I am disappointed by mainstream reviews of Anathem. Reviewers appear unable to judge speculative fiction on its own terms. They want books to be about character psychology. To which my response is, let them read Proust. I, for one, find Proust dreadfully boring. When Michael Dirda writes that much of the book sounds like "transcribing intellectual conversations that sound like really nerdy Caltech grad students schmoozing at 3 a.m." and he takes that to be a bad thing, he hits the nail on the head: books like this are not written for people like him, much like A la recherche du temps perdu was not written for people like me.
"Sex is referred to, but never seen," complains Dirda, while lamenting the detailed descriptions of "buildings, machines, and events." To me, when Stephenson does write about sex (the Baroque Cycle is full of it), I usually skip it. Not that he does a bad job of describing sex, but I already know how straight people go about rubbing their private parts together, purely from hearsay, but that's quite enough, thank you very much. Same with fights: Stephenson does a great job of fight scenes, but there's little novelty in the idea of a fight scene.
The descriptions of buildings, on the other hand, are fascinating. They have nothing to do with the usual scene-setting descriptions in an otherwise psychological novel. They are an integral part of the theme of the book.
The theme is long-run institutional design. How do you design an institution such that it would not only be useful if it were to persist, but such that it will in fact persist? It is the problem tackled in The Federalist, and it is the problem faced by those seeking to design nuclear waste disposal facilities or the Clock of the Long Now.
Much of the intricate description of building layouts in Anathem is an attempt to solve such a problem by architectural means. The architecture of Saunt Edhar's largely determines the modes of communication. There are overlapping interests and mutual expectations related to sticky architectural features that are hard to renegotiate because of limitations imposed by those same architectural features.
I suppose Dirda probably finds the Federalist Papers rather dry, too, lacking as they are in sex scenes and character development. But Anathem is a novel of ideas, and any particular reviewer's failure to appreciate those ideas is no reason to condemn it.
