Re: The Distribution Problem: An Ambiguity?

[CJ FINE <C.J.Fine@BRADFORD.AC.UK>]

I agree with Bob and Nora.

John's argument assumes first that logical connectives in tanru are
distributable, and secondly that this is a syntactic rather than purely
semantic operation. I dispute both of these.

In the first place, there are several reasons for thinking that internal
connectives (jeks) cannot be factored out in the way that eks and giheks
can. (I remember somebody saying this explicitly on the net several months
ago - I think it was Bob quoting pc, but I'm not sure.)

Thus
        lo nanmu je ninmu cu broda

manifestly does not have the same truth condition as

        lo nanme .e le ninmu cu broda

(I got worried with this when Iain Alexander (whom I presume John means
when he says Iain Hamilton?) pointed out that

        ro lo nanmu ja ninmu cu broda

means almost exactly the same as

        ro lo nanmu .e ro lo ninmu cu broda
)

So even when we are not using true tanru (these are technically kanxe, not
tanru) the assumption that the connective can be distributed is false.


A fortiori, we should not expect that

        melbi je nixli ckule

can necessarily be expanded to any

        melbi ckule je'ipaunai nixli ckule

(Note that when the kanxe is the whole of the selbri of a bridi, expansion
does seem to be well defined; thus
        ko'a broda je brode .ijo ko'a broda gi'e brode

but as I have shown, the selbri of a selgadri is just as problematic as a
seltanru or tertanru.

Secondly, the different interpretations of such connected tanru seem to me
to be well within the range of interpretations of simple tanru.

We know that
        nixli ckule

could be a school for girls or, as Bob says, a school with girl-like
properties. Equally, then

        cmalu je nixli ckule

which unambiguously means [cmalu je nixli] ckule,
indicates a school which is (small and girl). Whether it happens to make
sense to expand this as

        cmalu ckule gi'e nixli ckule

is no more independent of the particular words used than is the question
of whether

        cmalu ckule

means
        ckule gi'e cmalu
or      ckule belo cmalu
or even ckule befilo cmalu!

I thus argue that there is an ambiguity, but it is of the form essential
to tanru. The attempt to make such expressions formally expansible is a
chimaera, at least with our present understanding of jeks.

In passing, I disagree that "labno je remna" is a good tanru for
"werewolf". I suggest that a werewolf is neither "labno" nor "remna"
but precisely "labno joi remna".

                Colin