What can science prove?

[Frank]

The statement that "all fish now living on earth eventually die" is neither scientific nor conclusive, if science has anything to do with observable results (and I'm pretty sure it does). The only way that statement could be considered scientific and conclusive is for all fish now living on earth to be under controlled observation until they die. Although that's possible in theory, it's not feasible in practice because of the sheer number of (i) fish that would have to be found and (ii) tracking devices built—even if we restricted it to just tuna fish.

That's a fair point, but I predicated the deductive argument with a qualifier (e.g. "assuming the fundamental laws of nature remain unchanged"). This is a built in assumption with any conclusion reached by science. I've never heard a scientist qualify a discovery by saying something like, fish are aquatic species, and biological organisms, and as such (according to the laws of nature) they will necessarily die; but I can't say that for sure because the laws of nature could change tomorrow (or we might not exist at all - and this could all be an illusion).

Nonetheless, if you like, I turn to my other hypothetical of steel. When the guy who invented steel first came up with the idea of combining iron, coke, nickel, and other minerals in a smelting process to create a stronger and lighter metal, there was obviously a chance it wouldn't work. But once he was able to successfully produce steel, and show a method by which we can consistently manufacture steel, the formula for steel became a conclusively proven scientific fact (not an almost proven scientific fact). It may be true that occasionally producers don't get it right, or for some reason they make a bad batch (or whatever); but that doesn't change the validity of the formula for steel. Even if there may be other ways to make steel, or we devise improved ways of producing steel (or better types of metal), it doesn't matter. It doesn't change the validity of that formula as a conclusively proven method for producing steel.

I'm not saying we can't sit around and imagine all sorts of ways everything could change (or whether its possible that somewhere in the universe the laws of physics and chemistry are different). What if a wormhole pops into existence (it's possible right, even if there's only a one in several trillion chance of it happening). Or maybe we're all a figment of a giant globs imagination ... or whatever. But science operates under an assumption that we do exist and the laws of nature won't arbitrarily and unrecognizably change spontaneously (unless you're a quantum physicist and we're discussing fun stuff like virtual particles).

So yes you can point to something like the possibility that we don't really exist; but scientists generally accept that as a matter of a priori knowledge, and that does not dilute the conclusiveness of certain proven facts (like the chemical formula for steel), provided we're operating under the conditions the formula was designed for (e.g. a steel plant on earth).