Smolin's Cosmological Natural Selection: How does it help the problem?

[hadsed]

Alright, admittedly I'm a noob at LQG, but I have a basic question. The cosmological natural selection, as I understand it, is that universes are born from singularities (or something close) inside of black holes and 'bounce', inflate, and become universes like our own.

But what spawned this madness? You get into an infinite regression; there had to be an original universe, right? Is there any sort of explanation for this? I understand it's just a theory at the moment, but why pursue it if you get stuck into another 'rut' and get questions like this, similar to the reason why people tried to get out of Big Bang models (what came before the big bang type questions)?

 

[marcus]

there had to be an original universe, right?...

Does it seem intuitively obvious to you that there must have been an original or "first"?

In any case Smolin's CNS is a separate theory from LQG. It does not follow from LQG and you can have CNS with some other QG besides Loop. The ideas are independent from each other.

CNS is a conjecture, and a major point about it is that it is testable. You can falsify it by astronomical observations.
Another point is that if CNS were true (I think it cannot be true because a recent astronomical observation seems to go against it, but suppose it were true) then it would partly explain why the 30-some numbers that plug into the standard particle model are what they are.

Basic physics is determined by around 30 numbers (the "dimensionless parameters of the standard model" including cosmology).
These numbers come out the same whatever system of units you use. They are the essential physical constants.

The CNS conjecture is that these 30 numbers are locally optimal for astrophysical black hole production. That you can't find one that you can make a small change to that would get us more black holes. The CNS conjecture says that any small change you make in any of the basic physics numbers will not cause more, and will probably cause less black holes. The numbers are optimal, at least for small changes. Like being on a hilltop.

Why the 30 basic numbers of physics are what they are is a big puzzle. If anyone could get a handle on that they would get Nobel. It's a big big problem. But whatever explanation has to be EMPIRICAL. It has to be testable. Make some prediction that could be falsified if the explanation isn't right. Now CNS tried to do this. But unfortunately I think it has been shown to be the wrong explanation! A neutron star of about twice the mass of the sun was observed.

 

[Fra]

Alright, admittedly I'm a noob at LQG, but I have a basic question. The cosmological natural selection, as I understand it, is that universes are born from singularities (or something close) inside of black holes and 'bounce', inflate, and become universes like our own.

But what spawned this madness? You get into an infinite regression; there had to be an original universe, right? Is there any sort of explanation for this? I understand it's just a theory at the moment, but why pursue it if you get stuck into another 'rut' and get questions like this, similar to the reason why people tried to get out of Big Bang models (what came before the big bang type questions)?

Like Marcus said, Smolins ideas of evolution of law have nothing to do with LQG.

As an advocate of some of these ideas, I think one should also point out the difference between smolins more specific CNS idea, and the underlying more general (but less specific) idea that laws do evolve.

The idea that laws evolve aren't madness, it's a position you can arrive at by analysing and considering a number facts of nature, science and the history of science. I think one should understand the underlying motivation that lead smolin to form CNS and a falsifiable prediction, rather then just looking at the surface. There are several talks that elaborate on the logic behind this. Some points I think are better expressed by Unger.

"On the reality of time and the evolution of laws"
-- http://pirsa.org/08100049/

"Laws and time in cosmology"
-- http://pirsa.org/10050053/

As I see it, if you take the accurate perspective that science and information processing are done by observers, scientists which are in effect physical systems. There is no good reason to expect that the action of complex systems follow a LOGIC different than smaller systems. In this perspective, it's the knowledge about law, or EXPECTATION about laws, that does rule our actions, not any REALIST level of law! (Here social laws are a good analogy on which Unger is an expert; but unfortunately it takes some imagination to see how this translates to the physics, so I think it's easy to miss the points). Therfore the actual laws we have are nothing but expectations, the notion of absolute timeless fixed laws are just old realist-type fantasies; this things doesn't fit in a proper inference perspective. No finite observer can in finite time establish confidence in such things. Therfore ongoing evoltion, and "infinite regression" IS how nature most probably does work if you take this view.

So our problem is not to unravel the eternally true timeless laws, it's to understand the logic and mechanism of how "laws" or more correctly "expectation of laws" does manifest, and evolve in nature, and how these things are encoded in terms of physical systems.

These are deep ideas, and good ones, but they are also difficult to implement. But the main problem is not so much people seems to work on it, for some reason.

This is actully the next radical step away from realism. Anyone that think we lost realism with QM, will get another shock once you accept Ungers poitns above, because QM still have a lot of realism implicit in the background structure, or equivalence classes thereof. Once version of this argument aims to also remove structural realism that suggest that not even laws of nature are objective and fixed.

However, here in the last point, I think Smolin himself isn't radical enough. To secure that it seems that tha laws of physics as we know them have been the same throughout the universe, he thinks the only way to have them evolve is when the universe are born, here comes his CNS hypothesis etc.

However I think there is a different way to implement these ideas AND still to secure the de facto effective objectivity of law we konw of so far. Therefor though I like Smolin and Ungers, I was never a big fan of his specific CNS, but the basic logic of the evolving law is still good.

/Fredrik

 

[Fra]

why pursue it if you get stuck into another 'rut' and get questions like this, similar to the reason why people tried to get out of Big Bang models (what came before the big bang type questions)?

If you mean that we instead will end up with a set of meta laws that describe HOW the laws evolve?

My personal idea of this is that there will be not meta laws at all in the sense of realist laws, instead the "meta law" or whatever gives an expectation of evolution only to a limited extent (part of this is of course intrinsitcally undecidable) will be given by optimala inference! There will exist, for a given observing system an optimally rational expectation of the total evolution. Expected evolution, is something different than actual evolution, but the conjecture is that any system acts in consistency (to the decidable extent) with it's expectations.

This yields us a system of interacting and evolvling "inference systems", which correspond to observers, and an evolution will prefer a certain population of these, which specific properties - ideally here we need to find the replacement for or explanation for the effective Standard model of particle physics we have; which will be at the low complexity scale of observers.

As I see it, the answer to your question is that when we scale the complexity of observer down, to reach the unification limit, there is no observer around to ask these questions. Ie. the inside observers simply aren't able to encode and phrase these objections. It would still be a major boost if we could understand how the law emerges in this way. Most probably it could also help explain several parameters that today are just empirical.

But no model will be able to escape from some of the fundamental undecidability that we are all constrained to. This is again why optimal inference is the best we can do.

/Fredrik

 

[Naty1]

but why pursue it if you get stuck into another 'rut'

The most basic reason is that a new theory is never complete, never the theory of everything...so some sort of "rut" is usually present. Something is always missing...some question is left unanswered. Scientisits have to start somewhere...Einstein started wihout gravity in special relativity, then "graduated"...

...universes are born from singularities (or something close) inside of black holes and 'bounce', inflate,...But what spawned this madness?

well, think about it...There are only two choices...neither of which is scientifically appealing: either something, such as a universe, started from 'nothing' or there is an eternally cycling phenomena thru which universe have always been evolving ...starting from 'something'...maybe just one, maybe an infinite number of them...

Perhaps the worst case would be to discover that cosmic censorship is real and that what is really fundamental and ultimately most fascinating will always be hidden from us behind a horizon of one sort or another...

 

[hadsed]

Marcus, I read your thread on the neutron star after I posted this, very interesting. I kind of liked CNS a bit because it seemed to make sense, but that's science I suppose.

Fredrik, if I'm understanding you correctly, are you saying that asking these sorts of questions ends up being nonsensical?

Naty I was always under the assumption that the reason people started to shy away from the BBT was because it got us stuck in a situation where we couldn't explain 'before the BB'. I think I see your point though, now that I think of it I don't see any other way except for cyclical or BB-like.. the latter seems more unsettling to me though. I agree that it would be a terrible thing to never be able to know for sure.

I suppose my naivety in the field shows, and I probably shouldn't have used the word 'madness'. I've hardly begun the maths to understand LQG entirely, so I'd always assumed CNS came out of that. Oops. I suppose I should have put a little more thought into this before going wild and making a thread. The points you guys make seem to be obvious now. Thank you for the articles.

 

[marcus]

...so I'd always assumed CNS came out of that. ...

Well there can be different kinds of "coming out of" that do not necessarily involve a rigorous mathematical derivation.

You know of John Archibald Wheeler, who was Feynman's PhD thesis advisor and co-authored some with him. Wheeler was an early person to think about quantizing GR. The Wheeler-DeWitt equation is an early quantum gravity equation. People back then---1970?---already imagined that a quantum version of GR might get rid of the singularity and that a bounce might occur. So Wheeler actually conjectured that there might be black hole bounce and that it might result in a big bang out the bottom, so to speak---another expanding region of spacetime, not intersecting with ours.

But the Wheeler-DeWitt model still suffered from a singularity. It failed to resolve the singularity. So all that was still pure conjecture---guessing that some QG might eventually resolve the singularity.

Smolin had been postdoc at Princeton IAS and was exposed to Wheeler (a senior guy there.)
I imagine that what Smolin did was around 1992-1993 he realized that string theory was going to run into a landscape problem----having millions of variants of physics instead of an unique solution. There was already or soon would be a paper by Andy Strominger (a top guy) warning of this (already in mid 1990s, people just didn't pay attention to the impending landscape until later.)

So Smolin realized that the laws of physics, or at least the constants, could CHANGE going through a black hole, during the momentary quantum confusion where all structure was uncertain. This is very much at the level of conjecture (say some kind of string theory were true, and say that QG really does resolve the singularity into a bounce...etc...)

So Smolin put Wheeler's idea (of black holes making new tracts of spacetime) together with the idea that the new tracts could have slightly different physics. And he suddenly had a Darwinian situation that could explain tuning and make predictions about the tuning (what it is optimized for, if it is in fact optimized). The big decisive point was that it was FALSIFIABLE which a lot of multiverse fantasies are not---being falsifiable it could be considered as real science. Smolin is the one who took that step. Wheeler's original conjecture did not have the testability feature.

So in a fuzzy heuristic way you can say that CNS DOES come out of the quest for quantum gravity. And Smolin working on LQG at the same time surely had something to do with motivation and inspiration.

That's how I would reconstruct it. Smolin may give an account somewhere.

 

[Fra]

Fredrik, if I'm understanding you correctly, are you saying that asking these sorts of questions ends up being nonsensical?

What I meant is rather that such questions are not physical, and arise due to mixing contexts.

The questions we ask here, today, needs to be somehow "translated" back to the big bang if we imagine such things, and then one will realize that the context for formulating questions back then is far more constrained as there were no complex observers and in particular the state spaces is smaller, as seen by an inside observer (to be understood as subsystem, say the first systems of space-confined energy).

In this scaling or translation process, a complex observer can distinguish more questions than a simple observer can. Thus when a complex observer tries to project his questions to a simpler system the projected question is non-physical.

It's a bit analogous to a human wondering why dogs aren't bored to death, when they just sleep and eat. The resolution is that a dog is simply too stupid to formulate these concerns. So the whole paradox likes in in appropriate projections. Simpler organisms are characterized by more simple states.

Similarly, the most basic systems formed during some big bang event (or created during extreme high energy domains) would simply be UNABLE to distiguish - and thus respond to - differences between the 4 forces. It probalby takes a certain complexity of systems (dropping the energy) to be able to distinguish and thus act according to complexity in the environment.

So to rephrase again, the point is that once the questions are properly "scaled" they in fact dissapear.

Edit: I'm don't mean to imply that dogs can't think, i just used them as an example. But most would agree that human brain capacity is superior to a dogs.

/Fredrik