Re: Semantic precision and (vs?) context / pragmatics / culture

And Rosta wrote:
>Sai Emrys, On 04/07/2006 00:57:
>> To those of you who have "semantic precision" or some variant as a
>> desideratum...
>>
>> How do you deal with the issue of context / pragmatics? Namely, it
>> seems to me to be inevitable that any finite piece of communication
>> must be underspecified, and rely at least partly on
>> a) linguistic context
>> b) social context (especially for social 'messages' that overlie it)
>> c) indexing rather than specifying the referent of any named object.
>>
>> ... at least. I'm sure the cognitivists in the audience can flesh this
>> out more.
>>
>> To clarify (c), what I mean is that no thing can be (AFAICT)
>> completely defined in a contextless manner. Inevitably at some point
>> it's just a reference to "that thing we both know that I'm pointing to
>> enough for you to recognize which thing I mean". (Again, philosophers
>> in the audience can chip in here about things like qualia,
>> inten(s/t)ionality, etc.)
>>
>> Do you consider this a problem? Do you even agree with what I said above?
>
>A Lojban slogan, credited to John Cowan, is "the price of infinite
precision is infinite verbosity". Which is true, for a certain sort of
precision, namely the sort that your message is asking about -- the
closeness of correspondence between sentence meaning and some state of
affairs that the sentence/utterance is intended to describe. Any
representational system, especially a 'digital' one, has a level of
granularity (e.g. number of pixels in a picture;; number of words in a
lexicon) that limits the maximum closeness between the representation and
what is represented.
>
>When I state "semantic precision" as a desideratum, I mean that a
propositional thought should be linguistically expressible without
distortion or ambiguity. That is, the limitations of the language should not
force the speaker to say something that is not quite what the speaker wishes
to say.
>
>--And.
>=========================================================================
Given that Ithkuil is a language with semantic precision as a design goal,
the way I work around the "problem" you identify, Sai, is to essentially
concur with And's statement about a granularity limit, i.e., I don't
consider Ithkuil, for example, to ultimately be semantically precise, but
simply *far more* semantically precise than natlangs.  I agree with your
mention of context as an intrinsic feature of any attempt at semantic
precision, which is why Ithkuil has several rather unique morphological
categories (one of which is named surprisingly...CONTEXT!) which attempt to
establish an ontological versus socio-epistemological framework for
understanding the usage of any given noun or verb.  It's all part of trying
to narrow the gap between the ideal and what is possible in terms of
semantic precision.

-John Q.