Citing online image copies with inadequate information or provenance

Question

In a recent blog post, Elizabeth Shown Mills asked the question Do You "Just Trust" Citations Offered by Digital Providers?

EE hopes your answer is no. Today’s image demonstrates why. Not only do we need to double-check the factual details, but we also need to consider whether the “ready-made citation” actually covers all essentials.

Find My Past recently added the Index to New England Naturalization Petitions, 1791-1906, NARA publication M1299. Unlike Ancestry and FamilySearch, which supply a ready-made citation which can be cut-and-pasted into one's software, Find My Past merely offers a rudimentary transcription which cites the title of the collection and the microfilm number (which spans several rolls), and nothing else.

The display of the image itself says "New England Naturalizations, 1791-1906 Image" with no indication of what roll it comes from, or where it might be on the virtual microfilm roll (no "n of how-many-images"). They do not offer navigation through the entire virtual roll, so it isn't possible to scroll backwards to see whatever information might be on the cards at the start of the roll. Unlike the England and Wales Census images where the RG numbers are visible in the image itself, there is nothing showing in the frame. The only clues are in the name of the image when downloaded, which might be completely arbitrary. I could save the URL as a rudimentary record as to the path I took to get there, but that could change at any time.

Here's an example of the transcription available from Find My Past:

• First name(s) Fred
• Last name Abelein
• Birth year 1860
• Birth country Germany
• State Massachusetts
• NARA publication Index to New England Naturalization Petitions, 1791-1906
• NARA publication number M1299
• Record set New England Naturalizations, 1791-1906
• Category Immigration & Travel
• Record collection Naturalizations
• Collections from United States & Canada

Contrast the information one gets from FamilySearch.org:

• Name: Fred Abelein
• Event Type: Naturalization
• Event Place: Massachusetts
• Birth Year: 1860
• Birthplace: Germany

Affiliate Publication Title: Index to New England Naturalization Petitions, 1791-1906 , Affiliate Publication Number: M1299 , Affiliate Film Number: 47 , GS Film number: 1429717 , Digital Folder Number: 4639023 , Image Number: 00366

Citing this Record "New England Petitions for Naturalization Index, 1791-1906," index and images, FamilySearch (https://familysearch.org/pal:/MM9.1.1/VX5B-52P : accessed 07 Jul 2014), Fred Abelein, ; citing Massachusetts, NARA microfilm publication M1299, roll 47, National Archives and Records Service, Washington D.C.; FHL microfilm 1429717.

In the grand scheme of things, this specific example is not really important. The Index is merely a finding aid to other records, and if I wanted to record where I found the lead to Fred's naturalization record, I could simply discard the Find My Past discovery in favor of the one from FamilySearch. I don't anticipate finding any records on FMP that I haven't already seen on Ancestry or FamilySearch. It isn't always easy to tell if they are all using the same index and images provided by NARA. (How to find out that information should probably be a separate question, but on FamilySearch and Ancestry.com, the images show three cards; Find My Past returns the single image for that search result but nothing else, so I don't know if those few cards with images of the reverse can be accessed on FMP -- there is no name on the reverse that could be searched for.)

But what if this were an actual historical record, not simply a finding aid, and (for whatever reason), I could only find it on Find My Past, or Brightsolid's image copy was superior to all the other vendors? How are users supposed to cite the image copy when the vendor gives practically no information about where the image can be located within the virtual microfilm roll or within the collection, or where their image copies come from?

---

For those not familiar with this collection, a document describing Publication M1299 is available for download at Archives.gov. (This supplies the Roll numbers which Find My Past didn't bother to put in their abstract.)

This image, like others, is particularly annoying because there is a note at the bottom of the card that says (over). There was more information on the other side of the index card. As far as I can see, the opposite sides of these cards were not filmed a lot of the time. The NARA description (see link to PDF at bottom) says:

The index consists of 3x5 inch cards arranged by state and thereunder by name of petitioner, arranged according to the soundex system. The index refers to the name and location of the court that granted the certificate of naturalization, and to the volume and page number (or certificate number) of the naturalization record. The printed cards have spaces, often left blank, for other information from the naturalization papers such as place and date of birth.

Of the handful of cards I've found in this collection, only one of the people whose card said (over) had an image of the reverse side in the microfilm. The information on that card was the date a certificate was issued (which may not coincide with the date of naturalization, the field actually printed on the front side of the card).

One of the things on my To-Do list is to assemble a set of citations for the cards where I have no second image so I can email NARA and confirm that the second side of the card is lost because no image was made. This is the practical reason I need to be able to specify exactly what images I have been looking at; simply attaching an image to an email won't give an archivist much information about where the "original" (in this case, original = the surviving derivative image copy) might be found in their collection.

Answer 1

An interesting question. I am sure I cannot give a full answer, but perhaps some pointers to the things that I would think about might help drive some other aspects out.

Firstly, from my perspective (from the UK), why do I want to cite my sources?

1. In order to find them again;
2. To assist me to evaluate this source against another in the event of discrepancy;

Those 2 aspects may help me assess whether a provided citation works. Another test that I apply is to imagine someone with a different derivative of the same original source. If I used a microfilm of a parish register, could someone find the same event in an on-line digitisation? Or vice versa.

On that basis, I get to be inherently suspicious of a lot of the citations provided by Ancestry or FamilySearch. They keep quoting detail that bulks out the citation but seems to be of little practical use. For instance, NARA microfilm and roll. Can I use Ancestry or FS to enquire on NARA microfilm and roll? If I can, then it's useful. If I can't, then what's the point of it? (Of course, if I'm sat in NARA and they still have microfilms, that might be different...)

However, in fairness, there is perhaps some use for the NARA microfilm and roll. If I navigate to an image and I'm not sure if it's the right one, then if the NARA microfilm and roll are different - then it isn't the right one. Except... Except... The other month I was reading the Canadian Archives catalogue about its census microfilms, and that informed me that one (at least) of its censuses had been filmed twice. Different microfilms, same original.

Then there are the "image X of Y" values. That's fine until the provider discovers the odd missing image, inserts it, and your image counts are useless!

Browsing up and down a dataset are crucial needs - but all too often I've gone to "image 1" to find it's not the first image of the original at all, but the first image in some arbitrary selection.

So my tentative conclusion is, I think, that details of the intermediate microfilm or digital files are of dubious value, and nothing like as useful as the details that identify the original.

So what might be of use in an on-line site? If I think about finding the record again, and I cannot enter those film, folder, and roll numbers as search parameters, then I need to record what the search path is. On that basis, the FMP abstract (they have a bad habit of referring to them as transcriptions) probably suffices to find the image in FMP again. And it possibly also suffices to find the same image in the other providers.

Recording the search path is the closest I can get to an answer right now....

The big hole in the above is that we get no ability in an abstract as abbreviated as FMP's to assess the usefulness of this source. But - are the others much better? Do they describe who created the imaged source? Frequently, I find myself investigating the catalogues of Record Offices to work out where the original has come from and supply the missing items. But is it cheating if I put those deduced values into the citation of the on-line derivative?

All comments gratefully received!