On Philosophy

September 23, 2007

Evidence And The Future

Filed under: Epistemology — Peter @ 12:00 am

Certain skeptical possibilities we can deal with via the principle of epistemic indifference, which states that if we can’t distinguish between two states of affairs then we are justified in assuming that either is the case, and that their supposed differences are completely unimportant, at least as far as we are concerned. This handily does away with the possibility that there are extra parts of reality which do not interact causally with it (at least from our point of view, causation, being a description of events, admits of the possibility that there are other descriptions in which our universe and the extra parts do interact, it is just that the only account of causation we can justify, and which matters to us, does not include them). But there is another kind of skeptical possibility that it is not quite equipped to handle, skepticism about the future. Is it not possible that all emeralds are grue (green before 2020, blue afterwards), not green, as the classic example goes? Clearly all of our evidence supports one just as much as it supports the other. Now we might attempt to apply the principle of epistemic indifference here, pointing out that if we can’t tell the difference between grue and green then the difference must not matter. In one sense this is correct, here and now the difference between grue and green is irrelevant. However, there are ways to distinguish them (by waiting), and thus it is clear that they aren’t the same from all perspectives, just our current perspective.

The answer to such skeptical challenges, in which we entertain the possibility that the future may be radically different than the present for no particular reason, is to deny that the evidence equally supports the hypothesis that emeralds are green and that emeralds are grue. The evidence supports the hypothesis that emeralds are green and contradicts the hypothesis that emeralds are grue. Assuming, that is, that you understand evidence as I do.

As I use the term evidence refers to a falsifiable claim that is thought to be a reflection of objective reality. And I would note that the possibility that the claim is not a reflection of objective reality is itself falsifiable, and so evidence is falsifiable in two ways, we can show that it contradicts other evidence (or claims that follow from other evidence), or we can show that other evidence supports the claim that it does not in fact track objective reality. Our senses and our memories of events are the prototypical examples of evidence. We think that they reflect objective reality and admit that it is possible for them to be in error. Thus it is permissible to reason on the basis of them (and eventually to determine using them when they are in error).

I would consider the general principle, that nothing changes without a discoverable reason behind that change, to be another example of evidence. Obviously this principle seems to reflect objective reality, as what events occur and their relationships to each other certainly seems objective enough. And it is certainly falsifiable, if we were to observe events that could have in no way been predicted, or radical fluctuations in physical law, then we will know this general rule is wrong. And obviously this general rule prevents the possibility that emeralds are grue, because that would be a change in the way emeralds reflect light without any corresponding physical change (or a physical change that doesn’t follow any of the established patterns of physical change).

Obviously this doesn’t rule out the possibility of someone else taking the same principle and adding a single exception for emeralds and their color, and treating that principle as evidence instead. But neither does this treatment of evidence rule out someone taking their senses as completely distorted, and systematically interpreting them as indicating a completely different kind of world from that which seems like a natural interpretation to us, to the extent that their world doesn’t even contain other people. Both such alternate interpretations of evidence are epistemologically equivalent. Assuming that they both admit of falsification (required for them to be seriously considered evidence) then they will eventually converge (although possibly not in a practical time frame) and support the same claims. And we expect that eventually the person who takes the alternate principle as evidence will revise it, when emeralds don’t change color (or we will if they do). Thus in the long run the claims considered justified by people taking either general principle as evidence will converge to the same claims. This is why both positions are epistemologically equivalent (neither deserves condemnation as being a product of bad thinking). We might of course remark that someone who adopts such a principle as evidence is reasoning in an unnatural fashion, but we certainly can’t say that they are wrong for doing so, or are epistemologically in error. After all we can’t say with absolute certainty that they are wrong and we are right, even though we are completely justified in acting as if our claims are true and theirs false.

This I take to be a solution to the problem of induction, when combined with another claim of mine, made elsewhere. The problem of induction is, of course, the claim that we can’t know anything about causal relationships, first raised by Hume (although there are other formulations, but they all are essentially the same, and can be solved by the approach outlined here). The first step is to reject the idea of causation as something that exists in addition to the universe and pushes matter around. Causation is instead best understood as a way of describing patterns of events. Thus we can know about causation by knowing what general laws hold (physical laws being descriptions of patterns of matter and energy over time, and not concerned directly with causation). And we can justify such laws by appeal to the evidence of observed events, which makes them more or less likely than other laws. Of course this by itself only justifies the claim that the patterns described by the laws have held so far. But when we add our new general principle, that things (including patterns of events) don’t change without reason, as evidence we can conclude that we are justified in believing that these laws hold for all times. With that the problem of induction is no longer a problem, at least in terms of what we are justified in claiming, although it does reveal that we can’t know anything about causation with absolute and irrefutable certainty. But that is true in general.

Create a free website or blog at WordPress.com.