Vanuatu Voices Introduction Paper

lexibank / vanuatuvoices

Sound-Comparisons Vanuatu

Other

3 stars 1 forks source link

Vanuatu Voices Introduction Paper #23

Open LinguList opened 2 years ago

LinguList commented 2 years ago

The basic paper to introduce VV data should provide:

[ ] CLLD app with lexical data, phoneme inventories, and computed featueres
[ ] #25
[ ] additionally, features like the distance of Proto-Oceanic can be computed

LinguList commented 2 years ago

Further plans can be discussed here.

xrotwang commented 2 years ago

So, the plan would be to enrich CLDF creation, creating

a CLDF StructureDataset with the phoneme inventories
a CLDF StructureDataset with the computed features

as well? If so, should cognate detection be added as well?

LinguList commented 2 years ago

That would be cool, I assume.

xrotwang commented 2 years ago

And would the paper come with a derived dataset for inventories, features, etc. or would this be added to the CLDF data in this repository?

LinguList commented 2 years ago

When discussing how that paper should be structured, we realized it should have features calculated as in lexibank, so the paper would also be similar, with the difference being sound data and that the data was collected from original field work. Somehow, such a dataset CRIES for automatically inferred features, I think.

LinguList commented 2 years ago

I would say that exactly this should be done, yes. Extended CLDF data should be added to the repository.

xrotwang commented 2 years ago

I guess, cognate detection would be really nice, because that would provide another opportunity to highlight the soundfiles: Having an audio-player next to the alignments of a cognate set could be interesting!

xrotwang commented 2 years ago

Would be a small step towards linking sound change and sound signal - if only visually.

LinguList commented 2 years ago

Yes, we could definitely try it. And if we use a conservative threshold, we'd emphasize: these are the "good" cognates, where we have confidence, more will be added through manual annotation some time in the future, or similar.

maryewal commented 2 years ago

Linking the sound files and cognate sets would be excellent! Timo and I will discuss the features when we meet early next week together and will report back.