Darwinism, Reproduction and QM

apeiron

Do you have some references regarding this failure? My impression is that it has become the mainstream approach over the past decade.

While I would agree that the detailed machinery offered by Zurek and others is too clunky to be satisfactory, the general idea of the Universe as a system that is decohering events over random spatiotemporal scale seems the right one. It puts the "observer" everywhere and nowhere in the system.

WaveJumper

Decoherence doesn't address the issue of what happens to the other states when a single outcome is selected through loss of information to the environment and it is presumed those waves are real(in most formulations). MWI is too much of a baggage with its trillions of universes, i reject it on Occam's razor grounds.

"Leak of information" is a rather weak explanation as to why a certain single eigenstate is preferred. It is in fact no explanation at all for classical reality, but just a mechanism that might be useful in a future theory with greater explanatory powers. The only way decoherence makes sense as it is, is when coupled with MWI where every probability is realised in a different universe. But i am not convinced that believing in ghosts, the Loch Ness monster and blood sucking aliens is not easier to swallow. I can imagine Einstein asking Pais:

"Do you really think a new 72 billion light years across universe is created everytime a dung beetle moves its antennae?"

Freeman Dyson

-Penrose

Has anybody heard of quantum darwinism?

http://www.scientificblogging.com/ne...ntum_darwinism

apeiron

I thought you were referring to Zurek who coined the term....

Quantum Darwinism is a theory explaining the emergence of the classical world from the quantum world as due to a process of Darwinian natural selection. It is proposed by Wojciech Zurek and a group of collaborators including Ollivier, Poulin, Paz and Blume-Kohout.
http://en.wikipedia.org/wiki/Quantum_Darwinism

Penrose was instead exploring the idea that gravity was responsible for the collapse.

http://en.wikipedia.org/wiki/Penrose_interpretation

One could be called a subjective approach, a collapse of the wavefunction from the outside by the "observation" of a system.

The other is objective in that the collapse happens due to what is going on "inside" the wavefunction itself.

apeiron

Wavejumper, you are putting your finger on what I find "too clunky" about decoherence modelling to date. But I don't see MWI as the fix at all.

The way I view it is that the classical universe is a "QM uncertainty dissipating structure" (so I am expecting an even more thermodynamic slant to the eventual interpretation machinery).

The universe is a system of constraints. At this constraint is felt with increasing definite effect as scale grows. So while the constraint or "collapsing observation" is weak near the planckscale, it would grow with powerlaw vigour with physical scale. Or perhaps exponential. And so in our classical realm, QM uncertainty is decohered on a very fine grain generally, but also potentially can have fractally large scale - escape decoherence for quite a while.

An analogy would be an ideal gas. Take a box with a bunch of gas particles at gaussian equilibrium. Insert a much hotter or colder particle and quite quickly it will be "decohered" to the ambient average state. The idea gas is a system of constraint that can't necessarily get you right away, but will get you on some emergent average scale.

That would be crazy. As there is no evidence to suggest it happens.

But just about as crazy is what must be true when you look into the night sky, and see a distant star. QM tells us that there is a dance, a collapse over a sum of histories, between some excited stellar atom and the photoreceptor in your eyeball. To account for known QM effects, this has to be a retrocausal link with a "nonlocal" aspect.

So classical reality is probably best viewed (via systems decoherence, rather than environmental decoherence, perhaps) as a mesh of such interactions, such collapse events. Some of the events have vast scale (point-to-point event across millions of lightyears). But statistically, the great majority of events are quick and local collapses. Two atoms in the star are far more likely to make that photon connection.

Maybe even virtual particle interactions stabilise classical reality before things get that far, creating the true baseline? That seems to be where some theories like Wilczek's condensates are pointing.

Anyway, the natural commonsense expectation of a sound interpretation of QM would seem to revolve around the idea of one generally classical universe which becomes a sort of homogenising, wavefunction dissipating, equilibrium structure. There is no observer as such as observation, or constraint, is present everywhere. The history of where the universe has been becomes the general shape of what can happen next. Then QM uncertainty intrudes on the fine grain to make things a bit unpredictable and creative - in a predictable average sort of way.

Well, my expectation that this is a natural approach is probably due to my familiarity with systems science in biology and neuroscience where this kind of anticipatory or forward-modelling logic is the norm.

Moonbear

I haven't read anything else in the thread, but you have a seriously flawed premise here. Nature has no such tendency toward complexity, and that would have nothing to do with evolution. Evolution can and does happen in any direction.

apeiron

Of course there is a trend to complexity in nature. This is what dissipative structure theory, maximum entropy production principle, entropy degrader approaches, and other stuff is all about.

Order exists because it accelerates disorder. That is what life and mind are all about.

What you perhaps mean is that evolution itself - the darwinian selection mechanism - is essentially uncreative and homeostatic. And modern theoretical biology would agree. That is why they split biological systems in to evo and devo.

Development is about the self-organisation into complex dissipative structures. Then evolution is about the constraints exerted by an environment that limit the possibilities.

So a tree could sprout limbs and leaves in many branching patterns. But all sorts of accidents of circumstance in an actual forest limits it to some actual pattern of branching.

The tree - viewed as devo - wants to be as complex as possible in its branching to dissipate as much as it can. Then evo forces may knock off branches, chew its leaves, shade it out, starve its roots - act in blind and undirected fashion.

Coldcall

Apeiron,

The evidence regarding the failure of "decoherence" to address the measurement problem is in the public domain. Even Zurek, one of the founders of "decoherence" admits its not a solution to the measurement problem.

And if you think it does solve the MP then you have been misled in a big way.

Coldcall

Apeiron has answered your post well. In fact i find it laughable you actually made this post because clearly you have never read any studies on natural complexity, natural self-organisation and the wider areas of chaos theory.

I suggest you go to the library.

WaveJumper

Yes, there doesn't appear to be a way to resolve the measurement problem without admitting that our conscious activities(also referred to as measurements/observations) 'collapse' wavefunctions to 'particles'. Most of the interpretational problems of QM start and end at the double slit.

Quantum entanglement and Bell's conclusion that if QM is right, reality cannot be both local and realistic is very anti-realistic. "Local realistic" is the perception of the average Joe on the street as to how the world is. This view, however, isn't supported by experiements. I've yet to see a sensible explanation of what a non-local but realistic world is supposed to be; this cannot be a feature of a purely materialistic world.

apeiron

My own feeling was that Zurek's early papers (along with what others like Gell Mann were saying) was creating the right framework, but then along the line the thinking became contorted trying to cash out new formal machinery.

For me, the big problem is always in anchoring observation back to a static, located, human observer when reality is dynamic and self-organising (I presume from a systems science standpoint). So observers and their measurements have to be generalised in that direction, taking the nod from global boundary constraints thinking.

So in this view, decoherence would be about expanding light cones of QM potential. When spatiotemporal scale is still small, the potential has little context and so is less likely to encounter some crisp collapsing context. But as scale grows, it become rapidly more likely that collapse will occur.

This is hard to explain unless you can think about QM potential as a vagueness. There is not even a wavefunction crisply existent until the scale, the field of view, has grown enough to take in, say, a pair of particles who could frame some definite exchange.

It is a phase transition view I guess. When scale is small, you may have in effect a particle surrounded by a vague QM potential to "do something". The particle's gravity, EM, give it a QM potential or "presence" that propagates as a spherical boundary moving at speed of light. But it is a very raw QM state - like a chaotic jostle of dipoles in a hot bar magnet.

Then the scale grows large enough so that a second particle comes within exchange range. At that point, a crisp wavefunction can exist. There is a global boundary condition that can create constraint of that vaguer potential. General limits to what can happen are created and then something does happen. It is like the sufficient cooling that allows a crisply divided local~global state of order in a magnet.

In effect, the wavefunction and its collapse are two faces of the same thing. The wavefunction was not "always there and evolving" in an independent sense. Instead there was a rawer potential for somethingness developing, then a crisp QM wavefunction/crisp classical collapse did something with that spreading potential. We only impute an evolving wavefunction after the fact.

OK, I'm thinking aloud here as this was the general picture I took from Zurek's early writings about decoherence, combined with what I was hearing at the time from quantum vagueness guys like Chibeni (that stuff seems to have died a death sadly), and Cramer's convincing arguments for retrocausality. Plus, as I say, what seems obvious from a phase transition, systems science, way of looking at reality.

Zurek seems to be working in the right area on this....
http://arxiv.org/PS_cache/cond-mat/p.../0701768v2.pdf

But I think the key thing missing is the idea that QM information starts vague and needs a classical context to turn it into crisp QM probabilities, even if the crisp QM probabilities are still of the crisply entangled, uncertain and superimposed probabilities on the wavefunction side of things.

Boiling it down, the usual framing of the measurement problem is that we have an evolving wavefunction forever in search of the machinery that forces its collapse. The difficulty in seeing why the wavefunction should collapse (because no internal mechanism or hidden variables are permitted) leads people to say collapse requires consciousness, or perhaps in many worlds fashion, never happens.

Decoherence is broadly the attempt to put the collapse machinery back out there in the physical world. And really it would be good to have it happening as a global boundary constraint - that is, something that is presence and active over all classical spatiotemporal scales. (technical note: global means thermodynamic macrostate rather than "largest size").

Then what I take this to require is that the collapse machinery in fact manufactures the wavefunctions out of rawer QM potential. So it is the collapse that causes the wavefunctions, not the wavefunctions and that must produce a collapse.

Freeman Dyson

I will throw another thing out there that I have been interested in over the years:

Synchronicity

That events are tied together by purpose. By meaning. That is the connecting principle. Only meaningful and purposeful things happen. Classical objects form from quantum states because they are meaningful/purposeful objects.

Kind of out there I know..

disregardthat

Science does to a certain degree assume a realistic perspective, but my opinion is that this is only because of the context in which scientific predicates usually are understood. An important insight is that scientific predicates only says something about our perception, and the scientific body of knowledge represents the structure of perception. Realism posits that the world is mind-independent, and that scientific predicates makes sense in the absence of a mind understanding it and giving it meaning. But this is a senseless view in my opinion. The world independent of the mind can not be said anything about. It makes no sense to make a distinction between objects in a mind-independent world, it makes much less sense to talk about properties of objects, not even spatial and temporal properties. These are all conceptual characteristics. We see these characteristics of objects because through perception our minds interpret sensations with spatial and temporal structure. I find the anti-realistic view much more appealing. It also rid itself of many problems the realist view stumble into.

The results of QM are very interesting, and we couldn't find a better tool to propagate these points through with. My knowledge of QM is however limited, but it is my understanding that objects under observation more or less behaves as we would expect. Perhaps you, apeiron, are more knowledgeable about Kant's metaphysical theories about the necessities for experience than me and hopefully you have a comment on this. I find it incredibly interesting how Kant's points are manifested through the results in quantum mechanics.

apeiron

Experiment has already put tight parameters around such connectedness. So while QM supports nonlocal (making the global context meaningful and not just an a-causal void) it also limits the nature of the connection in very strict fashion.

You could argue that the psi research literature does the same from a different angle. If any kind of spooky stuff exists in a mind-entangling complex way, the signal is so small as to be swamped by experimental artifact and experimenter fraud.

So yes. Any theory can be entertained. But synchronicity in any Jungian sense has a stack of negative findings against it now.

apeiron

The realist vs anti-realist choice, as you put it, does offer also a third path other than a binary either/or.

You can instead say we need to model in terms that include both - both modeller and modelled, observer and observed.

Which is the essence of what Pattee, Rosen, Salthe and others in the semiotic, systems science, camp would be doing.

disregardthat

I don't think that the anti-realist is rejecting realism on behalf of his own perspective, that would be violating the very "ideology". The realistic perspective is a perfectly "valid" perspective in the anti-realistic sense as it is a coherent system of beliefs as any. But I agree with you if I understand you correctly. But instead of talking about a "third way", we can remove the clear-cut distinction between realism and anti-realism. In a way, anti-realism incorporates both perspectives. The reason I find anti-realism appealing is the rejection of subscribing to any specific set of beliefs or any specific ideology.

apeiron

We would agree then that "all is modelling". We start from a position of subjectivity and have to operate on that basis. The question then becomes how to we move towards the impossible ideal of an "objective" understanding.

The third way I'm talking about is very much concerned with just how to do this properly. And a central issue is how to make the epistemic cut, how to make a separation between observer and observed, given that we are stuck in a position of subjectivity.

Mainstream physics does just jump to realism. Or rather, being based on a positive, pragmatist, epistemology, it agrees all is modelling, and just models the observables. The observer is placed outside the description.

Well, with GR, the observer became part of the model to an important extent. With QM, the role of observers was made both crucial and obscure.

I guess a lot of people imagine that a ToE would do away with the need for observers perhaps. Their partial inclusion in physical models is an embarrassment and the urge is to find deeper theories that are just about naked observables. Meaningless information.

So the alternative would be to instead get observers and meaning-making into a ToE.