What evidence is there for evolution being non-random?

[Meadmaker]

There is no problem; I agree with you. But what does the definition of "by chance" have to do with "random"?

I thought they were nearly synonymous.

 

[Taffer]

Uh huh, care to cite where these predictions occur. Please tell us where the Asian Longhorn beetle will have it's next outbreal, or even better the Emerald Ash Borer, please tell us where the American Chestnut with resistance to disease will arise.

Show your work, where are these amazing predictions, what level of accuracy.


Please tell me where the rain will fall next week, that way I will know if the flea population will rise due to lack of moisture. It will also help in firefighting forest fires.

What model and what accuracy, bold claims Lone Star.

Once again, I must remind you I am not claiming the ability to predict, only that it is theoretically possible.

 

[Paul C. Anagnostopoulos]

I thought they were nearly synonymous.

I don't know, I think it means more like "purposeless." These words are so nuanced.

chance:
1 a : something that happens unpredictably without discernible human intention or observable cause b : the assumed impersonal purposeless determiner of unaccountable happenings : LUCK c : the fortuitous or incalculable element in existence : CONTINGENCY
2 : a situation favoring some purpose : OPPORTUNITY *needed a chance to relax*
3 : a fielding opportunity in baseball
4 a : the possibility of a particular outcome in an uncertain situation; also : the degree of likelihood of such an outcome *a small chance of success* b plural : the more likely indications *chances are he's already gone*

It's okay to call evolution random, without further qualification, as long as you also say that the environment is random.

~~ Paul

 

[Meadmaker]

It's okay to call evolution random, without further qualification, as long as you also say that the environment is random.

~~ Paul

Works for me.

Anyone have a problem with that?

 

[Paul C. Anagnostopoulos]

Works for me.

Anyone have a problem with that?

Well, actually, I guess you have to say that evolution is random, but in lock-step with a random environment. That makes it clear that evolution isn't random with respect to the environment.

Edited to add: Note that I have no idea whether it is reasonable to call the environment random. I'm just demanding that if you're going to call evolution random, you should mention that it is embedded in a random environment.

~~ Paul

 

[CapelDodger]

Well, actually, I guess you have to say that evolution is random, but in lock-step with a random environment. That makes it clear that evolution isn't random with respect to the environment.

Absolutely. Evolution is driven by the need to conform to an ever-changing environment.

Edited to add: Note that I have no idea whether it is reasonable to call the environment random. I'm just demanding that if you're going to call evolution random, you should mention that it is embedded in a random environment.

~~ Paul

There are aspects of the environment that are entirely uncoupled from life on earth. Orbits, plate tectonics, stuff like that. From life's point of view they're random. Some aspects of the environment are coupled to life - all that oxygen in the atmosphere, for instance. Any fool could have predicted that back when photosynthesis first kicked in, so not random .

 

[Taffer]

But while the selection may or may not be non-random, predicting which enviroment an organism will be in five years is very random.

So? Selection is non-random. That is my entire point.

 

[Taffer]

I think you'll find these cosmologists and astronomers would disagree with you.
http://www.newscientist.com/article/mg18124365.100-chaotic-heavens.html
http://www.newscientist.com/article/dn6796-moon-behaving-badly.html
http://www.newscientist.com/article...mean-for-the-future-of-the-solar-system-.html

Yeah, I've admitted my mistake and retracted my statement. Teach me to try to understand physics.

 

[Taffer]

I should say that, when I say "selection is non-random", I am specifically talking about the way selection acts on given levels of fitness of various alleles in a population. Whether or not reality follows selective pressures (we have genetic drift, after all), is beside my point. If you wish to call evolution random because it contains random elements, then that's fine. I just find it important to distinguish random mutation from environment-driven selection.

I am happy with Paul's nomination. "Evolution is random in a random environment" is perfectly fine by me. It makes it clear that evolution is driven by the environment in a non-random (i.e. determinied) way.

ETA: I should also say that my problem is not with any strict definition, but with misinterpretations which arrise because of a lack of definate definitions.

 

[mijopaalmc]

ETA: I should also say that my problem is not with any strict definition, but with misinterpretations which arrise because of a lack of definate definitions.

The question is: why should we be beholden to the misinterpretations of people who want to deliberately undermine evolution for religious and political gain when we form our interpretations?

The fact is that natural selection still only works by conferring a probability of "survival" on an individual. For example, if that probability is any higher that .2 in a subpopulation of 100, it is extremely improbable that evolution will function in any sort of deterministic way, meaning that all individuals with a given fitness compliment will "survive" while all other other of a different fitness complement won't. Only when the subpopulation is on the order of 10 individuals (i.e., between about 15 and about 85) does the probability of no individuals "surviving" become significant (i.e., have a probability that allows it to happen at least once in evolutionary time). In fact, at a probability of "survival" of .1, the probability that no individual in a subpopulation of is .348678. In other words, in a subpopulation of 10, no individuals with a probability of .1 will "survive" more than one third of the time, which is about 13127 times the the frequency of the probability that no individuals "survive" in a subpopulation of 100 with the same probability of individual "survival". This is why it is so important (or as I said, "interesting") that Dr. Adequate left out the final sentence of the MathWorld article on "stochastic" and that drkitten didn't seem to think that the last sentence of said article to be clear.

Stochastic is often used as counterpart of the word "deterministic," which means that random phenomena are not involved. Therefore, stochastic models are based on random trials, while deterministic models always produce the same output for a given starting condition.

While the wording is not the clearest due to its lack of grammatical parallelism, it distinguished between a deterministic process (i.e., a process in which the same input always has the same output) and a stochastic process (i.e., a process in which the same input has different outputs).

In short evolution is a stochastic process because individuals with the same fitness complement (i.e., same inputs) either "survive" or don't (i.e., different outputs). Now, it is up to the reader to decide whether he/she wants to exchange "random" for "stochastic" but the there is absolutely no way of getting around the stochasticity of evolution.

 

[Meadmaker]

Now, it is up to the reader to decide whether he/she wants to exchange "random" for "stochastic" but the there is absolutely no way of getting around the stochasticity of evolution.

I agree. Paul puts the randomness (I refuse to use the word stochasticity) on the environment, and that's fine with me.

The really interesting question, to me, is why anyone would want to try.

 

[mijopaalmc]

[...]I refuse to use the word stochasticity[...]

You just did.

Why not use it?

 

[Taffer]

The question is: why should we be beholden to the misinterpretations of people who want to deliberately undermine evolution for religious and political gain when we form our interpretations?

Because it does not give a complete picture. If we call evolution "random" with no other qualifiers, it doesn't not describe what processes can lead from random inputs into an ordered output.

The fact is that natural selection still only works by conferring a probability of "survival" on an individual. For example, if that probability is any higher that .2 in a subpopulation of 100, it is extremely improbable that evolution will function in any sort of deterministic way, meaning that all individuals with a given fitness compliment will "survive" while all other other of a different fitness complement won't. Only when the subpopulation is on the order of 10 individuals (i.e., between about 15 and about 85) does the probability of no individuals "surviving" become significant (i.e., have a probability that allows it to happen at least once in evolutionary time). In fact, at a probability of "survival" of .1, the probability that no individual in a subpopulation of is .348678. In other words, in a subpopulation of 10, no individuals with a probability of .1 will "survive" more than one third of the time, which is about 13127 times the the frequency of the probability that no individuals "survive" in a subpopulation of 100 with the same probability of individual "survival". This is why it is so important (or as I said, "interesting") that Dr. Adequate left out the final sentence of the MathWorld article on "stochastic" and that drkitten didn't seem to think that the last sentence of said article to be clear.



While the wording is not the clearest due to its lack of grammatical parallelism, it distinguished between a deterministic process (i.e., a process in which the same input always has the same output) and a stochastic process (i.e., a process in which the same input has different outputs).

In short evolution is a stochastic process because individuals with the same fitness complement (i.e., same inputs) either "survive" or don't (i.e., different outputs). Now, it is up to the reader to decide whether he/she wants to exchange "random" for "stochastic" but the there is absolutely no way of getting around the stochasticity of evolution.

Y'know, I'm fast loosing interest in this debate. You are right, we can call evolution what we like. Therefore, I will call it the (more accurate) non-random selection of random variation in a population. This is how population and evolutionary genetics describes it, and as that is my field of study, that is how I will describe it. If some wish to call it "random in a random environment", that is fine by me, but I suggest that the use of the word "random" can lead to misunderstandings.

 

[mijopaalmc]

Y'know, I'm fast loosing interest in this debate. You are right, we can call evolution what we like. Therefore, I will call it the (more accurate) non-random selection of random variation in a population. This is how population and evolutionary genetics describes it, and as that is my field of study, that is how I will describe it. If some wish to call it "random in a random environment", that is fine by me, but I suggest that the use of the word "random" can lead to misunderstandings.

What I object to, however, is being told that it is incorrect to describe evolution as random, which has been what a significant number of posters have been saying. It seems that you are calling evolution "non-random selection of random variation in a population" purely by convention, a convention that I realize was developed long ago, but a convention nonetheless. The problem is that, when you violate the convention, you are told flat out that evolution isn't random and called a creationist for pointing it out.

Why is it so important that evolution be called non-random?

What understanding do we gain from calling it such given the association that being non-random has (e.g., that evolution has hard and fast rules for who survives and who doesn't)?

Why is it so wrong to insist on a more rigorous use of the term, which would mean that it is correct and even preferable to refer to evolution as random?

 

[davidsmith73]

Uh huh, and you would win the Nobel Prize for chaos theory. The problem is this, how many factors effect an organism? Thousands? How do you rule out that a particular trati that is later beneficial is not already being selected for from thousands of candidates?

By tightly controlling your experimental conditions of course. The adaptive mutation experiments using bacteria seem to manage fine. Clone your bacteria (genetic factors controlled) and grow them on culture (environment easily controlled)

ps. I don't think this would have anything to do with chaos theory as we know it.

 

[Taffer]

What I object to, however, is being told that it is incorrect to describe evolution as random, which has been what a significant number of posters have been saying. It seems that you are calling evolution "non-random selection of random variation in a population" purely by convention, a convention that I realize was developed long ago, but a convention nonetheless. The problem is that, when you violate the convention, you are told flat out that evolution isn't random and called a creationist for pointing it out.

It is not "convention", it is the commonly accepted description in the field.

Why is it so important that evolution be called non-random?

Like I said, if you wish to call it "random in a random environment", then by all means go ahead. But it is incorrect to call it simply "random".

What understanding do we gain from calling it such given the association that being non-random has (e.g., that evolution has hard and fast rules for who survives and who doesn't)?

Because it is natural selection which makes evolution viable. With only random mutation, evolution would not happen nearly quickly enough to ever evolve anything. But that is not all that evolution is.

Why is it so wrong to insist on a more rigorous use of the term, which would mean that it is correct and even preferable to refer to evolution as random?

Because a) it is random in a random environment (see paul's posts), and b) it leads to misconceptions. Oh, and c) it suggests a lack of understanding of the inner workings of evolution.

 

[Paul C. Anagnostopoulos]

The really interesting question, to me, is why anyone would want to try.

Because creationists insist on saying "Evolution is random, so God musta done it." The implication is that we're here, we're not random, so we can't be a product of evolution. As you say, however, we're just as random as anything else.

Also, as has been said before, "evolution is random" is misleading to most people. Really, it is. They have no reasonable model in their heads of how evolution actually works. They have a grasp on random mutation, but no grasp of selection.

You're harping on "evolution is random" to the point where I'm convinced you have a hidden agenda.

~~ Paul

 

[Paul C. Anagnostopoulos]

Why is it so wrong to insist on a more rigorous use of the term, which would mean that it is correct and even preferable to refer to evolution as random?

Ah, since "random" is technically correct, we should use it alone, unqualified, so as to cause utter confusion? A few extra words of clarification just obfuscate the matter, do they?

You're harping on "evolution is random" to the point where I'm convinced you have a hidden agenda.

~~ Paul

 

[Dancing David]

Once again, I must remind you I am not claiming the ability to predict, only that it is theoretically possible.

And I am saying that it is not theoreticaly possible. To predict the behavior of complex systems (albeit even ones with simple rules and thousands of plares) is not possible in a deterministic way. You can use probability and low levels of accuracy.

I believe that you stated that certain things were not random because they could be modeled. I am saying different. The interaction of natural selection in the enviroments that organisms find themselves in involves hundreds if not hundreds of thousands of variable factors and millions and in case billions and trillions of plares.

How do you model something like that?

You stated that the process of evolution was non-tandom. I am saying that there is constrained variability is a system that is so large and so complex that it's interaction will be random in outcomes of where players exist and mutation interact.

A spider with a certain 'new' trait lays five thousand eggs. The hatchlings becomes air borne plankton. When you start to model : what do you model? X% will successfully ballon, Y% will ballon to and enviroment where the trait might prove beneficial, Z% of the enviroment will express the character that can be expoited by the critter, A% is the chance that the trait will carry on in the spider or it's offspring long enough for the enviroment to have the character for the trait to be beneficial, B% is the chance that the trait will not be detrimenta in the current enviroment the spider lands in, C% is the chance that another dominant trait will not mask the specific trait...... and so on.

Now when you need to model the chance that a spider with this possible beneficial trait will survive long enough and it's offspring carry the trait and the enviroment changes or has a character where the trait will be beneficial you will do what?

Model the spiders in the first generation.

Five thousand x(X% x Y% x Z% xA%.....) just to get the number of spiders that might have the trait in sn enviroment that they and thier offspring will carry the succesful trait to the next generation (in the first wave)

Then you have to do so for each of the members of the following generations and so on...


You sated that this was non-random, I know that the correct word is stochastic or possibly chaotic, how is this determined as a system. How can you predict where a player in generation N will be in enviroment G and then have a succesful radiation R of beneficial trait BT?

Can you really say that is a non-random system?

Remember an accurate model must then generate the nviroment and all it's players into the future as well.

What is the BT needs a specific metabolite to be carried in a prey species for the trait to be expressed succesfuly, then the model is dependant on all the players in the prey species as well, and if the prey species exists by eating only a single species of plant, then you have to model them as welll.... and so on.


Where is this model non-random? What level of accuracy exists within it to predict thee expression of a known and given beneficial trait. A series of probabilty expressions interactions in a matrix of other probability expressions?

 

[Dancing David]

So? Selection is non-random. That is my entire point.

The expression of traits is in an interaction of of individuals in complex enviroments, how is that not 'random'. If you can't predict individual players behaviors and the behaviors of the enviroments then you are left with what? A probability?

Explain to me how you will model

A. Traits that will be selected for.
B. Enviroments where the trait will be beneficial.
C. The rate at which a critter will find itself in enviroment B.
D. The rate of expression for given enviroment D that will express the quality B.
E. The level of the trait in given population P that happens to meet all the above criteria.
F. The level of individuals that meet all the above criteria and breed.
G.The level of survival of the progeny.

It is imporatant to remember that natural selection occurs in an aggregate of individuals and the agrregate of individuals that comprise an enviroment.

Now explain to me how selection (the pasing of beneficial traits to the future and thier expression in an enviroment) is not random?