18 February 2017

Against Parsimony

If you can't explain it simply, you don't understand it well enough.

Some things in life, of course, are just plain difficult to understand. Often enough this is their ineluctable nature (and ours); but if a particular subject seems to offer itself up for mastery, try learning as much as you possibly can about it and see if that doesn't thicken the plot just a bit.

The more you know, the more you don't know.

This second piece of popular wisdom carries a very different message from the first: a depth of understanding sufficient to permit the formation of simple explanations is highly destructive of the willingness to accept them; and thus in the ethical sense, I would say, also to dispense them.

Hence there is a third colloquialism which is an essential corollary to the first:

Just because you can doesn't mean you should.

1 comment:

Stefan Kac said...

E.O. Wilson
Consilience (1998)

"Economic theory is not Ptolemaic, not so structurally defective that a revolution in conception is needed. The most advanced of the micro-to-macro models are on the right track. But the theorists have unnecessarily handicapped themselves by closing off their theory from serious biology and psychology, comprising principles drawn from close description, experiments, and statistical analysis. They have done so, I believe, in order to avoid entanglement in the formidable complexities of these foundation sciences. Their strategy has been to solve the micro-to-macro problem with the fewest possible assumptions at the micro level. In other words, they have carried parsimony too far. Economic theories also aim to create models of the widest possible application, often crafting abstractions so extreme as to represent little more than exercises in applied mathematics. That is generality carried too far. The result of such stringency is a body of theory that is internally consistent but little else. Although economics, in my opinion, is headed in the right direction and provides the wedge behind which social theory will wisely follow, it is still mostly irrelevant."

(p. 202)