Sunday, November 20, 2005

Astrology, Myers-Briggs, the Sibuxiang, and You

Sibuxiang Today, I want to have a short conversation about astrology and its kin. First, I should say that I think that it is possible to do very good astrology. Perhaps this sounds surprising to those of you who know my views on the subject, but let me tell you what I think constitutes good astrology. If you ever pick up a copy of the Chicago Weekly News, chances are you'll find horoscopes by Rob Brezsny. Here are some examples I e-mailed a friend one night:


Gemini: Since your life has more than a slight resemblance to a chess match these days, I'm calling upon grandmaster Victor Korchnoi to talk a little strategy. "I like to coax my opponents into attacking," he says, "to let them taste the joy of the initiative, so that they may get carried away, become careless, and sacrifice material." Please meditate on how you might adopt this approach to use in your next gambit, Gemini. It's time, in my astrological opinion, for some smart mischief."

Pisces: Imagine that your life is a detective story. The goal is not to solve the crime, but to solve the mystery of why you're here on earth so you can carry out the special mission you've come to accomplish. Sometimes you go for months without even looking for clues. You sleepwalk through the world, reacting blindly to the tricks that the gods use to try to wake you up. Then there are those phases when hot leads and fresh evidence pop up all over the place, convincing you beyond a doubt that magic is one of the fundamental properties of reality. This is one of those times, Pisces.

Capricorn: I'm guessing that your imagination is both excited and perplexed; that your senses are heightened yet on the verge of being overwhelmed. According to my analysis of the astrological omens, you're going through a phase that at its best might be called a medley and at its most challenging a hodgepodge. It has resemblances to the sibuxiang, a mythic Chinese animal with the head of a dragon, antlers of a deer, tail of a lion, hooves of an ox and body covered with fishy scales. So is there anything you can do to ensure that you use this time to stir innovative solutions to long-standing problems? Here's one suggestion: Once a day for five minutes close your eyes and visualize yourself pulling off rodeo tricks while riding an exuberant sibuxiang.

What I like about these horoscopes are how creative they are, and how they juxtapose very different ideas. When someone reads such horoscopes, they learn things like the fact that Victor Korchnoi was a chess player. Further when advice is given, it is open-ended and capable of many different interpretations, but nonetheless it provides a framework to think through issues that might arise. For instance, perhaps a Pisces will learn to see the world in a slightly different way as a result of reading the horoscope above. That in itself might make reading the horoscope worthwhile.

“But what about the actual astrology?,” you might ask. “Don't you have to bring in the stars?” I suppose Rob might in fact look to the stars for his horoscopes. But I hardly think you need to bring in the stars. I'm analytically-minded enough to be highly skeptical that it is ever possible for a star thousands of light years away to influence most of the lives of people on earth in any significant fashion. That just seems sort of obvious to me. I mean, the mere fact that so few astrologers pursue astronomy degrees certainly counts against the validity of the field. But I suppose an astrologer thinks about things more indirectly, e.g. that some force coordinates both the stars and our lives, and that by looking at one we learn something about the other. But, of course, one would want to see the evidence that this relationship actually holds.

But the evidence is everywhere, might be the reply. See, it works: you're a Gemini, and you are SO a Gemini! But here we get into a whole mess of other issues. The fact of the matter is that human beings are incredibly complicated creatures. There are so many sides to our personalities that come out in particular circumstances, and so many ways of interpreting the very same action in the same circumstances, that I bet if I were a Sagittarius, you'd be telling me I'm such a Sagittarius. Further there is the added distortion that people who actually believe in horoscopes/fortunes/astrology often only look/remember horoscopes when they apparently seem to work. If I open a fortune cookie, and it says, "You will discover a great fortune," I might laugh it off, add the mandatory "in bed" clause, and then forget about it. But man, if I came home that night to find a large fortune on my doorstep, I may well suspect a coordinated project, and I'll certainly never forget that horoscope for the rest of my life. I take this to be actually a manifestation of a more general phenomenon. Human beings are notoriously bad at producing reliable statistical inferences from small sets of evidence, at least when measured against the canons of probability theory. This reminds me of Tversky and Kahneman's psychological experiments in the early 70s, and Nisbett and Ross's experiments in the early 80s. In both cases people were asked to make judgments based on statistical data, and in both cases people in the non-control group seemed to consistently draw conclusions that were not warranted by the data. This was taken to show that human beings are naturally prone to error. Of course, the unique feature about these studies was that the statistical data that people were generalizing from was particularly abstract, and perhaps not the best indicators of covariation, as Hilary Kornblith points out in his book, Inductive Reasoning and Its Natural Ground. The fact is that we are actually pretty good at generalizing about the natural world, what we are not good about is generalizing data in strange and unusual contexts that share little prima facie resemblance with the world that we evolved in tandem with.

I think it is this very feature that is exploited by people who believe in astrology and try to convince others to do the same. For, the fact is that astrology is an incredibly out-of-the-ordinary context, since the detachment of the data from any known causal properties means that we have very little ground for detecting covariation, and the sort of faith-based acceptance just flies in the face of most of the logical apparatus we usually think with. So, almost by definition astrology is an area in which we would be prone to errors in grouping kinds. Thus by reliabilist considerations alone we should consider dismissing astrology.

But why am I talking about astrology at all tonight? Well, it's connected to online dating actually. One of the online dating services I have used from time to time now incorporates personality assessment features based on Myers-Briggs categories. People who want to look for a suitable mate log on, answer about twenty questions, and then get a listing of the people that they are compatible with. Naturally enough I'm quite skeptical about these sorts of tests, not only because it is difficult to believe that a twenty question test is sufficiently extensive to determine a Myers-Briggs distribution, but also since, like astrology, I seriously question the existence of natural kinds corresponding to the types that are identified. I mean, don't get me wrong, I do think that there are people who take a Myers-Briggs test and record such extreme scores that the test probably does indicate a lot about their personalities. I actually have gone on dates with people who swore by Myers-Briggs, and may have well made their decision to date me based on the results from their tests! But, in my own experience, I find the tests that determine Myers-Briggs categories – especially the trivial online ones - to be very inconsistent. I remember the first time that I took such a Myers-Briggs test in high school, I had it narrowed down to two questions, and I couldn't decide which way to go on either, since the phrasing of the question seemed ambiguous. So I decided I'd score myself twice and see what the results were. The scores were radically different, and depending on how I chose, I would either be an extrovert or an introvert. Of course, the test didn't have an in-between category, and it presumably wouldn't produce such decisive outcomes if it did. And even when it does produce decisive outcomes the parameters seem quite suspect to me - for instance the thinking/feeling distinction, which seems to presume that the concept of feeling and the concept of thought could be easily differentiated. Intuiting/sensing? Well, maybe all of these have technically definable meanings, but most people don’t think of their results in this way.

Maybe I need to learn a little more about the test. I admit that I've been too skeptical about the idea of personality tests in general to even do much exploration on the subject. But, as I say I do now use a dating service that incorporates mandatory personality screenings. The problem with these tests is I always come out with a different sort of personality. I've taken about ten tests in recent years (both on-line and on paper), and here are some of the results that I’ve recorded (really!):

ESFP
INFP
INFJ
INTJ
ISFP
ENFP
ENFJ

(There may be more, but the rest escape me.)

Looking over this list it seems I have been diagnosed with opposite characteristics in all four of the categories! I have been assured by some believers in this test that this weird outcome is the result of having so moderate responses to the test, and that I need to take more extreme positions to really pin down, "What I am." But that seems viciously circular: the test is not working for you, since you are not working for the test. But I suppose there may well be a difficulty based on self-assessment, and that certain aspects of my personality may not be apparent to me, although they would be apparent to other people. I'll concede that, although I refuse to admit that this "me" that other people see is more of a "me" than the "me" I see. I am an anti-reductionist after all, and I find the concept of some "real me" that I don't have access to, to be deeply problematic. Anyway, one could look at these results and assume that I'm coming closer to the ideal me. So let's generalize over the tests. I/E seem about evenly divided, and so does J/P. The only results that seem to side strongly one way or the other is the intuition vs. sensing, and thought vs. feeling. Apparently, I am a feeling/intuiting sort of guy. But even this strikes me as deeply problematic. I mean, I'm so invested in analytical arguments that I teach logic at the graduate student level, and when I’m teaching I can be relentlessly analytical. It’s true on a number of important subjects I'm more likely to follow my feelings than my thought, but on other subjects I'm strongly in the thought category. (Don’t worry, I realize the irony in my use of these concepts! – I should add that at least I’m not seeking to determine categories empirically, but only using normal everyday concepts.) Perhaps this is the ultimate problem I have with these tests: they just generalize too much. Hmm, I'm reminded of a passage in Plato's Philebus that I translated today. The argument is a little abstract, but the basic idea is that people are too quick to move to single concepts, and they don't pay adequate attention to the differences that these concepts subsume. As some of my previous posts suggest, I think this holds in a lot of areas of human thought, and personality tests don’t seem to be an exception. I’m guessing the error is often missed, because of our tendency to make erroneous generalizations from limited sets of evidence. Thus, someone will say that they are a T because a test has told them they are a T, and then will come to believe it even if they might well be defined as an F in most areas, because they just assume it must be right.

There is a further rather deep philosophical point here to. When one gets results for Myers-Briggs, what one is determining is the average response to a set of questions. That is many particular decisions are made by the test-taker, and all of these are reduced to a single parameter. But a lot is lost in the reduction. For instance, I know someone who would be the paradigm of an extrovert personality, in the sense that this particular person loves being around lots of people, telling everyone exactly how she feels, and can at times find it difficult to be quiet or alone. But this is her personality around her friends. You throw her into a situation with lots of people she doesn’t know, and a crowd with particular characteristics, and she’ll be as quiet as a church mouse. That is her “personality” can be brought out in different ways with different people. This is a very common phenomenon. A person can be an introvert with some people, and all you need to add is one person with a slightly different personality, and you’ll see the person happily and immediately become an extrovert. All of this variation, this “landscape” of the psyche if you will, gets lost when you reduce such a person to an E or an I, even with indexes of variation. I take this to be a deep point, because most people, although aware of these facts, tend to like clear categories to put people in and ignore these complications. But it seems to me that it is these complications that most clearly bring out our individualities as human beings. Like the sibuxiang, individual people are not just E's or I's but veritable alphabets of different types.

Anyway, perhaps it is a bit antagonistic for me to subsume astrology and Myers-Briggs tests under the same category. But as the discussion above shows, I think that many of the same problems exist with both. This was even made apparent to me tonight from the language employed in the results of tests. On one site, a particular characteristic "rules" the other three as one's astrological sign might be "ruled over" by a particular planet (Admittedly, it was a very Jungian site). Or, a picturesque name is given to one of the divisions, and then it is used as the subject of sentences like a Chinese zodiacal name: "The growth teacher is equally skilled . . . " Or, when people talk about the Myers-Briggs test they say things like, "You are SO a P." But despite the fact that I think that there are a lot of similarities here, and that people who get too into online personality profiles are probably guilty of the same sort of irrationality as people who are too into horoscopes, I should say that I don't think they're exactly on par. In fact, even with my skeptical leanings I've recently found a purpose for Myer-Briggs personalities. Although one can easily get personalities wrong, I suppose it is worthwhile for someone to note that at least 16 different types of personalities do exist. This is something that one can easily lose sight of, and just assume that everyone’s thoughts and motivations operate like one’s own, or at least in one unified sort of way in contrast to one’s own personality. Hence, knowing the different types may make it easier to understand why someone is particularly offended when their partner is late for an event, or perhaps one can find a way of dealing with a communication problem with a coworker by just exploring the other places that person might be starting from, so to speak. I suppose it could even help one understand what has gone wrong in a dating relationship, and what steps can be taken to overcome the difficulties.

But, this brings me to the second sort of problem I have with the personality tests on this dating service. The entire reason this particular online dating service uses a variant on the Myers-Briggs assessment is that it can be used to pair "compatible" personalities. And once you get diagnosed as having one particular personality trait, the test tells you what personality-types you are compatible with, and what personality-types you should avoid. So, with my INFP result I went down the list and sort of guessed the personalities of people I have dated, and then looked at the compatibility story. Suddenly, there in front of me was the story of a recent break-up! Every part of the problem that it described was just what happened. For a moment I was suddenly convinced, maybe I am an INFP after all! But not leaving this to chance I went back and adopted my ENFJ personality (e.g. answered a few questions I waffled about in a different way), and then I had a similar experience, a few more checks with other personality types and I realized that basically I'd have this experience regardless of what my initial personality type was. The fact was, like horoscopes, the accounts were so general, and my guesses were so vague, that it was possible to get the same feeling of affirmation from whatever personality type I started with.

Not to stay entirely on the negative, I also looked at the types of personalities I am supposed to be compatible with, and found the results very disappointing. For instance, as an INFP, I'm supposed to be compatible with someone who is all about balance, and staying indoors, and not going partying at night, and staying at home reading a book. Although I admit that I'm not entirely adverse to people like this, I couldn’t help reading the profile without thinking it was a description of BO-RING. I mean, I suppose I’m the sort of personality that likes to have things a little out of balance, a little challenging. For instance, I decided not to be an architect at one phase of my life in part because it was too comfortable. I was good at it - I had unusual abilities at the age of six, something that my teachers recognized and tried to get me to enroll in vocational school drafting courses as a consequence, and after working on it the next decade or so, I can say with very little ego that I was really good at it. I think of it as my true talent, actually. Maybe someday I’ll go back to it . . . anyway, I decided to do something completely different because it just wasn’t challenging. I mean, I suppose that if I was in some great firm somewhere like Renzo Piano's, testing all sorts of new materials and the like, I would find the challenge I was looking for. But that seemed so far from the little country town I lived in. Well, I suppose I shouldn't stress this point too much since there were other reasons why I changed professions. The important general point is that I would probably intentionally try to subvert the compatibility suggestions this dating service recommends, because “easiest” is not a synonym for “best.”

But I suppose that the problem in all of this is that, as with most things in life, you have to take what you hear with a grain of salt, so to speak. If you don't take a person's status on the Myers-Briggs scale as a definitive result, if you're willing to be flexible in your assessments, and view the world not through a single personality lens, but to see a person as composed of many different aspects – overlapping and interconnecting lenses, if you will - and to see people through all of them, the Myers-Briggs test may end up being a useful way for a person to come to understand the world in a deeper way.

But maybe this is just my P personality shining through. In the meantime, I'm still waiting for results that will tell me about famous chess players, or imaginary Chinese animals!

P.S. As an aside, in some cultures and religions astrology plays a deep and significant role within the overall framework. In these cases there may be a significant point to the activity, even though it doesn’t stand on any scientifically verifiable facts about the world, so I don’t mean to trivialize these traditions in what I said above. Further, I admit I particularly like the astrological story that is associated with my Sanskrit name, and being the sort of guy who likes to read ancient texts and to spend hours identifying stars in the night sky, I am not completely adverse to studying astrology. And I suppose there is a place for a mild bit of superstition. Many completely rational people will suddenly find themselves holding their breaths or crossing their fingers during a baseball game. Or I find that love relationships often get that little extra something from unusual coincidences that bring people together. But now a piece of the F aspect of my personality might be coming through! And this also goes the other way too: Myers-Briggs tests, accurately administered, and for the right purposes may very well be worthwhile. It is just these trivial appropriations that annoy me.