Richard Lang: #9 - What Does the DNA say?
Are the trees right about my relationship to Richard Lang? What can be learned from my DNA🧬?
This post is the ninth, and final, post in a series of posts prompted by my research into the life and times of Richard Lang and his family.
Parts #1-#7 of the series are focused predominantly on Richard Lang himself, his background and his family with Part #8 looking at the enslaved people associated with the family. If you would like to go back to the beginning, you can find Part #1 here:
As I said at the beginning of this series, I first became interested in Richard because he was continually appearing in Ancestry trees as my 4th great grandfather; the father of my 3rd great grandfather, William Lang. However, nowhere have I found documentary evidence sufficient to confirm that link between William and Richard. Richard definitely had a son William Lang but are the two William’s the same person?
I decided to turn my attention to DNA. What could that tell me in conjunction with the other evidence I have. Could Richard Lang be my 4th great grandfather as so many trees suggest? The connection was speculative and hypothetical, at best. However, I decided to use it as a starting point - a hypothesis to be tested. Could I prove it wrong?
Hypothesis
The hypothesis was that I descend from Richard Lang, son of Robert Lang and Millicent Myrick1, via their son William (and William’s wife Martha ‘Patsy’ Adams), as illustrated below:
Approach to Puzzle
To test this out, I decided on the following steps:
Compare William Lang and William Lang - Before launching into the DNA evidence I reviewed the documentary evidence associated with William (who married Martha ‘Patsy’ Adams) and William (Richard’s son). Is there anything to point to them being the same person?
Check backwards from me to William Lang and Martha ‘Patsy Adams - I reviewed my relationship back to William and Martha in light of DNA and documentary evidence. I was already confident that this was solid but decided to double check for the sake of completeness.
Check for connection forwards to me from Robert Lang and Millicent Myrick - Review DNA data as it relates to a possible connection forward to me from Robert and Millicent.
Check for connection forwards to me from the Benson family - Review DNA data as it relates to a possible connection forward to me from the Benson family. If Richard is my 4th great grandfather, then his wife Sara Benson would be my 4th great grandmother and, accordingly, I would expect to have DNA matches who descend from her parents.
A Need for Caution
This investigation required me to look closely at shared DNA matches and the extent to which I can cluster and triangulate matches around relevant ancestors. While I knew I had already identified a DNA segment triangulation2 to Robert or Millicent, most of the DNA data I had was predominantly reliant on clusters of shared matches3 which, in the absence of segment triangulation, are less reliable at pinpointing an ancestor in particular. Some, or all of these matches could match to one another other on differing lines.
While they appeared to cluster around Robert and Millicent, the matches could potentially connect to me on a different line, or even multiple lines. Sometimes people connect on multiple lines genealogically and finding the genetic connection can be tricky. I was conscious that some lines in my tree don’t go back very far yet. The more incomplete my tree, the greater the likelihood that the shared DNA is indicative of a genetic link coming from a line I have yet to discover.
On top of this, because I was trying to solve a puzzle six generations back from me (at the level of 4th great grandfather), some of my potentially relevant matches share a relatively small amount of DNA with me. The lower the centimorgan (cM) value of the shared matching, the higher the probability that a match is invalid or indicative of a connection too far back in time to be identifiable. I decided to work only with those matches sharing 15cM or more with me, at least in the first instance4.
William Lang and William Lang?
The hypothesis that I descend Robert ⇒ Richard ⇒ William ⇒ Permelia … etc., and that Richard Lang is my 4th great grandfather, relies on William Lang, who married Martha ‘Patsy’ Adams, being the son of Richard.
I reviewed documentary evidence associated with William (who married Martha ‘Patsy’ Adams) and William (Richard’s son) to see if there is anything to point to them being the same person.
William who married Martha ‘Patsy’ Adams
I haven’t been able to find much information about this William. He was probably born in South Carolina, at least that is where his eldest children were born and where he most probably married5. He would have been about the same age as Richard’s son William.
It appears that William and Martha had moved to Alabama by 1835 as their youngest child was born there that year6. William and Martha next moved to Kemper County, Mississippi. That was sometime between 1835 and 1840. There appear to be a number of people named William Lang in Kemper County during this time period. The 1840 United States Federal Census for Kemper County, Mississippi shows a William Lang living alongside Travis Stone in Kemper County7. Travis Stone is William and Martha’s son-in-law. For this reason, I think this record belongs to the William Lang who married Martha ‘Patsy’ Adams.
William appears to have died sometime between 1840 and 1850. According to the 1850 United States Federal Census, Martha was living with her children Richard, Martha and Emily in Kemper County, Mississippi in 1850. William is not on this census entry. This census record shows that Martha was living nearby to her daughter Permelia and her family, and nearby to her son Robert Lang and his family8.
William, son of Richard Lang
I haven’t been able to find much information about Richard’s son William and most of the information I do have is from secondary sources. As seen in earlier parts of this series of posts:
William appears to have been born in about 1780 in South Carolina
He moved with his family to East Florida in about 1782 but is not specifically named in any of the census records I have found, although could have been one of the two sons mentioned in the 1787 and 1793 census’9
I can’t find any mention of this William after this. He may have returned to South Carolina as an adult and be the same William that married Martha ‘Patsy’ Adams. He may not have.
One in the Same, or not?
Whether or not Richard’s son William is the same William as the William who married Martha ‘Patsy’ Adams remains an open question. As far as I have been able to determine at this point, it is possible they are the same person. It is also possible they are not. There just isn't enough documentary evidence to determine this.
Is it significant that William and Martha's first son was named Robert, their second son William and their third son Richard?
Working Backwards to William Lang and Martha ‘Patsy’ Adams
Documentary evidence supports my connection back to William and Martha. I haven’t done a detailed DNA analysis of specific segments back to William or Martha. I do have a significant number of matches that cluster around them.
On Ancestry (which doesn’t have segment data available), there are several matches who descend from each of their children. I also have DNA matches (matching me and one another) who trace their ancestry back to Martha ‘Patsy’ Adam’s father, William, her grandfather, James, and her great grandfather, also James.
I have numerous shared matches on Ancestry who cluster around Permelia Lang and Travis Stone (my second great grandparents). I do have matches whose shared segments triangulate to one of Travis’ parents thereby adding support to my connection to Travis and Permelia.
For the reasons given above, in the absence of segment triangulation, I can’t be absolutely certain how these shared DNA match cluster groups genetically link to me. I share DNA with them all but I can’t be sure it is via William and/or Martha’s line. I could share other family with them that I have yet to discover? Notwithstanding that, and taking into consideration both traditional sources of evidence and the DNA-related evidence I do have, I am confident that this part of my tree – myself back to William Lang and Martha ‘Patsy’ Adams, is genealogically correct and reasonably well supported by genetic evidence.
Working Forwards from Robert Lang and Millicent Myrick
Match Clusters
Some years ago, I connected a significant number of my DNA matches to my research tree with Robert and Millicent as the most recent common ancestral couple (MRCA). Some of these matches are grouped as clusters of people who match me and one another. For reasons already given, I couldn’t be certain that the DNA connection I share with them is indicative of a shared genetic connection to Robert or Millicent. The matches may link to me in some other way as well. Perhaps a line I have yet to discover in my somewhat incomplete tree.
Triangulation Group
Then I managed to form a small chromosomal segment triangulation group around Robert Lang and Millicent Myrick. On a chromosome map on the DNA Painter website, it looks like this:
This is a little misleading as Match 1 (the blue one) is a second cousin and so connects to me at grandparent level. The rest (the green ones) all connect to me via Robert and Millicent (assuming Richard is my 4th great grandfather, as per the hypothesis being tested, or one of his brothers). In addition, as a group triangulating to Robert and Millicent, it should be considered a triangulation group of 4 rather than 6 because matches 2 and 5 are siblings and should be counted only once.
On the basis of current knowledge (my tree and the match’s trees which I have checked for accuracy), there doesn’t appear to be any other ancestor that we could triangulate to. Of course, my tree is incomplete so I can’t be absolutely sure. However, on the face of it, these matches don’t appear to fit together in any other way. Matches 2 to 6 all descend from Richard’s brother, Robert.
I can place some reliance in the trees of these matches. Traditional sources agree Robert is the son of Robert and Millicent and his descent lines appear to be reasonably well sourced - 1 match descends via his daughter Mary Ann Lang, 1 (the two siblings) via his son Benjamin, and 2 via his son Robert.
Although these triangulated matches link via 1 child of Robert and Millicent only. I do have other DNA matches who link to the other children of Robert and Millicent but they either don’t link on this segment or no segment information is available.
I also have groups of matches who trace their ancestry back to Richard and Sara Benson. Most descend via their son Isaac and 9 of his children (William, Nathaniel, Willis, Isaac, Robert, George, Susan, Priscilla and Elizabeth). Several descend via their son, David via 3 of his children (Ann, Richard and Margaret). A few descend from their daughter, Sara via 1 of her children (Richard).
Checking Relationship Probabilities
Consistent with the hypothesis I am testing, I connected myself to the triangulated matches in my research tree using Richard as my link to Robert and Millicent. In other words, I created a speculative relationship to me for each of the matches based on the hypothetical connection of Robert ⇨ Richard ⇨ William ⇨ Permelia … etc.
My next step was to consider how probable my speculative relationship to them is. Putting match 1 (my second cousin) aside, the total centimorgan (cM) value for the DNA I share with each of the remaining matches ranged from 15cM to 29cM. Building a WATO, or WATO Plus, tree on DNA Painter was not an option because all were less that 40cM10 .
However, as this was a small group, it was easy to check the probability of each of the speculative relationships using the Shared cM Project Tool on DNA Painter. I did that for all 6 (even though, technically, only 4 count to form the triangulation group as explained above).
When I first did this, some time ago now, I used v.4 of the Shared cM Project 4.0 (26 March 2020). At that time, they were the most up to date probabilities available on DNA Painter:
For the 5 matches who connect direct to Robert and Millicent, the speculative relationships ranged between 55%-60% probable. This suggested that Robert and Millicent could be my 5th great grandparents. It didn’t get me any closer to knowing how I link to them, though. The same probabilities would be at play if I descended from a sibling of Richard’s. It says nothing about whether Richard is my 4th great grandfather or not.
So far so good, or so I thought …
More Up to Date Probabilities
As I said, I initially undertook this analysis some time ago. When I reviewed my data for this post, I decided to run the probabilities again with the Shared cM Project 4.0 tool v4 Beta with updated probabilities. Yikes, did I get a surprise:
Probabilities are drastically reduced for the 5 who connect direct to Robert and Millicent now ranging between 3% - 11%. Still possible but a lot less likely.
Looking more comprehensively across the matches’ trees and my own tree, the DNA does not appear to be coming to me from anywhere else. However, again, I am conscious of the incompleteness of my tree. I have not yet been able to extend the line beyond my 2nd great grandmother, Eliza, whose surname remains unknown. Similarly, I have a significant gap in my tree beyond my 2nd great grandmother Nancy whose surname I am still unsure of despite what numerous other people’s trees say! Who knows what is to be found behind those brick walls? On the other hand, I do also have a set of DNA matches who trace their ancestry back to Millicent’s grandfather, Francis Myrick. Again, I can’t be sure this is where the DNA is coming from either.
Looking for Benson Connections
If Robert and Millicent are my 5th great grandparents, the analysis so far neither supports nor casts doubt on the hypothesis that my descent is via their son Richard Lang and his son, William. I could descend from one of Richard’s brothers.
If the two William’s are the same person, Richard and Sara would be my 4th great grandparents. Under these circumstances, I would expect to have a cluster of DNA matches with shared Benson ancestors. Unfortunately, I don’t know who Sara Benson’s parents were. Well not with any certainty at least. However, my intention was to check my matches for potential common Benson ancestors and see if I could work forwards to link to Sara.
Finding matches with Benson ancestors has not been easy. They certainly aren’t jumping out at me like the Lang one’s did. I do have a number of DNA matches who trace their ancestry back to Robert Benson born about 1685 in Greenville, Virginia (married to Frances Prou) but a lot more work is needed to analyse this data and they may well having nothing to do with my Lang puzzle. More analysis needed here.
So … Relative? Ancestor? or No Connection at All?
My gut feeling is that Richard is not my 4th great grandfather. If he were, I should have more connections to the Benson family, although not knowing who Sara’s parents are tends to hinder that line of enquiry. It is an area where more analysis is needed.
Despite the lower probabilities offered by the Shared cM Project 4.0 tool v4 Beta, all thing’s considered, including my connections to the Myrick’s, I think there is a reasonable chance I descend from Robert and Millicent but not necessarily via Richard. At least, I am not discounting it yet. If that is the case, then I probably descend via one of Richard’s brothers in relation to whom documentary evidence is elusive. Again, more research needed.
If you know of any concrete evidence that may help me find a more definitive answer, please do get in touch! In the meantime, I think the most probable answer to my question is that Richard is my 4th great grand-uncle. But I am still far from certain and I certainly can’t prove it.
This post draws information from my family history website on the weare.xyz platform.
You may like to check out my other Substack publication here:
Notes:
Some trees suggest that Millicent is Millicent Higginbotham not Myrick. I can’t find evidence to support this. On the basis of what evidence I do have, I am reasonably sure she was a Myrick.
A segment triangulation is where 3 or more people each share DNA with one another in the same place on the same chromosome and can all trace their family tree back to the same common ancestor or ancestral couple. That is, they each share an overlapping segment on the same chromosome with one another.
This is where 3 or more people who each share DNA with one another can all trace their family tree back to the same common ancestor or ancestral couple. Either the segment data is not known or it shows that they share with each other on different places on a chromosome or on different chromosomes.
The probability that a match is either identical by chance (false positive) or too old to be identifiable (common ancestor too far back in time) increases as the amount of DNA shared between matches (cM value) decreases. So, what constitutes too small to bother with? The answer, like everything else in genetic genealogy, is “it depends”. How risk-averse are you?
Various studies have been done to determine the probabilities, some using simulations and some using empirical data. Results vary. A genie friend of mine, Kaye Bachelor, asked AI to pull together a summary table drawing on data from the DNA testing companies and information from the International Society of Genetic Genealogy (ISOGG)’ website. This is what AI produced. The chart illustrates, by cM size, what the likelihood is that a match will be a valid one (Identical by Descent - IBD) or a false positive (Identical by Chance -IBC).
If we rummage about in low value matches we have no way of knowing which ones are worth pursuing and which ones aren’t. It can be a huge time-waster if more productive avenues can be followed. In my opinion, working with anything below 7cM (possibly 8cM) should be a ‘no-go’ zone. It’s bonkers to go there! However, above that may depend on the circumstances (e.g., availability of higher matches) and a good dose of caution is needed. You may be wasting your precious time! In setting my threshold at 15cM for this investigation, you can see that I tend to sit at the risk-averse cautious end of the spectrum!
By the way, I should also add that there’s a whole different story for those working with endogamy.
Assumptions based on a variety of records from Ancestry.com
The National Archives in Washington D.C.; Record Group: Records of the Bureau of the Census; Record Group Number: 29; Series Number: M653; Residence Date: 1860; Home in 1860: Kemper, Mississippi; Roll: M653_584; Page: 763; Family History Library Film: 803584 via Ancestry.com. 1860 United States Federal Census [database on-line]. Lehi, UT, USA: Ancestry.com Operations, Inc., 2009.
Year: 1840; Census Place: Kemper, Mississippi; Roll: 216; Page: 5; Family History Library Film: 0014841 via Ancestry.com. 1840 United States Federal Census [database on-line]. Lehi, UT, USA: Ancestry.com Operations, Inc., 2010.
The National Archives in Washington, DC; Record Group: Records of the Bureau of the Census; Record Group Number: 29; Series Number: M432; Residence Date: 1850; Home in 1850: Kemper, Mississippi; Roll: 374; Page: 184a via Ancestry.com. 1850 United States Federal Census [database on-line]. Lehi, UT, USA: Ancestry.com Operations, Inc., 2009.
WATO and WATO plus on DNA Painter work best when most matches share 40cM or greater. When I first undertook this investigation, WATO Plus was not available.
Have you tried What Are The Odds trees or are all the connections too distant?
Jane this is great research. I don't know as much as I should about DNA so I loved reading the steps of your research. It's a shame that you don't know who Sarah's parents were. That would have given you a good start. I hope you're able to get the breakthrough you need to further your research. Just shows how frustrating family history research can be at time.s