Writing is an incessant process of discovery.

When Emily Dickinson's poems were published in the 1890s, they were a best-seller; the first book of her poems went through eleven editions of a print run of about 400. So the first print run out of Boston for a first book of poems was 400 for a country that had fifty million people in it. Now a first print run for a first book is maybe 2,000? So that's a five-time increase in the expectation of readership. Probably the audience is almost exactly the same size as it was in 1900, if you just took that one example.

It was one of the great mass achievements of American civilization, and we did it because we thought if you were going to have a democratic form of government, people had to be able to read and understand complicated ideas on their own.

The ideal of universal literacy, in the West anyway, was first of all a Protestant idea - that everybody had to be able to read to save their soul. That idea got transposed into an idea of the importance of literacy for democratic citizenship.

Fiction writers have their own world, and poets have their own world, and literary criticism has sort of passed over into cultural studies in the university, and so on. They seem more disconnected from each other than they did when I first began to write.

The problem with describing poets as legislators is that at that level of politics - politics as political invention - poets have no special skills and are not apt to.

It's clear that there has to be some play between the vitality of invention in economic life and some regulation of it, and in some ways the great ideological wars of the 20th century that cost so many lives had to do with whether to have managed economies directed by government or economies directed by the free movement of capital, which is only partially subject to government regulation.