How could a quantum system "predict the future"?

  • I
  • Thread starter sayetsu
  • Start date
  • #1
29
1

Summary:

(Quote included. I guess "nonlocality" in involved.)

Main Question or Discussion Point

"How that quantum system 'knows' that it must evolve into an eigenstate can be viewed in (at least) two ways. One is to assume that the quantum system somehow “knows” that an interaction will take place in the future. This is actually less troubling than it sounds, since this kind of so-called nonlocality is part of quantum physics anyway, and does not lead to any observable violations of macroscopic causality.
What is nonlocality? How could a quantum system (like a superposition - is that right?) "know" an interaction will take place?

Source: https://www.quora.com/I-heard-that-when-we-observe-a-quantum-system-we-cause-its-wave-function-to-collapse-Whats-so-special-about-us-as-humans-making-that-measurement?q=How do I avoid solipsistic belief if I can't confirm there any observers causing quantum wave-function collapse but me?
 

Answers and Replies

  • #2
PeroK
Science Advisor
Homework Helper
Insights Author
Gold Member
12,317
5,117
Summary:: (Quote included. I guess "nonlocality" in involved.)

What is nonlocality? How could a quantum system (like a superposition - is that right?) "know" an interaction will take place?

Source: https://www.quora.com/I-heard-that-when-we-observe-a-quantum-system-we-cause-its-wave-function-to-collapse-Whats-so-special-about-us-as-humans-making-that-measurement?q=How do I avoid solipsistic belief if I can't confirm there any observers causing quantum wave-function collapse but me?
I would ignore anything on Quora. Unless you already know the answer it's impossible to tell the expert opnion from the nonsense on there.
 
  • #3
9,331
2,306
Non locality is the idea effects can happen instantaneously. Quantum systems do not know interactions will take place anymore than you would. Nothing is special about measurement, it's simply an interaction wth a quantum system. Collapse is an interpretation thing, even though some books do not present it that way.

If you really want to understand this sort of thing you first need to study an actual interpretation:
http://quantum.phys.cmu.edu/CHS/histories.html

Thanks
Bill
 
  • #4
DrChinese
Science Advisor
Gold Member
7,243
1,053
What is nonlocality? How could a quantum system (like a superposition - is that right?) "know" an interaction will take place?
This discussion probably belongs in the Interpretations subforum.

Your "Nonlocality" is better labeled as "Quantum Nonlocality" because it is defined/constrained by the rules of quantum mechanics. Three things:

1. Whatever nonlocality exists does not support a mechanism for faster than light signaling.
2. No one knows the underlying mechanism. There are quantum interpretations that intend to supply an answer.
3. Whether any kind of nonlocality is exhibited at all is therefore quantum interpretation dependent.
 
  • #5
PeterDonis
Mentor
Insights Author
2019 Award
28,263
8,040
This discussion probably belongs in the Interpretations subforum.
Indeed. Moved.
 
  • #6
vanhees71
Science Advisor
Insights Author
Gold Member
2019 Award
14,381
5,974
This discussion probably belongs in the Interpretations subforum.

Your "Nonlocality" is better labeled as "Quantum Nonlocality" because it is defined/constrained by the rules of quantum mechanics. Three things:

1. Whatever nonlocality exists does not support a mechanism for faster than light signaling.
2. No one knows the underlying mechanism. There are quantum interpretations that intend to supply an answer.
3. Whether any kind of nonlocality is exhibited at all is therefore quantum interpretation dependent.
1. and 3. are for sure true.

Concerning 2. that's not right. Anybody who has understood the content of a QM 1 lecture knows how "it works". What you call non-locality are in fact long-ranged correlations described by entanglement. A much better naming is dubbed by Einstein, and it's "inseparability".

A quantum system, consisting in some well-defined sense by two parts, which can be observed at far distant places, can be "inseparable" if some observables of each of the two parts are entangled.

Nowadays the most simple example is the entanglement of the polarization states of two photons originating from a process called parametric downconversion, where a strong laser beam interacting with certain sorts of birefringent crystals (usually a beta-barium borate (BBO) crystal) leads to the emission of two polarization (end momentum) entangled photons. Waiting long enough and making sure that nothing disturbs the photons on their path to the detectors the detectors can be located at a very far distance. The polarization state of each of the single photons is maximally indetermined, i.e., each of the observers just sees perfectly unpolarized photons when the experiment is repeated very often. But comparing the measurement of the polarization at both places with the polarization direction of both detectors the same (or perpendicular) one finds a 100% correlation between the measurement results (with the details depending on which of the entangled states you have prepared).

Some people think there's something "nonlocal" here, particularly if they take the socalled collapse hypothesis literally. As @DrChinese emphasized with item 1. there's nothing "nonlocal" here in the sense of "instantaneous interaction at a distance", i.e., nothing is nonlocal which would violate the causality structure of relativistic spacetime, and indeed by construction the successful relativistic sorts of quantum theory are local relativistic QFTs, where here "local" has a precise mathematical meaning, namely that local observable operators commute at space-like distances of their arguments. This holds particularly for the Hamilton density, and thus it implies that there are indeed by construction no "non-local interactions" or "causal faster-than-light effects" but as any QT also relativistic QFT allows for the stronger-than-classical "non-local correlations" described by entanglement as in the example with the two photons created in, e.g., a parametric-down-conversion process. That's why I prefer Einstein's notion of "inseparability" for this rather than "non-locality".

In other words: Entanglement is a property of an state and has nothing to do with interactions of the measured objects with the measurement devices (which are always local in the above QFT sense!).
 
  • #7
EPR
Gold Member
153
25
1. and 3. are for sure true.

Concerning 2. that's not right. Anybody who has understood the content of a QM 1 lecture knows how "it works". What you call non-locality are in fact long-ranged correlations described by entanglement. A much better naming is dubbed by Einstein, and it's "inseparability".

A quantum system, consisting in some well-defined sense by two parts, which can be observed at far distant places, can be "inseparable" if some observables of each of the two parts are entangled.

Nowadays the most simple example is the entanglement of the polarization states of two photons originating from a process called parametric downconversion, where a strong laser beam interacting with certain sorts of birefringent crystals (usually a beta-barium borate (BBO) crystal) leads to the emission of two polarization (end momentum) entangled photons. Waiting long enough and making sure that nothing disturbs the photons on their path to the detectors the detectors can be located at a very far distance. The polarization state of each of the single photons is maximally indetermined, i.e., each of the observers just sees perfectly unpolarized photons when the experiment is repeated very often. But comparing the measurement of the polarization at both places with the polarization direction of both detectors the same (or perpendicular) one finds a 100% correlation between the measurement results (with the details depending on which of the entangled states you have prepared).

Some people think there's something "nonlocal" here, particularly if they take the socalled collapse hypothesis literally. As @DrChinese emphasized with item 1. there's nothing "nonlocal" here in the sense of "instantaneous interaction at a distance", i.e., nothing is nonlocal which would violate the causality structure of relativistic spacetime, and indeed by construction the successful relativistic sorts of quantum theory are local relativistic QFTs, where here "local" has a precise mathematical meaning, namely that local observable operators commute at space-like distances of their arguments. This holds particularly for the Hamilton density, and thus it implies that there are indeed by construction no "non-local interactions" or "causal faster-than-light effects" but as any QT also relativistic QFT allows for the stronger-than-classical "non-local correlations" described by entanglement as in the example with the two photons created in, e.g., a parametric-down-conversion process. That's why I prefer Einstein's notion of "inseparability" for this rather than "non-locality".

In other words: Entanglement is a property of an state and has nothing to do with interactions of the measured objects with the measurement devices (which are always local in the above QFT sense!).
Why does QFT appear to be shunned given that it's the most comprehensive quantum theory?

We already know QFT is not a realistic theory('particles' are not there with properties but just a momentary iteration of the respective field). What is keeping people from moving on from outdated models and concepts?
 
Last edited:
  • #8
9,331
2,306
Why does QFT appear to be shunned given that it's the most comprehensive quantum theory?
It's not shunned - its just not generally required for EPR type experiments - although of course you can use it to analyse them if you want. But perhaps it's wise to keep it in mind given we know as per chapter 3 of Ballentine a number of features of standard QM tacitly assumes the Galilean transformations as an approximation, which are inherently non local. But as Landau points out in Mechanics it's the same in non-relativistic mechanics anyway.

We already know QFT is not a realistic theory('particles' are not there with properties but just a momentary iteration of the respective field). What is keeping people from moving on from outdated models and concepts?
You are interpreting QFT. What QFT says is particles are like a knot in a quantum field, but you get standard QM as the classical limit anyway, which is all you need.

Thanks
Bill
 
  • #9
EPR
Gold Member
153
25
I don't get the obssession with QM over QFT. QFT is a relativistic theory, i.e. it observes locality(the speed of light limit) while it sheds realism('particles' aren't fundamental but an emergent property of fileds). Yet, i've seen dozens of threads here arguing endlessly if reality in QM was either local or realistic. From the point of view of qm's extention, reality is local but non-realistic.
And it so happens, that QFT is the biggest success story in physics in its staggering accuracy(QED). It's almost a theory of everything, while QM is a mathematical tool for low energy statistics. Bump the energy and it becomes invalid(e.g. scattering experiments, esp with high speed particles).
 
Last edited:
  • #10
9,331
2,306
In other words: Entanglement is a property of an state and has nothing to do with interactions of the measured objects with the measurement devices (which are always local in the above QFT sense!).
Another way of looking at it is once systems, particles, etc become entangled they are now a single quantum system, not two separate systems. When you observe that system you interact with that single system and get an outcome - for example a particle with spin up and another particle, perhaps even light years away, with spin down, because you have set it up so they are correlated ie if one is spin down the other is spin up or conversely. You can't deduce, without other assumptions (generally called interpretations) that one particle has, in a non-local way, influenced the other particle because there may be no separate particles before the observation - only this single entangled system. Of course you are perfectly entitled to adhere to an interpretation that makes assumptions that allow you to draw that conclusion. But it is an assumption you are making - an assumption not forced on you by the facts. My view for what it's worth is non-locality is not a useful concept for correlated particles in the theory standard QM is a limiting case of, Quantum Field Theory, because that uses the so called Cluster Decomposition Property which you can only make sense of if you exclude correlated systems. But that is my view - their are all sorts of others and they fit the facts just as well - but for me introduce unnecessary and 'weird' ideas. However only correspondence with experiment is what science requires - with occasional exceptions that are really more philosophy than science and not really suitable for this thread.

Thanks
Bill
 
  • #11
vanhees71
Science Advisor
Insights Author
Gold Member
2019 Award
14,381
5,974
Why does QFT appear to be shunned given that it's the most comprehensive quantum theory?

We already know QFT is not a realistic theory('particles' are not there with properties but just a momentary iteration of the respective field). What is keeping people from moving on from outdated models and concepts?
I think you missed the most successful developments of the 2nd half of the 20th century! SCNR.
 
  • #12
vanhees71
Science Advisor
Insights Author
Gold Member
2019 Award
14,381
5,974
I don't get the obssession with QM over QFT. QFT is a relativistic theory, i.e. it observes locality(the speed of light limit) while it sheds realism('particles' aren't fundamental but an emergent property of fileds). Yet, i've seen dozens of threads here arguing endlessly if reality in QM was either local or realistic. From the point of view of qm's extention, reality is local but non-realistic.
And it so happens, that QFT is the biggest success story in physics in its staggering accuracy(QED). It's almost a theory of everything, while QM is a mathematical tool for low energy statistics. Bump the energy and it becomes invalid(e.g. scattering experiments, esp with high speed particles).
Now you are talking!
 
  • #13
vanhees71
Science Advisor
Insights Author
Gold Member
2019 Award
14,381
5,974
Another way of looking at it is once systems, particles, etc become entangled they are now a single quantum system, not two separate systems. When you observe that system you interact with that single system and get an outcome - for example a particle with spin up and another particle, perhaps even light years away, with spin down, because you have set it up so they are correlated ie if one is spin down the other is spin up or conversely. You can't deduce, without other assumptions (generally called interpretations) that one particle has, in a non-local way, influenced the other particle because there may be no separate particles before the observation - only this single entangled system. Of course you are perfectly entitled to adhere to an interpretation that makes assumptions that allow you to draw that conclusion. But it is an assumption you are making - an assumption not forced on you by the facts. My view for what it's worth is non-locality is not a useful concept for correlated particles in the theory standard QM is a limiting case of, Quantum Field Theory, because that uses the so called Cluster Decomposition Property which you can only make sense of if you exclude correlated systems. But that is my view - their are all sorts of others and they fit the facts just as well - but for me introduce unnecessary and 'weird' ideas. However only correspondence with experiment is what science requires - with occasional exceptions that are really more philosophy than science and not really suitable for this thread.

Thanks
Bill
But it's not an interpretation that the interactions described by standard relativistic QFTs are local by construction, i.e., causal effects can propagate at most with the speed of light (in vacuum) by mathematical construction!

With the rest I completely agree, and that's why I plead for the use of the word "inseparability" (coined by Einstein in his personal version of the EPR paper which he didn't like much, because it does not properly reflect his view; Einstein's quibble was in fact about inseparability):

A. Einstein, Quantenmechanik und Wirklichkeit, Dialectica 2
(1948) 320.
https: //dx.doi.org/10.1111/j.1746-8361.1948.tb00704.x

I don't know, whether there's an Enlish translation though :-(.
 
  • #14
EPR
Gold Member
153
25
When the field gets a local boost in energy, with the energy propagating through the field we observe a moving 'particle'.... Setting the 'measurement' issue aside(as it's quite vague), is this a valid description?
Is this how the double-slit experiment is treated in QFT?
 
  • #15
vanhees71
Science Advisor
Insights Author
Gold Member
2019 Award
14,381
5,974
No, "particles" are rather particular quantum states, the socalled asymptotic free Fock states. The measurement of particles is not vague at all but a highly developed experimental technique. Just look at the amazing huge detectors at CERN (ALICE, ATLAS, CMS, LHCb). They are complicated master pieces of current detector development but far from being vague in any sense!

I'm not aware of any QFT-treatment of the double-slit experiment. That's an interesting question. Usually it's treated for non-relativistic particles using the Schrödinger equation in an effective way using the Green's function solving with the boundary conditions imposed by the slits, but even in this apparently simple treatment there are still interesting recent new developments:

https://doi.org/10.1088/1367-2630/aac92c (open access paper)
 
  • #16
1,168
210
There are some novel interpretations beside already mentioned, like ER=EPR .
Have no idea how far they got with that.
 
  • #17
EPR
Gold Member
153
25
I was talking about 'measurement' in the usual twin-slit treatment. Does QFT resolve the so called measurement problem of QM, or does it 'smooth it out' by applying 'particle' creation and annihilation operators(but is still fundamentally stuck in probabilities)?
Or are you implying that since QFT usually treats high energy partcles, it's closer to classical relativistic theory than QM and thus suffers less from the inherent quatum ambiguity and uncertainty prior to measurement?
How does QFT explain why turning on a detector at one of the slits of double-slit experiment produces 2 single stripes instead of an interference pattern?
From what i have researched, it doesn't help much with these questions and the issue is as unresolved as it was 70 years ago. I.e. QFT paints a much more comprehensive picture of how the world is, but the deep questions of QM remain unresolved.
 
  • #18
vanhees71
Science Advisor
Insights Author
Gold Member
2019 Award
14,381
5,974
I don't know, which "measurement problem" you mean in connection with the double-slit experiment, which have been realized many times with no specific measurement problems.

I invoke relativistic QFT in these arguments only in connection with the claim that there are non-local interactions in connection with Bell experiments on entangled fast distant parts of a quantum system. By construction there are no non-local interactions within relativistic QFT, and since relativisitc QFT describes the observed properties (correlations) of these entangled states very well, there's no contradiction between relativistic spacetime/causality structure and these observed properties of entangled states.

I don't see any deep questions of QM unresolved. QM is very successfully applied to almost all known phenomena. The only exception and a very tough problem is of course gravity.
 
  • #19
EPR
Gold Member
153
25
The measurement problem that the quantum particle in the course of evolution as described by the Schrodinger equation appears to exist in all of its possible states(is delocalized and is in an idefinite state), but during measurement, the particle is always detected in only one of its possible states. The MP is carried over from QM to QFT, no?
 
  • #20
vanhees71
Science Advisor
Insights Author
Gold Member
2019 Award
14,381
5,974
The particle is at some time ##t_0## in the state it has been prepared or has been observed to be in, and then the Schrödinger equation tells you, given the complete Hamiltonian of the system, how this state evolves (in the Schrödinger picture of time evolution). The system is always in one and only one pure or mixed state and never in several states at once. This is often written in bad pupular-science books to describe superpositions of state vectors, but the state vector is independent of the choice of the specific basis you express its components in. That's so with any kind of vectors, but specifically also with the vectors of the quantum-mechanical (rigged) Hilbert space.

The state (described a positive semidefinite trace-one self-adjoint operator) describes the statistical properties for the outcome of meausurements, i.e., it gives the probability to measure any possible value for the measured quantity. To measure a quantity means by definition (i.e., construction of the measurement device) that the measurement gives a unique result, i.e., a specific value for the measured quantity. There's nothing mysterious in that, because that's how for neary 100 years QT is used to describe real-world observations and measurements.

On this fundamental level, there's no difference between QM and QFT. QM is an approximate formulation of quantum theory for some subclass of systems (i.e., those which are describable within non-relativistic physics with a conserved number of particles). QFT can be used to describe non-relativistic or relativistic systems with non-conserved particle (or quasi-particle) numbers.
 
  • #21
EPR
Gold Member
153
25
Agreed but if it's all just statistics of a single quantum system, what traverses the 2 slits and interferes with itself? The Hilbert space appears to correspond to something underlying the double-slit setup. Can we just brush it aside as something irrelevant in QFT?
The field must have some reality during the unitary evolution(possibly the only reality). The detector simply highlights the resultant measurement issue of 'particles'.
I understand you can do physics just fine without looking at these issues, but some physicists aren't contend and looking to resolve the conundrum.
Why does the photon field create particles in 2 stripes when there's a detector on?
 
  • #22
PeterDonis
Mentor
Insights Author
2019 Award
28,263
8,040
QFT paints a much more comprehensive picture of how the world is, but the deep questions of QM remain unresolved.
A better way to say it might be that, for people who see these "deep questions" as meaningful (some people, like @vanhees71, apparently do not), QFT doesn't say anything new about them compared to ordinary non-relativistic QM.

some physicists are looking to resolve the conundrum.
And as far as physics is concerned (as opposed to interpretations--one can always concoct new interpretations, but that doesn't help with resolving any physics questions), it's not going to be resolved in the framework of current QM/QFT. The only way for physics to resolve such conundrums (if you think there are conundrums to be resolved) is to come up with at least two different extensions to the current theory (in this case QM/QFT), that make different predictions about the results of some experiment, and then run the experiment to see which one is right. Nobody currently knows how to do that for QM/QFT.
 
  • #23
vanhees71
Science Advisor
Insights Author
Gold Member
2019 Award
14,381
5,974
It's a bit like in the tale of the emperor's new clothes: It's obvious that the emperor is naked, but only the child who just looks at the facts as they are realizes it.

It's the same with these apparent "problems" (measurement process, "reality", etc.): Quantum theory as any other kind of physical theory is a description of what we observe objectively (i.e., independent from who is performing the observation and restricted to reproducible observations) in terms of a few fundamental laws in quantitative mathemtical form. That this is possible at all is itself an observational fact, and it's quite amazing (as Wigner stressed in a famous essay).

It's an observational fact that quantum theory works (as far as it applies, and it seems to apply very well to everything except the gravitational interaction), i.e., it predicts the randomness of outcomes of measurements of the observables given the state of the system and it provides quantitatively the probabilities for the outcome of measurements which are in accordance with observations on ensembles (which are just stochastically independent systems prepared in a defined quantum state), using the frequentist interpretation of probabilities (note that there is as much philosophical confusion about the meaning of probabilities in general as there is about the particular case of quantum theory, providing probabilities for the outcome of measurements).

Since it's very well defined by the concrete measurement devices and experiments built by experimental physicists in their labs and the probabilistic description of quantum theory with astonishing significance is confirmed by all these experiments and observations my conclusion is that the "measurement problem" is like the emperor's new clothes: It's just not existent, and this holds true for the apparent idea that this view is "not realistic". The notion of "realism" has been so many meanings in the philosophical discourse that it becomes nearly unusable in the natural sciences. One should just remember what science (at least physics) is about: The objective observation of natural phenomena and the induction of general laws in terms of mathematical theories and models, which can be tested again in new experiments and are in principle falsifiable by experiments, which then leads to the development of ever better models and theories.
 
  • #24
EPR
Gold Member
153
25
I agree with all this and have stressed before that map and territory are not the same thing. Models imply conclusions which are always tentative as what Bohmian mechanics as a model teaches is that a new model may imply different things about how the world is. It's not possible to know if a new theory in the future will not supercede Quantum Theory and imply different ideas and concepts about what and how the world is. Every conclusion in that direction is tentative and somewhat speculative even if some physicists hold the current framework and what it seems to imply as ultimate Truth.
 
Last edited:
  • #25
ftr
622
43
probabilistic description of quantum theory
I think the question is that is nature truly fundamentally is probabilistic in the sense of CI. It is clear the interpretations schemes are to prove otherwise, and the different ideas are held by so many physicists( which you seem to dismiss them as philosophical which they beg to differ, otherwise why go through the trouble!). Even some like t'hoft and Weinberg expressing great unease about the whole thing, indicating that there is a deep problem with the setup of QM (EPR is one of them which you keep dismissing as a mystery). So I think your characterization of the issue as an open and shut case is very unfair, to say the least.

https://en.wikipedia.org/wiki/Interpretations_of_quantum_mechanics
https://www.preposterousuniverse.com/blog/2013/01/17/the-most-embarrassing-graph-in-modern-physics/
https://www.google.com/search?q=poll+about+quantum+mechanical+interpretation&oq=poll+about+quantum+mechanical+interpretation&aqs=chrome..69i57.17879j0j8&sourceid=chrome&ie=UTF-8
 

Related Threads on How could a quantum system "predict the future"?

Replies
9
Views
2K
Replies
13
Views
1K
Replies
22
Views
2K
Replies
4
Views
6K
Replies
17
Views
3K
Top