In the thread Rationality and Objectivity, Stephen Speicher wrote:
There is a distinction between that which is metaphysically possible and that which is epistemologically possible.
It is metaphysically possible for a pig to have wings of such a nature that the exercise thereof would enable the pig to fly. I suppose one might also say that, metaphysically speaking, it is possible that a pig could possess some other form of propulsion system.
But, if we include into our epistemology all things metaphysically possible, then, apart from axiomatic statements and perceptual knowledge, we can never be certain that any of our knowledge is, in fact, knowledge. Such a standard would ignore the objective nature of knowledge, in the manner in which I previously discussed truth as a relationship between a proposition and the facts of reality. Either we have knowledge, or we do not. We cannot impose a standard of omniscience upon knowledge, since man is man and not some ghostly god. All things metaphysically possible is not the same as all things epistemologically possible.
I want to first thank Stephen for raising this issue. He does make several crucial identifications, but I think he errors in distinguishing between the metaphysically possible and the epistemologically possible.
What Stephen has identified is not a distinction between metaphysical possibility and epistemological possibility, where “metaphysical possibility” must be excluded from our epistemology. Instead, he has identified a distinction between potentiality and the exercise of potentiality. A pig has the potential to have wings and fly – that is a proper identification based on the nature of the pig, the nature of wings, the nature of air, etc. But is it possible that pigs have wings and fly? No! There is no evidence to support that claim. It is completely arbitrary.
MinorityOfOne addressed this point in a blog entry concerning skepticism – skeptics jump from the observation that it is possible for man to error, to the conclusions that it is possible man is always in error. But that is a non sequitur – just because it is possible for man to error (i.e., just because he has that potential), is not evidence that he is in fact in error (i.e., that he has actualized that potential).
In the same way, it is possible for pigs to have wings and fly, but it is not possible that they have wings and can fly. There is evidence for the first proposition (the evidence being: knowledge about the relevant identities); there is no evidence for the second. In neither case are we permitted to retreat into the arbitrary.
This points to Stephen’s error: There is no need to exclude knowledge of potentialities from our epistemology. There is a vital need, on the other hand, to recognize that knowledge of a potentiality is not evidence that potentiality has been actualized.