Actually, the famous Gauquelin finding had nothing to do with Sun signs. If it had, it wou
Actually, the famous Gauquelin finding had nothing to do with Sun
signs. If it had, it would have been easily dismissed by skeptics
and debunkers because they could attribute the effect to some
unknown seasonal variation. (One poster suggested temperature
changes affecting fetal temperature as a possible "cause.")
The so-called "Mars effect" is that *champion* athletes (not
athletes in general) have Mars above the ascendant (i.e., just
risen) or just past culmination (i.e., just past the planet's
highest point in the sky) with a frequency greater than expected
by chance. The same is true of military leaders. By tradition,
Mars is associated with both athletics and the military. Also by
tradition, planets on an angle (rising, culminating, descending,
or anti-culminating) are considered to be more powerful, to have
a greater effect on personality, ability, and life circumstances.
This Mars effect has been replicated now by two skeptics'
organisations, one in Belgium (Belgian Committee for the
Scientific Study of Paranormal Phenomena) and one in America. In
the latter case, the group providing the replication was CSICOP
(Committee for the Scientific Investigation of Claims of the
Paranormal). They originally published tainted results which
seemed to represent a failure to replicate, but after years of
controversy they admitted to adding into the sample non-champion
athletes to dilute the originally positive result that they had
obtained. In a recent issue of the Skeptical Inquirer (their
mouthpiece), they published an article by Suitbert Ertel which
showed that the American sample used in the CSICOP study does
indeed show the Mars effect. As a result of the controversy
resulting from their cover-up of the originally positive finding
in the Mars/athlete study, CSICOP has ceased conducting
scientific investigations (and so their name is no longer
The Mars effect was only one of many obtained by Michel and
Francoise Gauquelin. They found similar effects for different
professions, with a planet traditionally associated with skills
useful in those professions being angular (rising or culminating)
with significantly greater frequency in the charts of the most
successful in those fields. For example, Jupiter was angular with
a probability greater than chance for actors and writers, Saturn
for scientists and physicians, the Moon for writers and
politicians, and so on.
In later work, they found that the link between character/ability
and planetary angularity was even more direct than the link with
careers. In other words, success in a particular career was
associated with the angularity of a certain planet because the
planet lent certain traits to the person that tended to
facilitate success in that career.
The Gauquelins also found some interesting results regarding
heredity. They found that a child with a certain planet angular
was more likely to have a parent with the same planet angular
than chance would predict. This effect has been replicated, and
the researchers have also shown that the effect is only present
when the birth is natural; the effect disappears if the birth is
induced (by drugs) or if a C-section is performed. (One so-called
failure to replicate has been shown to be due to the failure to
take account of the nature of the birth; the sample contained
many unnatural births, as any modern western sample would.
Removal of the induced births and C-sections yields the heredity
effect in the remaining sample.)
Many other researchers have found positive results in
astrological studies, and some of the results have been
replicated on independent samples. One example is the work of
Furze-Morrish on compatibility. He showed that married couples
who, according to their own judgement, were compatible had more
positive (benefic) aspects (trines and sextiles) between planets
in their two charts (i.e., in the "synastry" comparison -- a
comparison of the positions of planets in two charts) than
negative (malefic) aspects (squares and oppositions) as compared
to couples who described their marriages as full of conflict. He
later replicated this effect and extended it to compatibility
among school boys. (An "aspect" is an angular separation between
two planets that is an integer division of the circle; a trine is
one-third of the circle, or 120 degrees; a sextile is one-sixth;
a square is one-fourth; an opposition is one-half.)
Some recent work by Nick Kollerstrom and Mike O'Neill looks very
promising. They studied "Eureka moments," moments in which
scientists made significant discoveries in a flash of
inspiration. Based on John Addey's interpretation of aspects
involving division of the circle by 5 and 7, they predicted that
such aspects would be predominant among transiting planets (i.e.,
the planets at their current positions at the time in question)
when Eureka moments occurred. This prediction was strongly
confirmed. They recently (in unpublished work) showed that the
effect is even stronger if tighter orbs are used, that is, if
only aspects very close to exact are taken into account. By
tradition, closer aspects are more powerful. The researchers also
found that the scientists who made discoveries in flashes of
inspiration were more likely to have aspects in the 5th and 7th
harmonics (i.e., ones that divide the circle by 5 or 7) in their
birth charts than scientists who did not make discoveries in this
way. They have recently extended their findings to moments in
which inventions first worked, and they have shown that trines
were also more common at such moments. This work should be
published within a year or two. The original Eureka study is
published in the booklet "The Eureka Effect" (Nick Kollerstrom
and Michael O'Neill, London: Urania Trust, 1989).
One effect that has been replicated numerous times involves the
traditional correspondence between certain planets and certain
metals (e.g., the Sun and gold, the Moon and silver, Mars and
iron, Saturn and lead). When an aspect occurs between two
planets, metal ions associated with those planets behave
differently than they do when no aspect is present. For details,
see the following references, or read the recent postings on
alt.astrology with subject lines beginning, "Planets and
Faussurier, A. Conscience Ecologique et Cre'ativite' Humaine,
Fyfe, A. Uber die Variabilitat von Silber-Eisen-Steigbildern,
Elemente der Naturwissenschaft, Vol. 6, pp. 35-43 (Easter
Fyfe, A. Moon and Plant, Society for Cancer Research, Arlesheim
Switzerland 1967, pp. a7 b37.
Hammerton, C. Repetition of Experiment made by L. Kolisko in
relation to Observable Effects in Salts of Metals
Corresponding to the Planets, Astrology (UK), Vol. 28, No.
2, pp. 46-48 (1954).
Kolisko, L. Workings of the Stars on Earthly Substance, Parts 1 &
2, Stuttgart 1928.
Kolisko, L. Das Silber und der Mond, Orient-Occident Verlang,
Kolisko, L. Der Jupiter und das Zinn, Mathematisch-Astronomische
Sektion am Goetheanum (Doirnach), Stuttgard 1932 (available
in English as Workings of the Stars on Earthly Substances,
Part 4, Jupiter and Tin).
Kolisko, L. Gold and the Sun, Kolisko archive (published
privately), Stroud UK 1947 (a study of the total solar
eclipse of 20 May 1947; a study of the total solar eclipse
of 29 June 1927 is given in Workings of the Stars on Earthly
Substance, part 2; of 19 June 1936 in Gold and the Sun,
London 1937; and of 15 February 1961 in Die Sonnenfinsternis
vom 15 Februar 1961, Stuttgart 1961).
Kolisko, L. Spirit in Matter, Kolisko archive, Stroud UK 1947.
Kolisko, L. Saturn und Blei, Kolisko archive, Stroud UK 1952.
Kollerstrom, N. Astrochemistry: A Study of Metal-Planet
Affinities, London: Emergence Press, 1984.
Kollerstrom, N. The Correspondence of Metals and Planets --
Experimental Studies, The Astrological Journal, Vol. 18, No.
3, 1976, pp. 65-72.
Kollerstrom, N. Chemical Effects of a Mars-Saturn Conjunction,
The Astrological Journal, Vol. 19, No. 3, 1977, pp. 100-105.
Schwenk, T. 1949, quoted in W. Pelikan, The Secrets of Metals,
Anthroposophic Press, Spring Valley, NY, 1973, pp. 23-25.
Voss, K. Neue Aspekte, No. 5 (1965); summarised by R.C.
Firebrace, Confirmation of the Kolisko Experiments, Spica,
Vol. 4, No. 4, pp. 4-8 (1965).
To find additional research on astrology, look to the
alt.astrology FAQ file. Here is an excerpt from it, describing
some of the sources of relevant research studies:
9) Where can I find scientific research on astrology?
Answer: Brief summaries of a few scientific studies (written by
Thomas David Kehoe) are available at the ftp site
[hilbert.maths.utas.edu.au] in the files "gauquelin" and
"jung.synastry," which can be found in the directory
The most famous research is that of Michel and Francoise
Gauquelin. Some of their findings have been the focus of decades
of scrutiny by skeptics, and their results have held up under
this scrutiny. Some of their studies have been successfully
replicated with different samples and by independent researchers.
The highly publicised CSICOP "failure to replicate" on an
American sample for the "Mars effect" (the appearance of Mars in
certain sectors with greater-than-expected frequency for eminent
athletes) has been shown to replicate the effect when the
athletes are ordered by eminence (see the Winter, 1992 issue of
the Skeptical Inquirer). (The CSICOP researchers included far
fewer eminent athletes in their sample than did the Gauquelins,
and this washed out the Mars effect when the sample as a whole
was considered. When the athletes are divided into groups
according to an objective criterion of "eminence," the Mars
effect emerges among the most eminent.)
Some of the Gauquelins' research is summarised in the following
books: Michel Gauquelin, "Cosmic Influences on Human Behavior"
(3rd edition, published in 1985 by Aurora Press, P.O. Box 573,
Santa Fe, NM 87504); Michel Gauquelin, "Planetary Heredity"
(published in 1988 by ACS Publications, Inc., P.O. Box 16430, San
Diego, CA 92116-0430); Francoise Gauquelin, "Psychology of the
Planets" (published in 1982 by ACS Publications, Inc.).
A preliminary report of a study showing the relationship between
inspiration in scientific discovery and certain angular
separations of planets appears in a booklet entitled "The Eureka
Effect," by Nicholas Kollerstrom and Michael O'Neill. It was
published in 1989 by Urania Trust, 396 Caledonian Road, London N1
1DN. A complete report on this study and some additional data on
inventions will appear sometime in the next few years.
The Astrological Association of London publishes a scholarly
journal devoted entirely to astrological research. It is called
Correlation. (See the resource list for the address and phone
number of the Astrological Association; see # 18 for information
about the resource list.) Prior to its first publication in 1981,
research articles appeared in The Astrological Journal, also
published by the Astrological Association. If you are in Britain,
all issues of this journal are available at The Astrology Study
Centre (396 Caledonian Road, London N1 1DN), the Oxford and
Cambridge University libraries, the Scottish National Library in
Edinburgh, the National Library of Wales in Aberystwyth, Trinity
College in Dublin, the Warburg Institute, London University, the
British Library in London, and the York University library. In
the USA, these journals are available at the Heart Center
library, 315 Marion Avenue, Big Rapids, MI 49307. Astrologers in
your local area may have copies of these journals as well.
Astrological research appears occasionally in academic journals
of psychology. A literature search (e.g., of the database
"Psychological Abstracts") for articles containing the keyword
"astrology" or "astrological" (or "astrolog?" where "?" is a wild
card) would turn these up.
Because of the difficulty in publishing astrological research (or
any unorthodox research), much remains unpublished. Among such
studies are those described in postgraduate dissertations on
astrology. A list of these (up to 1981) appears in the December,
1982 issue of Correlation. For more recent dissertations, check
Dissertations Abstracts at a university library. (Our very own
Mark Urban-Lurain did a multivariate analysis of the birth data
of members of Alcoholics Anonymous for his Master's thesis at
Michigan State University.)
From: Jan Willem Nienhuys
>Anyone want to comment?
Yes but only to parts. The whole piece is too long.
>The so-called "Mars effect" is that *champion* athletes (not
>athletes in general) have Mars above the ascendant (i.e., just
>risen) or just past culmination (i.e., just past the planet's
>highest point in the sky) with a frequency greater than expected
>This Mars effect has been replicated now by two skeptics'
>organisations, one in Belgium (Belgian Committee for the
>Scientific Study of Paranormal Phenomena) and one in America. In
>the latter case, the group providing the replication was CSICOP
>(Committee for the Scientific Investigation of Claims of the
>Paranormal). They originally published tainted results which
>seemed to represent a failure to replicate, but after years of
>controversy they admitted to adding into the sample non-champion
>athletes to dilute the originally positive result that they had
>obtained. In a recent issue of the Skeptical Inquirer (their
>mouthpiece), they published an article by Suitbert Ertel which
>showed that the American sample used in the CSICOP study does
>indeed show the Mars effect. As a result of the controversy
>resulting from their cover-up of the originally positive finding
>in the Mars/athlete study, CSICOP has ceased conducting
>scientific investigations (and so their name is no longer
Not a lie, at least when you equate "scientific" with "astrological".
The story is as follows. Originally 128 champs showed a nonsignificant
Mars excess. CSICOP and Gauquelin agreed that this was not enough.
The CSICOP has been (to my opinion) imprudent by adding more athletes
without consulting Gauquelin. As they had started to select the very top
from the available data (but were unable to obtain all birth times
because of privacy regulations) they were caught in a double bind.
If the results had turned out favorable the Gauquelins would have
applauded their results. In the other case they were prepared to
cry "dilution!". Actually the second case happened. The basic reason
was that the Gauquelins never bothered to explain how good a champion
must be to be considered really good. Their original sample included
268 Italian aviators. Not quite a strenuous sport for the millions,
more something for viscounts and dukes that can afford private airplanes.
Later Gauquelin collected data about 600 Italian first division soccer
players. No result. He then raised the norms: the 98 ones that had played
at least once in an international game "showed the Mars effect".
In the Belgian test the norm was raised to 20 international games. What
not many people know (but Ertel, who has found this, does) is that *prior*
to the Belgian test the Gauquelins already had collected data about
Belgian soccer players; their files contained 76 ones that fell just
below the "20 internationals norm", and that showed a Mars effect of
only 10 percent (17.2 expected). I don't know how the Belgians got the
idea that "20 internationals" had to be the limit of excellency.
From the enormous number (198) of Belgian cyclists in the Belgian test
it might be concluded that even during the Belgian test the Gauquelins
were not too strict about how excellent a champion must be. But *after*
the American test they raised the standards. They complained that
the data base for basketball players was too large (1000 U.S. champions).
The reason? In their own data they had 33 basketball players of whom
only 3 were born in the appropriate Mars sector. So they thought they
had reason to distrust basketball players. After the test they suggested
that only "Olympic gold medal winners" were champion level.
The results of Ertel clearly show that the Gauquelins biased their data.
(I don't think the bias was deliberate, because probably both of the
Gauquelins did not understand the difference between exploratory research
and testing. Otherwise they would have been ashamed to commit their
post-hoc data selection after the American test).
Ertel's position is that the Gauquelin data still show a Mars effect, in
the sense that the effect is stronger in groups of more excellent
athletes. However, Ertel's analysis is poor. I have reanalysed it,
and (a) the effect is absent if one only looks at the French champions
(b) the effect is present if one does not distinguish between champions
whose result were and weren't published by the Gauquelins. As Ertel's
sources were partly used by the Gauqelins themselves to establish who
was good and who was not (which gave a bias in their results) the
Ertel eminence effect is very dubious. (c) Moreover, Ertel himself has since
"discovered" that this eminence effect is "reversed" among the very high
top (after some more results I expect him to find a sinusoidal behavior...).
And even though Ertel knows the Gauquelin data are biased, he keeps
forgetting the importance of the fact that the Gauquelins never bothered
to formulate what was a champion. All along the Gauquelin's idea has
been that the effect only shows up if you take your champions good enough.
The story about "cover up" refers to the treatment of the outcome of
the Zelen test. Originally this was a test to find a demographic
explanation for Gauquelin's results. The test came out as expected (by
Gauquelin), namely that this explanation didn't work. No results were
"covered up", but responsible CSICOP people were kind of slow in
recognizing the results.
The way Ertel gets "results" is by very carefully redefining what
he means by "key Mars sector". There are many slightly different
definitions possible, and he takes the one that gives the best
answers. All this is post hoc.
Gauquelin's data originally comprised 2087 athletes. Expected number
of Mars athletes was 359, with a standard error of 17. Lots of physicists
don't think a deviation of less than 5 sigma merits serious investigation.
That would mean that an experimental finding of over 446 Mars athletes
starts to be interesting. Initially Gauqulin found 452, but after
recomputation (Correlation 4, 1983) there were only 435. The whole
effect is therefore so much at the border of significance that it hardly
woth so much effort.
Ertel thinks different. For him is 1.65 sigma already significant.
In a recent publication he even translates that into "It is highly
probable that there exists a relation between the two phenomena"
(in that same publication it turns out to be a silly computational
mistake, but subtle or not so subtle errors in computation or
experimental design apparently have no place in the minds of
More about this in the forthcoming Proceedings of the Third
EuroSkeptics Congress (Amsterdam 1991).
From: York H. Dobyns
To: All Msg #79, Oct-07-92 02:00PM
Subject: Statistical Evidence (was Re: Fwd: Astrology...)
Organization: Princeton University
From: ydobyns@phoenix.Princeton.EDU (York H. Dobyns)
In article <firstname.lastname@example.org> email@example.com writes:
[...accusations by JWN of lies in the original posting, and of bias
on the part of Ertel and the Gauquelins, deleted; my concern is
only with some numeric, statistical assertions:]
>Gauquelin's data originally comprised 2087 athletes. Expected number
>of Mars athletes was 359, with a standard error of 17. Lots of physicists
>don't think a deviation of less than 5 sigma merits serious investigation.
Maybe "lots" of physicists don't, but this physicist hasn't met many of
them. I *certainly* would not put up with someone who handed me a piece of
apparatus and said "Oh, by the way, we tested the output and in terms of
our measurement uncertainty it was only 4 sigma out of spec, so we figured it
must be OK and didn't bother doing any more measurements." The overwhelming
majority of papers I've seen in physics are content to use at most 95%
error bars or the equivalent: that happens to be about 2 sigma for a
one-dimensional parameter measurement. Sometimes the conservative researcher
reports a 99% confidence interval instead, that's about 2.6 sigma. I find
this statement of JWN's utterly outrageous.
>That would mean that an experimental finding of over 446 Mars athletes
>starts to be interesting. Initially Gauqulin found 452, but after
>recomputation (Correlation 4, 1983) there were only 435. The whole
>effect is therefore so much at the border of significance that it hardly
>woth so much effort.
[...JWN concludes with snide remarks directed at Ertel's use of 1.65
sigma--the standard 95% one-tailed confidence level--as a significance
criterion, and with a sneer against "significance-fetishists."]
Significance-fetishists, eh? Well, let's apply a proper Bayesian
approach to the numbers that Jan finds so unimpressive--statistical
significance is a concept that doesn't even appear in that
formalism. The data in JWN's posting reproduced above report a
set of 2087 Bernoulli trials in which the theoretical expectation
is that 359 should fit a given criterion. In fact 435 cases fitting the
criterion are observed. What can we say about p, the probability that
one of these random trials fits the criterion (i.e., the probability
that one of these outstanding athletes has Mars in a "meaningful"
position by astrological standards)?
The null hypothesis ("There is no Mars effect") is that p=0.1720.
Call this H0.
A completely uninformed alternative ("There might be a Mars effect,
but there is no information we can use to predict its magnitude in
advance of the experiment") is simply 0
E-Mail Fredric L. Rice / The Skeptic Tank