Too much conviction can be a dangerous thing. It certainly was for Marine Major Mark Thompson, who had a friend approach The Washington Post on his behalf, in late 2014, about Thompson’s story of how he had been wrongly accused of sexual misconduct while teaching at the US Naval Academy. Thompson claimed the Naval Criminal Investigative Service had failed him; the academy was corrupt; and, above all else, the two female midshipmen who had accused him of having sex with them, while they were his students, were lying. Thompson wanted the Post to prove his innocence. When the Post finally published his story this March, it proved something very different.
“A Marine’s Convictions,” written by staff writer John Woodrow Cox, transforms the hairpin turns in Thompson’s case into a gripping investigative narrative that plays out almost in real time, as Cox discovers new evidence, as well as other shocking developments that the investigators missed. It’s also a valuable model for how to report a story when the source has a vested interest in the outcome, and the truth is murky at best–stories, for instance, involving sexual assault.
Thompson’s case was highly complex. Not only did it come with 3,500 pages of court documents and a befuddling dearth of evidence, but also with two opposing decisions about what Thompson had done. At his court martial in 2013, Thompson was acquitted of rape, but convicted of five other lesser offenses, including having sex with the two students, a violation of military law. Then in a stunning reversal the following year, a board of inquiry, made up of three Marine Corps officers, determined he’d done nothing wrong. The ruling couldn’t reverse the convictions, but it saved Thompson’s job and pension. It also led him to the Post, which he hoped would clear his name.
Cox resolved to remain skeptical of everyone involved. Even after months of reporting, he still wasn’t sure where the story was going, or what it was about. And then, at the encouragement of his editor, he wrote it all down in a first-person narrative that draws readers into his sense of pursuit and lays his reporting bare. Published at 8,000 words, complete with seven chapters and an epilogue, the piece reads more like a true crime thriller than a typical front-page news story.
When reporters embrace how little they know, resist forming conclusions, and share their doubts with their readers in a form that breaks with convention, they may wind up getting closer to the truth.
The payoff amounts to more than a riveting plot: Since publication in March, Cox’s discoveries have prompted the military to launch a new investigation into Thompson’s case, as well as the removal of another US Naval Academy instructor who’s been implicated in the growing scandal.
All of which says something interesting, and paradoxical, about the nature of reporting: If conviction itself creates bias, uncertainty can be a powerful tool. When reporters embrace how little they know, resist forming conclusions, and share their doubts with their readers in a form that breaks with convention, they may wind up getting closer to the truth.
One of the key insights to emerge in the wake of the infamous Rolling Stone story last year about a fraternity gang-rape at the University of Virginia–a gang-rape that never actually happened–was the danger of confirmation bias. Writing for this publication, Judith Shulevitz defined this type of bias as “our innate urge to see only evidence that confirms beliefs we hold to be self-evident, and dismiss facts that challenge those convictions.” She went on to describe how this bias had infected nearly every decision made by The Rolling Stone writer, Sabrina Rubin Erdely, and her editors in the course of reporting the story, from why she chose the rape survivor she did to whom she chose to interview.
In other words, critics said, the problem with the piece was foundational. Before Erdely ever started writing, nevermind reporting, it seemed she had already written the piece in her mind. “Erdely told Rosin that she’d gone all around the country looking for rape survivors and was delighted when she stumbled on Jackie,” Shulevitz wrote, referring to Hanna Rosin, one of the first journalists to question the story’s validity. “She was obviously traumatized, and her story illustrated everything Erdely knew to be true—that frat boys rape girls and universities are indifferent to rape survivors.” Shulevitz argued that this sense of knowing was powerful enough to allow Erdely and her editors to cut corners on basic reporting duties, such as giving the accused a chance to respond.
For Cox, that story became a cautionary tale, a road map of what to avoid while he reported on Thompson’s case. It led him to fiercely examine nearly every decision he made and why, down to which expert to call.
“In that Rolling Stone story, it seemed like the nutgraf might have been written early on, and that it was looking for a narrative that fit it,” Cox says, referring to what is usually the third or fourth paragraph in a story that explains what the piece is about, as well as its significance. “And then, in the interviews the author gave in the days after the story ran, there was some pretty clear bias that she had towards the institutions, towards the fraternity culture. Often our biases come from legitimate sources, they come from things that are real. We’re not just inventing them. But as a reporter, especially if you’re doing any kind of investigative work, it’s so dangerous to let those biases creep in because they can inform you from day one, just by saying, I’m going to call this expert, but not that one,” Cox continues.
To avoid those kinds of errors, Cox became almost maniacally detached from the story’s outcome. He promised himself he wouldn’t develop a hypothesis. He committed to follow the facts and assemble the story later.
The facts, of course, were muddled. Thompson claimed he was innocent. He made a legitimate case for why he couldn’t have received a fair trial and had a compelling alibi. But Cox had to wonder: Why would the women–one of whom, Sarah Stadler, was named in the press and had to give graphic testimony during the court martial–subject themselves to public scrutiny, and the risk of losing their careers, if none of it was true?
The accusations had surfaced in 2012, months after a drunken night of strip poker at Thompson’s home in Annapolis, when one of the women, who has not been named in the press, came forward and claimed that Thompson had raped her. But Stadler said the tryst had been consensual in her case, part of an ongoing relationship with Thompson.
“Sometimes when you report a story so thoroughly, you realize there is no story, or the narrative isn’t as interesting as you hoped it would be in the beginning….You’re way better off reporting yourself out of the story than you are reporting the wrong story.”
The military’s investigation raised other red flags. Ten minutes of audio were missing from the testimony of the alleged rape victim. Thompson’s lawyers discovered that after playing strip poker at Thompson’s house, Stadler had gone to a bar and kissed another student–a revelation that challenged not only the investigators’ competency, but also Stadler’s credibility. She claimed she had been so drunk she had no memory of any of this. What’s more, Stadler’s cell phone had never been found, leaving a major gap in digital evidence.
Meanwhile, Thompson denied everything. He claimed he hardly knew Stadler, nor the other woman. It was their word, and their reputations, against his.
“One thing I knew was that that’s not unusual, especially in cases involving sexual assault, alcohol, and long periods of time. People’s memories can change. Anybody who studies these things knows that happens. Still, it was worth checking out. The idea was, report this story to the bone without the expectation that I would discover something revelatory,” Cox says, adding that he aimed to over-report story, even if it meant discovering something that proved it wasn’t worth publishing.
“Sometimes when you report a story so thoroughly, you realize there is no story, or the narrative isn’t as interesting as you hoped it would be in the beginning,” Cox says. “I think that happens a lot. You’re way better off reporting yourself out of the story than you are reporting the wrong story.”
The pace was slow. He had 3,500 pages of court documents to read, a lot to learn about the military justice system, and, since he wanted to give everyone a chance to tell him he’d gotten something wrong or mischaracterized a fact, a lot of phone calls to make.
Luck and time were on his side, as with getting in touch with Stadler’s mother. Hoping to catch her on the phone so she didn’t have time to rehearse her statements, Cox called her for two or three months, without ever leaving a voicemail, before she finally picked up. His patience paid off. During the phone call, Stadler’s mother caught the federal prosecutor on Thompson’s case in a lie. Their conversation also led to a phone call with Stadler and a pivotal discovery. Not only would Stadler find her long-missing cell phone, but its memory was intact: It came with 650 text messages that proved Thompson had been lying to Cox since day one. Thompson had claimed he “hardly knew” his accusers. But according to some of the messages, Thompson knew Stadler very well.
They were the sorts of breakthrough revelations Cox had not expected to find. They also convinced Cox’s editor to employ a first-person narrative–an unusual choice for a major “front-page” news story, but one that made the entire piece, from the reporting to the final, shocking line, much more transparent to readers.
“I thought it was a situation where the reporting advanced the story. The first draft I read, you had to read between the lines to figure out that a lot of the really interesting information in the piece was there because of John’s reporting,” says Cameron Barr, the Post’s managing editor. “I thought it would be more compelling for the reader if we could use the first person without having to use tortured phrasings like ‘this reporter,’ etc.”
Barr also knew it was an opportunity to show readers how difficult it can be for news organizations to determine a source’s credibility. One of the major flaws in The Rolling Stone story had been the assumption that the rape survivor was automatically credible. The goal, in Barr’s mind, was to make the reporting process, which in this case was a subplot in its own right, as transparent as possible.
“I wanted to tell the story in a narrative fashion that enabled us to disclose more forthrightly that Major Thompson came to us,” Barr says. “I wanted that to be clearer much earlier in the piece, so we could bring the reader along on this journey where we’re presented with a set of facts, a version of what happened from the Major, and by virtue of what John did, we learned something that sheds a lot of new light on that.”
Indeed, last week, as a result of the Post’s investigation, the military filed two new charges against Thompson, including one count of making a false official statement and another of conduct unbecoming of an officer and a gentleman. If convicted, Thompson could face up to five years in prison, dismissal from the service, and loss of his retirement benefits.
More than a year after Erdely’s story delivered a blow to the fragile trust that exists between news organizations and readers, not to mention rape survivors, Cox’s reporting may offer a kind of correction. By embracing uncertainty, and dumping his notebook on the table for readers to see, Cox and his editors made media transparency a cornerstone of the piece. In doing so, they also brought readers and the case closer to the truth.Damaris Colhoun is CJR’s digital correspondent covering the media business. A reporter at large in New York, Colhoun has also written for The Believer, The New York Times, The Guardian, and Atlas Obscura. Find her on Twitter @damarisdeere.