From www.astrology-and-science.com         Click here to return to home page

Phillipson interview of researchers
Part 2. Results of scientific research into astrology

8. The picture emerging from research

OQQ8.1
What has been going on in the world of astrological research, and what kind of picture is emerging?

Researchers: If only subjective values matter, see Figure 1 again, the objective results that have been reported to date will be irrelevant whether positive or negative, as will much of this interview. Otherwise we can summarise the emerging picture as follows:

If we are to accept the claim that astrology provides reliable knowledge and accurate diagnoses, it has to do this under conditions where errors in our reasoning cannot intrude. Yes, the chart may fit the person, but does it fit better than other charts? Yes, clients may identify with their reading, but can they pick their own reading out of several? Yes, the prediction was a hit, but how many were misses, and is the hit rate better than that achieved by tossing a coin or by informed guessing?

Many such tests have been made, some by scientists, some by astrologers, and some by scientists and astrologers working together. Many have been made by ourselves since the 1970s when we first started our researches. But over the whole range of approaches the answer has been generally negative. The occasional promising result has not been confirmed, which illustrates the importance of not depending on just one study. Half a century of research into astrology, using techniques incomparably more powerful than those available to the Babylonians and Greeks, has failed to reveal effects (or at least effects commensurate with astrological claims) beyond those due to ordinary causes such as errors in reasoning. Which of course is very disappointing to those of us who have been astrologers. On the other hand it does not deny that future tests may be more positive, or that astrology may be fruitful in subjective ways.

EQQ8.2
Could you give an example of a promising result that was not confirmed?

Researchers: What got newspaper astrology columns off the ground was the apparently amazing accuracy of the British columnist R H Naylor. Unlike columns today, his column included predictions of national and world events as well as birthday predictions, then made by birth date rather than by sun sign. In the 5 October 1930 edition of the London Sunday Express, he made this prediction among several others:

"Earthquakes will occur, mostly near deep-sea levels, and affecting peninsulas, in the autumn quarter of 1930. They may not actually occur in October -- though from the 8th to the 15th is a real danger point -- but they will be exceedingly likely in November or December. British aircraft will be in danger about the same date."

On 5 October 1930 the great British airship R101 crashed in a storm near Paris. There were 46 dead and 8 survivors. Newspapers showed pictures of terrible wreckage. Naylor said this about his success:

"My prediction last week was based on a very simple observation. It can be proved that, whenever the new moon or full moon falls at a certain angle to the planet Uranus, aircraft accidents, electrical storms, and sometimes earthquakes follow. Now ... the configuration referred to [Full Moon opposition Uranus] occurred on October 7; the destruction of the R101, therefore, prematurely fulfilled the indication."

At first sight Naylor's hit seems quite amazing. Which is why his column took off. But the hit is ambiguous. Naylor's forecast said 8-15 October, during which time no aircraft crashes, electrical storms or earthquakes were reported by British newspapers, so no hits can be counted there. The crash occurred when the Moon was 33 degrees from conjunction, a long way from modern orbs of a degree or so. And later studies of air disasters did not confirm any link with the Moon and Uranus. So was it a hit or a coincidence? In fact an isolated hit is no more meaningful than an isolated chart factor. It is the totality of research which matters.

Less ambiguous were the forecasts of Edward Lyndoe, Naylor's rival in The People, who in 1939 consistently saw no war, e.g. "I see absolutely no signs of a Great War during 1939" (1 January), "The Nazis attacking Britain? Don't make me laugh! Not a sign in my charts" (25 June), "Hitler will not do it!" (27 August). Then, ingeniously, after war was declared, "A madman against the stars!" (3 September).

So were the columnists accurate or not? To find out, the London investigative magazine Picture Post tested the accuracy of the top five newspaper astrologers (including Naylor and Lyndoe) against nine outstanding events of 1939-1941 such as the German invasion of Poland and the collapse of France. Bear in mind that these astrologers were specialists in event prediction, so for events as momentous as these their accuracy should be as good as it gets. But they scored no better than informed guessing, the most notable result being "that the astrologer so often fails to make any reference whatever to the event in question" (from the 6 September 1941 issue).

Later, the issue of 27 September 1941 contained this letter from the predictive astrologer P J Harwood: "Like other astrologers, I have made my mistakes ... but, on the other hand, a large number of very close hits have been scored. I am sending you a copy of my booklet, When the War Will End, and if you are really interested in it I should be obliged if you could give it some publicity." To which the editor pithily replied: "Readers would be wise not to make their plans for peace celebrations too definite. Mr Harwood's booklet foretold invasion in May 1941, a separate peace with Italy in July 1941, and considered Russia unlikely to be implicated seriously in war" (all wrong, including his prediction that the war would end on Christmas Eve 1941).

Interestingly, in the days when few tests existed, most astrologers held that astrology was demonstrable and therefore testable. Just try it, they said, and you will experience for yourself how well it works. In fact those of us who were astrologers said exactly the same -- it was how we got hooked in the first place. But the advent of scientific tests and their predominantly negative results has caused a U-turn. Today many astrologers hold that astrology is not testable after all, thus denying the bad news, as if this did not also deny that astrological correspondences could be confirmed by experience or even discovered in the first place.

OQQ8.3
Astrologers who say astrology is not testable may do so because they see it as a divinatory tool more akin to a ritual that prepares the mind to intuit what needs to be said. In their view the working of this tool is not necessarily testable in a scientific way.

Researchers: Their view is hard to understand. It is like saying we don't know how gravity works, therefore we cannot test the fall of apples. The issue is whether the astrology ritual works better than a control ritual, e.g. by providing new information or by improving the clients self-esteem. Much is testable here.

NQQ8.4
Alternatively astrologers who say astrology is not testable may be referring to astrology as a world view. They may simply mean that an entire world view is not testable.

Researchers: This hardly helps. There are very many world views available in philosophy and religion, all of them untestable. But as shown in the journals devoted to metaphysics and the philosophy of religion, they can still be critically examined for consistency and how well they fit with the observed world. At the end of the day we may have many reasons, some of them empirical, for preferring one over another. We do not have to buy a flight ticket without knowing the destination.

Furthermore, astrologers in their daily practice do not invoke an entire world view. Nor do they agree on what this view is. Instead they invoke various sub-views, or what we called "real intrinsic meanings" such as Leos are generous or squares are difficult. If astrologers cannot test these sub-views, nothing about them can be discovered including their truth or falsity. To hide this awkward point some astrologers have retreated into obscurity, for example by defining astrology as a sacred science or as the language of individuality or as the study of potentials. Their obscurity also hides the underlying distinction between subjective and objective astrology, which (once again) is essential to avoid confusion.

9. Gauquelin, sun signs, the whole chart

OQQ9.1
You are certainly painting a gloomy picture. What would you consider to be the most convincing research in favour of astrology, and what are the weaknesses in it?

Researchers: Astrologers tend to quote the Gauquelin results as the most convincing evidence in support of astrological claims. (These results showed that eminent professionals tended to be born when the planet relevant to their occupation was just past rise or culmination; this tendency was later called the Mars effect, but depending on the occupation it could equally well have been called the Moon, Venus, Jupiter, or Saturn effect.) Indeed, two of us have worked for some years following them up.

Gauquelin's work was certainly the most rigorous of its time. But his planetary effects, even though independently confirmed by us, are too tiny to be of the slightest practical value. For example, given that Mars tends to be just past rising or culminating in the charts of eminent sports champions, this information is of no practical value unless your client is eminent (say 1 in 20,000 of the population), in which case the information is already superfluous.

Furthermore, Gauquelin's other results showed no effect for half the planets, or for signs, or for aspects, and work by one of us (Ertel) has disconfirmed initial hints of a planetary link with character traits, all of which is contrary to astrological claims. So the relevance of the Gauquelin results tends to be overstated. On the other hand some astrologers claim that the Gauquelin results are at best peripheral to astrology. Here again our subjective-objective distinction resolves the conflict.

OQQ9.2
But doesn't the Gauquelin research deserve more acknowledgement of its implications? Such as the statement made in Recent Advances that the Gauquelin results "provide, for the first time, rigorous and objective evidence about the basic fundamentals of astrology, upon which everything else depends ... something that astrologers themselves, despite millennia of study, have consistently failed to do" (page 394).

Researchers: A lot of research has occurred since that statement was made in 1977, and the results require it to be modified. There are two separate issues at stake here. First is the reality of the Gauquelin findings. Research by one of us (Ertel) has found that, despite some deficiencies, the basic findings withstand rigorous tests. Second is their relevance to astrology. Research by another of us (Dean) has found that they may have an ordinary explanation, hitherto unsuspected, in which case it would be premature to conclude that they are relevant to astrology. Real yes, relevant perhaps not. The reason is a simple one -- the Gauquelin data show evidence of manipulation by parents, for example there is a consistent deficit of births on the 13th. But if parents can manipulate dates then why not hours and thus planetary risings and culminations? So the Gauquelin findings might be due to parents, not planets.

OQQ9.3
Are there no other promising areas of research?

Researchers: Twenty years ago they seemed quite numerous. There was John Addey's harmonics ("promises to revolutionise astrology" said Recent Advances), Donald Bradley's Jupiter Pluvius (the tendency for heavy rainfall to occur if Jupiter aspected the local meridian when the Moon entered sidereal Capricorn), John Nelson's radio propagation quality (worsened by hard heliocentric aspects), the Mayo-Eysenck sun sign zigzag ("possibly the most important development for astrology in this century" said Phenomena), and Vernon Clark's matching experiments (see later). Even the mathematics (of aspects, of orbs, of probabilities, and so on) seemed promising.

Interest in research grew accordingly, only to fade away as artifacts were discovered. As in the above cases, once artifacts were controlled the supposed astrological effects disappeared. The most recent example is Gunther Sachs's best-selling book The Astrology File (Orion 1998), a supposed proof of sun sign effects. One of us re-analysed his results but found only artifacts. The whole massive study (it involved several millions of cases) was not careful enough.

OQQ9.4
Let me come back to some of the research areas you mentioned right at the start, starting with sun sign columns. What work has been done here?

Researchers: Lots. When labels and other cues are removed, people cannot pick the sign that is supposedly theirs. Yes, some columns might be uplifting, but the point is: Does the use of sun signs add otherwise unattainable truth and uplift to sun sign columns? Or do they merely con columnists and readers into believing that their "twelve thoughts for the day" are more meaningful than if they appeared in say Kahlil Gibran or a desk calendar? Research indicates the second.

Sun sign delineations have the distinction of being more widely tested by astrologers and researchers than any other factor in astrology. Altogether something like a hundred tests have been made involving a total of several millions of cases. The usual approach is to compare the distribution of births (of bakers, bankers, extraverts, and so on) across sun signs with that expected by chance. Unfortunately the latter is affected by astronomical and demographic variables to such an extent that the expected distribution can vary a great deal from country to country, from year to year, from place to place within the same country, and from one social group to another. When these variations are not recognised (as is usual), they are easily mistaken for sun sign effects. But once they are controlled no evidence for sun signs is found.

For example, in 1980 Michel Gauquelin compared the biographical details of famous people with their Sun, Moon and Ascending signs. He searched thousands of biographies to find people who had the qualities attributed to sign X, but they showed no tendency for their Sun, Moon or Ascendant to be in X rather than in any other sign. Other controlled studies have consistently supported this finding. In short, tests more rigorous and sensitive than any astrologer's experience have consistently provided no evidence whatsoever that signs mean what they are supposed to mean.

Interestingly, a weak but statistically significant link between sun sign delineations and extraversion scores was reported in 1978 by Mayo, White and Eysenck (Social Psychology 105, 229-236), advance notice of which was hailed by astrologers as "possibly the most important development for astrology in this century" (Phenomena 1977, 1.1, 1). The effect disappeared when people unfamilar with sun signs were tested, so it had a simple explanation -- prior knowledge of astrology. Ask Sagittarians (who are supposedly sociable and outgoing) whether they like going to parties, and their answer might be tipped by astrology in favour of yes rather than no. The bias may be unconscious and very slight but in large samples it can become surprisingly consistent. Of course it has nothing to do with sun signs as normally conceived.

OQQ9.5
Moving to the whole chart, has any work been done to test the accuracy of individual astrologers and their procedures?

Researchers: Again, lots. Can astrologers match birth charts to their owners? Can clients tell which chart reading is theirs? Unfortunately in each case the results have been no better than chance, see Figure 2

Figure 2. How astrologers (54 studies) and clients (11 studies) performed when put to the test

Left. According to astrology books, birth charts accurately match their owners. So in studies where astrologers have to match charts to owners, the hit rates should pile up close to 100% (the hit rate is the proportion of correct matches), perhaps something like that shown in grey. But in 54 studies made during 1950-2005 totalling 742 astrologers and 1407 charts, the mean hit rate was 51.7% as shown in black, or not usefully better than tossing a coin. Included are studies where only top astrologers were used, or where more information was given than in a normal consultation, or where artifacts might reasonably be suspected, so the results are optimistic. Nevertheless the astrologers could not usefully tell the difference between authentic charts and controls.

Right. Similarly, in studies where clients have to pick their own chart reading out of several, the hit rates should pile up close to 100%, perhaps as shown in grey. But in 11 studies totalling 423 clients where we can be reasonably sure that cues such as sun sign were absent, the mean hit rate was only 50.2% as shown in black. The client's own chart reading fitted them no better than somebody else's.

OQQ9.6
I thought some early results, namely those of Vernon Clark, were dramatically better than chance.

Researchers: Vernon Clark's results are in the upper third of the results shown in Figure 2 left. However, his results and those of most other matching tests have fatal problems due to their small sample sizes, typically 10 birth charts or 10 pairs of birth charts. Imagine a hand of 10 playing cards picked at random. Even though there are equal numbers of red and black in the pack, we seldom end up with equal proportions in our hand. Our hand has been affected by sampling variations. Of course the bigger our hand the closer we get to the proportions that exist in the pack.

The same with birth charts. Given that the stars only incline, charts will tend to be of two kinds, those that fit their owners (call these red) and those that don't (call these black). Suppose like Vernon Clark we want to know the proportion of owners with red charts. So we collect ten owners and send them to astrologers. If the astrologers know their stuff they will quickly discover which charts are red and which are black. But the reds in such a small sample will be determined much more by sampling variations than by the proportion in the population (which is what we want to know). Worse, for any given study we cannot remove the sampling variations, just as we cannot remove the sampling variations for any given hand of cards. They are there to stay. So our results will not tell us what we want to know. Yes, we may have three or seven reds instead of the five predicted, but so what?

Vernon Clark and nearly everyone else used few birth charts and many astrologers, which is precisely the combination guaranteed to produce this kind of uninformative outcome. Figures 3 and 4 show why such results cannot be taken at their face value.

Figure 3. How a small sample size scatters the hit rate.

The upper curve shows the hit rates from 54 studies of astrologers matching birth charts to their owners. The hit rates show a marked scatter from a low of 40.0% to a high of 72.0% (grey dots = Vernon Clark's results). At first sight the higher values seem encouraging, but as we shall see, the scatter is due to small sample sizes, typically 10 birth charts or pairs of birth charts per study (the range is 4-120, Vernon Clark used 10).

The lower curve shows what happens when the studies using 120 birth charts (black dots) are subdivided into 24 studies of 10 charts each. The same marked scatter emerges even though the original hit rates (50.2% and 50.5%) are close to 50%. Further subdivision increases the scatter. Even worse, if we now report only the higher hit rates and ignore the rest (astrology editors dislike negative results), we will be creating support for astrology where none actually exists.

Figure 4. How large sample sizes home in on the truth for astrology (left) and graphology (right)

Left. Hit rates for 54 studies of astrologers matching birth charts to their owners, grouped by sample size (number of birth charts or pairs of birth charts per study). As the sample size increases from left to right, the hit rates show a general decrease in scatter as shown by the trend in bar lengths, and a closer approach to the truth. The length of each bar is q2 standard deviations, which is the range within which just over 95% of the results from a large number of repeat studies are expected to fall. In each case the chance level of 50% falls well inside each range, which suggests that these astrologers were generally unable to match charts usefully better than tossing a coin, a point confirmed below.

Right. Decreasing scatter and increasing accuracy is a general consequence of increasing sample size. Thus the same trend is shown by 109 studies of graphologists matching handwritings to their owners, but this time the trend is more uniform due to the larger number of studies. Mean hit rate for the 109 studies (55.8% sd 5.6%) is higher than for the 54 astrology studies (51.7% sd 5.8%) due to personal information inadvertently present in some handwriting samples, which the graphologists took into account even though they are supposed to ignore content. But neither mean differs significantly from 50% (p = 0.30 for graphology and 0.77 for astrology). Tossing a coin would be easier, quicker, and generally just as accurate.

So what can be done? The solution is either to use many charts, as for the rightmost dot in Figure 4 left, or to submit the collective studies to what is known as meta-analysis. Meta-analysis subtracts the sampling variations to see if there is anything left (something not possible with an individual study), and therefore reaches sounder conclusions.

In this case meta-analysis of the 54 studies confirms what we already know from Figures 3 and 4 -- the sampling variations are so massive that there is nothing left for astrology to explain including the apparently encouraging results of Vernon Clark. Yes, the differences between results might seem to reflect differences in test quality, but such a view is mistaken. Here the differences between results reflect only sampling variations. Differences due to test quality may exist but they are swamped by sampling variations. In other words the reported hit rates offer no support whatever for the accuracy of judgements based on birth charts. Tossing a coin would be just as accurate.

(An earlier meta-analysis when the number of studies was 43 is given in visual form in Astrology in the Year Zero pages 146-147 and in Correlation Northern Winter 1998 page 75, but is not included here because Figures 3 and 4 are less technical and easier to understand. However, it suggests that future studies will show no improvement in hit rate, and that a publication bias exists against negative results.)

NQQ9.7
Have any tests been made of horary, medical, or business astrologers?

Researchers: Not systematically, but the isolated tests that have been made have not been encouraging. To be sure, some successes have been claimed, but on inspection the methodology has proved to be suspect. Tests of money market astrologers are particularly suspect because there is no way of knowing how many market investors are actually using astrology, so we cannot decide whether successes are due to astrology or to investors using astrology. Ironically, once a majority of investors are using astrology, the market cannot fail to contradict astrology. Thus if astrology says sell on the expectation that the market will go up, and all the astro investors sell, the market is likely to go any way but up.

At first sight horary astrology seems ideal for testing because it is supposed to give clear-cut yes or no answers. Thus Charles Carter could say of horary charts cast for him that "they have usually been downright wrong and never strikingly right" (Astrological Journal December 1962, a quote we repeat later in 18.12). But as noted by Koen Van de moortel (Astro-Logics 1997/2002, self-published), horary astrology has so many conflicting opinions about technique that "it's much easier to toss for it, if you really want a cosmic answer" (page 102). Koen points out that each opinion is based on years of experience, so if horary astrology actually worked the opinions should agree. But they generally disagree, often spectacularly. In other words researchers have no clear place to start, so perhaps it is best to wait until horary astrologers put their house in order.

NQQ9.8
Have any astrologers challenged scientists to test them?

Researchers: Such cases are not numerous but they have happened. For example one US astrologer was so confident he could match chart to appearance that he ran a newspaper ad challenging scientists to test him. Which they did, not once but twice, under conditions he was completely satisfied with. But each time he scored no better than chance. One French astrologer was confident he could pick cause of death but again the results were no better than chance. The crucial point here is not the poor results but how the astrologers could be so confident of having a skill they did not in fact possess. Reasoning errors score yet again.

10. Isolated vs multiple factors, time twins

NQQ10.1
Tests of one, or just a few, isolated factors have always seemed problematic, simply because they do not reflect the way in which astrologers read charts. Any competent astrologer will avoid using factors in isolation, and will try to look at each in terms of its relationship to the rest of the chart.

Researchers: But sun sign astrologers by definition never do this, and they might not agree that they are not competent. In fact even the most aggressive advocates of the everything-is-affected-by-everything-else viewpoint, such as the US astrologer Dr Glenn Perry, can still fill their writings with statements such as "individuals with Venus square Pluto are distrustful of love." The same is true of every astrology textbook, as shown by the following examples from reputable authors:

- Aquarians are gregarious and enjoy social interactions.
- Leos have yellow bushy hair, Aquarians are never short.
- Positive signs are characterised by extraversion.
- Planets mostly below the horizon indicate introversion.
- Saturn rising indicates an inhibited personality.
- Neptune in fourth house is artistic and musical.
- Easy Mercury-Mars aspects have good eyesight and hearing.
- Hard Moon-Uranus aspects incline men to divorce.
- Adverse Mars transits incline to accidents and injuries.
- Bucket patterns inspire or teach or become an agitator.
- Progressed Sun-Venus contacts usually indicate marriage.

The statements are respectively from Hand's Horoscope Symbols page 236, Mann's Round Art pages 147 and 159, Larousse Encyclopedia page 223, ditto page 133, Ebertin's Combination of Stellar Influences page 192, March and McEvers The Only Way to Learn Astrology Volume 1 page 261, Hone's Modern Textbook page 189, Llewellyn George's A to Z page 153, ditto page 465, March and McEvers Volume 2 page 118, and Davison's Technique of Prediction page 55.

Ironically Perry claims that testing such statements is meaningless, but evidently it is not considered meaningless for textbooks to publish such statements in the first place. Nor is it considered meaningless to welcome positive evidence for isolated factors, e.g. Gauquelin. In our view whatever is good enough for textbooks and for welcoming is good enough for testing.

NQQ10.2
Has anyone tested a large database, say one of sports champions showing a Mars effect, to see if invoking additional factors such as midpoints and dignities makes the effect stronger?

Researchers: Yes, and no it didn't. Invoking additional factors necessarily reduces the hits because the frequency of having A and B is always less than the frequency of having A alone (unless of course all have B, in which case adding B accomplishes nothing).

If we make it A or B instead of A and B, the problem is that, given enough additional factors, some can always be found that will make the effect stronger purely by chance, in the same way that shuffling cards enough times will sooner or later give you a full house. So we need to divide the sample in half to allow the results from one half to be tested on the other half. Unfortunately this immediately increases the sampling variations, whose effect was so decisive in Figure 2.

Even if we stay with the Gauquelin factors the outcomes are no more promising. Thus one of us found that the probability of an eminent birth did not increase when two or more relevant planets were in Gauquelin sensitive zones, or when the birth time was more precise, or when the planets were closer to the earth, all of which is contrary to what might be expected (but not if the effects are due to parents not planets, see 9.2).

In general the utility of any chart indication depends on two things, namely (1) the effect size, the correlation between the indication and reality, and (2) the base rate, the rate of occurrence in the general population. For areas of client interest such as illness or managerial ability or musical ability or lack of confidence, the base rate in the general population is typically 10% or less. For such low base rates the indication (whether by astrology, graphology, palmistry or whatever) will increase accuracy above the base rate only if the effect size exceeds about 0.4 or 0.5. Otherwise using the indication will make the prediction worse. But Figure 2 indicates a mean (and statistically nonsignificant) effect size for the whole chart of only 0.05, which also happens to be the effect size for Mars and eminent sports champions. So the chances of exceeding 0.4 or 0.5 do not seem good.

NQQ10.3
This seems to be an argument against ever using multiple factors.

Researchers: The argument here is only that the effect size for factors A and B will be less than for factor A alone. But there are many other ways in which multiple factors could operate such as A or B, or if A then B else C. If collective multiple factor X gave an effect size of 1, the effect size for X and B would still be less, but then B would not be needed. The real argument is not against using multiple factors but against using not-needed factors, which in the case of eminent sports champions seems (as far as we know) to be everything except Mars in key sectors. Indidentally the last is what might be predicted if the effects were due to parents and not planets.

OQQ10.4
Perhaps a refined and balanced technique is the only one which will work in astrology.

Researchers: The idea that only a refined and balanced technique will work seems incompatible with the high level of disagreement about techniques, all of which are claimed to work. If they actually work, the technique is evidently unimportant, and astrologers should be piling up hits in Vernon Clark experiments instead of scoring at chance level. But if they actually do not work, astrologers are evidently unable to tell. So the supposedly uniquely-true technique referred to in your question would be either redundant or unidentifiable. Nevertheless let us speculate to the contrary. Let us say okay, so elephants don't fly, but given the right conditions maybe tigers might fly, or maybe giraffes might fly. Clearly this won't do. There comes a time when we have to grasp the nettle.

OQQ10.5
And conclude that nothing in the jungle can fly?

Researchers: The point is that nothing in the jungle is flying. Maybe many things can fly, but this is back to speculating forever. We still have to grasp the nettle, namely that when nothing is actually flying, nothing is actually flying. People do not travel to Heathrow on the off-chance that somebody will suddenly discover aeroplanes.

EQQ10.6
How complex does your research get? Is it possible to design tests of sufficient complexity to allow for all the permutations of meaning between the various chart factors?

Researchers: Your question implies that astrologers are somehow better equipped to deal with complexity than scientific researchers are. But because researchers are more careful than astrologers, this is like saying that only sloppiness produces good astrology, which seems ludicrous. It will certainly be news to those who set astrology exams. Indeed, the inability of astrologers to agree over most things would seem to arise directly from their inability to deal with complexity.

Actually the complexity issue is a non-issue because (1) it has never stopped astrologers from practising, and (2) we can test astrologers directly as in Vernon Clark tests. If astrologers cannot perform better than chance then the complexity said to be responsible for their success (but which is supposedly beyond the grasp of scientific researchers) does not exist. Before we worry about complexity or any other detail we need to know if it actually delivers, else we end up chasing phantoms.

Nevertheless imagine that a level of complexity exists as yet undreamt of, and that it actually does deliver. Can tests be designed to cope with it? The answer is yes. Examples of approaches that take any complexity in their stride are multiple discriminant analysis, which finds the factors that discriminate best between a set of charts and controls, and probabilistic modelling using item response theory, which does everything an astrologer can do short of using ESP but includes what no astrologer can do, namely address astrology's inherent uncertainty directly. Both techniques require a computer. A scan through back issues of Correlation will lead directly or indirectly to examples of tests that address the complexity issue.

Another approach is to bypass complexity issues altogether, notably by a test of time twins (persons born close together in time and place), which avoids all problems of how the various chart factors should be interpreted and combined. A failure to validate sun signs might be dismissed as a failure to use the correct interpretation or to properly allow for competing chart factors, but a failure to show that time twins are significantly alike is less easily dismissed. In a city of one million people about 40,000 will have a time twin within 1 minute, so time twins are numerous enough. Here is the calculation:

The occurrence of time twins follows what is known as a Poisson distribution, see any statistics textbook. Once we know the mean number of births in a given interval, say ten minutes, for a particular location, say London, the number of births with at least one other birth in that interval can be obtained using a hand calculator. Time twins are surprisingly numerous. Thus in a city of one or ten million people with a typical birthrate of 2% per year, about 4% or 32% of the people will have at least one other person born there within one minute, and 32% or 98% will have at least one born within ten minutes.

EQQ10.7
Some striking cases of time twins have been reported, such as Samuel Hemmings and King George III. Have you looked into these?

Researchers: One of us has. Legend says that the prosperous London ironmonger Samuel Hemmings and King George III were born at the same hour and died at the same hour after lives showing striking similarities such as being married on the same day. But a careful check of contemporary records showed that the ironmonger's name was not Samuel Hemmings but Richard Speer, and of the events only the simultaneous death could be verified, the rest being most likely fabricated. Several other well-known cases of time twins could not be verified, either through lack of records or evident fabrication. But the number of exact time twins existing even in Western history is so enormous (hundreds of millions) that many striking cases are to be expected by chance alone, so the reported cases are unremarkable.

More systematic tests of time twins have recently been made. For example Roberts and Greengrass, in their The Astrology of Time Twins (Pentland 1994), collected a total of 128 people born on six dates, but they found no clear parallels in personality scores, appearance, handwriting, names, interests, occupation, or life events. The strong similarities predicted by astrology were simply not there. Even so, they claimed that the proportion of "close resemblers" increased as the birth interval decreased, which would provide some support for astrology. But an independent re-analysis (Journal of Scientific Exploration 1997, 11, 147-155) found that their division of the data had given samples very uneven in size, a point concealed by their use of proportions, which violated the technical requirements of their test. When the data were divided more evenly, or when tests were applied that did not require dividing the data, the effect disappeared. Roberts and Greengrass also claimed that personality differences measured within dates were smaller than the same differences measured between dates, which would again provide some support for astrology. But the re-analysis found they had overlooked the natural changes in personality with age, which immediately explained their result. Furthermore their data showed evidence of contamination by prior knowledge of astrology. In other words, any apparent time-twin effect was clearly an artifact. It was another case of astrologers being insufficiently careful.

Now look at ordinary twins. The time interval between births is much the same whether the twins are identical or non-identical. Therefore according to astrology they should be equally alike. But they are not. Identical twins are very alike, but non-identical twins are generally no more alike than ordinary siblings. This is one of the oldest arguments against astrology, but it is generally ignored by astrologers.

NQQ10.8
Although similarities should be found in twins if astrology means anything, how widely do you cast your net in defining similarity? For instance, if one twin is an athlete and the other is a welder, is this allowed as a "hit" since both are expressions of Mars?

Researchers: It does not matter how we define a hit because we can always generate controls (e.g. by computer) to suit our definition. What matters is not the definition but the difference between twins and controls. But time twins lead to another problem, namely the way astrologers explain misses by appealing to birth time errors. For example an error of one minute can affect the exactness of aspects to angles and house cusps, and therefore can in principle explain differences between supposed time twins. The argument is that, if the birth times were known accurately enough, the misses would disappear. But then so would the rationale for applying astrology to the average client, whose birth time would rarely be known to this suddenly-necessary level of accuracy, let alone to the readers of sun sign columns. Astrologers seem to want it both ways, which along with the practical difficulty of finding exact time twins may explain why the area is relatively neglected -- it requires too much work and produces too much embarrassment.

11. Open-mindedness, influence of world views

OQQ11.1
The subject of open-mindedness came up earlier (in 2.5). What is it, and how is it cultivated?

Researchers: Open-mindedness means a willingness to explore new ideas and arguments. Everyone (especially scientists) thinks they are open-minded, in the same way that everyone thinks they have a sense of humour. But when evidence exists both for and against a belief, most people do not show low levels of conviction, which logically they should, but high levels of conviction either for or against, which logically is indefensible. Logically they should have open minds but in fact they have the opposite. Open minds, like the Scarlet Pimpernel, are damned elusive.

So what is happening? An open mind requires us to embrace uncertainty. But for most of us uncertainty is something we hate. Our need to avoid uncertainty explains the popularity of superstitious beliefs, which can be seen as attempts to reduce anxiety using ineffective techniques when effective ones are unavailable. Or as Bertrand Russell said, "What men want is not knowledge but certainty." No wonder astrologers are amazed when we express no interest in where the astrological chips fall. To them it seems inconceivable that we should not be either for or against.

How to cultivate open-mindedness? (1) Explore new ideas and arguments. (2) Have critical colleagues watch over you and act on their criticisms.

NQQ11.2
What you say seems to imply that it's impossible for a group of people to share the same prejudices, and to reinforce those prejudices in one another. Which is -- obviously, I think -- implausible.

Researchers: You are right. Which is why it is so important to avoid cliques of yes-persons and to place ideas in the open marketplace for examination, as science does but astrology rarely does. Whereas science reserves its highest praise for those who prove their predecessors wrong, astrology drums critics out of the corps. Criticism in astrology is simply not welcome. In our case we try to avoid prejudice by sending our pre-publication material to astrologers for critical comment, but in most cases the response is a deafening silence. For example we invited four astrologers with research experience to check our answers to your questions, just in case we were missing something, but without success.

NQQ11.3
Would you agree that, if people who think of themselves as scientists consider astrology at all, they tend to feel hostility towards it?

Researchers: No. We certainly don't feel hostility towards astrology. True scientists will consider many things during their professional lifetime, and they have to consider them dispassionately, otherwise there is no point. But scientists, like the rest of us, are likely to show a wide range of attitudes towards astrology, just as they would towards religion, medicine, or politics. So if they do feel hostility, we might ask why. Is it because they see it as involving a commercial exploitation of untruths such as sun sign astrology? Or because they see astrology as the opposite of good science, a closed system impervious to disconfirming evidence? Or because they cannot abide ideas contrary to their own? Their hostility (except in the last case) might be justified. But as scientists it is their job to rise above it.

NQQ11.4
And you believe that it is possible to have a genuinely disinterested approach to the subject of astrology's validity?

Researchers: Yes, it is certainly possible, as shown by Gauquelin and Eysenck. That is the short answer. But as we said in reply to an earlier question (11.1), the open-mindedness required to be disinterested (i.e. not taking sides) is elusive. It requires constant vigilance. People who investigate astrology often hold strong views otherwise they might have no motivation, but what matters is whether they are willing to consider other views, and to change their own views according to the evidence. Here is an example:

For decades Fred Hoyle strongly advocated the steady-state theory of the universe, but when the accumulating evidence supported the rival Big Bang theory, he changed over. He followed where the evidence led. End of story. Unfortunately astrologers tend to see research that supports their beliefs as okay no matter how badly done. They also tend to see research that contradicts their beliefs as incompetent or misconceived or simply stupid, no matter how well done. It seems difficult if not impossible for astrologers to be disinterested and open-minded. Very few are like Fred Hoyle.

OQQ11.5
Perhaps it is still necessary to question the underlying world views of researchers. How would you describe your own world views and how they fit in with astrology?

Researchers: As scientific researchers our world views (e.g. the view that being careful will be productive) are only tentative. Show us a better way of doing things and we will explore it regardless of our world views. Success is the arbiter, but so far astrology has failed to deliver. It has not delivered the results promised by astrologers, nor has it provided the evidence required for careful people to adopt an astrological world view.

Astrology may of course be more compatible with some world views than with others, but by itself this means little. For eighteen centuries the doctrine of the four elements was compatible with prevailing world views, but the doctrine is now known to be wrong. Astrology (like UFOs, channelling, hobgoblins or magic spells) still has to make its own case regardless of its compatibility with a particular world view. Make a convincing case and we will explore it.

NQQ11.6
But this would not deny the possibility that astrology might change present beliefs about the world?

Researchers: Absolutely, just as we cannot deny the possibility that the discovery of elves might change present beliefs about the world, or that a gust of wind might save us if we jumped off a cliff. But for astrology to change present beliefs about the world will require an objective astrology based on sound arguments, convincing evidence, and an underlying theory agreed to by astrologers worldwide, which are precisely the things it presently lacks. At the moment claims such as "astrology is the science of the future" are a clear contradiction in terms. A subjective astrology is of course incapable by definition of changing science-based beliefs about the world. However, this might no longer apply if the beliefs became religion-based as in the Middle Ages, as might happen if Islamic fundamentalists were to realise their aim of conquering the world.

OQQ11.7
How can you be sure that the results of research aren't just reflecting your world views back at you?

Researchers: Presumably an example might be researchers whose world view requires isolated factors to be meaningful, so they test only isolated factors despite protests from astrologers. The answer to your question is easy -- we use as many approaches as possible regardless of what a particular world view might dictate. As we said before, show us a better way and we will explore it. But turn your question around -- how can astrologers be sure their claims aren't just reflecting their own world views back at them? An example might be astrologers who ignore reasoning errors despite protests from scientists. The question is one that only research can resolve (and in our view has probably already resolved), using as many approaches as possible, yet astrologers show little interest in such research.

12. Reactions of astrologers to research outcomes

NQQ12.1
In the discussion so far, you keep returning to the theme that astrology fails to deliver. As far as I can see, the Gauquelin findings do not support claims that astrology, as practised, is true. But they also close off the option of concluding that all ideas of astrological influence are false. Surely this is a significant point?

Researchers: We don't see why. One swallow does not make a summer. The fact that Claim A seems to be supported does not mean it is useful or beyond ordinary explanation, or that Claims B-Z have to be true. The fact that flying elephants might conceivably exist, say in tornado areas, does not improve our chances of catching the next flying elephant shuttle to Heathrow. Similarly, we cannot deny the possibility that surfing in Hawaii affects the waves in Australia, but would we bet on it? Until astrologers can replace their ten-a-penny (see 1.3) armchair proclamations with good empirical evidence, or engage in rigorous debates like this one, why should we believe any of it?

Nevertheless, suppose that all the research ever done has got it wrong, and that we have a world where astrology works to the extent claimed in astrology books. Hunger and hardship have disappeared because economic trends and climate are predictable. Science has disappeared because horary astrology answers any question. So has competitive sport for the same reason. Cars and planes are hazard-free because assembly times conducive to accidents are routinely avoided. Crime, war, illness and divorce are unknown because predictable. Every person is empowered, self-actualised, spiritually enlightened, and knows their individual purpose and direction. Abuse of astrological knowledge is prevented by restricting it to those whose charts reveal due merit. This is astrology world. Now compare astrology world with the actual world. Bearing in mind that astrology has had two thousand years to get it right, can we conclude that it really does deliver? Probably not.

NQQ12.2
Perhaps there are only two possibilities: either astrology is bunk, or it is a subtle and elusive art. Hopefully readers of this interview will be gathering some basis upon which to form their own conclusions.

Researchers: There could be more than two possibilities. Thus astrology could act as a focus for discussion without any claim to validity, or it could act as a projective tool like inkblots, neither of which would be bunk or elusive. But the straightforward claims made in astrology books contradict the idea that astrology is a "subtle and elusive art", always assuming that there is a single such art that astrologers would agree on. For example the rear cover of the reprint of Ronald Davison's best-selling book Astrology (CRCS 1987) says that its simple keyword system "allows even beginners to start interpreting charts immediately with great accuracy." Nothing subtle or elusive here, not even for beginners.

NQQ12.3
I wonder how well the book would have sold if the back cover had promised, "this subject baffles people who have studied it for years, and you have no chance of ever understanding it fully"!

Researchers: Presumably not well at all even though it might be less misleading. But the point is, this is not just another astrology book. Ronald Davison (1914-1985) was a highly respected British astrologer. He succeeded Charles Carter in 1952 as president of the Astrological Lodge, serving in that position until 1982, and was editor of Astrology: The Astrologer's Quarterly from 1959 to 1983. A review of his book in the Astrological Journal for 1963 comments "But the test of a first-class production perhaps is that it should have something to teach the experienced student too, and here again Mr Davison scores some very good points. There is in his descriptions a great fund of original observation and the maturity of judgement of one who has listened attentively to many viewpoints and separated the wheat from the chaff."

Furthermore, in the CRCS reprint, the US psychologist and astrologer Stephen Arroyo describes how, in the 1960s, he and his friends would look up their chart factors: "In an amazing percentage of cases, we would find reliable descriptions of the person I was talking with, and in fact I well remember the looks of fascination and reflection on people's faces as they revealed how impressed they were with the accuracy that this kind of astrology could provide. In fact, in numerous cases, they could not hide the strong impact made from insights into their motives and needs that they had never really considered before in that light. I have continually recommended Ronald Davison's Planets in the Signs keyword system for over twenty years to those ... who want to develop astrology as a reliable science of human nature." Again, nothing subtle or elusive here.

Indeed, almost every astrological textbook makes it clear that astrology is not an occult subject, and that anyone with diligence and application can learn to set up and interpret birth charts. Complicated and demanding yes, subtle and elusive no. Or so we are told.

NQQ12.4
If astrology fails to deliver, it means that all the astrology books ever printed, all the magazines, and all the claims, arguments and speculations therein, to say nothing of astrological organisations, conferences, meetings, lectures and courses -- that all this could be completely without foundation. Do you really think this could be the case?

Researchers: We think such a view is simplistic. To start with, it ignores the distinction between subjective and objective astrology. If we prefer subjective astrology, then astrology does not need to be true, in the same way that a religion does not need to be true (although we may need to believe it is true). In this case the emerging picture is hardly relevant.

But if we prefer objective astrology, then you could be right. Of course, until many more astrologers investigate astrology with proper care, or put themselves forward for testing, we should not be too hasty. We must wait and see. Nevertheless the emerging picture does suggest that people can justifiably discount the objective claims of astrologers in the same way that they can justifiably discount the objective claims of phrenologists. Which of course is not exactly good news for the writers of astrology books.

OQQ12.5
How do astrologers generally react to the picture that is emerging from research?

Researchers: Most astrologers seem unaware of it. Those who are aware of it either (1) dismiss it on the grounds that scientific research is irrelevant to astrology, or (2) they admit its relevance but claim its methods are presently inadequate for unravelling astrology's secrets.

Re (1), the philosopher Thomas Kuhn noted that when an idea is in crisis, its supporters retreat behind a smokescreen of speculation that sounds good but is actually empty. This is precisely the situation with modern astrology. Rather than demonstrate their claims under artifact-free conditions, or specify what research would be relevant or how controversies and disagreements might be dealt with, astrologers retreat behind a smokescreen of speculation about the nature of truth, reality, perception, language, and so on. Talk yes, actual progress no.

Re (2), recall that the claims of astrology are grandiose, and that almost no area of human affairs (individual, collective, past, present, future) is supposed to be exempt. In other words we are supposed to believe simultaneously that astrology, like gravity, is writ most exceedingly large, while its influence is most exceedingly difficult to demonstrate. Scientists tend to part company with astrologers at this point. How can astrology be so difficult to demonstrate when astrologers are so readily convinced that it works?

EQQ12.6
One answer might be as given by Stephen Arroyo in his Chart Interpretation Handbook (CRCS 1989). He says "statistical studies in astrology have been almost universally pointless" because "only experiments with living people in a clinical situation can fully show astrology's value and validity in its guidance, counselling and psychotherapy applications." What do you think of this view?

Researchers: Arroyo does not give examples of such experiments for others to try out, nor does he show how they have resolved conflicting claims, so we have no reason to believe him. Arroyo seems unaware that phrenologists said the same thing about an actually invalid phrenology, and that "living people in a clinical situation" is precisely the situation where reasoning errors (Barnum, Dr Fox, hindsight, placebo, Polyanna, and so on) rage most out of control. In fact clinical studies of the kind he advocates have been made, but they have revealed nothing not explainable by reasoning errors and other artifacts.

Indeed, scattered throughout the astrological literature are accounts by astrologers who had accidentally used the wrong chart during a client consultation. One of us (Smit) had the same experience, and another of us (Dean) deliberately used wrong charts. According to Arroyo, because "living people in a clinical situation" fully demonstrate astrology's validity, the error should have been instantly apparent. In fact nobody noticed. In Smit's case he had always been told that charts uniquely fitted their owners, so he was profoundly shocked -- it showed that "astrology's validity" was effectively meaningless.

On this point, listen to what Donald Bradley said in a 1964 issue of American Astrology: "How many times have you worked with erroneous birth data and found admirably apt indications for everything that happened in the native's lifetime? We've all had this jarring experience ... Give me some false data and ... the chances are good that I'll be able to find a convincing configuration, progression, transit, key cycle, revolution, direction or dasa that is appropriate ... with multiple confirmation too, making everybody cluck about how marvelous astrology is. Too many times have we found that somebody was really born in 1923 and not 1924; or a rural doctor ... wrote pm instead of am on a birth certificate; or someone ... was still using an Old-Style birthdate; or a birth hour should have been recorded in daylight-saving time -- and so forth. But even though the information was seriously in error, the gears of the chartwork seemed to click off just fine. ... But is it science? That's the big question, and on this question hangs the whole disposition of astrology's worthwhileness."

Or as Rob Hand says in the Nov-Dec 1989 issue of the Astrological Journal, "I'm sure you've all experienced, those of you who do any number of consultations, the horrible and demoralising phenomenon of giving a brilliant reading from the wrong birth data! It's one of those little classic embarrassments we don't like to talk about. ... nevertheless, we have to agree that convincing readings of the wrong birth data are a real phenomenon." Or as Geoffrey Cornelius says in his book The Moment of Astrology 1994, "The entirely 'wrong' horoscope produced by misinformation or gross error not infrequently (but not always) works just as if it is a 'right' horoscope" (page 259).

NQQ12.7
Some astrologers claim that the lack of proof for astrological effects has shown only the ineffectiveness of the measuring tools. They claim that the tools used to test astrology are like trying to catch plankton with a shark net. Scientists draw in the net, but it reveals nothing. Which is proof that plankton don't exist, right?

Researchers: This is a classic example of woolly thinking. Scientists would conclude that objects above the mesh size did not exist, not that objects below the mesh size did not exist. Furthermore they would be careful to use the same nets as astrologers and in the same way. If the nets reveal nothing, how can astrologers claim the opposite? This is basically the bottom line. It bears thinking about.

13. White crows, prestige, resources, could results improve?

OQQ13.1
Do you see any ways in which the scientific approach might be modified to increase the possibility of detecting genuine astrological effects?

Researchers: Increasing the sensitivity is straightforward -- just increase the sample size. Alternatively we can increase the signal to be detected, say by testing only the best astrologers, or by selecting only extreme cases so they are unambiguous. But such approaches have already been well explored without success. In fact the variety and extent of studies to date is seldom appreciated, especially by astrologers who claim that the scientific approach is inadequate. To be sure, the original studies may be hard to find, but increasing numbers of reviews now exist. Ignorance is no longer an excuse.

OQQ13.2
How likely is it that the emerging picture will become more favourable to astrology?

Researchers: The totality of research results to date involves hundreds of diverse studies, some positive but mostly negative. Even the positive studies tend to be incommensurate with astrological claims, for example a positive study in which astrologers scored 55% hits vs 50% expected by chance is incommensurate with the near-100% expected on the basis of what we read in astrology books. To overturn such a weight of negative evidence would require an avalanche of new studies where the results were consistently and dramatically positive. If Figure 2 is anything to go by, this is not going to happen.

Nevertheless could astrology deliver the necessary goods? For twenty-five years we have tried to find out. We have ransacked the literature and tested the most promising claims, both as astrologers and scientists. We have gathered our own data and analysed it using the most powerful methods available. We have looked at related areas in astronomy, philosophy, psychology, parapsychology, sociology, and statistics, in each case going back to the original academic literature, something that astrologers rarely do. We have lectured at conferences, run critical debates, issued challenges, and held prize competitions (all now closed) with prizes up to $US5000 for evidence in support of astrological claims. As far as we know nobody else has been as systematic or as thorough. The point is, if astrology could deliver the necessary goods then somewhere in all this it should be shining through. But we found nothing that could not be explained by reasoning errors and other artifacts. Yes, a great way to spend twenty-five years!

OQQ13.3
Were you disappointed?

Researchers: To some extent. But had we been working for a research institute, we would have been more than disappointed, we would have been fired, or at least diverted to more productive areas. Negative results means no grant money, little contribution to knowledge, little chance of publication, and no academic advancement. So nobody could have wanted positive results more than us. Which is not to say that negative findings are unimportant, for example it is useful to know that eating lettuce does not send you mad. But to continue pursuing a topic with grave theoretical difficulties and grave lack of supporting evidence does seem rather futile.

But before we slash our wrists there is one beacon of hope. Many of these hundreds of negative studies would be instantly overturned if an astrologer could be found who delivered the goods under conditions where reasoning errors and other artifacts did not apply. In short, all it needs is one white crow. As the famous psychologist William James said in 1897, "If you wish to upset the law that all crows are black, you must not seek to show that no crows are; it is enough to prove one single crow to be white." Of course the research studies in astrology collectively cover a wider field than could be overturned by a single white crow, but it would be a good start.

For twenty-five years we have tried our best to find a white crow, as have those skeptic groups around the world who are currently offering a total of over one million dollars to anyone who can demonstrate genuine paranormal powers, astrology included, but without success. (For how to apply, visit the websites www.csicop.org and www.randi.org.) There are many tens of thousands of astrologer crows but few seem willing to be proven white -- and the few who have come forward have so far proven to be black. This might seem remarkable in view of the whiteness that shines out of astrology books, but such is the case. It might of course seem unremarkable to anyone familiar with phrenology.

Recently one of us (Smit) discovered a rare Dutch book describing a test of Leo Knegt (1882-1957), one of the Netherland's most famous astrologers. Knegt was given the anonymous birth data of ten people and had to describe their character and circumstances. Because he was given no details whatever, this was far harder than a matching test -- he had to describe each person from scratch. Unknown to Knegt each person had a distinctive life, e.g. a university professor, a fantastic swindler, and an unsuccessful job-seeker, so to achieve a hit he would have to be very specific. And he was, ten times out of ten. For the university professor Knegt correctly predicted an emotional but pioneering intellectual, for the swindler he correctly predicted fraud and an unexpected scary end (it was actually suicide), and for the unsuccessful job-seeker he correctly predicted she would eventually find a job on a passenger ship.

Not only was this the kind of specificity that many modern astrologers claim is impossible, it was achieved using only nine planets (the year was 1933 and Pluto had only just been discovered). Unfortunately the test had none of the features such as controls that we would consider essential today, so these results are difficult to assess. So we have only one option -- we must try and repeat the test.

OQQ13.4
Did you try to do this?

Researchers: Yes. Smit compiled a new test using five of the ten cases for which Knegt had been successful. This new test asked astrologers to match case histories to birth data, a much easier task than the one given to Knegt even without the fewer number of charts. Furthermore, because Knegt had been successful, it could not be argued that the new test was poorly designed -- the charts had clearly delivered the goods that astrologers want. If Knegt could succeed in such a difficult task, other astrologers should succeed in this much easier task using the same data. You could hardly get a fairer test!

OQQ13.5
Sounds reasonable enough. What happened?

Researchers: In order to reach as large an audience as possible, the entire membership and council of the Astrological Association (over 1500 astrologers and students) were canvassed through the AA's newsletter Transit. Only two astrologers responded, so readers of the smaller but more active international monthly Astrologers' Forum were canvassed, which increased the total to 23. Their performance was poor (this is the result labelled Smit in Figure 2) and their agreement was even slightly worse than expected by chance. Astrologers will have to do better than this if they are to stop high-profile skeptics concluding publicly that white crows do not exist, and that astrology has nothing to contribute to science or philosophy. So, to answer your original question, is it likely that the emerging picture will change? Don't ask us, ask astrologers about white crows. The ball is firmly in their court.

EQQ13.6
The conclusion that astrology has nothing to contribute to science or philosophy would no doubt be disputed by astrologers.

Researchers: The problem here is the spectacular never-ending disagreement among astrologers on techniques, validity, and how astrology works. This enduring disagreement tells us that astrology has nothing to contribute to science and philosophy except enduring incoherence. Indeed, the most plausible reason for this enduring disagreement is that there is no real phenomena to be explained, as would be the case if astrology was the result of reasoning errors. Such an astrology could never contribute to science and philosophy except as an example of how not to do it. And this seems to be precisely its present status:

Go to any university library or large public library and look at modern works of science or philosophy. They almost never mention astrology. This is true even in the two fields closest to astrology, namely psychology and astronomy. Of the two dozen histories of psychology published since 1970, only two mention astrology, and even then only as an example of pseudoscience. No modern astronomical theory has been even slightly influenced by astrology. Astrologers do not even pose meaningful questions for astronomers to investigate. Evidently astrology, despite having ridden for eighteen centuries into the hearts and minds of learned people on the unity of Greek ideas, is seen as having made no contributions to psychology or astronomy worth mentioning.

The same is true of philosophy. Astrology either receives no mention, as in The Oxford Dictionary of Philosophy (OUP 1994), or is mentioned as an example of superstition, as in Bertrand Russell's classic History of Western Philosophy (Allen & Unwin 1961), where in the Renaissance "the first effect of emancipation from the Church was not to make men think rationally, but to open their minds to every sort of antique nonsense [like astrology]" (p.489). Evidently in philosophy, as in psychology and astronomy, astrology is considered worthless by people looking for productive ideas. The reason is simple.

EQQ13.7
Prejudice?

Researchers: No, track record. Astrology has been spectacularly unfruitful in guiding enquiry, mainly because it makes much less sense than existing theories. Which is why scientists and philosophers ignore astrology except for historical purposes. Nevertheless if astrology explained things for which there was no other conceivable explanation, or if it led to useful new discoveries such as new energies or channels of communication, they would immediately become interested.

Similarly, if cosmic connections with human mental processes were ever discovered, we might find astrology being treated in textbooks as an ancient belief with some kernel of truth. But for the moment there is not even that.

NQQ13.8
The popularity of astrology shows that many people see it as interesting, even uplifting, and worth spending time on. So how can it not be a productive field of study?)

Researchers: Astrology is certainly productive in the sense you mention. For that matter so is flower arranging, guitar playing, and pottery. But we meant productive in the sense used by scientists and philosophers.

A good example is Darwin's theory of evolution, which has been quite spectacularly productive. It not only provides a scheme for unifying the diversity of life, it also raises clear questions for research to address. How do new characteristics arise in populations? What are the mechanisms of inheritance? What criteria decide when a characteristic confers advantage? These and many other questions make Darwin's theory vulnerable from many directions, for example the mechanisms of heredity might have made it impossible for advantages to arise and spread. Yet attacks by armies of biologists, geneticists, morphologists, physiologists, ecologists and others have revealed the need for only minor departures from Darwin's original theory.

By contrast, astrology has been spectacularly unproductive in guiding inquiry. This does not deny that astrology may be productive in other ways, but these other ways are of little interest to scientists and philosophers.

NQQ13.9
Perhaps the problem here is a lack of resources. Most serious astrologers are simply too busy making a living to devote time to research, or to gaining the interest of scientists or philosophers.

Researchers: But astrologers have had more than two thousand years to prove their value. How much longer do they want? Furthermore, as we already noted (13.1), there now exists a large amount of research conducted by astrologers, sympathetic scientists, and critics. If astrology really was productive it would be more than apparent by now.

NQQ13.10
Recently in both the UK and USA there have been discussions about the possibility of introducing astrology into university courses. Despite what you say, this seems to indicate that astrology now has some academic respectability, or even prestige.

Researchers: This "academic respectability" might be misleading. First, undergraduate courses in astrology would most likely be very profitable. It is easy to miss the financial motive and see the outcome as an academic seal of approval. Second, the blessings might be decidedly mixed. Unlike ordinary schools, universities are at the top of the intellectual ladder. They set the standards. Standards mean rigour, and rigour means requiring students to be critical. Liabilities now hidden would be minutely explored on a scale never experienced before. Astrologers would be in a court ruled by empirical evidence and critical thinking skills, i.e. precisely those areas where they are most disadvantaged. Academic staff would not only have to respond publicly to criticism, they would also have to propose a research agenda, supervise research students, and publish positive test results in high-quality journals (weak or negative results might be fatal). A teaching position could be the hottest of seats. Who would be brave enough to accept?

That said, courses could still escape the requirements of critical thinking and empirical testing if they focussed on history, or religion, or philosophy, or counselling, or any other area of astrology that does not require it to be true. Interestingly the courses you refer to seem to do just that, so it would be quite wrong to suppose that astrology is now about to acquire "academic respectability."

As for "prestige", there is a problem here because real prestige is deserved. It is the result of disciplined work, of critical thought, of willingness to put beliefs to the test, and of following where the chips fall, none of which is presently true of astrology. Prestige cannot be created out of nothing. Astrology cannot aspire to prestige without striving for the virtues on which it is based.

NQQ13.11
Whilst I wouldn't want to claim that all astrologers possess the virtues you mention, I certainly believe that I have met a number who do. So what you say on this point strikes me as one-sided.

Researchers: Some astrologers do have those virtues in a general sense, for example astrologers such as John Addey were extremely hard working and certainly willing to put beliefs to the test. But what is missing is a good grasp of what scientific discipline demands. Being hard working and being scientific are not necessarily the same thing, and being the former does not make up for not being the latter. In our experience very few astrologers have those virtues in a scientific sense, which is what matters. After all, we can have all the right virtues for the history and philosophy of phrenology, but this does not make the practice of phrenology respectable and prestigious. The same applies here.

NQQ13.12
In general terms, many astrologers might see what you have been saying as less a statement of impartial evidence and more a statement of evidence for the prosecution. Of course I accept that you may sometimes choose provocative language in order to get a response. Nevertheless the one-sidedness seems out of place at times.

Researchers: Our focus is scientific research, and our standards are scientific ones. Our aim is not to defeat astrologers or their system but merely to point out the need to be careful and to show what happens when this need is neglected. We agree that our comments may seem somewhat provocative. They might even seem prosecutory to those inclined to feel guilt. But given the huge neglect by astrologers of scientific caution, and the disastrous effect of this on their credibility, we think our comments are justified. In fact many scientists might find our attitude towards astrology to be overly charitable. They would note how the grandiose claims of astrology far outweigh the flimsy observations, ill-considered speculations and armchair proclamations offered in lieu of evidence, and how the general ignorance of reasoning errors can explain everything. So they would see no reason to elevate astrology beyond pseudoscientific piffle. If you find us prosecutory, you should try the others!

14. Parallels between astrology and poetry

OQQ14.1
Perhaps we might look again at what you have defined as "subjective astrology." Could you give an example of research into this?

Researchers: In subjective astrology we are concerned with astrology as the language of images, metaphors and similes. The art form closest to astrology is poetry because both rely on words whereas music and pictures do not. Indeed, some astrologers hold that astrology is a form of poetry in disguise. So our example involves poetry.

We can describe the best poetry quite simply. It is beautiful, it conveys the poet's passion and feeling, and it reveals inner meaning. The same is true of the best astrology. It is beautiful, it makes us care, and it reveals inner meaning. But note how subjective everything is. Who is to say what is beautiful, worthy of passion, and meaningful? But this is art, not science. Here we are concerned only with feeling. What matters is that our hearts are moved. Now the logical next step:

Could an understanding of poetry improve the practice of astrology? If poetry moves hearts better than prose does, could we improve a chart interpretation by writing it as a poem? These are the questions some of us looked at. The results were rather fruitful. For astrologers they suggested three useful things: (1) If your technique feels right, be it ever so humble, then for you it is right. (2) By the same token, it will not feel right for clients unless they share your particular sensitivities and understanding. (3) Good poetry takes time. Before making an interpretation you should allow the chart to incubate in your unconscious, allowing its factors to translate spontaneously into poetry.

NQQ14.2
Could you give more details? Your interest in poetry certainly throws a new light on your researches.

Researchers: To start with, we compared various poetic descriptions with the keywords pioneered by the famous British astrologer Margaret Hone, as in Table 1.

Table 1. Astrological keywords vs poetic descriptions.

                      Poetic descriptions with apologies to
Hone keywords         Byron, Tennyson, Shelley, Wordsworth et al
---------------------------------------------------------------------
SO  Power, vitality   The sun, bright eye of your universe.
MO  Response,         A faded Moon in dim, silver twilight.
    fluctuation       Azure Moon, rippling in verdant skies.
ME  Communication     Nor all your tears wash out a word of it.
MA  Energy, heat      Mars, hectic red and pestilence stricken.
                      The Martian came down like a wolf on the fold.
JU  Expansion         In fair round belly with good capon lined.
NE  Nebulousness      Neptune, brackish with the salt of human tears.
                      The still, sad music of humanity.
8th solar house       Setting sail beyond the sunset.
Conjunct Venus        That deep in sweet embraces dwell.
Unaspected Moon       Alone and palely loitering.

For the present purpose (i.e. stirring the emotions) the accuracy of these poetic descriptions is of little consequence. What matters is what our imagination can do with them. They do not have to make sense to make sense. Which is bad news for hard-nosed objective researchers, but good news if you are into writing subjective astrology books. Any use of poetic descriptions would of course be as an addition to existing keywords, not as a replacement. Thus to describe the Sun as "bright eye of your universe" will be ambiguous without the context of power and vitality. Otherwise it could mean the telly.

NQQ14.3
How could this provide useful insights into the consultation?

Researchers: When groups of people are asked to rank various poems in order of appeal, it is usual to find that the same poem can be ranked top by some and bottom by others. This is not something peculiar to poetry -- it is true also of pictures and music. In other words there are marked individual differences in taste, which means you cannot possibly please all the people all the time.

Research by others has found that the main factors influencing the appeal of a poem are as follows: (1) Its merit. Good poems are more popular than bad poems, for example a poem may scan so badly that nobody likes it. Of course merit may be strongly decided by culture, in the same way as we may prefer Gershwin to a Zulu war chant, or vice versa if we are Zulu warriors. (2) IQ, especially verbal IQ. Poetry is more popular with bright people. (3) Personality. Particular personalities tend to prefer particular types of poem as shown in Table 2, which to be reliable would need to be much longer but it gives you the idea.

Table 2. The types of poem preferred by particular personalities.

Extraverts            Into the silence of an empty night I went,
prefer simplicity:    And took my scorned heart with me.

Introverts            Thou art not more lovelier than lilacs
prefer complexity:    Nor more fair than single white poppies.

Unstables             Come, let us make love deathless, thou and I,
prefer emotion:       Seeing that our footing on the Earth is brief.

Stables               Here lies a most beautiful lady,
prefer restraint:     Light of step and heart was she.

In our experience much the same applies to the appeal of astrology. For example, other things being equal (which they rarely are), we have found that stable extraverts tend to prefer the direct simplicity of Margaret Hone whereas unstable introverts tend to prefer the obscure complexity of Dane Rudhyar. So when astrologers promote this or that amazing new method, they have to take into account the personality of the astrologer or client. Otherwise they will be performing a mischief in areas where astrologers, if anyone, should know better. To feel the poetry in a chart the astrologer should not be hindered by technique.

OQQ14.4
What of those clients who come to an astrologer seeking straightforward answers to questions on health, wealth and romance? How would your poetry findings apply to them?

Researchers: Probably not at all, because this would no longer be subjective astrology. It would no doubt be okay if clients wanted only sympathy or spiritual uplift, but if they also wanted something factual then we are back to objective astrology and the dismal emerging picture. Poetry might only annoy them.

Actually your question uncovers a fundamental dilemma. Astrology seems unlikely to feel right unless astrologers and clients share a belief in objective astrology. Otherwise why bother with accurate charts? The same dilemma applied in phrenology. Phrenology worked because people believed (incorrectly) that it worked. If phrenologists had said (correctly) that "phrenology uses interesting superstitions to stimulate self examination", or that "phrenology is basically tea and sympathy", would clients have bothered?

OQQ14.5
How would you resolve this dilemma?

Researchers: The dilemma is without doubt an agonising one. But it is one that only astrologers, not researchers, can resolve.

Click here for rest of Part 2 (takes one download plus one more for the index)

From www.astrology-and-science.com         Click here to return to home page