A Death in Belmont (12 page)

Cohen finished in a blaze of indignation: “Smith is here on the basis of murder, rape, and robbery. Are you going to take his life? You have it within your hands tomorrow afternoon. Are you going to take his life on the basis of the people at the country club? Are you going to take his life on the basis of the children who ‘attended school' with the district attorney? Are you going to take his life on the basis of evidence that was evolved in a police station? Have courage! I know you will!”

This was a jury's dilemma at its most terrible: to decide between two plausible but utterly contradictory propositions. Either Roy Smith killed Bessie Goldberg but was too apathetic to conceal it—failing even to get rid of her address in his pocket—or he didn't kill Bessie Goldberg and was the victim of bad luck and the distortions of the Middlesex DA's office. Furthermore, in some criminal cases it is possible to be found guilty of some crimes and not others, but this case was all or nothing. Either Smith killed, raped, and robbed Bessie Goldberg or did none of those things; logically Smith could not be innocent of one of those crimes and not of the others. The man who killed Bessie Goldberg also raped her; if the rape was not plausible, then neither was the murder. And neither crime would have happened without the robbery, which was described by Richard Kelley as the motivation for the murder in the first place.

Even Beryl Cohen would have to admit that Smith made an
excellent suspect in the murder of Bessie Goldberg, but that wasn't the question before the jury. The question was whether Roy Smith
killed
Bessie Goldberg, not whether he
could
have killed her. In a civil case—in which a man's life or liberty is not at stake—the jury uses something called a “preponderance of evidence” to decide whether something is true. The jury only has to decide if something is more likely than
not
likely to be true; the chances have to be at least slightly better than fifty-fifty. By that standard a jury could easily find Roy Smith guilty of killing Bessie Goldberg.

In a criminal trial, however, a jury cannot send someone to prison—or, in Smith's case, to the electric chair—unless he is guilty “beyond a reasonable doubt.” It is a phrase that almost everyone thinks they understand but no one can explain. The classic definition was penned Chief Justice Lemuel Shaw, who incorporated it into his instructions on circumstantial evidence in the 1850
Webster
case. “What is reasonable doubt?” Shaw asked. “It is not a mere possible doubt; because everything relating to human affairs, and depending on moral evidence, is open to some possible or imaginary doubt. It is not sufficient to establish a probability, though a strong one, that the fact charged is more likely to be true than the contrary. The evidence must establish the truth of the fact to a reasonable and moral certainty.”

Trying to define “reasonable doubt” is like trying to define a color; you end up having to explain the very words you use in your definition. If a jury doesn't intuitively understand the phrase “reasonable doubt,” they're not going to be helped much by the term “moral certainty.” Over the years, courts have struggled to clarify, in some senses, the obvious: “It is not a fanciful doubt, it is not an imagined doubt, and it is not a doubt a juror might conjure up to avoid performing an unpleasant duty,” Black's Law Dictionary
offered. Nor is it “free-floating existential doubt,” a legal scholar added. One court finally acknowledged that there may be no better definition of the phrase “than the phrase itself.”

The central problem with the idea of reasonable doubt—with the entire premise of man-made justice—is that it tries to determine an objective truth with subjective tools. It will never be known in any absolute sense whether Roy Smith killed Bessie Goldberg; that is simply a given, the starting point of any determination of Smith's guilt. Lacking the ability to “see men as God does,” as one legal scholar put it, the courts must settle for some lesser version of the truth. That lesser version is known as “legal” guilt. It is possible to be actually innocent—to have truly not killed Bessie Goldberg—and to still be found legally guilty of her murder. Legal guilt simply means that twelve unbiased people heard the evidence against you and decided that you killed Bessie Goldberg beyond any reasonable doubt. Maybe you did
not
kill her but—strictly speaking—that is not a concern of the court. The concern of the court is that you receive justice, and justice is defined as having received a fair trial. Fair trials, in turn, have been described in the Constitution; everything else is in the hands of the jury.

The disadvantage of this system is that innocent people can be found guilty and put to death. Between 1973 and 2000, more than one hundred people have been released from death row—over 3 percent of the current death-row population—because they were later proved to be innocent. As one defense attorney noted, the Federal Aviation Administration would never tolerate an airline that lost one plane out of a hundred; why should the justice system? DNA evidence, which was not available until 1989, was responsible for thirteen of the exonerations. With an error rate as low as one in a trillion, DNA evidence is virtually all-powerful in court and can
win convictions with almost no supporting evidence. If it can send people to the electric chair, though, it can also save them from it, and that is a power that state prosecutors are loath give up. One case that wound up before the Missouri Supreme Court involved a man who was due to be executed despite the chance that a court-ordered DNA test would prove he was innocent. The assistant district attorney in the case argued that the new evidence should not be considered by the court. The presiding judge, Michael Wolff, was incredulous.

“To make sure we're clear on this,” Judge Wolff asked, “if we find that DNA evidence absolutely excludes somebody as the murderer, then we must execute them anyway?”

The answer, according to the assistant DA, was yes. A jury had determined that the man was legally guilty, and if the state began to question jury decisions, the whole system would fall apart. You don't endanger the ship to save one drowning sailor, in other words. Ultimately the gap between legal guilt and actual guilt is the gap between human perception and objective truth, and that is a gap that has ruined lives—and tormented philosophers—for millennia. In a famous parable the Greek philosopher Plato described a cave where bound prisoners are held in complete darkness except for a campfire behind them that projects shadows on the wall. The shadows were made by passing objects in front of the fire. Everything these men knew about reality was based on the shadows; the objects themselves remained unknowable and maybe even unimaginable.
That
is a modern jury. They are allowed to see evidence from the crime, but they can never turn their heads to see the crime itself. They must come to a conclusion based only on the evidence—the shadows—that they are allowed to see.

“This is a strange picture,” the Socratic student Glaucon com
mented after hearing about the cave for the first time. “And strange prisoners.”

At 11:25 a.m. on November 23, 1963, twelve such strange prisoners retired to a jury room in the Middlesex Superior Courthouse to decide whether Roy Smith had, in fact, killed Bessie Goldberg.

T
WENTY MILES TO
the north in the failing mill town of Lawrence, Massachusetts, the phone was ringing in the apartment of a twenty-three-year-old woman named Joann Graff. Graff lived alone in a one-room apartment and worked in an industrial design shop; on Sundays she taught classes at a local church. Graff answered the phone and spoke for a few moments to a friend named Mrs. Johnson, who invited her to dinner that night with some other families from their congregation. Graff favored plain dresses and wire-rim glasses, and a last-minute invitation to a Saturday-night church dinner was about as spontaneous as she ever got. She said yes and hung up the phone.

One hour later—at 12:30 in the afternoon—Graff's landlord knocked on the door. Graff opened up and gave him the fifteen dollars' rent and then closed the door behind her. In the time that the door was open, though, the landlord noticed that the breakfast dishes were done and that a religious book lay open on the kitchen table. At some point in the next three hours but probably toward the end of
that period, a second person knocked on Joann Graff's door. She must have let the person in, because Graff's neighbors heard nothing through the thin rooming-house walls. Whoever the intruder was, he knew what he was doing. He forced Graff diagonally across her own bed without the neighbors hearing, and he twisted two stockings and a black leotard around her neck without the neighbors hearing, and finally he stripped her and raped her and killed her without her neighbors hearing anything at all. Then he ransacked the apartment—though he left money for the gas bill sitting untouched on the kitchen table—and closed the apartment door behind him when he left.

The police were able to pinpoint the time of the murder to within minutes. A Northeastern University student named Ken Rowe, who lived one floor above Graff, told police that at 3:25 that afternoon, a man in a brown jacket and green pants had knocked on his door and asked where Graff lived. Rowe said the man looked like he was in his late twenties and wore his hair carefully combed back with grease. Rowe directed the man downstairs and shut the door. At that point Graff was presumably still alive.

Five minutes later Mrs. Johnson picked up the phone and started dialing Joann Graff's number. It was exactly 3:30 in the afternoon. In Cambridge the jurors in the Roy Smith case were just closing in on an agreement about his guilt. In Boston a National Guard gunner was locking a 105 mm round into the breach of a howitzer and waiting for permission to fire. By the time the thud of the concussion had died out, Joann Graff's phone had stopped ringing, and another Boston woman was dead.

 

AT
4:50
THAT
same afternoon, the members of the Smith jury, led by foreman Jim Bird, filed into Judge Bolster's courtroom and
took their seats. Both alternate jurors had been dismissed from the case because they were no longer needed. Roy Smith sat unshackled in his locked defendant's box with Beryl Cohen near him at the defense table and his mother, Mollie Smith, and his twenty-two-year-old sister, Betty, behind him in the front row. (Cohen had bought their plane tickets and found them a place to stay with a friend of his for the duration of the trial.) Cohen had undoubtedly told Smith that if the jury returned a guilty verdict he would use President Kennedy's assassination as the basis of an appeal. Attorneys believe they can read a jury's verdict on their faces, and both Cohen and Kelley must have scrutinized the jurors as they walked in. Jurors who keep their eyes lowered and avoid looking at the defendant come bearing bad news, lawyers believe; it is no different in the courtroom than in life.

The clerk of the court asked the jury and the defendant to rise, and Beryl Cohen and Roy Smith and the twelve men of the jury got to their feet. Rain tapped against the tall windows. The court officer asked the jury whether they had reached a verdict, and the foreman said that they had.

What say you, Mr. Foreman? the court officer demanded. Do you find the defendant guilty or not guilty?

Guilty, foreman Bird responded.

Of what crimes? the clerk demanded.

Of murder in the first degree, Bird responded.

Bird told the court that the jury had also found the defendant, Roy Smith, guilty of the charge of larceny but innocent of the charge of rape. There is nothing in the law that insists a verdict must be logical, so Richard Kelley could not challenge the jury on the rape acquittal. Smith could have gone to the electric chair if they had found otherwise, so it may have been a compromise verdict that
reflected some doubt about his guilt. When a jury departs from the evidence and delivers a verdict that reflects some broader sense of justice, it is called “jury nullification.” Juries have used nullification to spare white men who were obviously guilty of lynching blacks, as well as to spare battered women who were obviously guilty of killing their husbands. Jury nullification, although treated warily by the courts, has traditionally been one way in which the “conscience of the community” can counteract the effect of an unpopular law.

Jury verdicts are not open to questioning by the court, so there was no way to find out why Smith was exonerated of the rape or how the jury resolved in their minds the logical problems that gave rise to. Foreman Bird did, however, make a point of recommending that Smith not be sentenced to death. After the verdict was read, Smith remained standing but was described by newspaper reporters in the courtroom as “visibly shaken.” Leah Goldberg thought he looked utterly impassive, as though he expected this and didn't much care. The court quickly moved to sentence, and then Smith was handcuffed by the bailiff and led past his mother and sister, both of whom were sobbing, and past the lawyers and court clerks and the rest of the spectators.

At 5:30 on the afternoon of November 23, 1963, Smith walked out of the building under a cold rain to begin serving a term of life imprisonment without the possibility of parole.

T
WELVE WOMEN WERE
strangled and beaten to death in their homes over the course of a year and a half in the Boston area, and only one of the murders—Bessie Goldberg—was considered solved. Logically the remaining murders were committed either by one skilled murderer or many lucky ones. During the early sixties three out of four Boston homicides resulted in an arrest, and roughly 60 percent of those arrests resulted in a conviction. After a year and a half and the biggest manhunt in Massachusetts history, the eleven remaining Boston stranglings had an arrest rate of zero and a conviction rate of zero. The numbers were way, way off the bell curve, particularly for a city that had been on high alert for the past year and a half. Murderers are often impulsive, psychologically unstable people who, among other things, are not particularly good at thinking ahead. The ones who are, like Ted Bundy, quickly achieve legendary status for their ability to kill virtually anywhere at any time. The rest—like Roy Smith—are usually arrested within days.

During the eighteen months of “Boston stranglings,” three other women were also strangled who did not quite fit the pattern. One, a sixteen-year-old black girl named Donna Saunders, was found manually strangled in a back alley in Boston. Another was a sixty-year-old “heavy drinker” named Margaret Davis who was found manually strangled in a cheap hotel room. The third was a thirty-seven-year-old black woman named Modeste Freeman who was found nude in someone's backyard, strangled to death with a sweater and beaten almost beyond recognition. None of these murders were considered classic “Boston Stranglings” because the victims were found outside their homes. By comparison, however, two of those three murders were solved within days: Saunders was killed by a high school boy who was furious because she wouldn't kiss him, and Davis was killed by a man she met in a bar and went with to the hotel.

A success rate of two out of three is typical for a homicide unit in an American city; one out of twelve is not. Statistics alone would argue that there was a single skilled killer out there, but the great diversity of victims—old, young, black, white—would argue against it. Much of what was known about the psychology of sexual predators pointed to at least two men, one who killed older women and one who killed younger ones. Part of the problem in establishing a pattern was that the murders spanned a patchwork of town, city, and county jurisdictions, and the various authorities involved were unable—and reluctant—to share information. Without a careful analysis of all the crime scenes, investigators could never hope to create a psychological profile of the killer or, conversely, demonstrate that it was the work of several men.

The inefficiency of the investigators was rewarded one final time January 4, when a young white woman named Mary Sullivan was
found strangled in her Beacon Hill apartment. Sullivan had graduated from high school on Cape Cod a year earlier and had moved to Boston to work as a secretary in a financial firm. Her two teenage roommates arrived home around six o'clock on a snowy evening to find their dead friend propped up in bed with a leotard and two scarves tied around her neck and a broom handle shoved into her vagina. A greeting card that had been placed by her left foot read “Happy New Year!” The scarves she had been strangled with were brightly colored and done up in huge bows, as if she were someone's birthday present, and her legs were spread apart to shock her discoverer as much as possible. Mary Sullivan was nineteen years old.

The public outrage had still not died down two weeks later when Ed Brooke, the attorney general of Massachusetts, announced that his office was taking over the investigation of the Boston Strangling cases. Brooke had managed, as a Republican in a staunchly Democratic state, to become the highest elected black official—and the only black attorney general—in the country. No longer would the murders fall under nine separate jurisdictions within the state, Brooke announced; they would now be handled by a new office called the Special Division of Crime Research and Detection. To the people of Boston the special division quickly became known as the Strangler Bureau.

The Strangler Bureau was headed by an independently wealthy Bostonian named John Bottomly, who had graduated from Deerfield Academy and Harvard and gone on to head the Eminent Domain Division of the Attorney General's Office. Bottomly had absolutely no experience in criminal investigations but was a ferociously methodical man who was not intimidated by the 37,500 pages of documentation that had been generated by the ongoing investigation. It was a stack of paper ten feet high that, somewhere,
held the secret of who was killing and raping women in Boston; Bottomly just had to be clever enough to see it.

Taking direction from Bottomly was Michael Cullinane, acting captain of state police detectives, Lt. John Donovan, head of Boston Homicide, Lt. Andrew Tuney of the Essex County DA's office, and Detectives Phil DiNatale and James Mellon of the Boston police. They were joined later in the spring of 1964 by a low-ranking metropolitan police officer named Steve Delaney who had taken an amateur interest in the strangling cases. Delaney had gotten hold of a spreadsheet of all the murders and spent his spare time playing detective with the cases, trying to see patterns. One day Cullinane asked Delaney whether he thought the stranglings were the work of one man or several, and Delaney answered without hesitation that they were the work of one man. Cullinane asked why. Well, Delaney said, the murders had happened in half a dozen different police jurisdictions, all of which had an excellent record for solving murders, and none of them had been solved. That, to Delaney, suggested the work of one very sick and skillful murderer. Attorney General Brooke got wind of this demonstration of logic and asked Cullinane why he had to hear it from a beat cop instead of a homicide detective. Cullinane's response was to give Delaney a spot on the Strangler Bureau.

The murder files were kept in a vault at the Eminent Domain Office in the State House, just off Boston Common. Photographs of the murder scenes and lists of evidence were pinned to a board that was kept facing away from the vault entrance so that people who stopped in would not have to see them. Investigators like Delaney could walk around to the other side and see, at a glance, the entire sweep of the city's tragedy. Delaney was given the job of sorting through the crime scene reports to determine if there was a pattern
to the killings. It was an overwhelming task: The Eminent Domain Office was filled with scores of cardboard boxes, and each box was packed with files that no one had gone through yet. Within weeks, though, Delaney came to the conclusion that there was an unmistakable pattern among the killings—one that pointed to a single killer. His opinion was echoed by a little-known FBI report that came out several years later and was based on a much-less-hurried analysis. “All of the murders and assaults happened at multiple dwelling apartments with brick fronts, of transient character where many people were constantly coming and going,” the report stated. “There were always two or three exits to these apartment houses. Ninety percent of these apartment houses were sorely in need of repairs in one way or another. The victim was positioned so that, upon entering the room from any door, the body would be in full view. The sexual organs were noticed and the legs spread apart.”

Without exception the women had also died in their own homes after apparently letting their killer in voluntarily and putting up little, if any, struggle. It was as if the killer or killers were able to gain entry under some pretext and then attack the women from behind, when they could offer no defense. The attacks were quiet and deadly and probably over in minutes. Some women were sexually assaulted with objects and others were raped and still others bore injuries to their breasts, but none was grossly mutilated. The point of these crimes was sexual humiliation, and murder was an unavoidable result of that process. This was not a disorganized killer who was obeying voices in his head; this was a completely functional killer who walked out onto the street afterward and had no problem blending in with the people around him. This was a man who could be married, have a job, raise children—a man who, ultimately, was outwardly indistinguishable from the detectives who were trying to catch him.

One of the more distracting theories was that one of the killers might be a homosexual, which led detectives to throw a blanket of scrutiny over clubs that catered to that clientele. According to Delaney, police officers infiltrated known homosexual bars, hoping to hear something about the murders, and wrote down the license plate numbers of cars parked in the area. One particularly obsessed investigator gathered so much data on homosexual nightlife that he single-handedly began to slow down the investigation. He was finally issued a cease-and-desist order. Another distracting theory was that the murders couldn't possibly be the work of one man. A single uncatchable man was far more embarrassing to the police than a crime wave, and as a result there was reluctance on the part of police investigators—particularly Lt. John Donovan—to see links between the crimes.

Donovan, however, had his own theories on the apparent pattern in the killings. The victims were always women, he said, not because there was a lone, woman-obsessed killer out there, but simply because women are more likely to be the target of sexual attacks. The women were killed in their homes because that is where they were most likely to be found alone. There were no screams because the women were being strangled, and they were strangled with stockings, leotards, and scarves because those things could easily be found on the woman who was being attacked.

Police did have one thing going for them, and that was a possible description of at least one if the killers. Joann Graff's upstairs neighbor told police that the man who knocked on his door shortly before the murder was in his late twenties, of medium build, and wearing green pants and a dark jacket. That almost exactly described the man who had knocked on Marcella Lulka's door shortly before the Sophie Clark murder a year earlier; the only dif
ference was that Lulka had remembered the man's hair as “honey-colored” rather than dark. Neither of those people had actually been attacked, but their descriptions also matched that of the suspect in a sexual attack in Boston that had all the signs of a failed Boston Strangling. Around noon on February 18, 1963, a German-born waitress named Erika Wilsing got up to answer a knock on her apartment door and was confronted by a dark-haired man in his midthirties who said he was there to check her apartment for leaks. The man was wearing a waist-length jacket and green work pants and was of medium height and a muscular build. Wilsing let the man in and turned her back for a moment; the next thing she knew, he had her down on the floor with an arm across her throat. Wilsing was a big, strong woman, and she managed to lock her jaws on one of his fingers. The attacker loosened his grip long enough for her to scream, which got the attention of some workmen on a neighboring roof. Moments later her attacker was gone.

A police illustrator drew a sketch of the killer based on those descriptions, but it didn't even lead to a suspect, much less to an arrest. It was a measure of the bureau's desperation (or incompetence, depending on whom you asked) that they reached way out beyond the bounds of ordinary police work. A psychic named Paul Gordon had previously told police that a very disturbed young man named Arnold Wallace, who was suspected in his own mother's death, was also the Boston Strangler. Gordon was given an injection of sodium amytal, which induces a state similar to hypnosis, and was questioned extensively by several members of the Strangler Bureau. Gordon's testimony about the crime scenes was at times wildly off and at times strangely accurate. Among other things, he described a photograph of a ballerina hanging in Ida Irga's apartment that had never been mentioned in the press. One possibility
that the investigators considered was that Gordon himself was the murderer but didn't consciously realize it, and was only able to talk about it under a chemically induced hypnosis. The other possibility was that he was in fact psychic—whatever that meant—and that Arnold Wallace was indeed the killer, but investigators were never able to link him to the murders in any convincing way.

Within weeks of Brooke's announcement, the newly formed Strangler Bureau invited another, more famous psychic to Boston. Peter Hurkos was a Dutch Jew who said that he had become psychic very abruptly after injuring his head in a fall from a ladder. The accident had happened in the midst of World War II, and Hurkos claimed that his new “radar brain” helped him evade the Nazis after they overran Holland. “I see pictures, like on TV,” Hurkos explained. “I cannot help them, they come and go in my brain. They show me what I'm looking for.” After the war was over Hurkos applied his strange powers to solving murders and missing-persons cases and eventually wound up in Los Angeles to work on a movie about psychics. Hurkos was an enthusiastic self-promoter who had already put one innocent man in custody for a multiple murder committed by someone else. Three weeks after Mary Sullivan's murder, he flew into Providence, Rhode Island, and was driven by John Bottomly to a hotel outside Boston, where he checked in under a false name.

Like Gordon, Hurkos was inexplicably correct on several details of the crimes known only to the police. The focus of his suspicions was a lonely door-to-door shoe salesman named Daniel Moran who had already caught the attention of the police for writing odd, threatening letters to women. Moran was unquestionably disturbed; he tried to join a monastery but was refused, he supposedly showered with his shoes on, and he had once told his personal physician he was worried that he was out killing women during periodic black
outs. He confessed to nothing when police detained him, though, and in due time his alibis proved to be largely unassailable.