[Index] [Prev] [Next]

Comments to Semios-L

Steven Skaggs :
Wed, 19 Jun 1996 10:57:39 -0400

Part 2
reply to Katya Mandoki's comments

 If after all
>you've read, worked upon and discussed Pierce, you still feel not
>totally competent in peircian semiotics (I have exactly the same feeling,
>after having been working for years on this) something must be wrong in
>the way it is built. At least semantically. Peirce had great intuitions,
>was very original, honest, hard thinking, but didn't know how to write
>very well. Saussure didn't even write: he was interpreted.

Well in my case Katya, it may have a great deal to do with the limited
capacities of the reader : )

>> The lesson here is that objects are not 'real' (and certainly not
>> necessarily physical), but are position markers in the semiosic chain.
>Yes, this perspective of "position markers" in a process is right. But
>again, if objects are signifieds, they are certainly real; maybe not
>material, but real. Socially or individually real.

This question of 'reality' is sticking to our discussion like peanut butter
to the roof of one's mouth. Reality is a condition that is mediated by
signs. We do not know any way of being other than sign exchange. We might
barely imagine existence divorced from mediated relational action (Peirce's
Secondness), but haven't a clue what a reality based solely upon qualities
would mean (Peirce's Firstness). All we have in our experience is Thirdness
in which what we consider to be awareness is a constructed situation
mediated through signs.

>If you are talking about states and quantum mechanics, let's follow this
>metaphor all along and say that signifier and signified are states in
>which signs may be found (like wave or particle). The signifier, being
>more stable, can be more closely related to the particle state, the
>signified to the wave state. No. It doesn't work. Waves are not effects
>of particles as signifieds are effects of signifiers. Pity. There is a
>causal relation between Sr->Sd, whereas there is no such thing between
>waves and particles. Or am I wrong? Hope I am...

Yeah...doesn't quite work....but there is an interesting analogy to be made
between what the physicists call the 'collapse of the wave function' and
what we would call the 'getting of a signified'. There's a thought
experiment in quantum mechanics called Schrodinger's Cat. It attempts to
explain, in an almost koan-like manner, the state of being wave/particle.

Imagine you have a box and in the box is cat. There is a device hooked up
to the box the shoots a photon at a half-silvered mirror. If the photon
gets through the mirror, nothing happens. If the photon reflects from the
mirror, it triggers a photoelectric switch which injects the cat with a
lethal dose of sodium penathol. After the photon is released, you look down
at the box. What is the state of the cat: is it living or is it dead? At
that moment it seems to exist (to you) in both states. [think wave/particle
or signifier/signified]. Then you lift the lid of the box and look inside.
At that moment the wave function is said to "collapse". One condition or
the other obtains with 100% certainty. (Being a lover of animals we will
choose to say the cat purrs contentedly).

It seems to me that semiotics is much like this situation. We can envision
two or three states in the abstract but our experience is such that we
constantly collapse the wave function. This is also one reason terminology
is so difficult.

>A model should be simple and deep, rather than complex and shallow.

Yes, agreed...Ockham's Razor: discard anything that is superfluous.
But I hope that I've made the point that for certain kinds of work, the
tripartite Peircian system yields some insights that the dyadic model
doesn't. Seems to me they're both useful tools, one just needs to pick up
the right tool for the job.

>If this sounds reasonable, the peircian chain may be approached as
>sr -> sd/sr ->sd meaning that any signified, can in turn function as

Yes, you're right. Both Saussure (and Barthes for one stressed this) and
Peirce view semiosis to be an onrushing current. (Incidentally, the chain
you represent above, although it has three parts, would not be equivalent
to Peirce's triadic structure of an individual sign. he would have the
interpretant being the sign vehicle for an additional
object-sv-interpretant cluster and so on...)

And here we come to a major agreement area: semiosis is chainlike.

And here, wow, I think we may find what I lost two paragraphs
>ago in the quantum metaphor: indeed there is an interchangeable sd/sr
>state. Not in the sense that denies the sd as an effect of sr, but in the
>sense that any specific sign may take the form of either sr or sd, that a
>sd may be in turn sr in another relation.
>And this model gets the dynamic dimension so explicit in Peirce (lacking in

Yes, very nice.

One of the implications in both Saussure and Peirce is that there are these
little lumplike nodes that can be discriminated as separate parts. Whether
they be signifier, signified, sign vehicle, referent, or interpretant,
there is the implicit assumption that these things are discrete. I wonder
if that is the only way of looking at the matter.

For one thing, it seems to me that they may, if we continue the lump or
particle metaphor, be infinitely reducible. That is, a sign vehicle is
comprised of parts, those parts comprised of parts etc...(see Eco 1976
Theory of Semiotics : discussion of the 'matter of the sign vehicle'). Same
with interpretant. An interpretant is the 'coming to a provisional decision
based upon evidence put forward as the result of previous provisional
decisions'...Object? same thing... an object (or referent) holds its
position not only within a sign structure but also because of previously
constructed interpretants, etc....

If this is so, we have a couple of different ways to view signs. One is to
think of the process as a flow structure of some kind in which perhaps the
various 'parts' do considerable more blending than has been the case in
traditional semiotic analysis. Just what such a model might look like I'm
not sure. What tools of analysis it might provide are just as
questionable...yet it seems a worthy thing to ponder.

The other approach - and the one that I have played around with a bit - is
looking at the recursive nature of the parts (while allowing the
particulate nature to remain). This approach seems promising to me because
recursive yet discrete particulate systems can describe flow-type dynamics
while still remaining accessible to language. Think of pouring sand down a
hill: we need a way of touching upon the flowing nature of the sand as well
as the individual grains.

When we discuss the parts, we will be discussing loer-level wholes. The
whole thing begins to look like a nest or Russian toy dolls. We must frame
our analyses precisely: specifying what level we're regarding at every

That, I suspect, will be as much a trick as taming the current
terminological brambles.


Steven Skaggs

>> McCarthy: "Even machines as simple as thermostats can be said to have

<< Searle: "What beliefs does your thermostat have?"

>> McCarthy: "My thermostat has three beliefs - it is too hot in here, it
>>is too cold in here, and it is just right in here."

[Index] [Prev] [Next]

Comments to Semios-L