What evidence is there for evolution being non-random?

[mijopaalmc]

Interesting observation about Richard Dawkins:

According to this site, he hasn't published anything academic since 1998. Also, a his career progressed, it seems that published more and more in the popular press and less and less in the academic press. This doesn't mean that he wasn't/isn't a brilliant scientist, but I think it does mean that we should look with skepticism on comments made by or about him with respect to the state of scientific knowledge concerning the "randomness" of evolution.

 

[articulett]

Interesting observation about Richard Dawkins:

According to this site, he hasn't published anything academic since 1998. Also, a his career progressed, it seems that published more and more in the popular press and less and less in the academic press. This doesn't mean that he wasn't/isn't a brilliant scientist, but I think it does mean that we should look with skepticism on comments made by or about him with respect to the state of scientific knowledge concerning the "randomness" of evolution.

I think that makes him a whole lot more recent than the guys you were quoting. Moreover, that site doesn't look like it's even been updated since 1998! (Dawkins has published quite a bit in the last 9 years.) It is interesting that you would choose that site to support any contention at all.

And look, here's a current blog on the topic:

http://evolutionlist.blogspot.com/

Ever since Darwin, the primary "engine" of evolution has been considered to be natural selection. However, if one takes a closer look at this, it is clear that natural selection is not an "engine," it is an outcome. If evolution is defined as change in the characteristics of the members of a population over time and natural selection is defined as unequal non-random survival and reproduction (or, more parsimoniously, differential reproductive success), then the underlying cause of the changes that are differentially preserved over time is the real "engine" of evolution by natural selection.

And what might this "engine" of change be? Exactly what Darwin said it was in the Origin of Species: the "laws of variation" of which naturalists of his time were almost "completely ignorant." That is, given that some variations are heritable and that they can be passed from parents to offspring in the process of reproduction, then it is the processes that cause such variations that are the real "engine(s)" of evolution, including evolution by natural selection.

PZ Meyers, Carl Zimmer, and Sean Carroll also publish currently on the topic.

And here is Behe, again, at the Dover trial:

Irreducible complexity, on the other hand, is an argument to show that Darwinism, the presumptive alternative to design, is an unlikely explanation. However, one also has to be careful to remember that Darwinism isn't positively demonstrated by attacks on the concept of irreducible complexity. Darwinism can only be positively supported by convincing demonstrations that it is capable of building the machinery of the degree of complexity found in life. In the absence of such convincing demonstration it is rationally justified to think that design is correct.

Biologists will continue to use whatever descriptions or words best show how complexity evolves. To most of them, that means conveying an understanding natural selection. Whether they use words like random or non-random probably have more to do with the audience they are addressing and the questions they are asked. However, when they speak, they will ALWAYS be aware of this meme that wedge proponents have planted in the minds of the general public. So whatever your goals may be for describing evolution as a stochastic process, you will need a means of conveying how the complexity comes about if you expect any biologist to utilize your models and definitions. Otherwise, I think they'll stick with the words and models that are working--and refine and hone those. Your description and models are too easily abused and useless in regards to conveying incremental changes over time.

ETA: Not that it is relevant to this thread, but it is relevant to your derail: Here is a more recent CV of Dawkins: http://www.simonyi.ox.ac.uk/dawkins/CV.pdf.

 

[Schneibster]

I don't believe that.

You have a problem, then.

There is an experiment (not an appropriate subject for going into detail on in this thread, there are many references available on the 'Net and a pretty in-depth description on this forum written by yours truly) called the Delayed Choice Quantum Eraser and a related experiment called the Aspect experiment after its creator, and designed quite deliberately as a realization of the Einstein-Podolsky-Rosen experiment by which Einstein sought to show that quantum mechanics could not be a complete description of quantum dynamics, which show quite conclusively according to the overwhelming majority of physicists that Heisenberg Uncertainty is not a matter of us being unable to measure a value, but a matter of the particle in question not actually having a value of a parameter. If the spin of a particle (for example) on its X axis has been measured, then these experiments show not merely that we cannot measure the value of the spin on its Y axis, but that it does not have a value of spin on its Y axis at that time. If we later measure the spin on its Y axis, it will have one, but if many such measurements are made, they will give a quantum distribution, which is a random distribution, not a distribution that agrees with the law of the conservation of angular momentum. It's as if the value of the spin about that second axis was reset somehow by the earlier measurement of the spin about the X axis.

Random IS. The ultimate most basic constituents of reality show random behavior, and in fact if they did not, we would not see Newton's three Laws of Motion. More to the point, if they did not, we would not see the Second Law of Thermodynamics. This is not a guess, or an assertion; it is a fact. The derivation of the Second Law of Thermodynamics from the random character of quantum interactions is called the Fluctuation Theorem, and you will note that this is a mathematical theorem, not a scientific theory. It is therefore rigorous. The specific requirement for randomness is imposed by two requirements: ergodicity, and time reversal symmetry. These two cannot exist in combination without randomness.

Determinism exists; but it is a matter of the outcome of ensembles of randomly interacting particles. Be certain that you know precisely what you are talking about not believing.

ETA: Forgot to make the main point!

If you know the starting point has to have a truly random distribution, unconstrained by conservation laws, then by definition the same starting point need not create the same ending, if the outcome is deterministic.

 

[Schneibster]

The reaction against the word "random" is entirely due to creationist abuse of that word to imply that evolution without an intelligent designer is on par with the tornado in the junkyard analogy--as per the many examples provided (particularly the Deepak Chopra one).

And is unwise, since it appears that by the scientific definition of random, evolution is random. Just not by the common definition, which means disorderly.

 

[Schneibster]

What I find amusing is that having asked several of my colleagues (who are physicists, engineers, mathematicians and computer scientists): "Is evolution random?" most come to the opinion that it is not, though has random components. They are all considerably older than 6-years of age.

1. The plural of "anecdote" is not "evidence."
2. What definition of "random" did you ask them to use before you asked them whether evolution is random?

 

[Ivor the Engineer]

You have a problem, then.

There is an experiment (not an appropriate subject for going into detail on in this thread, there are many references available on the 'Net and a pretty in-depth description on this forum written by yours truly) called the Delayed Choice Quantum Eraser and a related experiment called the Aspect experiment after its creator, and designed quite deliberately as a realization of the Einstein-Podolsky-Rosen experiment by which Einstein sought to show that quantum mechanics could not be a complete description of quantum dynamics, which show quite conclusively according to the overwhelming majority of physicists that Heisenberg Uncertainty is not a matter of us being unable to measure a value, but a matter of the particle in question not actually having a value of a parameter. If the spin of a particle (for example) on its X axis has been measured, then these experiments show not merely that we cannot measure the value of the spin on its Y axis, but that it does not have a value of spin on its Y axis at that time. If we later measure the spin on its Y axis, it will have one, but if many such measurements are made, they will give a quantum distribution, which is a random distribution, not a distribution that agrees with the law of the conservation of angular momentum. It's as if the value of the spin about that second axis was reset somehow by the earlier measurement of the spin about the X axis.

Random IS. The ultimate most basic constituents of reality show random behavior, and in fact if they did not, we would not see Newton's three Laws of Motion. More to the point, if they did not, we would not see the Second Law of Thermodynamics. This is not a guess, or an assertion; it is a fact. The derivation of the Second Law of Thermodynamics from the random character of quantum interactions is called the Fluctuation Theorem, and you will note that this is a mathematical theorem, not a scientific theory. It is therefore rigorous. The specific requirement for randomness is imposed by two requirements: ergodicity, and time reversal symmetry. These two cannot exist in combination without randomness.

Determinism exists; but it is a matter of the outcome of ensembles of randomly interacting particles. Be certain that you know precisely what you are talking about not believing.

ETA: Forgot to make the main point!

If you know the starting point has to have a truly random distribution, unconstrained by conservation laws, then by definition the same starting point need not create the same ending, if the outcome is deterministic.

Interesting stuff quantum weirdness

I still fail to see how identical initial conditions - in every way (not just that which is known to us) - can lead to different outcomes. This may be a mute point, since identical to me would also be impossible to achieve.

Couldn't the various experiments performed at the quantum level be interpreted as putting a limit on what we can know? I.e. There are some events that will always be truly unpredictable or "random" to us.

 

[Ivor the Engineer]

1. The plural of "anecdote" is not "evidence."

True, just an informal observation.

2. What definition of "random" did you ask them to use before you asked them whether evolution is random?

I asked them how they would explain evolution. No one I asked said evolution is random.

My own opinion is that the word "random" implies a lack of memory, the exact opposite of what someone explaining evolution (to inform rather than deceive) is trying to convey.

Even to a scientist there are random processes with memory and random processes without memory. So the statement "Evolution is [a] random [process]" is ambiguous.

 

[Schneibster]

Interesting stuff quantum weirdness

I still fail to see how identical initial conditions - in every way (not just that which is known to us) - can lead to different outcomes. This may be a mute point, since identical to me would also be impossible to achieve.

When dealing with individual quanta, identical is omnipresent. You can't tell one electron from another; they don't have enough characteristics.

Couldn't the various experiments performed at the quantum level be interpreted as putting a limit on what we can know? I.e. There are some events that will always be truly unpredictable or "random" to us.

No, that's precisely what these experiments prove is not the case. It's not that we don't know it; it's that it doesn't have it. Not having a spin about one axis when spin exists about another is impossible for ordinary objects; it is therefore unimaginable. But for quanta, it's commonplace. It's not a limit on what we can know; it's a limit on what IS.

You really should read the descriptions of these experiments. They're quite compelling.

 

[aircloud]

Quantum Electrodynamics

Hawking sais that Quantum electrodynamics can determine everything that happens in Biology an Chemistry.
Determinism is similar to the Calvinist idea of the sovereighty of God but only from Gods perspective is everything determined.
Quantum mechanics determines either the velocity or the postion of a particle.
But as I have explained by adding equations you can get both postion and velocity of a particle using the Green Salad equation.
STEVE

 

[Schneibster]

My own opinion is that the word "random" implies a lack of memory, the exact opposite of what someone explaining evolution (to inform rather than deceive) is trying to convey.

I'm not sure what you mean by "memory," but if you mean dependence upon initial conditions, you need to know that the butterfly effect is alive and well.

Have you taken a look at nonlinear dynamical systems theory lately? They're doing all kinds of things with it. It's not just about fluid mechanics any more.

Even to a scientist there are random processes with memory and random processes without memory. So the statement "Evolution is [a] random [process]" is ambiguous.

Whatever memory means, I don't think it's the way that most scientists would divide processes up with respect to different kinds of randomness. I'd say a more likely division would be either into chaotic and non-chaotic, or constrained and unconstrained, random systems. Evolution is a constrained random system.

 

[Schneibster]

Oh for Ed's sake. Just what we need, the nutjob in here.

 

[Ivor the Engineer]

When dealing with individual quanta, identical is omnipresent. You can't tell one electron from another; they don't have enough characteristics.

...That we have observed. Which is my point. We may never be able to tell the difference between two electrons because we do not have access to the information that would allow us to do so. Or are you claiming there is no information that could be unavailable for us to observe?

No, that's precisely what these experiments prove is not the case. It's not that we don't know it; it's that it doesn't have it. Not having a spin about one axis when spin exists about another is impossible for ordinary objects; it is therefore unimaginable. But for quanta, it's commonplace. It's not a limit on what we can know; it's a limit on what IS.

You really should read the descriptions of these experiments. They're quite compelling.

I've read about the Delayed choice quantum eraser experiment. Clearly I'm not drawing the same conclusions from it that you are. I shall try again later.

I'm not really that far away from your position, just I would express it as "a limit on what IS for us".

 

[Schneibster]

...That we have observed. Which is my point. We may never be able to tell the difference between two electrons because we do not have access to the information that would allow us to do so. Or are you claiming there is no information that could be unavailable for us to observe?

Actually, other than position and momentum, yes, I am. You can look it up in any decent physics textbook, and even a number of popular treatments of the subject. That's what "elementary particle" means.

We're straying rather far afield. I'd like to suggest that if you want to discuss this in detail, I'm game, but another thread might be appropriate.

I've read about the Delayed choice quantum eraser experiment. Clearly I'm not drawing the same conclusions from it that you are. I shall try again later.

The history of the experiment goes back to a gedankenexperiment devised by John Wheeler called the Delayed Choice experiment. I'd suggest you might want to have a look at the DCQE information I posted in this forum a while back; you might have some questions that might be worth reviving the thread for. For a small consideration, I'll search it up for you.

I'm not really that far away from your position, just I would express it as "a limit on what IS for us".

That's quite far, from my perspective. The limit isn't us; it's the character of reality, and it's a necessary character because otherwise we wouldn't (and couldn't) see what we see.

 

[mijopaalmc]

As far as I understand, memory, in this context means Markov, or at least Markov processes, to my recollection Markov processes are sometimes called "memoryless processes".

 

[jimbob]

...That we have observed. Which is my point. We may never be able to tell the difference between two electrons because we do not have access to the information that would allow us to do so. Or are you claiming there is no information that could be unavailable for us to observe?



I've read about the Delayed choice quantum eraser experiment. Clearly I'm not drawing the same conclusions from it that you are. I shall try again later.

I'm not really that far away from your position, just I would express it as "a limit on what IS for us".

But what other chioice is there? The experimental designs would allow quantum effects to be non-random, only if these effects had (magically) predetermined the experimenter to perform the experiment at the particular time.

As Schneibster has said, the implication of the experiments is not that we can't measure both position and momentum (which would be a logical and simple interpretation of the uncertainty principle) but that the particles themselves do not "know" both their own values, which is far more interesting.

The MASER and quantum tunneling also suggeste this too.

 

[articulett]

And is unwise, since it appears that by the scientific definition of random, evolution is random. Just not by the common definition, which means disorderly.

But the whole post was about Dawkins/biologists use of the word and all of it is in response to the creationist obfuscation--which uses the common meaning of the word. It's all and attempt to allow people to comprehend the basics of evolution so that it is not equated with the impossible tornado in a junkyard analogy. Calling evolution random does not distinguish from the tornado in the junkyard analogy.

You may say evolution is random and that may be technically correct, but it's not informative as to how natural selection preferentially selects the next generation. I never hear the term non-random to describe evolution except as a reaction to the common understanding of random and misunderstandings built upon that. And most biologists go out of their way to emphasize the preferential survival of some organisms over others in explaining the concept or when people ask how evolution is "non-random"?

Isn't it the case that with the way you are describing random, every process is random and the evolution of anything is random? That might be a fine conclusion for all things that people don't have an agenda to misinform people about. It might be a great starting point to teaching people about chaos theory and entropy and probabilities and how order comes from randomness or why Brazil nuts end up on top. But this definition makes it a poor descriptor of the evolutionary process.

No matter what words biologists use it will always be in an effort to address the Behe brouhaha--the tornado in a junkyard analogy. All of the examples containing the word non-random or "opposite of chance" were in relation to the common use of the word. I really don't think that saying, "evolution is random, but order comes from randomness all the time" addresses the reason biologists use these words. They only use these words in response to a misperception of the common understanding of these words.

And as far as I can tell, there is no scientific agreement on the use of the word random even though you insist there is. There is no "singular" definition. I've heard some physicists say that nothing is "truly" random as everything is determined by the forces acting upon it. I guess biologists could say, "well evolution IS random if you are asking can it be described by or related to a probability chart, but if you mean 'disordered', then you need to understand that the environment selects certain traits through culling the genomes so that the reproductive success of changing genomes is altered through time."

Although stochastic models of evolution might be useful for physicists or some might insist that by the "true" meaning of the word evolution IS random or that there is NO evidence for evolution being non-random. I just don't think the rest of the world or other scientists are going to find this sort of conjecture or modeling useful...even if it's "true". It's just too prone to semantic abuse by those with a vested interest in having people not understand, as I hope was evidenced by the Behe quote. Unless more info. is added to the equation it just seems to mean nothing other than "evolution can be described by a probability chart". I think you'd be really hard pressed to find scientists who agree with this statement: "there is no evidence that evolution is non-random." I don't think most scientists have problems with the way Dawkins or the Berkeley site or Talk Origins explain the concepts --but then, again, they aren't addressing physicists and they are careful to define their terms.

If mijos OP question just boils down to, "what is the evidence for evolution not being able to be described by a probability chart?" then it's a really lame question. If it was about Dawkins' as he later contends, then the answer is, Dawkins is responding to the creationist obfuscation and the common understanding of that word. My mistake is in assuming mijo wanted the latter--as soon as I realized that per his definition of random, this thread and the evolution of everything is random I conceded that the evidence for the evolution of this thread being random is the same as the evidence for the evolution of anything being random no matter what definition you use.

 

[articulett]

http://www.discovery.org/scripts/viewDB/index.php?command=view&id=2539

According to Behe, the biochemist, recognizing the signature of the intelligent agent is like recognizing that the faces of the U.S. presidents carved into South Dakota's Mount Rushmore is the work of humans, not chaotic, random events.

Blame Behe if you don't like the response of biologists. Did Mount Rushmore come about via Random processes? Per your definition, it seems a case could be made that it did.

 

[mijopaalmc]

aeticulett-

The problem I have with your repetitious citation of Dawkins is that, quite frankly, you are giving undue weight to a non-expert in the field in which he is making comments. Dawkins, as far as I can tell from reading his CV, is not a mathematical evolutionary biologist. Although I do realize he is a, perhaps the, preeminent evolutionary biologist, I think that the biologists that I have cited above are better able to to evaluate the randomness of evolution as they were also statisticians.

In short saying that, because Fisher et alia stopped publishing in the 1980's, Dawkins (who stopped publishing in academia 10 years ago*) is more relevant ignores the fact that Dawkins is a non-expert in the mathematical aspects of evolutionary theory whereas Fisher et alia are much cited experts in the field.

*Reading Dawkins' CV make it obvious that he published almost exclusively in the popular media or publications that only reviewed his previous research after 1997/1998.

 

[articulett]

aeticulett-

The problem I have with your repetitious citation of Dawkins is that, quite frankly, you are giving undue weight to a non-expert in the field in which he is making comments. Dawkins, as far as I can tell from reading his CV, is not a mathematical evolutionary biologist. Although I do realize he is a, perhaps the, preeminent evolutionary biologist, I think that the biologists that I have cited above are better able to to evaluate the randomness of evolution as they were also statisticians.

In short saying that, because Fisher et alia stopped publishing in the 1980's, Dawkins (who stopped publishing in academia 10 years ago*) is more relevant ignores the fact that Dawkins is a non-expert in the mathematical aspects of evolutionary theory whereas Fisher et alia are much cited experts in the field.

*Reading Dawkins' CV make it obvious that he published almost exclusively in the popular media or publications that only reviewed his previous research after 1997/1998.

But I thought your whole question was in regards to Dawkins saying that natural selection was the opposite of random? http://www.simonyi.ox.ac.uk/dawkins/writings/blindwatchmaker.shtml

I don't even know if there are experts in the "mathematical aspects of evolutionary theory" other than in regards to assortive mating and gene sorting via game theory. And I can't imagine how any of their models could be useful without information generated by the reams of data generated by molecular DNA. And yes, Dawkins is known for making science understandable to lay people--but his scientific texts and articles are in wide distribution around the world and printed in multiple languages. I'm just pointing out that ways of explaining evolution will evolve depending on the audience, our understanding of evolution, our understanding of how evolution is best understood, and our realization of how various arguments are used to obfuscate rather than clarify.

I think it's pointless to point to old literature and pretend that current biologists are ignoring it. Scientific understanding evolves, and what works is added on to honed and modified-- less useful models become obsolete. Remember, the facts are the same. You are arguing that evolution be defined in a way that has been abused by creationists and that has not been useful in conveying the facts to the general public. You can insist that it's technically correct or the right way, but that's a semantic argument. We use the best tools we have to further understanding and, as far as I can tell, saying, "there is no evidence for evolution being non-random" conveys very little information about what evolution is or the way order arises from complexity. Hence, even if can be construed as true per some statistical definition or understanding--it still conveys no valuable information in itself. It is more confusing then clarifying. Why? Because it is indistinguishable from the very arguments creationist use to obfuscate the understanding of evolution.

I am pretty sure that Dawkins has added much more to the understanding of evolution to lay people and scientists than any of the authors you quote, and to me your statements in regards to clarity in conveying understanding of evolution is on par with Behe's--who is also not 6 years old and also a biochemist. Nobody is ignoring the research or the definitions of random--you just aren't saying anything that conveys anything that can't be conveyed in a simpler manner with clearer words. If you were, I'm sure people scientists would be honing your explanation or the ones you quote.

Remember--facts are the same; Definitions and explanations are not. The latter will evolve towards the one that is best at conveying the former. I think you are arguing for an obsolete simplistic way of defining something that can be defined in a much clearer and more useful manner. So long as your description of evolution is so prone to creationist abuse and so poor at describing how complexity evolves from randomness, you just aren't likely to get anybody to care that you think it's meaningful, useful, and true. Until you have the probability charts that explain the specifics of what you are trying to convey, who cares that evolution can be described by a probability chart? What can't be? Where's the meaning?

 

[Schneibster]

This is basically over, articulett. You're cherry-picking; if you want to respond, respond to my long post. I took your post apart, line by line. If you've got a response, now's the time.