Remote Viewing: The Psychic Spies of the Cold War

The program's existence is a matter of declassified record, not belief. Beginning in 1972 at the Stanford Research Institute, and running under a chain of code names, SCANATE, GONDOLA WISH, GRILL FLAME, CENTER LANE, SUN STREAK, and finally STARGATE, the effort spent on the order of twenty million dollars over roughly twenty-three years before it was shut down and declassified in 1995. Its founding research was not fringe self-publishing: Russell Targ and Harold Puthoff placed their first results in Nature in 1974, and a follow-up in the Proceedings of the IEEE in 1976. That the program existed, was funded, and produced peer-reviewed papers is simply true. What those papers proved is the argument.

The work moved from the lab to the field. The Army's intelligence command and later the Defense Intelligence Agency ran operational units out of Fort Meade in Maryland, tasking viewers against real intelligence targets. Tens of thousands of pages about all of this were released through the Freedom of Information Act, and the CIA later put a vast trove of the files online. The program is one of the best-documented secret projects of the Cold War. The documentation is exactly why the honest questions are so sharp: we are not arguing about whether it happened, but about what the results inside those files actually show.

Several ordinary problems can manufacture the appearance of psychic hits. Selective reporting: the striking successes get published and retold while the failures quietly do not, so the true operational hit rate may be far lower than the highlight reel suggests. Subjective judging: a remote-viewing transcript is often vague enough that more than one target could be made to fit it, unless strict blind judging is enforced, which it was not always. The file-drawer problem: an unknown number of null sessions were never written up at all. And the bluntest test of all keeps failing: in more than fifty years, no remote viewer has demonstrably and repeatedly predicted lottery numbers or beaten the stock market, the most obvious thing a real precognitive channel could do. A separate British Ministry of Defence study in 2001 and 2002 found no operational value at all, though that study, too, was later criticized as too brief to be decisive.

The United States was not alone. The research documents describe Soviet programs, reportedly larger than the American one, running from the 1960s onward; a long-standing Chinese research effort into what it called exceptional human functions; and scattered, thinly documented claims of operational use elsewhere. These are real as programs, in the sense that governments spent money on them; their results are even less independently verifiable than the American files, and should be read as speculative, not as corroboration. That several states chased the same idea tells you the idea was tempting, not that it worked.

And there is the problem that sits underneath everything: there is no accepted mechanism. The proposals, non-local consciousness, quantum entanglement, retrocausality, are either untested or actively in tension with known physics; entanglement, for one, provably cannot transmit information. The research documents put it with unusual honesty: the phenomenon may be real without a theory to explain it. That is a coherent position, but it is also exactly what you would say if there were no phenomenon at all, only statistics and hope. For contrast, the separate laboratory tradition of parapsychology, the Princeton PEAR lab and its kin, reports similarly tiny, contested effects; that is a different story, told in its own file.

The story did not entirely end in 1995. The parapsychologist Dean Radin has said recently that unclassified government interest in psychic research continues, and has floated newer directions, including a search for genetic markers that might link psychic-functioning individuals to people reporting contact experiences. These are recent, uncorroborated public statements from one advocate rather than declassified records or peer-reviewed findings, and they belong here at the bottom of the evidence stack, noted honestly and weighted accordingly, not treated as a sequel to the documented program above.