Arson convictions
challenged by science
By Maurice Possley
Tribune staff reporter
Published October 18, 2006, 8:28 PM CDT
~~~~~~~~~~~~~~
Cases Referenced: Amanda Hypes / Louis Taylor / Daniel Dougherty / Dennis Counterman / Todd Willingham
Wrongful Conviction Factors: Junk Science
~~~~~~~~~~~~~~~
Flames shot from a first-floor window and smoke curled from the second floor as Amanda Hypes pulled up to her frame house near Alexandria, La., in January 2001.
Hypes called to her oldest child, 10-year-old Sadie. "I yelled from the stairs, 'Sadie, just come down the stairs!'" she told investigators the following day. "I yelled and I yelled… She wouldn't answer me.
Hours later, when firefighters entered the smoldering home, they found Sadie, her 6-year-old brother, Luke, and their 3-year-old sister, Jessica, dead in the debris.
In April 2002, a Rapides Parish grand jury indicted Hypes on charges of arson and murder based on a California fire expert's findings—an analysis conducted more than a year after the blaze was extinguished and the house was razed. Prosecutors said they would demand the death penalty.
Hypes remained in jail for more than four years awaiting trial until this June, when a judge dismissed the indictment and ordered her released. He ruled that the initial arson finding by Louisiana authorities was based "merely on an old wives' tale" and that "every shred of evidence to prove or disprove a possible crime was destroyed and placed in a pile."
The arson prosecution of Hypes is one of a growing number where defense experts are challenging cases based on science that undermines long-held theories. They represent a new wave of courtroom challenges that are causing older cases to be re-examined and are pitting expert against expert in the name of science and justice.
The cases highlight not just the evolving science behind arson investigations, but two crucial issues facing the field: an ongoing split on whether to accept scientifically established fact and a lack of training to bring investigators up to date on the latest thinking.
"The current training process is, in most cases, deficient in teaching fundamental knowledge that can be applied to all fires," said Douglas Carpenter, a respected fire investigator.
Earlier this year, Carpenter and Richard Roby, both of Maryland-based Combustion Science & Engineering Inc., along with Jose Torero of the University of Edinburgh, which has long pioneered fire research, called for development of a more advanced science curriculum for fire investigators.
The leading guide to fire science is a manual known as NFPA 921, a publication of the National Fire Protection Agency, an international group dedicated to fire safety. First published in 1992, it is a guide to the scientific debunking of old and unproven arson theories.
"I am not sure how far we have come since the introduction of NFPA 921," said Carpenter. "We certainly have raised awareness, but I do not think the [investigation of fires] has advanced as much as some think it has."
While some in the forensic community equate the significance of the new scientific findings with the advent of DNA, there are key differences. For one, fire science is nowhere near as precise, and in fact makes a strong case that investigators should rarely point to fire-scene evidence as 100 percent proof of arson.
With some investigators refusing to embrace new science, a spate of high-profile court cases is emerging where fire evidence is being strenuously debated. Among them:
Louis Taylor is serving a life prison term for the worst fire disaster in Arizona history—a 1970 blaze in Tucson that killed 29 people. Taylor, then 16, was convicted of setting a series of fires in a hallway of the Pioneer International Hotel to distract authorities while he burgled guestrooms. In a bid to get Taylor a new trial, private fire experts have re-examined the case and say prosecution testimony about the fires has been scientifically debunked.
Dennis Dougherty is on Death Row in Pennsylvania after being convicted of setting a fire in 1985 that killed his two sons, Daniel, 4, and John, 3, in Philadelphia. Newly appointed lawyers expect to ask for a new trial next month, citing expert opinion that the state's proof of arson at trial in 2000 was scientifically wrong. Gerald Hurst, an arson expert, reviewed the case and found the state's proof was based on unreliable evidence.
For the past two years, state prosecutors pursued the death penalty against Dennis Counterman for allegedly setting a 1988 fire in Allentown, Pa., that killed his three sons, even though a fire investigator hired by the prosecution said there was no basis for the original finding of arson. On Wednesday, after 18 years behind bars, Counterman accepted a plea deal under which he maintained his innocence, but admitted prosecutors had evidence that could convict him. He was freed immediately, and the case was closed.
The increasing focus on arson prosecutions was spurred in part by a Tribune investigation that showed Cameron Todd Willingham was executed in Texas in 2004 based on scientifically invalid arson evidence.
Prompted by the Tribune report, Barry Scheck, co-founder of The Innocence Project, formed an arson review committee to re-examine cases where defendants say they were wrongfully convicted of arson.
The Willingham case was the first to be reviewed, and four experts, including Carpenter, concurred with the Tribune findings and called for Texas authorities to re-examine that case and hundreds of other arson prosecutions in that state.
Since the formation of the review committee, The Innocence Project, which has used DNA testing to free scores of wrongfully convicted defendants, has received about 30 requests to review individual arson cases, according to a project spokesman.
The prosecution of Hypes began with the discovery by Louisiana state fire investigators that her home's concrete slab was flaked and chipped—a phenomena known as spalling. A day after the fire, the house was razed and the slab washed clean in an attempt to find more spalling.
Until disproved by scientific testing, spalling for decades was considered proof of arson, the result of heat so intense that it could only have come from a fire fueled by an accelerant. But in recent years, tests have proved that spalling can be caused by fires that involve no accelerants.
Five days after the fire that killed her three children, Hypes was interrogated for hours by Rapides Parish sheriff's investigators in an attempt to get a confession. Hypes insisted she was innocent.
"There was never a chemical in my hand," she said, according to a transcript of the interrogation. "There was never anything to light a fire in my hand. There was nothing. …How this is happening, I'll never know."
When an investigator told her she had murdered her children, Hypes replied, "No, they were not murdered. My babies were not murdered."
"Gas poured all over them," the investigator said, even though there was no evidence of gasoline found in the house. "You won't get a bond, three counts of first degree murder. You won't ever get out."
After Hypes continued to insist she did not start the fire, the interrogation ended and she left to bury her children.
More than a year later, she was indicted after the prosecution hired John DeHaan, a private fire investigator from California, to re-analyze the fire.
In an interview with the Tribune, DeHaan said, "I came along and I said spalling doesn't have anything to do with it. The original investigation was pretty dreadful, and there were very few things I could salvage from that investigation."
DeHaan examined grand jury testimony of witnesses who had been in the house before the fire and described the household furnishings. He also examined the testimony of witnesses who saw the fire after it was discovered. DeHaan concluded there were separately ignited blazes in a first-floor bedroom and the kitchen.
His report, based in part on calculations of the amount of heat that would be generated by the furniture, said that because the fire moved quickly—consistent with the use of an accelerant, though none was found—and because he could find no accidental cause, the fire must have been arson.
As part of his report, DeHaan submitted his findings to David Icove, a fire expert at the University of Tennessee, who conducted a computer modeling and said he agreed with DeHaan's conclusion that the fire was started in two separate places—an indication the blaze was deliberately set.
Earlier this year, defense lawyers J. Michael Small and James Doyle requested the indictment be dismissed, alleging that prosecutors violated grand jury secrecy by providing its testimony to DeHaan.
After Judge Donald Johnson granted the motion, Hypes' lawyers asked that she be released from jail while the ruling was appealed. They also called three fire experts to testify about the prosecution's arson evidence.
The first witness, George Barnes, a retired agent from the federal Bureau of Alcohol, Tobacco, Firearms and Explosives, said he initially declined to work for Hypes because he had never before testified for a defendant in a criminal case.
But after examining the evidence, he changed his mind. "I took this case because the methodology and conclusions were outrageous," he testified. "There was very little if any adherence to the scientific method, and conclusions were based on wives' tales that had long since been proven to be incorrect. And based on the reports, I saw absolutely no indication of an incendiary fire."
Carpenter also was called as a defense witness and severely criticized Icove's computer modeling, noting that the computer program showed temperatures in a hallway where there was no fire to be higher than in rooms where there were flames.
He said Icove's "report is wrong. His modeling is wrong. And any conclusion based on it is wrong."
The defense also summoned John Lentini, a fire investigator who reviewed the Willingham case for The Innocence Project as well. "I was just shocked that this far down the road, where we have done so much work to try and dispel the mythology, that I was looking at a case based on spalling," Lentini testified, calling the examination of the Hypes fire "a case that has been botched worse than any fire investigation I've ever seen."
Lentini said the damage to the kitchen, the living room and a downstairs bedroom were the result of a fire phenomenon known as "flashover," where large areas suddenly explode in flames due to a build-up of gases. Once flashover occurred and Hypes' house was razed, he added, there was no way to determine if the fire was started in more than one place.
"What we've got is John DeHaan…trying to get to an arson determination without any evidence," Lentini testified.
On June 27, after hearing the testimony, the judge ordered Hypes released on bond. "The court finds that proof is not evident that arson even exists," Johnson ruled.
Thomas Walsh, the Rapides Parish prosecutor handling the case, appealed, but last month, the Louisiana appellate court upheld the indictment's dismissal.
Walsh said that should further appeals fail, the case would be "returned to our jurisdiction, and we would be able to go back to the grand jury to resubmit it."
mpossley@tribune.com