Miranda does her violin practice in the library/guest room; while she’s getting things ready, I spend some amount of time looking around at the shelves. Most of my books up there are nonfiction, and in particular I have a bunch of math books up there.
In order to keep my books under control, I have to periodically give books away. Which raises the question: why haven’t I given more of the math books away? I gave a bunch away when I left academia, and haven’t missed any of them; do I have any reason to keep around the remaining ones?
I’m increasingly thinking that the answer is, in general, no. Which makes me a little sad, but only a little; I’ve made my peace with that. What I was surprised to realize today, however, was that I actually do have a good reason to keep some of them around.
I reread books somewhat frequently; not as much as I like, but it’s still an important part of my life. I’m not in general in the habit of rereading math books these days, but I got to thinking: why not? In some cases, there’s a good reason: I might have read them for largely utilitarian reasons, they might demand a level of sustained concentration that I don’t want to invest my time in these days, or they might demand a familiarity with material that has slipped through my grasp over the last five or six years. Also, in general I read nonfiction because I think I’ll learn something, because it will change the way I think about something or do something, and that’s not likely to happen if I reread those math books.
So I’m not, for example, likely to reread my old textbooks. (Though I was amused to see Miranda pick up Basic Algebra I today. She decided it was a bit much for her, however.) But those aren’t the only reasons why I read nonfiction: sometimes, it’s just for the pleasure of the words, of the argument, of revisiting ideas that are old friends.
And some of the math books that I have qualify very well under those criteria. Local Fields, for example, is a fabulous book, and without the rigidity of presentation that charactizes a traditional textbook. I seem to recall thoroughly enjoying Bott and Tu as well; why not go through it again?
And, for that matter, why reject the idea of learning something new? The proof of the Weil Conjectures brought together some of the most important ideas of the twentieth century, beautiful ideas whose impact will still be making itself felt at the end of this century, and I never learned etale cohomology well enough to follow the proof to the end; maybe I should remedy that? For that matter, progress has hardly stopped since I, say, left grad school; are there any well-written monographs that have appeared over the last decade that I’d enjoy? (Any suggestions, Jordan?)
This post has not been revised since publication.