Connecting William Bowdoin: Finding my Alabama ancestor’s way home through history and DNA

The first part in a two-part series. (Part 2.)

This article is distilled from a longer research paper, which is shared below: “Connecting William Bowdoin: The Problem of the Ancestry of William Bowdoin (b. 1802) of Autauga County, Alabama, Addressed through Records and DNA, and a General Overview of the Family of William Bowdoin (b. ca. 1740)“

Genealogists have been searching for the ancestors of my fourth great-grandfather, William Bowdoin (b. 1802), for at least a century. Many people have adopted the suggestions of Ancestry’s algorithms for his parents and assumed the problem was solved, but even though the same parents are given for him in a great majority of trees online, this “consensus” is very clearly incorrect. In this article and the attached research paper, I undertook a project to thoroughly examine the facts of William Bowdoin’s life and family through historical methods, cutting through all the mess of bad assumptions and conflations that has gathered there, and then to approach the problem with fresh eyes through genealogical DNA analysis.

William Bowdoin was born in North Carolina in 1802. He came to Alabama in the early days of the state and settled near Deatsville in Autauga County (today in Elmore County). The name Bowdoin is spelled like the illustrious Bostonian family of French Huguenot origin, but here in the South, it’s pronounced simply “Bowden.” With a name so distinctive—and William Bowdoin was literate, and consistently used the Bowdoin spelling—it has always been frustrating as a genealogist that we could never connect him to his ancestry.

There was another southern family that spelled the name that way too. In 1982, a genealogist named U. Bowdoin Marsh published a book about his Alabama Bowdoin family, which settled in Coffee County in South Alabama’s Wiregrass region in the 1840s. My dad had a copy of Marsh’s book when I was growing up, and it was frustrating too that we couldn’t connect to his family even though he seemed to have done a lot of research, and the answers seemed just beyond the frontier.

In the “Ancestry era” of genealogy, in the past ten or fifteen years, Ancestry.com’s “hint” algorithms (“little green leaves”) suggesting “potential parents” for ancestors in your family tree have resulted in a majority of family trees settling into a “consensus” or “orthodoxy” about many ancestral connections. Especially with autosomal DNA testing and the introduction of Ancestry’s ThruLines tool, many people have taken the suggestions at face value and assumed they were correct. These “suggestions” are based on nothing more than similarity between names and dates in your tree and other people’s trees, and may not fit your facts at all; but the more people who adopt them, the stronger the algorithm’s “suggestion” gets, and the more people assume they are true. With William Bowdoin, at least two “suggestions” have taken hold and come to dominate in a majority of trees. In William’s case, these “consensus” suggestions are very easily proven false.

Not the son of William Bowdon (b. 1773) and Nancy Wiggins

A large majority of Ancestry family trees—I just checked, and there were 193 grouped under these facts—have the information that William Bowdoin’s parents were William Bowdon (b. 1773) and Nancy Wiggins, who settled in Coffee County, Tennessee. This is very clearly incorrect. They also identify my ancestor as “Rev. William B. Bowdoin”—but he was never a minister, and had no middle name or initial. More on that below.

There are several straightforward reasons why William Bowdon and Nancy Wiggins were not the parents of William Bowdoin (b. 1802):

• William Bowdon (b. 1773) in Coffee County, Tennessee, had another son named William Bowdon, shown living near him on the 1840 census, who was clearly not the one in Autauga County. This William was aged between 20 and 30 in 1840, so born after 1810. My William (born 1802) had been living in Alabama since the 1820s.
• William Bowdon (b. 1773) was living in South Carolina in 1802, as shown by a daughter born in South Carolina in 1802. He could not have been the father of a man born in North Carolina in 1802.
• This William (b. 1773) is documented in South Carolina as late as 1816, in Tennessee by 1819, and there is no record he was ever in Georgia. He was probably not the father of my William’s sister, Eliza (Bowdoin) Hackman, who was born in Georgia in 1817.
• Neither William Bowdoin nor Eliza Bowdoin was named in the will or any of the estate records of William Bowdon (b. 1773) in Coffee County, Tennessee.

However this “consensus” got started, it has no support from facts.

William Bowdon (b. 1773) was the son of Travis Bowdon (b. ca. 1750), often called Travis “Francis” Bowdon, though the “Francis” was not part of him name, but a result of people misreading “Travis” in handwritten sources. As it happens, this Travis Bowdon is the other major suggestion of the Ancestry “consensus” for being the father of William (b. 1802). This has clear and obvious problems as well:

• Travis Bowdon (b. 1750) also had another son named William—the one just named above, William Bowdon (b. 1773). The fact that William (b. 1773) was Travis’s son and not William (b. 1802) in Alabama is clear from the fact that Travis’s father, William Bowdon (b. 1720), named the grandson William in his will in 1773.
• Travis too was living in South Carolina in 1802 when William (b. 1802) was born in North Carolina.
• Travis died between about 1803 and 1805 in South Carolina, and could not have been the father of Eliza Bowdoin, who was born in 1817.

Not “Rev. William B. Bowdoin”

For as long as I have been around, and my dad says when he was young too, the legend has been passed around the family that William Bowdoin was a Methodist minister. But doing genealogy, it became clear that there is no contemporary evidence he was a minister. At some point in the past fifty or sixty years, some descendant placed a new, replacement marker in front of William’s original tombstone, with the inscription, “William B. Bowdoin, ‘Rev. Billy’, 1802–1870″—and set the legend in stone.

In genealogy, there is often conflict between family tradition and documented fact. The problem is, people tend to take what’s inscribed on a person’s tombstone as fact, by default; and the identification of William Bowdoin as “Rev. William B. Bowdoin” has proliferated in trees, as above. But in all the several dozen times his name appeared in documents—deeds, the census, his marriage record, church records—he never used a middle initial, and there is no reason to think he had one. This is a classic case of name creep.

And yes, I said church records. There are contemporary church records in the Autauga County circuit of the Methodist Episcopal Church, South, that show plainly that William Bowdoin was not a minister. The 1846 obituary of William’s wife Emily records that he was then a class leader; which in the context of an Evangelical church today, would be like a small group leader. Later, towards the end of his life, he served as an exhorter in the church. In the Methodist church of the day, an exhorter was a lay member who was authorized to hold meetings for prayer and exhortation. He was given a license by his pastor, which had to be renewed annually by the Quarterly Conference of the circuit, with an affirmation of the recipient’s good character. The office of exhorter was a predecessor of the Methodist office of lay speaker, but church documents were careful to stress that an exhorter was not a preacher and could not act in that role.

In today’s Evangelical churches, in which church offices are often limited to pastor or elder and deacon, and in which the only “license” anyone speaks of is a license to preach, I suspect descendants of William Bowdoin may have remembered his service as an exhorter and assumed he was a preacher. But no, he was not a preacher.

Picking up the ancestral trail

If the popular theories for William Bowdoin’s parentage, shown in so many family trees, are wrong, where do we go from there? We start with the things we are sure of:

• William Bowdoin was born in North Carolina in 1802. He came to Alabama as a very young man, no later than 1824, the year he married Emily Elizabeth Cooper in Montgomery County. He was either already settled in Autauga County, or settled there just after his marriage.
• William’s sister, Eliza Bowdoin, was born in Georgia in 1817. She came to live with her brother in Alabama as a young girl, and was living there by 1830, according to the census. She married J. Henry Hackman in Autauga County in 1838. She was married fewer than ten years before Mr. Hackman died, in about 1846, around the same time as William’s wife Emily. She and her children went to live with William, and she remained there until her children married.
• William and Eliza, who were literate, spelled their name Bowdoin, which suggests they could have been connected to the same Bowdoin family that settled in Coffee County, Alabama.

And from there, we can turn to DNA.

DNA lights the way

Fortunately, the trail of DNA is still very warm. We have some very good leads to start us on our way.

Y-DNA

Y-DNA is the DNA of the male Y-chromosome. Only males in the direct, patrilineal line of a family can take the test. The Y-chromosome is handed down, mostly intact, from father to son, so that a Y-DNA tester today has the same Y-chromosome as his father, grandfather, and every other man in the paternal line for dozens of generations. And he will match other men who descend from that same paternal line.

W.B., a man from my Bowdoin line, descended from William Bowdoin (b. 1802) and his son Reddin Read Bowdoin (b. 1831), took the Y-DNA test. He has a very close match to D.B., a man descended from the Bowdoins of Coffee County, Alabama. This match shows scientifically that the two men shared a genealogically recent common paternal ancestor. They are estimated, by two different estimates, from both the Y-111 test and the Big Y-700 test, to have been born as recently as about the year 1770—one to two generations before William Bowdoin (b. 1802).

D.B.’s known ancestors, in the timeframe of William Bowdoin (b. 1802), include:

• Turner Bowdoin (b. ca. 1812, Putnam County, Georgia);
• His father, William Bowdoin (b. ca. 1786, Granville County, North Carolina);
• His father, James Bowdoin (b. ca. 1764, Granville County, North Carolina);
• His father, William Bowdoin (b. ca. 1740, probably Virginia or Maryland).
• His father, William Bowdon (b. ca. 1720, probably Virginia or Maryland).

Turner Bowdoin, born after 1802, could not have been our William’s father. William Bowdoin (b. 1786) would have been only about sixteen years old in 1802 and did not marry his wife until 1806. He also had another son named William, William A. Bowdoin (b. ca. 1807). James Bowdoin (b. 1764) also already had a son named William, the one born 1786. William (b. 1786) was his oldest son; he did not have another son old enough to have had a child in 1802. In addition, he was living in South Carolina in 1802, and he and his wife had no more children born after about 1803. So we know that none of these men was the common ancestor between W.B. and D.B.

In that case, the late time-to-most-common-ancestor (TMRCA) estimate makes it very likely that the common ancestor was the next one in line, William Bowdoin (b. 1740). The estimate is based on a statistical model, and the common ancestor could conceivably have been earlier, back to William Bowdoin (b. 1720) or even beyond; but the further away from the average date of about 1770, the less likely.

There are other reasons to think the common ancestor was probably William Bowdoin (b. 1740). Of the southern branches of the Bowdoin family, his family is the main one that spelled the name Bowdoin. His father and brothers largely used the spelling Bowdon. And more importantly, the other type of DNA genealogy research, autosomal DNA, shows a large number of matches between the descendants of William Bowdoin (b. 1740) and the descendants of William Bowdoin (b. 1802).

Autosomal DNA

I have been conducting a major historical and DNA study of the Bowdoin family, and am publishing my results in two research papers, the first of which is attached to this article. Using the DNA matches of my grandfather, Robert P. Richardson (1926–2018), a great-great-grandson of William Bowdoin (b. 1802), and with the help of the Ancestry Pro Tools, I have catalogued more than 120 autosomal DNA matches with other descendants of our William (b. 1802), and more than 500 matches with descendants of William (b. 1740). These matches belong not only to my grandfather, but are shared matches with many of our cousins from William (b. 1802). Combined with the Y-DNA evidence, this is conclusive proof that William Bowdoin (b. 1740) was our ancestor.

William Bowdoin (b. ca. 1740)

William Bowdoin was born around 1740, likely in Virginia, the son of William Bowdon (b. ca. 1720) and his wife Elizabeth, “Betty”. He came to Granville County, North Carolina, with his parents, probably about 1760. He married his wife, Mary, in the early 1760s. (Contrary to the “consensus” of trees, there is no evidence his wife was “Martha Elizabeth Macon”.) All of his children were probably born in Granville County. During the American Revolution, he furnished supplies to the Continental Army.

In about 1788, William moved his family to Randolph County, North Carolina. In July and August 1788, he served as a delegate to the Hillsborough Convention, the first of two conventions in North Carolina to debate ratification of the new United States Constitution.

In about 1818 or 1819, William moved from North Carolina to central Georgia, probably Jasper or Putnam county, where two of his sons, James and Josiah, were already living. While in Georgia, he married a second time, to a widow named Mary (no other name known). Then about 1820, he continued his move on to Conecuh County, Alabama, where his youngest son Pleasant was living. In Alabama, his health rapidly declined, and he passed away 31 Jan 1821 in Conecuh County.

These facts are laid out in a contentious estate lawsuit which several of his grandchildren, children of his deceased daughters, filed against his surviving sons, alleging the daughters’ families had not gotten their rightful share of the estate. The case at least names William’s children and presents a clearer picture of his family than we would have otherwise, since he left no will.

William had these eight children, as named in the lawsuit:

1. James Bowdoin, born about 1764 in Granville County, North Carolina; died after 1840 in Monroe County, Georgia. Married 1785 in Granville County to Priscilla Williams. They were the ancestors of the Bowdoin families in Coffee County, Alabama, and of D.B., our Y-DNA match, as mentioned above.
2. Elizabeth (Betsy) Bowdoin, born about 1766 in Granville County; died about 1804 in Randolph County, North Carolina. Married John Macon 1786 in Granville County.
3. Martha (Patsy) Bowdoin, born 1768 in Granville County; died 1800 in Randolph County. Married Isaac Odell, probably about 1790.
4. John Bowdon (his preferred spelling of his branch of the family), born around 1770 in Granville County; died about 1850 in Richmond County, North Carolina. He was married and had descendants, but there is no record of his wife’s name.
5. Travis Bowdoin, born about 1772 in Granville County; died after 1860 in Graves County, Kentucky. Married about 1790 to a woman named Betsy, no maiden name known.
6. Mary (Molly) Bowdoin, born around 1773 in Granville County; died around 1810 in Randolph County. Married Gideon Macon about 1788 in Granville County.
7. Josiah Bowdoin, born about 1780 in Granville County; died about 1857 in Meriwether County, Georgia. Married (1) an unknown first wife, around 1800, (2) Nancy Freeman, 1826 in Henry County, Georgia; (3) Nancy Spencer, 1831 in Jasper County, Georgia.
8. Pleasant Bowden (his preferred spelling), born about 1785 in Granville County; died about 1851 in Montgomery County, Alabama. Married (1) wife unknown, probably around 1810 in North Carolina; (2) Mrs. Charlotte (Butler) Williams, 1844 in Montgomery County, Alabama.

A process of elimination

William Bowdoin (b. 1740) was almost certainly the grandfather of William Bowdoin (b. 1802). This means that one of his sons was William’s father. Using these criteria, I attempted to narrow down the possibilities:

1. The father of William (b. 1802) would not have another son named William.
2. He would have been married and having children by around 1802 when William was born, and still married, to a woman of childbearing age, in 1817 when Eliza was born.
3. Census records, if available, would be consistent with him having a son born in 1802 and a daughter born in 1817.
4. He was living in North Carolina in 1802 and in Georgia in 1817.

Based on these criteria, we can reject the follow for certain:

1. James Bowdoin (b. ca. 1764) definitely had another son named William (b. ca. 1786). William (b. 1786) was also too young to have been our William’s father. James and his wife were too old to have been Eliza’s parents in 1817. They were living in Georgia in 1817, but in South Carolina in 1802.
2. John Bowdon (b. ca. 1770) also had a son named William (b. ca. 1807). It appears unlikely that John was married in 1802. He never lived anywhere other than North Carolina, and was not living in Georgia in 1817. He had a will and estate proceedings, which do not name a daughter Eliza, and state that his son William was dead.
3. Travis Bowdoin (b. ca. 1772) was married before 1802 and living in North Carolina then. There is no definite evidence that he had a son named William, but some suggestion that his son Westwood may also have been named William. But Travis was still living in North Carolina in 1820, and never lived in Georgia.

Of the remaining two:

4. Josiah Bowdoin (b. 1780) looks most promising. He was married by 1802 (he had another son born in 1801), and in North Carolina at that time. He sold his North Carolina land in December 1816 and moved to Georgia, just in time to have a daughter be born there in 1817. There is no record that he had another son named William. The 1820 census shows him with several daughters under the age of 10.
5. Pleasant Bowden (b. 1785) was probably too young to have had a son in 1802, but it is conceivable. He was in North Carolina in 1802 and in Georgia by 1817. He did have several daughters under the age of 10 on the 1820 census. There is no record he had a son named William.

Conclusion

Based on this evidence, I have concluded that Josiah Bowdoin was very likely the father of William Bowdoin (b. 1802). I have a lot more to say about this research in my full research paper; and I will have a lot more still to say in Part 2, when I examine the DNA. Expect those results within the next week; I am still tidying up the DNA results for sharing! In the meantime, I would love to share the paper with you and would love your feedback.

Read the research paper this article is based on:

“Connecting William Bowdoin: The Problem of the Ancestry of William Bowdoin (b. 1802) of Autauga County, Alabama, Addressed through Records and DNA, and a General Overview of the Family of William Bowdoin (b. ca. 1740)“