Exegesis Volume 3 Issue #60

From: "Mark A. Melton"
Subject: Is astrology a statistical science?

Exegesis Digest Sat, 24 Oct 1998

Date: Fri, 23 Oct 1998 15:14:21 -0700
From: "Mark A. Melton"
To: exegesis
Subject: Is astrology a statistical science?

Subject: Re: Exegesis Digest V3 #59
Date: Thu, 22 Oct 1998

Tallman said (in part): ....
 > I don't know whether the mechanisms at work in the
 > celestial influence of storms is known or unknown. The rest, mass trends,
 > and cyclic emergence and decline theories are, as far as I know, magical
 > processes.
 > What constitutes experimental science? Is that a field that is subject to
 > investigation by testing an hypothesis by experiment? Or does that mean a
 > field in which the aspects thereof can be manipulated to observe results?
 > Or both?
 > I think that astrology could well become an experimental science. It lacks
 > the proper protocols, and it lacks a theoretical base, even the bare
 > structure within which one can be built.....

An experiment is the formal process of carrying out some manipulative PROCEDURE, a "TRIAL," on EXPERIMENTAL MATERIAL, and observing the OUTCOME. Chemistry and physics are the best examples. On the cutting edge of these fields, the mechanisms are not always well known, or known at all, but there are hypotheses. The hypotheses are more often wrong than not, but some hypothesis is required to guide the design of an experiment. The construction of hypotheses to test by experiment, and the exploration of possible mechanisms and inferences by mathematics is indeed often beautiful to the specialist who understands them. The element of aesthetics in science is definitely present; scientists prefer the beautiful theory over one not so beautiful, so the dichotomy, either beauty or science, is false. I see no reason why astrology, or some aspects of astrology, cannot be both.

Following R.A. Fisher's approach, (a well known British statistician) a statistical, sampling experiment can be defined in which the experimental material is not manipulated, as it is in chemistry, but carefully _selected_ by some randomization scheme to isolate certain influences, cancel out others, investigate correlations among attributes (variables) and their interactions. Very elaborate experimental designs have been cooked up, such as Latin squares, randomized blocks, split-plot designs, factorial experiments, and so on. Originally many experimental designs were for agricultural studies and the language reflects that. They may combine statistical sampling of natural factors plus manipulative experiments, and this whole field, statistical experimental design, is quite technical. The book I am looking at, "Experimental Designs, 2nd Ed.," by Wm. G. Cochran and G. M. Cox, was published 41 years ago --so you know where I'm coming from!

We cannot manipulate astrological experimental material --people, for the most part-- in the way chemists manipulate chemicals, to put it as simply as possible. I have sometimes wished I could cross an Aries man and a Virgo woman and see what the result would be, but you just can't do that! We can to an extent pick and choose the people we include in our sample, and if done systematically this could be called a statistical or sampling experiment. Of course, this selection usually must be done without knowledge of the individual charts, or if we choose charts, without knowedge of the individuals' personal traits.

If human cloning becomes commonplace, perhaps we could have individuals with exactly the same genetic makeup born at different months and years and isolate the effects of nurture and astrology --at least eliminating the effect of heredity (yes I know, astrological features often look as though they follow Mendelian laws of heredity --see below). Human cloning is not now ethical, but might be possible. Perhaps it will not be too long before identical twin embryos are frozen then implanted and born at widely different times. This is still in the "gray area," at least for me, but it would tell us a lot about what is hereditary and what is nurture and/or astrology.

I have tried some statistical experiments in astrology but have found the sampling problems are almost overwhelming. Even in this day and age getting accurate birthtimes is not as easy as it should be. European birth data are often rounded off to the nearest quarter-hour. In my opinion such data are almost worthless. (1)

Another point is the planets are exceedingly non-random in their motions, hence we are always dealing with intact samples. Almost my entire client data-base has Pluto in Leo or Virgo. Doing timed charts for large families also shows that the in-group variance is always lower than random --the same signs 3-4 signs keep showing up over and over. If the sign-house equivalences are considered the family effect is even more striking.

My approach has been to use astrology backwards to test my own ideas --call them hypotheses if you wish. Starting with my own family first, I tried to "predict" their times of birth, not knowing in advance anything about their birth times. In most cases the subjects did not know either and had to look it up. The object was to find a particular time of birth, then see if that predicted time could be verified by birth certificate or other recorded data. I was within one minute for my father, to the exact minute for my sister, but I was off completely on my mother. I have also tried this on a few good friends and former students and was within a minute in several cases, and off quite a bit in others --in the latter case I guessed a person had Aries rising, when he actually had Mars rising in Cancer. Usually when I was right I was very close to the B.C. Time, and when I was wrong I was clear out in left field.

Unfortunatelly, this approach is limited by the number of people one knows well and who can find their own birthtime reliably. Also it is not repeatable: Once I've done it I obviously cannot repeat for that person. Other astrologers can try to replicate my "trials" but the results would be inconclusive --no other astrologer knows my family, for example, like I do. To be a science the experiment must be repeatable by anyone.

Hope this is interesting!

Mark A. Melton FIRST LAW OF PREDICTION: Hindsight is an exact science.

Notes: (1) I am looking for a female b. 3 April 1966 at 6:02:30 a.m. MST near Denver, Colorado -would like to have a synopsis of life history.


End of Exegesis Digest Volume 3 Issue 60

[Exegesis Top][Table of Contents][Prior Issue][Next Issue]

Unless otherwise indicated, articles and submissions above are copyright © 1996-1999 their respective authors.