Jeremy Pallant

The article in question is entitled <a href="http://www.sciam.com/article.cfm?articleID=000D4FEC-7D5B-1D07-8E49809EC588EEDF&catID=2" target="_blank">15 Answers to Creationist Nonsense</a>.

echidna

Maybe it gets more response than the last time. From last month …

<a href="http://iidb.org/cgi-bin/ultimatebb.cgi?ubb=get_topic&f=58&t=000974" target="_blank">15 answers to creationist nonsense/monkeys on a typewriter</a>

Shadowy Man

I read this article in the recent Scientific American. I thought it was very good, except for his answer to number 8.

(emphasis added)


Doesn't this argument just help the Intelligent Design scenario? He is saying that random letters were selected, but if they agreed with what he a priori wanted the phrase to be, he would keep them, and throw out the ones that didn't.

Am I missing something here?

[ July 11, 2002: Message edited by: Shadowy Man ]</p>

madmax2976

The idea I believe is to emphasize that evolution is not a random process in that it "selects" things that "fit" so to speak. In this case, one algorithm did the "mutating" and another algorithm did the selecting. In natures case, Natural Selection does the selecting.

Creationists often misrepresent evolution as a "random" process, but this is not so and this example points that out.

Corwin

Many of the engines that drive evolution, such as mutation, are random.

The 'control' mechanism, natrual selection, is not.

One can make an argument that this non-random element is a proof for Intelligent Design, but Occam's Razor suggests that since the criteria seems to be 'that which continues to survive long enough to breed, and has an advantage when it comes to breeding...' that a natrualistic explanation works just fine. (Since if lifeforms didn't develop by these criteria, we wouldn't be here to care... so the criteria would be meaningless since there would be nobody to assign meaning to them.)

It's a bit convoluted I know... it would probably be less so if I'd had more coffee this morning.

Tom by the Bay

I also enjoyed the SciAm article. But I would have to agree with Shadowy Man the program in question leaves something to be desired. The program introduces a non-random element in an artificial way. The program's purpose is to create a specific phrase by selecting those random mutations which more closely match it. In other words, the program "knows" the answer a priori and is making directed changes in order to reproduce it. A creationist might argue that God (or whatever) does the same thing, so the program proves nothing about evolution.

Nature doesn't work that way. It is trying to solve a problem (or set of problems) anyway it can. Every species is a separate solution to the problem of survival. I think a more appropriate computer program would be one in which a random set of N characters is generated and compared against, say, the rules of English (or some other language that isn't as horrid as English). For instance, sequences with only consonants would be selected against, while those with a combination of consonants and vowels would be selected for. By adding other rules, it might be possible to make a program that, after a number of generations, would converge to a proper English phrase - but not always the same phrase! In this way, it might be possible to demonstrate that a consistent set of rules (environment) can produce a proper answer (more fit organism). But, as in nature, there will be many answers.

-tom

[ July 11, 2002: Message edited by: Tom by the Bay ]</p>

Shadowy Man

Tom: Well said. I agree completely.

echidna

I agree with Tom & SM, it’s kinda like Evolution playing “hotter and colder”.

The program seems to follow Lamarkianism.

(Tangent : Is it my imagination or is SciAm slipping its standards ? That whole environmental debate with Lomborg was quite unscientifically addressed IMO.)

bluefugue

It's the same thing when, in The Blind Watchmaker, Dawkins wrote a program that used random letter combinations and cumulative selection to achieve the phrase "Methinks it is like a weasel" in about 60-odd generations (whereas, starting from scratch each time, it would probably have taken trillions of tries to get it). Dawkins admitted that this is different from evolution because natural selection has no "target phrase." His point, however, was simply the power of cumulative selection to quickly achieve results that seem in retrospect to have been "improbable." I recall a debate in the evo-cre forums in which mturner argued the presence of a "target phrase" basically invalidated the whole example. This seems rather harsh to me -- as an example of the power of cumulative selection, I think it is still valid. It just doesn't tell the whole story. But the main point is that if you get to save your "work" every generation, you can quickly build up quite complex systems. It is perhaps true, however, that the question of teleology (is it set up ahead of time by an intelligence, or only applied retroactively to a blind process?) is not settled by the "Methinks it is like a weasel" program.

excreationist

AiG have written a point-by-point response to this called <a href="http://www.answersingenesis.org/news/scientific_american.asp" target="_blank">15 ways to refute materialistic bigotry</a>

They also have <a href="http://www.answersingenesis.org/news/0711sciam.asp" target="_blank">an article</a> saying that Sci Am. wants AiG to immediately remove that previous article because of copyright reasons otherwise they'd sue or something.