The Universe is Deterministic

[sol invictus]

"Stochastic" usually means there's some non-determinism in the system, but many results might be rather predictable anyway (the average, for example).

I think the correct distinction to draw here is between deterministic (the future is completely determined by the present) and non-determinstic. Words like "random" just get in the way.

 

[arthwollipot]

"Stochastic" usually means there's some non-determinism in the system, but many results might be rather predictable anyway (the average, for example).

I think the correct distinction to draw here is between deterministic (the future is completely determined by the present) and non-determinstic. Words like "random" just get in the way.

Agreed, which is why I didn't use it.

 

[Singularitarian]

I've created this thread because I disagree with these last two statements completely (and didn't want to derail the thread they were posted in).

While it may be impossible to prove, given our current understanding of physics, whether or not the universe is deterministic or not, there are no compelling reasons to therefor conclude that it isn't. All things in our experience have causes -- all of them. So why then is the prevaling opinion that the universe nondeterministic?

The general theories of relativity are mostly accepted to show indeed that the universe is determined. Somehow events themselves are written into the fabric of space and time.

You are absolutely right. You should read ''the frozen river,'' by Dr. Greene.

 

[edd]

GR is not a theory of everything.

 

[sol invictus]

The general theories of relativity are mostly accepted to show indeed that the universe is determined. Somehow events themselves are written into the fabric of space and time.

GR is a classical theory. Like every classical theory, it's deterministic (except at singularities where it breaks down).

But GR does not describe the world, because GR does not include quantum mechanics (or the other forces of nature for that matter, although it's not a problem to add them at the classical level).

One thing GR does tell us is that interactions should be local. That closes off one of the loopholes in Bell's theorem and proves just the opposite of what you say: that the universe must be non-deterministic, or one must give up something really basic about reality.

 

[Paul C. Anagnostopoulos]

A question: If we give up locality, do we introduce retrocausality? Even though two particles may interact instantaneously, the no-communication theory says that you can't use this to transmit information. And if you can't transmit information, there is no issue with retrocausality. Yes? No?

~~ Paul

 

[sol invictus]

A question: If we give up locality, do we introduce retrocausality? Even though two particles may interact instantaneously, the no-communication theory says that you can't use this to transmit information. And if you can't transmit information, there is no issue with retrocausality. Yes? No?

If you want to make the universe both causal and consistent with QM, Bell proved you must do one of two things:

(1) allow instantaneous interactions between particles. Any such theory can and will allow transmission of information faster than light. Therefore if relativity is correct, it can also be used to transmit information back in time, and the theory is not consistent. If relativity is not correct it might be OK.

(2) give up a more fundamental postulate about reality; for example, give up the idea that there was only one result to the experiment. My favorite example, and the one I think is probably correct, is the so-called "Many Worlds Interpretation".

Is the MWI deterministic? It is in the sense that the wavefunction (which is the full description of the world) evolves deterministically. But on the other hand it is (almost?) completely impossible to predict the result you will observe in some experiments. So in any functional sense it is non-deterministic. Perhaps it's best thought of as analogous to chaos: deterministic but not fully predictive.

 

[Paul C. Anagnostopoulos]

(1) allow instantaneous interactions between particles. Any such theory can and will allow transmission of information faster than light. Therefore if relativity is correct, it can also be used to transmit information back in time, and the theory is not consistent. If relativity is not correct it might be OK.

Then why does Wiki say:

Bell test experiments to date overwhelmingly show that Bell inequalities are violated. These results provide empirical evidence against local realism and in favor of QM. The no-communication theorem proves that the observers cannot use the inequality violations to communicate information to each other faster than the speed of light.

~~ Paul

 

[sol invictus]

Then why does Wiki say:

"Bell test experiments to date overwhelmingly show that Bell inequalities are violated."

That means local deterministic "real" theories are ruled out by experiment.

"These results provide empirical evidence against local realism and in favor of QM."

Because QM is not such a theory.

"The no-communication theorem proves that the observers cannot use the inequality violations to communicate information to each other faster than the speed of light."

QM does not allow information to be transferred faster than light. But QM is either non-determinstic or not "real", depending on whether you choose the Copenhagen interpretation or many worlds.

What would allow superluminal info transfer is any deterministic non-local theory that you might try to replace QM with.

 

[Paul C. Anagnostopoulos]

If you want to make the universe both causal and consistent with QM, Bell proved you must do one of two things:

(1) allow instantaneous interactions between particles. Any such theory can and will allow transmission of information faster than light.

...

QM does not allow information to be transferred faster than light.

These two statements appear to me to directly contradict each another. I'm sure it's a failure of understanding on my part. If it's to be consistent with QM, it cannot allow superluminal transfer. But then (1) says it does.

Could you explain post #89 again, but use more words?

~~ Paul

 

[wuschel]

"Bell test experiments to date overwhelmingly show that Bell inequalities are violated."

Would not such experiments always carry the underlying prior assumption of a source of true randomness?

 

[The Man]

If you want to make the universe both causal and consistent with QM, Bell proved you must do one of two things:

(1) allow instantaneous interactions between particles. Any such theory can and will allow transmission of information faster than light. Therefore if relativity is correct, it can also be used to transmit information back in time, and the theory is not consistent. If relativity is not correct it might be OK.

(2) give up a more fundamental postulate about reality; for example, give up the idea that there was only one result to the experiment. My favorite example, and the one I think is probably correct, is the so-called "Many Worlds Interpretation".

Is the MWI deterministic? It is in the sense that the wavefunction (which is the full description of the world) evolves deterministically. But on the other hand it is (almost?) completely impossible to predict the result you will observe in some experiments. So in any functional sense it is non-deterministic. Perhaps it's best thought of as analogous to chaos: deterministic but not fully predictive.

There is of course the Transactional Interpretation of Quantum Mechanics that is deterministic (although in a more time symmetrical fashion), non-local and does not allow for superluminal info transfer, but of course it has its own problems like free emission and self interaction. As Sol notes these interpretations generally force you to give up some intuitive description, whether it be determinism, “Real” (as Sol puts it) or free emission and self interaction. That is not to say of course that we may not eventually find a better interpretation, but for now this is what we have to work with.

 

[sol invictus]

These two statements appear to me to directly contradict each another. I'm sure it's a failure of understanding on my part. If it's to be consistent with QM, it cannot allow superluminal transfer. But then (1) says it does.

Could you explain post #89 again, but use more words?

I'm sorry, it's my fault. The sentence "If you want to make the universe both causal and consistent with QM, Bell proved you must do one of two things:" should read "If you want to make the universe both deterministic and consistent with QM, Bell proved you must do one of two things:"

Does that clear it up?

 

[sol invictus]

Would not such experiments always carry the underlying prior assumption of a source of true randomness?

No, not at all. Experiments can't assume things, they just collect data.

Bell proved that any theory in which (say) the two particles have a definite spin before they are measured gives results that differ from QM, unless there is instantaneous communication between them (or you give up "reality"). Since then, those predictions of QM have been confirmed, thereby ruling out any such theory. So Bell didn't assume anything either, he just contrasted the predictions of QM in some specific experiment to those of any deterministic, local, "real" theory.

And in fact it actually goes far beyond proving the source must be random, it says the particles the source produces cannot even have a definite spin.

There is of course the Transactional Interpretation of Quantum Mechanics that is deterministic (although in a more time symmetrical fashion), non-local and does not allow for superluminal info transfer,

It does allow for superluminal info transfer. Not even its creator disputes that, as you can see e.g. here, section IV. Cramer thinks that's OK and doesn't lead to inconsistencies, but I think he's wrong.

Also one thing to bear in mind in all this is that we already have a relativistic formulation of QM that works extraordinarily well, and it's not consistent with the transactional interpretation. I don't see any mention of quantum field theory anywhere on Cramer's site (perhaps I missed it), and that's extremely suspicious. QFT is the most successful theory in the history of science and it bears directly on this question; it cannot be simply ignored in these discussions.

 

[wuschel]

No, not at all. Experiments can't assume things, they just collect data.

...

And in fact it actually goes far beyond proving the source must be random, it says the particles the source produces cannot even have a definite spin.

I guess you misread me.

I wasn't referring to the randomness - or lack thereof - of the stuff being measured. Rather, I had in mind the measurement itself. The underlying assumption is that e.g. two supposedly arbitrary and independent measurements are indeed independent and arbitrary.

In a fully deterministic universe everything would follow "rules", even the measuring equipment, even if these "rules" are not entirely obvious.

The resulting observations would differ from expectation of "realism" in as far as there would develop the equivalent of a "Moire pattern".

Forgive me if this sounds a bit weird, but then, this is QM! ;-)

 

[sol invictus]

I wasn't referring to the randomness - or lack thereof - of the stuff being measured. Rather, I had in mind the measurement itself. The underlying assumption is that e.g. two supposedly arbitrary and independent measurements are indeed independent and arbitrary.

No such assumption is needed.

In a fully deterministic universe everything would follow "rules", even the measuring equipment, even if these "rules" are not entirely obvious.

That's precisely what Bell proved impossible. Or rather, he proved that any such theory satisfies a certain inequality, no matter what those "rules" are, and further, that QM violates that inequality. The only loopholes are as I've said: either one measurement can instantly affect the other or there is more than one result of each measurement.

It's a very powerful theorem.

 

[wuschel]

The only loopholes are as I've said: either one measurement can instantly affect the other.

Would that still be a requirement, if instead measurements were not truly independent to begin with (I use "entanglement" as a figurative analogy?)

What if our Universe does not allow us (on the quantum scale) to conduct two truly independent measurements?

 

[Paul C. Anagnostopoulos]

I'm sorry, it's my fault. The sentence "If you want to make the universe both causal and consistent with QM, Bell proved you must do one of two things:" should read "If you want to make the universe both deterministic and consistent with QM, Bell proved you must do one of two things:"

Does that clear it up?

Let me see if I understand. If we accept QM and Bell and want a deterministic universe, we have two choices:

(1) Allow instantaneous particle interactions, which allows superluminal transfer of information.

(2) Give up some fundamental postulate about reality.

Assuming I have that right, I have some more questions:

The no-communications theorem only comes into play when we haven't selected (1), right?

Does superluminal information transfer have any practical use? It would have to be the case that we could entangle particles that were already far apart, right? Otherwise we have to wait for them to travel apart before doing the transfer, in which case we haven't saved any time.

Is there a current concensus? Are we going nondeterministic, (1), or (2)?

~~ Paul

 

[sol invictus]

Would that still be a requirement, if instead measurements were not truly independent to begin with (I use "entanglement" as a figurative analogy?)

What if our Universe does not allow us (on the quantum scale) to conduct two truly independent measurements?

The detectors can be 20 lightyears apart when they measure the particles, and they could have been manufactured on two opposite ends of the universe and then brought in from very far away until they meet them. And yet, if one measures spin up the other must measure spin down (conservation of angular momentum), but the particles themselves could not have had definite spin beforehand (else Bell's inequality would not be violated).

So the only way out is if the measurement at one can determine the measurement at the other, but that requires faster than light influence.

Let me see if I understand. If we accept QM and Bell and want a deterministic universe, we have two choices:

(1) Allow instantaneous particle interactions, which allows superluminal transfer of information.

(2) Give up some fundamental postulate about reality.

Right.

Assuming I have that right, I have some more questions:

The no-communications theorem only comes into play when we haven't selected (1), right?

The no-communication theorem I know of applies to QM and only to QM. In the standard interpretation QM is non-deterministic, hence neither (1) nor (2) holds. In the MW interp. it's deterministic, but (2) holds.

Some proposed extensions of QM (like the transactional or Bohmian "interpretations") add extra laws of physics to QM that make it deterministic, and in those (1) holds.

Does superluminal information transfer have any practical use?

It would almost certainly render the universe inconsistent with the rules of logic. I'm not sure I'd call that a "use".

It would have to be the case that we could entangle particles that were already far apart, right? Otherwise we have to wait for them to travel apart before doing the transfer, in which case we haven't saved any time.

All particles are entangled to some degree with many many others all the time. It's just not usually detectable.

Is there a current concensus? Are we going nondeterministic, (1), or (2)?

My impression is that most physicists that have an opinion go with (2), specifically with MW.

 

[Paul C. Anagnostopoulos]

Thanks, Sol!

It would almost certainly render the universe inconsistent with the rules of logic. I'm not sure I'd call that a "use".

Could you expand on this a bit?

~~ Paul