Tuesday, July 31, 2012

Getting Benjamin Seaver (1791-1825) Right on FamilySearch Family Tree

Last week, I posted Sorting Out Benjamin Seaver (1791-1825) on FamilySearch Family Tree and lamented that the erroneous spouses of Benjamin still showed up in the FamilySearch Family Tree, and his father was still himself instead of Benjamin Seaver (1757-1816).   

Reader Geolover commented:

"The couple relationships that you believed you deleted for Benjamin b. 1791 are still there because at least one person is in the system as child of each of the relationships.  Using the small arrow at lower left of each couple box you can see the child(ren) in the system for each couple.

"You must research the actual parentage for each child before proceeding to remove the child from the couple. The child may be assigned to correct father or to correct mother. You need the Person ID number (PID) for each correct parent of each child. When you go to the child-parent relationship screen you can remove biological relationship for wrong parent and have the option to change parent to correct parent.

"Here is where it gets sticky. There are probably duplicated individuals listed as children of Seaver relatives. The listed child(ren) of wrong parents under Benjamin Seaver b. 1791 may be composites of more than one actual person. There may be multiple wrong vital dates/places assigned to any individual child.

"What seems like simple procedures can quickly turn to quicksand.  The child-parent relationship screen can be found by mousing over a child, clicking on "view" link at right."

I appreciate the response, and the specific recommendations to solve my multiple-marriage problem.  I did that - I deleted the relationships between Benjamin Seaver (1791-1825) and his erroneous wives and children, and the "problem" disappeared.  There may be some "loose women" in the database who used to be identified as Benjamin's wives, but at least they aren't married to him any longer.  This was actually fairly easy to do.

The more difficult task was getting Benjamin (1791-1825) to be the son of Benjamin Seaver (1757-1816).  I did it using this sequence (I think!):

*  Delete the relationship between husband Benjamin Seaver (1791-1825) and Martha Whitney (1764-1832).
*  Delete the relationships between Benjamin Seaver (1791-1825) and the erroneous children listed for him and Martha.
*  Go to Martha Whitney (1764-1832) and click on the "Add or Find Spouse" link, put Benjamin Seaver in the search fields, and pick the right one.  Then Add the Marriage record for them.

That went well, although I can't recall now if I had to do something to attach all the children to Benjamin and Martha again.  

I then cleaned up all of the Event and alternate name data to eliminate duplicates and obviously erroneous spellings, edited the locations to standardize place names, and now have the families for the two Benjamin Seavers in decent shape.  

They look like this all cleaned up - here's the Pedigree chart:

Here's the screens for Benjamin Seaver (1791-1825) (three screens shown):

In the screen above, you can see the Discussion section.  On 2 June 2011, I added this comment in the Discussion section:

"There are 10 children listed for Benjamin Seaver (1791-1825) and Abigail Gates (1797-1869), who married in 1817.
The first six, born between 1797 and 1812, are the children of Benjamin Seaver (1757-1816) and Martha Whitney (1764-1832), the parents of Benjamin Seaver (1791-1825). The last four - Abigail, Lucinda, Isaac and Benjamin are rightly in the family of Benjamin Seaver and Abigail Gates. Records for all of these births are in the Westminster MA Vital Records book and in the Town Records book. How does something like this get fixed?"

So I added this as a Comment today:

"I fixed this situation by deleting the erroneous relationships and managed to get Benjamin Seaver (1757-1816) as father of benjamin Seaver (1791-1825). I will be adding source records as time permits."

All of the Changes I made for Benjamin Seaver are in his "Person Changes" log. There are over 50 Changes performed to get to where I am for Benjamin Seaver (1791-1825):

For some of those records, I got fairly snarky with my Reason for deleting the Relationship.  For example, here is one Reason I used:

"This Benjamin (1791-1825) was not the son of Benjamin Seaver (1791-1825) and Martha Whitney (1764-1832). He was not his own father. Very careless and stupid error made by someone unskilled in the genealogy arts."

I'm sure that I could have been more efficient in doing all of this.

As you can tell, there is a definite learning curve required to understand how to fix relationships of this nature.  The steps are logical, but the user needs to figure out the best process to use.  

The next step for the two Benjamin Seaver families is to go into the FamilySearch Record collections and put Source information into the "Source Box" for each person in the two families.  I'm not going to attach sources for records not yet on FamilySearch because it has to be done by a copy/paste process.  I'm still hoping that the genealogy software vendors (like RootsMagic) will find a way to put source citations into the FamilySearch fields for each person.

The problem with all of the above exercises in FamilySearch Family Tree is that it took a significant amount of time to perform - in this case, probably two to three hours to correct these two families.  

I'm still bothered by the "loose women" that I detached from Benjamin Seaver... do I need to find them and delete them as persons?  How do I do that?  [Note: I don't have their Person IDs any longer].

Copyright (c) 2012, Randall J. Seaver

1 comment:

Geolover said...

Randy, thanks for your brave post and acknowledging my pointer to first steps for fixing relatives of Benjamin Seaver (b. 1791).

You ask about what to do with the "loose women" whom you detached along the way. Let's say you mean Martha Whitney (1764-1832).

One thing one needs to check when detaching persons as spouse or child is the person screen for those persons.

Under Benjamin Seaver (b. 1791) there were several boxes for wrong marriages, with one or more child(ren) attached to each marriage.

In any tree one has to start with the later generation -- looking at each child's page, for instance, to see whether more than one Benjamin Seaver was attached as parent, and more than one mother. This can give you a clue that there is more than one version of Benjamin Seaver (b. 1791) in the system, and perhaps more than one version of Martha Whitney (1764-1832). You need to look at each of the individual parents' screens from the child, to determine whether there are similar persons with different Person ID numbers (PIDs).

It's only under parents' screen or general search that you can tell whether there is a duplicate version of any or each of Benjamin Seaver's children, similarly by looking at each one's individual screen and noting the PID.

The problem with all this background work is that two key elements are not now installed in the Family Tree software.

1) The most useful in the above circumstances, if you find duplicate versions of Martha or of Benjamin's children or of Benjamin, is a future feature: merging individuals.

2) Another useful feature where individuals can be deleted, also has not been installed. This needs to be used with a great deal of care because one needs to determine that the person did not exist in fact and that the person is not connected with other persons in the system that you might be able to connect correctly to spouse(s), parents, children.

I have found some invented persons among some of 'my' families, but more often I have found only vaguely-defined persons whose identity (e.g., actual parents) I have in a database and therefore know that the person is actually a duplicate of a person shown elsewhere in Family Tree.

Logical further work in such circumstances is stymied by lack of software in Family Tree that would enable merging duplicates and getting rid of actually invented persons who are not otherwise connected.

My approach has been to leave the nonexistent persons there for future work, and to leave duplicates in place for when the software will be available to merge them.

This makes for messy trees, and this is going to become even more weird when more persons get access to FS-Family Tree and start to do the usual tree stuff.

In addition, there will be software installed allowing users to upload material from GEDCOM files uploaded to either PRF or AF. Supposedly there are going to be safeguards to minimize adding duplicates, but I expect that all of my work (and yours) to straighten out mistaken family groups may be defeated if someone sees "oh, this person isn't shown married to the Martha Whitney as I have in my tree."

This is one use of the "watch list" which will notify you when someone makes a change regarding an individual.

When the additions-from-GEDCOM begin to reverse what I have already fixed, I am done with it. There already are jillions of mistakes in what was migrated from new.FamilySearch, and I have no interest in spending time in wikiwars.