Cosmological natural selection

[bascule]

Ever since I read about Lee Smolin's fecund universes theory in Richard Dawkins' book The Ancestor's Tale I've been enthralled with the idea that our universe evolved by a process which shares a certain degree of similarity to the biological process of natural selection.

 

Unfortunately, I later discovered, at least in the context of Smolin's theory, that there is no selection mechanism. Rather, fecund universes work more like evolution prior to sexual reproduction, a diffuse cloud of self-replicating universes with no real selection mechanism.

 

Well, apparently I'm wrong: cosmological natural selection can involve a selection mechanism:

 

http://arxiv.org/abs/0706.3379

 

The paper's author says:

 

I just posted a paper on the arXiv [...] that amongst other things, implies that Cosmological natural selection may be the more appropriate term after all, since the true replicators, namely the `Space-Cells' really do undergo replication, variation and selection. Universes are just the `survival machines' for the Space-Cells

 

Neat stuff!

[foodchain]

I downloaded the pdf and plan to read more of it at a different point in relation to this post. I like this direction very much. It might be a bit difficult for tests in a great many areas, but testing I think would be possible simply via the fact it seems the universe and its laws such as the elements don’t seem to be immutable, not to mention that evolution does take place. It might also be nice to see a new set of math tools developed, or even refined.

[pcollins]

Neat stuff!

 

It's also a first draft that seems to build on a copy of Sean Carroll's GR book and some pop-sci stuff. This Dirk character spends a lot of time on definitions and analogies, little time in actually presenting formal arguments that come close to explaining, let alone justifying, his claims in the abstract.

[foodchain]

It's also a first draft that seems to build on a copy of Sean Carroll's GR book and some pop-sci stuff. This Dirk character spends a lot of time on definitions and analogies, little time in actually presenting formal arguments that come close to explaining, let alone justifying, his claims in the abstract.

 

Not to mention that its a first draft but really think about what its going on about really, it makes something like relativity seem very simple in comparison overall. The math tools to even define a hypothesis in such would I imagine be somewhat to drastically different the norm for most of physics. As far as sci-fi goes, well, people thought that about most of the major ideas that dominate physics currently.

[Martin]

... at least in the context of Smolin's theory, that there is no selection mechanism. Rather, fecund universes work more like evolution prior to sexual reproduction, a diffuse cloud of self-replicating universes with no real selection mechanism.

...

 

no "real" selection begs for some definitions

 

I think you may mean no competition for scarce resources.

 

I don't see why one should expect Smolin CNS to have all the features (including sex) that we observe in biological natural selection. the aim of the theory is to come up with a testable hypothesis, not to exhibit as perfect an analogy to biology as possible!

 

In the CNS picture many or most universes "self-destruct" genetically because they can't have any babies------like if Lambda is too big and expansion accelerates too fast, so galaxies and stars don't have time to condense. Or if matter is all unstable and decays before stars can form and collapse to black holes.

 

the competition is to see who can have the most babies-----those with good genes for reproduction will come to dominate the population (explosion) percent-wise.

 

it is definitely a NATURAL SELECTION MODEL, it is simply not Malthusian---there isn't the familar cap on population we expect to see because of limited resources. Dominance is defined PERCENTAGEWISE instead of in absolute numbers, because the total is always growing.

 

this growth in the number of universes has been conjectured to be related to the increase in entropy and the arrow of time---I'm not sure I care about that, but a lot of people were discussing it on cosmicvariance blog lately.

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

 

I can see that it is fun to add features to the CNS scenario so that you get MORE ANALOGIES with bio evo. You can add predators, sex, whatever.

 

But for me the focus is on testable hypothesis, with a side order of Occam.

 

and I think Smolin picture is SUFFICIENT to generate a testable hypothesis (in particular about masses of neutron stars but more generally) about the optimality of the present values of the 30-some dimensionless constants of particle physics and cosmology.

 

I could be wrong. You might be able to convince me that it is not sufficient to make a falsifiable prediction about reproductive optimality. But that is what I care about (not further analogies with bio.)

[bascule]

The main selection mechanism I'd see is universes incapable of producing black holes, or ones which lead to a lineage which is increasingly incapable of producing black holes.

 

Is that what this paper is suggesting?

 

To me a selection event occurs when an individual dies without reproducing.

[CPL.Luke]

and why pray tell would a blackhole createanother universe, or have anything to do with the universe forming.

[fredrik]

I read one of Smolins related papers but haven't read up on them in detail, but I also think the evolutionary thining needs to be taken seriously.

 

In my thinking, two obvious "scarce resources" in an evolutionary "learning" model is memory and processing power. The individual implementing the most efficient learning strategy that makes optimum use of given memory and processing capacity will dominate. But there is also another detail, which would be that optimum is still relative. Clearly, in a a world where special cases apply to the environment one can make a more efficient strategy, at the expense of loss in generality. A lifeform that makes a mistake (high risk) and run into a scenario where his model is not flexible enough to evole, it dies. In my thinking, "processing" is also part of the defining relation of time.

 

The last time I did a quick reading on Smolins alternative to the anthropic principle, my impression was that he is not even trying to develop the ideal of falsification mechanisms. Of course the anthropic principle is not the option, but I still feel smolins suggestion is not radical enough. But at least progress is made in the right direction. But perhaps I could read the papers more carefully. The first paper I read, didn't give me a major incentive to read more. And applying to the "scarce resources" of memory and processing power, it's impossible to ready every single published paper.

 

/Fredrik

[pcollins]

Not to mention that its a first draft but really think about what its going on about really, it makes something like relativity seem very simple in comparison overall. The math tools to even define a hypothesis in such would I imagine be somewhat to drastically different the norm for most of physics. As far as sci-fi goes, well, people thought that about most of the major ideas that dominate physics currently.

 

There is no new math here. In fact, Dirk's 'arguments' and 'analogies' are impenetrably wordy and suspiciously lacking in mathematical development of any kind. I seriously don't see anything of value here, and I hope that's because this first draft is supposed to qualitatively signal some upcoming research.

[foodchain]

There is no new math here. In fact, Dirk's 'arguments' and 'analogies' are impenetrably wordy and suspiciously lacking in mathematical development of any kind. I seriously don't see anything of value here, and I hope that's because this first draft is supposed to qualitatively signal some upcoming research.

 

 

I think that’s the point. As I study string theory or what not I think one of the reasons it falls apart is because its math is used far more then experimentation. Its not a fault of math, but the reality of it as I see it. I mean I was out reading on string theory, or debates back and fourth between smolin and another fellow and they were talking about negative dark matter or energy, and carrying over to change this. Now don’t get me wrong, but I hope they don’t mean carry over in a math sense to change reality in a theory.

 

Going from conservation laws alone, I have a hard time thinking that nothing existed prior to the big bang, another aspect is the relationship of matter and energy, or in what definite or only allowable forms of energy can exist, such as an electron for example.

 

Now this may not be an entirely new direction, but I don’t really know of any major thrusts of science to study various ways in which the universe and its physical laws or properties could have evolved. Every new major new theory in physics after sometime and pursuit by professionals ends up offering new equations and tools. Its not as if e=mc2 was just simply plucked from the aether one day either for example.

 

We do have man made elements, found anti particles, and happen to look for symmetry, supersymmetry, breaking of such, a standard model that relates something of an ecology really, or you had to have b to get to c from a for example.

 

Now not to say I know a lot about physics, I don’t. I know somewhat about biology, but nothing of any real professional level. What I do have though is a vast interest in it all, and I do spend a good deal of time reading about a diverse amount of subjects, and I don’t see why this should be contested, or any real hypothesis. I mean I object to string theory because it lacks in major ways to falsify, but I object to it only on that basis, not that it should be studied or pursued by interested people or scientific professionals.

[Martin]

To me a selection event occurs when an individual dies without reproducing.

 

in a population explosion, not limited by resource scarcity, one genotype can overwhelmingly dominate by having a a larger reproductive rate

 

another genotype can become a negligible demographic even if it can reproduce, simply by not reproducing as prolifically

[pcollins]

I think that’s the point. As I study string theory or what not I think one of the reasons it falls apart is because its math is used far more then experimentation.

 

The extent of math used in string theory is hardly the cause of its lack of verification. It struggles because of key predictions it makes are not testable at this time. The math, if anything, does nothing but allow physicists to make plentiful and more importantly less ambiguous predictions.

 

I think you mean to say that it's better to fit theory directly to experimental results rather than to try and infer it in the derivative manner theoreticians do. I have no dog in that fight, but I'll point out it has little to do with Dirk's paper. Everything he's describing is in the realm of theoretical physics. The absence of math in this context makes it extremely difficult to evaluate his proposals in any meaningful sense.

 

Its not a fault of math, but the reality of it as I see it.

 

That's where experimentation comes in.

 

I mean I was out reading on string theory, or debates back and fourth between smolin and another fellow and they were talking about negative dark matter or energy, and carrying over to change this. Now don’t get me wrong, but I hope they don’t mean carry over in a math sense to change reality in a theory.

 

I'd need more context about the debate to say anything, but I would hesitantly treat with great importance the correspondence between two theoreticians on a highly conjectural matter.

 

Going from conservation laws alone, I have a hard time thinking that nothing existed prior to the big bang, another aspect is the relationship of matter and energy, or in what definite or only allowable forms of energy can exist, such as an electron for example.

 

Yes, the fine-tuning problem which motivates this entire discussion. When somebody comes up with something that at least gives a hint that it's testable, then I'll be concerned about what if anything came before the Big Bang. As for Dirk's piece, I don't see how it at all illuminates a possible answer to the question.

 

Now not to say I know a lot about physics, I don’t. I know somewhat about biology, but nothing of any real professional level. What I do have though is a vast interest in it all, and I do spend a good deal of time reading about a diverse amount of subjects, and I don’t see why this should be contested, or any real hypothesis. I mean I object to string theory because it lacks in major ways to falsify, but I object to it only on that basis, not that it should be studied or pursued by interested people or scientific professionals.

 

I object to this paper because it doesn't say much that is obviously meaningful. Dirk spends a good chunk of 15 pages drawing analogies and constructing definitions. By the time you get to his argument, it's inaccessible. Maybe that's because it's a first draft, but the lack of math is really a stumbling block. It also triggers my bullshit warning light.

[Martin]

and why pray tell would a blackhole createanother universe, or have anything to do with the universe forming.

 

John Archibald Wheeler's idea.

grand old man of American physics

Richard Feynman's PhD thesis advisor

longtime figure at Institute for Advanced Studies at Princeton.

 

it was pretty daring when Wheeler proposed it circa 1970s IIRC

but as time goes on missing pieces of the puzzle keep surfacing.

 

basic idea is classical: singularities do not occur in nature, they are places where human theories break down. the way you fix a singularity is to improve the theory so it doesnt break

 

in 2001 and years following, BB singularity fixed and replaced by bounce

(work continues on this extending result to more and more general cases, and the new model must be empirically tested by CMB observation and other)

 

if the model checks out, then BB non-singularity was preceded by contraction phase.

 

from 2005 onwards, investigation into quantum model of BH, tentative results removing singularity there---time evolution continues at the BH pit---bounce occurs similar to the BB bounce---possible new spacetime region, re-expansion. these two bounces, if they occur, MAY or may not be connected. some similar mathematics

suggests possible connection.

 

inflation supplies most of the required energy (matter)----this is already a feature of prevailing inflation scenarios going back to Alan Guth ("universe is the ultimate free lunch")

 

it all still has to be checked by astronomical observation, and this is becoming an important scientific agenda----testing quantum gravity cosmological models.

 

there are career opportunities now for QG phenomenologist. it is an emergent field. I know someone who went into it recently and is doing very very well.

Also observational astronomers can publish, propose instruments, get funding etc. for things they couldnt a few years back.

 

QG is looking good and especially QC (quantum cosmology), but the connection between BB non-singularity and BH non-singularity might be WRONG. There might not be bounces, might be something else, and if there are bounces they might not connect etc etc. It all has to be worked out and checked and that is where the fun is.

[foodchain]

The extent of math used in string theory is hardly the cause of its lack of verification. It struggles because of key predictions it makes are not testable at this time. The math, if anything, does nothing but allow physicists to make plentiful and more importantly less ambiguous predictions.

 

I think you mean to say that it's better to fit theory directly to experimental results rather than to try and infer it in the derivative manner theoreticians do. I have no dog in that fight, but I'll point out it has little to do with Dirk's paper. Everything he's describing is in the realm of theoretical physics. The absence of math in this context makes it extremely difficult to evaluate his proposals in any meaningful sense.

 

 

 

That's where experimentation comes in.

 

 

 

I'd need more context about the debate to say anything, but I would hesitantly treat with great importance the correspondence between two theoreticians on a highly conjectural matter.

 

 

 

Yes, the fine-tuning problem which motivates this entire discussion. When somebody comes up with something that at least gives a hint that it's testable, then I'll be concerned about what if anything came before the Big Bang. As for Dirk's piece, I don't see how it at all illuminates a possible answer to the question.

 

 

 

I object to this paper because it doesn't say much that is obviously meaningful. Dirk spends a good chunk of 15 pages drawing analogies and constructing definitions. By the time you get to his argument, it's inaccessible. Maybe that's because it's a first draft, but the lack of math is really a stumbling block. It also triggers my bullshit warning light.

 

Sorry to not go quote by quote but I don’t think my post will stray to much.

 

What I mean about the math with string theory is a lot is possible in math, heck just about anything you want really. The problem with math is just simply being able to refine it via experimentation, to find out what’s real and what’s imaginary, that’s my only point really.

 

As to the context of the paper itself. It goes on about something in which the standard physical laws that govern the university might have come about. I don’t know if it goes into any degree of acceptable detail, but one thing is that study into such might lend a hand to a reverse engineering process of some type to give insight. For instance, I know a good deal of attention is paid towards the minor fraction of time after the big bang, because its plays a role in the development of the universe. Well, what about the stuff prior to those few seconds, how about a thousand years before that?

 

As to the more direct application of evolution, well, its just a word, or a title to describe something that takes place in the physical universe, and I would imagine by physical laws or such that evolution is guided by if anything. So if evolution can exist naturally, then why not attempt to use this as already offered by nature to examine nature and or reality. Evolution is studied basically by phylogeny of various forms, or tracing change, such is already evident in nucleosynthesis.

.

 

Again, as I am not totally sure as to the amount of time giving to the study of the universe in such a format, I mean I hear the term tossed around in terms of the evolution of galaxies, or elements, but not physics in general. Also, it does not seem as if physics, or what it studies is a bunch of unrelated phenomena, as a matter of fact I hear it a lot that everything is just energy, differentiated by form.

[Widdekind]

Yes, the fine-tuning problem which motivates this entire discussion...

 

'Creationists' often cite the complexity, of biological systems (e.g., human eye), as 'proof' of Divine Design. Yet, gradual evolution, through variation & selection, applied 'patiently & persistently', for long periods of time, can also account, for those complexities.

 

Now, 'Fine Tuning' sounds strikingly similar, of a like line of logic. For, some cite such a set of 'perfected parameters', as 'proof' of 'something special' having happened -- essentially, Divine Design. Yet, again, evolution, through variation & selection, could, equally again, account for Cosmic parameter 'perfectification'.

 

According to the DC documentary How the Universe Works, cosmo-genesis may be cyclical -- every Big Crunch begets an ensuing Big Bang. Such suggests, or sketches, the contours of a 'Cosmos-selection system':

 

Big Bang

Big Crunch

('unfit' universe 'dies')

'the stuff of space-time' is modified

, to some degree, by the fiery inferno

('clay re-thrown'; parameter values

varied

)

next

Big Bang

Such a scheme seems entirely 'natural' -- 'selection' by, 'variation' from, Big Crunch -- since most FT arguments allege, that 'typical' parameter value 'settings' would generate universes which were wildly unstable, suffering from infinitesimal lifetimes. Our Cosmos, then, would be one of the first 'long lived' universes, after an 'endless infinity of ephemeral fizzling failures'.

 

However, what would happen, if some universe 'failed' in a 'Big Stretch' scenario, expanding forever -- unlike 'natural' selection, such a scenario attributes infinite lifetime, to a 'failure'*...

*

Er go

, no Cosmological Constants ???

 

Wow, the author seems to suggest, that the space-time fabric, of our Cosmos, is a "dermal skin", composed of Planck-scale "skin cells", whose "replication & division" accounts for the observed expansion -- to wit, "growth" (!) -- of our Cosmic space-time. Spatially, such a structure is "cellular"; temporally, such a structure is "fibrous", the world-lines of the "skin cells" stranding together, like extruded spaghetti. Such seemingly says, that space-time itself is "living", a "living tissue membrane", in some informational-processing-sense -- an "H R Geiger" space-time:

 

[Widdekind]

First, the article suggests, that the 'space cells', comprising the fabric of our space-time, store internal information. Thus, in the 'space-time skin' analogy, 'space-time skin cells' have an internal 'DNA genetic code', by which they process (and store) information.

 

Second, if 'space cells' persist through time, then what would make our Cosmic Time discrete ? If 'space cells' are distributed, discretely through space, but persist through time -- so that space-time is 'cellular' spatially, and 'fibrous' temporally -- then perhaps Cosmic Time is continuous, as opposed to the three discrete spatial dimensions ??

[Widdekind]

Gardner & Shostak's Bicosm book is devoted to this topic.