Ping Pong is the perfect metaphor to describe how to use both documentary records and DNA to break down genealogy brick walls.
The strategy begins by analyzing evidence from documentary records and then using your findings here to guide the analysis of your DNA matches. Next, use the patterns observed within your DNA matches to inform your subsequent searches in documents. Like a ping pong game, this informational rally continues back and forth between documents and DNA until your brick wall is reduced in strength and height to a six-inch string net revealing the ancestor hidden on the other side.
I’ve had tremendous success using “Genealogy Ping Pong” to break down my own brick walls. It works whether your unknown ancestor is a parent or a five-times great grandparent.
Most of us are generally comfortable with documentary research due in great part to websites like Ancestry, FamilySearch, and Fold3 who have made record discovery easier, but DNA analysis is often more intimidating given the complexity of DNA inheritance and the variety of tools available for deep analysis of matches. Indeed, I’ve successfully used many of these impressive DNA analysis tools, but the one strategy I use the most are genetic networks. In my opinion genetic networks are the easiest to use, and it requires no additional external tools. Genetic networks are also known as clusters, shared matches, or in common with matches.
In this blog post, I use a case study from my own research to illustrate how to alternate between documentary evidence and DNA to resolve genealogical problems and identify the name of an unknown ancestor. But first, let me describe genetic networks in greater detail so we’re all on the same page.
Genetic Networks
Simply put, genetic networks refer to a group of people who share DNA. Within your list of DNA matches, you have many genetic networks where a group of matches share a common ancestor or ancestral couple. When reviewing a match, genetic networks become viewable by clicking on the shared matches tab in Ancestry DNA and MyHertiage or the in common with matches tab in FamilyTreeDNA and 23andMe.
The figure below oversimplifies the process, but it visualizes how genetic networks are formed, i.e., some of your matches (with red checkmarks) are the same as some of your match’s matches (also with red checkmarks). When you extract these matches from your respective lists, or you click on the shared matches tab, it creates a genetic network of shared matches.
Put another way, let’s assume I have 14,925 matches. Also assume I am viewing one of my matches, which I will call “MW”, and he has 23,457 matches. Ancestry DNA indicates he is a possible third cousin. When I click on the shared matches tab while viewing my match with MW, I see that MW and myself have 45 matches in common. That is, 45 of my 14,925 matches are also found within MW’s list of 23,457 matches. These 45 matches constitute a genetic network for me and MW. It is probable that most of the 45 matches inherited DNA from the same common ancestor that I and MW also share.
Finding the most recent common ancestor among the 45 matches can be challenging. Not all matches have detailed family trees linked to their DNA tests. If they do have a tree, it may not go back far enough to easily identify the most recent common ancestor. You may have to build out their respective trees for them or contact them to help you accomplish the same. It takes time and effort, but the payoff is bigger than you can imagine. Let me show you how with a case study.
Case Study: Who was William Hill’s First Wife
My three-times great grandmother, Susan Hill, married James H. Wilson in 1837 in Tuscarawas County, Ohio. Probate records for Susan’s father, William Hill (1775-1836), clearly established the generational link.[1] However, Susan (Hill) Wilson’s mother remained elusive in part because she died prior to 1823,[2] and she married William Hill somewhere in Pennsylvania where marriage records are not widely available until the early 1900s. William Hill migrated from Pennsylvania to Ohio about 1811, and there were too many William Hills in early Pennsylvania records to know which one was likely him.
Documentary Rally #1
To determine who Susan Hill’s mother was, I began by searching documentary records pertaining to William Hill in Ohio. I concentrated on the typical record groups: census, tax, land, probate, civil court, marriage, etc. in Tuscarawas County, Ohio, where William died. Here is a summary of some of the more important pieces of discovered evidence:
- Personal property taxes place William Hill in Warwick Township as early as 1826.[3]
- William died in 1836 naming nine children in his will.[4]
- William married three times in Tuscarawas County, Ohio – in 1823 (Mary Kittweiller), 1826 (Susan Whitman), and 1834 (Catherine (Allen) Stewart).[5]
Subsequent research on Susan (Hill) Wilson’s husband, James Wilson, identified a court case where Susan’s brother, William Hill Jr., sued James Wilson for non-payment for a horse purchased in 1862.[6] According to the case, William Hill Jr. lived in Monroe County, Ohio but prior to the discovery of this court case, William Hill Jr.’s whereabouts were unknown.
Similar documentary research on William Hill Jr. in Monroe County, Ohio revealed he arrived there about 1830, and two of his earliest land purchases were from Samuel Carlton and John Winland in 1830 and 1840, respectively.[7] At this point in my research, William Jr.’s move to Monroe appeared random as no other known relatives were living in Monroe County, and it was 80 miles from Tuscarawas County.
DNA Rally #1
After building out the respective family trees for Susan (Hill) Wilson’s siblings, I used the EGGOS Search Strategy (Earliest Generation Group of Siblings) to identify shared matches between myself and descendants of Susan’s full siblings (see the blog post or YouTube video to learn more about the EGGOS Search Strategy). The resulting shared matches represent genetic networks where matches in the clusters either share Hill ancestry, ancestry from William’s first unknown wife, or both. Given my research objective, I was hoping to identify matches associated with Susan’s mother’s line.
Across Susan’s five full siblings who had descendants, I reviewed all matches within the identified genetic networks concentrating on their linked family trees. When I reviewed these shared matches, I performed three “Forest Management” tasks with the family trees for these matches within the genetic network: pruning, fertilizing, and harvesting. Forest Management principles are described below and in my YouTube learning module of the same name. I typically allocate a solid block of time (i.e., 4+ hours) when engaging in Forest Management so that I can more easily identify family tree patterns and correlate information in my head in real time.
- Pruning. To quickly become familiar with the family trees in the genetic network, I engage in gestalt pattern recognition to identify patterns across family trees rather than within a specific tree. To do this, I quickly move from match to match within the cluster having linked family trees and take mental note of frequently occurring surnames and geographies of interest. In this example, I concentrated on the matches’ ancestors born in Ohio and Pennsylvania because this is where my Hill line resided. This quick tree review process aids in the identification of branches in the matches’ family trees that are potentially relevant to my research question. The branch identification process is called “pruning”, which helps make the fertilizing phase more efficient.
- Fertilizing. Using Ancestry and FamilySearch trees, I quickly build out any match’s trees that are incomplete or missing and need assistance in growing their trees (fertilizing). I don’t physically construct a separate tree for them. Rather, I use other people’s online trees to “visualize” their tree, which saves time and effort. Depending on the number of trees I review, I either take mental or physical notes. If I recognize any surname or geographic patterns from the earlier pruning phase, I spend more time fleshing out these lines in this phase. My experience has taught me that this is where I will likely discover a most recent common ancestor. While building out my matches’ trees, I go far enough back to cover at least one or two generations past my ancestor of interest. In this example, I tried to go back to the early 1700s because Susan’s mother was likely born in the late 1700s. Although other people’s trees are not always accurate, I address their accuracy only if necessary and if a particular branch is related to my research aims.
- Harvesting. During the prior two pattern recognition phases, the review of family trees were quick. Now during the harvesting phase, I go back through all my matches’ family trees. Now, I am more methodical and take physical note of surnames, especially those I recall seeing frequently during the earlier pass-throughs. If possible, I also look for common birth counties within states for my matches’ ancestors, e.g., Tuscarawas County, Ohio or counties in Pennsylvania. Within the genetic network, I’m searching for matches where I can find the most recent common ancestor. A common ancestor might be found either in my tree, likely descending from the ancestral couple of interest (e.g., William Hill); or I may find common ancestors within several of the matches within the genetic network, which constitutes a sub-cluster within the overall network and helps to connect the other matches to one another. The common ancestors in these smaller sub-clusters represent ancestral couples worthy of future research effort (i.e., harvesting).
To aid in both the mental and physical note taking associated with pruning and harvesting, I find it helpful to use Ancestry’s DNA match map feature, which permits the easy identification of the birth locations appearing in my match’s tree. The map is found at the bottom of the page when reviewing a match. (Note: I only review my match’s tree here (i.e., “Wilson’s tree” in the image below), not my own tree (i.e., “Your tree” in the image below), which keeps the map less cluttered and focused on the match. I use the dropdown menu in the upper lefthand corner of the map to select only my match’s tree. See below.)
Using the above Forest Management technique, I discovered one very important pattern in the shared matches between myself and descendants of Susan (Hill) Wilson’s siblings. There were 14 matches who descended from Heinrich and Dorothea Winland. Graphically, the Hill and Winland matches are displayed below. Four distinct lines are found on both the Hill and Winland sides, with the respective lines converging on either William Hill Sr. and his unknown wife or Heinrich and Dorothea Winland.
The trees associated with the Winland matches suggested this family migrated from Lycoming County, Pennsylvania to Monroe County, Ohio about 1815. Monroe County is where William Hill Jr. resided as an adult (see Documentary Rally #1). A great clue!
Documentary Rally #2
After discovering the DNA connection with the Winlands, I returned to documentary records focusing on Lycoming County, Pennsylvania believing that perhaps my Hills also came from Lycoming prior to arriving Ohio.
For my initial investigation of Lycoming, I prioritized census, tax, and land records for their ability to situate a person to a time and place. Land records also have the potential to establish family relationships through inheritance or transfers.
Census records found no evidence of William Hill in Lycoming in 1790 through 1810 (recall that William appeared to arrive in Ohio around 1811). However, the Winlands were found in Nippenose Township at this time. Tax records were unavailable for this time frame. Land deeds were also not online at FamilySearch preventing easy access, but Pennsylvania land patents were available on Ancestry and online at the Pennsylvania State Archives.
And just like that, I found William Hill. An 1805 land patent for 400 acres for William Hill in Nippenose Township and another 400 acres in the same township in 1807.[8] In support of his 1807 application, an affidavit made by William Antis stated that William Hill first improved the land in 1803, since produced two crops of grain, and currently resides on the land with his wife and two children (see image below). The mention of his family composition was an important clue.
I next compared William Hill’s family mentioned in the 1807 affidavit to those listed in his will. If the “Nippenose” William Hill” was my William Hill, then his will was missing a child born in 1807 or earlier (see below).
I needed more evidence. Living out of the state, I hired a local genealogist to search the land deeds since they were not available online or through correspondence. Two transactions were found where William Hill sold his two land patents in 1808 and 1809, with the latter transaction stating that “William Hill therein named who is now out of the county”.[9]
Based on the birth locations of William’s known children in the above table, it appeared my William Hill and his family left Pennsylvania between 1809 and 1812. The 1809 land deed’s mention that William Hill was now outside Lycoming County fits with this timeline. It also explains why William did not appear in the 1810 census in Lycoming County, and census records in Ohio did not begin until 1820.
Before I discovered the above information, the only place of residence I had for William Hill was in Tuscarawas County, Ohio, where he first appears in 1823.[10] I could not definitively find him in the 1820 census in Ohio in part because I was looking for a family of eight – two adults and six children, which was derived from his will (see prior table). But what if I included the unknown second child identified in the affidavit? With this new theory, I found only one William Hill in 1820 census records fitting this family composition having two adults and seven children, and it was in a county neighboring Tuscarawas County. A transcription of the 1820 census for Cadiz Township in Harrison County, Ohio is found below with each child (and birth year) annotated and correlated with the appropriate tick marks. It is a match for the “revised” household if we assume the missing child is a male.
With this new knowledge, I created a locality guide for Harrison County. Given the early 1800’s time frame, very few record groups have the potential to name a husband’s wife, and so I concentrated on land, probate, and court records where it possible to find a wife’s name mentioned. William did not own land while in Harrison County, but one court case provided the answer to my research question. In an 1816 Common Pleas court case, William Hill and his wife, Elizabeth, sued Jonathan Heman for defamation of character.[11]
DNA Rally #2
Technically, the genealogy ping pong game was over. The brick wall had crumbled, and the identity of William Hill’s wife was partially revealed – Elizabeth. But the ball was in my court, and I had the energy for one more rally. Could I theorize or confirm Elizabeth’s surname?
Although William Hill married four times, he only had children who produced descendants with two wives – Elizabeth (last name unknown) and Susan Whitman (see the first table). Including myself, I have access to 13 other Ancestry and FamilyTreeDNA kits for cousins descending from James Wilson and Susan Hill and two kits descending from two other children of William and Elizabeth Hill. All these kits possessed genetic networks of varying sizes that included both descendants of Heinrich and Dorothea Winland and descendants of William and Elizabeth Hill. None of the Winland networks included descendants of William Hill and Susan Whitman. Indeed, where descendants of William and Elizabeth Hill and descendants of William Hill and Susan Whitman were both members of the same genetic network, these groups did not include descendants of Heinrich and Dorothea Winland (see graphic below).
The tentative conclusion is that William Hill’s first wife, Elizabeth, is likely the daughter of Heinrich and Dorothea Winland. The shared cM for Winland descendant matches rage from 20 to 42 and fall into the expected range for 5th cousins, which is the assumed relationship for most of these matches if the common ancestors were Heinrich and Dorothea Winland.[12]
While the win for the genealogy ping pong game went to me, there is much more to be done in both documentary research and DNA analysis to increase the confidence that Elizabeth is in fact the daughter of Heinrich and Dorothea Winland. It is probable that the missing child from the 1807 affidavit had died prior to the writing of William Hill’s will, but this needs to be explored as well. These examinations will be the focus of future blog posts.
Summary
Genealogy brick walls are made of “bricks” for a reason. It’s only through persistence and a deliberate methodology that they can be broken down.
I had no reason to look for William Hill in Lycoming County prior to DNA analysis. It would have been inefficient to research every William Hill in Ohio and Pennsylvania during the requisite time frame. Indeed, William did not appear in any record group in Lycoming County that researchers typically review first, especially on Ancestry or FamilySearch. By all accounts, it appears that William came of age or moved to Lycoming after 1800 and left before 1810 making him nearly impossible to discover in census records.
To take the fear and frustration out of DNA match analysis, we need specific research questions. An analysis of documentary records can better frame that question and make the query into DNA matches more focused. The incremental informational rallies between documentary research and DNA analysis helps to illuminate records and matches that may have been missed or overlooked on our initial passes through them.
I hope you consider incorporating “Genealogy Ping Pong”, the EGGOS Search Strategy, and Forest Management principles into your genealogical toolbox to break down your own brick walls.
Sources
[1] Ohio Wills and Probate Records, 1786-1998, William Hill (1836), Tuscarawas County, Will Records, Nos. 1-2-3-4, p. 87; database with an image, Ancestry (www.ancestry.com, accessed 2 September 2023).
[2] The date of death for William Hill’s first wife is based on the observation that William married Mary Kittweiller on 11 December 1823. Source: Ohio, County Marriages, 1774-1993, William Hill and Mary Kittweiller (11 Dec 1823), Tuscarawas County, Ohio, record 600, p. 75; database with an image, Ancestry (www.ancestry.com, accessed 2 September 2023).
[3] Tuscarawas County, Ohio, Personal Property Taxes, William Hill (1826), Warwick, p. 156, image 578 of 825; database with an image, FamilySearch (www.familysearch.org, accessed 6 September 2023), film 4849164.
[4] See footnote 1. Named children include: Mary, intermarried with Adam Carnahan; Ann Gearing; William Hill; Elizabeth, intermarried with Michael Harper; Sally, intermarried with William Wilson; Susan Hill; Nancy Hill; Matilda Hill; and John Hill.
[5] Ohio, County Marriages, 1774-1993, William Hill and Mary Kittweiller (11 Dec 1823), Tuscarawas County, Ohio, record 600, p. 75; database with an image, Ancestry (www.ancestry.com, accessed 2 September 2023). And Ohio, County Marriages, 1774-1993, William Hill and Catherine Stewart (9 Feb 1834), Tuscarawas County, Ohio, record 1838, p. 230; database with an image, Ancestry (www.ancestry.com, accessed 2 September 2023). And Moravian Church Records, William Hill and Susan Whitman (18 Oct 1826), Sharon, Tuscarawas County, Ohio.
[6] Meigs County, Ohio, Court of Common Pleas, William Hill vs. James Wilson (1862).
[7] Monroe County, Ohio, land deed, William Hill from Samuel Carlton (1830), Book D, p. 167, Recorder of Deeds, Woodsfield. And Monroe County, Ohio, land deed, William Hill from John Winland (1840), Book 2, p. 38, Recorder of Deeds, Woodsfield.
[8] Pennsylvania, Land Warrants and Applications, 1733-1952, William Hill (1805), Lycoming County, Nippenose Township; database with an image, Ancestry (www.ancestry.com, accessed 31 August 2023). And Pennsylvania, Land Warrants and Applications, 1733-1952, William Hill (1807), Lycoming County, Nippenose Township; database with an image, Ancestry (www.ancestry.com, accessed 31 August 2023).
[9] Lycoming County, Pennsylvania, land deed, William Hill to John Barker (1808), Book 6, p. 287-288, Recorder of Deeds, Williamsport. And Lycoming County, Pennsylvania, land deed, William Hill to Richard Salmon (1809), Book 9, p. 13-14, Recorder of Deeds, Williamsport.
[10] The first record for William Hill in Tuscarawas is the 1823 marriage record for Mary Kittweiller. Source: Ohio, County Marriages, 1774-1993, William Hill and Mary Kittweiller (11 Dec 1823), Tuscarawas County, Ohio, record 600, p. 75; database with an image, Ancestry (www.ancestry.com, accessed 2 September 2023).
[11] Harrison County, Ohio, William & Elizabeth Hill vs. Jonathan Heman (16 May 1816), Court of Common Pleas, Book A, p. 127; original held at the Harrison County Genealogical Society, Cadiz.
[12] Bettinger, Blaine T. (2020). The Shared cM Project 4.0 Tool v4, DNA Painter, accessed 31 August 2023 at https://dnapainter.com/tools/sharedcmv4.
One thought on “Genealogy Ping Pong: Alternating Between Documentary Research and DNA”