Friday, November 15, 2013

Darned Overly Helpful Indexers

Records say the darnedest things

We depend upon records to reveal the “truth” about our pasts.

Yet sometimes records have anomalies.
Some are amusing or humorous.
Some are interesting or weird.
Some are peculiar or suspicious.
Some are infuriating, even downright laughable.

Yes, “Records Say the Darnedest Things.”

Records Say the Darnedest Things: Darned Overly Helpful Indexers

It was a usual practice back in the days of indexing the U.S. censuses to use other census years to help decipher hard to read text. I admit; I did it too. But I always wondered if the extra work would go for naught, knowing that an arbitrator might not be as diligent. I always wished I had a way to enter a note that was sent to the arbitrator, explaining how I had deciphered the entry and asking that they give extra weight to my efforts.

I recently came across an entry in the 1870 census wherein the indexer had found a creative way to send such a note to the arbitrator: The indexer included the note in the name field:


The indexer had found Bricius spelled differently in the 1860 census, so had made a note of it as part of her name. Now Bricius’s name is “Bricius (Briceus-186o Cs)”. Likewise with “Reymond (Raymond-186o Cs)”.

This entry is doubly disturbing. Why in the world did the arbitrator not fix this?

And how did FamilySearch allow it? It would be a simple computerized check for FamilySearch to detect this aberration. It could have automatically been flagged for intervention. Perhaps it could have been sent back to a FamilySearch missionary or another arbitrator. Two arbitrators would need to agree that the entry was an acceptable exception to the rule.

If anyone at FamilySearch is listening, perhaps these suggestions could be incorporated into next year’s new version of FamilySearch indexing.

Darned those overly helpful indexers!


  1. If I had seen this as an arbitrator, I would understand that the person obviously wanted me to cross reference this with an earlier document, but then to do what? To knowingly put something that wasn't written on the record and expect someone else to understand why I was doing it? To do so lacks transparency, and could lead to even further confusion and error. In a case like Reymond/Raymond, we might be able to say safely that the E spelling is incorrect. But in the case of Bricius/Briceus, I would have no idea which is correct.

    The current systems on most websites allow corrections of the records after they are posted online. Most of the time, the individuals who are doing this are researchers from the families in questions, who know how to answer these questions better than I can, as a stranger with no exposure to a family's history. Although we know these errors exist, taking care of them at an arbitrator level, I feel, is irresponsible. Especially since tracking spelling variations is important data. Tracing how names change and the variants that exist only helps us to do better research, and become a better researcher.

  2. The example you give isn't a case of deciphering hard-to-read text. The indexer apparently could read the text in the 1870 census and so recorded it as Bricius. Since indexers are supposed to record the actual record as written by the census taker, whether the name was spelled differently in 1860 is irrelevant. Furthermore any search engine worth its salt would find Briceus if Bricius were searched for and vice versa, so the extra info entered by the indexer is really unnecessary from any standpoint. So, yes, darned (I would say) incompetent indexer, but also darned incompetent arbitrator. And what a bummer for the other indexer (assuming she did it correctly) who got dinged for indexing it right.

    Now if the indexer couldn't tell from the 1870 census whether the fifth letter was "i" or "e" and looked at the 1860 census to see whether it was clear there, that seems entirely appropriate to me. But in that case his entry should have been something like "Briceus (e clr in 1860 Cs)", assuming the arbitrator would remove the material in parentheses, a dangerous assumption. As an indexer, I have often wished for some way to explain a particular choice, but I don't think this example is a good way to do it.

    Regarding FamilySearch and computerized checks...there are many things they could work into their software with very little effort that would prevent items such as the one in your example. Checking birth dates of children, for example, to see that they come after the birth dates of their parents. It just kills me to see examples in Family Tree of children who, according to their birth dates, are older than their parents. It would be so easy for FamilySearch of put a check for this and other similar issues in the Family Tree software, that I just can't understand why they don't do it.

  3. I spent a long time looking for records for my deceased brother under "Robert", now most arbitrators would agree on the spelling, turns out it was actually Robart. Even my mother changed the spelling to fit in with her new life, with her new husband "Robert". I think possible alternative spellings should be included.

  4. Oh, and I found two different family name changes after WW 1, from the German spelling to the English spelling - for obvious historical reasons.