Property talk:P6254
Documentation
lexeme described or discussed in this item
List of violations of this constraint: Database reports/Constraint violations/P6254#Scope, SPARQL
List of violations of this constraint: Database reports/Constraint violations/P6254#Entity types
Inverse of?
editI'm not sure the constraint inverse property (P1696) really make sense. Do we really want each item to point to thousands if not millions of lexemes ?
Cdlt, VIGNERON (talk) 17:07, 14 April 2019 (UTC)
- Yeah, I don't think it's suitable for (or meant for) uses such as Q7560#P6254 (currently 5 lexemes). --- Jura 17:56, 14 April 2019 (UTC)
- That something is inverse property (P1696) doesn't mean, that the items show to each other. For example: part of (P361) is inverse of has part(s) (P527). Nevertheless, there exists many items A with "A part of B" but B has not the property "B has part A". That two properties are inverse of doesn't mean that they should point to each other.
- What would be more relevant for your question is the property constraint: [1].
- Now answering your question: The german Wikipedia want to use the property female form of label (P2521). Yet, it seems that this porperty should be deprecated and be replaced by the lexemes: Wikidata:Properties for deletion#female form of label (P2521). If the german Wikipedia want to use the lexemes in the articles (instead of "female form of label"), it is necessary that the item contains at least the german lexeme.
- --Eulenspiegel1 (talk) 20:12, 14 April 2019 (UTC)
- Still, P6254 is not for "female form of label" in German. --- Jura 20:15, 14 April 2019 (UTC)
- Currently, it's not. Yet, the plan is: P6254 links to the lexeme and in the lexeme is written the female form of label. Thus, the article uses P6254 to get the female form of label which is written in the corresponding lexeme. --Eulenspiegel1 (talk) 20:32, 14 April 2019 (UTC)
- Please find another way to do this. --- Jura 21:46, 14 April 2019 (UTC)
- Why? Also as you can see in the Deletion Discussion: The Catalan Wikipedia also need the female form. Also I think links to lexemes are a much smarter and eleganter way to do this than a list of labels. --Eulenspiegel1 (talk) 22:13, 14 April 2019 (UTC)
- Properties are proposed for specific purposes. User tend to rely on that being stable. If you need want one for a new purpose, just propose one. I will support it. I'm the one who initially proposed the current property with monolingual string format. --- Jura 22:17, 14 April 2019 (UTC)
- The purpose of P6254 is to link between an item and the corresponding lexeme. This should be stable. And this stable quality is used. --Eulenspiegel1 (talk) 22:23, 14 April 2019 (UTC)
- Your use is not supported by the property as it was proposed. mother (Q7560) doesn't describe or discuss the lexemes listed at Q7560#P6254 --- Jura 22:28, 14 April 2019 (UTC)
- OK, you convinced me. Yet, there should be another way to link the items to their corresponding lexemes. --Eulenspiegel1 (talk) 22:33, 14 April 2019 (UTC)
- What use do you plan? Most items don't have predefined corresponding lexeme. I can imagine 20+ corresponding Russian lexemes in Q7560#P6254. --Infovarius (talk) 21:45, 15 April 2019 (UTC)
- I have thought about it and I think about a 5th column in the header section. Currently, we have:
Language | Label | Description | Also known as
- As a 5th column there should be "Lexeme".
- @Infovarius, I already explained the use: The link to the lexeme should replace the properties female form of label (P2521) and male form of label (P3321). For details, see Wikidata:Properties for deletion#female form of label (P2521).
- In the topic section of mother (Q7560) there are only two russians labels written: мать and мама.
- I suggest that you don't link every lexeme of a language, but only the lexeme which is written under "label". --Eulenspiegel1 (talk) 23:11, 15 April 2019 (UTC)
- If you plan to replace the properties female form of label (P2521) and male form of label (P3321) then you'd better propose replacement properties with the same labels but value-type=Lexeme. --Infovarius (talk) 16:06, 19 April 2019 (UTC)
- What use do you plan? Most items don't have predefined corresponding lexeme. I can imagine 20+ corresponding Russian lexemes in Q7560#P6254. --Infovarius (talk) 21:45, 15 April 2019 (UTC)
- OK, you convinced me. Yet, there should be another way to link the items to their corresponding lexemes. --Eulenspiegel1 (talk) 22:33, 14 April 2019 (UTC)
- Your use is not supported by the property as it was proposed. mother (Q7560) doesn't describe or discuss the lexemes listed at Q7560#P6254 --- Jura 22:28, 14 April 2019 (UTC)
- The purpose of P6254 is to link between an item and the corresponding lexeme. This should be stable. And this stable quality is used. --Eulenspiegel1 (talk) 22:23, 14 April 2019 (UTC)
- Properties are proposed for specific purposes. User tend to rely on that being stable. If you need want one for a new purpose, just propose one. I will support it. I'm the one who initially proposed the current property with monolingual string format. --- Jura 22:17, 14 April 2019 (UTC)
- This is *NOT* an inverse of 'item for this sense'. I am removing the constraint. Read the proposal discussion - this property should ONLY be used on items for specific dictionary entries, not for generic entities. ArthurPSmith (talk) 20:37, 17 April 2019 (UTC)
- @ArthurPSmith: Why should it only be used on dictonary entries, e.g., the scientific article “PLUS THAT I BELIEVE IT IS AN INNOVATION....…” On the marginal construction with plus at as an initial conjunction in modern spoken Danish (Q58482515) discusses the lexeme plus at (L36654). I have used P6254 to indicate that relation. — Finn Årup Nielsen (fnielsen) (talk) 09:57, 26 April 2019 (UTC)
- Yes that's fine - I really meant items about lexemes themselves, for which these examples qualify. ArthurPSmith (talk) 13:50, 26 April 2019 (UTC)
- @ArthurPSmith: Why should it only be used on dictonary entries, e.g., the scientific article “PLUS THAT I BELIEVE IT IS AN INNOVATION....…” On the marginal construction with plus at as an initial conjunction in modern spoken Danish (Q58482515) discusses the lexeme plus at (L36654). I have used P6254 to indicate that relation. — Finn Årup Nielsen (fnielsen) (talk) 09:57, 26 April 2019 (UTC)
- Agree with the above. This property (P6254) is [piece of writing about grammar] -> [a word discussed by the writing]. item for this sense (P5137) is [a word] -> [the concept described by the word] (not a dictionary entry about the word). Deryck Chan (talk) 08:50, 10 May 2019 (UTC)
Ongoing additions
editFrom time to time some users add this property to wrong items. I suspect that this is because constraint violation is (wrongly) shown at corresponding lexemes. There's T223372 for this which is not fixed and the problem persists... --Infovarius (talk) 17:11, 21 June 2019 (UTC)
Using on items
editPlease discuss: Property talk:P443#Deprecate for Q-items?. --Infovarius (talk) 00:00, 1 January 2021 (UTC)
Linking official names to lexemes
edit@Jura1 @ArthurPSmith @Infovarius Would it be OK to use this property as a qualifier in official name (P1448) ? Such as in Q35757#P1448? There's a need to connect places to their official name lexemes. Currently I don't know how else to property label a lexeme as one corresponding to an official name of the place. Vojtěch Dostál (talk) 10:52, 22 February 2023 (UTC) Also @Mormegil: who proposed this usage.--Vojtěch Dostál (talk) 10:53, 22 February 2023 (UTC)
- @Vojtěch Dostál: What about lexeme sense (P7018)? Or wasn't there a new property for something like this with words from a monolingual text phrase? ArthurPSmith (talk) 20:56, 28 February 2023 (UTC)
- lexeme sense (P7018) seems better. --Infovarius (talk) 21:17, 5 March 2023 (UTC)
- @ArthurPSmith @Infovarius Done with help from Mormegil. See eg. Q56414974#P1448 Vojtěch Dostál (talk) 07:36, 7 June 2023 (UTC)