Updates / additions to Mande languages about dev

[drammock]

I've said this before, but I don't think we should update existing records based on new knowledge. Instead, we should add a new record for that language, tied to a particular resource, and assign it a higher trump number than the older and allegedly inaccurate records. It's more in keeping with the concept of a doculect, and I think it will serve us well in the long-term to hew closely to that concept. Consequently I'm labeling this as new inventory rather than inventory error.

[drammock]

@bambooforest can this issue be closed?

[bambooforest]

@drammock -- let me update a few things, e.g. we have two IDs in the bibtex file for the same reference here. Also, by now Vydrin has published the description (draft) that he sent us before:

• https://www.oxfordhandbooks.com/view/10.1093/oxfordhb/9780199609895.001.0001/oxfordhb-9780199609895-e-66

• https://link.springer.com/chapter/10.1007/978-3-319-44330-0_5

So I can add this as a new inventory. Where should we start adding new inventories -- under which existing raw-data or do we start a new "source"? If we're going for semantically coherent sources, PH, UZ, and GM could be basically all be one, since they were collected and curated in the same manner and have no coherent genealogical or geographic basis (well, perhaps GM, which focused on Africa and later SE Asia).

I don't think we should add to UZ because that work is not being funded by UZ. We could add to PH and keep the existing structure, using InventoryIDs that start after the last current one (in ER), but then we should update the aggregation script to sort the output from within sources, if it doesn't already do so.

Let me know what you think. I'll make a PR for updating this current issue.

[drammock]

I'd be inclined to start a new source and name it after Neuchâtel... NE or UN?