The Philip Experiment: Can You Create Your Own Ghost?

Most hauntings begin with a claim: something is here. The Philip Experiment began with the opposite premise — nothing is here, and we are going to build it anyway.

A group of eight volunteers, led by mathematician and psychical researcher Dr. A.R.G. Owen and coordinated by his wife Iris Owen, invented a fictional seventeenth-century English aristocrat named Philip Aylesford. They gave him a biography, a doomed love affair, a suicide. They were explicit that he had never existed. Then they sat around a card table and waited.

Origins: A Deliberate Fiction

The Toronto Society for Psychical Research (TSPR) was not a group of thrill-seekers. Dr. A.R.G. Owen was a Cambridge-trained mathematician and geneticist; the membership included an industrial designer, an accountant, a housewife, a bookkeeper, and a sociology student. Their hypothesis, drawn from earlier work by British psychologist Kenneth Batcheldor, was blunt:

Poltergeist phenomena are not produced by the dead. They are produced, unconsciously, by the living — and belief is the switch.

To test this, the group needed a subject that could not possibly be a real spirit. So they wrote one.

The Biography of Philip Aylesford

Every detail was checked against history to make sure it was wrong. The group even commissioned a sketch of Philip so they would share a common mental image.

Phase One: A Year of Nothing

For roughly twelve months the group meditated. They sat in a room decorated to evoke the seventeenth century, discussed Philip's life, and tried to visualise him. Reports describe occasional feelings of presence, but nothing measurable. No raps. No movement.

This failure matters. It is the control period that later critics rarely quote.

Phase Two: The Batcheldor Switch

In 1973 the group changed its method on Batcheldor's model. Out went solemn meditation; in came singing, joking, ghost stories, and dim red lighting. The mood shifted from laboratory to party.

Within weeks, the table began to rap.

One rap for yes. Two raps for no.

The answers were consistent with Philip's invented biography — and, crucially, only with it. Asked questions outside the fiction, the table hedged, contradicted itself, or fell silent. When a sitter privately believed something false about Philip, the table sometimes produced that false answer.

Reported Phenomena

• Sharp, percussive raps from the tabletop, audible across the room
• The table sliding, tilting, and on several occasions "walking" toward latecomers
• The table reportedly climbing stairs during a 1974 filming session
• Cool breezes localised over the table surface
• Lights dimming and brightening in apparent response to requests
• A recorded session for Canadian television in front of a live studio audience¹

Phase Three: Going Public

In 1974 the group performed for a television audience of roughly fifty people. The table rapped on request and moved across the studio floor with the sitters' hands resting lightly on top. The footage survives and remains the most-cited primary artefact of the experiment.

Two follow-up groups were run by TSPR — "Lilith," a French-Canadian spy, and a fictional man named "Sebastian" — reportedly reproducing raps more quickly than Philip's group had.

The Skeptical Case

A fair reading has to take the prosaic explanations seriously.

1. The ideomotor effect. Sitters resting hands on a table produce unconscious micro-pressures. Under expectation and social contagion these can tilt, slide, and knock a lightweight table.
2. No blind controls. Sitters knew the questions and Philip's biography. There was no protocol in which the group was ignorant of the expected answer.
3. No instrumentation. No strain gauges, no force plates, no infrared. The evidence is film, audio, and testimony.
4. Deliberate permissiveness. Batcheldor's method explicitly encouraged "artifact" — allowing minor cheating to build belief. This is methodologically fatal for proving a force.
5. Absence of replication under controls. No controlled laboratory replication with disinterested observers has reproduced the reported table locomotion.

What Survives Scrutiny

Strip away the contested claims and a smaller, stranger residue remains.

• The group produced no phenomena at all during the year of sincere, earnest meditation.
• Phenomena appeared only after the group adopted relaxed, disinhibiting, belief-friendly conditions.
• The "personality" that emerged was bounded by the fiction the group wrote — it did not know things the group did not know.
• Nobody in the group reported consciously producing the raps, and several were disturbed by them.

That pattern — phenomena tracking group psychology rather than any external entity — is the experiment's actual finding, and it is the same pattern reported in a century of poltergeist case files.

[1] The 1974 Toronto broadcast footage is the principal surviving audiovisual record; several later copies circulate with degraded audio.

The Philip Experiment did not prove ghosts exist. It may have proved something more uncomfortable: that a group of ordinary people, under the right social conditions, can generate an apparently autonomous personality with a voice, a temper, and a will — and can become genuinely unsure whether they are the authors of it.

Whether the raps came from unconscious muscles or from something that has no name yet, the group built a haunting out of nothing but agreement. Every field investigator who has walked into a house where a family is certain something is there should find that deeply relevant.