Archiver > GENEALOGY-DNA > 2005-06 > 1118087160

From: (John Chandler)
Subject: Re: [DNA] Middle Eastern ancestral markers on new Euro 1.0 test
Date: Mon, 6 Jun 2005 15:46:00 -0400 (EDT)
References: <>
In-Reply-To: <>

Andrew wrote:
> No it is not. The term "selective advantage" clearly indicates that there is
> a cause underlieing the effect - something non-random so to speak, meaning
> that in terms of probability, the probabilities of survival for this group
> are higher than for others.

You are right in the sense that an observed selective advantage is
somehow unsatisfying if you don't know the reason for the advantage.
However, I've been speaking statistically all along. When I say that
the survival rate is observed to be higher in one group than in others,
I've already discarded by implication any cases where the difference
is not statistically significant. I say again: the observation of
a signficant difference in survival rate *is* the operational definition
of selective advantage.

> > As for attempt "b" above, you are much closer, but you left out an
> important proviso -- majority haplotype in a large population.
> Well you still have not defined large and small. The examples you give are
> always useless extremes (50 or "billions").

The problem is that you have been tilting at a straw man all this time
instead of reading what I wrote. I clearly specified very early on
that 1000 was a "large" population for this purpose, and I gave
another example of 2000 being "large". The reference to billions was
not population size, but rather years necessary for the events you

> Could you offer any convincing arguments for any of the above assertions? As
> far as I can see war, famine, pestilence, small populations, and occasional
> periods with declining populations are perfectly normal in Iberian history
> and pre-history.

You still don't have it. You are referring to populations in the
plural as if you could break up Spain (or the Basque region) into
village-sized groups that would drift genetically in unison. True, a
village-sized group can drift genetically, or even disappear, but the
composite of many such groups is a large population that averages all
those little drifts out.

> what we were discussing was whether you *need* immigrations in order to
> explain a big change in the haplotype population.

And the answer is yes, unless you want to postulate a bottleneck.

John Chandler

This thread: