Why citations in family history are more complicated than other citations.

Citations are widely used in academia, where they serve two purposes: they attribute ideas to their originatorsReputation is the currency of the academy, so it is hard to overstate the importance of attribution. and they allow those interested to locate the sources themselves. Their presence also serves rhetorical purpose, but that purpose is largely independent of the content of the citations themselves.

Family history research is largely outside of academia. The academy operates by having several researchers in partial competition with one another investigating the same subject, performing peer reviews, refuting incorrect claims, and advancing by building off of one another’s insights. Family history has more subjects than it has researchers, sidestepping that organization; and if any one subject becomes interesting to enough people that it is enters the domain of academia, we call it history instead of family history. There are academics who study family history, but most family history research is not performed by them.

Family history’s position as research that is performed outside of academia leads to multiple differences between its use of citations and that used in other disciplines. The following are some of those differences that I have noticed.

• Citation content is used to indicate the likelihood of errors in the information provided by a source. This means that every intermediary should be listed. Unlike in academia, most of those intermediaries will not be trusted, fund repositories accountable to the source authors and they will make mistakes. Even those that are of academic-database quality need to be identified to distinguish them from less-reliable intermediaries

• Individual family history sources often lack the funding to be properly maintained and tend to disappear with time. Hence, if I am aware of three copies I should share where all three are located, not just the most original one.

• Family history often cites huge documents such as censuses in order to discuss very specific information. Often the academic citation part selectors of volume, number, and page are inadequate for this purpose. Loose-leaf manuscripts, mislabeled pages, repeated row identifiers in tables or document serial numbers, and illegible or missing markings are all common enough to deserve orderly treatment.

• Many sources are not part of any publication or archive. Academic citation standards might list these with minimal information as “‍personal correspondence‍” with a name and date. Family history wants more than this: the context in which these ephemeral sources were obtained can do much to indicate their likely validity and replicability.

• Often a family historian can only access a derivative document, not the original. In that case, it is important to cite the creators of the derivative and whom they said were the creators of the original. There is not upper limit to these chain lengths: I might only have access to a translation of a translation of a transcription of a summary of a print of a woodcut of a manuscript.

None of these issues is unique to family history; any of them could, in principle, appear in any work. For example, I once cited in a computer science paper a card catalog entry for a manual for the use of a mechanical device intended to assist in the execution of an algorithm because the manual didn’t appear to actually be available and I couldn’t find any direct reference to the algorithm itself. But that was an exceptional occurrence so I simply described it in the text of the paper I was writing itself (and, at the request of the peer reviewers, removed most of the details before final publication). In family history, this sort of thing happens so frequently that trying to handle each case individually leads to disorder.

Various parties have tried to get a handle on family history citations. A few examples:

• Academic citations are digitized as key:value pairs, with keys like author, editor, title, pulbication date, and so on. I’ve spoken with people at several large software history companiesI have not been given permission to share which companies in particular who were parts of teams assigned to extend that model to family history citations; they report that after the teams passed a thousand distinct keys, they gave up the process as unmarketable.

• An exploratory group at the Family History Information Standards Organization attempted to derive a model of layered citations: a sequence of academic-style key:value pair set citations where each subsequent set is the source of the one before it. This showed some promise but never reached full expressivity.

• Elizabeth Shown Mills published an 800+-page illuminating her process for crafting citations, with hundreds of examples both presented and discussed in some detail. This helped popularize the idea that citation creation is a principle-driven art form. Its sheer size also helped scare some people away from the entire subject.

I am unaware of any effort to date that makes family history citations as orderly and well-understood as they are academic citations.

In my next few posts I hope to lay out some principles for the creation of structured family history citations. Along the way we’ll explore several topics that help run the digital world we live in.