Archiver > GENEALOGY-DNA > 2009-05 > 1243673733

From: "Lancaster-Boon" <>
Subject: [DNA] A major question in this field: archaeology, languages, mtDNA,Y-DNA, atDNA
Date: Sat, 30 May 2009 10:55:33 +0200

Hi everyone

In a recent review article I had the chance to see published in JOGG
( I hit upon an important question about a groups of
assumptions which underlie much of what is written in this field.

It is frequently vaguely assumed or stated that Y DNA distributions and
variations will reflect the paths of ancient migrations. I say vaguely,
because the following correspondences are all sometimes treated as if they
are one in the same, so that they can be used as proxies for each other:

1. The spreading of new technologies over time.

2. The spreading of major chunks of population, which would also BTW of
course be identical to the spreading of Y-DNA, mt-DNA, and at-DNA.

3. The spreading of languages within their language families.

The most important technological change in pre-history or history was
probably the development of farming, and the idea that populations were
changed massively as farmers bred and colonized new areas is associated with
the authors Cavalli-Sforza (also famous in the DNA field) and Ammerman.

Connecting this to the spread of languages on the other hand, is actually a
separate proposition (although the difference is often ignored I think by
people focussing on DNA). It is more associated with Colin Renfrew and
perhaps more recently Peter Bellwood.

While looking at the literature, including the non genetics literature, a
pattern seemed to emerge at least in the examples I was looking at...

1. The idea that the first farmers spread out and changed the population
profiles of areas away from their point of dispersal seems to hold good when
we look at data from many fields. This is true for at least the long period
of time between the first consolidation of the technological package, and
the period where it had become widespread in areas similar to its original

2. Increasingly, most archaeologists seem to accept that the spreading
happens in bursts. Once areas of a similar climate are colonized, the
package settles down and adapts to new conditions (ferments as one French
archaeologist, Guilaine puts it) and a new expansion begins. We would
intuitively expect each new expansion to show a bit less of the genetic
patterns of the first farmers from the original core area. And again, this
is indeed able to be consolidated with all the data if you look at it all.

3. The idea that languages dispersed with farmers (out of the Middle East at
least, which was the area I was reviewing) seems very weak. For example in
the Fertile Crescent you only have to go away from the generalist reviews
and look at the basic literature being referenced to, in order to see that
Semitic is thought to have been spread not by farmers, but by pastoralists,
and the same is often said about Indo-European. This makes sense because
pastoralists are more mobile and also more able to rule over sedentary
peoples. They are peoples who links the others in many ways, and so I am not
making a cliché about mongol hordes here. They played a positive role also,
and one very much involving communication.

4. Given that populations can be analysed by looking at their genetics, how
does language hold out there? Is there any link between languages and
genetics? Sometimes yes, but the cases I found were either:

a. In areas or periods where there was no farming.

b. Simply corresponding to the movements of farming technology, BUT, and
here is perhaps a controversial observation, my strong impression was that
the correspondence between farming (or other technologies) and genetics was
much more clear.

To take examples from the review:

* Chadic speakers speak an Afroasiatic language, but genetically they are
very difficult to connect to any Afroasiatic population. Their genetics and
their lifestyles and technologies correspond with other people who live in
the same area but who speak unrelated languages.

* Many parts of Eurasia show signs of having received early waves of
farmers, and this seemed more apparent to me in the genetic data than many
people might realize, even in odd cases like Northern Portugal which are
quite far from their source for this technology. But no one would ever claim
that there was a linguistic link in any evidence we have unless they were
making the purely circular argument that there must have been such a
language which is now lost.

5. Looking at genetics in more detail, I noticed signs, at least as it seems
to me, that there might be a useful distinction to be made between 3 types
of DNA. I guess I can't be the first person to state this clearly, but
anyway I don't remember anyone stating clearly, so it seems worth stating

a. Y lineages seem more mobile, and show movements of technologies,
sometimes even when the actual amount of migration might have been quite
small. In linguistics we might compare Y lineages to loanwords instead of
whole languages.

For example as Underhill and Kivisild remarked not so long ago in a review
paper, E-M35, a Y lineage shows signs of a movement from Africa to the
Middle East to Europe which does not show up in mitochondrial data for the
same period, but which happens to match a period archaeologists would be
reasonably happy to say that there was such a movement.

b. Mitochondrial DNA is perhaps most sedentary, and shows the oldest
connections between peoples.

For example concerning the Chadic speakers mentioned above atDNA as in the
Tischler paper only seems to confirm what the Y DNA says, but then along

Černý V, Fernandes V, Costa MD, Hájek M, Mulligan CJ, Pereira L (2009)
Migration of Chadic speaking pastoralists within Africa based on population
structure of Chad Basin and phylogeography of mitochondrial L3f haplogroup,
BMC Evolutionary Biology, 9:63 doi:10.1186/1471-2148-9-63

...which is still not a finished work, but which apparently claims to show
that the mtDNA shows the Afroasiatic connection.

Similar cases where mtDNA seems to shows the oldest relationships can be
named from many areas.

c. Autosomal DNA is of course the new big thing and rightly, because it is a
better for proxy for real "population".

However what seems important to remark about this is that we are not always
most interested in just knowing what ingredients are in the soup, but also
what order the ingredients were put in. Y DNA and mt DNA remain very
important then.


Having come to the above ways of thinking, I have noticed that I am in
conflict with the ideas of some people that we can treat it as a null
hypothesis that genetic haplotypes (implying large chunks of populations)
move together with languages. A lot of proxifying seems to be getting done
which might be fundamentally distorting the results of people's analyses.

I think the people who defend these ideas should put their thoughts together
more carefully. I wonder if anyone has a clear enough position to try to
publish it in JOGG? Maybe my article even helps bring the question into
focus, in order to help this case get presented.

I also wonder if the mathematicians can think of any tricky ways to test the
relative strengths of correlations for something like the geographical
spread, AND changes in spread, for any of the variables in consideration.

Best Regards

This thread: