Introduction

As I’ve said previously, Centurial is a refreshing approach to genealogy software and I still believe that.

That said, this is the first post in a series on the important and useful features that Centurial is missing, in no particular order, so look out for more of these in the future! In this post, I’m going to discuss a missing feature that has been a more-or-less constant thorn in my side: researcher interpretations.

What do I mean by interpretation?

Interpretations are basically what it says on the tin: the researcher applying their external knowledge and expertise to the literal claims made by the source material.

Use Cases

I have identified two broad ways that interpretations could be applied.

1. Intrepreting the value: This is pretty straightforward — the researcher adds a layer effectively changing the value of a claim made by the source, using some external knowledge or reasoning. Despite being quite a simple concept, there are numerous useful ways this could be applied. Here are just a few:

   • Correcting typographical/spelling errors: People make mistakes, but we, as human researchers, can frequently recognize those mistakes and correct them, making our data more useful for drawing conclusions. Right now, there is no faculty to do that, beside modifying the claims made by source, which is a huge no-no: we want to record the claims a source makes verbatim, and with this mechanism, we would be able to modify values without losing the original information.

   • Expanding name abbreviations: Centurial already has some mechanism for name abbreviations: it will recognize initials as evidence for the full name where appropriate. However, sometimes abbreviations are not so clear-cut: Wm for William and Jno for John, among many others, come to mind. Sure, they could add a list of abbreviations and use them to match names, but that ends up being a game of whack-a-mole, because abbreviations can vary across time, place, cultural background, and many factors. With “interpretations”, we sidestep that issue entirely.

   • Expanding place names: In practice, sources rarely give us the full information required to deduplicate and standardize place names. For example, my father’s birth certificate says Mineola, NY — but to standardize that, I generally find myself entering something like Mineola, Nassau, New York, USA, which is basically the standard format used by Ancestry.com. However, in principle, that is also a big no-no, because I should be recording exactly what the source says.

2. Interpreting the fact: This one comes to a principle I’ve come to that I think is very important, but one that I think Centurial completely misses: we should as much as possible be recording the actual claims that a source makes.

   In many or even most cases, sources don’t make explicit claims about the classification of name parts. Some exceptions might be a death certificate, with separate fields for first names, middle names, and surnames. But consider a newspaper article that lists the name of the person as simply John Smith.

   Right now, Centurial requires us to separate John and Smith into the “Given Names” and “Surname” field, respectively. However, this can absolutely lead us astray. For example, take this field from a death certificate I came across in my research:

   

   This appears to state that the deceased’s surviving spouse’s birth name is Lena Hazel. My first instinct is to divide that up with Lena as a given name and Hazel as her surname.

   The problem is that the source doesn’t actually say that. It just says “name”. So by splitting it up, I’m not only muddying the source’s waters with my own interpretations, but also, most importantly, I’d be wrong. You see, requesting her death certificate from many years later shows that her birth name is actually Lena Hazel Ireland. It turns out that Hazel is not her surname at all.

   That’s why it’s important to keep interpretations separate from what the source actually claims.

   With this feature, I would enter the name as a generic “Name” claim (which, admittedly, does not currently exist), and later add my interpretation that splits that claim into multiple fields.

Conclusions

“Interpretations” are the first type of data that I think are missing from Centurial. It’s not a completely breaking limitation, but I do feel it on a day-to-day basis, and believe that Centurial would get a huge power bump (from its already high level) through this addition.

By adding interprations, Centurial would essentially be turning part of the “proof argument” concept into a true, structured entity within the system, which is extremely useful. Currently, proof arguments seem almost like an afterthought, despite the fact that they are a core concept within the Genealogical Proof Standard.

To be fair, I’m much more convinced about value interpretations — they are an absolute must-have for this kind of software. Fact interpretations, on the other hand, are quite a bit more meta and I’m still rather on the fence about their overall utility.

Hopefully though, someday we’ll have this ability.