Sunday, November 10, 2013

First Genetic Relative with the Williamsons

For reference here is my paternal line with wives: Albert Thompson & Ida Williamson -> Ray Thompson & Ina Finks -> Charles Thompson & Elizabeth Seelye -> my dad -> me.

Although I have a few genetic relatives I believe are related to my Williamson family I haven't had anyone that actually contained my Williamsons in their tree..until last night. Hunting down a lead on some possible Williamson related people from Kentucky, I unexpectedly ran right into the Williamsons.

The shared ancestors would be Alden Williamson (b 1750) and Isabelle Thompson (my other Thompson family).

Now, not too long ago, I would have been jumping up and down about this, but now I'm a little reserved.

What is going against it?


  • There are references to known maternal relatives of my father in this match's tree. Farther back in time but still present are the Fulkersons that are related to my Seelye family. 
  • The match is over two chromosomes and each match individually is pretty small. They are individually small enough that they don't necessarily stand up on their own.
  • There are no known overlapping relatives..just a group of matches I haven't figured out yet. One of them is a group I only have a single tree for and that is the tree that contains a big east coast family with the Elmer family in it...also very active where my Seelye related family lived.
  • I may see false positives from Virginia that are really related to my Finks family (Ray Thompson's wife). They have a long history there and not all of it is known.
  • Also as I said in my post about our Y breakdown there is a non-paternal event somewhere between my grandfather and Albert Thompson as my grandfather effectively does not match the Y of his paternal uncle (Ray Thompson's brother). So it may be that my grandfather is not related to his father Ray Thompson at all...which is our connection to the Williamsons.


So, those things give me pause. What do I think this match has going for it?


  • Autosomal DNA is weird, sometimes small segments are handed down intact for generations and other times they get whittled away bits at a time. So those small match sizes may be a whittled down representative of a solid match.  
  • This is the first time the Williamsons have been named. It's the first time I've found them in the family tree of a genetic relative, unlike my grandmother's family which I usually find in any family tree...including those I know to be paternal matches. So I have a lot of history of false positives for my Seelye family.
  • Circumstantially, I see a lot of activity surrounding places my Williamsons lived in Kentucky including a string of overlapping matches that seem to share family there.
  • Although I think it is unlikely and I'm pretty resigned that we don't have a Thompson Y, I should say that I haven't yet tied the Indiana Thompsons to another Thompson family either. They are much, much more likely to connect to a Thompson family..but we just haven't tied that one up yet.
  • Although our Ys don't match up, I have a gut feeling that Ray Thompson is the father of Charles Thompson, but is likely not the son of Albert Thompson. Maybe it's because Ray looks so different to me from his brothers or that my grandfather looked a lot like Ray when he was older. Maybe it's because I can see my cousin's features in my grandfather and Ray. It's a hunch. Logically, I can say the Finks/Thompson match up is the most likely break given the odd circumstances surrounding it but those circumstances may also be the best thing they have going for it. For all intents and purposes Ray thought he was my grandfather's father and Ray's family took responsibility for my grandfather and raised him.

So basically I have a hunch with added weight from circumstantial evidence and this finding in a genetic relative's tree, that make me think the family picture in this post: http://thompsonhunt.blogspot.com/2011/02/thompson-photo-1927.html is an accurate depiction. Ida Williamson is there with my 2 year old grandfather (her grandson) on her lap. Her son Ray is in the back row looking like a serial killer. She is Ray's mother, but Ray's father is.. is not pictured.

Hopefully this is one of those things that can be resolved with 23 and me testing from my Indiana Thompsons who would share Ida Williamson as a great grandmother in this scenario or with further testing of my Seelye family to see if this match is totally on their side of the family.

Friday, November 8, 2013

Continuing Thoughts about Migration

I was tooling around the FTDNA site checking on my DF95 results and happened into the "maps" area. Since I'm always excited about map speculation, I checked it out. It's very basic. Since I'm in R1b it has that group highlighted with a nice line going to the  general vicinity of France. What was interesting was the placement of Haplogroup I.

Pardons for the small image but this is what they have there (undoubtedly old):


What I found so interesting about this particular map was the arc that I takes and that it seems broadly similar to the path of Z18 and many of it's subgroups. Is it possible that our Z18/Z14 only group in the last post was caught up in the current of I as it intersected and headed north? Maybe better put, is it possible that the circumstances that led I to go north on that path, also led Z18/Z14 to do the same?

I don't really know enough about I and it's movements or timeframes. It would be interesting to see if they coincide in some way or if this map is just so old that it's not a relevant picture of a migration pattern.

For interest, here is a gradient map of haplogroup I M253 from the Wikipedia:






Monday, November 4, 2013

458.2 Cluster Bomb..and thoughts about migration

I'm unsure about the continued relevance of my own Y DNA to the search for the Thompson family, but I am hopeful that it may yet lead me to some conclusion about my grandfather and his male line. So, we march on.

What you see here is really just a gathering of ideas for my own benefit. Please don't take it all too seriously as new things are discovered every week and I'm an average person trying to make sense of things the best way I know how. I am likely neither fair nor balanced.

For reference, here is a quick chart stolen from the Z18 and subgroups page:



What you can see there (click it for a larger view) is R1b at the top. That is a huge group of men in Europe. It breaks down into a few groups, the two here are P312 and U106. P312 is the larger group and is roughly labelled "Celtic" by many people because it trends towards areas where Celtic culture was prevalent. That is the branch I think many Thompsons would expect to be on. We are on U106, a large group but smaller than P312 which seems most prevalent in "Germanic" speaking areas today.

U106 breaks down into a few groups, but our focus would be people who are Z381 and people who are Z18. Z381 is the lions share of U106. L48 underneath that is roughly the largest group with many subgroups underneath it. We are down the smaller sibling group of Z381,  Z18. Beneath Z18 you have Z14 which leads to a child Z372 and then Z372's child L257. Z14 has other children, L343 and DF95.

We are Z372- so we remained in Z14, clustered with like individuals until just recently.

We got our news about the DYS458.2 in our Z14 Cumberland cluster. Some have it and some don't. Oddly (to me) the group that has the 458.2 is larger than those without. The TMRCA for 458.2+ and 458.2- is around 1300 years ago..so around 700CE. Hey it's on this side of the BC/CE line!

Several of us tested for DF95..my results are due next week but it is probably a "given" that I'm positive because multiple members of my cluster are also positive. So DF95 occurs in both 458.2- and 458.2+. The age for DF95 was given as about 1500 years ago, so 500CE. It came first before our 458.2 split.

So within a few hundred years of DF95 SNP we develop a DYS458.2. Not too shabby.

DF95 is a brother to Z372. We know they don't go together because multiple DF95 people are Z372- and at least one Z372 is DF95-. So they evolved separately from Z14 (the parent of both DF95 and Z372). 

Z372 is considered a pretty Norse leaning SNP. DF95 so far since it's basically only been found in Cumberland seems to lean to the South side of the Baltic and North Seas.

I have made mention that the DYS458.2+ seems odd to me because it covers more ground than the original flavor (DYS458.2-). That could be lack of sampling. Maybe we're just missing men from Poland who should be 458.2-. It's odd though in that general east to west migration model that the parent of a group would only occur in the western half of the area covered by the child group. Of course, I am a simple person so simple migration paths make more sense to me, but people rarely do things in a simple way.

So I decided to step back and look at Z18 (at least our side of the house) and try to get my bearings and wrap my head around what I'm seeing. This is what I gathered:


z18+ z14- (so these are the ultimately Z18 individuals not down any known path). This isn't based on frequency so there may only be one person in each place. There are very few Z18 Z14- people. There again, a parent group is much smaller than the child group.

France
Switzerland
Scotland
Germany
Belgium

Here's a map I made with google maps




A fairly West leaning and continental group with pretty good coverage North to South. To me that is kind of striking, Z18 is a major (if relatively small) child of U106 and U106 is pretty wide spread in Europe trending towards Germanic speakers, certainly occurring in these countries with good representation. It's the Westerly lean that strikes me for this group. That, and it doesn't seem to crowd around the North and Baltic sea so much mainly because there just aren't representatives from the Netherlands, England, Denmark, Poland and Norway currently. 


It's hard to make much of this group because it's so small. If Z18 is roughly 3000+ years old these men would have been part of multiple migrations over a long, long time. Someone who was U106+ mutated to Z18 and his children went on to be about 1/3rd as large as the Z381 population. Z381 mutation and Z18 mutation are siblings on the tree, but could be separated by many centuries and a lot of geography.

So here's the most concise map I could find for 1000BC: 



That's our list of suspects for Z18. Modern continental Z18 (everything else negative) today seems to occupy the part of the world labelled "Celts". There is great danger for error in assigning current geo-locations for 3000 plus year old DNA. Especially from a single sample form each country. For fun...Lets leap forward. Here is a map from 500BC:



So at 500BC we're starting to see things we would recognize today. The Teutonic peoples have split up into Scandinavians and Germans and pushed south into the Celtic world (this idea is based on language studies I think...along with archaeological evidence). Z18 would be well established somewhere in this world. Jumping forward, Here's a simple map of Europe around 200BC also stolen from timemaps.com:



So, somewhere in that great unwashed world, maybe labelled Celts in 1000BC or  "Gaul" or "Germans" in 200BC,  Z18 came into being in a single U106 man.  The people who don't diverge end up, thousands of years later in those modern countries from the first map above. Who were the original Z18 family? Who knows, there are a lot of "Germano-Celtic" tribes generally in that area and without actual remains tested, how could anyone say?

There are other migration theories as well. I'm taking a position today and thinking about it as static, but I've heard that you can look at the fallout of a genetic group (where they ended up at some geographic barrier) and try to bring the dots back to a point. In the case of Z18, maybe we guess came from the east because they are peppered to the west. Maybe they hit the alps to the south and the ocean to the west and called a halt.

Also good to remember that at this time on these maps you already have Z14 in the mix since it is about 3000 years old. Also, what we know most about the world is coming from those two colorful groups at the bottom of that last map. The Greeks and Romans. Lots of grains of salt to be had.

Anyway, it's speculative at best but interesting to note that we're lacking our friends in Scandinavia and Poland and the Netherlands. So people were obviously on the move, but what got them moving?

Here is a nice map of Europe around 30BC taken again from timemaps:



Here is a map from 200AD:




Facing this red swarm, some groups are going to be okay, but very obviously a lot of people are going to die, be displaced and invade other territories in the face of an organized superpower like the Romans. At this point (200AD) we may have Z372 branched off from Z14 and maybe L257 branched off from Z372? I think Z343 would also be in play with it's 1900 year TMRCA.



Z14+ (Child of Z18). These are people who have not tested positive for down stream SNPs.

England
Netherlands
France
Scotland
Sweden
Germany
Ireland
Denmark
Wales
Finland
Poland
Switzerland

Here's a map of that area. Keep in mind that I'm outlining modern nation boundaries that obviously did not exist, so the map is way off and also again may be missing representatives from specific nations. As a for instance, the total area of Sweden is not inhabited by armies of Z14 men..the populations hug the coasts there.




So this is the big child of Z18..Z14. Z14 is estimated to be about 3000 years old. So these people who tested Z14 and nothing more, had 3000 years to get where they are going and it's a lot more geography and more people than Z18. 

Cumberland was in this group until a few weeks ago. Here we have the same group of Z18 countries but with many more nations added and a more familiar pattern that crosses the North and Baltic Seas. The Scandinavian presence is much larger here and the range is pushed east by Poland and Finland. I'm discounting the contribution of  the isles because of the bias towards Isles testers and because it's one of those physical barrier destinations that are likely to get people all bunched up. My guess is that we're missing some testers from Norway and Belgium.

Again, when thinking about history for the maps here, Z14 and Z18 have an overlapping range and timeframe. Z14 appears to be very popular, leaving it's parent in the dust. Z14 also seems to have a classic "Germanic/Baltic/North Sea" spread you'd expect from a descendant of U106. I may be doing a disservice by not having exact geo-pins for these people. Maybe I'll tackle that later. On to the children of Z14.

Z372+ (but nothing more) which includes

Netherlands
Scotland
Sweden
England
Italy
Norway



Again, this map exists without regard really for number of people in a place or downstream SNPs like L257..and realizing that people are likely missing. This is a rough map of current nations for Z372. This group is for obvious reasons thought of as a Norse group. There is a member from Italy that is not shown on the map. I don't want to forget about them! So here is a map of viking settlements that includes Italy:



I think the general theory on this group (except for Italy) is conveyed in this viking invasion map stolen from the BBC's primary school history site:


It would be easy to see the Baltic as a squishy border for Z372. A Z14 person maybe in that fluid area becomes Z372 and carries it's popularity north and west? This group I think is what one might expect from a child group. The geographic area covered by Z14 is sort of "halved" by it's child Z372.



East Anglia Cluster L343 (for the moment seems to define the "East Anglia Cluster" and is an SNP that occurs in multiple clades) DF95-, Z372-

England
Ireland
Scotland
Germany


It's a decent sized cluster but so far pretty specific and I think the current theory on this group is summed up by this map stolen from Wikipedia:


EAC-L343, again seems to live in a subset of areas where Z14 exists. Roughly also heading west, likely from the first East Anglia Cluster-L343 man...maybe living in Denmark or  Germany?

When looking at this anglo-saxon invasion map, and the one above, it may be helpful to note that the broader groups all the way up the tree to Z18, U106, Z381, P312 are also going to be represented in these migrations. Genes don't always favor political or social boundaries. You may be more likely to run across P312 in France but it doesn't mean the others are absent...just maybe not as popular. 

Also the Romans were great movers of people. So you may note that where U106 and it's children have a great foothold in England is also where Rome had a great foothold in England.

Okay, so we are not part of Z372 and haven't tested positive for L343 (that I know of) but I think it's good to look at them as examples of possibly narrowing the geographic locations of some of these Y chromosome groups. So far, it looks like Z372 is the sibling of DF95.

We are a part of DF95..which may come to define our Cumberland cluster. I'm going to cheat because I've had a long time to look at my matches and I'm pretty sure my matches from Denmark and Switzerland would also be DF95+. I'm going to cheat in another way. I know the group is split by DYS458.2, but I'm ignoring that here:

DF95+
England
Ireland
Netherlands
Wales
Poland
Germany
Norway
Belgium
Possibly Denmark (my matches at ancestry and Ysearch Peterson and Hansen)
Possibly Switzerland (my matches at Ysearch Steiner)




The DF95 (and possibly CTS2206+ based on a recent test result) group is basically the Cumberland cluster at this point. Here we are, found pretty much everywhere Z14 is found but we lack the Finns and the Swedes that are so prominent in Z14 and in Z372. The Corsons are in this group and they may well be Swedes but they may also be from the Netherlands so it's hard to tell. We're missing France too. All of that could be lack of testing. The best guess is DF95 crops up around 500CE.

Here is the super rough map of Europe around the time DF95 is thought to have popped into existance in some Z14 male:



If you zoom in, you'll see the Frankish state, rough Saxons and Angles and Danes. Other tribes are listed but this is well into a movement of people during the Germanic Migration period. It would be hard to pick just one. Also, 500CE is a really rough guess based on the differences in the STRs of our cluster. DF95 may be much younger or older once more data is available. 

There is also room for DF95 (or CTS2206) to grow beyond our cluster as there are still a few Z14 groups out there who have not tested for it yet like the Polish cluster, the Finland cluster and the Z14* cluster..

Another point of interest I want to spend time on here is that  we're vastly outnumbered. The other branches of U106 dominate all these areas. Although Cumberland as a whole is large for Z14 groups..it's small in the context of the rest of U106 and it's really spread out. 20 to 30 individuals represent that DF95 chunk of the planet. And again, that current map doesn't represent one movement but 1500 years of movement.

For the next two breakdowns I need to remove Steiner and Switzerland from my list because although I suspect they have a DYS458.2 (Genebase refused to give a value for their DYS458 saying it was an "anomaly" and would not be helpful for matching), I'm not certain they have one so they are out. Booo.

458.2- (Cumberland Cluster Members who do not carry the DYS458.2 mutation. They are DF95+)
England
Netherlands
Ireland
Possibly Netherlands or Sweden



This group, in yellow overlaying the green DF95 image, is the most likely parent group for 458.2+. It likely includes the Corsons who are either Dutch or Swedish. It's very small. It could be that we just don't have enough samples or it could really define a home range for this group. I think the general feeling for Cumberland A..if we're not thinking about more recent migrations from Belgium and the Netherlands to England and Ireland (which are still a good possibility) is the same Anglo Saxon invasion of England that appears to be played out in L343. I don't know that the Saxons were particularly involved in Ireland up front but they definitely became involved as did the vikings from Norway and Denmark.


It's also very good to note that just like Z18 Z14- above I'm making assumptions and lots of them based on a small group of people.


458.2+
England
Netherlands
Wales
Poland
Germany
Norway
Belgium
possibly Denmark (Peterson match at Ancestry known to have DYS458.2)



I've lazily included Wales in a lot of these maps but it bears special note here because it's specifically listed an it's the second time I've run into it. The Edwardsons from Malpas England also have a firm belief that they are Welsh.

The General feeling for our entire cumberland DF95 group, I think, was an origin point in Poland moving west, but the 458.2 results cramp that style for me right now. It's possible that we're just missing 458.2- Polish people. Truthfully though, I would have expected them to be the majority of 458.2- people, not totally missing. The same thing for the Germans. I wouldn't expect 458.2 to be the majority group but it seems that way. Again, vastly outnumbered by other U106 people, but instead of sweeping cleanly west we seem to go in all directions. We cover nearly the same amount of ground as the Z14 only group. So the parent 458.2- group is the small more defined group and we are basically still most places Z14 was.

The guesstimated date for 458.2 is around 700CE or about 1300 years ago..very very roughly. Here is a rough map of Europe around 700CE:



The Anglo Saxons are in Britain presumably with some 458.2- cumberland A members along for the ride and somewhere on this map a single man gets a 458.2+ and becomes pretty prolific. In simple terms, I would look for that mutation to pop up in an area where Cumberland A lived. England, the Netherlands and Ireland. I harbor a lot of doubt that 458.2 cropped up in England or Ireland and then made it's way to Germany and Poland.

Alright. So up until now I haven't really looked at where there is clumping going on. Where do the highest numbers occur. For this, I'll only use the knowns.

England gets 5-6. (UK could be anywhere in there).
Poland 3
Germany 2
Belgium 1
Netherlands 1
Norway 1
Wales 1

I imagine we're missing a lot of data from the Netherlands and Belgium I think as they have projects for Y DNA that have been a bit closed..or at least prohibitive  in the past. We're very lucky to have our friends from Poland to give us balance in an Anglo-centric testing environment. It would be hard for me to say that this group originated in England and then moved to Poland and Germany just because of the little bit I know about England. It is more of a destination country.

For comparison here is he Z372+ (everything else minus) crowd:

Sweden 10
Norway 4
Scotland 3
England 3
netherlands 1
Italy 1

Part of the bulk of Swedes comes from the Scandinavia cluster which is heavily Swedish. Here again though, can I take the sheer number of testers to equal origin? I can say the numbers are high in places with Scandinavian influence but does that mean Sweden is the root or could it be one of the other countries. Knowing that Sweden is a source country and a people that are well traveled, that is exactly the temptation. The high saturation in Scandinavia makes England and Scotland easy to understand.

What does it mean to us in DF95 for migrations and sources though?

I'll pick on the "Swede Cluster" here. They are Z14+ and have members from England, Ireland, Wales and Denmark and a TMRCA of about 1600 years ago...so about 400CE, not too distant from the guessed origin date for DF95 but they are DF95-.

England 3
Denmark 1
Wales 1
Ireland 1

Looking at that distribution, I would still guess that England has a testing bias and Denmark is closer to the source..maybe it's of note that there are several Sweets in the East Anglia L343 cluster...Angles being from Germany/Denmark and all and here is the Swede cluster with a Swett.

Again to me it seems fairly direct for them. They cover a lot of countries but they could easily be categorized as likely Anglo Saxons or "Danes" which are kind of hard to tell apart.

So again, looking at distribution for all of DF95 458.2+ or - I have to wonder are we thinking of a model of source country west or are we thinking of a model of least population to most at some barrier? Could the barrier on the west be the Atlantic ocean and the Irish? Could the barrier on the East be the Slavs?

Poland west. The assumption I would make here is that we're just missing a few Polish testers who are DYS458.2-. There are fewer of them so by the numbers it would be harder to get them fom Poland than it is to get their English speaking counterparts. A similar assumption could be made for Germany. We just really haven't seen them yet but we could in the near future. Z14 exists in Poland and it's a nice place to branch south and continental and North to Scandinavia meeting up at the Isles.

England East..believe it or not English speakers did travel east to serve in armies for foreign kings. It's possible if unlikely that they originate in England and travel east. Gut feeling is that, that is very unlikely.

Germany west and east. Okay there are a few ideas I have for this. DF95 crops up in Germany..ish in an anglo saxon stronghold and moves west with the angles and saxons and jutes and moves east at a later date when the saxons controlled portions of Poland or through regular German immigration to Poland. After all German immigration to the U.S. brought all kinds of culture along with Y DNA to my country why wouldn't the same movements of people head east like the Volga Germans who lived in Russia.

The Danes. Like I said, I think we're missing a couple of Danes. This is the lawn sprinkler theory. You have your Saxons, Angles and Danes. Hard to tell apart at the Y as there is a lot of cross over there. You have one movement from the area west with the Anglo Saxon invasion. A second movement to the west with the Vikings and the great Heathen Army which took over the eastern side of England and set up the Danelaw. There is at the same time a colonization of the southern and eastern Baltic that puts the Danes in a nice position to cover a lot of ground both east and west in a similar timeframe and with long lasting effects. I think the Danes are running parts of Estonia in the 1300s. They have a long and I think sometimes overlooked relationship with the rulers of Poland too. They were in a nice position to hammer england, bother the Netherlands and control the south Baltic.

Those last two may be combined really as waves of people from the area of Denmark and Germany would have moved both east and west in search of better things and more opportunity.

I haven't forgotten our guy in Norway. Denmark also was connected through trade and administration with Norway. The Sagas (you can laugh now) at a minimum show that the Danes, Swedes and Norwegians were definitely working together to run that great Heathen Army around England and Attack the Franks and Attack the Slavs. By Attack, I also mean trade, farm and colonize too.

So there are my two arguments. We're looking at a nice East West migration from Poland..or maybe even further east and south a bit. Or my favorite today, we're looking at the same Anglo Saxon and Viking movements that would bring Z14 and DF95 men both west and east.

On top of that when I think about the study showing that the Danes used Polish mercenaries in their armies, I'm even more convinced of the relationship and the ability of the Danes to move men around their world. There are suspected Jomsviking beheadings in England..Jomsvikings being Polish Scandinavians is pretty darn interesting.

When I see the 458.2- people not including Poland as a major player but still containing the Netherlands it makes me think that we may be looking at a couple of movements at different points in time from the same general place in the middle.

The saxon boat burial in England looks just like a buried viking ship to almost anyone.

It could be the rum talking and I realize I'm favoring the Danes today and I'll likely favor the Hanseatic League or German Mennonites tomorrow but I have one of many "gut" feelings that our Scandinavian brother clade Z372 may just be the "northern" Scandinavian clade.

It will be interesting to see what further DF95, CTS2206 and CTS2207 testing bring up. It would be interesting to see another fracture in our cluster like the 458.2 divide.

To see and to speculate about.

In hind sight now having written all this, I think my block maps are crude and oversimplified. I should attempt pinpoints where possible. These ranges may not seem too extreme when looking at pinpoints. Well, that will have to wait for another day.




Edit 7/19/2015

This post is a few years old now and although I think it has good ideas, I thought I should add some new information that I've picked up from the U106 and Z18 projects with a big heap of my usual speculative ramblings.