On 06/12/2007, at 2:56 AM, Nick Arnett wrote:
>
>
> I'm talking about the Santa Fe Institution people and those doing  
> related
> work.  Kauffman, Waldrop, Holland, Arthur, Lewin, etc.

Right, now I'm a lot closer to understanding what you're alluding to  
(but it's the Santa Fe Institute...). I followed the early A-life  
stuff very carefully, and think that Chris Langton's work is  
fascinating. Kauffman is very interesting too, but he's got a lot of  
work to do. I'll discuss it below a bit.

> Just looked to see who Dembski is.  I guess I need to say clearly  
> that I am
> not nor ever have been a proponent of intelligent design.  I find  
> that whole
> idea and movement rather nauseating.  It seems to me to be rather  
> obviously
> based in fear, not science.

Yes. But when you start saying things like "complexity poses  
challenges to Darwinian models" without providing examples, you're  
echoing (inadvertently it seems), one of the major battle cries of the  
ID "movement". Gets my hackles up, because it's vacuous at best and  
downright unscientific more likely.

>  I feel the same way about people who assume
> that anybody who is unsatisfied with Darwinian -- natural selection  
> as the
> over-reaching mechanism of speciation -- must be proponents of  
> intelligent
> design.

It depends what people are saying. If one actually proposes some other  
model, then it can be evaluated, and there's a discussion, and we're  
doing science. But if you look back over the last few posts, you'll  
see that I've been trying to ask what models you're talking about, and  
it took several attempts to even get you to even mention some  
scientists by name. It appeared extremely evasive from where I'm  
sitting, and that again is a red-flag to pseudoscience. Bear in mind  
that outside of Brin-L, I spend a lot of time discussing evolutionary  
biology, and I'm well used to cranks hijacking discussions. People who  
make sweeping statements without getting involved in specifics are 9  
times out of 10 cranks, or at the very least don't know what they're  
talking about.

> It's like it's impossible to engage in debate without first
> rejecting the lunatics.

Which is what I've been trying to do by asking you exactly what you  
were alluding to. It's a lot easier to engage in a debate if you  
actually engage in it.
>
>> Emergence isn't trivial, it's actually an important insight, one of
>> those (like natural selection) that seems so damned obvious in
>> hindsight that it's hard to imagine not understanding it. However
>> you're right in that pointing out that a system exhibits emergence
>> doesn't tell you much about it unless you bother to discover the
>> nature of the simple causes and how they generate complex results.
>
>
> I wasn't saying that emergence is trivial.  I was saying that it is  
> trivial
> to describe emergence.  As I think you're saying, figuring out the
> implications of emergence is challenging.  There's a lot to be  
> discovered by
> those who can figure out the mathematics that will allow us to model  
> many
> kinds of emergent phenomena, which currently seem to be beyond- 
> astronomical
> in magnitude.

Emergence has applications in ecosystems, crowd control, city design,  
animal behaviour, surveillance, neural nets, and so on. Modelling  
those systems through a few simple rules is a challenge, but not  
beyond our capacity. Interestingly, some of the most successful work  
has come out of games and movies - SimCity exhibits some emergence,  
and CGI crowd/battle scenes
>
>
> So... perhaps I can answer your question this way... we don't know  
> much much
> of evolution is driven by simple rules that are inherent in the  
> universe
> (thus the anthropic principle) v. how much is driven by competition.

Why are those two different things? Evolution as currently accepted  
*is* driven by simple rules. If you have inheritance and differential  
breeding rates, that lead to changes in gene prevalence over  
generations. That's all evolution is. There is a lot in the complexity  
of ecosystems (or in maintaining simple systems in a stable way) that  
could be understood and explained by emergence. Especially interesting  
is actually the pre-evolutionary field of abiogenesis, where  
hypercycles may turn out to explain how a set of complex interactions  
of molecules could bootstrap out of the prebiotic chemical soup.  
Indeed, emergence and hypercycles may go a long way to explaining how  
biochemical systems like the Krebs cycle could have appeared and been  
incorporated. Look at one of those huge posters of the human metabolic  
pathways, and it absolutely screams emergence. Here:

http://expasy.org/cgi-bin/show_thumbnails.pl

Kauffman argues that the complexity of systems may result as much from  
emergent phenomena and complexity, non-linear dynamics, maybe chaos,  
as they do through natural selection. Well yes, they probably do. But  
I think these are two related but not totally overlapping areas -  
natural selection explains long-term genetic drift, and emergence and  
complexity explain a lot of ecosystem dynamics. This is an emerging  
(heh) field, and he and his colleagues have a lot to do to convince  
the rest of the field. Thing is, if they're right, they'll be able to  
show it.
>
>
> Saying it another way... complexity says that the interactions of  
> lots of
> agents gives rise to unpredictable (so far) phenomena.  At the  
> simplest
> mathematical levels, it is meaningless to describe those  
> interactions as
> competitive or cooperative, but at higher levels of observation, such
> behaviors appear to emerge.

Yes. But that's describing behaviour, not evolution (which is simply  
changes in gene frequencies in a population over time). Now, there's  
some speculation that DNA has a bit more going on than just a gene  
carrier: - it's been postulated that interactions of genes can act in  
a self-organising way, or even as a form of calculating device, a  
genetic computer. But this is controversial, and it's going to take a  
lot of work to show this. Interesting line of study, however.

Charlie.
_______________________________________________
http://www.mccmedia.com/mailman/listinfo/brin-l

Reply via email to