Help with inheritence.

[Nick Thompson]

Carl,

Thanks for pitching in,

If we are going to dig our way out of this one, it will probably be through
a hypothetical.  
Imagine that I have ten cultivars of corn, and I am curious about their
productivity.  Now imagine
 further that I have 10 bags of fertilizer and I am curious about their
relative
 effectivness.  So I have ten levels of corn and ten levels of fertilizer
and I lay down the
corn in rows and the fertilizer in perpendicular bands.  

Now lets imagine some possible results. We could get significant effects
for cultivar,
for fertilizer, of for both, and, or for an interaction between cultivar
and fertilizer.  If there
were an interaction, and we didnt have enough corn or fertilizer to make
replications
of our experiment, then we would be unable the separate the interaction and
error.
If we did the replications, however, and statistically demonstated the
interaction,  in'
absense of main effects for cultivar, we would be in no position to make a
decision
which cultivar to plant unless we knew which fertilizaer we had available.  

There would be no way to select for cultivar.

Now remember that it is I who is confused about this and it is stupid ...
and perhaps
even rude ... therefore forme to assert my position so firmly but...

I think we would come to the same conclusion if the rows and columns were
alleles at locus one and alleles at locus two, or alleles at locus one and
environments,
or mating partners, or ..... or ..... or.  Dominance, epistasis,
imprinting, etc., all render
selection impossible.  just to the extent that they reduce the corrrelation
between the
possession of a trait and its effects.  

Think about poor nature, trying to select for Malaria resistance in the
congo.  Even
though the heterozygote is clearly superior, there is no way select it
beyond 50 percent
because there is no gene for the heterozygote condition.  Ditto sex.  there
is no gene
for being male.  

I know I am in effect just repeating myself.... perhaps I am SHOUTING even,
but
I think what I am saying is a statistical necessity.  When one selects, one
uses the
traits of the parents as an index for the traits of their offspring.  Any
process that
makes parental phenotype an inferior, less reliable index of ofsprong
phenotype
makes selection that much harder.  

Nick

NIck

 


Nicholas S. Thompson
Professor of Psychology and Ethology
Clark University
[hidden email]
http://home.earthlink.net/~nickthompson/
 [hidden email]


> [Original Message]
> From: Carl <[hidden email]>
> To: <[hidden email]>; The Friday Morning Applied Complexity
Coffee Group <[hidden email]>

> Date: 1/25/2005 11:19:10 PM
> Subject: Re: [FRIAM] Help with inheritence.
>
> Nick,
>
> I have not yet understood why hyperplasticity or dynamism in a genome
> must necessarily interfere with inheritance.  It certainly may mediate
> it. An applied complexity practitioner might find it reasonable to talk
> about the robustness of a gene network relative to the non-canonical
> regulatory and genetic transfer mechanisms you speak of.  Under some
> circumstances these mechanisms might even reinforce inheritance.  My own
> view (which I don't think is particularly radical anymore) is that
> genes probably get expressed relative to the network environments they
> find themselves in (which they may in their turn have a hand in
> creating).  These notions (e.g. robustness of expresssion in the face
> of regulatory or network variability) do not seem to me to be
> antagonistic to inheritance, rather, they challenge and expand our
> notions of what 'selection' is and means when we can no longer guarantee
> 1:1 mappings between groups of traits and gene groups or vice-versa.
>
> I don't think I would have a problem with the idea of
> that-which-is-inherited being somehow vetted through some sort of
> genomic or epigenetic dynamics.  You might feel better considering
> that those dynamics may be responsive to the same environment in which
> the macro-scale selection occurs.
>
> Caveat: All this blowing bubbles in the standard model is by way of
> finding some theory to help us write some interesting (and yes, useful)
> programs that may illuminate some larger issues about complex systems.
> We're not necessarily saying earth works this way.
> Or that it should.  Yet.
>
> Carl
>
> Nicholas Thompson wrote:
> >
> >
> > Dear Friamers,
> >  
> > Everything that you-all are having me read  undermines the notion of
> > natural selection.  But it does so by undermining genetic inheritance,
> > the idea that because of genetic material passed from parent to
> > offspring, offspring will differentially resemble.  I would have

thought

> > that of all the premises on which natural selection was based, the
> > premise of family resemblance was the most secure.  I have two
> > questions, Does complexity have an alternative theory of
> > inheritance?  Does complexity have an evolutionary theory in which
> > inheritance plays not significant part?
> >  
> > The full text of my rumination on this subject follows:
> >  
> >
> >             It has become a standard critique  contra the neo-Darwinist
> > synthesis  that it is based on an unrealistic, over-simplified, and
> > perhaps ideologically based view of molecular genetics and
> > development.   Books by Margulis, Caporale, Elldredge, Gould, and

others

> > emphasize the great complexity of the events that take place in the
> > creation of the gamete genome, in the growth and development of the
> > cells subsequent to fertilization, and in the day-to-day production of
> > protein products by the bodies cells.  Some authors scoff at what they
> > call the “bean-bag” genetics of neo-Darwinist and its gratuitous
> > assumption t hat the genome consists of  a set of mutually independent,
> > randomly reassorting heritable units.  The critique cannot be ignored:
> >  The simple fact of chromosomes would seem to render such a conception
> > absurd, even if one ignored  the recent discoveries of high levels of
> > interaction among elements of  the nuclear genome and between that
> > genome and the mitochondrial genome during cell division.
> >
> >  
> >
> > But exactly how is the above critique a critique of Darwinism.

Darwin’s
> > theory is stated in the form of a hypothetical, a series of conditions
> > which if met lead to adaptation and evolution.  We are thus led, by
> > affirming the consequent, to the conclusion that these conditions are
> > the cause of adaptation and evolution.  The conditions are:
> >
> >  
> >
> > (1)If  the reproductive potential of the members of a species is
greatly

> > in excess of the capacity of their worlds to support them
> >
> >  
> >
> > AND
> >
> >  
> >
> > (2) The members of this species vary in the possession of traits
> >
> >  
> >
> > AND
> >
> >  
> >
> > (3) Those traits are heritable (i.e., they are possessed differentially
> > by the offspring of individuals that bear them)
> >
> >  
> >
> > AND
> >
> >  
> >
> > (4) some of these traits offer an advantage in the struggle for
> > existence implied by (1)
> >
> >  
> >
> > THEN
> >
> >  
> >
> > (5)  traits of the type mentioned in (4) will in time come to
> > characterize the species producing adaptation of the species to the
> > conditions of its existence.  This process, applied to many species,
> > will in time produce evolution, since evolution is just (on my account)
> > the fact that over history,  different species have tended to adapt to
> > their different conditions of existence.
> >
> >  
> >
> >             Like all cumulative chains of premises, natural selection
> > theory is a logical chain which depends on ALL of its elements  for its
> > logical success.  In other words, if any of the 4 premises above is
> > false, than natural selection as understood by Darwin cannot occur.  
> > Which of the premises of natural selection does the developmentalist
> > attack.  Surely not the Malthusian Premise, Premise (1).  Nor also the
> > variational premise (2) nor the reproductive advantage premise (4).    
> > The developmental critique is in fact an att ack on the inheritance
> > premise.  But the inheritance premise is the only one of the four that
> > is NOT a hypothetical.  It is often said that Darwin’s explanation of
> > inheritance was very bad, and got worse as he elaborated it through the
> > several editions of the Origin.  But his knowledge of the FACTS of
> > inheritance was very secure:  Darwin knew from close personal

experience
> > and avid reading that inheritance of traits from parents to offspring
> > was a possibility. And a century of research in breeding and
> > hybridization in plants and animals has not diminished our confidence
> > that (at least) some traits can be differentially represented in
parents
> > and offspring.   For the developmental critique to succeed in its
> > attempt to undermine Darwinism, it must do so by undermining not the
> > theory of inheritance, but the fact of inheritance, and that fact would
> > seem to be the most secure of Darwin’s premises.
> >
> >  
> >
> >             True, undermining the fact of inheritance would have a
> > devastating effect on natural selection theory.  If natural selection
is

> > to work, then the parents have to serve, in effect, as representations
> > of the offspring.  It must be true that decisions to breed or not to
> > breed parents, based on their phenotype, must be reflected in the
> > phenotypes of the offspring in the next generation.  Any developmental
> > process that interferes with this representation from parental to
> > offspring generation, interferes with natural selection.  Mutation
> > interferes with natural selection.  Do minance and epistasis interfere
> > with natural selection.  Genomic imprinting interferes with natural
> > selection.  Horizontal transfer of genes from other organisms would
> > interfere with natural selection.  Environmental muting or triggering

of
> > traits interferes with natural selection.  In fact, to the extent that
> > any mechanism interferes with the isomporphism between variations in
the

> > parental generation and variations in the offspring generation, that
> > mechanism interferes with natural selection.  It does so only because
> > */it interferes with inheritance/*.
> >
> >  
> >
> >             So let is grant for the moment that the main challenge of
> > developmentalists for natural selection theory is a challenge against
> > Mendelism, not a challenge against Darwinism.  Does complexity theory
> > have anything to offer to mitigate this challenge.  Is there a
> > complexity variant of inheritance theory?
> >
> >  
> >
> >  
> >
> >  
> >  
> >  
> > Nicholas S. Thompson
> > Professor of Psychology and Ethology
> > Clark University
> > [hidden email] <mailto:[hidden email]>
> > http://home.earthlink.net/~nickthompson/
> > [hidden email] <mailto:[hidden email]>
> >  
> >
> >
> > ------------------------------------------------------------------------
> >
> > ============================================================
> > FRIAM Applied Complexity Group listserv
> > Meets Fridays 9AM @ Jane's Cafe
> > Lecture schedule, archives, unsubscribe, etc.:
> > http://www.friam.org