His students don't see it as an obsession, not in the crazy single-minded sense. More like a passion.
Questions about the evidence against four young men convicted of a Fairbanks murder have gnawed at Brian O'Donoghue since his newspaper days, when readers told him there was something wrong. For more than seven years, the case has generated a rich vein of real-life assignments for his journalism students.
And now the work may help one of the defendants win a new trial.
"I have many times wished I could walk away from it, but there was always one more door to open," said O'Donoghue, an assistant journalism professor at the University of Alaska Fairbanks.
One of his students discovered an unauthorized experiment dreamed up by jurors to test the credibility of an eyewitness identification: The jury walked outside during deliberations to check how far away people could be recognized.
They apparently weren't concerned that their experiment took place on a summer's day in Anchorage, while the murder of 15-year-old John Hartman happened in the shadows on an October night in Fairbanks.
Prosecutors said the jury experiment didn't matter. The defense said it was everything. The judge ruled it was grounds for a new trial. An appeals court confirmed jury misconduct but agreed it didn't matter, so no new trial. Now the case is before the Alaska Supreme Court.
A rogue jury is just one lingering issue in a case where lack of physical evidence is worrisome, O'Donoghue said.
And so over the years, he has set students to dig through court records and interview defendants and examine alternate suspects. He studied the science of confessions, eyewitness identifications and boot print evidence. Students created a Web site loaded with police interviews, jailhouse interviews, trial summaries and an interactive map and timeline of downtown Fairbanks the night Hartman was killed.
"The real life application for a journalism student is pretty amazing," said UAF grad Jade Frank, who made a spring break side trip in March 2004 to an Arizona prison. There, she nervously interviewed three of the convicted killers.
Sure, maybe O'Donoghue is "slightly obsessed" to have stuck with the project so intently, Frank said, but it's for the good. Students would rather be researching a murder than writing "news" stories about a building dedication or a politician coming through town, she said.
"If he were to call me up and say that there had been a new break in this case and ask me to help out, I would drop everything to work for him again," Mark Evans, another of O'Donoghue's former students, wrote in an e-mail. At one point, Evans hightailed it down to the state prison in Seward to interview one of the convicted men and someone the class decided was an alternate suspect. Nothing big came of it.
Prosecutors stand by the convictions and say they'll continue defending their work in court, refusing to talk about the case for this story. After all, they say, three separate juries looked at the evidence and all agreed the young men were guilty of second-degree murder and assault. Fairbanks police didn't return a call.
O'Donoghue says he started out to confirm the convictions were valid, to prove doubters wrong. But despite, as the student Web site puts it, "a painstaking reconstruction of what took place the night of the murder and in the trials," he and the students remain unsatisfied.
The night John Hartman was killed, Fairbanks was especially rowdy. Alaska Permanent Fund checks had just come out. A couple of hundred people were dancing and partying at a big wedding reception at the Eagles Hall downtown.
On that night in October 1997, a group of young men -- prosecutors say it was the four now convicted -- went on the prowl. They mugged a man then, prosecutors say, randomly accosted Hartman, who was kicked in the head and sexually assaulted. He died the next day.
"A kind of a wilding thing," is how assistant attorney general William Hawley summed it up to the Supreme Court on Tuesday.
The murder stirred up so much furor in Fairbanks that the trials were moved to Anchorage. Racial tensions over the case -- all the defendants except Kevin Pease are Alaska Native and Hartman was white -- continue to this day.
Once or twice a year, protesters still take to Fairbanks streets.
Last year, Jerry Isaac, president of Tanana Chiefs Conference, spoke at a rally for the convicted killers and called for the cases to be reopened.
"We have four young men who may have not committed the crimes they were charged with, and they've wasted 10 years out of their lives and still going. And all I want is answers," Isaac said Friday in a phone interview.
O'Donoghue is a former reporter for both the Fairbanks Daily News-Miner and the Anchorage Daily News. He ended up as Fairbanks editorial page editor and says his interest in the Hartman murder was fueled by numerous letters to the editor saying "this case stinks." He says he switched jobs and began teaching at the university in 2001 in part to have time to research it. The project resulted in a seven-part series in the News-Miner last year.
Critics focus on the absence of physical evidence. No fibers, no hair, no blood, nothing tying the men to the crime other than one iffy boot print.
The state relied on confessions by two of the suspects -- one not yet 18 -- both of whom recanted, O'Donoghue said. And defendant Marvin Roberts, a good student, had an alibi.
"In America, none of them should have been convicted off of what was produced," the journalism professor said.
Many cases are built without DNA, fingerprints or other crime scene evidence, despite expectations for physical proof elevated by what prosecutors have come to call the "CSI effect."
All four defendants have lost their appeals in state court. The push for a new trial based on the jury experiment is on behalf of only one defendant, Pease, said his attorney, Lori Bodwell. George Frese and Eugene Vent were tried earlier, and Roberts lost his shot at a new trial before the jury experiment was revealed.
At the time of the killing, the four ranged in age from 17 to 20. They are serving sentences ranging from 33 to 79 years.
Sharice Walker was a single mom of two taking O'Donoghue's investigative reporting class in the fall of 2002, in the early days of the Hartman project.
She and another student examined jurors to learn what evidence convinced them to convict. A few jurors didn't want to talk, but then a jackpot. In an interview, one juror asked Walker: "Did anyone tell you we did a little demonstration?"
During deliberations, outside the Anchorage courthouse, one of the jurors paced out a few hundred feet to test whether the key witness for the state, Arlo Olson, could have seen what he testified he saw.
The night Hartman was killed, Olson was "half shot" he told jurors. He saw a group of four men mugging a man. Police say the same group stomped Hartman to death later that night. Olson eventually came forward to say he recognized the muggers from pictures that already had been published in the paper. Police said that was the link they needed -- the muggers and the killers were the same, they said.
No bystander witnessed the murder. Olson was the only person to put the four together that night, and, as a police sergeant once testified, he was 550 feet from the mugging -- nearly two football fields away.
Other jurors later confirmed the experiment, and two said it bolstered Olson's credibility, according to the students.
"Give credit where credit is due. ... We interviewed some of the jurors and they never mentioned it," Bodwell, Pease's defense attorney, said Friday.
That prosecutors might be annoyed with O'Donoghue and his students is understandable, but defense attorney Bodwell also has bones to pick. She said she's frustrated that the professor hasn't given her recordings of his interviews with Olson. In the series, O'Donoghue quoted Olson saying "I didn't want to testify" and "I told them I wasn't sure." Statements like that could help the defense, she said.
But O'Donoghue said he can't help one side or the other. He's a journalist, and won't turn over raw notes or recordings to anyone.
"My job is not to be an agent for the defense. My job is not to be an agent for the police," he said.
WILL IT EVER END?
While O'Donoghue uses elements of the Hartman case in almost every class, these days it's often just to illustrate a point about public records, or the value of follow-up interviews.
"It is not one of our primary focuses," said Cassandra Johnson, a student who went to the Supreme Court hearing and shot photos, one of which ran in the next day's News-Miner.
O'Donoghue is thinking about a book, but says the story is still too much in flux for that.
In the Hartman project, like most investigations, one person led to another, then usually to a dead end. It took him two years to find a woman who had reportedly warned Hartman's older brother about trouble the night of the murder, he said. By the time he talked to her years later, she couldn't remember.
"It really has been a pretty absurd Kafka-esque situation of retracing things that would have been easy for police to look into at the time," he said.
Find Lisa Demer online at adn.com/contact/ldemer or call 257-4390.