Wikidata:Property proposal/part of speech

part of speech

Originally proposed at Wikidata:Property proposal/Term

   Withdrawn

Description	function of a word or a phrase in a sentence
Represents	part of speech (Q82042)
Data type	Item
Domain	as a statement or qualifier for items that are instance of (P31)word (Q8171)
Allowed values	instance of (P31)part of speech (Q82042)
Example	technics (Q23001701) → noun (Q1084) Israel (Q801)demonym (P1549)Israeli / part of speech noun (Q1084)
Planned use	Migrate existing instances of demonym (P1549)/P794 (P794) to demonym (P1549)/this new property as appropriate; Add statements to any item instance of (P31)word (Q8171)

Motivation

I've been preparing for the deprecation of P794 (P794) when I came across the use of P794 (P794) to qualify Israel (Q801)demonym (P1549)IsraeliP794noun (Q1084). I thought, surely there is a "part of speech" property for this purpose? No, there isn't. A look at the results of this query and the property list suggests that such a property is needed but doesn't exist yet. So as a responsible Wikimedian I came here to propose one.

I expect this property to be used in two ways:

1. As a property for items that instance of (P31)word (Q8171) to state whether they are noun (Q1084), adjective (Q34698), verb (Q24905) etc
2. As a qualifier for demonym (P1549) and other properties which take a word or a phrase but there may be a need to qualify the part of speech.

Deryck Chan (talk) 12:53, 2 November 2017 (UTC)

Discussion

•  Oppose subclass of (P279)noun (Q1084) or subclass of (P279)verb (Q24905) is enough David (talk) 15:10, 2 November 2017 (UTC)

  Hi David, thanks for the suggestion. I don't think subclass of (P279)noun (Q1084) will serve the intended purpose:

  1. It's not a subclass of (P279) because we can see that the domain of the standalone use of this proposed property is instance of (P31)word (Q8171) but noun (Q1084)subclass of (P279)word (Q8171), adjective (Q34698)subclass of (P279)word (Q8171), verb (Q24905)subclass of (P279)word (Q8171).
  2. subclass of (P279) can't handle the use case where we use this as a qualifier: Israel (Q801)demonym (P1549)Israelisubclass of (P279)noun (Q1084) sounds wrong. --Deryck Chan (talk) 17:24, 2 November 2017 (UTC)

•  Support I think this property will be useful. Giovanni Alfredo Garciliano Díaz ★ diskutujo 17:34, 2 November 2017 (UTC)
•  Comment can you describe how this fits in with the Wiktionary integration that is planned for Wikidata?? ArthurPSmith (talk) 19:25, 2 November 2017 (UTC)
•  Strong oppose With the Wikidictionary integration we will have a new datatype that can actually hold words. ChristianKl (talk) 00:36, 3 November 2017 (UTC)
  • @ChristianKl: I thought that'd be a separate part of Wikidata? Will we then switch the datatype of demonym (P1549) to this new datatype (which I presume is Lexeme)? Deryck Chan (talk) 16:24, 3 November 2017 (UTC)

• We will have new namespaces in addition to the Qxxxx and Pxxxx namesspaces that we have currently. Those namespaces will contain information about individual words (lexemes). We won't have items with "instance of (P31) word (Q8171)" after that point inside the Qxxxx namespace. T

When it comes to dealing with properties like demonym (P1549) that a discussion we can have when we have the new datatypes in half a year. ChristianKl (talk) 16:40, 3 November 2017 (UTC)

• Withdrawing, pending implementation of dictionary content in Wikidata which will introduce a new datatype. Deryck Chan (talk) 12:28, 4 November 2017 (UTC)