Wikidata talk:Wikiproject Lieder

From Wikidata
Jump to navigation Jump to search

P31 for Cycle

[edit]

Hi, @Robertgarrigos, Moebeus:.

Are cycles always interpreted as a single piece? That is, similar to an opera with arias ?. In other words, is the cycle a piece of music with parts? or is it a serie/cycle of several lieder?

The reason for my question is that I am preparing a set of changes to improve the music items, and avoid the ambiguity of certain concepts that causes misuse or overlap. I need to understand this aspect of lieder.

Strictly speaking, a set of pieces with some common feature or grouping criterion is cycle (Q245596) or musical series (Q24992028), but it is not a composition.

Thanks, --Amadalvarez (talk) 13:26, 5 June 2022 (UTC)[reply]

Currently we model song cycles as p31 = "musical work/composition" with form = song cycle, etc. We use "has part" to list individual songs/parts. This is in line with how these are treated by MusicBrainz, IMSLP, Allmusic. For a report of the different forms used with "musical work/composition", see: https://www.wikidata.org/wiki/Wikidata:WikiProject_Music/Statistics/Works_by_form
Moebeus (talk) 13:42, 5 June 2022 (UTC)[reply]
Great !. Thanks. Amadalvarez (talk) 05:06, 10 June 2022 (UTC)[reply]
Thanks @Moebeus. I agree. Robertgarrigos (talk) 16:15, 15 August 2022 (UTC)[reply]

where to put the lyrics

[edit]

@Amadalvarez, Moebeus:

I now have a doubt on which property use to put the lyrics on: either based on (P144), which allows us to use derivative work (P4969) on the poem item, or has lyrics (P6439), which doesn't require a derivative. I like P144 because of the derivative requirement and thus would make it easy to search for, but I'm not an expert on searching. What are your thoughts on this? Thanks. --Robertgarrigos (talk) 16:14, 15 August 2022 (UTC)[reply]

Inverse properties are what we're trying to move away from, so I would maintain has lyrics (P6439) is the more future proof way. Using a gadget like "relateditems" will make the relationship clear enough imo, while removing the need to maintain data "on both sides". Moebeus (talk) 17:30, 15 August 2022 (UTC)[reply]

new subclass for Lied

[edit]

@Amadalvarez, Moebeus, Beat_Estermann:

After some conversations with Beat Estermann (Q63216809) I'm proposing a new subclass lyrico-musical work (Q114586269) as P31 for Lied and Lieder cycle as analogy of dramatico-musical work (Q58483083) for Opera works. I believe Lieder have enough entity to have a subclass for it and I would extend it to Lieder cycle also. This would be an example of a Lied: Die Forelle (Q630565).

If there is no opposition to this new subclass, I will apply the changes to other Lieder and cycles. Robertgarrigos (talk) 15:29, 11 October 2022 (UTC)[reply]

As of today (), I decided not to change those values but add the new one. Is this a better approach? Robertgarrigos (talk) 15:29, 21 October 2022 (UTC)[reply]

Lied or Kunstlied

[edit]

Hi @Amadalvarez @Moebeus @Beat Estermann @Anna Costal,

There are two items, art song (Q4797274) & Lied (Q216860), which represent virtually the same concept. We decided to use Lied (Q216860) as our main form, but we might need to discuss whether it is necessary/possible to merge those two items. What would be your thoughts? Robertgarrigos (talk) 10:51, 21 October 2022 (UTC)[reply]

While merging of similar concepts is often made impractical/impossible by sitelinks, there is nothing stopping you from picking one over the other in your datamodel if there is consensus. Enforcing the model can be achived with a combination of constraints and Wikidata usage instructions, letting users know to use one and not the other. For example, you could "empty out" all uses of art song and move them over to Lied, and then make sure art song stays unused by setting up a bot to automatically switch all future uses to Lied. Moebeus (talk) 11:20, 21 October 2022 (UTC)[reply]
I agree with @Moebeus. It's not just a linguistic matter but pointing to different concept, as in frwiki, where Lied (Q216860)=chant & art song (Q4797274)=genre. This causes a sitelink for each one.
To me is important keep in mind that are (almost) sinonimous, using said to be the same as (P460) on both of them, and using the instructions suggested by @Moebeus. Thanks ! Amadalvarez (talk) 14:09, 21 October 2022 (UTC)[reply]
Thank you, @Moebeus and @Amadalvarez. You are very helpful! :-) Robertgarrigos (talk) 16:38, 21 October 2022 (UTC)[reply]