Astrology, Barnum, and me

Oh dear, oh dear. Two of my favorite news sources – you know, those that we usually rely on for accurate reporting – seem to have done credulous puff pieces on an astrologer. More to the point, they took this kind of nonsense seriously. For example, Tonya Mosley on NPR asked the astrologer, Chani Nicholas, whether astrology was a pseudoscience. Her reply:

Astrology comes out of scientific roots and because we are studying the movements of planetary bodies over time and coming up with facts about them, it’s not to say that there is nothing scientific about astrology, but, and also, obviously it’s speaking a symbolic language. And so it’s not necessarily quantifiable in the same way as we would study other things. And then we have to get into the conversation about what is science and empirical knowledge? What are we looking at and what are our biases? So, people can call it whatever they want.

There was no follow-up to this gobbledygook; since I may call it whatever I want, I will call it bullshit. Astrology, sort of, came out of scientific roots, but it evolved into astronomy and is today no more scientific than alchemy, which no one would consider scientific even though it evolved into chemistry. When I last looked at the NPR website, the comments were uniformly negative, some very much so.

The Times, if possible, wrote an even puffier piece. Here, in the interest of giving equal time, a sample:

I’ve heard two main criticisms of astrology: that it’s fake and that it’s narcissistic. Many things are both, but astrology seems to bear the brunt of all this ire. A friend of mine always says that “astrology is fake until it’s real” — that is, until it confirms a presupposition or dovetails with a future outcome. (Yes, she’s a Gemini.)

Fake “until it confirms a presupposition or dovetails with a future outcome”? Almost a definition of confirmation bias.

PT Barnum is supposed to have said that there is a sucker born every minute. Considering how many more people there are today, it is probably worse than that. Anyway, herewith, below the fold, as a public service, something I wrote concerning astrology around 20 years ago (I cribbed it from an old Word Perfect file, so it may not be precisely the same as what is in the link. I have not linked the references, but you may find them in the linked book).

[Sir Karl] Popper considered astrology to be a belief system that is not falsifiable, even though it is empirically based. Indeed, he notes that rationalists have attacked astrology for its assertion that Heavenly bodies could influence terrestrial events. The astrologers’ assertion, however, is correct; both the moon and the sun are Heavenly bodies, and they influence the tides on the earth. Do the Heavenly bodies influence anything else, such as our futures? Popper argues that astrologers make vague predictions, explain away their failures, and, in that way, make their theory untestable.

Most physical scientists reject astrology because they can see no mechanism whereby the positions of the Heavenly bodies can affect actions of humans on earth. The nearest bodies, the sun and the moon, can cause oceanic tides, but these are unimpressive in the sense that the oceans are several kilometers deep, whereas the tides rise and fall only a few meters above and below the mean water line. The tide does not depend on the value of the gravitational attraction of the moon on the water, but rather on the difference between the gravitational attraction at the surface of the water and at the bottom. Partly for this reason, shallow bodies of water have virtually immeasureable tides. The difference of the moon’s attraction between the top of your head and your feet is similarly immeasurable, even if the moon is directly over your head, so it is very unlikely that a tidal force could affect your behavior. Objects, such as stars or planets, that are more distant than the moon or smaller than the sun influence even the tides in the deeper ocean immeasurably.

Additionally, according to the French astronomers Philippe Zarka and François Biraud, when astrologers construct their charts, they use obsolete data. [Henarejos, 1999] Owing to the slow rotation of the earth’s axis, or precession of the equinoxes, the alignment of the sun, moon, and planets among the constellations gradually drifts with respect to the seasons. When astrologers construct your birth chart, they use the alignment of the planets as it appeared on your birthday 2000 years ago. On the other hand, when they say we are entering the age of Aquarius, they use the alignment of the planets as it is today. They can’t have it both ways: “They have to choose!” say Zarka and Biraud. But they do not.

An astrologer’s response to these arguments is that astrology has been empirically demonstrated to work, so there must be something missing from my theory. But has it?

Geoffrey Dean [1986-1987, 1987] is a former scientist with CSIRO, the Australian equivalent of the U.S. National Institute of Standards and Technology or the National Research Council of Canada. Unlike many critics of astrology, he has studied it in detail and is sympathetic. Dean does not consider newspaper astrology columns to be “serious” astrology, so he set out to find serious astrologers; he claims that there are about as many serious astrologers as there are psychologists. The comparison is apt, since Dean argues that a majority of Western astrologers concentrate on psychology and counseling, not prediction, and he cites many studies that show astrology’s inability to predict.

How well does astrology perform as a counseling tool? Dean quotes one astrologer who discusses a meek person who has five planets in Aries. Having planets in Aries is supposed to signify aggression, and having all five planets in Aries is very unusual, so the meek person ought to be very aggressive. Thus, the astrologer looks for other signs that indicate meekness. That is, he looks for an ad hoc hypothesis with which to explain away his failed prediction. If he cannot find such signs, he employs another ad hoc hypothesis, specifically, the hypothesis that sometimes the opposite of what he predicts also happens:

[I]f a person has an excess of planets in a particular sign, he will tend to suppress the characteristics of that sign because he is scared that, if he reveals them, he will carry them to excess. But if on the next day I meet a very aggressive person who has five planets in Aries, I will change my tune: I will say that he had to be like that because of his planets in Aries. [italics in original]

In other words, if you have five planets in Aries, then, as a direct result, you will be either aggressive or not aggressive. Since everyone is either aggressive or not aggressive, or sometimes aggressive and sometimes not aggressive, the prediction is necessarily accurate, but it also is not testable and therefore valueless. Indeed, Dean quotes one astrologer, who recognizes that astrology cannot in reality make credible evaluations, as saying, “Any good I’ve done as a consultant, and I have done some good, had less to do with my being a good astrologer than with my being a good person.” Dean therefore considers astrology to be potentially useful, even if it is completely untrue. Unfortunately, he does not evaluate the astrologer’s statement that he has “done some good,” a claim for which we have only the astrologer’s word and no independent evaluation.

To demonstrate that astrology is, in his word, untrue, Dean describes a number of tests, such as giving right and wrong charts to subjects and asking them to pick which is right. That is, astrologers whom Dean considers serious are asked to evaluate two charts for each subject: the correct birth chart and also the chart of someone else or a chart in which signs have deliberately been changed or reversed. Each subject is then asked which of the two evaluations more accurately describes him or her. In no study cited by Dean could subjects discern their own charts more often than would have occurred by chance. Dean notes, by contrast, that subjects can recognize themselves with greater probability in valid personality tests, so the problem is with the astrology, not with the subjects or the methodology. Dean concludes that astrology is not real but can still help people as long as it “satisfies” them. But he provides no evidence that astrology is effective even in this limited way.

In the meantime, Geoffrey Dean has recanted his article, Does astrology need to be true? And published a thirty-year update in The Skeptical Inquirer. I am looking forward to reading it in more detail.

Note: Ms. Nicholas has apparently just published a book on astrology, which I will not dignify by providing a link.