Is QM Deterministic in MWI and Time Reversible?

[entropy1]

Is QM causal? Specifically, is it time reversable? Does that equate determinism? Specifically, in the case of MWI?

Edit:

Is QM deterministic? Specifically, in the case of MWI? Is it time reversable?

 

[A. Neumaier]

Is QM causal? Specifically, is it time reversible?

The second question is not at all related to the first.

 

[entropy1]

The second question is not at all related to the first.

Yes. I probably should replace that by "determinism".

Is QM deterministic? Specifically, in the case of MWI? Is it time reversable?

 

[vanhees71]

Quantum mechanics is causal and indeterministic. Time-reversal invariance is not accurately realized in nature since it's empirically shown not to hold in processes involving the weak interactions. In fact the weak interactions violate C, P, T, and CP symmetry but within the accuracy of current measurements not CPT (the latter finding would be a very big surprise since it would mean that we have to consider physics not only beyond the standard model but beyond local relativistic QFTs).

 

[PeterDonis]

Moderator's note: Thread moved to QM foundations and interpretations forum.

 

[WWGD]

Yes. I probably should replace that by "determinism".

Maybe you can go back in time and rephrase the question to test this ;)?

 

[entropy1]

Maybe you can go back in time and rephrase the question to test this ;)?

Well, I guess the question is so bad, that it'd better not be answered. So QM is deterministic. There.

With "causal" I ment inference-like. But Wikipedia held a different conception.

In this world I haven't started my long desired study physics. So next life.

 

[atyy]

Whether QM is causal or not is a matter of definition and interpretation.
https://arxiv.org/abs/1503.06413https://arxiv.org/abs/1602.07404

 

[Demystifier]

Is QM deterministic?

It's interpretation dependent.

Specifically, in the case of MWI?

MWI is deterministic.

Is it time reversable?

It's interpretation dependent. MWI is time reversible (up to certain effects of weak interactions, which are irrelevant in the present context).

 

[vanhees71]

Of course it depends on how you define "causality" and "determinism". I'd say the most lucid discussion can be found in the introductory chapter (Prologue) of Schwinger's "Quantum Mechanics":

Causality: The state of a system of time $t$ is determined by the state at previous times. In QT it's even the more strict causality being "local in time", i.e., the state at time $t$ is determined by the state at one previous point in time.

Determinism: All observables take determined values, independent of the state of the system. This is obviously not the case in QT, but the state describes the probabilities for measurement results of observables. There's no state, where all possible observbles of a system take a determined value.

Thus QT is causal (as any sensible physical theory must be, because why should it be possible to do physics if there are no laws to find?) but indeterministic.

Concerning time-reversal symmetry, I'd say it's a matter of observation, whether nature obeys a symmetry or not. As long as weak interactions can be neglected the fundamental laws are time-reversal invariant, while the weak interaction breaks this symmetry. Of course that's no contradiction to any fundamental law of Q(F)T.

 

[entropy1]

Determinism: All observables take determined values, independent of the state of the system. This is obviously not the case in QT, but the state describes the probabilities for measurement results of observables. There's no state, where all possible observbles of a system take a determined value.

What if the state is deterministic too?

 

[A. Neumaier]

Causality: The state of a system of time t is determined by the state at previous times. In QT it's even the more strict causality being "local in time", i.e., the state at time t is determined by the state at one previous point in time.

These are special cases only, for systems that have a specific notion of state.

The general principle of causality is that changes in a system lead to changes in the responses only at later times, in every admissible frame of reference. In short: effects appear later than their causes. This is a necessary condition for any meaningful use of physics.

 

[entropy1]

In short: effects appear later than their causes. This is a necessary condition for any meaningful use of physics.

That was I thinking: if QM is causal, then you could write this causality logically as A -> B. But if A -> B, then NOT-B -> NOT A. So that seems close to retrocausality to me! If the universe is causal, then it can be retrocausal to!

 

[A. Neumaier]

That was I thinking: if QM is causal, then you could write this causality logically as A -> B. But if A -> B, then NOT-B -> NOT A. So that seems close to retrocausality to me! If the universe is causal, then it can be retrocausal to!

This is nonsense! Causality has nothing to do with logical implication. The latter is indepedent of any notion of time.

Negating logical implications do not reverse the time direction: Suppose that change A made in a system has effects B. Then there is no way to make the effect not-B without prior to it making the change not-A. Thus causality is preserved.

 

[vanhees71]

These are special cases only, for systems that have a specific notion of state.

The general principle of causality is that changes in a system lead to changes in the responses only at later times, in every admissible frame of reference. In short: effects appear later than their causes. This is a necessary condition for any meaningful use of physics.

Sure, I referred to the special case this question was about, i.e., quantum theory. The only question is, why you think it's special? Because there's no satisfactory quantum theory of gravity?

 

[A. Neumaier]

Sure, I referred to the special case this question was about, i.e., quantum theory. The only question is, why you think it's special? Because there's no satisfactory quantum theory of gravity?

1. For quantum gravity we do not even know whether there is a deterministic dynamics for the state.

2. In the quantum theory of open systems, where we know a lot, one may have stochastic Schrödinger equations, where the state is determined only up to noise in the Hamiltonian.

3. On a more basic level, cause (change of the system) and effect (change of the response) refer to the results of preparation and measurement, and only the first but not the second is covered by the notion of a state.

 

[vanhees71]

Ad 1. Yes, so we don't know, whether there's a physical description of quantum gravity possible yet.

Ad 2. Sure, if you choose to describe open systems, often a stochastic description can provide a sufficient effective description. Here however we choose not to describe the complete dynamics (for whatever reason; in this case it's because of complexity) and substitute it by some stochastic description. This does of course not imply that the fundamental laws of physics are acausal or stochastic.

Ad 3. What has this to do with the question whether QT is causal or not?

 

[A. Neumaier]

Ad 1. Yes, so we don't know, whether there's a physical description of quantum gravity possible yet.

Ad 2. Sure, if you choose to describe open systems, often a stochastic description can provide a sufficient effective description. Here however we choose not to describe the complete dynamics (for whatever reason; in this case it's because of complexity) and substitute it by some stochastic description. This does of course not imply that the fundamental laws of physics are acausal or stochastic.

Ad 3. What has this to do with the question whether QT is causal or not?

Ad 1. because of black hole thermodynamics, quantum gravity is possibly intrinsically open, hence covered by 2. rather than 1.

Ad 2. If you argue that fundamental laws are necessarily described by a deterministic dynamics of a state, you need to take into account that no system is truly (i.e., fundamentally) closed, except for the universe as a whole. Bu you always claimed that a state for the universe is meaningless. How does that square with each other?

Ad 3. Causality is about the relations between cause and effect, and effects are described by measurement results, not by states. But according to the traditional interpretations, the Schrödinger equation only describes effects on the state, not on measurement results. Thus the causality of the Schrödinger equation is not a fully convincing replacement for causality in quantum mechanics.

 

[vanhees71]

Ad 1. Yes, maybe after all we have to give up physics as we know it, and causality is only an apparent approximate property. Who knows?

Ad 2. In this sense yes, there are no closed systems.

Ad 3. That's why I do not embrace the collapse hypothesis. Also a measurement is just described by the fundamental laws as any other interaction. After all a measurement device is nothing else than a piece of matter made of the the standard-model particles as any other thing around us.

 

[A. Neumaier]

a measurement is just described by the fundamental laws as any other interaction. After all a measurement device is nothing else than a piece of matter made of the the standard-model particles as any other thing around us.

But you predict the results of measurements using a simplified description based on the system alone, and not the fundamental laws for large pieces of matter made of the the standard-model particles. Because then you would end up with the measurement problem in its full difficulty.

 

[vanhees71]

Sure, I use effective descriptions of measurement devices and experimentalists use these descriptions even to build them. That's, however no contradiction to the statement that this effective description is just an approximation of the underlying "microscopic" dynamics which one cannot resolve anyway and which you don't need to resolve. To describe how a silicon-pixel detector works you don't need to describe the detailed interactions of a particle hitting $\mathcal{O}10^{24}$ silicon atoms let alone the quarks, gluons, leptons etc. particles these are made of on a more fundamental level.

 

[A. Neumaier]

Sure, I use effective descriptions of measurement devices and experimentalists use these descriptions even to build them. That's, however no contradiction to the statement that this effective description is just an approximation of the underlying "microscopic" dynamics which one cannot resolve anyway and which you don't need to resolve. To describe how a silicon-pixel detector works you don't need to describe the detailed interactions of a particle hitting $\mathcal{O}10^{24}$ silicon atoms let alone the quarks, gluons, leptons etc. particles these are made of on a more fundamental level.

But to show that the effective description is in agreement with the fundamental laws you need at least one truly closed system (i.e., a universe) into which the whole experimental setting is embedded. Without this, there is nothing to which the fundamental description applies that could be used as a starting point for deriving the effective theory. The measurement problem appears at this level.

 

[vanhees71]

Obviously the systems we use to test fundamental physics are closed to a good enough approximation. Locality of interactions is our friend ;-)).

 

[A. Neumaier]

That's, however no contradiction to the statement that this effective description is just an approximation of the underlying "microscopic" dynamics which one cannot resolve anyway and which you don't need to resolve.

How can this be if there is no underlying "microscopic" dynamics? The latter would be the dynamics of the whole universe, the only truly closed quantum system, hence the only quantum system to which Schrödinger's equation applies exactly. But according to your often stated view, the latter is not a valid quantum system...

 

[A. Neumaier]

Obviously the systems we use to test fundamental physics are closed to a good enough approximation.

These are tiny systems consisting of beams of particles during the time they are sill unobserved. But their detectors are huge and never even approximately closed.

 

[vanhees71]

Of course there's microscopic dynamics, described by a Hamiltonian including all relevant degrees of freedom and all relevant interactions. These cannot and need not be resolved and that's why we use effective theories to describe the relevant macroscopic degrees of freedom. They alone are of course never building a closed system, but you get a description that is accurate enough to describe the relevant dynamics. E.g., the motions of planets around the Sun in our solar System are well-described by classical Newtonian point mechanics. You don't need to solve the problem of Suns and planets as consisting of the particles of the Standard Model, which is neither possible nor useful in any way.

 

[A. Neumaier]

Of course there's microscopic dynamics, described by a Hamiltonian including all relevant degrees of freedom and all relevant interactions. These cannot and need not be resolved and that's why we use effective theories to describe the relevant macroscopic degrees of freedom. They alone are of course never building a closed system, but you get a description that is accurate enough to describe the relevant dynamics. E.g., the motions of planets around the Sun in our solar System are well-described by classical Newtonian point mechanics. You don't need to solve the problem of Suns and planets as consisting of the particles of the Standard Model, which is neither possible nor useful in any way.

It is a matter of principle, not of the ability to resolve it in practice. Laplace's deterministic universe could also not be resolved in practice.

If the standard foundations are exact and complete, there must be an exactly closed system. This system has to be the universe. All other systems are open and hence do not satisfy the Schrödinger equation exactly.

 

[vanhees71]

The observable universe is not closed either if you define it in such a strict way. In this sense you may conclude nothing in nature satisfies the Schrödinger equation exactly.

Now physics is not mathematics, and all natural laws, the Schrödinger equation one of them, are always valid up to the accracy they are empirically successful, and the Schrödinger equation is very successful within its range of applicability (much of atomic, molecule, and condensed-matter physics).

 

[A. Neumaier]

The observable universe is not closed either if you define it in such a strict way. In this sense you may conclude nothing in nature satisfies the Schrödinger equation exactly.

Now physics is not mathematics, and all natural laws, the Schrödinger equation one of them, are always valid up to the accracy they are empirically successful, and the Schrödinger equation is very successful within its range of applicability (much of atomic, molecule, and condensed-matter physics).

The closed system is the universe including its unobservable parts.

There are many who believe like me that it is possible to formulate a fundamental theory which exactly describes Nature in general terms - though we will never know exactly all details. Those working towards this goal are not satisfied with your operational approach, which kills all foundational curiosity by declaring it to be irrelevant or even futile.

 

[vanhees71]

Well, it's only relevant if it leads to observable consequences, making it testable against other models. I also don't think it's irrelevant to find a consistent description of all phenomena (including a quantum desription of gravity), but I don't think it's to be accessible with scholastic like approaches about pondering philosophical issues of existing theories. This even failed for the much simpler question about the nature of dark matter. All the ideas like SUSY and other extensions of the Standard Model so far have not lead to anything observable, and I think that even concerning this issue we need clear empirical input before we find the right ansatz for a new theory for "physics beyond the standard model". The same seems to hold the more for quantum gravity, where to find observable consequences of any quantu description seems to be even more challenging, because there's not any hint at such an observable, while dark matter seems to be rather omnipresent.