How does Bell's inequality rule out realism?

[stevendaryl]

I never said that I would argue this way. I just countered atyy's claim that answers to such questions would be implied by physical theories, which is wrong.

It's wrong according to one (in my opinion) very idiosyncratic interpretation of what "consequences of a physical theory" means. It's not wrong according to most people's interpretations of those words. Most physicists would say that our theories of physics do imply what happens when there are no physicists around.

I think it is completely mainstream. It is just rigorous argumentation. Interpretations don't follow from mathematical equations. We need to specify how they are supposed to relate to observable facts. We always emphasize this in introductory courses in theoretical physics, so I consider it completely mainstream. I don't define any arbitrary boundary. The boundary between science and faith is precisely at the point, where questions become inaccessible in principle to the methods of science.

I would agree with the sentences "Interpretations don't follow from mathematical equations. We need to specify how they are supposed to relate to observable facts." I think the rest of it is pretty eccentric, though. You're defining science in such a narrow way that there would be no reason for anyone to care about science. People care about science because they want to understand the universe, not because they want to be able to understand the experimental results of scientists. If you have a definition of science under which it makes no predictions about what is true when there are no scientists around, then that makes science useless, as far as I'm concerned.

 

[rubi]

It's wrong according to one (in my opinion) very idiosyncratic interpretation of what "consequences of a physical theory" means. It's not wrong according to most people's interpretations of those words. Most physicists would say that our theories of physics do imply what happens when there are no physicists around.

It is definitely not implied by the theory. Otherwise you would be able to give a logical proof of a statement of the form "Theory => Interpretation". However, we (including me) decide to use this interpretation, although no scientific argument can be made in favour of this interpretation. That is completely fine. Nevertheless, there is a sharp difference between physics and metaphysics.

 

[stevendaryl]

It is definitely not implied by the theory. Otherwise you would be able to give a logical proof of a statement of the form "Theory => Interpretation".

A theory is not simply a set of equations. A theory is a machine for making predictions of the form: In such and such situation, such and such will happen (or will happen with such and such a probability). So a minimal amount of interpretation is a required part of the theory. Schrodinger's equation is not a theory. Schrodinger's equation together with the claim that "the allowed energy levels of a hydrogen atom are the energy eigenvalues for the corresponding Schrodinger's equation" is a theory.

 

[rubi]

A theory is not simply a set of equations. A theory is a machine for making predictions of the form: In such and such situation, such and such will happen (or will happen with such and such a probability). So a minimal amount of interpretation is a required part of the theory.

That's what I've been saying the whole time. Mathematical equations on their own do not yet relate to physical situations. In order to have a physical theory, you need to specify how the equations relate to observable phenomena. When you have done this, you can test the theory and you might or might not find out that the theory does relate well to observable phenomena. So there is a scientific way to test, whether your theory (equations + relation to observable phenomena) is good or not.

You can also specify, how the theory relates to unobservable phenomena. However, there is no scientific way to test such a specification. Hence, any such specification must be a metaphysical one.

 

[stevendaryl]

That's what I've been saying the whole time.

I thought you were saying the opposite. You were saying that the interpretation was not part of the theory. You were saying that there is no way to prove "Theory => Interpretation".

The issue of disagreement was whether a theory of physics can say what happens in the absence of physicists. I say that yes, a theory can in principle describe what happens in any circumstance you could specify, including what happens after all the physicists are dead. The equations for the evolution of our sun don't depend on the existence of physicists.

Now, you're certainly right that you can't test a prediction of physics unless there is a physicist around to do the testing.

 

[rubi]

I thought you were saying the opposite. You were saying that the interpretation was not part of the theory. You were saying that there is no way to prove "Theory => Interpretation".

By interpretation, I was referring to the the parts that do not directly relate to observable phenomena, such as statements about the existence of a moon that isn't observed or the fate of the univserse after the death of physicists. Personally, I wouldn't consider the relation between equations and observable phenomena to be an interpretation, but this is probably just semantics.

 

[Demystifier]

I don't know any physicist who would count metaphysics to be a part of physics.

I don't know any physicist who, when thinking about physics, does not use any metaphysics. Every physicist tries to understand the meaning of mathematically formulated physical laws, and understanding of their meaning is impossible without metaphysics. In other words, there is no physics without metaphysics. And there is no sharp boundary between them.

 

[atyy]

As DrChinese said, this is a consequence of the theory and not religion. However, statements about the existence of the universe after the death of physicists are not consequences of the theory, so they belong to the realm of religion. Same for the existence of the moon when nobody is looking.

It is not a "pure" unreligious consequence of the theory. If only observations count, you have never observed things at a distance, since you are a localized entity. So for the Bell inequalities to be violated at spacelike separation, you have to make the religious assumption that when Alice tells you her result obtained at spacelike separation, that what she tells you now at non-spacelike separation reflects her past reality.

 

[atyy]

That's the theory, and it is science - not religion. Backed by experimental evidence, not faith.

It is not backed by "pure" experimental evidence. No one has ever seen the Bell inequalities violated at spacelike separation.

 

[Ilja]

Nevertheless, there is a sharp difference between physics and metaphysics.

I would suggest you to study elementary scientific methodology, in particular Popper's critical rationalism, which is essentially the scientific methodology accepted by the mainstream. It does not have a clear and certain boundary between physics and metaphysics.

It contains a criterion to distinguish empirical theories from metaphysical theories. Empirical theories have to make empirical, testable predictions. But this is about theories as a whole, and not about particular statements. So, empirical theories can contain a lot of statements which cannot be tested by any experiment.