Re: News: Replaying evolution.

John Harshman wrote:

dkomo wrote:

John Harshman wrote:

dkomo wrote:

Ye Old One wrote:

Replaying evolution
By Patrick Barry
June 2nd, 2008

http://www.sciencenews.org/view/generic/id/32801/title/Replaying_evolution_


Scientists show that happenstance mutations matter



WTF? The only thing this experiment shows is that *all* mutations matter, whether happenstance or not. Evolution is sequential. The way a species evolves in the future depends very much on how it evolved in the past. A fish will evolve as a fish because it followed the fish path in evolution in deep time. A fish won't evolve in the future like a mammal.

Duh!



Duh, indeed. Since a fish evolved in the past like a mammal, what exactly convinces you that another fish won't evolve like a mammal in the future?


You mean "evolved in the past *to* a mammal". Actually, the sequence was more like fish -> amphibian -> reptile -> mammal. The founder fish as it started down this long evolutionary trajectory was still a fish, or what passed for a fish in those long ago times, and evolved *like* a fish for relatively short time periods, a few hundred to a thousand generations or so.


What could be more like a mammal than a mammal? I really don't see a point here. What do you mean by "evolved like a fish"? And how is that different from "evolved like a mammal"? You should also remember that, speaking cladistically, mammals *are* fish.


By "evolving like a fish" I simply mean that the organism has a large
collection of fishy components (structures) as its phenotype. In the
near term, these will constrain how the fish can evolve. It will evolve
by gradually changing some of these components -- evolve like a fish.

We wouldn't expect the fish to acquire tits in a few generations and
begin to suckle its young. If it did that, it would "evolve like a
mammal" from that point because mammary glands are part of a mammal's
phenotypical kit.

Or are you just trying to invoke Dollo's Law, and are stating it poorly enough that I can't tell?


I don't know Dollo's Law.


Then what are you trying to do? Dollo's Law merely states that evolution is irreversible in detail.

Google is your friend:
http://en.wikipedia.org/wiki/Dollo's_law


That's a nice law, but it hasn't reckoned with the greatest local
entropy reversers in the known universe -- humans. What humans do is
still part of evolution. Example of violating Dollo's Law: reverse
genetic engineering an Emu into a dino-Emu, which would be part
dinosaur, part Emu. This was explained in a Discovery Science Channel
documentary last night.

If Stephen Jay Gould were alive today, he would be smiling. Maybe even
gloating.

No he wouldn't. Sorry.


He probably would. He was that sort of guy.


I mean he was smart enough to see that the aforementioned research didn't prove his replay thesis.


That depends on what he meant by his replay thesis. I think that he was referring to contingency, and the aforementioned research is a clear case of contingency.

New research suggests that the famous evolutionary biologist was right
when he argued that, if the evolution of life were “wound back” and
played again from the start, it could have turned out very
differently.



This research shows nothing about the randomness of the mutations in those microbes.

This is all much ado about nothing.


Agreed. All that's shown here is that slight differences in initial conditions can result in big differences in evolutionary outcome. Is anyone surprised? Now in fact I think that's what Gould meant by "rewinding the tape of life". Obviously he couldn't have meant it as a literal analogy, since no matter how many times you rewind a tape, it still ends the same way. And a big effect from slight changes in initial conditions is exactly what "contingency" means.


If this is all Gould is saying then it is trivial. I'd be embarrassed to even mention it, let alone support it as some kind of important statement about life.


[snip some boring parts]


Analogy was between counterfactuals in alternative human history and
Gould's "contingencies". Answer the following question, then: how are
the Gould contingencies in evolution different *in principle* from the
counterfactuals in human history? How is a Gould contingency different
from Mrs. O'Leary's cow kicking over the lantern in the barn and burning
down Chicago? Try not to be boring with your answer.

Great fun this replaying the tape of human history, as is replaying the tape of life. But both are fantasies. The evolution of life could only have taken a single path, which is exactly the path it actually did take.


I find this much less interesting than Gould's argument, if I can make sense of your claim. The only way I can do so is to suppose you are arguing for an absolutely deterministic universe, in which the initial condition absolutely determines the condition at all later dates. In such a universe, we can agree that there was only one possible path, given the initial condition. But this is much more trivial than Gould's argument. Nor do I think the premise is true. You have admitted that quantum events can have macroscopic consequences. If you want to retain your claim (if I'm right about your claim), you have to assume that quantum events are deterministic. Do you assume that?


Absolutely not. But I can claim that these random quantum events will
make little difference in the grand evolutionary scenario.


--dkomo@xxxxxxxx




.

Relevant Pages

  • Re: News: Replaying evolution.
    ... A fish will evolve as a fish because it followed the fish path in evolution in deep time. ... You should also remember that, speaking cladistically, mammals *are* fish. ...
    (talk.origins)
  • Re: News: Replaying evolution.
    ... Scientists show that happenstance mutations matter ... A fish will evolve as a fish because it followed the fish path in evolution in deep time. ...
    (talk.origins)
  • Re: News: Replaying evolution.
    ... A fish will evolve as a fish because it followed the fish path in evolution in deep time. ... You should also remember that, speaking cladistically, mammals *are* fish. ...
    (talk.origins)
  • Re: News: Replaying evolution.
    ... Scientists show that happenstance mutations matter ... A fish will evolve as a fish because it followed the fish path in evolution in deep time. ...
    (talk.origins)
  • Re: Trilobites
    ... everything alive today evolved from a common ancestor. ... They went extinct before mammals evolved, ... That MUST MEAN EVOLUTION IS A LIE ... but there's no reason why they would evolve into mammals. ...
    (talk.origins)