My internet access is not the free-for-all it usually is (since dialup is sort of slow…), but there are several comment of various ages that I’ve been meaning to reply to. Sorry for putting all these in the one post. The headings go to the original posts.
Sarah‘s comments on Dalabon are relevant to Bardi, especially for spatially specific vs non-specific aspects of the system. Nyalab and balab are spatially specific, and so is nyoon in my modern data, but not always clearly so in the old data. They get local case marking. Jarri does weird stuff. Thanks for the ref to Enfield’s article.
In this case, James, I’m a bit more sanguine about assuming that there aren’t other terms lurking in the system, since there’s so much textual data where all of these terms show up (with varying frequency), and these terms only.
Grammar organisation, form and function
James, for organising around speech acts I was thinking along the lines of, for example, putting all the information about questions in one place. The topic of questions includes intonation, morphology, syntax, information structure, and pronouns. There’s information about questions in five chapters of the grammar, depending on whether they’re content questions, polar interrogatives, whether the focus of the discussion is the forms used, the intonation pattern that they appear with typically, and so on.
DIY language of the week
Sally, in the semantics literature I’ve seen on this, in Hindi at least the verb on its own has the supposition. That is, if someone said ‘He killed Kenny’, the presumption would be that Kenny died, unless it was explicitly cancelled. Not sure about this for Bardi. In the texts there’s sometimes passages like “he killed him, and he dropped to the ground dead”, or “he killed him, finished”.
I just checked for inanimates and the only possibilities are the phrase booroo inggardina aalga ‘the day finished’ (the sun ground-entered) and some complex predicates, which don’t count because the preverb modifies the semantics of the light verb.