I had promised to send pictures of the reconstruction to various members of the Patterson family as soon as they were available, and Joanna made it easy, by taking digital photos and e-mailing them to Jon, who is more computer savvy than I am. Jon, in turn, relayed them to three of Leoma’s children—Frankie Davis, Barbara Atkins, and Ronnie Patterson (who appears in Leoma’s youthful photo as a small and solemn-faced baby sitting on her lap)—and to two of her grandchildren, J.R. Roach (the son of Leoma’s oldest daughter) and Nancy Albert (the daughter of her youngest daughter).
Nancy was the first to weigh in. “I was shocked,” she said. “As soon as I saw it, I thought, ‘That looks like my grandmother.’ ”
J.R. Roach—at forty-four, the oldest of the grandchildren—was next. He was far less positive. “I don’t see a whole lot of resemblance,” he said.
Barbara was the daughter Jon and I thought the reconstruction most resembled—“dead ringer” (no pun intended) was the phrase Jon used. So I was startled when Barbara said flatly, “That’s not her.” A moment later her brother Ronnie got on the phone. “That don’t look anything like our mama,” he said.
The strong difference of opinion took me by surprise, though maybe it shouldn’t have. Memory is a tricky and unreliable thing, facial reconstructions are approximations at best, and resemblance is in the eye of the beholder. What’s more, there had always been disagreement within the family about the original identification: Frankie Davis and her daughter Nancy were convinced that a turquoise ring found with the remains in 1979 had belonged to Leoma (“I saw a picture of my grandmother wearing that ring,” Nancy said), but other descendants insisted that the only ring Leoma ever wore was one that her brother had fashioned for her from a silver dime.
Within the Anthropology Department, opinion was more consistent, though less passionate. Kate Spradley, the Ph.D. student who had taken the digital measurements of the skull, took one look and said, “That looks like the family members.” Lee Jantz, who runs the bone lab, studied the reconstruction closely, comparing it to our best photo of Leoma’s face. Lee—who knew about the GenQuest report—raised her eyebrows and said slowly, “That is very interesting!” Then, after a pause, she added, “But the DNA says it’s not her. So it’s not her. Can’t be.”
She had a point. The DNA test had been conclusive. Or had it? Slowly—faintly at first—alarm bells began to ring in the back of my mind. I dug out the GenQuest report and pored over it. “Based on the above data,” it read, “the profiles obtained for Pearl Smith and Frankie Davis have identical sequence variations. It is likely that Pearl Smith and Frankie Davis are from the same maternal lineage.” In other words, these two sisters had the same mother: not exactly a news flash. The bombshell was what followed: “In comparison to the bone sample,” the report went on, “Pearl Smith and Frankie Davis can be excluded as having the same maternal lineage.” There: that part sounded conclusive, and exclusive. The woman whose teeth and bones I had sampled was not the mother of Pearl and Frankie; was not, in other words, Leoma Patterson. It was that line in the report that had led us to stop referring to 05-01 as “Leoma” and to start calling her “Not-Leoma.”
But as I read and reread the report, other sentences slowly caught my eye, and my attention. “The sample tested was a piece of bone.” Why just a piece of bone, rather than either of the teeth? Then—in a small footnote to the data about the bone sample, I read this: “The presence of multiple probes prohibited the identification of a single mitochondrial DNA profile. Possible explanations include that the sample is degraded. It is not recommended to use this information for sole identification or comparison purposes.” Finally, at the bottom of the report, was this note: “Direct sequencing of the HVI and HVII regions is recommended to determine conclusive results.”
The more I studied the GenQuest report, the more contradictory it seemed. On the one hand, it seemed to say that the woman in the grave was not Leoma; on the other, it seemed to say that the conclusion wasn’t particularly conclusive, wasn’t necessarily reliable. DNA isn’t my field of expertise; hand me a shattered skull or a knife-marked rib, and I feel confident that I’ll see what’s forensically important. Talk to me about nucleotides or restriction enzymes or sequencing, and it soon starts sounding like Greek to me. I decided to consult an expert. Luckily, one was right at hand, unpacking her books and files.
At about the same time Joanna Hughes began gluing depth markers onto our baffling skull, Graciela Cabana, Ph.D., had arrived in Knoxville and begun settling into her new office beneath Neyland Stadium. Graciela was an important addition to the Anthropology Department: a molecular anthropologist, one specializing in DNA. I desperately hoped she could give me an objective reading and clear explanation of the GenQuest report.
We met in the office of Pam Poe, an administrative assistant in the Anthropology Department, because Graciela’s office didn’t yet have a desk, a phone, or even a chair. Jon and I were joined by Lee Jantz, who oversees the Body Farm’s donation program, in addition to the bone lab, as well as Lee’s husband, Richard Jantz (co-developer of ForDisc), head of the department’s Forensic Anthropology Center. This case was a head-scratcher, and Lee and Richard were getting caught up in it, too.
Graciela began slowly, cautiously. “I can’t quite tell,” she said, “what GenQuest did and how they did it.” She had studied the report itself; what’s more, she had called the lab and—without indicating exactly why she was calling—asked the lab’s chief scientist some questions about their procedures and protocols. The fact that she remained puzzled even after talking to the senior scientist? That didn’t sound good. “For one thing, I can’t tell which sample they used,” she explained. “It appears they used only one bone sample, which surprises me. With old material like this, teeth are almost always a better source of DNA than bones.” I had known this—the hard shell of enamel on the tooth encapsulates the DNA and protects it from degradation better than bone; that was why I’d included two teeth in the set of four samples. Next, Graciela focused on the lab’s inability to establish a single profile for the dead woman. As she went on, her caution gradually dropped away. “This report is not just inconclusive,” she finally said, to the astonishment of us all, “it’s completely worthless. The only thing you can tell from this report,” she concluded, “is that the sample was degraded or contaminated. Maybe both.”
The air in the office nearly crackled with electricity. Suddenly everything—every possibility—was up for grabs again. Originally, the woman in the grave had been known as Leoma; after the GenQuest analysis, we had taken to calling her Not-Leoma. Now, in the space of a few minutes, she had suddenly become Maybe-Leoma. I couldn’t help thinking of CSI again. On television, DNA testing provides an instant, ironclad answer; in this real-life case, though, DNA testing had merely—and hugely—muddied the water: It appeared to provide an answer, but in fact all it did was raise a whole new set of questions—not just for me, but for Leoma Patterson’s relatives. The family had mistrusted the original investigation’s findings, which is why they contacted me in the first place, but I couldn’t imagine that this new twist would do much to restore their confidence. And it certainly wasn’t likely to ease their pain.
Now what? The answer seemed both obvious and frustrating: Start over—get another DNA analysis, by a different and more careful laboratory. But where? Graciela herself was trained to work with challenging DNA samples, but she didn’t even have a desk or a chair yet, let alone a sophisticated laboratory. Luckily for us, one of her Ph.D. colleagues, Jason Eshleman, was also a molecular anthropologist—a particularly good one, she said. After earning his doctorate, Jason had co-founded a company called Trace Genetics, which quickly established a reputation for extracting DNA from the most challenging samples possible: bones thousands of years old. “If anybody can get DNA for you,” said Graciela, “it’s Jason.” I thanked her for the hel
p and vowed to contact him.
Meanwhile, there was one final technique I wanted to try using the resources available to me at UT. It would require dismantling Joanna’s beautiful reconstruction, and that seemed a shame. She’d spent weeks creating the sculpture, and what I wanted to do would destroy it in a matter of minutes. But there was nothing more that the reconstruction itself could tell us; besides, we had plenty of photos of it, taken from virtually every angle except the back of the head. Joanna began peeling off the clay with remarkable cheerfulness, and soon I found myself confronted once more by the bare skull. It looked smaller and more exposed somehow, now that I had seen it with a face. It also looked more cryptic, almost defiant. “Who am I?” the skull of 05-01 seemed to say. It was both a challenge and a taunt.
After the reconstruction was deconstructed, Joanna, Jon, and I left the bone lab with Maybe-Leoma, climbed up the long flight steps flanking Neyland Stadium, and crossed Stadium Drive into the basement of the Communications Building. There, Tom Owen, the head of the university’s television and photographic services, listened with interest as I described the twists and turns in the case and asked for his help. Immediately agreeing, he led us down the hall to UT’s television studio, where his staff sprang into action.
First we attached the skull to a stand that would allow us to raise, lower, tilt, and rotate it in any direction or plane. Then we set a music stand beside the skull, to hold an eight-by-ten-inch enlargement of the best photo we had of Leoma—the one showing a young, slim-faced mother holding a somber baby on her lap. The studio crew rolled two mammoth cameras into position, one to focus on the skull, the other on the photo. Images filled a bank of monitors in the studio’s control room. Some monitors showed just the skull, larger than life; others showed just the photo. But one monitor, the one I was most interested in, combined the two images, superimposing one atop the other: Leoma Patterson’s youthful face appeared strangely translucent, and a skull showed through the flesh, like some gruesome special-effects image in a horror film.
For over an hour we fiddled with the cameras, the photo, the skull: zooming in, zooming out, cocking the photo, rotating and angling the skull by fractions of millimeters. You know how portrait photographers always have you tuck your chin and tilt your head in a way that feels very unnatural and awkward? Well, the photographer who had taken this portrait of Leoma and her baby was no exception, and replicating that pose with the skull clamped to a wobbly, Rube-Goldberg-like assemblage of rods and knuckle clamps proved next to impossible. Finally, after dozens of adjustments and compromises—and with Joanna’s hand wrapped in black to conceal it—we decided we had matched the image’s sizes and angles as closely as humanly possible. The angles weren’t a perfect match; we’d have needed Leoma herself to recreate them exactly—that, or a camera with a lens identical to the one that had taken the photo—and we didn’t have either of those. As a result, the studio camera couldn’t reproduce the focal length and aspect ratio of the still photo with perfect accuracy. But it was damn close, and it would certainly show us if the basic shape and key landmarks of 05-01 aligned with Leoma’s features . . . or if the fit was so obviously bad that the skull would figuratively shout, “I’m not her”: if the cheekbones extended beyond the margins of the face, for instance, or the chin was far too short or way too long, or if the skull’s eye orbits showed through the top of Leoma’s forehead.
After we’d fiddled and fussed, and then fiddled some more, I finally asked the technician in the control booth to show us only Leoma’s photo, then—very slowly—to dissolve to the skull. As the bone began to show through, I told him to stop. As I studied the superimposed images, I began to feel the same tingle of excitement I had felt when I first saw Joanna’s facial reconstruction. Clearly visible through the flesh of Leoma Patterson’s strong, rounded chin was the center of the skull’s strong, rounded chin—the facial landmark named “pogonion.” The cheekbones and the jaw lined up remarkably well, too. Near the top of the face, the skull’s nasal bone ran directly beneath the midline of Leoma’s nose, and the cranial landmark known as “nasion”—the indentation where the nasal bone meets the forehead—aligned precisely with the dip between Leoma’s eyes.
Just weeks before, a DNA lab report had me convinced me that the skull was not Leoma Patterson’s. Now I felt myself doing another U-turn. “It’s her,” I said. Half a dozen heads in the television studio nodded in agreement. I walked out of the studio feeling elated. The superimposition was powerfully persuasive.
And yet . . . And yet . . . It was persuasive, but it wasn’t proof.
Our only hope for proof was to go back to the DNA. We’d have to try climbing the twisted ladder one more time. And by now I realized all too well how devilishly twisted that ladder could be.
Click here for photos from Chapter 2
Chapter Three
Twisted Ladder, Twisted Case
THE AIRPORT SECURITY screener peered closely at the two small objects in the envelope I had taken from my pocket. After a moment he called over another screener, and then another, and then a supervisor. Within minutes I was surrounded by armed Transportation Safety Administration guards who questioned me closely about the two human teeth I was carrying—my “carrion luggage,” to make a terrible pun. Luckily, I was on my home turf—Knoxville’s McGhee-Tyson Airport—and half the TSA guards there knew me. They weren’t questioning me because they were suspicious; they were questioning me because they were curious. After I had explained why I was flying to Texas with two teeth from a dead woman in my pocket, they waved me on and wished me luck in San Antonio, the next stop in my quest to determine, once and for all, whether 05-01 was Leoma Patterson after all. Like a runner in a relay race—a scientific race to find the truth—I was about to pass the baton to the next runner. In this case, the baton I was handing off consisted of the two teeth, and tucked inside those teeth, I hoped, was enough mitochondrial DNA to settle the matter.
Millions of people first learned of the forensic potential of DNA during the televised murder trial of O.J. Simpson. Blood from the scene where Nicole Simpson and Ronald Goldman were slaughtered—a sidewalk leading from Nicole’s condominium—matched O.J. Simpson’s blood type and DNA, a parade of prosecution witnesses testified. In addition, blood matching Nicole’s DNA was found on a pair of socks in O.J.’s bedroom, and blood matching Ron Goldman’s DNA was found in O.J.’s Bronco. The odds against anyone other than O.J., Nicole, and Goldman being the sources of those bits of blood-borne DNA were astronomical, experts explained, ranging as high as nearly one in ten billion.
The reason for such overwhelming mathematical probability, the jury learned during days of mind-numbing scientific testimony, is the immense size of the DNA molecule, coupled with the near-infinite ways in which its biochemical building blocks can be arranged. Its now-famous shape, called a double helix, resembles a ladder whose uprights have been twisted so they spiral or corkscrew around one another. The three billion “rungs” of this corkscrew ladder consist of pairs of chemical bases named adenine, guanine, cytosine, and thymine (abbreviated A, G, C, and T). For biochemical reasons I won’t go into (because I don’t begin to understand them!), every “rung” with an A on one side always—no exceptions—has a T on its other side; similarly, G is always paired with C. So if you were magically reduced to the size of an atom, and you began climbing this three-billion-rung ladder, the first five rungs you ascended might be constructed this way: AT, TA, CG, TA, GC. Think Legos—billions of Legos—spiraling miles into the sky.
Statisticians and evolutionary scientists have an old saying about monkeys and typewriters that’s meant to illuminate how random variation, given enough time, produces specific meaning. One variation of that saying goes like this: Put a hundred million monkeys at a hundred million computer keyboards for a hundred million years, and one of them will write the complete works of Willi
am Shakespeare. Well, I don’t actually believe that—I suspect it’s far more likely that they’ll bring the Internet and the entire computer-dependent world to its knees, and in far less than a hundred million years. But I can believe this, and can grasp it at a very rudimentary level: Stack up three billion pairs of ATs, TAs, CGs, and GCs, one atop another—or string three billion beads on a genetic necklace, or strike three billion two-note chords on a piano keyboard—and you’re going to get one-of-a-kind combinations. Combinations named O.J. Simpson. Nicole Simpson. Ronald Goldman. Bill Bass. Leoma Patterson. All it takes to match a forensic DNA sample to a known individual—and to match it with rock-solid certainty—is an uncontaminated forensic sample, an uncontaminated sample from the known individual, and impeccable laboratory procedures. In the O.J. trial, the defense didn’t challenge the scientific validity of DNA analysis, or of the matches, or of the statistical probabilities. Instead the defense argued—and the jury seemed to believe—that the incriminating samples were either planted by detectives or contaminated by careless lab technicians. So despite the staggering odds of guilt that the prosecution and its experts invoked—billions to one!—O.J. walked. The case hadn’t been made beyond a reasonable doubt.