It might be useful for this word to actually attach to exactly the last
mentioned term/word in a mekso phrase (including PA, VUhU, etc.; I guess
that there could be issues of it attaching to another poi'a, but that too
may be desired by the speaker and is not really that big of a problem (it
should not be forbidden)).

Another important issue is grouping. Currently, the word attaches to
exactly the most recent PA. With the aforementioned expansion, it would
attach to the most recent word in a mekso expression. But what happens if
the speaker wants to attach it to the most recent occurrence of a unit of
mekso expression that is longer than one word? For example, what if one
wants to attach it to the number twenty-three rather than simply the number
three (in decimal)? I propose another word which begins a bracket for these
units; then, poi'a would attach to everything mentioned between it and
the last unmatched occurrence of this bracket-initializing word, treating
the intervening mekso subexpression as a single whole and indivisible unit
for the purposes of the clause and referent of ke'a.

Comment #2: Re: Expanded functionality

Curtis W Franks (Sat Jun 27 07:39:43 2015)

krtisfranks wrote: > It might be useful for this word to actually attach to exactly the last > mentioned term/word in a mekso phrase (including PA, VUhU, etc.; I guess > that there could be issues of it attaching to another poi'a, but that too > may be desired by the speaker and is not really that big of a problem (it > should not be forbidden)). I have mentioned this possibility in the notes. I personally advocate such expanded functionality at the present time, but I do not want to redefine the word in case doing so breaks it somehow. > > Another important issue is grouping. Currently, the word attaches to > exactly the most recent PA. With the aforementioned expansion, it would > attach to the most recent word in a mekso expression. But what happens if > the speaker wants to attach it to the most recent occurrence of a unit of > mekso expression that is longer than one word? For example, what if one > wants to attach it to the number twenty-three rather than simply the number > three (in decimal)? I propose another word which begins a bracket for these > units; then, poi'a would attach to everything mentioned between it and > the last unmatched occurrence of this bracket-initializing word, treating > the intervening mekso subexpression as a single whole and indivisible unit > for the purposes of the clause and referent of ke'a. See: xe'au and na'au.

if sei clauses were allowed inside numbers (which they should but they
dont) then they would parse like current UI parse:

([<pa (?re ui?) ci> BOI mi] VAU)

Also if sei is undesirable then a parallel to it but creating adverbials
might be necessary. i.e. like xoi but with the grammar of sei.

what is "wthe " in the def.?

Comment #5: Re: mistype

Curtis W Franks (Sat Mar 19 15:48:12 2016)

gleki wrote: > what is "wthe " in the def.? Indeed it was a typo; I expect that it was an artifact of using a smartphone. I fixed it to be "the". Thank you.