The undying zombie astrology thread has attracted another latecomer to the party, this time Curtis Manwaring of Astrology X-Files, an astrology software developer who put together a seemingly testable hypothesis and added it as a comment on that thread. I’m moving my response to its own post, because frankly, nobody seems to be reading any of the follow-ups that have linked to it, and would rather continue the fight there. I’m tired of the single zombie thread, which is responsible for the vast majority of my database difficulties, causing me to hack my website to absurd degrees as a result. If it keeps attracting newcomers, I’ll close it, and add a comment saying “this post is closed, please visit any of the posts linked on page 9 of the comments if you want to continue the discussion.”
The meat of Curtis’ comment appears to be a way to test astrology, or at least one aspect of it. My problem with the suggestion is the same that I’ve had with the concept of astrology as a whole — it depends on a foundation that is simply not there. It builds on hypotheses that have simply never been proven, but rather always taken for granted. For instance, the hypothesis that there is any sort of correlation between the planets’ movements and people’s individual lives. Beyond this, much of what he suggests appears to disagree with other astrologers in the thread — even if you exclude Jamie “Darkstar” Funk of Dark Star Astrology (who has since attempted to shed his association with his ridiculous arguments here by changing his name). And to make matters worse, it appears to misunderstand statistical significance, the importance of sample sizes, and the importance of controlling for variables.
This is, as all my discussions against unfalsifiable and self-perpetuating memes, a long one. Grab a coffee.
I could suggest a course of action if there is anyone so inclined (I’m too busy with software development to take this one on now). One of many problems is that much of what is said as effects is not easily nailed down in a way that can be falsified, so a scientific proof will have to address this issue. For this area, the realm of the concrete particular is more desirable than spiritual / psychological astrology to avoid subjectivity as much as possible.
I like the idea of actually attempting to avoid subjectivity in studying astrology, but I’ve noticed that rarely will you find two astrologers who agree on what particular configurations of events actually mean. None of the present day body of “knowledge” of astrology is based on any such falsification, so to be totally intellectually honest, if we were to attempt to prove any aspect of it in concrete particulars, one would literally have to start over — to scratch everything that’s purported to be true about astrology and start from scientific first principles. The problem then, when doing so, is that one will literally find that every aspect of astrology that one believes presently, turns out to be either false or unfalsifiable. What is one then left with? Is it actually called astrology then?
As to causes, even if a correlation could be proved it still would not necessarily rise to the level of efficient causation, but more along the lines of material causes (see Aristotle’s 4 causes) at best. One issue that I haven’t seen mentioned anywhere here (I confess I don’t have a lot of time and haven’t read everything here), is that not all charts were said to be equally telling (chrematistikos) according to Vettius Valens (he was an astrologer from the Hellenistic era who was roughly a contemporary of Ptolemy). What Valens said is that some nativities were more “fated” than others and these charts one can more easily “read” and that some charts fall into a category that are more difficult to discern because they are less “telling” (without getting into the technicalities of how that is determined here). I just highlight this issue because these charts are the best cases in which I can imagine any sort of proof because if it is true that some charts are more readable than others, then it is also the case that these are the cases that can prove whether there is anything going on or not.
The idea that any one chart is not as “telling” as another, is encompassed but not directly referenced by the idea that no two astrologers can agree on what a chart actually means. As it is not a science, and as most astrologers employ as much art as science even in creating the charts much less interpreting them, any attempt at producing a scientifically derived canonical body of astrological knowledge would injure pretty much every astrologer’s practice in some way. Having a scientifically derived body of knowledge, though, levels the playing field. It would no longer depend on who’s doing the astrological predictions, so much as flow from scientifically proven fact when certain aspects come into play in a person’s life.
If some charts are telling and some aren’t, that lends to selection bias. If you get a hit, that chart must have been telling. If you get a miss, either the chart wasn’t telling or you did it wrong — either way, selection bias will not allow for the possibility that astrology, as a field, was wrong.
Picking a chart and simply reading it is not the way to go about doing this because there are no boundaries set for truth or falsification that can be measured and the key is measurement. Science has had a long battle distinguishing the measurement and the measured from Heisenberg to Plato…
And there are other problems regarding falsification of particular charts. Namely, each set of aspects has not been individually studied to discover actual correlations between the aspects and some particular about the person’s life who is supposedly influenced by said aspects. Scientific controls are difficult to pin down with regard to astrology; I will certainly agree with you on that. But they are not impossible. They are more possible if you start from first principles.
There are also a number of assumptions that would have to be dismissed such as life being random or completely under one’s control. It is not logical to expect everyone to suddenly wake up at 3 am in Chicago and suddenly have a craving for studying calculus for example. By this I mean that certain patterns would be expected to begin with, so a baseline of what is “random chance” would have to be established to determine the norm.
Life is not completely random, nor is it completely under one’s control, but this is because we live in a deterministic universe. Cause always precedes effect, and if we knew all the variables, we could calculate out things like how a person’s life might play out. We’d need far more variables than what planets existed — we’d need everything, from every charge state and position of every molecule, and an understanding of how every molecule might affect every other. I suspect there is no such thing as free will — that we are chemicals and our present consciousnesses are products of our upbringings, and since we were in turn brought up by people whose consciousnesses are determined by their upbringings in a direct cause-and-effect sort of way, we’re pretty much doing exactly what we would do in a perfect full-universe simulation.
That is a digression. What I’m saying is, we can throw out “completely random”, and “complete free will”, but we shouldn’t throw out “deterministic” just because you’re studying variables that have negligible actual effect on the local population. Such as the position of Pluto when a person was born. Pluto doesn’t exert any influence on us. The only two celestial bodies that exert any influence on us are the sun and the moon, without which we probably wouldn’t be here to discuss this nonsense.
Again, in studying what baseline behaviour is, and determining what falls under the realm of “influence”, you would be best served by throwing out every bit of “knowledge” about astrology so far with the exception of the ephemerides — the movements of the planets and other heavenly bodies — because those are merely the easiest and most accurate way to predict where a planet might be at a given date. Teach a group of people how to build charts, then have those people create charts for a very large number of people from a very random selection set — let’s say ten thousand people selected in such a way as to prevent any sort of race, gender, location or affluence biases. Make the charts blinded — assign each person a number. Give these charts to your statisticians to attempt to find trends. At the same time, have these people interviewed by psychologists, and a good sense of the person’s actual psychology. Then have the psychologists and the statisticians compare notes. See if there is a statistical correlation between certain configurations and certain psychologies (and we’re not talking something that could be attributed to random noise — you’d need a large number of people with a specific psychology to suggest a specific event).
Once you have some possible hits, do not add them to your scientifically derived body of astrological knowledge yet. This is how astrology likely got its first start — by picking hits and saying “this must be how to interpret the signs”, then building on it from there. Once you have those possible hits, you need to attempt to replicate, using another identically configured test with another random but fair selection of equal size. Once you have replication of certain patterns, then and only then should you start looking for (again, a large sample of) people that fall under that pattern in order to see if there is a statistically significant relationship between people who have that pattern in their charts, and the specific effects you believe you’re looking for. If you can blind this, by having the tests carried out by different sets of statisticians, chart-makers and psychologists, that would be of course best.
Let me suggest that given that a chart with angular trigon lords and with the lot of spirit in the 10th sign from fortune, that when the zodiacal releasing reaches the 10th sign from fortune (if it does so in their lifetime) that they will reach their peak effectiveness and culmination of their actions (sometimes called fame). Obviously we would have to define “famous” in some way. I have done a few case studies such as what happens with actors (such as John Travolta who reached his peak and he was known for Pulp Fiction (1994), Saturday Night Fever and Grease(1977) when in his 10th from fortune period). George W. Bush reached his peak period starting in 1998 shortly before becoming president. Hitler reached his 10th from fortune period and the next day the Nazi’s won 108 seats in congress. Einstein was also in his 10th from fortune period when reaching peak fame. With more clearly defined rules, there is more hope for falsification which also allows for real verification.
Let ME suggest that your sample size is infinitesimally small. To achieve statistical significance, you’d have to have a very large, diverse and representative sample of people that match your selected criteria (e.g. who have angular trigon lords and a lot of spirit in the tenth sign, whatever that means), who are selected in such a way that other variables (e.g. country or region of birth, family cash-on-hand, etc.) are controlled for. Find me a few thousand people spread out amongst many family backgrounds and prove to me that there is no other variable that might account for the effect you’re claiming, and show me the data that shows that this particular chart orientation is statistically significantly correlated with the particular effect you’re claiming, and you might have something.
Rather than picking some people you know to be famous, then trying to find a correlation between them, pick a configuration and see if your totally random sample shows any trends that another control sample of completely random configurations does not. Repeat the test until you’re sure you’re not seeing noise in the data. Otherwise, you could be seeing a “mars effect” — a “statistically significant” result that actually isn’t significant at all.
I recuse myself of the responsibility of actually carrying out any of these suggestions, as I would be considered a hostile entity having argued strongly against astrology in the past. My protestations that my results are untarnished by bias, would naturally go unheeded, by virtue of my actual bias against things that people believe strongly, but show none of the intellectual curiosity or rigour necessary to make an honest assessment.
I am a software engineer and have a background in classical philosophy, statistics, mathematics, astronomy and the history of science.
I appreciate the note, but can’t help but think it is secondary to your argumentation. A question comes to mind — which did you study first? Astrology, or each / all of these fields of study? How much did you know about astrology, and believe about astrology, before you studied these fields?
You’re trying to set yourself up as an authority in several fields as though it means you have vetted the philosophical, statistical, mathematic and historical aspects of your field, being as it is the precursor to astronomy. Problem is, you’re still, in my estimation, starting from the principle that there is something to astrology, you just have to find an effect or a correlation or a mechanism. Normally one has one of the three before assuming they have anything at all. In a field that’s been investigated for as many thousands of years as astrology, one would think you’d have something by now.
I have no doubt in my mind that your astrology software (which costs a pretty penny by the way) is extremely effective at calculating out charts and events and such. They are very likely extremely well coded, polished pieces of software. And they’re probably very accurate in analyzing exactly where planets are, and exactly what charts would be drawn given the data at hand. My problem is not with the generation of the pseudo-random data, because it all has the patina of science, with complicated formulas and inscrutable charting formats. My problem is with the idea that all this pseudo-random data has been shown to have ANY correlation with reality whatsoever. As far as I can tell, it has not, and will not.
To this charge, every astrologer produces a small handful of hits, usually of famous events and celebrities. No astrologer has taken it upon themselves to increase their sample size beyond that small handful, and I can understand why — it’s a lot of work to do one chart, and when you get a hit, you’ve “gotten it right”, and it’s very exciting. I need more than people getting 25% hits throughout history before I’ll believe there’s any sort of correlation. I need way bigger samples, way better controls, and blinding, before any specific configuration-to-influence seems even remotely plausible. And once we have a configuration-to-influence correlation, I’ll need some plausible mechanism. That can come last, but it CANNOT be omitted in the pursuit of the truth about this universe.
You’re making the claim that it’s true. The burden of proof is on you. Otherwise, in view of the lack of evidence ever presented, I call hogwash on the whole self-perpetuating endeavour.