Programmer unleashes bad-grammar bot on Twitter


Caitlin Dewey,The Washington Post
Tuesday 11 December 2012 09:51 GMT
Twitter: No place for bad grammar
Twitter: No place for bad grammar (Getty Images)

For a soft-spoken programmer from Buffalo, Nate Fanaro gets a lot of hate mail.

Every day, his Twitter queue fills up with messages telling him to die or delete his account. "I find you extremely annoying," one caller said in a voice mail. "You make little girls cry. What's your problem?" said another.

Fanaro is not a hacker. He doesn't take down Web sites or swipe credit card numbers. Rather, the 30-year-old prankster is the creator of the Twitter grammar bot @CapsCop, an automated account that finds people who tweet in all caps and, within seconds, fires a snarky correction back at them: "Give lowercase a chance," perhaps, or "On Twitter, no one can hear you scream."

The technology behind such bots is simple, which helps explain why so many tech-savvy grammarians have launched their own. Programmers need only write a script to search Twitter's data and respond to certain phrases, and they're well on their way to Twitter infamy. Some accounts reply to users directly, while others retweet the offending messages.

Teachers, parents and other curmudgeons have long blamed texting and social media for the general decline of the English language. Considering the widespread disregard for grammar in certain corners of the Internet, they could "b 4given" for thinking that kids these days can't write. (Because while we're sometimes talking about outright mistakes, we're also talking about the grammar-agnostic spirit that has evolved online.)

Although Twitter may seem like a stronghold of sloppy writing and acronym-happy Internet slang, a number of vigilantes are hilariously and controversially fighting back.


Bots such as Fanaro's ping unsuspecting Twitter users with sarcastic corrections. Anonymous copy editors such as @fiercek send gentle revisions to work tweeted by writers and reporters. One of the newest accounts, a wildly popular project by Buzzfeed reporter Andrew Kaczynski, seeks to publicly shame users who tweet things like "speak English your in America omg." "I think you mean 'you're' in America. That's embarrassing," @YourInAmerica tweeted back to that one.

Since its launch in late November, Kaczynski's account — which exclusively targets the phrase "your in America" — has attracted 18,000 followers and plenty of praise from media outlets such as Latina magazine, which lauded him for launching a "grammar crusade" against "outraged nativists." But Kaczynski, who can be found tweeting pictures of adorable hamsters and politicians' Christmas trees in his spare time, treats the attention like so much comment-box blather.

"I didn't set out with a purpose or anything," he says. "I'm not personally offended by it. It's just . . . funny."

"Funny" is a fair description of Twitter grammar. Many of the platform's ungrammatical but widely used conventions — such as confused homophones ("your the man") — are, indeed, laughable to readers schooled in what linguists call "standard English."

But the vigilantes that froth over them can be hilarious, as well: @StealthMountain, the most popular of the bots, exists solely to tweet "I think you mean 'sneak peek' " to users who type "sneak peak." @YourorYoure, which dates back to April 2009, pings users with a simple "[Wrong!]" when they misuse every first-grader's most-hated contraction. Eric Mortensen, the bot's creator, says he made it after seeing a co-worker's rage at an e-mail that confused the two.


Despite the online kerfuffle, most linguists agree that neither texting nor the Internet defile the English language. Consider the headline, says Tim Stowell, a linguist at the University of California at Los Angeles who studies syntax and specialized speech. Much like telegrams, diaries and cookbooks, headlines come with their own wonky set of grammatical rules. But people don't leave out pronouns, articles and conjunctions in spoken sentences just because headlines do.

In fact, Stowell says there is no evidence that any form of "specialized speech" has corrupted spoken or written English, and plenty of recent studies have come to the same conclusion. In September, researchers at Coventry University in Britain ruled that there's no link between text-message conventions, which are also used on Twitter, and bad spelling or grammar in other forums. A 2009 study from the University of Alberta concluded that text-speak should be viewed as a dialect that people can switch into and out of.

Why the panic, then? That might have less to do with Twitter and more to do with who's tweeting, Stowell says. According to the Pew Research Center's 2012 Twitter Use report, 26 percent of Internet users younger than 30 tweet, versus only 9 percent of users between 50 and 64. That makes Twitter's grammar abuses the perfect fodder for a little generational outrage.

Even Fanaro, the @CapsCop creator, says the concern about grammar on Twitter is much ado about nothing. While he has dedicated a fair chunk of time to correcting grammar on Twitter — his three-year-old bot now boasts an accompanying Web site and iPhone app, and has tweeted more than 1 million times — he doesn't take his own message "too seriously."

Fanaro doesn't correct people's grammar offline, and he admits to posting a few caps-locked tweets himself.

"But when I write an e-mail, I read it over and over and over before I click send," he says. "And I think that's something we could all keep in the back of our minds."

Join our commenting forum

Join thought-provoking conversations, follow other Independent readers and see their replies


Thank you for registering

Please refresh the page or navigate to another page on the site to be automatically logged inPlease refresh your browser to be logged in