Response to James Brockbank’s
“The sceptical attack of Dean et al. on astrology”

by Arthur Mather

In his article, Brockbank outlines what he sees as the five main sceptical arguments, which he considers to be “attacks”. (He uses the Latin-based sceptical, presumably in its philosophical sense of doubting everything, whereas skeptics themselves use the Greek-based skeptical to emphasise its etymological sense of requiring evidence.) He lists the different types of astrology and then attempts to show that most are not amenable to conventional testing. He further holds that existing empirical tests are invalid because they have tested textbook astrology, not astrology as practised by astrologers.

Unfortunately there are serious problems with Brockbank’s approach that render his conclusions invalid. His five “sceptical attacks” are summarised below together with my response.

1. There is nothing to explain how astrology works.
Brockbank - The sceptical position specifies that there can only be Theories of Accuracy or Theories of Satisfaction. This is an attempt to limit astrology to what can be empirically tested (via Accuracy), or to the dustbin category of Satisfaction. The former is not applicable to most of astrology, and the latter (by implication) is inadequate. There could be non-scientific theories which would be much more appropriate to these other areas.

Mather - The problem lies in the claims made by astrologers; a claim that is useful, unlike “X may or may not indicate Y” which is not useful, is always testable. It is only where the astrological process can be beneficial regardless of its accuracy, as when it encourages discussion or reflection, that claims to accuracy could be dispensed with -- and which would then bring us into the realm of Satisfaction. In fact psychology has ample resources for addressing the diversity of topics possible under Satisfaction, but their application to astrology seems premature when more fundamental issues (such as conceptual problems and reasoning errors) are still ignored by astrologers.

2. There is no consensus on basic issues.
Brockbank - The lack of consensus is not a problem; different roads could lead to the same conclusion.

Mather - It would only be possible for widely differing systems to reach the same conclusion through the guiding agency of the astrologer, which would deprive astrology of any intrinsic validity. Existing tests have shown no useful agreement between different astrologers on what a given birth chart means.

3. Astrological symbolism is unsystematic ... anything goes.
Brockbank - There is a basis for reasoning in astrology that can provide as specific, or general, an interpretation as is required.

Mather - There is an underlying structure (or grammar) to astrology which should impart order to an otherwise chaotic situation; this would depend on the skill of the astrologer, but as just noted the required order is not being achieved despite 2000 years of practice. Of course, a particular reasoning can be specific, but this does not necessarily make it right.

4. It is an impossibly big task to interpret a birth chart.
Brockbank - The size of the task is not a problem because we can look at manageable sub-sets, which limits the number of (random) possibilities.

Mather - This point is reasonable, but does not in itself indicate any validity in astrology. It does not explain why, contrary to what every astrology book confidently implies, astrologers routinely fail under blind conditions to usefully match charts with their owners.

5. There is no empirical evidence to support astrological claims.
Brockbank - The types of astrology amenable to testing are limited to one or two, which contests the general relevance of Accuracy as a yardstick. Empirical studies of astrology, being without the input of an astrologer, are invalid as they do not reflect practice. But testing astrology as practised may not be feasible.

Mather - This fails to recognise that all claims are amenable to testing in their own terms; the area of astrology is irrelevant. Accuracy is an entirely appropriate measure for specific objective claims to which tests have been matched. Astrologers themselves have in fact shown it is perfectly feasible to test astrology as practised; however the results have been manifestly incommensurate with astrological claims. As noted, Brockbank is also dismissive of Satisfaction as an alternative measure, but he nevertheless accepts this by implication when he says later that Sun sign columns are beneficial if people find them worthwhile.

Brockbank - Chides astrologers for making unjustified and excessive claims, and sceptics for treating astrology too uniformly and not as practised.

Mather - Unfortunately both textbook astrology and astrological practice have been rigorously tested without any evidence that the outcomes are due to other than non-astrological factors.

In summary, Brockbank’s review is a significant, if flawed, step towards grasping the nettles revealed by research. A more rigorous and informed approach, with due attention to actual research findings in place of unsupported claims, could give the debate a most welcome boost. I look forward to future developments in this direction.


[Brockbank has submitted various point to Mather
for his consideration; hopefully a dialogue will ensue
and I will be able to publish the fruits of it here -
so watch this space - GP]