Is there any reason the Universe must have a wave function?

[C Rob]

I believe it does, but I'm having a debate with someone and I'm trying to prove why the universe must have a wave function. I was under impression the best equations for describing our universe involve it having it's own wave but he's asserting it doesn't need one. Can anyone help me? Super newb btw.

 

[marcus]

I believe it does, but I'm having a debate with someone and I'm trying to prove why the universe must have a wave function. I was under impression the best equations for describing our universe involve it having it's own wave but he's asserting it doesn't need one. Can anyone help me? Super newb btw.

Hawking popularized the phrase "the wave function of the universe".
What was usually meant was merely a quantum wavefunction for the SCALEFACTOR or roughly speaking the size of the universe.
I don't know what you and your fellow debater mean by the phrase. How to answer would depend on what you mean by it.

In cosmology there is a quantity called the scalefactor which intuitively is the average distance between galaxies. The symbol is a(t). The function a(t) is increasing with time.

a(t) is not the size of the universe, nor is it the size of the observable universe---we don't know what the size of the universe is, it might not even be well-defined, it might be infinite. So a(t) is just CALLED the size of the universe in popular books.
a(t) is well-defined mathematically in the context of the Friedmann metric and the Friedmann equations tell you how it evolves with time.

Naturally you can QUANTIZE the Friedmann equation and then instead of a normalized classical length observable a(t) you get a wavefunction describing our information about that length.

It is very modest and low-key. That is what Hawking and buddies usually meant when they wrote about "the wave function of the universe". No big deal. Not grandiose at all.
The actual reality of it is plain and simple. Universe expanding means the scalefactor is increasing and we have a wavefunction for that.

The quantum wavefunction version works a little bit better than the classical version when you get back near the big bang and want to push on to before the big bang.

Nowadays, the quantum cosmology people like Ashtekar and Bojowald do not talk about
"the wave function of the universe" because it sounds self-important and it's misleading. They talk about the wavefunction of the scalefactor, or the wavefunction of the inverse scalefactor. Sounds more dry and technical. Basically the same thing with less hype.

 

[Haelfix]

As often the case, people have different usage for the word.

The simplest concept for the "wavefunction of the universe" is the idea that you take the normal wave function of a particle, and then add every particle in the universe to it, probably in some uber complicated superposition. Presto, wave function of the universe. Obviously the states lives in a Hilbert space that is huge, poorly understood and probably illdefined mathematically. In some sense something like it is probably trivially true, but not particularly informative as it doesn't really tell you anything.

The above version is often truncated down to something that fits into a cosmological model, like the FRW metric. Obviously in this new context we are talking about tiny portions of the 'wave function of the universe', but this too is sometimes called 'the wave function of the universe'.

Slightly different but related in principle is the wave functional defined in the Wheeler De Witt equation, which encodes all the information about matter, geometry and field configurations over *the* spacetime of the universe (whatever that is). The hamiltonian constraint annihilates this wave functional in this picture. This is a slightly more interesting, and arguably deeper concept and relates to recent work Hawking and others did, and is amongst the foundations of quantum gravity.

Yet another version relates to the Many World interpretation of QM, about which I don't know much about.

You should come away with the feeling that these concepts are vague and have a good likelihood of being theoretical fictions. We are very far away from concrete statements about testable or observable parts of nature.

 

[Chronos]

A mathematical artifact, IMO. How big would the detector need to be? It's the sort of thing physicist probably joke about at the local pub.

 

[Demystifier]

I believe it does, but I'm having a debate with someone and I'm trying to prove why the universe must have a wave function. I was under impression the best equations for describing our universe involve it having it's own wave but he's asserting it doesn't need one. Can anyone help me? Super newb btw.

If QM is the fundamental theory describing everything, then Universe must have a wave function.
If QM is not so universal theory but applicable only as an effective description of micro phenomena, then the wave function of the Universe is meaningless.

Personally, I think the former is the correct view.

 

[marcus]

If QM is the fundamental theory ... then Universe must have a wave function.
If QM is not so universal theory ... then the wave function of the Universe is meaningless.
...

Demy,
how about a third possibility. The QM setup involves a system, and an observer outside the system who can make observations. QM accurately describes what the observer can know. The wave function represents the observer's information.

But this setup doesn't work with the universe.

The logical difficulty with supposing a quantum state of the universe is that the universe is all-inclusive. There is no room for a classical outside observer. Because any wanna-be observer is just another piece of the universe.

QM is not about Nature it is about what we can say about Nature, according to the ancient Bohrish wisdom.

And traditionally the observer who is saying things is supposed to be outside of the box.

James Hartle (among others) has thought about this and other fundamental difficulties with defining a quantum state of the universe, I believe, and has concluded that some modification of QM is needed to handle the problem. You may have read what Hartle has proposed, or perhaps you have some ideas of your own.

Personally I don't take a position on this. In the research that I'm familiar with, the wavefunction of the universe is simply the quantum state of the SIZE, which you need to run a quantum model of the big bang which goes back in time to before the big bang---quantizing the Friedmann equation of cosmology. In that sense, it is routine for people to plot the wavefunction of the size and calculate with it.

 

[cmos]

An isolated physical system will have a wavefunction. So particle-in-a-box has a wavefunction. Now make it two-particles-in-a-box. The wavefunction of that system is distinct from that of the first system. Same goes for many-particles-a-box; distinct wavefunction from that of the previous two systems. Now let our isolated system be the universe. Therefore, by the same reasoning, the universe has a wavefunction.

Disclaimer: I feel that decoherence makes the above argument invalid for macroscopic systems. Maybe someone else can shed some more light on this.

 

[vanesch]

I believe it does, but I'm having a debate with someone and I'm trying to prove why the universe must have a wave function. I was under impression the best equations for describing our universe involve it having it's own wave but he's asserting it doesn't need one. Can anyone help me?

That depends on whether you think quantum mechanics is applicable to the universe or not.

 

[Haelfix]

There are many proposals to get semi tractable notions of a 'wave function of the universe'.

In fact most of quantum cosmology research is about that question in one guise or the other (for instance common calculations are done in mini superspace +wkb approximations + choice of boundary conditions).

Alas the state of the art is far from satisfying. People are very far from a realistic calculation.

 

[Fra]

hidden information vs hidden structure

My choice of conceptual associations are along these lines.

Bohr's idea
- The wave function represents the observer's information.
- Information about what? - About the system!

If the systems is a box in a lab, this can be handled. But what if the box is the entire universe?

If we consider the idea of "the screen", then the screen is somehow the communication channel between observer and the unknown. Wether the outside is finite, or infinite is for the observer to try to figure out, if possible.

Do we even have to answer a priori what is on the other side of the screen? Isn't the whole point that that is part of this "information".

Lets say the wavefunction represents the information of one side of the screen, relative to the other side.

Then it seems the wavefunction of "the universe", relative to any choice of system/observer is different for each observer, and the world universe have no other meaning than the wavefunction of the observers environment (ie. what can be seen through "the screen"), and to me it's sound that there are no a priori structures here. These structures should be part of the information.

But like Rovelli argues in his relational QM, this doesn't prevent them from trying to communicate their views. And this very process results in a physical interaction. This means that due to the physical interactions, this subjectivity doesn't imply any inconsistencies no more than the stick and clock readings of different SR observers differ.

It seems the problems start when we presuppose universal a priori structure of the universe, on which we define the information.

So, when we ask what is the "wavefunction of the universe" what strikes me first is the preimposed structure implied in the world universe. How about instead asking what is the "wavefuncion" or "state of information" of an observer. And let the information itself, describe this "universe".

Something doesn't smell right to me, when one asks a seemingly innocent question like what do you know about A. The structure implicit in A is hidden structure. Instead just ask "what do you know". (implicitly meaning about whatever you know anything about).

The observer seeing the world through the screen is IMO a nice abstraction here. The observer tries to make sense of whatever comes through the screen, without preconceptions about what's on the other side.

/Fredrik

 

[C Rob]

"The logical difficulty with supposing a quantum state of the universe is that the universe is all-inclusive. There is no room for a classical outside observer. Because any wanna-be observer is just another piece of the universe"-Marcus

So...we don't know if the universe has a wave function. Assuming it does, since it's in a definite state, the wave has been narrowed, right? An observer(or interaction) is necessary. Marcus is this why you don't like it, because it implies an observer(God)? I'm am not versed in physics, so please don't destroy me too thoroughly.

 

[marcus]

"The logical difficulty with supposing a quantum state of the universe is that the universe is all-inclusive. There is no room for a classical outside observer. Because any wanna-be observer is just another piece of the universe"-Marcus

So...we don't know if the universe has a wave function. Assuming it does, since it's in a definite state, the wave has been narrowed, right? An observer(or interaction) is necessary. Marcus is this why you don't like it, because it implies an observer(God)? I'm am not versed in physics, so please don't destroy me too thoroughly.

why should I want to scold you? and on what grounds?
I don't study up on questions like this.
As I said in my first post#2, I think, the research I read just deals with a wave function for the SIZE (or a scalefactor that plays the role of the size)
nobody I'm aware of worries about some kind of all-describing wavefunction

Jim Hartle would be the closest---I mean of people who are still active in research. Maybe the others know some besides.

Unless you can construct the quantum state and calculate with it, I don't see the point.
But you are welcome to speculate as far as I'm concerned. (might be more appropriate in philosophy forum if it is not an empirical line of thought you are pursuing however)

 

[Haelfix]

The holy grail of quantum cosmology is in fact to find such a beast in the full superspace, and uniquely pick out the initial conditions and the boundary conditions.

But it probably won't happen until a full microscopic description exists of quantum gravity.

So people are content with doing various surgeries and approximations to the full wavefunction to get some sort of analytic control (see the no boundary proposal by Hawking et al, or the tunneling universe by Vilenkin). Those two are probably the most intensely studied ones in recent years.

 

[George Jones]

That depends on whether you think quantum mechanics is applicable to the universe or not.

The standard position in cosmology today is that quantum fluctuations early in the universe produced the clusters of galaxies which we see in pictures from Hubble and other telescopes. Many cosmology books mention this, but not all books mention the interpretational difficulties that this entails. For example a superposition of fields seems to lead, via Einstein's GR equation, to a superposition of spacetime geometries.

See the thread

https://www.physicsforums.com/showthread.php?t=246423

My paragraph on decoherence is probably too short or just plain wrong. Elaborations and corrections are most welcome, as are answers to chronnox's questions. I have been meaning to read Schlosser's Reviews of Modern Physics paper

http://arxiv.org/abs/quant-ph/0312059

on decoherence, but I have not gotten around to it.

 

[marcus]

The standard position in cosmology today is that quantum fluctuations early in the universe produced the clusters of galaxies which we see in pictures from Hubble and other telescopes. Many cosmology books mention this, but not all books mention the interpretational difficulties that this entails. For example a superposition of fields seems to lead, via Einstein's GR equation, to a superposition of spacetime geometries.

http://arxiv.org/abs/quant-ph/0312059

...

Parts of the Schlosshauer paper were very helpful. It's a well written review---he was careful to be clear and accessible.

I see the point you emphasize here. If decoherence of a quantum state happens by interaction with the environment, and we consider the state of the whole shebang, then what is the environment?

Does a path integral, or sum over histories, approach sidestep such difficulties?

 

[marcus]

C Rob asked if the Universe must have one grand all-encompassing all-describing wavefunction. I think that's the drift. At first I wasn't sure what was meant (a lot of quantum cosmology deals with socalled minisuperspace models, reduced down to a few parameters---indeed classical cosmology focusses a lot on reduced models).

but I think here the issue is the necessity of one grand comprehensive quantum state.

Must there be one? Rovelli argues there need not be:
http://arxiv.org/abs/quant-ph/0604064
Relational EPR
Authors: Matteo Smerlak, Carlo Rovelli
Foundations of Physics 37 (2007) 427-445
(Submitted on 10 Apr 2006)

"We study the EPR-type correlations from the perspective of the relational interpretation of quantum mechanics. We argue that these correlations do not entail any form of 'non-locality', when viewed in the context of this interpretation. The abandonment of strict Einstein realism implied by the relational stance permits to reconcile quantum mechanics, completeness, (operationally defined) separability, and locality."

The quantum state (technically a vector in a Hilbertspace) represents an observer's information about the world. Experience has shown it is an excellent way of representing uncertainties as well as knowledge, and the difficulties of determining various things at once---the limitations of a real observer. So good.

Rovelli says there doesn't have to be a single unique overall correct state. He says let every real observer have his own separate Hilbertspace and his own current idea of what the state vector (wavefunction) is. And let the observer keep track of how that changes as he makes measurments and determines more about the world. then observers can consult and find out what facts they can agree on.

Fra brought this up in an earlier post. So anyway Rovelli's is one possible view.

I see this as a question in the foundations of quantum mechanics, or foundations in general.
I tend to agree with Rovelli. I don't think Hilbertspaces and wavefunctions are physically real. I think they are manmade devices for describing the behavior of information. Wavefunction QM is about the observer and his information. It doesn't have some ultimate objective reality like it's the ground of existence. So there doesn't have to be One Big wavefunction. But others certainly differ on that score.
==========================

The practical issue in quantum cosmology, I think, is not to solve foundational or interpretational problems once and for all, but something quite different. Namely to get a mathematical formalism that joins largescale General Relativity geometry with extreme geometry of high curvature, very small scale, very high density matter.

In other words the practical problem is to remove the singularities of General Relativity in an expanded framework and be able to smoothly relate largescale geometry to smallscale (probably chaotic fluctuating) quantum geometry. And to find matter in the picture.

So from my perspective about the most advanced effort in this direction currently is what Ambjorn and Loll discuss in their July 2008 SciAm article. The link is in my signature. That is one way that quantum cosmology could turn out looking----if they can work matter into the picture successfully.

the quantum average of all the fluctuationg geometies turns out to be a smooth classical one. And it is a model that you can compute with----get numbers out of----suggesting the possibility of eventually testing predictions.

I see an effective quantum cosmology taking shape there which does NOT require postulating the existence of some single Overall wavefunction. (at least I don't think it does, please let me know if you think a big overall quantum state is required by the Ambjorn Loll approach.)

 

[Chronos]

Sounds stringy to me, marcus.

 

[Fra]

I think Marcus highlights a very important point and that is what the objective is. What is the problem we are answering to?

It was some month ago I listened to it but I think it was in one of the introductory talks available as mp3's here http://www.fuw.edu.pl/~kostecki/school.html, where Rovelli also made an analogous point to which I agree. This was to keep the real problem in mind, do not loose focus. He suggested that the only reason to study toy models, say 3D gravity, is if one has reasons to infer from it to 4D. Otherwise Rovelli said it's a waste of time according to work on unrealistic toy models. So his suggestion was to focus on the real, full problem, and never loose focus. Still of course, each one of us might still have different ideas of what that is.

Marcus raises some practical problems om of quantum cosmology. I can see the clash here when discussing this if we do not have the same focus.

For me personally I have quite a different perspective, which explains my philosophical angle. My core focus is exactly the fundamental issues, not only of QM, but even deeper, that of the scientific method, and the notion of physical law. And beyond this I see many possible excellent "practical application" beyond physics, and then I'm thinking mainly of AI, information theory and decision theory etc.

This is why bypassing the foundations questions to me, implies missing the point. Or missing my subjective point that is :)

This actually made me very curious on how differing views we all have on what the problem is. If we don't agree on the problem, we are unlikely to agree on the solution?

Rovelli says there doesn't have to be a single unique overall correct state. He says let every real observer have his own separate Hilbertspace and his own current idea of what the state vector (wavefunction) is. And let the observer keep track of how that changes as he makes measurments and determines more about the world. then observers can consult and find out what facts they can agree on.

I don't like or understand all of Rovelli's reasoning, but the above is a simple and excellent view IMHO.

I also think it holds at least a possible interesting solution to the unification problem if we simply identify the negotiation of disagreement about opinion with physical interaction? Rovelli associates communication of competing or contradictory information between observers, which physical interactions.

In this sense, the class of valid observers(or why not try to replace this world by matter?), having different information about the world is the reason for their interactions. So physical interaction between two systems, could be just an expression of differing information.

Interactions serve to exhacnge information, and thus something equilibrate, and decrease the "distance between them" if we think of distance as the distance between their opinions as if lived on a statistical manifold. The step to infer an attractive force here is not far.

So the direction of investigation I am personally look at is along those lines where one tries to characterize and find a phenomenology along the notion of "comparing competing views". If this is right, then the fundamental interactions might be encoded here. It would IMO we well in line with rovelli's INITIAL stance, not necessarily his further development which I don't understnand.

I think the point to take rovelli's idea further is an analysis of the consequences of the differing ideas. This is what I thought Rovelli was aiming at, before started to read his book, and I associated the spin networks as state of opinion, in a sense. But then the observer-observer interactions would be interactions between two networks, in some way. But I soon got the impression that this is not how he himlsef sees it, and I got sceptical about his development and I stopped reading.

/Fredrik

 

[Demystifier]

Demy,
how about a third possibility. The QM setup involves a system, and an observer outside the system who can make observations. QM accurately describes what the observer can know. The wave function represents the observer's information.

I view this as a variant of the second possibility.

 

[Atakor]

Hello,

First of all thanks for the link to Rovelli's talk, it's always a pleasur to listen to him.
( and even more to assist to his conferences).

Fra, indeed, such questions are highly dependant on the personnal philosophy (I think whether we are conscious or not about it).. for me, this leads to the notion of "reality".
if reality is what we see, say the wave function relative to an observer, that means that reality is relative or observer-dependant.
Personnaly I think that the 'real reality' is just the Invariants of such transformations.
just like in SR for ex.

Thus the Real wave function (if any) must be related to local wave functions, from which we can infer it _without getting out of the Universe_.
Does such a view contradict something ?

 

[Fra]

Hello Atakor!

First of all thanks for the link to Rovelli's talk, it's always a pleasur to listen to him.
( and even more to assist to his conferences).

Yes, to meet someone would be even more interesting, but that isn't practical for many, but to listen to talks by someone you know from books and papers are interesting! During longer car trips I tend to stuff an mp3 player with interesting lectures.

Personnaly I think that the 'real reality' is just the Invariants of such transformations. just like in SR for ex.

Thus the Real wave function (if any) must be related to local wave functions, from which we can infer it _without getting out of the Universe_.
Does such a view contradict something ?

The essence of what you express here is as far as I understand it, pretty much right in line with rovellis thinking! One might think the analogy to relativit is brilliant, it's the analogy Rovelli refers to his relation QM papers.

Unfortunately I personally (many will disagree I am sure) don't think it's that "easy" (that's not to imply that it's "easy", just that I think it's more complex)

I have noticed that it is difficuly to communicate the argument though. But in short it has to do with certainty in these invariants. At some level of reasoning, if I may be bold to say this, the "mistake" I think Rovelli makes in his reasoning here is the very same reason that causes him in this relational QM arguments to silently avoid reasoning around the physical basis of probability.

IMO, the suggested reasoning is I think a good start, but it is not finished yet, there is more.

I have an apparently very odd way of reasoning about this, but my vision is that laws of physics are similar in structure to a particular form of uncertain inductive reasoning.

If we consider that there is something such as invariants, then a question within the context of inductive reasoning of any given observer appears: Is there a risk-free induction or deduction of these invariants from the point of view of this observer? I have thought about this a lot and as far my head allows, I can't see there is. If I get Rovelli right, he seems to imagine (in the same manner we argue about existence of limits - ie probability) that well the observer can't know this, but he could something like know it probabilistically and he can further know the probability with certainty.

I can't wrap my head around this, and until them, I maintain what opinion that something is wrong. However, there is no question in my mind that it "could work" and could improve our theories, but we are still then pushing one fundamental probleam ahead of us.

So in this sense, I try to look for a fundamental reasoning that stands up to high standards of reasoning.

I'm not sure if my objection is clear. I've briefly argued about this before, but few seem to be bothered by it. So my take on this, is not to reject rovelli's ideas. I think he has argued more sound than many indeed, but perhaps it can be further improved. And as long as I see ways to get to grips with these problems, I can't motivated myself to work out lenghy constructions based on what I have reason to think is a slightly flawed foundation.

So my objection has to do with that all understanding, and that includes human science, as well as relational information between two molecules, are subject to the same structure of inductive reasoning, where ALL information has to be formed during these constraints. The priors can't be pulled in from under the table. At best they coudl be randomly assigned, and that's they way I look to solve this.

Consider the simple problem: How can one experimentally determined a probability free of uncertainty? And the usual FAPP response is not enough. Then consider that the notion of probability usually founding the logic of inductive reasoning ,and also part of the scientific method.

It seems rovelli thinks it's not a risk to ignore this, and go on with the normal fuzzy meaning of probability - and note - thus also ignoring some risks in his reasoning. I don't share this risk assessment, I even think there is more physics hiding in this.

/Fredrik

 

[Happablapp]

I think that if the universe has a wave function then so must any subset of it have one as well. And if subsets of the universe behave as waves, then there must be a way of modelling the behaviour of any system using interference of waves. For example, there must be an interference pattern that exactly predicts the circular path of a photon around the event horizon of a black hole. If anyone ever invents such an interference pattern, this could be useful in unifying gravity with electromagnetism.