Disclaimer: this is a shameless re-post of Laura Miller’s Salon.com article on metadata and iBook sales, plucked from Three Percent, the blog of Open Letter Books. As Chad Post of Three Percent points out, the article is grounded in the idea that, whereas the 20th century was about “sorting out supply,” or ordering information according to what is being produced, the 21st will be all about demand, or organizing information according to the way in which it is used.
Overall, I think Miller makes some very insightful observations about the relative merits of various e-reader platforms, and offers a refreshing, jargon-free overview of the data that drive the Amazon.com recommendation and categorization algorithms (which are largely based on user input). What she didn’t discuss, and what would really require separate consideration, are the implications of the iPad itself for books with less traditional formats, particularly those that are image-heavy like graphic novels. What will the purveyors of rare books do, if anything, in light of this new technology?
The article also brought to mind the perennial complaint of some of my fellow bookworms, who acknowledge the utility of digital texts but bemoan the loss of “chance” literary encounters. Miller:
However far we want to take it, metadata is essential to helping readers find books when they don’t have a real, physical bookstore to wander through.
Interesting, but I actually don’t see metadata as a surrogate for bookstores. If anything, iBooks is – in its imperfection – the closest thing we have to that experience in the digital sphere. I have been to a few REALLY good bookstores in my life, but I have to say that the majority of brick and mortars I have patronized have left much to be desired, and did in fact organize their shelves according to sweeping genre distinctions (“Literature”) and then… alphabetically by author. I don’t know – the clumsy process of wading through a long list of texts on a site like gutenberg.org actually does feel a lot like browsing through the stacks (a pastime with which I am well acquainted), with the added benefit that the books are free to keep (and print and mark and not return). Of course, I also look forward to a time in which more developed algorithms help me find texts I might not have come across on my own, and hope to be able to sort through this treasure trove of information according to both supply AND demand; that these models are not mutually exclusive. I suspect that they are not.
There is a limit, however, to my love of the algorithm. Miller mentions the ways Amazon’s Whispersync could be used to provide user data on reading patterns.
Why not, Lamere suggests, aggregate the data and tell us which books are read all the way to the finish, bought but never read at all or, conversely, reread repeatedly? Which books are most frequently abandoned, and is it always at about the same point — a particularly description-heavy chapter or patch of tedious exposition?
While Miller expresses ambivalence about this idea, I’m going to go out on a limb and say that it pretty much horrifies me. Let me save you the trouble, Amazon. Lots of people don’t make it through In Search of Lost Time. Does this mean that the volumes should be trimmed? Or that that the next time a writer sits down with his or her editor, they should be confronted with user pattern data and told that descriptions can be a real downer? There are plenty of books out there already designed with the marketplace in mind, and this is not American Idol. As we try to balance possibility and responsibility on the fly, we can’t help but note that it’s a fine line we walk.