R.G. Jahn, B.J. Dunne and R.D. Nelson- Engineering Anomalies Research

8/3/2019 R.G. Jahn, B.J. Dunne and R.D. Nelson- Engineering Anomalies Research

1/30

Journal of Scientif ic Explorat ion. Vol.1. No. I. pp. 21-50. 1987 0892-3310/87 $3.00+.00Pergamon Journals Ltd.Printed in the USA. 1987 Society fo r Scientific Exploration

Engineering Anomalies ResearchR . G . J A H N . B. J . D U N N E , and R. D. N E L S O N

School o f Engineering/Applied Science. P rinceton U n i v e r s i t y . Princeton, N J 0 8 5 4 4

Abstrac tAnomalous consciousness-related phenomena of possible rele-vance to basic physical science and modernengineering practiceare addressedexperimentally an d theoretically in an effort to identify those devices, systems,an d processes most likely to display operator-related anomalies in their per-formance, and to illuminate the characteristics of such aberrations. Threeinterrelated sectors of effort are pursued: th e design, implementation, oper-ation, and interpretation of experiments in low-level psychokinesis; the de-velopment of analytical methodologies fo r quantitative assessment of pre-cognitive remote perception data:and the development of theoretical modelsuseful fo r correlation of the experimental data, design of better experiments,an d explicationof the phenomena on fundamental grounds.The primary effect observed in the psychokinesis experiments is a marginalbut replicableshift of the mean of output count distributionswith respect toempirical baselines or theoretical expectations, with no discernible alterationsin an y higher moments. Over large data bases, these mean shifts can com-poun d with considerable statistical regularity to high levels of significance,depending on the particular operator, th e direction o f effort, and other pre-vailing experimental conditions. In many cases, individual operator "sig-natures" of achievement are found to transfer across various experimentaldevices, including some driven by deterministic pseudo-random sources.Quantitative analysis of a large data base of remote perception experimentsreveals similar departures from chance expectation of the degree of targetinformation acquired by anomalous means. Digital scoring techniques basedon a spectrum of 30 binary descriptors, applied to all targets an d perceptionsin the experimental pool, consistently indicate acquisition of substantial top-ical and impressionistic information about remote geographical locationsinaccessible by known sensory channels. The degree of such anomalous in-formation acquisition appears independent of the spatial separation of thepercipient from the target, up to global distances, and also independent ofthe temporal separation of the perception effort from the time of target spec-ification by the agent, up to periods of precognition or retrocognition ofseveral days.In an attempt to illuminate these empirical results, a theoretical modelhas been proposed that invokes quantum mechanical metaphors to describethe interaction of consciousness with its environment. B y representing con-sciousness by quantum mechanical wave functions and its physical environ-ment by appropriate potential energy profiles, Schrdinger wave mechanics

The Princeton Engineering Anomalies Research program has been supported over the pasteight years by grants from The McDonnell Foundation, The John E. Fetzer Foundation, Inc.,The Ohrstrom Foundation, Helix Investments, Ltd.,The Pillsbury Corporation, The ExplorersClub, The Institute for Noetic Sciences, The International Foundation for Survival Research,Inc., and by several individual gifts.21


2/30

22 R. G. Jahn et al.may be used to define eigenfunctions and eigenvalues indicative of psycho-logical and physical experience, both normal an d anomalous, in a form ap-plicable to the experimental designs.The experimental results in hand, along with th e generic predictions ofth e theoretical model, suggest numerous short and longer term practical ap -plications of the phenomena, and raise basic issues about the role of con-sciousness in the establishment of reality.

IntroductionScholarly research into a broad range of anomalous consciousness-relatedphenomena o ver the past century has produced an array of p rovocative results,but none that can be regarded as fully convincing in the traditional scientificsense. N or has this research yielded sufficient empirical correlations to support

any existing category of theoretical model fo r description and comprehensionof such effects, let alone to refine such models to functional utility. Neverthe-less, the potential implications for many fields of human endeavor are suffi-ciently profound an d pervasive that efforts for demonstration an d resolutioncontinue in several disciplines. A m o n g these, the various fields of engineeringscience are neither im m une from the imp lications n or impotent to co ntributeto th e search. Throughout its three major domains of modern activitytheprocessing of energy, of materials, and of informationengineering engagesa multitude of p hysical devices, systems, an d situations that may be p otentiallyvulnerable to such anomalous interactions. In particular, those involving sen-sitive m a n / m a c h i n e interfaces, low-level signal processing units of the micro-processor genre, elaborate data-storage systems, devices utilizing random orpseudo-random noise sources, and very large-scale integrated circuits wouldseem to merit attention.The research reported herein consists of three components, conceptuallydistinct, but in practice interrelated. The first is an ensemble of experimentsin low-level psychok inesis the interaction of human consciousness with somephysical device, system, or process resulting in statistical behavior differentfrom that expected on the basis of known science. Th e second addresses theprocess of precognitive remote perceptionthe acquisition of informationabout geographical targets remote in distance and time and inaccessible byany kn o wn sensory means. The primary interest here is the development ofanalytical m ethodologies fo r quantitative determination of the degree of in-formation obtained by such processes. The third segment is an e f f o r t to developa theoretical model to support the experimental program and provide som einsight into th e basic nature of the p h e n o m e n a .Particular experiments have been selected fo r their immediate and longerterm relevance to the practice of modern engineering science, and for theiramenability to controlled and systematic laboratory study. Physical and tech-nical parameters are the primary concern throughout; systematic investigationof psychological or physiological correlates is secondary to the accumulation


3/30

Engineering anomalies research 2 3of very large data bases by a relatively small n u m b e r of participants. A ll op-erators are a n o n y m o u s volunteers, non e o f w h o m claims extraordinary abil-ities, and no screening, training, o r induction techniques are employed.

Psychokinetic Interactions With Random Physical SystemsThe several experiments in low-level psychokinesis (PK), although diversein character and scale, all embody some type o f random physical processwhose distribution is established empirically and, when possible, theoretically.Various hum an operators then attempt to distort those distributions in pre-stated directions. To guide them in that task, each experiment provides someform of feedback, usually a visual display, that tracks the degree of shift fromthe baseline distribution.One such experiment belongs to a genre of random event generators (REG)widely used in contemporary studies of this class of phenomena (Krippner,1977; Radin. May, & Thomson, 1985; Schmidt , 1970; Stanford, 1977). Theparticular device employed in this program is based upon a commercial m i-croelectronic noise source whose output is transcribed by appropriate circuitryinto a random train o f positive and negative pulses, suitable for sampling andco unting [Fig. 1 (a)]. For most formal experiments, the device is set to generate"trials" of 200 pulses each at a rate of 1000 per second, and to count anddisplay the nu m ber o f those pulses which conform to the regular alternation:

+ , , + , , + , , etc. Various display and recording units show the operatorthe results of the counting and insert them on-line into a digital database andcomputational system.Figure 1(b) shows the experimental arrangement as seen by the operator,who sits a few feet from the device and its suppo rting equipm ent, fo r example,a computer terminal, a strip printer, and various fail-safe counters that guar-antee th e integrity of the data. The operator attempts to influence th e processto produce a higher number o f counts (PK+) or a lower number of counts(PK -), or to generate a baseline (BL), in accordance with pre-recorded inten-tions. In the protocol followed for the largest subset of our data base, data aregenerated in "runs" of 50 trials, accum ulated in "sessions" comprising a min-imum of f i v e runs. While session lengths are left to the preference of theoperator, a complete experimental "series" requires a full 7500 trials, or 50runs in each of the three directions of intention. (A few early series consistedof 5000 trials, or 100 runs per intention.) To preclude any artifactual bias,the protocol requires the operator to intersperse sequences of each of th e threeintentions, PK+, PK -, and B L, with all other experimental conditions heldconstant.A n example of the type of data obtained in this experiment is shown inFig. 2(a) as a distribution of scores for some 5000 baseline trials (i.e., onemillion pulses, or bits) taken by one operator, superimposed on the theoreticalGaussian approximation to the appropriate binomial statistics. With reference


4/30

24 R . G. J ah n et al.

(b) REG EXPERIMENTAL A R R A N G E M E N TFig. 1. Random Event Generator.

to the same theoretical distribution, Fig. 2(b) displays the results o f the sameoperator's efforts to shift this distribution toward higher or lower numbers ofcounts over the same number of PK+ and PK- trials, and Fig. 2(c) shows thebest Gaussian fits to these data. The effects found in this experiment areusually confined to such marginal shifts of the mean of the distribution, withno perceptible changes in the standard deviation, higher moments , or othercharacteristics of the distribution.


5/30

Engineering anomalies research 25

Fig. 2. REG frequency-of-count distributions.


6/30

26 R . G . Jahn et al.To track the consistency of these small shifts of the mean and to displaytheir statistical significance as a function of the data base size, the accumulateddeviation of the counts f r o m the chance mean are graphed as a function ofthe number of trials processed. Figure 3 employs the same data as Fig. 2, insuch cumulative deviation plots for each of the three intentions of the operatorrelative to the theoretical mean. A ll three experimental traces display thestochastic variations to be expected in this sort of random process, but whereasthe baseline curve meanders close to the theoretical expectation, the PK+ andPK- traces display almost linear systematic deviations f r o m the chance meanthat compound to progressively larger values as the number of trials accu-mulates. The dashed parabolas are the loci of the f i v e percent chance expec-tation of reaching that accumulated deviation at that number of trials, andthe scale at the right indicates the range of terminal chance probabilities. Th eterminal values of the means of these PK+ and PK- data, 100.264 and 99.509

respectively, d i f f e r from chance expectation by several standard error units,with the composite achievement unlikely by chance to the order of 10-6 .Such cumulative deviation graphs are found to be quite operator specificand hence are referred to as "signatures." Figure 4 shows such signatures fora few of the many other operators working on this same experiment. Someoperators achieve PK results in only one direction, some in neither, some inboth, and some show inverted results. The PK+ and PK- achievement patternsf o r a given operator are typically asymmetrical, and are o f t e n f o u n d to bedependent on the conditions under w h i c h the operator is performing theexperiment, such as the pulse counting rate, whether each trial in the run isinitiated manually or automatically, or whether the operator chooses or israndomly assigned the direction of e f f o r t . For example, Fig. 5 displays thesensitivity of o ne operator's performance to the "volitional" and "instructed"modes of data generation. In the volitional mode, the operator chooses the

Fig. 3. REG cumulative deviations from theoretical mean.


7/30

Engineering anomalies research 2 7

Fig. 4. REG cumulative deviations f r o m theoretical mean: Various operators.direction of effort and completes f i v e runs (250 trials), or some multiple thereof,before changing the intention. In the instructed mode, a random numbergenerated before each run assigns the direction of effort. Note that in the case


8/30

28 R . G . Jahn et al.

Fig. 5. REG cumulative deviations from theoretical mean: Instructed/volitional (operator 55).

shown, the PK+ and PK - results are essentially reversed, th at is, those in theinstructed mode are opposite to the operator's intention. A complete graphi-cal and statistical compendium of op erator signatures an d their dependenceon such parameters is available in a technical report (Nelson, Dunne &Jahn, 1984).Despite these variations in individual operator performance and in theirsecondary dependence on experimental conditions, the overall R EG data basealso displays a significant statistical trend. Figure 6 and Table 1 show thecombined results of the entire formal data base, comprising 87 completedseries, and totalling over 250,000 trials per intention (>150 million bits).These data were generated by 33 different operators on two different machinesover a period of approximately seven years. Again, the grand baseline meanremains close to the theoretical value, and the PK + and PK- data, despite


9/30


Fig. 6. REG cumulative deviations f r o m theoretical mean: A ll data, 33 operators.occasional reversals associated with particular operators and conditions, trendtoward increasingly significant deviations in their stated directions. Theprobability of the indicated overall result occurring by chance is less than2 x 10 - 4 .Other instructive features of the overall R EG data base appear in the dis-tribution of terminal scores of the com plete series. For example, Fig. 7 showshistograms and analytical fits of all 87 series z-scores fo r PK+, PK-, andbaseline efforts. W hile the mean values of these three distributionsare con-sistent with the terminal values of the cumulative deviation traces of Fig. 6,it is notable that the distributions of the PK + and PK- series scores bo th havelarger than expected variances, at significance levels of 0.06 (PK+) and 0.01(PK-) . Conversely, the distribution of baseline series scores is substantiallycompacted around the theoreticalmean and totally devoid of any scores out-side of the one-tailed significance criterion, z > 1.645. The correspondingreduction in the variance of the baseline score distribution is significant at p= 0.01. Recalling that baseline data are generated under conditions identicalto the PK series, save for the absence of a stated directional intention on thepart o f the operator, one is led to hypothesize that a conscious o r unconsciousmotivation to achieve a "good" baseline may actually produce a third PKcondition that entails an anomalous constriction of the distributionof scores.This issue is discussed in greater detail in a technical report (Jahn, Nelson, &Dunne, 1985).Experiments such as these inevitably raise the question of the focus of theinteraction between the consciousness of the operator and the machine. Inparticular, it is reasonable to ask whether the physical behavior of the noisesource itself is affected during the PK efforts, and if so, in what way.O neobvious strategy for addressing this question is to replace the source unit byother elements and compare results. Several similar microelectronic noise


10/30

TA B LE IREG data summary by operator

Opr.101416192 02 1293033364 14 24 449535559646 5666870

#Series

153313112125122313221213

#Trials55,1008.0007,5002,9507.5502.7002.5005,0004,0005,00013,4502,7005,3004,9507,55043,3005,1005,5002,6007,9504,6507,700

Mean100.082100.070100.070100030100.087100.04499.912100.02699.86899.978100.023100.09499.78199.87199.937100.018100.04699.940100.207100.00399.9559 9 9 6 3

PK +

z-Score2 .7290.8850.8560.2321.0640.321-0.6250.262-1.178-0.2180.3730.691-2.255-1.284-0.7780.5440.461-0.6251.4890.041-0.429-0.459

PK

Prob.*.003.188.196.408. 1 4 4.374( .266).397

( . 1 1 9 )( .414).355.245( .012)(.099)( .218).293.322( .266)068.484( .334)( .323)

# Series # Seriesp < .05* p < .5

3 121 2_ 2_ 1_ 2_ 1

_1 _ 1 3__ (1) _-(1) 12 ( 1 ) 7 1_1_1_1 1

#Trials55,0507,8007,5002,8007,4502,3002,5005,0002,5005 ,25015,0502,3006,2005,0507,45043,4003.9004.5002,4007.0505,3507,300

Mean99.89699.87299.763100.04299.979100.156100.04699.93999.928100.06899.984100.03199.918100.07299.981100.02899.923100.04299.95699.930100.00599.922

z-Score-3.459-1.603-2.9030.313-0.2621.0560.322-0.606-0.5120.695-0.2730.212-0.9140.722-0.2360.818-06840.403-0.303-0.8300048-0940

# SeriesProb.* p < .05*

3 x 10-4 4(1).054 1.002 1( .377) -.397 (.146) -( .374) -.27 2 .304 ( . 244) .392 1)( .416) .180 ( .235) -.407 (.207) -(1). 247 ( .344) -.381 .203 (.481) .174

# Seriesp< .5

133322114212521112

3

RG.Jnea


11/30

Table I ( c on t inue d )PK+

# # # Series # Series # Opr. Series Trials M e a n z-Score Prob.* p < .05* p < .5 Trials M e a nz-Sc80 2 7,500 100.185 2 .272 .012I I 7,500 99.871 -1.582 2 4,100 100.193 1.745 .041I 2 5,250 100.06084 I 2,500 100.102 0.724 .235 2,500 99.765 -1.66385 I 5.100 100.088 0.891 .186 5.300 100.18088 I 2,750 100.169 1.254 .105 3.000 100.084 090 I 2,500 100.088 0.619 .268 2,500 99.968 -092 2 7,400 99.941 -0.722 (.235) 1(1) 5,400 100.087 0.90393 2 5,000 100.144 1.440 .075 2 5,000 100.0194 4 10,000 100.056 0.796 .213I 2 10,000 99.939 -0.8696 I 2,250 100.073 0.492 .311 2,750 100.085 0.6397 I 2.500 100.110 0.781 .218 2,500 99.935 -All 87 262,650 100.037 2.666 004 10(4) 53 259,800 99

Baseline # Series# # # Series M ean # Opr. Series** Trials Mean z-Score Prob. p < .05 > 100 Trials z-Sc

10 13 59,100 100.011 0.389 349 8 110,15014 3 7,250 99.936 -0.774 .219 15,800 17516 3 7,500 100.024 0.292 .385 15,000 2.65819 I 2.500 100.044 0.311 .378 5.75020 3 7,500 99.956 -0.537 .296 0 15.000 21 I 2.500 100.032 0.229 .409 5,000 -0.48029 I 2,500 100.076 0.540 .295 5,000 -0.67030 I 2.650 99.978 -0.162 .436 0 10,00033 I 2,500 100.053 0.376 .353 6,500 -0.60736 2 4.500 99.928 -0.679 .249 10,250 -0.650


12/30

Table I (con t i nued)

Opr.4142444953555964656668708082848588909293949697A l l

#Series**312231011121321111122411

76

#Trials9,8002.5003,0005.0007,50040,0001,1005,5002,5007,5004,9507,5005,5005.0002,5005,5502,9002,5004,9505,0009,5002,5002,500

243.750

Mean100.01299.96499.982100.07899.961100.004100.11999.908100.17299.964100.158100.081100.160100.014100.092100.10799.91599.82099.90099.82799.96199.914100.104

100.004

Baseline

z-Score0.166

-0.255-0.1370.784-0.4770.1130.559-0.9651 .213-0.4441.5720.9981.6760.14406531.129-0.649-1.270-0.997-1.730-0.543-0.6110.7350.282

Prob..434.400.446. 2 1 7.317.455.288.167. 1 1 2.328.058.159.047.443.257.129.258.102.159.042.294. 271.231.389

# Series# Series Meanp < .05 > 100_ 1 0 1 2 2 6 1 0 1 1 1 2_ 2_ 1 1- 1 000 0_ 2 0 1 41

#Trials28,5005,00011,50010,00015,00086,7009,00010,0005,00015,00010,00015,00015,0009,3505,00010,4005,7505,00012,80010,00020,0005,0005,000

522.450

z-Score0.4550.364-0.860-1.417-0.386-0.1940.798-0.7341.2840.599-0.3280.3272.7240.6951.688-0.6980.3990.600-1.1350.8811.173-0.1400.8783.614

d P K

Prob.*.325.358(.195)(.078)(.350)(.423).213( .232).100.275( .371).372.003.244.046(.243).345.274( .128).189.120( .444).190

2 X 10 -4

# Series # Seriesp < .05* p < .5_ 4_1 -(1)-(1)2M l) 6_ 2 1_1 2

1 1 21 1

1_ 11(1) 1_ 21 3 110 (4) 56

N u m b e r s in parentheses indicate results opposite to in tent ion .* In some early series baselines were generated by experimenters, and these are not included in the table.

32

RG.Jaea


13/30


Fig. 7. REG frequency of series z-scores: A ll data, 33 operators.


14/30

34 R. G. Jahn et al.units have indeed been so employed, with no discernible consequences onthe overall pattern of data.In an attempt to explore this issue more aggressively, a categorically differentrandom source has been developed that m ay properly be termed "pseudo-r a ndom " in character. This device employs a feedback array of 31 microelec-tronic shift registers that produces a determinate repeating sequence of 2X 109 bits at a set clock frequency. In the mode most c o m m o n l y employed,this determinate sequence cycles continuously with a repetition period ofabout 60 hours, so that the only remaining non-deterministic aspect of theexperiment is the t ime of incursion initiated by the operator. This pseudo-random source can be switched into the standard R EG apparatus at an ap-propriate location, replacing the commercial microelectronic noise diode andits conditioning circuitry but leaving all subsequent sampling, coun ting, anddisplay circuitry, feedback, and software identical to the standard version.From the perspective of the operator, this system is virtually indistinguishablefrom that of the standard REG, save for an identifying code printed on thestrip tape, and the experimental protocols employed are identical. The resultsof 29 experimental series employing this pseudo-random source are also sta-tistically significant with a probability against chance of .003, (Fig. 8 andTable 2), and the individual operator signatures show strong qualitative sim-ilarities to those achieved on the standard REG.To pursue further the question of how device-specific are such signaturesof achievement, a substan tially different experimental device called a "RandomMechanical Cascade" (RMC)has been employed (Nelson , Dunne , & J a h n .1987). This apparatus, some 6' X 10' in dimension, allows 9,000 " polystyrenespheres to trickle downward through a quincunx array of 330 J* diameternylon pegs, whereby they are scattered into 19 collecting bins across the bot-

Fig. 8. Pseudo-REG cummulative deviations from theoretical mean: A ll data, 10 operators.


15/30

T A B L E 2Pseudo-REG data su mmar y by operatorPK+

# Series#Series Opr. # Series # Trials Mean z-Score Prob. p < .05 p < .5 # Trials Mean z-S10 9 23,000 100059 1.269 .102 2 4 22,00014 I 2,300 100.339 2.300 .011 I I 2.700 100.1971 5 I 2,500 99.788 -1 .499 (.067) 2,500 99.802 -16 I 2,500 100.385 2.724 .003I I 2,500 100.12241 2 5,600 99.928 -0.758 ( .224) - - 4,400 99.93455 9 23.400 99.988 -0.251 (.401) 3 21,600 99.98670 2 5,450 99.933 -0.69$ (.243) 1) I 4,550 99.921 -0 .75180 I 2,500 100.082 0.583 .280 I 2,500 88 I 2,500 99.953 -0.334 (.369) - - 2.500 99.74894 2 5,400 100.208 2.161 .015I 2 4,600 99.9

A ll 2 9 75 ,150 100.037 1.418 .078 5(1) 13 6Baseline

Series# Series Opr. # Series # Trials Mean z-Score Prob. p < .05 Mean > 100 * Trials z-S10 9 22,500 99.957 -0.904 .183 3 45,000 314 I 2,500 100.019 0.133 .447 I 5.000 0.491 5 I 2.500 100.183 1.293 .098 I 5,000 -0.0716 I 2.500 99.875 -0.882 .189 0 5.00041 2 5,000 100.138 1.384 .083 2 10,000 -0.15855 9 22,500 99.939 -1 . 29 1 .098 2 45,000 070 2 5,000 100.139 1.392 .082 2 10,000 -0.00780 I 2.500 100.089 0.631 .264 I 5,000 0.4988 I 2,500 99.886 -0.809 .209 0 5,000 102494 2 5,000 99.717 -2.834 .002I 0 10,000 1.734 .0A ll 2 9 72.500 99.969 - 1 . 1 7 0 .121I 12 145.000 2


16/30

36 R. G. Jahn et al.tom, filling them in close approximation to a Gaussian distribution (Fig. 9).The growing population of each of the bins is tracked photoelectrically anddisplayed via LED co unters at the bottom of that bin, and recorded on-linein an appropriately coded computer file. The experimental protocol calls fo rth e operator, seated on a couch approximately eight feet from the machine,to attempt o n volition or instruction to distort th e distribution to the right

Fig. 9. Random mechanical cascade.


17/30

Engineering anomalies research 37(PK+) or to the l e f t (PK-), or to generate baselines. A ll data are acquired inconcomitant sets of three runs, one under each of these intentions, to controlagainst unforeseen artifactual influences. The temperature and humiditywithin the RMC apparatus are routinely recorded to assess any possible cor-relations with th e experimental data.Figure 10(a) displays as cumulat ive deviations th e data of all 3072 runs bythe 22 operators who have completed at least one formal series of 10 or 20runs per intention . O nce again, th e total aberration is statistically significant,to the order of 3 x 1 0 -6 . but in this case only the left-going efforts are inde-pendently significant. As can be seen in Table 3. most of this asymmetry isdue to the characteristic contributions of two operators who happen to haveexceptionally large individual data bases. W i t h these tw o omitted, th e com-

Fig. 10. RM C cumulative deviations from fitted baseline mean: (a) A ll data, 22 operators, (b) Alldata, 20 operators.


18/30

38 R. G . Jahn et al.T A B L E 3R M C data summary by operator

Opr.

141 6204 14244495 1S35 563646668697079849 19394A l l

1 01 41 6204 14244495 1535 563646668697079849 19394A l l

#Series

1 71327311112011121411133

76

1 713273

20

21411133

76

#Runs

2702030209030201 01 01 030071 010401 1409

1 01 63 130

1024

2702030209030201 01 01 03007

1 01 0401 1409

1 01 63 130

1024

BLmean

10.032810.03361 0 . 0 1 1 110.000210.02059.991310.041910.021910.004710.028310.027210.020710.016410.021010.022410.046310.016610.021410.023710.029010.014210.004110.0245

10.032810.03361 0 . 0 1 1 110.000210.02059.991310.041910.021910.004710.028310.027210.020710.016410.021010.022410.046310.016610.021410.023710.029010.014210.004110.0245

Mean

10.029710.028410.024610.022610.02401 0 . 0 1 8 110.036110.029110.006610.021410.028310.034210.02199.990710.018010.025310.027810.044810.034810.032510.02199.997710.0260

10.019510.02599.999410.025910.02419.999010.034210.021210.000210.019510.024410.012610.024110.001710.02211 0 . 0 1 2 110.012210.032910.017510.039510.00459.984110.0190

SD*

.0508.0497

.0496.0505

.0533.0452.0457.0462.0456.0682.0497.0458.0476.0306.0453.0679.0541.0649.0376.0462.0492.0557.0506

.0536

.0418.0507

.0448

.0488.0502

.0484

.0369.0322.0632.0494

.0391.0506.0590.0469.0680.0454

.0394

.0516

.0566.0440

.0467

.0503

t -Score*PK+

-0.987-.0472

1.4971.9750.6203.248-0.5710.4880.133-0.3210.3740.7820.364-3.142

-0.608-1.0231 . 3 1 01.0790.9370.3060.875-0.6360.978

PK--4.076-0.822-1.2562.5660.6970.838-0.713

-0.058-0.436-0.444-0.985-0.5500.482-1.038-0.031-1.664-0.6270.872-0.3750.743-1.221-2.350

-3.473

Prob.(right)

(.162)(.321).073.031.268.001(.287).319.448(.378).354.232.362(.006)(.273)(.165).099. 1 5 6.187.382.194(.265).164

(left)3 x 10-5

. 2 1 1

. 1 1 0(.009)(.244)(.204)

.242

.478.337

.334

.163

.301(.321)

.163

.488

.064

.267(.204)

.358(.235). 1 1 6.013

3 X 1 0 - 4

# Seriesp


19/30

A ll

Engineering anomalies researchTable 3 (cont inued)

39

Opr . #Series *Ru n s B Lmean # Pairsof runs S D * /-Score* Prob. # Seriesp


20/30

40 R. G. Jahn et al.

Fig. 11. Cum ulative deviations: 3 devices, (operator 10).


21/30

Engineering anomalies research 41cases condition-specific, they seem not to be nearly so device-specific. Suchempirical evidence w eakens phenomenological interpretations involving con-sciousness interacting directly with th e random physical process itself, fo rexample, with th e f l u x of thermal electrons in the REG, or o f the polystyreneballs in the R M C , and favors models that deal with aspects generic to all ofthese systems, fo r example, th e i n format i on implicit in their output distri-butions.

Precognitive Remote PerceptionTh e second major class of experimentation concerns the anomalous ac-quisition of information about remote geographical targets, inaccessible byany k n o wn sensory channel. The particular protocol followed is a variation

o n n u m e r o u s similar studies elsewhere ( Du n n e & Bisaha, 1979; Hansen,Schlitz, & Tart, 1983; Puthoff & Targ, 1976; Schlitz & Gruber, 1980; Tart,Puthoff, & Targ, 1979), and is termed precognitive remote perception (PRP).Essentially, one participant, called the "percipient," is asked to generate adescription of an unknown location where a second participant, called the"agent" is, was, or will be situated at a prescribed time. Initially the percipientrecords his impressions about th e target in a free-response, stream-of-con-sciousness style, and then encodes them in some structured form amenableto analytical processing.Most of the experiments reported here were conducted in a precognitivemode, wherein th e percipient's impressions are recorded before th e agentvisits the target and, in many cases, before the target is even selected. Twomodes of target selection have been employed, with no discernible effect onthe experimental results. In the "instructed" mode, the target for each ex-perim ent is randomly selected from a large pool of potential targets previouslyprepared by a third person not otherwise involved in the experiment or itsevaluation, and maintained so that no percipient or agent has access to it. Inth e "volitional" mode, the target is arbitrarily selected by the agent at thetime specified for its visitation.Figures 12-14 show a few examples of typical targets with portions of thecorresponding free-response descriptions; more extensive data are presentedelsewhere (Dunne, Jahn, & Nelson, 1983; Dunne, Jahn, & Nelson, 1985;J a h n , 1982; J ah n & D u n n e , 1986, 1987; Nelson, Jahn, & Du n n e , 1986). A tpresent, the data in hand consist of 334 perceptions of this sort that rangefrom virtually photographic accuracy, through varying degrees of correspon-dence to the details and overall ambience o f the scene, to total irrelevance.In some cases, details that are central to the agent's view of the scene areignored by the percipient, while minor aspects are escalated in importance.

In other cases, there are spatial inversions or other geometrical distortions.Frequently, the more impressionistic aspects seem to be perceived more ac-curately than the analytical details.


22/30

42 R . G . J ahn et al.

Fig. 12. PRP Target: N W R R station. G lencoe. Illinois. Th e percipient wa s some five or six milesaway. The perception, generated 35 minutes precognitively, reads, in part:". . . I see a train station, one of the comm uter train stations that's on the expressway. . . I see a train coming . . . See just th e front end of the train station. See a little bi twithin it. I think there are wooden planks on the floor. I hear like the clicking . . . offeet or shoes on the wooden floor . . . There are posters or something up . some kind ofadvertisements or posters on the wall in the train station. I see the benches. Getting theimage of a sign, but I think it's probably the sign of what station it is. It's about 8 or 10letters in the word. Maybe something like Clydeburn or Clayburn. Have the impressionof this wooden floor being som ewhat littered, just sort of dirty. I see the tracks. N o trainon the tracks right now. Empty tracks.. . ."


23/30


Fig. 13. PRP Target: Ruins of Urquardt Castle, Loch Ness, Scotland. The percipient was in N ewYork City, some 3500 miles away. The perception, generated 14 hours retrocognitively,reads:"Rocks with uneven holes. Also smoothness. H eight. Ocean. Dark. Dark blue. Whitecaps. Waves boo ming against rocks? On m ountain or high rocks overlooking water. Darkgreen in distance. Gulls flying? Pelican on a post. Sand. A lighthouse? Tall structure.R o u n d with a conical roof. High windows or window space with a path leading up to it.Or a larger structure or a castle." (Here there is a sketch of a castle abutment on thetranscript.) "Old. Unused. Fallen apart. Feeling musty, or dark. Moss or grass growingin walls. Wood draw bridge? A black dog? Snow. Ice capping a mountain. High largecavernous hall. Castle."

The principal effort in this study has been to devise analytical methods toextract from such subjective evidence some quantitative measure of the degreeof anomalous information acquisition. For this purpose, a code, or alphabet,of simple descriptive queries is employed, which can be addressed to all targetsand all perceptions. These descriptors, 30 in number, range over a spectrumfrom quite factual discriminations, fo r example, whether the scene is indoorsor outdoors, whether trees are present, or whether there are automobiles, tomuch more subjective aspects, such as wh ether the ambience is noisy or quiet,confined or expansive, hectic, or tranquil. Encoding of the target is norm allyperformed by the agent at the time of visitation, and of the perception by thepercipient after the free-response impression of the target h as been recorded.With the target and perception thus encoded, a variety of analytical scoringm ethods are invoked, described in detail elsewhere (Dunne et al., 1983; Jahn,Dunne, & Jahn, 1980; Jahn et al., 1982), that yield numerical scores indicativeof the information content of each perception relative to its correspondingtarget. Most of these methods acknowledge the a priori probabilities of thevarious descriptors, that is, that m ore scenes tend to be outdoors than indoors,that more tend to have people in them than not, etc.; therefore, a perception


24/30

44 R . G . Jahn et al.

Fig. 14. PRP Target: Tretiakovskaia Gallereia, Moscow, U.S.S.R. The p ercipient w as in W isconsin,some 4500 miles away. The perception, generated 24 hours precognitively, reads, in part:"H ave the sensation of being in a very quiet, sombre, subdued so n of atmosphere. . . .Any color impressions I get are the samegreys, browns, dark subdued colors. I f e e l anoldness. . . . I 'm th ink ing o f a large church o r something, o r a castle. Some k i n d o fbuilding. I t seems to be quite large. Sensation o f sounds echoing, subdued colors. . . . Ise e several, maybe two to four round balls that seem to be on top of something. M a y b eit's some kind of decoration. Like on top of something that 's of a generally square shape.Almost like a square co lumn with a ball on t o p . I have a very clear picture suddenly o fan old building. It's quite large. There are windows with, like, arches. They may not beexactly arched: the arches come to a point on top. almost. Very impressive. It's a lightgrey color, very ornate. It comes to a point of some sort. bu t it's not a regular point. Likewhere it should be round on top it comes to a point. I 'm not sure if it 's windows or theshape of the building itself.. . . Great big double doors.... Just saw those square pillarswith the balls on top again. They seem to be almost like an entranceway. one on eitherside.. . ."

that is accurate a bo ut less likely aspects achieves a higher score than one thatcorrectly predicts more likely features. The scores are all normalized in somefashion, fo r example, by various chance expectations or by perfect scores. Insome recipes ternary or quaternary descriptor responses are also employed,whereby the agent and the percipient can effectively reject or equivocate ona question, or express gradations of its importance.The most powerful aspect of this coding approach is that unlike traditionalimpressionistic ranking procedures, digital scoring algorithms can be appliedto compare any perception with a very large number of alternative targetsnot just the 5 or 10 that could b e compared by a hum an judge. The distributionof the mismatch scores, that is, the off-diagonal matrix elements of the per-ception/target array, has sufficiently Gaussian characteristics to serve as anempirical "chance" reference fo r statistical quantification of the correctly


25/30


26/30


27/30

Engineering anomalies research 47is no statistically significant dependence o n this temporal parameter. Identi-fication of the specific process whereby th e consciousness o f a percipient isapparently able to access points remo te in both space and t ime from its cu rrentphysical location is well beyond our present understanding, but data such asdisplayed in Figs. 16 and 17, if sustained in further experimentation, willseverely delimit the physical mechanisms that could legitimately be invoked.Clearly there is also interest in the dependence of the yield of such exper-iments on the psychological characteristics and strategies of the percipientsand agents, both separately and as paired participants, but as noted earlier,this category of correlation has no t been extensively pursued, in part becauseth e data base is still far too small for such analyses to be effective. Nonetheless,th e compounding data are routinely examined to gather impressions aboutth e efficacy of the descriptor questions, th e variability in individual interpre-tation of and response to the descriptors, and the physical and psychologicalcorrelates of the most successful trials.

Theoretical ConsiderationsExperiments like those outlined above beg for some form o f theoreticalmodel to help correlate data, design more incisive experiments, and interpretth e observed effects in more fundamental terms. The literature o f psychic

research abounds with attempts to transpose various physical formalisms tothis purpose: electromagnetic models, thermodynamic models, mechanicalmodels, statistical mechanical models, hyperspace models, quantum me-chanical models, and others (Bohm, 1971; Chan, 1977; Costa de Beauregard,1979; Feinberg, 1975; Kogan, 1968;Persinger, 1979;Rauscher, 1979;vonLucadou & Kornwachs, 1979). Although these comprise an interesting bodyof effort, none of them seems fully competent to accommodate experimentaldata like those described above. Indeed, it appears that no simple applicationof existing physical theory is likely to prevail. In order to encompass th eobserved effects, a substantially more fundamental level o f theoretical modelwill need be deployed, one which more explicitly acknowledges th e role ofconsciousness in the definition of p hysical reality.The model that has so far proven most serviceable for our purposes takesthe position that reality, or experience, is constituted only in the interactionof consciousness with its environment, and thus that any physical theory, orany other scheme of conceptual organization, can only properly address theinteraction, not the environment or the consciousness, per se. Similarly, itregards th e co m m o n concepts an d formalisms of physical theories as no morethan useful organizing strategies adopted by the consciousness to order andprocess the information it acquires from the environment. Therefore, theseshould be as much reflective of the characteristics of the consciousness as ofthose of the environment or, more precisely, they should reflect the charac-teristics of the interaction of the two.


28/30

48 R. G. Jahn et al.In this spirit, the model at tem pts to apply , via metaphor, th e concepts andformalisms of elem entary quantum mechanics to a representation of the in-teraction of consciousness with physical systems and processes, in a form tha tcan accomm odate both "normal" and "anomalous" behavior. Central to thismetaphor is the assignment to consciousness of a partially wave-like characterwhich can manifest itself in various interactions, much like th e wave/particleduality of atomic scale phenomena . More specifically, by representing con-sciousness by q u a n t u m mechanical wave functions, and its physical environ-ment by appropriate potential energy profiles, Schrdinger wave mechanicsmay be used to derive eigenfunctions and eigenvalues that can be associatedwith both th e psychological an d physical experiences of the consciousness/environment interaction. To bring this metaphor to pragmatic utility, it isnecessary to relate certain mathematical aspects of the formalism, such as the

coordinate system, th e quantum numbers , and even th e metric itself, withvarious impressionistic descriptors of consciousness, such as its intensity, per-spective, approach/avoidance attitude, balance between cognitive and emo-tional activity, and receptive/active disposition. With these in hand, the genericprinciples of quantum mechanicsuncertainty, indistinguishability, corre-spondence, exclusion, etc.as well as a number of specific computat ionalapplications, such as the central force field and atomic structure, covalentmolecular bonds, barrier penetration, and quantum statistical collective be-havior, become useful analogies fo r representation and correlation o f a varietyo f consciousness-related physical effects, bot h normal and a n o m a l o u s , andfo r th e design o f exper iments to study these more systematically.The full text and mechanics of this model are developed elsewhere ( J ah n& D u n n e , 1986, 1987), along with its application to part icular experimentalsituations. Even in its comprehensive form, since the various associations itinvokes are largely intuitive an d empirical, and since few quantitative scalesof consciousness properties yet exist, no mo re than semi-quantitative corre-lations can so far be made. Notwithstanding, comparison of our laboratorydata and the informal testimony of our operators with appropriate aspects o fth e model have substantiated o ur confidence in the potential utility of thequantum mechanical metaphor and confirmed its value in providing a viableperspective and a facile language for the design of better experiments and theinformal representation of the operators' cognitive and emotional attitudesand strategies. Beyond this, the model suggests a number of testable hyp otheses,some o f which are currently under study. For example, the postulated wave-mechanical nature o f consciousness/environment interactions implies thatth e combined efforts of two or more consciousnesses attending to the sametask ma y display constructive o r destructive interference patterns, rather th ansimple linear superpositions. To test this, experiments utilizing the R EG andR M C devices are underway to explore the effects of multiple operators ad -dressing the same task simultaneously, compared to their individual signaturesof achievement, and preliminary results appear to support the model. Y etother experiments are investigating th e effects o f spatial separation between


29/30


30/30

50 R. G . Jah n et al.(Technical Note PEAR 85003). Princeton Engineering A nom alies Research. Princeton U ni -versity, School of Engineering/Applied Science.Feinberg, G . (1975). PrecognitionA memory of things future. In L. Oteri (Ed.). Q u a n t u m physicsand parapsychology (pp . 54-64). N ew York: Parapsychology Foundation.Hansen, G. P., Schlitz, M. J., & Tart. C. T. (1983). S u m m a r y of remote viewing experiments.Unpublished manuscript.Jahn. R. G. (1982). Th e persistent paradox of psychic phenomena: A n engineering perspective.Proceed ings IEEE, 70 , 136-170.Jahn, R. G., & Dunne, B. J. (1986). On the quantu m mechanics of consciousness, with applicationto anomalous phenomena. F o u n d a t i o n s of P h y s i c s . 1 6. 721-772.Jahn . R. G. . & Dunne . B. J. (1987). M ar g ins of reali ty. San Diego. N ew Yo rk . London: HarcourtBrace Jovanovich.Jahn, R. G.. Dunne, B. J., & Jahn E. G. (1980). A nalytical judging procedure fo r remo te perceptionexperiments. Journal of Parapsychology. 44 . 207-231.Jahn, R. G., Dunne. B. J., Nelson, R. D., Jahn. E. G., Curtis, T. A., & Cook, I. A. (1982).Anal y t i ca l j u d g i n g procedu re for r e m o t e pe r c e p t ion exper i m ent s . II: Ternary c o d i n g a nd g e n -eralized descriptors. (Technical Note PEAR 82002). Princeton Engineering Anomalies Re-search, Princeton University, School of En gineering/Ap plied Science.Jahn, R. G. , Nelson, R. D., & Dunne, B. J. (1985. August). Variance e f f e c t s in REG ser ies scoredistributions. Proceedings of the 28th Annual Convention of the Parapsychological Association,Tufts University, Medford, MA.Kogan, I . M . (1986). Information theory analysis of telepathic communication experiments. R a d i oEng ineer ing . 23 . 122.Krippner, S . (Ed.). (1977). A d v a n c e s i n par aps y c ho log i c a l r e s e ar c h: V o l . 1 . psy chok i nes i s . N ewYork: Plenum Press.Nelson, R. D . , Dunne. B. J. , & Jahn. R . G . (1984). A n R E G exper i m ent w i t h large d a t a basecapabi l i ty . I I I : O p era to r re la ted a n o m a l i e s . (Technical Note P E A R 84003). Princeton Engi-neering Anomalies Research. Princeton University. School of Engineering/Applied Science.Nelson. R. D. , Dunne, B. J . , & J a h n . R . G . (1987). Operator-re la ted a n o m a l i e s in a r a n d o mm e c h a n i c a l ca sca d e exper i m ent . (Technical Not e P E A R 87008). Princeton EngineeringAnomalies Research. Princeton University. School of Engineering/Applied Science.Nelson. R. D., J a h n . R. G . , & D u n n e . B. J . (1986) . Operator-related anom alies in physical systemsand information processes. Journal o f th e Society fo r Psychical Research . 5 3. 261-286.Persinger, M . A . (1979). ELF field mediation in spontaneous psi events: Direct information transferor conditional elicitation? In C. T. Tart, H. E. Puthoff. & R . Targ, (Eds.) (1979). Mind atlarge (pp. 189-204). N ew York: Praeger Special Studies.Puthoff, H. E., & Targ, R. (1976). A perceptual channel for information transfer over kilometerdistances: Historical perspective and recent research. Proceedings IEEE. 64 . 329-354.Radin, D . J., M a y . E. C., & . Thomson, M . J. (198 5, August). P si e x p e r i m e n ts w i t h ra n d o m n u m b e rgenerators: Meta-analysis P a r t 1. Proceedings of the 28th Annual Convention of the Para-psychological Association Tufts University, Medford, MA.Rauscher, E. A. (1979). Som e physical models potentially applicable to remote perception. In A.Puharich, (Ed.). Th e Iceland papers (pp. 49-83). Amherst, W I: Essentia Research Associates.Schlitz, M., & Gruber, E. (1980). Transcontinental remote viewing. Journal of Parapsychology.44 . 305-317.Schmidt H . (1970). A PK test with electronic equipment. Journa l of Parapsychology. 34 , 175-181.Stanford, R. G. (1977). Experimental psychokinesis: A review from diverse perspectives. In B. B.Wolman (Ed.), H a n d b o o k o f parapsychology (pp. 324-381). N ew York: Van Nostrand Rein-hold.Tart C. T., Puthoff, H. E., & Targ, R. (Eds.). (1979). Mind at large: IEEE symposia on the natureo f extrasensory percept ion. New York: Praeger Special Studies.vo n Lucadou, W., & Kornwachs, K. (1979). D e v e l o p m e n t o f th e sys tem-theore t i c approach topsychokines is . Paper presented at the Parascience Conference. London.Walker, E. H. (1975). Foundations of paraphysical and parapsychological phenomena. In L.Oteri, (Ed.), Q u a n t u m phys ics and parapsychology ( p p . 1-49). New York: ParapsychologyFoundation.

R.G. Jahn, B.J. Dunne and R.D. Nelson- Engineering Anomalies Research

Documents