Wikidata:Property proposal/pronunciation variety

pronunciation variety

Originally proposed at Wikidata:Property proposal/Lexemes

   Done: pronunciation variety (P5237) (Talk and documentation)

Description	qualifier for IPA transcription (P898) or spoken text audio (P989) to indicate the associated spoken language variant
Data type	Item
Domain	qualifier for IPA transcription statements on lexemes
Example	dog (en) IPA transcription (P898) /dɒɡ/ → Received Pronunciation (Q7985) dog (en) IPA transcription (P898) /dɑɡ/ → cot–caught merger (Q28401088)
Planned use	only use as qualifier

Motivation

IPA transcription (P898) should work fine to indicate pronunciations, but we need a qualifier for the spoken language variant associated with that pronunciation. ArthurPSmith (talk) 13:17, 17 April 2018 (UTC)

Discussion

• I  Support this property but what related qualifiers could this be used with? For instance we have 44 subproperties of location (P276), might one of these be used to indicate the typical region of some pronunciations? Does this also relate to spoken text audio (P989)? -- JakobVoss (talk) 13:31, 17 April 2018 (UTC)

• The location information would be appropriate on the pronunciation variant item I would think, not so much on the lexeme itself. And I guess spoken text audio (P989) makes sense also for this, thanks, I've added it in the description. ArthurPSmith (talk) 13:41, 17 April 2018 (UTC)

• At the test Wiki this property is used for forms. I must admin the data model requires some time to get used to it. -- JakobVoss (talk) 13:48, 17 April 2018 (UTC)

• Good point, I moved this to the Forms section. ArthurPSmith (talk) 15:30, 17 April 2018 (UTC)
• By the way, I don't like the way rhyming was proposed in that test wiki - I would think each form should give an item or string that describes its ending, so you could automatically match up all words in a language that rhyme, rather than requiring each rhyming word to link to each other. But there certainly may be technicalities regarding rhyming that I'm not familiar with and that make this more complicated... ArthurPSmith (talk) 19:13, 17 April 2018 (UTC)
• Yes, I'd also oppose a rhyming property as this can better be queried automatically depending on which type of rhyme you want. -- JakobVoss (talk) 06:38, 20 April 2018 (UTC)

• I agree with you as I think too that the type of rhyming can easily be infers with a query. Cdlt, VIGNERON (talk) 18:18, 24 April 2018 (UTC)

 Question why not use P134 (P134)? NMaia (talk) 23:24, 21 April 2018 (UTC)

@NMaia: That property (which is tagged for deletion) links a parent language to dialects of it, which is quite a different sort of relationship to the language variety associated with pronunciation of a single word. ArthurPSmith (talk) 13:05, 23 April 2018 (UTC)

We could broaden its scope to have it be used as a qualifier as well, I don't see why not. NMaia (talk) 14:09, 23 April 2018 (UTC)

It's not a matter of broadening scope, it's a completely different meaning. ArthurPSmith (talk) 19:38, 23 April 2018 (UTC)

@ArthurPSmith, VIGNERON: I created the property although we should not make it common to create properties such easily without additional discussion. I hope this is ok, given the infancy of the lexicographical data community at Wikidata. -- JakobVoss (talk) 20:52, 1 June 2018 (UTC)