From: Damien Broderick (thespike@earthlink.net)
Date: Fri Dec 27 2002 - 21:20:00 MST
Ben Goertzel suggests:
> 1) Mikhail Bakhtin's theory of "polyphonic reality" (introduced in his
> analysis of Dostoevskian fiction).
A fundamental component of any sophisticated semiotics:
< Bakhtin's dominating theme, dialogism, insists from the outset that
communication acts are reciprocal, mutually modifying, and occur between
speaker/listeners who are themselves riven by, and constituted in, a kind of
endless internalized discussion. This corrective to any positivist
engineering analogy seems to me quite unanswerable, and I shall take its
strictures into account in my own modifications of the Jakobson schema. A
vivid if somewhat simplified comparison between the two models has been
advanced by Tzvetan Todorov, who matches each node in the Jakobson flow
chart with one derived from Bakhtin's `translinguistic' account. I give it
here in superposed form, where the italicized [here, lower] foci are
Bakhtin-via-Todorov (Todorov, 1984, p. 54):
CONTEXT
OBJECT
MESSAGE
UTTERANCE
ADDRESSER<-------------> CONTACT <-------------> ADDRESSEE
SPEAKER INTERTEXT LISTENER
CODE
LANGUAGE
That this comparison is one of analogy rather than strict homology is
stressed by Todorov. The term he labels `intertext' functions at a different
level of abstraction from Jakobson's `contact' or `channel', which
originally denoted activities designed to test the fidelity or
signal-to-noise ratio of transmission. (In my own borrowing it expands to
incorporate genres and tropes, systems of convention that tend to hold
utterances and their interpretations within predictable bounds). What's
more, the `language' node is meant to exceed any information-theoretic
formalism encapsulated in the term `code'. Generative human utterances are
able to create an unpredictable, endless supply of sentences (and
higher-order categories of discourse) that implicitly exceed the constraints
of a code, while depending to a very large degree on the stability of the
underlying codes in which they are notated. Here, too, I shall adopt the
Bakhtinian perspective in my use of the Jakobson model, identifying Code
with Language as a Semiotic System - which is to say, always a
socially-embedded system that is able, paradoxically, to produce fresh
utterances that outrun the already-said. >
(THEORY AND ITS DISCONTENTS, 1997, pp. 138-9)
This might have some bearing on any theory of AI communications.
> 2) Some theories of quantum-electrodynamics in curved spacetime
> different observers in different reference frames may literally
> perceive different realities: a particle may be existent in one observer's
> reality, but nonexistent in another's.
True, certainly, of any theory with tachyons, where observer1 in frame1 may
see a multitude of tachyons while observer2 in frame2 sees only one, or none
at all. Very odd. (And of course one reason for disbelieving in tachyons.)
> I tried once to create a unified physics theory in which each particle is
> viewed as the Feynman sum of "every possible subjective reality's view of
> that particle." I never finished formulating the theory
Drat. I tried once to create a unified physics theory in which each photon
was matched by every possible superposed fauxton. I probably finished
formulating the theory, but that was in another and now decoherent history.
:)
Damien Broderick
This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:41 MDT