in article
[email protected], John Edser at
[email protected]
wrote on 2/25/04 9:04 AM:
>>>> JE:- I _was_ making a _valid_ and specific point re: group selection which I shall repeat: "A
>>>> larger group is NOT necessarily naturally selected over a smaller group".
>
>>> GH:- This is definitely true, just as larger individual organisms don't always have higher
>>> fitnesses than smaller individual organisms. Indeed, the
> only
>>> trait that can be universally advantageous (i.e., not contingent upon environmental conditions)
>>> is higher fitness itself.
>
>>> JE:- I think it would be of enormous benefit to sbe readers if you could outline the logic as to
>>> why this is the case.
>
>> J:- You need to see the logic behind a tautology?
>
>> JE:- No tautology was proposed, AFAICS.
>
>> J:- How do you define fitness?
>
>> JE:- That is the unanswered question.
>
>> J:- Obviously it's whatever improves an organism's reproductive success.
>>
>> JE:- Reproducing exactly what?
>
> GH;- In the context of individual level selection, we are simply talking about old organisms
> producing new organisms that are much like themselves (i.e., with some heritability).
>
> JE:- I think you have to be more specific.
GH:- You asked what it was that reproduces and I answered with complete specificity in the context
of individual level selection.
> I defined reproduced organisms as only _fertile_ reproductions, i.e. I have specifically excluded
> infertile immatures as valid units of fitness. I suggest that when Darwin talked about parents
> raising their offspring, he was talking about the same thing; _fertile_ forms. Once genes were
> identified then surely this is just obvious because genes trapped within a sterile body cannot be
> selected for or against only using a sterile body.
GH:- Are you concerned that I may have implied a role for the reproduction of unfertile forms. That
would be a bizarre source of confusion that must come from an agenda that had nothing to do with my
answer above.
>> J:- Which is the same as having an advantage.
>
>> JE:- Fitness gain causes the advantage. Unless fitness can be exactly defined no advantage can be
>> measured.
>
> GH:- It is more logical for me to think of the advantageous trait causing a fitness gain, rather
> than the other way around. How could you ascribe fitness to an organism without considering all of
> its traits?
>
> JE:- I can only conclude you have reversed cause and effect reducing evolutionary theory to just a
> non testable status in the process.
>
> I agree you have to include all traits. This means each _individual_ trait's fitness must be
> _dependent_.
GH:- Traits can't really have fitness because they don't actually exist, except as abstractions.
Aside from this minor point, I agree that the fitness of an individual involves the integrated
affects of all its traits.
> JE:- What must now be defined is exactly what each trait is dependent on.
GH:- I presume you mean the fitness of each trait, which is a slippery concept for me, given the
abstract nature of traits. Still, I will go along with your logic for now. I don't see, however, why
this is a necessary next step.
> JE:- Remember, we are trying to provide the logic as to why a larger group is not necessarily
> fitter than a smaller group.
GH:- OK.
> JE:- We have concluded that a trait that expands a population when selected at the organism group
> level (but not at the individual level) does not necessarily provide a increase in fitness as
> measured at the _group_ level of selection.
GH:- This is coming out of the blue to me. I am certainly not part of your "we" that ever arrived at
such a conclusion. I argue that a response to selection at any level generically tends to increase
fitness at that level in a constant environment.
[snip]
> GH:- Regarding the meaning of fitness, I have posted several times on sbe that I view fitness as a
> useful concept that has no real existence.
>
> JE:- If this was the case then Darwinian evolution by natural selection was never a testable
> process and no valid scientific view of evolution exists.
GH:- That may well be the case in your conceptual framework for science, which I do not share.
> GH:- Fitness is not an entity or a process.
>
> JE:- Fitness is just an objective count of _something_ allowing the prediction that some
> biological entity will be selected for/against.
GH:- So then, you agree with me that fitness does not have any natural existence.
> GH;- It cannot have an exact and correct definition in the absence of something real that is being
> described.
>
> JE:- Yes, fitness must be able to be counted in reality, i.e. within _nature_ e.g. not just within
> a computer simulation.
GH:- I agree that operational definitions of fitness surely ought to be quantifiable.
> GH:- It can only have a vague definition, or an exact definition limited to use in a carefully
> proscribed context.
>
> JE:- No. Only _contesting_ and _exact_ definitions of fitness allow scientific research into
> evolution. Different propositions of fitness must be eliminated by the Popperian process of
> refutation until just one exists on the table.
GH:- I disagree. This is a completely worthless exercise for something like fitness, which has no
natural existence.
> GH:- Fitness is a concept that has often been defined in exact and concrete ways for the purpose
> of model building, but we should not confuse that useful procedure with the notion of fitness as
> reality.
>
> JE:- I agree that any over simplified model of fitness cannot validly contest and win against any
> theory that provided n objective measure of fitness in-the-first-place.
>
> Since you have already posted that "fitness as a useful concept that has no real existence" where
> this statement stands in stark _contradiction_ to what you subsequently wrote: "but we should not
> confuse that useful procedure with the notion of fitness as reality" then sbe reader's can have no
> idea what you mean by "fitness". How can such a contradictory notion of fitness allow you to
> validly conclude that a larger group is not necessarily fitter than a smaller group?
What contradiction? Those statements seem unambiguously consistent to me. To review: fitness does
not really exist, but it is a heuristically useful concept. Fitness should, and usually is,
operationally defined in concrete terms for the purpose of modeling. The diversity of selection
models has appropriately resulted in a plethora of concrete, operational, measurable definitions;
but none of these concrete definitions should blind us to the fact that fitness is merely a vague
conceptual tool in the general theory.
Cheers,
Guy