For many years, I believed a story passed down in my family: that Francis Stephan, my great-grandmother Julia Stephan Maier’s only brother, had disappeared and was believed to have drowned in the Ohio River. His body was never recovered. The tale was told with such conviction that I never questioned it—until DNA told me otherwise.
Francis (born around 1848) appeared in both the 1850 and 1860 U.S. Federal Censuses, living with his parents, Anton “Anthony” Stephan and Catharine Elsasser, first in Mason County, Kentucky, and then in Brown County, Ohio. Family lore claimed that Francis had disappeared and that my great-grandmother Julia had never recovered from the loss. She rarely spoke of him.
My assumption that he died young began in childhood. I was told that Francis had probably drowned. As a child, I imagined him as a boy swept away by the river—because to me, only children drowned. It made sense at the time. Later, as an adult researcher, I found him in the 1860 census and revised the story in my head: he must have drowned sometime after that. I shifted the timeline and believed he perished as a teen, but I didn’t question my core belief that he had died young. I wrote about this family story in March of 2014. If you want to read it, you can find it here.
When I began exploring genetic genealogy in early 2014, I had only a few close DNA matches to work with. Y-DNA and mitochondrial testing were still the dominant tools, while autosomal DNA testing through AncestryDNA and other services was just gaining traction. Over time, I tested both my father and my maternal aunt and began organizing matches along grandparental and even great-grandparental lines. Gradually, a distinct cluster emerged: a group of cousins connected to my great-grandmother Julia Stephan Maier. Their trees bore the familiar surnames of my mother’s cousins—Kölsch, Specht, Burbrink—and were clearly descended from Julia’s sisters.
But another group of matches raised questions. These individuals shared DNA with both me and my aunt, as well as with my known Stephan cousins, yet their family trees didn’t include any version of the Stephan surname. Several of the strongest matches carried the surname Ballard. After creating extensive trees for these individuals, I found that they shared a common ancestor named Sarah “Sallie” Stevens, who had married Lawrence Ballard. This led me to a surprising discovery: Sallie was the daughter of Frank Stevens, who had married Mary Groh (or Grow, or Grey) in 1874 in Brown County, Ohio. Frank and Mary had seven children together before relocating to Bardstown in Nelson County, Kentucky.
The timeline and geography aligned perfectly. Frank Stevens appeared where Francis Stephan had left off. The surname had simply been Anglicized. It now seemed likely that my family’s story about a drowning was not about a teenager at all, but rather a man who disappeared much later. I revised my assumption again: perhaps Francis (now Frank) had drowned between 1889 (after his last child had been born) and 1897 (when Mary Stevens remarried). But DNA had more surprises in store.
New DNA matches eventually revealed something even more unexpected. In 2018, new evidence emerged. A separate cluster of DNA matches—unconnected on paper to the Stevens/Ballard line—shared significant DNA with my known Stephan relatives and me. These matches traced back to a man who appeared suddenly in another state in the late 1890s. He had no clear history before his marriage there, and his reported age and birthplace shifted from record to record. He had a very common surname, and I was told by a family member that they knew very little about this ancestor.
When I compared timelines, geography, and—most importantly—shared DNA across both descendant groups, the conclusion became unavoidable: the man who had once been Francis Stephan, later known as Frank Stevens, had taken on yet another identity and formed a second family.
This was not speculation based on name similarity or circumstantial location alone. The genetic evidence connected the two family groups in a way that traditional records never could. Descendants from each line share DNA at levels entirely consistent with a single common ancestor in Francis’s generation. No alternative explanation fits the data.
Rather than drowning in the Ohio River, Francis had disappeared from one life and quietly begun another. Within a relatively short distance of his first home, he remarried and fathered additional children under a new name.
To this day, I do not know what his first wife believed. She may have been told he drowned. She may have suspected abandonment. She may have chosen silence to protect her children. My great-grandmother’s lifelong reticence now makes more sense. What had once seemed like unresolved grief may instead have been the quiet aftermath of desertion.
Family lore often forms around absence. When someone vanishes, a story fills the vacuum. “He drowned” is simpler, kinder, and more socially acceptable than “He left.”
Through autosomal DNA testing, I was able to identify both of Francis’s families and reconstruct a far more complicated—and human—story than the one I inherited.
For genealogists, this case illustrates something profound. When an ancestor changes names, relocates, or intentionally severs ties, traditional records may fail us. DNA, however, does not forget. It links the descendants of both chapters of a life, even when the paper trail is fragmented or deliberately obscured.
In this instance, DNA did more than solve a mystery. It corrected a long-standing family narrative and restored the full truth of a man whose story had been simplified by time. Francis Stephan was not a boy lost to the river. He was a man who walked away—and whose descendants, unknowingly, carried the proof. I still haven’t located any photos of Frank Steven, and if any of his descendants are reading this story, I would love to hear your feedback in the comments section.