Randomness in Evolution: Valid and Invalid Usage

[Ichneumonwasp]

All that is required for evolution to occur is a source of change. What Darwin and those who follow tell us, is why the evolution we see looks the way it does. The answer was descent with modification (heredity with mutation) and natural selection. Sure evolution proceeds whether the change is random or not. The question is, is it? And if it is, does natural selection "suppress" the random nature, to make it as reliable as, say, gas laws?

Walt

No suppression necessary, and that seems an odd way of wording it. Gas laws describe an emergent property of the interaction of gas atoms/molecules. You cannot see 'pressure' in any one atom or in any single interaction. It is the mass interaction that allows the emergent property of 'pressure' to arise. That can be measured accurately and reliably.

The same is true for 'evolution'. The change in allele frequency over time is a function of variation and selection; there are plenty of mathematical formulas describing different systems and at least one general formula. Taffer has posted it several times. That change is an emergent property as is pressure.

 

[sol invictus]

The fact that evolution is random doesn't effect the fact that all life...

Tell us, mijo: do you still agree with that? Is evolution random?

YES or NO?

If you won't answer, you've changed your mind (which is quite OK).
If you say yes, we're going to have a great time with it (since then so are smoke detectors).
If you say no, the thread is over.

 

[Walter Wayne]

The answer:
Yes.

There is no need to "suppress" the "random nature" in the same way that there is no need for a gas to "suppress" its "random nature". Populations and species are predictable like gases because the scales at which their characteristics change is much larger and slower than that of their constituents. Most species live at least hundreds of thousands of years, most individual creatures in a species live or die based upon decisions made on the order of minutes to seconds. Most species have between millions and billions of members, at their peak, yet the random elements are individuals or even the genes of these individuals. When you look comparative numbers there is no other way to see it.

Conservatively species work on a scale that is a billion times larger than their components(a high end estimate is probably more like 10^15, true its no 10^23....but when your numbers are that big does it really matter? )

It looks like you are mistakenly applying characteristics of a large number of independent trials to a process which has heredity as one of its major components.

Sure species may peak at millions or billions of numbers. Consider that humans are above 6 billion now, but are thought to have number less than 10 thousand less than 100 thousand years ago. We suspect this because of how little genetic variation we have. What happened to that small population is inherited by us.

The article that Jimbob linked to actually gave us a nice example of randomness. When lines of bacteria were allowed to adapt to cold temperatures, two-thirds of them did worse when returned to high temperature. And one third did as well or better. In a competitive environment, the strains would have different advantages should the environment change again.

The mechanisms for short term randomness are there, and heredity will cause such changes to leave an imprint on future generations.

Walt

 

[Walter Wayne]

No suppression necessary, and that seems an odd way of wording it. Gas laws describe an emergent property of the interaction of gas atoms/molecules. You cannot see 'pressure' in any one atom or in any single interaction. It is the mass interaction that allows the emergent property of 'pressure' to arise. That can be measured accurately and reliably.

The same is true for 'evolution'. The change in allele frequency over time is a function of variation and selection; there are plenty of mathematical formulas describing different systems and at least one general formula. Taffer has posted it several times. That change is an emergent property as is pressure.

Well the use of the term suppress was probably a bad choice. What I meant was that the random properties of the actions are not reflect at the macro level.

So in gas laws, why do the underlying interactions manifest at the macro level as pressure which is vary consistent. It is not just that it is a mass interaction, it is the nature of those interactions. Mainly that collisions with the container are for the most part independent with each other. There is no heredity in the system. If heredity existed, say that molecule container collisions were mostly likely to occur where the previous collision happened, the emergent property wouldn't be a nice even reliable pressure, but a series of knocks concentrated at certain points.

The properties of molecular motion that leads to "pressure" are not the same properties of the processes that constitute evolution.

Walt

 

[Ichneumonwasp]

The properties of molecular motion that leads to "pressure" are not the same properties of the processes that constitute evolution.

Walt

Oh, yes, of course not. But that does not mean that the analogy does not hold. Both are describable at higher levels of abstraction where the random elements 'wash out' allowing the mathematical descriptions. I just think we all get lost in the details because it is easy to see molecules as random but the system as ordered, while with evolution the random elements occur at so many different levels -- at the level of DNA and the level of organisms -- that we tend to lose sight of the fact that the 'system' is 'evolution as a whole'.

Natural selection pushes the random elements of variation into a skewed alignment -- which is one reason why we tend to see (overall) the emergence of more complexity (though, not always).

 

[Walter Wayne]

Oh, yes, of course not. But that does not mean that the analogy does not hold. Both are describable at higher levels of abstraction where the random elements 'wash out' allowing the mathematical descriptions. I just think we all get lost in the details because it is easy to see molecules as random but the system as ordered, while with evolution the random elements occur at so many different levels -- at the level of DNA and the level of organisms -- that we tend to lose sight of the fact that the 'system' is 'evolution as a whole'.

Natural selection pushes the random elements of variation into a skewed alignment -- which is one reason why we tend to see (overall) the emergence of more complexity (though, not always).

When the property that leads to the predictable of pressure (memoryless system vs. a hereditary one) then the analogy most definitely doesn't hold. I mentioned to Paul previously why a large biased doesn't necessarily lead to the result you think in systems that do have memory. Even if it is unlikely that a species may "lose", it plays the survival game many times.

Walt

 

[Ichneumonwasp]

When the property that leads to the predictable of pressure (memoryless system vs. a hereditary one) then the analogy most definitely doesn't hold. I mentioned to Paul previously why a large biased doesn't necessarily lead to the result you think in systems that do have memory. Even if it is unlikely that a species may "lose", it plays the survival game many times.

Walt

But that is not the analogy. In a large container with many different species of gas molecules we do not describe where the argon is going to be in relation to the chlorine when using gas laws. The analogy concerns the fact that there will be a change in allele frequency over time (so the analogy is between the fact there there will be a change in allele frequency over time and pressure). That is what evolution is. Which alleles win and which lose is all part of the randomness of the system, with most of that randomness (at least the selection part) existing simply because of our ignorace. Now, the winners and losers are not completely random since we can make plenty of short-term predictions with excellent accuracy because we know more about the randomness in our macro environment than we do about gas molecules shifting about.

 

[zosima]

It looks like you are mistakenly applying characteristics of a large number of independent trials to a process which has heredity as one of its major components.

Sure species may peak at millions or billions of numbers. Consider that humans are above 6 billion now, but are thought to have number less than 10 thousand less than 100 thousand years ago. We suspect this because of how little genetic variation we have. What happened to that small population is inherited by us.

10^5 at minimum is still many times larger than the individual, but the argument was just supposed to be demonstrative. Scientists use orders of magnitude
to determine which level of explanation is significant and to help make general arguments about where to look. The argument doesn't exclude heredity, in fact it subsumes all lowerer levels of explanation. It just says changes in the species happen with much larger numbers and on much longer time scales than the individual, so insisting that we have to constantly consider the individual is being, at best, incorrigibly reductionist.

The article that Jimbob linked to actually gave us a nice example of randomness. When lines of bacteria were allowed to adapt to cold temperatures, two-thirds of them did worse when returned to high temperature. And one third did as well or better. In a competitive environment, the strains would have different advantages should the environment change again.

The mechanisms for short term randomness are there, and heredity will cause such changes to leave an imprint on future generations.

Walt

Short term randomness, what, huh? The article completely makes my point.
Lets start with the title:
"Fast-Reproducing Microbes Provide a Window on Natural Selection"
Thats right, to see natural selection on the human scale we had to watch microbes for 40,000 generations(a strong argument that change might be on the order of 10^5) and run the experiment for 18 years!

The specific example you talk about wasn't natural and it wasn't random. It was the equivalent of taking the microbes from one ecological niche to another. In nature there would be nothing random about going from 68 degrees to 102 degrees(sadly it doesn't say if we're talking F or C here...real scientific). On the other hand if you contend that this sort of temperature change is the result of weather variation, then it would happen seasonally. Also, non random
Also you could even make the argument in individualistic terms. You talked using the language of the level of explanation we've been advocating. You were talking about external pressures causing predictable changes on different proportions of a large population.

Not to mention the fact that the article doesn't mention the word "random" once.
In fact...(emphasis added)

scientists can now zero in on the precise genetic changes that unfold during evolution

Microbes can reproduce several times a day, and a billion of them can fit comfortably in a flask.

One striking lesson of the experiment is that evolution often follows the same path. “We’ve found a lot of parallel changes,” Dr. Lenski said.

Billions of individuals, thousands of generations, precise changes, parallel paths.

The example you gave with the hot and cold is the example of an advantageous trait being linked to a disadvantageous, and you are conflating the fact that the result was unexpected and previous not understood with random. It is absolutely not random, if scientists were to study the specific gene and protein that conferred resistance I'm sure they would find that there is no way it could confer an advantage to one without a disadvantage to the other. This sort of thing is what they're talking about when these say precise changes and parallel paths and it is exactly the crux of the position we are defending.

Were you just hoping I wouldn't read the article? It seems like every time a random evolution advocate links a supposed piece of evidence for their case, it involves a very selective reading of an article that as a whole completely rebuts their point.

 

[zosima]

.
Thanks for this zosima.

I really shouldn't have jumped in; there is a lot in the earlier posts you wrote in this thread, and for us to avoid talking past each other would take some time and patience.

Anyway, as the thread now seems to be about something quite different, I shall take my leave.

No stress
Ya if you want to have a calm and enlightened discussion or evolution, this probably isn't the thread to do it. Everyones been lobbing artillery in this thread for at least the last 5 or 6 pages.

 

[mijopaalmc]

zosima-

The journal article referred to the in the article that jimbob posted does in fact give an excellent example of what those on this thread who argue that evolution is random mean when they use the word random. While there was a general trend toward increased fitness at the temperature for which the strains were adapted, the actual change in fitness for the adapted strains varied widely and, in two cases was not even statistically significant, even though each of the six replicate strains was adapted to the same environment. In other words, that the strains became better adapted to their environment is indisputable and shows the orderliness and in that people like to call "non-random", but how the strains adapted to their environment demonstrates the essential randomness of the specific adaptation.

 

[Belz...]

Mijo... if evolution were random there'd be no evolution.

 

[mijopaalmc]

Mijo... if evolution were random there'd be no evolution.

The article I just citd demostrates that evolution by natural selection is random in the way that I have defined it and that it does occur.

Again, evolution can occur even without natural selection, in other words, if it is "random" as in "equiprobable".

 

[Walter Wayne]

10^5 at minimum is still many times larger than the individual, but the argument was just supposed to be demonstrative. Scientists use orders of magnitude
to determine which level of explanation is significant and to help make general arguments about where to look. The argument doesn't exclude heredity, in fact it subsumes all lowerer levels of explanation. It just says changes in the species happen with much larger numbers and on much longer time scales than the individual, so insisting that we have to constantly consider the individual is being, at best, incorrigibly reductionist.

I not considering the individual, I am considering the fact that the variation in early generations will determine the bias in much later generations.

Short term randomness, what, huh? The article completely makes my point.
Lets start with the title:
"Fast-Reproducing Microbes Provide a Window on Natural Selection"
Thats right, to see natural selection on the human scale we had to watch microbes for 40,000 generations(a strong argument that change might be on the order of 10^5) and run the experiment for 18 years!

Um, you realize that is a number of tests, not one done over 18 years.

The specific example you talk about wasn't natural and it wasn't random. It was the equivalent of taking the microbes from one ecological niche to another. In nature there would be nothing random about going from 68 degrees to 102 degrees(sadly it doesn't say if we're talking F or C here...real scientific). On the other hand if you contend that this sort of temperature change is the result of weather variation, then it would happen seasonally. Also, non random

If a the evolution of the strain show random characteristics in a predictable environment, it would be even stronger support that in less predictable situations the results will be random as well.

Also you could even make the argument in individualistic terms. You talked using the language of the level of explanation we've been advocating. You were talking about external pressures causing predictable changes on different proportions of a large population.

I'm not sure what you are refering to when you say "make the argument in individualistic terms." Actually I'm not sure what the paragraph is getting at.

Not to mention the fact that the article doesn't mention the word "random" once.

No it doesn't, it mentions that among 18 different lines, 12 showed one pattern, and 6 showed another. Do they need the word in there explicitly?

In fact...(emphasis added)
[quotes from article]
Billions of individuals, thousands of generations, precise changes, parallel paths.

The term precise genetic change doesn't say anything about randomness or not. The fact that I can spot a deletion at location X, doesn't mean that the change is predictable. Nothing special about seeing the precise result after the change has happened. Then look at the language you quote after that. "Evolution often[/n] follows the same path," and "we've found a lot of parallel changes." Not Evolution follows the same path, or we found only parallel changes. Both statements were qualified.

Now the words "often" and "a lot" does indicate that the probabilities are indeed skewed. In the experiment I outlined that skew was about 2:1. If you have been following my argument, you'll note I specifically discussed the accumulation of skewed probabilities over time. We are the accumulation of many, highly biased events. Even though each change might have a most likely outcome, the odds of them all happening are incredibly small.

...It is absolutely not random, if scientists were to study the specific gene and protein that conferred resistance I'm sure they would find that there is no way it could confer an advantage to one without a disadvantage to the other.

The fact is the strains developed (at least) two different ways of dealing with the change. Some developed one protein that confered one characteristic, and others developped a different protein that confered a different characteristic. That is a perfect example of unpredictability of changes at the gene level manifesting as significant variation at the phenotype level.

Were you just hoping I wouldn't read the article? It seems like every time a random evolution advocate links a supposed piece of evidence for their case, it involves a very selective reading of an article that as a whole completely rebuts their point.

No, I was hoping you would read that, and consider it in conjunction with my previous arguments and realise that the unpredictability, even with a 2:1 bias, might actually manifest as significant variation when looking at evolution of an ecosystem. I thought I had explained some of the properties of systems with memory (heredity) and how they differ from system without. Apparently not well enough.

As for selective reading - how one gets from only a 2:1 bias after "thousands of generations" as you say, to evidence of non-randomness, I don't know.

Walt

 

[mijopaalmc]

It is also interesting to not that the Bennett and Lenski article was published in the very same PNAS issue from which articulett draws the Ayala article that says "evolution is not random". Their findings directly contradicts Ayala's statements.

 

[Ichneumonwasp]

It is also interesting to not that the Bennett and Lenski article was published in the very same PNAS issue from which articulett draws the Ayala article that says "evolution is not random". Their findings directly contradicts Ayala's statements.

Haven't read them, but that sounds like it was very likely a conscious editorial choice. Good observation.

 

[mijopaalmc]

Ayala: Darwin's greatest discovery: Design without designer

Bennett and Lenski: An experimental test of evolutionary trade-offs during temperature adaptation

 

[Ichneumonwasp]

Ayala: Darwin's greatest discovery: Design without designer

Bennett and Lenski: An experimental test of evolutionary trade-offs during temperature adaptation

Thanks.

I don't see any direct contradiction there, though. They are discussing different systems and different ideas. Ayala uses the word 'random' in a way different from the way you do (but he is quite specific about his definitions, so I don't see any problem there), and he discusses general trends not specific examples as in the Bennett and Lenski paper.

 

[mijopaalmc]

It seems that Bennett and Lenski were also discussing general trends:

We used experimental evolution to test directly the important and commonplace evolutionary hypothesis that adaptation, increased fitness within the selective environment, is accompanied by trade-off, a loss of fitness in other nonselective environments. Specifically, we determined whether trade-offs at high temperature generally and necessarily accompany genetic adaptation to low temperature. We measured the relative fitness increment of 24 lineages of the bacterium Escherichia coli evolved for 2,000 generations at 20°C and the relative fitness decrement of these lines at 40°C. Trade-offs at the higher temperature were examined for their generality, universality, quantitative relationship, and historical contingency. Considering all 24 lines as a group, a significant decline in fitness was found at 40°C (mean decline = 9.4%), indicating the generality of the trade-off effect. However, in a lineage-by-lineage analysis, only 15 of 24 showed a significant trade-off, and one lineage increased fitness at high temperature. Thus, although general, trade-offs were not universal. Furthermore, there was no quantitative association between the magnitude of adaptive fitness increment at 20°C and fitness decline at 40°C, and no effect of lineages' historical thermal environment on either their improvement at 20°C or the extent of their trade-off at high temperature. We do not yet know the underlying mechanisms responsible for the trade-off, but they are sufficiently prevalent to drive a general effect. However, approximately one-third of the experimental lineages achieved low-temperature adaptation without detectable high-temperature trade-offs; therefore, it cannot be necessary that every change conferring benefit in cold environments has a negative effect on function in warmer environments.

 

[sol invictus]

The fact that evolution is random doesn't effect the fact that all life...

Tell us, mijo: do you still agree with that? Is evolution random?

YES or NO?

If you won't answer, you've changed your mind (which is OK, really).
If you say yes, we're going to have a great time with it (since then so are smoke detectors).
If you say no, the thread is over.



So??

 

[mijopaalmc]

sol invictus-

I have answered your questions several time. I still believe that evolution by natural selection is random. However, evolution by natural selection displays predictability and regularity; that doesn't make it non-random.