[
Lists Home |
Date Index |
Thread Index
]
You better duck. ;-)
A critique of Grice's Cooperative Principle says:
"For a taxonomy to be useful for our computational purposes, it should be
based on functional classes that correspond to explicitly used
information processing constraints and mechanisms. These constraints
must have operational definitions: definitions based on simple
primitives that can be implemented in hardware.... divide pragmatic
implicatures into two subclasses, direct and indirect, based on whether
new assumptions are needed to interpret an utterance as relating to the
previous conversation... the Cooperative Principle is real in some
way beyond a possible role in learning, and inferences derive from
it do occur in language... Relavance arises from the nature of
communication: a speaker demands resources from a hearer, creating
an implication that what the speaker is saying is worthwhile for the
hearer to attend to. Relevance results from having a large enough
effect on the hearer's cognitive environment with a small enough
processing effort." "Grice's Maxims: Do the Right Thing", Robert
E. Frederking
which is suspiciously one of the earliest web maxims. The use of the
http verbs comes down to implications: making the act of using
the network protocol imply the fewest facts possible, and that
accords well with TAG decisions about resource retrieval.
So is it the case that the SOAP approach increases the implicatures
of retrieving a resource? Is it the case that good web service
design whether REST or SOAP relies on making each implication
assumed discoverable prior to the commitment to assuming
responsibility for acquiring/learning the content?
len
From: Peter Hunsberger [mailto:peter.hunsberger@gmail.com]
On 2/24/06, Bullard, Claude L (Len) <len.bullard@intergraph.com> wrote:
<snip/>
> What I am still sorting out is if there
> is a clean separation of the pragmatic and the semantic layer.
> I sort of doubt it but I'm still learning.
Sure there is: the pragmatics layer is the set of symbols for which
*everyone* agrees on a single precise meaning. The semantics layer is
everything else....
(Running and hiding...)
<snip/>
> The point
> of pragmatics, I guess, is a formal means to establish a protocol
> of measures, aka, "right rock; wrong rock".
I sort of conceptualize it as the rules for how discovery is done on
semantics. Sounds sort of similar?
|
