Message-Id: <m0rAOYU-00007EC@xiron.pc.helsinki.fi>
Date:         Wed, 23 Nov 1994 20:39:22 +0000
Reply-To:     ucleaar <ucleaar@UCL.AC.UK>
Sender:       Lojban list <LOJBAN%CUVMB.bitnet@FINHUTC.hut.fi>
From:         ucleaar <ucleaar@UCL.AC.UK>
Subject:      Re: small universe consequences
To:           Veijo Vilva <veion@XIRON.PC.HELSINKI.FI>
In-Reply-To:  (Your message of Tue, 22 Nov 94 12:11:08 EST.)
Content-Length: 2957
Lines: 67

Bob:
> ucleaar@ucl.ac.uk said
>
>    ...This is the crux. I don't think the grammar says you have to
>    consider context. General principles of communication, not language-
>    specific, say you have to consider context. ...
>
> One of the rules of Lojban is that if you have not specified a
> numbered place, you should translate the utterance as if that numbered
> place contained {zo'e}, which is an unspecified value that makes the
> bridi true.

My understanding is that "zohe" is an existentially quantified variable
which is specified only as being an instance of a maximally unrestricted
category - that is, the sentence "gerku" means "zohe gerku zohe" which
means "Ex Ey: x a dog of species y".
But "zohe" doesn't make the bridi true: what makes it true is the
existence of some x and some y such that x is a dog and y is its species.
[I mean "breed", not species.] It may not be true. Certainly the grammar
doesn't care whether it's true.

> (I am not sure whether you want to call this grammar rule
> or something else.)

I would call my version of it a grammar rule. The rule of zohe insertion,
and its semantics, are rules of grammar.

> For example,
>
>     mi klama loi zarci
>
> is:
>
>     mi klama       loi zarci   zo'e            zo'e            zo'e
> I go/went/will go  to a market from someplace, via some route, via some means
>
> For {zo'e} to work in communication between two people, the choice of
> the unspecified value must be such as to make the bridii true in the
> context.  If the value makes the bridi true in some different context,
> people will consider the utterance false in the context that counts.

A sentence can express a range of propositions: the grammar determines
the set of propositions each sentence can express. (That's (linguistic)
semantics.) Next, the hearer chooses one of these propositions (this
actually involves completing an incomplete proposition, since the
grammar delivers propositions without reference assigned). This is
pragmatics. The choice (& completion - reference assignment) of
proposition is indeed done by considering the discourse context.

The process you describe (i.e. restricting range of possible reference
of zohe) comes at the stage where the hearer chooses & constructs
a proposition on the basis of the underspecified proposition derivable,
by the rules of grammar, from the utterance. But this initial act
of interpretation is not governed by grammar rules - it is governed
by more general cognitive & communicative principles.

So yes, of course successful communication requires the hearer to
derive from the utterance a proposition that takes context into
account, but no, this will have no effect on what the rules of
grammar say.
- What is relevant to the discussion we're straying from is that
the communicational requirement that context be considered has
no influence on the fact/proposal that the grammar specifies LO
as -specific & LE as +specific.

----
And