(Originally posted: 8/3/2011)
This misconception is something that I’ve touched on already in other posts. Its is also more debatable, since there are schools of thought that disagree with my position. However, I think the best way to think about what science is and what it tries to do is to give up on the idea of proving anything, and I also think most scientists who’ve thought about the issue would agree with me. An explanation follows the jump.
What do I mean when I say that it’s a misconception that scientists prove things? Simple: we don’t. Science is the process of showing that certain explanations are wrong, not that they are right. This idea traces back to the phiosopher of science Karl Popper, who popularized the idea of falsification. The approach to science boils down to this, more or less:
Science is supposed to be interested in explaining the objective world. However, our knowledge of the objective world is incomplete (otherwise, we’d have no need to explain it). To say that a particular idea has been proven to be true, we must have 100% confidence that it reflects the true nature of that objective reality. The corrollary of this is that we must have 0% confidence in all other statements, or rather 100% confidence that all other statements are false. So, in other words, to ever really prove anything, we must completely disprove all other possible explanations. Not just the ones that we can think of, or the ones that have been proposed before. Remember, our knowledge is imperfect, and there may be (in face, almost certainly are) explanations that have not been dreamed of yet. How can we ever disprove ideas that don’t even exist yet? But to prove anything, we have to!
Instead, what scientists do is set out to disprove ideas (either hypotheses or theories). It’s easy to disprove a statement. If it predicts circumstances in the observable world that are directly contradictory to what you see, you can usually consider it disproven! To say “I have 0% confidence that Statement A is true” carries no implications for the truth of Statement B!
To ground this a bit more, consider this example: In conventional wisdom, it has been proven that atoms exist. (This, despite the fact that they are theoretical constructs, as I’ve discussed in an earlier post.) Why? Well, experiments have been done, and the atomic theory of matter has held up to those tests. But can we really say, with 100% confidence, that the theory describes the real state of the universe? Is there not at least a infinitesimal chance that, say, a genie has manipulated the results of every experiment? Or that a vast cabal of scientists has secretly conspired to falsify every report? Sure, the likelihood is vanishingly small, but it is nonzero. And in that case, we cannot have 100% certainty in the alternative!
So, when all is said and done, what do scientists do? They gather as many possible explanations as they can–both plausible, unlikely, and just downright silly–and then begin discarding them. An experiment is an attempt to disprove a theory, not to prove it! When all the experiments are over, and all the disproven theories are discarded, whatever’s left is tentatively accepted as probably being true.
This, unfortunately, does not make for an easy process. There may be several competing (perhaps even mutually exclusive) theories that cannot be disproven. The lone theory that has survived may be thoroughly unsatisfying. And even if the theory that survives is satisfying, powerful, and widely accepted, it may also be wrong. Just because we’ve failed to disprove it doesn’t mean it must be correct! After all, humans are fallible!
So, the ultimate conclusion here is that scientists don’t prove things true. We fail to prove things false, and reluctantly agree to accept that theory as true, pending future attempts. Scientists should be happy when they disprove a theory, because that means that one more red herring has been eliminated, and we’re that much closer to the truth. The unfortunate drawback though, is that we’ll never really know when we get to the truth, because we have no way to distinguish a true theory from a false one that hasn’t been disproven yet!
On the other hand, there is something good to be said for this perspective on science: it means that scientists will always have a task ahead of them–the job is never done!