Re: The Cost of Substitution



Guy A Hoelzer hoelzer@xxxxxxx wrote:-

I know what a substitution is.

JE:-
You can define "a substitution" to be anything you wish so what you
claim to
"know" is not under question. Please supply YOUR definition of "a
substitution" so that _we_ can know what _you_ mean by it.

This is not as ambiguous as you make it out to be. There are two
definitions for substitution in population genetics. The first
definition,
which students learn in the Genetics class, is that "substitution" is
synonymous with "point mutation", where one base is exchanged for a
different one in a DNA sequence.

JE:-
One substitution event is defined by the definition: a "point mutation"
where "one base is exchanged for a different one in a DNA sequence" which is
an inductive inference. The cost at this level becomes a simple deduction
from it: one point mutation as a COST. It cost one to get one.

The second definition, which is clearly
to
applicable one in this discussion, is that a "substitution" is synonymous
with a "fixation event", where the frequency of a mutation increases to
100%.

JE:-
Ok. Now you are defining the cost for ONE POPULATION. The cost remains
deductive from the size of the proposed population. It cost one population
to substitute one population. If no population constant is defined then it
become impossible to calculate the actual cost.

There is admittedly a little more ambiguity in the 100% figure.

JE:-
Any ambiguity is entirely dependent on a failure to define a population
constant for the proposed simplified model.

To
allow for the fact that new mutations are constantly appearing and going
extinct, a figure of 95% is often used instead.

I asked what the claimed cost might be.

JE:-
I stated that the cost will always be (i.e. without any exceptions,
making
this claim entirely refutable) just a simple deduction from whatever
anybody
defines "a substitution" to be. I have provided my definition. Until you
provide your definition, the cost of substitution remains _impossible_
to
deduce.

This cost makes no sense given the meaning "substitution."

snip<


JE:-
Incorrect. Please see above. You may disagree (dislike) with the "sense" it
makes but you cannot validly claim it makes "no sense".


JE:-
Felsenstein's latest attempt to define substitution remained ambiguous
because it did not define the size of the population that is to be
substituted as a _constant algebraic term_. Using his most recent (or
any
previous) definition this can vary from 1 (alone providing a zero cost
of
substitution) to any other number and then back again. The cost of
substitution increases linearly with the size of the population.
Please
refer to my (as usual unanswered) criticism of the definition of
substitution that Felsenstein has only recently posted.

My intuition tells me that the notion of a substitution cost must
depend on
the direction of change in population size, but I can't really know
this for
sure until I understand what the claimed cost is about.

JE:-
Populations can only increase, remain stable or decrease. I argue that
selection forces a constant increase.

This is very strange to me. Is it your position then that selection is
not
operating in any instance where a population is decreasing in size?

JE:-
No. It means that the supposition of just a negative modeling population
constant forcing a population to constantly decrease in size means
extinction for that population. If one form remains constant or increases
its own TDF (as I have defined it), i.e. does not decrease its TDF on a
constant basis then the entire population becomes extinct except for just
this one form. If every form lowered it's TDF on a constant basis no form
would exist on this earth and life would become extinct despite the fact
that on just a comparative (relative) basis one form within such a doomed
population would be selectable over another. This is known as the relative
gain absolute loss argument.


Only if the population increases or
minimally remains stable can any cost of substitution argument make any
sense because the remaining option of a constant decrease only allows
extinction. My point was and remains: Felsenstein must supply his so far
_entirely missing population constant_ as an algebraic term for his zero
cost argument to be rational. This constant must minimally allow the
population to remain the same size or constantly increase. It cannot
allow
the population to constantly decrease simply because such a population
becomes extinct. IOW, any cost of substation _simplified model_ must
MINIMALLY define a population to remain _constant_ because only this
disallows extinction while providing a single measure of zero
reproductive
excess. Because Felsenstein did not define any population constant he
allowed himself TWO and not just the ONE point of zero excess for the
SAME
problem. This allowed himself _two contradictory_ Galilean frames of
reference to calculate his zero cost. Quite obviously, if the point of
zero
reproductive excess can be moved up or down "ad hoc" as just a variable
Felsenstein allowed it, then "anything goes" because the cost of
substitution remains relative to nothing defined by anybody.



I await a clear answer to my question. If you can't tell me the
currency of
the cost, then your claim that there is a cost seems empty to me.

JE:-
I have stated the units of substitution that I use within Felsenstein's
MODEL on a number of occasions: the reproduction of ONE HAPLOIID FERTILE
FORM (see below). If this population constant is defined to be 100 as
the
critical Galilean frame of reference and the population becomes reduced
to
50 within an adverse environment then the cost of substituting one gene
for
another is reproducing 50 fertile forms because 50+50=100.

This is an energetic cost. Right?

JE:-
The energetic cost is deleted within the simplified model. See Below.

It is not just zero as Felsenstein argued because the population
constant
cannot be switched from 100 to 50 "ad hoc". We are not doing magic
tricks we
are supposed to be doing science.

If the population constant of 100 is not just a minimal constant size
but a
_constantly expanding_ population then reproducing 100 fertile forms is
now
over and above the size of the original population. Using the same
example,
where the original population is 100 then the constant adds 100
increasing
that population to 200. In this case, i.e. defining a _population
increase
constant_ of 100, when the population became reduced to 50 then the cost
of
substitution is now reproducing 150 and not just 50 fertile forms.

Just so you know my view, despite your frequent arguments to the contrary,
I
think your reliance on constants is unnecessary and often burdensome.

JE:-
I am sure the Spanish Inquisition made a similar case against Galileo (who
probably invented frames of reference out of sheer desperation) after they
forced him to recant his refutable and rational argument against them.


IMHO
it turns your argument here into an uninteresting one.

JE:-
I am sure the Pope would agree :-)

This is just my
opinion and I'm sure I will read more of your advocacy for the importance
of
constants in such arguments, but please don't take this as a request for
your argument here. I am just trying to keep my own position clear in
this
discussion.

JE:-
I am well aware of your position. My position is that your position (and
Felsenstein's position) remains irrational.

Here is (yet another) analogous example: the cost of filling one
particular
cup of water (in any units of water that you wish to nominate) is the
total
amount of the same water units required to fill just that particular
cup.

OK. So in this analogy the currency is water. Again, what is the
currency
of cost when substitutions happen in populations?

JE:-
In this analogy I _strictly defined_ the currency as: units of water.
What
exactly measures the substitution of one unit of water by another and
what
measures the cost of this must be in exactly the same units. The cost of
filling the cup, refilling it after it was partly or completely empty or
for
keeping a constantly expanding cup full must be in exactly the same
units:
units of water. In this analogy you can imagine one unit of water to be
colored red substituting for another that is colorless turning the water
different shades of pink depending on the level of substitution.
Attempting
to keep a constantly _decreasing__ cup full is just a waste of time as
is
attempting to calculate the cost of substitution in units of water for a
cup
which can expand or contract as just another variable.

Zero excess was defined by Felsenstein to mean that the cup remains
full as
water is added and removed.

JE:-
A comment here would have been appreciated.

OK. I take this to imply that the cost currency for you is the energy
invested in producing more offspring than would otherwise be necessary.

JE:-
No. The _model_ cost is strictly in units of water and not the amount of
energy required to provide them. That is a related but entirely separate
issue.

It cannot be separate because it is impossible to have one without the
other.

JE:-
Yes, "it is impossible to have one without the other" just as it is
impossible to have a penny with just one side. Such things are termed
RELATIVE OPPOSING suppositions. We have discussed these before in the sexual
selection Vs the natural selection controversy They remain relative opposite
suppositions and NOT the SAME supposition. What all relative opposites have
in common is that they must employ _exactly the same_ Galilean frame of
reference.


Please see my other recent post where I address this notion directly.
Reproductive excess is not required for substitutions to occur in
populations.

JE:-
The above makes no sense unless you provide at least one population
constant. When one is defined then a rational zero cost of substitution
becomes maximally equal to the population constant in exactly the same
units
as defined for the what measures "one substitution". Felsenstein allowed
two
and not just the one point of zero excess by refusing to define any
population constant whatsoever. Quite obviously, allowing two zero
points of
zero reproductive excess for the same problem remains hopelessly
irrational.

Of course, it you can expand the cup "ad hoc" then the same level of
water
required to fill the cup remains insufficient INCREASING Felsenstein's
original definition of zero excess.

Of course, populations do change in size all the time, so it is
important to
consider this effect in this context.

JE:-
Yes but in order to be able to do so, even within just a simplified
model, a
population constant has to be defined. Without one costs become
arbitrary
simply because the definition of what constitutes "a substitution" was
always just arbitrary.

What constitutes a substitution has little arbitrariness about it (95% vs.
100%), and I see no reason at all that population size can't be a variable
rather than a constant in this model.

JE:-
The absolute size (not just a relative percentage size) of the population
critically matters to the cost of substitution. The cost of substituting a
MINIMAL population of just 1 is just zero simply because 1 represents the
point of zero excess AND the critical point of non extinction. OTOH the cost
of substituting a population greater than one increases additively with the
absolute size of that population because the point of zero excess increases
on just an additive basis.



ReMine focuses almost entirely on the case of human evolution, where
the size
of the human population has been growing at an astronomic rate for a
very
long time. In that context, it would be essential to include
population size
change into account.

JE:-
Yes, as an expanding population _constant_. If you attempt to reason
without
at least one population constant the cost of substitution row can never
be
addressed rationally.

Or maybe there is no cost to address, which still appears most likely to
me
at the moment.

JE:-
Then please provide your own model to illustrate what you are claiming.

At the moment the subject under discussion was and remains the validity of
FELSENSTEIN'S model, i.e. NOT your model, Edser's model or ReMine's model.
Felsenstein assumed TWO and not just the ONE zero point of reproductive
excess in order to pay for his entirely misrepresented "zero" cost. IOW his
zero cost argument was little more than old fashioned slight of hand simply
because it assumed no (boring) Galilean frame of reference. Either agree or
disagree that this was the case. If you disagree, then please provide the
missing argument that allowed Felsenstein to validly employ two and not just
the one zero point of reproductive excess to calculate a zero cost for the
EXACTLY THE SAME problem.
[snip]

I know this is your opinion, but the quality of your argument does not
seem
to support your smear.

JE:-
"Smear"? Did Edser employ an irrational argument to prohibit the
publication ReMine's paper? No, but Felsenstein did. In the paper that
was
rejected ReMine correctly criticized Felsenstein's zero cost argument as
false because it was demonstrably based on a contradiction. The only
ethical
thing for Felsenstein to have done was to excuse himself from refereeing
ReMine's paper on the grounds of a conflict of interest. Did he do so?
No,
he did the opposite. He deployed the argument from authority to smear
ReMine
as a creationist. It remains arguable that if Felsenstein had excused
himself (as ethics required him to do) ReMine's paper would mostly
probably
have been accepted for publication. As ReMine correctly pointed out, the
fact that he is a creationist remains _irrelevant_ to the issue
addressed by
the rejected paper.

Has anybody else here supported ReMine's entirely rational criticism of
Felsenstein's irrational zero cost argument? No, just Edser.

It is possible that there is no conspiracy, rather ReMine's argument may
not have been rational in the first place. Food for thought?

JE:-
Guy,
Firstly I NEVER alleged a "conspiracy". What I have always alleged are three
things:

1) The consistent use of just a false epistemology.

2) Consistent evasion by Felsenstein of any rational criticism of his
models.

3) Sycophantic behavior towards Felsenstein on this particular issue by
almost all of the other Professionals who post here displaying a deplorable
standard of integrity on their part.


Has anybody else here besides Edser (including ReMine) pinpointed
_exactly
how_ Felsenstein secretly paid for his misrepresented "zero" cost of
substitution? No. Has anybody here provided a rebuttal against Edser's
proposition that ANY rational argument must assume at least one Galilean
frame
of reference? Nobody at all....

Can a REASONABLE person assume that gene centric Neo Darwinists are
simply
closing ranks to protect their own? Most definitely.

They could. They would be wrong, but they could do that.

JE:-
Are you prepared to examine the actual facts in an unbiased and way or just
dismiss what happened here using the politically tried and tested fence
sitting exercise of "yes" AND "no"?


Perhaps you can tell me what the units are of this cost.

JE:-
I have been defining this for more than three years now: the units of
substitution are not individual genes they are individual fertile
forms (in
Felsenstein's example individual haploid fertile forms). IOW they are
Darwinian but most certainly not gene centric Neo Darwinian,
selectees.

OK. If I understand correctly, you are claiming that individuals must
produce more offspring to permit substitutions in the population than
they
would otherwise optimally produce.

JE:-
No. They must reproduce enough fertile forms to equal a defined
population
constant. Felsenstein, yourself or anybody else MUST supply a proposed
algebraic population constant. Not only does it become impossible to
argue a
cost structure without one, it even remains impossible to make sense of
what
"one substitution" is supposed to mean.

Now that you have a concrete definition of substitution, maybe some of
your
concerns will be answered.

JE:-
As I have outlined above: you have not addressed my most important
objections.

I REPEAT: Please explain how TWO zero points of reproductive excess and not
just the ONE could be validly employed to provide Felsenstein's zero cost
for exactly the same problem? If you don't know but you do actually care,
why don't you ask Felsenstein? Yet again I have to emphasize that this has
become a question of integrity for all sbe posters.

Regards,

John Edser
Independent Researcher

edser@xxxxxxxxxxxxxx
'





.

Loading