Meticulous Sourcing

I'm of mixed opinions about the sourcing capabilities and overall reliability of submissions to Ancestry.com's Personal Tree. On one hand, an owner can link directly between entries in his/her tree and Ancestry.com sources that support the data in the tree, like census images, military records, or passenger ship manifests. On the other hand, unless I'm mistaken, the system only allows the owner to present one form of a name or date when various sources suggest alternate possibilities.

Let me just say that I am quite pleased with the Carr/King/Rodriguez Family Tree that I found at Ancestry.com But it has a few sourcing problems, which I will use to illustrate my issues with Personal Tree. I don't mean to pick on the poor tree. It's fine. I must also say that some of these links will be by subscription only. Sorry for any inconvenience. (That said, I encourage you to get a subscription to Ancestry.com.)

The tree in question has a listing for Edward Bane Carr born in 1855. If you opt to View details, the tree shows two quills which, when clicked, each indicate that the full name or exact date of birth is sourced to five different censuses. To my mind, this sourcing ought to indicate that the full name and exact dob were derived from these census records, which is not the case. In actuality, the data were found on Edward's gravestone. The census records were routinely less precise, showing a middle initial here and there, various ages, and maybe a month and year of birth.

Seems to me that Ancestry.com doesn't leave a tree owner any alternatives. Ancestry.com encourages the tree owner to pick the best option to display -- best form of a name, best date and place of birth or marriage or death -- and then connect all sourcing to each, whether or not the sources support the portrayed data or vary from what is presented in the tree.

There is no allowance for gray. The census might say Edward B. Carr, or Edward Karr, maybe even Edw.'d Carr, but only the gravestone says Edward Bane Carr. The gravestone might say Edward was born in 1861, but he appears in the 1860 census as five years old. Which date do you present in your tree? Only the one you agree with?

I realize that you can always check the original records by clicking the accompanying hyperlink, a fine feature offered by Ancestry.com, but I feel that a family tree should carry all the variations, right or wrong, and show sources for each. That is a more robust, less misleading method of sourcing. And it is leaves a more clear record for future researchers to see how decisions were made based on conflicting information.

The discovery of serious errors in an online family tree should destroy a researcher's confidence in the entire project. Even minor errors here and there ought to lower one's sense of a tree's reliability. At the very least, occasional errors should throw up red flags and raise questions about sources and methods.

To cite the Carr/King/Rodriguez Family Tree again, Edward's grandfather, John Carr, is shown as being born in Mercer County, Virginia in 1798 and dying in Mercer, Virginia (Loudoun County) in 1874. The place of birth is certainly wrong, while the place of death is very likely incorrect. John lived in Mercer County for many years, based on census records, and he was buried there, but the county wasn't formed until 1837, so he couldn't have been born there. (The troubling issue here is that even if his death certificate said he was born in Mercer County, a bit of research would have exposed that the county didn't even exist in 1798.)

The coincidence of John's dying in a place called Mercer, but in another state and far from where he lived and was buried, makes the place of death seem highly dubious. I suspect that the researcher had a research note saying that John died in Mercer but couldn't find a Mercer County in Virginia in 1874 -- the county became a part of West Virginia in 1863 -- so he/she erroneously picked a city named Mercer in Virginia as the place of death, even though it was far away from where John lived and was buried. While I can't say with certainty that John died in Mercer County, West Virginia, I'm not ready to adopt Loudoun County as his place of death, either. The obvious error with the place of birth causes me to doubt the place of death.

So, what to do? Toss out the baby with the bath? No. The Carr/King/Rodriguez Family Tree is a goldmine of data from an obscure graveyard. I simply advise caution with online trees, whether you are creating one as an owner or using someone else's for research purposes. To be generous, they can be fraught with minor errors and imprecise sourcing. Always sample the data for the accuracy of spelling and data, then decide how much you are willing to rely on it for your research needs.

While researching this piece, I came across the blog Genea-Musings, which also discusses personal trees at Ancestry.com. You might want to have a look.