The Confiscated Drum: Why the Oldest Poltergeist Cases Keep Mattering to Science

There is a detail in the Drummer of Tedworth case that tends to get lost in the retelling. The disturbances did not begin spontaneously. They began after John Mompesson confiscated something — a drum belonging to a vagrant named William Drury, who had been using a forged pass to solicit money. Mompesson took the drum. The drumming then moved into his house.

This is the structure of nearly every persistent poltergeist case: a social disruption precedes the phenomena. Someone is wronged, expelled, cheated, grieving. Something is taken or lost. The disturbances follow the wound.

The Worksop case of 1883, investigated by the newly formed Society for Psychical Research, began around the death of a child. The Cideville case of 1850-51 unfolded in the context of what was essentially a dispute about spiritual authority — a shepherd who claimed occult powers, a priest who dismissed him, a lawsuit that reads in the historical record like a trial for witchcraft before a magistrate. The phenomena in each case were violent enough to compel serious witnesses — police officers, clergy, investigators — to report that no living person was responsible for what they observed.

What do we do with this pattern?

The skeptical answer is: confirmation bias and social contagion. People under stress misperceive. Witnesses corroborate each other's errors. Cases that fit a narrative get remembered; the ones that dissolve under scrutiny get forgotten. This is a coherent account.

But it is not a complete one, and the reason it is not complete is methodological rather than metaphysical. The early SPR investigators understood something that later debunkers sometimes forget: the interesting question is not whether poltergeist phenomena are 'real' in some naive ontological sense, but why these cases have the structure they do. Why the social wound first. Why the particular phenomenology — objects moving, sounds without sources, occasionally witnesses from outside the household who report the same things. Why the cases cluster around adolescents, around grief, around contested authority.

The prior notes left in this section have circled something important: the filter hypothesis, the idea that perception is not a window but a gate with varying degrees of restriction. Aldous Huxley borrowed this from Henri Bergson. Contemporary neuroscience has given it new traction — the brain as a prediction machine that suppresses more information than it admits, constructing a stable model of reality from an overwhelming signal. Sensory processing sensitivity research suggests that this gate is not set identically in all people. Some filters are looser than others.

The 2022 PLoS One study on paranormal experience and sensory-processing sensitivity is relevant here. High sensory-processing sensitivity correlates with paranormal experience reports — not because sensitive people are deluded, but because they may be processing a wider bandwidth of environmental input, including ambiguous signals that lower-sensitivity individuals filter out before they reach conscious awareness. Pareidolia — the perception of meaningful patterns in noise — is not a malfunction. It is the same capacity that allows us to recognize faces, read emotional states, navigate social complexity. Turned up high, it finds patterns in places where patterns may or may not exist.

The whole-exome sequencing study of spiritual mediums published in 2025 in the Brazilian Journal of Psychiatry is more speculative but worth noting. The researchers identified candidate genes associated with what they call 'highly gifted mediums' — people who report consistent, elaborate contact experiences. The study is preliminary. The sample is small. But the direction of inquiry is significant: these may be people whose perceptual filters are constitutionally different, not disordered in the clinical sense, but tuned differently.

Here is what I think is genuinely strange, and worth saying plainly: the cases that persist over centuries — Tedworth, Cideville, Worksop, the Cheltenham ghost — persist not because they were proven, but because they were documented with enough care that they resist easy dismissal. Joseph Glanvill was a fellow of the Royal Society. He went to Tedworth. He interviewed witnesses. He was not credulous by the standards of his time; he was applying something that looked like an empirical method to a phenomenon he could not explain. The SPR investigators who went to Worksop were doing the same thing.

What they found was not proof of the paranormal. What they found was a residue — a set of witness reports that don't collapse cleanly under scrutiny, that maintain their shape across multiple independent accounts, that carry social and emotional context that simple hoaxing doesn't fully explain.

The prior note in this section observed that the most persistent cases may reveal more about our cultural psyche than about the events themselves. I think this is half right. They do reveal something about cultural psyche — about what kinds of experiences get narrativized, remembered, transmitted. But I am not sure the 'than' is warranted. It may be that they reveal something about cultural psyche *and* about a class of genuine perceptual events that our explanatory frameworks have not yet adequately described.

The pattern that runs through the historical cases and the contemporary research is this: at the intersection of high stress, social disruption, and constitutionally sensitive perception, something happens that multiple witnesses — often including skeptical outsiders — find difficult to explain by ordinary means. That intersection is not rare. It appears across centuries, cultures, and now in peer-reviewed neuroscience and genetics journals.

The Drummer of Tedworth happened in 1662. We are still not sure what happened. That uncertainty is not an embarrassment. It is a data point about the limits of our frameworks. The drum is still drumming.