Record placename does not match actual document
Reuben C Harris, "Tennessee, County Marriages, 1790-1950" • FamilySearch
The record title is correct, but the Event Place seems to have nothing to do with it, and is incorrect.
Answers
-
Thank you for letting us know about the inconsistency. We have sent this to be reviewed.
0 -
This continues to happen all over the place. Is the intent for us to continue reporting every time we run into a discrepancy like this (that will take lots of time because I run into it about every 15 minutes) or does someone in FamilySearch engineering have the task to audit all these source records and correct them?
3 -
@EarlMarshall if it isnt documented it didnt happen - nobody seems to have picked up on this, so I shall venture what I think I know on the basis that if I'm wrong, people will rush to correct me and we'll find out the truth. 😉
As you may well know, the classic reason for a mess-up on the Event Place is the failure of the background auto-standardisation routine. Whether that's the case here, I don't know - we used to be able to see something called "Event Place (Original)" which was basically the placename as indexed. Between "Event Place (Original)" and the "Meta Data" for that part of the collection, the background auto-standardisation routine was supposed to come up with the standardised "Event Place". However, this didn't always work - often when "Event Place (Original)" was duplicated elsewhere in the world.
We were asked to report apparent background auto-standardisation routine issues and then the guy who was collecting the reports for the engineers (a heroic task I would add!) told us that the engineers had said that they had enough examples and didn't need any more reports, thank you.
My personal belief is that Engineering will not correct the source records themselves - instead, they will fix the algorithm and wait for the next run of the background auto-standardisation routine to correct the data in the source records.
So that's my understanding - Engineering don't want any more reports and everything will eventually get corrected automatically. Or, at least it will get closer to the truth.
All comments gratefully accepted and no, I have no idea how your index ended up as "Essex, Essex, New York, United States". Seeing "Event Place (Original)" might have helped but, as I say…
0 -
@Adrian Bruce1 I'm not sure there was an original Event Place here. The document never once lists a location, so there probably wasn't one indexed, either.
1 -
@BraydenGraves - strictly speaking, I suspect that you're right.
So far as I can see looking at the Ancestry images that seem to match, I don't see Event Places on the individual records. Or at least, not on the pages I looked at. Ancestry says it was Wilson County, Tennessee so that may be what is in the FamilySearch Metadata for this bit of the collection. So blank Event-Place-Original plus "Wilson, Tennessee" Metadata, somehow turns into "Essex, Essex, New York, United States". (My understanding is that it's nearly always Event-Place-Original plus Metadata as the input to the production of standardised Event-Place.)
How does this precise issue happen? Metaphorical shrug of shoulders at this point…
That collection appears to have lots of "Essex, Essex" places, by the way…
0 -
I'm pretty sure the index always had a place field; it just never came from an indexer, but was added in pre- or post-processing, based on the cataloging information for the film.
Authors: Wilson County (Tennessee). County Court Clerk (Main Author)
Tennessee. County Court (Wilson County) (Repository)How or why the automangler algorithm got from "Wilson County (Tennessee)" (or whatever formatting the pre- or post-processing applied to that) to "Essex" anywhere, nevermind in New York, is not something I can fathom.
3
