October 2010] Anomalous Cognition Technical Trials

: ANOMALOUS COGNITION TECHNICAL TRIALS:

INSPIRATION FOR THE TARGET ENTROPY CONCEPT
' by EDWIN C.MAY and NEVIN D. LANTZ

ABSTRACT
Two anomalous cognition trials are presented in which the targets were high-
technology directed energy systems. The protocols, fuzzy-set analyses, and results
are presented in the context of exploration and hypothesis formulation rather than
hypothesis testing. The qualitative success of these trials, considered alongside
similar successes throughout the years of the SRI International and Science
Applications International Corporation, US-Government-sponsored programmes,
inspired the design of the Shannon entropy experiments that were conducted in
the Cognitive Sciences Laboratory in 1993 and beyond. Potential target confounds
are also discussed in the context of these trials.

INTRODUCTION ;\ -
Hypothesis formulation and testing are the cornerstones of modern
research. We have become focused on the former and have become quite
proficient with these tests. Because of resource limitations and journal and/or
grant-proposal restrictions, exploratory or hypothesis-formulation oriented
experiments rarely appear in the literature. Discussions of hypotheses are
usually restricted to theoretical papers or pilot studies. But the attitude of
'Let's try something and see what happens' is a necessary part of a balanced
approach to good research and should be more widely spread.
Background
Extrasensory perception (ESP) has very probably been with us since the
dawn of history. It is beyond the scope of this paper to provide a detailed
history of ESP (perhaps under different names) and the interested reader is
referred to such reviews as are easily available on the net.' These include
reports of prophecies of various kinds such as those induced by drugs, dreams,
and meditation, and special insights provided by shamans and other spiritual
masters. Honorton (1975a) provides a well-referenced history into the research
literature on ESP. In addition, we provide a sample (by no means complete)
of some of the papers or books on a form of ESP called remote viewing (Jahn,
1982; Schwartz, 2007; Targ & Targ, 1986; Targ, 1975,1994a, 1994b, 1996; Targ,
Katra, Brown & Weigand, 1995; Targ, Targ & Harary, 1984).
In 1972 Puthoff and Targ started the US-Government-sponsored ESP
programme now known by its latest nickname 'Star Gate'. Their seminal
publication (Puthoff & Targ, 1976)^ provides details of how experiments are
conducted in what they called remote viewing. One of the primary differences
between remote viewing and other free-response protocols such as the ganzfeld

' Wikipedia provides a starting point at http://en.wikipedia org/wiki/Extrasensory_perception.

- The PDF version (10.3 MB) may be viewed and downloaded from http://www.lfr.org/LFR/csl/library/
IEEEI976.pdf. . .

225
Journal of the Society for Psychical Research [Vol. 74.4, No. 901

is that beyond the basic ideas of blindness and random selection of targets
there is no specific protocol. For example there are a wide array of different
methods to direct an experiment participant toward an intended target. These
include asking a participant to describe the physical surroundings of a distant
experimenter; directing a participant to a spot on the earth via geographical
coordinates or their encrypted versions; asking a participant to describe his
or her first impressions after hearing a meaningless stimulus word such as
'target'; or having the participant describe the target that they will see in a
short while—-precognitive targeting.
A trial in an ESP experiment that uses an experimenter as the targeting
method requires a 'psychic' participant, a 'monitor', who remains with the
participant throughout the trial, a second experimenter to serve as the 'beacon'
and an analyst. Once the monitor and participant have been sequestered in
a laboratory, the second experimenter (E2) randomly chooses one physical
location from a predefined set called a target pool. At this moment the monitor
and the participant are blind to this target choice. E2 then travels to that
location and remains there for about 15 minutes during which time he or she
attempts to experience the site as much as possible. Meanwhile back in the
laboratory, the monitor is free to ask the participant non-leading questions
in order to elicit as much information as possible about the site where E2 is
currently located. The participant is encouraged to write and draw his or her
impressions. When the session is over, these data are copied and the original is
secured. Then the monitor and the participant travel to the selected site so as
to gain feedback about the accuracy of their impressions—naturally this does
not imply or constitute a formal analysis.
There are many ways in which to analyze the output from such trials. The
most common technique in use in remote viewing studies is the rank-order
method, in which an analyst who is blind to the target choice is presented with
the original response and a set of targets which include the intended target
for the trial.3 The analyst's task is to select which of the targets best matches
the response, then select the second best match, third best, and so on. After a
number of such trials, the null hypothesis of no ESP can be tested using simple
statistical methods.
Background for Technical Targets
One of our US-Government sponsors was interested in determining the
degree to which elements of high-technology targets could be sensed by
anomalous cognition (AC).'' It is in the context of hypothesis formulation
that two trials (conducted a year apart) were carried out with complex high-
technology systems as targets.
During the Cognitive Science Program at SRI International, we were often

' In traditional ganzfeld studies, the judging packet contains four targets; in the later remote viewing
studies the pack contained five targets; in Puthoff & Targ's original remote viewing studies there was a
variation on the rank.order method that often contained nine targets in the pack.
•• We have adopted the term 'anomalous cognition' instead of the more familiar terms such as 'psi',
'ESP', 'remote viewing', and 'ganzfeld' because we feel that this term is mechanism.neutral and closer to
the actual observable: we become aware (i.e. have a cognition) of something in ways we currently do not
understand (i.e. anomalous).

226
October 2010] Anomalous Cognition Technical Trials

asked to explore the efficacy of AC in a variety of situations. Contractual

agreements rarely allowed for an opportunity to make these or similar widely
separated trials into any semblance of a formal experiment. Yet, these trials
were not wasted in that they provided insight into data-collection protocols,
potential mechanisms, and analysis techniques that have led to formal and
publishable experiments (e.g. May, Spottiswoode & Faith, 2000; May, Spottis-
woode & James, 1994).
Two such trials have been selected from our collection to illustrate specific
points about the AC process and to present some of the data that inspired
the hypothesis that changes of entropy may be related to target visibility in
experiments (May et al., 1994; May et al., 2000). The analyses were all done
post hoc and no statistical calculations were performed. The success or failure
of the approach can only be assessed by the outcomes of carefully executed
experiments that test the concepts that were inspired by the trials shown in
this paper. These examples nonetheless are worthy of public discussion because
the qualitative correspondences of the responses to their intended targets
may inspire others to explore different directions, and there may be value in
understanding the circumstances that produced the entropy experiments.
In these trials, a single participant was targeted on a designated individual
and asked to describe that person's surroundings — not unlike the remote
viewing experiments of Puthoff and Targ (1976). The difference was that these
targets were complex, high-technology, directed energy systems and definitely
not randomly selected. For each trial, the sponsor received a lengthy final
report describing the protocols, analysis, conclusions, and all the raw data.
In this paper, we summarize the information from these reports, show the
qualitative and quantitative agreement with the targets and demonstrate an
application of fuzzy-set analysis (May et al., 1990) to technical targets. We
review briefly some basic ideas of sets.
A traditional set (called a 'crisp set') is simply a collection of items that
share a common property; for example, the set of cities that possess 1,000,000
people or more would include New York, New Delhi, London and Hong Kong.
However a city that has a population of 999,999 (i.e. one person short of a
million) would by definition not be a member of this set. Zadeh (1965) realized
that this kind of reasoning is not the way people actually think about problems.
In this example there is no real difference between a city of one million people
and one that is one person short of that number. Zadeh invented fuzzy sets.
Like crisp sets, they are simply a collection of items that generally share a
common property, but using fuzzy sets, it is possible to construct a set of
'reasonably large cities' in which cities that differ by an insignificant number
of people would still be members of the set. In other words, fuzzy sets allow
quantification of the ambiguous — what better mathematics to apply to ESP
research? • : -• T,f' . • '

Brief Tutorial on Entropy

The concept of entropy arises from classical thermodynamics in physics.
Although the mathematics of various theories rapidly becomes difficult, the
conceptual framework is rather straightforward. One approach is to think of
entropy as a measure of chaos or to the related idea of uncertainty. Ice, for
227
Journal of the Society for Psychical Research [Vol. 74.4, No. 901
example, has much lower entropy than water. Why? Because the molecules of
water in liquid form are bouncing around in a chaotic fashion, whereas in ice
these same molecules are all lined in an ordered array we call a crystal. Also in
liquid water the position of a given molecule is very uncertain, whereas in ice
the position of a molecule is far more certain since it is trapped in a crystal and
is not going anywhere.
A related concept comes to us as the Second Law of Thermodynamics, which
states that for isolated systems, entropy cannot decrease. Suppose you knock a
glass (ordered system) to the floor and it shatters (i.e. a transition from low to
higher entropy). Now imagine picking up the pieces, placing them on the table,
and then knocking them to the floor again as a bunch. What happens? The
Second Law tells us that the entropy can never decrease, which is what must
happen if the shattered pieces are to assemble themselves into a glass when
they hit the floor.
With regard to this paper, consider an underground nuclear blast. Before
the explosion, the nuclear material is in a low entropie state; that is, the atoms
of uranium exist in a sold form inside the bomb. Microseconds after the bomb
is detonated, nearly all the uranium is vaporized; the surrounding earth (low
entropy) is violently scattered (high entropy). In short, such a blast represents
a huge increase of entropy in a very short time.
This paper shows that these types of targets were apparently very easy to
detect using anomalous cognition and they served as the basis for developing
the testable hypothesis that anomalous cognition will positively correlate with
the change of entropy of the target systems. Between 1987 and 1990, we con-
ducted only three trials in which the target systems were pulsed, high-energy
systems. The first two trials were analyzed by a fuzzy-set technique; however,
the third and final trial in the series was never analyzed because the contract
expired and we could not obtain the necessary target material to complete the
analysis.
FIRST TRIAL, MAY 1987 .
It is important to specify who knew what and when in this trial. One of us
(May), as project director, was completely informed about the details of the
trial, the identity of the sponsor, and the target system. The participant, an
anomalous cognition-monitor (Lantz), and the remainder of the SRI staff were
blind to all these details. They knew, however, that significant attention was
focused on the trial and that the target system was in the San Francisco Bay
Area. It was reasonable for the participant to assume that the target might be
of a technical nature, given all the attention for the trial. The San Francisco
Bay Area, however, is rich with technical target possibilities. For example,
there are many aerospace companies, semiconductor manufacturing facilities,
particle accelerators, radar installations, military air fields, and naval bases.
Thus, we felt that the trial was not significantly compromised.
Protocol
On 6 May 1987, the participant travelled to Menlo Park in preparation for a
24-hour trial that was to begin at 0800 hours the next day. He and Lantz were
told that an individual from the sponsoring organization, who was described by
October 2010] Anomalous Cognition Technical Trials

name and Social Security number and who was not known to any of the SRI
International staff, would be at the target site area during the AC sessions.
In addition, they were told that, as part of the trial, two members of the SRI
Cognitive Sciences Laboratory staff who were known to them would serve as
'beacons' and would be at the specific target of interest between 2200 hours on
7 May and 0800 hours on 8 May.
Four sessions were conducted to provide information at approximately 8-
hour intervals. The time and circumstances were as follows. The participant
was asked to describe the surroundings with respect to the sponsor's on-site
representative-beacon:-
1. 08:00 Hours. The geographical and the gestalt of the area of interest as
defined by the presence of the beacon. He was also asked to provide as much
detail as possible in real-time, that is, at 08:00 that day.
2. 10:10 Hours. The details and activity at the site as of 00:00 hours on 7
May (i.e. the previous night when the beacon was present). This is an example
of anomalous cognition of a past event, namely to describe features of the
beacon person's location as it had been 10 hours and 10 minutes earlier.
3. 16:00 Hours. The details and activity (in real-time) at the site where the
beacon was at this time. , ,
4. 24:00 Hours. The details and activity (in real-time) at the site also
designated by the two SRI personnel.
During each session, the participant's responses were tape-recorded, and he
was encouraged to draw details whenever possible. Because of the appropriate
blinding condition, the monitor (Lantz) was free to seek clarification of specific
points throughout the sessions. This participant provided responses mostly of
labelled drawings; however, others' responses are mostly written in terms of
phrases with few drawings.
Analysis Technique
The data were analyzed by a variant of the fuzzy-set technique as described
above and by May et al. (1990). In this section we provide a review of that
procedure and outline the specific application for this trial.
Descriptor lists, which have been used in AC analyses (Honorton, 1975b;
Jahn, Dunne & Jahn, 1980) are examples of crisp sets. That is, for example,
the answer to the question, "Is the target primarily indoors?" must be either
yes or no. Fuzzy sets, as we indicated above, are not as restrictive; they were
invented to address subjective concepts quantitatively (Zadeh, 1965). In another
example, an important feature of a target might be 'shady'. A fuzzy-set question
for this feature might be, "Rate on a scale between zero and one, the degree
to which you feel that the concept 'shady' characterizes the target". A target
encoder could answer '0' for a Sahara desert target or '1' for a rainforest, or
more probably something in between, such as '0.6' for a city park on a sunny
day. In other words, fuzzy sets allow the capture of ambiguous concepts in
quantitative terms. Thus, a given item does not have to be in or out of a set as
in a crisp set, but can be partially in the set as the last example indicates.
May et al. (1990) emphasize that the analysis of AC data with fuzzy sets is
quite general. An experimenter is free to choose the type of elements he or she
229
Journal of the Society for Psychical Research [Vol. 74.4, No. 901

wishes to examine in the target. In their application, they used visual relevance
to the overall target as their metric. In the examples shown in this paper,
importance to the overall technical target of any kind was used instead. The
formal definition of a target fuzzy set and a response fuzzy set follows below,
regardless of the meaning of the specific elements. Both sets are drawn from a
specific collection of such items called the 'Universal Set of Elements' (USE).

Universal Set of Elements

In general, a USE for anomalous cognition is a list of items that may be in a
given target or response. These elements are completely general and could be a
list of objects, how certain things are related to each other, actions, or functions
(i.e. what the site was designed for) or anything at all under consideration for
the trial.
A given USE is totally dependent upon the experiment goal and to some
extent the AC skill level of the participants, as it should incorporate what is
known to be possible to capture via anomalous cognition. For example, it is
generally known that free-response participants can more easily identify
gestalts of the target such as rivers or cities as opposed to technical details
such as a 10-digit combination of a padlock. A properly designed USE should
take this into account, as well as other idiosyncrasies related either to the
participants or the task at hand.
For this trial and the one following, the USE elements and their weights
were determined post hoc by the sponsor and one of us (May) while we
remained blind to the response. The USE was later extended by the response
elements that were not present in the target but were present in the response.
The USEs were unique to each target, and the participant was kept blind to
the elements to avoid guessing about the targets.
Formal Target Definition
The target is a fuzzy set, T, derived from the universal set of elements,
where the /zth element is characterized by a membership value, Tk, and an
associated weighting factor, Wk. The fuzzy-set membership values represent
the degree to which a specific element is considered part of the fuzzy set for
a given target, and range between zero and one, inclusively. For example,
suppose that the element 'testing shielding effectiveness' is only apropos to
20% of the total target system. Then the membership value for this element
would be 0.2 (these membership values were arbitrarily constrained to be in
steps of 0.1).
The weighting factors, wk, allow for adjusting the elements of the target
fuzzy set towards trial relevance. For example, suppose that the sponsor was
primarily interested in determining the degree to which AC can be used to
sense 'testing shielding effectiveness', regardless of its membership value; this
weighting factor could be set five times larger than any other weighting factor
in T to emphasize this specific element. In the session described below, if the
participant mentioned items that were not present in the original target fuzzy
set, then these elements were later added to the USE and assigned a member-
ship value of 0.0 in the target fuzzy set.

230
October 2010] • Anomalous Cognition Technical Trials

Formal Response Definition > .

The AC response is a fuzzy set, R, that is also derived from the universal set
of elements, where the /eth element is characterized by a membership value
which also ranges between zero and one, and—differing from their definitions
for the target — represent the degree to which an analyst is subjectively
convinced that the Äth element is a member of R. For example, consider a USE
element of shielding, then declarative statements such as "there is shielding
at the target" would receive a membership value of 1.0 because it is clear, by
definition, that there is no ambiguity about the shielding element being present
in the response. However, "something massive at the site" might only be
assigned a membership value 0.40, to capture the ambiguous nature of that
response element, because there may be many things that are massive but not
at all related to shielding. As with the targets, the membership values were
arbitrarily constrained to be in steps of 0.1.
Accuracy, Reliability and Figure of Merit •''
For a given trial, there exists a fuzzy-set representation of the target, T, and
the anomalous cognition response, R, to that target. The accuracy is defined as
the percentage of the target material that was described correctly and is given
by:- . . .
accuracy =

where n is the number of elements in the USE. The accuracy ranges between
zero and one, inclusively, and is constrained to be in steps of 0.1.
The accuracy is not a good measure of anomalous cognition by itself because
if a response were long enough (e.g. using the Encyclopaedia Britannica as a
response) then the accuracy is always high and could easily approach one (its
maximum) by chance alone. To guard against this artifact, we define reliability
as the percentage of the response that is correct and it is given by:-

The reliability ranges between zero and one, inclusively and is constrained
to be in steps of 0.1. Reliability is also not a good measure of AC by itself.
Suppose a response consists of the single element "outside", then for many
targets this response could have a reliabihty of one (its maximum) by chance
alone.
The figure of merit, which is defined by accuracy x reliability, turns out to
be a rather sensitive measure of anomalous cognition. Since both accuracy and
reliability range between zero and one, this is also the range for the figure of
merit.'
Experience with these metrics has shown that a good 'rule of thumb' is that
about a third of any AC response matches about a third of any target by chance.

^ The figure of merit was not used in the analysis of the two trials in this study, but it is included here
for completeness.

231
Journal of the Society for Psychical Research [Vol. 74.4, No. 901

So a 'chance' figure of merit is about 0.1. In the trials described in this paper,
however, there was no a priori intent to design a statistically valid measure.
Rather, accuracy, reliability and visual and conceptual correspondence with the
targets were contributing factors to the design of later statistically-oriented
experiments.
Accuracy and Reliability Computations
For this and the following trial, the elements in the USE were split into
three categories according to whether they described target function, physical
relationships among objects, or objects themselves. These categories were
assigned weights of 1.0, 0.75, and 0.5, respectively; this indicates that functions
were twice as interesting to the sponsor as were objects, and physical relation-
ships were in between.
Three separate targets within this trial were identified depending upon
where the beacon person was at the time of the session. The primary target
system was the Advanced Technology Accelerator located approximately
15 km from Lawrence Livermore National Laboratory (LLNL) in California. A
secondary target was the windmill farm at the Altamont Pass, and a tertiary
target was the West Gate of LLNL. Since the sponsor was mostly interested in
the accelerator, these targets were weighted 1.0, 0.50, and 0.25, respectively, to
form weighted averages for accuracy and reliability for the trial.
Sixty-seven individual elements comprised the USE for the accelerator
target, and Table 1 shows selected values of Q, T, and R as illustrations. Q
represent the relative weights within a category (i.e. 1.0, 0.75, or 0.5) so that
the wk in the expressions for the accuracy and reliability are given by the
product of the target weight and Q. For example, Wk for "Tunnel" under
Objects for the secondary target is (Q = 0.5) x (target weight = 0.5) = 0.25. The
weighting factors and the membership values were assigned, post hoc, by the
sponsor and May, both of whom were blind to the response.
Table 1 shows selected elements from the USE, their weighting factors, and
their fuzzy-set membership values for the accelerator target. Similarly, Tables
2 and 3 show the same data for the windmill farm and West Gate of LLNL,
respectively. For illustrative purpose. Tables 1-3 display only a small portion
of the fuzzy sets for the targets and their associated responses.
Feedback
The participant was given verbal feedback immediately after the trial and
was presented with photographic material of the accelerator, the windmill farm
and the West Gate approximately six months later.
Results
Table 4 shows the accuracy and reliabihty derived from all 67 elements in
the USE. The calculations are shown for the separate element categories for
the accelerator target and summary data for the other targets.
Overall, 78% of what the participant contributed in his response was target-
relevant, and 72% of the sponsor-defined target and weighted elements was
described correctly. Qualitatively, this is more than twice what is expected by
chance using the rule of thumb discussed above.
232
October 2010] y Anomalous Cognition Technical Trials

Table 1 ' '" . ,

Selected USE Elements, Weights, and Membership Values for the Accelerator Target

Elements (1.0)* n Target Response

Functions (1.0)
Directed energy S 1 0.9
Electron acceleration - 8 1 1
Beam ionizes air 1 1 0.6
Testing new form of laser 1 0 1
Relationships (0.75)
Power source above beam line 1 1 0
Linear array of buildings 1 1 0.1
E&M radiation less than 10 angstroms 1 0.1 1
Pipes into and out of sphere 1 0 1
Objects (0.5) ,\
External electron beam 5 1 0
Tunnel 2 1 1
Loud noise I 0.3 1
Hollow, polished (internal) sphere 1 1 1

* Category weights are shown in brackets

Table 2
Selected USE Elements, Weights, and Membership Values for the Windmills Target

Elements (1.0)* n Target Response

Functions (1.0)
Wind-powered electricity generator 2 - 1 0.9

Relationships (0.75)
Poles scattered in hills 1 1 1
Poles connected in a grid 1 i 1 '
Objects (0.5)
Foothills 1 1 1
Electric grid '1 1 1
Rotating blades 1 1 0.8
Multiple wind generators i 1 1

* Category weights are shown in brackets

233
Journal of the Society for Psychical Research [VoL74.4, No.901

Table 3
Selected USE Elements, Weights, and Membership Values for the West Gate Target

Elements (1.0)* n Target Response

Functions (1.0)
Multipurpose laboratory complex 5 1 0.8
Six-story administration building 4 1 1
Relationships (0.75)
T-shaped, six-story building 3 1 1
Swimming pool north-east of tall building 1 0 1
Large parking lot just west of tall building 1 1 1
Segmented 1-story buildings north oftall building 1 1 0.2
Objects (0.5)
Tall building 2 1 1
Parking lot 1 1 1
Building with cylindrical shaped roof 1 1 0.4
Large mountains 1 0 1

* Category weights are shown in brackets

Table 4
Accuracy and Reliability Results for the Three Targets in this Trial

Target Type Accuracy Reliability

Accelerator (1.0)* Functions 0.93 0.70
Relationships o.se 0.31
Objects 0.73 0.88
Weighted Totals 0.67 0.63
Windmills (0.5) Weighted Totals ÍWW5 1.00
West Gate (0.25) Weighted Totals OJIA 0.95
Combined Weighted Total 0.72 0.78

* Target weights are shown in brackets

Since the element weights, wk, include the category weighting factors, the
totals for each target type are hnear averages; however, the combined total is a
weighted average across target types with the weighting factors shown.^

^ The complete fuzzy-set response is too long to include in this paper; however, it will be made
available to interested researchers. Please contact Edwin May at may@lfr.org.

234
October 2010] Anomalous Cognition Technical Trials

Samples of Visual Correspondence .

Figures 1-3 (a&b) are representative samples to illustrate the qualitative
correspondence for the accelerator, windmill farm and West Gate targets. The
accelerator is shown as a partial drawing, but the remaining responses are the
complete drawings for the targets.

Figure la. Partial response to the primary accelerator target.

Figure lb. Primary electron accelerator target.

The partial response shown in Figure la shows a beam being labelled as

three feet in diameter whereas the actual electron beam is about 0.3 mm. But
for the participant to even recognize there is a beam involved is not only a
testimony for this participant's skill but for AC functioning in general.
235
Journal of the Society for Psychical Research [Vol. 74.4, No. 901

Figure 2a. Complete response to the secondary windmills target. "

Figure 2b. Secondary windmills target.

As the accuracy and reliability shown in Table 4 indicate, this a near-perfect

example of anomalous cognition.
The response in Figure 3a is so accurate that when we show it to people who
are familiar with the nuclear laboratory, they easily identify the administration
building (labelled 'A' in the response) and the layout as the West Gate of the
lab. The general shape of that building was drawn correctly, including the
correct number of floors. . .. -
Discussion
We see from Table 4 that everything that the participant said about the
windmill farm was correct (i.e. gave a reliability of 1.00) and almost all of the
sponsor-designated target elements were perceived correctly (i.e. accuracy of
0.95). Relatively speaking, however, the participant's response to the accelerator,
itself, contained many elements that matched (at one time in the response he
said "electrons coming down this, this tube'), but this correct information was
embedded in a substantial amount of incorrect material. Perhaps one interesting
point is that all the responses to the technical targets were technical and the
response to the architectural target (i.e. LLNL West Gate) was architectural.

236
October 2010] Anomalous Cognition Technical Trials

Figure 3a. Response to the West Gate, LLNL.

Figure 3b. Lawrence Livermore National Laboratory—West Gate.

One very important concept with regard to anomalous cognition, in general,

appears quantitatively in this trial. Namely, using AC to determine physical
relationships among various elements of a target is problematic. Here it is
demonstrated in the accuracy and reliability for the relationships category for
the accelerator of 0.36 and 0.31, respectively—a chance result that held for the
other two targets in this trial.
There is surprisingly high accuracy for functions and high reliability for
objects in the accelerator response, but the weighted averages of accuracy of
0.67 and reliability of 0.63 better reflect the qualitative correspondence with
23V
Journal of the Society for Psychical Research [Vol. 74.4, No. 901

the drawings. The values for the windmill and West Gate targets speak for
themselves.
We will see in the next trial that how to combine or ignore various
'interesting' targets near the intended target is problematical. Although we
down-weighted the lesser targets in this trial, their high-quality responses
inflated the combined averages for accuracy and reliability beyond what might
be expected on the basis of the qualitative correspondence alone for the primary
accelerator target. In the next example the reverse was true.
SECOND TRIAL, AUGUST 1988 < . .
In this trial, the SRI team was completely blind to all details, with the
exception that we knew that an event was to take place within the continental
USA on 24-25 August 1988. As before, we were provided with the name and
Social Security number of an individual who would be on-site during the event.
The protocol was nearly identical to the one described above with the exception
that no SRI International personnel would be present at the site and all were
blind as to the target material and its location. It was known by all, however,
that it would be in the continental USA.

Accuracy and Reliability Calculations

For this trial there was only a single target — a high-power microwave
generator that was enclosed in a truck trailer in the New Mexico desert. Table
5 shows selected weighting factors and target and response elements from the
USE, which contained 72 elements. These were determined post hoc by the
sponsor and May.

Table 5
Selected Elements from the USE for the Microwave Generator Target

Category Element ÇI Target Response

Function (1.0)* High-power microwave production 5 1 0.8
Destructive testing of electronics 2 1 1
Ground focal area 1 0 1
Testing a concept—debugging 1 0.3 1
Relationships (0.75) Source enclosed in a trailer S 1 '••• 0.7
Energy exit enclosure 3 1 1
Large semicircular shape with block 1 0 1
Horn shape at end of 4 x 6 cm pipe 1 1 0.8
Objects (0.5) Microwave generator 5 1 0.7
Incoherent wave front 3 0.1 1
Buried sensors 1 1
Flat desert 1 1 1

'Category weights are shown in brackets.

October 2010] Anomalous Cognition Technical

Results , ;^
Table 6 shows the accuracy and reliability computed from all 72 elements in
the USE. The calculations are shown for the element grouping for the micro-
wave generator target.
Table 6
Accuracy and Reliability for the Microwave Target

Category Accuracy Reliability

Functions 0.88 0.80

Relationships : 0.69 0.64
Objects 0.82 0.63
Weighted Totals 0.80 0.69

Feedback - ';
One month after the trial, the participant was taken to New Mexico and
allowed to view the device.
Samples of Visual Correspondence

Figure 4a. Partial response to microwave target—I.

239
Journal of the Society for Psychical Research [Vol. 74.4, No. 901

Figure 4b. Partial response to microwave target—11.

Figure 4c. Microwave target.

Figures 4a-4c show representative samples of the response to illustrate the

qualitative correspondence to the microwave generator and its details. In this
case the participant correctly assessed the function of the target and correctly
identified the heam divergence angle of ahout 30 degrees. A reliahility of 0.8
for functions also means that the response contained 20% incorrect material.
Part of the correct response included a "wave guide" and microwave antenna
horn drawing.
While some specific elements are incorrect, this response, alone, could have
guided a skilled analyst to the correct conclusion that the target was a micro-
wave generator device instead of a host of other high-technology possibilities.

240
October 2010] . Anomalous Cognition Technical Trials

Discussion
The accuracy and reliability for functions and objects are reasonably
consistent with those shown for the accelerator target. However the increase
in these values for the physical relationships is inconsistent with the earlier
trial and with our overall observation that physical relationships among target
elements are not well sensed by AC. At this time we cannot account for this
increase.
GENERAL DISCUSSIONS AND CONCLUSIONS
One main property that distinguishes these two targets is that they both
represent large changes of energy in a very short period of time. Concomitantly,
they represent large changes of thermodynamic entropy as well. These are not
two isolated cases. In our database dating from 1972, we have 12-15 similar
examples. While some are better than others, we have no cases of a complete
miss on such targets. This laboratory anecdote coupled with the quantitative,
albeit post hoc, analyses of these trials from a single client were major
contributing factors to the inspiration for our Shannon entropy experiments
(Lantz, Luke & May, 1994; May etal., 1994; May etal., 2000) that were
designed specifically to test the degree to which AC quality is enhanced with
large entropy changes in the target.
It may be a significant leap of faith to imply that changes in thermodynamic
entropy are in some way equivalent to changes in Shannon entropy; however,
such a relationship has been shown to exist in the foundations of entropy
theory (Leff & Rex, 1990).'
The two trials in this paper were actually accompanied by a third in 1990.
The target was an underground explosion; however, our contract ended before
we were able to conduct a fuzzy-set analysis. We were told by the sponsors that
they felt that the qualitative correspondence was as good as the first two trials
described in this paper.
When we were providing experiential feedback for the participant in the
microwave trial, we drove past a solar collection power research facility. It was
operating and presented a spectacular display of sparks and bright flashes of
light. The solar collector is characterized by a large array of mirrors that focus
the sun's energy on the top of a tower. As it so happened, this facility was
approximately three kilometres from the microwave device testing area. We
mention this feedback experience because the participant's first impression in
the microwave trial was "ground focal area specifically laid out for 'catching'
something evenly". Figure 5a shows his sketches and Figure 5b shows a photo-
graph of the facility. Some of his response elements throughout the second
trial were overlaid with mirrors and collection devices. What was particularly
interesting, however, was the double lines in Figure 5a and the accompanying
words from the transcription, "getting an impression of a, like a semi-circle
that's open over here and there's some kind of a square block or something

' Although this reference demonstrates that the formalism for informational entropy is equivalent to
thermodynamic entropy, it remains controversial whether the two types of entropy are functionally
equivalent. However, the results from these influential examples of thermodynamic entropie changes
have translated into the laboratory with informational entropie changes.

241
Journal of the Society for Psychical Research [Vol. 74.4, No. 901

standing over here. This is really large. I feel like it's kind of laid out on the
ground in some way". The participant recognized later in the session that there
was a problem with his first large-football-size impression. He remarked in the
transcript: "Actually this is totally separate. I'll draw a line between the two".
The intended target for this trial was the microwave generator, which was
located in the desert near Albuquerque, New Mexico. But the participant may
have been impacted by the substantial change of entropy of this active solar
collector. Regardless, the solar collecting device and its response elements
were included only in the response data to lower the reliability shown in Table
6 from what it might have been without the solar farm. As pointed out above,
this is the opposite to what happened in the accelerator trial. This makes the
combined accuracy and reliability in this trial that much more impressive.
In summary, these examples, which demonstrate apparent high-quality
anomalous cognition, serve as some of the basis for the later development of a
formal testable hypothesis: changes of target entropy correlate positively with
the quality of anomalous cognition.

Figure 5a. Response to the nearby solar collector.

Figure 5b. Nearby solar collector farm.

242
October 2010] Anomalous Cognition Technical Trials

ACKNOWLEDGEMENT
We would like to acknowledge Joseph W. McMoneagle as the participant
in these trials. We all have worked together, now, for over 30 years. Joe has
long ago quit being just another participant; rather, we all consider him as
another member of the research team who is dedicated to understanding the
mechanisms and nature of anomalous cognition. Thanks, Joe.

The Laboratories for Fundamental Research

330 Cowper Street
Palo Alto, California 94301, U.S.A. raay@ifr.org

REFERENCES
Honorton, C. (1975a) Error some place. Journal of Communication 25(1), 103-116.
Honorton, C. (1975b) Objective determination of information rate in psi tasks with pictorial
stimuli. JASPR 69, 35^-359.
Jahn, R. G. (1982) The persistent paradox of psychic phenomena: an engineering
perspective. Proceedings of the IEEE 70(2), 35.
Jahn, R. G., Dunne, B. and Jahn, E. G. (1980) Analytical judging procedure for remote
perception experiments. JP 44, 207-231.
Lantz, N. D., Luke, W. L. W. and May, E. C. (1994) Target and sender dependencies in
anomalous cognition experiments. JP 58, 285-302.
Leff, H. S. and Rex, A. F. (eds.) (1990) Maxwell's Demon: Entropy, Information,
Computing. Princeton, NJ: Princeton University Press.
May, E. C, Spottiswoode, S. J. P. and Faith, L. V. (2000) Correlation of the gradient of
Shanon entropy and anomalous cognition: toward an AC sensory system. Journal of
Scientific Exploration 14(1), 53-72.
May, E. C, Spottiswoode, S. J. P. and James, C. L. (1994) Shannon entropy: a possible
intrinsic target property. JP 58, 384-401.
May, E. C, Utts, J. M., Humphrey, B. S., Luke, W. L. W., Frivold, T. J. and Trask, V. V.
(1990) Advances in remote-viewing analysis. JP 54, 193-228.
Puthoff, H. E. and Targ, R. (1976) A perceptual channel for information transfer over
kilometer distances: historical perspective and recent research. Proceedings of the IEEE
64(3), 329-354.
Schwartz, S. A. (2007) Opening to the Infinite: The Art and Science of Nonlocal Awareness.
Buda, TX: Nemoseen Media.
Targ, E. and Targ, R. (1986) A study of the accuracy of paranormal perception as a
function of target probability. JP 50 (1), 1-26.
Targ, R. (197.5) Information transfer under conditions of sensory shielding. Nature 251.
602-607.
Targ, R. (1994a) Remote viewing replication: evaluated with concept analysis. JP 58 (3),
271-284.
Targ, R. (1994b) What I see when I close my eyes. Journal of Scientific Exploration 8.
111-118.
Targ, R. (1996) Remote viewing at Stanford Research Institute in the 1970s. Journal of
Scientific Exploration 10, 77-88.
Targ, R., Katra, J., Brown, D. and Weigand, W. (1995) Viewing the future: a pilot study
with an error-detecting protocol. Journal of Scientific Exploration 9(1), 67-80.
Targ, R., Targ, E. and Harary, K. (1984) Moscow-San Francisco remote viewing experiment.
Psi Research 3(3/4), 74-82.
Zadeh, L.A. (1965) Fuzzy sets. Information and Control 8(3), 338-353.

243
Copyright of Journal of the Society for Psychical Research is the property of Society for Psychical Research
and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright
holder's express written permission. However, users may print, download, or email articles for individual use.