Using AI to Rediscover Jean Toomer
I was a teenager when I discovered “Cane” at the local library. I recall immersing myself in the novel, which is structured as a series of short stories and poems about Black life at the beginning of the 20th century. “Karintha” stayed with me all these years because of the images it called up in my mind:
Her skin is like dusk on the eastern horizon,
O can’t you see it,
O can’t you see it,
Her skin is like dusk on the eastern horizon
. . . When the sun goes down.
The girl in the story was about my age at the time. Some of the things that happened resonated with me and last week I decided to revisit the story and other stories from Jean Toomer’s novel. I started with “Karintha” and used text prompts (in Midjourney) that came from my reading of the story: from a child running in a dusty road to a “growing thing ripened too soon”. I imagined her has the child and as the woman she would become.
The process to create these images is a collaboration… with a machine. I had to reinterpret the text from “Karintha” in a way the AI could better understand. It took a few tries and I could probably re-spool the tool to generate more if I had the time. I selected two I liked (see above) that illustrate the text from the story. The I moved on to more of Toomer’s work. The next few images are inspired by “Tell Me”, “Her Lips Are Copper Wire”, and “The Lost Dancer.”
Each of these images became part of a series of images inspired by Toomer’s work. They are unique visualizations of Black American literature made possible by natural language processing AI. I also used this process with song lyrics (ex. Stevie Wonder) and “Song of Solomon” by Toni Morrison.
But back to Jean Toomer who is likely the least recognized or talked about writer in literary circles today. There is space for this emerging type of analysis in the humanities, as well as graphic design and illustration. More important (to me), is the opportunity for readers to unpack things they’ve read and attempt to capture their ideas about the texts in visual images generated by Midjourney and other NLP AI tools.