How to graph endogamy trends in tree?

+3 votes

As I've continued to build out profiles for my ancestors and compared DNA testing results with my cousins, I've stumbled upon some things that I find interesting and/or unusual.   

I feel like there should be another way to graph or display a tree that shows additional details which show endogamy patterns of intermarriage between certain families or in geographic areas over time,  I'm just not sure how to add the meta data needed to do a specialized tree or graph that shows the intermarriage/location contexts in a way that is more intuitive.

I realize that others may not find this additional information useful or interesting, so I don't wish to make the profile too confusing or distracting.

Here are a couple of more concrete examples of what I am talking about if that helps to explain it better:

1. My Paternal great-grandmother's parents and were apparently 1st cousins, which leads to some genes from shared common ancestors being passed on more readily / frequently to descendants.   Obviously I can mention this in their profiles, but you would to read the text of the profile to realize that this occurred.

2. I have some 1st cousins who I share a great deal of matching DNA with, and others who I share very little with, sometimes on the very edge of predicted matching DNA as shown in on:

Those that I match with the most, often prove to be related to me through both of their parents, having a common ancestor, or both their parents sharing a common ancestor for my Mother and Father.  

Is there any way to mention or graph relationships which points out the closeness of matching DNA segments between individuals as well as the relationship name/degree?


Please point me to anything you know of that would help me to explore my data with the additional intermarriage/endogamy & location context if you know of such a thing.


in The Tree House by Troy High G2G1 (1.6k points)
Lately I've been thinking about this too.  For me it's less about strict endogamy and more about identifying cousins of my mother who are related to cousins of my father; both of my grand fathers came from the province of Friesland in the Netherland, which has a lot of small villages where 2x and 3x cousin marriage was common.

I suspect that you'll need to display your family web in a 3D way so that you can rotate the node-graph to view the relationships.  You could imagine this like a tube with the oldest generations at one end of the tube, the youngest generation at the other end, and all the people in between linked to two nodes above them and one or more nodes below them.

This "tube" structure could be flattened to a doughnut shape with the oldest generations defining the center hole of the doughnut and each younger generation defining a new outer edge of the doughnut.  One drawback of flattening the graph is that you may need to connect two people who are on opposite sides of the doughnut; when displayed as a 3D tube the connection crosses through the empty center of the tube.

2 Answers

+1 vote
I don't have a real answer to your question about how to graph this type of relationship. The problem stems from the inherent cycles that form resulting in the graph not being easily done in a planar manner. To be completely accurate, your particular description isn't actually endogamy but simple intermarriage. With true endogamy, this happens multiple times in the same lines which makes the problem even worse. For example, my paternal grandparents were first cousins once removed. My grandmother's parents were second cousins. Her paternal grandparents were first cousins.

Anyway, back to how to graph. I agree with Erik Oosterwal in that it would really take a 3D approach due to lines crossing so much. You might be able to use colored lines for connections but that would only work with one, maybe two, intermarriages. In a true endogamous community it would  rapidly become unworkable in a 2D representation.
by Doug McCallum G2G6 Pilot (404k points)
+1 vote

This is something I've been working on for a few weeks.  Here's where it's at now:

It's still rather rough.  Right now I have the "start" and "end" hard coded, and it prints out the tree at several stages of the processing.  The big hurdle is figuring out how to make it more efficient.  You'll notice that in the later generations there are quite a few duplicate people.  Some duplication is unavoidable, without going 3-D or crossing lines, but by just switching the order of a few siblings here and there it could be reduced.  The challenge is coming up with an algorithm that will do that for any given tree.  


by Ben Griffith G2G6 (9.0k points)

Related questions

+4 votes
1 answer
70 views asked Oct 12, 2019 in Genealogy Help by Jillaine Smith G2G6 Pilot (759k points)
+2 votes
1 answer
+1 vote
2 answers
144 views asked Mar 24, 2013 in Genealogy Help by anonymous
0 votes
1 answer
239 views asked Sep 25, 2012 in Genealogy Help by Pat Kersey G2G Rookie (130 points)
+5 votes
0 answers
21 views asked Nov 7 in Photos by Troy High G2G1 (1.6k points)
+3 votes
2 answers
+4 votes
1 answer
102 views asked Jul 25 in Genealogy Help by Robin Lee G2G6 Pilot (653k points)

WikiTree  ~  About  ~  Help Help  ~  Search Person Search  ~  Surname:

disclaimer - terms - copyright