Is evolution stochastic?

[recursive prophet]

There is an ongoing debate-over 2k replies-on RichardDawkins.net as to whether natural selection is stochastic or deterministic. At some point I may post a poll here on the question, but I'd be interested to know what those who follow this thread think. Also, if this has been covered in detail elsewhere I'd appreciate someone directing me to that thread. Thanks.

Split from: Evolution: the Facts. (http://forums.randi.org/showthread.php?t=88911)

[Dr Adequate]

Well, that's just semantics, which is why it's possible to debate it for 2000 posts without reaching agreement.

We know what natural selection does, we can understand the theory, make the predictions, and do the math, without answering that question or ever even thinking about it.

It's a quibble about words. If people started drawing different predictions from the two alternatives, then it would be a real issue. But so far as I know, they don't and it isn't.

[articulett]

It's mijopaalmc's old thread (goes by mjpam there)--

I had posted this clip about the God Delusion http://richarddawkins.net/article,210,Stephen-Colbert-Interviews-Richard-Dawkins,The-Colbert-Report-Richard-Dawkins (fun clip-- Dawkins on Colbert--a must see) where he once again explains that natural selection is the not random. And this was Mijo's reply.



The problem with using this quote and the other quotes you have posted is that they only assert that natural selection is non-random while providing very little justification for such an assertion. Those that do provide minimal justification use a different sense of "random" than the sense that is well-accepted by the scietific community. You really need to pay attention to the context of the quotes.

Here are some of the threads you may be looking for, RP:
http://forums.randi.org/showthread.php?t=82155
http://forums.randi.org/showthread.php?t=109957

You can google his name and get a nice list of all the threads he opens with insincere queries where he proceeds not to listen to anyone who takes the time to actually answer him.

[Dr Adequate]

Oh, him again. Couldn't he think of something else to be wrong about?

[recursive prophet]

There are non-chaotic physical systems where this isn't the case. The key word is "significantly".

I design and develop power transistors; these rely on the law of large numbers to average out quantum uncertainties, which make them work with very large numbers of electrons. Because they are power devices, they have to be large enough to manage high electric fields, so "normal" semiconductor physics (dealing with conduction bands and Fermi-Dirac statistics is valid, in a way that is getting less so for very small devices.)

I would argue that this means that the selection pressures, and thus the "direction" of evolution is subject to change due to random events.

If the direction is subjected to random change, then surely this is "random".
Articulett: Thanks for the links. I had read one of them before way back, but at this point it is so buried in the back pages I gave up finding it again. I don't have time to read them all right now, but as the quote above shows, apparently not everyone here disagrees with 'mijo.'
Oh, him again. Couldn't he think of something else to be wrong about?
Dr Adequate: On the Dawkins site the vast majority participating in the discussion agree with mijo, only we prefer to use the term stochastic as it is less ambiguous. If you're so certain the probability maths prove selection isn't in fact stochastic, (and I saw your graph) I would urge you and others who believe the same to join in the debate on Dawkins. I assure you it is far more detailed than what I have read here, and mijo is totally supported by 3 of the most knowledgeable users on the Evolution and Natural Selection Forum. They have made mijo's case far better than he did, though I believe he is pretty on target.

Several others on the ENS boards, like myself, started out leaning toward the deterministic side. But at this point most of us are convinced it is in fact stochastic, and that it's not just semantics. I am trying to recruit those who think otherwise as those arguing selection isn't stochastic are literally getting so soundly refuted I'm embarrassed for them. So come on over, and see if you can help Articulett and the 2 others still standing out. I try and keep an open mind on this, but at this stage it's difficult; the case for stochastic selection seems overwhelming from what I've read so far, and I've covered all 2000 replies.

[articulett]

RP is Mijo II. As if people "believe in" evolution being stochastic or determined. Pick one! And Believe! One or the other!

No, RP, most don't agree with Mijo there... just like on this forum, the smart people posted and left...and those who tried to educate were treated very poorly unless their words could be twisted to support Mijo's semantic game. (Though MZbiologist gets a nod from me for sticking with the nutters longer than anyone else.) I am not involved in the silly thread except to mock on occasion.

And everyone who explained "how evolution is not pure random chance" (the thread's title, btw) like Dawkins in that clip is shouted down by a couple of people with the intelligence of Mijo (I leave it to the reader to make their own determination). All who explain that it's a semantics argument are accused of calling it "determined" or accused of using the wrong definition of random. It is the same semantic games he played here. As if you could "win" a semantics argument. --As if determining what adjective to stick on evolution was more important than conveying understanding of what evolution is. (And I think we've all drawn our own conclusions as to Mijo's interest in and understanding of evolution.)

Start your own thread if you are interested in the best adjective to stick on evolution, Recursive Prophet. Most of us here are tired of the silly game... you and Mijo will have to recruit elsewhere. This is an thread about Evolution: The Facts. It's for people interested in how evolution works-- not for people interested in how they can describe it so that it sounds indistinguishable from the creationist straw man.

You can not win a semantic game, silly! Really. Though... as with Mijo... I'm sure this will not penetrate. (It is a good thing you have found each other.)



ETA--BTW-- probably not a good idea to quote jim-bob... he's seen as being maybe a hair more respectable and comprehensible than mijo around here. Either work on sounding more intelligent yourself or hitch your wagon to a smarter bunch. Coherence counts.

[biomorph]

There is an ongoing debate-over 2k replies-on RichardDawkins.net as to whether natural selection is stochastic or deterministic.

they've got the server space. It's a popular site. So?

At some point I may post a poll here on the question,
Opinions don't change the facts.

The point of the poll is?

but I'd be interested to know what those who follow this thread think.

Read away

Also, if this has been covered in detail elsewhere I'd appreciate someone directing me to that thread. Thanks.

You could say "detail", yeah, you could, we have a little.

And you could say "if" too.....you could.

Ah, you have.

From memory I recall a coupla, well, sorta backwater posts a while back.....cant' think where though, or when......:p

from my present understanding of the forum, the search thingy works sorta ok, you might find something there.........

[six7s]

the search thingy works sorta ok, you might find something there.........

It depends on what you're looking for (http://www.google.com/custom?q=%22anything+that+backs+up+my+woo%22&cof=S%3Ahttp%3A%2F%2Fforums.randi.org%3BAH%3Acente r%3BLH%3A75%3BL%3Ahttp%3A%2F%2Fforums.randi.org%2F images%2Fmisc%2Fsearchlogo.gif%3BLW%3A849%3BAWFID% 3A31746880203d5407%3B&domains=forums.randi.org&sitesearch=forums.randi.org) :p

[biomorph]

LOL thanx...:p:p

however searching the term "confirmation bias" and the like gives one or two more results....

[Dr Adequate]

Dr Adequate: On the Dawkins site the vast majority participating in the discussion agree with mijo, only we prefer to use the term stochastic as it is less ambiguous. If you're so certain the probability maths prove selection isn't in fact stochastic, (and I saw your graph) I would urge you and others who believe the same to join in the debate on Dawkins. I assure you it is far more detailed than what I have read here, and mijo is totally supported by 3 of the most knowledgeable users on the Evolution and Natural Selection Forum. They have made mijo's case far better than he did, though I believe he is pretty on target.

Several others on the ENS boards, like myself, started out leaning toward the deterministic side. But at this point most of us are convinced it is in fact stochastic, and that it's not just semantics. I am trying to recruit those who think otherwise as those arguing selection isn't stochastic are literally getting so soundly refuted I'm embarrassed for them. So come on over, and see if you can help Articulett and the 2 others still standing out. I try and keep an open mind on this, but at this stage it's difficult; the case for stochastic selection seems overwhelming from what I've read so far, and I've covered all 2000 replies. As I recall, mijo was pretending that evolution is random, which was stupid. If he is now reduced to saying that natural selection is stochastic, I guess that's slightly less misleading, though he would do better to say that evolution is a naturally occuring instance of a stochastic algorithm, then he wouldn't be wrong at all. But then, as his whole objective seems to be to describe evolution as inaccurately as possible, that would hardly suit him, would it?

I quite understand why the people on the Dawkins website have abandoned him as uneducable. They are right. He's been flogging this obsession for years, and it's not gotten any less dumb.

[recursive prophet]

As I recall, mijo was pretending that evolution is random, which was stupid. If he is now reduced to saying that natural selection is stochastic, I guess that's slightly less misleading, though he would do better to say that evolution is a naturally occuring instance of a stochastic algorithm, then he wouldn't be wrong at all. But then, as his whole objective seems to be to describe evolution as inaccurately as possible, that would hardly suit him, would it?

I quite understand why the people on the Dawkins website have abandoned him as uneducable. They are right. He's been flogging this obsession for years, and it's not gotten any less dumb.
Well, I guess ad homs are in vogue here also among those who don't believe selection is inherently stochastic. The only informed poster I know of on Dawkins who abandoned the thread-IMO because he was getting his ass kicked and was smart enough to realize it-was lbq, unless Articulett can name even one other. And who are the 'smart people?' Those who agree with your viewpoint? She keeps trying to maintain most agree with her, but if you go to the link Articulett posted above and read back few pages of the replies-especially those by susu.exp-you will find some very detailed arguments supporting the view that selection is stochastic.

[six7s]

And who are the 'smart people?Try this as a simple test:

Hands up those who can recognise, compare and contrast the following terms from Dr A's reply above:
pretending that evolution is random, which was stupid.
saying that natural selection is stochastic, I guess that's slightly less misleading
better to say that evolution is a naturally occuring instance of a stochastic algorithm, then he wouldn't be wrong at all

[articulett]

Isn't that funny? I mean the garbled part about "those who don't believe selection is inherently stochastic"? In his brain there is a way evolution MUST be described, and it has nothing to do with conveying understanding of the process. It's like trying to have a conversation with someone who is too busy having a conversation with the voices in their head. It's reminiscent of this:

N(nutter): Evolution is a noun
S(smart person): Well, yes, but it's a process-- calling it a noun doesn't really describe it.
N: A process is a noun.
S: Yes... but most people describe evolution as a 2 part process-- mutation and
N: Mutation is a noun
S: Yes. And then there is the 2nd part-- the part that selects (a verb) from the mutations
N: Selection is a noun...you're saying it's a verb! (little voice in head chalks up a "win" point)

Anyhow, that's what I mean by him being Mijo II--sorry he derailed this. I'm not going to stretch out his stay any longer by addressing his straw men. I'm sure everyone has as much information as they need to draw their own conclusions.

[mijopaalmc]

As I recall, mijo was pretending that evolution is random, which was stupid. If he is now reduced to saying that natural selection is stochastic, I guess that's slightly less misleading, though he would do better to say that evolution is a naturally occuring instance of a stochastic algorithm, then he wouldn't be wrong at all. But then, as his whole objective seems to be to describe evolution as inaccurately as possible, that would hardly suit him, would it?

I quite understand why the people on the Dawkins website have abandoned him as uneducable. They are right. He's been flogging this obsession for years, and it's not gotten any less dumb.
The focus of the debate has shifted, because very few people on the RDF actually disagree that evolution is a stochastic process. In fact, articulett is the only one claiming that evolution (the compound process of mutation, genetic drift, natural selection, etc.) is not a stochastic process. Rather, the predominate argument of the the thread is whether natural selection is a stochastic process. Given that the founders of the Modern Synthesis formulated a hypothesis that natural selection is a stochastic process and that that hypothesis has not been falsified over the past 50 years, it speaks volumes about your knowledge of evolutionary biology that you think I am "ineducable".

[articulett]

Try this as a simple test:

Hands up those who can recognise, compare and contrast the following terms from Dr A's reply above:
pretending that evolution is random, which was stupid.
saying that natural selection is stochastic, I guess that's slightly less misleading
better to say that evolution is a naturally occuring instance of a stochastic algorithm, then he wouldn't be wrong at all


I'm guessing everyone who reads it... except for the "not so smart people".

Let's contrast the level of clarity in RP's or Mijo's posts. How is it that Dr A manages to say so much with so few words, while you can't find a single useful point in pages and pages of Mijo or RP's posts? Or rather, I can't. Maybe someone can. (But I'm skeptical.)

Mijo cannot understand the last sentence because he needs to be able to boil evolution down to the creationist straw man-- the idea that evolution is a theory of random chance. He's much more interested in the semantics than conveying understanding of the process which he seems utterly clueless and absolutely incurious about.

[articulett]

Don't derail this thread Mijo... and I will use my words to characterize my own argument.

You haven't got a clue. Go play on the thread you derailed over there. Don't embarrass yourself here.

[articulett]

The focus of the debate has shifted, because very few people on the RDF actually disagree that evolution is a stochastic process. In fact, articulett is the only one claiming that evolution (the compound process of mutation, genetic drift, natural selection, etc.) is not a stochastic process. Rather, the predominate argument of the the thread is whether natural selection is a stochastic process. Given that the founders of the Modern Synthesis formulated a hypothesis that natural selection is a stochastic process and that that hypothesis has not been falsified over the past 50 years, it speaks volumes about your knowledge of evolutionary biology that you think I am "ineducable".


I believe it was in your silly "why is there a discontinuity in the fossil record?" thread where Dr Adequate could reach his conclusion. He was your advocate and patiently answered your questions in great detail with illustrations and links. And your conclusion was "so scientists can't explain the discontinuity, eh?" I think many people concluded at that point that you were ineducable... others waited until you started and derailed even more threads without ever listening to any of the detailed answer people gave--because you just assume you already know the answer.

I think it speaks volumes for your density that you don't understand why most smart people feel that you are "ineducable". The evidence is all over this forum and assorted other skeptic forums where you spew the same garbled nothingness.

I think Dr Adequate's many writings and skeptwiki are testament to his knowledge on evolutionary biology... Nothing needs to "speak volumes"-- he's written volumes--clear, coherent, beautifully written, much praised, "volumes". You just spin out semantic blather. There's not a gem of wisdom, truth, depth, in any of your words. You insult others so that you can feel better than them, but you are only fooling other fools.

[mijopaalmc]

Well, that's just semantics, which is why it's possible to debate it for 2000 posts without reaching agreement.

We know what natural selection does, we can understand the theory, make the predictions, and do the math, without answering that question or ever even thinking about it.

It's a quibble about words. If people started drawing different predictions from the two alternatives, then it would be a real issue. But so far as I know, they don't and it isn't.

The "semantic argument" argument is an old and tried red herring. You and articulett would be the first to point out that the creationist argument that evolution cold be wrong because evolution "is only a theory" is a straw man, because the definition of "theory" is different within the scientific community that it is in everyday conversation. Nevertheless you and articulett are ignoring that the publications in which Dawkins asserts that evolution is non-random are aimed toward a popular audience and therefore use the popular definition of "random". You two are also ignoring that within the peer-reviewed literature, "random" is most often used in a way that is synonymous with "stochastic".

[mijopaalmc]

Don't derail this thread Mijo..

How is discussing a fact about evolution that has been well-established since the 1950's derailing a thread entitled "Evolution :the Facts"?

and I will use my words to characterize my own argument.

And you will deny that you have said thing that it is easy enough for other to prove you said.

You haven't got a clue.

I am more aware of the past 50 years of research in evolutionary biology than you are. You choose to rely on press releases and popular accounts of evolution instead of the preponderance of the peer-reviewed literature.

Go play on the thread you derailed over there. Don't embarrass yourself here.

For the millionth time, the thread that you have posted your senseless rants in was merged with a thread that I was participating in which had an OP the specifically addressed the usage of "random" and "stochastic" in evolutionary biology.

Also, you only think that I am "embarrassing" myself because most people here have similar backgrounds to you and therefore agree with you. However, as you have been shown multiple times at RDF, the theories that Dawkins and others presented in their popular publications:


assume infinite population sizes
ignore genetic drift


They are thus a good starting point for understanding but fail to cover many cases within evolutionary biology. Putting these simplified models as completely descriptive of evolution (as you do when you insist that Dawkins' statements refute the fact that evolution is a stochastic process.

[articulett]

so sad... but so funny. (It almost makes me miss Kleinman.)

(I encourage others not to feed any trolls and perhaps the mods will split off the derail.)

[mijopaalmc]

I believe it was in your silly "why is there a discontinuity in the fossil record?" thread where Dr Adequate could reach his conclusion. He was your advocate and patiently answered your questions in great detail with illustrations and links. And your conclusion was "so scientists can't explain the discontinuity, eh?"

That was not my conclusion. If had bothered to read what I had written in that thread, you would have seen that my problem in understanding was based on the assumption that popular representations (e.g., the opening sequence from "The Simpsons", "The Descent of Man", etc.) were accurate representations of how scientists understood evolution to work. Since the error was in my understanding of evolution and not the theory of evolution, I never concluded that scientists had no explanation for the "discontinuity" in the fossil record.

I think many people concluded at that point that you were ineducable... others waited until you started and derailed even more threads without ever listening to any of the detailed answer people gave--because you just assume you already know the answer.

I think it speaks volumes for your density that you don't understand why most smart people feel that you are "ineducable". The evidence is all over this forum and assorted other skeptic forums where you spew the same garbled nothingness.
No, "[t]he evidence is all over this forum and assorted other skeptic forums" is that very few people who spend their time defending evolution on the internet have not read Fisher, Wright, Haldane, Maynard-Smith, Kimura, Ohta, Nei, Mayr, Gould, and many other evolutionary biologists’ peer-reviewed publications. Instead, they settle for the popular publication of evolutionary biologists and pretend that the present anything close to a complete representation of how evolutionary biologists understand evolution.

I think Dr Adequates many writings and skeptwiki are testament to his knowledge on evolutionary biology... Nothing needs to "speak volumes"-- he's written volumes--clear, coherent, beautifully written, much praised, "volumes". You just spin out semantic blather. There's not a gem of wisdom, truth, depth, in any of your words. You insult others so that you can feel better than them, but you are only fooling other fools.

Oh, so writings of an anonymous person on the internet outweigh The Genetical Theory of Natural Selection or The Causes of Evolution?

[mijopaalmc]

so sad... but so funny. (It almost makes me miss Kleinman.)

(I encourage others not to feed any trolls and perhaps the mods will split off the derail.)

Why not actually read the evolutionary biologists I have listed and build an argument that they say that evolution is not a stochastic process?

[The Atheist]

I've just asked for this discussion to be moved to a new thread - not really appropriate for this one.

Thanks

[mijopaalmc]

I've just asked for this discussion to be moved to a new thread - not really appropriate for this one.

Thanks

How so?

That evolution and natural selection are stochastic processes has been a well-established fact since the 1950s. Just because there is a contingent here that prefers discussion of simplified theories of evolution and natural selection to the theories that have been laid out and subsequently not falsified doesn't mean that evolution's being a stochastic process is not a germane to a discussion of the facts of evolution.

[Paulhoff]

How so?

That evolution and natural selection are stochastic processes has been a well-established fact since the 1950s. Just because there is a contingent here that prefers discussion of simplified theories of evolution and natural selection to the theories that have been laid out and subsequently not falsified doesn't mean that evolution's being a stochastic process is not a germane to a discussion of the facts of evolution.
Tell us what parts of evolution that are not random, and what parts are.

Paul

:) :) :)

[Dr Adequate]

How so?

That evolution and natural selection are stochastic processes has been a well-established fact since the 1950s. Just because there is a contingent here that prefers discussion of simplified theories of evolution and natural selection to the theories that have been laid out and subsequently not falsified doesn't mean that evolution's being a stochastic process is not a germane to a discussion of the facts of evolution. So you've abandoned the whole "evolution is random" claim? Good.

So, if you've abandoned your "simplified" (i.e. wrong) "theories of evolution", and you are now willing to agree with the rest of the world about the nature of evolution, what are you complaining about?

[mijopaalmc]

Tell us what parts of evolution that are not random, and what parts are.

Paul

:) :) :)

Let's start with mutation, genetic drift, and natural selection. Some, such as mutation and natural selection, may be biased to one outcome or another. For instance, the probability of mutation is not uniform across the genome (there are hot spots of mutation with in the genome) and natural selection is defined as the differential probability of reproduction across different phenotypes. Unfortunately, bias does not equate to non-randomness, and no-one has produced credible evidence that all individuals possessing each phenotype within a population either reproduce or fail to reproduce.

[Dr Adequate]

The "semantic argument" argument is an old and tried red herring. You and articulett would be the first to point out that the creationist argument that evolution cold be wrong because evolution "is only a theory" is a straw man, because the definition of "theory" is different within the scientific community that it is in everyday conversation. Nevertheless you and articulett are ignoring that the publications in which Dawkins asserts that evolution is non-random are aimed toward a popular audience and therefore use the popular definition of "random". You two are also ignoring that within the peer-reviewed literature, "random" is most often used in a way that is synonymous with "stochastic". Ah, I see. I wondered why you'd briefly started telling the truth, but now I understand. Having admitted that evolution is a stochastic process, you're going to pretend that this means the same as "evolution is random".

Do let us know if you fool anyone. (Creationists don't count, as they are already wrong.)

[Paulhoff]

For instance, the probability of mutation is not uniform across the genome
That would be random right there for one.

Paul

:) :) :)

[mijopaalmc]

Ah, I see. I wondered why you'd briefly started telling the truth, but now I understand. Having admitted that evolution is a stochastic process, you're going to pretend that this means the same as "evolution is random".

Do let us know if you fool anyone. (Creationists don't count, as they are already wrong.)

This is just a silly argument because I have been explicitly that, within the context of the discussion on this thread (and the thread that started nearly two years ago), "random" is synonymous to "stochastic" only in so far as both mean "based on probability". If you have evidence that there is only one outcome for thte reproduction of each phenotype within a population, you should present it, as it falsify the Modern Synthesis and probably win you a Nobel Prize in the process. However, pretending that I am making some sort of unorthodox statement about evolution because I say that it is accurate to call it "random" if you carefully explain what you mean by random before is as misleading as you would like others to think calling evolution "random" is in the first place.

[MG1962]

Can I jump into this gunfight for a minute. Back in the 80's there was a lot of interest in the "Evolutionary Pump" theory. Googling brings nothing up, so I am guessing the concept died or has a new name.

If it has a new name, is this an example of evolution operating in a non random way. If it doesn't exist any more, what explanation is there for say the dramatic and quick evolution of the whale and the seeminly frozen state of shark evolution.

**Ducks back into the saloon before being hit by a richocette**

[Dr Adequate]

Oh, so writings of an anonymous person on the internet outweigh The Genetical Theory of Natural Selection or The Causes of Evolution? Since your mishmash of nonsense appears in neither of those works, which I note that you did not write, a more apt question would be whether my writings outweigh the writings of another anonymous person on the Internet. To which the answer is yes, they are indeed superior to your tedious attempts at obfuscation --- but then, one could say the same of complete silence.

[six7s]

Mijo, long time no see

It was great to see your back! :D
That was not my conclusion. If had bothered to read what I had written in that thread, you would have seen that my problem in understanding was based on the assumption that <snippedForTheSakeOfBrevitySanityAndCosicanity/>
As ever, you seem to be having significant problems with language, this time omitting and mixing up pronouns

Allow me:

That was not my conclusion. If <insert>I</insert>had bothered to read what I had written in that thread, you <insert>I</insert>would have seen that my problem in understanding was based on the assumption that blah, blah,blah

As ever, I am happy to help

To illustrate this, here's a link you might want to follow to see a listing all of the places where your insight is still required (http://www.internetlastpage.com/)

[mijopaalmc]

Since your mishmash of nonsense appears in neither of those works, which I note that you did not write, a more apt question would be whether my writings outweigh the writings of another anonymous person on the Internet. To which the answer is yes, they are indeed superior to your tedious attempts at obfuscation --- but then, one could say the same of complete silence.

Nice misdirection!

If you insist that these works do not say that evolution is a stochastic process, you have obviously not read them nor apparently have you read any mojor evolutionary biologist since the 1920s. What it seem you have done though is substitute explanations written for a popular audience for what evolutionaryu biologists actually say in their peer-reviewed publications.

[paximperium]

Nice misdirection!

If you insist that these works do not say that evolution is a stochastic process, you have obviously not read them nor apparently have you read any mojor evolutionary biologist since the 1920s. What it seem you have done though is substitute explanations written for a popular audience for what evolutionaryu biologists actually say in their peer-reviewed publications.
Would you care to name five "major" evolutionary biologists within the last 20years who believe and support your wonderful little claim and cite their works which you seem to be such an expert of?

[Dr Adequate]

This is just a silly argument because I have been explicitly that, within the context of the discussion on this thread (and the thread that started nearly two years ago), "random" is synonymous to "stochastic" only in so far as both mean "based on probability". And in so far as they do not mean the same thing, they are not synonymous.

If you have evidence that there is only one outcome for thte reproduction of each phenotype within a population, you should present it, as it falsify the Modern Synthesis and probably win you a Nobel Prize in the process. I made no such claim, and you will deceive no-one into believing that I did.

However, pretending that I am making some sort of unorthodox statement about evolution because I say that it is accurate to call it "random" if you carefully explain what you mean by random ... It is perfectly true that if, every time you say "evolution is random" you also explain that when you say "evolution is random" you don't mean that evolution is random, then people will not take your statement that evolution is random as meaning that evolution is random --- assuming that they understand and believe your explanation of how you don't mean what you're actually saying.

They may think that you've gone off your head, but that's your problem.

You might also truthfully claim to be an aadvark so long as you carefully explain that by "aardvark" you don't mean aardvark; this too would allow you to grasp at the shadow of a silly lie while claiming your statements to be true in substance. But I have no idea why you'd want to.

[Dr Adequate]

Nice misdirection!

If you insist that these works do not say that evolution is a stochastic process ... Of course, I insist on no such thing.

What I said was that "your mishmash of nonsense appears in neither of those works". As everyone reading this thread can see what I wrote, you will not deceive anyone into thinking that I said something else.

I don't see why you are whining about "misdirection", since you identify none in my post, but when you follow this vacuous complain with deceitful garbage like that, you pile hypocrisy on mendacity.

... you have obviously not read them nor apparently have you read any mojor evolutionary biologist since the 1920s. What it seem you have done though is substitute explanations written for a popular audience for what evolutionaryu biologists actually say in their peer-reviewed publications. What a strange fantasy. Though I guess it's no stranger than the other stuff you've made up.

You should note that whether or not you can control your instinct for falsehood, you are unlikely to deceive me as to what I have and haven't read.

[articulett]

He actually can twist the words of a few people and their definitions and allude that they are saying what he is saying... but all he's really saying is "evolution is random" because he desperately needs it to be so.

He's been shown a peer reviewed study from 2007 that says: Evolution is deterministic, not random, biologists conclude from multi-species study http://www.bio-medicine.org/biology-news-1/Evolution-is-deterministic--not-random--biologists-conclude-from-multi-species-study-1268-1/

--and yet he maintains with a straight face that the authors of this study are using their words wrong and being less clear than he is!

He cannot understand anything he reads... none of the papers he quotes... he just reads the words so that he can prove to himself that some scientists somewhere think it's useful or informative to say "evolution is random". (His new tactic is to use "stochastic" as a synonym for random... just like creationism morphed into ID with the hopes of drawing more respect.)

I challenge anyone to find one great clear sentence of Mijo's-- or even the point of what he's trying to say. In his mind there's this great contingent of scientists discovering that evolution is really "random" (however one wishes to define that term) because (gasp!) in can be modeled using probabilities and no one can "prove" that unbiased processes are not "random"-- (though few people who are trying to clarify would refer to a biased process as "random"--just as they wouldn't refer to a loaded dice as random and they'd distinguish the randomness in poker from the randomness of roulette.) Per Mijo's definition, elections are as random as lotteries! Every "thing" is random just as every "thing" is a noun. Truly he never says anything more informative about evolution than calling it a noun. Sure, evolution IS a noun, but why the hell would anyone who cares about describing the process to anyone need to emphasize that? They wouldn't. Unless their goal was to obfuscate understanding.

In his mind he's trying to prove that creationist misconceptions are true--this nutty idea that evolution is a "theory of chance"--it's the best the creationists have these days:

From a review of a creotard book:

A detailed critique being out of the question, the strategy adopted here is to focus upon two themes that are characteristic of the book's treatment of evolutionary biology: chance and typology.

The first theme, which occurs repeatedly as a leitmotif, is that familiar old war horse, "Mere Chance." It first appears in the preface with the statement that since Charles Darwin's time "... chance ruled supreme. God's will was replaced by the capriciousness of a roulette wheel." In a later passage is found the assertion: "The driving force behind the whole of evolution was the purely random process of natural selection" (p. 60). Equating natural selection and the origin of adaptations with "problem solving by trial and error" and "gigantic random searches" is a repeated theme (e.g., pp. 61 and 308).

Pejorative appeal to naive notions of "chance" is typical of creation science literature and is a clear sign that Denton's book is not to be taken as a serious book on biology. http://ncseweb.org/creationism/analysis/review-evolution-theory-crisis

I know that people here aren't fooled, but it does seem to confuse a few people (see Recursive Prophet). When people understand how natural selection is NOT pure chance, their "intelligent designer" becomes superfluous. Mijo derailed a thread on the Dawkins forum that was entitled "How Evolution is Not Pure Random Chance" to tell everyone that it really is "pure random chance" (per his convoluted reasoning.) And he asserts that all these important people agree with him. It makes me laugh, but I'm lucky enough to understand why it is so laughable. Some people are gobbling up this garbage and sounding just as muddled as the folks putting it in their heads. --Which wouldn't be so awful if it didn't keep them from understanding some of the most profound things humans have come to know. They bad mouth those who share the facts while never saying anything intelligible at all. Mijo actually thinks he's being more scientifically "rigorous" than Dawkins??!! Or Dr Adequate?!? He imagines the people he quotes are supporting his nothingness?!?

Ask yourself why someone is so desperate to "win" the equivalent of the statement "evolution is a noun"? Why would they need to characterize anyone who was being more clear as someone who is claiming that "evolution is a verb". This is exactly what Mijo is doing. Just substitute "random" for "noun". Biologists Ken Miller has this to say about such a tactic:

I have no idea why someone would take a term like natural selection and say it is random... Natural selection is a distinctly non-random process that acts as a sieve through which genetic changes are filtered. Just as a sieve filled with various rocks will not end up filtering out its contents randomly, natural selection does not filter organisms randomly. http://forthesakeofscience.wordpress.com/2008/10/25/why-natural-selection-is-not-random/

I have no doubt all respected biologists would agree. And none of it has anything to do with the "proper" definition of words (I find it hysterical that Mijo imagines himself the definer of the "correct definitions"-bwa ha ha ha ha--as if!)

Learn from the real experts folks--not the ones who imagine themselves to be such.

[recursive prophet]

YOu have a truly Shakespearean bent Articulett; Mush Ado About Nothing. Would you like me to send you a link to the reply by your ally MZ where he states that while he knows evolution is random, it doesn’t apply to selection? And it is you who keep claiming we say selection is random, when in fact the claim is that it’s stochastic? Do you understand the difference? And for the record, I count at least 9 regular posters on the Dawkins ‘random’ thread who argue selection is stochastic compared with 3 who claim it isn’t.

As Mohammed won’t go to the mountain, I posted susu’ latest reply to MZbiologist, only one of two 'determinists' left on the thread, who continues to argue not that selection isn’t necessarily stochastic and now claims that it’s a philosophical issue. The second is one of susu's earlier comments where he gives a summary of his viewpoint. Note how he makes a detailed reply to each of MZ’s points, and does not emulate his ad hom insults.

I've edited out the very confusing copy/paste I posted earlier. Sorry if this was confusing, but I lacked time to write in the code difference and wasn't even sure it would work as those quoted aren't members here combined with restrictions on new posters. I'm grateful to mijo for doing it for me, and I will use his formant in future quotes from the Dawkins site, or just post links after I hit 15 replies and can do so. I’d love to read a comprehensive response to the points susu has made that mijo posted on the page 2, and again invite any ‘determinists’ to weigh in at Dawkins. Of course those who agree selection is stochastic are also welcome, but that side really doesn’t need any help. :rolleyes:

[sol invictus]

YOu have a truly Shakespearean bent Articulett; Mush Ado About Nothing.
<enormous, garbled, and incomprehensible post immediately following that comment snipped>


The debate here was over an almost totally meaningless semantic distinction, with a creationist subtext. If you want to resurrect it , I suggest you begin by defining "stochastic". Then we can have a poll and see who agrees with the claim that "evolution is stochastic" under your definition.

Note that there need to be three options ("true", "false", and "not even wrong"), as there are many statements ("evolution is random" is one) which are partially true and partially false, particularly when your definition is as stupid as mijo's was.

Oh, and by the way - the answer is E and 3, as every schoolchild knows.

[The Atheist]

How so?

Already done. You can always take it up in management if you disagree since it's already moved.

[mijopaalmc]

Here's what the quotes in Recursive Prophet's post above should look like:

A strict Popper view denies that anything can ever be confirmed as proven. This is a limited and naive view. In any given system, I can prove many things.

No, you only can disprove things, though you prove that one of the alternatives must be true in the way of doing that. You can also prove that things are possible, simply by noting a case where they have happened. It´s not a proof of a hypothesis, it´s just data.

I can prove a gene is essential by knocking it out.

Actually, you disprove that a gene is non-essential, by knocking it out.

I can prove an enzyme can cut a substrate.

In this case you note that it can happen. You can not prove it will happen, though if it repeatedly does the hypothesis that it will do so becomes rather certain. I can "prove" the sun came up this morning, but that´s not a hypothesis, it´s data. A hypothesis has to make predictions about the future, like "the sun will keep coming up every morning".

Here´s a simple question:
I´ve got 4 cards in front of me. all of them have a number on one side and a letter on the other. You can only see one side and the cards are showing E, 4, F and 3. I claim that on the cards with a vocal there is an even number on the other side. Which two cards will you turn to figure out if I´m right?

I can prove that a mutation "can cause" a change in phenotype. What Popper is stating, and Eversbane is adhering to, is the notion that I cannot prove a mutation "does cause" an observed phenotype outside my experimental system. Many observed effects can have multiple causes, just as many phenotypes can be affected by many genes. This type of Popperian conclusion is completely fair.

It does not rule out "proving" a cause leads to an effect in an experimental system. Eversbane implied a philosophical view that nothing is ever provable or certain. That is not the case.

No theory is ever provable or certain. A theory is a set of axioms, from which predictions can be derived. It is proven wrong, when a prediction turns out to be false. It has to be consistent with all observations prior to its formulation. And of course to be a theory it has to have been tested lots of times. Now, once it stops being predictive it can be proven, but since it makes no further predictions it´s not a hypothesis, lest a theory. Your answer to eversbane implied that you are willing to stretch the meaning of hypothesis and theory to include things that make no predictions.

Nope. There is no claim that anything "vanishes." QM simply is irrelevant.

There´s your claim that macroscopic systems are not just approximately deterministic, but actually deterministic. That is a claim that the stochasticity does not merely have effects too small to be relevant (as noted, that´s the position David has taken), but that it has NO effects. And since you haven´t so far conceeded that point and withdrawn to the position that stochasticity doesn´t have big enough effects to matter (and I´ve offered you the opportunity to clarify your position in this way several times) concluding that you think it vanishes is fair.

Its a personal interpretation of yours, that it is relevant.

No. The effects of QM are a mathematical necessity. And they can be observed. 13.1% of humans deaths occured through direct effects of QM processes in 2005. If you think something that kills one in 8 members of a population is irrelevant, you need to substantiate it.

It does not matter how many people might share your interpretation. Other interpretations are not ruled out. Your claim that they are, is bogus; it does not follow.

My claim that you can not build deterministic systems out of a finite number of stochastic components is simple mathematics. You can debunk this if you want to: Let Xn be the nth number rolled on a fair, 6-sided die.
Now produce Y=f(X1,X2...XN), so that
a) f explicitly depends on at least one of the Xn
b) N is finite
c) f is a deterministic variable.

Do that, and I´ll concede my point (you´ll probably get a field medal in the bargain as well).
But I didn´t state you were claiming something akin to 1+1=3 for naught - your mission, if you choose to accept it - is showing a proven theorem in probability theory to be wrong.

A great deal in this thread, can be viewed philosophically as Reductionist versus antireductionist, specifically as it applies to the theory of natural selection, but NOT the theory of evolution.

Mjpam, Susu, and others, keep trying to imply that we MUST only talk about evolutionary theory as a whole. This is a forced straw man.

I wouldn´t say that. But we can not neglect interactions and the point I´m stressing is that drift has an effect on selection and therefore selection is stochastic.

They have in large part, throughout this thread, denied that any valid reductionist view exists.

Not at all. The difference is not really reductionism vs. non-reductionism, but rather superposability vs. non-superposability. You can break evolution up into different processes happening at once. However you argue that the effects of these processes can be added - superposition. We´ve argued that you can not do so - the processes feed into each other and their combined effect - quite literaly - is more than the sum of the parts.

I also note that Davids accusation of Marios and me has been that we are too reductionist.

Mjpam foolishly argues that the "modern synthesis" is all that exists or matters. This is a very naive faith position.

Nope. It´s a scientific position, because as has been pointed out, the modern synthesis in its central points has not been falsified. We have included neutral and near neutral theory, though, as well as system theoretic considerations - it´s not the status quo of the 50s (we´re merely pointing out, that some of the arguments brought up against us have been known to be wrong since then).

I'm a molecular biologist. That essentially makes me a reductionist. However, I also fully understand the antireductionist view. Mjpam and others have indicated throughout this thread, an inability to understand the reductionist view, and even go so far as to deny its existence. That lack of vision and understanding has led to apologetic argumentation full of fallacies and derision.

The funny thing is that there´s nothing more reductionist than population genetics. And that´s where we´ve been arguing from. You are the one who has argued against looking at gene frequencies and advocated something as "holistic" as phenotypes.

What Mjpam and others are doing, is insisting that the Modern Synthesis only covers Group Selection. Nothing could be further from the truth. Anyone who has come to this thread and naively believed Mjpam is some type of authority, has been duped.

None of us is advocating group selection. What Mayr states here is that selection acts on genes through the phenotypes of individual organisms. It´s not a refutation that selections happens in populations. Quite the opposite: Organismal selection defines the unit of sampling:
The frequency of an allele in a population is n/2N for diploid organisms. N is the number of organisms (and 2N is the number of loci under consideration: 2 per individual organism), n is the number of times the allele is present (0,1 or 2 times for each individual organism).

There is no and there can be no selection on a single organic individual (because that single individual has a fixed allele frequency). There can be selection within an organism - cells with different genomes and different rates of mitosis and apoptosis as in cancer for example: Cellular or somatic evolution also includes selection (in this case the frequency of an allele would be m/2M defined analogously, with M being the number of cells in the organism and m the number of times a cell includes the allele in question).

[recursive prophet]

The debate here was over an almost totally meaningless semantic distinction, with a creationist subtext. If you want to resurrect it , I suggest you begin by defining "stochastic". Then we can have a poll and see who agrees with the claim that "evolution is stochastic" under your definition.
Will you accept the wiki definition below?

"A stochastic process is one whose behavior is non-deterministic in that a state's next state is determined both by the process's predictable actions and by a random element. [emphasis mine] Stochastic crafts are complex systems whose practitioners, even if complete experts, acknowledge that outcomes result from both known and unknown causes. Classical examples of this are medicine: a doctor can administer the same treatment to multiple patients suffering from the same symptoms, however, the patients may not all react to the treatment the same way. This makes medicine a stochastic process."

From my understanding following the thread on Dawkins. it's proposed that 2 of the random elements in selection-and one is all that is needed to comply with the definition above-are genetic drift and reproductive frequency variance. But if you go to RD.net and read just the 6 pages of replies on part 3, it is all covered pretty extensively by scientists working in related fields, such as Susu.exp, Marios, and Dlx2. And there are many other quite knowledgeable posters contributing to the debate, and I can tell you 'mijo'-mjpam-has many there with far more background than I who are in definitely in agreement with his position. Saying he is 'stupid' or that his points are 'silly' doesn't make it so, nor does it constitute a rebuttal. Or does it in this forum?

[articulett]

I told you MZbiologist was good.

"Susu" is the so called expert that Mijo name drops in every post... Susu has said that per his definition of random elections are "pure random chance"--after all--they can be modeled using stochastics. Yes, he's one of those guys who describes everything as random--He reminds me a little of Schneibster mixed with Jim-Bob.

Luckily we can put them on ignore here. A couple of bozos have fallen for it on Dawkins net--but Susu is the best they've got--and what mijo quoted might be his clearest.

I am glad that their overconfidence coupled with incompetence allowed them to display their bozofacedness on two forums for us to enjoy. Please forgive me for playing a role in the return of Mr. "evolution-is-random". I noticed T'ai Chi has been posting again so, with Recursive Prophet, I think we have the "evolution-is-random" trifecta. If only we hadn't all played this silly game before.

I shall put them all on ignore so I can keep my claws retracted.

[The Atheist]

I shall put them all on ignore so I can keep my claws retracted.

Thank god for that.

Ultimately, Arti will create the perfect forum - one where only her own words are visible, because everyone else is on ignore.

"Claws"? More like tentacles, I would have said.

[recursive prophet]

Thank god for that.

Ultimately, Arti will create the perfect forum - one where only her own words are visible, because everyone else is on ignore.

Ultimately? :rolleyes: I suspect that point was passed long ago.:boggled:

[Dr Adequate]

Will you accept the wiki definition below?

"A stochastic process is one whose behavior is non-deterministic in that a state's next state is determined both by the process's predictable actions and by a random element. [emphasis mine] Stochastic crafts are complex systems whose practitioners, even if complete experts, acknowledge that outcomes result from both known and unknown causes. Classical examples of this are medicine: a doctor can administer the same treatment to multiple patients suffering from the same symptoms, however, the patients may not all react to the treatment the same way. This makes medicine a stochastic process."

From my understanding following the thread on Dawkins. it's proposed that 2 of the random elements in selection-and one is all that is needed to comply with the definition above-are genetic drift and reproductive frequency variance. But if you go to RD.net and read just the 6 pages of replies on part 3, it is all covered pretty extensively by scientists working in related fields, such as Susu.exp, Marios, and Dlx2. And there are many other quite knowledgeable posters contributing to the debate, and I can tell you 'mijo'-mjpam-has many there with far more background than I who are in definitely in agreement with his position. Saying he is 'stupid' or that his points are 'silly' doesn't make it so, nor does it constitute a rebuttal. Or does it in this forum? No, round here the endless posts in which mijo's points have been shown to be silly constitutes a rebuttal.

What is silly, stupid, and, indeed, deceitful, is pretending that any true statement about evolution justifies the statement that: "evolution is random". It does not, any more than the statement about medicine given in the wiki article would justify the statement that "medicine is random"; or true statements about the law of mass action would justify the claim that "chemistry is random"; or true statements about quantum electrodynamics would justify the statement that "optics is random"; or true statements about chaos theory would justify the statement that "gravity is random".

[my_wan]

Adding air to the tire on your automobile is a stochastic process. When you connect an air tank with higher pressure than your tire to the tire air will pass both directions randomly across that connection. The pressure only increases in your tire because the air molecules are more likely to pass into the tire than out of it due to less air there to begin with.

When a creationist talks about something being random it essentially means God didn't do it. In reality randomness can be viewed as a way to quantify missing knowledge of the system, not missing information or causes. In this sense there is no such thing as randomness in an absolute sense. With this definition a stochastic process is any process where you can predict an average result without knowledge or ability to predict the individual events causing that result. You could make Quantum arguments against this, but then again the very definition of a Quantum system entails incomplete knowledge yet the results remain predictable. Whether or not that knowledge is incomplete due to incomplete knowledge or the information itself simply doesn't exist is an open question.

The fact remains that creationist insisting on attaching a certain ontology to the notion of randomness to maintain the statement, "It couldn't have all happened by accident", doesn't help them. Even accidents are not accidents in physical law, rather accidents only in in our incomplete knowledge to effect the outcome.

[mijopaalmc]

Dr. Adequate-

"Random" and "stochastic" share a technical definition and therefor can be used as synonyms for that particular definition. Creationists are most often not using the definition I am, so I don't see the problem with using them intercahngeably as long as I am explicitly clear as to how I am using them, as I have been on this board.

[Dr Adequate]

Dr. Adequate-

"Random" and "stochastic" share a technical definition and therefor can be used as synonyms for that particular definition. Bollocks.

Creationists are most often not using the definition I am, so I don't see the problem with using them intercahngeably as long as I am explicitly clear as to how I am using them, as I have been on this board. It is perfectly true that if, every time you say "evolution is random" you also explain that when you say "evolution is random" you don't mean that evolution is random, then people will not take your statement that evolution is random as meaning that evolution is random --- assuming that they understand and believe your explanation of how you don't mean what you're actually saying.

They may think that you've gone off your head, but that's your problem.

[Eos of the Eons]

Evolution is as random as the climate in any given area and the changes it causes, which is to say that it is only mostly random if you take snapshots over time and and leave out everything that happened in between.

Overall though...Yeah, a few things are random, but the existing genes that are selected for over time due to many variables is not a random process.

Hm. I could go on and on... redundantly and repetitively.

[T'ai Chi]

Yes, evolution is a random process: http://www.statisticool.com/evolution.htm

Basically, Evolution = NS(RM, OS)

where NS (ie. Natural Selection) is a non-trivial function and is not random, RM is random mutations and is random, and OS is other stuff, some random, and some that is not random.

That is, following Dawkins and many others, evolution is the non-random selection of random variation.

In probability theory, it is known that if X is a random variable, then for a non-trivial function f, f(X) is a random variable.

This topic has been put to rest literally decades ago.

[articulett]

Yes, put to rest by you, T'ai -- an admitted creationist.

Creationists need to believe that evolution is a theory of random chance, because that sounds like a god must have been involved. So long as they don't understand or cannot explain natural selection and how it culls and ratchets the "fittest" from the randomness, they can pretend that it all sounds too impossible. Those who understand natural selection don't have an obsessive need to tell others how it's really "random".

The real question is, does anyone other than you take you seriously?

[Dr Adequate]

Yes, evolution is a random process Don't tell silly lies.

This topic has been put to rest literally decades ago. And yet you still feel obliged to babble out creationist nonsense.

[sol invictus]

Will you accept the wiki definition below?

"A stochastic process is one whose behavior is non-deterministic in that a state's next state is determined both by the process's predictable actions and by a random element."

With that definition, yes - evolution is a stochastic process (note that that is not quite the same thing as saying "evolution is stochastic"). I'd be surprised if anyone here disagrees with that, given that definition.

Saying he is 'stupid' or that his points are 'silly' doesn't make it so, nor does it constitute a rebuttal. Or does it in this forum?

I didn't see anyone say that mijo was stupid (other than you). Anyway, mijo wasn't saying "evolution is a stochastic process" - he was saying "evolution is random", and justifying that with a very stupid definition of random, one which coincides with no one else's. I can't think of any reasons for him to keep insisting on that other than a) he has an agenda and wants to make evolution sound like something it's not, or b) he's incredibly stubborn and can't admit when he's wrong.

[Dr Adequate]

I didn't see anyone say that mijo was stupid (other than you). Well, I wouldn't say that mijo is stupid, because that would break the forum rules. I will, however, say that his argument is stupid, and I'll follow that up by saying that it is vacuous, dishonest, boring, pathetic, and futile.

[technoextreme]

And everyone who explained "how evolution is not pure random chance" (the thread's title, btw) like Dawkins in that clip is shouted down by a couple of people with the intelligence of Mijo (I leave it to the reader to make their own determination).
This is what you get when you have an issue that needs a surgeons touch to explain the issue and you get Jason Vohres. The best example of how dam complicated evolution is given here: http://www.princeton.edu/main/news/archive/S22/60/95O56/index.xml It is evidence that evolution resulted in feedback mechanisms to regulate random mutations.

[Paul C. Anagnostopoulos]

In case it's of any use, here are the definitions from various Penguin dictionaries:

Mathematics:
random: A random sample of r items from n is a selection in which each item has an equal chance of selection.

stochastic: Implying random variation, generally used to describe systems which are not deterministic rather than systems which are deterministic apart from a random error.

Physics:
random: not defined

stochastic process: A process resulting from the random behavior of its generators.

Biology:
random: not defined

stochastic: not defined

I would say that calling something simply "random" implies that it is more-or-less uniformly random. Evolution clearly is not.

Much ado about not much.

~~ Paul

[articulett]

This is what you get when you have an issue that needs a surgeons touch to explain the issue and you get Jason Vohres. The best example of how dam complicated evolution is given here: http://www.princeton.edu/main/news/archive/S22/60/95O56/index.xml It is evidence that evolution resulted in feedback mechanisms to regulate random mutations.

That's a good description... and it makes sense that genomes would "evolve" similarly to phenotypes... with some parts being more plastic and some parts highly conserved.

But creationists have jumped on this one in a funny way too... like god designed genomes to evolve or something like that. There was an interesting commentary on Pharyngula when the article first came out--though, this isn't really "new" information.

[technoextreme]

But creationists have jumped on this one in a funny way too... like god designed genomes to evolve or something like that.
Well it isn't that hard to explain that fallacy. Scientists are using control theory which primairly was developed by engineers/mathematicians to explain their observations. Therfore one can jump to a faulty conclusion that since your thermostat works on the same principle it means that someone had to design those proteins. Of course they are ignoring the fact that control theory dictates how stable dynamic systems behave and it would be very odd if it would deviate from that theory because that would mean that something is horribly wrong with that theory. Its not new because the same concepts have been applied to chemotaxis and homeostatis with similar observations. Ironically, engineers are stealing the those observations to use back into control theory.

[recursive prophet]

When a creationist talks about something being random it essentially means God didn't do it. In reality randomness can be viewed as a way to quantify missing knowledge of the system, not missing information or causes. In this sense there is no such thing as randomness in an absolute sense. With this definition a stochastic process is any process where you can predict an average result without knowledge or ability to predict the individual events causing that result. You could make Quantum arguments against this, but then again the very definition of a Quantum system entails incomplete knowledge yet the results remain predictable. Whether or not that knowledge is incomplete due to incomplete knowledge or the information itself simply doesn't exist is an open question.

The fact remains that creationist insisting on attaching a certain ontology to the notion of randomness to maintain the statement, "It couldn't have all happened by accident", doesn't help them. Even accidents are not accidents in physical law, rather accidents only in in our incomplete knowledge to effect the outcome. [Emphasis mine]


My Wan: You can’t imagine how grateful I was to read your insightful reply. I would beseech you to join in the discussion on Dawkins, as those proposing that NS isn’t stochastic are outnumbered big time and on the ropes. I am but a science/math dilettante with no formal training and an overtaxed memory. I have no dog in this fight, but I can tell you that the RD random thread has overall been a fascinating read. Also, given much of the sentiment I’ve witnessed so far, I suspect this topic will have a short shelf life here, and possibly be locked.

You make a valid point re: the parameters of randomness. I image it as being like the islands of increasing order in the entropy stream. Ultimately disorder will prevail, but splotches here and there can buck the trend for while. So the key question would seem to be what is the practicality of assuming a stochastic, versus deterministic, versus hybrid process in various areas of research. There are some excellent posts on RD that cover this area, as well as many other questions I’d never considered before. As soon as I reach the prerequisite 15 replies I’ll be posting a number of links rather than clutter the thread with long comments. I find it somewhat amusing the way so many on both sides of the debate dismiss their opponents as fools, though I can document on RD that it is VERY lopsided. Most ad homs come from those arguing against NS being a stochastic process. They could really use your rational help.

For me there is an inescapable dichotomy here; if a process is not deterministic, it is by definition stochastic. Einstein believed the universe was deterministic, but advances in quantum mechanics have indicated otherwise. A few on RD argued the quantum world is irrelevant, but boy did they get hit with some very detailed refutations on that one, which I may also link here.

I find it ironic so many jump on the ways creationists use randomness. For me, the idea of a deterministic universe implies there is a ‘determiner,’ and would be far more compatible with ID than a random one? How could randomness co-exist with an omnipotent ruler of the heavens and earth? I would appreciate enlightenment on this, as I don’t ever engage in debates with believers on any aspect of faith. I envy those with so much time to spare doing so.
No, round here the endless posts in which mijo's points have been shown to be silly constitutes a rebuttal.

What is silly, stupid, and, indeed, deceitful, is pretending that any true statement about evolution justifies the statement that: "evolution is random". It does not, any more than the statement about medicine given in the wiki article would justify the statement that "medicine is random"; or true statements about the law of mass action would justify the claim that "chemistry is random"; or true statements about quantum electrodynamics would justify the statement that "optics is random"; or true statements about chaos theory would justify the statement that "gravity is random".
Well pilgrim, I’m not from ‘round here. So I’m either a drifter from the planes of cyberspace just looking for answers, or a closet fundamentalist in partnership with ‘mijo’ (translation from Spanish ‘my son’)seeking only to muddy the waters of evolution. We might be members of a secret cabal, struggling to undermine any faith in science. When the economy truly collapses and people are desperate, we will collectively set the clock back and have the population under our control via the myths built around the icons already in place.

We will live high, as the masses will rely; on our message from on high.
Eat hey and pray, is what we we’ll say. There’ll be pie, in the sky, when you die.

Even talking about evolution will become a crime, and we have triangulated the locations of all our antagonists online so they can be graduates in the first reeducation class. For soon, we the Cognoscenti, will rule the metaverse!![Begin maniacal laughter soundtrack.] :jaw-dropp

I need to do a word search and see how many times I’ve seen one side in this debate or the other compare their opponents with creationists. However, only those arguing evolution is not stochastic, from what I’ve witnessed, have accused their opponents of actually being creationist. Seems quite a histrionic charge, IMHO, but I’ve seen it made often.

Take a second look at the reply by susu, where he addresses MZ’s points one by one. There are countless follow ups just like that on the thread, and rarely do they get a reply anywhere near as conclusive. MZ has gotten much better in the last month or so, but still leaves many key statements unacknowledged and questions unanswered. Usually the replies are those such as appeared here; they respond to a couple points and ignore everything else. I was hoping someone might respond to susu’s post in a detailed way, but can’t say I’m surprised no one did.

Evolution is not random, eh? So answer this, DA. Is mutation random? If you say no, I won’t bother you with further questions. If yes, then how can a process be partly random? Isn’t that akin to being partly pregnant? Isn’t each throw of a loaded dye still stochastic?

Technoextreme: Yes, evolution is incredibly complicated and those wishing to dismiss the question of how the process functions as trivial or just semantics remind me of an Einstein quote which I shall paraphrase: “Always make things as simple as possible, but not more so.”

Tai' Chi: Just for the record, are you a confessed creationist, or can't you say as you're part of the Cognoscenti Cabal I mention above? If so, I hope you remember the secret digital handshake and you got the latest email. You'll need to lay a freshly slaughtered lamb on your keyboard at midnight on the next full moon. :rolleyes:

[articulett]

Well it isn't that hard to explain that fallacy. Scientists are using control theory which primairly was developed by engineers/mathematicians to explain their observations. Therfore one can jump to a faulty conclusion that since your thermostat works on the same principle it means that someone had to design those proteins. Of course they are ignoring the fact that control theory dictates how stable dynamic systems behave and it would be very odd if it would deviate from that theory because that would mean that something is horribly wrong with that theory. Its not new because the same concepts have been applied to chemotaxis and homeostatis with similar observations. Ironically, engineers are stealing the those observations to use back into control theory.

Information begets better information assimilators, replicators, and mutators... data used to build computers begets better computers that assimilate, replicate and mutate (tweak) more and better data which begets better....

DNA is information in a code and operates the same. Information either gets passed on with the chance to evolve into the future-- or it dies out.

All the atoms that build complex things were around long before the complex things were... it takes time for the information to evolve-- and that does not happen "randomly". This is true whether we are talking living things, technology, science, languages, etc.

[sol invictus]

Evolution is not random, eh?

No one said that either. The mistake is thinking it must be one or the other. Is a red and white shirt red? Is it white?

So answer this, DA. Is mutation random? If you say no, I won’t bother you with further questions.

Of course it's not random! There are all kinds of non-random mechanisms that control it, determine when and where it can happen, and whether it gets repaired or passed on.

Of course it's random! Mutations can be caused by a cosmic ray, and what could be more random than that?

Get the point?

If yes, then how can a process be partly random? Isn’t that akin to being partly pregnant? Isn’t each throw of a loaded dye still stochastic?

Why, after carefully defining "stochastic" the way you did, are you asking that? And why are you conflating it with "random"?

Were you aware that most smoke detectors rely on the decay of a radioactive substance to function? What could be more random than when a radioactive atom decays... so are smoke detectors random?

[Dr Adequate]

[Emphasis mine]


My Wan: You can’t imagine how grateful I was to read your insightful reply. I would beseech you to join in the discussion on Dawkins, as those proposing that NS isn’t stochastic are outnumbered big time and on the ropes. I am but a science/math dilettante with no formal training and an overtaxed memory. I have no dog in this fight, but I can tell you that the RD random thread has overall been a fascinating read. Also, given much of the sentiment I’ve witnessed so far, I suspect this topic will have a short shelf life here, and possibly be locked.

You make a valid point re: the parameters of randomness. I image it as being like the islands of increasing order in the entropy stream. Ultimately disorder will prevail, but splotches here and there can buck the trend for while. So the key question would seem to be what is the practicality of assuming a stochastic, versus deterministic, versus hybrid process in various areas of research. There are some excellent posts on RD that cover this area, as well as many other questions I’d never considered before. As soon as I reach the prerequisite 15 replies I’ll be posting a number of links rather than clutter the thread with long comments. I find it somewhat amusing the way so many on both sides of the debate dismiss their opponents as fools, though I can document on RD that it is VERY lopsided. Most ad homs come from those arguing against NS being a stochastic process. They could really use your rational help.

For me there is an inescapable dichotomy here; if a process is not deterministic, it is by definition stochastic. Einstein believed the universe was deterministic, but advances in quantum mechanics have indicated otherwise. A few on RD argued the quantum world is irrelevant, but boy did they get hit with some very detailed refutations on that one, which I may also link here.

I find it ironic so many jump on the ways creationists use randomness. For me, the idea of a deterministic universe implies there is a ‘determiner,’ and would be far more compatible with ID than a random one? How could randomness co-exist with an omnipotent ruler of the heavens and earth? I would appreciate enlightenment on this, as I don’t ever engage in debates with believers on any aspect of faith. I envy those with so much time to spare doing so.

Well pilgrim, I’m not from ‘round here. So I’m either a drifter from the planes of cyberspace just looking for answers, or a closet fundamentalist in partnership with ‘mijo’ (translation from Spanish ‘my son’)seeking only to muddy the waters of evolution. We might be members of a secret cabal, struggling to undermine any faith in science. When the economy truly collapses and people are desperate, we will collectively set the clock back and have the population under our control via the myths built around the icons already in place.

We will live high, as the masses will rely; on our message from on high.
Eat hey and pray, is what we we’ll say. There’ll be pie, in the sky, when you die.

Even talking about evolution will become a crime, and we have triangulated the locations of all our antagonists online so they can be graduates in the first reeducation class. For soon, we the Cognoscenti, will rule the metaverse!![Begin maniacal laughter soundtrack.] :jaw-dropp

I need to do a word search and see how many times I’ve seen one side in this debate or the other compare their opponents with creationists. However, only those arguing evolution is not stochastic, from what I’ve witnessed, have accused their opponents of actually being creationist. Seems quite a histrionic charge, IMHO, but I’ve seen it made often.

Take a second look at the reply by susu, where he addresses MZ’s points one by one. There are countless follow ups just like that on the thread, and rarely do they get a reply anywhere near as conclusive. MZ has gotten much better in the last month or so, but still leaves many key statements unacknowledged and questions unanswered. Usually the replies are those such as appeared here; they respond to a couple points and ignore everything else. I was hoping someone might respond to susu’s post in a detailed way, but can’t say I’m surprised no one did.

Evolution is not random, eh? So answer this, DA. Is mutation random? If you say no, I won’t bother you with further questions. If yes, then how can a process be partly random? Isn’t that akin to being partly pregnant? Isn’t each throw of a loaded dye still stochastic?

Technoextreme: Yes, evolution is incredibly complicated and those wishing to dismiss the question of how the process functions as trivial or just semantics remind me of an Einstein quote which I shall paraphrase: “Always make things as simple as possible, but not more so.”

Tai' Chi: Just for the record, are you a confessed creationist, or can't you say as you're part of the Cognoscenti Cabal I mention above? If so, I hope you remember the secret digital handshake and you got the latest email. You'll need to lay a freshly slaughtered lamb on your keyboard at midnight on the next full moon. :rolleyes: Two questions occur to me. One is whether you have a point, and the other is whether you are familiar with the phrase "word salad".

[articulett]

No one said that either. The mistake is thinking it must be one or the other. Is a red and white shirt red? Is it white?



Of course it's not random! There are all kinds of non-random mechanisms that control it, determine when and where it can happen, and whether it gets repaired or passed on.

Of course it's random! Mutations can be caused by a cosmic ray, and what could be more random than that?

Get the point?



Why, after carefully defining "stochastic" the way you did, are you asking that? And why are you conflating it with "random"?

Were you aware that most smoke detectors rely on the decay of a radioactive substance to function? What could be more random than when a radioactive atom decays... so are smoke detectors random?

Sure... by their lame-o reasoning all roads lead to random and random becomes a useless word--no more descriptive than "noun". They will say that you are claiming things are "determined" whenever you point out that calling evolution random is misleading at best.

Here--watch me prove that Chemistry is random by Mijo reasoning:

Chemistry is about atoms and everyone knows electrons have random spin-- and nothing can be partly random... if there's any randomness anywhere it it then it is random-- so Chemistry is random! Cool eh. Oh, and Cemistry is a noun. Isn't pedantic insanity fun? If you attempt to clarify you are claiming (per mijo reasoning) that chemistry is "deterministic" --for shame!

[articulett]

Two questions occur to me. One is whether you have a point, and the other is whether you are familiar with the phrase "word salad".

See, that's why I use the ignore button. As my sig illustrates, the incompetents are too incompetent to realize they are the incompetent ones. He's probably over at the Dawkins site telling everyone he and mijo are "winning the debate" here. Ha!

RP thought my_wan was agreeing with him? I though s/he was echoing the common understanding that nothing is really random... except on the quantum level. The bozos define EVERYTHING as random... and then invoke QM to pretend that physicists agree. (The ambiguity in the word "random" sure seems to be used to prop up a lot of woo.)

We live in a very deterministic universe on a macro level where all events are caused by the outcomes in a preceding chain of events. People who want to be clear, don't call this "random". They may use a stochastic model to illustrate some such events, but they do not do the convoluted thing that Mijo does to then claim this means those processes are random!

Why would anyone be so insistent on describing evolution so vaguely, unless they didn't have a clue about how natural selection works and/or they wanted to ensure no-one else did? I find the obsession intriguing, --and the obsessed are fun to talk about even if it's impossible to engage in dialogue with them.

[technoextreme]

Information begets better information assimilators, replicators, and mutators... data used to build computers begets better computers that assimilate, replicate and mutate (tweak) more and better data which begets better....

DNA is information in a code and operates the same. Information either gets passed on with the chance to evolve into the future-- or it dies out.

All the atoms that build complex things were around long before the complex things were... it takes time for the information to evolve-- and that does not happen "randomly". This is true whether we are talking living things, technology, science, languages, etc.
That probably isn't the right comparison. The control systems that Im talking about are mathematical constructs that can take on many different form. This includes chemical, electrical, and mechanical. It doesn't necessairly have to be directly related to programing. It is very weird because I learned that integrators give zero steady state error in a stable system but how that intergrator is created is irelevant. And you were right. This isn't new. The concept that evolution tends to keep favorable mutations is in this paper which is a year old.
http://www.hamilton.ie/systemsbiology/files/2007/ControlAndSystemsBiology.pdf
Technoextreme: Yes, evolution is incredibly complicated and those wishing to dismiss the question of how the process functions as trivial or just semantics remind me of an Einstein quote which I shall paraphrase: “Always make things as simple as possible, but not more so.”
Oooo for christ's sake I just read that quote on a system's biology presentation. Though the definition of complicated is a bit odd here. Bang-bang controllers are the simplest types of controller available. Effectively the proteins are either fixing or not fixing themselves.
http://webber.physik.uni-freiburg.de/~jeti/papers/antritts_vorlesung.pdf

[articulett]

Well...it's off topic, but all the atoms that existed in your computer existed for eons, right? but they couldn't come together and make your computer in the past. Why not? Various information systems had to evolve. They did not have to evolve to make your computer... but information has to evolve to make better "information processors"... just as DNA has to evolve to make better "DNA processors" (copiers/assimilators/tweakers/recombiners etc.)

Or Languages... they evolve from the bottom up (not Tower of Babel style) and when assorted languages come together via a blending of cultures or the alterations of youth culture, then they start to evolve together... the stuff that people write and say and utilize gets copied and morphs into the language of the future, and the other stuff becomes obsolete.

Information HAS to code for better information processors... technology has to get "better" (more memory, gigabytes, RAM, etc.)... it can't go "backwards"... it's not "random", it's the only way information CAN get passed on. The only way a branch on a tree can grow is up or out.

If any stretch of DNA does not have what it takes to build an organism that can get it copied... it dies out... the only surviving DNA is the DNA that is best at getting itself copied. If a stretch of DNA is better at getting itself copied by being prone to mutation (by coding for beneficial variety for example) it will be copied preferentially including whatever it is that makes it mutable; whereas a piece of code that makes a key protein or needs to be a certain shape or distance for optimal functioning of replicators will find itself preferentially multiplied in organisms that evolve mechanisms to "conserve" it.

It's just the logical extrapolation of the "selfish gene" or "selfish meme". In this way, information systems are very different than normal physical systems-- because it's the information that evolves-- the products or creatures they code for "die out"--but the successful information is copied exponentially in direct relation to how well the team it's on (genome) builds replicators (organsms)--and evolves.

If someone is bent on calling natural selection "random", how can they even begin to grasp this information that is so essential for understanding how wondrous complex "things" come to be?

How cool is it to live in a time when we can understand this? In the past we had to relegate such mysteries to god, but now we can untangle them by working backwards. Future generations could have no knowledge of how the internet came to be-- past generations would surely think it was magical and miraculous... but all complexity... all emergent properties... they all work by the same basic "evolutionary" algorithm. Information that is best at getting copied, evolves--whether viruses, computer viruses, the Internet --or life. If it doesn't, it stops existing.

I am disgusted with those who think it does justice to call this "random". To me they sound like they haven't got the first clue about the process. They could learn and teach this information to others, but their need to "win" their semantic game makes them obfuscatory clowns.

[articulett]

Sickle Cell trait is a mutation that makes cells prone to an irregular shape which has a survival advantage in regions where Malaria is popular... Heterozygotes have about half of their cells which have this trait, and it doesn't have any disadvantage, but when an offspring inherits two copies of the gene... the shape of the blood cells makes them prone to clotting disorders, infarctions, and infections. They inherit the disease from ancestors who preferentially survived in Malaria prone regions due to having Sickle Cell trait. The protection of heterozygotes and the disease in homozygotes is directly related to the shape of blood cells... but all genes code for proteins and all proteins have "shapes" -- so physics can't help but be intimately involved with genetics... and evolution...

A genome has been compared to musical notes... the 4 base pairs are responsible for all the forms of life just as 8 basic notes are responsible for all of music-- and it's all based on physics at the most basic of levels. Even first life (abiogenesis) has a strong physics and is thought to do with minerals that stick together when water washes over rocks... and the physical properties of water molecules and carbon. The thing that makes life different from other physical properties is that it's information that "gets itself" copied by building information copiers. It's information that evolves--not the specific animal or product they code for--those are just "snapshots". You will die the same species you are born as... but parts of your DNA can be part of what evolves with time.

So even though physics is involved... it's important to distinguish the difference-- because classical physics does not involve copying information so that more "refined" physical specimens can inhabit the future. Information can be considered the directions for building "things"--physics deals with the things themselves. Yes, the code is also "physical"-- that is true for my writing here and it's true for DNA. But don't confuse the code for what it coding for. (The notes are not the music). I can physically change a recipe (the code) which will change the product it codes for. But it's the product that determines whether the code is passed on... and to what extent. Information, in a sense, is always competing to make itself more likely to be copied.

I agree with what you are saying technoextreme... I'm just trying to bypass the nuttery of those who'd call this "random". And I wanted to acknowledge your pdf.

From a physics perspective, "life" is more or less defined as "that which locally pumps away entropy" (at least if we treat machines as extensions of the life forms that built them). Information allows us to assemble matter in increasingly "complex" ways before entropy destroys the individual forms the information passes through.

[recursive prophet]

With that definition, yes - evolution is a stochastic process (note that that is not quite the same thing as saying "evolution is stochastic"). I'd be surprised if anyone here disagrees with that, given that definition.

I didn't see anyone say that mijo was stupid (other than you). Anyway, mijo wasn't saying "evolution is a stochastic process" - he was saying "evolution is random", and justifying that with a very stupid definition of random, one which coincides with no one else's. I can't think of any reasons for him to keep insisting on that other than a) he has an agenda and wants to make evolution sound like something it's not, or b) he's incredibly stubborn and can't admit when he's wrong.
So what is your definition of stochastic where it wouldn’t apply to evolution? And I certainly never said mijo was stupid. As to his hidden agenda, what are you proposing? The Cognoscenti Cabal I mentioned in my last post? Is there no possibility you and a few others here might be wrong?
No one said that either. The mistake is thinking it must be one or the other. Is a red and white shirt red? Is it white?
Afraid this is an over-simplification.
Why, after carefully defining "stochastic" the way you did, are you asking that? And why are you conflating it with "random"?
I'm not conflating them; others have here, repeatedly, so I was replying in kind.
Were you aware that most smoke detectors rely on the decay of a radioactive substance to function? What could be more random than when a radioactive atom decays... so are smoke detectors random?
No, I wasn’t aware of that, but radioactive decay as a process is of course random though the detectors that employ it are not.
“Selection is a process that produces differences in the probability of changes of allele frequencies. Selection is not that change in allele frequencies itself. That´s crucial, because a change in allele frequencies is not in itself evidence of selection."-Susu.exp


Two questions occur to me. One is whether you have a point, and the other is whether you are familiar with the phrase "word salad".
Yep to both; my point is that if mutation is random, how can it result in a deterministic evolution? As to word salad, it's a term generally employed by the effably inadequate.

If a neutral mutation occurs in the genome that in effect has no positive or negative effect then I would say that it is "invisible" to a NS filter. The idea of the filter can only work on something that gives benefit or is detriment to survival and reproduction.

That's where Nearly Neutral evolution comes in (scoot back and have a look at susu's graph/equationsre: sickle-cell anaemia). That's a nice exposition of how you won't have a clear grasp of selection without modeling it fully as a stochastic process.

Beyond that, you've got a problem in any situation where selection pressures aren't static (i.e. almost nontrivial models) - selection pressures will depend on other species genepools (coevolution), the abiotic environment (which can only be approximated as a deterministic variable if it's very, very nearly static) and the species own genepool (frequency-dependent selection). Pretty much everything beyond GCSE evolution requires a grasp of how stochasticity will change/destroy the equilibria found in a simplified deterministic model. Science relies on being able to treat real-world problems with simplified models, but beyond a trivial level you're expected to grasp that these are simplifications and be able to explain when they are justifiable simplifications and when they might be ignoring significant phenomena.

Marios

[Reality Check]

Yep to both; my point is that if mutation is random, how can it result in a deterministic evolution? As to word salad, it's a term generally employed by the effably inadequate.
It sounds like you are at the point that mijo was many, many, many posts ago.
You are assuming that biologists think that evolution is deterministic, i.e. a specific mutation will always result in a specific change in a species. This is wrong.
You are assuming that mutation is random. This is almost correct but there are exceptions. Dr Adequate should be able to give you a more comprehensive answer.
You are asuming that evolution is mutation alone. This is wrong since there is also natural selection.

So the answer to your question is:
If mutation is random then natural selection results in a predictive (but not deterministic) evolution.

[six7s]

Well...it's off topic, but all the atoms that existed in your computer existed for eons, right? but they couldn't come together and make your computer in the past. Why not? Various information systems had to evolve. They did not have to evolve to make your computer... but information has to evolve to make better "information processors"... just as DNA has to evolve to make better "DNA processors" (copiers/assimilators/tweakers/recombiners etc.)It might be off-topic, but its damned interesting!

[articulett]

Thanks six7s, I'm glad someone thinks so.

(Surely I'm not the only geek around.)

[six7s]

... technology has to get "better" (more memory, gigabytes, RAM, etc.)... it can't go "backwards"... Tell that to the poor suckers who 'upgraded' from Windoze XP to Vista ;)

[my_wan]

Yes, evolution is a random process: http://www.statisticool.com/evolution.htm

Basically, Evolution = NS(RM, OS)

where NS (ie. Natural Selection) is a non-trivial function and is not random, RM is random mutations and is random, and OS is other stuff, some random, and some that is not random.

That is, following Dawkins and many others, evolution is the non-random selection of random variation.

In probability theory, it is known that if X is a random variable, then for a non-trivial function f, f(X) is a random variable.

This topic has been put to rest literally decades ago.

When a pool cue hits another ball is the angle and speed the other ball recoils random? No. Air molecules work essentially the same way. Yet we say that the individual speed and angles of the molecules are completely random. Why? Not because they are but because randomness describes what happens just as well as if we measure the mass, direction, and speed of every molecule of air. Randomness only defines the information about cause that we don't have, or need to know, to get the right answers. Randomness is not a thing in itself.

Saying evolution couldn't have happened all by accident is like saying that temperature couldn't have been an accident of all those different speeds of air molecules, very few of which are actually the right mass and speed to get that temperature. If you have the odds against a molecular interaction of 1 in 1x10^100...1000 what does that really mean? When we say your odds of winning the lottery is 1 in 175,711,536 does that mean it's impossible for anybody to win the lottery in your lifetime? The same applies to abiogenesis odds. With countless many individual interactions you have to multiply that by all the molecules on Earth, times the number of times per hour, day, year, billion years, etc. The longer the time the better those odds get. Then you have to consider that there are probably as many planets in this Universe as your given odds. That makes it essentially certain to happen no matter how big you make the odds against it.

So irrespective of your post being technically correct the very meaning (as in thing) generally attached to the term random is bogus.

[sol invictus]

So what is your definition of stochastic where it wouldn’t apply to evolution?

I'm happy with the one you gave. But the point that mijo (and you, it appears) was unable or unwilling to grasp is that short statements like "evolution is random" are almost useless, because evolution is a term that describes a huge set of extremely complex and varied phenomena and processes that takes place over enormous span of time. If you want one word to apply to all of that, it needs to be so general as to be almost meaningless.

And I certainly never said mijo was stupid.

You attributed that statement to posters here, when none had made it. You need to read more carefully.

Is there no possibility you and a few others here might be wrong?

About this? No, not really - at worst we'd be not even right. These kind of semantic questions are too empty and useless to be truly wrong or right about.

Afraid this is an over-simplification.

You consider that to be an over-simplification, but not statements like "evolution is stochastic"? How completely absurd.


No, I wasn’t aware of that, but radioactive decay as a process is of course random though the detectors that employ it are not.

So you believe that a process which relies in a fundamentally important way on random events but which produces outcomes which are highly predictable in some aspects is "not random"?

[Earthborn]

Or Languages... they evolve from the bottom up (not Tower of Babel style) and when assorted languages come together via a blending of cultures or the alterations of youth culture, then they start to evolve together... the stuff that people write and say and utilize gets copied and morphs into the language of the future, and the other stuff becomes obsolete.I'm pretty sure most linguists are convinced that all languages have evolved from a single ancestor language and have diversified. You are right that many languages are disappearing and languages are blending and there is a trend for towards a common second language for all people (which doubtlessly will be English), but the opposite trend is also occurring: increasing technological specialisation requires people to speak increasingly in specialised jargon, and there is also an increasing number of communities speaking in artificial languages (such as Klingon or Quenya).

Information HAS to code for better information processors... technology has to get "better" (more memory, gigabytes, RAM, etc.)... it can't go "backwards"... it's not "random",It is however "Intelligent Design". Perhaps you are right if you mean to say that some form of "Intelligent Design" is likely to evolve as it would be a better "information copier".

If someone is bent on calling natural selection "random",Perhaps you have noticed that there are people who have called evolution "random", but far fewer have called natural selection "random".

how can they even begin to grasp this information that is so essential for understanding how wondrous complex "things" come to be?IMHO the wondrous thing about evolution isn't its complexity. A scrapheap is complex, but just a dump. The wondrous thing about evolution is that it produces things that seem highly ordered. "Randomness" is I think essential in understanding how this came to be, because random things become highly ordered and predictable on a larger scale or over a longer time.

I am disgusted with those who think it does justice to call this "random".Perhaps you get disgusted too easily.

[Earthborn]

From a physics perspective, "life" is more or less defined as "that which locally pumps away entropy" (at least if we treat machines as extensions of the life forms that built them).May I suggest you present this definition in this thread (http://forums.randi.org/showthread.php?t=131035)? It is probably the best I have heard.

[GreyICE]

I would agree with the statement that natural selection is stochastic (stochastic - a system that is governed by both predictable processes (survival of the fittest, et al) and unpredictable, chaotic processes (mutation).

I would disagree with the statement that evolution is random.

Dunno about the Dawkins debate, sounds to me that someone smashed his face in an indefensible position, flailed about for a while, retreated to a defensible position, decided to claim that he was at said defensible position for the entire time, annoyed everyone who wanted to talk to him, and then claimed internet victory. Reading it, it seems to be full of the worst sort of nonsense.

Also, articulett has the same avatar on both forums. Since that saves me paragraphs of pain, I nominate it as a public service.

[TraneWreck]

Why would you argue about evolution in these terms? It seems like you would want to know whether ANYTHING was determined or stochastic before you started applying those terms to specific systems.

It isn't clear to me that there is any instance of "randomness" in the universe. I know people point to quantum theory, but the link between quantum indeterminacy and the sort of macro-scale abitrary events suggested by the opening question are dubious at best. And further, much of quantum theory involves a sort of "black-box" mathematical understanding where we can very accurately predict results without understanding the causal mechanism. This seems less like a proof of randomness than a mystery beyond our current abilities. To apply it to evolution is a massive leap of judgment.

This discussion seems like nothing more than an intellectual sounding smear against evolution. If the universe as a whole is deterministic, then so is evolution. If you can prove it isn't, then maybe we have something to discuss.

[SirPhilip]

There is an ongoing debate-over 2k replies-on RichardDawkins.net as to whether natural selection is stochastic or deterministic. At some point I may post a poll here on the question, but I'd be interested to know what those who follow this thread think. Also, if this has been covered in detail elsewhere I'd appreciate someone directing me to that thread. Thanks. Deterministic in what sense. Living things have observable characteristic agency. Philosophical implications aside, anything that is living, while relative to the environment, move essentially by themselves. It's analogous to Shrodinger's Cat: are biological systems actually fundamentally alive or dead. The answer is both, possessing agency and predictability. Agency, despite it's existence, however is fundamentally untenable.

[Dr Adequate]

Yep to both; my point is that if mutation is random, how can it result in a deterministic evolution? OK, let me ask you a more specific question: do you have a point that doesn't involve arguing against a witless straw man of your own construction?

As to word salad, it's a term generally employed by the effably inadequate. I asked you if you knew what it meant, not if you had dumb fantasies about the people who use the term.

[sol invictus]

Evolution is not random, eh? So answer this, DA. Is mutation random? If you say no, I won’t bother you with further questions. If yes, then how can a process be partly random? Isn’t that akin to being partly pregnant? Isn’t each throw of a loaded dye still stochastic?

No, I wasn’t aware of that, but radioactive decay as a process is of course random though the detectors that employ it are not.


You learned something, I see. Good.

Why would you argue about evolution in these terms? It seems like you would want to know whether ANYTHING was determined or stochastic before you started applying those terms to specific systems.

I think the only sensible definition of random is "unpredictable". Hence regardless of the fundamental nature of the collapse of the wavefunction in QM, it is random, as are some - but only some - aspects of evolution.

OK, let me ask you a more specific question: do you have a point that doesn't involve arguing against a witless straw man of your own construction?

Is there such a thing as a witty straw man?

[articulett]

Is there such a thing as a witty straw man?

Well, there's this one: rm-jqKgkAyc

Okay, maybe it's not witty... but it sings a ditty--that must count for something, right?

[articulett]

or maybe this?

[recursive prophet]

Why would you argue about evolution in these terms? It seems like you would want to know whether ANYTHING was determined or stochastic before you started applying those terms to specific systems.
This is a reasonable point, but it’s like saying we’d need to have a proof for Riemann’s Zeta Hypothesis before we can really understand the primes. The question is which system is falsifiable. That would be determinism.
It isn't clear to me that there is any instance of "randomness" in the universe. I know people point to quantum theory, but the link between quantum indeterminacy and the sort of macro-scale abitrary events suggested by the opening question are dubious at best. And further, much of quantum theory involves a sort of "black-box" mathematical understanding where we can very accurately predict results without understanding the causal mechanism. This seems less like a proof of randomness than a mystery beyond our current abilities. To apply it to evolution is a massive leap of judgment.

This discussion seems like nothing more than an intellectual sounding smear against evolution. If the universe as a whole is deterministic, then so is evolution. If you can prove it isn't, then maybe we have something to discuss.
Nomenclature standardization is extremely important in science. So the question at hand does have semantic aspects, as people are using terms such as drift, phenotype, and stochastic in non-conventional ways. But there is also the issue what working assumption is most widely used and practical in evolutionary science. If you check out the links below all this is covered in detail, or check the sting of comments by 2 of the key debaters on RD.

Deterministic in what sense. Living things have observable characteristic agency. Philosophical implications aside, anything that is living, while relative to the environment, move essentially by themselves. It's analogous to Shrodinger's Cat: are biological systems actually fundamentally alive or dead. The answer is both, possessing agency and predictability. Agency, despite it's existence, however is fundamentally untenable.
I guess you could apply Schrödinger’s analogy here, Phillip, and I recall someone else doing so way back on RD, along with Heisenberg’s Uncertainty. Yet like assuming both evolution and selection are stochastic as your working paradigm is more practical and defendable, I think it makes sense to begin with the working hypothesis that the biological systems are in fact alive. Read the arguments made below.

OK, let me ask you a more specific question: do you have a point that doesn't involve arguing against a witless straw man of your own construction?
How is that? Seems the 'not really stochastic' crowd uses a lot of 'strawmen' allegations in their replies, along with ad homs.
I asked you if you knew what it meant, not if you had dumb fantasies about the people who use the term.
However did you guess? And unlike the way you ignored my query on mutation, I did in fact answer your so very polite and thoughtful question. I wouldn't bother with what's presented below were I you. Just more word salad served up by strawmen. [I'd be such a happy fella, I'd be rich as Rockefeller, if I only had a brain..hopping down the road..]

Beginning of pt. 3 of RD How evolution isn’t purely random thread: http://richarddawkins.net/forum/viewtopic.php?f=4&t=66951

Latest susu posts: http://richarddawkins.net/forum/viewtopic.php?p=1608732#p1608732

Segments of the dialog between susu, MZ, and David. I would welcome any constructive comments on basic flaws in susu's reasoning. If there is not a mention of MZ or David, assume susu is being quoted, as just below. To really appreciate all that has been presented, go to the first link above.

The difference is not really reductionism vs. non-reductionism, but rather superposability vs. non-superposability. You can break evolution up into different processes happening at once. However you argue that the effects of these processes can be added - superposition. We´ve argued that you can not do so - the processes feed into each other and their combined effect - quite literally - is more than the sum of the parts.

None of us is advocating group selection. What Mayr states here is that selection acts on genes through the phenotypes of individual organisms. It´s not a refutation that selections happens in populations. Quite the opposite: Organism selection defines the unit of sampling:
The frequency of an allele in a population is n/2N for diploid organisms. N is the number of organisms (and 2N is the number of loci under consideration: 2 per individual organism), n is the number of times the allele is present (0,1 or 2 times for each individual organism).

There is no and there can be no selection on a single organismal individual (because that single individual has a fixed allele frequency). There can be selection within an organism - cells with different genomes and different rates of mitosis and apoptosis as in cancer for example: Cellular or somatic evolution also includes selection (in this case the frequency of an allele would be m/2M defined analogously, with M being the number of cells in the organism and m the number of times a cell includes the allele in question).

If you argue that drift and selection are one thing, then the combination is stochastic. If you argue that the two are separate then you have to note that drift affects(!) selection and renders selection stochastic. No matter if you view selection/drift as one process or selection and drift separately, selection is stochastic.

See, selection can be defined as separate from drift either through stating that selection happens whenever the sampling is not causally unaffected by phenotypes (the view in the article), i.e. drift is what happens under neutrality. In that case drift would be a special case of selection (and indeed it´d be hard to draw a distinction and selection would be stochastic without any further discussion). Another way would be to divide the change your model predicts into an expected value and a centered drift component. The expected value would not be a random variable if the fitness of individuals was independent (!) of allele frequencies and only in such a case would selection be a deterministic component. I don’t think there’s any other way of describing selection as deterministic. But if you argue that selection is independent of any allele frequency, you have to make the case that a single allele can provide the same benefit to fitness, regardless of the remainder of the genome (even finding out that it would always provide a benefit wouldn’t be sufficient, it´d have to be the exact same benefit).

Maynard-Smith was far later (he was Haldanes student in the late 40s). While the work of Wright, Fisher and Haldane was covering most bases in the 20s and 30s as a paleontologist interested in macro evolutionary patterns and speciation I don’t consider the synthesis "finished" until Mayrs "Systematics" ('42), Simpson’s "Tempo and Mode" ('44) and Renschs "Evolution above the species level" (published in German in '47, a revised English version had to wait until '60) had been published. This "second wave" connected the results of the first one to different fields in biology.

The philosophical claim, if I understand you correctly is either viewing selection and drift as separate or as one and the same process. I´ve stated that my argument holds either way and in that case I do not make such a claim. I´ve stated before that I like to view the two as one process - that´s a philosophical claim. But I´m also arguing that if you keep them separate, selection is still always stochastic under the condition that no gene provides the same fitness effect on every thing that is, ever was and ever will be alive. And if there was such a gene (of which I´m extremely doubtful), it´d be the extreme exception and by no means sufficient to call selection as a whole deterministic (because for all other alleles it is stochastic).

MZbiologist wrote: You are correct that it needs to be demonstrated that fitness was independent of non selected alleles.

Of all of them. It needs to be demonstrated that the same effect on fitness is produced regardless of the other alleles. That adds another requirement: it has to be able to function independently - i.e. in the absence of other genes. It has to be an autonomous self-replicating unit. Which means that the fitness effect it provides is total (if it was absent as well, you´d have no genome and no organism - a fitness of 0). So you just have to show that there is an allele which is present in every living thing, which can live independently from other genes and without which all organisms would die. Good luck.

I have the strong suspicion that there´s no gene that would produce the same fitness effect regarless of the genome it´s part of, yes. Especially as noted above, it´d have to be a gene that no organism would be viable without and therefore present in all organisms. I´d place bets that even the closest you could get to that would allow some synonymous 3rd base substitutions. What you propose is so outrageous that I do demand evidence.

MZbiologist wrote: All you have done, is shifted the burden of proof.

The andromeda nebula is made of cheese eaten by pink unicorns wearing spandex pants singing "Living la vida loca" all day. If you expect me to substantiate that, I´ll say that you´re shifting the burden of proof and my view is just as valid as yours, it´s just a different view...

I´m declaring that we have empirical evidence that contradicts the notion above: Deterministic models fail to predict the outcome correctly. In physics I have noted there is an experiment - based on Bells inequality - that could test whether a hypothetical model could exist - we knew that we couldn´t find rules deeper than the stochastic QM ones before, but the EPR experiments have shown that there are no hidden rules either. You can´t predict who will get cancer accurately (though you can give probabilities). And since cancer is the direct result of QM processes, we know that there are no rules that would allow even a being that could break the laws of physics and exactly find out both momentum and position of all particles in the universe at some point could not predict these. It´s not philosophy: It´s empirical science. If there were hidden deterministic rules, Bells inequality would hold - but it doesn´t. I state that observed "determinism" is a result of the law of large numbers.

DavidMcC wrote: Another way of putting my point about evolutionary pressure is to make an analogy with Ohm's law (V=I*R). If it was still not possible to measure voltages, you presumably would claim that there is no such thing as an electric field.

Voltages aren´t the electric field (the Electric field is a vector field, a voltage is a difference in electric potential which is a scalar field).
More to the point: There is no such thing as an electric field. All fields are mathematical abstractions used in models to allow us to make predictions. But you can get rid of the idea of fields without changing the predictions of physical theories (it just makes the math more complicated - Newton worked out gravity without the notion of "fields" and that was one of the things that made his mechanics accessible to only a few people, "fields" are a great teaching tool, because they simplify the maths a lot).
Voltages, i.e. differences in electric potential don´t exist either (because the scalar potential field is as much a mathematical abstraction as the electric field), a voltmeter basically measures current (i.e. charge per time or really electrons per time).

Now what these things are is useful mathematical abstractions. They are helpful in making predictions and ultimately that´s what science delivers: Instersubjectively testable predictions. A theory is a huge amount of such predictions rolled into a framework that allows you to deduce the predictions. And within a theory it makes sense to give concepts labels for easy reference. But it´s a key mistake to "reify" such abstractions and think that they have some existance outside of theory.

MZbiologist wrote:The vast majority of biological science utilizes hypothesis testing in this way, but you obviously know this. So basically, you are saying "lets replace one assumption with another."


Not at all: Determinism has been falsfified and stochasticity is not an assumtion anymore.
MZbiologist wrote:Again, this is an absurd level of reductionism. Falsification of deterministic behavior for particles has no meaning at macro scales. Genes, phenotypes, and their organismic vehicles, are not entangled wave particles.

Entanglement is important for running the EPR-type experiments, but what they show holds for all quantum objects. Now organisms, DNA molecules actually are built out of quantum objects, I hope we can agree on that. And you can not build a deterministic system from stocahstic components, it´s mathematicall impossible. Now you can argue that the stochasticity doesn´t produce effects we can not ignore, but that makes the system approximately deterministic, not actually deterministic. You´re not arguing that, because that option has been offered to you repeatedly - your argument has been that there are actually deterministic systems.

Now on a molecular level quantum effects matter and they certainly matter for mutations, the classical case of a mutation produced by radiation is fully dependent on quantum effects.

MZbiologist wrote:If you are going to insist that Bell's quantum results are relevant, you need demonstrate such is the case with more than philosophical assertion. That means you need to offer empirical evidence that quantum considerations percolate up to the macro scale and prevent any and all systems from ever being deterministic.


Nope. To prove the system stochastic I only need to show that it does contain a single stochastic component. Now, I have to show that the quantum effects percolate up to the macro scale if I want to argue that this produces effects on the macro scale that make deterministic models useless (i.e. I want to show that the errors you make if you construct a deterministic model are too big).

Since mutations are produced by quantum effects, so are somatic mutations. And some somatic mutations lead to cancer. Cancer kills organisms. And thus reduces the number of offspring they would have had if they hadn’t got cancer. So you do get a stochastic component into the number of offspring an organism has. This is the justification for the formula given earlier which plots fitness against the probability of fixation. As has been noted the effects of this are pronounced in small populations and in cases where changes in frequency also affect selection, i.e. when there’s co evolution and frequency dependent selection. Do we have empirical evidence? Yup: Every time two clonal copies of a cell take different amounts of time before they split again.

MZbiologist wrote:To argue quantum effects are relevant to selection, you MUST demonstrate that a stochastic quantum event is part of an unbroken chain of causation that affects a macro level "choice." Good luck with that.

An uranium nucleus decays producing a high energy photon, send out in some direction stochastically. Provided it goes in a certain direction it meets an oxygen atom in a DNA strand, just while it´s single stranded during mitosis. It breaks up the molecule through ionizing the oxygen atom, and removes the attached BP. This results in a loss of functionality of a gene coding for the receptor that takes the "order" for apoptosis: That cell won´t die if it´s getting the signal for programmed cell death. Since the cell and it´s daughter cells do not die as readily as other somatic cells, they stay around, with time forming a tumor. That tumor results in the death of the organism the cell was a part of, which just happens to be a cat with the "turn left" allele. This increases the expected number of offspring mice with the "turn left "allele will have (their fitness), but does not affect the expected number of offspring for mice with the "turn right" allele. Thus the relative fitness of the two alleles in the mouse population is altered. Tada! An unbroken chain from a quantum event to selection.

[GreyICE]

Does anyone actually know the definition of stochastic? Anyone? I'm starting to think the answer really is no, since the entire thread has demonstrated a complete inability to define the term.

(hint: weather is stochastic. As is pressure)

[articulett]

Because stochastic processes are sometimes referred to random processes (because they involve random inputs), some people pretend that they mean the exact same thing.

In statistics a basic definition of "stochastic" is: Involving or containing a random variable or variables

Stochastic is generally used in regard to modeling potential outcomes. Here's an example of a stochastic model of elections.

http://www.columbia.edu/~ks20/Erik-sig-academic.pdf

Obviously, this does not mean that elections are "random" nor that it makes any sort of coherent sense to describe elections as stochastic if one were trying to explain what elections were and how they differ from, say, lotteries.

Basically, Recursive Prophet and Mijo are claiming that, because evolution contains "randomness", it's fair to call it "random" or "stochastic" (which they assert is synonymous with random.) If you find this misleading or unclear, they will imagine that you are saying "evolution is deterministic". This is mijo's favorite delusion, in fact. Oddly enough, it's a favorite straw man of creationist in their description of evolution (see Thai Chi, Behe and Dembski). They'd like people to believe that evolution is a theory of "random chance". It isn't.

[recursive prophet]

Does anyone actually know the definition of stochastic? Anyone? I'm starting to think the answer really is no, since the entire thread has demonstrated a complete inability to define the term.
GreyICE-I presented my working definition from Wiki below earlier, and none refuted it.

"A stochastic process is one whose behavior is non-deterministic in that a state's next state is determined both by the process's predictable actions and by a random element."
Because stochastic processes are sometimes referred to random processes (because they involve random inputs), some people pretend that they mean the exact same thing.
So tell us, Articulett, is evolution stochastic or not? You’ve OFTEN claimed it’s not random. How are you using random-definition please-when you make this statement? And show us the quote where I have said random=stochastic.

Stochastic and random are synonyms; look in any dictionary. But as you ostensibly know, that doesn't mean they are equal or should be applied in the same way in describing NS. Much as I hate to insert your much overused term, this is indeed your favorite 'straw man.' The title of this thread asks if evolution is stochastic, not random. So is it? Simple question; perfect for a mind like yours. :D

[GreyICE]

[QUOTE]
Stochastic and random are synonyms; look in any dictionary. But as you ostensibly know, that doesn't mean they are equal or should be applied in the same way in describing NS. Synonym - I do not think that word means what you think it means.

[recursive prophet]

GreyICE: Ooookay. What do you think I think synonym means? I'm not talking about different names for a single taxon, if that's what has you confused, but a word that is 'similar' to another word. Not exactly the same, generally, or it would be superfluous.

[my_wan]

GreyICE: Ooookay. What do you think I think synonym means? I'm not talking about different names for a single taxon, if that's what has you confused, but a word that is 'similar' to another word. Not exactly the same, generally, or it would be superfluous.

Saying "Stochastic and random are synonyms" is like say soup and water are synonyms. By 'similar' independent of taxon would mean cat is a synonym of dog, i.e., four legged animals? If you go with the definition of 'embodiment' of something then consider random embodies unpredictability whereas stochastic embodies predictability in spite of containing random determinants, like temperature. If temperature was random like the underlying determinants spontaneous human (etc) combustion would be quiet common.

Random entails a variable in which the value is unpredictable, with the information given, though restrained by the degrees of freedom. Stochastic entails a process that contain random variables as determinants yet have results that are quiet predictable.

[sol invictus]

GreyICE: Ooookay. What do you think I think synonym means? I'm not talking about different names for a single taxon, if that's what has you confused, but a word that is 'similar' to another word. Not exactly the same, generally, or it would be superfluous.

Actually you gave a definition of "stochastic process", not "stochastic". Do you think "stochastic" and "random" are close enough that the statement "evolution is a stochastic process" implies the statement "evolution is random"? If so, why don't you give us the definition of "random" you're using, so we can discuss it?

And are you going to address this?
Evolution is not random, eh? So answer this, DA. Is mutation random? If you say no, I won’t bother you with further questions. If yes, then how can a process be partly random? Isn’t that akin to being partly pregnant? Isn’t each throw of a loaded dye still stochastic?

No, I wasn’t aware of that, but radioactive decay as a process is of course random though the detectors that employ it are not.

You completely reversed your position on this, the most essential question in this debate. That's fine, but you should acknowledge it when you suddenly flipflop 100% on the central issue under discussion. It's confusing for your interlocutors otherwise.

[articulett]

I'm sure he'll get right back to you as soon as he checks with the sources that he imagines are knowledgeable, and looks in his thesaurus for words like ostensibly and superfluous that he thinks makes him sound smart.

Plus, he has to come up with something that he imagines is a clever put down.

You know-- the mijo technique of obfuscation and pedantic nothingness.

[GreyICE]

Saying "Stochastic and random are synonyms" is like say soup and water are synonyms. By 'similar' independent of taxon would mean cat is a synonym of dog, i.e., four legged animals? If you go with the definition of 'embodiment' of something then consider random embodies unpredictability whereas stochastic embodies predictability in spite of containing random determinants, like temperature. If temperature was random like the underlying determinants spontaneous human (etc) combustion would be quiet common.

Random entails a variable in which the value is unpredictable, with the information given, though restrained by the degrees of freedom. Stochastic entails a process that contain random variables as determinants yet have results that are quiet somewhat predictable.

Fixed. Unless you think the weather is 'quite predictable.'

[mijopaalmc]

This is getting really old. The argument that "stochastic" and "random" are not synonymous seems to be based on the presupposition that "stochastic" in "stochastic process" has a different meaning that "stochastic" when it stands alone. As far as I have been able to discern from nonspecialist English dictionaries and reputable scientific sources this is simply not the case.

For instance The American Heritage Dictionary defines "random" and "stochastic" as:

random (răn'dəm)

adj.

Having no specific pattern, purpose, or objective: random movements. See synonyms at chance.
Mathematics & Statistics. Of or relating to a type of circumstance or event that is described by a probability distribution.
Of or relating to an event in which all outcomes are equally likely, as in the testing of a blood sample for the presence of a substance.

idiom:

at random


Without a governing design, method, or purpose; unsystematically: chose a card at random from the deck.



stochastic (stō-kăs'tĭk)
adj.


Of, relating to, or characterized by conjecture; conjectural.
Statistics.
Involving or containing a random variable or variables: stochastic calculus.
Involving chance or probability: a stochastic stimulation.




Moreover, the National Research Council defines a "stochastic process" as:

stochastic process (http://books.google.com/books?id=WxQhVsraqBQC&pg=PA154&lr=&as_brr=3&as_pt=ALLTYPES&source=gbs_search_s&cad=0#): Opposite of a deterministic process in probability theory. Instead of dealing only with one possible "reality" of how the process under time (as is the case, for example, for solutions of an ordinary differential equation), in a stochastic or random process there is some indeterminacy in its future evolution described by probability distributions. This means that even if the initial condition (or starting point) is known, there are many possibilities the process might go to, but some paths more probable and others less.

As one can easily see, even though definition (3) of "random" does restrict random events to being equiprobable, (3) is not the definition that "random" shares with "stochastic". Instead, "random" and "stochastic" share definitions (2) and (2)(b), respectively, which can also be found in the NRC's definition of "stochastic process" as "in a stochastic or random process there is some indeterminacy in its future evolution described by probability distributions". Thus, the insistence that "random", "stochastic", and "stochastic process" are not in some sense interchangeably if one takes the time to carefully define the terms is a red herring.

[GreyICE]

The key that you are looking for is that random by definition (check your own definitions) says that there is no governing purpose. Stochastic processes can have direction and purpose, but contain a random element.

[mijopaalmc]

The key that you are looking for is that random by definition (check your own definitions) says that there is no governing purpose.

That is one of the definitions of "random".

Stochastic processes can have direction and purpose, but contain a random element.

Just because stochastic processes have long-term asymptotic behavior does not mean that they are not random.

[GreyICE]

That is one of the definitions of "random". Okay, counting failure. That's pretty damn sad.



Just because stochastic processes have long-term asymptotic behavior does not mean that they are not random. Yes it does. They are not random.

[six7s]

That is one of the definitions of "random".Mijo, only you (and, perhaps, the patron saint of bollocks) know why, where, when and how often you have started threads on this topic

That you continue to ignore the many, many simple, cogent, concise and coherent replies and, instead, persist in stretching the truth to the point of distortion suggests that you are wasting time looking to the virtual world for help with a real world problem

[mijopaalmc]

GreyICE-

Systems that are described as "directionless" cannot be simultaneously described as "equiprobable" or "probabilistic". Therefore, to avoid contradiction, one must choose a definition of "random" to use in a specific context. In the context of the peer-reviewed literature on evolutionary biology, the definition that I highlighted had been used for over half an century.

[Third Eye Open]

Random = not predictable
Evolution = Predictable
Evolution != Random

[mijopaalmc]

Random = not predictable
Evolution = Predictable
Evolution != Random

Where is "random" defined as being necessarily unpredictable?

As I have pointed out before, if we use either definitions (2) or (3) from The American Heritage Dictionary, the long term behaviors of a system so described becomes predictable, yet both are valid definitions of "random". Therefore, the definition of "random" does not necessarily depend on the system being predictable.

[mijopaalmc]

Mijo, only you (and, perhaps, the patron saint of bollocks) know why, where, when and how often you have started threads on this topic

That you continue to ignore the many, many simple, cogent, concise and coherent replies and, instead, persist in stretching the truth to the point of distortion suggests that you are wasting time looking to the virtual world for help with a real world problem

I don't ignore the posts to which you refer; I consider them irrelevant, because the belie a deep ignorance of evolutionary biology since the early 20th century.

[GreyICE]

GreyICE-

Systems that are described as "directionless" cannot be simultaneously described as "equiprobable" or "probabilistic". Therefore, to avoid contradiction, one must choose a definition of "random" to use in a specific context. In the context of the peer-reviewed literature on evolutionary biology, the definition that I highlighted had been used for over half an century.

Your first statement is just demonstrably false. Dice rolls have no specific pattern. Dice rolls are equiprobable. Your second statement is false too. That's not a definition of random that is ever used, random systems don't pick and choose outcomes.

[Third Eye Open]

Where is "random" defined as being necessarily unpredictable?



An example of something that is random and predictable would be nice.

If something is predictable, how are we predicting it without knowing the causes behind it?

If we know the causes behind something, how can you still call it random?

[mijopaalmc]

Your first statement is just demonstrably false. Dice rolls have no specific pattern. Dice rolls are equiprobable.

If you roll a fair die many times (or many dice simultaneously) the mean of the all those rolls will most likely be around 3.5. If you repeat the trial (rolling one die many times or rolling many dice simultaneously) many time, you will obtain a distribution of means centered around 3.5N (where N is the number of dice), and the mean of the sampling distribution approaches the expected value of the stochastic process. In other words, the convergence of random variables toward the expecdted values of the stochastic process contradicts the idea the randomness lacks direction.

Your second statement is false too. That's not a definition of random that is ever used, random systems don't pick and choose outcomes.

Nice anthropomorphization! Bias does not imply choice.

[articulett]

Oh, he can call everything random.

If any part of it has anything to do with probabilities it's "random" to mijo.

It doesn't matter if it's biased, predictable, or covered in chocolate-- Mijo can use semantics to tell you that it's "random". And he'll insist that you just don't understand probabilities if you point out that such a claim is uninformative and misleading at best.

See my sig.

[GreyICE]

If you roll a fair die many times (or many dice simultaneously) the mean of the all those rolls will most likely be around 3.5. If you repeat the trial (rolling one die many times or rolling many dice simultaneously) many time, you will obtain a distribution of means centered around 3.5N (where N is the number of dice), and the mean of the sampling distribution approaches the expected value of the stochastic process. In other words, the convergence of random variables toward the expecdted values of the stochastic process contradicts the idea the randomness lacks direction. You will never roll a 3.5. Ever. Even if you roll 100 billion trillion times. Feel free to try again. The fact that you can find a range of values for a random event doesn't make the event non-random, nor will the number of trials in anyway influence the randomness.



Nice anthropomorphization! Bias does not imply choice. I don't think that word means what you think it means either.

[mijopaalmc]

So, articulett, is a sequence of fair coin tosses not random because you are more likely to get 50% heads and 50% tails instead 100% heads or 100% tails?

[mijopaalmc]

No, the values will never converge towards 3.5. In fact you will never roll a 3.5. Ever. Even if you roll 100 billion trillion times. Feel free to try again.
Look up expected value (http://en.wikipedia.org/wiki/Expected_value). Again, the mean of the sampling distribution of means converges on the the expected value of the stochastic process with an ever decreasing variance.

I don't think that word means what you think it means either.
The better word is "anthropomorphism (http://www.google.com/url?q=http://www.answers.com/anthropomorphism&r=67&sa=X&oi=dict&ct=d&usg=AFQjCNG-LNiQ3FhrpkmpMmgICfGwmZLEgA)".

[Third Eye Open]

So, articulett, is a sequence of fair coin tosses not random because you are more likely to get 50% heads and 50% tails instead 100% heads or 100% tails?

The ratio of heads to tails is predictable because we know the coin is fair and has two sides.

The result of the next toss is not predictable because the coin is fair.

[TraneWreck]

This is a reasonable point, but it’s like saying we’d need to have a proof for Riemann’s Zeta Hypothesis before we can really understand the primes. The question is which system is falsifiable. That would be determinism.


I am certainly no mathematics expert, so it's entirely possible I don't fully understand the Riemann Zeta hypothesis, but from what I have read, it deals with the distribution of prime numbers, not whether they exist. Prime numbers are defined (divisible by only 1 and itself) independently of a Zeta hypothesis proof. In order for evolution to be stochastic, by contrast, it must be proven that the universe actually operates in a non-deterministic way. I am not convinced of this.

Quantum indeterminancy, for example, merely says we don't understand the causal mechanism of electron location (among other claims). It doesn't positively assert that there is no cause. This distinction is conceptually crucial, though it may have limited practical relevence.


Nomenclature standardization is extremely important in science. So the question at hand does have semantic aspects, as people are using terms such as drift, phenotype, and stochastic in non-conventional ways. But there is also the issue what working assumption is most widely used and practical in evolutionary science. If you check out the links below all this is covered in detail, or check the sting of comments by 2 of the key debaters on RD.


I'm fully willing to accept that what's useful as a matter of practicality for evolutionary scientists may not necessarily be conceptually satisfying. I see weather along the same lines. It's clearly practical to deal with it as a stochastic system, but I find the claim that there are random elements present in weather patterns dubious.

Weather is unpredictable because it is complex, not because it's random. If we had enough information, it would be perfectly deterministic. I don't know if human technology will ever be advanced enough to keep track of all relevent factors influencing the weather, but assuming we could, would you still view weather as a stochastic system?


None of us is advocating group selection. What Mayr states here is that selection acts on genes through the phenotypes of individual organisms. It´s not a refutation that selections happens in populations. Quite the opposite: Organism selection defines the unit of sampling:
The frequency of an allele in a population is n/2N for diploid organisms. N is the number of organisms (and 2N is the number of loci under consideration: 2 per individual organism), n is the number of times the allele is present (0,1 or 2 times for each individual organism).

There is no and there can be no selection on a single organismal individual (because that single individual has a fixed allele frequency). There can be selection within an organism - cells with different genomes and different rates of mitosis and apoptosis as in cancer for example: Cellular or somatic evolution also includes selection (in this case the frequency of an allele would be m/2M defined analogously, with M being the number of cells in the organism and m the number of times a cell includes the allele in question).


I'm jumping into this discussion a little late, so I apologize if I miss something obvious.

That being said, I don't understand the sentence I bolded. Earlier you said selection acts on phenotypes. An organismal individual is the expression of that phenotype, thus it seems that selection has to act on those individuals. I read those pragraphs a couple of times, but I still don't see the argument.


If you argue that drift and selection are one thing, then the combination is stochastic. If you argue that the two are separate then you have to note that drift affects(!) selection and renders selection stochastic. No matter if you view selection/drift as one process or selection and drift separately, selection is stochastic.

See, selection can be defined as separate from drift either through stating that selection happens whenever the sampling is not causally unaffected by phenotypes (the view in the article), i.e. drift is what happens under neutrality. In that case drift would be a special case of selection (and indeed it´d be hard to draw a distinction and selection would be stochastic without any further discussion). Another way would be to divide the change your model predicts into an expected value and a centered drift component. The expected value would not be a random variable if the fitness of individuals was independent (!) of allele frequencies and only in such a case would selection be a deterministic component. I don’t think there’s any other way of describing selection as deterministic. But if you argue that selection is independent of any allele frequency, you have to make the case that a single allele can provide the same benefit to fitness, regardless of the remainder of the genome (even finding out that it would always provide a benefit wouldn’t be sufficient, it´d have to be the exact same benefit).


Ok, I think my problem with this debate is becoming more clear. You are understanding "drift" as separate from selection because it affects organisms with no concern for their genetics. Thus, how can evolution be deterministic because drift is unpredictable. Do you view drift as the random component necessary for a system to be stochastic?

I don't see how evolution and selection can be segmented thusly. Why are drift elements random? Let's say a meteor impacts the planet. Organisms that survive do so because of "luck," they happen to be standing in a place protected from the blast. Their phenotype was irrelevent to their survival.

There is nothing random about the meteor hitting the planet, and there was nothing random about the way the blast pattern progressed. Nor was there anything random about the way organisms were affected.

Looking at the system as a whole, if enough information were obtained, every step would be 100% deterministic. Given our position as limited creatures, it seems stochastic. Again, this deterministic/stochastic split seems like more of a practical distinction than an ontological or metaphysical reality. And the argument about the actual metaphysics will take place at the quantum level, making evolution irrelevent to the discussion.



The philosophical claim, if I understand you correctly is either viewing selection and drift as separate or as one and the same process. I´ve stated that my argument holds either way and in that case I do not make such a claim. I´ve stated before that I like to view the two as one process - that´s a philosophical claim. But I´m also arguing that if you keep them separate, selection is still always stochastic under the condition that no gene provides the same fitness effect on every thing that is, ever was and ever will be alive. And if there was such a gene (of which I´m extremely doubtful), it´d be the extreme exception and by no means sufficient to call selection as a whole deterministic (because for all other alleles it is stochastic).


I don't understand why it's necessary to produce such a gene for evolution to be deterministic. It's all situational. The survival of that gene is dependent upon the environment it finds itself in, but it's ability to survive, thrive, or parish will be equally determinsitic in all situations. It seems that once again you're in the position of arguing that drift is by nature random. I see no reason why this has to be.


An uranium nucleus decays producing a high energy photon, send out in some direction stochastically. Provided it goes in a certain direction it meets an oxygen atom in a DNA strand, just while it´s single stranded during mitosis. It breaks up the molecule through ionizing the oxygen atom, and removes the attached BP. This results in a loss of functionality of a gene coding for the receptor that takes the "order" for apoptosis: That cell won´t die if it´s getting the signal for programmed cell death. Since the cell and it´s daughter cells do not die as readily as other somatic cells, they stay around, with time forming a tumor. That tumor results in the death of the organism the cell was a part of, which just happens to be a cat with the "turn left" allele. This increases the expected number of offspring mice with the "turn left "allele will have (their fitness), but does not affect the expected number of offspring for mice with the "turn right" allele. Thus the relative fitness of the two alleles in the mouse population is altered. Tada! An unbroken chain from a quantum event to selection.

I wanted to deal with a few more of your interesting points, but I thought I'd limit my responses a little (I know, I already went on for a while).

This was a nice argument you put together here. It seems to me that this all boils down to whether those quantum events are actually random, or if it's simply the case that we can only currently understand them as random. In a practical sense, I have no disagreement with you, the best modern science only allows us view these systems as stochastic.

I began by asserting that I didn't see why eveolution was relevent to this debate, and your post has cemented my position. The place this debate will take place involves your claim about Bell's Inequalities and whether they support your claim that even a being knowing the position of all particles in the universe could predict things like cancer.

I would happily engage in that discussion with you (I have already learned a great deal in the short time I've spent reading through this string), but I think that it's off-topic as far as the OP of this thread is concerned.

[mijopaalmc]

The ratio of heads to tails is predictable because we know the coin is fair and has two sides.

A fair coin flipped 100 time will yield a sequence with 51 heads and 49 tails with a probability of .0780. The sequence with 50 heads and 50 tails is only more likely to occur, with a probability .0796.

The result of the next toss is not predictable because the coin is fair.

The next toss of a weighted coin is also unpredictable. It is only after a sequence of tosses that you will be able to guess at whether the coin is weighted, given that a weighted coin has a different probability distribution than a fair coin.

[Third Eye Open]

A fair coin flipped 100 time will yield a sequence with 51 heads and 49 tails with a probability of .0780. The sequence with 50 heads and 50 tails is only more likely to occur, with a probability .0796.



The next toss of a weighted coin is also unpredictable. It is only after a sequence of tosses that you will be able to guess at whether the coin is weighted, given that a weighted coin has a different probability distribution than a fair coin.

You seem to be confusing 'predictable' with 'know absolutely'.

A weighted coin is more likely to land on one side than the other. We can predict that it will land on that side more often.

A 'fair' coin has no influences acting on either side, so it is unpredictable which one it will land on.

I suppose you would call the weather unpredictable and random as well.

[GreyICE]

Look up expected value (http://en.wikipedia.org/wiki/Expected_value). Again, the mean of the sampling distribution of means converges on the the expected value of the stochastic process with an ever decreasing variance.


The better word is "anthropomorphism (http://www.google.com/url?q=http://www.answers.com/anthropomorphism&r=67&sa=X&oi=dict&ct=d&usg=AFQjCNG-LNiQ3FhrpkmpMmgICfGwmZLEgA)".
You're trimming my posts to make your strawman. Can you guess why you're on ignore now?

[Paulhoff]

So, articulett, is a sequence of fair coin tosses not random because you are more likely to get 50% heads and 50% tails instead 100% heads or 100% tails?
No, the chance of getting a head or tail is 50%, the pattern of heads and tails will be random.

Paul

:) :) :)

[mijopaalmc]

You're trimming my posts to make your strawman.

I did not trim your post; I opened to the response window before you posted your edit.

Can you guess why you're on ignore now?

Because it is easier to make a grand rhetorical gesture of putting me on ignore and announce it to the whole thread that it is the admit that I could have possibly not seen the addition you made to your post.

By the way, the addition that you made to your post does not actually change my response to you except that I would have said that you also need to review the central limit theorem (http://en.wikipedia.org/wiki/Central_limit_theorem), which does inded demonstrate that, as the number of trials increase, the expected value becomes the most likely value. In fact, for an infinite number of trial, the expected value become the only certain value.

[articulett]

Awesome explanation Tranewreck! But it will fall on deaf ears.

Mijo and RP have this bizarre need to have this nutty, non-explanatory, misleading notion be true through semantics: Evolution is a theory of random chance and life got here randomly. See, all the creationists mischarachterize evolution this way, because it keeps people from actually understanding natural selection. Natural selection is the de-randomizer... it is the thing that gives the appearance of design and renders god superfluous. When you see someone calling evolution "random", you know the person doesn't really understand natural selection though they think they do.

I don't trust people who are more interested in what adjective to stick on evolution than how evolution actually works. In my experience they are either dumb or dishonest creationists. They can only read papers or your argument or the words of others to the extent that they can use it to prove that evolution is "stochastic" (which to them means it's "random")-- (however vague and retarded that might sound.)

They use pedantic words and language they don't understand in a horrible mishmash and then claim that your failure to agree is because you aren't as educated as them, or scientifically rigorous, or "wise" (yes RP and Mijo think they are wise--*snort*), or because you don't understand probabilities (yes, they have imagined-expertise here as well.). Meanwhile, perfectly intelligent people come in and explain things beautifully as you have done thinking that these guys actually care about understanding evolution and why we use terminology a lot more precisely than they are.

But they do not care nor can they learn about that. The just "need" evolution to be "random". If you tell them "random" is misleading--they will assert straw man views of your position and then proceed to fight that straw man. They will tell you to prove that predictable things can't be random or biased things can't be random! And feel as smug a fundy who thinks they've made a point by your being unable to disprove their god.

And they will go on as long as they have an audience. Mijo does this at multiple skeptic sites... some are even fooled for a while. You cannot make a dent with him. RP seems his equal to me. I think they are winning the "semantic game" in their head when they get the last word. But, in truth, I suspect most people just learn to ignore them like they do DOC and the other woo and/or engage them for entertainment purposes.

If nothing else, your response is very well written and can serve to inform others. It cannot make a dent with RP or Mijo because they imagine themselves experts here to teach us all how evolution is really "random".

Remember, per their definition,-- every THING is random. Elections, Chemisty, smoke alarms, whether seat belts save lives, you, life, cooking, etc. Their definitions are literally that meaningless, vague, and useless. Per their definitions, evolution IS random-- but so is everything... just like everything is a noun. That's why the intelligent folks are far more precise even as the dolts imagine themselves being "academically rigorous" and using the "true and proper definitions" and speaking on some more "technically right" definition. They aren't. They are the usual self-important garbled buffoons you will see on every skeptic's forum. Engage them for entertainment purposes only. Keep your expectations very low. Don't let their bluster peeve you, and when you are tired of trying to talk to them, you can put them on ignore and, instead, talk about them.

The smart people know who the smart people are. Sometimes it's best to let birds of a feather flock together, so to speak.

[mijopaalmc]

For the last time, articulett, neither recursive prophet or I are creationists. It is not a mischaracterization of evolution to call it a stochastic process or even random, if you properly define what you mean by "random".

[Walter Wayne]

Random = not predictable
Evolution = Predictable
Evolution != Random
Evolution is predictable?

Tell me, when our numbers had dwindled to less than ten thousand less than one hundred years ago, would you have predicted us being as widespread as we are, or perhaps disappearing? Heck ,how long ago would an observer think a social fire using animal would arise and manage to live in as many diverse climates as we have? How obvious was it.

Walt

[sol invictus]

For the last time, articulett, neither recursive prophet or I are creationists. It is not a mischaracterization of evolution to call it a stochastic process or even random, if you properly define what you mean by "random".

So mijo - are smoke detectors random?

[mijopaalmc]

So mijo - are smoke detectors random?

I've answered your question many time; you jst didn't like the answer I gave you.

You can accurately describe the as orderly and predictable, neither of which preclude them from random.

[Walter Wayne]

Just because stochastic processes have long-term asymptotic behavior does not mean that they are not random.
Yes it does. They are not random.
Actually provided that the asymptote approached is not predictable, then yes it the process could be random. You could run the process once and see it approach X, and then run an identical process and see it approach Y.

Now something random and memoryless won't be asymptotic, but one that does have memory might be. An example of memory in a stochastic system is heredity.

Walt

[Walter Wayne]

You seem to be confusing 'predictable' with 'know absolutely'.

A weighted coin is more likely to land on one side than the other. We can predict that it will land on that side more often.

A 'fair' coin has no influences acting on either side, so it is unpredictable which one it will land on.

I suppose you would call the weather unpredictable and random as well.
If it is simply enough to predict that a coin will land on one side more than other to say it is predictable, why isn't predicting that both events will be roughly equal in number not enough to call it predictable?

Walt

[articulett]

I don't care if Mijo is a creationist. I just don't trust people who argue like them. Clueless people who are obsessed with defining evolution as random as opposed to discussing what evolution is and how it leads to the appearance of design may as well be creationists to me. They are as incoherent, dishonest, blustery, and as tangential as them.

Despite Mijo's repeated assertions to the contrary, there is no respected biologist running around claiming "evolution IS random"-- though Mijo will twist and redefine their words so that he can pretend they are. Creationists do this as well. I can't imagine why anyone actually interested in evolution or who actually understood evolution would do this-- and in my experience their expertise on the subject seems to be entirely in their minds.

I didn't say RP is a creationist... that's their straw men-- I'm saying I can't tell the difference between their blather and creationist blather. All the honest and intelligent people I know go out of their way to clarify the difference. They use words to convey information-- not to obfuscate understanding and pretend that someone intelligent thinks it's meaningful to describe evolution so that it's indistinguishable from the creationist straw man.

Pretty much every respected biologists goes out of their way to explain how evolution is not random--not the creationist straw man. But mijo is obsessed with proving that the straw man definition is secretly "accurate" or "true" or "valid". I let the reader be the judge as to why he is more obsessed with an adjective than conveying understanding about evolution. I leave the reader to assess why his curiosity in science and evolution is entirely focused on this semantic silliness. He doesn't want to know anything new-- he thinks he knows everything there is to know on the subject-- despite being a dolt who could not understand the discontinuity in the fossil record 2 years ago-- even after multiple very detailed explanations!

[sol invictus]

I've answered your question many time; you jst didn't like the answer I gave you.

You can accurately describe the as orderly and predictable, neither of which preclude them from random.

Yes or no?

[sol invictus]

You realize this simple question totally demolishes any lame vestige of a point you might have had, right?

You have three choices:

a) you agree with the statement "smoke detectors are random" (totally absurd), or

b) you agree with the statement "smoke detectors aren't random" (then neither is evolution), or

c) you argue that neither statement suffices because smoke detectors have both random and non-random elements, in which case it is crushingly obvious to everyone that the same applies to the vastly more varied and complicated phenomenon of evolution.

Pick your poison, mijo.

[six7s]

That you continue to ignore the many, many simple, cogent, concise and coherent replies and, instead, persist in stretching the truth to the point of distortion suggests that you are wasting time looking to the virtual world for help with a real world problemI don't ignore the posts to which you refer; I consider them irrelevant, because the belie a deep ignorance of evolutionary biology since the early 20th century.So... you deny ignoring the replies. Fairy nuff

I note that you deny neither that you persist in stretching the truth to the point of distortion nor that you are wasting time looking to the virtual world for help with a real world problem

There might be hope for you yet :)

[articulett]

You realize this simple question totally demolishes any lame vestige of a point you might have had, right?

You have three choices:

a) you agree with the statement "smoke detectors are random" (totally absurd), or

b) you agree with the statement "smoke detectors aren't random" (then neither is evolution), or

c) you argue that neither statement suffices because smoke detectors have both random and non-random elements, in which case it is crushingly obvious to everyone that the same applies to the vastly more varied and complicated phenomenon of evolution.

Pick your poison, mijo.



The same goes for elections.

And gravity.

And whether seat-belts save lives.

To mijo poker is as random as roulette. He's made the term worthless for conveying information. He wants to pretend that everyone who doesn't pick option "a" is choosing "b". He denies option "c" so he can fight the straw man and imagine himself victorious.

[mijopaalmc]

You realize this simple question totally demolishes any lame vestige of a point you might have had, right?

You have three choices:

a) you agree with the statement "smoke detectors are random" (totally absurd), or

b) you agree with the statement "smoke detectors aren't random" (then neither is evolution), or

c) you argue that neither statement suffices because smoke detectors have both random and non-random elements, in which case it is crushingly obvious to everyone that the same applies to the vastly more varied and complicated phenomenon of evolution.

Pick your poison, mijo.
You want a simple answer in which I would have to admit that I was wrong in some way. You are not going to get that, because I am not the one who disagrees with the past half century of evolutionary biology.

[articulett]

Bwahahahha-- mijo considers himself deeply knowledgeable about evolutionary biology!

This, from a dude who probably still can't explain the supposed "discontinuity" in the fossil record.

Hysterical.

Though I suppose Behe imagines himself similarly knowledgeable. The dumb ones are just too damn dumb to realize they are the dumb ones--and too busy telling themselves that their communication problems have to do with other peoples' lack of education or expertise!

What a treasure trove of irony! (I especially like it when they pull out the thesaurus and think they are using words properly and cleverly not realizing how buffoonish they are coming across.)

[sol invictus]

You want a simple answer in which I would have to admit that I was wrong in some way.

That's OK, mijo - the fact that you won't answer is conclusive.

I'll just keep asking that question occasionally in these farces of threads you inhabit, just to remind everyone that you have no credibility or integrity.

[articulett]

Evolution is predictable?

Tell me, when our numbers had dwindled to less than ten thousand less than one hundred years ago, would you have predicted us being as widespread as we are, or perhaps disappearing? Heck ,how long ago would an observer think a social fire using animal would arise and manage to live in as many diverse climates as we have? How obvious was it.

Walt

Irrelevant.

Evolution is predictable in that we can determine that the transition from water to land happened about 375 million years ago... and then go look in the strata for that period of time and discover what that transitional species looked like! Tiktaalkik.

We can predict that everything alive today has a common ancestor and rest assured that we'll be filling in the pieces with each new discovery and each new genome mapping. We can predict that every species will evolve or perish.

And they always have.

Anyone can play semantic games to imagine that it's informative to call evolution random, but not a single respected biologist would say life evolved as a result of random chance. Natural selection is not random chance unless you've defined "random chance" in such a way that everything is random chance.

Biologists know that when you have someone referring to evolution as "random chance", you either have a creationist, someone taught evolution by a creationist, or someone who thinks they understand evolution but is lacking in an understanding of natural selection--possibly purposefully so.

[six7s]

Evolution is predictable in that we can determine that the transition from water to land happened about 375 million years ago... and then go look in the strata for that period of time and discover what that transitional species looked like! Tiktaalkik.

We can predict that everything alive today has a common ancestor and rest assured that we'll be filling in the pieces with each new discovery and each new genome mapping. We can predict that every species will evolve or perish.

And they always have.

Anyone can play semantic games to imagine that it's informative to call evolution random, but not a single respected biologist would say life evolved as a result of random chance. Natural selection is not random chance unless you've defined "random chance" in such a way that everything is random chance.

Biologists know that when you have someone referring to evolution as "random chance", you either have a creationist, someone taught evolution by a creationist, or someone who thinks they understand evolution but is lacking in an understanding of natural selection--possibly purposefully so.If only you had taken the effort to explain this so concisely and clearly months and months ago, we could have been spared mijo-only-knows how many train wreck threads


Oh wait... you did...

So why is he still littering teh interwebs with his inanities? :confused:

[recursive prophet]

Train Wreck: I just put in quite an effort-I’m old-to post your last reply on RD so that susu and hopefully Marios can reply to your points directly. I do hope my nick appeared over susu’s comments just due to the way the quote feature works here, and you and others realize I didn’t write them; that was pure susu. In my wildest dreams I was never half that smart. He studies fossils in Bonn, and has an amazingly broad spectrum of knowledge. See for yourself.

You would not believe how appreciated your thoughtful reply was, and it was posted here ( http://richarddawkins.net/forum/viewtopic.php?p=1618366#p1618366) on the RD thread. You can read the replies there. While I can guess at his response to some of your points, I look forward to his reaction to your questions about the weather. Could we ever know which butterfly would flap it’s wings at the right moment to trigger the chain reaction? Nothing short of omnipotence could accomplish that, could it?

The first 2 quotes are from me, regarding R-Zeta. And reread my previous post; I didn’t in any way imply Zeta would prove primes exist; just that we couldn’t dismiss efforts to unlock their mysteries lacking the ‘assumed’ Zeta proof. They are up to 500 billion non-trivial zeroes last I checked-there’s an online project under a Brit immigrant-Sebastian something-and elsewhere to extend out Riemann as far as possible, and it could tell us loads about the primes. But a proof would be ….ineffable. Zagier:."prime numbers grow like weeds among the natural numbers, seeming to obey no other law than that of chance,..and just the opposite: there are laws governing their behavior, and they obey these laws with almost military precision.”

Thanks again for taking the time to read susu’s comments TW. From what I can ascertain you may be one of very few to do so, and your ‘due diligence’ was inspiring. Thanks-rp.

Mijo, only you (and, perhaps, the patron saint of bollocks) know why, where, when and how often you have started threads on this topic

That you continue to ignore the many, many simple, cogent, concise and coherent replies and, instead, persist in stretching the truth to the point of distortion suggests that you are wasting time looking to the virtual world for help with a real world problem
Sis7s: I sure haven’t seen many such replies. Links please? What I mostly see is comments like yours, appealing to both the crowd and authority while attacking the man instead of his expressed viewpoint. I find the level of hostility directed toward mijo personally on this forum both interesting and ominous. On RD there are a few who do likewise, but his most persistent attacker there is Articulett, whose persistence leaves some to believe she’s in love with him.

I have stated many times previously I’m a complete novice in evolutionary science, and my agenda is simply to learn. I have read many pages of replies by mijo/mjpam, and while I’ve seen him be abrasive, there is NO question he is far less so than those opposing his views. Its all out there people; you can click his profile and read all his comments, as you can do with any user. As with the ones on this thread, he often just ignores insults and argues his point.

Now, were mijo the lone voice making the kind of distinctions he has regarding the process of evolution, as many of his anti-disciples would have you believe, one such as I would likely assume there was some flaw in his reasoning that in my ignorance I simply couldn’t fathom. This is far from the reality, from the evidence I have seen. He is generally respected-certainly by me-but more significantly by the 3 posters on the ENS forum that in my estimation and many others are among the most knowledgeable; Marios, susu, and Dlx2. My level of cognizance wouldn’t even make the cut for the forth tier among the many grad students there that maintain a consistent cohesion within the group in advancing the understanding of their viewpoint. They are very determined. :D In contrast, the few who embrace the majority view I’ve seen endorsed here, seem unable to avoid contradicting one another. They could use help.

When I first started reading the random thread on RD, it was my understanding that while evolution as a whole was stochastic, natural selection was not. For quite a while my mind was like an old couch as the debate drove on-it bore the impression of whoever sat on it last. Now I lean heavily toward a profoundly stochastic process, having been convinced-as were others-by the extensive analysis. But I have read some compelling arguments to the contrary, and that’s why I came here looking for other perspectives.

I would welcome any serious commentary, and will pass it on. I shall not attempt to address even the valid various replies right now, as nearly all have already been covered by those for more incisive than I. They are appreciated, and will be responded to in time, generally in broad strokes as the volume necessitates.

IMPORTANT NOTICE: ALL the replies Trane Wreck was responding to except the first 2, were written by susu.exp on RD, not myself. I be but an aged, humble messenger, seeking nothing more than greater understanding. If you can contribute to that, I’d love to read your deconstructions of what I’ll be posting here, and invite you to also join in at RD.

And do let us remember, we are all one eyed apes who have fallen out of trees; navigating our way through the land of the blind. We are not kings, but outcasts. I sense the energies of some here would be better employed for the cause fighting on the front lines with the creationists, which is where many largely post anyway. This forum should be left to those more comfortable with the very abstract, and who have the time to read complex explanations and respond in kind. Emotions should be conserved for the bedroom. There are very few simplistic answers to anything, but plenty of like minds seeking them.

[my_wan]

Just because stochastic processes have long-term asymptotic behavior does not mean that they are not random.

I'm going to repeat GreyICE's objection. Temperature is an asymptotic behavior resulting from random variables. Temperature is not random. Just because some uncertainty remains does not make it random. Uncertainty and randomness are not the same thing. The Uncertainty Principle in Quantum Mechanics has nothing to do with the randomness either. In fact randomness can be tested for using the Poisson distribution. When a graph deviates from a Poisson distribution we know it contains non-random elements even if we don't know how or why, much less able to predict it.

[articulett]

If only you had taken the effort to explain this so concisely and clearly months and months ago, we could have been spared mijo-only-knows how many train wreck threads


Oh wait... you did...

So why is he still littering teh interwebs with his inanities? :confused:

The woo win points in their head every time they get the last word-- they think that their argument was so good, that no one can address it.

Every time Mijo argues his semantic silliness and gets others to join in-- evolution becomes more "random" in his mind.

All the woo do that. They have their little thing they "need" to be true, and they prop it up by spinning the delusion under the pretext of conversation or debate. People who care what evolution is or care to explain it to others, aren't hung up on proving that it's "random" or "stochastic". I think it's fun to figure out the "belief" they need to prove to themselves.. and then screw around with it. It always boils down to a single claim. Mijo needs the creationist straw man about evolution being random to be "true". Apparently their world rests on "proving" some inane thing to themselves. The more you read them the better you get at discerning the little belief they are trying to protect, though they lie and deny it and never present evidence for it. They imagine they will be taken more seriously that way I guess.

They must not realize how much they sound like each other.... how after a while all the ways people spin to confirm their biases boils down to the same semantic games. Whether the belief is about their own expertise or their astral travel experience or what "atheists think"-- it's all the same "blather". People tend to fool themselves in very predictable (nonrandom :) ) ways! Of course, I'm sure Mijo could tell us how "random" it really is. What isn't, to Mijo?

[Walter Wayne]

Lost my first reply so this will be short ...
Evolution is predictable in that we can determine that the transition from water to land happened about 375 million years ago... and then go look in the strata for that period of time and discover what that transitional species looked like!I said predict. You haven't predicted what evolution will do, but have discovered what it has done. You've "predicted" the past; congratulations.
We can predict that everything alive today has a common ancestor and rest assured that we'll be filling in the pieces with each new discovery and each new genome mapping.
Predicting the past again.
We can predict that every species will evolve or perish.
Finally a prediction of the future. And you immediately predict two possibilities. Multiple future possibilities is pretty much the hallmark of unpredictability.
And they always have.
Predicting the past again I see.

If you wish to accuse me of semantic games, don't play silly games where you make such sweeping multiple choice predictions. It's as impressive as predicting a die roll will come up as an integer between 1 and 6, but not 7. Obviously we shouldn't refer to die rolls as random.

Walt

[articulett]

What's your point, Walter?,

Why don't you answer Sol's questions above-- and if you don't have an answer or you think smoke alarms are random-- then you need to converse on your level. I suggest Mijo or Recursive Prophet.

I think there are still some folks playing the semantic game over on the Dawkins thread, but I consider it beneath me. I think Dawkins, Ken Miller, and just about everybody else is more than clear. I cannot tell the ramblings of the leftovers from the ramblings of Behe. I leave them to engage each other with their deep questions about which adjective to stick on evolution. I don't care what words you use. I care that people understand the process and why Dawkins, et. al. do not consider evolution a theory of random chance.

I like you okay, Walt, but I have to put you on ignore so that I don't get distracted by your many questions that you do not really want answered. Why don't you and Mijo and RP answer Sol's multiple choice question first?

ETA, and we PREDICTED where we could find a transitional species. We can predict what the sex of a baby is via amniocentesis. It's the fact that science is deterministic on that allows us to predict weather and do just about everything we've learned to do. Play your apologetics with the stupid folks. Asking the impossible of evolution, doesn't make it random, walt--except for those who desperately need to believe that it's useful to describe it as such.

[Earthborn]

An example of something that is random and predictable would be nice.The lottery. It is as random as randomness can be, but it is very safe to predict that you won't win it.
Many things called "random" behave -- in the long run, or on a large scale -- very predictably. Much more predictably than evolution in fact.

If something is predictable, how are we predicting it without knowing the causes behind it?By extrapolating past outcomes.

If we know the causes behind something, how can you still call it random?There are a few possibilities:
Some of the causes may be random
Knowing the causes may not help predicting very specific outcomes (such as knowing Newtonian physics doesn't help us predicting which lottery ball comes out of the machine next) and a probabilistic approach is much more useful.

[articulett]

So how do you answer Sol's 3 question query?

[Earthborn]

ETA, and we PREDICTED where we could find a transitional species.I thought the discussion was about how predictable the process of evolution is, not what predictions the theory of evolution makes.

[articulett]

You realize this simple question totally demolishes any lame vestige of a point you might have had, right?

You have three choices:

a) you agree with the statement "smoke detectors are random" (totally absurd), or

b) you agree with the statement "smoke detectors aren't random" (then neither is evolution), or

c) you argue that neither statement suffices because smoke detectors have both random and non-random elements, in which case it is crushingly obvious to everyone that the same applies to the vastly more varied and complicated phenomenon of evolution.

Pick your poison, mijo.

I like this test. Everyone who refuses to answer or who chooses "a" can be funneled through the ignore feature into dealing with each other. And you get a quick sense of peoples "hidden needs" so-to-speak. It also gives you a quick assessment on the general intelligence level of the person you are trying to converse with.

I just want to congratulate you on this tool you've developed-- (evolved) (nonrandomly) :fg: I shall use it to select whom to read and whom to file under "mijoesque".

win-win for everyone.

[articulett]

I thought the discussion was about how predictable the process of evolution is, not what predictions the theory of evolution makes.

It's not ... that would be another semantics game. You are free to play with those interested in semantic games. I suggest mijo, walter wayne, and recursive prophet.

Oh, btw, how do you answer Sol's question?

[Earthborn]

The process by which smoke detectors operate is random. The result is that they work very reliably and predictably as is so often the case with random processes: in the long run (which in the case of smoke detectors doesn't take much time) they tend to become very predictable.

[six7s]

Sis7s: I sure haven’t seen many such replies. Links please? You know how to "click his profile and read all his comments" on JREF and RD - here you can see the threads he started too (http://forums.randi.org/search.php?searchid=1790693)... check 'em out... there is a pattern and it ain't pretty

I find the level of hostility directed toward mijo personally on this forum both interesting and ominous. Curious

Ambiguous as all hell... but curious

On RD there are a few who do likewise, but his most persistent attacker there is Articulett, whose persistence leaves some to believe she’s in love with him. Some? Who? The deluded? Well, really, are you surprised that IDiots would spout such crap?

I have stated many times previously I’m a complete novice in evolutionary science, and my agenda is simply to learn.Then quit trying to teach on the subject

I have read many pages of replies by mijo/mjpam, and while I’ve seen him be abrasive, there is NO question he is far less so than those opposing his views. It really doesn't matter if he comes across like a pit-bull terrier on meth or a king charles spaniel on valium ... it's not how he talks but what he says that counts

Zero ain't much of a count

[articulett]

Thanks six7s...

RP is a bit delusional from my perspective. He may be the only one who sees things as he does. But it's common for that type to speak as if the rest of the world views things through their biased lens. You need only read his first few posts to get a grasp on his actual motives. However, I am hardly interested in what the mijo-esque think of me; I suspect it pales in comparison to my opinions of them.

And for the record, I tend only to fall in love with the witty, brilliant, and heterosexual. I believe that leaves Mijo out on all 3 counts. I think RP may be the only one suffering from that particular delusion which I suppose is no more bizarre than his many others.

[articulett]

The process by which smoke detectors operate is random. The result is that they work very reliably and predictably as is so often the case with random processes: in the long run (which in the case of smoke detectors doesn't take much time) they tend to become very predictable.

So no real answer, eh?

Umm, did you know that smoke detectors don't replicate and pass on the best mutations and recombinations exponentially? By golly, what would you call a process that did such a thing resulting in a honing and pruning of the information over time? Smart people don't call it "random". All the smart people I know describe it more or less like Dawkins did in the clip at the beginning. Or like answer "c". None have the obsessive need to call it random.

I guess that puts you in the conversation pit with others who are similarly vague.


Breach of Membership Agreement removed.

[The Atheist]

Note to recursive:

Maybe now you can see why I wanted this ridiculous thread split off the evolution thread.

Now would be a really good time to let this thread die a natural death so we can bury it.








About 500 feet deep.

[Earthborn]

So no real answer, eh?Not one I expected you to like, no. I agree with Sol Invictus that "smoke detectors are random" is an absurd statement, but for a different reason. Only processes can be fairly called "random" and a smoke detector is not a process. It is just a thing. A thing that has been "intelligently designed" and deliberately constructed, and therefore not random at all.

Umm, did you know that smoke detectors don't replicate and pass on the best mutations and recombinations exponentially?Yes, I know that. I fail to see the relevance of pointing this out.

By golly, what would you call a process that did such a thing resulting in a honing and pruning of the information over time? Smart people don't call it "random".I think smart people understand that some random processes can do such things, for example genetic algorithms or learning algorithms.

None have the obsessive need to call it random.I don't see anyone having an obsessive need to call evolution random, but I do see some obsessively trying to deny that evolution can be called random for some definitions of "random".

I guess that puts you in the conversation pit with others who are similarly vague.Just because you make no attempt at trying to understand what I write does not make me vague.

[GreyICE]

I don't see anyone having an obsessive need to call evolution random, but I do see some obsessively trying to deny that evolution can be called random for some definitions of "random".

Not, however, one found in a dictionary. "Today, random means blue!"

The word you are actually looking for is stochastic, which is why this goddamn thread should have been over about 6 pages ago. I am shocked and appalled CFLarson hasn't contributed, it would make this thread complete.

[Dr Adequate]

Lost my first reply so this will be short ...
I said predict. You haven't predicted what evolution will do, but have discovered what it has done. You've "predicted" the past; congratulations. No, the present. Like every scientific theory.

Predicting the past again. No, the present. Like every scientific theory.

[Dr Adequate]

I think smart people understand that some random processes can do such things, for example genetic algorithms or learning algorithms. But those aren't random processes any more than the behavior of a smoke detector is a random process.

I believe I am a "smart person", but if I am not, I have at least been employed as a computer scientist, and if I went around pretending that a genetic algorithm was a "random process" I would get laughed at so much by all the other computer scientists.

[recursive prophet]

I'm going to repeat GreyICE's objection. Temperature is an asymptotic behavior resulting from random variables. Temperature is not random. Just because some uncertainty remains does not make it random. Uncertainty and randomness are not the same thing. The Uncertainty Principle in Quantum Mechanics has nothing to do with the randomness either. In fact randomness can be tested for using the Poisson distribution. When a graph deviates from a Poisson distribution we know it contains non-random elements even if we don't know how or why, much less able to predict it.
And yet this proves what? I shall post this on RD to see if someone there understands the point here, but I don’t. The last sentence above is particularly confusing, but I’ll leave open the possibility it’s my lack of comprehension. Too bad a few others here don’t have a little more ‘uncertainty’ about there conclusions.
The process by which smoke detectors operate is random. The result is that they work very reliably and predictably as is so often the case with random processes: in the long run (which in the case of smoke detectors doesn't take much time) they tend to become very predictable.
Welcome aboard, Earthborn. Don't pay any mind to the trolls here; it's there nesting season I guess. Taking the long view, as it seems you're doing with the smoke detectors, is not something the myopic are likely to embrace though, so best bet is to just ignore them. If you read this thread you'll have to admit they're often good for a laugh, talking about ignoring other users like mijo then persistently responding to his every post.

My method is to only deal with those who present some actual content other than the personal attacks. There are a few here I will soon reply to such as GreyICE. But for the many who thrive on insults and attempting to deny others the right to express themselves, I don’t rave on about ignoring them; I just do. It’s easy. But you will get the jabs when they see you’re not in their little clique. Its both funny and sad. Hehe. Sigh.

Try reading the string of replies by susu in my post on the previous page. Much is covered there, but only Trane Wreck, as far as I can ascertain, has taken the time to do so.

[my_wan]

Lost my first reply so this will be short ...
I said predict. You haven't predicted what evolution will do, but have discovered what it has done. You've "predicted" the past; congratulations.

Predicting the past again.

Finally a prediction of the future. And you immediately predict two possibilities. Multiple future possibilities is pretty much the hallmark of unpredictability.

Predicting the past again I see.

Is this not a prediction?

BXdQRvSdLAs
http://www.youtube.com/watch?v=BXdQRvSdLAs

If you wish to accuse me of semantic games, don't play silly games where you make such sweeping multiple choice predictions. It's as impressive as predicting a die roll will come up as an integer between 1 and 6, but not 7. Obviously we shouldn't refer to die rolls as random.

Walt

I didn't see your original post but you are playing semantics games here. A prediction doesn't have to be about the future if the past being predicted is not known at the time of the prediction, or the theory unambiguously requires those observed facts. In other words it remains predictive so long as the theory is requires those predictions regardless of whether it has been observed or not. God did it doesn't pass this test because no matter what is observed "God did it" still fits.

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

Finally the OP is about "stochastic", just because something is unpredictable does not make it random or stochastic. Need a good example? Try a pendulum attached to a second pendulum.

[articulett]

Hurray Ken Miller!

Yes, that's Ken Miller who says the the following:

"I have no idea why someone would take a term like natural selection and say it is random”, said Miller when reached for an interview.

"Natural selection is a distinctly non-random process that acts as a sieve through which genetic changes are filtered.” Just as a sieve filled with various rocks will not end up filtering out its contents randomly, natural selection does not filter organisms randomly."

http://forthesakeofscience.wordpress.com/2008/10/25/why-natural-selection-is-not-random/

If you cannot follow the Ken Miller Clip or the Dawkins clip (at the beginning of this thread) or you don't agree or imagine you are being more accurate despite all these people telling you you are not, you are playing a semantic game...

Even if you really think you are saying something deep and important and valid. If evolution is "random", then everything is random. The term becomes meaningless.

All of science works because, at least on a macro level, everything is, in fact, deterministic. When things appear random, it's because we don't have all the inputs. We really mean unpredictable. We can't predict the airplanes of the future, but we can predict that they will be better than the ones in the past. We can't predict which planes will crash, but we know some will, and that we will improve our designs based on both our failures and our successes. Evolution is the same. To call this "random" makes you sound like a bonehead to me. Or a creationist. Not that I can tell the difference.

Oddly, you can gauge a person by the interest they show in such clips and links. Those pretending to be interested in evolution, but who are really interested in "semantics", fail to show the curiosity or get the thrill from those clips that people who are actually interested in evolution do. I guess they imagine themselves smarter than the experts. Intelligent discourse is wasted on them. They think they can "win" a semantics game. Tsk.

[my_wan]

And yet this proves what? I shall post this on RD to see if someone there understands the point here, but I don’t. The last sentence above is particularly confusing, but I’ll leave open the possibility it’s my lack of comprehension. Too bad a few others here don’t have a little more ‘uncertainty’ about there conclusions.

It would help if I knew more about what you do understand, even if it only loosely relates. However I will try articulate and you can let me know how I've done.

The last sentence you said the most confusing, I'll start there. The last sentence was: "In fact randomness can be tested for using the Poisson distribution". Any purely discrete random variable when plotted on a graph will produce a Poisson distribution.
http://en.wikipedia.org/wiki/Poisson_distribution
Looks somewhat like an upside down bell. The the graph itself representing the cumulative values of many identical such random variable is also called a bell curve, normal distribution or Gaussian distribution.
http://en.wikipedia.org/wiki/Normal_distribution

For instance if we measure the mass and speed of air molecules in an isolated box and graph enough of them we'll find our graph is a normal distribution. What happens though if we do this and find it doesn't graph that way? Then we know the box is not isolated and it is exchanging energy with the outside. In other words something non-random is happening, like heat leaking in a particular direction. In many circumstances we may not know why, but we can know when non-random variables are buggering the data this way. The theory of evolution is in a sense the theory of what is buggering the data in random mutations. Unless animals just "accidentally" mate evolution is not random but is stochastic, due to the random variables that the non-accidents are buggering.

You may post this over on RD also.

[articulett]

To RP and Mijo, random=stochastic. They are synonyms to them.

A chaotic system is not random to you, but it is to them. It's ALL random. Allele frequencies are more or less stochastic, but phenotype is strongly deterministic--http://www.bio-medicine.org/biology-news-1/Evolution-is-deterministic--not-random--biologists-conclude-from-multi-species-study-1268-1/

Of course, peer reviewed scientists are not defining words as vaguely as RP or Mijo. They are pointing out that each step of the mutation is selected for in a phenotypic change-- there's direction. The question is what percentage of any genome is there because it was selected for or because it just happened to ride along in a genome that didn't cause it to be selected against. Unless you are specific with what exactly you are referring to, it really isn't any more useful to call evolution "stochastic" than it is to call elections "stochastic". I mean you can, but... why? What information is being conveyed? Is it more informative than saying "evolution is a noun"?

I know you understand this. But they can only understand enough so that they can prove that evolution is "stochastic"--which to them is the same as "evolution is random".

They will use your words as "proof" that evolution is random on a thread entitled "how evolution is not pure random chance". To them, anything that is stochastic is random-- otherwise it must be deterministic. Do you understand the semantic game they are playing here?

I suggest you find out his answers to Sol's 3 choices to assess what he does and doesn't understand and how exactly he's defining words. He is very abusive to those whose words he cannot construe into the idea that "evolution is random" despite his claim that he "has no dog in the fight".

[articulett]

Oh, and RP, it's "their" conclusions. And you might try adhering to your own advice.

I don't think all the education in the world is going to let you suddenly see the light and declare, "Oh, I see why Dawkins et. al. are so adamant in saying that evolution is NOT random chance! I finally understand natural selection. Thanks for bearing with me!" Your interest in not understanding others seems to be on par with Mijo's. And in two years, he's as clueless as he was when he came here. He's earned his reputation.

[six7s]

Do you understand the semantic game they are playing here?I don't

Although it seems like a semantic game, it doesn't seem to have any codified rules, nor does it seem to be much fun...

Why, I wonder, do they continue to turn up let alone play when they are getting their arses whipped at every turn, have yet to score even a penalty and are down a gazillion to zip - hardly surprising when they field z-grade players with the mascot - Rear Admiral Random (retired) - as the captain

[articulett]

Like creationists... if some serious scientific types are taking them seriously, they imagine they must be winning points, I guess. They get off on getting the last word.

They are trying to prove that the creationist straw man that says "evolution is a theory of random chance" is true. Every respected biologist knows that is a straw man view of evolution designed to obscure understanding of the fabulous way natural selection brings about the appearance of design.

I think our understanding of evolution is one of the coolest things humans have come to know, and I have nothing but disgust for those who feign interest when their real goal is to obfuscate. I love explaining it to people... but like Dawkins, etc.-- I'm sick of explaining things to people who have a vested interest in not understanding --or people who imagine themselves more expert than the experts.

It does get hard to be nice to people who are clearly ignorant, but imagine themselves experts. I can't tell the stupid from the woo-pid. :)

They waste the time of and are really nasty to those they might learn from while fawning over those whose words they can twist to prop up their delusions of their imagined expertise.

[cyborg]

I think smart people understand that some random processes can do such things, for example genetic algorithms or learning algorithms.

As Dr A tried to allude to algorithms are entirely deterministic.

Whoops. If Genetic Algorithms are in some way a good analog of evolutionary processes then there's some 'splainin' to do on the whole importance of randomness - a quality that is impossible, by definition, to have with a deterministic system.

[sol invictus]

Not one I expected you to like, no. I agree with Sol Invictus that "smoke detectors are random" is an absurd statement, but for a different reason. Only processes can be fairly called "random" and a smoke detector is not a process. It is just a thing. A thing that has been "intelligently designed" and deliberately constructed, and therefore not random at all.

That's a transparent evasion based on a semantic nitpick. But very well, here are your new choices:

a) the detection of smoke by smoke detectors is random

b) the detection of smoke by smoke detectors is non-random

c) the detection of smoke by smoke detectors has both random and non-random elements.

You said you believe something very close to a), which is totally absurd. Would you like to change your answer?

[sol invictus]

S
No, I wasn’t aware of that, but radioactive decay as a process is of course random though the detectors that employ it are not.

So you choose option b), while mijo just runs away. Too bad you guys can't get your story straight even on this totally trivial example.

What kind of pigheaded arrogance does it take to think you can accurately describe something as complex as evolution in one word, when you can't even agree how to apply it to the simplest of everyday objects?

@Walter Wayne - evolution predicts many very specific facts about the future. For example, expose 12 colonies of bacteria to a low level of antibiotic, and all 12 will evolve resistance to it. That's a precise, specific, and extremely useful prediction about the future that has been experimentally verified. There are a very large number of such examples.

[Dr Adequate]

As Dr A tried to allude to algorithms are entirely deterministic. Huh?

Whoops. If Genetic Algorithms are in some way a good analog of evolutionary processes then there's some 'splainin' to do on the whole importance of randomness - a quality that is impossible, by definition, to have with a deterministic system. Still not following you, sorry.

[sol invictus]

I think cyborg's point is that one can run perfectly good simulations of evolution, or make use of genetic algorithms, using deterministic pseudo-random number generators. Which highlights the fact that fundamental randomness (if it even exists) is totally irrelevant to evolution.

What's more pertinent to this "debate" is the question of the unpredictability of some aspect of evolution. In some cases and in some aspects the outcome is highly predictable; in others it's not. Some things are too complex to be described by a one-word sound bite, shocking as that may seem to some here.

[articulett]

In the standard theory of computation, the term algorithm stands for a deterministic algorithm.

In algorithm design, nondeterministic algorithms are often used as specifications, but they are called "nondeterministic" algorithms-- not "random" algorithms. All algorithms are considered deterministic unless otherwise specified.

Having random components, does not a "random process" make. Even if Earthborn wishes to believe it were so.

[Dr Adequate]

I think cyborg's point is that one can run perfectly good simulations of evolution, or make use of genetic algorithms, using deterministic pseudo-random number generators. Ah yes, very true; but it wasn't what I was alluding to, which was what puzzled me.

[six7s]

They are trying to prove that the creationist straw man that says "evolution is a theory of random chance" is true. Every respected biologist knows that is a straw man view of evolution designed to obscure understanding of the fabulous way natural selection brings about the appearance of design.


I guess Ken Miller is respected, yes?

I have been a-Googlin' and found some stuff that I find interesting


Ken Miller's "Random and Undirected" Testimony
www.evolutionnews.org/2006/07/ken_millers_random_and_undirec.html (http://www.evolutionnews.org/2006/07/ken_millers_random_and_undirec.html)
...Miller is a widely promoted theistic evolutionist, and thus served as the plaintiffs leadoff expert witness for biology, evolution, and theistic evolutionism during the Kitzmiller trial.

<snip/>

On the second day of the Kitzmiller trial, Miller was confronted about theologically charged statements about evolution in one of his biology textbooks, which stated that "[e]volution is random and undirected." Miller defended himself by claiming that the theological language about evolution in his textbook was a "mistake," and was added by his co-author, and that the statement "[e]volution is random and undirected" appears only in the 3rd edition of his "elephant textbook," Biology. Miller said, "that statement was not in the first edition the book, it was not in the second edition, it was not in the fourth edition."

<snip/>

Q. Sir, is evolution random and undirected?

A. I don’t think that is an appropriate scientific question. First of all, evolution most definitely is not random. There are elements of evolutionary change that are unpredictable, but the principal force driving evolution, which is natural selection is most definitely a non-random force, and then the second part of your question, undirected, that requires a conclusion about meaning and purpose that I think is beyond the realm of science. So my answer for different reasons to both parts of your question is no. Or excuse me, perhaps more aptly put, science, science cannot answer the second part of the question. I think that’s a more accurate way to put it.

(Transcript of Testimony of Kenneth Miller, Day 2 of Kitzmiller Trial (Sept. 27, 2005), pgs. 4-8.)

<snip/>

Miller was then confronted with the fact that the phrase "Evolution is random and undirected" appears in the 1995 edition of his textbook:

Q. Sir, in your 1995 edition of Biology, I believe it’s the Elephant Book?

A. That’s correct. It’s generally known by that name.

Q. Did it not state in that book, “It is important to keep this concept in mind. Evolution is random and undirected, and the part “evolution is random and undirected” was in bold print?

[…]

A. … So yes, sir, it does say that.

(Transcript of Testimony of Kenneth Miller, Day 2 of Kitzmiller Trial (Sept. 27, 2005), pgs. 4-8.)

What's most interesting is Miller's explanation for the presence of that statement. Here it is:

Q. Now, you testified previously that that’s not a scientific concept, correct?

A. I did indeed, sir.

Q. Why was it in your book?

A. It was in my book because as I’m sure you’ve also looked at, that statement was not in the first edition the book, it was not in the second edition, it was not in the fourth edition, it was not in the fifth edition. It was not ---

Q. My question is why is it in this edition?

A. I’m trying to set the context so I can give a full and complete answer to your question. So the interesting thing is that this is the only edition of any of the books that we have published, and probably eleven different editions, that contains that statement, and the reason for that quite simply is that I work with a co-author whose name is Joseph Levine, and Joe and I work together on many of the chapters in the book, but many of them we write separately and individually, and this was a statement as I was going through Joe’s chapters, and I feel very badly about that. When this was first pointed out to me, the third edition of this book was in print, I immediately went to Joe, I said Joe, I think this is a bad idea, I said I think this is a non-scientific statement, I think it will mislead students. Joe agreed. We immediately took it out of the book, and that’s why I emphasized that it did not appear in subsequent editions. So what you’re looking at, sir, is a mistake.

(Transcript of Testimony of Kenneth Miller, Day 2 of Kitzmiller Trial (Sept. 27, 2005), pgs. 4-8.)



YouTube: How To Shut Up Pesky Creationists (http://www.youtube.com/watch?v=dK3O6KYPmEw&feature=related)
v=dK3O6KYPmEw&feature=related
Easy to understand evidence for a common ape-like human/chimp ancestor. One of thousands of collaborating pieces of evidence, but this one is very conceptual. Whip it out anytime a creationist starts preaching.

Watch the entire lecture "The Collapse of Intelligent Design" given by Ken Miller at www.youtube.com/watch?v=JVRsWAjvQSg (http://www.youtube.com/watch?v=JVRsWAjvQSg)



YouTube: Ken Miller on Intelligent Design (http://www.youtube.com/watch?v=JVRsWAjvQSg)
JVRsWAjvQSg
Ken Miller's talk on Intelligent Design at Case Western University. Ken Miller basically rips Intelligent Design apart in a 2 hour long exposé of the claims of intelligent design and the tactics that creationists employ to get it shoehorned into the American school system.

[articulett]

Thanks six7s. So the questions to me is, why would anyone persist and insist in describing evolution in a way that's indistinguishable from the creationist straw man despite expert opinion like this all over the web.

Yes, this is the best that the creationists have, because it appears to work on a lot of confused types and it's hard to undo the dumb. It's their favorite canard. I'll try to rescue those infested with the meme, but when they are beyond help, I want nothing to do with them. I consider them as much of a lost cause as Behe. I find their vacuity and mendacity shun-worthy.

Ken Miller is an excellent person to have on the side of truth, humor, and intelligence. Plus he doesn't scare away people with that evil "atheist" label. Whatever his god does, it stays in Miller's head. He is the author/editor of some very fine Biology textbooks, and I am repulsed that mijo thinks he knows better or uses terms more accurately than him; mijo still cannot discuss anything intelligent about the supposed "discontinuity in the fossil record" despite tons of data uncovered in the last two years that would be rather thrilling to anyone actually interested in the knowing such a thing as he came here pretending to be interested in knowing.

If you have so many damn questions, why the hell are you never interested in the answers, Mijo? Unless, they are the answers you "want", of course. Why aren't you as admiring of Ken Miller as all those who are really interested in the subject? You get smart people running around thinking you actually want answers to your inane loaded questions but you just want to reject the answers so you can assume the creationist straw man is true. You and your ilk deserve to try to make sense of each other.

[Third Eye Open]

If it is simply enough to predict that a coin will land on one side more than other to say it is predictable, why isn't predicting that both events will be roughly equal in number not enough to call it predictable?

Walt

Sorry, should have added what is in bold

You seem to be confusing 'predictable' with 'know absolutely'.

A weighted coin is more likely to land on one side than the other. We can predict that it will land on that side more often.

A 'fair' coin has no influences acting on either side, so it is unpredictable which one it will land on on the next toss.

I suppose you would call the weather unpredictable and random as well.

[mijopaalmc]

If it is simply enough to predict that a coin will land on one side more than other to say it is predictable, why isn't predicting that both events will be roughly equal in number not enough to call it predictable?

Walt

Sorry, should have added what is in bold

You seem to be confusing 'predictable' with 'know absolutely'.

A weighted coin is more likely to land on one side than the other. We can predict that it will land on that side more often.

A 'fair' coin has no influences acting on either side, so it is unpredictable which one it will land on on the next toss.

I suppose you would call the weather unpredictable and random as well.

And the ting that you are missing is that you cannot predict on which side a weighted coin will land on the next toss. You can only say that it will land on one side more often than the other.

[paximperium]

And the ting that you are missing is that you cannot predict on which side a weighted coin will land on the next toss. You can only say that it will land on one side more often than the other.
So? How does that in any way equate to random?

[mijopaalmc]

So? How does that in any way equate to random?

How does it not?

In both systems, you cannot predict what the next toss will be, but you can say that over the long term you will end up with a specific proportion of heads to tails. What that ratio is different from system to system but the fact that the proportion for any finite number of trials approaches the expected value for the proportion as the number of trial increases does not change; it is happens whether the coin is fair or biased. Thus, the insistence that only fair coins are random is not borne out by anything but a superficial analysis of the systems.

[articulett]

So? How does that in any way equate to random?

You'll be sorry you asked. (Get ready to have the "stupid" heaped upon you.)
(to Mijo, anything related to probabilities in any way is random... really.)

[mijopaalmc]

You'll be sorry you asked. (Get ready to have the "stupid" heaped upon you.)

I'll leave the heaping of "stupid" to you. After all, you are the expert in such matters.

[paximperium]

How does it not?

In both systems, you cannot predict what the next toss will be, but you can say that over the long term you will end up with a specific proportion of heads to tails. What that ratio is different from system to system but the fact that the proportion for any finite number of trials approaches the expected value for the proportion as the number of trial increases does not change; it is happens whether the coin is fair or biased. Thus, the insistence that only fair coins are random is not borne out by anything but a superficial analysis of the systems.
You call that random? What nonsense.

Let's use a dice analogy for mutation alone.

The throw of a die is mutation in evolution.
However, you can never throw a 1 because the organism dies in vitro.
You can never throw a 2 since the organism dies within a day of birth.
You can throw 3,4 and 5.
However, since 6 exist in a very high variable mutation zone, you are twice as likely to throw a 6.

So, in this analogy you can get 3, 4, 5 and 6 and 6.
Do you consider that random?

Now take the fact that 5's are ten times likelier to live and develop offspring compared to 3, 4 or 6.
Do you consider that random?

Do you consider it random to have a higher number of 5's than any other number after 10,000 rolls?

[articulett]

It's all been said before Pax... for over 2 years... on multiple forums. He can't answer sol's 3 option query on post #164 because he needs to believe evolution is random. If you say "nonsense", he'll accuse you of calling evolution deterministic and ask you to prove it, malign you, divert and claims some scientist is saying what he is saying (they aren't), pretend that you have an aversion to the word "random", or some other idiocy of the type you've come to know and love on the JREF or maybe he'll just call you a liar as he's repeatedly done to me.

He claims not to be a creationists (but so does Behe) while using their exact straw man view of evolution and the exact arguments they use to obfuscate understanding of natural selection. In fact, his whole history here has been a series of loaded creationist straw men. That's why most have him on ignore. Engage him for entertainment purposes only. The more you try to give him a clue, the more he'll take it out on you.

I predict Mijo will be able to clearly convey natural selection and stop using the creationist canard that "evolution is random" the day Behe does. His inanity is that breathtakingly entrenched. Adjust your expectations accordingly and warn other smart folks who might think he truly is curious about evolution. I don't know where his imagined expertise comes from, but I'm guessing he pulled directly from his ass like Behe.

I don't think he can actually even follow what you are writing. If you can't prove that something is nonrandom-- and it has anything to do with probabilities whatsoever-- Mijo will insist that it's random. Moreover, he will insist that he's being "technically rigorous" and that scientists all over talk this way and that it's meaningful to describe evolution as a "theory of random chance"--that Dawkins is being pedestrian when he insists that it's not.

No smart people agree of course, though many creationists do. Most feel as the Miller quote above illustrates. But, so long as evolution is a "theory of random chance" that says "life evolved randomly", it also sounds too super duper unlikely-- unless you stick god in. This is wahy obfuscating understanding of natural selection by referring to evolution as a theory of randomness is the creotards favorite "weapon" of ignorance promotion. And, look, some people fall for it. (Hint: If you can't answer sol's question--consider yourself afflicted.) I am most impressed with Sol's question screener. It can tell you peoples intelligent level, biases, and conversation worthiness rather quickly. Refusal to answer or vague answers or answer "a" means they get to play their little semantic games with each other and let the grown ups talk amongst themselves. :)

[recursive prophet]

That's a transparent evasion based on a semantic nitpick. But very well, here are your new choices:

a) the detection of smoke by smoke detectors is random

b) the detection of smoke by smoke detectors is non-random

c) the detection of smoke by smoke detectors has both random and non-random elements.

You said you believe something very close to a), which is totally absurd. Would you like to change your answer?
The OP here posed the question; Is evolution stochastic? If you replaced random above with stochastic, I would pick a. “You act as if stochastic meant "devoid of rules and mechanisms". It doesn´t. It means that for an initial condition there is more than one possible outcome and the probabilities with which these outcomes occur can be described through probability distributions. If you shoot an electron through a double slit at a detecting surface, I can give you the probabilities for hitting any specific spot. And if you do it with a lot of electrons, you´ll see the probability distribution approximated by the frequency distribution.” (susu)

Walter Wayne: Welcome to the thread, and I apologize for the behavior of some who seem to have some weird form of emotional connection the question I posed in another thread on evolution that was moved and begat this thread. Not sure what to make of it, but feel free to join in the similar thread I posted a link to below. I’ve tried to get some of the users there to join in here also, but there are far fewer histrionic posters there. A gaggle of very informed grad students have made the case for a stochastic evolutionary process as the accepted working hypothesis in science today quite convincingly. Still, I struggle against complacency and so am still seeking out those with contrary opinions.

paximperium: I don’t really understand your analogy, but if you go to the link below and work your way back to pt. 1 and 2 of the thread linked there you will find quite a few very complex dice analogies on evolution and how it is stochastic. I diverge a bit from mijo on this in that I’m not comfortable defending the definition he uses in relating random with stochastic, though I can see why he has resorted to that approach here.

I’m working on an index of parts 1,2, and 3 of the thread on RD now, and when I come across one particularly compelling dice metaphors I recall I’ll post it here for you.

I am hoping others might still weigh in on the series of replies I posted by susu here (http://forums.randi.org/showthread.php?postid=4301273#post4301273). Please ignore those who continue to lambaste those asking questions, even to the ridiculous point of speculation about our intentions. You will note these misstatements of our positions and other oft repeated calumnies have been ignored in my posts, and I will continue to do so. Those with an actual question/viewpoint will get a response.

Let us remember, we are all among the one eyed apes who have fallen out of trees, navigating our way through the land of blind myth. Yet we did not become kings, but outcasts. This forum should be for those comfortable with the very abstract, and who have the time to read complex explanations and respond in kind. Emotions should be conserved for the bedroom. There are very few simplistic answers to anything, but plenty of like minds seeking them.

[The Atheist]

Man, I'd love to see you guys at a craps table.

[sol invictus]

The OP here posed the question; Is evolution stochastic? If you replaced random above with stochastic, I would pick a.

As would I, using the definition for "stochastic process" you gave earlier - and I haven't seen a single poster in this thread disagree with that. So if that's all you want to know, why are you still posting here?

On the other hand your little friend seems to think "stochastic" and "random" are the same:

"Random" and "stochastic" share a technical definition and therefor can be used as synonyms for that particular definition.
It is not a mischaracterization of evolution to call it a stochastic process or even random, if you properly define what you mean by "random".


You, on the other hand, have said smoke detectors are not random but are stochastic:

No, I wasn’t aware of that, but radioactive decay as a process is of course random though the detectors that employ it are not.

and you agree that the two terms should not be equally applied:

Stochastic and random are synonyms; look in any dictionary. But as you ostensibly know, that doesn't mean they are equal or should be applied in the same way in describing NS.

So to summarize - and I'm being as fair as I can here - your position regarding smoke detectors (and evolution) seems to be precisely c) on my list; that it is a process with both random and predictable elements. You do not agree with the statement "evolution is random", but you do agree with "evolution is a stochastic process" (again, according to wiki's definition).

If that's a fair summary of your position, we are in agreement, and I'm curious whether anyone here other than mijo disagrees with that.

[recursive prophet]

As would I, using the definition for "stochastic process" you gave earlier - and I haven't seen a single poster in this thread disagree with that. So if that's all you want to know, why are you still posting here?

On the other hand your little friend seems to think "stochastic" and "random" are the same:
Actually, the question I wanted to get around to asking before regarded whether or not natural selection was stochastic. Do you think it is? I started with evolution in general to see if there was any agreement on that here, and note Articulett never did answer my question as to whether or not she believed it was. My guess is she realizes that now saying it IS stochastic would contradict many of her earlier statements, on RD for sure and mijo has called her on it.

As to my 'little friend,' I can tell you he looms larger in my eyes by the day as I watch him let the relentless torrent of 'belittling' comments roll off his back like water off a duck and stick to the actual arguments made.

I think mjpam made a mistake letting himself try and respond to the way random and stochastic were being continually conflated on these boards, as I did in my earlier reply to you. The fire detector itself is an object that consists of some stochastic components. The process is stochastic, not the unit itself, as the process of evolution is a stochastic one. Unlike the fire detector, an object, evolution is a word FOR a process, and I think this has been part of the confusion here regarding evolution itself being stochastic.

As to why I’m still posting here I’m hoping to draw some more informed users such as Trane Wreck and My Wan with different viewpoints into the debate. Do I not have the right to do so?
You, on the other hand, have said smoke detectors are not random but are stochastic: and you agree that the two terms should not be equally applied:
Yes. This is the position of most on RD, but it depends on how you define random, and that very definition is being challenged right now by My Wan on Wiki.

So to summarize - and I'm being as fair as I can here - your position regarding smoke detectors (and evolution) seems to be precisely c) on my list; that it is a process with both random and predictable elements. You do not agree with the statement "evolution is random", but you do agree with "evolution is a stochastic process" (again, according to wiki's definition).

If that's a fair summary of your position, we are in agreement, and I'm curious whether anyone here other than mijo disagrees with that.
No to my choice on your list. If we use the word random I would pick d), none of the above. Smoke detector, unlike evolution, does not describe a process. And I too am curious if any here will deny evolution is stochastic, and you’d have to show me where mijo ever did so.


It appears at this point I have likely reached all that are interested in pursuing this subject here on JREF. One transplant, My Wan, has already made a valuable contribution to the RD thread on this same topic. It has caused much reexamination of my thinking, and her input alone has made my effort here worthwhile. If I could get Trane Wreck and a couple others to join in it would be gravy.

For those interested, you can jump into the recent discussion on RD here: http://richarddawkins.net/forum/viewtopic.php?p=1621596#p1621596 and read My Wan’s post there just a few down along with links to the Wiki dispute. If you do so you will see the questions posed in this thread are anything but as clear cut as many posting here would maintain. ..rp

[jimbob]

As I recall, mijo was pretending that evolution is random, which was stupid. If he is now reduced to saying that natural selection is stochastic, I guess that's slightly less misleading, though he would do better to say that evolution is a naturally occuring instance of a stochastic algorithm, then he wouldn't be wrong at all. But then, as his whole objective seems to be to describe evolution as inaccurately as possible, that would hardly suit him, would it?

I quite understand why the people on the Dawkins website have abandoned him as uneducable. They are right. He's been flogging this obsession for years, and it's not gotten any less dumb.


Dr A.

I have been accused on this thread of being incomprehensible, so I would like your opinion as to if/where you disagree with me in the following:

It is wrong to describe natural selection as "random", because that has connotations of "haphazard and unbiased". However it is more wrong to describe natural selection as "nonrandom", because that implies that all "beneficial traits" would be transmitted. I prefer "probabilistic" because beneficial traits make it more likely that an organism will reproduce, and thus the trait will be transmitted.

One can sometimes even assign values as to how beneficial a trait would be.

For example: Outside the malaerial belt and without medical intervention, two carriers of sickle-cell anaemia would suffer about a 25% reduction in their reproductive success; but inside the malaeria-belt these two carriers would have their reproductive succes improved by a factor of several times as about 50% of their offspring would be resistant to malaeria.

I would argue that with large populations, in "sufficiently-stable" ecosystems, over "moderate" timescales, evolution isn't "very" random (randomness could be ignored). However for small populations, random events can be important in the course of the evolution of their offspring.

Over very long (geological) timescales, I would also argue that large-scale random events are again important. The obvious one is the KT event, but there are many smaller events too. For example there are estimates that the Toba eruption about 70,000 years ago reduced the human population to several thousand individuals, which would have put it at risk from other, smaller "events".


I would also say that after a large set of extinctions, the ecosystem has changed significantly, so surviving organisms are less-well "optimised" and thus any change is more likely to be beneficial, so evolution would speed up. This would mean that there is a period when lots of new traits are evolving, so the selective pressures from the surrounding organisms in the ecosystem are in a state of flux.

In other words, in a rapidly changing environment, changes in one organism can drastically change the fitness landscape for other organisms.

On top of these responses to catastrophic events, I would also say that sometimes a "disruptive" mutation could arise within a stable environment, and cause dramatic changes to the fitness landscape.

I would argue that we have seen some evidence for this in Lenski's Long term evolution experiment. There were no significant differences in the initial environments in each of the jars, nor in their initial populations. However there was a seemingly (?) neutral mutation that occured in one population, after several thousand generations, which allowed a previously undocumented metabolism of citrate to evolve several thousand generations later. And repeatedly so. This mutation completely changed the very simple ecosystem in the experiment, and there is every reason to expect similar mutations to have similarly large effects in the wider environment. Mutations of pathogens being some obvious candidates for chance mutation having a large effect on an ecosystem.

(an earlier post about the LTE here (http://forums.randi.org/showthread.php?postid=4123374#post4123374))





When talking about the course of evolution, I don't see why we should avoid talking about the importance of random events.

[articulett]

But how would you answer Sol's query (#164)? And being evasive and not answering constitutes an answer identical to Mijo's.

Why shouldn't people who answer the question similarly be having the discussion with each other whether than playing semantic games with those who define terms differently?

[recursive prophet]

When talking about the course of evolution, I don't see why we should avoid talking about the importance of random events.
Nor do I, jimbob, and if you'd care to converse with others who share this view you really should check out this thread on RD.net: http://richarddawkins.net/forum/viewtopic.php?f=4&t=66951

I wouldn't hold my breath waiting for any meaningful reply from Dr. A; all you're likely to get are more ad homs-his standard response from what I've seen, to those that disagree with him.

Lenski's work has been thoroughly reviewed on the RD thread also, and I posted an email reply I got from him looking for clarification on some of his statements after they were quote mined by the only one there who has consistently argued that the question of stochastic natural selection is strickly in the realm of philosophy or semantics. You'll find many informed users-mostly grad students working in related fields-discussing this issue. We generally just ignore those such as Articulett, while users like my_wan-who joined us recently and can make a rational case-are engaged enthusiastically. Come on over!

[jimbob]

But how would you answer Sol's query (#164)? And being evasive and not answering constitutes an answer identical to Mijo's.

Why shouldn't people who answer the question similarly be having the discussion with each other whether than playing semantic games with those who define terms differently?


I wouldn't describe the operation of a smoke detector as random, for the same reason that I wouldn't describe the oporation of a steam engine as random.

Both rely on statistical processes large numbers of random events (atomic decay or brownian motion) but with such large numbers of individual random events that they can be averaged out and regarded as nonrandom.

If I was being pedantic I would opt for "c) the detection of smoke by smoke detectors has both random and non-random elements." But with the caveat that the numbers are so large and the system linear enough that the randomness is unimportant.

This is different from evolution, where there are situations where the randomness is averaged out. But there are also important situations where it isn't. These important situations can't be ignored if you attempt to answer how humanity evolved, or if you attempt to answer how any particular ecosystem evolved.


If you are considering "sufficiently large" populations, in "sufficiently stable" ecosystems, and over "moderate*" timescales then randomness is unimportant. This is not always the case in evolution.


In summary:

If random variations are insignificant in determining the outcome of a system, then it can be described as nonrandom. This is the case for a smoke detector, so it can be regarded as nonrandom. If random events would have a significant effect on the outcome of a system, then it can't be described as nonrandom. Random events are significant in the course of evolution, so evolution can't be described as nonrandom.



*"moderate" being thousands of years to millions of years...

[articulett]

So it all depends on the semantics of what you call the outcome.

Is an eye an outcome... or a specific organism...? is a biped an outcome or a specific biped? More semantics, eh?

Most would just answer c. But then again most people understand the importance of natural selection while those who don't, tend to mistake those of us who do for people who "avoid talking about the importance of random events."

Ken Miller (clearly more of an expert than you) in the bolded quote from six7s above would answer "b"--which is also correct because in the world of people defining things precisely, "b" is included in "c"-- "a" isn't. I think all evolutionary experts would agree with Ken Miller's testimony in contrast to Behe's need to "talk about the importance of random events."

The randomness is so easy to understand that even a creationist can understand it. It's the nonrandom aspects the ignorant have trouble with. You pretty much suck at describing natural selection in a coherent way--and this is due to your need to sound like Behe--a recognized blowhard-- rather than Miller-- a recognized expert.

If you want to be taken seriously by the experts, you might try sounding more like the experts and less like those you don't want to be mistaken for. Just a suggestion, of course.

[Darat]

Several reports from this thread. I've issued a couple of warnings however on further reading I'd rather not have to split the posts that do breach the Membership Agreement but also discuss the actual points under discussion so a reminder to all participants - do not continue to breach Rule 12 and Rule 11 else further action will be taken.

If you can't discuss the points raised in this thread without resorting to personal insults or personalising the discussion simply don't participate.

[jimbob]

So it all depends on the semantics of what you call the outcome.

Is an eye an outcome... or a specific organism...? is a biped an outcome or a specific biped? More semantics, eh?

I would consider the "shape" of an ecosystem to be an outcome that is significantly affected by random factors.

The toba eruption about 70,000 years ago nearly wiped out humanity; it reduced the population to such a level that the species must have been vulnerable to a run of bad luck. If this had happened, there would have been significantly different ecosystems, and thus evolutionary pressures for many species.

The black death probably helped precipitate the destruction of feudalism in europe, so it indirectly helped pave the way for the growth of Western Europe, and thus had a great effect on the evolution of many other species (e.g. the Dodo).

The KT impact, was vital in allowing the rise of mammals. After the impact, the ecosystem was in a state of flux, so selective pressures kept changing as organisms evolved, and changed the selective environment. Depending on which organisms acquired which traits, different types of ecosystem would arise, with differnet selective pressures.

I would also say that as some asteroid orbits are chaotic, whether and where a chaotically orbiting asteroid strikes the Earth is also affected by random factors. A hit/miss is a significant difference.

Flowering plants took a long time to evolve after plants colonised land, so the evolution of them wasn't inevitable for millions of years. Their rise certainly had a great effect on the shape of ecosystems, and the course of evolution.

The evolution of grasses also had a great effect, again as it took such a long time for them to evolve, they weren't inevitable, nor was it inevitible that another organism would fill exactly the same ecological niche, as some other plant could have partly filled this niche and prevented it being filled in that way.

The chance discovery of antibiotics has had a great effect on the evolution of antibiotic-resistant bacteria.

The evolution of bird-life on the Galapagos would have been different if some different birds to the ancestors of the Darwin's Finches had arrived earlier, and partly filled the niches occupied these birds today.




Most would just answer c. But then again most people understand the importance of natural selection while those who don't, tend to mistake those of us who do for people who "avoid talking about the importance of random events."

Ken Miller (clearly more of an expert than you) in the bolded quote from six7s above would answer "b"--which is also correct because in the world of people defining things precisely, "b" is included in "c"-- "a" isn't. I think all evolutionary experts would agree with Ken Miller's testimony in contrast to Behe's need to "talk about the importance of random events."

The randomness is so easy to understand that even a creationist can understand it. It's the nonrandom aspects the ignorant have trouble with. You pretty much suck at describing natural selection in a coherent way--and this is due to your need to sound like Behe--a recognized blowhard-- rather than Miller-- a recognized expert.

If you want to be taken seriously by the experts, you might try sounding more like the experts and less like those you don't want to be mistaken for. Just a suggestion, of course.

If you looked I said

I wouldn't describe the operation of a smoke detector as random, for the same reason that I wouldn't describe the oporation of a steam engine as random.

Both rely on statistical processes large numbers of random events (atomic decay or brownian motion) but with such large numbers of individual random events that they can be averaged out and regarded as nonrandom.

If I was being pedantic I would opt for "c) the detection of smoke by smoke detectors has both random and non-random elements." But with the caveat that the numbers are so large and the system linear enough that the randomness is unimportant.

which *is* "b" or "c but the randomness is unimportant"

[sol invictus]

If you are considering "sufficiently large" populations, in "sufficiently stable" ecosystems, and over "moderate*" timescales then randomness is unimportant. This is not always the case in evolution.

It's not always the case for smoke detectors, either. Try predicting precisely which curl of smoke will set it off. Or how to stop it most efficiently (not counting taking out the batteries!). Or how to describe the cases where it has malfunctioned. Etc.

Some aspects of the macroscopic behavior is predictable, and some are not. The whole thing is based on random microscopic processes, which sometimes alters the macroscopic behavior in unpredictable ways.

If random variations are insignificant in determining the outcome of a system, then it can be described as nonrandom. This is the case for a smoke detector, so it can be regarded as nonrandom. If random events would have a significant effect on the outcome of a system, then it can't be described as nonrandom. Random events are significant in the course of evolution, so evolution can't be described as nonrandom.

That's based on a rather random judgment about what you personally at the moment choose to regard as "significant". You've failed to make a meaningful distinction.

[Molinaro]

So you believe that a process which relies in a fundamentally important way on random events but which produces outcomes which are highly predictable in some aspects is "not random"?

That's a win right there.

Why did the thread continue past this point? Shouldn't making the point so clearly be enough to get the OP and their supporters to agree? :rolleyes:

Edit to add: Wow... I posted after only reading the first 2 pages, and loh and behold my comment lands 3 pages later, right next to more smoke detector talk!

[articulett]

In fact, Dr A answered it in his very first response only some people didn't like the answer.

The original question was in fact split off from a derail regarding the facts of evolution.

I, too, think Sol nailed it at the get go and then hammered the point home.

This thread should be closed and locked.

[jimbob]

Sol, the exact operation of smoke detectors) is affected by random factors, but fire detectors have a threshold level below which they will not alarm, so it will alarm when there is a certain level of smoke, either ionisation current or optical obscurance/reflectance within certain tolerances.

The smoke detector will alarm at a certain level of smoke within a certain tolerance. if oyu are only interested in whether the alarm will detect a fire or not, random factors are unimportant.

They will alarm within a certain time for specified fire conditions (EN54-9 (http://www.google.co.uk/search?hl=en&client=opera&rls=en&hs=aZq&sa=X&oi=spell&resnum=0&ct=result&cd=1&q=EN+54-9&spell=1) is the EU standard, I can't remember the UL or Australian standards - prior to designing semiconductor devices I designed fire detectors and sensors)


If random variations are insignificant in determining the outcome of a system, then it can be described as nonrandom. This is the case for a smoke detector, so it can be regarded as nonrandom. If random events would have a significant effect on the outcome of a system, then it can't be described as nonrandom. Random events are significant in the course of evolution, so evolution can't be described as nonrandom.
That's based on a rather random judgment about what you personally at the moment choose to regard as "significant". You've failed to make a meaningful distinction.

No, it is based on a subjective judgement.

However I consider that the evolution or not of a sophisticated-tool using social-species to be significant, just as the rise of the mammals after the KT impact to have been significant, ditto for the evolution of grasses.

Do you consider the KT impact to have been significant? Do you consider that its effects on the Earth's ecosystem were also affected by random factors?

[sol invictus]

Do you consider the KT impact to have been significant? Do you consider that its effects on the Earth's ecosystem were also affected by random factors?

What, you don't think a giant asteroid smacking into it will affect a smoke detector?

Your argument is completely absurd. Of course random events like that can have macroscopic and important implications for evolution - just as they can for everything else in the world. Which is the whole point - this discussion is an incredible waste of time (unless you have an ulterior motive).

[jimbob]

Originally Posted by jimbob
Do you consider the KT impact to have been significant? Do you consider that its effects on the Earth's ecosystem were also affected by random factors?
What, you don't think a giant asteroid smacking into it will affect a smoke detector?

Your argument is completely absurd. Of course random events like that can have macroscopic and important implications for evolution - just as they can for everything else in the world. Which is the whole point - this discussion is an incredible waste of time (unless you have an ulterior motive).

Not really Sol,

My point is that calling evolution nonrandom is incorrect. Calling it "random" is misleading, but slightly less wrong.

The KT impact and other events have been part of the actual evolutionary history, and part of the "normal operation" of evolution.

Nothing to do with the "normal operation" of smoke detectors.

That was only an extreme case.

In Lenski's Long Term Evolution experiment, we have got as close as we can get to "rerunning the tape of evolution" because he has frozen samples from each population at different stages.

Because of this, we can see that "rerunning the tape of evolution" doesn't always produce the same results.

In one population, citrate metabolism in e.coli has been observed to evolve and I would say that this is the best candidate, so far observed, for the chance evolution of "disruptive" trait that completely changes the selective landscape.

We know that the cause was a mutation, and not a subtle difference in the environments of the jars. We also know that this mutation occurred between 15,000 and 20,000 generations into the experiment, because descendents from the "right" population repeatably evolved this ability when using samples more recent than 20,000 generations, and that (with the exception of the parent population) they didn't when using samples older than 15,000 generations.

This experiment is incredibly elegant.

It shows that in the evolution of this trait, externally-applied differences in the environment were unimportant, compared to the internally-applied changes to the environment caused by a rare mutation.

If externally-applied environmental differences were the cause, then there wouldn't be a trigger point between 15,000 and 20,000 generations.

This is still a very simple environment, and with more interactions between more different organisms, the potential for "disruptive traits" changing the selective landscape would be greater.

As I said before, saying "evolution is random" is liable to mislead, but saying "evolution is nonrandom" is contradicted by the evidence.

I wouldn't want to use either phrase, but one is only misleading and allows the subtleties to be explained, whilst the other is wrong.

I would prefer to say that traits, or collections of traits evolving within a certain timescale range from "very likely" to "impossible". Sometimes it might be possible to quantify the probability of occurrence within a certain timeframe.


Describing evolution as "nonrandom" strikes me as the most arrogant anthropocentricism, because this would mean, that as humans have evolved, our evolution was inevitable, so maybe the creator set up the Earth so that worshipers were inevitable.* In fact we are the result of chance occurrences in an uncaring universe. Something amazingly well-adapted to its environment would have evolved, but not necessarily any sentient organism.


*In the UK, I would suggest that most Anglicans and other similarly "liberal" pseudo-secular religious people subscribe to this view. It is one that I have heard a lot.

[jimbob]

Or the short version to the above post:

I contend that the course of evolution is significantly affected by random influences, so that the course of evolution is not inevible*. If we were to have identical universes at the same moment in time (say the KT impact), with only future random events differing, then 65-million years later, the ecosystems would probably look quite different.

There was nothing inevitible about the evolution of mammals, nor of dinosaurs, nor indeed any liiving organism. Following each extinction, the flora and fauna have recovered their complexity, but they were also significantly differnent. There are similarities, of course, herbivores and carnivores, for example, but there have also been differences.


*EDIT: Don't we all laugh at the astounding predictions as to what humans will evolve into? One of trhe reasons why that is so ridiculous is because future evolution hasn't been determined yet because it is subject to future random influences.

[recursive prophet]

What, you don't think a giant asteroid smacking into it will affect a smoke detector?

Your argument is completely absurd. Of course random events like that can have macroscopic and important implications for evolution - just as they can for everything else in the world. Which is the whole point - this discussion is an incredible waste of time (unless you have an ulterior motive).
Thanks to an impressive deconstruction of the thread addressing this topic on RD.net by my_wan, I now understand much better why you and others here believe this whole discussion is a waste of time. Saying it is absurd, or constructing smoke detecting straw men to attack was not really helpful, and I did answer your quiz on this page. It was similar to Jimbobs reply; a smoke detector is a device, not a process, as is evolution. If you ever bothered to reply to either of us regarding this point, I missed it.

So, unless someone posts something new on this thread or poses a question, I believe it has achieved my purpose in posing the OP question here. my_wan made a persuasively precise rebuttal to much of the debate on RD, and much was revealed in his analysis. However susu has yet to reply to his revelations of the last 2 days, and I haven't gone back over the many posts both have linked, so if something new comes up I will post it here. For those wishing to read my_wan's commentary, here's a link to a good place to begin: http://richarddawkins.net/forum/viewtopic.php?p=1622718#p1622718 His replies and the responses to them go on for another 6 pages, but I personally found them very informative, if largely limited to arguments made against one specific user there. Thanks again for sharing my_wan.

[sol invictus]

It was similar to Jimbobs reply; a smoke detector is a device, not a process, as is evolution. If you ever bothered to reply to either of us regarding this point, I missed it.

Then you have a very short memory indeed. I replied, and you replied to that, and I replied, and you replied...

[articulett]

Then you have a very short memory indeed. I replied, and you replied to that, and I replied, and you replied...

He's trying revive it at Dawkins too... he imagines an epic battle over this semantic inanity.

My_Wan handed him and his "buddies" their collective asses (it was beautiful), but it just made them feel more confident that somebody of caliber took them seriously.

He has a very short and misleading memory indeed. Mijo-esque.

[recursive prophet]

Thanks to an impressive deconstruction of the thread addressing this topic on RD.net by my_wan, I now understand much better why you and others here believe this whole discussion is a waste of time. Saying it is absurd, or constructing smoke detecting straw men to attack was not really helpful, and I did answer your quiz on this page. It was similar to Jimbobs reply; a smoke detector is a device, not a process, as is evolution. If you ever bothered to reply to either of us regarding this point, I missed it.
Then you have a very short memory indeed. I replied, and you replied to that, and I replied, and you replied...
Yes, it’s short, but not to that degree. I’m referring to the one I made to your updated quiz-use the link below or just scroll up.
http://forums.randi.org/showthread.php?postid=4309092#post4309092

No, RP, most don't agree with Mijo there... just like on this forum, the smart people posted and left... [emphasis mine]

So, given the fact that Articulett has more replies on this thread than anyone else and is still here on page 5, what am I to deduce from her 2nd reply on page 1 of this thread?

Ahhhh, Articulett. You make me think of that old Dylan classic. "I wish that for just one time you could stand inside my shoes, and just for that one moment I could be you.... Then you would know, but you don't, and sadly never will.

I deduced long ago from your comments that you rarely bother to read what others say. Just one of countless examples I could-and will if you continue-bring up is your recent comment on the similar RD thread. As on this one, you continue to say is not worth posting on and that all the 'smart people' have left. Yet on both threads YOU are the most active poster. The evidence is there for anyone to see. Your quote: “I just wanted to say this in case RP didn't ever thank your allowing him to recruit you into this unhinged derailment.” http://richarddawkins.net/forum/viewtopic.php?p=1636369#p1636369

The next link is to one of 3 posts I sent before you did the one above on which I expressed sincere gratitude. In this one I concede he has literally ‘kicked ass,’ but reserved final judgment until I’ve done what your words suggest you never do..read everything carefully. I'm also certain you didn't read the 6 pages of my_wan's comments, let alone the counters. Just the difference in our number of posts tells much of the story. YOu like to 'express yourself,' I like to read new ideas. Here’s the link for any interested:
http://richarddawkins.net/forum/viewtopic.php?p=1635155#p1635155

Articulett, I really do wish you would stick to arguing with the creationists and theists. How can you possibly rationalize what your posts here and on RD reveal as a true obsession with this topic? Can you actually not realize every time you reply on this or the other thread on RD you just help keep it alive? That you never say anything new? That anyone who gives it a moment’s thought will ask why you keep beating on what you proclaim to be a dead horse? At least read your own words if nobody else’s. You clearly didn’t even read my last post before making your latest inaccurate claim, and please spare us the ‘ignore button’ BS. You only embarrass yourself with it. I predicted on the RD thread before my_wan made his first post there that he would genuinely shake things up, and later that he had exceeded my expectations. I LOVE having misconceptions shattered, and will be pleased and impressed by my_wan's argument whether it holds or is itself refuted. Don't even try and understand why. You would never get it. Just please stop whining about how lame the questions are; simply don't read or reply to this or the RD thread anymore, as you have promised to do countless times. But if you MUST, do it here; the discussion on RD is far more in depth and your disparaging remarks are more of a distraction there. Please??

[jimbob]

So does anyone disagree with the following statements, and if so, where and why?

"Until relatively recently in the Earth's history, there was no inevitibility in the evolution of a social animal that used complex tools requiring more than three different types of operation to make them."

"During the history of life on Earth, the course of evolution has been significantly altered by random events."

"Furthermore the fitness landscape and the consequent direction of selective pressures are sometimes (infrequently) subject to change due to random mutations that provide such a significant advantage that the entire ecosystem is disrupted. This means that the selective pressures on other organisms alter to such an extent that completely different traits begin to evolve in the descnedants of other organisms in the ecosystem in response to this disruptive mutation."

[recursive prophet]

Jimbob: I doubt any here will disagree with any of your statements except possibly the last which I find interesting. They will likely just consider them irrelevant as regards the OP question, and to the process of evolution as a whole. Why? Well, say a meteorite did hit and destroyed all mammals. This would not have any impact on whatever changes the evolutionary process had achieved up until that moment other than the fact they would all be canceled out when they died. That makes their ultimate fate random, but it’s still not relevant to the process that got them to the point they were at when the meteor hit. I’m guessing though.

It seems to break down to a philosophical question, like the tree falling in the forest that no one hears. My intuition on randomness is similar to yours. I agree with those that believe the universe is random, not deterministic, though there be far more advanced minds than my own on both sides.

Sol wrote: “What kind of pigheaded arrogance does it take to think you can accurately describe something as complex as evolution in one word, when you can't even agree how to apply it to the simplest of everyday objects?”

Can’t believe I forgot this before. Show me where anyone claimed stochastic-the word used in OP question-or random-the term used by many posters here-was a one word description of evolution.

We’re talking about a process, and whether it is stochastic or deterministic is but one aspect of that process, as is the much overlooked genetic drift and its effects on small populations. You were the one that conflated a process-evolution-with an object-smoke detector.

No one I know of argues everything is random and that determinism doesn’t exist. The stochastic lumpers just believe it is a more preeminent force than the splitters. Similar divisions occur within the determinist community. The question was never stochastic or deterministic, all or nothing. Sadly both sides seem bent on creating this straw man. The OP question was aimed at nothing more than seeing what the consensus was on JREF, and contrary to your earlier comment I would bet many on this site would say evolution, and particularly natural selection, were in fact deterministic. Quite a few did so on RD, some quite convincingly.

Think this subject appears played out. Never intended to create a new thread on this, of course, and am sort of amazed it’s gone on for 6 pages. I’m going over the entire 2400 replies on RD and may post a thread later with all the key points made condensed. My_wan has brought the discussion to a halt as I suspect most-like me-are going back into the archives looking for data to support/deny his many assertions. I see a few slips he made, but it will take a while to piece it all together, and I’m now very open minded to the idea that he has got most of it right, and that at core we have a ‘reducible randomness' vs. 'irreducible randomness' debate, and that neither side is wrong as both positions are based on plausible but improvable assumptions. We shall see.

[cyborg]

A meteorite crashing into the Earth is only a "random" event due to you ignorance of what bits of rock are out there and what their paths are.

It's not like this hasn't been pointed out before but sure - you go ahead and ignore it again.

Fundamentally I just don't think you understand what a model is and what it isn't.

[Paulhoff]

http://www.ted.com/index.php/talks/paul_sereno_digs_up_dinosaurs.html

Paul

:) :) :)

[recursive prophet]

A meteorite crashing into the Earth is only a "random" event due to you ignorance of what bits of rock are out there and what their paths are.

It's not like this hasn't been pointed out before but sure - you go ahead and ignore it again.

Fundamentally I just don't think you understand what a model is and what it isn't.
OK. How about you tell me your definition of a model, and what about them I fail to comprehend? Fundamentally. As to the old chestnut on nothing being actually random if we achieved omnipotence and could know say, a photons speed and position, I would ask if you are suggesting such an entity exist or is probable? There will always be ignorance of key facts, like whether randomness is irreducible. As overall deterministic patterns can not be verified in selection, the approach of most of those currently working in evolutionary biology who have posted on the ‘sister’ RD thread is that it is in fact a stochastic process. If we’re going to use ‘random’ we have to agree on a working definition. I've learned from reading, and watching mijo get caught in that semantics trap. And of course even if you win that game, your opponents will create a "your whole argument is based on semantics" straw man, and quickly tromp it.

In a way I suspect this comes down to a possible recent shift in biology toward greater reliance on mathematical models. With number crunching capability beyond anything conceivable 3 decades back, old paradigms die hard and only truly disappear when the last of the ‘old school’ profs retire. I'm not saying I'm totally convinced that this is the case in the debate here and on RD over the nature of evolution and NS. But I’m certain, after many hours re-reading, that those making the argument that the entire process is stochastic have done a much better job making their case. The challenges by susu, Dlx2 and Marios are out there waiting now; we shall see if they are addressed.

[cyborg]

OK. How about you tell me your definition of a model, and what about them I fail to comprehend?

This:

As to the old chestnut on nothing being actually random if we achieved omnipotence and could know say, a photons speed and position, I would ask if you are suggesting such an entity exist or is probable?

This fundamentally shows that you do not understand that the model of evolution does not rely on any particular physical reality. You can talk about evolution without knowing the nature of photons - at all.

I pointed out earlier that genetic algorithms must BY DEFINITION be deterministic.

If the pseudo-randomness employed by such algorithms means they are doomed to be an insufficient analog for biological evolution then I demand an explanation as to why.

Otherwise this is a continuation of the appeal that somehow the properties of the subatomic should dominate how we label the models of large macro scale events.

Sorry, no. Not going to happen. I'm going to continue to not give one iota of a **** about the properties of atoms when talking about evolutionary systems.

[articulett]

Few can cut through the BS like cyborg.

Bravo!

[CFLarsen]

Which should eliminate further responses.

[Walter Wayne]

If the pseudo-randomness employed by such algorithms means they are doomed to be an insufficient analog for biological evolution then I demand an explanation as to why.
There are several reasons why using genetic algorithms are insufficient for making statements about biological evolution. Many of the significant differences between them I outlined in a previous incarnation of this argument.

But, when you look at the development of genetic algorithms in computer science, it is obvious why it an insufficient analog with respect to this particular argument. During the development of the method it was found that care had to be taken in constructing the algorithm in order for it to always find the solution sought. If not constructed properly genetic algorithms does not find all solutions, and could also converge to local minima of the error function that didn't represent a solution.

Of course, the results of said algorithms were still deterministic. Given the same seed for the PRNG they always found the same solution. But if this seed was changed, the particular solutions found varied. A random input would actually produce a random output.

Of course, with tuning, you can get the algorithm to given the same solutions for any particular seed.

If genetic algorithms are a good analog, you still have to ask which genetic algorithm best represents the nature of evolution? The one whose output is invariant with the starting point of the PRNG, or the one that converges to different solution given different "random" input?

Walt

[cyborg]

During the development of the method it was found that care had to be taken in constructing the algorithm in order for it to always find the solution sought. If not constructed properly genetic algorithms does not find all solutions, and could also converge to local minima of the error function that didn't represent a solution.

You don't think local minima exist in biology?!?

Yep, it's certainly a problem - but then what you do about it depends on what you are using it for and not all uses are for modelling biological processes.

Of course, the results of said algorithms were still deterministic. Given the same seed for the PRNG they always found the same solution. But if this seed was changed, the particular solutions found varied. A random input would actually produce a random output.

Right,

For the final time - in unamibigous Unix terms:


/dev/random > random_text

random_text | ./evolution.pl > result1
random_text | ./evolution.pl > result2

diff result1 result2


Given ANY choice of random device connected to a Unix box - be it a hardware based, software based, faith based - do you ever expect to see a difference here?

Please explain to me how the output is random.

Of course, with tuning, you can get the algorithm to given the same solutions for any particular seed.

Huh?

If genetic algorithms are a good analog, you still have to ask which genetic algorithm best represents the nature of evolution? The one whose output is invariant with the starting point of the PRNG, or the one that converges to different solution given different "random" input?

Again: huh? We've already established that if input is different then output can vary and if it is not then output cannot vary. I don't really understand the distinction you're trying to make here.

[jimbob]

A meteorite crashing into the Earth is only a "random" event due to you ignorance of what bits of rock are out there and what their paths are.

It's not like this hasn't been pointed out before but sure - you go ahead and ignore it again.

Fundamentally I just don't think you understand what a model is and what it isn't.


There are plenty of physicists who would disagree with you:

Sorry, I meant a single quantum event influencing the outcome of a die roll; perhaps some relevant (i.e. relevant to classical motion, not dice specifically) empirically determined data.

Yes, that is what I'm looking for. I'm having trouble finding something that specific (that's not related to consciousness).

Thank you.

LindaOK, here you go:

http://findarticles.com/p/articles/mi_m1200/is_n14_v138/ai_8986262

This discusses how, in classically chaotic situations like rolling dice, the uncertainty of the underlying quantum description of the system increases to permit the chaotic behavior.



A significant proportion of Earth-crossing asteroids are in chaotic orbits: (google search results (http://www.google.co.uk/search?client=opera&rls=en&q=earth+crossing+asteroid+chaotic+orbit&sourceid=opera&ie=utf-8&oe=utf-8))

This means that their orbits are very sensitive to perturbations, and that their orbits are in-principle completely unpredictible over timescales of (say) 20-million years or longer.

If you find it hard to accept that the orbit can be significaltly altered by quantum events (indeed simply uncertainty in the asteroid's position and momentum) then maybe a slightly more "large-scale" set of events might be easier to grasp:

Chaotic orbits are so sensitive to perturbations that even the gravitational influence from an individual animal's movement would significantly alter the eventual orbit of the asteroid. We know that single photons can cause cancer or mutations, so the random and undecided future absorbtion of a photon of ionising radiation could affect the movement of an animal (by killing it).

Of course it is impossible in principle to calculate what this effect would be on the asteroid, because other random effects are bigger, as are non-random but unpredictible events, but it means that (unless something happens in the meantime) the Universe itself hasn't yet "decided" whether a particuar Earth-crossing asteroid will collide with the earth within the next 20-million years.

[recursive prophet]

If genetic algorithms are a good analog, you still have to ask which genetic algorithm best represents the nature of evolution? The one whose output is invariant with the starting point of the PRNG, or the one that converges to different solution given different "random" input?

Walt
Excellent point, Walt; garbage in; garbage out. Perhaps it’s something Cyborg fails to grasp about models? Perish the thought! Some new methods have come a long way in far exceeding the PRNG state, and I discussed this weeks ago with susu. Haven’t gotten up to pt. 3 yet in my condensation of the RD thread, but when I do I’ll try and remember to post a link here. OR, you can just go to the link below and drop back about 5 pages.

This:

This fundamentally shows that you do not understand that the model of evolution does not rely on any particular physical reality. You can talk about evolution without knowing the nature of photons - at all.
OMFG! Do you mean to tell me the map isn’t actually the terrain? And that we can just forget about those silly QM shadows and evolution will go on? M.erde! (If you’ll pardon my French) If only I had known. You are indeed a fundamentalist.

I pointed out earlier that genetic algorithms must BY DEFINITION be deterministic.
Could you please expatiate on this? I did ask for your definition of a model.
If the pseudo-randomness employed by such algorithms means they are doomed to be an insufficient analog for biological evolution then I demand an explanation as to why.
Afraid you’re conflating my words with others, or this is your interpretation of what I have stated. Just as you have islands of increased organization in the stream of entropy, you have deterministic events within stochastic processes. Do you see a dichotomy?
Otherwise this is a continuation of the appeal that somehow the properties of the subatomic should dominate how we label the models of large macro scale events.

Sorry, no. Not going to happen. I'm going to continue to not give one iota of a **** about the properties of atoms when talking about evolutionary systems.
Wow. You sure kicked that straw man’s butt!! However, while Bell’s Theorem and the possible QM implications for evolution did come up a few times as an aside, care to send me a link to where anyone has built their case for stochastic selection based on QM? Everyone knows that’s a dead end pending far more research. But wrap it in straw and put the arms on hinges so they flap up and down and soon you can convince yourself it’s a concise ‘model’ of what I am trying to relate. You should understand I don’t create any hypotheses on this subject. I am strictly an inquiring student, passing on lessons from others for further critique.

I used Heisenberg’s uncertainty as a mere example of the unknowable, not to introduce particle physics into the discussion. It is completely unnecessary for making the case that selection is stochastic; the macrocosm abounds with examples. susu’s recent reply to my_wan sums up the perspectives of his many allies on the much larger, more detailed debate on the RD thread.

Just because MZ thinks we´re talking about a philosophical difference here, doesn´t make it so. He´s been making non-philosophical claims and retreated to this when repeatedly being shown wrong. His argument is misrepresenting at least my view, when he states that we are trying to argue that evolution must be viewed as a whole. I´ve got no problem breaking it down into component processes (his later arguemnta however have this problem; there´s no differentiation between selection, drift and non-selective sorting).

My point is that the selection component is stochastic.
When dXi/dt is the change of the frequency of allele i in time it can be written as the sum of several processes:
dXi/dt= Selection + Mutation + drift +...
dXi/dt is a random variable, because several of the terms are random variables (any sum containing at least 1 random variable is itself a random variable).
The Selection component is a random variable under the condition that it explicitly depends on any X - i.e. the frequency of any allele. And that is a given, because the selection component has to be 0 if Xi is 1 or 0 (the only way for the selection component to be independent of Xi itself would be if it was constantly 0 - but then there is no selection - you’ve got a neutral case).

Those interested can read susu’s entire response along with comments by others currently working in studying evolutionary biology. Just go back 1 page. http://richarddawkins.net/forum/viewtopic.php?p=1648236#p1648236

And a post from last August by Marios.

one cannot not know with certainty whether a specific individual possessing initial conditions will reproduce.

I think that point was grasped a while back - I think the contention is that the stochastic effects are just 'noise' distributed around a deterministic mean and can be safely neglected. In some cases that's true (in the appropriate limits), but only if you dramatically simplify the system.

One usual simplification is assuming that the (abiotic) environment selects the organism, but that the organism has no effect on the environment - remove that and you have non-linear equations, even if the environment is otherwise totally static, stochastic fluctuations in the population will lead to stochastic fluctuations in the environment leading to stochastic fluctuations in the selective pressures leading to further changes in the population leading to further changes in the environment ... The concept is known as niche construction (similar to an extended phenotype, but focusing on permanent changes to the environment).

Marios

For any interested in actually discussing the issues raised in the now close to 100 pages of the Why evolution is not purely random thread on RD, I would invite you to join in. my_wan made some very interesting arguments, and I learned a lot going back over all the earlier exchanges again. Now that I have done so, I'm convinced that my_wan in fact created yet another straw man to attack. In this case it was constructed with what one poster had stated re: the discussion being entirely philosophical, and completely ignored all the other comments aimed at demonstrating how selection was in fact stochastic. He has several standing rebuttals to his analysis awaiting him, including the one linked above by susu. Somehow, I doubt he will be back, content with having soundly defeated the figure he planted in the cornfield for the benefit of crows passing by. Read it for yourself.

As a long time admirer of James Randi, and one closer to his age than I am most of the posters here, I must confess it saddens me to see what I perceive as the ‘gestalt’ of this forum. Brash and confrontational are IN, and there appears to be far more emphasis on sounding clever and belittling those with opposing views than any form of collective search for greater understanding. That’s not to say it doesn’t happen here; just that it seems far from the dominant motif. And so it goes. :(

Just noticed your reply to Walt, Cyborg, and will comment on it next time. I'm sure that will make your day. :cool:

[jimbob]

Genetic algorithms are good models for evolution *but* they are only partial models.

All the evolutionary algorithms that I have seen in engineering have been aimed at selecting for particular predefinerd notions of fitnesss, e.g. antenna-efficiency, or the ability to perform a particular set of functions in an FPGA (google search here (http://www.google.co.uk/search?client=opera&rls=en&q=evolutionary+algorithms+FPGA&sourceid=opera&ie=utf-8&oe=utf-8)).

When you use these algorithms, the working iteration does what it has been selected to do. This is not surprising. Because these variants do not replicate themselves, an external agent needs to actively select those variants to form the seed of the next generation. This active selection by an external agency is different to how evolution works in life. It is also misleadingly close to what some ID proponents would like evolution to actually be.

These ones for example (http://www.creationsafaris.com/crev200709.htm)
ETA: Their bolding

Can Humans Use Evolution? 09/10/2007
Evolution is being used. A press release from University of Wisconsin-Madison was titled, “Using evolution, UW team creates a template for many new therapeutic agents.” How does one use evolution? It continued, “By guiding an enzyme down a new evolutionary pathway, a team of University of Wisconsin-Madison researchers has created a new form of an enzyme capable of producing a range of potential new therapeutic agents with anticancer and antibiotic properties.”

<snip>
Such stories do nothing but obfuscate. Evolution has nothing to do with it. These scientists had a goal, and purposely selected enzymes with the properties they desired. Darwin team, the referee just blew the whistle. The penalty for your foul is to yield those two points to the ID team. The title now reads, “Using selection criteria by intelligent design, UW team discovers a template for many new therapeutic agents.” Much better; play ball. Whoops; we just remembered the Darwin team has no goal. Game over by forfeit.



My answer the statement hidden above is that the original press release could have been better:

“Using evolution, evolutionary approaches UW team creates a template for many new therapeutic agents.


Living organisms are only selected by reproducing. What they *do* is reproduce, because that is what they are selected for. Many different strategies have evolved to do this.


Evolutionary algorithms show that you don't need a designer to get to a particular goal, and that complex systems can be developed if they are selected for. If a system reproduces, then viable variants will "select" themselves, and the selection criteria will be "that which reproduces". Unsurprisingly, in such a case, very complex "seeming-designs" can evolve.

[cyborg]

All the evolutionary algorithms that I have seen in engineering have been aimed at selecting for particular predefinerd notions of fitnesss,

*SIgh* and as I pointed out before this is not true in all cases. Having personally dabbled in evolutionary simulations I know that it is *incredibly* easy to not construct a particular predefined notion of fitness. You just don't define one.

Do you see a distinction here or not?

[cyborg]

OMFG! Do you mean to tell me the map isn’t actually the terrain? And that we can just forget about those silly QM shadows and evolution will go on? M.erde! (If you’ll pardon my French) If only I had known. You are indeed a fundamentalist.

I fail to see how.

I am not the one continually referring back to the nature of fundamental particles as important components in understanding evolution so I want to know why you are.

Could you please expatiate on this? I did ask for your definition of a model.

A model is an abstract system created to be an analog of another system - usually physical but not necessarily so - that removes features of the system being modeled (otherwise it would be the system) not considered important enough to affect conclusions about the modeled system allowing the model to be more easily reasoned about.

E.g. ignoring the particulars of how any given genetic variation comes about still allows us to reason about the effects of genetic variation.

Afraid you’re conflating my words with others, or this is your interpretation of what I have stated. Just as you have islands of increased organization in the stream of entropy, you have deterministic events within stochastic processes. Do you see a dichotomy?

I don't understand what you're trying to say here but if you're asking if I see a deterministic/random dichotomy then I don't - and have continually argued against it.

Wow. You sure kicked that straw man’s butt!!

That is the argument presented.

If you can explain to me why the continued discussion on QM is relevant to evolution I will listen. So far all I can see is appeals that somehow the quality of the randomness is very important without any particularly good arguments as to why.

However, while Bell’s Theorem and the possible QM implications for evolution did come up a few times as an aside, care to send me a link to where anyone has built their case for stochastic selection based on QM?

This thread and others like it.

It isn't me who brings it up. I simply don't see how understanding QM helps us to understand evolution so the only time you'll see me talking about them in the same breath is in these discussions.

I used Heisenberg’s uncertainty as a mere example of the unknowable, not to introduce particle physics into the discussion.

Right so can we stop talking about particle physics now and assume that I understand what "unknowable' entails?

[jimbob]

*SIgh* and as I pointed out before this is not true in all cases. Having personally dabbled in evolutionary simulations I know that it is *incredibly* easy to not construct a particular predefined notion of fitness. You just don't define one.

Do you see a distinction here or not?

Do you mean that they self-replicate within the environment of the simulation?

[cyborg]

Do you mean that they self-replicate within the environment of the simulation?

It doesn't exactly take a master programmer to figure out a simple way to model mitosis - it's just copying data.

[articulett]

Cyborg

I'd say RP falls between mijo and jim-bob on the penetrability scale...

make of that what you will--and temper all expectations accordingly.

The more you try to clarify, the nuttier the straw men will get-- do you really want to play "semantics of the silly" again? I do love reading the responses, but I suspect engaging them makes them feel like someone takes their nonsense seriously. They cannot learn because they think they understand more than those who would take the time to give them a clue.

RP is trying his damndest both here and at Dawkins.net to gather in people for yet another round of "evolution-is-random". It's amusing, but he can get rather nasty to those who don't corroborate his imaginary expertise. And he will run to the mods when you dish back... plus, plowing through the assorted word salads... how do you stomach it?

I don't think anyone you respect is paying attention to his hyperbolic ("OMFG...") insults of others, so there is no need to defend against his mischarachterizations. I think the smart people know who the smart people are. Let those who would answer Sol's query different than you, talk at each other. Your knowledge is wasted on them, because they are certain that they know more than you. Is there any evidence that such people can get a clue? Their point is always the same-- to prove that evolution is "random"... whatever vague thing that might mean to whomever and attack all those who aim for a little more clarity. I've learned not to trust people who are more stuck on the semantics of describing evolution rather than conveying understanding about the process. Remember, the incompetents are too incompetent to realize they are the incompetents.

Let them prove to each other how random evolution is due to meteors or stochastics or QM or same starting points having different ending points or because it's related to probabilities etc. Per their semantics, evolution IS random-- and so are elections, smoke detectors, poker, and everything else. Who can have a discussion with someone who uses words so vaguely and uselessly?

Your intelligence can be utilized on people who can engage in dialogue clarify, and learn. You make a coherent point in every post, and they use tons of words, and their point is lost amidst the verbiage. I mean what is their point except to prove to themselves that some scientist (other than Behe?) thinks it makes sense to describe evolution as "random" or in terms of randomness as they fight tangential straw men and pretend that those who are more clear are refusing to acknowledge the "random" elements?

[articulett]

ugh!-- they do not know the difference between stochastic, chaotic, and random... to the buffoons, it's ALL random if it any part of it is random.

[recursive prophet]

Right so can we stop talking about particle physics now and assume that I understand what "unknowable' entails?
We sure can, and I regret the example I used to make the point that we will always face 'unknowables,' or 'undecidables' as Gödel showed even mathematics contained. While compiling some thoughts on your last reply I came across an exchange right up your ally, Cyborg. Are you a Root somewhere? As you seem familiar with UNIX models and confident you’re on top of how these models function, tell me your response to susu’s reply to my_wan below. I have dabbled with gnu, and on another forum rms is actually one of my few ‘friends’ there. But my understanding of the mechanics involved are extremely abstract, and as relatively ancient as I am. My point? I’m not asking your opinion to make one. Susu is a biologist, and perhaps he is feeling the elephant’s trunk. You tell me. Unlike many here, I only provoke to seek answers.

I cut the remainder but I wanted to comment on your characterization of stochastic models being more powerful. Not necessarily, it depends on what you call powerful. Stochastic models gain computational ease by throwing away details of the process.

I disagree, for the reason mentioned above: You need to run deterministic models only once, stochastic models need repeated runs. Stochastic models are not necessarily simpler than deterministic ones, either. As mentioned in my reply to RP, there are several questions in evolutionary biology where both deterministic and stochastic models exist, both applicable to the same data, both having the same detail. But the deterministic ones are generally not used in publications, except as a heuristic - they run on a home computer, while the stochastic ones run for a couple of weeks to months on supercomputers. In terms of computations, I don’t know a single case where deterministic models take more computation power than stochastic ones. Computational ease is not the reason to opt for stochastic models.

[cyborg]

I don't think I can comment on that in anything other than generalities.

I don't know what distinction here is being made between "stochastic" and "deterministic" models - especially if, as the above seems to apply, they are applying to equivalent models.

It seems like "deterministic" and "stochastic" are simply being used in place of "low/high problem space/complexity/number of runs".

[jimbob]

Cuborg, my point is if you are simulating a system that can replicate itself, then you are simulating an accurate part of evolution. If the system has some form of externaly imposed selection, then it is a less-useful model of evolution.

However lets look at evolution in real life.

Lenski's Long Term Evolution experiment (http://en.wikipedia.org/wiki/E._coli_long-term_evolution_experiment#Results) started off with 12 sets of bacteria that were as near-identical as possible, originating from one parent bacterium. Different subsets of the resulting populations have been subjected to different conditions and the bacteria have evolved in response to their environment.

The most obvious example is the evolution of citrate metabolism in one population of e.coli.

One strain developed citrate metabolism, whilst the other 11 did not.

Maybe there was a subtle difference in the environment of this one flask?
There might have been, but the experimental setup showed that these differences were unimportant in the evolution of this trait, as were any differences in the initial populations. The only important factor seems to have been a potentiating mutation between 15,000 and 20,000 generations.


In other words the most important difference between this and the other populations was a set of chance mutations that meant that this one population had a different course of evolution compared to the other populations.

From the abstract to Lenski's paper (PDF (http://myxo.css.msu.edu/lenski/pdf/2008,%20PNAS,%20Blount%20et%20al.pdf)):

This potentiating change increased the mutation rate to Cit+ but did not cause generalized hypermutability. Thus, the evolution of this phenotype was contingent on the particular history of that population. More generally, we suggest that historical contingency is especially important when it facilitates the evolution of key innovations that are not easily evolved by gradual, cumulative selection.

More generally, we suggest that historical contingency is especially important when it facilitates the evolution of key innovations that are not easily evolved by gradual, cumulative selection.

In other words, chance is important, and this is not creationists talking, but highly respected scientists.

Maybe there could be less name-calling in this thread, and a recognition that there is a discussion about this within the field of evolutionary biology, but I'd imagine that the consensus is tending towards this view too.

[articulett]

Every organism that has evolved has the same starting point-- their common ancestor. But everything that happens after that serves to differentiate them--

Even identical twins don't inhabit the exact same space or have the exact same experiences-- this is true whether you are talking identical twin bacteria or identical twin humans. The differentiating factors are all deterministic even if we don't have all the information to predict the subsequent outcomes... they all are a continuous chain of events determined by what has come about so far. Lenski is not saying what you are saying even if you imagine he is.

Of course, this has all been said before.

You're out of your league Jim-Bob... and you can't catch up because you imagine you're winning some race but in fact, cyborg et. al. are almost a full lap ahead of you. Like Mijo, you imagine you understand what you are reading, but to anyone who can actually read a scientific paper, you are clearly not getting the same information from the paper that the majority is. Your entire point or argument rests of fuzzy semantics, insulting others, and misconstruing others so that you can pretend they agree with you or malign them for a position they do not hold.

We can all read Lenski's paper. We just don't think it supports your notion that "randomness" is the key to understanding evolution or whatever point you imagine you're making. Moreover, I don't think anyone who actually understands or teaches about natural selection would conclude that you understand it as well as you imagine you do.

Having random components does not a random process make. (Even if per your loose definition it does.) In the world where precision counts, it's just goofy and obfuscatory. Most understand evoution as a two part process-- one produces variety, and one selects from the variety and multiplies the winners.

In evolution, the genes that end up in organisms that copy them preferentially are the key to survival... whether the variety is generated "randomly" or not is NOT as relevent as how the beneficial variations are accumulated over time. Randomness is NOT essential to evolution-- variety is. We refer to the generators of that variety as "random", but we really mean unpredictable and unbiased... it happens without regard to whether it benefits the organism containing it or not. But most biologists concede that even mutation and recombination (the sources of variety) are not completely "random". Genomes themselves have evolved hot spots and highly conserved areas.

[recursive prophet]

I don't think I can comment on that in anything other than generalities.
I don't know what distinction here is being made between "stochastic" and "deterministic" models - especially if, as the above seems to apply, they are applying to equivalent models.
It seems like "deterministic" and "stochastic" are simply being used in place of "low/high problem space/complexity/number of runs".Thanks for the reply Cyborg. I will post it on RD and relay the response. Off the top of my pointy head I would say the standard definition, and think you’re onto it with scn, especially number of runs which susu mentions. But my question, to you or any reading this, regards his opinion that the stochastic models take much longer to run, and are what are mostly used in the field today in published research. Can you verify or deny that part?

Looking back over this page I reread my very sarcastic retort to you, and I’d like to apologize for that. I was reacting to your assuming from my posts I didn’t know a model from reality. Though no doubt an argument could be made that in fact I don’t, it’s hardly a feasible presupposition for any meaningful exchange of ideas, which is what I seek. I will have other questions, and look forward to your response to the above.
Cuborg, my point is if you are simulating a system that can replicate itself, then you are simulating an accurate part of evolution. If the system has some form of externaly imposed selection, then it is a less-useful model of evolution.
Interesting observation and I look forward to Cyborg's reply. I may be way off base, but I see possible Number Theory reflections here. “All consistent axiomatic formulations of number theory include undecidable propositions;” Gödel declared, and “all claims of consistency in Number Theory are in fact inconsistent.” Still, the role of science is to choose what assumptions form the working hypothesis, and this is at the core of the debate on the RD thread.
The most obvious example is the evolution of citrate metabolism in one population of e.coli.There is a discussion at least 3 pages long covering this very experiment, as I mentioned before, on RD. However, both sides claimed his experiment and other statements supported their position. IMHO the e.coli debate-like most-was won hands down by the ‘stochastics,’ though in this case it’s hard to ignore QM’s effects in this process. Eventually I’ll get around to condensing that exchange and post a link here, but if you go to one of the links I’ve already posted and go back to about p.3 where it was first introduced, you can read it now.

I emailed Lenski requesting he assign one of his grad students to participate in the discussion. After reading some links I sent him where his intention was in dispute, he made a few comments and politely declined to participate. I fear he read too far in, and came across some posts by an extremely abusive user there. This user still pops in regularly. Repeating the same mine-quote appeals to authority and spewing out ad homs and insults like some benzedrine puff adler. Assuring all who enter with comments they are wasting their, time while posting more on that same thread than any other. Very curious, and certainly not the kind of thing that would encourage a respected scientist to join in. Quite a few others, as renowned as Lenski, have posted on the ENS forum at RD, including this same thread, a year ago.

In other words, chance is important, and this is not creationists talking, but highly respected scientists.And I suspect it is becoming more prevalent, but have no proof this is the case. It’s one of the many things I hope to learn in my quest to grok the evolutionary process.
Maybe there could be less name-calling in this thread, and a recognition that there is a discussion about this within the field of evolutionary biology, but I'd imagine that the consensus is tending towards this view too.Again I agree, JB. Your final comment is why I told the Lenski story. My conjecture as to his reasons for declining to participate is unlikely, but the point I make is still valid. For the sake of our own edification we should ignore obvious trolls and just reply to those seeking actual information or to share it.

I’ve long opposed any kind of censorship online. Yet today, for the first time in years of participating in numerous different forums, I reported another poster. I’m relating this fact as one of many prevarications this individual has written about me is that I’m quick to report others. This was the first time EVER! What amazes me still is that consistently pernicious posters such as this one don’t get laughed off the boards. All the evidence you need is right here in writing! Like any absurd reference to my ‘imagined expertise’ when I have clearly stated at least 3 times I’m a complete novice. Do I need to create a sig stating this? Hmm. Not a bad idea. :rolleyes:

Hopefully the mods will be able to eliminate the need for me to make any further response to calumnies aimed at me here, and we can have an informative exchange and learn more about the battle between determinism and chaos. As by far the brightest, most knowledgeable person I know personally-along with Einstein-believe the universe is inherently deterministic, my mind is still open as to whether NS is mostly stochastic. I can say with complete certainty the argument for it not being so was completely overwhelmed on the RD thread, and hopefully at some point you and others here will have the opportunity to read it. I’m condensing the entire thread, but have just finished Part 1 and got it down from 40 to about 10 pages by snipping all ad homs and redundancy.

[cyborg]

Cuborg, my point is if you are simulating a system that can replicate itself, then you are simulating an accurate part of evolution. If the system has some form of externaly imposed selection, then it is a less-useful model of evolution.

I haven't ever argued anything different.

In other words, chance is important, and this is not creationists talking, but highly respected scientists.

Again, I haven't ever argued anything different.

See my previous shell scripted argument. I don't think I can make it any clearer than that.

[cyborg]

But my question, to you or any reading this, regards his opinion that the stochastic models take much longer to run, and are what are mostly used in the field today in published research. Can you verify or deny that part?

Not really. I would have thought a stochastic model would take longer to run purely on the basis that you would run it multiple times in order to get conclusions from it.

[recursive prophet]

Below is a the last reply by Marios on RD. I found it informative, and decided to share. I thought Cyborg and JB might find it of interest as we've been discussing genetic algorithms.

The best way to discover how genetic algorithm work is to program them yourself and see what they produce. They're really not that hard.

I'm sitting in front of a genetic algorithm right now (trying to replicate Azevedo et al (2006) - evolution of robustness stuff) - it's important to understand the difference between genetic algorithms as they are understood by Machine Learning/Industrial Optimisation bods and by evolutionary biologists. In Machine Learning people are generally given fixed, external fitness functions (this is how much aluminium costs and how much each seat lost costs - please optimise this function) - real ecological, biological considerations are kind of irrelevant. There are people who work in both machine learning and evolutionary biology (like me, grants depending), but mostly people work with one or the other - depending on whose model of how GAs work (and at what level) you'll get a very different interpretation of their advantages and limitations.

P.S. You're right, if you can, knocking out a bit of a code and seeing what happens is always better than arguing from intuition. Where I'm situated at the moment I have to read stuff by philosophers of biology, computer scientists and mathematical biologists and there's a definitely a progression in terms of clarity and generality which goes: "Natural Language arguments" << "Computational Modelling" << "Mathematical Analysis"- of course, difficulty goes the other way.

Marios

Any comments?

[cyborg]

Seems pretty congruent with what I've been saying.

[jimbob]

Cuborg, my point is if you are simulating a system that can replicate itself, then you are simulating an accurate part of evolution. If the system has some form of externaly imposed selection, then it is a less-useful model of evolution.

I haven't ever argued anything different.


Surely a corollory of this is that that self-replication is an integral part of natural selection. Indeed I have argued that in any finite system self-replication is necessary and sufficient for natural selection to occur. I seem to recall you disagreeing vehemently with this.



In other words, chance is important, and this is not creationists talking, but highly respected scientists.

Again, I haven't ever argued anything different.

See my previous shell scripted argument. I don't think I can make it any clearer than that.

I *can* read various different types of code, but I don't see what your point is.

Do you actually disagree with any of my points in my previous summation? If so where and why?

So does anyone disagree with the following statements, and if so, where and why?

"Until relatively recently in the Earth's history, there was no inevitibility in the evolution of a social animal that used complex tools requiring more than three different types of operation to make them."

"During the history of life on Earth, the course of evolution has been significantly altered by random events."

"Furthermore the fitness landscape and the consequent direction of selective pressures are sometimes (infrequently) subject to change due to random mutations that provide such a significant advantage that the entire ecosystem is disrupted. This means that the selective pressures on other organisms alter to such an extent that completely different traits begin to evolve in the descnedants of other organisms in the ecosystem in response to this disruptive mutation."

Do you understand that if an asteroid orbit is chaotic, then beyond a certain timescale, we can't know where it will be, not just because of insufficiently accurate measurement, but because this hasn't been "decided" yet?

EDIT: I think I do see your point now: deterministic systems with random inputs could reach different outputs every time they are run?

I think of evolution as slightly different, as having different levels of random feedback, so identical inputs can *still* give different outputs.

[cyborg]

Surely a corollory of this is that that self-replication is an integral part of natural selection. Indeed I have argued that in any finite system self-replication is necessary and sufficient for natural selection to occur. I seem to recall you disagreeing vehemently with this.

I don't think you ever understood what I was saying and after the amount of time I tried explaining it I don't think going over it again is going to achieve a thing.

I *can* read various different types of code, but I don't see what your point is.

It's pretty simple: why is it meaningful to say "the output is random"?

Do you actually disagree with any of my points in my previous summation? If so where and why?

No point in attempting to address them.

EDIT: I think I do see your point now: deterministic systems with random inputs could reach different outputs every time they are run?

No. Not really. The point was quite simple: it doesn't matter how "random" your input is - the same input is the same input is the same input. Using the same "random" input file on the same program won't yield different results. If it does something is wrong.

I think of evolution as slightly different, as having different levels of random feedback, so identical inputs can *still* give different outputs.

Doesn't matter in my example. The randomness in the system is coming from *somewhere*. If it's the same series of random events and you DON'T get the same output that just implies there's another layer of randomness you didn't account for somewhere.

[recursive prophet]

Cyborg: Thanks again for taking time to read my post and reply. I love congruence! To hopefully further it, I have pasted in below a recent exchange between Marios and McSwan, a new user who ostensibly works with software codes for these models. It seems consistent with some things you've been saying, but as misstating another's position seems to be a primary source of dispute on both RD and JR boards, I wanted to be sure. I would also welcome input from others, agree or disagree.

"deterministic" technically doesn't exist in nature due to the chaos theory, ie even the smallest change can change the results of a bigger experiment. (- eg a butterfly wing flap causes a cyclone, we can't measure to infinity, therefore the parts we do not measure can affect an experiment).

That's still deterministic though - nonlinearity will explode any stochastic effects (something which several of the posters on this thread have refused to accept) - but a highly non-linear deterministic system is 'technically' deterministic.

Computer simulation and mathematics can be deterministic because everything can be known about the experiment and measured to perfection, whereas that is an impossibility in nature.

What some people (DavidMC) have been trying to argue for is a sort of highly convergent pseudo-determinism (theoretically stochastic, but on a constant, highly canalized environment).
"deterministic" technically doesn't exist in nature due to the chaos theory, ie even the smallest change can change the results of a bigger experiment. (- eg a butterfly wing flap causes a cyclone, we can't measure to infinity, therefore the parts we do not measure can affect an experiment).
]The only reason I can fathom that how you have or even consider determinism in (natural) evolution is by changing it's definition.

I don't - you'd have to take that up with the people who do. I'm happy with dinky little deterministic models so long as we're clear that their purpose is being simple and easy to grasp, not accurate modelling.
Marios

[jimbob]

Articulett, Lenski's experimental design directly addressed your statements.

I agree that for evolution to occr, all you need a source of variation and natural selction, however when you look into this, without some (limited) form of imperfection in the copying, then the evolution would be limited to optimal combinations of existing traits.

Anyway, this is not what happens in biology, where at least some mutations are random. I'd actually go further and say that without random mutations, the only way for new traits to arise would be if they arose due to directed "mutations", which ID proponents would be happy accepting, but I would not.


Surely a corollory of this is that that self-replication is an integral part of natural selection. Indeed I have argued that in any finite system self-replication is necessary and sufficient for natural selection to occur. I seem to recall you disagreeing vehemently with this.

I don't think you ever understood what I was saying and after the amount of time I tried explaining it I don't think going over it again is going to achieve a thing.


As you seem to be keener on playing the points game, I'll remind you that you were the person with the problems with logic on this subject:

Apart from being unclear by trying (and failing) to use equations correctly:

This is the pure logic of what you are saying:

S = Self-Replication
P = Perfect copying
E = Evolution

This is what you are telling me:

S and P entails not E
S and not P entails E
not S and P entails not E
not S and not P entails not E

Therefore:

not E is entailed by (S and P) or (not S and P) or (not S and not P)
not E is entailed by (S and P) or (not S or (P and not P))
not E is entailed by (S and P) or (not S)
not E is entailed by P or (S or not S)
not E is entailed by P
E is entailed by not P

S does not determine E.

Self-replication is irrelevant.

Or translated, you were arguing that: if imperfect self-replication is necessary and sufficient for evolution; then logically self-replication is not needed.

I have no idea why you have this dogmatic objection to the statement that imperfect self-replication is necessary and sufficient for evolution.

You need a source of variation and you need natural selection.

The imperfect copying (aka replication) is needed for a single initial variant to produce different offspring. This is the source of variation.

This is what happens in asexual reproduction, and it is still present in sexual reproduction, which evolved later. With sexual reproduction, and a population that has combinations of differnet traits, perfect self-replication could get to the optimal combination of existing traits, but no new traits could evolve.

You also need natural selection.

If a system does not reproduce itself, then something else must do the copying, and must select the variants to be copied. The selection process is then imposed externally, and the selection criteria are arbitarily chosen. This is what happens when one uses evolutionary algorithms to solve engineering problems, the selection is artificial.



I *can* read various different types of code, but I don't see what your point is.

It's pretty simple: why is it meaningful to say "the output is random"?

Do you actually disagree with any of my points in my previous summation? If so where and why?

No point in attempting to address them.

I am not sure whether you and articulett disagree with my statements.

If you aren't playing semantic games, your debating style is pretty eccentric.



EDIT: I think I do see your point now: deterministic systems with random inputs could reach different outputs every time they are run?

No. Not really. The point was quite simple: it doesn't matter how "random" your input is - the same input is the same input is the same input. Using the same "random" input file on the same program won't yield different results. If it does something is wrong.


But that is what happens in evolution, as shown in the duplicated runs in the long-term evolution experiment.

Nothing is wrong, it is just that any model that presupposes that this shouldn't happen is wrong.

The computer programme is obviously not a good model of evolution.


I think of evolution as slightly different, as having different levels of random feedback, so identical inputs can *still* give different outputs.

Doesn't matter in my example. The randomness in the system is coming from *somewhere*. If it's the same series of random events and you DON'T get the same output that just implies there's another layer of randomness you didn't account for somewhere.

Bingo: the feedback is sometimes random. That is my point.


You are not talking about an evolutionary system.

Lenski's experimental design showed that identical starting conditions, and identical external environments could lead to non-identical outcomes, because the system acts upon itself, with random elements.

It took tens of thousands of generations for a distruptive* internally-generated event to occur, which is why I state that over moderate timescales and large populations, the random alterations in the selective landscape can usualy be considered unimportant.

This was an internally-generated event, which shows that the random events are not *just* asteroid strikes etc, but also significant mutations.



*Where a disruptive event is one that changes the selective landscape so much that the nature of selective pressures are altered..

[cyborg]

If a system does not reproduce itself, then something else must do the copying, and must select the variants to be copied.

No. This is not the way things must be. The "something else" doing the copying does not have to be in any way active about what it copies.

Why do you think it must be the case otherwise? From a purely mechanical perspective you simply can't imagine a copying machine that entities can freely walk in and out of but it itself can't move at all? In what sense would that device be choosing what gets copied?

But that is what happens in evolution, as shown in the duplicated runs in the long-term evolution experiment.

*Sigh*

Do you really think that when trying to perform the same experiment physically that you realistically have control over every aspect of the physics of the situation?

No? You don't? Then why do you think that you are performing the same run of evolution when it's *clearly* different and therefore NOT the same run?

Remember: the computer model being created is something over which there is complete control and which making the same run in every exacting detail every time is precisely what the computer does.

The computer programme is obviously not a good model of evolution.

What you are saying is that no program is a good model of evolution because it behaves like a computer program.

This was an internally-generated event, which shows that the random events are not *just* asteroid strikes etc, but also significant mutations.

Right, so, again:

Doesn't matter in my example. The randomness in the system is coming from *somewhere*. If it's the same series of random events and you DON'T get the same output that just implies there's another layer of randomness you didn't account for somewhere.

Clearly what it is that causes *any* particular mutation to occur is covered by the above.

So irrespective of what computer model you chose to model this and what method you chose to generate a mutation if you reproduce the same scenario such that the same mutation reoccurs at the same time with the same population profile you would expect the same outcome to occur right?

Otherwise you're not saying that the mutation wasn't important at all and in fact that the impact of that mutation was arbitrary - but it had to occur to have its arbitrary effect.

There is a difference between saying that an event occurred that had a causative effect but its occurrence was arbitrary and that an event occurred which had an arbitrary effect.

Clearly in the first case in the model I can simply pick and choose when these events occur and I'd expect them to have a particular impact depending on when I do so.

In the second case when I choose to make the event occur isn't really important just whether or not the system chooses to act in a particular way to it or not.

Still, in both cases, I can have full control over the randomness of the choice and can always choose to reproduce the exact same set of random events again.

[jimbob]

If a system does not reproduce itself, then something else must do the copying, and must select the variants to be copied.

No. This is not the way things must be. The "something else" doing the copying does not have to be in any way active about what it copies.

Why do you think it must be the case otherwise? From a purely mechanical perspective you simply can't imagine a copying machine that entities can freely walk in and out of but it itself can't move at all? In what sense would that device be choosing what gets copied?

The entities themselves are choosing themselves. Only those entities that make it to the copying system would actually get copied. This self-replication is analogous to that which viruses do, where only those that make it into a compatible host-"cell" actually have any chance of reproducing.

Assuming finite resources, and assuming that on average a reproducing entity reproduces more than one entity (i.e. population increase due to reproduction), then there will be a limit to the number of entities that could reproduce. This leads to competeition and natural selection.

But that is what happens in evolution, as shown in the duplicated runs in the long-term evolution experiment.

*Sigh*

Do you really think that when trying to perform the same experiment physically that you realistically have control over every aspect of the physics of the situation?

No? You don't? Then why do you think that you are performing the same run of evolution when it's *clearly* different and therefore NOT the same run?



No, but the experimental design showed that an environmental differences were not the cause of the differnet results. The externally alpplied environmental differences were unimportant as far as the differences in the outcome.

How do we know that? Because there were frozen samples every 500 generations, and when the experiment was rerun, only those samples from one population after 15,000 to 20,000 generations evolved citrate metabolism (with increasing frequency towards later generations).

This shows that an important difference occured between 15,000 and 20,000 generations in one poppulation.

Should some subtle externally-applied environmental factor have been important, then this pattern would have been different.

Different outcomes are possible even between evolutionary systems where initial and external inputs have no important differences (i.e. differences that would, of themselves, cause different outcomes).





Remember: the computer model being created is something over which there is complete control and which making the same run in every exacting detail every time is precisely what the computer does.

The computer programme is obviously not a good model of evolution.

What you are saying is that no program is a good model of evolution because it behaves like a computer program.


No, I am saying that evolution doesn't behave like a computer programme. This is different. You can model it, but the "universe" of the simulation is predetermined by the inputs, and identical inputs would have the same outputs.

You would need to have a computer model that had random (not pseudorandom) internal-feedback, which would mean that identical starting conditions with identical external inputs need not produce identical outputs.



This was an internally-generated event, which shows that the random events are not *just* asteroid strikes etc, but also significant mutations.

Right, so, again:

Doesn't matter in my example. The randomness in the system is coming from *somewhere*. If it's the same series of random events and you DON'T get the same output that just implies there's another layer of randomness you didn't account for somewhere.

Clearly what it is that causes *any* particular mutation to occur is covered by the above.

So irrespective of what computer model you chose to model this and what method you chose to generate a mutation if you reproduce the same scenario such that the same mutation reoccurs at the same time with the same population profile you would expect the same outcome to occur right?


I would if I thought that there was no randomness in natural selection. However I have always said that natural selection is probabilistic in its nature, and that (by Malthusian reasoning) it is obvious that the vast majority of organisms do not reproduce. This means that on average there is little chance of any particular organism reproducing. Given that most organisms tend to resemble their parent (or parents), even organisms with average fitness are unlikely to reproduce. Any particular mutation, even one that would improve the chances of reproduction, is unlikely to spread beyond its first generation, unless the population is growing or it is very beneficial.


Otherwise you're not saying that the mutation wasn't important at all and in fact that the impact of that mutation was arbitrary - but it had to occur to have its arbitrary effect.

There is a difference between saying that an event occurred that had a causative effect but its occurrence was arbitrary and that an event occurred which had an arbitrary effect.

Clearly in the first case in the model I can simply pick and choose when these events occur and I'd expect them to have a particular impact depending on when I do so.

In the second case when I choose to make the event occur isn't really important just whether or not the system chooses to act in a particular way to it or not.

Still, in both cases, I can have full control over the randomness of the choice and can always choose to reproduce the exact same set of random events again.

And this is where we get back to the "rerunning the tape of evolution" question.

I agree that if identical events happened, then identical events would have occured. That is tautological.

The evidence from Lenski's experiment suggests that if you were able to create multiple universes, and ensure that they were identical at one time (say just after the KT impact), then you would not expect to see identical evolutionary outcomes 65-million years late, because the systems could have almost certainly would been significantly affected by different random events (that had not been predetermined at the time of the KT impact).


If the system is such that identical starting conditions and external inputs can give differnt outcomes, then that system is random. I am arguing that evolution is such a system.

[articulett]

You can argue that all you want... no one else thinks it's clear to describe evolution as random.

Elections have identical starting points and different outcomes per your definition... so do smoke detectors... it doesn't mean it's useful or informative to call such things random. You CAN, but don't expect to converse with those who find that bizarrely vague and misleading. Lenski isn't calling evolution random-- you are using his information and your vague definition to conclude that he is.

It's exactly what creationists do. You may as well ignore the selection process completely. You've given a definition which makes evolution indistinguishable from a lottery when it's more akin to a cumulative process of elimination rounds.

[jimbob]

Articulett, I *don't* think it is "clear" to simply describe evolution as random.

However it is less unclear than describing it as nonrandom, because to a numerate person, this is obviously wrong.

Do you actually disagree with me when I say that chance events havce had an important effect on the course of evolution?

What about when I say that chance events were the reason that one population of e.coli evolved citrate metabolism whilst this did not occur in eleven other populations with "sufficiently-similar" environments to this population.

[cyborg]

The entities themselves are choosing themselves. Only those entities that make it to the copying system would actually get copied. This self-replication is analogous to that which viruses do, where only those that make it into a compatible host-"cell" actually have any chance of reproducing.

Right - but last time I went over this you wanted to argue that it was somehow important that the factory was integrated into the product when clearly that is not the case.

Should some subtle externally-applied environmental factor have been important, then this pattern would have been different.

IOW the level of control in the experiment is limited. You don't seem to appreciate the difference is that the internal factors that you can't control mean that the experiment is different every time despite the best efforts otherwise.

You would need to have a computer model that had random (not pseudorandom) internal-feedback, which would mean that identical starting conditions with identical external inputs need not produce identical outputs.

As I said previously I make no assumption whatsoever about the model because it really doesn't matter. At whatever point in the model you choose to invoke the need for a random data point that point is a point at which you can apply the technique I outlined above to reproduce the same effect each and every time regardless of how "random" you try to make it.

I would if I thought that there was no randomness in natural selection. However I have always said that natural selection is probabilistic in its nature, and that (by Malthusian reasoning) it is obvious that the vast majority of organisms do not reproduce. This means that on average there is little chance of any particular organism reproducing.

Right - you apply the reasoning to the populations and not the individuals and the fate of the individual is not important per se.

However particular individuals will be selected for particular reasons even if those reasons have arbitrary variables. If those arbitrary variables are the same you would expect the same individuals to be selected and hence the situation to play out in the same way.

As I have said before it is important to consider the level at which you are analyzing the system.

I agree that if identical events happened, then identical events would have occured. That is tautological.

That at least you can agree with.

The evidence from Lenski's experiment suggests that if you were able to create multiple universes, and ensure that they were identical at one time (say just after the KT impact), then you would not expect to see identical evolutionary outcomes 65-million years late, because the systems could have almost certainly would been significantly affected by different random events (that had not been predetermined at the time of the KT impact).

In other words if there is some variables you don't have control over that can affect the outcome then you can't fully control the outcome and hence it can - will - be different.

When you can control these variables then the outcome is the same everytime - as you agree above.

In a computer simulation you can control these variables. In reality it may not be possible to do so.

If the system is such that identical starting conditions and external inputs can give differnt outcomes, then that system is random. I am arguing that evolution is such a system.

I could equally argue the same for the computer I am using right here.

We don't because the subtle differences in the underlying physical implementation of the abstract computational model isn't enough to cause any processing errors. Nonetheless I cannot make the computer produce this text in precisely the same physical way again as I cannot actually produce the same physical outcome even if the abstract outcome I am interested in is identical.

It is in fact a physical system specifically designed to reduce the amount of errors arising from random physical effects that would cause it to deviate from the theoretical ideal to near zero as possible.

[jimbob]

I suspect a semantic difference, or possibly a philosophical one.

When you can control these variables then the outcome is the same everytime - as you agree above.

But you can't control some of the variables, because the process is affected by random factors, like radioactive decay.

You can't control when a nucleus decays, and it seems very likely that the moment of decay is not predetermined. It happens randomly. If the universe doesn't "know" when something will happen, only a supernatural (i.e. fictional) ability could know in advance when this will occur.

ETA: Yur statement about computer screens is why I use the word "significant". If you are interested in different paths of electrons, and photons, then the differences would be significant. If not then they aren't.

You could also look at what the system does. Evolutin produces variations in organisms, differences in organisms are sigfnificant in what it "does". A computer programme produces a set of (digital) outputs for a set of (digital) inputs. The digital outputs are the same (Ones and zeros, even if the actual "voltages" are different.

[articulett]

Articulett, I *don't* think it is "clear" to simply describe evolution as random.

However it is less unclear than describing it as nonrandom, because to a numerate person, this is obviously wrong.

Do you actually disagree with me when I say that chance events havce had an important effect on the course of evolution?

What about when I say that chance events were the reason that one population of e.coli evolved citrate metabolism whilst this did not occur in eleven other populations with "sufficiently-similar" environments to this population.

Mutations were random, selection was not. Yes, chance events play a role in evolution, but things do not evolve randomly... they evolve by a slow, cumulative, and exponential mutliplication of beneficial alleles.

Just as Presidents are not elected randomly, neither do species evolve randomly. Lotteries select winners at random, chess tournaments do not. All the "random events" in a chess tournament do not make chess a random game or a random process--though per your definition it would.

It makes more sense to describe things accurately. Nonrandom is more accurate than random, that is for sure. But most people describe the process in detail if they want to convey information. They don't worry about what singular adjective to stick on it. Most people would say if there is any part that is not random, then the process is not fully random. However you guys say, anything that has any randomness IS random. That just makes EVERYTHING random and the term useless. That's why no one takes such people seriously.

Some people seem obsessed with defining evolution as random This is particularly odd given that creationists tend to be the primary promoter of this canard and all biologists will argue that such a position is misleading-- wrong-- a straw man. They will agree with Miller-- not with you. They will agree with Dawkins not Mijo. They will agree with me and Talk Origins and Doctor Adequate and Cyborg and Jerry Coyne and Francisco Ayala-- not the self appointed experts who continue to imagine it's valid or informative to call it random--people for whom no amount evidence will change their mind.... people who hear others saying things they aren't saying... who imagine Lenski or any one else would agree with their conclusions about the way to describe evolution.

Trees start from a single seed... the same source, the same environment, but the branches don't grow at random... there's an algorhithm... and algorithms are deterministic. Google does not return results randomly. Having random components does not a random process make. If per your definition it does, then everything is "random" and the term is completely useless for conveying information--and so becomes a word used primarily by woo so they can imagine they are conveying information when they aren't really saying much at all.

[articulett]

And yes... it's semantics... it's a philosophical go nowhere question and conversation... as Dr Adequate pointed out at the beginning of this thread.

But why in the world would people be bent on describing something in a way that is indistinguishable from known dishonest woo like Behe? That's what makes me curious. Is your goal to clarify understanding, or prove that your way of saying something is meaningful and clear and useful to anyone other than those who "need" evolution to be "random" whatever vague meaning they attach to that?

[cyborg]

But you can't control some of the variables, because the process is affected by random factors, like radioactive decay.

Right.

And you can't tell how every particle in the air is going to behave from one moment to the next but you can still make a statement about its pressure.

You can't tell how every particle of smoke in the air is going to behave from one moment to the next but you can still make a statement about the operation of a smoke alarm.

Your argument is that because random factors can have a significant impact that the randomness in and of itself is important rather than the consequence of that random event. My argument is that it is really quite irrelevant how mathematically random anything is at all because it is not the randomness you're trying to analyse - just how a particular system behaves in given circumstances.

You don't really care that you don't know how each particular air molecule is behaving because you're only interested in the aggregate behaviour leading to the phenomena of pressure.

You don't really care how every particle of smoke in the air is behaving or how precisely the radioactive material in the detector is behaving because you're only interested in the aggregate effect leading to the activation of the alarm.

Similarly in evolution it is not really interesting as to how it is any particular shift in form arises. It is the aggregate behaviour of populations that is being analysed.

[jimbob]

Right.

And you can't tell how every particle in the air is going to behave from one moment to the next but you can still make a statement about its pressure.

You can't tell how every particle of smoke in the air is going to behave from one moment to the next but you can still make a statement about the operation of a smoke alarm.

Your argument is that because random factors can have a significant impact that the randomness in and of itself is important rather than the consequence of that random event. My argument is that it is really quite irrelevant how mathematically random anything is at all because it is not the randomness you're trying to analyse - just how a particular system behaves in given circumstances.

You don't really care that you don't know how each particular air molecule is behaving because you're only interested in the aggregate behaviour leading to the phenomena of pressure.

You don't really care how every particle of smoke in the air is behaving or how precisely the radioactive material in the detector is behaving because you're only interested in the aggregate effect leading to the activation of the alarm.

Similarly in evolution it is not really interesting as to how it is any particular shift in form arises. It is the aggregate behaviour of populations that is being analysed.

I'd agree if the we weren't discussing a highly nonlinear system.


You can make a statement about air pressure, but if weather is governed by chaotic factors, then you do need to care about every individual molecule.

There are systems where differences in individual components can be aggregated together (say calculating air pressure), and there are systems where differnces in individual components multiply and increase over time.

The Long Term Evoluiton experiment has shown that the differences multiply, indeed that significant differences can arise even where any initial differences were unimportant (with regards to the different outcomes).

[jimbob]

I am not interested in discussing how evolution*could* work, but how it *does* work in biological systems.


But why in the world would people be bent on describing something in a way that is indistinguishable from known dishonest woo like Behe? That's what makes me curious. Is your goal to clarify understanding, or prove that your way of saying something is meaningful and clear and useful to anyone other than those who "need" evolution to be "random" whatever vague meaning they attach to that?

I agree that you can ignore randomness over "moderate" timescales, but you can't over "long" timescales.

There is a big difference between what Behe says, and what I am saying.

Behe accepts "microevolution" works, he then argues from incredulity that somehow "macroevolution" couldn't work to produce "irreducibly complex" structures, giving the (amusing (http://scienceblogs.com/pharyngula/2007/06/now_behe_is_thrown_to_the_wolv.php)) example of the bacterial flagellum of something that couldn't evolve.

The randomness is not an important part of his argument, because he has neglected the law of large numbers. He has also neglected the weak anthropic principle, in that *we* perceive something special about our existance, and that we only do this because we exist. The evolution of humanity was an unlikely outcome, but then, so was any other possible outcome.

[cyborg]

You can make a statement about air pressure, but if weather is governed by chaotic factors, then you do need to care about every individual molecule.

So why isn't anyone going around saying "weather is random"?

There are systems where differences in individual components can be aggregated together (say calculating air pressure), and there are systems where differnces in individual components multiply and increase over time.

Which says nothing about the determinism of the system.

The Long Term Evoluiton experiment has shown that the differences multiply, indeed that significant differences can arise even where any initial differences were unimportant (with regards to the different outcomes).

So in weather we do need to care about every single molecule but in evolution we can stop caring and declare this detail unimportant and chalk up our ignorance as "random" and be done with it?

[recursive prophet]

I found the similarities with some of the recent points jimbob has made with ones presented on the RD thread quite relevant to that discussion, so I copied and posted them there and invited commentary. If I get some I will relay it here. Below is a link to that post. Those interested might also scroll up to the 2 previos replies by susu.exp. In the one just above mine he again presents his left/right-cat/mouse model to show how selection becomes intertwined with different predator/prey species while remaining inherently stochastic.

Interesting debate, jimbob and Cyborg. Do carry on.
http://richarddawkins.net/forum/viewtopic.php?p=1671730#p1671730

[Edit]As I just remembered susu's re-post of the model linked above had no explanation, I pasted in a brief one below. It involves analysis of cats that have a hypothetical 'turn right/left' allele when in pursuit of a mouse, and the relationship with mice possessing a right/left escape allele. When I come across the link where he lays it all out I'll post it here if requested.

"This is the cat and mouse phasespace, given as frequencies of the "turn left" alleles. It´s colored according to the final state a deterministic chaos model produces. In a stochastic chaos model, each innitial state has no certain end state, but probabilities for winding up in a number of final states. Most importantly: The blue (0.5;0.5) state vanishes, as it is only a matter of time, before the stochastic variation leads to the state of the system falling outside of the blue area and after fixation has occurred for both cats and mice, no further change happens."

[sol invictus]

Can any of the randomians in this thread formulate an experiment or result which would

a) falsify the claim that "evolution is stochastic" (or "random", take your pick), or

b) confirm it

?

If not, the debate is not scientific - it's purely semantic, and really, really, really boring (akin to arguing whether that furry thing over there is a "cat" or a "chat").

[recursive prophet]

Sol-Your reply will be posted on RD while both Marios and susu are about. My own response would be that if you stretch the definition enough, anything outside the realm of mathematics can be classified as semantics. Perhaps you agree with Marios that going beyond the maths into the effable inevitably becomes at least philosophical? Is not much of science based on unproven hypotheses on some level? I would invite you to join in the discussion on RD, and will post any replies to your brief comment there. Hopefully you’ll find the replies below of interest.

@Cyborg: Below are the replies from Marios and susu to some of your comments, posted this morning. Look forward to your take on all they’ve laid out. Very thorough lads. Marios is working on his doctorate at Bath, and susu is at Bonn, both in evolutionary biology. There are at least 4 others doing post-grad work, and there is a high level of agreement between them. Yet, I’d be happy to entertain the possibility they have all got it wrong and your take is the true path, or I just didn't comprehend your intent. We shall see, maybe.


And you can't tell how every particle in the air is going to behave from one moment to the next but you can still make a statement about its pressure.

Because the ideal gas laws are linearisable. Somewhere online there ought to be the A level derivation for the ideal gas laws, but if you want the short answer it's that stochastic deviations in linear systems cause a linear deviation in some final state. I.e. stochastic deviation e -> final deviation k.e. In a number of linear systems you can then 'cancel out' symmetric noise.

Unfortunately most real-world/interesting systems are non-linear - either you use limited and difficult mathematical techniques designed to deal with non-linear systems (the hard part of a serious maths degree) - or try to find some way to justify an approximation to a linear system.

Almost all the examples of mathematics that a non-mathematician will see are linear - but almost all the real-world/serious research systems that a mathematician will have to model are non-linear.

Your argument is that because random factors can have a significant impact that the randomness in and of itself is important rather than the consequence of that random event. My argument is that it is really quite irrelevant how mathematically random anything is at all because it is not the randomness you're trying to analyse - just how a particular system behaves in given circumstances.

In very simplistic systems, you can do that. In real systems, you either deal with nightmarish mathematics or you have to try to justify why your real-world system can be approximated by a set of very crude assumptions. Applied mathematics is almost a department all on it's own, even before you start looking at the issues raised within any given scientific discipline. Last year one of my assignments was modelling a dialysis machine - the model we used was unbelievably crude and unrealistic - and yet it still wasn't (and isn't) possible to go straight to a deterministic solution without making some disgusting simplifications (effectively assuming the fluid in one side isn't moving because the other side is moving faster). To go beyond that we literally had to just simulate it (which is art form all in itself). And that's just water with solute moving past water without.

Before anyone tries to say "Yeah, but that's just MATHS, not SCIENCE" - maths is the language of science. It's how we communicate models to each other without any ambiguity. If you have a *strong intuition* that you can't explain in maths but nonetheless want to try to communicate in words, then you're engaging in philosophical conjecture.

Marios
Your argument is that because random factors can have a significant impact that the randomness in and of itself is important rather than the consequence of that random event. My argument is that it is really quite irrelevant how mathematically random anything is at all because it is not the randomness you're trying to analyse - just how a particular system behaves in given circumstances.

You don't really care that you don't know how each particular air molecule is behaving because you're only interested in the aggregate behaviour leading to the phenomena of pressure.

Such aggregates are still random variables, unless you´ve got an infinite number of components. Throw a lot of fair coins and the frequency of heads is a stochastic variable (following a binomial distribution with parameters 0.5 and N). Only in infinity do you get a non-stocahstic variable: 0.5. Apart from that you´ve got an expected value of 0.5 and a SD of 0.5*N-1/2.

More crucially: The law of large numbers only applies to random variables.

You can make a statement about air pressure, but if weather is governed by chaotic factors then you do need to care about every individual molecule.

There are systems where differences in individual components can be aggregated together (say calculating air pressure), and there are systems where differences in individual components multiply and increase over time.

There are systems where you can - and generally must - do both. You can still use aggregates, the difference non-linearity makes is that random variables in your equations do not all vanish if you move to infinity (as noted for evolution in my reply to David, your drift term may go to 0 on its own, but that does not in general hold for functions of the drift term). In other words: The two points here are not mutually exclusive, in fact selection is always an agglomerate process, but in cases where you´ve got non-linearity it is also stochastic.

So why isn't anyone going around saying "weather is random"?

Meteorologists are actually saying that weather is stochastic:
http://www.nonlin-processes-geophys.net/15/761/2008/npg-15-761-2008.html (I haven´t had time to digest it fully, but it may help with the cat-mouse problem).

So in weather we do need to care about every single molecule but in evolution we can stop caring and declare this detail unimportant and chalk up our ignorance as "random" and be done with it?

Cannard. The key point is that we have to take into account that the behaviour of individual components is stochastic, and that this stochasticity does generally not get lost in non-linear systems. The idea that this leads to "being done with it", is probably the worst bit. No it doesn´t. Stochastic explanations are not simple, they tend to be far more complicated than deterministic ones.

Here´s a simple deterministic process:
In each step add 1.
You get 1;2;3...
Now for a simple stochastic one:
In each step throw a fair coin and add 1 if it is heads and 0 if it is tails (numbers in brackets: probabilities):
You get 0(.5),1(.5);0(.25),1(.5),2(.25);0(.125),1(.375),2( .375),3(.125)...
The number of possible pathways is growing exponentially with each step rather than staying constant (and 1) in the deterministic system. I can tell you the outcome for step 487 in the determinitic one (it´s 487), but my calculator quits on me when I try to figure out the probabilities for that step in the stochastic case (it´s 2-487*487!/((487-n)!n!) with n being the value we want the probability for).

Now imagine doing this for a process where even deterministic approximations are hazardous to mathematicians (and in the realm of non-linearity there are quite a few things people can mull over for a decade or two). That´s what I mean when I say "naive and insulting", its people who take down a molehill with a Bucket-wheel excavator and look over to those trying to get at Mt. Everest with a toothpick and having the nerve to say "You guys are just using the toothpick stuff as an excuse not to do the real work".

As Marios has pointed out, if we can in any way justify it, we´ll try to linearize our system and to find a deterministic approximation (in the above analogy, we try to shrink Mt.Everest to molehill size and build a Bucket-wheel excavator from toothpicks). It makes it easier (though both shrinking the mountain and building the Bucket-wheel excavator are still pretty hard work).

[recursive prophet]

@sol invictus: Below is the response of susu and Marios to your questions. I look forward to reading your reaction, one that will hopefully address all the key points made.
The link is a good place to jump in if you’d care to see all the current debate on this at RD. Seems arguing at cross purposes is an inherent flaw in cyber-communication.

@Cyborg: Still looking forward to reading your thoughts on the responses I posted earlier from RD. This topic has gotten fairly technical there, and is driven by a cadre of those who have argued non-determinism from early on, quite successfully. It appears that in this far more generalized forum-as opposed to one specifically dedicated to evolution and natural selection-there isn’t much interest in what apparently seems to many a pretty esoteric debate. I can only hope this hybridization project I’ve initiated won’t discourage jimbob and Cyborg from continuing their thought provoking exchange. Que sera.

https://richarddawkins.net/forum/viewtopic.php?f=4&t=66951&sid=2d74ce0122e385a818c0f2611d668240&start=475

Can any of the randomians in this thread formulate an experiment or result which would

a) falsify the claim that "evolution is stochastic" (or "random", take your pick), or

Rabbits in the precambrian. There simply isn´t a non-stochastic theory of evolution, so the only way to falsify stochasticity of evolution would be to falsify evolution.

b) confirm it

Well, I´m pretty sure I can falsify any deterministic model, through an example where it breaks down. SCA being maintained above selection-mutation balance for instance. It´s confirmed even through the simple fact that twins do not always have the same number of offspring.

If not, the debate is not scientific - it's purely semantic, and really, really, really boring (akin to arguing whether that furry thing over there is a "cat" or a "chat").

Well, the ball really is on the non-stochastic side. If I´m asked what the stochastic theory of evolution looks like, I can simply point to the textbooks and 150 years of scientific papers. The other side has nothing of that sort, so far I haven´t seen a single model from that side. I´ve seen assertions, but no models. Since the term "stochastic" has a hard definition and the theory of evolution fits that definition, I don´t think I´m arguing semantics. But people have certainly claimed that the LLN makes it non-stochastic and they are indeed leading a boring and useless semantic game (because the LLN only works for stochastic systems).
Easy to see how you got confused, but the replies you quoted me on were all from Cyborg.

Sorry - it was an off the cuff reply.

I may even add in the classic Lars reply re: the role of maths. Remember that?

I don't - reminder?

Can not the philosophical argument often trump the mathematical one on a purely pragmatic level?

No, I don't think so - it doesn't get any more pragmatic than maths. To argue that philosophy gives us pragmatic answers in science is like saying we can know scientific truths without any discourse - which violates the structure of modern science (you can't just _know_ a truth, you have to communicate your observation to others and convince them - the most appropriate language for that is the least ambiguous language - maths).

After all, nothing as exactly precise as that which can be defined mathematically is to be found in nature, right?

But it's not about nature - there are no druids in science, no one gains insights directly from nature, everything you do begins and ends with human language - nature's only involvement is in the experimental process which attempts to falsify one hypotheses in favour of another. There may well be things in nature which cannot be spoken of - not in ambiguous natural language or in unambiguous mathematics - but that's science, that's philosophy or theology. If you have a truth in mind which you cannot communicate except as a gut instinct, then you're in the realm of philosophy, not science.

a) falsify the claim that "evolution is stochastic" (or "random", take your pick), or

There are several different arguments all helpfully overloaded on one another.

One of the arguments is one of basic comprehension (i.e. you can argue if you want that everything is *secretly deterministic* - but can you _show_ it?) - that's not falsifiable (but then it isn't interesting).

There's a direct deductive argument about whether or not the model of evolution as it stands can be integrated over stochastic fluctuations so that they drop out of the model - they can't, that's not experimental question anymore than asking whether the differential of x^2 is 3.x - it just isn't.

The only sensible empirical argument is Simon Conway Morris' (unfortunately rejected by the people arguing for determinism/pseudo-determinism) - evolution works as per standard textbooks, but convergence is highly prevalent because morphospace is highly canalized - it's generally rejected as ridiculously vague and hard to falsify, but it is theoretically falsifiable.

If not, the debate is not scientific - it's purely semantic, and really, really, really boring (akin to arguing whether that furry thing over there is a "cat" or a "chat").

Yes, that's why we're not actually arguing about that. We're focusing on the slightly more interesting "But don't stochastic perturbations just smooth out?" "But won't the model always stabilise around _the optimum_?" arguments. They are wrong, but it's a wrongness that many people who think they support evolution don't grasp - you can show someone the worked answer, but you can't make them read it.

Marios

[jimbob]

Can any of the randomians in this thread formulate an experiment or result which would

a) falsify the claim that "evolution is stochastic" (or "random", take your pick), or

b) confirm it

?

If not, the debate is not scientific - it's purely semantic, and really, really, really boring (akin to arguing whether that furry thing over there is a "cat" or a "chat").

How about:


Taking an e.coli bacterium, allowing it to divide for a very few genrations then splitting the resulting population into twelve ancestral strains (all with only a few generations to limit the number of mutations from the parent bacterium) and then allowing them to grow in environments that are as similar as possible, whilst freeze-drying samples from each population every 500-generations (or 75 days), so you can attempt to "rerun the tape of evolution" if something interesting happens.

And then leave running for at least twenty years.... (Since 1988)


Does this sound familiar?

Admittedly this couldn't really "falsify" the claim that evolution is random, but it could falsify the claim that evolution is nonrandom.

For this to occur, you would need to see one popluation spontaneously evolve a trait that is not seen in the other twelve populations (or even better that has not been seen in any strains of e.coli anywhere else, where the lack of this trait has been used as one way of identifying colonies of e.coli)...

PDF by Lenski: here (http://myxo.css.msu.edu/lenski/pdf/2004,%20Plant%20Breeding%20Reviews,%20Lenski.pdf)


Second, this experiment seeks to examine the repeatability of evolution by having 12 replicate populations, all of them founded from the same ancestor and maintained in the same environment. Which phenotypic and genomic aspects of evolutionary change are repeatable in this system, and which are haphazard? How can we understand the causes of parallelism and divergence of replicate lines? The issue of the predictability of evolutionary change - or lack thereof - has long been of interest. The question was well captured by the late palaeontologist Steven Jay Gould (1989) in a thought experiment: "I call this experiment 'replaying life's tape.' you press the rewind button and, making sure you thoroughly erase everything that actually happened, go back to any time and place in the past -say to the seas of the Burgess Shale. Then let the tape run again and see if the repetition looks at all like the original..." Gould went on to say, however, that "The bad news is that we can't possibly perform the experiment." Of course, we could never run an experiment on the vast temporal and spatial scales imagined by Gould. But on much smaller scales, our experiment with E.coli allows us to address the same question. We do so by allowing 12 scenarios to play out simultaneously, rather than sequentially, as the notion of replay implies, but the issue of repeatability is fundamentally the same.

Lenski's experiment is aimed at addressing this very issue, and has answered it.

In one population, a mutation arose that changed the course of evolution in that population, and this *is* repeatable in samples taken after the mutation occured, but is not repeatable from samples taken before the mutation occured.

This shows that it wasn't minor seemingly-insignificant differences in the environmental conditions of the flasks, but was due to an internally-generated change in one of the flasks (a fortuitous mutation). Environmental differences between the flasks were insignificant. As were the genetics of the initial populations. The important factor was chance.

Those might be my words, but the above quote from Lenski shows that this question was what the experiment was aimed at addressing.