grammar words

[solpahi]

This group consists of:
gadri
gismu
valsi
sumti
cmavo
lujvo
tanru
bridi
gerna
(+ others?)

xorxes recently said:

I'm not advocating these by any means, I think grammar words are particularly messy and could do with some unbloating, but this is how I read the current definitions:
zo nelci cu gismu lo ka ce'u ce'u nelci kei lo nelci ce'o lo se nelci zoi ke nel ke jo'u zoi ke nei ke jo'u zoi ke nelcy ke jo'u zoi ke nelci ke

lo'u mutce nelci le'u cu tanru zo mutce zo nelci lo ka ce'u xi pa mutce lo ka ce'u xi pa nelci ce'u xi re kei kei lu mi na mutce nelci lo tersu'i stura be zo tanru li'u

doi la .xorxes. what are your ideas for unbloating these words? I can see gismu5 and tanru5 getting removed.

There is also the question of what "meaning" means, e.g. valsi2. It has usage as ka, du'u and as concrete. I've seen {ma valsi ti}, and also {lo du'u broda cu smuni zo broda} (and what's the exact difference between {sinxa} and {smuni}), but gismu2 is more clearly ka (a relation). Let's discuss.

[xorxes]

On Fri, Jun 13, 2014 at 10:57 AM, selpahi notifications@github.com wrote:

This group consists of:
gadri
gismu
valsi
sumti
cmavo
lujvo
tanru
bridi
gerna
(+ others?)

rafsi
slaka
jufra
lerfu?
bangu?
rimni?

doi la .xorxes. what are your ideas for unbloating these words? I can see
gismu5 and tanru5 getting removed.

I don't think gismu should strictly have any places other than x1, because
the relationship between the word, its predicate meaning and its arguments
is not particularly gismu-ish. "gismu" is a morphological class, not a
semantic class. It would make more sense to have a word such as "x1 is a
predicate expressing relation x2 among arguments x3", which is basically
what the definition of "bridi" says it means, but we don't have such a
word. Perhaps "bridi" was initially intended to be such word, but somehow
it turned into "predication", and "selbri" took the place of the predicate,
with the place for the relation expressed by that predicate disappearing,
so that the place structure of "bridi" is now another incoherent mish-mash.

As for "tanru", yes, at least tanru5 should probably be dropped.

There is also the question of what "meaning" means, e.g. valsi2. It has

usage as ka, du'u and as concrete. I've seen {ma valsi ti}, and also {lo
du'u broda cu smuni zo broda} (and what's the exact difference between
{sinxa} and {smuni}), but gismu2 is more clearly ka (a relation). Let's
discuss.

I think smuni and sinxa are just the converse of each other, like pritu and
zunle, or nelci and pluka. If there is a difference between them then it's
probably too subtle to bother with.

valsi is a very specific kind of sinxa. Something like "ko'a sinxa ko'e
gi'e selci ko'i noi bangu" ("selci" rather than "kantu" because that would
be a morpheme).

I don't have a problem with propositions being meanings, but they are not
usually the meaning of a single word, unless the word is being used as a
sentence. I also don't have problems with properties, relations or
concretes being meanings. The evil answer to "ma valsi ti" is of course "zo
ti fa'u zo ta go'i va'o lo nu do fa'u mi cusku".

mu'o mi'e xorxes