Intra-Chart Validation, A Method Within Zodiacal Releasing

Astrological Research: Some Reservations about Statistics &

Some Suggestions for a New Program

Robert H. Schmidt

The Current Crisis in the Field of Statistics

Developments in the theory of statistics over the past half century call into question the appropriateness, in many cases, of the statistical approach that was the norm throughout most of the 20th century and continues to be taught to this day, particularly in the “soft” sciences. Often called the “frequentist” approach, it is by far the one most commonly used by astrologers when doing their research. As a matter of fact, it appears that many astrologers are not even aware that this approach has come under close scrutiny in recent years and that there are other statistical paradigms currently under serious consideration. It is above all the revival of what are called “Bayesian” methods that has brought the statistics community to a veritable state of crisis.

Let me quote from a recent paper in The American Statistician, August 2006, Vol. 60, No. 3. In his paper Calibrated Bayes: A Bayes/Frequentist Roadmap (based on an important invitational address at the 2005 Joint Statistics Meeting ), the author Roderick J. Little writes:

"The lack of an agreed inferential basis for statistics makes life “interesting” for academic statisticians, but at the price of negative implications for the status of statistics in industry, science, and government. The practice of our discipline will mature only when we can come to a basic agreement about how to apply statistics to real problems. Simple and more general illustrations are given of the negative consequences of the existing schism between frequentists and Bayesians. (p. 213)

Currently it seems to me that the pragmatist approach to inference predominates. Pragmatists might argue that good statisticians can get sensible answers under Bayes or frequentist paradigms; indeed maybe two philosophies are better than one, since they provide more tools for the statisticians toolkit….Since the Bayesian and frequentist philosophies can differ even on simple problems, at some point decisions seem needed as to which is right. I believe our credibility as statisticians is undermined when we cannot agree on the fundamentals of our subject as Efron (2005) noted:

The physicists I talked to were really bothered over the 250-year old Bayesian-frequentist argument. Basically, there’s only one way of doing physics, but there seems to be at least two ways of doing statistics, and they don’t always give the same answers.

A prominent Bayesian (Berger 2000) writes in a similar vein:

Note that I am not arguing for an eclectic attitude toward statistics here; indeed, I think that the general refusal in our field to strive for a unified perspective has been the single biggest impediment to its advancement. (p. 214)"

Does this sorry state of affairs sound familiar to astrologers?

This is not the place to go into the terms of this controversy in any detailed way. Suffice it to say for the purpose of my exposition here that the Bayesian methods take into account and figure in any prior knowledge that one may have about the issue under investigation, whereas the frequentist approach does not.

My reason for quoting from the article above is to make a simple point: Astrologers have been trying to gain recognition for their own discipline by applying the mathematical tools of a discipline (statistics) that is itself in crisis and consequently worried about its own status and future.

Never mind the irony of this situation. The fact is that there is precious little statistical evidence for the validity of astrology at present, despite the continued efforts of a considerable number of dedicated researchers using the frequentist approach. And given the state of affairs in the statistics community, it is far from clear that even that “precious little” will ultimately survive more sophisticated scrutiny if and when the current controversy is resolved and a sounder statistical science emerges.  

Now, I do not wish to discourage anyone from mastering the usual frequentist methods and learning how to apply them to problems of astrological research. As a matter of fact, the same Bayesian (Berger) cited above by Little went on to say that “any unification that will be achieved will almost certainly have frequentist components to it.” The more modern astrologers understand this approach, the better prepared they will be if the statisticians ever get their own act together, especially if astrologers undertake to study the Bayesian approach. (Astrologers should also take a look at some other approaches as well, such as entropy maximization, which was in fact used by the late astrologer Theodor Landscheidt in some of his work.) But I do wish to alert the community to the strong probability that much of what has been done up to this point in the area of astrological research may not hold up.

Hellenistic Astrology Provides the Opportunity for a New Approach to Research: Intra-Chart Validation

The research my wife and I have been doing in testing out the techniques of Hellenistic astrology has led me to question whether the statistical approach to astrological validation in the usual sense is the most appropriate one. At the same time it has suggested a new approach to the validation of astrology, which for the moment I am designating by the terms “intra-chart” and “inter-chart” validation.

A strange Hellenistic timing procedure (reported in Book IV of the Anthology of Vettius Valens), which we call “zodiacal releasing,” was the occasion for investigating this new approach. Zodiacal releasing is one of the central Hellenistic “time-lord” procedures. These are the Hellenistic analogues of the Indian dasa systems, although the Hellenistic procedures seem to work best in the tropical zodiac. Instead of concentrating on isolated moments in a life, time-lord procedures study spans of time that are assigned to the rulership of the planets (and sometimes the signs) according to some orderly schedule.

One application of zodiacal releasing is in the study of a person’s career or working life. With this time-lord procedure a person’s entire working life can be divided into a series of distinct chapters, and each of these chapters can be further subdivided into a series of distinct episodes. Valens provides us with a sketchy set of interdependent interpretive principles—it took a fair amount of work to make them fully serviceable—that allows us to make statements about these chapters and episodes, such as whether a given period should be successful for the native, whether the native is likely to change jobs during a given period, whether he or she should attain public recognition during a given period, etc. It is the fact that this same procedure with its attendant interpretive principles can be applied in sequence to all the numerous distinct chapters and episodes in some person’s life that opens up the possibility of validating it to some measure within the context of a single chart—hence the term “intra-chart” validation.

The first thing I learned from the study of time-lords is that there is a different way of looking at events. It is perhaps part of the insidious influence of modern physics that modern astrologers tend to view all events as occurrences at some instant or moment of time—like a car accident, the launching of a space-ship, the christening of a child. This is also why they rely almost exclusively on transits, progressions, and directions for timing. The thinking is that a human event should be more or less synchronous or in one-to-one correspondence with the encounter of a transiting, progressed, or directed planet or other sensitive point with some natal position, either by conjunction or by aspect.

The Hellenistic approach to timing by means of time-lords is very different. It understands events to occupy a span of time and have duration. For the Hellenistic astrologer, the event of marriage, for instance, is far more than simply saying “I do.” It involves the courtship, setting the date, squabbles between the soon-to-be in-laws, the best man being delayed or losing the ring, the actual marriage ceremony, the banquet, subsequent consummation and honeymoon and so forth.

Now, through the study of zodiacal releasing we repeatedly saw that the different moments in the full description of an event (as explained above) were meaningfully grouped together under the different or expected periods of time-lordship, even though there was not necessarily any synchronous astronomical event at all. This struck us as far more compelling evidence for the validity of this procedure than, say, several one-to-one correspondences between particular moments in these extended events and transits would be evidence for the validity of transits.

By the way, Hellenistic astrologers did use transits, progressions, and directions (in fact they pioneered them), although almost always in the context of time-lords. To the extent that this is the correct way of dealing with such “notoriously inaccurate” dynamic techniques as transits, it opens up the possibility of validating them as well within the framework of our approach.

Again, in the natal charts we studied in considerable detail from the perspective of zodiacal releasing, we found that the procedure was extremely successful in blocking out the entirety of the life into meaningful chapters and episodes. This was very compelling. Of course there is the risk of some subjectivity in all this. But to our astonishment we found that, to a great degree, biographers themselves blocked out the life in nearly the same fashion, and often used phrases to characterize these periods that were perfectly consistent with the significations of the planetary time-lords for these periods. Thus, by keeping an eye on what the biographers themselves do, we were able to minimize the risk of subjectivity significantly.

For a while I was bothered by the problem of how one could practically model such a subtle and complex astrological scenario, given the fact that one would have to quantify biographical data in such a way as to make it amenable to any kind of statistical treatment and validation. But then I realized that no amount of statistical analysis would render the results we had any more compelling than they already were. Statistics could only confirm what common sense and natural intuition was already telling us: that the correlations between the biography and the time-lord periods could not have happened accidentally throughout the course of an entire life. A correlation between one episode in the native’s life and the concurrent time-lord description might have been accidental, but how could that be the case when the repeated application of the same technique to the successive episodes in the entirety of a given life continued to work?

I summed up this experience to myself in the following way: Isn’t seeing a single technique with several interdependent interpretive principles work well over the entire course of a single person’s life is more evidential than seeing the procedure account well for a few select periods in a number of different lives? Or isn’t seeing all of Valen’s interdependent interpretive principles for this procedure working well together in a single chart more evidential than seeing one of them work in some number of charts? And by extension: Wouldn’t it be a far more compelling validation of astrology in general and Hellenistic astrology in particular to see the whole machinery of Hellenistic astrology with its numerous interdependent techniques, methods, and procedures working in a single chart than seeing some of these work piecemeal to a comparable degree in some number of different charts?

It now seems to me that the use of statistics to discover astrological correlations in the manner of Gauquelin, or even its use in the attempted verification of some astrological hypothesis about the meaning of certain chart factors by examining the lives of persons who have those signatures in the own charts, may be fruitless. What is needed is evidence of the workings of astrology that is strongly compelling from the outset. 

In order to recognize such evidence when it comes their way, I believe that astrologers need to cultivate their natural human sense of what is plausible and implausible, what is evidential and what is not, in the context of astrological validation. Number crunching does not help develop this sense; if anything, in distancing us from what we see in individual natal charts, it has a tendency to dull this sense. Far better would be to find something compelling and immediately evidential in a single chart and then find appropriate ways of further testing this primary evidence. And at the moment I believe that Hellenistic astrology is the best candidate for providing such evidence because of its value in addressing the totality of a life.

However, in lieu of the application of statistical reasoning, what can we reasonably do to test compelling data from a single chart on the off-chance that we have been deluding ourselves? What kind of intra-chart testing could we perform? One way is to generate hypotheses that serve to test the data we have in a single chart against itself. This is a device that has been employed even in regular statistical investigations to give the data more credibility. A good example of this is what Ertel did with the original Gauquelin findings on the “Mars effect.” Since the Gauquelin data purported to show that eminent athletes tended to have Mars in certain sectors of the natal chart, Ertel reasoned that the more eminent the athlete, the more likely that he would have Mars in one of these sectors. This hypothesis was not part of Gauquelin’s original analysis, so when it turned out to be true, it added more credibility to Gauqeulin’s results.

We have made use of similar devices in our testing of the correlations within a given chart. For instance, since certain configurations of planets and signs can recur during the course of a life in the application of zodiacal releasing, it stands to reason that when these constellations of astrological factors repeat themselves identically or partially, we should expect some biographical parallels in the life. We have in fact found this to be the case, and it has given us all the more confidence in our approach.

Once we have found compelling evidence in one chart, we can look to another chart that has certain formal similarities to it from the perspective of the time-lord procedure we are testing. If an analysis of this second chart separately turns out to be equally compelling, then we can compare these two charts with each other to learn whether there is some formal similarity in how these two individuals lived out their lives, as the logic of the technique would suggest. To the extent that this inter-chart analysis pans out, we have all the more reason to trust the technique. Here again we have had striking results by comparing the charts of figures as diverse as Einstein and Hitler. Zodiacal releasing makes quite satisfying sense of their lives and careers individually, but quite striking parallels also emerge when we compare these two charts and lives with each other, as we would have expected from certain chart factors these two individuals have in common.

A final observation. When we have done our studies of zodiacal releasing and other Hellenistic time-lord procedures, we have always tried to look at the events in the native’s life as if they could have been predicted using this procedure, even though we were reading the charts in retrospect. The more we learned about the details of the native’s life, the more we felt that we could have made fairly accurate forecasts about what would happen to these natives in the future. In other words, the more we were able to take prior knowledge into account the better we felt we would have been able to predict. Now this should seem obvious to most astrologers, but this simple feature is missing in most astrological research. However, taking prior knowledge into account is a hallmark of the Bayesian approach mentioned above.

We have all learned that the more we know about a client or research subject, the more we can narrow the possibilities of what is likely to happen with that person in the future. As a result, our astrological forecasting becomes more accurate. There is nothing wrong in doing this and it is not cheating as long as we are not giving back to the client information that we have already extracted from them. There is a comparison I like to use here. The differential equations used in physics to describe the behavior of some system are useless until certain “boundary conditions” have been established. Without determining these boundary conditions, the differential equations are too general. By learning more about the client’s life we are simply determining these “boundary conditions,” so that our interpretation of astrological factors in a chart become more specific to that individual. From this point of view, blind readings or tests are more or less meaningless in the validation or invalidation of astrology. What we should be testing is how increasingly accurate are the astrological predictions given increasing levels of prior knowledge about the life.

In brief, this is the program we have set out for ourselves. My wife Ellen Black will be exemplifying some features of this program in her presentation of some episodes in the career of John Kerry.