I need help untangling a family tree!
So for some reason Name
Mary (Kelly) SummerschenGPKJ-ZL7 has a bunch of husband's listed and the husband's have a bunch of wives with the same name would anyone be able to fix this for me? I am very confused 😕
Answers
-
Philadelpiah was/is a very Irish city. That means Kelly is a very common surname in Philadelphia records. There are over 40 sources attached to Mary Kelly for various husbands and children with those husbands.
It appears you have been busy the last day or so. Have you made sure all those records belong to Mary GPKJ-ZL7?
1 -
Well the thing is I'm on mobile so it's very difficult to view the records (at least for me) And there are also multiple Mary's I will go back through and look through those records see if I can get anything sorted out
0 -
I'm an experienced researcher, and I wouldn't try working on a new family on my phone or even my tablet. I have 2 large monitors to display everything in detail when I'm working on new profiles.
If no one else has edited the family members, and you just created them this week, you might consider deleting them and starting over. No one else can delete, but you can, since you created them.2 -
It won't let me delete profiles and these people show up in my records hints but it says relationships unknown
0 -
@ChloeSlease I've looked at the problem you have run into a couple of times now. The time period and place are really outside what I generally work with, but I've decided I should give a few suggestions.
First, you have to decide whether you really want to take care of this tangled up family or, since the record hints don't show you are actually related to them, just walk away, quit attaching hints to these people and just mark the hints ignored, and leave this for somebody else.
Second if you decide you are up to this, I completely agree with Áine Ní Donnghaile. You cannot do this on a mobile device. That would be like trying to move Mount Everest with a teaspoon. If you don't have a large screen computer you can use, I'd suggest you find a local FamilySearch Center to go to where ideally you can log into three side-by-side computers so you can view a profile on one, have the profile's change log on another, and see sources on the third. You can, of course, just use one screen with multiple tabs and jump between the tabs.
Third, if you are not really, really, really comfortable with the change log, take some time to look through some change logs and get very familiar with what is there and how to use the filters. Be sure you are using the current change log and not the new experiment one.
Fourth, view this presentation which covers how to figure out who people are supposed to be:
Fifth, you need to realize that all hints and all possible duplicates on these people have a very high probability of being flat out wrong because they are based on information on the profiles that are wrong.
Now that you are ready to start, get yourself a note book where you can write down and keep track of every ID number you work with so you don't loose any of them.
The next question is where to start. Do not start with Mary or any of her husbands or with any of the husbands or any of their wives. Start with the children.
I'll run you through an example of what you need to do with everyone in a second post so this one doesn't get too long.
1 -
The reason for starting with the children is that they are less likely to have badly corrupted information so it will be easier to figure out who they were supposed to be. Also, it is usually best to start with the youngest child because the records for that child will often be the most recent records for the family.
In this case, this is Child Kelly ID P95H-VSH ( https://www.familysearch.org/en/tree/person/details/P95H-VSH )
Click on Show All in the Latest Changes box and go to the very first entries which show:
Now a common problem when people attach hints is that they often never really look at the source to see what the source says, never check the original source to see what was not indexed, and never add all the information in the original source to the profile. There are currently two sources on this child which are two different death certificates. Going to the image of the documents a couple of critical facts can be seen:
- These two death certificates are for two different children. One of them has the notation "Twins."
- The only information given on the parents is their names. William Kelly and Mary ". There is no indication as to whether the mother was to be recorded with her maiden or married name.
- Scrolling through the film, when parents' names are recorded, the mother always has the same last name as the father or is just shown with the ditto mark. Therefore we can conclude that it is her married name.
- William? Kelly? Mary? There is no way to tell from these two documents who these people are. These names would be way too common.
- There is very important place information that should be added to the profile that may help to figure out who these girl's parents are. This is the family's home address and the cemetery where they were buried. If this is added to their profiles, other researchers can see the information immediately and use it as clues to keep working.
So the next step is to fix the profile and create one for her twin by:
- Following FamilySearch's directions for stillborn children and removing the "first name" of Child.
- Removing her from the set of random parents she was put with and creating her real parents as documented with names and birth places only.
- Adding birthplace information with family residence at time of birth (I'm assume she was born at home), death information with residence at time of death, and adding the burial information with cemetery.
- Creating a new profile for her sister.
- Moving the source for her sister to that profile.
- Removing the duplicate and unhelpful residence in the Other Information section.
This turns the profile from this:
into this:
The alert note is to state the two girls are twins and not to merge them. Of course, a possible duplicate notice popped right up and I had to mark this as Not-A-Match.
So that is the first step in fixing the tangled mess you found. You could either just leave this as it is and hope another researcher someday correctly determines who this William Kelly and Mary are or you could research further. If I were to do that, I would probably check the 1900 census and see who was living at 3121 Edgemont Street (or is that Edgemont Lane? I did check google maps and there is a 3121 Edgemont Street but no 3121 Edgemont Lane so I think I'm reading that little two letter scrawl correctly) and I would try to check the New Cathedral Cemetery and see who the girls are buried next to for hints about their family.
Good luck on working on this project! It can be really fun, very satisfying, and teach you a lot about working in Family Tree.
1 -
Just one more straightforward example. This time for Baby Kelly P9Y3-F9S ( https://www.familysearch.org/en/tree/person/details/P9Y3-F9S ). This profile also appears to have been created through attaching a hint placed on random parents.
This profile has a single source which is again a death certificate that only shows father's name, mother's first name, and their address. Again it is a William Kelley (different spelling! Is this real or just a spelling variant?) and Mary living in Philadelphia. Checking the document, this one doesn't even say where the parents were born. It does give their address and the cemetery the infant was buried in.
Referencing the document and going through the same process:
- Following FamilySearch's directions for unnamed children and removing the "first name" of Baby.
- Removing her from the set of random parents she was put with and creating her real parents as documented with names only.
- Adding birthplace information with family residence at time of birth (I'm assume she was born at home), death information with residence at time of death, and adding the burial information with cemetery.
Turning this:
into this (Note the parents' ID numbers. These are completely new profiles, not hijacked ones.):
Cleaning this up brought up another hint which was for a register of deaths but which unfortunately did not have any additional information.
Could this be the same William Kelly/Kelley and Mary as for the twins? Could they have moved six miles from one home to another between 1900 and 1904? Could they have had their girls buried in different cemeteries? Of course. Do I have even a shred of evidence to support they were and did? Absolutely not!
Again the first place I would go if I was going to pursue this is the 1900 census and see who was living at 604 Delancy.
This is another illustration of the danger of just attaching whatever hints are offered. They can be very good, but they can also by wildly wrong. Particularly when a source does not contain sufficient information to properly identify people.
To paraphrase Socrates: The unexamined Hint is not worth attaching. ( https://en.wikipedia.org/wiki/The_unexamined_life_is_not_worth_living )
A separate issue is that even with fixing the records for these children, there is a danger that William Kelly and William Kelley with wives Mary and Mary will show up as possible duplicates and someone will merge them with checking to see if they really are duplicates. I don't know of any way to block that short of marking them as not a match and I don't want to do that because there is a chance they are.
These two example were pretty simple because the sources were very clear and no merging was involved. When there are multiple merges things get more complicated because you have to evaluate each merge and undo any that were incorrect.
If you have any questions if you proceed with this project, feel free to post them here.
2 -
(I meant, of course, "will merge them without checking to see if they really are duplicates." There is always one more typo and I am still hesitant about editing posts with images.)
I did check William Kelley and no duplicates are showing up but he does have five record hints. One for the birth of a child, two hints for the 1900 census with different families and two hints for the 1920 census with different families. It will take a lot of work to determine if any of the hints are for the William Kelley I just created!
1 -
And it's bugging me, days later, that I made a typo in my original reply.
0 -
I've spent a few hours weeding before the plants I want get to tall. Time to take a break. So I might as well look at one more thing.
How did William Kelly and Mary Ann (Kelly) Summerschen get to be a couple? Copying William's ID number and using that in the Find command of my browser window, there are seven instances in Mary Ann's change log. From earliest to latest:
- When added to Mary Ann as a spouse (no reason given, likely while attaching a hint)
- When linked as father to Kelly who has now been removed.
- When Kelley was added as a child.
- When a Gilbert Kelly was added as a child. (Have to take a look at him!)
- When Gilbert was moved to a different marriage for one of them.
- When I removed Kelly as their child.
- When I removed Kelley as their child.
So this is also pretty clear that this couple relationship (with no events) was only created because of those hints for children and can just be deleted. I've taken care of that.
Looking at Gilbert here: https://www.familysearch.org/en/tree/person/details/9674-W1N he was moved to the family of a different William Kelly and a Mary Ann Johnson. There are four source for him that document his one month of life. His parents are just listed as Wm Kelly and Mary A. He has a sister born just 5 months before him. There was no reason given for him being moved to this family. I suspect he does not belong there, either.
The rest of this I'll need to leave to someone who researches in this area and time.
1 -
Piece of trivia for the day: In the 1920 census for Philadelphia there are 21 entries for a William Kelly or W Kelly with a wife Mary or wife M. All but three of these are attached to Family Tree profiles.
Compiling a list of all of these with their full street addresses then comparing these to birth records and the street addresses on those would be a great way to figure out which children go where.
1 -
I do a LOT of Philadelphia area research because my County Tyrone family settled there. But, I can't view Philadelphia birth certificates from home to verify addresses of those. Those certificates require access at a local affiliate library or FSC.
0
