WIP: Kalathi (NP: morphosyntax basics)

[So Haleza Grise]

I had to write a bunch of code to get the right characters at the right codepoints.

Yeesh, that sounds like a lot of work. I use texmaker on Mac and it doesn't seem to have any font installation settings. Your hard work paid off though, it is a really beautiful font and it looks great.

[nebula wind phone]

With XeLaTeX (which texmaker appears to support) you don't need to do anything special for font support, you can just use any font installed on your system. Which in particular means you can use the SIL fonts, Charis or Gentium, with basically no hassle, and those both have very nice IPA characters.

Stone Sans Phonetic was extra work, but only because it's pre-Unicode.

[So Haleza Grise]

With XeLaTeX (which texmaker appears to support) you don't need to do anything special for font support, you can just use any font installed on your system. Which in particular means you can use the SIL fonts, Charis or Gentium, with basically no hassle, and those both have very nice IPA characters.

Stone Sans Phonetic was extra work, but only because it's pre-Unicode.

Thank you for this! I have Gentium.

[Sevly]

wow so this was a pleasure to read! i love the way you've described the language, and the use of colour in the examples is a deceptivey simple but important touch that makes everything both clean and informative. i'm really curious to see more of the morphosyntax, especially how the (written) word boundaries are mismatched with a surface-level understanding of the conceptual categories—really seems ripe for deep analysis

tl;dr this is good. very.

[chris_notts]

The big open question for me right now is what limits to put on verb serialization. For language-specific grammar reasons, I want to limit them to ones where each verb has the same subject. (Object markers and oblique markers are affixes that attach to a single verb; subject markers are clitics that attach to a whole verb sequence; so a sequence with two different subjects would need two clitics, and I don't like that.)

This isn't necessarily true. There are languages with serial verb constructions either with a single subject inflection despite the SVC being switch subject, or with the same subject inflection on every verb even though there are underlyingly different subjects. The crucial point is that you want tightly integrated SVCs with strictly contiguous verbs, and in many languages which have these the SVC as a whole has its own transitivity and argument structure.

The tricky thing with such tightly integrated structures is that over time they tend to merge to form compounds (single phonological words), or in the case of asymmetric SVCs into verbal affixes.