Diary 2024-03-18

Tags:
By Max Woerner Chase

Starting this post early because I had stuff that can go in it on my mind...

I had a look at just a few phonology sections on Wikipedia to refresh my memory, and from looking at them it looks clear to me that "just subset the IPA charts" is not going to work well. My inclination is to say "people can come up with whatever rows and columns they like, although I'd like to provide some sensible defaults". And "it should be possible to nest heading cells in verious ways, and I'm not confident right now that I can figure out the proper manifestation of the nesting automatically". What goes in cells should be relatively freeform, but I'd like the ability to choose between "centered, left-justified, right-justified, some text left- and some text right- justified" on a per-cell basis. As far as IPA symbols, that should be entirely freeform, but it would be good to have some helpers for stuff like aspiration and devoicing.

And those helpers look to be a minor nightmare to put together, because the IPA symbols have a lot of stuff going on, in terms of unicode representation. And Wikipedia isn't helping, with the frankly bizarre styling it inflicted on the table I'd like to use, which seems to make it entirely unsuitable for printing. (Also, there's something going on with the click consonants that doesn't look right. I don't have plans for click consonants in Tsetatsatal, but it would offend my sensibilities if I ignored them or got them wrong. That said, it looks like so much of the details of click phonemes is represented through diacritics, that it looks to me like the sensible course of action initially would be to have a separate helper just for getting the IPA character for the place of articulation, and punt on filling in the rest of the details for now.)

Those digressions aside, I think I have to acknowledge that, while it would be cool to build up some of the IPA consonant representations in a sort of refinement process, it makes much more sense to create a hardcoded lookup table. Sadly, even this isn't totally straightforward, because there's a lot of stuff going on in the various IPA tables I'm looking at, and they also don't agree with each other in some significant ways. Stuff like "do we call out dental plosives as distinct from alveolar plosives?" or "so what should the deal with sibilants be?" and especially "so what does it mean in these charts when the line between cells ends halfway between?" At this stage, it seems like it would make more sense to simply embed the IPA chart in the supporting cells and be all "copy out the symbols that you want".

Plausible other things to work with would be stuff like "combining mark/spacer applier" or "superscript converter".

Looking at what I can find easily, there are a few issues, which I'll need to take some time to remedy. I'll get to it in a few days, hopefully. For now, I want to get off my computer.

Good night.