Re: Theories, models, and simplifications
From: John Edser (edser_at_tpg.com.au)
Date: 01/31/05
- Next message: John Edser: "Re: Theories, models, and simplifications."
- Previous message: EKurtz99: "Re: Theories, models, and simplifications."
- Messages sorted by: [ date ] [ thread ]
Date: Sun, 30 Jan 2005 22:17:16 -0500 (EST)
"Perplexed in Peoria" <jimmenegay@sbcglobal.net> wrote:
> > JE:-
> > Note: Hamilton's Rule does not prohibit
> > _any_ number being applied to any of its
> > three variables including zero and negative
> > numbers, e.g. negative r which is just a
> > biological absurdity.
> JM:-
> I agree with you that there can be a distinction between
> what is mathematically permitted and what is biologically
> permitted. But you have to come up with a better example.
JE:-
This is the best example possible. This is because
Hamilton's model remains totally headless (does not have a
minimum of just one constant term or even just imply one).
All a 100% relative model needs to provide to become
refutable is to prohibit just one single value for just
one variable employed by the model. None at all are
prohibited or can be prohibited within Hamilton's
Rule even if IBD is replaced by a mean relatedness.
Do you agree or disagree that the rule was and remains
at 100% relative. Please (finally) provide an answer
so that discussion can progress.
> JM:-
> If we are talking about r as the probability of identical
> descent, then a negative r is absurd both mathematically
> and biologically. Negative probabilities make no sense
> mathematically.
JE:-
Yes. Only NAS has said so publicly. To
my knowledge Felsenstein still refuses to
comment. Why don't you ask him to do so?
My argument: IBD still remains _absolutely_
required to establish Hamilton's gene as
a valid independent selectable parent which can
force Darwinian organism fitness altruism, simply
remains ignored. If no independent gene
level of selection exists because r is just
a comparison to a mean value then please explain how
rb could ever be selected over c when Hamilton's
gene is being counted over organism generations
and not gene generations?
> JM:-
> On the other hand, if we are talking about r as the
> regression coefficient of the 1970 paper, then negative
> r can make sense both mathematically and biologically.
> See the paper itself for the details.
JE:-
Yes, it is quite clear that NAS can mathematically derive
the rule using r as the mean relatedness of one population
where -r becomes a valid deviation from the mean. Of course
this mean relatedness in rb still requires enormous wastage that
normal reproduction never needs to pay. If you couple
these ignored costs (Felsenstein calculated one example
to be 50%!) to Neo Darwinian incompetence re: the
common confusion as to what the maximal relatedness
allowed for via any proxy reproduction within Hamilton's
Rule is, (it is 0.25) does the level of Neo Darwinian
misrepresentation now become apparent.
Maximal proxy relatedness is commonly allowed to
be 0.5 within Hamilton's Rule. This is not correct.
It is only possible within Haldane's Pub Rule
which is NOT the same rule. This common error makes
an enormous difference:
______________________________________________________
When you factor in the _empirical_ fact that all gene
fitnesses must be assumed to be epistatic until at
least one non epistatic gene fitness becomes documented
within nature, the maximal relatedness possible within
any proxy reproduction is r^e = 0.25^2 = 0.0625.
This means the best that rb can come up with
is 16 proxy reproductions for each natural reproduction
even if zero wastage occurs (which it never does)!
The rule cannot work even as just a 100% relative rule.
______________________________________________________
Either: Neo Darwinians are just incompetent,
deceitful or worst of all, they are both. No
other way exists to explain how such a gross
distortion and utter misuse of a model could
happen in the name of the biological sciences.
> > JE:-
> > Therefore
> > any refutable test of the rule is just a statement
> > from any authority on the subject, e.g. Felsenstein,
> > as to which number for which variable is prohibited
> > by the rule.
> JM:-
> You seem to be assuming that a restriction on the range
> of one of the variables *by itself* is the only possible
> point of refutation for a model or theory. However,
> in most theories, the point of refutation is a restriction
> on some combination of variables taken together.
IF, like Hamilton's Rule, you only create
a headless model THEN only the one value of just one
variable needs to be prohibited. If say, two constants
existed within the parent theory and the process of over
simplification deleted one of them so that one remained
then a combination of variables would now become necessary.
>snip<
> JM:-
> An attempt to refute the rule using only three variables,
> under the adaptationist assumption that nature is already
> perfect (and that hence rb>c should be true for ANY social
> behavior, because the gene has already increased in frequency
> to fixation) wouldn't work. It is quite possible that the
> mutation that eliminates behaviors with rb<c just hasn't
> come along yet. So THAT test would tend to refute
> adaptationism rather than refuting Hamilton.
JE:-
Adaptionist nature has *NEVER* assumed that
"nature is already perfect" it only assumes
that biological forms continually attempt to
maximally reproduce themselves. It is that simple.
This event is measured as a maximand Darwinian
fitness assumed to be heritable to some (unknown)
extent. ANY maximand fitness can only be selected
to increase and not decrease.
Perfection at the gene level, i.e. one gene
increasing to fixation is *NOT* "perfection"
at the organism level because all organism
level fitnesses require epistatic gene
fitnesses and not non epistatic gene fitnesses.
Not one single non epistatic gene fitness has
ever been documented within nature. Hamilton's
rb fitness is just a heuristic but the cost
c is not, it remains empirical. It is absurd
to compare a heuristic to empirical reality
in order to measure how something can be
naturally selected FOR within *NATURE*.
> > JE:-
> > Until this entirely missing point of
> > refutation is supplied forthwith, an _irrefutable_ rb
> > cannot validly contest a _refutable_ c within the rule.
> > Either agree or throw out Popper. You cannot have it
> > both ways.
> JM:_
> No comment, since I have supplied a point of refutation.
JE:-
No point of refutation for Hamilton's Rule has
been made by yourself just, more non verifications.
Please provide one prohibited value
for r, b or c. If you cannot do so them you
are required to say that this is the case
or just throw out Popper. You cannot
have it both ways.
> JM:-
> But that doesn't necessarily mean that I agree with (or
> even understand) some of the assumptions you seem to be
> making here. Also, I wish you had omitted that "forthwith".
> The people you converse with here on sbe are not robots
> or lackeys that you can order around. Stop making demands
> of people. I demand that you stop doing this.
JE:-
Yet again, you seem not to wish to understand
that the scientific method is making this demand.
I am simply articulating it. Please stop
shooting the messenger. Your amusing
comment points out the absolute self
contradiction of suggesting to anybody
that they must absolutely stop making
demands. ANY enquiry MUST make DEMANDS
to be an enquiry, no exceptions. It
appears all anybody wants here is face
saving whitewash.
Hoelzer has stated that b/c is a constant.
Do you agree/disagree? If you agree:
WHAT IS THIS MISSING CONSTANT? WHY WILL
NOBODY HERE EXCEPT MYSELF DEFINE IT?
Regards,
John Edser
Independent Researcher
PO Box 266
Church Pt
NSW 2105
Australia
edser@tpg.com.au
- Next message: John Edser: "Re: Theories, models, and simplifications."
- Previous message: EKurtz99: "Re: Theories, models, and simplifications."
- Messages sorted by: [ date ] [ thread ]
Relevant Pages
|
