The Signal and the Noise
One tweeter boasted of a "game-changing victory" for crowdsourcing in the early hours of the Boston area manhunt. But what began as a low-grade fever on social media spiked with the wrongful naming of a bombing suspect. All the while, Nieman Visiting Fellow Hong Qu was testing his new tool Keepr as a screen for credibility and posting early results on Nieman Reports as the story unfolded. Qu and journalist Seth Mnookin, who tweeted live from the manhunt, write about how smartphones and their unprecedented power to publish require new journalistic tools and practices, while other Nieman Fellows consider the intersection of social media and journalism in the aftermath of the attac
@sethmnookin I was at home on the night of Thursday April 18th when I received a text alert from the Massachusetts Institute of Technology about a shooting on campus. I co-direct MIT's Graduate Program in Science Writing, and the Institute's students and faculty were already on edge. The Boston Marathon bombings had occurred just three days earlier, and it had been less than two months since the entire campus was locked down after a false report of a shooter was called in to the police. This latest message didn't say much more than, "Gunshots were reported. Please stay clear of the area until further notice." The FBI had released photos of the bombing suspects that afternoon, so I thought there might be a connection.
In the late 1990s, I had been a police reporter for The Palm Beach (Fla.) Post, and one of my primary responsibilities was to listen to local police scanners to make sure the paper didn't miss anything. As soon as I got the shooting alert from MIT, I turned on a Boston scanner and began monitoring the situation. I also followed developments on Twitter, but I didn't turn on my TV or check news organization websites. I knew they wouldn't have anything yet.
The number of units being called in and the urgency in the dispatchers' voices made it clear this was a big deal. The level of response was out of proportion to a simple shooting, even one on a college campus. It was the type of response that suggested there was either a significant ongoing threat or this incident was, in fact, related to the bombings—or both.
I've been at MIT for almost two years, but I still think of myself primarily as a journalist. I realized this was a situation in which my training as a reporter might allow me to add value to the coverage. That's why I drove to Kendall Square in Cambridge, where MIT is located and the shooting had occurred.
By the time I arrived, there were roughly 20 journalists gathered around a police perimeter. Before I left home, I had noticed that Taylor Dobbs, a journalism student at Northeastern and the son of a friend of mine, was tweeting from the scene. We hadn't met before, but I sent Taylor a message so we could connect in person. He was there with another Northeastern student, Brian D'Amico, a photographer.
Around 12:15 a.m., word came that Sean Collier, the MIT officer who had been shot, had died. I still had the police scanner going on my iPhone when things suddenly went from being relatively quiet to a state of total chaos. A pack of Massachusetts State Police cruisers started booking it down the street, and the three of us—Brian, Taylor and I—got in my car to follow them. There was talk on the scanner of a carjacking and of a possible suspect on the run.
We ended up at a gas station in Cambridge on nearby Memorial Drive, which we learned was where the carjacking victim might have escaped. Then Taylor
Photo: Brian D'Amico
10:48 p.m.
Gunshots at MIT building 32 (Stata Center). Area cordoned off; unknown if injuries.saw a tweet that said the stolen car, a black Mercedes SUV, had been tracked to Watertown. Seconds later, we heard, "Shots fired on officers in Watertown" over the scanner, and it was total chaos again. The police at the gas station took off toward Watertown, and we followed.
We got to Watertown a little before 1 a.m., moments after the shootout between police and Tamerlan and Dzhokhar Tsarnaev, the brothers suspected of carrying out the Marathon bombings, had ended. You could smell gunpowder and see smoke in the residential neighborhood near the Watertown and Arsenal Malls. The next morning we learned that hundreds of rounds of ammunition had been fired and homemade bombs had been deployed.
Photo: Andrew Kitzenberg
12:53 a.m.
Reports of massive explosions and grenades. Can smell smoke in Watertown near Arsenal Mall.@hqu I used Keepr, a social media monitoring software tool I am developing as a visiting fellow at the Nieman Foundation, to capture Seth's tweets from Watertown. Keepr's algorithm extracts credible real-time information from raw Twitter feeds by pulling the 100 most recent tweets from Twitter's API and counting the words and phrases that occur most frequently. It also discovers Twitter users with the most mentions.
During the manhunt, I used Keepr to identify reliable sources who appeared to be tweeting from the scene. I used four factors as indicators of credibility: disclosure of location, preferably via geocoding (Taylor Dobbs had activated the geocoding feature on his iPhone Twitter app that night, but Seth had not), multiple source verification (the tweets cited information from primary as well as other sources), original pictures or video, and accuracy over time.
Photo: Jennings Raske
1:00 a.m.
“All units retreat. There are explosives on scene.”
Boston PD officer down; Watertown ambulance transporting.Keepr's algorithm detects good sources during breaking news events by focusing on high numbers of tweets during short periods of time, tweet bursts containing frequent mentions of specific usernames (this is how Keepr found @sethmnookin, whom I had not been previously following), and extremely rapid rates of follower growth. Wikipedia founder Jimmy Wales followed Seth that night, tweeting at 2:39 a.m., "When I followed you awhile ago, you had 10,000 followers. Now 30,000. Be late, be right, and be safe."
During those chaotic early morning hours, Seth practiced a new form of networked journalism, one that combines the speed and immediacy of social media with best journalistic practice. Seth was plugged into the cellular grid with his smartphone, but he was also plugged into a spontaneous self-organizing online group that was both consuming and participating in his coverage.
In addition to reporting what was happening, Seth held a kind of rolling, live-streamed press conference in which he answered followers' questions, corrected misinformation spreading via social media, and distributed important public safety updates from the police. He also crowdsourced reporting assistance, enlisting followers' help with everything from identifying the most useful police scanner to locating an iPhone charger.
Social media groups can go terribly wrong, of course, as with the Reddit online community that mistakenly identified a missing Brown University student as a suspect. But professional journalists get things wrong, too. The Reddit misinformation was retweeted by several prominent journalists, and CNN, The AP, and The New York Post, among others, all made embarrassing errors.
This style of breaking news coverage will likely appeal to a new generation of consumers willing to tolerate high levels of uncertainty and constantly changing information in exchange for 'watching' an event unfold live. These audiences will self-organize to support and amplify anyone reporting in this way, regardless of whether that person has press credentials. So, as 2013 Nieman fellow Betsy O'Donovan argues in "The (New) Industry Standard," news organizations would do well to promote a greater understanding of journalistic standards and ethics among the general public.
@sethmnookin I parked my car at the intersection of Dexter and Nichols Avenues, where the police were focusing their search for someone they described as "suspect number two." When Taylor, Brian and I arrived, we were the only reporters on the scene. There were, however, perhaps over 100 cops, many with their guns drawn.
As we were trying to get our bearings, the police were also trying to get theirs. There were cops just wandering around the streets, and you could hear over their radios, "We need to go to Laurel Street," which was two blocks south of where we were and, as it turned out, near where Dzhokhar Tsarnaev was later discovered to be hiding in a boat.
I had a small reporter's notebook with me, but early on realized that tweeting would be a more effective way to take notes. Everything I wrote would be time stamped, and I wouldn't need to worry about not being able to
Photo: Seth Mnookin
1:45 a.m.
Arrest going on now. “He’s got shit in his pockets. Get down that street now!”
pic.twitter.com
/rnpWrZ2H2Gread my messy handwriting after the fact. This worked out even better than I expected. Because I knew my notes were going to be public, I spent more time thinking about whether something was important or informative or whether I was simply writing things down because I was nervous or had nothing else to do. One thing I regret is not taking more pictures and recording more video, especially at the outset, when there was no one else at the scene.
Minutes after we arrived in Watertown, it became clear the police were preparing for an arrest; the second suspect, it seemed, was in a car on Nichols Avenue. Suddenly, there was shouting all around us, some of which was directed at the suspect, whom the police feared was wearing an explosive vest, and some of which was directed at us, ordering us to get down the street as quickly as possible. It took more than 10 minutes, but the police eventually got the suspect out of his car and ordered him to strip.
Around 2:30 a.m., Massachusetts State Police spokesman David Procopio told reporters, "We have two suspects in custody." One was the person who had been apprehended in the gun battle with police (we learned later that
Photo: Taylor Dobbs
2:58 a.m.
“One more suspect at large. Two accounted for should be revised. One accounted for. One at large.”
- state police PIO.this was Tamerlan Tsarnaev, who had been killed); the other was the man who had stripped and been loaded into a police car. But the latter turned out to have no connection to the Marathon bombings or that night's shootout, and about 30 minutes later Procopio gave an update: "One more suspect at large. Two accounted for should be revised. One accounted for. One at large."
By 3:30 a.m., I felt the situation was becoming more stable. The police hadn't found the second suspect, but there were no reports of further gunfire. I checked and saw I had dozens of missed calls and e-mails from TV and radio producers. It wasn't until after 6 a.m. that I finally went home to see my wife and kids. By that point, Good Morning America had offered to put me up in the Four Seasons—provided that I promised not to appear on any other shows before going on the air with them. By the time the show was broadcast, its producers had decided they weren't interested in me, after all.
Historically, TV and radio have had the biggest competitive advantage in breaking news. But fluid, chaotic situations are also precisely those in which the transfer of information from cameraman to reporter to producer to anchor is most prone to error. Plus, when you're on the air, providing steady updates isn't an option—it's a necessity. And needing to fill airtime can cause problems of its own. On Twitter, if there's a new development every minute, you can update every minute; if nothing is happening, you can wait.
The Watertown manhunt illustrated the fact that there are times when a traditional journalist can do his job more efficiently and effectively on Twitter than in any other medium. I've worked in virtually every type of print outlet, from webzines to newspapers to newsweeklies to monthly glossies. I've blogged on my own website, on blog networks, and on traditional news sites. I've written books. I've been on the network's morning news shows and The Daily Show, on shock jock drive-time radio and Fresh Air. But for those three or four hours when a gunman was on the loose and a neighborhood was under siege, Twitter was the most efficient way to get information out to the public.
@hqu Twitter coverage of the manhunt in Watertown is a remarkable milestone for journalism. Even more remarkable are the implications for ordinary citizens who, without a press pass, can report news and influence coverage. For the latter group, this event instilled a newfound sense of power and responsibility in how they verify and disseminate news. Tools and processes for assessing source credibility need to catch up with social media technology and culture, especially in dangerous environments in which the public relies on reporters to provide actionable news updates with minimal misinformation and fallout.
Legacy media have a crucial role to play, both in providing original reporting and in curating social media. Seth's followers increased from a little over 8,000 before the night of the Watertown manhunt to over 45,000 during the few hours he was tweeting from the scene. That amplification was achieved largely through prominent journalists and celebrities (political commentator Keith Olbermann, New York Times reporter Brian Stelter, and actor/comedian Kumail Nanjiani, to name a few) and major news outlets (including ABC News and Foreign Policy) following and retweeting him. As 2013 Nieman fellow Ludovic Blecher explains in "Curation Is the Key," online breaking news forums, controlled and curated by professional journalists, can add value—and readers—to media brands.
In the long run, news organizations to which the public turns for good judgment in adjudicating news will accrue goodwill and command attention. The pace of the news cycle is quickening, but the fundamental responsibility of journalists to gather and disseminate reliable news hasn't changed, nor will it be supplanted by savvy social media auteurs.
There is a reflexive reaction to pit emergent social media behavior against traditional journalistic practices and norms. This defensive posture is counterproductive, for both sides. Rather than pointing out flaws to favor one model over the other, we should appreciate the interplay between them, an interdependence that ultimately produces a more participatory, accurate and compelling news cycle.
Social media is not going away. Even though the business models of the mainstream news industry are experiencing creative destruction, demand for good storytelling from trustworthy news sources isn't going away either.
Seth Mnookin is co-director of the Graduate Program in Science Writing at MIT. Hong Qu, a 2013 visiting fellow at the Nieman Foundation, is working on Keepr, an application to help journalists and other users better follow stories through Twitter.
MORE COVERAGE OF THE BOSTON MARATHON BOMBINGS
Reporting on Radicalization
By Souad Mekhennet
New Challenges, New Rewards for Journalists on Social Media
By Borja Echevarría de la Gándara
Curation Is the Key to Bringing Social Media and Journalism Together
By Ludovic Blecher
The (New) Industry Standard: Making Citizen Broadcasters into Citizen Journalists
By Betsy O'Donovan
In the late 1990s, I had been a police reporter for The Palm Beach (Fla.) Post, and one of my primary responsibilities was to listen to local police scanners to make sure the paper didn't miss anything. As soon as I got the shooting alert from MIT, I turned on a Boston scanner and began monitoring the situation. I also followed developments on Twitter, but I didn't turn on my TV or check news organization websites. I knew they wouldn't have anything yet.
The number of units being called in and the urgency in the dispatchers' voices made it clear this was a big deal. The level of response was out of proportion to a simple shooting, even one on a college campus. It was the type of response that suggested there was either a significant ongoing threat or this incident was, in fact, related to the bombings—or both.
I've been at MIT for almost two years, but I still think of myself primarily as a journalist. I realized this was a situation in which my training as a reporter might allow me to add value to the coverage. That's why I drove to Kendall Square in Cambridge, where MIT is located and the shooting had occurred.
By the time I arrived, there were roughly 20 journalists gathered around a police perimeter. Before I left home, I had noticed that Taylor Dobbs, a journalism student at Northeastern and the son of a friend of mine, was tweeting from the scene. We hadn't met before, but I sent Taylor a message so we could connect in person. He was there with another Northeastern student, Brian D'Amico, a photographer.
Around 12:15 a.m., word came that Sean Collier, the MIT officer who had been shot, had died. I still had the police scanner going on my iPhone when things suddenly went from being relatively quiet to a state of total chaos. A pack of Massachusetts State Police cruisers started booking it down the street, and the three of us—Brian, Taylor and I—got in my car to follow them. There was talk on the scanner of a carjacking and of a possible suspect on the run.
We ended up at a gas station in Cambridge on nearby Memorial Drive, which we learned was where the carjacking victim might have escaped. Then Taylor
Photo: Brian D'Amico
10:48 p.m.
Gunshots at MIT building 32 (Stata Center). Area cordoned off; unknown if injuries.saw a tweet that said the stolen car, a black Mercedes SUV, had been tracked to Watertown. Seconds later, we heard, "Shots fired on officers in Watertown" over the scanner, and it was total chaos again. The police at the gas station took off toward Watertown, and we followed.
We got to Watertown a little before 1 a.m., moments after the shootout between police and Tamerlan and Dzhokhar Tsarnaev, the brothers suspected of carrying out the Marathon bombings, had ended. You could smell gunpowder and see smoke in the residential neighborhood near the Watertown and Arsenal Malls. The next morning we learned that hundreds of rounds of ammunition had been fired and homemade bombs had been deployed.
Photo: Andrew Kitzenberg
12:53 a.m.
Reports of massive explosions and grenades. Can smell smoke in Watertown near Arsenal Mall.@hqu I used Keepr, a social media monitoring software tool I am developing as a visiting fellow at the Nieman Foundation, to capture Seth's tweets from Watertown. Keepr's algorithm extracts credible real-time information from raw Twitter feeds by pulling the 100 most recent tweets from Twitter's API and counting the words and phrases that occur most frequently. It also discovers Twitter users with the most mentions.
During the manhunt, I used Keepr to identify reliable sources who appeared to be tweeting from the scene. I used four factors as indicators of credibility: disclosure of location, preferably via geocoding (Taylor Dobbs had activated the geocoding feature on his iPhone Twitter app that night, but Seth had not), multiple source verification (the tweets cited information from primary as well as other sources), original pictures or video, and accuracy over time.
Photo: Jennings Raske
1:00 a.m.
“All units retreat. There are explosives on scene.”
Boston PD officer down; Watertown ambulance transporting.Keepr's algorithm detects good sources during breaking news events by focusing on high numbers of tweets during short periods of time, tweet bursts containing frequent mentions of specific usernames (this is how Keepr found @sethmnookin, whom I had not been previously following), and extremely rapid rates of follower growth. Wikipedia founder Jimmy Wales followed Seth that night, tweeting at 2:39 a.m., "When I followed you awhile ago, you had 10,000 followers. Now 30,000. Be late, be right, and be safe."
During those chaotic early morning hours, Seth practiced a new form of networked journalism, one that combines the speed and immediacy of social media with best journalistic practice. Seth was plugged into the cellular grid with his smartphone, but he was also plugged into a spontaneous self-organizing online group that was both consuming and participating in his coverage.
In addition to reporting what was happening, Seth held a kind of rolling, live-streamed press conference in which he answered followers' questions, corrected misinformation spreading via social media, and distributed important public safety updates from the police. He also crowdsourced reporting assistance, enlisting followers' help with everything from identifying the most useful police scanner to locating an iPhone charger.
Social media groups can go terribly wrong, of course, as with the Reddit online community that mistakenly identified a missing Brown University student as a suspect. But professional journalists get things wrong, too. The Reddit misinformation was retweeted by several prominent journalists, and CNN, The AP, and The New York Post, among others, all made embarrassing errors.
This style of breaking news coverage will likely appeal to a new generation of consumers willing to tolerate high levels of uncertainty and constantly changing information in exchange for 'watching' an event unfold live. These audiences will self-organize to support and amplify anyone reporting in this way, regardless of whether that person has press credentials. So, as 2013 Nieman fellow Betsy O'Donovan argues in "The (New) Industry Standard," news organizations would do well to promote a greater understanding of journalistic standards and ethics among the general public.
@sethmnookin I parked my car at the intersection of Dexter and Nichols Avenues, where the police were focusing their search for someone they described as "suspect number two." When Taylor, Brian and I arrived, we were the only reporters on the scene. There were, however, perhaps over 100 cops, many with their guns drawn.
As we were trying to get our bearings, the police were also trying to get theirs. There were cops just wandering around the streets, and you could hear over their radios, "We need to go to Laurel Street," which was two blocks south of where we were and, as it turned out, near where Dzhokhar Tsarnaev was later discovered to be hiding in a boat.
I had a small reporter's notebook with me, but early on realized that tweeting would be a more effective way to take notes. Everything I wrote would be time stamped, and I wouldn't need to worry about not being able to
Photo: Seth Mnookin
1:45 a.m.
Arrest going on now. “He’s got shit in his pockets. Get down that street now!”
pic.twitter.com
/rnpWrZ2H2Gread my messy handwriting after the fact. This worked out even better than I expected. Because I knew my notes were going to be public, I spent more time thinking about whether something was important or informative or whether I was simply writing things down because I was nervous or had nothing else to do. One thing I regret is not taking more pictures and recording more video, especially at the outset, when there was no one else at the scene.
Minutes after we arrived in Watertown, it became clear the police were preparing for an arrest; the second suspect, it seemed, was in a car on Nichols Avenue. Suddenly, there was shouting all around us, some of which was directed at the suspect, whom the police feared was wearing an explosive vest, and some of which was directed at us, ordering us to get down the street as quickly as possible. It took more than 10 minutes, but the police eventually got the suspect out of his car and ordered him to strip.
Around 2:30 a.m., Massachusetts State Police spokesman David Procopio told reporters, "We have two suspects in custody." One was the person who had been apprehended in the gun battle with police (we learned later that
Photo: Taylor Dobbs
2:58 a.m.
“One more suspect at large. Two accounted for should be revised. One accounted for. One at large.”
- state police PIO.this was Tamerlan Tsarnaev, who had been killed); the other was the man who had stripped and been loaded into a police car. But the latter turned out to have no connection to the Marathon bombings or that night's shootout, and about 30 minutes later Procopio gave an update: "One more suspect at large. Two accounted for should be revised. One accounted for. One at large."
By 3:30 a.m., I felt the situation was becoming more stable. The police hadn't found the second suspect, but there were no reports of further gunfire. I checked and saw I had dozens of missed calls and e-mails from TV and radio producers. It wasn't until after 6 a.m. that I finally went home to see my wife and kids. By that point, Good Morning America had offered to put me up in the Four Seasons—provided that I promised not to appear on any other shows before going on the air with them. By the time the show was broadcast, its producers had decided they weren't interested in me, after all.
Historically, TV and radio have had the biggest competitive advantage in breaking news. But fluid, chaotic situations are also precisely those in which the transfer of information from cameraman to reporter to producer to anchor is most prone to error. Plus, when you're on the air, providing steady updates isn't an option—it's a necessity. And needing to fill airtime can cause problems of its own. On Twitter, if there's a new development every minute, you can update every minute; if nothing is happening, you can wait.
The Watertown manhunt illustrated the fact that there are times when a traditional journalist can do his job more efficiently and effectively on Twitter than in any other medium. I've worked in virtually every type of print outlet, from webzines to newspapers to newsweeklies to monthly glossies. I've blogged on my own website, on blog networks, and on traditional news sites. I've written books. I've been on the network's morning news shows and The Daily Show, on shock jock drive-time radio and Fresh Air. But for those three or four hours when a gunman was on the loose and a neighborhood was under siege, Twitter was the most efficient way to get information out to the public.
@hqu Twitter coverage of the manhunt in Watertown is a remarkable milestone for journalism. Even more remarkable are the implications for ordinary citizens who, without a press pass, can report news and influence coverage. For the latter group, this event instilled a newfound sense of power and responsibility in how they verify and disseminate news. Tools and processes for assessing source credibility need to catch up with social media technology and culture, especially in dangerous environments in which the public relies on reporters to provide actionable news updates with minimal misinformation and fallout.
Legacy media have a crucial role to play, both in providing original reporting and in curating social media. Seth's followers increased from a little over 8,000 before the night of the Watertown manhunt to over 45,000 during the few hours he was tweeting from the scene. That amplification was achieved largely through prominent journalists and celebrities (political commentator Keith Olbermann, New York Times reporter Brian Stelter, and actor/comedian Kumail Nanjiani, to name a few) and major news outlets (including ABC News and Foreign Policy) following and retweeting him. As 2013 Nieman fellow Ludovic Blecher explains in "Curation Is the Key," online breaking news forums, controlled and curated by professional journalists, can add value—and readers—to media brands.
In the long run, news organizations to which the public turns for good judgment in adjudicating news will accrue goodwill and command attention. The pace of the news cycle is quickening, but the fundamental responsibility of journalists to gather and disseminate reliable news hasn't changed, nor will it be supplanted by savvy social media auteurs.
There is a reflexive reaction to pit emergent social media behavior against traditional journalistic practices and norms. This defensive posture is counterproductive, for both sides. Rather than pointing out flaws to favor one model over the other, we should appreciate the interplay between them, an interdependence that ultimately produces a more participatory, accurate and compelling news cycle.
Social media is not going away. Even though the business models of the mainstream news industry are experiencing creative destruction, demand for good storytelling from trustworthy news sources isn't going away either.
Seth Mnookin is co-director of the Graduate Program in Science Writing at MIT. Hong Qu, a 2013 visiting fellow at the Nieman Foundation, is working on Keepr, an application to help journalists and other users better follow stories through Twitter.
MORE COVERAGE OF THE BOSTON MARATHON BOMBINGS
Reporting on Radicalization
By Souad Mekhennet
New Challenges, New Rewards for Journalists on Social Media
By Borja Echevarría de la Gándara
Curation Is the Key to Bringing Social Media and Journalism Together
By Ludovic Blecher
The (New) Industry Standard: Making Citizen Broadcasters into Citizen Journalists
By Betsy O'Donovan
