ARTS AND CULTURE
Are Canadian Novels So Boring That Even a Machine Could Write Them?
Man studies web development, sends CanLit army into fits of giggles.
Quickly, which of these is not a plot description for a real Canadian novel?
(1) A woman returns to her hometown to find her missing father, but instead gets lost in the wilderness … and madness.
(2) A woman stands at the sink of a simple cabin to reconnect with the false, nostalgic simplicity of her youth.
(3) A lonely archivist heads into the woods of Northern Ontario and begins a sexual relationship with a bear.
Numbers 1 and 3 are real books, namely Margaret Atwood’s Surfacing, and Bear by Marian Engel. The made-up plot is 2, which was the “brainchild” of a computer algorithm created by Adam Brady of Toronto.
But truth be told, #2 probably does describe dozens of Canadian novels, which is why literature lovers are chattering over the recently launched CanLit Premise Generator.
Try it and see: The generator whips up instant plot synopses for made-up Canadian novels, especially the wistful kind full of trees and dogs and emotions. (The plot conclusion “… finding community where one would least expect” seems to come up more often than random chance, and that is how Canadian novels are in real life, too.) Not only does the CanLit Premise Generator make auto-satire, you can use it to get halfway to an arts council grant!
The generator’s inventor, Adam Brady, has been (wistfully?) studying web development at HackerYou (which is in Toronto, not a cabin in the woods, but OK) and was merely trying out his burgeoning skills when he developed the algorithm and code for the premise generator this month. His friends, writers Paul Fowler and Chuck Kotowych., helped with the concept. Launched on Family Day, the page quickly got passed from person to person until days later The Globe and Mail’s books editor was recommending it for a Giller.
(Meanwhile, back in England, Victoria Glendenning must be smiling: In 2009, the Giller Prize invited the British novelist to judge the annual CanLit contest. She panned the preponderance of “unbelievably dreadful” Canadian novels with protagonists who “in emotional turmoil … sit, brooding, on Muskoka chairs.” Yeah, so?)
Brady says his takedown of Canadian literature was motivated by a love for it, not disdain. “I love most Canadian writers. My current favourites are probably André Alexis and Michael Ondaatje,” he said in an interview with Billy. “But like any genre, (CanLit) comes with a set of conventions and tropes that sometimes make it ripe for parody.”
He explains that the algorithm simply smooshes together a protagonist, an action and a result. “It mixes up everything from a set of lists that we have,” Brady says.
Want to contribute ideas for the matrix? You can do so here. Among Brady’s favourite recent suggestions: the mayor of Timmins, Ont. as a main character, “Vancouver office job” as a setting, and “opens a craft brewery” as a plot point.
Brady estimates the possible permutations have reached 900,000, but that’s only growing as more new elements are plugged into the matrix. With so many combinations there’s no telling what you’ll get each time you hit the “write me another” button. (See below: I even got one about my former roommate!)
Could this lead somewhere more lucrative for Brady? Will he perhaps find community where he would least expect? A man from Toronto builds an algorithm, and parlays it into a job coaching struggling Canadian novel writers in the mountainside idyll of the Banff Centre?
In truth, Brady says, the ultimate payoff would be finding out that someone actually had been inspired by his generator to write a genuine book. “That would be amazing. I wouldn't be able to contain my excitement,” he says.
Whoa, now. Excitement? Be careful, we’re pretty sure too much of that will cost you a Giller.