by Harold Henderson | Jan 12, 2012
As genealogists, we have all spent time learning which records to trust and which not to trust. We might start out learning that a photocopy of an original death certificate is good, whereas an entry in an on-line database transcribed from a poorly typed 70-year-old transcription of the death certificate is not so good.
But that's not quite right. Later we learn that the quality of the information matters more than the record itself. Thus an original death certificate is good for information about the death, but not so good for information about the birthplace of the parents of the deceased.
But that's not quite right either. After a while we realize that it also makes a difference who provided the information on the death certificate - the widow? a child of the deceased? a stranger?
And still we are not quite right. Eventually we learn the underlying fact of research life: no record and no class of records deserves our trust.
One of the first articles I read in the National Genealogical Society Quarterly was Tom Jones's account of Calvin Snell of Lake County, Ohio. In his 1849 will Calvin named three children. Good information, right? As Jones wrote, "Wills are a bedrock source in genealogy, much sought and highly regarded." He then marshalled evidence proving that two of the named children were not Calvin's and the third was in doubt. Worse yet, Calvin's will omitted about a dozen of his own children.
My own experience is less dramatic but has the same moral. I once encountered a research subject whose marriage record named one set of parents and whose application for government benefits named another. Both of those sources would rank high on any generic scale of trustworthiness, but at least one of them was dead wrong.
Even the best kind of record can be wrong, even if it contains the best kind of information from an informant who should have known the truth. This ties in to another general fact about genealogy -- it's not about averages, it's about the particular case. Exceptions are our life.
This may seem like a strange learning process, where we learn something and then keep finding out that we did not quite get it right. Is this evidence that genealogy is futile? Or that we are all a bit slow?
No, the learning process is like climbing a spiral staircase. It is not that we learn wrong things and have to unlearn them, but our previous knowledge is refined. As we make the next turn up the staircase, we can look down and see our former vantage point and its partial truth within a larger perspective.
Genealogy is not the only discipline where such spiral staircases happen. In a similar but more technical way, Einstein's relativistic view of the physical world included and explained Newton's earlier classical view. The first lesson is unlikely to be the last, but what was good in that first lesson is conserved in the later corrections.
In practical genealogical work, four things follow:
1. Seek independent backup for every clue. The crusty old Chicago City News Bureau adage applies: If your mother says she loves you, check it out. Even partial confirmation is better than nothing. Relying on a single source - or, worse, a single reading of a single word in a single source - for a critical conclusion is like building a sprawling house on a narrow foundation. It hurts just to look at it. The number of sources is not as important as their independence and their quality, evaluated individually.
And what if backup is not forthcoming and inconsistencies start to swarm like mosquitoes on a summer evening? Don't brush them off. Gather as much evidence as you can on the subject, and be prepared to use indirect evidence as well. Again, the Calvin Snell article is a good guide even if your case isn't so extreme.
2. Don't be a selective skeptic. I once gave a talk about genealogy blogs, and afterwards an audience member complained that you could never believe anything on a blog. When I spoke on property records, a real-estate veteran volunteered inside information on how many errors are made in the creation of those records as well. Their skepticism isn't misplaced -- it just wasn't broad enough. Those sources' ability to err does not make them unique. It makes them typical.
3. Don't be an evidence snob. It's natural, but unwise, to look down on our earlier views and practices with lofty contempt. I used to rely heavily on on-line family trees; now I rarely consult them under any circumstances. Both approaches are wrong. If any source can err, any source can contain a nugget of truth. Your long-sought clue may appear in disreputable circumstances.
My wife's great-great grandmother Jennie Cochran was a North-Carolina-born mystery who lived and died in Pittsburgh. Family members sought her origins off and on for decades in the pre-Internet era, even hiring a professional genealogist, who made no progress. An unsourced family tree on Rootsweb (now apparently taken down) connected her back to her birth family. Once we knew where to look, the puzzle pieces all fit together and made a sound conclusion.
Despised pseudo-sources can be valuable if you consider them as clues rather than gospel. Remember that there are two kinds of evidence: evidence that leads us on or points us in a direction, and evidence that goes into the final proof argument. Of course some evidence will do both. But don't shut yourself off from leads, even when you know the proof will have to come from somewhere else. The analogy to criminal investigations is obvious.
In another case, with limited time and with only derivative sources available, I wanted to learn something about the parents of a Pennsylvania woman variously named as Louisa, Lucetta, Luisette, Lisette, and Lecetta. Family stories offered two possibilities for her surname: it was said to be either Battren or Maidenform. (Sometimes genealogists are better off without a sense of humor.) A published reading of her Emerich in-laws' church's burial ground included an immigrant called Nicolaus Van Medenfort. That led to an unsourced mug book that profiled one of the grandsons of "Nicholas Von Madenford." The profile named Nicholas's children, including "Lucetta." In place of bizarre conjectures we now had a plausible and testable hypothesis (not a proven conclusion!). The rest of that research story has yet to be told, but if I had disdained those non-original sources, it would have had no beginning.
4. Don't stop learning. If you're scared of heights, you may want to switch metaphors right now. Genealogy learning does differ from an actual spiral staircase in one interesting way: I see no reason to think that it has a top.
Start your free trial today to learn more about your ancestors using our powerful and intuitive search. Cancel any time, no strings attached.