Using sources that may be dodgy

I started this train of thought in my last post, re: ACGS Yearbooks. Here’s a little more.

The perpetual problem in family history research is that so many of the available narratives are unreliable–sometimes intentionally so, more often because the authors and compilers haven’t exercised a very high standard of evidence analysis. Often no distinction is made between what has been proven and what is believed, and enthusiasm counts for more than accuracy. It hardly matters whether you’re looking at your grandmother’s genealogical notes, a magazine article about a local historical site, or the efforts of an amateur historian–rarely will such sources prove entirely sound.

People make mistakes. (About half the family trees I’ve seen on Ancestry.com give the wrong name for the father of John S. Magruder. His father was Nathaniel; his uncle was Nathan. Read Nathaniel’s will if you don’t believe me.)

People hide secrets. (Is the birthdate in a family Bible accurate? Or does it hide an out-of-wedlock conception?)

People find one record and stop looking. (The wife’s not in the will, therefore the wife was deceased.)

People assume public records are accurate. (Try comparing ages for one person across 2 or 3 19th-century census records!)

People look for evidence to confirm what they already believe, or want to be true, ignoring gaps and contradictions. (Among those associated with a particular family, I’m looking for African-Americans named Basil Mullin, so I want to believe a slave named Basil, with no surname given, must be another Mullin.)

People like heroes, villains, and good stories. (We’d like it if Roderick McGregor had freed all his slaves in his will. Sadly, he didn’t.)

That’s the reality, so what should we do? The question is not whether to use sources that may be unreliable–often they are our only starting point. The question is how to use them. And the answer is twofold: cautiously and proactively.

Here’s an example from my own research.

In the 1917 Year Book of the American Clan Gregor Society, Caleb Clarke Magruder, Jr., published a 26-page article titled “Nathaniel Magruder of ‘Dunblane’,” in which he summarized and quoted from the wills and estate records of Nathaniel Magruder (d.1786) and a large number of his descendants. C.C. Magruder was one of the better researchers among the early ACGS writers, and I am interested in this particular family, so I am quite happy to have this article. It has pointed the way toward several interesting lines of inquiry, clarified some family relationships about which I was hazy, and given me a large panorama of the family’s property-transference, slave-owning, and other matters.

But that’s not the same as accepting C.C. Magruder as the final word. Here’s where caution and proactive research come in.

I had already read a few of the wills he quotes, so I spotted one large error right away: he says Roderick McGregor did free all his slaves, when in fact he freed only one family. This is very clear in the will and in the estate inventories. (I’ll refrain from theorizing how this mistake could be made, but there it is; and others have quoted it as authoritative.) This alerted me to treat the article with caution, but you shouldn’t need any special red flags. You should always follow up with your own legwork. (See Will of Roderick McGregor and related pages, for images and transcriptions.)

I discovered another error quite by chance… or, rather, by casting a wide net. On subjects that interest me, I read every published source I can find, and I often start with the index. That’s how I discovered–in Letitia Woods  Brown’s Free Negroes in the District of Columbia 1790-1846–that one Elizabeth Magruder made her will in the District in 1827, including detailed provisions regarding her slaves. When I first found that information, I didn’t know who this particular Elizabeth Magruder was. C.C. Magruder’s article clued me in that she was Nathaniel Magruder’s unmarried daughter (and John S. Magruder’s sister). Why Elizabeth made her will in Washington, D.C., despite the fact that her legal and apparent residence was the family plantation, Dunblane, in Prince George’s County, may be explained by the will itself: her executor and all her heirs are found in the family of one of her sisters, a D.C. resident. Another lesson here: check other jurisdictions, wherever your target has relations, business interests, or property. (See Will of Elizabeth Magruder and related pages, for the rest of the story.)

The last issue I’ll talk about took a while to pin down. Because Roderick McGregor’s wife, Ann, is not mentioned  in his will, C.C. Magruder states that  she predeceased him. It’s a perfectly reasonable assumption in a time when divorce was so extremely rare. But once all the clues are assembled, it seems very probable that Roderick and his wife Ann were legally separated or divorced at the time of Roderick’s death. I picked up this trail in the 1850 census, which shows Roderick living alone and Ann living with a woman who is likely her mother. It took me a while to feel  confident that this Ann McGregor was Roderick’s wife–I think I first drew the connection from the genealogy in Sue Emerson’s Magruders in America. Sue doesn’t give sources, so at that stage I had to call it a working hypothesis and keep digging. Eventually, I found Ann McGregor’s will, from 1871, which confirmed her identity. My last confirmation came from details in an 1850s land record. Again, the lesson is: cast a wide net. Search every category of records you can reasonably reach, for any and all information about your target. (See Interrelations page in the Magruder-McGregor pages, for details of this story.)

One way to contextualize any article you are reading is to consider what sources were and were not consulted. C.C. Magruder’s article on Nathaniel was based primarily on a survey of wills in Prince George’s County and the District of Columbia. You immediately know that you can and should look for other categories of records to confirm and expand the information.

And then, since we are people, too, and prey to all the mistakes I listed at the top of this post… distinguish carefully between what you can prove and what you believe to be true, and then be thorough and transparent in documenting and revealing your sources, allowing others to judge how firm the basis for that belief might be. It is useful to read each other’s theories; it is the opposite of useful to read a theory presented as if it were fact.

No time to do all this? Fine, just practice that very last piece of advice, letting the rest of us know that (so far) your only source is an uncorroborated article or family tree, and more work remains to be done.

In Evidence Explained, in the chapter on “Fundamentals of Evidence Analysis,” Barbara Shown Mills offers 6 descending “Levels of Confidence” that a conscientious researcher can communicate to readers.

Certainly: no reasonable doubt, sound and adequate evidence

Probably: more likely than not, good evidence but not quite certain

Possibly: some evidence, but far from proved

Likely: enough evidence that the odds are in its favor

Apparently: you’ve formed an impression but not tested it

Perhaps: you think an idea is plausible, but have not tested it

This is a useful vocabulary, whether or not we all conform to Mills’ exact  distinctions. It allows us to share what we’ve found without making too-large claims about what our research “proves.”

Leave a Comment

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s