Re: cmavo hit-list

[Logical Language Group <lojbab>]

la xorxes. cusku di'e

> Here is a list of those cmavo that I do not intend to memorize. Some
> I won't use because I think they are too specific, others because I don't
> understand what they mean. The list is of course a draft and I could
> change my mind about any or all of them.

Note:  When I say "no argument" I don't mean I don't disagree, but simply
that I don't choose to present counterarguments.

>         ce'a lau tau zai
>         ga'e ge'o je'o jo'o lo'a na'a ru'o se'e to'a (character shifts)

No argument.  The language definition needs them for completeness, but probably
few will use them.

> These could be used in mathematics, where Greek letters are very
> much used as symbols, but I don't intend to use Lojban for mathematics.
> A formula that involves more than two or three symbols quickly becomes
> almost incomprehensible if written down in words (be it English or Lojban).
> If written down in symbols, it can't be directly read out in grammatical
> Lojban anyway, so what's the point of having such an elaborate mekso
> system?

In addition to formal mathematics, there is also "intuitive" (not intuitionist)
or "household" mathematics, which we do occasionally need: "Now, let's see:
500 ml of milk is half a liter, so I need to use all of this carton and
part of the next".  How's that in Lojban?

>         jo'i me'o ku'e ma'o mo'e na'u ni'e nu'a pe'o te'u vei ve'o
>         fu'u ge'a fe'i pi'i su'i vu'u fa'i gei ju'u pa'i te'a cu'a
>         de'o fe'a ne'o va'a pi'a re'a ri'o sa'i sa'o si'i fu'a ti'o

MEX words.  See above.

>         la'i le'i lo'i
> 
> The predicates that need sets as arguments usually work just as well
> with masses, but I admit that le'i and lo'i are too traditional to
> be ignored, so I will remove them from this list (I already know them
> by heart anyway). {la'i} I do find extremely useless, because the set
> of things that share a name usually share nothing else.

But "la'i" is +specific, so it's more useful than you think.

	le mi patfu pu traji lo ka to'ercitno fo la'i kau,n.
	My father was superlative in property old-age among the-set-of Cowans.

Here I mean the Cowans in my family, not all the things which share the
name.

>         nu'o pu'i (modals)
> 
> I don't understand them, unless they mean "ka'e jenai puca'a" and
> "ka'e je puca'a". If they do mean that, then pu'i doesn't say anything
> more than pu, and instead of nu'o one can say ka'e without much
> loss. If they mean something else, I don't know what that is.

I can't comment yet.

>         jei li'i si'o mu'e pu'u za'i zu'o (abstractions)

No argument.

>         nu'e (vocative)
> 
> Why is this a vocative? It doesn't seem to have anything to do
> with the others.

It's on the border between COI and UI.  A promise is a promise >to<
someone, though.  You can't have promising-in-the-air the way you can have
pleading-in-the-air: the distinction between "e'o" and "pe'u" is precisely
that "pe'u" has a specific referent, the person of whom something is
being asked.  The slogan might better be "pe'u do ko sarji la lojban.".

>         fi'a (question FA)
> 
> This job is already done better by cu'e, I think.

"cu'e" asks which modal place an argument belongs to, whereas "fi'a" asks
which regular place an argument belongs to.  They don't overlap; they could
be made to, but I think that loses too much meaning: "cu'e" is already
extremely vague, and specific questions are better asked with "BAI ma".

>         ze'o zo'a zo'i  (location tenses)
> 
> They seem to be very similar to to'o, te'e, fa'a. The tense paper
> says something about ones relating to the speaker and the others not,
> but why then isn't this a problem with other FAhAs, which could also
> relate to the speaker or to some other point?

They could, but the Lojban system simply doesn't have a compact tense for
"left of non-speaker point".  It could have had one.  Here I think you
have the strongest case, and I >may< think about a few cmavo "dying in the
arse" here.

>         bu'a bu'e bu'i  (logically quantified predicate variables)
> 
> I don't know how to use them.

The paper is unwritten, but we have at least one example so far:

	ro bu'a zo'u la .aniis. cu djica le nu bu'a .inaja bu'a
	For all X predicates, Anyi desires the event-of X-happens only-if X-happens.
	Anything Anyi wanted to happen, happened.

Here the "ro bu'a" in the prenex acts as a quantification of "bu'a", although
the parser thinks it is a gadri-less description.   This is an exceptional
use peculiar to the prenex.  We cannot rewrite this as "*la .aniis. cu
djica le nu ro bu'a .inaja bu'a", which is ungrammatical, because the
nu-clause has no selbri.

>         go'a go'e go'u nei (pro-bridis)
> 
> It seems an excess to have three pro-bridis for bridis before
> the last one. Only one I think would be more than enough. Also,
> I can't think of any reasonable use for nei.

No argument.

>         da'e da'u de'e de'u do'i
> 
> They also seem too many. It would be nice if they could refer to
> only part of a bridi, like the inside of an abstraction.

They do not refer to bridi at all, but to utterances.  It is important
not to compare this series to the go'i-series; the go'i-series is in
fact parallel to the ri-series.

>         si'e (portion selbri)
> 
> Seems redundant to mei. What's the difference between {pimusi'e} and
> {pimumei}? [From the definition in the cmavo list it would appear
> that {resi'e} would be a half, but the examples of the list suggest
> it is {pimusi'e}.]

"pimusi'e" is a half; the place structure relates a mass to a mass:

	lu'o la .djan. pimusei ro lei re prenu
	Massified John is-half-of the-whole-of the-in-mind-mass-of persons.

>         na'o  (typically)
> 
> I don't understand how it differs from ta'e.

The term "subjective tense/modal" and "tense/modal" seem to be used {unsinnlich}
in the cmavo list.  I believe that "ta'e" only refers to the behavior of
animates who have habits.  "na'o" is the general term.

>         dau fei gai jau rei vai (hex digits)
> 
> What a waste of top quality cmavo...

No argument.

>         vu'i sei se'o fu'e fu'o
> 
> These are on probation until I work out what they mean :)

What's so obscure?

"vu'i" is like "lu'i", but tells us the set is well-ordered.  Since there
are no gadri, fortunately, for constructing ordered sets, but some place
structures demand them, "vu'i" provides the bridge for when you want to
specify a sequence by its items.  It's redundant to "LE BRIVLA be", but
so are all LAhEs for properly chosen LE and BRIVLA.

"sei" allows us to manufacture new UIs.

"se'o" is for claims made on the basis of uninspectable internal mental
processes.

"fu'e/fu'o" allow us to manipulate attitudinal scope, as in the Lord's
Prayer, which sets certain attitudinals and leaves them on.

> (I left out UIs and BAIs, because it is hard to say what is useful
> there and what isn't. I think there are quite a few that I will never
> use, but I'm not yet sure which ones they are.)

Probably true.  Both sets are over-inclusive by design.

> Comments on all this very welcome.

No charge.

-- 
John Cowan		sharing account <lojbab@access.digex.net> for now
		e'osai ko sarji la lojban.