The Dolors of Psychology

Margaret Wente, my favorite columnist at the Globe and Mail, has an excellent column on the recent travails of social psychology: "Social psychology's credibility crisis". I mention this because this bears out a view I came to personally some twenty years ago. After noticing that every time I read a book on neurosis it seemed to make me more neurotic, I decided that perhaps there was something deeply wrong with modern psychology. Deciding it was like a religion, or perhaps a cult, I simply stopped believing in it and stopped using the associated technical terms like "ego", "id", "subconscious" and so on. I'm not denying any actual facts here, just the weird theories scaffolded around them. As a result, I no longer have neuroses! Heh.

Anyway, back to Margaret's column:

The bombshell came last August, after an inquisitive scientist at the University of Virginia, Brian Nosek, launched something called the Reproducibility Project. It was designed to test how reliable psychology studies really are. Hundreds of researchers replicated 100 published psychological experiments to see if they could get the same results. Their success rate was a dismal 36 per cent. There’s been a lot of squabbling over those results. But the Reproducibility Project was just the latest in a string of doubts and scandals that have plagued the field. And now, a lot of people, including scientists who have invested years of their professional lives in these theories, feel as if they’re in a train wreck.

Yep, 64% of the studies they examined are simply ... wrong. I don't know about you, but I demand a slightly higher level of accuracy. When I gave concerts, the standard of accuracy was probably 99% and even that one percent some audience members would notice!

Let me link this to a recent post on music. The post on polymeter I just put up a few days ago cited a couple of examples where the tension between the two meters was what drove the piece forward. Now, according to psychology, this isn't possible because they believe that in any case where there are two different meters or textures or whatever, one will inevitably be perceived as "figure" and the other as "ground". This is the classic example:

This will be perceived as either a vase or two people facing one another in profile, but not both simultaneously. But music doesn't work like that. In both of the examples I cited it is the tension from hearing both of the layers that gives the music its drive. Music simply is not analogous to the visual environment.

Here is another example from Margaret's essay:

The prevailing theory is that willpower is a finite resource, like a muscle that can be exercised to exhaustion. This is among the most widely accepted findings in social psychology. But is it true? Maybe not. Prof. Inzlicht was part of a replication team that chose one classic experiment on the depletion theory of willpower, and repeated it in 24 different labs. And what they found was – nothing. Only three of the labs produced any significant result at all, and one of those was negative.

Psychology often seems to rely on some crude analogy and then cobbles up some statistically smelly studies to "prove" it. Why not think of willpower as being a virtue of habit as Aristotle would have characterized it? You learn it by observing a model and become good at it with practice until it becomes a habit--oddly enough, just as a muscle becomes stronger with use.

I'm pretty leery of everything science tends to claim about music and aesthetics--but I am twice as leery if it comes from the pseudo-discipline of modern psychology. Psychology, by the way, was invented by Aristotle in the 4th century BC and his psychology is still pretty good.

Let's listen to a little music to close. How about a bulerías by Sabicas illustrating the drive coming from the skillful use of polymeter:

UPDATE: I changed the clip for a different one.