Coming from a (sort of) linguistics background, I've found the thread on
smysl and znachenie interesting. The obvious connections for me are with the
langue/parole, competence/performance distinctions of structuralist
semiotics. Sorry if I rehearse the obvious below - it's new to me!
Jay does a brilliant job of respresenting the structuralist (and post-stru)
positions on semantics: the sign as 'a meaning-potential, a distributional
set
of probabilities for winding up as some meaning, or an abstract position in
an abstract system of 'meaning-contrasts' that distinguishes systematically
the meaning-potentials of one sign vs. another in the same system of signs
(e.g. two words)'.
Both positions are predicated on the first, structuralist distinction
between the 'surface' of language - 'performance' or actual speech acts -
and the 'depth' which is usually represented as some form of grammar -
language stripped bare of the unfortunate messiness that is real human
(inter)action. In Saussure and Chomsky for example the underlying 'langue'
is structured system within whose (potentially 100% efficient) connectivity
true 'meaning' resides. All actual speech acts, and indeed all actual
languages, are potentially reduceable to this underlying structure of
logical connectivity. Indeed human communication is a matter of translating
to and from this logical system (note NOT necessarily referents) and the
messy rhetorical texts of actual speech. Hence the decades of writing
'scripts', 'scenarios' and logical grammars which will enable machines to
understand human speech. The problem, as natural language programmes found
in the 70s and 80s, is that there is no language-free language, no
non-representational representation: the problem of meaning simply recedes
further the more you try to strip it away.
If I understand it correctly, the AT approach refuses to privilege grammar
over rhetoric, langue over parole, competence over performance, znachenie
over smysl etc. This seems essential if language is to be re-cognised as a
form of human action, and a specific tool for mediating human action. But
its specificity is surely worth remarking on: human sign systems have rules
which can be represented in a formal (grammatical) fashion, which seem to be
assimilated alongside but in quite a different fashion to vocabulary and
instances of everyday speech. The oft-cited English children who learn the
rule '-ed for the past tense' and apply it to all verbs, including irregular
verbs, are exploiting this aspect of language - it would be hard to imagine
learning a language without it. The systemic nature of language seems as
inescapable as its mediational nature: also if we forget it we forget how
power operates in language which is through these accumulated utterly
non-innocent rules.
I'm afraid I think Jay's slot-filling account falls back into the regression
problem - where is the meaning-making activity which creates the system of
slots? This is clearly not just ths same activity which populates the
'slots' in the actual speech-act situation, but an organised history of
previous speech acts and their contexts. If the history of speech acts is
organised so as to enable future speech acts, how is it organised? What
means do we have of describing this organisation which is different from the
AT/speech act theory we have to describe the actual mediated activities of
human subjects? Forgive me if this is clear to others on the list.
Agreed:
> There is really no clear line between denotation and
connotation,
BUT Bakhtin teaches us, there are genres. Genres are more than just an
'accumulation' of speech acts, they have rules, limits, boundaries. What is
the AT analysis of genre?
Helen
This archive was generated by hypermail 2b29 : Tue May 23 2000 - 09:21:11 PDT