Rob wrote:
> Metaphor is too complex to deal with, IMHO, prior
to the controlled language
> JS describes. The problems of metaphor, at
least as I see it, are distinct
> from the problems of linguistic mappings from
reality to FOL. Metaphor
> happens after that step, when you MUST use
inference to reach assertions
> that cannot be directly read from the axioms.
> Rich
I agree pretty much. But I don't see the harm in
chipping away metaphor-by-metaphor (or analogy/simile/xyz) as a way of
enhancing understanding/translation. Human discourse is heavily saturated with
figurative references ("a stitch in time, saves nine" is not about
sewing). Cliches ("No pain, no gain", "You have to break some
eggs to make an omelet") so favored by managers who are
analysis-challenged are similarly figurative (but I would believe,
'translatable' expressions). My bottom-line is there's a danger of getting too
literal-minded about meaning that we lose the intended meaning. So why not incrementally
macro our way out of many of these expressions (as well as the host of others)
when we try to use machines to parse-and-compile human expressions? [the many faces
of "the" being an example starting point]
Rob
*********
Someone (John Bottoms?) earlier brought up
preexisting knowledge, i.e., the S[0]=t0 state of a trajectory of
inferences.
-Rich
Sincerely,
Rich Cooper
EnglishLogicKernel.com
Rich AT EnglishLogicKernel DOT com