# Different MWinterpretations

1. Mar 6, 2008

### confusedashell

MWI keep haunting me like a ghost and I cant seem to wrap my head around it.
It seems everyone claims different versions of this absurd interpretation.
Today I just learned that Deutsch doesn't support splitting, but rather preexisting universes and we exist in one and we never split from this superdetermined "cosmis strip"?
Other's just support the math but claims the universes arent really real, like ours.
Then we got the ones who say we split all the time...

How can there be all this different interpretations of something that isn't even supported by any evidence yet/ever?

2. Mar 7, 2008

### JesseM

Where did you hear this? It was my understanding that the MWI does involve splitting--see here for example.
None of the interpretations of QM is supported by any evidence, nor could they ever be, since "interpretations" by definition all make the same physical predictions (though I suppose we could imagine that someday someone will come up with a new theory of physics that contains testable MWI-like elements or Bohm-like elements or transactional interpretation-like elements).

3. Mar 7, 2008

### Demystifier

My impression is also that the adherents of the MW interpretation do not really agree what exactly this interpretation says. It is perhaps the vaguest interpretation of QM. But this is not necessarily its drawback; it can also be its virtue. Namely, the vagueness can be exploited to further develop the current form(s) of this interpretation, possibly in a form that could be acceptable to a wider physicist community.

4. Mar 7, 2008

### Fredrik

Staff Emeritus
I don't think the argument on that page is very convincing. I'm not even sure what they're saying. For example, what's this supposed to mean?

Code (Text):

If you try to treat the worlds as pre-existing and separate
then the maths and probabilistic behaviour all comes out wrong.
Also the differentiation theory isn't deterministic, in contradiction
to the wave equations which are deterministic, since many-minds
says that:

AAAAAAAAAAAAAAABBBBBBBBBBBBBBB         --------------> time
(Worlds differentiate)
AAAAAAAAAAAAAAACCCCCCCCCCCCCCC

occurs, rather than:
BBBBBBBBBBBBBBB
B
AAAAAAAAAAAAAA                         (Worlds split)
C
CCCCCCCCCCCCCCC

according to many-worlds.

To me, this looks like they're saying that if a system is in the state $|\psi\rangle=\frac{1}{\sqrt{2}}(|0\rangle+|1\rangle )$, then the change of the density matrix when a measurement is performed must be

$$|\psi\rangle\langle\psi| \rightarrow \frac{1}{2}|0\rangle\langle 0| + \frac{1}{2}|1\rangle\langle 1|$$​

and can't be

$$\frac{1}{2}|\psi\rangle\langle\psi| + \frac{1}{2}|\psi\rangle\langle\psi| \rightarrow \frac{1}{2}|0\rangle\langle 0| + \frac{1}{2}|1\rangle\langle 1|$$​

That obviously doesn't make sense, so they probably meant something else.

Also, their claim that if worlds don't split it contradicts the wave equation seems very strange to me. It seems to me that what they're describing as "worlds split" is precisely what I would describe as "worlds don't split".

The way I see it, there are two classes of interpretations of QM. You either think the state vector is a complete and objective representation of all the properties of the physical system and changes with time according to the Schrödinger equation, or you don't. If you do, you're an adherent of some version of the many-worlds interpretation, whether you realize it or not.

I'm not going to elaborate on what I think the MWI says in this thread, because I did it in this one, so if anyone is interested, they can check out posts #11 and #23 in that thread. (The link goes to the forum at the website of the James Randi Educational foundation. I have the same username there).

I am not an expert on the MWI, so I might be wrong about some of this. I hope someone will tell me if I am.

5. Mar 7, 2008

### Fra

Perhaps the missing "wrappability" is the observation you need to wrap it up

For me, the "interpretation" that I CAN wrap my head around, and that keeps me on track is the one I'd stick to. I figure wrappability is relative. If the idea is wrong, or it's something wrong with my head simply doesn't matter to me, because it makes no difference as I'm stuck with this head.

/Fredrik

6. Mar 7, 2008

### vanesch

Staff Emeritus
I think this is a healthy attitude, after all, interpretations are nothing else but "pictures" or "stories" that we like to tell to give some kind of meaning to the formalisms we use. So the main idea is to get a kind of more or less coherent story that we can relate to that helps us get a feeling for what the formalism tells us.

Personally, as many know, I prefer indeed MWI, because in my opinion, it sticks closest to the formalism (and as such, helps to understand most easily the "strangeness" of the results), but if you feel better with another one, be my guest.

7. Mar 7, 2008

### vanesch

Staff Emeritus
I think there is a part on which all MWI-ers agree:

* the wavefunction (or, relativistically, the unitary structure) describes a genuine objective ontological physical situation

* an observer is described by an observer body state, which appears in one of the terms of this wavefunction, and is hence only aware of the states of other things in this term, as these are the only ones that can have modified/evolved/... this particular body state.

As such, given that there will be several terms in the wavefunction which contain variants of that bodystate, we can say that to each of these terms corresponds an "experienced world", and these worlds will, thanks to decoherence, never have any influence anymore on one another.

Where MWI-ers differ, is on points like:

* the "mind-brain" connection between bodystates and "subjective experiences" (does this play an essential role, or are these superfluous concepts which have nothing to do with physics here?), things like solipsism or other philosophical questions.
* the mechanism of the emergence of probabilities, and the related issue of what exactly determines the "experienced" bodystates. Should we "count" states, or do the probabilities come in "by axiom" ? Is this problem solved by arguments like Dewitt's or Deutsch's, or is there a difficulty ?

On the other hand, words like "splits of worlds" and so on are not precise, and can differ from one person to another, and hence, they can give rise to different stances on the issues.

8. Mar 9, 2008

### Fra

I can sometimes appreciate elements of interpretations other than mine as well. And it can be interesting to hear how other people reason. I think my interpretation doesn't fit into any of the big ones, like MWI. But for comparasion here is how I personally think differentely on vanesh listed points.

Here I disagree. And in this sense my "interpretation" really does have the naive goal of improving QM by acknowledging what doesn't make sense. I don't think unitarity is fundamental in the traditional sense. My interpretation is that "unitarity" is emergent, and that processes that are in an abstract sense "close to equilibrium", will be unitary. This yields to straight contradictions with current experiments, because they can be explained to be effectively unitary. But the possibility of non-unitary processes also is a potential for expansion. This goes hand in hand, with the critics against the deterministic evolution of probabilities.

That events sum up to 100% follows from definition of probability theory. But that's not the question since this is physics. The question is how the axioms of probability map to the physics, and how this formalism deform during possible physical interactions. That's not a mathematical question at all.

I'm not sure exactly what MWI:s mean here, but if I understand this right, I think this is important. I'm not spritual in any way, but I think that "subjective expectations" is exactly the type of reasoning that might restore the consistency when I'm at the same time talk about non-unitarity. The point would be that the unitarity is relative too.

I guess many would call me a solipsist, but most criticts on this seem to mix this up with spiritual stuff and consciousness. I see a possibility that we may come up with a consistent formalism for this. Subjective information, and subjective probabilities and expectations.

These are IMO also a very important questions. By no means do I consider this to be irrelevant to the progress. IMO, this would make up an importany key to the interpretations.

/Fredrik

9. Mar 9, 2008

### reilly

If I'm not mistaken, unitarity can always be imposed on a QM system by embedding the system in a higher dimensional space, adding more particles -- usually done with the concept of an optical potential. Newton's 3rd law says this embedding is always possible. Further, the optical potential idea is widely used, for example:in all the various averaging techniques when dealing with reservoirs in stat. mech. It's been substantially used in nuclear physics, etc. Whether it's a good thing to do so is another matter.

Note that deterministic determination of probabilities is an old,old idea. The odds of drawing a particular hand in poker have been constant for over a century -- deterministic indeed. Like Markoff chains and so forth.

What are the clues, if any, that say reassessing the application of probability to the quantum world is worth doing? If so, how do you even start; what's the problem at hand?

By the way, I applaud your pushing the boundaries in a way that's civil and clear.
Regards,
Reilly

10. Mar 9, 2008

### Fra

If we talk informally, I think you are right here (and even if there is a way that you're formally wrong, I still get your point, which is enough :) But the problem I have with this, is that in general this added structure needed to recover unitarity is not unique as far as I can see.

I think it definitely has a place but often it's done in a way that is ambigous, and the problem is that it's ignored. I would like to elevate the PROCESS of inducing these higher dimensions into the physical interactions.

Clearly it's easier to make up higher dimensionsn and then reduce the information. Than to show how new dimensions can emerge from deviations of the lower ones.

Edit: Another major issue with this is also that it inflates the apparent degrees of freedom in the models. IE. the model has more degrees of freedom than the model it's supposed to fit. I even think that the degrees of freedom is the constraint that should prevent arbitrary new structures. It effectively means an overhead. An example of this is COULD BE the space time continuum. The question is for example if the contiuum has any physical correspondence.

Maybe I could reflect on this later, but... I assure you I have asked myself this question. and to me there are many subjective observations that motivates this effort. One of them is also, to ask yourself, what is the motivation for NOT reassessing old ideas on a regular basis? What is the estimated COST for walking in the wrong direction for X years? Here is also a risk assessment to make. I can not, at this point, motivate others. Maybe when I have come a little longer.

/Fredrik

Last edited: Mar 9, 2008
11. Mar 9, 2008

### Fra

The real reason why I think this is that I think that the observers understanding itself must be directly or indirectly encoded in the observers microstructure itself - where else would it be encoded?

Consider the extreme example, that the model is more complex than the system it's trying to describe. That wouldn't make sense. This is why my personal interpretation is that the information capacity of the observer put constrains on the modelling itself - as done by the observer.

I think when we humans make models, we can often "afford" to have a huge overhead, but if we consider a subatomic particle, I can't see how they can afford this luxury. I like to think that the physical strucutures selected by natural selection does have something to say about the design of optimum models with no overhead.

If we increase the complexity to retain unitariry, the complexity increases. But what happens when our (the observers) complexity is reached? I mean, supposed I can only count to 1e100? When thta limit is reached, there is not place to expand into, I have to start to remodel and I personally have hard to se how this remodelling in the general case can be unitary - general unitarity to me, implies that the observer, who is also the physical basis for modelling, has no information capacity limit. This make no sense to me.

/Fredrik

12. Mar 10, 2008

### vanesch

Staff Emeritus
Ah, I should add 2 things. The first one is that unitarity is not discussable in terms of "interpretation". In other words: "finding an interpretation for quantum mechanics" is an exercise that STARTS with, as a GIVEN, the axioms of quantum mechanics, of which unitarity is part. You can modify this if you want to, but then you're not *interpreting* the formalism of quantum mechanics, you're inventing a new one. One should distinguish "interpreting quantum mechanics" and "this is how I think the world is like". They are two different exercises. In "interpreting quantum mechanics" the exercise is to make up a story AROUND the formalism of quantum mechanics, independent of whether this has anything to do with how you really think the world might be. At least, to me. It is my principal motivation for adhering to MWI, because it respects entirely the mathematical formalism. So MWI, to me, sounds like the story that fits best with it. Totally independent of whether the real world is like that!

But it was not that what I meant with "unitary structure". The difficulty with saying that it is the "wavefunction" which is ontologically real, in a relativistic setting, is that the wavefunction is dependent on the frame of reference. If we want something relativistically "ontological" then you have not to take a "specific wavefunction", but rather "the entire structure of all the transformed wavefunctions in all possible reference frames, and their transformations in between them". This is what I call the "unitary structure". These transformations are BTW unitary transformations, and the time-translations are nothing else but the unitary time evolutions as seen in a specific frame.

13. Mar 10, 2008

### vanesch

Staff Emeritus
There's a difference between a kind of representation of a model by, say, a binary memory or something, and the abstract mathematical concept.

Nobody (I presume) has any difficulty considering Euclidean space as a model of "physical space", in, say, Newtonian mechanics, nevertheless, each mathematical point of it represents an infinite amount of information (to specify a point in an Euclidean space requires 3 real numbers, and a real number is already not "specifiable" with a finite amount of information).

Should this stop us from thinking of space (in a Newtonian setting) as an Euclidean structure ?

14. Mar 10, 2008

### Fra

Yes you have a good point here. But I happen to think that the choice of interpretation of a given theory is an important part for improving the theory. With this I mean that interpretations may be part of the scientific process, at least for an individual scientist so to speak.

If I were to make an interpretation without any ambition of improving anything, I'd prefer the shut up and calculate one. Ie to treat the model as a calculation scheme, you follow a receipe, put numbers in, and get numbers out. But to me that's engineering or applied science, and personally my take on physics is via the scientific method. I am interested in the dynamics of science. I probably have an odd angle to this, what do I know. But I like to hear how others think of this.

I understand your point and it's a good one. I just personally have hard to see a crystal clear distincion and I try to discuss the interpretations in a creative manner. My motivation for discussing any of this in the first place is that I wish to improve my own understanding of the world. I didn't mean to pick on MWI as such, and my critics is possibly beyond interpretation only like you say. But I can't help it.

/Fredrik

15. Mar 10, 2008

### reilly

Fra -- Are there differences enough between CM and QM to claim that Newton's Laws escape your Applied Science classification? Are they not the basis for computational recipes?

Regards,
Reilly

16. Mar 10, 2008

### Fra

Reilly, I'm not sure I understood this question?

I think, today, the application of newtons laws and classical physics to engineering and is "applied science" or engineering to me. However, to discovery or the "invention" of newtons law's, back in the days when they were invented, was not necessarily applied science.

OTOH, one if one describes the scientific method in terms of recepices, then I suppose even the scientific progress can be considered to be applied science. An application of the "theory of scientific method". I hold the opinon that even the scientific method may evolve. It constitutes the fundamental frameworks of how the general scientific thinking tends to work. What the "typical" reasonings goes like. It seems CM -> QM was a revolution in that sense. But I think we still await another step like this.

I see no clear cut separation between knowledge and method to acquire knowledge, because the knowhow of howto acquire knowledge is actually knowledge itself. Blurring the ontological and epistemological perspectives.

But I have no idea of that was what you questioned? I've have a feeling it wasn't. I'm sorry, maybe you can elaborate the question?

/Fredrik

17. Mar 10, 2008

### Fra

Anyway my intention wasn't to make any distinctions between science and engineering. That's not really interesting. I guess my focus was on the dynamics of knowledge, not knowledge itself. To study the principles of learning and adpation, and their impact of physical interactions, and look for their commong denominators. There is a lot of human intuition to feed into this. As kids, we don't just learn stuff. We learn howto learn. So what is REALLY the distinction between states and dynamics? could the pattern of dynamics be considered a state too? it seems so.

/Fredrik

18. Mar 11, 2008

### Fra

Yes, but I still have the opinion, interpretation or idea or however we should label it :) that even mathematical concepts "live somewhere". I consider mathematics to a human invention - or discovery. I personally see no distinct measurable difference between discovery and invention. For the body of mathematics to have any meaning, we must have a mathematician reading it, to understand it.

Similarly I really do think that one of the missing elements in many current models is the constraint on the models themselves that I THINK is imposed by the observer.

Even in terms of modelling, I think this is done by an observer too. A simple observer, simply wont be able to wrap his head around an overly complex model. And in that case, from his POV, it's not a good model.

OK, so what possible implications could this idea? The first one is the one of renormalizations. IMO, there might be a deep physical interpretation behind the complexity scaling that might correspond to the information capacity of an observer. And this scaling can not possibly in the general case be information preserving. I have a feeling this general scheme is not yet exploited to it's full power. I think this should be integrated at a more fundamental level.

The other potential I see is that of linking energy or mass with information capacity. This means that an infinite information density of space, might suggest an infinite energy density of empty space. I don't have the solution but I personally think think these issues are related. It does make sense to me.

Another thing I'm working on a little bit albeit slowly, is to understand how the superposition principle can be a result of a selective process under an information capcity constraint. The normal decoherence expositions, does not reveal IMO the full insight. I think this may resolve the confusion on classical vs QM statistics.

I also see a connection between the QM action, renormalization and gravity.

I personally, think that all these things are related and it's probably easiser to solve them at once, rather than one at a time. I could be wrong, but I have enough confidence to believe in this no matter if it seem naive.

It's not a problem to use newtons mechanics and his view of things. It sure makes things easier, althought slightly less correct. But that is insignificant in everyday life.

I do have problems to pictures to see euclidian space as *physical*. And you're right that the real numbers themselves seem non-physical in this sense. This is why I am trying to see howto build states from combinatorics. Which will recover the continuum in the infinite capacity limit, or at a fixed information capacity limit at a given confidence level. This could mean that a finite observer, may be unable to distinguish between a continuum of states and discrete state spectrum.

/Fredrik

19. Mar 11, 2008

### Fra

So MAYBE the question of wether the worlds in discrete or not at the planck scale has no answer, perhaps we could understand WHY it could be both ways, and what the relation is. Maybe there's a new duality between discreteness and continuum to be found? :)

/Fredrik

20. Mar 12, 2008

### vanesch

Staff Emeritus
Ah ! The Pythagorean dream