I am an active user contributing to map Turkmenistan on OpenStreetMap. Recently, I edited Balkan Province of Turkmenistan to test a new set of admin_level tags before noticing I had to discuss it beforehand. Sorry for that. I’ve chosen that province because it’s the least densely populated of the country and thus easier to revert.
Today, current admin_level tags for Turkmenistan are the following:
4=Provinces (Welaýat);
5=Districts (Etrap) within a Province;
8=Cities (Şäher) and Towns (Şäherçe);
9=Rural Councils (Geňeşlik), Boroughs within municipalities (UlyEtrap) and Villages (Oba)
10=Neighborhoods (Ýaşaýaş toplumy) and Districts within municipalities (Kiçi Etrap/Etrapça)
11=Subdistricts (Etrabyň etrap)
This set has several problems in my opinion. The main problem, I think, is that it is not representative of the law of Turkmenistan when it comes to municipalities.
Municipalities of Turkmenistan as units are Cities, Towns, and Villages. However, Cities and Towns are lawfully equal to Rural Councils (a group of several villages or a single village representation), although Cities are in fact more “powerful” than Towns, which are more “powerful” than Rural Councils. Currently, Cities and Towns are on the same admin_level, while Rural Councils are on the same admin level as villages, which makes no sense. Also, there are Cities that are promoted by law to a district status (in the case of Balkan Province, Balkanabat and Türkmenbaşy). It’s not represented in admin_levels, and they are then considered as simple Cities. And finally, we have two levels, 6 and 7, that do not serve any purpose in the admin_level tag for Turkmenistan.
I’ve been directed here by fellow contributors apm-wa and Aleksandar Matejevic to discuss it and maybe fix it.
My proposition to reorganize the admin_level tag for Turkmenistan is the following:
4=Provinces (Including Ashgabat, capital city and lawfully a “Province-like” City, not part of any province)
5=Districts (Including cities with district-status, there are 7 in the country)
6=Cities, Towns and Rural Councils (that last one would be a relation instead of a node until now)
8=Villages
9=Boroughs within municipalities
10=Neighborhoods and Districts within municipalities
11=Subdistricts
apm-wa was concerned by my proposition about “complexity” and proposed another set, which is the following one, as formulated by them:
place=province (there are 5 in Turkmenistan, though some folks INSIST on including Ashgabat because it has the legal privileges of a province), admin_level=4
place=provincial district (has a governor), admin_level=5
place=city, admin_level=6
place=town, admin_level=7
place=Rural Councils and Municipal Boroughs with mayors, Villages, admin_level=8
place=Neighborhoods, Hamlets, Municipal Districts, admin_level=9
place=Subdistricts (what in Russian are called микрорайон), admin_level=10
Let me know how everyone thinks about this topic. Be critical, and I hope we will be able to find a consensus.
I don’t know about the concrete situation in Turkmenistan but would like make a general observation here: the admin_level value should reflect on the administrative hierarchy of the governmental entities. This is not necessarily the same as the hierarchy created through place tags. The place tags should reflect the human perception of a place. There usually is no 1:1 matching between admin_level and place type and I would advise against attempting it.
Regarding Ashgabat, I wouldn’t be surprised at all when it ends up as place=city+boundary=administrative+admin_level=4. This is a perfectly normal situation for capitals seen in many countries (example: Berlin vs Munich, Tokyo vs. Osaka).
In that spirit, your proposal looks quite reasonable to me. And with that I leave the discussion to the locals.
Turkmen law sets some specific criteria for what by law is designated a “village”, “rural council”, “town”, and “city”. As Saamat indicates, it further splits cities into “cities subordinate to a district” (districts being the rough equivalent of counties) and “cities with district status”. My position has been that, because the Turkmen government loves to tweak these lists, keeping them up to date is a losing proposition so simply designating municipalities as “villages”, “rural councils”, “towns” and “cities” suffices.
Furthermore, when I began mapping Turkmenistan in 2015, I discovered that whoever had mapped before me had skipped admin levels 6 and 7. I have no idea why. This meant that when I began mapping Ashgabat, I ran out of admin levels when it got to the microdistricts. I had to add an extraordinary admin_level=11. I was not happy about it but the alternative was to do a wholesale overhaul of the admin_levels and I was too busy collecting raw data (plus working full time in Ashgabat) to undertake that. Hence my proposal as Saamat has provided above, which reduces the total number of admin levels to 10 (the norm). If Saamat wants to go to the work of changing rural councils to relations, that’s fine. I would have done that if I could have figured out how to do it.
I remember you telling me so in private messaging. Consequently, I would edit my set and tag villages as admin_levels=7, boroughs=8, etc., removing the 11th tag. I can handle creating relations for Rural Councils.
About your previous comment and Turkmen government tweaking lists of municipalities, I acknowledge it changed a lot in the past few years and is subject to change again. Yet, I think we can now keep track of it as official documents are now partly shared on the website of the Assembly of Turkmenistan.
The problem I have with that is that boroughs within cities have presidentially appointed mayors (häkim) so setting them at a lower admin level than villages (which have zero governance and are subordinate to the rural councils, which in turn are subordinate to a provincial district, a city, or a town) is illogical. The boroughs of Ashgabat, Arkadag, and Turkmenbashy have mayors, so should be higher admin levels than villages. Rural councils and villages should be lumped together. They are the same admin level.
Based on recent research into the administrative structure of Turkmenistan, the hierarchy is defined by three distinct levels:
First Level: Provinces (welayatlar) and the city of Ashgabat (due to its special status).
Second Level: Districts (etraplar), including rural districts, city districts, and cities with district status.
Third Level: Cities, towns, and rural councils (gengeşlikler).
While Saamat’s structural proposal aligns with this three-tier hierarchy, I highly value the extensive contributions and localized knowledge provided by apm-wa. To ensure the data is as robust as possible, I suggest a hybrid approach:
Administrative Relations: Utilize same admin_level value for the relations that share same administrative power/level.
Place Tagging: Incorporate apm-wa’s specific place-tagging methodology within these relations to clearly distinguish between a city, a town, and a rural council.
Example: Third Level administrative relation will have admin_level=6 value and proper place=city/town/village tag so we can distinguish the difference between them, or query specific place type if needed.
To ensure a smooth transition and maintain community consensus, I suggest the following next steps for Saamat:
Wiki Update: Before making any bulk edits, please update the OSM boundary=administrative Wiki page to reflect these changes. This ensures transparency and provides a reference point for all mappers.
Global Update: Once documented, proceed with updating the admin_level values across Turkmenistan to align with the new schema.
Just to be sure, before making any changes on the wiki, I hear apm-wa’s argument about Neighborhoods, and thus I would swap Boroughs and Villages and set them to 7 and 8 respectively. I’m not that sure about setting Councils and Villages at the same value, though. Rural Councils being a higher structure/entity, including one or several villages, makes me feel like they should be on a different value than Villages.
Here’s what my draft currently looks like:
4=Provinces (Including Ashgabat)
5=Districts (Including cities with district-status)
6=Cities, Towns and Rural Councils
7=Boroughs within municipalities
8=Villages
9=Neighborhoods and Districts within municipalities
This makes much more sense. I am not sure that ONLY having rural councils as relations of villages is advisable. I would prefer leaving a node for the rural council in the village designated as the seat of the rural council so travelers have an idea of where to find the town hall (few of which have been mapped precisely). It has to be there somewhere, and is usually only identifiable by driving around the village and finding the nondescript structure with a small brass sign on it that is the “town hall” (the cell phone store is invariably much easier to find). Where the town hall has been located (e.g., the Owadan depe geňeşligi, which I geolocated) a node designating the town hall itself works, but those are pretty rare in Turkmenistan.
I have no problem with rural councils and villages having the same admin level–neither controls its own budget or affairs (they are at the mercy of the district administration, same as towns and cities subordinate to a district). I can live with this proposal, though, if you all really feel strongly about equating cities, towns, and rural councils.
I would move Neighborhoods (Turkmen: ýaşaýyş toplumy, Russian: жилой комплекс, in English “residential complexes”) to admin level 10, along with subdistricts. Each Neighborhood is subordinate to a municipal district (admin_level=9) and each is always contained inside (i.e., is smaller than) one municipal district. They are crucial to navigation (which is what maps are for, right?) and in most cases are recognized by the mayor’s or district governor’s office, so need to be on the map.
Careful with that. We in the US have places called “city” that have little to no population. I would say that the place tag represent the relative scale of a place, generally independent of boundaries.
Hi, I actually planned to do so some days ago already, but I’ve been cut short as I saw a group of edits tagged #OzonGeo changing Rural Councils’ relations to just settlements’ borders. I forgot to talk about it. So I thought about it, and I think it’s not so important actually, and I might add relations for both the settlements themselves and their rural council.
I already changed the admin_level set in boundary=administrative and I’ll be able to start changing it soon on OPM.
Indeed. In my current general U.S. neighborhood the “City of Falls Church” has a population of 14,258 but the “Town of Vienna” (Virginia) has a population of 16,473.
That said, however, Turkmen law is quite definitive about what constitutes a “city” (şäher), “town” (şäherçe), or village (oba): minimum population level for each category (listed in the above linked wiki article), quantity of certain amenities like hospitals, clinics, and theaters. I’ve had to update the map periodically when cities have been downgraded to towns and vice versa.
To add to the headache, sometimes a name is used for both a city and a town (Türkmenbaşy is a good example, and if you think that doesn’t confuse people…) Village names are routinely recycled but namesakes are located in different districts, which is why I have been a bit obsessive about adding addr:district=* to every village I’ve mapped.
What Turkman law calls a place shouldn’t be too relevant for what goes into the place tag. You can do as the US people and put it into border_type. What should go into the place tag (node) is how the average Turkman inhabitant qualifies the place and that should be quite stable. I think that is exactly the point that @ZeLonewolf is trying to make and that I wanted to bring over as well.
