Different interpretations? No, different theories

In summary, the conversation discusses the differences between Everett's paper on the "Relative State" Formulation of Quantum Mechanics and other interpretations such as the Copenhagen interpretation. It is argued that these are not just different interpretations, but different theories that give different predictions. The issue of testable predictions is also discussed, with the conclusion that in order to promote an interpretation to a theory, it must have different predictions that can be tested in principle. The concept of Many Worlds is also brought up, with the idea that it may be inevitable when analyzing observers and measuring devices using the Rules of Quantum Mechanics.
  • #36
Fredrik, I think this way you get something different, not MWI.
 
Physics news on Phys.org
  • #37
martinbn said:
Fredrik, I think this way you get something different, not MWI.
It's definitely not Everett's MWI, but I think many worlds are unavoidable once we make the assumption that the state vector of the universe represents all the properties of the universe. The argument is what I said in the quote in #16.

Is there even such a thing as "Everett's MWI" or "The MWI"? All attempts to recover the Born rule have either failed or required additional assumptions that are essentially equivalent to the Born rule.
 
Last edited:
  • #38
Fredrik said:
It's definitely not Everett's MWI, but I think many worlds are unavoidable once we make the assumption that the state vector of the universe represents all the properties of the universe. The argument is what I said in the quote in #16.
Or, to be stated differently, if everything is quantum mechanical (sounds more unambiguous).
 
  • #39
Fredrik said:
It's definitely not Everett's MWI, but I think many worlds are unavoidable once we make the assumption that the state vector of the universe represents all the properties of the universe. The argument is what I said in the quote in #16.

Is there even such a thing as "Everett's MWI" or "The MWI"? All attempts to recover the Born rule have either failed or required additional assumptions that are essentially equivalent to the Born rule.

On the other hand, there is a sense in which the Born rule in a limiting sense is built into the definition of the hilbert space for quantum mechanics. Two wave functions [itex]\Psi_1(x)[/itex] and [itex]\Psi_2(x)[/itex] are considered equal, as members of the hilbert space, if [itex]\int |\delta \Psi|^2 dx = 0[/itex], where [itex]\delta \Psi = \Psi_2 - \Psi_1[/itex]. So if you have a branch of the wave function whose norm is 0, you can ignore it.

One of the arguments I've heard for recovering the Born rule is to consider the quantum problem of recording an infinite sequence of measurements of some experiment that has amplitude [itex]\dfrac{1}{\sqrt{2}}[/itex]. The argument is that the Hilbert space measure of the histories where the frequencies don't equal [itex]\dfrac{1}{2}[/itex] is zero, and so they don't exist (the definition of Hilbert space mods out by things of measure zero).

The mathematics of this derivation is pretty screwy, though. To get a "wave function" that can record an infinite number of measurements, I think you need something like a nonseparable Hilbert space.
 
  • #41
JK423 said:
Or, to be stated differently, if everything is quantum mechanical (sounds more unambiguous).
Does it really sound more unambiguous? I don't think it's clear what "is quantum mechanical" means. If we want your statement to mean the same as mine, plus that the entire universe can be assigned a state vector, then I think we would need to explain the words "is quantum mechanical" by a statement like mine.

Without an explanation, your statement could be interpreted as only saying that every subsystem of the universe including the universe itself can be assigned a Hilbert space and a state vector. Since we already know that an atom can be assigned a Hilbert space and a state vector, and don't know if that state vector describes all the properties of that atom, it's not clear that "the universe has a state vector" implies that that state vector describes all the properties of the universe.

I agree that my statement isn't as precise as we'd want it to be, but I think that some ambiguity is unavoidable when we're talking about the meaning of mathematical terms in the theory.
 
  • #42
stevendaryl said:
One of the arguments I've heard for recovering the Born rule is to consider the quantum problem of recording an infinite sequence of measurements of some experiment that has amplitude [itex]\dfrac{1}{\sqrt{2}}[/itex]. The argument is that the Hilbert space measure of the histories where the frequencies don't equal [itex]\dfrac{1}{2}[/itex] is zero, and so they don't exist (the definition of Hilbert space mods out by things of measure zero).

As I said in another post, there's a similar philosophical problem with understanding the empirical meaning of classical probability. A coin flip having probability [itex]\dfrac{1}{2}[/itex] doesn't mean that when you flip the coin 100 times, you're going to get 50 heads. But you can say that the set of all possible histories in which an infinite sequence of coin flips don't yield 50% heads has measure zero. So you can understand the meaning of "probability 50%" for a single flip in terms of "probability 0" for an infinite sequence of flips.
 
  • #43
JK423 said:
Regarding the non-circular derivation of Born's rule, Zurek has done serious work on this with the concept "envariance". He seems to have derived Born's rule non-circularly:

http://prl.aps.org/abstract/PRL/v90/i12/e120404
http://pra.aps.org/abstract/PRA/v71/i5/e052105

Are you all aware of this?
Yes. He uses that the Hilbert space of a composite system is the tensor product of the Hilbert spaces of the subsystems. (I think all of these derivations do). I think the best motivation for that assumption is the argument made by Aerts and Daubechies in 1978. (pdf). I believe that all of their axioms about propositional systems can be derived from the usual Hilbert space version of QM with the Born rule.
 
  • #44
Fredrik said:
Does it really sound more unambiguous? I don't think it's clear what "is quantum mechanical" means. If we want your statement to mean the same as mine, plus that the entire universe can be assigned a state vector, then I think we would need to explain the words "is quantum mechanical" by a statement like mine.

Without an explanation, your statement could be interpreted as only saying that every subsystem of the universe including the universe itself can be assigned a Hilbert space and a state vector. Since we already know that an atom can be assigned a Hilbert space and a state vector, and don't know if that state vector describes all the properties of that atom, it's not clear that "the universe has a state vector" implies that that state vector describes all the properties of the universe.

I agree that my statement isn't as precise as we'd want it to be, but I think that some ambiguity is unavoidable when we're talking about the meaning of mathematical terms in the theory.
Yes, ok, we agree. We just ascribe different meaning to the words. When i say "everything is quantum mechanical" i mean that quantum mechanics is "all there is", so there are no extra properties not described by QM.
 
  • #45
Fredrik said:
Yes. He uses that the Hilbert space of a composite system is the tensor product of the Hilbert spaces of the subsystems. (I think all of these derivations do). I think the best motivation for that assumption is the argument made by Aerts and Daubechies in 1978. (pdf). I believe that all of their axioms about propositional systems can be derived from the usual Hilbert space version of QM with the Born rule.
So, since you are aware of this, are there any objections against it? Why isn't it considered the ultimate proof that Born's rule can be derived from quantum theory? For some reason you don't accept it.
 
  • #46
JK423 said:
So, since you are aware of this, are there any objections against it? Why isn't it considered the ultimate proof that Born's rule can be derived from quantum theory? For some reason you don't accept it.
What I said is the objection. The argument is circular. It's roughly like this:
QM with the Born rule → the rules for propositional systems → QM without the Born rule + tensor products → QM with the Born rule​
What Zurek did is perhaps the ultimate proof of that last implication. But the relevant implication
QM without the Born rule → QM with the Born rule​
has never been proved, and I don't think it can be done.
 
  • #47
Fredrik said:
What I said is the objection. The argument is circular. It's roughly like this:
QM with the Born rule → the rules for propositional systems → QM without the Born rule + tensor products → QM with the Born rule​
What Zurek did is perhaps the ultimate proof of that last implication. But the relevant implication
QM without the Born rule → QM with the Born rule​
has never been proved, and I don't think it can be done.

Thank you for the explanation.

But, isn't the mathematical structure, underlying quantum theory, supposed to be taken axiomatically? Why do you say that it's based on Born's rule? For example, i could argue that, based on how the interactions of nature act on subsystems (which is experimentally confirmed) i am led to axiomatically consider the tensor product structure of subsystems. I do not pre-assume Born's rule, it's just the way nature works. Isn't that analogous to asking, "Why Hilbert space?" ?
 
Last edited:
  • #48
JK423, can you reply to my post #26? Thanks!
 
  • #49
I have! #28. I was waiting for your reply! :-p
 
  • #50
The central idea of a MWI is that the state vector of the universe describes everything that's happening. If it does, then we should be able to explain why experimenters get results that are consistent with the Born rule. To even begin do something like that, we need to add assumptions about how to deal with subsystems. So we add an assumption (the tensor product stuff) that can be derived from the Born rule. Then we can try to recover the Born rule from that.

Zurek has claimed that he has done that. I have been really bothered by the apparent circularity that comes from the fact that the tensor product stuff can be derived using the Born rule. But right now I'm not sure that's even a problem. I need to think about this some more.

There is however another issue with Zurek's derivation and similar attempts, the issue of why states are to be associated with probabilities in the first place. That may be a more serious problem. I need to think about that too.
 
  • #51
To my current understanding, the tensor product structure (TPS) has nothing to do with Born's rule. Since both Born's rule and TPS are correct, it's natural that you may derive one from the other (since they seem not to be independent), but it's wrong to say that TPS is a consequence of Born's rule since TPS would be true even if Born wasn't born! The opposite would be correct.The mathematical structure (i.e. TPS) i think is more fundamental than empirical rules (i.e. Born's rule).

The real problem, i think, is the other thing that you say, that there has to be some association with probabilities. My understanding ends here, i need to study Zurek's work more to understand what he's done.
 
  • #52
JK423 said:
Does the CI assume that the observer is classical (or better, non-quantum mechanical)? I think yes, because if not then we are lead to Everett's view (which simply says that everything is quantum mechanical).
Niels Bohr liked to emphasize that a measurement by definition has a result. The result is indicated by some component of the measuring device. The possible final states of the indicator component must be easily distinguishable by a human. If not, we wouldn't consider what just happened a "measurement". This means that the experiment must make the quantum state of the indicator component for all practical purposes indistinguishable from a classical superposition.

It doesn't mean that the indicator component doesn't have a quantum state. It just means that if there's an experiment in which it behaves in a noticeably non-classical way, we would consider it a specimen, not a measuring device.

These are statements about what sort of thing we would consider a "measurement". They say very little (if anything at all) about the properties of measuring devices or the domain of validity of quantum mechanics.
 
Last edited:
  • #53
JK423 said:
...it's wrong to say that TPS is a consequence of Born's rule since TPS would be true even if Born wasn't born! The opposite would be correct.The mathematical structure (i.e. TPS) i think is more fundamental than empirical rules (i.e. Born's rule).
What makes you say that? I don't see any reason to think so.

The tensor product stuff isn't even part of any specific quantum theory. It's just a prescription for how to define new quantum theories from existing ones.
 
  • #54
Fredrik said:
Niels Bohr liked to emphasize that a measurement by definition has a result. The result is indicated by some component of the measuring device. The possible final states of the indicator component must be easily distinguishable by a human. If not, we wouldn't consider what just happened a "measurement". This means that the experiment must make the quantum state of the indicator component for all practical purposes indistinguishable from a classical superposition.

It doesn't mean that the indicator component doesn't have a quantum state. It just means that if there's an experiment in which such an indicator component behaves in a noticeably non-classical way, we would consider it a specimen, not a measuring device.

These are statements about what sort of thing we would consider a "measurement". It says very little (if anything at all) about the properties of measuring devices or the domain of validity of quantum mechanics.

It seems like you're right, i confirmed this from a work of Zurek's where he explicitly mentions that

Indeed, since the ultimate components of classical objects are quantum, Bohr emphasized that the boundary must be moveable, so that even the human nervous system could be regarded as quantum, provided that suitable classical devices to detect its quantum features were available.

This is a really strange point of view! Bohr was Everettian without even knowing it :-p.

Fredrik said:
What makes you say that? I don't see any reason to think so.
The tensor product stuff isn't even part of any specific quantum theory. It's just a prescription for how to define new quantum theories from existing ones.

In nature, particle states interact via TPS. Why is this fact dependent on what you will find IF you make measurements? The quantum state exists and evolves even if you don't measure it, it's independent of the Born's rule. Now, the fact that IF you measure it you will indeed find Born's rule is due to the fact that in nature particles interact the way they do, not the other way around.
 
  • #55
JK423 said:
Before trying to answer your question, i need to know if the following hypothesis that i make is correct:
Does the CI assume that the observer is classical (or better, non-quantum mechanical)? I think yes, because if not then we are lead to Everett's view (which simply says that everything is quantum mechanical). If the observer is assumed to be non-quantum mechanical, then doesn't this mean that quantum mechanics fail at some point? Isn't the failure of quantum mechanics, in the description of the observer, in principle testable?
Yes is my answer to all these questions. But I would still like to challenge you to propose a CONCRETE thought (gedanken) experiment where the difference would be seen explicitly.
 
  • #56
JK423 said:
I have! #28. I was waiting for your reply! :-p
Sorry, I haven't noticed it. See my reply in the post #55 above!
 
  • #57
Demystifier said:
Yes is my answer to all these questions. But I would still like to challenge you to propose a CONCRETE thought (gedanken) experiment where the difference would be seen explicitly.
I would give it a lot of thought if that was the case, but as you can see from the posts above, it's not. Copenhagen interpretation does not say that the observer is non-quantum mechanical. For CI, an observer is quantum if he is being observed but something else if he is observing. All this is so vague that i am not sure what CI is about, and i don't know what to prove! In order to prove something, i need to know the rules.. and the rules seem so vague in CI.
 
  • #58
JK423 said:
I would give it a lot of thought if that was the case, but as you can see from the posts above, it's not. Copenhagen interpretation does not say that the observer is non-quantum mechanical. For CI, an observer is quantum if he is being observed but something else if he is observing. All this is so vague that i am not sure what CI is about, and i don't know what to prove! In order to prove something, i need to know the rules.. and the rules seem so vague in CI.
So compared with your first post on this thread, now you have changed your opinion. First you thought that CI is a theory different from MWI, now you think that CI is not even a well defined theory. Am I right?

If so, then I rephrase my challenge. Propose a thought experiment for which the measurable predictions of MWI are unambiguous, while those of CI are not!
 
  • #59
Demystifier said:
So compared with your first post on this thread, now you have changed your opinion. First you thought that CI is a theory different from MWI, now you think that CI is not even a well defined theory. Am I right?

If so, then I rephrase my challenge. Propose a thought experiment for which the measurable predictions of MWI are unambiguous, while those of CI are not!

Yes you're right. To my current understanding CI is just MWI without explicitely saying it.. Otherwise it's nonsense.
How is it possible that an observer collapses wavepackets, but when observed he is quantum and he doesn't actually collapse anything, just unitary evolution?
If you accept that QM holds universally, even to observers, then all this is nonsense.

I need you to give me a clear definition of what CI is.. Can you? Then i'll try to take on your challenge :redface:
This is necessary, because otherwise i may assume something that leads to the response "hey, CI doesn't say that".
 
Last edited:
  • #60
JK423 said:
I need you to give me a clear definition of what CI is.. Can you? Then i'll try to take on your challenge :redface:
This is necessary, because otherwise i may assume something that leads to the response "hey, CI doesn't say that".
There is no definition of the CI that wouldn't make a lot of people go "hey, CI doesn't say that".

JK423 said:
Yes you're right. To my current understanding CI is just MWI without explicitely saying it.. Otherwise it's nonsense.
It's clear that you're making the assumption that QM describes what's happening to the system even at times between state preparation and measurement. There's nothing in QM that forces us to make that assumption. I would say that this assumption is the starting point of a definition of a MWI, so what you're saying sounds to me like "if we assume the MWI, then it's nonsense to also assume something that contradicts it". This is obviously true, but I need to point out that you're making an assumption that isn't necessary.

It's also possible that QM isn't a description of what's actually happening. I would take that as the definition of the CI, and also as the definition of an ensemble interpretation, because that just seems to be a different way to say the same thing. This interpretation says that QM is just a set of rules that assigns probabilities to possible results of experiments. Edit: Since it doesn't make any claims about what is "actually happening", it's questionable if it should be called an interpretation at all.

JK423 said:
How is it possible that an observer collapses wavepackets, but when observed he is quantum and he doesn't actually collapse anything, just unitary evolution?
Unitary evolution only applies to systems that are isolated from their environments. Observers, by definition, are not. If a system has an environment A that's isolated from its environment B, then "system+A" evolves unitarily. This sort of thing is taken into account in decoherence calculations.
 
Last edited:
  • #62
Fredrik said:
There is no definition of the CI that wouldn't make a lot of people go "hey, CI doesn't say that".

Oh boy aren't that the truth as a number of 'discussions' I have had about it show.

Even the MSI I hold to has variations - it seems part of the landscape that within most if not all interpretations you have different schools of thought.

Thanks
Bill
 
  • #63
Fredrik said:
What I said is the objection. The argument is circular. It's roughly like this:
QM with the Born rule → the rules for propositional systems → QM without the Born rule + tensor products → QM with the Born rule​
What Zurek did is perhaps the ultimate proof of that last implication. But the relevant implication
QM without the Born rule → QM with the Born rule​
has never been proved, and I don't think it can be done.

But you do have "QM without the Born rule + tensor products → QM with the Born rule", so if one were happy to start at "QM without the Born rule + tensor products" it wouldn't be circular.
 
Last edited:
  • #64
Fredrik said:
But the relevant implication
QM without the Born rule → QM with the Born rule​
has never been proved, and I don't think it can be done.

Hmmmm.

That depends on what you mean by QM without the Born rule.

If that includes non contextuality then Gleasons Theorem implies the Born rule.

What I think it is fairer to say is the observable axiom (ie the eigenvalues of a Hermition operator give the possible outcomes of an observation) is not sufficient by itself to derive the Born rule - however some reasonable added assumptions such as additivity of expectations or non contextuality are.

Thanks
Bill
 
  • #65
bhobba said:
Hmmmm.

That depends on what you mean by QM without the Born rule.
And that depends on what I mean by "the Born rule". I consider it to be not just the formula, but the statement that the probabilities it assigns correspond to measurement results.

bhobba said:
If that includes non contextuality then Gleasons Theorem implies the Born rule.
Gleason's theorem tells us that probability measures on the lattice of closed linear subspaces of the Hilbert space correspond bijectively to state operators, and that the formula for that bijection is a generalized version of the formula that makes up the purely mathematical part of the Born rule. So when we drop the Born rule, what we're really dropping isn't the formula, but the assumption about in what way the subspaces of the Hilbert space are significant.
 
Last edited:
  • #66
JK423 said:
I would give it a lot of thought if that was the case, but as you can see from the posts above, it's not. Copenhagen interpretation does not say that the observer is non-quantum mechanical. For CI, an observer is quantum if he is being observed but something else if he is observing. All this is so vague that i am not sure what CI is about, and i don't know what to prove! In order to prove something, i need to know the rules.. and the rules seem so vague in CI.
That's because you after a classical explanation of quantum behavior. Of course this can't be done and no such explanation exists except fancy, religious ideas like the MWI.

As usual, most of the mystery is reflected and highlighted in the double slit experiment - esp. the one done with large molecules like C60. This experiment is the best evidence to date that the observed classical properties and behavior at our scales(of table and chairs and walls) are JUST a manifestation of an underlying reality.

You may continue to imagine a billion different worlds and magical guiding waves that have access to the SE, but the truth is only one - classical behavior is a completely contextual secondary manifestation.

http://www.uam.es/personal_pdi/ciencias/jcuevas/Teaching/double-slit-C60.pdf

Quantum interference experiments with large molecules
Olaf Nairz, Markus Arndt, and Anton Zeilinger
Institut fur Experimentalphysik, Universitat Wien
 
Last edited by a moderator:
  • #67
Maui said:
That's because you after a classical explanation of quantum behavior. Of course this can't be done and no such explanation exists except fancy, religious ideas like the MWI.

Calling it "religious" is pretty ridiculous. You might as well call it "poopy-headed" for all the information it conveys.
 
  • #68
atyy said:
But you do have "QM without the Born rule + tensor products → QM with the Born rule", so if one were happy to start at "QM without the Born rule + tensor products" it wouldn't be circular.
Yes, I think that last part is correct. But now I think that my statement that "QM without the Born rule + tensor products → QM with the Born rule" isn't accurate enough. I think that the real problem with Zurek's derivation is that it relies not only on "QM without the Born rule + tensor products", but also on an assumption about how the probabilities assigned by the formula correspond to measurement results.
 
  • #69
Maui said:
That's because you after a classical explanation of quantum behavior. Of course this can't be done and no such explanation exists except fancy, religious ideas like the MWI.

MWI is about as far from a "classical explanation" as you could possibly get. Really, your comment makes no sense whatsoever. In my opinion.
 
  • #70
stevendaryl said:
Calling it "religious" is pretty ridiculous. You might as well call it "poopy-headed" for all the information it conveys.
Religious dogma is usually defined as something that is taken on faith without experimental evidence. Like the MWI. Sorry if i hurt religious feelings about the trillion worlds.
MWI is about as far from a "classical explanation" as you could possibly get. Really, your comment makes no sense whatsoever. In my opinion.
Your commment makes no sense either, as I didn't say that the MWI was a classical explantion, but that JK423 seemed to be intent on reducing the CI to a complete account of classical behavior, whereas as Bohr himself stated - the role of physics is what we can say about nature, not how nature is. The purpose of the CI is not to make sense to you or to JK423, but to provide the best framework for making predictions. Yes, it doesn't make sense classically, but neither do the other so called "interpretations".
 
Last edited:

Similar threads

I Are there any accepted relativistic interpretations of quantum mechanics?
2
Replies
45
Views
4K
I QFT made Bohmian mechanics a non-starter: missed opportunities?
11
Replies
376
Views
16K
I Why can't there be a Universal Interpretation of Quantum Mechanics?
Replies
14
Views
2K
I Can I take the transactional interpretation seriously?
Replies
4
Views
1K
A Game theory, quantum physics and the Bell inequality (paper)
Replies
2
Views
2K
I Understanding the Fundamental Difference in Interpretations of QM
3
Replies
76
Views
5K
I For and Against the Copenhagen Interpretation
Replies
28
Views
3K
A Implications of quantum foundations on interpretations of relativity
6
Replies
179
Views
13K
A Loophole-free test of local realism via Hardy's violation
2
Replies
37
Views
4K
I Realistic interpretation of QM
4
Replies
109
Views
9K

Hot Threads

Recent Insights

Back
Top