Property talk:P5980
Latest comment: 5 years ago by Liamjamesperritt in topic Proposal for use on senses
Documentation
object sense
(qualifier) statement applies only to listed senses of statement's object
(qualifier) statement applies only to listed senses of statement's object
| Data type | Sense | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Example | no label (L22603) → no label (L8627-S1) no label (L22604) → no label (L3340-S1) no label (L22605) → no label (L3537-S1) | |||||||||
| See also | object form (P5548), semantic derivation of (P12410) | |||||||||
| Lists | ||||||||||
| Proposal discussion | Proposal discussion | |||||||||
| Current uses |
| |||||||||
| Search for values |
[create Create a translatable help page (preferably in English) for this property to be included here]
Scope is as qualifier (Q54828449): the property must be used by specified way only (Help)
Exceptions are possible as rare values may exist. Exceptions can be specified using exception to constraint (P2303). List of violations of this constraint: Database reports/Constraint violations/P5980#Scope, SPARQL
Allowed entity types are Wikibase lexeme (Q51885771), Wikibase sense (Q54285715), Wikibase item (Q29934200): the property may only be used on a certain entity type (Help)
Exceptions are possible as rare values may exist. Exceptions can be specified using exception to constraint (P2303). List of violations of this constraint: Database reports/Constraint violations/P5980#Entity types
Proposal for use on senses
edit@Denny, ArthurPSmith, Jura1, Deryck Chan, KaMan: @Vive la Rosière, EncycloPetey, Tobias1984:
This is following on from the discussion at Wikidata:Property proposal/periphrasis. This proposal is to use object sense (P5980) instead of the proposed periphrasis property.
Currently, using object sense (P5980) as a qualifier for derived from lexeme (P5191) and combines lexemes (P5238) at the Lexeme level is good and important, however it is often the case that a whole lexeme is not derived from a single sense of an ancestor word, but that each of the lexeme's senses are derived from a different sense of the ancestor word. Since it is currently only used at the Lexeme level, there is no way at the moment to describe what ancestor senses a sense is made up of.
Proposal: Could we add an additional use-case for object sense (P5980) where the subject is a Sense (i.e. used also at the sense level)?
- (This is in addition to it's current usage as a qualifier at the Lexeme level.)
- This probably should only be used on Senses of Lexemes that contain a derived from lexeme (P5191) or combines lexemes (P5238) statement.
- This would give us a tool for defining morphosemantic relations between word senses.
- I believe this would also solve what Wikidata:Property proposal/periphrasis is trying to solve, without having to create a new property.
Alternative: Create a new property that fulfills this morphosemantic function.
Cheers,
Liamjamesperritt (talk) 12:27, 5 May 2019 (UTC)
- Liamjamesperritt: I don't feel strongly about this. But I'm not sure what this proposal is about - are you proposing to change the scope of P5980 or the proposed "periphrasis" property? P5980 is called "derived from sense" and you're proposing to move something something from Lexeme to Sense... Deryck Chan (talk) 14:49, 6 May 2019 (UTC)
- @Deryck Chan: Apologies for the lack of clarity. The proposal is to use object sense (P5980) on senses rather than implementing the proposed "periphrasis" property. I'm not proposing we move it to the sense level, but just that the Sense level usage should be allowed in addition to the Lexeme level. Regards. Liamjamesperritt (talk) 04:08, 7 May 2019 (UTC)
- So an example where it's being used now is lifeboat (L37082), pointing to one sense of "life", and it applies to the whole lexeme "lifeboat". Do you have an example where it would apply just to a sense? But I don't see any reason why it couldn't be used at the sense level if appropriate right now. ArthurPSmith (talk) 20:43, 6 May 2019 (UTC)
- @ArthurPSmith, Deryck Chan: I think in a lot of cases, the whole lexeme is defined by a single sense, but there are some where that is not the case. I think adverbs formed from adjectives are common instance of this. For example: offensively (L12735) combines lexemes (P5238) offensive (L12734) and -ly (L28890). Even though the sense -ly (L28890-S1) applies to the whole lexeme, the sense offensively (L12735-S1) is derived from offensive (L12734-S1), whereas the sense offensively (L12735-S2) is derived from offensive (L12734-S2). Liamjamesperritt (talk) 04:36, 7 May 2019 (UTC)
- Thanks Liam. Yes your proposal makes sense. I'm out of my depth here and am happy to defer to your expertise on this matter. Deryck Chan (talk) 10:44, 7 May 2019 (UTC)
- Yes, I support this use of this property. ArthurPSmith (talk) 13:03, 7 May 2019 (UTC)
- @ArthurPSmith, Deryck Chan: I think in a lot of cases, the whole lexeme is defined by a single sense, but there are some where that is not the case. I think adverbs formed from adjectives are common instance of this. For example: offensively (L12735) combines lexemes (P5238) offensive (L12734) and -ly (L28890). Even though the sense -ly (L28890-S1) applies to the whole lexeme, the sense offensively (L12735-S1) is derived from offensive (L12734-S1), whereas the sense offensively (L12735-S2) is derived from offensive (L12734-S2). Liamjamesperritt (talk) 04:36, 7 May 2019 (UTC)
- I think it might work for some of the uses (samples at Wikidata:Property proposal/periphrasis), but not all of them (maybe you can demonstrate the contrary). The combination suggested here could only work in one direction. As the underlying properties aren't meant for writing definitions, I'm not sure if the result would be very consistent. --- Jura 07:13, 8 May 2019 (UTC)
- @Jura1: You are correct that it will only work one direction, and that is how I believe it should be. If you define the words in both directions, you will end up with recursive definitions. For example: traveller (one who travels) --is defined by--> travel (go from place to place) + -er (person that performs the root action). If you then go the other direction: travel (go from place to place) --is defined by--> traveller (one who travels) + <action performed by agent>, then not only is that redundant (as the second direction can be inferred from the first) but then the full definition for 'traveller' becomes recursive; i.e. "A traveller is a person who performs the action that is performed by a traveller". The better way to do it, is to only go one direction ('traveller' --> 'travel'), and then use a different word to define 'travel', e.g. travel (go from place to place) --is defined by--> go (move; change location) + abroad (in different directions; over a wide area). That way the meaning of words begins to become apparent. Liamjamesperritt (talk) 12:19, 8 May 2019 (UTC)
- I don't think it's an issue. It tends to happen, whether it's done in a gloss, a full definition, or a structured way. --- Jura 12:19, 14 May 2019 (UTC)
- @Jura1: I'm still not convinced by the idea of the periphrasis property. It seems a little bit too messy. Maybe we can re-think it a little bit and tighten up the proposal? FYI, all the examples given in the proposal (excluding bidirectional definitions, which I still think are unnecessary) can be achieved with object sense (P5980), as I'll demonstrate below:
- @Jura1: You are correct that it will only work one direction, and that is how I believe it should be. If you define the words in both directions, you will end up with recursive definitions. For example: traveller (one who travels) --is defined by--> travel (go from place to place) + -er (person that performs the root action). If you then go the other direction: travel (go from place to place) --is defined by--> traveller (one who travels) + <action performed by agent>, then not only is that redundant (as the second direction can be inferred from the first) but then the full definition for 'traveller' becomes recursive; i.e. "A traveller is a person who performs the action that is performed by a traveller". The better way to do it, is to only go one direction ('traveller' --> 'travel'), and then use a different word to define 'travel', e.g. travel (go from place to place) --is defined by--> go (move; change location) + abroad (in different directions; over a wide area). That way the meaning of words begins to become apparent. Liamjamesperritt (talk) 12:19, 8 May 2019 (UTC)
- Example 7 can be easily inferred from the above information, and is redundant.
- I feel like this is the most effective way to model these definitions, as the meaning of every single example the result of morphological derivation. Please feel free to give an example for which object sense (P5980) will not suffice. Liamjamesperritt (talk) 08:41, 30 May 2019 (UTC)
- I understand your reasoning, but personally I prefer an approach people can use to write definitions without knowing the detailed derivation or without subscribing to a specific view on derivation. --- Jura 10:53, 31 May 2019 (UTC)
- @Jura1: I suppose that object sense (P5980) does not explicitly refer to definitions, and that is what you are trying to achieve, so I understand your reasoning as well. In that case, here's a potential example where object sense (P5980) doesn't suffice:
- Example: kennel (shelter for dogs) --(definition (PXXX))--> shelter (structure that provides cover or refuge) [qualified with --(attribute for definition (PXXX))--> dog (mammal used as pet and related to the wolf)] + [qualified with --(link for definition (PXXX))--> for (intended to belong to or be used by)]
- Do you see this as a potential usage for this definition property? Liamjamesperritt (talk) 12:21, 31 May 2019 (UTC)
- There are a few usual ways of writing definitions. Depending on the objective and the sense, users prefer one or the other. I think lexemes could easily include any of them in a structured way. We already have proposals/properties for some of them, but lack others. The one you outline could fit the hyperonym property we already have. The tautological or periphrastic one is slightly different and what the other proposal is for. Anyways, I don't really see reasons why we should limit users to one or the other ways of writing definitions nor do I have such an agenda for other contributors. --- Jura 18:07, 1 June 2019 (UTC)
- Fair enough, I suppose. Liamjamesperritt (talk) 03:00, 8 June 2019 (UTC)
- There are a few usual ways of writing definitions. Depending on the objective and the sense, users prefer one or the other. I think lexemes could easily include any of them in a structured way. We already have proposals/properties for some of them, but lack others. The one you outline could fit the hyperonym property we already have. The tautological or periphrastic one is slightly different and what the other proposal is for. Anyways, I don't really see reasons why we should limit users to one or the other ways of writing definitions nor do I have such an agenda for other contributors. --- Jura 18:07, 1 June 2019 (UTC)
- I feel like this is the most effective way to model these definitions, as the meaning of every single example the result of morphological derivation. Please feel free to give an example for which object sense (P5980) will not suffice. Liamjamesperritt (talk) 08:41, 30 May 2019 (UTC)