Multilingualism in Serbia

Dear Serbian OSM editors,

Can you please clarify the Serbian part of the Multilingual names here:

https://wiki.openstreetmap.org/wiki/Multilingual_names#Serbia

"The default name= tag should be filled in using Serbian Cyrillic, unless the feature is in an area where the language and script of the sizable ethnic minority is in official use (e.g. Subotica or Preševo). In that case, name:sr=* must be added, as well as the tag specifying the language of the default tag (e.g. name:hu=, name:sq=, etc.)."*

This pharagraph mentions where there is a sizeable minority the local name tag should be added. But it is not clear what to do with the name tag. It has to be always in Serbian cyrillic?

Thank you in advance.

Andras

I think that contents of name tag is a point of contention, so we might as well re-discuss the issue.

I think that all name tags in Serbia (Kosovo excluded) should be written in Serbian Cyrillic, except possibly for business/brand names of foreign origin.

I have personally reverted naming of “Суботица / Subotica / Szabadka” for a couple of times. Having a multi-part primary name is,

  1. inaccurate: that’s not a “name”, it’s an aggregation of three names, and nobody ever calls this with three words, it is unpronoundable.
  2. impractical: how are end users supposed to consume this? No software is able to transliterate that. No navigation voice is able to pronounce that. It’s not of much use after all.
  3. inconsistent: most municipalities of Vojvodina are officially bilingual or multi-lingual. Several hundreds of villages have “a sizable ethnic minority”. It would be particularly unwieldy to try to force multi-language naming on all of them.

Let’s keep it simple, let’s keep the primary name in Cyrillic. Please.

P.S. I personally do not even use Cyrillic, and I’d prefer having primary names in Serbian Latin for practical reasons. But Cyrillic is the primary official script in Serbia, it makes sense to use it consistently, and as long this is the agreed convention I will try to enforce it.

Thank you for your answer. I accept your decision. However I would like to share my thoughts also.

Multilingual names are a quite sensible topic especially in OSM. In Western Europe we can see several place where they use it. They put more than one languages to the name tag. The base of OSM is the ground mapping rule I suppose. We should map exactly what we can see on the ground on the signs. Sometimes of course this is not possible. And as Vojvodina is an autonomous region with 6 official languages I feel that there is a place to discuss this.

Let’s have one example. Navigation. If I wanna travel through Vojvodina by car and let’s say I am travelling from Subotica until Beograd. As I am a Hungarian the best option for me is when I see Hungarian names in my navigation. So far it is easy I can set the navigation to Hungarian. Let’s pretend that I won’t use the motorways I would like to see the countryside. Until Backo Gradiste the signs will be multilingual so I will know where I am. Then I need to stop and set back the language to the default because from now on only I can read in Serbian. But I feel that that - for orientation - for any nations it is easier when the name has two names and on the signs too.
Answering to your points:

  1. We argued a little bit with the Hungarian community on this also. At this point I need to tell you that I am only a cartographer, not an OSM expert. I am not so familiar with the sytaxes of the OSM. They said also that “Суботица / Subotica / Szabadka” is incorrect as that is Суботица divided Subotica divided Szabadka. Technically it should be Суботица;Subotica;Szabadka and that means that the name contains three different names. Also I don’t want to say anything about the Serbian Cyrillic and Latin, that is totally your call.
  2. Maybe it’s also true, but in this case what happens in South Tyrol? Everything is Italian and German.
  3. You are right again. I just did a small research on Google Streetview and I noticed that even where there are only 4-5% Hungarian population there is a Hungarian name on the sign. I was shocked when I realized this, this is really no necessery for this in my opinion. The best option would be maybe when the Serbian would be the primary and secondary name would be only shown where the minority(the secondary language) has the majority. If I counted well that would mean 75 Hungarian, 15 Slovak, 16 Romanian, 10 Croatian and 3 Rusyn muncipalities. Just one option for consideration. :slight_smile:

I guess until now we practiced what is common in other parts of the world - if there is multi language community and it uses names in two or more languages, all languages go to name tag. In our case, Serbian Cyrilic goes first and others follow. I guess it is omitted from description you are referring to by error. At time it was defined, we were more focused at specific name tags for languages.

That is also consistent with on the ground rules, as our law says that in multi language communities signs are also multi language. Serbian Cyrillic is mandatory and others are allowed. In reality that is the case, signs do have text in Serbian and language of minorities.

However, we established rule that in such cases additional name tags specific for each language are mandatory.

That is most practical and most useful solution. I see no reason to change that and favor Serbian Cyrilic. Especially now when we do have customized Serbian language rendering that allows user to chose which script he wants on map tiles (see https://openstreetmap.rs/map/)).

The only exception is Kosovo and Metohija region. For some reason, OpenStreetMap officials allowed them to break all rules and all logic and force Albanian language by forceful albanisation of all names. Even names that are clearly not Albanian are transliterated to Albanian and used as such. We are forbidden to enter Serbian language names in name tag. If Albanians agree to add Serbian in name tag that is exclusively Latin script. They forbid using Cyrillic on OpenStreetMap even if their law allows it on the ground. By Serbian law and law of so called Kosovo, Albanian and Serbian languages are to be equally used on whole Kosovo and Metohija region. OSM does not follow that, they alow albanisation of all names.

What you describe is definitely not so common in other parts of the world. Let us just browse European countries and regions with strong bilingual tradition:

  • Belgium: Except for Brussels (whose name=“Bruxelles - Brussel”), all cities have the name according to the primary speaking community
  • Switzerland: Same as Belgium, with exceptions for the country name and a few select canton names
  • Åland (Swedish-speaking territory in Finland): all names in Swedish
  • Catalonia and Balearics: all names in Catalan
  • Belarus: all names in Russian (the country’s lingua franca); Belarussian names are “hidden” behind name:be tag.
  • North Macedonia: all names in Macedonian, even for apparently Albanian-majority places
  • Bosnia and Herzegovina: primary names in Serbo-Croatian Latin, while Cyrillic is behind name:sr tags
  • Ireland: primary names in English; only “Londonderry / Derry” has the double name due to long-standing historical divide. Irish Gaelic names are behind name:ga tags.

The only region where I see systematical double naming is South Tyrol, and even there numerous hamlets and toponyms have only German name. I would say that bilinguality there is particularly strong and often mandated by law.

Allow me to explain about Brussels: by law, the rather small territory called “the 19 communes” is bilingual NL/FR, with both languages to be treated equally. As the matter is delicate, politically and historically, this law is strictly seen to. The present scheme is the solution acceptable to all, and is applied to all names in this area, streets, buildings, whatever.

Now I am a little bit confused.

I agree that this is not so common around the world. But it’s getting more and more used. And I can hope that in the future more countries will agree on using this.

Some more examples:

  • Slovenia, near Koper and Lendava: Slovenian / minority language
  • Austria, in some places of Burgenland and Karnten: German/minority language
  • Italy: not only in South Tyrol, but near Trieste and in Sardinia: Italian / minority language
  • Germany, South Brandenburg: German - Sorbian
  • There are a few examples in Wales, UK also, but mainly the names are Welsh

Dear Sebian Editors,

I have just noticed that recently there were again changes from some users in the name tag in Vojvodina.
Now Croatian names have been added to the name tag.

I would really like to suggest to sit together, make a discussion about this topic. Until that is not done there will always be name wars like this. There should be a solution which is acceptable for all of us. That is pretty clear that all names on the settlements signs cannot go to the name tag as it would be totally confusing.

But multilingualism has to be possible in an autonomus region, hasn’t it? :slight_smile:

In hoping of further successful cooperation:

Andras

Links?

Problem is name tag itself. It is not appropriate for practical use. This is why name tag fro each language exists.

In my opinion name tag should be abandoned or repurposed. It just causes name tagging wars.

Bački Monoštor and Donji Tavankut

I personally like the / divider in the name tag. When there are more than one local name I really believe that is the best solution.

Can you name some specific cases when this causes troubles?
I only use openstreetmap.org (where there is no option to choose a displaying language) and the OSM based OsmAnd app for Android.

All it does is cluttering the map for no good reason. Plus, except for Hungarian names (that truly differ), most other “names” across Vojvodina are mere transliterations of one another.

I use OSM-based Maps.me, and it displays settlement and street names in both Cyrillic and Latin (next to each other), and POI names in Latin only. It seems to use a rather smart algorithm how to choose among name:en, name:sr-Latn and direct transliteration from Cyrillic.