Re: The status of Infertile forms ( was: What is the evolutionary

in article erttdj$2pnj$1@xxxxxxxxxxxxxxxxxxx, John Edser at
edser@xxxxxxxxxxxxxx wrote on 2/25/07 10:06 PM:



Guy A Hoelzer hoelzer@xxxxxxx wrote:-


JE:-
But this remains the case on just logical grounds (see below).

My hesitation in agreement was based on examples like mules, where the
offspring have nothing to do with the parents' reproductive strategies
beyond possibly being a failed attempt to create a reproductive offspring.
Sometimes infertile offspring do not remain an integrated part of a fertile
form, which is why I did not fully agree with your statement that they
"must" remain so.

JE:
As previously indicated my "must" remains a proposition of logic which
confines choices to just two alternatives but does not force either choice.

This tactic can be useful in some instances and misleading/misguided in
others. In any case, it would be helpful for me, and I suspect many others
on sbe, if you would more clearly distinguish statements of formal
hypotheses from less formal arguments you make about the the world. Formal
hypotheses are challenges for refutation, to use your Popperian framework,
and they need not represent the way you think the world actually works. An
hypothesis does not require much justification on your part, but an argument
is all about justification.

If a sterile organism such as a mule cannot be selected to assist a fertile
form (preferably its own parents) for _whatever reason_, then its own
independent fitness must remain zero.

As an argument about reality then, you seem to agree that "must" is too
strong.

There are only two choices: be selected to be reduced to just a fitness sub
part (here the sterile form assumes a dependent nested fitness within a
fertile form) or remain a fitness nothing as far as nature is concerned.

I'm glad to see that you consider this an option.

I agree that a sterile form such as mule may not be able to be selected to
assist a fertile form (if it could do so the infertile form obtains for itself
a dependent nested fitness which is better than just a zero independent
fitness).

Of course, fitness (especially reproduction) may not be a goal for all
individuals, including mules.

However, this does not change these two choices: remain independent
with only a zero fitness or be reduced to a dependent part of some other
fertile form. It seems the only possible way nature can select for a nested
fitness is via hormones requiring an enclosed space, e.g. sterile eusocials.

JE:-
The "must" I used is only based on logic: genes within an infertile form can
only possess a zero fitness relative to an infertile body or any meta level
above this except for a fertile body no matter how well adapted they allow a
meta infertile form to become and irrespective of how much longer these
genes increase the lifespan of an infertile form.

This is IMHO badly flawed logic. It seems to be based on the false
assumption that every organism must have or contribute to reproductive
fitness in some way. You conclude that if they aren't going to reproduce
themselves, they must be play a role in another's (the parents) fitness.
This is simply not true. I already gave you a concrete example that
demonstrates the falsehood of your logic. Infertile mules rarely play any
role in the reproductive success of their parents.

JE:-
The fact "infertile mules rarely play any role in the reproductive success
of their parents: does not exclude the logic that I have detailed. Please
review the argument presented above. To refute it all that is required is
the provision of a 3rd alternative for the sterile form. To recap: I reason
that only two logical alternatives can empirically exist: a dependent nested
fitness within a fertile form or fitness independency which only provides a
zero fitness.

Your logic still escapes me, but I acknowledge that this does not mean your
model lacks logic.

You could explicitly assume that they do for the sake of a model, but that
doesn't make it so.

JE:-
I was not assuming a simplified/oversimplified model. What I carefully
articulated above represented an empirically refutable theory contingent on
my definition of fitness which can empirically and objectively separate
fitness dependency from fitness independency providing the required rigor to
underwrite the basic organism concept on which the science of biology
remains based.

I can only give you the benefit of the doubt on this claim.

JE:-
Why did you exclude selection OF the group in which each group fitness
is just the sum of the fitness of each fertile individual which
comprises any one group?


Because under this description there is no evidence that the group
actually exists as a possible unit of selection.

JE:-
But you failed in the past and have failed here to explain WHY this is the
case. Your agreement that "under this description there is no evidence that
the group actually exists as a possible unit of selection" without providing
any explanation as to why this is the case only confirms a belief. Surely
you realize that since you rejected my argument as to why, the validity of
your agreement as to why remains contingent on you providing an alternative
theory with worked examples (which you have not provided).

I have no such obligation, but I also think that I have provided far more of
what you request in past posts than you suggest. Perhaps you are not
reading or pulling together comments I have made in other threads. You
have, for example, indicated that you are as unsure of my views relating to
complex dynamics and emergence as I am of your paradigm (I use this word
because you have indicated that you think you have something more than a
model). My arguments about physical existentialism are contained in those
posts. To summarize, higher order entities/processes (e.g., you are an
entity/process; so is the biosphere/natural selection) emerge when top-down
influences (e.g., entrainment) organize the dynamics at lower levels of
organization. Your description of a group of organisms pointed to ways
that it lacked dynamical boundaries (even fuzzy ones), while failing to
indicated any way in which the dynamics might be bounded. I replied by
saying that "under this description there is no evidence that the group
actually exists as a possible unit of selection." I did not take the bait
of trying to argue about group selection in the absence of real groups only
to be "proven wrong" in the end because the groups didn't exist.

In my opinion, you will not be able to understand multilevel selection
theory until you understand the physics of emergence regarding higher order
agency.

The theory and the worked examples as to why that that I have provided here
(which you reject out of hand) remain empirically testable: ANY additive
fitness _proves fitness independence_ thereby refuting fitness dependence
(and vice versa). Here is my explanation as to why: addition only employs a
reversible set intersection whereas a NON additive fitness must assume a non
reversible set nesting which whenever it becomes reversed provides a
required refutation.

I agree with your points about reversibility, however they lead me to argue
that non-reversibility = independence and reversibility = dependence. I am
still unsure whether our difference here is trivially semantic or
substantive.

JE:-
But I have been posting the same argument to sbe for nearly 5 years now.
Previously you rejected it (I think the closest was about 3 months ago).

I have always argued that fitnesses at different levels of selection must
be independent, so I'm sure that I never posted anything consistent with
the claim that group selection can occur when group fitness was merely the
sum of its individual fitnesses. I have consistently argued against this
idea over the years on sbe and I challenge you to find anything I have
posted contrary to this.

Example:
To: Sci.bio.evolution moderation account <sbe@xxxxxxxxxxxxxxxxxxx>
Date: Jan 29 2006:

------------------------------ repost ---------------------------------

Guy Hoelzer hoelzer@xxxxxxx wrote Jan 29 2006

I agree that every level of selection, such as the individual organisms,
must have a phenotype (including the trait of fitness) that is a non-
additive result of single gene influences.

JE:- Do you agree that "the trait of fitness" as measured at the gene
level is always dependent on another level and is therefore not an
independent fitness?

No. I have tried to convey to you that I do not see dependence/independence
as a black and white issue.

JE:-
You did _convey_ this. The problem was is it made no sense to me. The reason
why is this: If you allow "dependence/independence" to be complimentary and
not contradictory (it cannot be both) it becomes impossible to distinguish
mono-centricity from poly-centricity.

Thank you for acknowledging that I did convey this and have been
consistent.

JE:-
Your consistency only appears to me to be that you prefer sitting on the
fence resolving nothing.

I do not feel any obligation to pretend that dependence/independence is
anything other than a continuum, nor do I see any value in doing so.

JE:-
If fitness dependence cannot be empirically differentiated from fitness
independence via SOME testable evolutionary theory then "anything goes"
simply because what remains the one same system and what constitutes a
Darwinian competing system to it, cannot be empirically differentiated.

I disagree. Continua are the things we measure. They are in fact the grist
of empiricism. We are slaves to the data when it comes to continua, so it
is hardly "anything goes". You might ask, how does one measure the degree
of dependence/independence? That would be a great question. I don't know
if there are any metrics developed for measuring this directly, but I don't
see any major obstacles in doing so. A simple approach might be to estimate
the fraction of interactions that occur among members of a hypothetical
group compared with the number of interactions between group members and
agents hypothetically on the outside. There may also be ways to measure
agency by estimating the coherence of an hypothetical group as it interacts
with its environment, including other agents that are part of its (the
group, not the group's members) environment. One could also estimate the
strength of top-down influence over the behavior of group members. All of
these approaches could constitute empirical tests of boundedness, agency,
and internal mechanism (metabolism) that would establish or discredit the
real existence of a group.

Assuming that dependence/independence is a binary quality is not useful
IMHO. I can only see this assumption taking us down a rabbit hole. We
would eventually emerge back into the light in the same place we started.
It would be a waste of time IMHO.

IMHO, no two things or processes in the
universe are 100% absolutely independent.

------------------------------------ end repost ------------------------
----

JE:-
Your final comment in the above repost: "no two things or processes in the
universe are 100% absolutely independent" is a get out of jail free card. It
allows you to only subjectively dictate what differentiates group fitness
interdependence from group fitness dependence without having to be
empirically accountable.

I disagree, and I think I have explained my reasons in sbe dialogues with
you before.

JE:-
You have not provided:

1) WHY you have rejected my argument out of hand.

I had explained WHY before and again here.

2) An alternative argument which can be tested empirically.

I think I have gone further in my argument about the empirical testability
of the existence of group agents here than I have in the past, although it
is false to say that I have not addressed this in the past. I doubt I have
convinced you of the validity of my view here, but I hope that you will
grant me the same benefit of the doubt as I grant you.

3) How a science of evolutionary theory can possibly exist if researchers
remain free to define fitness independence and fitness independence in any
way that they like.

If you didn't already know better, I hope you will acknowledge that this is
a misleading smear of my position. I think that I am thinking in a more
focused and determined way about fitness dependence/independence than most,
even if you and I disagree about how to do so.

Surely it is obvious that whenever fitness independent forms are not able to
be objectively differentiated from fitness dependent forms then an
evolutionary theory which remains testable against nature does not exist.

I don't agree at least in part because I'm not sure what you mean. I also
try to avoid using the rhetorically loaded word "obvious" in making
arguments because I think it typically undermines the substance of an
argument.

In such an unhappy situation only a bunch of beliefs exist which remain
subject to nothing which can be described as an empirical test against nature
allowing bias and inevitable political misuse.


The only possible reason why "under this description there is no evidence
that the group actually exists as a possible unit of selection" is because
additive fitness associations can only produce _fitness independent_
associations. Associations between what? You haven't said what you are
referring to.

JE:-
Just, any group of biological entities, e.g. any group of organisms, any
group of genes, any group of cells....

I am talking about independence of fitness at different levels of
organization. For example, individuals in a group could have high
fitnesses, but the group could have low fitness, and vice versa.

JE:- This remains empirically possible if the fitness of a group is not just
the simple sum of the fitness of each part. If most individuals in such a
group have a higher fitness then any additive group must also have a higher
fitness. However, if any group fitness remains NON additive then it can
become geared to become either much higher or much lower than the mean
individual fitness. The cost: a loss of fitness independence.

Every time I try to make the discussion more clear by emphasizing that group
fitness may not be a function of individual fitnesses at all you seem to
steer it back to the additive/non-additive issue. Can we try to avoid the
semantics about "dependence" and focus on a complete absence of a functional
relationship?

JE:-
I will continue to steer discussion towards and not away from *ANY*
empirical test simply because the epistemology of science requires me to do
so. I am happy to consider other empirically based alternatives as to WHY an
additive in fitness group cannot constitute one selectee. So far you have
not provided one.

to reiterate: without being able to _empirically_, i.e. objectively
differentiate between fitness dependency and fitness independency (which
includes fitness interdependency) using anybodies unambiguous theory, no
basis exists for the organism concept within Darwinism. Without this concept
the science of biology cannot exist. If you argue that it can, what are you
replacing it with?

I don't see the relevance of any of this to the point under discussion, but
then we are probably just talking past one another.

The fitness of a honey bee colony may critically depend on the number of
bees in the colony because it must be able to keep the queen warm on the
middle when it gets cold in the environment. Individual bees may die while
keeping the queen warm at night as a sign of a healthy colony. Similarly,
the fitness of a colony may suffer if there is too much individual
reproduction. The factors effecting function at these two levels of
organization may be utterly and qualitatively different, and I view fitness
as intimately connected with function. Everything I am talking about is
just as amenable to empiricism as your paradigm, as far as I can see.

This means that selection between additive in fitness parts must be
occurring before selection between groups possibly can.

This strikes me as rubbish, but then I may not understand your argument.

JE:- My hypothesis which I have often repeated here remains precise and
empirically refutable, i.e. remains rigorous. I define any additive in
fitness group to consist entirely of independent in fitness wholes which
_intersect_ their fitnesses allowing the largest fitness to reduce all the
other fitnesses to intersecting (reversible) subsets of itself. This
represents a testable logic of natural selection. This is the reason why I
state that selection between additive in fitness parts (which actually
constitute wholes) must be occurring before selection between groups of them
can. The facts of this matter is that any additive in fitness parts of a
population are not fitness dependent parts of one selectee but empirically
fitness independent wholes. Reversing this proposition is empirically
unsustainable.

As I think about levels of selection, it is clearly possible to have
selection at the group level without any selection occurring at the
individual level.

JE:-
Only when supposed as competition between non additive in fitness wholes
which empirically represent an individual. When employing selection via
competing additive in fitness wholes within an additive meta population,
selection has already operated on these "parts" (competing Darwinian
individuals) before it could possibly occur between populations of them. The
first additive in fitness level is just the fertile organism level. It
dominates every other level simply because of this fact. If you can provide
a lower additive in fitness level which is empirically based then please
provide one.

If selection remains equal for each Darwinian selectee then it has operated
before selection at the additive group level has, e.g. extinguishing one
additive in fitness population in favor of another group. This is routinely
but incorrectly employed to represent an act of group selection.

Selection at different levels are independent processes that need not
influence one another, although they may often influence one another in
nature.

JE:-
But this "influence" can employ two _contradictory_ pathways:

1) A non mutualised influence directed from above (using nested sets of
fitness): Here an enforced fitness loss at a lower nested set level provides
a fitness gain for the largest fitness set it remains within. A schematic
example: if A is entirely nested within B (not the reverse) then A CAN be
selected to sacrifice fitness for B simply because selection only occurs as
a contest between B's, i.e. competition remains entirely at the B level of
selection. Note: the opposing nested set fitness configuration provides an
empirically based refutation.


2) A mutualised exchange influence not imposed from above (using intersected
sets of fitness): Fitnesses reversibly intersects within an additive in
fitness proposition allowing those receiving a bad fitness deal to simply
leave and find a better one. A schematic example: If A only reversibly
intersects with B then selection between A and B operates before selection
between additive in fitness populations of either do. Selection between
groups constituting additive in fitness selectees can produce a grouped
selective force but not a grouped selectee. These groups can only act as one
(of many) selective forces on Darwinian individuals e.g. the Baldwin effect.
Note that V. C. Wynne-Edwards confirmed this in his final book on this
subject which occupied most of his life (this last book remains ignored).
Wynne-Edwards recanted only allowing group selection via groups but not on
groups (which of course is not group selection). These groups can act as a
selective force but not as a selectee.

I predict that an additive fitness association can never allow the selection
of a fitness loss in order to sustain a fitness gain at a meta additive
level. A non additive set nesting association represents a different kettle
of fish because any inner nested set remains stuck within the largest one so
it can only be selected via the fitness of the larger set. OTOH additive
fitness associations remain entirely reversible allowing those suffering a
reduced fitness to move somewhere else.

To which of the above do you refer?

I am referring to neither one. I am trying to steer the discussion to the
case of no influence.

JE:-
No influence? How is such a thing possible? It seems to me that everything
affects everything else given chaos theory where just one butterfly flutter
can trigger the weather to change. What is required from is an empirical way
to separate two entirely _contradictory_ effects as far as _fitness_ is
concerned: dependent and independent (which includes interdependent).

Here is a riddle for you. How can a human religious group that prohibits
reproduction and advocates suicide for its group members have high fitness at
the group level? [Remember - group fitness refers to the propensity for
persistence and spawning of daughter groups.]

JE:-
At best it can only have an entirely illusionary "high fitness at the group
level".

Fitness can be empirically estimated. I am not talking about anything
illusory.

At worst it remains obvious that such groups have a lower fitness
compared to others.

There's "obvious" again. I'm afraid that didn't persuade me at all. It
looks like you trying to avoid reason.

I know of no human group which "prohibits reproduction".
What human groups do is regulate this in some way. It seems clear to me that
group selection is not required here given the fact that mean total
productivity per individual increases as group size increases but at a cost
which has to be paid.

I made the riddle extreme to make my point clear. Of course, I am happy to
see that you think the suppression of reproduction is really a continuum of
effect.

Groups can evolve "suicide" but as a rare last resort mutualised group
defense mechanism without any group selection required, if and only if, the
costs on average remain less than the gains where both may be units of risk
(like an insurance company). I am happy to expand on this argument which
appears to me to remain neglected within evolutionary theory. Suicide as a
way of life cannot evolve for the same reason that cannibalism as a way of
life cannot evolve: the individual costs are greater than the individual
gains.

Strange you should mention cannibalism. In fact, it is a way of life in
many species. One of my dissertation chapters documented filial cannibalism
(eating of one's own offspring) as a customary way of life in a marine fish
species. I have also suggested in publication that one reason male-only
parental care may be so common among fishes is that fathers would often
have the option to fine tune the number of offspring they eat relative to
the number they raise to fledging. Maybe this was an empirical test that
refuted your paradigm. There are many studies by different investigators
working with different species confirming that filial cannibalism by fish
fathers is the norm.

Your definition of group fitness as "the propensity for persistence and
spawning of daughter groups" remains a forever ongoing proposition and
therefore non testable.

Would you say the same about all single celled organisms?

Group persistence can always be increased at the
expense of group reproduction and vice versa allowing a reduction of one to
be _explained away_ as some ongoing future increase for the other. For your
argument to make rational sense at some stage you will have to define
criteria which require either "propensity for persistence" to become more
important than a "spawning of daughter groups" or vice versa. It should be
self evident that group reproduction cannot be selected to become reduced in
order to increase group persistence simply because selection for increased
group persistence (group survival) cannot win against group reproduction.
What serves group selection the best is the same logic as to what serves
individual selection the best: lean and mean survival strategies which free
up limited resources in order to maximize reproduction.

Fine. I see no problem for group selection in any of this.

This in turn means group selection must always remain subject to
individual selection disallowing the evolution of organism fitness
altruism in nature via selection between additive in fitness groups.

JE:-
The above is deductive from the inductive inference that additive groups of
fitness cannot constitute one selectee, just the one selector (the
contradictory opposite proposition).

It suggests that the existence of the group as a potential target of
selection is a figment of our imagination.

JE:-
I argue that group selection between additive in fitness groups remains
100% subject to individual selection, i.e. it can act as an amplifier for
Darwinian selection acting at the fertile organism level. The common
argument that it can contest and win against the Darwinian fertile
organism level was and remains _entirely_ false.

It seems to me that what you are talking about has nothing to do with
multilevel selection. You are making a straw man by defining groups as the
sort of thing that selection does not grab hold of, then pointing out that
selection doesn't grab hold of them.

JE:-
No. I stated clearly and unambiguously that an additive in fitness group
cannot possibly represent a single selectee but it can constitute one (of
many) selectors. On the other hand a non additive in fitness group can and
does constitute one selectee, e.g. one group of somatic cells empirically
represents a single organism fitness. The empirical test: provide a
verification of just one organism which is the simple sum of the fitness of
any of its parts. None exist. Not a single additive somatic gene or somatic
cell fitness can be found. No additive organ fitness exists. No body part
fitness can be just added to another to provide an empirically based
additive fitness. The very first additive in fitness level is the organism
level which must become fertile to reproduce itself. This is why the
organism induction remains essential to the science of biology. I cannot see
how it is possible to make this critical but basic distinction any clearer.

Clarity may not be the problem. I just don't accept your premises. I have
tried to be clear about the problems I have seen.

JE:-
But my premises remain empirically based. What are you replacing them with
and why are you rejecting them out of hand?

Your assumptions are not empirically based, as far as I can tell. They may
have some logical justification that I don't understand, but these are the
ones I object to. For example, assuming that dependence/independence as a
binary factor has no empirical basis.

You should try considering the sort of groups that selection can get
traction with, such as groups with fitnesses that are either non-linear
functions of their individual's fitnesses or groups with fitnesses that are
not functions of their individual's fitnesses at all.

JE:-
The only empirical example that I know of groups which have fitnesses which
remain non-linear functions of their individual's fitnesses are organism
parts grouped within one fertile organism. To my knowledge all organism
group fitnesses remain linear. I you can provide documented examples which
refute this then please do so.

You just keep dismissing and ignoring all such examples. You have been
provided with many such examples over the years. I would include the
example in my riddle of human religious groups.

JE:-
Please provide an example. In this case each group can only constitute a
single selectee and not a group of selectee's.

Following on my riddle above, the fitness of a religious group could be
independent of the fitnesses of its component individuals in the sense that
their may be no link between the two.

JE:-
I can see no way that "the fitness of a religious group" can be
"independent of the fitnesses of its component individuals" simply because
the very existence of these groups remains dependent on just the existence
of the individuals concerned.

This is false. In fact, it is undeniable that all members of every
religious group die, yet the group lives on. Survival of the group MAY be
supported by reproduction of its component individuals, but this need not be
the case.

The only valid question here is the nature of
their fitness relationship. For some reason it is this which remains evaded.

It is evaded because it is false. I am tempted to ask why you keep avoiding
the "obvious fact", but then that would be rhetorical bullying.

I prefer to think of group selection acting when there is absolutely no
correlation between individual fitnesses and group fitnesses.

JE:-
If the fitness of the group is NOT just the simple sum of the fitness of
the parts then that group must constitute one Darwinian selectee.

Then maybe you are a multilevel selectionist after all.

JE:-
No, simply because selection at the very _first additive and independent
fitness level_ determines the final fitness. All so called higher
multi-levels of fitness that remain additive evolve to become fitness
mutualistic to just this first additive in fitness level. Nested higher
levels of fitness remain lethally contradictory forcing a resolution as to
who is the boss: e.g. the transformation of fitness dependent to fitness
independent cancer cells. These newly fitness independent cells have no
other choice but to contest the fitness of the body they are within. Healthy
cells constitute a dependent nested (not independent intersected) fitness.
Once these cells transform to have an independent fitness a selectee a war
inevitably results which destroys both the cancer and the body.

IOW what we are looking at here is one organism and any infertile
offspring as just a single unit of selection. Quite clearly the fitness of
each cell or each organ does not simply add up to provide the fitness of
one organism no matter how you define fitness. I might add neither does
the fitness of the genomic genes. Therefore it is not possible to argue
that any of these parts have an independent fitness.

We are having a semantic problem here. If the fitnesses of higher units
was constrained to be merely the sum (or average) of the fitnesses of their
components, then I would describe the fitnesses of these different levels
in the hierarchy as dependent.

JE:-
But the semantics I use remain empirically refutable. How can it be claimed
to be empirically true that "the sum (or average) of the fitnesses of their
components"... "would describe the fitnesses of these different levels in
the hierarchy as dependent" when empirically these can be proven to remain
fitness interdependent (just reversible intersecting sets fitness) and
therefore fitness _independent_?


They are independent if they have no influence on one another.

JE:-
But this is empirically impossible (see above).

You showed above that you don't understand my argument. You provided no
reason whatsoever to the contrary.

They are independent to a degree if the fitness at one level is only one of
several factors influencing fitness at the other level.

JE:-
I put it to you that factors which appear to you "at one level" as "only one
of several factors influencing fitness at the other level" are actually
contained within a fitness. IOW, the "several factors influencing fitness"
at a higher level must all be contained within a fitness at the lower level.

"Must"? Are you stating an hypothesis again? Or are you saying that you
reason this to be physically or logically impossible?

Here is another example: the fitness of a religion may critically depend on
its ability to attract new members. This might be strongly influenced by
the quality and extent of ornamentation of churches and clergy. However,
the survival and reproductive output of church members may have nothing at
all to do with church and clergy ornamentation.

If the fitnesses at different levels were not so constrained, I would call
them independent. You seem to be using "dependence" in the opposite way.

JE:-
My argument is that you have these critical propositions in reverse where my
claim can be empirically substantiated.

Put another way, any such argument constitutes an oversimplification of
empirical reality which must be corrected before it can be validly
applied.

In my opinion fitnesses at nested levels of organization are often
intertwined and only semi-independent, but it is important to recognize
that being semi-independent is not the same as being entirely dependent.
It is autonomy by degree.

JE:-
Fitness "autonomy by degree" means nothing to evolutionary theory if you
cannot/refuse to define what does and what does not constitute an autonomous
fitness. My definition: any additive fitness (no matter how you define
fitness) represents an autonomous fitness. Therefore, any non autonomous
fitness is any non additive fitness which can only be represented as a
nested fitness.

Please supply your definitions.

Done above.

JE:-
I can see no definition/definitions that you have provided above which
enables me or anybody else here to be able to empirically discriminate
between independent and dependent fitnesses!

If the fitness at one level of organization has no influence over fitness at
another level, then they are completely independent. If fitness at one
level determines fitness at another (apparently existing) level, then they
are completely dependent. If fitness at one level is influential over
fitness at another level, but does not determine that other fitness
entirely, then these fitnesses lie between the extremes of the continuum of
dependence/independence. As I explained above, the existence of different
levels is subject to empirical testing. In addition, fitnesses can be
empirically estimate, allowing for the empirical testing of
dependent/independence of fitness relationships.

You may not agree with my paradigm, but it is no less scientific or testable
than yours, as far as I can tell.

Guy


.

Quantcast