Add in my access to the Internet, and through that to the catalogue of the Cambridge University Library up the road, and it seems that I can find any information I need as quickly as I can think of it. There are moments when it seems that the whole world has been turned into one vast digitised encyclopaedia by the information revolution. There are no longer serious limits to the amount of information which can be stored and retrieved. If CD-Roms prove too small there will be newer, yet more gigantic storage media.
If these prove too small, then there are already whole libraries available across the Internet; every poem ever written in English before 1900; 1000 years of medieval civilisation in the Patrologiae Latinae. Soon there will be more: complete editions of Hansard from the beginning; everything ever printed in this country in the 19th century; the archives of the KGB. All of them will be stored in microscopic laser-cut pits on slivers of aluminium: the human eye sees nothing there but a slightly dulled mirror which, as I turn it, shows sudden prismatic flares like straightened, vivid rainbows.
This dazzling blankness is one reason to be suspicious of the idea of "information". The term comes from computer science, where it has a fairly precise meaning. A compact disc, say the computerate, holds about 460 times as much information as a floppy. The hard drive on my ancient laptop holds a third as much information as a compact disc. But in what sense are these statements true? Looking around this room I can see an extraordinary variety of compact discs. One holds the complete works of Cannon's Jug Stompers. Does it contain the same amount of information as the Encyclopaedia Britannica, as the video disc of The Chant of Jimmy Blacksmith (which I have never managed to make play) or as is contained in two Beethoven piano sonatas? And how could any of these be compared with the English Poetry CDs from Chadwyck Healey, or they with the Duke Nukem 3D Platinum Edition? In another rack, one disc containing 21 Bible translations nestles against the catechism of the Catholic Church.
Here is all the knowledge necessary for salvation. But it isn't information. In fact none of these discs are worthwhile for the "information" they contain, any more than animals are valuable for their genes. DNA can also be understood as a sequence of numbers; in human DNA there are about 3.5 billion of them. This is close enough to the number of holes on a compact disc: in the library of the future you could fit each reader in among the CD-Roms he reads. I put this idea to show its absurdity, though it is one of the most influential myths around. People who would reject the idea of an immortal soul as a religious superstition treasure the hope that they are really software which just happens to be running on fleshy machines but whose nature is to be immortal and capable of running on anything. What's more, they regard this hope as scientific.
The point is not just that human beings aren't software, though we aren't. It's that software isn't software either in that sense. It's all embodied and what it contains is not information, but meaning. The limitless mind of God - should it exist - may be able to apprehend the world as pure information. But all mortal or created beings see the world in the light of their limited purposes. Without a common web of purpose binding the reader and the writer we might as well use all these CDs as bird-scarers - when of course, they convey an urgent message to the birds they frighten. This reflection suggests an answer to the question I first started with: how big can an encyclopaedia be? The answer is that the size has nothing to do with physical limits, nor even with the number of facts inside it. It is constrained by tone and by indexing. The word itself brings this to light, for the root of it is the Greek idea of "a circle of arts and sciences essential to a liberal education."
Circles may spread. An encyclical letter is meant to cover the whole round world. But they are essentially harmonious. There is a concord within them. The differing arts and sciences must fit reasonably well together: they must also be presented to the same sort of reader in the same kind of tone. So the boundaries of a library are not set by technology but by the human labours of the indexers and the compilers and even the writers. This is not a matter of information leaping from one form of hardware to another, but of the slow organic growth of culture making sense and giving meaning. Bigger will almost always mean worse, even when it comes in a smaller package. The value of an encyclopaedia is determined as much by what it excludes as by what it contains. The reason that the Routledge Encyclopaedia is worthwhile is that it shuts out billions of words of bad or confused philosophy and contains only a few million words of scrupulous lucidity - or so I hope. I haven't, of course, got round to reading it yet.