What evidence is there for evolution being non-random?

Meadmaker said:
I had never noticed this phrasing before in this thread, but this is an excellent way of saying it.
Click to expand...
I am sure this phrasing will really nail it next time a creationist asserts that it is laughable that random mutations have led to the emergence of man.
 
steenkh said:
I am sure this phrasing will really nail it next time a creationist asserts that it is laughable that random mutations have led to the emergence of man.
Click to expand...

Not likely, but it happens to be correct, nonetheless.
 
Articulett, do you agree with my statement that the selection pressures are nonrandom, but the selection is probabilistic.
 
jimbob said:
Articulett, do you agree with my statement that the selection pressures are nonrandom, but the selection is probabilistic.
Click to expand...

I know that you directed your question towards articulett, but when I see a sentence-long summation of a topic as complex as natural selection, I generally like to clarify.

Are you saying that what the selection pressure are (e.g., change in food supply or predation) are not random but how the act at the level of individual replicators to effect a change in allele frequencies is random in so far as when two individuals with identical collections of alleles are subjected to the same selective pressures one may survive and reproduce while the other my perish?
 
mijopaalmc said:
I know that you directed your question towards articulett, but when I see a sentence-long summation of a topic as complex as natural selection, I generally like to clarify.

Are you saying that what the selection pressure are (e.g., change in food supply or predation) are not random but how the act at the level of individual replicators to effect a change in allele frequencies is random in so far as when two individuals with identical collections of alleles are subjected to the same selective pressures one may survive and reproduce while the other my perish?
Click to expand...


Essentially, yes. each generation of an organism in a stable environment will be more optimised for that environment, but there are many ways and how this optimisation is performed is another matter.

A "fitter" organism is one with an increased chance of breeding, that is all I mean by "fitter".

An example:
Whether a peppered moth is eaten by a bird depends on whether it is seen by a bird. Better or worse camoflague might affect the probability of being seen, reducing the distance where there is a 70% chance of the bird seeing it, say, but whetehr the bird appears within this visible distance is essentially random, in that it has nothing to do with any trait that the moth has, given that it is in the same environment as potential predators. It can alter its behaviour to escape from some predators, but this will put it within the range of others.

This is for an individual moth. If you have many moths and many birds, one can begin to treat them as populatitions. One can then analyse differences in traits in the moths and see how these correlate with reproductive success.

In principle, a mutation arises in an individual, and this modulates the individuals reproductive chances. The probability of the mutation survivintg depends on the population and the reproductive chances of that population.

In a stable population I am assuming that there is on average one reproducing offspring per parent. A hypothetical mutation increases the reproductive success by 10%. If the average brood is ten per parent, with nine dying, then this mutation in one child will only have a 110%times10% chance of reproducing. If the mutated offspring survives to breed, then there is a better than 50% chance of the mutation surviving in at least one child.

I am assuming that this survival is modelled by a poission distribution, thus lambda will now be 1.1 as opposed to 1, for the non-mutated organism.

If the mutation survives for ten generations, then the mean breeding population is now 1.110 which is 2.5.., as opposed to 1 for the non-mutated offspring.

The more generations the mutation survives for, the more likely it is to survive further...

The chance of the mutation surviving will be the integral of the probable populations over the generations. And I can't be bothered to do the maths for that...

Of course, it gets easier in an oscillating population, for example lemmings. If the mutation occurs during a crash, it is very unlikly to survive, it it occurs during a boom, it is.

​
.

In a stable environment, the unpredictible aspects like predation and weather can still work as a selection pressure, if the events are frequent enough; they will then act to cause optimisation to that pressure.

However on geological timescales, the environment and thus the selection pressures change, and with mass extinctions, so many species are removed that many ecological niches are left unoccupied. Which niche is occupied first might preclude or open up other niches, and this, with the small initial populations would depend heavily on initial random starting conditions.

It is therefore quite likely that in such a stable environment, a mutation that affects the chances of reproduction is more likely to move away from "the optimum" than towards "the optimum". Thus most mutations that have an effect will be deleterious.

However, if the environment changes (e.g. the arrival of a new potential food supply like nylon) then the parent organisms are further from the "new optimum", and mutations are more likely to be beneficial.

New species are more likely to emerge to emerge when the ecosystem changes, and the old isn't so well adapted.

What is nonrandom about natural selection, it that over the generations there will be optimisation to the environment.
Click to expand...
 
jimbob said:
Articulett, do you agree with my statement that the selection pressures are nonrandom, but the selection is probabilistic.
Click to expand...

I said this before... it's not incorrect, but it's just confusing and imprecise as many have said. Moreover, Mijo thinks of probabilistic as a definition of "random". Other scientists don't. I just don't think you understand natural selection or can convey it very well although you think you do and can. I think I'll stick with the many experts who have weighted in on this thread and in peer reviewed papers. I'll stick with the experts who have conveyed the understanding to me and to many--people like Dawkins, Ridley, Jones, Ayala, Pigliucci, Berekely, Talk Origins, etc.

They all consider the environment the de-randomizer--the elimination rounds successive generations carrying the information (subject to random changes) must pass through. You don't call the elimination rounds random. Football outcomes are not random because of snow. Snow is considered part of the game... so are injuries... the environment is part of the elimination round so-to-speak. You "randomites" just don't seem to get this--ever. Your explanations do not show how winning combinations get multiplied exponentially (as in the butterfly article). You aren't able to convey how this gives the appearance of design...so it's a loser of an explanation though there doesn't seem to be any way to convince you of this.

If it was useful, it would be used. The best ways to explain information will evolve based on our incoming knowledge, creationist obfuscation, what works the best, and what conveys the most important part of the process--namely, natural selection. The randomness is the easy stuff. But even you don't seem to get natural selection. And if you do, you can't convey it simply like Cyborg, Foster Zygote, Dawkins, Ayala, Talk Origins, Paul A., Dr. Adequate, etc. Maybe if you watched the Dawkins video I linked or read the Selfish Gene you would understand and be able to understand why your definition is almost as vague and wishy washy and uninformative as Mijo's... But I don't think you want to know that. I think you both (and Meadmaker) want to believe you understand natural selection and believe you know what the creationist obfuscation techniques are and believe you can pick out creationists, but you are wrong. And I don't think any evidence will be enough to prove it to you.. Dawkins is clear. You 3 are not.
 
Last edited:
It's not jimbob who doesn't understand natural selection, aritculett; it is you who doesn't understand it. We all know that the scientists you site say that evolution is non-random, but the also say the adaptive variation increase the probability of their possessors' survival and reproduction. I know that is hard for you to understand, but such a statement is a probabilistic statement and in so far as "random" means "[o]f or relating to a situation characterized by a probability distribution or random variables", these respected scientists are saying something quite nonsensical and self-contradictory.

What is so hard for you to understand about that?

If you do respond, could yo please refrain from telling me that my definition "makes everything 'random'"?

It doesn't, and you just display your ignorance of probability theory and statistics when you say so.
 
Jimbob, as I said earlier, you cannot think of selection like a line-up of phenotypes where all those meeting the genetic criterion move forward.

You are still stuck on the idea that because organism A and B have the same phenotype but one survives and the other does not that this entails the selection of whether or not A or B was chosen is random. It does not. It entails that your view of how selection works is too simplistic.
 
The hyperfocus on "randomness" makes you guys miss the forest for the trees. Really. Google results appear "designed". There's an algorithm that involves randomness. But you can't understand the "design" just by focusing on the random aspects. The same goes for evolution. What brings the order?--The appearance of design? Natural selection with the best replicators from the many experiments exponentially multiplying their winning combinations. It really isn't so different from Google results.

Selection produces the order and begets more of the selected.
 
Last edited:
What do you mean by "hyperfocus"?

Where is my reasoning incorrect in post #2325?

Once a mutation arises in a population, the best way of calculating how likely it is to sread, is probabilistically.

http://users.rcn.com/jkimball.ma.ultranet/BiologyPages/P/Populations2.html

If a population cycle can rise or fall a hundredfold then even if a beneficial mutation has arisen and the organism has bread, it is still unlikely to spread

An organism has a mutation that increases its chances of having breeding offspring tenfold. tis organism breeds, at the peak of the population (when the population is momenterily flat. It has ten offspring that breed, unlike the average, which is one.

These offspring are in the trough year, and the average number of offspring per organism is now 0.01. Each of these ten ofspring have a ten times better chance of breeding (i.e. 10%). In one generation the chance of this mutation not surviving is 0.9^10 (34%)

Where is the error in this reasoning?
 
Jimbob... the question was about the non-random aspects of evolution.. not the best way to calculate the likelihood of anything spreading. All you can see is the probability aspect... not how the order we perceive comes about. Your sentence about population cycles makes no sense. Did you watch the Dawkins clip I referenced earlier? I know you think you are making sense... but you're not really saying anything. Answer this question-- how does the order arise? How is it that all the butterflies now have the mutation mentioned in the MSNBC link? Why are there so many more male butterflies now? That is NOT random. That is the answer to Mijo's question. That is not "probabilistic"-- either... I mean, you could use that word, but it doesn't explain anything...

The error in reasoning is that you have no ability to describe the answer to the question cohesively. You are not conveying how one creatures or multiple creatures can drive the evolution of each other. You can use your definition to say that a foot ball team one a game probabilistically. I mean, I guess it's true--ish-- but it doesn't mean anything. It conveys nothing. You and Mijo and Meadmaker are not conveying a cohesive definition. Dawkins, Cyborg, Ayala, and most of the people who have dropped by this thread (with the exception of creationists) have. Have you read a review of Behe's book? Do you know what's wrong or bizarre about his reaosoning? You guys have the same problem with yours. You are just so muddled sounding but so sure that you are saying something informative. If it was useful to say it that way, I think people would be using it. But, from what I can see, you may as well say that a football teams winning was random or probabilistic or that poker is a random game or probabilistic game. Normal people might say these are aspects of the game, but it is inadequate for explaining the observed results.

Remember the part cyborg wrote about a priori reasoning. You could use probabilities to determine outcomes of a football season (and bookies do)-- but if you wanted to describe what DID happen after the fact, the probabilities are no longer part of the story... you are MISSING that part of the explanation in regards to what evolution is and why things look "designed". I can't tell if it's because you don't understand it or if it's just because you have a need to describe things in regards to probability the way Mijo needs to related evolution to being "random".
 
Last edited:
You're wrong articulett.

If you can assign a probability to the outcome of a football game then the outcome is decided by a coin flip - as we all well know.

All that stuff that happens in between the start and end of the game? Inconsequential. You're missing the purity of the randomness there.
 
It is interesting articulett that you continue to insist that other posters are inadequate of explain themselves and yet you are the one who continues to make errors in your explanations. As cyborg points out you can certainly use probability to determine the likely outcome of a football game or a hand of cards in poker. As far as evolution is concerned natural selection sees to it that the success of a species is not purely random. While probability does play a huge role in evolution it does not mean that evolution is somehow simply a random outcome. As with most scientific theories one cannot simple break down evolution to a paragraph or two of description. You can only look at the basic tenets of evolution to see its implications and meanings. If you have the time and dedication to become an expert you might be able understand the intricacy of evolution, in this case dealing with probability and randomness, and how all the details and processes of evolution fit into a basic set of tenets. In short it takes years to truly understand a scientific theory in all of its glory and details. The nature of the questions and observations on this thread goes way beyond the casual interest in evolution. So it fits that details of how probability and randomness truly do work in evolution is missing, but we can fall back on evolutions basic tenets, one part selection pressure, one part randomness as such through genetic drift, but I have a feeling that will not be enough for most of you. :cool:
 
Dumbledore said:
It is interesting articulett that you continue to insist that other posters are inadequate of explain themselves and yet you are the one who continues to make errors in your explanations. As cyborg points out you can certainly use probability to determine the likely outcome of a football game or a hand of cards in poker.
Click to expand...

Er... articulett knows this as well as I do. The point I've been making is that just because you model something with probabilities doesn't mean the thing you're modelling is actually acausal.

So it fits that details of how probability and randomness truly do work in evolution is missing, but we can fall back on evolutions basic tenets, one part selection pressure, one part randomness as such through genetic drift, but I have a feeling that will not be enough for most of you. :cool:
Click to expand...

I already explained the most salient parts.

Non-deterministic mutation with respect to the genome.

Deterministic selection with respect to the genome.

It is in fact irrelevant to the evolutionary algorithm as to whether or not there is any true acausality in the non-deterministic aspects - which is why genetic algorithms can work perfectly well despite the fact that computers, by definition, cannot be random.
 
Dumbledore said:
It is interesting articulett that you continue to insist that other posters are inadequate of explain themselves and yet you are the one who continues to make errors in your explanations. As cyborg points out you can certainly use probability to determine the likely outcome of a football game or a hand of cards in poker. As far as evolution is concerned natural selection sees to it that the success of a species is not purely random. While probability does play a huge role in evolution it does not mean that evolution is somehow simply a random outcome. As with most scientific theories one cannot simple break down evolution to a paragraph or two of description. You can only look at the basic tenets of evolution to see its implications and meanings. If you have the time and dedication to become an expert you might be able understand the intricacy of evolution, in this case dealing with probability and randomness, and how all the details and processes of evolution fit into a basic set of tenets. In short it takes years to truly understand a scientific theory in all of its glory and details. The nature of the questions and observations on this thread goes way beyond the casual interest in evolution. So it fits that details of how probability and randomness truly do work in evolution is missing, but we can fall back on evolutions basic tenets, one part selection pressure, one part randomness as such through genetic drift, but I have a feeling that will not be enough for most of you. :cool:
Click to expand...

And you like some many before you are playing fast and loose with the definition of and usage of "random" and in turn are misrepresenting the arguments of the posters who argue that evolution is a stochastic process. Yes, evolution is "random" in the sense that is governed by probabilities of survival and reproduction (which even the scientists who vehemently deny that evolution is "random" admit), but that doesn't mean that it can't have an orderly result in the sense that organisms adapt to their environment. In fact there is a class of stochastic processes, known as Galton-Watson processes that behave just as evolution would be expected to (i.e., if the mean number of copies of a given allele per individual possessing that genes in the previous generation then the allele will eventually die out). This doesn't mean that evolution has to be a Galton-Watson process, but it does mean that the out-of-hand dismissal of evolution being a stochastic process because stochastic processes supposedly can't give orderly results is completely unwarranted.
 
Dumbledore... try reading the thread.

Yes... the most common explanation is "random mutation coupled with natural selection"-- Scientists refer to the latter as the "non-random" aspect of evolution.

Cyborg was using humor. The problem with the explanations of the people who do not understand evolution is that they hyperfocus on the randomness and the probability and not the process that connects the events producing an "incremental" effect and the appearance of design.

The environment sifts through all the organisms created and "chooses" which copy themselves and how much. Those with the "winning genomic combinations" get the most copies of themselves and the most chances for some descendant somewhere picking up one of those rare, but beneficial, mutations that gives it's progeny an edge as the environment continually sifts.

The random part is easy. Natural Selection is a bit more difficult... the fact that the population of male butterflies has made a comeback and that they all carry the same mutation is not random or probabilistic:
http://www.msnbc.msn.com/id/19733274/

The protective mutation was random. But the evolution-- the process-- the fact that the males made a comeback and they all carry this mutation is due to the fact that only the male who had the mutation survived to produce tons of male offspring who also survived who produced... etc.

Go back and read before you jump into a thread where you are so poorly informed.
 
The concept of selection intensity divides the deaths before breeding into "fortuitous" i.e. no selection pressure on the trait under discussion and "adaptive" i.e. a selection pressure on this trait. The intensity is the ratio of "adaptive" to "fortuitous". Not my word but John Maynard Smith's.

With natural selection, the selection intensity might get as large as 10%. This means that it would take about 10 times longer for a trait to evolve than if the intensity had been 100%.

You can do the sums if you accept this treatment, without a tacit acceptance of the random or "fortuitious" nature of much "deselection", you can only use handwaving.
 
You must log in or register to reply here.

Back
Top Bottom