Decoherence of the Universe as a whole

[Karl Coryat]

Hello, I would like to ask if there is a standard explanation of decoherence in the universe as a whole.

I can see how decoherence is responsible for the classical behavior of a measured macroscopic object in its environment. But what about the big picture? Is it thought that the entire universe began spontaneously decohering immediately after the Big Bang, and if so, what would be considered the 'measured object' and what was the environment?

I'm obviously missing something, otherwise the first stars never could have formed. Links to good papers on this topic would be appreciated; I'm surprised that there are essentially only two Google results from searching the phrase "decoherence of the universe." Thank you.

 

[Dmitry67]

It is very interesting and deep question.
Decoherence requires the arbitrary basis of the decoherence. It is not clear how to define such basis in early universe.

 

[Karl Coryat]

I thought of this after reading a sentence in a 1993 paper by Zeh: "The universe as a whole never decoheres." In context, he was illustrating closed vs. open systems, but it got me wondering if the question had a good answer.

 

[tom.stoer]

My interpretation is that decoherence always works if you separate a system (here: the universe) into a subsystem and the environment (= universe minus subsystem).

As there is no observer outside the universe (as the universe is all there is) the universe does not feel decoherence.

But any observer in the universe may define her own subsystem, e.g. a "spherical screen" around herself. This need not be a physical object but just a mathematical artefact. Therefore the universe w/o the small ball enclosed by the screen is the subsystem, whereas the ball is the environment.

This works even w/o a physical (or human) observer; it works once one draws a boundary such as the sphere; and it works for small and large spheres equally well. Normally (e.g. in a lab) the subsystem is a small ball (e.g. the interior of a Pauli trap), the environment is everything outside the defining sphere (the Pauli trap itself, dust, air molecules, photons, the laboratory, ...); in the case of the universe it's just the other way round, you turn it inside out.

The reason why I use the sphere here is the holographic principle. It says that strictly speaking it is enough to describe not the wohle physical subsystem but only its boundary Hilbert space living on the sphere. So any measuring device, observer or environment defined by a sphere does not interact with the whole subsystem (which could be rather large :-) but only with the boundary Hilbert space.

This preserves locality and explains why decoherence is an observer or subsystem specific phenomenon.

According to decoherence there is no reason to separate the subsystem and the environment geometrically (like I did with the sphere); it's just one way to do it to get in contact with the holographic principle.

Of course what I call a sphere need not be a geometrically perfect sphere, S² (topologically) will do.

 

[Karl Coryat]

Thank you Tom, that explanation helps a lot. I'm wondering, though, if you (or anyone else) have an idea or opinion on what the first event, or "catalyst," of decoherence within the universe might have been -- if it is even possible to consider such a question. How do we get from a universe that presumably begins as whole and undivided, to one with subsystems and the resulting discrete matter? If it is simply a function of anisotropy, don't we need decoherence before anisotropy can occur?

Alternatively, perhaps this question plays into Hawking's top-down cosmology. If I'm understanding it correctly, he proposed that the universe began in a superposition of configurations (as a way to approach the fine-tuning question). Yet we still need to get from these superposed universes to our decohered universe somehow. I apologize if my language is off, but I haven't seen a lot written on this topic. Perhaps I should ask on the cosmology forum?

 

[tom.stoer]

This is a difficult question.

Let me rephrase in order to see that I understand correctly: Let's start with a highly (maximal ?) symmetric state of the universe at some time very short after the big bang (or whatever it may be). According to qm this state will evolve according to a unitary operator U = exp(-iHt); regardless what H and t is the state will stay symmetric (H is assumed to respect the symmetry).

The problem now is that we introduce subsystems according to symmetry considerations (we may define a sphere in a physical why, e.g. in a certain evacuated region of space) which breaks the symmetry. But in practice it's the other way round: something already broke the symmetry such that we will decide afterwards where to place the sphere.

The question is how this symmetry of the initial state is broken. According to my argument presented above nothing can break the symmetry as the universe as a whole is not subject to decoherence. But an observer defining a boundary Hilbert space cannot break the symmetry, either. At least the symmetry of the ensemble of all possible definitions of subsystems remains unbroken.

Is this your question?

Then perhaps the only solution (which I can think about) is spontaneous symmetry breaking. But I am not sure and I have to admit that you hit a weak point.

 

[George Jones]

I don't have this stuff straight in my mind. A post I made a while back:

https://www.physicsforums.com/showthread.php?p=1810866#post1810866.

 

[Dmitry67]

In any case, deterministic theory can't break a symmetry - unless it is multi-history.
So either MWI... or BM, but in BM assymetry must be encoded from the very beginning in the configuration of BM particles, it is just hidden until some moment of time. I find it ugly because it means that initial conditions at BB have huge (hidden) entropy. So Mukhanov is right - MWI is the best candidate.

 

[tom.stoer]

But I don't like MWI; please find another solution for me :-)

 

[Dmitry67]

But I don't like MWI; please find another solution for me :-)

Do you agree that collapse interpretations are history now?
Then the choice is between SM, BM and MWI.

 

[tom.stoer]

collapse is (ancient) history; what are SM are BM?

 

[Dmitry67]

collapse is (ancient) history; what are SM are BM?

SM = Stochastic Mechanics
aka macroscopic realism
aka Shut up and calculate

BM = Bohmian Mechanics
aka dBB
aka Demistifier's Mechanics :)

 

[tom.stoer]

OK, let's summarize:

Copenhagen / collapse is over.

MWI is metaphysical speculation because it introduces experimentally inaccessable ontological entities and is therefore unacceptable according to Ockhams razor.

BM has been discussed here for a while; I am not convinced that is has the potential to become applicable in quantum field theory.

Shut up and calculate is quite successfull - as long as everybody agrees not to aks certain questions.

My conclusion is that the quest is still open :-)

 

[Karl Coryat]

Thank you all; this has been very interesting. I like the idea of viewing the universe as an inside-out version of a small-system experiment. Perhaps experimentally probing the precise meaning of the 'Heisenberg cut' in these small systems could lead to an answer to the cosmological question.

 

[Dmitry67]

MWI is metaphysical speculation because it introduces experimentally inaccessable ontological entities and is therefore unacceptable according to Ockhams razor.

Not more than cosmology: as expansion accelerates and we will never contact anything outside the cosmological horizon, any claims that Universe is infinite or there is matter outside our Hubble space are also unacceptable :)

But seriously, talking about infinite Universe, there is a strange equivalence between MWI and any single-history theory: in truly infinite Universe ANY configuration is implemented. BM can be made more logical that way: claiming that initial configuration of BM particles does not contain infinite information, on the contrary, all configurations are implemented, and thus information is very low.

I find it very elegant if LG 'big bounce' is true: BB (t=0) (*) entropy is at minimum, at t>0 entropy increases, at t<0 it increases too, and as all possible states are implemented universe wavefunction U is symmetric: U(t)=U(-t). Universe before the Big Bounce is the same as Universe after it, they are equivalent (**)

(*) t is cosmological time, whatever it means. I understand that close to BB the very notion of time might be different or dissapear.
(**) Of course, observers at t<0 persieve time as going to the direction where entropy increases, hence, away from BB, info -inf.

 

[tom.stoer]

I do not fully agree with you regarding the Hubble sphere; depending on the dynamics of the universe invisible regions can become visible in principle (even so that may not help in practice because it may take some time ...); this is different for MWI, as the perpendicular branches are invisible and experimentally unaccessable even in principle, not only in practice.

But I see an even more serious problem with an infinite universe: ANY configuration (with probability > 0) is implemented INFINITLY MANY TIMES! And this does not happen in parallel branches but in our universe!

So there are infinitly many copes of Dimitry and Thomas arguing about MWI. And there are also infinitly many copies of you agreeing with me that MWI is metaphysical speculation ... I am really glad about that ...

 

[Demystifier]

BM has been discussed here for a while; I am not convinced that is has the potential to become applicable in quantum field theory.

Why not?

 

[Dmitry67]

Tom, you are not the first one who noiced it, and Max Tegmark had even calculated an avergae distance to our exact copies :)

So if you believe in randomness+single history, then between infinite number of your copies some percentage of tom.stoers agree with my logic while another % do not. Branching occur... almost like in MWI... :)

 

[RUTA]

Do you agree that collapse interpretations are history now?
Then the choice is between SM, BM and MWI.

You also have blockworld and backwards causation models in the no-collapse alternatives.

 

[Dmitry67]

The only backward causality int. I am aware of is Transactional Interpretation.
But TI is a collapse int., even Cramer does not use that word explicitly; he talks about 'emitters' and 'absorbers'. But it is just different wording: if you can tell absorber from non-absorber (or in general event from non-event) then you can tell measurement device from any other system so it is equivalent to CI.

 

[RUTA]

The only backward causality int. I am aware of is Transactional Interpretation.
But TI is a collapse int., even Cramer does not use that word explicitly; he talks about 'emitters' and 'absorbers'. But it is just different wording: if you can tell absorber from non-absorber (or in general event from non-event) then you can tell measurement device from any other system so it is equivalent to CI.

TI is a no-collapse account, since it implies a blockworld, i.e., the coexistence/co-reality of the future, present and past. In TI, the future must coexist with the present since outcomes are to be linked to sources. And since the present is the future of the past, the past must coexist with the present. Thus, TI specifically, and backwards causation in general, imply blockworld. And, in a BW nothing “happens,” so there can be no collapse unless one wants to invoke some sort of “metatime.” Kastner, Elitzur and Dolev are working on that approach, btw.

Kastner has a paper forthcoming in Studies In History and Philosophy of Modern Physics where she will explain this: http://philsci-archive.pitt.edu/archive/00004707/01/transactional_qle2.pdf. Elitzur and Dolev are also shopping a version that Avi presented at Imperial College last week. Here is an older reference to their approach: Elitzur, A.C. and Dolev, S. (2005). Quantum phenomena within a new theory of time. In A. Elitzur, S. Dolev and N. Kolenda (eds.), Quo vadis quantum mechanics? Berlin: Springer Verlag, 325 – 349. But, I digress, back to the issue here.

Check out the special issue of Studies In History and Philosophy of Modern Physics, Volume 39, Issue 4 November 2008, with papers from Price’s conference, Time Symmetric Quantum Mechanics, University of Sydney, July 2005. Aharonov and Vaidman present their versions of the two-vector formalism therein. I don’t know if Peter Lewis has a paper therein or not, here is a reference to his no-collapse account:

Lewis, P. (2007). “Towards a local hidden variable theory.” Foundations of Physics, 37, 461 – 469.

Our Relational Blockworld is therein. See also: Stuckey, W.M. and Silberstein, M. and Cifone, M. (2008). “Reconciling spacetime and the quantum: relational blockworld and the quantum liar paradox.” Foundations of Physics, 38 (4), 348 – 383.

Huw has a book on the subject: Price, H. (1996) Time’s arrow and archimede’s point (Oxford: University Press, Oxford), p. 260.

 

[Dmitry67]

We had a discussion of TI few month ago. The major problem of TI is the same as CI (dispite the 'no-collapse' claim above) - it provides NO definition of what an absorber is. Absorbers have the same magic the measurement devices are in CI. Photocamera is an absorber, while lens and mirror are not absorbers. The same mirror is an absorber for the gamma-rays it can't reflect... Once you have an algorithm to tell an absorber from non-absorber in TI, you can use the same to define measurement device in CI. They are twins: they exploit the same 'magic of measurement' - just wording is different.

Besides that, TI has another serious problem.
What is so special about the light (photon)? Is it an only of 16 SM particles with a special role of an 'absorbtion'? If yes, why it is so special? If not, what's about absorbtion of gluons? How retarded waves are defined for non-abelian charges? One person on this forum who had a chance to talk to Cramer in person mentioned that even he does not have a consistent view.

 

[RUTA]

We had a discussion of TI few month ago. The major problem of TI is the same as CI (dispite the 'no-collapse claim above') - it provides NO definition of what an absorber is. Absorbers have the same magic the measurement devices are in CI. Photocamera is an absorber, while lens and mirror are not absorbers. The same mirror is an absorber for the gamma-rays it can't reflect... Once you have an algorithm to tell an absorber from non-absorber it TI, you can use the same to define measurement device in CI. They are twins: they exploit the same 'magic of measurement' - just wording is different.

Besides that, TI has another serious problems.
What is so special about the light (photon)? Is it an only of 16 SM particle with a special role of 'absorbtion'? If yes, why it is special? If not, what's about absorbtion of gluons? How retarded waves are defined for non-abelian charges? One person who had a chance to talk to Cramer in person mentioned that even he does not have a consistent view.

You're preaching to the choir. We think it's ludicrous to talk about "backwards causation" since you need a blockworld and, as so nicely stated by Geroch,:

There is no dynamics within space-time itself: nothing ever moves therein; nothing happens; nothing changes. In particular, one does not think of particles as moving through space-time, or as following along their world-lines. Rather, particles are just in space-time, once and for all, and the world-line represents, all at once, the complete life history of the particle. Geroch, R.: General Relativity from A to B. University of Chicago Press, Chicago (1978), pp 20-21.

We just had a long exchange with Kastner about TI. She champions it because "TI can derive the Born rule." But, as I told her, Cramer has no reason for introducing his "transaction" other than to end up with the Born rule, and a 1-for-1 swap between ad hoc rules is not a "derivation."

 

[Dmitry67]

Yes, I agree - BlockWorld is inevitable even in good old SR. I don't understand how one can accept SR without accepting BW, so for me BW is an old commonly accepted concept since 1905.
And yes, "There is no dynamics within space-time itself: nothing ever moves therein; nothing happens; nothing changes" - I totally agree, there is just one Universe wavefunction, defined by TOE equations, and this Universe wavefunction is, as a solution, static in some N-dimensional space. And Universe itself is just a set of equations and nothing more (MUH), initial conditions must be null or very simple (this is why I don't like BM)

 

[RUTA]

Yes, I just wanted to mention that for me BlockWorld is not related to TI: BlockWorld is inevitable even in good old SR. I don't understand how one can accept SR without accepting BW, so for me BW is an old commonly accepted concept since 1905. Strange that you mentioned it in the context with TI, did I miss something?

Those in the BCQM camp don't like to admit that they've a blockworld, it defeats the whole purpose of having a dynamical story to tell about the weirdest of QM phenomena. But, Price and Cramer both have admitted that, of course, the dynamical story is just that -- a story. We challenged their ability to do so in the case of the Quantum Liar Paradox (“Why Quantum Mechanics Favors Adynamical and Acausal Interpretations such as Relational Blockworld over Backwardly Causal and Time-Symmetric Rivals,” Michael Silberstein, Michael Cifone & W.M. Stuckey, Studies in History & Philosophy of Modern Physics 39, No. 4, 736 – 751 (2008)).

And yes, "There is no dynamics within space-time itself: nothing ever moves therein; nothing happens; nothing changes" - I totally agree, there is just one Universe wavefunction, defined by TOE equations, and this Universe wavefunction is, as a solution, static in some N-dimensional space.

I think Rovelli's Relational QM is a variant of this.

 

[rkastner]

It is not at all true that "Cramer has no reason for introducing his "transaction" other than to end up with the Born rule". His 1986 paper gives a sustained argument for why TI should be considered, addressing other aspects (e.g., the issue of complexity of the wave function which is different from classical fields; the existence of advanced solutions which heretofore have been ignored; and a possible solution to the measurement problem). I suggest folks read the basic literature before arriving at conclusions.

 

[Dmitry67]

It is not at all true that "Cramer has no reason for introducing his "transaction" other than to end up with the Born rule". His 1986 paper gives a sustained argument for why TI should be considered, addressing other aspects (e.g., the issue of complexity of the wave function which is different from classical fields; the existence of advanced solutions which heretofore have been ignored; and a possible solution to the measurement problem). I suggest folks read the basic literature before arriving at conclusions.

Do you know his (or yours) optinion regarding the question I asked before:

What is so special about the light (photon)?
Is it an only of 16 SM particles with a special role of an 'absorbtion'?
If yes, why it is so special?
If not, what's about absorbtion of gluons?
How retarded waves are defined for non-abelian charges?

 

[RUTA]

It is not at all true that "Cramer has no reason for introducing his "transaction" other than to end up with the Born rule". His 1986 paper gives a sustained argument for why TI should be considered, addressing other aspects (e.g., the issue of complexity of the wave function which is different from classical fields; the existence of advanced solutions which heretofore have been ignored; and a possible solution to the measurement problem). I suggest folks read the basic literature before arriving at conclusions.

You're absolutely right, I should've said, "There is no GOOD reason for subscribing to TI" and left Cramer's intentions out of it. We've discussed the technical aspects of the 1986 paper at length, so you know where I stand and why :-)

 

[Galteeth]

I do not fully agree with you regarding the Hubble sphere; depending on the dynamics of the universe invisible regions can become visible in principle (even so that may not help in practice because it may take some time ...); this is different for MWI, as the perpendicular branches are invisible and experimentally unaccessable even in principle, not only in practice.

But I see an even more serious problem with an infinite universe: ANY configuration (with probability > 0) is implemented INFINITLY MANY TIMES! And this does not happen in parallel branches but in our universe!

So there are infinitly many copes of Dimitry and Thomas arguing about MWI. And there are also infinitly many copies of you agreeing with me that MWI is metaphysical speculation ... I am really glad about that ...

"But I see an even more serious problem with an infinite universe: ANY configuration (with probability > 0) is implemented INFINITLY MANY TIMES! And this does not happen in parallel branches but in our universe!"

Hi, I'm not a scientist.

Maybe, I am not following your meaning. I don't understand why that would be the case. If I had a blank piece of paper, I can write a great many things. I could write random scribbles, I could draw pictures, or i could write nonsense letters. But let's say I write a story. I could write a story that makes absolutely no sense, with unicorns and talking snakes and people being dead and then reappearing in the next paragraph with no explanation.

But let's say that i do write a story that makes sense. And in this story there's a character who's wondering about his own nature, and somehow realizes that he is only a series of marks on paper.

Now lets's say it's not paper, but say a digital file with the same information possibilities as the piece of paper. And it's running a program to create all possible arrangements of information. My story, the one where the character has been wondering about his existence hasn't changed. It exists in one of the many, many files this program has generated.
Now does the physical presence of the saved file change the character of my story? In other words, does the coherency of the character's thoughts change whether or not the fiLE actually exists?
No, because the "sense" that the story makes is a property of the information in the story. My character doesn't have to worry about unicorns or random scribbles because that's not his story.

 

[RUTA]

But I see an even more serious problem with an infinite universe: ANY configuration (with probability > 0) is implemented INFINITLY MANY TIMES! And this does not happen in parallel branches but in our universe!

I don't see why that has to be true. As I tried to get JesseM to understand, you're assuming that every conceivable possibility is instantiated in an infinite universe. For example, suppose a phenomenon is coded in the universe (by some "universal statistical law" USL) to have a 50-50 outcome, call it A or B. According to your assumption, in an infinite universe all conceivable distributions will be instantiated in accord with the USL, which entails that in some region(s) scientists will always measure A (or B). In those places, the scientists will deduce that the phenomenon is not statistical, but definitive. In fact, only those regions who see 50-50 outcomes within statistical limits will deduce the USL. So, if you subscribe to this belief about the nature of statistical laws, how can science be done with any confidence? There is an alternative, though. That, in fact, what will be observed in any region of the infinite universe will be in accord with the USL. That means there is no place that sees all A or all B. In fact, there is no place that will empirically deduce the "wrong" USL within experimental limits (by assumption). So, as a scientist, why not assume this instantiation of the USL?