Re: le/lo

[bob@MEGALITH.RATTLESNAKE.COM]

And Rosta <a.rosta@uclan.ac.uk> wrote:

Unless:

   > As an experiment, translate each of your uses of {le} with the long
   > gloss; similarly with {lo}; and pay attention to context.  The
   > problems of specificity will or will not make themselves felt, as the
   > case may be.  When they do make themselves felt, you have to use the
   > other standard Lojban methods for specification, such as expressing
   > color or number or tense.

   I suspect you misunderstand specificity. It is not a question
   of whether the addressee can identify the referent. It is a question
   of whether the speaker is predicating something of a particular
   referent at all. It's more like identifiability-*in-principle*
   than identifiability-*in-practise*.

Veridicality is an intrinsic characteristic of {lo}; it is an operator
that says `one-or-more-of-all-the-things-which-really-are'.
Specificity is not intrinsic.  And sometimes there can be more than
one.  But sometimes there *is* only one.  In this case, the general
and the specific merge.  Specificity is a sometime side-effect of
veridicality.

In other words, I have been speaking about *both*
identifiability-in-principle and identifiability-in-practice.

It is a question of whether the addressee can identify the *context*,
as well as the *referent*.

Suppose there is exactly *one* object in principle and practice.  To
my way of speaking English, it is often a bad translation to refer to
that object as `a'.

This is a matter of what you consider the best translation of a Lojban
utterance into English.

For example, there is just one original Mona Lisa painting.  To refer
to `a Mona Lisa' conveys something quite different to an English
speaker than to refer to `the Mona Lisa'.  (And, no, I am not talking
about an entity that is _named_, although that is what the English
usage suggests; I am considering the situation in which I wish to make
predications about members of a category that meet the veridicality
test, in this case, the one and only member of the category.)

What this discussion keeps coming down to, I think, is the question of
what people consider a fair statement of context and a fair
translation from that context.  I say:

    For the purposes of this discussion, there is just *one* real cat
    in the whole universe.  And I conclude, that as a side effect of
    this, you can identify the cat to which I am referring, since
    there is no other.

Others say,

    Hmmm...  in a context in which there is exactly one cat in the
    universe in both principle and practice, the best translation is
    always to refer to that cat in English as `a cat', not ever as
    `the cat'.

    In exactly the same way, we always translate so as to refer to the
    one original Mona Lisa painting in existance as `a Mona Lisa', and
    never as `the Mona Lisa'.

    We only use `the' in translation so as to refer to something that
    is not necessarily the Mona Lisa when we are designating some
    entity as the Mona Lisa for the purposes of a discussion.

It goes without saying that if your context is always that of the
whole universe, not reduced by any conversational or other context,
then a predication about `cat' is about identifiability-*in-principle*
and `a' becomes a preferred translation, since in that whole universe
there is more than one cat.  But that circumstance is a different
context than the one I am discussing.

--

    Robert J. Chassell               bob@rattlesnake.com
    25 Rattlesnake Mountain Road     bob@ai.mit.edu
    Stockbridge, MA 01262-0693 USA   (413) 298-4725