Abiogenesis

Post Reply New Topic

Blimey! I see this thread has descended into complete chaos. Nobody seems to understand what anyone else is trying to say. Misinterpretations on top of misinterpretations. Hmmmm! Not sure whether to try to pick up the pieces or leave the thread to die a natural death.

I always appreciate your posts ...! !

I am guessing you think this is a point in favour of intelligent design. If that is what you think, can you please explain why? I don't want to have to guess at your argument.


You will have to explain why you think I need that.


Same question as before: what is my mistake?


There are definite reasons to believe that magic doesn't exist. It violates rules for which we do have evidence, like the first and second laws of thermodynamics. The same goes for unicorns and fairies because they are supposed to be magical creatures. Are you saying self-organization would contradict something we presently accept as true?

You talk of "informationally enclosed" as if you assume there is a law of conservation of information analogous to conservation of energy. Is that an assumption you make? If you do, what reason do you have?


That very much depends on what alternative hypotheses you consider and how probable they are. We are discussing something that we know has happened. If the only two alternative explanations for what has happened are self-organization and an intelligent designer, and I have no positive evidence for either, should I assume an intelligent designer? Why? I would have to ask how probable the existence of the designer is. How do I explain the intelligence of the designer if I am supposed to avoid reference to self-organization? If I assume a designer for the designer, I have to explain the origin of the intelligence of that other complex designer.

The problem gets worse if I assume, as many intelligent design advocates do, that there is a law of conservation of information. Then the designer must be more complex than whatever he/she/it has designed. That only increases the problem of explaining where that information comes from.


We are arguing the difference between absence of evidence and evidence of absence, so you have to be clear about what you mean by problem. Are you saying that evolutionary theory has so far failed to explain the origin of these higher levels? Or are you saying that it is impossible for evolutionary theory to provide an explanation? Is it clear how different the two claims are?

The reason why there's no complete answer as to where space-time comes from is that the answer requires a theory of quantum gravity. Such a theory has proven difficult to develop because it's difficult to reconcile general relativity and quantum field theory for any given energy level, though the most popular and developed attempt to do this is is string theory. On the other hand if you just try to naively put the two together, you can get an effective theory of quantum gravity but it only works up to a certain energy scale, after which everything becomes infinite and all predictions become unphysical nonsense. The reason for this is that at the quantum gravity scale, or the Planck scale, space-time becomes fuzzy with no clear definition of space and time. This is because as virtual particles at this are created and annihilated at this scale, they curve space-time according to general relativity. So space-time becomes foamy. This brings me to a partial answer to your question, space and time are probably emergent from something more fundamental where neither space nor time exist as commonly understood. As to where the universe came from, there is a mechanism for that in the current inflationary theory called bubble nucleation. That is, a small bubble appeared in the quantum foam mentioned above, which could happen all the time due to the foamy nature. This bubble was initially one Planck length across or around 1.616*10^-35 m. The bubble then inflated, or expanded incredibly rapidly to become our universe. Sound familiar, this was the Big Bang. Again this is only a partial answer because a full explanation requires a complete theory of quantum gravity like what string theory tries to do, but string theory is not yet a complete theory and as of yet no one knows how inflationary theory is realized in string theory.



I can see that. I think I may be partly responsible for the thread initially going off track.

The evidence seems to point in such a direction.
It's the only way to explain the origin of the hierarchical modular organization of life.

(1) Self-organization has never been observed. See Self-organization vs. self-ordering events in life-origin models; All observation show that organization requires either direct interaction by an intelligent agent or an intelligently designed algorithmic system capable of organizing (e.g. life, computers, etc); there is no observed exception.

(2) If you read one of Dawkins book you notice that his whole focus is on selection? He never bothers to demonstrate that the right kind of variation takes place to find selectable variations/solutions. In his mind solutions are trivial and always available in just the right way. Experiments (and common sense, when one finally realizes this and starts to think about it) show differently: Some (to us obvious) variations (=solution to a problem) are never realized: Variations never realized can't be selected.

Also note that the (perceived) usefulness of a variation does nothing to make the appearance of this solution any more likely. Especially misleading are solutions which seem easily reachable at macroscopic level (phenotype) without knowing what it would require at the microscopic level (genotype), only solution which are easily reachable at the genetic level will ever be realized.

Only after a variation is realized in some individual can one measure the fitness of this individual; Note that one does not test the variation it self? It's all done indirectly in context.

In general: To test a given solution is trivial, however to generate any non-trivial solutions is in itself non-trivial. This is just common sense and common experience.

(2) Experiments with evolutionary algorithms has never resulted in to a solution with a hierarchical modular organization. Functionality is always spread out and the structure is often very complex and often barely understandable. In many ways such a process is alien to us and generates only unique solutions which no human would do. The designs generated like this look like nothing we would design. Notice that life does appear to look designed by a human-like intelligence?

(3) All experiments with variation-selection processes (computer simulations) stops dead-end against a complexity barrier. At best it can (over time) exhaust the trivial solution space, but after that it just repeats know solutions. It's not creative, it just an finite algorithm designed to find a local/temporal optimum within a very narrow variation set. This is how you get tigers and lion and all the other big cats from the same ancestors, each population wandering trough the same limited set of available configurations directed by natural selection.

This variation-selection process is however not capable of open-ended evolution. It can't move beyond the trivial. It can not cause from within a reorganization, only limited reconfiguration which the organization allows. Note that our knowledge of the link between genotype en phenotype is still very limited? People who know little can imagine anything to be possible; We already know that not everything it possible, but the exact limitations are not clearly known.

Because you introduced not only energy but also information (by painting the wall black/white); so the system is not really informationally closed, the information is only front-loaded into the system to become expressed when it opens up to light.

Making it open to light will only change the likely hood of some micro-events, those where the lack of light was the limiting factor, not all events; especially not very specific macro-events which require more or different information.

The more uniquely specific the macro-event that you want to cause is, the more unlikely its chance occurrence will be. Trough either direct interaction or front-loading the system one can make its occurrence more likely (inevitable) by adding all required organization information in just the right way, of just the right kind.

Yes.

Different kind of information: Meaningful information. Which is itself is a verily new concept. Specifically: Algorithms: Encoded in matter, to be interpreted and executed by matter. (This kind of information has no law of conservation, so that's not a concern.)

The problem is that the only known source of algorithms is human intelligence. Natural processes can not generate algorithms, only a "interpretor" (a machine) will be able to run them if they are encoded correctly. The encoding and the interpretor are bound only by the meaning of the algorithm in one way from encoding trough the interpretor; From meaning towards encoding requires (as much as we know) intelligent action.

No. Without positive evidence one should not assume anything.

In the end it may seem that one has the choice between either intelligent design (e.g. "we are cylons" ) or that we are extremely unlikely lucky that chance came up with just the same solution as an intelligent designer would have every fraking time; Neither option is very appealing for any scientist; but at least the ID option allows rational research to continue.

Rightly so, however this would require researching a designer; Given that at best we only see the artifacts left behind (life, us) and not the designers we can't say much about the nature of these designers.

Depending on your definition of information such a law does or does not exist. Some kinds of information are conserved.
True. But beyond the question "Is life as we know it designed?" To answer that question we need not know much about the designer(s) that did it.

Define the term evolution? What exactly do you mean with it?

That life did evolve is a fact. Life was once different from now, there has been an evolution of life. But by the same token computers have evolved as well. Computers where once different from now. To demonstrate that evolution occurred does not prove much about the method (the how) by which it has evolved.

Fact is that we now discover that the variation/selection method can't explain all we can see. It can explain some, but not everything. It's a required part, but it's not sufficient.

Sartresue, are you talking about cosmology, such as the Steady-state model? Or about biological organisms?

Isn't it speculated that the total sum of energy in the universe is zero, if one takes the negative energy of gravity into account? Well, I'm not a physicist.

A person that has the ability to speculate that has quite the imagination. But even if so, consider this: for purely alternating current, the sine wave has an equal amount of waveform below ground reference (i.e. "zero") as it does above. However, the amount of usable energy is not "zero" as that would be an improper form of averaging.

Also, as per the above equation of "0 = 1 - 1", that is improper, as it is actually "1 = 1", via

0 = 1 - 1
0 + 1 = 1 - 1 + 1
1 = 1


And this is similar to the Accounting Equation,

Assets = Liabilities + Equity

Whereas, you can set this equal to zero just the same as the other above,

Assets - Assets = Liabilities + Equity - Assets

Zero = Liabilities + Equity - Assets.


However, even with the equation set to zero, all of the transactions, debits and credits, still exist. Regardless of playing with numbers.

Only absence of evidence again. Think of Fermat's theorem. For centuries there was an absence of evidence that it is true. It would have been a mistake to conclude that it is therefore false. The Riemann hypothesis is 150 years old without proof that it is true. Would you conclude that it must be false, or that it is a hard problem to decide whether it is true or false?


I took the time to read it. I found 14 pages of stating as a fact that self-organization can't happen by either lawful processes or chance. Throughout those 14 pages, the only supporting references were other papers by the same authors. Then they no longer claim the impossibility of self-organization as fact, but as a hypothesis.

Their opinion on connectionism is quite interesting:

That looks like they reserve self-organization for whatever has free will. That is a rather restrictive definition. At first sight it seems far narrower than what they keep repeating elsewhere, that self-organization consists of setting dynamically inert or dynamically incoherent switches, but the restriction comes from what kind of switch setting they exclude from their definitions of self-organization. That's why they end up with this:


They presuppose a living cognitive agent being involved in any switch setting. That puts them in a rather curious position with regard to development from egg to adult. Is that self-organization? It definitely looks like a program by their definition, but any configuring of switches by environmental inputs (development is full of gene-environment interaction) probably would be a physicodynamic process by their definition, and they don't want to call that self-organization. Perhaps that is the reason why they don't discuss development.

In summary, I don't buy this. Abel & Trevors have carefully defined self-organization so that it can't possibly be involved in the origin of life, but once you say that only a living cognitive agent can organize its own activities, you are no longer discussing abiogenesis anyway.

A major problem with saying that only a living cognitive agent can organize its own activities is that it gives intelligence (or cognition) a magical ability that is assumed, but never explained. How does intelligence escape the supposed limits?


And how do you know what solutions are trivial and what solutions are non-trivial? You need an a-priori definition to avoid a circular argument. The common experience may be a consequence of unnoticed circularity. And you should be careful with appeals to common sense. If common sense were reliable, we wouldn't need science. Appeals to common sense are usually appeals to abandon science.


Ironically, that is why some of the results of evolution seem to be irreducibly complex. The geneticist Muller predicted that in 1918, and Behe fell for it anyway.


No. That is not something I ever noticed. How do you come to that conclusion? What are your criteria for identifying design?


They can also imagine many things to be impossible when the limits are in the simulations, not the process. Going by empirical data, I would think that biological evolution had a really hard time going beyond what you probably would consider trivial. It took a looong time until there was more than bacteria on Earth. Then still a fairly long time to multicellular life.

Say your intuitions about what is trivial and what is non-trivial are right. Then consider the history of life on Earth. Look at the time between successive generations of bacteria, at how many bacteria you have on a whole planet, and think about them being bacteria for billions of years. Looks like they had a hard time getting beyond some barrier. You say evolutionary algorithms hit a complexity barrier, and you conclude that shows the limits of evolution. I look at the history of life, and I think the limits may depend on what genotype-phenotype coupling evolution has to work with.


I thought you might say that, but in doing so you changed the issue being discussed. I objected to your claim that any state improbable in a closed system will still be equally improbable in an open system. That is not true. Energy flow through a structured environment will create conditions that are vanishingly improbable in a homogeneous environment at equilibrium. The study of abiogenesis always involves reconstruction of the conditions of the early Earth, the structure. If you want to call the structure of the early Earth front loading, you can only do that by assuming the strong anthropic principle, and the front loading would have happened at the origin of the universe, not at the origin of life on Earth. That becomes a discussion of cosmology, not of abiogenesis.


Of course not all previously improbable events become more probable. We are not talking about the infinite improbability drive here.


And so you say that depends on front loading very specific conditions? Doesn't that depend on you deciding on just one goal state in advance? That looks to me like playing chess and deciding in advance that you want to check mate in one predetermined position, instead of achieving check mate in any way that fits the rules of the game. If you decide on a narrowly defined goal, of course your probabilities are very small unless you set up very specific conditions.

There is a related problem with irreducible complexity. All the examples I have seen have been arguments from personal ignorance. Someone can't imagine a system simpler than what they see, they declare abiogenesis depends on that system coming into being in a single leap, then they calculate a vanishingly small probability of that happening. Then when they are shown that there are simpler precursors, they just do the same thing again. The problem with the approach is that they have no idea how to find out how much must happen in a single leap and they pick a goal state for that single leap that turns out to be arbitrary. I think your reasoning has a similar problem.


That is exactly what I mean.


I don't understand. I asked what information is conserved. You first offer me meaningful information, then give me algorithms as an example, then you say these are not conserved. Could you please make that clearer?


What is your criterion for deciding what an intelligent designer would have done?


Without researching the designer, you can't calculate the probability of design. without that probability, you can't identify design by making it the default that must be true if the one alternative you consider (abiogenesis or evolution, depending on context) appears improbable. You must find an independent criterion for identifying design. I have never found any criterion for that in the ID literature. Do you have one?


Only if you have an independent criterion for identifying design.


You talked about hierarchical modular organization and claimed that could not have evolved. It is your claim, so the definition of evolution that matters is your definition.

Would you say that the transition from prokaryotes to eukaryotes involved a new level in the hierarchy of organization? Would you say the transition from unicellular to multicellular eukaryotes involes a new level? Then I think you would have to say neither of these transitions was caused by any evolutionary process. Do you claim that, or have I misunderstood your position?

If you don't want to go through all this point by point, here are what I think are my most important questions:

a) What information is subject to a law of conservation of information?

b) Can intelligence violate that conservation of information?

c) If it can, how?

d) How do you identify design?

e) You said self-organization contradicts something we accept as true. What is the contradiction?

I can't answer to the whole post today, I will try to do so later this week.

In the mean time, a few answers and remarks.

I think we have a slight minor miscommunication regarding information conservation. I really did not make that statement, but I can understand how the confusion could have occurred. What I intended was the idea that a system can be open or closed for external prescriptive or organizational information. This is not information in the sense of Shannon information or physical information.

You are correct that we have little knowledge about intelligence. We are only just beginning to be able to do any serious research in this direction. A proper definition is hard to give at the moment. Why intelligence is different, why it has what you call a "magical" property. Most likely it may only seem magical because we don't understand it yet?

Answers to your most important questions:
a) As I pointed out above, the kind of information I'm taking about is not (in any relevant or meaningful way) subject to a law of conservation.

b) "Life is consistent with, but undecidable from physics and chemistry." (Niels Bohr) With regard to intelligence you can say the same. So intelligence can't violate any hard physical limit.

c) It can not.

To make this point crystal clear: I am not a creationist. The world/life was not created in 6 days, I do not believe in an immortal soul. There is no live beyond death; With death life itself ends and the person is gone. There is no life possible without some living physical body.

When discussion intelligent design the idea that this designer is a supernatural god always pops up and confuses the discussion. If anything this designer must be more natural then us. If we are build then we are not natural but in fact artificial. If this universe was created then nothing we see is truly natural. Only the (original) designer would be natural occurring (but then this designer needs to be before/beyond this universe).

I don't believe in supernatural gods. If we take the history of the concept god and want to have a timeless definition then it can be defined as anyone/anything more powerful then I or the personification of power for the powerless. "Mother is the name for God on the lips and hearts of all children." Only in this sense was the Pharaoh of Egypt truly a god, and so there are many gods.

The meaning of the word God, with capital, then becomes the hypothetical most powerful god; The one who is so powerful that no-one and nothing is more powerful. It does not mean that this hypothetical God can break the laws of nature. An interesting point can be made that God is the only true atheist: He really is without a god above him.

d) Assuming that we are intelligent, we can compare the technology that we design with what we find in living organisms. If they are similar it stands to reason that they have a similar origin.

Is life designed as we would do it? I say yes. The similarities are obvious. There is a clear hierarchy of modular design, obvious functional abstraction in interchangeable parts.

[The fact that we are at all able to transplant organs from one individual to another in it self is a very strong indication for design. There is no reason why natural selection would have favored this above all other possibilities. There is no hypotheses to explain this in terms of random variation. Random mutations at genetic level would not cause this.]

Can we explain these similarities in design of life and our own technology in any other way, by any other means? Can we explain how a non-intelligent process could possibly be able to abstract functional complexity in to mostly independent modules and functionally organize a system just like an intelligence would do? The whole idea is absurd*: Unintelligent processes do not generate the same patterns as intelligent agents would. If it looks a duck, walks like a duck and talks like a duck it most likely is a duck. So, if we are honest, we cannot explain it without an external intelligence.

* The absurdity is almost as crazy as the YEC idea that the world is really ca. 6000 years old, and all evidence against this is just their god who likes to pretend that its older. He buried the dinosaur bones, etc.

I offer a testable prediction for positive evidence: If life is the result of design by an intelligent designer we should find evidence of cross-lineage reuse, not just in-lineage reuse. Some designs or structures that where developed for one lineage, should show up in unrelated lineages developed later on by the same designer.

Modular design is by the way not something which a variation-selection process could have just stumbled upon by accident.. It's not some concrete functional part that we can add to a system to make it modular. It's not something you can point at and say: See, there!. Modular design is a design pattern: An abstract notion about how to structure stuff. Its also often not even the most efficient way to do things: it represents a cost. But often run rime is way cheaper then design time, so we accept the inefficiency at run time.

Adding billions of year does not change the intrinsic obstacles of any process. It's basically an appeal to luck. Without reason, logic, or science to back it up. Without even proving that it's possible or probable. It often ends up being no much more then a hollow and meaningless "Well, however unlikely, it happened because here we are!" which is almost equivalent to the old "Every house has a creator so the universe/earth/life/we must have one too". Neither one is really proving anything, they are just affirming a conviction.

e) I need to think more about this question to give you a proper answer.

No. Not at all. The wing shapes of vultures and sailplanes resemble each other more in foil section, aspect ratio and wing loading than the wing shapes of vultures and ostriches or of sailplanes and Concorde. The reason is function. If form follows function (your fitness landscape has only one or a few peaks) then similar form can follow from similar function without similar origin.

The same goes for convergent evolution, another case of similar solutions despite (phylogenetically) different origin.


Look up Unintelligent Design. There is a lot that human bioengineers would design differently. Starting with the human spine, lungs and vision. Here is a link to a list.


I don't follow your reasoning. Are you saying evolution should produce organs that are so different from one individual to the next that transplantation would be impossible? Why would you expect that? Or do you have some other reason for thinking transplantation favours intelligent design?


Yes. Form follows function. And often enough our technology copies biology: http://en.wikipedia.org/wiki/Biomimetics


Does this change your opinion?
Spontaneous evolution of modularity and network motifs

Spontaneous emergence of modularity in cellular networks

Spontaneous Emergence of Modularity in a Model of Evolving Individuals

Spontaneous emergence of modularity in a model of evolving individuals and in real networks

An Analytically Solvable Model for Rapid Evolution of Modular Structure

Evolution of Complex Modular Biological Networks

I'm stopping now because it takes me a lot longer to paste in the links and abstracts than it takes to find the references. You can find more by following up the ones I already gave you, or by searching for "evolvability AND modularity".


a) "Cross-lineage reuse" is expected in evolutionary theory. There is even more than one way in which it can happen. There is convergence and parallelism.
b) If you want to explain all cases of cross-lineage reuse by intelligent design, you must assume frequent intervention by the designer(s). If you postulate best practice, defined as each species getting the best solutions available to the designer(s) at the time the species is being designed (you are talking about intelligent design), then why do modern molluscs lack hemoglobin and myelinated axons, why don't we have lungs and colour vision like birds and a retina constructed the right way round like molluscs? If intelligent design is supposed to explain cross-lineage reuse, then cross-lineage reuse doesn't happen nearly as often as you would expect.

I think your testable prediction has been falsified, unless you add auxiliary hypotheses that explain why the designer(s) would not apply the best available solutions to all species being designed at the time. If you do that ad hoc, your prediction becomes too vague to discriminate between evolution and intelligent design. You would also have to discuss the intentions of the designer(s), something you wanted to avoid.


I write a program that produces 0s and 1s with probability 0.5 for each. Please calculate the probability of getting ten 1s in a row. If you like, assume that I don't use an algorithm to generate pseudo-random numbers, but that my computer gets its random numbers from having a small amount of a radioactive isotope and a Geiger counter.

I promise I could write you a program that would be almost guaranteed to come up with a string of ten consecutive 1s. I would not alter the probabilities of the individual events and I would not have the program lie about what random numbers it gets. Can you work out how I would do this?

Stanley Miller did a famous experament on this

Dr Jonathan Sarfati, PhD physical chemistry, speaking on Stanley Miller's experiment:

[youtube]http://www.youtube.com/watch?v=evuR14mWBRI[/youtube]

Quote from Dr Sarfati:


Nope, sorry it wouldn't. Not if only the conditions of the early earth were recreated in the laboratory and no other intervention happened otherwise. In that case, it would prove that life could of arose naturally in those conditions. If you go back to the start of this thread, you'll see that that's what some scientists have recently done.

You have not a clue about organic chemistry, do you? Anyhow, that is not the main point of the video in reference to Stanley Miller's experiment.

Here you make the mistake of assuming the conclusion you need to prove. You can't then just point to life as evidence of what an non-intelligent process can generate: You need to prove that that we see in life, all the features of life, can be explained by a non-intelligent process. Before you make an appeal to what a non-intelligent process is capable of, it would be quite handy if you could back that up with evidence that this is indeed within the limits of what it can do.

The question is Is life the result of intelligent design or non-intelligent process?

ID does not point to life, but to human technology and argues that similar results imply similar origins.

Can you make the case that the features of life can be explained by any non-intelligent process without assuming this to be true? Can you prove that convergent evolution is the result of a non-intelligent process and not intelligent reuse? Like I said, evidence of (historical) evolution does not imply cause, either intelligent of otherwise. The cause of this observation of evolution needs to be proven independently.

As far as I know there is no independent evidence that a non-intelligent process can cause a similar organization. None.
Sure, on details. Remember that we don't know all the requirements that went before the design, neither do we know know the exact limits of what is possible given the technology (biology) available. Maybe we wouldn't do it any differently it we would know more?

What we will not change is that way life is organized into a hierarchy of connected by mostly independent modules. This is the design aspect where Intelligent Design is talking about: The only know source of such an organization is intelligence.


Until now any evolutionary process (simulation) has never resulted in any "organs"; like I said before: A process of genetic variation and selection on phenotype results in a solution where the functionality is spread out throughout; Not abstracted, organized in to connected but independent (sometimes even interchangeable!) modules.

Because of what we see in experiments. All know experiments with (non-intelligent) evolutionary processes result is something without organs. There is not even a viable hypotheses to explain the origin of organs. Variation on genetic level will not cause them, selection on phenotype does not favor them (rather against in most cases).

We have a natural tendency to organize system in just this way because we can't do it any other way. Our minds are just not capable of doing it any other way; especially when the systems we want to design are complex.

A non-intelligent process would not have similar tendencies, limitations, etc.

That's evidence in favor of ID, biology = technology, unless you can prove that no intelligence is needed. Form follows function is a design principle, not evidence for a non-intelligent origin.

I will review it. Most of it I have already read. The problem in most cases is that they assume that an non-intelligent cause must have cause what we see and then proceed to explain the modularity we see in those terms without ever really proving the cause. Which is just bad science if you ask me. You can't assume a cause and then use the result as evidence for this cause. That's crazy talk.

Can you think of any test which could differentiate the two causes?

e.g. Our retina is constructed just right given what was required and possible. Note that oxygen transport in humans (and all mammals) is by blood? In lifeforms where this was not a limiting factor the retina was constructed differently because it could be constructed differently. Also, we tend to live in places which are much brighter then those other lifeforms do which do not use blood to transport oxygen: Our retinas, especially for high resolution colour-vision, uses a lot more energy and oxygen because of this much higher brightness.

See how reasoning from design can give a scientific explanation for the construction of retinas where evolution can't?
(Unless you now will re-use the above an retool an evolutionary explanation... )

I think not, but it might not be as conclusive as I assumed.
How do we know that was the best available? Also, humans don't always use the best available, do we? And sometimes we just try something new, which may or may not fail when tested. NB: ID does not require an infallible, all knowing designer.
True, you would not want to do so, unless its obvious: form follows function.

To get ten 1s on a row is not all that unlikely. It would be pointless however because it would not prove anything relevant.
Please explain why this would in anyway be relevant?

But lets make it interesting. Let say that those 1s and 0s are bit of a unicode string. Now, how long would it take before it would generate any original 180 page story in any human language which most readers will assume to have been written by a very intelligent writer? Can you think of a fitness function which does not require intelligence?