Sie sind auf Seite 1von 15

0

The 'Maharishi Effect' of


Transcendental Meditation
Invincible or Invisible?

Joantine Berghuijs
January 2006
Paper for the
Master Seminar 'Philosophy of Science for Religious Studies'
Instructor: Prof.dr. W.B. Drees

1
2
3
4
5

Contents
Introduction
The International Peace Project in the Middle East
Analysis of Controversies
How to Address Heterodox Claims
Conclusions
References

1
3
5
10
13
14

1. Introduction
Some time ago, my attention was drawn to statements that meditation techniques, and especially
Transcendental Meditation, have shown to have statistically significant effects. These effects comprise
not only reducing anxiety, but also producing episodes of respiratory suspension, correlated with
reports of experiences of 'pure' consciousness - as often described in traditional texts - and promoting
self-actualization. The results found would support claims that meditation can be a good tool for
promoting unusual levels of individual growth, as made by many religions. (Shear, 2001, p.287-288,
292). There seems much less likelihood, the author adds, that future results will corrobate the more
extravagant claims of supernormal abilities, like levitation and invisibility (ibid., p.292).
Another of the more extravagant claims adherents of Transcendental Meditation (TM) make, is that it
can reduce accidents and crime and promote peace among the surrounding people, when a certain
number of practitioners meditate together. This is called the 'Maharishi Effect'. Assuming that the
effects on the physical level have been proven, could it be possible? If so, would that not be a
challenging alternative for or at least a complement to political and armed struggle? Has this claim
also been investigated? Yes, it has been, and passionately. I came across a fervent discussion, in- and
outside the scientific world, a discussion that continues until the present day.
In this paper I analyse this discussion. Of course, the most interesting question is about the truth of the
claim. In order to get near to an answer to that question, I will show you the observations and
arguments brought forward by the proponents and the opponents. I will also reflect upon criteria to be
used by the scientific community in assessing extraordinary claims.
This exercise will lead us to the conclusion that however challenging the presented data in support of
the Maharishi Effect are, they are suspect for many reasons, ranging from strict scientific criteria to
social factors that are related to credibility.

Transcendental Meditation
This paragraph is based on Chryssides (1999,292-303).
Transcendental Meditation was founded by the Maharishi Mahesh Yogi, who originated within the
Hindu tradition. He studied physics at Allahabad University and graduated in 1940. For 13 years
thereafter he studied Hindu mystical techniques under Swami Brahmananda Saraswati (1869-1953),
'Guru Dev'. In the late 1950s he introduced TM to the West. TM teaches to transcend the physical
world to gain fulfillment from within, leading to pure 'Being' or 'Bliss-consciousness'. The knowledge
to do this has been brought to humanity by Krishna, the Buddha, Shankara, and Patanjali, but it has
been lost with the passage of time, Maharishi teaches. Therefore he has modified the techniques given
by these men, and made them more consistent.
The technique is very easy, and can be mastered within a few days. There is a formal initiation
ceremony in which the student appears before the teacher with a clean white handkerchief, flowers and
payment. The teacher gives the student a personal mantra (which has to be kept secret) on which he or
she should meditate twice daily for 15 to 20 minutes at each session. After some individual and group
sessions during the next few days, the course is completed. The initiation ceremony is compulsory. A
mantra taken from a book will not work, because it may be the wrong kind of mantra, and the power
of the mantra lies in the sound, which you do not know exactly.
TM makes no claim to be a form of religion, preferring to be regarded as a 'technique', although it has
clear ritual elements in the form of the initiation ceremony and the mantras derived from Indian
religious traditions.
The Maharishi has described how the TM technique works. According to the third law of
thermodynamics, when physical temperature or activity decreases, a system becomes more orderly.
The same is true, he teaches, of the mental world. During meditation, the Self becomes calmer and
better ordered. The brain shows a marked increase in alpha waves. As one penetrates beneath the level

2
of conscious awareness, one discovers subtler levels of thought, until finally one is able to penetrate
the source of all thought, which is Being itself, pure consciousness, totally calm and still. It is said that
the benefits of TM emerge on a number of levels. On the physical level, breathing rate and metabolic
activity will slow down, and there is a drop in the levels of certain stress-related chemicals in the
blood. A lot of medical disorders are said to be improved by TM. In 1993 a number of medical doctors
in Britain expressed their support for TM. On a societal level, TM claims that where sufficient
numbers of people practise TM in an area, one can expect fewer accidents, and reduced crime rates.
Peace can be brought nearer. This is known as the 'field effect' or the 'Maharishi effect'. It reflects the
belief that consciousness is like a 'field'. TM claims that the Maharishi effect can be and is
scientifically verified; its literature refers to some 500 studies, some by well-accredited American
universities. These claims have not gone unchallenged.
In 1976 the TM-Sidhi programme was introduced. 'Sidhi' literally means 'perfection', and the Sidhi is a
disciple who possesses advanced supernatural spiritual powers, including 'yogic flying', to which the
media have drawn considerable attention.
Investigations
There have been several attempts to prove the Maharishi effect, mainly during the 1980s and 1990s
(e.g., Orme-Johnson et al, 1988; Gelderloos et al, 1990; Hagelin, J.S. et al, 1999). These comprised
experiments with group meditation during a period of weeks or months, and gathering data on events
like car accidents, terrorist attacks or war deaths. Remarkably strong correlations were presented
between group sizes and number of events, suggestive of positive influence of the meditation group.
On their websites and in their publications, TM-adherents always point at the same handful of
experiments, which were published in refereed journals.
I shall focus my analysis on the discussion around the most discussed of these experiments, the
'International Peace Project in the Middle East; the effects of the Maharishi Technology of the Unified
Field', which was published in the Journal of Conflict Resolution (Orme-Johnson et al, 1988), be it
after reading by extra referees, and with feelings of ambivalence on behalf of the editor (Russett,
1988). It was followed by several comments (Duval, 1988; Schrodt, 1990; Orme-Johnson et al, 1990).
Fales and Markovsky (1997) drew more theoretical approaches into the discussion; this article was
answered by Orme-Johnson on the TM website on Invincible Defence Technology provisionally and
in a more popular language (www.invinciblemilitary.org/articles/heterodox_theory_rebuttal.html ).
These articles seem representative for the discussion on the Maharishi effect.

2. The International Peace Project in the Middle East


Orme-Johnson et al (1988) state that, according to Maharishi, violent conflict is an expression of stress
in collective consciousness. There is a reciprocal relationship between individual and
collective consciousness. A level of collective consciousness corresponds to each level of
social organization.
The authors introduce the following model for the Maharishi effect (ME), expressed as the population
size influenced by the TM techniques, as:
ME = aN1 + bN22
in which
N1 = the number of independent meditators distributed throughout the population
N2 = the number of individuals practicing TM-Sidhi collectively in one place
a and b are empirical constants about the size of 102 for N > 100.
The model is said to be based on a field theoretical model describing the coherent superposition of
amplitudes; the authors compare the working to that of coherent physical systems like lasers. They
conclude that known fields like electromagnetic and gravitational fields, and strong and weak
interactions, are of an inappropriate scale to mediate direct human effects. They look for an
explanation of the unified field of consciousness in the ultimate scale of superunification, the Planck
scale of 10-33 to 10-44, where the fundamental forces and matter fields are said to become fully unified.
They see striking parallels between the description of the unified field by the objective approach of
modern science and pure consciousness by the subjective approach of the ancient Vedic tradition, and
they suggest that these parallels may reflect different perspectives on the same fundamental reality.
The experiment carried out to confirm the theory was carried out in Jerusalem from August 1 through
September 30, 1983.
The independent variable was the meditation group size. Meditators from Israel were invited to
participate. The number of days of their participation differed per person. Meditation sessions were
twice daily, once in the morning and once in the late afternoon. From August 15 to August 27 an
advanced course was offered as an extra incentive to participate. N 2 ranged from 65 to 241. N1 was
38,000 for Israel and 2,000 for Lebanon, populations were 429,000 for Jerusalem, 5,304,000 for Israel
(including West Bank and Gaza Strip) and 7,905,000 for Israel and Lebanon together; their
populations added up 7,905,000. Based on the model equation, the group size would have to be 197 to
create coherence for Israel and Lebanon combined, and 122 to affect Israel alone.
The experimental hypotheses were lodged in advance of the study with independent review boards of
scientists in North America and Israel. Upon arrival in Israel, the authors met with Israeli scientists to
further reduce the list of dependent variables proposed earlier.
The dependent variables used after completion of the experiment were daily time series data, from
publicly available sources:
Jerusalem: auto accidents, fires
Israel: total crimes per day, stock market index, national mood
Lebanon: war deaths, war intensity scale
The national mood and war intensity scale were based on blind scoring of news bulletins in the
Jerusalem Post by independent raters.
In addition, composite indices were formed for each locale, taking arithmetic means of the
standardized variables for Jerusalem, Israel and Lebanon; individual variables were inverted when
necessary, so that a positive deflection in an index indicated a positive change in the quality of life.
The Overall Composite Index was composed of all standardized variables, excluding war deaths,
because war deaths seemed to be highly correlated with war intensity scale. Finally, the
Variability Index was a measure of the variation each day of the 6 variables in the Overall Composite

4
Index. Increase of coherence would have to be reflected in an increase of the Overall Composite Index
and a decrease of the Variability Index.
Upon analysis, the group size was divided in four quartiles (the first being the lowest). Results showed
a decrease, measured in the fourth quartile as compared to the first, of 45% for war intensity, 76% for
deaths in Lebanon, 12% for crime in Jerusalem, 8.8% for crime in the rest of Israel, 30% for fires in
Jerusalem, and 34% for auto accidents in Jerusalem. The last four decreases mentioned were only
significant for one of the two statistical methods used. The Composite Indices for Jerusalem, Israel and
Lebanon and the Overall Composite Index showed significant increases, with values between .75 and
1.7 standard deviations for the fourth quartile as compared to the first, whereas the Variability Index
decreased significantly by .99 of the standard deviation. For the Jerusalem and Israel composite
indices, the effect reached significance by the 2 nd quartile, when the group size lay between 125 and
157, whereas for the Lebanon Composite Index the effect did not reach significance until the 4 th
quartile, when the group reached 180-241. These results, the authors state, provide general support for
the prediction made from the model, that Israel would be affected by a group size of 122 or more,
whereas Lebanon would not be affected until the group size exceeded 197. The results show that
combining variables into composite indices generally resulted in more clear-cut effects. The authors
think this is logical if TM produces a global effect influencing all variables, because adding the
variables would enhance the signal-to-noise ratio.
With respect to causality the authors state that, although the group size was not completely random,
the quartiles in which it was broken were essentially randomly distributed over the duration of the
experiment. In that case the increased impact for the larger quartiles would support causal
interpretation. In addition, cross-correlation and transfer function analyses showed that none of the
dependent variables led the group size, whereas there was consistent evidence of the opposite. A
possible alternative explanation, being that knowledge of the war had an immediate influence on level
of group participation, was rejected, because conflict events in Lebanon were almost always reported
in the newspapers the day after they occurred. Also, participants came from all over Israel, usually for
a week or more; the number of 'drop ins' from the immediate Jerusalem area was quite small.
Nevertheless, there were indications of a lag between peaks in group size and effects on the Lebanese
conflict.
The authors claim that the results support the interpretation of an underlying unifying influence being
produced on many diverse systems simultaneously. They suggest an immediate priority for social
scientists and policymakers to investigate the large-scale application of this simple and nonintrusive
technology to resolve international conflicts.

3. Analysis of Controversies1
Theory
There was a lot of critique on the theory:

The key terms are not or only roughly defined (consciousness, collective consciousness,
pure consciousness) (Fales and Markovsky, 1997, p.516).

Orme-Johnson (year unknown) replies that the means by which anyone can directly
experience the state of pure consciousness is operationally defined by the TM technique, and a large
body of research over the last 35 years has provided objective physiological definition of the state.

The theory/hypothesis is outside the current scientific paradigm, absurd (Russett, 1988,
p.773/4), or anomalous (Duval, 1988, p.816).

Orme-Johnson et al (1990, p.758) points out that William James and others proposed theories
of collective consciousness almost a century ago, which at the time could not be tested.

The plausibility of the theory is low. Detailed experimental evidence is lacking for Planck
scale phenomena. Particle and nuclear physicists find the suggestion that some quantum-mechanical
properties of physical fields match characteristics of consciousness obscure. Parallels between the
Vedas and contemporary unified-field theories are nothing more than arbitrary formal isomorphisms
(e.g., five special Vedic terms called transmatrans, and five spin-types in quantum mechanics),
metaphors and vague analogies (Fales and Markovsky, 1997, p.518).

Orme-Johnson et al (1990, p.757) cite Russett (1988) stating that the criteria for plausibility
are unclear in the social sciences. Therefore, they say, the study must be judged primarily on the basis
of its experimental rigor and replicability (p.757); in fact, the effect was replicated in six subsequent
interventions (p.766). Replying to Fales and Markovsky (1997), Orme-Johnson (year unknown) states
that (1) the basic theoretical tenets of the Maharishi Effect comprise the Perennial Philosophy, the
most ancient and ubiquitous knowledge of the human race, and (2) the central proposition that
individuals directly interact at a distance is currently the focus of several active research programs.
Furthermore, they say, the criticized physicist John Hagelin is in fact an award winning Harvard PhD
in theoretical physics. His work, in the tradition of Planck and Einstein, is on the leading edge of
modern science and among the most frequently quoted in physics. This, together with the fact that he
has engaged in hundreds and hundreds of hours of discussions of consciousness with Maharishi, the
worlds foremost authority of consciousness, makes him uniquely qualified to formulate an integration
of consciousness with modern science.

The theory does not explain (a) how group meditation affects the unified field; (b) how these
effects cause changes in the actions of individual human beings; and (c) how those individual actions
have their claimed social impact (Fales and Markovsky, 1997, p.518).

Orme-Johnson (year unknown) replies that the practice in science is that theories are created,
tested, and modified according to the data.

There is no rationale for the threshold built into the model; the cut-offs at N = 100 are
arbitrary. Such a discontinuity produces an awkward behavioural model (Fales and Markovsky, 1997,
p.516).
1

In this chapter, the check marks


give the points of critique;
give the response of the TM-researchers on the critique, and

give my own analyses of the controversies.

Orme-Johnson et al (1990) say that the threshold was presented as a sufficient


condition for measurable improvements, not as a necessary condition for any improvement. (p.758/60)

It is clear that the theory is neither (adequately) connected to natural science, nor to the social
sciences, by lack of definitions of its terms and vagueness of its contents. It is outside the current
paradigm of how we understand the world to work.
Many of the counterarguments used on the website by Orme-Johnson (year unknown) are probably
meant for a non-scientific audience; appealing to the authority of the physicist Hagelin, to the
authority of Maharishi himself, or to the very general conception of the Perennial Philosophy, is
rather ridiculous. In the draft of the definitive reply to Fales and Markovsky (1997), on which OrmeJohnson is still working, and which I received from him, he plans to first demonstrate that the
alternative explanations are not viable, and then address the points about the theory. Up till now he has
concentrated on the first task (Orme-Johnson, 2005).
I think Orme-Johnson (year unknown) is right that in science theories are created, tested and modified
to explain data, instead of the other way around. In this case, fluctuations of data that might quite well
be normal, are supposed to be explained by a new theory.
Design and execution of the experiment
There were three main points of critique on the design and execution of the experiment:

The participants were not randomly assigned to the task of meditating (Duval, 1988, p.815). The
treatment days were not random (Schrodt, 1990, p.752).

Orme-Johnson et al (1990) reply that in field studies such as this one it is not practical nor easy to
implement random assignment experiments. It is unconscionable to withhold people from meditating,
especially if they believe in the effects. The authors looked through 89 articles in 13 issues of the
Journal of Conflict Resolution and concluded that their article might be the first to actually evaluate
experimentally a means of directly reducing conflict on the international scale (p.765/6).

Of course a real random test could be realized, especially if the meditators were
convinced of the effects. Well informed, they would be glad to cooperate in an experiment that would
definitively prove the Maharishi effect to be real. Furthermore, I think it was not necessary to review
89 articles (how chosen?) to prove that their own experiment was extraordinary.

There was no connection between the model and the measurements: the time-series analysis
employed each days number of Sidhi meditators rather than its square, and the number of non-Sidhi
meditators was not included in the test. Also, periods with group size below the threshold were not
counted as zero; there was no information about the distribution and (absence of) activity of the
38,000 non-Sidhi meditators in Israel and the 2,000 in Lebanon. (August being the vacation month in
Israel). Without the non-Sidhi numbers mentioned, the experimental group was not powerful enough
(Fales and Markovsky, 1997, p. 519). If the material world is affected (through the unified field),
inanimate Maharishi-Effect-detectors should be placed at varying distances. The measurement unit of
the model is the number of people. This has a paradoxical complication: the radius of the effect
depends on the density of the population (ibid., p. 516).

Orme-Johnson (year unknown) replies that in fact the study in question did empirically evaluate
the square of the number of Sidhi-meditators.

Obviously, the critique has not (yet) been completely answered. Furthermore, the
graphics used in the original article do indeed suggest a direct correlation between the number of Sidhi
meditators and, for instance the Composite Index of Quality of Life (see fig. 1).

fig. 1. Graphic from Orme-Johnson et al (1988, p.795).


MTUF = Maharishi Technology of the Unified Field

The independent variable was measured incorrectly. If one uses distance instead of political
boundaries, the 1% -effect (of N2, see model equation) would never have been reached, because the
metropolitan areas of Amman and Damascus (population 1.2 million each), and other cities, all closer
than Beirut, are ignored. In other articles on the Maharishi Effect, authors use geographical distance
(Schrodt, 1990, p. 748).

Orme-Johnson et al (1990) reply that they have always used political units, because they reflect
greater homogeneity and stronger internal lines of influence. With greater group size, it was assumed
that among those nations contiguous with Israel, the Lebanese would be influenced first, because at
the time they were interacting far more intimately with the Israelites than were any other groups.
Furthermore, just as with radio waves, field effects are affected by local conditions and so an uneven
spread of the effects is not unusual. And finally, the 1% threshold was presented as a sufficient
condition for measurable improvements, not as a necessary condition for any improvement. The
important issue was whether there would be an effect at all (p.758/60)

Orme-Johnson et al (1988) do indeed speak of a collective consciousness


corresponding to each level of social organization: family, city, state, nation, and world. Also, the 1%
threshold was given as the threshold for a measurable effect (p.782).

8
Results and (statistical) interpretation
The struggle over statistical methodology and interpretation of the data was intense. It was about:

Complexity of the statistical method used: Duval (1988) says that, paradoxically, as our techniques
grow more sophisticated, we must place a greater faith in the competence of the researcher (p.816).
Schrodt (1990) calls Box-Jenkins methods suspect, unnecessarily complex, and so flexible, that the
likelihood of finding significant effects by chance is high. A simple additional test should have
checked this. Robustness checking should have been done by testing for the effects of pseudointerventions on random days when the treatment was not actually used. The robustness of the
technique on the dependent variable could have been tested using time-series which would not be
affected by the treatment, for instance war deaths in Angola (p.750, 752).

Orme-Johnson et al (1990) present additional reanalysis of the Lebanon war index which
combined the two war variables. They present tests showing that the relatively simple model used was
adequate. (p.762/3). As pseudo-interventions the authors present 2 tests, randomizing the sequence of
the independent and the dependent variable; this produced no significant results (p.765).

In the reanalysis Orme-Johnson et al (1990) use a combination of the two war


variables. This seems strange when we remember the dropping of one of the two in the original
analysis because they were highly correlated. The pseudointerventions in the form of just mixing up
the data do not strike me as meeting the demands of Schrodt (1990).

Lag effects : Schrodt (1990) remarks that the lag structure makes little theoretical sense without an
a priori theoretical justification. Lags effects at 0 and 10 days are significant, and the lags in between
are not (p.751). Also, inferring causal direction from correlational time-series has been proven
dangerous (p.751).

Orme-Johnson et al (1990) reply that tests confirmed that none of the dependent variables led the
independent variable (p.762). There were strong results for lags 0 and 1. The results found for lags 5
and 10 may have been due to chance, although those of lag 5 were quite robust (p < 0.05 for all
models tested) (p.764/5).

I agree with the critique, but of course, in a new experiment, it would be easy to
suggest lag effects beforehand.

Reversed causality: Schrodt (1990) calls the arguments against the possibility of
reversed causality insufficient. He thinks meditators are more likely to participate when the Maharishi
effect is operating; there may be autocorrelation. Israeli public is intensely well informed about the
war in Lebanon. Only a purely randomized design could have avoided this problem (p.749/50).

Orme-Johnson et al (1990) reply that Schrodt (1990) conjectures that at reports of violence the
meditators might hurry to join the group. But, in reality, the relation between group size and war
intensity was the other way around. Moreover, the number of drop insfrom Jerusalem and
surroundings was quite small; and during the planned 13 days period of the advanced TM course the
number of war deaths dropped very much, from 33.7 per day during the 13 days before and after the
course to 1.5 per day during the course (p.761).

I think the direction of causality was not proven, but certainly made plausible; Duval (1988,
p.815) also admits this.

Significance: Schrodt (1990) points at the fact that significance levels reported in the article are
the probabilities that the reported coefficients are not equal to zero; they are not the probabilities that
the model is correct (p.750).

I am not impressed with this critique. Is this not always the case with significance levels?

Alternative interpretations: Fales and Markovsky (1997) suggest a number of factors not
accounted for, that might have influenced the results: mainly religious or vacation holidays and a
number of political events. They state that they do not have to explain the data; plausible alternative
explanations are enough (p.519-21).

9
Orme-Johnson (year unknown) accuses Fales and Markovsky (1997) of not explaining the data,
and giving post hoc explanations which do not meet the minimum requirement of an explanation of
the strong relationships found.

I will get to the question who should explain the data and who should explain away the critique in
chapter 4.

General

Schrodt (1990) thinks that poor design and measurement errors ('sloppy science') have led to selfdeception; most (if not all) claims for scientific support of the paranormal phenomena disappear once
a reasonable set of controls is introduced into the procedure; the human brain is quite adept at seeing
patterns where they do not exist. Introduction of statistical techniques, he states, has magnified these
problems (p.746/7). He warns that confirmation of the theory would render suspect virtually the whole
of modern social scientific concepts of causality; therefore insistence on additional specifications,
controls and statistic tests would not have been unreasonable. The article would not have appeared in
World Politics or Foreign Affairs because those journals employ stricter criteria of scientific
scepticism (p.753).

Orme-Johnson et al (1990) reply that there is no question of seeing patterns where they do not
exist, because the major results are obvious from the raw data, without use of statistics. They raise
objections against the term paranormal, because of the negative connotations that paranormal
research typically invokes (p.757).

Fales and Markovsky (1997) say that the authors of the Israel study would not send them the raw
data set (p.524).

Orme-Johnson (year unknown) complaints that in fact the authors were sent all the data in graphic
form. Moreover, the authors were told that they would also be sent the data in spreadsheet form as
soon as they publicly retracted false statements that they had made about the research in television
interviews and in the popular press. He dismisses the article of Fales and Markovsky (1997) as a
thinly disguised attempt to censor research on the Maharishi effect.

It seems very reasonable to me that complete data sets of an experiment to be published


should be available for other scientists; in fact, today the redaction of the Journal of Conflict
Resolution demands as one of the submission requirements that authors of empirical articles will be
expected to make their data sets available on the internet to the scholarly community for replication
purposes by the month of publication of the article (www.yale.edu/unsy/jcrsubmit.htm).

In spite of all the legitimate objections against the theory of the Maharishi Effect and the
measurements and their interpretation, I am still impressed and puzzled by the strong correlations
found. If this is not fraud, what is it? The results have not been explained away by the suggestions of
the critics. We cannot confirm, though, that the major results are obvious from the raw data, because
nobody outside the research team has seen the raw data. Furthermore, all indices were based on scored
scale parameters, with possible subjective influences. In addition, the experimenters do not make
themselves more credible by refusing to hand over their complete data set, and by appealing to
authority instead of holding on to facts. Moreover, in none of the articles, they give clear indications of
including independent scientists in their team. It would add to their credibility, for instance, if the
counting of meditators would have been done by a really independent person.

10

4. How to address heterodox claims


It is not surprising that the publications concerning the Maharishi effect have promoted a more general
discussion about how the scientific community should address theories and connected data that are
outside the current scientific paradigm. The key questions are: when should a heterodox theory
deserve serious attention (in time, cooperation, and journal space); when can data presented as
evidence be considered to support the theory; and how can a heterodox claim gain credibility? In this
paragraph I give some considerations on these issues, and then reconsider the Maharishi effect study.
Theory
In principle, I agree with Fales and Markovsky (1997), that new scientific theories have to meet the
following criteria:

The theorys terms must be defined with enough precision and specificity, and related to the
scientific field in question;

The theory should be compelling in view of prior knowledge;

The theory should include a description of how it works (p. 512).


The problem is that, if theories meet these criteria, they are not heterodox. Therefore it is necessary, if
we want to look at heterodox theories at all, if we want to stay open-minded, to drop one or more of
these demands, and maybe replace them by others. This applies especially in cases where we have to
do with subjective emotional and spiritual elements, and their relation to the objective material world.
We know those relations are real, for instance, no one denies the existence of psychosomatical
diseases. But how exactly the relation between psyche and body works, is not clear. Should we
therefore dismiss as a study that presents a correlation between, for instance, stress factors and asthma,
as not scientific? The same goes for the Maharishi Effect theory. Can it be called science?
Popper (1998) states that a good scientific theory is falsifiable, that is: it should be testable, it should
forbid certain things to happen, and confirmations should only count if they are the result of risky
predictions (p.7).
The Maharishi effect is testable.
The theory forbids opposite effects, I think. Critics have looked up the crime rates in and around
Fairfield, Iowa, where the Maharishi University of Management is located. This is a city with
2,000 active meditators in a population of 10,000. For the period 1991-1998 there was an overall
increase in violent crime and property crime, both for Fairfield/Jefferson County and for Iowa
(Weldon, 2004).
The predictions from the theory were risky in principle, but not very precise. There would be a
'measurable effect' on a number of parameters. What raised my suspicion was the way these
parameters were selected. The description of this process (Orme-Johnson, 1988, p.787) left open
the possibility that, during the experiment, a number of variables that showed no effect, could have
been discarded, because the set of dependent variables was chosen from a bigger set, proposed
earlier.
Measured with the criteria of Popper (1998), the theory is shaky at least, in terms of science.
I would suggest that a heterodox theory is worth attention of the scientific community if it connects a
number of objectively measurable and well defined variables, which makes it testable. The proponents
have the additional task to provide confirmation of the claimed correlation by tests, carried out in a
team comprising sceptic experts in the field under consideration.
It is clear that The Maharishi Effect studies do not meet the standards put forward by the critics and
theorists cited nor the alternative criteria I have suggested.

11
Confirmation
The editor of the Journal of Conflict Resolution Russett (1988) cites one of the reviewers who would
only consider the article seriously for publication if it were conducted by an independent scientific
body such as the National Academy of Sciences. The editor replies that for a study as heterodox as this
one, it would be almost impossible to imagine the NAS being willing to support it without some prior
evidence as produced by the normal scientific review procedure. This would mean the virtual
impossibility for evidence ever to appear in print. In replications of this experiment, he stresses,
safeguards against personal bias are necessary, for instance in the team composition (p.775).
It is my opinion that if a heterodox theory is to be given attention, it should have been tested under
strict extra conditions, especially if the working mechanism is unknown. A good example of measures
to prevent 'leakage' of information is given by Radin (1997, p. 73-89) in describing the extremely
strict protocols accompanying Ganzfield-experiments (telepathy).
Should the burden of refuting alternative explanations indeed lie with the proponents of the theory, as
Fales and Markovsky (1997, p.514) say? Lakatos (1998) states that there is no refutation without a
better theory (p.25).
In my opinion, it depends. If a theory is orthodox, there will probably not be many critics; if there are
orthodox critics, they will want to compete and get the possibility to do it. In that case, Lakatos is right
(heterodox critics will not be listened to). If a theory is heterodox, it has to provide in advance,
confirmations as suggested above; if it does not, it is not worth attention; if it does, and there is no
alternative theory, the critics have to come up with a better model, and not just with assumptions or
possibilities. In that case, Lakatos is right again. Another consideration, related to the Maharishi
articles, is the following: if a heterodox theory tries to explain observations that could very well be
quite normal fluctuations of events, it would seem strange that an alternative explanation would be
necessary to refute it. But if the fluctuations are predicted and confirmed, and correspond to the
theory, we have to reflect on the question if we have a normal explanation for those normal
fluctuations.
The Maharishi Effect studies do not meet the criteria I have proposed above. Especially the important
demand of taking up critics in the team, and the extra strict safety measures to be taken in the case of
examining paranormal phenomena, are missing.
Credibility
Even after a theory is published and confirmations added, it can be qualified as pseudoscience
afterwards. In the opinion of Thagard (1998) a theory is pseudoscientific if and only if
1.
it has been less progressive than alternative theories over a long period of time, and faces
many unsolved problems; but
2.
the community of practitioners makes little attempt to develop the theory towards
solutions of the problems, shows no concern for attempts to evaluate the theory in relation to
others, and is selective in considering confirmations and disconfirmations;
progressiveness being a matter of the success of the theory in adding to its set of facts explained and
problems solved. (p. 32).
It must be said that Orme-Johnson tries hard to refute the objections raised against the Maharishi
studies. But up till now, he mainly makes additional statistical tests of the original dataset. There is no
real alternative theory, only suggestions that can rather easily be refuted with the dataset used. This
places more importance on the dataset itself, and how it was realized. And then the issue of
trustworthiness comes up.

12
In my opinion, proponents of a heterodox theory can gain or lose credibility by their behaviour. They
lose it if they do not give access to the raw data set for scrutinizing by others, if they use unscientific
counterarguments against their critics, and if related claims are never proven. Even if this behaviour is
displayed apart from the scientific discussion, it makes them less reliable as research partners. These
are not scientific, but sociological facts about how the world works.

13

5. Conclusions
The Maharishi Effect study (Orme-Johnson et al, 1988) presents an extravagant and compelling claim.
The theory behind the effect is vague and the relations with natural sciences are spurious.
I have indicated that it is in the nature of heterodox theories that they do not conform to all the criteria
put forward for theories to be accepted by mainstream science. To be worth hearing by the scientific
community, I suggested some alternative criteria:
A heterodox theory is worth attention of the scientific community if it connects a number of
objectively measurable and well-defined variables, which makes it falsifiable. The proponents
have the additional task to provide confirmation of the claimed correlation by tests, carried out in a
team comprising sceptic experts in the field under consideration, and taking strict precautions
against bias and leakage of information. If these criteria are met, and in absence of a competing
theory (not being a mere suggestion), the heterodox theory should be heard.
The presented results of the experiment carried out in Jerusalem in 1983 show impressive correlations
between meditator group size and a number of variables related to the quality of life in Israel and the
Lebanon. Nevertheless, these results are suspect because of a number of factors: the research team was
not independent, extra strict precautions to be taken in the case of examining paranormal phenomena
are missing, and the dataset was not published. In addition, the proponents have lost credibility by
their subsequent attitude.
Recently, a new article was published (Davies and Alexander, 2005), re-examining the data of the
original experiment and a number of replications in the following 2.5 years. Of course the design of
those experiments could not be changed with respect to the critiques given later on. Instead the authors
concentrate on construction of new parameters for violence, cooperation and casualties, using nine
international and regional news sources and an 'experienced Lebanese coder blind to the hypotheses
and techniques employed'. It is hardly surprising that the correlations found are as excellent as before.
What is surprising, is that the TM-community does not carry out a brandnew experiment, committing
themselves to all the restrictions and precautions their critics have advised them.
The Maharishi Effect studies do not meet the standards put forward by the critics cited, nor do they
meet the criteria I have proposed above. As long as this is not the case, the Maharishi effect is not an
invincible, but an invisible defence technology.

14

References

Chryssides, G.D. 2001. Exploring New Religions. London, New York, Continuum.
Davies, J.L., and C.N. Alexander. 2005. Alleviating Political Violence Through Reducing Collective
Tension: Impact Assessment Analysis of the Lebanon War. Journal of Social Behavior and Personality,
2005, 285338.
Duval, R. 1988. TM or not TM? A comment on 'International Peace Project in the Middle East'. Journal of
Conflict Resolution 32 (4): 812-817.
Fales, E., and B. Markovsky. 1997. Evaluating Heterodox Theories. Social Forces 76 (2): 511-525.
Gelderloos, P., K.L. Cavanaugh and J.L. Davies. 1990. The Dynamics of U.S.-Soviet relations, 1979-1986:
A Simultaneous Transfer Function Analysis of U.S.-Soviet Relations: A Test of the Maharishi Effect.
Proceedings of the American Statistical Association, Social Statistics Section, p. 297-302.
Hagelin, J.S., Orme-Johnson, D.W., Rainforth, M. et al. 1999. Results of the National Demonstration Project
to Reduce Violent Crime and Improve Governmental Effectiveness in Washington, D.C. Social Indicators
Research 47: 153-201.
Lakatos, I. 1998. Science and Pseudoscience. In: Curd, M., and J.A. Cover (eds). 1998. Philosophy of
Science; The Central Issues. New York, London, W.W. Norton & Company. p. 20-26.
Orme-Johnson, D. 2005. Re-analysis of International Peace Project in the Middle East (unpublished draft)
Orme-Johnson, D. (year unknown) Critique of 'Evaluating Heterodox Theories'.
www. invinciblemilitary.org/articles/heterodox_theory_rebuttal.html (visited at 21-12-2005)
Orme-Johnson, D., C.N. Alexander and J. Davies. The Effects of the Maharishi Technology of the Unified
Field. Reply to a Methodological Critique. Journal of Conflict Resolution 34 (4): 756-768.
Orme-Johnson, D.W., C.N. Alexander, J.L. Davies et al. 1988. International Peace Project in the Middle
East; the Effects of the Maharishi Technology of the Unified Field. Journal of Conflict Resolution 32 (4):
776-812.
Orme-Johnson, D. W., M. C. Dillbeck, C. N. Alexander, H. M. Chandler, and R. W. Cranson. Effects of
Large Assemblies of Participants in the Transcendental Meditation and TM-Sidhi Program on Reducing
International Conflict and Terrorism. Journal of Offender Rehabilitation 2002/2003 (1/2/3/4).
Popper, K. 1998. Science: Conjectures and Refutations. In: Curd, M., and J.A. Cover (eds). 1998.
Philosophy of Science; The Central Issues. New York, London, W.W. Norton & Company. p.3-10.
Radin, D. 1997. The Conscious Universe; the Scientific Truth of Psychic Phenomena. New York, Harper
Collins.
Russett, B. 1988. Editor's Comment. Journal of Conflict Resolution 32 (4): 773-775.
Schrodt, P.A. A Methodological Critique of a Test of the Effects of the Maharishi Technology of the Unified
Field. Journal of Conflict Resolution 34 (4): 745-755.
Shear, J. 2001. Experimental Studies of Meditation and Consciousness. In: Jonte-Pace, D., and W.B.
Parsons. 2001. Religion and Psychology: Mapping the Terrain. London, New York, Routledge. p.280-294.
Thagard, P.R. 1998. Why Astrology is a Pseudoscience. In: Curd, M., and J.A. Cover (eds). 1998.
Philosophy of Science; The Central Issues. New York, London, W.W. Norton & Company. p. 27-37.
Weldon, J. 2004. Transcendental Meditation in the New Millennium. Part Two: Does TM Really Work?
Christian Research Journal 27: 6.

Websites:

www.invinciblemilitary.org/articles/heterodox_theory_rebuttal.html (21-12-2005)
www.yale.edu/unsy/jcrsubmit.htm (10-1-2006)
http://www.skepsis.nl/hoppen.html (11-1-2006)