Saturday, July 17, 2010

Experimental

Vanity + curiosity + internet access = a non-scientifically sound literary experimentation

The first mentions of I Write Like piqued little more than a dismissive notice. Increasing mentions of the novelty program, including this article that made my home page's list of "Most Popular" tidbits, garnered more interest. Intriguing to see how a computer program could compare vocabulary, syntax, and punctuation to attempt to classify writing samples by a sample of published authors. Just for grins, I decided to enter in the most recent ohbutno post. Highlight. Click Copy. Switch tabs. Click Paste. Submit, and "Voila!" A wail of, "Nooooo!" follows the revelation that the (admittedly incomplete) list of authors found a best match in Dan Brown.
Convinced that the program has a glitch, I enter the preceding post. And get a little puffed up over the William Shakespeare response. Perplexed by the inability to reconcile two such disparate authors, the idea to enter a month's worth of posts takes root. And I take action in a flurry of point and click punctuated by hisses, groans, moans, gleeful crowing, and a, "Say what?!" or two. Out of thirteen posts for the month of July, the program produced the following results:
David Foster Wallace: 23%
Dan Brown: 15%
William Shakespeare, Jonathon Swift, Arthur C. Clarke, Arthur Conan Doyle, Ursula LeGuin, Bram Stoker, Kurt Vonnegut: 7.75% ea.
And this particular post? Put Dan Brown in a tie with Wallace at just over 21%. I really should have quit while I was ahead.

No comments: