Darwin, Kant, and Hamilton



J

Jim Menegay

Guest
For a young person growing up in bourgeois Western society,
one of the vicissitudes of the process of socialization is
that perfectly natural self-serving behaviors will be met
with the query: "What if everyone did that?".

It is a difficult question to even understand, let alone
answer. I have always been partial to Dunbar's response in
Heller's "Catch 22": "Well, if everyone did it, then I would
be a damned fool to do any differently!"

I have heard that this query can be blamed, ultimately,
on Emmanuel Kant. (If I have heard wrong, don't expect me
to actually READ Kant to educate myself - I have recently
gone thru the ordeal of reading Hamilton, and that was
bad enough.)

It has recently occurred to me that in Darwin's Natural
Selection, Nature never asks Kant's question. Organisms are
expected to behave self-servingly, even though "if everyone
did that" the self is ill-served. Ironically, Nature sees to
it that everyone does do that, though She doesn't take the
inevitable consequences of Her own actions into account. As
a result, every individual may suffer from an evolutionary
change that was to the benefit of each individual. To claim
that this is paradoxical is to commit the fallacy of
composition, as that fallacy is defined by economists.

I has also recently occurred to me that in Hamilton's kin
selection model, Nature DOES ask a watered down version of
Kant's question. In Hamilton's model, Nature asks the
question: "What if everyone whose genome is correlated with
yours did that?". Or better, "What if everyone did that to
the extent that their genome is correlated with yours?".

Please note the use of the words "correlated with", rather
than "similar to". I have just been thru this issue with
Hoelzer and McGinn. If everyone is similar, then no one is
correlated.

Food for thought, even if Dr. Wilkins will disapprove of the
anthropomorphism.
 
Jim Menegay <[email protected]> wrote:

> For a young person growing up in bourgeois Western
> society, one of the vicissitudes of the process of
> socialization is that perfectly natural self-serving
> behaviors will be met with the query: "What if everyone
> did that?".
>
> It is a difficult question to even understand, let alone
> answer. I have always been partial to Dunbar's response in
> Heller's "Catch 22": "Well, if everyone did it, then I
> would be a damned fool to do any differently!"
>
> I have heard that this query can be blamed, ultimately, on
> Emmanuel Kant. (If I have heard wrong, don't expect me to
> actually READ Kant to educate myself - I have recently
> gone thru the ordeal of reading Hamilton, and that was bad
> enough.)

Kant' proposed a "rational test" of a moral claim - did it
match what he called the categorical imperative? That is,
could it be said that you acted this way in the knowledge
that if all acted that way, the result would be good. Your
precis is good enough.
>
> It has recently occurred to me that in Darwin's Natural
> Selection, Nature never asks Kant's question. Organisms
> are expected to behave self-servingly, even though "if
> everyone did that" the self is ill-served. Ironically,
> Nature sees to it that everyone does do that, though She
> doesn't take the inevitable consequences of Her own
> actions into account. As a result, every individual may
> suffer from an evolutionary change that was to the benefit
> of each individual. To claim that this is paradoxical is
> to commit the fallacy of composition, as that fallacy is
> defined by economists.
>
> I has also recently occurred to me that in Hamilton's kin
> selection model, Nature DOES ask a watered down version of
> Kant's question. In Hamilton's model, Nature asks the
> question: "What if everyone whose genome is correlated
> with yours did that?". Or better, "What if everyone did
> that to the extent that their genome is correlated with
> yours?".
>
> Please note the use of the words "correlated with", rather
> than "similar to". I have just been thru this issue with
> Hoelzer and McGinn. If everyone is similar, then no one is
> correlated.
>
> Food for thought, even if Dr. Wilkins will disapprove of
> the anthropomorphism.

I'm reading _The Narrow Roads of Gene Land vol I_ right now,
and the way WDH tells it, there is no danger of
anthropomorphising at all :) He even discusses the use of
words like "altruism" and the misunderstandings this cas
caused, and so realises that without scare quotes they are
dangerous words...

Nice metaphor/analogy. I'll steal it at an appropriate time,
of course.
--
Dr John Wilkins [email protected]
http://wilkins.id.au "Men mark it when they hit, but do not
mark it when they miss"
- Francis
Bacon
 
Hi Jim,

in article [email protected], Jim Menegay at
[email protected] wrote on 6/24/04 10:28 PM:

> For a young person growing up in bourgeois Western
> society, one of the vicissitudes of the process of
> socialization is that perfectly natural self-serving
> behaviors will be met with the query: "What if everyone
> did that?".
>
> It is a difficult question to even understand, let alone
> answer. I have always been partial to Dunbar's response in
> Heller's "Catch 22": "Well, if everyone did it, then I
> would be a damned fool to do any differently!"
>
> I have heard that this query can be blamed, ultimately, on
> Emmanuel Kant. (If I have heard wrong, don't expect me to
> actually READ Kant to educate myself - I have recently
> gone thru the ordeal of reading Hamilton, and that was bad
> enough.)
>
> It has recently occurred to me that in Darwin's Natural
> Selection, Nature never asks Kant's question. Organisms
> are expected to behave self-servingly, even though "if
> everyone did that" the self is ill-served. Ironically,
> Nature sees to it that everyone does do that, though She
> doesn't take the inevitable consequences of Her own
> actions into account. As a result, every individual may
> suffer from an evolutionary change that was to the benefit
> of each individual. To claim that this is paradoxical is
> to commit the fallacy of composition, as that fallacy is
> defined by economists.
>
> I has also recently occurred to me that in Hamilton's kin
> selection model, Nature DOES ask a watered down version of
> Kant's question. In Hamilton's model, Nature asks the
> question: "What if everyone whose genome is correlated
> with yours did that?". Or better, "What if everyone did
> that to the extent that their genome is correlated with
> yours?".
>
> Please note the use of the words "correlated with", rather
> than "similar to". I have just been thru this issue with
> Hoelzer and McGinn. If everyone is similar, then no one is
> correlated.

And I think your argument is a good one, although I also
think it succeeds by dropping some of the most cherished
notions about KIN selection held by its more fervent
advocates.

> Food for thought, even if Dr. Wilkins will disapprove of
> the anthropomorphism.

This is indeed an interesting idea. However, I think you
might be misreading Hamilton's model a bit. The model does
not have Nature asking a question. It does not even posit
that Nature cares about the future. Instead, Hamilton's
model illustrates a mechanistic way in which Nature tends to
blindly explore the population level consequences of
mutations that affect fitness-altering social behavior. In
my view, this was the genius of Hamilton's model. He was the
first AFAIK to model a dynamical evolutionary system that
included a positive feedback loop, thus potentiating an
engine of evolutionary change. Viewed in this way, the model
of kin selection shares fundamental similarities with models
for the evolution of any signaling system, including the
coevolution of mutualistic species interactions.

More food for thought (I hope),

Guy
 
[email protected] (Jim Menegay) wrote in message news:<[email protected]>...
> For a young person growing up in bourgeois Western
> society, one of the vicissitudes of the process of
> socialization is that perfectly natural self-serving
> behaviors will be met with the query: "What if everyone
> did that?".
>
> It is a difficult question to even understand, let alone
> answer. I have always been partial to Dunbar's response in
> Heller's "Catch 22": "Well, if everyone did it, then I
> would be a damned fool to do any differently!"
>
> I have heard that this query can be blamed, ultimately, on
> Emmanuel Kant. (If I have heard wrong, don't expect me to
> actually READ Kant to educate myself - I have recently
> gone thru the ordeal of reading Hamilton, and that was bad
> enough.)
>
> It has recently occurred to me that in Darwin's Natural
> Selection, Nature never asks Kant's question. Organisms
> are expected to behave self-servingly, even though "if
> everyone did that" the self is ill-served. Ironically,
> Nature sees to it that everyone does do that, though She
> doesn't take the inevitable consequences of Her own
> actions into account. As a result, every individual may
> suffer from an evolutionary change that was to the benefit
> of each individual. To claim that this is paradoxical is
> to commit the fallacy of composition, as that fallacy is
> defined by economists.
>
> I has also recently occurred to me that in Hamilton's kin
> selection model, Nature DOES ask a watered down version of
> Kant's question. In Hamilton's model, Nature asks the
> question: "What if everyone whose genome is correlated
> with yours did that?". Or better, "What if everyone did
> that to the extent that their genome is correlated with
> yours?".
>
> Please note the use of the words "correlated with", rather
> than "similar to". I have just been thru this issue with
> Hoelzer and McGinn. If everyone is similar, then no one is
> correlated.
>
> Food for thought, even if Dr. Wilkins will disapprove of
> the anthropomorphism.

This is essentially the Tragedy of the Commons, as outlined
by Hardin in 1968, and LLoyd George a long while before
that. And yes, it can be solved by relatedness, for exactly
the reason you mentioned. Well done!
 
"Guy Hoelzer" <[email protected]> wrote in message
news:[email protected]...
> [Hamilton] was the first AFAIK to model a dynamical
> evolutionary system that included a positive feedback
> loop, thus potentiating an engine of evolutionary change.
> Viewed in this way, the model of kin selection shares
> fundamental similarities with models for the evolution of
> any signaling system, including the coevolution of
> mutualistic species interactions.
>
> More food for thought (I hope),

Unfortunately, there are also some fundamental differences.
If you look at the literature that grew out of Hamilton and
Axelrod's (1980?) paper, you will see that there are
enormous complexities in even figuring out whether one
strategy is superior to another. The space of possible
strategies is infinite-dimensional. And if you take a look
at the Hofbauer and Sigmund book, you will see that there
are huge difficulties in understanding the dynamics, even if
you are offered a small finite collection of strategies "on
a platter".

Agreed that signalling and mutualism are the next stage. But
I have no illusions that these phenomena will be understood
in the sense of being rigorously modeled any time soon.
Nonlinearity makes everything much, much more difficult.
AFAIK Prigogine's school has still not extended the minimum
entropy production theorem to systems that are far from
equilibrium. And they are working with effectively infinite
populations with a fixed dynamics. The evolution of
signaling strikes me as considerably more difficult still
because the invention of new ways to cheat is a
nondeterministic source of continuing perturbations.

Hamilton's rule for unilateral altruism was "easy".
Reciprocity and signaling are going to be very hard. IMHO.
 
Guy Hoelzer <[email protected]> wrote:

> GH:- Greetings

JE:- Thanks for your reply.

> >> JM:- I have always been partial to Dunbar's response in
> >> Heller's "Catch 22": "Well, if everyone did it, then I
> >> would be a damned fool to do any differently!" I have
> >> heard that this query can be blamed, ultimately, on
> >> Emmanuel Kant. (If I have heard wrong, don't expect me
> >> to actually READ Kant to educate myself - I have
> >> recently gone thru the ordeal of reading Hamilton, and
> >> that was bad enough.) It has recently occurred to me
> >> that in Darwin's Natural Selection, Nature never asks
> >> Kant's question. Organisms are expected to behave self-
> >> servingly, even though "if everyone did that" the self
> >> is ill-served. Ironically, Nature sees to it that
> >> everyone does do that, though She doesn't take the
> >> inevitable consequences of Her own actions into
> >> account. As a result, every individual may suffer from
> >> an evolutionary change that was to the benefit of each
> >> individual. To claim that this is paradoxical is to
> >> commit the fallacy of composition, as that fallacy is
> >> defined by economists.

> > JE:- The fallacy that is common to Kant, Hamilton and
> > "economists" is:
> > _____________________________________
> > A relative gain for an absolute cost where the relative
> > gain < absolute cost.
> > _____________________________________
> >
> > "If everyone did it" but it only produced a relative
> > gain < absolute cost, then everyone _loses_ including
> > the relative winner.

> GH:- The meaning of "if everyone did it" escapes me
> here,..

JE:- The meaning of "everyone" IS unambiguous: it means all
concerned, no exceptions. So, why did you write that it
"escapes me here"?

> GH:- but I think the issue of relative vs. absolute
> fitness costs is an important one to resolve. Let's take a
> closer look at the relative fitness model when relative
> fitness gains come at the expense of absolute fitness. A
> classic example from the

> subsumes the relative fitness model. The Fisherian model
> shows that a mutation, say for very attractive epigamic
> traits at the expense of gamete viability or production,
> will increase in frequency in a population if it reduces
> the number of offspring produced by others more than it
> reduces the number of offspring produced by the mutant.
> So, the population size is declining as the attractive/low
> fecundity allele increases in frequency. This process of
> selection MIGHT lead the population to extinction, but it
> MIGHT not. The population size at the new state attractor
> of the system would depend strongly on the relative
> fitness outcomes of interactions among mutants, which
> could conceivably even be synergistic. In that case the
> population size would be drawn to higher numbers as the
> attractive/low fecundity allele approaches fixation.

JE:- Your argument refutes the position that a relative
increase can win over an absolute decrease. It can't. In the
situation where "the relative fitness outcomes of
interactions among mutants which could conceivably even be
synergistic" is just a case where absolute fitness INCREASED
because the relationship had become MUTUALISTIC.

> GH:- When an altruism mutation causes relative fitness to
> increase (rb>c) and absolute fitness to decrease, the the
> "winner" wins relatively. This is true whether selection
> is pushing the population to extinction or not.

JE:- To my knowledge, nobody has ever suggested, including
myself, that it was not true that: "When an altruism
mutation causes relative fitness to increase (rb>c) and
absolute fitness to decrease, then the "winner" wins
relatively." What REMAINS to be debated is: can just a
relative win, at just one point in time be sufficient to
determine that the gene has spread? The answer is NO simply
because Hamilton's rule cannot discriminate between a
reduction of altruism (+c) which requires just a relative
measure and mutualism (-c) which requires an absolute
measure. Mutualism and altruism stand in absolute
contradiction to each other.

--------------quote----------------------

1) 22/01/2004:

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 -------------------- The rule cannot
discriminate between a reduced +c and -c because no general
term exists within it that represents the total Darwinian
fitness of the actor. Without it organism fitness altruism
(OFM) cannot be separated from organism fitness mutualism
(OFM) within the rule proving the rule is biologically
unintelligible.

Darwinian fitness is only implied as the negative value cmax
(the maximum cost c to the actor). This value does not
appear anywhere within the rule as a general term so
altruism can never be established. Hamilton's rule cannot
know if just a relative win at just one point in time
produces an absolute cost > or < relative win to recipients
so the rule cannot determine if the altruistic gene is
moving towards extinction or not! If an accountant audited a
companies books where that company only indicated that it
made a big profit for one week but omitted to say that it
made a massive loss over that financial year, (relative gain
at an absolute loss) or the company decided to enter debits
as credits because the accounting rule allowed it (the rule
could not discriminate between a reduction in non profitable
losses and mutualistic (traded) gains, then that companies
books would be deemed fraudulent.

> > JE:- This is the logic of war which is explicit within
> > Van Valen's incomplete Red Queen hypothesis. Nature does
> > not use such a logic because it leads to mutualised
> > extinction.

> GH:- John - If I didn't know better I'd swear you were a
> group selectionist at heart. How, in your view of
> evolution, could individual selection avoid driving
> populations to extinction sometimes?

JE:- Of course populations can become extinct no matter how
hard each individual within it attempts to increase its own
Darwinian fitness. A population becomes extinct simply
because all of the individuals within it make a continual
absolute fitness loss. Why? because none of them are up to
the Darwinian task within that environment (possibly because
they employed Hamilton's faulty logic but I doubt if nature
is as stupid as we are). My point is a VERY simple one:
Extinction cannot be SELECTED FOR within any rational
evolutionary theory. At the Darwinian level of selection
this means exactly: No Darwinian selectee can be selected to
reduce its own Darwinian fitness,
i.e. it cannot be selected to reduce the total number of
fertile forms it reproduces into one population.
However it MUST do this to prove OFA. This is why OFA
refutes Darwinism.

> GH:- If selection acted exclusively at the individual
> level, then individuals should not act "for the good of
> the species", right?If the logic of natural selection
> biases evolution toward an adaptive response to
> extinction, then natural selection must have acted on the
> population level.

JE:- False. Natural selection produced organism fitness
mutualism which can be _erroneously_ interpreted as a group
selective effect.

Regards,

John Edser Independent Researcher PO Box 266 Church Pt NSW
2105 Australia

[email protected]