John Edser  wrote or quoted:
> Tim Tyler  wrote:-

> > TT:-
> > The issue of to what extent selection and drift are responsible
> > for observed features of organisms is quite subject to
> > experimental investigation - the hypothesis that selection is involved
> > predicts a lot of convergent evolution - whereas the hypothesis
> > that drift
> > is involved predicts much greater morphological diversity.
> 
> JE:-
> The proposition that has to be able to be tested
> to refutation is _very_ explicit: Can the process
> of random sampling error cause evolution entirely
> on its own without selection where evolution is
> defined as ANY gene freq. change in a deme?
> 
> The short answer is that such a proposition
> CANNOT be tested to _refutation_ it can only be
> tested to _non_ verification. This means that
> all that can be achieved is the elimination
> of selection for a significant period of time,
> in the experimental way that I have described.
> Drift cannot be eliminated unless the experimenters
> have an infinite population at their disposal
> (which of course they do not). A random gene
> freq. change due to random drift must
> always exists and cannot _cannot_ be eliminated.
> This being the case, all that can be done is
> provide an expanding population that is only
> subject to random processes (drift mutation
> etc). Of course, this experimental population
> will always produce a gene freq. change and thus,
> according to the definition "evolution".

So you (finally!) admit that sampling errors cause
evolution - as it is conventionally defined.

> However, my prediction is that such gene
> freq. changes can only produce the dissolution
> of experimental genomes. This does NOT
> refute the "any gene freq. change in a deme"
> definition but it would show that such a
> definition remains irrational as well as
> just an ubiquitous iron man theory.

....and you want to redefine evolution.  That's fine.

You might want to specify your new proposed definition -
and then people can decide which one they want to use.

Also - if you are using your own definition - you might
want to remind people of that occasionally - in order
to prevent misunderstandings between yourself and other
people who are using the word in the conventional sense.

> > TT:-
> > Recall the experiment where you fire ball bearings "at
random" into
> > a grid of vertical spikes on an inclined plane - and the ball bearings
> > wind up forming a normal distrubution in buckets at the bottom of
> > the slope?
> > That's a classic example of random processes producing a non-random
> > pattern.
> 
> JE:-
> If a supposed "random process" produces a "non-random
> pattern" then the proposition it was just
> a random process, stands _refuted_.

In that case you will be happy to hear that drift is
not a random process - by your definition.

All that was ever claimed for the randomness of mutations
is that they are random with respect to what is beneficial.

They are not "totally" random.  Some parts of the genome
are more subject to mutation than others - for example.
-- 
__________
 |im |yler  http://timtyler.org/  tim{at}tt1lock.org  Remove lock to reply.
---
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 10/2/04 5:06:41 PM