Approved For Release 2000/08/08 : CIA-RDP96-00789ROO2600090002-3 NOTES: RESEARCH PLAN The diagram on page 2 is a reasonably complete collection of research/applied items for consideration in the overall plan for the Cognitive Sciences Program. The. items are not necessarily independent, nor are they intended to be the final word. In addition, they are in random order of priority. In short, iiiis "straw- person" will serve as points for discussion. This document includes comments on each of the items shown on page 2. Notes: Research Plan Approved For Release 2000/08/08 : CIA-RDP96-00789ROO2600090002-3 Approved For Release 2000/08/08 : CIA-RDP96-00789ROO2600090002-3 z 0 (D ,a Anomalous Phenomena (Mental) Cognition Det2or j Transmission gasic Central Nervous System NeuroNet models Autonomic Responses inter-spedes Communications other Animals Applim! other Physiology (Skin) personality (Behavloral/Self ROPOrt-o-sort/man) perceptual Models psychological Models (MotIvation/Emotion) Selection pirecVCorrelational) Environment- Physical (GMF) Environment-psychologic*J (Set and Setting) Environment- Physiological (comfort) Artficial Response IYpe (AUdio/Vldeo/Left Hand) Redundancy (Multiple Pass/MuMple Detectors) Communication Analysis Mixed Internal Noise Source Training (Macroscopic/Operant) Session Protocols Micro Macro Mixed source Basic Informational (Entropy/Meaning) piezoeiectic strain Gauge other Thermodynamic Restive Stain Gauge Vector/Scalar Potential metal (Bending) Pendulums (Unear/Torsion/Blologica!) Applied Mechanical systems (Swis/interferometers) Boundaries Applied Definitions Inertial Systems Human Sander Demarcation (coordinates/Beacon) External Noise Source inverse (Search) Atoms Mixed Nuclei (Mossbauer Effect) Physical Characteristics Photons (size/Compositlon) Type (Statc/Dynamic) Cells (Algae, Blood) One-in-*n" (Forced Bacteria (Mutation - Choloe/Sinary Search) Salmonella) Quantum Systems (Neutron/Photon interference) Crystal Structure Molecular Structure QR Spectra of HzO) Theory (Quantum Measurement/Zeno) Decision Augmentation Random Number Generators Theory (Eleotronle/Nuclew) Worm Holes (4-Dimensions) App.110A Vector/Scalar Potential Magneto Domains Propagation Stochastic Causality ML=d Electrons ApAllAd Living Systems (Humans/Fish -Behavioral) Mloromachlnes Perturbation Approved For Release 2000/08/08 : CIA-RDP96-00789ROO2600090002-3 Approved For Release 2000/08/08 : CIA-RDP96-00789ROO2600090002-3 1. Anomalous Cognition Anomalous Cognition (AC) is described within the context of a sensorial analogy. Information about the source propagates via some tranwziKOon mechanism to a defector. In this model, the detector is usually thought of as human, but other animals or machines could possibly be used. The research items are divided into three broad categories; (1) Basic-primarily to understand the phenomena, (2) Applied- primarily to improve the output of AC experiments, and (3) Mixed-both basic and applied. 1.1 Detector For human detectors, basic research involves internal physiological/physical mechanisms. For the point of discussion, all other variables (e.g., environment, psychology, etc.) can be thought as contributors to the detector efficiency and fall under the applied domain. 1.1.1 Basic 1.1.1.1 Central Nervous System (Cn The central nervous system is the centerpiece for the detector basic research. It is important to identify all possible sources of artifact in the previous pilot study and to replicate the earlier (Y-phase shifts. Once an effect can be verified, the next step is to identify specific evoked responses to remote stimuli. One technique is to use a.classical conditioning protocol to "train!' the visual system to respond to a direct (i.e., in the ears of the subject) audio stimulus. If this can be accomplished, it is a simple addition to intermix direct audio stimuli with remote and sensorially isolated audio stimuli. The advantage here is that the visual system's response to the local audio stimuli can be used to search for a similar signal in a potentially noisy response to a remote audio stimulus. Another interesting potential is to examine sensorially handicapped individuals. For example, is itpossi- ble to observe an ERF to an AC stimulus with someone who was blind from birth. In addition, we could look at individuals who became blind at different times in life. We could also consider split-brain individ- uals, anq people with varying degrees of lateralization. 1.1.11 Neuronet Model~ Neuronetwork technology is enjoying limited success in understanding traditional sensorial processing. This research item is intended to explore that particular success applied to AC perception. 1.1.13 Autonomic Re=nses All known senses appear to respond to external stimuli initially below the level of conscious awareness. In the visual system, for example, a hard-wired response to photons striking the retina occurs appro)d- mately 100 ms after the on-set of the stimuli. Can a similar response to AC stimuli be identified? Can that autonomic response be used to reduce noise in the overall response? Notes: Research Plan Approved For Release 2000108/08 : CIA-RDP96-00789ROO2600090002-3 For Release 2000/08/08 : CIA-RDP96-00789ROO2600090002-3 jC be used to direct the behavior of members of other species? Are some species better communi- than others? 1.1.1.5 Qther Animals if an autonomic response (or AC--evoked response) canbe identified in humans, it is very likely that such a physical sensorial system would not be unique to humans. What other animals possess similar sensorial systems. One might argue that "lower" animals might require AC as a survival mechanism more than humans to. MEGs routinely are used in animal studies. 1.1.2 Applied 1.1.2.1 Qther PWsiology (~3kin, VVilliam Braud at the Mind Science Laboratory has been demonstrating electrodermal activity (EDR) in humans in response to AC stimuli. Can variations on this theme be used to identify correct from incor- rect elements in a verbal AC response. 1.12.2 P~,rsonalily (L3ehavioral/Self R=ort-O-SQrtfM`BT1 The search for personality that is predisposed toward a natural AC ability has been one of the major efforts in research parapsychology. (Hint: Thus the name-) Not surprising, however, to first order it has failed. (No slur intended -it is a very difficult problem!) Self-report (i.e., ask the person to answer tricky personality-revealing questions about him/herself) methods are very unreliable. However, the Meyers Briggs 71Ype Indicator (MBTI) and the Q-Sort have shown some statistical significant results, but the reliability has been poor. We have examined, in detail, a promising behavioral technique (i.e., ask the person to do things rather than answerquestions) called the Personality Assessment System. The PAS has enjoyed great success in certain government applications, and while it is difficult to administer, it continues to be promising to help us in the selection problem. Another approach to this general problem'is to examine independent measures of various types of cre- ativity to search for correlations with AC functioning. 1.1.23 Perceptual Models Are there developed and standard models of perception that can help in understanding AC perception? For example, the parallel between subliminal perception, of which much is known, and AC perception, of which little is known, is striking. Can subliminal training techniques be ported to the AC domain? 1.12.4 PEychological Models As the name parapsychology also implies, much has been done in attempting to construct/apply psycho- logical models to understanding AC. Standard models have not been particularly successful in describ- ing non-AC behavior and thus its application to AC is problematical. Some success with Notes: Research Plan 4 Approved For Release 2000/08/08: CIA-RDP96-00789ROO2600090002-3 Approved For Release 2000/08/08: CIA-RDP96-00789ROO2600090002-3 -data collec- extroversion/introversion scales has been observed, but I think it is driven mainly by the AC tion technique rather than anything basic to the functioning. 1.1.2.s Selection (12irect/Correlatiomlj How are individuals, who possess a natural AC-ability or are susceptible to training, found? An example of a direct method is simply to ask many people to participate in an AC experiment and pick the best. This is particularly inefficient (i.e., approximately 1%). One promising correlational technique is that high scoring individuals on the Stanford Hypnotic Susceptibly Scale are also good AC detectors. 1.1.2.6 Environment-Physical (OME) Interesting new analysis of the geomagnetic field (GMF) has shown some weak correlation with AC abil- ity. Is this real and can it be used to enhance the output quality? Are there other physical correlates? In addition, are there other mental.Ld/or physical human activities that correlate with the GMF or re- lated'phenomena (e.g., DST-Change in Storm Time). 1.11.7 Environment-j%T-hological (Set and & "in The field of envirorqnental psychology deals with understanding the impact of our near-environment on human performance (e.g., color of the walls, music). Are there other aspects of the so-called setting that are important (e.g., lack of clutter)? The psychological set determines how the staff interact with the participants. What role do white lab coats play? Previous research has shown that a warm, supportive and inviting persona is important, but "touchy-feely" is too much. What are the important variable to illicit excellent AC functioning? Are there variable with known diurnal variations that correlate with AC functioning? 1.1.2.8 Environnient-Phy,,&logical (Comfort) Some of the obvious variables are: temperature, humidity, hung er and body stress. Are there more sub- tle issues (e.g., left hemisphere distraction) that can improve the functioning? 1.1.2.9 Artificial Can a machine/computer be taught to detect AC data? How can current expert systems be applied to understanding AC phenomena? ype (Audio/Visual~Left Hand) 1.1.2.10 Re5ponse T M-aditional AC responses have been mostly written and drawn. These methods are not necessarily opti- mal infon-nation processing modes for all people (e.g., some people can describe verbally better,than in writing). If AC functioning is predominantly a right-brain activity, asking highly lateralized people to write/draw with their left hands may give more accurate responses. Notes: Research Plan 5 Approved For Release 2000/08/08 : CIA-RDP96-00789ROO2600090002-3 Approved For Release 2000/08/08 CIA-RDP96-00789ROO2600090002-3 One variation on this theme was developed by Budzinski. He found that "simple" right-brain messages were better "understood" if the left-brain was occupied. Can this technique be adapted to AC experi- ments.? 1.12.11 Redundancy (Multiple Pass[Multiple Detg:~ Little formal investigation of redundancy appears in the literature. Sophisticated fuzzy-set analysis pro- vides a legitimate method of applying redundancy. 1.1.2.12 Communkation AC techniques have been used in the past to "send" messages with 100% accuracy (e.g., a single 50-bit binary message was flawlessly sent using a multiple pass procedure). Can modem and efficient tech- niques (e.g., sequential sampling) improve the efficiency of such systems? 1.11.13 Creativily Do independent measures of various creative abilities correlate with AC functioning. 1.1.2.14 Analysis The development of analysis techniques for AC data has traditionally been a major part of the Cognitive Sciences Program. Simple techniques such as rank-order have the advantage of being understood by everyone, but those same techniques grossly underestimate the information transfer bit-rate. Yet com- plicated techniques such as fuzzy set analysis are difficult to understand and thus are prone to criticism. Nonetheless, adaptive fuzzy set analysis has been shown to be verypromising in developing a person-de- pendent lexicon that can significantly improve the AC measure. Can neuronetworks and/or genetic algorithms be applied to the fuzzy set approach to AC analysis? 1.1.3 MIxed 1.1.3.1 Internal Noise Source The general problem of "internal" imagery is a current and major topic in traditional cognitive science. With somd limited success AC practitioners have been able to separate internally generated "noise" from target-related data. Can we improve on this? What training from traditional cognitive sciences can be adopted? 1.1.3.2 Training (L4acroscopic/012emni) M-aining has been particularly difficult. The good news is that we have not observed a so-called decline effect (i.e., subjects start well and converge to the mean), but we also have not generally been able to demonstrate learning. By macroscopic I mean specific protocols-lesson plan. By operant I mean can response-neurons (hopefully identified in item 1.1.1.3 above) be trained by biofeedback techniques to improve the quality of AC data? Notes: Research Plan 6 Approved For Release 2000/08/08 : CIA-RDP96-00789ROO2600090002-3 Approved For Release 2000/08/08 : CIA-RDP96-00789ROO2600090002-3 1.1.33 Session ProtQWh Session protocols is a general area that includes feedback parameters, session structure (i.e., how long, how often, technique), and data knowledge (Le., who knows what when). Another important aspect of session protocol is when does the session begin and/or end? One test of this is the "hooded" feedback protocol. In the falcon experiment, the AC participant is blind- folded jest after the session. With possible variations on a theme, the blindfold is removed at specific times during a physical visit to the target site. Some the sites revealed are decoy targets and occurbefore and after the actual feedback target. 1.2 Transmission Understanding how data propagates from the source to the detector is a basic research problem. One driving issue is that precognition (i.&, data is received before-in 3-space-the source emits it) is real. 1.2.1 Basic 1.2.1.1 Decision Augmentation Tbf,= Decision Augmentation Theory (DAJ) portents that the usual decision processes are augmented by sta- tistical amounts of information-leakage from the future. One application of this model to 15-years of random number generator data has been particularly successful. 'Me model should be expanded and tested in a number of different experimental domains. 1.2.1.2 Worm Holes (4-Dimensions) Recent research in general relativity has suggested the possibility of realizable time travel through worm holes in the 4-D manifold. Aboundary cyndition that emerged is that causality must be stochastic-not unlike the conclusion forced by the precognition data. What is needed is some smart GR theorist to calculate things for us. 1.2.1.3 Vector/Scalar Potential Prol2agation The E&~, vector/scalar (AyD) potentials are under intense investigation. After the Aharonof (sp?) /Bohm effect demonstrated that A,(D are fundamental aspects of E&M theory, it is hoped that A,(D waves could propagate. To date, theory and experiments have not confirmed such propagation. Howev- er, if there were a possibleA, (D propagation, it is likely that its properties would emulate some aspects of AC. Thus we should look into it. 12.1A Stochastic Caumft Besides worm holes in 4-D, are there other implications of stochastic causality? Notes: Research Plan 7 Approved For Release 2000/08/08 : CIA-RDP96-00789ROO2600090002-3 Approved For Release 2000108/08: CIA-RDP96-00789ROO2600090002-3 1.3 Source ti tive def* i The source of AC data has implications for basic and applied research. lb date, no quan ta in tion of the source exists. Characterizing the source is the topic under discussion. 1.3.1 Basic 1.3.1.1 Informational (EntropyaAeaaW4 Informational entropy is a macroscopic aspect of the source (i.e., target), whereas meaning is a perceptu- al aspect of the target. (It occurs to me at this writing, meaning cannot be a target property!) By entropy, I mean Shannon type at this early stage of the thinldng. Is there a form of "collected" entropy or informa- tion compression that can be adapted to AC target descriptions? 13.11 Other ThermQftami Cari-classical thermodynamics teach us about the limitations of the information transfer from the source? We cannot assume at this stage that information transfer occurs without energy transfer. What are the thermodynamic consequences of this? 13.13 Vector/Scalar Potential Specifically, what are the source requirements that might emit Ad) waves? 1.3.2 Applied 132.1 Boundaries What are the source boundaries? Are they hard-edged (probably not)? From a pragmatic point of view, the judge must determine where to stop the analysis. For example, if the Golden Gate Bridge is the target, does the City of San Francisco count since it is visible from the bridge? 13.2.2 Defmition~ Traditional AC research has focused on visual and or functional aspects of the target.. Ilere. are other dimensions that may be important. What are optimal target definitions for all these domains? 1323 Human Sender In the so-called telepathy mode, the source information is "transmitted" by the mind of an observer. A major amount of data (i.e., the Ganzfeld) has been collected under this assumption. It is obviously not a necessary condition in that most of the SRI AC data was collected without a human sender, but to what degree is a "sender" involved? 13.2.4 Demarcation (_Coordinates/Beacon) The universe isa near-infinitecollection of potential targets. How is the intended target tobe specified? Virtually everything that has been tried (e.g., beacon, coordinates, encrypted coordinates, abstract words, etc.) have been successful. What does this mean? Are some better than others? Why? Notes: Research Plan Approved For Release 2000/08/08 : CIA-RDP96-00789ROO2600090002-3 Approved For Release 2000/08/08 : CIA-RDP96-00789ROO2600090002-3 1.32.5 Egemal Noise Source Once the source of AC data has been identified, targets can be selected on the bases of the lack: of noise located at the source. Can this Vfpc of noise be used as masking? 13.2.6 Inverse (Seardi~ The forward direction for AC experiments involve the demarcation of a target in an attempt to obtain information about it. The inverse problem (i.e., as in physics, a very difficult one) involves giving all information about the target except is demarcation. Where is that target? 1.3.3 MIxed 1-3.3.1 Physical Characteristics (Size/ComW~itiQjn Are atoms "visible" via AC techniques? How can mixed-sized targets be optimally detected? 13.3.2 Type (Staticffl~mamic) Are there specific target types that constitute "easier" AC sources? If so, why? 1333 Qne-in-"n" (F-Qrced Choice/Bina1y Searsh) In early experiments (circa 1930s), most of the effort involved guessing a target symbol from a limited set (usually 5). There is a massive database that shows consistent statistical evidence for an anomaly, but the effect size is quite small (e.g., 0.02). There are a number of possibilities for this. Maybe there is a basic research reason such as entropy. There are internal reasons as well. For example, when an individual knows the set of targets, strong internal imagery resultsfrbin memory and iniqination that masks a puta- tive AC signal. Notes: Research Plan 9 Approved For Release 2000/08/08 : CIA-RDP96-00789ROO2600090002-3 -oved For Release 2000/08/08 : CIA-RDP96-00789ROO2600090002-3 2. Anomalous Perturbation Anomalous perturbation (AP) is the putative mental interaction with the physical world. Generally, tht! potential target systems are divided into two size domains, micro and macro. Most of the items shm~' under Perturbation on page 2 have been used as AP targets. As I have said many times before, the literz- ture is inconclusive mostly in that the researchers are simply unaware of potential confounding in- fluences. Research of AP should begin with an attempt to define more specifically what will be accepted as a genu- ine AP effect. In parallel, detailed calculations should be conducted on specific historical experiments tL) identify possible "normal" perturbations that may have occurred. Notes: Research Plan Approved For Release 2000/08/08 : CIA-RDP96-00789ROO2600090002-3 10