Progress Report No. 3 Coveri:;g the Period 1 April to 1 August 1974 Stanford Research Institute Project 3183 PERCEPTUAL AUGMENTATION TECI21'IQUES Approved For Release 2001/06/09 : CIA-RDP96-00791 R000300030006-7 >l Approved For Release 2001/06/09 : CIA-RDP96-00791 R000300030006-7 The purpose of the program is to determine the characteristics of those perceptual-modalities through which individuals obtain information about their environment, wherein such information is.not presented to any known sense. The program is divided into two categories of investigation of approximately equal effort, applied research and basic research. The purpose of the applied researca effort is to explore experimentally the potential for applications of perceptual abilities of interest, with special attention given to accuracy and reliability. The purpose of the basic research effort is to identify the characLer_stics of individuals posse=:sing such abilities, and to identify neurophysiological correlates and baz_c mechanisms involved in such functioning. Approved For Release 2001/06/09 : CIA-RDP96-00791 R000300030006-7 Approved For Release 2001/06/09 : CIA-RDP96-00791 R000300030006-7 A Applied Research Remote Viewing (a) Project Atlas Remote Viewing A remote-viewing experiment has been carried out on a client- designated target of interest, a European R&D test facility. The experiment, carried out in three phases, had as its goal the determination of the utility of remote-viewing under operational conditions. In phase I, map coordinates were furnished to the experimenter ded being the designation of the taroct the only additional information provi The experimenters then carried out a remote viewing as an R&D test facility. Subject 1=k(Px e) on a double-blind basis. The results of the experiment with uation experiment were turned over to client representatives for data eval. figure 1 shows the level of detail for a sample early effort ws the suoject's first effort at drawing layout, and Figure 2 sho at building a gantry crane he observed, both resuit: being obtained on a double-bl i-nd basis before exposure to client-held information. An artist's conception of the site as known to the client (but not to contract personne_)prior to the experiment is shown in Figure 3. experiment would have Were the results not promising, the a e stopped at this point. The results were judged to be of sufficiently good quality, however, that Phase Ii was entered in which the subject was made witting by client representatives. A second round of experimentation ensued with ?articipation Numerical designations for subjects are discussed in Section B. Approved For Release 2001/06/09 : CIA-RDP96-00791 R000300030006-7 Approved For Release 2001/06/09 : CIA-RDP96-0079 R0003 30006-7 0 UD Approved For Release 2001/06/09 : CIA-RDP96-00791 R000300030006-7 i1 /TTY CNITRUCTIO'T Approved For Release 20(''I~O CI~U T~ 9r6-fl~~~2~~~36006-7 - Approved For Release 2001/06/09 : CIA-RDP96-00791 R000300030006-7 Approved For Release 2001/06/09 : CIA-RDP96-00791 R000300030006-7 Approved For Release 2001/06/09 : CIA-RDP96-00791 R000300030006-7 of client representatives. The Phase II effort was focussed on the generation of physical data which could be client-verified, providing a calibration in the process. The end of Phase II gradually evolved into the first part of Phase III, the generation of unverifiable data not available to the client, but of interest nonetheless. Evaluation of the data by the client is under- way. (b) Costa Rica Remote Viewing Experiment - lUoj Cti ~~ 1'. :. utzu ?s N.:ii ~i4.L ~..r a. t,cu ..... .,,. ,....,..b ..~,.. ...,.....,`. ,. ___ .~ i involving a Central American target series. In this experiment, one of the experimenters (Dr. Puthoff) spent a week traveling through Costa Rica on a co ibination business/pleasu:?e trip. That is all that was known to the subjects about the traveler's itinerary. The experiment called for Dr. Puthoff to keep a detailed record of his location and activities, including photographs, each day at 1330 PDT. Si:: daily responses were obtained from Subject 1, five from Subject 4. The results Were of high quality and are presently being evaluated in detail, containing as they did a large amount of material. I r,) / (J Samples of that data are as follows. Of the five daily responses obtained from Subject 4, two were in excellent agreement, two had elements in common but were not clear correspondences, and one was clearly a miss. In the first of the two matches, Dr. Puthoff was driving in rugged terrain at the base of a volcano (Figure 4) and -he subject's response was "large bare table mountain, Approved For Release 2001/06/09 : CIA-RDP96-00791 R000300030006-7 Approved For Release 2001/06/09 : CIA-RDP96-00791 R000300030006-7 jungle below, dark cool moist atmosphere," a match both with regard to topography and ambience. In the second match the subject submitted that all she got was a "picture of Dr. Puthoff sitting in a beach chair by a pool," which was entirely correct. During the course of the Central America experiment, on one occasion when the test subject was unavailable, an experimenter volunteered a drawing of an image he obtained at the beginning of one of the daily ex- periments. (The target for that day was an airport, an unexpected target associated with a side excursion at midpoint of the week's activity.) The match was good, as shown in Figures 5 and.6. The transcript data will be examined further to determine =ine structure, resolution, etc. (c) Local Targets with Feedback In this series of experiments, designed to give immediate data to experimenters, a subject is asked to take part in a remote viewing experiment under the following conditions. The subject and two experimenters (one of vhom was R.T.) are in a first floor laboratory in building 30 at SRI. A second experimenter (H.P.) leaves the area and proceods to a remote location of his choosing. None of the experimenters with the subject knew of the remote target location. H.P. and R.T. are in two-way radio communication via walkie-talkie, (a) to provide the experimenter at the target location real-time data and (b) to give the subject immediate feedback after he has made his assessment of the target. Approved For Release 2001/06/09 : CIA-RDP96-00791 R000300030006-7 Approved For Release 2001/06/09 : CIA-RDP96-00791 R000300030006-7 co d QW Y Q W W Q~ ,q`I lily !~i li f 1 }l~~. i Q LL ?- -. Q Approved For Release 2001/06/09: CIA-RDP96-00791 R000300030006-7 Approved For Release 2001/06/09 : CIA-RDP96-00791 R000300030006-7 -? 11.2 /7. FIGURE: 6 SKETCH PRODUCED BY SUBJECT FROM SAN ANDRES, COLOMBIA, AIRPORT USED AS REMOTE VIEWING TARGET Approved For Release 2001/06/09 : CIA-RDP96-00791 R000300030006-7 Approved For Release 2001/06/09 : CIA-RDP96-00791 R000300030006-7 By this means the subject has an opportunity to learn to separate real from imagined images. This is not considered to be a demonstration-of -ability test, but rather a 'training step on a gradient scale of ability. In many of these experiments we monitor physiological correlates as discussed in Section B.3 W. (Nine of these experiments have been completed to date, seven with the measurements of physiological correlates.) The following is a sample of an experiment with Subject 4. In this experiment we monitored physiolocial. correlates of the remote viewing activity. As is apparent in the following text, the subject initially had only a fragmentary picture of the remote site, but with what we judge to be a small amount of feedback, the subject was able to put images together into a correct description. Accompanying the verbal description presented below is a pho7-ograph of the actual scene at the remote location (Fig. 7). he e%)erimen er with the subject (R0T,) was, as always, kept ignorant of the target location o prevent guidance in the questioning. The capital letters signify walkie-talkie communication. LT. t is now _5:35. S-4: ...very strong diagonal.... like a zigzag that goes this way, vertically. R.T.: 5-4's FIRST IMPRESSION IS OF A VERY STRONG DIAGONAL ZIGZAG THAT'S GO G VERTICALLY. OVER, (Talking on walkie talkie to H.P.) H.P.: TIRE IS A STRONG ZIGZAG AT MY PLACE, BUT IT IS NOT VERTICAL BUT RATHER HORIZONTAL; BUT IF SHE IS LOOKING FROM THE AIR, THAT'S EXACTLY WHAT IT WOULD LOOK LIRE . OVER. R.T.: Can you tell what the zigzag is attached to? Whether it's part of a building or a fence on the ground? Approved For Release 2001/06/09 : CIA-RDP96-00791 R000300030006-7 Approved For Release 2001/06/09 : CIA-RDP96-00791 R000300030006-7 R.T.: It's 12:41. S-4: My head gets in the way now that he's said that it's horizontal. I usually think of a fence. R.T.: Why don't you go up and look down and view the whole thing from above and see if you can get the whole gestalt of where he is. S-4: .....definitely a non-yegetation...almost no vegetation around. It's mostly concrete and whatever that zigzag is--either water or steel--shiny, zigzag,..definitely shiny. R.T.: 7267, THE ZIGZAG IS A SHINY THING WHETHER IT'S STEEL OR S-4: Water.. R.T.: WATER, WE CAN'T TELL. IT'S SHINY AND THERE'S VERY LITTLE VEGETATION ENO VEGETATION AROUND.... S-4: Mostly concrete. . . R.T.: IT'S MOSTLY CONCRETE... S-4: He's standing on concrete.... R.T.: YOU'RE STANDING ON CONCRETE. OVER. H.P.: IT CERTAINLY IS TRUE THAT THIS IS SHINY AND IN MY NEAR VICLNITY IT IS BARREN AND CONCRETE OR CONCRETE-COLORED EARTH. SHE SAID THAT IT LOOKED LIKE STEEL 'OR WATER. CAN SHE MAKE THE DIFFERENTIATION BETWEEN THE TWO? R.T.: He wants to know whether it looks more like steel or water. S-4: It seems to have movement--that's why I would deduce that it's water. R.T.: What if you try to look at the whole thing. S-4: I'm trying to get an eagle's eye view. That's a waterworks. R.T.: Why does it look like a waterworks? In what way? S-4: There seems to be a man-made layout of channels and connections to conduct Approved For Release 2001/06/09 : CIA-RDP96-00791 R000300030006-7 Approved For Release 2001/06/09 : CIA-RDP96-00791 R000300030006-7 R.T.: S-4 SEES MOVEMENT IN THE ZIGZAG THING, SO SHE THINKS THAT IT'S WATER, AND A KIND OF LAYOUT OF CHANNELS AS THOUGH IT WAS A MANY -MADE WATERWORKS WITH WATER RUNNING IN ZIGZAG CHANNELS. OVER. H.P. THAT IS PRECISELY CORRECT. IT IS A ZIGZAG MAN-MADE WATER CHANNEL WITH CONCRETE SIDES. OVER. The above is an excerpt from an early experiment, and is typical, rather" a sample of exceptionally good quality. That experiment continued with four more site idescriptions, three of which were of equal quality. One experiment of this nature has been'carried out with Subject 1, one with Subject 2, two with Subject 3, and five with Subject 4. A number of descriptions were essentially free of error and with no feedbacb other than verification following the remote viewer's description. i 1, A complete analysis is to be carried out on these transcripts following more experimentation. To date it appears that the viewing is weak in the following areas: (a) perspective and dimension are often distorted (an eight foot tower is taken to be 50 feet tall, a 20 foot separation between buildings may appear to be 100 feet, etc.) and, (b) written material generally cannot be read. RCS ,, (d) Local Targets with Azimuth Bearing In two remote viewing experiments, the second of which was r (~ clearly correct from a descriptive standpoint, an effort was made to determine whether in driving the subject around the area it would be possible to determine the location of the target team by triangulation with a bearing compass. The triangulation lines were essentially uncorrelated with each other and with the target location, and therefore provided a,p14i l result. Approved For Release 2001/06/0 ~4r6P96- 1 R00 Approved For Release 2001/06/09 CIA-RDP96-00791R000300030006-7 B. Basic Research yield data indicating the feasibility of the application of paranormal abilities to operational needs, fifty percent of the effort is devoted 1. Identification of measurable characteristics possessed by gifted individuals (20%); 2. Identification of neurophysiological correlates which relate to paranormal activities (20%); 3. Identification of the nature of paranormal phenomena and energy filled during the course of experimentation: 1) establish--.and apply criteria to differentiate between those for whom paranormal ability is considered to be functional and those for whom it is not; 2) obtain sufficient medical and psychological data to establish baseline profiles against which (a) one individual may be compared with another, and (b) an individual may be compared to himself at different times to determine whether para- normal functioning occurs in an altered neurophysiological state, 3) specific validation experiments must be conducted with sufficient control to ensure that all conventional communication paths are blocked, and with outcomes sufficiently unambiguous to determine whether paranormal functioning occurred; 4) obtain neurophysicological data during experimentation to Approved For Release 2001/06/09 :CIA-RDP96-00791R000300030 Approved For Release 2001106/.09-;..GIA-RDDR96-A97-94R009300630006-7 determine those correlates, if any, which relate to paranormal activity. In the following paragraphs, each of these items is considered in turn and the progress to date reported. The milestone chart for the basic re- search program is shown in Table 1. The work is progressing in accordance with the schedule prepared for this program, and the remaining time and funds are sufficient to meet all program objectives. 1. Criteria for the Determination of Gifted Individuals One of the key issues in the program is the establishment of criteria capable of differentiating individuals apparently gifted in paranormal functioning from those who are not. Three experimental paradigms were chosen to act as screening JI/ chance expectation). The tests are (a) remote viewing of natural targets, (b) reproduction of simple line drawings hidden from the subject but viewed by an experimenter, and (c) determination of the state of a four-state elec- tests on the basis that these tests had been useful for such purposes prior to this program ( in the sense that certain apparently gifted individuals did exceedingly well on at least one of the tests, wbe yeas the results of unselected volunteers did not differ significantly from tronic random stimulus generator. The first test constitutes a so-called "free-response" paradigm in which the subject originates freely about contents of his awareness; furthermore, the channel in general may involve both direct perception of the remote site and perception of the mental contents of an observer L Approved For Release 2UD-fT6OG-:-CIA-RDPg6-00791R000300030006-7 Cp V C In .~ W N N ro ID i cn r t 0 N C19 O G CD 1 n P. H P - N N ? 0 r .d Si It) Si i P . T.' N . w W N C 0 10 rr 0 Id 0 Si ? d CD CU N? H N H 0 O Si rt P) rD Cl) CA Ha H C E all C)) rr (DD a (D 03 0 (D (D rt 0 ~ G W 0 r H 'J p, 0 H Cs) rt (DD rr CD (D o It) . CD d F- rt r -'? 90 ' ' X rr O 0 C) ri ?' H. 09 (D H- d J CD CD' P' (D (D N n rn :3 0 P. In (M 09 (n Y? > F3 l 0 0 rat n F - rt n d ? 0 ? tt n H n G 0) F . CA C n rr p.. . / rr w aq rr W CD Id o H 0 1-' F6 F4 1 (D X (D H- n 1 ::j :j o 0 H 0 P (D ~ O 0 F-A n C=1 r r, rt tH to F~ (D n CD 1 rt fal (D N A F4 (D ;j ri (D ct cr N D C C) rD ti 1-4 # H Si N C1q F-~ n rt CD F + rt w 11 (D H- F'? rr Cl) 0 CD CD n0 0 (a H rr N rt Fs? 0 O Clq N f.'. CD CD Qq rn H n 4 1 were observed. Of these 63 events, 42 were distributed among the activity periods, 21 among the no-activity periods, a correlation significant at the p = 0.004 level. Subjects 2 and 6 also interacted.with the device. Although subject efforts and observed perturbations sometimes coincided, activity was generally low and did not appear to be a signature of correlated activity under control. ':A controlled ten-trial run with Subject 2 and two such runs with Subject 6 yielded non-significant results. Given the limited availability of the instrument and somewnat u- noisy environment, from our best effort we nonetheless for Subject 1 the observed number of precisely timed events in pilot work coupled with the statistically significant (p 0.004) correlation between subject effort and signal output in controlled runs indicate a * With the exception of the first run where 25-second trials were used. Approved For Release 2001/06/09 : CIA-RDP96-00791 R000300030006-7 Approved For Release 2001/06/09 : CIA-RDP96-00791 R000300030006-7 highly -probable cause-effect relationship. Thus it appears that a gifted subject can interact with a second derivative magnetic gradiometer of sensitivity- 10-9 Gauss/cm2 from a distance of four meters. Further work would be required to determine absolutely the precise nature of the interaction, although given the equipment design the generation of a magnetic field is the most probable mechanism. (c) Experiments with Laser-Monitored Torsion Pendulum In this series of experiments we examine the possibility that a subject may be able to exert a physical influence on a remotely located physical system. The target is a torsion pendulum suspended by a metal fiber inside a sealed glass bell jar. The pendulum consists of three 100 gram balls arranged symetrically at 1200 angles on a 2 cm radius. The entire apparatus is shock mounted, and protected from air currents by the enclosing bell jar. The angular position of the pendulum is measured by means of an optical readout system. The system consists of a laser beam from a lo. power argon laser reflected from a small mirror on the pendulum onto a position sensing silicon detectortt 1.5 meters from the pendulum. The detector yields an output voltage proportional to spot position. The output from the detector is monitored by a chart recorderttt which provides a continuous'sine wave record of pendulum position. a. Spectra Physics Model 262 tt United Detector Technology Model SC/10 ttt Brush Model Mark 200 Approved For Release 2001/06/09 : CIA-RDP96-00791 R000300030006-7 Approved For Release 2001/06/09 : CIA-RDP96-00791 R000300030006-7 The system exhibits a sensitivity of approximately 10 microradians. Under typical experimental conditions random accoustical fluctuations drive the pendulum in its torsional normal mode of 10 second period to a level - 100 microradians angular deviation. During control runs the pendulum executes harmonic motion with a maximum variation amplitude of ? 10 percent over an hour period. Sudden vibrational perturbations in the environment produce oscillation of the pendulum in the vertical plane at a frequency of 1 Hz, as contrasted with the torsional mode in the horizontal plane at 0.1 Hz. Approved For Release 2001/06/09 : CIA-RDP96-00791R000300030006-7 Approved For Release 2001/06/09 : CIA-RDP96-00791 R000300030006-7 The subject is asked, as a mental task, to affect the pendulum motion, the results of which would be available as feedback from the chart recorder The subject is then encouraged to work with the pendulum from a distance of 1 meter, observing effects being produced. If satisfied that there is a .possibility of producing effects (typically following.a week's activity, a couple of hours per day), an experiment is begun. As in other experiments, subject efforts to increase or decrease oscillation amplitude are determined by an experimenter utilizing the uni- versal randomization,protocol described in (a). Each experiment lasts one hour and consists of six 5-minute work periods alternated with six 5-minute rest periods. In later work, the subject is removed to a room 12 meters down the hall with three intervening office spaces to determine whether effects can be produced from a remote location. The subject is provided feedback at the remote location either by closed circuit video or by a second chart recorder in parallel with the recorder in the enclosed target laboratory. The remote aspect was instituted both to prevent artifactual effects from body heat, etc., and also to determine whether energy -can be coupled via the remote viewing channel to a remote location.} Both experimental evidence and theoretical work indicate that distance may not be a strong factor in paranormal phenomena. See, for example, E.H. Walker "Properties of Hidden Variables in Quantum Theory: Impli- cations for Paraphysics:', U.S. Army Ballistic Research Laboratories, Aberdeen Proving Ground, Maryland. Approved For Release 2001/06/09 CIA-RDP96-00791 R000300030006-7 il Approved For Release 2001/06/09 : CIA-RDP96-00791 R000300030006-7 In pilot studies we observed.considerable evidence indicating that a gifted subject located in the same room is able, by concentration, to increase or decrease pendulum motion on command while sitting quietly one meter from the bell. jar. The change-to-baseline ratio is often 5:1 or better so the effects are not small. A sample chart showing a rest period followed by a decrease period is given in Figure 14. Vibrational artifacts can be ruled out on the basis that when such inputs occur, a marked 1 Hz oscillation signal due to vertical motion is superimposed on the 0.1 Hz torsional motion. What is especially interesting are the decreases which take the motion below that generally observed due to en- vironmental noise driving. Such observations indicate the application of a constraint which couples energy out of the pendulum motion. Similar o servations have been observed with the subject removed to the second location 12 meters away. Although less pronounced (change-to-baseline ratios typically 2:1), the effect remains easily observable. The universal randomization protocol is used throughout to determine increase/decrease periods. Control run data are being collected to be sub- jected to the same analysis. Multiple recording is used throughout to rule out artifacts due to recorder effects. Finally, an electrometer with the base of the bell jar serving as one electrode.. is monitored to record acoustic vibration independently. Due to the potential significance of such findings; considerable data is being taken in order that the matter can be subjected to statistical analysis over a large sample involving hundreds of work periods. A few hundred data samples have already been collected for this purpose, an Approved For Release 2001/06/09 -GtA-RDP96-0fi76R90 030006=7 l -t-mi -~-'-"~-- Appr Approve i J ~1 1T1. gi- tL-`-- =a Approved For Release 2001/06/09 : CIA-RDP96-00791 R000300030006-7 (d) Experiments with Geiger Counter As part of a continuing search for mechanisms involved in para- normal phenomena, a series of experiments were conducted with Subject 1 to determine whether a geiger counter in the 'y - ray mode (i.e., beta shield in place) would register subject-directed efforts. The output of a geiger counter, fed into a Monsanto Model 1020 counter/timer, indicates a background count due to cosmic rays 35 counts/minute. Experimental protocol requires the subject to try to change the registered count by concentration on the geiger counter probe from a distance ^- 0.5 meters. Each run consists of 15 60-second trials, with 10-second separations between the trials. Preceding each subject run is a control run of equal duration. In four runs to date the results, shown in Table 6, indicate no effect of statistical sigificance, neither'the mean nor standard * OCDM[ Item No. CD V-700, Model No. 66, Electro-Neutronics, Inc., Approved For Release 2001/06/09: CIA-RDP96-00791R000300030006-7 Approved For Release 2001/06/09 : CIA-RDP96-00791 R000300030006-7 TABLE 6 Geiger Counter Experiment Control Runs Ex erimental Runs Run Mean Standard Deviation Mean tanclagri gDegirLfj 1 36.07 5.73 35.33 6.00 2 34.87 6.23 33.87 7.27 3 33.87 5.88 34.00 5.25 4 35.20 5.09 35.67 5.77 Approved For Release 2001/06/09 : CIA-RDP96-00791 R000300030006-7 Approved For Release 2001/06/09 : CIA-RDP96-00791 R000300030006-7 5. Basic Research Summary The basic research program to date has been spread over a number of subjects and over number of activities, generating a considerable amount of data. It was deemed desirable in the first half of the research program to cover as much material as possible in a horizontal development in order to determine the best subjects and the fruitful directions for concentrated effort in the second half of the program. We intend to concentrate on analysis of the large amounts of data already obtained while subjects are involved in extramural medical and psy- chological testing. Based on the findings, a few carefully-chosen items will be culled for final specific testing following discussion with client representatives. Approved For Release 2001/06/09 : CIA-RDP96-00791 R000300030006-7 Approved For Release 2001/06/09 : CIA-RDP96-00791 R000300030006-7 Randomness Tests of Four-State Electronic Random Stimulus Generator The design objective was to build a four-state machine, with each state equally likely to occur on each trial, independent of the past sequence of states. If the machine meets this objective, it should not be possible to devise.a rule for future play that significantly differs from chance. A simple example of such a rule would be to select the machine state observed in the preceding trial; if this strategy were to produce scores significantly above chance (25 percent hits), we would reject the hypothesis of randomness of the machine under test. Before experimentation, four machines, purchased from Aquarius Electronics, Albion, California, were extensively tested for randomness. Data were analyzed on a CDC-6400 computer, and three machines finally selected for use in screening met established criteria for randomness. In developing randomness tests, we:-are guided in part by a knowledge of the machine logic. When one of the four choice keys or the pass key is depressed, the current machine state is displayed; then a brief time after release of the key, a new machine state is established (but not shown to the subject) by sampling the instantaneous state of a high-speed four-state elec- tronic counter. For the machine to be random., the times of dwell of the counter in each of the four states must be precisely equal; otherwise, the distribution of outcomes will be biased. The first randomness test is thus based on tallying the number of occurrences of each of the four states. This Approved For Release 2001/06/09 : CIA-RDP96-00791 R000300030006-7 tePISRMfjg(f R'e@ ?P gQ2AkQ/Q$0;sCI p~i9?-( 7 ;1 F~0@(,2? 4Q9 0~~7s . To test for .this second possibility we also tally the distribution of outcomes in each group of 100 trials, then compute a likelihood ratio test statistic (see below) for each group. Under the null hypothesis of equal likelihood of the four states, these statistic values are distributed approximately as chi-square with three degrees of freedom and their sum for m groups distributed, approxi- mately as chi-square with three m degrees of freedom. This test may also detect stable bias, but is not as powerful for this purpose as the first test. Variable bias of still shorter period, if substantial, can be tested for by tallying the frequency with which the previous machine state is repeated; an overall repeat ratio ("ell") significantly above 0.25 is indicative of such bias. If for any. reason the machine were to fail to sample the counter to establish a new state, the previous machine state would be repeated. To test for this possibility, we tally the number of repeats following the depression of each key. A repeat ratio significantly greater than 0.25 should be con- sidered a danger signal. We also tally the initial machine states following reset and the tran- sitions between states., In each case, the number of occurrences of each of the four possible outcomes should be approximately equal. When repeats are deleted from the sequence of trials ("nondiagonal transitions"), the four states should also be approximately equal infrequency. In testing the null hypothesis of four equally likely outcomes of a trial, a likelihood ratio test is used. The statistic a- Alexander Mood, Introduction to the Theory of Statistics (McGraw Hill, New York, 1950). Approved For Release 2001/06/09 : CIA-RDP96-00791 R000300030006-7 Approved For Release 2001/06/09 : CIA-RDP96-00791 R000300030006-7 4 n/4 ni #n ( n i=1 under the null hypothesis is distributed approximately as chi-square with three degrees of freedom,'with rejection for large.values of this statis- tic. The computer program used in testing randomness includes a subroutine for computing the probability of a chi-square value as large or larger than that observed. In testing the null hypothesis that the probability of a repeat is 0.25, the binomial probability of obtaining the observed number K or more repeats in N trials is computed. For K greater than 1000, a normal distribution approximation is computed, assuming the statistic K-1/2 N ( N - 0.25 3/16 to be approximately normal with mean zero and standard deviation one. The typical test pattern used was six passes followed by twenty-five choices of one color, repeating this for each of the four colors. In this way each of the five keys other than rest were given approximately equal use. Typically, 2,000 to 6,000 trials were made; in each sitting. In the absence of any unusual results in the randomness tests, a minimum of 10,000 trials were made before using a machine with experimental subjects. With 10,000 trials, the expected fraction of re- peats is 0.25 with a standard deviation of 3/200 = 0.00866. A sample computer listing of the results of randomness tests on Machine 4 is included in Table 1. Of the four machines tested, three were found suitable for use in screening activity. The fourth machine was returned to the manufacturer for adjustment. Approved For Release 2001/06/09 :CIA-RDP96-00791 R000300030006-7 Approved For Release 2001/06/09 : CIA-RDP96-00791 R000300030006-7 Number of Trials Chi-Sq. Binom. Prob. 1.996 728 764 765 790 3047 2.573 0.46 777 784 773 863 3197 6.745 0.08 776 796 '810 773 3155 1.158 0.76 787 852 803 805 3247 2.877 0.41 Nondiagonal transitions Diagonal Binomial Prob. transitions Yellow 2774 705 0.2541 0.313 Green 2755 674 0.2446 0.748 Blue 2761. 706 0.2557 0.250 Red 2742 667 0.2433 0.793 1614 0.2323 0.953 12646 3127 0.2473 0.763 Randomness in groups of 100 trials: Chi-sq. = 299.6141 D.F. = 345 Prob. = 0.9628 Approved For Release 2001/06/09 : CIA-RDP96-00791 R000300030006-7 Approved For Release 2001/06/09 : CIA-RDP96-00791 R000300030006-7 PERSONAL OBSERVATIONS ON THE USE OF THE FOUR-STATE ELECTRONIC RANDOM STIMULUS GENERATOR * The following notes are based solely upon my experience and I there- fore make no claim that they are generalizable to other persons. Since I am still learning about ESP phenomena, I am confident that additional work in this area will expand, modify, and refine the perceptual processes discussed below. While I have tried to describe these experiential processes with as much precision as possible, the use of seemingly precise language should not leave the impression that the perceptions themselves were equally precise. To the contrary, I found these perceptions to be delicate, transient and ephemeral--and yet, at the same time--and somewhat surprisingly--unmistakably real. 1. Perceptual Processes Working with the ESP machine proved to be a venture into unfamiliar perceptual territory which functioned according to new and different rules. It took some time (five hours or so with the ESP machine) to begin to learn not only which perceptual processes would work but, equally important which would not work. There was clearly a learning process in finding those delicate and subtle internal cues that would allow me to make perceptually based choices. After approximately 1000 trials with the ESP machine, five dominant perceptual modes emerged. Subsequent * Prepared by a policy research analyst at SRI, who was a high-scoring subject (p < 10 -6 with the four-state electronic random stimulus generator. Approved For Release 2001/06/09 : CIA-RDP96-00791 R000300030006-7 Approved For Release 2001/06/09 : CIA-RDP96-00791 R000300030006-7 work with the machine seemed to essentially, expand and refine these perceptual processes that emerged initially. Direct Knowing (Used approximately 5 to 15 percent of the time)--This perceptual cue came as a "gift" that I did not have to, work for. This is not to say that this "cue" was always right, but when there was a direct perception of the appropriate response unmediated by any of the other cues described below, my chances of being right seemed quite high (say 75 percent of the time). Internally,'this was simply the feeling that I should push one specific button and the knowing was almost immediate. If it were not immediate then, typically, one of the other cues would be used. "Closure Cues" (Used perhaps 75 percent of the time)--This cue manifested itself in a variety of ways; a sense of "fullness" with respect to a particular button, an internal anticipation of the bell ringing, a sense of "hardness"' or "firmness" and a sense of being "locked into" the correct response:. The validity of this cue could be tested by acting and thinking as if I were going to push a particular button and then noting the extent to which these "closure cues" became present. This sense of active intentionality--both physically and psychologically-- seems important in that it allowed me to sort out many real from imagined perceptions. Also, this cue often gave a kind of veto power; i.e., it did not necessarily assure me as to the right answer but it would tend to tell me if I had picked the wrong one, the aforementioned cues. i.e., I would not experience Pattern Recognition (Negligible use initially, but then used approximately 75 percent of the time during Phase IV)--Although I used this perceptual mode very infrequently during the initial stages of the Approved For Release 2001/06/09 : CIA-RDP96-00791 R000300030006-7 Approved For Release 2001/06/09 : CIA-RDP96-00791 R000300030006-7 experiment, it emerged rather naturally toward the end. This was similar to the "direct knowing" but not isolated to a single button; rather, there was a sense of the next two to three buttons that would be the correct responses. These perceptual cues were obtained in a less objective/ rational way and in more of a meditative state, highly concentrated but without specific focus on a particular button. Interestingly, in using this perceptual process, I was able to go somewhat faster and have greater access to all of the buttons in an equivalent way (see the second point under Section 2 next page). Thus, this mode had the advantage of loosening habituated perceptual patterns but it also made selections less amenable to conscious control and testing. This process proved to be either highly accurate or highly inaccurate. Accuracy seemed to be a function of the degree to which I could become synchronized with the evolving pattern of machine selected choices--and it was easy to get out of phase/sequence with this pattern. Rational Guessing (Used approximately=5 percent of the time)--Although I virtually never did try to superimpose some rationally predicted pattern upon the random, machine selection of buttons, I would sometimes temper my selections (very seldom for the better) by noting that one button had come up too often for it to be likely on the next trial or, conversely, it had come up so seldom that it should be given special consideration as a likely possibility on the next trial. Again, although this was a tempting strategy, I found that random processes were not amenable to rational anticipations and my rational guesses seemed often to be wrong. Tension/Vector Analysis (Used approximately 75 percent of the time)-- Here the cue was manifested as a sense of tension(s) pulling in one direction or another with the selection buttons as the locus for that tension. The cue was also manifested as a feeling of "emptiness" and Approved For Release 2001/06/09 : CIA-RDP96-00791 R000300030006-7 Approved For Release 2001/06/09 : CIA-RDP96-00791 R000300030006-7 conversely as a sense of it fullness." To describe this process further, it felt analogous to vector analysis in physics where, in sorting out competing tugs and pulls, one finds the "dominant" vector; i.e., the one with the strongest "pull" or the one that best "balances" the other vector tensions. Figure A-1 illustrates this phenomenon. Although the tension/vector cues were very useful and among the most reliable of all the cues, I found them to be at times quite mis- leading. The source of confusion stemmed from the role of time as a variable rather than a constant in extrasensory reality (discussed in more detail under section "Comments on Perceptual Processes"). If my assumptions as to the temporal nature of my perceptions did not fit with the actual nature of those perceptions. then the perceptions were quite misleading. (Recall that precognition refers here to a button that will be selected in the future--typically the next trial). The nine-cell matrix shown in Figure A-2 may clarify the complexity of the perceptual process, the need for discriminating awareness and the possibility for error. Out of nine possible combinations of the assumed/actual nature of perceptions, only three are matched or congruent and yield accurate understandings. Each of these primary cases is discussed below: ? Clairvoyant--Here the feeling which allows sorting and selection is like that described in Figure A-1. ? Precognitive--The feeling, sorting, and selection is like that described in Figure A-1 with clairvoyance; the primary difference being a shift in the time dimension to refer, not to the present target of the machine, but to the one to be selected next. To act on this perception I would press the pass button to bring the future into the present and then press the button that corresponded to my precognitive per- ceptions. ? Clairvoyant and Precognitive--The perception is of. a pattern of buttons, distributed through time, that are and will be selected by the machine--the "pattern" usually consisted of two to three buttons. Again, the time variable was most Approved For Release 2001/06/09 : CIA-RDP96-00791 R000300030006-7 Approved For Release 2001/06/09 : CIA-RDP96-00791 R000300030006-7 LIGHT PULL/ UNBALANCED FIGURE A-1 ILLUSTRATION OF TENSION/VECTOR ANALYSIS IN OPERATION With Button C being the one selected using these cues. ACTUAL NATURE OF PERCEPTIONS ASSUMED NATURE OF PERCEPTIONS Clairvoyant Clairvoyant Precognitive and Precognitive Clairvoyant Correct Misperception Misperception Perception Precognitive Misperception Correct Perception Misperception Clairvoyant and Misperception Misperception Correct Precognitive Perception FIGURE A-2 MATRIX SHOWING CORRECT PERCEPTION AND MISPERCEPTION IN THE USE OF TENSION/VECTOR CUES VIA THE INTERFACE BETWEEN ASSUMED AND ACTUAL NATURE OF PERCEPTIONS Approved For Release 2001/06/09 : CIA-RDP96-00791 R000300030006-7 Approved For Release 2001/06/09 : CIA-RDP96-00791 R000300030006-7 troublesome--typically with greater difficulty in determining the order in which the buttons would appear as targets and lesser difficulty in determining which buttons were targets. Confusion and error would arise when I assumed the tension/vector perceptions were clairvoyant when in fact they were (say) clairvoyant and precognitive. To explain how this felt, refer back to Figure A-l. If the actual sequence of correct answers-were Buttons B and D, and if I were assuming the perceptions were clairvoyant only, then it was not uncommon to have the perception that the intervening button (C) was the correct choice. The rationale for this perception was that it felt like a balance point between Buttons B (present target) and D (next target). In retrospect, when I am more rationally aware of the room for error in the use of this cue mechanism, I am somewhat surprised as to how useful it was in operation. It should be clear from the preceding descriptions that selections were made by a variety of processes which were used sometimes in isolation and oftentimes in combination. A typical sequence in the selection process was: (1) Check for "direct knowing" cues, if not there, then (2) Use "tension/vector" cues, then (3) Make final selection with "closure cues. t' ?2. Comments on Perceptual Processes Rather than work rapidly, I chose to work, deliberately, consciously, and therefore slowly. I would typically take five to thirty seconds to select a button--enough time to have a firm and conscious sense of my internal cues and what I thought they meant. The typical sequence would be as follows: Approved For Release. 2001/06/09 : CIA-RDP96-00791 R000300030006-7 Approved For Release 2001/06/09 : CIA-RDP96-00791 R000300030006-7 ? Observe various cues ? Rationally interact with cues to sort them out ? Select a button and press it ? Integrate feedback from response ? Clear mind and become quiet. Except during "pattern recognition," when all buttons seemed equally accessible, I found that the top two buttons on the machine were much more accessible than the bottom two. Three plausible explanations emerge to account for this. First (and least likely I think) is a psychological predisposition against the bottom two buttons--perhaps because of the i color of the buttons or because of the pictures associated with the tar- gets. Second is the possibility that the circuitry of the ESP machine in some way favors the top two buttons or obscures the bottom two. Third (and most plausible to me) is the possibility that to the extent I used the "tension/vector" cue, then the bottom two buttons would be without a vector below them--making it more difficult to "bracket" the bottom two buttons with this perceptual process. In later phases of the experiment, I was more able to access the bottom two buttons and this seemed to cor- respond with increasing use of the "pattern recognition" cues and the decreasing use of tension/vector cues. The longer I worked with the ESP machine, the more apparent it be- came that, in an extrasensory perception reality, time becomes fluid. In other words, although the experiment was designed to test clairvoyance (selecting the current target) only, I found that the perceptual cues would oftentimes be equally applicable to precognition (selecting a future target--usually the next one). Therefore, making a correct selection required doing two things; first, finding the correct "pattern" of buttons that would be randomly selected by the machine (typically the pattern consisted of two to three buttons) and second, associating a time component Approved For Release 2001/06/09 : CIA-RDP96-00791 R000300030006-7 Approved For Release 2001/06/09 : CIA-RDP96-00791 R000300030006-7 with the buttons in that pattern. Stated differently, the same cues discussed above held equally well for precognition or for clairvoyance-- so the problem of making a selection was compounded by the additional difficulty of having to determine whether a perceptual cue was associated with the button that had already been selected by the machine or the button that would be selected in the next or even subsequent trial. I definitely felt that if I could consistently separate clairvoyant from precognitive dimensions of identical cues, that I could substantially increase the accuracy of overall scores. The cues were not always consistent in their presence and meaning. For example, I might be obtaining good results with the use of tension/ vector cues and then find them becoming ambiguous, with a commensurate decline in my score. Then I would rely more heavily upon other cues. Or, the cues might work well for clairvoyant perceptions for a while but then shift to operate for precognition--then I would-have to "recalibrate" myself to the cue mechanisms. So, it was a fluid, dynamic perceptual process which required flexibility and patience. Highly significant scores and perceptions seemed to go in spurts of ten trials or so, then I would fall back to a chance level until I could resynchronize myself ith the machine and the character of my perceptual cues. I tend to agree with the notion that it might be more appropriate to call these processes "extraconceptual perception" rather than "extra- sensory perception." The perceptual cues were definitely present and they had sensory dimensions even though they do not fit into our traditional sensory categories. Just "where" and "how" these sensory cues were present is not clear to me--but these are essentially conceptual rather than sensory issues. Approved For Release 2001/06/09 : CIA-RDP96-00791 R000300030006-7 Approved For Release 2001/06/09 : CIA-RDP96-00791 R000300030006-7 3. Problems in Perceptual Translation A basic problem in using the ESP machine was not so much the obtaining of perceptual data as the translating of those data into sufficient information to allow the action of selecting the correct button. While the act itself is so simple as to be trivial, the information processes (gathering, filtering, dynamically translating) underlying that act seemed to me very substantial. It is within this unseen and unrecorded portion of the ESP testing process that most of the "action" takes place. From this vantage point I would like to suggest two impediments that might partially account for relatively low scores. First, I am still not fluent in the "language" of extrasensory perceptions--analogously, it is like hearing many separate commands in Russian (or another unfamiliar language), each time spoken in slightly different ways and with different intonations and inflections. The call for action may be clearly heard but the translation of that command into operational reality is an imprecise process until the language can be better understood. Second is the problem created by shifting back and forth between rational and intuitive knowledge processes during the course of the experiment. In selecting a single button I would use intuitive knowledge processes for perception and oftentimes, rational or semirational knowledge processes to inJ-~rpret those perceptions. This is not to say that the rational component is absolutely necessary, but it did seem to be useful for me. In any event, since the experiment covers thousands of trials (button selections) it required thousands of translations from one knowledge mode to another. Although the rational mode did seem helpful for inter- Approved For Release 2001/06/09 : CIA-RDP96-00791 R000300030006-7 Approved For Release 2001/06/09 : CIA-RDP96-00791 R000300030006-7 pretation, it was also I 'costly" (i.e., by shifting to a rational mode) I could be thrown slightly off-balance in maintaining contact with the subtle and delicate intuitive processes--thereby introducing an additional element of ambiguity and error). Related to the problem of differential knowledge processes is the problem of having to translate between states of consciousness in order to act upon extrasensory perceptions. LeShan* analyzed the experiential properties of what he has termed Clairvoyant Reality and found that while certain events (such as telepathy, precognition, and clairvoyance) are ""normal" to this reality, certain other events (such as being able to take directed action toward a goal) are "paranormal." For me this was manifested experientially as the feeling that when I obtain extrasensory perceptions, I am so much a part of, and immersed in the Clairvoyant Reality that in order to act, I must causally separate myself from the Clairvoyant Reality and enter the dualistic, subject/object Reality that LeShan terms "Sensory Reality." Encouragingly, the "pattern recognition" process seemed to offer a means of both perception and action, which did not require the same degree of transfer between these subtly different states of con- sciousness. The preceding points suggest that one difficulty in testing and assessing extrasensory perception may be the apparent need to translate it into an output that is not isomorphic with the perceptions themselves-- a person must translate the perceptual "language" to a familiar form, across rational and intuitive dimensions, and relatedly, from one state of awareness to another. Is it possible, then, that our means for testing * Lawrence LeShan, The Medium, The Mystic, and the Physicist (Viking Press, New York, 1974). Approved For Release 2001/06/09 : CIA-RDP96-00791 R000300030006-7 Approved For Release 2001/06/09 : CIA-RDP96-00791 R000300030006-7 ESP may not be highly congruent with the nature of the phenomenon, and this may inherently reduce the significance of the test results that can be obtained? 4. Two Views of the ESP Process I suspect that, to an external observer, my work with the ESP machine might appear as fairly consistent scoring slightly above chance--the logical inference could then be made that a small. amount of extrasensory perception was mixed with a substantial amount of pure guessing.' While the scoring data may support this inference, my awareness of the input process does not. Consider the following: on the first run, a person could get six "hits" out of twenty-five by pushing buttons at random; then on the second run, he could get six "hits" out of twenty-five by using extrasensory perception. To the statistician who looks only at the output, the scores are identical--they are no more than would occur by chance-- and the logical inference would be that the input processes were identical or at least very similar. However alike-they might appear externally, internally they could feel like quite different runs. In the second instance, the chance level of scoring would be the result of an imperfect but operative extrasensory perception process. Obviously, then, measure- ment of ESP by statistical output alone obscures the nature and extent of the extrasensory input. A relatively modest score on the ESP machine can--I think--substantially understate the amount of learning and perceptual development that actually occurs. The foregoing is consistent with my impression that my scores, though statistically significant, still did not reflect the actual amount of learning that had occurred. Approved For Release 2001/06/09 : CIA-RDP96-00791 R000300030006-7 Approved For Release 2001/06/09 : CIA-RDP96-00791 R000300030006-7 5. Supportive Mind Set There emerged, after a time, what seemed to be a series of pre- conditions to good performance in terms of mind set. These were: A high level of motivation seemed essential. The task of pushing one of four buttons over thousands of trials could be rather boring--enough to allow one's attention to wander. With each trial, it was necessary to have a high level of motivation to ensure adequate levels of concentration and focused attention. ? Although motivation, concentration, and attention were important, it was also necessary not to be too concerned with the success or failure associated with each selection. If I became "attached" to the outcome of a previous trial, whether a success or a failure, it could divert a significant amount of attention from the present trial. Therefore, each trial must be separate/fresh/clear/unconditioned by the actual success or failure of previous trials and separate from the imagined successes or failures of upcoming trials. ? A relatively stable, undisturbed emotional state also seems important. I noticed the most substantial fluctuation in my scores when I was emotionally stressed (angry, hassled, and so on). ? Feeling rested physically also seemed important. This was particularly true if I were to work with the machine for an hour or two--as this required a substantial amount of energy. ? A positive attitude--a feeling that I could do well and could always score at least at the chance level--was also important. A corollary to this was that I found I did better when I "always liked myself" even if I did poorly. Self-deprecation seemed to be a sure way of rapidly diminishing the accuracy of the perceptual processes. 6. The Environment There were attributes of the surrounding environment that seemed to enhance the accuracy of my selections. The more significant factors seemed to be the following: Approved For Release 2001/06/09 : CIA-RDP96-00791 R000300030006-7 Approved For Release 2001/06/09 : CIA-RDP96-00791 R000300030006-7 ? It was helpful to have a relatively quiet working environ- ment. Or, if there were noises, to have them of a sort-- fairly constant ones that remained in the background--that could be readily filtered out of my consciousness. My impression was that external sensory information--particularly sounds--could readily overload/override subtle and delicate internal sensory information. ? It also seemed to help to have low light levels--I would always turn out the overhead lights in the testing room. I experimented with closing my eyes to further reduce external sensory stimulation and I found that this would increase the sensitivity of sensory cues, but this increase in sensitivity was offset by a lack of visually based feedback to verify the accuracy of the selections. As a consequence, I chose to keep my eyes open. ? I found it essential to work with the ESP machine by sitting somewhat above it so that I could look down on the face of the machine. For some reason, perceptual discrimination seemed much more difficult when I would sit at a lower level which placed the buttons in a plane more nearly horizontal to my .face and upper body. 7. Transferability of Processes The perceptual learning gained in this experiment seemed generally transferable to other situations where I might use ESP abilities, in particular, telepathy, precognition, and clairvoyance. The inference is that a process or faculty is being developed which has numerous appli- cations in other situations which would rely upon ESP. Analogously, just as jogging could exercise muscles to make a person more adept at playing football, dancing, swimming, and the like, the use and development of these It psychic" muscles seems to have some degree of transference to Approved For Release 2001/06/09 : CIA-RDP96-00791 R000300030006-7 Approved For Release 2001/06/09 : CIA-RDP96-00791 R000300030006-7 8. Conclusions I found the experiment to be a very substantial learning experience in which, I feel, I learned much more than was reflected in the scores. It allowed me to begin to identify an ability which I presume was largely latent within--never having had a prior opportunity for overt expression. Finally, it suggests to me that this must be a common ability among many people that they simply do not recognize--primarily because they have never had the opportunity to explore it as a legitimate and "real" i phenomenon. Approved For Release 2001/06/09 : CIA-RDP96-00791 R000300030006-7 Approved For Release 2001/06/09 : CIA-RDP96-00791 R000300030006-7 SAMPLE OF RAW DATA - MEDICAL EVALUATION OF SUBJECT 1 Approved For Release 2001/06/09 : CIA-RDP96-00791 R000300030006-7 SGFOIA3 Approved For Release 2001/06/09 : CIA-RDP96-00791 R000300030006-7 Next 14 Page(s) In Document Exempt Approved For Release 2001/06/09 : CIA-RDP96-00791 R000300030006-7