> [mailto:CONLANG@listserv.brown.edu] On Behalf Of Jim Henry
> > Perhaps it would be useful to make a distinction between
adverbial
> > prepositional phrases and adjectival prepositional phrases
(I'm not
> > doing that here, because Gogido has no distinction between
adverbs and
> > adjectives, but it could be a neat feature for some other
language).
>
> I think Larry Sulky made that distinction in his Konya and/or
Ilomi, with
> an inflection or derivation of the prepositions. He thought
it would help
> resolve ambiguity to always mark whether a prepositional
phrase
> applies to its immediate preceding noun, or to the verb
wherever
> it might be.
Sasxsek too doesn't distinguish adverbs from adjectives because
they are considered the same class. To distinguish whether a
qualifier applies to another qualifier rather than the head
noun/verb, I simply double the "-i" suffix (which also means
inserting an epenthetic <r>) so you may have something like
this.
timiri bruni haus = dark brown house ("dark"
applies to "brown")
timi bruni haus = dark brown house (the house is dark
AND brown)
> Similarly, it might make sense to have a way to
> mark whether a given prepositional phrase applies to the
immediately
> preceding noun or to some other earlier noun...? I'm not sure
how
> often real ambiguity as opposed to theoretical would result
without
> such marking; e.g. in "she killed him with the gun in the
library",
> "in the library" could theoretically apply to "gun" but
> obviously, from the pragmatics of the situation, applies to
"killed".
This is a case where I think proximity plays a role, but we're
probably more likely to say "from the library" if we are
referring to the origin of the gun.
> > *On the subject of theta-role marking, I had another idea
> for sentence
> > structure which I don't think I've seen before, and I wonder
what
> > langs, if any, employ it. The idea is to have the theta-role
> > assignment order be integrated into the meaning of every
verb. Or,
>
> Sounds like Lojban, maybe Loglan as well. I'm not sure, but
I
> think it's one of the aspects of the language that make it
particularly
> hard to learn, memorizing the purely word-order based argument
> structure of each predicate word.
This is the one thing I hate about Lojan/Loglan. Learning the
vocabulary also means learning the argument list for each one.
There is a certain pattern that most fall into, but it's still
not very intutive. I have a loglang of my own in the works
where I take it down to where each lexical has only one
argument.