Tell me a story

Tue 04 Oct 2005 09:52 PM

Last May, Carl Sachs asked me what I thought the difference was between a story and a theory. I replied along these lines: A story specifies what its world is like. A theory conjectures what our world is like. Put differently, a theory is a story which we take to be about our actual world.

It now occurs to me that there is an asymmetry between familiar philosophical accounts of stories and of theories: One talks of a story specifying a single world. One asks: What is it like in Sherlock Holmes' London? Of course, the answer cannot be entirely determinate. There are things that the story does not tell us.

According to the semantic conception of theories, a theory can be identified with its models. One might say that the theory is the set of worlds that behave in accord with the theory. The theory is right if the actual world is one of the worlds in that set.

So we are tempted to say that there is the world described by the Holmes stories, although it is described incompletely, but there are many worlds specified when specifying a scientific theory. This difference may simply be superficial. If it is not, then either: (A) A theory is not a story; or (B) it is wrong to think of a theory as the set of its models.

(I suspect the difference is superficial. That, after all, is why this is mentioned in a blog rather than in a paper.)