Re: Theories, models, and simplifications
From: Perplexed in Peoria (jimmenegay_at_sbcglobal.net)
Date: 02/02/05
- Next message: Robert Karl Stonjek: "Article: RNAi Screens Seek Cancer Genes"
- Previous message: John Edser: "Re: Theories, models, and simplifications"
- In reply to: John Edser: "Re: Theories, models, and simplifications"
- Next in thread: Perplexed in Peoria: "Re: Theories, models, and simplifications"
- Messages sorted by: [ date ] [ thread ]
Date: Wed, 2 Feb 2005 02:24:41 -0500 (EST)
"John Edser" <edser@tpg.com.au> wrote in message news:ctolrj$2ik5$1@darwin.ediacara.org...
>
>
> > > JE:-
> > > Hoelzer has stated that b/c is a constant.
> > > Do you agree/disagree?
>
> > JM:-
> > John,
> > Among the many disagreeable features of your methodology, this tactic
> > is probably the most reprehensible. You take something that someone
> > has written, misinterpret it, and then demand that innocent bystanders
> > agree or disagree.
>
> JE:-
> Jim, you are sprouting rubbish to evade the question.
> You seem to think that yourself and other people here
> can say what they like and validly refuse to take
> any responsibility for what they have said. Evasion
> is not a part if the scientific method. All questions
> have to be answered and all terms defined. To my
> knowledge I have never knowingly refused to answer
> a question that makes logical sense. I have defined
> all my terms. OTOH you and others still refuse to define
> c and have the gall to suggest that I have a different
> understanding of b and c compared to everybody
> else.
But John, you DO have a different understanding! And you
prove it below where you respond to my question regarding
whether Hamilton's adaptive altruists have more children
than non-altruists.
> Please provide:-
>
> a) Your definition c.
> b) What you think was Hamilton's definition of c.
>
> Until you answer these two critical questions
> discussion cannot proceed in an orderly way.
Let me take these two in reverse order.
Hamilton never, so far as I know, used the variables "b"
and "c". (They were introduced by later commentators
on Hamilton - perhaps by Maynard Smith. Does anyone
in sbe know for sure who introduced these symbols?)
The symbol which (in 1964) Hamilton used for what we now
call "-c" was script-delta of "a". The script-delta represents
the difference between two alleles. "a" itself is used to
represent a portion of the donor's fitness - specifically
the fitness with all *reception* of social effects ignored.
That is, Hamilton meant for the script-delta of a to *exclude*
any differences between genotypes that might be associated
with (correlated with) different *reception* of altruism.
And on p44 of NROGL, he provides a little table or matrix
in which he distinguishes four cases, depending on the signs
of c and b. (The symbol he uses for "b" is script-delta of
T_nought). Interestingly, he uses the terms "altruistic" and
"selfish" as we might expect. However, he does not use the
terms "mutualistic" or "spiteful". Instead, he simply says
"selected" (for mutualistic) and "counter-selected" (for spiteful).
I agree with Hamilton's definition of c. So does the rest
of the neo-Darwinist establishment. You, on the other hand,
seem to be under the impression that "c" should represent
the quantity that would be represented by script-delta of
a_dot (in Hamilton's notation). Hamilton uses a_dot to
represent the total fitness of the organism - without subtracting
out its *reception* of social effects from others.
> _______________________________________________________
> As far as the specific question I asked is concerned
> all you have to do is:
>
> 1) Define b and c
Various definitions of b and c work with Hamilton's rule,
as long as you are consistent and use the same time scale
for both and the same fitness units for both.
The definition of b and c that was implicit in my "Peshawar"
example would be this:
"c" is the cost of behaving altruistically as measured in
units of Edserian fitness, and as calculated over the lifetime
of the donor. That is, it is the cost (counted in fertile children)
of behaving altruistically, with the assumption that any
"return altruism" still would happen even if the donor were not
to behave altruistically.
"b" is the benefit of being the recipient of altruism, as
measured in the same units and over the same period.
Neither "b" nor "c" can actually be measured for a real-world
organism, because the definitions involve a subtraction between
an actual measured fitness and a hypothetical measured fitness.
Nevertheless, using simple statistical procedures (i.e. partial
regression), we can provide good estimates for the average b
and average c in a population.
> 2) State if b/c is or is not a constant
> ACCORDING TO YOUR OWN DEFINITIONS OF
> b AND c.
It is definitely not a constant, as I understand the term
constant. "b" and "c" and their ratio differ from individual
to individual (using the "hypothetical" definition of b and c).
The average ratio or ratio of the averages may well vary between
sub-populations if you use partial regression to measure the
averages.
> If this is too much to ask of anybody then the
> scientific method cannot work here.
> _______________________________________________________
>
> >snip<
>
> > JM:-
> > 1. My claim that carriers of the "gene for altruism" have higher
> > Darwinian fitnesses than do non-carriers.
>
> JE:-
> Then the claim that the actor was altruistic in
> paying c for b resources transferred to a number
> recipients stands refuted.
No, the claim that Edser is using definitions of "b", "c",
and "altruism" that are the same as everyone else stands
refuted.
John, the theory of kin selection that you have been
attacking for four lonely years does not say what you
seem to think that it says. You have been tilting at
a windmill. You have been wasting your own, and
everyone else's time.
> If the Darwinian
> fitness of the actor increases and does not decrease
> then altruism stands refuted.
No, you simply have not appreciated the difference between
Hamilton's "a" and his "a_dot". You are trying to distinguish
altruism from mutualism using script-delta of a_dot. But
everyone else thinks that the distinction should be made
using the script-delta of a.
> The only possible explanation
> is that the act was mutualistic but the rule gave a
> false reading.
This makes no sense at all. The rule hasn't even been
used yet. We are talking about whether a particular
situation should be called "altruism" or "mutualism",
not about whether the behavior is likely to be favored
by selection.
> This is not surprising because the
> sign of c employed by the rule to separate altruism from
> mutualism remains entirely arbitrary. This is so because:
>
> -------------- quote ----------------------------
> sent Fri 28/01/2005 3:42
>
> JE:-
> What people seem desperate to evade is
> the fact that Hamilton's rule only
> takes an _arbitrary_ fitness sample from the
> donor's total fitness. The reason why all
> Neo Darwinian fitnesses remain ongoing is because the
> total fitness of the donar has been _artificially_
> deleted. This is also the
> reason why no defined time frame exists for
> measuring fitness within the rule and why
> the sign of c must remain arbitrary. The only
> _non_ arbitrary time frame that exists is the time
> taken for one parent to complete reproducing
> fertile forms into one population.
> ----------------------------------------------------
>
>
> The above was confirmed by O'Hara:
>
> --------------quote----------------------
>
> JE:-
> What is the difference between
> a reduced positive c and a negative c?
> If c was an abolute measure of fitness
> then yes, a real difference exists. However
> c is only a relative fitness cost and not
> an absolute fitness cost, so what is the
> difference?
>
> BOH:-
>
> As far as the rule is concerned, none.
>
> ----------- end quote --------------------
I still have no idea what you mean when you say that
"the sign of c is arbitrary". Perhaps there is some
meaning there that indicates that you understand the
distinction between a and a_dot, but you think that it
is Hamilton that is confused on the issue.
It doesn't much matter though. It is now clearly revealed
(to me at least) that the "Hamilton logic" that you have
been criticizing is not the logic that the real Hamilton
employed. You have been critiquing (perhaps correctly)
a garbled version of Hamilton that you have constructed
in your head.
No doubt, you will be reluctant to accept my diagnosis
here. Your understanding of what Hamilton said and mine
are clearly different. But your understanding of what
Hamilton said may still be in agreement with NAS's, BOH's,
etc. If your understanding is in agreement with theirs,
then I am dead wrong. But if my understanding is in
agreement with theirs, then there are only two possibilities:
1. You have been tilting at a windmill.
2. You are the only person who understands what Hamilton
really meant, and everyone else is defending a windmill.
The real Hamilton argument really is wrong and deserves
to be attacked.
> > JM:-
> > 2. My example of the orbit of Mercury as a counter-example to
> > your insistence that a refutable model must contain a constant.
> > (Incidentally, you seem to attribute this idea to Popper.
>
> JE:-
> What I attribute to Popper is that any theory
> must be refutable. The view that a constant term
> must exist for a view to be refutable is my own
> argument which can easily be proven by anybody who
> can reason their way out of a paper bag.
>
> Example:-
> If E=Mc^2 but c is just a variable then
> then the equation remains entirely arbitrary,
> i.e. it makes mathematical but not scientific
> sense. Only because c is a constant does
> Einstein's equation make any sense.
>
> When a constant does not appear then
> one must be referred to by implication
> or the view is just arbitrary.
> You have agreed that when a constant is
> deleted you zoom into a view like a
> microscope. This type of model
> allows you to see more clearly some
> innards of the theory but it cannot
> replace that theory. When the theory is
> applied to nature the view must be taken
> out from under the microscope so that
> the deleted constant reappears. Theories
> can only be validly applied to nature in
> their entirety.
But you haven't even addressed my counterexample to your
theorem. You have simply reiterated your "proof". Don't
you see that a counter-example must be countered on
its own terms? If it can't be countered, then there
must be a flaw in the "proof".
> > JM:-
> > I have
> > "The Logic of Scientific Discovery" handy now. Could you point out
> > where he discusses this idea? Or is it your own "improvement" of
> > Popper? Also, could you point out where he distinguishes refutations
> > from non-verifications? I apparently don't yet understand this
> > distinction.)
>
> JE:-
> It is just VERY BASIC.
Well, if YOU don't have a copy of Popper handy just now,
that is fine. But I really would like to have the
reference, when you can find it.
> A refutation is a prohibited observation of an
> theory but a non verification is an allowed
> observation from the same theory that was not
> observed.
OK, but when you requested a refutation of Hamilton's
rule, I provide you with a whole collection of
prohibited observations. You, dismissed this as
a non-verification. So, either you are changing
your definition of a non-verification (which I doubt)
or you didn't understand my suggested refutation.
Or else I still don't understand some subtlety here.
Here again is my proposal for a refutation of
Hamilton:
If it is observed that rb > c for a behavior and
the behavior is decreasing in frequency, then
Hamilton is refuted.
What is wrong with that?
> Cause and effect can be reversed in
> any theory. The reversal is called the anti-thesis.
> The anti-thesis provides points of refutation
> for the thesis. The sun centric and earth
> centric theories are contradictory. Only
> one of them can be correct. The constants
> used in one cannot be used in the other.
> Some observations of one must refute some
> observations of the other otherwise the
> views cannot be empirically separated.
>
> > JM:-
> > 3. My suggestion that r is the "missing constant" in Hamilton's
> > rule. I make this suggestion somewhat hesitantly, since my
> > understanding of what you mean by a "constant" is still new and
> > untested. But, based on what I understand of your thinking, it
> > seems to me that saying "r is the constant in Hamilton" makes at
> > least as much sense as saying "fitness is the constant in Darwin".
>
> JE:-
> Firstly is not "my" view of a constant
> it is what any constant is within the
> sciences.
Strongly disagree. You dismiss as incompetent the authors
of textbooks who call some things "exogenous variables"
that you would prefer to call a "constant". Then after
dismissing the majority of scientists as misguided,
incompetent, or just plain wrong, you have the nerve
to tell us all what the situation is "within the
sciences". John, you may well have some ideas for
worthwhile reforms in scientific methodology and
terminology. But don't try to draw the mantle of
authority around your positions and claim that they
are orthodox.
[snip stuff about John's disagreement with EK]
>
> Relatedness r cannot be a constant if r is
> just a probability or employs a statistical
> mean.
I respond to this in another post. So, I will
ignore it here.
Incidentally, I apologize for including #3 ("constant r")
and #4 (critique of THE EXPERIMENT) in my list of
things you have not responded to. You were already
in the process of responding to these two in different places.
Therefore, I will snip your responses here, and answer
the responses you have made elsewhere.
[snip remainder, except for this:]
> This will verify that random variation
> cannot provide evolution as Dr Moran
> insists that it can.
I ought to simply stand aside, and let Larry respond to this
himself. But I cannot resist pointing out that Moran
defines evolution as pretty much ANY change that takes
place. He has never suggested that random variation
can provide ADAPTIVE change. Perhaps you and he simply
differ on what kinds of changes should be included under
the umbrella of that "power word" - EVOLUTION.
- Next message: Robert Karl Stonjek: "Article: RNAi Screens Seek Cancer Genes"
- Previous message: John Edser: "Re: Theories, models, and simplifications"
- In reply to: John Edser: "Re: Theories, models, and simplifications"
- Next in thread: Perplexed in Peoria: "Re: Theories, models, and simplifications"
- Messages sorted by: [ date ] [ thread ]
Relevant Pages
|
