The Annals of Unsolved Crime

by Edward Jay Epstein

  CHAPTER 18

  THE PURSUIT OF DR. SAM SHEPPARD

  The brutal murder of Marilyn Sheppard in her bedroom in 1956, and the arrest of her husband, Dr. Samuel Sheppard, captivated the public’s attention: Did a stranger break into the Sheppard home and batter his wife to death, as Dr. Sheppard claimed, or did Sheppard himself commit the murder, as the prosecution claimed? The trial so stirred the popular imagination that it inspired one of television’s longest-running weekly series, The Fugitive, as well as movie of the same name.

  A husband killing his own wife is a familiar crime to police. When a wife is murdered at home, suspicion commonly falls on the husband. If the husband has no verifiable alibi, and there are no witnesses, police tend to classify the crime as a possible domestic homicide. Nor is such a scenario a difficult case for prosecutors to make in court. A husband living at home has both opportunity and a presumed motive to “terminate the marriage.” The problem for investigators is eliminating the possibility that an intruder entered the home and committed the murder.

  The Marilyn Sheppard murder occurred on the Fourth of July holiday weekend in the lakefront suburb of Bayview, Ohio, just outside of Cleveland. At 6:00 a.m. on July 4, 1956, police were summoned to the home of Dr. Samuel Holmes Sheppard, one of Cleveland’s most prominent orthopedic doctors. When they arrived, they found Sheppard’s wife Marilyn dead in the upstairs bedroom. Her half-naked body had been beaten so savagely and her face slashed so many times that she was barely recognizable. The entire room, including the walls, floor, and closet doors, had been splattered with blood. No murder weapon was found at the scene, but downstairs police found signs of an apparent robbery. Dr. Sheppard’s medical bag had been overturned with its contents strewn on the floor. Dr. Sheppard’s sports trophies were smashed. Desk drawers had been pulled open. Yet, according to Dr. Sheppard, nothing was missing. So police considered it a possibility that the robbery had been faked to cover up the murder.

  The time of Marilyn’s death was established as approximately 4:30 a.m. Asked to account for his movements, Sheppard said that after a dinner party that night he fell asleep on a day bed in the downstairs den. He was awakened by his wife’s scream and, rushing upstairs, saw an unknown person. When he got to her bedroom, he was knocked out from behind. When he regained consciousness, still woozy, he took Marilyn’s pulse and found that she was dead. At that moment, he said that he heard someone leaving by the back door. He ran after him along the Lake Erie shore and caught up with a “bushy-haired” person. But he was then knocked out again.

  Police almost immediately doubted his story because there were no signs of forced entry. Instead, they assumed it was a domestic homicide. Since they believed that Dr. Sheppard was the only other person in the house, they made little more than a perfunctory effort to recover fingerprints, blood samples, hair, and fiber evidence. If it were a domestic homicide, collecting such evidence would serve no purpose other than to prove that Sam and Marilyn Sheppard shared the home. The police at the scene were so convinced that this was a domestic homicide that detectives dispatched to the hospital to question Sheppard were instructed to get a full confession. One detective, playing the “bad cop,” told him: “I don’t know about my partner, but I think you killed your wife.” Despite such tactics, Sheppard steadfastly insisted he was innocent and was himself a victim of the same assailant who murdered his wife.

  Nevertheless, on July 29, 1954, Sheppard was arrested for the murder. The trial, which began in October 1954, became a media circus. Few trials in American history had generated as much lurid coverage. The prosecution spent weeks bringing out the erotic details of Sheppard’s three-year tryst with Susan Hayes, a nurse at his hospital, to establish that he had already betrayed his wife. His defense lawyers focused on gory crime-scene evidence. They produced expert witnesses who testified that Marilyn’s broken teeth indicated that she had bitten her attacker in the struggle. But Sheppard had no open wounds. The defense also produced a report by one of America’s most eminent neurosurgeons, Dr. Charles Elkin, who examined Sheppard after the murder and found that he had suffered a cervical concussion and nerve injury, that he had weak nerve reflexes that were impossible for him to fake, and that his injuries could not have been self-inflicted. Two other witnesses found by his private investigators testified that they had seen a bushy-haired man near the Sheppard home on the day of the crime. Sheppard, taking the stand in his own defense, described his fight with a “bushy-haired intruder.” The jury did not believe Sheppard’s story and, on December 21, 1954, brought in a verdict of second-degree murder, and he was later sentenced to life imprisonment.

  Sheppard spent nearly ten years in prison before the U.S. Supreme Court overturned his conviction on the grounds he was denied due process. When he was retried in 1966, he was quickly acquitted.

  As a free man, he never returned to medicine. Instead he wrote a book, wrestled professionally under the name “The Killer,” became an alcoholic, and died of liver disease in 1970, at the age of forty-six.

  The mystery did not die with him. His son reopened the case in 1998 on the basis of new DNA evidence, which did not exist at the time of the crime. The new DNA tests showed traces of blood on Sheppard’s clothes that was neither his nor his wife’s. The same blood was also found on a closet door only a few feet away from where Marilyn was murdered.

  There are three principal theories of the murder. The police theory is that Dr. Sheppard murdered his wife, staged a fake robbery, invented the bushy-haired man, and inflicted his injuries on himself. Next, there is the handyman theory, in which Richard George Eberling, who worked in the Sheppard house, is alleged to have murdered Marilyn Sheppard. Eberling had his own house-cleaning company, which serviced the Sheppard home. This gave him access to the home on the night of the murder. He also had reportedly stolen items from some of the houses he cleaned, which might have provided a motive, since a ring belonging to Marilyn was found in his possession. He indeed admitted entering the Sheppard home on the day of the murder, saying he had cut himself and was seeking bandages. In 1984, he was convicted (along with another man) of murdering Ethel May Durkin in Lakeville, Ohio, so he was capable of homicide. DNA tests on him proved inconclusive, and he died in prison in 1998. Finally, there is the “neighbor theory.” Its main proponent is F. Lee Bailey, who was Sheppard’s attorney in his retrial. Bailey contends that there is evidence that Marilyn Sheppard was having an affair with her neighbor, and that the neighbor’s jealous wife might have killed her.

  My assessment is that the new DNA evidence supports Dr. Sheppard’s story that there was a third person in the house. If blood was left during the attack, it means that some unidentified person was in the bedroom or had access to the home. If the bushy-haired intruder was capable of knocking Sheppard out twice, it was unlikely to be the neighbor’s wife. Since there were no signs of forced entry, the intruder almost surely had easy access to the house. In my view, the person who had these qualifications, Richard Eberling, was the killer.

  The lesson here is the powerful role played in police investigation by a phenomenon called “confirmation bias.” In social psychology, it is defined as the tendency of people to favor information that confirms their beliefs over information that undercuts their beliefs. In the Sheppard case, the hypothesis of the police was that the husband had killed his wife. The investigation then focused on inconsistencies in his story that tended to confirm their hypothesis, while rejecting contrary evidence, such as that of an intruder. The consequence was a miscarriage of justice.

  CHAPTER 19

  THE KILLING OF JONBENET RAMSEY

  In 1996, the reported kidnapping of JonBenet Ramsey, a six-year-old star in the world of child beauty pageants, set off a monthlong media feeding frenzy reminiscent of the 1932 Lindbergh kidnapping. In both cases, a high-profile child was taken from its bed while the parents were at home, and was then found dead. And in both cases, police could find no signs of forced entry, identifiable fingerprints, or credib
le witnesses to the putative intruder except for a handwritten ransom note. With Lindbergh, who was a national hero, the police focused on the intruder, but in the Ramsey case, the police focused on the family.

  The JonBenet Ramsey investigation began on the morning of December 26, 1996, after her father, John Ramsey, reported her missing from her home in Boulder, Colorado. He told police that the last time the child star was seen by anyone in the family was when he carried her to bed at 10:00 on Christmas night. He showed them a handwritten ransom note that he had found in the house that said that JonBenet had been abducted by a “group” representing a “foreign faction.” It demanded that $118,000, the exact size of John Ramsey’s annual bonus, be delivered to the kidnappers. At 1:05 p.m. that day, before the ransom money could be paid, Ramsey found his daughter’s body covered in a white blanket in the wine cellar.

  The medical examination established that her wrists had been tied above her head, and her mouth covered by duct tape, and that she had been garroted by a nylon cord. From the advanced state of rigor mortis, the time of death was between 10:00 p.m. on December 25 and 6:00 a.m. on December 26. The autopsy determined that she was killed by either strangulation or a skull-fracturing blow to the head, and that there were indications that she had been sexually assaulted.

  The only solid clue for the investigators was the ransom note. Forensic experts found that the three sheets of paper used in it, as well as the pen with which it was written, came from a table near the kitchen in the Ramsey home. This meant that someone inside the house had taken the time to write a lengthy letter before or after the strangling of JonBenet.

  Police found a footprint made by a hiking boot in dust and a palm print on the door of the wine cellar that could have come from an outsider, but they could not date them to the night of the kidnapping. They also found a pubic hair in the blanket in which JonBenet was wrapped that could not be matched to any family member, but it also could have been left in the blanket at an earlier time or resulted from the accidental contamination of the crime scene, which was not initially sealed off. So, even with modern DNA tests, there was no certain evidence of an intruder.

  The investigators were also unable to find an escape route. There was an opened basement window, but there were no footprints in the snow outside the window. So the investigation homed in on the activities of the three family members who were in the house—John Ramsey; JonBenet’s mother, Patsy; and JonBenet’s brother, Burke. Despite an intensive effort, however, the police were unable to match the handwriting samples of any family member to the ransom note, or to find any other evidence implicating them.

  Meanwhile, the family hired lawyers to protect their interests and file lawsuits against the media outlets that were reporting police “leaks.” So the investigation ground to a halt.

  It took nearly twelve years for the district attorney’s office to officially exonerate the family members on the basis of the DNA. Even though the investigation officially resumed, the Boulder police chief observed, “Some cases never get solved.” There were a number of false confessions, such as that of John Mark Karr in 2006, but none of these confessors matched the DNA profile established by the FBI.

  The theories fall into either the domestic-violence category or the unknown-intruder category. The former theories raise the suspicion, which is common when a murder occurs in a household and there are no witnesses, that the ransom note was fabricated to cover the involvement of a Ramsey family member in the death of JonBenet. It was alleged that Patsy Ramsey may have altered her handwriting to avoid it being matched to the note, but this is hardly evidence. The intruder theories posited that some outsider who knew the layout of the Ramsey house, possibly a neighbor, business associate, or relative, murdered JonBenet as an act of revenge or anger against her parents and wrote the ransom note to divert police attention. Finally, there is the sexual-predator theory. Because JonBenet appeared in beauty pageants, there is the possibility that a sexual predator stalked her, cased the house, broke in somehow, and assaulted her.

  My assessment is that the abduction and murder were committed by an outsider. As the district attorney explained: “The match of male DNA on two separate items of clothing worn by the victim at the time of the murder makes it clear to us that an unknown male handled these items.” This lack of a match convinces me that someone broke into the house. It turns out that there were thirty-eight registered sex offenders within a two-mile radius of the Ramsey home, and, with the attention that JonBenet received as a child star, it is likely that the unknown DNA came from a sexual predator. If he was a known sex offender, he may well have left the note not to collect a ransom but to mislead the police investigation.

  DNA is a double-edged sword. It can prove the innocence of an outsider, as it did in the case of the bogus confessor John Mark Karr, but its presence cannot serve as evidence for insiders, such as the Ramsey family members, whose DNA would be expected to be found in the house. While DNA analysis provides a substantial advance over fingerprints in identifying individuals at a crime scene, it still requires a positive match to a person who is not expected to be at the scene of the crime.

  CHAPTER 20

  THE ZODIAC

  A series of attacks on young couples in Northern California in the late 1960s terrified the public after a serial killer, called “The Zodiac,” taunted newspapers with coded letters, a signature symbol of a circle bisected by one horizontal and one vertical line, and bloody artifacts from his crimes. Even though the murders themselves stopped in 1970, Zodiac letters extended the journalistic fascination with the case.

  The first two victims were high school students, Betty Lou Jensen and David Faraday. Both were shot to death on a secluded lover’s lane in the town of Benicia at about 11:00 p.m. on December 20, 1968. Police found no motive, witnesses, or clues.

  A similar incident occurred on July 4, 1969, in the parking lot of a park in Vallejo, California. Michael Renault Mageau, nineteen, and Darlene Elizabeth Ferrin, twenty-two, were shot by an unknown man. Ferrin died, but Mageau survived.

  Up until August, these attacks had not been connected by police. But on August 1, 1969, letters were sent to the Vallejo Times-Herald, the San Francisco Chronicle, and the San Francisco Examiner that linked the attacks. Each letter contained one-third of a 408-symbol cryptogram, which, the writer claimed, revealed the identity of the killer. He not only challenged these newspapers to investigate his murders but warned that unless each paper printed his letter on their front page he would shoot “a dozen people over the weekend.” The Chronicle immediately published its third of the cryptogram, and there were no murders that weekend. But the press had now become deeply involved in the killer’s activities. Two amateur cryptographers, Donald and Bettye Harden, managed to crack the code on August 8, but it contained no names or clues, other than to say that the killer was collecting “slaves” for the afterlife.

  The next letter received by the San Francisco Examiner provided non-public details about the two previous attacks. It was signed “The Zodiac,” which, as with the “Jack the Ripper” case, provided the media with a vivid name for their headlines.

  The next killing came on September 27, 1969, near Lake Berryessa in Napa Valley. This time the killer wore a Zodiac costume, consisting of a black hood, clip-on sunglasses, and a bib with the same cross-hair Zodiac symbol that was in the letters. When he appeared in the park, he carried a gun in one hand and held pieces of a clothesline in the other. He then approached Bryan Hartnell and Cecilia Shepard and ordered Shepard to tie up Hartnell. He then stabbed them both repeatedly, leaving them in a pool of blood. He then drew his Zodiac symbol on Hartnell’s car door. Shepard died from her wounds, but Hartnell survived to tell the story.

  The final murder attributed to the Zodiac was in San Francisco on October 11, 1969. According to three teenage witnesses who watched the attack, a taxicab stopped and the passenger shot the driver, Paul Lee Stine, to death. The passenger then calmly cut off part of his victim’s
bloodstained shirt and fled the scene on foot.

  Police speculated that this murder was committed by the Zodiac merely to get bloody swatches from Stine’s shirt. He sent these swatches with his next round of letters to newspapers in which he demanded that a lawyer meet him. He named two of America’s most celebrated defense lawyers, F. Lee Bailey and Melvin Belli, and said that either would be acceptable to him and that they should await his call on the Jim Dunbar television show. Belli came on the show in front of a television audience in Daly City, California. Although Belli followed the Zodiac’s directions, the Zodiac did not show up. However, Belli himself later received in the mail a swatch from Stine’s shirt with a letter asking for his professional help. That was the last Belli heard from the Zodiac, and the last bloody swatch to appear. Letters of unknown provenance continued to be received by newspapers with coded messages, but none could be deciphered. The Zodiac, or the letter-writers claiming his identity, eventually took credit for no less than thirty-seven murders. This led to a new journalistic enterprise: mining the police cold case files for similar killings over the past decade. Although this pursuit continued for a decade, none of the unsolved cases could be tied by evidence to the Zodiac. The police were also stymied by their inability to match evidence at the four crime scenes, or to match the crime scenes to the letters. Even when DNA analysis became available in the 1990s, investigators were unable to match DNA found in the saliva on the stamps on the letters to any crime scene or suspect.

  The only suspect identified by a victim was Arthur Leigh Allen, a convicted child molester. At the time of the Zodiac murders, he was in his mid-thirties, lived with his parents, and was a gun collector. In 1991, Michael Mageau, the first survivor, identified Allen from a picture on his 1968 driver’s license as the man who shot him. Allen also worked only a few miles from where victim Darlene Ferrin was killed, and he wore a Zodiac-brand watch (with the same symbol found in the letters). However, Allen’s fingerprints did not match those taken from the fourth crime scene, Stine’s taxicab. Nor did Allen’s DNA match the DNA found on the envelopes of the letters, when DNA testing became available. And Sherwood Morrill, head of the Questioned Documents Section of California’s Criminal Identification and Investigation Bureau, could not match the handwriting in the Zodiac letters to Allen. So no charges were ever filed. With no other eyewitness identifications, and no murder weapon, the trail was cold.