure - a man in whom immense knowledge simply congeals, achieving nothing - we're inclined to take Will's side.

There is a sting in the tail, though, and one that's relevant to the intellectual culture promoted by the internet. George Eliot adds this: "Young Mr Ladislaw was not at all deep himself in German writers; but very little achievement is required in order to pity another man's shortcomings." So Will could also be taken to represent the dangerous facility of internet knowledge; the speed with which it can confer a superficial expertise on the inexpert.

That's the real cause for anxiety with internet intellectualism, rather than that its standards of scholarship are poor (though they often are). The Gutenberg Project actually boasts that "we do not write for the reader who cares whether a certain phrase in Shakespeare has a ':' or a ';' between its clauses". But that will almost certainly change as the technology improves and academic use grows. What won't alter - indeed, what will probably be amplified - is the ability to find information completely isolated from its context.

You can already see the consequences of this in bookshops, with the increasing prevalence of what you might call Google books; impressive in their range of reference, making discipline-defying connections between different branches of knowledge. They give every appearance of massive erudition, but appearance is often all it is, because their authors are not so much well-read as well-surfed - not the same thing at all.

Perhaps that doesn't matter, in the long run. After all, if facts are readily available to everybody with an internet connection, mere possession of knowledge will eventually matter less than the ability to manipulate it into new and pleasing patterns. But, much as I look forward to Google's virtual library, I suspect there may be a reaction against the magical swiftness with which evidence can be assembled for almost any proposition. It will soon be time to found a Campaign for Slow Scholarship.