Text 1196, 92 rader
Skriven 2004-12-27 12:58:00 av Joe Felsenstein (1:278/230)
Ärende: McGinn's disproof of Hard
=================================


In article <cq341u$1kfj$1@darwin.ediacara.org>,
Jim McGinn <jimmcginn@yahoo.com> wrote:
>
>Joe Felsenstein wrote:
>> In article <cp0ptg$192u$1@darwin.ediacara.org>,
>> Jim McGinn <jimmcginn@yahoo.com> wrote:
>> >JMc:
>> >> > Hamilton's arguments don't work.  They just appear
>> >> > to work because those that think they do work are
>> >> > using Hamilton's arguments themselves to verify its
>> >> > conceptual accuracy.
>>
[me]
>> This is the same old objection from McGinn.  We go
>> through the same process every time.  We come up with a
>> simple model, in which it can be shown that Hamilton's
>> relatedness measure is the one to use to derive the
>> condition under which a rare allele predisposing to an
>> altruistic behavior will be expected to increase in
>> frequency.
>
[McGinn:]
>Your argument is, an has always been, that we should take your word
>that this is what it shows.  I choose not to take your word for it.

Nonsense.  I can define a model in which Hamilton's rules work.  But
when I try to do so, McGinn diverts the discussion onto whether these
assumptions "lack causal validity".  It turns out that he objects to
all models that anyone has ever used in theoretical population genetics!
In other words, his statements about Hamilton not understanding
Hamilton's concepts is not a statement just about Hamilton.  It applies
to all theoretical population geneticists!

[So I said:]
>> McGinn's objection is not specifically to anything
>> Hamilton did.  It is to all results derived from models
>> in theoretical population genetics, all of them, including
>> Hardy-Weinberg proportions.  So why don't we discuss
>> Hard-Weinberg proportions?
>
[McGinn:]
>Go for it.

If I list a set of assumptions and show that Hardy-Weinberg
proportions spring from them (and that is easy with a simple
discrete-generations random mating model), does that mean that
McGinn is either going to agree, or is going to show that these
assumptions do not in fact lead to those conclusions?

If so, I'd be happy to make the argument.

Or is McGinn just going to demand that I show that these
assumptions have "causal validity"?  i.e., that those assumptions
are true in real life?  Of course the model will at best be an
approximation to real life.  All theoretical population genetics
models are.  They therefore won't be literally true in real life.
The models can be made to be better and better approximations to
real life, but will always be loftily dismissed by McGinn.

Is there a single model, anywhere in theoretical population
genetics, that McGinn thinks has this "causal validity"?  (If so,
then perhaps we are misunderstanding McGinn's argument.)

If there is not a model that he likes, then all he is doing is
demanding that our arguments be exact descriptions of real life.
In which case he is off by himself (as no one else here seems to
agree with his arguments).  And in opposition to the entire
development of theoretical population genetics over the last
century.

And of course, I repeat, McGinn's argument has nothing specific
to do with Hamilton, or Hamilton's relatedness quantities, or
Hamilton's argument, or Hamilton's alleged misunderstandings, or
Hamilton's alleged mistakes.  McGinn's objections apply with equal
force or, more properly, lack of force, to Hardy and Weinberg.

So we're waiting to hear McGinn tell us what model in theoretical
population genetics he would approve of.  And waiting.  And waiting ...

-- 
Joe Felsenstein         joe@removethispart.gs.washington.edu
 Department of Genome Sciences and Department of Biology,
 University of Washington, Box 357730, Seattle, WA 98195-7730 USA
---
ū RIMEGate(tm)/RGXPost V1.14 at BBSWORLD * Info@bbsworld.com

---
 * RIMEGate(tm)V10.2á˙* RelayNet(tm) NNTP Gateway * MoonDog BBS
 * RgateImp.MoonDog.BBS at 12/27/04 12:58:13 PM
 * Origin: MoonDog BBS, Brooklyn,NY, 718 692-2498, 1:278/230 (1:278/230)