Tracing Your Roots: Did My Ancestor Work in a Prior Enslaver’s Home?

Example of 1870s domestic service: Eugen Keller and an unnamed nanny in Pernambuco, Brazil, circa 1874, where slavery was still legal at the time. (George Ermakoff/Wikimedia Commons)
Example of 1870s domestic service: Eugen Keller and an unnamed nanny in Pernambuco, Brazil, circa 1874, where slavery was still legal at the time. (George Ermakoff/Wikimedia Commons)

Repeating patterns in Reconstruction-era census records point to possible connections during slavery.


Dear Professor Gates:

I’m trying to determine if my third great-grandmother (from my mother’s paternal side of the family) was a slave or if her mother was. In the 1880 census in Lytle’s Fork of Scott County, Ky., she is listed as Polly Roberts, a servant in the household of a white man by the name of Silas M. Berry. Among others listed in the household is a David Roberts. He is also a servant, and there is nearly a 50-year age gap between the two, so I’m not sure what their relationship could be.

I dug further and found that in the 1870 census, Silas Berry has a Polly Moss (age 16) in his household, in addition to others with the same surname. Although the age doesn’t seem to match up (22 in 1880, 16 in 1870), I’m trying to determine if these records are for the same person.

I tried checking back further, but I’m really confused on how to make sense of the 1860 and 1850 Slave Schedules. Any tips you can provide would be appreciated. —Nelly Dee

You are right to suspect that Polly Roberts and Polly Ross are the same person. True, the ages don’t align perfectly, but it is important to keep in mind that ages can vary across records—sometimes as much as 10 or 15 years. This is particularly true for former slaves who may not be completely sure about when they were born.

Likewise, surnames can vary for those who were enslaved. Our previous column “Were Slaves Surnames Like Brands?” noted:

[T]here were a variety of ways that former slaves adopted surnames, only one of which was adopting the surname of their slave owner. There were instances of an enslaved family passing down a surname through several generations. Sometimes people kept a surname of a previous slave owner, and sometimes they had a name chosen for other reasons, such as their occupation or a long-standing family connection to a name. A window into the complexities of slaves’ surnames can be found by reading the article “A Perspective on Indexing Slaves’ Names” (pdf), by David E. Patterson, in the American Archivist.


Noting associates of your ancestors (such as household head Silas Berry) can help to bridge such inconsistencies. Based on the connection to Berry in both 1880 and 1870, it seems likely that the Polly recorded in both records is the same person. To bolster this assumption, we looked at other names in the household as well to see if they also repeat under different surnames.

Striking Consistencies in Census Records … 

As you mentioned, Silas Berry’s household in 1880 records him living in Lytles Fork, Scott County, Ky. He was born about 1797 in Virginia. Others in the household include your Polly Roberts (born about 1858) and David Roberts (born about 1810). The other servants in the household should be noted, too. They include Thomas Locus (born about 1864), Martha Locus (born about 1865), Bettie Jenkins (born about 1875) and York Jenkins (born about 1870).


When we examined the household of Silas Berry in 1870, we saw that a number of his servants that year have first names that match those in his household in 1880, but they were all recorded under the surname Moss. This group includes Polly Moss (born about 1854), Thomas Moss (born about 1862) and Martha Moss (born about 1863). David Huble (born about 1805) is also a good match for David Roberts in 1880.

There are a few explanations as to why the surnames would change. One could be that the Moss name was inherited from a former slave owner and that by 1880, the servants had adopted surnames of their own choosing. Since you cannot be certain about the circumstances, you will want to investigate all possibilities.


… and Clues in Slave Schedules 

We started by searching the 1860 Slave Schedules for any slave owners in Scott County with the surnames that are associated with your Polly, namely Berry, Roberts, Moss and Huble. Since the enslaved were not identified by name in the slave schedules, a strategy for reading these enumerations to locate your Polly would be to determine her age, as well as the genders and ages of other servants in the Berry household, and compare those with the genders and ages of those listed households of potential former slave owners.


Silas Berry was recorded as a slave owner in the 1860 U.S. census. The record includes a 3-year-old female (born about 1857), who would be a close match in age for what we know about your Polly Roberts. There is no good match for David in this household; however, on the same page there is a 36-year-old male (born about 1824) owned by B.Y. Glap “for Silas Berry” that is a possible match. This would mean that this person was owned by Silas Berry but was currently working for B.Y. Glap. We did not locate any other records for Glap in the slave schedules, so perhaps he did not own any slaves himself.

Since you know that Silas Berry lived after the end of slavery, you are probably not going to find mention of his former slaves in a probate file, unless he left something to them and identified them as such. You may want to examine deeds in deeds in Scott County, Ky., since slave sales are often recorded in deed books and you may be able to locate a record of Silas Berry having purchased slaves. You may also want to investigate further for a possible connection to the surname Moss.


A Jane Moss owned slaves in Scott County in 1860, although none of the slaves in her household in 1860 are a good match for your Polly Roberts/Moss. Locating her in the 1860 U.S. census reveals that Jane Moss was born about 1780 in Maryland and had a 16-year-old also residing in her household named Ellen Moss, who was born in Kentucky. There is also a record for her having resided in Scott County in 1850 in the household of her husband, Thomas Moss, who was born about 1773 in Virginia. This tells you that Thomas Moss likely died between 1850 and 1860, perhaps leaving a probate record that may mention his slaves.

Continue the Search in Probate Records 

We located an administration bond for the estate of Thomas Moss dated Oct. 27, 1859 (on; subscription required). William Parker was named the administrator of his estate, and John Davis and Cobb Parker the sureties. In the same collection was an administration bond for Jane Moss dated April 21, 1864. James H. Davis was named the administrator and John David and John Moss the sureties. This means that both Thomas and Jane Moss died prior to the end of the Civil War. These are just the bonds for the administration of the estates, but there are likely other probate documents, such as a will or inventories of their estates, that you could research.


To find probate records in Scott County, Ky., you can search digitized records available through The site’s searchable index did not yield results, but you could browse the scanned index to see if the database missed something. You could also contact the county clerk to see if that office has a probate file for either Thomas or Jane Moss. Follow this same approach for any slave owners with the Huble surname, too. Once you have identified potential slave owners, search for any records you can for the slave owner, since these may reveal more about your ancestors.

Henry Louis Gates Jr. is the Alphonse Fletcher University Professor and founding director of the Hutchins Center for African and African American Research at Harvard University. He is also chairman of The Root. Follow him on Twitter and Facebook. 


Send your questions about tracing your own roots to

This answer was provided in consultation with Meaghan E.H. Siekman, Ph.D., a senior researcher from the New England Historic Genealogical Society.Founded in 1845, NEHGS is the country’s leading nonprofit resource for family history research. Its website,, contains more than 1 billion searchable records for research in New England, New York and beyond. With the leading experts in the field, NEHGS staff can provide assistance and guidance for questions in most research areas. They can also be hired to conduct research on your family. Learn more today about researching African-American roots.



Sometimes the ages are wrong, perhaps due to a misunderstanding, or inattention on the part of a census keeper. I found a census record which gave the wrong age to my great-great grandmother, but was able to find other records which gave a different and more logical date (four years after her older brother was born, not the same year). Good luck!