Teasing Secrets from the Dead: My Investigations at America's Most Infamous Crime Scenes

by Emily Craig

  The answer lies in the unique hydrology of the “duck ponds” created between two barges in a tow. The motion of the barges creates eddy currents in these open spaces, sucking up the water and anything else on or slightly below the surface, down to a depth of eighteen inches. Luna's body had somehow gotten caught in a duck-pond eddy and been pulled upstream with the barges. Now, my fellow investigators and I are all going back to other Jane and John Does that have been found in the Ohio River, looking downstream as well as upstream for missing persons reports they might match.

  Of course, the toughest problem in identifying human remains is also the simplest: Where do you start? According to Kym Pasqualini, the founder and coordinator of the NMCO, the number of reported missing adults topped 43,000 in March 2003. When you add the number of missing children to the list, the total comes to a staggering 97,297.

  Those numbers are daunting enough when you've got a victim with an unusual biological profile or a special piece of surgical hardware whose serial number might somehow be traced. But what about the thousands of victims who basically resemble thousands of others?

  Such was the problem with the Jane Doe in Baraboo, Wisconsin, the one whose facial reconstruction I described in the prologue to this book. This was the young woman whose body parts had been butchered and flayed and carefully wrapped in grocery bags, which someone had then thrown into the Wisconsin River. Since months of searching for her by conventional means had failed, Sauk County Detective Joe Welsch and Wisconsin Special Agent Elizabeth Feagles had come to me, hoping I could do a facial reconstruction on her skull.

  By the time I became involved with the case, forensic scientists in Wisconsin had already done a complete analysis of the remains, determining that the young woman so brutally butchered had been a young Black female, about twenty to twenty-five years old, probably about 5'2'' and weighing 120-130 pounds. Wisconsin fingerprint expert Mike Riddle had even managed to lift prints from her decomposed hand-an almost superhuman feat that left me awestruck.

  But all this science hadn't gotten them very far. According to the NCIC database, more than 1,500 women who fit that profile had been reported missing since early that summer. Getting the prints was terrific-but where could they find a match? Most people who aren't criminals don't have their fingerprints on file. If Joe and Liz had had any idea where to look, they could have tried to lift prints from one of the young woman's possessions. But until they had some idea of who their victim was, they were stuck.

  Like so many other unidentified victims, the Baraboo Jane Doe was so frustratingly ordinary. Her teeth were perfect, with no restorations. She had no tattoos or scars and no evidence of previously broken bones. The D.A. hoped that her skull would hold some critical forensic evidence-some cut marks that might someday be matched to a weapon. That was why we were using the rapid prototyping technology to create a perfect replica of her skull. But so far the skull itself had yielded far too little information about this woman's identity.

  When Liz and Joe came to me, I was their last resort. They hoped desperately that my facial reconstruction would give them a visual image that they could circulate throughout the state. If all went as we intended-and we all knew that it might not-someone would see the image I created, recognize the victim, and come forward.

  So as I began my facial reconstruction that Labor Day weekend, I knew the stakes were frighteningly high. Until the police knew who Jane Doe was, they would never find her killer. If a serial killer was out there somewhere, we had given him virtual license to try again. If the killer were someone more ordinary-a boyfriend, spouse, relative, or friend-he might literally get away with murder, and a particularly brutal murder at that. There was one last chance to keep that from happening-and it was all up to me.

  I tried to keep the image of this woman's mutilated flesh out of my mind and concentrate on the skeletal details. Although it was unusual to be starting with a laminated paper skull rather than one made of human bone, everything else about this reconstruction was perfectly ordinary-just like the victim. As always, I began by cutting tissue markers-small sections of rubber that mark the depth of tissue in various parts of the victim's face. I make my markers from the standard pink erasers that go into mechanical pencils-long thin tubes of rubber that I buy at the office supply store and cut to size with an ordinary X-Acto knife.

  The length and positioning of these markers is based on standard anthropological formulas that tell me how deep the flesh is likely to be on a person's cheeks, forehead, chin, and elsewhere, based on his or her sex, race, and estimated weight. Carefully following these formulas, I glue close to two dozen markers at specific points all over the skull, in the middle of forehead, the bridge of the nose, the point of the chin, and other key places. Then I connect them with clay, using the bone structure as my guide.

  The most tedious part of the job comes right at the beginning. Cutting the markers to the right length and placing each one in its precise position is a painstaking task made all the more stressful by my awareness that the slightest mistake might compromise the accuracy of my final result. Some of those little rubber cylinders are no more than an eighth of an inch long, so as I worked on the Baraboo case that Labor Day weekend, I needed a sharp knife and a steady hand. Soon, however, I became absorbed in the soothing-if somewhat boring-mechanics of cutting the twenty-three markers, numbering each one of them with a sharp pencil, and laying them all out in numerical order. After about an hour, I was ready to go back to the skull.

  I'd already mounted the laminated prototype on a converted camera tripod, which I'd fitted with a big eyebolt that fit up inside the spinal cord opening known as the foramen magnum. My tripod has a large ball joint at its base, which allows me to rotate and tilt the skull until it is perfectly level, a position known as the Frankfurt horizontal. In this position, the eye sockets appear to be aimed straight ahead and I can draw an imaginary level line from the bottom of the eye orbit to the ear hole known as the external auditory meatus. I grabbed the small carpenter's level that I use for this task and centered it over the bottom of the prototype's eye orbits.

  Then I reached for the mandible, which the Milwaukee team had also made out of laminated paper, and fit it into sockets located just in front of the ears, the temporomandibular joint. I fiddled with the paper jaw until it fit perfectly, opening and closing in a smooth gliding motion so that the teeth of the upper and lower jaw fit together in normal occlusion. I didn't want my statue gritting her teeth-she'd be harder to recognize that way-so I put a small plastic strut betwen her upper and lower teeth for that tiny bit of separation that creates a more natural look. Then I adjusted the mandible until I had created a slight bit of distance between it and the skull, to mimic the normal separation created by the articular cartilage and a small fibrous disk called a meniscus. I knew that each tiny detail might make the difference between a face that someone might recognize and one that looked just slightly “off.”

  If my Jane Doe had had unusual teeth, they might have helped someone recognize her, so I would have had her bare those striking teeth in a smile-a complicated procedure that would have required still more manipulation of the jaw, since when a person smiles, the jaw drops and pulls back a little. Then, when I added the clay, I'd have had to make the statue's nostrils flare a bit, crunch up the flesh under her eyes, and flatten the flesh across her upper teeth to almost nothing-subtle but crucial touches that could make a huge difference in the final product.

  Luckily, I didn't have to do that here-this woman's mouth would be closed. The replica's teeth were perfectly shaped and placed, but they were coated with the same honey-brown resin that covered the rest of the skull, which would hardly give a natural look to the final result. Besides, there was nothing unusual about the woman's teeth, so once her mandible was seated correctly, I started gluing on the tissue markers.

  Simply out of habit, I always start at the forehead, dipping the eraser into some all-purpose glue and holding it in place for a few minutes
until the glue starts to set. It took me the rest of that afternoon to glue each marker onto the skull.

  As the sun was beginning to set I started on the eyes. Each artist has his or her own method, but I tend to do the eyes as soon as possible, mainly because I don't like to see those empty sockets staring at me hour after hour after hour. It's also easier to adjust the eyes before I put on the clay.

  Running my fingers lightly over the replica's eye sockets, I found the place where the palpable ligaments would be inserted-the tiny ligaments that anchor the corners of each eyelid. The insertion points are located by means of subtle bumps that I couldn't see, but that my fingertips found immediately. I marked each one with pen because that's where the corners of the eyelids would go, and I wanted to remain aware of that positioning through the rest of my work with the eyes.

  If I'd been using an actual skull, I'd have put some cotton into the eye sockets to protect the fragile bones for further forensic analysis. On a replica, that wasn't an issue, but I did need to keep the eyes from falling backward into the sockets. Folded-up Kleenex worked quite nicely.

  Then, using a small block of clay, I made a little pedestal for the first artificial eye, which I'd bought from a surgical supply house that makes eyes for people who need prosthetic implants. These false eyes look eerily realistic and come in all sorts of colors-for a Black woman, I had chosen the darkest brown available, with a slight yellow tinge to the surrounding “white” sclera. I stuck the eye onto its little clay pedestal and quickly mortared it into place with more strips of clay.

  Soon both eyes were in and I began adjusting their position. I wanted my sculpture to have a perfect gaze-each eye centered precisely in its orbit, protruding just the right distance in relation to the surrounding bone. The eyes should be level, too, and they should look together in the same direction. One of my tricks is to shine a single bright desk lamp into the eyes and look at the reflection. In a perfect gaze, the light is reflected exactly the same way in both eyes, so I spent half an hour adjusting first one eye, then the other. My reward was a steady, earnest gaze resembling that of a living person.

  I took a late dinner break and went back to apply the clay. Here was where my artistic intuition came into play. Although I am ultimately a scientist, I've learned over the years that simply following the mathematical formulas isn't enough. If my facial reconstruction is ever to come to life, I have to venture beyond the formula and allow my intuition to guide me to create all those individual little details that ultimately distinguish each face from every other. I have to make creative leaps-but leaps that are entirely supported by scientific data. It's this fusing of art and science that makes the difference between a scientifically correct but somehow vague face and a vivid, lively image that someone might actually recognize.

  Luckily, my intuition had lots of data to work with. When Joe had brought me the skull, he'd also handed over several photographs taken at autopsy and a copy of the autopsy report. The pathologist had found that this young woman was basically healthy, with an average amount of well-distributed subcutaneous fat. Since her genitalia were still present, he'd known she was a female. He'd estimated her age based on the youthful condition of her internal organs-mature, but showing no age-related changes in the heart, reproductive organs, or arteries.

  He had also determined that this woman was African American, based on the color of her very dark skin. Of course, skin color can undergo rapid and dramatic changes after death, but this woman had other Negroid features as well. The pathologist had mentioned her black, coarse, and extremely curly body hair. And despite the fact that her face was no longer visible-the flesh had literally been cut away from the bone-the anthropologic analysis of her skull told us that she had once had wide-set eyes; a well-rounded or “bossed” forehead; and a wide flat nose. Until I finished applying the clay, I would have said “African American,” too.

  But when my sculpture was done, something in the facial contours caught my eye. Somehow, the extreme flatness of the mid-face and the almost vertical shape of her front teeth and jaws made hers look different from the other African-American skulls I'd seen. Certainly, this woman wasn't White. But I couldn't quite believe she was a Black American, either.

  What other choices were there? I e-mailed my concerns to Dr. Leslie Eisenberg, a consultant to the pathologist who had done the autopsy and one of the foremost anthropologists in the country. For a time, Leslie considered my speculation that the woman might be one of the native Hmong tribespeople from the mountains of Vietnam who had relocated to Wisconsin as a result of the Vietnam War. The Hmong have dark skin, too, and relatively flat facial features, but unlike this woman's, their hair is usually straight.

  Maybe Indonesian, I suggested, and Leslie politely considered that possibility, too. Eventually, we both concluded that this woman was Black-but the unusual combination of features continued to bother me.

  At least my reconstruction was done. I took another look at her innocent young face and wondered if we would ever find out who she was-and who had killed her. But my work wasn't done yet. I wanted my finished product to look as much like a person as possible. A forensic sculpture will never be as accurate as a sculptured portrait or “bust” that has been made from a photograph or taken from life. Because I have only the shape of the skull to guide me, any reconstruction I create will be an approximation at best, a caricature at worst: an artificial face with enough similarities to the victim that it might trigger recognition in someone, but far from a perfect portrait. So I wanted to give all possible help to the man or woman who might see my work, to increase the chance that he or she would recognize our victim.

  I placed a long black wig on the sculpture and combed it carefully into place. Later, I'd try out two or three other wigs, taking photographs of each version. Meanwhile, I “dressed” my statue in a pink striped T-shirt, and for an extra touch of realism I dabbed fresh lipstick over the lips, just enough to add a little color and shine.

  I'd created the sculpture at home, but I arranged to meet Joe and Liz at my office before work that Tuesday morning, exhausted but excited after my three-day working weekend. Liz, I knew, had been skeptical from the beginning about Joe bringing in an outsider-a Kentucky artist to solve a Wisconsin case. Like most law enforcement investigators, she was a bit territorial, especially about such a high-profile case-and I later learned that she'd worked several other cases in which facial reconstructions had proved futile. Though she had little faith in this effort, Joe was eager but reserved, his eyes continually wandering to Liz to check out her reactions.

  “Come on back to my lab,” I said after the introductions were made. I pointed to my sculpture, sitting on the counter in the center of the room, and waited.

  When they saw it, the expression on their faces didn't change and neither of them said a word. Each of them glanced quickly at me and then back at the model. Joe reached out to touch the hair, and Liz gently placed her hand on his wrist to stop him.

  I was sure they hated it. Nervously, I broke the silence with a lengthy description of the digital photographs I had taken of my work, assuring them that the computer printouts minimized the little flaws and surface irregularities that showed up so glaringly in the clay. “You can see here-” I began, pointing to the pictures lined up on the counter beside the model. Liz shook her head and I fell silent again.

  “And I've got these other wigs-” I began once more, reaching for them. This time Liz held up one palm for silence.

  I searched desperately for my well-worn explanations of the limits of forensic sculpture. How it could never be portrait-quality-we just don't have the data for that. How, nevertheless, many people seem able to leap over the crude quality of a forensic image and jump to a flash of recognition, particularly when a loved one is involved. How often I had seen forensic images succeed-and, to be honest, how often I had seen them fail.

  But to my utter surprise, the two of them began to smile and then to laugh.

  “This is
amazing,” Joe said softly.

  “More than amazing,” Liz agreed. She turned to me. “I thought you'd do some kind of Gumby-like thing-I don't know, something that looked weird and unnatural. But this really looks like a human being. We might really find her with this.”

  Three months later, nurse-practitioner Shari Goss saw the four photos of my facial reconstruction posted on the bulletin board of her neighborhood grocery store and burst into tears. “I know her,” she told the astonished grocer. She had recognized Mwivano Mwambashi Kupaza, a young exchange student from Tanzania who had been living in Madison, Wisconsin, for the past three years. Kupaza was the twenty-five-year-old cousin of forty-year-old Peter Kupaza, Goss's ex-husband. After seeing the poster, Goss called the police in her rural Wisconsin hometown of Wesby to give them the young woman's name.

  Joe and Liz were then able to find a photograph of Mwivano, which resembled my reconstruction almost exactly. They went on to match the prints lifted from the remains they had found with fingerprints lifted from medical records that Mwivano had touched when she signed them. Finally, we had our positive ID.

  The story that Joe and Liz eventually put together was heartbreaking. They believed that Peter had raped Mwivano, who became pregnant and then had an abortion. About two years later, he allegedly killed her and dismembered her body in his home, packing it in plastic bags and carrying it to the river.

  Ironically, no one had ever filed a missing persons report on Mwivano Kupaza. Her friends and relatives in Tanzania believed she was still in the United States. Her U.S. community of friends and fellow students thought she had returned to Tanzania.