The more stuff there is, the harder it is to find.
David Weinberger, one of the authors of the Cluetrain Manifesto and author of the best one phrase description of the internet, Small Pieces, Loosely Joined, has turned his attention to the question of how best to organise information unconstrained by physical storage.
The short version of his solution is not to organise it all – because Everything is Miscellaneous. Supermarkets have to be organised because everything has to be somewhere – and while a few things may be in more than one somewhere, the overwhelming majority of things is in a single somewhere, with that somewhere determined by what kind of thing it is. Milk is near cream; bananas are near apples; pork is near beef; beer is near nappies. Well, maybe not the last one, which is something of an urban legend, though like all the best legends, with a grounding in truth.
Similarly, libraries are organised, almost invariably, thematically – not just to be able to find a specific book again (though that is obviously pretty important), but the more easily to be able to find things like it. All such classifications are inherently arbitrary (which doesn’t stop at least some of them being useful), if not always quite as gloriously arbitrary as the Celestial Emporium of Benevolent Knowledge.
Everything is Miscellaneous is about how that whole approach to sorting things, built up over centuries, collapses in a world where anything can be next to anything else without duplicating physical storage. It is also about the problem that, if anything can be anywhere, things can be hard to find. The solution to that is not to invent yet another single top-down categorisation, somehow better than all the others – partly because it won’t be, and partly because if even if it were nobody would make the necessary investment in classification.
One way of dealing with that is to have lots of ways of classifying things. Pandora, a wonderful service for playing music which is like other music you like (but from which everybody outside the US is banned), uses almost 400 attributes in its analysis to determine ways in which a given piece of music is like another one, ranging from syncopation to G-funk synth line. Another is to let the classification emerge. If thousands of people classify pictures on Flickr, an approach to searching based on keywords will emerge. If millions of people make links between web pages, relationships will emerge – and the founders of Google will get rich.
Weinberger writes well, and the book is a good read. But it’s even quicker and easier to watch it instead. There is a brilliant five minute video on YouTube which explains the whole thing – Information R/evolution.
As for HD30.2 .W4516 2007, that’s the Library of Congress classification for Everything is Miscellaneous – the classification of a book about the death of classification.