Computer Algorithm Recognizes Your Inner Faulkner A new writing "analyzer" has become all the rage on the Web. "I Write Like" analyzes your prose and tells you what famous author your writing most resembles. Turns out Robert Siegel's wordsmithing resembles that of David Foster Wallace. And Michele Norris' prose? Would you believe she writes like Mark Twain? Robert and Michele talk with Dmitry Chestnykh, programmer and founder of the site.
NPR logo

Computer Algorithm Recognizes Your Inner Faulkner

  • Download
  • <iframe src="https://www.npr.org/player/embed/128572009/128572963" width="100%" height="290" frameborder="0" scrolling="no" title="NPR embedded audio player">
  • Transcript
Computer Algorithm Recognizes Your Inner Faulkner

Computer Algorithm Recognizes Your Inner Faulkner

Computer Algorithm Recognizes Your Inner Faulkner

  • Download
  • <iframe src="https://www.npr.org/player/embed/128572009/128572963" width="100%" height="290" frameborder="0" scrolling="no" title="NPR embedded audio player">
  • Transcript

A new writing "analyzer" has become all the rage on the Web. "I Write Like" analyzes your prose and tells you what famous author your writing most resembles. Turns out Robert Siegel's wordsmithing resembles that of David Foster Wallace. And Michele Norris' prose? Would you believe she writes like Mark Twain? Robert and Michele talk with Dmitry Chestnykh, programmer and founder of the site.

MICHELE NORRIS, host:

A simple little website called I Write Like has really taken off in popularity this week.

ROBERT SIEGEL, host:

You just type or paste in some prose. The site analyzes the text and tells you that your writing is like that of, well, H.P. Lovecraft.

NORRIS: Or Margaret Atwood.

SIEGEL: Or William Shakespeare.

NORRIS: The site is the work of a computer whiz named Dmitry Chestnykh. And we reached him on a Skype line in Montenegro, and he tried to explain how it all works.

Mr. DMITRY CHESTNYKH (Creator, I Write Like): It's a simple algorithm, actually. It's called Bayesian classifier.

NORRIS: Dmitry Chestnykh says the technology is similar to the way an email program might be able to tell that one message is from your mother and another is simply spam.

SIEGEL: But perhaps like a spam filter, I Write Like's results can seem a little, well, random, at times.

NORRIS: Well, it can, Robert. I put a bit of text from my upcoming book, "The Grace of Silence," into the website. Here's part of the passage I inserted.

(Reading) Before the alarm clock goes off, before the smell of coffee or bacon find its way to your bedroom door, there's often a sound outside your window that jolts you out of slumber during the winter months in Minnesota. It's the scrape of a snow shovel against wet cement.

SIEGEL: Okay, you put that passage into I Write Like and it tells you that your prose most resembles what author?

NORRIS: Stephen King.

SIEGEL: Stephen King. Ah, the resemblance is unmistakable.

(Soundbite of laughter)

NORRIS: Pretty scary, huh?

SIEGEL: I tried something that's a bit more prosaic. I submitted part of a memo I wrote in response to an inventory request from NPR, an inventory of recording equipment that I might or might not have in my possession. Part of this memo reads as follows:

(Reading) Second, I do not have the following items: E3414 Shock Mount Fishpole Pistol Grip Battery Pack 12/1 Card Reader.

NORRIS: And I Write Like said you write like whom?

SIEGEL: Like David Foster Wallace.

NORRIS: The cadence, the rhythm, it really does sound like David Foster Wallace.

SIEGEL: In fact, I submitted lots of other things I had written that weren't inventory memos, and they all came out David Foster Wallace.

(Soundbite of laughter)

NORRIS: Well, there you go. Well, for his part, the site's author says he's working to make some improvements.

Mr. CHESTNYKH: I need to put more works of different writers to make it more accurate.

NORRIS: Dmitry Chestnykh. His site is called I Write Like.

SIEGEL: Stephen King.

NORRIS: If you say so.

Copyright © 2010 NPR. All rights reserved. Visit our website terms of use and permissions pages at www.npr.org for further information.

NPR transcripts are created on a rush deadline by Verb8tm, Inc., an NPR contractor, and produced using a proprietary transcription process developed with NPR. This text may not be in its final form and may be updated or revised in the future. Accuracy and availability may vary. The authoritative record of NPR’s programming is the audio record.