La Réunion Parish and Civil Registrations Are Now Online
"Mastering Genealogical Proof" and French Genealogy - Part 7

"Mastering Genealogical Proof" and French Genealogy - Part 4 - Elizabeth Shown Mills Responds


U - Very Long climb

We are extremely grateful that Elizabeth Shown Mills, of great renown in just about every field of genealogy and author of "Evidence Explained", has taken the time out of her busy schedule to respond to our citation queries. We give her response here in full, with many thanks.


Hello, Anne, 

My apologies for the delay in answering your questions. ... I’ll intersperse comments amid your questions below . . .

1) Wouldn't it be better not to give English titles for French records, except perhaps in parentheses, if desired?

 Yes, indeed. That’s what EE recommends in Chapter 2’s basic instruction  (“Fundamentals of Citation”) at 2.23: copy the title exactly, then put an English-language translation in brackets.  EE also demonstrates that in over two dozen other sections (3.10–3.12, 6.52–6.54, 6.56–6.57, 7.16, 7.18, 7.37, 7.39–7.42, 7.44–7.45, 9.49–9.51, 9.54, 9.57, 11.61, 11.67, 11.69), using examples from France, Germany, Italy, Mexico, Norway, Spain, Sweden, and Switzerland.

However, EE does not pedantically say You must. Good researchers have differing opinions on this point and most style guides offer options. EE follows the “options” school—one especially needed in the field of genealogy where so many people use relational databases with built-in templates. They can, as you have suggested, use the free-form capability in their software to create the best citations possible, but then those citations rarely convert when they change programs.  

EE’s principle exception to the “quote the title exactly” rule is census records—the category you used as an example in your blog post. Again, this respects the preferences of the majority of English-language users who organize their source lists by categories and then list censuses generically by year and country/province/etc. (as illustrated at EE 6.3). If they identify censuses by the exact titles, then those censuses end up scattered throughout their bibliographies under names such as “Listes nominatives de recensement de population, 1836–1896,” “Riveli di beni e anime,” “Padrons de veins,” and a jillion other names they can never remember.

2) French records concerning the population, such as census returns or civil registrations are organized according to the following administrative hierarchy: Département, Arrondissement, Canton, Commune. This is slightly different for large cities such as Paris and Lyon, which have  arrondissments which function as communes. I do not understand why "Evidence Explained" omits the canton in some cases.

Can you cite the section in which EE omits the canton?   [It was section 9.49, see below] You mentioned this also in your blog of 25 November 2010, but I’m having trouble finding an example that omits it.  The “France” section of EE’s census chapter (6.52) carries two examples, with the following jurisdictions shown:

1856   Orne (department), Argentan (arrondissement), Mortrée (canton), La Bellière (commune), Quartier Tetre-aux-Allains

1836   La Mayenne (department), Laval (arrondissement), Évron (canton)  

 The latter example does not include commune and quartier, but when this record was retrieved that data was not available at the website that provided the image. The QuickCheck model on p. 241 diagrams the Mayenne example; again, there, the canton is included.

The addition of the unique INSEE code for the commune would eliminate almost every possibility for mistakes. This would allow for the shorter citation to give only the commune with the INSEE code and the Département.

I can’t agree that it is better to use just an INSEE code and department name (or the equivalent in other countries), rather than explicitly naming all the relevant jurisdictions.  The INSEE code would useful as a supplement (just as U.S. researchers might add zip codes to locales), but I cannot recommend it as a shorter substitute.  My reasons are these: 

1.       Any time we type a set of numbers, we increase the “possibility for mistakes.” Over three decades of editing manuscripts and fact-checking citations—over half of those years at a scholarly journal that published this country’s best researches and writers—I found far more errors in number sets than in any other aspects of citation.

2.       Researchers need to know the full jurisdictional trail, especially those who are not experts in a specific region.

3.       Codes and abbreviations, across time, are immensely unstable. As historical researchers, we are recording information for generations to come. Shortcuts that we take now will create problems for future users of our work

3) As you know, the recording of births or baptisms, marriages and deaths or burials in France were copied or often written anew by a clerk, or greffier; this copy was contemporary and was signed by the same people who signed the parish registration. This copy was held in local administrative offices. Both copies of parish registrations recorded before 1792 were placed under the authority of the state and ordered to be sent to the newly created Departmental Archives. …  Departmental Archives are now filming the clerk's copies and putting them on their websites.

Because of this increased access to duplicate sets of registrations, I would suggest that, in citing parish and civil registrations, no standard English or French term should be used. The French title of the particular register or film seen is what should be given. … To use a generic term will only create confusion. 

I agree, always and in every situation, that the exact title is preferable. On the other hand, the ideal is not always the reality. Amid EE’s range of examples, the ones that demonstrate citations to series, by generic names, are materials that we personally received from those repositories, whose staffers cited them for us in that generic fashion.

4) I would warn against using a website page as the website identification. The ultimate home page should be, I suggest, the website cited. At times, this may be a Departmental Archives website but it may also be a general Departmental website, with the archives section having a sub-page, along with sports, libraries, taxes, and other local administrative activities.

My response has to be: It depends. With many websites, it is impossible to locate a document by citing only the homepage or root. From there, we have two options: we might cite the path  to follow in order to drill down to the record, or we might cite the URL for the actual record—presuming it’s not a dynamic URL.  In the unstable world of the Internet, both of these options pose problems.  Site administrators frequently change the paths—or they change the URL. Either change can make an online record difficult to locate from a generic citation to the website’s home page. Some online providers are now promoting what they proclaim to be stable URLs—and insist that we do use the specific URL for that document—until, of course, they discover a reason why the URL has to be changed. 

It seems to me that the only safe course we have as users is to include every detail about the record and the site that might be needed to drill down to the record itself.  But that, of course, causes others to gripe about overlong citations.

“A Sourcing Suggestion – INSEE Codes,” The French Genealogy Blog, 25 November 2010, raised two other issues:


“As each department seems to source differently, [using their preferred styles] can be thorny. … Do we respect the sourcing style of the department or do we modify it – risking their wrath, perhaps – for the sake of consistency?”

This is an issue on which virtually every style guide ends up punting. EE addresses the issue in several places, particularly 3.2. There I give basically the same advice other style guides give (but in my own words, of course):

“Because each archive has its own preferences for citing its materials, you should ask each facility whether it has a citation guide it would like you to follow. If you choose to use each facility’s recommended style, you can expect to have considerable inconsistencies within your citations. If you prefer to use a consistent style, you will still benefit by studying the format recommended by the archive whose material you have used. From it, you will learn which pieces of information the archive needs you to cite in order to relocate the material.”

And, of course, another reality exists here: The staff member who is assigned to create these citations, at one or another archive, is often a low-level staffer with minimum research experience, little enthusiasm for the chore, and limited perspective about records and citation issues outside the holdings of their own repository.

As researchers and writers, we cannot please every entity, every person who helps us, or every person who uses what we produce. All we can do is to be as informed as possible, thoughtfully weigh the factors involved, and try not to create problems for others. But, of course, if we ask editors at any press, their answer will be: “You do what our guide says.” 


Evidence Explained, in “stating that ‘Worldwide, these records are more likely to be consulted via the Family History Library microfilm’ suggests giving the FHL microfilm number as well. We suspect that users are gradually switching to the immediacy and free use of the websites of the departmental archives … in preference to the long wait for the rented FHL films.”

Your observation is on the mark, of course. That shift is inevitable and desirable. However, the section from which you’ve quoted (EE 9.49) was one that specifically addressed the issue of FHL film. It was not meant to discount the value of using online material. The preceding section (EE 9.48) deals with online records, using an English example.

Realistically, it simply is not possible for any guide to address every type of media for every type of record for every country. EE offers over 100 examples and discussions for international records, but it cannot do justice to any country touched upon in the “International ” sections of each chapter.

Ideally, what is needed is a comprehensive style guide for each individual country—one that attempts to covers all types of records created in that country for use by historical researchers. One day that will happen. I suspect Anne Morddel would be a good person to make that happen for France.[*]

I notice that you’ve promised to update your readers when I responded to your questions. If you want to publish this as a Q & A, please feel free to do so. 

Kind regards,



And there you have it. Again, many thanks Ms. Shown Mills.


©2013 Anne Morddel

French Genealogy