Publishing is a word that, like the book, is almost but not quite a proxy for the “business of literature.” Current accounts of publishing have the industry about as imperiled as the book, and the presumption is that if we lose publishing, we lose good books. Yet what we have right now is a system that produces great literature in spite of itself. We have come to believe that the taste-making, genius-discerning editorial activity attached to the selection, packaging, printing, and distribution of books to retailers is central to the value of literature. We believe it protects us from the shameful indulgence of too many books by insisting on a rigorous, abstemious diet. Critiques of publishing often focus on its corporate or capitalist nature, arguing that the profit motive retards decisions that would otherwise be based on pure literary merit. But capitalism per se and the market forces that both animate and pre-suppose it aren't the problem. They are, in fact, what brought literature and the author into being.
The story of the book as technology—the book as revolutionary, disruptive technology—must be told honestly, without triumphalism or defeatism, without hope, without despair, just as Isak Dinesen admonished us to write. A great challenge in producing such an account, however, is the “availability heuristic.” This is a model of cognitive psychology first proposed in 1973 by Nobel laureate Daniel Kahneman and his colleague Amos Tversky, which describes how humans make decisions based on information that is relatively easy to recall. The things that we easily recall are things that happen frequently, and so making decisions based on the samples we have at hand would seem to make sense. The sun rises every day; we infer from this that the sun rises every day. A turkey is fed every day; it infers that it will be fed every day—until, suddenly, it isn't. Heuristics are great until they aren't. A person sees several news stories of cats leaping out of tall trees and surviving, so he believes that cats must be robust to long falls. These kinds of news reports are far more prevalent than ones where a cat falls to its d**h, which is the more common event. But since it is less reported on, it is not readily available to a person for him to make judgments.
Publishing is tremendously susceptible to the availability heuristic for two significant reasons. First, prior to recent innovations, man*scripts not published were unavailable for an*lysis. So the universe of knowledge we have about books, literature, and publishing excludes that universe of books that were never published. It also mostly excludes those books that were commercial or critical failures. One doesn't see books that don't sell, not on store bookshelves or in friends' houses, not on Top Ten lists, not on Twitter, not in the Times (London, New York, Irish), and so on.
There are books in the data set now, such as Leaves of Gra**, that were self-published, and others, such as Moby-Dick, that were ignored in their time but reappeared through good luck. The novelist Paula Fox published, vanished, published again. Her reappearance is a triumph of publishing. But what about all the unrediscovered Paula Foxes? Or, for that matter, what about all the books I published at Soft Skull in the 2000s that had been rejected by ten, twenty, thirty, sixty publishers? And what about the man*scripts I rejected at Soft Skull that I would subsequently see published by prestigious publishers large and small? Is this proof of the effectiveness of the existing system for the production and dissemination of literature? It's quite clear that while we do our best, our output is as much proof of the awfulness of the system as it is of its strengths. Much like Patty Hearst, we cannot bear to consider the alternative.