Friday, July 20, 2007

Theories, "Truth," and Missing The Point

Michael O'Hare makes an argument (or rather, presents Nelson Goodman's argument) that a model need not be true if it is useful. He gives the example of the Ptolemaic earth-centered universe, which worked just fine for centuries telling people where to look in the sky to make observations.

The point seems to be that if a model is useful, it need not be true. But the geocentric model was abandoned because it was no longer useful, because more and more epicycles and exceptions had to be added to account for the growing number of observations that defied the model. He says:

This is behavioral evidence, the best kind, that he credits the Ptolemaic theory with a utility for a purpose that is not operationally very different from truth.
But the "behavioral evidence" kept mounting that the truth and the model weren't matching up--and hence the model had to be changed.

He concludes:
It's not always necessary to pick one truth from a set of propositions that appear to be inconsistent, and to have what appear to be inconsistent models at hand for different purposes is not the same as post-modernist relativism or mushy contingency.
Um, actually, yes it is necessary. Sometimes we don't have enough evidence to make a decision, as in the early days of quantum physics. Is light a particle? A wave? A particle with wave-like properties? A wavicle? Something completely different? As evidence comes in, one proposition gains support, another fades away. Saying "We don't have enough evidence to choose between two competing models" is much different from saying "We don't need to decide between two competing models."

Of course, this is coming on the tail of a discussion about religion and how much deference/respect (if any) it is due. If his point is that the geocentric model works "well enough" for the astrologers and so we shouldn't laugh at them for using a model that works "well enough" for them, okay. (I laugh at them for completely different reasons.)

Likewise, if his point is "I believe in my religious beliefs because they make me feel better and thus work 'well enough' for my purposes, whether they're true or not," again, okay. At least it's honest.

But scientific theories are "useful" only to the extent that they are true or that suggest experiments pointing toward the truth. Saying "It's not true but it's useful" is equivalent to saying "It's not science." Nothing wrong with that, there are lots of things that aren't science.

But don't call it something it's not.

No comments: