So there is now a new formula with which boffins can identify an author's "literary fingerprint". Having exhaustively analysed texts by Melville, Hardy and DH Lawrence it is presumably now possible to identify texts by these authors without looking at the title page or spine of the book.
These exercises are always depressing. That there is some algorithm which helps us determine who wrote what falls into the same kind of category of futility as those scientific studies that claim to have determined the formula for female beauty or what makes a really good sandwich. The sadness of the claim resides in the way in which human intuition, and the actual experience of reading, is to be removed from the various processes of life. It is like explaining to the utterly humourless what the key elements of a joke are.
For any reasonably well-read person should be able to tell whether a text is by Hardy, Melville or Lawrence almost at a glance even if they haven't read it before. And that a computer has been programmed, presumably with some effort, to recognise that if a writer uses words like "thrust", "flame", "being" and maybe "fuck" with any inordinate frequency then he may well be DH Lawrence should really occasion little more than a slow, sarcastic handclap. Oh, well DONE. Do you remember when, years ago, some dismal piece of doggerel (which began, as I recall, with the lines "Shall I die?/Shall I fly?") was, on the basis of word-frequency, claimed to be a hitherto undiscovered work by Shakespeare? A few people were impolite enough to point out that it was far too shit for Shakespeare to have written, but on the whole news agencies and those with a tin ear for poetry went along with the assertion. It even made it into an edition of his collected works, but I think now has been quietly dropped.
Concordances and other word-crunching tools like this have their uses – it can be revealing to note what words crop up most and least often in a writer's vocabulary – but to trust literary attribution to a machine is to forget that even a note for the milkman (if there are still milkmen) is still about human writing to human. Machines just don't get it.
One intriguing notion from the scientists' latest wheeze, though, is the claim that "the writing of a text can be described by a process where the author pulls a piece of text out of a large mother book (the meta book) and puts it down on paper. This meta book is an imaginary infinite book which gives a representation of the word frequency characteristics of everything that a certain author could ever think of writing."
When I first read that I thought it was rather nicely Borgesian. But then, when I thought about it a bit more, I realised the idea was meaningless garbage. And anyone who can write a phrase like "word frequency characteristics" without being in some way ashamed of himself is really better not writing anything at all, or having anything to do with other people's writing. I suggest they read more good books.
