Documenting my family's past for future generations. My family tree includes the Smith/Mansell families of Alabama and Oklahoma, the Castle/Day families of Kentucky and Oklahoma, the Wheat/Ming families of Texas and Oklahoma, and the Bell/Roberts families of Mississippi, Tennessee, and Oklahoma.

Friday, July 17, 2020

Roberts Family Revisited

Jackson County, Tennessee, was created in 1801, formed from the eastern part of Smith County. It is located in Middle Tennessee about 80 miles east of Nashville. Ancestry users often confuse it with the city of Jackson in Madison County in West Tennessee. In fact, it's almost a giveaway that Ancestry users haven't really done their own research (ie., have borrowed names and places from other users) if they identify a Jackson County person as one from Jackson, Madison County. Jackson County is best known now as the location of Historic Granville (granvilletn.com). Formed in 1999 to preserve the history of the riverboat town, the attractions include museums, a Pioneer Village, the Sutton General Store, music events, heritage days, and a genealogy festival. My brother and I visited a few years ago.


Granville Museum

Sutton General Store

My brother having his first pimiento cheese sandwich (and an RC Cola)
at Sutton General Store


An 1872 courthouse fire destroyed many records of early Jackson County. The 1810 census is missing, as well as marriage and probate records from 1801 to 1871. I suspect that is the reason that I have had such a hard time identifying the parents of my 2nd great-grandparents, Stephen and Elzina (Huff) Roberts. They first appear on the 1850 census as a married couple, so since the previous censuses don't list children by name, I have no way of knowing to which families they belonged.

I've been following genealogist Amy Johnson Crow's series about her research method, which she calls WANDER. First, she says to identify WHAT you really want to know. The next step is to ANALYZE what you already have, and then NOTE what you are missing. I have been trying to apply this method to the question, "Who are the parents of Elzina Huff and Stephen Roberts?" So let's analyze what I already know.

As I said above, the first time that either Stephen or Elzina appears in records is on the 1850 census. They are living in District 15, Jackson County, Tennessee. Stephen's age is reported as 40 (birth year 1810) and his occupation as farming. His wife is enumerated as Elsy, age 30. (Every census shows a different spelling of Elzina's name. Elzina/Elzinia/Elzinah is the spelling of her name on her 1874 divorce complaint, which she "signed"--with her mark--so that is the name I have used in my tree. More about the divorce later.) Children Polly, age 7; Samuel, age 4; and Nancy, age 1 are enumerated with their parents. All members of the family were born in Tennessee.

On the 1860 census the family is enumerated in District 15 of Jackson County with the nearest post office listed as Granville. Stephen is age 44 (birth year 1816) and his occupation is given as Day Laborer. Elcena is age 34. Children are Mary (Polly), age 16; Nancy, 11; Henry, 9; John, 4; and Thomas, 2. 

The 1870 census, again in District 15, Jackson County, shows Stephen age 50 (birth year 1820); Eliza, age 45; Nancy, 20; Henry, 18; Ellis (John), 16; Permelia, 14; Thomas, 11; Caleb, 7; and Nathan, 2. Included with the Roberts family at this household (Dwelling Number 176) are George Huff, age 26, and his wife Mariah, age 19. 


1870 Jackson County census
George's wife Mariah is on the next page

A couple of weird things show up on this census. First, a new daughter has been added between John Ellis and Thomas--Permelia, age 14. I think this is my great-grandmother Cornelia, who would have been 5 on this census. As hard as it is to believe that the census taker could get both the name and age wrong, I think that is what happened. I have looked for a Permelia, just in case she existed, but I have never found her again on any document, and if you look carefully at the 1860 census, you can see that there is not room between John and Thomas for another child, but there is room between Caleb and Nathan. In addition, Cornelia will show up on the divorce complaint as a child of the marriage and Permelia will not. 

The other weird thing is the addition of George and Mariah Huff to the household. I have puzzled over this for years, especially after I realized that newlyweds George Washington and Mariah (Lambert) Huff were enumerated in their own home (Dwelling Number 63) on this same census. My guess has always been that even though the census taker mistook Cornelia's name and age, he took his job seriously and enumerated the people he found at the Roberts household on September 3, 1870. George and Mariah were just visiting. The burning question is, How are they related to Stephen and Elzina? If I knew that, I would have a big clue as to the parents of (probably) Elzina. Was George her cousin, her brother? Maybe I do have a clue here to analyze later.

The next big piece of information I have about Stephen and Elzina is Elzina's divorce complaint from 1874. In it she seeks divorce from Stephen, with whom she had lived "tolerable harmoniously from the time of the marriage except at times when defendant would indulge in the use of ardent spirits." As stated in the complaint, Stephen was "a habitual drunkard when he could get whiskey" at which time he was "in the habit of abusing the complainant." 


1874 divorce complaint, page 2 of 7


Elzina lived with this state of affairs for 35 or 36 years, as she states in the complaint, until May 1872 when Stephen "willfully and maliciously abandoned [her] and children without any just or noble cause." She names their children as "John E. Roberts, Thomas J. Roberts, Calip (Caleb) L. Roberts, Cornelia Roberts, Nathan Roberts, Nancy Roberts and Henry Roberts, the two last being of age." She states that Stephen left her and the children "without any provisions to subsist on whatever" and "for more than two whole years before the filing of this bill ... has the defendant ever furnished...one item of support...either directly or indirectly." Sadly, she further states, within those two years two of the children, John E. and Caleb, had "departed this life." 

I hadn't read the 7-page handwritten document in a while. It isn't easy to read due to the handwriting and the legal language. However, analyzing it for more detail, I found a couple of statements that may help me in my search for Elzina's parents. While I had noticed before that Stephen and Elzina lived "in Jackson County on the waters of Martin's Creek," I had never noticed Elzina's statements that she was "born and raised in Jackson County Tennessee within two miles of where she now lives" or that she and Stephen "intermarried in Jackson County Tennessee near where [she] now lives." If I could ever determine from land records exactly where Elzina lived on the 1850, 1860, and 1870 censuses, I could look for neighbors within two miles who could have been her parents. Unfortunately, those land records probably don't exist as they fall in that window of 1801-1871 for which documents are scarce because of the courthouse fire. 

Moving on to the last piece of information I have about Elzina--the 1880 census of Caldwell County, Kentucky. I have tried and tried to determine why the Roberts family (minus Stephen) moved to this particular place. I have looked for people with any related family name living close to the family on this census, and I can't find any obvious connections. Living in the Harmony District, Caldwell County, Kentucky, are Elmira, age 54, divorced; Nancy, 24; James H., 18; Thomas J., 14; Cornelia, 13; and Nathan J., 12. John Grider, farm laborer, is also enumerated with the family. (Later he would marry Nancy.)

Oh, and by the way, in 1880 Stephen is still back in Jackson County, age 67 (birth year 1813), claiming to be a widower. He is living with his nephew Nathan, son of Stephen's brother Caleb. 

The 1880 census is the last mention I can find of Elzina in written documents. As to her children--I don't know what happened to James H. (Henry). Thomas stayed in Kentucky and is enumerated with wife and children on the 1900 census in Trigg County (next door to Caldwell), where he died in 1936. Cornelia married Thomas Jefferson Bell in Chickasaw Nation, Indian Territory, in 1893 and is enumerated there with her husband and children on the 1900 census. James Nathan married first, Isabelle Anderson, and then Callie Grotts in Troy, Indian Territory, in 1903. Later Cornelia and Nathan will both live in Hughes County, Oklahoma, and Nathan will attend Thomas's funeral in Trigg County, Kentucky, in 1936. Cornelia died in Dustin, Hughes County, in 1938, and Nathan in Barnard, Hughes County, in 1951. Cornelia is buried at Dustin, and Nathan about 15 miles south of Dustin at Lamar. I requested death certificates for Cornelia and James Nathan from the State of Oklahoma, hoping they might show the maiden name of their mother. No luck--apparently there was no consistent reporting of deaths in Hughes County, even as late as 1951.

Before we leave the paper trail of Elzina Huff Roberts, one last observation--on her date of birth as revealed by the census record. She is 30 on the first census in which she appears by name, the 1850, making her birth year 1820. By 1860 she has only aged 4 years to 34. In 1870 she is 45, and in 1880 she is 54. These last three censuses are fairly constant, so I have always estimated Elzina's birth date to about 1826. However--in the divorce complaint, dated 1874, she claims that she and Stephen had been married 35 or 36 years. That would mean they married in 1838, which would make Elzina 12 when they married. Their first child, Polly, was born in 1838. I guess it's just hard for me to believe that Elzina could have been 12 or 13 at the birth of her first child. Now I tend to think it's more probable that she was born about 1820. Stephen's birth dates on the census range from 1810 to 1820 with no consistency except that he is always older than Elzina, usually about 10 years. Most researchers give his birth year as 1813.

You know how you can look at something for years and then all of a sudden really see it? I just realized that on my Ancestry page for Elzina I have actual birth and death dates. The birth date is 8 September 1826 and the death date is 16 October 1901, place given as Troy, Indian Territory. However, I have no idea where these dates came from. I think they might have come from a descendant of James Nathan. It sounds like they could be from a headstone, but in the past (and then again this week) I looked and looked in the vicinity of Troy (now located in Johnston County, OK) and found no Findagrave entry for Elzina, or anyone with those dates.

N in Amy Johnson Crow's research method, WANDER, stands for Note What Is Missing. Wow, I feel like I have tried everything but I'm sure there is something I've forgotten. I don't know what land records are available for Jackson County. That might be helpful if Elzina and Stephen were living on family land. I have looked and looked at the census records, hoping to find new clues. Amazingly, last week the census records in combination with Ancestry's Thru-Lines gave me a new piece of information. 


Ancestry Thru-Lines for Stephen Roberts

I always wondered what happened to Mary (Polly) and Samuel, the two oldest children of Elzina and Stephen. Samuel never appears again after the 1850 census where he was age 4, so I assume he died young. Polly is on the 1850 and 1860 censuses with the Roberts family, but I was never able to determine what happened to her or where she was after 1860. Well, as it turns out, she wasn't very far, as I have discovered since beginning this post. 

I decided to check Ancestry's Thrulines for Stephen and Elzina while I was analyzing what I know about the Roberts family. I found four descendants of Polly that showed in their trees that she had married John Clemons. Clemons is one of those Jackson County names that you recognize, although I didn't know my family had any connection to them. Each of the descendants of John and Polly share 7 cm's with me. (By the way, on August 1 Ancestry is deleting all matches who share less than 8 total cm's. If I had not found these matches in July, they would have been gone in August.)

John Clemons was born in 1815 and had been the head of household on censuses since 1840. His first wife was named Delilah (some trees show her as a daughter of James Pharris and Alcy Moreland). They had Alexander, Caldonia, Harriet, Ollie Frances, Mary, Martha, and Albert, before Delilah died sometime in the early 1860's. John and Polly must have married about 1865, because Polly gave birth to Sidney in 1866, followed by Margaret, Moreland, Cora Jane, and Thomas. (Sidney's death certificate confirms that his mother was indeed Polly Roberts. Tom's death certificate names "Mollie?" as his mother. Polly and Molly were both very common nicknames for Mary.)

On the 1870 census the Clemons family is enumerated at household #171, while the Roberts family resides at household #179. John, age 55, is the head of household, followed by Mary, age 25. The rest of the family consists of Martha and Albert, the children of John's first wife, and Sidney and Margaret, who are both presumably Polly's. John is listed on the 1800 Census Mortality Schedule, which says he died in February 1880 at age 69. Polly must have died within months, because on the 1880 census the younger children, Sidney, Margaret, Moreland, Cora, and Thomas, are all living with their half-brother Alexander and his family. The marriage of John and Mary/Polly could have easily slipped through the cracks. In fact, in the indexing of the census entry Mary is listed as one of the "inferred children" of John. I'm not sure I would have ever discovered it without the trees of my DNA matches.  

I know a lot about how my great-grandmother Cornelia's life turned out. She was always an unhappy person, I think. She married a kind man and was awful to him, according to her granddaughters. I always felt sorry for her--brought up in this dysfunctional family with a drunken father, brothers who died, and a mother who was trying her best to keep the rest of them from starving to death. But poor Polly! Reading between the lines of the stark data of her life, you can see that she took her chance to get out of an unhappy situation by marrying a man 30 years her senior. As a recent widower in her neighborhood, he was probably looking for someone to take care of his existing children. She then had five children in ten years and died. It is to be hoped that her children remembered something about her, but their loved ones obviously didn't even know her name for her children's death certificates. How sad!

Death certificate of Sidney Clemons, only child to
show mother as Polly Roberts


I used to correspond with several descendants of Jackson County families, but we were all stuck at the same place and eventually we lost contact. I thought it couldn't hurt to contact some descendants through Ancestry, especially those who live in or close to Jackson Co. I found a DNA match to me who had a tree on Ancestry that included some of the same names I have in my tree--not a descendant of Elzina and Stephen, but of some of their possible ancestors. She looked like a friendly person with a carefully constructed tree, so I messaged her. After several messages back and forth, we identified a possible common ancestor, but in one of her emails she mentioned an ancestor of hers that I had always had questions about. In answering my questions, she mentioned her ancestor's first husband--John Clemons! It turns out that her ancestor was John's first wife and my 2nd great-aunt Polly was his second. Many of her first cousins are involved with the genealogy festival at Granville, and one of John and Polly's descendants runs the town! I'm so glad I contacted her. 

Back to the question of the day: Who were the parents of Elzina and Stephen?

I wish I could remember who first told me, or where I first read, that Elzina's parents were William Nathan Huff and Susanna Toney. Almost everybody who has Elzina in a tree identifies these two as her parents. The trouble is, there is absolutely no proof. 

There was a Susanna Huff in Jackson County. Her birth dates vary widely on the censuses from 26-44 in 1820 to 60-69 in 1830 to 50-59 in 1840. Depending on which one you believe, she might have been too old to be the mother of Elzina, whose birth date falls somewhere between 1820 and 1826. No proof exists for her maiden name of Toney either, as far as I have found, although Toneys did live in Jackson County. She is never named in conjunction with her husband. In fact, Betty Huff Bryant, probably the leading expert on Martin's Creek, remarks in her book, Building Neighborhoods, that "George Huff is possibly the wandering husband of Susanna Huff." Whether Susanna's husband was dead or "wandering," she is named as head of household on the 1820, 1830, and 1840 censuses. There are children in her household (enumerated by age in 1830 and 1840) who could be Elzina.

However, as far as I can tell, there is no such person as William Nathan Huff. There is a William Huff and there is a Nathan Huff, but they are not the same person, and no documentation exists to prove that a person by either name married Susanna. An article written by a Huff researcher in The Jackson County Family History Book: 200 Years of Memories in 1989 may provide the origin of "William Nathan Huff." In trying to reconcile the available documentation with the information passed down to descendants, the writer of the document surmised that William Huff, the probable husband of Susanna and father of Sam Huff, had a second name, Nathan. It appears there was a custom among German communities of giving two names--one for casual use and one for legal records. Since descendants insisted that the husband of Susanna and father of Sam was called Nathan, that person must have been named William Nathan Huff. The identification of William as Susanna's husband seems to be circumstantially based on legal and land records and family legends.

The aforementioned article names the children of William and Susanna as Sam, George, Nathan, Sara and Elender. Elzina is not named as one of their children, although many Huff researchers make much of the fact that Sara/Sally was Elzina's sister, and that the two sisters married brothers, Caleb and Stephen Roberts. Some also put Nathan "Tash" Huff in the same category as Elzina: born too late to be the son of William and Susanna. I'm not sure to which George the article refers, but George Washington Huff (who visited Stephen and Elzina in 1870) was the son of Sam, not William.

The Roberts side has been even harder. I had no family story about the parents of Stephen. There were several Roberts men in early Jackson County, any of whom could have been the father of Stephen and Caleb, including an older Stephen Roberts that was living and buying land there in 1813. Most trees on Ancestry show the parents of Stephen as Reuben Roberts (1744-1841) and Mary "Millie" Asher (1759-1847) who lived in Warren County, Tennessee. I have never actually taken this identification seriously because they were too old to be Stephen and Caleb's parents. 

So where does all this get us, in terms of figuring out the parents of Elzina and Stephen? Well, I want to talk about names. It was common in Southern families to name children after their ancestors, and sometimes this naming convention followed a specific pattern: first son for father's father; second son for mother's father; third son for father; fourth son for father's eldest brother; first daughter for mother's mother; second daughter for father's mother, etc. While I don't think Elzina and Stephen followed this tradition strictly in naming their children, I do think there might be some clues in the names they gave them. 

The children of Stephen and Elzina Roberts were: Mary, known as Polly; Samuel; Nancy; James Henry; John Ellis; Thomas Jefferson; Caleb; Cornelia; and James Nathan. These are some of the possibilities that have occurred to me, based on these names:
1. Samuel was named for Elzina's father. Elzina was not really the daughter of William and Susanna but of their son, Samuel. Samuel married Lucinda Hardcastle in 1844 when he was 38 years old. It is entirely possible, and likely probable, that Sam was married before. Elzina could be his daughter by his first wife, possibly raised by his mother, Susanna. Susanna wasn't Elzina's mother, but her grandmother.
2. Although Henry was usually Henry, on the 1880 census he is listed as James H. That means that the Roberts parents used the name James twice for their children, for Henry and for Nathan. Does this have any significance? One of my cousins has broached the idea that Elzina was the child of James Pharris, who proudly claimed to have had four wives and fathered 20 children. He wasn't clear about whether any of those children were born out of wedlock. This would be one explanation for my many Pharris DNA matches. There was a James Roberts in Jackson County who could have been Stephen's father, but his will does not list Stephen or Caleb as his children.
3. John was listed as John, Ellis, and John E. on various censuses. I had never come across anything to explain the name Ellis, until I recently saw the tree of a DNA match who showed the father of Reuben Roberts to be Ellis B. Roberts. If true, this would certainly give credence to the identification of Reuben Roberts as the ancestor of Stephen.
4. Caleb, of course, is named for Stephen's brother. This is the one name that is not in question.
5. My great-grandmother Cornelia's middle name has been given as "Dee" and as "Orange," which I thought was highly unlikely until I noticed that her headstone reads "Cornelia O. Bell." Could ancestors of this family come originally from Orange County, North Carolina, or is there some other significance to that name?
5. While usually enumerated as Nathan or Nathan J., the youngest Roberts son went by Nathan or "Naith" during his life. His headstone reads "James N. Roberts." Nathan, of course, was the name that descendants remembered as the husband of Susanna, usually known now as William Nathan Huff. He could have been Elzina's father or grandfather. However, and this is interesting, one of James Nathan's children, buried in the same cemetery, is Napoleon "Rube" Roberts, and he had a son named Reuben!






All headstones from Lamar Cemetery, Hughes Co. OK
(courtesy of Connie on Findagrave)


The D in Amy Johnson Crow's WANDER system stands for Discover, as in new records or new resources. Now we turn to what my DNA results can contribute to the question of Elzina and Stephen's parents. 

I have many matches to Huff descendants, as does my brother. Some of his are larger than mine, and some of mine are larger than his. (Which is a good reason for getting siblings tested.) Recently I worked through my matches on Ancestry, color coding by family. Since the Jackson County families are so interrelated, I made a group that I just called "Jackson County people." 

Many of my Huff matches led back to Common Ancestor William Nathan Huff, but of course, nobody had any sources for him. In many cases my Pharris matches were theoretically explained by the identification of William Nathan's mother as Nancy Pharris. As I have noticed for a long time, many of my unexplained matches are related to the Embry family--some of whom lived in Jackson County, but most of whom lived in Butler County, Kentucky. I have puzzled over this for a while now.

This is another case of having had something at your fingertips that explained something, but having forgotten you had it. I knew I had read something attached to a tree on Ancestry called "The Search for Sam Huff's Father." My new DNA cousin with connections to Granville supplied me with the original article, which was published in The Jackson County Family History Book: 200 Years of Memories. It says, "One Huff researcher in Ohio says the legend in her family is that the older Leonard Huff (...who was on Martin's Creek with the Pharis family by 1803) moved to Butler County, Kentucky, and all his sons accompanied him except two--[Leonard] "Knight" and William--and those two remained in Jackson County, Tennessee." Well, that explains Huff and Pharis connections in Jackson Co., as well as the connection with Butler Co. Kentucky Embrys. I still haven't figured out if I actually have an Embry ancestor, or if the Embry name comes in downstream with the many Huffs and Pharrises that married into that family.

My largest Huff match is a Huff on his father's side and a Pharris (with an Embry ancestor) on his mother's side. He is in his 90's and is the descendant of Sam Huff through his son George Washington Huff--the same George Huff who was visiting the Roberts family the day the enumerator came around to take the census in 1870! What is even more amazing, he is only one generation removed from George and two from Sam, since George was his father's father. He falls within the parameters of being my 1/2 2nd cousin, twice removed, which he would be if Elzina was Sam's daughter, and she and George were half-siblings. 

One ancestor I had tentatively eliminated long ago was Susannah (Toney?) Huff, as existing documentation said she was too old to be Elzina's mother. I wasn't even positive that she was Sam's mother, although the author of "The Search for Sam Huff's Father" thought she was. Again, he listed the children of William and Susanna as Sam, George, Nathan, Sara, and Elender. No Elzina, even though Sara/Sally was supposed to be her sister. One thing I hadn't done in a long time was to search for Toney matches. The last time I looked I didn't have many and those could be explained by trees that identified William Huff and Susanna Toney as the parents of Elzina. Well, times change. Just doing a simple search for Toney as a "Surname in Match's Tree" gave me over 70 results. And very few of them were tied to Susanna, the wife of William Huff; most of them went back to the same Toney family living in Goochland and New Kent counties in Virginia. I haven't made the connection yet between this family and Susanna, but that is something to work on.

These are all possibilities to Evaluate, the E in Amy Johnson Crow's WANDER system. Have I answered the "What Do I Want to Know" question about the parents of Elzina and Stephen Roberts? No, but I have Analyzed everything I do know, even things I forgot I knew. I have Noted what is missing and Discovered new resources, something I didn't think possible. Now I need to Evaluate and Repeat as necessary. 

One simple thing I hadn't really taken into account--evaluated--is the lack of records in Jackson County. Many marriage documents and wills do not exist. Maybe there aren't as many secrets in Jackson County as I thought; there are just undocumented marriages and legitimate children. Everybody has made guesses about Jackson Co. people in their trees, because there is no paper proof; so can there ever be a definitive answer? 

Maybe the answer lies in GPS--the Genealogical Proof Standard--the set of "best practices" for genealogists. In 1989 the Huff researcher who wrote "The Search for Sam Huff's Father" took the facts that had been documented along with the stories that had been passed along and reached what he felt was a logical conclusion. Maybe I can add a third component, DNA testing, arrive at a logical conclusion about the parents of Elzina and Stephen Roberts, and use GPS to make my case. Maybe that's the best I will ever be able to do.

 

1 comment:

  1. Hi, Becky are you still pretty active on the Roberts history? I have kinda been reading some of your stuff off and on the past two years. And would love to know what all you have on the Robert's/grotts/gollahone if any

    ReplyDelete