What evidence is there for evolution being non-random?

[Meadmaker]

Because creationists insist on saying "Evolution is random, so God musta done it."

But it's the second part of that statement that is incorrect. I'm baffled as to why people spend so much time attacking the first.

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

~~ Paul

To the extent that I have any agenda at all, I've been quite up front about it. I'm curious about what underlies the aversion to "randomness" or "chance" in descriptions of evolution. I think there's more to it than your, and others', protestations about "random" being oversimplified and/or misleading. I think you are aware that any brief description of evolution, or any other complex phenomenon, will necessarily be inaccurate, just because it can't be described accurately in a small space.

I think the best expression of what I think underlies that aversion was probably in this post:

http://www.internationalskeptics.com/forums/showpost.php?p=2604758&postcount=145

 

[Dancing David]

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.

Show to me that it is determined.

Remember a trait may be neutral or mildly detrimental (in that it is not detrimental to succesful reproduction to the point that is extinguished) and only becomes beneficial when the organism finds itself in an enviroment where the trait is beneficial to reproductive success at a rate greater than the trait being detrimental.



How is that deterministic?

How would you model it?

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.[/QUOTE]

 

[Dancing David]

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.



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.

That is great. Given the fact that we are talking about random players in a random enviroment, why is that a usefull distinction to make. the theory of natural selection occurs in relation to the historic past and the current present.

The theory of natural slection is not about a limited set of players in a closed system with a number of known variables. You may use that to model limited sets of expression of traits of a limited numebr of players.


But what does that have to do with the expression of the theory of natural slecetion in real enviroments?


You have stated that it is a non-random and determined selection process.

Show me how that is true and not just an assertion on your part?

 

[Dancing David]

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.

I don't believe that is what the random players have stated, i am stating that the interaction of depermined players in a small complex ssystem is random.

You can not predict the collison of four balls in a box, you can not determine the position of the balls unless you put them on tracks and govern thier rate of motion. The interaction of all the balls is completely determined.

But the interaction of all the balls is not.


When you can predict the given expression of a trait in a given population in an uncontrolled enviroment then you can say you have a determined system.

 

[Vorticity]

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

~~ Paul

Now Paul, that's unfair. We're not all T'ai Chis.

I've spent a lot of time in these threads harping on "evolution is random" (before I gave up), and I know for sure that I don't have a hidden agenda. It's just that for those who study stochastic processes and the like, it's always a bit of a "wince moment" whenever we hear someone insist that "evolution is not random", when it very obviously is, by the technical definition of the term. I fully understand that this can cause confusion in the general public, because the common definition and the technical definition are not the same. But that does not suddenly make it non-random.

 

[Dancing David]

Taffer:

I apologise for all my extended responses.

The statement about the four balls in the box is the meaningful argument.

The motion and elasticity of the balls can be detrmined, but the system is not deterministic.

The interaction of determined individuals is not always deterministic.

I have not satted the ID argument that "all evolution is random", it is constrained variability in a complex interaction that creates randomness.

 

[Paul C. Anagnostopoulos]

I've spent a lot of time in these threads harping on "evolution is random" (before I gave up), and I know for sure that I don't have a hidden agenda. It's just that for those who study stochastic processes and the like, it's always a bit of a "wince moment" whenever we hear someone insist that "evolution is not random", when it very obviously is, by the technical definition of the term. I fully understand that this can cause confusion in the general public, because the common definition and the technical definition are not the same. But that does not suddenly make it non-random.

Yeah, but notice how you gave up. Meadmaker hasn't.

To the extent that I have any agenda at all, I've been quite up front about it. I'm curious about what underlies the aversion to "randomness" or "chance" in descriptions of evolution. I think there's more to it than your, and others', protestations about "random" being oversimplified and/or misleading. I think you are aware that any brief description of evolution, or any other complex phenomenon, will necessarily be inaccurate, just because it can't be described accurately in a small space.

Indeed, which is why the terse statement "evolution is random" is about as misleading as you can get.

So I'll make a deal with you. If you explain your reason for harping on this, I'll explain mine. Here's mine:

Simply calling evolution random, with no further explanation, leads many people to believe there has to be another source for "intelligent" evolution, since they think we clearly aren't random.

~~ Paul

 

[CFLarsen]

Simply calling evolution random, with no further explanation, leads many people to believe there has to be another source for "intelligent" evolution, since they think we clearly aren't random.

Hear, hear.

 

[Meadmaker]

Indeed, which is why the terse statement "evolution is random" is about as misleading as you can get.

Personally, I think it's less misleading than "evolution is not random".

So I'll make a deal with you. If you explain your reason for harping on this,

I think there's a psychologically interesting element about some people's aversion to describing evolution as "random" or "by chance", that goes deeper than any fear of inaccuracies in descrption. In my last post, I linked to an earlier post in which I gave two different descriptions of evolution. Factually, they were identical (or I tried to make them so), but they had different psychological impacts. I'd like to hear your comments on that post.

 

[joobz]

Indeed, which is why the terse statement "evolution is random" is about as misleading as you can get.

Personally, I think it's less misleading than "evolution is not random".

Then can we agree to start using the more accurate "evolution is not purely random" description.

That may help to avoid Paul (and my) aversion to this evolution is random statement, since it exists a colloquial meaning of being totally random without method.

And as has been mentioned, "evolution is not random" reads to mean that there is NO underlying process that is not deterministic.

 

[aggle-rithm]

But it's the second part of that statement that is incorrect. I'm baffled as to why people spend so much time attacking the first.

So you'll waste all your time and energy on insignificant questions and never get around to the big ones. Conspiracy theorists often use the same tactics.

 

[mijopaalmc]

Simply calling evolution random, with no further explanation, leads many people to believe there has to be another source for "intelligent" evolution, since they think we clearly aren't random.

So you're saying that we should describe evolution not just inaccurately, but contrafactually, because an accurate and factual description will confuse people?

Note this is not the same as the simplification that takes place in General Chemistry and General Physics courses, where rules that are described as true all the time are in fact true only some of the time (e.g., Newtonian mechanics as a simplification and good approximation of relativistic mechanics at non-relativistic speeds). It is an outright fabrication seemingly to protect against a straw man argument of people who don't and more often that not don't want to understand.

Probably the easiest to explain the way the stochastic nature of evolution seems deterministic over evolutionary time is to say, in the general case, over any finite but arbitrarily large number of trials, you (in the impersonal sense) is relatively unlikely to any one result, even the most common one. You are however almost guaranteed of getting a small range of values relative to the range of all possible values. Furthermore, as the range of all possible values increases, the range of "guaranteed" value decreases in relative size. Therefore, as the number of trials goes to infinity, the range of "guaranteed" values converges on one single value, and that is why, over an infinite number of trials, there is a given probability of event happening, while, over a finite number of trials this probability seems to fluctuate.

For instance if your were to toss a fair coin 10,000 times the probability of getting exactly 5,000 heads is .007979, but you can be 95% certain that you will between 4902 and 5098 heads a relative error of 1.96%. Similarly, if you toss a fair coin 1,000,000 times the probability of getting exactly 500,000 heads is .0007979, but you can be 95% certain that you will between 499,020 and 500,980 heads a relative error of .196%. Thus, even though it is not proven here and I myself don't know the exact form of the rigorous mathematical proof, it is reasonable to say that as the number of trial goes to infinity, the probability of getting a head approaches .5, despite the fact that for an arbitrarily large but finite number of trials the probability of getting half head and half tails approaches zero. I will leave others to refine this argument, because in its current form, I don't know how to reselve the apparent contradiction.

The point of that rather lengthy explanation, is that, even though there is random mutation in evolution and natural selection is a stochastic filter, many iterations of these processes cause a convergence in the mean fitness of a population.

 

[drkitten]

For instance if your were to toss a fair coin 10,000 times the probability of getting exactly 5,000 heads is .007979, but you can be 95% certain that you will between 4902 and 5098 heads a relative error of 1.96%. Similarly, if you toss a fair coin 1,000,000 times the probability of getting exactly 500,000 heads is .0007979, but you can be 95% certain that you will between 499,020 and 500,980 heads a relative error of .196%. Thus, even though it is not proven here and I myself don't know the exact form of the rigorous mathematical proof, it is reasonable to say that as the number of trial goes to infinity, the probability of getting a head approaches .5, despite the fact that for an arbitrarily large but finite number of trials the probability of getting half head and half tails approaches zero. I will leave others to refine this argument, because in its current form, I don't know how to reselve the apparent contradiction.

Er, there is no contradiction, apparent or otherwise, to resolve. The amount of variation around the actual probability of 0.5 gets smaller and smaller until it disappears under the noise.

The point of that rather lengthy explanation, is that, even though there is random mutation in evolution and natural selection is a stochastic filter, many iterations of these processes cause a convergence in the mean fitness of a population.

... and that's why evolution isn't "random."

For approximately the same reason that casino profits aren't generally considered "random." When you walk into a casino, you may be gambling. The casino is not -- it knows that it's going to make money today, and even how much. It just doesn't know from whom.

 

[Dancing David]

Then can we agree to start using the more accurate "evolution is not purely random" description.

That may help to avoid Paul (and my) aversion to this evolution is random statement, since it exists a colloquial meaning of being totally random without method.

And as has been mentioned, "evolution is not random" reads to mean that there is NO underlying process that is not deterministic.

Yeah, okay, but I have a huge issue with this.

There are two kinds of determinism

1. The scientific notion that there are effects that can causaly related to other events (although I disagree with the root of this as well, in a system of sifficient variables you may not be able to determine the causal factors beyond a correlation.)

2. The philosophical notion of determinism, which is akin to the watchmaker's universe. The universe is set at a given point and can be determined to where it will be to a certain degreee of accuracy. Given the nature of deviation in timekeeping of the watch, it can be predicted as to where within a certain percentage the hands of the clock will be. So too there are foolish people who claim that given the data of a set of interactions at a given moment in time they can predict with accuracy as to the following conditions at a later point in time.


I find major problems with both uses of the word determined in the context of the theory of natural selection.

In case one, when people say things like "Humans evolved to have intelligence becasue of the obvious benefits." We can discuss possible scenraios for the development of the larger brain and complex structure of the brain, but it is difficult to do more than speculate.

So there are causal relationships to the process by which an idividual organsim will have reproductive success. But after the fact to say "This is what determined the path of natural selection for this organism." is va very difficult proposition.


I apologise because it is the philosophical use of the word determinism that drives me crazy. We can have causal factors in a system, but that does not mean you can determine the later state of the system.

This is the part of natural selection that further adds to the 'random' nature of evolution. An organism and the alles are blind to what enviroments they mat encounter and the changes that will occur in an enviroment.

So yes there are causal relationships that lead to the process of natural selection of reproductive success, which is why I use the term constrained variability. A pig(bearing alleles) is not going to suddenly levitate and have new benefits from levitation, there is constraint in the expression of the genome through morphology. Nor is a pig going to be able to know that it and it's progeny will suddely find the Lost Mine of Twinkies and need to develop the alleles for sugar digestion and better dentition.

I refer to the process as random, because of the expression of traits and survival in unknown enviroments, but that doesn't mean that pigs can fly and prognosticate the location of the Lost Twinkie Mine.

 

[Soapy Sam]

Nice straw man. The internet is a created non-biological entity and therefore not at all an analog to an evolved (and evolving) biological entity. It is interesting that this thread seems mainly to be motivated by the desire to combat the creationist straw man that order cannot arise from disorder, or as they put it, "evolution cannot happen because, as evolutionists say, it is random" yet the "non-randomites", who, I might add, have not presented a cogent argument in support of their position, resort to straw men such as the above to trivialize the argument that "probabilistic", "random", and "stochastic" are proper descriptors of evolution.

Well spotted.

What is your explanation for the origin of the internet then?

 

[mijopaalmc]

Well spotted.

What is your explanation for the origin of the internet then?

Do we really have to go over this again?

The internet is not analogous to biological life forms because it is a created entity. Furthermore, this is akin to the creationist straw man "evolution is an incomplete theory because it doesn't explain abiogenesis". Evolution is a theory describing how life developed after it came into existence and therefore doesn't to need to explain how it came into existence. Similarly, just because my description of evolution doesn't necessarily describe how the internet could come into existence (ignoring for a minute that the internet is a created entity and therefore not analogous to biological life forms) doesn't mean it isn't an invalid description as it needn't describe that event just as evolution needn't describe abiogenesis. In other words, I don't have explain how the internet came into existence for my description to be valid, sound, or true and therefore won't except to say that the internet was created by humans and the history is available online.

Honestly, I would have thought that, for people who complain so much about how creationists frequently set up straw men and claim that evolution is disproven when they demolish them, the "non-randomites" would be extra sensitive when they are using the same techniques themselves.

 

[Vintij]

A theory in science is often accepted as fact because it can be tested and goes through alot to become a theory in the first place. To say that evolution is just a theory, is incorrect, it is a fact.

The theory of evolution goes back even further than darwin, though it was darwin who predicted the mechanism of evolution which to this day is still a vital part of it. But not the main aspect. See now that technology has allowed us to study genes, we have discoverd things the darwin could not observe, like genetic drift and phenotypes. So not only was darwins theory the most logical at the time and accepted as fact, the "theory" has become even more of a fact with new discoverys. So if anything, evolution is a fact, the mechanism of evolution, (generally believed to be natural selection) is a theory, as well as modern synthesis (the new neo-darwin mechanism). But evolution in general, IS a fact and can not be disproven. It is the mechanism of evolution or (the cause) that we are trying to determine at the present time.

Anyone who does not believe in evolution, obviously has no interest in knowledge and likes to believe in the fantasy that earth and humans were created in 7 days. And that dinosaurs existed only a few thousand years ago.

 

[Meadmaker]

A theory in science is often accepted as fact ...And that dinosaurs existed only a few thousand years ago.

Welcome to the board. Certainly, most people here share your views. In fact, I'm willing to assert that every single participant in this particular thread shares your views.

 

[Vintij]

Welcome to the board. Certainly, most people here share your views. In fact, I'm willing to assert that every single participant in this particular thread shares your views.

Thank you, I find it very rewarding to be in the presence of fellow thinkers. Its hard to have a decent conversation with my friends, which is why....Ill assume, that everyone is on this board because it is an outlet for there own ideas and theorys to be heard, rather than brushed off as boring science talk to the majority of non-thinkers in the world.

 

[Myriad]

There's another reason to object to the overarching statement "evolution is random," besides the misrepresentation of the import of that statement by Creationists. It really can be a barrier to understanding evolution.

The movement of molecules in a gas is random. Simple fact. And there are many phenomena one can examine in which understanding that fact is paramount to explaining the observed behavior. Brownian motion, for instance, or explaining the functioning of a diffusion pump.

However, there are many other phenomena for which considering the random motion of gas molecules is not at all helpful. Predicting the weather, for example. Attempting to calculate or model anything by the position and motion of individual air molecules isn't very helpful. The important variables in, say, the formation of a tornado aren't the random positions and motions of individual molecules, they're pressure and wind veolocity, properties that are statistical aggregates of the positions and motions of vast numbers of air molecules.

So, if the question is, "how does a tornado form?" an answer that begins, "when you have a some air molecules moving randomly..." is not going to be a good answer. It's the wrong level of description for the phenomenon you're interested in.

Similarly, if the key question someone has about evolution is, as it so often is, "how could evolution have generated humans starting only with microbes?" then an answer that begins "when you have some genomes mutating randomly..." is not a good answer. It's the wrong level of description for the phenomenon you're interested in.

If your starting point for understanding evolution is "it's a random process," it announces right off the bat that you're going to be looking at the process at the wrong level of description to explain the "big picture" aspects effectively, such as how microbes led to Cambrian fauna or humans separated from apes. (Likewise if your starting point for understanding air is "gas molecules moving randomly" it means that you're at the wrong level of description to explain effectively how a sailboat works.) It's not the technical accuracy that's at issue, it's the practicality. You could describe and model the airfoil effect of a sailboat sail based on analysis of individual randomly moving air molecules, and you could explain evolutionary history based on random individual genetic events, but that explanation is going to be very hard to understand. Dawkins wrote a whole book (Climbing Mount Improbable) in order to put across such an explanation.

Just as we have pressure and temperature and wind veolocity to describe the behavior of masses of air, and these aren't random at the level of many of the phenomena we're interested in (such as sailboats or tornados), we have allele frequency and reproduction rate and selection pressure to describe the behavior of populations of genomes, and these also aren't random at the level of many of the phenomena we're interested in, such as species divergence and development of complexity.

The trap is describing evolutionary change in terms of what happens to individuals, when it's counterproductive to do so. And when you're focusing on the random (individual) events, you're already close to that trap. "So you see, the mommy bird that lays the weaker-shelled eggs has fewer baby birds hatch, and the one that has the random mutation to lay stronger-shelled eggs has more babies hatch." But (thinks the student/audience), what if the strong-shelled bird's nest gets hit with a random bolt of lightning? Then all the stronger eggs would die and evolution doesn't work! So aren't you just making up stories to make it look like it works, when it really depends on luck? Not only that, that story clashes with our favorite Western myths, which say the individual with the disadvantage will find a way to overcome that disadvantage and win.

What makes this difficult, of course, is that just as understanding some phenomena involving gases does require treating gas as individual randomly moving molecules, understanding some evolutionary phenomena does require dealing with singular genetic events such as emergence of new alleles via point mutation, crossover mutation, and similar events. So a thorough understanding does require appreciation of the aspects of the process that are random. I just don't think that's a good starting point. Most people have an intuitive understanding, at least, of pressure and temperature before learning the thermodynamic descriptions of those phenomena; they have little reason to doubt that those phenomena exist just because they're statistical aggregates of individually random events. The same isn't true of the role of random variation in evolution. It's a mistake, albeit a common one, to emphasize that aspect first and foremost.

Respectfully,
Myriad