Re: proposal for new member of SE

[Logical Language Group <lojbab@ACCESS.DIGEX.NET>]

la .and. cusku di'e

> I see I was ignorant of the FA-suppletion rules. I assumed that FA-marked
> sumti had no effect on the place number of the other sumti, except for
> removing the FA-sumti from the sequence. So Fi-A B C D E F would
> be 3 1 2 4 5 6. And Fi-A B C fa-D E F would (no doubt gardenpathing
> the hearer) be 3 2 4 1 5 6.

Here's an attempt, which may not be quite right but is
the best I can do at this point, to state the full set of rules.

0) BAI-tagged and tense-tagged sumti are not relevant to any of the following
rules, which act as if they were not present.

1) A FA-tagged sumti is always assigned to the place specified by the tag,
even if that place is occupied (in which case there are effectively two
or more claims: the sumti forced to occupy a single place are not identified).

2) If the first sumti in a bridi is not FA-tagged, it is x1 if it precedes
the selbri and x2 if it follows the selbri.

3) All other sumti are assigned to the smallest unfilled place whose number
is greater than that of the immediately preceding sumti.

Therefore, "fi A B C D E F" has assignments "x3 [rule 1] x4 [rule 3] x5
x6 x7 x8", and "fi A B C fa D E F" has assignments "x3 [rule 1] x4 [rule 3]
x5 x1 [rule 1] x2 x6 [rule 3; x3-x5 are filled]".  The assignment of F to x6
would no doubt be a garden path: by this point, the listener has probably
forgotten what A was!

The parenthetical clause of rule 1 explains why your "le brito be fa la .and."
doesn't work: the underlying bridi is "(t) cu brito fa la .and" which says
that "(t)" is British and so is And, but doesn't equate them.

> What is the rationale behind the FA-suppletion rules?

As the textbook says:

        It is usually a good idea for the speaker to explicitly mark
        each place when he/she scrambles them this drastically;
        we are merely presenting a convention that allows
        interpretation when the speaker is inconsiderate.

Most often, FA is used either for explicit VSO order, or to skip several
places without explicit "zo'e" operations.  It is convenient to be able to
continue counting from the last explicit marker, but placing two sumti into
the same place is oddball enough that it should not happen by default:
only explicit FA marking can force this to happen.

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