HANS IN LUCK: THE CURRENCY OF EVIDENCE IN PARAPSYCHOLOGY.
ABSTRACT: Personal as well as scientific evidence that leads to the conviction that paranormal phenomena exist is often based on personal experiences. It is important to formulate criteria that constitute personal or scientific evidence. In this article, the author discusses 2 models, a classical and a nonclassical, of how to produce evidence, each leading to different experimental paradigms. The usual classical criteria for scientific evidence are effect oriented. Experimental results of parapsychology seem unable to fulfill these requirements. One gets the impression that an erosion of evidence rather than an accumulation of evidence is taking place in parapsychology. This results in a discrepancy between personal and scientific evidence. A person who reports a paranormal experience gets the impression that the scientific description of it is inadequate and that the relevant aspects of the experiences are given away. This is called the Hans-in-luck-syndrome. The nonclassical model for scientific evidence is development oriented instead of effect oriented. It takes into account the inherent entanglement of psychophysical systems and the fact that such systems have their own history. In such systems, evidence cannot simply be accumulated because the conditions that produce evidence change during the development of the system.
In this article, I define evidence as a direct, intuitive certainty or clear recognition that something exists beyond any doubt because of given empirical data or theoretical considerations. Thus, evidence is an example of so-called "qualia," which cannot be considered as true or false; it is completely subjective, like emotions, for example. It can be described as an insurmountable constraint or force that leads to a firm conviction. Furthermore, evidence is irreversible. If something has become evident for someone, it cannot be denied anymore. For instance, if someone has understood the proof of Pythagoras's theorem, she or he cannot doubt it anymore. Nevertheless, it is necessary to distinguish between personal and scientific evidence. In principle, scientific evidence is based on personal evidence because each member of the scientific community must use her or own evidence to accept a scientific finding. But scientific evidence is usually regarded as more substantial than personal. This has to do with th e fact that scientific evidence is the result of a sometimes long-lasting historical process to which many individual scientists have made their contribution. However, the history of scientific errors shows that it is not always useful to give too much preference to scientific evidence.
Personal evidence is event-dependent at a specific time in a person's life. In most cases, it occurs at a specific and certain moment. An example is the experience that Hans Berger, the inventor of electroencephalography (EEG), had in 1893, when he was a student of astronomy. The experience inspired him to study the interaction between mental phenomena and physiological processes. Riding a horse on the narrow edge of a steep ravine, Berger fell into the path of a mounted battery and came to lie almost beneath the wheel of one of the horsedrawn guns. The battery came to a stop just in time, and he escaped injury. In the evening of the same day, Berger received a telegram from his father asking about his well-being, the only time in his life that he had received such a query. This inquiry resulted from Berger's sister having told her parents that she knew for a fact her brother had been involved in an accident. Berger (1940) later wrote, "This is a case of spontaneous telepathy in which at a time of mortal dan ger, and as I contemplated certain death, I transmitted my thoughts, while my sister, who was particularly close to me, acted as the receiver" (p. 6).
This is a good example of how personal evidence may yield scientific evidence. However, as we know after 70 years, electrical potentials of the brain cannot be an explanation of what Berger considered as telepathy: There is abundant evidence that electrical brain potentials exist, but the evidence for telepathy is still in question. This does not mean that there was no scientific evidence for telepathy a century ago. On July 17, 1882, Henry Sidgwick gave the following statement in his famous presidential address to the Society for Psychical Research, of which he was one of the founding fathers:
Scientific incredulity has been so long in growing, and has so many and strong roots, that we shall only kill it, if we are able to kill it at all as regards any of those questions, by burying it alive under a heap of facts. (We must keep "pegging away," as Lincoln said.) We must accumulate fact upon fact, and add experiment upon experiment, and, I should say, not wrangle too much with incredulous outsiders about the conclusiveness of any one, but trust to the mass of evidence for conviction. The highest degree of demonstrative force that we can obtain out of any single record of investigation is, of course, limited by the trustworthiness of the investigator. We have done all that we can when the critics has [sic] nothing left to allege except that the investigator is in trick. But when he has nothing else left to allege he will allege that. (p. 12)
If one compares the scientific status of telepathy and EEG today with regard to Sidgwick's strong statement, we have to ask ourselves as parapsychologists why it is not possible to accumulate the same "amount" of scientific evidence for telepathy as for EEG.
Accumulation of Evidence
It is not the purpose of this article to summarize the state of the art of philosophy of science, but before I discuss the accumulation of scientific evidence I should summarize general rules of scientific enterprise. Here, I formulate four rules that are generally accepted among scientists and that, if consequently applied, serve as a self-regulating system:
1. Any statement is admissible (openness).
2. Any statement must be demonstrated (demonstrability).
3. Any demonstration must be questioned (criticability).
4. Any fact must be taken into account (completeness).
It is known that these requirements are not always used consequently, and both parapsychologists and skeptics claim that their antagonists lack one or more of these principles. The procedure itself does not yet create any evidence but is a prerequisite for it. Usually, scientific evidence is not created in one single step. McCroskey (1969) described this process in a model that contains both empirical and sociological aspects of the production of evidence. Symbolically, the classical model of evidence is given by the following schema (McCroskey, 1969):
The model assumes that the influx of empirical data D yields evidence E by a cyclic process in which wrong assumptions are canceled out by a scientific judging procedure. The procedure filters out irrelevant or misconceived or wrong data (lower loop in the scheme). If data D increase, evidence may arise like a step function or a quantum jump. However, there are two sociological factors (upper loop in the scheme) that also control the production of scientific evidence. The more general factor that is specially relevant in parapsychology and anomalies research contains aspects of the (scientific) community, such as existing paradigms, prejudices, experience, and plausibility. The second factor reflects characteristics of a single investigator, such as her or his prestige, credibility, and, increasingly, her or his political correctness. It is obvious that in the case of parapsychology these two sociological factors have antagonistic influence on the development of evidence E that a certain phenomenon, for inst ance, telepathy, exists. The model also describes that new evidence has an influence on these sociological factors. The second loop creates a certain "hysteresis function" (indicated in the box). This means that if once a phenomenon or a concept has become "scientifically evident" and if a new paradigm P has been established, a greater amount of empirical data is necessary to remove the evidence (or the paradigm) than it was necessary for establishing it initially.
In any case, the model confirms the idea of Sidgwick (1882) that "we must accumulate fact upon fact, and add experiment upon experiment." I call this the classical experimental paradigm. In parapsychological experiments it includes the following 11 features:
* accumulation, amplification, long runs
* preparation of "pure" situations, far from real life
* optimizing the effect size, psychological variables = error variance
* data reduction to a psi effect
* separation of effect and its meaning (fancy display)
* pure chance, optimizing the random event generators, or REGs (output switching)
* single or double-blind design
* experimental tricks, hidden agenda
* direct evaluation after the experiment
* identical replications
* orthogonal variables
I discuss these points later in comparison with the proposed nonclassical model.
The model above says little about the speed of the process. Take, for example, the case of telepathy: There is a low a priori probability for its existence in relation to the currently accepted paradigm. Using a Bayesian analysis, Beauregard (1978) showed that in such a case even moderate skeptics and open-minded believers can contribute little to the probability that evidence for psi arises as a result of experimental data. The increase of the a posteriori probability, which is necessary to surmount the "quantum jump" for evidence, seems to be only large enough if the a priori possibility is not too small. This is the case in normal science but not in parapsychology.
The following is Beauregard's (1978) model in more detail:
p = a posteriori probability for psi (successful experiment);
A = a priori probability for psi;
L = probability for a conventional explanation (no psi).
p = 1/(1 + L*(1/A - 1)).
For a moderate skeptic: L = .5, A [much less than] .5.
p = 2/(1/A + 1).
[right arrow] p very small, p - A very small.
For an open-minded believer: A = .5, L [much less than] 1.
p = 1/(1 + L).
[right arrow] p can become large, however if L = 1, p is still p = .5, which means that the experiment has no validity to falsify the assumption of the believer.
To my conviction, parapsychology has--since the days of Henry Sidgwick-not yet succeeded in establishing indisputable scientific evidence that psi exists. If one observes the recent development of parapsychology, one gets the impression that "erosion of evidence" rather than "accumulation of evidence" occurs. This results in a discrepancy between personal and scientific evidence. People who have paranormal experience get the impression that the scientific description is inadequate and that the relevant aspects of their experiences are given away. This does not exclude the possibility that individual researchers (including me) are personally convinced that psi exists, but the scientific community in general is far from reaching this belief. Today only a minority of mainstream researchers will accept the existence of psi.
However, when I talk to the general public about psi experiences, the reaction is usually quite different. If I cite the experience of Hans Berger, given above, nearly 80% of the auditorium is convinced that Berger's personal evidence is acceptable, and they wonder why scientists still have problems accepting this. (For more than 10 years, I have given at least 500 public lectures [or at least 50 per year] about parapsychology for a general audience in Germany. Usually, I present Berger's story as an example for a spontaneous case. As a possible explanation, I propose fraud, chance, or telepathy and ask the audience for their opinion. About 70%-80% vote for telepathy.)
Hans in Luck
In a recent meta-analysis, Bierman (2000) showed that nearly all psi experiments that had been performed since the days of J. B. Rhine exhibit a significant interexperimental decline effect. Especially the three recent, nearly identical replications of the PEAR (Princeton Engineering Anomalies Research Laboratory) experiments at Freiburg (FAMMI; Freiburg Anomalous Mind/Machine Interactions group), GieSSen (GARP; Giessen Anomalies Research Project), and at PEAR itself (the largest database ever available, and the most carefully conducted experiment, to my knowledge) could not confirm the main hypothesis. Given such a large database, including a large number of different experimental designs and settings, it seems highly improbable that this insignificant result could be explained by psychological factors such as loss of motivation, exhaustion, or experimenter expectation. Even though this psychological interpretation is in principle unfalsifiable, it is reasonable and legitimate to assume that the interexperi mental and possibly also the intraexperimental decline effect exhibit an essential characteristics of psi phenomena.
If one considers this situation after more than a century of parapsychological research, one is not only inclined to agree with John Beloff (1993) that there is only one reliable result of parapsychology, namely, that its experiments are not replicable, but it also brings to mind an old German fairy tale named "Hans in Luck," of which I relate the relevant last episode:
As Hans was going through the last village, there stood a scissors-grinder with his barrow, and as his wheel whirred he sang, "I sharpen scissors and quickly grind, my coat blows out in the wind behind."
Hans stood still and looked at him. At last he spoke to the scissors-grinder and said, "All's well with you, as you are so merry with your grinding." 'Yes," answered the scissors-grinder, "the trade has a golden foundation. A real grinder is a man who as often as he puts his hand into his pocket finds gold in it. But where did you buy that fine goose?"
"I did not buy it, but exchanged my pig for it."
"And the pig?"
"That I got for a cow."
"And the cow?"
"I took that instead of a horse."
"And the horse?"
"For that I gave a lump of gold as big as my head."
"And the gold?"
"Well, that was my wages for seven years' service."
"You have known how to look after yourself each time," said the grinder. "If you can only get on so far as to hear the moneyjingle in your pocket whenever you stand up, you will have made your fortune."
"How shall I manage that?" asked Hans. "You must be a grinder, as I am, nothing particular is wanted for it but a grindstone, the rest finds itself."
The Hans-in-luck-syndrome, as I call it, consists of a tendency to give away valuable things for inferior objects, without realizing the loss, and moreover feeling satisfied with that. If one deals with parapsychology for more than 30 years, as I do, one cannot avoid having the impression sometimes that some parapsychologists suffer from the Hans-in-luck-syndrome. But I want to emphasize here that I believe that Hans Berger was certainly not one of these, nor the other two great Hanses in German parapsychology, namely, Hans Driesch and Hans Bender.
Many parapsychologists have had their own personal psi experiences or at least have heard impressive spontaneous reports from credible friends. As "reward," they get their personal evidence, but because it is not a scientific one they are not satisfied and try to change it. But what do they obtain instead?
For the: [right arrow] they get: experience [right arrow] report (spontaneous cases) report [right arrow] category (telepathy, ESP, psychokinesis [PK]) category [right arrow] operationalization (qualitative) operationalization [right arrow] experiment (quantitative) experiment [right arrow] effect effect [right arrow] statistics (effect size, z, p value) statistics [right arrow] meta-analysis (file drawer, quality rating) meta-analysis [right arrow] belief system belief system [right arrow] social data social data [right arrow] clients clients [right arrow] money
Thus, one can reformulate the grinder in "Hans in Luck" the following way: "You must be a counselor, as I am, nothing particular is wanted for it but a telephone, the rest finds itself, and you can hear the moneyjingle in your pocket whenever you stand up, at least." Parapsychological counseling is indeed a growing market, and as John B. Hasted (also no Hans in luck) phrased it: "Parapsychology will sell soap!"
On a more serious note, counseling for people with exceptional human experiences is an important application of parapsychology. It is important, however, that one can decide whether there is scientific evidence for psi phenomena or not. One story may serve as an example. A young man was hospitalized in the psychiatric ward because he claimed to have precognitive dreams and tried to warn the police to avoid accidents. It turned out, that, indeed, before he went to the police, he had had a precognitive dream about a fatal accident of his father, who was at that time on a business trip. He reported the dream to his mother, but she did not inform the father and the accident happened a short time later. I asked the psychiatrists whether they had taken into account this spontaneous experience of the patient. They replied: "Yes indeed, we asked ourselves whether the dream about a fatal accident of his father was an oedipal fantasy."
Accumulation or Growing?
As an alternative to the classical model of evidence, which seems to be rather infertile in the case of parapsychology, as I discussed above, I propose a new nonclassical model of evidence. It is a further application of the model of pragmatic information (MPI) that I have developed to describe psi phenomena (Lucadou, 1995a, 1995b). In contrast to the usual observational theories, the MPI does not start at the description level of quantum theory but at a very general system-theoretical level. From the fundamental complementarity of structure and function in any description of a given system, an uncertainty relation can be derived. This can further be applied to the concept of pragmatic information, leading to the fundamental equation:
I = R * A = B * E = n * i.
This equation describes the partitioning (product) of reliability R and autonomy A of an organizationally closed system that interacts with its environment by the exchange of pragmatic information I. This exchange can be called a measurement, which yields empirical data. B describes the "confirmation" and E the "novelty" of the pragmatic information in the environmental system, and i is the minimum action that the pragmatic information exerts on the specified system during a measurement.
Only self-referential systems can become organizationally closed. In macroscopic systems, this requires a certain degree of complexity and informational exchange inside the self-referential system. The ratio of internal and external pragmatic information defines the boundaries of the organizationally closed subsystem.
Independent measurements on organizationally closed systems exhibit characteristic correlations that mirror the symmetries and the conserved entities that generate the system. These correlations are nonlocal in nature and can be isolated from causal correlations.
The MPI assumes that psi effects are manifestations of such nonlocal correlations. However, they cannot be used as a carrier of signals that causes rigorous limitations on every psi effect. Any operationalization that would use the nonlocal correlation as a signal transfer would make the correlation vanish. In my opinion, the MPI provides a natural explanation for the inter- and intraexperimental decline effect in parapsychology. (I should mention here, that on the basis of the MPI, I made a clear-cut prediction about the outcome of the FAMMI replication study. It was kept in the minutes before the final evaluation began, but, unfortunately, it is not mentioned in the final research report.)
Because the MPI is a general system-theoretical description of interacting (self-referential) systems, it can also be applied to the system that creates scientific evidence. In contrast to the classical model described earlier, it puts emphasis mainly on the aspect of "growing of evidence" rather than on "accumulation." Accumulation means that external criteria are responsible, whereas growing means that there exists an internal dynamic that "creates" evidence. It cannot simply be accumulated because the conditions that produce evidence change arbitrarily during the development of the system. Furthermore, in this model, evidence is considered as a measure of "fitting." As an equivalent graphical scheme of the nonclassical model, there is no "black box" with input and output control mechanisms but a kind of hierarchically nested system of interacting surfaces comparable with a cell or an organ. If the surface of an inner cell fits with the surrounding cell, evidence is created. From the point of view of the s urrounding cell, however, the initial data D are no more "naked" but they are included in an organizationally closed subsystem. The surrounding cell, which has adapted the original data, becomes again a piece of new data for its own surrounding cell and so on. If there are many of such internested hierarchical levels in which each level fits perfectly with the others, one does not get a "heap of evidence" but a well-organized growth of evidence, which finally can be regarded as paradigm P. But the model also includes the reverse action of pragmatic information from outside to inside. Thus, the existing paradigm acts on the hierarchically lower shell (e.g., the experimental setting) and influences its "surface" in relation to the including cells (e.g., the data) and so on.
Thus, according to the nonclassical model of evidence:
I = Reliability (R) * Autonomy (A)
= Confirmation (B) * Novelty (E)
= n * I
Evidence = Fitting
Fitting = Imax = maximal Novelty including maximal Confirmation
The equation above describes the action of a given pragmatic information I on a system. Because R and A or B and E are corresponding pairs of complementary observables, nothing can be said on a partitioning of an outside pragmatic information I to reliability R and autonomy A "inside" of the system without further specifying the "measurement." Evidence emerges if the loss of action is minimal, which means fitting. Any piece of pragmatic information I (e.g., of the data D or of the paradigm P) that interacts with the system also produces pragmatic information, but with a new partitioning of novelty E and confirmation B, which can be interpreted as the "reaction" of the subsystem or supersystem. Therefore, it is an important issue to specify the subsystem or supersystem, or to be more precise, the boundaries of the system.
If the data D contain "anomalous information" (psi), this can be understood as a mark of the autonomy A of the data subsystem. The recurrence of the same data is a mark of the reliability R of the data system. Any interaction of this system with other "observers" (e.g., the experimental setting) must be described as an exchange of pragmatic information I. However, not only is the data system delivering pragmatic information but also the observers "produce" pragmatic information that manipulates the data system, for instance, by their operationalization or (in higher supersystems) their expectations. For instance, if these observers would "expect" more confirmation B of the subsystem for a specific "surprising" phenomenon x, such that the product E (x) * B (x) would exceed the product I = A * R, the phenomenon may produce a displacement in such a way that instead of the expected something else x' happens for which E(x) * B(x) = A * R hold. Again, I must emphasize that "expectation" in this context means a pra gmatic criterion; namely, a (potential) measurement of x and not only a mental attitude. An example is the installing of a video-recording system, which indeed may record a larger amount of pragmatic information I = B * E than the system could afford. From this example, one can see that the model is able to give a natural explanation of the remarkable elusiveness of psi phenomena.
A closer inspection shows that the model reveals four distinct phases for the time development of the growing of evidence E: surprise, displacement, focusing, and establishment. Hans Berger's initial experience may serve as an example: If Hans Berger would not have been surprised, he would never had changed his scientific interest from astronomy to physiology. We do not know in detail which hypotheses he considered to explain his initial personal evidence, thus we cannot say much about his displacement phase, but finally he came out with the idea that the brain may work like a radio emitter and receiver. Now he focused himself on this assumption and found an operationalization that was surely not adequate for telepathy. We know this is not a good model for his initial experience, but finally the scientific community and history established the evidence that electrical brain potentials exist. The use of EEG has grown to a well-established paradigm. P. Berger's initial personal evidence of telepathy, however, has vanished as scientific evidence. This sounds negative, but there is no doubt that if telepathy would work as Hans Berger assumed, people would not need portable phones today.
It is clear that the nonclassical model requires new experimental requirements because the basic assumption of the classical model, namely, the accumulation of effects and finally evidence, is highly questionable in parapsychology.
A New Experimental Paradigm
The present status of scientific evidence in parapsychology and the theoretical considerations does not mean that parapsychological phenomena cannot be investigated by scientific methods anymore or that scientific evidence for psi has died out, but the resulting experimental paradigm may appear contraintuitive for a conventional parapsychologist.
The basic assumption of the MPI and the nonclassical model of evidence is that psi is no classical signal but merely a (nonlocal) correlation and that neither the signal nor the evidence of its existence can be accumulated. This means that one has to give up Sidgwick's (1992) optimistic program. Thus my first recommendation is this: Do not treat psi as a signal!
The following 11 requirements are more or less consequences of this principle:
* no accumulation, short runs
* close to the physical process, fluctuations, many channels
* correlations with physiological and psychological variables
* complete recording of the processes (no data reduction)
* no independent events (Markov chains)
* simple display (instruction unequivocally, no quirks)
* organized closure of the experiment (spatially and temporally)
* evaluation with "distance" (let it ripen!)
* conceptual replications (identical, not possible)
* nonorthogonal variables
It should be mentioned that these 11 requirements are not independent from each other. Therefore, I discuss them not necessarily in the given order.
The idea that short runs may do better than long runs (Point 1) is not new, but the reason given for it is mainly psychological. It is assumed that long runs are boring. But here I have a different reason: Long runs can be used to code a signal if the expected effect size is large enough.
Also the second requirement (Point 2) has to do with the idea that psi is not a signal. Most experimental procedures in parapsychology use REGs, and it is assumed that the physics behind it is irrelevant for the psi process. This means that the deviation from a given expectation value is an indication of a psi signal. However, if psi is considered to be a correlation in an entangled psychophysical system, it is important to get as much information as possible about the interface between the psychological and the physical description of the system. According to the MPI, the psi correlation needs as many channels as possible to establish itself, as it is described in Brunswick's (1956) lens model (Point 3). In this model the psychophysical correlation is established in a network of many possible channels. If, owing to the experimental conditions, one of the channels is blocked (for instance, by violation of my fundamental rule), the system seeks to use another channel, which is not expected to show a psi effec t. Thus, the psychophysical system is permanently changing its internal structure to maintain psychophysical entanglement. This does not mean that predictions are no more possible; one could, for instance, predict the number and strength of the sum of all correlations but not of a single one.
On the physical side, the correlations are linked with (statistical possible) fluctuations of the system. The information about it should not be given away by data reduction of the REG process as it is the case in the PEAR experiment because of sampling or output switching (Point 4). To reconstruct this information afterward requires sophisticated techniques, as the analysis of subsets of the FAMMI data by Atmanspacher and Scheingraber (2000) and by Pallikari and Boller (1999) shows.
Moreover, one of the most important results of the MPI is that pure random sequences are not good at all as random source, at least in PK experiments. Instead one should use random sequences that show internal dependencies (correlations), which, so to say, have their own "history." They are called Markov chains (Point 5). I discuss this approach in more detail in Lucadou (2000).
Point 2 holds, of course, also for the psychological and physiological part of the psychophysical system (Point 3). At present, this is generally accepted among parapsychologists as a psychological requirement and need not be discussed further. For the same reason, many experiments use fancy displays. Again the motivation is to avoid boredom. However, this conflicts some times with the requirement of the MPI, which maintains that it is important to establish an organizational closure of the psychophysical system. The crucial point is what is called in the MPI "pragmatic information of the display." It means that one must attribute a clear-cut meaning to the display taken from the context of one's instruction. This has been shown in one of my own PK experiments 15 years ago (Lucadou, 1986). From this point of view, it seems useful to keep everything that is presented to the participant as simple and unequivocally as possible (Point 6). This also includes avoiding presenting any misinformation to the participa nt. It seems plausible that such strategies give rise to suspicion and prevents entanglement in the psychophysical system.
This requirement (Point 6) does not mean that the participant must be informed about each detail of the experimental setting; on the contrary, both the experimenter and the participant should be blind in relation to the experimental hypotheses (double blindness). But double blindness is not sufficient in the light of the MPI. Because the operational knowledge of previous experiments has a real influence on any future experiment, the experimenter and the participant have also to be blind in relation to the operational result of previous experiments. I called this triple blindness (Lucadou, 1990). This is difficult to realize because, again, it has nothing to do with the subjective knowledge of the experimenter or the participant. This is one of the reasons why identical replications are not possible (Point 10).
One possibility to obtain triple blindness is to delay the evaluation of an experiment as long as possible (Point 9) and to perform parallel replication studies in the meantime. In this case, the knowledge of the first experiment cannot be used potentially in other studies. But a delayed evaluation has also another advantage: It keeps the organizational closure between the experimenter and the psychophysical system as small as possible while the organizational closure between participant and experimental setting is preserved (Point 8).
If we compare the experimental requirements of the nonclassical model with the classical ones, it must be said that many traditional experimental methods have to be modified. Many of the new requirements have been used in parapsychology beforehand, mainly with a more intuitive substantiation, and they were often very successful. Some examples include the initial card-guessing experiments of J. B. Rhine, John B. Hasted's initial metal-bending experiments, or Hans Bender's "Platzexperimente" or his investigations of Christine Mylius or of the metal bender Silvio. They did not contain long series, reliable randomization, reduction of data, or fancy displays, and they did not allow identical replication or accumulation of effects. In those classical experiments, no considerations were made to check whether the dependent or independent variables of the experiments were orthogonal (statistically independent) among each other (Point 11). In "modern" experiments, however, one strives to use only factorized (orthogon al) variables to minimize the variance of the investigated correlations. This, however, does not hold for nonlocal correlations. Here it is advantageous to use "superpositions" of variables (nonorthogonal variables), because it is only in this case that a difference between a local and a nonlocal correlation can be measured. In other words, the Bell inequality shows only differences from the classical case for nonorthogonal situations.
It is true that from the viewpoint of the classical model these experiments showed considerable methodological flaws, and of course they were not convincing for the scientific community in general. But now one may be able to understand that these experiments did not show inpressive effects because of flaws or artifacts but possibly because the investigators had some intuitive insight in the structure of psychophysical systems, long before the MPI revealed it.
(1.) This paper was an invited address delivered at the 43rd Annual Convention of the Parapsychological Association, Freiburg, Germany, August 17-20, 2000.
ATMANSPACHER, H., & SCHEINGRABER, H. (2000). Investigating deviations from dynamical randomness with scaling indices. Journal of Scientific Exploration, 14, 1-18.
BEAUREGARD, L. (1978). Scepticism, science and the paranormal. Zetetic Scholar, 1, 3-10.
BELOFF, J. (1993). Lehren aus der Geschichte der Parapsychologie [The lessons of history]. Zeitschrift fur Parapsychologie und Grenzgebiete der Psychologie, 35, 129-144.
BERGER, H. (1940). Psyche. Jena, Germany: Gustav Fischer.
BIERMAN, D. J. (2000). On the nature of anomalous phenomena: Another reality between the world of subjective consciousness and the objective world of physics? In P. van Loocke (Ed.), The physical nature of consciousness (pp. 269-292). New York: Benjamin Publication.
BRUNSWICK, E. (1956). Perception and the representative design of psychological experiments. Berkeley: University of California Press.
McCROSKEY, J. C. (1969). A summary of experimental research on the effects of evidence in persuasive communication. Quarterly Journal of Speech, 55/I, 169-176.
LUCADOU, W. v. (1986). Experimentelle Untersuchungen zur Beeinflu[beta]barkeit von stochastischen quantenphysikalischen Systemen durch den Beobachter [Experimental investigations on a possible influence of an observer on stochastical quantumphysical systems]. Frankfurt: Herchen Verlag.
LUCADOU, W. v. (1990). Was man nicht wiederholen kann-zum Problem der Replizierbarkeit bei Experimenten mit komplexen Systemen [What cannot be repeated: The problem of replication in experiments with complex systems]. Zeitschrift fur Parapsychologie und Grenzgebiete der Psychologie, 32, 212-230.
LUCADOU, W. v. (1995a). The model of pragmatic information (MPI). European Journal of Parapsychology, 11, 58-75.
LUCADOU, W. v. (1995b). Psyche und Chaos-Theorien der Parapsychologie [Psyche and chaos-Theories of parapsychology]. Frankfurt: Inselverlag.
LUCADOU, W. v. (2000). Backward causation and the Hausdorffdimension of singular events. Proceedings of Presented Papers: The Parapsychological Association 43rd Annual Convention, 138-147.
PALLIKARI, F., & BOLLER E. (1999). A rescale range analysis of random events. Journal of Scientific Exploration, 13, 25-40.
SIDGWICK, H. (1882). Presidential address. Proceedings of the Society for Psychical Research (1882-1883), 1, 7-12.
|Printer friendly Cite/link Email Feedback|
|Author:||LUCADOU, WALTER V.|
|Publication:||The Journal of Parapsychology|
|Article Type:||Statistical Data Included|
|Date:||Mar 1, 2001|
|Next Article:||DOES PRECOGNITION FORESEE THE FUTURE? SERIES 2, A LABORATORY REPLICATION AND SERIES 3, A WORLD WIDE WEB REPLICATION.|