In 1848, in a small wood-frame house in Hydesville, New York, two sisters aged fifteen and eleven, Margaret and Kate Fox, announced that they had made contact with the spirit of a murdered peddler buried in the cellar. The spirit communicated with them by producing audible knocking sounds in response to questions, one knock for yes, two for no, or a knock in sequence for each letter of the alphabet. Within two years, the Fox sisters were touring American concert halls. Within ten years, something like three million people in the United States identified as spiritualists. Within forty years, spiritualism had become a serious cross-Atlantic movement, sponsored by eminent physicists, novelists, and members of the British aristocracy.
Margaret Fox confessed in 1888, to a large audience at the New York Academy of Music, that the raps had always been produced by her and her sister cracking the joints of their toes. She demonstrated the technique from the stage. The movement continued as if she had not spoken. She later recanted the confession. It did not matter. Spiritualism had, by then, become too large and too ramified to be contained by the exposure of its founding frauds.
What I want to explain in this piece is why the Victorians were, specifically, a ghost-haunted generation, and why the particular institutional expression of that haunting — the Society for Psychical Research, founded in London in 1882 — ended up being more interesting than either its defenders or its debunkers usually give it credit for.
The mourning economy
Death was extraordinarily visible in nineteenth-century life in a way it has not been since. The infant mortality rate in England in 1840 was approximately 148 per thousand live births. Something like one in every seven children would not reach the age of one. Women routinely died in childbirth. Cholera epidemics, tuberculosis, typhoid, and occupational injury killed adults at rates that would, today, be considered symptoms of a humanitarian emergency. The average urban household of 1860 had, by our standards, an astonishing quantity of funeral experience.
The material culture of the period reflected this. Mourning clothes were standardized, and the duration of mourning was calibrated by relationship: two years for a husband, one for a parent, six months for a sibling. Hair from the deceased was woven into jewelry. Post-mortem photography, which became inexpensive in the 1850s, produced a small industry of posed images of deceased children and adults. Households kept lockets, letters, and watchcases that contained memorial relics. The dead were, in a specific and infrastructural sense, still around.
Into this environment arrived the spiritualist claim that the dead were also still around in a less metaphorical sense. For a household that had lost two children in a single typhoid season, the prospect of a medium who could transmit a message was not absurd. It was, for a great many people, the only available form of closure.
The scientific frame
What is easy to miss, from a twenty-first century vantage, is how new the category of "the natural" was in the 1850s and 1860s. The same decades that produced the Fox sisters also produced the discovery of the electromagnetic spectrum, the formulation of the germ theory of disease, Darwin's Origin of Species, the first telegraph cables across the Atlantic, and the first photographs of the surface of the moon. The list of phenomena that had, within living memory, been classified as impossible or supernatural and were now routine scientific objects was extremely long. If invisible electromagnetic waves could carry a telegram across an ocean, the intuition that there might be other invisible channels, through which other kinds of information might travel, was not, on its face, irrational.
This was the frame in which serious physicists started attending séances. Sir William Crookes, who isolated thallium and is on the periodic table, investigated the medium Florence Cook in the 1870s and concluded that she was producing genuine materializations. Oliver Lodge, a pioneer of radio transmission, became convinced that his son Raymond, killed at Ypres in 1915, was communicating through mediums. Alfred Russel Wallace, the co-discoverer of natural selection, was a committed spiritualist for the last four decades of his life. These were not credulous people. They were people who had, in their professional lives, seen the map of possible phenomena expand repeatedly, and who were not prepared to draw a new line a priori.
The Society for Psychical Research
In January 1882, a group of Cambridge-educated researchers (Henry Sidgwick, Frederic Myers, Edmund Gurney, and others) founded the Society for Psychical Research with the explicit goal of investigating paranormal phenomena with scientific rigor. The SPR's stated methodology was straightforward: interview witnesses, cross-examine them, corroborate where possible, and publish both the results and the methods. Its Proceedings and Journal were peer-reviewed, in the sense that other SPR members reviewed them; they were circulated to universities; they were cited in the mainstream press.
The SPR's first major project was the Census of Hallucinations (1894), which surveyed 17,000 people and reported that approximately one in ten of them had, at some point, experienced a sensory impression of a person who was not physically present, and that a measurable minority of those experiences coincided temporally with the death of the person perceived. The statistical analysis, performed by Henry Sidgwick's wife Eleanor (also a pioneering mathematician), argued that the coincidences exceeded what chance would predict. The methodology was questioned at the time; it is still questioned. The attempt, however, was serious.
What the SPR also did, which is less often remembered, is produce a string of the most rigorous debunkings of spiritualist mediums ever performed. The same researchers who took the question seriously were also the researchers who exposed Eusapia Palladino's rope-trick mechanisms in 1895 and Douglas Blackburn's thought-transference fraud in 1908. The SPR's reputation for credulity is, in my reading, inaccurate. What it was actually doing was testing. And testing is, by its nature, going to produce negative results in most cases and positive results in a few. The SPR's positive results are what survived in public memory. Its extensive record of negative results has mostly not.
The photograph problem
The spiritualist movement produced a subgenre of photographic evidence that is now, taken as a visual archive, genuinely strange. Spirit photography (images in which the subject is joined by a translucent figure of a deceased relative) became commercially available in the 1860s. William H. Mumler of Boston was the first major commercial practitioner. Mary Todd Lincoln sat for a Mumler portrait in 1872 and received an image in which her deceased husband was standing behind her with his hands on her shoulders. The image is still extant. It is not, taken as a photograph, convincing; the "ghost" is a double exposure, and Mumler was eventually tried (though acquitted) for fraud. What is striking is how many of these photographs were produced, how much money changed hands for them, and how many grieving clients reported that they were, despite the obvious technique, a comfort.
Sir Arthur Conan Doyle, a committed spiritualist from about 1916 onward, endorsed a number of spirit photographs in the years after his son Kingsley's death in the 1918 flu pandemic. His endorsement of the Cottingley fairy photographs (see elsewhere in this archive) was part of the same pattern. What Doyle was demonstrating, in a difficult way, was the motivational structure of Victorian credulity: the photographs were not convincing, but the need was. He knew his son was dead, and he was a father. The photographs were what was available.
What ended it
Victorian spiritualism did not end with a single exposé or scientific refutation. It declined, gradually, over the three decades between about 1920 and 1950, and the decline had several causes. The First World War had killed sixteen million people, and the bereavement demand for spirit communication had been, for a period, almost inexhaustible; the 1918 flu pandemic added millions more. By the mid-1920s, professional mediums had become a scaled commercial industry with a predictable fraud profile, and a new generation of debunkers (Harry Houdini being the most famous, though not the most effective) had begun systematically exposing the mechanical techniques involved.
More importantly, the late Victorian mourning economy itself declined. Hospital-based childbirth reduced maternal mortality. Antibiotics, once they arrived in the late 1940s, reduced the infectious disease load dramatically. Cremation, which had been uncommon in 1890, was routine by 1950. Mourning clothes became optional, then vestigial, then extinct. The generation that buried its children also, eventually, died. The generation that came after did not have the same quantity of absence to explain.
The SPR still exists, by the way. It has been publishing its Journal continuously since 1884. It maintains an archive in Cambridge that is probably the largest collection of primary-source paranormal investigation in the world. I have been there, briefly, and it is not what one might expect. It is quiet. The files are neatly labeled. The researchers who produced them believed that they were doing science, and in a certain way, with unrepresentative subjects and unfalsifiable claims, they were: they were keeping notes. The ghosts did not arrive. The notes did. The notes are still there.