Drakkith said:
How would you ever know that your answer is absolutely, 100%, correct?
That's what I see as the issue. You can make all the predictions you want, but you can never know whether your interpretation about what is going on is true or not.
Fair enough.
No doubt it is extremely useful to be able to predict with as much accuaracy as possible. In fact, all of technology is based on that. And, questions about why and, in rigor, what, are naturally outside such a "craft of prediction".
However, I don't see why we should constrain Science to only be such a craft.
The official argument (which you just presented) is that the matching ratio of predictions is the metric which allows the results of "such a craft" (I don't want to call it Science for I believe it ought to be consider a part of it) to be given a certainity. We can't be 100% sure of any physical explanation but we can be X% sure, where the fact that X is an (approximatedly) known figure what matters.
That makes sense, but so does locking up in my room because I can never know if I will be hit by a car, or some such, if I leave. So, is it really the case that Science cannot do better?
IIUC it was at the dawn of quantum physics where we (perhaps I should say you, I am not a physicist :) decided to give up and be happy with just making predictions, but, has such a position real epistemological value, or is it just a reaction to the ongoing failure to make any
sense of the microscopic world? According to an article on Scientific American that I was reading last week, we don't even really know what the quantum field theory actually says, least what it means.
But, if not Science who is going to answer the fundamental question of what is really going on? it's great that we can make quantum computers, atom teleporting machines, spacetime bending warp travel ships, and all that, but that's technology and Science was never just the fuel of that.
So, what is better, not having absolutely any clue about a given thing, or having an hypothesis for which its certainity level has to be given by a metric which is not at all based on matching predictions. One which in fact cannot be related to any prediction at all?
You are of course right that all that matters is whether we know or just believe, for it is that which separates Science from everything else, but giving up on how to test it just because we haven't figure out how to do it yet, except for the narrow case of predictions against experiments, is leaving a big hole in knowledege that just can't and won't be filled by anything else.
I would argue that matching predictions cannot be the only way to asses the certainity of an explanation, and, if such a certainity can be qualified in some other way, then it is a scientific explanation.