[YG Conlang Archives] > [jboske group] > messages [Date Index] [Thread Index] >
xorxes:
> la and cusku di'e
>
> > > (Hopefully we won't waste {lo'e} on this. It is not the
> > > kind of thing we say all the time such that it requires
> > > a special gadri.)
> >
> >It's not a very gadri-ish meaning, so for that reason isn't
> >a prime candidate for {lo'e}, but I don't see why the form
> >{lo'e} is so precious that we mustn't waste it. Monosyllabic
> >cmavo are ultraprecious (& with hindsight, many of those
> >were squandered), but oodles of disyllabics are available
>
> {lo'e} has a couple of practical advantages over other
> disyllabics: (1) it is recognized by parsers, so that you can
> check your grammar if you use it (I almost never actually
> use the parsers, but anyway)
A decent parser would allow for the possibility of experimental
cmavo & take input from a mahoste file that lists experimental
cmavo and their selmaho.
> and (2) it appears on word lists, so that people new to the
> language can at least know it is a gadri
Fair point, though proper word lists would include the more
favoured experimental cmavo. I suppose there must be a silent
constituency out there who don't want to participate in jboske
but who do want jboske to sort out the language for them.
> >I don't want to insist on differences if the differences are
> >trivial. There certainly are differences:
> >
> >* loi'e makes claims only about worlds where there is only
> >one broda
> >* lo'ei creates an underspecified proposition {co'e tu'o
> >du'u co'e}, where the content of co'e has to be glorked
> >
> >The pragmatic contexts where they would tend to be used
> >are probably different. For example, with lo'ei, although
> >the higher co'e could be construed as jetnu, so effectively
> >vacuous, gricean principles would implicate something non
> >vacuous. So, for example, I would tend to interpret
> >{mi citka lo'ei plise} as habitual -- "It is habitual
> >that mi citka lo plise" -- because otherwise one could
> >equally well say {mi citka lo plise}
>
> How would you tend to interpret {mi citka loi'e plise}?
> Doesn't it also suggest habituality?
I think the interpretation would be pretty context dependent.
I might be in a restaurant and be choosing my pudding (=dessert,
afters, sweet). In the world of the menu there is only one apple
-- the rest of the world is populated by the one orange, the
one tiramisu, the one profiterol, the one millefeuille, the one
madeleine (oops! I'm getting carried away!), etc. So in this
context I could say {mi citka loi'e plise} in a not at all
habitual sense.
> The problem is that the unmarked form should have been
> lo'ei/loi'e, not {lo}
IMO, lo'ei would not be a good unmarked form because it creates
too much underspecification about the predicates involved
(as a necessary concomitant of being a quick and dirty way
to handle 'buried quantifiers'). Maybe tu'o could have been the
unmarked form.
As for whether loi'e should have been the unmarked form, I don't
know. Certainly it seems to me to be the one that is needed the
most often.
> >Similarly, while I might say {loi'e merko -president cu
> >nanmu}, {loi'e merko -president cu -quaker} implicates that
> >generally US presidents are quakers, given the reasonable
> >premise that when you abstract away from the tokens to get
> >the type, it is the properties that are more common among
> >the tokens that survive
>
> And the same applies to {lo'ei merko -president}, doesn't it?
Only in the sense that {lo'ei merko -president cu nanmu} is
likely to be construed as "It is generally the case that
there is an American president who is a man".
> In other words, it seems to me that the abstraction to worlds
> where there is only one broda is very similar if not identical
> to the vagueness introduced by not looking at the extension
Okay, that may be so, but my understanding of {lo'ei broda} is
that it introduces vagueness in the co'e in {co'e tu'o du'u lo
broda cu co'e}. The vagueness is of course drastically constrained
by context, but there is more to {lo'ei} than not looking at the
extension.
Maybe I've misanalysed lo'ei, though. Maybe {lo'ei broda cu brode}
means something more like:
da zo'u ro mu'ei tu'o du'u da broda cu da brode
That would reduce a lot of the vagueness, but I haven't yet
checked out whether it covers all the cases it should.
--And.