i'au is not a true famyma'o at all. All it does, is two things:

1. Automatically insert all famyma'o required to return to the top level of
the sentence. In other words, everything .i will add except the last
vau. (It's not meant to be more powerful than .i; in other words, it
will not close lu...li'u, to...toi, tu'e...tu'u, or fu'e...fu'o, and
perhaps others I'm not aware of)

2. Any attitudinal-groups (UI/NAI/CAI) that attach to it scope over the
entire sentence. Thus, it does behave much like that last vau with
respect to them, even though it doesn't actually add it.

One interesting result is that mi broda do i'au i'au i'au i'au zo'o is
grammatical. i'au works even when there isn't actually any famyma'o for
it to add.

Comment #2: Re: Current (and hopefully final) understanding of the grammar of i'au

Curtis W Franks (Thu Jul 16 21:01:29 2015)

spheniscine wrote: > i'au is not a true famyma'o at all. All it does, is two things: > > 1. Automatically insert all famyma'o required to return to the top level of > the sentence. In other words, everything .i will add except the last > vau. (It's not meant to be more powerful than .i; in other words, it > will not close lu...li'u, to...toi, tu'e...tu'u, or fu'e...fu'o, and > perhaps others I'm not aware of) > > 2. Any attitudinal-groups (UI/NAI/CAI) that attach to it scope over the > entire sentence. Thus, it does behave much like that last vau with > respect to them, even though it doesn't actually add it. > > One interesting result is that mi broda do i'au i'au i'au i'au zo'o is > grammatical. i'au works even when there isn't actually any famyma'o for > it to add. I agree with (1) and the later observation but I kind of disagree with how (2) should work. If .i'au really just takes the bridi level back to that of the original bridi, then it just closes all of the most recent nesting. Any UI immediately following it should apply to the last full structure in the original bridi (at level 0), which is what would happen if all of the necessary famyma'o had been inserted. The only way to apply the scope to the whole bridi is via vau or putting UI immediately after .i. For example, "lo broda be lo brode be lo brodi .i'au .ui" has ".ui" apply to all of the "lo broda" construct (rather than just lo brodi).

Comment #3: Re: Current (and hopefully final) understanding of the grammar of i'au Jonathan (Fri Jul 17 05:57:11 2015) Well, part of the problem with that is that the original purpose of i'au was indeed meant to also represent that final vau (It could even have been a true famyma'o, closing the "sentence" construct), for the purposes of attaching an attitudinal to the entire sentence as an afterthought. That is still its most common usage, and I wish to preserve that meaning in the use-case represented by do sidju mi lo nu mi zenba lo ni ricfu i'au ui. Perhaps though, it could just have the attitudinals scope over everything *previous* to it in the sentence. Comment #5: Re: Current (and hopefully final) understanding of the grammar of i'au Curtis W Franks (Sat Jul 18 06:07:44 2015) spheniscine wrote: > Well, part of the problem with that is that the original purpose of i'au > was indeed meant to also represent that final vau (It could even have > been a true famyma'o, closing the "sentence" construct), for the purposes > of attaching an attitudinal to the entire sentence as an afterthought. That > is still its most common usage, and I wish to preserve that meaning in the > use-case represented by do sidju mi lo nu mi zenba lo ni ricfu i'au ui. > > Perhaps though, it could just have the attitudinals scope over everything > *previous* to it in the sentence. Sure, it can function howsoever we (as a community) desire (with you, as the creator wielding considerable power in the process, at least initially). But, morally, I feel like (1) and (2) are in conflict, which is why I said "should" instead of "could". I just do not see how, if I were to have a whole bunch of nested bridi levels and if I then closed them all and followed that closure with .ui, that that .ui would apply to anything other than the last whole structure uttered- that is how UI normally works- all that happened is that we jumped back to the a main bridi level. But I guess that this word could also start a metalinguistic comment for UI which says "hey, if any UI immediately follow this word, apply them to the whole bridi; otherwise, get on with your life". Concerning your last point: The other day, while showering, I realized the desirability of having a universal UI terminator that would wrap up the scope of any open UI's. Basically, UI would apply to the last whole construct mentioned as usual and, if that construct were still open, then their scope would continue until the terminator, thus not applying to the remainder of the construct (and the terminator would kill all UI, not just the most recently mentioned one). I think that your idea is a specific case of this one, wherein the construct is the whole bridi itself (so the word preceding the UI would be either .i or vau). I could not quite word it/make it work technically, but I still think that it could be useful. Comment #6: Re: Current (and hopefully final) understanding of the grammar of i'au Jonathan (Sun Jul 19 12:22:46 2015) Actually I'm beginning to think you may have a point. Consider lo nu ko'a broda ko'e lo nu ko'i ko'o brode i'au po'o broda brode po'o would be more naturally interpreted to attach to that last construct, so why should it work if that nested-NU occurs at the end? This does complicate things somewhat. Both these usages want something short (hence my booking of the precious, and atypical, monosyllabic cmavo iau for this), but cmavo space is an issue. Comment #9: Re: Current (and hopefully final) understanding of the grammar of i'au Curtis W Franks (Sun Jul 19 17:54:32 2015) spheniscine wrote: > Actually I'm beginning to think you may have a point. Consider > > lo nu ko'a broda ko'e lo nu ko'i ko'o brode i'au po'o broda brode > > po'o would be more naturally interpreter to attach to that last > construct, so why should it work if that nested-NU occurs at the end? > > This does complicate things somewhat. Both these usages want something > short (hence my booking of the precious, and atypical, monosyllabic cmavo > iau for this) .i'au and .iau can be different but closely related. Comment #7: Re: Current (and hopefully final) understanding of the grammar of i'au Jonathan (Sun Jul 19 12:25:06 2015) s/ should it work / should it work differently Comment #8: Re: Current (and hopefully final) understanding of the grammar of i'au Jonathan (Sun Jul 19 14:34:28 2015) This brings us to the problem of attaching a UI to a sumti (especially important with po'o and ji'a, as the meaning of a sentence can change a lot depending on what they're attached to) "fa po'o lo broda cu brode" is commonly used to mean "lo broda ku po'o brode", but may be somewhat problematic... the "correct" meaning is probably "fa, and no other places". This meaning may be awkward (fi'a is hardly used), but replace fa with a sumtcita and the problem becomes clear: "ca po'o lo broda cu brode"... "only when broda did brode happen". However, the ku version is often inconvenient, especially if the sumti has several nestings. There is actually a solution we found. "ke po'o lo broda cu brode" fronts the UI without needing to attach to the sumti place. This is an extension of ke useful to do complex things with connectives like "ke ko'a ce ko'e ke'e ce ko'i". Unfortunately, the camxes parser has a bug, and doesn't parse it correctly. (The more-experimental zantufa and alta parsers do parse them, however.) Comment #4: Re: Current (and hopefully final) understanding of the grammar of i'au Jonathan (Fri Jul 17 06:20:27 2015) Well, in other words, i'au is *not* cu'au xi da'a. It grammatically acts *like* cu'au xi da'a, but treats attitudinal scope differently.