Re: Underestimating 'r'


Catherine Woodgold wrote:
> "Perplexed in Peoria" (jimmenegay@xxxxxxxxxxxxx) writes:
> > "Catherine Woodgold" <an588@xxxxxxxxxxxxxxxxxxx> wrote in message news:diq3mu$1irv$1@xxxxxxxxxxxxxxxxxxxxxx
> >> ... It turns out that if the rate of A
> >> in the population is high, then the best strategy
> >> for the Aa organism if it wants to promote the
> >> rate of A may be to commit suicide (or to
> >> stay alive only if by doing so it can cause
> >> a lot of harm to its relatives!)
> >
> > But there is no reason why an Aa organism would
> > 'want' to promote the rate of A. It has an
> > equal 'interest' in promoting the rate of a.
> ...
> > However, even people like me who LIKE teleological
> > language become somewhat nervous about the use of
> > 'want' and 'interests' in a case like this.
>
> Maybe some people "want" to use teleological language
> at certain times, and other people "want" to use it
> at other times. I happened to feel like using it
> right then, even though I figured somebody would
> probably object to it. How about just translating what I
> said into non-teleological language and assuming
> that's what I meant?

The trouble is that in a case like this, there is no
unique translation. The teleological language is ambiguous
as will be seen below.

> > In fact, if the frequency of A in the population
> > is high, the organism will be more interested in
> > promoting the success of the rarer allele a than
> > in promoting the success of the common allele A.
>
> I disagree with this statement. Can we make
> it more clearly stated and testable and then
> test it?

Well, make it more clearly stated so we can model it,
anyways.

> Do you mean this: If A is common and a is rare, then
> (averaged over many possible genes A and a) an organism
> with Aa is more likely to behave in a way that promotes
> the prevalence of a than it is to behave in a way that
> promote the prevalance of A.

Nope. That is not what I meant. Though your interpretation
is one possible interpretation. That is what I meant above
by saying that the teleological language is ambiguous.

What I meant was that when an Aa individual encounters an
aa individual, 'r' is slightly larger than it would be
when an AA individual is encountered. Two reasons for this:
1. The Aa individual is more 'a-ish' than the general
population, and less 'A-ish' than the general population.
So, if you see his goal as making the rest of the population
look more like himself, then his goal should be to increase
'a' in the population and decrease 'A'. In fact, if you
use the regression definition of 'r', it is clear that
you get a larger 'r' with aa as the recipient than with AA.
2. Since our focal individual has the rare gene 'a', it
is likely that his relatives do too. Thus, an aa individual
is more likely to be a relative than is an AA individual.

However, when you interpret it your way, then you are
no longer just dealing with motive - you have to consider
method and opportunity as well. And alibi. The rare
gene 'a' is probably innocent of receiving favoritism
in this population - if it did receive favoritism it
probably would not be rare.

But I disagree with several points in your analysis:

> I hereby argue that it's the other way around: if a is
> rare, it's more likely to be a recent mutation and
> to just behave in some random way. Since A is common,
> it must have gotten common somehow so it's more
> likely to behave in a way that promotes its own success.

Disagree. Most genes that become common do so without
promoting themselves. They become common because they
provide fitness to the organisms that carry them. (Hmmm.
Does that count as self-promotion?).

> A is more likely to have been tested in pairs with
> other genes -- that is, it's probably been
> paired with a lot of other genes in the
> past, yet it's still common. So A may tend to cause some
> behaviour in Aa that promotes A, while a might behave in a
> way that promotes either A or a.

Well, it is likely that 'A' has some effect beneficial to
the organisms that carry it. Whereas 'a' is possibly
detrimental.

But it is extremely unlikely that either 'a' or 'A' would
cause any kind of 'green beard' behavior. On the other
hand, it is quite possible that some combination of genes
in the total genome might cause altruism to individuals
with a large 'r' with respect to the donor.

At the heart of our disagreement is your apparent assumption
that the locus a/A has something to do with altruism. I was
assuming that it is some arbitrary locus in the genome.


.

Relevant Pages

  • Re: Underestimating r
    ... >> for the Aa organism if it wants to promote the ... If A is common and a is rare, ... (averaged over many possible genes A and a) ... with Aa is more likely to behave in a way that promotes ...
    (sci.bio.evolution)
  • Re: I wonder who was stupid enough to hack the masmforum server ?
    ... One thing that scientists have learned is that genes are surprisingly ... genes could jump from organism to organism, ... buy seeds and chemicals and implements but there is a monsoon failure ...
    (alt.lang.asm)
  • Re: Lottery tickets
    ... >> Are winning lottery tickets different from lottery ... > that are contained in the same organism, ... No one would claim that the non-IBD genes in the same ... content) as a kind of causation. ...
    (sci.bio.evolution)
  • RE: Some answers by Hamilton
    ... > on their own genes, or do they behave as if each of their genes is ... IF "the outcome is that the organism acts largely as a whole" ... so the simultaneous level of selection that Hamilton must be referring ... What kind of glue is this? ...
    (sci.bio.evolution)
  • Re: Sbe Peer Reviewed Papers
    ... Darwinian fitness (the total number of fertile ... Any of Haldane's genes has a 0.5 chance of ending ... and not the IBD relatedness of their offspring. ... supposed to cause organism fitness altruism. ...
    (sci.bio.evolution)