From: Christopher Healey (CHealey@unicom-inc.com)
Date: Wed Feb 02 2005 - 11:58:19 MST
This would be true if that "left-wall" was a static barrier set at zero complexity. But it is not.
Evolution effectively ratchets this "left-wall" forward to higher base levels of complexity. This occurs because the complex functional adaptations (CFAs) that develop tend to break badly under random mutation. I don't mean to assert that decreases in complexity don't happen, but the dynamics are definately biased. A slight bias is all that is necessary, though. In other words, the accretion of complexity occurs because of a localized random walk, whose results are pruned more aggressively when they cross under the threshold for maintaining the proper CFAs. Crossing below that point hits a sharp downward knee in the fitness curve. So globally, we see a trend directed toward complexity. The fuzzy-bound has moved toward the right.
What's kind of neat, is that as complexity continues to increase, not just including more CFAs, but also supersystems of CFAs, you could expect that supersystem- and CFA-disabling mutations would cause much sharper drops in fitness. This should create a strong and sustained selection pressure for specific complexity to form as a way of minimizing highly-disabling mutations. I believe we call it sex.
With mechanisms having evolved to prevent the spread of highly-disabling mutations, as well as increasingly negative phenotypical consequences against mutations that break CFAs and supersystems of CFAs, one might even expect complexity in species to increase at an exponential rate (accreted complexity feeding back to bias against decreasing complexity).
Of course, I am not an evolutionary biologist, so I might be misunderstanding some of the literature.
From: email@example.com on behalf of Phil Goetz
Sent: Tue 2/1/2005 9:20 PM
Subject: Re: Gould and complexity
Yes, and the point of the book is that this increase
in the complexity of the most complex species is due
to a random walk with a left wall, not to any
inherent drive in evolution towards increasing
This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:50 MDT