Temporal and spatial precision of areas

In history, countries have changed shape, both geographical and societal, in the course of time. How precise do we want to be for relationships? I’ll illustrate with the case of Czechoslovakia. We have the area XC, which is suitable for both the democratic First republic (1918-1938), the Second republic (1938-1938) which was reduced in are by the Munich agreement, the after-war Third republic (1945-1948) and the totalitarian ČSSR (1948-1989), which, during its existence also changed political forms. While many of these differed in shape and political systems, I’m happing with using the one area entity we have for all of them, but:

What about occupied areas? During WWII, today’s Czechia was mostly occupied by Germany and officially (by German and collaborating officials) presented as the puppet state Protektorat Böhmen und Mähren.

Now, I want to add a relationship to a group that was created there in 1941. Should I use Germany, into which it was de facto integrated by force? Or is it Czechoslovakia, because that’s what it was taken from by force? Is it its own entity, even though it was a puppet state with practically no autonomy?

Each of these is a bad choice in my view - It never was Germany and even wartime Germany didn’t consider it a part of Germany (yet). It certainly wasn’t Czechoslovakia, because that had ceased to exist apart from the government in exile, which did have some impact on the war and its outcome, but didn’t really govern the country at the time. And lastly, creating a new entity for a puppet state created by hostile action seems morally wrong. Without the intention of sparking a political debate, I can’t help but note that I can’t imagine submitting a new area ticket with a separatist Luhansk republic and for the purpose of this debate, there isn’t much difference between that and the Protectorate.

Similarly, the second “half” of Czechoslovakia, Slovakia, was a puppet state of Germany, but it came to be by choice of its government, which collaborated with Germany and declared an autonomous state (autonomous from Czechoslovakia, that is). The Protectorate’s government, on the other hand, had little continuity in willingness, with the legitimate Czechoslovak government. Should wartime Slovakia be a separate entity? It has little continuity with the present-day Slovakia, from which it is separated by decades of inclusion in Czechoslovakia. Wartime Slovakia and present day Slovakia also differ in shape significantly, just as the protectorate and present day Czechia. Additionally, the entity Slovakia in MB has its begin date set to 1993-01-01, which is the correct date of creation of the modern day Slovak Republic, after the peaceful dissolution of Czechoslovakia. Releases and relationship from before this date probably shouldn’t be listed under this entity.

Note that while we do have style guidelines for countries - a country is whatever is in ISO 3166-1 (or -3), these are countries that predate the ISO standard. I don’t think this disqualifies them from being used as countries per se, which is why I’m starting this topic. Also, we have a country entity, the Kingdom of Netherlands, which I understand to exist for the purpose of being the parent entity for various domains of the Kingdom, which are also countries. This is a different case from that of Czechoslovakia and its historical fractions, but it is a precedent for having a country entity that is not in the standard.

There’s a related topic, but I didn’t want to necro a 4 year old post.

1 Like

Hi, here is just a technical point of view on how this is handled in the database, to help with having a more general understanding and managing expectations for when it will come to propose possible improvements in the database design and editing guidelines.

Areas in the database are gracefully maintained by @drsaunde as a volunteer for now and we don’t have the human bandwidth to be committed to maintain historical areas, even though a few ones have been added in the past and some areas previously existing in the database have been marked as ended. See the full list of ended areas.

As you mentioned, historical facts can be debatable so we should rather rely on a specialized third-party database for handling areas’ data, or a general-purpose third-party such as Wikidata which was the topic of last year’s GSoC project even though it isn’t in production and it didn’t consider historical areas, this is a small step forward.

4 Likes

Looking forward to that automation. Even if it might not outright solve this particular problem, there are so many areas much less significant than countries, that are missing and that I’ve wanted to use, that I’ve decided to not bother our sole area maintainer with them. Instead I use the next best thing (the smallest superset present) in combination with “credited as” where possible, until a third party area maintainer is implemented.

this is a small step forward.

It’s a small step for MB, but a giant leap for @drsaunde!

In any case - the question still stands, I’m not leaning towards any of the solutions I could think of, including adding new areas.

You probably already know this, but I didn’t see it mentioned in the thread. To add missing areas, go to: Browse projects - MetaBrainz Tickets
Select Area Request, then click on create.

Dr saunders handles this. A little behind at the moment, but usually catches up eventually. I usually drop a note on why I want it added, so that I remember when I get the email about it being added, why I wanted it in the first place.

Update: I’m just seeing this GSOC '23 link and it has all this in there. Also, explains why area request have fallen behind a little bit.

1 Like

Yeah, I’ve used this option in the past, but the matter at hand is more a matter of style, rather than missing areas. It could be argued and decided that such area relationships would be solved by using the next best thing, just as well as that new areas are to be made.