This month — nearly 25 years after the blaze — a federal magistrate recommended that Lee should either be given a new trial or released from prison outright. A federal judge must approve the recommendation, and prosecutors are expected to file papers this week arguing that Lee's conviction should stand.
His case is one of dozens around the country to come under scrutiny because of entrenched but now-discredited beliefs about how arson can be detected. The Arson Research Project at the Monterey College of Law in California has highlighted at least 31 convictions based at least partly on debunked fire investigations, including that of a Texas man executed in 2004, and experts believe there are many more.
"There was just no science behind" the old assumptions about arson, said Paul Cates of The Innocence Project, a group that works to overturn wrongful convictions, primarily through the use of DNA. "A lot of this was just guesswork and voodoo."
Lee, now 79, has consistently maintained his innocence.
A clothing store owner in New York City, Lee had taken his volatile, mentally ill 20-year-old daughter to a northeastern Pennsylvania religious retreat at the suggestion of the family pastor. Early on July 29, 1989, the cabin they shared became engulfed in flames. Lee escaped, but his daughter's body was found in the ashes, curled in the fetal position.
When firefighters showed up, they found Lee sitting stoically on a bench outside the cabin. Inside the wreckage were clues that led authorities to suspect foul play.
At the time, investigators were taught that unusually hot and intense fires indicated the use of an accelerant and that arson could be confirmed by the presence of deep charring or shiny blistering of wood as well as "crazed glass," tiny fractures in windows.
Research conducted in the 1980s debunked these and other notions about arson. By 1992, the National Fire Protection Association had published new standards to guide fire investigations.
But acceptance did not come right away.
"Most arson investigators' heads exploded, and they just went nuts for the next seven or eight years trying to discredit that document," said John Lentini, one of the nation's leading experts in fire analysis and a defense consultant for Lee.
At Lee's trial, a fire marshal and other prosecution experts said the physical evidence overwhelmingly pointed to arson. In addition to the telltale burn patterns, they said, testing revealed accelerants on Lee's shirt, pants and a jug found in the cabin.
Lee's attorney didn't challenge the arson finding, arguing instead the fire was set by Lee's troubled daughter to commit suicide. The jury didn't buy it.
A quarter-century later, U.S. Magistrate Judge Martin Carlson said scientific progress had invalidated the conviction.
"Over the past two decades, there has been a revolution in fire science," he wrote June 13. "It is a revolution that has toppled old orthodoxies, and cast into doubt longstanding assumptions regarding fire scene analysis."
Monroe County prosecutors concede the point but say other evidence points to Lee's guilt, including the accelerants found on his clothing.