As by air and earth and water, so also man lives by fire.
—George R. Stewart, Fire
When there’s a fatal fire and someone survives, the survivor will be charged with arson and murder.
—Gerald Hurst, Ph.D.
BY ITS VERY NATURE, arson is a difficult crime to prove. In the past it has also been a difficult crime to disprove. A perpetrator can easily start a fire when he is somewhere else—the material he requires is easily obtained, and the fire itself destroys all the evidence.
In his 1906 text Criminal Investigation, Hans Gross lists some of the agents used to start fires: candles, saltpeter-impregnated strips of tinder, and matches fixed to alarm clock clappers. He relates that in America:
A particularly dangerous and common method . . . is to utilize an ordinary electric bell. The sounding part of the bell is replaced by a thin, balloon-shaped glass filled with sulphuric acid. The hammer strikes on the glass and breaks it, the sulphuric acid runs into a vessel placed beneath and filled with a mixture of . . . chloric acid and sugar. This produces fire that can easily be converted into a conflagration.
According to Gross, there was once a miller’s boy who started a fire in the cottage of a farmer he had a grudge against. What makes the fire noteworthy is that it was not until nine months after the boy had gone away that the farmer’s house, situated next to the mill, caught fire and burned to the ground at midday, a time when everyone was working in the fields.
What the clever, evil-minded lad had done was this: He first attached a strong spring to a window in the slanted roof of the mill. This way when the window flipped open, it would throw whatever might be sitting on it onto the thatched roof of the farmer’s house next door. The boy held the window shut with a string tied to the window on one end and to a ring in the wall on the other. Then he sealed the window with pitch. Next, he spread a flammable material along the path underneath the string. Then he fastened a magnifying glass (a “burning glass”) to his apparatus in such a way that at a certain time of year the rays of the midday sun would be focused on his flammable material. Not until a year later was the sun in such a position that its rays were focused through the glass onto this fuel. When it caught fire, in turn it ignited the pitch and the string. As the string burned, it released the spring, which flipped open the window and tossed the burning pitch onto the farmer’s roof.
According to Gross, this demonstrates “how with a little skill and ingenuity, most extraordinary things may take place.”
The crime of arson is defined as “a malicious burning of property.” There are more than 300,000 structural fires in the United States every year. About 75,000 of these are estimated by the FBI to be “of criminal or suspicious origin.” The consensus is that 15 to 25 percent of fires involving property are arson. In times of economic distress, this figure rises. The reasons for the crime of arson are varied, and the arson investigator must first determine the arsonist’s frame of mind in order to know where to look for him. Statistics show that about 90 percent of arsonists are male, and half are under the age of eighteen.
The most common motives for arson are:
—Hate or revenge. Setting fire to a church or synagogue or to the school or orphanage of an ethnic or religious group the arsonist dislikes is classified as a hate crime. Setting fire to the store you have just been fired from or the home of the girl who has just rejected you is an act of revenge. Almost half the arson fires set in this country are motivated by revenge.
—Profit. After revenge, profit is the second most common motive for arson. There are many ways to profit from burning up your own or someone else’s property: collecting on insurance; getting rid of a structure you do not want and that the city or state will not permit you to tear down; getting rid of undesirable tenants; extortion; and slum clearance (so you can put up that high-rise condominium). In the Berkeley, California, hills in the 1980s, a series of grass fires was set by a part-time firefighter who needed the overtime pay.
—Thrills. Some people are sexually aroused by the sight of fire. Most of the pyromaniacs are men whose fire-setting began when they were quite young, possibly only eight or nine years old. The fire-setting is often compulsive—fires are repeatedly set in the same place or type of place, at the same time of day, and in the same manner.
—Egoism. People, sometimes even firefighters, have been known to set fires in order to appear heroic as they rescue those in danger or help put the fire out.
—Social protest. Some people are so caught up in a social or political cause that they blow up abortion clinics, recruiting stations, or federal buildings in the mistaken belief that they are performing a useful social service.
—Concealment of another crime. After a murder or robbery, what better way to get rid of the evidence than to burn it up?
Since the fire itself destroys evidence (and the longer it burns the more it destroys), an arson investigator will try to get to the scene as quickly as possible and will carefully question anyone—firefighter or civilian—who was there earlier. Some of the things the investigator will want to know are: How much smoke was there early in the fire? What colors appeared in the flames and in the smoke? Was there anything unusual about the room in the house—the presence of something unexpected or the absence of something expected?
Different materials produce flames of different colors as they burn. Smoke too will have a characteristic color. Gasoline and kerosene burn yellow or white and give off black smoke. Except for treated wood, wood generally burns yellow or red and gives off a grey smoke. If there are places on the wall where two or three pictures have been recently removed, this might be a sign that someone expected a fire.
The body of lore, rules, adages, and techniques surrounding arson and its investigation has accumulated gradually over years of practical experience by firefighters and arson investigators. But until recently much of this informed body of knowledge had not been tested scientifically. And when it was tested, much of it turned out to be incorrect.
For decades firefighters and arson investigators believed (and taught their students to believe) that:
—Fires always burn up, not down. After all, heat rises.
—Fires that burn very fast have been set using accelerants, such as kerosene or gasoline, and are therefore the results of arson.
—Accelerant-fueled fires burn hotter than normal fires.
—Separate burn holes on the floor or uneven burn patterns in one room or throughout the house indicate multiple points of origin. Accidental fires start only in one place.
—Melted copper wire or melted steel is a sign that accelerants were used.
—Bed springs and furniture springs do not collapse in a normal, or accidental, fire.
—Blistering on the walls is a sign of a hotter-than-normal fire.
—“Crazed glass”—glass that is replete with small irregular cracks—is a sign of an extra-hot fire.
None of these things is necessarily true, and some of them are quite false. But it was not until the 1990s that scientific evidence refuting the received wisdom of the ages was developed. Even today people are still being tried and convicted of arson, and murder, on the basis of faulty science.
On October 20, 1991, a wildfire in the hills of East Oakland, California, spread to nearby residential areas and burned down nearly three thousand homes before it was extinguished. Four arson investigators decided to use this tragedy as an opportunity to test some commonly held beliefs about arson fires. They found, to their surprise, that much of the conventional wisdom regarding the unmistakable signs of arson was simply not true. Fire investigator John Lentini, one of the four, has written extensively about their findings and become a vocal advocate for them. He often testifies in cases in which the prosecutors are trying to convict someone of arson by using evidence now understood to be invalid.
“Crazed glass” was supposed to be the sign of the use of accelerants. This turns out not to be so. A very hot fire will cause glass to melt but not to craze. Investigators found crazed glass in houses at the periphery of the Oakland Hills fire. All of them were houses that the fire departments had sprayed with water. It turns out that the heat from the fire alone did not craze the glass. The crazing occurred when the glass was doused with cold water when it was very hot.
So the “expert” testimony that had put people in prison over the years is wrong. Starting a fire does not cause crazed glass, but trying to put it out does.
Investigators also learned that melted copper wire was routinely found in most of the Oakland houses, and that some coils of steel (bedsprings) that appeared to have lost their tempering—which had been thought to have suffered extended high heat—actually lose tempering at a fairly low heat. And some coils which seemed to be melted, indicating an unreasonably high temperature, were actually heavily oxidized by the heat. They appeared to mimic melting unless inspected microscopically.
A phenomenon that earlier fire investigators had either missed or discounted was something called “flashover.” When a fire burns in a comparatively closed space, say a room, the heat and gases rise to the ceiling and build up there until a critical point is reached. Then, in a nearly instantaneous flash, the whole room is ablaze. Often the room then burns down from the ceiling, not up from the floor. This can happen in a matter of minutes, causing a fire that burns rapidly and without the aid of accelerants or multiple points of origin.
During the night of April 7, 2003, Rose Kate Roseborough was sleeping on a couch in the living room of her home in Ashland, Ohio, when a fire started upstairs where her eleven-month-old twin daughters, Lucie and Julia Bursley, slept. She tried to reach their room and rescue them, but the smoke and heat were too intense. She ran out of the house screaming hysterically, unable to get to the telephone because it had been left upstairs in the twins’ room. A neighbor called the fire department, but it was too late to save the twins, who died in the fire.
Kevin Rosser, the emergency medical technician who answered the call, claimed to notice “large-particle soot” on Roseborough’s face at the scene. She was later arrested and brought to trial for arson and two counts of murder. Rosser, claiming expertise in fire examination, testified that because large-particle soot is formed only at the beginning of a fire, Roseborough must have set it herself. The defense attorney asked for a Daubert hearing (a hearing on the validity of scientific claims given in testimony), but his request was refused. In denying it, the trial judge cited a case he called Tomlinch, concluding that it “was easy to tell he had chopped off his thumb; he had chopped off his thumb because it was gone.” Roseborough was convicted and sentenced to life without the possibility of parole.
Fortunately for Roseborough, her attorney found a real expert before her appeal. At a deposition on April 23, 2008, Gerald Hurst, an expert on fire chemistry and dynamics, testified that EMT Rosser had gotten it wrong. In fact, during a room fire the gradual depletion of oxygen and build-up of soot cause the large soot particles to form toward the end of the fire, not at the beginning. Hurst testified that not only did many other experts share this knowledge, but that anyone with a master’s degree in chemistry could have figured it out. On January 6, 2009, the original verdict was set aside by the appeals court, and a new trial was ordered for Roseborough.
After immigrating from South Korea, Han Tak Lee, his wife Esther, and their two daughters settled in Queens, New York, where Lee ran a clothing business. Lee’s older daughter, Ji Yun, suffered from mental illness and in the past had attempted suicide. On Friday, July 28, 1989, Lee and Ji Yun had a loud argument over her refusal to take her medication. Then, when Ji Yun threw a clock through a screened window the police and the pastor of the Lees’ Pentecostal church were summoned to the apartment. The pastor suggested that Lee take Ji Yun to the Hebron Camp, a religious retreat run by the Korean Assembly of God Church in Stroud Township, Pennsylvania.
Ji Yun went with her father to the Hebron Camp that very day, followed later by her mother and sister. That night she had a screaming fit, and two pastors staying at the camp had to help Lee restrain her. A few hours later the Lees’ cabin caught fire.
By the time the fire department arrived, the interior of the cabin was completely ablaze and flames shot out of the windows. Han Tak Lee sat on a bench a short distance away, staring silently into space. Or, as the Stroud Township Police Department put it in their report, Lee “remained complacently seated throughout the fire.”
In his statement to the police, Lee said that he fell asleep after spending the evening in prayer. He was awakened by the smell of smoke and could see that his daughter’s bedroom was on fire. He ran outside to see if she was there, but she was not. He then ran back inside, threw their suitcases out the door, and banged on the bathroom door. He got no response from his daughter. Finally, overcome by smoke, he ran out the back door.
Trooper Thomas Jones of the Pennsylvania State Patrol was the lead investigator for the incident. His report says that the girl’s body was found in the hallway of the cabin, and that “Near the rear door of this hallway, a spill or flow pattern was observed. At this point, it was discovered that the fuel filter on the furnace had been tampered with and unscrewed, spilling fuel on the floor.”
The report goes on to say that “The window on the rear wall exhibited very fine crazing of the glass and was burnt clean.” Further, “spill patterns and deep char patterns” were found on the floor in various places that were “inconsistent with a normal fire.”
And as for Han Tak Lee, he “remained almost emotionless and while in view of this officer made no attempts to console his wife.”
On August 4, 1989, Han Tak Lee was extradited from his home in New York to answer charges of arson and murder filed against him in Stroud Township, Pennsylvania. The state assembled several expert witnesses against Lee. The first was Daniel Aston, a certified fire-protection specialist. The state based its case almost entirely on the findings in his report of December 20, 1989. Aston compared the timeline of the fire, as he reconstructed it, against a standard time-versus-temperature graph developed in 1918. Although the curve on this graph was originally meant to describe the operation of a furnace used for testing the fire resistance of various materials, it had somehow worked its way into at least one standard textbook as a means of determining whether a fire was “normal.” As fire expert John Lentini put it, “Despite numerous experiments on real fires which show that this . . . time/temperature curve has no relation to reality,” it was an important part of Aston’s reconstruction of the fire.
Aston assigned a total weight of 7,788.28 pounds to the amount of “Class A combustible materials” in the cabin. He did this by adding the estimated weights of the cabin’s contents—the sofa (60 pounds), the carpet (386.9 pounds), the mattress (32.85 pounds), and so on. He then determined the energy content of this weight of fuel, made a few incorrect assumptions about how wood burns, and created a table plotting the “normal” burn rate of such a structure against the “actual” burn rate of the cabin. He concluded that it would have required not only all of the 62 gallons of fuel missing from the cabin’s fuel tank (assuming it was full) but an additional 12.2 pounds of gasoline. If Aston was right, the cabin would have been awash in flammable liquids when the fire started. And the cabin, according to Aston, burned 81.6 percent more severely than “normal.”
When asked how many fires he had examined, Aston replied, “Perhaps in my career I’ve probably experienced some 15,000 fires that I’ve been called upon to determine the cause for.” Now, Aston had been a part-time fire examiner for about twenty years. And if we take his number seriously, it means he investigated an average of fifteen fires a week in addition to holding down a day job designing sprinkler systems. According to John Lentini, a busy full-time fire investigator may work fifteen fires a month. The defense attorney did not do the math and failed to challenge Aston on his numbers.
Convinced as well that the fire must have been started intentionally, Lee’s own attorney tried to make the case that Ji Yun had started the fire herself in an attempt to commit suicide. One assumes that if Lee had started the fire he would welcome this theory. In fact he was not at all happy with it—his daughter would not have committed suicide because it was against her religion.
“You know from common sense that this fire was started effectively,” the district attorney began his summation to the jury. “Whoever did this knew what they were doing.”
Unfortunately “common sense” doesn’t work well when you know nothing about the subject. In fact the jury relied on the testimony of three dubious “experts”:
Aston, who added in his testimony that “A shiny alligatoring pattern generally indicates a very quick fire, a dull alligatoring pattern indicates a slow, very slow fire.” In fact, the alligatoring pattern on the walls indicates nothing at all about the speed of the fire.
Trooper Jones, who misinformed the jury about the significance of the crazed glass (testimony that Aston quoted and reinforced) also testified that the lack of communication between the several burn holes in the floor showed that the fire was intentionally set. In fact, this means that the roof caved in.
A chemist named Thomas Pacewicz, who testified that a “substantial quantity” of gasoline was found. On cross-examination, when asked why there was no mention of gasoline in his written report, Pacewicz said it was because “There wasn’t a sufficient quantity of accelerant to confirm by our instruments.” So a “substantial quantity” of gasoline was still not enough to confirm with his instruments?
Han Tak Lee was convicted of murder on September 17, 1990, and sentenced to life in prison. Several appeals have been turned down. John Lentini, who investigated the case, had this to say about it:
The quality of the evidence presented by the Commonwealth speaks for itself. Fuel loads calculated to six significant figures, hydrocarbon “ranges” being interpreted as evidence of a mixture, furnace operating instructions being touted as normal fire behavior, and a host of other “old wives’ tales” were used to convict Han Tak Lee.
Agreeing with Lentini, David M. Smith, a former bomb and arson investigator in Tucson, Arizona, said, “That’s a perfect example of a system run amok.”
At about four in the morning of Monday, June 30, 1986, a fire broke out in the apartment outside of Cleveland, Ohio, where two-year-old Cynthia Collins slept. Kenny Richie, a twenty-one-year-old friend of the girl’s mother, was baby-sitting Cynthia while her mother was out with a friend. He was asleep and admittedly drunk in the next room. The girl was killed in the fire.
The next day Assistant State Fire Marshal Robert Cryer decided that the fire had been caused by arson—he believed that some of the burn patterns in the apartment indicated the use of accelerants. A search for empty accelerant containers turned up nothing. Nonetheless Richie was accused of felony murder on the theory that he had tried to kill his ex-girlfriend, Candy Barchet, and her new boyfriend, Mike Nichols, who lived in the apartment below. A state forensic chemist testified that gas chromatograms from the Ohio Arson Crime Laboratory showed gasoline residue on a sample from the living room carpet, and that a sample of wood from the balcony indicated the presence of paint thinner. Because the spaces between the boards had burned more than the tops of the boards, he concluded that liquid accelerant had flowed into the cracks.
William Kluge, Richie’s’s court-appointed attorney, with a budget of only $3,000 for expert testimony, hired L. Gregory DuBois of the firm CTL Engineering. DuBois had no experience or accreditation in arson investigation, did not interview the fire marshal, and did no testing of his own on the samples the state used to support its claims. Added to the fire marshal’s opinion was the testimony of Peggy Villearreal, the next-door neighbor, who said she had heard Richie threaten to burn down the apartment building.
Despite the clear, undisputed evidence that Richie had tried desperately to get into the bedroom to save the child, Richie was convicted and sent to Ohio’s death row, where he remained for the next twenty years.
In 2005 the Federal Court of Appeals for the Sixth Circuit found that “counsel’s incompetent handling of the sole forensic expert in this case fell far below the wide range of acceptable professional standards.” The court ordered that the State of Ohio retry Richie within ninety days or release him.
The court had learned that after the conviction, Peggy Villearreal had changed her story and said that she had never heard Richie threaten to burn anything. She also remembered that little Cynthia had a habit of playing with matches and had once set fire to the couch. But the principal factor in the court’s decision was the testimony of several forensic experts that the evidence of arson cited at the trial was incorrect. Tony Café, an Australian arson expert who believed that an article of his had been misunderstood by the prosecution’s experts and misquoted at the trial, filed an amicus brief in an attempt to put things right. He pointed out that flashover can make a fire look as though accelerants were used, and further that even without accelerants, the spaces between boards burn more than the surface because there is a good supply of air coming up through the cracks. As for the chromatograms showing gasoline on the carpet: “I cannot see any evidence whatsoever in the chromatograms that indicate the presence of an accelerant. Simply put, the chromatograms for the samples from the fire debris do not resemble the chromatograms from the standard gasoline or standard paint thinner.”
And if that was not strong enough, Café added:
I am sure that most of the world’s leading forensic scientists in this field would be horrified if they saw the chromatograms used to convict Kenny Richie. If Kenny Richie were executed on the basis of this scientific evidence, then these chromatograms will become historical documents, examined by scientists all over the world and used to show just how wrong forensic evidence can be. It would be a great tragedy for the future of forensic science.
Amnesty International joined in, calling the Richie case “one of the most compelling cases of apparent innocence that human rights campaigners have ever seen.”
But the state did not give up without a fight, and two years later, in 2007, the court had to repeat this same ruling. Ohio had appealed the first ruling to the Supreme Court, saying that since Richie (actually his lawyer) had not raised the question of “ineffective assistance of counsel” in earlier appeals, he could not introduce it now. The Supreme Court said that he could and sent the case back to the Sixth Circuit. There Judge R. Guy Cole Jr., wrote in the majority opinion that “the deficient performance of Richie’s counsel undermines our confidence in the outcome of his trial.” Again the Court gave Ohio ninety days to release Richie or give him a new trial.
This time Ohio released Richie, but not without one final dig. Richie had to plead “no contest” to a charge of child neglect and agree never to return to Putnam County, Ohio. As soon as he was released, Richie went back home to Scotland, though he later returned to the United States.
The old beliefs about how fires burn and about the telltale signs of arson are with us yet. People are still convicted of arson and even of murder based on unscientific or disproved theories. As new textbooks are written and newly trained investigators enter the field, this will gradually change. Until then people will continue to face the double horror of losing loved ones and of being held responsible for their deaths.