How to Write and Publish a Proof Argument with DNA Evidence
Updated 11 November 2020
Have you solved a case using DNA evidence and want to write a proof argument? Selecting an organizational structure, preparing figures and tables, requesting permissions, and reviewing DNA standards are important steps in the process.
Perhaps you are finished with your proof argument and wondering how to submit it for publication in a genealogical journal. Several authors whose DNA cases were published in the National Genealogical Society Quarterly (NGSQ) shared insights and will be quoted throughout the article.1 To see a list of articles including DNA evidence in the NGSQ since 2001, see:
DNA Case Studies in the NGSQ – Google Sheets Document
A few of these articles are available to read on the Board for Certification of Genealogists’ Genealogical Work Samples website here: https://bcgcertification.org/learning/skills/genealogical-work-samples/. Elizabeth Shown Mills’ article about Zilphy Watts is available on her website here: https://www.historicpathways.com/articles.html
Studying these articles is a valuable way to prepare for writing your own DNA proof argument.
Proof Statements, Summaries and Arguments
Proof statements are simple phrases that state a genealogical fact. In order to prove something in just one statement, you should be able to provide a footnote with one or two citations as documentation for the statement. The information should not be conflicting. DNA evidence is sometimes used in a proof statement when a relationship is very obvious, like an autosomal DNA match between a parent and child.
Proof summaries are used when proving a genealogical fact requires more explanation than one statement. You may need a bulleted list of evidence, an explanation of conflicting evidence that is easily resolved, or two to three paragraphs of explanation. DNA evidence can be used in a proof summary. An example of this is when Y-DNA results are used and show a close genetic distance between two men whose relationship can also be proved with traditional records. An example of using autosomal DNA evidence in a narrative pedigree with proof statements and summaries can be found in Angela Packer McGhie’s presentation, “Using DNA Test Results to Confirm a Pedigree,” 20 October 2020 at Family Tree Webinars.
Proof arguments are longer than proof summaries. Genealogists write proof arguments for complex cases that include a large body of evidence. Proof arguments often include multiple pieces of conflicting information that are not easily resolved or where multiple hypotheses exist and must be disproved. Within proof arguments you may also have proof statements and proof summaries that help build the case. DNA evidence in a proof argument may include many test takers and evaluation of several types of DNA.
To read more about types of proofs and writing proof arguments, see Chapter 7 in Mastering Genealogical Proof by Thomas W. Jones.2
Organizational Structure of the Proof Argument
After several phases of research, all of which have produced notes, reports, and preliminary conclusions, you are now ready to assemble your results into a proof argument. You may be able to reuse the writing and tables you have already created in research reports as you completed each phase of research. Outlining your evidence can be a helpful first step. As you list your evidence and arrange it logically, consider one of the following ways to organize the evidence.
Documentary Evidence First, Then DNA Evidence
Proof arguments with DNA evidence about a question several generations back often begin with documentary evidence and end with DNA evidence. Following this model, start with an introduction that presents your research question. Then present each piece of traditional evidence in a logical order for the first half of the article. To meet the GPS element about reasonably exhaustive research, include all “reliable information potentially relevant to the research question.”3 Be sure you have tapped all relevant non-genetic sources.
After the documentary research has been presented, the second half of the article states that DNA was used to test the hypothesis reached with traditional research. The DNA section often begins with background information about DNA inheritance and methodology, goes on to discuss the methodology used, present the test-takers, their lines of descent, and DNA analysis. Following the body is a conclusion that sums up the evidence.
Weaving DNA and Documentary Evidence
Not all proof arguments start with documentary evidence and end with DNA evidence. Some choose to go back and forth between the two types of evidence. Karen Stanbary wrote “Rafael Arriaga, a Mexican Father in Michigan: Autosomal DNA Helps Identify Paternity.”4 Hers is the only case study in the NGSQ that deals with an unknown parent of a test-taker. The test-taker, Joanne, was adopted by her mother’s sister and no record named the father. After a short opening paragraph, Karen discussed which types of DNA were applicable to the research question. She then wove DNA findings with documentary evidence throughout the article. She used pedigree intersection of close matches to find several candidates, then eliminated all but one through targeted testing. Karen shared several nuggets of advice for incorporating DNA evidence into a proof argument:
Weave the information from the genetic evidence correlated with the documentary evidence into the argument by generation. Be very careful with word choice. Be wary of jargon and labels. Do not overplay the evidence. Consider Elizabeth Shown Mills’ qualifying words for conclusions from Evidence Explained. Always keep in mind you must detail the meaningful evidence and your reasoning that ties the pieces of evidence together.
Jill Morelli’s article, “DNA Helps Identify “Molly” (Frisch/Lancour) Morelli’s Father,” which was an unknown grandfather case, followed an organizational structure similar to Karen’s. After five paragraphs of background information, the rest of the article included DNA evidence interwoven with documentary evidence until the possible candidates were narrowed to one father for Molly.5
Patricia Lee Hobbs’ NGSQ case, “DNA Identifies a Father for Rachel, Wife of James Lee of Huntingdon County, Pennsylvania,” was the first in which DNA first identified a distant ancestor, as opposed to an indirect case confirmed with DNA.6 She began with the fact that no traditional clues led to parents for Rachel, wife of James Lee, and shared the basics of Rachel’s life after marriage. After this short intro, she jumped right into DNA clues that led to the Addleman family, Rachel’s family of origin.
DNA Evidence First, Then Documentary Evidence
Occasionally, genetic evidence is presented early in the article followed by documentary evidence. This is the case in B. Darrell Jackson’s article, “George Craig of Howard County, Missouri: Genetic and Documentary Evidence of His Ancestry.”7 The Y-DNA test results presented at the beginning suggest a common ancestor between the test-takers and a significant Y-match within a close number of generations. The match’s earliest known ancestor, William Craig d. 1759, had a patrilineal descendant who possibly matched the George Craig of Howard County whose origins were unknown. Darrell’s exhaustive research on George Craig of Howard, Missouri, paired with genetic evidence, allowed him to bridge George’s post-1806 timeline with the pre-1805 timeline of William Craig’s grandson, George Craig of Virginia. The presentation of Y-DNA results up-front set the stage for the traditional research that followed.
Ultimately, the organizational structure of the proof argument is up to the author and should build the case logically, leading the reader through each evidentiary item. It is not a good idea to present research in the order it was found, as stated in standard 61 of Genealogy Standards.8
Figures and Tables
The DNA section of a proof argument typically contains a diagram showing test-takers’ descent from the common ancestor. Karen Stanbary advises, “Include an extended source-cited traced descendant chart incorporating all significant matches used in the argument.” Some NGSQ articles have charts spanning two pages, with the base test-takers who descend from the research subject shaded and matches descending from the hypothesized ancestor not shaded.9
Below is a simple diagram showing descent from a hypothesized ancestor in the NGSQ style with simple lines and names. To make a chart like this using Lucidchart, start with boxes and connecting lines, then select each box by clicking Select>All Shapes, then change the outline color to white. In this example, I have privatized the test-takers’ names, as well as the names of their parents. If privatization of the test-takers’ parents is necessary, because they are living, you may still want to identify their gender if applicable to DNA inheritance. In the example below, I have followed LaBrenda Garret Nelson’s example to state the privatized person’s name as “Leta’s daughter” and “Vernon’s son.”10
Another option to privatize living parents is to say “living female,” as Melinda Daffin Henningfield did in her March 2019 article.11 Melinda shared that she decided to privatize the names of her test takers, although all of them provided written permission to be used in the publication. She shared two reasons for this decision:
1. I am a retired nurse practitioner. As a former medical provider, I feel strongly about protecting living people’s privacy. I asked each test taker if they would mind being privatized and each of them said that would be okay. Most of them did not care one way or another.
2. Think of the reader. If you, as a writer, provide them with thirteen individuals with a variety of names, say, Joan Rickets, Joe Richards, Jack Ramsey, Betsy Damsel, Butch Dayton, Barbara Harris, Ben Becket, Sally Timmons, Sarah Gordon, Jimmy Craft, Theresa Hall, Charles Marks, and Cindy Gunderson. As you walk the reader through the analysis and correlation of the evidence, he/she is furiously thumbing back and forth in the article from the figure to the narrative (and back to the figure again) trying to remember if Sarah was a descendant of Mary… or John and Elizabeth? How about Butch? It is really difficult for the reader to keep straight who is a descendant of whom. In the analysis and correlation it is really important for your reader to be able to differentiate between the two groups. On the other hand, if you tell your reader that Mary’s descendants are numbered 1–5 and have the letter A, John and Elizabeth’s descendants are numbered 6-13 and have the letter B, and Mary’s descendant’s are shaded in the figure, the reader can tell at any place in the narrative who is descended from whom. I think it is much easier for the reader to keep the test takers straight and it protects their privacy if they are privatized in this way.
Whether the testers are privatized or not, it’s imperative that readers are able to visualize their lines of descent. Providing a diagram is the simplest way to do this. Besides descendant diagrams, you may need to use other types of diagrams to show lineages and pedigrees in a visual format.
The following figures are often used to in DNA proof arguments:
– Lines of descent from a common ancestral couple
– Linage of a single tester to the research subject
– Hypothesized relationships to the research subject
– Pedigree of a tester
– Pedigree of a DNA match
– Triangulated matches showing bars with overlapping segments
Parent-child links shown in a diagram are documented with at least once citation per link. Recent case studies in the NGSQ have also used author-created Ancestry trees to document the parent-child links. This saves space in the printed journal, however this doesn’t mean the authors haven’t reviewed the parent-child documentation thoroughly. The first NGSQ author to include an author-made Ancestry tree in a DNA article was Melinda Daffin Henningfield. She shared the following about the decision to cite her Ancestry trees for parent-child links:
When I sent the manuscript to the Q, it was 25 pages. In addition, it had 2 figures and 6 tables (each of those might take a page). In addition, the parent-child links were in a separate appendix that was 6 pages long and was in 10 point type. The Q has a page limit of around 20 pages. I knew I had to do something to save some space and came up with this idea. (I don’t have a website, so that was not a possibility). The editor thought it was a good idea and so we went with it.
LaBrenda Garrett Nelson’s case study, “Parents for Isaac Garrett of Laurens County, South Carolina: DNA Corroborates Oral Tradition,” used both in-text citations for parent-child links and her Ancestry tree which provides additional documentation for each link. She cited her Ancestry tree again, later in the case study, to document two of the test takers’ pedigrees in a discussion about pedigree completeness.12 I asked LaBrenda about her decision to include the Ancestry tree. She said, “I had read every modern DNA article published in the NGSQ for guidance, including how others met the requirement in Standard 2 to document each parent-child link. I followed Melinda Henningfield’s lead in her award-winning article that included a documented Ancestry tree.”
Incorporating DNA evidence requires that shared CM amounts, expected amounts of shared DNA, triangulation of segments, and other numerical data be presented. Often the best way to include this information is in a table or figure. Karen Stanbary shared that figures were helpful to illustrate complex hypotheses in her article. She said the best part about preparing research for publication was benefiting from the wisdom of her editor, Thomas W. Jones, who converted dense reader-unfriendly narrative into beautiful figures that clearly illustrated her points.
Proof arguments with DNA evidence often include the following types of tables to present DNA data:
– Total shared cM that base testers share with each other
– Shared segments among testers
– Total shared cM and proposed relationships between testers
– Total cM shared among base testers and DNA matches
– Triangulated segments
– Tree completeness of testers
– Y-STR values of testers
Genealogy Standards, 2nd Edition, includes several standards regarding the use of DNA evidence.13 The new DNA standards were added in 2019, so all the articles in the NGSQ prior to that date did not have a set of DNA standards to use. Jill Morelli’s December 2018 article was the first to utilize the new DNA standards, which at that time weren’t published.14 Jill said, “I had no map other than the very knowledgeable help of Karen Stanbary who was writing/reviewing the standards as I wrote the article. The article came out about three months before the standards were adopted and distributed.”
The standards can be your guide you as you put together a proof argument. Standard 52 suggests that genealogists should address factors that can affect determining a genetic relationship, like the possibility of more than one common ancestor for a tester and DNA match. One strategy to mitigate this possibility is pedigree evaluation to see if the pedigrees of each person are complete to the generation of the common ancestor, and possibly 1-2 generations beyond.
Standard 53 encourages genealogists to use a sufficient number of testers and matches. If we haven’t tested enough people to rule out conflicting hypotheses, our conclusion may not meet the GPS. Targeted testing or asking cousin matches already identified in DNA databases to share their results with us can help.
Standard 54 states that assembled research results should include enough detail and identifying information that others can verify the conclusions. If some of our matches have not transferred to GEDmatch so readers can verify the matches we cite, we may need to ask some of our testers and matches to transfer. In LaBrenda’s article, some of the test takers did not transfer to GEDmatch, but others did. For those who did not transfer, her citation states the matches were analyzed with the tools at FamilyTreeDNA and the results were verified by the editors.15
Inclusion of DNA Evidence
You may be wondering when DNA evidence should be included as part of a reasonably exhaustive search. In research planning, genealogists should consider whether genetic sources might aid with our question, as standards 12 and 14 suggest.16 However the use of DNA is only required when we want to state that a relationship is biological.17 Not all indirect cases require the use of DNA, and in fact, there are many cases where DNA would not apply or be useful. One NGSQ author stated that mitochondrial DNA could have been helpful to her case, had a matrilineal descendant existed.18
Nancy Niles Wehner authored “Parents for Richard M. Vaughan (1844-1921) of Howell County, Missouri,” where she used autosomal DNA and documentary evidence to connect a man who was estranged from his parents during the Civil War to his family of origin.19 Nancy said that she typically starts by exhausting documentary sources before adding DNA to her cases. She also shared valuable advice about when to use DNA evidence:
If the answer can be found in documentary sources, whether directly or a good solid indirect case, then going to the additional effort involved in doing a DNA study just is not worth it. The key for the indirect case is “solid” as sometimes you can have a possible indirect case but it seems kind of thin, a bit forced, or subject to too many caveats or possible challenges. Then turning to DNA is probably appropriate to bolster the case.
Ask yourself if DNA evidence is truly needed to answer your research question. Yes, it’s the “in” thing to do right now but the reality is that not all research questions need DNA evidence to make a case. After all, we’d been doing without it for a long time before DNA became an option for us.
If DNA is important to your case, then be prepared to take the time to make sure that the number of DNA donors for your study is sufficient to draw meaningful conclusions. Looking back, I would consider the Richard Vaughan case to be barely adequate; but at the time it was every donor I had available to me. The further back in time you are trying to use the DNA evidence for the more donors you will need to look at since the possible 0 cM shared DNA for relationships out past about 2C1R is going to eat into your pool of donors. My current study to connect Richard’s father, Thomas Henry Vaughan, to his parents and siblings has 16 DNA donors to date and I’m looking for a few more.
LaBrenda Garret Nelson advised, “You should not even think about including DNA until you have exhausted non-genetic sources that point to hypothesized relationships.”
Benefits and Challenges of Publishing in a Journal
I asked several NGSQ authors to share their favorite part about publishing their work and some of the challenges they encountered in the process. Several mentioned the process of gathering permission, editing, receiving peer review, and revising were both challenging and beneficial. Jill Morelli suggested reviewing the guidelines of the journal you plan to submit to. The NGSQ guidelines can be found here: Guidelines for Writers and The Editorial Process. These documents also provide excellent tips and resources for writing case studies.
Satisfaction of a Conclusion
Jill Morelli said the best part was the satisfaction of proving her conclusion: “I had known about the problem of identity of my husband’s grandfather for 30 years, and even tried to find some relatives of the Wood family but was unsuccessful. It is always satisfying to get to a conclusion.” The difficult part of preparing research for publication was contacting test takers – “My husband did not know any of the test takers–so we couldn’t draw on a long history of building trust–basically all the contact with the 5 test takers were “cold calls.” Two of those test takers, in the first minutes of the phone call said they wouldn’t test. I listened, listened some more and eventually both decided to test.”
Contacting DNA Donors
Nancy Wehner also shared that locating and contacting DNA donors was challenging. Once they responded, it was also challenging getting the written permission forms back in order to use their results in a publication. She advised: “If you need permissions, get them up front before committing your time and money to an individual’s testing and analysis. People have the right to withdraw permissions at any time, but I’ve never had one do that where I got permissions to start with, but have had a couple bail on me after the testing by then refusing to give permissions (live and learn).”
Time and Money
Vic Dunn, author of “Determining Origin with Negative and Indirect Evidence: Cylus H. Feagans of Virginia and West Virginia,” said meeting the GPS can be time consuming: “I had reached the conclusion about Cylus H. Feagans’ parentage early in my genealogy life and had put it aside for about a dozen years with plans to publish at some point. However, when I did start writing it, I realized it didn’t meet the GPS since I needed to eliminate all the other Feagans parental candidates which ended up being a significant task. … The article was basically complete when I decided the argument could be bolstered by adding the DNA evidence.” His advice? “Be prepared to spend a lot of time and possibly money, particularly if you have to pay for some of the tests like I did.”20
Patti Lee Hobbs also mentioned the time commitment that incorporating both documentary and DNA evidence requires. She advised, “don’t forget the documentary research and be proactive in gathering more DNA evidence. Some will try to use too little DNA evidence to prove the case and they don’t seem to know how to gather more. Gathering the DNA evidence takes a lot of time.”
Elizabeth Shown Mills, author of the first NGSQ article to use autosomal DNA and mitochondrial DNA, “Testing the FAN Principle Against DNA: Zilphy (Watts) Price Cooksey Cooksey of Georgia and Mississippi,”21 shared that the process of peer-review can be validating: “When we attempt to reconstruct the past, no matter who we are, we all have doubts about the accuracy of our conclusion. (If we don’t have those doubts, we should!) Peer-review provides outside perspectives and critical analyses that help us refine our arguments and re-evaluate conclusions.”
Elizabeth also provided valuable insight into one of challenges into publication–the length restraints of a journal article. She said:
[Validating parent-child links] can generate a huge amount of documentation. At that point, we have to judiciously cull and tighten the evidence for each generational link. Currently, authors and editors tend to handle this challenge by attaching note references to every “fact” asserted in tables and charts, with a separate set of reference notes to document each of those assertions. In complex cases, writers may also provide the editors with a supporting file of documentation that goes beyond what is provided in the tables or charts. That supporting evidence will then be evaluated by peer-review and editorial-review, after which the editor presents the shorter version to the journal’s readership, with the implicit promise that the supporting evidence validates the assertions that could not be more robustly documented in the paper itself. (On the other hand, this is not necessarily a DNA-specific problem. As NGSQ editor, once upon a time, I had an author submit a manuscript in a box 14” x 11” x 5” inches. The manuscript was the top item. Below it were five file folders, labelled “Footnote 117, File A,” “Footnote 117, File B,” etc., all the way through File E.)
Improving our Writing
Patti Lee Hobbs shared that one of the benefits of writing for a journal is that it will help us become better. She said, “My main aspiration as a genealogist is to write and publish. One can write for themselves, of course, but writing for someone else especially for those with more exacting standards will drive us to our best work.”
She also shared the challenges technical writing presents: “I have a love-hate relationship with writing. It is painful for me; and when I’m in the midst of it, I feel terribly inefficient. But the results are so gratifying! When we minutely examine every little piece to put it together, we start seeing things we didn’t see before. It forces us into a frame of yet further discovery beyond what we’ve already found more superficially. I’m a goal-oriented person and knowing that producing a package of genealogical proof lies at the end spurs me through the difficult, tedious parts.”
LaBrenda Garrett Nelson said the best part about publishing was “finally producing documented conclusions regarding my Garrett ancestors in a setting where I would receive expert feedback about ways in which I could tighten the analysis. Also pushing the envelope by showing how DNA can be particularly useful in the case of communities that left few records and little direct evidence.”
Peer review and expert feedback is a compelling reason to submit to a journal, but can you also receive helpful feedback from friends? Can hired editors and friends be as honest as anonymous peer reviewers in a double-blind process? Elizabeth Shown Mills said, “Friends will offer a bit of constructive advice—guardedly. Very guardedly. They worry about hurting our feelings. Peer reviewers don’t. Good peer reviewers understand that their first obligation is to scholarship and to the integrity of the journal—and anonymity allows them to be honest.”
I will end with thoughts from Thomas W. Jones, author of the first NGSQ case to use only autosomal DNA, “Too Few Sources to Solve a Family Mystery? Some Greenfields in Central and Western New York.”22 His advice for those publishing a proof argument that includes DNA evidence is: “Do not ignore any potentially useful documentary evidence, keep an open mind about possible conclusions, and ensure that you have asserted the only possible answer to your research question.”
Will you write a proof argument with DNA? I hope to see many more case studies with DNA evidence published in the NGSQ and other journals soon. I plan to submit my own someday. Studying the DNA standards and case studies published in the NGSQ has taught me how to follow the footsteps of the authors who shared their insights in this article. I hope it will do the same for you.