Request for Style: Genealogy

I have just begun using Zotero and find it very useful for genealogy research - especially the plug-in for Word that allows for footnotes and bibliography.

The standard for citing evidence in genealogy is contained in a book by Elizabeth Shown Mills called "Evidence Explained - Citing History Sources from Artifacts to Cyberspace."

It would serve a large group of researchers if you were to include this style as an option.

    note that ESM essentially is Chicago Manual of Style with additional information on what to do with item types that aren't specified in CMoS.
  • Thank you, I didn't know that, and it gives me the educated choice.
  • Please see the following comment concerning the ESM citation style:
  • Since Zotero can't even theoretically follow the ESM style precisely (because she has something like 100 item types) I would suggest just calling it Chicago Manual of Style - Genealogy and stay away from that issue.

    Genealogist should probably find themselves a better guru than someone who puts her profits (which I assume are already rather substantial) before the benefit of researchers in her field, but that's up to you guys.
  • On 1 January 2012, adamsmith wrote in two threads: "ESM essentially is Chicago Manual of Style with additional information on what to do with item types that aren't specified in CMOS. ... I would suggest just calling it Chicago Manual of Style-Genealogy."

    Adam, this assertion is a serious misunderstanding that needs correcting. Evidence Explained is *not* 'essentially CMOS with additional information.' Yes, Evidence-style citations to *published* materials and academic papers do follow the basic citation principles we learned as students, whether we learned them as CMOS, MLA or whatever—but not CMOS’s scientific style. EE uses the reference-note format that is essential to most work in original documents and common to many different style guides.

    What you refer to as ‘additional information’ constitutes the bulk of EE's 885 pages, as well as the very reason EE exists. If CMOS (or MLA or APA, etc.) covered the range of original and complex records used by historical researchers — or if their citations supported the degree of precision needed to understand the quality of the source being cited as an authority — then EE would not need to exist. Indeed, if any of the older guides came anywhere close to filling the needs of those who work daily in original historical records, then Zotero users would not continually ask Zotero to consider EE.

    The homepage at offers a more detailed explanation of the differences between EE and the other citation guides we all know and respect. If you have time, a quick scan of the chapters list will show how far EE goes beyond the basic sources other manuals cover. Clicking on the links provided for each chapter will give you a contents list for that chapter — a quick survey of the breadth of material we use and the quirks many of them involve. It is these quirks, of course, that make our needs a challenge for citation-generators.

    If you or other members of Zotero’s team would like a digital download of EE so that you can better understand our concerns, please email
  • Thanks for coming back to this discussion & sorry for the harsh words above, which were in reaction to a misunderstanding of the actual situation as presented in the other thread that you've since cleared up.

    I really don't see how anything I say indicates I'm misunderstanding this. I say above that EE contains ~100 different item types (and that's probably an understatement depending on how you count), so it's not like I'm unaware of the breadth of items you cover.

    The Chicago Manual's footnotes style is a reference-note style and looking at your examples, you do follow the basic logic of CMoS (things like the use of italics, periods in the source list vs. commas in the note, long notes and short notes etc. - these are all very close to CMoS and very different from e.g. APA or MLA. Nothing wrong with that, of course, it's a natural choice for historical research) and then you provide guidelines on how to handle a wealth of different sources within that general framework. Quite clearly, many people working in genealogy are finding that useful, so that statement isn't taking away anything from your book.

    But while
    a) it's probably possible to improve some item handling and, perhaps more importantly, provide some guidelines on how to enter data into Zotero,
    b) Zotero will implement some improvements and additional fields for handling archival resources, likely in version 4.2,

    I don't see how Zotero will ever be able to accommodate the detail and nuance without completely overloading its interface, item types, and citation styles.
  • adamsmith,

    I definitely agree with your final statement: it would be difficult for Zotero to "accommodate" the breadth of record types and details needed for each type, without "completely overloading" its current interface.

    In the meanwhile, I also appreciate your recognizing that EE is not 'essentially CMOS with additional information.' Across 35 years of publishing in academic journals and at university presses, I've had the pleasure of tailoring citations to most styles that use reference-note format. That experience — as well as my work with numerous university press editors who have needed help in structuring citations to complex records used by other historians — has left me keenly aware of the comparative differences. It would not be fair to either CMOS or EE to identify EE as “Chicago Manual of Style–Genealogy.”

    Should you have any questions in the future about citations to original historical material or their many derivative forms online, I would be happy to help in any way possible.

  • edited November 15, 2013
    It would not be fair to either CMOS or EE to identify EE as “Chicago Manual of Style–Genealogy."
    I think @adamsmith means that with the current state of Zotero and CSL, we can't really cover most of what makes EE different from CMoS. So if somebody attempts to make a CSL style for EE, it would still look a lot like CMoS, in which case the name "Chicago Manual of Style–Genealogy" might not be such a bad choice for such a style.

    Personally I think that adapting Zotero and CSL to fully support EE might not be worth the effort, and that it's probably better for users to focus on dedicated genealogy software instead.
  • Rintze, thank you for adding your perspective. However, a style called CMOS-Genealogy would be a misnomer from Day 1. As both a genealogist and a historian, I'm straining hard to think of any sources used by genealogists that historians do not also use. Good genealogists probably use them more often, but the records are the same. Of the 1000+ endnotes in my latest university-press book(a history, not a genealogy), there's not a single type that would be considered unique to genealogy.
  • edited February 18, 2017
    I've just come across this forum and others that reference Zotero and genealogy citation. AdamSmith, you've asked for some information about genealogy citation format and data elements that I don't think have been answered. I have attempted answer your questions in the following:

    Genealogy and Zotero (posted in parts due to length limitation)

    Genealogists access an enormous variety of original (primary) source documents as well as derived (secondary) sources in the course of their research. The original sources are often documents created by a government entity, but original sources might also include a personal letter, or a diary. Compiled sources include authored family histories and indexes of various types where original records have been collected, transcribed, sorted, listed, and published. Finally, genealogists also may cite traditional bibliographic materials such a reference books, maps, and journals.

    Citing these documents correctly is challenging, because each individual source may appear in a variety of formats. For example, a marriage record may be:

    1. The original marriage license signed by the officiant on the day of the wedding
    2. A certified copy obtained from a County Records office years later
    3. A microfilmed image found online,
    4. An entry in a book indexing the marriages of a county.

    Why is this important? Handling, copying, microfilming, transcribing, indexing, or publishing records may create errors in the new version of the sources. Therefore, genealogists want citations that indicate just how “original” the sources are. They prefer that a citation document whether the source that provided a data element is from an original certificate or from an online image of a microfilmed copy. This preference greatly complicates the citation process.

    Citations generally include at least three types of data:

    1. Information about the original source (description of the record type).
    2. Information about the compilation in which the original source has been collected and how it was made available to the researcher.
    3. Personal information about a person and/ or family that the source documents.
    Here are some of the issues that would need to be overcome if Zotero were to be made a useful tool for genealogists.

  • edited February 18, 2017
    Sources that genealogists use have multiple unique data elements that vary significantly from one source type to another. In the examples shown below, I’ve tried to bridge this gap by creating standard field labels that both record types can share, but it is difficult to do so because the events different record types document are so variable.

    Here are a few issues that arise with trying to create standard citation templates:

    RECORD NUMBER: In the examples below, note it takes 3 fields to cite what is, in effect, an original record number for the census record (Enumeration District; Sheet #; and Family #) because the census lists multiple families and people on one sheet. However, it takes only one number to cite the original record number for a death certificate (the state assigned death certificate number). Also note that for the census, the ED, Sheet # and Family # aren’t actually a record number, per se, but are the only data items that can be used to create a virtual record number.

    VARIABLE NEED FOR DATA FIELDS: Some sources, such as the Federal Census, are so familiar and standardized that less information is needed to create a citation. For the census, most know that Federal Census records originate with the original records microfilmed many years ago at NARA (National Archives and Records Administration). These microfilm images are available in a variety of formats and are used by a variety of content providers. So technically, a lot of the NARA information is superfluous.

    By contrast, a city, county, or state may issue a death certificate. It may also be an original copy that the researcher holds, or it may be an electronic copy have been initially archived by any of these jurisdictions, so for precision, fields need to be included to allow this information to be documented—but these details are unnecessary for a census citation.

    NUMBER OF CITATION TEMPLATES: There are several hundred citation formats in the standard genealogy citation reference, Evidence Explained (EE) authored by Elizabeth Shown Mills. A significant number of these are variations of a master format. For instance, EE offers about 13 variations for the citation of US Federal Census records. These cover the various ways in which the census can be accessed (i.e.: online images, CD-ROM, microfilm, etc.) and to vary what data element in the citation is being emphasized (i.e., member of a household, vs. the head of household).

    EXAMPLE 1: To cite a US Census Record, the following data would need to be captured. (Note that I have arbitrarily created names for data field labels.)

    DATA FIELD LABEL.........................DATA
    RECORD TYPE (orig)......................United States Census
    RECORD TYPE sub (orig)................Population schedule
    SOURCE (orig)................................Federal Census Records, 1900
    SOURCE REPOSITORY (orig)..........NARA
    SOURCE LOCATION (orig)..............Washington, DC
    COMPILATION................................NARA Microfilm
    COMPILATION sub a.......................Film T623
    COMPILATION sub b.......................Roll 1219
    COMPILATION TYPE.......................Microfilm
    DATABASE (derivative)...................United States Census, 1900
    DATABASE SOURCE (derivative).....FamilySearch
    DATABASE TYPE (derivative)..........Online database with images
    EVENT TYPE...................................Decennial census
    EVENT COUNTRY...........................United States
    EVENT STATE.................................North Carolina
    EVENT COUNTY.............................Surry
    EVENT LOCALE..............................Pilot Township
    RECORD # (orig)............................Enumeration Dist 114
    ----RECORD # sub a (orig)...............SHEET 10A
    ----RECORD # sub b (orig)...............Dwelling 168
    ----RECORD # sub c (orig)...............Family 171
    DATE EVENT..................................1900
    DATE RECORD...............................1900
    NAME-person of interest................John Doe
    NAME-informant............................ - - -
    NAME-head of household..............William H. Doe
    DATE ACCESSED..........................17 February 2017
    URL......https: // /61903/1: 1: MS17-KSZ

    The citation this data would generate might look like this:

    1890 Federal Census, Surry County, North Carolina, population schedule, Pilot Township, enumeration district (ED) 114, Sheet 10A, Family 171, John Doe, Household of William H. Doe; database with images, FamilySearch ( accessed 10 October 2016); citing NARA microfilm T623, roll 1219.

    EXAMPLE 2: To cite a North Carolina death certificate, the following data would need to be captured:

    DATA FIELD LABEL.........................DATA
    RECORD TYPE (orig)......................Death certificate
    RECORD TYPE sub (orig)............... - - -
    SOURCE (orig)...............................Department of Vital Statistics
    SOURCE REPOSITORY (orig).........State Dept Archives & History
    SOURCE LOCATION (orig)..............Raleigh NC
    COMPILATION...............................NC death certificates, 1931-1994
    COMPILATION sub a......................Mt. Airy, Surry, NC, Deaths vol. 36B
    COMPILATION sub b...................... - - -
    COMPILATION TYPE......................Microfilm of original records
    DATABASE (derivative)..................NC Deaths, 1931-1994
    DATABASE SOURCE (derivative)....FamilySearch
    DATABASE TYPE (derivative)..........Online database with images
    EVENT TYPE..................................Death
    EVENT COUNTRY...........................United States
    EVENT STATE................................NC
    EVENT COUNTY.............................Surry
    EVENT LOCALE.............................Mount Airy
    RECORD # (orig)............................DCN 36619
    ----RECORD # sub a (orig)............ - - -
    ----RECORD # sub b (orig)............ - - -
    DATE EVENT..................................04 Oct 1969
    DATE RECORD..............................04 Oct 1969
    NAME-person of interest...............John Doe
    NAME-informant...........................Sarah Williams Doe
    NAME-head of household.............. - - -
    DATE ACCESSED..........................8 December 2014
    URL......https: // /61903/1: 1: FPSB-HZ6

    The citation this data would generate might look like this:

    NC Death Certificate; Surry County, North Carolina, Department of Vital Statistics, State Dept. of Archives & History, Raleigh NC; Death certificate #: 36619, John Doe, (4 October 1969); NC; Deaths, 1931-1994, database with images, Family Search (https: // /61903/1: 1: FPSB-HZ6), citing NC death certificates, 1931-1994.

  • edited February 18, 2017
    The ideal would be for Zotero to gather data when the researcher is looking at an electronic copy of a record on one of any number of genealogy websites. I don’t know if uniform metadata standards exist for online genealogy records, but I tend to think that the large online genealogy sites (i.e., FamilySearch, Fold3, etc.) have some type of metadata functionality. They have, to some degree, automated the screening, analysis and linking of data from records to individual. Whether or not there is an industry-wide standard is unknown to me.

    However, even if there are metadata standards, I doubt that other sites, such as state vital records “look up” sites do.

    Therefore, automatic collection of essential data for citations might be extremely limited and a Zotero genealogy application might be limited with some sites requiring manual data entry. This would greatly reduce its effectiveness.


    Zotero is designed to integrate with MS Word. Genealogists use Word, but this is not the primary place where data and citations are stored. Most genealogists archive their data in one of the various commercial genealogy software programs (FamilyTree Maker; RootsMagic, Legacy Family Tree, Ancestral Quest, etc.) There are well over a dozen.

    Most of these programs have some utility for citations—but they can be clunky and often rely upon manual transfer and re-entry of data.

    To be most effective, Zotero would have to be able to interface with one or more commercial genealogy software programs.


    Though EE has gone a long way to standardizing citation, there is still wide variation in opinions about how citations should be formatted, what data elements should be included, and how they should be arranged.

    As noted above, FamilyTree Maker has a different standard, as does Roots Magic, as does FamilySearch. They may (or may not) follow Evidence Explained generally, but there are numerous exceptions.

    It’s also very likely that people would not only disagree with the names of the data labels I have created, but more importantly what particular data element might fit with a specific data label. These wide variations in opinion will complicate coming up with a Zotero standard.
  • edited February 18, 2017

    Finally, there is a guy Jeff Marca PhD who has done some tremendous thinking about genealogy citation. He's published his ideas and works on his web page "Simple Citations: Making Life Easier for Family Historians." His approach has reduced the number of citation templates to four (Master, Census, Traditional, and Non-traditional). He created add ins for use with RootsMagic and has made them available for any interested user free of charge.

    His formats aren't as detailed as EE, but I don't think he intended them to be. they accomplish the basics of citation with great efficiency. His strategy leaves out data about the databases consulted that is standard, widely known, or can be readily found on the web if the more streamlined format of his citation isn't sufficient. If there is still interest in doing this for genealogy, I would suggest that you involve Jeff Marca in the discussion.

    The examples I posted above would look like this in Marca's format:

    United States, Bureau of the Census (1 June 1900). 1900 U.S. Census: William H. Doe family; Sheet 10-A. John Doe. Pilot Township, Surry, NC; enumeration district 114, 171. FamilySearch (https: // /61903/1: 1: MS17-KSZ: accessed 10 October 2016).

    North Carolina, Department of Vital Statistics (4 October 1969). Death certificate: John Doe; Mount Airy, Surry, NC. FamilySearch; NC Deaths, 1931-1994 (https: // /61903/1: 1: FPSB-HZ6: accessed 8 December 2014). Certificate# 36619.

    These are just a few thoughts off the top of my head. I would be happy to answer any additional questions or participate in any users group you might put together.
  • edited February 18, 2017
    Thanks -- for the time being, this doesn't really change my assessment above, unfortunately. Zotero is not going to create 20 different new fields and something close to as many item types to accomodate this. Maybe this could be something to reconsider once custom fields and item types are introduced (my understanding is that that's a medium-term goal), which would allow for something like a Genealogy set/template. That would still leave import into those custom fields, which likely would require an enormous amount of work and also leave the question of citations open -- I've never quite understood how custom fields would translate into citation styles.

    Working with historians and archivist, my impression is that for most _citation_ purposes, a description of a reference by author, date, title, which archive/database, which collection, exact location of record is sufficient. Zotero is mostly able to accomodate that (and will be better set up to do that once we have an archival collection variable), but in the near future that's likely going to be all there is.
  • (this was posted prior to the Jeff Marca comment -- fewer number of templates does make things easier, but still likely doesn't get us anywhere close to a one-to-one adoption).
  • Adamsmith:

    Are you with Zotero?

    If so, I would welcome the chance to discuss this with you. I THINK something could be done if we only do some out of the box thinking. If it's possible to talk on the phone, please let me know how we can set up the conversation.
  • I don't work for Zotero but I do co-run the citation style system it uses and Zotero consults with me and my colleagues there on new item types and fields as they're needed. Both because of my limited time (I have a full time academic job) and because I think it's important to have such discussions in a venue where other people can chime in -- both for transparency and because of the additional expertise that draws in -- doing this over the phone is not an option for me, sorry.
  • edited February 19, 2017
    Maybe I can add something here that might help. My apologies for the stream-of-consiousness form.

    The birth and death record examples gordongrant provides are both data _and_ metadata. Zotero primarily is used to capture, edit, store, and assist with the proper citation of metadata. That John Doe lived at a certain address on a certain date with named people in the household is data. The source of the data can be described as having an institutional author, the federal government; a title, federal census records of a particular year and place; a locator, film roll and frame numbers, and a publisher, in this case the publisher is the website company. Given the information I mentioned, anyone could go to the source to verify that the citation is accurate. Except that, perhaps, some publishers imaging quality may be better than others; an image of a census page is an image. This isn't like a 3rd edition of a book where there may be both revisions of the text and pagination. In general, proper citation only requires enough metadata to: 1) give credit to the creator and publisher of the item and 2) allow anyone to locate the item. Genealogy may be unique in that its conventions require citation details far beyond what is needed to identify the source. I can sympathize with that concept and have had intensive personal experience working genealogy data. A non-genealogy example is the requirement of some citation styles to omit the issue number when citing a journal article if pagination doesn't restart with each issue. (I hate that because it makes me spend more time locating the article. That said, I can nonetheless find the article.)

    I no longer work with genealogy data. When I did, I used specialty software (The Master Genealogist) with a database that allowed capture, storage, and citation of sources to the level of detail you desire. That was many years ago and I don't know if the software is still maintained.

    There is a fork of Zotero (along with an extension of CSL) for users who require multilingual citations and references to legal and court documents. This fork facilitates proper citation by authors of law journal articles and legal documents. Maybe, someday many of the specialty aspects of this Juris-M (nee multi-lingual Zotero) will be incorporated into mainstream Zotero. Maybe someday, there will be a fork of Citation Style Language -- the xml-based "language" that supports the citation fields and item types that are used by Zotero, Papers, Mandeley and other bibliographic software. Maybe someone will make the effort to extend Zotero to capture and store metadata desired by North American genealogists. Even then, it seems to me, that would only fill a small part of genealogists' needs. Genealogy requires storing that Jane Doe was born on a certain date and location to named parents, that she wed Rich Roe on a date and at a place, etc. There should be a (metadata) source attached to each of those facts (data points). Zotero is designed for working with the metadata and not the data itself. Zotero is for citing the source of an idea, assertion, or quotation but not to store the actual assertion or quotation itself. edit-- Well, maybe the quotes may be stored in a Zotero note but Zotero isn't really set-up to insert that stored quote into a narrative along with the source citation.

  • While I'm still hopeful someday there will be a professional fork, I'm feeling like in the meantime CMOS would be sufficient for my family genealogy work. I would be interested in joining a Zotero group that is sharing specific examples of how folks have mapped Zotero's existing fields to genealogical records (census, death certificates etc.).
  • There is a book and discussion group for those of us using Zotero for Genealogy

    I'm still in early days of deciding how to do all of this myself, and am using CMOS17 myself. I plan to modify the CMOS style a bit so that some fields show up in the citations for all item types, but although XML is orderly and easy to understand, it's far from easy to figure out what to modify.

    At any rate, thought I would point you in the direction of a genealogy-oriented group.
  • It is rather remarkable how complex citations for genealogy have become. Elizabeth Shown Mills says that she can't think of any sources that are exclusive to genealogy rather than relevant to history as well. And she's surely right. So why does she make it so complicated when most historians manage with far simpler models.

    The practice described in Evidence Explained is very thorough, extremely complex and time consuming, sometimes inconsistent, and quite unnecessary for most purposes. After all, a UK census enumeration schedule, for example, can be found by anyone from a few basic bits of information including year, class/piece, folio and page. Professional historians have been doing this for decades. It's unnecessary hand-holding to tell the readers that they can find it at such and such a place on Ancestry, or Findmypast, or The Genealogist, or wherever.

    Through and accurate citations are vital, but there is such a thing as too much unnecessary information that only clouds the issue.

    I thank gordongrant for the reference to Jeff Marca's site which brings some much-needed common sense to this question. And, of course, to Jeff himself.
  • As I know little about the inner workings of Zotero, but have an interest in seeing it used in genealogical writing, I'd like to as a question.

    Does Zotero have the capability to store a complete citation reference note as a single field. Perhaps the templating capability of Zotero would allow this. If so; one could create an Evidence Explained style/type citation for a document and store it, along with one's other material in Zotero. I think this would result in the same EE text being used in a bibliography, but this is better than having to edit every exported citation and dealing with assembling a bibliography in the target document. Much of what is needed in an EE-style bibliographic entry appears to also be present in the citation reference note.
  • edited December 7, 2023
    Thank you gordongrant and thebookbloke for your kind words. I just came across this thread and, while your posts are not recent, still want to mention that a book of my approach, "Simple Citations for Genealogical Sources" is currently in the process of being published and will be available for sale within a few weeks. This has been a long-term goal for me as the feedback I've received from those who have used it has been extremely favorable.

    The "Simple Citations" model has long been defined on the web site at: and the book will provide full details on how the system was developed, as well as explain the theory behind it.

    The book consists of just nine chapters, plus appendices. In keeping with my belief that we should keep things simple; the book is not large and contains approximately 160 pages. Here’s an abbreviated Table of Contents:

    Table of Contents
    Chapter 1: Quick Start Guide
    Chapter 2: Citations
    Chapter 3: Other Citation Systems
    Chapter 4: The Simple Citations Model
    Chapter 5: Syntax Conventions
    Chapter 6: Practical Guide to Using the Templates
    Chapter 7: Using the Census Template
    Chapter 8: Using the Traditional Template
    Chapter 9: Using the Non-Traditional Template

    Appendix A: Entering Data into RootsMagic 7 and RootsMagic 9
    Appendix B: Formatting Issues with Software Applications
    Appendix C: Entering Repository Data
    Appendix D: Entering DNA Data

    Examples for use with both FamilyHistorian 7 and RootsMagic are provided throughout the book.

    Again, thanks for your kind words.

    Jeff La Marca, Ph.D.
Sign In or Register to comment.