Story image
Where Did That Information Come From?

Many of us have been told that official documents are the “be all and end all” when it comes to information. But is this really true?


Anyone who has started to dig into their family trees will tell you the information on forms is only as good as:

  1. The person who filled it out (census taker, coroner)
  2. The person who provided the information (family member, neighbor)
  3. The person who transcribed the information (data entry worker at Ancestry)

I hadn’t been researching my family for very long when I noticed little things like names being misspelled either on the census itself or in the index provided by Ancestry. My great-grandmother “Tylene” was listed as “Tilene” on the 1940 census and the entire family’s last name was spelled “Louis” instead of “Lewis” on the 1920 census. Little things like that you expect to come across. But this last one takes the cake.

Mary Jo Clubb Williams is my maternal third great-grandmother. I started researching her after coming across her name on someone else’s family tree, as she was not listed on her son Pleasant’s death certificate. In fact, Pleasant’s father Solomon was listed as “Sol,” Sol’s birthplace was listed as “unknown,” mother unknown, mother’s birthplace unknown. Super helpful, right? All of that information was provided by an informant, Mrs. Nannie Gregory. There are Gregorys in my family, but as far as I can tell, she’s not one of them.

Death certificate for Mary Jo Clubb Williams

But back to Mary Jo Clubb Williams. I found her death certificate to be even less helpful than Pleasant’s. I like to think that the informant didn’t sign the death certificate because they were ashamed at the information they didn’t know.

So, back to other sources. Let’s try comparing the info to what’s on other records. They were just as helpful:

  • Death certificate – state of birth, Illinois
  • 1910 census – state of birth, Indiana
  • 1900 census – state of birth, Illinois
  • 1880 census – state of birth, Indiana
  • 1870 census – state of birth, Tennessee
  • 1860 census – state of birth, South Carolina
  • 1850 census – state of birth, South Carolina
  • Son Peter’s birth record – state of birth, Illinois

As far as I can tell, Mary Jo was born in South Carolina. She and her family were there until at least the 1860 census. She married Peter Solomon Williams in 1862 in Marshall County, Kentucky. I have yet to find out when she moved to Kentucky and why, but that’s her on the South Carolina census. Why did her state of birth jump all over the place on the census? Good question! It could be that various neighbors answered questions from the census takers and didn’t know. It could be that she and Solomon weren’t home and the kids tried to answer the questions. Heck, it could be that she didn’t remember what state she was born in. On the 1870 census, she’s listed as not being able to read or write, but at this point, I’m taking everything with a grain of salt.

I’ve also not found anything to back up the claim that Mary’s parents were from Ireland. Thus far, I’ve found data showing her father, Joseph, as bring born in either South Carolina or Germany. Of course, the censuses have listed all of the same states as Mary, with the addition of Kentucky.

It’s enough to make one’s head spin, if not laugh at the absurdity of it all.

However the information is recorded, it can be listed as primary, secondary and indeterminable information:

  • Primary Information: Information is primary if it was made orally or in writing (or even pictorially) by someone in a position to know firsthand (such as an eyewitness or a participant) and recorded in a timely manner while memory is fresh. The informant may have provided faulty information, but nonetheless the information is considered primary information. “Primary” does not ensure accuracy.
  • Secondary Information: When we know the informant but know that the informant was not furnishing an eyewitness or first-hand account, it is secondary information. The neighbor Patty, for example, says Mary next door had her baby for Patty heard an infant crying all night. Patty’s information is secondary; she did not witness the birth nor was she a participant in the event. It may be correct but it is nonetheless based on secondary information.
  • Indeterminable Information: When we can’t identify who created the record or furnished the information, that is, when the informant is unknown, we consider the information as indeterminable. Put another way, if we don’t know who the informant was on a census record, we cannot judge whether the information given was primary or secondary. It is therefore considered indeterminable information. We can seek other records which might product secondary information.
  • Primary information generally carries more weight than secondary or indeterminable information.

It’s important to remember that people are, well, human. Census takers didn’t always ask how to spell names or make the trek all the way down to the house at the end of the street, instead asking the neighbors for information. Mary Jo was a widow, and it could be that she passed away while family was gone, leaving a neighbor to be the informant on the death certificate. Sometimes, there’s a plethora of documents to help you discern fact from fiction. Other times, you have to rely on one or two because nothing seems to agree. It’s our job as researchers to take the time to look at all of the documentation and figure it out!


  1. Kentucky, Death Records, 1852-1964 [database on-line]. Lehi, UT, USA: Operations Inc, 2007.
  2. Rose, C. (2014). Genealogical proof standard: building a solid case. San Jose, CA: CR Publications.