Chapter 8 A CASE OF IDENTITY

On February 11, 1909, the U.S. State Department approved the Crown’s extradition warrant. On the fourteenth, Detectives Pyper and Warnock escorted Slater, with his baggage, sealed by U.S. Customs, onto the steamer Columbia. Arriving in Scotland on the twenty-first, the ship sailed up the river Clyde. To avoid the throng, straining for a glimpse of the notorious suspect, that was anticipated at Glasgow, Slater was removed from the ship at Renfrew, some five miles away, to be taken the rest of the way by car. As he disembarked, in handcuffs, a member of the Columbia’s crew kicked him.

At Glasgow police headquarters, Slater’s baggage was unsealed. There, among his neatly folded, carefully packed clothing, was the little hammer—not much more than a tack hammer. To the police, the hammer appeared to have been washed, as did a fawn-colored waterproof coat, which bore dark stains. There was an array of hats, including two cloth caps, though nothing resembling the Donegal cap Mary Barrowman had invoked. Nor did police find the checked trousers, fawn spats, or brown boots that some witnesses had described the “watcher” as wearing as he eyed Miss Gilchrist’s house.

“In the fierce popular indignation which is excited by a sanguinary crime, there is a tendency, in which judges and juries share, to brush aside or treat as irrelevant those doubts the benefit of which is supposed to be one of the privileges of the accused,” Conan Doyle wrote in Strange Studies from Life, his nonfiction survey of three murders in nineteenth-century England. “Far wiser is the contention that it is better that ninety-nine guilty should escape than that one innocent man should suffer.”

The Glasgow police appeared to have no such scruples. The case against Slater was weak, and they knew it. The brooch clue—the spark that had ignited the manhunt—had long since fallen away. But they had settled on their man, and they would have him. As a result, the case would need to hinge almost entirely on witness identification. But even where there is no intent to deceive, eyewitness memory is a risky proposition. It is patchy, fungible, and highly susceptible to suggestion. Though the inherent unreliability of eyewitness testimony would not be demonstrated scientifically until the late twentieth century, it was already well known anecdotally in Edwardian Britain.

A decade before the Gilchrist murder, another wrongful conviction had driven the point very publicly home. In 1895, a London woman accused Adolf Beck, a down-at-the-heels Norwegian dandy, of having swindled her out of jewelry by posing as a nobleman. On arresting Beck, the police learned that a man playing just this confidence game had swindled almost two dozen women in recent years. Police convened a lineup, known in Britain as an identity parade. Many of the former victims identified Beck—the only man in the lineup with a mustache and distinguished gray hair—as the swindler.

Beck protested that he was a victim of mistaken identity: he had been in South America, he said, when the earlier crimes occurred. But damned by eyewitness testimony, he was convicted and sentenced to seven years’ penal servitude. Paroled in 1901, he was rearrested, tried, and convicted on a similar charge soon afterward. Only in 1904 did police discover the real culprit: a gray-haired Viennese man named Wilhelm Meyer. Meyer, who was living in England under an alias, superficially resembled Beck. He confessed, Beck was pardoned, and the case endured as a seminal cautionary tale.

“It is notorious,” Conan Doyle would write in 1912, “that nothing is more tricky than evidence of identification.” Cognizant of this fact, police and prosecutors in the case against Oscar Slater made certain that it would not fail.

On February 21, 1909, at Glasgow’s Central Police Station, Slater was displayed before witnesses in an identity parade. Standing alongside the dark-haired, olive-skinned suspect were eleven other men: nine pale pink Scottish plainclothes policemen and two pale pink Scottish railway officials. Not every witness could spot the man who had been the “watcher” outside Miss Gilchrist’s home, but those who made an identification chose Slater immediately.

“To expect a row of Glasgow constables and railwaymen to offer ‘cover’ to the identification of a German Jew, of unmistakable foreign appearance, was very much…like attempting to conceal a bull-dog among ladies’ poodles,” the journalist William Park wrote acidly years later. It also did not hurt that several witnesses had been shown Slater’s photograph before the lineup, common practice at the time.

On the twenty-second, Slater, accompanied by his solicitor, Ewing Speirs, was formally charged with Miss Gilchrist’s murder. He was remanded to Glasgow’s Duke Street Prison to await trial. “Slater impressed everyone with his coolness and courtesy,” Peter Hunt wrote. “He asked Mr. Speirs to thank the police for their kind treatment.” Soon afterward, Speirs told the newspapers: “The more I see of Slater the more convinced I am of his innocence. I do not say this, remember, as the agent of his defence. As man speaking to man as I have done with Slater, I cannot help feeling that some dreadful mistake has been made by someone. He is not at all the type of man who would associate with such a revolting crime.”

Before long, the trial venue was changed to Edinburgh, the capital, and Slater was transferred to Edinburgh’s Calton Jail. At the trial, scheduled to begin that spring, the Crown would be represented by James Hart, the procurator fiscal for Lanarkshire, the county that included Glasgow.*1 Hart, who pursued the Slater prosecution with uncommon zeal, would prove to be one of the great malign forces in the case. The defense team included Speirs and the barrister Alexander Logan McClure, who would handle the courtroom arguments.

On April 6, Slater was indicted; his hammer, his raincoat, and one of his hats were sent for testing to Dr. John Glaister of Glasgow University. Glaister, one of the leading forensic medical experts in Scotland, had led the autopsy of Miss Gilchrist’s body. His testimony at trial—including a catalogue of Miss Gilchrist’s devastating injuries and the assertion that Slater’s little hammer could have caused them all—almost certainly helped bring about Slater’s conviction.

THE QUESTION THAT VEXED Conan Doyle has persisted for a century: Why, when the police knew within a week that the brooch clue was false, did they pursue Slater anyway? There was a reason, and it resides in an unfortunate accident of history.

At the time of the Gilchrist murder, the identification of criminal suspects was at a crossroads. Standing at that pass, a detective on the trail of a criminal had two choices. There was the way forward, a nascent, rationalist twentieth-century science that would come to be called criminalistics. There was the way back, the murky nineteenth-century pseudoscience known as criminology, rooted in the work of Cesare Lombroso and his ilk. In taking the path of criminology, the Glasgow police doomed Slater. Then again, as was confirmed long afterward, framing him for the Gilchrist murder had been their objective from the very start.

The Victorian age has been called the Age of Identification, and the name is apt. The technology that had spawned the era’s cities also fostered mobility: railways and faster steamships let ordinary people cross borders with ease. The trouble was, they let criminals do likewise. And the cities themselves, in their seething anonymity, offered criminals safe havens in which identity became fluid: one had only to adopt an alias and dissolve into the crowd. As a result, the anxieties of the era focused on the need to identify criminals at a distance. But individual identification—finding the uniquely correct needle in a dense cosmopolitan haystack—is no mean feat, and for the Victorians, an urgent question was how to go about it.

The identification of any criminal suspect involves the reading of signs: at the crime scene, on the victim, or on the criminal himself. Today, the best-known way of doing this is through forensic sciences like ballistics, fingerprinting, serology, and toxicology. These are the reconstructive sciences of the postmodern age, letting investigators reestablish past events after the fact, sometimes long after, as DNA fingerprinting has done since its introduction in the 1980s.

But in the Victorian era, forensic science was in embryo: the very concept of a “crime scene” did not exist until the end of the nineteenth century. And forensic investigation as we now understand it—involving rigorous professional protocols, up-to-date scientific procedures, and state-of-the-art police laboratories—did not begin to come into its own until the 1930s and ’40s.

Yet the need for criminal identification is as old as mankind. Think of Cain, who committed the first recorded homicide and was marked ever after by God. How, then, did people identify criminals before the mid-twentieth century? The answer lies in the locus of the signifiers: if the signs used today are read mainly off the crime scene, those of the past were read directly off the criminal.

AN INVESTIGATOR IN PURSUIT of a suspect has three chances at identification. He can identify the suspect after the fact, through forensic analysis of the scene. He can identify him during the fact, via eyewitness testimony. Irrational as it sounds, he can also identify the suspect before the fact, in a preemptive strike meant to safeguard the community. Which technique an investigator employs depends partly on circumstance and partly on the technology available. It also depends—tellingly—on the era’s prevailing attitudes toward crime, criminals, and punishment.

In antiquity and long afterward, Western culture viewed crime as sin. In the interest of public safety, criminals, once known, had to be marked: think of Hester Prynne, and of the vivid capital letter chalked onto the coat of the murderous Peter Lorre in M, Fritz Lang’s 1931 thriller. During the Middle Ages, offenders were often given visible stigmata that signified the nature of the offense. “Branding and ear-boring, as a means of marking the deviant status of the criminal, had been statutory punishments in England from at least the late fourteenth century,” one historian has written, adding:

A labour statute of 1361 declared that fugitives were to be branded on the forehead with “F” for “falsity.” The Vagabonds Act of 1547…ordered that vagrants should be branded with a “V” on their breast. Ear-boring was introduced in 1572, when a statute was passed requiring all vagabonds to be “grievously whipped and burned through the gristle of the right ear with a hot iron.” By an Act of 1604, incorrigible rogues were to be “branded in the left shoulder with a hot burning iron of the breadth of an English shilling with a great Roman ‘R’ upon the iron.”

The marks offered threefold social control. Visible brands alerted members of the honest public. In principle, they also deterred people contemplating a life of crime. And in an era before widespread literacy and comprehensive penal records, they could be “read” by law officers as signs of prior conviction: criminal suspects were routinely strip-searched in pursuit of them.

In medieval England, an enfranchised system of street justice also dealt with undesirables. It was tied to the concept of the outlaw, a word denoting not a criminal per se but a person deemed to be outside the law’s protection. In legal cases of the era, a criminal defendant (or the subject of a civil action) who repeatedly failed to appear in court—and whom the authorities could not locate—could be declared an outlaw. Once a man had been “outlawed,” any citizen encountering him had the right to do with him as he pleased, including commit homicide.^*2 “Outlawry was the capital punishment of a rude age,” two twentieth-century historians have written. “To pursue the outlaw and knock him on the head as though he were a wild beast [was] the right and duty of every law-abiding man.”

But by the Enlightenment, the perception of crime and criminals had changed. Crime was now seen as a misguided ethical choice, which allowed for the possibility, after removal from society and sufficient contemplation, of rehabilitation.^*3 For citizens of the era, though, the essential question on encountering a stranger remained the same as ever: “Who are you, with whom I have to deal?,” as the philosopher Jeremy Bentham put it in the early 1800s.

Because Enlightenment sensibilities deemed the mutilation and mob justice of a prior age inhumane, the state now began keeping meticulous dossiers on criminals. The main function of these records was to identify recidivists. A check of the files in a police department or prison could reveal whether a suspect had been convicted before, much as the brands had done in earlier times. But the system had a fundamental flaw: it was utterly useless if a suspect had changed his name, and this fact bedeviled law officers for decades.

In the 1870s, Alphonse Bertillon, a civilian employee of the French police, sought a better way to identify repeat offenders. Exploiting the relatively new medium of photography, he created what we now call the mug shot: full-face and profile images of a convict, affixed to a card. To this card, he added copious data about the convict’s bodily dimensions. At a police station or prison, a newly arrived convict would be rigorously measured and the results compared against the sets of measurements already on file. A match, Bertillon argued, would prove identity even if the convict was using an alias.

The system, known as bertillonage, was widely adopted by police departments in Britain and the United States. But while it did identify some recidivists, it was unwieldy, requiring intensive training to administer. By definition, it did not work with juvenile offenders, who were likely to have grown between measurements. And its very nature meant that it could be used only after the fact, when the suspect was already well in hand.

Lombroso’s “scientific criminology” was designed to circumvent these problems. Anchored firmly in Victorian prejudice, it was a diagnostic approach that unapologetically trained the lens of the majority culture onto the Other. But even more than its underlying bias, the great hazard of Lombroso’s method was this: under the system, criminal identity was no longer read but was instead constructed.

BEFORE VICTORIAN TIMES, preemptive identification was simpler. When stranger met stranger, a set of well-known class signifiers—accent, attire, bearing, coiffure—reliably broadcast to each whether the other could be trusted or should be given a wide berth. No matter that these signifiers could not actually identify criminals: they successfully identified the Other, and that, for bourgeois citizens, was more than good enough. In the view of the seventeenth- and eighteenth-century upper crust, it was far better to have a warning system that overgenerated rather than undergenerated, and simply to throw all the lower-class babies out with the criminal bathwater.

But with the coming of modernity, time-honored signifiers began to blur. As foreigners thronged their cities, Victorians, who could unerringly tell Cockney from the Queen’s English, found their ear for dialect of little avail. Even more worrisome, almost anyone seeking upward mobility—or illicit gain—could manipulate the old signals, adopting particular accents or modes of dress to counterfeit class identity.

For the Victorian bourgeoisie, a modern identification system was needed, and if the old signifiers had broken down, they would simply invent new ones. It is here that Lombroso’s “scientific criminology” enters the fray. If public order depends on social control, its guiding principle went, then it is vital to be able to recognize whom one needs to control. And so Lombroso set out to produce a field guide to the common criminal.

Where criminalistics, rooted in real science, would focus on the crime scene, criminology focused on the criminal. Like many intellectual enterprises of the period, it was inspired by Darwinian theory, which coursed through the age like an electric charge. But in the hands of Lombroso and his fellows, criminology proved to be Darwinism of the darkest kind.

To these criminal anthropologists, criminality was inborn—an innate predisposition that no amount of reform could undo. What was required, they argued, was a way to identify habitual criminals (along with those men and women, guilty of no crime, who possessed inherited criminal tendencies) through a set of anatomical signifiers. These signs were broad enough that they could be read from afar, like the topmost line of an eye chart.

In his magnum opus, Criminal Man, published in Italian in 1876, Lombroso wrote of having spied the link between physiognomy and criminal character in the 1860s, when he performed an autopsy on a known malefactor. “This was not merely an idea, but a revelation,” he wrote. Whipping himself into a lather of gothic melodrama, he continued:

At the sight of that skull, I seemed to see all of a sudden, lighted up as a vast plain under a flaming sky, the problem of the nature of the criminal—an atavistic being who reproduces in his person the ferocious instincts of primitive humanity and the inferior animals. Thus were explained anatomically the enormous jaws, high cheek-bones, prominent superciliary arches, solitary lines in the palms, extreme size of the orbits, handle-shaped or sessile ears found in criminals, savages, and apes, insensibility to pain, extremely acute sight, tattooing, excessive idleness, love of orgies, and the irresistible craving for evil for its own sake, the desire not only to extinguish life in the victim, but to mutilate the corpse, tear its flesh, and drink its blood.

Lombroso’s work did not appear in English until 1911, but Victorian Britons knew it from secondary sources. Among them were the writings of Havelock Ellis, the English physician and eugenicist, who had helped popularize the term “criminology” in the 1890s. Even more influential for British criminology of the day was the work of Francis Galton, an ardent eugenicist and a cousin of Charles Darwin. Seeking to ensure the purity of the British gene pool, Galton experimented with composite photography: he superimposed criminals’ faces atop one another, producing what he hoped would be an image of the ur-criminal, with features common to the entire criminal class. Once identified, members of that class could be kept from breeding.

Galton’s work, like Lombroso’s, overtly married criminal anthropology to the eugenic program, a common coupling then. A fringe benefit of their systems was that once classifications of criminal features were drawn up, they could be extended to any unwanted group, be it Gypsies, Jews, or other immigrants. This—a social enterprise that the criminologist Paul Knepper would call “the racialization of crime”—the Victorians enthusiastically set out to enact.

Excluding immigrants was easy, as bans could be legislated. In the United States, the first major law restricting immigration, the Chinese Exclusion Act, was passed by Congress in 1882 and signed into law by President Chester A. Arthur. In Britain, Parliament passed the Aliens Act in 1905. The act denied entry to “undesirable immigrants,” a conveniently elastic term understood as code for Eastern European Jews. It is striking to note the conflation of foreignness with criminality, a contrivance used to justify identifying, marginalizing, and punishing the convenient Other. Today we call it “profiling.”

THE FIRST TRULY CRIMINALISTIC approach to identification began in the late 1800s with the work of Hans Gross. An Austrian jurist who was fascinated by the application of the new science to the solution of crime, he published his monumental work, Handbuch für Untersuchungsrichter, Polizeibeamte, Gendarmen, in 1893.

Gross’s work represented a significant advance over the dark anthropology of Galton and Lombroso.^*4 To their subjective, racialized approach, he brought the rigor of scientific method: instead of reading imagined signs off the criminal’s body, investigators would read them from the locus of the crime. His handbook ranged over such subjects as “What to Do at the Scene of Offence,” “Search for Hidden Objects,” “Construction and Use of Weapons,” “Reproduction of Footprints,” and “How to Register and Describe Traces of Blood.” But it did not see English translation for more than a decade, appearing only in 1906 under the title Criminal Investigation: A Practical Handbook for Magistrates, Police Officers and Lawyers.

In 1908, when detectives confronted the Gilchrist murder, the old criminological methods of Galton and Lombroso and the nascent criminalistic ones born of the scientific revolution existed side by side. This was their forensic watershed, and to their credit, they did try the new methods. But in the Gilchrist case these methods were still so primitive that they proved either immaterial or unworkable. As a result, Slater was consigned to the mercy of criminology, which neatly constructed guilt where none had existed before.

It is not clear whether Glasgow detectives knew Gross’s work, which had appeared in English just two years earlier. They were familiar with fingerprinting, introduced in Britain at the turn of the twentieth century.^*5 Dusting Miss Gilchrist’s flat for prints, they found a suspicious one, on the workbox in the spare bedroom. But the best fingerprint technology in the world is only as good as the database against which a print can be compared, and the department’s files, barely a decade old, yielded no matches.

With criminalistic techniques of little help in identifying Miss Gilchrist’s killer after the fact, police were left to fall back on the two alternative means of criminal identification. One was identification during the fact, by means of eyewitness testimony. This is where the spate of neighbors’ statements, methodically solicited, about the “watcher” outside the Gilchrist home came in. It is also where the testimony of Lambie and Barrowman, manipulated to damning effect, was allowed to do its work.

But more than anything, the police reverted to the most pernicious means of identification of all: criminology, or the fingering of the criminal before the fact. It was this method, so closely bound up with the racialization of crime, that ensured the identification, pursuit, and conviction of Oscar Slater. As it transpired, the Glasgow police had begun their identification of Slater well before Miss Gilchrist’s murder.

CRIMINOLOGY’S CHIEF FAILING IS that it is the bluntest of blunt instruments. Because it can’t work after the fact, it can’t identify individual culprits. It can only tag a targeted person as belonging to a particular class—ethnic, social, religious, and so on. But given the anxious preoccupations of the Victorian age, the method’s great failing was also its great strength. At a time when the salient question between strangers was no longer “Who are you?” but “To which group do you belong?” criminology worked brilliantly as a means of preemptive social control, training a spotlight on members of marginalized populations.

In a criminalistic investigation, detection precedes identification. By reading the “infinitely little” traces at the crime scene, the investigator homes in on the culprit’s identity. That is the logical order of things.

Victorian criminology reversed the process. Criminology sees only the big taxonomic picture—the foreigner, the gambler, the pauper, the Jew. This approach, an unsavory exercise of the diagnostic imagination, is the time-honored refuge of the bigot. By criminology’s hall-of-mirrors logic, detection now follows identification, a topsy-turvy arrangement that recalls the Queen’s biting line from Lewis Carroll’s 1865 classic, Alice’s Adventures in Wonderland: “Sentence first—verdict afterwards.”

With criminology as their primary tool, the Glasgow police knew that they could never use it to prove Slater’s guilt. But they could use it, masterfully, to construct his guilt. And thus, by the bourgeois imperatives of the day, the apprehension of Slater was a grand success, whether he had killed Miss Gilchrist or not. For if Oscar Slater was no murderer, then he was at the very least a convenient Other writ large.

To the police, the brooch clue was a singular stroke of luck, for it netted a man of the kind Edwardian Glasgow wanted off its streets anyway—one who, to borrow the words of the American defense lawyer Eleanor Jackson Piel, was “available and disposable.” That the clue foundered scarcely mattered, for Slater’s capture and conviction remained a fourfold coup: in one blow, the city would be rid of a foreigner, a Jew, a gambler, and a member (at least intermittently) of the lower classes. Slater might as well be hanged for a sheep as for a lamb, the prevailing sentiment seemed to run, and of course he very nearly was.

That Slater’s guilt had been rigged from the start was confirmed in 1927, when the Scottish journalist William Park interviewed him after his release from Peterhead. “For some time before the murder the police were watching his house to get him on a charge of immoral housekeeping,” Park wrote to Conan Doyle that year. He continued:

He saw Lieut. Douglas & other officers watching him and was quite well aware of it. Just close on his arrest he saw repeatedly his men in observation.

I find…a statement by Gordon Henderson, club master at the Sloper Club,^*6 that the police called there on Wednesday 23d Dec. asking for Oscar Slater. This was two days before McLean reported the pawnticket….

This gives us a new theory altogether. Slater was being watched for another crime & was rolled into the Gilchrist case as a handy sort of fellow to convict….As far back as 1911 Slater disclosed this fact of the police watching & letting him [get] away from Glasgow so as to roll him into “flight from justice.”…The police admitted they were at Slater’s house two hours or so before he departed and did not arrest him….

The further we go into this terrible business we see nothing but pure manufacture of a case: deliberately operating beforehand to make a prosecution.

In the end, then, it all came down to this: Oscar Slater arrived in Glasgow in the autumn of 1908. He was almost certainly known to the police from his previous stays. This time he was targeted on arrival and his movements observed. Then came the Gilchrist murder and, for the police, the happy coincidence of the brooch. That was all the pretext they needed to identify, pursue, and arrest Slater. When their case proved weak, police and prosecutors shored it up with dubious eyewitness statements, suborned perjury, withheld exculpatory evidence, and all the inflammatory illogic that the criminological method allows. At trial, the judge told the jury that Slater “has not the presumption of innocence in his favour…of the ordinary man”—branding him an outlaw in all but name.

The case, a capstone to a century “virtually hypnotized by class,” as the historian Peter Gay has written, turned out to be about class in both senses of the word: it centered not only on Slater’s threadbare background but also on the set of damning, classifying labels that the majority culture had long affixed to him. It would fall to Conan Doyle to bring to the case the criminalistic approach it badly needed. It was this approach—scientific, rationalist, exquisitely abductive—that would ultimately redeem Slater, one of the most convenient “convenient Others” of his age.