Sunday, January 23, 2011

More on the FamilySearch Family Tree Update

After I posted FamilySearch Family Tree Update on Friday, one of the comments I received was from Geolover (who is anonymous, knowledgable and savvy about genealogy companies and databases), who said:

"You say, 'My opinion is that when this FamilySearch Family Tree is released to the general public, if it works as described at the Blogger Day and the note above, it will eventually be the largest, most accurate and best sourced family tree available to the genealogical community.'

"Unfortunately the Pedigree Resource File (PRF) and Ancestral File (AF) material is largely lifted from the old, mostly sourceless, Family Group Sheets submitted by LDS members and others. In most cases it is impossible to discover whether there was some underlying evidentiary documentation.

"Tree nuclei based on the Extracted Records part of the IGI would be littered with the "same name = same person" problem, depending on how the "make a tree" program was constructed. In any event the Extracted Records are very scattered, the individual databases have many errors and omissions, and often the source-citations for them are very incomplete. One particularly noticeable omission is the names of the baptismal sponsors from those records that originally listed them.

"As you are well aware, there is a vast difference between "source" and "evidence." Most material from AF and PRF is not any more evidentiary than the vast majority of trees on Ancestry.com and other tree-hosting sites, which are compiled from other trees (cited as sources on Ancestry.com), from the old IGI consisting mainly in extracts from Family Group Sheets, and largely uncited books, publications, and web sites.

"From user comments thus far, it seems that there is still a problem in the newFamilySearch tree with making evidence-based corrections where the original submitters of genealogical assertions are unknown or deceased. As in other tree sites, there is also a large component of users who do not understand the process of evaluating evidence.

"So will this enterprise present an overall improved environment for fact-based genealogy? It is hard to see this as being so right now, but certainly time will tell."


I agree with almost everything that Geolover mentioned, and appreciate the detailed response.  As FamilySearch pointed out, they have enabled Discussions in new FamilySearch, but, as Geolover explained, you apparently need the "owner" of the person to agree to discuss issues.

To me, the critical information in the post from the FamilySearch response was:

"The data in Family Tree will also become editable by all registered patrons. Critical to the success of this plan is the provision of key tools that: 1) allow a user to be notified when designated parts of the tree that they care about are changed, 2) allow users to communicate about those changes and share the info they have, 3) and tools allowing rollback of changes deemed incorrect. The goal is to increase genealogical soundness by enabling greater collaboration in a shared tree context."

To expand on my own understanding, the FamilySearch Family Tree would (I believe):

*  Not be what "new FamilySearch" is today with "ownership" of persons and information.  No researcher would own any person in the Family Tree.

*  Permit any registered use to add, delete or edit content on any person in a wiki environment, including a a Person page, Discussions, Editing History, event conclusions, alternate event assertions, source citations, person notes, documents, images, stories, research notes, etc.

*  Would be seeded (Geolover used the words "tree nuclei") with the assertion data from the new FamilySearch databases (AF, PRF, IGI), but sources for those assertions would be removed. 

*  Be Conclusion based. The wiki environment, and the opportunity for source citations, documents and discussion, should create conclusions based on the source information and analysis of evidence ("Our Tree"), rather than a collection of assertions (many "My Trees," as currently in new FamilySearch). 

I'm sure that the FamilySearch Family Tree will not sprout perfectly overnight, it will probably take years for (hundreds of?) thousands of researchers to add their families to it.  It will also take some time for each researcher to learn how to interact with other researchers, and the tree itself, likely through genealogy software and direct entry in the tree. 

I tend to see things as a glass half full - I want genealogy websites, family trees and research processes to work and work well for the benefit of all researchers.  I want to participate in making it better, and I think all researchers want it all, and we want it now. 

1 comment:

Lineagekeeper said...

The FamilySearch database will undoubtedly be the largest and best evidenced data in the world ... one day. But as noted, it will take many, many years of effort and cooperation between submitters before that day arrives.

Critical to its success is the release of a promised but never realized API that allows the uploading of sources directly from a user database. The current manual sourcing feature is terribly slow and difficult to use. Additionally, there is little evidence of current users taking the time to read the sources, coupled with the source report data that seems to shift user sources to the bottom of often very lengthly source lists on events.

Additionally, controls of some type must be included to stop the mindless incorrect combinations of records by users. Today, it is difficult if not almost impossible to correct errors and incorrect linkings and combinations that were easily introduced by programs such as Roots Magic, Ancestral Quest and new to the direct interface scene, Legacy, in addition to the erroneous combinations made manually by users.

Frustration by users who try to submit well-sourced data only to see it combined with subsequent gross errors by careless users is a deterrent to ongoing participation and effort in achieving the promise of the database.

It will be interesting watching how the problems are resolved. Human 'moderators' will undoubtedly be part of the process at some point to resolve the apparent inability of many users to add and combine data responsibly.