Some people eat, sleep and chew gum, I do genealogy and write...

Saturday, September 14, 2013

Is a unified online family tree program possible?

I think the clear answer to the question, posed in the title to this post, is definitely yes. The more specific question is whether or not the present FamilySearch.org Family Tree can function as the theoretical unified world-wide family tree? I think the answer to this second question is a qualified yes. In saying this, I know I am disagreeing with some of the more vocal members of the online genealogical community. The reason I make both statements has to do with my own experience and the nature of wiki programs. Almost all of the complaints I have read about FamilySearch Family Tree program revolve around an untrue and unsupportable assumption that a unified Family Tree in the form proposed by FamilySearch will devolve into a pile of junk. The reason why I would take exception to this assertion is based on the nature of wiki programs and my personal experience.

I would point out that the FamilySearch.org Family Tree has many (perhaps almost all) of the features of a wiki program. I have very extensive experience with wikis and I know they can function, even with very controversial topics. Those features that Family Tree has in common with most online wikis are the same features attacked by the online commentaries as precursors to failure. I believe they are making some very limited and invalid assumptions in drawing those conclusions. The concept and execution of the FamilySearch Family Tree program is basically sound. It will only fail if it is abandoned by the those who programmed it and who are designated to maintain it.

Most of the criticism is couched in terms of ownership. The detractors of FamilySearch Family Tree are certain that it will fail because their data will be corrupted. In assuming ownership of concurrently owned genealogical information, they are reacting in a highly predictable way. Every single time I have taught a class on wikis, particularly the FamilySearch Research Wiki, I get exactly the same criticism, always by people who either are hearing about a wiki for the first time and/or have never worked with one on a day to day basis to see how and why they operate as they do. In a wiki, such as the Family Tree implementation, no one owns the data. It is not your family tree. It is everyone's family tree.

What does FamilySearch Family Tree have in common with a wiki? Is FamilySearch Family Tree a wiki program? A wiki is defined as a web application which allows people to add, modify, or delete content in a collaboration with others. See Wikipedia: Wiki. The usual form of a wiki looks more like Wikipedia and the FamilySearch Research Wiki, than a family tree program. The main difference between Family Tree and the other wiki programs is in the structure of the data and the fact that it is moderated. Neither of these differences change the basic functionality of the program as a wiki. Why does FamilySearch not use the "wiki" term in conjunction with Family Tree? Simple. The negative reaction to the strange word "wiki."

Explaining how and why a wiki works in practice is relatively difficult.  One of the basic reasons why wikis work involves reliance on basic human nature. Wikis work because those who care about the stewardship of the data are generally the ones contributing, correcting, modifying, and deleting content. All of the others, including the detractors who publicly admit not participating on the Family Tree program, simply do not care enough about the content to make changes or add more content. So concerns with the integrity of the content are misplaced. As it turns out in real practice, very, very few of the imagined difficulties with wiki programs actually occur.

For the past few years, I have been quite vocal in my opinion that the only online unified family tree program that would work would have to be modeled after a wiki-type program. With a cooperative program such as the FamilySearch Research Wiki, there are always some overhead issues that need to be resolved with both the format and the content. By structuring the wiki, as in Family Tree, at least one of those issues is eliminated. I not only expect FamilySearch Family Tree to work, I see it working very well every time I go onto the program. Any present limitations in the program are derived from the data set used to seed the implementation of the program. The data is duplicative, inaccurate in many respects and poorly formatted, but all of these issues are addressed by Family Tree and are being resolved.

I do not agree with the nay-sayers. Family Tree is working so far and with some additional improvements will live up to my expectations completely. So what is needed?

It has been mentioned quite a few times that the content of Family Tree could be "frozen" as to certain individuals and families. I think this is going to become necessary. I would also think that each fact and source should have a rating system: something like Five Stars or whatever with cumulative input. Sort of like GoodReads.com where users can make comments and rate the data based on reliability and consistency. This would give the users some basis for doing further research and also deter those who would rely on poorly rated items. As an example, the passengers on the Mayflower have been extensively researched for over two hundred years. Any new information about these individuals must pass an extensive review process by professional genealogists before being generally accepted. It would seem appropriate to freeze individuals such as these and not allow any changes absent an appeal to a review committee or board. Although such a board could not be expected to know everything about all of the potential frozen individuals, it would be possible to refer the information to those who supplied the basic information for the frozen individuals in the first place. By taking this relatively minor step, much of the controversy and expected disagreements will be entirely eliminated.

Many of the detractors to the Family Tree program assume that there will be controversy about some individual's ancestry. This may well happen. But my experience with the FamilySearch Research Wiki argues strongly that such controversy will be scattered and minor compared to the goals of the overall program. The program already has a very elegant way of handling such controversies.

Most of the proposed additions to the program will also contribute to the consistency and believability of the program. I have often blamed the addition of all of the duplicate information from all of the sources added to New.FamilySearch.org as the basic challenge and problem of the program, but, upon reflection over the past few years, I do not think anything different could have been done. It is and was inevitable that we would have to confront the amassed information from previous submissions. That we have to do this now is no fault of FamilySearch or the Family Tree program. It is the cumulation of years of poor genealogical practices. In an ideal world, the data dumped into New.FamilySearch.org and thereby transferred to Family Tree would have been more consistent and accurate. But now, we have the tools, and good tools at that, to address these ongoing problems and solve them.

Don't give up the ship before it sails out of the port. I am more than certain that the basic format of the Family Tree program will and has created a mechanism that will revolutionize the genealogical community. If you could see the number of photos, sources and stories already available, just for my family alone, you would start to believe what I am saying. The tide will turn and the ship will safely and securely sail into the unknown world of our genealogical future. The program works as it is supposed to.

What about my qualified response in the first paragraph? My qualification comes from the belief that the present managers of Family Tree will stay the course and not cave in to the criticism of the few (or many or whatever). The issues with Family Tree will be resolved. My support of the program is only conditioned on the continued general support of the program by FamilySearch. As time passes, I think the concerns of the detractors will vanish like a morning mist in the heat of the noon day sun.

4 comments:

  1. James, you say "I would also think that each fact and source should have a rating system: something like Five Stars or whatever with cumulative input. Sort of like GoodReads.com where users can make comments and rate the data based on reliability and consistency."

    Since the vast majority of persons who upload trees and genealogical material in other forms to the web think that other trees and non-documented genealogical accounts of sundry types are acceptable tree components, how would a rating system by FT users have any import as to accuracy? The 100 or so folks who've incorporated wrong genealogical stuff in trees regarding two of my ancestral lines, published in at least 5 books, would not be especially inclined to give a positive rating to something disagreeing with their beliefs. Indeed, they are more likely to change what's in FT to agree with what they think is verified by each publication.

    ReplyDelete
    Replies
    1. You are making an assumption that a unified tree would work the same as a user owned tree. They are not the same. You can put anything you want into your own online family tree, but anything that goes into a unified tree needs a consensus to remain online.

      Delete
  2. Interesting you are as optimistic as you are given your comments a few days ago about people not seeing the necessity for multiple sources...

    For what it's worth, I do believe in the possibility of a unified tree. I see, though, the necessity for dealing with several aspects.

    1. As I have said several times recently, the unified tree cannot displace individual researchers' own trees. It has to live alongside them and that means the product managers of FS FT need to figure out how to interface the two sets of data in an easy and controlled manner. Otherwise, it's not practical for my stuff to go in.

    2. Wikipedia dispute resolution works at an informal level because the data is presented as text, which can have nuances built in. For instance, take "Robert de Brus, 1st Lord of Annandale" (an ancestor of King Robert I of Scotland, a.k.a. Robert The Bruce). Right now, the Wikipedia article (https://en.wikipedia.org/wiki/Robert_de_Brus,_1st_Lord_of_Annandale, accessed 15 Sept 2013) says "An early modern historiographical tradition that he was the son of a Norman noble named Robert I le Brus or de Brus who came to England with William the Conqueror in 1066 and died ca. 1094 has been found to be without basis". Now, how do we get that into an obvious place in the FS FT? Are the existing discussions going to be able to cope with the several screens of Talk that we find on Wikipedia? (I'm not saying they can't - I'm simply asking, and suggesting this as a typical example of the issues and workload)

    Actually the above example is simple. The consensus there is that we don't know his ancestry but need to record the dodgy suggestions to stop them being recreated all the time. A trickier aspect is where the consensus says "X is probably descended from Y". Like many tree software, FS FT has no ability to record unconfirmed relationships other than in free-format text. It's not just a case of marking the relationship as "possible" since it could be the case that "X definitely has a father named Y and it's possible that the Y in question is this chap named Y over here..." (I'll let the reader work out the extra possibilities themselves). So if we are to record the consensus in a formal way, we need (un)certainty options that we don't currently have. If we do have them, then the ordinary genealogists can work out and record the subtleties of the consensus themselves, without recourse to arbitration.

    3. One unified tree can work if people contribute professionally. I have my concerns about some of the stuff people want to input, stuff driven not by professional research motives - the Wikipedia talk page mentioned above says "the hideously large-sized peerage crowd help such nonsense spread like a virus all over the internet and onto wiki." This reinforces your suggestion for freezing. But then - who does the freezing? And I think back to postings on Dick Eastman's blog about another "One Tree" system, which has curators - some of whom are held to be the problem, not the solution.

    I guess I think the sun will rise but I'm unconvinced about how many hours remain to noon!

    ReplyDelete
    Replies
    1. Good points and all of this is something that has to be resolved and discussed.

      Delete