Randomness in Evolution: Valid and Invalid Usage

[Wowbagger]

Here is a thread to discuss how the word "random" applies, or does not apply, to the Theory of Evolution, depending on how you define the word, and stuff like that.

I have started this thread, so that no others need to get derailed on this (often semantically and mathematically confusing) topic. (I hope it's not too late!)

Here is a summary of my current position, for starters:

The word "random" is often used, in different contexts, when describing evolution. But, it does not need to be used. And, its usage often contributes to confusion, which is why I like to avoid it, myself.

Although, I will admit there are both valid and invalid uses of the term.

Valid usage includes, but might not be limited to:

* Describing mutations where the word "indifferent" could be a substitute. For example, instead of saying "mutations are random to the life form's survival", you could say "mutations are unconsciously indifferent to the life form's survival". (This is more or less the context I have seen Richard Dawkins use.)

* Describing a model of evolution, where our knowledge is not perfect. Such usage implies that the Evolution is actually deterministic behind the scenes. But, information about its initial conditions are lost to time and thermodynamics; and the number of variables involved in the current conditions are too many for us to handle. Therefore, we resort to simplified models that utilize either stochastic algorithms and/or random variables. As the models improve, the less we rely on these things.

Invalid usage includes, but might not be limited to:

* "Random chance", "blind chance", "happy accident", etc. Any term that implies Evolution is all about complete and utter randomness. Evolution is an algorithm, and one that was practically inevitable to crop up somewhere in the Universe (likely more than one place). To imply evolution is random, in this manner, is to misconstrue its nature. In part, because it implies lack of predictive power.

* Appeals to quantum uncertainty. While quantum uncertainty might have some small impact on the course of evolution, it would be unfair to "out" Evolution as a theory of randomness because of this, simply because quantum uncertainty makes an impact on all of the other sciences, as well. Also, most quantum fluctuations are averaged out (or "smeared out") in large scales, anyway.

* Referring to Arthur Dent's daughter. She has nothing to do with this!

In any definition of the term, the more we study about life, the more precision we can make in our predictions, and the less randomness plays a role. If Evolution was supposed to be a theory about randomness, you would think the opposite trend would take place.

Your thoughts?

 

[Walter Wayne]

* Referring to Arthur Dent's daughter. She has nothing to do with this!

...

Your thoughts?

I'd like to point out that Arthur Dent's daughter, being responsible for the extinction of human kind and all other earth species, is by no means a bit player in evolution.

I will get to the rest of the post when I have more than just a lunch break to post, but thought that such an egregious error had to be correct as soon as possible.

Walt

 

[Belz...]

I've always thought that "random" either means we have not enough information to predict the outcome or that it implies acausal processes.

 

[Vorticity]

Sweet Jesus. Not again.

This thread will come to a bad end.

 

[shadron]

* Describing mutations where the word "indifferent" could be a substitute. For example, instead of saying "mutations are random to the life form's survival", you could say "mutations are unconsciously indifferent to the life form's survival". (This is more or less the context I have seen Richard Dawkins use.)

I'll cite http://genomebiology.com/2005/6/6/R50, which says:

Among three sources of evolutionary innovation in gene function - point mutations, gene duplications, and gene shuffling (recombination between dissimilar genes) - gene shuffling is the most potent one.

The first two are, perhaps, covered by the point above; their result is, indeed, indifferent to the survival of the individual sufferer. The third, being a naturally occurring random process, is not covered, but needs to be; it is the major "reason" for sexual reproduction being a valuable process to a genome. And it make cats interesting, as well.

 

[Taffer]

From the previous thread:

You seem to have missed the posts where cyborg has said that acausality is necessary to randomness:

If, as cyborg claims, acuasality is necessary to randomness, then randomness cannot be without acausality; therefore, one must have acausality to have randomness, acuasality defines randomness, and random means acausal.

Belz understood it. All truely random events are acausal. But for "random" to mean "acausal", all situations of acausality must be considered random. However, that does not follow from the premise "all random events are acausal".

You are trying to say:
If A then B
∴ If B then A
∴ A = B

Which is not a valid argument.

 

[Almo]

I would say that evolution as a process harnesses random events (mutations) as a means of generating variety.

 

[Wauthan]

I'd say that mutation is random within limits. If you mutate too much, or against selective pressure, you can't have kids. If you mutate too little, or get "locked inside" a specialized biological niche, you can't adapt to changes in the enviroment.

Mutation might be random but evolution follows rules. Squeeze mutation through evolution and you end up with sort-of-randomness.

 

[Belz...]

You are trying to say:
If A then B
∴ If B then A
∴ A = B

Which is not a valid argument.

It's amazing how many people make that mistake.

 

[mijopaalmc]

From the previous thread:



Belz understood it. All truely random events are acausal. But for "random" to mean "acausal", all situations of acausality must be considered random. However, that does not follow from the premise "all random events are acausal".

You are trying to say:
If A then B
∴ If B then A
∴ A = B

Which is not a valid argument.

You're still missing the point (why am I not surprised?). At no point did either cyborg or I say that acausality was the only characteristic of randomness. However, as cyborg has said several time in the other thread acausality is a necessary condition for randomness. Since randomness simply would not exist without acausality, as acausality is a necessary condition for randomness in part defines randomness, acausality in part defines randomness, and therefore, random means in part acausal.

 

[Third Eye Open]

I think a better word for the things most people call 'random' would be 'chaotic'.

 

[Taffer]

You're still missing the point (why am I not surprised?). At no point did either cyborg or I say that acausality was the only characteristic of randomness. However, as cyborg has said several time in the other thread acausality is a necessary condition for randomness. Since randomness simply would not exist without acausality, as acausality is a necessary condition for randomness in part defines randomness, acausality in part defines randomness, and therefore, random means in part acausal.

You're still missing the point (why am I not surprised?). Something doesn't mean something else unless the are equivalent. Belz understood this. Unless you are now trying to say random only partly means acausal, in which case you would only be partly correct, wouldn't you?

This is pointless anyway. Let us grant you that "randomness" equates to "acausal". What does this get us? It means nothing to evolution by natural selection - the acausal element only creates variation, variation which could arise by any means at all and evolution would still occur.

What is your point with all this?

 

[mijopaalmc]

You're still missing the point (why am I not surprised?). Something doesn't mean something else unless the are equivalent. Belz understood this. Unless you are now trying to say random only partly means acausal, in which case you would only be partly correct, wouldn't you?

I''m sorry, but are you saying that a necessary condition does not provide part of the meaning to its possessor?

This is pointless anyway. Let us grant you that "randomness" equates to "acausal". What does this get us? It means nothing to evolution by natural selection - the acausal element only creates variation, variation which could arise by any means at all and evolution would still occur.

What is your point with all this?

Randomness is not necessarily acausal. One of the line of reasoning is therefore that none of the definitions of "random" is "acausal".

 

[Taffer]

I''m sorry, but are you saying that a necessary condition does not provide part of the meaning to its possessor?

Ok, I admit I'm slightly confused. Aren't you claiming that evolution is random, and don't you agree that randomness is acausal?

Randomness is not necessarily acausal. One of the line of reasoning is therefore that none of the definitions of "random" is "acausal".

What? How does that answer my question "what does this mean for evolution?"

 

[Paul C. Anagnostopoulos]

Sweet Jesus. Not again.

No kidding.

Evolution is a stochastic process, because at least some of the mechanisms involved are stochastic (that is, random). However, it is misleading to call evolution a "random process, full stop" because selection is not random with respect to the local environment.

Note that the definition of random is "not deterministic."

~~ Paul

 

[Walter Wayne]

Here is a thread to discuss how the word "random" applies, or does not apply, to the Theory of Evolution, depending on how you define the word, and stuff like that.

...

Although, I will admit there are both valid and invalid uses of the term.

Valid usage includes, but might not be limited to:

* Describing mutations where the word "indifferent" could be a substitute. For example, instead of saying "mutations are random to the life form's survival", you could say "mutations are unconsciously indifferent to the life form's survival". (This is more or less the context I have seen Richard Dawkins use.)

"random to the life form's survival" isn't something I've ever heard someone say, and odd since non-random doesn't imply that it isn't indifferent.

* Describing a model of evolution, where our knowledge is not perfect. Such usage implies that the Evolution is actually deterministic behind the scenes. But, information about its initial conditions are lost to time and thermodynamics; and the number of variables involved in the current conditions are too many for us to handle. Therefore, we resort to simplified models that utilize either stochastic algorithms and/or random variables. As the models improve, the less we rely on these things.

I'd agree, but I'd say it can also be used to describe evolution, not just the model for reasons I'll mention later.

Invalid usage includes, but might not be limited to:

* "Random chance", "blind chance", "happy accident", etc. Any term that implies Evolution is all about complete and utter randomness. Evolution is an algorithm, and one that was practically inevitable to crop up somewhere in the Universe (likely more than one place). To imply evolution is random, in this manner, is to misconstrue its nature. In part, because it implies lack of predictive power.

Random chance seems appropriate to me. Random, as I've mention before, does not mean that all possibilities are equal, and I've manage to explain that easily to layman with the simple sum of two dice example. If one acknowledges randomness I wouldn't see an objection to happy chance or bad luck, as in "it was bad luck that apes ever stood upright."

I agree with your objection to blind chance, as it implies that it was unifluenced by surroundings (the environment).

However, while evolution is a process that crops up elsewhere, it is characteristics the constituent processes that make biological evolution what it is. A different means of descent with modification and a selection process unlike our "biosphere" might create a non-random evolution.

* Appeals to quantum uncertainty. While quantum uncertainty might have some small impact on the course of evolution, it would be unfair to "out" Evolution as a theory of randomness because of this, simply because quantum uncertainty makes an impact on all of the other sciences, as well. Also, most quantum fluctuations are averaged out (or "smeared out") in large scales, anyway.

Quantum uncertainty is the engine of mutation. And those mutations have an undeniable and significant impact on the course of biological evolution. The "smeared out" argument doesn't apply because of the vary nature of the biological system. Even if you believe a certain mutation is bound to happen, whether it is beneficial or not depends on the dynamic environment. One that is beneficial at one time, may be deleterious at another. And once a mutation becomes fixed, it influences the environment of those around it.

Biological evolution is not the nicely behaved type of process that we are generally familiar with where variations "average out in the long run".

Your thoughts?

I'd say from the technical point of view, current knowledge implies evolution is random. A particular mutation influences not only the individual it is in, but it influences the environment of individuals around them, possibly affecting there "selection process". This compounding affect will result in large variation at the macro-level rather than the "smearing-out" people are accustommed to discussing.

From the laymans point of view, this will be random in almost every sense I can think of other than that all possibilities are equal. I would go further and state, that even if our knowledge of mutation and selection changed and we found out they were determistic, evolution would still be random in the laymans sense. The innumerable variables were small details can have significant affects would make it "random".

As an example, if it turns out that fundamental particles interactions are determistic in nature, thermal noise while fundamental determistic would be random for all practical purposes.

I agree it is often best to avoid the term, but in most senses biological evolution is random.

Walt

Edited to add: Crap, more text than I intended.

 

[Walter Wayne]

On the subject of causality, people who deal with the subject of randomness do not equate it with acausal in anyway.

The simplist example off the top of my head is a random number generator. The random action of electrons creates a voltage, which is compared to 0V and a logical "1" or "0" is produced based on whether it is above or below.

The voltage, determined by electron position is said to be random. The logic 1 or 0, generated by the determistic comparator-circuit, is said to be random. This is because no amount of information about the system will tell you the state a few moments from now. The linking of random and acausal is not something that is done by those who actually study such systems.

Walt

 

[Taffer]

No kidding.

Evolution is a stochastic process, because at least some of the mechanisms involved are stochastic (that is, random). However, it is misleading to call evolution a "random process, full stop" because selection is not random with respect to the local environment.

Note that the definition of random is "not deterministic."

~~ Paul

See, I have no problem with this. It is misleading to call evolution random, full stop. As I said before, random variation could be replaced with designed variation, and evolution would still happen.

I don't quite understand what your point is, mij.

 

[mijopaalmc]

See, I have no problem with this. It is misleading to call evolution random, full stop. As I said before, random variation could be replaced with designed variation, and evolution would still happen.

I don't quite understand what your point is, mij.

Overall, I want to know how evolution by natural selection is not "random" by the definition "[o]f or relating to a type of circumstance or event that is described by a probability distribution".

In reference to what you and I have been discussing most recently, I am arguing randomness is not necessarily acausal, so evolution by natural selection's possessing a causal structure does not imply in any way that evolution by natural selection is non-random.

 

[Taffer]

Overall, I want to know how evolution by natural selection is not "random" by the definition "[o]f or relating to a type of circumstance or event that is described by a probability distribution".

In reference to what you and I have been discussing most recently, I am arguing randomness is not necessarily acausal, so evolution by natural selection's possessing a causal structure does not imply in any way that evolution by natural selection is non-random.

Ok. So evolution is random.

So what?