Beating the Devil's Game: A History of Forensic Science and Criminal (31 page)

While the investigation continued, Jeffreys traveled to the FBI’s academy at Quantico to show them what was involved with the process, and at the age of thirty-six he became an international celebrity. Many people became interested in how they could cash in. But Jeffreys’ fame at this point was minor in comparison to what it was soon to become.

Back in Narborough, the police were determined to find the right perpetrator, so the men of Narborough and villages nearby within a certain age range were asked to voluntarily provide a blood sample. More than 4,500 men agreed to do it, most of whom were eliminated via conventional blood tests (since DNA analysis was expensive and time-consuming). But the real goal was to ferret out any man who would not willingly submit, because he might have something to hide. Yet after all this processing, the police were disappointed to realize that they had failed to identify the Footpath Murderer. Then in September 1987, they learned about a suspicious incident that had occurred over a month earlier.

A baker named Ian Kelly claimed that he’d provided his own sample to the police as a substitute for his friend and fellow baker, Colin Pitchfork. He’d received a falsified passport and had gotten away with it, so he bragged about it during lunch at a pub. A female manager overheard Kelly and passed this information along to the police. They knew that Pitchfork had been arrested numerous times for indecent exposure, so given this background and his attempt to circumvent their investigation, on September 19, they brought him in for interrogation. He admitted that he’d been out looking for girls to whom he could expose himself when he happened across the victims. There had been no witnesses, so he’d seized the opportunity to rape them. To ensure that he could not recant, the police sent Pitchfork’s blood for DNA testing, and the results proved that Pitchfork’s genetic profile was indistinguishable from that of both semen samples. He became the first person to be convicted of murder based on genetic fingerprinting. Pitchfork drew a life sentence, while Kelly received a suspended sentence for obstructing the investigation.

The Pitchfork case sparked headlines around the world, as well as a book by bestselling writer Joseph Wambaugh,
The Blooding
, inspiring a great deal of attention from the law enforcement community. It seemed that a potentially foolproof method was at hand for solving crimes involving biological evidence. The rush was on in many different places to apply DNA technology to more crimes. Lifecodes, located near Westchester, New York, became the first private lab in the United States to offer RFLP testing for criminal incidents.

Around this same time, crime investigation took several large steps in other directions. The FBI had developed a national computer database called the Violent Criminal Apprehension Program (VICAP), slating it to become the most comprehensive computerized database for linking and solving homicides nationwide. Police departments around the country were invited to record solved, unsolved, and attempted homicides; unidentified bodies in which the manner of death was suspect; and missing-persons cases involving suspected foul play. By 1985, the Bureau had also set up the National Center for the Analysis of Violent Crime (NCAVC).

State and local agencies built up automated fingerprint identification systems (AFIS) via computers, and the FBI expedited an exchange of information among law enforcement agencies by introducing a standard system of fingerprint classification that harmonized the transmission of information from one AFIS system to another. The program digitally encoded scanned prints into a mathematical algorithm based on their characteristics and the relationships among their features. Each image received a corresponding file of demographic data. In contrast to the days and weeks it once took a fingerprint analyst to complete, within seconds the computer could compare a set of prints against a half million others.

Nevertheless, for cases with biological evidence, the process of identification was about to evolve. After the initial success with DNA typing in England, Lifecodes quickly developed the technology in the States. In 1987, Florida’s assistant state attorney, Tim Berry, contacted Lifecodes’s forensic director, Michael Baird, about a rape case that was going to court. He wanted to know what DNA identification could do for him.

The case had begun in May 1986, when a man entered the Orlando, Florida, apartment of Nancy Hodge and raped her at knifepoint. She managed to see his face before he left, grabbing her purse on his way out. During the succeeding months, the man raped more women, taking care to conceal his identity, and on his way out he always took something that belonged to them. In the course of six months, investigators believed this man had raped more than twenty-three women, but he proved to be maddeningly elusive. Eventually, however, he made a mistake: He left behind two fingerprints on a window screen. When another woman reported him as a prowler, the police chased him for two miles. Once they had him in custody, they found that his prints matched those from the window screen. Finally, they had their man: Tommie Lee Andrews. In addition to the fingerprint evidence, Nancy Hodge, a victim who had seen him, also identified him.

Yet proving Andrews to be a serial rapist with a high number of victims was going to be much more difficult. Although Andrews’s blood group matched semen samples taken from several of the victims, blood group analysis was not precise and thus there was potential in each case for reasonable doubt. Berry hoped a DNA match might strengthen the prosecution’s position, as well as close numerous cases and send Andrews to prison for a much longer stint.

Blood samples from Andrews and semen samples from the rapist went to Lifecodes. Within two months, the results came back: The bar codes from the rape samples were too highly consistent for the semen to have originated with anyone other than Andrews. The odds were stacked against it.

Nevertheless, DNA testing had not yet been accepted into court and before it could be used, it had to go through a pretrial hearing. Ever since the
Frye
test in 1923, any new scientific technology introduced as testimony had to pass the test of acceptability within a relevant scientific community. That way the courts avoided admitting evidence based on whim or supposed science that was actually devoid of objectivity and rigorous controls. DNA analysis had to prove itself scientifically sound in method, theory, and interpretation, and be positively reviewed by peers.

The hearing was long and complex, but finally the judge admitted the DNA evidence. However, Berry made a misstep by stating impressive odds for the samples to have come from Andrews—one in ten billion—and he could not substantiate how these odds had been derived. He withdrew the figure, hoping to repair the damage, but the jury hung. Andrews went to trial for the second rape charge and this time he was convicted. Months later, the first rape charge was retried and the DNA evidence was brought in with more clarity and power. After that trial, Andrews’s prison sentence stretched from his initial twenty-two years for rape to 115 years for serial rape. He became the first person in United States history to be tried and convicted with DNA evidence.

From there, DNA gained increasing acceptance in the courts, although challenges were aimed at the way samples were interpreted, as well as at shoddy handling of specimen evidence. Without safeguards in place for proper scientific examination, the labs put prosecutors at a disadvantage, because defense attorneys were learning about the vulnerabilities in the system. Manhattan-based attorneys Barry Scheck and Peter Neufeld cofounded the DNA Task Force of the National Association of Criminal Defense Attorneys. Their goal was to debunk DNA typing in courts across the country, and failing that, to at least limit its application. They evaluated laboratories and evidence technicians, offering support to any attorney faced with this evidence in the courtroom. Because many different things can occur between the collection of a sample and the final interpretation, the courts were forced to review DNA testimony on a case-by-case basis. The FBI reported the first RFLP-processed case from its own lab in 1989, publishing guidelines to set standards for quality assurance.

That year, in
People
v.
Castro
, the technology was successfully challenged for the first time. On February 5, 1987, Vilma Ponce and her two-year-old daughter, Natasha, were stabbed to death in their home in a Bronx apartment building and a speck of blood was found on the watch of a neighbor named Joseph Castro, who was the building’s handyman. Ponce’s common-law husband, David Rivera, identified him as the man he’d seen leaving the building on the day of the murder, covered in blood. Right afterward, Rivera had discovered the door to his apartment unlocked and open, and the police arrived to find his pregnant wife stabbed fifty-eight times, while the child had received sixteen wounds.

Castro was the obvious suspect, especially after an acquaintance of Ponce’s reported that Castro had pestered Ponce with sexual advances, which she had spurned. He was also the worker who had installed the lock on her door, which proved to be defective. However, information turned up that Rivera had been violent with Ponce in the past, breaking her jaw, which complicated the case. Bronx prosecutor Rise Sugarman hoped that the blood from Castro’s watch would prove to be from one of the victims, because it would add weight against Castro. Lifecodes tested it, along with samples from the victims, and reported that it matched Vilma’s at three locations on three chromosomes.

However, during the course of a twelve-week hearing, the defense, aided by Neufeld and Scheck, pointed out that the lab had made a technical error, which invalidated the DNA results. The defense’s expert was Dr. Eric Lander, an MIT scientist, who helped the attorneys get a concession from Lifecodes that they did not in fact utilize mathematical standards for reporting the odds but instead had interpreted the results subjectively, via observation.

As the issues grew more complicated, scientists from both camps agreed to meet outside the legal arena to discuss the problems. Together they devised a two-page statement that admitted that the results in the Castro case were too poor to make a definitive analysis either way, and that most scientists would agree with them: The results should be considered inconclusive. One of the prosecutor’s experts even returned to the stand to qualify his original testimony.

The court decided that the testing could be used to show that the blood on the watch was not Castro’s (to exclude him), but it could
not
be used to claim that the blood matched one of the victims (to include her as the source). Given the room this decision left for reasonable doubt, the DA offered a deal. Castro pled guilty to both murders for a lesser sentence, which to some extent vindicated the Lifecodes analysis.

Yet thanks to cases like this, whose individual failings some viewed more generically as the failings of DNA technology, and to alleged statements in the
New York Times
in 1990 by prominent scientists against DNA testing, the courts backed up. Although four of the scientists named insisted they had made no such statements (their written objections were ignored), the courts grew more conservative, allowing the use of DNA to
exclude
suspects as the source of origin but not to make claims that the suspect’s DNA was a match to collected evidence. Prosecution experts had to work hard to prove that DNA analysis could perform as promised.

Improved methods over the next few months increased the testing accuracy, and the technology that could demonstrate the chance that a specific sample matched a specific person showed statistical odds so overwhelming that the courts gradually allowed it for stronger claims. In 1992, the National Research Council Committee of the National Academy of Sciences affirmed the use of DNA analysis but recommended tight regulations over collection and processing. In addition, they advised that the testing be done by objective parties with no stake in the outcome.

That same year, Connecticut prosecuted a murder case with DNA evidence alone—a first. Carla Almeida, a twenty-two-year-old masseuse, had turned up missing on April 18, 1988, and a client with whom she’d been booked, Tevfik Sivri, said that she’d come and gone. Criminalist Henry Lee went to Sivri’s home and found nothing out of the ordinary, except that the place appeared a little too clean as if in an effort to remove evidence. He searched the house extensively, then reached down to feel the carpet. It was damp. Using a blood reagent, he showed that the carpet had been soaked in blood at some point. When it was cut away, a large pool of blood was revealed on the bare floor. Lee used a calculation of his own devising to prove that the loss of this much blood meant the loss of life. The police then found a spot of Carla’s blood in Sivri’s car, proven with DNA analysis. Although the state had no body, they went to trial against Sivri. The judge approved the analysis as reliable, and the jury convicted Sivri.

However, in 1994, he got a new trial. But by that time, Carla’s remains had turned up. She’d been shot in the head and dumped near a tree farm. Sivri was again sentenced to life.

REVELATIONS

In July 1984, items of clothing from a missing nine-year-old girl, Dawn Hamilton, were found in a tree in the woods outside Fontana Village, Maryland. Then the police found the girl’s body lying facedown not far from the path, still warm to the touch, but clearly dead. She had been raped, strangled, violated with a stick, and bludgeoned with a rock. An anonymous call to the police sent them to investigate Kirk Bloodsworth, because the caller reported seeing Bloodsworth with the victim the day before she was murdered. Despite his insistence that he was innocent and did not know the girl, Bloodsworth was arrested. Other witnesses offered police enough detail for a sketch, and from it yet another witness identified Bloodsworth. All five appeared at his trial to state that he was the man they had seen with the victim, and a shoeprint found near the victim proved to be his size. For further circumstantial evidence, prosecutors said that Bloodsworth had told a friend that he had done something terrible that day that would adversely affect his relationship with his wife, and during the interrogation, he had mentioned a bloody rock. The case was stacked against him so it was no surprise when the jury convicted him. They also sentenced him to death in the gas chamber.