There's also a way that you have of being precise but also allusive, that works well for me - it's something about the open-hearted way you frame your queries. Instead of feeling daunted or discouraged, I feel excited to give whatever it is a try. This takes a lot of editorial wisdom and confidence - to know just how to get the writer to take that extra chance.

My mind has an obsessive, neurotic quality, but I also have a very hard work ethic. I know that a lot of people think, oh, you have a wild imagination. I don't, really. If ten people are sitting around and someone says, "Hey complete this sentence in a funny way," I'm never ahead of the pack.

My idea about collections is that you write as hard as you can for some period and what you're really doing during that time is hyper-focusing on the individual pieces - trying to make each one sit up and really do some surprising work.

Both [Donald] Trump and Bernie [Sanders] got to this idea of the vanishing middle class sooner and with more passion than more mainstream politicians, and benefited from it. The difference, of course, is that Bernie understood this in a more compassionate framework, and talked about it in conjunction with a revitalization of another part of the American project, which is the notion that we are all created equal, and that our laws and culture and action ought to reflect that.

I'm not sure I would call it agony but there is a kind of cyclic frustration. You get one story right and then here comes another one. When does that end? What I'm trying to do is get it to end right now, by recognizing that that cycle is writing. That is: trying to understand the frustrations and setbacks (and agony) as part of a bigger chess game you are playing with art itself.

I'm repledging myself to human-scale values. As a fiction writer, the best data comes through the senses and is then processed through many revisions. We have to learn to be intelligent assessors of the data coming in to us and what it's doing to our mental process.

I was trained in seismic prospecting. We'd drill a deep hole and put dynamite in the bottom and blow it up remotely, which would give you a cross-sectional picture of the subsurface, which tells you where to drill.