John Edser wrote:
>>>JE:-
>>>The definition of evolution as  "a change in gene
>>>frequencies" does not exclude random changes in
>>>gene freq. Thus the definition is incorrect because
>>>it allows a _non_ testable view of evolution.
> 
> 
>>BOH:-
>>Not so, see for example Mueller et al. (1985) Genetics 110: 495-511.
>>The specifically test changes in allozyme frequency changes to see if
>>the changes could be caused by genetic drift (sampling error is even
>>easier to test for).
> 
> 
>>JE:-
>>Incorrect. All random patterns, without
>>exception, can be validly suggested
>>to be caused by _either_ random or
>>non random processes. This is because non random
>>processes commonly produce random patterns
>>but random processes can only produce random
>>patterns, by definition. This being the case,
>>it is logically _impossible_ to refute
>>any claim of evolution being caused by just
>>a random process. 
> 
> 
> BOH:-
> Rubbish.  Read something simple on stochastic processes.  Random 
> processes can be characterised by their properties (means, variances 
> etc.), and then thesecan be compared with the data.
> 
> JE:-
> Do you understand the difference between a random pattern
> and a random process?
> 
Yes.  What evidence do you have that Mueller et al. observed a random 
pattern?

> I did not say that random patterns could never be observed,
> I simply said that all observed (verified) random patterns can
> be assumed to be produced from *EITHER* a random OR non random 
> process; no exceptions. The observation on a random pattern 
> is *NOT* definitive for a random process, alone, causing it. Thus 
> the assumption of a random process (not a random pattern!) can be 
> refuted but it cannot be verified, ever. 

Indeed.  And this is what Mueller et al. did.

This is the reason why
> random observations are simply thrown out as inconclusive within 
> the sciences. Statistics was invented to separate random
> patterns (not random processes!) from non random patterns.

But now we are also able to examine processes too.  I repeat - look at 
the literature on stochastic processes.


> 
> BOH:-
> To bring it to the biology - the amount of variation due to genetic 
> drift depends on the effective population size (N_e).  Mueller et al. 
> measured N_e in the butterfly, and then used that to see if the changes 
> in allozyme allele frequencies were small enough to compatible with the 
> variation expected from genetic drift.  In several cases the variation 
> was much larger than expected.
> 
> JE:-
> Typically, you confuse a random 
> pattern with a random process. While
> the refutation and verification of either
> a random or non random pattern is possible,
> it is not possible to verify that only a random 
> process caused any verified random pattern, ever.
> 
How can you refute a random pattern?  Surely all you can do is refute 
that it was created by a random process, which means that you are 
refuting the hypothesis that it was created by a random process.

Bob

-- 
Bob O'Hara

Rolf Nevanlinna Institute
P.O. Box 4 (Yliopistonkatu 5)
FIN-00014 University of Helsinki
Finland
Telephone: +358-9-191 23743
Mobile: +358 50 599 0540
Fax:  +358-9-191 22 779
WWW:  http://www.RNI.Helsinki.FI/~boh/
---
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 12/5/03 6:03:21 AM