Standing on the Shoulders of Other Researchers: How I Broke an 87-Year-Old

I am ridiculously proud of myself.

Like… obnoxiously, want-to-explain-Virginia-Chancery-records-to-random-strangers proud.

Because I think I finally broke one of the biggest brick walls in my family tree.

And of course it was a Johnson.

Because apparently having McDonalds on my mother’s side wasn’t enough genealogical punishment. I also needed Johnsons on my dad’s side.

When you are researching names that common, genealogy gets dangerous fast. You can find the right name, in approximately the right place, at approximately the right time, attach him to a completely reasonable-looking family…

…and be completely fucking wrong.

That is basically what happened to my third-great-grandfather, Joseph James Johnson.

Except the story gets much better than simply “I found his mother.”

Because I eventually discovered that people in my family had been looking for this answer since at least 1939.

And in 2026, eighty-seven years later, I finally got to send it back to them.

The Johnson Problem

My paternal line goes:

Me
My dad
Mathilda Rosetta Caldwell, my grandma
Sarah Emeline Johnson
John Meshack Johnson
Joseph James Johnson

For years, the documentary trail for Joseph James Johnson was both incredibly vivid and maddeningly incomplete.

We knew a remarkable amount about the end of his story.

Joseph was born in Virginia around 1812. He made his way southwest into East Tennessee, where he married Eletha Rowden in Roane County on November 8, 1840. By 1850, Joseph, Eletha, their children, and members of the extended Rowden family had moved west into Missouri and settled in the Big Richwoods of Miller County.

There, Joseph became a Justice of the Peace in Richwoods Township.

And then the Civil War came.

Missouri was not simply Union or Confederate during the war. It was a violently divided border state where neighbors could find themselves on opposite sides, guerrilla warfare continued outside formal military structures, and political loyalties could become deadly.

According to a biographical history preserved by the Miller County Museum, Joseph was one of only a handful of men in Miller County who voted for Abraham Lincoln in the 1860 presidential election.

The same account says that during the war Confederate troops passing through stole five horses from Joseph’s farm. Joseph fired into the group and wounded two of the soldiers.

Several years later, Confederate bushwhackers returned to the Johnson farm, called Joseph outside, and shot him in his own front yard.

According to the family and local-history accounts, Eletha was pregnant when Joseph was killed.

Their youngest daughter, Martha Ella Lee Johnson, was born posthumously.

Martha Ella’s exact birth date is one of those maddening little discrepancies that comes with nineteenth-century genealogy. Her death certificate gives March 21, 1865, while census records created during her lifetime variously point to 1865 or 1866. So I am not going to pretend I can resolve the precise date yet.

What is consistent is the larger chronology: Joseph was killed while Eletha was pregnant, and Martha Ella was born afterward. She entered the world already carrying the consequences of something that had happened before she could ever know her father.

And more than 150 years later, her descendants are still here.

I DNA match descendants of Martha Ella’s line, along with descendants through other children of Joseph and Eletha. Those matches fall into the same larger Johnson genetic network that ultimately helped me test the family I would later reconstruct from the Virginia court records.

The DNA cannot tell me whether Martha Ella was born in March 1865 or sometime later. Historical records have to answer that question.

But the DNA gives me something different and extraordinarily powerful: living descendants of this posthumous daughter genetically connect back into the same Joseph-and-Eletha family cluster as descendants of their other children.

Joseph and Eletha are not theoretical boxes on a family tree.

Their family is still visible in our DNA.

And the violence did not end neatly with Joseph’s death. My grandmother told me that afterward the family was essentially “chased out” of central Missouri and scattered. That is oral history, and I cannot yet document every element of it. But the dispersal itself is real.

Joseph and Eletha’s son John Meshack Johnson, my great-great-grandfather, eventually ended up far to the southeast in Ripley County, Missouri. And this is where genealogy produces one of those coincidences that makes me stare at the screen for a while. Ripley County’s county seat is Doniphan, Missouri. It was named for Colonel Alexander William Doniphan. And Doniphan was John Meshack Johnson’s distant cousin.

Not through the Johnsons. Through his mother. Eletha Rowden was the daughter of Meshack Michael Rowden and Sarah McNabb. Sarah McNabb’s ancestry leads back to John Pigg Sr. and Jane Shippingfield. I descend from their son John Pigg Jr. Alexander William Doniphan descended from their daughter Elizabeth Pigg. So Doniphan was not some famous historical figure randomly grafted onto a family legend. There really was a collateral family connection.

And John Meshack Johnson eventually lived near a Missouri town bearing his distant cousin’s name. Did John know that? I have absolutely no idea.

But Grandma knew there was a Doniphan connection. For years, I remembered her associating it with her Johnson family, and that memory became one of the things that made a completely different Virginia Johnson ancestry seem plausible to me.

Now I am increasingly unsure that Grandma ever told me the connection came through the Johnsons themselves. I was a troubled kid and teenager listening to my grandmother explain a sprawling nineteenth-century family tree involving Johnsons, Rowdens, McNabbs, Piggs and Doniphans. She may very well have told me the truth. I may have been the one who simplified it.

And that is another lesson in oral history: sometimes the family story does not become distorted over five generations. Sometimes it gets distorted in one child’s memory.

What we knew about Joseph’s adult life was unusually rich. We knew where he lived. We knew whom he married. We knew he served his community as a Justice of the Peace. We knew how he voted. We knew the political violence surrounding his death. We knew his children. We could follow their descendants. And now DNA gives us another independent layer of evidence connecting those descendant branches back into the same family. That much was solid.

But who Joseph’s parents were, where exactly in Virginia he came from, and what family he belonged to before he appeared in Tennessee?

That was darkness. We had the end of the book. The first chapter was missing.

And into that empty space, generations of online genealogy had inserted a completely different Johnson family.

The DNA Kept Telling Me No

The first serious problem with the Quaker Johnson theory was DNA. There was a faint signal around one proposed descendant, but “faint” eventually became impossible to ignore.

I have tens of thousands of DNA matches. If this family really represented my direct Johnson ancestry, where were all the descendants?

I started doing descendancy research instead of ancestor collecting.

That distinction matters. Rather than saying, “Here is a man who could be Joseph’s father,” I started asking: If this man really were Joseph’s father, where are his other descendants among my DNA matches?

Mostly, they weren’t there. I built test trees. I traced descendants. I rearranged hypotheses.

Ancestry Thrulines occasionally got excited, only for me to discover it had wandered into another Johnson family entirely.

Then I contacted someone who had researched the proposed Quaker Johnson family much more extensively than I had. His research actually questioned one of the crucial relationships supporting the whole theory.

He did not tell me who Joseph’s parents were. He did something almost as valuable. He stopped me from wasting another six months trying to prove something that probably wasn’t true.

So I deleted the relationship from my working tree.

Joseph became a brick wall again. Which sucked.

But an honest brick wall is better than a fictional ancestor.

Then I Found a Wrong Tree With the Right Document

This is why I will never become one of those genealogists who says publically wrong family trees are worthless. Public trees are not evidence. They can contain breathtaking amounts of garbage. But sometimes the garbage contains a treasure map.

I was searching Ancestry trees containing Joseph James Johnson when I found one that had Joseph’s family partially wrong. But they had attached Virginia Chancery Court records.

And they had another Johnson attached to the family.

Ewel.

I practically levitated.

Then I opened the original 1841 court papers. Not somebody’s transcription. Not somebody’s family tree. The actual handwritten documents.

And there was Holland Johnson.

The person who had built the tree apparently interpreted Holland as a man. But the court case made something very different clear: Holland Johnson was a woman. And Ewel and Joseph were her sons.

That one realization cracked the whole thing open.

Holland Had Been Sitting There the Whole Time

I followed that 1841 Chancery litigation backward into an earlier dispute over the estate of Jacob Johnson Sr. Now I could see an extended family instead of isolated names. Jacob Johnson Sr.’s heirs included Holland and her siblings.

That 1841 case connected Holland directly with Joseph and Ewel. The 1824 inheritance dispute connected Jacob Johnson Sr. to his children, including “my” Holland.

Between both Chancery litigation cases, I had something I could test.

So I built the family outward. Not upward. Outward.

Holland’s siblings. Their children. Their grandchildren. Joseph’s children. Ewel’s son. Their descendants.

And then something happened that had never happened with the Quaker theory: the DNA became elegant. Descendants of Joseph appeared. Descendants of Ewel appeared. Descendants through Holland’s extended family appeared.

And they shared matches with one another.

That was the moment when I stopped thinking: Maybe.

And started thinking: Holy shit. This is actually it.

The documents and the DNA were finally telling the same story!

But Someone Had Already Been Looking for Ewel

While researching Ewel, I contacted one of his descendants. I will call him A. because I am anonymizing living relatives in this story.

I shamelessly contact people who have done good genealogy. Usually I write some variation of:

“Hi, I am some stranger on the internet who has been obsessing over your dead relatives. Thank you for attaching actual sources. Also I think we are related.“

This time I told A. that I believed I had identified Ewel’s mother as Holland Johnson and that Ewel and my Joseph James Johnson were brothers. A. sent me family material. Then he forwarded an email his father had received in 2007.

That email changed the emotional meaning of the entire project.

The woman who wrote it, whom I’ll call S., descended from another one of Joseph James Johnson’s children and we are DNA matches on Ancestry. S. told A.’s father that her own mother, Goldie Johnson, had been searching for Ewel decades earlier.

Then I saw Goldie’s actual letter. It is dated January 21, 1939. Clayton, Oklahoma. She addressed it to the Chief Statistician of the Bureau of the Census.

And there, in Goldie’s own handwriting, was the mystery I had just spent months solving. She wrote that she was trying to compile a genealogical survey of Joseph Johnson, her ancestor in Miller County, Missouri. She said Joseph had an older brother. His name, as she remembered it, was something like “Euel or Ewal.” She knew that he had left Virginia around the same period. And she believed that he had become a criminal lawyer in Texas.

The letter was never mailed.

S. later found it in a family footlocker.

And then she picked up where her mother had stopped.

2007: They Almost Solved It

S.’s 2007 email is simultaneously wonderful and maddening to read now. She had already found Ewell Johnson marrying Susan Rowden in Meigs County, Tennessee in 1846. She suspected he had migrated west with Joseph and the Rowdens. She spent hours looking for Ewel in Texas census records. She contacted Rowden researchers. She even contacted the Texas Bar Association because of the family tradition that Ewel had become a lawyer.

She was asking exactly the right questions: Who were Ewel and Joseph’s parents? Where in Virginia had they lived? Was this Ewel actually Joseph’s brother?

She simply did not have the records I now have.

A.’s father read Goldie’s letter in 2007 and wrote back to his son: This was enough for him. He believed Joseph was Ewel’s younger brother.

But they still lacked the legal records that put the family together.

They had gotten unbelievably close. They had oral history. They had family correspondence. They had the Tennessee marriage. They had Texas descendants. They had an independent tradition on Joseph’s side saying that Joseph had an older brother named Ewel who went to Texas.

What they did not have was Holland.

Ewel Really Did Go to Texas

Another piece of the story survived in Ewel’s own family.

His son Elijah S. “Lige” Johnson left a handwritten autobiographical note stating that he was born March 25, 1850 in Missouri and “come to Texas in the year 1860.”

Then I found the 1869 Tarrant County voter registration. Ewel appears there and reports nine years of residence in Texas. 1869 minus nine. 1860.

Then there is Elijah’s obituary. When he died in 1929, the newspaper reported that he had been born in Cedar County, Missouri on March 25, 1850 and had come to Texas with his family at age ten.

Three independent pieces of evidence.

The handwritten family record.
The contemporary voter registration.
The later obituary.

All pointing to the same migration year.

And then I went digging through Cedar County, Missouri court records and found what had happened immediately before they left.

Ewel Didn’t Just Decide to Move

In 1859, Ewel’s financial life appears to have collapsed. Multiple creditors were pursuing him through Cedar County Circuit Court. There were judgments. Garnishment proceedings. Debt collection.

And finally an execution against his real estate. On November 4, 1859, the sheriff sold Ewel’s land at public auction. Then, almost immediately afterward, the family disappears from Missouri.

By 1860 they are in Texas.

That does not permit me to write a fictional movie script and pretend I know exactly what Ewel was thinking as he loaded a wagon. But it gives the migration a completely different context.

The Panic of 1857 had destabilized credit. The Missouri-Kansas border region had been experiencing years of political violence and market disruption. A rural physician operating partly through credit could have been extremely vulnerable in that environment. Those historical forces are context, not direct proof of causation.

What I can prove is simpler: Ewel was being pursued by multiple creditors in 1859. His property was taken under execution. His family went to Texas in 1860.

And suddenly the old family story of a man vanishing west does not look mysterious at all.

It looks human.

Then I Had One More Problem

I knew what had happened. But the person whose research had helped prove it was no longer active online. S. had not logged into Ancestry for more than a year. My email bounced. I contacted The Miller County Historical Society who remembered her, but she had stopped coming around some time before COVID.

I found her Facebook account, but there was no indication she was actively using that either.

And this is where I got strangely nervous. Was it creepy to contact her family? Because genealogy people understand that “Hello, I have solved the mystery your great-grandmother began investigating in 1939” is an exciting message. Normal people may reasonably interpret that as: Who the hell are you and why do you know who my great-grandmother was?

So I asked a genealogy group. The response was overwhelming: Reach out. Be respectful. Explain yourself. Give them an easy way not to respond. But try.

So I did.

I found S.’s granddaughter, whom I’ll call T., and sent her a gentle Facebook message. She responded within hours. S. is still living, but she has been experiencing health problems and has stepped away from genealogy and the internet. T. was gracious. She wanted the material.

And suddenly I was doing something that felt much bigger than updating my Ancestry tree. I was sending Goldie’s answer home.

Eighty-Seven Years Later

Goldie wrote her letter in 1939. It was never mailed.

Her daughter found it. Her daughter spent years continuing the search. She contacted another branch of the family in 2007. Those people preserved the correspondence.

Nearly twenty years later, one of their descendants forwarded it to me.

I found the Virginia Chancery records.

DNA testing, which Goldie could never have imagined, confirmed the family structure through multiple descendant lines.

Digitized Tennessee, Missouri and Texas records connected the migrations.

And in 2026, I reached Goldie’s great-granddaughter and sent the evidence back into the family.

Eighty-seven years.

That is the part that gets me. Because genealogy is so often presented as a solitary activity. One researcher. One tree. One breakthrough.

That is not what happened here.

I Did Not Solve This Alone

I did a lot of work. And yes, I am still ridiculously proud of it. I recognized that the DNA did not support the Quaker theory. I killed my own hypothesis. I reopened original documents. I figured out Holland was a woman. I traced the earlier court case. I reconstructed the family. I used descendancy research instead of blindly trusting Thrulines. I traced multiple DNA branches. I found Ewel’s Missouri court cases. I connected the Missouri records with Texas evidence. I contacted descendants. I pushed the research back into public repositories.

But none of this happened by myself.

A researcher studying the Quaker Johnsons helped me realize I was probably wrong. Someone else built a partially incorrect tree but attached the correct Virginia lawsuit. Goldie wrote down what she knew in 1939. S. saved that letter and continued her mother’s research. A.’s family preserved emails from 2007. Ewel’s descendants preserved Elijah’s handwritten autobiography. Someone preserved the newspaper obituary. Archivists preserved courthouse books. The Library of Virginia digitized Chancery files. FamilySearch digitized records that once would have required travel and microfilm. Thousands of descendants took DNA tests. A Facebook genealogy group convinced me not to chicken out about contacting the next generation. And T. answered.

That is what broke the brick wall.

Not one brilliant discovery. A chain of people preserving little pieces of truth.

Even Wrong Research Has Value

This experience has completely changed the way I think about bad family trees.

I still hate tree rot. Obviously. One bad relationship gets copied into fifty trees, then those fifty trees become “evidence” for another hundred people, and suddenly someone who never met a Quaker in his life has six generations of beautifully formatted Quaker ancestors.

But I no longer think incorrect research is necessarily useless. Someone was wrong about Holland. But they attached the document. That document changed everything.

Someone else could not prove the Quaker relationship. Their failure saved me months of work.

S. and her mother could not prove Ewel and Joseph were brothers. But they preserved the question.

Sometimes genealogy advances because someone proves something. Sometimes it advances because someone says: I tried this. It does not work. Sometimes because somebody writes: I think these two men were brothers, but I cannot prove it.

That is valuable information. Maybe especially when we admit what we do not know.

This Is Why I Keep My Research Public

I have become almost evangelical about this.

Make the public tree. Attach your sources. Put things on FamilySearch. Write the blog post. Scan and transcribe the weird handwritten letter. Upload the photograph with names. Save the email. Back everything up. Keep copies on more than one platform. Give the local historical society a copy. Let the Internet Archive crawl your website.

And please, for the love of every future genealogist who will curse your name otherwise, label the people in old photographs.

Our older family historians did this work when genealogy meant driving to courthouses, writing letters to county clerks, scrolling microfilm until your eyes crossed and hoping somebody answered a stamped envelope. Goldie wrote to the Census Bureau. S. called the Texas Bar Association.

I sat in my house with a laptop, downloaded Virginia court cases, compared DNA matches and found descendants across several states in an afternoon.

Those are radically different tools. But it is the same project. We are passing information forward. And the researchers who come after me will have tools I cannot imagine either.

Somebody Is Going to Research Us Too

That may be the strangest thing genealogy teaches me. The people in these records did not think they were historical figures. They were not walking around in 1859 thinking: Someday my fourth-great-niece will analyze this sheriff’s sale as evidence of nineteenth-century migration patterns.

Ewel was probably thinking: Fuck. I lost the farm.

Goldie was probably just a young woman who wanted to know what happened to her family.

S. was probably sitting at a computer in 2007 completely frustrated that she could not find this damn man in Texas.

They were living ordinary days. Just like us.

And someday somebody is going to look backward at our lives and realize that we lived through history too. They will wonder what we knew.

What frightened us.
Why we moved.
Why relationships broke apart.
What stories our parents told us.
What parts we remembered wrong.
And whether anybody bothered to write any of it down.

So write it down. Save the letter. Keep the weird email. Publish the imperfect tree with good notes explaining what is proved and what is still hypothetical. Reach out to the distant cousin.

Because sometimes an unanswered question can survive for eighty-seven years.

And sometimes you get to be the person who finally sends the answer home.

Sources and Web Citations

The following links include the primary records discussed in this article as well as my working research profiles. My Ancestry and FamilySearch profiles are actively maintained and will continue to be updated as I locate, transcribe, and organize additional evidence.

Working Family Profiles

Holland Johnson, Ancestry Family Tree Profile.
Working research profile maintained by Jessica Murray, including attached records, relationships, DNA-supported research, and source documentation.
https://www.ancestry.com/family-tree/person/tree/210099939/person/432805198041/facts

Holland Johnson, FamilySearch Family Tree Profile, GHWZ-974.
Collaborative public profile; additional records and citations are being added as research is completed.
https://www.familysearch.org/en/tree/person/details/GHWZ-974

Virginia Chancery Court Records

Heirs of Jacob Johnson Sr., Patrick County, Virginia Chancery Cause 077-1824-001.
Library of Virginia, Chancery Records Index. This estate litigation identifies Holland Johnson among the heirs of Jacob Johnson Sr. and was essential to reconstructing the extended Johnson family.
https://old.lva.virginia.gov/chancery/case_detail.asp?CFN=077-1824-001

John Boyd v. Holland Johnson et al., Patrick County, Virginia Chancery Cause 141-1858-009.
Library of Virginia, Chancery Records Index. The proceedings discussed in this article connect Holland Johnson with her sons Ewel and Joseph Johnson and document the family’s land dispute.
https://old.lva.virginia.gov/chancery/case_detail.asp?CFN=141-1858-009

Cedar County, Missouri Circuit Court Records

M. Church & Co. v. El R. Johnson, May 3, 1859.
Missouri Circuit Court, Cedar County, Circuit Court Record Books, May 3, 1859 Term, p. 614. Garnishment proceeding involving Ewel R. Johnson. Digital image, FamilySearch, Image Group Number 008194255.
https://www.familysearch.org/ark:/61903/3:1:3Q9M-CSLF-ZKWN?view=explore

Washington Kington v. Euel R. Johnson, May 3, 1859.
Missouri Circuit Court, Cedar County, Circuit Court Record Books, May 3, 1859 Term, p. 616. Judgment against Euel R. Johnson for debt, with securities associated with his appeal. Digital image, FamilySearch, Image Group Number 008194255.
https://www.familysearch.org/ark:/61903/3:1:3Q9M-CSLF-ZKQJ?view=explore

Sherrill Hartly & Co. v. Ell R. Johnson, November 4, 1859.
Missouri Circuit Court, Cedar County, Circuit Court Record Books, November 4, 1859 Term, p. 674. Sheriff’s execution and sale of Ewel R. Johnson’s real property. Digital image, FamilySearch, Image Group Number 008194255.
https://www.familysearch.org/ark:/61903/3:1:3Q9M-CSLF-ZK8G?view=explore

A note about the records

Names in nineteenth-century records were frequently spelled inconsistently. Ewel R. Johnson appears in various records as Ewel, Euel, El, and Ell Johnson. I have preserved the spelling used in each individual source citation rather than silently standardizing the original record.