GENEALOGY-DNA-L Archives

Archiver > GENEALOGY-DNA > 2008-05 > 1209757760


From: "Lancaster-Boon" <>
Subject: [DNA] Genetic Adam theory
Date: Fri, 2 May 2008 21:49:20 +0200


Dear Joe

My answers are inserted...

---
From: "Joe Flood" <>
Subject: Re: [DNA] GENEALOGY-DNA Digest, Vol 3, Issue 521
Date: Fri, 2 May 2008 23:38:52 +1000
References: <>

Yes I mean the theory that all male lines go back to a single human male. As
far as I know this is not proven, in fact I think it is unlikely. I am
always uncomfortable with theories which appear to have a religious basis
and which everybody "knows". If someone can refer me to a concise proof
which handles the objections I have raised (plus a few others that spring to
mind) I would very much like to see it.

Andrew: I did not read anything in what you wrote which seemed like a
refutation. That's why I wrote to define the theory better for further
discussion. Concerning whether the theory is proven or not see below...

I refer to the relatively modern period of the last five hundred years
because that is the only time in which you can see the principles in action,
in practice. By and large surnames follow Y-DNA; and the principles of their
extinction or otherwise are mathematically identical to the principles
governing Y-DNA survival, with about the same timing. There is a minor
complication with bastard children, as you suggest, but it is no greater
than the complication arising from NPEs which work in the opposite
direction. There is no reason that surnames should converge or diverge
faster or slower than Y-DNA.

Andrew. I do not want it to sound like I am playing a word game. But this
last sentence sounds like "theories which appear to have a religious basis
and which everybody 'knows'" in the passage above. Surnames and Y-DNA
clearly move according to vastly different principles. That they approximate
at all is a wonderful coincidence for genealogists, but if you have broad
experience in surname projects you'll know how complex the relationship can
be.

Andrew: The second point to make here is that Ken's points, both for better
and worse, were indeed based off of what can be seen in the short run, and
extrapolating back. This is the way in which the theory is proved beyond any
doubt in the mind of most people working on this. Again, I did not see you
really offer an alternative interpretation of observable DNA data. See below
for more...

What I am suggesting for surnames- and by extension for Y-DNA - is that
there are large numbers of very long very narrow lineages with no branches
to speak of.

Andrew: Well, I still not convinced that you are "talking the same language"
as, for example, Ken. What do you mean by branches? What Ken means is SNP
mutations. These are very rare but it would be even more strange if they did
not happen at all for >100,000 years. Are you saying that there are men out
there with no SNP's on their Y chromosome since their furthest back human
male ancestor? On what basis?

In fact if my observation is correct, I would say they form the considerable
majority of clades descending from a given point in time (but of course, not
the majority of descendants). Of course there are a few huge clades, like
say Nhiall - these can be observed in the last few hundred years also. My
guess is that the distribution of clade size is lognormal, rather like the
distribution of city sizes but more skewed.

Andrew: Better to let Ken tackle the distribution you assert. But more
importantly, is this just a feeling, or are you really making statements
based upon a model you are not sharing?

When we run random sampling of the population and then start running
clustering routines, of course we are going to pick up the big clades
straight away. But we are going to ignore or even miss sampling most of the
small (narrow) clades, which could be extraordinarily ancient. Therefore all
the action goes into picking up on these big clades and analysing them and
the impression is given that there are just a few original ancestors around
within a given time span when there could be many thousands.

Andrew: Once again I think you are on a very interesting track, but perhaps
not totally aware of the work done before and of what Ken was referring to.
Clustering routines refers to what we do with the microsatellites or DYS
type "markers". These are mutations that can jump back and forth very
easily. Ken was talking about SNP mutations. The pattern of these is not
based on running clustering routines, but only on seeing whether the SNPs
are strictly nested in a logical family tree. You should keep in mind that
Ken and others doing similar things have spent more time seeking any sign of
small but ancient clades than they have one the big common clades. There
certainly are some. But they still all have SNP mutations which they share
with a big part of humanity, in a strictly nested "family tree" structure,
as if they sat in the same family tree of male lines.

Best Regards
Andrew



This thread: