WikiDiscuss

---

> On Tue, 29 Mar 2005 09:12:22 -0800 (PST), John
> E Clifford wrote:
> > Consider
> > even {du'o} which is the most generous case I
> can
> > think of outside the causals.
>
> There are plenty of generous cases: {bai},
> {cau}, {de'i},
> {di'o}, {du'i}, {fa'e}, {ga'a}, {gau}, {ji'e},
> {ji'o}, {kai}, {koi},
> {ma'i}, {mau}, {me'a}, {pa'a}, {se pi'o}, {se
> ra'a}, {ra'i},
> {ri'i}, {si'u}, {ta'i}, {ti'i}, {ti'u}, {tu'i},
> {va'o}, {se va'u}, {zau},
> {zu'e}, to list some.
>
> > It does appear
> > that your pattern works here, sort of: the
> sumti
> > is the person who know and what he knows is
> the
> > main predication.
>
> Right.
>
> > But that is not quite what the
> > original says; the original is using {djuno}
> as a
> > stand in for {certu}, though with a
> proposition
> > (the main bridi) standing in for a
> generalized
> > event description, combined with {xusra} to
> give
> > the more specific proposition. Notice, for
> > example, that we do not take the {du'o x}
> away if
> > the proposition happens to be false
>
> We don't? If we find out that {la tom klama le
> zarci} is false,
> we nevertheless keep insisting that {la tom
> klama le zarci
> du'o la djan} is true?

No, but there is an issue of what to make of the
sentence. Is the point that John claims the base
or that it is true. If the former, then we will
still insist {la tam klama le zarci du'o la djan}
even if {la tam klama le zarci} is false. If it
is the latter, then, of course, the whole is
false and John is not reliable. Transformations
tend to make things look like the first case,
whereas what is meant is very often the second.
(This is irrelevant if you are not actually using
transformations to give meanings. In that case,
I think it is at least as informative to see how
the transformation fails as how it succeeds.)

> >(indeed, one
> > major use for this sort of thing is to put
> forth
> > somewhat suspect stuff) quite differently
> from
> > {djuno}. As far as I can see {x djuno lo du'u
> y
> > broda} is not even implied by {y broda du'o
> x}
> > and the same seems to be the case with many
> such
> > transformations.
>
> Be that as it may, I'm not really looking for
> exact
> transformations but for a guide to the meaning.
> In {la tom klama le zarci du'o la djan}, what
> role does
> John play in the going? Is he a knower of
> anything at all,
> a knower of anything somehow related to the
> going, or a
> knower that the going takes place? I'm arguing
> that
> it can't mean that we go ta a market that John
> knows
> or that we are known to John, but John doesn't
> know
> that we are going to the market. It puts John
> as a
> knower that the going takes place. If the
> knowing is
> restricted to one of the other sumti, then
> {du'o la djan}
> needs to be attached to it.

There is another possible example (or a prettying
up of this) that makes a different case. Suppose
there is a job that Tom can do if he gets some
expert advice which John can give him, so he
takes the job du'o la djan, with John as the
knower (how to do the work), not with John
knowing that he has taken the job: Tom might say
{mi cpatu'i du'o la djan} in accepting the
contract. This actually seems a more systematic
use of the form {du'o} than the current one,
which looks to be related to {xusra} (from which
we do not have a sumtcita, oddly and who relation
to the evidential {ju'a} is tenuous in both
ways). In general, I have found that the fast
way to figure out what a BAI means on the basis
of the associated brivla is to give the place a
name ("knower" in the case of {du'o}) and then
take the sumti in apposition after "with name."
When this results in gobbledygook, the BAI is
almost certainly either never used or is used in
a sense pretty remote from the brivla's. Of
course this latter also happens with fairly
intelligible cases, as the {du'o} example shows.

>
> > Oh, I have been reading you as saying that
> the
> > transformation — if once we could figure out
> > what it is — *is* the meaning of the
> sentence.
>
> No, I have only used the transfromations to
> show why
> some of the examples provided seemed odd to me.
> The meaning of the sentence is often different
> with
> regards to the things actually claimed. The
> transformations
> help to understand what is related to what and
> how.

Maybe the relations become clearer, but the
transformations do not usually actually show the
relations involved. It is that latter step that
is dangerous (although the former presents some
problems as well — see the {du'o} example
again).

> > > All variations will appear on the
> dictionary,
> > > so we want
> > > examples for all, even if they are unlikely
> to
> > > end up being
> > > used.
> >
> > And what I am saying is that we should not
> put in
> > these oblique forms unless they have some
> usage.
>
> Actually, I sort of agree with that. In some
> cases, only the
> oblique form has usage ({se pi'o} for example),
> so the example
> for the {pi'o} entry would use {se pi'o}. There
> shouldn't be
> separate entries for the SE-converted BAIs. All
> the useful
> transformations should be under the same head
> word.
>
> > And in particular, we should not prejudge
> what
> > their uses might be on the basis of some
> abstract
> > rule which ignores the rather considerable
> role
> > of human ingenuity that still creeps into
> Lojban.
>
> We are defining the language, we should be as
> precise
> in our definitions as we can. That is not to
> say that we
> can't define some words as having ample
> meanings if
> that's what we want.
>
But the made up forms are not in the language yet
(even in posse, since the connection with brivla
is not so close as to automatically offer up the
oblique forms), so they are no part of the
desciption. When they get used, then they need
as good a definition as we can give, by whatever
means works.