What are the Different Interpretations of MWI and How Do They Differ?

[confusedashell]

MWI keep haunting me like a ghost and I can't 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?

 

[JesseM]

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"?

Where did you hear this? It was my understanding that the MWI does involve splitting--see here for example.

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

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).

 

[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.

 

[Fredrik]

Where did you hear this? It was my understanding that the MWI does involve splitting--see here for example.

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?

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.

 

[Fra]

MWI keep haunting me like a ghost and I can't seem to wrap my head around it.

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

 

[vanesch]

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.

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.

 

[vanesch]

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.

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.

 

[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.

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

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.

* 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.

* 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.

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.

IMO I think future progress will say something about this.

* 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 ?

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

 

[reilly]

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 [/B]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.

/Fredrik

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

 

[Fra]

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

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.

Whether it's a good thing to do so is another matter.

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.

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?

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

 

[Fra]

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.

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

 

[vanesch]

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.

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.

 

[vanesch]

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.

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 ?

 

[Fra]

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.

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

 

[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

 

[Fra]

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?

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

 

[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

 

[Fra]

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

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 won't 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.

T
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 ?

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

 

[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

 

[vanesch]

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.

Ah ! The Pythagorean dream

 

[Fra]

Mmm not sure about that. Of course real numbers are still in principle, physically realized as limiting cases or series of rational ones. That's how the real numbers are defined anyway. And though it may seem a bit farfetched, I think we are forced into number theoretic questions here. And I think it may have relevance to physics.

IMO, I start by considering the concept of "distinguishable states". The concept of "distinguishable state", is defined withing the realm of an observer, which renders the distinguishability concept fundamentally local or observer relative. Then I try to consider how to LABEL distinguishable states. Here numbers are nice. But IMO numbers are nothing but labels. They numbers themselves have not physical significance as such. We could equallly well use a, b, c or any other symbols and still be able to defined the "equivalent" of real numbers.

Well that's how twisted my head is. I just can't "accept QM" and make up a strict interpretation. It sort of makes little sense to me. As soon as I TRY that, I find myself trying to improve things.

/Fredrik

 

[Ken G]

Then I try to consider how to LABEL distinguishable states. Here numbers are nice.

That actually makes a lot of sense to me. I see real numbers as human conceptions (integers too, but less abstracted from reality), so I don't think "nature" has the least idea what a number is. It is a template we lay over nature, and then forget our role in doing so because it is so successful. If so, that means we should keep careful track of what we are doing, or we miss what nature is doing. For that reason, I think your "distinguishable states" idea is pure genius. One must be careful not to endow the observer with too much influence, however, for a loss of objectivity would be disastrous.

 

[Fra]

One must be careful not to endow the observer with too much influence, however, for a loss of objectivity would be disastrous.

I personally have no problems to emphasise the observer view. Because to me, that is THE physical view. But then, here people seem to disagree.

They way I personally see it, I don't "loose" the objectivity - we never really "had it" in the first place. I rather see the "effective" objectivity we are used to, and what I think our intuition tells us to trust, as emergent in the relations between subjective views.

Either you think the fundamental view is the objective class of relations between subjective views, and then the problem is the one of CHOOSING the view. This is the birds view of the frog world.

or you think that each subjective view is fundamental, and then the problem is to try to figure out if first of all there seems to exists any objective relations at all, and if so what they are. This is the frogs view of the bird, with the different that there is no a priori reason to assume that the bird exists. So this is to me, less deterministic than the first case BUT also less speculative, which is the main reason I like it.

To me it's easy to see what viewpoint that is more realistic. I'm into the second view where objectivity is emergent, not fundamental.

/Fredrik

 

[Fra]

for a loss of objectivity would be disastrous

Yes, that's like the first think one would suspect, and there is the challange. To understand exactly how we can formally loose objectivity, and still avoid "disaster"!

I think the disastrous situation would quickly destabilise, beucase how is the degree of disaster defined? It seems the observer itself limits how disastrous it can get.

/Fredrik

 

[vanesch]

I personally have no problems to emphasise the observer view. Because to me, that is THE physical view. But then, here people seem to disagree.

They way I personally see it, I don't "loose" the objectivity - we never really "had it" in the first place. I rather see the "effective" objectivity we are used to, and what I think our intuition tells us to trust, as emergent in the relations between subjective views.

Either you think the fundamental view is the objective class of relations between subjective views, and then the problem is the one of CHOOSING the view. This is the birds view of the frog world.

or you think that each subjective view is fundamental, and then the problem is to try to figure out if first of all there seems to exists any objective relations at all, and if so what they are. This is the frogs view of the bird, with the different that there is no a priori reason to assume that the bird exists. So this is to me, less deterministic than the first case BUT also less speculative, which is the main reason I like it.

To me it's easy to see what viewpoint that is more realistic. I'm into the second view where objectivity is emergent, not fundamental.

/Fredrik

I can't say I disagree

 

[Ken G]

I personally have no problems to emphasise the observer view. Because to me, that is THE physical view.

It depends on what one means by "view"-- the concept of what is "physical" and what is "objective" are closely related (though not necessarily what is "real", we agree there). For example, relativity allows for "reference frames", but it doesn't matter which observer or clock is put in which frame. That is how it maintains objectivity. To lose that would indeed be disastrous, because all of science is built on the principle of objectivity.

They way I personally see it, I don't "loose" the objectivity - we never really "had it" in the first place.

I must agree that objectivity is a construct of science that we don't actually "have", but science is all about idealization. I think to lose the idealization of objectivity would hamstring science. But that doesn't mean the observer "view" is not important, I think your idea that one must look at what an observer can distinguish is crucial. What I mean is, if there are things that I can distinguish that you cannot, or vice versa, we cannot try to build that into the description of reality or we cannot distinguish science from philosophy.

I rather see the "effective" objectivity we are used to, and what I think our intuition tells us to trust, as emergent in the relations between subjective views.

I don't disagree as much as you may think-- indeed, I think what is objective is a subset of what is subjective, whereas many scientists seem to believe the converse. But I do recognize that the restriction to what is objective is the crucial "devil's bargain" of science-- it both limits us and empowers us. So a theory of distinguishability must take that into account, in my view.

To me it's easy to see what viewpoint that is more realistic. I'm into the second view where objectivity is emergent, not fundamental.

We actually agree there-- the issue is whether science comes prior to the emergence of objectivity, or as a result of that.

 

[Fra]

It depends on what one means by "view"-- the concept of what is "physical" and what is "objective" are closely related (though not necessarily what is "real", we agree there). For example, relativity allows for "reference frames", but it doesn't matter which observer or clock is put in which frame. That is how it maintains objectivity. To lose that would indeed be disastrous, because all of science is built on the principle of objectivity.

But to address your concern which I consider important indeed - the loss of objectivity I refer to, is not total. To acknowledge that we don't know for sure, does IMHO not mean we have no clue and we are lost! We can still guess. And often a very good "scientific guess", is FAPP almost as accurate as knowing for sure. Life is risky game :)

What I am in favour for, would imply that we replace the concept of objectivity with "subjective degrees of objectivity", by a similar token that determinism in classical physics is replaced by probability in QM.

As I think of it, objectivity is emergent when a group of subjective views are in agreement. The question then is how does a group of observers come into agreement? To me this is driven by communication which is just another word for interactions.

But I mean this in a quite radical way.

Consider how say a thermodynamical equilibration process looks like, say temperature equilibration. Two bodies meet, with different temperatures, then they will finally attain the same temperature, and reach an "agreement", and further responses internally in the bodies we could for example have phase transitions and so on. Of course traditional stat mech is not formualated in a relational manner. It usually refers to objective equiprobability hypothesis that arbitrarily singles out a microstructure. What I'm saying is that this microstructure should be seen as a dynamical object.

This is a bit like how I see this. Communication results in tuning of the transceivers. And observer can be seen as a transceiver. And there is a selection for the transceiver design - ie the transciever is a dynamical object.

In part the transciever design in my thinking encodes the local rules of reasoning, which i consider to be remotely associated to the action, beeing further responsible for transition "probabilities" etc.

I must agree that objectivity is a construct of science that we don't actually "have", but science is all about idealization. I think to lose the idealization of objectivity would hamstring science. But that doesn't mean the observer "view" is not important, I think your idea that one must look at what an observer can distinguish is crucial. What I mean is, if there are things that I can distinguish that you cannot, or vice versa, we cannot try to build that into the description of reality or we cannot distinguish science from philosophy.

Perhaps part of the the dynamics we experience can be explained just due to this? I am well aware that I see things differently than others. I have different data at hand than others, and vice versa. There is no contradiction at all in that. The contradiction is only when something assumes that everyone must always make perfectly consistent conclusions.

I'm not trying to cripple science. I am not worried about that. I realize that what I suggest just crippled the structures we rely on, but I don't think it's as bad as it seems. I'm not removing structures, I just want to replace assumptions with flexible DYNAMICAL objects.

I don't disagree as much as you may think-- indeed, I think what is objective is a subset of what is subjective, whereas many scientists seem to believe the converse. But I do recognize that the restriction to what is objective is the crucial "devil's bargain" of science-- it both limits us and empowers us. So a theory of distinguishability must take that into account, in my view.

I obviously don't have all the answers even to my own questions, but the idea I mentioned will address this issue. But maybe differently than you think. There will not be a solid objectivity like in classical physics - ie there is no objective manifold etc. Neithre do I think there is an objective hilbert space. etc. But there should come something even better.

We actually agree there-- the issue is whether science comes prior to the emergence of objectivity, or as a result of that.

This is an interesting question and now we do enter philosohpy of science. I picture that even the scientific method in a sense evolves. And I think that there are parallells between science and the laws of physics. What I mean with this is that the scientist themelsevs are of course just frogs in this game. This is one of my problems when reading stuff from many theoretical physicists. They talk about birds and frog views, like they were a big hawk seeing both the frogs and the birds without realising that they are just in one of the frogs.

I see the scientist as a frog, and transciever and the very rules of reasoning of "science" is IMO opinon encoded in the transceiver design. This is how something that is effectively false for one observer, can consistently be effectively true for someone else. The scientific society OTOH is more like a "collection of frogs" in communication :) So if one likes, one can consider this collection of frogs for "an observer", and then we do get an objectivity that at least is invariant within the local group of scientists, but still this local group does not exhaust the universe, and even if it did we would still have issue with the intercommunication within this now non-local group.

(With locality here, I really mean local information, not locality in the sense of space).

Edit: In this sense, information can induces something like distance measures. The distance between two states can loosely be related to the subjective probability that they are mixed up. And the measure defining the "subjective probability" is in my thinking encoded in the microstructure of the observer, at best as a "mirror image" of the outside world.

/Fredrik

 

[Fra]

I'm not trying to cripple science. I am not worried about that. I realize that what I suggest just crippled the structures we rely on, but I don't think it's as bad as it seems. I'm not removing structures, I just want to replace assumptions with flexible DYNAMICAL objects.

I suspect you'd be worried about stability. How do we know that the dynamical references are stable enough to not bring us chaos. We don't. But I think it's plausible that it does. And here I think the concept of intertia enters. I am still thinking about this but the intertia is what should make the dynamical references sufficiently stable, but still flexible.

So what IS intertia? I am thinking of it related to the measure of confidence of a state. In a simple way one can consider it as the observers interal "counting" of the distinguishabl counts in favour of this _distinguishable_ state. Note that two indistinguishable states are counted as the same. When the observer is exposed new contradicting information, there is a way to rate them. And the relative confidence of the current view relativ to the perturbation now imposes an intertian in the sense of resistance against change.

IMO this logic at least seem clear, consistent and plausible. But the realisation of this into a new formalism remains. So i have good confidence in this idea. The problem is to mature it and formalise it.

/Fredrik

 

[Ken G]

I'm not trying to cripple science. I am not worried about that. I realize that what I suggest just crippled the structures we rely on, but I don't think it's as bad as it seems. I'm not removing structures, I just want to replace assumptions with flexible DYNAMICAL objects.

I understand your goal, but in these general terms I'm not sure if you are helping science evolve, or doing something other than science. It is true that the science of the Greeks had some similar attributes to the modern counterpart, and some very different ones, but I don't view that as an "evolution" of science, I view it as a combination of what we now call science and something we might call "natural philosophy". The Greeks didn't distinguish those, and suffered a lot of scientific schizophrenia as a result. I view the breakthrough of Galileo as being the splitting of empirical science from natural philosophy, the idea that science doesn't need us to shepherd it, it needs us to ask the right questions and then get out of the way. So that's why you and I come from very different perspectives on "objectivity", even though the very fact that we both see it as an emergent property, rather than a fundamental property of reality, separates us from the popular view.

So in a nutshell, I'm hearing the advantages of including what we can actually distinguish as an integral part of how we objectively interact with reality, rather than as a minor complication. But it seems to me it has to be done in a way that preserves the tradeoff between what is objectively testable and what is self-actualizable. I think we hamstring our science by straying too far away from what we can test objectively (i.e., can agree on), even as we can also hamstring our science by clinging too closely to a philosophical reliance on positivism (I marvel at the convulsions some people will go through to marry quantum mechanics to classical pictures of reality, just as they try to marry particles and waves, instead of simply recognizing that each plays a different role in our process of conceptualizing reality).

This is one of my problems when reading stuff from many theoretical physicists. They talk about birds and frog views, like they were a big hawk seeing both the frogs and the birds without realising that they are just in one of the frogs.

My favorite is the "theory of everything". I realize those words may not be intended to be heard literally, but they certainly sound that way. Have we learned nothing from the long history of this endeavor?

So if one likes, one can consider this collection of frogs for "an observer", and then we do get an objectivity that at least is invariant within the local group of scientists, but still this local group does not exhaust the universe, and even if it did we would still have issue with the intercommunication within this now non-local group.

But this does not seem to be responsive to the amazing fact that even in relativity, where the descriptions of what is happening seem to be so totally diverse, the equations themselves keep their same form for all observers. That is a remarkable coincidence if reality is malleable to the observer's condition. Still, what I will agree with is already somewhat of a radical view-- that the art of science is the art of seeing what makes sense to us to see. In other words, science is merely a projection of reality onto what we are ready to understand of it. Still, that projection preserves a property of objectivity-- the nontrivial set that survives that projection is far more rich and profound than we had any business hoping. That fact is what I call "science".

And the measure defining the "subjective probability" is in my thinking encoded in the microstructure of the observer, at best as a "mirror image" of the outside world.

This sounds a lot like what I think of as the "objective projection" that is used as science's defining tool of perceptiveness. This approach gives us a facility for navigating the waters of philosophically murky waters like quantum mechanics-- why must a particle go through one slit or the other unless the physical setup supports some means to ask it that question?

 

[Fra]

But this does not seem to be responsive to the amazing fact that even in relativity, where the descriptions of what is happening seem to be so totally diverse, the equations themselves keep their same form for all observers. That is a remarkable coincidence if reality is malleable to the observer's condition.

They way I see that is that while it's not deterministically predictable, it is not entirely unexpected. The "concidence" is in my view, since the microstructures of the interacting group of observers in a certain sense has a common origin. They have evolved togther, and it seems highly implausible that there would be selection for inconsistencies so to speak. Inconsitencies would imply highly deforming interactions, that will correct them - a kind of selforganisation.

So in despited of that subjectivity seems to be the more fundamental view, the emergence of objectivity is plausible and hence expected! It's just that it's not certain, and not deterministic.

I think that this spirit should eventually translate into dynamics. But since dynamics of information is in my thinking not deterministic - like in QM. What would probably be more or less like plain QM is the _expected_ dynamics_. And IMO a possible, plausible reason for actuall expecting "emergent determinism" (as in QM) is simply because the opposite is more unlikely.

So I think we can infer the dynamics from this.

It seems we share sufficient views to at least be in partial agreement. What's beyond this is still in progress as far as I am concerned. These are ideas only.

The distinguishability concept, and definition of observer my means of internal, external and interfacing degrees of freedom is somewhere where I start. From that I associate the interfacing degrees of freedom with the capacity of the communication channel. This restricts the possible interactions. The internal degrees of freedom corresponds to information capacity of the observer. The external degrees of freedom is a misnomer because it's unknonwn. The only think that is known about the environment, is the result of the history of interactions, that has formed the microstructure of the internal structure. This is a dynamical structure. where the degrees of freedom are also dynamical.

The observers "knowhow" about it's environment is partly stored in the properties/design of it's very microstructure, and partly in the STATE of the microstructure. So here I make a crucial distinction that deviates from the normal stat mech way. The CHOICE of the microstructure itself! does contain information! This is subtle but IMO important. The microstructure evolving is I think, possibly natures way of solving the problem of optimizing their chances given the constraints at hand.

In classical mechanics, we innocently consider a microstructure and then try to consider the information encoded in the STATE of the microstructure. But rarely is it acknowledge that when we defined the microstructure, information are put in. I think in a proper approach these things must be explicitly accounted for to the degree possible.

/Fredrik

 

[Fra]

I understand your goal, but in these general terms I'm not sure if you are helping science evolve, or doing something other than science.

If you're unsure wether this will all lead to something constructive, then I'm unsure too, but that seems an unvoidable part of the game of life :) I don't think anyone knows, then it would all be trivial. But this is my best bet. It's my gain and my loss, I'm not speculating with other peoples resources.

/Fredrik

 

[Ken G]

Granted. And I wasn't saying "don't try", I was saying "by all means try-- but in my opinion, as you consider the role of how the observer thinks about what is happening affects how we describe what is happening, you will end up with a more powerful tool if you maintain the arbitrariness of the observer." In other words, "time flies when you're having fun", but a clock always gets it right and without that science would be anemic.