>>7
Suppose a model fits extremely well with the data, statistically speaking. So we are inclined to accept it, fine.
But there are several places where this relies on at least some level of subjective human experience. Each of the scientists must share the common interprative experience of realising that the data fits the model. Merely understanding a number relies on representing the concept of that number in some way in our brains - something we don't have a full understanding of.
Similarly, what is this inclination, once that realisation has taken place, to accept the model? It is based on logic, sure. Falsifiability and all that. Again I'd contend that we cannot possibly posit that logic as a universal structure because it is simply something we discuss with each and seem to comunally understand with each other.