For months, our magazine has been subject to accusations that stories we published by an American soldier then serving in Iraq were fabricated. When these accusations first arose, we promised our readers a full account of our investigation. We spent the last four-and-a-half months re-reporting his stories. These are our findings.

When Michael Goldfarb, a blogger for The Weekly Standard, left me a message on a Tuesday afternoon in mid-July, I didn’t know him or his byline. And I certainly didn’t anticipate that his message would become the starting point for a controversy.

A day earlier, The New Republic had published a piece titled “Shock Troops.” It appeared on the magazine’s back page, the “Diarist” slot, which is reserved for short first-person meditations. “Shock Troops” bore the byline Scott Thomas, which we identified as a pseudonym for a soldier then serving in Iraq. Thomas described how war distorts moral judgments. To illustrate his point, he narrated three disturbing anecdotes. In one, he and his comrades cracked vulgar jokes about a woman with a scarred face while she sat in close proximity. In another, a soldier paraded around with the fragment of an exhumed skull on his head. A final vignette described a driver of a Bradley Fighting Vehicle who took pride in running over dogs.

Goldfarb said he had been contacted by tipsters who thought these scenarios sounded concocted by a writer with an overactive imagination—or perhaps by a total fabulist. He asked for evidence that might answer these complaints, “any details that would reassure that this isn’t fiction.” Among other things, he wanted the name of the base where the author had mocked the disfigured woman.

The same afternoon, we contacted the author, asking permission to answer Goldfarb’s queries. We thought we could provide details that might answer these concerns without revealing the author’s identity and violating the compact we formed when granting him a pseudonym. He agreed. I told Goldfarb that the insults to the woman had occurred at Forward Operating Base (FOB) Falcon. A day later, Goldfarb sent a link to an item on the Standard blog. It quoted an anonymous source who said the story sounded like a collection of the “This is no bullshit … stories soldiers like to tell.” Goldfarb called on the military blogosphere to do “some digging” and for “individual soldiers and veterans to come forward with relevant information.”

By the weekend, the Standard’s editor, William Kristol, published an editorial that, without evidence, pronounced the Diarist an open-and-shut case. Kristol wrote, “But what is revealing about this mistake is that the editors must have wanted to suspend their disbelief in tales of gross misconduct by American troops. How else could they have published such a farrago of dubious tales? Having turned against a war that some of them supported, the left is now turning against the troops they claim still to support.”

In prior months, our magazine had been coming under attack from the left for criticizing the war but failing to champion withdrawal (not to mention for initially supporting the war). So it was disorienting to find ourselves criticized from the right, too, for supposedly slandering the troops.

But, regardless of the Standard’s ideological motives, the doubts about “Shock Troops” resonated. All over the blogosphere, people who presented themselves as experts claimed that the events described in the piece could never have happened. Some of these assertions were vague and meaningless—“They are not ‘Shock Troops.’ They are our best and bravest,” Kristol wrote—as if our soldiers were plaster saints immune from the traumas of war. But others were more specific and troubling. Denizens of FOB Falcon insisted that they had never seen a woman who matched Thomas’s description; some familiar with the Bradley asserted that it couldn’t be maneuvered to kill dogs; others claimed that any exhumation of bones would be reported up the chain of command as a matter of course.

Did we have a Jayson Blair on our hands—or, closer to home, another Stephen Glass, the fabulist who did so much to tarnish this magazine’s reputation ten years ago? The facts in dispute would be difficult to untangle: Our writer’s identity required protection, he was far away, and the events themselves occurred in a war zone.

We published an online statement pledging an investigation. That weekend, members of the editorial staff assembled at my house to divide up the task of re-reporting his stories. It was the beginning of a project that, for long stretches, superseded our day jobs—and led us to some uncomfortable conclusions.

By now, the identity of Scott Thomas is publicly known. He is Scott Thomas Beauchamp, age 24. He first came to our attention nearly a year ago by way of Elspeth Reeve, one of three reporter-researchers who work at TNR as essentially yearlong interns and whose responsibilities include fact-checking. When she sent along a piece from her friend Scott in Iraq, we were intrigued. The introspective writings of a low-ranking soldier seemed valuable. When, before publication, Beauchamp asked for a pseudonym, we granted it. We felt that a soldier in a war zone could write most honestly about his feelings and experiences under a penumbra of anonymity.

His first piece, a Diarist titled “War Bonds” published in our February 5 issue, described the woes of an Iraqi boy named Ali who adopted the moniker “James Bond.” Soon after James Bond chit-chats with American soldiers, Beauchamp learns that thugs—most likely insurgents—cut out his tongue. This first piece didn’t receive much attention, but the attention it did receive was positive. Hawks, in particular, liked that it sympathetically described the plight of sensitive young soldiers on the front line.

Several weeks passed before Beauchamp sent us another story—one recounting dialogue between soldiers in a guard tower, which we rejected. During that time, he took leave in Germany with Reeve. The two had been casual friends at the University of Missouri and resumed a relationship online, which quickly turned into something serious. During Beauchamp’s leave, he and Reeve left Germany and, without telling anyone at the magazine, married at a lawyer’s office in Virginia. A day after the ceremony, Reeve returned to TNR’s office to share the news.

Beauchamp visited our office during his brief stay in Washington. We shook hands, and I encouraged him to send more pieces. Soon after, in May, another arrived. The story described what a fellow soldier called the “zombie dogs” of Baghdad, the homeless mutts who devoured human corpses, many of them victims of sectarian violence.

Another piece followed, about wartime humor among soldiers—the piece that ultimately became “Shock Troops.” Beauchamp wrote that humor was essential to the soldiers’ humanity, but that “the jokes were dark, violent, and would be seen by [my] former self, I assume, as in bad taste.” That draft included the story of mocking the disfigured woman in the “chow hall.” And it concluded with the following paragraph featuring his pseudonymous friends:

There are other examples I could give of just how dark and sad humor during a war can be. There was the time that Jibson wore the top of a human skull as a hat during a mission. All of Short’s dog hunting stories (I think he’s up to 17 kills). The time we hid a pink dildo in a very conservative Christian kids gear before an inspection (don’t ask how we got a pink dildo in Iraq). The point remains, each world is set on a sliding scale of morality that’s determined by setting and necessity. The only constant is the desire to laugh. And sometimes, in certain situations, you’ll laugh at anything no matter what.

Naturally we wanted to learn more about the dog-hunting and the skull— although, in hindsight, the genesis of these anecdotes in such a nonchalant aside should have provoked greater suspicion. Beauchamp revised the piece, and we sanded down the prose. A month after he submitted the first draft, after several revisions, it entered into galleys.

Fact-checking is a process used by most magazines (but not most newspapers) to independently verify what’s in their articles. Beauchamp’s anonymity complicated this process. Because we promised to protect his identity, we were reluctant to call Army public affairs to review his claims. What’s more, the fact-checking of first-person articles about personal experiences necessarily relies heavily on the author’s word and description of events.

But there was one avoidable problem with our Beauchamp fact-check. His wife, Reeve, was assigned a large role in checking his third piece. While we believe she acted with good faith and integrity—not just in this instance, but throughout this whole ordeal—there was a clear conflict of interest. At the time, our logic—in hindsight, obviously flawed—was that corresponding with a soldier in Iraq is logistically difficult and Reeve was already routinely speaking with him. It was a mistake—and we’ve imposed new rules to prevent future fact-checking conflicts of interest.

Facing the difficulties of verifying the piece, but wanting to ensure its plausibility before publication, we sent the piece to a correspondent for a major newspaper who had spent many tours embedded in Iraq. He had heard accounts of soldiers killing dogs with Bradleys. These accounts stuck with him because they represented a symbolic shift in the war. Iraqis regard dogs as annoying pests. At the beginning of the conflict, Americans made great efforts to befriend these mistreated mutts. It seemed telling that Americans now treated dogs with as little regard as Iraqis did. He considered Beauchamp’s dog-hunting anecdote plausible.

But the reporter doubted the tale of the disfigured woman. What would a woman with the disfigurements described by Beauchamp be doing in a war zone? This became the focal point of our fact-checking. We asked Reeve to push Beauchamp for corroboration of this woman’s existence. In an e-mail, she relayed his answer (throughout this story, we’ve withheld the names of soldiers who never gave us permission to use them):

OK, talked to Scott. He said it looked like the lady’s injuries were cosmetic, though he had no way of knowing her medical history, of course. I asked him if there was anyone around who had seen her. He was in the tower but he shouted over to his buddy [name withheld], asking if he remembered the woman in the DFAC with burns on her face. I heard a guy shout yeah. I asked Scott to ask [name withheld] to describe her. Scott shouts, “Hey, can you say what you remember that woman looked like?” I heard, “Yeah, I remember that butt-ugly woman in the DFAC [dining facility].” So there’s that. Scott said that if he had to guess, the woman was a contractor, and had gone home after her injury and then decided to come back. Her scars looked long-healed. But again, he stressed he had no way of knowing her real story.

Reeve also asked a National Guard medic who had served in Iraq if he had seen burn victims in chow halls. He replied, “[N]ot many … but a couple.”

With first-person narratives, of course, especially in war zones, there are limits to what can be independently verified. The editor who worked on the piece spoke with Beauchamp to push him further on the physical description. During a phone call, Beauchamp assured the editor that he had accurately described the incident with the woman. Because of his corroboration, and because he wrote two other pieces with no apparent problems, we gave him the benefit of the doubt.

I hadn’t worked with Stephen Glass, who made up stories out of whole cloth, but I knew the lessons derived from that scandal. Fabulists are often nabbed by the little lies, the asides they assume that no one will check. As we began our re-reporting of Beauchamp’s pieces, we searched for the easily verifiable bits of information that would serve as crucial benchmarks. And, on the first full day of our investigation, it didn’t look good for Beauchamp.

In his second story, he described dogs eating the brain of a corpse. He ended with a slice of dialogue with a soldier he called Hernandez:

“I took his driver’s license,” I said. “You did?” questioned Hernandez. “Yeah. It said he was an organ donor.” We chuckled in the dark for a moment, and then looked out the window into the night. We didn’t talk again until we were back at our base.

But do Iraqis have driver’s licenses that allow for organ donation? We called the Iraqi Embassy. Apparently, licenses have never contained such information. The question then shifted: Was the license the punchline to a joke? Or did Beauchamp intend the sentence to be read literally?

We called him a day later. He answered on his cell phone, a fuzzy connection with seemingly interminable delays.

TNR: “Tell me a little about the driver’s license story with Hernandez.”

Beauchamp: “What exactly do you want to know about it?”

TNR: “I read it, but what was the deal there? What happened?”

Beauchamp: “We came across the body and the dogs were eating its brains out, and it became obvious through, I guess, the shells we found, that it was an I.P. [Iraqi police] execution, and that’s about basically the story.”

TNR: “What about the driver’s license?”

Beauchamp: “Oh, that was just a joke I made.”

He had survived one test. And, better than that, Reeve provided us with a contemporaneous e-mail from him that described the same event. Beauchamp wasn’t a reporter, but these served the same function as a reporter’s notes.

There were other quick-and-dirty tests. In “Shock Troops,” Beauchamp tried to set his appalling behavior in autobiographical context: “I’ve never thought of myself as a cruel person.… I once worked at a summer camp for developmentally disabled children, and, in college, I devoted hours every week to helping a student with cerebral palsy perform basic tasks like typing, eating, and going to the bathroom.” And, indeed, Teka McDonald in the human resources department at the St. Louis-area Life Skills camp confirmed that Beauchamp had worked there in the summers of 2004 and 2005. We spoke with Andrew Hogan, who also helped take care of the student with cerebral palsy. Beauchamp’s account checked out. At this early stage, we felt comfortable that, at the very least, we weren’t dealing with another Glass.

During the first week of the investigation, I reached Beauchamp with regularity on his cell phone. My calls with him often began the same way. “You’re not a professional journalist,” I would tell him. “If you got anything wrong or exaggerated things, people will understand; it’s better to admit error than get caught in a lie.” Every time, he stood by his stories.

He also added details to his accounts. The woman Beauchamp said he had mocked loomed large within his circle of friends. They called her “Crypt Keeper” or “Mandrake’s Bride.” The bones, meanwhile, had been uncovered while filling sandbags in a small section of his combat outpost. (I received a photo of Beauchamp holding a bone in one hand while obscuring the name on his uniform with the other.) He provided us with the names of the soldier who wore the skull and the driver who ran over dogs. And he solicited corroborating accounts from five other soldiers.

During our first call, he passed the phone to a soldier who had driven Bradley vehicles, the kind Beauchamp had written was used to kill dogs. The driver would only talk to me after I assured him that I would never print his name. He hadn’t witnessed the specific incident with the woman in the dining facility, although he had frequently seen her. But he had watched the soldier wear the skull fragment. “It fit like a yarmulke. It might still be floating around,” he told me. “We tossed it around each others’ vehicles.” He also had witnessed dog-hunting: “I have seen [the driver that Beauchamp wrote about] roll over dogs. I don’t keep track of how many times he’s done it—but it was multiple times.” In a series of subsequent e-mails, this Bradley driver elaborated at great length:

How you do this (I’ve seen it done more than once) is, when you approach the dog in question, suddenly lurch the Bradley on the opposite side of the road the dog is on. The rear-end of the vehicle will then swing TOWARD the animal, scaring it into running out into the road. If it works, the dog is running into the center of the road as the driver swings his yoke back around the other way, and the dog becomes a chalk outline. In this particular instance, as a Humvee gunner, I was up out of the vehicle, able to hear the dog actually scream as it was pulled into the tracks on the left side of the Bradley; a paw got stuck in the front drive sprocket, spinning like a stick caught in the spokes of a bicycle for a few rotations before joining the rest of the wet, meaty mess in the tracks. I can still verbally mimic the sound I heard the dog made.

It was an overwritten message, to be sure. But it added to Beauchamp’s original description. We received other e-mails from soldiers. The authors made it clear that they didn’t want their names appearing in print. They were only providing details because Beauchamp had asked them to come forward. Some excerpts:

Soldier A: “I would like to say something about Mandrake’s Bride. … [W]e first saw the lady in Kuwait and it is very true. She had burns on her head and its strange but in a way most people thought it was humorous. It might sound sick but I guess that’s all we really have here is to laugh when we can and day dream of home.”

Soldier B: “The crypt keeper, yes I saw her, the skin of her face had something wrong with it, burn, maybe some sort of surgery and her hair was like a thinning mullet with chunks missing, she was wearing DCUs [Desert Camouflage Uniforms] if I remember correctly but like Beauchamp said I can’t remember seeing a unit patch on her which makes me think she was a civilian.”

Soldier A: “While digging we came across several bones and a guy named [name withheld] said he was part Indian and danced around the bones to show he was peaceful and he did a proper burial procedure.”

On another call, Beauchamp passed us to someone who identified himself as a non-commissioned officer. Very nervously, the NCO told us that he was aware of Beauchamp’s article and that the men in his unit were all very proud of him. When we asked him about the soldier wearing the skull, he became silent. We asked him again. He said that he couldn’t comment. Answering the question, he said, would reflect on his leadership. He then quickly ended the call.

The nature of these contacts wasn’t ideal: Beauchamp was soliciting his own witnesses. But, once Beauchamp established the initial contact, we tried to communicate with these soldiers independently. We always considered the possibility that they were lying to cover for their friend, but there was no way for us to know that for certain, and we couldn’t dismiss what they told us. They were not only Beauchamp’s buddies, but, in some instances, the only witnesses to the events described.

On the Standard website and elsewhere, there was speculation that Scott Thomas might not be an active-duty soldier at all. The Standard described a lengthy “semiotics-based analysis” arguing that he “fits the profile of a creative writing program graduate.” I tried to convince Beauchamp that he could buy credibility and knock down these specious claims with one gesture: revealing his name. A week after the initial call from Goldfarb, Beauchamp finally agreed.

In the early hours of July 26, we exchanged instant messages with Beauchamp, who reported a meeting scheduled for later in the day about wearing “skulls on their head in sector.” Beauchamp didn’t know what to make of this session. But he wanted his statement, which announced his identity and defended the factual basis of his piece, posted as soon as possible. We published his statement on our website at about 6 a.m. Thanks to instant-messaging, we watched the early phase of the Army investigation in real time.

Beauchamp instant-messaged us that officers had “made people sign sworn statements saying that brads don’t intentionally hit dogs and that no mass grave was found” at his combat outpost— “in fact, that no human remains at all were found there.” Beauchamp said he was under enormous stress. “[I] wanted to get out of the room alive,” he told us. He signed statements but tried to phrase them carefully. “[I] think i worded it pretty well enough to buy me some more time without contradicting myself.”

Earlier that morning, we had received an e-mail from a soldier in Beauchamp’s unit who had mentioned seeing the disfigured woman in Kuwait. It was the second time a soldier had placed her there. During our instant-messaging, we pushed Beauchamp on this:

TNR: where did you see the crypt keeper?

Beauchamp: are you there?

TNR: yes

Beauchamp: the last thing i got was “where did you see the crypt keeper”

TNR: yes

Beauchamp: the dfac on falcon or chow hall, as it IS commonly called

TNR: what about kuwait?

Beauchamp: brb [be right back]

Nine minutes of silence

TNR: you there?

Ten minutes of silence

Beauchamp: ok just did a sworn statement

TNR: about?

Beauchamp: saying that i wrote the articles

TNR: ok

Beauchamp: theyre taking away my laptop

TNR: fuck is this it for communication?

Beauchamp: yeah and im fucked

TNR: they said that?

Beauchamp: because you’re right the crypt keep WAS in Kuwait FUCK FUCK FUCK this is bad isnt it

TNR: yes where in kuwait?

Beauchamp: it did happen in kuwait Camp Beuhring

TNR: why didn’t you tell us that?

Beauchamp: i thought it was on falcon till somebody here convinced me that it wasnt i just talked to [Soldier A] and he convinced me that it was in kuwait when i thought it was on falcon fuck

TNR: if what you’re saying is true it’s not the end of the world

Beauchamp: ok

TNR: as long as we can confirm it

Beauchamp: good i have to go like NOW though im so sorry

TNR: are you gonna be able to talk again?

Beauchamp: i hope so but i dont know thank you again for everything

TNR: i didn’t do anything what did you sign?

After that, the Army, by its own admission, didn’t permit Beauchamp to speak to TNR for over a month. It was the worst moment to lose contact. He had admitted a major mistake, using an event that occurred in Kuwait before he ever set foot in Iraq to describe the psychological impact of war. We published a statement announcing this error soon after. But did it stem from an intentional manipulation of fact or an innocent slip of memory, as his instant message seemed to indicate? He had also gone from a pseudonym to a soldier with a name and a face and a personal history—all of which were about to become grist for the bloggers.

Within an hour of our instant-message exchange with Beauchamp, one of the soldiers in his unit with whom we had previously spoken sent an account of the Army’s investigation:

[S]lew of events: initially, the whole platoon was called in and we received certificates of having been instructed in some sort of equal opportunity training we never got. Then everyone was dismissed; everyone, of course, except the platoon’s four Bradley drivers. … What we had to do, then, was write and sign a sworn statement … saying that we’d never seen or committed the act of randomly causing destruction with our Brads, and that we’d found no “mass grave” site at [combat outpost] Ellis. … [I]t bottomed out to us saying that we’d found “unidentified remains.” [Captain] cheerfully edged us into calling them “animal” remains “so that there’s no implication of them possibly being human.” I changed mine to what he wanted. SCOTT changed his to “remains that people had said were animals.”

A pattern began. Beauchamp’s behavior was sometimes suspicious—promising evidence that never arrived—but so was the Army’s. Beauchamp had corroboration, but his confusion over Iraq and Kuwait was troubling. And we were running out of leads; one of the few remaining was a former member of Beauchamp’s unit named Kristopher Kiple.

Beauchamp had described Kiple to me as the figure in his story who stabs his mashed potatoes in disgust at the sight of the disfigured woman and cracks jokes at her expense. When the “Shock Troops” controversy emerged, Kiple was in the process of leaving the military and was being held at a base in Germany. He told me the Army had removed him from Iraq on mental health grounds. Once in Germany, he had gotten into trouble for “out on the town stuff” and “resisting arrest.” We’d left messages on his MySpace page for him to call. Several days after Beauchamp went incommunicado, Kiple called me on a Saturday morning.

Kiple understood that he didn’t make the ideal witness, given his current predicament. But he did recall the events Beauchamp described. “I remember the woman,” he told me. “She didn’t go to Iraq; she was in Kuwait. She was bald with strands of hair—her hair was gray just a little bit. Her face was kind of mangled. It looked it like it was scarred or something. It wasn’t recent. It happened in the past. She looked recovered. She wore a brown uniform, BDU [Battle Dress Uniform], with pocketed pants. It didn’t have any rank. She looked like a civilian contractor or something. She looked like an American. We saw her about every day or every other day—maybe fourteen times. Usually, mostly during lunch chow—twelve, one p.m. Yes, we called her Mandrake’s Bride, some crazy mythology that Scott and one of our buddies made up for her. I don’t remember some of the shit that they used to talk about her.”

In his story, Beauchamp had written about a joke he and his buddies had made suggesting that Mandrake’s Bride appear in a calendar of battle-scarred women, which Beauchamp dubbed “IED Babes.” This had become one of the most controversial parts of his account—would soldiers really say something so despicable? Without my prompting, Kiple raised the subject. “I remember the calendar distinctly—the ‘IED Babes’—because I thought that was the funniest thing in the world.” I pushed him to describe the scene. “We were really poking fun at her. It was just me and Scott Beauchamp the day that I made that comment. We were pretty loud. She was sitting at the table behind me. We were at the end of the table. I believe that there were a few people a few feet to the right. We were pretty loud about it. Nobody said anything. Mandrake’s Bride heard it and got up and left.”

I told Kiple that if he was lying, it would only hurt his pal. He replied, “I was nervous, questioning myself. Some of it’s a blur. But it happened.” He told me that he was about to leave the Army, and, when he did, we would have permission to quote him by name.

We had a hard time prodding members of Beauchamp’s unit to talk further. By coming down hard on Beauchamp, the Army clearly provided a cautionary tale about the perils of cooperating with the press. As soon as Beauchamp went public, according to Reeve, the military immediately prevented him from calling even his family, who enlisted the help of a home-state politician to restore a line of communication; Beauchamp began working longer shifts and was isolated from his comrades.

Without new evidence to be gleaned, we began to lay out the evidence we had assembled. It wasn’t just the testimonials from the soldiers in his unit. Among others, we had called a forensic anthropologist and a spokesman for the manufacturer of Bradley Fighting Vehicles. Nothing in our conversations with them had dissuaded us of the plausibility of Beauchamp’s pieces.

But we also found some reason to doubt Beauchamp’s reliability: In 2006, he had written a personal blog, Sir Real Scott Thomas, which we only discovered after the controversy erupted. He appeared an angst-ridden young man prone to paroxysms: “I shoot, move, communicate, and kill … the deaths that I inflict secure the riches of the empire.” With his excited prose and tendency toward overstatement, his blog did not inspire journalistic confidence. We had good reasons never to assign Beauchamp another piece. Overall, however, when we considered the totality of what we had amassed, we didn’t have enough information to retract the ones we had published. There was the one significant mistake, but we needed to know more about how it had occurred. And that information could only be obtained from another conversation with Beauchamp, something the Army wouldn’t allow. On August 2, we laid out what we had learned in a second carefully worded statement on our website.

On August 1, six days after the “skulls on their head in sector” meeting, the Army concluded its investigation. Two days later, a public affairs officer announced that Beauchamp’s piece had been “refuted by members of his platoon and proven to be false.” The Army didn’t announce this to The New York Times or even The Weekly Standard, let alone in a public report. It first gave the story of Beauchamp’s supposed fraudulence to a former porn actor turned blogger named Matt Sanchez. Apparently, the Army wanted the matter to quietly fade away. Several days after Sanchez’s scoop, the Standard reported, based on an anonymous military source, that Beauchamp had signed a statement admitting that all three of his pieces were “fabrications containing only ‘a smidgen’ of truth.”

Many reporters who wrote stories about the Beauchamp affair noted the thinness of the Army’s report. Spokesmen from the military curtly confirmed the findings to reporters. When pressed, they declined to elaborate. Goldfarb’s allegation that Scott had recanted his stories, however, was more difficult to blunt. Both the Times and The Washington Post repeated the Standard’s anonymously sourced accusation. Either the Army source had lied to the Standard or Beauchamp had lied to us. When Beauchamp had described his statements to us, it seemed like he was walking a fine line, trying to satisfy his commanders while staying on the side of the truth. But, without the actual documents in hand, we had no way of judging. Through his wife and lawyer, we made the first of many requests for these statements, which Beauchamp was legally entitled to obtain for us.

My colleagues and I placed calls throughout the military’s public affairs apparatus in Baghdad and Washington, hoping to set up back channels. We asked officials to provide us any conclusive evidence, even off the record, that would give us faith in the Army’s findings.

We never received this cooperation. But conservative bloggers who were fixated on this controversy—one arrived unannounced at TNR’s offices with a video camera, another later attempted to organize an advertiser boycott of the magazine—were treated differently. After we had posted an online statement explaining that we had been unable to communicate with Beauchamp—who, according to Reeve, was under orders not to speak with us—and pleading with the Army to make him available to us, General David Petraeus’s spokesman, Steven Boylan, told the Standard, “We are not preventing [Beauchamp] from speaking to TNR or anyone.” One of our editors called Boylan’s office on a near-daily basis to set up a phone call with Beauchamp; every time, they told us they were working on our request. After several weeks, we stopped hearing back from them. The Army later confirmed to us that it had, indeed, prevented Beauchamp from speaking.

In the meantime, I told reporters from the Times and the Post that we’d fully cooperate if they wanted to dig further and visit Beauchamp’s combat outpost. The Post never responded, and an Iraq-based reporter from the Times declined.

On a Thursday morning in early September—over a month after that final, hectic instant-messaging session—the Army finally brokered a call with Beauchamp. Over in Iraq, Beauchamp sat in a room with his squad leader and a public affairs officer. He spoke on speakerphone.

In contrast to previous conversations, Beauchamp sounded uneasy, at times even catatonic, as he repeated variations on the same: “I just want to not think about this anymore and just basically do my job. And that’s all I really want to do.” More worrisome, he refused to even talk about his stories with us.

Since then, some have said that we shouldn’t have continued with the conversation under these conditions: Beauchamp was in an impossible position. We were prodding him to confess in front of superiors who might punish him. But TNR had been out of contact with him for six weeks; we had to find out whatever we could.

Beauchamp told us that the Army had scheduled calls with other news outlets in which he would say that he had no interest in further discussing his article and to demonstrate that the Army wasn’t censoring him. We asked him to cancel those interviews, because we believed that he owed us answers first. The exchange had left us shaken. How could we stand by Beauchamp’s story if he himself was refusing to do so? We began to think that we had no choice but to retract his story. But, then, Beauchamp reached out to us through his wife. He said that, during our call, he’d spoken under duress. We worried that Beauchamp was just conveying to us what we wanted to hear, but his protestation did seem plausible: The statements he made to us had seemed unusually formulaic and coached. (Our suspicions about the latter were later borne out when the Army included what seemed to be talking points at the end of a transcript of our call, which was leaked to the Drudge Report: “official statement: I don’t want to do any more interviews. I want to concentrate on my job as a Soldier right now. It’s more important to me and not only that—but it’s more important to my country and the other Soldiers around me. And that’s what I’m going to do.”)

Most importantly, Beauchamp said he might want to publish another statement standing by his story. We put off any decision to retract and began working out the timing and terms of our next conversation with him.

The following Monday, September 10, the conservative blogger Confederate Yankee posted an interview with Major John Cross, the executive officer of Beauchamp’s battalion who led the official Army investigation. This surprised us: We had repeatedly requested to speak to someone with substantive information on the investigation and were never told of Cross’s availability. After reading the exchange with Confederate Yankee, we booked time with him later in the week.

In our interview, surprisingly, Cross bolstered Beauchamp’s credibility. He stated that Beauchamp had never recanted, flatly refuting what Goldfarb and others reported. In fact, he agreed that Beauchamp had carefully crafted his signed statements in an attempt to avoid contradictions. And he admitted that, in his investigation, he had neglected to interview a substantial portion of Beauchamp’s platoon.

Then there were the underlying facts of the case. Even though he argued the events in Beauchamp’s articles never happened, Cross conceded that bones were found in the area surrounding Beauchamp’s combat outpost. He guessed that the bones came from animal carcasses. Bradleys, he told us, unintentionally hit dogs. Indeed, dogs flock toward Bradleys. We weren’t sure what to make of these statements.

Was Beauchamp a liar? we asked.

“Well, I can’t state, you know, when you talk about lying, it’s making a statement, oral or written, with the intent to deceive. What I did, like I said before, was check into the veracity of the allegations made.”

We pressed him.

“I can’t say whether or not he wanted to deceive people, and that’s as far as I’m going to say on that point.”

Despite all the commotion he caused, Beauchamp had returned to serving with his unit. We asked Cross how we should weigh the testimonials we received from Beauchamp’s unit.

He answered: “Yeah, I would definitely tell you it’s a minefield. Um, one that I wouldn’t want to find myself in.”

Beauchamp’s writings had originally appealed to us because we wanted to publish a soldier’s introspections. We still believe in this journalistic mission, especially as the number of reporters embedded in Iraq dwindles. But, as these months of controversy have shown, telling the story of what is happening in Iraq through a soldier’s eyes is a fraught project. The more we dug into Beauchamp’s writings, the more clear it became that we might have been in the realm of war stories, a genre notoriously rife with embellishment. It is telling that Beauchamp and his comrades gave the disfigured woman mythological names—Crypt Keeper, Mandrake’s Bride—and made her the subject of telling and retelling.

For the past four-and-a-half months, we’ve been reluctant to retract Beauchamp’s stories. Substantial evidence supports his account. It is difficult to imagine that he could enlist a conspiracy of soldiers to lie on his behalf. And they didn’t just vouch for him—they added new details and admitted gaps in their own knowledge. If they were simply lying to protect him, they likely wouldn’t have alerted us to Beauchamp’s Kuwait mistake. Furthermore, our conversation with Cross confirmed important underlying premises—the existence of bones, Bradleys running over dogs.

Finally, we had obligations to the writer, whatever anxieties we might have had about these pieces. For long stretches, the military prevented Beauchamp from defending himself against his accusers. Even when he was allowed to speak with us, he did so under obvious duress. And the Army’s behavior—its initial efforts to bury the results of its investigation, not to mention the four months and counting it has taken to process our Freedom of Information Act request for those results—made us reluctant to rush to judgment.

But, after our re-reporting, some of our questions are still unanswered. Did the driver intentionally run over dogs? Did he record his kills in a little green notebook? We’ve never been able to reach the driver. And Beauchamp told us that he’d procure a page from the notebook, but that has not materialized. This is a plausible anecdote, and several soldiers in Beauchamp’s unit had heard stories about dog-hunting, but only one had actually seen the driver Beauchamp wrote about intentionally hit dogs. He is one of Beauchamp’s friends, and, over the course of a number of e-mail exchanges with him, our faith in him has diminished.

Several weeks after the monitored call in September, we finally had the opportunity to ask Beauchamp, without any of his supervisors on the line, about how he could mistake a dining hall in Kuwait for one in Iraq. He told us he considered the detail to be “mundane” given the far more horrific events he had witnessed. That’s not a convincing explanation. If the event was so mundane, why did he write about it—and with such vivid detail? In accounting for the inaccuracy of a central fact, he sounded defensive and evasive.

Beauchamp has lived through this ordeal under the most trying of conditions. He is facing pressures that we can only begin to imagine. And, over the course of our dealings with him, we’ve tried to give him the benefit of the doubt. Ever since August, we’ve asked him, first though his wife and lawyer and later via direct e-mail and phone calls, to personally obtain the sworn statements that the military had him draft and sign on July 26. And, ever since then, he has promised repeatedly to do just that. We are, unfortunately, still waiting.

In retrospect, we never should have put Beauchamp in this situation. He was a young soldier in a war zone, an untried writer without journalistic training. We published his accounts of sensitive events while granting him the shield of anonymity—which, in the wrong hands, can become license to exaggerate, if not fabricate.

When I last spoke with Beauchamp in early November, he continued to stand by his stories. Unfortunately, the standards of this magazine require more than that. And, in light of the evidence available to us, after months of intensive re-reporting, we cannot be confident that the events in his pieces occurred in exactly the manner that he described them. Without that essential confidence, we cannot stand by these stories.