Visualizing Complicated Relationships: Working with Pedigree Collapse, Multiple Relationships, and Endogamy

Visualizing Complicated Relationships: Working with Pedigree Collapse, Multiple Relationships, and Endogamy

Steve Little, one of our Research Like a Pro with DNA 3 Study Group members, wrote this guest blog post about how he uses diagrams to help visualize the multiple relationships with his DNA matches. He is from a community of mild endogamy — Appalachia. I hope his insights will help anyone dealing with multiple relationships with DNA matches, pedigree collapse, and endogamy.   -Nicole

Don’t panic! At first glance, the swirling tangle of lines below may appear to be a meaningless and unmitigated mess. However, this is not the case when we consider how it was assembled. Approached in small steps, simple, smaller, easily-understood diagrams can be assembled and combined to simultaneously display information that may not be appreciated when spread across multiple diagrams. Sometimes the whole is greater than the sum of its parts.

But a tangle of confusion is the dilemma that some genetic genealogists may face, especially when first dealing with cases in communities where pedigree collapse, multiple relationships, and endogamy are present. At least, that was my personal experience, as I first began to grapple with my “mildly endogamous” Appalachian heritage. This post intends to help clarify for the learning researcher how to differentiate between the concepts of pedigree collapse and multiple relationships–even when they may be present simultaneously–and to introduce some methods for teasing apart and clarifying complicated genetic relationships, and how they may be visually depicted to assist in research planning and to more deeply appreciate their complexity.

Figure 1: “Adam” LITTLE and “Betty” LITTLE (in green near bottom) are cousins related in several different ways, perhaps more than a dozen, including by pedigree collapse and by multiple relationships.

 

All 32 of my third-great-grandparents had settled into one Appalachian county by 1820 (many earlier), and 60 of my most-recent ancestors were born, lived, and died there, in Ashe County, North Carolina. My parents left Ashe County in the mid-1960s, the first of my direct ancestors to leave Ashe in 140 years; sometimes they say they escaped (left of their own free will) and at other times they will say that they were exiled (“asked” to leave); both are in jest. The only people who love Ashe County more than the people who live there are the people who have left there.

Anyway, a love of place and of family was instilled in me early. In the 1980s, as a teenager, I began digitizing an aunt’s decades of solid genealogy research; with the birth of my first son in the 1990s, I began my own family history research; and with the advent of DNA tests, I began the work of confirming the documentary research that my aunt and I had done.

But, with such deep roots in one county, I found myself encountering DNA challenges that weren’t apparent to many of my more cosmopolitan colleagues. For example, while they were able to produce neat and tidy Leeds Charts, mine looked like an explosion at the Benjamin Moore paint store: a riot of color; while their Gephi network graphs images looked like beautiful and meaningful color-coordinated galaxies, mine looked like a Jackson Pollock painting had been run through a shredder and swept into a clump.

So this year, 2022, was the year I threw myself deeply into learning everything I could about how to apply DNA analysis to untangle and tease apart complicated relationships in my Appalachian heritage. And it’s been a great year. Some highlights and golden nuggets would include these folks that helped me tremendously this year and what they taught me:

  • Research Like a Pro with DNA” from Diana Elder, Nicole Dyer, Robin Wirthlin, and the Family Locket team was the greatest value, bang-for-the-buck, during a year in which I took five classes, courses, or institutes; this multi-month course is the foundation I would recommend to a friend. My key takeaway was “having a plan and working the plan,” as I paraphrased the process to myself; recognizing the great value of putting one’s research into written form was made real to me during my work in this course; finally, this course gave me the confidence to advance to some of the more focused and specialized training that followed, including:
  • Diahan Southard (Your DNA Guide)’s “Endogamy & DNA Course” in April exquisitely clarified for me the critical differences between multiple relationships and pedigree collapse; she also drove home for me the point that documentary research and DNA analysis complement and inspire each other;
  • Leah Larkin (TheDNAGeek)’s “When Your Tree Is a Banyan: Untangling Endogamy,” also during spring 2022, really refined for me the idea that how the researcher approaches DNA analysis in endogamous communities should be guided by how much endogamy is in a community (in three “levels,” measured relatively looking at average segment size with a match and total shared DNA with that match) and to adjust one’s research strategy accordingly, looking for larger and larger segments as communities move from “mild,” to “moderate,” and “strong” levels of endogamy;
  • Kimberly Powell‘s chapter “The Challenge of Endogamy And Pedigree Collapse” is essential reading for the researcher working with pedigree collapse and multiple relationships; this chapter alone is worth the price of the book in which it is found, Advanced Genetic Genealogy: Techniques and Case Studies (ed: Wayne; 2019), but of course, the whole text is a great resource. So, during Kimberly’s course “Tools & Strategies for Tackling Tough Research Problems” at the Genealogical Research Institute of Pittsburgh (GRIP) 2022, I took every opportunity to steer conversations toward pedigree collapse, multiple relationships, and endogamy. It is the steps from Kimberly’s chapter that I work through when approaching challenging DNA problems concerning pedigree collapse and multiple relationships.
  • Karen Stanbary‘s course at the Institute of Genealogy and Historical Research (IGHR) 2022, “Planning for and Conducting Research Using DNA and Documentary Sources,” impressed upon me the value of applying the Genealogical Proof Standard (GPS) to DNA analysis and provided practical experience using the GPS to guide report preparation;
  • Paul Woodbury, during a brief one-on-one session at IGHR, shed some light on research strategies on a particularly thorny research problem, suggesting that targeted testing would be helpful, and introducing me to the concept of “genetic pioneers,” that is, folks who may have left the endogamous population, and prioritize them for research; in my research, identifying matches of several folks who were adopted out of Ashe county were particularly revealing and helpful.
  • I could not be more comfortable with reading a textbook, watching a webinar, and/or engaging in independent study, but something magical happens in small peer groups over time, and I was blessed to experience three or four of those magical learning experiences this year. Like skilled Sherpas, peer group leaders such as Lisa Stokes took the fear out of citations and even inspired their mastery, and Sam Williams led by example while charting paths forward through tough research problems. But of course, it was my peer group friends from which I learned the most over the past year. Thank you all–you know who you are!–for your time, patience, insights, feedback, encouragement, and support. I apologize for falling out of touch, and I hope we can work together again.

 

I can’t introduce the concepts of multiple relationships, pedigree collapse, and endogamy better than have already been done here earlier, so instead, I will focus on how I have encountered them in my Appalachian genetic genealogy research. But, if these are new concepts to you, you must start here:

 

As was mentioned earlier, all 32 of my third-great-grandparents had settled into one Appalachian county by 1820 (many earlier), and 60 of my most-recent ancestors were born, lived, and died there, in Ashe County, North Carolina. And when my parents left Ashe County in the mid-1960s, they were the first of my direct ancestors to leave Ashe in 140 years. These deep roots make doing genetic genealogy a bit more challenging. This increased difficulty makes sense when we remember what the goal is of many genetic genealogy research questions: who is the common ancestor that is shared between two DNA testers? So to risk stating the obvious, in incidents where DNA matches share multiple relationships, there will not be only one common ancestor.  And incidents of pedigree collapse render less reliable one of our best research tools: drawing assumptions about possible common ancestors based on the amount of shared DNA between two matches, which may be misleadingly high among descendants of pedigree collapse.

Although these challenges may seem daunting, they are not impossible to overcome. However, more effort will be required to address these situations, and it is essential to carefully consider assumptions and remain open-minded. This is especially important because research problems can often be more difficult and require more time to solve than initially anticipated, and solutions may be more complex than initially thought.

My experience is in Appalachia, and while I have little experience working with Polynesian, Arcadian, and Jewish genetic genealogy, I hope that these and other closely-knit communities will benefit from my reflections.

 

Multiple Relationships: A “Simple” Example

What may initially appear simple can actually be deceptively complex. For example, consider the relationship between this researcher and his first cousin, “Frank” SHEPHERD.

relationship chart depicting two first cousins their and shared grandparents

Figure 2: It would be easy to assume that the shared DNA between “Writer” LITTLE and his first cousin “Frank” SHEPHERD is inherited from their shared grandparents. But, especially in some communities, relationships are often more complex than they first appear to be.

Because “Writer” LITTLE and “Frank” SHEPHERD share a set of grandparents, and because their amount of shared DNA (865 cM over 38 segments, the largest of which is 116 cM) is almost exactly the expected amount suggested by the Shared centiMorgan Project, one might assume that all of those shared segments were inherited from their LAWRENCE-HOUCK grandparents. And in more cosmopolitan communities, that assumption may be more likely to be true.

But in even mildly endogamous communities, that assumption cannot be trusted. In this case, while the mothers of “Writer” LITTLE and “Frank” SHEPHERD were known to be full sisters, there was no commonly known relationship between their fathers “Adam” LITTLE and “Earl” SHEPHERD. However, both documentary research and DNA analysis uncovered a previously unknown relationship: the brothers-in-law by marriage are also, it turns out, half-fourth cousins, sharing a third great-grandfather, Absolom BOWER (1786-1857), as the diagram below depicts.

relationship chart depicting an example of multiple relationships, a case in which two people are maternal first cousins and paternal fifth cousins

Figure 3: An expansion of the diagram above shows that “Writer” LITTLE and his maternal first cousin “Frank” SHEPHERD share multiple relationships, including through their fathers. In other words, two sisters, “Charlotte” and “Dawn” LAWRENCE married and had sons with two fourth cousins, “Adam” LITTLE and “Earl” SHEPHERD.

This example demonstrates why the researcher cannot assume that all chromosome segments shared between matches are inherited only from the most recent common ancestor. That bears repeating: just because we may have discovered the most recent common ancestor between two matches, that does not mean that we have found every common ancestor between two matches. Especially when conducting DNA segment triangulation and analysis, the researcher needs to be mindful that a segment may have mistakenly been misattributed to a more recent common ancestor, and that re-evaluation and correction are a part of the research process.

Now, let’s consider a slightly more complex example. Still focusing on “Adam” LITTLE, let us examine the relationship between him and a different cousin, “Betty” LITTLE.

 

Multiple Relationships: Another Example (or two, or three, … YIKES!)

Living testers “Adam” LITTLE and “Betty” Little share multiple relationships. RootsMagic 7 contains a “Relationship Calculator” tool which suggests they may share more than a dozen unique relationship paths. The simplified relationship diagrams below show just two of those relationships: figure 4, as first cousins once removed through their shared LITTLE-BARE ancestors, and figure 5, as second cousins through their shared BOWER-BARE great-grandparents. Further below, figure 6 shows together in one diagram the information depicted separately in figure 4 and figure 5.

Figure 4: Among their several different relationships (perhaps more than a dozen), the closest between “Adam” LITTLE and “Betty” LITTLE is as first cousins once removed.

Through a separate set of grandparents, “Adam” LITTLE and “Betty” LITTLE are also second cousins, as shown below in figure 5.

Figure 5: “Adam” LITTLE and “Betty” LITTLE are also second cousins through a separate set of grandparents, Bawly BOWER and Emma Jane BARE.

Shown immediately below, figure 6 is a relationship diagram that combines the two diagram above, yet in which persons are only depicted once. And in which another layer of complication becomes apparent: the surname BARE seems to be appearing a suspicious number of times (three times, so far).

relationship chart depicting several instances of pedigree collapse shared between two cousins

Figure 6: Instances of multiple relationships do not always occur in isolation. It may be the case in some communities that multiple simultaneous instances of pedigree collapse and multiple relationships are more the rule than the exception. Here, two cousins, “Adam” and “Betty” LITTLE share multiple instances of pedigree collapse AND multiple relationships among their ancestors.

Pedigree Collapse: An Example

As if the multiple relationship situation described above were not complicated enough, there is also an instance of pedigree collapse in this case. Our research subjects “Adam” LITTLE and his cousin “Betty” LITTLE share multiple ancestors, one of whom is James Eli “Bawly” BOWER (1863-1960). It turns out that Bawly BOWER married his half-first cousin once removed Emma Jane BARE (1863-1940). When cousins marry, that means a grandparent (and perhaps both) will appear twice in a pedigree chart. And that is what you see below in figure 7, a pedigree chart in which Absolom BOWER (1786-1857) appears twice: first as the great-grandfather of Bawly BOWER (with his first wife), and then again as the grandfather (with a second partner) of Bawly’s wife, Emma Jane BARE.

(And, YES! Careful readers will have noticed that this Absolom BOWER (1786-1857) is the same person mentioned in the initial example at the beginning of this post; he is the shared third great-grandfather of “Adam” LITTLE and “Earl” SHEPHERD, fourth cousins. To say Absolom BOWER was a bit of a notorious character in Ashe County history is perhaps an understatement.)

Figure 7: When cousins marry their shared grandparent(s) will appear twice in a traditional pedigree chart, first as an ancestor of the first spouse, and then again as an ancestor of the second spouse. In this case, Gypsy Fredor BOWER is both a descendant of Absolom BOWER and his first wife, Susannah KOONTZ, and a descendant of Absolom BOWER and his later partner Catherine Ann “Caty”/”Katy” SHEETS.

But wait! There are more complications to note. The pedigree chart in figure 7 is for Gypsy Fredor BOWER (1901-1991). If you look again at figure 6, the combined multiple relationship diagram for “Adam” LITTLE and “Betty” LITTLE, you will notice that on the same generational level as Gypsy Fredor BOWER, you will also see her brother, George Cecil BOWER (1893-1987). That means an identical pedigree chart for George Cecil BOWER would also show Absolom BOWER two additional times, as it does in figure 7. That means that if we were to expand the diagram in figure 6 to include the instances of pedigree collapse emanating from Absolom BOWER, his genetic heritage would descend along four relationship paths between the two cousins. That is what we will illustrate below.

But first, we must address yet another complication that sharp-eyed readers will have noticed in figure 6: There are three women in the diagram with the surname BARE. In cosmopolitan regions, the reoccurrence of a surname might not be notable; how strongly would you suspect that two people named SMITH in New York City were related to one another? But our assumptions in endogamous populations are different. While we don’t assume that everyone with the same surname is related, red flags to do deep research is warranted. In this case, documentary evidence strongly suggests that all three of these BARE women are descendants of an early Ashe County pioneer and settler, Henry BARE (1779-1849).

Figure 8: A customized pedigree chart for “Adam” LITTLE. The chart creation tool Exploring Family Trees at https://learnforeverlearn.com/ancestors/ creates pedigree charts that depict instances of pedigree collapse by only showing each ancestor once, even if they would fill two positions in a traditional pedigree chart (as when cousins marry, and their grandparent(s) would appear twice). The tools work by rendering GEDCOM data uploaded by the user. The data in this custom file contained ONLY persons in line affected by instances of pedigree collapse.

Simultaneous Multiple Relationships and Pedigree Collapse in One Relationship, One Diagram

A combined diagram is never the first I create. And not every research problem benefits from a combined diagram. But sometimes a combined diagramis very helpful in expressing the complexity of some relationships. (And I often find it helpful to be reminded that, like life, sometimes genetic genealogy is messy. And sometimes a messy diagram has its own beauty. And that beauty is a reflection of the family it depicts, also messy and beautiful.)

Figure 1, again: “Adam” LITTLE and “Betty” LITTLE (in green near bottom) are cousins related in several different ways, perhaps more than a dozen, including by pedigree collapse and by multiple relationships. This diagram is both simple and complex. It is simple in that it does NOT depict every ancestor of the DNA testers, but rather only those ancestors in instances of multiple relationships or pedigree collapse. It is complex in the sense that it combines several instances of multiple relationships, and it shows sources of pedigree collapse only once (whereas in a traditional pedigree chart they would be shown twice).

A combined diagram also has practical usages. In the selection of targeted DNA testing candidates, this diagram is helpful in two ways: identifying ancestors whose descendants may be useful testers, and then also to eliminate lines from consideration, e.g., if you needed to find a testing candidate that didn’t have BARE DNA. The usefulness of keeping the complexity of a relationship foremost in mind should not be understated; it tempers us against too quickly latching onto the first possible solution that presents itself, cautioning us to be thorough in our research.

And that brings me to a few concluding thoughts. Later, I hope to be able to share with you some of the more technical tools and methods that I use in my approach to doing genetic genealogy in Appalachia such as segment triangulation and chromosome mapping. But visually introducing those challenges is an important first step.

Figure 10: About 26 million people live in these subregions of Appalachia across a dozen states (Appalachian Regional Commission).

So as I acknowledge the complexity and challenges that genetic genealogy in even mildly endogamous communities entails, the scope of this issue needs to be emphasized. These messy and complicated relationships are not unique or rare. It is not something special between “Adam” and “Betty.” This is the largely underemphasized point: nearly every shared match between them will also be as complex. And this is not unique to the BARE, BOWER, and LITTLE families. In fact, many, if not most, families in Ashe County will have roots as deep as these. You can probably sense what needs to be acknowledged next: mild endogamy is not unique to Ashe County, NC. There are 26 million people living in Appalachia, and while most will not have roots only in one county for over 100 years, many will. And so the need is great.

Which brings me to my final point: another part of the challenge of genetic genealogy in endogamous communities is that best practices, great tools, and established methodologies are in the process of being developed. I am not aware of anyone that is offering a “second-semester” course in genetic genealogy in endogamous communities. I, and I suspect many, many others, would welcome an advanced endogamy and pedigree collapse course, a course already assuming familiarity with key terms and concepts, deep working knowledge of DNA tools, and moderate experience working with DNA tools and documentary research to resolve complex cases, but looking for advanced instruction in methodology (steps) in these cases, with emphasis on working through case studies in class. In the meanwhile, I would welcome collaborating with others in this work. If you have a success story to share, or a tip, tool, or technique you think others would like to know about, please share it. Let’s lean into these beautiful messes.

Takeaways

  • Carefully consider assumptions and remain open-minded.
  • Just because we may have discovered the most recent common ancestor between two matches, that does not mean that we have found every common ancestor between two matches.
  • Be mindful that a segment may have mistakenly been misattributed to a more recent common ancestor. Re-evaluation and correction are a part of the research process.
  • In some communities, multiple simultaneous instances of pedigree collapse and multiple relationships are more the rule than the exception.
  • A combined diagram is very helpful in expressing the complexity of some relationships. In the selection of targeted DNA testing candidates, combined diagrams are helpful in two ways: identifying ancestors whose descendants may be useful testers, and then also to eliminate lines from consideration.
  • Messy and complicated relationships like Adam and Betty are not unique or rare. Nearly every shared match between them will also be as complex.
Back to blog