WonderCat

The first time I taught literature without a reading list, I thought it was my job to make personalized recommendations for my students. This was an overwhelming task that my students helped me realize was unnecessary (they did this primarily by not reading the things I suggested). Though most didn’t want my recommendations, they did report feeling stuck when it was time to decide what they would read. I winced as I observed them Googling for ideas, and then I had a vision. I dreamed of a tool that would allow students to see what other students had read and what experiences and narration technologies they had encountered with those stories. More specifically, I imagined a network graph that would illustrate the connections between specific students, the texts they’d chosen to read, and the experiences and technologies they reported for those texts.

My thinking was that if students could see what other students had read and also see the experiences those students reported having with those stories, this could inspire some enthusiasm for the LIMITLESS ENJOYMENT TO BE FOUND IN THE ARCHIVE OF HUMAN CREATIVITY!! (not to overstate the power of reading or anything). I found collaborators and we are working together to create WonderCat.

The structure for this tool was inspired by the invention-finding method at the center of Angus Fletcher’s Wonderworks. In his introduction, Fletcher describes this method by narrating Aristotle’s departure from Plato’s Academy over two thousand years ago. Aristotle, Fletcher explains, renounced “the academy’s commitment to reason and reason alone” and:

Ambled off to chat with beekeepers, vivisect bird eggs, catalogue mortal emotions, classify wildflowers, and chronicle stage shows, laying the empirical foundations of zoology, physiology, psychology, botany, and dramaturgy. And somewhere in the course of these investigations into the nuts and bolts of life, he discovered the method for finding all of literature’s inventions. […] Aristotle’s account of the invention-finding method appears in a cragged paragraph that falls about one-third of the way into a treatise known as the Poetics. […] As knotty as the prose paragraphs of the Poetics can be, they manage to preserve the invention-finding method with enough clarity for us to recover its essential action. That action consists of two linked steps: first, identify what literature does, and second, work backward to uncover how literature does it. The what is the specific psychological effect of a literary work; an effect that’s typically linked to emotion. The how is the unique literary invention that drives the effect; an invention that’s usually engineered from one of the core elements of narrative: plot, character, storyworld, and narrator. In theory, this two-step process of reverse engineering literature is simple. In practice, it’s a bit more difficult. But the more we practice, the easier it gets (14-15).

Wonderworks models a new (but actually very old) way of engaging with literature; this is what I most appreciated when I first read it. Fletcher’s chapters aren’t organized around interpretive arguments. Each chapter introduces a literary technique through the story of its invention and then outlines experiences that technique can prompt in the humans engaging with it. And the end of almost every chapter comes with a suggestion of other stories that use the same technique. By organizing chapters around specific inventions, Fletcher reveals connections between creative works emerging from different historical moments. One of my favorite chapters explores the experience of bias reduction through the technique of the double alien, which Ursula K. Le Guin wielded masterfully in The Left Hand of Darkness, an innovation on Jonathan Swift’s use of the technique in Gulliver’s Travels. At the end of the chapter, Fletcher suggests other narratives that use the same technique:

You can find Ursula Le Guin’s bias-busting blueprint in Paul Theroux’s The Great Railway Bazaar: By Train Through Asia (1975), Marjane Satrapi’s Persepolis 2: The Story of a Return (2000), and many other modern travelogues. Or if you’d like to take the next step into science fiction and discover how new narrative technologies can manipulate our expectations, try playing the video game Bioshock (2007) (325).

While Fletcher’s chapters provide specific examples of literary inventions and suggest other more recently published stories using those inventions, I recognized that this was only the beginning. I realized that the effects and techniques Fletcher identified (I call them experiences and technologies) could form the basis of a tool that could expand to include experiences and technologies from many readers. In this structure, Darwin’s use of suspense in On the Origin of Species could be connected to Jordan Peele’s Get Out.

When I first had this idea, I didn’t know how to create a network graph. I learned quickly that we would need a relational database. We would essentially be creating lots of different tables (or spreadsheets) that would be linked by the terms in our controlled vocabularies (glossaries). We are chronicling the experience of building WonderCat in detail elsewhere, so I’ll just share here that it is designed to function in three interwoven ways.

• First, it is a discovery tool. If you’re looking for something to read, you can peruse the visualizations in the tool to find stories according to the experiences they prompted and the technologies identified by WonderCat users. We are adding features that make it possible to filter creative works by publication dates and locations, making it more useful for classes focused on specific literary periods.
• WonderCat is also an open pedagogy project. It is growing with user contributions. All users are encouraged to create an account and enter their story experiences. Each experience entered is included in the visualizations (unless the experience is set to private). WonderCat makes it possible for you to put yourself into the visualization to see how your experiences connect to other users. It also makes it possible for you to keep your own private set of nicely organized experiences with story.
• Finally, the WonderCat editorial team publishes glossaries that are always expanding to more accurately describe the way literature works in the world. We recognize that the terms we currently have in our taxonomies do not capture the full range of human experiences with story (how could they?!), nor does it include every narration technology that exists. WonderCat users are encouraged to propose new terms for these taxonomies and are credited publicly for their contributions at the website.

WonderCat is a database, not a recommender system, but my description of relational data might sound a little like the systems that deliver recommendations in streaming services and online shopping sites (“because you watched…”, “customers who bought…also bought…”). The difference is that on those sites, the recommendations are computed with proprietary algorithms drawing on your behavior and demographic data. Those algorithms are built around assumptions about your identity and can’t know your actual experiences (if you want to think more about how algorithms use your behavior and demographics to make predictions about you, check if you have a Google adsense profile). I think it’s better to have some agency in these choices and so I’m quite proud that WonderCat is a database that you can search, not a recommender system making predictions about what you will like.