Wikidata:Property proposal/grammatical gender

grammatical gender

Originally proposed at Wikidata:Property proposal/Lexemes

   Done: grammatical gender (P5185) (Talk and documentation)

Description	grammatical gender of the word
Represents	grammatical gender (Q162378)
Data type	Item
Domain	Lexemes and Forms
Allowed values	masculine (Q499327), feminine (Q1775415), neuter (Q1775461), common (Q1305037), animate (Q51927507), inanimate (Q51927539), dependent (Q51927608)
Example	chien (fr) → masculine (Q499327) joc (ca) → masculine (Q499327) wandeling (nl) → feminine (Q1775415) vovse (L128) (da) → common (Q1305037)
See also	Wikidata:Property proposal/has grammatical gender

Tubezlob (🙋) 19:19, 17 April 2018 (UTC)

Discussion

Statement directly in the lexeme for the main form, and in forms for specific forms. Tubezlob (🙋) 19:19, 17 April 2018 (UTC)

•  Support --Denny (talk) 19:34, 17 April 2018 (UTC)
•  Support obviously, but the "allowed values" list should probably be expanded (or maybe we should accept all items under grammatical gender (Q162378), BTW is
  ⟨ feminine (Q1775415)    ⟩ instance of (P31) ⟨ grammatical gender (Q162378)    ⟩
  but is
  ⟨ animate (Q51927507)    ⟩ subclass of (P279) ⟨ grammatical gender (Q162378)    ⟩
  , is this on purpose?). Cdlt, VIGNERON (talk) 06:30, 18 April 2018 (UTC)

  @VIGNERON: No it's my mistake (I created the three last items). instance of (P31) is better? Tubezlob (🙋) 19:46, 18 April 2018 (UTC)

•  Support. YULdigitalpreservation (talk) 12:45, 20 April 2018 (UTC)
•  Support NMaia (talk) 22:58, 21 April 2018 (UTC)
•  Support Robin van der Vliet (talk) (contribs) 21:18, 22 April 2018 (UTC)
•  Support but of course with all the values, see --Barcelona (talk) 08:29, 24 April 2018 (UTC)
•  Weak oppose. Do we intend to use the same property for noun classes in languages (e.g. Swahili) where the noun classes aren't typically called "genders" by English-speaking grammarians? Deryck Chan (talk) 09:25, 24 April 2018 (UTC)

  @Deryck Chan: This seems to be a controverse between authors:

"A noun may belong to a given class because of characteristic features of its referent, such as sex, animacy, shape, although in some instances a noun can be placed in a particular class based purely on its grammatical behavior. Some authors use the term "grammatical gender" as a synonym of "noun class", but others use different definitions for each.

Many authors prefer "noun classes" when none of the inflections in a language relate to sex, such as when an animate–inanimate distinction is made. Note however that the word "gender" derives from Latin genus (also the root of genre) which originally meant "kind", so it does not necessarily have a sexual meaning." — Wikipedia, Grammatical gender

I think we shoud use the same property for the global defintion of gender (that some people call "noun class"). So now we should choose between "grammatical gender" and "noun class". I think that "grammatical gender" is better, because it's the common term in use, "noun class" is more obscure. Tubezlob (🙋) 10:24, 24 April 2018 (UTC)

I also understood grammatical gender and noun class to be synonymous. I don't care about the label used. --Denny (talk) 16:06, 24 April 2018 (UTC)

•  Support --Fralambert (talk) 12:15, 24 April 2018 (UTC)
•  Oppose - already covered by the "grammatical features" functionality of Lexemes. Deryck Chan (talk) 09:19, 26 April 2018 (UTC)

  @Deryck Chan: I don't know. This seems wired for me to put all in the same field: gender, number, person, case, tense and mood, etc. The use of several properties give us the opportunity to put constraints. We've always done this in Wikidata. Tubezlob (🙋) 12:40, 26 April 2018 (UTC)

  • @Tubezlob: See e.g. this demo item. It doesn't hurt to have case, gender and number in the same field because one can click into the features to see what grammatical categories they belong to. I guess having "grammatical features" as one field circumvents the noun class vs gender debate and tense/voice/mood boundary debate. Actually most dictionaries state them in the same line (e.g. this Biblical Greek concordance simply writes V-RPA-NMS for verb / participle / present / active / nominative / masculine / singular) and that doesn't seem to impede understanding. Deryck Chan (talk) 08:31, 27 April 2018 (UTC)

  @Deryck Chan: OK but I'm not entirely convince 😉 Does anyone else have any opinions on the subject?

  And that's only for forms, isn't it? To use the example again, "chien" in French is masculine (Q499327) and not a form of another word. How to state this without the property? Tubezlob (🙋) 11:48, 28 April 2018 (UTC)

  @Tubezlob: See this ongoing discussion with Lea. Deryck Chan (talk) 20:43, 29 April 2018 (UTC)

•  Done
  --- Jura 09:16, 23 May 2018 (UTC)