Re: Directional evolution



On Dec 18, 1:17 am, Tim Tyler <seemy...@xxxxxxxxxxxxxx> wrote:
Bill Morse wrote:
Gould has argued that the apparent increase in complexity with time is
simply the result of a drunkard's walk from the fixed wall of zero
complexity . Others (myself included) think that increased complexity is a
direction that offers new niches and so is a direction that will be taken
by evolution. [...]

Gould's model resembled the expansion of a gas, next to a wall.

However, common sense suggests the model of a gas expanding next to a
wall - into a *vacuum*. It seems that this model would be better on
the grounds that species may well expand semi-randomly into vacant
adjacent niches - but that one thing that is certain is that the
niches are /initially/ all *empty*.

I think I can modify Gould's analogy just a little for clarity.
One can think of evolution like the diffusion of solute through a
solvent. The solute molecules will tend to collect in spaces that are
analogous to niches. However, the solute molecules still move outward
with time, they just get hung over in these spaces.
Solid state solutions are probably the best analogs to the
adaptive landscape. In particular, think of an impurity atom (dopant)
diffusing through a hot crystal from an initial starting point. The
impurity atoms will collect at certain lattice points. But they will
still diffuse outward from the starting point. Maybe the nich of self-
aware, high-tech intelligence is merely an extreme point on the
crystal, far from the initial doping site.
Or think of the annealing of a crystal, where the defects in a
crystal move through the crystal and interact. These defects tend to
go annihilate each other, and collect to form very complex networks of
defects. The local order of some of these networks can be very high,
and has been said to approach biological complexity. Modeling this
process is sort of a quagmire. However, I think the process is in many
ways analogous to biological evolution. but in some ways the current
models. Despite the formation of stereotypical forms of network
defects, the overall motion of the defects has to go outward from the
center.
Eventually, very complex and large defect networks form. The
existence of such networks inhibits the formation of othe defects.
This is why tempered steel is so hard. One can't break the sword
because the complex defects from tempering prevent the defects that
later threaten to break the sword. So in a way, tempered steel has the
"network" niches filled. Competition between defects is what makes a
defect hard. Almost a Wagnerian metaphor!
Some of the models of abiotic genesis, on the surfaces of solids,
really resemble a type of annealing model.

.



Relevant Pages

  • Re: Why does Software have bugs?
    ... Why does Software have bugs? ... Requirements defects exist in the SRS because it was improperly formed. ... The complexity of current software applications can be difficult to ... Thus an array indexing error is still an array indexing error whether it was inserted in response to original requirements or in response to a change in requirements. ...
    (comp.software.testing)
  • Re: Data Encryption on TurboImage/MPE
    ... Complexity: ... infected mobile computers re-attached to internal networks. ... a sudden need for encrypting data where it resides. ... If the hypothetical HR Dept. has its data on a host, ...
    (comp.sys.hp.mpe)
  • Re: Making money on open source, if not by selling _support_, then how?
    ... You obiously don't understand the complexity of cutting 2x4's to length. ... be scanned and analyzed for defects. ... I don't program trimmers, but others in my department do, and I am ...
    (comp.lang.ada)
  • CFP: Workshop on Computation in Genetic and Biochemical Networks
    ... Workshop on Computation in Genetic and Biochemical Networks ... complexity lies not within the individual gene products but rather ... biologists trying to understand biological function, ... Michael Lones, Department of Electronics, University of York ...
    (comp.ai.genetic)
  • CFP: Workshop on Computation in Genetic and Biochemical Networks (CoGBiN 2006)
    ... Workshop on Computation in Genetic and Biochemical Networks ... complexity lies not within the individual gene products but rather ... biologists trying to understand biological function, ... Michael Lones, Department of Electronics, University of York ...
    (sci.nonlinear)