The Many-Worlds Interpretation of QM

[Maui]

This is not about your description of Kim's experiment. You claim that Kim's experiment refutes the MWI, correct? I just ask you to back this claim up by a reference.

I claim that in the Kim experiement it's the which-path information that causes wavefunction collapse. This is also what the authors claim and is visible from the setup and the results. From it follows that NEITHER detectors NOR interactions between photons in this particular setup creates or destroys the interference pattern.

It seems that since you cannot explain the results of this experiment in terms of the MWI(you would have if you could), you hope you can just skip the technical details of the experiement and appeal to authorities. Physics does not deal with interpretations and their refutals, so you are free to continue to stick to whatever interpretation you think is peer-reviewed or favored by a your chosen group. I am done here.

 

[mfb]

I claim that in the Kim experiement it's the which-path information that causes wavefunction collapse.

To rephrase this, you use a non-MWI interpretation (collapses), and get results different from MWI.
How does this invalidate MWI?

It seems that since you cannot explain the results of this experiment in terms of the MWI(you would have if you could)

Just let the wave function evolve unitarily (and calculate amplitude-squares after the experiment if you like), it will give the same result.

 

[Maui]

To rephrase this, you use a non-MWI interpretation (collapses), and get results different from MWI.
How does this invalidate MWI?

You can control the collapse and when it happens, so yes it's experimentally observed. See the paper i cited earlier.

Just let the wave function evolve unitarily (and calculate amplitude-squares after the experiment if you like), it will give the same result.

This is not an interpretation of the repetive no interference results when which-path information is available. It's a restatement of the shut up and calculate approach, which is most definitely not an interpretation of quantum theory.

 

[Fredrik]

I think that some (many?) opponents of MWI get the impression that supporters claim that, and argue against that self-made claim.

As I wrote that line, I was thinking about the many false claims about what MWI supporters believe that had already been made in this thread, and I was wondering if I was doing the same. That's why I used weak phrases like "it seems to me".

The main reason why it seems that way to me is that even the MWI supporters are talking about "the preferred basis problem" as if it really is a problem. I don't see any reason why someone who isn't assuming that there's only one set of worlds that make up the universe would consider "the preferred basis problem" anything more than an interesting feature.

There's also the fact that no one ever talks about worlds that don't correspond to possible measurement results (not even to dismiss them as irrelevant), or worlds that correspond to a different choice of decomposition of the universe into subsystems. OK, the guy who wrote that article did, but he was using it to argue against the MWI.

 

[Fredrik]

No, there exist interpretations that are not part of the perdictive theory and these interpretations can be tested for viability in experiements.

If you mean that there are interpretations that make different predictions than the theory, then I would consider them different theories. And this is irrelevant to what we've been talking about anyway.

If you mean that experiments can test an assumption that has no effect on the theory's predictions, then this is obviously incorrect.

Look at the quantum eraser by Kim et al and you'll notice how there are detectors(D0--D0) at all possible photon paths all the time but you only get quantum behavior(interference pattern)when the which-way information cannot be obtained(D1 and D2).

So? This is just what the theory predicts. So it's obviously consistent with all interpretations of the theory.

No, the role of the interpretations is to tell us more than the theory(qm). They don't have any other role and they actually tell us a lot more about the inner workings of the world that is not visible in the theory. If one believes it though.

Right. But they don't change the theory's predictions, and experiments only check the accuracy of the predictions. Experiments test theories, not interpretations.

 

[Maui]

If you mean that there are interpretations that make different predictions than the theory, then I would consider them different theories. And this is irrelevant to what we've been talking about anyway.

If you mean that experiments can test an assumption that has no effect on the theory's predictions, then this is obviously incorrect.

I stated clearly what i mean but you keep shifting the focus and misinterpreting my words.

So? This is just what the theory predicts. So it's obviously consistent with all interpretations of the theory.

Incorrect. The interpretation is not the theory(of quantum mechanics). I must have pointed out this very obvious fact more than a dozen times now. If you think otherwise, show me a textbook that talks about the MWI on equal footing with quantum theory. The theory surely doesn't predict > 99 trillion worlds.

Right. But they don't change the theory's predictions, and experiments only check the accuracy of the predictions. Experiments test theories, not interpretations.

This might have been the case in the 1950's but experiments get more sophisticated and their outcomes can only be interpreted as 'which-path information causing wavefunction collapse' or by appealing to superdeterminism, which seems to be the new religion of physics these days.

 

[Fredrik]

I stated clearly what i mean but you keep shifting the focus and misinterpreting my words.

What are you talking about? I haven't shifted anything. The only thing we've been talking about is whether an additional assumption that's added to the theory without changing its predictions can be falsified by experiments. It can't. This is an obvious fact that has has nothing to do with technological advances or how well we understand the theory.

And I don't always know what you mean. What did you mean by an "interpretation that's not part of the predictive theory"? What does it have to do with the possibility of disproving the MWI by experiment?

Look at the quantum eraser by Kim et al and you'll notice how there are detectors(D0--D0) at all possible photon paths all the time but you only get quantum behavior(interference pattern)when the which-way information cannot be obtained(D1 and D2).

So? This is just what the theory predicts. So it's obviously consistent with all interpretations of the theory.

Incorrect. The interpretation is not the theory(of quantum mechanics). I must have pointed out this very obvious fact more than a dozen times now. If you think otherwise, show me a textbook that talks about the MWI on equal footing with quantum theory. The theory surely doesn't predict > 99 trillion worlds.

How am I supposed to answer this? Your answer has nothing to do with what I said, or with your previous statement that I was responding to. You said that we only get quantum behavior when which-way information cannot be obtained. So when I quoted that statement and said that "this" is just what the theory (i.e. QM) predicts, I was obviously talking about the quantum behavior and the which-way information. So when you're quoting that and saying "incorrect", your statement can only be interpreted as saying that QM's prediction is something other than this, so that the experiment actually falsifies QM. I don't think that's what you meant, but that's what you're saying. So it's impossible to respond.

What do you mean by the suggestion that you have pointed out that the theory isn't the interpretation? That's one of the things I've been saying. An interpretation is defined by non-mathematical, non-scientific assumptions that are made in addition to the assumptions that define the theory and determine all the predictions. That's why they aren't testable even in principle. It doesn't matter how sophisticated our experiments get.

"> 99 trillion worlds" is not a prediction. A prediction is an assignment of a probability to a possible result of a measurement.

 

[bhobba]

If I understand it correctly (and it's certainly possible that I don't), it says that decoherence does select a preferred basis, given a decomposition into subsystems. But there's no preferred decomposition. I don't know if the article explains clearly why it considers that a problem. Personally, I think the only "problem" with it is that it prevents us from saying that a preferred basis identifies "the worlds that make up the universe"

Get where you are coming from now.

My opinion is the breaking of a system into subsystems is entirely arbitrary - basically you choose it for convenience. You can choose not to break it into subsystems, or into as many as you like with any boundaries you like. If however you choose a not particularly good one then the interaction Hamiltonian, Lagrangian, or whatever formalism you use is more complicated, however if you break it into subsystems such that the interaction term (or terms) is non existent or at least simple then you have chosen a good one. You would, or should anyway, obtain exactly the same results either way - its just a suitable choice will make life easy.

As far measurements are concerned before the measurement apparatus is used generally there is no interaction so breaking the system into apparatus and what is being measured looks a good choice. Once they interact then the interaction Hamiltonian will single out the preferred basis. I think any breaking of what is being measured and measurement apparatus into subsystems, or even if to break it at all, is arbitrary - you can choose not to do it - or break into anything you like - but your analysis becomes a lot more complicated - or easier - depending on you choice.

Thanks
Bill

 

[bhobba]

No, the role of the interpretations is to tell us more than the theory(qm). They don't have any other role and they actually tell us a lot more about the inner workings of the world that is not visible in the theory. If one believes it though.

Yes - sometimes. But there is also a role for simply mapping the mathematical objects of the theory to something out there. The MSI does that and is not particularly illuminating about inner workings. Others like BM give a lot of detail about those inner workings. I would say the MWI is actually fairly close to the MSI in what it does - it really does take the formalism to its logical conclusion - its just the way it does it is unappealing in its 'weirdness' to many.

Thanks
Bill

 

[bhobba]

I claim that in the Kim experiement it's the which-path information that causes wavefunction collapse.

But what does that have to do with MW since it doesn't have wavefunction collapse.

That said prior to decoherence creating a mixed state whose basis is the new worlds it splits into which path information is used. But be careful here - MWI being a decoherence based interpretation does not consider probabilities assigned by the Born rule to states independent of decoherence to be of any real value - they are called pre-probabilities in such interpretations and are basically of academic use only - they play no real part in the interpretation so which path information is not really the way MWI looks at it.

Thanks
Bill

 

[mfb]

You can control the collapse and when it happens, so yes it's experimentally observed. See the paper i cited earlier.

Sorry, but you need a collapse interpretation to say that. It is obvious that the interpreted results of collapse interpretations are not compatible with MWI. No, there was NO observed process of collapses. Collapses are just a possible interpretation of the results.

The main reason why it seems that way to me is that even the MWI supporters are talking about "the preferred basis problem" as if it really is a problem.

I don't see it as a real problem, but it would be nice to have an obvious way to get bases which have some similarity to classical physics.

There's also the fact that no one ever talks about worlds that don't correspond to possible measurement results (not even to dismiss them as irrelevant)

I'm sure I am not the first one to do that.

or worlds that correspond to a different choice of decomposition of the universe into subsystems

Well, I don't care about them.

 

[HomogenousCow]

MWI seems like a theoretical dead-end to me, it is impossible to test, it adds nothing to the theory and seems like a non-answer to me. How does one mathematically define the phrase "The Universe just splits into two"?

 

[mfb]

MWI seems like a theoretical dead-end to me, it is impossible to test, it adds nothing to the theory and seems like a non-answer to me.

The same can be said about all other interpretations of QM.

How does one mathematically define the phrase "The Universe just splits into two"?

You do not have to. But if you like, with decoherence: In your favorite basis (actually, in many bases), the wavefunction splits in two parts with negligible cross-terms.

 

[bhobba]

MWI seems like a theoretical dead-end to me, it is impossible to test, it adds nothing to the theory and seems like a non-answer to me. How does one mathematically define the phrase "The Universe just splits into two"?

Well it actually doesn't split in two, or even split really - it simply keeps evolving.

After decoherence a superposition is converted to a 'improper' mixture that has the form sum pi |bi>|ri> where |bi> is the outcome of the observation and |ri> is the rest of the world. Now in a collapse interpretation pi gives the probability of the outcome |bi> and after the observation the outcome and the rest of the world has state |bi>|ri> that keeps on evolving. This is the collapse of the wavefunction issue and depending on your interpretation may or may not be an issue. If it was a real mixture it wouldn't be an issue since a mixture is the outcomes randomly presented but having the outcome there prior to observation - its like selecting a random marble out of a hat that has been placed there. However its an 'improper' mixture meaning its mathematically the same but is not physically the same - its observationally exactly the same but has not been physically prepared the same way. You can assume it is and hold it has that outcome prior to observation but in reality how it selected that outcome is a mystery. I hold to a modified MSI that does just that - simply assumes it observes an outcome already there but what it is has probability pi. But what I have really done is hidden the collapse issue - however its still there - just swept under the rug.

MWI is a twist on this - the mixture keeps on evolving as is and each of the |bi>|ri> continues to evolve. However to an observer in the |ri> its as if an observation was made and the outcome |bi> occurred. No collapse occurred and everything continues to evolve deterministically. Of course we are left with the issue of how probabilities enters into it and various researches such as David Deutch have their own take on it that you can read about.

Very very elegant - no collapse, totally deterministic - all issues resolved. But can you stomach all these exponentially increasing multiple you's - that's the question.

Thanks
Bill

 

[bhobba]

The same can be said about all other interpretations of QM.
You do not have to. But if you like, with decoherence: In your favorite basis (actually, in many bases), the wavefunction splits in two parts with negligible cross-terms.

That's true - you don't have to evoke decoherence but it does help in explaining why observers in each of the worlds experiences it like a collapse. Its the way I prefer to look at it.

Thanks
Bill

 

[Fredrik]

My opinion is the breaking of a system into subsystems is entirely arbitrary - basically you choose it for convenience. You can choose not to break it into subsystems, or into as many as you like with any boundaries you like. If however you choose a not particularly good one then the interaction Hamiltonian, Lagrangian, or whatever formalism you use is more complicated, however if you break it into subsystems such that the interaction term (or terms) is non existent or at least simple then you have chosen a good one. You would, or should anyway, obtain exactly the same results either way - its just a suitable choice will make life easy.

I would expect the results (the preferred bases) corresponding to two different decompositions to not be exactly the same. In fact, I would expect them to have almost nothing in common. For example, I don't see any reason to think that "the silver atom in my Stern-Gerlach experiment + everything else" and "an iron atom in the sun + everything else" should give us similar results. That article seems to confirm that, by showing that there's always a decomposition such that the two subsystems do not interact with each other. If we choose that decomposition, no preferred basis will emerge.

What I expect all this to mean (this is what I've been thinking for a few years, so it doesn't have anything to do with that article) is e.g. that a claim that a classical world is being split in two is never objectively true or false. It can be true relative to one decomposition, and false relative to another. Worlds can be splitting in one way relative to one decomposition, in an entirely different way relative to another.

I also expect this "non-objectiveness" to go very far beyond this. The worlds relative to one decomposition could perhaps be very different from the worlds relative to another decomposition. Note that a decomposition doesn't have to be "this chunk of matter + everything else". Every countable orthonormal basis and every bijection from $\mathbb Z$ (the set of integers) into $\mathbb Z\times \mathbb Z$ defines a decomposition, as described (briefly and not very pedagogically) in the article. Basically, you take a basis $\{e_i\}$ and relabel the basis vectors $\{e_{ij}\}$. Then for each j, $\{e_{ij}\}_{i=1}^\infty$ spans a subspace, and if I understand this idea correctly, the original Hilbert space can be viewed as a tensor product of (something isomorphic to) that subspace and (something isomorphic to) its orthogonal complement. These spaces can then be considered the Hilbert spaces of "subsystems".

There could be decompositions such that a description of what's happening is completely different from what we'd get from a typical decomposition. It seems entirely plausible to me that such a description could describe worlds where you and me don't even exist. The "exponentially multiplying" copies of you will only exist in the descriptions corresponding to a specific sequence of decompositions. Relative to other decompositions, you may not be splitting in the same way, and you may not even exist. Since these ideas haven't been explored (as far as I know) it's hard to tell how deep this goes. Is it possible that the same chunk of matter that we think of as "bhobba", is a Coca-Cola machine relative to another decomposition? I have no idea, and I think its funny that I can't even tell if such things can be ruled out.

Note that I'm not saying that this is what's actually happening. I still prefer to think of QM as just an assignment of probabilities to possible results of experiments. I'm just saying that it seems to me that the consequences of assuming that QM describes what's actually happening are very different from what people seem to think.

I don't consider this a reason to dismiss the MWI. I don't consider it a reason to favor the MWI. I just think it makes the MWI much more fascinating than it seemed at first.

 

[Maui]

What are you talking about? I haven't shifted anything. The only thing we've been talking about is whether an additional assumption that's added to the theory without changing its predictions can be falsified by experiments. It can't. This is an obvious fact that has has nothing to do with technological advances or how well we understand the theory.

The assumption is falsified if it cannot explain experimental results which can otherwise quite easily be explained without making the assumption.

And I don't always know what you mean. What did you mean by an "interpretation that's not part of the predictive theory"? What does it have to do with the possibility of disproving the MWI by experiment?

The MWI is not part of quantum theory. Do you understand thus far? It's your fantasy that it's somehow the same as quantum theory but it is not. It's fantasy proposed by a propminent figure, that's all.

How am I supposed to answer this? Your answer has nothing to do with what I said, or with your previous statement that I was responding to. You said that we only get quantum behavior when which-way information cannot be obtained. So when I quoted that statement and said that "this" is just what the theory (i.e. QM) predicts, I was obviously talking about the quantum behavior and the which-way information. So when you're quoting that and saying "incorrect", your statement can only be interpreted as saying that QM's prediction is something other than this,

Definitely no, quantum theory does not make predictions in the sense of classical mechanics but assigns probabilities. You most definitely cannot predict outcomes 1:1 but only approximately for large ensembles. So the fact that you misrepresent qm as predicting the results of the quantum eraser is enough for me to stop responding to your posts(especially since human choice enters the experimental setup which cannot be modeled by the SE). QM does not describe how you get no interference pattern when which-path information is available, this can only be 'explained' via superdeterminism and the SE.

Sorry that was my last post arguing religious beliefs.

 

[Maui]

Sorry, but you need a collapse interpretation to say that.

No, you can choose whether to look at detectors that can in principle provide which-path information or you can choose to look at the detectors that don't provide the information. By doing this many times, you can choose what behavior you want to observe - you can 'cause' collapse just by choosing the right set of detectors.

 

[Nugatory]

The assumption is falsified if it cannot explain experimental results which can otherwise quite easily be explained without making the assumption.

That's not sufficient to falsify an assumption; it shows that the assumption is unnecessary but does not show that the asumption is untrue. Falsification requires a stronger criterion: the assumption must lead to a prediction that can be shown to be false.

 

[Fredrik]

The assumption is falsified if it cannot explain experimental results which can otherwise quite easily be explained without making the assumption.

See nugatory's reply.

The MWI is not part of quantum theory. Do you understand thus far?

Of course, that's why I've been saying that many times.

It's your fantasy that it's somehow the same as quantum theory but it is not.

This is a lie. I don't have a problem with you making incorrect statements about QM or its interpretations, or with you using words like "prediction" and "falsified" incorrectly, but now you're lying about me. That's very far from acceptable behavior in this forum.

Definitely no, quantum theory does not make predictions in the sense of classical mechanics but assigns probabilities.

Why are you telling me this? This is something I'm saying all the time. I'm sure I've said it more times than anyone else in this forum.

You most definitely cannot predict outcomes 1:1 but only approximately for large ensembles. So the fact that you misrepresent qm as predicting the results of the quantum eraser is enough for me to stop responding to your posts(especially since human choice enters the experimental setup which cannot be modeled by the SE). QM does not describe how you get no interference pattern when which-path information is available, this can only be 'explained' via superdeterminism and the SE.

Wait, you're actually suggesting that what I said in my previous post implies that QM doesn't assign non-trivial probabilities to possible results of measurements? That's so far from the truth that I don't even know where to begin.

 

[kith]

It seems that since you cannot explain the results of this experiment in terms of the MWI(you would have if you could) [...]

Sure I can. Whenever collapse interpretations say collapse happens, the MWI says that decoherence occurs. For every which-path information you get, there is a complementary world where you register that the other path has been taken. These points of view can't be distinguished by experiments. If you don't understand this, you don't understand the mathematics of the MWI.

[...] you hope you can just skip the technical details of the experiement and appeal to authorities.

Yes, I appeal to authorities. Kim's experiment has been done almost 15 years ago. Since then, numerous articles on the MWI have appeared in peer-reviewed physics journals. You claim that all these people are wrong. This is an extraordinary claim which has to be backed up by a reference. And no, Kim's original paper is not supporting your claim.

 

[bhobba]

That's not sufficient to falsify an assumption; it shows that the assumption is unnecessary but does not show that the asumption is untrue. Falsification requires a stronger criterion: the assumption must lead to a prediction that can be shown to be false.

True. But in this case LESS assumptions are made - no collapse assumption is assumed - instead each of the outcomes of the mixture after decoherence continues to evolve with each observer in each element of the mixture experiencing that outcome - it literally takes the formalism at face value.

Thanks
Bill

 

[Quantumental]

True. But in this case LESS assumptions are made - no collapse assumption is assumed - instead each of the outcomes of the mixture after decoherence continues to evolve with each observer in each element of the mixture experiencing that outcome - it literally takes the formalism at face value.

Thanks
Bill

So how are you going to get worlds out of the formalism and how are you going to get probabilities out of the worlds?
I posted 2 papers, one that highlights the factorization problem, no one has rebutted it.
The second paper shows the inherent circularity of relying on decoherence for a basis, when decoherence itself is probabilistic. So you have to magically introduce (just like collapse) the born rule into decoherence and then hope that FAPP will give you worlds.

I'm so tired of people claiming that MWI hass less assumptions and more parsimony when it isn't even a working interpretation yet!

I hate collapse interpretations, I literally dispise them to the end of time, but at least they can have their interpretation work (its obviously wrong, but still). MWI cannot even do that.


So please do me the favor of reading both of those papers and then come back with a rebutle before boasting about how MWI is elegant again.

 

[bhobba]

So how are you going to get worlds out of the formalism and how are you going to get probabilities out of the worlds?
I posted 2 papers, one that highlights the factorization problem, no one has rebutted it.

The first paper was a philosophical analysis and philosophy is notorious for not being able to get a consensus on anything and to be blunt their knowledge of the physics often leaves a lot to be desired. Sorry mate but despite your cajoling you are not going to get too many physics types to take it seriously. Care to give us a run down on its argument? If it sounds promising you may get more people interested in reading it.

MFB rebutted the second.

So you have to magically introduce (just like collapse) the born rule into decoherence and then hope that FAPP will give you worlds.

David Deutch and others beg to differ. You may not agree with it but that is a matter of opinion - saying its magic is not a rebuttal.

I'm so tired of people claiming that MWI hass less assumptions and more parsimony when it isn't even a working interpretation yet!

You know what I am tired of? Statements like the above not backed up by anything. Care to actually tell us why?

I personally think MW is BS - but its a fully valid theory that takes the formalism literally.

Lets be clear exactly what it does. Decoherence, completely deterministically, transforms a superposition into a mixed state. Each element of the mixed state then evolves deterministically - no collapse - no assumption - simply the normal deterministic evolution of a quantum state. How you get the worlds is dead simple - by definition each element of the mixed state is a world. Exactly how is that not a working interpretation?

Thanks
Bill

 

[Fredrik]

So how are you going to get worlds out of the formalism

By an assumption that we can consider a part of the definition of the MWI. It seems to me that this assumption should say that every 1-dimensional subspace of the Hilbert space is a world. It annoys me that the stuff I've read (admittedly not that much) about the MWI doesn't explicitly define what they mean by a "world".

and how are you going to get probabilities out of the worlds?

I think that once a decomposition into subsystems has been chosen, Gleason's theorem gives us the probabilities, but not their significance. It's not clear to me to what extent that problem has, or can, be solved. It certainly hasn't been completely solved.

I posted 2 papers, one that highlights the factorization problem, no one has rebutted it.

I have however written several comments about it. My "rebuttal" is that I don't consider "the factorization problem" a problem at all. (See e.g. my posts #90 and #116). It's only a problem for those who think that at a given moment, there's exactly one set of worlds that make up the universe. The factorization "problem" is however a very interesting feature, that should be highlighted by a proper definition of the interpretation. So I find it pretty weird that this paper is the first mention of this feature that I've seen outside of my own forum posts.

The second paper shows the inherent circularity of relying on decoherence for a basis, when decoherence itself is probabilistic. So you have to magically introduce (just like collapse) the born rule into decoherence and then hope that FAPP will give you worlds.

As I said before, the low quality of the published papers I've read makes me reluctant to read an unpublished preprint.

I'm so tired of people claiming that MWI hass less assumptions and more parsimony when it isn't even a working interpretation yet!

I agree that it still needs work. I don't think that anyone has ever written down a definition of "the MWI" that I would find satisfactory. So maybe it shouldn't be called an interpretation yet. It's an idea that probably can be developed into an interpretation.

I hate collapse interpretations, I literally dispise them to the end of time, but at least they can have their interpretation work (its obviously wrong, but still).

I find those pretty ridiculous too.

 

[martinbn]

Fredrik, as you very well know, 1-dim subspaces are the states of the system. If you call them worlds, it seems to me, that you are just renaming them, that cannot make any difference.

 

[Fredrik]

Fredrik, as you very well know, 1-dim subspaces are the states of the system. If you call them worlds, it seems to me, that you are just renaming them, that cannot make any difference.

The difference is the idea that they all represent something that actually exists, rather than something that could exist. (States should usually be thought of as equivalence classes of preparation procedures, but I guess that wouldn't make sense when we apply QM to the whole universe).

If you (or anyone) know a better definition of "world", I would be interested in hearing it. All I know is that terms like that need to be defined before we can claim to have defined a many-worlds interpretation.

 

[bhobba]

The difference is the idea that they all represent something that actually exists, rather than something that could exist. (States should usually be thought of as equivalence classes of preparation procedures, but I guess that wouldn't make sense when we apply QM to the whole universe).

If you (or anyone) know a better definition of "world", I would be interested in hearing it. All I know is that terms like that need to be defined before we can claim to have defined a many-worlds interpretation.

That's the thing with MW - they think of a state like an electric field - something real - not like for example Ballentine does in his excellent book. Of course you then have the arguments he brings to bear about such a view but I don't think they are deal breakers within the context of MW.

We also have the issues you raised about decoherence in selecting a preferred basis (ie a different decomposition may single out a different basis or even none at all) but the consensus view is that decoherence solves that issue. It certainly does if we make the assumption of decomposing the system in the obvious way - ie observational apparatus and system being observed. The decoherence guys also admit that this is still an area of active research and I have zero doubt it will be clarified in time.

Now taking that into account I think what I wrote, namely the |bi>|ri> where the |bi> are the basis vectors after decoherence and the |ri> the state of the rest of the world can be taken as the worlds the MW guys talk about. They come about deterministically via the decoherence process (ie tracing over the environment) in transforming a superposition into the form of a mixed state sum pi |bi>|ri> and evolve deterministically after decoherence.

From the discussion here I think we can say that MW is a valid interpretation and a very elegant one at that. There are some issues that are controversial such as how probabilities come into it and the preferred basis problem but exactly what's going on there will become clearer with further research.

Because of that I see nothing that justifies some of the hysteria in this thread that it has been disproved or that it is not even a 'working' interpretation.

Thanks
Bill

 

[Fredrik]

We also have the issues you raised about decoherence in selecting a preferred basis (ie a different decomposition may single out a different basis or even none at all) but the consensus view is that decoherence solves that issue.

I would say that it doesn't even address the issue.

It certainly does if we make the assumption of decomposing the system in the obvious way - ie observational apparatus and system being observed.

I think this is like saying that the ether must be comoving with me, because the obvious way for me to choose a coordinate system is to choose the inertial coordinate system that's comoving with me.

What if two different experiments are performed at the same time? Then there are two "obvious" decompositions.

Now taking that into account I think what I wrote, namely the |bi>|ri> where the |bi> are the basis vectors after decoherence and the |ri> the state of the rest of the world can be taken as the worlds the MW guys talk about. They come about deterministically via the decoherence process (ie tracing over the environment) in transforming a superposition into the form of a mixed state sum pi |bi>|ri> and evolve deterministically after decoherence.

Since any complex number times |bi>|ri> represents the same state as |bi>|ri>, I think it's slightly more accurate to say that each 1-dimensional subspace spanned by a member of this basis is a world. This basis has the property that the state operator of the universe is approximately diagonal in it. If it had been exactly diagonal, I think it might have made sense to say that the worlds corresponding to the basis vectors are the only ones. But it's only approximately diagonal, so I don't think it makes sense to say that $|b_i\rangle|r_i\rangle$ represents something that actually exists, while $|b_i'\rangle|r_i'\rangle$ such that $\langle b_i'|b_i\rangle \langle r_i'|r_i\rangle$ is extremely close to 1, doesn't.

I think people who write about the MWI are aware of this. They sometimes refer to the worlds identified by a preferred basis as "the classical worlds" rather than "the worlds". I think it would be appropriate to say that all 1-dimensional subspaces are worlds, and that the ones identified by the basis are the classical worlds, or perhaps more accurately, the most classical worlds. They are the ones in which the classical description of what's going on (we measured the spin and found it to be "up") is as close as possible to being accurate. I think it would be even more accurate to say that the 1-dimensional subspaces corresponding to the members of the preferred basis associated with the chosen decomposition are the classical worlds associated with that decomposition.

I don't think this is a problem that can be "solved". It's just a matter of accepting that at any instant, there are many inequivalent ways to view the universe as consisting of classical worlds. To me, this makes the MWI less crazy, not more.

 

[martinbn]

The difference is the idea that they all represent something that actually exists, rather than something that could exist. (States should usually be thought of as equivalence classes of preparation procedures, but I guess that wouldn't make sense when we apply QM to the whole universe).

If you (or anyone) know a better definition of "world", I would be interested in hearing it. All I know is that terms like that need to be defined before we can claim to have defined a many-worlds interpretation.

I still don't see the difference. It seems simply as more words. If you try to explain what exists means here, I think you'll end up with nothing new just different names.

 

[mfb]

If you (or anyone) know a better definition of "world", I would be interested in hearing it. All I know is that terms like that need to be defined before we can claim to have defined a many-worlds interpretation.

"Real-wavefunction-interpretation"?

From the discussion here I think we can say that MW is a valid interpretation and a very elegant one at that.

:)

What if two different experiments are performed at the same time? Then there are two "obvious" decompositions.

And there are "obvious" decompositions describing both experiments at the same time.

I think people who write about the MWI are aware of this. They sometimes refer to the worlds identified by a preferred basis as "the classical worlds" rather than "the worlds". I think it would be appropriate to say that all 1-dimensional subspaces are worlds, and that the ones identified by the basis are the classical worlds, or perhaps more accurately, the most classical worlds. They are the ones in which the classical description of what's going on (we measured the spin and found it to be "up") is as close as possible to being accurate. I think it would be even more accurate to say that the 1-dimensional subspaces corresponding to the members of the preferred basis associated with the chosen decomposition are the classical worlds associated with that decomposition.

I don't think this is a problem that can be "solved". It's just a matter of accepting that at any instant, there are many inequivalent ways to view the universe as consisting of classical worlds. To me, this makes the MWI less crazy, not more.

That sounds good.

 

[Fredrik]

I still don't see the difference. It seems simply as more words. If you try to explain what exists means here, I think you'll end up with nothing new just different names.

I think that perhaps you expect too much from an interpretation. The starting point of this interpretation is that QM is more than just an assignment of probabilities, it also describes what's actually happening. Since you have been trained to think like a scientist, you're probably immediately thinking that such statements are nonsense unless we define them. You may also have thought about how to define some of them and only come up with useless definitions like "a theory describes what's happening if it makes good predictions about the results of experiments". This definition is useless because if this is how the statement is defined, it doesn't tell us anything.

We are so used to requiring everything to be defined that we sometimes forget that some concepts are simply more fundamental than the things we use to define them. For example, we like to define "integers" in ZFC set theory. But a set theory is defined by a set of axioms that are statements in a formal language, and the formal language is defined by specifying an alphabet of symbols and some other stuff. How can you even write down an alphabet with say five symbols, if the concepts 1,2,3,4,5 are undefined? It seems that we are using those concepts before we have defined them.

This is, unfortunately, how we need to think of the statements that define an interpretation. You don't define what it means for a theory to describe what's happening. You already know what it means. Similarly, if the interpretation says that every 1-dimensional subspace represents something that actually exists, you need to just accept that you already know what that statement means. If you allow such statements to have the meanings your intuition is already assigning to them, then they do tell us something that the theory doesn't.

Is this science? Hell no. But it's not supposed to be. Is it nonsense? Probably not.

 

[martinbn]

Fredrik, I think now I see better what you mean, but it still seems that your proposal doesn't give anything knew as an interpretation, except for new names. Of course, that's my problem, I am not very familiar with any of them.

 

[Quotidian]

There's an interesting Scientific American review of Everett's Many Worlds here. Not that it is very relevant, but he was not a likable, warm human being. He died an emotionally distant, chain-smoking alchoholic aged 51; the article remarks that his son found his body and felt for his pulse, then realized it was possibly the only time he had ever actually touched him.

A lot of his career was spent callibrating the maximum effectiveness for where to drop nuclear weapons. His ethics were also questioned by some of his employers.

Most of his peers thought the actual theory was nonsense and refused to even consider it. However the article notes that attitudes have changed in the last 20 years. I do wonder if this is simply because we have generally developed a much higher tolerance for nonsense. That would not at all surprise me.

I thought a very telling quotation from Everett was this one:

The Copenhagen Interpretation is hopelessly incomplete because of its a priori reliance on classical physics ... as well as a philosophic monstrosity with a “reality” concept for the macroscopic world and denial of the same for the microcosm.

(page 3)

Why 'a denial of the same'? I think it is because the then-prevailing Copenhagen interpretation posited the centrality of the act of observation in the 'collapse of the wave-function'. This, of course, undercuts the notion of objectivity, and is the source of a lot of the debate around the whole issue. (Einstein was said to have remarked in relation to this notion 'does the moon continue to exist when nobody is looking at it?')

So the solution was to posit an uncountable number of worlds. This was preferable to allowing the heresy that consciousness might have a fundamental role in the scheme of things even if it seems such an affront to the natural principles of science.

Strange old world.

This one, anyway.

 

[bhobba]

This was preferable to allowing the heresy that consciousness might have a fundamental role in the scheme of things even if it seems such an affront to the natural principles of science.

I think consciousness causes collapse is basically rubbish - but heresy - no.

It attracted the attention of some pretty high powered and reputable people such as Von-Neumann and Wigner.

The interesting thing however is that MWI was the first interpretation to incorporate decoherence - it was in its infancy then but it certainly signalled the start of its rise. In fact from some early work on decoherence by Zurek Wigner abandoned the consciousness causes collapse idea as no longer being necessary - and he was correct - it isn't - some quite simple interpretive assumptions avoid it eg decoherent histories. Indeed in a poll only about 7% ascribe to it nowadays so one could argue it really is in decline - and MWI either with or without decohenrece polls 20%. My ensemble interpretation polls about the same as consciousness cause collapse at a lousy 7% - looks like shut up and calculate aren't that popular either - which may be good or bad depending how you look at it - I personally think its a good thing because it shows people actually are thinking about it.

Thanks
Bill

 

[Quotidian]

The 'consciousness causes the wave-function collapse' idea is rubbish because...

As regards 'decoherence', I have never read an intelligible account of this idea in English, and actually doubt that there is such a thing. Although I am open to persuasion.

 

[bhobba]

The 'consciousness causes the wave-function collapse' idea is rubbish because...

Well imagine a quantum experiment recording the positions of an interference experiment but the results are stored in a computer. The apparatus is disassembled and centuries later the results viewed. Are you seriously going to tell me that's when the wave-function collapsed and the positions manifested? Indeed exactly what quality distinguishes a computer memory recording results and human memory doing the same thing? You could probably come up with some contrived scheme placing the human brain on a pedestal with some privileged status but exactly for what gain? Do you want the world to be this weird consciousnesses created reality? Is solipsism appealing to you? If it is I can't prove you wrong but there is a reason virtually everyone rejects it - but hey if that's what floats your boat be my guest - I - and I suspect most people won't be there with you.

As regards 'decoherence', I have never read an intelligible account of this idea in English, and actually doubt that there is such a thing. Although I am open to persuasion.

Well obviously guys like Wigner did - which should be a clue there is something to it. Its standard textbook stuff - see for example Schlosshauer - Decoherence: and the Quantum-To-Classical Transition:
https://www.amazon.com/dp/3642071422/?tag=pfamazon01-20

Thanks
Bill

 

[Quotidian]

Indeed exactly what quality distinguishes a computer memory recording results and human memory doing the same thing?

A computer doesn't read or interpret anything whatever. Quite apart from the fact that the computer itself is only an instrument and will only do what it is programmed to do by a human intelligence, whatever data it contains does not amount to information unless it is interpreted by a human. in his lectures on Mind and Matter, Schrodinger argued that there is a difference between measuring instruments and human observation: a thermometer’s registration cannot be considered an act of observation, as it contains no meaning in itself.

We, of course, can argue about that. A computer, however, cannot.

Is solipsism appealing to you?

Solipsism only enters into the equation if you think that yours is the sole intelligence in the universe. I would not be so vain. You will note that we all share the same basic perceptual apparatus, metabolism, brain functioning, and so on. On a slightly higher level of the 'stack' we also share a common culture, language, background assumptions about what is or is not real, and so on. And these things do indeed constitute the fabric of reality.

As regards the book linked to on 'decoherence'. The topic might be 'standard textbook stuff', however my observation was that I have never encounted a comprehensible account of it in English. The reader reviews and editorial information on those titles do nothing to address that.

 

[bhobba]

A computer doesn't read or interpret anything whatever. Quite apart from the fact that the computer itself is only an instrument and will only do what it is programmed to do by a human intelligence, whatever data it contains does not amount to information

So you think until it's printed out of memory or displayed on a screen and interpreted by a human its not information? Like I said you can probably come up with some weird view giving some kind of privileged status to human consciousness - and it looks like you did. All I can say is your view is not held by any computer scientist I am aware of. I would suggest if you promulgated such a view in a computer science class your audience will disperse rather quickly amid quite a bit of chuckling. Still I can't prove you wrong - its simply a very weird view that is at odds with how the majority view such things.

As regards the book linked to on 'decoherence'. The topic might be 'standard textbook stuff', however my observation was that I have never encounted a comprehensible account of it

Well since that is not the reaction of the vast majority of people exposed to it have you considered that the issue may reside in you rather than the material?

You might like to detail exactly what you don't understand. For example what's the issue with tracing over the environment transforming a pure state into an improper mixture?

Thanks
Bill

 

[kith]

[...] in his lectures on Mind and Matter, Schrodinger argued that there is a difference between measuring instruments and human observation: a thermometer’s registration cannot be considered an act of observation, as it contains no meaning in itself.

The question is, why should the difference between "observation by a human" and "registration by an apparatus" in QM be more important than in classical physics? In consciousness-based interpretations and (arguably) the Copenhagen interpretation, this is the case. In the MWI, it is not.

As regards the book linked to on 'decoherence'. The topic might be 'standard textbook stuff', however my observation was that I have never encounted a comprehensible account of it in English.

It's quite difficult to explain decoherence without using a bit of math. I haven't come up with an easy but accurate explanation of it yet. Maybe other people here have?

However, in order to understand how the measurement problem is solved within the MWI and how the worlds are introduced, I think decoherence is a detour. See my post #3 for Everett's initial approach.

 

[Quotidian]

you can probably come up with some weird view giving some kind of privileged status to human consciousness - and it looks like you did.

Right. So believing that 'observation requires an observer' is weird, but entertaining the idea of 'infinite proliferating worlds' is not.

In consciousness-based interpretations and (arguably) the Copenhagen interpretation, this is the case. In the MWI, it is not.

Again - at what cost? Your post #3 says 'There's nothing what singles out one outcome, so we can interpret each of them as belonging to a different world'. But are there actually 'different worlds'? Does it matter? And if it doesn't matter, then does this construction have anything to do with reality? Or is it just like a 'thought-experiment', a kind of conceptual model to do away with something that is 'conceptionally unsatisfying for many people'?

 

[HomogenousCow]

At the end of the day, without any real distinguishing evidence to go by, I think the subject of interpretations is a non-subject, for example if I choose to believe that there are parallel universes which cannot affect us in any way, you cannot prove me wrong but I cannot prove myself right either.
Or you know..if you believe there are omni-powerful beings present in the universe..Oh wait I have a warning point already.

 

[bhobba]

Right. So believing that 'observation requires an observer' is weird, but entertaining the idea of 'infinite proliferating worlds' is not.

MW is equally as weird IMHO as consciousness causes collapse. But weirdness is in the eye of the beholder. My Ensemble interpretation also has a weird aspect - namely exactly how is the improper mixture of decoherence converted into an actual mixture. Observationally they are identical - but their physical preparation is entirely different. Every single interpretation is weird in some way - you simply pick the one that is the least weird to you. Most people, because of the difficulties I alluded to with computers, and other reasons, would say consciousness causes collapse has too much baggage associated with it to be taken seriously - but that doesn't mean its not valid.

BTW in QM observation means any device capable of registering in the macro world - it's a misunderstanding thinking a priori observation requires an observer despite the semantic closeness of the terms. If that is what you believe you are not the only one to be confused by it.

Thanks
Bill

 

[bhobba]

Again - at what cost?

You asked me to justify:

I think consciousness causes collapse is basically rubbish - but heresy - no.

Notice the words - I think.

In any interpretation you are making a judgement - every single interpretation has baggage - it's purely a matter of which you think has the least baggage. But that is your choice - its not subject to experimental verification. Opinions are like bums - everyone has one - it doesn't make it correct.

Thanks
Bill

 

[bhobba]

I think the subject of interpretations is a non-subject

Not really. In order to apply QM you need a way to map the mathematical elements of its formalism to stuff out there. That's the job of an interpretation. Even if you are in the shut-up and calculate camp you still have an interpretation - its called the Minimal Statistical Interpretation.

Think of good old Euclidean Geometry. How to map it is usually very obvious physically and no issues arise. The interesting thing about QM is that mapping is not obvious so you have all these different interpretations and threads like this. Its in the nature of the material.

Interestingly in the past, say about the time Feynman was in his heyday, the shut-up and calculate view was predominate but these days its all over the place. As mentioned previously this may be a good or bad thing depending on how you look at it.

Thanks
Bill

 

[Quotidian]

Very true. Wise words indeed. It is just that us philosophically curious lay-persons are often informed by the likes of Lawrence Krauss and Stephen Hawking that 'physics has superseded philosophy'. Yet the question of 'interpretations' is very much a philosophical question - and one which, no disrespect intended, few physicists seem to have much of a handle on.

I think I will go back to Kant, Schopenhauer and Buddhist studies. Thanks for your responses.

 

[bhobba]

Very true. Wise words indeed. It is just that us philosophically curious lay-persons are often informed by the likes of Lawrence Krauss and Stephen Hawking that 'physics has superseded philosophy'. Yet the question of 'interpretations' is very much a philosophical question - and one which, no disrespect intended, few physicists seem to have much of a handle on.

I think I will go back to Kant, Schopenhauer and Buddhist studies. Thanks for your responses.

Most physicists are like myself and are a bit anti-philosophy along the lines of Feynman.

But note - that itself is a philosophy.

I don't think its fair to say, like Lawrence Krauss and Stephen Hawking says (and I have read what they say and they do say stuff like that), that science now supersedes philosophy. What is fairer to say is science makes extensive use of a tiebreaker in an argument - philosophical or otherwise - actual observation - that's the real issue. Truth is only held provisional while observation supports it. That's the key difference.

Thanks
Bill

 

[Fredrik]

In my opinion, philosophy has a very limited relevance to science and mathematics. It can handle questions like "What is science?", "What is mathematics?" and "What is the best way to define the term 'theory of physics'?". To note that terms like "state" can be defined in a theory-independent way is to do philosophy. To think about what whether QM can or can't be interpreted as a description of what's actually happening, is also to do philosophy.

Unfortunately I don't think philosophers are doing any of those things well. So I can't help wondering if philosophers have any relevance to science.

The only reason I'm not saying that the philosophy of science and mathematics should definitely be left to physicists and mathematicians is that they're not doing these things well either. Philosophers probably just don't understand the subject well enough. But physicists and mathematicians have an attitude problem. They think that this sort of stuff is beneath them.

 

[dx]

To note that terms like "state" can be defined in a theory-independent way is to do philosophy. To think about what whether QM can or can't be interpreted as a description of what's actually happening, is also to do philosophy.

I would like to mention this again, since it is a key point in the copenhagen interpretation (a term I reserve for Bohr's views alone)

"What is actually happening to the system" is clearly a meaningful concept as long as the behavior of the system can be observed without the act of observing the system influencing the system.

But in quantum mechanics, the existence of the quantum of action h implies that there is a lower limit to the interaction between the measuring bodies and the objects under investigation. One can try to control this interaction, by observing the measuring bodies themselves, but in that case those measuring bodies themselves become part of the system being observed, and the additional measuring bodies introduced will again have an uncontrollable interaction with the system.

So the point is, "what is happening to the system" cannot be separated from the question of the observation of "what is happening to the system"

 

[Fredrik]

I would like to mention this again, since it is a key point in the copenhagen interpretation (a term I reserve for Bohr's views alone)

"What is actually happening to the system" is clearly a meaningful concept as long as the behavior of the system can be observed without the act of observing the system influencing the system.

But in quantum mechanics, the existence of the quantum of action h implies that there is a lower limit to the interaction between the measuring bodies and the objects under investigation. One can try to control this interaction, by observing the measuring bodies themselves, but in that case those measuring bodies themselves become part of the system being observed, and the additional measuring bodies introduced will again have an uncontrollable interaction with the system.

So the point is, "what is happening to the system" cannot be separated from the question of the observation of "what is happening to the system"

QM says that any isolated system (or at least any isolated system that can be made to interact with a measuring device) has a state that changes with time as described by the Schrödinger equation. I would say that the main purpose of an interpretation of QM is to provide a possible answer to the question of what is actually happening to an isolated system, while it's still isolated.

Isn't what you're suggesting, and what those Bohr quotes in post #15 are suggesting, only that a measurement breaks the isolation? What is happening to to the system during the measurement can't be separated from the observation of what's happening to the system. I don't think this is relevant to what I had in mind, because when I was talking about the question of whether QM describes "what's actually happening" to a system, what I had in mind was an isolated system.