Re: Felsenstein is a liar
From: Jim McGinn (jimmcginn_at_yahoo.com)
Date: 02/22/05
- Next message: r norman: "Re: 50% mutation"
- Previous message: Jim McGinn: "Re: Labels and Obfuscations"
- In reply to: Jim McGinn: "Felsenstein is a liar"
- Next in thread: Josh Hayes: "Re: Jim doesn't like Felsenstein"
- Reply: Josh Hayes: "Re: Jim doesn't like Felsenstein"
- Reply: William Morse: "Re: Jim doesn't like Felsenstein"
- Messages sorted by: [ date ] [ thread ]
Date: Tue, 22 Feb 2005 13:27:22 -0500 (EST)
William Morse wrote:
> Jim McGinn wrote:
>
> >
> > Prediction: William Morse will not respond to this
> > post. And if he does the questions I asked him at
> > the end of the post will be ignored.
>
> I think this is a piece of an earlier follow. The first prediction
was a
> good one, and it might still be the better part of valor for me to
ignore
> the challenge, but what the heck:
I'm surprised by the lack of dismissiveness in your
tone. It's a good thing.
> > These are the questions:
>
> > 1) All other things being the same, are genes that
> > are IBD different from genes that are not IBD in
> > terms of their causative properties? Yes, No?
>
> No. In fact this is a point I have been arguing in another thread on
> "green beard" effects.
I'll take a look at it.
>
> > 2) If they are different in terms of their causative
> > properties what is the PROXIMATE mechanism by which
> > they produce this difference in causative properties?
> > I mean, do genes somehow know whether or not they are
> > IBD and act accordingly? Yes, No? How do they know
> > this?
>
> I agree with you on this point.
I was wondering if maybe somebody was going to hypothesize
(or maybe conjecture is the right term) DNA molecules that
practice the same dance steps. You know, those that
prefer the jitterbug prefer others that jitterbug, etc. :)
> > 3) If they are not different in terms of their
> > causative properties then why even bother to
> > distinguish between them?
>
> I wish I knew the history of the term Identical By Descent. In
another
> thread, NAS argues the IBD a la Malecot is different from Wright's
> relatedness. I have seen a discussion that includes a mathematical
> derivation indicating they are the same, which I have previously
posted a
> reference to. But to get back to the subject at hand.
>
> The important idea in Hamilton's inclusive fitness concept is the
> probability of an individual sharing an allele with another
individual.
I think this sentence is a throwback into the vagueness
and ambiguity of Hamilton's thinking. Nobody disputes
that sharing alleles and the probability thereof is an
important aspect of Hamilton's model. This is true for
all genetics based models. But to say that it is "the"
imporant idea is simply wrong.
> Let us say, for a particular gene A, that there are alleles A1, A2,
A3,
> A4, and A5. Only version A5 results in behavior that benefits others
at
> the expense of the individual.
The way you word this opens up more questions than it
answers. Specifically, how do we quantify the "benefits
to others " and the "expense to the individual."
Nevertheless I think I get the gist of what you are
saying: Only version A5 results in an organism that tends
to produce altruistic behavior.
> Given that we know that a benefactor is
> A5, what is the probability that the recipient is A5? Obviously, the
more
> close the relationship between the two, the more likely it is that
the
> recipient is A5.
No, this is not obvious in anyway whatsoever. There
could be many different factors involved with why A5
exists in the genepool. To suggest that they are
determined by relatedness (ie, your phrase, " . . .
the more close the relationship between the two, . . .")
requires a us assume a rather cartoonish understanding
of adaptation. (Look up the term preadaptation.)
The premise you are proposing here (the one gene
equals altruism premise) is a premise that has
been thoroughly discussed and rejected by none
other than myself here in this NG. If I remember
correctly I think it was discussed (and rejected)
just last December. And this wasn't the first
time. This premise always rears it's ugly head
whenever Hamilton's rule is discussed.
> Now we can discuss the question of the likelihood of
> sharing the allele A5 other than by descent, but that question is
> irrelevant to Hamilton's rule.
We're all supposedly supposed to take your word
for it? You must think I was born yesterday. I
don't think there's a chance in hell you'd get me
to accept this as a valid statement. And that's
because I know for a fact that it is wrong.
> The reason to consider relatedness in
> terms of Hamilton's rule is that it affects the likelihood of sharing
a
> particular allele A5.
As I stated above, it's one of many factors.
(Don't feel bad, Hamilton made the same mistake.)
> If you are with me so far,
Well, Bill, unfortunately I'm not with you. I'm
way ahead of you. As I stated previously you
really need to read a thread thoroughly before
you jump in the middle of it. Everything you
stated in these last few paragraphs has already
be considered and resoundingly rejected. If you
don't believe me then I can only suggest you do
a search through google groups.
> I will try to explain why anyone bothers with
> the expected prevalence of A5 alleles in a relative (i.e. the
prevalence
> due to IBD) vs. the expected prevalence of A5 alleles in the
population
> as a whole
Don't bother. We've already been over this
ground.
(BTW, If the answer to the first question is no then
the answer to the third question can only be no.
The only other option involves spiritualistic
understanding of cause and effect.)
> - but it isn't because of a difference in their causative
> properties.
Is it because of a differences in their
noncausative properties (sarcasm intended)?
Jim
- Next message: r norman: "Re: 50% mutation"
- Previous message: Jim McGinn: "Re: Labels and Obfuscations"
- In reply to: Jim McGinn: "Felsenstein is a liar"
- Next in thread: Josh Hayes: "Re: Jim doesn't like Felsenstein"
- Reply: Josh Hayes: "Re: Jim doesn't like Felsenstein"
- Reply: William Morse: "Re: Jim doesn't like Felsenstein"
- Messages sorted by: [ date ] [ thread ]
Relevant Pages
|
