Do You "Just Trust" Citations Offered by Digital Providers?

 
 
 

 

25 June 2014

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.

The census enumeration maps offered by one of our most-valued providers of digitized records suggests that we use the following citation:

"United States Enumeration District Maps for the Twelfth through the Sixteenth US Censuses, 1900-1940," images, FamilySearch (https://familysearch.org/pal:/MM9.3.1/TH-1942-34868-10041-54?cc=2329948&wc=92V7-16X:1077257201 : accessed 23 Jun 2014), Roll 24, Louisiana, Acadia-Rapides 1900-1940 > image 667 of 881; citing NARA microfilm publication M1882, National Archives and Records Administration, Washington, D.C.

EE would cite that image in a significantly different way:

Enumeration District Maps for the Twelfth through Sixteenth Censuses of the United States, 1900–1940, Microfilm Publication A3378 (Washington, D.C.: National Archives, 2003), roll 24, Acadia–Rapides, Map: “Natchitoches Parish, Louisiana, 1900," frame number not visible on digital image at “United States Enumeration District Maps for the Twelfth through the Sixteenth US Censuses, 1900-1940," FamilySearch (https://familysearch.org/pal:/MM9.3.1/TH-1942-34868-10041-54?cc=2329948&wc=92V7-16X:1077257201 : accessed 23 Jun 2014), image 667.

What differences do you see—and why do you think they do or don't matter?

Jade
Jade's picture
Citation errors

At first glance I see you caught a wrong micropublication number and that the original citation failed to identify the specific document cited.

While I sometimes catch wrong microfilm numbers on the part of FamilySearch, it is very common there for nature or title of document not to be described in citations.  There is also a strong trend there to give wrong titles for volumes and even of collections.

The determined and careful researcher must check everything.

 

Jade

EE
EE's picture
You caught two critical

You caught two critical points, Jade, and your policy of double-checking, regardless of who the provider is, is exactly on target. As I noted in today's follow-up post (http://bit.ly/1lSlhY6) this same situation exists with many providers. Citations rarely rank high on a provider's business plan. As a result, providers who do offer citations often have them done by employees with limited research experience, or even interns.

We really, really, really appreciate their recognition of the fact that source-identification is needed. Even if some of their citations have problems, they are conveying to new researchers the fact that where-we-get-our-stuff does matter. Still, serious researchers also recognize the value of thinking through each citation for themselves.

The Editor

stanm
stanm's picture
Some differences that matter ...

The missing publication title and incorrect publication number are important. I tried a Google search on "NARA M1882," and the match was for Schedules of the 1935 Special Censuses of Puerto Rico: The Agricultural Schedules, 1935. This is certainly a cause for confusion!

I'm curious as to why a semi-colon doesn't separate the two parts of this layered citation.

EE
EE's picture
Stanm, you clearly have an

Stanm, you clearly have an eye for detail.  No one else caught that "missing" semi-colon between two parts of a layered citation in the model that EE gave:

Enumeration District Maps for the Twelfth through Sixteenth Censuses of the United States, 1900–1940, Microfilm Publication A3378 (Washington, D.C.: National Archives, 2003), roll 24, Acadia–Rapides, Map: “Natchitoches Parish, Louisiana, 1900," frame number not visible on digital image at “United States Enumeration District Maps for the Twelfth through the Sixteenth US Censuses, 1900-1940," FamilySearch (https://familysearch.org/pal:/MM9.3.1/TH-1942-34868-10041-54?cc=2329948&wc=92V7-16X:1077257201 : accessed 23 Jun 2014), image 667.

The issue here harks back to the mantra that citation is an art, not a science. Here, the citation needed to say exactly what I've highlighted above: the frame number is not visible on the digital image. Having said that, if we then put the commonplace semicolon and then put "digital image" again in the second layer as an identifier for what comes thereafter, then we'd be creating unnecessary redundancy:

Enumeration District Maps for the Twelfth through Sixteenth Censuses of the United States, 1900–1940, Microfilm Publication A3378 (Washington, D.C.: National Archives, 2003), roll 24, Acadia–Rapides, Map: “Natchitoches Parish, Louisiana, 1900," frame number not visible on digital image; digital image, "United States Enumeration District Maps for the Twelfth through the Sixteenth US Censuses, 1900-1940," FamilySearch (https://familysearch.org/pal:/MM9.3.1/TH-1942-34868-10041-54?cc=2329948&wc=92V7-16X:1077257201 : accessed 23 Jun 2014), image 667.

If, to avoid redundancy, our first layer were to simply say "frame number not visible," then we would be implying that it was the NARA film itself that did not have a visible frame number.

Would you handle this differently?

The Editor

stanm
stanm's picture
Thank you Elizabeth for the

Thank you Elizabeth for the explanation, that "citation is an art" and duplicating "digital image" makes the citation unpleasant to read. I was tempted to use this in place of your highlighted phrase: "unnumbered frames; digital image," but that would imply that the frames of the original microfilm were unnumbered which we can't say for sure using only the online source.

EE
EE's picture
You're right, stanm. We can't

You're right, stanm. We can't!

The Editor

DebNC
DebNC's picture
Differences in Citation
  1. The treatment of the title is different in that you italicized it and put the "of the United States" last, putting emphasis on the fact that it is a map;
  2. There are two different microfilm publications cited: theirs states M1882, Roll 24; yours states A3378, Roll 24, and you include where the microfilm is located;
  3. They put the web address following the title, while you put it at the end of your citation;
  4. They split up the microfilm number and the image number and include image number/of total number of images, while you keep it together and only include the image number you are describing;
  5. They end with the repository that holds the microfilm, and you put that information between the microfilm publication number and the roll number.

I have one question on how you described the web address for this document: in an earlier post, I had thought you mentioned that it is unnecessary to record the complete web address, but that we should only post, Family Search (https://familysearch.org: accessed 23 June 2014).

EE
EE's picture
Good observations, Debra.

Good observations, Debra.

Re the "treatment of the title," the basic issue here is that two different titles are involved: (1) the title that NARA gave to its own microfilm publication and (2) the title that the provider gave to its database.

Re, the web address: EE's reason for placing it at the end is that we have a two-layered citation. First, we need to cite the record itself. That is a microfilm publication that has no web address. Having cited that in full, then it's time to cite the medium through which we viewed this record—the website.

Re the splitting of the "microfilm number and image number," the provider is correct in separating the two because they are peas and apples. The image number refers to the image in their database, not the image in the microfilm. In Layer 1 of the citation, the microfilm publication, we do need to cite a frame number; but we can't because the provider's image appears to be cropped and doesn't show a frame number. (This is one the issues we discussed in today's follow-up to yesterday's question; see http://bit.ly/1lSlhY6). In Layer 2, our citation of the website that provides a digital image of the original publication, citing the image number helps to locate their image at their site.

Re their placement of the identification of "the repository that holds the microfilm," this is another issue covered in today's post; but you've raised an angle not covered there. The provider's citation of their source-of-the-source isn't actually to a repository. If we were citing an original record that existed in only one place, we would cite the repository. In this case, however, the repository that owns this set of records had it microfilmed and then published the film. Therefore, the digital provider is referring to the film. But do see our broader comments in today's post about citing source-of-the-source.

The Editor

EE
EE's picture
Long URLs

Debra, in Message 5, wrote:

"I have one question on how you described the web address for this document: in an earlier post, I had thought you mentioned that it is unnecessary to record the complete web address, but that we should only post, Family Search (https://familysearch.org: accessed 23 June 2014)."

Debra, this is a choice that's generally up to each person to decide for themselves. The pros and cons might be summarized this way:

Pro: FamilySearch has labored mightily to create stable URLs that will take users to the exact image they desire.

Con: If anyone makes a typo in retyping or doing a cut-and-paste of that URL—if they accidentally delete even one character—then the URL becomes unworkable and users are often left with no obvious way to find what's needed.

The alternative is to cite the base URL for a provider and then provide the path by which the image can be found. The added advantage, in citations to complex original records of the type we often see at FamilySearch, is that recording the path also means we are recording useful information about the organization of that set of records.

The Editor

ACProctor
ACProctor's picture
Are Digital Providers being demonised?

I don't want to defend the citations that the digital providers offer since I (try to) craft my own. The level of detail is known to be lacking, and some don't offer any help at all which I consider to be a worse problem.

It's not just the big providers either since there are many small sites offering images and trancriptions created by volunteers. I can't think of a single one of these that offer citation help, and yet I am truly grateful for all the work they've done to make the data available.

Surely, this is not a problem specific to digital providers. Any citation may be deficient, and there is no magic spell for knowing whose can be trusted.

EE
EE's picture
Are Digital Providers being demonised?

ACProctor,

No. Demonization is not the issue. Nor do we consider critical analysis to be demonization. We greatly appreciate all providers who make an effort to include source citations and have, for years, worked with more than a few of them to advance this boon.

The issues (from our perspective) are these:

  1. As new researchers, we all learn by seeing as well as doing. We learn standards and practices by seeing what others provide. In this regard, digital providers are also providing immensely important education. They have, arguably, more power to teach than any conventional educator because their teaching-by-example reaches millions on a daily basis.
  2. Most new researchers whose academic training did not include historical research often say they have difficulty understanding what is essential to record and what is not essential. For them, it is much easier to do a copy-paste of whatever a provider offers. Eventually, all researchers begin finding conflicts in their evidence and are perplexed as to what the "facts" of a situation actually are. If the provider's suggested citation has captured all essential elements, then the researcher is equipped to analyze findings and decide what is the most trustworthy. If the provider's suggested citation has not adequately reflected the nature of the record, then the trusting user may be misled and will often stay perplexed, not understanding where the problem lies.
  3. Digital providers, whose engineers are not typically trained in historical research, also say that they have trouble understanding what is essential from our perspective. Over the years, in both working sessions and informal discussions between these engineers and "subject-matter experts," the first obvious thing is usually this: we speak different languages and see different needs. Our object is to help each other learn the parameters that each field must work within.
  4. When suggested citations are analyzed—whether they come from a digital provider, a library, an archive, or a researcher—we all learn from discussion.

The Editor

ACProctor
ACProctor's picture
I accept that some

I accept that some researchers may be unsure of what to provide when citing something from a digital provider, and so there's a reliance on what that provider offers for copy-and-paste. However,  my point was that this is a small part of a larger issue.

There are digital providers (big ones too, who I won't name) who not only offer no citations to copy but offer little or no information about the origin of their data either. I believe that provision of this information is more important, especially in those cases where the provider's citation is either incorrect or absent.

Further than this, there are many small local projects who transcribe historical records (e.g. parish records) and offer them for free in all manner of database and textual forms. These are great resources but they exhibt the same issues as those bigger digital providers. Rather than expect them to offer ready-made citations, though, provision of comprehensive information about the origin of the data should be a mandatory consideration for them.

As an instance showing the importance of this, there are collected transcriptions available for parish records from my home town/city. These are provided in a database format on CDs. They great but there is no connection declared with any original records, and no images whatsoever. The lack of images may be imposed by the local archive or the local parishes -- I am unsure -- but even if I travel back there to look at an original document, I have no concrete evidence of where and when the transcription was taken from. If there is a discrepancy then there's no way to report it back and have it confirmed. The same transcriptions are indexed by sites such as FamilySearch but they can offer no source information other than the fact that the information came from the local Family History Society.

Apologies for the long discourse but I just feel that the bigger issue is the leap from original documents to digital data -- which is now overwhelimg this field -- and that it isn't specific to the large commercial digital providers.

 

rraymond
rraymond's picture
Another problem with the FamilySearch citation

Elizabeth fixed another mistake that I don't see explicitly called out by other commenters. Not only did "the provider" mess up the format of the publication information, it neglected to include the publication date. (It left it off the collection page citation as well. https://familysearch.org/search/collection/2329948 .)

When you double-check a provider's citation, as Elizabeth wisely counsels, if a NARA microfilm publication date is missing, here's how to find it:

  1. Go to www.archives.gov
  2. Make your way to the web page where you can order microfilm. The path changes occasionally. Today:
    a.       Click on Shop Online (the brown box on the far right). A shortcut that works today is: http://www.archives.gov/shop/
    b.      On the right-hand side, part way down is “Buy Reproductions and Microfilm.” Underneath that title, click on Order Reproductions.
  3. On the menu bar along the top, click on Microfilm.
  4. Under “Search & Browse…” enter the publication number, such as A3378, in the search box. Click the Search button.
  5. Click on the search result. If the microfilm publication contains too many rolls, there may be multiple entries; click any one of them.
  6. There may be a PDF link on the right-hand side under the title “Publication Details.” If there isn’t, indicate “n.d.” in the citation. If there is, then click on “View Important Publication Details.”
  7. Open the PDF document.
  8. Scan the first several pages of the document for publication information. In the case of A3378, the first page is a typical title page and contains publication information at the bottom of the page. Many publications don’t indicate a date and you will need to specify “n.d.” in the citation.

---Robert Raymond
FamilySearch