The UNKNOWN question
See https://www.familysearch.org/tree/person/sources/98TN-MPL
3rd source
An extracted IGI record
In the Extraction Program, the forms had blanks for parents, grandparents; when there was no information on that, the IGI system put "UNKNOWN", which is now goofing up the Source Linker. It's trying to match UNKNOWN against names of people. UNKNOWN is not a person, it's a blank in the form. I keep seeing this, "Unfinished Attachments • Dismiss" and I hit Dismiss - after investigating. You will see this 5 times on this source page alone. Will we have to hit Dismiss 100 million times, or could somebody change the Source Linker to just make it not do that?
Answers
-
If you know who that Unknown person is suppose to be, such as the mother, you can still go ahead a link that unknown to the proper individual that it should have been.
1 -
@Gordon Collett Why would you attach a source to a person (the mother to use your example) when neither she, nor anyone else, is named as the mother in the source?
If I understand it right, it’s not that the mother was recorded as “Unknown” on the original, in which case I’d agree with you, but that the mother wasn’t recorded at all. Neither was the father, nor siblings, nor (to take it to extremes) the man next door. “Unknown” was a fabrication of the indexer or extraction program.
0 -
Because the linking from sources from Family Tree is a two way street. If records are attached, then you can either jump from Family Tree to the record via the source page or, you can click on the pedigree icon in the historical records database and jump to Family Tree.
So if I have a set of records supporting the birth and parents of a child and one of the records does not record the mother name but only has "Unknown" while all of the other records show who that mother is, I would consider that sufficient evidence for attaching Unknown to the mother in order to completely attach that source record and take one step towards the ultimate (but probably unobtainable goal) of having every record in FamilySearch's historical records database attached to the correct person in Family Tree, even, as far as possible, this page of the historical records database:
0 -
Hi Gordon, Sorry, It looks like I have not made clear what, I think, is wrong.
I agree with you in the examples you show. I have seen plenty like them myself. The original (death certificate, or register, or whatever) names the individual and has a space for the mother and a space for the father and one, or both, of them are recorded as "Unknown". I agree that "Unknown" represents, and should be attached to the mother, or father as appropriate.
The example quoted by Woody is different. In this case the original is a birth or baptism (certificate or register or whatever) where the only people mentioned were the child, the father and the mother, as you might expect. The index was created (apparently) using a pro-forma with spaces for, "name", "mother's name" and "father's name". The problem is the same pro-forma has been used when creating the index entry for each person named on the original. Everything works fine when the form is filled in for the child. But... When you use this form for the father he is now in the "name" field and the "mother's name" field has been filled with "Unknown" because his mother, the child's grandmother, is not named on the original. The same is true for his (the father's) father, and the mother's parents. "Unknown"s have been created for people who were not just not named on the original, but were never intended to be named on the original.
@Woody Brison Can you confirm that I have understood the problem?
@Gordon Collett Have I done a better job this time?
0 -
Interesting. In the indexing batch C05829-1, looking at the first five pages, every single page looks like this:
I wonder what all these Unknowns do to the hinting routines. I hope they are just ignored. As far as what to do, I'm not sure there is any harm in attaching these unknowns to the correct person, not as a source for them but as a link to the source about the grandchild. Again that would depend on what effect that has on the hinting routine.
In this situation, I would say the source linker does not need to be modified, but the error in creating the index needs fixing. By any chance has anyone looked at the actual microfilm? Do the actual record books have a spot for grandparents that was just never used in that parish? Or was the original indexing project set up wrong? Or is this another instance of a post-processing routine that went awry?
0 -
Every person (so far) had a mother and a father. So by induction every father had a father etc. We could take one of these records like what I cited and create an entire tree to an infinite distance of UNKNOWNS and attach them all to the shared tree. Exactly what would that accomplish? The way to understand this is with information theory. Information proceeds from an event outward via media to observers. If there's no media, there's no information flowing.
ColinCameron it appears to me you have understood.
Gordon Collett you are very helpful. At some point tho, we need to brush aside regular reserve, and storm the Bastille with torches and pitchforks, and get the entrenched bureaucracy to admit that the Source Linker is spinning eccentrically, out of control. It needs a thorough redesign going back to the charter.
0 -
Who knows, maybe a redesign is in the works. The source linker is pretty old and does not yet incorporate the new style sheets FamilySearch is moving to. Can only hope it is the next page to be updated now that the sources, source box, search pages, and others are done.
Unfortunately, the only time people start with the torches and pitchforks is when something is updated.
Rather than forming a mob, I think this is a case where someone needs to to set up a Go Fund Me page for FamilySearch to be able to hire a thousand programmers and researchers to fix the indexes. In this case to get rid of all these grandparents. In others to fix all the incorrect Event Place names incorrectly derived from incorrect Original Event Place names. I know of at least one batch where the original place name wasn't bad even though it was not really correct, but now it shows that all the children in the parish where christened in the cemetery.
0 -
It looks like somebody's been working on it. Not only does it still claim, under the Sources tab, where a source "has" UNKNOWN 4 times, that there are Unfinished Attachments; now it Suggests such sources on the Details tab under Research Help.
0 -
Hoping that whoever might rework the Attach Source function, here's another little problem:
Look at the Sources tab of any Dossier. Just about every last source listed has a Title like "New England Baptisms 1620-1933", or "Genealogical Records of Europe, 2000 BC to 2000 AD" or "Wales Parish Records, Surnames Jones, Hughes, Williams".
It's nice to know where the source can be found, but it would be nicer to know what it is, like "London Christening of John, son of Henry and Jane Wilkins." The date need not be part of the title since it's immediately to the left.
You see, in the context of a Sources tab, I want to know what that record is. At present, to find out, I have to open it, scroll down, oops scrolled too far, scroll up, figure out who's being born or married or what; then scroll back up, close the tab; then as I look around, now lessee, that was the marriage of John & Jane, now which one was it again? Open it up again, go thru all that once more, and hope I can remember why in Conniption I even wanted to know.
I need to see the data in tab form, all presented neatly so I can quickly navigate. So I can get these people figured out and arranged correctly. So they will stop invading my sleep at night.
0 -
@Woody Brison, I agree that the auto-generated titles are useless -- which is why I frequently rewrite them. I also never accept the auto-suggested titles when attaching unindexed images as sources. Now, granted, I can't seem to stick to any sort of consistent format (or language!) in my rewritten titles, but I try to include enough information to know what event was recorded where, without needing to look at the details.
But perhaps this topic needs a thread of its own: we've drifted pretty far from the problem discussed upthread.
0