African American Genealogy: Researching Poorly Documented Ancestors: Harriet Roberts (Northampton County, NC)

The basis of this article is a research report supplied to a client who is keen to expand his North Carolina-based family tree. Researching poorly documented ancestors sometimes requires out-of-the-box thinking. While the focus of this article is on free people of color in rural North Carolina, the approaches outlined are equally applicable to Caucasian ancestors.

The Research Process


The client relayed his interest in the researcher investigating his ancestor Harriet Roberts (1822- before 1870) who was the wife of Augustus Dudley Walden (1825 – Before 1870) of Northampton County, North Carolina.

The research process involved researching publicly available records including, but not limited to: US Federal and State Census records, colonial tax lists, available Northampton County, North Carolina deeds and marriage records, available family records, local history books via Google Books and, colonial-era newspaper articles, and various subscription-based online people finding services to place living DNA matches into the client’s family tree. The researcher also accessed the client’s Ancestry DNA matches for further clues.

Harriet Roberts

I was unable to records for Harriet prior to her marriage to Augustus Walden. Research for Harriet was going to involve working from the known into the unknown.

Marriage documents lookup

The marriage bond for Augustus Walden and Harriet Roberts. Source Citation: Description: Series: Marriage Bonds (1812 – 1867) | Source Information North Carolina, U.S., Marriage Records, 1741-2011 [database on-line]. Provo, UT, USA: Operations, Inc., 2015. Original data:North Carolina County Registers of Deeds. Microfilm. Record Group 048. North Carolina State Archives, Raleigh, NC.

While the above marriage bond proved the pending marriage of Harriet and Augustus, it provided no details about parents for either marriage party. The last roll of the dice, in terms of marriage records, was going to be a marriage license. This document has not surfaced yet.

My research for Harriet had a challenging start. However, based on Harriet’s age at the time she was married, I was satisfied this was her first marriage. By this, I mean that Roberts was her maiden name – and not a surname from a previous marriage. When it comes to researching female ancestors, surnames are everything. In order to have any hope of identifying who Harriet’s parents were…I had to know what her correct maiden name was.

U.S. State and Federal Census Records

I began my census records research with the 1850 Census to see whether there were any Roberts family living near Harriet and Augustus at the time of the census. Ransom Roberts and his family were living less than a mile from Harriet and Augustus:

U.S. 1850 Census for Northampton County, North Carolina. This image shows the households of Augustus Dudley Walden and Harriet Roberts-Walden above, and Ransom Roberts’ household below. Source Citation: Year: 1850; Census Place: Northampton, North Carolina; Roll: 639; Page: 66a

I had previously relayed to the client that he shared a number of DNA matches that link Harriet to Ransom. That they were family is not in question. I was tasked with trying to reveal what the nature of that familial relationship was. Was Ransom Harriet’s father? There are a number of family trees on Ancestry that claim this. However, the researcher wanted to exhaust all avenues before concurring with this assessment.

Looking for other Harriet Roberts in Northampton and Halifax Counties, North Carolina

DNA strongly suggested that Harriet was a descendant of a Virginia born Margaret Roberts who arrived in Northampton County, North Carolina from Virginia prior to the outbreak of the American Revolution. I attempted to trace all of Margaret’s known descendants to determine if there were any other Harriets within the Roberts family. There was one, a great-granddaughter of Margaret’s – and born a generation too early to be the client’s ancestor Harriet.

As with the client’s other colonial-era family in North Carolina, available records are patchy at best. I was not able to complete a full descendant tree for Margaret’s descendants. However, enough work was completed to determine that Harriet was not a common name among the Roberts. Any record bearing the name Harriet was going to be an important discovery. Unfortunately, I was not able to find such records.

Enter Stage Right: Property Deeds, Deeds of Sale, and Probate Records

I next began to look at Roberts’ family deeds and probate to see if there were any such documents that involved Augustus Walden as a witness, an administrator, or a guarantor. As the client was aware, sons-in-laws and brothers-in-law often acted as witnesses, guarantors, bondsmen, etc. for their in-laws within his family. This was a family dynamic stretching back to early colonial Virginia within his family. No such records involving Augustus were found online despite an exhaustive search. Nor did I find any records that named Harriet either under a maiden name or her married name.

Having exhausted all other avenues, the researcher zeroed in on Ransom Roberts.

Ransom Roberts

Ransom Roberts Household Composition

The first question I sought to solve was whether Ransom had any daughters.

A Snapshot of Ransom’s household over various census points. This image shows how I place information from the U.S. Federal Census from 1800 to 1840 into a family tree.

The image above tracks the composition of Ransom’s household in the 1820, 1830, and 1840 US Federal Census records. The 1820 Census had one adult female who was Lavinia MNU (maiden name unknown), Ransom’s wife. There are two females in Ransom’s household in the U.S. 1830 Census: Ransom’s wife, Lavinia, and their daughter, Mary. I will return to Mary shortly. In 1840, there were 3 females: Lavinia, Mary, and an unknown minor-aged female. Harriet could very well be the unknown young female in this household in 1830. Only further research would reveal whether this was a correct assessment…or not.

The 1850 Census sheds some more light on Ransom’s children:

In 1850, Ransom is shown with his wife and his children – all of whom feature in his estate records. Nicholas would have been a grandchild and remains unresearched at this time. However, there is Mary, exactly where she should be. It isn’t surprising to see the absence of Harriet. Harriet had already married Augustus by 1850 and was with him and their eldest children by the time of the 1850 Census. What this document confirms, is that Mary was Ransom’s daughter.

Ransom Roberts 1858 Estate sale in Northampton County, North Carolina

The image above is one page from an extensive probate file for Ransom’s estate. You will see the names of his children. These match the 1850 Census. Neither Harriet nor Augustus Walden were purchasers at this estate sale. While a will has not been found for Ransom, the estate sale documents, in my research experience, confirm who Ransom’s children were. Harriet nor Augustus was not among them. If Harriet had been Ransom’s daughter, not buying anything from her father’s estate would have been odd. This is especially so as she and Augustus literally lived just down the road from Ransom and his widow.

At this stage of the research, I believed Ransom was Harriet’s uncle.

Researching Ransom Robert’s siblings

The next stage of research involved researching Ransom’s siblings to see if any of them had a daughter named Harriet.

Ransom shown in-situ with his parents and his siblings. I would like to thank researcher Paul Heinegg for his excellent research which is freely available via his Free Africans of Virginia, North Carolina, South Carolina, Maryland, and Delaware website via His seminal work provided concrete clues to research.

The majority of Ransom’s siblings were fairly straightforward to research, with the exception of his brother, James. James aside, the researcher was able to name Ransom’s sibling’s children – and account for Ransom’s siblings’ whereabouts. Harriet was not a name that appeared among any of the siblings’ children. A number of Ransom’s siblings had also left North Carolina for Vigo County, Indiana around the time Harriet was born. The sole person left to research was the one who was the most difficult to research due to a scarcity of records: James.

Enter James Roberts

Before we look at James, I would like to point out a basic research issue around James. James was a heavily used name among the Roberts family. Almost every branch of the family contains the name James in every generational line of the family. Many of these James have been confused with one another, or merged, in various online trees and in some lineage books. It took some time to untangle as many of the various James Roberts in Northampton County, North Carolina. In time, the researcher was able to create a basic page for the James who was Ransom’s brother:

Apparently, there is an existing will for James. It is referenced by the researcher Paul Heinegg in his book Free African Americans of Virginia, North Carolina, And South Carolina, Maryland, and Delaware. I was unable to find this will. What was found was the rest of James’ probate file and involved a complicated land sale arranged by his widow, Martha. No minor children were cited anywhere in the probate file. Given almost all the pages were about the sale of James’ land, this is not surprising.

There was a growing picture that Harriet was orphaned. This would explain why it has been so difficult to definitively identify her parents. Ransom died in 1826. Martha died some time around 1830. Harriet would have been between 4 to 6 years old by the time both of her parents had died. Which makes it more plausible that Harriet could be the second young female in Ransom’s household in 1840.

However, in 1830, there was clearly a young female in James and Martha Robert’s household. The researcher believes this to be Harriet. This finding is echoed by DNA matches for the client on Ancestry.

Enter Stage Left: Genetic Genealogy

The client has around a dozen or so DNA matches that link Harriet back to Ransom and James’ overall family group. Among them, there were some strong hints that link back to James. This is covered later. However, before looking at these matches and the ThruLines they generated, it is important to again relate that ThruLines should be viewed as potential hints – no more and no less. The researcher can’t stress this enough. The researcher spent some time going through numerous family trees from the client’s DNA matches. Thankfully, the majority of the lines back to Harriet were correct, well documented, with good citations.

20+ DNA matches, who were individually researched, provides us with a picture that looks like:

A snapshot of the client with DNA matches who descend from two of James Roberts’ brothers: Hansel and Willis. I won’t supply a breakdown of each and every step in the DNA triangulation process to spare you reading a book-length article.

DNA matches for descendants of James’ brother, Hansel Roberts.
DNA matches for descendants of James’ brother, Willis Roberts.

At the very least, I can confirm he has found Harriet’s most immediate family. She is a granddaughter of Jonathan Roberts and Mary Scott. The latter person, Mary, brings in an additional Scott line to the client’s ancestry which is seen among his Scott family DNA matches.

Chipping away at Jonathan and Mary’s children, only one child remained who could be Harriet’s father, and that is James.


It’s worth bearing in mind that Northampton County, North Carolina is a very rural county to this day. When researching early arrivals to an uninhabited or sparsely inhabited place, there are a few considerations to bear in mind. There was no form of local government when the early settlers of Northampton County, North Carolina settled in that place. By this I mean there was no centralized place to record baptisms/christenings, marriages, or similar activities in that community’s early years. Church services were most likely held in people’s homes until a church could be built. These are but a handful of examples that makes researching early rural communities difficult.

As the community developed, challenges remained. Relative poverty is one consideration. Could a groom afford a marriage bond? Could his intended bride’s family afford one? Could a circuit-riding preacher be counted on to record marriages, births, and deaths? And if a circuit preacher could record such vital information, where would he have recorded it?

While we, as researchers, endeavour to provide a solid paper trail to document familial relationships – there are times these records simply don’t exist. So we have to rely on whatever paper trail we can uncover, and then use deductive reasoning and critical thinking to fill in the gaps. It’s not ideal, however, poorly documented ancestors didn’t know they would have descendants looking for them over a century later. And then write our discoveries and research rationale and add this information as a report to the poorly documented ancestors in our tress to enable other researchers to follow our information and logic stream.

We are available for your genealogy research project!

Whether you’re new to genealogy, don’t have the time to pursue your own research, or have a stubborn brick wall you just can’t break through: you can hire one of our experienced research team who will happily work with you to achieve your genealogy research goals. As the first Black-owned genealogy research company, our African American genealogists excel at African American genealogy! As African Americans with caucasian ancestors – our genealogists are just as experienced in researching European-descended American ancestry.

The short video below covers what we do and how we do it:

How to contact us

For more information about our genealogy research services, our contact form, and our research service contact, please visit:

3 thoughts on “African American Genealogy: Researching Poorly Documented Ancestors: Harriet Roberts (Northampton County, NC)

  1. Brian!
    I was just about to sign off when this post popped up, but I just HAD to read through it, quickly. Great work – but I must say, we should have been working together, because it sounds like we’ve both had our heads buried in the same record sets, and we’ve come to similar conclusions. As I think I’ve mentioned to you, I’ve also gone blurry-eyed working through the available court records, looking for any mention of Augustus, Harriett, and these other Walden and Roberts suspects…lol.

    There is one thing I want to point out to you, but you are about to have a show and I need to run out, so I’ll reach out a little later. Thanks for sharing your work, publicly. It never occurred to me to do that with client work – but it’s a good idea!

    Keep blazing!


Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: