Exegesis V4 Issue 25

Exegesis Volume 4 Issue #25

From: Andre D
Subject: Re: Exegesis Digest V4 #22

Exegesis Digest Sat, 27 Mar 1999

Date: Sat, 27 Mar 1999 23:53:46 +1200
From: Andre D
To: Exegesis
Subject: Re: Exegesis Digest V4 #22

Time has been and is against me, as I go into pre-training as a post-hole digger over the next eight months. Thus I have had to ignore your reply to my last post, even though it contained important points. Nevertheless, one last contribution.

It seems commonplace in forums such as this to insist on the need to reach agreed definitions. As there has been little dissent to date, I take it that these have been reached. So as an overall suggestion I think that the step we have to undertake now is to convert the "big" questions that William has posed with the help of critical comment from Rog and others into a series of smaller, more manageable questions. (I fancy Roger will not agree with such an apparent reductionism, but *if* this process occurs I also think that Roger's or anyone else's dissenting voice will be a vital part of that process).

> >[Rog] raises an interesting question: does intuition as we apply it to
> >astrological practice rely solely on our own experience...
> >or can it also operate on data from our immediate sense of the
> >astrological mechanism?

Well (one possible way) if experience is one of the factors in intuition, then we will find that astrologers get steadily better as we age. Do we have evidence that this is the case? And on the other hand, if an ability to _sense_ the astrological mechanism is part of it, then there will be some astrologers who have an outstanding "knack" right from the start of their training - but a knack that may well fluctuate, whereas that based on experience will tend to show little fluctuation.

> >Concerning the idea that astrology can only produce results approximating
> >chance: I know of no conclusive or compelling basis for this assertion.

Agreed. I take it you are referring to statistical studies, especially those conducted by "objective" researchers. In fact, if chance is the basis, then at least *some* of those studies should have found significant results, simply by chance!

> >My approach is straightforward: I recognize that astrology
> >historically purports to be able to make prediction possible < snip >
> >If there exists an astrological mechanism, *and* the construct (astrology)
> >that has been developed to interpret the results of said mechanism is valid,
> >then I do not see why prediction is impossible, at least at the gross level
> >of perceived physicality. I am assuming that this gross level is a
> >statistical result of particle (quantum?) physicality, and so I can say
> >nothing about what the astrological mechanism might or might not be there.

I think there is a second problem here, which I address in another context below, which is "what is the size of the predicted effect?". Those in the forum familiar with statistics will know the concept of "variance". It is commonplace in psychology, for example, to be satisfied with mechanisms and causal factors that only have modest or tiny effects. Personality variables for example (admittedly at the lower end of the scale), typically have validities of only 0.2-0.3 (on an absolute scale of 0 to 1), meaning they only explain 4-9% of the total variation in the behaviour that was measured. I am unaware of *any* non-trivial claims in psychology that approach 1.0.

Similarly, I suspect that the effect sizes of the astrological mechanism are similar to or less than those commonly treated in psychology. Considering that there seems to be some overlap between astrological predictions about personality and predictions based on personality tests, one might argue that astrology lies roughly in the same predictive area.

> >Any investigation of astrology *must* be founded on the celestial phenomena;
> >otherwise it is not astrology that is investigated. The celestial data
> >*must* be developed and organized at the beginning of any research project.
> >It is bootless to take a history and then generate the astrology, which is
> >what I suspect is the usual informal protocol.

Agreed. Moreover, the social sciences are well supplied with methodologies and protocols, very hard won, that can serve our purposes, with intelligent adaptation required in some cases. It is witless for (poorly resourced) astrologers to attempt to reinvent the wheel on the one hand, or to continue using invalid research methods on the other.

> >The current practice of "psychological astrology" seems to be quite
> >successful, in general. < snip >
> >The most useful interpretation, in my view, is that which identifies and
> >points out how the client is unique. < snip >

Well said. This is exactly my own view, with qualifications (by definition, the unique cannot be described. Rather, unique combinations of *common* social experience can be, after a fashion, but unique combinations of unique elements simply cannot be described or talked about. This is where I find Roger's views definitely congenial with my own).

I think it needs to be added that sociology, psychology and so on point out how the client is *not* unique. This is valid because it *does* seem that people behave in ways that are common with others, as well as different from others. We share socially constructed realities, as well as organic commonalities. Thus, the social sciences have predictive validity. However, the contention we have to propose and defend as astrologers is that the social scientific account of the person is incomplete: astrology supplies part or all of the missing picture. This also operates in reverse: astrology is not a complete account of the person either.

> >I would hope that all readers of this post regard my views as taken for
> >granted in professional practice, as perhaps even overly simplistic.

I hope so too. Yet many astrologers in my experience seem unaware that detecting and describing uniqueness is one of the primary, distinguishing features of astrology; and that if this is accepted as true, it places limitations upon the generalisability or utility of astrological knowledge (in the manner I just described above). Shareable knowledge (shared in textbooks or online discussions about astrological meanings, or shared with the client etc.) must be social; it cannot be individual in any ideal sense.

This does point to questions about the nature of astrological practice with the client, and how this must differ from the practice of (say) a clinical psychologist who is, happily, in a far less complex position regarding the application of her/his knowledge.

> >Again, I return to the original question in this series of posts: does the
> >astrological mechanism require the presence of life (sentience?)? That
> >would seem to me to be a good place to start in a search for some insight
> >into the nature of the astrological mechanism. < snip >

Good, fundamental question, but *big*. Where on earth do we start to answer this? I think suggestions on how to tackle this are needed - although I take the next point you made, regarding financial fluctuations, and assertions of Mary's, as relevant.

> >Robert Hand has made the assertion that some form of astrology is effective
> >< snip > success that is done in a scholarly manner would be welcome.

Actually, purely as a *starting* point, one might proceed by assuming that the various astrological claims *are* valid; or one might rather, similarly to the way that "weight" is used in the scientific literature, only accept those claims as tentatively valid that have the most support. For example, a critical part of my current work depends on a literature connected with the social influence of minorities; and whether minorities (to be successful) employ distinctively different persuasive strategies from majorities. Unfortunately the literature, so far as I can see, is deadlocked pro and con. Science does not differ from astrology in having undecided or disputed points.

At any rate, one can then start making deductions from that point. Inconsistencies should soon become apparent, and then quite well focused hypotheses become possible. IMHO. (Incidentally, the argument about cycles and time that I advanced a few posts ago is one that is based on interpreting certain astrological claims, and I believe it does generate certain hypotheses that *are* testable: if appropriate methodologies are employed).

But the best approach would seem to be to proceed empirically, as Dale has advocated. Whilst I agree wholeheatedly with Dale, I also imagine that this is not to the taste of most. Better to stick with the square wheel that rolls slowly and with difficulty, then to throw it away and begin again from first principles, eh? ================================== William, Rog and others: I have greatly enjoyed these discussions. I intend to remain a subscriber, and will watch developments with interest. Perchance I will find time to contribute occasionally too.

Francis, I have printed off the (lengthy) paper you referred to. My interest would be to attempt to evaluate the soundness of the statistical approach, which would take more time than I have at present. Perhaps someone else can oblige? My first impressions though: whilst I agree with Lehman and Brady that the key question is "whether the models will continue to hold up over time", I am less happy with the interpretation of the p values found (they are employing the now disputed null-hypothesis testing strategy in reporting that "any results of p < .05 could be considered significant", about which I wrote in my first post to exegesis) and their use of the entire "population" of available tests (p-significance is used with samples, not populations). Their finding could be spurious, something that could have been demonstrated by drawing random samples of tests and than "cross-validating" against other samples. Future samples will constitute just such a cross-validating sample - so long as it is kept separate and not combined with the existing "population" - but we are in for a considerable wait, in terms of the Ashes anyway! Also, having pointed to p < .05 as a possible significance level, they did not directly discuss those (several) results that failed to meet the level.