[ExI] Is Evolution Random?

[Bryan Bishop]

On Thursday 06 December 2007, citta437 at aol.com wrote:
> Is evolution random?

I am no statistician, but how could anything be truly random? I once sat 
down to think about this, to try to prove to my father that nothing can 
be random: if something was random, then my random number generator on 
my box (which I know is pseudo, yes) would output an entire (mutant) 
cow instead of a number, for it is random, or a mass of computronium. 
But this would be in violation of cause-and-effect, thermodynamics, and 
other particularly important laws that we have so far held to be true.

There is a finite set of 'choices' (whether or not they are 
truly 'choices' in a deterministic universe is another matter) from 
which a process can select. This doesn't allow selection from an 
infinitey of choices. "All consumers must be specialized to some 
degree."

> To understand the randomness claimed for evolution by scientists, as

What scientists claim that evolution is random?

> Another way to say this is just that the changes that get encoded in
> genes occur with no forethought to the eventual needs of the organism
> (or the species) that carries those genes. A gene change (for
> instance, a point mutation -- a mistake at a single locus of the
> genetic structure) may change in any way permitted by the laws of
> molecular biology, according to the specific causes at the time. This
> may result in a phenotypic change that may be better suited to
> current conditions than the others about at the time. However, it
> probably won't. So far as the local environment is concerned, the
> change is the result of a random process, a black box that isn't
> driven with reference to things going on at the level of the
> environment. It's not really random, of course, because it is the
> result of causal processes, but so far as natural selection is
> concerned, it may as well be."

Just how useful is it to say that something is random in one context, 
but in the greater context it is indeed not random? I am interested in 
exploring the results of this idea of the contextualization of 
randomness.

> Is it safe to say then that Evolution can both be random in some ways
> as in genetic mutation and deterministic in the molecular level?

Perhaps, when considering evolution in terms of only biology, one can 
ignore the underlying physics of why various stray photons may be 
causing particular mutations, or why the one-in-a-billion transcription 
error due to faulty diffusion gradients makes another such mutation, 
and then call *that* random. Hasn't this, then, been the aim of the 
astrologers for a long, long time? To be able to connect all that 
happens in the skies to the world down below. Astrophysicists probably 
have a better chance at this, yes.

- Bryan