Long but Profound......................
Arguing for an Observational Theory of
JOOP M. HOUTKOOPER
Centre for Psychobiology and Behavioral Medicine at the Justus-Liebig-University of Giessen,
Otto-Behaghel-Strasse 10, D-35394 Giessen, Germany
Abstract—The problem of devising a theory for paranormal phenomena
(psi) may be separated into, first, the basic physical mechanism and second,
the psychological aspects of how and when people are able to elicit psi.
Observational theory addresses primarily the first aspect, the basic physical
mechanism of psi. A problem is that the known types of physical interaction
do not fit the existing data.
The measurement problem in quantum mechanics can be used to hypothesize
an observer who adds information at the collapse of the wave function.
For each random event one of the possible outcomes becomes realized as the
event is being observed. The basic tenet of observational theory is: the statistics
of single events become biased if the observer is motivated and prefers
one of the possible outcomes over the other.
Features of observational theory are: its predictive power, proven with regard
to retroactive psychokinesis (PK); parsimony: the same mechanism explains
ESP as well as PK; fruitfulness, as there is a direct relationship with
quantum mechanics; and openness to crucial experiments. The falsifiability
criterion poses some problems: these are discussed in relation to Lakatos’
methodology of research programmes.
Observational theory is a solution to the problem of the mediation of psi,
thus providing a framework for the psychological problem of the manifestation
of psi. Henceforth, the observation of the outcome should be regarded as
formal part of parapsychological experiments. There is reason for cautious,
patient optimism about the acceptance of observational theory.
Since the early 1970’s, an approach to the theory of paranormal phenomena
has been developed, which has become known as “observational theory”.
What, then, is observational theory? As I will show, it is an attempt to explain
the basic mechanism of the paranormal phenomena, that is, telepathy, clairvoyance,
precognition (collectively called ESP) and psychokinesis (PK). The
generic term for these phenomena, psi, hitherto has been a collective term
without a clear basis. Observational theory, however, does provide a theoretical basis for the use of this term.
Observational theory can best be characterized
by its ‘central axiom’: “The act of observation by a motivated observer of
an event with a quantum mechanically uncertain outcome influences that outcome”.
For the explanation of psi, observational theory draws upon the ‘measurement
problem’ in quantum mechanics. This will be outlined in detail, but first
the general problem of a theory of paranormal phenomena is considered.
Paranormal phenomena have puzzled many scientists and therefore it is not
surprising that the literature is replete with theories of paranormal phenomena.
Parapsychological theories have been reviewed comprehensively by Stokes
(1987), who categorizes the multitude of theories that have been proposed
since the early 20th century. Apart from discussing observational theory,
Stokes mentions, for instance, electromagnetic theories, space-time theories,
neuropsychological theories and theories based upon mind-brain dualism.
These theories are accompanied by the problems they entail, such as the coding-
decoding problem when considering electromagnetic radiation as the
medium for transfer of information in ESP. All theories are attempts to explain
the occurrence of certain aspects of paranormal phenomena, that is, theories
are of varying scope and of varying merits.
This article is concerned with observational theory, about which I hope to
make clear that this theory provides insight in a mechanism which can explain
all categories of paranormal phenomena in way that has a certain relationship
with present day physics. Although observational theory deals mainly with the
physical aspects of psi, it may provide a way to come to a comprehensive theory
of paranormal phenomena, phenomena which then will no longer belong to
the catogory of the ‘anomalous’.
For the understanding of the problems parapsychology is concerned with,
the distinction made by Irwin (1994) is revealing. Irwin distinguishes two theoretical
1. How are psi phenomena at all possible? This is called the problem of the
mediation of psi, or the physical problem of psi.
2. How are people able to produce psi phenomena? This is called the problem
of the manifestation of psi, or the psychological problem of paranormal
One reason why progress in parapsychology has been so slow may well be
that the one problem cannot be solved without at least a fair idea of the other.
Therefore, observational theory may serve as a catalyst, even if it is incomplete
in the sense that not all of the problems it entails have been solved.
The purpose of this article is to show what the fruitful aspects of observational
theory are—and where it could be more useful if observational theory is
Physics and Psi
About Physics and Psi
In the prevalent view, physics leaves little room for paranormal phenomena.
My present objective is to make the most of the room there is within established
physics, instead of invoking completely new principles.
To grasp the difficulty of understanding psi within the framework of
physics, take for instance precognition: A future event that cannot be inferred
from present knowledge, such as an event produced by a random generator
(RNG), is predicted by a person. No physical interaction (electromagnetic,
weak, strong nuclear, or gravitational) appears to be able to explain the foreknowledge
of such a future event, which is fundamentally random and to our
normal understanding unpredictable. Even within parapsychology the phenomenon
of precognition has met with considerable skepticism: see, for instance,
As it is not excluded by the laws of physics that future events produce their
traces in the present, attempts have been made to explain precognition within
physics, for instance, by invoking advanced-wave solutions of the Maxwell
equations, but although it promises a mechanism in principle, it simply does
not work for the time scales of hours to weeks and even longer that are observed
for precognition. Moreover, if information transmission by such a
mechanism would indeed occur, physical apparatus should be realizable to
demonstrate the effect in a physics experiment and in existing experiments
such effects should have been observed.
In the literature, many aspects of physics, from electromagnetic radiation to
neutrinos and tachyons, have been proposed as responsible for the mediation
of paranormal phenomena. Here, only a few specific aspects will be mentioned,
whereas for the broader overview the review by Stokes (1987) is an excellent
reference. In order to place observational theory in its proper framework
within physics, I will first briefly outline the quantum mechanical
The Measurement Problem
In quantum mechanics, a physical system is described by its wave function.
However, the wave function cannot be observed directly, but, in the conventional
view, is identified with the statistical distribution of observed events.
To give a concrete example of conventional quantum mechanics, suppose
we have a Stern-Gerlach device, which produces electrons that upon observation
have their spin either up or down. The spin of each electron is easily detected,
since electrons with different spin are deflected differently in the apparatus.
The quantum mechanical description of one electron before observation
is a wave function that does not assign a definite up or down spin direction but
only states that the probabilities of spin up and spin down after observation are
both 50%. The wave function is, according to the usual interpretation, the
Arguing for an Observational Theory most complete description of the system, so we cannot ascribe a definite spin direction up or down to the electron before the observation. Upon observation, however, each electron obtains either spin up or spin down.
Conventional quantum mechanics considers the statistical distribution of
the electrons and identifies this with the probability distribution predicted by
the wave function. The postulate of statistical determinism ensures this (d’Espagnat,
1976). However, the step from the wave function to the observed individual
electron has not been taken without some reservations. This is the subject
of the “measurement problem”, a central issue in the interpretation of
quantum mechanics (Bell, 1987; d’Espagnat, 1976). The measurement problem
has been debated between prominent theorists who were keen on understanding
how measurement results come about, instead of being satisfied with
the mathematical formalism and the ensuing statistical predictions.
Some of these theorists (Wigner, von Neumann) have at some time suggested that only
upon arriving in the consciousness of the observer, the probabilistic event becomes
concrete, such as a spin becoming either up or down. Other quantum
physicists have discarded such a view involving consciousness as pure mysticism.
From the viewpoint of physics, consciousness is an ill-defined and illfitting
concept and it is only rational to regard it as a bad apple in the basket
and keep it outside of the theory proper as long as possible.
A way to quantify the measurement problem is to compare the information
content of the electron wave function before the electron has been observed, in
the 50% up, 50% down state, with the situation after observation, when the
electron is with certainty in one state and not in the other. The usual Shannon
information is larger in the second case. As long as the information which has
been apparently added to the electron is random, the statistical distribution
will remain as predicted. If, however, the added information is biased, we may
observe a biased statistical distribution of the electron spin.
The last half-century has seen numerous attempts to resolve the measurement
problem of quantum mechanics (QM), but no consensus has emerged.
Some recent proposals are known as ‘decoherence’ theories. Quantum decoherence
is the phenomenon that through interactions with the environment the
information required to produce the characteristic quantum interference effects
is almost immediately lost in the many degrees of freedom of the environment.
Decoherence theories assert that this produces a definite state in the
measuring instrument immediately after interaction with a quantum system,
although the formalism does not tell us which state has thus become definite.
Jeffrey Bub calls this an “ignorance interpretation”. While he discusses different
versions of this approach, he argues that the measurement problem is not
resolved by “this manoeuvre” (Bub, 1997). With regard to the relative popularity
of problems and solutions, Bub remarks at the beginning of the section on
“For most physicists, the measurement problem of QM would
hardly rate as a ‘small cloud’ on the horizon. The standard view is that Bohr
had it more or less right, and that anyone willing to waste a little time on the
subject could easily straighten out the muddle philosophers might get themselves
into” (Bub, 1997, p. 212). All irony apart, Bub alludes here to a lecture
by Lord Kelvin on the status of physics at the end of the 19th century (Kelvin,
1901). The two ‘small clouds’ of that time led to the relativity theory and to
Thus far a synopsis has been given of the measurement problem, as it can be
discussed within physics. The problem of the transition from the indefinite to
the definite seems unsolvable in present-day physics. But new resources become
available when we consider the problem in a wider context. Thus, the
role of the measurement problem alters fundamentally if new experimental
data are brought to bear upon it. This is what parapsychological research findings
will effect if they are considered within the conceptual framework of observational
Before we evaluate observational theory in this respect, we will consider another
aspect of the relationship between physics and parapsychology, namely
the concept of non-locality, which has been the cause of some misunderstandings
about observational theory.
In discussing theories of paranomal phenomena, parapsychologists often
refer to the ‘non-locality’ of quantum phenomena. The phenomenon of the
non-locality or non-separability of wave functions is best known from the Einstein-
Podolsky-Rosen paradox, a gedanken experiment that was originally
proposed to show the absurdity of quantum mechanics in 1935. Until 1964 this
remained a moot point, as it seemed to have no consequences for experimental
outcomes. In 1964, John Bell showed that the essential spatial extendedness,
or ‘entanglement’, of wave functions could have experimental consequences
and this was confirmed by Alain Aspect and his coworkers in 1982 for the first
time and in many experiments since.
An interesting corollary is that an issue which at first has to do with the understanding
of the theory and how to paint a satisfactory mental picture of it
became an experimental issue, with the possibility of quantum mechanics to
be rejected by experimental outcomes, as new experiments were invented. As
it is, quantum mechanics was vindicated and the reality of entangled quantum
states which have a non-local character confirmed.
Some attempts to explain psi have focussed on this non-locality of the wave
function. However, despite the fact that “non-local correlations” do indeed
exist, it is hard to imagine a wave function describing, in a precognition experiment,
the compound system of the human subject on the one hand and a random
number generator of which the outcomes are to be predicted on the other
hand. Furthermore, according to the formalism of quantum mechanics, no
transfer of information is possible via non-local, or EPR-, correlations. Carrying
the “non-local explanation” of psi to the extreme, sometimes the argument
that all elementary particles in the universe have been connected with each
other at the big bang is seriously put forward. Indeed, that may be the case, but
if an apparent information transmission is possible between one collection of
particles and another, why doesn’t this effect show up in physics experiments?
And, moreover, why would this specific prediction come true and why doesn’t
it connect to any other process in the universe?
Lack of specificity is a bad property of a theory. A theory of psi based upon
non-local correlations might be called a permissive theory: Everything is possible,
but this implies that nothing specific can be predicted.
But, one might ask, what about the relationship between the observer and
the observed system; can’t we regard this as a non-local correlation? Here, the
“colloquial” approach we follow reaches its limits. However, the necessary
technical treatment shows that the relationship between observer and observed
system is fundamentally different from a non-local correlation. Exacerbating
the conceptual muddle, non-local correlations have become associated with
‘holism’, a panacea which again sadly fails on the specificity argument.
At this point we may surmise that, from the point of view of present-day
physics, the only possible (that is, imaginable within present-day physics)
mechanism for psi lies in the measurement problem and in hypothesizing a
role for the conscious observer in the collapse of the wave function. This will
be argued for in the remainder of this paper.
Parapsychologists and Physicists
Scientific parapsychology has traditionally been practised mostly by psychologists
and scientists from various other disciplines, some with a definite
preference for a mind-brain dualism. For them, quantum mechanics is the bad
apple (or an esoteric gem) in the psychological basket. The other side of the
same coin is that for quantum theorists “consciousness” is the bad apple in the
physical basket. The use of the word “quantum” in a more or less mystical
sense is not any less frequent than the misuse of the term “paranormal” for any
unusual, arousal-provoking story. (There has been no reason for quantum
physicists to abandon the word “quantum” anymore than there should be a reason
for parapsychologists to abandon the words “paranormal” and “parapsychology”.)
These problems, that physicists and parapsychologists have with each
other’s technical terms, are paralleled by the two main problems of parapsychology,
namely of the mediation and the manifestation of psi.
The History of Observational Theory
The basic idea of observational theory was first formulated by Evan Harris
Walker (1973, 1975), who tried to identify quantum processes in the brain
with the quantum uncertainty in observed random events. In his view, the essential
connection between the human observer and the observed random
process can be represented by ‘hidden variables’ they have in common. Hidden
variables theories were fashionable at that time as candidate solutions of
the measurement problem of quantum mechanics.
Furthermore, Helmut Schmidt (1975) proposed a simple mathematical
model of psi, based on the premise that the probabilities of random events become
biased through being observed by the subject in an experiment. Although
the theories by Walker and Schmidt primarily explain the influence of
an observer on a quantum mechanical random process, it was Schmidt who
showed that the same mechanism can explain precognition and ESP in general.
Implicitly, Schmidt supposed “Von Neumann’s hierarchy” (d’Espagnat,
1976) to apply to unobserved—quantum mechanical—random events. Therefore,
in these theories the role of the observer in quantum mechanics is the central
Not surprisingly, the generic name ‘observational theories’ was coined
shortly afterward (Houtkooper, 1977), followed by a review by Millar (1978).
This author scrutinized observational theory with regard to the interpretation
of quantum mechanics, both from the experimentalist’s viewpoint and from
the viewpoint of the philosophy of science (Houtkooper, 1983). To my initial
surprise, the most fruitful view was provided by the orthodox Copenhagen interpretation
of quantum mechanics, or, in Bub’s terminology, the “orthodox
Dirac-von Neumann interpretation”.
The conclusion was that the phenomena which are usually regarded as representing ‘backward causation’, such as precognition
and retroactive PK, can be understood by assuming that random
events which have not yet been observed remain in a state of ‘indefinite reality’.
They only obtain a definite status upon observation. In this view, which is
compatible with quantum mechanics, ordinary causal relationships between
events can be maintained. The cost of maintaining causality consists of having
to give up realism with regard to quantum mechanical random events.
In the 1980’s observational theory has been discussed in various articles, of
which those by Schmidt (1984) and Walker (1984) may be mentioned, although
not reviewed here. Applications in experiments have been scarce and
the elaborate protocols have apparently not been conducive to the occurrence
of psi in these experiments (Houtkooper et al., 1989; Johnson & Houtkooper,
1988; Schmidt & Dalton, 1999).
The Unification of Paranormal Phenomena
The idea that PK and ESP belong to the same class of phenomena goes back,
at least, to the mediumistic scéances of the 19th century. Observational theory
for the first time provides a relationship and even a common mechanism for
both phenomena. Schmidt (1975) has made this very explicit by ‘constructing’
a paragnost, which exhibits precognition, using a scheme involving a source of
PK. PK is to be regarded as the basic phenomenon, acting on quantum mechanically
random events. Assuming Von Neumann’s hierarchy means that unob-
served random events remain in a state of ‘indefinite reality’. Upon observation
the particular outcome of a random event is determined. This explains the
apparent success of experiments with retroactive PK.
In a precognition experiment, the target is a picture or symbol, determined
randomly some time before the experiment proper. The experiment consists of
a subject guessing, using his or her intuition or imagery, what the target might
be. After this, the correspondence between target and guess is determined. The
outcome of the experiment is therefore determined at the moment this correspondence
is observed and the preference of the observer of the correspondence
between target and guess must be a major factor in determining the success
of the experiment. In other words, precognition is simply retroactive PK
in disguise. The same explanation holds for the other ESP phenomena, telepathy
and clairvoyance. Therefore, the paranormal phenomena have a unified
explanation within observational theory.
The ramifications of Schmidt’s unification theorem, such as the a priori
pairwise stochastic independence of target, guess and correspondence, and the
application to psi in real-life situations, may serve to further develop observational
Observational Theory and Lakatos’ Methodology of Research Programmes
When considering pros and cons of observational theory, we may regard observational
theory as a research programme in the sense of Lakatos (1978).
Lakatos’ view of the acceptance and rejection of scientific theories is different
from Popper’s: If theories have apparently been falsified, they protect themselves
against falsification by secondary theories or “auxiliary hypotheses”.
The basic theory which is thus defended is the “core theory” and it has some
resemblance to Kuhn’s “paradigm” (Kuhn, 1962). To cite an example of
Lakatos: The anomalous precession of the perihelion of the planet Mercury
was well known in the 19th century; instead of regarding this as a falsification
of Newton’s law of gravitation, astronomers hypothesized a new planet between
the sun and Mercury as the cause and searched for it, but in vain. The
cause of not finding it was blamed on the difficulty of observing a small object
close to the sun. It was Einstein’s general relativity theory which explained the
observations without the need for an extra planet.
Therefore, according to Lakatos, theories are accepted or superseded, not by
the outcome of a crucial experiment or observation, but by their connections
with other theories and their explanative power on the one hand and their need
for protection to explain away the falsifications and to defend themselves
against conflicting experiments on the other hand. These characteristics are
called, respectively, the positive and negative heuristics of a theory. The theory,
as it develops its different heuristics, is called a research programme.
Observational theory can be best characterized by its central axiom: “The
act of observation by a motivated observer of an event with a quantum
mechanically uncertain outcome influences that outcome.”
This is the core theory in the sense of Lakatos. By a “motivated observer” is
meant an observer with a conscious or unconscious preference for one specific
outcome over the other possible outcome(s).
Thus, we may regard observational theory as a research programme in the
sense of Lakatos. Next, we examine the positive and negative heuristics of observational
theory, in order to show to what extent observational theory can be
considered a fruitful approach and to what extent it has some unwanted characteristics.
Positive Heuristics of Observational Theory
1. The phenomenon of retroactive PK was predicted by observational theory:
The moment of observation is crucial instead of the moment of generation
of the random events. In experiments with retroactive PK, or
‘retro-PK’, random events are generated and recorded some time before
the proper experiment is conducted. In the experiment, the random
events are presented to the subject, possibly without he or she being
aware of the fact that the events are not being momentarily generated.
As it is assumed, based on “Von Neumann’s hierarchy” (d’Espagnat,
1976), that the randomness of the recorded events is not compromised
by having been stored, this delay of the observation should make no difference
for the PK effect. In fact, this idea has been corroborated by experiments
with prerecorded trials. Moreover, in these experiments with
retroactive PK some good experimental corroboration has been found
for the influence being exerted at the time of feedback to the subject, and
not at the time of generation of the prerecorded trials (Schmidt, 1976).
This means observational theory has proven its power to predict new
phenomena and that the central axiom has found direct support.
2. Precognition is accounted for without needing an (ad hoc) extra mechanism.
Schmidt (1975) showed the equivalence of precognition and PK if
one assumes an observational model. As clairvoyance and telepathy are
equivalent to precognition in real time, the various paranormal phenomena,
psychokinesis, telepathy, clairvoyance and precognition, can be
unified into one basic mechanism. In other words, observational theory
is parsimonious in comparison to theories which invoke different principles
for different paranormal phenomena.
3. Observational theory predicts specifically the possibility to manipulate
effects, such as in repeated retroactive PK, so that detailed models can
be formulated and tested. Such a model, the hierarchical model
(Houtkooper, 1983), was devised on the basis of indications from experiments.
It implies a diminishing influence of the observations by later
observers of the same random events, through a “psi-absorption” parameter
assigned to observers. The designs based upon this model promise
experimental results with a better reproducibility than had been hitherto
4. The use in PK experiments of random processes, like dice and RNGs, instead
of sensitive instruments to measure small static forces, can now be
justified on the grounds of the consideration of psi effects as effects acting
upon the quantum mechanical uncertainty in these random processes.
Experimental corroboration is found in PK experiments with a gifted
subject who could reproducibly influence the motion of cubes rolling on
a horizontal surface, but not the motion of a ball rolling down an incline
(Forwald, 1977; Pratt & Forwald, 1958).
5. The experimental finding that psi effects are independent of the complexity
of the random process involved (Schmidt & Pantas, 1972), can
be explained by the act of observation as the crucial event at which a psi
effect is mediated. Devices used as PK target systems, like numerous
dice in one throw, noise-based RNGs and even the speed of radioactive
decay of uranium, are highly non-transparent to the PK agent. The success
of these devices in ensnaring ostensible psi-effects has hitherto
been a source of concern for theorists. Observational theory predicts on
the one hand the success of devices which feature complex and even microscopic
randomness and on the other hand the scarcity of controlled
experiments in which macroscopic forces have been measured and revealed
6. Observational theory bears directly upon quantum mechanics, in that
the existence of psi effects contradicts the principle of statistical determinism
(d’Espagnat, 1976). The important limiting principle is that the
deviations from this principle all involve the observation of quantum
mechanically uncertain variables by a motivated observer. Thus, observational
theory has a direct conceptual relationship to quantum mechanics.
7. The unsolved problems in observational theory appear to have their analogues
in quantum mechanics, namely the problem of what constitutes
an observation (in quantum mechanics: a measurement) and the problem
of how to determine the order in time of observations (in quantum
mechanics: the order in time of spatially separated measurements). That
is, the direct relationship to quantum mechanics may yet prove to be
fruitful for quantum mechanics as well as for observational theory.
8. Implicitly, parapsychological researchers still use a transmission paradigm
in their design of experiments. The major drawback of this view,
or rather, prejudice, is that conditions of observation of random events
are not noted. An example is the very successful series of ganzfeld experiments
reported by Honorton (Bem & Honorton, 1994; Honorton et
al., 1990), where it is unclear who, either sender or experimenter, is the
first to observe the target-guess correspondence. Ganzfeld experiments
are experiments on the ostensible transference of information between a
‘sender’ and a ‘receiver’, in which the receiver is put in a situation of
mild sensory deprivation. The receiver speaks out his or her impressions,
the ‘mentation’. Usually in these experiments, the sender may
hear this mentation. Even if this is the case, and the sender listens to the
receivers’ mentation, it is unclear whether the sender is the first to know
if the session was a success. Accepting the possibility of observational
theory means that the importance of the observation of outcomes of an
experiment has to be stressed, as a formal part of the experiment. Also, if
the transmission paradigm is regarded as an alternative to observational
theory, observational theory forces it to reveal itself openly, including its
auxiliary assumptions, such as the issue of real time transmission—
Parker and Persson (1999) are taking up this challenge—and the properties
of sender and receiver.
9. Not only PK, but also ESP experiments, can and should be designed
using observational theory. Both a blind-matching and a ganzfeld experiment
have been designed and carried out by the author in collaboration
with others: First, the ganzfeld experiment reported by Houtkooper et al.
(1989) aimed, among other things, at showing the role of feedback in
ganzfeld experiments. An observational part was added to the design of
an otherwise ordinary ganzfeld experiment. Results of the experiment
revealed no psi, so that corroboration for neither observational nor
transmission paradigm was obtained. The second was the blind-matching
experiment reported by Johnson and Houtkooper (1988). Likewise,
it was unsuccessful in revealing straightforward psi-effects, so that the
effect of later observations could not be ascertained. Lack of experience
and familiarity with the somewhat cumbersome observation procedures
may in part have been the cause of these disappointing results, but nevertheless
these experiments may show that observational theory can very
well be applied to experimental designs.
10. The apparent effects of experimenter psi can be understood as effects of
the experimenter as a later observer (Weiner, 1982; Weiner & Zingrone,
1986, 1989) of the data of his or her experiment. Apart from normal,
psychological experimenter effects, apparent effects of the experimenter
have been observed in situations where normal sensory channels do not
apply. These are ostensible cases of parapsychological experimenter effects,
or ‘experimenter psi’ (Houtkooper, 1994; Irwin, 1994; Schmeidler,
Negative Heuristics of Observational Theory
We should be well aware of the negative heuristics of a research programme,
because these would eventually be the cause of its possible obsolescence and
1. The Central Axiom is, by itself, non-falsifiable: In principle, all experimental
outcomes that can be discussed have at one time been observed.
Also, motivation can hardly be reliably measured, apart from its end-effects.
This means we may have to get hold of specific properties of specific
observers and study their individually different effects. Moreover,
the influence of later observers of the same data (which may be in the
form of the overall outcome of the experiment) has to be constrained in
some way, because otherwise no experiment is ever finished. This requires
viable and fruitful models of the influence of later observers, such
as the hierarchical model by Houtkooper (1983), which is one possible
solution to the problem of later observers.
2. The number of parameters even the simplest model requires makes it
hard to falsify in a practical experiment. To keep the observer’s motivation
and disposition constant is a requirement that is hard to satisfy.
These problems, albeit technical, need to be overcome if progress is to be
made. Research with outstanding subjects is one way of solving this and
another would be to find conditions which lead to reproducible results.
3. Even in the case of the hierarchical model, which can be considered to be
one of the simplest models, parameters can easily be adapted to account
for nearly any experimental outcome as long as the effects are as weak as
they usually are. Moreover, auxiliary hypotheses can be put forward, for
instance concerning the observation of outcomes by experimenters, to
protect the model against falsification. The further sophistication of experimental
designs is the obvious way to overcome these difficulties.
Observational theory is a viable research programme for paranormal phenomena.
Its main properties are a) predictive power (with regard to retroactive
parsimony (the paranormal phenomena, ESP and PK, can be explained
by one and the same mechanism), c) fruitfulness (a direct relationship
to quantum mechanics), and d) openness to crucial experiments distinguishing—
as far as that is possible within Lakatos’ philosophy of science—between
the predictions of, on the one hand, observational theory and, on the
other hand, alternative theories and models.
Thus, as shown by the positive heuristics, the questions to be answered by
parapsychological experiments become more interesting. In Lakatos’ terminology,
observational theory must be regarded as a positive problem shift.
The negative heuristics of observational theory stem mainly from the requirement
of falsifiability. This criterion should be applied to specific theories
and models, but not to research programmes or core theories.
This points to the necessity to develop specific theories, such as the hierarchical model. Furthermore there is the very practical problem of small effects and rather imperfect
reproducibility. Although observational theory is no cure for all problems, it
offers an explanation for the mediation of psi. This provides experimental research
involving the manifestation of psi with a more fruitful basis.
Research programmes are superseded by superior ones. The alternatives to
observational theory are unspecified and ad hoc transmission ‘principles’ for
ESP and likewise field models for PK, or, more recently discussed is Decision
Augmentation Theory (DAT): May, Utts, and Spottiswoode (1995) proposed a
mechanism for the mediation of psi in that the subject basically uses his or her
intuition to decide between targets in an ESP task, whereas PK is explained by
waiting to start a run until the random event turns out favorably with regard to
the target—whereby PK is explained as ESP. May et al. find support in existing
data on the dependency of effect size on run length (May, Spottiswoode et al.,
1995), which has been criticized by Dobyns and Nelson (1998). Is DAT a superior
An experiment which would discriminate between DAT and observational
theory is quite obvious: One has to separate the subject who pushes the startbutton
from the subject who observes the outcome. Experiments with retroactive
PK are a case in point: Schmidt’s (1976) experiments are hard to explain
as an effect of DAT, except as a contorted case of an experimenter effect. Furthermore,
DAT has some elements in common with observational theory, such
as ‘state selection’, but in contrast, DAT lacks a physical basis for ‘intuition’,
whereas observational theory indicates the point in conventional physics
where psi can be fitted in. This point in physics is the observation of a quantum
mechanically random event by an intent observer. The conclusion must be that
DAT is conceptually less fruitful and finds as yet less experimental support
than observational theory.
As observational theory indicates the point in physics where “the action is”,
it is clear the same indication specifies where the psychology cum neurophysiology
of paranormal phenomena should aim, namely at the same intent observer,
in parapsychological experiments, as he or she observes a quantum mechanically
random event which is the the subject of his or her intention.
Parapsychology will not make progress if it doesn’t come to grips with specific
models of the underlying phenomena. Certainly, the difficult area of relationships
with psychological variables will have a hard time if it doesn’t take
into account the best available assumptions about the basic mechanism of psi.
Observational theory provides such a basic mechanism and this should bear its
fruits if applied in the design of future experiments.
How likely is it that experimental parapsychologists take heed of observational
theory in the future? A note of caution is in order: The very idea of a theory
that specifically includes the experimenter himself will have profound difficulties
in establishing itself. This is an element that is new to science and one that
implies requirements of experimental methodology that still have to be explored.
In fact, the only example—that I know of—of an intimate interaction between
the researcher and object of research is in primatology (De Waal, 2001). In
parapsychology, the situation is more extreme: Every wish or motivation of the
experimenter may well influence the result of an experiment decisively.
The methodological difficulties this entails are by no means close to a solution.
Nevertheless, parapsychology fits well in the picture painted by Dijksterhuis
(1950) of the evolution of science: The ancients started their scientific endeavours with astronomy. Physics developed at a much later stage. Biology became
a science after that and only in the second half of the 19th century was
The obvious ‘inward’ movement, namely from the
cosmos to the mental phenomena of human being, inexorably seems to lead to
a science where the researcher is part and parcel of his or her object of research.
To accept this role will be a psychological stumbling block, but, despite our
very wish and need to have an objective science, it is the challenge apparently
posed by paranormal phenomena.
This paper is based on a talk given at the 17th Annual Meeting of the Society
for Scientific Exploration at the University of Virginia, 28–30 May 1998.
I wish to thank Dennis Dieks, Kamiel Houtkooper and Dean Radin for helpful
and stimulating comments on earlier drafts of this paper and the Institut für
Grenzgebiete der Psychologie und Psychohygiene for financial support that
has facilitated this study.
Bell, J. S. (1987). Speakable and Unspeakable in Quantum Mechanics. Cambridge, United Kingdom:
Cambridge University Press.
Bem, D. J., & Honorton, C. (1994). Does Psi exist? Replicable evidence for an anomalous process
of information transfer. Psychological Bulletin, 115(1), 4–18.
Bub, J. (1997). Interpreting the Quantum World. Cambridge, United Kingdom: Cambridge University
Dijksterhuis, E. J. (1950). De Mechanisering van het Wereldbeeld. [The Mechanization of the
World-Picture]. Amsterdam: Meulenhoff.
Dobyns, Y. H., & Nelson, R. D. (1998). Empirical evidence against Decision Augmentation Theory.
Journal of Scientific Exploration, 12, 231–258.
D’Espagnat, B. (1976). Conceptual Foundations of Quantum Mechanics (2nd ed.). Reading, MA:
Forwald, H. (1977). A PK experiment with a single ball rolling on a decline. European Journal of
Parapsychology, 2(1), 4–14.
Honorton, C., Berger, R. E., Varvoglis, M. P., Quant, M., Derr, P., Schechter, E. I., & Ferrari, D. C.
(1990). Psi communication in the ganzfeld: Experiments with an automated testing system and
a comparison with meta-analysis of earlier studies. Journal of Parapsychology, 54, 99–139.
Houtkooper, J. M. (1977). A study of repeated retroactive psychokinesis in relation to direct and
random PK effects. European Journal of Parapsychology, 1(4), 1–20.
Houtkooper, J. M. (1983). Observational Theory: A Research Programme for Paranormal Phenomena.
Lisse, The Netherlands: Swets & Zeitlinger.
Houtkooper, J. M. (1994). Does a meta-analysis demolition effect exist? Abstracts of 18th International
Conference of the Society for Psychical Research (Bournemouth, England, 2–4 September
Houtkooper, J. M., Gissurarson, L. R., & Haraldsson, E. (1989). Why the ganzfeld is conducive to
ESP: A study of observational theory and the percipient-order effect. European Journal of
Parapsychology, 7(2–4), 169–192.
Irwin, H. J. (1994). An Introduction to Parapsychology (2nd ed.). Jefferson, NC: McFarland.
Johnson, M. U., & Houtkooper, J. M. (1988). The effect of anxiety on scoring and observer-effects
in a blind matching test with targets of specific and emotional implications to the subjects. In
D. H. Weiner & Morris, R. L. (Eds.), Research In Parapsychology 1987, 56–59. Metuchen,
NJ: Scarecrow Press.
Kelvin, Lord. (1901). Nineteenth century clouds over the dynamical theory of heat and light.
Philosophical Magazine, 2, 1–40.
Kuhn, T. S. (1962). The Structure of Scientific Revolutions. Chicago: University of Chicago Press.
Lakatos, I. (1978). The Methodology of Scientific Research Programmes. Cambridge: Cambridge
May, E. C., Spottiswoode, S. J. P., Utts, J. M., James, C. L. (1995). Applications of Decision Augmentation
Theory. Journal of Parapsychology, 59, 221–250.
May, E. C., Utts, J. M., & Spottiswoode, S. J. P. (1995). Decision Augmentation Theory: Toward a
model of anomalous mental phenomena. Journal of Parapsychology, 59, 195–220.
Millar, B. (1978). The observational theories: A primer. European Journal of Parapsychology,
Parker, A., & Persson, A. (1999). Using the ganzfeld in qualitative process research: An interim
report. Proceedings of the 42nd Annual Convention of the Parapsychological Association,
(Palo Alto, CA, August 4–8, 1999), 270–277. Abstract in Journal of Parapsychology, 63, 219.
Pratt, J. G., & Forwald, H. (1958). Confirmation of the PK placement effect. Journal of Parapsychology,
Schmeidler, G. R. (1977). Methods for controlled research on ESP and PK. In Wolman, B. B.
(Ed.), Handbook of Parapsychology (pp. 131–159). New York: Van Nostrand Reinhold.
Schmidt, H. (1975). Toward a mathematical theory of psi. Journal of the American Society for
Psychical Research, 69, 301–319.
Schmidt, H. (1976). PK effect on pre-recorded targets. Journal of the American Society for Psychical
Research, 70, 267–291.
Schmidt, H. (1984). Comparison of a teleological model with a quantum collapse model of psi.
Journal of Parapsychology, 48, 261–276.
Schmidt, H., & Dalton, K. (1999). PK tests with repeated efforts on prerecorded targets. Journal of
Parapsychology, 63, 275–283.
Schmidt, H., & Pantas, L. (1972). Psi tests with internally different machines. Journal of Parapsychology,
Steinkamp, F. (1999). Testing clairvoyance and precognition by manipulating probabilities: A
conceptual assessment of the experimental literature. Journal of Parapsychology, 63, 99–130.
Stokes, D. M. (1987). Theoretical parapsychology. In Krippner, S. (Ed.), Advances in Parapsychological
Research 5 (pp. 77–189). Jefferson, NC: McFarland.
De Waal, F. B. M. (2001). The Ape and the Sushi Master. Cultural Reflections by a Primatologist.
London: Allen Lane the Penguin Press.
Walker, E. H. (1973). Application of the quantum theory of consciousness to the problem of psi
phenomena. In Roll, W. G., Morris, R. L., & Morris, J. D. (Eds.), Research in Parapsychology
1972 (pp. 51–53). Metuchen, NJ: Scarecrow Press.
Walker, E. H. (1975). Foundations of paraphysical and parapsychological phenomena. In Oteri, L.
(Ed.), Quantum Physics and Parapsychology (pp. 1–44). New York: Parapsychology Foundation.
Walker, E. H. (1984). A review of criticisms of the quantum mechanical theory of psi phenomena.
Journal of Parapsychology, 48, 277–332.
Weiner, D. H. (1982). The once and future psi source: A review of the literature on future-observer
effects. Paper presented at the Southeastern Regional Parapsychological Association Conference,
Chapel Hill, NC, February 1982. Abstract in Journal of Parapsychology, 46, 58–59.
Weiner, D. H., & Zingrone, N. L. (1986). The Checker Effect revisited. Journal of Parapsychology,
Weiner, D. H., & Zingrone, N. L. (1989). In the eye of the beholder: Further research on the
“Checker Effect”. Journal of Parapsychology, 53, 203–231.