surprise as method

In any given project, if you’re working hard and keeping your eyes open, there ought to be a few surprises. It’s a bad sign if there aren’t any– a sign that you’re being a sort of ideologue and trying to compress everything you find into your pre-conceived gridwork rather than letting the world prove you wrong.

Almost always, I find that the surprises are the key to any project, and can/should be used to organize the project as a whole.