A Second Helping A Review of the A.I.R. Report - 2 of 4
Paul H. Smith
"AN EVALUATION OF
REMOTE VIEWING: RESEARCH
AND APPLICATIONS"
This document is the second of a four-part review of the CIA- sponsored report by the American Institutes
of Research (AIR) of its evaluation of the U.S. government's twenty-four
year long remote viewing program. Part One, Bologna
on Wry Bread, covers the operational intelligence portion of the
program. Part Two, A Second Helping,
points out that the research reviewed by the AIR was inadequate as a basis
for a fair assessment of remote viewing. Part Three, Scraps
and Crumbs, examines the AIR's faulty evaluation of that research.
Part Four, has additional notes and corrections.
In Part 1 of this review I discussed some of the highlights of the
AIR/CIA report that was responsible for the demise of the STAR
GATE remote viewing program. I focused primarily on the operations
half of the unit. As promised, Part 2 will concentrate on the research
portion of the program. As Part 1 explained, two experienced scientists
were retained to do the evaluation: Dr. Jessica Utts, a nationally-known
expert on statistical analysis and a supporter of parapsychology research,
and Dr. Ray Hyman, a professor of Psychology at the University of Oregon,
and among the most widely-known skeptics of parapsychology.
Utts and Hyman were to conduct a thorough
review of "all laboratory experiments and meta-analytic reviews conducted
as part of the research program," which amounted to about 80 reports, a
number of which summarized several experiments each (p. E-2). The scientists
would be assisted by a couple of AIR associates, an additional statistics
consultant, and AIR's president, Dr. David Goslin.
All experiments available for review were
conducted over an approximate ten-year period by Dr. Ed May, who had assumed
responsibility for the experimental side of the remote viewing program
at SRI-International in the mid-1980's after the departure of Dr. Hal Puthoff,
who had lead the program since it's founding in 1971. In the early '90s,
May and his experiments moved to Science Applications International Corporation
(SAIC).
On the surface, AIR's review of the research
program is a more credible effort than was its evaluation of the operational
unit. The review process was to all appearances well documented, the rationales
employed seemed well thought out, and a seemingly equitable point/counterpoint
format between pro-psi Utts and anti-psi Hyman adopted in an attempt to
bring consensus to the differing conclusions arrived at by the two primary
evaluators. However, the evaluation turned out to be nothing so much as
a comedy of errors, with both sides—AIR and the STAR GATE
researchers—in starring roles.
To best sort out this muddled situation,
we will explore the shortcomings of the research effort first, to provide
a context in which to understand where AIR failed in its evaluation.
THE RESEARCH PROGRAM
Dr. Ed May and I are on the same side on
this issue, so it's not overly pleasant to have to criticize the SAIC research.
Nonetheless, there are things that must be brought out to understand what
really happened during the AIR review.
I will begin with a brief summary of the
ten experiments ultimately examined by AIR reviewers. Fortunately, Dr.
Utts provided summaries in her portion of the AIR report. In the interest
of space I have condensed these summaries still further, but retain the
essentials:
-
Purpose. Two-fold:
-
(a) determine if a "sender"(i.e., someone
at the site) was necessary to help the viewer access the target or if the
viewer could obtain information merely by being focused on the site through
a coordinate or other mechanism;
-
(b) Determine whether "static" targets—i.e.,
the photos—would be easier or harder to perceive than "dynamic" targets—i.e.,
short video clips.
-
Target. Photos from the pages of National
Geographic sealed in envelopes; alternatively, short video clips.
-
Experiment 2:
-
Purpose. Discover if viewers can correctly
determine computer-generated "binary targets"—"Is it one or is it zero?"
"Is it yes or is it no?" If so, this might lead to answering questions
such as, "Is there a bomb in this building or not?"
-
Target. A computer-driven random number generator.
-
Experiment 3:
-
Purpose. Using a magnetoencephalograph (MEG),
attempt to detect anomalous brain signals of remote viewers.
-
Target. A flashing light observed by a "sender."
-
Experiment 4:
-
Purpose. Determine if remote viewing can be
used in an information-sending capacity.
-
Target. Specially designed or chosen targets
with distinct characteristics. Presence or absence of each characteristic
represented either a "1" or a "0." If a characteristic was perceived and
reported by the viewer, a "1" was recorded; if the characteristic was not
perceived to be present, a "0" was recorded. Binary numbers could thus
be constructed by tabulating presence or absence of target characteristics.
If successful, information could be "sent" in a manner roughly analogous
to Morse code.
-
Experiment 5:
-
Purpose. Test three novices to see if they
could remote view.
-
Target. National Geographic photos placed
on a table in another room.
-
Experiment 6:
-
Purpose. Could lucid dreaming be used as a
tool to enhance remote-viewing?
-
Target. National Geographic photos contained
in opaque envelopes placed next to the bed where person was attempting
to achieve a "lucid dreaming" state.
-
Experiment 7:
-
Purpose. Determine if a person becomes "physiologically
aware" of being watched, even though he/she is not consciously aware of
being watched.
-
Target. The subject him/herself. He/she is
seated in a room with a video camera aimed at him/her. Galvanic skin response
was then measured to determined if it increased during periods of observation.
-
Experiment 8:
-
Purpose. Using an electroencephalograph (EEG),
attempt to identify interruptions in alpha brain-waves when a remote viewing
target is flashed on a computer screen in another room.
-
Experiment 9:
-
Purpose. Determine if viewers could describe
a target briefly displayed on a computer monitor. (This is the remote viewing
portion of Experiment 8.)
-
Target. Target (not further described in the
report, but perhaps the aforementioned video clips) was displayed briefly
on a computer CRT in another room.
-
Experiment 10:
-
Purpose. An improved version of Experiment
1. An equal number of static and dynamic targets were employed, no "senders"
were used, and all attempts were done at SAIC in California, instead of
from the participants homes, as was the case with Experiment 1.
-
Target. Selections from a pool of various
photos and video clips. [Summaries were excerpted from pp. 3-33 to 3-41
of the AIR report.]
As listed in the AIR report, the three assigned
missions of the STAR GATE-affiliated research program
were to: (1) Demonstrate through scientific experiment the existence of
the remote viewing phenomenon; (2) Determine the cause and effect mechanism
through which the phenomenon functions; and (3) Explore methods and techniques
to enhance the operational effectiveness of the phenomenon [p. 2-1]. These
goals, incidentally, were essentially unchanged from the days of the GRILL
FLAME effort, as enumerated in a report I recently saw dating from 1977.
Let us evaluate these experiments in terms
of the three stated missions of the research effort—in effect, the intended
purpose for which research money had originally been appropriated.
Mission 1: DEMONSTRATE EXISTENCE OF
THE REMOTE VIEWING PHENOMENON
As designed, seven of the SAIC experiments
would provide useful support to the existence of the remote viewing phenomenon,
and one would have been of marginal value. Two would not have given useful
support in demonstrating the RV phenomenon. Experiment 3 (which was unsuccessful
because of faulty experiment construction) might have been of marginal
value but would not in itself have provided unambiguous support for the
existence of RV. Had this experiment been a success, any anomalous brain
signals detected might still have been the artifact of some other common
element in the viewers' experiences, backgrounds, or training. However,
isolating and identifying the signal might ultimately have led to useful
information which could potentially provide later support to the existence
of RV.
Experiment 2, which focused on computer-generated
"binary" targets, might demonstrate a paranormal effect, but not in the
sense of classic remote viewing. The experiment's results may actually
display some sort of "dowsing" effect (though some would argue that RV
and dowsing are but different sides to the same coin), or perhaps even
a psychokinetic (PK) effect, since it would be difficult to determine if
the viewer were merely anticipating the correct answer, or in some way
influencing the number generation process.
Experiment 7 could be useful in demonstrating
the existence of some sort of paranormal linking effect between observed
and observer. But the experiment would not have been useful in supporting
the existence of remote viewing. No useable information could be transferred
across space and/or time using the demonstrated effect.
Mission 2: DETERMINE CAUSE AND EFFECT
RELATIONSHIP
None of the SAIC experiments, even when
successful, would have provided any substantial answers to the cause-and-effect
relationship for the remote viewing phenomenon. Only Experiments 3 and
8 would have provided even marginal information bearing on cause-and-effect,
and they would have merely demonstrated an anomalous effect without identifying
a causal linkage.
Mission 3: DEVELOP MORE EFFECTIVE RV
OPERATIONS METHODS
Because of their design, seven of the SAIC
experiments could have provided no benefit whatsoever in developing new
or better operational methods or techniques. Experiment 2 showed potential,
were it to lead to a reliable "yes/no" selection technique. However, the
experiment only involved trying to "second-guess" a machine. A real-world
binary problem, such as, "Is Gen. Dozier in Italy?" or "Will Hezbollah
attack the Statue of Liberty tomorrow?" involves much different selection
mechanisms than tapping a computer key, is of much different psychic texture
than "0"s and "1"s, and has far greater ultimate consequences—and therefore
dramatically greater emotional loading in the viewing process—than do yes-or-no
type questions on a computer.
Experiment 4, an attempt to use RV to transmit
coded information by identifying specific characteristics of a target,
uses remote viewing not as an intelligence collection tool, but as a communications
method. This would by definition be of no use for operational RV; however,
if such a communications ability could be reliably developed, it would
have great utilitarian value—to include undetectable transmission of intelligence
from a denied area.
As explored in Experiment 6, lucid dreaming
might possibly provide added value to the remote viewing process (though
I personally have my doubts). Therefore, this experiment at least had the
potential to benefit operational remote viewing.
When we tabulate the results, this is what
we find:
| MISSION # |
Relevant |
Maybe |
Irrelevant |
| Mission 1—Proof of Phenomenon |
7 |
1 |
2 |
| Mission 2—Determine Cause/Effect |
0 |
1 |
9 |
| Mission 3—Operations Enhancement |
0 |
3 |
7 |
By far the majority of the ten experiments
focus on proving the existence of the phenomenon—the first mission. The
other two missions were essentially ignored. In fact, one experiment—determining
whether someone is physiologically aware of being watched—is interesting
from a parapsychology standpoint, but has almost nothing to do with remote
viewing (one individual prominent in RV research did suggest that the experiment
might be a preliminary step toward determining if one could be aware of
being targeted by a remote viewer). Another three—Experiments 2,3,and 4—are
only indirectly related to RV, particularly RV as an intelligence collection
tool.
The research program's first error was
fundamental—it failed to evenly address all aspects of this three-fold
mission, concentrating instead almost exclusively on the first of the specified
goals. This would have been forgivable, had the program indeed successfully
proved beyond any doubt the existence of remote viewing as a paranormal
phenomenon. However, as demonstrated by Ray Hyman's conclusion that something
was happening, but it was too early to assume it was psi [pp. 3-75, 3-76],
this goal eluded the program. To be fair, this effect was certainly amplified
by AIR efforts (discussed below) to "stack the deck" against STAR GATE.
Nonetheless, the whole research emphasis was generally out
of sync with the stated purpose of the STAR GATE effort.
Perhaps the rationale was something like
this: "Until we can prove the existence of the phenomenon, there's no point
in trying to establish the cause-and-effect; and if these first two questions
aren't answered, it seems pointless as well to bother much about how to
enhance the operational effectiveness of something we haven't proved to
exist, nor know how it works." At any rate, the bulk of the experiments
focused on trying to convincingly demonstrate an effect, and few went beyond
that decidedly preliminary step. While statistically, at least, some remarkable
effects were demonstrated, both Utts, the supporter, and Hyman, the skeptic
agree that nothing irrefutably conclusive was proven. Utts believed that
the effects nonetheless demonstrated the strong possibility of a psi-based
effect. Hyman and the AIR researchers concluded there was not enough evidence
to say even that.
Would the results have been better had
May concentrated more on true RV experiments, and tried more concertedly
to address the other two missions? The answer to this is a qualified yes.
Notably, the experiments more closely approaching a classical remote-viewing
model were the most successful, with Experiment 10 producing quite impressive
results. Those which departed most from the model tended to be the least
conclusive. Additionally, had more experiments been designed to enhance
operational methods or develop new techniques, they would in and of themselves
have provided additional proof for existence of the phenomenon. If RV technique
gets good enough to work nearly every time, producing solid information
under a variety of conditions, the phenomenon is essentially proved—accomplishing
two of the research missions for the price of one. (As they say, nothing
succeeds like success.)
Cause-and-effect research would, however,
have been less productive. Of course, if in some brilliant moment of discovery
a verifiable causal relationship were found and demonstrated, the skeptics
would have to retreat. But such an event is highly unlikely. Thus far,
there is not even a worthwhile hypothesis as to what the phenomenon is
in terms of the "physical" world—if it even has such a connection (though
there are one or two interesting ideas waiting in the wings to emerge).
We do have a pretty good idea what the basic nature of remote viewing is
NOT: It is unlikely to be electro-magnetic in any sense, as demonstrated
by the successful remote viewings done in electro- magnetically shielded
Faraday cages, or those which are precognitive or retrocognitive, seemingly
in violation of the accepted laws of physics which radio waves or other
electromagnetic phenomena obey.
Since we have no other good candidate to
account for information transmission of the nature and quality good remote
viewing produces, we are pretty much left in the dark as to where to start.
It makes far more sense to work on practical applications and leave the
fundamental underpinnings for those with more time, money, and no need
to answer to a house full of skeptics. Regrettably, the wavering focus
of the SAIC effort was inadequate for fair assessment of remote viewing
in its own right.
I should point out here that the experimental
focus was not entirely up to Dr. May and his team. Representatives for
a contracting agency write the statement of work and draft the contract
that specifies what will be done in the course of the research. A review
of the DIA contracts shows that much of the work performed at SAIC was
indeed specified by the DIA representative.
Still, there is a lot of behind-the-scenes
give-and-take before the formal document is drafted, and the government
representative must rely heavily on the expertise and advice of the contractor
in the process of deciding what can or should be done in the course of
the contract. Further, there is an added degree of flexibility built into
the contract to allow researchers to explore promising directions that
may not necessarily have been foreseen during the original contracting
process. This flexibility is necessary and desireable to allow examination
of serendipitous discoveries or unforseen effects, but it is also a point
vulnerable to exploitation by researchers with their own agendas to pursue.
Ultimately, both parties share responsibility for the direction a research
program takes, right or wrong.
As an additional consideration, the SAIC
work was a follow- on to previous research done via a still-classified
connection with an agency which mandated more generalized research. Remote
viewing was only one of several phenomena to be explored. PK, for example,
was always of interest in prior research programs and, as the random number
generation experiment shows, some vestiges of interest may have remained
in the SAIC experiments. This interest in general parapsychology seems
to have bled over into the DIA/SAIC remote viewing research.
May's broader-ranging experimental focus
did produce some interesting and perhaps even ultimately useful research.
Unfortunately, there was not a more rigorous attempt made to route the
SAIC research further away from this general focus and concentrate more
intently on what should have been STAR GATE's RV-centered
research agenda. Ultimately, the overly-eclectic approach increased vulnerability
to pointed criticism which Ray Hyman and AIR were only too eager to provide.
In fact Dr. Hyman does give lip service
to Ed May's difficulties in not being "free to run the program to maximize
scientific payoff," because May was required to "do experiments and add
variables to suit the desires of his sponsors," resulting in "an attempt
to explore too many questions with too few resources. . . The scientific
inquiry was spread too thin." (3- 46) Of course, as just mentioned, there
was much room for negotiation in the contracting process, and May could
certainly have argued for a more narrow focus. The evidence suggests it
was more the other way around. In fact, several people in a position to
know have suggested that Dr. May saw the RV research contracts as an opportunity
to explore some of his own parapsychological interests at the same time
as pursuing the official purposes for which the research was contracted.
However that may be, Hyman's gratuitous
comments are no exoneration in this matter. If Hyman recognized the eclectic
nature of the research AIR was to evaluate, he is certainly well- qualified
enough as a scientist to realize that the limited numbers of experiments
were inadequate to answer the question EITHER WAY as to whether or not
remote viewing had any efficacy as an intelligence collection tool. That
Hyman persisted (as discussed below) in pretending that they did seems
intellectually dishonest.
PROTOCOLS
The bias in favor of wider parapsychology
research was not the only problem with the SAIC experiments, however. Curiously,
May and his colleagues seem to have followed rather anachronistic procedures
in conducting even the experiments which were more purely remote viewing
in character. My first quarrel is with the target pool.
Remote viewing, both experimentally and
operationally, has been pursued for more than two decades. While a lot
has been learned, some of the most valuable data—that accumulated by the
operational RV unit in its various incarnations—has hardly been considered
in the research process. The operational data set includes brilliant successes
that point to improved ways of doing things, as well as ignominious failures
which can be just as instructive. There was a fair amount of well-structured
experimentation at Ft. Meade in targeting and cuing methods, RV data documentation
and analysis, accessing target details, and so forth. Unfortunately, the
operations activity was kept mostly separate from the research program
until after the 1992 transition to STAR GATE, and
even then the connection existed primarily to provide subjects for some
of the SAIC experiments. The vast database from the Ft. Meade unit of thousands
of documented sessions—both training and operational—remains largely unmined.
One pronounced difference between RV targeting
in the SAIC research effort and that in operations was that operations
focused on "live" targets, while the SAIC experiments used two- dimensional
images, both static photographs (pictures gleaned from the pages of National
Geographic) and short, live-action video clips. The thinking at SRI was
that the video clips might provide increased "change" values, adding variety
to the target material, perhaps making it easier for viewers to detect
and report. Similarly, photos were selected that displayed significant
"change in entropy"—that is, contrast and variety in shapes and in color
and value patterns that again theoretically would make detection and reporting
easier.
In comparison, daily operational remote
viewing missions at Ft. Meade accessed targets in real time "on the ground"
(or water, or whatever), not in a photograph. What photos that were provided
were not used as targets, but only for later feedback or to guide analysts.
There was plenty of evidence that the operational viewers were indeed accessing
the sites themselves and not merely the feedback folders (in operations,
feedback was usually pretty lean and sporadic anyway). When a viewer accurately
describes several significant structural or functional details that are
completely lacking from feedback packages yet which are later confirmed
to be at the site, it becomes obvious very quickly that "real" remote viewing
is occurring. This literally happened scores, even hundreds of times.
However, at Ft. Meade there was some experimentation
with photos as actual targets. This was conducted both as an in-house training
exercise, and at one or two other times as part of one of the rare instances
when the operations unit was asked to participate long-distance in an SRI
experimental series during the mid-to-late '80s. Across the board operational
viewer results dropped off when targeted against "static" photographic
targets. At the time, video clips were not avalable as an option (or so
I presume, as participating viewer received only terse feedback), so I
can render no judgement as to whether they would have been more effective.
Indeed, to a remote viewer accustomed to
accessing actual sites in four-dimensional space, a static photograph is
not a representation of the Statue of Liberty in New York harbor or Mount
Pinatubo during an eruption. It is in reality only a colored piece of paper
in a manila envelope. It's not surprising that results from operational
viewers suffer when targeted under such circumstances.
To be sure, an experienced viewer CAN access
a photograph—the positive results of several of the SAIC's experimental
RV sessions demonstrate this. But if the focus had been on "real"—and therefore
naturally dynamic—sites as opposed to two- dimensional representations,
May and his colleagues might not have had to bother about testing the use
of "dynamic" moving images (the videos) to provide greater change and variety
to improve remote viewer detection; or about mapping the "change in entropy"
of the static images to enhance researchers' ability to decode viewer results,
as was done for these experiments. Perhaps there were experimental control
reasons why such a fixed target pool was desired. In my mind, however,
the drawbacks far outweigh the possible benefits.
Another troublesome aspect of at least
one of the SAIC experiments was the apparent need to experiment further
with "senders"—individuals sent to the target site to act as a "beacon"
or a "transmitter" for the remote viewer. Indeed, one of the stated purposes
of the experiment was to determine if a "sender" was necessary. Senders
and beacons were used in the early SRI experiments, and continued to be
used for beginner trainees at Ft. Meade, simply as a way of providing a
connection with the site that the novice viewer could easily grasp. Both
at SRI and Ft. Meade, however, the need for senders in advanced remote
viewings was surpassed long ago. The introduction of coordinates as a targeting
mechanism, and later (to avoid any hint of contamination) encrypted coordinates,
made senders/beacons obsolete. No degradation in response quality resulted,
and in fact, accuracy seemed even to be enhanced. The encrypted coordinates
provided the added benefit of defusing one of the most popular (if improbable)
criticisms of coordinate-cued RV—that some viewer might just "memorize"
what was at the end of all the geographic coordinates in the world, and
cheat.
The need for beacon or sender was already
discounted by the late '70s and early '80s, and was certainly well established
at the time Ed May took over as primary researcher. Though the sender/beacon
personnel were dispensed with later in the SAIC ten-experiment sequence,
it was puzzling why the researchers felt the need to thus "reinvent the
wheel" at the start.
In the end, the main problem with the SAIC
experiments was not that they were particularly poor experiments, but that
they should have been better. More importantly, the experiments could—and
really should—have focused more particularly on remote viewing, guided
by the three missions that Congress had decreed when earmarking funds for
the program. As it was, the primary consequence of the SAIC program was
to provide a very tempting strawman for the AIR bull (at the behest of
the CIA) to gore and trample, hoodwinking the general public into believing
that AIR had a live matador at its mercy. In reality, the matador wasn't
even in town.
But now, after I have spent several pages
"blaming the victim," it's time to turn my attention to the perpetrator.
This is part 2 in a series of 4.
© 1996 Leonard Buchanan on behalf of Paul Smith aka "Mr. X"
|