[YG Conlang Archives] > [jboske group] > messages [Date Index] [Thread Index] >
Nick:
> I commend you guys for your premisses, and blast you for your confusion
>
> > So here is what xorxes and I agree on:
> >
> > set: lu'iro lu'ile lu'ila (=lo'i/le'i/la'i)
> > quantified: PA(ro) PAle PAla
> > substance: lu'oiro lu'oile lu'oila
> > collective: lu'oro lu'ole lu'ola (=(piro)loi/lei/lai)
> > 'unique': lu'airo lu'aile lu'aila
>
> The ontology is certainly a sound basis for anything to be said on gadri
Good.
> > Stuff still up in the air:
> >
> > * Is a gadri for Prototypical (or Representative Imaginary, or
> > whatever) logically sound?
> > If it isn't, then lo'e/le'e are short for lu'airo/lu'aile
> > If it is, then CLL compatibility might argue for lo'e/le'e as
> > the Prototypical
>
> By 'logically sound', what do you mean? Because 'prototype' seems to be
> a psychological, rather than logical, construct
I meant that, as I had explained in other messages (which quite
pardonably you may have been to busy to read, or to overloaded to
digest), it is subject to certain logical ambiguities. I have
offered a solution in a message called "***RO** lo'e cinfo cu
xabju le friko".
> > * Is there any way to make any of the above a bit shorter?
>
> Here comes trouble..
>
> > * Is there any scope to fiddle with default interpretations
> > of bare gadri?
>
> By the fundamentalist imperative, not much
I agree. But we know that CLL is partly broken, that usage is
partly broken, that the proposal is excellent, and that quite
possibly it mends more broken usage than it renders broken.
> > I propose, as the solution that would allow the greatest
> > economy in usage conditioned only by considerations of
> > meaning, that bare lo/le/la should be interpreted as if
> > preceded by lu'oiro
>
> And in a fundamentalist imperative where CLL compatibility and past
> usage outweigh perceived utility and shortness, this is of course
> unacceptable
You can't assume past usage is your side here. A lot of past usage
would be more consistent with the new proposal than with CLL.
But yes, there is a departure from CLL, in a change to the
rule for interpretation of bare lo/le/la when there is no outer
or inner PA. But fundamentalism would likewise insist on loi/lei/lai
as being fuzzy conflations of Substance and Collective; it is not
acceptable for someone to be Fundamentalist on some things and not
on others, without making their Fundamentalism suspect.
> > That would then give the following table of shortest forms:
> >
> > set: lo'i le'i la'i
> > quantified: PA PAle PAla
> > substance: lo le la
> > collective: loi lei lai
> > 'unique': lu'airo (~lo'e) lu'airole (~ le'e) lu'aila
>
> Doing the following:
>
> * Regarding 'substance' as the basic meaning of {loi}
> * Allowing that the piano carriers define 'substance' as well as
> 'collective' --- they just define 'collective' better --- and the "if
> one of us then all of us" definition (CLL p. 123) fits substance and
> not collective
> * Regarding the definition lo = su'o pa as inviolable
> * Allowing the statistical article to go back to gismu
> * Rejecting the DeLong/LeChevalier line of {lo prenu remei cu bevri},
> because if {remei} is a mass we're back to square one (it could still
> be one person doing the carrying), and if {remei} is a collective...
> well, that might work, but it needs a lot of thinking
Does Fundamentalism say that remei must be a lojbanmass rather than
a Collective? That is not such a problem, since it is nonsensical
to construe it as a Substance, so the only coherent interpretation
is as a Colective.
> I come up with this rather messier paradigm --- which does not throw
> existing Lojban out:
>
> set: lo'i le'i la'i
> quant: (PA)lo (PA)le (PA)la
> substance: loi l[e]i l[a]i
> collective: lu'oi ro lu'oi le lu'oi la
> coll., alt. lo romei be lo lo romei be le lo romei be la
> (if {mei} becomes collective not substance, which is not impossible)
> unique: lu'ai ro lo lu'ai ro le lu'ai ro la
> prototype: lo'e le'e
> mode: lo fadni belo'i lo fadni bele'i lo fadni be la'i
>
> It makes infinite sense to me that the default quantifier between LAhE
> and LE be {ro}, as is clear by inspection; but whatever
I agree that this would be a fundamentalist alternative, except
arguably for the restriction of lVi to Substance. That is,
fundamentalist wrt CLL, though not to past usage.
> > The default inner PA for bare lo/le/la would be tu'o
>
> The default inner PA for substances shall indeed be tu'o, that's the
> whole point of substances, and I reject the perpetuation of error.
> Inasmuch as CLL perpetuates this confusion, this part of it I reject
CLL exx of Prototype=lo'e are also error, IMO.
Default outer quantifiers for lVi are also error, IMO.
CLL is as compatible with lVi = Collective as with lVi = Substance.
Once you correct all the error in CLL, my proposal is about as
fundamentalist as yours. To summarize:
Your treatment of lo/le/la is more CLL-compatible, because yours
requires no change, while mine requires change to the rule that
applies when outer and inner PA are simultaneously absent.
My treatment of loi/lei/lai is more CLL-compatible, because both
proposals are equally compatible with CLL meaning, both require
change to default outer PA, mine is more compatible with canonical
exx, and mine does not require change to inner PA.
My treatment of lo'e/le'e (if they are Unique) is more consistent
with canonical exx.
The evidence is too uncertain to argue that one proposal is definitely
more consistent with prior usage than the other.
But my proposal gives a much more user-friendly outcome, in terms
of the match of brevity and frequency.
> > An overt
> > PA would change the interpretation from substance:
> >
> > quantified: loPA = su'oloPA lePA = rolePA
> > substance: lo(tu'o) le(tu'o)
> >
> > An indeterminable amount of prior usage would be invalidated,
>
> Too much: you're making all references to individuals change,
Truthconditions of most uses of lo would not be affected.
It wouldn't make any difference to la, as far as I can see.
I need to think more about le, but it wouldn't affect truthconditions.
> and I content we still do mostly want to talk about individuals
My proposal allows you to talk about individuals. You just have
to make your quantification explicit, which is not a bad, thing,
because people tend to forget about them.
> > but that is a small price to pay, because lo/le/la would at
> > last be assigned to a role where they would properly have highest
> > frequency
>
> A Lojban Mark II statement. I can't countenance so much of a change
I don't think it is more of a change than your proposal, and when
two proposals are much or a muchness in terms of CLL-conformity,
it seems reasonable to use user-friendliness as a criterion in
choosing a solution.
--And.