Originally Posted by stevea
No, that's not tenable. You're assuming that there is a "correct" hypothesis that is correct in some way over and above being one of a set of hypotheses that correctly predict all known data, but that assumption is unwarranted. If we have a hypothesis that predicts all known data, then what other properties do we require for it? If we find, on amassing further data, that there is data that our current hypothesis does not predict, then we must replace it with a new hypothesis that does predict that data in addition to all the already known data; but that, too, is in accordance with Occam's Razor and the scientific method.
So what, precisely, are the properties required of your "correct" hypothesis other than correctly matching all the data?