When to Use Shortened Form of a Citation

Mike, the key thought in creating short forms for censuses is caution. Census citations are detail intensive; leaving out any one of those details could be regrettable. EE's examples for citing the originals or images in "short form" do drop the identification of the microfilm publication and roll number, and the archive where originals are located (details that need appear only in the first full reference note). They also abbreviate common words, but the rest of the census detail almost always needs to be repeated.

Alas, BBailey,  no one can nail jello to a tree. Nor can we make a hard, fast, rule about short-form citations that someone can apply anytime, anywhere. As researchers, we always have to appraise each situation and choose the best course of action.

The issue you’ve raised is one that needs to be dealt with in two parts. Let’s put each in a separate message.

ISSUE 1: 

You and Mike have asked for a rule to apply when citing censuses. However, neither of you have defined the most critical issue: Where are you using the citation?   Whether to use a full citation or a shortened citation is a decision we make at the time that we do a piece of writing. For example:

• If we are writing a book, we may want to do a full citation the first time a source is cited in each chapter. Otherwise, depending upon whether we include a bibliography (aka source list) or how complex that bibliography is, readers might have to comb through back notes for dozens or hundreds of pages to identify the source fully. A good way to get a feel for handling this situation is to study the citation practices followed in books published by university presses.
• If we are writing an article for a journal, or a biographical sketch of an individual, then the long vs. short debate seems simple. But, of course, census citations are seldom simple—which is why you've asked the question and why we're dealing with censuses, specifically, in a separate message.

Your “lumper vs. splitter” analogy—an argument that rages among users of relational databases—suggests that you are not actually thinking about writing but are in the data-entry stage. If so, the one hard and fast rule here is a simple one: data entry is not the time to make a long-form vs. short-form decision.

Yes, when we use a relational database software,  at the point that we enter the identification of a new source, the software does prompt us to create a short-form citation. But, from that point on, we don’t make a choice between a full reference note or a short citation.  It’s not an issue until we tell the database to print a report. At that point, the database itself makes the decision based upon what portions of the data we choose to include in the report.

If you are using a relational database, EE has to ask two further questions:

1. Are you asking how to create a shortened form that your database can then consistently apply to its automated reports?
2. Or, do you follow the better practice of moving your “automated reports” into word-processing software to rewrite or edit it in accordance with standards of good writing?

In short: does your question center upon the data-entry process or the writing process?

ISSUE 2:

For this, let’s go back to Mike’s original posting and look at the three scenarios he presents:

• different census year, different county (same state), different family, or
• different census year, different family (same county/state), or
• different census year only, (i.e. same county, same family)

Mike then adds:

>It would seem that #1 would require full citation for each footnote of this type. But, what about examples #2 and #3?  Do these still require full citations; or can I use a shortened version.  

EE will answer these two questions with a question:  If we do a full citation for one census year and then cite the same family in a different year, how could we get by with using a short citation to the different year, when each census is a different source?  (The 1880 census of, say, Custer County, Colorado, has one particular set of identifying details. The 1900 census of Custer has a different set of identifying details. We have to use a different microfilm publication for each. Each year appears on a different roll number. The enumeration district names or numbers are likely to be different. &c &c &c.)

Mike, then wrote:

>The short form examples of EE tell me how to format the citation, but they always [use] a shortened version of the same (but full) citation.

Yes, they do because, of course, that’s what a short-form citation does.

EE’s examples do not address the situations covered by the three bulleted scenarios because those three scenarios all involve three different sources. Each one, the first time it’s cited, has to be cited in full. Each one, if used again, can be shortened. EE’s census chapter (pp. 235–308) demonstrates how to do that for many different types of censuses—U.S. and international. But each census is a different source.

BBailey also asked:

>Lets take, for example, the 1870 U.S. census for Tarrant County, Texas.  If your research cites five households within the county in that census, do you use a full footnote for the first instance of each household?  Or just one full footnote per county per census year?  

Again, the first consideration is where are we creating the citation. For database entries, we use one practice. In the writing stage, our practice might differ according to what kind of product we are producing.

To create an example, let's say that (a) we are writing an article or a research report; and (b) we are writing about three individuals who migrated into Tarrant County before the Civil War. A number of scenarios then present themselves:

1. All three are enumerated on the same page.
2. All three are in the same precinct, same post office, but on different pages.
3. All three are in different precincts, but same post office.
4. All three are in different precincts and different postal districts.

A full citation to just one person might be this:

      1870 U.S. census, Tarrant County, Texas, population schedule, Precinct 1, Post Office: Fort Worth, p. 537 (stamped), dwelling 937, family 939, E. C. Bradford; National Archives microfilm publication M593, roll 1605.

A shortened citation to that same person might be this:

     1870 U.S. census, Tarrant Co., Tex., pop. sch., Precinct 1, P.O. Fort Worth, p. 537 (stamped), dwell. 937, fam. 939, E. C. Bradford.

If, say, the second time we cited this census, our text stated that Bradford lived next door to H. C. Childress. In that case, our shortened citation to that census we have already cited in full might be this:

     1870 U.S. census, Tarrant Co., Tex., pop. sch., Precinct 1, P.O. Fort Worth, p. 537 (stamped), dwell. 937, fam. 939, E. C. Bradford; and dwell. 938, fam. 940, H. C. Childress.

Or, let's say, the second time we need to cite this census, we are not talking about Bradford at all. We are only talking about Childress. In that case, we could also use the shortened citation:

     1870 U.S. census, Tarrant Co., Tex., pop. sch., Precinct 1, P.O. Fort Worth, p. 537 (stamped), dwell. 938, fam. 940, H. C. Childress.

If our next reference to the census switches to a different page, we can still use the short form, repeating every element up to the page number, then filling out the rest of the short form with the specific details.

If our next reference to the census switches to a different post office, then we could use the short form up to the point of the post office identification. Then we would complete the short form citation with the specific details needed.

If our next reference to the census switches to a different precinct, then we could use the short form up to the point of the precinct's I.D., then we would complete the short form citation with the specific details needed.

If, however, this were a year in which Tarrant County's returns are split across two different microfilm rolls, and our persons-of-interest appeared on different rolls, then a short-form citation to one roll would not work for citing someone on a different roll.

As with all citations, we have to thoughtfully consider the situation involved and decide exactly what elements will be needed in each case.

BBailey: You're right. EE  does try to stay out of the weed patches that bear the names of specific software. But let's follow your train of thought here to the next station--or the next two of them.

1. If the date that we access a digital image of a census page is considered a part of the "master source" in Whatever Software, then we would be creating a new master source every time we go online to re-use that census.  Yes?
2. If the online provider is considered a part of the "master source," then we would again end up with multiple master sources for the same census, depending upon the provider used. No?
    

BBailey: No, EE would not recommend creating a master source for every census household, every entry in register, or every obituary in some online database. However, this being New Year's Eve, EE will let you mull the problem while we make merry elsewhere. In the next couple of days, we'll use this as fodder for the daily tip at EE's Facebook page.

BBailey:

One further question: Have you tried the approach used for the 1850 Source List Entry at EE 6.20 and 6.25?

Thanks for clarifying BBailey. We love our software, but we do get frustrated with their wrinkles!