Previously I discussed a principle of epistemic indifference, which stated that if it is impossible to tell the difference between two different situations (to have evidence that supports one being the case and not the other) then the differences between them are irrelevant, even though we can’t know which is the case. This is primarily a response to certain hypothetical scenarios that purport to throw the whole of reality into doubt, by supposing it to be a dream, or something similar. By appealing to our principle of epistemic indifference we can respond to such proposals by pointing out that whether the world is a dream or not simply doesn’t matter, and that for the sake of simplicity that we might as well just suppose that it isn’t.
This is not the only principle that can be used to defuse such worries. Another way to get around such problems is to argue that we can know something so long as we don’t have reason to doubt it. Obviously this rids us of the possibility that the world is a dream; no one seriously entertains that possibility (I hope), and no one has reason to believe it, and thus the fact that we can conceive of the possibility that the world is a dream, and recognize that all our evidence supports that hypothesis as much as it does the claim that the world is real, does not prevent us from knowing that the world is real.
Admittedly this principle is just as handy as that of epistemic indifference when it comes to dealing with such fabricated doubts. But it has a few problems as well. The first is that it handles these cases only as long as we are all basically in agreement as to what is the case, and are entertaining these doubts only to discover what we really know. Suppose that we encounter some mystic who really does believe that the world is a dream. If the mystic is as convinced that the world is a fantasy of some sleeper as we are of its reality then the mystic can apply this principle to demonstrate that they know that the world is a dream despite the fact that they can’t produce evidence against the possibility that it is real. If knowledge is strong, and entails truth, then we have a problem, because clearly both claims can’t be true. Even if knowledge isn’t strong the whole situation is a bit ridiculous, because it seems to make what we know dependant on who we have come into contact with, which seems irrelevant to matters of justification. (Assuming that if someone believes in a claim, and we know it, then that gives us reason to entertain the claim seriously, even if we personally don’t have any other motivation to take it seriously. If other people’s beliefs don’t give us reason to take a claim seriously then that makes the second problem, below, that much worse.)
The second, and more serious, problem is that this principle can be extended beyond where it was supposed to apply. Consider a scientific theory that is well accepted. Now let us suppose that someone proposes a different theory, which makes basically the same predictions (and thus is supported by all the same evidence). Obviously if they agreed completely then we would be in one of those situations where the differences doesn’t matter. Thus we will further stipulate that there is a possible experiment which could tell us which theory was correct, even though the conditions in the experiment would never be reproduced naturally. In spite of this proposed rival it would seem like scientists who believe the established theory could still claim that they know that it is true and that the rival is false. Since no one has done the experiment yet they don’t have any reason to doubt the established theory. And since they already know that the established theory is correct there is no reason to perform the experiment. Similarly, consider someone who believes a logically complicated set of propositions about the world. And let us suppose that those propositions contain a subtle contradiction that can be exposed through a twenty step proof. But so long as the person doesn’t actually learn of this proof they can invoke the principle to claim that they know this complicated set of propositions to be true, and then using the fact that they know them to be true to justify never investigating whether a contradiction exists.
Obviously it is possible to improve the principle by making exceptions. For example, we could stipulate that it could only be invoked where every possible effort has been made to distinguish the two possibilities empirically and failed. That would seem to rule out the situations that give rise to the second problem (although not the first). Obviously such an addition is ad hoc, and we could argue that methodologically such fixes are prohibited because every claim can be repaired if tacking on exceptions is allowed, but that is all beside the point. The central problem with the principle is not really any particular situations it might give rise to, but rather that it falsely creates certainty where none is warranted. It is clear that we can’t be certain (can’t know) that the world is not a dream. It is also obvious that we aren’t bothered by this lack of certainty, and that we don’t entertain the possibility seriously. The principle discussed here tries to reconcile these facts simply by denying our lack of certainty, by fiat, and thus explaining our apparent confidence by supposing that we can be certain. And really the solution is just to accept that certainty (and knowledge) is impossible with respect to these matters, and that the fact that we aren’t bothered by that lack simply shows that our lack of certainty is irrelevant in these cases because the differences don’t matter. So my final argument against the principle is not really an argument at all, simply an observation that we are only led to such principles by certain commitments to labeling everything we strongly endorse as knowledge, when really the proper response is simply to drop those commitments.