Guy Hoelzer  wrote:

> in article b3qr9s$1i2s$1{at}darwin.ediacara.org, John Wilkins at
> john.wilkins{at}bigpond.com wrote on 3/1/03 9:40 AM:
> 
> >> How would you describe the distinction between sampling and sorting?
> >> Or are they synonyms in your lexicon?
> > 
> > No. Sorting is biased sampling - it orders the prevalence of the
> > alleles. As you note, some sampling can be random (i.e., drift) or
> > neutral (Kimura-style neutral evolution).
> 
> The use of the term "sorting" in the literature with which I am most
> familiar began with John Avise.  He and his colleagues used the term
> "lineage sorting" specifically to refer to the RANDOM
sampling of lineages.
> I don't think the word "sorting" necessarily implied random
sampling before
> Avise got hold of it, but the cat is out of the bag.

Well it has a life outside taxonomy and paleontology, too. But I was
taking it from the shift Gould made (and later seemed to either revoke
or publish in his Big Book material unchanged from an earlier period)
from "species selection" to "species sorting" under
pressure from folk
like Hull and, to an extent, his collaborator Lloyd. This goes directly
to this issue. Hull and others (even including Eldredge) argued that for
it to be selection, there had to be variation (i.e., a population of
alternatives) and this variation had to be hereditable. Arguably,
species had the latter (on Gould's account), but there was never a good
claim for them having the former.

Gould had two options - to redefine "selection" or to rename the
phenomenon. He initially chose the latter, but the Big Book preserves
the former. I think the latter is the right path.
> 
> >>> But selection itself requires two further conditions - variation in
> >>> hereditable traits, and that those traits have some
differential bias
> >>> that correlates with hereditability. This differential bias we call
> >>> fitness.
> >> 
> >> This definition confounds the traditionally recognized
distinction between
> >> the act of selection and the population's response to selection.
> > 
> > How?
> 
> It does because the process of selection does not depend upon the
> heritability of traits, although a population's response to selection does.

If organisms have traits that are not hereditable that are subjected to
eliminative pressures, how is the overall frequencies of a population
changed? Evolutionarily such change is merely noise. Selection *is* the
population's response to differential influences of reproduction of
hereditable traits.
> 
> > I am merely saying that genes and traits must be correlated, and
> > have a hereditability coefficient, for selection to engage.
> 
> Again, this requirement is at odds with the traditional view.  The
> mechanism underlying phenotypic variation is irrelevant to the process of
> selection. There must be heritability if the population is respond to the
> process of selection with adaptive evolution.  Furthermore, any mechanism
> of inheritance will do, so it does not require the involvement of genes.
> If it did, Darwin and Wallace could not possibly have conceptualized the
> idea of natural selection.

They did not need molecular genes, to be sure. But they did have a
notion of what came to be called biometrical traits - which by
definition are highly hereditable. When Mendelian genetics was folded
into Darwinian theory, they were not molecular either, but they
satisfied the conditions of Darwin-Wallace selection.

You are quite right that any mechanism of heredity will do - selection
occurs on cultural transmission in my world view. It will occur also on
prions if there are sufficient variations in a population. But all this
means is that there is a strong correlation between the mechanism of
heredity and the economic interative traits.

This is known as the Hull-Dawkins distinction - there must be a
replicator and an interactor - the first a mechanism of hereditability
that correlates closely with the interactive properties of the resultant
entity. This is a generalised notion of evolution that does not rely on
any particular mechanisms. I have some concerns about how exclusive this
is for all evolutionary processes, but it is a solid distinction for the
bulk of evolutionary processes involving selection.
> 
> > I am not saying that for selection to occur some trait has to go to
> > equilibrium; that is an effect of selection.
> 
> I did not have anything relating to equilibrium in mind.  In fact, I
> personally think that there is no such thing as an equilibrium system in the
> universe.  They only exist in our simple models.

I agree. Put another way, if something is stable over very long periods,
it is not a physical, concrete thing, but an abstract thing. Only
classes of things can remain unchanged, not things themselves.
> 
> >>> Another way to think of it is that you have a variable
for alleles that
> >>> can take any value from one to zero for variability.
> 
> I assume you mean "frequency" rather than "variability."

No, variability - polymorphism.
> 
> >>> Selection occurs when it is greater than zero but less
than one for a
> >>> locus in a population. You cannot non-arbitrarily draw the line
> >>> anywhere in particular, but if the variability is extreme enough we
> >>> just do not call it selection any more.
> 
> You lost me.

No polymorphism, no selection. All polymorphism, no hereditability (or
else there would not be a unique morph for each organism).
> 
> >> I am not sure what you are getting at here.  Every individual has a
> >> unique phenotype, even within clonal populations.  Are you suggesting
> >> that selection cannot operate under this condition?
> > 
> > Not quite. I am suggesting that for selection to occur, the variability
> > has to be between total identity of all organisms (otherwise what
> > happens to them is accident only, and does not get passed on in any
> > correlated manner) and total variability of all organisms (in which case
> > there is no hereditability to select *on*).
> 
> Now I am hopelessly lost.

This is my goal in life - to confuse folk...

-- 
John Wilkins
"Listen to your heart, not the voices in your head" - Marge Simpson
---
þ RIMEGate(tm)/RGXPost V1.14 at BBSWORLD * Info{at}bbsworld.com

---
 * RIMEGate(tm)V10.2áÿ* RelayNet(tm) NNTP Gateway * MoonDog BBS
 * RgateImp.MoonDog.BBS at 3/3/03 7:41:02 PM