Provability (that is, being able to be proved by deductive means) is not generally taken to be a requirement for asserting a fact. To prove something is to demo

Provability (that is, being able to be proved by deductive means) is not generally taken to be a requirement for asserting a fact. To prove something is to demonstrate that it could not be otherwise (that it is necessary, given the logic employed and that that the premises are true). For example, I might assert that "Tomorrow, the gravitational constant be the same as it is today." This seems like a fact: we have pretty strong theoretical justification for it, given certain empirical evidence, an assumption of regularity in nature, and the continued predictive success of the general theory of relativity. However, if tomorrow the gravitational constant is slightly different to what it is today, then although no (deductive [1]) logical law has been contravened (as would have to be the case if a provable statement turned out to be untrue), my assertion was clearly not a fact. However, proofs (as in mathematical or logical proofs) are obtained through deductive reasoning, not inductive reasoning, as my statement was. Thus, I would argue, there is nothing wrong with presenting something as a fact which is not provable. Otherwise, we wouldn't be able to assert very much at all: the restricted epistemology obtained would rule out all of science, for starters, and might even cast doubt on certain classes of observation statements. Of course, you may have meant something other by 'provable' than I do, in which case clarification would be appreciated, since you presumably meant something by your assertion that facts must be provable, and it would be nice to know what it was! [1] It is controversial as to whether a general inductive logic can be formulated.