On 02-15-98 Andrew Cummins wrote to Day Brown... 
 AC> It has never been demonstrated by observation or demonstrated by 
 AC> rigorous theory that self-replicating objects can gain complexity 
 AC> through a process of Natural Selection and Mutation (long term). 
 AC> Indeed, any such process appears to favor a reduction of complexity. 
Iterate fractle formulae on a computer; you can start with a very 
simple formula such as A + B + C = D, where A= any integer, 
B= an integer to some power, C= a product of B+A, and then, D is 
used to replace one of the other terms in another calculation to 
derive a new 'D'. 
 
You can ascribe various colors to various magnitudes of elements 
in the formula, and locate points according to the relative size 
of elements to each other, and what do you get? images of such a 
degree of complexity that you would be hard put to even imagine 
how the simple formula can produce them, *even after you know* 
what the simple formula is. 
 
The DNA is an iteration of simple relationship, and the complex 
way it self replicates looks to be an example of this iteration 
over time that does indeed, gain complexity.  I grant you that 
there may be formulae and iterations that reduce to simplicity. 
But those that do, do not self replicate, so we have a form of 
a self-fulfilling prophecy, in that any that do replicate will 
do so, and evolve all that we see as a by-product.  
___ 
 * OFFLINE 1.58 * Death to all fanatics! 
--- Maximus 3.01
---------------