Citations need to identify the original document

LegacyUser

James Cobban said: The purpose of a citation is to permit researchers to verify that you have properly interpreted a source document when incorporating it into the conclusions. However most of the suggested "citations" from both Family Search and Ancestry are of the form "database such and such". Period. Full stop. For example I am currently looking at a death certificate from Michigan. The suggested source citation is "Michigan Department of Community Health, Division for Vital Records and Health Statistics; Lansing, Michigan; Death Records". That is clearly not a citation. It could be a source title, but the suggested source title is a reference to the online nominal index database. A citation ideally identifies the specific page in a source document. Obviously the "Michigan Department of Community Health, Division for Vital Records and Health Statistics; Lansing, Michigan" must have a mechanism by which its staff can locate a specific record, but you certainly could not identify it from what is presented here. The image and the URL looks like a reference to a particular frame on a particular reel of microfilm, but I cannot see any way to identify WHICH reel of microfilm or how to obtain access to it, because that information is missing from the "citation", never mind how to gain access if desired to the original paper document to see if there is information which has been obscured by the conversion to microfilm. See https://www.ancestry.ca/interactive/6...

LegacyUser

Tom Huber said: First, very few census records' original documents are on film, but are paper copies at the archive (in the case of U.S. Federal Census, the NARA) of the state involved. The citation should cite the crucial information. Most states have an index they can access, which means for death records, you need the deceased person's name and date and place of death.

Some states have imaged their records, using Family Search and Ancestry, depending upon the state (for instance, Idaho uses Ancestry). The images serve as a copy of a paper copy. This is becoming more and more common among many organizations and governments. In some cases, the organization does this to protect the paper records, which in a number of instances are becoming fragile and in serious danger of loss.

Now to citing these kinds of records. If the record has been digitized and is available on a specific cite (it doesn't matter if it is FamilySearch, Ancestry, or an organization's site), then the citation should contain the crucial information, but generally, a bibliographical entry is probably preferred. As far as I can tell, there is no "standard" reference that the users in this forum agree upon.

For me, I prefer something along these lines:

U.S. Census Bureau, Census: U.S. Federal Census, 1880 (Online database), NARA Series T9, Roll 178, E.D. 46, Pg 420A, 4 Jun 1880, National Archives and Records Administration, 7th & Pennsylvania Ave., NW, Washington, DC 20408.

LegacyUser

David Newton said: What's the most important thing for a source citation to do? Provide enough information to reliably find the document cited. Any source citation which does not do that is worthless. That means the vast majority of source citations in online trees are worthless. That includes "citations" created by the tree providers themselves.

Take a "citation" for records from a church near me which are deposited at the local record office:

""England, Warwickshire, Parish Registers, 1535-1963," database with images, FamilySearch (https://familysearch.org/ark:/61903/1... : 26 February 2019), Emily Hales, 03 Feb 1876; from parish registers of the Church of England, database and images, citing Baptism, Milverton, Warwickshire, England, United Kingdom, Warwick County Record Office, England."

So the reference for the online bit is just about passable, but pretty much everything else is worthless. Where is the full site postal address of the repository? Why is the repository not named correctly? (NB this is a "good" parish register citation, it at least identifies the correct record office!) Where is the internet address of the catalogue of the repository? Where is the proper document catalogue reference number for the original document? Where is the page number and entry number reference for this specific entry in the original document? All of those need to be present for a proper citation.

Could I find that record were I to go to the original archive when it re-opens on 2nd January 2020? Yes. However I would only know exactly where to look because I am familiar with the local area and know which church is being referenced. That should not be necessary as the information should be within the source citation. In this case does Baptism, Milverton refer to the christening registers of St James the Great, Milverton or to the christening registers of St Mark, Milverton? The date is after St Mark's opened so it could be either. There is no information in the citation which says. As it happens in this case the entry is one of the first from the registers of St Mark, Milverton.

What would a proper archival catalogue reference be for this case? DR0610/1 Register of baptisms (parish of New Milverton), 1875–1886. Then a page and entry number reference to pin down the individual entry. In this case page 2 number 9.

Move to a place name with more parishes and this problem gets even worse. Consider the town of Leamington Spa for example. By 100 years ago in that town a christening could have taken place in All Saints' (the original ancient parish), St Mary's, St Mark's (same St Mark's as above since that part of Milverton was functionally part of Leamington and became so administratively after 1894), St Paul's, St John the Baptist's, Holy Trinity, Christ Church, St Alban's, St Luke's or St Saviour's. That's just a medium-sized town.

As with all online providers Familysearch has a serious reference quality problem. Ancestry seem to be slowly doing something about this, as you will more and more often find proper archival document references in their databases. Findmypast have always tended to have very thorough references and I have no idea about MyHeritage. Familysearch doesn't seem to be making any progress on the subject however. This isn't a matter of programming resources but of data mangement.

LegacyUser

David Newton said: NARA is particularly bad at providing online catalogues as an archive and they are absolutely appalling at providing a consistent citation format for their whole collection set.

Compare that to the UK's archives at Kew and the contrast is very strong. TNA have a very well-written, comprehensive, but also comparatively brief document on how to cite their collections. However Familysearch have completely ignored the authoritative citation guidance. It is thus impossible to trace an England and Wales census citation back properly to the archival catalogue. It's an easy, albeit big job to create proper TNA citations and they've thoroughly botched it.

LegacyUser

Juli said: My usual citation format skips most of the useless frou-frou:

FamilySearch Film #NNNNNNNNN Image P of Q (Place, Denomination, Type, Year): entry/line number, Name1 Name2, Date (link).

Alternately, if it's a waypointed collection, I can use the waypoints:

FamilySearch - CollectionName - Waypoint1 - Waypoint2 (etc.) - Image P of Q: entry/line number, Name1 Name2, Date (link).

Since the church and civil registers that I'm mostly citing are not published, 99% of the fields provided by genealogy websites and software are completely inapplicable. There is no author. There is no publisher or publication date or anything else to do with a process or event that never happened. Inventing an author or publisher is not going to help anyone find the record.

I have successfully used such citations to re-find register entries in other repositories, and even in other copies of the register. The film and image numbers are not useful in such cases, but the date and entry/line numbers are. (And of course the specific location and church and such that I include in the parenthetical part.)

LegacyUser

David Newton said: If it's on a FSFT general circulation microfilm it has most certainly been published. Publishing is the issuing of copies of a work to the general public, and so making copies available at FHCs around the world definitely qualifies. However I'm inclined to agree that "publishing" information of that type isn't especially helpful.

What is necessary however are the original archival document references where it is an archival document. Why? Simply put so you can walk into the archive and order the original document to consult. That's the whole point of such a source citation: directions to allow viewing of the original document. Oh and just because something is an archival document does not necessarily mean it is authorless.

Citations need to be independent of digital records provider. I should be able to use the citation to find a record equally well on Familysearch or Ancestry or Findmypast, assuming that each has an available copy of the work.

LegacyUser

Juli said: The Hungarian civil registrations are in large part straight to digital (no microfilm available). But even when it does involve microfilm, the "publisher" is unclear and totally irrelevant to finding the record again.

The auto-citations are Totally Useless:

"Hungary Civil Registration, 1895-1980," database with images, FamilySearch (https://familysearch.org/ark:/61903/1... : 3 November 2019), Palotay József Béla, 2 Jul 1942; citing Marriage, Budapest, Pest-Pilis-Solt-Kiskun, Hungary, Archiv der Stadt Budapest (Archive of the City), Hungary.

Why is "Archiv der Stadt" in German, of all things? Which district of Budapest?? (There are over 20.) If it's a marriage, who's the bride, for heaven's sake? If I showed up at a civil registrar's office or the city archives in Budapest with that citation, I'd come away empty-handed, and if the URL got corrupted (as it doubtless will be by the standard truncation that GetSat applies), it'd take hours of browsing through each district's marriages in July of 1942 to find the record again -- basically the same as if I didn't have the citation.

Compare my citation format options (with my usual abbreviations 'cause I'm lazy):

HCR - PPSKK - BP XI - Házasultak 1942 (jún) - Image 9 of 149: 411. dr. Palotay József & Stettner Mária, 1942 jul 2 (https://www.familysearch.org/ark:/619...).

FS Film 4429758 437/580 (BP XI házasultak, 1942): 411. dr. Palotay József & Stettner Mária, 1942 jul 2 (https://www.familysearch.org/ark:/619...).

Even if the URL went AWOL or FS went belly-up, I would be able to find the marriage record at any archive or on any website that had the register books available in some format or other. Notice how there's no mention of publishers and authors and such, because such things are irrelevant to the purpose of identifying the record.

LegacyUser

David Newton said: Civil registration does tend to be a special case. An equivalent for England and Wales would be, 1875 Dec qtr Warwick vol 6d p 577. That plus the name of the person would be enough to get a birth certificate from the GRO. The transcription's got all the information on Familysearch in this case, but again the "citation" is useless.

LegacyUser

Robert Raymond said: Hi James, Thanks for your continued campaign to improve FamilySearch citations. It mirrors my own internal one and I appreciate it very much.

It turns out that complete citations require indexing fields that serve no purpose beyond the citation. Decision makers must choose between indexing more genealogical collections or indexing citation metadata, resulting in fewer collections. We have fewer indexing volunteers than we would hope. We appreciate all the indexing you are doing to solve the problem you are raising.

So while there is great value in citations, there is also great value in a greater number of records whose citation metadata can be gleaned from the images by the sophisticated researcher.

You and I might make a different decision than FamilySearch's decision makers. But if we continue raising the issue, they at least can see the importance that users place on having complete metadata.

LegacyUser

David Newton said: What should be the top priority? Sorting out the gross errors in quite a few collections. There has been another post pointing out that records from Cambridgeshire have found their way into a collection for Hampshire for example.

LegacyUser

James Cobban said: Thank you for your good wishes. I remind FamilySearch that you are invited to copy information from my web site including transcriptions, with the exception of the Family Tree itself. All information is visible not only in human readable (HTML) form, but also in XML and JSON.

FamilyTree has been doing a better job than Ancestry. In particular for the past several years your volunteers have been including the registration number of vital statistics from Ontario. By contrast Ancestry does not even get the year right for many vital statistics records.

I have been forced to largely duplicate the functionality of the FamilySearch and other genealogy sites for several reasons:

1) I started my project 30 years ago.
2) I wanted my family tree presentation to supersede the need for a traditional published family tree, so that the page on a family looked as much like a traditional publication with all of the advantages of the Web added, such as hyperlinks instead of "see page 127".
3) A major goal of my research was to be able to actually measure the progress of the research. That is to be able to state what percentage of the information in a particular source document has actually been incorporated into the family tree. This requires implementing a reverse link from each item in a source document to the member of the family tree that it documents.

If I could get FamilySearch to work with me to incorporate the capabilities of my site into FamilySearch I could focus on the research rather than maintaining my own site.