spheniscine wrote:
> In (iuro'o toi'e uinai), does toi'e attach to just (ro'o) or (iuro'o)?

Hmmm... I am not sure. I suppose that it is a matter of what we
want/practicality. Tell me what you think of this: It is my opinion that it
should apply to only ro'o since it is the single immediately previous UI
cmavo and the definition does not mention clusters; this has a practical
side to it as well: if you want it to apply to a cluster, under this
interpretation, you can always force them into one unit via use of fu'e
and fu'o, whereas if it automatically applied to the entirety of the
immediately previous cluster, then there would be no easy way (aside from
bracketing just one individual cmavo in the cluster with it via fu'e and
fu'o) to make it apply to just one part of the cluster, which is
potentially desirable. I also prefer to bracket longer units, rather than
smaller ones.

Comment #4: Re: Question

Jonathan (Fri May 22 06:49:37 2015)

krtisfranks wrote: > spheniscine wrote: > > In (iuro'o toi'e uinai), does toi'e attach to just (ro'o) or (iuro'o)? > > Hmmm... I am not sure. I suppose that it is a matter of what we > want/practicality. Tell me what you think of this: It is my opinion that it > should apply to only ro'o since it is the single immediately previous UI > cmavo and the definition does not mention clusters; this has a practical > side to it as well: if you want it to apply to a cluster, under this > interpretation, you can always force them into one unit via use of fu'e > and fu'o, whereas if it automatically applied to the entirety of the > immediately previous cluster, then there would be no easy way (aside from > bracketing just one individual cmavo in the cluster with it via fu'e and > fu'o) to make it apply to just one part of the cluster, which is > potentially desirable. I also prefer to bracket longer units, rather than > smaller ones. I think that might be potentially problematic. What of modifiers like nai and cai then? In my understanding, UI-cmavo groups naturally stick together like rice; by default, they clump together and all simultaneously apply to the last lexical item. Additionally, I don't think fu'e / fu'o is an adequate solution, as in my understanding fu'e / fu'o act as anchors for UI-cmavo to attach to (similar to how a UI-cmavo attaching to a le or ku would affect the entire construct. For example, if I wanted to give a long speech about the things I hope (a'o) for, I might use group the entire thing under a fu'e ... fu'o, and attach a'o either to fu'e or fu'o. Thus, trying to use fu'e / fu'o to apply UI-cmavo to other UI-cmavo may cause ambiguity or undefined behavior, since which are the UI-cmavo that attach and which are the UI-cmavo that are attached to? Additionally, fu'e / fu'o requires forethought and is thus difficult in practical usage. Perhaps a solution is to create yet another TOIhE cmavo, that acts as a "null" parenthesis (similar to how ke works for tanru) to separate UI cmavo. For example, in (au *koi'e ui toi'e iu), iu would apply to ui and not au, while without koi'e, it'd apply to both.

Comment #5: Re: Question Curtis W Franks (Fri May 22 07:51:15 2015) spheniscine wrote: > krtisfranks wrote: > > spheniscine wrote: > > > In (iuro'o toi'e uinai), does toi'e attach to just (ro'o) or (iuro'o)? > > > > Hmmm... I am not sure. I suppose that it is a matter of what we > > want/practicality. Tell me what you think of this: It is my opinion that > it > > should apply to only ro'o since it is the single immediately previous > UI > > cmavo and the definition does not mention clusters; this has a practical > > side to it as well: if you want it to apply to a cluster, under this > > interpretation, you can always force them into one unit via use of fu'e > > > and fu'o, whereas if it automatically applied to the entirety of the > > immediately previous cluster, then there would be no easy way (aside from > > > bracketing just one individual cmavo in the cluster with it via fu'e > and > > fu'o) to make it apply to just one part of the cluster, which is > > potentially desirable. I also prefer to bracket longer units, rather than > > > smaller ones. > > I think that might be potentially problematic. What of modifiers like nai > and cai then? In my understanding, UI-cmavo groups naturally stick > together like rice; by default, they clump together and all simultaneously > apply to the last lexical item. > > Additionally, I don't think fu'e / fu'o is an adequate solution, as in > my understanding fu'e / fu'o act as anchors for UI-cmavo to attach to > (similar to how a UI-cmavo attaching to a le or ku would affect the > entire construct. For example, if I wanted to give a long speech about the > things I hope (a'o) for, I might use group the entire thing under a > fu'e ... fu'o, and attach a'o either to fu'e or fu'o. > > Thus, trying to use fu'e / fu'o to apply UI-cmavo to other UI-cmavo may > cause ambiguity or undefined behavior, since which are the UI-cmavo that > attach and which are the UI-cmavo that are attached to? Additionally, > fu'e / fu'o requires forethought and is thus difficult in practical > usage. > > Perhaps a solution is to create yet another TOIhE cmavo, that acts as a > "null" parenthesis (similar to how ke works for tanru) to separate UI > cmavo. For example, in (au *koi'e ui toi'e iu), iu would apply to ui and > not au, while without koi'e, it'd apply to both. Write up that definition. I will support it. Even if other solutions work with already-present words and modifications/additions to the existing grammar, an explicit unambiguous and clearly-functioning mechanism could be helpful, at least in the experimental stage.