Quantum Mechanics without Measurement

[craigi]

He only discusses his interpretation.

As I have said he believes his interpretation is realistic, but if it really is that is an open issue.

I like CH, but its not my favorite because I find it a bit more complex than I think necessary, with frameworks and what not. I simply assume after decoherence the improper mixed state is a proper one - easy as far as I am concerned without this baggage of frameworks, histories, blah, blah, blah.

I am the wrong person to ask about if an interpretation is non counterfactual etc. Terms like that to me is philosophical verbosity. I can't even remember without looking it up exactly what it means.

My view is much simpler. QM is basically the most reasonable general probability model for physical systems that allows continuous transformations or equivalently entanglement. Its entanglement with the environment and measurement apparatus that leads to observations - properties exist because of that, and systems don't actually have properties apart from that. So, just prior to observation outcomes are actualized via dechoerence - but before that - blah. Is that counterfactual definite - maybe, maybe not - I will let others judge. As I said I am not into that sort of thing.

Thanks
Bill

CFD, is really dBB's realm. Particles with definite properties independent of measurement. Though in CH, measurement doesn't play a central role, it's not considered to be CFD. For most interpretations, this is one of the first things they throw out. dBB really wanted to retain that, so made concessions elsewhere.

 

[Jilang]

DrC has a useful page that effectively proves the mathematical impossibility of LHV – it just doesn't work! (i.e. unless 'someone' wants to refute mathematics as well...)

http://www.drchinese.com/David/Bell_Theorem_Easy_Math.htm

OK let's leave out the negative probabilities then. Working through the example in the link you could consider that each of the eight scenarios are equally likely for the particle up until the point it's measured. At that point scenarios (1) and (8) are wiped out by the process of measurement as they can never be measured with that result. The probability of measuring a coincidence would then be 6x.333/8. Which is 0.25.

 

[stevendaryl]

wow... just wow... if this is not refuting QM & Bell's theorem, then what is??

I don't know why you say that. He definitely is not refuting QM. The whole point is to give a more sound treatment of QM. He's definitely not refuting Bell' theorem. It's a theorem, after all. What he's doing is questioning the definition of "local realism". What should be the appropriate formal definition of an informal, intuitive idea is up for debate. In contrast, the experimental predictions of quantum mechanics and mathematical theorems are not up for debate (although their implications or meaning might be).

He's up against a whole world of professional and rigorous experiments, working flawlessly every time... what on Earth will this man say when Anton Zeilinger and Alain Aspect receive the Nobel Prize in Physics – the whole world is wrong and he is right, even if he can't prove it?
Gosh

He's not questioning any of that.

 

[craigi]

Apologies, if this is treading old ground, but earlier in this thread, some suggested that they found the idea of non-classical logic applied to quantum physics unsatisfactory. My question is that is this really something that is specific to CH?

I think we all notice something unsatisfactory when we first learn about it, but then we dismiss it without demanding an answer. If we consider a single spin half particle,

A: Sx = +1/2
B: Sx = -1/2
C: Sz = +1/2
D: Sz = -1/2

Classical logic would tell us that since,

C or D = 1

(A and C) or (A and D) = A

but we know that from the QM forumlation that the LHS is always false and that the RHS is sometimes true.

Does this not demonstrate that classical logic cannot be applied to QM without extra rules, independent of interpretation?

 

[stevendaryl]

I think that negative probabilities means that it is simply impossible to mimic this feature by any means of classical tools. LHV requires 'something' to be 'there' all the time, definitely. If the probability of 'something' to be 'there' is negative – it means it's not 'there', i.e. it's not definite.

Negative probability would not be "not there"--that would be "zero probability". The thing about negative probability is that since probabilities have to sum to one, if some probabilities are negative, then other probabilities have to add up to greater than one.

 

[stevendaryl]

Negative Probabilities

Here's a sketch of a way that negative probabilities can make some sense, which is sort of reminiscent of relativistic quantum mechanics.

Suppose you have a string running from left to right. Impose a coordinate system, with the x-axis going horizontally and the y-axis running vertically. If the shape of the string is random, then you could perhaps describe the situation by giving a probability distribution at each point of the x-axis: $P(x,y) dy =$ the probability that the string runs through some point in the interval $(x,y \pm dy)$. If the string is mostly horizontal, and the curves are not too wild, then you can expect that:

$\int P(x,y) dy = 1$

That is, if you pick a value for x, the string has to cross at some point, so the probability of it crossing must add up to one.

But now consider a wilder shape for the string, as shown in the picture below. As you can see, the string passes point $A$ once, but passes point $B$ 3 times, and passes point $C$ 5 times. So we can no longer describe the random string using a probability distribution of the form $P(x,y)$. However...

What is interesting about the pattern of crossings is that the number of crossings is odd (assuming that the far left-end of the string and far right-end of the string are fixed in place). That's because for every section where the string is running right-to-left, there must be a corresponding section where the string is running left-to-right. In the picture below, we color the left-to-right sections blue, and the right-to-left sections red. Then instead of a probability that adds up to 1, we have some kind of density function

$D(x,y) dy$

which can be either positive or negative. Positive density means that blue sections are more likely than red sections. Negative density means that red sections are more likely than blue sections. The constraint now is that

$\int D(x,y) dy = 1$

where $D(x,y)$ counts the excess of blue sections over red sections.

I've toyed with the idea of a local hidden variables theory where instead of probabilities there were densities, which could be either positive or negative. I couldn't find any way to make sense of that for the EPR experiment, though.

 

[atyy]

Apologies, if this is treading old ground, but earlier in this thread, some suggested that they found the idea of non-classical logic applied to quantum physics unsatisfactory. My question is that is this really something that is specific to CH?

I think we all notice something unsatisfactory when we first learn about it, but then we dismiss it without demanding an answer. If we consider a single spin half particle,

A: Sx = +1/2
B: Sx = -1/2
C: Sz = +1/2
D: Sz = -1/2

Classical logic would tell us that since,

C or D = 1

(A and C) or (A and D) = A

but we know that from the QM forumlation that the LHS is always false and that the RHS is sometimes true.

Does this not demonstrate that classical logic cannot be applied to QM without extra rules, independent of interpretation?

Can't one say that since Sx and Sz cannot be simultaneously measured, (A and C) doesn't exist, so classical logic is fine? ie. Can one say that as long as one knows how to make a classical/quantum cut, and what a measurement is, then QM satisfies the rules of common sense?

 

[craigi]

Can't one say that since Sx and Sz cannot be simultaneously measured, (A and C) doesn't exist, so classical logic is fine? ie. Can one say that as long as one knows how to make a classical/quantum cut, and what a measurement is, then QM satisfies the rules of common sense?

Sure, but that is pretty much how CH works. I think there other ways to forbid classical logic here too. TTere's no way around it.

 

[stevendaryl]

Apologies, if this is treading old ground, but earlier in this thread, some suggested that they found the idea of non-classical logic applied to quantum physics unsatisfactory. My question is that is this really something that is specific to CH?

I think we all notice something unsatisfactory when we first learn about it, but then we dismiss it without demanding an answer. If we consider a single spin half particle,

A: Sx = +1/2
B: Sx = -1/2
C: Sz = +1/2
D: Sz = -1/2

Classical logic would tell us that since,

C or D = 1

(A and C) or (A and D) = A

but we know that from the QM forumlation that the LHS is always false and that the RHS is sometimes true.

Does this not demonstrate that classical logic cannot be applied to QM without extra rules, independent of interpretation?

I think that many people would say that the meaningful statement is not "The particle's spin in the z-direction is +1/2" but "When I measured the particle's spin in the z-direction, the result was +1/2". If you restrict the meaningful statements to observations and measurements, then classical logic applies.

But to me, that's exactly the same sort of modification of logic that CH is doing. So I guess I agree with you.

 

[craigi]

I think that many people would say that the meaningful statement is not "The particle's spin in the z-direction is +1/2" but "When I measured the particle's spin in the z-direction, the result was +1/2". If you restrict the meaningful statements to observations and measurements, then classical logic applies.

But to me, that's exactly the same sort of modification of logic that CH is doing. So I guess I agree with you.

So this restricts this logic, by sacrifcing CFD and giving the observer a causal role, which is the CI. I would suggest that when you say many people would say this, it's because the CI became ingrained in the teaching of QM for so long. I think if it weren't for that, it would seem at least as solipsistic as any other interpretation.

 

[atyy]

So this restricts this logic, by sacrifcing CFD and giving the observer a causal role, which is the CI. I would suggest that when you say many people would say this, it's because the CI became ingrained in the teaching of QM for so long. I think if it weren't for that, it would seem at least as solipsistic as any other interpretation.

CI is not solipsistic. The main correction to CI needed is the error in von Neumann's proof that hidden variables are impossible. With that, it's possible to take CI under the assumption that QM is incomplete. The nice thing is that CI has a common sense reality, because of the classical/quantum cut. Since QM is not a complete theory, there is no problem with needing a subjective observer. So CI sits at the centre of all interpretive possibilities, eg. dBB takes QM to be incomplete, but CH and many-worlds take it to be complete and nonreal or real.

 

[Maui]

So this restricts this logic, by sacrifcing CFD and giving the observer a causal role, which is the CI. I would suggest that when you say many people would say this, it's because the CI became ingrained in the teaching of QM for so long. I think if it weren't for that, it would seem at least as solipsistic as any other interpretation.

That which we call 'reality' is always brain generated(projected) and we know from SR and QM that reality is not classical. From neuroscience we know that brains can generate false impressions and memories of things that were not there(relevant to out viewpoint as opposed to someone suffering from schizophrenia). Given that knowledge already plays a part in certain areas of qm and reality is always experienced as classical, the role of the brain in the perception of the so called outside world should at least be examined. I wouldn't be astonished if it holds big surpizes on the nature of the fundamental particles(I would actually be truly astonished if one day it turns out that classicality is fundamental).
Yes, I know, there is a 100% agreement between observers about observations, but the observers themselves are of the same quantum nature and must rise to classicality in the same way as rocks, chairs, atoms and particle trajectories do. Weird indeed, but these are experimentally solid facts.

 

[atyy]

Does CH really solve the measurement problem? If the observer is included in the framework, who chooses the single framework?

 

[craigi]

Does CH really solve the measurement problem? If the observer is included in the framework, who chooses the single framework?

As in MWI, there is no measurement problem to solve in CH, so yes it does. The measurement problem is primairily a CI issue, though later interpretations such as Von Neumann, fall foul of it too. I think our dBB'ers and Ensemblers probably don't see a measurement problem either. Demystifier? Bill? Can you confirm?

I think that in CH everything in the classical environment, including the observer is part of a single framework and other frameworks occur in sufficiently isolated quantum systems. My terminology might be a bit off, but if I'm right this is where decoherence plays a big role.

 

[atyy]

As in MWI, there is no measurement problem to solve in CH, so yes it does. The measurement problem is primairily a CI issue, though later interpretations such as Von Neumann, fall foul of it too. I think our dBB'ers and Ensemblers probably don't see a measurement problem either. Demystifier? Bill? Can you confirm?

I think that in CH everything in the classical environment, including the observer is part of a single framework and other frameworks occur in sufficiently isolated quantum systems. My terminology might be a bit off, but if I'm right this is where decoherence plays a big role.

I agree that MWI and dBB have no measurement problem. The ensemble interpretation is just CI in density matrix language, and has a measurement problem because it has a classical/quantum cut.

Does CH really not have a measurement problem? If CH requires a single framework to be chosen, then who chooses the framework? Or must the observer lie outside, and choose the framework? Or do all frameworks occur?

 

[craigi]

I agree that MWI and dBB have no measurement problem. The ensemble interpretation is just CI in density matrix language, and has a measurement problem because it has a classical/quantum cut.

Does CH really not have a measurement problem? If CH requires a single framework to be chosen, then who chooses the framework? Or must the observer lie outside, and choose the framework? Or do all frameworks occur?

I might be wrong here, but as I understand it, in the EI, since the wave function describes an ensemble rather than an individual system, there is no collapse, so there is no measurement problem.

Again, I could be wrong here, but in the CH, I think this single framework is the entire classical world containing all obeservers. I don't have a reference for this, so I could be way off the mark.

 

[atyy]

I might be wrong here, but as I understand it, in the EI, since the wave function describes an ensemble rather than an individual system, there is no collapse, so there is no measurement problem.

Again, I could be wrong here, but in the CH, I think this single framework is the entire classical world containing all obeservers. I don't have a reference for this, so I could be way off the mark.

EI has collapse. This is in the assumption that an improper mixture can be treated as a proper mixture.

My question in CH is there is more than one possible framework, so if all observers are in all frameworks, then who chooses which framework is "reality"? In other words, the single framework rule requires a choice, whose choice?

 

[forcefield]

Let $P(i, j)$ be the probability that Alice measures spin-up along axis $i$ and Bob measures spin-up along axis $j$.

Why do you need Alice and Bob ? Why don't you just describe it in terms of measurement settings and measurement results ?

 

[craigi]

EI has collapse. This is in the assumption that an improper mixture can be treated as a proper mixture.

My question in CH is there is more than one possible framework, so if all observers are in all frameworks, then who chooses which framework is "reality"? In other words, the single framework rule requires a choice, whose choice?

I'm far from an expert on this. I haven't read Griffiths' book and I'm really just extrapolating from what I know here, so please someone tell me if I'm wrong, but I'm going to try to explain this, as I understand it.

Decoherence causes the entire classical world, that is you and everyone on the planet to be connected in a non-quantum manner, though constant exchange of particles and virtual particles in thermodynamically irreversible processes.

When I say planet, I'm being conservative here, because I don't understand to what extent dechorence is taking place on a cosmological level. Certainly our solar system is going to be undergoing decoherence, due to the stream of particles from the sun. Perhaps the lesser intensity of particle exchange going on between stars is a candidate for a less significant form of decoherence and solar systems can be sufficiently isolated to be considered separate quantum systems. Outer space is very cold, but it's not absolute zero. I just don't know if it's cold enough to prevent decoherence on a cosmological scale, but let's stick to planet, where we can be confident that all histories on the macroscopic scale are consistent.

So you asked about observers. In CH, no observer has a privilged role. Humans, animals, plants, computers, rocks and so on, are all just entities in the classical world, all kept consistent with each other. So there is no special role for an "observer", and no "choice" to be made. If there is a privilged role for anything, it's simply the classical world, which is so active that all quantum behaviour dissipates almost instantanously.

Within that, we have systems that become isolated from the classical world, or environment, on a quantum mechanical level. This is where other "histories" are. Typically, these systems are small or weakly interacting, such as the famous QM experiments. Some of which require careful isolation from their environment, often thermal isolation. If this isolation is broken, we lose the quantum behaviour of the system as it decoheres with the classical world and its histories become consistent with it.

This is largely just a description of decoherence, but I think it's necessary in order to explain where there separate sets of histories lie and to answer your question about the role of observers.

Arguably, an aspect of the measurement problem is still relevant, in that there could be a yet unknown mechanism selecting a pure state from probabilities of a mixed state, that would seem more plausible than the process being intrinsically stochastic in nature.

 

[stevendaryl]

Why do you need Alice and Bob ? Why don't you just describe it in terms of measurement settings and measurement results ?

That's what it is: $P(i,j)$ is the probability that one particle will have spin-up along axis $i$ and the other particle will have spin-up along axis $j$. "Alice" and "Bob" are just traditional.

 

[atyy]

I'm far from an expert on this. I haven't read Griffiths' book and I'm really just extrapolating from what I know here, so please someone tell me if I'm wrong, but I'm going to try to explain this, as I understand it.

Decoherence causes the entire classical world, that is you and everyone on the planet to be connected in a non-quantum manner, though constant exchange of particles and virtual particles in thermodynamically irreversible processes.

When I say planet, I'm being conservative here, because I don't understand to what extent dechorence is taking place on a cosmological level. Certainly our solar system is going to be undergoing decoherence, due to the stream of particles from the sun. Perhaps the lesser intensity of particle exchange going on between stars is a candidate for a less significant form of decoherence and solar systems can be sufficiently isolated to be considered separate quantum systems. Outer space is very cold, but it's not absolute zero. I just don't know if it's cold enough to prevent decoherence on a cosmological scale, but let's stick to planet, where we can be confident that all histories on the macroscopic scale are consistent.

So you asked about observers. In CH, no observer has a privilged role. Humans, animals, plants, computers, rocks and so on, are all just entities in the classical world, all kept consistent with each other. So there is no special role for an "observer", and no "choice" to be made. If there is a privilged role for anything, it's simply the classical world, which is so active that all quantum behaviour dissipates almost instantanously.

Within that, we have systems that become isolated from the classical world, or environment, on a quantum mechanical level. This is where other "histories" are. Typically, these systems are small or weakly interacting, such as the famous QM experiments. Some of which require careful isolation from their environment, often thermal isolation. If this isolation is broken, we lose the quantum behaviour of the system as it decoheres with the classical world and its histories become consistent with it.

This is largely just a description of decoherence, but I think it's necessary in order to explain where there separate sets of histories lie and to answer your question about the role of observers.

Arguably, an aspect of the measurement problem is still relevant, in that there could still be a yet unknown mechanism selecting a pure state from probabilities of a mixed state, rather than being intrinsically stochastic in nature.

But if the sun is needed then isn't there still a part of the universe that isn't quantum?

If the whole universe is quantum, then given that there are multiple frameworks (a framework is a family of consistent histories), and that a single framework must be chosen, who chooses that framework?

 

[Autochthon]

But if the sun is needed then isn't there still a part of the universe that isn't quantum?

If the whole universe is quantum, then given that there are multiple frameworks (a framework is a family of consistent histories), and that a single framework must be chosen, who chooses that framework?

This may be way off base but, Wouldn't the framework be defined by overlapping wavefunctions. ie. I make a measurement, the wavefunction of that measurement acts as a "filter" on family of histories?

 

[stevendaryl]

But if the sun is needed then isn't there still a part of the universe that isn't quantum?

If the whole universe is quantum, then given that there are multiple frameworks (a framework is a family of consistent histories), and that a single framework must be chosen, who chooses that framework?

To me, it seems like choosing a framework is purely pragmatic. You want to know the answers to specific questions, then you choose a framework for which those questions have answers.

 

[atyy]

This may be way off base but, Wouldn't the framework be defined by overlapping wavefunctions. ie. I make a measurement, the wavefunction of that measurement acts as a "filter" on family of histories?

As I understand it, there is no you sitting outside the system making a measurement. If there is then measurement is reintroduced as a fundamental concept, and the measurement problem is not solved.

So given multiple frameworks, and the need to choose one framework, who chooses the framework? Does the single framework rule reintroduce the measurement problem?

 

[craigi]

But if the sun is needed then isn't there still a part of the universe that isn't quantum?

If the whole universe is quantum, then given that there are multiple frameworks (a framework is a family of consistent histories), and that a single framework must be chosen, who chooses that framework?

For the first part, I just don't know if we can consider the solar systems of the universe as a collection of isolated macroscopic quantum systems, each with their own classicality. Perhaps gravity plays the dominant role in decoherence at this scale. I'm not sure if anyone understands this, yet. However, the universe is filled with neutrinos and photons, which have coherent quantum properties and their own histories, independent of our classical world.

For the second part, I think what you're talking about is how CH phrases the reduction from mixed to pure state, that I was referring to.

So given multiple frameworks, and the need to choose one framework, who chooses the framework? Does the single framework rule reintroduce the measurement problem?

I need to read up on this.

 

[stevendaryl]

As I understand it, there is no you sitting outside the system making a measurement. If there is then measurement is reintroduced as a fundamental concept, and the measurement problem is not solved.

So given multiple frameworks, and the need to choose one framework, who chooses the framework? Does the single framework rule reintroduce the measurement problem?

I thought the idea was that all frameworks are equally valid, but are not equally useful. A framework in which cats are in a superposition of dead and alive is perfectly valid, although it would be useless.

 

[atyy]

I thought the idea was that all frameworks are equally valid, but are not equally useful. A framework in which cats are in a superposition of dead and alive is perfectly valid, although it would be useless.

All frameworks are equally valid, but as I understand it there are multiple frameworks and one must choose one, because different valid framework as are not compatible. If one must choose one framework from many, who chooses it?

We can't talk about usefulness for the measurement problem, since there are no observers, for whom the choice is useful.

 

[stevendaryl]

All frameworks are equally valid, but as I understand it there are multiple frameworks and one must choose one, because different valid framework as are not compatible. If one must choose one framework from many, who chooses it?

We can't talk about usefulness for the measurement problem, since there are no observers, for whom the choice is useful.

I don't quite understand the issue. You can use any framework you like. But the questions you can ask depend on the framework. So choose a framework that is appropriate for the questions you want answers to.

You can choose a different framework than I choose, and presumably the formalism works for both of us.

 

[atyy]

I don't quite understand the issue. You can use any framework you like. But the questions you can ask depend on the framework. So choose a framework that is appropriate for the questions you want answers to.

You can choose a different framework than I choose, and presumably the formalism works for both of us.

But choosing is subjective. Hasn't the subjective nature of the classical/quantum cut been reintroduced?

If there are no observers, who chooses? Can the multiple frameworks coexist without an observer to make a choice? It's not obvious to me they can, since they are incompatible.

Edit: According to Gell-Mann and Hartle's http://arxiv.org/abs/1106.0767, in the no observers case, one can consider one history from each framework or realm to be "real", but because the frameworks are incompatible, each "real" history has nothing to do with any other "real" history. They consider this a challenge to the notion of reality in CH, so they introduce negative probabilities, which according to them makes sense in an extended Bayesian framework (in the sense of de Finetti). After they do this, they say that there is a single "fine-grained history" which is real. Their comments seem to support that the measurement problem has not been solved in CH, because in the absence of an observer, there is no one to choose a single framework from the multiple valid, but incompatible frameworks.

 

[Jilang]

Atyy, I am stuck already with this paper which looks very interesting, QM as a classical stochastic theory with negative probabilities. What is a fine grained history in simple words?

 

[craigi]

But choosing is subjective. Hasn't the subjective nature of the classical/quantum cut been reintroduced?

If there are no observers, who chooses? Can the multiple frameworks coexist without an observer to make a choice? It's not obvious to me they can, since they are incompatible.

Edit: According to Gell-Mann and Hartle's http://arxiv.org/abs/1106.0767, in the no observers case, one can consider one history from each framework or realm to be "real", but because the frameworks are incompatible, each "real" history has nothing to do with any other "real" history. They consider this a challenge to the notion of reality in CH, so they introduce negative probabilities, which according to them makes sense in an extended Bayesian framework (in the sense of de Finetti). After they do this, they say that there is a single "fine-grained history" which is real. Their comments seem to support that the measurement problem has not been solved in CH, because in the absence of an observer, there is no one to choose a single framework from the multiple valid, but incompatible frameworks.

As I understand it, you can choose whichever framework suits your purpose, on a theoretical level. The Single Framework Rule, ensures that inferences can only be generated from compatible properties. For example, spin x isn't int the same framework as spin z on an electron.

When it comes to making measurements, we must use the framework of the classical world. What we can do is manipulate the classical world, for instance, by changing the orientation of our detector, to modify which other frameworks are compatible with it. Just before measurement the frameworks combine through the process of decoherence and the quantum system goes from a superposed state to a mixed state. A pure state is then selected stochastically.

I don't consider that the classical world has a special role here. It's just a large framework where the superposition has dissipated and has a strong propensity to merge with other compatible frameworks. Any other framework could have equally played that role. Observers just happen to live in the this large framework. I don't see how they have a more significant role than that, under this interpretation.

 

[stevendaryl]

But choosing is subjective. Hasn't the subjective nature of the classical/quantum cut been reintroduced? If there are no observers, who chooses?

Who says there are no observers? You're an observer, I'm an observer. The goal of the CH is that there is no special physics associated with an observer.

Can the multiple frameworks coexist without an observer to make a choice? It's not obvious to me they can, since they are incompatible.

I think that all possible frameworks exist simultaneously, in the MWI type way.

 

[atyy]

As I understand it, you can choose whichever framework suits your purpose, on a theoretical level. The Single Framework Rule, ensures that inferences can only be generated from compatible properties. For example, spin x isn't int the same framework as spin z on an electron.

Yes, but this assumes an observer exists to choose a framework.To solve the measurement problem, observers cannot be fundamental, so let's say there are no observers. Do all frameworks coexist then? How can they if they are incompatible?

 

[atyy]

Who says there are no observers? You're an observer, I'm an observer. The goal of the CH is that there is no special physics associated with an observer.

To solve the measurement problem, observers cannot be fundamental.

I think that all possible frameworks exist simultaneously, in the MWI type way.

I can accept this as a solution to ehat happens at the fundamental level at which no observers exist. But now since different frameworks don't interact, can't we just throw all but one away?

 

[stevendaryl]

I can accept this as a solution to ehat happens at the fundamental level at which no observers exist. But now since different frameworks don't interact, can't we just throw all but one away?

We can ignore all but one. But the point is that there is nothing about the physics that would select one.