Spontaneous Human Combustion

by Brian Dunning
May 17, 2011

Today we're going to point our skeptical eye at one of the mainstays of the paranormal: spontaneous human combustion (SHC). The idea is that people can, while simply minding their own business, burst into flames, with no external source of ignition. It's not a medically recognized phenomenon, and no explanation exists that can reasonably account for any but a few of the many stories. Thus, it's found a firm home in the world of the strange, that subdivision of Earthly phenomena that is studied and promoted by only a few fringe researchers and outsiders. That doesn't mean it's wrong though; and we're going to look at it as closely as we can.

Spontaneous human combustion is a little different from most paranormal phenomena, in that it's a claim of no external source for the fire. That people have burned up is the fact that's not in question; the question is the theory of what caused them to burn. In this case, believers are asserting that there was no conventional cause. Their job is, in effect, to prove a negative. Proving a negative is different from the null hypothesis. The null hypothesis for an unexplained fire is simply to say that no cause is known, which is different from stating authoritatively that there is no cause. Thus, the burden of proof still rests on the claimant, even in this case. Science does not allow us to make the jump from "the cause of the fire is unknown," therefore "the cause is known and it's spontaneous human combustion." Science allows unanswered questions; indeed, science exists because of unanswered questions. The lack of an answer proves only that we don't know something yet, it does not prove the existence of the paranormal.

Stories of SHC generally fall into one of two categories. The first type is the discovery of a body that burned while nobody was present, usually almost completely to ashes but for a few bits like the hands or feet. Even bones are burned away. The second type is a dinner party or some other event, where many witnesses all see one person suddenly go up into flames for no evident reason, and the flames are usually extinguished before the person can be killed. For each of these two types, there are a few very prominently repeated examples that you'll find on the Internet or in books. We'll give two examples of each.

Since this is my show, I'm going to invoke Host's Privilege and formally declare the two kinds. Spontaneous Human Combustion of the First Kind is when there are no witnesses to what happened:

• The most famous such case is that of 67-year-old Mary Reeser, whose remains were found by a friend in her St. Petersburg, FL home in 1951. Only her foot remained, still in its slipper, while the rest of her body had been reduced completely to ashes, along with the chair in which she'd been sitting. Her case is sometimes referred to as "the cinder woman".
  
  
• Another example you're likely to find in the books is that of 92-year-old John Bentley. A meter reader found Bentley's foreleg and his walker straddled atop a hole burned into his bathroom floor, and Bentley's ashes on the floor of the basement below.

Most sources cite something like 300 such cases of the First Kind, and they all follow this same basic pattern. A person, usually elderly, often overweight, frequently mobility challenged, is found burned almost completely to ashes, bones and all. Their surroundings show scorching but are usually not burned.

Now, in recent years, a pretty good theory has been publicized that adequately explains all (or most) reliably documented cases of the First Kind, and that's the wick effect, of which a candle is the most familiar example. The flame on a candle's wick is small, but its temperature is very hot; thus it has a powerful melting effect within its tiny sphere of influence. This melts the wax into liquid, which is drawn up the wick, where it vaporizes and burns. The wick itself does not burn due to the cooling effect of the vaporization; but once the wax is gone, the wick burns away as well.

The application of the wick effect to human corpses is not supposition, but proven fact. In 2001, the Journal of Forensic Sciences published an account of a test performed at the State of California's Bureau of Forensic Services in which a pig carcass was wrapped in a blanket and provided with a source of ignition. After a number of hours, the smoldering fire was extinguished and it was discovered that the part of the pig that had burned so far, bones and all, had been reduced to ash. The experiment was repeated on the BBC television program QED. The body burns very slowly, with only a tiny flame or even no visible flame at all; and like a candle, the heat is so localized that very little else in the vicinity is affected by it.

In 1991, a pair of hikers in Oregon came across the body of a murdered woman in which a wick effect fire was still taking place. The middle portion of her body had been completely burned away, including the pelvis and spine, while the slow-burning smoldering fire was still taking place in both legs and the upper torso. The killer was later captured and confessed to having lit the corpse on fire using lighter fluid. Like the other victims, the woman was overweight, with a high fat content that is believed to have provided ideal conditions for the wick effect to take place. Mary Reeser and John Bentley were both overweight, had been wearing flammable clothes (exceptionally flammable in Mary Reeser's case), and both were smoking at the time of death. The condition of both corpses and the rooms in which they were found was perfectly consistent with what we'd expect to find if the wick effect had occurred.

However, not everyone accepts the wick effect explanation. Author Larry Arnold is among its most vocal critics. His 1995 book Ablaze! The Mysterious Fires of Spontaneous Human Combustion asserts that the cases mentioned above, and many others, have no natural explanation. Arnold wrote two other books as well, The Reiki Handbook about energy healing, and a report on what he believed were the psychic causes of the Three Mile Island nuclear accident. His explanation for SHC follows the same type of fringe reasoning. Arnold proposes that a particle that he called a "pyrotron" strikes the victim's body and ignites it from within. His pyrotron is unknown to science, of unknown origin, is undetectable, and has no describable properties, except that it seems to have something to do with kundalini yoga. Though the publication of his book has made Larry Arnold something of a go-to guy expert on SHC, his explanation is clearly unacceptable from any reasonable scientific perspective. Science does not allow simply making up a subatomic particle and calling that a mechanism for anything.

Other researchers have proposed various other explanations, including methane, which is one byproduct of bacterial action in the gut. And, as every flatulent college student with a match knows, it's flammable. One problem with the methane hypothesis is that cows produce even more gas than humans, and if it were true, we'd expect spontaneous cow combustion to be common. But we don't have any reports of this. One explanation for the discrepancy is that cows do not participate in the triggering activities. Cows don't wear a lot of flammable rayon acetate nightgowns, and their nighttime place of rest is a rarely an overstuffed chair beside a crackling fireplace. They spend much less time smoking than humans, as their hooves lack the manual dexterity needed to operate a cigarette lighter (explaining the lack of Saturday night barnyard hilarity).

Spontaneous Human Combustion of the Second Kind is when the event is witnessed and we have accounts of what took place. These accounts are quite different than those of the First Kind. Slow, smoldering fires are never the case; they are always a large sudden ignition with active flames. When the victims survive, the burns (which can be serious) are on the skin, never the deep, complete reduction to ashes seen in the First Kind.

• In London in 1982, Jeannie Saffin, a severely mentally handicapped elderly woman, was sitting at a table with family when her upper torso suddenly caught on fire. They extinguished the flames and paramedics took her to a burn unit, where she died eight days later of lung damage from inhaling the fire.
  
  
• In 1938 in Chelmsford, 22-year-old Phyllis Newcombe's dress suddenly caught on fire as she was going downstairs at a dance. Other revelers extinguished the flames but she, too, died at the hospital from her burns.

People catching on fire is not especially uncommon. It happens all the time. The only thing differentiating the cases classified as SHC is that no source of ignition was found; the fires are said to have been spontaneous. Other than that, there's nothing especially remarkable about them. The fires burned in a familiar manner, and the injuries are what would be expected. But these cases of the Second Kind are also rare; probably more rare than the First Kind. The reason is that these are unsolved, whereas the First Kind cases are generally solved, at least to the satisfaction of the investigators. These two cases of the Second Kind are famous only because there was no source of ignition found. No cigarettes, open flames, or sparks were found near either Jeannie Saffin or Phyllis Newcombe; but it's not scientifically permissible to conclude that their combustions were spontaneous. Maybe they were; but just because we didn't find the cause hardly means that there wasn't one.

Structure fires or brush fires sometimes go unsolved as well, but I think you'll have a hard time finding a fire inspector who will invent the term "spontaneous structure combustion" as if the lack of a determined cause means there wasn't one. Spontaneous Human Combustion of the Second Kind should not be allowed to exist as a category; instead we should call them what they are: Unsolved deaths by fire. Similarly, SHC of the First Kind has never been found to be spontaneous either. Those are simply the rare cases where a natural death in isolation has been followed by a slow combustion from some nearby source of ignition.

The wick effect is an interesting tidbit of science, albeit somewhat gruesome. I find that the logical pitfall of calling either type of SHC "spontaneous", and instead recognizing why they're not, is even more interesting.

By Brian Dunning
Follow @BrianDunning