Max Tegmark, Lee Smolin-polite conversation at Peter's

[marcus]

Max Tegmark, Lee Smolin--polite conversation at Peter's

interesting exchanges at Woit's blog
http://www.math.columbia.edu/~woit/wordpress/?p=310#comments
raising an important issue with the recent Tegmark et al, and Wilczek, papers.
links to those papers are here:
https://www.physicsforums.com/showthread.php?t=104167

These recent papers touched on multiverse ideas but failed to discuss CNS or justify the omission. Max Tegmark posted a message on Peter's
and the following ensued. Perhaps the most substantive is Smolin's post which is also the latest, so if you wish to cut to the chase just scroll down to the end of this post.

----quoting from Not Even Wrong---
Who Says:
December 17th, 2005 at 2:26 pm
Dear Max,
I thought it was a serious flaw in your paper that it failed to address Smolin’s cosmological natural selection proposal (CNS) and the accompanying arguments presented in

Scientific Alternatives to the Anthropic Principle
http://arxiv.org/abs/hep-th/0407213

—exerpt from abstract—
...We show however that it is still possible to make falsifiable predictions from theories of multiverses, if the ensemble predicted has certain properties specified here. An example of such a falsifiable multiverse theory is cosmological natural selection. It is reviewed here and it is argued that the theory remains unfalsified. But it is very vulnerable to falsification by current observations, which shows that it is a scientific theory...
—endquote—

Since the CNS proposal is the simplest and most directly testable multiverse model, and it has not yet been refuted by observation as far as I know. I think it is incumbent on you to describe CNS and to say why you think it can be ruled out. Or explain why it is not testable by current astronomical observations, if you believe it is not.

It might further the discussion if you did this.

Cheers,
;-)
==========================

woit Says:
December 17th, 2005 at 3:26 pm
Hi Max,

Thanks a lot for taking the time to write here, I’d certainly enjoy talking to you in person about these issues sometime. Here are some comments, I hope they address the questions you asked.

First, some context. My background is in mathematics and particle physics, not cosmology, and what motivates me is the idea of trying to find an improvement of the standard model, using new mathematical ideas. I think Wilczek would classify me as the sort who is trying to get somewhere by “pure thought”, something which he doesn’t believe will ever work. I’d claim that, given the lack of any new helpful experimental input, particle theorists don’t really have any choice at the moment except to go this admittedly much more difficult route. I’ve never found string theory a particularly appealing idea for unification (although it gives interesting insights into strongly coupled gauge theories and has led to some important new mathematics), and think the way it has completely dominated mathematically-minded particle theory for the last twenty years has been a disaster for the subject. By now it should be clear that it is simply a highly speculative idea that failed, and the field desperately needs to acknowledge this and move on to trying other things. The whole string theory landscape program, especially in its “anthropic” version, seems to me to be a retreat from the very idea of doing science, motivated by a refusal to admit failure. For more about this, read my recent review here of Susskind’s new book.

Like just about any particle theorist, I’ve spent some time looking at what is going on in cosmology and hoping it will provide some new insights into how to get beyond the standard model. So far it seems to me cosmology has provided some interesting hints, but unfortunately they’re no more than hints. We’re agreed that what needs to be done here is to find ways of confronting our mathematical models with the real world. When dealing with models that involve a statistical ensemble of universes and observables that are only probabilistically determined, sure, one has to take into account selection effects.

The scientific part of my critique of your paper was not that I thought any of it was incorrect. The major part of it, your anthropic calculation of the selection effect for the dark energy density was more complicated than I have the time or interest to follow. You end up with a probability distribution for the dark energy, with the observed value near the peak. What was unclear to me was exactly what conclusions you are claiming can be drawn from this. Ignoring the axion cosmology prior, it seems you’re just claiming that there’s an anthropic window in the dark energy density, and observations show we’re in the middle of it.

But, as a particle physicist, what I really want to know is what is left over when you remove the selection effect. Exactly what does your calculation says about the axion cosmology model? Have you provided any new evidence for it other than that it’s not obviously inconsistent? Can you rule out a flat dark energy density prior in favor of the axion cosmology one? Your paper didn’t seem to seriously address these kind of questions, or the more general question of when you can expect to get genuine, falsifiable predictions out of this kind of calculation.

That’s the scientific critique, there’s also a more general critique, one I made in my posting. Especially the early part of your paper seemed to me a heavily ideological push for the idea of anthropic determination of the standard model parameters, without anything to really back this up. Given the on-going disaster in particle theory these days due to the string theory landscape, I think this is really unhelpful.

Anyway, thanks again for writing in. If you’d like to write something more than a short comment in response, I’d be more than happy to put it up as a new posting, so people would be more likely to see it and comment on it if they wish.

Peter

========================

Aaron Bergman Says:
December 17th, 2005 at 3:48 pm
For my part, I think probabilistic arguments across multiverses are nonsensical. Otherwise, I don’t see a satisfactory response to the doomsday argument. So, you can file me under (a).

========================

Who Says:
December 17th, 2005 at 3:52 pm
Peter to Max: “If you’d like to write something more than a short comment in response, I’d be more than happy to put it up as a new posting, so people would be more likely to see it and comment on it if they wish.”

That would be great! Please consider doing this, Max!

============================

Max Tegmark Says:
December 18th, 2005 at 10:29 am
Hi Peter,

Thanks for your thoughful response. I’m glad to hear that you didn’t feel any of it was incorrect.

I agree that it will be interesting of one can make stronger statements about the axion model - as described in the paper, this will require
improved astrophysics calculations that I hope somebody will make. You described the early part of the paper as a heavily ideological push for the idea of anthropic determination of the standard model parameters. The intent was rather to push for an open mindset on the issue, since we frankly don’t know how many parameters will ultimately be computable from others.

Regarding the CNS-critigue by “Who”: The pre-inflationary axion model is a complete physical theory, whereas Smolin’s “Cosmological natural selection” is not (we lack a mathematical description of how black holes spawn universes; see also http://arxiv.org/abs/hep-th/0407266) . Moreover, my guess is that the hypothesis is ruled out by the low observed fluctuation level (~1/10^5), since raising it would lead to more black holes.

Finally and most importantly, I’m glad that many of you disagree with my views! As argued in http://arxiv.org/abs/physics/0510188, I think that diversity in the physics community is more useful than an ideological monoculture, since it motivates physicists to tackle unsolved problems with a wide variety of approaches.

;-)

===========================

Who Says:
December 18th, 2005 at 12:51 pm
Hi Max,
personally I view your “low observed fluctuation level” not as a fundamental physical constant but more of an ad hoc result. It could be symptomatic of more basic parameters established prior to the beginning of expansion. But this is largely a matter of preference. It is personal opinion on your part (as you say) and equally on mine!

I would have expected you to mention CNS in your paper and to say why, in your opinion, it is probably ruled out. This would have given others a convenient opportunity to argue that it has not yet been ruled out.

There have been quite a few papers in the past year relating to the question of how black holes might spawn universes. Much of the work (QG modeling black hole collapse) was by people who reported in the Friday (14 October) session of Loops ‘05. The main topic was LQC, but the same people work on quantum models of gravitational collapse and, for what it’s worth, the bounce mathematics has turned out to look rather similar in both cases.

You say “we lack a mathematical description” and cite a 2004 paper by Leonard Susskind responding to Smolin, but, as I say, there has been quite a bit of mathematical description since then. I also do not believe Susskind’s rebuttal stands unchallenged.

Be that as it may, my point is that the CNS idea has not been disposed of and should have been discussed. If you think it is not falsifiable, or that it has already been falsified, then you should at least give your reasons.

In case you would like to look up the past year’s QG papers modeling gravitational collapse and bounce, I will post some arxiv numbers later. Some of the relevant authors, if you want to look up their recent work yourself, would be Abhay Ashtekar, Martin Bojowald, Viqar Husain, Oliver Winkler, Leonardo Modesto, Parampreet Singh.

To be fair, one should note that Smolin’s CNS proposal is testable without reference to any specific mathematical description of black hole collapse and bounce. The CNS conjecture is that some reproductive/evolutionary mechanism has fine-tuned the constants for black hole production. CNS challenges us to find even one fundamental constant (hope you will pardon me if I decline to view your “low observed fluctuation level” as a fundamental constant) which if it were better tuned would result in substantially greater black hole abundance.
That is something one can use to test—and possibly falsify—CNS, even before one has a complete mathematical theory of the conjunction of black hole and big bang events.

Thanks,
;-)

==========================
Lee Smolin Says:
December 18th, 2005 at 1:19 pm
Dear Max and Who

Thanks for mentioning the CNS proposal. The issue of the level of delta rho/rho-the fluctuation level-was discussed in detail and resolved in the first paper published on the model, Did the universe evolve?, Classical and Quantum Gravity 9 (1992) 173-191, summarized in Life of the Cosmos, ps 309-10.

Given that this was the first theory based on the “landscape*” I would hope that people would look at the literature before dismissing it.

The point is that in simple one field inflation models the fluctuation level is proportional to the inflaton self-coupling constant. However the number of e-foldings in inflation is related to (if I recall right) the inverse of the square root of the inflaton self-coupling. The result is that the cost of raising the self-coupling to produce primordial black holes or make more galaxies is that there are fewer e-foldings in inflation and the resulting universe is exponentially smaller. The result is that CNS predicts that the self-coupling should be as small as possible consisent with some steady level of black hole production. i.e. many more black holes are produced by a slow, but steady rate of black hole production in an exponentially larger universe than by a burst of primordial black hole production in an exponentially smaller universe.

This seems to characterize our universe where most black holes are made as supernova remnants. It also leads to a prediction, which is that inflation is governed by one parameter and not by a complicated potential with more than one parameter in which the fluctuation amplitude and number of e-foldings would be independent. So were there any evidence for an inflaton potential governed by more than one parameter the theory would be ruled out. I made this prediction in 1992 and so far it has held up.

I can endorse WHO’s comments: recently we have very good detailed support from the quantum theory of gravity that black hole and cosmological sinigularities bounce. There is not yet a detailed description of the parameters mutate in a bounce, but this is certainly plausible given current views of the “landscape”.

Max, if these were your main objection to CNS, would you now agree that the theory is viable?

I would go further, CNS is the only landscape theory proposed so far that makes falsifiable predictions, Should this not make it the leading
candidate for an explanation of the choices of parameters in the case that the landscape is real?

Thanks,

Lee

*and indeed the origin of the term, which comes from “fitness landscape”.

----------end quote-------

 

[Spin_Network]

There are some interesting questions raised by Tegmark, I recall a recent documentary programme here in the UK, by M Rees :http://www.channel4.com/science/microsites/W/what_we_still_dont_know/index.html

I actually made a post about here on PF, but cannot find it?

Tegmark made a lot of contributions in the programme, I will watch it again as I have the video of it, but for now I can recall what the programme left me asking myself.

1)In the inflationary model, where did all the extra-dimensions evolve from?

2)When one invokes dimensional "re-winding" from, now-to-then, B Bang, space and time are seperate, a SINGLE dimension cannot have more than one paramiter of 'extra quantities', thus it either space or time?

3)The expansion of Space without "extra-dimensions", can follow the CNS of smolins e-foldings, but the sequence of events follow via Blackhole singularities, and not Universal Singularities.

There was a really interesting visual effect contained in the documentary, it involved Observers and the rewinding of Spacetime and dimensional constraints. One can see that in a 3+1 (spacetime), the time paramiter actually take the role of "Space", in that it is singular with respect to observation.

P.S actually I recall the link above was in my original posting, the Lady sitting down in the first column, was where in the video the interesting thing happened, there was a cyclist (I believe it was meant to be a young einstien), cycling "backwards"..but the weird thing was the cyclist pedals were still pedalling in the "forwards motion"!

 

[Spin_Network]

There are some interesting questions raised by Tegmark, I recall a recent documentary programme here in the UK, by M Rees :http://www.channel4.com/science/microsites/W/what_we_still_dont_know/index.html
I actually made a post about here on PF, but cannot find it?
Tegmark made a lot of contributions in the programme, I will watch it again as I have the video of it, but for now I can recall what the programme left me asking myself.
1)In the inflationary model, where did all the extra-dimensions evolve from?
2)When one invokes dimensional "re-winding" from, now-to-then, B Bang, space and time are seperate, a SINGLE dimension cannot have more than one paramiter of 'extra quantities', thus it either space or time?
3)The expansion of Space without "extra-dimensions", can follow the CNS of smolins e-foldings, but the sequence of events follow via Blackhole singularities, and not Universal Singularities.
There was a really interesting visual effect contained in the documentary, it involved Observers and the rewinding of Spacetime and dimensional constraints. One can see that in a 3+1 (spacetime), the time paramiter actually take the role of "Space", in that it is singular with respect to observation.
P.S actually I recall the link above was in my original posting, the Lady sitting down in the first column, was where in the video the interesting thing happened, there was a cyclist (I believe it was meant to be a young einstien), cycling "backwards"..but the weird thing was the cyclist pedals were still pedalling in the "forwards motion"!

Just a quote from the programme :"The cosmological constant can be considered as the intrinsic mass and volume of empty space, which Einstein had suggested was not zero. It turns out that this number needs to be set to an accuracy of one part in a trillion, trillion trillion, trillion trillion, trillion trillion, trillion trillion, trillion. Any minuscule variation and life is a non-starter."

Which make the whole idea of Extra-Dimensions quite irrelevent?

Now Smolins CNS is deemed to be derived from before the BB, quote:
Smolin offers a theory of the universe that is radically different from anything proposed before. He departs from contemporary physicists to explore the idea that the laws of nature we observe may be the partial result of a process of natural selection that occurred before the Big Bang.

Now my understanding is that to emerge from a point prior to the Big Bang, all the fluctuations are really no different from the selective "Vacuum State" in stringtheory? the natural selection = the vacuum state?..and thus the extra-dimensions evolved from the stringtheory camp, cannot yeild any predictive quantities.

 

[marcus]

There are some interesting questions raised by Tegmark, I recall a recent documentary programme here in the UK, by M Rees :http://www.channel4.com/science/microsites/W/what_we_still_dont_know/index.html
...

...
Now Smolins CNS is deemed to be derived from before the BB, quote:

Smolin offers a theory of the universe that is radically different from anything proposed before. He departs from contemporary physicists to explore the idea that the laws of nature we observe may be the partial result of a process of natural selection that occurred before the Big Bang.
...

It is interesting that this is a quote from a BBC program, if I understand what you are saying. I did not watch the program myself. Is this really a quote from Martin Rees? If so I am glad that he is not ignoring CNS in this one case. In the recent Tegmark et al paper, which Rees co-authored, it is NOT discussed, I believe.

========================================

Anthony Aguirre has joined the discussion at Peter's

----quote from Woit blog----
Anthony Aguirre Says:
December 19th, 2005 at 3:39 pm
Hi Lee,

This is a nice argument, but does not terribly ease my misgivings about CNS, which are approximately 5-fold.

First, the initial field value will also come into the number of e-foldings, and (absent a concrete model) that may or may not be correlated with the inflaton coupling.

Second, other things could lead to ultra-numerous black holes, e.g. a very high baryon/photon ratio. This physics should be unrelated to inflation, and variable by tuning the coupling constants. Or, we could introduce tilt into the perturbation spectrum to create scads of primordial black holes. Or is there a size requirement?

Third, I’ve never been clear on the ‘number of black holes’: per unit what? If per unit photon (or dark matter particle), we can easily maximize this by increasing the baryon (or dark matter) to photon ratio. If per unit volume, this is of course time-dependent. And in all cases we must worry about infinite universes (note that a universe starting out finite does not mean that it cannot spawn an infinite universe either, as in open inflation). These are more-or-less the same semi-intractible issues eternal inflation has to deal with.

Fourth, CNS still requires, even if we accept the nascent models for creating baby universes from black holes, some way of passing-down constants with small variations.

Last, CNS does not answer the question of why the universe supports life, unless black holes are alive. It merely transforms the question into ‘why does a high black-hole formation rate correlate with life?’. An alternative that *would* explain things, I would say, were if advanced civilzations were responsible for creating the baby universes. But that it pretty science-fictiony and I suspect you would rather avoid going down that path.

Thus I would be very pleased if CNS could be made to work and create strongly-peaked probability distribution for the observed parameters. (Note, in regard to our recent paper that CNS is not, as I see it, an alternative to the scenario we present, but a special case of the ‘multiverse’ explanation in which p_prior is strongly peaked around certain parameter values related to black hole formation. Unless axions are involved in black holes, I’m not sure it would impact the argument). But it seems to me that there are even more missing pieces, and about as many very difficult problems of principle to deal with, than in the eternal inflation scenario.

cheers,

Anthony
----endquote----

Aguirre is assistant director (to Tegmark) of the new Templeton-funded "Foundational Questions" foundation FQX and also one of the co-authors of the recent paper about the 31 fundamental dimensionless constants by Tegmark, Wilczek, Rees, Aguirre.

 

[marcus]

smolin's reply is actually useful exposition (point for Galilean dialog)

Smolin has replied to Aguirre:
------quote from Woit's blog-----

Lee Smolin Says:
December 19th, 2005 at 10:14 pm
Hi Antony,

Thanks for your criticisms. Briefly, CNS works well for the parameters of the standard model of particle physics, and I don’t think this can be disregarded. Every variation of those parameters for which definite conclusions could be reached, the conclusion was that the number of black holes made decreased. The case is not as strong for some of the cosmological parameters. I don’t take these as important as we don’t which if any cosmological parameters are determined by parameters of a theory. But there are some answers to your points, which I’ll repeat and then address:

“… the initial field value will also come into the number of e-foldings, and (absent a concrete model) that may or may not be correlated with the inflaton coupling….”

Fine, but the mechanism that determines the initial field may be independent of the inflaton coupling to leading order, if so then this decouples, i.e. for any initial field value my argument works.

What about “… a very high baryon/photon ratio…?”

If I recall correctly, this affects nucleosynthesis, and many other things, as this is a cold or tepid big bang. Ill check.

“…Or, we could introduce tilt into the perturbation spectrum to create scads of primordial black holes..”

The question again is what the consequences are for other processes, Ill think about this.

“…I’ve never been clear on the ‘number of black holes’: per unit what?”

The answer is per unit bounce, which means per unit black hole in the previous universe.

“Fourth, CNS still requires, even if we accept the nascent models for creating baby universes from black holes, some way of passing-down constants with small variations….”

Granted. This remains to be filled in. But still it is possible to deduce falsifiable predictions. This is not surprising, Darwin could make predictions in spite of being ignorant of DNA and its replication.

“…Last, CNS does not answer the question of why the universe supports life, unless black holes are alive. It merely transforms the question into ‘why does a high black-hole formation rate correlate with life?’…”

There is a clear correlation. You only get a lot of black hole production in universes with many massive stars. You only get many massive stars when the initial mass function (IMF) is power law out to many solar masses rather than exponentially damped. While this is incompletely understood, there is evidence the high end of the IMF is power-law because there are processes which efficiently cool giant molecular clouds (GMC’s) and because the GMC’s are well shielded from star light. Both the cooling and shielding involve carbon and oxygen. (This is described in some detail in Life of the Cosmos). Hence the chemical complexity life requires may exist because it is needed for a universe to efficiently produce stars massive enough to leave black hole remnants.

“…An alternative that *would* explain things, I would say, were if advanced civilzations were responsible for creating the baby universes. But that it pretty science-fictiony and I suspect you would rather avoid going down that path….”

I don’t. But Harrison (of Harrison-Zeldovich) has written a paper which goes down that path, as has Louis Crane.

I hope this helps,

Thanks,

Lee

---endquote---

 

[marcus]

=====quote from Woit blog====
Anthony Aguirre Says:
December 20th, 2005 at 6:38 pm
...
...
Lee:

This is probably a discussion that we should have over a nice meal sometime, and hopefully I can make it out to PI soon to do so. I would love to understand how CNS could work, and would be happy to be convinced. Perhaps my biggest stumbling block now, as I think about it, is the invocation of inflation. If inflation is involved, then it is easy for inflation to be future eternal, and if this were possible for *any* set of constants, then universes involving eternal inflation, which spawn infinitely many black holes, would immediately dominate the ensemble, and we are back where we started. If you have a good idea for allowing inflation while precluding the possibility of eternal inflation, that would be wonderful (as much as I enjoy thinking about eternal inflation, I would not mourn it’s passing, I have to say But, for example, in the case with possible multiple vacua, I see no way to avoid it from occurring at least for some parameter values — and I think evolution would find a way to exploit this mechanism for infinite reproduction!

cheers,

Anthony

---------------------------
Moshe Says:
December 21st, 2005 at 9:25 am
Peter, since this discussion is going in your territory, I am curious what is your opinion of cosmological natural selection as an alternative to anthropic reasoning. Feel free to ignore this if you haven’t looked at it in sufficient detail.

Happy holidays!

best,

Moshe

-----------------------------
woit Says:
December 21st, 2005 at 9:46 am
Hi Moshe,

I haven’t looked at CNS in enough detail to have an intelligent comment on it. The crucial question for anything like this is whether it generates testable predictions. Here Smolin claims it does, and his discussion here with Aguirre looked like it could be very illuminating, but I haven’t taken the time to really understand the arguments on both sides (life is short, and it’s exam period here these days…)

In general, what I really care about and am willing to invest time in trying to carefully understand, are new physical ideas that explain something about particle theory, or new mathematical ideas that might somehow be useful in better understanding particle theory. This means there are a lot of topics in cosmology and quantum gravity that I’ve never studied in a really serious way, CNS is one of them.

-----------------------------
Lee Smolin Says:
December 21st, 2005 at 9:52 am
Hi,

To Antony on eternal inflation. I understand the arguments that go from inflation to eternal inflation but I find myself unconvinced of them, especially in the context in which inflation would arise in a small region to the future of the bounce of a black hole singularity. I would trust inflation in general much more if we had solved the cosmological constant problems. The cosmological constant problem arises not in a fundamental theory but only at the effective field theory level and, as pointed out by Dreyer in hep-th/0409048 there may be a physical mechanism such as proposed by he or Volovick which dynamically adjusts the vacuum non-perturbatively close to zero, in which case there is no inflation. Of course then I would lose the argument I made, but there is no problem in principle as the horizon problem is already solved by the fact that the universe arises from a bounce.

On the other hand, WE KNOW that there are many black holes-perhaps as many as 10^19 in our Hubble volume. And we have good theoretical evidence that black hole singularities bounce. So only one element is missing for the CNS scenario-that the changes in vacuum resulting results in small changes in standard model parameters.

For the time being we must assume this-although it is possible this assumption may be tested soon. But once we do we get quite a bit.

My impression, if I can say so, is that many cosmologists undervalue the positive successes of CNS. It EXPLAINS otherwise mysterious features of our universe such as the setting of the parameters to make carbon and oxygen abundent-not because of life but because of their role in cooling GMC’s. It also EXPLAINS the hierarchy problem and the scale of the weak interactions-because these can also be understood to be tuned to extremize black hole production. Further, it EXPLAINS two otherwise improbable features of glaxies: why the IMF for star formation is power law and why disk galaxies maintain a steady rate of massive star formation.

Moreover CNS makes a few real predictions: that the upper mass limit of neutron stars is less than 1.6 solar masses and that inflation is governed by one parameter. Made in 1992 these predictions could easily have been falsified but they have stood up.

CNS makes these genuine explanations and predictions without having to invoke the AP, whereas eternal inflation requires invoking strongly the AP just to be plausible.

It seems to me CNS has a much longer list of successes than eternal inflation. which so far as I can tell explains NOTHING about the particle physics standard model parameters and which makes no fallsifiable predictions.

There are always many theories with attractive features, my understanding of how science is supposed to work is we are to pick out of the list of otherewise attractive theories those few that genuinly explain and predict over those that don’t.

The problems I set out to solve were 1) why the neutron is a bit heavier than the proton, 2) why their difference is comparable to the electron mass, 3) the value of the fermi constant, 4) the value of the strange quark mass, 4) the hierarchy problem. These are real problems which have been around for decades. So far as I know, CNS is the only explanation proposed so far that is both genuinly explanatory and genuinly falsifiable in terms of ongoing observations.

If you ask me to discount all this because of a highly speculative theory of the very early universe that explains nothing and makes no falsifiable predictions, that may easily disappear when the cosmological constant problem is solved, I am afraid I don’t think this is consistent with the methodology of science as I understand it. I understand reasonable people may differ, and I have the greatest respect for you, Max and other cosmologists, but I must admit also I am puzzled by your apparent judgements of what is more and less reliable in present theory.

Also, to Michael Bacon: Methods from quantum computation are being used to solve the problem you mention, which is how distinct particle states emerge from a spin foam. See D. W. Kribs, F. Markopoulou, Geometry from quantum particles, gr-qc/0510052.

Thanks,

Lee

=======endquote=======

 

[marcus]

Frank Wilczek joined the discussion at Peter's blog today.

His post is in this thread
http://www.math.columbia.edu/~woit/wordpress/?p=310
where earlier we saw posts from Max Tegmark, Anthony Aguirre, and Lee Smolin.

there are currently about 97 posts on this thread and Wilczek's is around post #95

So far, he has not said very much. Only to encourage people to read his actual papers on these issues (what issues? explanation vs excuses?)

I am very much in favor of reading Wilczek's papers. He is has a flair for writing and often gets original-sounding ideas---can be fun to read.
He has a recent one called

http://www.arxiv.org/abs/hep-ph/0512187
Enlightenment, Knowledge, Ignorance, Temptation
Frank Wilczek
10 pages, 5 figures. Summary talk at 'Expectations of a Final Theory', Trinity College, Cambridge, September 2005. v2: date of earlier conference corrected (2003->2001)
"I discuss the historical and conceptual roots of reasoning about the parameters of fundamental physics and cosmology based on selection effects. I argue concretely that such reasoning can and should be combined with arguments based on symmetry and dynamics; it supplements them, but does not replace them."

 

[marcus]

http://arxiv.org/astro-ph/0511774
Dimensionless constants, cosmology and other dark matters
Max Tegmark, Anthony Aguirre, Martin J Rees, Frank Wilczek

---quote from Tegmark et al page 1, Introduction---

B. The origin of the dimensionless numbers

So why do we observe these 31 parameters to have the particular values listed in Table1? Interest in that question has grown with the gradual realization that some of these parameters appear fine-tuned for life, in the sense that small relative changes to their values would result in dramatic qualitative changes that could preclude intelligent life, and hence the very possibility of reflective observation. As discussed extensively elsewhere [9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22], there are four common responses to this realization:

1. Fluke: Any apparent fine-tuning is a fluke and is best ignored.

2. Multiverse: These parameters vary across an ensemble of physically realized and (for all practical purposes) parallel universes, and we find ourselves in one where life is possible.

3. Design: Our universe is somehow created or simulated with parameters chosen to allow life.

4. Fecundity: There is no fine-tuning, because intelligent life of some form will emerge under extremely varied circumstances.

Options 1, 2, and 4 tend to be preferred by physicists, with recent developments in inflation and high-energy theory giving new popularity to option 2.

Like relativity theory and quantum mechanics, the theory of inflation has not only solved old problems, but also widened our intellectual horizons, arguably deepening our understanding of the nature of physical reality. First of all, inflation is generically eternal [25, 26, 27, 28, 29, 30, 31], so that even though inflation has ended in the part of space that we inhabit, it still continues elsewhere and will ultimately produce...

...More dramatically, a common feature of much string theory related model building is that there is a “landscape” of solutions, corresponding to spacetime configurations involving different values of both seemingly continuous parameters (Table 1) and discrete parameters...
---endquote---

 

[Kea]

---quote from Tegmark et al---
1. Fluke: Any apparent fine-tuning is a fluke and is best ignored.
2. Multiverse: These parameters vary across an ensemble of physically realized and (for all practical purposes) parallel universes, and we find ourselves in one where life is possible.
3. Design: Our universe is somehow created or simulated with parameters chosen to allow life.
4. Fecundity: There is no fine-tuning, because intelligent life of some form will emerge under extremely varied circumstances.

These options bothered me a lot. I would have thought the most popular option amongst physicists was:

5. There is new physics in emergent complexity, which will eventually help us understand the apparent fine tuning without resorting to a simple multiverse.

 

[Chronos]

Agreed. Kea. There is a simple set of axioms that will explain this unverse, in my opinion. I think we can do this without appealing to 'super' universes

 

[Chronos]

I relayed an observation to Dr. Smolin that appears to refute his CN selection model. He has yet to reply. Of course that does not mean anything, it's just disappointing.

 

[Spin_Network]

I relayed an observation to Dr. Smolin that appears to refute his CN selection model. He has yet to reply. That does not mean he is wrong. But I find it disturbing.

Cool ..so mentors are allowed to ask questions,not just give answers!

I know you have not a reply yet, but will you eventually share this "answer" to CN problem here on PF?

 

[Spin_Network]

It is interesting that this is a quote from a BBC program, if I understand what you are saying. I did not watch the program myself. Is this really a quote from Martin Rees? If so I am glad that he is not ignoring CNS in this one case. In the recent Tegmark et al paper, which Rees co-authored, it is NOT discussed, I believe.

========================================

Sorry Marcus, I must have missed this fact earlier, its actually from here:

http://www.channel4.com/science/microsites/W/what_we_still_dont_know/findoutmore.html

As you can see(scroll down a little) it is the overview of Smolins book.

I have been following your postings here and there (N-E-W), and your thread here shows you are tackling interesting problems,thanks again for a really interesting thread, and of course your contributions at N-E-W, "who" knows..you might get a reply directly here from L Smolin

 

[marcus]

I relayed an observation to Dr. Smolin that appears to refute his CN selection model. He has yet to reply. That does not mean he is wrong. But I find it disturbing.

could you give me the link, please?

my impression is that the size of the errorbar on a neutron star mass measurment can allow for some wiggleroom. It might take several observations. I would very much like to see your most recent find.

 

[Mike2]

These options bothered me a lot. I would have thought the most popular option amongst physicists was:
5. There is new physics in emergent complexity, which will eventually help us understand the apparent fine tuning without resorting to a simple multiverse.

I wonder if there is not some sort of contraint on entropy that is at work in this CNS rule. I think there may be a conservation of information at work in the fine tuning of the parameters so that complex structures can emerge. And it probably has something to do with the expansion of the univerese.

At farther and farther distances, space is traveling faster and faster away from us. Eventurally, space is traveling as fast and faster than the speed of light. At that distance we will never see what is beyond that distance because the photons cannot travel towards us faster than space is receding away from us. This is called the cosmological event horizon. Effects at this distant horizon are very similar to the effects seen at the event horizon of a black hole. As object approach the cosmological event horizon they become red shifted and they appear to slow down, until they disappear altogether.

Some have suggested that the entropy calculated for the event horizon of a black hole can also be calculated for the area of the cosmological event horizon as well. If this is the case, then we can see that as the expansion of the universe accelerates, the cosmological event horizon gets smaller so that the entropy calculated for the material inside it decreases as well. This may be responsible for the emergence of complex structures such as life. Curious that life arose on Earth at about the same time that the expansion of the universe started to accelerate.

A similar effect may be responsible for the fine tuning of the physical parameters. Or indeed, such a conservation of information may be exactly what gives rise to the emergence of particles in the first place. I think it would be fair to say that there can be no information contained anywhere outside the universe. If no information can be lost from the universe as a whole, then there must be a conservation of information at work for the universe as a whole. This must have effects at the local scale if it has any meaning at all. For if it did not have any effects on local interactions, then the accumulation of local effects would not obey such a conservation of information rule. So this conservation rule would have to apply only to portions of the universe that are in causal relationship to each other. For if we lose all information about some portion of the universe, then that information cannot be accounted for so as to have local effects. If we lose information due to cosmological expansion, then structures must arise to compensate for that loss. I suspect that there is information contained in the curvature of spacetime. So that as the universe expands and flattens out, information is lost and particles must arise to compensate or that lose.

"Earth and sky fled from his presence and there was no place for them... Then I saw a new heaven and a new earth..."

 

[marcus]

the discussion at Peter's has involved posts from Lee Smolin, Frank Wilczek, Anthony Aguirre, and Max Tegmark. earlier I quoted a passage from a Tegmark et al paper where completeness required mentioning CNS as one explanation in response to the perception of fine-tuned constants, but where the authors neglected to discuss this possible explanation.

so far the discussion seems to indicate that CNS has not been ruled out, since cogent arguments to justify ignoring it were not forthcoming. but we will just have to wait and see. here is the passage from the recent Tegmark et al paper:

http://arxiv.org/astro-ph/0511774
Dimensionless constants, cosmology and other dark matters
Max Tegmark, Anthony Aguirre, Martin J Rees, Frank Wilczek

---quote from Tegmark et al page 1, Introduction---

B. The origin of the dimensionless numbers

So why do we observe these 31 parameters to have the particular values listed in Table1? Interest in that question has grown with the gradual realization that some of these parameters appear fine-tuned for life, in the sense that small relative changes to their values would result in dramatic qualitative changes that could preclude intelligent life, and hence the very possibility of reflective observation. As discussed extensively elsewhere [9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22], there are four common responses to this realization:

1. Fluke: Any apparent fine-tuning is a fluke and is best ignored.

2. Multiverse: These parameters vary across an ensemble of physically realized and (for all practical purposes) parallel universes, and we find ourselves in one where life is possible.

3. Design: Our universe is somehow created or simulated with parameters chosen to allow life.

4. Fecundity: There is no fine-tuning, because intelligent life of some form will emerge under extremely varied circumstances.

Options 1, 2, and 4 tend to be preferred by physicists, with recent developments in inflation and high-energy theory giving new popularity to option 2.

Like relativity theory and quantum mechanics, the theory of inflation has not only solved old problems, but also widened our intellectual horizons, arguably deepening our understanding of the nature of physical reality. First of all, inflation is generically eternal [25, 26, 27, 28, 29, 30, 31], so that even though inflation has ended in the part of space that we inhabit, it still continues elsewhere and will ultimately produce...

...More dramatically, a common feature of much string theory related model building is that there is a “landscape” of solutions, corresponding to spacetime configurations involving different values of both seemingly continuous parameters (Table 1) and discrete parameters...
---endquote---

I should mention that none of the references cited by Tegmark et al are to CNS papers, for instance here:
"...discussed extensively elsewhere [9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22], there are four..."
So one can see that the authors plainly did not intend to discuss CNS, for instance in their response 2. (Response 2 evokes an explanation or non-explanation for the relevant constants which is logically different from CNS. Their discussion of response 2. fails to apply to CNS.)

======quotes from Woit blog======
Who Says:
December 26th, 2005 at 6:26 pm
Peter,

let me be a bit more specific about where I think the omission occurred and then either you or Wilczek can judge if my objection is valid. In the Tegmark et al paper...
...

Peter, you say an obvious answer to why CNS was not mentioned is that
"His [Wilczek’s] paper with Tegmark is not a review of the various attempts to get scientific predictions out of anthropic arguments.” But I would reply that CNS is not an attempt to get scientific predictions out of anthropic arguments. The theory has no anthropic character at all—it doesn't refer to life or consciousness, or depend logically on their existence.

However the paper does review 4 possible ways of reacting to the “realization that some…parameters appear fine-tuned for life.”

I believe that until or unless CNS is refuted it should be included among the possible responses, and that it does not fit neatly into anyone of the four listed by the paper. I suppose that according to the CNS view it can be considered a fluke that conscious life can arise where evolution has tuned the constants for abundant black holes. And that in this view there is a single universe capable of branching where, as in the case of a stellar mass black hole, gravitational collapse leads to a bounce. So it is perhaps closest to responses 1 and 2—but not adequately represented by either!=====================
Anthony Aguirre Says:
December 26th, 2005 at 11:45 pm
To “who”:

I agree with Aaron Bergman that CNS is subsumed in ‘multiverse’, (and this is why I did not think it required separate listing either in the present paper or my earlier one with a similar list) though I also agree with you that, as I noted before, there is an aspect of ‘fluke; that the conditions for life must coincide with the conditions for black-hole formation (or else life all lives in rare universes and we are back to anthropics to explain why we do not observe the most common type of universe).

Lee:

For me, and I think for many, the argument goes:

1) observations imply inflation.
2) Inflation implies at least the possibility of eternal inflation.

breaking eather link would be extremely interesting. Breaking the first, by creating a viable alternative to inflation for explaining the CMB fluctuations, etc., would obviously be interesting, but has not, in my opinion, been even nearly done, with all due respect to the cyclic and VSL folks.

Breaking the second link, which you suggest, would also be interesting and I would be happy to entertain such a possibility — I just do not see how to do it. You may be right that somehow eternal inflation will ‘go away’ if we understand vacuum energy, but I’m not sure how this could happen without inflation itself going away.

There are some, I think, who entusiastically embrace the eternal inflation picture. But I think many others others, like myself, find it interesting that observations seem imply it, perhaps despite our wishes.

cheers,

Anthony===================
Lee Smolin Says:
December 27th, 2005 at 1:10 pm
Dear Anthony,

Thanks, I am aware of this point of view but I haven’t found the literature completely convincing. Perhaps you could tell me the paper where the argument “inflation implies eternal inflation” is most clearly and convincingly presented, and I’ll study it.

Worry about the realness of the vacuum energy is not the only issue I’ve had with this argument. Another is how well defined are the frameworks in which calculations are done. Some are very heuristic, others depend on assumptions about the interpretation of the wave function of the universe and measures on infinite numbers of universes that don’t seem to make sense when scruitinized.

Another kind of worry is that extending from inflation to internal inflation requires believing that the mechanism is reliable at scales presently outside our horizon that were formerly way below the Planck scale. If there is a universal Planck scale cut off, as in DSR, it could alter the physics presently outside of our horizon. In fact, there are apparent anomalies in the CMB spectra near present Hubble scales-both the low power and the axis of evil. If real they support the idea that Planck scale effects have a non-trivial effect on inflation, which would change things outside our present horizon.

On the other hand, we know our universe produces large numbers of astrophsyical black holes, a fact that depends on features of star formation and galactic dynamics that do not seem otherwise necessary.

Thanks,

Lee

ps to Who, thanks for pressing the argument.

=============

 

[marcus]

Anthony Aguirre replied:
========quote from Woit blog==========
Anthony Aguirre Says:
December 27th, 2005 at 3:06 pm
Hi Lee,

Absolutely agree that many of the arguments for eternal ’stochastic’ inflation (driven by quantum fluctuations) are somewhat unrigorous, even at the level of effective field theory in curved spacetime. I find them reasonably compelling but consider it possible that there is something fundamentally wrong with them. The argument for eternal inflation with multiple minima, however, seems extremely strong to me — the spacetime can be described more-or-less exactly, and a very straightforward computation implies that the (physical) inflating volume increases when spatial sections are chosen so as to make the background spacetime homogeneous. One must only accept the Coleman-De Luccia picture of the decay of the false vacuum for this to be true for at least some multiple-well potentials.

Now, the questions of how *generic* eternal inflation is, is I think an interesting one: are there non-fine-tuned inflation models that explain the observations and are not eternal? I don’t know of any decent study on this question (and may well undertake one myself). In the context of CNS, however, even the *possibility* of eternal inflation is very dangerous, however, because unless eternal inflation is *forbidden* by the ‘meta’ laws that govern which types of inflation are possible in the landscape of possibilities, it must only be realized once to ‘take over’ the ensemble by creating an infinite number of black holes out of one. The only escape from this, it would seem, would be to have other channels to create infinitely many black holes from one ‘parent’ — but then you would be in the same extremely thorny boat of comparing infinities that eternal inflation is already in; further, which type of universe would come out ‘winning’ in this competition between infinities would seem quite independent of mundania such as the IMF…

One interesting way that EI might be essentially wrong is if the QFT in curved spacetime description is essentially wrong, for example for ‘holographic’ reasons (see the recent Banks paper). In this view, all of the zillions of ‘other universes’ are just refigurings of the same degrees of freedom inside the horizon (this realizes, in some sense, a version of the ‘hall of mirrors’ mentioned by ‘dissident’ above.) I find this view very hard to understand, but perhaps it is right and would change the way we think about EI, perhaps in a way less troubling (to me at least) in regard to CNS.

cheers,

Anthony
====endquote====

 

[Chronos]

could you give me the link, please?
my impression is that the size of the errorbar on a neutron star mass measurment can allow for some wiggleroom. It might take several observations. I would very much like to see your most recent find.

You are correct in noting previous candidates for over weight neutron stars have have not been statistically signficant. The first serious challenger emerged a few months ago. The CNS 1.6 solar mass limit was ruled out up to the 2 sigma limit in this paper:

http://www.arxiv.org/abs/astro-ph/0508050
Title: A 2.1 Solar Mass Pulsar Measured by Relativistic Orbital Decay
Authors: David J. Nice, Eric M. Splaver (Princeton), Ingrid H. Stairs (UBC), Oliver Loehmer, Axel Jessner (MPIfR), Michael Kramer (Jodrell Bank), James M. Cordes (Cornell)

ABSTRACT: PSR J0751+1807 is a millisecond pulsar in a circular 6 hr binary system with a helium white dwarf secondary. Through high precision pulse timing measurements with the Arecibo and Effelsberg radio telescopes, we have detected the decay of its orbit due to emission of gravitational radiation . . . . Interpreted in the context of general relativity, and combined with measurement of Shapiro delay, it implies a pulsar mass of 2.1+-0.2 solar masses, the most massive pulsar measured.

I've been following the citations to this paper, but, no one has challenged this finding. There was, however, some interesting discussion in a recent paper co-authored by Brown [one of the sources for the CNS mass limit prediction] :

http://www.arxiv.org/abs/astro-ph/0510380 :
Title: Double Neutron Star Binaries: Implications for LIGO
Authors: Chang-Hwan Lee, Gerald E. Brown

Of course PSR J0751+1807 is but a single example, which is not altogether satisfactory in my book - where there is one, there should be others. Perhaps LIGO will be able to resolve the issue.

 

[marcus]

thanks. this is definitely one to watch! good on you for flagging it!

...
http://www.arxiv.org/abs/astro-ph/0508050
Title: A 2.1 Solar Mass Pulsar Measured by Relativistic Orbital Decay
Authors: David J. Nice, Eric M. Splaver (Princeton), Ingrid H. Stairs (UBC), Oliver Loehmer, Axel Jessner (MPIfR), Michael Kramer (Jodrell Bank), James M. Cordes (Cornell)
ABSTRACT: PSR J0751+1807 is a millisecond pulsar in a circular 6 hr binary system with a helium white dwarf secondary. Through high precision pulse timing measurements with the Arecibo and Effelsberg radio telescopes, we have detected the decay of its orbit due to emission of gravitational radiation . . . . Interpreted in the context of general relativity, and combined with measurement of Shapiro delay, it implies a pulsar mass of 2.1+-0.2 solar masses, the most massive pulsar measured.
...

besides the Chang-Hwan Lee, Gerald E. Brown
paper you mentioned
http://www.arxiv.org/abs/astro-ph/0510380
there ishttp://www.arxiv.org/abs/astro-ph/0510379
a related paper by the same two authors plus the late Hans Bethe

if this 2.1 figure is confirmed, they point out, it plays havoc with their model of NS binaries. Here is a quote from page 21.

"...J0751+1807 : We consider the measurement of a 2.1M⊙ neutron star mass in this neutron star white dwarf binary a serious challenge to our maximum neutron star mass (Nice et al. 2005). It will be clear in our Section 8 that in the evolution of neutron-star, white dwarf binaries, sufficient mass is furnished during the red giant evolution of the white dwarf progenitor, often in conservative mass transfer so that if accepted by the neutron star, then most of them would have masses in the vicinity of the quoted mass in J0751+1807 or higher, as found by Tauris and Savonije (1999). These authors did not introduce the propeller effect, whereas Francischelli et al. (2002) found that in evolution of double neutron star binaries this effect often cuts the accretion down by an order of magnitude. We have given our reasons earlier 4 that the maximum neutron star mass cannot be far above 1.5M⊙. We should mention that the existence of a 2.1M⊙ neutron star would wreak havoc with our scenario ..."

 

[Chronos]

Good find, marcus. I expected more discussion than what I found, and you found it. Call me lazy. I only read the latest paper by Lee, et al. But, even assuming obese NS's are confirmed, is it really that big a problem for CNS? It looks to me like you can raise the bar on the mass limit without unduly harming the CNS conjecture. Besides that, it resolves some issues I have with the mass gap between neutron stars and black holes. I'm still hoping Dr. Smolin will comment on this.

 

[selfAdjoint]

...plus the late Hans Bethe

Awe-inspiring, out on the arxiv in October of this year! He died in March at the age of 98. Can it be his very last paper?