Re: Theories, models, and simplifications.

From: Perplexed in Peoria (jimmenegay_at_sbcglobal.net)
Date: 01/26/05 

Date: Wed, 26 Jan 2005 11:41:13 -0500 (EST)

"John Edser" <edser@tpg.com.au> wrote in message news:ct4l3o$1gh0$1@darwin.ediacara.org...

> JE:-
> Jim, do you realise that Felsenstein can
> derive Hamilton's Rule from classical
> group selection. Ask him yourself. Now
> what does that tell you?

It tells me that Hamilton's rule, in the 1970 version, at least,
is a tautology. It is a mathematical relationship that must
be true for any real or fictitious social behaviors as long
as you get the fitness accounting right.

It surprises me not at all that if someone makes artificial
assumptions regarding patterns of behavior and population
structure, and then shows in their model that those behaviors
are favored by selection, then they get back Hamilton's
rule when they try to figure out what "r" is in their
situation.

Hamilton's 1970 model is mathematics, not biology. I made
exactly this point in my response to NAS's submission
for McGinn's $10,000.

However, the theory of kin selection is biology, not mathematics.
Hamilton claims, in his biological theory, that there are no
biological factors in operation that make the math inapplicable.
For green beard altruism - another theory that might be suggested
by the math - there clearly ARE biological factors in operation
that leave the math correct, but make it inapplicable to biology.

A similar situation exists regarding Darwin. The math of NS
(Fisher's math, say, or the replicator equation) is tautological.
But Darwin's theory of evolution says more than that the math
of NS works as math. It also says that the math works as applied
to biology - that selection is the source of the adaptation that
we see, and that there is a sufficient source of new variation
that NS can, over time, construct remarkable adaptations.

So, if you take Darwin's theory to be that NS works as math, then
I would have to say that it is an irrefutable theory - a tautology.
Same goes for Hamilton's rule, considered as math.

If you take Darwin's theory to be that NS is the sole cause of
change in evolution, then I would have to say that Darwin is not
only refutable, he is quite clearly refuted - many times. Same
goes for Hamilton's theory, if you take it to be that kin selection
is the only cause of altruism in nature.

But if you take the theories of Hamilton and Darwin to be that
the processes that they describe are important in the explanation
of altruism or adaptation, then I'm not sure that Popper even
applies to the situation. How do you refute a claim that something
is "important"?

I'm really curious to see how it is that your definition of fitness
(and no other) makes NS both valid and refutable. I would be able
to figure it out from your description of the refuting experiment
if only I could understand how you distinguish a selected reduction
in fitness from a non-selected one.

Relevant Pages