Citation quirk for FamilySearch

mbcross, those quirks do frustrate us.

Your Paragraph 1:

EE's Velcro Principle comes in handy in these situations.  To quote from EE4 3.16:

THE RULE THAT HAS NO EXCEPTION
The mix-and-match nature of citations to online images has one cardinal rule: Details that describe one entity must not be attached to a different entity. Think of this as a Velcro Principle for online citations: what’s meant to stick together should stick together.

If we choose Emphasis on Record, then all details we use to describe the original record stay together in Layer 1, while all details that identify the website or database stay in Layer 2. For example:

• An imaged page may display an original page number, while the website’s frame around that image may state an image number. The number of the image created by the website cannot be used in place of the document’s own page number, or vice versa.
• The title of the website’s database can never be substituted for the record title in the layer that identifies the record. Any user’s effort to find that database title within the original record set would fail, because the website’s database title will not exist within the original record set.

In the situation you reference in paragraph 1, you would identify the record book by the label on its cover. If FamilySearch divides the book and calls it S, Sa, Sb, [or whatever], then that information goes in the layer in which you identify FamilySearch.

Your Paragraph 2

The link you provide in your full citation takes me straight to the image. No problem there. However, the paragraph 1 situation is not discernible there and other problems arise:

• When I use the FS path above the image, that refers to a database that your citation does not identify.

• When I use the image group number (004395464) and key that into the catalog search box, I get no result at all from the catalog. It is as though the image group number does not exist.

• When I use the catalog search box to query for Wharton County, Texas, marriages, the cataloging data does not include 1942 marriage records

• If I choose the option "1909–1934," even though it does not encompass 1942, the more detailed cataloging data does include 1942.  There, we do see records from 1942 offered under the 1909-1934 group, with the image group number you specify.

Unfortunately, quirks like this are a common occurrence with online records. 

How does it affect you? Well, the nice citation that you created will work so long as one uses only the URL and image number. If you (or a user of your citation) make a typo in the URL—or, if its changed at some future date—then the backup information you provide (the image group number) won't be usable without experimentation as I did above—until and unless FamilySearch corrects the catalog.

Because of this, it would be best for your citation to identify the database in which that image group appears:

Wharton County, Texas, Marriage records, vol Q: 179, Kurtz-Derrich, 21 September 1942; imaged, "Texas, County Marriage Records, 1837–1965," FamilySearch (https://www.familysearch.org/ark:/61903/3:1:S3HY-65B9-3W5 : accessed 27 Feb 2024) > Image Group Number 4395464 > image 233 of 376.

Now, Matthew, we're at the heart of the issue (well, one of them): what do we do when image producers image the inside pages of original record books without imaging the cover?

And this leads us right back to “The Rule That Has No Exception,” discussed above for layered citations: Details that describe one entity must not be attached to a different entity.

Your draft citation (with coloration to differentiate the layers) is this:

Wharton County, Texas, Marriage records, vol. P: 257, Kurtz-Derrich, 21 September 1940; digital images, FamilySearch (https://www.familysearch.org/ark:/61903/3:1:S3HT-X9XS-L1T : accessed 27 Feb 2024), Image Group Number 4395432 > Image 311 of 354

The image that you have used is this:

​

From that image, we can discern the following:

• the record is from Wharton County, Texas
• the page number is 257
• the record is for a Kurtz-Derrich marriage that took place on 21 September.

What we cannot identify from this image is the name of the book.  For that information, we typically scroll back to the first images for the book. However, in this case:

• The first images for this image book show no cover and no title. The first imaged page is the first page of the index.
• Our identification of the book, when using this website, comes only from FamilySearch. Specifically, we find it under the “Catalog Record” tab that appears in FamilySearch’s frame for the image. There, the “Film/Digital Note” informs us that this is “Marriage records, vol. P, 1939–1941.” 
• However, if we take information from FamilySearch’s frame and insert it into Layer 1 of our citation (our citation to the original record), then we are “borrowing a citation” from our source of these images. We do not know, from our own examination of the book, that it actually is Vol. P. (And yes, image providers do make errors, even FamilySearch.)  
• We can verify, from our own examination of the images, that this set of images begins with 1939. However, this image group ends at p. 300 with 1940 rather than 1941, as stated in the cited title for the volume "... vol. P, 1939–1941."

This leads us back to your question:  What do we do when FamilySearch arbitrarily divides a volume into parts and assigns new numbers or new identification for the additional parts?

As you’ve already determined, if we go to the next image group in FS’s series (IGN 4395434), we find the same issues: The image set begins with image 301—suggesting, logically, that it is an extension of the prior image group (IGN 4395434).  However, the identity of the volume has changed:

​

According to the Film/Digital Note in FamilySearch’s frame, we are now using a different volume, one called “Marriage records, vol. Pa. 1939–1941.”  What we don’t know is this:  What is on the cover of the original volume? When someone goes to the record office that holds this volume, what volume title do they look for?  Many times, record offices themselves split “fat” volumes into parts when they fall apart and need rebinding. We might assume from these images that FamilySearch divided the original volume into different image groups and renamed the parts—but that’s an assumption or a conjecture until and unless we examine the original volume. We don’t cite assumptions or conjectures as fact.

Bottom line: 

We cannot identify the original volume from what we see on the images. We can only cite what FamilySearch tells us.  Thus, the need for The Rule That Has No Exception: details that describe one entity must not be attached to a different entity. Details assigned by the provider should not be inserted into our identification of the original record.

Solution:

The solution here is simple. Instead of crafting our citation to emphasize the original book that we have not used and cannot identify of our own knowledge, we cite what we used—the database:

“Texas, County Marriage Records, 1837–1965,” FamilySearch (https://www.familysearch.org/ark:/61903/3:1:S3HT-X9XS-L1T : accessed 27 Feb 2024), Image Group Number 4395432 > image 311 of 354 > Kurtz-Derrich, marriage, 21 September 1940, Wharton County, at p. 257 of book with no cover image; citing “Marriage records, vol. P, 1939–1941.”

Does all this matter, so long as we're citing the FamilySearch image?  Not really, at this moment. But what happens when FamilySearch loses its permission or license to offer these records and the images are taken down?  In that case, any effort to verify the authenticity of an image in circulation among researchers will require consulting the originals in the record office that holds the records. At that point, the exact identification of the volume we have used does matter significantly.