Hard Evidence

It's arson, bomb and booby trap week at one of the nation's toughest forensics schools.

Photographs by John B. Carnett

Burn patterns left by the accelerant tip students off to arson; forensic anthropologist Joanne Devlin in firefighting regalia.Photographs by John B. Carnett

A morning mist clings to the foothills as 15 crime scene investigators from across the country approach a shrapnel-pierced Pontiac Bonneville outside Knoxville, Tennessee. Minutes before, a fiery blast engulfed the car's passenger area, exploding the side windows and sending the crazed-glass windshield arcing, slo-mo, 50 feet through the air.

Scene commander Tom Sparks, a beefy lieutenant with the Hartselle, Alabama, police department, designates a sketch artist and photographer to record the vehicle exactly as found and assigns five more CSIs the task of dismantling it for evidence regarding the cause of the blast. Another seven line up with Sparks along one side of the scene perimeter. "Step," he bellows. They move one stride forward before bending to mark potential clues. Within minutes, the sodden ground blossoms with orange evidence flags.

Back at the Bonneville, Joy Smith, a tall, blonde evidence specialist from Modesto, California, peers into the red-clay hole where the passenger seat and floorboard had been. She searches for anything remotely bomb-like in the surrounding jumble of shredded wiring, metal and plastic. "Now I see the wisdom of spending time at Radio Shack," she says. "It all looks like car parts to me."

Thirty minutes into the investigation, crucial clues emerge. From inside the passenger door a team member pulls a chunk of galvanized steel with threads on one side and the raised imprint "1 1/4" on the other. "Looks like we've got a 1 1/4-inch pipe," he says. Others pull matching bits of steel shrapnel from the perforated headliner. A cobalt-blue sheen marks the shorn edges
of several pieces--the signature of a high-power explosive.

From the shredded driver's seat, Ohio crime technician Matt Dulaney digs out a curl of flattened metal and holds it to his nose. "Gunpowder," he says. "Doesn't a windup clock have a round spring?" The perimeter searchers, for their part, have found shreds of duct tape that, all agree, could have held a pipe bomb and detonator together. The CSIs confront a grizzled former Marine turned explosive ordnance expert. "Very good," he says, nodding. "I put a pound of C4 in the pipe, used a clock timer, and shoved it all under the passenger-side seat."

Time to move on: Knoxville bomb-squad commander Van J. Bubel has other surprises in store this morning. He directs the group's attention to a shoe bomb laced to the elegantly turned foot of a fashion mannequin standing across the field.

"She's just like that guy on the airplane," says Bubel, a detonator cord in hand, "only smarter."

Ka-boom.

So goes a typical day in Week 7--arson, bombs and booby traps--at the National Forensic Academy, a joint project of the National Institute of Justice, the University of Tennessee, Oak Ridge National Labs and a host of state and local law enforcement agencies. Students are now more than halfway through a 10-week course that includes units on postmortem fingerprinting, blood spatter, skeletal scatter, grave detection, cybercrime and weapons of mass destruction, and wraps with students resolving a gauntlet of mock crime scenes under the demanding eye of an FBI evidence recovery team.

Now in its second year, the National Forensic Academy aims to establish high national investigative standards for a field that sorely lacks them. The truth is, unlike the highly specialized lab scientists on TV's CSI, most U.S. crime scene investigators come from the rank and file of local police departments and sheriffs' offices. Their training varies as widely as the budgets of their municipalities. The result: Countless cases get dropped when lack of expertise results in missed clues and spoiled evidence; other cases get shredded in court la O.J. Simpson, when defense attorneys attack less-than-perfect crime-scene procedures.

Not that the academy's cadets are greenhorns. Virtually all the men and women in this, the school's fifth session, have already attended a half-dozen or more courses in evidence collection and served years on the CSI beat. Modesto evidence specialist Smith, to name one, arrived with more than 450 hours of training and eight years of field experience under her belt. "But you can't compare sitting in a classroom listening to someone lecture out of a book to coming here and getting hands-on training from the best people in every field," she says.

Indeed, the academy has already earned an international reputation for the sharp realism of its training exercises and the unprecedented caliber of its faculty. One has only to consider the macabre list of school supplies: Each class works with a half-dozen human cadavers, two sets of skeletal remains and several pints of fresh blood. Academy coordinators Jarrett Hallcox and Nathan Lefebvre also scrounge up two cars and a couple of condemned houses for each class to blood-spatter, burn, and bomb.

The academy's hallmark style of extreme authenticity stems in large part from the University of Tennessee's world-renowned forensic anthropology department. Its outdoor anthropological research station--widely known as the Body Farm--is the only place where exercises in grave detection and body recovery involve actual (donated) human remains.

The value of working with "the real stuff" can't be overstated, says Hallcox. "Some of these investigators come from areas where body exhumations are once-in-a-lifetime events. But after our training, they'll be able to meet the challenge with genuine experience." There are limits, Hallcox admits. "We didn't use a real body for the shoe bomb," he explains apologetically during bomb week, "because Nathan and I would have gotten stuck with the cleanup."

Students also spend three days at the state medical examiner's office in Nashville, where they learn to fingerprint cadavers (a chunk of Silly Putty helps roll prints off decomposing fingers) along with trickier techniques such as lifting prints off thighs and buttocks. The latter can prove crucial to cracking homicides that include rape or physical struggle. But no students in the class had ever mastered the skill before Week 4, when Art Bohanan--inventor of the portable "superglue" fuming technology featured on CSI--taught them to warm the body part to around 70°F, fume it with cyanoacrylate (heated superglue), dust it with magnetic powder, and lift the clearly visible print with contact paper.

The academy's world-class faculty also includes renowned forensic anthropologist and Body Farm founder William Bass, who in Week 5 taught the class how to extract fingerprints from the sloughed off "glove" of skin sometimes found next
to a decomposed cadaver. (Soak the tissue overnight in a bucket of water, and slip your own hand inside the stocking-like glove.) Paulette Sutton, the widely published protegee
of Herbert MacDonnell (the father of bloodstain-pattern analysis in North America), runs blood-spatter week; and forensic biochemist and time-of-death expert Arpad Vass of Oak Ridge National Labs does triple duty with bloodborne pathogens (Week 1), human decomposition (Week 5) and weapons of mass destruction (Week 9).

Day two of burn and bomb week finds the National Forensic Academy's forensic anthropologist Joanne Devlin striding away from a kerosene-doused Chevy Citation she just torched. A burned-bone expert, Devlin teaches fire-fatality and arson investigation, with special emphasis on the recognition and recovery of charred skeletal remains. Her expertise lies in the interpretation of the burn-altered signs of bullet wounds and other trauma. She has also become the academy's designated arsonist, having spent the previous weekend burning down a four-room farmhouse for a Wednesday field practicum.

Today Devlin and her co-instructors intend to teach these CSIs to make preliminary, on-the-scene determinations concerning the possibility of arson--whether for profit (insurance scam), for revenge or to cover up murder. Such quick determinations often prove crucial, because getting results from a crime lab can take weeks to months, and clues missed at the start may be washed away or demolished in the interim.

Already, the class has watched Devlin, veteran fire investigator Mike Dalton and special agent Dennis Kennamer of the Bureau of Alcohol, Tobacco, Firearms and Explosives burn two furnished "burn cells" (mock rooms with one wall left open for viewing) to illustrate the aftermath of accelerant-fueled arson and the results of a lighted wastepaper basket strategically placed in the corner of a room.

The most obvious clues of criminal intent include gasoline trails or "splash and dash" burn marks on carpeting, furniture and walls. Window glass provides other clues: Long shards directly inside the windowsill point to a prior break-in, while small chunks of crazed glass suggest a heat-related shattering. Where did the fire start? Look up: "Lightbulbs have the obliging tendency to bubble and extend toward intense heat, as if to say, 'Look here, dummy,'" explains Dalton.

Earlier the same morning, the class looked without a flinch at Devlin's PowerPoint presentation from hell. The close-up photographs featured one fire fatality after another, each graphically illustrating the telltale signs that distinguish victims who perish during a fire from those already dead when the fire began--the latter being a red flag for possible homicide. A mask of soot around the nose and mouth, for instance, paints the picture of a fire victim still gasping for breath when engulfed. A face-down position suggests an attempt to crawl to safety or huddle from overhead smoke. A face-up victim raises more questions. On the other hand, Devlin warns her students against mistaking the drawn-up "pugilistic pose" of a severely burned corpse as a sign of struggle or self-defense. In fact, the pose results from the contraction of cooked muscle.

After a break to let the torched Chevy cool, Devlin herds students to the smoldering vehicle and cajoles them to reach into the char to gauge the fragility of the cremated remains. The student investigators have more difficulty with Devlin's mock human victims--three animal carcasses were burned in this car--than with the half-dozen cadavers they handled in previous weeks at the state morgue and Body Farm. They hang back. So Devlin pushes open the trunk and picks up a molar from the blackened remains of a raccoon. "What if this is the tooth you need to make a positive ID?" she asks, pinching it to dust between her thumb and forefinger. "Oops."

As students begin handling the charred bones, Devlin points out that though the hands, feet and facial characteristics have burned away, the underside is intact: There is not so much as a singed hair where the body rested against the trunk floor.

Devlin also wants her CSIs to experience the difficulty of distinguishing charred bones from other fire debris. Among her class exhibits she includes a dark version of "Where's Waldo?"--a trough of charred skeletal remains mixed with look-alike fire debris such as burned and crumbled ceiling tiles.

The previous weeks of fieldwork have already sharpened the students' powers of observation in ways they had not imagined possible. During a Week 5 daylong exercise in "surface scatter," the class divides into two teams, each assigned to recover a separate set of 30 skeletal fragments in different sections of the Body Farm. Instructors planted the weathered bones in the wooded enclave's thick underbrush, just as wild animals might scatter the remains of a homicide victim.

"The bones looked just like sticks and chunks of wood," says Baton Rouge crime tech Pammy Anderson. Nonetheless, Anderson's team found all but one of its scatter set while the other team found every bone plus an ulna (forearm) left by the previous class. It's a matter of utmost pride: The score of the two teams combined surpassed that of any previous session.

By Thursday night, the academy's fifth class is ready for some mindless entertainment, having doffed class uniforms (black boots, combat pants, and polo shirts emblazoned with a skull, gun and fingerprint) for jeans and sweats. By 9 p.m., most of the crew has settled, beer and pizza in hand, in front of the TV at one of the corporate apartments that serve as the school's upscale dorms. It's time for America's favorite prime-time drama, a show that some in the room love and some hate but all agree features a lot of "in your dreams" stuff: CSI. The show has also heightened the public's expectations of what CSIs can do and how fast they can do it.

For starters, several rush to point out, CSI's college-educated, city-roaming cast of characters would, in real life, belong to the ranks of don't-get-your-hands-dirty "lab rats" who work within the confines of state and regional crime laboratories. Some of these labs do, in fact, field mobile units to assist local police with the occasional scene investigation, "but most of the time, we're on our own," says Tim Horne, an investigator with the Orange County, North Carolina, sheriff's office. "We collect it, and 90 percent of the time, we process it ourselves." And in
a typical, medium-size law-enforcement agency such as Horne's, in-house processing means whatever the local investigators can pull off in the ad hoc evidence room.

As for the technology employed on CSI, this audience agrees that, for the most part, it's real, even if pricey, exaggerated and needlessly flashy. Hooting begins as they watch an audiovisual expert in the fictional Las Vegas crime lab zoom in for a close-up of a mole on the neck of an out-of-focus figure in a confiscated snuff film. As every investigator learns when dealing with security camera videos, you can't focus an already out-of-focus picture. (Digital sharpening can produce an image that looks more focused but at the cost of detail and accuracy.) Nor, investigators point out, can you get blood to fluoresce in broad daylight, something the fictional Warrick accomplishes after the next commercial break.

But the biggest beef this class has with Hollywood's glitzed-up version of their work is the speed with which the prime-time CSIs get their results. "They scan in a fingerprint and presto, up comes the name of a convicted felon," scoffs Houston crime technician Christopher Duncan. "I wish!"

In reality, fingerprint matching takes days to weeks using AFIS, or Automated Fingerprint Identification Systems, the computer database that searches for matches against the prints of persons arrested in a given state or region. Even then, the computer database spits out not one but an array of close matches, leaving it to the investigator to make the painstaking side-by-side print comparisons.

As for getting a match for a DNA sample lifted from a crime scene, try months to over a year, depending on the backlog of cases being run through such state and national databases as CODIS, or Combined DNA Index System. "All you can do is submit your evidence and take a number," says Horne. "Our case may be important, but so are those of every other agency in the state."

All agree that the show has wildly distorted crime victims'
expectations as to what investigators can or will do. "One lady demanded to know why I wasn't swabbing her windowsill for DNA," relates Mississippi detective Craig Burdett. "Even if I could get a sample, we're not going to run a $500 DNA test over a $50 stolen TV."

Whether they come from rural sheriffs' offices or big-city police departments, every one of these crime scene investigators knows the frustration of begging for funds to pay for outsourced tests such as DNA fingerprinting, as well as the basic chemicals and equipment needed for evidence processing. "Just because we know how to do it doesn't mean we'll get the materials," explains Horne. "So while we appreciate all the cutting-edge stuff we've been learning, the best is when they give us the Wal-Mart version."

Which explains why the class's hands-down favorite technique is an on-the-scene print-lifting method learned in
Week 3. It employs superglue and cigarette ashes in a jerry-rigged print-fuming chamber made from a Styrofoam cup. They also enthuse over recipes for fingerprint-lifting gels and strips cooked up using dollar-store items like glue sticks, glass cleaner and a dozen-odd types of duct, masking and adhesive tapes. "Our department's a lot more likely to let us buy a $3 stick of Elmer's blue glue than a $25 bag of chemicals," Horne says.

Eager to apply their new tricks, the students mull the convictions that might have been: "For years, I've been trying to get prints off the cheap sandwich bags our druggies stuff with marijuana and crack," says crime technician Steve Smith of Montgomery, Alabama. (Apparently, higher-grade Ziploc bags give up their secrets more easily.) But now Smith knows a correspondingly cheap trick that will bring out prints on the flimsiest of plastic. Using an ordinary aquarium as a fuming chamber, he will heat a few drops of superglue to create a cloud of whitish fumes that adhere to the print's amino acids. He'll then gently stretch his evidence across an embroidery hoop and spray the print (faint white from the superglue) with a fluorescing dye so that it pops up bright orange for a photograph clear enough to run through the AFIS database.

Others brood over the killers they might have put behind bars had they known then what they know now. Tim Carnahan of Burlington, Kentucky, describes a case in which a young woman was bludgeoned to death in her garage after a wild chase that started at the front door and wound throughout the house. "We had a good idea who did it," says Carnahan. "But we didn't know how to read the blood spatter to determine the weapon, or even the number of attackers. Next time will be different," he vows. (A suspect has since emerged and Carnahan plans to revisit the blood-spatter evidence.)

Already, alumni of the academy's inaugural year, 2001, have begun to make their mark. Back in Cocke County, Tennessee, detective Derrick Woods prepares for grand jury testimony with full confidence that he has a murder conviction all but in the bag. "I told the guy flat out that it couldn't have happened that way," he says of a shooting to which he responded a week after graduating from the academy in the summer of 2002. When Woods arrived on the scene--a disheveled mobile home--he found a corpse crumpled in front of a couch and a suspect. "The individual told me he'd pointed the gun at the victim just to scare him," Woods recalls. "He claimed that the victim jumped up and grabbed the gun," which went off accidentally during the ensuing struggle.

The shot was at close proximity all right, says Woods. "But there was no blood above the couch. It was on the side wall, and when I looked closely I saw that both the direction and depth of the spatter pointed down." Woods says his academy training told him that the victim had to have been shot at an angle from above. "When I confronted the individual with what I saw, he admitted I was correct."

For session four graduate Bobby Moore, a Lynchburg, Virginia, investigator, the puzzle pieces began falling together even before he left Knoxville. Moore describes a shooting that occurred 6 months before he left for the academy. Police found the victim, a middle-aged woman, shot in the head and sitting upright on the floor in a room barely heated by a wood-burning stove. Crime-lab tests on the gloves she wore came back positive for gunpowder residue, suggesting she'd been handling a gun, though no gun was found at the scene. Even more confusing, bleeding from her massive head wound had produced a strange pattern of staining: strips of blood-soaked clothing alternating with completely blood-free fabric.

"It was one of those cases that just didn't add up," Moore says. "When I came to the academy, I left behind a lot of uncertainty as to what happened and exactly where this woman had been when she was killed." By the time Moore got back, he says, "I could see the whole scene play out in front of me."

Moore applied his new understanding of gunshot residue and bloodstain pattern analysis to reconstruct how the victim, shot from the front at close range, had tumbled forward onto the wood-chip-littered floor, then raised her gloved hands to her face, smearing them with gunshot residue from her skin. Blood pouring from the wound soaked through her clothes, except where folds of fabric had crumpled together when she fell. Consistent with this scenario were the splinters and wood chips Moore had noticed in the victim's hair--a sign that at some point she had been on the unswept floor. "What I found really interesting," says Moore, "is that someone had then lifted her up off the floor to look at her." And in so doing, had unfolded the pleats of clothing that had remained clean. "Only someone who cared about the victim would have done that."

On his return from the academy, Moore went to the prosecutors who were considering pressing charges against the dead woman's boyfriend. "I could explain a lot of things to them," he says, "and we were able to line up all our evidence in a row." Faced with the overwhelming case against him, the boyfriend pleaded guilty to second-degree murder.

Such stories validate the academy's mission of raising the caliber of crime scene investigation in this country through effective training. Already, 66 graduates have returned to their communities not only to use what they have learned but to disseminate it to colleagues. Still, with classes kept small to maximize hands-on training, there's little hope of teaching even a single representative from each of the nation's approximately 18,000 local law enforcement agencies.

"We see ourselves as a model," says Hallcox, "and a possible avenue for setting national training standards in many aspects of crime scene investigation." The Department of Justice appears to agree, if its award of an additional $1 million in hard-won federal funding is any indicator. The money will subsidize police departments and sheriffs' offices that can't afford the $6,500 tuition, and provide seed money for the first research grants awarded by the academy's umbrella group, the National Forensic Science Institute at the University of Tennessee.

Not that real-life crime investigation will ever resemble the seductive wizardry that has turned blood-spatter analysis into prime-time entertainment. "In real life, it's down-on-your-hands-and-knees dirty business," says Anderson. "Ninety percent of the time, what we do is tedious," she adds. "But that other 10 percent makes it all worthwhile."

Jessica Snyder Sachs, author of Corpse_, writes the Crime Seen column for_ Popular Science.