Re: TEXH: Specificity & Definiteness

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

John says
+++++++++>
 BTW, nobody so far has given an example of -specific +definite in English.
lojbab has produced one case in Lojban:

        A: mi klama le zarci
        B: le zarci be ma

        A: I'm going to the house.
        B: Whose house?

Here A's use is +specific -definite; he knows what he means, but doesn't
necessarily expect B to know, and indeed B does not.  However, when B
uses the construction, it is -specific +definite; B expects A to know the
referent, but B doesn't know the referent!  Needless to say, this case is
highly anomalous, and arises because B is parroting A's usage without real
understanding.
>++++++++

I do not think this is either -spec or anomalous.

On the contrary, since A has already introduced +spec market(s) [not houses!]
it is (they are) specific in the discourse, even though B does not know
their identity.
It is not anomalous, because I can't think of a way to ask B's question without
a formally specific expression other than 'ma se zdani', (ma ve zarci) which
is only pragmatically connected to A's statement anyway.

I suspect that in
le zarci ki'a
the +spec is illusory, because the sumti is being used
extrasemantically - almost as quoted text.

I repeat, that +/- spec is to do with whether the reference
is to specific individuals, .iju their identity is known to
anybody even in principle.
+/- def, which is not systematically expressed in lojban,
is to do with whether the speaker expects the identity to
be known to the hearer.
Therefore -spec +def would be highly anomalous, in that
the speaker is not referring to specific individuals, but
nevertheless expects the hearer to know which ones are
referred to!

Lojbab says:
++++++++>
 I think context has a lot to do with what "le" means.  If the referent has
not been mentioned before, it generally means that the speaker has a
specific one in mind, or at least could identify a specific if asked.
The listener need not know what the referent is, but if not, the listener
can ask "le ki'a" and the speaker oughta be able to answer.  Since the
essence of language in most usages is communication, the speaker better
have in mind trying to indicate the specific referent if it isn't obvious,
through restrictive relative clauses, pointing, or whatever.
>+++++
I don't really agree to the last bit

if I say
mi terve'u le karce
I certainly have in mind which car, and if you ask
le karce ki'a
I might say
le karce poi mi'o lamprudo'i viska ke'a
"the car you and I saw yesterday"

but I might also say
la vogzl pe li jy
[which of course in English means 'a J-reg Vauxhall' and is
probably completely unintelligible to an American who
doesn't know that a J registration is 1991-2]
which is not an answer about the identity of the
car, but about its characteristics. Now of course it
may be that I am deliberately being obtuse, and know
that you want information about the identity of the
car. But it may also be that this is the best I can do -
there is no useful information I can give you about
the identity except maybe
le karce poi selve'u fi mi
'the car I bought'

The point is that Bob's remark seems to imply that if
I used 'le' in the first place, I'm being naughty if I
can't usefully identify it to you. I absolutely reject this.
I use 'le' precisely to mean that it is a definite car
and I am not saying anything about whether I expect you
to know which car I mean. If I wanted to indicate that
I did expect you to know, I would either qualify it with
a poi, or use bi'unai.
Of course, in many cases you will know, and I would
expect that - but I do not need to express this, and 'le'
does not do so.

+++++++++>
 Thus "le" is first of all specific and descriptive.  It is second of all an
statement implying listener knowledge and/or the right to ask for such
specification.
>++++++++
Yes to the first. No to the second.

        Colin