ANALYZING THE GELLER EFFECT
by John G. Taylor, Ph.D., Department of Mathematics,
King's College, University of London.
THE GELLER EFFECT involves the bending or breaking of objects,
usually metal, by certain people in situations in which such results
are scientifically impossible. The overt methods used are either
a gentle stroking of the object or mental concentration on it
without any direct contact with the specimen being distorted.
At least fifty cases have been reported in this country, and
similar numbers have been noted in other countries after visits
by Uri Geller, the majority of the subjects being children. The
effect itself is clearly a challenge to science, and requires
a thorough investigation. This is especially attractive since
the materials being distorted have structures that are reasonably
well understood. What is more, the amounts of energy involved
must be considerable because they cause metal objects to bend
or break, so whatever mechanism is involved cannot be too elusive
to be observed by reasonably sensitive apparatus. The prognosis
for a complete analysis of the effect is therefore good.
The first question concerns the validity of the effect. Many
instances of it have been reported under highly variable conditions.(1)
The emergence of subjects with metal-bending powers has allowed
more careful analysis with controlled surroundings. The repeatability
of the effect due to the persisting powers of the subjects has
permitted simple tests to be designed; they quite clearly validate
the Effect.(1,2) That does not mean that all cases of metal bending
have to be accepted as belonging to the Geller Effect, but sufficient
precaution can be taken to prevent fraud. The pressure being
applied during stroking of a specimen can always be measured by
the experimenter either holding one end of the object or attaching
it to a suitable spring balance. Less certain, but still reasonable,
is the very careful direct observation during the bending process
itself. The majority of the results to be reported here involved
the latter process, but only with subjects whose ability to bend
material was already validated.
The next step in the scientific analysis of the Geller Effect
is to determine the range of the phenomenon itself. This involves
such things as the range of materials that can be affected, the
variation of ease of bending according to the shape of the material,
and the comparison of the powers possessed by different people.
More technical questions, such as what associated phenomena -
temperature change, current flow, etc. - occur in the metal or
other material during distortion, also need to be answered, as
do the possible forms of radiation that could be used to transmit
energy to the specimen. There are also related disturbances that
may be helpful to consider, and that are brought about specifically
by the metal-bending subjects. In this paper we attempt to present
preliminary data on these matters obtained from work with a number
of young subjects (all under the age of sixteen) as well as with
Uri Geller. (A microanalysis of the objects is underway.)
The final step is to determine the mechanisms involved in the
Geller Effect. That is inextricably tied to the results of the
experiments mentioned above, but it was felt appropriate to attempt
as descriptive a discussion as possible before soaring to the
rarefied atmosphere of theoretical analysis. Clearly the choice
of experiments is partially guided by the mechanisms available.
This, must not prevent all possible mechanisms from being investigated
without prejudice. The existence of the Geller Effect alone shows
that the scientifically impossible can sometimes occur.
The range of materials that can be bent or broken is broad. The
metals include copper, aluminum, brass, several forms of steel,
tin, lead, zinc, and silver. The first four of these have been
affected by subjects under conditions in which the pressure applied
to metals was measured or when there was no direct contact. The
remainder have been distorted under less strict conditions but
still when only gentle pressure appeared to be applied, and by
children whose powers on the first four metals had been validated
under strict conditions. Plastics have also been distorted by
children. The plastics were tensile test specimens of polystyrene
and polycarbonate. Ionic crystals have also been affected without
direct contact.(2) Single crystals of both LiF and NaCl were fractured,
the former with marked disintegration, and both fractures were
made by Uri Geller. Silicon has been affected, as reported earlier.(2)
Other materials have also been reported as being distorted; one
was wood, though the experimenter has not observed this directly.
It is clearly important to know the exact range of materials
that are amenable, since important clues to the effect could result
from such knowledge. In particular it is of value to learn if
it is only materials with a certain degree of regularity in their
structure that are at risk.
The next question is that of the effect of the shape of the specimen
on ease of distortion. The results would indicate whether the
Geller Effect involves the whole specimen being distorted or only
a portion of it. To test this feature, standard specimens of
different sizes were used with two subjects (A and B) whose bending
powers had been validated on other specimens.(2) The first test
involved five specimens of aluminum, each with a cross section
of 0.6 x 0.2 cm. Every specimen was stroked gently by subject
A for a period of seven minutes. The specimens were tested successively
with a rest of two minutes in between each. The resulting degree
of bending is given in Table 1. There is a clear relation of length
of specimen to ease of bending, the best case being for that with
a length of 19.4 cm. This result was supported by a similar test
performed at an earlier time by the mother of subject A; the results
are also given in Table 1. The times are not exactly comparable
with those of the later test, but it was found that a strip about
20 cm long was bent most easily. Tests were also done under parental
supervision with subject B. The results are shown in the table;
the maximum length for the copper specimens used was 13.5 cm.
Tests were also set up, again under parental supervision, with
subject A using copper strips 2.6 cm wide. Here again the longest
strip bent most rapidly. It is clear that the length of the specimen
is of considerable importance in the bending process; the influence
of cross section is not so apparent, and obviously more work needs
to be done on this.
Table 1. Variations in the length of metal specimens
and the ease of bending them
Material Details
Sub- Super- and cross of
ject visor section bending
A Experi- Aluminum Length (cm) 5.0 12.4 19.4 50.0
100.0
menter' 0.6*0.20 Degree of 0 0 32 14 6
cm bending
Time (min) 7 7 7 7 7
A Parent ditto Length (cm) 5.1 9.4 18.6
Degree of 0 125 178
bending
Time (min) 8 4.75 2.75
B ditto Copper Length (cm) 5 10 13.5 15.5
0.6 x 0.2 Degree of 0 4 1.1 1.0
CM. bending
Time (min) 25 15 10 1.0
A ditto Copper Length (cm) 5.4 10.1 17.3 19.9
2.6 x 0.15 Degree of 0 0 4 15
cm bending
Time (min)
There are also data about the subjects with metal-bending powers.
At least fifty such cases have been reported to the experimenter
by reliable witnesses, though he has personally investigated only
twenty-five of these cases, and of these only six were observed
in action, making it possible to say they were genuine. There
are only four adults among the fifty persons reported, all the
others being under the age of sixteen; no adults are present at
all among the six observed most closely by the experimenter.
The youngest child observed is seven and a half (a four-and-a-half-year-old
has been reported as having metal-bending ability, but the child's
alleged talent was not directly studied by this experimenter).
The average age of the children is about twelve, and they are
roughly divided equally between boys and girls. The outstanding
feature of these statistics is the complete exclusion of adults
from the select group; its explanation might be found in physiological
changes normally occurring at puberty.
It is also of value to investigate what is taking place in the
specimen itself during bending. At the macroscopic level, measurements
have been made of temperature changes by taping a 0.3-cm-long
thermocouple to each specimen. No temperature changes of more
than 2 degrees C have been recorded during the bending process.
This change is on the order of that expected when a specimen
is touched by the hands. Microheating in the interior of a specimen
could have escaped notice by such a measurement, but microanalysis
to date indicates the absence of any such effects.
In a similar fashion any possible current flow occurring during
bending was measured by means of wires soldered to metal specimens.
Specimens of various shapes were used, the main ones being copper
strips, each with a cross section of 0.6 x 0.2 cm and a length
of 10 to 15 cm. The wires were attached at the mid point
and near one end of each specimen. Changes of potential were
measured on either D.C. or A.C. millivoltmeters. The sensitivity
of the measurements was increased by a factor of 1000 by a suitable
voltage amplifier. In three cases of bending of the Cu strips,
one case involving a bend of over one hundred and eighty degrees,
no deflections of the apparatus were recorded on the most sensitive
scales. The corresponding potential differences were below 25
micro Volts at D.C. and 0.1 micro Volts over the frequency range
of 1 Hz to 3 MHz.
Any radiant energy transferred to the specimen during bending
was also accessible to measurement. The first type of radiation
investigated was of the ionizing variety. A portable radiation
monitor Type 5-40 with an x-ray scintillation probe Type 5-42,
made by Mini Instruments, sensitive to 3-100 keV photons, or a
larger Airmac radiation monitor 1021B was used in various tests
with known subjects, the radiation probes being placed as close
as possible to the specimen being affected. No readings above
background radiation were observed, to within the accuracy of
the equipment, in three separate sessions when bending occurred
with five subjects. Nor were there any effects, during these
tests, on the discharge rate of a gold-leaf electroscope. For
example, in one test the electroscope deflection decreased twenty
degrees in twenty minutes while bending occurred; a similar reduction,
to within 5%, occurred due to normal charge leakage. The only
way in which ionizing radiation could have been involved was if
it were so highly focused as to have evaded the radiation probes.
Ultraviolet radiation was tested for by the use of metal strips
coated with sodium salicylate and enclosed in quartz tubes. The
strips were treated as specimens to be distorted, the attempts
being made by subjects in reduced illumination so as to observe
any fluorescence, appearing as a violet glow, produced by ultraviolet
radiation. The strips were also placed nearby when standard specimens
were bent. There were some violet flashes seen while the tubes
were rubbed (though no bending of the treated strips occurred).
but roughly the same number of flashes, and with the same level
of illumination, were obtained by purely frictional effects at
a different time. Otherwise no fluorescence occurred. The total
fluorescence in all cases was less than that produced by illuminating
the tube with a helium vapor lamp (Hg vapor calibration lamp A26-4812
at 5 W, UV Products, U.S.A.) at a distance of 20 cm; the amount
of ultraviolet radiation that may have been involved during the
bending sessions was clearly far too low to be significant.
A similar complete lack of success was encountered when we tried
to detect infrared radiation from 2 micro to 1 mm wave length
by means of a germanium crystal detector connected to an oscilloscope.
In two separate tests the detector was positioned next to the
specimens being bent. In all cases there was at most a half-division
change of the oscilloscope beam during bending; this is in contrast
to the two-division deflection caused when hands were passed in
front of the detector at about the same distance, 10 cm, as the
specimens that were being bent. The possibility of static magnetic
field effects during bending was also ruled out for numbers of
subjects by using a milligaussmeter (Type M Magnetometer, Newport
Instruments) in two sessions; no deflections were observed, while
bending occurred, to within the noise level of the instruments
on the most sensitive scale.
It is possible that one or another of the above forms of radiation
could still be involved with the Geller Effect, but in too narrow
a beam. To test this hypothesis, direct attempts were made by
subjects to affect the various measuring instruments. No modifications
above those ascribable to experimental error were obtained with
the infrared or ultraviolet detectors (except for breakage of
an axle on the chopper blade in front of the germanium crystal)
nor with the milligaussmeter. This was not the case with the
Geiger counter.
The first clear evidence that a Geiger counter could be affected
occurred when the portable radiation monitor was tested with Uri
Geller. In the presence of two independent observers, Geller
caused the counting rate to be increased by a factor of twenty-five
above the background radiation on three separate occasions, and
by a factor of 50, 100, and 500 times above it on one time for
each. The duration of each of these high count rates was about
three seconds, except for the highest rate, which lasted five
seconds. In each case Geller held the probe in his hands and
appeared to exert some physical force, though a test afterward
showed that distortion of the probe was very difficult to achieve.
Since this result was so unexpected, the effect on the Geiger
counter was tested by the more satisfactory method in which the
subject was not allowed to touch the radiation monitor at all;
the output was also recorded automatically by allowing each output
pulse to charge up condenser whose slower discharge across a resistance
was recorded on a Servoscribe millivoltmeter. The results obtained
supports those found with Geller. The case of a sixteen-year-old
girl, whose metal-bending powers were suspected but had not been
validated was outstanding. She obtained deflections up to twenty
times greater than the background radiation. A group of seven
subjects who were not metal benders produced deflections about
forty times greater than background radiation. The smoothness
of the back ground readings over three hours of monitoring was
an indication of the satisfactory functioning of the instrument,
which had thoroughly overhauled the preceding week.
None of the subjects was able to influence significantly the
rate of discharge of an electroscope. There is clearly a paradox
here, since the latter result indicates that it is most unlikely
that ionizing radiation which is the cause of the modification
of the Geiger counters. The problem, then, is to discover what
the cause actually is, for it may well be closely linked to the
mechanism producing the Geller Effect itself.
A final phenomenon, which is clearly relevant, is the subject's
power to move objects. This was observed with Geller, who caused,
a compass needle to rotate smoothly from its equilibrium position
through forty degrees and remain deflected for a period of four
seconds. This occurred three times, the third time in the presence
of three other witnesses. It was not possible to duplicate this
phenomenon by moving the body in such a way as to cause the compass
needle suspension to vibrate, since any such motion added a vertical
oscillatory component to the rotation of the needle; none was
observed with Geller. Again, the cause of this may be related
to the Geller Effect.
In conclusion, we can say that the Geller Effect depends on the
gross shape of specimens, especially their length; it involves
very little change of temperature or flow of current in the specimen,
and is very unlikely to be caused by ionizing, ultraviolet or
infrared radiation. Nor are static magnetic fields involved.
Further investigations of these questions are now underway.
Other phenomena affecting Geiger counters and rotating magnets
also need further investigation to determine their mechanisms
and the light they may throw on the Geller Effect and on the extrasensory
powers of human beings in general.
REFERENCES
1. J. G. Taylor, Superminds. A Scientist Looks at the Paranormal
(New
York: Viking Press, 1975).
2. J. G. Taylor, "On the Geller Effect," Psychoenergetic
Systems No. 5 (yet to be published).
Back to books list.
Back to main index.
Back to Uri Geller's home page.