Hi all,
There's a lot here to unpack, so let me start from the beginning:
FTDNA's TMRCAs are likely to be significantly better than what's currently on the Big Tree. There are several reasons for this to do with larger numbers of kits, utilising new TMRCA calculation methods (including some of mine), being kept up-to-date and using additional data from STRs. The split of R-FGC12993 can therefore be dated to the first few centuries AD, while R-A321 is late medieval (probably 14th or 15th century).
>First of all, I assume that the gap and lack of other associated surnames suggest isolation. ?Is this a valid assumption?
Long blocks of unbroken SNPs can arise from several different causes. Blocks are broken only when a population is either sufficiently large, or sufficiently well tested, to show up in commercial databases like FTDNA's. Long blocks can therefore either indicate a population whose other branches have died out (or nearly died out), or a population that only exists in regions that are not well-tested. The further we go back in time, the more likely it is for branches to have died out (the Galton-Watson process), while haplogroups splitting closer to the present will be more likely to remain in the same place as their origin.
For surname-era haplogroups, this can also extend to the fraction of the family moving out of Europe (Europeans don't test as much), socio-economic factors within the family itself (which dictate whether or not the family tests), or an historic change in reproductive success: it is true that richer people had more sons surviving to reproductive age, but this is only a weak correlation, so I wouldn't necessarily link it to the events of 1424.
Instead, I would invoke the alternatives. For example, we would normally expect that long branches like this did not form in the British Isles, and only split once they arrived here. We see a lot of R-U106 haplogroups in the British Isles that date to the period between the post-Roman Germanic migrations and shortly a couple of centruies after the Norman conquest. (Some day I should make an updated graph of these.) So we might use this long block as weak evidence that the ancestors of R-A321 only arrived in the British Isles at some point in the few centuries preceding its formation. This is, as I say, weak evidence - especially if we invoke a Scandinavian origin, as Scandinavia is also well tested.
If we look at the Time Tree for R-FGC12993, we can see a lot of activity with a typical European pattern before 1000 AD, and a lot of activity in the British Isles after 1400 AD. Allowing for uncertainties in the TMRCAs, this probably indicates a generally European population before the Viking Age, and a wholesale migration to the British Isles before 1400 AD. Such generalisations across whole haplogroups are just that: general. They may not apply to R-A321 or any other specific sub-clade. But they do also give weak evidence that a wholesale migration of R-FGC12993 occurred around or shortly before 1400 AD. We know the Kincaids have been there since 1238, but it appears they and their immediate ancestors and cousins didn't flourish during the surrounding centuries.
None of this says as much as I'd like about the gap between R-FGC12993 and R-A321, but that's the nature of such gaps. This gap is sufficiently close to the present that it's likely to be filled in with future testing. Currently, we only have one test for every 800 men in the British Isles, and one for every 1200 men in Norway and Sweden. Roughly speaking, that means we respectively sample only about 1% and 0.6% of the lines that existed in 1700. Everywhere else the situation is worse. So there is definitely mileage in future testing. This remains our best opportunity for progress here.
Cheers,
Iain.
