Falsification of eternal inflation

[Calrid]

I never said that fewer assumptions always lead to the truth. What I said was that fewer assumptions usually lead to the truth. And it's obviously only a valid statement as long as the theories in question aren't already falsified.

And that's a terrible, terrible example anyway because not only was the old theory of four elements falsified long ago, it wasn't all that well-defined in the first place.

Which is a qualitative claim that has nothing to do with science.

I don't think any example is going to satisfy you with such premises as you have so meh.

It's reasonable to have some debate. What is unreasonable is people insisting that their ideas hold special status such that any other ideas must have evidence to be considered reasonable, without any justification for making such a claim.

Ironically that is precisely what you are doing.

I'm not making any claims, quite the opposite, I am saying we cannot make any claims atm both are equally likely to be true.

No, they are materially different theories. Wave function collapse is never absolute in MWI, whereas Copenhagen Interpretation assumes that under unspecified conditions, the wave function collapses. The problem with testing the Copenhagen Interpretation is precisely the fact that the theory is ill-defined: this means that any time what you're testing looks like MWI, you can just shift the boundary of collapse in CI to save CI from being falsified. Collapse in CI, in other words, is just a "god of the gaps". MWI, on the other hand, is exactly specified and can, in principle, be falsified (whereas the wavefunction collapse postulate of CI cannot be falsified).

No they are not materially different, ie the wave function has been shown to exist in all possible states by experiment, they are conceptually different.

It's not a "god of the gaps" because CI makes no judgement about what exists in those gaps it can only say that what happens between emission and measurement is undefined, ie it is agnostic about an entity before measurement. The wave may exist as a definite entity but such a contention cannot be known. Hence we say that it's description is a figurative one not a pictorial one.

The paper, by the way, basically shows two things:
1. The boundary of collapse is gradual, as predicted by MWI.
2. The wavefunction collapse occurs even though no measurement of the wavefunction is performed (an interaction is turned on, the results of which are not recorded).

Have they shown that there really is gradual collapse? I Wouldn't know I can't read the paper. AFAIK anyway Copenhagen is agnostic on what measurably happens at collapse anyway as it has to be by definition.

Doesn't MWI contend that collapse doesn't happen anyway because the wave function is real and the wave function actually exists as a complete description of physical reality before measurement? Hence it is deterministic.

The results are not recorded and never can be. This isn't even philosophy it's pure arm waving.

It is kind of pointless discussing that paper since I can't read it, but I bet plenty of people have contended that its results do not prove CI is wrong, or I would of heard about it. I suspect your conclusions are arguable and a matter of interpretation.

 

[Chalnoth]

Ironically that is precisely what you are doing.



Your conclusions are arguable and a matter of interpretation.

Uh, what? No they aren't. That's what the experiment shows. There is a measurement of the resulting wavefunction, obviously. That is how the progressive decoherence is observed. It is the interaction in the middle that involves no measurement.

 

[Calrid]

Uh, what? No they aren't. That's what the experiment shows. There is a measurement of the resulting wavefunction, obviously. That is how the progressive decoherence is observed. It is the interaction in the middle that involves no measurement.

So we still don't know what exactly happens at measurement, but we'll claim we do anyway? it still sounds inductive to me?

Kind of a pointless contention anyway because I can't read what you are talking about. An interesting analogy to the wave but ultimately a pointless discussion.

 

[dm4b]

Well, the main issue here is that we don't know how to sensibly talk about a universe without a spacetime. But even if you want to consider a universe starting literally out of nothing, the default expectation would be for it to start in a generic state, not in a very special one. And a generic state is, by definition, a high-entropy state. So even with a universe beginning out of nothing, you run into the exact same entropy considerations.

Looks like I missed a few pages of activity on this thread!

The human mind has the tendency to feel that what it cannot imagine, is not there. Just because spacetime could have been "created" at the moment of the Big Bang, does not mean the Universe was created out of nothing.

And, if we understand so little about how the Universe could have come about in this way, as to say it was created out of nothing, we surely can't say what the default expectation would be. Also, is entropy going to exist outside time/space? I would think a more highly ordered state would be just as likely, which might make a lower entropy beginning to the Universe more likely.

So, I disagree, I don't think we necessarily would have the exact same entropy considerations.

But, you're right, we do lack the ability at this point and time to talk sensibly (in physics terms anyhow) about a reality (NOT our Universe, which wouldn't be around anyhow) without space and time. But, then again, I'm not so sure reality is restricted by what we humans can talk about, let alone imagine.

I am reminded of an Alan Watts saying when I see some of the challenges in physics today: "A persistently insolvable problem should always be considered to perhaps be a question being asked the wrong way".

 

[Calrid]

And, if we understand so little about how the Universe could have come about in this way, as to say it was created out of nothing, we surely can't say what the default expectation would be.

I think it's rather arrogant to assume we have as yet any concrete answers.

Also, is entropy going to exist outside time/space? I would think a more highly ordered state would be just as likely, which might make a lower entropy beginning to the Universe more likely.

Entropy is such a contentious term, it only really makes sense in very specific situations we can measure with any consistency, when it comes to commenting on overarching reality it is such an ill defined term. Hell I don't recall the name, but one scientist went mad trying to pin it down to a system specifically.

For a start define ordered? Isn't that a human conceit? What is more ordered about a human being and the chemicals that make up his structure in whatever state they happen to be in. Or a diffuse mostly hydrogen gas cloud and a star? Think about it?

But, then again, I'm not so sure reality is restricted by what we humans can talk about, let alone imagine.

Quite we tend to be very anthropocentric. As I said before I'm not sure reality gives a toss about our speculations on it, or the conditions we try to force on it because of our biases.

I think Bohr had it right we might not have the language or maths to describe reality, that I hope will not always be the case.

I am reminded of an Alan Watts saying when I see some of the challenges in physics today: "A persistently insolvable problem should always be considered to perhaps be a question being asked the wrong way".

Nice way of putting it, I'm sure Bohr would of approved.

 

[George Jones]

It's not another way of describing the Copenhagen interpretation, because it actually describes what happens at the boundary of collapse, while the Copenhagen interpretation does not. Despite the use of the word "interpretation", the two are not the same theory, because MWI drops the assumption of collapse. Its predictions about the boundary of collapse have also been experimentally verified:
http://prl.aps.org/abstract/PRL/v77/i24/p4887_1

I have access to the paper but I have not read (or even downloaded) it. From its abstract:

The decoherence phenomenon transforming this superposition into a statistical mixture was observed while it unfolded

This seems to agree with my limited understanding. Decoherence takes a quantum state to a statistical mixture of classical states, and each classical possibility is a branch for the MWI.

A statistical mixture of classical states is represent by a diagonal density matrix, but the decomposition of a given diagonal density matrix into a classical mixture of states is highly non-unique! Which decomposition is used for the branching?

 

[Calrid]

Gah I so wish I could read it.

The Devil is in the details.

Be damned if I'm spending money on something if it doesn't conclusively disprove CI though. I doubt it does though, I suspect it just throws up more questions. If it does then surely science missed out on the moment where once and for all CI was disproved, I have as yet seen no breaking news.

 

[George Jones]

Gah I so wish I could read it.

Don't you have access to the paper (hardcopy or ecopy or both) through your school?

I'm a "mature" student studying maths & physics p/t AAMOI.

 

[marcus]

Hi George and Cairid,
Cairid did you know of the concept *Kairos* in Christian theology?
"In the New Testament kairos means "the appointed time in the purpose of God", the time when God acts (e.g. Mark 1.15, the kairos is fulfilled). It differs from the more usual word for time which is chronos (kronos)."
http://en.wikipedia.org/wiki/Kairos#In_Christian_theology

Your handle could be interpreted "of or pertaining to the Hour" (of fulfillment or of some significant event.)

I have no particular interest in theology, I just happened to notice the nice analogy between "the collapse of the wave function" and the Greek concept of *kairos*, the moment of fulfillment. Just an odd association that came to mind.

 

[Chalnoth]

Looks like I missed a few pages of activity on this thread!

The human mind has the tendency to feel that what it cannot imagine, is not there. Just because spacetime could have been "created" at the moment of the Big Bang, does not mean the Universe was created out of nothing.

And, if we understand so little about how the Universe could have come about in this way, as to say it was created out of nothing, we surely can't say what the default expectation would be. Also, is entropy going to exist outside time/space? I would think a more highly ordered state would be just as likely, which might make a lower entropy beginning to the Universe more likely.

So, I disagree, I don't think we necessarily would have the exact same entropy considerations.

Oh, we absolutely would. And here's why: entropy is proportional to the logarithm of the number of states that replicate the same overall properties of the system. So the number of configurations that look like a low-entropy state are vastly, vastly outnumbered by the number of states that look like a high-entropy state. The argument for a universe out of nothing starting in a generic state, then, has nothing to do with the physics of what came before. It's just simple counting.

In order to get a low-entropy start of the universe, you'd have to have some sort of physical process that makes the much smaller number of low-entropy states more likely, despite their very small number.

 

[Chalnoth]

Gah I so wish I could read it.

The Devil is in the details.

Be damned if I'm spending money on something if it doesn't conclusively disprove CI though. I doubt it does though, I suspect it just throws up more questions. If it does then surely science missed out on the moment where once and for all CI was disproved, I have as yet seen no breaking news.

A quick Google search found a free version:
http://www.atomwave.org/rmparticle/ao%20refs/aifm%20refs%20sorted%20by%20topic/decoherence%20refs/BHD96.pdf

But by the way, it doesn't falsify the Copenhagen Interpretation per se, because the Copenhagen Interpretation is, in regard to its treatment of wavefunction collapse, unfalsifiable, due to the fact that the conditions of collapse are completely unspecified. So in this regard, CI is "not even wrong". MWI, on the other hand, could have been falsified but wasn't.

 

[dm4b]

In order to get a low-entropy start of the universe, you'd have to have some sort of physical process that makes the much smaller number of low-entropy states more likely, despite their very small number.

Well, and that is sort of the key point I was trying to make.

Since we have no idea what physics would be required to "create" time itself, I don't think we can make any definitive claims. It may very well be that the "mechanism" that brought about the Universe (and spacetime), only allows the very Universe that we happen to have.

Also, once again, this in no way implies that the Universe came out of "nothing" either.

 

[Chalnoth]

Well, and that is sort of the key point I was trying to make.

Since we have no idea what physics would be required to "create" time itself, I don't think we can make any definitive claims. It may very well be that the "mechanism" that brought about the Universe (and spacetime), only allows the very Universe that we happen to have.

Also, once again, this in no way implies that the Universe came out of "nothing" either.

My point is that even if we don't know how to sensibly talk about what came before, we can sensibly talk about the beginning itself.

In any event, whatever the physical process that started off our region of space-time, that physical process has to somehow produce low-entropy states with higher frequency than would be implied by their entropy. One idea that I rather like is this one:
http://arxiv.org/abs/hep-th/0410270

It proposes that if we take seriously the idea of a small positive cosmological constant, then the asymptotic future of a universe is one that is very large and very empty, with extremely high entropy. But, it also has a very large volume, which means a low entropy density. So if only a tiny fraction of this volume undergoes a quantum fluctuation to become a new region of space-time, then that will actually not be a terribly large change in entropy, so it is not so strongly disfavored.

 

[dm4b]

So if only a tiny fraction of this volume undergoes a quantum fluctuation to become a new region of space-time

That does sound like an interesting proposal, within the framework of eternal inflation.

I'm not real up on the details of eternal inflation, so just wondering:

(1) How exactly does the new region of spacetime get "created" under eternal inflation? What's the mechanism for bringing about a new "Universe"? What math is used to describe this mechanism?

(2) How is this new region "connected" to our region of spacetime (assuming it spawned off our region of spacetime because of the above mentioned quantum fluctuation)?

(3) Does the theory allow any communication, or transit, between our spacetime and the newly created region of spacetime?

Thanks in advance for any info you can provide on these.

dm4b

 

[Chalnoth]

That does sound like an interesting proposal, within the framework of eternal inflation.

This isn't the same sort of idea as normal eternal inflation. It is similar in that it is future eternal, with new regions undergoing inflation all the time. But the difference is that they look at fluctuations in the late universe that can produce new inflating regions, instead of considering inflation itself to continue eternally all the time.

(1) How exactly does the new region of spacetime get "created" under eternal inflation? What's the mechanism for bringing about a new "Universe"? What math is used to describe this mechanism?

A universe dominated by a small cosmological constant has a temperature, albeit a very small one. This temperature will lead to fluctuations in all sorts of matter fields including, very occasionally, the inflaton field.

(2) How is this new region "connected" to our region of spacetime (assuming it spawned off our region of spacetime because of the above mentioned quantum fluctuation)?

It actually becomes disconnected almost instantly after formation. The new universe, from the outside, looks like a microscopic black hole that pops into existence then immediately decays away into nothing. From the inside, though, and entire new universe can form: the new universe sort of "pinches off" and becomes its own region of space-time.

(3) Does the theory allow any communication, or transit, between our spacetime and the newly created region of spacetime?

Nope.

 

[marcus]

...
In order to get a low-entropy start of the universe, you'd have to have some sort of physical process that makes the much smaller number of low-entropy states more likely, despite their very small number.

The Loop bounce does that.
Since when quantized in loopy fashion, gravity becomes repellent at very high density.
Uniformized states became more likely for a brief interval around the bounce.

Ashtekar's most recent review paper: http://arxiv.org/abs/1005.5491

BTW I see that the 2004 Carroll-Chen paper you cite has accumulated 33 citations:
http://arxiv.org/cits/hep-th/0410270
That's better than I would have expected, pretty respectable under the circumstances.
However, a good many of the cites look to be of the polite mention-in-passing variety. I don't see the Carroll-Chen idea catching on, being developed by other researchers, picked up by phenomenologists and so forth.

 

[Chalnoth]

The Loop bounce does that.
Since when quantized in loopy fashion, gravity becomes repellent at very high density.
Uniformized states became more likely for a brief interval around the bounce.

Ashtekar's most recent review paper: http://arxiv.org/abs/1005.5491

I remain extraordinarily skeptical of the bounce approach. The symmetry around the bounce would seem to require extreme fine tuning at the bounce, despite claims to the contrary. After all, if the picture is genuinely generic, then this approach predicts that if we reverse our time coordinate, then the finite past of our universe looks like a collapsing universe, which, according to LQC, predicts an inflationary expansion into the past behind the bounce (with the subsequent increase in entropy that this brings along with it).

So it seems like the idea has a contradiction built-in: if it's genuinely generic, then it predicts there will be an increase in entropy into the past, but that increase in entropy into the past requires massive fine tuning.

BTW I see that the 2004 Carroll-Chen paper you cite has accumulated 33 citations:
http://arxiv.org/cits/hep-th/0410270
That's better than I would have expected, pretty respectable under the circumstances.
However, a good many of the cites look to be of the polite mention-in-passing variety. I don't see the Carroll-Chen idea catching on, being developed by other researchers, picked up by phenomenologists and so forth.

I don't see how this is relevant.

 

[skydivephil]

With regards to inflation being eternal into the future but the not the past I have a question.
If we label the present T and future eternal inflation is true then there is a real point eternally into the future,lets call that point Te. IF Te is real then oberveers at Te will consider point T an eternity into their past. So for them the past is eternally far away. How could we know that we are not them? Can someone illucidate?

 

[dm4b]

(1) How exactly does the new region of spacetime get "created" under eternal inflation? What's the mechanism for bringing about a new "Universe"? What math is used to describe this mechanism?

A universe dominated by a small cosmological constant has a temperature, albeit a very small one. This temperature will lead to fluctuations in all sorts of matter fields including, very occasionally, the inflaton field.

Thanks for the info Chalnoth. I'm still confused on one question, which is quoted above.

As I understand it, inflation did not start at t=0 in our Universe. If so, inflation did not create the Universe. Is that correct?

If it is correct, I don't see what the mechanism is to actually create another region of spacetime, which spawns off our Universe. So, why does a quantum fluctuaton in our spacetime, bring about an entirely separate region of spacetime? What is the mechanism for this? What math desribes that mechanism and process? What physics is used to describe the creation of spacetime?

Your second response seems to be starting to answer this. But, I'm still not understanding the physics of how one region of spacetime can get "pinched off" our Universe?


In addition, it also seems like eternal inflation is a chicken/egg type question. If it does indeed claim that new Universes are created, or spawned off from, inflation fields created by quantum fluctuations in existing Universes, well, what started it all? Where did the first Universe come from? It doesn't go eternally into the past, which means there is a t=0, right? Well, what kicked it all off?

 

[Chalnoth]

Thanks for the info Chalnoth. I'm still confused on one question, which is quoted above.

As I understand it, inflation did not start at t=0 in our Universe. If so, inflation did not create the Universe. Is that correct?

Well, that depends upon what you mean by "t=0". If you mean the time of the singularity if you naively extrapolate a Lambda-CDM model back in time, yes, it is correct that inflation started before then. That time t=0 would have just been a particular time late in inflation, just before reheating, not any special time. The start of inflation would have come earlier. We don't know exactly when so it's not so easy to label a specific time t=0 that would have been the beginning of inflation.

Inflation is still responsible for the generation of the observable universe.

 

[marcus]

...
So it seems like the idea has a contradiction built-in: if it's genuinely generic, then it predicts there will be an increase in entropy into the past, but that increase in entropy into the past requires massive fine tuning.

I'm curious, Chalnoth. How do you define the entropy of the universe? The quantity which you think would be increasing as you go back in time through and beyond the bounce?

 

[dm4b]

Well, that depends upon what you mean by "t=0". If you mean the time of the singularity if you naively extrapolate a Lambda-CDM model back in time, yes, it is correct that inflation started before then. That time t=0 would have just been a particular time late in inflation, just before reheating, not any special time. The start of inflation would have come earlier. We don't know exactly when so it's not so easy to label a specific time t=0 that would have been the beginning of inflation.

Inflation is still responsible for the generation of the observable universe.

hmmm, doesn't really sound like you answered my questions, but that's okay, I kinda figured there may not be an answer for those particular ones.

Sounds like eternal inflation leaves as many questions as answers, as far as a "where did the Universe come from" theory.

 

[Chalnoth]

I'm curious, Chalnoth. How do you define the entropy of the universe? The quantity which you think would be increasing as you go back in time through and beyond the bounce?

I don't see how it matters in this case, because we define entropy as increasing into the future in our universe. If the bounce is generic, then that is enough to show that the theory predicts entropy increasing into the past before the bounce.

 

[Chalnoth]

hmmm, doesn't really sound like you answered my questions, but that's okay, I kinda figured there may not be an answer for those particular ones.

Sounds like eternal inflation leaves as many questions as answers, as far as a "where did the Universe come from" theory.

Well, I felt I'd go with the first one, since the rest depended on that.

 

[marcus]

I don't see how it matters in this case, because we define entropy as increasing into the future in our universe. If the bounce is generic, then that is enough to show that the theory predicts entropy increasing into the past before the bounce.

I don't understand, entropy has to be defined or there is no argument.
BTW I assume that by "generic" you mean what Loop cosmology people call "robust". That is it happens pretty much with any choice of parameters, does not require fine adjustment or special choices.

Indeed that has been found. The bounce is a robust feature.

However the theory does not predict that the pre-bounce collapse looks just like the post-bounce run backwards.

I suspect that is the source of your misunderstanding.

 

[Chalnoth]

I don't understand, entropy has to be defined or there is no argument.

I don't see that it is necessary. This is a point about the symmetry of physical law. As long as LQC is time reversal invariant, then it should equally describe the universe's behavior backward in time as forward.

Nevertheless, you could use the simple definition used to describe the entropy during inflation of any Hubble volume as being proportional to the horizon scale. With this definition, entropy would indeed be decreasing as the universe collapses.

BTW I assume that by "generic" you mean what Loop cosmology people call "robust". That is it happens pretty much with any choice of parameters, does not require fine adjustment or special choices.

Indeed that has been found. The bounce is a robust feature.

If the previous universe doesn't look like a universe that came about as a result of inflation with the time coordinate set the wrong way, then the bounce isn't robust. If it does look like that, then it doesn't explain the low entropy boundary.

 

[marcus]

I never heard anyone say that LQC model of bounce is time reversal invariant. What Ashtekar says is that you have a classical universe collapse (often called a "big crunch") leading up to the bounce. A crunch does not look like a bang movie run backwards. The Loop bounce has an obvious "arrow of time" directionality.

About microscopic reversibility of laws, note that the distinction between microscopic and macroscopic law becomes hazy during what A. calls the "quantum regime" or "Planck regime" phase of the bounce. It is definitely something for theorists to explore and get sorted out.

 

[marcus]

About the need to have a definite mathematical meaning for the entropy. You do need this. Otherwise you have no assurance that the word has any meaning at all in the regime under consideration.

 

[Chalnoth]

I never heard anyone say that LQC model of bounce is time reversal invariant. What Ashtekar says is that you have a classical universe collapse (often called a "big crunch") leading up to the bounce. A crunch does not look like a bang movie run backwards. The Loop bounce has an obvious "arrow of time" directionality.

About microscopic reversibility of laws, note that the distinction between microscopic and macroscopic law becomes hazy during what A. calls the "quantum regime" or "Planck regime" phase of the bounce. It is definitely something for theorists to explore and get sorted out.

That's not what I'm saying. What I am saying is that the expansion of our space-time would look like a collapse run backwards. Because of that, one should be able to do the same analysis backwards in time with our universe as the "starting" point.

 

[marcus]

What I am saying is that the expansion of our space-time would look like a collapse run backwards.

I think that is wrong, thought I made that clear earlier. I'm too lazy to look back and find where I made that point.

You are welcome to think that a classic universe collapse looks like the start of expansion run backwards. You must suit yourself. I don't want to argue with somebody's deep-set mental images.

 

[marcus]

If anybody else besides Chalnoth (who seems to have a conceptual hangup) is following this thread, the obvious reason that classical expansion from a smooth intitial condition does not look like a classical crunch run backwards is basically just what Two-Fish said a few posts back.

As long as grav. is universally attractive the forward progression of time tends towards the grav field---the geometry of the U---becoming more lumpy and pock-marked.

The geometry gets more and more inhomogeneous as stuff coagulates and as structure forms.

I've said this before in this and parallel threads.

So the picture of a crunch has lots and lots of inhomogeneity.

On the other hand starting when the classical regime takes over in the big bang you have lots of uniformity, for 100s of thousands of years. Lots of homogeneity. It takes time for structure to begin to nucleate and start gathering.

So the two movies are quite different. One movie is not going to bear any resemblance to the other movie run backwards.

I will try to discuss this, and how this happens in bounce cosmology ( a research area that seems to be taking off) in that other thread.

"Eternal Inflation" is a much older idea (1980s? 1990s?) and this thread is supposed to be about "Falsification of eternal inflation". So I think to be considerate of people interested in that kind of thing we ought to discuss bounce cosmology, how it implements this, and the definition of entropy, and so forth in that other thread.

Here is the "Definition of entropy of the universe?" thread's link, so we can get out of this "eternal inflation" one:
https://www.physicsforums.com/showthread.php?t=487703

 

[eloheim]

Hey just a quick (edit: okay maybe not) question in regards to this issue and the OP's question: (And I have to admit I haven't made it through this whole thread yet so I hope I'm not repeating what others have already said.)

Is the "youngness paradox" an example of the anthropic principle? My understanding (via Brain "String-bean" Greene) is that Steven Weinberg suggested in his '87 paper on the topic that, should there be a multitude of different universes with different cosmological constants, then the probabilities say we should inhabit one with a constant that has a value just low enough for us to be able to exist, and not much smaller.

Does the OP's criticism of eternal inflation follow the same path, by pointing out that, if true, then, given the odds, we should inhabit a universe that is just [STRIKE]young[/STRIKE] old enough for life to exist?

(Hoping I'm on the right track) I take it that the reason this argument is presented as evidence against E.I. is that it seems we could have come about a lot earlier? And that is because there have been stars and galaxies, presumably with planets like ours, for a long time now?

My question then is, "How long do universes consistent with E.I. allow for the emergence of life?" If the answer were something huge, like a trillion, trillion, trillion, trillion, etc. years, then wouldn't the "youngness paradox" actually bolster the case for E.I. because, relatively speaking, we did evolve extremely quickly?

 

[offroff]

eloheim,
You understand me I think. My original post and the youngness paradox is very similar but according to the other guys in this thread it's flawed reasoning.

But if the youngess paradox is valid after all, then I don't think your scenario helps E.I. at all. You just have to consider that there will be so many universes like ours out there. Almost all of them one second or more younger than ours. The universes older than ours is just not relevant because there are so few of them.

Anyway I think the youngness paradox is too naive. And I think it shows that physical cosmology is ridicoulously naive (because guys like Guth takes the youngness paradox seriously).