How can I remove, or amend sources

Natalie Goodwin_2

These recent sources that have been added to my family, are so incorrect that they are offensive

Rimmer Household

Hi Natalie

We are so very sorry that this has happened to you. There is a way to detach incorrect sources.

There is a way you can detach the incorrect sources by following these instructions:

https://www.familysearch.org/help/helpcenter/article/how-do-i-detach-a-source-from-a-person-in-family-tree

Also there FamilySearch also encourages us to communicate to those who have contributed: See here:

https://www.familysearch.org/help/helpcenter/article/how-do-i-send-a-message-to-a-recent-contact-in-familysearch-messaging

We hope that you are able to correct this situation and wish you all the best

Miriam Rimmer

Natalie Goodwin_2

I had messaged the person who added the erroneous source offering to work with her to correct the information. There has not been a reply to date.

The errors are complicated in that it relates first off to my grandfather's death. He however married three times and had three different families children being half siblings. The source has resulted in two of the families having the wrong mother. It is possible the source was indexed or transcribed with no check on the whole family unit and may have been taken off a newspaper family notice. One family line has a spelling error in the surname, 8 siblings or half siblings with errors, no surname on 1 plus other errors. The whole thing is a mess. I can foresee future generations of these families not knowing a correct family line.

Julia Szent-Györgyi

If these wrong sources are what I think they are -- GenealogyBank obituaries -- then there are more places/ways for errors to arise than there are people mentioned in an obituary. First, there was the bereaved person communicating with the newspaper employee or funeral director: he or she may not have reported the facts straight, because the family lore was wrong, or he/she misremembered, or got confused. Then there's the possibility of misread handwriting and typos; some of these were likely caught in one or more proofreading passes, but at some point, everyone involved had seen the text once too often, so they were no longer seeing what was there, but what they expected to be there. (I've worked as a proofreader. It's just the way the human brain works.) After the obituary was published, if there was an egregious error, the paper may have issued a correction -- but that wouldn't do anything to the original publication.

Eventually, the newspaper was digitized (scanned or photographed). There are errors possible even in this process: things like page numbers or dates of publication can be entered wrong in the digital image's metadata. Then, there are two main ways for the digital images to be turned into machine-parseable (searchable) text: OCR (optical character recognition) and indexing by humans. Both of these are error-prone processes. OCR often has great trouble with names, because the software tries to fit them into its dictionary. OCR also has trouble with multi-column text, like what's found in newspapers. Human indexers also often have trouble with surnames that are unfamiliar to them, or with spelling variations of familiar names. (Again, it's the way the human brain works.)

The GenealogyBank obituaries go a step further: after the OCR process, they've been run through a parsing algorithm that tries to assemble the text into people and relationships. In other words, they've been machine-indexed. There are some obituaries where the parser got even the name of the deceased wrong, never mind the rest of the relationships. (English -- like all human language -- is too variable for easy machine processing.)

The surprising thing about the GenealogyBank obituary indexes is not that they're error-ridden. It's the fact that sometimes, they do actually get things mostly correct.

Gordon Collett

So taking into account all the potential errors in these obituaries Julia has explained, put on top of this someone who is not very familiar with the family and accepts the indexed version of the obituary and doesn't question how the source linker lines people up so just adds people and attaches people as Source Linker shows them and you can end up with a mess. It doesn't help that so many people in obituaries are listed with just their first names.

What I would advise you to do is to take the indexed sources for this obituary, compare these to the actual obituary (Can you see this there on the page for the indexed sources? I think there might be some restrictions on the GenealogyBank images. If you can't, I'd be happy to take a screen shot of it and post it here.), then move everyone into the right families, and detach and re-attach the entries for each person to the actual person. You can edit the title of each source to make it clear exactly who each one is. You can also add notes to explain anything in the index you can't correct.

This type of thing can easily happen with census records also. A user in Family Tree who does very good work there that I run across often because he is my wife's 8th cousin recently used the 1900 Norwegian census to create a duplicate of my wife's grandfather as the child of his uncle because the census listed the uncle as the head of household, her grandfather as a dependent child, and her grandfather's mother (sister to the uncle) as household help. A quick fix of family relationships and merging away the duplicate took care of this. It will take a bit longer for you to fix your problems because it sounds like the obituary has a lot of people in it.

Do make sure that the person who attached the obituary did not take a lot of people from the obituary and add them to Family Tree as deceased who are still living. That also tends to happen with these obituary sources and needs to be corrected promptly.

If you need any help getting these sources fixed, please post the Family Tree ID of your grandfather for specific suggestions and directions.