31
Thanks, dubbhism. Lots of good stuff in your latest post.

I think it's important to distinguish between cultural astronomy (ethnoastronomy,) solar system physics, and horoscopic astrology. Nobody doubts that the moon affects tides, for example, but this doesn't explain why a moon in Leo is different from a moon in Capricorn. There is also a lot of star lore held by different societies around the globe, but most of them are not even close to horoscopic astrology.

Even if we postulate some sort of physical, mechanistic causal mechanism whereby planets influence human behaviour, we are still left with all of the non-natal forms of astrology like horary and mundane. Even predictive astrology based on nativities is problematical, because your planetary positions at Time 1 no longer exist at Time N. Most of the currency of horoscopic astrology-- signs, houses, degrees, constellations-- do not exist up in the sky. They are what the astrologer-- through agreement with like-minded astrologers-- maps onto the heavens. Signs were invented by the Babylonians, for example, not discovered as having any real existence. They exist in a theoretical cultural domain.

Then we run into all kinds of problems of definition in running a big statistical study. For example, David Cochrane is one astrologer doing some fun stuff with big data sets, but I think he assumes a modern astrology platform. Some of his work deals with conjunctions to the sun as defining human personality traits; whereas in a traditional interpretation such a planet would probably be combust (unless it were cazimi) and thereby weakened.

I'm not saying that the many, many required definitions and parameters couldn't be addressed; it's just that they should be, prior to crunching all those numbers. Otherwise the researchers leave themselves open to all kinds of potential bias.

I think we'd need large-format multi-variate statistical tests, not just the old planet=trait binaries.

We cannot forget that astrology deals with people: their hopes, fears, talents, mistakes, and all the rest of it. Many things that influence human lives wouldn't even show up in a natal horoscope without supplementary data, such as gender, family religion or atheism, and skin colour.

To use your Dutch example, it matters if a native of Amsterdam is Muslim or ancestral Reform church, but you cannot read this off a horoscope.

Then statistics deals with probabilities, so if your planet=trait correlation holds 75% of the time, that might be highly significant, but it doesn't explain the other 25% of your sample.

Most quantitative research these days is done by teams. I think a team comprised of successful professional astrologers, quantitative social scientists (who do get a background in stats,) and an astronomer might get somewhere, but I don't see any one person having the required expertise.

32
A mundane chart for the Battle of Cannae consisting of a progressed Canlunar (Moon's entry into Sidereal Cancer, the MC progressed by the travel of the Moon from that time to the time of an event) with an eclipse chart previous to the event and located there around it. Charts are shown in the Tropical zodic. Locations of points are shown in right ascension. I do not use signs nor houses but am forced to show them by software. These charts are read by what planets or points are being swept by or are in tight aspect to the angles.

The progressed Mars/Uranus midpoint rose at 2:15 pm, LMT.

Tens of thousands lost their lives there then, thousands more were wounded, tens of thousands more were captured by their enemies. Tens of thousands more experienced victory over an enemy.


Image

33
Thanks a lot for that unique astrology, i'm not sure that your analysis would change Cicero's mind, but the plot certainly thickens.

It is my impression that there are quite a few extraordinary things happening during the big battles between Rome and Hannibal. Personally i focus on asteroids like Hannibal (conjunct Uranus during the first big victory against Rome at the Trebia) and Roma, but also Chaos, Eris, Sedna and Pholus. The role of Neptune in the Battle at Lake Trasimene might also be interesting to you: the largest ambush in military history (record still standing) under the cover of fog.

For a short characterization of 4 battles between Hannibal and Rome based mainly on asteroids see http://www.dubbhism.org/2017/01/star-wa ... -rome.html

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

@waybread, thanks for your thoughts, i think all your concerns are valid.

Just one last thing about what i like in Big Data: techniques like Deep Learning are (or can be used) relatively theory agnostic. So when you say that we have to be careful about letting bias creep in before we even start, my suggestion would be that with Big Data, you don't have to start per se with very specific assumptions about possible correlations. Instead you might get a big, high quality dataset and let unassuming algorithms hunt blindly for all kinds of patterns. Fully automated research (or data-based preparation for more serious work) so to speak. You could in principle search fairly exhaustively for certain simple correlations this way. All you need is a little bit of a focus and some general assumptions like perhaps the canonical mythological meanings as they seem related to certain heavenly bodies, the qualities of certain aspects and maybe one or two other parameters.

For example, i looked at 13 years worth of crime data (calls for service) as provided by the Police Department of the City of Chicago https://data.cityofchicago.org/Public-S ... /ijzp-q8t2, focusing on the theme of 'violent crime' meaning crime categories as diverse as public peace, robbery, homicide, assault and sexual crimes.

Unfortunately this particular data set is not high quality (calls for service are not the same thing as actual crimes so there's a lot of noise) but by focusing on hard aspects between the most common physical objects (planets, dwarfs, some centaurs and asteroids) the algorithm was still able to single out Mars-Pluto conjunctions and conjunctions of Sun + Pholus, Nessus, Eris and Sedna as indicators for a range of violent crimes, and confirm these results in similar datasets published by the Police Departments of San Francisco and other major cities in the USA (these datasets are even less quality than the Chicago set, but still).

A professional data scientist (which i am not) could take such a procedure a few steps further and automate the research in such a way that - in theory - every possible angle between every included object (or any other quantifiable parameter for that matter) can be checked against a given high quality data set. That's a somewhat different approach. I think it has potential. But you do need lots of good data, a big computer and serious programming and statistics skills to get it right. A potential problem with Deep Learning is a certain black box element, which is bad for transparency, but i think this can be solved good enough.

Although the patterns i noticed in the open source American Police Department data are worthless for science because of bad confidence intervals, the correlations that did pop up would seem archetypically correct so this particular project still had value for me because it suggests new directions for research that may some day even be scientifically bona fide.

By the way, i actually found two separate ways to track the influence of these bodies i mentioned in the Police data: either by looking at hard aspects or using Golden Aspects. For this last method i calculated the total number of Golden Aspects or Phi angles between the Sun, Pluto, Mars, Chiron, Pholus, Nessus, Eris and Sedna in 24 hour intervals for those 13 years of data (1.5 degree orb) and fed the algorithm a 'phi angle count' and a 'phi angle index' (as in Barbault-index, everything both geocentrically and heliocentrically) producing similar/better results.
~~~~~~~~~~~~~~~~~~~~~~~~~

http://www.dubbhism.org/search/label/astrology

34
dubbhism, I just prepared a carefully crafted response to your post, only to accidentally delete it. Right now I don't have the patience to recreate it, except to say that, no.

You're familiar with the "garbage in, garbage out" problem in research?

Obviously a statistical study has a minimum legitimate sample size, but bigger is better only if the data and methods still meet the basic research design criteria designed to minimize bias.

A correlation below 50%, even if statistically significant, is still a poor predictor.

I shudder at the thought of using "canonical" mythological meanings, because a study of mythology reveals just how syncretistic and contradictory the myths are.

Surely you're familiar with how much real estate around the zodiac a 10-degree orb for the sun in relation to another planet using only the major aspects would occupy: 160 degrees out of 360, by my calculation.

Plus, how does Big Data address the OP, which is how or why astrology should work, if it does?

35
Surely you're familiar with how much real estate around the zodiac a 10-degree orb for the sun in relation to another planet using only the major aspects would occupy: 160 degrees out of 360, by my calculation.
Orbs are subjective, quantitative calibrations of the results. In the 'automated' approach i was suggesting, variation in orb size can be automated.
I shudder at the thought of using "canonical" mythological meanings, because a study of mythology reveals just how syncretistic and contradictory the myths are.
All i'm saying is that in general, Mars-Pluto conjunctions would not seem to be related to themes like love, peace and harmony. That's what i mean with "canonical". Same thing for the dwarfs. If you want to research whether astrology works or not, assumptions about classic Graeco-Roman mythology or any other body of mythology are not needed per se in my opinion, but allowing for a modicum of *meaning* (even if based on nothing else but a broad consensus/working hypothesis in the astrological community) would seem fairly essential to me.
Plus, how does Big Data address the OP, which is how or why astrology should work, if it does?
In my opinion Deep Learning and a few related techniques can be used in a relatively theory agnostic way. That's good for transparancy (just like using data produced by 'objective' parties like Police Departments). Therefore, it's not impossible (merely speculating here) that relevant new kinds of data may be collected using this technique, potentially leading to solid new insights. If that's too indirect for this thread i'm sorry.

Anyway Deep Learning is not magic, it's just fancy statistics. Here's an example of one of the discussions going on in this field https://www.quantamagazine.org/new-theo ... -20170921/
~~~~~~~~~~~~~~~~~~~~~~~~~

http://www.dubbhism.org/search/label/astrology

36
Thanks, unique_astrology and dubbhism, for your thoughtful posts.

On that now 82-page thread at Astrodienst on this topic, I argued mid-way through the discussion, that I don't think statistics are going to help us very much, but if others think they will, I will be interested to see the results of their studies. I'm not opposed to attempts to quantify astrological outcomes on principle, rather, I think that to do it correctly, astrologers (or whoever does the analysis) are going to have to start thinking in terms of specific definitions, outlining their assumptions, and explaining how they mitigate at least the most obvious potential biases.

This is how academics set up their research. If they don't, they get clobbered by their critics.

Then whose astrology would you use? I have a lot of admiration for astrologer David Cochrane, a really original thinker, but I don't think he'd get to square one with traditional astrologers, because his assumptions seem to be thoroughly modern. See, for example, https://www.youtube.com/watch?v=3pOkwi63qOw

My course background in statistics was a long-ago and far away required course, but during my career I worked a lot around social scientists and scientists who used stats routinely. I read their papers and attended their conference and seminar presentations. My poor knowledge of stats was a work-around, and I did learn how to get my head around the introductions and conclusions of papers in a lot of different fields.

Let's take well-known actors as an example.

There are a lot of different multivariate statistical tests out there. Some of which don't require mythological assumptions. For example, the Astro-DataBank is full of the horoscopes of American film actors. It would be possible to select out the Rodden AA individuals, have all kinds of horoscope data bytes in digitized form by degree, and just see whether given "actor" signatures showed up significantly. Then you could compare them to a control group of the same size. (I. e., kind of a Gauquelin-type study for the 2010s.)

The problem that I see, having looked at the horoscopes of a lot of actors and other creative people, is that I'm not convinced that a horoscope discriminates between actor vs. non-actor. Rather, I think it shows the kind of actor someone will become once s/he's on that trajectory. But then I didn't get into minor aspects, fixed stars, asteroids, Arabian parts, and so on; in the way that a big data project could.

But if my surmise is correct, and a horoscope shows how someone will go about acting, painting, &c; rather than as a simple career choice, or perhaps a simple "fame factor," then we've got a real problem, if ithe problem in fact boils down to each artist's unique creativity.

The other problem is that succeeding in a field like acting depends on much besides raw acting talent. Getting discovered or getting the big break might depend upon a transit. Having good social connections in show business. Or conversely, a talented person living his life in West Overshoe, Iowa, so that all of that talent went into the high school play, or maybe becoming a gifted high school teacher, never to surface into the Rodden data base.

Maybe, just maybe, Big Data will show some blips and burps such that a given correlation is indeed statistically significant. But if it explains only at the level of 65%, then that still leaves 1/3 of the sample unaccounted for. Not helpful if an astrologer gets a read-my-chart question like, "Can I have a successful acting career?"

This starts to nudge us into more qualitative research. Which is the current direction that a lot of social scientists are taking.

And then what is the role of the astrologer? The minute the talk turns to statistics, suddenly the astrologer becomes invisible.

38
The minute the talk turns to statistics, suddenly the astrologer becomes invisible.
But waybread, only a few posts ago i gave an example of (excellent) peer reviewed research by Renay Oshop published in a fine scientific journal, based on big data and solid statistics.

What exactly is not good enough for you in the statistics of that particular research project?
~~~~~~~~~~~~~~~~~~~~~~~~~

http://www.dubbhism.org/search/label/astrology

39
dubbhism wrote:
The minute the talk turns to statistics, suddenly the astrologer becomes invisible.
But waybread, only a few posts ago i gave an example of (excellent) peer reviewed research by Renay Oshop published in a fine scientific journal, based on big data and solid statistics.

What exactly is not good enough for you in the statistics of that particular research project?
Mercury retrograde is an interesting problem. In both traditional and western astrology, it is associated with communication and transportation miscues.

However, from an astronomical perspective there is no such thing as retrogradation. Mercury (or any other planet) never actually goes backwards in the sky. Apparent retrograde motion is based on a Ptolemaic, geocentric perspective. Not Mercury's actual orbit. https://en.wikipedia.org/wiki/Apparent_ ... ade_motion

Then let's do the mythology. Mercury (Hermes, and before him, the Babylonian god Nabu) was the messenger or scribe. The connection of this planetary god with communication is really ancient. The planetary god Mercury "going backwards" in the sky seems highly symbolic of communication that is not direct, or is thwarted in some manner.

So if Mercury retrograde is not an accurate description of the planet's orbit but it does correlate with mythologically derived "inaccurate communication" in a big data study, what does this say about the cultural cosmos?

The paper by Renee Oshop looks like a set of slides, apparently set up for a "Kepler" lecture or conference presentation. It didn't have the usual format of a journal article, notably missing a bibliography.

My Mercury has progressed into Taurus, so maybe that explains why it gets sludgy these days. Hopefully you can help me out.

However, I didn't get this last bullet of Oshop's first slide, did you?

"Just the ones corresponding to Mercury retrograde were pulled out and analyzed (like a single person's voice.)"

Did this mean there was no control group? Mightn't misspellings occurring during some other astrological event (like Venus retrograde or Mercury combust sun) have a statistically indistinguishable signature from Mercury retrograde? Or was this handled in the "linear regression" slide?

I also didn't get the "transformed data" slide. Did you? The author seems to have transformed the data into a 30-day running average. I'm really unclear how this could relate to Mercury retrograde periods, which are relatively short.

What were the axes showing on the "beautiful Mercury retrograde patterns" graph showing? They weren't labeled.

The "Sum of Corresponding Waves" slide seemed to show a 13.2% increase of misspellings during Mercury retrograde periods, but since there seems to be no control group, it's hard to gauge the significance. Then presumably 86.8% of words were unaffected! In other words, is it highly likely that words will not be misspelled during Mercury retrograde periods?

I didn't get the "sound" graph that followed. Did you?

In the "linear regression" chart, did it look like the modern outer planets had a vastly bigger effect than Mercury? But note that they are retrograde a whole lot of the time, so are we looking at a data artifact here? Again, astronomically planets do not go retrograde.

Then what do you make of his disclaimers in the "some more details" slide? Or was the author just trying to show how his method could be used, not specifically making a case in support of astrology?

Anyway, sorry to be as dumb as a post about the significance of this presentation. If you can explain what I'm missing, I would seriously appreciate it.

As I've said before, if researchers can come up with useful statistical studies of astrology, using big data or more sophisticated statistical tests I would like to know about it.

40
Therese, if these don't help, I'll try to answer any questions.

http://libguides.usc.edu/writingguide/qualitative

https://en.wikipedia.org/wiki/Qualitative_research

https://measuringu.com/qual-methods/

But basically, quantitative research in astrology would have to proceed in some fashion like a social science, unless we could somehow do it as neuroscience or cognitive psychology. Astrology wouldn't qualify as a physical, natural, or medical science.

The term "social science" in fact stems from an era when researchers who studied human society wanted to put it on a scientific footing. Over the past century, a lot of research on people in groups, or the individual in society was quantitative. Who hasn't taken a questionnaire, for example? The results of those surveys were typically aggregated and plugged into statistical tests.

Sociology and economics are probably the most quantitative of the social sciences; and anthropology, the least. Psychology, which some modern astrologers feel is the closest discipline to their field, in fact moved far away from Jung and Freud in favor of carefully controlled clinical and laboratory studies.

But in the past few decades. the subjects of social science research, together with some of their researchers, began to realize that a scholar's "God's eye" view or top-down view of a group of people might utterly miss or misunderstand what life looked like to them.

Consider that, from the comfort of my middle class existence, I might devise a questionnaire about the issues facing homeless people in my nearest city. I might even go out on the streets and convince dozens of homeless people to take it. Then I could come home, aggregate the responses, crunch the numbers, and give an acclaimed academic paper on homelessness-- all the while knowing very little, from the ground up-- on the experience of homelessness.

Qualitative research generally started with the idea of engaging the research subjects in the design of the study; and creating some benefit out of the research for the participants. When I agree to respond to questionnaires, I am often disgruntled by how little they match my experience. But what if I could help design them?

Then oftentimes you couldn't collect data in the same way as one could in using, say, census data. (Or the Astro-DataBank.) Homeless people, to extend my example, are often users of illegal drugs. They might be very wary of talking to me if I looked too authoritative. I might need to befriend a few of them, ask if they would introduce me to their friends, and from these friends, ask for introductions to their friends, and thereby extend my network of subjects. This is a qualitative method called a "snowball sample," which is very different from quantitative researchers' random or stratified samples.

Maybe I would just ask my research subjects to tell me their stories, based upon just a few open-ended questions. From their narratives, I might apply some simple methods of textual analysis.

I note also that business schools do a lot with case studies, which draw on students' factual learning while engaging their ability to use it analytically in real-world problem-solving.

I think this type of research has broad implications for research in astrology. I mean, what if someone conducted open-ended interviews with some really good professional astrologers about their experiences of chart-reading? What happens in the moment when the chart suddenly begins to describe a human life? Assuming astrology does work to some degree, I think we'd have to start with astrologers themselves, not with presumed causal or synchronous effects of planets on people.

If we ask astrologers what happens in the moment when the chart begins to describe a human life, we start focusing on the individual nativity, the horary moment, the political election outcome. Astrology isn't about what's up in the sky so much as it is about the process of connecting with individual human lives.

41
hey waybread, i wasn't referring to the Mercury Retrograde slides, i was referring to this article:

Twitter Followers Biased to Astrological Charts of Celebrities (Published in the Journal of Scientific Exploration)
https://www.dropbox.com/s/cpm5359n308wm ... hop.pdf?dl

The Mercury Retrograde theme is very interesting, but not published as a scientific article. It was a lecture for an astrological convention i think.
~~~~~~~~~~~~~~~~~~~~~~~~~

http://www.dubbhism.org/search/label/astrology

42
waybread wrote:
So if Mercury retrograde is not an accurate description of the planet's orbit but it does correlate with mythologically derived "inaccurate communication" in a big data study, what does this say about the cultural cosmos?
In general i would assume that meaning insofar as it's being mediated through the interpretation of astrological signs and symbols is essentially a multivalent phenomenon that can, but doesn't have to, mirror certain physical phenomena 'exactly'. Multiple perspectives would seem simultaneously valid, even when they're not pointing to the 'exact' same physical phenomenon and/or metaphysical meaning. To me personally there is nothing inherently contradictory about this; i don't think Laplace's daemon or any other mechanistic 'billiard balls' model of reality which might seem incompatible with this view is still relevant. Remember "It from bit", but also complexity theory, where we have similar patterns popping up all over the place simultaneously.

In general i personally think excluding certain (physical) phenomena, techniques or symbols beforehand can be counterproductive in astrology. I generally prefer inclusive thinking to exclusive thinking, even though it may seem impractical. For example, i think both the tropical and the sidereal zodiac have their use, both the geocentric and heliocentric approach have their use, both Western and Eastern (in fact all) bodies of mythology may be relevant. And even meaning based on more basic semantic/morphological commonalities, associations, parallels etc. may be relevant, even from a highly individual 'topocentric' perspective. But while the countless individual perspectives have their relevance, we also have to include various mundane perspectives, like the 'global' historical perspective that's offered by for example Richard Tarnas.
~~~~~~~~~~~~~~~~~~~~~~~~~

http://www.dubbhism.org/search/label/astrology