What Twitter's For: Twitter Vs. Traditional Media on Covering the Capture of the Alleged Boston Bombers

Back in the day, whether you learned journalism in a class or from an editor, you probably heard something like the following: "If your mom tells you she loves you, get two confirming sources." Back in that very same pre-Twitter day, the immovable object of confirming a story stood athwart the irresistible force of getting the story out -- preferably before any other competing news source. As far back as the 1920s, that ever-recurring opposition of mandates created the tension in Ben Hecht's brilliant play (then movie) The Front Page.

As the standoff in Watertown unfolded Thursday night and especially early Friday morning, then came to its conclusion Friday evening, I found myself glued not to the prInternet or TV news, but to Twitter. I have often asked what the hell Twitter is supposed to be for, and that night, morning and evening I got my answer. In the initial hours of the confrontation and capture, the Twitter feed reported everything more quickly than the live TV feed and the the Internet versions of print media -- and ultimately more accurately too.

Unsurprisingly, rumors and speculation ran rife on Twitter. The Reddit-fed idea that one of the suspects was a missing Brown student made for a particularly sad and virulent example, even on @dannysullivan's watertown list, which only journalists on the scene could join. (Note that Reddit apologized for that rumor.) But those rumors were just as effectively quashed, and a surprising percentage of the events reported on Twitter proved accurate, unlike those on, for example, the New York Times website, which initially reported that one suspect was in custody and two were the object of an FBI and BPD manhunt. (Actually, one suspect at the time was dead, one was being sought by the police.) And the Twitterverse's mistakes were at least entertaining, like the "naked man" mistaken for one of the shooters, who turned out to be a local caught in the perimeter and forced to strip by the cops on the scene, and whom journalist and Harvard professor Seth Mnookin described as "in a very wrong place at a very wrong time." (Speaking of Mnookin, many of us following his informative on-the-scene tweets, including -- randomly enough -- Anonymous, were also rooting for him to find a charger for his cellphone, and not to have his car, which had a large black backpack in the rear seat, blown up by the police.) In the standoff's final hours, virtually every aspect -- the gunshots, the boat, the use of flashbangs, the negotiation, the surrende reported on Twitter turned out accurate, and appeared seconds to minutes to hours before it appeared on a newspaper website or television.

In contrast, in the early part of the confrontation, Thursday night and Friday morning, television feeds continued to show the same video clips over and over, and the websites of print media remained unupdated for hours after Twitter had reported events belying the media sites' accounts. (The traditional media reporting Friday evening did follow the events much more closely than they had in the early hours of Friday morning.) Nor, as noted, did these long delays add to the traditional media's accuracy; partially accurate and inaccurate rumors and speculation were no less common on the traditional media than on Twitter, but on Twitter, refutations of false reports came far more swiftly. Twitter's mix of true reports, rumors and outright falsehoods came unmediated by producers and editors to anyone following the story; the traditional media's narrative, on the other hand, dribbled out, and the "news judgment" that theoretically comprises the value added by traditional media's intermediaries did not result in less-flawed reporting. In the choice between "faster" and "more-accurate", the traditional media chose "none of the above."

I'd add that the traditional media did a much better job reporting that final standoff with Dzohkhar Tsarnaev than they did the initial armed confrontation of the Tsarnaev brothers with the police, most notably the incredibly competent Scott Yount of New England Cable News, whose feed I watched as the police surrounded then captured the suspect last Friday. By good fortune, the NECN team had a much better vantage point than the other news organs, and I admired Yount's detailed and descriptive reports, all but free of speculation not supported by facts on hand. By that point, the Twitter feed largely just repeated the same information going out via television news and print/web outlets. Perhaps the traditional media outlets had geared up for this final showdown by then; perhaps they simply had more of their employees working in the early evening than in the wee hours of the morning. Whatever the explanation, television and prInternet sites did much better reporting the final hours than they did the first confrontation of the bombing suspects with police.

So, in the end, I think Twitter outshined the television and print/web-based media at reporting the breaking story on a moment-to-moment basis. The lack of filtering in social media, interestingly, made for a more, not less accurate feed, as rumor competed for bandwidth with fact and largely (though by no means completely) lost out. Perhaps a little paradoxically, not having editorial funnel-points and a need for confirmation got the story out both more quickly and more accurately. When in the midst of a breaking story -- a true breaking story, one not previously expected and therefore one that the media cannot prepare to cover -- Twitter does outperform the traditional media. I don't know what this might imply about the future of news coverage, if anything, but the next time a truly startling and unexpected story breaks, I won't turn on the TV to follow it; I'll turn to Twitter.