Meaning of grammatical gismu

[Colin Fine <C.J.Fine@bradford.ac.uk>]

I've just been thinking again about grammatical terms in Lojban.
 In trying to come up with place structures for 'terminator'
(famyma'o/fa'orma'o .e'u) and 'logical connector' (I now suggest
lojyki'ima'o rather than trying to use jorne), I have realised that
there is an inconsistency among the terms we have as gismu.

Specifically,
 cmavo  x1 is a structure word, grammar exemplified by
word x2, with meaning/function x3 in language x4

refers to a cmavo by its abstract function, whereas

gadri   x1 is an article/descriptor labelling description x2
in sentence x3, language x4, semantics x5

refers to one actually in use in a sentence.

To put it another way, I claim that

zo mi cu ni'i cmavo

because mi is indeed a structure word, with sensible values
for the other sumti (specifically

zo mi cu cmavo zo ko'a lo nu sinxa le cusku kei la lojban.)

But
zo le cu gadri

is not necessarily true - in fact is only true if in some extralinguistic
way I am indicating that the particular instance of 'le' to which
I am referring is actually being used in a sentence to introduce a
description.

In particular,

lu le zdani li'u zo'u zo le pe ri cu gadri fi no da

I think that this is a real problem, but the problem is not in the
 definitions, but in the way we are inclined to use them.

I suggest that in fact
zo le cu gadri
actually is false, unless 'zo le' is anaphoric for a particular instance of
'le' in use, and for the general case we need a tanru or lujvo:

zo le cu gadma'o

gadma'o x1 is an article/descriptor, grammar exemplified by
word x2, with meaning/function x3 in language x4

However, if this is correct, then the 'language' place of 'gadri' is
inappropriate.


Thoughts, anybody?

        Colin