Harry Mathews and the productive pleasures of constraints

Last week I urged you to go read the interview with Harry Mathews that the Paris Review published in 2007. A week later, I find I'm still carrying it around in my shoulder bag, flipping through it and thinking about it--especially about this exchange:

INTERVIEWER

Did Ashbery introduce you to any writers whose work you did read?

MATHEWS

Yes, thanks to John I began reading Raymond Roussel. Roussel had methodical approaches to writing fiction that completely excluded psychology. In the American novel, what else is there? If you don’t have psychology, people don’t see the words on the page. What was really holding me up was this idea that you had to have character development, relationships, and that this was the substance of the novel. Indeed, it is the substance of many novels, including extraordinary ones. But I had tried writing works involving psychology and characters and all that, and the results were terrible. In Roussel I discovered you could write prose the way you do poetry. You don’t approach it from the idea that what you have to say is inside you. It’s a materialist approach, for want of a better word. You make something. You give up expressing and start inventing.

INTERVIEWER

Which of Roussel's methods interested you?

MATHEWS

One method he used for short stories involved making the first and last sentence identical except for one letter. Each word has one meaning in the first sentence and a different one in the last. A word like train might be a choo-choo to start with and a trailing skirt-end afterward. In the longer works, he would take fragments of nursery rhymes and parrot them phonetically and then use the new words to construct a story. For instance, the song "J'ai du bon tabac" becomes "Jade tube onde aubade."

INTERVIEWER

What is the point of such a method? What does it achieve?

MATHEWS

It's very liberating. It allows you to make up something that you never would have if you didn't have this nasty problem to solve. For example, in Selected Declarations of Dependence I gave myself the task of writing a story using the one hundred and eight-five words that were found in forty-six proverbs. This is a forbiddingly small vocabulary. It was hard to know what to do with them. Then I started putting words together and a few words would lead to a sentence and then eventually it became this sweet love story. It was as though you were wandering through a jungle and suddenly you came into a clearing that is a beautifully composed garden. It's extraordinary, the feeling it gives you.

I happened to read that passage at what seemed to be the perfect moment: for the first time since grade school,* I was trying to write a song. I've been taking piano lessons lately, the most recent step in a process of re-learning that began eighteen months ago after twenty-five years away from the instrument. I can read music, and I can handle the keyboard tolerably well, but I've never really understood how music works, why I'm playing the notes I'm playing. So that's what I'm asking of my teacher: to simultaneously help me improve my technique and my understanding of music itself.

Which led him, two weeks ago, to give me the assignment of writing a new melody to the chords of "Amazing Grace." I spent hours on it . . . hours of sheer joy. And, to bring this back to Mathews, what it reminded me of most was attempting to write poetry back in my undergraduate days. The presiding spirit of the poetry program, Mary Kinzie, was a formalist at heart, and her approach to teaching poetry was to combine extensive reading with boot-camp-style immersion in formal structures. Whitman was verboten; Shakespeare's sonnets were to be gnawed to the marrow.

And it was effective. Form, for a beginning poet, forces attention, innovation, and persistence--and that's what I felt as I was struggling with my nascent melody: I had, essentially, a steeplechase course of chords through which I had to run it, and while my instincts would have led me to simpler formulations, to a quick resolution of the tension inherent in the first couple of chord changes, being forced to hew to the number of changes and measures I'd been given was oddly liberating. Mathews is startlingly right: when I discovered a way to handle it, it was like nothing so much as coming unexpectedly into a beautiful clearing.

Which makes me understand Mathews's desire to impose structure and limitations on the novel. Good god, why doesn't everyone want to do that? To deal with the illimitable, wholly protean form that we call the novel . . . how can anyone do that and not go insane?

{If you want to hear my song, I think this link should allow you to download a poor performance of it. I make no claims for it other than that it was a lot of fun to work on--and that I find myself whistling it sometimes.}