I Does quantum mechanics obey causality?

[n01]

My intuition tells me that it does not given that physical phenomena don't obey the principle of sufficient reason under quantum mechanics (a dogma many still hold certain). A lucid definition of the PoSR can be found here. Meaning, that some events are non-localized and the distinction between localized phenomena and global phenomena gets significantly blurred.

I don't see where else to post this as it's more of a meta-physical question.

 

[mfb]

Quantum field theory obeys causality. The commutator between spacelike separated points in spacetime vanishes, which means they don't influence each other.

 

[bhobba]

Quantum field theory obeys causality.

Its associated with the cluster decomposition property:
https://www.physicsforums.com/threads/cluster-decomposition-in-qft.547574/

Normal QM being based on the Galilean transformations doesn't. To fully understand the connection see the following beautiful book by the great theoretical/mathematical physicist Lev landau:
https://www.amazon.com/dp/0750628960/?tag=pfamazon01-20

I don't know why but that is the only book I know that gives the details, admittedly in Landau's usual terse style. Its very important but other books simply do not delve into it.

You will find QM dynamics developed from Galilean Relativity in Chapter 3 of Ballentine. Its the reason for Schrodinger's equation etc so is foundational and inescapable in ordinary QM. One must go to QFT to rectify it.

Thanks
Bill

 

[pr3dator]

Yes there is causality just sometime it is reversed

 

[bhobba]

I don't see where else to post this as it's more of a meta-physical question.

Its actually a deep physical question going to the heart of modern physics and its foundations from symmetry. It was quite possibly the deepest discovery of 20th century physics that symmetry is of such importance to physics like in the 19th century mathematicians discovered its importance to mathematics and especially geometry. I suspect there is a deep secret there waiting to be uncovered - but only time will tell.

Thanks
Bill

 

[bhobba]

Yes there is causality just sometime it is reversed

There is no causality in standard QM as can be seen by the form of the Hamiltonian obeying the Galilean transformations.

You are thinking of the retro-causal transnational interpretation. Its simply that - an interpretation.

Thanks
Bill

 

[pr3dator]

You have right.
One of the possible right interpretation.

 

[n01]

Its actually;y a deep physical question going to the heart of modern physics and its foundations from symmetry. It was quite possibly the deepest discovery of 20th century physics that symmetry is of such importance to physics like in the 19th century mathematicians discovered its importance to mathematics and especially geometry. I suspect there is a deep secret there waiting to be uncovered - but only time will tell.

Thanks
Bill

Yes,
I'm keenly interested on this from a quasi-computational perspective, and in some sense it's a tautology. So, let me elaborate if I don't start sounding metaphysical. Given that every physical law is either computable or non computable, then within such a system there will arise situations or "state of affairs" that could not be explained within the system itself. This is basically Godel's Incompleteness theorem stated in a nutshell.

My hunch is that QM and the logical conclusions derived at by Godel are in some deep sense intertwined and manifest in reality (I mean, how can they not be... unless we're talking about higher dimensions; but, even then those higher dimensions would require another higher dimension to maintain deterministic causality of each sub-dimension).

Hope that doesn't sound too outlandish as it does to me.

 

[jerromyjon]

This is basically Godel's Incompleteness theorem stated in a nutshell.

I take Godel a little more loosely, as in the system being used can't encompass all of reality, not that any given reality can't be modeled completely by a complete system. In my humble opinion, it is due to the introduction of supposed "infinite" degrees of freedom into a "finite" system which leads to the inconclusive result...

 

[Paul Colby]

Quantum field theory obeys causality. The commutator between spacelike separated points in spacetime vanishes, which means they don't influence each other.

Is this a definition of causality or merely a requirement of such? There are those amongst us that might claim entanglement experiments demonstrate superluminal "cause". I am very much not one of them.

 

[mfb]

"they don't influence each other" is what I would consider as definition of locality - causality then arises if you consider the evolution in one time-direction.

There are non-local interpretations of quantum mechanics - but you don't have to follow those, there are also local interpretations.

 

[stevendaryl]

Quantum field theory obeys causality. The commutator between spacelike separated points in spacetime vanishes, which means they don't influence each other.

People keep saying this, but to me, it's only addressing half of the quantum formalism. In quantum mechanics, you apply mathematics in two different ways:

1. The state and/or the field operators evolve with time.
2. You use the state to compute probabilities of outcomes.

The issue with whether QM is causal was never about #1, it was about #2. When you observe an outcome, does the state change? If so, how? The fact that field operators at a spacelike separation commute (or anticommute) doesn't have any obvious relevance to the question of whether Von Neumann's collapse happens, and whether that collapse (if it does happen) violates the no-FTL rule.

#1 would be sufficient to prove that causality is never violated, if that were the only process involved in QM (as is the case in many-worlds). But that conclusion doesn't follow from QFT, it requires an interpretation of QFT.

 

[mfb]

If there is a causal interpretation (and there is: MWI for example), and a non-causal interpretation, I would call the underlying theory causal. If it would be non-causal, then causal interpretations would be impossible.

 

[Paul Colby]

IMO, the entire reason "Does QM obey causality" is a question is due to a lack of a well defined theoretical or operational definition of the term "causality". The appearance of probability in a theory complicates this question.

 

[DrChinese]

My intuition tells me that it does not ...

I don't see where else to post this as it's more of a meta-physical question.

There is no physical evidence that at the quantum mechanical level there is any cause which specifically determines an outcome to a measurement (excepting repeated measurements when the systems is already in a known state). I doubt anyone here would dispute that statement.

It is a metaphysical question (as you say), because there are interpretations in which there are such causes. However, those interpretations cannot be discerned from non-causal interpretations by virtue of physical evidence. It is by inference and personal preference alone that such can be made.

 

[n01]

If any of you are interested in entertaining this thought and possibly looking at it from a more specific example of the origins of causality in QM...

I have always been puzzled by what exactly determines a wave function collapse? Or stated another way, what keeps quantum systems evolving as opposed to simply collapsing on itself?

 

[vanhees71]

There is no wave-function collapse. So you don't bother you with it :-).

 

[n01]

There is no wave-function collapse. So you don't bother you with it :-).

Well, yes according to the MWI; but, in that case what determines that our universe is the way it is as opposed to other worlds?

 

[vanhees71]

I don't believe in the religion of MWI. I'm a proponent of the minimal statistical interpretation. "Collapse" is just the interaction of the measured object with the measurement device, which usually is macroscopic, and we coarse grain over many microscopic degrees of freedom which leads to the classical behavior of the relevant observables of measurement apparatus. Thus, in general we don't know the state of the total quantum system (measured system + measurement apparatus), and tracing over the measurement apparatus leads to a mixture for the state of the observed system.

 

[jack476]

Yes. Quantum mechanics is probabilistic, but it's not random. The probabilities are deterministic.

 

[mfb]

I don't believe in the religion of MWI. I'm a proponent of the minimal statistical interpretation.

Do you call all interpretations religions?

Thus, in general we don't know the state of the total quantum system (measured system + measurement apparatus), and tracing over the measurement apparatus leads to a mixture for the state of the observed system.

Even worse, in the minimal statistical interpretations there is not even such a state, because you cannot give the wave function a physical reality (otherwise you need some way to get rid of it).

 

[Paul Colby]

Even worse, in the minimal statistical interpretations there is not even such a state, because you cannot give the wave function a physical reality (otherwise you need some way to get rid of it).

Is there an operational definition of or measure of something having "a physical reality?". IMO, the fact these so called interpretations can't be distinguished by the theory or experiment is proof they are simply added baggage that add nothing to the science.

 

[mfb]

The problem is that you need an interpretation to make experiments: you have to relate the theory to what you get as experimental result. This is trivial for all theories apart from quantum mechanics, so it seems to appear only there.

 

[Paul Colby]

The problem is that you need an interpretation to make experiments: you have to relate the theory to what you get as experimental result.

So you classify the Born rule as interpretation? It's what's observed to happen in experiments. Geiger counters count and photoelectrons are collected, at random.

 

[zonde]

The commutator between spacelike separated points in spacetime vanishes, which means they don't influence each other.

I would like to understand to what extent this statement is justified.
I will use an analogy. Parent has two different candies. Two kids ask him for candies, one asks first and the other one later. Parent gives random candy to the kid who asks first and the other one to other kid. Basically either kid can't really influence which candy he will get and which candy will get the other kid by asking fist or second. But just the same the two events are not independent.

So just because commutator vanishes we can't really claim that the two spacetime points are independent, right?

 

[atyy]

I would like to understand to what extent this statement is justified.
I will use an analogy. Parent has two different candies. Two kids ask him for candies, one asks first and the other one later. Parent gives random candy to the kid who asks first and the other one to other kid. Basically either kid can't really influence which candy he will get and which candy will get the other kid by asking fist or second. But just the same the two events are not independent.

So just because commutator vanishes we can't really claim that the two spacetime points are independent, right?

It's the usual answer.

1. Commutator vanishing implies signal locality, which QM respects.
2. Commutator vanishing does not imply local causality, which QM violates.

 

[mfb]

So you classify the Born rule as interpretation? It's what's observed to happen in experiments. Geiger counters count and photoelectrons are collected, at random.

Depends on how you use the rule, but in general: yes.

It's what's observed to happen in experiments. Geiger counters count and photoelectrons are collected, at random.

That is an interpretation already. How can you be sure it is random (all deterministic interpretations)? How can you be sure that your measurement result is the only one that happened (MWI)? How do you interpret small deviations from the Born rule (which will typically happen - within the uncertainty)? And why do you choose the Born rule?
All those questions do not have testable laws of physics as answer - they are interpretations.

 

[Yubal M]

Correct me if I am wrong, but the mainstream dominating thinking in Quantum Mechanics (QM) since the 1930s is the Copenhagen Interpretation (CI) conceived mainly by Niels Bohr. The non-causality of quantum effects is the heart in the CI. But since then (and even simultaneously with the CI) there are alternative models of QM which diminish the non-causality in QM and render greater causality to the quantum effects. The price for this is, as far as I get, that they assume some unidentified pilot waves, unidentified quantum fields, physical realities outside of our space-time. But all these do give some sense to the weirdness (non-causality) of QM as revealed in the CI.

An experimental evidence for a pilot wave pattern was found around 2 years ago in a macroscopic scale: https://www.quantamagazine.org/20140624-fluid-tests-hint-at-concrete-quantum-reality/

Please note that it's not any direct quantum effect found in a macroscopic scale, but just a pattern appearing in a macroscopic scale which is identical to a predicted pattern appearing in an alternative quantum model to the CI.

2 examples (books) for such alternative different models:
a. Wholeness and the Implicate Order, by David Bohm (The conceiver of the Pilot Wave model)
b. The Transactional Interpretation of Quantum Mechanics -- The Reality of Possibility, by Ruth E Kastner (the idea first conceived by John G Cramer in 1986).

 

[Paul Colby]

That is an interpretation already. How can you be sure it is random (all deterministic interpretations)

Even classical statistical mechanics has this issue which is also an application of probability theory. Is it "truly random" (what ever that means) is not a question for physics unless it is a testable assertion. The Born rule states one uses probability theory to connect the formalism with experiments and says nothing about the results being or not being truly random. This connection has been tested with nearly every experimental application of the theory. For me the Born rule is not interpretation but an established fact. In this narrow sense it's no different than the application of probability in statistical mechanics.

 

[bhobba]

Is there an operational definition of or measure of something having "a physical reality?".

Yes there is. Its when decoherence occurs. The question that is left up in the air, not addressed, explicitly or not explicitly stated etc etc, depending on what you are reading, and how careful they are, is how does a improper mixed state become a proper one. Colloquially its - why do we get any outcomes at all. THE book that carefully and exhaustively explains all this is Schlosshauer's standard text :
https://www.amazon.com/dp/3540357734/?tag=pfamazon01-20

Thanks
Bill

 

[bhobba]

Correct me if I am wrong, but the mainstream dominating thinking in Quantum Mechanics (QM) since the 1930s is the Copenhagen Interpretation (CI) conceived mainly by Niels Bohr

Yes - but with our modern understanding of decoherence it has morphed a bit because it leaves a question open. That question is, since in Copenhagen, QM is a theory about observations that appear here in an assumed common-sense classical world, how does it explain such a world that it assumes in the first place. Great progress has been made in rectifying that blemish, but some issues remain. I think the modern form of Copenhagen would be Consistent Histories:
http://quantum.phys.cmu.edu/CHS/histories.html

It's a nice interpretation with a lot to like. For me however it has the feel of defining your way out of problems which is why I hold to the ignorance ensemble interpretation. It's just a minor variation on the ensemble interpretation applying it only to the mixed state after decoherence.

I think studying interpretations is very interesting, and sheds a lot of light on the formalism. Particularly interesting is seeing exactly what the formalism implies and doesn't. For example since we have interpretations without collapse such as MW the formalism doesn't have collapse, even though at first sight you think it does. But it must be borne in mind, and this is VERY VERY important, no interpretation is better hasn't any other - choice is purely a personal thing depending on what appeals to you. The other thing is, very difficult questions in QM such as if its random or not, are trivial in specific interpretations and IMHO are best approached that way - ie discuss them in various interpretations and not generally.

Thanks
Bill

 

[bhobba]

Wholeness and the Implicate Order, by David Bohm (The conceiver of the Pilot Wave model).

David Bohm was a great physicist. When he was being that he was very good, but when he wasn't - well you get writings like the above. Its nearly, but not quite, metaphysical mumbo jumbo that borders on junk science, not Bohm's finest hour.

Thanks
Bill

 

[bhobba]

So you classify the Born rule as interpretation? It's what's observed to happen in experiments. Geiger counters count and photoelectrons are collected, at random.

That's actually an interesting question.

You see the Born rule mentions probability, and as John Baez, correctly IMHO, asserts, many QM interpretations are simply arguments about the meaning of probability:
http://math.ucr.edu/home/baez/bayes.html

John's comment about frequentest probability, while correct as it stands, needs further fleshing out, but this is not the thread to do it.

Thanks
Bill

 

[vanhees71]

Do you call all interpretations religions?Even worse, in the minimal statistical interpretations there is not even such a state, because you cannot give the wave function a physical reality (otherwise you need some way to get rid of it).

Yes, everything that's not testable by observations/experiments is not physics and thus in some sense free to individual believes like religion.

Of course, a state is not described by "a wave function" but a statistical operator in Hilbert space. Wave functions are representants of states for a very small subset of situations, where nonrelativistic descriptions with a conserved particle number are applicable.

For me, the quantum-theoretical state is very real. It's first defined operationally as a (an equivalence class of) preparation procedure(s). Given the state, you have probabilistic information about the system. If you have even a pure state, it's the situation, where you have maximal possible knowledge about the system. Contrary to the situation in classical physics, this complete state determination does not imply the determination of all observables. This indeterminacy of many observables in a given state is a real property of nature.

I avoid to use the word "reality" in the context of physics, because this notion is spoiled by a plethora of different philosophical meanings, which usually are not even clearly defined. As I said, from a physics point of view, reality is everything that's obejctively observable or even quantitatively measurable.

 

[n01]

Of course, a state is not described by "a wave function" but a statistical operator in Hilbert space. Wave functions are representants of states for a very small subset of situations, where nonrelativistic descriptions with a conserved particle number are applicable.

Isn't that the same thing, only that you're committing the wavefunction to a determinate state by measurement or observation? This would be, by all intents and purposes, eliminating entanglement and nonlocality states of wavefunctions, yes?

 

[vanhees71]

How do you come to this conclusion? I've not said anything along these lines. I just describe standard quantum theory in the minimal interpretation, which of course includes entanglement and long-ranged correlations associated with them. I don't like to call them less precisely "nonlocality" since this is often mixed up with non-local interactions, which are very problematic to say the least, and there's no evidence from observations that they are needed. The most successfull theory is the Standard Model of elementary particle physics which has incorporated the locality of interactions in its foundations. Nevertheless, of course, it includes the possibility for entangled states of systems with "parts" showing strong correlations. This is well-established nowadays (mostly realized with polarization-entangled two-photon states as can be prepared on demand by shining a laser on certain birefringent crystals in a process called parametric down-conversion).

 

[StevieTNZ]

This is a recently published article which may be of interest: http://phys.org/news/2016-08-quantu...e=menu&utm_medium=link&utm_campaign=item-menu

 

[Mentz114]

This is a recently published article which may be of interest: http://phys.org/news/2016-08-quantu...e=menu&utm_medium=link&utm_campaign=item-menu

I think this is it http://arxiv.org/abs/1602.02767

 

[bhobba]

I think this is it http://arxiv.org/abs/1602.02767

From the above:
'Explaining observations in terms of causes and effects is central to all of empirical science. Correlations between entangled quantum particles, however, seem to defy such an explanation.'

Scratching head. Of course it does. Correlations must be removed from a causal explanation of anything and that is all Bell is. As I often post, standard QM is not local - its in its very foundations because it obeys the Galilean transformations. In fact as chapter 3 of Ballentine shows its the only thing needed to deduce the dynamics. And, just as an aside, as Landau's beautiful book, Mechanics, shows it's all that is required even in classical mechanics - if you have not read that masterpiece please get a copy - you will not be disappointed (the real foundation is the principle of least action which follows from QM - strangely at first sight the basis of classical mechanics is QM). You must go to QFT for locality to be an issue and when you do that its based on the cluster decomposition property:
https://www.physicsforums.com/threads/cluster-decomposition-in-qft.547574/

For it to make sense it can't apply to correlated systems so they must be removed ie it does not apply to Bell type correlations.

One thing I have found out about physics since posting here is sometimes even professional physicists get confused about basic things. To be fair even I was confused about a lot of stuff before posting here - I look back and wince at some of the things I thought back then. But it does show you must be careful of what is written in professional papers.

Added later:
I have read a bit more of that paper. I sort of see what they are getting at, but it is a bit confused about fundamental things and that needs to be disentangled before getting to its 'meat' so to speak. I seem to recall it has been discussed here before, but if people really want to chat about it best to start a new thread.

To get back to the original question, that too has been discussed many times - the answer depends of what is meant by causality (Schrodinger's equation shows the state is causal) but the standard formalism is non committal about the cause of observational outcomes and is best discussed with reference to some interpretation otherwise in answering it you just go around in circles.

Thanks
Bill

 

[mikeyork]

The notion of causality is rooted in our perception of space-time. There is currently some discussion of whether space-time exists in reality or is something an observer creates to bring some sort of consistency to their perception.

 

[bhobba]

The notion of causality is rooted in our perception of space-time.

Scratching head.

Its rooted in cause and effect. It's usually specified by some kind of differential equation.

QM has Schrodinger's equation so initial state determines final state. In that sense its causal. But the act of observation - that is a matter of interpretation - the formalism is silent on it.

Thanks
Bill

 

[mikeyork]

Your mention of "initial" and "final" state implies time. Schrodinger's equation involves derivatives with respect to space and time since it describes evolution in time.

 

[bhobba]

Your mention of "initial" and "final" state implies time

Your logic is astray. I wasn't going to spell it out, but decided to. Differential equations have a set of parameters that can be in anything not just time or space. What they are determines other values - not just space and time. It'simply colloquial to call them initial and final.

Thanks
Bill

 

[mikeyork]

Your own example of Schrodinger's equation (the context in which you mention "initial" and "final" states) shows the space-time context explicitly. Differential equations show analytic continuity. I consider this is a different concept from causality. For instance causality need not be continuous, but describe discrete but related events. That context also implies some notion of time though, just not continuous co-ordinate time.

 

[bhobba]

Your language is astray

Fair enough.

Take the example of inflationary models. They are expressed in the language of differential equations 'prior' to the emergence of space and time. But here we run into problems with language because prior makes no sense without the concept of time. That's why math, not English, is the best language to discuss this.

Thanks
Bill

 

[mikeyork]

Our posts are crossing each other. Editing complicates things even more.

You wrote "prior makes no sense without the concept of time". This sounds very much like "causality makes no sense without the concept of time". So I think we agree once we get the language issues out of the way and in the absence of time-dependence focus on the math.

 

[bhobba]

Our posts are crossing each other. Editing complicates things even more.

Yes - I apologize for the editing I did before.

My point is that causality has with modern theories gone beyond notions of space time etc.

Thanks
Bill

 

[David Lewis]

Parent has two different candies. Two kids ask him for candies, one asks first and the other one later. Parent gives random candy to the kid who asks first and the other one to other kid. Basically either kid can't really influence which candy he will get and which candy will get the other kid by asking first or second. But just the same the two events are not independent.

What if the candies are not of one type or another until after a child sees what kind of candy he got?

 

[bhobba]

You wrote "prior makes no sense without the concept of time". This sounds very much like "causality makes no sense without the concept of time". So I think we agree once we get the language issues out of the way and in the absence of time-dependence focus on the math.

Yes. With these kind of foundational issues English is the enemy.

If, in terms of math, you can explain your context that would be fantastic. Hopefully I have done that - but feel free to ask for any clarification.

Thanks
Bill

 

[martinbn]

Fair enough.

Take the example of inflationary models. They are expressed in the language of differential equations 'prior' to the emergence of space and time. But here we run into problems with language because prior makes no sense without the concept of time. That's why math, not English, is the best language to discuss this.

Thanks
Bill

What do you mean by this? Inflationary models do not deal with anything 'prior' to space-time.