Border Reiver Deep Ancestry
(Modes of Interpretation)
Oxford Ancestors Approach
The Oxford Ancestors method of interpreting "deep ancestry" seems guided by the
1) Y chromosome DNA profiles, called "haplotypes", mutate too often to distinguish
geographical origin among them when they belong to the same class or "haplogroup".
2) Since R1b prevails among the indigenous population of the British Isles, then any Briton
who belongs to R1b is considered a descendant of that population until proven otherwise.
3) OA's customers are divided into one of several "Tribes of Britain", each of which is
implicitly associated with a particular haplogroup.
The bar graph below shows how OA might interpret our Border Reiver DNA profiles.
The "Celtic", "Anglo-Saxon / Danish" and "Norse Viking" options are well-known among
OA customers. (The interpretation of all J/J2 haplotypes as "Ancient Roman" was inspired
by the anecdote of one OA customer who claimed that Bryan Sykes suggested a Roman
origin for his J2 haplotype. To our knowledge, OA has no actual "Tribe of Britain" called
The "North Sea Celtic" Approach
Population geneticists have observed high proportions of the R1b haplogroup in all parts
of the British Isles, with the highest appearing in Western Ireland and Wales. Since the
Basques of Spain exhibit similarly high frequencies of R1b, scientists concur that R1b
is peculiar to the aboriginal population of Western Europe. They also seem to believe not
only that R1b populations have existed in Europe since the Paleolithic, but that they have
largely stayed put ever since.
That is, perhaps, an oversimplification. The fact that R1b is indigenous to Western
Europe is not incompatible with the likelihood that distinct tribal groups with high
proportions of R1b have moved from one place to another within Western Europe for
thousands of years. The highest frequencies of R1b are found among isolated communities
like the Basques or the Gaelic Irish not because they alone represent the Paleolithic heritage
of Western Europe, but because they are isolated. They simply have not received the same
infusion of "new blood" from other areas that characterizes other parts of the region.
There are substantial proportions of R1b in Germany, Italy, Switzerland, France,
The Low Countries, Denmark - even in Norway and parts of Eastern Europe. Although
R1b is not quite as common in these locales, it often comes very close. In fact, among
the top ten highest match frequencies in YHRD for the basic R1b modal values
(DYS # 19=14/389i=13/391=11/392=13/393=13), four fall in The Netherlands and
one falls in Denmark. The rest fall in Ireland, Spain and England. R1b has almost
certainly been present along the North Sea coast of Continental Europe since the
Paleolithic as well.
Since we can assume R1b was at least as common across the whole of Western Europe
one or two thousand years ago as it is now, there is no way to tell how much of the R1b
in, say, England is derived from the Paleolithic natives of Iberia, from the Celts of
Switzerland and the Rhine Valley, or from the Saxons and Frisians of the Western
The only way to guess which types of R1b have a longer history in some locales than in
others is to identify where those types of R1b represent a higher proportion of the total
R1b in the population. A number of DNA genealogy researchers have recognized that a
significantly higher ratio of R1b haplotypes with DYS390 marker values of 23 occur in
Teutonic countries such as Germany, The Netherlands and Denmark than in areas such
as Spain, Portugal, Ireland and Wales.
One researcher, Alan Foster, has dubbed this variant "North Sea Celtic". Below is a table
that attributes "North Sea Celtic" Border Reiver DNA profiles to an Anglo-Saxon rather than
an indigenous "Celtic" population.
13.75 percent of the Border Reiver sample could be classified as "North Sea Celtic".
"North Sea Celtic" Plus I1 Subclade Analysis
Another researcher, Ken Nordtvedt, has analyzed the comparative frequencies of
I1 and R1b haplotype variations within the European samples of YHRD (or the Y
Chromosome Haplotype Reference Database). He estimated the total number of I1
and R1b haplotypes in each geographical area (e.g., Norway, France, Northern Italy),
and used each total as a denominator to calculate the proportions that exhibited
Like Mr. Foster, he discovered that the highest ratios of R1b haplotypes with DYS390
marker values of 23 fell among countries with a long history of Teutonic occupation -
such as Germany, Denmark, The Netherlands and so on. His study of I1 haplotypes,
on the other hand, has divided I1 haplotypes into approximately five subgroups.
Two, which he calls H1 and H2, exhibit their highest frequencies in parts of continental
Europe with a strong Germanic past.
Two others, however - H4 and H5 - are found most often in Finland and Scandinavia.
These, he implies, express a Norse origin when found in a person of Western European
descent. A few of the Border Reiver haplotypes have formally been assigned to, or
estimated as, I1. However, many more that were estimated only as "I" are actually
"I1", and some of these are probably Norse rather than Anglo-Saxon. The primary
distinguishing marker values (at least for those markers available in our Border
Reiver DNA profiles) are DYS390=23, DYS19=14 (mostly) and DYS385a,b = 14,14
or 14,15 or (possibly) 13,14.
Our "deep ancestry" table below has been supplemented in accordance with Dr.
About 4.3 percent of the Border Reiver haplotypes could be classified as "Ultra-Norse".
Haplogroups K2/N/P/Q/R/R1 Counted As "Norse"
Several researchers, like David Faux of the Shetland Islands Project, have suggested that
some haplogroups of Central or Northern Asiatic origin should be interpreted as indicators of
Norse paternal ancestry when found among individuals with roots in the British Isles.
Haplogroups N and N3, which are found among Finns, Saami and other largely
Uralic populations, very likely entered the British gene pool as an admixture among
Three other such haplogroups include K2, P, R1 (found mostly in Central Asia, R2 (which is found
most often in India and Pakistan), and Q (which is found in the Shetland Isles, Iceland, Norway,
and among other European populations with possible Central Asian admixtures, including
some Jews and Croatians).
Like R1a, which is also thought to have originated in Central Asia, these haplogroups
may have been absorbed by Heruls, Rugians and Goths and other Baltic peoples during
their wanderings near the Black Sea early in the first millennium. When these peoples
returned to the Baltic Sea, they brought these haplogroups with them.
Such haplogroups may also have been brought to Scandinavia by prehistoric Indo-
European invaders like the Kurgans or even the Indo-Aryans.
Finally, haplogroups such as Q, which is common among the Evenks and other Siberian
tribes, may have been introduced by Norse interaction with such nations as Bjarmaland,
which is now called Permia and is adjacent to Siberia.
The chart below has been augmented to reflect this interpretation.
About 1.1 percent of the Border Reiver sample belong to one of these haplogroups.
Since surnames did not become common practice in Britain until after the 12th
century, two things must be true.
1) The linguistic origin of the surname reflects the time and place where the
surname was acquired, not necessarily the "deep ancestry" of who acquired it.
2) Since lineages precede surnames, the two cannot be identical. Many persons of
different lineages may have acquired the same surname, and many persons of the
same lineage may have acquired different surnames.
Examples of surnames that may not reflect a shared genetic lineage are:
1) Place Names - such as the names of towns, estates, or physical features of
2) Patronyms - in other words, surnames based on the given name of the father.
These could have been acquired by many persons whose only relationship is
that their fathers had the same first name.
3) Occupational Names - such as Baker, Walker, Carpenter, Smith, and
so on. These are acquired by persons whose only relationship is a
In spite of the essentially non-genetic nature of surnames, some of them are associated
with old Scottish or English families with a known national origin. Just out of curiosity,
we will total up the Border Reiver DNA profiles by probable surname origin.
The resulting chart will provide an interesting "control" in our survey of geographical
origin as determined by DNA.
The patronymic names could easily have multiple origins. Therefore they were grouped
separately, even though some reputedly identify specific Celtic, Anglo-Saxon or Norse
families. The Anglo-Saxon word construction of these patronyms may tempt us to classify
most as Anglo-Saxon, but the very strong tradition of patronymic names among both
Scandinavians and Gaelic Scots suggests that many may simply be Anglicizations.
Two of the occupational names are English - Turner and Chamberlain - but strong
identification of Taylor, Forrester and Hunter with Scottish families dissuade us from
classifying them as Anglo-Saxon.
(NOTE: We have assigned the surnames to these categories on the basis of research.
Unfortunately, there is often considerable disagreement about their origin. If any of
these assignments appears incorrect, please contact us - but be nice, any mistakes
Maximum Norse Ancestry
We have prepared tables for each and every one of our Border Reiver haplotypes that show,
not only the locales where matches were found in the Y Chromosome Haplotype Reference
Database, but also the frequency of such matches in each locale.
The locales are sorted in descending order of match frequency, and those that appear at
the head of the table may represent the possible source of each DNA profile.
Of course, many haplotypes score matches in multiple locales. There is also a danger of
misidentifying matches with haplotypes that actually belong to different haplogroups, as
the DNA profiles in YHRD are not formally identified by haplogroup and do not include
certain STR markers - such as DYS388 - that are especially helpful for discerning the
true haplogroup of a given profile.
As a result, there is no surefire way to determine which of those locales - or the overall regions
that contain them - represent the ultimate origin of a given haplotype.
Hence, we can only speculate. We can hypothesize in one direction, and then in another, and
perhaps derive a semblance of the truth from multiple perspectives.
In this section, we will propose a "what if" scenario, and quantify its implications. We will
supplement the haplotypes that have already been attributed on the basis of haplogroup to Norse
ancestry with all apparent I and R1b haplotypes whose top three match regions include Norway
or Sweden, or both.
Such haplotypes include most I1, some I2a, and a good portion of I2b, amounting to 14.2 percent
of the Border Reiver sample. They also include at least a third of the total R1b.
Lastly, we will include all our G and G2 entries. G and its subclades are thought,
like R1a, to have originated among the Central Asiatic Indo-European peoples, and G2 (at least)
is occasionally found among Scandinavians. Such haplotypes comprise 1.4 percent of the total.
This will represent a "hypothetical maximum" of Norse ancestry among the Border Reivers.
The actual percentage of Norse ancestry is unknown, but may be simulated based on two premises -
all R1a in Britain is of Norse origin, and the proportion of R1a in Norway 1,200 years ago matches
that in Norway today. Since approximately one third - 34 percent - of Norwegian Y-DNA haplotypes
belong to R1a, we can estimate the percentage of Norse ancestry among the Border Reivers by
multiplying the percent of R1a by three - 3.7 x 3 = 11.1 percent.
Maximum Anglo/Danish/Flemish Ancestry
Three major population groups that migrated to Northern England and Southern Scotland between 500
and 1500 A.D. are the Angles, the Danes, and the Flemish.
The Angles invaded Britain in the 6th century and conquered most of eastern Scotland as far north as
the Firth of Forth, as well as large portions of Cumbria, Dumfries and Galloway. The name "Dumfries", as
a matter of fact, is Celtic for "Hill of the Frisians". The Angles were especially numerous in what became the
East Marches of England and Scotland. The English East March is composed primarily of Northumberland,
while the Scottish East March includes the ancient kingdom of Berenicia, which was a co-dominion of Angles
The Danes first invaded eastern England in the 9th century, and eventually controlled a large
portion of England called the Danelaw, which ranged from East Anglia to the southern part of Northumbria.
The Flemish were a people of mixed Celtic and Teutonic descent who resided in Flanders. Unlike
many Northern European peoples, the Flemish were able to repel the Viking invaders of the early middle
ages - and in the process developed a formidable military aristocracy. Flemish nobles later established
alliances with the Normans, which resulted in many nobles - like the Bruces - settling in Normandy and
eventually joining the invasion of England in 1066. After the Norman invasion, Malcolm Canmore and his
successors invited Norman and Flemish nobles into Scotland to modernize the infrastructure of the Scottish
nobility with continental European innovations like feudalism and chivalry. Trade with Flanders was also
encouraged, and eventually many Flemish merchants and weavers emigrated to Scotland to help manage
the burgeoning woolens industry.
Although the Angles, the Danes and the Flemish arrived in Britain under very different circumstances, they
all originally came from the marshy coastline of the North Sea and were very similar peoples. It is extrremely
hard to distinguish genetically among the descendants of the Angles, the Danes and the Flemish, so they must
be considered together.
The chart below combines the R1b DYS390=23 haplotypes and the "I" type haplogroups, which others
have already attributed to a largely Anglo-Saxon or Danish (what Capelli et al. call "invader") origin, with
any other Border Reiver haplotypes that count among their top three YHRD match regions *
Denmark, Germany, the Rhineland, the Netherlands or Belgium.
Such haplotypes include all I - which comprises 20.5 percent of all Border Reiver haplotypes. The remaining
49.6 percent includes all R1b DYS390=23, most R1b DYS390=22, and much R1b DYS390=24, including the
Western Atlantic Modal Haplotype itself, and a few R1b DYS393=12 haplotypes. This represents our
"hypothetical maximum" of Anglo-Saxon, Danish and Flemish ancestry among the Border Reivers.
The actual percentage of Anglian, Danish and Flemish ancestry is unknown, but may be approximated
based on the following premises:
1) All R1a in Germany or Denmark is of Norse or Slavic origin, introduced after the 6th century.
2) No more than 3.1 percent of all Border Reiver results - or 28 percent of the 11.1 percent likely Norse
total - are Norse "I" haplotypes.
3) 50 percent of the original Anglo-Danish stock was R1b, and 50 percent was I -
hence, the total Anglo-Danish R1b must equal the total projected Anglo-Danish I.
The resulting calculation would yield this: (20.4 - 3.1) x 2 = 34.6 percent.
Maximum Celtic Ancestry
We mean by Celtic, of course, any descendants of the Paleolithic aborigines of Britain who came
originally from the Ice Age refugia in Iberia and Northwest France, and eventually acquired the Celtic language.
We also mean the descendants of any real Celts who migrated to Britain from the Celtic areas of continental
Europe, such as the Rhineland and the Alpine regions. The Amesbury archer, a 2500 year old Briton whose
remains were found near Stonehenge, may have been one of those Celts, as a chemical analysis of his teeth
indicated that he grew up in Switzerland. This group will contain all R1b haplotypes, which comprise 67.5
percent of the entire Border Reiver sample, and all I2b haplotypes, which comprise 5.5 percent - as well
as any I2a2 haplotypes we are able to identify, as these, like WAMH, are also associated with Basque
and Irish populations.
This will represent a "hypothetical maximum" of Celt-Iberian ancestry among the Border Reivers.
This actual percentage of Celtiberian ancestry may be approximated from the total percentage of R1b,
minus the projected actual percentage of Norse R1b (about 3.4 percent) and the projected actual percentage
of Anglian, Danish or Flemish R1b (about 17.4 percent).
The resulting calculation would yield this: 67.5 - (17.4 + 3.4) = 46.7 percent.
The Issue Of "Neolithic" J2 In Britain
The interesting paradox about "Celtic" DNA is that, although it easily comprises the vast proportion of
the Border Reiver DNA profiles, it is restricted to a much narrower range of haplogroups. It consists largely of
R1b. Even the inclusion of "I" haplogroups or subclades such as I2b and I2a2 as "indigenous" groups is
J2 haplotypes may also have come to Britain during the Neolithic expansion of agriculture, and would
therefore have been absorbed into the eventual "Celtic" populace. But the known source of J2 haplotypes
provided by the half-millennium of Roman occupation overshadows, in our eyes, the purely theoretical input of
Neolithic J2. The assumption that agriculture must have been brought to Britain by J2 individuals from the
Middle East also contradicts the Migration vs. Cultural Diffusion model which the theory of Neolithic J2 was
introduced to support. The same thinkers who believe that all R1b in Britain has been there since Paleolithic times
tend to believe that all J2 in Britain has been there since Neolithic times. That would make both R1b and J2
Yet the J2 did come from elsewhere. If these thinkers - known in archaeology as "Immobilists" - believe
that the Celtic language was introduced by cultural diffusion rather than invasion, with few if any continental
"Celts" entering Britain, then why can't they believe that agricultural technology could also have been introduced
through cultural diffusion, without the actual settlement of Middle Eastern immigrants on British soil? Material
technology is, if anything, less intrinsic to a culture than language, and therefore more readily passed from one
society to another without the actual incursion of a foreign people.
For the above reasons, we on the Border Reiver DNA Project reject the assumption that most J2 in Britain
is indigenous in origin. Other factors support our conclusion. For instance, the uneven distribution of J2 in
the British Isles, especially its relative absence in the Scottish Isles and most of Ireland, suggest that it has
been imperfectly diffused, indicating recent arrival. We believe that J2 most likely came to Britain starting
with Mediterranean trade before the Christian Era, and increased during Roman occupation.