The DENNISON DNA Project: Patrilineage 2

DENNISON DNA Patrilineage 2

The three American DENNISONs appear to have emanated from Pittsylvania County, Virginia, in the late 1700s. Even though Pittsylvania lies just to the SE of the Blue Ridge mountains, just outside the Valley of Virginia where the Scotch-Irish DENNISONs of Patrilineage 1 settled, the yDNA results shows that these two groups of Virginia DENNISONs are completely unrelated.

It is highly unlikely that the Patrlineage 2 DENNISONs are Scotch-Irish, for a number of reasons, beginning with the fact that most British DENNISONs are English, not Scottish, thus the vast majority of American DENNISONs have English, not Scotch-Irish roots.

Second, Pittsylvania County was overwhelmingly settled, not by the Scotch-Irish whose initial settlements were beyond the mountains, but by tidewater Virginians slowly trickling west. This was true of Pittsylvania's parent county, Halifax, to the east, and of its child county, Henry, to the west. And the same east to west pattern applied to the next northern tier of counties, from Lunenburg, to Bedford, to Franklin County (where I happen to live). Furthermore, these early settlers of Virginia’s piedmont, and “south side”, continued west mostly in the wake of the Scotch-Irish, progressing to Tennessee and Kentucky a decade or two later.

Third, the early settlers of these cismontane counties, including these DENNISONs, do not, by and large, bear Scotch-Irish surnames, nor are their given names confined to the restrictive Scotch-Irish set, and they don’t appear to observe the customary Scottish and Scotch-Irish onomastic (child-naming) pattern either.

Positive yDNA Evidence of an English Origin

Finally, we have a first piece of DNA evidence that England is the original home of this early Virginia family of DENNISONs, in the form of a new member of Patrilineage 2, Lee Daniels—a Londoner whose ancestors have always been English back at least to Thomas Daniel of Buckinghamshire, born say 1762. This evidence is provisional, though, for two reasons. First, and most obvious, is the different surname, and second is the fact that there is a genetic distance of 6 between Lee and the DENNISON-surnamed members of this patrilineage.

As for the surname, experienced FTDNA project administrators are finding ample confirmation in their data, of the well-known phenomenon of the MPE (Misattributed Paternity Event), where a son ends up with a different surname from his biological father. The rate of MPEs per generation in historical western societies is variously estimated at between 1-7%, depending on the place and period, and since the probabilities are cumulative, at a rate of 3%, the chances of an MPE over, say, 8 generations, would be about 20%. Not all MPEs result in a changed surname, but if we guess that this is true of half of them, we might expect to find one divergent surname for every 10-20 in a yDNA surname project, and that is about what project administrators are finding in their data.

Moreover, it’s probable that many patrilineage members with divergent surnames are being overlooked—stranded in the projects whose surname they match—first, and understandably, because we all have a considerable investment in the assumption that the surname we bear represents the ancestral “tribe” we belong to, and second, thanks to an unfortunate set of genetic distance guidelines published by FTDNA (the premier yDNA testing company), which have misled many, if not most, project administrators.

According to these guidelines, two living men tested on FTDNA’s 37-marker panel whose results come in at a genetic distance of 6 or more, are “unrelated”, although for a GD of 6, it is allowed that a person might be considered a member of the patrilineage if a “tweener” can be found. What is meant by this, is that if an additional testee can be found who is GD 5 or less with both the accepted members and with the person in question, that the latter may thereby be linked with the group.

But these guidelines fail to take into account either the fact that two people who match only loosely also have a common surname (in most cases), and furthermore, that in many cases there is strong genealogical evidence of a relationship. And when these factors are taken into account, examples of people who belong in patrilineages even though they have at best a GD 8-10 relationshp with the others, are not hard to find. The fact is, there are many cases where as many as 5 mutations have occurred in lines of descent since the founding of a surname line, which may go back as many as 800-900 years, and if there happen to be two such lines of descent within the same patrilineage, each with 5 mutations different from the other, then even a GD of 10 is possible, although to be sure one would certainly want to find a “tweener” for such a tenuous relationship.

In the instant case, the GD is only 6—well within the bounds of probability, requiring only that there have been two lines of descent since the founder of DENNISON Patrilineage 2, each with 3 different mutations. Furthermore, we have, as further confirmation, a 67-marker comparison between Lee (D-11) and Dale Dennison (D-10), with just one additional marker divergence. Although in the TMRCA chart below, I have calculated that there is a 50% probability that these two have a common ancestor within the last 510 years, the FTDNA Tip calculator lowers this to 460 years, even after factoring in that we know there was no common ancestor since 1750. And there is also the consideration that Lee Daniels is completely unrelated to any of the 54 other members of the DANIEL(S) surname project who have been tested on 37 markers or more and who share the usual British haplogroup R1b.

Although a tweener would be desirable in this case, especially since the surname is different, in my judgement the odds greatly favor the proposition that Lee Daniels belongs to the same patrilineage as the existing DENNISON members. As I say, there are any number of such instances to be found in the 5000+ FTDNA surname projects, but I have yet to hear of an authenticated case of a false positive, where a person with a different surname and a reasonably close GD can be shown by historical research not to belong to a patrilineage—although I am sure that such cases exist. Let us by all means keep our minds open to both alternatives here, as we wait for a suitable tweener to emerge to confirm the relationship beyond any reasonable doubt.

Finally, if, as I think, Lee Daniels belongs with this group, we also have to recognize the possibility that the original surname for this group was DANIEL(S), not DENNISON. All we really have here is evidence of a probable close relationship between Lee and a single rather closely related line of DENNISONs. We don’t as yet have any history on either ancestry before the middle of the 18th century that would give us some basis for choosing one surname as the original one over the other. However, the fact that Lee matches to none of the 54 other members of the DANIEL(S) project does suggest that the original surname was DENNISON. If a match should turn up between Lee and a newly tested DANIELS, it may or may not tip the balance toward DANIEL(S) as the original surname, or for the inclusion of Lee in the DENNISON group; for example, the new DANIELS might prove to be of the same descent as Lee—from Thomas Daniel of Buckinghamshire.

DENNISON Patrilineage 2 Descendancies