Friday, August 31, 2012

FamilySearch Citation Report Card

DSC01056At the FamilySearch Blogger Dinner of this year’s annual Federation of Genealogical Societies conference, FamilySearch spokesperson, Michael Hall presented a report card of sorts on FamilySearch’s citations.

I have to give a big conflict of interest here. I serve as an advisor to those implementing citations at FamilySearch. I’m partly responsible for what’s good and what’s bad. I also have to maintain good working relationships. But while I can’t give an unbiased view of FamilySearch’s citations, I am in a position to understand why certain decisions were made.

Hall’s presentation was augmented by this handout:

Introduction

FamilySearch adheres to the principles of “Evidence Style” as presented in Elizabeth Shown Mills’s book Evidence Explained: Citing History Sources from Artifacts to Cyberspace and in associated reference sheets such as QuickSheet: Citing Online Historical Resources: Evidence! Style and QuickSheet: Citing Ancestry.com Databases and Images: Evidence! Style. Applying these principles can result in differences between FamilySearch citations and similar examples in these reference works. Some of the differences are operational and some are strategic.

Evidence Explained is used because it is the only citation guide that supports derivatives and covers the many kinds of manuscripts used by genealogists. Other citation styles could be considered, but none support derivative sources and none cover all the different kinds of sources used by genealogists. An MLA style guide, for example, is not going to tell you which citation elements are necessary to locate a birth certificate created at the county level and filed at a state archive. Elizabeth Shown Mills took an existing style guide, Chicago, and extended it to meet the needs of genealogists.

Derivative Sources

Genealogists depend on derivatives of manuscript sources. Derivatives are copies of various sorts such as transcriptions, abstractions, and microfilm or digital images. Genealogists can seldom visit the churches and archives housing the many manuscript sources they use. Consequently, they must depend on derivatives.

A genealogical citation, therefore, should specify the location of both the online derivative and the offline original. A complete citation basically combines a citation for an online derivative and one for the offline original. The two are joined using the word “citing.” Any information already found in the citation to the derivative is dropped from the citation to the original.

For example (with “citing” bolded),

        1. "United States Census, 1790," index and images, FamilySearch (https://familysearch.org/pal:/
MM9.1.1/XHKG-MTH : accessed 07 Aug 2012), Paul Raymond, Richmond, Berkshire, Massachusetts; citing p. 503, NARA microfilm publication M637, roll 4.

Collections, Records, and Images

FamilySearch needs to provide users with citations to its record collections, records, and images.

Collection citations are displayed on collection pages. Up until recently they have been stored in the FamilySearch Research Wiki and have cited the offline original but not the online derivative. Today (or very soon, if not already) they are stored in a secure location and cite both the online and offline sources. Some of these citations need improvement, which FamilySearch will do over time.

Sample record citations are presented in the Research Wiki, and for a long time no citations were displayed on record detail pages. FamilySearch began displaying citations to the online derivatives on the record pages several months ago. Today, for our most popular collections, FamilySearch cites both the online derivatives and the offline originals. Over time, FamilySearch will make this improvement to additional collections.

FamilySearch is not yet displaying image citations, and we have no progress to report at this time regarding such an effort.

Multiple Archives

Collections composed of records from multiple archives are difficult to cite. Collection citations in the wiki cited something generic like “County courthouses in Ohio,” rather than attempting to list all the counties in Ohio. Such citations have little value. While users need to know the coverage of record collections, it is impractical to list dozens or hundreds of archives in collection citations. Of necessity, coverage information must be handled elsewhere. In such cases FamilySearch follows the Evidence Style practice of citing just the online derivative.

While a collection may contain records from multiple archives, each record comes from a single archive, and the record citation can specify it. Unfortunately, our current technology does not allow FamilySearch to do so. However, we can include the FHL microfilm number. Using the film number, researchers can look up the archive in the FamilySearch catalog. Developing the technology to do the lookup automatically is expensive and will take a considerable amount of time, but this functionality should be available at some point in the future.

Conclusion

FamilySearch has made considerable progress in our goal to present users with complete, genealogically sound citations for collections and records. Additional work is needed for images and for citations to collections with records from multiple archives.


This leaves plenty of questions to be answered. I may answer some of the larger ones.

1 comment:

  1. I disagree with this post.
    The citation process is clunky and confusing.
    Requested and required data fields are not intuitive.
    Editing them is very difficult.
    Basing them on "Evidence Explained" is a mistake. The citation process detailed in this book suffers form the same faults. I believe the author, Elizabeth Shown Mills, states that there are over 100 different citation formats in her system. Crazy! I have gven up trying to use her formats.
    A great system known as Simple Citations (http://www.simplecitations.com/overview.html)
    SHOULD become the international genealogy standard. They are simple and consistent from one source type to the next.

    ReplyDelete