> Roger Odisio:
> > ...
> > And I disagree with Laibman. Are you sure that's what he said? Did he
> > ever write that anywhere? Silly may be too kind of a word for such a
> > statement.
>
> No, it's quite correct. For an observation to refute a
> theory, there must be a theory as to why the observation is
> important and therefore refutes. As Einstein said, "theory
> tells us what we can observe." Phenomena must be ordered
> before they can mean anything. And so on.
But according to Doug, Laibman said you can refute a theory *only* with another theory. I took that to mean facts are irrelevant and I suspect that's why Doug called his statement silly. Upon reflection however, that interpretation doesn't sound like Laibman to me. He certainly understands, I think, the interactions necessary between fact and theory during verification, some of which you mention.
So I'm going to punt.
RO