When I started college, in September 1966, at Chico State (in Northern California), one of the seven courses I was taking in the first semester (crazy, yes) was Honors English, taught by Lennis Dunlap. (The course continued under the same rubric in the Spring semester.) “Mr. Dunlap,” as we were told to call him—he didn’t have a PhD—was an extraordinary figure: the most immensely confident teacher I ever had at any level. He dressed beautifully (he was said to have an independent income), in marked contrast to most professors, and had the posture of an equestrian. He came from the South (Tennessee, I think), as did his very charming wife, but he had a cosmopolitan manner. He was a superb teacher, though his wit could be cutting.
Early in the semester, we were assigned to write a short extempore essay about Saul Bellow’s novella Seize the Day. In the next class session, Mr. Dunlap read my essay aloud, praising it without reservation. Some of my fellow students, alas, hated me thereafter, but I practically floated out of the classroom that day. Before long I was grading papers for Mr. Dunlap (papers written by students in other courses he was teaching); he paid for this out of his own pocket.
Our text for the year-long course was an anthology titled The Forms of Fiction (Random House,1962), edited by Mr. Dunlap and a certain John Gardner (yes, that John Gardner), who had taught at Chico from 1959 to 1962. He and Mr. Dunlap had become good friends, and they worked together on a literary mag. From Gardner, Mr. Dunlap predicted great things (quite rightly). In addition to that anthology, we read other books for the class (mostly in translation, and not restricted to fiction; a Beckett play or two, for instance).
If you can find an old copy of The Forms of Fiction, you should acquire it. The first section of the anthology focuses on “Form and Its Relationship to Meaning” as seen in “The Modern Writer’s Use of the Sketch, Fable, Yarn, and Tale.” The epigraph on the title page is from Coleridge: “. . . nothing can permanently please, which does not contain in itself the reason why it should be so, and not otherwise.” Sixty years on, I wondered how professors today (working against the grain) might go about teaching the forms of fiction. And that in turn led me to think about AI and fiction, a subject that I normally avoid because I find it infinitely depressing.
Hugh Kenner, from whom I learned more than from any other writer, was very interested in it; one of his late books was a collaboration with Charles O. Hartman, Sentences (1995):
The child of the playfully perverse marriage of a 19th century grammar school book and a contemporary computer technology, Sentences is a surprising lyrical romp. The noted literary scholar and poet Charles O. Hartman discovered a book titled Sentences for Analysis and Parsing, which was used at the Thayer Street Grammar School in Providence, RI. Hartman took this somewhat didactic text and ran it through the TRAVESTY computer program developed by Hugh Kenner and Joseph O’Rourke. Working with Kenner, Hartman then ran the result through his own DIASTEXT program, based on the “diastic” writings of poet Jackson Mac Low. The result is a wonderfully eccentric poetry that resembles something between Gertrude Stein and Surrealist automatic writing. An inspired collaboration between technology and human literacy, Sentences is an innovative delight from beginning to end.
I must retrieve my copy from the stacks. Kenner was a tough customer, and the prospects of AI intrigued him early on. I continue to feel mostly a sense of overwhelming dread at the prospect of What’s Coming, but I have dutifully read what could be called optimistic takes on the subject: Dennis Yi Tenen’s Literary Theory for Robots: How Computers Learned to Write, published last year, for instance, and—just out, from a scholar I admire—Webb Keane’s Animals Robots Gods: Adventures in the Moral Imagination.
There has been one glimmer of optimism. Faithful readers of my Prufrock column may recall that last June, I wrote about reading this or that book filed under nonfiction and finding that, before you have even finished the introduction, “your mind, unbidden, is spinning a tale.” I gave a couple of examples—Rosamund Johnson’s Red Tape: Radio and Politics in Czechoslovakia, 1945-1969, for instance, a subject rich with potential. Something like this happens to me literally every day, sometimes several times in a single day. This has been a regular feature of my mental life for many years. I’m not bragging about it (“Oh, what an interesting fellow I am!”) nor apologizing. Just the facts. Another example of a “prompt” I gave in that column was Descartes’ Meditative Turn: Cartesian Thought as a Spiritual Practice, by Christopher J. Wild; what struck me was a line on the back of the book: “a dramatic conversion experience in southern Germany in the winter of 1619.” The germ of a novel!
Then something interesting happened between last June and now. One day, out of the blue, it occurred to me that a writer of such a bent (a writer for whom such random “prompts” were energizing) might be able to use AI, judicially, as a tool to generate juicy material. Maybe the prospects aren’t entirely bleak after all.
Dear Mr. Wilson-
I recently graduated from an MFA program that required a thesis containing both a full book manuscript and a scholarly introduction+bibliography. I chose to highlight form in my scholarly work because I was specifically drawn to the form of the novella, and I also worked with AI on my project.
Regarding form, I enjoyed 13 Ways of Looking at the Novel by Jane Smiley. In addition, though it arrived at lessons on form through specific craft elements, I enjoyed The House of Fiction by Gordon and Tate. My advisers said that my interest in looking at form and genre analytically--in the way you write in this column-- was just one aesthetic approach to writing among many. I found that interesting. Would you be able to comment on "understanding form as an aesthetic preference"?
Regarding AI, due to personal desires to avoid plagiarizing, the tool did not generate any of my sentences; it simply aided exploratory thinking. In hindsight, I think I would have preferred working with a human creative partner, but at the time, I only felt gratified by the speed and efficiency of on-demand engagement with an assistant. AI makes an adequate substitute for exploratory conversations and brainstorming, but it doesn't fulfill or replace the special event of creating with another human person. As an analogy: in a pinch, bananas can work as binder in baked goods, but eggs really do take the flavor, richness, and consistency to a new level.
Love your column on Prufrock. Thank you for writing.