Many Worlds Interpretation existence

[martinbn]

But in non-local BM, the ontology (namely particle positions) is local. What is strange about that?

That part is not strange, although it seems like a step back towards classical thinking. The strange bit is the action at a distance.

 

[martinbn]

What you call "black box" is the most successful and comprehensive model about the objectively observable part of Nature. Admittedly it's incomplete since we haven't been able to quantize the gravitational interaction, but you can't expect Nature to be as you like. She just behaves as she does, and as physicists we have to figure out how to describe this behavior as good as we can. You can't expect more from the natural sciences than the best possible description of phenomena that can be objectively quantified and observed.

But even without gravity it has problems. It has the measurement problem. QFT is mathematically not sound.

 

[Boing3000]

How is absolute simultaneity compatible with relativity?!

How is it not ?

 

[vanhees71]

But even without gravity it has problems. It has the measurement problem. QFT is mathematically not sound.

Hm, and I thought perturbative renormalized QFT makes a lot of sense (up to 12 digits or even more confirmation of predictions by experiment)!

 

[martinbn]

How is it not ?

Not sure how to respond to this. Given that relativity of simultaneity is one of the most famous things about relativity, if you claim the opposite you need to explain, not ask questions.

 

[martinbn]

Hm, and I thought perturbative renormalized QFT makes a lot of sense (up to 12 digits or even more confirmation of predictions by experiment)!

And yet there is no mathematically rigorous QFT in 3+1 dimensions.

 

[Boing3000]

Not sure how to respond to this. Given that relativity of simultaneity is one of the most famous things about relativity, if you claim the opposite you need to explain, not ask questions.

They are not opposite at all, one is the consequence of the other. I have explained what this "now line" is, and despite the ugly language, those absolute proper geodesics exist and define that "absolute proper simultaneity".

Experimenter are actually using them and try to measure it. I am pretty sure they use relativity quite extensively, especially to synchronize clocks.

Relativity allows us to define an absolute proper universe, whatever our particular relative perspective on it. And non-local entanglement is totally compatible with it. That's FTL that is not (hypothetical tachyon aside)

 

[Demystifier]

That part is not strange, although it seems like a step back towards classical thinking. The strange bit is the action at a distance.

I agree that action at a distance is strange. But to me, the known alternatives look even stranger. See Sec. 5.3 of my https://arxiv.org/abs/1703.08341

 

[Demystifier]

And yet there is no mathematically rigorous QFT in 3+1 dimensions.

I would say that QFT on 3+1 dimensional lattice is rigorous.

 

[name123]

Not sure how to respond to this. Given that relativity of simultaneity is one of the most famous things about relativity, if you claim the opposite you need to explain, not ask questions.

The concept of absolute simultaneity in reality is compatible with relativity (because Presentism is), but there would be no experimental means of establishing it (it would be a metaphysical concept). If simultaneity was given an experimental definition, it would be relative, and observers could disagree on what was simultaneous.

 

[PeterDonis]

How is it not ?

It's easy for people to think they disagree if they are using words to mean different things.

If "absolute simultaneity" means "there is a preferred notion of simultaneity built into the laws of physics", then that is incompatible with "relativity" if "relativity" means "the principle of relativity", i.e., "there is no preferred inertial frame built into the laws of physics".

But if "relativity" only means "we observe Lorentz invariance to the accuracy we can measure", then that is not necessarily incompatible with "absolute simultaneity", since one can always claim that the preferred notion of simultaneity is in principle unobservable. But doing that means you are talking about LET, which is a banned topic here on PF.

 

[LeandroMdO]

I would say that QFT on 3+1 dimensional lattice is rigorous.

It's worth pointing out that lattice theories are defined in Euclidean space, via the analogy between the path integral and the partition function from thermodynamics. For the most part, this means that such theories are not relativistic quantum lattice theories at all, but rather the classical statistical mechanics of discretized field theories. These theories are of course related to their quantum counterparts by a Wick rotation, which means a great deal of information about the continuum Minkowski theory is encoded in statistical properties of the classical ensembles, and this can be extracted if you know where to look for it. Whether or not this procedure is rigorous is a separate question, however.

That said, I don't see the relevance of the mathematical difficulties with rigorous quantum field theory for the present topic. Aside from the particular example you chose, the central message that QFT ought to make sense at least as an approximate effective theory for something more fundamental is essentially unimpeachable.

 

[vanhees71]

And yet there is no mathematically rigorous QFT in 3+1 dimensions.

There is no mathematically rigorous QFT beyond the level of an effective theory, and indeed it's pretty likely that at one energy scale the Standard Model fails, and one needs at least another effective theory which works there.

 

[Blue Scallop]

The preparation would determine the correlation between the two, but not the polarisation of B's photon.
Prior to measurement of A's photon, the polarisation of B's photon would be undetermined.
Upon measurement of the polarisation of A's photon the polarisation of B's photon would no longer be undetermined.

So how did the measurement at A's detector determine the polarisation of B's photon ?

vanhees71.. may I know what is your answer to the above question by name123?
I'm kinda confused about your position.. do you believe that in entanglement experiments like Clauser and Aspect experiments.. it's due to common causes like red and blue socks being determined from the origin? But in the experiments that violate Bell's Inequality, there were correlations. Or maybe you believed there were correlations but it was not "non-local" because there is no reality.. hence there is nothing to be local about (so spooky at distance is false)? (if so, then this is reasonable). Or do you indeed believe Bell's Inequality is violated due to some loophole that still proves it's like red and blue socks from the initial preparation (this was actually believed by Einstein (the EPR arguments) but disproven by Bell's Experiments that showed there were really correlations even if the entangled particles were light years apart).

 

[Boing3000]

If "absolute simultaneity" means "there is a preferred notion of simultaneity built into the laws of physics", then that is incompatible with "relativity" if "relativity" means "the principle of relativity", i.e., "there is no preferred inertial frame built into the laws of physics".

There is no "frame" involved here (and even less so a "preferred" one), and that maybe the reason of the confusion. "Absolute simultaneity" is defined by clocks once synchronized displaying the same time. Proper time is an absolute notion in relativity, and all frames agrees with that.

But if "relativity" only means "we observe Lorentz invariance to the accuracy we can measure", then that is not necessarily incompatible with "absolute simultaneity", since one can always claim that the preferred notion of simultaneity is in principle unobservable. But doing that means you are talking about LET, which is a banned topic here on PF.

I had to google LET to find out what you meant. And the "simultaneity" of entanglement correlation have nothing to do with LET (an obviously crackpot notion)

An more interesting case for experimentation would be if there are entangled properties that evolves with time. Spin orientation don't, but is there any that does ?

 

[martinbn]

To call this absolute simultaneity is very misleading. And what do you mean by clocks once synchronized? Once in what frame?

 

[Boing3000]

To call this absolute simultaneity is very misleading. And what do you mean by clocks once synchronized? Once in what frame?

Any "frame" of the one event called entanglement. The common start of the two world line.

 

[vanhees71]

vanhees71.. may I know what is your answer to the above question by name123?
I'm kinda confused about your position.. do you believe that in entanglement experiments like Clauser and Aspect experiments.. it's due to common causes like red and blue socks being determined from the origin? But in the experiments that violate Bell's Inequality, there were correlations. Or maybe you believed there were correlations but it was not "non-local" because there is no reality.. hence there is nothing to be local about (so spooky at distance is false)? (if so, then this is reasonable). Or do you indeed believe Bell's Inequality is violated due to some loophole that still proves it's like red and blue socks from the initial preparation (this was actually believed by Einstein (the EPR arguments) but disproven by Bell's Experiments that showed there were really correlations even if the entangled particles were light years apart).

This is a very good example for why I thik that quantum states are better interpreted in an epistemic sense and why the collapse hypothesis leads to problems with locality and causality.

My view is that the correlations, as all probabilistic relations of quantum systems, are described by the state of the system, and the state of the system is determined by preparation. The preparation in that case is when the entangled photon pair is created (e.g., by parametric downconversion by shooting a laser beam into a crystal). When A measures her photon's polarization state (her photon is defined by that it is registered by the detector at A's place), she immediately also knows the polarization state of B's photon (his photon is defined by that it is registered by the detector at B's place, which can be very far distant from A's detector). For B nothing has changed. He simply expects an unpolarized photon and gets with 50% probability the one or the other polarization when he measures it.

Let's now assume that A's detector is very close to the photon source, and B's very far, such that A measures her photon earlier than B. In other words, the measurement processes are assumed to happen as time-like separated events. Then "collapse" happens definitely at different times for A than for B: A changes the state of the photon pair due to her measurement result much earlier than B. Still, there is no contradiction by what's known to A and B concerning the outcome of their mesaurements. Both A's and B's photons are exactly unpolarized, i.e., the polarization state if maximally indetermined.

There's, of course, also no problem when the measurement processes are realized at space-like distances. Then you can always find a reference frame, where A and B register their result simultaneously or another reference frame, where A registers her result before B or again another frame, where B registers his result before A. Still there's no contradiction, because both, A and B always just find that their photons sent from the source of entangled photon pairs are precisely unpolarized.

When A and B compare their measurement protocols (always keeping detailed track about the time, when they registered their measurement outcome to be sure to relate always the pairs which where created together at the source), they find in any case the 100% correlations due to the preparation of the photon pairs in the polarization-entangled state.

Of course, here I made two assumptions: (a) the polarization measurements are local events as described by standard QED and thus the linked-cluster principle is valid, i.e., A's measurement cannot instantaneously affect B's photon and/or measurement apparatus (implied by microcausality) and (b) that all there is possible to be known about photons is what is described by quantum states, and since this is probabilistic knowledge (some may think only) it refers to ensembles of equally prepared quantum systems, i.e., the probabilistic information described by the prepared state can only be tested by collecting "enough statistics", i.e., using a sufficiently large ensemble.

The problems start, whenever you try to give more meaning to the quantum state then is implied by this minimal interpretation. Some think (in the past Einstein and Schrödinger were the most prominent physicists to do so) that this is not a complete description of nature since "in reality" (whatever "reality" means to them) all possible observables should have determined values always. It's not completely ruled out that maybe somebody one day finds some satisfactory theory, where this is the case, but Bell's work and the empirical precise findings with respect to it, imply that such a deterministic hidden-variable theory must be non-local, and so far there seems not to be a satisfactory such kind of theory in the relativistic realm.

 

[Blue Scallop]

This is a very good example for why I thik that quantum states are better interpreted in an epistemic sense and why the collapse hypothesis leads to problems with locality and causality.

My view is that the correlations, as all probabilistic relations of quantum systems, are described by the state of the system, and the state of the system is determined by preparation. The preparation in that case is when the entangled photon pair is created (e.g., by parametric downconversion by shooting a laser beam into a crystal). When A measures her photon's polarization state (her photon is defined by that it is registered by the detector at A's place), she immediately also knows the polarization state of B's photon (his photon is defined by that it is registered by the detector at B's place, which can be very far distant from A's detector). For B nothing has changed. He simply expects an unpolarized photon and gets with 50% probability the one or the other polarization when he measures it.

Let's now assume that A's detector is very close to the photon source, and B's very far, such that A measures her photon earlier than B. In other words, the measurement processes are assumed to happen as time-like separated events. Then "collapse" happens definitely at different times for A than for B: A changes the state of the photon pair due to her measurement result much earlier than B. Still, there is no contradiction by what's known to A and B concerning the outcome of their mesaurements. Both A's and B's photons are exactly unpolarized, i.e., the polarization state if maximally indetermined.

There's, of course, also no problem when the measurement processes are realized at space-like distances. Then you can always find a reference frame, where A and B register their result simultaneously or another reference frame, where A registers her result before B or again another frame, where B registers his result before A. Still there's no contradiction, because both, A and B always just find that their photons sent from the source of entangled photon pairs are precisely unpolarized.

When A and B compare their measurement protocols (always keeping detailed track about the time, when they registered their measurement outcome to be sure to relate always the pairs which where created together at the source), they find in any case the 100% correlations due to the preparation of the photon pairs in the polarization-entangled state.

Of course, here I made two assumptions: (a) the polarization measurements are local events as described by standard QED and thus the linked-cluster principle is valid, i.e., A's measurement cannot instantaneously affect B's photon and/or measurement apparatus (implied by microcausality) and (b) that all there is possible to be known about photons is what is described by quantum states, and since this is probabilistic knowledge (some may think only) it refers to ensembles of equally prepared quantum systems, i.e., the probabilistic information described by the prepared state can only be tested by collecting "enough statistics", i.e., using a sufficiently large ensemble.

The problems start, whenever you try to give more meaning to the quantum state then is implied by this minimal interpretation. Some think (in the past Einstein and Schrödinger were the most prominent physicists to do so) that this is not a complete description of nature since "in reality" (whatever "reality" means to them) all possible observables should have determined values always. It's not completely ruled out that maybe somebody one day finds some satisfactory theory, where this is the case, but Bell's work and the empirical precise findings with respect to it, imply that such a deterministic hidden-variable theory must be non-local, and so far there seems not to be a satisfactory such kind of theory in the relativistic realm.

Yes. Thanks for the details. Good to know at least you still believe in Bell's Theorem. I thought you were arguing in terms of Einstein EPR days where Einstein believed there was no correlations and there were hidden variables (akin to the red and blue socks being determined from beginning).

And yes. It's so difficult to visualize the wave function as ontic as this can violate relativity (except if one used the banned LET).

Whatever, the correlations in QM still work in QFT.. is it not.. what experiments in QFT (or QED) where there are long distant correlations like in the Aspect Experiments? Anyone knows?

Also remember the wave function in MWI lives not in physical space but in higher configuration space. And Demystifier calls it alocal...

 

[martinbn]

Any "frame" of the one event called entanglement. The common start of the two world line.

This is even less clear. What is a frame (in quotes) of one event?

 

[Boing3000]

This is even less clear. What is a frame (in quotes) of one event?

In any frame, there is only one event where entanglement start. I cannot be more clearer sorry.

 

[PeterDonis]

"Absolute simultaneity" is defined by clocks once synchronized displaying the same time.

When the clocks are spatially separated, how do you know they display "the same time"? Giving meaning to that statement requires a simultaneity convention, which is equivalent to choosing an inertial frame.

the "simultaneity" of entanglement correlation

What does this mean?

 

[PeterDonis]

In any frame, there is only one event where entanglement start

But then the entangled particles move apart, so they are spatially separated. See my previous post.

 

[Boing3000]

When the clocks are spatially separated, how do you know they display "the same time"?

Know ? I "know" it because I can compute any particle 4D path (and proper time along it), and pick the event pair whee both clock display the same time.

Giving meaning to that statement requires a simultaneity convention, which is equivalent to choosing an inertial frame.

No. The simultaneity of convention is when synchronized clock display the same time. There is no preferred frame involved here. Any frame will do.

What does this mean?

It means both measure can done at a perfectly valid absolute proper time. Simultaneity didn't disappear because different frame have a different perspective on some ordering. Events only happen once at one place, that's called causality.

 

[PeterDonis]

I "know" it because I can compute any particle 4D path (and proper time along it), and pick the event pair whee both clock display the same time.

Ah, ok. But then you have to deal with the twin paradox. You can separate your clocks and then bring them back together, and they won't read the same time--they will be sitting right next to each other, but reading different times. So which one is the "right" time--the time that is "absolute"?

 

[tom.stoer]

The problems start, whenever you try to give more meaning to the quantum state then is implied by this minimal interpretation. Some think (in the past Einstein and Schrödinger were the most prominent physicists to do so) that this is not a complete description of nature since "in reality" (whatever "reality" means to them) all possible observables should have determined values always. It's not completely ruled out that maybe somebody one day finds some satisfactory theory, where this is the case, but Bell's work and the empirical precise findings with respect to it, imply that such a deterministic hidden-variable theory must be non-local, and so far there seems not to be a satisfactory such kind of theory in the relativistic realm.

That's why I think that the only mathematical object which can be "real" in the sense that it "faithfully represents external reality" is the state vector itself.

The definition of "external reality" is quite simple. This morning I got up and went to the bathroom. I believe the bathroom did exist all over night as part of "external reality"; it did not "become real" by being observed or used in the morning. Newton's equations of motion, conservation of energy etc. describing or predicting the continuously existence of the bathroom, and the bathroom itself have - for me - more than a pure epistemic meaning.

When talking about the quantum world we have to accept different laws of nature, but we need not give up this simple view on "external reality". Quantum mechanics is - in contrast to what many lectures and textbooks are trying to explain - perfectly consistent with this paradigm, once we accept that the mathematical entity we have to use as a "representation of external reality" is not a collection of classical properties like position, momentum etc. but the state vector itself.

In that sense Everett's approach, upgraded by decoherence, is a rather logical consequence of a simple philosophical position plus standard quantum mechanics. The only problem is to accept the weird consequences.

For me, all statements that quantum mechanics must not or cannot talk about "external reality" are fundamentally flawed. This is Bohr's legacy which Everett at al. try to overcome.

 

[martinbn]

Know ? I "know" it because I can compute any particle 4D path (and proper time along it), and pick the event pair whee both clock display the same time.No. The simultaneity of convention is when synchronized clock display the same time. There is no preferred frame involved here. Any frame will do.It means both measure can done at a perfectly valid absolute proper time. Simultaneity didn't disappear because different frame have a different perspective on some ordering. Events only happen once at one place, that's called causality.

Your terminology is very strange. For example, if we take my world line that of my great-great grandfather, when he was born his clock showed 0, when I was born so did mine. You would say that these two events a simultaneous, then when he was 10year, and when I was 10year, are two events that are simultaneous as well.

 

[LeandroMdO]

In that sense Everett's approach, upgraded by decoherence, is a rather logical consequence of a simple philosophical position plus standard quantum mechanics. The only problem is to accept the weird consequences.

The problem with Everettian accounts is not weirdness. I can accept weirdness, and so can most physicists. The problem is that at this moment the role of probability in the interpretation is still unclear. The interpretation can be kindasorta described in words, but no one has, to my knowledge, shown a convincing derivation that observers in this interpretation should split in such a way that, if they perform experiments, Born's rule is observed. Though it's been almost 30 years, the situation is not qualitatively different from the one described here (also see a small, more recent update here).

 

[tom.stoer]

The problem is that at this moment the role of probability in the [Everett] interpretation is still unclear ... no one has, to my knowledge, shown a convincing derivation that observers in this interpretation should split in such a way that, if they perform experiments, Born's rule is observed.

I agree in principle that there are open issues.

But from my perspective the arguments against Everett's approach are not more convincing than the arguments supporting his relative state interpretation. So for me it's still draw.

In addition it's not only about one interpretation but about a comparison of different interpretations. And here I must say that there are some rather ridiculous interpretations on the market. Really outstanding is the minimalistic interpretation saying that "everything works fine and we must not care about how and why". I clearly prefer an ambitious position with some open issues we can work on instead of no position with only open issues we should not even think about.

 

[SPW]

As scientists we need to keep philosophical interpretations of unexplained phenomena to a minimum. Keep an open mind, question everything & sooner or later, usually much later, the answers will come. "Betting" on the probability of an unquantifiable observation adds nothing to the proof. I suppose it puts me in the "shut up and calculate" club. As Einstien once said "God does not play dice", by which he meant there must be something we don't yet know which lead to apparently paradoxical outcomes...

 

[tom.stoer]

As scientists we need to keep philosophical interpretations of unexplained phenomena to a minimum. Keep an open mind, question everything & sooner or later, usually much later, the answers will come. ... I suppose it puts me in the "shut up and calculate" club.

It doesn't b/c it seems that you are looking for explanations. :-)

 

[tom.stoer]

with Everettian accounts is not weirdness. I can accept weirdness, and so can most physicists. The problem is that at this moment the role of probability in the interpretation is still unclear. The interpretation can be kindasorta described in words, but no one has, to my knowledge, shown a convincing derivation that observers in this interpretation should split in such a way that, if they perform experiments, Born's rule is observed.

Though it's been almost 30 years, the situation is not qualitatively different from the one described here (also see a small, more recent update here).

Thanks for mentioning Kent.

Here's his introduction I fully agree with.

 

[vanhees71]

That's why I think that the only mathematical object which can be "real" in the sense that it "faithfully represents external reality" is the state vector itself.

The definition of "external reality" is quite simple. This morning I got up and went to the bathroom. I believe the bathroom did exist all over night as part of "external reality"; it did not "become real" by being observed or used in the morning. Newton's equations of motion, conservation of energy etc. describing or predicting the continuously existence of the bathroom, and the bathroom itself have - for me - more than a pure epistemic meaning.

When talking about the quantum world we have to accept different laws of nature, but we need not give up this simple view on "external reality". Quantum mechanics is - in contrast to what many lectures and textbooks are trying to explain - perfectly consistent with this paradigm, once we accept that the mathematical entity we have to use as a "representation of external reality" is not a collection of classical properties like position, momentum etc. but the state vector itself.

I couldn't agree more with this part of your posting. Indeed QT doesn't contradict this "simple philosophical position" at all. Only the notion of what a state is. In classical physics it's described by a point in phase space, in QT by the statistical operator of the system. The main difference between the classical description and quantum description is that in the classical case by the complete knowledge of the state one knows the values of all possible observables of this system (i.e., the values of all observables are always determined), while in QT only a well-defined class of observables take determined values, when any (pure or mixed) state of the system is prepared.

In that sense Everett's approach, upgraded by decoherence, is a rather logical consequence of a simple philosophical position plus standard quantum mechanics. The only problem is to accept the weird consequences.

For me, all statements that quantum mechanics must not or cannot talk about "external reality" are fundamentally flawed. This is Bohr's legacy which Everett at al. try to overcome.

I never understood the point of Everett's interpretation, let alone even in this modern form as "many-worlds interpretation" with the world branching into many parallel universes at any observation someone does on anything he or she observes, which also brings in the famous funny question by Bell, whether the observation of something by an amoeba is enough to cause a collapse (or in the MW interpretation the branching of the world) or whether you need some "more intelligent being" like a dog, monkey or human (who knows, how intelligent an amoeba maybe might be, but we just don't know it ;-)).

I think the logical consequence is the minimal interpretation, according to which QT is a formalism to describe probabilistically what we observe objectively in this real world, which you describe above. Indeed, as the name says, the quantum state describes the state of the system, and it's not a state vector but the statistical operator (or equivalently in the case of pure states, which are in general quite rare and must be carefully prepared in the lab, rays in Hilbert space).

 

[Boing3000]

Ah, ok. But then you have to deal with the twin paradox. You can separate your clocks and then bring them back together, and they won't read the same time--they will be sitting right next to each other, but reading different times. So which one is the "right" time--the time that is "absolute"?

This won't happens for photon, but for entangled electron, this would give an interesting setup, if half the electron are sent into an accelerator ring then slowed down for comparison (if entanglement could resit such a trip).
There is no problem for them sitting next to each other with at a different age. But then measuring the youngest batch before the oldest should lead to different result then the doing it the opposite.

 

[Boing3000]

Your terminology is very strange. For example, if we take my world line that of my great-great grandfather, when he was born his clock showed 0, when I was born so did mine. You would say that these two events a simultaneous, then when he was 10year, and when I was 10year, are two events that are simultaneous as well.

That would be, if both clock would have been synchronized. Have they ?
Events are absolute, and clock measure time. I had no idea why a relativist would have a problem with that...