Re: [lojban] le medomoi e le memimoi e le memi'omoi

[And Rosta <arosta@uclan.ac.uk>]

Jorge:
#>The lei/loi versus le/lo distinction doesn't matter if there's only
#>one of a given thing. But if you're not certain of this, but wish
#>to refer to a single thing, then lei/loi are the appropriate
#>choice.
#
#If only lei/loi were less marked than le/lo, I would agree. But
#I feel them as more marked, and I just don't think that {le du'u}
#is wrong enough to justify the effort of unlearning it.

I realize that this is for many people a pressing consideration in this 
and other cases.

But I think it goes against the spirit of Lojban to some extent, in that
the only practicable way for the designers of Lojban to ensure that
the encoded-markedness genuinely reflect conceptual markedness
would be to have 'draft stages' of the language revised repeatedly 
to on the basis of usage adjust encoded markedness to harmonize with 
conceptual markedness. There just wouldn't have been enough
people willing to engage in use of constantly changing language.

The upshot is that either your usage is guided by the encoded
markedness, which will result in more stylistic elegance and
harmony, or your usage ignores the encoded markedness and
says what you want to say and says it in the way you would
want to say it, which would result in us actually learning something
interesting.

#>Loi is more appropriate than lei for at least two reasons. Firstly,
#>there's no getting away from the fact that lei -- ko'a voi -- is 
#>nonveridical
#>and so makes truth-conditionally different claims from veridicals.
#
#If anything that would be an argument in favour of le/lei. I'm
#not even sure that I know what a real du'u is.

Okay. But this sort of issue would matter in contexts such as legislation
(or patents?), where the meaning encoded by the language is
binding.

#>Secondly, the logic of ko'a voi/poi in contrast to da poi, is such that
#>there is a  referent (which must be established contextually, irrespective
#>of whether it turns out that there is only one candidate referent).
#
#Both {ko'a voi} and {da poi} require that there be a referent, if
#that's what you mean. 

No. {da poi} requires a referent only for the sentence to be true.
{ko'a} requires a referent for the sentence to be interpreted.

Some linguisticians/logicians reserve the term 'referential' to 
the ko'a type.

#{ro da poi} is the only one that doesn't
#have existential import, but {[ro] le [su'o]} and {[piro] lei [su'o]}
#both do.
#
#>Hence
#>use of le ~ lei ~ ko'a voi/poi is always more context-dependent than
#>lo ~ loi ~ da poi.
#
#Maybe, but I still can't see that it makes any significant difference.
#Do you have any examples where this could cause a problem?

No, tho I haven't really tried to think of any. But I think of it more
as an ideological/philosophical issue about how, in communicating 
speaker meaning, we balance the burden of labour between
logical sentence-meaning and the hearer's cooperative Gricean
inferencing. Since all languages are equivalent with respect
to the hearer's cooperative Gricean inferencing, it seems more
interesting to me to concentrate on sentence-meanings.

#>On top of this, some uses of "le nu" are plain wrong.
#
#In the case of {nu}, I have to admit that sometimes I use {le nu}
#even though I suspect it is wrong. Maybe I should start paying
#more attention to that.

Yes, since your usage is currently the paragon model for others 
to emulate. Or so I would advise anybody seeking a model.

--And.