[YG Conlang Archives] > [jboske group] > messages [Date Index] [Thread Index] >
On Wed, Apr 30, 2003 at 08:29:09PM -0700, Jorge Llambmas wrote:
>
> > > They don't have to have identical grammar to be analogous. If they did,
> > > we wouldn't be having this discussion.
> >
> > That definition is too broad, as you well know, otherwise "fa" and
> > "ma" are "analagous" simply because they're both cmavo and they
> > both end in the letter 'a'.
>
> They are analogous to that extent, yes. But being analogous in that
> they can both be a simple-tense-modal is stronger than that they are
> both cmavo. It is not unreasonable to generalize from seeing that
> you can nai a simple-tense-modal to all simple-tense-modals. It goes
> against the grammar, I know, but for people who learn by imitation
> and generalization that doesn't count. And it is not unreasonable
> to expect that restrictions in Lojban are based on something more
> than "that's the way it is".
A simple-tense-modal is not conceptually a single cmavo when it
comes to pu/faha+modal. mi pu fa'a klama contains *one*
simple-tense-modal.
It makes no sense to say "they're both simple-tense-modals".
It is true that they both alone can be a simple-tense-modal; that
isn't relevant though, since KI alone is also a simple-tense-modal,
and it certainly isn't analagous.
> > > > I think you mentioned {mi pu na'eka'e broda}. This doesn't make
> > > > sense either, since either the whole tense should be negated or
> > > > not.
> > >
> > > It makes perfect sense: impossible in the past. This is different
> > > from possible in the non-past.
> >
> > I thought "na'e pu ka'e" is other than (possible in the past).
> >
> > Doesn't the na'e scope over the whole tense unit?
>
> I have no idea. Since it makes perfect sense with short scope,
> and na'e has short scope in tanru, for example, I expect short
> scope here as well. But tenses are fairly wacky, so you may
> very well be right. In which case, how do you do short scope
> na'e on tenses?
I think you'd need to float it:
na'epuku mi ka'e broda
I don't know that the na'e is actually scoped over the whole
simple-tense-modal, though (but the look of the grammar suggests
it should have to be, since you can't put it in the middle of the
simple-tense-modal). I don't know if CLL even covers this.
John/bab?
> > > No, it is not a matter of saving syllables. It is a question of not
> > > forbidding sensible strings just for the sake of forbidding them.
> > [...]
> >
> > It's not a matter of forbidding strings.
> >
> > The grammar specifies the list of allowed strings. If it's not
> > broad enough for you I think the burden of proof is on you to show
> > that it is worthwhile for it to be broader in a particular area.
>
> Different philosophies, I guess. I don't want arbitrary restrictions,
> you don't mind them.
At some level all restrictions are arbitrary.
As long as the restrictions are defined by relatively simply rules
I'm fine.
> > One possible argument is that it would allow saying things with
> > less syllables. Another is that it would remove reduce/reduce or
> > shift/reduce conflicts and/or simplify the grammar.
>
> Another is that if it makes sense, and it causes no problems,
> there is no reason not to allow it.
A lot of things make sense that aren't worth allowing.
Everyone can tell you what
le po'u la fidos. gerku cu citka
would mean. But why allow it? It saves nothing.
Like I was saying though; I think the burden is on the positive
claim that it should be allowed. "There's not reason not to" doesn't
count as a reason for it.
> > But increasing the number of allowed strings simply for that purpose
> > alone is not a valid argument. (At least not to me).
>
> To me, forbidding strings just for the heck of it is not acceptable,
> even if there are other ways to get to the meaning that the forbidden
> string would naturally have.
--
Jordan DeLong - fracture@hidden.email
lu zo'o loi censa bakni cu terzba le zaltapla poi xagrai li'u
sei la mark. tuen. cusku
Attachment:
biniflG5yrj2Y.bin
Description: application/ygp-stripped