GENEALOGY-DNA-L Archives

Archiver > GENEALOGY-DNA > 2005-12 > 1134890450


From: "John McEwan" <>
Subject: RE: [DNA] Re: Dal Riata Modal Haplogroup
Date: Sun, 18 Dec 2005 20:20:50 +1300
In-Reply-To: <00cd01c60390$9065c340$1802a8c0@WorkGroup>


Dear Robert

Thanks very much for your cogently argued hypothesis and supporting
evidence. I will be honest, I save all my key e-mails but the import of
your previous one did not really strike me until now. It may have been
when I was overseas. In fact I am sure I just got a link to the
Livingston Mclea site.

I think you are on to something and I have been fumbling around with a
different aspect of the same phenomena. If I understand you correctly,
what you are implying is the "pre scots" haplotype emerged from central
Europe and moved to England then to Scotland where it became a major
regional haplotype cluster.

Now I have read a little pre-history and this sounds very like part of,
or a variant of, the "Kurgan hypothesis" and indo-european language
(Gaelic in this case) spread along with chambered burials, horses and
pastoralism (circa ~5000 years bp) some time after agriculture
techniques spread. Yes, I have a copy of JP Mallory!

To recap my work (all available on www.geocities.com/mcewanjc ):
* In April I examined selected surnames from Ireland Scotland and
England in order to probe the origin of Dal Riadic surnames. The results
of this simple phase one analysis were clear. Dal Riadic surnames were
closer to Irish than "Britonic" based on DNA haplotypes. This suggests
but does not prove an Irish origin.
* A phase 1 cluster analysis was also done on the data (25 STR markers)
and this identified that the Irish R1b cluster was present a higher
level in Irish (26%) than dal riadic (19%) than scots Britonic Celt
(13%) surnames. In contrast the Scots R1b cluster was low in Irish (0%)
modest in dal riadic surnames (5%) and high in scots Britonic celt
surnames (17%). These differences were significant.
* I actually have just today tried to extract out of the new Irish Moore
et al paper the % of the Scots cluster in Ireland. The figure was ~5%
but like your 12 marker haplotype a few assumptions had to be made.
See www.geocites.com/mcewanj/clustersummary.htm
* I then did a phase 2 analysis in June using R1b data from Ysearch and
37 STR markers and this again identified the Scots and Irish clusters
but also several other groups that subsequently developed into the
Frisian R1b cluster. However, my emphasis was on the Scots cluster.
See www.geocities.com/mcewanjc/phase2.htm
* Then I got myself into really hot water using the wrong mutation rate,
but aside from the estimated dates you will see the same general
conclusion was reached that you have proposed. The so called Laginian
Dumnonii hypothesis had them rolling in the isles :-(
www.geocities.com/mcewanjc/originr1bkn.htm

I have still not revised it mainly to remind myself to slow down and
doublecheck things and partly because estimating dates is fraught.

The central observations were: the distribution of the Scots cluster was
a trail leading from southern England, and was also present on the
continent, older haplotypes were in England and the continent suggested
that the cluster had come into England (invader) or originated in
southern England (adopter) and traversed slowly up the country
(excluding Wales). The absolute size of the movement also strongly
suggested and event prior to the Roman invasion.

The key problem was dates. The dates I obtained were ~300BC, others
thought and compellingly argued they were 3X older (circa 5500yrs BC).
In reality great uncertainty exists. This has led to me going back and
doing a "global" phase 3 analysis which I am still summarising some 6
months later, but there are only 2 possibilities consistent with the
observations:
a) the Scots cluster are the remnants of the first native British
b) they are associated with either the introduction of agriculture or
some new technology (horses, bronze or iron ).... In the latter case
they could have been early adopters in southern UK that subsequently
spread. I personally think the event has to do with agriculture in some
way because the imprint is through the best arable land in the UK.

The current pattern observed (Wales low, Ireland low, England moderate,
Scotland high) is a consequence of subsequent events in particular the
Roman invasion of Britain, then the Angles, Saxons, and Normans all
directly or indirectly reducing the fraction of the prior inhabitants in
England as a proportion of the total population.

www.geocities.com/mcewanjc/scotsr1b.htm

This gets back to what needs to be done to resolve it? Well your summary
helps. If I ever get finished I can use the phase 3 analysis to directly
answer some questions raised from my previous analyses, but dates are
still a problem and have to be tackled at some point. Jim Wilson and
David Faux's SNPs help because they reduce the noise level and
uncertainty inherent in working with R1b.

Well this e-mail will stir up the hornets

Cheers

John McEwan







This thread: