PROJECT STARGATE

THE CLAIM

That the United States government funded psychic spying for over two decades because remote viewing works, and that the 1995 closure was a cover rather than a conclusion.

THE THEORY, AS ITS PROPONENTS TELL IT

The proponents' case starts from a fact that sounds like invention and is not: the CIA, the DIA and Army Intelligence really did fund remote viewing research continuously from 1972 to 1995, at Stanford Research Institute and later SAIC, and the files really were declassified in 1995. Anyone can read them.

From there the argument is institutional. Intelligence agencies are not sentimental about budgets. A programme that produced nothing would not survive three administrations, multiple funding reviews and a change of host institution. Its persistence, proponents argue, is itself evidence that something in it was working.

Specific results are offered as proof. Pat Price's description of a Soviet facility at Semipalatinsk, Joe McMoneagle's reported description of a large submarine under construction, and Rosemary Smith's location of a downed aircraft in Africa are the canonical cases, and supporters note that some of these were treated seriously inside the agencies at the time.

The closure is then read as concealment. The 1995 evaluation was commissioned by the CIA, conducted quickly, and used to justify shutting the programme and releasing the files. The argument holds that a genuinely useful capability would be moved and reclassified, not advertised as a failure, and that the public closure is the cover for a private continuation.

WHAT IS KNOWN

The programme existed, at the scale described, and cost roughly twenty million dollars over its life. The declassified files, including the SRI reports and internal assessments, are public and have been since 1995.

The 1995 evaluation by the American Institutes for Research engaged statistician Jessica Utts and psychologist Ray Hyman. Utts concluded a statistically significant effect had been demonstrated in the laboratory data; Hyman concluded the effect was not distinguishable from methodological flaws. Both agreed on the operational finding, which is the one that mattered: the programme had never produced intelligence that was actionable.

The canonical successes do not survive contact with the files. Descriptions were typically vague, generated in quantity, and matched to targets after the fact by analysts who knew the answer. The programme's own managers recorded the difficulty of getting anything specific enough to task a satellite against.

EVIDENCE FOR

• The programme is genuinely confirmed: agencies, budgets, host institutions and duration are all in the declassified record.
• Jessica Utts, a statistician of standing, concluded the laboratory database showed an effect unlikely to be chance, and she has not retracted that.
• Several serving intelligence officers took the work seriously at the time, which is a fact about the record even if it is not evidence about reality.

EVIDENCE AGAINST

• The programme's own evaluation, commissioned by the people paying for it, found no operational value in twenty-three years. That is the strongest possible form of negative evidence: the believers had the budget, the access, the time and every incentive to succeed.
• The laboratory effect Utts identified has never been independently replicated under conditions that exclude sensory leakage and post-hoc matching, and Hyman's methodological objections were never answered.
• Uri Geller was tested at SRI as part of this work and has been repeatedly demonstrated to use conjuring methods. His involvement is treated by proponents as supporting and reads the other way.
• The concealment argument is unfalsifiable as stated. A closed programme proves failure; a continued one proves success; no observation is allowed to count against the claim, which is a property of the argument rather than of the evidence.

ASSESSMENT

Debunked as to the ability, confirmed as to the programme, and the gap between those two sentences is the entire value of this file. It is the best available demonstration that a government funding something for decades is not evidence the thing is real. Agencies chase capabilities that would be decisive if they existed, and MKULTRA (OSA-001) shows the same appetite with a far darker outcome. Stargate ran for twenty-three years with sympathetic management and a classified budget, and the reason we can be confident it found nothing is that its own sponsors said so, in public, while releasing the files.

WHAT WOULD CHANGE THE GRADE

• The programme is CONFIRMED and the ability is not. Keep those separate: agencies, budgets, host institutions and twenty-three years are all in the declassified record.
• To reopen: a pre-registered, independently replicated protocol producing an effect that survives blinding and excludes post-hoc matching. Jessica Utts's statistical finding has stood since 1995 without such a replication, and Ray Hyman's methodological objections have never been answered.
• The continuation claim, that a working capability was quietly moved and reclassified, is not gradeable as stated: closure is offered as proof of failure and continuation as proof of success, so no observation can count against it.

FURTHER READING

• Stargate Project · Wikipedia
• American Institutes for Research, An Evaluation of Remote Viewing (1995) · the report that closed it
• CIA CREST archive · the declassified programme files, searchable online