Phenomena
The File on Psi Was Closed by Policy, Not by Data
14 minute read
I do not collect wonders. I collect procedures that institutions were willing to pay for when no one was watching. Psychic functioning — psi, in the dry name — is one of those procedures. The fashionable move is to laugh. Laughter is cheap. The file is not.
Annie Jacobsen’s Phenomena: The Secret History of the U.S. Government’s Investigations into Extrasensory Perception and Psychokinesis is a journalist’s book, not a parapsychology textbook. That is why I start there. She follows memoranda, contracts, and people. The United States did not dabble in a séance and go home. From the early Cold War into 1995 it ran a lineage of programs — Gondola Wish, Grill Flame, Center Lane, Sun Streak, Star Gate — because planners thought the Soviet Union might be doing the same work under the name of psychotronics. Competition, not mysticism, signed the checks.
What the laboratories actually did
At SRI International in the 1970s, Harold Puthoff and Russell Targ ran remote-viewing experiments with Ingo Swann, Pat Price, and others. The protocol that survived the anecdote is coordinate remote viewing: a viewer describes a place given only a coordinate or a sealed target, and a judge later ranks the transcript against the target and decoys. Some sessions, Price’s among them, became legends inside the program because the descriptions were specific. Legends are not data. The ranking method was the data, and it is the part a serious reader has to sit with.
Uri Geller passed through the same laboratories. His metal-bending and drawing trials are the part of the file a stage magician can most easily wound, and critics, James Randi and Ray Hyman among them, did wound the conditions. Jacobsen does not need Geller to be clean for the rest of the history to stand. A celebrity is a bad instrument. A protocol with a pre-registered target pool is a better one. The programs outlived the celebrity.
Princeton’s PEAR laboratory, under Robert Jahn, spent years on small engineering anomalies — intention and random devices — and published effects that were tiny and, in the lab’s telling, persistent. Tiny and persistent is exactly the shape of result that a culture trained on explosions will call nothing. It is also the shape of result that a bad protocol will counterfeit. Both sentences belong in the file. Only one of them is usually allowed out.
The review that did not say zero
Star Gate was reviewed and terminated in 1995. The American Institutes for Research evaluation is the document the joke never quotes in full. Statistician Jessica Utts, who had studied the data, concluded that the statistical case for a real anomalous cognition was strong enough that further research was warranted. Ray Hyman, the skeptic on the same review, concluded that the results were not proven and that methodological leaks could explain them. The program ended. The two reports did not agree. A state can close a shop for budget, embarrassment, or strategy. Closure is not a replication that returned zero.
The CIA’s later release of Star Gate material onto its reading-room archive lets anyone check the texture of the work: tasking, targets, evaluations, the military’s attempt to turn a fragile effect into an intelligence product. Some of that product was useless. Some of it was scored as better than chance by the program’s own analysts. The hits stay in the record as hits. The misses stay as misses. Together they are a research program that was interrupted, not a joke that closed the question.
The call
Take psi seriously again. That sentence does not mean believe every remote viewer, and it does not mean build a doctrine. It means return the question to methods that can fail in public. Pre-registration. Target pools sealed from the judge. Independent scoring. Publication of the misses. The same courtesy already given to unstable results in medicine and psychology.
Daryl Bem’s 2011 claims of feeling the future, and the replication war that followed, showed how unready the open sciences were for this argument. Unreadiness is not a refutation. Jacobsen’s contribution is historical: the most powerful state of the twentieth century already decided the question was operational. It kept the work in a compartment, then walked away while its own reviewers were still split. I am not split about the next step. The phenomena go back on the bench. Mockery is not a measurement.