[-empyre-] FWDing Peter Morse

[marcus bastos <bastos.marcus@gmail.com>]

---------- Forwarded message ----------
From: Peter Morse <pmorse@unimelb.edu.au>
To: soft_skinned_space <empyre@gamera.cofa.unsw.edu.au>
Date: Mon, 24 Oct 2005 02:04:15 +1000
Subject: Re: [-empyre-] C. S. Peirce and Code
As a bit of Peirce fan, and having followed this thread, I thought I
might add my five-cents-worth.

First, a cavil, Peirce is not Pierce and is prounced like "purse" (cf:
http://en.wikipedia.org/wiki/C._S._Peirce) - just thought the record
should be straight (though I am SURE I'll make some typos too!)

The discussion seems to have revolved around a few Peircean notions:
icon, index and symbol; firstness, secondness and thirdness. These
seem to me (and I should say I am not a specialist in this) somewhat
cognate, in that the icon "stands for" (cf. Wittgenstein's discussion
of "ostensivity") something by "resemblance," the index by "causality"
 or "deixis" (pointing-to) (but aren't these rather divergent?) and
the symbol by "convention." The sense of 1st, 2nd and 3rdness, seems
to be a phenomenological register of these sets of relationships, and
is intertwined with their epistemological purport (for a good
discussion of this see James Elkins:
http://jameselkins.com/Texts/Peirce.pdf and David McNeil's "Pictures
and Parables" (McNeil, David. (1985), `Pictures and Parables' in Bloc
(No.10, 1985, pp.10-18)  )). This needs a bit of drawing out,
especially as most people who come across semiotics are exposed mainly
to the Saussurian model and secondarily to the Peircean one. It
shouldn't be forgotten that there are many others - eg. Hjelmslev,
Morris, Bryson (in his way) and M.A.K. Halliday - in
systemic-functional semiotics (disclaimer - that's my Ph.D.
background, so I might have a particular bias here.)

Peirce developed an incredibly elaborate semiotics that eventually
extended to about (as I recall) 64,000 sign-types - variations of
qualisigns, sinsigns, legisigns and Rheme, Dicent sign and Argument 
(for a detailed discussion direct from the source see:
http://www.helsinki.fi/science/commens/dictionary.html and
http://www.chass.utoronto.ca/french/as-sa/ASSA-No10/No10-A2.html)
Obviously it's going to be impossible and irrelevant to discuss these
here, so the point I would make is that between Saussure and Peirce
one can find a very sharp distinction between what I would call a
"linguistic" semiotics (after all, Saussure was a philologist, first
and foremost) and a logical semiotics (Peirce, Hjelmslev et al.) - and
this in turn originates or delineates the fundamental differences of
thought that have proliferated throughout "continental" and
'empirical" philosophy (loosely the difference between the French
schools and British and American empiricism.) This opens a whole can
of worms, because none of these things are monolithic or homogenous
(especially with regard to German reception theory - Habermas, Iser,
Jauss etc.), so I will nervously evade the subject any further.

So: to digital writing. I thought the invocation of Peirce here is a
very important and interesting one, and has bearing upon all sorts of
discussions that have occurred about what a "hypertext" is, where
blogs "begin and end", what the significance of "code" is and so
forth. And I thought that the Peirce/Saussaure juncture might be a
good point by which one might start unpacking these ideas.

Having given this some thought I am not persuaded of the idea that
hypertext is revolutionary or a genuinely new mode of language. It
seems a bit old-fashioned to invoke the idea of a "new" language (a
utopian dream of expressing the previously inexpressible; a
narcissist's fantasy of public incomprehension; Orwell's "newspeak"
and "doublethink"), as this is something that permeates avowedly
Modernist tracts and practices, and yet the fact remains that these
things were written in language. What is new here? What is actually
meant when you say a "new language"? The language itself or the
conformation of the language? All language is necessarily evolving,
changing and diversifying  - it is a precondition of language that it
does so (otherwise why would there be so many different languages -
English monolinguals tend to be very limited in their ideas about
this), and clearly it has done so for millenia: it is circumscribed by
historical, socioeconomic and ideological circumstances, and
conversely gives rise to them via semiosis. "Modernity" is something I
think of as a late C19th confabulation (though arguably it extends
back through the enlightenment) -  Sterne's "Tristram Shandy" (1766)
(and indeed, the 'novel" itself) is a great exemplar of this
problematical idea - something that further problematises the dogma of
postmodernity or whatever comes "after" that. These are not
particularly good names for complex phenomena. I really fail to see
much distinction between a text written in a book and one published on
the web - the point being that when a text is read we have to ask
where the text resides - is it in the material medium or is it in the
mind? If it is in the medium or mind or in both (as I would argue it
is) then it is already necessarily hypertextual (or hypotextual), as
it has become incorporated into the the mental performance and
associationism of the individual and its context of reception. A good
example of this can be found in Greimas' notion of the actant (and
this is a straight-forward structuralist approach) - the actant can
be, for instance, the hyperfiction of the "individual" in a story -
all the words describing that "person" that are correlated by the
reader into a functioning characterisation. The cognitive processes in
this are incredibly complex, and they are certainly not more complex
in computer-based texts that exhibit hyperlinking. It seems to me that
the principal distinction here is one of speed and directedness -
firstly that it is far, far quicker to clink from link to link, but
that the association is more directed (at first) by authorial linking,
and then by various scales of metadata and search-engine linking that
become more abstruse or aleatoric the further you go. Of course this
doesn't apply (necessarily) to generative texts, so a distinction has
to be drawn.

At this point - the distinction between an "authored" text and a
generative text - we reach some idea of "code." This is a very muddy
term, replete with ambiguity, because what sort of code are we talking
about here? I interpret the way this term "code" has been used in this
thread so far in two ways: code is a language (and this may be
"natural language" - the use of the term "code" was a frequent feature
of semiotic texts since the 1970's), yet it is also the underpinning
computer-code that in and of itself is a type of artificial language
that enables computable manifestations of logical concepts (or icons
of them?). But where does this second case begin and end? That's a
very interesting question from a Peircean semiotic point of view. This
brings in not only Peirce but also Turing, as we have to ask what is
computable? Clearly artificial computer languages are metastructural
approaches to making the natural language possible or apparent. We
can't impute conscious meaning to generative texts, other than through
an assumed intentionality of a programmer/author to make certain
possibilities available for interpretation. Machines don't think, and
I don't want to get bogged down in any intentionalist fallacy.
Naively, I believe that the author is well and truly alive. - but s/he
does have limitations!

So back to Peirce. It seems to me that Peirce's ideas are useful here:
sure, we can accept his notion of the symbol as being entirely cognate
with the Saussaurian idea of the sign. It gets interesting when you
look at the index and icon. Peirce discusses the icon as pertaining to
its referent by a relation of similarity - and thus people generally
think of a picture, though this is extremely arguable. A better
example might be onomatopeia, though, again, we have a symbol that
triggers the phoneme that might be "similar" - Peirce's clearest
example of an "icon" is a mathematical expression, eg. a=b+n. This is
because it "expresses"/"pictures" a set of abstract relationships that
are only/entirely cognate with the equation - even though they are
symbolic - because they could not be "pictured" any other way. A very
sharp point, and one that clearly has implications for any discussion
of iconicity in relationship to algorithmic processes that underpin
generative "writing." This seems to me to picture a different semiotic
dimension to the text - not across the screen (left, right, or up,
down) but from the screen, through the software, the machine language
and into the processor. That's a genuinely material dimension that
maps digital writing to computer science and engineering and,
ultimately, to quantum physics. It's where writing begins, but
ultimately falls apart. On a simpler note a hyperlink in a hypertext
has a colour - that's the link that you clink, and it's well and truly
symbolic - though of course, has a pictorial, iconic, element. But
what does it re-semble? There's nothing like it in the "natural"
world.

Perhaps this has ended up being more a response about semiotics and
ontology than digital writing, but I am very much of the opinion that
it is the emergent practices that test the theories, rather than the
other way round. Nevertheless, I think Peirce has something sharp to
say today for a world he could not have foreseen.




Best,




Peter

____________________________________

Dr Peter Morse

Digital Media

Visual Media

School of Creative Arts

Faculty of Arts

University of Melbourne

Tel: +61 (0)3 8344 8358

email: pmorse@unimelb.edu.au

internet:

http://www.sca.unimelb.edu.au

http://www.petermorse.com.au

http://www.antarcticavirtua.net




Melbourne Chapter,

ACM SIGGRAPH

http://melbourne.siggraph.org