Problems with Many Worlds Interpretation

[Fra]

I think that is an interesting insight. You are essentially encoding solipsism into the theory itself-- saying that how we interact with a system, and what we are capable of understanding about a system, are actually parts of the system itself. That means that the subject/object separation, so valuable for so many eons of physics, has become fundamentally a block to obtaining a more powerful theory.

Yes exactly. But with emphasis on that it's now just how we humans interact with the system, it's how any parts of the universe interact by other parts. This insight, provides us with deeper way to understand the origin of any interaction, and thus to address unification since as you scale the complexity of the interacting parts down, it becomes impossible for them to execute complicate interactions. It just looks complicate from the point of view of an external observer due to that fact that the simple theory scales into a - not so simple theory. Without this insight, we end up with spaces of parameters which we don't konw why they are tuned as they are.

This solipsist inside view, at least provides a handle on these issues, something which can't be said on the realist views. You end up with a landscape of realities where you don't know WHICH reality you are in.

/Fredrik

 

[Dmitry67]

MUH isn't a theory at all, it is a means of achieving an aesthetic sense of cognitive resonance. Things like that were separated from science thousands of years ago, so I'm not sure now is the time to bring them back in.

You should not classify MUH as pure phylosophy, because it has a falsifiable prediction: TOE has no "word baggage" - everything will be described by math, no additional comments will be required.

 

[Ken G]

You should not classify MUH as pure phylosophy, because it has a falsifiable prediction: TOE has no "word baggage" - everything will be described by math, no additional comments will be required.

Pure philosophy makes falsifiable predictions too-- "philosophy will someday let us understand everything there is to know." But of course, that prediction doesn't really mean anything, because it is not currently testable, nor does it give us guidance on where to start looking to test it. What you're saying about MUH is not much better-- it might provide some guidance to those looking for the next great theory of physics (calling that a TOE is kind of silly, if you ask me), but it might just as easily provide false guidance as good guidance. Perhaps that means it is just as well that some people accept a MUH perspective, and some don't-- that latter group is more or less the purpose of solipsism when it is used constructively.

 

[Eqblaauw]

I think it's wrong to say that mwi is an inevitable outcome of qm:
At first the theory had the universe splitting continuisly. That didn't work,so they had to come up with a new interpretation. Nowadays everybody working on the theory (this is what deutsch wrote to me in a email) assumes there are many universes (or infinite, who cares about a universe more or less right), that are exactly the same until they differentiate. And hey it works again (quite a difference i would say, the first version is impossible to live with if you think it through, the second is still mental hospital stuff but less). I like to propose that al the universes that are just assumed as a solution to get the math right, are eaten by a little invisible man, and when he burps it causes interference
and tegmark is one immoral freak for introducing quantum suicide

 

[Fredrik]

AFAIK, most physicists do see MWI as simply 'no non-unitary collapse'. Otherwise, I think MWI would be taken a lot less seriously.

It's definitely more than this, because what you're saying sounds more like a statistical interpretation than like a MWI. At the very least, a MWI supporter has to believe that QM describes reality. (See my posts earlier in the thread if you're thinking about asking what that means).

I would consider the assumption that "QM describes reality" a definition of a MWI. Strictly speaking, this assumption doesn't imply many worlds. It doesn't imply anything, since it's not a mathematical statement. However, it strongly suggests that many worlds exist. This is how: Every measurement includes an interaction that (approximately) puts the combined system specimen+device into a mixed state, represented by a state operator with no off-diagonal terms (in a specific basis determined by the interaction). Let's write it as \rho=\sum_i p_i|\psi_i\rangle\langle\psi_i|. We know from experience that one of those terms represents what has actually happened, but there is nothing in QM that singles out one of the terms as different from the others. So if you are going to claim that QM describes reality, you better be ready to accept all of the terms as real.

We can of course change the original assumption to "exactly one of the terms describes reality", but then we're talking about a "Copenhagenish" interpretation rather than a MWI.

 

[Fredrik]

Do you have a citation for that? Skimming Wikipedia suggests otherwise.

As I understood it, Everett's key insight was to use relative states -- thus the "relative state formulation".

No I don't have a citation. I think you're right. Everett didn't insist that the Born rule should be recovered from branch counting. Or maybe he did, I don't know. What I'm trying to say is that I have no memory of reading anything that suggests that he is personally to blame for that idea. But Everett did remove the Born rule from QM. (I don't remember where I read that. If you doubt it, I guess I can start looking for original articles). This removed the theory's ability to make predictions, and therefore made it something less than a theory. He must have thought that there would be some way to recover it.

 

[Hurkyl]

Another quick-ish response.

A lot of people misunderstand the role of solipsism in science.

While I appreciate that you want to give your speech, you're responding to a straw-man. As far as I'm concerned, solipsism is fine -- it's a well-known irrefutable position. The problem is that you invoke it as if it changes anything in the discussion; I don't care if you think there's a world out there or if it's all in your head, because as far as I can tell, all of the important questions remain essentially the same.

And I really can't figure out why, at this level, you have in your head that MWI is doing anything differently than any other interpretation of quantum mechanics or any other physical theory.

Pretending we know what we does contribute to closedmindedness.

It's funny that you mention this, because it's one of my central points. Time and time again, we see that we should be listening to our physical theories rather than imposing our prior views of how reality behaves.

Einstein took the wave equation for light seriously, and worked out special relativity. There are lots of ways one could react -- one could

1. Retain absolute time and look for new physics that explain Einstein's weird results
2. Retain absolute time, and view Einstein's weird results as just being a method to get right answers, not something to be taken seriously as a description of reality
3. Accept Einstein's weird results, and use them to correct our understanding of reality

Why is #3 the right answer for special relativity, but #2 correct for quantum mechanics?

Kind of like those many worlds-- maybe the next theory that replaces quantum mechanics will benefit from a postulate along the lines of why the many worlds do not exist, a way to think of things where that which covers its tracks so well may be viewed as superfluous.

Maybe it will. And maybe the next theory of cosmology will benefit from a postulate of absolute time. However:

• The issue is irrelevant, because we don't have the next theory that replaces quantum mechanics or general relativity
• And besides, the better money by far is on the bet that science is moving in the right direction, rather than exactly the wrong one.

 

[Hurkyl]

But Everett did remove the Born rule from QM. (I don't remember where I read that. If you doubt it, I guess I can start looking for original articles). This removed the theory's ability to make predictions, and therefore made it something less than a theory. He must have thought that there would be some way to recover it.

Wikipedia skimming:

Everett (1957) briefly derived the Born rule ... and that its derivation was as justified as the procedure for defining probability in classical mechanics.​

I infer the latter clause to refer to interpreting probabilities as limits of frequencies, but I can't know for sure. It's stated that Gleason's theorem is a reproduction of Everett's work. I haven't fully absorbed the statement contained therein, though.

 

[Fredrik]

Wikipedia skimming:

Everett (1957) briefly derived the Born rule ... and that its derivation was as justified as the procedure for defining probability in classical mechanics.​

I infer the latter clause to refer to interpreting probabilities as limits of frequencies, but I can't know for sure. It's stated that Gleason's theorem is a reproduction of Everett's work. I haven't fully absorbed the statement contained therein, though.

I am very skeptical of physicists' "derivations" of the Born rule. The ones I've come across haven't exactly looked convincing. I think Gleason's theorem is the only way to justify the Born rule, but it seems to me that it can't be thought of as a derivation of the Born rule from the rest of QM. The rest of QM doesn't even suggest that probabilities are involved, let alone that they should be assigned by probability measures on the lattice of closed linear subspaces. But Gleason's theorem tells us that every probability measure on that lattice is of the form M\mapsto\operatorname{Tr(\rho P_M)}, where \rho is a positive, self-adjoint operator with trace 1, and P_M is the projection operator associated with the closed linear subspace M. Note that it doesn't just tell us how to calculate probabilities given a state. It also tells us what mathematical objects to think of as states.

 

[Ken G]

A
While I appreciate that you want to give your speech, you're responding to a straw-man. As far as I'm concerned, solipsism is fine -- it's a well-known irrefutable position.

Your words demonstrate that I was not responding to a straw man at all-- I was trying to explain solipsism to someone who misconstrues its purpose in exactly the way I expected. Your statement that it is an "irrefutable position" exposes exactly the misconception I attempted to dispell, as did Fredrik.

The problem is that you invoke it as if it changes anything in the discussion; I don't care if you think there's a world out there or if it's all in your head, because as far as I can tell, all of the important questions remain essentially the same.

Yes, indeed you have not understood my objectives at all, nor Fredrik's it would seem. There is actually a very constructive purpose in the challenge "if you would maintain that MWI is true, you must give me more than cognitive resonance, because I can get that using many different things other than mathematical logic that stems from an undecidable rationalist assumption" (like the one you correctly identified above). As I said, MWI is just one thing-- the ultimate rationalist dream about what physics is. The problem is, physics has never been that before, though the rationalists thought it was, time and time again. You seem to feel these past rationalist failures are harmless-- it "doesn't change the questions." The harm isn't in changing the questions, not at all-- it is in pretending that the questions are answered, when they are not. That's the whole point of solipsism-- to keep track of when the questions aren't really answered, so we don't close a door that might have the next answer sitting right behind it. You see that as a pointless mission, strangely. I counter that what is pointless is to mistake an interpretation of a theory as something true-- that is wrong on two counts:
1) the theory itself isn't even true, and
2) an interpretation of a theory is always less true than the theory, because interpretations are non-unique but effectively identical.
Indeed, the only value of an interpretation is it can either help you get the answers right (a personal and subjective issue), or, it might help you figure out how a theory needs to be changed by keeping some aspect of an interpretation. But since you don't know how it will need to be changed, it's best to be able to adopt any interpretation, and just see how things look from the new perspective. Marrying one interpretation just seems like utter folly to me, and I've never seen a convincing argument otherwise. Certainly not from the history of physics.

And I really can't figure out why, at this level, you have in your head that MWI is doing anything differently than any other interpretation of quantum mechanics or any other physical theory.

Why do you imagine I think that? Every objection I've given is to adopting MWI as truth about reality, rather than what it actually is-- an allowed interpretation of the QM theory. Why would we imagine MWI is anything other than what it is?

It's funny that you mention this, because it's one of my central points. Time and time again, we see that we should be listening to our physical theories rather than imposing our prior views of how reality behaves.

Time and time again, the postulates of our theories prove to not be quite the truth. That is innocent if we are only using them to predict outcomes, but it is awful if we think they are true and cease doing what science should always do-- maintain skepticism about our postulates and seek why they are wrong, not adopt faith-based world views that rely on them.

Einstein took the wave equation for light seriously, and worked out special relativity. There are lots of ways one could react -- one could

1. Retain absolute time and look for new physics that explain Einstein's weird results
2. Retain absolute time, and view Einstein's weird results as just being a method to get right answers, not something to be taken seriously as a description of reality
3. Accept Einstein's weird results, and use them to correct our understanding of reality

Why is #3 the right answer for special relativity, but #2 correct for quantum mechanics?

You are confusing the choosing of the postulates of a theory with forming an interpretation of those postulates. How did MWI help us choose the postulates of QM? You can ignore the rest of my "speech" if you have any decent answer to that.

 

[Fra]

I don't know if Ken and me are subscribing to the same "solipsism" or not but just one last input on this (try to answer from My perspective)

And I really can't figure out why, at this level, you have in your head that MWI is doing anything differently than any other interpretation of quantum mechanics or any other physical theory.

What I suggested above as "solipsism at it's finest" is actually a interpretation, that understands QM as a limiting case, thus the "solipsist extremist" like myself, is convinced that QM as it stands needs to be generalized. Not in order to explain what it already can do well, but in order to solve open issues with unification and gravity.

This is the reason for my preference. Before I started to think seriously about these things - say more than 10 years ago - I preferred a more axiomatic philosophy. Until I started to understand the problem with that perspective, in that it's completely ignorance of the process of inducing and developing the axiom system. Because this is the non-trivial part.

/Fredrik

 

[Hurkyl]

Why do you imagine I think that? Every objection I've given is to adopting MWI as truth about reality, rather than what it actually is

If you truly are not singling MWI out as being different, then most of what you've said is completely irrelevant. If your objection applies to all of physics, then it doesn't have any bearing on the discussion.

It also means you've missed my point about solipsism: the entire discussion about interpretations remains unchanged, except for everywhere formally substituting "truth about reality" with whatever refinement you find appropriate.

But I think you really are singling MWI out; e.g.

As I said, MWI is just one thing-- the ultimate rationalist dream about what physics is.

but I can't see anything MWI is doing differently from anything else.

That is, other than challenging a viewpoint on reality that is married to the idea of definite outcomes.

But since you don't know how it will need to be changed, it's best to be able to adopt any interpretation, and just see how things look from the new perspective.

One interesting point, though, is that perspectives aren't disjoint. MWI, for example, already includes all of the CI viewpoint, except for the specific point that collapse is the 'truth about reality' -- that outcomes are definite from the bird's eye view.


And while it is best to be familiar with many interpretations, it is certainly not true that one should weigh them all equally. The interpretations inspired by the physical theory, for example, should be given more weight than interpretations that run contrary to it. (e.g. special relativity should be given more weight than the equivalent Lorentz ether theory)


Similarly, interpretations more likely to promote understanding, and more likely to relate to future theories should also be given more weight. CI certainly fails on the latter count, and IMO it fails on the first count -- it looks like a quick-and-dirty hack meant to quickly connect quantum mechanics to the warm and fuzzy classical way of thinking about things... and it really does look like the product of an era that thought unitary evolution was theoretically incapable of being consistent with our observations of reality.

Time and time again, the postulates of our theories prove to not be quite the truth.

Right -- "not quite the truth". That's very different from "diametrically opposed to the truth".

 

[Hurkyl]

I agree that those two questions warrant negative answers, but I see all interpretations of quantum mechanics (other than deBB) as giving negative answers to those questions.

Every variant of the Copenhagen interpretation I have ever seen gives a positive answer to the first^* -- I don't know how an interpretation could justifiably be labelled "CI" if it does not.

Proponents of CI often (maybe even usually) give a positive answer to the second as well.


*: except they may object to using the phrase "time evolution" to describe it -- but that objection has never seemed convincing to me.


Grrr, and I've run out of time again. I'll take another shot at responding to this post later.

 

[Ken G]

If you truly are not singling MWI out as being different, then most of what you've said is completely irrelevant. If your objection applies to all of physics, then it doesn't have any bearing on the discussion.

What I am singling out about MWI is the fervor with which people hold to it. That's why I mentioned quantum suicide a few times. I think it's perfectly clear that the MW is more than an I to those who hold to it, it is a faith-based world view (faith in postulates, not religious faith). That is also why I mentioned rationalism so many times-- rationalism is a means of attributing truth, it is much more than a non-unique equivalent interpretation of a physics theory.

It also means you've missed my point about solipsism: the entire discussion about interpretations remains unchanged, except for everywhere formally substituting "truth about reality" with whatever refinement you find appropriate.

Two simple questions I would put to you to clarify the claims you are making on the MWI:
1) do you think that physics tells us there really are multiple worlds?
2) did thinking this help anyone create any postulates of any physical theory?

That is, other than challenging a viewpoint on reality that is married to the idea of definite outcomes.

MWI does a whole lot more than challenge viewpoints about definite outcomes! Just look at the language people use around it. MWI makes assertions about indefinite outcomes. Simply challenging beliefs about definite outcomes is the approach of solipsism. Ironic that you would now claim to hold the solipsistic high ground!

One interesting point, though, is that perspectives aren't disjoint. MWI, for example, already includes all of the CI viewpoint, except for the specific point that collapse is the 'truth about reality' -- that outcomes are definite from the bird's eye view.

Sure, that's because CI is a more minimal ontology than MWI. That's a strength of CI, not a weakness as you are painting it here.

The interpretations inspired by the physical theory, for example, should be given more weight than interpretations that run contrary to it. (e.g. special relativity should be given more weight than the equivalent Lorentz ether theory)

That could change in a heartbeat if the next great theory invokes an aether. You are not stating any kind of physical principle here, it's an expression of prejudice only. Actually, the pendulum tends to swing.

Right -- "not quite the truth". That's very different from "diametrically opposed to the truth".

I was referring to the postulates, not the interpretations. Interpretations have no relationship with truth, they have relationships with postulates, and their ontological stance is merely informal, like a picture. That's the problem with how people talk about MWI, right there-- they think it is something more than a picture, but they cannot justify treating it such without simply marrying the philosophy of rationalism.

 

[Ken G]

Every variant of the Copenhagen interpretation I have ever seen gives a positive answer to the first^* -- I don't know how an interpretation could justifiably be labelled "CI" if it does not.

OK, then we have a point of interest here: we have different interpretations of what CI is. The question you brought up before was:
Is "collapse" under time evolution part of the "holistic information required to predict and understand a system"?
I would say the defining difference between MWI and CI is the direction of the logic, as we can see by considering how the following puzzle is resolved: "A quantum system is in state S, not an eigenstate of A. I apply measurement A, and this places the quantum system in an eigenstate of A, which is nonunitary on the subsystem in question so is called a collapse of the subsystem." The puzzle is what do we do with the whole system. MWI asserts that the postulates of QM should be interpreted globally, independent of the measurement and its registration by the observer. That requires that the closed system evolve unitarily, so the apparent nonunitary evolution of the subsystem and its macro pointer is simply due to a projection from the "many worlds". This carries over into the classical world-- I flip a coin and both results occur, I just perceive one result for some reason that MWI has no description for. So everything is great on the postulate end, and the problems appear at the perception end of the story. This is exactly what a rationalist would expect-- good postulates, troublesome perceptions.

CI takes the opposite direction to the logic. Being primarily empiricist, CI simply takes the perceptions as the reality. So it has no difficulty at the perception end-- we perceive a nonunitary evolution of the closed system that we are part of, so the closed system evolves nonunitarily. No problem yet. But CI uses the postulate that closed systems evolve unitarily when not being observed, so it has a problem tracking back to the postulates-- the postulates seem incomplete, they have no account of measurement. But CI accounts just fine for decoherence-- it is perfectly aware that the action of an operator is to decohere a subsystem across eigenboundaries, if you will. It merely asserts that once the decoherence has occured, according to the postulates, we perceive only one outcome (which is true, we do), so the postulates have reached the end of their usefulness and we make no ontological claims about what happened there other than what we observe. Ergo, the postulates, and their associated wave functions, were never there to tell us what reality is doing, they were always there to help us predict what we observe. Empiricism at its best-- the map isn't the territory.

So now you see why I claim CI answers "no" to your question above-- the information used to predict an outcome is all contained in the wave function, there is nothing that has anything to do with collapse in the holistic information! (Note I take "collapse" to mean the collapse of our perceptions-- the decoherence across eigenboundaries is there in all interpretations, it's simply an observable element of the theory of quantum mechanics.) CI doesn't assert there is any more "holistic information" in the collapse than MWI does, it merely takes a different stance about what is the actual reality. CI asserts that there is no information in the collapse-- which outcome we perceive is unpredicted by the theory, there is no such thing as holistic information that points to what outcome will occur, only the statistical holistic information that is in the wave function (and is identical to the detectable part of the holistic information in MWI).

Proponents of CI often (maybe even usually) give a positive answer to the second as well.

I think you misunderstand the CI. You are referring to the issue:
Is there any "holistic information required to predict and understand a system" that is not encoded in the wave-function?
Let us take Bohr for example-- he claimed there was no such thing as the "quantum world." It's all just information, held at the classical level, about some ?. He steadfastly refused to adopt any quantum ontology at all, and the only classical ontology he would accept was what was empirically demonstrable. So he would have answered "no" to both those questions, if I were to try and "channel him". Bohr felt the wave function encoded all the holistic information required, indeed, all the holistic information in existence, to predict and understand a system. The predictions are simply statistical in nature-- and that is understanding them, as well as can be done. MWI does not predict the outcomes of the experiment either-- unless you rig it to claim that it has to be right even though you cannot demonstrate what prediction it has made beyond something statistical.

So the bottom line is, CI and MWI take all the same holistic information that is actually detectable and testable, and do something different with it. CI builds the reality from the perceptions, so is perfectly happy to leave the information as intrinsically statistical, and this simply makes the postulates incomplete in regard to registering the pointer. Empiricist expectation: the reality is what we experience, what we are doing in our minds is just a kind of game that works for some mysterious reason. MWI takes all that detectable holistic information and embeds it in a larger matrix of undetectable, essentially fantastical, holistic information, so as to preserve the completeness of the postulates. Classical rationalism-- the postulates are the reality, our means of connecting to nature is by thinking about it. But it is important to note that MWI has no accounting whatsoever for why we perceive what we do-- it would require some kind of theory of mind to even attempt it. CI would also need a theory of mind to explain the incompleteness of the postulates-- it's just a different role for the mind, based on whatever is missing-- the thoughts or the perceptions.

 

[Eqblaauw]

obviously no-one knows for sure, and every interpretation has a what if elements. So we simply cannot decide. Many mwiers are trying to convince people that phycisists are in favor of mwi, that is simply false. Wikipedia,s mwi is a fansite and blatend propaganda, selecting only very questionable favorable polls, and viewpoints in favor of mw. When you look at the talk page of the wiki mwi site you will see, another story that presents another viewpoint; a negative reception. I'm not saying it's correct, but i think it's less biased then the current page.
Let's wait a minute before we assume we have a bilion copies, and wait till we know more

 

[BruceW]

Also, different people think MWI means different things, which causes even more confusion on this topic.
In any case, this thread should really be in metaphysics or philosophy section, since there are no experimental results which discern between the different interpretations.

 

[Hurkyl]

What I am singling out about MWI is the fervor with which people hold to it.

I'm not interested in this topic of sociology.

Two simple questions I would put to you to clarify the claims you are making on the MWI:
1) do you think that physics tells us there really are multiple worlds?
2) did thinking this help anyone create any postulates of any physical theory?

For question 1: yes. I add two caveats:

• It "really tells us" in the same sense that, e.g., Special Relativity "really tells us" that simultaneity is relative.
• "World" refers to a property of quantum states, rather than any prior notion

Normally I wouldn't bother mentioning that first bullet point, but it seems important to you so there it is.

For question 2: I don't consider it important which statements in a theory we single out to be called "postulates". But I will try to answer in what I think the spirit of the question is.

MWI is quantum states evolving by unitary evolution. Decoherence has provided a compelling mechanism by which the appearance of classicalness can emerge from such a process. (if another mechanism was discovered, I'm not sure if I would include it under the scope of MWI research, or if I would prefer to restrict MWI specifically to a decoherence-based approach)

FWIW, to the extent of my awareness, every promising line of research towards the next physical theory that works both on quantum scale and on macroscopic / cosmological scale retains the quantum states + unitary evolution setup, and crucially relies on the above to provide plausibility that it has any chance of matching observation.

I'm not aware of any path forward that eschews the quantum setup, or embraces collapse as the means by which it expects to agree with observations.

On a less ambitious note, I did some searching and found a quote from Decoherence, the measurement problem, and [URL="https://www.physicsforums.com/insights/fundamental-difference-interpretations-quantum-mechanics/"]interpretations of quantum mechanics[/url]:

Bub (1997) termed decoherence part of the “new orthodoxy” of understanding quantum mechanics—as the working physicist’s way of motivating the postulates of quantum mechanics from physical principles​

MWI makes assertions about indefinite outcomes.

But those aren't new assertions -- they are assertions that are consequences of the basic setup of quantum mechanics: quantum states and unitary evolution.

Because of where they come from, every interpretation of quantum mechanics includes these assertions in some form -- even the CI. The only reason, e.g., CI doesn't pay too much attention to these assertions is because it invokes a collapse before they can become interesting.

Simply challenging beliefs about definite outcomes is the approach of solipsism. Ironic that you would now claim to hold the solipsistic high ground!

There's some sort of converse fallacy in there, but that aside, aren't you being inconsistent? I sure got the impression you had previously been defending the definite outcomes of CI a while ago.

Sure, that's because CI is a more minimal ontology than MWI.

Eh? In the most basic CI, they have identical ontologies: wave-functions correspond to what's "real".

Unless you include how wave-functions evolve over time, in which case MWI is the minimal one -- CI has both unitary evolution and collapse, MWI has only unitary evolution.

You are not stating any kind of physical principle here, it's an expression of prejudice only.

Eh? It's the most central tenet of empiricism -- that we derive knowledge from the results of observation and experiment!

Interpretations have no relationship with truth, they have relationships with postulates,

Interpretations, more or less by definition are means by which we give meaning to the elements of our physical theory.

And I take a rather formal bent on the word "truth" -- the concept of a truth value is, more or less, defined to be type of thing that propositions are mapped to under an interpretation.

Because of the baggage people like to associate to it, I don't particularly like using the word -- unfortunately it's rather awkward to talk about in English about the subject of semantics without using it.

That's the problem with how people talk about MWI, right there-- they think it is something more than a picture, but they cannot justify treating it such without simply marrying the philosophy of rationalism.

If you don't think interpretations are more than pictures, then stop thinking about MWI as being more than a picture then.

 

[Hurkyl]

I just perceive one result for some reason that MWI has no description for.

I'm not aware of MWI lacking a description for this. For any two distinct mutually exclusive outcomes X and Y, "I perceive X and I perceive Y" is fairly clearly an outcome no observer could ever perceive.

It merely asserts that once the decoherence has occured, according to the postulates, we perceive only one outcome

No, it asserts more strongly that the outcome is definite.

we make no ontological claims about what happened there other than what we observe.

This is the collapse, and specifically what MWI does not do. It's enough to claim that we perceive one outcome -- we don't have to take the extra step and tell reality that it has chosen a definite outcome.

So now you see why I claim CI answers "no" to your question above

What you have said above post amounts to a "yes" answer to the intent of my question. What I hadn't noticed until now is that you were picking at my phrasing to justify a negative response -- e.g. claiming collapse as something incomprehensible, and thus cannot be part of how we "understand" the behavior of the system.

 

[Ken G]

For question 1: yes. I add two caveats:

• It "really tells us" in the same sense that, e.g., Special Relativity "really tells us" that simultaneity is relative.

• And this is indeed the crux of the entire issue. Do you see this as a lesson of special relativity? I do not, actually. Instead, I see the lesson of special relativity to be that simultaneity should be thought of as an invariant equivalence class (essentially spacelike separation), it was wrong of us to think that simultaneity had a physical meaning in terms of a one-to-one matching of events "in the now". The core lesson of relativity is that physical theories must be built from invariants, like the above equivalence class, and not from coordinates, like a one-to-one matching of the "now" of different observers.
  
  So yes, I agree that buying MWI as a world view gives us that lesson in the same way that SR gives the lesson you cite-- and I see them both as false lessons for much of the same reasons. Saying that simultaneity is relative is mistaking a coordinate system, which is arbitrary, for a statement of physical fact, which should be invariant. The point is that the "lessons" of our theories should be restricted to what those lessons really are, without reading additional things into them. In the case of MWI, the lesson is that accepting the postulates as literally true requires imagining many worlds. That's completely true, and no more. That does not in any way adjudicate between the two clear choices: go ahead and accept the postulates as literally true (MWI) or conclude the postulates cannot be literally true because they lead one into what is essentially fantastical thinking, so instead they are just what they are: the postulates of a mathematical structure that allow us to make statistical predictions for an essentially mysterious reason. That's what all physical postulates are, and mystery is preferred over fantasy.
  

  MWI is quantum states evolving by unitary evolution. Decoherence has provided a compelling mechanism by which the appearance of classicalness can emerge from such a process.

  Decoherence doesn't say anything the least bit different in an MWI approach as in a CI approach. You don't even need an interpretation yet, decoherence simply limits the types of wavefunctions that either approach is going to need to interpret. The fundamental rift between MWI and CI happens later-- it happens when we notice that decoherence has destroyed correlations between what we do observe, and what we do not observe. Thus, decoherence is not a reason to favor MWI, it is the way to see that MWI and CI are physically equivalent-- we either don't see those "other worlds" because we don't cohere with them, or we don't see them because they do not exist. No physical difference at all. But MWI is a world view, a claim on reality, whereas CI is an honest admission of lack of knowledge about that reality. MWI is faith-based (faith that the postulates are true), so is closed, CI is solipsistic in the constructive sense, so is open to admitting what it simply does not know.
  

  FWIW, to the extent of my awareness, every promising line of research towards the next physical theory that works both on quantum scale and on macroscopic / cosmological scale retains the quantum states + unitary evolution setup, and crucially relies on the above to provide plausibility that it has any chance of matching observation.

  That claim is not supportable. You are essentially claiming that because rationalists believe in their postulates, this is somehow going to be needed for the next theory. String theory requires MWI? No. String theory is completely compatible with CI, for all the same reasons that quantum mechanics was. The history of physics is rife with the rationalist dream of a "TOE", and every generation thought they essentially had it, just a few minor tweaks is all they would need. Until, that is, they discovered that the new theory had postulates that were dramatically different in their ontological suggestions. But that won't happen this time, you say...
  

  I'm not aware of any path forward that eschews the quantum setup, or embraces collapse as the means by which it expects to agree with observations.

  None of the theories "embrace" collapse any more than they embrace MWI. All the theories we have now are completely ambivalent to these questions, and the next theory could just as easily be inspired by CI empiricism as by MWI rationalism. We just don't know, because right now we don't even have a theory that is any different in the two veins, so even imagining a theory that would care about that difference is difficult. Remember, the unitary postulate is not required as a claim against reality in any of these theories, it is merely a claim on how the mathematics works. As long as their is a Born rule, we cannot physically distinguish MWI and CI.
  

  On a less ambitious note, I did some searching and found a quote from Decoherence, the measurement problem, and [URL="https://www.physicsforums.com/insights/fundamental-difference-interpretations-quantum-mechanics/"]interpretations of quantum mechanics[/url]:
  

  Bub (1997) termed decoherence part of the “new orthodoxy” of understanding quantum mechanics—as the working physicist’s way of motivating the postulates of quantum mechanics from physical principles​

  But this line of reasoning is not relevant to the issue of distinguishing MWI and CI, because as I said, CI is completely consistent with decoherence. In fact, CI needs decoherence just as badly as MWI does-- all decoherence does is bring us to the place where we then need to make a choice in how we will interpret the meaning of the Born rule. What is clear enough is that the phenomenon of decoherence makes it impossible to physically distinguish CI from MWI. Do you disagree?
  

  But those aren't new assertions -- they are assertions that are consequences of the basic setup of quantum mechanics: quantum states and unitary evolution.

  No, that's exactly the point-- MWI is not a consequence of the postulates about unitary evolution, it is a consequence of an additional claim about what is actually evolving unitarily. CI has all the same unitary evolution in the wave function as MWI! The sole difference is that CI does not interpret the thing that is evolving unitarily as the reality itself, it interprets the thing that is evolving unitarily is the machinery for making predictions. In short, CI calls a postulate of a theory a postulate of a theory, and MWI calls it a claim on reality. Which is more honest about what we know?
  
  

  There's some sort of converse fallacy in there, but that aside, aren't you being inconsistent? I sure got the impression you had previously been defending the definite outcomes of CI a while ago.

  I'm not sure what you are saying here, in my view there is nothing subtle or ambiguous about the concept of definiteness. From the empirical standpoint, what is definite is what is observed, period-- it's completely straightforward. That's the whole point of the CI, in fact. I was saying that the MWI takes the rationalist tack of putting stock in the indefinite outcomes of observations-- the outcomes in the "other worlds" which are indefinite in the sense that we have no empirical evidence to say those outcomes ever do actually occur anywhere.
  
  

  Eh? In the most basic CI, they have identical ontologies: wave-functions correspond to what's "real".

  Absolutely not. This may be our core disconnect-- I maintain that you maintain a kind of caricature of the CI. The core tenet of the CI, as it seems to me it is described by Bohr and Heisenberg, is that quantum mechanics is an information theory. It is a theory about how we must update our knowledge of a system as one of two things happen: closed systems evolve in time (including decoherence), or a measurement is registered (which is a "manual" type of information updating). As such, no wave function is ever regarded as real in CI, it is always regarded as a mathematical device that enters into statistical predictions via the Born rule. The fundamental reality there is not the wave function, and indeed is regarded as inscrutable. Why would Born say "there is no such thing as the quantum world" if he thought a wave function was real?
  

  Unless you include how wave-functions evolve over time, in which case MWI is the minimal one -- CI has both unitary evolution and collapse, MWI has only unitary evolution.

  We agree there-- CI has two fundamental types of information updating, whereas MWI doesn't view it as information at all, it views it as the reality. The unification of these two aspects of CI is the main advantage of MWI, but it comes at a pretty significant cost-- it forces us to drop empiricism as the way of demonstrating scientific truth, in favor of the rationalistic desire to say "what unifies is what is real." It's the rationalistic dream that our access to truth is via our concepts, rather than by our experiences. But this is no surprise to anyone who understands rationalism.
  
  

  Interpretations, more or less by definition are means by which we give meaning to the elements of our physical theory.

  Not quite, because everything that matters in that sentence is all locked up in the word "meaning." Empiricism and rationalism use two completely different approaches to that "meaning"-- to a rationalist, the meaning of a theory is the reality itself, and to an empiricist, the meaning of a theory is a kind of mental game, like chess, whose outcome is found to be successful at predicting what is the actual reality-- the outcome of experiment. So when the CI "gives meaning" to the postulates of QM, it does so without making ontological claims on any reality other than the outcome of the observation, because there is no reality other than the outcome of the observation. That simple statement rules out many worlds a priori.
  

  And I take a rather formal bent on the word "truth" -- the concept of a truth value is, more or less, defined to be type of thing that propositions are mapped to under an interpretation.

  Exactly-- you are a rationalist. I already know that. There's nothing wrong with being a rationalist as long as you understand that you are one.
  
  

  If you don't think interpretations are more than pictures, then stop thinking about MWI as being more than a picture then.

  But I have no problem at all with MWI as a picture. In fact, I have no problem at all with MWI as a natural world view for a rationalist. All I'm saying is, people who adopt MWI don't think of it as a picture, they think of it as a world view, which means they are rationalists. That in turn means that they have bought off on a suite of perspectives that have been well studied throughout the history of both physics and philosophy, and its pitfalls and successes have been well documented. I happen to be pointing to the pitfalls, but I do not deny that theoretical physicists tend toward rationalism, and they have certainly had their successes. But if you look at someone like Einstein and his relativity, what you see is a careful balance between the aesthetic power of unifying principles like the need for the laws to be the same for all observers, and the rock-solid grounding that essentially says "build your postulates from that which all observers can agree on." Can all observers agree on the different outcomes of the observations that appear in the many worlds? There's a more violent departure from empiricism there than rationalists have previously been willing to make, so it ushers those pitfalls right to the fore.

 

[Ken G]

I'm not aware of MWI lacking a description for this. For any two distinct mutually exclusive outcomes X and Y, "I perceive X and I perceive Y" is fairly clearly an outcome no observer could ever perceive.

I already had a long post, so I won't go on here, but I wanted to address this issue. I just don't see how you can claim that MWI has any way to account for the particular outcome of an observation. In MWI, all outcomes occur in the "actual reality", so there simply is no accounting for us in "our branch" of the reality-- our placement in this branch is undescribed by MWI, so it has already lost what it claimed was its raison d'etre-- unity and completeness in the description. What answer can be given, that is fundamentally dictated by MWI, to the question, why did I perceive the result I did? Is there some attribute of me that requires I get "spin up" rather than "spin down", or are we to imagine that there is some additional process that decides what I'll get versus what someone else in another world will get? There's no avoiding the need for a "model of me" to enter the story. That's exactly what CI avoids-- by accepting, a priori, the outcome I get as the actual reality, there is immediately no need to account for why I got what I got-- I got it because that's what happened. Why that happened, and not one of the other possibilities, is regarded as fundamentally mysterious in CI. The question goes away because we have not set out to answer it, whereas MWI claims to leave nothing as fundamentally mysterious, it purports to be a complete description of what is actually happening-- but it doesn't deliver without a model of me.

That's where I am all for solipsism-- because it is simply being honest about what we can know, and what we can only pretend to know. The price for pretending is it might close a door on some new insight that we need to leave open (perhaps along the lines of what Fredrik was saying he wanted to do, who knows). So there's nothing wrong with the MWI as a way to picture QM, the problem is using it to make claim on a scientifically supported world view. It's no better than using Newton's laws to make a claim on a scientifically supported world view about a mechanistic/deterministic universe of particle trajectories, and look at how that world view panned out.

 

[Dmitry67]

so there simply is no accounting for us in "our branch" of the reality-- our placement in this branch is undescribed by MWI

Is there some attribute of me that requires I get "spin up" rather than "spin down", or are we to imagine that there is some additional process that decides what I'll get versus what someone else in another world will get?

MWI redefines the very notion of "me". You are not a world line with a single history. You are a tree with all histories. You remember only the past, so when any leaf look down to the root, it can track "his own" branch. However, leaves can't look up (you don't see the future) and around (communication is lost to the parallel branches) So any leaf has an illusion that it is unique.

I think your problem with MWI is just psycological, am I right?

 

[BruceW]

Ken G - To explain how MWI works:
lets say you make a measurement on a quantum system with two outcomes, and decide to colour your hair red or blue depending on the outcome.

Now, MWI says non-unitary collapse doesn't happen. So there is always some extremely small possibility of the state going back to what it was before you made the measurement. (In other words, you would go back to having zero knowledge of the outcome) (and your hair would go back to its natural colour). So now you ask "well, for some time, I had measured the outcome, so what colour was my hair at that time?" The answer to this is clear, it was a superposition of the two different colours. So now you ask "which was 'me'?" Well, clearly 'you' were not just one or the other, and 'you' weren't both. The only way to describe 'you' was as a superposition.

We all experience ourselves as being with either red or blue hair after one of these experiments, and we don't really care if our state is eventually returned to how it was before the experiment. So now a good question is "what is the probability of experiencing having either red or blue hair after the experiment?" and the answer is given by the Born rule.

So in my first example, I said 'you' were a superposition of hair colours. And in the second example, 'you' had just one hair colour. Both are viable definitions of 'you'. So in MWI, there is more than one possible definition of 'you'.

 

[mitchell porter]

Dmitry67 and BruceW, you should explain at some point where the unequal probabilities come from. On another website I recently proposed that MWI is like a car dealer who, asked to deliver 9 BMWs and 16 Rolls-Royces, instead shows up with one BMW with a "3" painted on it, and one Rolls-Royce with a "4" painted on it, saying you just have to square the numbers and you'll get the cars you ordered. If you are going to explain quantum mechanics by having a multiverse in which all possible outcomes actually exist, then you need to have an outcome which empirically occurs, say, twice as often, actually occurring twice as often in the multiverse.

 

[BruceW]

The unequal probabilities are given by the Born rule. As I just said: So now a good question is "what is the probability of experiencing having either red or blue hair after the experiment?" and the answer is given by the Born rule.

The Born rule is used in both MWI and CI, I don't understand your objection to it.

Just to make clear, I'm using the definition of MWI as simply 'no non-unitary collapse happens', which is the most widely-held definition of MWI, as far as I'm aware.

 

[Dmitry67]

Born rule was always the weak point of MWI, but in 2010 Born Rule was derived from QM by Max Tegmark in the assumption of an Infinite Universe.
http://arxiv.org/abs/1008.1066

 

[Eqblaauw]

Tegmark is known for saying 'Many worlds, or Many words', I think you're equally right when you say 'Many Gods or Many Words'. When you don't have a problem with many-worlds it is psychologial I would say.
And about the whole splitting, this is part of the reaction to an email that I send to David Deutsch:
Dear Edo Blaauw -- > Dear David Deutsch,> > I have a few questions about the many-worlds interpretation: > Are there according to most of the many-world believers many universes exactly the same until they differentiate, Yes. > or do most of the 'mwi-ers' believe we are continuously being split? That is an older view, of which I currently know of no supporters among people working on the theory.

 

[Fredrik]

The unequal probabilities are given by the Born rule. As I just said: So now a good question is "what is the probability of experiencing having either red or blue hair after the experiment?" and the answer is given by the Born rule.

The Born rule is used in both MWI and CI, I don't understand your objection to it.

Just to make clear, I'm using the definition of MWI as simply 'no non-unitary collapse happens', which is the most widely-held definition of MWI, as far as I'm aware.

He's objecting because it's very common to define the MWI as QM without the Born rule, applied to the whole universe. Of course, if you remove the Born rule, what you have left is no longer a theory, since it can't be used to make predictions. So someone who defines the MWI that way will have to find a way to derive the Born rule if they want to be taken seriously. There are lots of attempts to do that, but I haven't seen one yet that I find convincing. (Gleason's theorem is certainly convincing, but what it tells us isn't quite what the people who define the MWI this way are required to show. See my comments in post #159). I think more and more people are realizing that this approach just doesn't work.

On the other hand, I don't think there are any strong arguments against the idea that QM with the Born rule can be interpreted as describing many worlds. I think that if it describes anything at all, it would have to be either many worlds, or a single world where some of the normal rules of logic don't apply. However, there's also the possibility that it doesn't really tell us what's actually happening in every situation, i.e. that it's first and foremost a set of instructions that tells us how to assign probabilities to verifiable statements.

 

[BruceW]

He's objecting because it's very common to define the MWI as QM without the Born rule, applied to the whole universe. Of course, if you remove the Born rule, what you have left is no longer a theory, since it can't be used to make predictions. So someone who defines the MWI that way will have to find a way to derive the Born rule if they want to be taken seriously. There are lots of attempts to do that, but I haven't seen one yet that I find convincing. (Gleason's theorem is certainly convincing, but what it tells us isn't quite what the people who define the MWI this way are required to show. See my comments in post #159). I think more and more people are realizing that this approach just doesn't work.

Interesting, I didn't know that. Do you know why it is common to define MWI without any kind of Born rule? Maybe its because the MWI advocates wanted to get as far away from conventional QM as possible... I would personally have kept in the Born rule as a fundamental postulate of QM. (And just taken out the bit about non-unitary collapse).

 

[BruceW]

On the other hand, I don't think there are any strong arguments against the idea that QM with the Born rule can be interpreted as describing many worlds. I think that if it describes anything at all, it would have to be either many worlds, or a single world where some of the normal rules of logic don't apply. However, there's also the possibility that it doesn't really tell us what's actually happening in every situation, i.e. that it's first and foremost a set of instructions that tells us how to assign probabilities to verifiable statements.

I don't fully get what you're saying here. You were saying before that people who define MWI as QM without the Born rule would have to derive the Born rule. So what difference would there be between 'MWI with a derived Born rule' and 'MWI with a postulated Born rule'? (Apart from the fact that the derived Born rule is preferred, since it would mean there were less postulates in the theory).

 

[Ken G]

So now you ask "which was 'me'?" Well, clearly 'you' were not just one or the other, and 'you' weren't both. The only way to describe 'you' was as a superposition.

I don't believe that MWI asserts that. First let us not make a model of "me", let us simply make a model of "him." "He" is part of a system, but he is not a whole system. Thus "he" is not in a superposition state, he is projected from a pure state, and projections from pure states onto subsystems are mixed states. That much is true in either MWI or CI. The difference only appears when "I" enter the picture. because that is the only time that I have to account for a single outcome-- the place where the single outcome appears is purely in my perception. If there were no perceiving agents, MWI would be a complete description, but a description of what? What is a "description" when there are no perceiving agents interested in having something be described? The universe doesn't need physics, physics is for perceiving agents to try and understand their perceptions.

So now a good question is "what is the probability of experiencing having either red or blue hair after the experiment?" and the answer is given by the Born rule.

That is a question that is answered the same by CI or MWI, if MWI is using the Born rule (if it isn't, it's not even physics, as Fredrik pointed out). The important question is "what determined the outcome that I perceived?" Where is the complete accounting for that in MWI? I already said how CI accounts for that-- what actually happens determines what I perceive, so the perception is explained (and nothing explains what actually happened-- CI makes no pretense of being a complete description of a process it views as fundamentally mysterious.) How can MWI pretend it has removed the mystery there, without creating a model of me (which it does not do, and that's why we get incompletenesses like quantum suicide)?

So in my first example, I said 'you' were a superposition of hair colours. And in the second example, 'you' had just one hair colour. Both are viable definitions of 'you'. So in MWI, there is more than one possible definition of 'you'.

Now you are getting closer to creating a "model of me" that could work in MWI. But then we must ask-- does this model of me really work? It's simply not true that the model could have me be a superposition, as I said above, but I could perhaps be modeled as some kind of mixed state. That doesn't really work though, because it just pushes the question back to "if I'm a mixed state, why do I perceive myself as being in a definite state?" The need for a "model of me" has not been satisfied, that question remains as the fundamental incompleteness of MWI. CI, on the other hand, has no issue with its "model of me" (it asserts that I am a perceiving agent whose perceptions define reality, and the consistencies found in those perceptions are the domain of physics), but its incompleteness appears at the other end-- it cannot account for what happens, beyond a statistical description. So take your pick-- have trouble saying what determines what happens (CI), or have trouble saying why it happened to you. I don't see any great advance in creating the fantastical realm of the many worlds, other than allowing me to marry the postulates of a theory that will most likely be found to be incorrect at some future time.

 

[Ken G]

MWI redefines the very notion of "me". You are not a world line with a single history. You are a tree with all histories.

I don't think that will succeed very well as a model of me. It has two big problems:
1) it does not explain why I do not perceive such a tree, and
2) it does not account for how "I" diverge along different branches of the tree. What makes each branch the same "me", that there is someone in it with my genes? How much does my appearance or personality have to change before the branches are to be associated with different "me"s? And what about all the "yous", am I alone on my tree because the tree is what is actually me? This is the point I'm making, we must not imagine that MWI has a reasonable model of me without subjecting that model to scrutiny. Taking your model of me to the limit, all of reality is me. It's nice monism, but not really a very useful model of me.

 

[BruceW]

Now you are getting closer to creating a "model of me" that could work in MWI. But then we must ask-- does this model of me really work? It's simply not true that the model could have me be a superposition, as I said above, but I could perhaps be modeled as some kind of mixed state. That doesn't really work though, because it just pushes the question back to "if I'm a mixed state, why do I perceive myself as being in a definite state?"

'I' could have been a pure superposition of quantum states at an earlier time. 'I' as I perceive myself right now, am in my bedroom. But at an earlier time, my state could have been a superposition of 'in bedroom' and 'in kitchen'. As long as there is no information on whether I was in the kitchen or bedroom, then I could have been a superposition of both.

So It's true that I never see myself as a superposition, but that doesn't mean I don't have the ability to be in a superposition.

 

[BruceW]

I don't think that will succeed very well as a model of me. It has two big problems:
1) it does not explain why I do not perceive such a tree, and
2) it does not account for how "I" diverge along different branches of the tree. What makes each branch the same "me", that there is someone in it with my genes? How much does my appearance or personality have to change before the branches are to be associated with different "me"s? And what about all the "yous", am I alone on my tree because the tree is what is actually me? This is the point I'm making, we must not imagine that MWI has a reasonable model of me without subjecting that model to scrutiny. Taking your model of me to the limit, all of reality is me. It's nice monism, but not really a very useful model of me.

This is the thing with MWI, it doesn't give a precise definition of 'me, here, now'. Since if a few of my cells changed, then would I notice the difference? No. So If there were a pure quantum superposition of 'me plus a few cells' and 'me minus a few cells', then any linear combination of them could also be argued to be 'me, here, now'. But then if I went and counted all my cells, then the new definition of 'I' would have to be one or the other, not a superposition.

 

[Fredrik]

Interesting, I didn't know that. Do you know why it is common to define MWI without any kind of Born rule?

I can only speculate. My guess is that the two main reasons are a) since the whole MWI debate started with Everett, people like to use definitions that are similar to his, and b) people are used to viewing probabilities as "number of favorable outcomes"/"number of outcomes", so they were thinking that maybe that's exactly what the probabilities in QM are.

I don't fully get what you're saying here. You were saying before that people who define MWI as QM without the Born rule would have to derive the Born rule. So what difference would there be between 'MWI with a derived Born rule' and 'MWI with a postulated Born rule'? (Apart from the fact that the derived Born rule is preferred, since it would mean there were less postulates in the theory).

There wouldn't be any significant difference. I agree that a derived Born rule would be preferred. I just haven't seen any derivations that I consider valid. (Gleason's is valid, but his theorem doesn't explain why there's a need to assign probabilities at all, or why they should be associated with Hilbert subspaces). So I don't believe that there is a MWI without the Born rule.

Adrian Kent examined Everett's and similar approaches to the MWI in his Against many-worlds interpretations. His conclusion was very critical.

Everett’s original attempt at a many-worlds interpretation was an interesting and well motivated idea. Everett wanted simultaneously to simplify the axioms of quantum mechanics and to produce a realist physical theory (one which gives an explicit mathematical description of an external reality). However, it is clear that he did not succeed. Successive advocates of many-worlds interpretations, while usually claiming merely to clarify Everett’s position, have substantially abandoned his goals. No proponent of MWI (as far as we are aware) has actually produced a complete set of axioms to define their physical theory.​

 

[Ken G]

But at an earlier time, my state could have been a superposition of 'in bedroom' and 'in kitchen'. As long as there is no information on whether I was in the kitchen or bedroom, then I could have been a superposition of both.

And what about the "you" that is perceiving themself to be in the kitchen or bedroom-- they are both you? So what defines "you", is everyone and everything "you"? Remember, the pure state that you are projected from also includes entanglements with me, so you and I are not even separable in your model of you.

 

[Dmitry67]

I don't think that will succeed very well as a model of me. It has two big problems:
1) it does not explain why I do not perceive such a tree, and
2) it does not account for how "I" diverge along different branches of the tree. What makes each branch the same "me", that there is someone in it with my genes? How much does my appearance or personality have to change before the branches are to be associated with different "me"s? And what about all the "yous", am I alone on my tree because the tree is what is actually me? This is the point I'm making, we must not imagine that MWI has a reasonable model of me without subjecting that model to scrutiny. Taking your model of me to the limit, all of reality is me. It's nice monism, but not really a very useful model of me.

#1 is very easy.
You remember the past, not the future.
You don't communicate with other 'branches'
But as different branches share the same past, and your memory is always from the past, you have an illusion that world is a single-history.

#2 - I can answer what someone would give an answer, what a consiousness is.

 

[Eqblaauw]

No according to the current theory there are many-worlds each with only one timeline (partly the same until they differentiate)
once again:
David Deutsch:
Dear Edo Blaauw -- > Dear David Deutsch,> > I have a few questions about the many-worlds interpretation: > Are there according to most of the many-world believers many universes exactly the same until they differentiate, Yes. > or do most of the 'mwi-ers' believe we are continuously being split? That is an older view, of which I currently know of no supporters among people working on the theory.
Don't try to make people crazier then they are

 

[Ken G]

#1 is very easy.
You remember the past, not the future.

You said that I was a tree, not a linear branch of a tree. So my memories cannot be me either, in your model of me. My memories tell a story of nonunitary evolution, that's the whole problem. Memory is no kind of fix.

#2 - I can answer what someone would give an answer, what a consiousness is.

I agree-- maybe when someone can say what a consciousness is, MWI will be able to claim to be a complete description and we'll know what to do with quantum suicide. But that's just a pipe dream at the moment-- the incompleteness of the MWI world view persists at present.

 

[RandallBSmith]

I believe most physicists would be surprised to find an experiment that could distinguish Many Worlds vs. Copenhagen. (Trying to come up with such might be a useful way to understand these models. E.g.: Would transitioning into a universe in which someone spontaneously exploded disprove anything?)

If two theories make identical experimental predictions, then choose one based on philosophical taste or cognitive economy.

I see these polls as showing how smart and thoughtful people like to think about quantum mechanics, not about which model is more or less correct. An interpretation can be useful for efficient education or as a way to think about some quantum phenomenon ... stuff like that. Personally, I find one of the big disappointments about life to be that apparently we don't get to know which of two equivalent theories is "true," only that they both work to describe phenomena. Kind of a bummer, but oh well.

 

[Ken G]

Dear David Deutsch,> > I have a few questions about the many-worlds interpretation: > Are there according to most of the many-world believers many universes exactly the same until they differentiate, Yes. > or do most of the 'mwi-ers' believe we are continuously being split? That is an older view, of which I currently know of no supporters among people working on the theory.
Don't try to make people crazier then they are

Considerable elaboration is required to understand what Deutsch is saying. First of all, what is the point of having many universes that are exactly the same? I see no difference between that situation and one universe, it's an incoherent distinction. Second of all, if the many worlds are not being continuously split, how is unitarity being preserved? Maybe "split" is the wrong word, and "subsectioned" would work better for him, but I think most people intend the same meaning in those two words. Finally, just what "theory" is being worked on here? I was not aware that MWI was a theory, is there a suggestion of some kind of new prediction being generated?

 

[Ken G]

Personally, I find one of the big disappointments about life to be that apparently we don't get to know which of two equivalent theories is "true," only that they both work to describe phenomena. Kind of a bummer, but oh well.

Then let me suggest a less disappointing perspective: think of each theory, and its range of interpretations also, as a set of filters for looking at the world, each one revealing some different "color" if you will. Framed like that, the myriad of possibilities is not disappointing, it is enlightening. How boring is a world described completely by one approach, that's just a form of tunnel vision.

 

[RandallBSmith]

Then let me suggest a less disappointing perspective: think of each theory, and its range of interpretations also, as a set of filters for looking at the world, each one revealing some different "color" if you will. Framed like that, the myriad of possibilities is not disappointing, it is enlightening. How boring is a world described completely by one approach, that's just a form of tunnel vision.

Cute ... a nice "glass half full" way to look at our life with mystery.

 

[Dmitry67]

No according to the current theory there are many-worlds each with only one timeline (partly the same until they differentiate)
once again:
David Deutsch:
Dear Edo Blaauw -- > Dear David Deutsch,> > I have a few questions about the many-worlds interpretation: > Are there according to most of the many-world believers many universes exactly the same until they differentiate, Yes. > or do most of the 'mwi-ers' believe we are continuously being split? That is an older view, of which I currently know of no supporters among people working on the theory.
Don't try to make people crazier then they are

Before discussing it, you need to define what 'splitting' is exactly.
You will see that the definition of 'splitting' is fuzzy. The same as when 1 road has a fork, you can't locate a place where it splits up to 1 micrometer.

 

[Dmitry67]

You said that I was a tree, not a linear branch of a tree. So my memories cannot be me either, in your model of me. My memories tell a story of nonunitary evolution, that's the whole problem. Memory is no kind of fix.

Again you mix 'you' as a line, located in a branch, with a full collection of you (tree).
I suggest to prefix all objects with F (frog perspective, or a line) and with B (birds perspective, or a tree).

My-F memory-F tells a story of nonunitary evolution (F)
My-B (full collection of all) memories-B cover all possibilities and evolution-B is unitary
You-F don't have an access to memories-B
You-F feel always as You-F, never as you-B

 

[Hurkyl]

I just don't see how you can claim that MWI has any way to account for the particular outcome of an observation.

The important question is "what determined the outcome that I perceived?" ... what actually happens determines what I perceive,

I don't account for it. The question isn't important. The relevant point is that I consistently perceive a single outcome. One doesn't need to take such drastic action and insist on definiteness.Let's play a game. It's a classical universe, but with four-valued logic: <T,T>, <T,F>, <F,T>, and <F,F>. We are talking to an observer in the universe -- but we have the advantage of being external to the universe so that we can get the bird's eye view and see the four truth values.

To go along with the four-valued logic, everything gets treated the same way -- numbers, particles, observers, et cetera.

Your goal is to make the observer aware (with empirical observation!) that his perceptions are not determined.

In his responses, I will write anything definite normally, and anything indefinite with the angle brackets as I did with the truth values.I'll get started.

Me: Hi there!
Experimenter: Oh hi!
Me: What are you doing?
Experimenter: I'm flipping a coin.
Me: Oh, what was the result?
Experimenter: It came up <heads, tails>.
Me: Oh, so you didn't see a definite result, huh?
Experimenter: What do you mean? My result is clearly definite.
Me: Well, did you see heads?
Experimenter: <yes, no>.
Me: There you go!
Experimenter: What are you talking about? I <definitely, definitely did not> see heads.
Me: Let's try this again. Did you see heads?
Experimenter: <yes, no>.
Me: Did you see tails?
Experimenter: <no, yes>.
Me: So, did you see heads or did you see tails?
Experimenter: Yes.
Me: Har har, you're a comedian.
Experimenter: But in particular, I saw <heads, tails>.
Me: Well, flip it a few more times.
Experimenter: ...
Me: What are the results?
Experimenter: I got <53,51> heads, out of <101,100> flips in all.
Me: Er... 100 flips?
Experimenter: <No 101 flips, Yes>
Me: Er... can you make it 101 flips?
Experimenter: <I already did, ... Okay, 52 heads>​

 

[Eqblaauw]

Considerable elaboration is required to understand what Deutsch is saying. First of all, what is the point of having many universes that are exactly the same? I see no difference between that situation and one universe, it's an incoherent distinction. Second of all, if the many worlds are not being continuously split, how is unitarity being preserved? Maybe "split" is the wrong word, and "subsectioned" would work better for him, but I think most people intend the same meaning in those two words. Finally, just what "theory" is being worked on here? I was not aware that MWI was a theory, is there a suggestion of some kind of new prediction being generated?

what he says is very clear. If and how it works i don't know. David deutsch always said that mwi was actually a theory. I think when deutch says no-one who is an expert in this theory believes the splitting, but in many universes exactly the same until they differentiate that counts for something. There is also a dutch documentary about deutsch where he makes this claim. If you are really interested I can post the url.
To the post that says as a response to my post 'There is no actual hard-splitting' i would say no: there is no splitting

 

[Samshorn]

Considerable elaboration is required to understand what Deutsch is saying. First of all, what is the point of having many universes that are exactly the same? I see no difference between that situation and one universe, it's an incoherent distinction. Second of all, if the many worlds are not being continuously split, how is unitarity being preserved? Maybe "split" is the wrong word, and "subsectioned" would work better for him, but I think most people intend the same meaning in those two words.

I'm not an expert on what Deutsch thinks, but I believe he prefers the word "partitioned". He imagines a fixed (albeit possibly infinite) number of worlds, none being created or destroyed, and at any given time they are partitioned into equivalence classess of identical worlds, but these equivalence classes get re-shuffled, so two worlds that were equivalent before are not any more, and two worlds that used to be distinct can merge and become equivalent. So he imagines a constant re-partitioning of these worlds into these equivalence classes.

Finally, just what "theory" is being worked on here? I was not aware that MWI was a theory, is there a suggestion of some kind of new prediction being generated?

Yes, Deutsch argues (or at least he has in the past, not sure what he says today) that in fact a sufficiently sensitive observer actually WOULD perceive a superposition of states, because the worlds merge back together sometimes. He has suggested that this could be tested by some kind of artificial intelligence machine that is more sensitive to subtle memories than humans are. I guess there is some irony in the fact that a promiment proponent of MWI essentially agrees with all the past critics of MWI who always argued that an observer should see a superposition of states. What then becomes of all the iron-clad arguments of the other MWI proponents who claim that only an idiot would think that MWI implies the possible perception of superpositions?

 

[Eqblaauw]

Before discussing it, you need to define what 'splitting' is exactly.
You will see that the definition of 'splitting' is fuzzy. The same as when 1 road has a fork, you can't locate a place where it splits up to 1 micrometer.

well david deutsch agrees with many universes already there exactly the same until they differentiate so it's needless to think of what splitting may mean, he just doesn't agree with it, my dear semantic philosopher

 

[Hurkyl]

No. String theory is completely compatible with CI, for all the same reasons that quantum mechanics was.

I don't see how. CI's compatibility with QM depends on you not being part of the system, so that it makes sense to place a cut between it and the classical you.

CI has all the same unitary evolution in the wave function as MWI!

... but only between collapses. I do sometimes like to make the point that even a staunch CI still ought to learn some MWI to understand how the wave-function behaves between collapses.

The sole difference is that CI does not interpret the thing that is evolving unitarily as the reality itself,

If it looks like interpreting the thing as reality itself and it quacks like interpreting the thing as reality itself, then as far as I'm concerned it's interpreting the thing as reality itself.

As far as I can tell, nothing is observationally different between a person who says "Oh, the wave-function is a mathematical object that contains all of the information about reality, and we update that information via unitary evolution or sometimes collapse" and "Oh, the wave-function is a mathematical object that corresponds to a real entity, and the time evolution of that real entity agrees with unitary evolution or sometimes collapse" except for the particular choice of words they used.
And, for the record, information updated sometimes by unitary evolution and sometimes by collapse is somewhat more ad-hoc and unsatisfying than information that is updated consistently by unitary evolution. From this point of view, the reason CI needs collapse is because it's not asking the right questions.

empiricist, the meaning of a theory is a kind of mental game, like chess

No, that's syntax.

outcome is found to be successful at predicting what is the actual reality-- the outcome of experiment.

But this is mapping elements of the theory to elements of reality -- an interpretation, in exactly the same sense the rationalist is using it.

Empiricism and rationalism use two completely different approaches to that "meaning"

No, they use different approaches to gaining knowledge.

For an empiricist, knowledge can be derived from observation and experiment.
For a rationalist, knowledge can be derived through logic.

A scientist must be both an empiricist and a rationalist.

Can all observers agree on the different outcomes of the observations that appear in the many worlds?

Er, yes, this is fairly clear. This has never been a problem. I'll continue my game with a demonstration.