Robert Forkel

So, as a small experiment, I created CLDF data as proposed above using cldfbench. The repos is here: https://github.com/cldf-datasets/steinertthrelkeldmodals/ The code in [the python module](https://github.com/cldf-datasets/steinertthrelkeldmodals/blob/main/cldfbench_steinertthrelkeldmodals.py) turns raw data as in...

Oh, and regarding my claim that sqlite may be used to query this data, too: ```shell $ cldf createdb cldf/StructureDataset-metadata.json modals.sqlite INFO loaded in modals.sqlite $ sqlite3 modals.sqlite SQLite version...

I agree that this kind of data is common. The [clld database framework](https://clld.readthedocs.io/en/latest/) (which was one of the inspirations for CLDF) has entities `Unit`, `UnitParameter` and `UnitValue` for this exact...

CLDF allows adding more custom tables to a dataset. To put such an extension on the "standards track", you'd create such a public dataset and then we'd put it to...

Oh and the "multi-dimensional feature" idea is *not* really supported by any software yet. To some extent that's a relic of a time where we wanted to cram as much...

ad 1) Yes, I guess `Wordlist` would make sense - although a `StructureDataset` with single parameter `ModalExpression` would also make sense. ad 2) That would be the `UnitParameter_ID` in a...

@LinguList could you have a look?

I think we have to be careful with slice notation, since this always requires well defined segmentation. What would be the separator for phrases?

I think having parameters of different kinds in the same `parameters.csv` isn't too much of a problem, as long as they are "measured" for the same things - and that...

This kind of dataset would be interesting to investigate how semantics and grammar relate to each other, after all, right? So it seems separating these concerns into - say -...