# Decoherence of the Universe as a whole

1. Jul 21, 2010

### 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.

2. Jul 21, 2010

### 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.

3. Jul 22, 2010

### 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.

4. Jul 22, 2010

### 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, SÂ˛ (topologically) will do.

5. Jul 26, 2010

### 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?

Last edited: Jul 27, 2010
6. Jul 27, 2010

### 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.

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.

7. Jul 27, 2010

### George Jones

Staff Emeritus
8. Jul 27, 2010

### Dmitry67

In any case, deterministic theory cant 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.

9. Jul 27, 2010

### tom.stoer

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

10. Jul 27, 2010

### Dmitry67

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

11. Jul 27, 2010

### tom.stoer

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

12. Jul 27, 2010

### Dmitry67

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

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

13. Jul 27, 2010

### 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 :-)

14. Jul 27, 2010

### 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.

Last edited: Jul 28, 2010
15. Jul 28, 2010

### Dmitry67

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.

16. Jul 28, 2010

### 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 ...

17. Jul 28, 2010

### Demystifier

Why not?

18. Jul 28, 2010

### 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... :)

19. Jul 28, 2010

### RUTA

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

20. Jul 28, 2010

### 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.