Sic et Non self deconstructs

[_DrW]

The second law of thermodynamics will eventually have it's way. However, considering sets or assemblages of auto-catalytic polymers (proteins or RNA at the micro scale) as being "open" thermodynamic systems, taking in energy and materials from the environment, you will agree that they can ignore, or at least postpone, the ultimate consequences of the second law for a while - just as we are doing now.

The Second Law has its way all the time; biological processes don't postpone its consequences at all. Entropy isn't the bad guy. It's not even precisely "disorder". It increases in all spontaneous processes—whether the processes are ones that we find constructive or destructive. Entropy increases in death and decay; it also increases in birth and growth. If entropy didn't increase as life evolves we wouldn't be here. The idea that entropy has to be overcome by life is a creationist myth.

Your comments reflect the following passage from my biochemistry textbook by Albert Lehninger.

"-living organisms preserve their internal order by taking from their surroundings free energy, in the form of nutrients or sunlight, and returning to their surroundings an equal amount of energy as heat and entropy.

However, it is also true, as I stated, that in open thermodynamic systems at the micro scale, the entropy is decreased within sets of auto-catalytic polymers. This decrease is also found within living organisms.

Living organisms take in the energy they need to decrease their entropy, by eating food or photosynthesis, etc. ... Some energy is always wasted and some given off as heat, so in a wider context, the overall entropy is increased even when entropy decreases locally within an organism.

(Also see "Self-organization and entropy reduction in a living cell, Paul C.W. Davies, et al in Biosystems. 2013 Jan; 111(1): 1–10.)

If you decide to read the papers I cited on Constraint Closure, you will see that these approaches take into consideration the scale or granularity, as well as time dependence, of the system being described. The papers cited would also address some of the other questions you posed.

[_Physics Guy]

For a subsystem within a larger system to lose entropy is not a special feature of biology. It's what usually happens when entropy increases. The most basic and common case of entropy increase is heat flow from a warmer body to a cooler one, and when that happens the warmer body always loses entropy. The cooler one gains more. It's a kind of arbitrage in which entropy always shows a net profit, because temperature is literally the exchange rate between heat and entropy.

There is no need for biology to hold thermodynamics at bay in any sense; biology does not exploit any loopholes in thermodynamics. Biology is amazing but its amazingness is not in what it does with thermodynamics.

[_DrW]

For a subsystem within a larger system to lose entropy is not a special feature of biology. It's what usually happens when entropy increases. The most basic and common case of entropy increase is heat flow from a warmer body to a cooler one, and when that happens the warmer body always loses entropy. The cooler one gains more. It's a kind of arbitrage in which entropy always shows a net profit, because temperature is literally the exchange rate between heat and entropy.

There is no need for biology to hold thermodynamics at bay in any sense; biology does not exploit any loopholes in thermodynamics. Biology is amazing but its amazingness is not in what it does with thermodynamics.

Seems to me you are saying that physics has the same explanatory power with regard to biological self organization in the developing human embryo as it does in the formation of a snowflake.

As a physicist who claims that biology cannot hold thermodynamics at bay in any sense and that the evolution of biological systems is not non-deterministic, what is it that you find so amazing about biology?

I only ask because Kauffman and some of the other authors mentioned upthread, having approached the issue of self organization and evolution as biologists, appear to have a different view of the explanatory power of physical laws regarding biological evolution than what you seem to indicate on this thread. Kauffman titled his book be "A World Beyond Physics --", for a reason. He and other published authors do not believe that physics alone has the explanatory power required for a detailed understanding of biological evolution.

No one is saying that biology does not obey the laws of physics at the most fundamental levels. What they are saying is that new and fundamental concepts need to be layered over physics in order to gain predictive power with regard to biological evolution. Those concepts, including constraint closure, which has been adapted as a model for understanding consciousness, are being developed and published in peer reviewed journals.

Let me suggest that you have a look at Kauffman’s book entitled, A World Beyond Physics: The Emergence and Evolution of Life, or any of the papers on the concepts of self-organization and constraint closure in biology mentioned upthread. If you are still interested after doing so, I suggest we discuss the actual published works at issue rather than the short lay level synopses of the works as reported and commented upon here.

[_Arc]

This general approach to non-deterministic self-organization in biological systems has led to advancements over Tononi’s IIT theory of consciousness (mentioned upthread by Arc). One such advance, by Chang et al., is designated as ICT for the Information Closure Theory of consciousness. This paper, which appeared in September of this year, includes a formal mathematical description of the theory. A PDF version is available at https://arxiv.org/pdf/1909.13045.pdf

Thanks for the Chang ICT reference. I'll be interested to see how it extends the IIT of Tononi.

[_Gadianton]

I spent a little time with the ICT paper, hopefully enough to get a basic idea. First thing's first; as out of place in the modern world that Chalmers' argument for property dualism might seem, he gets cited even here:

It’s important to clarify that ICT does not intend to completely solve the hard problems of consciousness (Chalmers, 1995). Knowing the state of a conscious process does not allow us to answer "What is it like to be in this state of this process."

And the other big problem, identity (example: making identical replicas of persons or teleportation, is it still you?):

The problem of individuality is a significant theoretical weakness of the current version of ICT

Ayway, as I understand it, they're saying there's this theoretical way to separate a system from its environment, and if a highly separated system interacts substantially with its environment, then such a system is conscious -- the consciousness being the "information" content of the system. The "environment" in this case has as much to do with neighboring brain cells and tissue that isn't taking part in consciousness as what's out there in the world. This system is called NTIC.

They specifically reject functionalism: "ICT associates information with consciousness, functional features accompanied by consciousness are collateral consequence"

Okay, so I used functionalism as an example (it's the usual one) of multiple realization. The authors do accept multiple realizibility, and presumably NTIC could be done in silicon, but what on earth would that look like if we're rejecting functionalism? What isn't mind-brain identity but yet, not functionalism either?

Shelving that question for a moment. The ICT authors provide no limits on how NTIC could be wired up, other than to say doing it dirty in neurons isn't the only way. And so, the idea of "coarse graining" (coarse graining: temperature is to atomic motion as consciousness is to neural activity) is a secondary idea; it's important in that it's the mechanism found in nature and so we need to understand that to understand the relation to brain.

The authors specifically say "a crucial implication of ICT is that a pure feedforward network cannot produce consciousness because NTIC requires a form of memory". But what about recurrent neural networks "RNNs can use their internal state (memory) to process sequences of inputs. This makes them applicable to tasks such as unsegmented, connected handwriting recognition[1] or speech recognition.[2][3]"

So do they mean to imply RNN's are candidates for NTIC?

Neural nets are interesting constructs because they intuitively connect to natural processes, as in beehives or neurons. But neural nets, including RNN's are equivalent to computing. And so an RNN should also be out of the question as it's just another way of looking at computation, and computation is a subset of functionalism.

So there's "information" which holds an identity relation to consciousness and emerges in an NTIC, where some kind of functional states pace the information -- is it that information supervenes on neurons and functional states supervene on information (and these states are epiphenomenal)? And then, there is no telling how information explains the first-person experience of the system.

Well it's an interesting idea, I hadn't encountered that idea about separating a system from its environment and that alone is really fascinating. There are some really long articles about IIT that are pretty accessible but I didn't immediately run across that for ICT -- I think you said it was new, anyway.

[_Philo Sofee]

For Physics Guy. You said physics is determinism, but I just read Ilya Prigogines book "The End of Certainty" where he describes the physics as indeterminate and this is what causes the arrow of time. It's somewhat deep for an amateur as myself, but overall I believe I have grasped his point. How come Prigogine's ideas are not used and believed in physics.
Here is a link to him just to show that he is a valid scientist and thinker.
https://en.wikipedia.org/wiki/Ilya_Prigogine

Has anyone else read Prigogine and if so where is he wrong in his dissipative structures thinking concerning chaos indeterminism and the arrow of time?

[_Physics Guy]

I did my first postdoc with a former student of Prigogine's. Prigogine was a guy who did some really good things several decades ago and then went a little strange. It happens sometimes. I'd say he's probably worth reading because he might say something insightful but I wouldn't rely on his authority. Some of the things he asserted confidently may be things that only he ever believed.

Anyway, The End of Certainty was written in 1996. I don't think anyone could say that it has had a lot of impact. So I wouldn't read it as if it were anything like a cornerstone of modern science. I haven't read it myself and I don't remember anyone I knew ever being excited about it. My recollection is that the few times I did hear people talking about it they were kind of rolling their eyes.

[_Physics Guy]

I also tried reading that ICT paper but found it pretty disappointing. Lots of words, lots of arrows, little that I could recognize as actually meaning anything at all. For instance they never seemed to define "information", so it was hard to know what they were even talking about. They did talk about stochastic processes so I was guessing their concept of information was some kind of Shannon or Rényi entropy, but there are quite a lot of kinds of entropy. They didn't seem to define "stochastic process" and I couldn't tell why on earth they were interested in stochastic processes, anyway.

All they really seemed to do, in fact, was make a couple of definitions (involving undefined terms) and then claim that this was a big advance. I'm used to seeing people actually demonstrate some more non-trivial conclusions than that, with some actual examples, before claiming victory.

At the end of this paper I wanted to ask the authors, "So, is a cricket conscious? Is a thunderstorm conscious?" I had no idea from reading the paper what those answers would be. Worse, I had no idea from the paper what kind of data would be needed to answer the questions.

[Guest]

The idea that the laws of nature systematically favor the emergence of life, at least to some degree, is hardly unscientific.
[SNIP]
If natural law favors life, that still doesn't necessarily mean that natural law must have been designed to be like that by a benevolent God.

It’s my view that natural law indeed favors the emergence of life. The primary elements of organic chemistry are carbon and hydrogen (C and H). Here on Earth, a great deal of organic chemistry was accomplished before we arrived such that our main sources of organic fuels and organic raw materials came from previously living organisms. Elsewhere in the solar system, on Titan for example, temperatures are such that there are lakes of stable organic compounds including methane and ethane.

Along with C and H, the main elements in our coal and petroleum resources (derived from ancient life) are oxygen, nitrogen and sulfur. With sufficient relative concentrations of these CHONS in water, and with especially with clay as a catalyst, the spontaneous formation of monomers can occur. Amino acids (monomers of protein polymers) have been detected in interstellar space. Eventually it appears that self-organizing organic polymers can form from both amino acids and nucleotide monomers, and spontaneously create local transient negative entropy.

Add other elements such as calcium, potassium, phosphorus, chlorine, sodium, silicon and iron heteroatoms, and soon one can have fairly complex life. It’s interesting to note that all these elements required for life on Earth are formed in main sequence stars (comprising some 90% of the stars in the visible universe).

Carbon, oxygen and nitrogen are formed in relatively high abundance in main sequence stars in a positive feedback nuclear synthesis process known as the CNO cycle. Starting with hydrogen, nuclear synthesis in main sequences stars ends with iron, which is an important metal for oxygen transport in complex organisms. Elements heavier than iron are only formed in supernovae, from which all the naturally occurring elements are ejected into space.

As to life as we know it, when sufficient concentrations of these abundant and stable elements (starting with CHONS) are available in an enabling environment, a process of self-organization can begin.

Last time I posted about abiogenesis a few years ago, scientists had observed spontaneous formation, growth and division (reproduction) of fatty acid micelles, and the spontaneous self-organization of preferred RNA (more readily catalyzed and stable) sequences in a reaction mixture. Not life yet, but showing spontaneous, unguided formation of organic polymers on which living organisms are based. These exhibited rudimentary feeding from the environment, growth and reproduction.

[_Philo Sofee]

As to life as we know it, when sufficient concentrations of these abundant and stable elements (starting with CHONS) are available in an enabling environment, a process of self-organization can begin.

Last time I posted about abiogenesis a few years ago, scientists had observed spontaneous formation, growth and division (reproduction) of fatty acid micelles, and the spontaneous self-organization of preferred RNA (more readily catalyzed and stable) sequences in a reaction mixture. Not life yet, but showing spontaneous, unguided formation of organic polymers on which living organisms are based. These exhibited rudimentary feeding from the environment, growth and reproduction.

This is the area or at least the thinking of Kauffman in his book "At Home in the Universe." A most interesting idea this self-organization! It seems to me they are closing in on it. Exciting future discoveries may be in store for us all.