[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Subject Index][Author Index]
>Philidor 11 wrote:
>Randomness refers to the generation of new capabilities through >mutation.
You're using it as equivalent to the process of natural >selection itself.
Descent with modification: Sheesh, Chris.
No, no. I'm certainly not equating natural selection with randomness. I'm
refering to the old example (who came up with it now? Gould?) to explain
how average complexity (difficult concept that) can increase over time even
if the same number of lineages are becoming more and less complex....
Take a drunk and stand him next to a wall (on his right hand side). Let him
stagger along until he falls over (the idea is that he should stagger
randomly). Mark the position. Pick him up and let him do it again. Mark the
position again. Do it several thousand times. Because the wall is to the
drunk's right, all final resting places will be either against the wall or
to the left of it. Some will be a long way to the left. Increase the number
of attempts and the chances are that the distance of the furthest left
fall-over point will also increase.
In the case of evolution, the wall represents the fact that there is a
minimum complexity for life. Start here, as life must, then some lineages
will remain at minimum complexity, some will increase in complexity, but
none can decrease. Increase the length of time over which lineages have the
chance to stagger around and the complexity of the 'leftmost' product of
natural selection will increase by the law of averages.
This has to be the null hypothesis to explain the observed increase in the
complexity of life over time. Whether it's right or not, I do not know.
There was an article written recently in which someone claimed to have
tested the theory by determining the number of lineages that decreased and
increased in 'complexity' over time (complexity is fiendishly tricky to
define and I can't remember how it was done in this instance). I seem to
remember that the results gave tentative support for the random walk
scenario. Anyone got the reference?