I think I stated my original question rather poorly.  there is no doubt
that domestication is not a result of "regression to the means".  what I
should have stated is if anybody has looked into whether the secondary
characteristics such as piebald patterns (which I actually do believe
is a result from domestication), can be explained by "regression to the
means" instead.  I will also say that I think the experiments they did
are amazing, and I'm asking not because I question their work, but
because I want to extend my understanding of it.
 
Bob stated very clearly why tamenss is increased --> the population
tamness is being skewed towards one end of a bell curve.  there is a
very strong selective pressure that is imposed on the population.  at
the same time that tameness is increased, the population also showed
characteristics such as piebald patterns, curled tails, floppy ears etc.
these characteristics are quite common in domesticated animals, but
seldom seen in the wild.  now, wild animals are also under very strong
selective pressures (e.g.  if you are an albino moose, you'll be the
first to go because predators can single you out easily).  so the
question is, are these secondary characteristics due to tameness, or
due to the population regressing towards the means?  keep in mind no
one knows what the means is.
 
in the article I read, there is very little mention of the group of
control farm foxes.  to rule out "regression to the means" as an
explanation of secondary characteristics, the control group should be
bred from one starting group of foxes, be subject to the same human
contact and tests, bred for a low inbreeding coefficient, but NOT
selected for tameness.  if the resulting foxes at the same generation
also show similar characteristics in similar frequencies, then selecting
for tameness alone does not account for piebald patterns, regression to
the means does.  so knowing the setup for the control group of foxes
would eliminate this explanation.
 
personally, I don't believe the secondary characteristics are a result
of regression to the means, but I was wonderinf if there is an
experimentation related reason to eliminate the regression to the means,
not an alternative explanation for the appearance of secondary
characteristics that make sense.
 
another thing I didn't find from the article is whether the domesticated
foxes have more health problems than farm foxes.  reason I'm curious, of
course, is whether we can domesticate polecats and get a healthier breed
of ferrets out.  Bob, do you have thoughts on this?
 
// ***************************************
// Selina, Sprite, Sand, Bear, Dart & Clef
// in spirit - Birch, Dief & Storm
// http://www.cgl.uwaterloo.ca/~ssiu
[Posted in FML issue 4128]