Sharing conclusions isn’t viable.

It is always dangerous to make generalizations about what people do, but sometimes venturing into that dangerous territory serves rhetorical purposed. Thus, the next paragraph might be false in some instance, but it matches my current understanding and is a useful foil.

Online tools for collaborating on genealogical research share a mixture of structured conclusions with citations and unstructured text. There are lots of ways to organize these elements, from a wiki-like network of articles to massive family trees tagged with free-form explanations. But these elements are the current state of the art.

There are problems with both of these pieces.

Unstructured text is outside of a computer’s ability to assist, meaning that duplicate entries and contradictions may proliferate unnoticed. It also makes misunderstandings much more likely, imposes language barriers as research becomes more global, etc.

Conclusions, on the other hand, suffer because humans are “‍differently rational.‍” Frankly, some are non-rational and will see themselves related to some high-up or the other, all evidence to the contrary be hanged. But even within a fully rational group or a single person, conflicting but defensible beliefs can proliferate. I’m not talking here about blind men and an elephant style disagreements—those can be handled by free-form text. I refer to things like which father to choose for a possibly-illegitimacy child or selecting between equally-trustworthy sources that disagree with one another.

Two adjustments might make sharing conclusions more viable. But both have problems.

First, we might require some kind of adjudication, be it voting on disagreements or appointing a steward of each person. The problem with this is that it has a tendency to alienate researchers who disagree. If I cannot enter all of my conclusions in the shared tool then I’ll need to keep my own copy, and two copies is a recipe for inconsistency and redundant work.

Second, we might allow the family tree to turn into a network of alternative family trees. The quality of this adjustment depends on how the system treats an edit to a shared ancestor. When I change our common ancestors’ information, does it automatically change your view too or does it create a new alternative, splitting our trees into to? If the former, I lose control over my data and still need to keep a personal copy to prevent my research getting hijacked; if the later, we aren’t really sharing trees at all, just looking at each others’ trees from time to time.

I’d love to have a tool that supports actual collaborative genealogy. Currently I can’t find one that even lets me collaborate with my own parents effectively. My choices right now are to either walk through all of their research or to assume everything they said is true (which even they don’t do) and try to find new sources to push out the edges of a possibly-faulty branch.