Amazon has just announced a new line of e-reading devices. While the tech press is digesting the new information, one thing of interest has already emerged: ever disruptive, Amazon now plans to disintermediate the human reader. We are entering the era of post-human publishing. Polish up your chrome fittings, set all your dials, and read on.
The fascinating aspect of Amazon’s announcement is the emphasis on enhancements to a feature called X-ray. X-ray is an automated service that provides information about a text that is not apparent from the text itself. Some examples of X-ray have been with us for some time now. With highlighting, you can see comments made by other readers of the same text, in effect providing you with crowdsourced annotations. X-ray can also draw on other digital resources and bring them to bear on a work–that’s the really big news. Many of the examples I have seen in the tech press understandably focus on the use of X-ray for Amazon-marketed movies, which can be enriched by drawing on the Internet Movie Database (IMDb). I say “understandably” because movies are more popular than books, the IMDb has no precise parallel in the world of books (that is, no one place that aggregates so much information on every title), and, not coincidentally, Amazon owns IMDb, giving it special and exclusive access to use the reference in this way.
What this means as a practical matter is that we are entering a period when all books potentially come to us as annotated editions. At this point someone will say that Amazon’s algorithms can never replace the careful editorial work of a professor of literature, and that is true — today. But the point of thinking strategically is to see the tree in the seed. Amazon is a-planting.
As good as today’s editors are, there are things that machines do better. Big Data applications to books are likely to change the way we think about texts. The example above of crowdsourcing highlights is just one of many that no human editor could have accomplished without the aid of machines. And there are many others. What is the vocabulary of a particular book and how do its word frequencies vary with those of all books published that year? What are the statistics for the frequency of dialogue in a novel and how do they compare to other books, other authors, other periods? I leave it to others to determine how useful this kind of thing will prove to be. On the other hand, when the time comes for machines to serve as arbiters, they may be understandably biased toward findings like these.
There is a big difference, though, between an X-ray for a work of nonfiction (including all research literature) and fiction. For nonfiction, who would not like to have prior research, perhaps culled from the world’s great libraries, brought to bear on a text, provided that the interface is useful and that the (human) reader can control the scope and depth of the annotations? But with fiction there are other forces at work. Sometimes the aesthetic point of a novel is not to know what is going on.
Let’s take the example of Amazon providing a plot of each character’s appearance in a work. Now you know that John Doe (or Huckleberry Finn or Tom Jones) will appear on a specified number of pages. But suppose you are reading a book where withholding that information is precisely the point, suppose you were reading a mystery or a psychological novel where part of the reading experience is to reconstruct the events of the story in linear form even though they are presented in a non-linear fashion?
Keeping the reader in the dark is a time-honored narrative trick. In Henry Fielding’s “Tom Jones” (1749), there are instances in which the author uses Latin quotations and then deliberately mistranslates them. Very clever. Fielding’s joke here is that he knows he has two kinds of readers, those who know Latin and those who do not. If a computer reliably translated the quotations, the joke would be lost.
I first began writing about this topic in 2003 in an essay entitled “The Processed Book.” In that essay, I used the poetry of Ezra Pound as an example. Pound’s poetry is notoriously difficult to read because of the welter of allusions to earlier works. If you don’t pick up the references, it’s hard to know what is going on. For a human reader of Pound, part of the aesthetic tension is between knowing and not knowing, but if the poetic equivalent of IMDB were running in the background of Pound’s “Cantos,” everything would be spelled out; there would be no obscure allusions.
Thus technology shapes art. If Pound were alive and writing today, aware that a literary IMDb was everywhere present, he would have had to reach for a new style, a different kind of poetry. Amazon’s X-ray is not just a cool tool for your e-reader; it is a shaper of culture.
And it is also a reader of a different sort. By telling us which character will appear where, by documenting every allusion, by plucking the heart out of a mystery, the Amazon platform is now doing some of the reading for us. The text we approach for the first time has been predigested, as it were. The robots have gone through, grazed a bit, and left the remains for us.
I can’t say I much like being disintermediated. After all, the self is a mediator and I am not ready to toss my pound of flesh into a drawer and simply fire up the transistors. But I will respond to the sheer coolness of X-ray with the same spirit as someone who puffs on a cigarette. I know this is killing me, but . . .