Modern Software Experience

2011-07-17

reconstructive geneathology

web experiment

This week, I performed a fairly simple experiment to compare paper-published and web-published genealogy; I picked a name, created a genealogy from various web sources, and then compared the result to an printed genealogy from a respected Dutch genealogical series.

I picked a genealogy published in Nederland's Patriciaat (Dutch Patriciate), an annual publication of the Central Bureau of Genealogy (CBG). The name I picked is Van Wayenburg - also spelled Van Waijenburg. That genealogy was published in Nederland's Patriciaat No. 74, in 1990.
I used Google to find online genealogies. I was not surprised to find most that most of the genealogies I ended up using were published on either Genealogie Online, a large Dutch site, or geneanet.org, a French site quite popular in Europe.

fragmentation

One problem you might expect when trying to construct a genealogy by copying parts from a variety of online genealogies is that you end up with a bunch of unconnected fragments. I lucked out though, and ended up with everyone in a single fragment.
I'll readily admit that it is not merely luck, an important factor is that there is a lot of Dutch genealogy online.

conflicts

Another problem you might expect when consulting a variety of sources is that some of them conflict with each other. I lucked out again. I found small disagreements between some of them, but few are serious enough to mention.
Once again, there is a bit more going on here than mere luck. Many of the online publication probably got their data for this family, directly or indirectly, from Nederland's Patriciaat; it is such a well-respected publication, that most genealogist will not bother to do their own research, nor to double-check it, but simply copy it.
Moreover, Genealogie Online matches your database to those of others, and informs you about such matches. Geneanet provides a subscription that will notify you of new trees containing names you are interested in. Such features encourage collaboration, and thus early detection and resolution of any conflicts that might have been there.

errors

There are differences between the printed genealogy and the online genealogies. It is, in this particular case, fairly reasonable to assume that most of the online genealogies got their data directly or indirectly from the Nederland's Patriciaat.
Under that assumption, most of the differences can be categorised as minor transcription errors. For example, the printed genealogy contains the name Josepha, while several online trees give the name as Josephina.

The most serious difference I encountered concerned the place name Beek. There is a place with this name in the province of Limburg, and another one in the province of Gelderland. A fair part of the genealogy contains places in Limburg, but whenever Beek occurs, the printed genealogy indicates that it is Beek in Ubbergen, and Ubbergen is a municipality in Gelderland. Yet, several online genealogies had Beek in Limburg anyway.

sources

Practically none of the online genealogies lists any sources. That certainly does not imply that the authors did not consult any sources, it merely means that they did not chose to publish their sources. The Nederland's Patriciaat does not list its sources either.

It is tempting to simply conclude that, because some of the consulted online genealogies do list sources, that the web is a higher quality source than Nederland's Patriciaat, but the situation is not that simple.
Most online genealogies do not list sources, and that makes it hard to judge the quality of the work. Nederland's Patriciaat does not list sources either, but we know that its editors guard over the quality of the genealogies they publish.

When the printed genealogy reports a name as a name as Josepha, and some online genealogies report the name as Josephina, it is fair to wonder whether the printed publication is missing two letters, and the web genealogies got it right. I searched the Amsterdam archives, and quickly found several scans that clearly give the name Josepha, not Josephina.

conclusions

This is just one quick experiment, not an exhaustive study. I offer no conclusions, just a few observations.
One is that there really is a lot of Dutch genealogy online. I've been under the impression for some time now that a lot of the data originally published in Nederland's Patriciaat has found its way online, and that this experiment to build a genealogy from online sources resulted in an unfragmented one certainly has not weakened that impression.

Because they lack sources, the genealogies printed in Nederland's Patriciaat and most of the online genealogies I compared with are really geneathologies. Geneathologies should not be cited as evidence, but it is remarkable how much the various geneathologies are in agreement with each other. Then again, the most likely reason for that close agreement is that all the online geneathologies copied, directly or indirectly, from Nederland's Patriciaat, and it probably pleases the editors of Nederland's Patriciaat that a spot-check of one particular difference showed their publication to be right.

links