Wikidata:Property proposal/Position with respect to the noun
position with respect to the noun edit
Originally proposed at Wikidata:Property proposal/Lexemes
| Description | the position this lexeme must have in respect to the noun, normally to use with adjectives |
|---|---|
| Data type | Item |
| Domain | lexeme |
| Allowed values | before (Q79030196) and after (Q79030284) |
| Example 1 | red (L3271) → before (Q79030196) |
| Example 2 | abertzale (L82679) → after (Q79030284) |
| Example 3 | bueno (L230409) → before (Q79030196) |
| Planned use | add information about adjectives |
| Expected completeness | always incomplete (Q21873886) |
Motivation edit
This is relevant if we need to model a language with automatic tools, and also for translation. Where to place the adjective (or other words) is something difficult for many people and machines need to learn it. Theklan (talk) 15:03, 18 December 2019 (UTC)
Discussion edit
Support - Theklan (talk) 15:03, 18 December 2019 (UTC)
- Hmm, I'm not sure this is the right way to model this, although I agree something is needed to cover this aspect of language. Another aspect of this is the "adjective order" rules, see en:Adjective#Order - how could that be modeled? It seems like maybe we ought to have something like "classes" for lexemes; we could use instance of (P31) or a custom property to indicate? Not sure how to proceed on this... ArthurPSmith (talk) 20:10, 18 December 2019 (UTC)
- @ArthurPSmith: I don't know if this is the best way to do this, but in Basque this is a relevant aspect. And I suspect that it exists also in other languages. -Theklan (talk) 12:42, 23 December 2019 (UTC)
Comment This might vary by sense, cf. French grand (L9217):
(‘a great man/celebrity’) vs.
(‘a tall man’). With the proposal being that this be a property of lexemes, would there then be two different lexemes for French
,
and
(see wikt:fr:grand#Notes)?―BlaueBlüte (talk) 04:41, 26 December 2019 (UTC)
Comment Regarding before (Q79030196) and after (Q79030284), why do we have adverbs (instance of (P31)adverb (Q380057)) in the item space? (In which case, which language should they be taken from? And should that language be used regardless of the language of the lexemes described by this proposed property?)
Items should represent more abstract concepts (e.g., anteposition (placement of a dependent phrase before the one it depends on)instance of (P31)placement of dependent phrase), not the adverbs used to denote those concepts.―BlaueBlüte (talk) 04:41, 26 December 2019 (UTC)
Comment “with respect to the noun” might be inconveniently narrow, as the lexeme in question might also be a noun phrase or other nominal phrase. And speaking of which, are there languages that have non-arbitrary ordering of certain kinds of adjectives or similar? For example, adjectives that are always positioned after the noun, but before any other adjectives that are also positioned after the noun? Like, ‘ranks’ of adjectives? Maybe the range for this property should in that case be (positive and negative) numeric, e.g., negative numbers meaning before, positive numbers meaning after the ‘noun’, with the absolute value indicating how far (with respect to potential further adjectives as attributes to the same ‘noun’).―BlaueBlüte (talk) 04:41, 26 December 2019 (UTC)
Weak oppose if it were by sense this would be very good for open-source translations for now, but currently it needs some changes as User:BlaueBlüte says.
Comment after (Q79030284) and before (Q79030196) are currently instance of (P31) of adverb (Q380057). They should not be words, but concepts IMHO. — Finn Årup Nielsen (fnielsen) (talk) 00:33, 28 February 2020 (UTC)
- @Theklan, BlaueBlüte, Fnielsen:
Not done Closing as stale, unsupported. Feel free to create a new proposal if this still seems needed, referencing this discussion. ArthurPSmith (talk) 16:20, 11 January 2021 (UTC)
