To copy or not to copy: Escaping the telephone game in online research

Research Tips   •   October 21, 2019

Copyright, 2019, Allison Peacock

Creative Commons, BY-ND, attribution required.

Recently I received a text from a client who asked me a question to help him make sense of something he was seeing online in his own genealogical research. As I answered him, I was immediately reminded me of a childhood game I used to play a lot at birthday parties as a kid growing up in the 1960’s and 70s.

Remember the childhood game of Telephone?

A room full of children lines up and the first person whispers a message into the ear of the second person in the line. Then, the second person repeats the message to the third person, and so on. At the end of the line the message received is announced to the whole room. Do you ever remember the message surviving accurately? I don’t either! It was usually a hoot.

I thought I’d share my answer to his question about the metaphorical game of Genealogy Telephone played out every day on family tree research sites. It includes some common sense genealogy research tips on avoiding the spread of inaccurate facts online.

Q: I have put together my paternal lineage all the way back to the 1100’s, using public family trees found online. But some of the links seem unlikely—such as when it is reported that Frank A. was 4 years old when his child so-and-so was born. The genealogists in question are probably British [he was researching in Wales and England] and take this stuff very seriously, so I’m thinking they are passing on errors in recording or reading of the actual record. Thoughts?

A. Yes! I have tons. I see this every day.

It’s easy to use discovered family trees online as a lead in your own research; however, I don’t suggest ever copying over as fact results that do not have corroborating primary source material attached.

A primary source is a historical record created during the time period you are researching. This includes birth certificates, marriage records, death records, newspaper accounts, and first hand narratives written during someone’s life. Responsible researchers online attach these kinds of records to each individual to back up the facts in their profiles.

It can be a nightmare to undo a nest of erroneously attached family members when three generations later in your search you realize you’re on the wrong trail. Yet there are alternatives to completely ignoring unsourced trees.

The first thing to remember is that when actively researching unproven ancestors, making your tree private will ensure that unproven work isn’t passed on to casual researchers. Facts online spread like wildfire whether proven or not. Many casual researchers are what professional genealogists call “name collectors.” Give them a name to fill in a blank in their tree and they move on.

It’s human nature to want shortcuts, especially in the immediate gratification game of online interactions.

What I do when I see something interesting that I think might be a lead is this:

First, I make a note about the online hypothesis in the last known sure person’s profile in my own work, including a link to the tree or hypothesized family in question.

Next, I add the hypothesized related ancestor, then I immediately delete the relationship connection between the unproven parties. This ensures that I have the individuals recorded that I want to investigate further. And yet the unproven families or individuals are “floating” in the tree I’m working on and not attached as a proven relationship between individuals.

After doing more extensive research and proving or disproving the relationship, floating families can easily be connected later – or even deleted if proven unrelated with further research.

When I first run across a new family hypothesis connection I often keep an eye out for what someone else speculated in their tree by the details they included without sources. This sometimes helps me determine right away the accuracy of the tree I’m considering.

For instance, if someone speculated that a father was born on a certain date and it’s really specific, rather than “About 1800” look for the record with the name and that specific date on it. Then look and see if it verifies anything, or if it’s just a lead. Or better, is it someone you can rule out completely and move on from?

Sometimes these erroneous associations are obviously based on suggestions by the website’s hints algorithm. For instance, I was recently working on a case for a client who told me his great great grandmother was born in Tennessee although she lived most of her life in Missouri. I saw Tennessee as her birthplace all over the internet on research sites. As soon as I began looking for primary source materials for her in order to build out my own tree for this client I saw the problem.

My client’s ancestor was born Mary Farmer and she married a Mr. Pritchett, so for most of her life she was Mary Pritchett. However, in 1850, when Mary would have been about 52 years old, a woman named Mary Farmer was enumerated as a widow in Missouri with several children. This woman told the census taker she was born in Tennessee. This birthplace got attached to dozens of trees in spite of the fact that in the same census, my client’s Mary Pritchett was documented with her husband, and three youngest children still at home right in the county they’d lived in for decades.

Several obvious primary sources told me our Mary was born about 1798 in the Carolinas. I had to find them on my own rather than follow system “hints” but they were easily located.

Why the error?

Most people don’t read deeply enough. A website’s hint system throws up this widowed, middle aged Mary Farmer as a hint and we’re off to the races. Never mind that our Mary hasn’t been Mary Farmer for decades. I found 80% of the trees online had this place of birth and many had the wrong census record attached to her. (This is a great example of how attached sources do not necessarily an accurate tree make!)

More surprisingly, an equal number of trees had her born in Tennessee after copying someone else’s tree, yet the correct census record for her. This record affirmed North Carolina as her birthplace.

With these kind of errors based on system hints, you can sometimes see right away that an hypothesized fact or person is incorrect. This way you can quickly write off either the wrong facts, or the whole candidate, saving yourself a lot of time.

Unfortunately, when this happens, you can assume much of the fruit from the same tree is bad. But don’t despair. With the floating tree method above to help you evaluate these hints, you’ll have plenty of leads to explore.

And if you’re a die hard genealogy researcher, this is enough to keep you happily researching for proven facts and unknown family members until the wee hours of the morning with the rest of us!