Brice writes:
> Fred was an Atheist who had an epiphany. He came to reverse his opinion
> on that based upon scientific evidence. He later exposed some 'holes' in
> evolutionary theory, one of which was sometimes called the 'two mountains'
> hypothesis, or theory, or proposal. I do not have the reference here so I
> cannot go into detail. As well read as Wirt might be, I cannot help but
> wonder if he has ready anything by Hoyle, except maybe how to play
> card games.
Asking as to whether I would hire Fred Hoyle, he wouldn't be on my short list,
not because he wasn't intelligent, but because he strenuously worked towards
the answer that he wanted, selectively culling that evidence that fit his
predispositions and ignoring the rest. In that manner, he exhibited all of the
worst sins of the creationists, and that's no way to do science.
As to his "two mountains" hypothesis, I had never heard of it before. Nor have
apparently many others. In the several billion webpages on the internet, this is
only page that I could find that referenced any similar comment, and it's not
really clear that the comment is even associated with Hoyle:
http://iidb.infidels.org/vbb/archive/index.php/t-148104.html
However, the correspondent in that discussion gets the answer wrong to the
question posed. The question is: how does an evolving population pass from
one peak of optimization to another without first going through a valley of
lower quality.
This is a question that occurs in the first week of every class on evolutionary
optimization and I've written about it often. Here's one of my papers from 14
years ago, written in an IEEE journal to an engineering audience, where I
answer the question. It at least has the benefit of being brief and to the point:
http://aics-research.com/research/notes.html#IID
But its second benefit is that it's easy to demonstrate that it's not a problem
and is easily surmounted when code and behavior are separated. Selection
operates against the behaviors it sees in the context of its current
environment; mutation operates on the inherited code without regard to
selection or the environment in which it will be expressed
Wirt Atmar
* To join/leave the list, search archives, change list settings, *
* etc., please visit http://raven.utc.edu/archives/hp3000-l.html *
|