Should protectorates and Palatinates be mapped as part of the British Empire? The wiki says that a protectorate is an independent country controlled/protected by the British and a Palatinate is land controlled by a British noble.
And if they should be mapped as part of the British Empire what admin_levels should they be tagged with? I’m currently mapping colonies with admin_level=3.
There seems to be a country=empire tag in use for empires that consist of multiple countries.
Consider how OSM represents present-day possessions. OSM has a separate relation for the British Overseas Territories distinct from the United Kingdom, because the overseas territories aren’t considered “part of” the UK, even if it’s the UK’s sovereign territory. If I’m not mistaken, relations like the United Kingdom of Great Britain and Ireland should include, obviously, Great Britain and Ireland, but not the British Overseas Territories, and not Crown Dependencies like Jersey and Guernsey.
These and the former realms would’ve been considered countries or country-equivalents in their own right – not only Canada or India but also, say, Bermuda and British Hong Kong. I think these boundaries should have admin_level=2. The admin level system is an attempt to draw parallels across different systems of administration. It’s OK that the United States and the Dominion Canada are both admin_level=2 even though the Dominion of Canada is subordinate to a larger entity.
There doesn’t seem to be a relation yet for the British Empire as a whole. That would be a good candidate for country=empire as well as admin_level=1.
Not every empire is mapped this way. The Ottoman Empire was divided into provinces with separate vassals and tributary states. The main relation is currently tagged as an ordinary country, even though it includes those tributary states. I anticipate that the structure will get more complex once we attempt to detail the various kinds of administrative divisions under the Ottoman Empire.
In the OSM wiki says that the only areas with admin_level=2 should be the countries in the ISO 3166-1 with independent=yes, and Kosovo, the SADR, and Taiwan
There’s a sign saying that some areas currently tagged as admin_level=2 shouldn’t be so, and I assume this refers mostly to the british territories.
This documentation is currently written for OSM. In the present day, ISO 3166-1 adequately describes what most people regard as countries, with a relatively small number of edge cases here and there. The admin_level documentation is a little clearer that the purpose of this key is to reflect levels of nesting. We can use OSM’s approach as a point of reference, but that ISO standard only dates back to 1974. I suspect we’ll have to chart our own course based on our experiences modeling earlier time periods, and then document our approach on a separate page.
The British Overseas Territories probably account for most of the exceptions that the warning box alludes to. But I think it’s only being modeled as an administrative boundary for convenience; in reality, it’s more or less a collection of disparate territories, like the United States Minor Outlying Islands, which are tagged as a statistical boundary. We could model both as statistical boundaries, or even as collection relations, which would give us additional flexibility.
In my opinion, if we treat these entities as countries in their own right, and then each dependency as a subnational boundary, it creates problems in the 19th century. At that time, it wasn’t uncommon for a single European monarchy to rule over disparate realms that were, practically speaking, administered from one of those realms.
A rule of thumb that sometimes comes up in OSM is: hypothetically, would it appear on a welcome sign? Would one say they are entering or leaving British Overseas Territories, or the Commonwealth, as they sail toward or away from Bermuda? Whereas maybe the Holy Roman Empire would be a different story, but it can already be mapped as an empire.
I’d prefer a different rule of thumb. More along the lines of “was historically able to conduct its own foreign policy”. Although I think it would create the same problems you allude to, with the monarchies.
Depending on the author of the map, these monarchies are sometimes represented as one single country (usually Austria-Hungary or 17th Century Spain) or as different countries (the random collection of Habsburg ruled domains). But really they are the same kind of chimeric state.
As for the British Overseas Territories in OSM: they are modelled as an administrative boundary without an admin_level, while each territory is its own admin_level=2. I’d prefer if they were mapped as, say, admin_level=3 (individually, not as the Overseas Territories) while inside the United Kingdom relation, which is how other similar cases are mapped in OSM, such as Hong Kong, the United States Minor Outlying Islands, Puerto Rico or Clipperton Island.
And that’s how I’d prefer these cases were handled in OHM, with the rule of thumb I mentioned at the beginning. I don’t really care if it ends up being admin_level=1 or 2, but one of those should represent states capable of conducting their own foreign policy, and the meaning shouldn’t change with time (at least not too much).
I recognize this is completely utopic and there’d be an insane number of exceptions (I’m having trouble fitting some situations into this rule already).
But it’s important we get some kind of clear definition.
Yes, history is a mess, I would like some clarity as you say and some generic rules on the wiki covering some of these topics for new mappers to better understand what’s going on. I guess I’m also a new mapper on OHM (but experienced on OSM) and this is why I’m trying to raise this topic.
The first thing that comes to mind for me is the status of a country during wartime or some other disaster. Lately, in OSM, the general approach has been to keep the status quo antebellum, but that might run up against a desire to show the state of the world more dynamically.
Of course, during such events, the boundaries might change for other reasons anyways (geometry, name), so additional copies may be no big deal. But rather than overload admin_level with multiple meanings, what if we were to make the sovereignty of a country in a separate tag (I suppose independent=yes), so that maps can choose to distinguish by that characteristic or not? Then admin_level is more likely to mean just “how many relations is it nested in” and we don’t have to wonder about other meanings.
OSM isn’t handing all these cases consistently. The U.S. Minor Outlying Islands is a boundary=statistical without an admin_level; Baker Island is at level 2. But Puerto Rico is at level 4, even though both are included in the United States boundary and only Puerto Rico has its own ISO 3166-1 code. I’m unsure of the reason for this discrepancy, other than possibly an oversight. I’m pretty sure the U.S. community in OSM would object to promoting Puerto Rico to level 3, as though it could encompass multiple states.
I agree that, as long as the dependency lies within the sovereign country’s border, it should have a numerically higher admin level than the sovereign country. I think the difficulty with the British dependencies is that, even at the height of the British Empire, the UK home countries were usually counted as a country per se (correct me if I’m wrong about that). We should have separate relations for the home countries versus the overall empire. That either requires demoting the home countries to a more minor level or promoting the empire to level 1.
The British Empire ended in 1997 after handing Hong Kong over to China. Then the British Overseas Territories were founded. But I would consider adding those territories to the United Kingdom relation just as the US has done because they are the owners of that land and the British Overseas Territories is just a name.