Monday, October 22, 2012

Dueling Birth Dates: Is Your Database the Solution or the Problem?

Yesterday there was a thought-provoking discussion on LinkedIn's "Genealogical and Historical Research" group, based on a real-life genealogy question, "How do you decide how to enter an approximate birth year when you don't have the birth record and other sources vary?"

The question, and some of the answers, assume that we need to decide how to enter a birth year given varying evidence. Is this true?

First of all, how many problems of relationship or identity turn on knowing an exact birth date? Some do, for sure, but it's not a given.

Second, if not, why this urgency about deciding? Isn't it a sign that the tail is vigorously wagging the dog? Old-timers are used to filling out pedigrees and family group sheets; an increasing majority of genealogists are wedded to entering data into their database programs from Personal Ancestry File or Family Tree Maker on up.

Paper or electronic, I've used many of these forms and database programs; for years I spent much of my genealogy time breaking down the information I had into small enough components to enter each one into the program, and then tweaking it so that the outputs would be understandable. Some were better than others with problems of this sort.

But is a smoothly running database the reason why we started researching our families? I think not. The database is a tool, and doing our genealogy so that it will fit into the tool is not very different from a carpenter trying to saw a board using a hammer . . . because that's his favorite tool, and saws are too much trouble.

The real genealogical question here is how we deal with conflicting evidence of any kind. The right way doesn't have much to do with any form or database that I'm acquainted with (and if my acquaintance isn't wide enough, let me know). It has to do with listing out the different birth dates and where they came from, and evaluating each of those sources for evidence of reliability. Do we have the original source? Do other entries show some bias or impairment in the record creator? Did the informant have an incentive to deceive? And so on. For a checklist of ways to approach this task, read the last seven pages of the first chapter of Elizabeth Shown Mills's Evidence Explained, or visit the same-name web site for any recent discussions (such as this one on a brick-wall problem).

If our family or our problem dictates that we come to a best possible conclusion, a table or other format may help focus our thoughts. But in the end there is no substitute for the fifth prong of the Genealogical Proof Standard: a "soundly reasoned, coherently written conclusion." No rule of thumb (such as relying cautiously on the earliest census record) is a good substitute for a well-documented, clearly reasoned, explicit statement explaining why our conclusion is the best, based on the weight of all the available evidence. Accept no substitute.

Harold Henderson, "Dueling Birth Dates: Is Your Database the Solution or the Problem?," Midwestern Microhistory: A Genealogy Blog, posted 22 October 2012 ( : accessed [access date]). [Please feel free to link to the specific post if you prefer.]


Yvette Hoitink said...

You voice exactly what has been going through my mind lately too. It's funny how tools shape our thinking, while it should be the other way around. I went from a genealogy program that only allowed one value for each event to a genealogy program that allowed multiple values for each event. That felt like a liberation :-)

Harold said...

Thanks, Yvette. There's a whole 'nother post to be written (by someone else!) comparing the current offerings on this basis.

Michael Hait said...

This conversation also ties in with a few of my previous blog posts concerning genealogical software--not to mention the article I wrote in the current issue of the APG Quarterly: "Why I Do Not Use Genealogy Database Software."

Harold said...

Doggone, I knew I was missing a link! Here's one:

Anonymous said...

I create a "date range" linked to the various sources that the dates came from. In the 1841 Scottish Census, everyone over the age of 15 was asked to round down their age to the nearest 5 years (for statistical purposes) so there's a problem to start with...and not all did. Jo :-)

Anonymous said...

I agree with imagespast I also encounter the same issue with rounding up ages for statistical purposes, I hope this issues will be solved soon.