Michael Persinger

Michael Persinger, working at Laurentian University in Ontario, applied weak and complex magnetic fields across the temporal region of volunteers in a sound reduced room and recorded what they reported. A substantial proportion described a sensed presence, a feeling that something or someone else was in the room with them, and some described it in religious terms. The equipment acquired the name God helmet in the press, which he did not choose and which fixed the popular understanding of the work.

His interpretation was that the sense of presence is generated by transient instability between the hemispheres, that the right hemisphere sense of self intrudes on the left and is experienced as another, and that religious experience in general has a normal neurological basis of this kind rather than an external cause. He set the hypothesis out in a paper in 1983 and pursued it for decades. It is worth being precise about what such a claim can and cannot establish. Showing that an experience can be produced by stimulating the brain shows that the brain is involved in having it, which no one disputes, and does not by itself settle what the experience is about, any more than producing the taste of salt by stimulation shows that salt does not exist.

In 2005 a Swedish group led by Pehr Granqvist ran a double blind version of the experiment, in which neither the participants nor the experimenters in the room knew whether the field was on. The reported experiences occurred, and they tracked measures of suggestibility and of absorption taken beforehand rather than tracking the field, with participants in the sham condition reporting them as often as those exposed. The authors concluded that suggestion, in a dark and quiet room, with an expectation established in advance, accounted for the results.

Persinger disputed the replication on technical grounds, arguing that the equipment had not been set up correctly, that the field patterns and timings were not those he used, and that the exposure durations differed. The Swedish group answered that they had used his apparatus and his parameters. The exchange did not converge, and the disagreement was never closed before his death in 2018. The atlas records it as an open dispute in the literature rather than as a refuted claim or an established finding, which is the accurate description and also the more useful one, since the design question at issue, how to blind an experiment whose whole subject matter is expectation, applies well beyond this case.