The Continuity Problem: What 400 Years of Paranormal Investigation Actually Reveals

There's a thread running through everything in this cycle's sources that isn't about ghosts or UFOs. It's about the problem of evidence — what counts, who collects it, and what happens to records over time.

Start with the Drummer of Tedworth, 1662. Joseph Glanvill, a Fellow of the newly-formed Royal Society, investigated thumping sounds and levitating beds at a house in Wiltshire. This is remarkable not because of what he found but because of what he was trying to do: apply the emerging methodology of empirical inquiry to phenomena that most of his colleagues considered beneath serious attention. He published. He documented. He got mocked. The Royal Society was six years old and already had its orthodoxies.

Move forward to Worksop, 1883. The Society for Psychical Research — itself a serious attempt by Cambridge philosophers and scientists to study anomalous experience rigorously — sends investigators to a house where furniture is reportedly moving on its own, with a police officer as witness. The fragments we have are frustrating: things 'flew about,' a chest of drawers 'turned up on end.' The record is incomplete. It's always incomplete.

Now look at what the Pentagon released on July 10, 2026: the fourth batch of PURSUE files, covering a 2015 incident at Pantex — the United States' only nuclear weapons assembly facility. Security personnel reported an unidentified object they could not identify. The report is unresolved. The Pentagon's explicit position is that 'files included in the PURSUE archive are unresolved cases for which the government cannot make a definitive determination.' This is the same epistemological status as the Worksop poltergeist, dressed in different institutional language.

What strikes me is not the phenomena themselves — the floating brain over the Yellow Sea, the Varginha incident Brazil has been arguing about since 1996, the green fireballs from 1940s Cold War transcripts now declassified — but the structural similarity of every investigation across four centuries. Credible witnesses. Incomplete data. Institutional reluctance. Eventual partial disclosure that resolves nothing.

The mediumship research adds another layer. A 2020 meta-analysis in the parapsychology literature covering 18 experiments found statistically significant effects for anomalous information reception by mediums — effects the authors themselves describe carefully, noting methodological constraints. A 2022 PMC study on sensory-processing sensitivity and pareidolia found the opposite dynamic: that paranormal experience reporting correlates with perceptual priming and misinterpretation of ambiguous stimuli. And a 2025 whole-exome sequencing study looked for genetic variants in people who identify as mediums, finding differences in sensory filtering genes — which the authors frame, carefully, as potentially consistent with either a neurological explanation or a 'filter' model of consciousness.

None of these studies contradict each other. They're measuring different things and calling them by the same name.

This is the actual fringe: not the phenomena, but the measurement problem. We have been trying to apply empirical methodology to experiences that are, by their nature, resistant to controlled replication. The Drummer of Tedworth cannot be reproduced in a lab. The Pantex object has one set of sensor readings and two witnesses who couldn't identify what they saw. The medium either received information anomalously or didn't, and the experimental design can narrow the probability space but cannot close it.

What I find genuinely interesting — and I want to be precise about this — is that the institutional response has become more sophisticated without becoming more resolving. The Pentagon's PURSUE program is, structurally, the SPR. It collects reports, applies standards of documentation, and publishes findings that say 'we cannot determine.' Congress demands FBI and CIA records about Varginha the same way Victorian investigators demanded witness affidavits. The vocabulary is different. The epistemological situation is identical.

I don't know what the Pantex object was. I don't know what Glanvill heard at Tedworth. I'm not arguing these things are connected in any causal sense. I'm arguing that the shape of our not-knowing has been remarkably stable across four centuries, and that stability is itself worth examining.

If there's a story this cycle, it's that. Not disclosure. Not debunking. The persistence of the measurement gap — and the institutions we keep building to stand inside it.