phillip smith  wrote:
 William Morse wrote:
 "Anon."  wrote in news:cl65sn$ave$1

>snip<

> > BOH:-
> > This is, of course, not what we want: we know that bipedalism has
> > evolved, and want to know whether it was by drift or something else.  To
> > estimate the probability that it was drift wot done it, we have to
> > calculate the probability that it some other factor than drift caused
> > bipedalism, and then use Bayes formula to invert the probabilities.  At
> > this point, I will only point out that probabilities of fixation of
> > alleles can be increased by giving them a positive selection coefficient
> > - we would need to know what the selection coefficients were to do the
> > calculations.
> > After all that, I can give you a list of 57 different parameters that
> > are necessary for the calculation, but are impossible to estimate.
> > There may be a different, and better approach to this analysis, if so
> > I'd be interested in hearing it.  One that needs only 32
> > impossible-to-estimate parameters would be fine.

>PS:-
> I would like to confuse the issue more by suggesting both selection and
> drift  occurred.

JE:-
Phillip Smith has not "confused" the issue he has clarified it.

The view that it is somehow reasonable for anybody to
_seriously_ propose that a complex trait (a trait based on
epistasis) like bipedalism could have evolved entirely
via variation provided by just a random process such as
sampling error (only because such a heuristic possibility
exists within Dr O'Hara's massively over simplified model) is
not rational because such a hypothesis cannot be tested
to refutation. Such a model view is only valid if it helps
test the theory it was derived from to refutation. A model is not
supposed to replace the theory the model was oversimplified from.
Doing so reduces refutable evolutionary theory to just a non
refutable mathematicians Mad Hatters Tea Party. No way exists
to delete random variation produced by random process from any
natural population even if Dr O'Hara can delete it from one
of his over simplified models. However, natural selection can
be entirely deleted for a _significant_ time because it is not just
a random  process. I have outlined an experiment that could do this.
This being the  case, no EXPERIMENT can be proposed that can test the
hypothesis that  bipedalism could have evolved entirely via variation
provided by just a random process to refutation, only to non verification.
This means even if after all natural selection has been deleted bipedalism
was observed to evolve the hypothesis is only verified. If it doesn't evolve
it is not verified but that is ALL of it. No point of refutation exists
for this hypothesis simply because any random variation caused by a random
process cannot be deleted from any natural population. Unless you can
delete the process you are testing from a natural population
you cannot test that process to refutation only to non verification.
If you delete selection but evolution still occurs via just
non removable variation then selection stands refuted and is not
just non verified. Refutation is definitive but non verification
is not definitive.

While model building is extremely useful to help suggest ways
to test the theory that every model _has_ to be a simplification
from, only that theory and _not_ any of its derived models can be
tested to refutation. Neo Darwinist model builders have taken model
misuse to new, dizzy heights. It appears that if you pay model
building specialists to just sit in rooms with computers
building endless models they no longer need to care about how these
models are supposed to be joined to refutable theories of reality.
That is OK because it is the job of biologists TO CARE.



Regards,

John Edser
Independent Researcher

PO Box 266
Church Pt
NSW 2105
Australia

edser{at}tpg.com.au
---
þ RIMEGate(tm)/RGXPost V1.14 at BBSWORLD * Info{at}bbsworld.com

---
 * RIMEGate(tm)V10.2áÿ* RelayNet(tm) NNTP Gateway * MoonDog BBS
 * RgateImp.MoonDog.BBS at 10/25/04 10:54:27 PM