Talk:Q6256

Descriptions in different languages

Descriptions in different languages are different enough to convey rather different meanings:

• English: distinct region in geography; a broad term that can include political divisions or regions associated with distinct political characteristics
• Spanish: región legalmente identificada como una entidad geopolítica particular, con un gobierno, administración y leyes propias

I think we should decide what this item mean and harmonize descriptions.--Pere prlpz (talk) 14:25, 18 July 2014 (UTC)

Not same as state

User:Infovarius comments in Special:Diff/329205979 that there is no difference between country (Q6256) and state (Q7275). Regardless of how the two items may currently be linked in the database, this isn't in correspondence with reality (which is why two distinct words and concepts exist in the first place). For instance, Scotland (Q22) is a country, but not a state (its state is United Kingdom (Q145)). The same goes for Greenland (Q223) (its state is Kingdom of Denmark (Q756617)), Curaçao (Q25279) (its state is Kingdom of the Netherlands (Q29999)) and others.--Anders Feder (talk) 22:38, 2 May 2016 (UTC)

Good examples! But I suppose they are rather exceptions from the equality... --Infovarius (talk) 20:42, 19 May 2016 (UTC)

That is a good way to put it, though they are well established. "Country" is more closely associated with and delimited by culture and shared history (it's a social term), while "state" is more closely associated with and delimited by international law and administrative boundaries (it's a technical/legal term).--Anders Feder (talk) 23:51, 20 May 2016 (UTC)

Administrative territorial entity?

This item is currently a subclass of administrative territorial entity (Q56061). This seems odd since the concept of administrative entity is usually used in context of individual countries (sovereign states), where it is a subnational entity. There are prominent definitions for international use, like this, which also considers administrative regions as units which a country is divided in. At the moment I didn't find prominent sources which would consider countries or sovereign states as administrative (territorial) entities. An unsourced English language Wikipedia article lists "country" as a term for administrative division, but this is probably inexact. An extended Wikipedia list on a separate page clarifies that "country" as a term may be used as a translation for certain types of subnational entities (e.g. in India and France). Subclasses of a country, like constituent states as subnational entities, may be considered as administrative entities, but for other kinds of countries like sovereign states this probably doesn't apply. I'd remove Q56061 from here, and similarly clear up administrative entity related classes for state (Q7275) and political territorial entity (Q1048835). --90.191.81.65 15:28, 7 June 2018 (UTC)

historical country

If the country doesn't exist now, but country existed someday, can we add the property "instance of" = country?

See for example, Republic of Central Lithuania (Q523380), where "instance of" = "country" and "historical country".

I invite to this duscussion @Maksim L. and editors of the object "historical country": @Bouzinac, Infovarius, Neo-Jay, Sian EJ, Yupik. Thus, I have two questions:

1. Can an object have two "instance of": "country" and "historical country" at the same time?
2. Can a country have "instance of" = "country" now, if that country existed in the past? Andrew Krizhanovsky (talk) 15:00, 22 December 2020 (UTC)

"Subclass of (P279): former administrative territorial entity (Q19953632)" has been changed from "Normal rank" to "Preferred rank" for "historical country" (Q3024240). That means "historical country" (Q3024240) is virtually not a subclass of country (Q6256), and we cannot find a historical country as a country in a query. They are treated as two separated classes: Q3024240 refers to the countries that existed in the past, while Q6256 refers to the countries that exist now. So, in my view, 1. an item cannot have both "instance of: country" and "instance of: historical country" at the same time; and 2. a country that existed in the past should have "instance of: historical country", not "instance of: country". --Neo-Jay (talk) 16:38, 22 December 2020 (UTC)

• country (Q6256) does not only refer to countries that exist now according to our main usage of the term. Qualifiers (like end time (P582)) exist to specify that a country doesn't exist anymore. ChristianKl ❪✉❫ 17:06, 22 December 2020 (UTC)

• If Q6256 also refers to those countries that once existed and have ceased to exist, do you mean that "historical country" (Q3024240) should be a subclass of "country" (Q6256)? Or do you mean that, if a country existed from Year A to Year B, this country was a "country" from Year A to Year B, and has been a "historical country" since Year B? In my view, we do not need to add the two complicated statements with qualifiers to a historical country. We may simply add "instance of: historical country" without qualifiers. This can simplify the queries for countries that exist now. --Neo-Jay (talk) 17:26, 22 December 2020 (UTC)

• Yes, historical country (Q3024240) should subclass country (Q6256). While saying that historical countries are no countries does simplify some queries it goes against the assumptions of the English languages which assume that "adjective_a X is a subclass of X". It makes things complicated at other places of Wikidata as well. country for sport (P1532) for example would need a changed value-type constraint (Q21510865). When defining a constraint for a property like value-type constraint (Q21510865) there's no reason to think about time. Forcing thinking about time into the concerns make things more complicated at that place. Being to be able to inherit all the schema that we do define for country (Q6256) is useful as well.

Pushing more information into instance of (P31) that can be better stated in direct statements makes queries easier that care for that specific information but is modeling that often leads to problems. ChristianKl ❪✉❫ 18:23, 22 December 2020 (UTC)

I also thought that "adjective_a X is a subclass of X" (see the edit summary of my edit on 22 June 2015). But I hesitate to say so now because there are exceptions (e.g., fictional country (Q1145276) is not a subclass of country (Q6256)). As for country for sport (P1532), I think that we can simply add historical country (Q3024240) as a value type constraint. If you think that historical country (Q3024240) should be a subclass country (Q6256), would you agree to revert this edit by Infovarius? This edit changes Q3024240's "subclass of: former administrative territorial entity" from "normal rank" to "preferred rank" and makes its normal-rank "subclass of: country" not queryable. --Neo-Jay (talk) 19:01, 22 December 2020 (UTC)

We currently have no property that stores the distinction between fictional and non-fictional entities directly in the property and use instance of (P31) to do that. For this model to work we have fictional country (Q1145276) fictional or mythical analog of (P1074) state (Q7275) to allow it to inherit models without using subclass of (P279). There's no equivalent for historical entities that replaces subclass of (P279) the way fictional or mythical analog of (P1074) does replace subclass of (P279).

Preferred rank doesn't completely stop querying. It stops wdt but not wd. I personally would prefer to see no instance of (P31) historical country (Q3024240) claims at all. As far as I remember past conflicts about using instance of (P31) historical country (Q3024240) and instance of (P31) country (Q6256) where resolved with the compromize of us listing both. ChristianKl ❪✉❫ 19:28, 22 December 2020 (UTC)

• I am agree with Neo-Jay that 1. object can be "instance of" "country" or "historical country" but not both at the same time. And 2. historical country should have "instance of: historical country", not "instance of: country".
• Qualifiers (like end time (P582)) could be used to specify that a country doesn't exist anymore. But these quilifiers will complicate scripts. --Andrew Krizhanovsky (talk) 15:36, 24 December 2020 (UTC)

  end time (P582) is a wrong qualifyer for once existed objects/states/real stuff etc such as Q3024240#P1963 : dissolved, abolished or demolished date (P576) is more fit to tell that something has disappeared. end time (P582) would suit more for purely thinkable concepts such as Q11514315#P1963. Bouzinac ^{💬●✒️●💛} 16:19, 24 December 2020 (UTC)

• Thank you for explanation the end time (P582). --Andrew Krizhanovsky (talk) 12:43, 26 December 2020 (UTC)

• Hello, Андрей. I like the current model when we show that e.g. USSR was a sovereign state until 1991 and it is now a historical state. This model uses "instance of: state" with normal rank and qualifier "end date" and "instance of: historical state" with preferred rank. This allows us to query list of sovereign states at any specific point of time (e.g. 1950). And at the same time it doesn't mess a query of current sovereign states. So I believe that the class of historical states is not now a subclass of (present) states but it can be regarded as a subclass "in some previous period of time". P.S. Honestly I don't know if we can trace not preferrable ranks along all over the superclass tree... --Infovarius (talk) 23:06, 25 December 2020 (UTC)

• Dear @Infovarius, thank you for the reply. Just a question: Why Soviet Union (Q151) is not an instance of country, but Republic of Central Lithuania (Q523380) is the instance of country? What is the difference? --Andrew Krizhanovsky (talk) 12:43, 26 December 2020 (UTC)

@AKA MBG: Because of ranks, I think. I've fixed. --Infovarius (talk) 19:07, 26 December 2020 (UTC)

Restrictions

Two issues with how country (Q6256) is used now:

1. Right now, querying for instances of country (Q6256) returns, for example, both United Kingdom (Q145) and Wales (Q25), and Kingdom of the Netherlands (Q29999) and Aruba (Q21203), which does not make much sense hierarchically.

   Either

   a) Wales (Q25) and Aruba (Q21203) should be instances of constituent state (Q1763527) (or similar), and not simultaneously be instances of country (Q6256)

   b) United Kingdom (Q145) and Kingdom of the Netherlands (Q29999) (and most other countries as well) should be instances of sovereign state (Q3624078), but not simultaneously be instances of country (Q6256) (unless, conceptually, there could exist a sovereign state (Q3624078) that was not also a country (Q6256); but currently sovereign state (Q3624078) is a subclass of country (Q6256), so this is not currently how we treat these concepts)

2. What thresholds should we have for when a state with limited recognition (Q15634554) also become a country (Q6256)?
   1. When e.g. more than 50% of other countries recognize it?
   2. When it becomes a UN member state?
   3. Others?

--Njardarlogar (talk) 10:12, 3 March 2024 (UTC)

I would say it has historical and cultural meaning rather than political. --Infovarius (talk) 18:23, 4 March 2024 (UTC)