sirdanilot wrote:1. Simply using the standard geminate symbol ː would be less confusing, imo. The fact that they aren't really that long phonetically doesn't matter, as you are describing phonemes.
I was describing long phonemes that are realized, most of the time, as long phones. I don't know where you got the idea that I was saying they weren't realized phonetically long. I was just kind of switching it Americanist transcription since I'm used to doing it when dealing with this kind of language.
sirdanilot wrote:2. The distribution of dentals is surprising and one would wonder how it got that way. /tθ'/ might have come from *θ' or something.
3. The absence of /tsʼ/ is remarkable in a Salish-style language, especially considering (2) and the presence of other affricates.
This dental affricate/fricative inventory is actually attested in Coast Salish. It comes about from *c *c̓ > θ t̓θ. That's what happened here. Also what happened here was *k k̓ x > č č̓ š.
sirdanilot wrote:I like this idea in general. Anyway, [e:r]/[er] is a bit hard to pronounce, and one would expect to keep [i:r]/[ir] in this place (or something like [ɪər] or [ɪːr] anyway). This is attested in Dutch at least. Sure, if you already have a /er/ sequence one could keep it, but assimilation of a vowel to something that is actually harder to pronounce sounds strange to me; after all, the entire point of the assimilation is to make the sequence easier to pronounce.
I don't really understand what you're talking about here.
sirdanilot wrote:The position of the glottal sounds a bit weird to me. It would give stuff like kʼaʔk, qʷuːʔq and so.
Yup. That's what I intended. Happens all the time in Salish. It's a common way of forming plurals that goes back to Proto-Salish.