Re: TECH: grammar updates

[John Cowan <cowan@SNARK.THYRSUS.COM>]

la kolin. cusku di'e

> > I'm fairly sure this is wrong, because I know that BAI and tenses are
> > not parallel, but I don't have a firm grasp on what's wrong with it, for
> > two reasons:  1) I haven't yet written the paper on place structures, 2)
> > I slept very badly last night.
>
> I know that there are differences, but I think the parallel does go quite
> a way.

Well, I know this much: excluding the question of what is or isn't asserted,

1)      le nanmu cu batci le gerku .izu'abo le verba cu cadzu le bisli
        The man bites the dog.  To the left, the child dances on the ice

means the same as

2)      le verba cu cadzu le bisli zu'a le nu le nanmu cu batci le gerku
        The child dances on-the ice to-the-left-of the event-of
                the man biting the dog.

(Imaginary Journeys examples 15.2 and 15.1), whereas

3)      le nanmu cu batci le gerku .imu'ibo le verba cu cadzu le bisli
        The man bites the dog, because the child dances on-the ice.

means the same as:

4)      le nanmu cu batci le gerku mu'i le nu le verba cu cadzu le bisli
        The man bites the dog motivated-by the event-of the child's
                dancing on-the ice.

Colin has complained about this behavior before, but the answer is
"Too late to change it now: the relearning cost would be too high."
One way to think about this inconsistency is that "mu'i" is short for
"fi'o mukti"; whereas "pu <sumti>" means "fi'o se purci <sumti>", that which
something is earlier than.

Nick formulated the rule as follows:  ".ipubo" means ".i pu la'edi'u",
whereas ".imu'ibo" means "mu'i la'edi'e .i".

> > Your account also fails to explain the "BAI gi ... gi" construction, which
> > works well for tenses (which do have an implicit 2nd place, viz. the
> > space-time origin), but may be very shaky for BAIs.
>
> I think that in the sense in which tenses do have an implicit second place,
> so do BAI, but certainly it is not nearly so well-defined.
>
> I think there is in fact a FORMAL parallel between tesnes and BAI in all their
> uses, though the semantics are consistently different.

Certainly: the grammar compels this -- both are instances of <tag> and <stag>.

> The more I think about it, the more dubious I am about
>       <stag> gi ... gi
>
> As you say, my account doesn't explain the BAI case - but I don't know
> what the 'tense' case is about either!

Imaginary Journeys is incomplete here.  The intention is that "<tense>gi...gi"
Imaginary Journeys (Section 15) is incomplete here.  The intention is that
"<tense>gi <LHS> gi <RHS>" (LHS=left-hand side; RHS=right-hand side)
express a relationship between the LHS, which is the space-time origin,
and the RHS, which is the point reached by the tense.  Sentences so joined
do not assert either LHS or RHS.

Theoretically, all such sentences can be converted to something of the
form "le nu <LHS> <selbri> le nu <RHS>", but often the value of <selbri>
is very unclear.  For "pu", it is "se purci", as explained above, but
what is it for "vite'eva" or "ca'oza'o"?  In effect, this construct allows
the (grammatical) tenses to be used directly as (semantic) predicates.

> It seems to me that
>       <tense> gi .. gi
> will always mean
>        ... .e/je/gi'e <tense> bo ...
> but is distressingly unparallel, in that the logical (or non-logical)
> connective is not expressed (indeed, cannot be expressed).
> What does
>       pugi mi gi do cu klama
> mean?
> Is it asserting that we both go?
>       mi .epubo do cu klama
> or
>       *ge mi gipubo do cu klama
>               (not I think currently grammatical)
> or is it another of these covert raisings as with *mo'u?
>       lenu mi klama cu balvi lenu do go'i

It is a covert raising, made useful because of the wide range of tenses.

--
John Cowan      cowan@snark.thyrsus.com         ...!uunet!lock60!snark!cowan
                        e'osai ko sarji la lojban.