Let’s try. I write down a sequence of numbers:
1 2 3 4 5 6 7 8 9 10 11 12 ?
What is the next number?Did you mean 13? wrong. It’s 1. For the scheme that I had in my head were the hours of the day, and after 12 comes again 1.
This is a simplistic example, but the basic idea behind science is the same.The universe constantly torpedoes us with experiments, all the time. And every single one of these could be the one that breaks the scheme, no matter how carefully you’ve studied it and how well it has worked so far.
This is called the sub-determination of empirical theories by the evidence.Other good articles on the subject are Underdetermination of Scientific Theory or The Subdetermination of Empirical Theories by Evidence.Until you have studied the whole universe, in all areas of space and time, one cannot seriously say beyond doubt that there is no contradiction out there.
Gravity is a great example.We thought for about 200 years that we had a really good theory that works great. Then came Einstein, who found circumstances in which this theory does not work. They searched for evidence and found it. We could say that the old theory has been replaced, but remains as a special case, or various other formulations of the same idea, but the central lesson is that even something as obviously universal as gravity ultimately falls under some extreme may be incorrect.
In practice, however, this distinction is usually purely semantic.Theories are not proven, but “accepted,” assuming that we may one day withdraw this acceptance in the face of new data. But that ignores that a good theory has been tested many, many times and it would be hair-splitting to say that it could actually be incorrect, especially if you are not actively working on a better theory that can also explain how the old theory is so could survive for a long time.
So this unprovability is not something that should be taken seriously outside the outermost edges of a field.We’re making real progress because a theory has proven itself – and that’s what makes a theory useful. A theory that makes assumptions that might be wrong but arenot, tells you something you didn’t know before. A theory that could never be wrong, no matter what experiment you do, makes no real predictions about the experiment.
And that means that all this is pretty okay and real scientists don’t often care about that detail.It is a question for the philosophers of science and as soon as something happens that is useful, it would be nice if they let us know about it. Until then… It’s better to leave it there.
tl;dr from the translator of the answer:
We talk in science about theories when we have specific ideas about gravity, evolution, etc.because it is never certain that we will not have a better idea of how this works. Gravity was once a more limited concept of Newton, which had to be expanded and is therefore no longer the concept of that time. That doesn’t mean he wasn’t right in his world. All scientific ideas are human ideas of how reality works. We cannot learn a divine truth about how gravity works EXACT, so good ideas about it, probably approximations, must suffice – and they are necessarily variable and/or inaccurate.